More Best Practices for
Rotating Equipment
Michael S. Forsthoffer
Butterworth-Heinemann is an imprint of Elsevier
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
Copyright © 2017 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means, electronic
or mechanical, including photocopying, recording, or any information storage and retrieval system,
without permission in writing from the publisher. Details on how to seek permission, further
information about the Publisher’s permissions policies and our arrangements with organizations
such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our
website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the
Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical
treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating
and using any information, methods, compounds, or experiments described herein. In using such
information or methods they should be mindful of their own safety and the safety of others, including
parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume
any liability for any injury and/or damage to persons or property as a matter of products liability,
negligence or otherwise, or from any use or operation of any methods, products, instructions, or
ideas contained in the material herein.
Library of Congress Cataloging-in-Publication Data
A catalog record for this book is available from the Library of Congress
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
ISBN: 978-0-12-809277-4
For information on all Butterworth-Heinemann publications
visit our website at https://www.elsevier.com/books-and-journals
Publisher: Joe Hayton
Acquisition Editor: Brian Guerin
Editorial Project Manager: Edward Payne
Production Project Manager: Susan Li
Designer: Victoria Pearson
Typeset by Thomson Digital
This work is dedicated to my father, William E. Forsthoffer,
who has been the ultimate mentor to me both
professionally and personally.
About the Author
In his 18 years of experience, Michael has had the opportunity to be involved
in design/selection of or field troubleshooting of all of the major types of rotating equipment. He spent 4 years with John Crane, 3 of which consisted of
being the on-site Seal Reliability Engineer at Hovensa Refinery, St. Croix, VI.
He worked as a Rotating Equipment Specialist for Forsthoffer Associates, Inc.
from 2009–15 and has been involved in selection of equipment for new and
revamp projects, troubleshooting, and site specific training worldwide for all
types of rotating machinery including, Pumps, Compressors, Steam Turbines,
and Gas Turbines. He is currently the President of Forsthoffer Associates, Inc.,
whose goal is to continue to bridge the gap between vendor and user by bringing
design and field knowledge to the user in a practical manner to assure maximum
rotating equipment reliability and safety.
ix
Preface
The objective of this book was to build upon Forsthoffer’s Best Practice Handbook for Rotating Machinery by adding more best practices that will optimize
plant safety, reliability, and profits. These best practices have all been demonstrated globally by the author and/or coworkers and are intended to resolve
machinery issues in a cost effective manner.
The format for each best practice, again is the same, and is as follows:
l
l
l
l
The Best Practice presented clearly
The Lesson Learned, which is the issue that was resolved by the Best
Practice
Benchmark, which describes generally where each best practice was used
and what it did to increase safety, reliability and/or profits
Supporting Material to give the information that lead to solving the problem
or lesson learned. This information is intended to aid in assuring management implementation of your recommendations.
The book is arranged in 12 chapters just like its predecessor with only the
last chapter being of a different subject matter (Reliability Optimization). Subjects covered include Projects, all the major types of rotating machinery in
individual chapters, and the major components of rotating machinery. Chapters
also included are pre-commissioning, start-up and Turnarounds, and Predictive
and Preventive Maintenance.
All of these Best Practices were not included in the previous book and if
there are any references to the previous book a brief summary of that reference
is included so the reader does not have to jump back and forth. The same idea
was in mind when including the supporting material for this book. If there was
the same or similar supporting material for a best practice from the previous
book, it was still included here so the reader can easily access the needed material.
In conclusion, the best practices presented throughout this book are intended
to optimize the reliability and safety of rotating equipment based in a practical
cost effective manner. If used properly, the hope is that you, the reader, will be
able to increase your rate of implementation for rotating machinery reliability
recommendations.
xi
Acknowledgments
The best practices contained in this book are a result of my 18 years of experience plus probably about the 1000 years of experience I have tried to soak up
from many of my mentors along the way in design and operation/maintenance
of rotating equipment.
The first and most important person to acknowledge is my father, William
E. Forsthoffer, who has over 50 years of experience and is recognized globally
as one of the best minds in the industry. If it wasn’t for all of the knowledge he
provided to me, whether on a job or at a pub jotting down equations on a napkin,
this book would not be possible. The majority of these best practices started as
a conversation between us two with him always providing a little extra insight
that I would not have thought of.
I also want to thank certain coworkers who have helped mentor me to this
point of my career (and I can always use more). Bob Linquist and Ken Laplant
of John Crane gave me what seems like decades of knowledge about mechanical seals and pumps in about a 4 year stint which aided in Chapters 2 and 8 of
this book. Dennis Campbell, globally recognized as one of the premier Auxiliary Systems designers has passed a lot of information to me over the last 5 years
that has led to a number of (as well as some supporting material) Best Practices
included in Chapters 7 and 9 of this book. Richard (Dick) Salzmann has been
a very helpful mentor in passing down centrifugal compressor application and
design knowledge to me over the years. A lot of his thoughts come out in the
material in Chapter 3 of this book. Also, Jimmy Trice has mentored me in the
way of field maintenance and operation over the past few years and provided
me with a practical way of getting to the root of field issues without beating
around the bush.
I also want to acknowledge my parents (William and Doris) and my siblings
(Brian, Eric, Dara, Jennifer, and Donna) who have always been there for me and
continue to be my best friends.
Finally, I’d like to acknowledge my nephew Ben, who always encouraged
me to get back to writing with his question, “Aren’t you done with that book
yet, Uncle Michael?” As of writing this, he is approaching his 10th Birthday and
I am sure he will be very successful at whatever endeavor he chooses.
xiii
How to Use This Book
This book is intended for all disciplines within the rotating equipment field with
a wide range of experience. For the reader who is new in their career in this
field and want to gain a basic knowledge of all types of rotating equipment and
their components, the supporting information can be used in a practical manner
to do so.
For plant personnel with more experience who want to troubleshoot and
resolve current rotating equipment issues, this book will be most effective.
To do this, the most efficient way included in this book are essentially two
table of contents. One lists the Best Practice, which is the resolution to a machinery problem, and the other lists the Lessons Learned or the actual problem.
The Lessons Learned list is a quick way to search for an existing problem you
are dealing with in the plant and see what the resolution (Best Practice) is for
the problem.
When you go to any specific Best Practice throughout this book you will see
they all begin with a page of the following format:
l
l
l
l
Best Practice—Resolution to a machinery problem in a cost effective manner to positively impact safety, reliability, and plant revenue.
Lesson Learned—The issue that lead to the Best Practice. This issue negatively impacted safety, reliability, and plant revenue.
Benchmark—Where this Best Practice has been applied and what its real
impact was on safety, reliability, and plant revenue.
Supporting Material—information to support why the best practice recommendation will positively impact safety, reliability, and plant revenue. This
is essential in making a presentation to management in order for immediate
implementation.
All Best Practice numbers will be written on the top corner of each page
(just like Encyclopedia) so you can quickly reference and turn to the proper
page of the book.
Whatever you do, don’t attempt to use this book like a Stephen King novel
and read it straight through, unless you suffer from severe insomnia!!!
xv
List of Best Practices
CHAPTER 1 PROJECT BEST PRACTICES
B.P. 1.1:
B.P. 1.2:
B.P. 1.3:
B.P. 1.4:
B.P. 1.5:
B.P. 1.6:
B.P. 1.7:
B.P. 1.8:
B.P. 1.9:
B.P. 1.10:
B.P. 1.11:
B.P. 1.12:
B.P. 1.13:
B.P. 1.14:
B.P. 1.15:
B.P. 1.16:
Establish a database of Lessons Learned with cost of unavailability in order for all
corporate plants to utilize
Organize periodic conferences with users in the same geographical area focused
on machinery continuous improvement
Never allow vendors to take exceptions to machinery specifications that result in
additional costs
Evaluate vendors on experience, scope of supply, and specifications exceptions
before quotation ($) has been submitted
Be proactive during the pre-award meeting by meeting and assessing the proposed
project engineer (get and review cv)—Be sure to state this requirement in the ITB.
Establish and enforce a drop dead date for process conditions to be finalized, as
early in the project as possible
Think “Outside the Flanges” when preparing datasheets. Include process and
system effect details into the datasheets
Assure Plant experienced personnel are involved in all phases of the project
Obtain CV’s of EP&C Machinery Engineers prior to EP&C Award
Assure construction specifications are included in ITB to Construction Contractors
Prepare FAT Scope to cost effectively duplicate field conditions as closely as
possible
Always benchmark Best Practice recommendations, showing results in: MTBF,
MTTR, revenue savings, safety, and emissions
It is essential to have experienced personnel involved in MOC and HAZOP studies
Have a cold eyes design review as early as possible (during VCM)
Assure all issues that expose the end user to safety, revenue lost and emissions
issues are documented by notes on the datasheets.
Perform auxiliary system component selection review at 40% engineering phase
CHAPTER 2 PUMPS
B.P. 2.1:
B.P. 2.2:
B.P. 2.3:
B.P. 2.4:
B.P. 2.5:
If an individual flow meter is not available, calculate flow using a process control
valve
Assure that all critical pumps are installed with an individual flow meter
Use Pipe Differential Temperature to determine whether or not a pump is operating in its “EROE”
Confirm NPSH Available in the field for bad actor pumps
How to determine EROE boundaries when unsure
xvii
xviii
List of Best Practices
B.P. 2.6:
B.P. 2.7:
Accurately define Suction Specific Speed for pumps with Double Suction Impellers and create new boundaries of EROE
When cost effective, Assure Driver and System have “End of Curve” Power and
NPSHA respectively
CHAPTER 3 COMPRESSORS
B.P. 3.1:
B.P. 3.2:
B.P. 3.3:
B.P. 3.4:
B.P. 3.5:
B.P. 3.6:
B.P. 3.7:
B.P. 3.8:
B.P. 3.9:
B.P. 3.10:
B.P. 3.11:
B.P. 3.12:
B.P. 3.13:
Favor dry (No Oil Injection) screw compressors for process applications below
5000 ACFM (8500 Actual m3/h)
When to use medium and high speed (>400 RPM) reciprocating compressors
Meet with Process Licensor and/or EP&C Process Engineers as early as possible
in the project to assure all operating and process conditions are on the data sheets
Always assure highest head required point in the lifetime of the plant is on the data
sheets for centrifugal and axial compressors
When considering a horizontally split compressor case for pressures greater than
40 barg (600 psig), assure that there is a minimum of 2-year operating experience
in a similar application
Assure centrifugal compressors are selected that have a rated operating temperature below 350 degrees Fahrenheit (approx. 180°C)
Require a pulsation audit by an experienced company immediately after installation of reciprocating compressor
Require a one-Piece Impeller for all sour gas services
Always require two pressure and temperature transmitters in the same plane at
the inlet and discharge of each compressor section (for both between bearing and
integral gear compressors)
Always check tilting pad thrust bearing clearance using a hydraulic jack and set
alarm and trip values based on this clearance
Impeller Design Pre-Bid Meeting Guidelines
Require bundle removal tooling be used during the performance/mechanical running testing period for barrel type (radially split) compressors
Size compressor driver for end of curve power at MCOS when greater flow = more
plant profit
CHAPTER 4 GEARS AND COUPLINGS
B.P. 4.1:
B.P. 4.2:
B.P. 4.3:
Confirm Gear no-load pressure exerted on bearings. If this value is less than 50
PSI, modify bearing design to assure shaft vibrations are at an acceptable value
during no load conditions
Always check the following when replacing gear couplings with dry couplings
Always match-mark hydraulic fit coupling to shaft in order to observe if the coupling slipped at all
CHAPTER 5 STEAM TURBINES
B.P. 5.1:
B.P. 5.2:
Always require single valve steam turbines to be supplied with a throttle valve
position indicator
If the driven equipment has additional flow range and there are no other plant
bottlenecks, consider additional steam turbine power
List of Best Practices
xix
B.P. 5.3:
Consider using Backpressure (or Extraction/Backpressure) turbines whenever
possible for process trains
B.P. 5.4: Operate condensing turbines at specified exhaust pressure
B.P. 5.5: Trend After 1st Stage (and after extraction stage for extraction turbines) in “Real
Time”
B.P. 5.6: Use after 1st stage (or after extraction pressure for extraction/condensing) pressure to determine blade/nozzle corrosion for condensing turbines
B.P. 5.7: Require a thrust analysis on all condensing turbine applications and install a thrust
balance device where necessary
B.P. 5.8: Always require ratchet type turning gear device on compressor drives
B.P. 5.9: Perform a rotor stability analysis per API 617 to confirm rotor stability for turbines with VHP inlet steam (above 100 bar or 1500 psi)
B.P. 5.10: Always purchase critical service steam turbines with electronic overspeed (two
out of three voting) backup system in order to avoid use of mechanical overspeed
trip systems
B.P. 5.11: Live trend steam seal gland condenser vacuum pressure in DCS
B.P. 5.12: Determine frequency of trip valve exercising based on steam system quality, increase frequency if steam quality is off-spec
CHAPTER 6 GAS TURBINES
B.P. 6.1:
B.P. 6.2:
B.P. 6.3:
B.P. 6.4:
B.P. 6.5:
Always use, if possible, two shaft gas turbines for mechanical drive applications
Conduct design audits on aero-derivative gas turbines that have zero, or limited
(less than 2 years in operation), mechanical drive experience
Require a compressor discharge temperature (CDT) transmitter for all gas turbines in order to accurately trend air compressor efficiency
Establish a washing procedure consisting of both on line and crank washing techniques
Use external (API-614) Lube Systems for critical mechanical drive applications
over 40 MW
CHAPTER 7 AUXILIARY SYSTEMS
B.P. 7.1:
B.P. 7.2:
Oil viscosity selection guidelines
Assure the vendor is provided with details of supply and drain interconnecting
piping (if they are not the supplier)
B.P. 7.3: Oil system console layout best practices
B.P. 7.4: Locate auxiliary pump auto start switch or transmitter in pump discharge header
B.P. 7.5: If an oil system sub-vendor is used A design audit shall be conducted with them
present, along with a shop audit of the sub-vendor
B.P. 7.6: Install high point vents on direct acting valves
B.P. 7.7: Do not install time delays in oil system trip circuits
B.P. 7.8: Install a Differential Pressure gauge across seal oil drainers when a balance line
DP gauge or transmitter is not installed on the compressor
B.P. 7.9: Always mark oil system control valves after a turnaround to give a baseline
condition and determine wear throughout a run (from turnaround to next scheduled shutdown)
B.P. 7.10: Install a bypass with a valve and orifice around accumulator isolation valve
xx
List of Best Practices
CHAPTER 8 PUMP MECHANICAL SEALS
B.P. 8.1:
B.P. 8.2:
B.P. 8.3:
B.P. 8.4:
B.P. 8.5:
B.P. 8.6:
B.P. 8.7:
Use plant, company, and industry lessons learned to properly select mechanical
seal and flush system and document details on data sheets in Pre-FEED stage
Do not use nitrogen bottles to pressurize plan 53A flush systems
API plan 52/53 fluid circulation guidelines
When replacing a mechanical seal, ALWAYS check throat bushing clearance and
replace if out of tolerance
If using an air type cooler, assure a fan is used to aid in cooling and promote fluid
circulation
API plan 23 configuration and operation guidelines
Utilize a constant flow control (Kates or equal) for external flush systems
(API 32 or 54)
CHAPTER 9 DRY GAS SEALS
B.P. 9.1:
B.P. 9.2:
B.P. 9.3:
B.P. 9.4:
B.P. 9.5:
Submit a seal gas system P&ID to vendor as early as possible in the project based
on plant, company and industry lessons learned
If sufficient Nitrogen pressure is not available for normal operation of a double dry
gas seal, utilize a nitrogen amplifier (booster compressor)
Use an amplifier (booster compressor) for start-up on tandem seals when the primary seal gas supply is taken from discharge of the compressor
Assure sensing lines for alarm and trip devices are as minimal as possible
Install a backpressure control valve in the primary vent with an electronic position
indicator
CHAPTER 10 CONSTRUCTION, INSTALLATION,
COMMISSIONING, AND TURNAROUNDS
B.P. 10.1: Conduct machinery pre-turnaround audits to determine scope of work during the
turnaround
B.P. 10.2: Conduct site specific training for all disciplines involved with machinery to better
understand how the major components are supposed to work and the effect that the
process and all other related systems have on the reliability of these components
B.P. 10.3: Review machinery instruction manuals prior to shipment from vendor
B.P. 10.4: Spare critical machinery rotor storage guidelines
B.P. 10.5: Assure vendor for epoxy grout is on site for initial pours and provides training
B.P. 10.6: Bring key millwrights and operators to factory acceptance test
B.P. 10.7: Have vendor service representative available at factory acceptance test
B.P. 10.8: Assure dry gas seal piping from source to the panel (and including the panel) is
stainless steel
B.P. 10.9: Perform initial functional testing on auxiliary systems prior to initial start-up of
the train
CHAPTER 11 PREDICTIVE AND PREVENTIVE MAINTENANCE
B.P. 11.1: Begin Root Cause Analysis (RCA) immediately when a change in condition of
one or more components has been observed
List of Best Practices
xxi
B.P. 11.2: Try to postpone pump maintenance until turnaround to assure that a spare pump is
always available
B.P. 11.3: Initiate site machinery instrumentation excellence program to assure all installed
instruments are calibrated and in working condition
B.P. 11.4: Utilize a company machinery database for lessons learned in order to improve
machinery reliability
B.P. 11.5: Conduct in-house training using supervisors within a unit to instruct young personnel on the importance of and how to perform key PM tasks
B.P. 11.6: Assure all oil system and seal gas control valves have a means of position indication
B.P. 11.7: Check and confirm oil system relief valve settings on the console during a turnaround
B.P. 11.8: Check the function of all main oil pump steam turbine (if you have one) components during turnaround
B.P. 11.9: Performance monitoring should be the responsibility of the machinery reliability
department
CHAPTER 12 RELIABILITY OPTIMIZATION
B.P. 12.1: Establish a methodology for identifying plant bad actors
B.P. 12.2: Establish cost of unavailability for critical equipment
B.P. 12.3: Bring component condition monitoring (CCM) philosophy into reliability centered maintenance (RCM)
B.P. 12.4: Root Cause Analysis (RCA) guidelines
B.P. 12.5: Guidelines to gathering facts when conducting a RCA
List of Lessons Learned
CHAPTER 1 PROJECT BEST PRACTICES
L.L. 1.1:
L.L. 1.2:
L.L. 1.3:
L.L. 1.4:
L.L. 1.5:
L.L. 1.6:
L.L. 1.7:
L.L. 1.8:
L.L. 1.9:
L.L. 1.10:
L.L. 1.11:
L.L. 1.12:
L.L. 1.13:
L.L. 1.14:
L.L. 1.15:
L.L. 1.16:
Failure to utilize Corporate Wide LL/BP Database has resulted in certain plants
within the company that have ongoing issues that other plants have already solved
Failure to hold regional conferences will result in continued firefighting and
lengthen the time for many plants in the region to resolve current machinery issues
Allowing exceptions to user specifications (especially ones that reduce cost) go
against company philosophy and will result in lower machinery reliability
Failure to evaluate the vendor experience, scope of supply, and specifications
exceptions prior to money being discussed will result in lower reliability
Project Engineers with limited experience has resulted in delays and reliability
issues
Process Conditions changing after the Vendors have placed their bids has resulted
in redesign of the machinery and significant cost adders/delay of schedule
It is the writer’s experience that approximately 80% of Machinery Failure Root
Causes are due to process variations not anticipated in the design phase
Failure to have the appropriate experienced personnel involved in the project result in decreased reliability and loss of revenue
Failure to vet the EP&C Machinery Engineers can result in acceptance of unreliable equipment
Failure to include construction specifications in the ITB to the Construction Contractor can result in possibly long delays in construction schedule
Improper specified FAT Scope can result in unexpected reliability issues during
initial start-up
Failure to benchmark Best Practices properly result in unresolved machinery
issues and continuation of revenue lost
Inexperience in MOC’s has resulted in Implementation of machinery Issue resolution being delayed or canceled
Failure to hold a cold eyes review at this stage has resulted in schedule delays
Not indicating to the Vendors the Plant exposure to unplanned shutdowns can
result in inadequate support
Failure to perform this review has resulted in delayed startups and unplanned shutdowns
CHAPTER 2 PUMPS
L.L. 2.1:
L.L. 2.2:
Not knowing and trending flow of critical and bad actor pumps will result in lower
pump component reliability
The Inability to accurately know the flow through each pump can result in unnecessary maintenance and risk of lost production
xxiii
xxiv
List of Lessons Learned
L.L. 2.3:
L.L. 2.4:
L.L. 2.5:
L.L. 2.6:
L.L. 2.7:
The inability to know if a pump is operating at low flow can result in wear and/or
component failure
Failure to identify the cause of fluid vaporization within the pump will most likely
result in multiple failures and increased maintenance costs or loss of production
Failure to operate within the EROE will result in component wear and failures
Failure to identify the Suction Specific Speed Value and Boundaries of Operation for
a Double Suction Impeller can result in significant damage and even shaft breakage
Inability to incorporate this Best Practice can result in pumps being bottlenecks in
allowing for more plant production
CHAPTER 3 COMPRESSORS
L.L. 3.1:
L.L. 3.2:
L.L. 3.3:
L.L. 3.4:
L.L. 3.5:
L.L. 3.6:
L.L. 3.7:
L.L. 3.8:
L.L. 3.9:
L.L. 3.10:
L.L. 3.11:
L.L. 3.12:
L.L. 3.13:
Selecting a Lubricated Screw or Reciprocating Compressor over a Dry Screw
Compressor has led to the following
Use of a medium or high speed reciprocating compressor in critical process units
have resulted in very poor reliability
Failure to assure all operating and process conditions are noted on the data sheets
will most likely result in lower reliability and possible loss of production
Not listing the highest possible head required on the datasheets has resulted in
compressors operating in the field with the Anti-Surge Valves Open for long
periods of time
Failure to check experience of horizontal split case in high pressure applications
has resulted in safety hazards due to split line leaks
Compressors selected with rated temperatures over 350°F will result in lower reliability in the form of either efficiency loss, or worst unplanned shutdowns
Failure to have a field pulsation study performed can result in years of operation
with premature component failures
Use of welded impellers in H2S service has resulted in catastrophic impeller damage and significant loss of production
Having only one pressure and temperature transmitter at each inlet and outlet may
give inaccurate
Failure to use a hydraulic jack during the setting of thrust clearance has resulted in
premature thrust alarms and trips
The Failure to review for impeller/blade experience prior to vendor acceptance
can result in extended FAT time, delayed field start-up and continuous safety and
reliability field issues
Failure to test the actual bundle removal tooling in the vendor’s shop can result in
significant delays in the field during a turnaround
Not having sufficient driver power can result in the inability to make more profit
or loss of production
CHAPTER 4 GEARS AND COUPLINGS
L.L. 4.1:
L.L. 4.2:
L.L. 4.3:
Gears operated at low load without proper bearing design can experience excessive vibration and potentially trip, causing down time and loss of production
There have been many experiences involving operating within a Natural frequency when converting to dry coupling
Failure to match mark the hydraulic fit coupling has resulted in severe failures that
could have been discovered and fixed during prior shutdowns
List of Lessons Learned
xxv
CHAPTER 5 STEAM TURBINES
L.L. 5.1:
Inability to know throttle valve position can result in reduction of speed for critical
equipment which equals reduction of rates/profit
L.L. 5.2: Lack of turbine power can be a bottleneck for plant production
L.L. 5.3: Condensing steam turbines will have moisture toward the exhaust side, which
could cause corrosion of the blades and nozzles in the last few stages
L.L. 5.4: Operation of condensing turbines at exhaust pressures lower than specified on the
data sheets will result in more moisture and higher rate of blade/nozzle corrosion
on the back end
L.L. 5.5: Failure to trend After 1st Stage Pressure can result in an abundance of fouling that
could cause an unplanned shutdown by tripping on vibration when the fouling
breaks off of the rotor
L.L. 5.6: Failure to identify and trend rate of blade/nozzle corrosion can result in unplanned
shutdowns and production lost
L.L. 5.7: Inability to request a thrust analysis for impulse type condensing turbines have
resulted in high thrust loading and excessive bearing pad temperatures/wear
L.L. 5.8: The use of continuous speed turning gear devices has resulted in premature dry
gas seal failures
L.L. 5.9: Failure to perform a rotor stability analysis for turbines in VHP steam service has
resulted in continuous vibration issues in the field and eventual turbine replacement due to the low reliability
L.L. 5.10: Machinery historical case studies are full of examples of failed turbines and personnel injury resulting from the failure of turbine overspeed trip devices during
the uncoupled overspeed trip checks
L.L. 5.11: Failure to monitor and trend gland condenser vacuum on special purpose (Unspared) steam turbines has resulted in gross contamination of the oil systems and
reduced bearing life
L.L. 5.12: Failure to exercise trip valves at the proper frequency resulted in catastrophic machinery failure, personnel lost time and loss of life
CHAPTER 6 GAS TURBINES
L.L. 6.1:
L.L. 6.2:
L.L. 6.3:
L.L. 6.4:
L.L. 6.5:
Large starting motors (sometimes over 50 MW) have inherently lower reliability
and have resulted in inability to start up on time
Failure to accurately audit the design of new mechanical drive gas turbines can
result in unexpected issues and project/startup delays
Inability to trend compressor efficiency has resulted in permanent gas turbine
power loss and reduction of rates
Improper washing procedures have resulted in ineffective washes with minimal
performance gain
Typical gas turbine lube systems (inside enclosure) do not allow for effective condition monitoring of the system and have resulted in poor system reliability
CHAPTER 7 AUXILIARY SYSTEMS
L.L. 7.1:
Inadequate oil viscosity in high temperature climates has resulted significant rotary pump wear and low system reliability
xxvi
List of Lessons Learned
L.L. 7.2:
Failure to coordinate details of interconnecting piping with the oil system vendor
has resulted in unplanned shutdowns and revenue lost
L.L. 7.3: Consoles that are crowded and do not allow easy access are often ignored by operators and not fully understood in terms of system function
L.L. 7.4: Improper location and setup of Auxiliary Oil Pump (AOP) Auto-start has resulted
in numerous unit trips and lost production
L.L. 7.5: Failure to design audit new oil system and component design from sub-vendors
has caused many start-up delays and trips of critical (un-spared) compressor
trains
L.L. 7.6: Failure to vent direct acting control valve sensing lines has resulted in delayed
response of control valves and unit trips
L.L. 7.7: Installing time delays on trip circuits do not go after the cause of failure but put a
“Band Aid” on it and will delay the time to troubleshoot the actual problem, resulting in profit loss for the facility
L.L. 7.8: Thrust bearing assemblies are frequently changed, without considering balance
system differential pressure trends only to find that balance device deterioration is
the root cause and compressor disassembly is required forcing a 5–7 day loss of
revenue
L.L. 7.9: Inability to know control valve position from beginning to end of a run (from
turnaround to next scheduled shutdown) has resulted in delays in troubleshooting
the root cause of oil system failure and loss of production
L.L. 7.10: Many unit trips have been traced back to opening the isolation valve to the
accumulator after regular maintenance to quickly
CHAPTER 8 PUMP MECHANICAL SEALS
L.L. 8.1:
L.L. 8.2:
L.L. 8.3:
L.L. 8.4:
L.L. 8.5:
L.L. 8.6:
L.L. 8.7:
Failure to utilize previous plant, company and industry lessons learned and include details on the mechanical seal data sheet will result in lower than optimum
seal MTBFs
The use of nitrogen bottles to pressurize plan 53 flush plans have resulted in unexpected seal failures when capacity runs low
Improper piping and reservoir setup for dual seals has resulted in MTBFs that are
12 months or less
Failure to check and replace throat bushing has resulted in poor seal MTBF due to
the seal chamber conditions not being ideal
The use of fin type air coolers without a fan have resulted in minimal to zero
temperature reduction and minimal to zero fluid circulation
Failure set up and operate/monitor API plan 23 seal flush systems have resulted in
seal MTBFs far below expected values (Lower than 12 months)
Manually controlling the flow or pressure of a plan 32 or utilizing just pressure
control on plan 54s have resulted in numerous seal failures due to inadequate flow
during process changes
CHAPTER 9 DRY GAS SEALS
L.L. 9.1:
Failure to provide seal gas system P&ID to vendor based on lessons learned has
resulted in unreliable systems that have caused unplanned shutdowns and revenue
loss
List of Lessons Learned
L.L. 9.2:
L.L. 9.3:
L.L. 9.4:
L.L. 9.5:
xxvii
Double dry gas seal systems as compared to tandem systems eliminate the following items to reduce complexity and optimize reliability
Failure to have a start-up gas has resulted in seal failures right after a start-up and
revenue loss
Improper sensing line setup for primary vent instrumentation has lead to inability
for the instruments to alarm and trip the machinery when operating at unsafe leakage levels
Inability to accurately monitor dry gas seals has lead to premature seal replacement that could have been saved for a planned shutdown
CHAPTER 10 CONSTRUCTION, INSTALLATION,
COMMISSIONING, AND TURNAROUNDS
L.L. 10.1: Inability to properly define turnaround work scope for critical machinery has often
resulted in overhauls that were not required
L.L. 10.2: Inability to properly train all disciplines on the importance of the process and
system’s effect on machinery and components will lead to reoccurring machinery
failures since the root cause may not be identified the first time
L.L. 10.3: Inadequate vendor instruction manuals have resulted in longer mean time to repair
equipment since communication with vendor for specific details is required
L.L. 10.4: Improper storage of spare critical machinery rotors have resulted in severe
turnaround delays when the spare rotor was completely corroded in critical areas
and unable to use
L.L. 10.5: Failure to specify the use of epoxy grout for all machinery installations, have an
approved grout procedure in place, and an experienced epoxy grout contractor
have caused significant project delays and foundations that required re-grouting
before or during the first scheduled plant turnaround
L.L. 10.6: Unfamiliarity with the equipment by plant personnel has resulted in delays during
unit turnarounds
L.L. 10.7: Inability to screen the service representative for your equipment can result in
inadequate help and delays in start-up
L.L. 10.8: Failure to assure all dry gas seal piping is stainless steel has resulted in severe
fouling of a compressor and delays in initial plant start-up
L.L. 10.9: Failure to functional test auxiliary systems has resulted in unplanned shutdowns
when the system did not recover quickly enough during a main pump driver trip
CHAPTER 11 PREDICTIVE AND PREVENTIVE MAINTENANCE
L.L. 11.1: Failure to identify component condition change early enough has resulted in
numerous unplanned shutdowns that could have been avoided
L.L. 11.2: Inability to have a spare pump available (because one pump is in the shop for
maintenance) has resulted in numerous instances of plants that had to significantly
reduce rates because of no pumps available for a particular service
L.L. 11.3: Inaccurate or non-working instruments have resulted in unplanned shutdowns
because a certain component was not accurately being monitored, causing lost
revenue
L.L. 11.4: Failure to utilize Corporate Wide LL/BP Database has resulted in certain plants
within the company that have ongoing issues that other plants have already solved
xxviii
List of Lessons Learned
L.L. 11.5: Inability to implement programs to carry out regular machinery PM tasks has
resulted into numerous unplanned shutdowns and revenue lost
L.L. 11.6: The inability to monitor control valve position in auxiliary systems has led to
many surprises and replacements soon after a turnaround. Monitoring of valve
stem position would have identified worn components and allowed replacement
during a turnaround
L.L. 11.7: Many unit trips have been traced back to improper setting of relief valves that
caused them to open at lower than set pressures, which required the auxiliary
pump to start. Starting of the auxiliary pump was either too late or caused control
valve instability resulting in a low oil pressure trip and a unit trip
L.L. 11.8: Steam Turbines used for main oil pump drivers have the lowest reliability of oil
system components and have been responsible for many oil system trips
L.L. 11.9: Reliability groups not incorporating operations and process engineering input
produce lower machinery MTBFs and less implementation of recommendations
CHAPTER 12 RELIABILITY OPTIMIZATION
L.L. 12.1: Inability to identify the plant bad actors accurately can result in recurring failures
L.L. 12.2: Failure to establish cost of unavailability has resulted in continuing failures since
resolution could not be implemented
L.L. 12.3: Failure to simplify RCM has resulted in personnel being flooded with paperwork
and machinery problems not being solved properly
L.L. 12.4: Ineffective root cause analyses have resulted in the inability to identify a root
cause of failure, which leads to repeat failures and revenue lost
L.L. 12.5: Failure look at the 5 components and 5 causes of failure when gathering facts has
resulted in long drawn out RCA’s that have not determined the root cause of the
problem
Chapter 1
Project Best Practices
B.P. 1.1: Establish a database of Lessons Learned with cost of unavailability in order for all corporate plants to utilize
Obtain daily revenue value (based on nominal market prices) and calculate
the time that specific issues (Lesson Learned) have resulted in a shutdown or
reduced rates and what the total loss of revenue was for each Lesson Learned.
This information should be tabulated in a company-wide database along
with the Best Practice that would be focused on eliminating the root cause.
Then all facilities within the corporate umbrella can utilize these Best Practices
to resolve ongoing issues, but most importantly assure that future projects incorporate the appropriate Best Practices.
Of course, this database should be used by the “Project Team” when a new
or upgrade project is in the PRE-FEED phase.
L.L. 1.1: Failure to utilize Corporate Wide LL/BP Database has resulted
in certain plants within the company that have ongoing issues that other
plants have already solved
Not having specific Lessons Learned in a corporate database segregates the
plants from each other and if one plant has solved a machinery issue, the others
may not know the solution and continue to lose daily revenue.
Not incorporating this BP will also result in new construction/upgrade projects that end up with the same exact issues that have been present since day 1 of
start up in other similar corporate wide projects.
BENCHMARKS
This Best Practice has been used by Forsthoffer Associates, Inc. since 1990 and
has been incorporated into companies with the following types of Plants with
the benefits listed earlier:
l
l
l
l
MEGA Ethylene Plants
MEGA Butyl Rubber Plants
Methanol Plants
MEGA LNG Plants
More Best Practices for Rotating Equipment. http://dx.doi.org/10.1016/B978-0-12-809277-4.00001-2
Copyright © 2017 Elsevier Inc. All rights reserved.
1
B.P. 1.1
More Best Practices for Rotating Equipment
SUPPORTING MATERIAL
As someone who has been involved with projects as a rotating equipment vendor, end user, and consultant since 2004 (and with experience of personnel
within the company since the 1970s), I have had the opportunity to see custom
designed rotating equipment projects from all industry viewpoints. Regardless
of your position, you will face the challenges of company profit optimization,
depleted workforce experience levels and time constraints.
The Vendor, EPC, and End Users all have different objectives and the more
that the End User can provide up front early in the project the better chance the
Vendor and EPC will understand their objectives. Vendor lessons learned are
detailed in Table 1.1.1.
Table 1.1.2 details the lessons learned by the End User.
Review Tables 1.1.1 and 1.1.2 and observe the similarities all imposed by
time and budget constraints. Also, observe how the involved individuals seldom
have the opportunity to observe how their client operates and what their objectives are.
TABLE 1.1.1 Vendor Lessons Learned
j Time constraints forced the acceptance of what was on the data sheet
j The tendency was to think inside the flanges of the compressor only and not consider the process
j Questions to the end user/contractor were minimal based upon competitive pressures and time constraints
j Copying from past jobs “cut and paste” was a necessity to minimize engineering
hours and Today (21st Century) is electronic cut and paste
j Contractor/end user questions diminished valuable engineering time. There was
little time or money for visits to client plants unless there were significant design
problems
TABLE 1.1.2 End User Lessons Learned
j Time constraints forced acceptance of what was on the process data sheet without
time to question the basis for the stated conditions
j The tendency initially was to think inside the machinery flanges, but eventually it
was understood that all equipment is directly influenced by the process
j Contact with the client (plant where the equipment will be installed) was minimal
based on project team pressures for schedule milestones
j Company specification contents were increasing rapidly since all company divisions
and plants were required to review specifications and therefore naturally contribute
something
j There was limited project budget for visits to client plants unless there were equipment design problems.
2
Project Best Practices
Chapter | 1
Since 1990, Forsthoffer Associates, Inc. has engaged in troubleshooting,
machinery selection and revamps, as well as site-specific operator, maintenance
and engineering training. There are other challenges but the similarities are
striking and the challenges are the same. These facts are noted in Table 1.1.3.
Based on my experience, I have learned, most of the time the hard way, that
all three of these groups (vendors, contractors, and end users) have the same
objective but different means of obtaining that objective. Table 1.1.4 presents
these facts.
It is important to remember these facts at all times during the entire project. The information contained in the following figure should be the basis for
convincing the Project Team that all decisions regarding equipment purchase
should be made on the basis of Process Unit life cycle cost and not capital cost
and/or schedule considerations. The specific objectives of the end user are presented in Table 1.1.5.
TABLE 1.1.3 Contractor/Consultant Lessons Learned
j Both vendors and clients have limited experience bases
j Decisions are made quickly, often without benefit of all the pertinent facts
j Most projects are run on the basis of minimum capital investment and not life cycle
cost
j Implementation of action plans is slow
j Vendor and end user’s interface infrequently—usually only during field failures
TABLE 1.1.4 The Objective—Maximum Profits
Everyone has the objective of maximum profits but the means to accomplish this end
is different:
j Vendor—designs for minimum cost
j Contractor—engineers and installs for minimum cost
j End user—must operate the custom designed equipment 24/7 for 30 years or more.
Therefore, the end users objectives can be directly opposed to the vendor’s and
contractors!!!
TABLE 1.1.5 End user—Specific Objectives for Maximum Profit
j Maximum machine reliability
j Minimum operating cost
j Minimum time to repair
These objectives result in.............................................. maximum up time
which will yield .............................................................. maximum revenue
and .................................................................................maximum profits
For the entire life cycle of the process unit!!
3
B.P. 1.1
More Best Practices for Rotating Equipment
TABLE 1.1.6 Daily Revenue
j Is the amount of revenue obtained in 24 h of operation
j Trip of an un-spared item = exposure to revenue loss
j Daily revenue values can range from 1MM$ to 5 MM$+
j Always justify Project Scope requirements on the basis of daily revenue loss
j Assign an Actual Daily Revenue Loss amount to each proposed Best Practice if it is
not implemented
The most important factor in life cycle cost considerations is daily revenue
and obtaining this figure should be the number one priority in the early stages
of the project. It will be a key fact in obtaining management support for your
project action plans. Table 1.1.6 presents these facts.
Therefore, the company life cycle revenue and profit, potential will be a result of incorporating all of your project best practice requirements into the project action plan at the first opportunity before the first project budget estimate is
prepared. Fig. 1.1.1 shows the advantages of incorporating this philosophy as
early as possible into the project.
This action should be taken when the project is first announced and the project team is assembled, before the Project Budget Estimate is calculated. The
approach taken during the first 3–6 months after the initial project kick off will
determine the level of reliability and life cycle cost savings for the entire life of
the process unit (over 30 years). Most important is the necessity of establishing immediate creditability with the project team so that your ideas are implemented.
Hopefully, the previously mentioned information will be of use in your project involvement in terms of lessons learned. The resulting best practices should
be developed into a project philosophy that will eliminate all the issues noted
FIGURE 1.1.1 The life span of rotating equipment.
4
Project Best Practices
Chapter | 1
earlier and will obtain and maintain your management’s support throughout the
entire project from the pre-feed phase to field operation.
Note that while this book is concerned with rotating equipment Best Practices, many of the principles in this Book are equally applicable to all assets
included in a project.
BP 1.2: Organize periodic conferences with users in the same geographical area focused on machinery continuous improvement
Many machinery problems (especially but not limited to GT’s) can be attributed to the ambient condition challenges where the equipment is installed.
For example, the hot dusty climate of the Middle East, the hot humid climate
in many Asian areas, and the very cold climate in Northern North America, just
to name a few.
A 2–3 day conference (or periodic one day events), where personnel from
each of the companies will present their issues and what they have implemented to resolve their issues will result in higher machinery reliability in
the region, since many of the other users have probably encountered the same
problems.
A report should be generated at the close of the conference and sent to
representatives of each of the companies attending the conference. The report
should detail the regional Best Practices discussed during the event and how
they have been implemented and what result they have had in increased reliability.
L.L. 1.2: Failure to hold regional conferences will result in continued firefighting and lengthen the time for many plants in the region to resolve current machinery issues
As in BP 1.1 of this book, not holding these conferences will definitely prolong the time that some plants take to resolve major machinery problems, since
another plant may have already solved the issue that you are currently dealing
with.
BENCHMARKS
This Best Practice has been used by the writer since the 1990s and has been
increasingly been implemented in the last 5 years since profit margins have
become less and less due to the global economy. These conferences have been
held by specific End Users, as well as conferences like the annual Texas A&M
Turbomachinery Symposium and have resulted in implementation of Regional
Best Practices in a much timelier manner.
SUPPORTING MATERIAL
Please refer to the Supporting Material for BP 1.1.
5
B.P. 1.3
More Best Practices for Rotating Equipment
B.P. 1.3: Never allow vendors to take exceptions to machinery specifications that result in additional costs
Specifications are detailed in a project for the point of assuring that the Corporate and Industry Best Practices will be incorporated in the project. Some
of these specifications increase reliability by either increasing or reducing the
number of components and vendors may take exceptions to this. The End User
Specifications should be the final say and the project team should not allow any
exceptions that can affect reliability, especially ones that cost more.
Items listed in the specifications are in there for a reason and should not be
allowed to be taken exception on. To name a few, following are items that could
be listed in the specifications that can affect reliability:
Machinery assets:
l
Limiting the maximum number of impellers in a compressor or pump that
will cause natural frequency/vibration issues when an additional case will
prevent these issues. Note: Lessons learned must be presented to Project
Management (see B.P. 1.1) to justify this action.
Auxiliary systems:
l
l
Using Centrifugal Pumps on Lube Oil Systems, since they eliminate the
need for Relief Valves and a Backpressure Control Valve.
Using a common DP Control Valve for the DGS Primary Seal Gas Supply.
L.L. 1.3: Allowing exceptions to user specifications (especially ones that reduce cost) go against company philosophy and will result in lower machinery reliability
Not fully utilizing the specifications developed by the End Users and/or
Global Experts/Committees will negate the reasoning behind having those
specifications to begin with.
BENCHMARKS
This Best Practice has been prominent with Machinery Assets (Number of
Compressor or Pump Casings in a Train), Oil Systems, and Gas Seal Systems
and has been utilized most recently in the last 5 years since the influx of instrumentation/additional components has been introduced to the industry. It is a fact
that if you can meet your reliability goals with minimal components your risk
of failures will go down.
SUPPORTING MATERIAL
Please see Supporting Material for BP 1.1
B.P. 1.4: Evaluate vendors on experience, scope of supply, and specifications exceptions before quotation ($) has been submitted
6
Project Best Practices
Chapter | 1
If all of the items listed earlier have been discussed and evaluated prior to the
Vendor providing the costs for the equipment the following will be achieved:
l
l
l
l
l
You will not receive a prototype(s) since experience of all components have
been fully vetted.
All of the specific Lessons Learned from the company will be incorporated
at minimal cost.
Vendors will be required to take minimum exceptions to your specifications
(see B.P. 1.3).
You will assure to get all testing as specified without additional cost adders.
Schedule will be minimally affected since the vendor is aware of the complete scope prior to the bid being submitted and accepted.
L.L. 1.4: Failure to evaluate the vendor experience, scope of supply, and
specifications exceptions prior to money being discussed will result in lower
reliability
When the aforementioned has been reviewed after the bid has been submitted with money attached to it, we have seen users not being able to get their
reliability optimizing BP’s into the project. Also, scheduling delays in the order
of months have been experienced when users give input to the vendors after the
bid has already been accepted and engineering has commenced.
BENCHMARKS
This Best Practice has been used by Forsthoffer Associates, Inc., especially for
Critical (Un-Spared) Trains since the 1990 and assures optimum safety and reliability and maximum revenue over the life of the equipment at minimal costs. It
has been incorporated globally in all Upstream and Downstream Projects.
Optimum machinery train reliability has resulted from this Best Practice
saving upwards of $2,000,000.00 (minimum) per year for Plants with Daily
Revenue’s greater than $1,000,000.00.
SUPPORTING MATERIAL
Classifications of Rotating Equipment
Once the ITB has been released to the quoting vendors, the work begins. The
first order of business is to prepare for audits required at this stage. If the equipment in question is prototype, design and manufacturing audits have already
been initiated and are ongoing. If the equipment contains multiple major components that do not have field experience, audits are required in this phase.
These facts are shown in Table 1.4.1.
The concept of pre-bidding is very powerful and rewarding to all three
parties in the bid process—vendors, contractors, and end users. Pre-bidding
requires that all technical details are discussed, with appropriate changes for
7
B.P. 1.4
More Best Practices for Rotating Equipment
TABLE 1.4.1 Design and/or Manufacturing Audit Requirements in the PreBid Phase
j Finalize audit results and prepare project team recommendations for prototype class
equipment
j Interview quoting vendors and determine requirement for design and/or manufacturing audits during pre-bid phase for major equipment with multiple component
inexperience
optimum safety and reliability made before a price is quoted. The advantages of
this approach are presented in Table 1.4.2.
The pre-bid meeting is frequently called a bid clarification meeting. This
title can be misleading and may not have the same advantages as a pre-bid meeting. The significant differences are noted in Table 1.4.3.
It is most important to confirm the requirements and details of the pre-bid
meeting with the contractor in the beginning of the project. Table 1.4.4 presents
the benefits of conducting a true pre-bid meeting and not a bid clarification
meeting.
TABLE 1.4.2 Technical Discussions Before Priced Bids
Eliminates competitive pressures on the vendor by:
j Allowing technical review before price
j Assures the same scope for each supplier
j Assures offering of the highest safety and reliability
j Is performed regardless of risk classification
TABLE 1.4.3 Bid Clarification Versus Pre-Bid Meeting Differences
j Pre-bid meetings—are conducted before a price is quoted and allow for modifications to technical offering
j Bid clarification meetings—are conducted after a price is quoted and may not allow
for modification to technical offering
TABLE 1.4.4 The Pre-Bid Meeting
j The highest equipment reliability
j The lowest life cycle cost
j Equal scope of supply for each vendor
j Shortest bid evaluation cycle time
8
Project Best Practices
Chapter | 1
TABLE 1.4.5 Who Really Manufactures it?
Vendors frequently use sub-suppliers for:
j Lower component costs
j Reduced vendor machine shop investment
j Greater schedule flexibility
j Reduced in-house shop load
A pre-bid procedure fact summary and a typical agenda for a compressor
train are contained at the end of this section. It is recommended that this information be used to justify these meetings with the project management team as
early as possible in the project, preferably in the pre-FEED phase.
Due to competitive pressures, past union agreements, and high in-house
manufacturing costs, vendors have been forced to use numerous sub-suppliers
for major component and auxiliary system manufacture and in some cases, design. This approach exposes the end user to potential delivery delays due to
sub-supplier manufacturing, quality and schedule issues. These important facts
are presented in Tables 1.4.5 and 1.4.6.
Based on the potential sub-supplier problems noted earlier, when should
they be audited? The suggested action is noted in Table 1.4.7.
The final recommendation therefore is to always have vendors define major sub-suppliers and their experience during the pre-bid phase. Please refer to
Table 1.4.8.
At this point, all details concerning vendor experience, scope, exceptions
and sub-supplier experience have been identified. If the objectives of the prebid phase and any required audits have been met, the bid evaluation phase will
be short and easy since there will be a true apples to apples comparison and the
TABLE 1.4.6 Potential Sub-Supplier Issues
j Component scrap due to inexperience
j Component scrap due to improper machine tools
j Component scrap due to improper handling
j Poor or nonexistent inspection
j Delay in shipment
TABLE 1.4.7 Audit Sub-Suppliers
j Experience for similar components is low
j Equipment risk class is high
j End user “lessons learned” warrant
9
B.P. 1.5
More Best Practices for Rotating Equipment
TABLE 1.4.8 Always Require Definition of Sub-Vendor and Experience
j Casing
j Impellers and/or blades
j Diaphragms
j Shaft
j Baseplate
j Auxiliary systems
j Control panels
lowest price vendor can be selected without any additional meetings or discussions.
B.P. 1.5: Be proactive during the pre-award meeting by meeting and assessing the proposed project engineer (get and review cv)—Be sure to state
this requirement in the ITB.
Assure that the Vendor’s Project Engineer has experience with multiple projects for machinery of similar applications, including both driven and drivers.
If a vendor is asked to provide a CV of the project engineer they will be sure
to assign somebody with experience.
If the Project Engineer has worked on projects within your company in the
past, interview the project team who worked with him to get their opinions on
the honesty and technical knowledge of the engineer.
L.L. 1.5: Project Engineers with limited experience has resulted in delays
and reliability issues
Limited Experience by the Project Engineer will tend to have multiple misinterpretations of the technical scope, which have ended in projects that have
been delayed by several months.
BENCHMARKS
This Best Practice has been used by Forsthoffer Associates, Inc., especially for
Critical (Un-Spared) Equipment since the late 1990s when an influx of new
young engineers had entered the industry and extended project schedules by
months! It has been incorporated globally in all Upstream and Downstream
Projects since 1999 by Forsthoffer Associates.
SUPPORTING MATERIAL
Pre-Award Meeting
After completion of the bid tabulation and approval of the selected vendor,
confirmation of the approved vendor’s proposal details are required prior to
10
Project Best Practices
Chapter | 1
TABLE 1.5.1 Pre-Award Meeting “Key Facts”
j Purpose—to assure agreed compliance
j With who?—the recommended vendor
j When?—ASAP after the bid tab is approved
j Where?—depends on complexity and risk class
j Confirm—marketing to engineering continuity
TABLE 1.5.2 The Pre-Award Meeting Agenda
j Assure the attendance of vendor marketing and project engineer
j Agenda to be prepared by contractor/end user
j Agenda contents:
j Scope of supply confirmation
j Clarification and agreement of all exceptions to specifications
j Resolution of pending design audit issues
j Confirmation of price and delivery schedule
j Agreement of minutes and action points
the award of an order. In my experience, there have been many times when
the vendor’s marketing and engineering departments have had significant differences of opinion in regard to what was actually sold. The purpose of the
pre-award meeting therefore is to confirm order content before a contract to
eliminate additional costs and delays during the equipment engineering and
manufacturing phases. Key facts regarding the pre-award meeting are presented in Table 1.5.1.
A suggested outline for a pre-award meeting is noted in Table 1.5.2.
B.P. 1.6: Establish and enforce a drop dead date for process conditions to
be finalized, as early in the project as possible
While it is important to assure all process conditions are considered, there
must not be any process changes after the Vendor has submitted their bid.
Having all final Process Conditions submitted to the vendor at once will assure that there are no changes in machinery scope.
Specific Process Conditions required, include but are not limited to:
l
l
l
l
l
Gas Process Conditions for Compressor (this can determine the Gas Seal
System components as well)
Steam Conditions for Steam Turbines
Fuel Conditions for Gas Turbines
Cooling Water Conditions
Dry Gas Seal external process gas conditions
11
B.P. 1.7
More Best Practices for Rotating Equipment
L.L. 1.6: Process Conditions changing after the Vendors have placed their
bids has resulted in redesign of the machinery and significant cost adders/
delay of schedule
Changes in Process Conditions after the bids have been placed by the vendors ended in larger casings having to be provided for Compressor and Turbines
and Dry Gas Seal failures, due to failure to indicate contaminants and/or saturated components in the gas analysis.
Also, several delays have resulted when process conditions have changed
during the Engineering phase of the machinery that was purchased.
BENCHMARKS
This Best Practice has been used by Forsthoffer Associates, Inc. since 1990 and
for Projects in all Industrial Plant Applications including but not limited to:
l
l
l
l
Ammonia Syngas and Refrigeration Compressor Trains
Methanol Syngas and Circulator Compressor Trains
LNG—PRC, ERC, and Mixed Refrig. Compressor Trains
Ethylene Plant—PGC (CGC), Propane, and Ethylene Refrig. Compressor
Trains
This approach has resulted in machinery being selected to proper specs with
minimal scheduling delays at minimum capital cost.
B.P. 1.7: Think “Outside the Flanges” when preparing datasheets. Include
process and system effect details into the datasheets
The vendors are trying to design the equipment to meet all of the user’s
needs; however they do not fully know the process and the objectives of the
end user.
If the vendor knows fully how the process is designed and what issues could
occur during the life of the plant they may be able to incorporate provisions into
the design of the equipment in order to keep it operating at maximum efficiency
and reliability.
Discuss the process flow diagram (PFD) and piping and instrumentation
diagram/drawing (P&ID) for each process with experienced Process Engineers
from the Process Licensors, Project Team and/or Plant Process Engineers and
Operators.
Define, with the concurrence of the experienced Process Engineers and Operators the potential upset and unusual process conditions.
Be sure that all upset and unusual process conditions are defined on the appropriate equipment data sheets.
Require that an experienced Process Engineer and Operator be part of the
Machinery Project Team and attend all planning sessions, specification and data
12
Project Best Practices
Chapter | 1
sheet reviews, as well as all vendor meetings during the Pre-Selection, Design,
and Test Phases.
L.L. 1.7: It is the writer’s experience that approximately 80% of Machinery Failure Root Causes are due to process variations not anticipated in the
design phase
Detailed Root Cause Failure Analysis (RCFA’s) show that the majority of
Root Causes of Failure lie in unanticipated process condition changes. Failure
to incorporate these upset and unusual conditions on the appropriate equipment
Data Sheets can:
l
l
l
l
Extend Project Schedule by requiring Machinery Re-Design
Result in significant cost adders for Machinery Re-Design
Reduce Machinery Safety and Reliability
Extend Plant Start-Up
BENCHMARKS
Since the late 1990s due to increased process unit size and potential Daily
Revenue Losses, we have required detailed Process reviews between Machinery
Specialists, Process Engineers, and Operators prior to issue of Specifications
and Data Sheets. This action has been implemented for all Projects large and
small.
Results from this Best Practice have been:
l
l
l
elimination of changes after issue of the Purchase Order
reduced machinery design time and early machinery delivery
no surprises during plant start up
Supporting Material
There are four basic function classifications of rotating equipment. Refer to
Table 1.7.1, which defines the classifications of rotating equipment.
Each Machinery Train or Unit is made up of all of the four Classifications.
The Safety and Reliability of the Train is directly related to the proper selection
TABLE 1.7.1 Classifications of Rotating Equipment
j Driven
j Drivers or prime movers (provide power)
j Transmission devices
j Auxiliary equipment
13
B.P. 1.7
More Best Practices for Rotating Equipment
TABLE 1.7.2 Major Types of Rotating Equipment
I. Driven equipment
II. Drivers—prime movers
A. Compressors
A. Steam turbines
1. Dynamic
2. Positive displacement
B. Gas turbines
Centrifugal
Screw
Axial
Rotary lobe
Induction
Integral Gear
Reciprocating
Synchronous
Diaphragm
Liquid Ring
C. Motors
Vari-speed
D. Engines
Internal combustion
B. Pumps
1. Dynamic
Diesel
2. Positive displacement
Centrifugal
Plunger
Axial
Diaphragm
Slurry
Gear
Integral Gear
Screw
Gas
Progressive cavity
C. Extruders
D. Mixers
E. Fans
III. Transmission
devices
IV. Auxiliary equipment
A. Gears
A. Lube and seal systems
Helical
B. Buffer gas systems
Double helical
C. Cooling systems
B. Clutches
C. Couplings
and design of each of these classifications. Failure to consider the proper experience, selection, and design of each Train component will result in lower Train
Safety and Reliability. Table 1.7.2 is a partial listing of some rotating equipment
types grouped according to Classification (Function).
Site Equipment Examples
Shown later are examples of typical site rotating equipment.
Figs. 1.7.1–1.7.4 show examples of each rotating equipment classification.
14
Project Best Practices
Chapter | 1
FIGURE 1.7.1 High-pressure centrifugal compressor. Courtesy of Dresser Rand.
FIGURE 1.7.2 Extraction—condensing steam turbine. Courtesy of MHI.
15
B.P. 1.8
More Best Practices for Rotating Equipment
FIGURE 1.7.3 Multiple, convoluted diaphragm-spacer coupling. Courtesy of Zurn Industries.
FIGURE 1.7.4 Horizontal oil console arrangement. Courtesy of Oltechnique.
B.P. 1.8: Assure Plant experienced personnel are involved in all phases of
the project
All major disciplines within the plant have input which governs the design
of the major components of machinery; therefore their inclusion is a must from
the beginning to the conclusion of the project.
Experienced plant personnel involved in the project shall include Process
Engineers, Machinery Engineers, Reliability Engineers, Electrical and Instrumentation Engineers, Maintenance Engineers, and Operations.
16
Project Best Practices
Chapter | 1
L.L. 1.8: Failure to have the appropriate experienced personnel involved
in the project result in decreased reliability and loss of revenue
When the appropriate experienced personnel have not been involved
throughout the project, the following results have occurred:
l
l
l
l
Delayed Bid Selection—adding weeks or months to the schedule.
Cost adders for items that were not defined or exceptions to specifications
that were not resolved during Bid Clarification.
Delays in the manufacturing schedule for incorporation of items not resolved during the Bid Phase.
Lower Field Reliability and possible Safety issues resulting from selection
of the low cost bidder.
BENCHMARKS
This approach has been used since 1975 in all Critical Equipment Selection
since 1975 that has produced Machinery of Optimum Safety and Reliability
(Compressor Trains = 99.5%+). Pre-Bid Meetings have been used in all of the
following Industries globally:
l
l
l
l
Upstream Oil and Gas—Offshore and Onshore
Refining
Chemical
Co-Generation
SUPPORTING MATERIAL
Suggested Vendor Pre-Bid Meeting Details Letter
The following are suggested letter contents. Comments for consideration are
noted in bold.
Please be advised that you will be asked to attend a pre-bid meeting at
____________ (Your or Contractor’s or Company … decision required) offices.
Note: If possible, the meeting should be held at the vendor’s offices. This
is advisable since more experienced specialists are immediately available
to answer any questions that may arise. This decision should also be influenced by the machinery risk classification. The higher the risk, the more
important the vendor office meeting is.
The pre-bid meeting will take place approximately _______ (2–4 weeks
after receipt of bid and must be coordinated with the project team … note
this decision will also be influenced by the machinery risk class).
Only technical details will be discussed. Please bring the technical, unpriced proposal for the equipment that you will quote (Trains include compressor, gear (if applicable) turbine and auxiliary systems). Your representatives at
the meeting must include an experienced application, instrument engineer and
any other personnel you require.
17
B.P. 1.8
More Best Practices for Rotating Equipment
The meeting objective will be to qualify your bid technically based on
component experience and to fully define scope of supply and approve exceptions to specifications. If necessary, the technical aspects for your equipment
may change as a result of the meeting discussions. In addition to our (contractor) equipment specialist the end user specialist ________ (or other assigned
engineer) will participate in these meetings. We emphasize that it is in your
interest to bring the most qualified personnel to the meeting since this will be
the only technical meeting prior to the final bid.
The following additional audits may be required as a result of your bid details and the use of major sub-suppliers.
____________ (The end user to identify sub-suppliers for ____________
manufacturing, handling and shipping audits based on vendor bid details)
At the conclusion of the meeting, all details will be summarized and you will
be asked to submit your priced proposal in accordance with the technical details, scope of supply and approved exceptions to specifications agreed to in the
meeting. Your proposal will be required in ________ (Normally 2 weeks but
may be longer based on complexity of equipment offering and machinery
risk classification) weeks after the meeting.
The pre-bid meeting agenda is attached.
Pre-Bid Procedure Fact Summary
The following is a brief summary of the salient procedure facts:
1. Required personnel experience—experienced rotating equipment specialists
from contractor, supplier(s), and client are required to participate in the prebid meetings. Note: End user “in-house” specialists are required.
2. Individual supplier meetings—individual meetings are held with each supplier, using notes from previous meetings, to assure equal supplier experience, scope, and exceptions to specifications.
3. Meeting duration—anticipated 1–2 days per major equipment train depending on machinery risk classification. Note: this includes compressors, drivers, and auxiliaries. Please refer to the typical agenda later.
4. Typical pre-bid meeting activity—Technical details are reviewed using
agenda requirements to assure proven component experience (impellers,
diffusers, rotor response, bearing, seal, and auxiliary system, etc.), scope
compliance and acceptable exceptions to specifications.
Modifications are made, as necessary, to assure that each vendor is offering
proven components within acceptable design limits.
Manufacturing capabilities are confirmed and sub-suppliers for all major
components and auxiliary systems are identified and their experience is confirmed for similar component manufacture.
At the conclusion of the meeting, notes are reviewed and each vendor is
instructed to submit a final priced proposal, in full accordance with meeting
notes that will be used for the bid evaluation.
18
Project Best Practices
Chapter | 1
Depending upon machinery risk classification, additional end user in-house
and/or independent 3rd-party design checks may be required. In addition,
separate vendor and sub-supplier machining, handling, and shipping capability audits may be required.
Typical Compressor Train Pre-Bid Meeting Agenda
Please note that the following agenda will be followed for each of the compressor trains being offered. Note: 1–2 days will be required for the meeting to
review all details based on unit risk classification.
1. Compressor experience review (vendor to include necessary reference
charts, tables, etc.)
l Casing experience and review of compressor layout drawing
l Impeller experience (flow and head coefficient)
l Individual impeller curve (location of rated point to impeller best
efficiency point)
l Impeller stress
l Rotor response
l Stability analysis (if applicable)
l Bearings—surface speed, load, and experience
l Thrust balance
l Seals—surface speed, balance forces, and experience
l Surge control and process control system
2. Steam turbine or motor experience review (vendor to include necessary
reference charts, tables, etc.)
l Turbine casing experience and review of layout drawing
l Stage nozzle and blade experience (profile, velocity ratio, BTU/stage)
l Blade attachment method and blade stresses
l Campbell and Goodman diagram review
l Rotor response
l Bearings—surface speed, load, and experience
l Thrust balance (reaction and hybrid types)
l Shaft seals
l Transient torsional response experience review (synchronous motors)
l Control and protection system
3. Gear experience (if applicable) (vendor to include necessary reference
charts, tables etc.)
l Gear box experience review and review of layout drawing
l Review of gear data sheet
l Gear calculation review (in accordance with API 613)
l Bearings—surface speed, load, and experience
l Thrust loading—single helical gears
l Pitch line velocity review
19
B.P. 1.9
More Best Practices for Rotating Equipment
4. Auxiliary system experience (lube, dry gas seal, and control oil system)
l Review of P&IDs
l Review of API 614 data sheets
l Review of typical arrangement drawings
l List of experienced system sub-suppliers
l Review of proposed dry gas seal supplier information
5. Scope of supply for compressor train (all components and auxiliaries)
review
6. Compressor train (all components) exceptions to specification
7. Meeting summary and action required
Note: Based on machinery risk, the following design checks may be
required:
l Aero-dynamic
l Thermodynamic
l Rotor response
l Stability analysis
l Seal balance
l Thrust balance
l Bearing loading
l Control system simulations
l System layout maintenance accessibility
B.P. 1.9: Obtain CV’s of EP&C Machinery Engineers prior to EP&C
Award
This will allow the project group to evaluate and compare the experience of
the Machinery Engineers selected by the EP&C’s.
Ideally, the EP&C machinery engineers should have some field experience,
as well as design experience.
Note that if this BP is followed, typically the EP&C will assure the most
experienced Engineers available are involved in the project.
L.L. 1.9: Failure to vet the EP&C Machinery Engineers can result in
acceptance of unreliable equipment
If an inexperienced EP&C Machinery Engineer is used, the following issues
(among others) may occur in the project:
l
l
l
20
Ease of maintenance may not be considered, resulting in longer Mean Time
To Repair (MTTR). IE. Trapped Barrel Compressors or Horizontal Split
Casings with Upward Nozzles.
May be talked into cheaper options, such as Field Performance Test, which
can result in long delays in startup.
Selection of Prototype machinery and/or components could be selected,
resulting in a learning curve, and potential poor reliability.
Project Best Practices
Chapter | 1
BENCHMARKS
This Best Practice has been used by Forsthoffer Associates, Inc., especially for
Critical (Un-Spared) equipment since the late 1990s when an influx of new
young engineers had entered the industry and extended project schedules by
months! It has been incorporated globally in all Upstream and Downstream
Projects since 1999 by Forsthoffer Associates.
SUPPORTING MATERIAL
The action taken during the pre-FEED phase (Front End Engineering Design)
relative to rotating equipment will set the stage for its availability and profit
improvement for the entire life of the process unit. However, the company must
take the initiative to assemble and brief the project team members immediately
upon project inception. Corporate responsibilities are outlined in Table 1.9.1.
The corporate action outlined in Table 1.9.1 will enable the specialist to acquire the project information that she or he needs to determine the degree of risk
involved for the purchase and manufacture of the critical (custom designed) equipment on the project. In addition, the specialist can prepare the Machinery Best
Practice List from Company and Plant Lessons Learned being sure to only select
those Best Practices that affect safety and produce significant revenue increases.
The manner in which Project Recommendations are presented will have a great
impact on Project Team trust and support. Table 1.9.2 presents these facts.
TABLE 1.9.1 Corporate Project Responsibilities
j Assemble the entire project team immediately upon project announcement
j The team should include existing plant maintenance, operations personnel or experienced personnel if the plant is a grass roots (new) installation
j Brief specialists regarding details of process, size of equipment special details, etc.
j Require specialist input immediately regarding equipment special project requirements (Best Practices)
TABLE 1.9.2 Guidelines for Project Plan Presentation
j Present Best Practices based on Lessons Learned
j Benchmark each Best Practice in terms of Safety and Revenue Increase
j Present envisioned equipment overview based on input data, estimation calculations, prior vendor discussions and best practices
j Define risk—safety and cost of un-availability
j Recommend action plan—include: audit, application best practice, and special test
requirements
j Define cost and additional schedule time based on recommended action plan
j Define company savings and increased profit over life of process unit
j Request decision to proceed with the proposed plan
21
B.P. 1.10
More Best Practices for Rotating Equipment
TABLE 1.9.3 Refrigeration Compressor Selection Case History
j Input from project team—propylene refrigeration duty data sheet
j Calculations and vendor discussions showed that duty required a prototype machine
in regard to rotor bearing span and shaft diameter (shaft stiffness)
j Risk class was determined as multiple component inexperience
j Vendors were invited to pre-screening design review meetings to determine action
j Based on meeting reviews with three vendors, it was determined that bearing span had to
be reduced and that two compressor cases, in series, were required for proven reliability
j Costs of second case were assembled along with supporting data and cost figures for
exposure to reduced availability and benchmarks of problems experienced with the
one case option (this “Lesson Learned” information was obtained from experienced
plant maintenance and operations personnel)
j The management presentation was successful and additional 5 $MM was approved
for purchase and installation of the second compressor casing
The format of this presentation can range from a discussion with the project
engineering manager and project manager to a formal power point presentation. Regardless of Project Team trust and type of presentation, time is of the
essence and the presentation must detail in clear and concise terms, the specific
requirements, schedule time and life cycle cost savings for the proposed plan.
An example of turning acquired pre-FEED information into an action plan is
presented in the case history in Table 1.9.3.
The previous action was possible because the project team provided early
information to the specialists, allowed pre-screening meetings to audit vendor
experience and took the specialist’s recommendations seriously.
This action took place before the budget estimate. The example outlined in
Table 1.9.1 will become more important, in the future, as the size of projects
increase and the exposure to loss of daily process unit profit can easily exceed
millions of dollars.
B.P. 1.10: Assure construction specifications are included in ITB to Construction Contractors
In order to do this the user must have a specification for construction of new
equipment, which many do not have.
This standard specification, once established, needs to be included in all
projects as early as possible (before the construction contractor is selected) to
assure the proper standards are followed.
L.L. 1.10: Failure to include construction specifications in the ITB to the
Construction Contractor can result in possibly long delays in construction
schedule
Issues that could occur can be but are not limited to the following:
l
22
Improper grouting procedure resulting in poor foundation and excessive
bearing force.
Project Best Practices
l
l
Chapter | 1
Improper Piping installation resulting in excessive piping stress and excessive bearing loads.
Missing components, IE Check valves on discharge of compressor, therefore a large volume of high-pressure gas can act on the compressor during
an ESD and result in backwards rotation.
BENCHMARKS
The writer has used the aforementioned approach in all Critical Equipment
Projects since 1990 that has resulted in “smooth, issue free project construction.” This approach has been used in all Upstream Oil and Gas Projects—Offshore and Onshore
l
l
l
Refining
Chemical
Co-Generation
SUPPORTING MATERIAL
Table 1.10.1 lists ITB instructions to vendors.
B.P. 1.11: Prepare FAT Scope to cost effectively duplicate field conditions
as closely as possible
In order to get accurate indication of how the compressor is going to operate
in the field the performance test should be run at conditions as close as possible
to the field conditions.
This should include a test gas that allows the equivalent speed (Test Speed)
to be at least 80% of the Field Operating Speed. If that is followed, then the
thrust load and thermal growth will be much closer to field operation.
L.L. 1.11: Improper specified FAT Scope can result in unexpected reliability issues during initial start-up
The writer’s experience has been that inability to properly specify the
FAT has resulted in Thrust Bearing displacement and pad temps operating
TABLE 1.10.1 ITB instructions to vendors
j Incorporate all project team accepted items (Design Audit, Best Practices, Pre-Bid
Meetings, and Test Requirements etc.)
j Include Pre-Bid Meeting Instructions (when, where and who attends)
j Include Design Audit Details (when, where, and who attends)
j Include Construction Specifications
j Define discipline and experience requirements for all participants in all scheduled
meetings
j Note penalty for noncompliance (e.g., bid not accepted)
23
B.P. 1.11
More Best Practices for Rotating Equipment
aforementioned alarm levels with premature failures, along with Inter Barrel
Split Lines opening up due to thermal growth (efficiency loss).
BENCHMARKS
Use of the aforementioned approach in all Critical Equipment Projects since
1990 has resulted in machinery that started-up without any delays due to unexpected operating conditions of the components and has been applied in the
following applications:
l
l
l
l
Upstream Oil and Gas—Offshore and Onshore
Refining
Chemical
Co-Generation
SUPPORTING MATERIAL
Testing Phase
The testing phase is the last phase in terms of vendor and sub-supplier design
and manufacturing involvement in the project and … the last chance to assure
the optimum availability of the finished product.
Remember that all of the equipment addressed in this section is most likely
custom designed and no matter how much accrued design and manufacturing
experience is present, the possibility of some abnormality, hopefully minor, is
high. Therefore, it is imperative that this phase be carefully observed and witnessed by the end user team. Table 1.11.1 lists important facts surrounding this
phase of the project.
I have included a shop test checklist at the end of this section that will be
valuable in planning and executing the shop test phase. Yes there certainly are
many opportunities to assure equipment reliability during shop test but there
are also a lot of potential lost opportunities if they are not justified to the project
team early, during the pre-FEED phase, of the project. The potential lost shop
test phase opportunities are noted in Table 1.11.2.
TABLE 1.11.1 The Shop Test
j Confirm vendor proper design and manufacture
j To match field conditions
j To witness assembly and disassembly using job special tools
j To have plant personnel observe test, assembly, etc. and take pictures for purposes
of emergency field maintenance excellence
j Review the instruction book
j Have the assigned vendor field service engineer observe the equipment he will
install
j Review all vendor field procedures
24
Project Best Practices
Chapter | 1
TABLE 1.11.2 Potential Lost Test Opportunities
The following opportunities will be lost if they are not justified at project inception:
j Possible full load test
j Unproven component tests
j Attendance at test by plant personnel
j Use of special tools
j Vendor permission for pictures
j Agreement that assigned vendor field service specialists will be present for tests
j Agreement that the instruction book is reviewed
j Agreement for formal field construction meeting to clearly define all vendor procedures from receipt of equipment on site to initial run in of equipment
The success of the shop test depends on a good test plan that is reviewed by
the end user and contractor and modified as requested well in advance of the
test. Table 1.11.3 presents these facts.
I actually began my career in rotating equipment on the test floor. And I
can still remember how we would see the witnesses come in with an intent to
completely participate in the entire test only to leave for a long “test lunch” an
hour or so later. Why did this occur? Usually because the concerned end user
and contractor witnesses did not have the opportunity to review the test set up
and the procedure prior to the test. As a result, I have always been a proponent
of a pre-test meeting.
Is it always required? I think it is but the detail and timing of the meeting
depends on certain factors. These factors are noted in Table 1.11.4.
TABLE 1.11.3 Shop Test Agenda Review—Key Facts
j The agenda is issued for review 2 months prior to test
j It incorporates agreed VCM scope
j Compressor performance test conditions are per ASME PTC–10 requirements
j A sample of test calculations and report format is included
j Vendor concurs with all end user and contractor comments prior to test
TABLE 1.11.4 When is a Pre-Test Meeting Required?
j If the equipment is prototype
j If the equipment is complex
j If a full load test is required
j If the test facility is new
25
B.P. 1.11
More Best Practices for Rotating Equipment
TABLE 1.11.5 Pre-Test Meeting—Key Facts
j Conduct the meeting prior to the test day
j Send the agenda to the vendor well in advance
j A typical agenda outline:
j Confirm test agenda requirements
j Confirm all test parameter acceptance limits
j Confirm instrument calibration
j Review test set up or concept drawing
j Review data reduction methods
j Confirm all test program agreements
If it is decided to conduct a pre-test meeting, the key facts are noted in Table 1.11.5.
Tables 1.11.6–1.11.8 define recommended test activity for the mechanical,
auxiliary equipment and performance shop tests respectively.
At the conclusion of all test activities, there is still important work to be
performed. These items are defined in Table 1.11.9.
What happens if the test is not successful? Approximately 50% of the tests
that I have either run or participated in over my career have not been successful
in regards to one component or more not meeting test requirements. Possible
rejected test action is noted in Table 1.11.10.
TABLE 1.11.6 Mechanical Test—Key Facts
j Per API and project requirements
j Confirm all components are installed
j Confirm all accessories are installed
j Monitor progress of test, look for leaks, etc.
j Do not accept test until all requirements are met
TABLE 1.11.7 Auxiliary System Test—Key Facts
j Must be per API and project requirements
j Confirm that the test agenda is followed
j Confirm all components are installed
j Confirm that all required instruments are installed
j Monitor the progress of the test—look for leaks, etc.
j Do not accept until all requirements are met
26
Project Best Practices
Chapter | 1
TABLE 1.11.8 Compressor Performance Test—Key Facts
j Per ASME PTC-10 requirements
j Reconfirm test speed is per PTC-10
j Confirm all instruments are calibrated and installed
j Confirm test gas purity
j Agree that conditions are stable prior to each test point
j Confirm vendor’s calculations for each test point
j Do not accept until all test requirements are met
TABLE 1.11.9 Post Test—Key Facts
j Confirm performance results, corrected to field conditions
j Confirm mechanical test acceptance
j Confirm auxiliary system test acceptance
j Inspect components and confirm acceptance
j Agree to any corrective action in writing
j Accept or reject test—any corrective action requires a retest!
TABLE 1.11.10 Rejected Test Action
j Immediately provide details to the project team
j Confirm if field conditions can handle the abnormality
j Determine if the “as tested” machine will meet all reliability requirements
j If the decision is to reject, inform the vendor and detail the reasons
j Do not accept unrealistic delivery delays
Finally, do not forget the importance of test report requirements. The test
report is a most important document that represents the “baseline performance
of the unit” and will be a benchmark for field operation acceptability. Test
report—key facts are noted in Table 1.11.11.
TABLE 1.11.11 Test Report—Key Facts
j The shop test is the field baseline!
j The test report must be detailed and complete
j Review the preliminary contents of the report before leaving the test floor
j Obtain the actual test results
j When the final report is received, check the results obtained at test against the final report
j Immediately contact the vendor if there are any differences
27
B.P. 1.11
More Best Practices for Rotating Equipment
Shop Test Checklist
1. Scope
j Appropriate industry specs included (ANSI, API, NEMA, etc.)
j In-house and/or E&C specs included
j Project specific requirements
h Performance test h All rotors h One rotor
h Test (equivalent) conditions
h Field (actual) conditions
h Mechanical test h All rotors h One rotor
h Test (equivalent) conditions
h Field (actual) conditions
h Unit test of all equipment (string test)
h No load h Includes auxiliary systems
h Full load h Does not include auxiliary systems
h Use of job couplings and coupling guards
h Testing of instrumentation, control, and protection devices
h Auxiliary system test
h Lure oil h Test press h Full press
h Control oil h Test press h Full press
h Seal oil h Test press h Full press
h Seal gas h Test press h Full press
h Fuel h Test press h Full press
h Flow measurement required
h Time base recording of transient events required
h Use of all special tools during test (rotor, removal, coupling, etc.)
h Shop test attendance (includes assembly and disassembly)
h Site reliability h Site maintenance h Site operations
h Review of instruction book during shop test visit
h Test agenda requirements
h Mutually agreed limits for each measured parameter
h Issue for approval 2 months prior to contract test date
2. Pre-test meeting agenda
j Meet with test department prior to test to:
h confirm test agenda requirements
h confirm all test parameters have mutually agreed established limits
h review all instrument calibration procedures
h review test set-up drawing
h review data calculation (data reduction) methods
h define work scopes for site personnel (assembly and disassembly
witness, video or still frame pictures, etc.)
j
Confirm assigned vendor service engineers will be in attendance for:
h Assembly
h Disassembly
h Test
28
Project Best Practices
Chapter | 1
3. Shop test activity
j review and understand test agenda prior to test
j immediately prior to test meet with assigned test engineer to:
h review schedule of events
h designate a team leader
h confirm test team leader will be notified prior to each event
h “walk” test set-up to identify each instrumented point
h confirm calibration of each test instrument
h obtain documents for data reduction check – if applicable (flow
meter equations, gas data, etc.)
j During test
Note: coordinate with test personnel to avoid interference
h review “as measured” raw data for consistency
h “walk” equipment—look for leaks, contract instrument, piping, and
baseplate vibration, etc.
h use test team effectively—assign a station to each individual
h ask all questions now, not later, while an opportunity exists to correct the problem
h check vendor’s data reduction for rated point—if applicable
j After test
h inspect all components as required by the test agenda (bearings,
seals, labyrinths, RTD wires, etc.)
h review data reduction of performance data corrected to guarantee
conditions
h review—all mechanical test data
h generate list of action (if applicable) prior to acceptance of test
h approve or reject
B.P. 1.12: Always benchmark Best Practice recommendations, showing
results in: MTBF, MTTR, revenue savings, safety, and emissions
In order to implement a solution to a major machinery issue (whether new
or existing equipment), the action plan needs to be identified as a cost effective
solution to the actual cause of the problem. Cost Effective, means it either has
to be justified by costing less over a period of time than the loss of revenue that
has been incurred (increased MTBF and/or reduced MTTR).
Likewise if the issue is a safety or environmental hazard, the project team
(and/or upper management) will most likely approve the action, as these two
items are the most important parameters considered by upper management.
L.L. 1.12: Failure to benchmark Best Practices properly result in unresolved machinery issues and continuation of revenue lost
Since 1990 FAI has been involved with many Field Troubleshooting assignments where we made recommendations that were ultimately followed,
however it was brought to our attention that these same recommendations
were being made by plant personnel without the same success rate. Failure to
29
B.P. 1.13
More Best Practices for Rotating Equipment
accurately indicate the consequences of not implementing the Best Practices
is the reason.
BENCHMARKS
The writer has used the aforementioned approach in all Critical Equipment
Projects since 1990 and especially since 2000 when MEGA projects became
common in the Industry. This approach has resulted in Safety and Reliability
issue free projects without significant cost adders and schedule delays. This approach was used for the following recent projects:
l
l
l
l
Refinery Hydrocracker Recycle Compressor
LNG Mixed Refrigerant Compressor
Methanol MAC and BAC Air Separation Train
Ethylene Refrigeration Compressor
SUPPORTING MATERIAL
Please see Supporting Material for BP 1.1.
B.P. 1.13: It is essential to have experienced personnel involved in MOC
and HAZOP studies
The studies mentioned previously are required, important, and can be effective as long as experienced personnel are involved.
This will result in much quicker acceptance and implementation of action
plans and eliminate second guessing to hazards that will never occur.
L.L. 1.13: Inexperience in MOC’s has resulted in Implementation of machinery Issue resolution being delayed or canceled
The writer has personally been involved in RCFA’s that have resulted in
action plans that have been delayed several months due to the MOC team not
fully understanding the technical reasons for making the recommendations.
BENCHMARKS
The writer has used the aforementioned approach in the last 10 years when
MOC’s have been a very common practice. By assuring the MOC team being
technically experienced, implementation time has been significantly reduced,
therefore reducing revenue lost.
SUPPORTING MATERIAL
Design and Manufacturing Audits
Design and manufacturing audits; as previously stated, are required based on
the equipment risk class and vendor and sub-supplier design and manufacturing
30
Project Best Practices
Chapter | 1
TABLE 1.13.1 Vendor Audit Requirements
j Detailed agenda, well in advance
j Design audit at vendor’s offices with follow-up at end users offices
j Manufacturing audit at vendor’s and/or sub-suppliers plants
j End user specialists must participate
j Conduct preliminary end user in-house checks prior to the design audit if possible
experience level. These audits can be conducted at any phase of the project but
the sooner the better. Prototype equipment requires that audits be conducted
during the pre-FEED or FEED phase of the project. Today, most projects are
defined as MEGA projects since the process units are the largest size ever built
and most probably will incorporate single equipment trains that are prototype in
nature. Therefore, many projects require that design audits (pre-screening) be
conducted immediately upon project start. Planning and conducting effective
supplier design and manufacturing audits require pre-planning and a significant
amount of work, but it is certainly worth the effort in terms of increased profits
and reduced project schedule. These salient points are noted in Table 1.13.1.
When supplier or sub-supplier manufacturing audits are required, suggested
action is shown in Table 1.13.2.
In Table 1.13.3, I have presented a suggested list of what the design audit
should include based on the risk classification.
After conducting the appropriate audits, prompt follow up regarding any action items is required to confirm acceptance of the supplier and/or sub-suppliers
and to maintain the project schedule. Table 1.13.4 presents these facts.
After completion of the required audits, regardless of what project phase in
which they are conducted, follow-up document review is essential to confirm
that all stated design and manufacturing requirements are met.
In Table 1.13.5, I have presented a typical design audit meeting agenda to
resolve a long-term compressor seal oil system problem. Always remember to
be sure to send agendas well in advance to allow the vendor sufficient time for
preparing the required material.
TABLE 1.13.2 Manufacturing Audit Guidelines
j Machining capabilities (max, size capability)
j Balancing capabilities (low speed and/or high speed, max rotor size)
j Size of assembly area
j Shop load status
j Testing capabilities (gas test, full load test, power limits)
j Handling capabilities (max lift, lay-down area)
j Shipping capabilities
31
B.P. 1.14
More Best Practices for Rotating Equipment
TABLE 1.13.3 Suggested Design Audit Activity
1.
Risk type
2.
Design Checks
1
2
3
4
j Aero-dynamic
X
X
?
*
j Thermodynamic
X
X
?
*
j Rotor response
X
X
?
*
j Stability analysis (if applicable)
X
X
?
*
j Seal balance
X
X
?
*
j Thrust balance
X
X
?
*
j Bearing loading
X
X
?
*
j Train lateral analysis
X
X
?
*
j Torsional analysis (if applicable)
X
X
?
*
j Transient torsional (if req’d)
X
X
?
*
j Control system simulations
X
X
X
X
j System layout—accessibility
X
X
X
X
1, Prototype; 2, multiple component inexperience; 3, single component inexperience; 4, proven
experience for all components; X, required; ?, optional; *, not required.
TABLE 1.13.4 Design Audit Summary and Follow-Up Action
j Prepare an executive summary of conclusions
j Immediately present to the project team for approval
j Inform vendor’s of results
j Prepare vendor follow up meeting agenda
Note action required and follow up as required to maintain project schedule
Document Review
It goes without saying that document review should definitely be timely, within
the project schedule and accurate. However, in addition there are other pertinent
facts, which are presented in Table 1.13.6.
B.P. 1.14: Have a cold eyes design review as early as possible (during
VCM)
If the Best practices aforementioned have been fully followed, there should
be minimal issues to review technically from the vendor after engineering has
begun; however performing a cold eyes review during the VCM will catch any
little items that have not been thought of prior to this time.
32
Project Best Practices
Chapter | 1
TABLE 1.13.5 Design Audit Agenda
Lube/seal oil system
1. Introductions
2. Purpose of meeting
2.1 Review study results, past modifications/failures, and recommendations to assist
client in resolving seal oil delta pressure trips on the subject compressor.
2.1.1 Supply client with recommendations, modifications required, and cost and
delivery.
3. Results of studies performed in the field
3.1 Client field reliability study
3.2 Vendor engineering study
4. Review seal design and requirements
4.1 Review seal components and function
4.1.1 Upgrades?
4.1.2 Modifications?
5. Review seal oil system component design
5.1 Sizing of components (pumps, coolers, filters, reservoir, etc.)
5.2 Review valve selection and sizing (including Cv)
5.3 Review control system
5.3.1 Upgrades?
5.3.2 Modifications?
6. Review comments to seal oil system component sizing study
6.1 List recommendations
6.1.1 Feasibility and reliability issues
7. Review final recommendations and feasibility
7.1 Assign tasks and schedules
7.2 Create final timeline up to delivery of parts and installation
8. Conclusion
TABLE 1.13.6 Effective Document Review Considerations
j Assure that required review time frames are realistic and then meet them!!!
j Thoroughly review all items
j Question all required items and follow up
j Be especially careful in the final phases of the project to assure that all required
vendor changes have been made
33
B.P. 1.15
More Best Practices for Rotating Equipment
L.L. 1.14: Failure to hold a cold eyes review at this stage has resulted in
schedule delays
The writer and his company have been involved with numerous projects
since 1990 and in all of our experience, there is always something caught during
the VCM Stage of the Project.
Many RCFA’s we have been called in to perform were the result of issues
that were forgot about during the project and would have most likely been
caught during a cold eyes review, which if done during the VCM would have
resulted in minimal if any delays.
BENCHMARKS
The writer has used the aforementioned approach in all Critical Equipment
Projects since the early 1990s.
This approach has resulted in Safety and Reliability issue free projects without significant cost adders and schedule delays.
SUPPORTING MATERIAL
Please see supporting material for BP 1.13.
B.P. 1.15: Assure all issues that expose the end user to safety, revenue lost
and emissions issues are documented by notes on the datasheets.
Stating these exposures specifically in the datasheets will allow the vendors
to better understand what happens outside of the flanges of their machinery (see
B.P. 1.7).
By understanding this, the Vendors will be able to implement a proactive
design approach accounting for all of the potential issues they can run into in
the operation of their equipment.
L.L. 1.15: Not indicating to the Vendors the Plant exposure to unplanned
shutdowns can result in inadequate support
When the Vendors don’t realize the risks of unplanned shutdowns, they will
typically lean toward the idea of performing maintenance, when that may not
meet the objective of the End User in terms of revenue, safety, and environment.
Many RCFA’s we have been called in to perform were the result of issues
that were forgot about during the project and would have most likely been
caught during a cold eyes review, which if done during the VCM would have
resulted in minimal if any delays.
BENCHMARKS
The writer has been using this approach in the last 5 years when it has become
very evident that the vendors want to know what can expose their machine to
issues and better understand how the process system works.
34
Project Best Practices
Chapter | 1
TABLE 1.15.1 Specification Format Effects
The selected specification format can produce the following effects on equipment reliability and project schedule:
j Interpretation error—due to complexity of specs = reliability risk and delay
j Additional preparation time—for contractor
j Additional review time—for vendors
j Additional meeting and/or document exchange time for contractor/end user review
of vendor exceptions to specifications
This approach has resulted in improved vendor support and significant reduction in plant exposure.
SUPPORTING MATERIAL
At this point, the information required for preparing the project specifications
should be available. The challenge will now be to format these specifications to
assure complete compliance by all quoting parties and to minimize the schedule
time. The selected format of the specifications will significantly affect the project schedule in terms of preparation time, vendor response time, and contractor/
end user/vendor time to review specification exceptions for approval. The effects of the specification format on reliability and project schedule are shown
in Table 1.15.1.
Let’s face it; specifications have become very complex and extensive. It is
my experience that this is due to the fact that most end users send specification
drafts out to their plants and affiliates for review prior to publication and the
result is that every participant has to contribute or it does not look good! The end
product is a thick specification. Are specifications of this type really necessary?
Another consideration regarding complex specifications is that the experience
level in all phases of the industry is decreasing and the time required and accuracy of specification reviews is being affected.
Is there a viable alternative? Over the last 15 years, I have been involved
with a number of medium to small companies that are beginning to build new
facilities and have the flexibility to decide on specification format. Considering that their budget for specification preparation is limited, but they clearly
recognize the importance of a sound specification based on industry standards
and global best practices, they have generally adopted a strategy as shown in
Table 1.15.2.
The approach noted in Table 1.15.2 will only be possible if specialist creditability with project management has been secured and maintained. If there is a
reluctance to depart from the established specification format, a recommendation would be to contact other associates who you have met at industry conferences or search the web for consultants who have experience in this regard and
35
B.P. 1.15
More Best Practices for Rotating Equipment
TABLE 1.15.2 The Streamlined Specification Strategy
j Use an established industry specifications (e.g. API or ANSI)
j Include industry data sheets that are completed to outline all details of scope of
supply and company lessons learned
j Attach a best practice list applicable for the specific project only
j Use a global consultant firm to review and comment
j Note in the cover letter to the specification package that strict compliance with all
requirements is mandatory—options are not acceptable
TABLE 1.15.3 Streamlined Specification Format—Benefits
j End user—low probability of misinterpretation, shorter vendor bidding cycle time,
and reduced review time for vendor exceptions resulting in lower cost, reduced
project schedule, and manpower
j Contractor—less preparation and review time resulting in lower cost, reduced
project schedule, and manpower
j Vendor—less review and exception preparation time resulting in lower cost,
reduced project schedule, and manpower
can support your effort. Believe me, there are many benefits to this approach
for all parties (vendors, contractors, and end users). Benefits to the streamlined
specification approach are shown in Table 1.15.3.
The streamlined specification approach as shown in Table 1.15.1 is really
a win for all participants but for larger, established companies, it definitely is
a culture change and will have its price in terms of learning curve and initial
expended revenue.
Any new or different approach will meet initial resistance (a “paradigm
shift”) and will require specific instructions to the bidder in clear, concise and
brief terms. Table 1.15.4 presents some guidelines. It has been my experience
that the guidelines presented earlier for a streamlined approach, if endorsed by
the project management team and implemented as noted, can result in considerable schedule savings and increased equipment reliability.
TABLE 1.15.4 Important ITB Message to All Vendors
j The consideration given to your bid will be based on your compliance with the
requirement to list the exceptions applicable to this project only that are necessary
due to manufacturing, not cost constraints
j Include the added cost to comply with requirements in your bid and not as an
option
j Blanket exceptions to industry specifications (API, ANSI, etc.) are not acceptable
36
Project Best Practices
Chapter | 1
B.P. 1.16: Perform auxiliary system component selection review at 40%
engineering phase
By doing this you will assure that all of the components contained in the Oil
and Dry Gas Seal Systems are of top reliability, sized properly, and as simple in
design as possible.
L.L. 1.16: Failure to perform this review has resulted in delayed startups
and unplanned shutdowns
In the writer’s experience, at least 80% of all machinery issues can be traced
back to issues within the corresponding Systems. For this reason it is most important to assure all components within these systems are selected properly.
Forsthoffer Associates, Inc. has been involved with many RCFA’s on issues
within Auxiliary systems that were traced back to improper component sizing/
selection or system design.
BENCHMARKS
The writer has used the aforementioned approach in all Critical Equipment
Projects since the early 1990s.
This approach has resulted in Safety and Reliability issue free projects without significant cost adders and schedule delays.
SUPPORTING MATERIAL
The Coordination Meeting
After the order is placed, the coordination meeting is the first contact between
the contractor, end user hopefully, and the vendor. This meeting is usually held
approximately 4 weeks after the order placement and should be held at the vendor’s shop.
In my experience the effectiveness of this meeting is significantly increased
if the end user’s rotating equipment specialist and/or consultant, senior operator,
and maintenance engineer are in attendance. Depending on the project management team, it may be necessary to “campaign” for the attendance of these valuable people. There can be the suspicion that incorporating these individuals will
add additional cost to the job. It is my strong opinion that the addition of these
individuals will reduce significantly the life cycle cost of the job by incorporating lessons learned and best practices into the job. Refer back to the pre-FEED
phase of the job and note that the same individuals were asked to contribute
input to the job in this phase.
My “best practice” is to include a senior operator and maintenance man
(millwright) in all phases of the project from pre-FEED up to and including
the test phase to assure that all company “lessons learned” are turned into “best
practices” for the project. The key facts for the VCM (vendor coordination
meeting) are presented in Table 1.16.1.
37
B.P. 1.16
More Best Practices for Rotating Equipment
TABLE 1.16.1 The VCM—Key Facts
j Purpose—to confirm scope and design
j Design confirmation amount is proportional to the risk class
j If there is any component inexperience, details must be finalized now!
j Location—vendor’s shop
j Attendance by: vendor specialists, sub-supplier specialists, contractor specialists,
end user specialists
j Timing—approximately 6 weeks after order
j Duration—2–4 days based on complexity and risk class
TABLE 1.16.2 VCM Agenda—Key Facts
j Agenda by vendor approved by contractor/end user
j Agenda to be issued for review 2 weeks before meeting
j Assure that all required design reviews are included
j Inform project team in advance of required attendance
j End user should take detailed minutes
j Review all minutes and acceptance required by all parties, with action point responsibilities and required dates noted prior to adjournment
Vendor coordination meeting key facts are shown in Table 1.16.2.
A VCM checklist is included in Table 1.16.3 for your use to assure that all
important facts are covered. Depending upon the risk class of the equipment,
this may very well be the last chance for vendor engineering contact prior to the
shop test phase.
TABLE 1.16.3 VCM Agenda Checklist
j Review and confirm process conditions
j Review and confirm aero, thermo, and mechanical design
j Conduct any required design and/or manufacturing audits
j Confirm all major connection locations
j Review machine and auxiliary layouts for maintenance accessibility
j Review preliminary test agenda
j Resolve any outstanding specification issues
j Review vendor and sub-supplier QC procedures (there may be a separate meeting
for this activity)
38
Chapter 2
Pumps
B.P. 2.1: If an individual flow meter is not available, calculate flow using a
process control valve
Assuming that there is an individual process control valve for the pump, it is
very accurate to calculate the flow through a control valve. All that is needed is
a pressure upstream and downstream of the valve, the position of the valve, fluid
S.G. and the valve type and trim. A pressure gauge may need to be installed
downstream of the valve.
This calculation can be programmed into the DCS and a tag number created
for flow that can be used in order to trend pump performance.
L.L. 2.1: Not knowing and trending flow of critical and bad actor pumps
will result in lower pump component reliability
At least 80% of centrifugal pump component failures can be attributed process condition changes that affect the flow in the pump and many failures continue to occur because pump flow is not being trended.
Since many pumps do not have individual flow meters installed it can be
very difficult to know where they are operating on the performance curve.
BENCHMARKS
Since the 1990s this best practice has been used for Ammonia, Ethylene, LNG,
Methanol Plants, and Refineries. Not only is it successful for pumps without
flow meters but is also a good check on calibration of flow meters in plants.
Specifically, this has been used in the Alberta Tar Sands industry where many
pump applications are slurries and the Flow meters wear out very frequently.
When this Best Practice has been combined with monitoring and keeping a
pump within its Equipment Reliability Operating Envelope (EROE), pump MTBF’s can exceed 80 months.
More Best Practices for Rotating Equipment. http://dx.doi.org/10.1016/B978-0-12-809277-4.00002-4
Copyright © 2017 Elsevier Inc. All rights reserved.
39
B.P. 2.1
More Best Practices for Rotating Equipment
SUPPORTING MATERIAL
Control valve liquid sizing coefficient
Control valve liquid sizing coefficient (Cv) is an important concept that must be
understood when dealing with any type of control valve on liquid service. Cv
valve sizing coefficient is defined by the following equation:
CV = Q (GPM)
S.G.
∆P
Where S.G., specific gravity, 0.85 (for oil); ∆P, value pressure drop (P.S.I.).
Solving this equation for gallons per minute (GPM) we see that:
Q (GPM ) =
Cv
S .G.
∆P
We can see referring back to “the concept of an equivalent orifice” that this
equation is similar to that of an orifice. Naturally the only difference is that a
valve is a variable orifice. Valves are sized using this concept of Cv (valve coefficient). Each valve has a maximum Cv. Depending on the type of internal valve
design, seats, plugs, and body; a valve will exhibit a certain characteristic. Refer
to Fig. 2.1.1, which is a graph of valve characteristics. Plotted on the Y-axis is
valve flow in percent of maximum flow and plotted on the X-axis is travel of the
valve plug in percent of rated travel. As we cover specific valve application later
in this section, the characteristics of particular valves will be discussed. Referring back to the relationship for valve coefficient, it can be seen that the valve
coefficient is dependent on flow rate, differential pressure across the valve, and
fluid characteristic.
As an example, suppose that a valve is sized to pass 20 GPM under normal
conditions of 150 PSI pressure drop. The fluid in this case is light turbine oil at
150°F (60 SSU). Solving for the valve Cv per the above equation, we arrive at
a figure of 1.51. If the valve pressure drop were to decrease to 100 lbs., and we
still required 20 GPM to pass the valve coefficient would be 1.84. This change
represents approximately a 22% change in the valve coefficient. Depending
on the characteristic curve of the valve in question, it would represent a given
amount of valve plug opening (increase of travel). In the same example, now let
us assume that the flow changes to 40 GPM with 100 lbs pressure drop across
the valve. The Cv now would be 3.69 or approximately 200% the previous value. Depending on the valve size, this coefficient may or may not be obtainable.
Refer to Table 2.1.1 which is a typical valve coefficient table showing valve
coefficients for % travel of a particular valve. When sizing all control valves, Cv
maximum, Cv normal, and Cv minimum must be calculated. A general rule is
that all of the above values should fall between 10% and 90% of the maximum
Cv for a particular valve selected.
40
Chapter | 2
Pumps
FIGURE 2.1.1 Control valve flow characteristics. Courtesy of Fisher Controls Inc.
TABLE 2.1.1 Typical Liquid Valve Sizing Coefficient Table
Valve Cv’s for different sizes and valve travel
% Travel
12.5
25
50
75
100
Valve travel (in.)
Body size (in.)
Port size (in.)
1/32
1/16
1/8
3/16
1/4
1
3/4
1.4
3.1
4.2
5.3
7
1
1
2.4
4.2
7
10
12
When dealing with viscous liquids as in the case of oil, valve coefficient
viscosity corrections must be made. For the example case mentioned above,
the correction factor for 220 centistokes (cSt) (1000 Sabolt Universal Seconds
viscosity SSU) would be approximately 1.5–2. Therefore the valve coefficient
required would be 1.5–2 times that required at normal viscosities (60 SSU for
light turbine oil at normal operating temperatures). Viscosity correction nomographs are available from control valve manufacturers for determining valve
sizes required under high viscosity conditions.
41
B.P. 2.2
More Best Practices for Rotating Equipment
A flow meter in every system
Considering the relationship discussed earlier it can be seen that every control
valve can be considered as a flow meter if the fluid differential pressure across
the valve, valve travel and, a valve characteristic chart is known. While not a
completely accurate flow measuring device, this concept can be extremely valuable while troubleshooting auxiliary systems. Obtaining the valve travel and
using the valve coefficient chart, the Cv can be obtained. Calculating for GPM
knowing the Cv, the pressure drop across the valve and the specific gravity of the
liquid can then yield the flow rate. It is important to note that with small valve travels
on the order of 1/4 in. maximum, an accurate means of measuring valve travel must
be obtained. It is the writer’s experience that many times travel indicators are not
furnished with the valve. It is strongly recommended that valve travel indicators be
supplied or retrofitted in the field.
B.P. 2.2: Assure that all critical pumps are installed with an individual
flow meter
It is very important to know where each pump is operating in relation to
its Best Efficiency Point (BEP) in order to assure that neither low flow (ReCirculation) nor high velocity cavitation is occurring within the pump. The only
way to accurately know this is by having a flow meter installed in the discharge
of each pump, not in the discharge header of the service.
This flow meter must be upstream of the minimum flow bypass line in order
to give the flow actually going through the pump.
This Best Practice is most useful in services where pumps are operated in parallel and it is difficult to know the exact condition of each individual pump. One pump
operating at a lower flow than the other indicates that pump is in worse condition,
since a worn pump will produce less head than the pumps in good condition. If a
pump is producing less head than the other pumps, it will naturally produce less
flow when run in parallel operation in order to match the head of the other pumps.
L.L. 2.2: The Inability to accurately know the flow through each pump can
result in unnecessary maintenance and risk of lost production
Without individual flow readings through pumps operating in parallel, it
may be difficult to determine if a pump needs maintenance.
If the service is critical (i.e. Boiler Feed or a unit Charge Pump) and a pump
is taken out of service for maintenance, there is no longer a spare. Therefore, it
is essential that the pump with deficiencies is worked on or a reduction in rates
could occur.
BENCHMARKS
This Best Practice has been successfully used since the mid-1990s during the
early stages of the project phase for critical pumps in parallel operation. It has
resulted in high reliability (greater than 80 months MTBF) in critical pumps
42
Pumps
Chapter | 2
since the operating point is always known and an accurate decision can be made
as to which pump is wearing more internally.
SUPPORTING MATERIAL
Effects of the process on pump reliability and MTBF
The effect of the process on machinery reliability is often neglected as a root
cause of machinery failure. It is a fact that process condition changes can cause
damage and/or failure to every major machinery component. For this discussion, the most common type of Driven Equipment—Pumps will be used.
There are two major classifications of pumps, positive displacement and kinetic, centrifugal types being the most common. A positive displacement pump
is shown in Fig. 2.2.1. A centrifugal pump is shown in Fig. 2.2.2
It is most important to remember that all driven equipment (pumps, compressors, fans, etc.) react to the process system requirements. They do only what
the process requires. This fact is noted in Table 2.2.1 for pumps.
Centrifugal (Kinetic) Pumps and their drivers
Centrifugal pumps increase the pressure of the liquid by using rotating blades to
increase the velocity of a liquid and then reduce the velocity of the liquid in the
volute. Refer again to Fig. 2.2.2.
FIGURE 2.2.1 Positive displacement plunger pump.
43
B.P. 2.2
More Best Practices for Rotating Equipment
TABLE 2.2.1 Pump Performance
j
j
Pumps produce the pressure required by the process
The flow rate for the required pressure is dependent on the pump’s characteristic
FIGURE 2.2.2 Centrifugal pump.
A good analogy to this procedure is a football (soccer) game. When the ball
(liquid molecule) is kicked, the leg (vane) increases its velocity. When the goaltender (volute), hopefully, catches the ball, its velocity is significantly reduced
and the pressure in the ball (molecule) is increased. If an instant replay “freeze
shot” picture is taken of the ball at this instant, the volume of the ball is reduced
and the pressure is increased.
The characteristics of any centrifugal pump then are significantly different
from positive displacement pumps and are noted in Table 2.2.2.
Refer again to Table 2.2.1 and note that all pumps react to the process requirements.
Based on the characteristics of centrifugal pumps noted in Table 2.2.2, the
flow rate of all types of centrifugal pumps is affected by the Process System.
This fact is shown in Fig. 2.2.3.
Therefore, the flow rate of any centrifugal pump is affected by the process
system. A typical process system with a centrifugal pump installed is shown in
Fig. 2.2.4.
The differential pressure required (proportional to head) by any process system is the result of the pressure and liquid level in the suction and discharge
44
Pumps
Chapter | 2
TABLE 2.2.2 Centrifugal Pump Characteristics
j
j
j
j
Variable flow
Fixed differential pressure produced for a specific flowa
Does not require a pressure limiting device
Flow varies with differential pressure (P1 − P2) and/or specific gravity
a
Assuring specific gravity is constant.
TABLE 2.2.3 Centrifugal Pump Reliability
j
j
j
j
Is affected by process system changes (system resistance and S.G.)
It is not affected by the operators!
Increased differential pressure (P2 − P1) means reduced flow rate
Decreased differential pressure (P2 − P1) means increased flow rate
FIGURE 2.2.3 A centrifugal pump in a process system.
vessel and the system resistance (pressure drop) in the suction and discharge
piping.
Therefore, the differential pressure required by the process can be changed
by adjusting a control valve in the discharge line. Any of the following process
variables (P.V.) shown in Fig. 2.2.4, can be controlled:
j
j
j
level
pressure
flow
As shown in Fig. 2.2.3, changing the head required by the process (differential pressure divided by specific gravity), will change the flow rate of any
centrifugal pump!
45
B.P. 2.2
More Best Practices for Rotating Equipment
FIGURE 2.2.4 Centrifugal pump control options.
Refer to Fig. 2.2.5 and it can be observed that all types of mechanical failures can occur based on where the pump is operating based on the process
requirements.
Since greater than 95% of the pumps used in this refinery are centrifugal,
their operating flow will be affected by the process. Please refer to Table 2.2.3,
which shows centrifugal pump reliability and flow rate is affected by process
system changes.
At this point it should be easy to see how we can condition monitor the centrifugal pump operating point. Refer to Table 2.2.4.
Driver reliability (motors, steam turbine, and diesel engines) can also be affected by the process when centrifugal driven equipment (pumps, compressor,
and fans) are used.
Refer to Fig. 2.2.6 and observe a typical centrifugal pump curve.
Since the flow rate will be determined by the process requirements, the power (BHP) required by the driver will also be affected. What would occur if an
8½ in. diameter impeller were used and the head (differential pressure) required
by the process was low? Answer: Since the pressure differential required is low,
the flow rate will increase and for the 8½ in. diameter impeller, the power required by the driver (BHP) will increase.
Therefore, a motor can trip out on overload, a steam turbine’s speed can
reduce or a diesel engine can trip on high engine temperature. These facts are
shown in Table 2.2.5.
Auxiliary System Reliability is also affected by process changes. Auxiliary
systems support the equipment and their components by providing ... clean,
46
Pumps
FIGURE 2.2.5
point.
Chapter | 2
Centrifugal pump component damage and causes as a function of operating
TABLE 2.2.4 Centrifugal Pump Practical Condition Monitoring
j
j
Monitor flow and check with reliability unit (RERU) for Significant changes
Flow can also be monitored by:
j Control valve position
j Differential Temperature across pump
j Motor amps
j Steam turbine valve position
cool fluid to the components at the correct differential pressure, temperature,
and flow rate.
Typical auxiliary systems are:
j
j
j
j
Lube Oil Systems
Seal Flush System
Seal Steam Quench System
Cooling Water System
The reliability of machinery components (bearings, seals, etc.) is directly
related to the reliability of the auxiliary system. In many cases, the root cause of
the component failure is found in the supporting auxiliary system.
47
B.P. 2.2
More Best Practices for Rotating Equipment
FIGURE 2.2.6 A typical centrifugal pump performance curve.
TABLE 2.2.5 Effect of the Process on Drivers
j
j
j
Motors can trip on overload
Steam turbines can reduce speed
Diesel engines can trip on high engine temperature
TABLE 2.2.6 Effect of the Process on Drivers
j
j
j
Is directly related to auxiliary system reliability
Auxiliary system reliability is affected by process condition changes
“Root causes” of component failure are often found in the auxiliary system
TABLE 2.2.7 Always “Think System”
j
j
Monitor auxiliary system condition
Inspect auxiliary system during component replacement
As an example, changes in auxiliary system supply temperature, resulting
from cooling water temperature or ambient air temperature changes, can be the
root cause of component failure. Table 2.2.6 presents these facts.
As a result, the condition of all the auxiliary systems supporting a piece of
equipment must be monitored. Please refer to Table 2.2.7.
48
Pumps
Chapter | 2
EQUIPMENT RELIABILITY OPERATING ENVELOPE
DETERMINATION
As noted in Table 2.2.8, process changes will vary the flow of any centrifugal pump.
If the centrifugal pump flow is too high or too low hydraulic disturbances
will be present that can change the pumped fluid pressure and/or temperature.
Since the majority of Mechanical Seal applications use the pumped fluid in the
seal chamber, the seal chamber pressure and/or temperature will be affected.
These changes will directly impact Mechanical Seal Life and Reliability.
Fig. 2.2.7 shows a typical centrifugal pump head vs. flow curve with the
following items noted:
j
the “Desirable Region” of Operation—Heart of the Curve or Equipment Reliability Operating Envelope (EROE)
TABLE 2.2.8 Process Effects on Centrifugal Pump Flow
Decreased Pump Flow:
j Increased P2
j Decreased P1
j Decreased S.G.
Increased Pump Flow:
j Decreased P2
j Increased P1
j Increased S.G.
FIGURE 2.2.7 Centrifugal pump head versus flow curve.
49
B.P. 2.2
More Best Practices for Rotating Equipment
TABLE 2.2.9 EROE Facts
1. The EROE flow range is +10% and −50% of the pump best efficiency point (BEP)
flow
2. All “bad actor pumps”—(more than one component failure per year) should be
checked for EROE
3. To determine that the pump is operating in EROE:
j Calculate the pump head required
j Measure the flow
j Plot the intersection of head and flow on the pump shop test curve
regions of Hydraulic Disturbances—on the upper portion of the curve
the Pump Components affected—on the lower portion of the curve
j
j
The “Heart of the Curve” is the flow region for any centrifugal pump that
will be free of Hydraulic Disturbances and where the seal fluid should be free
of vapor if the seal fluid conditions stated on the Pump and Seal Data Sheets are
present during pump field operation.
This Flow Region is also called the:
EROE—The Equipment Reliability Operating Envelope
Table 2.2.9 presents facts concerning the EROE.
In many pump installations, a flow meter is not installed and a suction pressure gauge is not installed. A calibrated suction pressure gauge can be installed
in the suction pipe drain connection (always present). Be sure to obtain a Management of Change (MOC) and Work Permit and any other plant required
permission prior to installing a suction pressure gauge as the pumped fluid
could be sour (H2S), flammable and/or carcinogenic.
If a flow meter is not installed, Table 2.2.10 defines the options available to
determine the pump flow so the EROE can be obtained.
The flow values in Table 2.2.10 can be determined by hand calculations using the equations available in any pump text (Power Equation and Pump Temperature Rise Equation).
It can be seen that the EROE will provide a reasonable guide that usually
will eliminate Hydraulic disturbances that can cause seal chamber pressures and
TABLE 2.2.10 Available Pump Flow Determination Options
1. Record control valve position, valve differential pressure, fluid S.G., and calculate
valve flow (pump flow)
2. Measure motor amps and calculate power
3. Measure pump pipe differential temperature and calculate efficiency
4. Obtain an ultrasonic flowmeter to measure flow
5. For items 2 and 3, locate the calculated value (power or efficiency) on the pump test
curve to determine pump flow
50
Pumps
Chapter | 2
TABLE 2.2.11 Factors That can Reduce Low Flow EROE Range
j
j
j
j
j
Pumps with suction specific speeds > 8000 (customary units)
Double suction pumps
Water pumps with low NPSH margin
Fluids with S.G. < 0.7
Pumps with Inducers
TABLE 2.2.12 If a Centrifugal Pump is Outside its EROE
j
j
Consult operations to determine if process changes can be made to operate in EROE
Define target EROE parameters for operations (flow, amps, control valve position,
delta T)
temperatures to change and lead to premature seal wear and/or failure. Note that
the stated EROE low flow range can be reduced if the pump or fluid have any of
the following characteristics noted in Table 2.2.11.
Therefore, we always recommend that the first step in seal condition monitoring be determination of pump operation within its EROE. If the “Bad Actor” Pump is operating outside its EROE, we recommend the action shown in
Table 2.2.12.
If seal reliability does not improve when operating within the EROE, further
investigation is required concerning the process conditions in the seal chamber
and/or flush system.
Pump parallel operation
One of the most common mistakes made in the field when checking pump performance of pumps operating in parallel, is that the total flow measured in the
discharge header is split in half by each of the two pumps in operation (assuming two pumps are running). This is only true if the pumps are both in the exact
same conditions (Fig. 2.2.8).
As you can see in Fig. 2.2.8, there can be assumed to be one performance
curve because both pumps are in the same condition. Therefore, for given process requirements (Head) the flow produced by both could potentially be double, if the System curve allows (very flat curve).
However, more times than not pumps will not wear at the same rate, even
if they operate for the same time in the same service (Fig. 2.2.9). This figure
shows the fact that one pump may be operating and putting out all of the flow
while the worn pump is running in shutoff! This is why it is very important to
have a measurement or at least a way to know if a pump is in better condition
than the other.
51
B.P. 2.2
More Best Practices for Rotating Equipment
FIGURE 2.2.8 Two identical pumps in parallel operation.
FIGURE 2.2.9 Non-identical pumps operating in parallel.
52
Pumps
Chapter | 2
B.P. 2.3: Use Pipe Differential Temperature to determine whether or not a
pump is operating in its “EROE”
We know by now that every Centrifugal Pump has a preferred operating
range for flow that we call EROE. This is great if we have an accurate way of
measuring flow (i.e., Flow meter or a control valve). However if we can’t accurately measure flow, then we can at least estimate where the pump is operating
on its curve.
When a pump operates at a flow below the low flow area of its EROE, you
will see that the efficiency rapidly drops off due to the recirculation occurring.
From this recirculation of the fluid, you will definitely see heat being transferred
to the pumped fluid.
A single stage centrifugal pump operating within its EROE will see a differential temperature definitely less than 5°C. Therefore, anything greater than
that value would indicate the pump is operating at a lower efficiency and most
likely outside of its EROE.
A multistage pump will have a higher normal temperature than a single stage
pump, so make sure to get a baseline pipe differential temperature when the
pump is known to be in good condition.
It is very important to check the temperatures just upstream of the suction
flange and downstream of the discharge flange. If you measure the temperatures
closer to the pump, you may see erroneous values. Also, make sure to mark the
spots you take the temperatures as the pipe temperature will vary at different
spots circumferentially around.
L.L. 2.3: The inability to know if a pump is operating at low flow can result
in wear and/or component failure
Have you ever seen a pump in the shop where the impeller wear rings are
essentially welded to the casing wear rings?
This is absolute evidence that the pump operated at low flow for an extended
period of time and was not noticed in the field.
The pump head required by the process is a function of Inlet Pressure, Discharge Pressure, and Specific Gravity. If any of these values change to increase
the head required, the flow will decrease. This will result in lower efficiency and
higher temperatures internally if the pump is operating outside of its low flow
EROE (typically less than 50% of BEP Flow).
BENCHMARKS
This Best Practice has been implemented in all types of plants since the late
1990s and has been successful in early detection of pump issues, allowing a
component failure to be prevented. By following this Best Practice, plants have
been able to increase Pump MTBF above 80 months and approach 100 months.
It is highly recommended for this BP to be included in the operations daily
rounds, so they can notify the personnel in charge of machinery when a deviation
53
B.P. 2.4
More Best Practices for Rotating Equipment
occurs. Operators are the front line of defense and will see change in operating
conditions before any other discipline in the plant.
SUPPORTING MATERIAL
See B.P. 2.2 for Supporting Material.
B.P. 2.4: Confirm NPSH Available in the field for bad actor pumps
Many times pumps operate in a condition that appears to be cavitation, but it
is not defined as being the classical (high velocity) cavitation or recirculation
(low velocity cavitation). One way to determine this is by confirming what the
NPSH Available is in the field and if it is sufficient to meet the requirement of
the pump. Unfortunately, it is difficult to accurately model the piping and system upstream of the pump in order to calculate the NPSH available in the field.
The easiest, most accurate way is to use a calibrated suction pressure gauge
as close as possible to the suction nozzle of the pump and measure P1.
Then use the following equation:
NPSHA(m) =
10.2 × ( P1 − Pv )
S.G.
Where NPSHA, meters; P1 is suction pressure in bara; Pv is the fluid vapor
pressure in bara; S.G. is the fluid specific gravity.
If pressure is in kPa, the constant is 0.102
If pressure is in kg/cm2, the constant is 10.01
If NPSHA is in ft. and pressure in Psia, the constant is 2.311.
L.L. 2.4: Failure to identify the cause of fluid vaporization within the pump
will most likely result in multiple failures and increased maintenance costs
or loss of production
BENCHMARKS
This BP has been followed since the mid-1990s for all bad actor pumps in order
to assure pumps operate at the highest reliability. It has played a part in maintaining pumps with a MTBF in excess of 80 months.
SUPPORTING MATERIAL
Start with a data sheet to completely define requirements
One of the single most important factors in selecting a pump to meet the requirements of a process system is to completely and accurately state all the
requirements on a data sheet. A centrifugal pump data sheet, courtesy of the
American Petroleum Industry (API 610) is supplied at the end of this Supporting Material.
54
Pumps
Chapter | 2
TABLE 2.4.1 Minimum Data Sheet Requirements
(P) (U)
j
Pump application and operating mode (single or parallel)
(P) (U)
j
Detailed operating conditions
(P) (U)
j
Accurate site and utility requirements
(M)
j
Pump performance
(P) (M)
j
Pump construction and experience
(P)
j
Spare parts required
(P) (M)
j
Driver details
(P)
j
QA inspection and test requirements
Note: to be completed by: P, purchaser; M, manufacturer (vendor); U, user.
Regardless of the source, all pump data sheets should contain the categories
of information shown in Table 2.4.1.
Completely define the operating conditions
Correctly stated operating conditions are essential for proper definition and
subsequent selection of a specific type and configuration of pump to meet the
specified conditions.
Once it is decided to install a pumping system, a sketch should be drawn to
define all of the components, which are required to be included into that system.
Some of the factors, which need to be considered in completing the sketch and
system design include the following:
Flow rate—All flow rates including minimum, normal, and rated should be
listed in the data sheet. Normal flow is usually the flow required to achieve a
specific process operation. The rated flow is normally a set percentage increase
over the normal flow and it usually includes consideration for pump wear and
the type of operation within the process system. It can amount to as much as
10% depending upon specific company practice. The minimum flow is important to identify in order to establish if a minimum flow bypass line is required
for process or mechanical design considerations.
Head required—The required head that the pump must develop is based
on the static pressure difference between the discharge terminal point and the
suction source, the elevation difference and the friction losses through system
components including suction and discharge side piping, pressure drop through
heat exchangers, furnaces, control valves and other equipment. It is represented
by the equation in Table 2.4.2.
Liquid properties—Viscosity, vapor pressure, and specific gravity each
play an important role in achieving the required level of pump reliability within
the operating system. Viscosity can impact pump performance to the extent that
55
B.P. 2.4
More Best Practices for Rotating Equipment
TABLE 2.4.2 Required Head Equation
10.2 × ∆P (at pump flanges)
S.G.
∆P = Total pressure difference between the discharge system and suction system,
measured at the pump flanges in barg
S.G. = Specific gravity of the liquid at pump temperature
H = Pump required head in meter-kg force/kg mass
Notes: If pressure is measured in kPa, constant = 0.102
If pressure is measured in kg/cm2, constant = 10.003
If pressure is measured in PSIG, constant = 2.311 and head required is measured
in ft-lb force/lb. mass
H=
it may not be justified to even use a centrifugal pump when the viscosity values
are greater than 7.5 cSt (50 SSU). The hydraulic institute has published curves,
which can be used to calculate the performance effects resulting from pumping
viscous liquid.
Vapor pressure and specific gravity influence the type of pump to use and
its mechanical design configuration. Vapor pressure is an important property
when determining whether there is adequate net energy available at the pump
suction to avoid vaporization of the liquid, which can lead to performance deterioration and possible shortened life expectancy of the pump.
Specific gravity is the liquid property used to calculate the amount of head
a pump has to produce to overcome the resistance of the suction and discharge
systems. It is also used as a guideline to determine whether a pump casing design should be of the vertical (radial) split or horizontal (axial) configuration
(refer to Table 2.4.3 for some guidelines).
NPSH available—Net positive suction head available is a characteristic of
the process suction system. It is the energy above the vapor pressure of the liquid,
measured at the suction flange of the pump, which is required to maintain the
fluid in a liquid state. In a centrifugal pump it is usually measured in feet of liquid
(refer to Fig. 2.4.1 for a typical method for calculating NPSH available. It is important to note that the pressure at the suction source cannot be considered equal
to the NPSH. In Fig. 2.4.1 it can be seen that the source pressure is the same as
the vapor pressure, indicating that the liquid is at its boiling point. When the vapor pressure is subtracted from the suction pressure the resulting NPSH available
TABLE 2.4.3 Casing Configuration Guidelines
Use radial split casing for:
j S.G. ≤ 0.7 at pumping temperature
j Pumping temperature ≥200°C (400°F)
j Flammable or toxic liquids at rated discharge pressures above 6896 kPa (1000 psig)
56
Pumps
Chapter | 2
FIGURE 2.4.1 Calculate available NPSH.
is 2.1 psi or 10 ft. When calculating NPSH available it is prudent to incorporate a
margin of safety to protect the pump from potential cavitation damage resulting
from unexpected upsets. The actual margin amount will vary from company to
company. Some will use the normal liquid level as the datum, while others use
the vessel tangent or the bottom of the vessel. Typical suggested margins are:
2 ft. for hydrocarbon liquids (including low S.G.), and 10 ft. for boiling water.
NPSH required
The NPSH required is the amount of energy required to keep the fluid in its
liquid state all the way to the beginning of the impeller vanes. Therefore, after
all of the pressure drop that occurs in the suction piping and components (especially in the nozzle and pump impeller eye) upstream of the pump impeller, we
want the pressure to still be above the fluid vapor pressure. See Fig. 2.4.2 for a
good depiction of this concept.
Defining the pump rated point for efficient operation
Since centrifugal pumps are not normally custom designed items of equipment,
it is important to assure that each vendor will quote similar pump configurations
for the specific operating conditions set forth on each application data sheet.
When establishing which pump characteristic and impeller pattern to select for
a specific application, certain guidelines should be followed (Table 2.4.4).
57
B.P. 2.4
More Best Practices for Rotating Equipment
FIGURE 2.4.2 NPSHR.
TABLE 2.4.4 Application Guidelines
j
j
58
When selecting a specific impeller pattern the rated flow should be no greater than
10% to the right of best efficiency point. This will result in operation at close to best
efficiency point during normal operation (refer to Fig. 2.4.3). Also, selecting a pump
to operate too far to the right of best efficiency point can result in the pump operating in the “break.” A pump is considered operating the “break” when it is pumping
maximum capacity and the total head is reduced while the suction head is held (the
impeller actually acts as an orifice to limit the flow).
Selecting a pump for the rated flow too far to the left of best efficiency point (oversized pump) can result in cavitation damage caused by internal recirculation (refer
to Fig. 2.4.4).
Pumps
Chapter | 2
FIGURE 2.4.3 Selecting a specific impeller pattern.
FIGURE 2.4.4 Suction recirculation flow pattern.
Carefully define critical component requirements
Pump reliability improvement can be achieved through proper specification, selection, and operation of components such as bearings, mechanical seals, and
drivers. Industry standards such as API Standard 610 for centrifugal pumps and
Standard 682 for mechanical seals contain minimum requirements, which if implemented, should result in improved reliability and extended on stream operating time. Some salient points about each of these components are highlighted
in Tables 2.4.5–2.4.7.
Guidelines to use when selecting pump style
The choice for selecting the type of pump to use for a given application can
vary with specific gravity, operating temperature, pressure conditions, liquid
composition, and available NPSH. Tables 2.4.8–2.4.10, 2.4.11, and 2.4.12 provide guidelines, which can help make the choice or selecting a style of pump
less complicated.
59
B.P. 2.4
More Best Practices for Rotating Equipment
TABLE 2.4.5 Bearing Application Guidelines
j
j
j
j
j
j
j
Centrifugal pumps require bearings to carry radial/axial loads
Bearing alternatives: anti-friction, hydrodynamic ring oil lubricated, or hydrodynamic pressure lubricated
Oil lubricated (anti-friction) bearings are used in majority of process pumps to carry
loads
Pressure lubricated hydrodynamic bearings are normally used for high pressure,
high horsepower, high speed applications
Criteria in API Standard 610 for pressure lubrication: when product of pump rated
horsepower and rated speed in revolutions per minute is greater than 2.0 million
Pressure lubrication systems can either be integral or separate, but should include,
as a minimum, an oil pump, reservoir, filter, cooler, controls, and instrumentation
Ring oil lubrication may be applied to hydrodynamic journal bearings in less severe
service [when dN factor is less than 300,000. A dN factor is the product of bearing,
size (bore) in millimeters and the rated speed in revolutions per minute].
TABLE 2.4.6 Mechanical Seal Application Guidelines
j
j
j
j
j
j
j
j
j
j
Mechanical seals are often used in pumps handling hazardous as well as nonhazardous liquids that must be contained within the unit
Single seal arrangement is most widely used in process industry
Single seal design consists of rotary face in contact with a stationary face
For most services, a carbon face mating against tungsten carbide is satisfactory
Seals offer the advantages of long life, low maintenance and high reliability
In general, seals handling light specific gravity liquids at low temperature and high
vapor pressure give most problems in the field
Materials for cold service seals must be suitable for temperatures of startup, cool
down and running; the atmospheric side must be held above 0°C (32°F) to prevent
ice formation; and there must be enough liquid at the seal surfaces
Successful operation of any seal depends largely on correctly specifying liquid conditions of vapor pressure, temperature specific gravity, etc.
API Standard 682 is excellent resource for overall mechanical seal application guideline
The pressure in the seal chamber (stuffing box) must be at least 25 psig above the
pump suction pressure.
Using the guidelines presented, let us now focus on three examples of how
to select a centrifugal pump for a given process system application (Fig. 2.4.5).
Before the appropriate pump and driver can be selected, it will be necessary
to completely define the process system operating conditions in which the pump
will operate. This will include the suction and discharge system resistance, the
head (energy) required by the system and the NPSH available (Table 2.4.13).
60
Pumps
Chapter | 2
TABLE 2.4.7 Driver Sizing Guidelines
j
j
j
j
j
j
j
Pump drivers are normally electric motor or steam turbine
Choice of driver type is usually based on plant utility balance plus reliability
evaluation of each type to perform within the operating system
Motors can be sized by several methods:
Name plate rating large enough to cover the complete range of pump performance
curve
Size of motor based on system curve analysis to establish maximum horsepower
required at intersection of system curve and pump performance curve
API Standard 610 has guidelines for sizing motor drives. A margin of 125%
is recommended for motors equal to or less than 18.6 kW (25 hp), 115% for
22.4–56 kW (30–75 hp) and 110% for motors rated 75 kW (100 hp) or more
Steam turbine drives are normally sized for the power required at pump rated
condition. This is possible because turbines can accommodate increased power
loads more readily than electric motors.
TABLE 2.4.8 Single Stage, Single Suction Overhung Impeller Characteristics
j
j
j
j
j
j
j
Most commonly applied centrifugal pump—most applications
Total head limited to 380 mm (15 in.) impeller diameter at 3600 rpm [approximately
183 m (600 ft.) head]. Larger diameter impellers operate at lower speeds
Low, medium, high temperature (with cooled bearings, stuffing box)
Relatively low NPSH required for single suction impeller
All process services with proper materials selection
Center of gravity of impeller is outside bearing span
Axial thrust
TABLE 2.4.9 Single Stage In-Line Pump Characteristics
j
j
j
j
j
j
j
Gaining acceptance as alternative to single stage overhung pump
Total head limited to approximately 122 m (400 ft.)
Low temperature applications only
Relatively low NPSH required
Limited to approximately 150 kW (200 H.P.)
Most designs utilized do not incorporate bearings (they use the motor bearings to
position the pump shaft)
Note: Designs are now available that incorporate an antifriction bearing in the
pump housing
61
B.P. 2.4
More Best Practices for Rotating Equipment
TABLE 2.4.10 Single Stage Integral Gear Centrifugal Pump Characteristics
j
j
j
j
j
j
j
Used for high head low flow applications
Total maximum head approximately 762 m (2,500 ft.)
Can be used for all temperatures
Lowest NPSHrequired (can use inducer)
Limited to 300 kW (400 bhp) maximum
All process services with proper material selection
Incorporates gear box to increase pump speeds as high as 30,000 rpm
TABLE 2.4.11 Horizontal Multistage Between Bearing Pump Characteristics
j
j
j
j
j
Used for high head medium flow applications
Double suction impeller for first stage for low NPSHR
Low, medium, high temperature (with cooled bearings, stuffing box)
No speed constraint
Thrust requires compensation (back to back impellers, balance device)
TABLE 2.4.12 Vertical Multistage Pump Characteristics
j
j
j
j
j
j
Used for low NPSH available applications
High head capability by adding stages
Low, medium, high temperature
Low, medium flow range
No speed constraint
Most nonabrasive process liquids with proper materials selection
FIGURE 2.4.5 Example No. 1 process system.
62
Pumps
Chapter | 2
TABLE 2.4.13 Calculate Process System Variables
Elevations above grade:
Discharge pressure calculation:
Pump centerline
3 ft.
Vessel pressure
Inlet nozzle to vessel
72 ft.
Static elevation head
Liquid surface in
suction vessel
(72−3) × 0.433 × 0.488
310 psig
= 14.5 psi
Max 32 ft.
Friction ∆P:
Min 22 ft.
Piping
10 psi
Orifice
2 psi
Control valve
30 psi
Pressure P, in bottom
of suction vessel
237 psig
Pressure drops:
Exchanger
15 psi
Suction piping
1 psi
Pd (Vesselpress + Elevpress + losses)
= 381.5 psi
Discharge piping
10 psi
∆P
Flow orifice
2 psi
381.5 – 240
141.5 psi
Control valve
30 psi
Exchanger
15 psi
Flow rate:
500 gpm
Specific gravity
0.488 at p.t.
Vapor pressure
251 psia
∆P × 2.31 141.5 × 2.31
=
S.G.
0.488
Suction pressure calculation:
NPSHAVAILABLE (for boiling liquid)
Vessel pressure
Pa surface + 14.7 – Pv +
237 psig
Static elevation head
Pd – Ps
= 670 ft.
Static elev diff – friction
(22 − 3) × 0.433 ×
0.488
= 4 psi
Suction line ∆P
–1 psi
Ps (Vesselpress +
Elevpress − loss)
= 240 psi
237 + 14.7 – 251 + 4 – 1 = 3.7 psi
3.7 × 2.31
NPSHAVAILABLE =
= 17.5ft.
0.486
When the process system is defined, the next step is to complete the tasks
presented in Table 2.4.14.
Based on an assessment of the process system requirements in Table 2.4.13
and the guidelines for selecting pump and driver we can determine that the
pump defined in Table 2.4.14 satisfies all of the guidelines.
For example No. 2 let us select a pump for a boiler feed water application
with operating conditions shown in Fig. 2.4.6.
63
B.P. 2.4
More Best Practices for Rotating Equipment
TABLE 2.4.14 Tasks for Selecting Pump and Driver
j
j
j
j
j
j
Select pump type based on guidelines:
Single stage overhung impeller
Multi stage axial or radial split casing design
Match NPSHR versus NPSHA
Calculate bhp based on pump efficiency
Determine driver hp rating based on API criteria
FIGURE 2.4.6 Example No. 2 pump selection. Courtesy of Union Pump Co.
This application requires a multistage axial split case pump based on the
criteria that the head (energy) required by the system exceeds the head (energy),
which can be provided by a 15 in. single-stage impeller (Table 2.4.15).
The pump selected is a Union Pump 3 × 4 MOC, five-stage axial split casing unit. Note that selecting a 9.50-in. diameter will result in the pump operating
at its best efficiency point (bep) at rated flow.
64
Pumps
Chapter | 2
TABLE 2.4.15 Example No. 2 operating conditions
Liquid
Boiler feed water
S.G.
0.93
P.T.
220°F
Ps
25 psig
Pd
650 psig
NPSAavailable
26 ft.
Flow rate rated
275 gpm
headrequired
1553 ft.
TABLE 2.4.16 Example No. 3 Operating Conditions
Flow rate
100 gpm
S.G.
0.98
P.T.
120°F
NPSHA
0 ft. at grade
Ps
1.96 psig
Pd
120 psig
Head
3.13
For our third example, we shall examine the selection of a pump type with a
constraint on NPSH available. A hot well condensate pump installed in a steam
turbine condenser system will be used to illustrate this example (Table 2.4.16)
for operating conditions (Fig. 2.4.7).
It is apparent that the NPSH available is a major constraint for selecting a
conventional horizontal pump for pumping condensate from the condenser hot
well. For this application, a vertical canned pump is the appropriate selection
(Fig. 2.4.8).
The feature about this design, which makes it suitable for use in this type
of service is the fact that the first stage impeller is located at the bottom end of
the shaft and the shaft length can be made sufficiently long to satisfy the NPSH
required by the pump. It is common practice to reference available NPSH to
grade elevation for this type of pump design. This allows for variations in design
of concrete foundation height and location of suction nozzle centerline from top
of foundation (Figs. 2.4.9 and 2.4.10).
65
B.P. 2.4
More Best Practices for Rotating Equipment
FIGURE 2.4.7 Typical centrifugal pump performance curve. Courtesy of Union Pump Co.
FIGURE 2.4.8 Multistage centrifugal pump. Courtesy of Union Pump Co.
66
Pumps
Chapter | 2
FIGURE 2.4.9 Application of vertical pump in condensate hot well service.
FIGURE 2.4.10 Example No. 3 NPSH reference.
67
B.P. 2.5
More Best Practices for Rotating Equipment
B.P. 2.5: How to determine EROE boundaries when unsure
EROE is defined as the safe operating range for a centrifugal pump where
it can operate without significant component wear. Typically, this range is from
−50% to +10% of the BEP for the pump, however it can change due to different factors.
The best way to assure safe operation is to operate the pump in a region of
the performance curve where there is a significant slope and stay away from
the flat region of the curve. In the flat region of a performance curve, a small
change in head required by the process will result in a large change in flow. A
flat region can be defined as an area of the curve that has less than 4% head rise
to shutoff (zero flow).
L.L. 2.5: Failure to operate within the EROE will result in component
wear and failures
Many times, pumps can operate it what is stated to be a safe operating range,
but the characteristic of the pump is that it has a very flat curve.
In those cases it is important to stay away from the lower flow regions,
which will result in a great change in flow for a small change in P1, P2, or S.G.
BENCHMARKS
This has been used successfully in all plants since the late 1990s and has contributed in maintaining pump MTBF’s above 80 months.
SUPPORTING MATERIAL
Fig. 2.5.1 shows a sketch of a typical pump performance curve and where it is
safe to operate. The red region is defined as any flow point on the curve where
you have 4% or less head rise to the shutoff head or zero flow region.
FIGURE 2.5.1 Safe operating area for a pump.
68
Pumps
Chapter | 2
B.P. 2.6: Accurately define Suction Specific Speed for pumps with Double
Suction Impellers and create new boundaries of EROE
The higher value of suction specific speed will result in lower velocity of the
pumped fluid entering the pump and therefore could allow for recirculation at
flows above −50% of the BEP.
For Imperial units (SI units for Suction specific speed result in a value about
2,000 higher than Imperial units), any Suction specific speed value above 8,000
will result in a narrower EROE Range, and an impeller with a suction specific
speed at about 16,000 can experience recirculation very close to the BEP of the
pump.
If the pump has a suction specific speed in Imperial units of 12,000, we
would incorporate an EROE of −30% to +10% of the BEP Flow.
L.L. 2.6: Failure to identify the Suction Specific Speed Value and Boundaries of Operation for a Double Suction Impeller can result in significant
damage and even shaft breakage
The writer has experienced many times where pump component failures in
Double Suction Pumps have occurred while operating in a range within −50%
of the Best Efficiency Point flow, and some that occurred very close to the BEP.
When Suction Specific Speed values were calculated, they were found to be
above 8,000 and some cases (where recirculation occurred close to the BEP)
revealed a suction specific speed close to 16,000.
BENCHMARKS
This best practice has been used since the early 2000s after experiencing many
times in Double suction pumps where failures have occurred outside of typical
EROE values. It is ideal to not purchase a pump with limited operating range,
but if it is already purchased and it is not seen that production will reduce greatly over the life of the plant, this best practice will help maintain these pumps and
their components at the highest possible reliability.
SUPPORTING MATERIAL
Piping
Piping accounts for the connection of the machinery to the environment surrounding the equipment. Improper piping assembly, like improper foundation
installation has resulted in reduced rotating equipment reliability.
General practices
Fig. 2.6.1 presents piping considerations that will result in proper installation of
equipment of high reliability.
69
B.P. 2.6
More Best Practices for Rotating Equipment
FIGURE 2.6.1 Piping considerations.
Piping must always be floated to the machine and NOT first mounted
on the machine. A machine is definitely not a pipe support. During construction, observe that piping is first mounted to vessels in pipe racks and
then and only then floated to the machine. Any other procedure is totally
unacceptable.
In bolting piping to machinery, lock pins on springs or pipe supports must
be kept in place until the system is filled with liquid or gas since they are designed to support the piping during the operation of equipment. Installation of
piping to equipment flanges must be performed with care. Bolts should be freely
removed from mating piping and equipment flanges without the use of force
(come along). Most importantly, flanges must be parallel on the machine surfaces of the mating and equipment flange within ±0.010 across the face diameter.
It is wise to always observe if piping has been removed or reassembled during
turnarounds. Frequently when this activity is performed, proper procedures are
not followed and excessive stresses are exerted on the equipment casing. Since
the equipment casing supports the bearings, which ultimately support the shaft
by antifriction bearings or a thin oil film in the case of hydrodynamic bearings,
improper piping assembly can significantly affect machinery operation. Keep
this fact in mind.
Suction specific speed
NSS, known as suction specific speed is determined by the same equation used
for specific speed NS but substitutes NPSHR for H (pump head). As the name
70
Pumps
Chapter | 2
TABLE 2.6.1 NSS Related to Flow Separation Probability
NSS
NPSHR
Inlet velocity
Inlet passage ∆P
Probability of
flow Separation
14,000 (High)
Low
Low
Low
High probability
8,000 (Low)
High
High
High
Low probability
TABLE 2.6.2 Recirculation as a Function of NSS
The onset flow of recirculation increases with increasing suction specific speed
implies, NSS considers the inlet of the impeller and is related to the impeller inlet
velocity. The relationship for NSS is:
N SS =
N Q
( NPSH R )3/4
Where N, speed; Q, flow − GPM; NPSHR, net positive suction head required.
NPSHR is related to the pressure drop from the inlet flange to the impeller.
The higher the NPSHR, the greater the pressure drop. The lower the NPSHR, the
less the pressure drop. From the previous equation, we can show the relationships between NPSHR, NSS, inlet velocity, inlet pressure drop, and the probability of flow separation in Table 2.6.1. Note that NPSHR for the previous equation
is at the Best Efficiency Point of the curve.
Based on the information presented in Table 2.6.1, it can be seen that flow
separation will occur for high specific speeds resulting from low inlet velocity.
The critical question the pump user needs answered is “At what flow does the
disturbance and resulting cavitation occur?” This is not an easy answer because
the unstable flow range is a function of the impeller inlet design as well as the
inlet velocity. A general answer to this question is shown in Table 2.6.2.
Another way of describing the statement in Table 2.6.2 is “The higher the
value of NSS, the sooner the pump will experience recirculation when operating at flows below the BEP.” Therefore, before an acceptable value of NSS can
be determined, the process system and pumped liquid characteristics must be
defined.
B.P. 2.7: When cost effective, Assure Driver and System have “End of
Curve” Power and NPSHA respectively
If an increase in flow through a particular pump process directly impacts
plant rates (more flow = more $), then the idea would be to operate the pump at
the highest flow possible!
71
B.P. 2.7
More Best Practices for Rotating Equipment
We define the general EROE range as +10% to −50% in flow from the pump
Best Efficiency Point, however what constitutes the +10% is the power available
in the driver and the NPSHA in the suction system piping and vessel height.
Therefore, if we have sufficient driver power and NPSHA, we can operate the
pump outside of the +10% and all the way to the right end of the performance
curve.
It is very important that this be discussed early on in the project and the production increase be monetized so as to show the payoff time for the increased
capital cost of a larger driver and a higher suction vessel.
L.L. 2.7: Inability to incorporate this Best Practice can result in pumps being bottlenecks in allowing for more plant production
BENCHMARKS
This BP has been used since the late 1990s in all plant applications where pump
production directly affected plant rates, thus eliminating the small equipment as
being bottlenecks.
SUPPORTING MATERIAL
See B.P. 2.2 for Supporting Material.
72
Chapter 3
Compressors
B.P. 3.1: Favor dry (No Oil Injection) screw compressors for process
applications below 5000 ACFM (8500 Actual m3/h)
If selected properly, a Centrifugal Compressor will provide the highest reliability, however will not be applicable at a proper efficiency below a certain
flow rate. Therefore, in these specific lower flow applications another selection
needs to be made.
The other selection in this application for years has been a Reciprocating
Compressor, however we all know the reliability of these to be lower due to
the number of high wear/maintenance components inherent with reciprocating
machinery.
Since being relatively new to the industry (were not really in use until the
mid-1960s, compared to reciprocating compressors being used in the mid-19th
century), users have been reluctant to use Dry Screw Compressors just because
of them being an unknown.
Furthermore, the addition of Lubricated Screw Compressors (oil injected
into the process) to the market in the 1980s has allowed users to purchase equipment that can produce about a 3 times higher compression ratio in one case and
therefore have a smaller footprint and less capital cost. However, it has been
seen that the oil injection can significantly affect the reliability of the machinery
and hence be much more expensive to operate.
A Dry Screw Compressor will allow the user to approach Centrifugal Compressor Reliability (greater then 99%) while maintaining high efficiency in low
flow applications.
L.L. 3.1: Selecting a Lubricated Screw or Reciprocating Compressor over
a Dry Screw Compressor has led to the following:
l
l
Lubricated screw compressor in sour gas service—never operated more than
24 h continuously.
Reciprocating compressors used when screw compressors should be used
resulting in extensive maintenance costs and sour gas leakage.
More Best Practices for Rotating Equipment. http://dx.doi.org/10.1016/B978-0-12-809277-4.00003-6
Copyright © 2017 Elsevier Inc. All rights reserved.
73
B.P. 3.1
More Best Practices for Rotating Equipment
BENCHMARKS
This best practice has been in use since the mid-1980s in upstream, refining,
and chemical plants. Following are typical Reliabilities for each major type of
Compressor if properly selected:
l
l
l
l
Centrifugal greater than 99.5%
Dry Screw greater than 99%
Lubricated Screw greater than 97.5%
Dry Reciprocating greater than 94%
SUPPORTING MATERIAL
This section overviews compressor types and their typical applications. The
two basic classifications of compressors are positive displacement and dynamic
compressors.
Positive displacement compressors are constant volume, variable energy
(head) machines that are not affected by gas characteristics.
Dynamic compressors are variable volume, constant energy (head) machines that are significantly affected by gas characteristics.
The type of compressor that will be used for a specific application therefore
depends on the flow rate and pressure required and the characteristic of the gas
to be compressed.
In general, dynamic compressors are the first choice since their maintenance
requirements are the lowest. The next choices are rotary type positive displacement compressors since they do not contain valves and are gas pulsation free.
The last choice is reciprocating compressors since they are the highest maintenance compressor type and produce gas pulsations. However, the final selection
depends upon the application requirements as discussed below.
Fig. 3.1.1 presents a flow range chart showing the various types of compressor applications as a function of flow (ACFM) and discharge pressure
(PSIG).
Table 3.1.1 shows the typical operating ranges for the various types of compressors used in the refining, chemical and gas processing industries.
Although the above table states Centrifugal Compressors can be used down
to 1200 m3/h (700 ACFM), this is by the use of 2D Type Impellers throughout
the compressor and will be at such low flows to each impeller that the efficiency
will be affected greatly. Below 8500 m3/h (5000 ACFM) is when efficiency will
begin to take a major hit (in the low 70% range). Note that Dry Screw Compressors will typically be above 85% efficient.
Screw compressors are the newest type of compressors. The dry screw compressor was developed in the late 1940s and did not experience wide use until
the 1960s for low to medium flow plant air services. On the other hand, reciprocating compressors were developed 100 years before (c. 1850) and centrifugal
compressors at the turn of the last century. In the 1980s the concept of the dry
74
Compressors
Chapter | 3
FIGURE 3.1.1 Compressor application range chart.
screw compressor was modified by continuously injecting a liquid (usually lube
oil), which enabled much higher compression ratios and simplified the mechanical design by eliminating the timing gears. Figs. 3.1.2 and 3.1.3 illustrate the
two major screw compressor designs. Screw compressors have many inherent
advantages over their competition in the low to medium flow range. Since they
are positive displacement compressors, like the reciprocating type, they will
draw a constant inlet volume (assuming constant speed), can meet the varying
differential pressure requirements of the process and are not significantly affected by gas density changes.
However, unlike their reciprocating cousins, they can accomplish the above
tasks by drawing a continuous, non-pulsating volume. As a result, pulsations
are minimized, suction and discharge valves and troublesome unloaders are not
required as well as high maintenance packing. Since there is only rotary motion,
all of the conventional sealing alternatives are available (including well proven
dry gas mechanical seals).
In addition, the rotary motion significantly reduces the “footprint” of the
screw compressor unit compared to other positive displacement alternatives.
The result is an efficient, reliable compressor type that is very competitive from
an initial cost and installed cost standpoint. The advantages of screw type compressors are presented in Table 3.1.2.
75
B.P. 3.1
76
TABLE 3.1.1 Typical Operating Range of Various Types of Gas Compressors
Machine
Type
m3/h
(ICFM)
Min
Max
T2 Max [°C
(°F)]
P1 Max [kPa
(PSIA)]
P2 Max [kPa
(PSIA)]
Min
Max
P/R Min
P/R Max
Rotary lobe
1–68,000 (1–40,000)
177 (350)
240 (35)
380 (55)
1.0+
2.4
Rotary vane
75–5,500 (45–3,300)
177 (350)
340 (45)
450 (65)
1.3
3.2
Rotary screw
80–34,000 (50–20,000)
177 (350)
1,000 (150)
4,250 (615)
2.0
6.0
Recip
1–17,000 (1 to 10,000)
427 (800)
6,900 (1,000)
69,000 (10,000)
3.0
50.0
Liquid ring
17–17,000 (10–10,000)
N/A
690 (100)
965 (140)
1.0+
10.0
Centrifugal
1,200–250,000 (700–150,000)
260 (500)
6,900 (1,000)
9,650 (1,400)
1.0+
3.4
500–250,000 (300–150,000)
427 (800)
13,800 (2,000)
41,400 (6,000)
2.0
10.0
125,000–600,000 (75,000–350,000)
427 (800)
210 (30)
1,030 (150)
1.0
10.0
Single stage
Centrifugal
Multi stage
Axial
More Best Practices for Rotating Equipment
Capacity
Compressors
Chapter | 3
FIGURE 3.1.2 Dry twin screw compressor. (Courtesy of Man/GHH)
FIGURE 3.1.3 Oil injected twin screw compressor. (Courtesy of Kobelco-Kobe Steel Ltd.)
As a result of their many advantages, dry screw and flooded screw compressors, have become a dominant force in low to medium flow process applications. They have also continued to grow as the preferred type of plant and
instrument air compressor in this flow range.
In the “upstream” exploration and production industry and in gas plants, the
screw compressor has become “the type to use.” For gas gathering, the depleted
fields of North America have utilized the advantages of the screw compressor
to provide highly reliable, cost effective service. Many “upstream” applications
77
B.P. 3.1
More Best Practices for Rotating Equipment
TABLE 3.1.2 Screw Compressor Advantages
j
j
j
j
j
j
j
j
j
Constant inlet volume flow
Can meet variable process differential pressure requirements
Can handle varying density and dirty gases
Pulsation free
Valves, unloaders and volume pockets not required
Mechanical seals used, packing not required
High reliability (close to centrifugal types)
Small “footprint”
Low installed cost
utilize gas turbines as prime movers for power generation, large pump, and
compressor drives. The screw is rapidly becoming the compressor of choice in
gas turbine fuel gas booster applications, which require medium flows, pressure
ratios, and the ability to handle varying gas densities. As the characteristics and
advantages of screw compressors become more widely appreciated, their use in
all types of applications will increase.
Throughout this section, the reader must remember that all screw compressors are positive displacement compressors.
Later, see Table 3.1.3 and Fig. 3.1.4, which define the differences between
Positive Displacement and Dynamic Compressors.
As will be discussed in this section, all types of positive displacement compressors present the designer with a challenge. This challenge is to provide
varying flow requirements to the process system in a safe and reliable manner.
You will discover that the screw compressor industry has met the challenge in a
safe, reliable, and most efficient way.
TABLE 3.1.3 Positive Displacement and Dynamic Characteristics
Positive Displacement (Screw, Lobe
and Reciprocating)
Dynamic (Centrifugal and Axial)
• Increases pressure by operating on
a fixed volume in a confined space
(constant volume if at Constant Speed)
• Increases Pressure by increasing fluid
velocity with rotating blades and
reducing the velocity in stationary
components.
• Need Self Limiting Device (Relief
valve or PSV) as flow will not change
and pressure will therefore increase
significantly
• Does not require self limiting device
as flow changes with pressure changes
• Not sensitive to changes within the
system.
• Sensitive to changes on the system.
78
Compressors
Chapter | 3
FIGURE 3.1.4 Positive displacement and dynamic typical performance curves.
Principles of Operation
Fig. 3.1.5 shows the basic principles of operation for a twin screw compressor.
Regardless of type, dry or flooded, the principle is the same.
As the screws separate at the suction end, the volume between the male
(drive) and female (idle) screw is filled with gas until the outlet screw flute
passes out of the suction volute, the suction volume is a function of the mating
screw volume and the speed of the compressor.
Once the screw flute passes through the section volute, the compression
phase begins and continues until the screw flute enters the discharge volute. The
designer determines the length of the compression phase by the machine specified requirements. The compressor ratio is a function of the volume reduction
ratio and the gas characteristics (specific heat ratio K = Cp/Cv). Details concerning performance will be discussed later in this chapter.
Fig. 3.1.6 shows the effects of operating any screw compressor on input
power at greater and lower than specified compressor ratio. This situation is
FIGURE 3.1.5 Principle of operation.
79
B.P. 3.1
More Best Practices for Rotating Equipment
FIGURE 3.1.6 Screw pv diagram.
always the actual operating mode in the field. Since power required is a function of volumetric efficiency (related to internal leakage through the screws) the
effects of off design operation (over or under pressurizing) are minimal (usually
less than 3%).
Screw Compressor Types
Twin Screw—Oil Free (Non Oil Injected)
Fig. 3.1.7 shows a single stage, oil free twin screw compressor with major components noted.
The oil free screw compressor was the first type developed for use as a plant
and instrument air compressor. Since no lubrication is introduced in the gas
stream, an external, close tolerance timing gear is used to separate the rotors.
Consequently, both screws must be sealed on each end to prevent oil from entering the gas stream.
80
Compressors
Chapter | 3
FIGURE 3.1.7 Single stage twin oil free compressor. (Courtesy of Kobelco–Kobe Steel Ltd.)
To maintain high volumetric efficiency by minimizing internal leakage (slip)
losses, the backlash of the timing gears (clearance between teeth) control the
screw rotor clearance to small valves. A major reliability factor in dry screw
type compressors is rotor deflection at high compression ratios. Typically, the
compression ratio for dry screw compressors is limited to up to 7:1. With special
screw rotor profiles, external cooling jackets and liquid injection (as great as
20% of inlet mass flow) higher compression ratios can be attained in one casing
(approx. 10:1) with discharge gas temperatures as high as 550°F. Flow range for
dry screw compressors is 300–40,000 ACFM. Please refer to Fig. 3.1.8 to view
the various screw compressor rotor profiles and their applications.
Dry screw compressors were the preferred type for plant and instrument air
services. However, in recent years, oil flooded screw compressors have gained
popularity in this field of application. This has been due to the development of
high efficiency oil separation coalessers and the lower cost for a flooded screw
type compressor unit. Dry screw compressors should be considered for sour
process services since sour gas can cause oil quality deterioration, which can
result in frequent oil changes and possible component damage. Facts concerning dry screw compressors are noted in Table 3.1.4.
81
B.P. 3.1
More Best Practices for Rotating Equipment
FIGURE 3.1.8 Screw compressor rotor profiles.
TABLE 3.1.4 Dry Screw Compressor Facts
j
j
j
j
j
j
82
External timing gears required
Normal compression ratio 5:1
Jacket cooling and/or rotor cooling and liquid injection allows operation up to
288°C (550°F) max ratio = 10:1 ∆P max = 2,585 kPa (375 psi) with special screw
profile
Multi staging may be required to prevent high discharge temperature and rotor
deflection 500–68,000 m3/h (300–40,000 ACFM)
Approximately ⅓ speed of centrifugal compressors and 3× speed of flooded screw
compressors
Seals required at each end of screws
Compressors
Chapter | 3
Twin Screw—Oil Flooded
A twin screw, oil flooded compressor is shown in Fig. 3.1.9 with major components noted.
The oil flooded twin-screw compressor, developed in the 1960s has become the
most widely used variety. The primary reason for its success is the ability to handle
very high compression ratios without external cooling. Compression ratios as high
as 25:1 are possible and ∆P’s of 5500 kPa (800 PSI) can be attained with special
screw profiles (6 + 8). Flow range for this type is 170–14,500 m3/h (100–8500
ACFM). Oil injection enables operation at high compression ratios but limits discharge temperature to 100°C (210°F). Use of synthetic oils allows discharges temperature to reach 120°C (250°F). The injection of oil also eliminates the necessity
of timing gears and reduces the number of shaft end seals required to one.
Another advantage of oil flooded screw compressors is the efficiency of its
capacity control system. The use of oil internal to the compressor allows the
use of a slide valve in the compressor section of the compressor. By varying the
stroke of the valve, the volume ratio can be varied. The result is that flow can be
adjusted between 10–100%.
Finally, flooded screw compressors have proven that they can tolerate dirty,
difficult to handle gases. The continuous oil injection serves as an “anti foulant
liquid” and prevents fouling build up. However, it is to note that these slide
valves are internal and issues have occurred in the field where the valve has either hung open or closed without indication other than flow meters downstream
of the compressor showing odd values.
FIGURE 3.1.9 Two-stage, tandem screw oil flooded compressor. Note: integral gear used for
speed increase of male rotors. (Courtesy of Kobelco–Kobe Steel Ltd.)
83
B.P. 3.1
More Best Practices for Rotating Equipment
TABLE 3.1.5 Flooded Screw Facts
j
j
j
j
j
j
j
j
External timing gears not required
Compression ratios as high as 25:1
Discharge temperature limited to 100°C (210°F) (mineral oil) 120°C (250°F)
(synthetic oil)
170-14,500 m3/h (100–8,500 ACFM)
⅓ speed of dry screw compressors
Only one shaft end seal required
Rotor cooling not required
Compatibility of oil with gas must be confirmed
Therefore, it can be seen that the screw compressor has many advantages
over the other types and can usually be purchased for a lower cost. The facts
concerning flooded screw compressors are detailed in Table 3.1.5.
However, the compatibility of the injected oil with the process gas must be
examined before purchasing a flooded type. There are two primary concerns:
j
j
Process system deterioration
Oil degradation
These are major concerns with reliability and why generally Oil injected
screw compressor reliability is significantly lower than Dry Screw Compressor
reliability.
Prior to purchase of a flooded screw compressor, the entire downstream
process system must be examined to confirm small quantities of oil will not
affect coolers, reactors etc. The efficiency of the oil separation systems used
is high 99.9%. However, upsets can cause oil to be transferred downstream. A
typical oil separation system used for a flooded screw compressor is shown in
Fig. 3.1.10.
The quoting screw compressor vendors should be consulted regarding modification to the standard system that can be made to meet requirements. Available
options are:
j
j
j
separate lube and seal oil injection systems
increased separator vessel retention time
self cleaning separator vessels upstream of the compressor
However, this activity must take place early in the project (phase 1) since the
cost of the unit will increase.
The other concern with flooded screw compressors is oil degradation caused
by interaction of the process gas with the injected oil. Please refer again to
Table 3.1.6. The retention time of the oil in the separator is typically 1–11/2 min.
Therefore, any contaminant in the process gas may not have sufficient time for
removal. There have been some bad experiences, particularly in gas recovery
84
Compressors
Chapter | 3
FIGURE 3.1.10 Oil separation system. (Courtesy of Man/GHH)
TABLE 3.1.6 Flooded Screw Reliability
In phase 1 of project
j Confirm compatibility of oil with process
j Confirm oil will not be degraded by process gas
j Visit installations to confirm vendor claims
applications where a sour, dirty gas (asphaltines) has required frequent oil and
oil filter change out. Our recommendation is to discuss these facts in detail
with the vendor prior to the order (Phase 1 of project) and require references of
where the equipment is operating. It has been our experience that this action,
unfortunately, is taken after installation of the compressor. These facts are presented in Table 3.1.6.
Performance Relationships
All screw compressors are positive displacement compressors. Performance
of screw compressors is calculated using the adiabiatic process. An adiabiatic
85
B.P. 3.1
More Best Practices for Rotating Equipment
compression process assumes that no heat is lost in the compression process.
The path of compression is defined by:
 K −1

K 
T2  P2  
=
T1  P1 
Table 3.1.7 defines the performance relationships used for screw compressors.
The relationships in Table 3.1.7 can be used to either estimate screw compressor performance for selection purposes or to determine field performance.
When determining screw compressor field performance, a trend of delivered
flow is most useful since all screw compressors are positive displacement
TABLE 3.1.7 Performance Relationships
j
j
j
CFM = (RPM) (DM3) (L/DM) (C0) where,
DM = male rotor diameter—inches
L/DM = length to diameter ratio of male rotor
C0 = rotor profile constant
4 + 6 profile = 15.853
3 + 4 profile = 18.082
4 + 6 ‘A’ profile = 17.17
4 + 6 ‘D’ profile = 16.325
3 + 4 ‘A’ profile = 19.372
6 + 8 ‘D’ profile = 10.364
P2 = P1
(T2 ) (k /k −1)
(T1 )
V2 (T2 ) (1 k )
=
V1 (T1 )
where, P = PSIA
T = °R
°R = °F + 460°
V = volume ( cubic FT min )
j
Adiabiatic H.P =
j
BHP =
P1Q1(P2/P1)
(K − 1)
229
K
Adiabiatic H.P
Adiabiatic efficiency (TOTAL)
Where, adiabiatic efficiency total = (adiabiatic efficiency) (mechanical efficiency)
 k −1

k 
j
 P  
T2 = T1  2 
 P1 
j
Adiabiatic efficiency =
86
∆T Adiabiatic
∆T actual
Compressors
Chapter | 3
TABLE 3.1.8 Field Performance Checks
j
j
Trend volume flow
Note: assure speed and control devices are at constant values.
Trend adiabiatic efficiency
Note: assure oil injection rates, jacket cooling temperature and rotor cooling flows
and valve positions are at constant values.
compressors and will deliver a constant volume flow. Reduction of flow is an
indication of an increase of slip (internal leakage). Care must be used in trending flow however. Rotor speed, suction throttle valve, slide valve and bypass
valve (if supplied) position must be constant.
Trending adiabiatic efficiency is another useful indicator of performances.
Again, care must be taken to assure oil injection (flooded screws), jacket water
temperatures, and rotor cooling flows (dry screws), in addition to valve positions are constant. These facts are presented in Table 3.1.8.
Mechanical Components
It is useful to always remember that any type of rotating machine contains five major components. It is important to know their function and monitor their condition.
Screw compressors are no exceptions. The five major component systems are:
j
j
j
j
j
rotor
journal bearing
thrust bearing
seals
auxiliary systems
Please refer to Fig. 3.1.11, which shows a typical dry screw compressor and
Fig. 3.1.12, which shows a typical flooded screw compressor. These figures will
be used to describe the function of the major mechanical components.
Rotor
As previously shown, there are various screw profiles available. Their use depends on the process condition and type of screw compressor (dry or flooded).
The most common profile is the asymmetric 4 + 6 (4 lobe male rotor and 6 lobe
female rotor). The major mechanical concern is rotor deflection. Therefore, the
L/D (rotor supported length divided by diameter) and number of lobes will vary
directly with capacity and pressure ratios. High flow rates may use only 3 male
lobes and high-pressure ratio applications may employ 6 male lobes.
We recommend that the contractor (or end user) require vendor rotor experience references during the bidding phase. These references should be contacted
to confirm reliable field operation for the proposed rotor configuration.
87
B.P. 3.1
More Best Practices for Rotating Equipment
FIGURE 3.1.11 Dry screw compressor. (Courtesy of Kobelco–Kobe Steel Ltd.)
Rotor speeds are low relative to centrifugal compressors. Consequently,
screw compressor rotors normally are rigid (they operate below their first critical
speed). Dry screw compressor tip speeds are 160 m/s (350 ft./s) and flooded
screw speeds are 30–50 m/s (100–175 ft./s).
FIGURE 3.1.12 Flooded screw compressor. (Courtesy of Kobelco–Kobe Steel Ltd.)
88
Compressors
Chapter | 3
Journal Bearings
The usual Journal bearings selected are roller type anti-friction bearings. Larger
compressors, above 220 kW (300 BHP) use sleeve or tilt pad bearings.
It is important to confirm proper bearing size and selection during the
bidding process. Anti-friction bearings should be checked to confirm that DN
number (diameter of bearing bore in mm multiplied by shaft speed) is within acceptable limits and the L-10 life is a minimum of 25,000 h. Sleeve bearing loads
(based on projected area) should be less than 1725 kN/m2 (250 PSI).
Thrust Bearings
Small screw compressors use angular contact anti-friction bearings. As stated
above, DN and L-10 life should be confirmed to be acceptable during the bid
phase.
Larger screw compressors above 220 kW (300 BHP) may use plain, tapered
land or even tilt pad thrust bearings for larger size, above 750 kW (1000 BHP).
Regardless of type, bearing loads should be less than 1725 kN/m2 (250 PSI).
Larger screw compressors use balancing devices to control the thrust load
(balance pistons).
Timing Gears
As previously mentioned, dry screw compressors require external timing gears
to assure proper rotor clearances. An increase in rotor clearance of 0.01 mm
(0.0004 in.) can reduce efficiency by 1%. The timing gears are precision ground
gears that require continuous lubrication. Small compressors, less than 220 kW
(300 BHP), can use self-contained ring oil lubrication. Larger compressors will
require a pressurized lubrication system.
Flooded screw compressors do not require external timing gears but do require significant amounts of oil that is injected into the screws (usually in the
bottom of the casing). As an example, a typical oil injection rate is 27 L/m
(7 GPM) per 170 m3/h (100 ACFM) flow rate for air compressors.
Sealing Devices
Dry screw compressors require that sealing devices be installed at each end of
each rotor. This is necessary to prevent timing gear oil from entering the process
stream and process gas entering the lube oil system. Flooded screw compressors
only require a sealing device on the drive screw-coupling end.
Regardless of type of screw compressor, there are many sealing alternatives
available. Some of the common types are:
j
j
j
j
labyrinth
restrictive carbon ring type
mechanical (liquid) seals
dry gas seals
89
B.P. 3.1
More Best Practices for Rotating Equipment
FIGURE 3.1.13 Typical oil flooded screw oil separation system. (Courtesy of Kobelco–Kobe
Steel Ltd.)
Auxiliary Systems
Small screw compressors can be supplied with self-contained lubrication (ring
oil) and sealing systems. Depending on operating conditions, a jacket cooling
system (dry screws) may be required.
Larger screw compressors will employ a pressurized lubrication system and
perhaps a liquid or gas seal buffer system.
All flooded screw compressors will require an oil separation system, which
is usually combined with the lubrication system. A typical P and ID for such a
system is shown in Fig. 3.1.13.
Experience has shown that bearing and seal component reliability is a direct function of auxiliary system component selection and design. The auxiliary
system(s) are the only major source of potential cost reduction for screw compressor vendors. Since screw compressors are relatively new, the specifications
have not reached the sophistication of reciprocating and centrifugal compressors. We recommend that all auxiliary systems be thoroughly reviewed in the
bidding phase, with references required and during the co-ordination meeting
for large compressors. Special care should be given in the pre-order phase to
gas/oil separator retention time and vendor experience with this item in similar
applications.
Capacity Control
Since screw compressors are positive displacement compressors, inlet volume
flow is constant. If the speed of the rotor remains constant, like reciprocating
compressors, various methods are used to allow capacity to be varied. The four
control methods used are presented in Table 3.1.9.
Variable Speed
As shown in the previous section, the capacity of the compressor is a function of
rotor profile and rotor speed. Variable speed is the preferred method of capacity
control for oil free compressors. It is not the preferred method for oil-flooded
90
Compressors
Chapter | 3
TABLE 3.1.9 Capacity Control Methods
j
j
j
j
Variable speed
Suction throttling
Slide valve
Bypass
compressors since injection oil would have to be varied with speed and the slide
valve method is available.
Suction Throttling
Suction throttling enables the mass flow of the screw compressor to be controlled by changing the inlet gas density. Care must be used to assure that the
maximum compressor ratio is not exceeded which could cause rotor deflection
and/or excessive discharge gas temperature.
Slide Valve
Oil injected screw compressors are usually fitted with a slide valve that allows
the suction volume to be varied between 10 and 100%. A slide valve inside the
screw case housing is shown in Fig. 3.1.14.
This stepless method of capacity control maintains high efficiency and is
the main advantage of oil injected screw compressors. The slide valve method
has recently been employed in oil free compressors for a limited number of
applications.
The operation of a slide valve is shown in Fig. 3.1.15. The slide valve is an
axial moveable segment of the casing cylinder wall. As it is moved from the
suction end toward the discharge end, flow is bypassed to the suction. Oil is
injected through ports in the valve cooling the recycled flow. This method of
control is the most efficient after the variable speed method.
Bypass Control
This method utilizes an external control valve to bypass excess gas back to
the suction. The recycled gas must be cooled. The control of the bypass valve
can be either pressure or flow. Bypass control is the most inefficient method of
capacity control.
Selection Guidelines
As previously mentioned, screw compressors have a flow range of 170–
68,000 m3/h (100–40,000 ACFM) and can produce compression ratios as high
as 25:1 (Oil Injected Screw Compressors). In order to optimize compressor
91
B.P. 3.1
More Best Practices for Rotating Equipment
FIGURE 3.1.14 Slide valve control.
efficiency, accurate process data must be provided to the quoting vendors. The
required input data is the same as for other types of compressors and is shown
in Table 3.1.10.
It is most important to accurately define gas contaminants (asphaltines, etc.)
and the percent per unit volume. Sour gas (H2S) must be accurately defined and
will determine material types and if a flooded screw can be used.
Dry or Wet Screw?
As discussed in this chapter, the technical and cost advantages of flooded
screw compressors, has made them the compressor of choice. However, special
modifications may be required for use of flooded screws in sour gas service.
The low retention time of oil in the separation vessel (60–90 s) do not allow
sour gas components to be vented from the reservoir oil. This can lead to oil
contamination, screw component damage and frequent oil changes. Reservoir
92
Compressors
Chapter | 3
FIGURE 3.1.15 Slide valve operation. (Courtesy of Kobelco–Kobe Steel Ltd.)
TABLE 3.1.10 Required Process Data
j
j
j
j
j
j
P1, P2
T1
Gas analysis (for each case)
Gas contaminants
Flow rate
Off design flow requirements
capacity can be increased, but at an extra cost. The use of dry screw compressors should be considered in sour gas service and evaluated against properly
designed oil flooded compressors. Vendor experience lists should be required
and checked.
Condition Monitoring
Condition monitoring requirements for screw compressors should follow the
principle of component condition monitoring. The following major components
should be monitored:
j
j
Rotor
Journal bearing
93
B.P. 3.2
More Best Practices for Rotating Equipment
TABLE 3.1.11 Minimum Condition Monitoring Required
j
j
j
j
j
Performance (rotor)
j Flow trend
j Efficiency check
Manual vibration readings (accelerometers/velocity)
Thrust bearing temperature measurement (rtd’s)
Thrust bearing
Seals
The suggested minimal condition monitoring requirements are presented in
Table 3.1.11.
B.P. 3.2: When to use medium and high speed (>400 RPM) reciprocating
compressors
Although it is generally recommended against using these types of
Reciprocating Compressors in process units due to low inherent reliability and
high maintenance costs, there is a very valuable application in which they are
recommended.
That would be in upstream services in the gas exploration fields where the
locations are remote with limited if any electricity available for a motor and
wells are constantly changing. These compressors are typically used in the field
as a package with a reciprocating gas engine, using the gas from the wells to
drive it.
A huge advantage in this service would be the smaller footprint of this machine allowing it to be portable and moveable from one location to another in
the gas fields. Most Users have at least one spare ready to go in case one of their
compressor packages has a failure at one of the wells.
L.L. 3.2: Use of a medium or high speed reciprocating compressor in
critical process units have resulted in very poor reliability
Due to the high speed, the packing, piston rings, bearings, and especially
the valves will wear at higher rates and the maintenance schedule increased. In
a petrochemical plant or refinery where a critical compressor being shutdown
equates to immediate loss of production, the larger frame/higher capital cost of
a low speed compressor will pay itself off in the first year of operation.
BENCHMARKS
This Best Practice has been utilized by many users in gas fields since the 1980s
as it was found that the mobility and ease of installation of these smaller skids
were ideal for this service.
94
Compressors
Chapter | 3
SUPPORTING MATERIAL
In the last 10 years, the used of medium speed (400–1000 rpm) and high speed
(1000–3600 rpm) lubricated reciprocating compressors has gained popularity
and initial acceptance mostly by project personnel. Granted, the capital costs are
lower and installation can and has been skid type in many cases.
Field personnel experience is significantly different and has resulted in
many hard lessons learned that now prohibit the use of this type of reciprocating compressor.
Maintenance costs, excessive pulsation and associated safety and mechanical issues have resulted in a wide aversion to the use of lubricated compressors
operating above 400 rpm. Typical component MTBF’s for high speed (greater
than 1000 rpm) are:
l
l
l
Packing—less than 12 months
Piston rings—less than 12 months
Valves—less than 12 months
Shutdown to repair pulsation related issues—less than 6 months
In this section the functions of each major component of a reciprocating
compressor are defined. That is, what the purpose of each component is or
“What It Does?” By understanding what each component is supposed to do, you
will be in a better position to know if it is performing its duty correctly. We will
present each major component starting with the crankcase, state its function, operating limits, and what to look for. After presenting each component’s general
information, we will present specific information concerning site compressors.
Frame and Running Gear
Fig. 3.2.1 presents a picture of a seven-throw crankshaft arrangement along with
a sectional view of two throws. The crankcase supports the crankshaft bearings,
provides a sump for the bearing and crosshead lube oil and provides support for
the crosshead assembly.
Typical crankcase condition monitoring and safety devices are:
j
j
j
j
j
j
Relief device—To prevent crankcase breakage in the event of explosion
(caused by entrance of process gas into the crankcase).
Breather vent—To allow removal of entrained air from the lube oil.
Crankcase oil level gauge—Allows continuous monitoring of crankcase
lube oil level.
Crankcase oil temperature gauge—Allows continuous monitoring of crankcase lube oil temperature.
Crankcase vibration detector (optional)—Provides information concerning
crankcase vibration useful in detecting dynamic changes in running gear.
Crankcase low oil level switch (optional)—Provides alarm signal on low
crankcase oil level.
95
B.P. 3.2
FIGURE 3.2.1
Rand)
j
j
More Best Practices for Rotating Equipment
Frame and running gear (Crankcase and crosshead). (Courtesy of Dresser
Main lube oil pump–shaft driven (optional)—Directly connected to crankshaft and usually discharges oil directly to crankshaft bearings, connecting
rod bearing, crosshead shoes and crosshead pin bushing via precision bore
in crankshaft and connecting rod bearing.
Main lube oil pump discharge pressure gauge (when supplied)—Allows
continuous monitoring of main lube oil pump discharge pressure.
An important reliability consideration is to assure that the crankcase is
securely mounted and level. This requires proper grouting and maintaining a
crack free (continuous) crankcase base support. Since the dynamic forces on
the crankcase and crosshead mounting feet can be very large, it is usually common to use an epoxy grout. Epoxy grouts provide high bond strengths and are
oil resistant. All reciprocating baseplates should be continuously checked for
any evidence of grout foundation cracks (discontinuities) and repaired at the
first opportunity.
Figs. 3.2.2 and 3.2.3 show plan, elevation, and side views of a two throw
balanced opposed crankcase assembly.
The crosshead assembly shown has the function of continuously assuring
vibration free reciprocating motion of the piston and piston rod. The crosshead pads (or shoes) and supports are usually made from Babbitt or aluminum
(smaller size units). Crosshead assembly lubrication is supplied via a pressuredrilled hole (rifle drilled) in the connecting rod which in turn lubricates the
crosshead pin bushing and crosshead shoes (Fig. 3.2.3).
96
Compressors
Chapter | 3
FIGURE 3.2.2 HDS off-gas running gear. (Courtesy of Dresser Rand)
FIGURE 3.2.3 HDS off-gas running gear. (Courtesy of Dresser Rand)
97
B.P. 3.2
More Best Practices for Rotating Equipment
Cylinder Distance Piece
Fig. 3.2.4 presents the functions of the cylinder distance piece.
The proper operation of the distance piece baffles and seals is essential to
maintaining reciprocating compressor safety and reliability in process gas applications. In most refinery process gas applications, a double compartment distance piece is used to assure contamination of the crankcase or cylinders does
not occur. Usually, the cylinder end compartment contains a partial N2 atmosphere since the packing rings is usually N2 purged.
Reliability considerations concerning this assembly are assuring proper
packing, partition packing, and wiper ring clearances.
Piston Rod Packing
Fig. 3.2.5 depicts a sectional and exterior view of a cartridge packing assembly.
The number of packing rings and type of arrangement is varied according
to the cylinder maximum operating pressures. It is important to note that the
packing does not provide an absolute seal, but only minimizes the leakage from
the cylinder. Shown in the upper portion of the section drawing in Fig. 3.2.5
FIGURE 3.2.4 Cylinder distance piece.
FIGURE 3.2.5 Cylinder packing.
98
Compressors
Chapter | 3
FIGURE 3.2.6 Left: oil scraper ring arrangement for 9 and 11 in. ESH/V, HSE units; right:
piston rod packing. (Courtesy of Dresser Rand)
is the vent port which carries the leakage gas either to a safe vent location
(atmosphere, flare, or fuel gas system) or back to the cylinder suction.
A means should be available to provide easy detection of excessive packing
clearances. Alternatives are:
j
j
j
Packing line flow switch
Packing line orifice and pressure switch (only if compressor pressures are
controlled to be constant)
Visual detection of gas flow (vent). Note: Flammable or toxic process gas
must be purged with N2 to attain a non-flammable mixture if the gas is to be
vented to atmosphere.
Fig. 3.2.6 shows additional packing assembly details.
The figure on the left side of the drawing is typical of a packing arrangement
used between sections of a distance piece. Mounted horizontally, the assembly
is equipped with a gravity drain and top vent.
The figure on the right side of Fig. 3.2.6 shows a four ring piston rod packing
assembly. The upper part of the drawing shows the lubrication connections that
are used when lubricated packing is required. Lube packing is normally used if
the lubricant is compatible with the process stream.
If dry packing is used, piston rod speeds are usually slower and PTFE materials are usually employed. In the lower half of the drawing, the vent connections are shown and perform as previously discussed. The cup supports and
positions an individual packing ring. Not shown is a purge connection, which
is inserted between the last and next to last packing ring (rings closest to the
distance piece).
Cylinder and Liner
Shown in Fig. 3.2.7 are the two most common cylinder arrangements; double
acting and single acting.
99
B.P. 3.2
More Best Practices for Rotating Equipment
FIGURE 3.2.7 Cylinder and liner. (Courtesy of Dresser Rand)
Most process reciprocating compressors are supplied with a replaceable
cylinder liner. All cylinders are either jacked for cooling H2O or finned for aircooling. Some older design cylinders use gaskets to isolate cooling water jackets from the cylinder. This design exposes the user to breakage from excessive
cylinder H2O entrainment if the gasket fails. Most reciprocating specifications
today do not allow gaskets to be used in the cylinder. A double acting cylinder is
designed to compress gas on both ends of the cylinder. (crank end and cylinder
head end) while a single acting cylinder is designed for compression only on
one end of the cylinder.
Reciprocating Compressor Cylinder Valves
There are many different types of reciprocating compressor valves. Regardless
of their design, all valves perform the same function … they allow gas to enter
the cylinder, prevent recirculation flow back to the suction piping and allow
gas to pass into the discharge system when the process discharge pressure at
the compressor flange is exceeded. Valves are the highest maintenance item in
reciprocating compressors. Their life is dependent on gas composition and condition, gas temperature and piston speed. Typical valve lives are:
j
j
Process gas service in excess of 1 year
H2 gas service—8–12 months
In hydrogen service, particular attention should be paid to cylinder discharge
temperature in order to obtain maximum valve life, cylinder discharge temperature for service with > 60% H2 should be limited to 250°F. Recently, light
weight, non metallic valves (PEEK) have been used successfully to increase the
valve life in H2 service above 1 year.
Fig. 3.2.8 shows a typical channel valve assembly.
100
Compressors
Chapter | 3
FIGURE 3.2.8 Reciprocating compressor valves—Channel type, Suction and Discharge.
(Courtesy of Dresser Rand)
The life of the channel valves shown is controlled by the spring force of the
valve springs. The channel arrangement reduces the forces on the valve seal and
usually results in increased valve life.
Fig. 3.2.9 depicts a ring or plate valve assembly. This type of valve is most
widely used.
Regardless of the type of valve, condition monitoring of valves is important to the profitability of any operation. The following parameters should be
monitored:
Type of valve
Suction
j
j
Discharge
j
j
Valve body temperature
Compressor volume flow rate
Interstage process gas temperature
Compressor volume flow rate
Changes in these parameters in excess of 10% from original (baseline) values should be cause for component inspection and replacement.
Piston Assembly
Fig. 3.2.10 presents a typical piston assembly consisting of the piston rod nut,
piston rod, piston, and piston nut.
Piston rod materials are hardened steel and can include metal spray in packing areas to extend rod life. Piston materials can be steel, cast nodular iron, or
aluminum. The most common being cast iron due to its durability. Aluminum
101
B.P. 3.2
More Best Practices for Rotating Equipment
FIGURE 3.2.9 HDS suction and discharge valves. (Courtesy of Dresser Rand)
FIGURE 3.2.10 Rod and piston.
102
Compressors
Chapter | 3
FIGURE 3.2.11 HDS off-gas-piston rod and piston. (Courtesy of Dresser Rand)
pistons are used in large cylinder applications (usually 1st stage) to minimize
piston rod assembly weight.
Fig. 3.2.11 shows piston rider bands (2) items 111 and piston rings (3) items 113.
Rider band and ring material is dependent on cylinder lubrication. If the cylinder is lubricated, carbon materials or compounds are used. If non-lubricated
service is required, PTFE materials or other Teflon derivatives are used. Note
also the piston-hollowed area for piston weight control. Rider band and ring
life is a function of piston speed, cylinder gas temperature and cleanliness of the
process gas. In many process applications, a strainer is required upstream of the
compressor to prevent excessive ring wear.
Condition monitoring of rider band and piston ring wear can be accomplished by measuring and trending the vertical distance between each piston rod
and a fixed point (known as rod drop). This can be accomplished either by mechanical or electrical (Bentley Nevada proximity probe) means. Of importance
in piston assembly design is rod loading and rod reversal.
Rod loading is the stress (tension or compression) in the piston road and
crosshead assembly caused by the ∆P across the piston. Rod load limits the
maximum compression ratio that a cylinder can tolerate. This is the reason that
many first stage cylinders are supplied with a suction pressure switch. Rod reversal is necessary so that the piston rod reaction forces on the crosshead pin
will change allowing oil to enter the pin bushing. If the position of the pin in
the bushing did not change (reverse) with each stroke, the bushing could not be
sufficiently lubricated and would prematurely fail.
Pulsation Dampeners
Since the action of the piston is noncontinuous, pressure pulsations will be
generated. Depending upon the piping arrangement, these pulsations can be
magnified to destructive levels. The use of pulsation dampeners, as shown in
Fig. 3.2.12 can reduce pulsations to 2% or lower.
103
B.P. 3.2
More Best Practices for Rotating Equipment
FIGURE 3.2.12 Pulsation dampeners. (Courtesy of Dresser Rand)
There are methods available to evaluate and simulate the effect of pulsation
dampeners prior to field operation. However, the variation between predicted
and actual results can be large and field modifications (installation of orifices or
pipe modifications) may be necessary.
Cylinder and Packing Lubricators
Whenever mineral oil is compatible with the process, lubricators will be
used. Lubricators can be either positive displacement or dynamic type. Attendees are asked to review lubrication details in the appropriate instruction
book. Lubricators will increase piston ring and packing life by reducing
friction.
Fig. 3.2.13 presents a typical lube oil system and its function.
ALL instruments in the lube oil system should be continuously monitored
(baseline and current conditions). Remember, component (bearing) failure will
occur if any major component in the system fails to function.
Fig. 3.2.14 shows a lube oil system containing a shaft driven main lube oil
pump with an internal relief valve.
This arrangement is a common one. Failure of the relief valve to seat can
cause a low lube oil pressure trip.
104
Compressors
Chapter | 3
FIGURE 3.2.13 Lube oil system.
Cooling System
The final topic to be covered is the cooling system. The cylinders, packing,
and process gas must be cooled to extend run time and minimize maintenance.
Fig. 3.2.15 presents a typical water-cooled circuit.
In addition to cooling, the temperature of the cooling water must be regulated so that moisture (condensate) will not form in the cylinder in wet gas
applications. It is recommended that the tempered water system temperature in
the cylinder be maintained a minimum of 10–15°F above the cylinder inlet gas
temperature.
Careful monitoring of the cooling circuit is essential in determining cooler,
jacket, and cylinder maintenance (cleaning) requirements.
105
B.P. 3.2
More Best Practices for Rotating Equipment
FIGURE 3.2.14 HDS off-gas-lube oil system. (Courtesy of Dresser Rand)
FIGURE 3.2.15 Cylinder, packing, and intercooler cooling water system.
106
Compressors
Chapter | 3
B.P. 3.3: Meet with Process Licensor and/or EP&C Process Engineers as
early as possible in the project to assure all operating and process conditions are on the data sheets
This Best Practice must be applied in conjunction with B.P. 1.6 as it is essential that the process conditions be finalized as early as possible as to not
delay the project.
All operating conditions need to be considered from initial plant start—up
to normal operating initially plus what is to be expected in 15–30 years from
initial start-up and beyond. Utilities conditions need to be considered as well
since Steam and Cooling can affect the efficiency of a train.
While it may be difficult to consider early on in the project, the change
in conditions over time must be looked at as well. After a number of years
(or sometimes before start-up of the plant) the plant will typically like to
produce more than rated case, but more power will be required if more head
and/or flow is required, so the Driver and Driven must both meet these requirements. Another issue in many plants today is Gas Curtailment as many
plants may be built based on a certain supply but either the supply becomes
less over time or other plants come online, increasing demand of the feed
gas. This will require a Dynamic Compressor (Centrifugal or Axial) to run
closer to surge, making the head rise to surge more important when selecting
a Dynamic Compressor.
Also, many compressor applications may need to start-up (initial and after
Process Unit Turnarounds) on N2 or another gas that is different from normal
operation in order to bring the unit online. This could greatly affect the performance of a Centrifugal Compressor if Gas density is 20% or more away from
the normal gas and if that is the case, certainly new curves will need to be drawn
for those conditions.
L.L. 3.3: Failure to assure all operating and process conditions are noted
on the data sheets will most likely result in lower reliability and possible
loss of production
If the train cannot not meet one of the process conditions that it may see
throughout the life of the plant, then production may not be met whether due
to lack of driver power or lack of ability in the driver to meet the head or flow
requirements for the specific gas conditions.
BENCHMARKS
This Best Practice has been used in the industry since the mid 1990s and
is even more important today when the profits of production are more and
more. When followed, this Best Practice can result in compressor reliability
of 99.7%.
107
B.P. 3.3
More Best Practices for Rotating Equipment
SUPPORTING MATERIAL
In this section we will cover the relationships that the COMPRESSOR VENDOR uses to determine the head produced, efficiency, horsepower required
and overall design for a particular compressor application. The END USER’s
or the PURCHASER’s objective is to deliver a specified amount of a given
gas to the process. Therefore, the data that the compressor vendor obtains is
required mass flow, inlet pressure, temperature conditions and gas composition. With this data a compressor manufacturer will calculate actual flow, the
ideal energy required and the horsepower required to achieve that objective.
The calculation for horsepower will require a specific compressor efficiency
as well as compressor mechanical losses, that is bearing friction losses, seal
losses, and disc friction losses. Gas characteristics are defined in this chapter
and useful relationships are presented to enable the reader to calculate various
compressor requirements. Once the VENDOR obtains the data, the gas head
can be calculated. Once the head and required flow are known, the impeller
can be selected.
The principle of impeller design is chiefly based on that of specific speed.
Specific speed is defined as the ratio of speed times the square root of the actual flow divided by head raised to the three quarters power. It can be shown
that increasing values of specific speed will result in increasing impeller efficiencies. Therefore, having been given the required flow and energy (head) the
only source of obtaining higher specific speed for the vendor is to increase the
compressor speed. This fact is very significant, because while compressors have
increased in efficiency over the years, the mechanical requirements have also
increased significantly, that is, higher bore impeller stresses, etc. resulting in
potential reliability problems. Therefore, the design of the impeller is a very fine
balance between the performance requirements and the mechanical constraints
of the components used in the compressor design.
Efficiency is presented as a ratio of ideal energy to actual energy as depicted
on a typical Mollier Diagram. In addition, the fan laws are presented showing
how increased impeller energy can be obtained via speed change in a compressor application.
Satisfying the Objective
The objective of the end user is to deliver a specified amount of a given gas. Refer to Fig. 3.3.1 and note that his objective can best be stated by the relationship:
Gas flow produced = Gas flow delivered.
This incidentally is the reason why most process control systems monitor
pressure in the process system and install a controller to either modulate flow
via a control valve (change the head required by the process) or vary the speed
of the compressor (change the head produced by the compressor).
108
Compressors
Chapter | 3
FIGURE 3.3.1 The objective: to deliver a specified amount of a given gas.
The vendor then, determines the head required by the process based on the
parameters given by the contractor and end user on the equipment data sheet.
It is very important to note that all possible sources should be used to confirm
that the conditions stated on the data sheet are correct and realistic. This fact is
especially true for dynamic compressors, since erroneous process conditions
will impact the throughput of the compressor.
Gas Characteristics
Table 3.3.1 presents the relationships used to calculate the design parameters for
the compressor. Note that the same relationships are used regardless of the type
of compressor (Positive Displacement or Dynamic).
The gas characteristics used in the determination of design parameters are
defined in Table 3.3.2.
Table 3.3.3 shows useful relationships used in compressor calculations as
well as the definitions for constants used.
Compression Head
The ideal gas head equations are again defined in Table 3.3.4. As previously
stated, polytropic head is the usual choice among compressor vendors.
109
B.P. 3.3
More Best Practices for Rotating Equipment
TABLE 3.3.1 Performance Relationships
To achieve the client’s objective the compressor vendor must calculate the actual flow
to the compressor inlet, the actual energy and work required
Actual flow
Volume flow rate m3/h (ft.3/min) = mass flow rate kg/h (lb/min) × density kg/m3 (lb/ft.3)
gasdensity
dependson P1,
T1, Z ,MW
Energy (ideal)
= m - kgf kgm
Compression
HEAD POLYTROPIC
 ft. − lbs 
 lb mass 
Energy (ideal) to compress
and deliver one LB of gas
from P1 to P2
Depends on P1,T1,
Z avg ,K avg ,MW P2 ,
effciency
Work
m − kgf  foot-lb's 
kg  lb 
mass flow 

kgm  lb mass 
hr  min 
Power kW (hp) =
 ft − lb's  
m − kgf 
3600
33,000 
 × efficiency (%)
h − kW 
 min − hp  
ideal energy
TABLE 3.3.2 Gas Characteristics
Compressibility (Z)
−
Accounts for the deviation from an ideal gas
Specific heat (C)
−
The amount of heat raise one mass of gas one
degree
CP and CV
−
Specific heat at constant pressure and volume
respectively
Specific heat ratio (K)
−
CP/CV
MW
−
Molecular weight
Polytropic exponent (n)
−
Used in polytropic head calculation
n −1 k −1
1
=
×
n
k
η polytropic
Impeller Types and Specific Speed
Various types of radial (centrifugal) impellers are shown in Figs. 3.3.2 and 3.3.3.
Open Impellers
Some open impellers are shown in Fig. 3.3.2. The advantage of open impellers
is their ability to operate at higher tip speeds and thus produce greater head than
110
Compressors
Chapter | 3
TABLE 3.3.3 Useful Relationships
Actual flow—m3/h (FT3/min)
m3 / h(ACFM) =
mass flow kg/h (lbs/min)
density kg/m3 (lbs/ft.3 )
where,
C = 3,600 =
m − kgf  FT − LBS 


h − kW  Min − H.P. 
HEADm − kgf  ft − lbs 


kgm
lb 
Densitykg/m3 (lbs/t 3 ) =
(P )
ZRT
HD =
ACFM = m3/h Nm3 /h ×
(101) (T )
P 289
Mass flow =
kg  lb 


h  min 
(14.7) (T) 

 SCFM ×

P 520 
Energy (Ideal) −
m − kgf
(ft − lb/lb Mass)
kgm
Use head equation, Polytropic is usually used
Eff’y = corresponding efficiency
(polytropic, isentropic, etc)
P = pressure—kPaa (psia)
Efficiency (%)
T = temperature—K (R*)
Derived from impeller test results—does
not include mechanical losses
K = °C + 273.1 (*R = °F + 460)
Z = compressibility
Work—kW (horsepower)
R = 1545/mol. wgt
Brake power = gas power + mech. losses
Nm3/h = Normal m3/h referenced to 17°C
and 101 kPA (SCFM = standard FT3/min
referenced to 60°F and 14.7 psia)
Gas power =
(HD)(Mass flow)
(C)(eff'y)
closed impellers. Open impellers can produce approximately 4,500–7,500 m-kg
force/kg mass (15,000 – 25,000 ft.-lbs force/lb mass) of head per stage. This is
because a side plate is not attached to the inlet side of the vanes, which results
in significantly lower blade stresses. The disadvantages of open impellers are
their lower efficiency due to increased shroud (front side) leakage and increased
number of blade natural frequencies resulting from the cantilevered attachment
of the blades to the hub. Most end users restrict the use of open impellers to plant
and instrument air applications since the high speeds and intercooling offset the
efficiency penalties caused by shroud leakage. Older design multistage centrifugal compressors frequently used open impellers in the first stages since the high
flows caused unacceptable side plate stresses in closed impeller design. Modern
calculation (finite element) methods and manufacturing methods (attachment
techniques—machine welding, brazing, etc.) today make possible the use of enclosed first stage impellers for all multistage compressor applications. Finally,
111
B.P. 3.3
More Best Practices for Rotating Equipment
TABLE 3.3.4 Ideal Gas Head Equations
Isothermal Head
HD
M − Kgf/kgm =
  P2  
847.4  1545 

 (T1)( Z AVG ) LN   
MW  MW 
  P1  
(FT − Lbf/Lbm)
Isentropic (Adiabatic) Head
HD
 K −1




847.4  1545 
 K 
 P2   K  
− 1
M − Kgf/kgm =

 (T1) 
 ( Z AVG )  
P 
MW MW
K −1
 1

(FT − Lbf/Lbm)
Polytropic Head
HD
M − Kgf/kgm =
 n −1




847.4  1545 
 n 
 P2   n  
− 1
(
)
(
Z
)
T

 1 
 AVG  
P 
MW MW
n −1
 1

(FT − Lbf/Lbm)
Where:
•
•
•
•
•
•
847.4
= Metric Gas Constant “R”
MW
1545
= Customary Gas Constant “R”
MW
MW = Molecular Weight
T1 = Inlet Temperature °K or °R
°K = 273.1 + °C
°R = 460.0 + °F
 Z + Z2
• ZAVG = Average Compressibility  1


2 
• K = Ratio of Specific Heats Cp/Cv
 K − 1  1 
=
=
 K   Polyn 
•
n −1
= Polytropic Exponent
n
•
•
•
•
Polyn = Polytropic Efficiency
Ln = Log to base A
P1 = Suction Pressure KPa (PSIA)
P2 = Discharge Pressure KPa (PSIA)
112
T2
T1
P2
Ln
P1
Ln
Compressors
Chapter | 3
FIGURE 3.3.2 Compressor impellers.
radial bladed impellers (whether open or enclosed) produce an extremely flat
(almost horizontal) head curve. This characteristic renders these impellers unstable in process systems that do not contain much system resistance. Therefore,
radial impellers are to be avoided in process systems that do not contain much
system resistance (plant and instrument air compressors, charge gas compressors, and refrigeration applications with side loads).
Enclosed Impellers
Enclosed impellers are shown in Fig. 3.3.3.
Note that the first stage impeller in any multistage configuration is always
the widest. That is, it has the largest flow passage. As a result, the first stage impeller will usually be the highest stressed impeller. The exception is a refrigeration compressor with side loads (economizers). Fig. 3.3.4 depicts a three-stage
rotor for a refrigeration application in a large LNG plant. Note the large axial
component of each impeller. Each impeller has a very large flow coefficient due
FIGURE 3.3.3 Enclosed impellers. (Courtesy of IMO Industries, Inc.)
113
B.P. 3.3
More Best Practices for Rotating Equipment
FIGURE 3.3.4 High Flow Refrigeration Compressor Rotor. (Courtesy of MHI)
to the breakthroughs in manufacturing (namely One Piece Impeller Design) that
many vendors, if not all, can apply today.
Dynamic compressor vendors use specific speed to select impellers based on
the data given by the contractors and end user. The vendor is given the total head
required by the process and the inlet volume flow. As previously discussed, at the
stated inlet flow (rated flow) the head required by the process is in equilibrium with
the head produced by compressor. Vendor calculation methods then determine how
many compressor impellers are required based on mechanical limitations (stresses)
and performance requirements (quoted overall efficiency). Once the head required
per stage is determined, the compressor speed is optimized for highest possible
overall efficiency using the concept of specific speed as shown in Fig. 3.3.5.
It is a proven fact that the larger the specific speed, the higher the attainable
efficiency. As shown, specific speed is a direct function of shaft speed and volume
flow and an inverse function of produced head. Since the vendor at this point in
the design knows the volume flow and head produced for each impeller, increasing the shaft speed will increase the specific speed and the compressor efficiency.
However, the reader is cautioned that all mechanical design aspects (impeller stress, critical speeds, rotor stability, bearing, and seal design) must be confirmed prior to acceptance of impeller selection. Often, too great an emphasis
114
Compressors
Chapter | 3
FIGURE 3.3.5 Impeller geometry versus specific speed.
on performance (efficiency) results in decreased compressor reliability. One
mechanical design problem can quickly offset any power savings realized by
designing a compressor for a higher efficiency.
Referring back to Fig. 3.3.5, calculation of specific speed for the first impeller by the contractor or end user will give an indication of the type of dynamic
compressor blading to be used. One other comment, Sundstrand Corporation
successfully employs an integral high-speed gearbox design for low flow, high
head applications or for low specific speed applications. The use of a speed increasing gearbox (for speeds up to 34,000 RPM) enables the specific speed to be
increased and therefore resulting in higher efficiency and less complexity than
would be obtained with a multistage compressor design approach.
Efficiency
Compressor efficiency, regardless of the type of compressor, can best be understood by referring to a typical Mollier Diagram as depicted in Fig. 3.3.6.
All produced heads shown on performance curves (isothermal, isentropic,
and polytropic) represent the ideal reversible head produced to compress a given gas from P1 to P2. This then is the theoretical compression path of the gas.
That is, the energy required to compress a gas if the efficiency is 100%.
However due to friction, sudden expansion etc., the efficiency is less than
100%. Therefore the actual compression path requires more head (energy) to
compress the gas from P1 to P2. The efficiency then is equal to:
Efficiency =
∆ E Ideal( E2 Ideal − E1 )
∆ E Actual( E2 Actual − E1 )
115
B.P. 3.3
More Best Practices for Rotating Equipment
FIGURE 3.3.6 Efficiency.
Note that ( ) is used to represent any ideal reversible path (isothermal, isentropic, polytropic).
Horsepower
Gas Horsepower is defined as the total actual energy (work) required to compress a given gas from P1 to P2 when compressing a given mass flow:


m − kgf  ft. − lbf 
Head ( )

 (Mass Flow − kg/h(lb/min))


kgm
lbm


÷(3,600(33,000) × Efficiency)
Note: ( ) must be for the same ideal reversible compression path. The brake
horsepower is the sum of the gas horsepower and the mechanical losses of the
compressor.
B.H.P. = G.H.P. + Mechanical losses
The mechanical losses are the total of bearing, seal, and windage (disc friction) losses and are provided by the compressor vendor. For estimating purposes, a conservative value of mechanical losses for one centrifugal or axial
compressor case would be 112 kW (150 H.P.).
116
Compressors
Chapter | 3
FIGURE 3.3.7 The Fan Laws.
The Fan Laws
These familiar relationships, sometimes called the affinity laws for pumps were
originally derived for a single stage fan, which is a low pressure compressor.
The Fan Laws are presented in Fig. 3.3.7.
As shown, if speed is changed, the flow, head and horsepower vary by the
first, second and third power of speed ratio respectively.
The reader must be cautioned however that the Fan Laws are only an approximation to be used as an estimating tool. Their accuracy significantly decreases with
increasing gas molecular weight and increase in the number of compression stages.
B.P. 3.4: Always assure highest head required point in the lifetime of the
plant is on the data sheets for centrifugal and axial compressors
This is to be followed in conjunction with B.P. 3.3 and B.P. 1.6 and is essential
to be sure that the machine is not operated close to surge throughout its lifetime.
This Best Practice will allow the vendors to propose a compressor that can
operate at this point without the surge control valve(s) being open during operation which = wasted energy.
Following this BP will also assure a compressor selection with an acceptable
head rise to surge (>5%).
L.L. 3.4: Not listing the highest possible head required on the datasheets
has resulted in compressors operating in the field with the Anti-Surge
Valves Open for long periods of time
That is not an acceptable way to operate, as it is a waste of Horsepower.
BENCHMARKS
This BP has been used since the mid 1990s to result in centrifugal and axial
compressor Reliability approaching 99.7%.
117
B.P. 3.5
More Best Practices for Rotating Equipment
SUPPORTING MATERIAL
See BP 3.3 for Supporting Material.
B.P. 3.5: When considering a horizontally split compressor case for pressures greater than 40 barg (600 psig), assure that there is a minimum of
2-year operating experience in a similar application
While a horizontal split compressor has a major benefit in capital savings, it
must be vetted properly to assure reliability is not going to be affected compared
to a barrel (vertical or radial split) type compressor.
Early on in the project (Pre-FEED), any vendor proposing a horizontal split
compressor should give a number of users with experience in a similar application. They can then be contacted to see if they have experienced any horizontal
split-line gas leaks during their operation.
L.L. 3.5: Failure to check experience of horizontal split case in high pressure applications has resulted in safety hazards due to split line leaks
BENCHMARKS
Since the late 1990s this Best Practice has been used in selection of centrifugal
compressors that yielded reliability of at least 99.7%
SUPPORTING MATERIAL
Centrifugal Multi-Stage Horizontal Split
A typical multi-stage horizontally split centrifugal compressor is shown in
Fig. 3.5.1. The casing is divided into upper and lower halves along the horizontal centerline of the compressor. The horizontal split casing allows access to the
internal components of the compressor without disturbing the rotor to casing
clearances or bearing alignment. If possible, piping nozzles should be mounted
on the lower half of the compressor casing to allow disassembly of the compressor without removal of the process piping.
Centrifugal Multi-Stage With Side Loads
This type of compressor is used exclusively for refrigeration services. The only
difference from the compressor shown in Fig. 3.5.1 is that gas is induced or removed from the compressor via side load nozzles. A typical refrigeration compressor is shown in Fig. 3.5.2. Note that this type of compressor can be either
horizontally or radially split.
Centrifugal Multi-Stage (Barrel)
A typical multi-stage, radially-split, centrifugal compressor in shown in
Fig. 3.5.3. The compressor casing is constructed as a complete cylinder with
118
Compressors
Chapter | 3
FIGURE 3.5.1 Centrifugal multi-stage horizontal split. (Courtesy of Mannesmann Demag)
FIGURE 3.5.2 Typical multi-stage refrigeration compressor.
119
B.P. 3.5
More Best Practices for Rotating Equipment
FIGURE 3.5.3 Typical multi-stage, radially split centrifugal compressor. (Courtesy of Mannesmann Demag)
one end of the compressor removable to allow access to the internal components. Multi-stage, radially-slit centrifugal compressors are commonly called
barrel compressors.
Impeller Types and Specific Speed
Various types of radial (centrifugal) impellers are shown in Figs. 3.5.4 and 3.5.5.
FIGURE 3.5.4 Compressor impellers.
120
Compressors
Chapter | 3
FIGURE 3.5.5 Enclosed impellers. (Courtesy of IMO Industries, Inc.)
Open Impellers
Open impellers are shown in Fig. 3.5.4. The advantage of open impellers is their
ability to operate at higher tip speeds and thus produce greater head than closed
impellers. Open impellers can produce 15,000–25,000 ft.-lbs/LB of head per
stage. This is because a side plate is not attached to the inlet side of the vanes,
which results in significantly lower blade stresses. The disadvantages of open
impellers are their lower efficiency due to increased shroud (front side) leakage
and increased number of blade natural frequencies resulting from the cantilevered attachment of the blades to the hub.
Most end users restrict the use of open impellers to plant and instrument
air applications since the high speeds and intercooling offset the efficiency penalties caused by shroud leakage. Older design multistage centrifugal compressors frequently used open impellers in the first stages since the
high flows caused unacceptable side plate stresses in closed impeller design.
Modern calculation (finite element) methods and manufacturing methods
(attachment techniques—machine welding, brazing, etc.) today make possible the use of enclosed first stage impellers for all multistage compressor
applications. Finally, radial bladed impellers (whether open or enclosed)
produce an extremely flat (almost horizontal) head curve. This characteristic
renders these impellers unstable in process systems that do not contain much
system resistance. Therefore, radial impellers are to be avoided in process
systems that do not contain much system resistance (plant and instrument
air compressors, charge gas compressors and refrigeration applications with
side loads.
121
B.P. 3.5
More Best Practices for Rotating Equipment
Enclosed Impellers
Enclosed impellers are shown in Fig. 3.5.5. Note that the first stage impeller
in any multistage configuration is always the widest. That is, it has the largest
flow passage. As a result, the first stage impeller will usually be the highest
stressed impeller. The exception is a refrigeration compressor with side loads
(economizers).
Dynamic compressor vendors use specific speed to select impellers based on
the data given by the contractors and end user. The vendor is given the total head
required by the process and the inlet volume flow. As previously discussed, at
the stated inlet flow (rated flow) the head required by the process is in equilibrium with the head produced by compressor. Vendor calculation methods then
determine how many compressor impellers are required based on mechanical
limitations (stresses) and performance requirements (quoted overall efficiency).
Once the head required per stage is determined, the compressor speed is optimized for highest possible overall efficiency using the concept of specific speed
as shown in Fig. 3.5.6.
It is a proven fact that the larger the specific speed, the higher the attainable efficiency. As shown, specific speed is a direct function of shaft speed and
volume flow and an inverse function of produced head. Since the vendor at this
point in the design knows the volume flow and head produced for each impeller,
increasing the shaft speed will increase the specific speed and the compressor
efficiency.
However, the reader is cautioned that all mechanical design aspects (impeller stress, critical speeds, rotor stability, bearing, and seal design) must be
confirmed prior to acceptance of impeller selection. Often, too great an emphasis on performance (efficiency) results in decreased compressor reliability. One
FIGURE 3.5.6 Impeller geometry versus specific speed.
122
Compressors
Chapter | 3
mechanical design problem can quickly offset any power savings realized by
designing a compressor for a higher efficiency.
Referring back to Fig. 3.5.6, calculation of specific speed for the first impeller by the contractor or end user will give an indication of the type of dynamic
compressor blading to be used. One other comment, Sundstrand Corporation
successfully employs an integral high-speed gear box design for low flow, high
head applications or for low specific speed applications. The use of a speed
increasing gear box (for speeds up to 34,000 RPM) enables the specific speed
to be increased and therefore resulting in higher efficiency and less complexity
than would be obtained with a multistage compressor design approach.
Critical Speeds and Rotor Response
The term “critical speed” is often misunderstood. In nature, all things exhibit a
natural frequency. A natural frequency is defined as that frequency at which a
body will vibrate if excited by an external force. The natural frequency of any
body is a function of the stiffness and the mass of that body. As mentioned, for
a body to vibrate, it must be excited. A classical example of natural frequency
excitation is the famous bridge “Galloping Gerty” in the state of Washington.
That bridge vibrated to destruction when its natural frequency was excited by
prevailing winds.
In the case of turbo-compressor rotors, their natural frequency must be excited by some external force to produce a response that will result in increased
amplitude of vibration. One excitation force that could produce this result is
the speed of the rotor itself. Thus the term “critical speeds.” The term “critical speed” defines the operating speed at which a natural frequency of a rotor
system will be excited. All rotor systems have both lateral (horizontal and vertical) and torsional (twist about the central shaft axis) natural frequencies. Only
lateral critical speeds will be discussed in this section.
In the early days of rotor design, it was thought that the rotor system consisted primarily of the rotor supported by the bearings. This led to the assumption that only the stiffness of the rotor supported by rigid bearings needed to be
considered in the analysis of the natural frequency. Countless machinery problems have proven this assumption to be false over the years. The concept of the
“rotor system” must be thoroughly understood. The rotor system consists of the
rotor itself, the characteristics of the oil film that support the rotor, the bearing,
the bearing housing, the compressor case that supports the bearing, compressor
support (base plate), and the foundation. The stiffness and damping characteristics of all of these components together result in the total rotor system that
produce the rotor response to excitation forces.
We will examine a typical rotor response case in this section and note the
various assumptions, the procedure modeling, the placement of unbalance, the
response calculation output, and discuss the correlation of these calculations to
actual test results.
123
B.P. 3.5
More Best Practices for Rotating Equipment
Critical Speeds
The natural frequency of any object is defined by the relationship:
FNATURAL =
K
M
Where K, stiffness; M, mass.
When excited by an external force, any object will vibrate at its natural frequency. If the frequency of the exciting force is equal to the natural frequency
of the object, and no damping is present, the object can vibrate to destruction.
Therefore, if the frequency of an exciting force equals the natural frequency of
an object, the exciting force is operating at the “critical frequency.”
Rotor speed is one of the most common external forces in turbo-machinery.
When the rotor operates at any rotor system natural frequency, it is said that the
rotor is operating at its critical speed. The critical speed of a rotor is commonly
designated as NC and the corresponding natural frequencies or critical speeds
are: NC1, NC2, NC3, etc.
Every turbo-compressor that is designed must have the rotor system critical
speeds determined prior to manufacture. In this section, we will follow the procedure for the determination of the necessary parameters to define a rotor systems
critical speed. The procedure is commonly known as determination of rotor response. Fig. 3.5.7 is a representation of a critical speed map for a rotor system.
It should be understood that all stiffness values are “calculated” and will
vary under actual conditions. As an exercise, determine NC1, NC2, and NC3
for the horizontal and vertical directions for each bearing in Fig. 3.5.7 (assume
bearing 1 and 2 stiffness are the same)
Critical speed
Horizontal (X)
Vertical (Y)
NC1
NC2
NC3
3,300 rpm
9,700 rpm
16,000 rpm
3,000 rpm
8,000 rpm
15,000 rpm
Based on a separation margin of ± 20% from a critical speed, what would be
the maximum allowable speed range between NC1 and NC2 in Fig. 3.3.7?
j
j
Maximum speed
Minimum speed
6600 rpm
4000 rpm
Remember, changing of any value of support stiffness will change the critical
speed. Plotted on the X-axis is support stiffness in pounds per inch. The primary
components of support stiffness in order of decreasing increasing influence are:
j
j
Oil support stiffness
Bearing pad or shell
124
Compressors
Chapter | 3
125
FIGURE 3.5.7 Compressor rotor critical speed map—no damping. (Courtesy of Elliott Company)
B.P. 3.5
j
j
j
j
j
More Best Practices for Rotating Equipment
Bearing housing
Bearing bracket
Casing support foot
Baseplate
Foundation
Note that this analysis of the critical speed does not include oil film damping. It is common practice to first determine the “undamped critical speeds” to
allow for necessary modifications to the rotor or support system. This is because
the effects of stiffness on the location of critical speed are significantly greater
than damping. Fig. 3.5.7 shows four distinct critical speeds. Operation within
±20 of actual critical speeds is to be avoided. Also plotted are the horizontal (x)
and vertical (y) bearing stiffness for each bearing. Note that these values vary
with speed and are the result of changes in the oil stiffness. Therefore, a change
in any of the support stiffness components noted above can change the rotor
critical speed. Experience has shown that critical speed values seldom change
from ±5% of their original installed values.
If a turbo-compressor with oil seals experiences a significant change in critical speeds, it is usually an indication of seal lock-up. That is, the seal does not
have the required degrees of freedom and supports the shaft acting like a bearing. Since the seal span is less than the bearing span, the rotor stiffness “K”
increases and the critical speeds will increase in this case.
The Rotor System (Input)
Fig. 3.5.8 shows a typical turbo-compressor rotor before modeling for critical
speed or rotor response analysis.
Since the natural frequency or critical speed is a function of shaft stiffness
and mass, Fig. 3.5.9 presents the rotor in Fig. 3.5.8 modeled for input to the
computer rotor response program. Fig. 3.5.9 is an example of a modeled rotor
and only includes the rotor stiffness (K) and mass (M).
In order to accurately calculate the rotor critical speeds, the entire rotor system stiffness, masses and damping must be considered. Table 3.5.1 models the
oil film stiffness and damping of the journal bearings at different shaft speeds.
FIGURE 3.5.8 Rotor response modeling—rotor. (Courtesy of Elliott Co.)
126
FIGURE 3.5.9 Rotor response input data—dimensions, masses, and unbalances. (Courtesy
of Elliott Co.)
127
B.P. 3.5
More Best Practices for Rotating Equipment
TABLE 3.5.1 Typical Compressor Oil Film Bearing Parameters
4 × 1.6 in. tilt 20.5 in. TB 3.0 in. shaftend 7.5–6.5 in. shaft Bendix coupling
Static bearing
load (lbs)
897
Diameter (inches)
4.00
Bearing station
12
Length (inches)
1.60
Bearing location
Thrust
Diam. assembly
clearance (inches)
5.7487E–03
Bearing type
Tilt pad
Diam. machined
clearance (inches)
8.7500E–03
Location of load
Between pads
Inlet oil temperature (°F)
120.0
Preload
0.343
Type of oil
DTE–light
(150SSU @100°F)
Fluid film stiffness
Damping
Speed
(rpm)
50 mm
No
KXX (Ib/in)
KYY (Ib/in)
WCXX (Ib/in) WCYY (Ib/in)
2500
0.114
1.3871E 06
7.5446E 05
7.7995E 05
4.6249E 05
3000
0.137
1.2984E 06
7.1330E 05
7.8487E 05
4.7587E 05
4000
0.183
1.1769E 06
6.6147E 05
8.0311E 05
5.0825E 05
4500
0.206
1.1341E 06
6.4543E 05
8.1400E 05
5.2564E 05
5500
0.252
1.0703E 06
6.2556E 05
8.3686E 05
5.6116E 05
6613
0.303
1.0230E 06
6.1679E 05
8.6656E 05
6.0354E 05
7000
0.321
1.0109E 06
6.1616E 05
6.7775E 05
6.1885E 05
8000
0.366
9.8751E 05
6.1898E 05
9.0798E 05
6.5935E 05
9000
0.412
9.7305E 05
6.2684E 05
9.4015E 05
7.0111E 05
10000
0.458
9.6556E 05
6.3864E 05
9.7461E 05
7.4430E 05
11000
0.504
9.636OE 05
6.5354E 05
1.0110E 06
7.8878E 05
12000
0.549
9.6610E 05
6.7094E 05
1.0490E 06
8.3434E 05
13000
0.595
9.7225E 05
6.9037E 05
1.0881E 06
8.8080E 05
14000
0.641
9.8144E 05
7.1149E 05
1.1283E 06
9.2801E 05
15000
0.687
9.9317E 05
7.3403E 05
1.1696E 06
9.7586E 05
Courtesy of Elliott Co.
128
Compressors
Chapter | 3
Note that it is essential that the type of oil to be used in the field (viscosity
characteristics) must be known. End users are cautioned to confirm with the
OEM before changing oil type as this will affect the rotor response. In addition
to modeling of the rotor and bearings, most rotor response calculations also
include the following additional inputs:
j
j
Bearing support stiffness
Oil film seal damping effects
Of all the input parameters, the effects of bearing and seal oil film parameters are the most difficult to calculate and measure. Therefore, a correlation
difference will always exist between the predicted and actual values of critical
speed. Historically, predicted values of NC1 (first critical speed) generally agree
within ±5%. However, wide variations between predicted and actual values
above the first critical speed (NC1) exist for NC2, NC3, etc.
When selecting machinery, the best practice is to request specific vendor
experience references for installed equipment with similar design parameters
as follows:
j
j
j
j
j
Bear span ÷ major shaft diameter
Speeds
Bearing design
Seal design
Operating conditions (if possible)
Once the rotor system is adequately modeled, the remaining input parameter is the amount and location of unbalance. Since the objective of the
rotor response study is to accurately predict the critical speed values and
responses, an assumed value and location of unbalances must be defined.
Other than bearing and seal parameters, unbalance amount and location is
the other parameter with a “correlation factor.” There is no way to accurately
predict the amount and location of residual unbalance on the rotor. Presently,
the accepted method is to input a value of 8 × A.P.I. acceptable unbalance
limit (4W)/N.
This results in a rotor response input unbalance of 32W/N.
The location of the unbalance is placed to excite the various critical speeds.
Typically the unbalances are placed as noted:
Location
To excite
Mid span
Quarter span (two identical unbalances)
At coupling
NC1
NC2
NC2, NC3
Failure to accurately determine the value and location of residual rotor unbalance is one of the major causes of correlation differences between predicted
and actual critical speeds.
129
B.P. 3.5
More Best Practices for Rotating Equipment
Rotor Response (Output)
The output from the rotor response study yields the following:
j
j
Relative rotor mode shapes
Rotor response for a given unbalance
Fig. 3.5.10 shows the relative rotor mode shapes for NC1, NC2, NC3, and
NC4. Usually, the rotor will operate between NC1 and NC2.
Rotor mode shape data is important to the designer because it allows determination of modifications to change critical speed values.
For the end user, this data provides an approximation of the vibration at any
point along the shaft as a ratio of the measured vibration data. As an example
in Fig. 3.5.10, determine the vibration at the shaft mid span if the vibration
measured by the probe C2 when operating at NC1 is 2.00 mils. From Fig. 3.5.10,
the vibration at the shaft mid span when operating at the first critical speed of
3327 RPM (50 in location) is:
1.00
or10 × the bearing vibration
0.1
Ten times the value at C2 or 20.0 mils!
Mode shape data should always be referred to when vibration at operating
speed starts to increase and your supervisor asks
“When do we have to shut down the unit?”
or
“Can we raise the radial vibration trip setting?”
In this example, the bearing clearance may be 0.006 or 6 mills. And an honest request would be … “We’ll replace the bearing at the turnaround, please run
to 7.0 mils vibration.”
Refer to Fig. 3.5.10 and remember:
j
j
j
The compressor must go through NC1
The shaft vibration increases at NC1 (usually 2X, 3X, or more)
The vibration at center span is approximately 10X the probe vibration
Therefore,
Vibration at the mid span during the first critical speed will be:
= (7.0 mils) ×
Probe value
= 140 mils!!
(2.0)
NC1 amplification
×
(10)
Mode shape difference
Normal clearance between the rotor and interstage labyrinths is typically
40 mils!! This vibration exposes the diaphragms, which are usually cast iron,
to breakage. One final comment … during shutdown, the rate of rotor speed
decrease CANNOT be controlled as in the case of start-up. It depends on rotor inertia, load in the compressor, the process system characteristics and the
130
Compressors
Chapter | 3
131
FIGURE 3.5.10 Rotor natural frequency mode shapes. (Courtesy of Elliott Co.)
B.P. 3.5
More Best Practices for Rotating Equipment
control and protection system. If the vibration at the probe locations is high, the
best advice is to stop the compressor fully loaded which will reduce the time
in the critical speed range as much as possible. Yes, the compressor will surge,
but the short duration will not normally damage the compressor. Figs. 3.5.11
and 3.5.12 present the primary output of a rotor response study.
Rotor response plots display vibration amplitude, measured at the probes,
versus shaft speed for the horizontal and vertical probes. Note that a response
curve must be plotted for each set of unbalance locations and unbalance amount.
FIGURE 3.5.11 Rotor response output at non-drive end bearing (NDE). (Courtesy of Elliott Co.)
FIGURE 3.5.12 Rotor response output drive end bearing (D.E.). (Courtesy of Elliott Co.)
132
Compressors
Chapter | 3
Fig. 3.5.11 shows the rotor response for the non-drive end (N.D.E.) set of
probes with the first set of unbalance. Fig. 3.5.12 shows the rotor response for
the drive end set of probes (D.E.). The operating speed range of this example is
6000–8000 rpm.
Measured Rotor Response
During shop test, the rotor response of every turbo-compressor rotor is measured
during acceleration to maximum speed and deceleration to minimum speed.
Values are plotted on the same coordinates as for the rotor response analysis.
The plot of shaft vibration and phase angle of unbalance versus shaft speed is
known as a BODE PLOT.
Bode plots represent the actual signature (rotor response) of a rotor for a
given condition of unbalance, support stiffness and unbalance. They indicate
the location of critical speeds, the change of shaft vibration with speed and
the phase angle of unbalance at any speed. A bode plot is a dynamic or transient signature of vibration for a rotor system and is unique to that system for
the recorded time frame. Bode plots should be recorded during every planned
start-up and shutdown of every turbo-compressor. As discussed in this section,
the bode plot will provide valuable information concerning shaft vibration and
phase angle at any shaft speed.
B.P. 3.6: Assure centrifugal compressors are selected that have a rated
operating temperature below 350 degrees Fahrenheit (approx. 180°C)
It is essential to evaluate all of the unpriced bids in a detailed manner very
early in the project. One of the items that may give one a vendor a cheaper
up front cost over another vendor is to stuff an extra impeller into one casing rather than going to a second casing. Especially with heavy gases this
will result in higher discharge temperatures (in each section if an intercooled
compressor), which could cause loss of productivity. Note that the discharge
temperature will greatly increase when the compressor is operated closer to
surge.
Remember also that the Users goal is to produce at least 100% rates 24/7
as long as possible, therefore a two case compressor, while costly up front, will
pay back quickly due to less planned and unplanned maintenance.
L.L. 3.6: Compressors selected with rated temperatures over 350°F will
result in lower reliability in the form of either efficiency loss, or worst unplanned shutdowns
The writer has witnessed on multiple occasions where the high temperatures can actually open up the diaphragms along the horizontal split and have
resulted in efficiency drops of 10% or more. This will rarely be seen during
the performance test because the gas will not be the same as in the field, so the
temperatures will never reach a high value. The fix for this can take significant
133
B.P. 3.7
More Best Practices for Rotating Equipment
time (up to 2 weeks) to implement and if done before a Turnaround would be
very costly in terms of production lost.
Also, there can be high vibrations and potential rubs due to thermal expansion of the rotor in relation to the casing. One case history has been witnessed
where the inter-stage sleeves expanded due to high temperatures so much
that they ended up bowing the rotor and had to reduce rates by more than
5% to maintain the vibration below 3.5 mils for a 6 year period until the next
Turnaround.
Finally, vendors may offer a high temperature cladding on the stationary
parts, however the writer would question the reliability of the cladding and the
potential for it to chip/break off during operation.
BENCHMARKS
This Best Practice has been in use since the 70s in order to maintain the highest
reliability possible for centrifugal compressors. If followed, centrifugal compressor reliability can approach and exceed 99.7%.
SUPPORTING MATERIAL
See supporting material for B.P. 3.5.
B.P. 3.7: Require a pulsation audit by an experienced company immediately after installation of reciprocating compressor
One of the most common issues experienced globally with reciprocating
compressors are high pulsations that can damage the machinery but even more
importantly could break small piping and tubing in the area, exposing a safety
hazard.
While the pulsation dampeners are designed based on what the piping is
expected to be like, it is very rare that it is exactly the same in the field, causing
pulsations to be greater than expected.
A proper field pulsation study will be able to determine whether more or less
cushion of gas is required on the suction, discharge, or both dampeners. Usually,
the modifications required after this study will be to add or change the size of
an orifice on the dampener.
L.L. 3.7: Failure to have a field pulsation study performed can result in
years of operation with premature component failures
BENCHMARKS
This best practice has been used since the 1990s as many reciprocating compressors had unacceptable pulsation values resulting in excessive bearing, piston ring, and valve wear and occasional safety hazards. Implementation of this
best practice resulted in significant increased reliability of all components, including the foundation.
134
Compressors
Chapter | 3
FIGURE 3.7.1 Pulsation dampeners. (Courtesy of Dresser Rand)
SUPPORTING MATERIAL
Pulsation Dampeners
Since the action of the piston is noncontinuous, pressure pulsations will be generated. Depending upon the piping arrangement, these pulsations can be magnified to destructive levels. The use of pulsation dampeners, as shown in Fig. 3.7.1
can reduce pulsations to 2% or lower.
There are methods available to evaluate and simulate the effect of pulsation
dampeners prior to field operation. However, the variation between predicted
and actual results can be large and field modifications (installation of orifices or
pipe modifications) may be necessary.
B.P. 3.8: Require a one-Piece Impeller for all sour gas services
Due to advancements in CNC machining all vendors have the capabilities
these days to machine an impeller out of one material, rather than welding the
front cover on the blades.
This can be very expensive and quite frankly unnecessary for most services,
BUT in sour gas services (H2S is present) the weld is susceptible to stress corrosion cracking and would be a weak-point. Therefore, it is very advantageous
to utilize this new technology for these specific services.
One-piece impellers can also be useful in very high flow/heavy gas
(MW > 40) applications such as large refrigeration compressors in LNG or
135
B.P. 3.8
More Best Practices for Rotating Equipment
Ethylene plants where the stress on the impellers (especially the first stage) can
be significantly high.
If going this route, however, be sure that the experience of impellers with at
least 2 years in operation has been checked on the vendor or sub-vendor who is
machining the 1 piece impellers.
L.L. 3.8: Use of welded impellers in H2S service has resulted in catastrophic
impeller damage and significant loss of production
Stress corrosion cracking is nearly impossible to indicate during operation
until pieces of an impeller break off and then it is too late (Anti-surge valve
SHOULD go wide open!!!). Obviously this is a catastrophic failure.
If you are operating in a service with H2S without a one-piece impeller, one
thing you can do is put a soft alarm (in BNC System 1) to warn you when the
phase angle shifts significantly (say greater than 30 degree). This will tell you
that the heavy spot on the rotor is changing, and especially if you do not expect
to foul in the application, it can either be stress corrosion cracking or maybe an
impeller slipping on the rotor.
BENCHMARKS
This technology has only been widely available in the last 10 years and has been
used by the author ever since in all services where H2S is present or in very high
flow and heavy gas applications. It will aid in providing the highest reliability
possible for these tough applications.
SUPPORTING MATERIAL
The Compressor Stage
To begin our discussion, let us observe a typical compressor stage shown in
Fig. 3.8.1.
A compressor stage is defined as one impeller, the stationary inlet and discharge passages known as the inlet guide vanes and the diffuser respectively and
the seals, namely the eye labyrinth seal and the shaft labyrinth seal. Each compressor stage at a given flow and impeller speed will produce a certain amount
of head (energy) and have a specific stage efficiency. It can be observed that any
dynamic curve (turbo-compressor or pump) has the characteristic of producing
increased energy only at a lower fluid flow assuming the inlet speed and the inlet
gas angle are constant.
Before we continue, a few important facts and relationships need to be presented. These relationships are: the definition of a vector, tip speed, flow as a
function of velocity and flow related to conditions and the concept of actual
flow. In addition, there is the important concept of an equivalent orifice. Refer
to Fig. 3.8.2.
136
Compressors
Chapter | 3
FIGURE 3.8.1 The compressor stage and characteristic curve.
FIGURE 3.8.2 Reduce it to an equivalent orifice.
Given any impeller configuration, specific areas can be reduced to equivalent orifices. The eye or inlet area, the discharge area between any two vanes,
the eye seal and the hub or shaft seal.
This concept makes it much easier to understand that for a give area gas flow
will change directly proportional to the differential pressure and the compressor
stage. It can be seen in Fig. 3.8.2 that there is an optimum design velocity for
the inlet of the impeller and the discharge of the impeller. These velocities are
controlled by selection of a proper inlet eye area and discharge area based on the
impeller flow requirement. Again, the concept of an equivalent orifice is helpful
to understand that the gas velocity is dependent on the geometry of the specific
impeller. Also, the process system can be reduced to a simple orifice. In any
process the suction side of the process and the discharge side of the process can
137
B.P. 3.8
More Best Practices for Rotating Equipment
TABLE 3.8.1 Facts and Relationships
• A vector describes magnitude and direction →
Tip speed V =
DN  (D )(N ) 

 US units
19,108  229 
Flow related to velocity Q = AV [‘Q’ = (A)(V)(60)]
• Flow related to conditions
(compressible flow)
QF − Ql ×
Pl Tf Z f
× ×
Pf Tl Z l
Where U = Tip velocity (m/s or ft./s)
f = Final condition
D = Diameter (mm or in2)
l = Initial condition
N = Speed (rpm)
P = Pressure (kPa or PSIA)
Q = Flow rate (m3/h or ft3/min)
T = Temperature (°K or °R)
°K = °C + 273
A = Area (m2 or ft.2)
°R = °F + 460
V = Velocity (m/s or ft./s)
Z = Compressibility
be conceived as an orifice placed at the inlet and discharge of the compressor
flanges for a given flow condition.
Table 3.8.1 presents the definitions of facts and relationships necessary for
the discussion that follows. Please note that the relationships presented are in
British Units, metric units are not presented in this section, but can be easily
derived referring to appropriate conversion tables.
Impeller With Side Plate Removed
To begin our discussion, assume that we are operating at the impeller design
point (as shown in Fig. 3.8.1) and that we have removed the side plate of the impeller and are examining the flow between any two vanes. Typical impellers are
shown in Fig. 3.8.3 and the schematic of any impeller for our purposes showing
the upper half of the impeller with the side plate removed is shown in Fig. 3.8.4.
In Fig. 3.8.4 we can see that only two velocities need to be considered to
properly describe the generation of head (energy). At the tips of the vanes there
are two velocities that are present. The blade tip velocity, identified as U and
the velocity relative to the blade identified as VREL. The blade tip velocity is the
function of the diameter of the blade and the blade rotational speed. The velocity relative to the blade (VREL) is a function of the area between the blades, the
flow rate at that location and the angle of the blade at the discharge of the impeller. Summing these two velocities, the resultant or absolute velocity defines the
magnitude and the direction of the gas as it exits the blade. For this discussion,
we assume that the velocity relative to the blade exactly follows the blade angle,
138
Compressors
Chapter | 3
FIGURE 3.8.3 Typical impellers. (Courtesy of IMO Industries, Inc.)
that is the slip is equal is zero. This assumption can be used since it will not
impact the final conclusion of our discussion.
Impeller Discharge Velocities
If we now resolve the absolute velocity noted in Fig. 3.8.4 (R) into x and y components, the x-axis projection of the component is the tangential velocity of the
FIGURE 3.8.4 Impeller with side plate removed.
139
B.P. 3.8
More Best Practices for Rotating Equipment
FIGURE 3.8.5 Impeller with side plate removed.
gas at the impeller discharge (refer to Fig. 3.8.5). Eulers’ energy equation states
“The energy created by any turbo machine is proportional to the product of the
tip speed and the tangential velocity.”
Let us now assume that the head required by the process changes such that the
flow VREL through the impeller reduces. Referring to Fig. 3.8.6 let us again examine the discharge velocity to see what happens at this reduced flow condition.
Assuming that the rotor speed is constant, it can be seen that the tip speed
value does not change since tip speed is a function of impeller diameter and
FIGURE 3.8.6 Impeller with side plate removed.
140
Compressors
Chapter | 3
FIGURE 3.8.7 Impeller with side plate removed.
shaft speed. However, the velocity relative to the blades (VREL) will be reduced
as a result of a lower volume flow passing through a fixed area, resulting in a
low velocity relative to the blade at the discharge. If we again sum the velocity
vectors to obtain the absolute velocity R (refer to Fig. 3.8.7), we can see that
the angle of the gas exiting the blade is significantly reduced and the X projection of the tangential velocity will be greater than the previous value (refer to
Fig. 3.8.8).
FIGURE 3.8.8 Impeller with side plate removed.
141
B.P. 3.8
More Best Practices for Rotating Equipment
Since the head (energy) produced by the blade is proportional to the tip
speed (unchanged) and the tangential velocity (increased) we can see that the
reduction of flow through the blade has resulted in increased head or energy
imparted to the fluid. Practically, this makes sense, since the slower the gas
proceeds through the vane, the more time it has to pick up energy imparted by
the blades and as a result will increase the energy produced within the impeller.
Therefore, it can be seen for all dynamic blades and impellers which increase
the energy of the fluid by the action of the vane on the fluid can increase fluid
energy only at a lower flow rate, assuming the speed of the impeller and the inlet
angle of the fluid to the blade remain unchanged.
Blading Types
Backward Lean
The previous discussion was focused on the characteristic of a backward leaning vane. Most turbo machinery vanes are backward leaning since they produce
a greater head rise from impeller design point to the low flow operating point.
The low flow limit of operation for centrifugal compressors is known as surge.
Head rise is defined as the head produced by the impeller at the low flow operating point divided by the head produced by the impeller at the impeller design
point. Today, the industry prefers backward leaning impellers with an external
or exit blade angle of approximately 40–50 degree. This blade angle will produce head rises in the range of 5–15% depending on the gas density.
Radial
Radial vanes are used in some older design open type first stage impellers, and
in some modern impellers that operate at a very low flow. Let us now examine
the effect of a radial blade on the performance curve. If we were to design an
impeller with radial blades let us examine again what would happen when we
changed flows from a rated point to a lower flow. At the rated point the blade
tip speed and velocity relative to the blade will be as shown. Refer to Fig. 3.8.9.
Notice that the velocity relative to the blade is completely radial assuming
zero slip and consequently the absolute velocity is the sum of the two vectors.
Again we project the tangential velocity on the x-axis projection from the absolute velocity and note its value as shown in Fig. 3.8.10.
At a lower flow, tip speed will remain constant (assuming constant shaft
speed) and the relative velocity will decrease as in the case of the backward
leaning blade. However, note that since the relative velocity follows the radial
blade path, the magnitude of the tangential velocity remains constant regardless
of the value of relative velocity. This is shown in Fig. 3.8.11.
Since the energy generated by the blade is the product of tip speed (unchanged)
and a tangential velocity (unchanged) the design head (energy) produced in a radial impeller will remain essentially constant. Therefore, the curve shape will be
significantly flatter and will possess much less of a head rise than that of a non
142
Compressors
Chapter | 3
FIGURE 3.8.9 Radial blading.
FIGURE 3.8.10 Impeller with side plate removed.
143
B.P. 3.8
More Best Practices for Rotating Equipment
FIGURE 3.8.11 Impeller with side plate removed.
radial vane. In reality though, the effects of friction will in fact produce a curve
shape that will increase from high flows to low flows but the effects will produce
much less of an energy increase. This value is typically approximately 3% head
rise or less. This is an important fact to remember since the operating point of any
dynamic machine will be the intersection of the head required and the machinery
curve head produced. A characteristic curve with a low head rise will have greater
sensitivity to process changes than a curve with higher head rise.
In summary, it should be noted that the previous discussion can be equally
applied to pump impellers since pumps also operate on a fluid (liquid). One very
important thing to remember from this discussion however is that regardless of
the type of liquid used in pumps, velocity relative to the blade will never change
since the fluid is incompressible. In the case of a turbo-compressor however
this will not be true since the gas is compressible and the velocity relative to
the blades at the discharge will change as a result of pressure and temperature
of that gas at the exit. Therefore, the statement that head (energy) produced by
a compressor impeller will remain constant at a given speed is not totally true.
Having previously discussed performance characteristics of a single compressor stage, we will now examine the effects that multistage compressor configurations have on the overall compressor performance curve.
The stage curve is defined as the curve representing the performance of one
stage, which consists of the impeller, the stage seals, the diffuser, the cross over
and return passage. Each compressor blade row or impeller stage has a specific
performance curve that is plotted as actual volume flow versus head (energy)
and actual volume flow versus efficiency.
144
Compressors
Chapter | 3
The amount of impeller stage head is limited by both mechanical and aerodynamic factors. Mechanical factors include impeller stresses and blade natural
frequencies. Aerodynamic factors are tip speed, optimum efficiency, mach number, and flow range considerations.
A compressor section is defined as the number of stages between turbocompressor casing nozzles. A section can contain one or more stages. Its performance is defined by a section curve. The number of stages per section is limited
by discharge temperature, process gas characteristics, casing and configuration,
and rotor stiffness considerations.
We will examine stage and section performance for an ideal designed compressor case, a case with fouled impellers, and a case with varying molecular weight.
Any multistage compressor is designed on the basis that succeeding impellers will compress a lower volume flow resulting from the compressibility of the
specific gas handled. Remembering that each application is designed for only
one operating point, the compressor designer matches each successive impeller
as closely as possible to achieve operation at the individual impeller design point
such that each individual impeller will operate at its best efficiency point. It is
important to understand that in many applications, operating at the compressor
section rated point does not mean operating at each individual compressor impeller best efficiency (design point). Many older designed multistage compressors were built using specific impeller designs. For any given application, the
succeeding impeller may not be at its optimum efficiency point (design point)
but may be significantly far from that operating point. A good rule of thumb in
selecting a multistage compressor is to assure that when operating at the compressor section rated point each impeller operates at approximately + or − 10%
of its design point. Today (2016) many compressor manufacturers are using
computer-aided design (CAD) to manufacture each impeller individually so
that each impeller operates at its design point and there is a minimum of mismatch between impellers. Therefore, it can be seen that the overall performance
section curve is the composite of the operation of each individual impeller performance curve.
If the performance of one or more of the impeller stages in a compressor section were to deteriorate as a result of fouling or increased labyrinth seal clearances, the overall performance curve will be affected. The amount of this effect
will depend on the performance deterioration of the individual impeller stages.
We have shown a case where the first and second impellers of a three-stage compressor section become fouled. The resulting section operating curve can deteriorate significantly in the terms of head (energy), flow range, and efficiency.
Another case to consider is the change in gas molecular weight. Remembering that once an impeller is designed, its energy production at a given flow point
is essentially fixed, we can see that pressure rise will change as gas composition
changes. It must be remembered that each gas composition requires a different
amount of head (energy) to increase its pressure level to a given amount. An
impeller or blade is designed for only one gas composition and therefore only a
145
B.P. 3.8
More Best Practices for Rotating Equipment
specific amount of energy is designed in a blade or impeller for a given application at a specific flow rate. In a multistage compressor, the effect of gas composition change can have a significant effect on the overall compressor curve. If
for instance, a gas of higher molecular weight were to be handled, the pressure
produced in the first stage of a multistage compressor section would increase.
Since the gas is compressible, this would result in a reduced volume to the second stage. This impeller was initially designed for a higher volume but now will
handle a lower volume. Also, a dynamic blade or impeller will produce higher
energy at a reduced flow, therefore compounding the effect of the molecular
weight increase with increased blade energy capability resulting in a further reduction in flow rate to the succeeding stage. This fact will result in a shifting of
the section curve from the design point toward the surge point and could result
in surge of a compressor with no significant change in system resistance.
The Stage Curve
Most of the dynamic compressors used today are multistage because the head
required by most processes are in excess of the head produced by one stage of a
dynamic compressor. Typical values for one stage of dynamic compression are:
ft − lb f
m − kg f
j
Centrifugal closed impeller 3,050
or 10,000
lb M
kg M
m − kg f
j
Centrifugal semi-open impeller 4,575–7,625
kg M
ft − lb f
or 15,000–25,000
lb M
m − kg f
ft − lb f
j
Axial blade row 915
or 3,000
kg M
lb M
Fig. 3.8.12 shows the components associated with one stage of a centrifugal
closed impeller.
FIGURE 3.8.12 A stage consists of an impeller, stage seals, diffuser (crossover and return
passage—if in a multistage configuration).
146
Compressors
Chapter | 3
TABLE 3.8.2 Factors Limiting Compressor Impeller Stage Head
Mechanical factors
Aerodynamic factors
j
Impeller stresses
j
Tip speed
j
Blade natural frequencies
j
Optimum efficiency
j
Mach number
j
Flow range considerations
For each impeller or blade stage, there exists a specific head versus flow
performance curve. In order to achieve maximum efficiency and flow range in a
multistage dynamic compressor, each stage should be operating at its maximum
efficiency point when the compressor is at its rated (guaranteed) point. This is
not true for many older designs where only a limited number of compressor
impeller or blade designs were available to select from. The amount of head
produced by one stage is limited by certain mechanical and aerodynamic factors. These factors are presented in Table 3.8.2.
As an example, open impellers can produce as much as 250% greater head
than a closed impeller because the impeller stresses at higher speeds are significantly reduced by omitting the impeller cover or side plate.
The Overall Curve
Frequently, there is much confusion concerning the terms compressor stage and
compressor section. Process engineers and operators usually use the term stage
to describe what properly is termed a compressor section. This is probably because process flow diagrams only show a stage and not the individual impellers
in the usual block diagram format. Fig. 3.8.13 defines a compressor section and
shows a typical section performance curve.
In order to properly define compressor performance, a performance curve is
required for each section. Each sectional performance curve is developed from
the individual impeller stage curves. An important fact to remember concerning
compressors is that the design of each succeeding stage is based on the predicted
preceding stage performance. If the preceding stage performance is not as predicted, the next stage will be affected. The greater the number of stages, the
greater this effect will be. It is commonly called mismatching. Any change in
gas density (pressure, molecular weight, temperature, compressibility) will affect the flow into the succeeding stage and thus effect the performance. The most
effective way to minimize the effect of mismatching is to require that each compressor stage operate as close as possible to its design or best efficiency point.
The number of stages per section for dynamic compressors are also limited
by performance and mechanical factors (refer to Table 3.8.3).
147
B.P. 3.8
More Best Practices for Rotating Equipment
FIGURE 3.8.13 Compressor section-definition/performance curve.
TABLE 3.8.3 The Number of Stages Per Section—Limiting Factors
j
j
j
j
Discharge temperature
Process gas characteristics
Casing configuration
Rotor stiffness
Some processes can cause accumulation of solid materials (fouling) within
the compressor impellers or stationary passages. Such a phenomena is usually
temperature related and can influence the number of intercooled sections for a
given application. Also, a large number of stages on a single rotor can reduce
the rotor stiffness and thus reduce the natural frequency (critical speed) of the
compressor.
Determining Section Performance
As previously discussed, the performance curve for any section is derived from
the individual stage curves of each impeller in that section. Fig. 3.8.14 shows an
example of a three-stage nitrogen compressor section.
Note how the inlet volume flow to each successive section is reduced
and how it depends on the impeller head produced by the proceeding stage.
Fig. 3.8.15 shows the effect of fouling the first and second stage impellers in the
same compressor.
Note how this affects the operating points on the third stage:  = original
operating point, ○ = new operating point in fouled condition. The same affect would occur if the interstage labyrinth clearance increased from erosion or
148
Compressors
Chapter | 3
FIGURE 3.8.14 Stage and section performance.
FIGURE 3.8.15 Stage and section performance (1st and 2nd impellers fouled).
149
B.P. 3.8
More Best Practices for Rotating Equipment
FIGURE 3.8.16 Stage and section performance (M.W. varies from 24–32).
vibration on the first and second stages. Any reduction of head produced in a
preceding stage will increase the volume flow in a succeeding stage and reduce
the overall head produced by a compressor section.
In Fig. 3.8.16, the effect of changing the molecular weight in the same compressor is shown.
Note that for molecular weight changes of less than 20%, the head produced by a dynamic compressor stage does not significantly change. The greater
the density of a gas, the greater the discharge pressure and the closer the surge
point to the design point. This is because once any dynamic compressor stage
is designed, the head (energy) produced for a given flow and speed is fixed.
The greater the density of the gas (proportional to molecular weight), the higher
the pressure produced and the lower the volume flow. Flow is inversely proportional to pressure (Boyles Law). Since surge is caused by low flow, a dynamic
compressor handling a denser gas will surge sooner.
In the case of reduced gas density, the opposite effect will occur. The discharge pressure will be reduced and the surge point will move to the left, farther
from the design point.
This concludes the chapter concerning individual stage and overall performance. One final comment. Individual stage performance curves are vendor
propriety information and are not reproduced. The only opportunity that an end
user has to review these curves are:
j
j
Prior to an order
During a design audit
150
Compressors
Chapter | 3
Readers are encouraged to review this information during the bid phase of a
contract to determine if all individual impeller operating points when operating
at the guaranteed point (rated) are as close as possible to the individual impeller
best efficiency points.
B.P. 3.9: Always require two pressure and temperature transmitters in the
same plane at the inlet and discharge of each compressor section (for both
between bearing and integral gear compressors)
One of the most common and sometimes frustrating events with trending
compressor performance is at initial start-up. The unit starts up and we calculate
the Head and Efficiency of each section at the operating flow and we find that
the machine is already 5% low in efficiency. Obviously this can’t be correct,
right? Right!!! The fact is that all Factory Acceptance Tests (FAT’s) are performed with four pressure and temperature transmitters at each inlet and discharge in the same plane and averaged to give an accurate gradient to calculate
head and efficiency and that is what is plotted on the performance curve. While
impractical for the field, having two of each transmitter in the same plane will
increase the field accuracy dramatically and give us a piece of mind and a very
accurate baseline to start trending performance.
Another benefit is that sometimes one of the transmitters can go bad and
rather than flying blind until turnaround, at least we can still trend performance.
While there is an added upfront cost of doubling the transmitters, the benefits of accurately knowing your performance at all times should outweigh that
cost.
L.L. 3.9: Having only one pressure and temperature transmitter at each
inlet and outlet may give inaccurate
Although the proper way to trend performance is to start when the compressor is in good condition (baseline after initial start-up or after turnaround),
many times the continuous trending is not available at that time and a manual
calculation was not performed in the heat of start-up. Then performance may
be checked a few months or even later for the first time and may show that the
efficiency is 10% off of the curve. This could indicate wear in the laby’s or possibly fouling, however if one pressure and temperature transmitters are installed
at each flange, this is within the margin of error. Therefore, it is at great risk to
try to state that this compressor is underperforming.
BENCHMARKS
The writer has used this best practice for the last 10 years since daily production
rates have skyrocketed and it is essential to keep units online as long as possible.
Compressors can take close to two weeks to perform a complete overhaul on
and if the rotor looks fine when you do this, you can lose upwards of $5MM per
day times 14 days of production.
151
B.P. 3.9
More Best Practices for Rotating Equipment
SUPPORTING MATERIAL
The Major Machinery Components
Think of all the machinery that you have been associated with and ask … What
are the major components and systems that are common to all types of rotating
equipment?
Table 3.9.1 presents the major component classifications for any type of
machinery:
Pumps
Steam turbines
Compressors
Motors
Gas turbines
Fans, etc.
j
j
j
j
j
j
Regardless of the type of machinery, monitor these components and you will
know the total condition of the machine.
Component Condition Monitoring
As previously stated, component and system functions must first be defined and the
normal values for each component listed. These facts are presented in Table 3.9.2.
Once the function of each component is defined, each major machinery
component can be monitored as shown in Table 3.9.3.
Baseline
Having defined all condition parameters that must be monitored, the next step in
a condition monitoring exercise is to obtain baseline information. It is important
TABLE 3.9.1 Major Machinery Components and Systems
j
j
j
j
j
Rotor
Radial bearing
Thrust bearing
Seal
Auxiliary systems
TABLE 3.9.2 Component and System Functions
j
j
j
152
Define the function of each affected component
Define the system in which each affected component operates
List the normal parameters for each affected component and system component
Compressors
Chapter | 3
TABLE 3.9.3 Component Condition Monitoring
j
j
j
j
j
Define each major component
List condition monitoring parameters
Obtain baseline data
Trend data
Establish threshold limits
TABLE 3.9.4 Base Line Condition
If you don’t know where you started, you do not know where you are going!
to obtain baseline information as soon as physically possible after start-up of
equipment. However, operations should be consulted to confirm when the unit
is operating at rated or lined out conditions. Obtaining baseline information
without conferring with operations is not suggested since misinformation could
be obtained and thus lead to erroneous conclusions in predictive maintenance
(PDM). Table 3.9.4 states the basics of a baseline condition.
It is amazing to us how many times baseline conditions are ignored. Please
remember Table 3.9.4 and make it a practice to obtain baseline conditions as
soon as possible after start-up.
Trending
Trending is simply the practice of monitoring parameter condition with time.
Trending begins with baseline condition and will continue until equipment
shutdown. In modern day thought, it is often conjectured that trending must be
performed by micro-processors and sophisticated control systems. This is not
necessary! Effective trending can be obtained by periodic manual observation
of equipment or using equipment available to us in the plant, which will include
DCS systems, etc. The important fact is to obtain the baseline and trends of data
on a periodic basis. When trending data, threshold points should also be defined
for each parameter that is trended. This means that when the parameter preestablished value is exceeded action must be taken regarding problem analysis.
Setting threshold values a standard percentage above normal value is recommended. Typically values are on the order of 25–30% above baseline values.
However, these values must be defined for each component based on experience. Fig. 3.9.1 presents trending data for a hydrodynamic journal bearing. All
of the parameters noted in Fig. 3.9.1 should be monitored to define the condition
of this journal bearing.
153
B.P. 3.9
More Best Practices for Rotating Equipment
FIGURE 3.9.1 Trending data.
Specific Machinery Component and System Monitoring
Parameters and Their Limits
On the following pages is contained information concerning what parameters
should be monitored for each major machinery component to determine its condition. In addition, typical limits are noted for each component.
These limits represent the approximate point at which action should be
planned for maintenance. They are not intended to define shutdown values.
The Rotor
Rotor condition defines the performance condition (energy and efficiency) of
the machine. Table 3.9.5 presents this value for a pump.
Radial Bearings
Tables 3.9.6 and 3.9.7 present the facts concerning anti-friction and hydrodynamic (sleeve) radial or journal bearing condition monitoring.
Thrust Bearings
Tables 3.9.8 and 3.9.9 show condition parameters and their limits for anti-friction and hydrodynamic thrust bearings.
154
Compressors
Chapter | 3
TABLE 3.9.5 Pump Performance Monitoring
1. Take value at minimum flow (shut off discharge valve)
2. Measure:
j
P1
j
Driver bhp
j
P2
j
Specific gravity
Where, P1 and P2 = psig, bhp = brake horsepower.
1. Calculate:
A. Head produced m-Kgp-  ft -lbr  ∆P(kPa) × .102  ∆P × 2.311

Kgm  lbm 
S.G.
S.G. 
m3
h ×S.G.  hd × gpm × SG 
B. Pump efficiency (%) =
 3960 × bhp 
360 × kW
hd ×
2. Compare to previous value if > –10% perform maintenance
TABLE 3.9.6 Condition Monitoring Parameters and Their Alarm Limits—
Journal Bearing (Anti-Friction)
Journal bearing (anti-friction)
Parameter
Limits
1. Bearing housing vibration (peak)
0.4 inch/s (10 mm/s)
2. Bearing housing temperature
180°F (85°C)
3. Lube oil viscosity
Off spec 50%
4. Lube oil particle size
j
Non metallic
25 µm
j
Metallic
Any magnetic particle in the sump
5. Lube oil water content
Below 200 ppm
Seals
Tables 3.9.8 and 3.9.9 3.9.10 presents condition parameters and their limits for
a pump liquid mechanical seal.
Auxiliary Systems
Condition monitoring parameters and their alarm limits are defined in
Table 3.9.11 for lube systems.
155
B.P. 3.9
More Best Practices for Rotating Equipment
TABLE 3.9.7 Condition Monitoring Parameters and Their Alarm Limits—
Journal Bearing (Hydrodynamic)
Journal bearing (hydrodynamic)
Parameter
Limits
1. Radial vibration (peak to peak)
2.5 mils (60 microns)
2. Bearing pad temperature
220°F (108°C)
3. Radial shaft positiona
>30° change and/or 30% position change
4. Lube oil supply temperature
140°F (60°C)
5. Lube oil drain temperature
190°F (90°C)
6. Lube oil viscosity
Off spec 50%
7. Lube oil particle size
>25 microns
8. Lube oil water content
Below 200 ppm
a
Except for gearboxes where greater values are normal from unloaded to loaded.
TABLE 3.9.8 Condition Monitoring Parameters and Their Alarm Limits—
Thrust Bearing (Anti-Friction)
Thrust bearing (anti-friction)
Parameter
Limits
1. Bearing housing vibration (peak)
j
Radial
0.4 in/s (10 mm/s)
j
Axial
0.3 in/s (1 mm/s)
2. Bearing housing temperature
185°F (85°C)
3. Lube oil viscosity
Off spec 50%
4. Lube oil particle size
j
Non metallic
>25 µm
j
Metallic
Any magnetic particles with sump
5. Lube oil water content
Below 200 ppm
Tables 3.9.12–3.9.14 present condition monitoring parameters and limits for
dynamic compressor performance, liquid seals and seal oil systems. One final
recommendation is presented in Table 3.9.12.
Predictive Maintenance Techniques
Now that the component condition monitoring parameters and their limits have
been presented, PDM techniques must be used if typical condition limits are
156
Compressors
Chapter | 3
TABLE 3.9.9 Condition Monitoring Parameters and Their Alarm Limits—
Thrust Bearing (Hydrodynamic)
Thrust bearing (hydrodynamic)
Parameter
Limits
1. Axial displacementa
>15–20 mils (0.4–0.5 mm)
2. Thrust pad temperature
220°F (105°C)
3. Lube oil supply temperature
140°F (60°C)
4. Lube oil drain temperature
190°F (90°C)
5. Lube oil viscosity
Off spec 50%
6. Lube oil particle size
>25 µm
7. Lube oil water content
Below 200 ppm
a
Thrust pad temperatures >220°F (105°C).
TABLE 3.9.10 Condition Monitoring Parameters and Their Alarm Limits—
Pump Liquid Mechanical Seal
Pump liquid mechanical seal
Parameter
Limits
1. Seal Chamber Pressure
>50 psia above Vapor Pressure at PT (3.5 bar)
2. Stuffing box temperature
Below boiling temperature for process liquid
TABLE 3.9.11 Condition Monitoring Parameters and Their Alarm Limits—
Lube Oil Systems
Lube oil systems
Parameters
Limits
1. Oil viscosity
Off spec 50%
2. Lube oil water content
Below 200 ppm
3. Auxiliary oil pump operating yes/no
Operating
4. Bypass valve position (P.D. pumps)
Change > 20%
5. Temperature control valve position
Closed, supply temperature > 130 (55°C)
6. Filter ∆P
>25 psid (170 kpag)
7. Lube oil supply valve position
Change > +/–20%
157
B.P. 3.9
More Best Practices for Rotating Equipment
TABLE 3.9.12 Compressor Performance Condition Monitoring
1. Calibrated: pressure and temperature gauges and flow meter
2. Know gas analysis and calculate k, z, m.w
3. Perform as close to rated speed and flow as possible
4. Relationships:
A.
C.
N −1
=
N
HEADPOLY
(T2 )
(T1)
(P )
LN 2
(P1)
LN
B. EFFICIENY poly
k −1
= k
n −1
n
n −1


 P2  n
n
m kgf 847.4  Ft-lbf  1545

− 1
=
× T1 ×
× Z aug ×  
=
=


  P1 

n −1
kgm
mw  Lbm  MW


5. Compare to previous value, if decreasing trend exists greater than 10%, inspect at
first opportunity
TABLE 3.9.13 Condition Monitoring Parameters and Their Alarm Limits—
Compressor Liquid Seal
Compressor liquid seal
Parameter
Limits
1. Gas side seal oil/gas ∆P
j
bushing
<12 ft. (3.5 m)
j
mechanical contact
<20 psi (140 kpa)
2. Atmospheric bushing oil drain temperature
200°F (95°C)
3. Seal oil valvea position
>25% position change
4. Gas side seal oil leakage
>20 gpd per seal
Note this assumes compressor reference gas pressure stays constant.
a
Supply valve, +25%; return valve, –25%.
exceeded. The following chapter will address the techniques used for PDM
analysis and root cause analysis techniques.
Now that the principles of turbo-compressor performance have been explained and hopefully understood, they can be implemented to observe internal turbo-compressor condition changes. Always remember that … we want to
know the internal, not the external condition. Table 3.9.15 presents the outline
of a case history that will show the value of performance condition monitoring.
158
Compressors
Chapter | 3
TABLE 3.9.14 Condition Monitoring Parameters and Their Alarm Limits—
Compressor Liquid Seal Oil Systems
Compressor liquid seal oil systems
Parameters
Limits
1. Oil Viscosity
Off spec 50%
2. Oil flash point
Below 200°F (100°C)
3. Auxiliary oil pump operating yes/no
Operating
4. Bypass valve position (P.D. Pumps)
Change > 20%
5. Temperature control valve position
Closed, supply temperature 130°F (55°C)
6. Filter ∆P
25 psid (170 kpag)
7. Seal oil valve position
Change > 20% open (supply)
>20% closed (return)
8. Seal oil drainer condition
(Proper operation)
9. constant level (yes/no)
Level should be observed
10. observed level (yes/no)
Level should not be constant
11. time between drains
Approximately 1 h (depends on drainer
volume)
TABLE 3.9.15 “The Long, Long, Long, Turnaround”
j
j
j
The first plan
The second plan
The third plan
The First Plan
I visited a refinery a few years ago to troubleshoot an existing turbo-compressor
problem. While I was on site, another process unit, a reformer, was scheduled
for a turnaround. Since I was already on site, I was invited to the pre-turnaround
meeting and became involved with turnaround activities.
During this meeting I learned that the recycle compressor was scheduled for
a bearing inspection only (radial and thrust). I asked why. The answer was that it
was normal practice. I asked if I could see the bearing condition monitoring data
(vibration, bearing displacement, bearing temperature, oil flow—valve position
and oil sample). I was shown to a room and told—“It’s in there.” In desperation
I concluded—“Oh, what the hey, we usually inspect bearings anyway.”
159
B.P. 3.9
More Best Practices for Rotating Equipment
The Second Plan
Well, as you might expect, the bearings were removed and they were pretty
badly damaged. Wipes on both journal and thrust bearing pads and indications
of particle rubs showed that we really needed to both inspect and flush the oil
system. It was later learned that an oil system accumulator had a continuous
nitrogen purge that was at a higher pressure than the oil system and, oh yes …
the bladder had a hole in it! Now back to the bearings … we didn’t know the
condition or trends of parameters. Were there upsets? High vibrations, temperatures, etc. Were there a lot of surges? Nobody could seem to remember! …
sound familiar? Well, as you might expect, we prepared a second plan—better
inspect the oil seals. After all, we didn’t have to get into the compressor so we
had sufficient time.
The Third Plan
No doubt about it, the seals were really bad. Bushing seals, both the atmospheric bushings and gas side bushings were wiped and the atmospheric side
showed typical evidence of high ∆T … probably due to start-up on low pressure
N2 (to save N2 costs). Since the gas was sweet, seal leakage was returned to
the reservoir via a degassing tank and there was no seal condition monitoring
trend data available. (Seal oil leakage—gas side, seal oil leakage—atmospheric
side, which could have been obtained by trending seal oil valve position). Am I
making my point?? Well, it was crunch time (decision time). Were the damaged
seals the root cause of bearing failure? Was the dirty lube and seal system or was
there another deeper cause inside of the compressor? I could go on and on but in
the interest of time let it suffice to say:
j
j
j
j
We decided to open the barrel and compressor—we had time, based on
schedule
We could not find tools and when we did, they did not work
Oh yes, a piece of a suction strainer (supposedly only for start-up) had been
lodged in the case/barrel interface for years.
When we finally got the compressor apart—2 days after the turnaround was
complete, we got lucky—the internals including the balance drum were
perfect!
If only we had established a performance condition monitoring program, as well as seal, bearing, balance drum, and lube/seal system condition
monitoring.
Incidentally, this was a major refinery that placed high value on reliability,
maintainability, root cause analysis, etc. They had all the books and had sent
people to the right workshops … even one of mine!
The objectives of turbo-compressor condition monitoring are presented in
Table 3.9.16.
160
Compressors
Chapter | 3
TABLE 3.9.16 The Objectives
Know turbo-compressor internal condition to determine:
j Loss of daily revenue
j Justification for turnaround activities
j Root cause
The Parameters
What parameters must be measured? Readers should be able to answer this
question readily at this point. Table 3.9.17 presents the reduced parameters—the
answers and Table 3.9.18 contains all the factors necessary for calculation.
Accuracy
Accuracy of data involves both calibration and location of instruments. Before
proceeding, we need to present some important facts concerning condition
monitoring. These facts are presented in Table 3.9.19.
TABLE 3.9.17 The Turbo-Compressor Performance Condition Parameters
j
j
j
j
j
j
Polytropic (or isentropic) head
Actual inlet flow rate
Polytropic (or isentropic) efficiency
Polytropic exponent
Horsepower
Speed (or guide vane position)
TABLE 3.9.18 The Data (Factors) Required
j
j
j
j
j
j
j
j
j
From gas analysis and equation of state
M.W.
K average
Z inlet
P1, P2
T1, T2
Flow rate
Speed (if applicable)
Guide vane position (if applicable)
161
B.P. 3.9
More Best Practices for Rotating Equipment
TABLE 3.9.19 Turbo-Compressor Performance Condition Monitoring Facts
j
j
Specific data (to confirm field guarantees) requires pre-planning to assure accuracy
Trends produce relative change in values
TABLE 3.9.20 Instrument Calibration Facts
All performance condition monitoring instruments must have known, accurate calibration values!
If:
j
j
j
∆T as measured is off 5%, efficiency is affected 20%!
Inaccurate gas analysis of 5% can affect efficiency 20%
Pressure gauge inaccuracy has much less of an affect on efficiency!
TABLE 3.9.21 Gas Analysis Guidelines
j
j
j
j
Take samples from the top of pipe
Measure gas temperature
Be sure to analyze gas at temperatures equal or greater than field conditions
Confirm laboratory experience and methods
Table 3.9.20 presents important facts concerning instrument calibration.
Inaccurate gas analysis procedures can produce some pretty wild results!
Efficiencies that exceed 100%! Table 3.9.21 presents guidelines for accurate
gas analysis.
Location and number of field instruments are just as important as instrument calibration. I cannot overemphasize the importance of this fact. Convince
management (plant management first, then project management) to pre-invest
in turbo-compressor performance and instrumentation. If you don’t accurately know what’s happening inside the patient, you can’t effectively prevent
problems! … That old medical analogy to a turbo-compressor again! Fig. 3.9.2
presents some guidelines from the ASME Turbo-compressor Test Code
(PTC-10) regarding the number and location of instruments.
The Field is Not a Laboratory!
Field turbo-compressor condition monitoring must be planned such that it
does not impact production rates. Never change operation to check performance but rather check performance at a given operating point and compare
to the curve.
162
Compressors
Chapter | 3
FIGURE 3.9.2 Performance instrumentation location guidelines (typical for multi-stage
compressor). (Courtesy of ASME PTC-10)
Establish a Baseline
Before we discuss trending or fully understand turbo-compressor performance
condition monitoring, the baseline condition must be defined. Since performance
condition monitoring, as well as any type of condition monitoring, is concerned
with relative change, the starting point must be established. Table 3.9.22 defines
the baseline condition.
I can’t remember how many times I have said … “I sure wish we had established the ‘baseline condition.”
163
B.P. 3.9
More Best Practices for Rotating Equipment
TABLE 3.9.22 The Baseline Condition
j
The baseline condition establishes the initial (‘new’ or ‘rebuilt’) condition from
which all changes (trends) are measured.
The baseline turbo-compressor performance condition should be established
once the process unit is on spec, immediately, not one day, week, month or year
later. The sooner the baseline is established the more data in your machine historical file. Many problems occur in the initial period of operation, be sure you
have recorded the data. The baseline condition then represents the first point
on the trend graph of any measured parameter. Table 3.9.23 recommends when
baseline conditions should be taken.
Trending
Trending requires that a parameter be monitored over a period of time to determine if significant change occurs. These factors are defined in Table 3.9.24.
It should be noted that significant change is not synonymous with alarm
point. If an initial value is small, a small change can be significant and still be
far away from the set alarm point!
Details concerning field performance testing will be discussed in the next
chapter. We have presented suggested turbo-compressor performance trend
parameters in the final figure of this chapter … Table 3.9.25.
TABLE 3.9.23 Baseline Performance Conditions—When?
Baseline performance conditions should be established when the process is on spec.
immediately after:
j New compressor start-up
j Turnaround (if internals, or thrust bearing was changed)
TABLE 3.9.24 Trending Guidelines
j
j
j
j
164
Establish baseline condition
Trend parameter (Y axis) versus time (X axis)
Compare related trends with same time scale (X axis)
A significant change in a parameter is a 25–50% change.
Compressors
Chapter | 3
TABLE 3.9.25 Useful Trend Parameters
Polytropic head ÷ speed2 (HdPOLY/N2)
Actual flow ÷ speed (Q/N)
j Polytropic efficiency (ηPOLY)
In certain cases,
j Molecular weight
j Inlet temperature
j
j
B.P. 3.10: Always check tilting pad thrust bearing clearance using a
hydraulic jack and set alarm and trip values based on this clearance
This is the only way to apply a load on the pads equal to what it is during
operation. You will most likely need to contact the vendor (who will gladly give
you this information) to obtain the design thrust bearing load at rated conditions
in order to know what pressure to apply the load with the hydraulic jack.
When you use the hydraulic jack to check the clearance at thrust load at
rated operating point, you can see a few thousandths more clearance than if it
was done manually (without aid of hydraulic jack). If this takes the clearance
out of recommended setting, then be sure to adjust the clearance with a shim
as necessary.
The reason for the extra clearance at higher loads (operating loads) is that
the number of components in the thrust bearing assembly will compress very
slightly and give you some extra clearance before wearing the bearing. Therefore, it is essential to use the clearance you obtained with the hydraulic jack (at
correct pressure setting) in order to set the alarm and trip set-points.
Note that it is key to push in both directions with the hydraulic jack when
performing this best practice in order to confirm the total clearance when the
assembly in each direction slightly compresses. In order to achieve this, it may
be necessary to furnish brackets to fit on either side of the compressor in order
to retain the hydraulic jack while pushing the rotor.
L.L. 3.10: Failure to use a hydraulic jack during the setting of thrust clearance has resulted in premature thrust alarms and trips
Many times it has been seen in the field where machines with tilting pad
thrust bearings are operating in alarm on axial displacement and very close to
the trip setting, while thrust pad temperatures have not increased and are well
below 100°C (212°F). This indicates that the thrust pads are not wearing since
no heat is being generated and that the assembly is most likely compressing
a bit.
This has resulted in a number of machines that have actually tripped on
axial displacement only to check or replace the thrust bearing and find that the
same thing is happening again after start-up and a few days have been lost in
production.
165
B.P. 3.10
More Best Practices for Rotating Equipment
The issues tend to arise more on smaller equipment where the rotors can be
moved relatively easily by Pry-bar and the thought is that a hydraulic jack is not
needed, however the BIGGEST maintenance mechanic in the plant still cannot
exert the operating load the thrust bearing will see.
BENCHMARKS
This best practice has been used since the early 1990s when there were a number
of instances discovered in the field where users were operating close to trip on
axial displacement without any increase in pad temperatures. It has saved a number of unnecessary trips and a lot of unnecessary maintenance and days offline.
SUPPORTING MATERIAL
In every rotating machine utilizing reaction type blading, a significant thrust is
developed across the rotor by the action of the impellers or blades. Also in the
case of equipment incorporating higher than atmospheric suction pressure, a
thrust force is exerted in the axial direction as a result of the pressure differential
between the pressure in the case and atmospheric pressure.
In this section we will cover a specific rotor thrust example and calculate
thrust balance for a specific case. We will see the necessity in some applications
of employing an axial force balance device known as a balance drum. In many
instances, the absence of this device will result in excessive axial (thrust) bearing
loadings. For the case of a machine with a balance device, the maintenance of
the clearances on this device are of utmost importance. In many older designs the
clearances are maintained by a fixed close clearance bushing made out of babbitt
which has a melting temperature of approximately 175°C (350°F), depending on
the pressure differential across the balance drum. If the temperature in this region
should exceed this value, the effectiveness of the balance drum would suddenly
be lost and catastrophic failures can occur inside the machine. Understanding the
function of this device and the potential high axial forces involved in its absence
is a very important aspect of condition monitoring of turbo-compressors.
We will also examine various machine configurations including natural balanced (opposed) thrust and see how thrust values change even in the case of a
balanced machine as a function of machine flow rate.
Finally, we will examine thrust system condition monitoring and discuss
some of the confusion that results with monitoring these machines.
The Hydrodynamic Thrust Bearing
A typical hydrodynamic double acting thrust bearing is pictured in Fig. 3.10.1.
The thrust bearing assembly consists of a thrust collar mounted on the rotor
and two sets of thrust pads (usually identical in capacity) supported by a base
ring (Michell Type).
166
Compressors
Chapter | 3
FIGURE 3.10.1 Double acting self-equalizing thrust bearing assembly (thrust collar removed). (Courtesy of Elliott Company)
FIGURE 3.10.2 Small Kingsbury six-shoe, two direction thrust bearing. Left-hand group
assembled, except for one shoe and “upper” leveling plate. Right-hand group disassembled. (Courtesy
of Kingsbury, Inc.)
The Kingsbury type includes a set of leveling plates between each set of
pads and the base ring. This design is shown in Fig. 3.10.2.
Both the Michell and Kingsbury types are used. Fig. 3.10.3 provides a view
of the leveling plates providing the self-equalizing feature in the Kingsbury
design. The self-equalizing feature allows the thrust pads to lie in a plane parallel to the thrust collar.
167
B.P. 3.10
More Best Practices for Rotating Equipment
FIGURE 3.10.3 Self-equalizing tilt-pad thrust bearing (View—looking down on assembly).
(Courtesy of Kingsbury, Inc.)
Regardless of the design features, the functions of all thrust bearings are:
j
j
To continuously support all axial loads
To maintain the axial position of the rotor
The first function is accomplished by designing the thrust bearing to provide
sufficient thrust area to absorb all thrust loads without exceeding the support
film (oil) pressure limit (approximately 500 psi).
Fig. 3.10.4 shows what occurs when the support film pressure limit is
exceeded.
The oil film breaks down, thus allowing contact between the steel thrust collar and soft thrust bearing pad overlay (Babbitt). Once this thin layer (1/16 in.) is
worn away, steel to steel contact occurs resulting in significant turbo-compressor
damage.
Thrust pad temperature sensors, located directly behind the babbitt at the
pad maximum load point protect the compressor by tripping the unit before
steel to steel contact can occur.
FIGURE 3.10.4 Evidence of overload on a tilt-pad self-equalizing thrust bearing pad.
(Courtesy of Kingsbury, Inc.)
168
Compressors
Chapter | 3
FIGURE 3.10.5 Thrust bearing rated load versus speed. (Courtesy of Kingsbury Corp.)
Fig. 3.10.5 presents different Kingsbury bearing size rated capacities as a
function of speed.
Fig. 3.10.6 shows how thrust pad temperature and thrust load are related for
a given thrust bearing size and shaft speed. Note that the greater the thrust load
(P.S.I.), the smaller the oil film and the greater the effect of oil viscosity on oil
flow and heat removal. Based on a maximum load of 3448 kPa (500 psi), it can
be seen from Fig. 3.10.6 that a turbo-compressor thrust bearing pad temperature
trip setting should be between 127 and 132°C (260 and 270°F).
Other than to support the rotor in an axial direction, the other function of the
thrust bearing is to continuously maintain the axial position of the rotor. This
is accomplished by locating stainless steel shims between the thrust bearing assembly and compressor axial bearing support plates. The most common thrust
assembly clearance with the thrust shims installed is 0.275–0.35 mm (0.011–
0.014 in.). These values vary with thrust bearing size. The vendor instruction
book must be consulted to determine the proper clearance.
The following procedure is used to assure that the rotor is properly positioned in the axial direction.
1. With thrust shims removed, record total end float by pushing rotor axially in
both directions, typically 6.35–12.7 mm (0.250–0.500 in.).
2. Position rotor as stated in instruction book.
3. Install minimum number of stainless steel thrust shims to limit end float
to specified value. An excessive number of thrust shims act as a spring
169
B.P. 3.10
More Best Practices for Rotating Equipment
FIGURE 3.10.6 The relationship between thrust pad temperature and thrust load. (Courtesy
of Kingsbury, Inc.)
170
Compressors
Chapter | 3
resulting in a greater than specified axial clearance during full thrust load
conditions.
Proper running position of the rotor is critical to obtaining optimum efficiency and preventing axial rubs during transient and upset conditions (start-up,
surge, etc.).
Impeller Thrust Forces
Every reaction type compressor blade set or impeller produces an axial force
toward the suction of the blade or impeller. Refer to Fig. 3.10.7.
In this example, the net force toward the compressor suction is 8900 N
(2000 lbs) for the set of conditions noted. Note that the pressure behind the impeller is essentially constant 344.75 kPa (50 psi), but the pressure on the front
side of impeller varies from 344.75–275.8 kPa (50–40 psi) because of the pressure drop across the eye labyrinth. Every impeller in a multistage compressor
will produce a specific value of axial force toward it’s suction at a specific flow
rate, speed, and gas composition. A change in any or all of these parameters will
produce a corresponding change in impeller thrust.
Rotor Thrust Balance
Fig. 3.10.8 shows how a balance drum or opposed impeller design reduces
thrust force. The total impeller force is the sum of the forces from the individual
FIGURE 3.10.7 Impeller thrust force.
171
B.P. 3.10
More Best Practices for Rotating Equipment
FIGURE 3.10.8 Rotor thrust force.
impellers. If the suction side of the impellers is opposed, as noted in Fig. 3.10.8,
the thrust force will be significantly reduced and can approach 0. If the suction side
of all impellers are the same (in series), the total impeller thrust force can be very
high and may exceed the thrust bearing rating. If this is the case, a balance drum
must be mounted on the rotor as shown in Fig. 3.10.8. The balance drum face
area is varied such that the opposing force generated by the balance drum reduces
the thrust-bearing load to an acceptable value. The opposing thrust force results
from the differential between compressor discharge pressure (PF) and compressor
suction pressure (P1) since the area behind the balance drum is usually referenced
to the suction of the compressor. This is accomplished by a pipe that connects this
chamber to the compressor suction. This line is typically called the ‘balance line’.
It is very important to note that a balance drum is used only where the thrust
bearing does not have sufficient capacity to absorb the total compressor axial
load. And the effectiveness of the balance drum depends directly on the balance
drum seal. Fail the seal, (open clearance significantly) and thrust-bearing failure
can result.
A common misunderstanding associated with balance drum systems is that
a balance drum always reduces the rotor thrust to zero. Refer to Fig. 3.10.9
and observe that this statement may or may not be true depending on the thrust
balance system design. And even if it is, the thrust is zero only at one set of
operating conditions.
172
Compressors
Chapter | 3
FIGURE 3.10.9 Rotor system designed four different ways.
Fig. 3.10.9 shows a rotor system designed four (4) different ways. Note how
the thrust always changes with the flow rate regardless of the design. Another
misconception regarding thrust balance systems is the normal or “active” direction of thrust. In many cases, the active thrust is assumed to always be toward
the suction of the compressor.
Observing Fig. 3.10.9, it is obvious that the “active” direction can change
when the turbo-compressor has a balance drum or is an opposed design. It is
recommended that the use of active thrust be avoided where possible and that
axial displacement monitors be labeled to allow determination of the thrust direction at all times.
Please refer to Fig. 3.10.10, which shows a typical thrust displacement
monitor.
These monitors detect thrust position by targeting the shaft end, thrust collar or other collar on the rotor. Usually two or three probes (multiple voting
arrangement) are provided to eliminate unnecessary compressor trips. The output of the probes is noted on the monitor as either + (normal) or − (counter).
However, this information gives no direct indication of the axial direction of the
thrust collar. The following procedure is recommended:
1. With compressor shutdown, push rotor toward the suction and note direction
of displacement indicator.
2. Label indicator to show direction toward suction of compressor.
Knowing the actual direction of the thrust can be very useful during troubleshooting exercises in determining the root cause of thrust position changes.
173
B.P. 3.10
More Best Practices for Rotating Equipment
FIGURE 3.10.10 Typical axial thrust monitor.
Thrust Condition Monitoring
Failure of a thrust bearing can cause long term and possibly catastrophic damage to a turbo-compressor. Condition monitoring and trending of critical thrust
bearing parameters will optimize turbo-compressor reliability.
The critical thrust bearing condition monitoring parameters are:
j
j
j
Rotor position
Thrust pad temperature
Balance line ∆P
Rotor position is the most common thrust bearing condition parameter and
provides useful information regarding the direction of thrust. It also provides an
indication of thrust load but does not confirm that thrust load is high. Refer to
Fig. 3.10.11.
All axial displacement monitors have pre-set (adjustable) values for alarm
and trip in both thrust directions. Typically, the established procedure is to record the thrust clearance (shims installed) during shutdown and set the alarm
and trip settings as follows:
174
Compressors
Chapter | 3
FIGURE 3.10.11 Typical axial displacement monitor.
Clearance
+ 10 mils (each direction)
2
Trip = Alarm Setting + 5 mils (each direction)
Alarm =
The above procedure assumes the rotor is in the mid or zero position of the
thrust clearance. An alternative method is to hand push the rotor to the assumed
active position and add appropriate values for alarm and trip.
The writer personally recommends the first method since an active direction
of thrust does not have to be assumed.
As noted, axial displacement monitors only indicate the quantity of thrust
load. False indication of alarm or even trip settings can come from:
j
j
j
Compression of thrust bearing components
Thermal expansion of probe adaptors or bearing brackets
Loose probes
It is strongly recommended that any alarm or trip displacement value be confirmed by thrust pad temperature if possible prior to taking action. Please refer
175
B.P. 3.11
More Best Practices for Rotating Equipment
back to Fig. 3.10.6 and note that the thrust pad temperature in the case of thrust
pad overload is approximately 121°C (250°F). If an axial displacement alarm or
trip signal is activated observe the corresponding thrust pad temperature. If it is
below 104°C (220°F), take the following action:
j
Observe thrust pads. If no evidence of high load is observed (pad and back of
pad) confirm calibration of thrust monitor and change settings if necessary.
The last condition monitoring parameter for the thrust system is balance
line pressure drop. An increase of balance line ∆P will indicate increased balance drum seal leakage and will result in higher thrust bearing load. Noting
the baseline ∆P of the balance line and trending this parameter will provide
valuable information as to the root cause of a thrust bearing failure. In many
field case histories, the end user made many thrust bearing replacements until
an excessive balance drum clearance was discovered as the root cause of the
thrust bearing failure. It is a good practice to always check the balance line ∆P
after reported machine surge. Surging will cause high internal gas temperatures,
which can damage the balance drum seal.
B.P. 3.11: Impeller Design Pre-Bid Meeting Guidelines
In Forsthoffer’s Best Practice Handbook for Rotating Machinery, B.P. 3.9
discussed key items to screen for impeller design during the Pre-Bid phase of the
project. This Best Practice expands on that by adding additional items to screen
as well as the format in which this is done. Note that BP 3.9 from Forsthoffer’s
Best Practice Handbook has only been added to so the reader doesn’t have to
switch back and forth between the two books, or even have to own both.
Centrifugal and Axial compressors are custom designed but should incorporate proven impellers or blades and gas path parts.
The success of the Factory Acceptance Test (FAT), field start up and process
life safe and reliable compressor reliability is directly related to impeller, blade
and gas path component integrity.
As a result, vendor requirements should be noted in the invitation to bid
(ITB) to provide the following parameters for each stage (impeller or blade row)
will assure optimum FAT results and field safety/reliability:
l
l
l
l
Flow Coefficient scatter curve experience, showing each proposed impeller
on the same chart as the flow coefficient the vendor has in operation for at
least 2 years in similar applications (similar gas and operating conditions).
Head Coefficient scatter curve experience the same as flow coefficient scatter curve experience is explained above.
Tip Speed Experience again on scatter curve showing the proposed impellers versus actual field experience of minimum of 2 years in similar applications (similar gas and operating conditions).
Impeller Inlet (eye) Mach Number Experience in the same manner as explained for tip speed experience above.
176
Compressors
l
Chapter | 3
Individual Impeller Curves need to be reviewed during pre-bid meeting
(note the vendors will show these to the users if asked, but will not let you
keep them as it is proprietary information) in order to assess the following:
l Head rise to surge for each impeller to be greater than 5% from rated
point
l Individual impeller operating point versus Best Efficiency Point Flow—
The rated operating point should be as close as possible to BEP. For
MW over 40 the rated flow can be within 5% of BEP, MW less than 40
can be within 10%. Note that if the rated operating point is not on the
BEP for each impeller, it is always preferred that it is to the left of the
BEP. This gives a cushion to allow for increasing the flow in the compressor for future operations without worrying about over loading the
impellers.
L.L. 3.11: The Failure to review for impeller/blade experience prior to
vendor acceptance can result in extended FAT time, delayed field start-up
and continuous safety and reliability field issues
Accepting vendor proposed impellers and blades without a review of experience and design details has led to many unexpected surprises during factory
acceptance performance testing and field start-ups.
Once impellers or blade rows are designed and operating, field changes are
difficult, time consuming, will produce revenue loss and difficult to confirm since
field instrumentation does not have the same accuracy as fat instrumentation.
BENCHMARKS
This best practice has been since the mid 1970s to achieve problem free FAT’s,
smooth start-ups and optimum centrifugal and axial compressor safety and reliability (99.7%+).
SUPPORTING MATERIAL
The Factors Involved
The parameters necessary to define a given fluid are presented in Table 3.11.1.
Note that only two parameters are necessary to define a fluid in the liquid state
since it is incompressible. On the other hand, 3 times that number are required
to define that fluid in its vapor state since the vapor is compressible.
Fig. 3.11.1 shows the relationships used to determine the head (energy) required to increase the pressure of a fluid in it’s liquid and vapor state. Note how
much the density of the fluid influences the amount of energy required to meet
a certain process requirement. When one considers that the additional amount
of head produced as a centrifugal compressor’s flow rate decreases from rated
point to surge point is on the order of only 10%, it can be seen that a small
177
B.P. 3.11
More Best Practices for Rotating Equipment
TABLE 3.11.1 What Factors Define a Given Fluid
Liquid (incompressible)
Gas (compressible)
Specific Gravity (S.G.)
Molecular Weight (M.W.)
Viscosity ()
Specific Heat Ratio (K)
Compressibility (Z)
Pressure (P, kPa or PSIA)
Temperature (T, °K or °R)
FIGURE 3.11.1 Fluid head.
change in gas density can result in a significant flow reduction and possibly
compressor surge.
The Effect on Turbo-Compressor Pressure Ratio
The pressure ratio produced by a dynamic compressor is affected by gas density. Table 3.11.2 shows that for a given compressor flow and speed the head
produced by a dynamic compressor is essentially constant. Therefore, any
change in MW, T, K, or Z will change the pressure ratio produced. This information is presented in tabular form for changes in molecular weight and inlet
gas temperature.
178
Compressors
Chapter | 3
TABLE 3.11.2 The Effect of a Gas Composition and Temperature Change on
the Turbo-Compressor Pressure Ratio
The effect of a gas composition and temperature change on the turbo-compressor
pressure ratio
Headi sentropic is related to pressure ratio by:
HDISEN
 K −1 
P K

 1545 
 K 
(Z )  2
−
− 1
(T )
 M.W  1  K − 1
P
 1



Assuming HDI SEN is constant for a given flow,
K

 K −1
P2 
(HDISEN )(M.W ) 
−  1+

 K 
P1 

 (1545) (T1)  K − 1 (Z ) 
Therefore the following table can be developed:
EFFECT OF GAS AND T CHANGES ON PRESS. RATIO
MOLECULAR WGT.
INLET TEMP.
PRESSURE RATIO
INCREASES
CONSTANT
INCREASES
DECREASES
CONSTANT
DECREASES
CONSTANT
INCREASES
DECREASES
CONSTANT
DECREASES
INCREASES
The Effect on the Compressor Head
It is commonly thought that dynamic compressor head produced is always constant for a given flow rate and speed. Fig. 3.11.2 presents this fact for the same
compressor operating on different gases (O2 and N2).
This statement is not true for a fluid in the vapor state since head in a dynamic compressor is produced by blade velocity and gas velocity. Gas velocity
will change will change with gas density since a gas is compressible. These
facts are presented in Table 3.11.3.
Please refer to Fig. 3.11.3 which shows the relationship between gas velocity
(Vrel) blade tip speed (U) and tangential gas velocity in a centrifugal compressor.
Since the head produced by any dynamic impeller is proportional to blade
tip speed and gas tangential velocity, reduced gas velocity through the impeller
(Vrel) will increase the head produced as shown in Fig. 3.11.3. This is the result
of increased gas tangential velocity for a given impeller diameter and speed. As
shown in Table 3.11.3, gas velocity (Vrel) will vary with gas density.
179
B.P. 3.11
More Best Practices for Rotating Equipment
FIGURE 3.11.2 The effect of gas composition change on HEAD.
TABLE 3.11.3 The Effect of a Gas Composition and Temperature Change on
Turbo-Compressor Head
The assumption that compressor head remains constant for a given flow with gas composition and temperature changes is not true because:
j
Head is generated by impeller tip speed and exit velocity relative to the blade
j
Gas composition and temperature changes affect the compression ratio
j
Volume flow rate changes with pressure, temperature and compressibility
j
Since the impeller exit area is fixed, a change in exit volume rate will produce a
change in velocity
Note: For changes on the order of 20%, it is common practice to assume head is constant for a given
flow and speed.
Fig. 3.11.4 presents the effect of gas density changes on impeller produced
head, surge point and choke point. It can be seen that curve shape is influenced
by gas density changes. Therefore, a low-density gas will always have a greater
flow range than a high density gas.
The Effect on System Resistance
Fig. 3.11.5 presents the effect of gas density change on the system resistance
curve. A slight change in the friction drop in pipes, fittings, and vessels results
from a change in gas density.
180
Compressors
Chapter | 3
FIGURE 3.11.3 Head produced α (U)(VT).
FIGURE 3.11.4 Turbo-compressor impeller head change and curve shape summary.
181
B.P. 3.11
More Best Practices for Rotating Equipment
FIGURE 3.11.5 The effect of process changes on the system resistance curve.
The Effect on Turbo-Compressor Flow Rate
The effect of gas density changes on actual mass and standard flow rates is
shown in Table 3.11.4. Note that gas density changes will change the operating
point of each compressor stage in a multistage compressor. Depending on the
impeller selection, this change could have an adverse affect on the operation of
a dynamic compressor causing surge and corresponding high vibration, temperature, flow changes, etc.).
The Effect on Power
As shown in Table 3.11.5, dynamic compressor required power increases directly with gas density up to the choke flow or stonewall region of the performance curve. In the choke flow region, the head produced by the compressor
approaches zero since the gas velocity is equal to its sonic velocity.
Fig. 3.11.6 shows the affect on compressor section performance resulting
from a change in the gas molecular weight. As previously discussed molecular
weight changes can result in compressor stage mismatching which can cause
significant mechanical damage to the compressor train.
TABLE 3.11.4 The Effect of Gas Composition on Turbo-Compressor Flow
Rate
j
The actual volume flow rate will vary as a result of the operating point change
which is the intersection of the turbo-compressor curve (pressure versus flow) and
the system resistance
j
The mass flow rate (lbs/min) will be the product of the new actual volume flow rate
(ft3/min) and the gas density (lb/ft3) at the new gas conditions (M.W., P, T, Z)
j
The standard volume flow rate (SCFM) will be the product of the new actual volume
flow rate (ft.3/min) at its pressure temperature corrected for standard conditions
(14.7 PSIA and 60°F)
182
Compressors
Chapter | 3
TABLE 3.11.5 The Effect on Power
 m − kgf 
 kg 
Head 
× Mass flow  
 hr 
 kgm 
Power (kW) =
+ Mech. losses (kW)
 m − kgf 
3,600 
×
η
(%)
 min − kW 
Power


 Ft − Lb) 
 Lb 
× Mass flow 
Head 


 Lb 
 Min 
(BHP) =
+ Mech. losses (BHP)
−
Ft
Lb




× η (%)
33,000 


 Min − H.P. 
FIGURE 3.11.6 Stage and section performance.
B.P. 3.12: Require bundle removal tooling be used during the performance/mechanical running testing period for barrel type (radially split)
compressors
The most efficient way to accomplish this is by first conducting the performance and mechanical running tests on the spare rotor. Then use the bundle
removal tooling to remove the spare rotor and install the main rotor for its mechanical running test. After a successful test (Hopefully!!), the compressor with
main rotor bundle is ready to ship.
183
B.P. 3.13
More Best Practices for Rotating Equipment
Assure that a maintenance representative who will be responsible for work
on this compressor is available at the test so they can observe the procedure.
Obviously, also make sure that all of the ACTUAL (what is being shipped with
your equipment) tooling required to remove the bundle is being used so you
have no surprises in the field.
If there are any issues with the tooling, it can be fixed before the compressor leaves the shop where the vendor is most equipped to do so. Fixing bundle
tooling issues 5+ years down the road in the field can be very cumbersome and
time consuming.
L.L. 3.12: Failure to test the actual bundle removal tooling in the vendor’s
shop can result in significant delays in the field during a turnaround
Remember, if it is not tested in the vendor’s shop, you will not use the tooling for the first time until the earliest 4 years (most users are trying to go to 5
and now 6 between Turnarounds) and if it is a clean service you may go a few
turnarounds before removing the bundle.
BENCHMARKS
This Best Practice has been implemented since the late 1990s and always assures that the tooling provided is capable in the field. The writer was involved
in a job in 2015 where the bundle removal tooling was used during the FAT and
Mechanical running test bundle removals at the vendor’s shop and an issue with
tooling was discovered. Modifications were made by the vendor without delaying the shipment of the compressor and saving a lot of grief/lost production in
the field when the bundle eventually had to be removed.
B.P. 3.13: Size compressor driver for end of curve power at MCOS when
greater flow = more plant profit
When compressor flow rate is directly proportional to production in the
plant, there are always benefits to have extra power available. End of the
curve flow will vary depending on the system resistance curve (compressor
will only meet what the system allows it to), gas properties and compressor
selection but can be over 20% more flow from the rated point (or significantly
greater) if everything is in your favor. This is only valid, however as long as
there are no bottlenecks elsewhere in the plant preventing you from producing more.
As the compressor operates for a number of years, the efficiency will eventually reduce by some amount either due to fouling (if possible) or opening of
the labys (namely balance drum and impeller eye labys). Also remember that
Gas and Steam Turbines can both foul as well, reducing their power output.
Therefore, another huge advantage of extra power is to assure you can operate
at normal plant rates when these issues occur.
184
Compressors
Chapter | 3
L.L. 3.13: Not having sufficient driver power can result in the inability to
make more profit or loss of production
BENCHMARKS
The writer has used this best practice since the mid 1990s for all drivers when
there are no other bottlenecks in the plant and has always resulted in the user
being able to make significantly more profits due to the extra flow they can process. Depending on the process and how much the rates can be increased, payoff
of the capital cost for more power is generally less than a year.
SUPPORTING MATERIAL
See B.P. 3.3 Supporting Material.
185
Chapter 4
Gears and Couplings
B.P. 4.1: Confirm Gear no-load pressure exerted on bearings. If this value
is less than 50 PSI, modify bearing design to assure shaft vibrations are at
an acceptable value during no load conditions.
When a Compressor train is under load (operating at normal conditions) and
contains a gear, the forces exerted on the gear itself are tremendous. Therefore,
the journal bearings generally contain a large surface area to minimize the pressure on the pads when the train is at normal operating load. Until it gets to a
relatively high loading, the journal bearings typically do not have much load
on them.
During start-up or if the compressor train is operated at minimal load,
there can be excessive vibrations if the load on the bearing is very low as the
rotor will tend to be unstable and basically just bounce around off the different
pads.
If the machine is already in the field, while the ultimate resolution is to have
a proper designed bearing for all loads, the short term fix would be to get up
to a load where the vibrations stop as quick as possible and do not stay below
that for any significant period of time. Note that B.P. 4.4 of Forsthoffer’s Best
Practice Handbook for Rotating Machinery states to limit the stroke of the oil
system backpressure control valve (spillback valve) to allow for more oil to the
train and increase the loading of the gear bearings.
L.L. 4.1: Gears operated at low load without proper bearing design can
experience excessive vibration and potentially trip, causing down time and
loss of production.
Numerous vibration trips have been experienced during start-up or low load
conditions.
BENCHMARKS
This best practice has been in use since 1990 and has aided in maintaining the
highest possible gearbox reliability (above 99.5%).
More Best Practices for Rotating Equipment. http://dx.doi.org/10.1016/B978-0-12-809277-4.00004-8
Copyright © 2017 Elsevier Inc. All rights reserved.
187
B.P. 4.1
More Best Practices for Rotating Equipment
SUPPORTING MATERIAL
Gear Reaction (Bearing) Forces
When considering reaction forces, one must consider the entire gear system
from the gear mesh to the gear foundation. The transmission of torque load
through the gear rotors is shown in Fig. 4.1.1 assuming a speed increaser.
The amount of torque transmitted depends on the operating condition (startup, rated load, off-design load, shutdown, etc.).
Fig. 4.1.2 shows a typical compressor torque versus speed curve.
Note that the start-up condition is always at low load and frequently the
shutdown condition will be at low load (if case is vented on shutdown). This is
an important fact to consider when gear vibration and/or noise is observed at
start-up, shutdown or off-design conditions. Table 4.1.1 presents this important
consideration.
Since the transmitted torque loads will be considerably less, the gear reaction forces will be considerably less and the component stresses and pressures
will be less. This is exactly why gear meshes are noisy on start-up, vibration
increases and bearings can become unstable.
Fig. 4.1.3 shows the reaction force transmission path of a gear radial bearing.
Fig. 4.1.3 shows that the entire gear unit system contributes to the support of
transmitted loads. A change in the load carrying capability of any of the items
noted in Fig. 4.1.3 can result in reduced gear unit reliability.
FIGURE 4.1.1 Gear rotor torque transmission path.
188
Gears and Couplings
Chapter | 4
FIGURE 4.1.2 Speed versus torque curve.
TABLE 4.1.1 Gear Unit Design Basis for Reaction Forces
All gear unit component stresses (pressures)
FORCE
AREA
are designed for rate (maximum) torque loads.
Therefore, loads during off-design conditions can be considerably less
Gear Reaction Forces at Bearings
Since between bearing helical gears are the most common type on site, only this
type will be covered. However, the relations discussed will also apply to internal
and external spur gears with minor modifications. Fig. 4.1.4 shows the reaction
forces present on a helical pinion tooth.
Once the total radial load WRT is known the individual radial bearing forces
can be determined by statics mechanics as shown in Fig. 4.1.5.
189
B.P. 4.1
More Best Practices for Rotating Equipment
FIGURE 4.1.3 Gear radial bearing force transmission path.
The axial load WA is calculated directly as noted in Fig. 4.1.4 and can be
applied to either the gear shaft, pinion shaft, or divided between the gear and
pinion shaft. Most gear designs absorb all thrust on the gear shaft (low speed)
since this usually results in the lowest thrust bearing losses.
Hydrodynamic Bearing Types
Regardless of the type of hydrodynamic bearing, all bearing surfaces are lined
with a soft, surface material made of a composition of tin and lead. This material is known as Babbitt. Its melting temperature is above 200°C (400°F), but
under load will begin to deform at approximately 160°C (320°F). Typical thickness of Babbitt over steel is 1.5 mm (0.060 in.). Bearing embedded temperature
probes are a most effective means of measuring bearing load point temperature
and are inserted just below the Babbitt surface. RTDs or thermocouples can be
used. There are many modifications available to increase the load effectiveness
of hydrodynamic bearings. Among the methods available are:
l
l
l
Copper backed Babbitt or “Trimetal”—to aid in heat removal
Back pad cooling—used on tilt pad bearings to remove heat
Direct cooling—directing cool oil to maximum load points
A typical straight sleeve hydrodynamic journal bearing is shown in Fig. 4.1.6.
Straight sleeve bearings are used for low shaft speeds (less than 5000 RPM)
or for older turbo-compressor designs. Frequently, they are modified to incorporate a pressure dam, in the direction of rotation. The pressure dam must be
190
Gears and Couplings
Chapter | 4
FIGURE 4.1.4 Helical pinion tooth reactions at pitch diameter.
positioned in the top half of the bearing to increase the load vector. This action
assures that the tangential force vector will be small relative to the load vector
thus preventing shaft instability. It should be noted that incorrectly assembling
the pressure dam in the lower half of the bearing would render this type of
bearing unstable. When shaft speed is high, other alternatives to prevent rotor
instabilities are noted in Fig. 4.1.7.
Shown are examples of anti-whirl bearings. The most common types of
these bearings are the 3 and 4 lobe design. Elliptical and offset bearing designs
do prevent instabilities but tend to increase shaft vibration if the load vector
191
B.P. 4.1
More Best Practices for Rotating Equipment
FIGURE 4.1.5 Determination of radial bearing loads.
passes through the major axis of the bearing. These types of bearings may have
to be rotated in the bearing brackets to prevent this occurrence.
The most common hydrodynamic bearing for higher speed applications is
the tilt pad journal bearing shown in Fig. 4.1.8. A tilting pad bearing offers the
advantage of increased contact area since the individual pads conform to the
shaft orbit. In addition, this type is also a highly effective anti-whirl bearing
since the spaces between the pads prevent oil whirl. Most end users specify tilt
pad radial and thrust bearings for turbo-compressor applications.
FIGURE 4.1.6 Straight sleeve bearing liner. Courtesy of Elliott Co.
192
Gears and Couplings
Chapter | 4
FIGURE 4.1.7 Prevention of rotor instabilities.
FIGURE 4.1.8 Tilting pad journal bearing assembly. Courtesy of Kingsbury, Inc.
193
B.P. 4.1
More Best Practices for Rotating Equipment
FIGURE 4.1.9 Typical journal bearing selection curve. Courtesy of Kingsbury, Inc.
Fig. 4.1.9 shows the mechanical frictional losses and oil flow requirements
for a tilt pad journal bearing as a function of shaft speed.
Note that the basis for horsepower loss and oil flow is an oil temperature rise
of 16.7°C (30°F). This is the normal design ∆T for all hydrodynamic bearings.
Also given in this figure is the data necessary to calculate bearing pressure at
the load point.
As an exercise calculate the following for this bearing:
Projected Area
APROJECTED = 5in. × 2in.
= 10 in.2
Pressure
= 3479 Lb force ÷ 10 in.2
= 347.9 psi on the oil film at load point
194
Gears and Couplings
Chapter | 4
Condition Monitoring
In order to determine the condition of any journal bearing, all the parameters
that determine its condition must be monitored. Table 4.1.2 presents the eight
parameters that determine the condition of a hydrodynamic journal bearing
along with typical limits. Attendees are advised to consult the manufacturers
instruction book for vendor recommended limits.
One important parameter noted in Fig. 4.1.10 that is frequently overlooked
is shaft position. Change of shaft position can only occur if the forces acting on
a bearing change or if the bearing surface wears. Fig. 4.1.10 shows how shaft
position is determined using standard shaft proximity probes.
Regardless of the parameters that are condition monitored, relative
change of condition determines if and when action is required. Therefore,
effective condition monitoring requires the following action for each monitored condition:
l
l
l
Establish baseline condition
Record condition trend
Establish condition limit
Fig. 4.1.11 presents these facts for a typical hydrodynamic journal bearing.
Based on the information shown in this trend, the bearing should be
inspected at the next scheduled shutdown. A change in parameters during
month 6 has resulted in increased shaft position, vibration, and bearing pad
temperature.
TABLE 4.1.2 The Eight Parameters That Determine the Condition of a
Hydrodynamic Journal Bearing Along With Typical Limits
Parameter
Limits
1. Radial vibration (peak to peak)
2.5 mils (60 µm)
2. Bearing pad temperature
220°F (108°C)
3. Radial shaft position (except for gearboxes where greater values are normal
from unloaded to loaded operation)
>30° change and/or 30% position
change
4. Lube oil supply temperature
140°F (60°C)
5. Lube oil drain temperature
190°F (90°C)
6. Lube oil viscosity
Off spec 50%
7. Lube oil flash point
Below 200°F (100°C)
8. Lube oil particle size
Greater than 25 µm
Condition monitoring parameters and their alarm limits according to component:
1. Journal bearing (hydrodynamic)
195
B.P. 4.1
More Best Practices for Rotating Equipment
FIGURE 4.1.10 Shaft movement analysis (relative to bearing bore). Courtesy of M.E. Crane
Consultant.
FIGURE 4.1.11 Trending data for a typical hydrodynamic journal bearing.
196
Gears and Couplings
Chapter | 4
TABLE 4.1.3 Vibration
j
j
Vibration is the result of a system being acted on by an excitation.
This excitation produces a dynamic force by the relationship:
FDYNAMIC = Ma
Where M is the mass (weight/g)
g is the acceleration due to gravity (386 in./s2)
a is the acceleration of mass M (in./s2)
j
Vibration can be (when look at rotor from Aerial view):
j
Lateral
j
Axial
j
Torsional
Vibration Instabilities
Vibration is an important condition associated with journal bearings because
it can provide a wealth of diagnostic information valuable in determining the
root cause of a problem. Table 4.1.3 presents important information concerning
vibration.
Fig. 4.1.12 defines excitation forces with examples that can cause rotor
(shaft) vibration.
Turbo-compressors generally monitor shaft vibration relative to the bearing bracket using a non-contact or “proximity probe” system as shown in
FIGURE 4.1.12 Excitation forces with examples.
197
B.P. 4.1
More Best Practices for Rotating Equipment
FIGURE 4.1.13 Non-contact displacement measuring system.
Fig. 4.1.13. The probe generates a DC eddy current which continuously measures the change in gap between the probe tip and the shaft. The result is that
the peak to peak unfiltered (overall) shaft vibration is read in mils or thousandth
of an inch. The DC signal is normally calibrated for 200 mV/mil. Probe gaps
(distance between probe and shaft) are typically 1 mm (0.040 mils) or 8 V DC
to assure the calibration curve is in the linear range. It is important to remember
that this system measures shaft vibration relative to the bearing bracket and
assumes the bearing bracket is fixed. Some systems incorporate an additional
bearing bracket vibration monitor and thus record vibration relative to the earth
or “seismic vibration.”
As previously discussed, vibration limits are usually defined by:
Vibration(mils p–p) =
12000
RPM
This value represents the allowable shop acceptance level. A.P.I. recommends alarm and trip shaft vibration levels be set as follows:
VALARM =
VTRIP =
24000
RPM
36000
RPM
In my opinion, shaft vibration alarm and trip levels should be based on the
following parameters as a minimum and should be discussed with the machinery vendor prior to establishing levels:
198
Gears and Couplings
Chapter | 4
FIGURE 4.1.14 Vibration severity chart. Courtesy of Dresser-Rand and C. J. Jackson P.E.
l
l
l
l
l
l
l
Application (critical or general purpose)
Potential loss of revenue
Application characteristics (prone to fouling, liquid, unbalance, etc.)
Bearing clearance
Speed
Rotor actual response (Bode Plot)
Rotor mode shapes (at critical and operating speeds)
Fig. 4.1.14 presents a vibration severity chart with recommended action.
A schematic of a shaft vibration and shaft displacement monitor is shown
in Fig. 4.1.15.
As mentioned earlier, vibration is measured unfiltered or presents “overall
vibration.” Fig. 4.1.16 shows a vibration signal in the unfiltered and filtered
conditions. All vibration diagnostic work (troubleshooting) relies heavily on
filtered vibration to supply valuable information to determine the root cause of
the vibration.
Fig. 4.1.17 presents an example of a radio tuner as an analogy to a filtered
vibration signal. By observing the predominant filtered frequencies in any overall (unfiltered) vibration signal, valuable information can be gained to add in the
troubleshooting procedure and thus define the root cause of the problem.
199
B.P. 4.1
More Best Practices for Rotating Equipment
FIGURE 4.1.15 Shaft vibration and displacement.
FIGURE 4.1.16 Vibration frequency.
200
Gears and Couplings
Chapter | 4
FIGURE 4.1.17 Radio tuner/vibration filter analogy.
B.P. 4.2: Always check the following when replacing gear couplings with
dry couplings.
l
l
l
l
Torsional Natural Frequencies
Lateral Natural Frequencies
Thermal Expansion capability of dry coupling connecting shaft
Coupling guard internal clearances to coupling hub flange O.D.
L.L. 4.2: There have been many experiences involving operating within a
Natural frequency when converting to dry coupling.
If the aforementioned was checked properly, it would have resulted in on
time start-ups without any issues.
BENCHMARKS
This B.P. has been in use since the early 1990s when converting to dry couplings
and has resulted in trouble free start-up and no delays due to coupling related
issues.
SUPPORTING MATERIAL
The Coupling Function
The function of a flexible coupling is to transmit torque from the driver to the
driven machine while making allowances for minor shaft misalignment and
shaft end position changes between the two machines. The design of the coupling should provide for transmission of the required torque at the required
201
B.P. 4.2
More Best Practices for Rotating Equipment
FIGURE 4.2.1 Shaft misalignment and axial position.
speed with a minimum of extraneous forces and perturbations exerted on either
the driver or driven shaft. Shaft misalignment exists when the centerlines of two
shafts joined by a coupling do not coincide. Fig. 4.2.1 shows the various types
of misalignment and shaft end position changes that can occur.
Each coupling type has a maximum tolerance of misalignment and axial
position change that is noted on the coupling drawing. Regardless of coupling
type, misalignment tolerance is stated in degrees and is usually ¼˚. Axial position change tolerance varies with coupling type. Gear type couplings have a
large axial position change tolerance compared to flexible element types.
Types
The following is a list of various types of flexible couplings:
j
j
j
j
j
j
j
Gear couplings
Continuous lubrication
Grease packed
Flexible membrane or flexible disc couplings
Single membrane type
Multiple membrane or multiple disc type
Couplings with elastomer insert flexible drive members
Gear Couplings
Gear type couplings are shown in Figs. 4.2.2 and 4.2.3. Gear couplings usually
include two separate gear mesh units. Each gear mesh unit consists of an external gear which fits closely into an internal gear. The internal gear can either
be part of the coupling hub assemblies or mounted on each end of the coupling
202
Gears and Couplings
Chapter | 4
FIGURE 4.2.2 Gear tooth coupling (grease packed). Courtesy of Zurn Industries.
FIGURE 4.2.3
Industries.
Continuously lubricated gear type coupling with spacer. Courtesy of Zurn
203
B.P. 4.2
More Best Practices for Rotating Equipment
spacer assembly. If the internal gears are hub mounted, then the external gears
are spacer mounted and vice versa.
Grease pack couplings (Fig. 4.2.2) are normally designed with hub mounted external gears and the internal gears are part of a sleeve type spacer which
serves as a retainer for the grease lubrication. The flange joint of the sleeve is
either precision ground to avoid lubrication leaks or has a gasket between the
two flange faces. The sleeve ends are fitted with “O” ring seals to keep dust out
and lubrication in.
In recent years, flexible element couplings have been used almost exclusively. However, many gear type couplings are still in use. They are the most
compact coupling for a given amount of torque transmission of all the coupling
designs. For this reason, they also have the least overhung weight. In addition,
the gear coupling can adapt more readily to requirements for axial growth of
the driver and driven shafts. Axial position change tolerances are on the order
of ½ in. or greater.
There is a common disadvantage in all gear type flexible couplings. Any
gear mesh has a break-away friction factor in the axial direction. This is caused
by the high contact force between the two sets of gear teeth. The result is that the
forces imposed on the driver and driven shafts are not totally predictable and are
sometimes higher than desired due to the quality of the tooth machine surfaces
and the inevitable build up of sludge or foreign material in the tooth mesh during extended service. These forces are detrimental to the ability of the coupling
to make the required corrections for misalignment but, more importantly, can
have a disastrous effect on the ability of the coupling to correct for thermal or
thrust force changes between the driver and driven machines.
Both coupling manufacturers and users have long been aware of this problem
and have used many methods to minimize the effect. Some of these methods are:
j
j
j
j
Reduction of the forces between the gear teeth by increasing the pitch diameter of the gear mesh. This is often self defeating in that it results in
increased size of the coupling and the coupling weight.
Reduction of the break-away friction factor by the use of higher quality gear
tooth finish and better tooth geometry and fit.
Reduction of sludge and foreign material build up in the gear mesh by finer
filtration of the coupling lubricant.
Reduction of sludge and foreign material build up in the gear mesh by incorporating self flushing passages and ports in the coupling to allow any
contaminants to pass through in the lubricant without being trapped in the
gear mesh area.
These steps have been only partially successful and the problem still exists
in many applications.
Coupling manufacturers are asked to quote the design break-away friction
factor of their coupling as built and shipped from the factory. Machinery train
designers then use this figure to calculate the maximum axial force that the
204
Gears and Couplings
Chapter | 4
coupling would be expected to exert on the connected shafts. From this information, the designers can decide if the thrust bearings adjacent to the coupling
are adequate to handle the axial loads within the machine plus the possible load
from the coupling resistance to any external forces.
There has been much discussion and some disagreement regarding the friction factor to be used when calculating possible thrust forces which can be
transmitted by the coupling. When the coupling is in reasonably good condition, factors from 0.15 to 0.30 have been considered reasonable. Since the factor
reflects the total force relationship, the coupling design can have a significant effect on the factor used. The factor is a function of the number of teeth in contact
and the contact areas of each tooth plus the quality of the tooth contact surface.
If we assume that the factor to be used is 0.30, then the axial force which must
be exerted in order to allow the coupling to correct for axial spacing changes
can be calculated as:
Fa =
0.30 × T
Dp /2
where Fa is the required axial force in kg (pounds), T is the design torque in
Ncm (in./lb.), and Dp is the pitch diameter of gear mesh in cm (inches).
We can assume then, that if we use a coupling with a 15 cm (six inch) pitch
diameter gear mesh transmitting 28,250 Ncm (25,000 in./lb.) of torque and a
break-away friction factor of 0.30, the axial force required to move the gear
mesh to a new axial position would be 11,300 N (2,500 lb.). Adjacent thrust
bearings must be capable of handling this force in addition to the machine’s
normal calculated thrust forces. Machinery train designers and users must be
aware of this and make provisions for it in the built-in safety factors of thrust
bearings and machinery mounting design.
The machinery user must know that the same phenomenon has an effect on
machinery vibration when machinery is operated with excessive misalignment.
The gear mesh position must change with each revolution of the shaft to correct
for the misalignment. This results in counter axial forces on a cyclic basis since
the mesh is moving in opposite directions at each side of the coupling. Vibration
detection and monitoring instrumentation will show that the resulting vibration
will occur at twice the running frequency of the shafts. Although the primary
force generated is axial, the resultant can show up as a radial vibration due to
the lever arm forces required on the coupling spacer to make the gear meshes act
as ball and socket connections. Axial or radial vibration in rotating machinery
which occurs at twice the frequency of the shaft rotational speed will normally
be an indication of misalignment between the two machines.
Fig. 4.2.3 shows a continuously lubricated, spacer gear type coupling. Spacers are usually required for component removal (seals, etc.). They also provide
greater tolerance to shaft misalignment. A common spacer size used for unspared (critical) equipment is 46 cm (18 in.).
205
B.P. 4.2
More Best Practices for Rotating Equipment
Flexible Membrane or Flexible Disc Couplings
Couplings in these categories do not have moving parts and derive their flexibility from controlled flexure of specially designed diaphragms or discs. They
do not require lubrication and are commonly known as “dry couplings.” The
diaphragms or discs transmit torque from one shaft to the other just as do the
gear meshes in a gear coupling.
The following features are common to all flexible disc or flexible membrane
type couplings:
1. None require lubrication.
2. All provide a predictable thrust force curve for a given axial displacement
range.
3. Properly applied, operated, and maintained, none are subject to wear and
have an infinite life span.
4. All provide smooth, predictable response to cyclic correction for minor
misalignment.
It should be noted that none of the aforementioned comments can be applied across the board to gear type flexible couplings. For this reason, more and
more special purpose machinery trains are being supplied with flexible metallic
element couplings in their design. Many users do not allow the use of gear type
coupling for critical (unspared) applications.
The following is a discussion of the various types of “dry” couplings with
comments pertaining to their application ranges and limitations. Fig. 4.2.4
shows a typical flexible disc coupling.
This is the most common type and is generally used for general purpose
applications (pumps, fans, etc.). The major consideration with this type of coupling is assuring the shaft end separation (B.S.E.) is within the allowable limits
of the couplings. This value is typically only 1.5 mm (0.060 in.) for shaft sizes
FIGURE 4.2.4 Flexible disc spacer coupling. Courtesy of Rexnord.
206
Gears and Couplings
Chapter | 4
FIGURE 4.2.5 Single diaphragm spacer coupling. Courtesy of Lucas Aerospace.
in the 1–2 in. range. At shaft sizes over 4 in. the maximum end float can be
6 mm (0.150 in.) or more. Exceeding the allowable end float will significantly
increase the axial load on the thrust bearings of the equipment and can fail the
coupling discs. A single diaphragm, spacer type coupling is shown in Figs. 4.2.5
and 4.26. Fig. 4.2.5 is a cutaway view and Fig. 4.2.6 presents a two-dimensional
assembly drawing.
This type of coupling is commonly used for critical (unspared) applications
where axial end float values are less than 5 mm (0.125 in.). This limit is based
on an approximate axial float of ±1.5 mm (0.062 in.). If end float is greater than
5 mm (0.125 in.), a convoluted (wavy) diaphragm or multiple type diaphragm
must be used. During disassembly, care must be taken when removing the spacer to not scratch or dent the diaphragm element. A dent or even a scratch that
penetrates the protective coating can cause a diaphragm failure.
Regardless of the type of diaphragm couplings, it is common practice to
“pre-stretch” these couplings to take full advantage of the maximum available
end float. Readers are cautioned to always require equipment vendors provide
axial shaft movement calculations in order to confirm that the coupling maximum end float is not exceeded. Fig. 4.2.7 graphically displays the various combinations of end shaft movement and the calculation method.
Fig. 4.2.8 is a picture of a multiple, convoluted (wavy) diaphragm spacer
coupling.
This type of coupling is used whenever large values of axial end float exist.
Axial end float values of ±22.2 mm (0.875 in.) or greater are attainable with this
type of coupling.
207
B.P. 4.2
More Best Practices for Rotating Equipment
FIGURE 4.2.6 Single diaphragm spacer coupling. Courtesy of Lucas Aerospace.
FIGURE 4.2.7 Rotor/case thermal movement—steam turbine.
208
Gears and Couplings
Chapter | 4
FIGURE 4.2.8 Multiple, convoluted diaphragm-spacer coupling. Courtesy of Zurn Industries.
As previously mentioned, gear type couplings provide the lowest value of
overhung weight (coupling moment) on the bearing. However, a dry type coupling will usually have a higher coupling moment because the flexible assembly
is farther from the bearing centerline than the gear teeth in a gear coupling. An
excessive coupling moment will reduce the second natural frequency (Nc2) of
a turbo-compressor and could move it close to or within the operating speed
range. A solution in these cases can be to use a reduced moment diaphragm
coupling as shown in Fig. 4.2.9.
In this design, the diaphragm is moved to the back of the hub and the flange
diameter is reduced thus significantly reducing the coupling moment. The reduced moment coupling approaches the gear coupling in terms of coupling moment value.
FIGURE 4.2.9 Reduced moment convoluted (wavy) diaphragm spacer coupling. Courtesy of
Lucas Aerospace.
209
B.P. 4.2
More Best Practices for Rotating Equipment
FIGURE 4.2.10 Jaw and spider coupling.
Couplings with Elastomer Insert Flexible Drive Members
This type of coupling is normally used only for low horsepower, general purpose
applications. Their limitations are based primarily on the wear factor and the
difficulty in maintaining shape and concentricity of the elastomer insert. These
items have a tendency to limit the maximum design speed at which such couplings can be operated. A typical “Jaw and Spider” type is shown in Fig. 4.2.10.
One exception is a special design used for synchronous motor driven compressor trains. A characteristic of synchronous motors is a variable oscillating torque that decreases linearly in frequency from 2× line frequency (50 or
60 HZ) at 0 RPM to 0 frequency at rated RPM. Fig. 4.2.11 shows a plot of motor
RPM versus transient torsional excitation frequency. The excitation frequency
FIGURE 4.2.11 Transient torsional excitation—Frequency versus motor speed.
210
Gears and Couplings
Chapter | 4
FIGURE 4.2.12 Holset coupling (exploded view) non-spacer type.
inherent in all synchronous motors will excite all torsional natural frequencies
present between 2× line frequency and 0 RPM.
When the motor torsional excitation frequency briefly coincides with a torsional natural frequency, torque values can amplify to as much as 5 or 6 times
full load torque. The “Holset” or elastomeric coupling shown in Fig. 4.2.12 significantly reduces the torque amplification by dampening out the response in
the elastomeric elements. The hardness of these elements is controlled to limit
the maximum amplification factor to an acceptable value (usually 2–3 × rated
torque).
The Coupling System
It has been the writers’ experience that if couplings are properly selected, the
root cause of failure, if it occurs, is in the coupling system.
The “coupling system” must continuously transmit torque safely between
the driver and driven equipment and must allow for changes in shaft misalignment and axial movement. The components that make up the shaft system are:
j
j
j
j
j
The driver shaft
Driver shaft/coupling fit
Coupling
Driven shaft/coupling fit
The driven shaft
211
B.P. 4.2
More Best Practices for Rotating Equipment
FIGURE 4.2.13 The coupling system. Courtesy of M.E. Crane Consultant.
j
j
j
The coupling spacer system
Lubrication (if required)
Cooling system (if required)
A schematic of a coupling system is shown in Fig. 4.2.13. The reliability of
the coupling is a function of the coupling system design and assembly. If any
of the items noted earlier are not properly designed or assembled a coupling
failure can occur.
Coupling assembly/disassembly errors and enclosed coupling guard design
are two important areas that are critical to coupling system reliability.
Coupling Installation and Removal
The most common methods of coupling attachment are:
j
j
j
Key fit
Spline fit
Hydraulic fit
Key fits are used whenever possible. They are the most common method of
shaft fit. It is important to assure keys and keyways are properly manufactured
to avoid problems with removal or breakage. Key fits will be used on equipment
that does not require coupling removal to remove shaft components (seals, bearings, etc.). Keyed fits are usually used on motors, gearboxes and most pumps
and small steam turbines. Since heat is usually required to remove keyed couplings, they will not be used where removal in the field is necessary. In these
applications either spline or hydraulic fits are used.
212
Gears and Couplings
Chapter | 4
FIGURE 4.2.14 Typical coupling hydraulic shrink fit.
Spline fits consist of a male (on the shaft) and female (in coupling hub)
finely machined mating gear teeth with line to line fit (no backlash). When assembled on the shaft, the fit is rigid and provides no flexibility. Spline fits are
commonly used in the gas turbine industry. They do not usually require heat for
removal.
Hydraulic fits are used where heat to remove the coupling hub is either not
available or not permitted. Usually, turbo-compressors will utilize hydraulic fits
for this reason, since hydrocarbon gas, usually present, requires a flame free
environment. Fig. 4.2.14 shows a typical coupling hydraulic shrink fit arrangement. Note that the entire torque load is transmitted by the shrink fit and that no
keys are used!
The equipment vendor calculates the required shrink fit based on the shaft
and coupling dimensions. Typical values of hydraulic shrink fit are 0.002 in./in.
of shaft diameter.
For ease of hydraulic fit assembly and disassembly, all shafts and coupling
hubs are tapered. Different shaft/coupling hub matching tapers are used. The
most common are:
j
j
j
½° taper
½″ per foot taper
¾″ per foot taper (shafts above 4 in. diameter)
Once the shrink fit is calculated, the value appears on the coupling drawing and is usually expressed as “drive” or “push” based on the shaft taper. This
is the axial distance the coupling must be moved up the shaft. The coupling
drive per 0.001 in. of shrink fit for the most common shaft tapers is noted as
follows:
Shaft Taper
Drive Per 0.001 Shrink
½°
½ in. per foot
¾ in. per foot
1.448 mm (0.057 in.)
0.610 mm (0.024 in.)
0.406 mm (0.016 in.)
213
B.P. 4.2
More Best Practices for Rotating Equipment
As an example, a hydraulic fit coupling with a 101.5 mm (4 in.) bore requires a 0.008 in. shrink fit (i.e., the bore diameter is 0.008 less than the shaft).
To expand the coupling bore 0.008 in., what is the drive if the shaft taper is:
Taper
Drive
½°
½ in. per foot
¾ in. per foot
11.582 mm (0.456 in.)
4.877 mm (0.192 in.)
3.251 (0.128 in.)
Since the load torque is completely transmitted by the shrink fit, one can
see the importance of assuring that the correct shrink fit (or drive) is obtained.
The shrink fit amount is directly proportional to torque load c­ apability. If the
shrink fit is 50% of the specified value, so is the torque capability! However,
industry specifications require that the shrink fit at minimum tolerances be
a minimum of 125% greater than the driver maximum torque. Observing
the calculated drives in the aforementioned example it can be seen that the
smaller the shaft, the more critical the correct drive becomes for a given shaft
taper. The coupling drive is measured by positioning a dial indicator on the
coupling hub and measuring the axial distance traveled during coupling assembly.
FIGURE 4.2.15 Hydraulic fit coupling. Courtesy of Dresser-Rand Corp.
214
Gears and Couplings
Chapter | 4
Fig. 4.2.15 shows a typical hydraulic coupling mounting arrangement used
by Dresser-Rand. All turbo-compressor manufacturers use similar arrangements. There are some slight differences which are:
Hydraulic oil enters the coupling hub and not the shaft.
An additional pump is used to move the hydraulic tool axially (Fig. 4.2.15
shows a nut which is manually turned to push the coupling axially).
j
j
The basic coupling mounting procedure is as follows: (readers must refer to
the specific vendors’ instruction book for the exact procedure)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Clean shaft end and coupling bore with light oil.
Remove all “O” rings from shaft end and coupling.
Lightly blue the coupling hub.
Push coupling on shaft without “O” rings and tap hub with wood to assure
tight fit.
While hub is on shaft, index coupling hub axial position relative to a machined surface on shaft (usually shaft end).
Remove hub and confirm contact area of blue is a minimum of 85%. If not,
correct as required.
When coupling contact of 85% is confirmed, clean shaft and coupling hub
and install shaft and coupling “O” rings.
Hand push coupling on shaft to indexed position in step 5.
NOTE: It may be necessary to use pump since “O” rings can provide
significant resistance to movement.
With hub at indexed (zero drive) position, use hand pump to push coupling axially to value noted on coupling drawing.
Coupling drive must be within tolerances noted.
NOTE: Pump pressures will be high. Be extremely careful when connecting pump and tubing. Be sure to secure pump so that hand jacking
cannot break tubing. Pressures typically required range from 103,000–
206,000 kPa (15,000–30,000 PSI) depending on shaft dimensions, coupling dimensions, and shrink fit.
When coupling is on shaft correct amount, do not remove dial indicator but
reduce pump pressure to zero and back off hydraulic tool slightly. Observe
that dial indicator does not move before removing tool.
Promptly assemble shaft end coupling nut.
Measure between shaft end dimension (B.S.E.) to assure it is as stated on
coupling drawing before assembling coupling spacer. If this dimension is
not correct, consult instruction book and O.E.M. if necessary before taking corrective action. Under no circumstances should coupling spacers be
added unless allowed by the coupling manufacturer or should equipment
axial shaft position be changed without O.E.M. consent.
When coupling is properly assembled check alignment using “reverse dial
indicator procedure.”
215
B.P. 4.2
More Best Practices for Rotating Equipment
NOTE: For coupling removal, consult vendor’s instruction book.
Under no circumstances should coupling be pulled or heated. Usually,
hydraulic pressure required for removal will be higher (5–10%) than that
required for assembly. If the value required exceeds 241,000 kPa (35,000
PSI), do not proceed until consulting O.E.M. for additional options concerning removal.
Incorrectly mounting a hydraulic coupling can cause catastrophic coupling and/or shaft end failure.
Enclosed Coupling Guards
Most turbo-compressor couplings are completely enclosed by a spark proof
(usually aluminum) coupling guard. This is because the couplings are continuously lubricated gear type or to prevent oil siphoned from the bearing brackets
by the windage action of the dry couplings. In either case, proper design of the
coupling guard is essential to maintaining coupling reliability. Many coupling
failures have resulted from high coupling enclosure temperatures, enclosures
full with oil and debris educted into the coupling guard from the atmosphere.
As a minimum the following must be checked by the O.E.M. and coupling
vendor during equipment design or field coupling retrofit from gear to dry type:
l
l
l
Proper coupling O.D. to guard and/or bearing bracket I.D. clearance.
Proper coupling guard baffle design to allow proper drainage. NOTE: All
enclosed coupling guards must be supplied with vent and drain.
Proper vent breather sizing and design.
Fig. 4.2.16 presents coupling guard dimensional design criteria for dry type
couplings operating in enclosed coupling guards. Note that in some designs, Do
may be the I.D. of the bearing bracket.
Recommend coupling guard skin operating temperatures should be below
93°C (200°F) to avoid coupling and coupling guard leakage problems. Under no
circumstances should coupling guard skin temperatures approach the flash point
of lubricating oil, 200°C (400°F) for new mineral oil.
Field Retrofits From Lubricated to Dry Couplings
Considering the advantages of dry couplings, many users are retrofitting their
older style lubricated couplings to dry couplings. Whenever considering a retrofit, the following action should be taken to maintain or increase coupling system
reliability.
1. Consult equipment or coupling vendor for proper selection of new coupling.
2. Consult equipment O.E.M.(s) (each affected vendor) to confirm:
a. Critical speeds will not be affected
b. Coupling guard design is acceptable
216
Gears and Couplings
Chapter | 4
FIGURE 4.2.16 Coupling guard temperature calculations. Courtesy of M.E. Crane Consultant.
3. Advise coupling vendor or any environmental considerations that may affect dry
flexible element life (environmental gases, temperatures, excessive dust, etc.).
B.P. 4.3: Always match-mark hydraulic fit coupling to shaft in order to
observe if the coupling slipped at all.
This will allow for observation that either installation/installation procedure
was incorrect. If this is observed, be sure to check shaft end taper is 0.5 in. per ft.
or less and that pusher was left on the hub for the vendors recommended period
of time after it has found its “Home” position.
L.L. 4.3: Failure to match mark the hydraulic fit coupling has resulted
in severe failures that could have been discovered and fixed during prior
shutdowns.
BENCHMARKS
This best practice has been used since the early 1990s and has caught many issues prior to them becoming catastrophic failures.
SUPPORTING MATERIAL
See supporting material for B.P. 4.2
217
Chapter 5
Steam Turbines
B.P. 5.1: Always require single valve steam turbines to be supplied with a
throttle valve position indicator.
While single valve turbines are typically used in spared applications, they
many times are a driver for critical pumps within the plant (Charge pumps,
BFW, oil system pump for critical compressor train, etc.) that can affect production. Therefore, they should be equipped with instrumentation to allow for
proper monitoring.
A position indicator on the throttle valve will allow for indication of both
driver and driven condition. If the throttle valve goes open for the same steam
conditions and speed, it most likely indicates that the driven equipment is in
poor condition. However, if the same occurs and the performance for the driven
equipment has not changed, then the steam turbine is the culprit. It could be
fouled or worn, but it may be something as simple as the hand valve(s) were
closed and the turbine couldn’t take the required steam flow to meet the new
process demands. Also, if the valve position has not moved and the speed has
decreased, this indicates the governor linkage is bound up (common with these
small turbines). Note that adherence to BP 5.15 (Always have a Tachometer
installed on single valve steam sturbines) from Forsthoffer’s Best Practice
Handbook for Rotating Machinery is required in order to catch this properly as
it might be thought that the pump is worn if the speed is not monitored and it
has decreased.
This is also a good gauge of how much power is available and, if used properly, can help plan for maintenance on the turbine. The rate of % opening of the
valve can be tracked and an estimate can be made as to when it would almost
be wide open. Obviously, if you can make the rates that the process is demanding and have more travel left in the throttle valve, you keep operating, but trend
the position.
In today’s day and age it is highly recommended to have an electronic positioner on the valve since the signal can be sent to the DCS and the value trended
with all of the other key parameters for the unit.
More Best Practices for Rotating Equipment. http://dx.doi.org/10.1016/B978-0-12-809277-4.00005-X
Copyright © 2017 Elsevier Inc. All rights reserved.
219
B.P. 5.1
More Best Practices for Rotating Equipment
L.L. 5.1: Inability to know throttle valve position can result in reduction of
speed for critical equipment which equals reduction of rates/profit.
BENCHMARKS
The writer has used the philosophy of equipping critical machinery with instrumentation for proper monitoring of rotor condition extensively in the last ten
years. Many instances have occurred where an issue was noticed in its infancy
stage in order to allow time for maintenance to occur in a planned manner and
save from significant loss of production.
SUPPORTING MATERIAL
Single Stage Turbine Guidelines
The five common problems with single stage turbines are noted in Table 5.1.1.
We will now discuss each problem in detail. Please refer to Fig. 5.1.1, which
has each problem area circled.
Bearing Bracket Oil Contamination
Please refer to Item 1 in Fig. 5.1.1.
The most common reliability problem with single stage steam turbines is the
contamination of the oil in the bearing housing with water. The root cause of
the problem is the ineffectiveness of the floating carbon ring shaft seal system
to stop.
Unless present site systems are modified to eliminate the root cause, the best
action plan is to minimize the effect of the contamination so a bearing failure
will not occur. Such an action plan is presented in Tables 5.1.2–5.1.4.
Slow Governor System Response
Please refer to Item 2 in Fig. 5.1.1. Another very common reliability problem
is the slow or non-movement of the governor system linkage during start-up
and normal operation during steam condition changes. It will appear that the
TABLE 5.1.1 Single Stage Steam Turbines Common Reliability Problems
j
j
j
j
j
220
Bearing bracket oil contamination (inadequate carbon ring steam seal design)
Slow governor system response (inadequate governor linkage maintenance and
governor power)
Hand valve(s) closed on critical services
Bearing bracket oil viscosity reduction and bearing wear (high pressure service)
Use of sentinel valves on turbine cases
Steam Turbines
Chapter | 5
221
FIGURE 5.1.1 Single stage steam turbines common reliability problems.
B.P. 5.1
More Best Practices for Rotating Equipment
TABLE 5.1.2 Bearing Bracket Oil Contamination (Root Cause)
Shaft carbon ring seal cannot positively prevent steam leakage
j
TABLE 5.1.3 Steam Turbine Bearing Bracket Oil Contamination Monitoring
Action Plan
Install oil condition site glasses in bearing bracket drain connection
Inspect once per shift
Drain water as required
Sample oil monthly initially
j
j
j
j
TABLE 5.1.4 How to Correct Carbon Ring Seal Ineffectiveness
Install steam eductor on each seal chamber leak off drain (between 4th and 5th
carbon ring)
Design eductor to pull 5–10 in. of H2O vacuum at this point
Alternative approach – install bearing housing isolation seal (“Impro” or equal)
j
j
j
governor is not responding because speed will not be controlled when it should.
Typical examples are:
j
j
Speed will continue to increase when throttle valve is opened, turbine will
trip on overspeed
Speed will increase or decrease when:
○ Steam conditions change
○ Driver equipment changes
These facts are presented in Table 5.1.5.
Since most single stage steam turbines are not supplied with tachometers,
it is difficult, if not impossible to condition monitor this problem. A condition
monitoring action plan is provided in Table 5.1.6.
TABLE 5.1.5 Slow Governor System Response
1. Rapid speed change and trip on start-up
2. Speed increase or decrease on steam condition or load condition change
3. Governor instability (hunting) around set point
Note: #1 usually occurs on “solo”, #2 occurs during steady state operation
222
Steam Turbines
Chapter | 5
TABLE 5.1.6 Slow Governor System Response Condition Monitoring
Action Plan
• Install tachometer on all single stage steam turbines
• Always test speed control on “solo run” (1)
• Monitor turbine speed once per shift. Take corrective action if speed varies +/- 5%
(200 rpm)
Note: (1) since load is very low, test acceptance is the ability to stabilize speed and prevent
overspeed trip when throttle valve is slowly opened.
TABLE 5.1.7 Causes of Excessive Governor Mechanical Linkage System
and Valve Friction
•
•
•
•
Linkage bushings not lubricated with high temp. grease
Valve steam packing too tight
Steam deposits in valve and/or packing after extended shut down (turbine cold)
Bent steam valve stem
TABLE 5.1.8 Slow Governor System Response Condition Monitoring
Action Plan
•
•
•
•
If problems occur, disconnect linkage and confirm ease of valve movement
Replace bushings and/or lubricate with “molycote” or equal
Clean deposits from valve and packing as required
If above action does not correct problem, replace governor (inspection and/or adjustment of governor droop is required)
The usual root cause of the problem is that the friction in the mechanical linkage
and/or valve stem packing exceeds the maximum torque force that the governor output lever can deliver. The governor designations TG-10, TG-13, and TG-17 simply
mean “turbine governor with … FT-LB torque.” Therefore, if a TG-10 governor is
installed and the torque required to move the valve steam exceeds the value of 10 FTLBs, the governor system will not control speed. Taking the governor to the shop,
will not solve the problem. Causes of excessive friction are shown in Table 5.1.7.
Table 5.1.8 shows the action plans to free the governor linkage when response is slow.
Hand Valve(s) Closed on Critical Services
Most single stage steam turbines are supplied with one or more hand valves in
the steam chest. Refer to Figure 5.1.2, Item 3. The purpose of the hand valves is
223
B.P. 5.1
More Best Practices for Rotating Equipment
FIGURE 5.1.2 Single valve turbine admission path.
to allow more or less inlet steam nozzles to be used during operation. Optimizing the steam nozzles used, maintains turbine efficiency during load changes.
However, the efficiency of single stage steam turbines is only 35% at best!
Therefore, adjustment of hand valves, other than during start-up or during slow
roll, should not be required.
We have witnessed many unscheduled shutdowns of critical (unspared)
compressor units because the general purpose steam turbine that is the main
lube oil pump driver, had the hand valves closed. An upset in the steam system
reduced steam supply pressure and caused the turbine and lube pump to slow
down. This was because, hand valves were closed and the throttle valve, even
when full open, could not meet steam flow requirements. When the speed of the
steam turbine decreased, the lube oil pressure dropped and guess what? … The
auxiliary pump did not start in time and the unit tripped.
Table 5.1.9 presents the recommended action plan in the refinery for single
stage steam turbine hand valves.
Bearing bracket oil viscosity reduction and bearing wear on high-pressure
single stage steam turbines.
224
Steam Turbines
Chapter | 5
TABLE 5.1.9 Single Stage Steam Turbine Hand Valve Recommendations
• Never throttle hand valves
• Hand valves should be open on main oil pump and auto-start steam turbines
TABLE 5.1.10 High-Pressure Single Stage Steam Turbine Bearing Problems
and Oil Viscosity Reduction
• Sleeve bearings (usually steam inlet end) wear out quickly
• Oil viscosity is reduced and difficult to maintain
Please refer to Fig. 5.1.1, Item 4. Observe the jacket in the bearing housings.
The purpose of this jacket is to cool the oil in the bearing bracket. When the inlet
steam pressure is high, the high temperature of the steam is transmitted to the steam
end inlet bearing through the shaft. Although the jacket in the bearing housing does
reduce the oil temperature in the bearing housing, it cannot effectively reduce the
oil temperature at the shaft/bearing interface. Table 5.1.10 presents these facts.
This problem is a design issue. A small single stage turbine is not provided
with an effective oil system to remove the heat between the shaft and bearing
when the turbine is operating on high temperature (up to 750°F) steam. The
solution is to require pressure lubrication for this application.
Naturally, it is difficult, and not cost effective to retrofit these turbines for
pressure lubrication. The field proven solutions to this problem are presented in
Table 5.1.11.
Also, another tool used to monitor the condition of the bearing bracket oil is
illustrated in Fig. 5.1.3. An oil condition monitor, which has a trade name of PK
glass among other names, will give a visual indication of water or other particle
contamination. The water can be drained out easily online.
Continued Use of Sentinel Valves on Turbine Cases
Please refer to Fig. 5.1.1, Item 5. Sentinel valves were used, years ago, as alarm
devices indicating that the steam turbine case (low pressure part) was under
excessive pressure.
TABLE 5.1.11 Eliminate Bearing Wear and Oil Viscosity Reduction (High
Pressure Service)
• Assuring bearing housing jacket passages are open (flushed)
• Consulting with turbine vendor for bearing material change
• Using special high temperature service oil (synthetic based oil)
225
B.P. 5.2
More Best Practices for Rotating Equipment
FIGURE 5.1.3 Oil Condition Monitoring Bottle.
TABLE 5.1.12 Prevent Excessive Sentinel Valve Maintenance
• Removing sentinel valves
• Assuring that inlet and exhaust casings are protected by properly sized and set pressure relief valves
These devises are not pressure relief valves and will not protect the case
from failure during over pressure events.
It is a known fact that the sentinel valves, wear, leak and require steam turbine shutdown for repair. Most large company specifications prevent the use of
sentinel valves and require full relief valve protection on the inlet and exhaust of
all single stage turbines. These facts are presented in Table 5.1.12.
B.P. 5.2: If the driven equipment has additional flow range and there are
no other plant bottlenecks, consider additional steam turbine power.
Many plants these days try to operate at the highest flows possible from the
initial start-up!!! Therefore, when more compressor flow = more $$$ then it can
be feasible to select a turbine with more power as long as it is specified early
on in the project.
It is also a good idea to max out turbine power to what the compressor could
produce in the future if it was uprated with new internals. The c­ ompressor
vendor can give you the HP required for operation at max flow within the same
compressor casing. This can only be justified if it is identified early on in the
project AND there are no other bottlenecks limiting production within the plant.
226
Steam Turbines
Chapter | 5
L.L. 5.2: Lack of turbine power can be a bottleneck for plant production.
Many times in the field it has been experienced when Turbine power is not
available to operate at desired process rates. This can be due to fouling or wear
in the compressor, fouling or wear in the turbine, or just trying to increase rates
to increase production. Either way you put it, this results in loss of production.
BENCHMARKS
The writer has used this best practice since 2000 when Mega Projects have
become the norm and have allowed pre-investment in additional driver power
based on additional revenue potential.
SUPPORTING MATERIAL
Steam Conditions
Steam conditions determine the energy available per pound of steam. Table 5.2.1
explains where they are measured and how they determine the energy produced.
Frequently, proper attention is not paid to maintaining the proper steam conditions at the flanges of a steam turbine. Failure to maintain proper steam conditions will affect power produced and can cause mechanical damage to turbine
internals resulting from blade erosion and/or corrosion. Table 5.2.2 presents
these facts.
Mollier Diagram or steam tables allow determination of the energy available in a pound of steam for a specific pressure and temperature. Table 5.2.3
describes the Mollier Diagram and the parameters involved.
TABLE 5.2.1 Steam Conditions
• The steam conditions are the pressure and temperature conditions at the turbine
inlet and exhaust flanges.
• They define the energy per unit weight of vapor that is converted from potential
energy to kinetic energy (work).
TABLE 5.2.2 Steam Condition Limits
Inlet steam conditions should be as close as possible (+/- 5%) to specified conditions
because:
• Power output will decrease
• Exhaust end steam moisture content will increase, causing blade, nozzle and diaphragm erosion.
227
B.P. 5.2
More Best Practices for Rotating Equipment
TABLE 5.2.3 The Mollier Diagram
Describes the energy per unit mass of fluid when pressure and temperature are known.
• Enthalpy (energy/unit mass) is plotted on Y-axis
• Entropy (energy/unit mass degree) is plotted on X-axis
• Locating P1, T1 gives a value of enthalpy (H) horizontal and entropy (S) vertical
• Isentropic expansion occurs at constant entropy (∆S = 0) and represents an ideal
(reversible) expansion
Refer to Fig. 5.2.1, which is an enlarged Mollier Diagram.
As an exercise, plot the following values on the Mollier Diagram in this
section and determine the corresponding available energy in BTUs per pound.
1. P1 = 600 PSIG, T1 = 800°F
2. P2 = 150 PSIG, T2 = 580°F
3. P1 = 1500 PSIG, T1 = 900°F
4. P2 = 2 PSIG, % moisture = 9%
BTU
LBM
BTU
h2 =
LBM
BTU
h1 =
LBM
BTU
h2 =
LBM
h1 =
Having plotted various inlet and exhaust conditions on the Mollier Diagram
to become familiar with its use, please refer to Table 5.2.4, which presents the
definitions and uses of steam rate.
Theoretical Steam Rate
The theoretical steam rate is the amount of steam, in kgs or lbs per hour required
to produce one horsepower if the isentropic efficiency of the turbine is 100%.
As shown in Table 5.2.4, it is determined by dividing the theoretical enthalpy
∆hisentropic into the amount of kJ/hr (btu’s/hr in one unit of power (kW or hp).
Actual Steam Rate
The actual steam rate is the amount of steam, in kg or lbs per hour, required to
produce one unit of power based on the actual turbine efficiency. As shown in
Table 5.2.4, it is determined by dividing the theoretical steam rate (T.S.R.) by
the turbine efficiency. Alternately, if the turbine efficiency is not known and the
turbine inlet and exhaust conditions are given (P2, T2, or % moisture), the actual
steam rate can be obtained in the same manner as theoretical steam rate but
substituting ∆Hactual for ∆Hisentropic.
228
Steam Turbines
Chapter | 5
FIGURE 5.2.1 Mollier steam diagram. (Courtesy of Elliott Company)
229
B.P. 5.2
More Best Practices for Rotating Equipment
TABLE 5.2.4 Determining Steam Rate
Uses:
j
Determine the amount of steam required per hour
j
Determine the amount of potential KW (horsepower)
Required:
j
Steam conditions
j
Theoretical steam rate table or Mollier Diagram
j
Thermal efficiency of turbine
Formula:
Metric Units
U.S. Units
Theoretical steam rate
j
T.S.R.(kg / kW - h) =
3600kJ / kW -HR
∆H ISENTROPIC
T.S.R.(lb / HP - h) =
2545BTU'S / HP -HR
∆HISENTROPIC
Actual steam rate
j
T.S.R.
Efficiency
3600kJ / kW -HR
=
∆H ACTUAL
A.S.R.(kg / kW - h) =
T.S.R
Efficiency
2545BTU / HP -HR
=
∆H ACTUAL HP / HR
A.S.R.(lb / HP - h) =
Turbine efficiency
j
T.S.R.
A.S.R
∆H ACTUAL
=
∆HISENTROPIC
Efficiency =
Turbine Efficiency
As shown in Table 5.2.4, turbine efficiency can be determined either by the ratio
of T.S.R. to A.S.R. or ∆hactual to ∆Hisentropic.
It is relatively easy to determine the efficiency of any operating turbine in the
field if the exhaust conditions are superheated. All that is required are calibrated
pressure and temperature gauges on the inlet and discharge and a Mollier Diagram or Steam Tables. The procedure is as follows:
1.
2.
3.
4.
5.
For inlet conditions, determine h1
For inlet condition with ∆S = 0, determine h2ideal
For outlet conditions, determine h2actual
Determine ∆hideal = h1–h2ideal
Determine ∆hactual = h1–h2actual
230
Steam Turbines
Chapter | 5
6. Determine efficiency
Efficiency =
∆Hactual
∆H ideal
However, for turbines with saturated exhaust conditions, the above procedure cannot be used because the actual exhaust condition cannot be easily determined. This is because the percent moisture must be known. Instruments
(calorimeters) are available, but results are not always accurate. Therefore the
suggested procedure for turbines with saturated exhaust conditions is as follows:
1. Determine the power required by the driven equipment or record turbine
power if a torque meter is installed. This is equal to the power produced by
the turbine.
2. Measure the following turbine parameters using calibrated gauges:
j
Pin
Pexhaust
j
Tin
Steam flow in (LBS/HR)
3. Determine the theoretical steam rate by plotting Pin, Tin, Pexhaust @ ∆S = 0.
and dividing ∆hisentropic into the constant.
4. Determine the actual steam rate of the turbine as follows:
j
j
Actual Steam Rate (A.S.R.) =
Steam Flow (LB/HR)
BHP required by driven equipment
5. Determine efficiency
Efficiency =
T.S.R.
A.S.R.
Table 5.2.5 presents the advice and values concerning steam turbine efficiencies. The efficiencies presented can be used for estimating purposes.
Refer to Figures 5.2.2 and 5.2.3 for typical efficiency values for multistage and
single stage steam turbines as a function of steam conditions, power and speed.
TABLE 5.2.5 Typical Steam Turbine Efficiencies
j
j
j
Quoted turbine efficiencies are external efficiencies, they include mechanical
­(bearing, etc.) and leakage losses
Turbine efficiency at off load conditions will usually be lower than rated efficiency
Typical efficiencies are presented for impulse turbine:
○ Condensing multi-stage
○ Non condensing multi-stage
○ Non condensing single state
231
B.P. 5.2
232
More Best Practices for Rotating Equipment
FIGURE 5.2.2 Efficiency of multistage turbines. (Courtesy of IMO Industries)
Steam Turbines
Chapter | 5
FIGURE 5.2.3 Efficiency of single stage turbines. (Courtesy of IMO Industries)
B.P. 5.3: Consider using Backpressure (or Extraction/Backpressure) turbines
whenever possible for process trains.
The use of Backpressure type turbines gives three distinct advantages as
follows:
l
l
l
Will eliminate issues with Blade Corrosion in the last few stages since the
steam should not be saturated on the back end.
Increased rotor rigidity due to less stages required.
Ease of monitoring turbine performance since steam is not saturated and efficiency can be calculated by ∆hact/∆hisen.
Of course plant steam balance needs to be taken into account in order to
implement this best practice. One suggestion would be to use Condensing Turbines for Steam Turbine Gen sets in the plant as there will be a spare gen set
usually and they operate at constant speed, which is generally more reliable.
L.L. 5.3: Condensing steam turbines will have moisture toward the exhaust
side, which could cause corrosion of the blades and nozzles in the last few stages.
There are many examples of blade/nozzle corrosion toward the back end of
condensing steam turbines. Unplanned shutdowns have occurred due to severe
corrosion of the nozzles, causing stationary blade failure.
233
B.P. 5.3
More Best Practices for Rotating Equipment
Note that more moisture will be present if the condenser is operated at more
of a vacuum (lower absolute pressure). See B.P. 5.4 for more details on this
point.
Condensing (and Extraction/Condensing) steam turbines are also very difficult to calculate efficiency unless you have a Torquemeter installed on the
train. This is B.P. 5.13 in Forsthoffer’s Best Practice Handbook for Rotating
Machinery.
BENCHMARKS
This Best Practice was brought to the writer’s attention in a recent project where
this was suggested by Process Engineering for the Process equipment. It was
immediately noted by the author the increased reliability that resulted by going
this route due to less risk of corrosion, increased rotor rigidity and ease of efficiency trending.
SUPPORTING MATERIAL
See B.P. 5.2 for Efficiency Calculations.
Types of Steam Turbines
In this section we will examine several types of Expansion Turbines commonly
installed in Refineries, Petrochemical Plants and other Installations.
All types of Expansion Turbines regardless of their design, perform similar duties. That is, they extract usable energy from a vapor and provide sufficient power to operate their load (driven equipment) at rated conditions. An
Expansion Turbine performs the opposite duty of a turbo-compressor. A turbo-­
compressor requires power to increase the energy of a vapor while an expansion
turbine obtains power to drive the turbo-compressor from the potential energy
of the vapor. In other words, in a turbo-compressor the blades work on the gas
and in an expansion turbine, the gas performs work on the blades.
There are many different types of expansion vapors. Steam, because of its
ease of generation and comparative cost effectiveness, is the most widely used
expansion vapor. However, many cryogenic (ethylene, hydrogen, etc.) and fired
vapors (gas generated) are also used. Fired vapors are used in a gas turbine. It
is extremely important to understand that the operation of a steam turbine and
an expansion turbine in a gas turbine are identical. We will examine the expansion of steam on a Mollier Diagram and also a hydrocarbon gas (ethylene) on
its Mollier Diagram.
The definition of single-stage, multi-stage, single valve, multi valve, back
pressure, and condensing turbines will be discussed along with the reason for
using these various types of steam turbines.
In addition, we will discuss extraction and admission steam turbines and explain why double and triple flow exhaust ends are used in large steam turbines.
234
Steam Turbines
Chapter | 5
FIGURE 5.3.1 Turbo compressor vs. expansion turbine.
We will conclude this section by discussing the various applications of the
different types of turbines discussed here. This will provide the attendee with
a working knowledge of when to use various types of steam turbines and why
those types are used for specific applications.
Fig. 5.3.1 defines the functions of a turbo-compressor and expansion turbine.
The turbo-compressor requires power and is termed a “driven” machine.
The expansion provides power and is called the driver or “prime mover.”
A function diagram of one row of impulse and reaction blading is shown in
Fig. 5.3.2.
FIGURE 5.3.2 Turbine blade types.
235
B.P. 5.3
More Best Practices for Rotating Equipment
The two basic types of blading sets used to extract energy from a vapor and
produce power are shown. In general, impulse blading has been widely used in
the steam turbine industry and reaction blading has been widely used in the gas
turbine industry. In recent years, the steam turbine industry has been designing a
“Hybrid Turbine” utilizing rugged impulse blading in the initial stages and high
efficiency reaction blading in the final stages.
The advantages and disadvantages of each blading type are noted as
­follows:
Type
Advantages
Disadvantages
Impulse
•
•
•
•
•
•
•
•
Reaction
Rugged
No thrust forces
Higher efficiency
Light weight per stage
Lower efficiency
Heavy weight per stage
More stages required
Significant thrust force per stage
Fig. 5.3.3 shows an example of a hybrid turbine design that incorporates
impulse blading in the first stage and reaction blading in all other stages.
The principle of operation of any type of expansion turbine is shown in
Fig. 5.3.4. The thermodynamic expansion process is exactly the opposite of the
ideal thermodynamic compression process used in turbo-compressors. An Isentropic (Adiabatic) reversible expansion is commonly used to determine steam
turbine performance.
There are many different types of expansion vapors used. Steam is the most
common. Table 5.3.1 presents some other commonly used expansion vapors.
FIGURE 5.3.3 Steam turbine with impulse and reaction type blading.
236
Steam Turbines
Chapter | 5
FIGURE 5.3.4 Energy extraction in a turbine.
TABLE 5.3.1 Types of Expansion Vapors
The Types of Expansion Vapors Available are Many. They can be Grouped as Follows:
STEAM
CRYOGENIC
FIRED VAPORS (GAS GENERATED)
• Ethylene
• Diesel Fuel and Air
• Hydrogen
• Natural Gas and Air
• Nitrogen
• Refinery Gas and Air
Naturally, the Vapors with the Highest Amount of Energy (Btu’s or Joules) Per Unit Mass
Will be Used.
Steam Turbine Types
Fig. 5.3.5 contains an assembly drawing of a Single and Multi-Stage steam
turbine.
Note that the single stage turbine shown actually has two blade rows. This
arrangement in known as a “Curtis” stage and is used in single stage and some
older design multi-stage turbines to reduce the blade loading. All single stage
steam turbines contain one Curtis stage.
The limiting factors for a single stage expansion turbine are shown in
­Table 5.3.2.
The approximate power limitations for single and multi-stage turbines are
noted in Table 5.3.3.
237
B.P. 5.3
238
More Best Practices for Rotating Equipment
FIGURE 5.3.5 Single stage vs. multi-stage steam turbines.
Steam Turbines
Chapter | 5
TABLE 5.3.2 Single Versus Multi-Stage
The Energy Extracted From the Vapor Is Limited By Aerothermal and Mechanical
Factors:
AEROTHERMAL
MECHANICAL
• Exhaust Moisture Content
• Blade Bending Stresses
• Nozzle And Blade Velocities
• Blade Attachment (Root) Stresses
• Blade Incident Angles
• Blade Disc Stresses
TABLE 5.3.3 Single Versus Multi-Stage (Continued)
As a Result:
• Single Stage—Limited To Approximately 2000 H.P. (1500 Kw)
• Multi-Stage—Above 2000 H.P. (1860 Kw)
The objective of any expansion turbine is to extract the maximum possible
energy from each pound of vapor to produce power. The inlet throttle valves
control the amount (mass flow) of steam admitted to the expansion turbine to
meet the power requirements of the driven equipment. However, to maximize
energy extraction, the losses across the inlet throttle valves must be minimum.
Pictured in Fig. 5.3.6 is a Single Valve and a Multi-Valve steam turbine.
The factors that determine the choice of a single valve or a multi-valve
steam turbine are presented in Table 5.3.4 and Fig. 5.3.7.
Schematics depicting a backpressure and condensing steam turbine are
shown in Fig. 5.3.8. In a backpressure turbine, the exhaust pressure is greater
FIGURE 5.3.6 Single valve vs. multi-valve steam turbines.
239
B.P. 5.3
More Best Practices for Rotating Equipment
TABLE 5.3.4 Single Valve and Multi-Valve Steam Turbines
The Choice Depends On:
• Steam Flow Requirements
• Operating Requirements
• Cost Of Steam (Efficiency)
FIGURE 5.3.7 Single vs. multi-valve performance.
FIGURE 5.3.8 Backpressure vs. condensing steam turbines.
than atmospheric pressure. A condensing turbine’s exhaust pressure is equal to
or less than atmospheric pressure. Most condensing turbines operate at a high
vacuum (3–4 in. Hg Vac) for maximum efficiency and energy extraction.
Tables 5.3.5 and 5.3.6 present facts concerning the definition and selection
of backpressure and condensing steam turbines.
240
Steam Turbines
Chapter | 5
TABLE 5.3.5 Backpressure/Condensing Steam Turbines
• The greater inlet to exhaust ∆P, the greater amount of potential energy per pound of
vapor
• Backpressure turbine exhaust pressures are above atmospheric pressure
• Condensing turbine exhaust pressures are below atmospheric pressure
TABLE 5.3.6 Backpressure/Condensing Steam Turbine
The Choice Depends on:
• The cost of steam
• The plant steam balance
• The capital cost of auxiliaries (condenser, condensate system, etc.)
Extraction and Admission turbines are used to optimize the cycle efficiency
for a given plant steam balance. Fig. 5.3.9 shows a Single Extraction and single
Admission steam turbine.
By using an Extraction or Admission turbine steam pressure can be efficiently reduced to a desired pressure level or excess steam can be utilized to
produce power. Either type of turbine is actually a series combination of either
backpressure or condensing type turbines. Each turbine section is supplied with
a throttle valve.
In Fig. 5.3.10 a Single Flow and Double Flow steam turbine are shown. In
large steam turbines the volume of exhaust steam is too large for one stage of
blading to accommodate due to excessive blade length. As a result, the flow is
FIGURE 5.3.9 Extraction (left) and admission (right) steam turbines.
241
B.P. 5.3
More Best Practices for Rotating Equipment
FIGURE 5.3.10 Single flow (left) and double flow (right) steam turbines.
divided into two stages. In some very large turbines, a triple flow back end can
be used for the same purpose.
Steam Turbine Applications
Tables 5.3.7–5.3.10 present lists of typical applications arranged according to
number of stages, number of valves, exhaust end design and extraction/admission design.
TABLE 5.3.7 Steam Turbine Applications
Single stage
Multi-stage
• Process Pump Drive
• Generator Drive
• Boiler Fan Drive
• Boiler Feed Pump Drive
• Generator Drive
• Cooling Water Pump Drive
• Lube/Seal Oil Pump Drive
• Turbo-Compressor Drive
TABLE 5.3.8 Steam Turbine Applications (Continued)
Single valve
Multi valve
• Process Pump Drive
• Boiler Feed Pump Drive >3000 H.P.
(746 KW)
• Boiler Fan Drive
• Compressor Drive >3000 H.P.
(1860 KW)
• Lube/Seal Oil Pumps
• Generator Drive
• Generator Drive (Small co-gen HRSG)
• Cooling Water Pump Drive
• Compressor Drive <3000 H.P. (1860 KW)
242
Steam Turbines
Chapter | 5
TABLE 5.3.9 Steam Turbine Applications (Continued)
Backpressure
Condensing
• Process Pump Drive
• Compressor Drivea
• Boiler Fan Drive
• Generator Drivea
• Boiler Feed Pump Drivea
• Lube/Seal Oil Pump Drive
• Cooling Water Pump Drive
a
• Compressor Drivea
• Boiler Feed Pump Drivea
a
Choice Depends On Power Level (Approximately >5000 Bhp (3700 Kw) and Plant Steam Balance.
TABLE 5.3.10 Steam Turbine Applications (Continued)
Extraction/Admission
• Generator Drivea
• Compressor Drivea
• Boiler Feed Pump Drivea
a
Approximately >15,000 H.P. (11,000 Kw).
This concludes the steam turbine types module. Attendees are reminded to
complete the workbook problem in this section and refer to the site-specific
turbine drawings in the manual.
B.P. 5.4 Operate condensing turbines at specified exhaust pressure.
Lower pressure at the exhaust of the turbine will obviously produce more
power. However, more power is not better when more moisture is produced as
well, which is the case when the condenser pressure becomes lower (more of a
vacuum).
If more power is desired, it is always better to maximize the energy at the
inlet of the turbine.
See B.P. 5.6 to determine if operation at a lower condensing pressure has
caused blade/nozzle corrosion.
L.L. 5.4: Operation of condensing turbines at exhaust pressures lower than
specified on the data sheets will result in more moisture and higher rate of
blade/nozzle corrosion on the back end.
BENCHMARKS
This best practice has been in use since the mid 1990s and has resulted in maximum reliability of condensing (and extraction/condensing) type steam turbines.
243
B.P. 5.5
More Best Practices for Rotating Equipment
SUPPORTING MATERIAL
See supporting material for B.P. 5.2.
B.P. 5.5: Trend After 1st Stage (and after extraction stage for extraction
turbines) in “Real Time”.
This requires a transmitter to be installed in these areas within the turbine,
therefore it should be specified early on in the project to have these instruments
installed (most vendors will supply this without asking these days).
While calculating and trending efficiency of the turbine is useful in indicating performance drop, it does not indicate why the performance has degraded (it
could be fouling or internal wear). If for a given steam inlet flow rate, the after
1st stage pressure has increased, this indicates that fouling has occurred within
the turbine. When it is known that fouling has occurred, you can typically wash
and regain a good amount of efficiency without having to open up the turbine.
By trending this value, you will know as soon as fouling is occurring, giving
you a better shot of washing it off and regaining as much efficiency as possible.
L.L. 5.5: Failure to trend After 1st Stage Pressure can result in an abundance of fouling that could cause an unplanned shutdown by tripping on
vibration when the fouling breaks off of the rotor.
It has been seen a number of times when the after first stage pressure was
not monitored and fouling accumulated so much that when it eventually breaks
off a large unbalance is produced and the turbine trips on vibration. This will
cause loss of production that most likely could have been avoided if the fouling
was detected earlier.
BENCHMARKS
This best practice has been in use since 1990 and has saved numerous unplanned
shutdowns by identifying and dealing with turbine fouling early on.
SUPPORTING MATERIAL
The Mechanism of Fouling
As mentioned earlier, one can reduce any blade row or impeller to a series of
equivalent orifices. Flow is a function of area and velocity.
Whenever any blade row or impeller is designed, the designer sets inlet and
discharge blade areas such that optimum velocities relative to the blade will be
achieved at teach location. By a series of tests and experience, designers have
defined optimum relative velocity rather well. Therefore the resulting inlet and
discharge areas will produce optimum velocities and corresponding optimum
impeller efficiencies. If however, the areas were to change, and flow passages
were to become rough and non-continuous, an impeller performance change
244
Steam Turbines
Chapter | 5
FIGURE 5.5.1 Fouling—the effect on the operating point. (A) Impeller—side view. (B) Impeller
—­side plate removed.
would result. Fig. 5.5.1 shows the effect of fouling on a closed centrifugal
­impeller.
Fouling is defined as the accumulation of debris in the impeller or blade passage that reduces the flow area and roughens the surface finish. The distribution
of the foulant on the impeller or blade row is non-uniform and usually changes
with time. Flow patterns within the impeller or blade cause unequal distribution.
In addition, the forces exerted on the foulant cause it to chip off with time as it
becomes dry and brittle. This results in a change in rotor balance and a change
in performance (head and efficiency).
The Effect of Fouling on the Operating Point
If we refer back to the previous example of a backward leaning centrifugal
compressor impeller, the effect of fouling can be understood. Fig. 5.5.2 shows
the effect of fouling on the relative velocity.
Since the area of the flow passage is reduced when the impeller is fouled,
VREL will increase, the flow angle ∝ will increase and therefore result in an absolute velocity (increased R) as shown in Fig. 5.5.3.
The increase in ∝ and R due to fouling will reduce the tangential velocity of
the gas as shown in Fig. 5.5.4.
Since the head (energy) produced by the impeller is the product of the impeller tip speed “U,” which does not change in the fouled condition, and the
tangential velocity which is reduced, the head produced will be reduced in the
fouled condition. In addition, the non-uniform distribution of the foulant will
reduce the efficiency of the impeller stage.
Fig. 5.5.5 shows the effect of fouling on the impeller stage curve. Impeller
fouling is the accumulation of material in the impeller passages that reduces
flow area and roughens surface finish. It reduces impeller head capacity and
efficiency.
245
B.P. 5.5
More Best Practices for Rotating Equipment
FIGURE 5.5.2 Impeller with side plate removed.
FIGURE 5.5.3 Impeller with side plate removed.
Note that the surge margin actually increases slightly in the fouled condition. This is because the cause of surge is low gas velocity. Since the area
of the flow passage is reduced, the gas velocity increases thus increasing the
surge margin. The surge margin is defined as the flow at surge divided by the
impeller design flow. However, the stage head produced by the impeller at any
flow rate is reduced. Therefore, for the same process system head required, the
impeller flow rate will be reduced thus forcing the operating point closer to the
surge line.
While this example has been shown for a centrifugal compressor stage, fouling has the same effect on a steam turbine stage. Steam velocity in the blades is
increased, efficiency is reduced and less power is produced.
246
Steam Turbines
Chapter | 5
FIGURE 5.5.4 Impeller with side plate removed.
FIGURE 5.5.5 Impeller fouling.
The Causes of Fouling in Steam Turbines
The causes of fouling are described in Table 5.5.1.
Fig. 5.5.6 shows a typical after 1st stage pressure curve for the high-pressure
section of an extraction/condensing steam turbine.
Measurement of steam flow and after first stage pressure in the high-pressure
section will enable plotting of the operating point on this curve. If the operating
247
B.P. 5.5
More Best Practices for Rotating Equipment
TABLE 5.5.1 The Causes of Fouling in Steam Turbines
• Boiler upsets resulting in deposits of calcium and/or silica on turbine blades
• Improper boiler feed water treatment resulting in calcium and/or silica deposits on
turbine blades
FIGURE 5.5.6 Typical “after 1st stage turbine curve.” (Courtesy of MHI)
248
Steam Turbines
Chapter | 5
point is above the curve, which is based on clean turbine steam path conditions,
fouling is present assuming flow and pressure measurement are accurate.
For this reason, trending is advised and confirmation of pressure/flow instrumentation is advised prior to taking corrective action (on-line or off-line water
washing).
B.P. 5.6: Use after 1st stage (or after extraction pressure for extraction/
condensing) pressure to determine blade/nozzle corrosion for condensing
turbines.
When trending after 1st stage (and after extraction stage for extraction condensing turbines) pressure in conjunction with B.P. 5.5, a decrease in pressure
for the same steam flow rate indicates that the back end blades/nozzles of the
turbine are opening up. This is indication of corrosion due to excessive moisture
in the steam toward the back end of the turbine.
If this has been observed, provisions should be made immediately to increase condenser pressure until the after 1st stage pressure stops decreasing for
a constant steam flow rate. This will give the user time to operate and plan for
work during the next scheduled turnaround.
Be sure to have spare nozzles available as these are generally where the corrosion will occur the most. It is highly recommended to upgrade to SS Nozzles
on the last 3 stages if corrosion has been observed toward the exhaust of a condensing type turbine.
Note that operating the condenser below specified pressure (See B.P. 5.4)
will increase the rate of corrosion.
L.L. 5.6: Failure to identify and trend rate of blade/nozzle corrosion can
result in unplanned shutdowns and production lost.
Many times plants are lucky and find out they have had blade/nozzle corrosion during the turnaround and not during an emergency shutdown, but it has
happened and will result in production lost. Even if you are lucky to find this
out during the turnaround, you will not be anticipating the corrosion (since you
were not trending the after 1st stage pressure) and hence will likely not have the
proper spare parts available (many times spare nozzles are not carried).
BENCHMARKS
The writer has used this best practice more and more in the last ten years when
plants have been pushed to operate at higher rates for more production and condenser pressures lowered on the back end of condensing steam turbines. It has
resulted in optimal reliability for trains driven by condensing (and extraction/
condensing) steam turbines.
SUPPORTING MATERIAL
See B.P. 5.5 for supporting material.
249
B.P. 5.7
More Best Practices for Rotating Equipment
B.P. 5.7: Require a thrust analysis on all condensing turbine applications
and install a thrust balance device where necessary.
Condensing turbines will incorporate reaction type blading in the last few
stages (at least) due to the increased efficiency. If incorporated in a turbine design that utilizes mostly impulse blading, the thrust may be unexpectedly higher
when operating due to the extra thrust forces added by the reaction stages.
A detailed thrust analysis will determine what type (if any) of balance device
would be required to balance out the thrust loading. Options include an undercut
in the shaft or increasing the inlet end steam seal diameter. The analysis and
conclusions need to be conducted early on in the project phase (before priced
bid has been submitted) in order to avoid large cost adders.
L.L. 5.7: Inability to request a thrust analysis for impulse type condensing
turbines have resulted in high thrust loading and excessive bearing pad
temperatures/wear.
This began occurring in the 1980s when impulse turbine manufacturers added reaction type blading on the back end. Bearing pad temperatures exceeded
110°Celsius at initial start-up and bearing life was not optimal until rotor modifications were made.
BENCHMARKS
This best practice has been in use since the 1990s to assure impulse turbine
thrust pad temperatures were within acceptable levels without the use of “Band
Aid” modifications (going after effect not cause). The best practice has produced steam turbines with trouble free thrust bearing operation and thrust bearing MTBF’s exceeding 100 months.
SUPPORTING MATERIAL
Rotor thrust balance
Fig. 5.7.1 shows how a balance drum or opposed impeller design reduces thrust
force. The total impeller force is the sum of the forces from the individual impellers. If the suction side of the impellers is opposed, as noted in Fig. 5.7.1,
the thrust force will be significantly reduced and can approach 0. If the suction
side of all impellers are the same (in series), the total impeller thrust force can
be very high and may exceed the thrust bearing rating. If this is the case, a balance drum must be mounted on the rotor as shown in Fig. 5.7.1. The balance
drum face area is varied such that the opposing force generated by the balance drum reduces the thrust bearing load to an acceptable value. The opposing
thrust force results from the differential between compressor discharge pressure
(PF) and compressor suction pressure (P1) since the area behind the balance
drum is usually referenced to the suction of the compressor.
250
Steam Turbines
Chapter | 5
FIGURE 5.7.1 Rotor thrust force.
This is accomplished by a pipe that connects this chamber to the compressor
suction. This line is typically called the ‘balance line’.
It is very important to note that a balance drum is used only where the thrust
bearing does not have sufficient capacity to absorb the total compressor axial
load. And the effectiveness of the balance drum depends directly on the balance
drum seal. Fail the seal, (open clearance significantly) and thrust bearing failure
can result.
A common misunderstanding associated with balance drum systems is that
a balance drum always reduces the rotor thrust to zero. Refer to Fig. 5.7.2 and
observe that this statement may or may not be true depending on the thrust
balance system design. And even if it is, the thrust is zero only at one set of
operating conditions.
Fig. 5.7.2 shows a rotor system designed four different ways. Note how the
thrust always changes with the flow rate regardless of the design. Another misconception regarding thrust balance systems is the normal or “active” direction
of thrust. In many cases, the active thrust is assumed to always be toward the
suction of the compressor.
Observing Fig. 5.7.2, it is obvious that the “active” direction can change
when the turbo-compressor has a balance drum or is an opposed design. It is
recommended that the use of active thrust be avoided where possible and that
251
B.P. 5.7
More Best Practices for Rotating Equipment
FIGURE 5.7.2 Rotor system designed four different ways.
axial displacement monitors be labeled to allow determination of the thrust direction at all times.
Please refer to Fig. 5.7.3, which shows a typical thrust displacement
­monitor.
FIGURE 5.7.3 Typical axial thrust monitor.
252
Steam Turbines
Chapter | 5
These monitors detect thrust position by targeting the shaft end, thrust collar or other collar on the rotor. Usually two or three probes (multiple voting
arrangement) are provided to eliminate unnecessary compressor trips. The output of the probes is noted on the monitor as either + (normal) or –(counter).
However, this information gives no direct indication of the axial direction of the
thrust collar. The following procedure is recommended:
1. With compressor shutdown, push rotor toward the suction and note direction
of displacement indicator.
2. Label indicator to show direction toward suction of compressor.
Knowing the actual direction of the thrust can be very useful during troubleshooting exercises in determining the root cause of thrust position changes.
B.P. 5.8: Always require ratchet type turning gear device on compressor
drives.
A ratchet type turning gear will rotate the shaft a certain amount of degrees
every few minutes (frequency determined in order to prevent rotor sag). The
slow rotation of this device will produce zero wear on the Dry Gas Seals.
L.L. 5.8: The use of continuous speed turning gear devices has resulted in
premature dry gas seal failures.
A slow speed of 100 RPM may not be enough to initiate lift off (when faces
do not contact) of the dry gas seals but is enough to cause excessive wear and
reduce the life of the DGS.
If a continuous peed turning gear is used, it is essential to work with the seal
vendors to assure DGS lift off speed is less than turning gear speed.
BENCHMARKS
This BP has been used extensively since the mid 1990s when the majority of
new compressors were sold with dry gas seals and has helped maximize DGS
reliability to exceed 96 months MTBF.
SUPPORTING MATERIAL
The term “critical speed” is often misunderstood. In nature, all things exhibit a
natural frequency. A natural frequency is defined as that frequency at which a body
will vibrate if excited by an external force. The natural frequency of any body is a
function of the stiffness and the mass of that body. As mentioned, for a body to vibrate, it must be excited. A classical example of natural frequency excitation is the
famous bridge “Galloping Gerty” in the state of Washington. That bridge vibrated
to destruction when its natural frequency was excited by prevailing winds.
In the case of turbo-compressor rotors, their natural frequency must be excited by some external force to produce a response that will result in increased
253
B.P. 5.8
More Best Practices for Rotating Equipment
amplitude of vibration. One excitation force that could produce this result is
the speed of the rotor itself. Thus the term “critical speeds.” The term “critical speed” defines the operating speed at which a natural frequency of a rotor
system will be excited. All rotor systems have both lateral (horizontal and vertical) and torsional (twist about the central shaft axis) natural frequencies. Only
lateral critical speeds will be discussed in this section.
In the early days of rotor design, it was thought that the rotor system consisted primarily of the rotor supported by the bearings. This led to the assumption that only the stiffness of the rotor supported by rigid bearings needed to be
considered in the analysis of the natural frequency. Countless machinery problems have proven this assumption to be false over the years. The concept of the
“rotor system” must be thoroughly understood. The rotor system consists of the
rotor itself, the characteristics of the oil film that support the rotor, the bearing,
the bearing housing, the compressor case that supports the bearing, compressor
support (base plate), and the foundation. The stiffness and damping characteristics of all of these components together result in the total rotor system that
produce the rotor response to excitation forces.
We will examine a typical rotor response case in this section and note the
various assumptions, the procedure modeling, the placement of unbalance, the
response calculation output, and discuss the correlation of these calculations to
actual test results.
Critical Speeds
The natural frequency of any object is defined by the relationship:
FNATURAL =
K
M
Where K, stiffness; M, mass.
When excited by an external force, any object will vibrate at its natural frequency. If the frequency of the exciting force is equal to the natural frequency
of the object, and no damping is present, the object can vibrate to destruction.
Therefore, if the frequency of an exciting force equals the natural frequency of
an object, the exciting force is operating at the “critical frequency.”
Rotor speed is one of the most common external forces in turbo-machinery.
When the rotor operates at any rotor system natural frequency, it is said that the
rotor is operating at its critical speed. The critical speed of a rotor is commonly
designated as NC and the corresponding natural frequencies or critical speeds
are: NC1, NC2, NC3, etc.
Every turbo-compressor that is designed must have the rotor system critical
speeds determined prior to manufacture. In this section, we will follow the procedure for the determination of the necessary parameters to define a rotor systems
critical speed. The procedure is commonly known as determination of rotor response. Fig. 5.8.1 is a representation of a critical speed map for a rotor system.
254
Steam Turbines
Chapter | 5
255
FIGURE 5.8.1 Compressor rotor critical speed map—no damping. (Courtesy of Elliott Company)
B.P. 5.8
More Best Practices for Rotating Equipment
It should be understood that all stiffness values are “calculated” and will
vary under actual conditions. As an exercise, determine NC1, NC2, and NC3
for the horizontal and vertical directions for each bearing in Fig. 5.8.1 (assume
bearing 1 and 2 stiffness are the same).
Critical speed
Horizontal (X)
Vertical (Y)
NC1
NC2
NC3
3,300 rpm
9,700 rpm
16,000 rpm
3,000 rpm
8,000 rpm
15,000 rpm
Based on a separation margin of ± 20% from a critical speed, what would be
the maximum allowable speed range between NC1 and NC2 in Fig. 5.8.1?
j
j
Maximum speed
Minimum speed
6600 rpm
4000 rpm
Remember, changing of any value of support stiffness will change the critical speed. Plotted on the X-axis is support stiffness in lbs/inch. The primary
components of support stiffness in order of decreasing increasing influence are:
j
j
j
j
j
j
j
Oil support stiffness
Bearing pad or shell
Bearing housing
Bearing bracket
Casing support foot
Baseplate
Foundation
Note that this analysis of the critical speed does not include oil film damping. It
is common practice to first determine the “undamped critical speeds” to allow for
necessary modifications to the rotor or support system. This is because the effects
of stiffness on the location of critical speed are significantly greater than damping.
Fig. 5.8.1 shows four distinct critical speeds. Operation within ±20 of actual critical
speeds is to be avoided. Also plotted are the horizontal (x) and vertical (y) bearing
stiffness for each bearing. Note that these values vary with speed and are the result
of changes in the oil stiffness. Therefore, a change in any of the support stiffness
components noted above can change the rotor critical speed. Experience has shown
that critical speed values seldom change from ±5% of their original installed values.
If a turbo-compressor with oil seals experiences a significant change in critical speeds, it is usually an indication of seal lock-up. That is, the seal does not
have the required degrees of freedom and supports the shaft acting like a bearing. Since the seal span is less than the bearing span, the rotor stiffness “K”
increases and the critical speeds will increase in this case.
The Rotor System (Input)
Fig. 5.8.2 shows a typical turbo-compressor rotor before modeling for critical
speed or rotor response analysis.
256
Steam Turbines
Chapter | 5
FIGURE 5.8.2 Rotor response modeling—rotor. (Courtesy of Elliott Co.)
Since the natural frequency or critical speed is a function of shaft stiffness
and mass, Fig. 5.8.3 presents the rotor in Fig. 5.8.2 modeled for input to the
computer rotor response program.
Fig. 5.8.3 is an example of a modeled rotor and only includes the rotor stiffness (K) and mass (M).
In order to accurately calculate the rotor critical speeds, the entire rotor system stiffness, masses and damping must be considered. Table 5.8.1 models the
oil film stiffness and damping of the journal bearings at different shaft speeds.
Note that it is essential that the type of oil to be used in the field (viscosity
characteristics) must be known. End users are cautioned to confirm with the
OEM before changing oil type as this will affect the rotor response. In addition
to modeling of the rotor and bearings, most rotor response calculations also
include the following additional inputs:
j
j
Bearing support stiffness
Oil film seal damping effects
Of all the input parameters, the effects of bearing and seal oil film parameters are the most difficult to calculate and measure. Therefore, a correlation
difference will always exist between the predicted and actual values of critical
speed. Historically, predicted values of NC1 (first critical speed) generally agree
within ± 5%. However, wide variations between predicted and actual values
above the first critical speed (NC1) exist for NC2, NC3, etc.
When selecting machinery, the best practice is to request specific vendor
experience references for installed equipment with similar design parameters
as follows:
j
j
j
j
j
Bear span ÷ major shaft diameter
Speeds
Bearing design
Seal design
Operating conditions (if possible)
Once the rotor system is adequately modeled, the remaining input parameter
is the amount and location of unbalance. Since the objective of the rotor response study is to accurately predict the critical speed values and responses, an
assumed value and location of unbalances must be defined. Other than b­ earing
257
B.P. 5.8
258
More Best Practices for Rotating Equipment
FIGURE 5.8.3 Rotor response input data—dimensions, masses, and unbalances. (Courtesy of Elliott Co.)
Chapter | 5
Steam Turbines
TABLE 5.8.1 Typical Compressor Oil Film Bearing Parameters
4 × 1.6 in. tilt 20.5 in. TB 3.0 in. shaftend 7.5–6.5 in. shaft Bendix coupling
Static bearing load (lbs)
897
Diameter (inches)
4.00
Bearing station
12
Length (inches)
1.60
Bearing location
thrust
Diam assembly
clearance (inches)
5.7487E–03
Bearing type
tilt pad
Diam machined
clearance (inches)
8.7500E–03
Location of load
between
pads
Inlet oil temperature
(deg F)
120.0
Preload
0.343
Type of oil
DTE–light
(150SSU @100°F)
Fluid film stiffness
Damping
Speed
(rpm)
50 mm
No
KXX (Ib/in)
KYY (Ib/in)
WCXX
(Ib/in)
WCYY
(Ib/in)
2500
0.114
1.3871E 06
7.5446E 05
7.7995E 05
4.6249E 05
3000
0.137
1.2984E 06
7.1330E 05
7.8487E 05
4.7587E 05
4000
0.183
1.1769E 06
6.6147E 05
8.0311E 05
5.0825E 05
4500
0.206
1.1341E 06
6.4543E 05
8.1400E 05
5.2564E 05
5500
0.252
1.0703E 06
6.2556E 05
8.3686E 05
5.6116E05
6613
0.303
1.0230E 06
6.1679E 05
8.6656E 05
6.0354E 05
7000
0.321
1.0109E 06
6.1616E 05
6.7775E 05
6.1885E 05
8000
0.366
9.8751E 05
6.1898E 05
9.0798E 05
6.5935E 05
9000
0.412
9.7305E 05
6.2684E 05
9.4015E 05
7.0111E 05
10000
0.458
9.6556E 05
6.3864E 05
9.7461E 05
7.4430E 05
11000
0.504
9.6360E 05
6.5354E 05
1.0110E 06
7.8878E 05
12000
0.549
9.6610E 05
6.7094E 05
1.0490E 06
8.3434E 05
13000
0.595
9.7225E 05
6.9037E 05
1.0881E 06
8.8080E 05
14000
0.641
9.8144E 05
7.1149E 05
1.1283E 06
9.2801E 05
15000
0.687
9.9317E 05
7.3403E 05
1.1696E 06
9.7586E 05
Courtesy of Elliott Co.
259
B.P. 5.8
More Best Practices for Rotating Equipment
and seal parameters, unbalance amount and location is the other parameter with
a “correlation factor.” There is no way to accurately predict the amount and
location of residual unbalance on the rotor. Presently, the accepted method is to
input a value of 8 × A.P.I. acceptable unbalance limit (4W)/N.
This results in a rotor response input unbalance of (32W)/N.
The location of the unbalance is placed to excite the various critical speeds.
Typically the unbalances are placed as noted below:
Location
To excite
Mid span
Quarter span (two identical unbalances)
At coupling
NC1
NC2
NC2, NC3
Failure to accurately determine the value and location of residual rotor unbalance is one of the major causes of correlation differences between predicted
and actual critical speeds.
Rotor Response (Output)
The output from the rotor response study yields the following:
j
j
Relative rotor mode shapes
Rotor response for a given unbalance
Fig. 5.8.4 shows the relative rotor mode shapes for NC1, NC2, NC3, and NC4.
Usually, the rotor will operate between NC1 and NC2.
Rotor mode shape data is important to the designer because it allows determination of modifications to change critical speed values.
For the end user, this data provides an approximation of the vibration at any
point along the shaft as a ratio of the measured vibration data. As an example
in Fig. 5.8.4, determine the vibration at the shaft mid span if the vibration measured by the probe C2 when operating at NC1 is 2.00 mils. From Fig. 5.8.4, the
vibration at the shaft mid span when operating at the first critical speed of 3327
RPM (50 in location) is:
1.00
or10 × the bearing vibration
0.1
Ten times the value at C2 or 20.0 mils!
Mode shape data should always be referred to when vibration at operating
speed starts to increase and your supervisor asks
“When do we have to shut down the unit?”
or
“Can we raise the radial vibration trip setting?”
In this example, the bearing clearance may be 0.006 or 6 mills. And an honest request would be … “We’ll replace the bearing at the turnaround, please run
to 7.0 mils vibration.”
260
Steam Turbines
Chapter | 5
261
FIGURE 5.8.4 Rotor natural frequency mode shapes. (Courtesy of Elliott Co.)
B.P. 5.8
More Best Practices for Rotating Equipment
Refer to Fig. 5.8.4 and remember:
j
j
j
The compressor must go through NC1
The shaft vibration increases at NC1 (usually 2X, 3X, or more)
The vibration at center span is approximately 10X the probe vibration
Therefore,
Vibration at the mid span during the first critical speed will be:
= (7.0 mils)
Probe value
= 140 mils!!
×
(2.0)
NC1 amplification
×
(10)
Mode shape difference
Normal clearance between the rotor and interstage labyrinths is typically
40 mils!! This vibration exposes the diaphragms, which are usually cast iron,
to breakage. One final comment … during shutdown, the rate of rotor speed
decrease CANNOT be controlled as in the case of start-up. It depends on rotor
inertia, load in the compressor, the process system characteristics and the control and protection system. If the vibration at the probe locations is high, the best
advice is to stop the compressor fully loaded which will reduce the time in the
critical speed range as much as possible. Yes, the compressor will surge, but the
short duration will not normally damage the compressor. Figs. 5.8.5 and 5.8.6
present the primary output of a rotor response study.
Rotor response plots display vibration amplitude, measured at the probes,
vs shaft speed for the horizontal and vertical probes. Note that a response curve
must be plotted for each set of unbalance locations and unbalance amount.
FIGURE 5.8.5 Rotor response output at non-drive end bearing (NDE). (Courtesy of Elliott Co.)
262
Steam Turbines
Chapter | 5
FIGURE 5.8.6 Rotor response output drive end bearing (D.E.). (Courtesy of Elliott Co.)
Fig. 5.8.5 shows the rotor response for the non-drive end (N.D.E.) set of probes
with the first set of unbalance. Fig. 5.8.6 shows the rotor response for the drive end
set of probes (D.E.). The operating speed range of this example is 6000–8000 rpm.
Measured Rotor Response
During shop test, the rotor response of every turbo-compressor rotor is measured
during acceleration to maximum speed and deceleration to minimum speed.
Values are plotted on the same coordinates as for the rotor response analysis.
The plot of shaft vibration and phase angle of unbalance versus shaft speed is
known as a BODE PLOT.
Bode plots represent the actual signature (rotor response) of a rotor for a
given condition of unbalance, support stiffness and unbalance. They indicate
the location of critical speeds, the change of shaft vibration with speed and
the phase angle of unbalance at any speed. A bode plot is a dynamic or transient signature of vibration for a rotor system and is unique to that system for
the recorded time frame. Bode plots should be recorded during every planned
start-up and shutdown of every turbo-compressor. As discussed in this section,
the bode plot will provide valuable information concerning shaft vibration and
phase angle at any shaft speed.
B.P. 5.9: Perform a rotor stability analysis per API 617 to confirm rotor
stability for turbines with VHP inlet steam (above 100 bar or 1500 psi).
This will confirm that any turbine being proposed will not undergo any instability issues in the field, which is a concern at these exceptionally high steam
pressures.
263
B.P. 5.10
More Best Practices for Rotating Equipment
L.L. 5.9: Failure to perform a rotor stability analysis for turbines in VHP
steam service has resulted in continuous vibration issues in the field and
eventual turbine replacement due to the low reliability.
BENCHMARKS
This best practice has been in use for new projects since the late 1990s when a
lot of turbines had inherent high vibrations when installed in VHP steam service. Optimal turbine reliability has resulted in these services.
SUPPORTING MATERIAL
Please refer to Part I sections 4.8.5 and 4.8.6 in API 617 Standard (latest edition
as of writing this book is 8th edition) for details on how to conduct the Level I
and Level II rotor stability analyses.
B.P. 5.10: Always purchase critical service steam turbines with electronic
overspeed (two out of three voting) backup system in order to avoid use of
mechanical overspeed trip systems.
This will ultimately allow for coupled overspeed trip tests as per B.P. 5.11 of
Forsthoffer’s Best Practice Handbook for Rotating Machinery. A coupled overspeed trip test can be performed at a speed below overspeed (and operating speed)
by easily adjusting the trip set-point. The set-point of the main system can be reduced first and checked, then when trip initiates, the main system can be set back
to original setting and the backup system reduced to check it operates properly.
L.L. 5.10: Machinery historical case studies are full of examples of failed
turbines and personnel injury resulting from the failure of turbine overspeed trip devices during the uncoupled overspeed trip checks.
BENCHMARKS
This best practice has been recommended to clients since 2005 when insurance
companies accepted checking overspeed with electronic governors and backup
system in a coupled state.
SUPPORTING MATERIAL
Total Train Control and Protection Objectives
Table 5.10.1 presents the total train control and protection objectives.
Regardless of the type of driven equipment, the objective of the control and
protection system is to assure that the required quantity of product or generated
power is continuously supplied maintaining the highest possible total train efficiency and reliability.
264
Steam Turbines
Chapter | 5
TABLE 5.10.1 Total Train Control/Protection Objectives
j
j
j
j
j
j
Meet driven equipment control requirements
Compressor—pressure or flow
Pump—pressure, flow or level
Generator—load
Meet above objectives when in series or parallel with other trains
Continuously protect entire train from damage due to:
j Overspeed
j Loss of auxiliaries
j Component mechanical failure
j Driven equipment upsets (surge, minimum flow, high load, etc.)
Fig. 5.10.1 presents a process diagram for a steam turbine driven compressor train.
Depending on the selected process variable and location, any PIC or FIC
will continuously monitor the selected process variable sending its signal as
an input signal to the turbine speed controller. For this example, assume the set
point is a flow controller located in the discharge line of the turbo-compressor
(FICD). The process system head (energy) requirements A, B, C are shown.
These different energy requirements can represent either increased pressure ratio requirements (suction strainer blockage exchanger ∆P etc.) and/or gas density changes (M.W. P or T). As the process head (energy) requirements increase
from A to B to C, the input flow variable will decrease if the turbo-compressor
speed does not change. However, as soon as the monitored process variable,
FICD ≠ flow set point, the turbine speed controller output will open the turbine
inlet throttle valves to provide more turbine power to increase the head (energy)
produced by the compressor to meet the additional process system head requirements and therefore maintain the desired throughput.
Adjusting the speed of the driven equipment is the most efficient control
method since there are no control valves required in the system. Therefore only
the exact value of head required by the process system is produced by the turbocompressor.
Also noted in Fig. 5.10.1 are the two major protection systems for the compressor and steam turbine, the surge protection and turbine overspeed protection
systems. The surge system has been previously discussed, the turbine overspeed
system will be discussed later in this chapter. In addition to the two major protection systems mentioned above, other typical protection systems for a rotating
equipment train are:
j
j
j
j
Shaft vibration
Bearing bracket vibration
Axial thrust displacement
Bearing temperature
265
B.P. 5.10
More Best Practices for Rotating Equipment
FIGURE 5.10.1 Total train control.
j
j
j
j
Process gas temperature
Lube oil pressure
Seal oil ∆P
Suction drum high liquid level (compressors)
Control
A turbine governor is a speed controller. Important facts concerning expansion
turbine governors are shown in Table 5.10.2.
Regardless of type, all controllers have three identical parameters:
j
j
j
Input
Set point
Output
Some familiar controllers are:
j
j
j
j
j
j
Pressure
Flow
Level
Temperature
Surge
Speed
As an example, refer to Fig. 5.10.2, which is a speed controller that may be
familiar.
In both cases, load change is inversely related to speed change. The controller compares input to set point and changes output appropriately.
266
Steam Turbines
Chapter | 5
TABLE 5.10.2 Control
j
j
j
The governor is the heart of the control system
The governor in simple terms compares input signal(s) to a set point and sends an
output signal to achieve the desired set point.
An example of a simple governor system is “cruise control” in a car
FIGURE 5.10.2 A control system analogy. (A) Cruise control. (B) Steam turbine governor.
In Fig. 5.10.2, we compare an auto “Cruise Control” to a steam turbine governor (typical single stage mechanical/hydraulic). Both are speed controllers
and have an:
j
j
j
Input
Set point
Output
The table below shows a comparison of these parameters.
Parameter
C.C. (Cruise control)
T.G. (Turbine governor)
Input
Set point
Output
Actual speed from speedometer
Selected by driver
To fuel control system
Actual speed from speed pick-up
Selected by operator
To steam throttle valve
Fig. 5.10.3 is a schematic of a steam turbine governor system.
Note that the set point can either be a manual set point, similar to a driver
setting a “speed” in a cruise control system or a process variable. Examples of
process variable set points would be:
j
j
j
Pressure
Flow
Level (pump applications)
There are many controller designs. Historically, the first controllers were
entirely mechanical. An example of a mechanical speed controller is shown in
Fig. 5.10.4.
267
B.P. 5.10
More Best Practices for Rotating Equipment
FIGURE 5.10.3 Steam turbine control. (Courtesy of M.E. Crane, Consultant)
FIGURE 5.10.4 A mechanical governor system.
268
Steam Turbines
Chapter | 5
Commonly called “Fly Ball Governors,” the input shaft from the driver
would rotate the weights through a gear set. As the weights rotated, centrifugal
force would move the weights outward, compressing the spring and thus moving the output linkage. The tension on the spring from the speed changer (set
point) would control the speed as the equilibrium point of the input and set
point values.
Many mechanical governors are still in use today on older, small single valve
steam turbines. The mechanical governor output force is limited and lead to the
development of the mechanical hydraulic governor pictured in Fig. 5.10.5.
The mechanical-hydraulic governor uses the same mechanical mechanism
to determine the output signal. However, the output shaft moves a pilot valve,
which allows hydraulic fluid (usually oil) to provide the output signal to the
throttle valve(s). The common Woodward “T.G.” and “P.G.” governors are examples of mechanical/hydraulic governors. These governors have internal positive displacement oil pumps driven by the governor input shaft.
All mechanical-hydraulic governors require hydraulic fluid and site preventive maintenance practices must include these governors. They are provided
FIGURE 5.10.5 A mechanical hydraulic governor system.
269
B.P. 5.10
More Best Practices for Rotating Equipment
FIGURE 5.10.6 Typical mechanical-hydraulic governor for turbine drive. (Courtesy of
­Elliott/Woodward)
with a sight glass to indicate the operating level of the hydraulic fluid. Typical
fluids used are turbine oil and automatic transmission fluid “ATF.” Governor
­instruction books must be consulted for specific hydraulic specifications. In
larger systems, the governor hydraulic fluid reservoir may not be large enough
to provide a sufficient fluid quantity to fill all of the speed governor oil lines.
Readers are cautioned that additional hydraulic fluid may have to be added
during initial start-up and whenever work has been done on the governor system
during a turnaround.
Fig. 5.10.6 is a representation of a mechanical-hydraulic governor system
for a multi-valve steam turbine.
The system shows a Woodward “P.G.–P.L.” governor system. These systems, common in the 1960s and 1970s are still in use today and have provided
extremely reliable service. However, both mechanical and mechanical-­hydraulic
governors receive their input signal via a gear arrangement. Therefore, they
cannot be repaired or removed while the turbine is operating. During the
1970s refinery, petrochemical and gas plant capacities increased significantly.
As a result, the lost product revenue for 1-day downtime for governor repair
became very large (typically $500,000 to over $1,000,000!). Therefore, there
was an ­urgent need for a governor system that could be maintained without
270
Steam Turbines
Chapter | 5
TABLE 5.10.3 Electro-Hydraulic Governors
j
j
j
j
j
j
Do not require a mechanical input signal
Provide extremely accurate control
Provide self diagnostics, fault tolerance and auto-start capability
Require actuator to convert electric output signal to control signal (hydraulic or
pneumatic)
Types:
j Analog
j Digital
Either type can be:
j Non-redundant
j Redundant
j Triple redundant
having to shut down the turbine. The electro/hydraulic governor met this need.
Table 5.10.3 presents the important facts concerning this system.
Since they did not require a mechanical (gear or shaft drive) input signal,
these governors could be exchanged while the operators kept the turbine in the
manual mode. As an analogy, exchanging automatic control valves is the same
procedure. In this case, the operator maintains process conditions by manually
throttling the bypass valve while the automatic control valve undergoes repair.
The first electronic governors were analog type, which required significant
maintenance to change out cards. Digital governors were introduced in the late
1970s and are the only type of speed control used today. As micro-processors
became popular, digital governors also offered the great advantage of redundancy. Redundant and triple redundant governors became very popular because
governors could now automatically transfer on line to allow control to be maintained while the other governor required maintenance. Operator assistance was
no longer required. Fig. 5.10.7 presents a block diagram for an electro-hydraulic
governor systems.
In the 1990s, the trend is to control all process and machinery functions
through the plant central distributed control system. A new chemical plant in
South America is presently designing a D.C.S. system that will control all critical system functions:
j
j
j
j
j
j
Turbine speed control
Process control
Surge protection
E.S.D. systems
On-line monitoring
Emergency pump auto-start
In this design, all critical functions are actuated on the basis of a two out of
three voting system.
271
B.P. 5.10
More Best Practices for Rotating Equipment
FIGURE 5.10.7 Electrohydraulic governor block diagram. (Courtesy of M.E. Crane Consultant)
As previously discussed, extraction turbines are used to optimize plant
steam balance and overall steam cycle efficiency. Table 5.10.4 defines the function of an extraction steam turbine control system.
Both mechanical-hydraulic and electro-hydraulic extraction control systems
are successfully operating in the field. Either design incorporates two or more
governors operating together to meet the control system objectives. Each governors output controls a specific set of throttle valves. In addition, each governor
in an extraction or admission system continuously receives an input signal from
the other governors in the system. Each governor will respond to this input
signal as required to meet all of the control objectives of the governor system.
Mechanical-hydraulic extraction or admission systems have proven to
require a significant amount of adjustment and maintenance due to the high
amount of friction in the systems. Please refer to Fig. 5.10.8, which shows a
mechanical-hydraulic single extraction governor system.
TABLE 5.10.4 Extraction Control
Function: satisfy driven equipment control requirement and provide required
­extraction steam quantity at desired flow or pressure
An extraction control system consists of multiple governors with feed back
272
Steam Turbines
Chapter | 5
273
FIGURE 5.10.8 Mechanical/hydraulic extraction control. (Courtesy of Elliott/Woodward)
B.P. 5.10
More Best Practices for Rotating Equipment
FIGURE 5.10.9 Electrohydraulic extraction control and protection system. (Courtesy of M.E.
Crane, Consultant)
As a result, all new systems incorporate electro-hydraulic governor arrangements as shown in Fig. 5.10.9.
Coupled with redundant features, these systems offer high reliability and
efficient process control. Regardless of the type of governor utilized, mechanicalhydraulic and electro-hydraulic governors must be supplied with a reliable
control oil system. Table 5.10.5 presents the function and frequent problem areas
of hydraulic control systems.
Usually, the hydraulic control system is integral with the lubrication system.
Typical pressure operating ranges for these systems are:
Low pressure
Medium pressure
High pressure
276–690 kPa (40–100 PSI)
827–4137 kPa (120–600 PSI)
Above 4137 kPa (600 PSI)
TABLE 5.10.5 Control Oil System
Function: Continuously provide cool, clean control oil to control and protection system at proper pressure, flow rate and temperature
Frequent problem areas:
j Main to auxiliary pump transfer
j Control oil valve instability
j Instantaneous flow requirement changes (need for accumulator)
274
Steam Turbines
Chapter | 5
TABLE 5.10.6 Steam Turbine Governor System Application Chart
Application-driven
equipment
Speed regulation %
Type of governor system
Spared pump
NEMA A ± 10%
Mechanical (older applications)
Mechanical (hydraulic)
Elector-hydraulic (optional)*
Non-redundant
Fan(s)
NEMA A ± 10%
Mechanical (older hydraulics)
Mechanical hydraulic
Lube/seal oil pump(s)
NEMA A ± 10%
Mechanical/hydraulic
Turbo-compressor
NEMA D ± 0.5%
Electro-hydraulic (post 1980)
Non-redundant
Optional-redundant, triple
redundant
Generator
NEMA D ± 0.5%
Isochronous (0% droop)
Mechanical/hydraulic
Present
Electro-hydraulic
Table 5.10.6 is an application chart showing type of governor classification,
speed regulation and type of governor used.
In general, NEMA A governors are used in general purpose (spared) applications and NEMA D governors are used in special purpose (unspared) applications.
Protection
In the writers’ experience the function of the steam turbine protection system is
often confused with the control system. The two systems are entirely separate.
The protection system operates only when any of the control system set point
parameters are exceeded and steam turbine will be damaged if it continues to
operate. Table 5.10.7 defines the typical protection methods.
A schematic of a multi-valve, multi-stage turbine protection system is shown
in Fig. 5.10.10.
This system incorporates a mechanical overspeed device (trip pin) to shut
down the turbine on overspeed (10% above maximum continuous speed). Centrifugal force resulting from high shaft speed will force the trip lever, which
will allow the spring loaded handle to move inward. When this occurs, the port
in the handle stem will allow the control oil pressure to drain and drop to 0.
The high energy spring in the trip and throttle valve, normally opposed by the
275
B.P. 5.10
More Best Practices for Rotating Equipment
TABLE 5.10.7 Protection
The protection system monitors steam turbine total train parameters and assures safety
and reliability by the following action:
j Start-up (optional) provides a safe, reliable fully automatic start-up and will shut
down the turbine on any abnormality
j Manual shutdown
j Trip valve exerciser allows trip valve stem movement to be confirmed during operation without shutdown
j Rotor overspeed monitors turbine rotor speed and will shut down turbine when
maximum allowable speed (trip speed) is attained
j Excessive process variable signal monitors all train process variables and will shut
down turbine when maximum value is exceeded
FIGURE 5.10.10 Typical steam turbine protection. (Courtesy of Elliott Co.)
c­ ontrol oil pressure will close suddenly (less than 1 s). In this system there are
two other means of tripping the turbine (reducing control oil pressure to 0):
l
l
Manually pushing spring loaded handle
Solenoid valve opening
The solenoid valve will open on command when any trip parameter set point
is exceeded. Solenoid valves are designed to be normally energized to close.
In recent years the industry has required parallel and series arrangements of
solenoid valves to assure increased steam turbine train reliability.
276
Steam Turbines
Chapter | 5
FIGURE 5.10.11 Overspeed detection. (Courtesy of Elliott Co.)
Fig. 5.10.11 shows two popular methods of overspeed protection used in
the past.
Today, most speed trip systems incorporate magnetic speed input signals
and two out of three voting for increased reliability. Table 5.10.8 presents the
devices that trip the turbine internally. That is, they directly reduce the control
TABLE 5.10.8 Internal Protection
j
j
j
Loss of control oil pressure
Spring force automatically overcomes oil force holding valve open (approximate set
point 50–65% of normal control oil pressure)
Manual trip (panic button)
Manually dumps control oil on command
Optional
Turbine excessive axial movement
277
B.P. 5.10
More Best Practices for Rotating Equipment
FIGURE 5.10.12 Steam turbine shut-off valves. (A) Trip and throttle. (B) Trip. (Part A: Courtesy of Gimple Corp. Part B: Courtesy of Siemens)
oil pressure causing a trip valve closure without the need of a solenoid valve
(external trip method).
Two popular types of steam turbine shutoff valves are shown in Fig. 5.10.12.
Both types use a high spring force, opposed by control oil pressure during
normal operation, to close the valve rapidly on loss of control oil pressure.
It is very important to note that the trip valve will only close if the spring
has sufficient force to overcome valve stem friction. Steam system solid build
up, which increases with system pressure (when steam systems are not properly
maintained) can prevent the trip valve from closing.
To assure the trip valve stem is free to move, all trip valves should be manually exercised on line. The recommended frequency is once per month.
All turbine trip valves should be provided with manual exercisers to allow
this feature. Table 5.10.9 presents facts concerning manually exercising a turbine while on line.
TABLE 5.10.9 Online Manual Exercise of Trip Valve
j
j
j
j
278
Trip valve is only as reliable as valve to move
Should periodically (minimum one per month) exercise valve to assure movement
Exercisers will not trip turbine
If valve does not move, must be remedied immediately
Steam Turbines
Chapter | 5
TABLE 5.10.10 Protection System Philosophies
j
j
Most domestic vendors rely only on trip valve to shut off steam supply. (Throttle
valves remain open)
European vendors close both trip and automatic throttle valve on trip signal
Protection system philosophies have tended to vary geographically with
steam turbine vendors. Table 5.10.10 presents these facts.
B.P. 5.11: Live trend steam seal gland condenser vacuum pressure in DCS.
It is critical that the seal gland condenser on special purpose (or general
purpose if equipped with a gland condenser) turbines be maintained lower than
atmospheric pressure in order to assure the steam is sucked down and not directed toward the bearing assembly.
By trending the vacuum pressure the user can see the rate at which the gland
condensing system (usually educator) is wearing and needs to be fixed. It is
highly recommended to use two eductors (with one as a back up which can be
switched to online) in conjunction with this Best Practice, in order to assure
optimal steam seal reliability.
A transmitter is required to be installed for this monitoring but is definitely
worth the small extra cost up front when considering the headaches and failures
that can occur with excessive water contamination in the oil system.
L.L. 5.11: Failure to monitor and trend gland condenser vacuum on special purpose (Un-spared) steam turbines has resulted in gross contamination of the oil systems and reduced bearing life.
BENCHMARKS
This best practice has been in use since the 1990s to prevent excessive contamination of the oil system (Greater than 200 PPM of H2O). When the gland
condenser pressure is properly monitored and trended, steam seal MTBF’s will
exceed 100 months.
SUPPORTING MATERIAL
Shaft End Seals
Facts concerning shaft end seals and sealing systems for critical service noncondensing and condensing turbines are shown in Figs. 5.11.1–5.11.3, and
Table 5.11.1.
The key to successful shaft end seal operation is to continuously maintain
a slight [2–4 cm (5–10 in.) H2O] vacuum in the last chamber of the seal. By
maintaining a vacuum at this location, atmospheric air will enter the seal thus
279
B.P. 5.11
More Best Practices for Rotating Equipment
FIGURE 5.11.1 Expansion turbine shaft end seals (Special purpose (unspared) turbines).
(A) Typical exhaust end seal. (B) Typical inlet end seal. (Part B: Courtesy of IMO Industries)
FIGURE 5.11.2 Gland seals and drains: noncondensing automatic-extraction turbine.
­(Courtesy of IMO Industries)
280
Steam Turbines
Chapter | 5
FIGURE 5.11.3 Grand seals and drains: condensing turbine. (Courtesy of IMO Industries)
TABLE 5.11.1 Steam Turbine Shaft Sealing Systems
• Function: prevent steam from escaping to atmosphere along the shaft and entering
the bearing housing
• Special purpose turbines usually employ a low vacuum [2–4 cm (5–10 in.) H2O
vacuum] to buffer atmospheric end labyrinth with air
• General purpose turbines usually do not employ a vacuum system and do not totally
prevent moisture from entering bearing housing
assuring that steam (moisture) will not enter the bearing bracket and contaminate the oil system.
Condition monitoring of the system vacuum is essential to maintaining moisture free lubrication oil. Many a turbine bearing has failed because of poor seal
system preventive and predictive maintenance practices. “THINK S
­ YSTEM”
and check all components of the seal system frequently.
281
B.P. 5.12
More Best Practices for Rotating Equipment
B.P. 5.12: Determine frequency of trip valve exercising based on steam
system quality, increase frequency if steam quality is off-spec.
Regardless of type of trip valve, there should be a manual type exerciser in
order to regularly check and assure that the valve stem can move when the valve
is needed to shut. This small movement CANNOT trip the turbine and is so
minimal that no effect on performance will be observed.
The reason for regularly exercising trip valves is because of the calcium,
silica and other particles contained within the steam that will want to make
there way through the ath of least resistance. The path of least resistance is between the trip valve stem and packing sine the pressure upstream is inlet steam
pressure and downstream is atmosphere. Therefore, the packing is basically a
filter and the higher the inlet pressure means that more sediment will form on
the packing. So, it is a matter of time until the sediment between the packing
and stem can bind the valve. Regular exercising will break up this sediment and
maintain the valve in good condition.
As mentioned, the higher the steam inlet pressure it would be recommended
to increase frequency of exercising. So for instance, for HP Steam (about 60 bar
or 900 psi) inlet it would be recommended normally to exercise the trip valve
on a monthly basis and for VHP steam it would be normally recommended to
do this on a weekly basis.
For this best practice, it is recommended that when steam quality does not
meet specifications, frequency of exercising be increased over normal recommendations, as more sediment will likely be present in the steam. During the
period that the steam quality does not meet specifications, frequency for VHP
trip valve exercising should be increased to daily and HP trip valve exercising
increased to weekly. Once steam quality has returned to meet specs, then the
frequency of trip valve exercising can return to the normal recommendations.
L.L. 5.12: Failure to exercise trip valves at the proper frequency resulted
in catastrophic machinery failure, personnel lost time and loss of life.
BENCHMARKS
This best practice has been in use since the late 1990s and has resulted in zero
lost time accidents and failure to trip incidents. When not followed, it has resulted in catastrophic machine outage in critical (Un-spared) machinery that has
exceeded 3 months for repair.
SUPPORTING MATERIAL
Protection
The function of the steam turbine protection system is often confused with the
control system. The two systems are entirely separate. The protection system
282
Steam Turbines
Chapter | 5
TABLE 5.12.1 Protection
The protection system monitors steam turbine total train parameters and assures safety
and reliability by the following action:
j Start-up (optional) provides a safe, reliable fully automatic start-up and will shut
down the turbine on any abnormality
j Manual shutdown
j Trip valve exerciser allows trip valve stem movement to be confirmed during
­operation without shutdown
j Rotor overspeed monitors turbine rotor speed and will shut down turbine when
maximum allowable speed (trip speed) is attained
j Excessive process variable signal monitors all train process variables and will shut
down turbine when maximum value is exceeded
operates only when any of the control system set point parameters are exceeded
and steam turbine will be damaged if it continues to operate. Table 5.12.1 defines the typical protection methods.
A schematic of a multi-valve, multi-stage turbine protection system is
shown in Fig. 5.12.1. This system incorporates a mechanical overspeed device
(trip pin) to shut down the turbine on overspeed (10% above maximum continuous speed). Centrifugal force resulting from high shaft speed will force the trip
lever, which will allow the spring loaded handle to move inward. When this
occurs, the port in the handle stem will allow the control oil pressure to drain
FIGURE 5.12.1 Typical steam turbine protection. (Courtesy of Elliott Co.)
283
B.P. 5.12
More Best Practices for Rotating Equipment
FIGURE 5.12.2 Overspeed detection. (Courtesy of Elliott Co.)
and drop to 0. The high energy spring in the trip and throttle valve, normally
opposed by the control oil pressure will close suddenly (less than 1 second). In
this system there are two other means of tripping the turbine (reducing control
oil pressure to 0):
j
j
Manually pushing spring loaded handle
Solenoid valve opening
The solenoid valve will open on command when any trip parameter set point
is exceeded. Solenoid valves are designed to be normally energized to close.
In recent years the industry has required parallel and series arrangements of
solenoid valves to assure increased steam turbine train reliability.
Fig. 5.12.2 shows two popular methods of overspeed protection used in the
past.
Today, most speed trip systems incorporate magnetic speed input signals and two
out of three voting for increased reliability. Table 5.12.2 presents the devices that trip
284
Steam Turbines
Chapter | 5
TABLE 5.12.2 Internal Protection
j
j
j
Loss of control oil pressure
Spring force automatically overcomes oil force holding valve open (approximate set
point 50–65% of normal control oil pressure)
Manual trip (panic button)
Manually dumps control oil on command
Optional
Turbine excessive axial movement
the turbine internally. That is, they directly reduce the control oil pressure causing a
trip valve closure without the need of a solenoid valve (external trip method).
Two popular types of steam turbine shutoff valves are shown in Fig. 5.12.3.
Both types use a high spring force, opposed by control oil pressure during
normal operation, to close the valve rapidly on loss of control oil pressure.
It is very important to note that the trip valve will only close if the spring
has sufficient force to overcome valve stem friction. Steam system solid build
up, which increases with system pressure (when steam systems are not properly
maintained) can prevent the trip valve from closing.
To assure the trip valve stem is free to move, all trip valves should be manually exercised on line. The recommended frequency is once per month for High
Pressure (40 bar) steam systems and daily for Very High Pressure (1000 bar +)
steam systems.
FIGURE 5.12.3 Steam turbine shut-off valves. (A) Trip and throttle. (B) Trip. (Part A: Courtesy
of Gimple Corp. Part B: Courtesy of Siemens)
285
B.P. 5.12
More Best Practices for Rotating Equipment
TABLE 5.12.3 Online Manual Exercise of Trip Valve
j
j
j
j
Trip valve is only as reliable as valve to move
Should periodically exercise valve to assure movement (minimum one per month for
High Pressure (40 bar) steam systems and daily for Very High Pressure (1000 bar +)
steam systems)
Exercisers will not trip turbine
If valve does not move, must be remedied immediately
TABLE 5.12.4 Protection System Philosophies
j
j
Most domestic vendors rely only on trip valve to shut off steam supply (throttle
valves remain open)
European vendors close both trip and automatic throttle valve on trip signal
All turbine trip valves should be provided with manual exercisers to allow
this feature. Table 5.12.3 presents facts concerning manually exercising a turbine while on line.
Protection system philosophies have tended to vary geographically with
steam turbine vendors. Table 5.12.4 presents these facts.
286
Chapter 6
Gas Turbines
B.P. 6.1: Always use, if possible, two shaft gas turbines for mechanical
drive applications
Two shaft gas turbines have the great advantage over single shaft gas turbines, since the power turbine shaft is separate from the gas generator shaft.
This allows for a SIGNFICANTLY smaller and much more reliable starting
motor.
Progress has been made within the last decade to increase the availability
of two shaft gas turbines for high power applications and currently there are
options in the industry for two shaft gas turbines producing over 100 MW ISOrated power.
L.L. 6.1: Large starting motors (sometimes over 50 MW) have inherently
lower reliability and have resulted in inability to start up on time.
BENCHMARKS
This best practice has been in use since about 2005 when Mega Plants were using very large single shaft gas turbines to drive process units and issues arose
due to the very large starter motors. When able to use dual shaft gas turbines,
starting motor reliability was optimized and in turn optimized gas turbine reliability.
SUPPORTING MATERIAL
Introduction
In this section, we will discuss functions and types of gas turbines. In my personal experience, the gas turbine is the most misunderstood rotating equipment
item. Due to its many support systems and various configurations, the gas turbine is often approached with mystery and confusion. In order to thoroughly
explain the gas turbine from a functional standpoint, we will build on prior
knowledge. We will also compare the gas turbine to an automotive engine in
terms of its combustion cycle. Having done this, we will then use a building
More Best Practices for Rotating Equipment. http://dx.doi.org/10.1016/B978-0-12-809277-4.00006-1
Copyright © 2017 Elsevier Inc. All rights reserved.
287
B.P. 6.1
More Best Practices for Rotating Equipment
block approach to explain the total configuration of a gas turbine and conclude
this introduction by discussing a brief history of its evolution.
Gas turbine classifications will then be presented, specifically:
j
j
j
j
j
Design type
Number of shafts
Drive and number of shafts
Cycle
Drive and location
We will discuss the major design difference between aero derivative and
hybrid (aero derivative gas generator/industrial power turbine). Single and multiple shaft gas turbines will be discussed and reviewed. The three major application cycles for gas turbines: simple, regenerative, and combined will be
presented and discussed.
Finally, we will present applications of different gas turbine types and provide the information concerning when the different types are used.
Comparison to a Steam Turbine
Fig. 6.1.1 shows a typical condensing steam turbine and an industrial type gas
turbine. The major difference is that a steam turbine is an external combustion
engine, while a gas turbine is an internal combustion engine. That is, the motive
fluid for a steam turbine is generated external (in the boiler) to the engine. In
the case of a gas turbine, the motive fluid is generated internal to the engine (air
compressor and combustor).
Fig. 6.1.2 presents the comparison of the gas turbine and steam turbine
cycles. The steam turbine cycle is known as the Rankine cycle. As shown,
the hot vapor is generated in the boiler, which is external to the steam turbine
(expander). In the gas turbine cycle known as the Brayton cycle, air is brought
into the engine by the axial compressor, combined with fuel and an ignition
FIGURE 6.1.1 A gas turbine versus steam turbine. (A) Steam turbine. (B) Mars gas turbine.
(Part A: Courtesy of General Electric Co. Part B: Courtesy of Solar Turbines, Inc)
288
Gas Turbines
Chapter | 6
FIGURE 6.1.2 Comparison—gas turbine versus steam turbine cycles.
source in the combustor to produce a hot vapor which then is expanded through
the HP (high pressure) turbine. The combination of the compressor, combustor
and HP turbine is commonly known as the gas generator. This is because the
function of the gas generator is to generate or produce a hot vapor from the combination of an air fuel mixture. Essentially, a gas generator can be considered
to have the same function as a boiler—both produce a hot vapor. One can think
of the gas generator then as a “rotating boiler.” After the hot vapor is generated, it then is expanded additionally in the power turbine. The power turbine,
therefore, serves exactly the same function as the steam turbine. That is, both
components are hot gas expanders.
Comparison to an Automotive Engine
Fig. 6.1.3 shows the similarities between an automotive engine and a gas turbine. If one considers that a gas turbine is only a dynamic internal combustion
engine, the understanding of the gas turbine becomes significantly easier. As
shown in the referenced figure, an automotive engine is a positive displacement
internal combustion engine having an intake, compression, combustion and exhaust stroke. A gas turbine engine is a dynamic internal combustion engine.
The process in this case is continuous and not intermittent, as is the case for the
automotive engine. Both engines have compression, combustion and exhaust
sections. When one considers the similarities of these engines in this manner, it
289
B.P. 6.1
More Best Practices for Rotating Equipment
FIGURE 6.1.3 Gas turbine versus automotive engine. (Courtesy of Dresser-Rand)
can be seen that both require starters, ignition sources, inlet air filters, inlet fuel
systems, cooling systems, and monitoring systems.
Building a Gas Turbine
We have already discussed turbo compressors and expansion turbines. A gas
turbine is a combination of these components, plus a combustor that produces
the hot gas for expansion. Fig. 6.1.4 presents these facts.
Fig. 6.1.5 shows a gas turbine configuration for a gas turbine with a regenerator. The function of the regenerator is to use gas generator exhaust vapors
to preheat the air exiting the air compressor, thus reducing the amount of fuel
required by the gas turbine. Figure 6.1.5 also shows the changes of gas temperature, pressure, energy, and the horsepower produced in the different sections of
the gas turbine. Note that the power produced in a gas turbine is typically three
times the output power. This is because the air compressor typically requires
two-thirds of the produced power.
FIGURE 6.1.4 Building a gas turbine.
290
Gas Turbines
Chapter | 6
FIGURE 6.1.5 Gas turbine configuration.
History of Gas Turbine Development
A brief history of gas turbine development is presented in Table 6.1.1. Gas turbines were initially used in the early 1940s for military purposes. In the 1950s,
gas turbines first entered mechanical service applications. Since gas turbines
are production type equipment and not custom designed, we usually refer to
the generation of gas turbines. The first generation of gas turbines begins in
the 1950s and progresses through the 1970s (second generation) 1980s (third
generation) and present day efficiency improvements.
Gas Turbine Classifications
In this section, we will discuss the different gas turbine classifications in terms
of design type, number of shafts, drive location, and cycle.
Classification by Design Type
Fig. 6.1.6 presents the industrial type of gas turbine. The two industrial divisions of gas turbines are shown. The older industrial type which was grass roots
industrial; that is, never built to function as an aircraft engine. The modern
291
B.P. 6.1
More Best Practices for Rotating Equipment
TABLE 6.1.1 Gas Turbines—History of Development
Year
Milestone
1900–30
Various works—expansion turbines (Delaval, Parsons, etc.)
1930
Sir Frank Whittle granted gas turbine patent
1943
First successful gas turbine (jet engine)
1950s
Industrial gas generators and power turbines used in pipeline service
(1st generation)
1960s
Improved efficiency through use of material and cooling
improvements. Use in power generation and industrial plants
1970s
j
j
j
j
1980s
j
j
j
j
1990s
j
j
Development of 2nd generation—larger sizes, higher efficiency
(higher firing temperatures)
First uses of aero derivative gas generators and power turbine for
‘off shore’ applications
Increased availability
Increased preventive maintenance cycle time
Extensive use of gas turbines in combined cycles for cogeneration
Aero derivative types gain further acceptance
Continued efficiency increase (higher firing temperatures)—retrofits
of 1st generation units
Development of 3rd generation gas turbines (use of advanced
materials, processing, coating and cooling techniques
Further acceptance of gas turbine as an industrial prime mover
Simple cycle efficiencies approach 45%. Firing temperatures
approach 1400°C (2500°F)
FIGURE 6.1.6 Gas turbine classifications industrial type. (A) Grass root’s industrial (never
built to fly). (B) Aero-influenced industrial (lighter weight hydrodynamic bearings). (Part A: Courtesy of General Electric Co. Part B: Courtesy of Solar Turbines, Inc)
292
Gas Turbines
Chapter | 6
TABLE 6.1.2 Industrial Type Gas Turbines
Advantages
Disadvantages
j
Longer cycle time between
maintenance
j
Longer maintenance times
j
Longer bearing life (hydrodynamic
bearings)
j
Large foot print
j
Greater tolerance to upsets
j
High specific weight
Lower efficiency (1st and 2nd generation)
Longer start sequences
j
j
approach to industrial type gas turbines (late 1960s) is the aero derivative influenced industrial gas turbine. This gas turbine design evolves from the aircraft
industry and is a lighter weight type of industrial turbine. Maintenance is easier
than the grass roots industrial since components are modulized and are changed
out as opposed individual parts in the grass roots industrial. Both industrial
types are differentiated from aero derivative types by the fact the radial bearings
are always hydrodynamic. Facts concerning industrial type turbines—advantages and disadvantages are presented in Table 6.1.2. In recent years, the high
emphasis on maintainability has favored the aero derivative type gas turbine as
opposed to either type of industrial gas turbine presented here. This is because
the maintenance times for the aero derivative gas turbines in the field are significantly reduced over industrial type gas turbines because the aero derivative
unit can be easily exchanged with a similar unit in the field. Therefore, the field
maintenance time is significantly lower for the aero derivative gas turbine (typically 72 h as opposed to 360 h + ).
A single shaft grass roots industrial gas turbine—General Electric model
7000 (Frame 7) is shown in Fig. 6.1.7. This turbine has been used for both
generator drives and mechanical drives. Nominal ISO horsepower is in the 100
(135,000 BHP) range. Efficiency is approximately 35%.
FIGURE 6.1.7 Single shaft industrial gas turbine. (Courtesy of General Electric)
293
B.P. 6.1
More Best Practices for Rotating Equipment
FIGURE 6.1.8 Mars gas turbine. (Courtesy of Solar Turbines)
Fig. 6.1.8 shows an example of a two shaft aero derivative gas turbine—
Solar Mars gas turbine used for mechanical drive applications (compressor and
pump drives). Nominal ISO horsepower is in the 11 MW (15,000 BHP) range.
Efficiency is approximately 35%.
Figs. 6.1.9–6.1.11 are examples of various aero derivative gas turbines. In
Fig. 6.1.9, a General Electric LM 2500 gas turbine is shown in two applications,
the first as a gas generator for a Dresser Rand DJ 270R power turbine. The
FIGURE 6.1.9 Gas turbine classifications aero derivative. (A) Used as gas generator only. (B)
Entire engine adapted (gas generator and power turbine). Note: power turbine is either turbo-prop
or bypasss fan drive in aero-engine version. (Part A: Courtesy of Dresser Rand. Part B: Courtsey
of General Electric)
294
Gas Turbines
Chapter | 6
FIGURE 6.1.10 The industrial RB211. (Courtesy of Rolls Royce)
other application uses the LM 2500s six-stage power turbine on a separate shaft.
Nominal ISO horsepower is in the 20 MW (25,000 BHP) range. Efficiency is
approximately 37%.
Fig. 6.1.10 is a drawing of a Rolls Royce RB211 two-shaft gas turbine. This
gas turbine has intermediate and high-pressure axial compressors mounted on
separate shafts for increased gas turbine efficiency. Nominal ISO horsepower is
in the 23 MW (30,000 BHP) range. Approximate efficiency is 38%.
FIGURE 6.1.11 LM6000 gas turbine. (Courtesy of General Electric)
295
B.P. 6.1
More Best Practices for Rotating Equipment
TABLE 6.1.3 Aero-Derivative Type Gas Turbines
Advantages
Disadvantages
j
Shorter maintenance times
j aShorter
j
Small foot print
j
Less tolerance to upsets
j
Low specific weight
j
Shorter bearing life (anti-friction bearings)
j
Higher efficiency
j
Faster start sequence
cycle time between maintenance
a
Note: maintenance cycle time is increasing and approaching industrial types
The newest aero-derivative gas turbine used for generator and mechanical
drive is the LM 6000 two-shaft gas turbine shown in Fig. 6.1.11. This gas turbine produces 45 MW (60,000 ISO horsepower), has an efficiency of 43% and
can drive a load on either or both ends.
The advantages and disadvantages of aero-derivative gas turbines are presented in Table 6.1.3.
Fig. 6.1.12 shows a hybrid type gas turbine, which is a combination of
an aero-derivative gas generator and an industrial power turbine. This design
offers the advantage of maintainability on the “hot section” of the gas turbine and high reliability in the power turbine. These facts are presented in
Table 6.1.4.
Aero-derivative and industrial facts are discussed in Table 6.1.5.
FIGURE 6.1.12
296
Gas turbine classifications hybrid type industrial. (Courtesy of Dresser Rand)
Gas Turbines
Chapter | 6
TABLE 6.1.4 Classification of Industrial and Aero-Derivative Gas Turbines
Usually based on power turbine type
Depends on types of bearings
j Anti-friction = aero
j Hydrodynamic = industrial
j With time, both types will converge to a “hybrid”
Present 3rd generation designs are moving in this direction
j
j
TABLE 6.1.5 Aero-Derivative Versus Industrial Facts
Item
Aero-derivative
Industrial
Casing weight
Light
Very heavy
Casing material yield
3 times higher yield strength
—
Rotor weight
15–20 times lighter
—
Bearing type
Anti-friction
Hydrodynamic
Bearing life
50,000 h
50,000–100,000 h
Start-idle times
1–2 min
15–30 min
Boroscope locations
More than industrial
The Number of Gas Turbine Shafts
Gas turbines are configured as single, dual, or triple shaft designs. The advantages and disadvantages of each type are presented in Fig. 6.1.13. Most modern gas
turbines are of the triple shaft design. Fig. 6.1.14 shows a single shaft gas turbine
where the gas generator and power turbine are mounted on the same shaft. This
figure also shows a dual shaft gas turbine, where the gas generator and power
turbine are mounted on different shafts. Single shaft gas turbines are usually limited to generator drive applications since the starting turbine load is significantly
less for a generator application, because generator is started under no load. Dual
shaft turbines are used for mechanical drive, pump, and compressor applications.
Gas Turbine Drive Configurations
Gas turbines can be designed as hot end drive, or cold end drive. Table 6.1.6
presents these facts. The majority of first and second generation gas turbines
were of a hot end drive. Most third generation gas turbines are of the cold and
dry type. A cold end drive configuration is a more reliable approach, in the writer’s opinion, since the coupling environment is significantly reduced in terms of
temperature. This results in a much lower axial expansion of the drive coupling
and subsequently increases the reliability of the gas turbine.
297
B.P. 6.1
More Best Practices for Rotating Equipment
FIGURE 6.1.13 The number of gas turbine shafts advantages/disadvantages.
FIGURE 6.1.14 The number of gas turbine shafts. (Courtesy of Solar Turbines, Inc)
Gas Turbine Cycles
Gas turbine cycles are presented in Fig. 6.1.15. There are essentially three
types of gas turbine cycles. The simple cycle, where the gas is exhausted directly to atmosphere. The regenerative cycle, where the exhaust gas is used
298
Gas Turbines
Chapter | 6
TABLE 6.1.6 Gas Turbine Drive Configurations
Hot end drive (exhaust end)
j
Majority of 1st, 2nd generation
Cold end drive (inlet end)
j
j
Disadvantages
Some 2nd generation
Most 3rd generation
Advantages
j
Longer drive coupling spacer
j
Shorter drive coupling Spacer
j
Driver coupling in hot environment
j
Minimized thermal expansion effects
FIGURE 6.1.15 Gas turbine cycles.
in an exchanger (regenerator) to preheat the compressor discharge air prior to
the combustor and the combined cycle where the exhaust gas is used in a heat
recovery steam generator (HRSG) to either generate steam for plant use or as
an expansion fluid is a steam turbine. Typical efficiencies are as follows:
j
j
j
Simple cycle 20–43%
Regenerative cycle 30–45%
Combined cycle 55–60%
Table 6.1.7 shows the different types of Gas Turbines to be used in different
applications.
299
B.P. 6.2
More Best Practices for Rotating Equipment
TABLE 6.1.7 Gas Turbine Applications
Industrial single shaft
j
Generator drive
j aLimited
mechanical
drive applications
(Compressor—pump)
Industrial dual shaft
j
Land based mechanical drive
j
Pipeline
Refinery, gas plant, petrochemical plant land based
Generator drive
j
j
Aero multi-shaft (dual or triple)
Co-generation
Off-shore
Generator drive
Compressor drive
Pipeline
Pump drive
Compressor drive
a
Usually used only when a multi-shaft alternative in the designed power range is not available
B.P. 6.2: Conduct design audits on aero-derivative gas turbines that have
zero, or limited (less than 2 years in operation), mechanical drive experience
Aero-derivative gas turbines are being used more and more these days for
mechanical drive applications in on-shore facilities, due to their higher reliability and ease of maintenance (change-out to an already factory tested replacement). However, in the last 10 years with plants becoming bigger and bigger
there has been a push to use these types of turbines for higher power requirements as mechanical drives.
This best practice is not to discourage the use of aeroderivative gas turbines
by any means, but rather assure that the vendors are utilizing a design for mechanical drive that will not introduce any new issues.
The design audit shall be conducted to evaluate the design criteria for all components mechanically, thermally and aerodynamically and confirm that they are
all within design limitations by the manufacturer. Many times it will be found that
some, if not all, of the components have been used in other turbines (maybe on a
smaller scale) that have satisfactory experience, therefore they can be accepted.
L.L. 6.2: Failure to accurately audit the design of new mechanical drive
gas turbines can result in unexpected issues and project/startup delays
BENCHMARKS
This best practice is nothing new to compressors and steam turbines in new applications, but was not used for gas turbines because the industrial versions had
significant mechanical drive experience for many years and they are production
300
Chapter | 6
Gas Turbines
TABLE 6.2.1 Suggested Design Audit Activity
1.
Risk type
2.
Design Checks
1
2
3
4
j
Aero-dynamic
X
X
?
*
j
Thermodynamic
X
X
?
*
j
Rotor response
X
X
?
*
j
Stability analysis (if applicable)
X
X
?
*
j
Seal balance
X
X
?
*
j
Thrust balance
X
X
?
*
j
Bearing loading
X
X
?
*
j
Train lateral analysis
X
X
?
*
j
Torsional analysis (if applicable)
X
X
?
*
j
Transient torsional (if req’d)
X
X
?
*
j
Control system simulations
X
X
X
X
j
System layout—accessibility
X
X
X
X
1, Prototype; 2, multiple component inexperience; 3, single component inexperience; 4, proven
experience for all components; X, required; ?, optional; *, not required.
units, not custom equipment like compressors and steam turbines. However,
with the recent success and increased reliability of aero-derivative gas turbines
this best practice has been followed in the last decade when aero-derivative gas
turbines were being used in more mechanical drive applications for higher and
higher power outputs. The results of this best practice for aero-derivative gas
turbines have been optimal gas turbine reliability and minimal project or startup delays.
SUPPORTING MATERIAL
See Table 6.2.1 for a typical design audit agenda for a compressor train.
B.P. 6.3: Require a compressor discharge temperature (CDT) transmitter
for all gas turbines in order to accurately trend air compressor efficiency
Most Gas Turbines will not be provided with a Compressor Discharge Temperature
Instrument, which is critical to determining and trending the efficiency of
the air compressor on the gas turbine. This is the only true way to accurately
know if a gas turbine power loss is due to the fouling of the air compressor
or other factors. Remember, just looking at power output from the gas turbine is not indicating the cause of power drop as power output will fluctuate
301
B.P. 6.3
More Best Practices for Rotating Equipment
normally due to ambient conditions (Temperature, Humidity, and Pressure),
pressure drop across inlet filtration, fuel conditions, etc. Of course fouling of
the compressor section is the most common cause of permanent gas turbine
power loss.
Therefore, trending compressor efficiency will allow for the user to define
the level of fouling in the compressor section and determine a proper schedule/
procedure for washing this section. The point is, we want to get after what the
cause of the efficiency drop is immediately and try to get the performance back
to where it should be as quickly and as easy as possible. It is a known fact that
the earlier a wash (whether online or crank) can be introduced when compressor efficiency decreases, the better chance you have to gain back maximum
performance.
Note: If your existing gas turbine is not equipped with CDT, it should be
easy to add since most vendors have a location for it where they installed a
thermowell during testing.
L.L. 6.3: Inability to trend compressor efficiency has resulted in permanent
gas turbine power loss and reduction of rates
Many users without the ability to trend turbine compressor efficiency will
not identify fouling quick enough and washing is much less effective.
BENCHMARKS
This best practice has been used since the mid 1990s to aid in full gas turbine
condition monitoring and has optimized gas turbine reliability and on-stream
time at maximum rates. A compressor discharge pressure (CDP) is also required to calculate compressor efficiency, but is typically provided on all gas
turbines.
SUPPORTING MATERIAL
Gas Turbine Performance
A gas turbine is a dynamic internal combustion engine. When we compare the
performance of a gas turbine to a steam turbine, it becomes immediately evident
that steam turbine performance is much easier to calculate since both the vapor
and the vapor conditions are fixed. When we examine the performance of a gas
turbine, we immediately see that the vapor condition is variable based on the
type of fuel used and the atmospheric conditions. This is true since the inlet
to the gas turbine engine is from atmosphere and any change of temperature,
humidity or pressure will affect the mass flow and consequently the power produced by the gas turbine. The gas turbine cycle (Brayton) is open.
As a result, steam turbine performance can be expressed rather easily in
terms of steam rate (pounds of steam per horsepower or kilowatt hour) and
external efficiency. Since the gas turbine vapor conditions are variable however,
302
Gas Turbines
Chapter | 6
gas turbine performance must be expressed in terms of heat rate, BTU’s per
horsepower or kilowatt hour, thermal efficiency and fuel rate. All of the above
also must be expressed in standardized terms.
A set of standardized conditions has been established by ISO (International
Standards Organization) to rate all gas turbines. We will discuss the various ISO
standard requirements and how the site rating is obtained by using vendor ISO
derating data for each turbine design. An actual gas turbine performance example
will be presented and the effect of varying inlet conditions (temperature, pressure and humidity) on performance will also be presented. Finally, the exhaust gas
composition will be discussed and the emission products examined. In addition,
various alternatives for meeting local emission requirements will be presented and
discussed.
Fig. 6.3.1 presents a comparison between gas turbine and steam turbine performance.
A gas turbine is an internal combustion engine in that the hot vapor is produced internal to the engine. The cycle is open since both inlet and exhaust
conditions are “open” to the atmosphere and vary with atmospheric conditions.
FIGURE 6.3.1 Gas turbine versus steam turbine performance. (Reprinted with permission
from Gas Producers Suppliers Association GPSA)
303
B.P. 6.3
More Best Practices for Rotating Equipment
TABLE 6.3.1 Gas Turbine Versus Steam Performance
Since the steam turbine (Rankine) cycle is closed, the vapor and vapor conditions
are constant!
Therefore performance can be expressed in terms of:
kg or lbs of steam
Steam Rate =
kW or bhp - h
j
j
j
actual work × loss factor
ideal work
j
External efficiency =
j
Mechanical and leakage losses
The steam turbine is an external combustion engine since the hot vapor is
produced external to the engine. The steam turbine cycle is closed in that both
inlet and exhaust conditions are controlled by the steam generation system
(boiler), therefore steam turbine conditions are constant and do not vary.
Table 6.3.1 presents performance parameters for steam turbines. Since inlet and exhaust conditions are controlled and the steam turbine is an external
combustion engine, steam rate and external efficiency can be used to express
performances.
Since the gas turbine Brayton cycle is open, vapor conditions are variable
and performance must be expressed as:
Heat rate
Thermal efficiency
Fuel rate
j
j
j
These facts are shown in Table 6.3.2.
TABLE 6.3.2 Gas Turbine Versus Steam Turbine Performance
j
j
j
Since the gas turbine (Brayton) cycle is open, both the vapor and vapor conditions
are variable.
Therefore performance is expressed in terms of:
kJ or btu
heat rate (ISO) =
kW or bhp-hr
j
thermal efficiency =
j
fuel rate =
kJ/kW-hr
heat rate(ISO)kJ/kW-hr
(heat rate) (kW)
fuel heating value (kJ/kg)*
or
or
btu/hp-h
heat rate(ISO)btu/hp-hr
(heat rate) × (horsepower)
fuel heating value (btu / lb)*
Note: * ISO conditions—standardized fuel, inlet conditions at design speed—no losses.
304
Gas Turbines
Chapter | 6
TABLE 6.3.3 Gas Turbine Performance ISO Conditions
j
Since gas turbine performance varies as a function of fuel and inlet conditions, a
set of standardized conditions has been established by ISO (International Standards
Organization) to rate all gas turbines.
ISO standard conditions are:
j
T inlet = 15°C (59°F)
j
Relative humidity = 0%
j
P inlet = sea level
j
Design speed of rotors
j
Inlet and exhaust losses = 0″H2O
j
Power losses = 0
j
Based on stated fuel heating value
j
Compressor bleed air = 0
j
Gas Turbine ISO Conditions
Since gas turbine performance varies as a function of fuel and inlet conditions,
a set of standard conditions has been established by the International Standards
Organization to define gas turbine performance. These facts are presented in
Table 6.3.3.
Gas turbine vendors publish performance data in terms of ISO power rating
and ISO heat rate. Typical vendor data is shown in Table 6.3.4.
Site Rating Correction Factors
Gas turbine site performance is directly affected by inlet air density and air
environmental conditions as shown in Tables 6.3.5 and 6.3.6 and Fig. 6.3.2 respectively.
Since produced power and heat rate vary as a function of inlet temperature,
pressure and inlet duct and exhaust duct pressure drop, vendors supply correction curves to convert ISO conditions to site conditions.
Table 6.3.7 Figs. 6.3.3–6.3.5 present an example of a typical gas turbine site
rating exercise.
The Effect of Firing Temperature on Power and Efficiency
A small increase in firing temperature has a significant effect on produced
horsepower on engine efficiency. These facts are shown in Fig. 6.3.6.
See a typical form with calculations built in for total Gas Turbine Condition
Monitoring. These are based on general gas turbine configurations and may
be slightly different from your specific turbine (but the form can be modified)
(Table 6.3.8).
305
B.P. 6.3
306
TABLE 6.3.4 Gas Turbine Performance ISO and Site Performance
Dresser-Rand
Turbo Products Division
Model
POWER
RATING ISO
Base Load
Gas Fuel (HP)
HEAT RATE
Lower Heating
Value (LHV)
(BTU/HP-hr)
POWER
SHAFT
SPEED
(RPM)
PRESSURE
RATIO
DR-22C
5,278
8,850
13,280
9.9
NO. OF
COMBUSTORS
Turbine
Inlet
Temp.
(C)
Exhaust
Flow
(kg/sec.)
Exhaust
Temp.
(C)
6
1,035
15.6
579
DR-990
5,900
8,350
7,200
12.2
1
1,082
20.0
482
DR-60G
18,750
6,840
7,000
21.5
1
1,216
45.6
482
DR-61G
31,200
6,777
3,600
18.8
1
1,235
69.0
523
DR-61
30,800
6,800
5,500
18.8
1
1,235
69.0
520
DR-63G
56,840
6,135
3,600
30.0
1
1,154
122.5
452
NOTES
ALL VENDORS PUBLISH ISO PERFORMANCE AND DE-RATING DATA SO THAT SITE PERFORMANCE (AT ACTUAL SITE CONDITIONS, FUEL AND LOSSES) CAN BE
DETERMINED.
TYPICAL VENDORS DATA.
Reprinted with permission of Turbomachinery International Handbook 1993 Vol. 34 No. 3.
More Best Practices for Rotating Equipment
AT ISO RATING CONTINUOUS
Gas Turbines
Chapter | 6
TABLE 6.3.5 The Effect of Inlet Air Density on Produced Power and Heat
Rate
j
j
j
j
A given engine design limits air volume flow capacity.
Produced power is a function of actual energy extracted per pound of vapor and
mass flow of vapor.
For a given engine therefore, produced power varies directly with inlet air density.
Produced power does become limited by low volume (stall and surge) flow.
TABLE 6.3.6 Gas Turbine Performance Effect of Inlet Conditions on
Performance
j
j
Care must be taken when selecting gas turbines to assure sufficient shaft power is
available at:
j High temperature conditions
j Fouled inlet conditions
Gas turbine applications tend to be “fully loaded” since gas turbines (unlike steam
turbines) are not custom designed.
FIGURE 6.3.2
Typical gas turbine output power and heat rate versus ambient temperature.
307
B.P. 6.3
More Best Practices for Rotating Equipment
TABLE 6.3.7 Typical Gas Turbine Site Rating Exercise
1. Scope
The purpose of this specification is to estimate the site shaft horsepower and heat rate
for a given set of site conditions.
2. Applicable documents
Figures 6.3.3–6.3.5
3. Requirements
3.1 The following site condition must be known:
A. Elevation (ft.)
B. Inlet temperature (°F)
C. Inlet duct pressure loss (inches of water)
D. Exhaust duct pressure loss (inches of water)
4. Procedure
4.1 Read the shaft horsepower (SHP) and heat rate (HR) for the site inlet temperature
(from Figure 6.3.3)
4.2 Read the elevation correction factor (δ) for the site elevation (from
Figure 6.3.4)
4.3 Site shaft horsepower:
A. Read the inlet correction factor (Ki) for the site inlet duct pressure loss (from
Figure 6.3.5).
B. Read the exhaust correction factor (Ke) for the site exhaust duct pressure loss
(from Figure 6.3.5).
C. Calculate the site shaft horsepower:
Site SHP = SHP (from Figure 6.3.3) × δ × Ki × Ke
4.4 Site heat rate:
A. No elevation correction factor (δ) is used for the heat rate.
B. Read the heat rate correction factor (Kh) from Figure 6.3.5 for the duct
pressure loss (sum of site inlet and exhaust duct pressure losses).
C. Calculate the site heat rate:
Site HR = HR (from Figure 6.3.3) × Kh
5. Sample calculation
5.2 Assume the following site conditions:
A. Elevation (ft.)
1000 ft
B. Inlet temperature (°F)
59°F
C. Inlet duct pressure loss (inches of water)
3.5 inches of water
D. Exhaust duct pressure loss (inches of water)
4.5 inches of water
308
Gas Turbines
Chapter | 6
TABLE 6.3.7 Typical Gas Turbine Site Rating Exercise (cont.)
5.2 Read the shaft horsepower (SHP) and heat rate (HR) with no inlet or exhaust duct
pressure losses for the site inlet temperature (from Figure 6.3.3).
59°F site inlet temperature
Shaft horsepower (SHP)
29,200
Heat rate (HR)
7,035 BTU/HP-HR
5.3 Read the elevation correction factor (δ) for the site elevation (from
Figure 6.3.4).
1000 ft. site elevation:
Elevation correction factor (δ)
0.964
5.4 Site shaft horsepower:
A. Read the inlet correction factor (Ki) for the site inlet duct pressure loss (from
Figure 6.3.5).
3.5 inches of water site inlet duct pressure loss:
Inlet correct factor (Ki)
0.9845
B. Read the exhaust correct factor (Ke) for the site exhaust duct pressure loss (from
Figure 6.3.5).
4.5 inches of water site exhaust duct pressure loss:
Exhaust correction factor (Ke)
0.991
C. Calculate the site shaft horsepower:
Site SHP = SHP (from Figure 6.3.3) × δ × Ki × Ke
= 20.200 × 0.964 × 0.9845 × 0.991
Site SHP = 27,463
5.1 Site heat rate:
A. Read the heat rate correction factor (Kh) from Figure 6.3.5 for the duct
pressure loss (sum of site inlet and exhaust duct pressure losses).
Duct pressure loss is the sum of 3.5 inches of water site inlet duct pressure loss
and 4.5 inches of water site exhaust duct pressure loss, equaling 8.0 inches of water
duct pressure loss:
Heat rate correction factor (Kh)
1.016
B. Calculate the site heat rate:
Site HR = HR (from Figure 6.3.2) × Kh
= 7,035 × 1.016
Site HR = 7,148 BTU/HP-HR
309
B.P. 6.3
More Best Practices for Rotating Equipment
FIGURE 6.3.3 Figure for typical gas turbine site rating exercise. (Courtesy of General Electric)
310
Gas Turbines
Chapter | 6
FIGURE 6.3.4 Figure for typical gas turbine site rating exercise. (Courtesy of General Electric)
311
B.P. 6.3
More Best Practices for Rotating Equipment
FIGURE 6.3.5 Figure for typical gas turbine site rating exercise. (Courtesy of General Electric)
312
Chapter | 6
Gas Turbines
FIGURE 6.3.6 The effect of increased firing temperature on produced Power and engine
efficiency.
TABLE 6.3.8 Gas Turbine Component Condition Monitoring
Equip. #
Date
Compressor Performance
P atmosphere (psia)
T ambient (°F)
P Disch. (psia)
T Disch. (°F)
Speed (RPM)
K
(K−1)/K
—
—
—
—
(n−1)/n
—
—
—
—
Polytropic Efficiency
—
—
—
—
What Action is Required based on %
difference from design efficiency?
Gas Turbine Performance
Fuel LHV (BTU/SCFH)
Flow (lb/h)
FG Pressure (psia)
(Continued)
313
B.P. 6.3
More Best Practices for Rotating Equipment
TABLE 6.3.8 Gas Turbine Component Condition Monitoring (cont.)
FG Temperature (R)
FG Density (lb/ft.3)
—
—
—
—
SCFH FG Flow
—
—
—
—
Site Heat Rate (BTU/h)
—
—
—
—
GT Efficiency
—
—
—
—
Output HP
What Action is Required based on Gas
Turbine Performance info. above?
% PT Speed
Bearing Condition
# 1 Journal Brg. Vib X
# 1 Journal Brg. Vib Y
Major Frequency Observed
Radial Shaft Position
# 1 Brg. Pad Temp. 1 (°F)
# 1 Brg. Pad Temp. 2 (°F)
# 1 Brg. Drain Temp (°F)
GG Thrust Brg. Displacement
Direction of Thrust
Pad Temp. Inlet End 1 (°F)
Pad Temp. Inlet End 2 (°F)
Pad Temp. Exhaust End 1 (°F)
Pad Temp. Exhaust End 2 (°F)
Thrust Brg. Drain Temp. (°F)
# 2 Journal Brg. Vib X
# 2 Journal Brg. Vib Y
Major Frequency Observed
Radial Shaft Position
# 2 Brg. Pad Temp. 1 (°F)
# 2 Brg. Pad Temp. 2 (°F)
# 2 Brg. Drain Temp (°F)
# 3 Journal Brg. Vib X
# 3 Journal Brg. Vib Y
Major Frequency Observed
Radial Shaft Position
# 3 Brg. Pad Temp. 1 (°F)
314
Gas Turbines
Chapter | 6
TABLE 6.3.8 Gas Turbine Component Condition Monitoring (cont.)
# 3 Brg. Pad Temp. 2 (°F)
# 3 Brg. Drain Temp (°F)
PT Thrust Brg. Displacement
Direction of Thrust
Pad Temp. Inlet End 1 (°F)
Pad Temp. Inlet End 2 (°F)
Pad Temp. Exhaust End 1 (°F)
Pad Temp. Exhaust End 2 (°F)
Thrust Brg. Drain Temp. (°F)
# 4 Journal Brg. Vib X
# 4 Journal Brg. Vib Y
Major Frequency Observed
Radial Shaft Position
# 4 Brg. Pad Temp. 1 (°F)
# 4 Brg. Pad Temp. 2 (°F)
# 4 Brg. Drain Temp (°F)
Brg. Inlet Oil Pressure (psia)
Brg. Inlet Oil Temperature (°F)
Viscosity (cst)
% Water in oil
Lube Oil Flashpoint (°F)
What Action is Required based on Gas
Turbine Bearing info. above?
Air Filtration Sys. Condition
Filter DP (psid)
Action of Self Cleaning Air System (Active
or Inactive)?
Action?
Variable Inlet Guide Vanes
Hyd. Supply Differential Press. (psid)
Guide Vane Position
Air Flow Set Point
Measured Air Flow
IGV Exhaust Temp. Reference
Measured Exhaust Temp. (°F)
Action?
(Continued)
315
B.P. 6.3
More Best Practices for Rotating Equipment
TABLE 6.3.8 Gas Turbine Component Condition Monitoring (cont.)
Cooling and Sealing Air Sys.
Cooling Air Pressure (psia)
Other Cooling Sys. Observations
Air Sealing System Supply Pressure (psia)
Fuel Gas System
Ambient Temp. (°F)
—
—
—
—
—
—
—
—
Relative Humidity
Fuel Supply Pressure (psia)
Fuel Flow Rate (lb/h)
Fuel Pressure Upstream of Shut Off Valve
(psia)
Fuel Shut Off valve position
Fuel Pressure Upstream of Fuel Control
Valve (psia)
Fuel Control Valve position
Fuel Supply Pressure to combustors (psia)
Action?
Combustion Monitoring System
Ambient Temp. (°F)
Relative Humidity
Exhaust Temp. 1 (°F)
Exhaust Temp. 2 (°F)
Exhaust Temp. 3 (°F)
Exhaust Temp. 4 (°F)
Exhaust Temp. 5 (°F)
Exhaust Temp. 6 (°F)
Exhaust Temp. 7 (°F)
Exhaust Temp. 8 (°F)
Exhaust Temp. 9 (°F)
Exhaust Temp. 10 (°F)
Exhaust Temp. 11 (°F)
Exhaust Temp. 12 (°F)
Allowable Spread (°F)
Top Spread 1 (°F)
Top Spread 2 (°F)
Top Spread 3 (°F)
316
Gas Turbines
Chapter | 6
TABLE 6.3.8 Gas Turbine Component Condition Monitoring (cont.)
Exhaust Thermocouple Alarm on?
Combustion Alarm on?
High Exhaust Temp. Spread trip?
Monitor Enable Activated?
Action?
Exhaust Temperature Control System
Compressor Disch. P (psia)
Fuel Stroke Reference
Ambient Temperature (°F)
Relative Humidity
Calculated Firing Temp. (°F)
Action?
B.P. 6.4: Establish a washing procedure consisting of both on line and
crank washing techniques.
As mentioned in B.P. 6.3, compressor fouling is the most common form of
efficiency loss in any gas turbine. Therefore, we first want to trend compressor
efficiency to indicate the moment that fouling is beginning to occur. Now, once
we have indication, what do we do?
It has been found that the earlier you act by washing the turbine, the better
you will be able to get back the efficiency lost. Also, the frequency and type
of washing is important. The best documented results for efficiency gain have
been from some combination of on line washing and crank washing. The best
frequency shall be based on trending of compressor efficiency gain from washing and first start with more frequent washes and continue to reduce frequency
until you see a difference in results (for example the compressor efficiency gain
back was less when you decreased crank wash from monthly to two months, so
go back to monthly or even 6 weeks and monitor).
L.L. 6.4: Improper washing procedures have resulted in ineffective washes
with minimal performance gain
BENCHMARKS
This is a new best practice, added in 2011 as a result of many users who have
stated that washing of their turbines was almost a waste of time since minimal
performance was gained back during their washing. This will optimize the reliability and maximize performance of the turbine over time.
317
B.P. 6.5
More Best Practices for Rotating Equipment
SUPPORTING MATERIAL
See B.P. 6.3 supporting material for details regarding what can cause gas turbine
efficiency loss.
See the paper provided from Texas A&M Turbo-Lab and presented during
METS II (Middle East Turbomachinery Symposium) in 2013 in Doha, Qatar.
This article outlines industry best practices for gas turbine washing (http://mets.
tamu.edu/images/files/prog/proc/mets13/METS2Tutorial5.pdf).
B.P. 6.5: Use external (API-614) Lube Systems for critical mechanical
drive applications over 40 MW
In general, the cause of many failures in critical machinery lies within the
Oil System due to the abundance of components that need to be monitored accurately. For Gas Turbines, this is even more true since the turbine is sold typically
as a module and the oil system is located within the gas turbine enclosure.
By having the Lube System external and in accordance with API 614 will
allow for proper monitoring and troubleshooting of issues within the system.
L.L. 6.5: Typical gas turbine lube systems (inside enclosure) do not allow
for effective condition monitoring of the system and have resulted in poor
system reliability.
BENCHMARKS
This best practice has been in use since the mid 1990s and has allowed for optimal lube system and train reliability.
SUPPORTING MATERIAL
The Development of an Auxiliary System
Thus far, we have defined the functions of major types of auxiliary systems to
be covered in this course and have seen that all these systems are similar in their
functional design objective which is, “to continuously supply clean fluid to each
specified point at the required pressure, temperature and flow rate.” In this and
the following three sections we will detail the development of an actual auxiliary
system in order to fully understand the function of each major component and
how it contributes to the total operation and reliability of the system. Only after
the function of an auxiliary system is thoroughly understood, can we proceed to
discuss specifications, testing, operation and preventive maintenance.
The application selected will be to develop a pressurized lubrication and
steam turbine control oil system for the critical equipment unit. This example was
selected since many class members will be familiar with this type and because it
provides a good foundation towards understanding fluid sealing systems.
In the exercise that follows, we will examine the total system function in
detail, define each major component function, present component sizing criteria
318
Gas Turbines
Chapter | 6
and discuss common pitfalls in the selection, operation, and preventive maintenance of these components.
System Requirements
In order to determine the system requirements, the following information is
needed.
A. System design
B. Critical equipment vendor data
C. Site conditions System Design
The system schematic can be defined in the end users specifications and data
sheets or can be the vendor’s design. Regardless of the source, this fact should
be finalized prior to a purchase order. In our case, the system design is defined
by Fig. 6.5.1, which is the result of the end user’s specification modified by the
mutually agreed vendor exceptions.
Critical Equipment Vendor Data
This data must be furnished by each critical equipment vendor and will contain
information as shown in Table 6.5.1. It is important to note that frequently, different vendors furnish different pieces of critical equipment in the same unit. In
this case, all vendors should agree to a common Tube oil type and common value of oil supply conditions if possible. Failure to do so only complicates system
design and requires additional components, which can reduce system reliability.
FIGURE 6.5.1 Lube oil system schematic.
319
B.P. 6.5
More Best Practices for Rotating Equipment
TABLE 6.5.1 Critical Equipment Vendor Data
•
•
•
•
Oil flow rate for each bearing or component
Bearing or component friction loss (Heat load—BTU’s/HR)
Required lube oil type
Required oil supply pressure and temperature ranges (minimum and maximum) to
each bearing or component
• Equipment coast down time
• Any special requirements (equipment cool off time, etc.)
Site Conditions
This information is required for the proper design of the system and should be
accurately stated. As a minimum, the data noted in Table 6.5.2 should be included. Frequently, this information is not completed until well into the project
(if at all) and only leads to cost adders, delivery delays and unreliable systems.
End user input in the pre-purchase order phase of the project will eliminate these
problems. In addition, determination of auxiliary system arrangements and module location at this time will usually result in simpler, more practical designs that
can increase system reliability. A typical auxiliary system vendor data sheet is
included at the end of this section. It has been completed to include both equipment vendor site data and end user required data for the present example.
Having obtained all the necessary data to establish the system requirements,
we can now define the function of this specific system by modifying the general
definition of a lubrication system. The resulting definition is:
To continuously supply clean lubricating oil to each bearing and control oil
to the steam turbine governor and control valve system at the required pressure, temperature, and flow rate.
We now proceed to determine the system parameters necessary to design
this specific system. Once these values are established, the system components
can be sized and selected.
TABLE 6.5.2 Site Condition Data
• Site environmental conditions
• All utility data
• Location of system modules (consoles) relative to critical equipment—distance and
elevation
• Area electrical classification
• Information or sketch detailing system arrangement (location of oil supply and drain
connections, component location on modules, required space for maintenance and
minimum size of modules)
320
Gas Turbines
Chapter | 6
Determining the System Parameters
Fig. 6.5.2 is a table summarizing input data for the system design in Fig. 6.5.1.
Power loss, heat load, oil flow and oil temperature rise are related as follows:
 BTU  
BTU  
Heat Load 
= 2545 
 × [Power Loss (H.p)]
 HR  
 H.p - HR  
Also,
BTU  
LBMass  
Heat Load 
= Mass Rate 
 HR  
 HR  
 Specific Heat  BTU  
×

  × [ ∆T(°F)]
 of the Fluid  LB- °F  
FIGURE 6.5.2 Lube system design parameters.
321
B.P. 6.5
More Best Practices for Rotating Equipment
FIGURE 6.5.3 Typical bearing curve.
Specific heat (Cp) for critical equipment lubricating oil is approximately
0.4 BTU/LB-°F that is, 0.4 British thermal units of heat are required to raise one
lb. of oil 1°F in temperature.
The above relationships are used to determine the amount of fluid required
to maintain a specified fluid temperature rise in any component (bearing, gearing, seal, etc.) that experiences frictional power loss. Component designers
use this relationship to establish component oil requirement curves. A component (a bearing in this case) is installed in a test rig and operated at a certain
speed (rpm) and bearing load (pounds per in2). The fluid flow is varied to
achieve the desired fluid temperature rise. The power loss is measured using
a torque meter or other means. A typical performance curve for a bearing is
shown in Fig. 6.5.3.
In the present example, the data contained in Fig. 6.5.2 was obtained by using curves similar to Fig. 6.5.3 for each component with the following information available:
Equipment speed
Bearing size
Acceptable fluid temperature rise
Bearing load
322
Gas Turbines
Chapter | 6
Bearing load (P.S.I) is the result of a force acting on a given bearing area.
Hence,
Bearing Load = (PSI) =
Force (LBForce)
Area (in.2 )
Since bearings are not custom designed for each application, there are ranges of acceptable bearing loads. For hydrodynamic (oil film) bearings, the maximum acceptable load P, is approximately 500 PSI. This relationship holds for
both radial (journal) and axial (thrust) bearings. Although this example treats
hydrodynamic bearings, the relationships noted also apply to anti-friction (ball,
roller bearings, etc.). Some critical equipment drivers (aero-derivative gas turbines) use high precision anti-friction bearings with pressurized Tube oil systems operating at speeds in excess of 10,000 rpm.
The last item in Fig. 6.5.2 to mention is oil supply pressure. Having determined
the required quantity of oil to each component, the supply pressure required to
force the flow through the component must be determined. Refer to Fig. 6.5.4.
A very useful concept is presented here. The basic orifice relationship is:
 ∆P 
Fluid Flow = (Constant) × ( Flow Coefficient × ( D 2 ) × 
 S.G. 
Where, D, orifice diameter; ∆P, pressure drop; S.G., specific gravity of the
fluid.
FIGURE 6.5.4 Equivalent orifice.
323
B.P. 6.5
More Best Practices for Rotating Equipment
All components can be reduced to equivalent orifices. A 1 in. pipe—10 feet
long, a bearing with a given clearance, the passage through a bearing housing,
etc. Using this concept can assist greatly in understanding the operation of components and systems.
For the present example, a supply pressure of 15 PSI is required for the flow
through each bearing.
Fig. 6.5.5A shows a schematic of a journal bearing in a bearing housing
with an inlet and drain connection. The equivalent orifice system is shown in
Fig. 6.5.5B. Addressing any component in this manner shows the many factors
that can change. Any equivalent orifice area can increase with wear. Some can
decrease area by differential expansion (as in the case of high temperatures
caused by excessive friction). Note that we reduced Fig. 6.5.5A to an equivalent
orifice system, that is, the change of one orifice in the system (wear, etc.) effects
all of the other orifices in the system.
Having established the individual end component (bearing, control system)
requirements, we can now determine the following system requirements:
System Heat Load
By adding the individual component heat loads shown in Fig. 6.5.2, the system
heat load is obtained as follows:
Qsystem = Qcritical equip + Qsystem components
where Q, is heat load in BTU’s/HR.
The system component heat load is a small percentage of the total system heat
load and is primarily the heat (BTU/HR) contribution of the system pump(s).
FIGURE 6.5.5 Equivalent orifices within a bearing.
324
Gas Turbines
Chapter | 6
The system heat load therefore becomes the duty for the system cooler,
which in our case required that 275,000 BTU/HR of frictional heat be removed
from the system to maintain the specific oil supply temperature of 120°F.
System Flow Rate
The system flow rate is equal to the total of the individual component flow
rates and a contingency flow to account for component (bearing, seal, etc.) flow
changes (wear, etc.)
GPM system = GPM components + GPM contingency
The system flow rate is determined by the pump capacity. Fig. 6.5.6 shows
typical flow versus system pressure curves for positive displacement and dynamic classifications of pumps. It can be seen that the capacity of a positive displacement pump remains essentially constant while the capacity of a dynamic
pump increases with decreasing system pressure. In the case of a properly sized
dynamic pump, the additional critical equipment component oil flow required
due to wear, etc. will be automatically available.
On the other hand, a positive displacement pump must be oversized to have the
contingency capacity available. As we will see, a modulating control valve will be
combined with a positive displacement pump to render it a variable flow system.
Most auxiliary systems using oil usually utilize positive displacement pumps
since they are relatively insensitive to oil viscosity. Therefore, positive displacement pumps will be used for the present example.
FIGURE 6.5.6 Oil system pump performance.
325
B.P. 6.5
More Best Practices for Rotating Equipment
Thus, the system flow rate for our example is calculated as follows:
n =∞
n =∞
System Flow Rate = ∑ n =1 GPM N + (0.2) × ∑ n =1 GPM N
where, n, each component.
72 GPM = 60 + (0.2)(60)
A flow contingency of 10–20% is usually used depending on the system size (At
least 20% for API 614 Oil System). The system flow rate determined earlier will
then be the rated point for the system positive displacement pumps used in this example. If dynamic pumps were used, the initial operating point would be the normal
point in Fig. 6.5.6 and the maximum required flow case would be the rated point.
Minimum System Operating Capacity
Having obtained the system flow rate, the minimum system operating capacity
can be determined. The minimum system operating capacity is calculated by:
Minimum System Capacity = Flow to the Critical Equipment × System Retention Time
System retention time is equal to the time (in minutes) the pumps will operate if no oil is returned to the reservoir (as in the case of a total pipe break) and
if the beak occurs when the reservoir is at minimum operating level.
Retention times vary from one to in excess of ten minutes and should be
selected considering the following facts:
A. The time to bring the critical equipment to a stop (coast down time).
B. The time to adequately degas returned liquids (assuming external degassing
facilities are not available).
For the present example, a retention time of 5 min is selected. Therefore the
minimum system operating capacity is:
Minimum System Capacity = 60 gallons/min × 5 min = 300 gallons
This information will be used to size the oil reservoir.
System Resistance
At this point, we have almost all the system information required for component sizing and selection. The only exception is system supply pressure at the
pump discharge. In order to obtain this value, we must calculate the total system
resistance for the maximum and minimum pressures at the pump discharge.
Refer again to the system schematic in Fig. 6.5.1. The total system resistance
or pressure drop is calculated by adding up the pressure drops, for a given flow
rate, from the critical equipment component supply points back to the pump discharge Table 6.5.3 presents the calculated system resistance for the rated flow,
“clean system” (minimum component pressure drops) case of our example.
326
Gas Turbines
Chapter | 6
TABLE 6.5.3 Lube System Pressure Drops
ITEM (location of
Component)
Pressure
drop
Comments
• Bearing Pressure Drop
15 PSI
Supply pressure—atmospheric pressure
• Unit Supply Header ∆P
5 PSI
Pressure drop in supply header
• Elevation of unit
relative to module
10 PSI
Pressure drop for 25 ft. elevation
• Module to unit interconnecting piping ∆P
5 PSI
Interconnecting pipe pressure drop
• Lube oil pressure
control valve (PCV)
pressure drop
100 PSI
PCV set for 100 PSI ∆P to supply control
oil pressure of 135 PSI at module and
assure oil pressure exceeds water pressure
in oil coolers
• Transfer valve and
module piping ∆P
5 PSI
• Clean filter ∆P
10 PSI
• Clean cooler ∆P
10 PSI
TOTAL PRESSURE DROP
160 PSI
Pump discharge pressure (clean system)
• Control System ∆P
120 PSI
Supply pressure—atmospheric pressure
• Supply Header ∆P
1 PSI
Pressure drop in control oil supply header
• Elevation of unit
relative to module
10 PSI
Pressure drop for 25 ft. elevation
• Interconnecting pipe ∆P
4 PSI
Interconnecting pipe pressure drop
TOTAL PRESSURE DROP
135 PSI
Pressure that is required at control oil take
off point on module
I. Lubrication System
II. Control Oil System
The following conditions are given:
A. Oil type—light turbine oil
150 SSU at 100°F
60 SSU at 150°F (operating)
1000 SSU at 40°F
B. Rated oil flow to system = 60 gallons per min
C. Pipe size selected for an oil velocity of 5 ft./s
D. System component arrangement and location of module (console) relative to
equipment known
Therefore, the minimum pump discharge pressure required will be 160 PSI.
The maximum pump discharge pressure will be determined by a similar calculation at maximum component pressure drop conditions (dirty system) and the
327
B.P. 6.5
More Best Practices for Rotating Equipment
TABLE 6.5.4 Lube/Control Oil System Requirement Summary
Item
System Requirement
• System Heat Load
275,000 BTU/HR
• System Flow Rate
60 gallons per min (to unit)
75 gallons per min (pump capacity)
• Minimum System Operating Capacity
300 gallons
• System Resistance
160 PSI (clean)
200 PSI (dirty)
minimum allowable oil temperature (maximum oil viscosity) of the system. For
the present example, this value is 200 PSI.
This exercise concludes the determination of the system requirements for
this example. A summary of requirements are noted in Table 6.5.4.
See the following Figs. 6.5.7–6.5.9 of an API 614 Lube Oil System specified
to optimize reliability and ability to monitor and maintain easily. Believe it or
not, this console is complete and ready to ship!!!
See the following typical spreadsheet used to identify and monitor the major
components of a lube oil system.
FIGURE 6.5.7 Best of Best Oil System View 1. (Courtesy of D. A. Campbell)
328
Gas Turbines
Chapter | 6
FIGURE 6.5.8 Best of Best Oil System View 2. (Courtesy of D. A. Campbell)
FIGURE 6.5.9 Best of Best Oil System View 3. (Courtesy of D. A. Campbell)
329
B.P. 6.5
Component/Item
Oil Reservoir
Level
Oil Temp. (°C)
Air in Oil? (Y/N)
Gas in Oil?
Oil Sample?
Pumps
Aux. Pump
Operating?
P2 (bar)
Suction Noise?
Suction Filter ∆P
(bard)
Vibration (µm)
Brg. Bracket Temp.
(°C)
Couplings
Noise?
Strobe Findings
Turbine Driver
Operating Speed
(RPM)
Trip Speed Setpoint
(RPM)
Vibration (µm)
Brg. Bracket Temp.
(°C)
Gov. Hunting?
Trip Lever Condition
Gov. Oil Condition
Motor Driver
Operating?
Vibration (µm)
Brg. Bracket Temp.
(°C)
330
More Best Practices for Rotating Equipment
Specified Value
Actual Value
Comments
Gas Turbines
Component/Item
Specified Value
Actual Value
Chapter | 6
Comments
Axial Shaft
Movement (µm)
Fan Noise?
Relief Valves
Passing?
Set Pressure (bar)
Pump P2 Press. (bar)
Check Valves
Aux. Pump Turning
Backwards?
Noise?
Back Pressure Valve
% Open
Stable?
Valve Noise?
Set Pressure (kPa)
Maintained Pressure
(kPa)
Transfer Valves
One Bank Operating?
Noise?
Coolers
∆T Oil
CW Valve Pos.
Cooler Operating?
Vent Valves Open?
TCV’s
% Open
Set Temp. (°C)
Stable?
Actual Temp. (°C)
Filters
∆P (bar)
Vent Valves Open?
331
B.P. 6.5
Component/Item
More Best Practices for Rotating Equipment
Specified Value
Last Filter Change
Accumulators
Pre-charged
Pressure (bar)
Last PM Date
Lube Oil PCV
% Open
Set Pressure (bar)
Actual Pressure (bar)
Stable?
Control Oil PCV
% Open
Set Pressure (bar)
Actual Pressure (bar)
Stable?
Lube Oil Rundown Tank (or Emerg. Pump)
Pump or Tank?
Pump Operating?
Tank Overflow
Lube Oil Supply Lines
Leaks?
Noise?
Vibration (µm)
332
Actual Value
Comments
Chapter 7
Auxiliary Systems
B.P. 7.1: Oil viscosity selection guidelines
It is very important in the early design stages of an oil system to select the
proper oil as it will affect the design and selection of the major components
(Reservoir heater, Pumps, coolers, filters, etc.) within the system. Following
are guidelines to aid in selection of the proper oil viscosity in order to optimize
train reliability:
l
l
l
If system will be in a climate that has average high temperatures greater than
30°C (86°F) for at least 6 months throughout the year, use ISO VG 46 oil.
This is acceptable even with Centrifugal type oil pumps.
If system will be in a climate where average high temperatures are not greater than 30°C (86°F) for 6 months, ISO VG 32 is acceptable.
If the train consists of a gearbox, ISO VG 46 shall be specified at minimum.
The user knows their climate better than the designer, therefore both the user
and designer should work together in the selection of viscosity grade for the oil
system early on. Then the designer will easily be able to select the proper-sized
components based on the appropriate oil.
L.L. 7.1: Inadequate oil viscosity in high temperature climates has resulted
significant rotary pump wear and low system reliability
BENCHMARKS
This best practice has been in use since the 1980s and has resulted in minimal unplanned shutdowns due to oil system issues and maximized process unit
revenue.
SUPPORTING MATERIAL
Critical Equipment Vendor Data
This data must be furnished by each critical equipment vendor and will contain information as shown in Table 7.1.1. It is important to note that frequently,
different vendors furnish different pieces of critical equipment in the same unit.
More Best Practices for Rotating Equipment. http://dx.doi.org/10.1016/B978-0-12-809277-4.00007-3
Copyright © 2017 Elsevier Inc. All rights reserved.
333
B.P. 7.2
More Best Practices for Rotating Equipment
TABLE 7.1.1 Critical Equipment Vendor Data
•
•
•
•
Oil flow rate for each bearing or component
Bearing or component friction loss [Heat load; kJ/h (BTU/h)]
Required lube oil type
Required oil supply pressure and temperature ranges (minimum and maximum) to
each bearing or component
• Equipment coast down time
• Any special requirements (equipment cool off time, etc.)
TABLE 7.1.2 Site Condition Data
• Site environmental conditions
• All utility data
• Location of system modules (consoles) relative to critical equipment—distance and
elevation
• Area electrical classification
• Information or sketch detailing system arrangement (location of oil supply and drain
connections, component location on modules, required space for maintenance, and
minimum size of modules)
In this case, all vendors should agree to a common lube oil type and common value of oil supply conditions if possible. Failure to do so only complicates system
design and requires additional components which can reduce system reliability.
Site Conditions
This information is required for the proper design of the system and should be
accurately stated. As a minimum, the data noted in Table 7.1.2 should be included. Frequently, this information is not completed until well into the project
(if at all) and only leads to cost adders, delivery delays, and unreliable systems.
End user input in the pre-purchase order phase of the project will eliminate
these problems. In addition, determination of auxiliary system arrangements
and module location at this time will usually result in simpler, more practical designs that can increase system reliability. A typical auxiliary system vendor data
sheet is included at the end of this chapter. It has been completed to include both
equipment vendor site data and end user required data for the present example.
B.P. 7.2: Assure the vendor is provided with details of supply and drain
interconnecting piping (if they are not the supplier)
It is highly recommended that the Oil System Vendor provide the interconnecting supply and drain piping as it is critical to know the pressure drop
334
Auxiliary Systems
Chapter | 7
across this piping in order to design and select the components within the
system.
If the vendor does not supply this piping, then it must be coordinated early
on in the project between the Oil System Vendor and interconnecting piping
designer/provider to assure components are sized/selected properly.
The Oil system vendor needs to know the orientation and elevation of the
equipment supply and drain porting, the size of piping and detailed layout of
piping to calculate the pressure drop of the system properly.
This should be finalized by the Vendor Coordination Meeting (VCM) in
order to make sure there are no surprises and the long lead time items are purchased based on the proper system pressure drop.
L.L. 7.2: Failure to coordinate details of interconnecting piping with
the oil system vendor has resulted in unplanned shutdowns and revenue
lost
It has been seen many times in the field where the pressure drop on the supply piping is more than expected by the vendor and the Back Pressure Control
Valve operates closer to fully closed position than expected. This means that the
auxiliary pump can be triggered to start sooner than expected by the designers.
Many units have shutdown unexpectedly in instances when the auxiliary was
unable to start in time and maintain the pressure above the low oil pressure trip
setting.
BENCHMARKS
This best practice has been used for projects and retrofits since 1990 to produce
oil unit trains of the highest reliability.
This best practice has optimized centrifugal compressor train reliability (above 99.7%) and machinery component MTBFs. (Greater than
100 months.)
SUPPORTING MATERIAL
Reservoir, Vessel, Piping, and Component Material Preferences
Any required reservoir, overhead tank design or material preference should be
stated. It is recommended that reservoirs and overhead tanks be constructed of
Austenitic stainless steel to assure minimal entrance of excessive debris into the
auxiliary system. In addition, any preference for piping and synthetic materials should be stated. Recent practice has been to require stainless steel piping
as well as reservoir and overhead tank material, while carbon steel slip-on flanges
have been acceptable for lube oil service piping. Experience has shown that
in systems containing water, such as a water seal systems, the stainless steel
flanges as well as pipes are required since a considerable amount of rust scale
335
B.P. 7.2
More Best Practices for Rotating Equipment
emanates from the flange pipe interface area below the flange gaskets. The subject of the material of the main components; filters, coolers, valves, etc. is a
purchasers preference. When one considers the potential damage resulting from
excessive debris in a system, the additional cost for non-corrosive components
can be justified in many cases. This issue should be thoroughly investigated
prior to auxiliary system purchase.
Supply Pipe Velocity Checks
The pump header, interconnecting console pipe, and piping to the unit should all
be checked for proper fluid velocity or pressure drop. Typical velocity values in
auxiliary system supply pipes are on the order of 4–6 ft./s velocity. Velocity is
derived from the following equation for incompressible flow.
Q ft./min = A × V
where A = Internal pipe area (m2 or ft.2), V = fluid velocity (m/s or ft./s).
Charts for standard pipe sizes and schedules are available to determine velocities (Fig. 7.2.1).
Note that schedule 80 is usually used for carbon steel pipe below 2 in.
Schedule 40 is used above 2 in. For stainless steel pipe, schedules 10 and 20 are
used respectively.
FIGURE 7.2.1 Typical pipe sizing chart.
336
Auxiliary Systems
Chapter | 7
Typical drain line velocities are 0.15–0.08 m/s (1/2–1/4 ft./s). Attention
is drawn to properly sizing drain pipes for installations where critical equipment is significantly elevated above reservoir. All drain pipes should be sized
with adequate area to preclude excessive air being entrained with the oil
to promote drainage back to the reservoir. An additional consideration for
supply headers at the unit is that supply headers are frequently sized for
one standard pipe dimension. In the case of large critical equipment units
(two or three bodies and driver), the amount of oil from the entrance to the
header to the last component decreases significantly. In an effort to minimize
pipe size, many vendors size headers small. Therefore, pressure drop in
the header is excessive and requires a higher supply header pressure than
anticipated in the unit design. Improper sizing of critical equipment supply headers could cause excessive flow across equivalent orifices (bearings)
thereby requiring all flow of the main pump and necessitating the operation
of the auxiliary pump.
B.P. 7.3: Oil system console layout best practices
The cause of major reliability issues in critical machinery has historically
been traced, more times than not, back to lube, seal, or control oil systems. Not
only do these systems rely on the proper design and operation of a large number
of components, they are typically not designed in a way that is conducive to accurately monitoring the components. Following are key guidelines to assure the
system can be maintained and monitored as easy as possible:
l
l
l
l
Locate the reservoir in the center of the baseplate and build around it by locating the pumps on one side, the coolers along one of the long sides of the
baseplate and the filters on the side opposite the pumps.
Utilize double 3-way Ball Type Transfer valves which allow for much less
piping than the 6-way types and free up a lot of real estate on the console.
Mount the instruments and gauges on the reservoir, as close as possible to
the location they are measuring. This will assure minimum runs of sensing
lines for the instruments and optimal accuracy.
Assure control valves and relief valves are easy to access and check the
condition of valve position indicators where applicable. They should be at
or below eye level and at least a meter (approx. 3 ft.) of open space around
them for easy monitoring.
L.L. 7.3: Consoles that are crowded and do not allow easy access are often ignored by operators and not fully understood in terms of system
function
The writer has experienced critical unit shutdowns caused by the steam turbine–driven pump when the local trip lever was accidentally hit due to limited
space when the personnel were climbing on the console for normal maintenance
activities.
337
B.P. 7.3
More Best Practices for Rotating Equipment
BENCHMARKS
This best practice has been in use since the 1980s and has been improved upon
since, to its current details as mentioned in this book. This approach aids in assuring +99.7% reliability on all critical equipment trains.
SUPPORTING MATERIAL
Console Layout and Component Arrangement
Having confirmed the acceptable component sizing and selection, the console
layout and arrangement of components must be reviewed. Methods of this review incorporate either the review of outline drawings of the proposed arrangement, a model review, or a cad 3D drawing review. Many vendors and users
have found that models or cad 3D drawings aid greatly in understanding and
reviewing maintenance accessibility and layout considerations.
Console Construction
Auxiliary equipment consoles or modules house most of the components present in the auxiliary systems. Their construction should be reviewed to assure
proper stiffness and facilities for installation on site. Many horizontal consoles
are constructed in a flexible manner that can result in bending or excessive
pipe strains introduced into components during shipment and at installation.
It is suggested that full length cross members be positioned as a minimum under pumps, coolers, and filters on the equipment baseplate (Fig. 7.3.1). If the
baseplate is to be grouted in the field, grout and vent holes should be specified
FIGURE 7.3.1 Console baseplate construction. (Courtesy of Fluid Systems)
338
Auxiliary Systems
Chapter | 7
and reviewed for accessibility to pore grout when equipment is installed on the
baseplate.
Maintenance Accessibility
Since equipment must be maintained and calibrated while the auxiliary system is in operation, it is important to provide ample personnel space such that
equipment can be maintained safely and reliably without damage to surrounding components. A rule of thumb is to provide approximately 1 m of space
around components for accessibility. Note that this is with the utility lines installed. The review of equipment on a model, cad 3D drawing, or an outline
should be made considering installation of all utility lines that will be installed
in the field.
Online Testing and Calibration Accessibility
Considering that many components (pumps, drivers, coolers, filters, control
valves, instrumentation) will be tested and calibrated with equipment in operation, accessibility for this operation must be considered.
In addition to reviewing the vendor manufactured skids, the placement of all
skids in the field must be reviewed for accessibility. Consideration of the skid
arrangement only to be complicated by installation against a column or wall in
the field will not obtain the objectives of total accessibility.
Utility Supply Arrangement
Care should be given to the routing of all utility (conduits, steam lines, water
lines) supply lines in order to maximize accessibility to the critical equipment
auxiliary systems.
Considerations for Component Disassembly
All components must be able to be dissembled quickly, easily, and safely while
the unit is operating in the field. To meet this requirement, sufficient space
around the auxiliary console must be available for such exercises as cooler
bundle removal, filter cartridge removal, and auxiliary or main driver removal.
In addition, consoles are frequently installed in congested areas and lifting arrangements should be reviewed beforehand to confirm if components can be
removed in a safe and easy manner.
Figs. 7.3.2–7.3.4 depict an actual oil system that incorporates the guidelines
mentioned in this best practice. You can see how this will aid in optimal train reliability by allowing for plant personnel to have access to monitor and maintain
the components very easily.
339
B.P. 7.3
More Best Practices for Rotating Equipment
FIGURE 7.3.2 Best of best oil system view 1. (Courtesy of D. A. Campbell)
FIGURE 7.3.3 Best of best oil system view 2. (Courtesy of D. A. Campbell)
340
Auxiliary Systems
Chapter | 7
FIGURE 7.3.4 Best of best oil system view 3.
B.P. 7.4: Locate auxiliary pump auto start switch or transmitter in pump
discharge header
While the start time of the auxiliary pump is critical to save the unit from a
low oil pressure trip, the location of the auto start instrumentation is just as critical as the signal has to be detected immediately and sent quickly to the pump.
The most effective way is by locating the instrument in the pump discharge
header with minimal sensing tubing as possible, since this will allow for quick
detection of the loss of pressure.
Note that if the pressure signal is sent to the DCS before going to the Motor
Control Center (MCC), there could be some significant delays in relaying this
signal. In this case it is highly recommended to have a priority interrupt on the
Low Pressure Alarm and Start AOP in order to minimize these delays.
L.L. 7.4: Improper location and setup of Auxiliary Oil Pump (AOP) Autostart has resulted in numerous unit trips and lost production
BENCHMARKS
This Best Practice has been in use since the early 1990s and has resulted in
trouble free start-up of the auxiliary oil pump when it was needed to operate.
341
B.P. 7.4
More Best Practices for Rotating Equipment
SUPPORTING MATERIAL
Auxiliary System Function Summary
The system objective, remember, is to continuously supply cool, clean fluid
to critical equipment components at the correct pressure, temperature, and
flow rate continuously for 36 months or 3 years. Some of the major components such as pumps, couplings, and drivers can meet this objective. Others
such as filter elements, control valves, and instrumentation possibly cannot.
Therefore, in order to meet our objective, the system must be designed such
that every major component is spared and or can be maintained while the
unit is operating. Refer to Fig. 7.4.1 and note that for our auxiliary system example every major component with the exception of the reservoir is
spared.
Note that even valves are spared with manual bypasses. Instrumentation
should be designed such that it can be calibrated and maintained or replaced
on site during operation. These conditions having been met, it remains to assure that the entire system can function together to meet the objective. The
objective of the system can be seen in picture form in Fig. 7.4.2. Fig. 7.4.2
plots system flow, pressure, and temperature to the critical components (the
bearings and the control oil system components). The objective is to maintain a steady value of these parameters with time. In this section then, we
will examine all of the major components operating together as a reliable
system to meet the objectives of an auxiliary system in critical equipment
service.
FIGURE 7.4.1 Typical lube oil supply system.
342
Auxiliary Systems
Chapter | 7
FIGURE 7.4.2 Auxiliary system dynamic response.
Auxiliary System Concepts—An Equivalent Vessel and Orifice
Every critical equipment component (bearing, seal, control system component,
etc.) can be treated as an equivalent orifice in a system with an upstream equivalent vessel—refer to Fig. 7.4.3.
As previously explained, a bearing can be treated as an equivalent orifice
with pressure (PS) upstream of the bearing and (PA) for atmospheric pressure in
the bearing drain. In Fig. 7.4.3, PS* represents the auxiliary system supply to the
equivalent vessel upstream of the component being discussed. The equivalent
vessel actually represents a small, infinitesimal section of the supply pipe to the
bearing. Steady state operation dictates that PS* = PS.
Let’s examine the different situations that can occur with an equivalent orifice and equivalent vessel in an auxiliary system. Refer to Fig. 7.4.4.
Fig. 7.4.4 shows the same auxiliary system as pictured in Fig. 7.4.3 but with
different operational cases noted on both the demand side that is downstream
of the equivalent vessel and the supply side which is upstream of the equivalent
vessel. The demand side flow is determined by the flow rate across the critical
equipment components, which as previously stated is determined by the classic
orifice equation. That is, any change in pressure drop across the orifice or in the
orifice size will cause a corresponding change in flow.
Examples of demand side changes are as follows:
1. Gradual orifice wear (bearing or seal wear)
343
B.P. 7.4
More Best Practices for Rotating Equipment
FIGURE 7.4.3 Reduce it to an equivalent vessel.
FIGURE 7.4.4 Auxiliary system operational cases.
344
Auxiliary Systems
Chapter | 7
2. Sudden increase in orifice diameter (loss of bearing lining—babitt)
3. Sudden orifice differential pressure increase (turbine servo piston rapid
movement)
Changes that can occur on the supply side of the equivalent vessel are:
1. Gradual supply system decrease (main auxiliary system pump driver speed
change, lower speed)
2. Sudden supply system increase (sudden control valve bypass closure)
3. Sudden supply system decrease (main pump shutdown)
4. Sudden supply system decrease (relief valve opening)
5. Sudden supply system increase (two pump operation)
6. Sudden supply system increase (pressure reducing valve failure—fail open
mode)
Referring to Fig. 7.4.4, remember that any increase in the demand side flow
will cause a corresponding decrease in pressure in the equivalent vessel. Any
decrease in demand side flow will cause a corresponding increase in equivalent
vessel pressure. On the supply side, any decrease in supply side flow will cause
a corresponding decrease in equivalent vessel pressure and any increase in supply side flow will cause a corresponding increase in equivalent vessel pressure.
We will now examine some of the previous cases and review the total functioning of the auxiliary system in Fig. 7.4.1.
Auxiliary System Function Examples
Control System Sudden Demand
If the steam turbine driver in our example were to experience a sudden load increase, the valve rack in the turbine would suddenly open. To cause this opening
the control oil from the auxiliary lubrication system would suddenly increase in
demand flow. That is, the flow across the equivalent orifice of the control system
would suddenly increase because the downstream pressure on the orifice would
become lower. Referring to Fig. 7.4.4, this action would result in a simultaneous
drop of PS since the demand flow to the control system would instantaneously
exceed the supply flow into the equivalent vessel. The result would be a sudden
drop in pressure PS that could cause a system trip if the pressure were to fall
below the minimum control system supply pressure. To prevent this occurrence,
an additional vessel or supply source must be available to increase PS* at the
same rate as the increased demand of flow to the equivalent orifice. In most control oil systems, an accumulator is usually used to supply this quantity of fluid.
Experience has shown that rapid movement of steam turbine governor valves requires an accumulator since the response rate of auxiliary system control valves
is usually not rapid enough to account for the sudden increase in demand flow.
Bearing Wear
In the case of gradual bearing wear, the diameter of the equivalent bearing orifice would increase. As a result, the flow across that equivalent orifice would also
345
B.P. 7.4
More Best Practices for Rotating Equipment
increase. Referring to Fig. 7.4.4, the supply flow to the orifice would increase, therefore increasing the demand side flow and thus causing a corresponding decrease in
the pressure of the equivalent vessel. The auxiliary system bypass valve sensing system pressure decrease would close to maintain the preset pressure value in the auxiliary system. Remember the flow is incompressible. Consequently, change in any
part of the system will be affected in another part of the system. The bypass control
valve will close to maintain the preset pressure thereby diverting excess pump flow,
which normally is recirculated, to the critical equipment to account for the increased
demand flow required by the bearing wear change. It is important to remember that
in all auxiliary systems, pressure is used to determine flow rate change. This is a result of the accuracy of pressure readings as opposed to flow readings. Since the fluid
is incompressible, rapid changes in flow produce corresponding changes in pressure
which result in fast reaction times when pressure is used for control.
Auxiliary Pump Auto Start
If the main pump (steam turbine driven) were to suddenly trip, the equivalent
vessel in Fig. 7.4.4 would experience a sudden drop in pressure since the supply
flow would be less than the demand flow. This concept is used in all auxiliary
system auto start systems. Sensing a sudden decrease in system pressure, a pressure switch sends a signal to start the auxiliary or stand-by pump immediately.
The pressure switch setting should be as high as possible but not too close to
cause spurious starts of stand-by pumps. If the stand-by pump rate of flow does
not equal the demand flow rate, the pressure in the equivalent vessel will continue
to decrease and the unit will shut-down. Therefore, the necessity of rapid standby pump start-up and rapid closure of bypass control valves can be appreciated.
Simply stated, the net effect of the stand-by pump acceleration and the closure
rate of the bypass valve must produce a supply flow to the equivalent vessel equal
to or greater than the demand side flow from that vessel to avert a system shutdown. Consequently, if the auxiliary pump auto start instrument is located too
far away from the pump discharge or has a long run of sensing lines, the ability
to start the aux pump up in time will also be affected. Therefore, it is essential to
have the auxiliary pump auto start located in the pump discharge header and have
minimal run of sensing line tubing. Frequently operators deduce that the system
is not designed correctly if a trip were to occur on low pressure in this scenario.
It must be understood that the response time of the bypass control valve is dependent on the condition of the valve (valve stem friction) and the condition of
the sensing line and controller (if furnished to that valve). Frequent checking of
control valve response time is recommended. Also in this example, a drift of the
pressure switch setting from a higher to a lower setting would also cause a slower
response time. Any restriction to the pressure switch, (a block valve partially
closed, etc.) would also cause a change of response time of the system.
Remember the concept of “sub-systems” when troubleshooting. In this example, the pump sub-system consists of the pump units, bypass control valve,
valve sensing ports, controller pressure switch, and auxiliary pump motor
346
Auxiliary Systems
Chapter | 7
starter. The malfunction of any component can result in the shutdown of the
critical equipment.
Two Pump Operation
Typically, auxiliary pump start-up is checked with the unit on line to assure the
integrity of the automatic pump start system. This action will cause a sudden
system pressure increase since the supply flow to the equivalent vessel will significantly exceed the demand flow from the vessel. As a result, system pressure
will instantaneously rise. If the opening rate of the bypass valve is not equal to
the rate of flow increase from the stand-by pump, system pressure can significantly rise and exceed the relief valve setting. This occurrence is undesirable
since the risk of not reseating the relief valves exists and as previously mentioned will introduce another equivalent orifice into the system that will reduce
the supply flow to the critical equipment. This action could cause unit shutdown
if the net supply flow is less than the demand flow of the critical equipment
under that condition. Therefore, the condition of bypass valve system must be
operationally checked to assure its response time in the event of sudden standby pump start with main pump in operation.
Transfer to an Empty Cooler-Filter Bank
Let us assume that differential pressure indication across the filters indicates
that filters are becoming clogged and need to be changed. Utilizing the transfer valve, operation can be diverted from the dirty filter cooler combination to
the stand-by filter cooler combination. Operators have checked to assure that the
system to be transferred to is properly filled and purged of air. In this case
the transfer is accomplished without any change of supply flow to the equivalent
vessel. Therefore, the pressure in the system remains constant. However, when
the filters were changed the filter vessel was not refilled with oil.
Approximately 8 months later the operating side filter indicates dirt retention and again it is anticipated that the system will be transferred without checking the condition of the bank of cooler and filters to be transferred to. The result
is the introduction of a large equivalent vessel into the system which receives
pump flow but does not distribute flow into the system until the vessel is full.
Assuming the capacity of the cooler, filter and interconnecting pipe is 120 gallons and the flow to the system is a maximum 120 gallons with both pumps
running, it would take 1 min to restore pressure to the system. In this example,
a trip of the critical equipment would definitely occur.
The previous examples are but a few of the cases that can occur in an auxiliary system. It is a good practice to always “what if” operation cases with any
system to fully understand the total function of the system involved. Remember
that in any auxiliary system, system responses are based on pressure signals
which sense rapid reduction of flow. The prime operation objective of any system is to equalize the supply flow rate and the demand flow rate in the system as
quickly as possible. If this cannot be accomplished, the instrumentation of the
347
B.P. 7.4
More Best Practices for Rotating Equipment
FIGURE 7.4.5 For successful auto-transfer of lube/seal oil pumps.
system is designed to shut-down the critical equipment in a time that will not
destroy the equipment.
Having now examined component selection sizing and function and the total
function of an auxiliary system, we are now ready to continue our discussion of
factors that determine auxiliary system reliability.
We will now divert our attention in subsequent chapters to the specification
of auxiliary systems, the auxiliary design audit, factory tests, and the installation and operation of the auxiliary system on site.
Transient Case Pump Auto Transfer
By far, the most common cause of auxiliary system related unit shutdowns is the
transient case of pump auto transfer. Fig. 7.4.5 presents the system requirements
for successful auto transfer.
When confronted with transient response problems, it is helpful to think of
system requirements in terms of lube oil flow per second. This is because transients usually last on the order of 3 s. Fig. 7.4.6 presents a plot of oil flow to a
unit expressed in gallons per second versus time in seconds. This plot shows a
typical transient for:
l
l
l
Main pump decreasing flow
Auxiliary pump increasing flow
Additional unit flow from the bypass valve
Since oil is incompressible, the requirement for a successful pump auto
transfer in Fig. 7.4.5 would be that at any time, the total flow to the unit must be
equal to or greater than the oil flow requirement. For this example, the oil flow
requirement is 1.5 gallons per second [90 gallons per minute (GPM)]. If the
total flow to the unit does not meet this value, a low oil pressure trip will shut
down the unit.
348
Auxiliary Systems
Chapter | 7
FIGURE 7.4.6 Transient response—lube/seal system.
It can be seen from Fig. 7.4.6 that any variation of auxiliary pump start-up
time or bypass valve closing time could cause a unit trip. An effective on-site
functional test program is required to assure proper auxiliary pump start-up
times and bypass valve closing times.
In view of the exposure to unit trips resulting from auxiliary pump starting
times and bypass valve closing times, an accumulator is often installed. Serious consideration should be given to the requirement for properly sized and
installed accumulators on all critical (un-spared) units.
B.P. 7.5: If an oil system sub-vendor is used A design audit shall be conducted with them present, along with a shop audit of the sub-vendor
Very frequently, the oil system is subcontracted out to another vendor. It is
essential that the Machinery vendor and the sub-vendor communicate properly
to assure there are no issues in design of the oil system. For this, a design audit
of the system should be conducted with a design engineer of the sub-vendor
present as early as possible in the project.
This design audit should be conducted at the sub-vendor’s facility in order
to conduct an audit of the sub vendor’s capabilities. The audit of the sub-vendor
should consist of a tour of the facility to see the machining capabilities and
workload as well as a review of previous systems they manufactured with similar flow rates and pressures that have been in service for more than 2 years.
L.L. 7.5: Failure to design audit new oil system and component design
from sub-vendors has caused many start-up delays and trips of critical (unspared) compressor trains
349
B.P. 7.5
More Best Practices for Rotating Equipment
Even at the present time, a design audit of oil systems is not common. Failure to conduct an audit can result in shop testing, start-up delays, and troublesome low reliability oil consoles, especially if the provider is unknown and has
limited experience with your application.
BENCHMARKS
This best practice has been in use since the early 1980s to assure optimum oil
console design and component selection that has resulted in critical machine
reliability exceeding 99.7%.
SUPPORTING MATERIAL
Design Audit Agenda
In this section, we will be dealing with the specific areas important to the confirmation of auxiliary system design and manufacture. To assure maximum effectiveness of these reviews, it is recommended that a prior agenda, mutually
agreed upon between OEM and user, be generated and supplied to both parties
well in advance of any meetings. In addition to detailing subjects of the discussion, the agenda should also define the attendees of the meeting. A well-defined
meeting is still ineffective if the participants are not familiar with the subject or
have a minimum amount of experience.
Confirmation of Scope
For the reasons mentioned previously, scope review approximately 1–2 months
after auxiliary system order placement is recommended. The major areas of
scope review are:
l
l
l
Schematic review
Data sheet review
Exceptions to specification
Schematic (P&ID) Review
The original system schematic (P&ID) (console and unit) as contained in the
equipment specification should be reviewed at this point to confirm all system
logic and instrumentation is as specified. That is, the schematic should be reviewed in the framework of a P&ID (process and instrument diagram). All comments should be noted and the system schematics corrected.
Data Sheet Review
The system data sheet should be thoroughly reviewed and be complete at this
point to include specific component details and desired manufacturers of major
350
Auxiliary Systems
Chapter | 7
components. This review goes both ways. That is, vendor required information
and user information must be detailed and correct on the data sheets. Frequently
utility information and site information is not complete. This absence can only
lead to reliability and communication problems in the field. As with any meeting, detailed minutes should be kept and every effort be expended to resolve all
open items prior to conclusion of meeting. Postponing decisions only creates
inefficiencies.
Exceptions to Specifications
All vendor exceptions to specifications must be reviewed and either be accepted
or rejected. The final, mutually agreed to list of vendor exceptions should become part of the job specification.
Component Sizing Audit
A typical component sizing audit form is included at the end of this B.P. We
will now review the major areas of this audit form and comment relative to the
specific items.
System Requirements
The first subject of discussion concerns confirmation of auxiliary system flow
rates, pressures, and heat loads required. This information determines the size
of all major system components. It must be correct and not subject to modification during the design process of the equipment. The component sizing agenda should emphasize the need to have all required information furnished and
confirmed by each critical equipment vendor. Attention is drawn to comparing
values noted. If significant discrepancies appear, question them! Remember all
critical equipment components are equivalent orifices and at a specified pressure will only pass a given flow. If the component oil flow specified is greater
than the amount the components will actually pass, the excess oil will be bypassed back to the oil reservoir and could create overheating problems in the
system. Conversely, if too low a value of component oil flow is specified, a
system may continuously operate with both main and stand-by pump in operation since the capacity of the main pump will have been sized too small for the
system.
Reservoir Sizing, Construction, and Sub-Component Details
Refer to Fig. 7.5.1 which is a schematic representation of an auxiliary system
reservoir. Reservoir size and levels as noted in Fig. 7.5.1 must be determined at
this time. Size will be a function of system flow which previously will have been
defined. The height of the reservoir should be such that in its final field location,
it will provide adequate gravity return from the main equipment.
351
B.P. 7.5
More Best Practices for Rotating Equipment
FIGURE 7.5.1 Schematic representation of an auxiliary system reservoir.
The construction of the reservoir should be checked at this time. The original equipment vendor should have a reservoir drawing that details the reservoir
internals available for review. Attention is drawn to the requirement that auxiliary return fluid should not be allowed to free fall to the surface of the liquid.
All returns should be through stilling tubes or sloped troughs. It is also wise to
confirm that internal design is proven and that the manufacturer has successfully
designed similar reservoirs in the past. Accessibility for cleaning should be confirmed and the location of return connections and pump supply nozzles should
be such that maximum residence time of system fluid is assured. Material of
construction should be confirmed at this point and all details of the following
reservoir sub-components reviewed:
l
l
l
l
Reservoir heater sizing calculations
Level control alarm
Connection locations and size
Additional instrumentation
Pump and Driver Sizing
Pump Performance
Regardless of the types of pumps used, centrifugal or positive displacement, the
performance curves should be reviewed at this point.
1. Positive displacement—Positive displacement pumps, furnished without
external timing gears, are mechanically sensitive to fluid viscosity. The performance curve should be checked at all operating points to confirm that
adequate rotor separation is present at low fluid viscosities. If an operating
352
Auxiliary Systems
Chapter | 7
FIGURE 7.5.2 Screw pump performance. (Courtesy of IMO Ind.)
point is at the end of within 20% of a pressure versus flow curve at a low
operating viscosity [7.4–10 cSt (50–60 SSU)], the pump vendor should be
contacted to confirm a correct selection has been made. Refer to Fig. 7.5.2
for an example of this case.
2. Centrifugal pumps—Since centrifugal pump performance must be corrected
for viscous fluid operation, pump sizing must confirm that the actual operating points are not close to the operating extremities of the corrected curve.
That is, any operating point should not be less than 20% of pump best efficiency point nor be more than 110% of best efficiency point of the operating
curve corrected for viscosity. Refer to Fig. 7.5.3 for an example. Operation
outside the stated boundaries, in addition to causing high revenue costs due
to lower efficiency, can jeopardize the reliable operation of the pump.
Pump Mechanical Requirements
Pump data sheets must be checked at this point to confirm that proper pump
case material design, bearings, seals, and pump flushing arrangements are provided as specified. In addition, it is recommended that all pumps be factory
tested prior to the auxiliary system test to confirm acceptable operation.
353
B.P. 7.5
More Best Practices for Rotating Equipment
FIGURE 7.5.3 The effect of oil viscosity on centrifugal pump performance.
Pump Unit Couplings
Couplings should be selected for the maximum driver horsepower and include
a sizing safety factor (usually 20–25%) above the maximum driver horsepower
rating. Spacer couplings are recommended in order to provide ease of maintenance and minimize the necessity to remove a pump or driver while the critical
equipment unit is operating. The type of coupling selected should be of high
quality and reliability and provide a minimum of 3 years continuous operation.
While either batch lube gear type couplings or dry flexible element types can
be used, the latter types are preferred for their low maintenance requirements.
Coupling material should be steel as opposed to cast iron to prevent breakage
during removal or during extreme temperature changes (as during a fire). Flexible elements should be stainless steel.
The coupling shaft fit configuration and amount of shrink fit should be
checked to confirm correct values.
Driver Sizing
Driver sizing must be confirmed to assure adequate delivered horsepower during
all operating conditions. Utility conditions to the drivers should be rechecked at
this point to assure that values are as stated on data sheets. As an example, steam
turbine data (inlet pressure and temperature and exhaust temperature) should be
checked so that all conditions as stated will exist on site. Similarly, minimum
starting voltage for motor drivers should be confirmed. Lower minimum starting
voltage values than stated on data sheets will cause stand-by pump start time to
be less than anticipated, shorten motor life, and could result in serious transient
auxiliary pump start problems that could cause critical equipment shutdown.
354
Auxiliary Systems
Chapter | 7
Driver sizing must be confirmed with specification requirements such that
driver horsepower equals pump horsepower times a specified service factor. Selection charts for expansion turbines should be checked and confirmation that
proper standard size electrical drivers have been selected should be checked. In
applications where viscous fluids are used, pump calculations for horsepower
corrections at maximum fluid viscosity must be confirmed. Attention is drawn
to realistic sizing of pumps and drivers concerning viscosity. If minimum site
ambient is below 40°F, for example, and a properly sized reservoir heater is
furnished, there will not be a requirement for high viscosity operation if it is
accepted that reservoir heater will bring the auxiliary fluid to a minimum pump
starting temperature prior to pump operation. A permissive temperature switch
could be installed to preclude the possibility of equipment start prior to acceptable temperature conditions.
Driver Mechanical Requirements
Data sheets for both main and auxiliary drivers should be checked to confirm
proper mechanical design.
Motor drivers should be designed as specified with attention being paid to
bearing design and motor housing design. Many smaller auxiliary systems have
utilized aluminum frame motors in the past. Due to the high coefficient of thermal expansion of aluminum (double that of steel), these motors are subject to
significant alignment changes with operating temperatures and could cause coupling misalignment problems.
Expansion turbine mechanical review should include governor system confirmation and safety system confirmation. Some safety valves furnished with
small expansion turbines are not designed for positive shut off. This can result
in operation of the turbine at lower speed once the equipment has been tripped.
Most steam turbines presently operating in auxiliary systems, do not have speed
indicators. To assure correct operating speed a stroboscope or hand-held tachometer, both of which can give inaccurate readings, are used. Particularly in
the case of dynamic pumps, turbine speed setting is important to assure proper
flow to the system. Therefore, any new installation incorporating expansion turbines should be equipped with speed indication.
Relief Valve Selection
Relief valve selection should be confirmed to qualify proper size, minimum
accumulation (the pressure required over the valve setting to provide full flow)
and chatter free operation. Relief valves should be located as close as possible
to the pump discharge line to minimize the possibility of air entrainment in the
line to the relief valve which can result in a delayed pump flow to the unit. This
would be the case if the RVs were mounted on the reservoir a significant distance from the pump discharge line.
355
B.P. 7.5
More Best Practices for Rotating Equipment
FIGURE 7.5.4 Typical lube oil supply system. (Courtesy of M.E. Crane Consultant)
Control Valve Selection
Control valve data sheets for each control valve in the system should be available
for review. Information furnished on control valve data sheets should be complete
in terms of valve sizing, actuator selection, and valve controller (if furnished).
Valve Cv—All operating valve coefficients (Cvs) should be stated on the
control valve data sheet. That is, the normal Cv, maximum Cv, and minimum Cv.
These values should be compared with the selected valve internals to assure that
all operating conditions fall within 10–90% of the maximum valve coefficient.
Failure to confirm this can lead to valve instabilities. When reviewing valve coefficients, the system design must be reviewed (system schematic) since certain
changes in the system could render the valve unstable.
Bypass Valve
For this application, the valve back pressure is atmospheric and the control
valve differential depends on the condition of the auxiliary system cleanliness
and any additional control valve setting (refer to the typical system schematic
Fig. 7.5.4).
l
l
Cv Minimum—The minimum valve coefficient in this application would be
with a dirty filter (filter ∆P) and one pump operating.
Cv Maximum normal—The normal valve coefficient in this application occurs with a clean system (minimum filter ∆P) and one pump operating.
356
Auxiliary Systems
Chapter | 7
FIGURE 7.5.5 Control valve flow characteristics. (Courtesy of Fisher Controls Inc)
l
Cv Maximum—The maximum valve coefficient would be with the main and
auxiliary pumps operating and the minimum pressure drop across the valve
(clean filter).
The maximum flow for this condition would be the normal bypass flow of
the main pump plus the total flow of the auxiliary pump.
Attention is drawn to determine the characteristic of the valve curve for this
application. The normal operating point would be approximately the minimum
Cv therefore, a valve characteristic that results in a fairly significant (15–25%)
valve travel for this small Cv would be desired (quick opening). Two pump
operation (maximum Cv) is an abnormal case. Therefore the valve should be
designed merely to pass this flow (refer to Fig. 7.5.5) at 90% or less of the valve
catalogue Cv.
Pressure Reducing Valve
In a centrifugal pump application, pressure reducing valves would experience
minimum, normal, and maximum Cvs similar to bypass valves with the exception that downstream valve pressure will change with increasing flow.
When pressure reducing valves are used to reduce pressure levels (control
oil pressure to lube oil pressure, seal oil pressure to lube oil pressure, etc.) the
valve Cv should be selected for all possible operating cases as mentioned previously for bypass valves. Care should be taken to assure all possible operating
cases are considered.
357
B.P. 7.5
More Best Practices for Rotating Equipment
Temperature Control Valves
The temperature control valve Cv will remain relatively constant under all
auxiliary system conditions. In the case of a two way valve however, the
valve must be sufficiently sized such that the full flow pressure drop across
the valve is less than the clean pressure drop across the cooler in parallel with
this valve.
Differential pressure control valves and level control valves are sized and
examined in the same manner as described previously for bypass and pressure
reducing valves. Details will be discussed in subsequent chapters. Viscosity corrections are required for all control valve sizing when operating viscosities exceed 7.4 cSt [50 Sabolt Universal Seconds (50 SSU)]. Significant size increases
are required for high viscosity operation approaching 220 cSt (1000 SSU) on
the order of 1½ to 2 times the selected valve coefficient without viscosity considerations.
Control Valve Sensing Line Snubber Devices (Dampers)
If these devices are furnished, a review of device design and confirmation of
proper installation should be confirmed. Such devices provide unrestricted flow
in one direction and restricted flow in another direction. The total auxiliary system operation must be reviewed in this light to confirm proper installation and
orientation.
Supply Pipe Velocity Checks
The pump header, interconnecting console pipe, and piping to the unit should all
be checked for proper fluid velocity or pressure drop. Typical velocity values in
auxiliary system supply pipes are on the order of 4–6 ft./s velocity. Velocity is
derived from the following equation for incompressible flow.
Q ft./ min = A × V
where A = internal pipe area (m2 or ft.2), V = fluid velocity (m/s or ft./s).
Charts for standard pipe sizes and schedules are available to determine velocities (Fig. 7.5.6).
Note that schedule 80 is usually used for carbon steel pipe below 2 in.
Schedule 40 is used above 2 in. For stainless steel pipe, schedules 10 and 20 are
used respectively.
Typical drain line velocities are 0.15–0.08 m/s (1/2–1/4 ft./s). Attention is
drawn to properly sizing drain pipes for installations where critical equipment
is significantly elevated above reservoir. All drain pipes should be sized with
adequate area to preclude excessive air being entrained with the oil to promote
drainage back to the reservoir. An additional consideration for supply headers
at the unit is that supply headers are frequently sized for one standard pipe dimension. In the case of large critical equipment units (two or three bodies and
358
Auxiliary Systems
Chapter | 7
FIGURE 7.5.6 Typical pipe sizing chart.
driver), the amount of oil from the entrance to the header to the last component
decreases significantly. In an effort to minimize pipe size, many vendors size
headers small. Therefore, pressure drop in the header is excessive and requires
a higher supply header pressure than anticipated in the unit design. Improper
sizing of critical equipment supply headers could cause excessive flow across
equivalent orifices (bearings) thereby requiring all flow of the main pump and
necessitating the operation of the auxiliary pump.
Transfer Valve Sizing
Transfer valve configuration and materials of construction should be confirmed
at this point. Transfer valve design should be checked to confirm tight shutoff.
Cooler Sizing
The cooler data sheet should be reviewed to confirm correct duty, confirmation
of correct cooling media details, fouling factors and materials of construction.
Filter Sizing
Filter information should be reviewed to confirm correct filter sizing for the
normal and the maximum viscosity case (in the case of viscous fluids). Additionally, maximum filter collapse pressure, internal filter cartridge design, and
cartridge sealing design should also be reviewed at this time.
359
B.P. 7.5
More Best Practices for Rotating Equipment
Instrumentation
All instrumentation should be reviewed to confirm proper selection, materials
of construction, and proposed installation locations. All instrumentation loops
should be reviewed to assure that critical instrumentation can be calibrated and
maintained while unit is in operation.
Console Layout and Component Arrangement
Having confirmed the acceptable component sizing and selection, the console
layout and arrangement of components must be reviewed. Methods of this review incorporate either the review of outline drawings of the proposed arrangement, a model review, or a cad 3D drawing review. Many vendors and users
have found that models or cad 3D drawings aid greatly in understanding and
reviewing maintenance accessibility and layout considerations.
Console Construction
Auxiliary equipment consoles or modules house most of the components present in the auxiliary systems. Their construction should be reviewed to assure
proper stiffness and facilities for installation on site. Many horizontal consoles
are constructed in a flexible manner that can result in bending or excessive
pipe strains introduced into components during shipment and at installation.
It is suggested that full length cross members be positioned as a minimum
under pumps, coolers, and filters on the equipment baseplate (Fig. 7.5.7). If the
baseplate is to be grouted in the field, grout and vent holes should be specified
and reviewed for accessibility to pore grout when equipment is installed on the
baseplate.
FIGURE 7.5.7 Console baseplate construction. (Courtesy of Fluid Systems)
360
Auxiliary Systems
Chapter | 7
Maintenance Accessibility
Since equipment must be maintained and calibrated while the auxiliary system is in operation, it is important to provide ample personnel space such that
equipment can be maintained safely and reliably without damage to surrounding components. A rule of thumb is to provide approximately 1 m of space
around components for accessibility. Note that this is with the utility lines installed. The review of equipment on a model, cad 3D drawing, or an outline
should be made considering installation of all utility lines that will be installed
in the field.
Online Testing and Calibration Accessibility
Considering that many components (pumps, drivers, coolers, filters, control
valves, instrumentation) will be tested and calibrated with equipment in operation, accessibility for this operation must be considered.
In addition to reviewing the vendor manufactured skids, the placement of all
skids in the field must be reviewed for accessibility. Consideration of the skid
arrangement only to be complicated by installation against a column or wall in
the field will not obtain the objectives of total accessibility.
Utility Supply Arrangement
Care should be given to the routing of all utility (conduits, steam lines, water
lines) supply lines in order to maximize accessibility to the critical equipment
auxiliary systems.
Considerations for Component Disassembly
All components must be able to be dissembled quickly, easily, and safely while
the unit is operating in the field. To meet this requirement, sufficient space
around the auxiliary console must be available for such exercises as cooler
bundle removal, filter cartridge removal, and auxiliary or main driver removal.
In addition, consoles are frequently installed in congested areas and lifting arrangements should be reviewed beforehand to confirm if components can be
removed in a safe and easy manner.
This completes comments concerning the component sizing audit. All
changes made during this meeting should be documented and followed up to
guarantee that final component design and arrangements are as specified and
agreed to in this meeting.
Factory Testing and Inspection
Having properly specified, designed, and manufactured the unit, it remains to
confirm proper arrangement and operation. This is accomplished during factory
361
B.P. 7.5
More Best Practices for Rotating Equipment
testing and inspection. The objectives then, of this phase are to confirm the
proper arrangement details and functional operation of the equipment.
Test Agenda
To meet the objectives of this phase the equipment must be thoroughly tested
prior to field installation. The test should confirm the functional operation of
all components as they will operate in the field. In order to assure a valid factory test, a test agenda should be prepared approximately 2 months before test
date and reviewed by the equipment purchaser. Specific areas of concern are as
follows.
Flushing
Component system flushing is required as an inspection point and should be
accepted prior to the initiation of the test. Additionally, all test agendas should
always be structured such that a limit for each item to be tested is specifically
defined in the test agenda. The flushing acceptance criteria must be mutually
agreed upon and be adhered to during the review of flushing the operation. Following is an industry best practice flushing procedure:
Auxiliary System Flushing Procedure
Courtesy of M.E. Crane Consultant
The following procedure is presented as a guide for field flushing of lube and
seal systems. In order to be fully productive, it is recommended that all requirements noted herein be strictly followed.
1. General
1.1 Flushing operation will be carried out by the designated party. (Contractor or Owner)
1.2 Cleanliness of oil console, equipment skid, overhead seal oil tanks,
piping systems and screens shall be determined by mutual agreement
between equipment vendor, contractor and owner.
1.3 Owner and vendor shall keep a log for general review of flushing progress. Master flow sheets shall be kept by owner and updated to progress. An entry shall be made during each shift.
1.4 In general the oil flush shall be performed using selected permanent
auxiliary equipment which is part of the vendor supply package. This
will include the following:
l Auxiliary oil pump (electrical) and main oil pump if possible.
l Main oil filter to be in position for all flushing.
l Main oil reservoir, degassing tank, overhead seal oil tanks and seal
oil traps.
l Skid piping.
l Selected instruments and controls.
362
Auxiliary Systems
Chapter | 7
2. Preparation
2.1 Any residual oil from factory testing must be removed from the reservoir and filters. Relief valves must have been checked prior to flushing.
2.2 The reservoirs, degassing tanks and filter casing must be wiped clean,
inspected and approved by the owner. All cleaning must be carried out
using a lint free cloth. When filters are open for cleaning, special care
should be taken to avoid contaminates falling into ‘clean’ side of filter
housing.
2.3 The filters must be verified to be in place and satisfactory for flushing.
Examination of filters will include checks for bypassing, inside or outside of filter housing.
2.4 All lube and seal oil interconnecting piping will be installed consistent
with normal operating conditions in accordance with bypass piping arrangements as mutually agreed upon between vendor and owner.
l All equipment supply and drain piping is required to be flushed during entire flushing operation. Location of all valves, bypasses and
screens shall be in accordance with marked-up P&ID’s of lube oil,
seal oil and control oil system.
l Add hand valves suitable to meet the operating requirements during
flushing to all piping supply points.
l All lines to the steam turbine throttle valve, servo motors and dump
devices will be flushed in accordance with a manufacturer’s requirements.
l Overhead seal oil tanks will be flushed by jumpering to compressor
reference gas lines between compressor, overhead tanks and drainers are required to be flushed.
2.5 Stainless steel 100 mesh screens with back up 60 mesh screens with a
number of spares must be fabricated with retaining gaskets and installed
at selected lube oil piping flanges. This fabrication involves cutting and
fitting the screen to the gaskets. Screens must be clearly and permanently tagged for ease of identification. Location for screens must be
agreed with by vendor and owner. Basically, they should be positioned
at all inlets to the machine. Locations immediately after risers must be
avoided. Additionally, 100 mesh screens with 60 mesh back up screens,
will be installed at the main oil return, degassing tank inlet and return
and reservoir oil fill connection. These screens will be in place during
all flushing operations. A drain should be fitted at lower point of return
lines ahead of screen in order to deal with a blocked screen. Also, a
pressure device (manometer-ie: simple length of plastic tubing) is to be
installed at this point to monitor any pressure build-up due to blockage.
2.6 The reservoir shall be filled with the lube oil specified for permanent
plant operation unless directed otherwise by the owner.
2.7 Lube oil flush should not occur until the compressor skid has been fully
grouted.
363
B.P. 7.5
More Best Practices for Rotating Equipment
2.8 The compressor rear bearing port cover and the load coupling guard
must be installed in order to minimize any oil spill during flushing.
2.9 The vent piping must be checked out for mechanical completeness.
2.10 Check for correct operation of the auxiliary lube oil pump on/off/auto
switch and the oil high temperature alarm before commencing flushing.
2.11 The instruments required for the flushing operation shall be identified
on a P&ID mark-up for flushing and will be calibrated for normal
operation prior to starting the oil system.
2.12 Add nitrogen bottles or instrument air connections at suitable tapping
points downstream of lube and seal oil filters.
2.13 Water to oil coolers shall be provided.
2.14 An auxiliary boiler will be provided to heat the oil to approximately
180°F.
3. Flushing procedure
3.1 The range of temperatures for the hot oil circulation flush shall be
120°F to 180°F. Before initial circulation the oil should be heated to
approximately 120°F.
3.2 The following parameters shall be documented in the log on an hourly
basis:
l Pump discharge pressure
l Bearing header pressure
l Oil reservoir temperature
l Bearing header temperature
l Oil filter differential pressure
l Filter in use
l Sections of piping being flushed
l Start time
3.3 The drain oil sight flow gauges shall be continually monitored for
flow at all places. The complete filling of the sight glass indicates a
flow blockage. Immediate action shall be taken to stop circulation and
clean filter screens. The debris obtained on the screen shall be collected into plastic bags, identified by screen location, machine number
and time.
3.4 The following schedule shall be used for flushing:
l Add 100 mesh screen with 60 mesh backup screens at oil reservoir
and degassing tank if applicable as stated in paragraph 2.5 and
flush through total system at 15 minute intervals until screens are
reasonably clean. Monitor closely the return lines on a continuous
basis to ensure system is not backing up with oil.
l Flush through total system at intervals of one hour until screens are
reasonably clean.
l Flush total system, alternating through systems section until
screens are clean. Supply lines shall be alternated to ensure a greater than 150 percent oil flow is maintained at all times.
364
Auxiliary Systems
Chapter | 7
l
3.5
3.6
3.7
3.8
3.9
3.10
Add 100 mesh screens with 60 mesh backup screens to lube and
seal oil inlet lines to all equipment. Also add 100/60 mesh screens
to overhead and seal oil tanks, reference lines and coupling guard
feed lines (if furnished).
l Repeat ‘flush total system’ above.
l Alternate flushing through each filter/cooler section, control valves
and their bypasses, overhead tanks, seal oil traps and reference gas
lines. Record Note: A differential pressure of 15 PSIG across either filter indicates the need for filter change.
l Bubble nitrogen through system at regular intervals. Record
l Flush through all instrument connections. Record
l Flush through all pressure control valve impulse lines. Record
l Thermoshock the system by use of the lube oil coolers at regular
intervals (varying oil temperature between 120°F – 180°F). Record
l Rap exposed piping with a fibre hammer at one hour intervals.
Record
When the 100 mesh screens meet criteria in paragraph 3.9/3.10 reinstate all instrumentation, orifices, pipe spools, etc. Arrange entire
system in normal, complete configuration with all controls, alarms,
etc., in operation.
Add 100 mesh white cloth (backed with 60 mesh screen) to all lube
and seal oil supply points on equipment bearings, seals and control oil
inlets.
Flush until the criteria as stated in paragraph 3.9/3.10 is achieved, but
for a minimum period of 24 hours.
Note: During final flush, alternate flushing through each filter/cooler
section, control valves and their bypasses.
Whenever practical, circulation shall be continued from the time of
startup until completion on a 24 hour per day basis. Any irregularities
shall be immediately reported to the vendor representatives and the
applicable owner representative as designated, which will be posted
on the accessory skid control panel.
Owner’s quality control representative will monitor all operations for
compliance and verify all records, test parameters and acceptance criteria on a surveillance basis. Final screen particle count will be verified and recorded by quality control.
The oil system acceptance criteria which shall be the basis for witness
approval parameters for contractor and/or owner shall be as follows:
Screen contamination shall be within the particle count limits according to the size of pipe it will be determined as follows:
20 Non-metallic particles on pipe 1″ to 2″
50 Non-metallic particles on pipe 2″ to 4″.
Particles shall not be metallic and shall display random distribution on
the screen.
365
B.P. 7.5
More Best Practices for Rotating Equipment
3.11 The acceptance of the system shall be after installation/inspection 100
mesh cloth covered screens, then circulating the lube oil an additional
four hours and re-inspecting the screens. The final four hours flush
should be with valves open allowing full flow through the system at
operating temperature. Final acceptance requires an oil analysis to determine metal content, viscosity and water content.
3.12 Following acceptance the system shall be restored by the owner for
normal operation including the following actions:
l Remove all screens
l Visually inspect overhead seal oil tanks, degassing tanks, drainer
modules, if required, wipe clean with lint-free cloth.
l Replace all components. Clean filter cartridges (less than 5PSI
pressure drop) can remain subject to owner approval. If new cartridges are fitted, the cleanliness of the system must be rechecked.
l Restore reservoir oil to normal operating level.
l Obtain oil analysis.
Confirm Arrangement Details
Prior to commencement of test, proper arrangement of all components, controls
and instruments should be confirmed to agree with design requirements. Any
discrepancies should be corrected prior to functional test.
Confirm Proper Test Fluid and Capacity are Present Prior
to Initiation of Test
Temporary Test Setup
If it is necessary to import utilities or switches for test that are not normally
furnished, as in the case of steam turbine, steam generator, and control switches,
these items must be confirmed prior to test initiation. In addition, a means of
confirming proper flow rates, temperature, and pressure during the test must be
provided. Supply lines to the unit must be provided with properly sized orifices
to duplicate unit flow requirements.
Functional Testing
Having confirmed proper test set up, calibration of all instrumentation, proper
fluid and test instrumentation, the functional test is now ready to be performed.
As a minimum the following tests should be performed on any auxiliary system
console:
l
l
l
Relief valve test (if supplied)
Transfer valve test
Auto start test of auxiliary pump with the following conditions:
l Main pump tripped
l Two pump operation (main pump in operation, standby pump started)
366
Auxiliary Systems
Chapter | 7
During all functional testing, any system pulsations or pressure drops above
specified values are reason for non-acceptance of test. All components not meeting requirements must be corrected and units must be completely retested.
The value of testing the auxiliary console with the unit should be seriously
considered. Since the console and the unit piping form the specific auxiliary
system, there is a significant benefit to testing both together. Additional costs
of such a test should be evaluated against the potential reduction of reliability
and loss of operation time in the field if any malfunctions exist that were not
determined by test of the console alone.
Testing the console without the unit assumes that a model of the actual
equivalent critical equipment orifices has been properly installed during the test.
There is no assurance that the actual manufacturer of the equipment does not
incorporate changes in equivalent orifices sizes. This would result in different
flows to the unit and different console responses for various transient operational
modes. Daily revenue should be considered in evaluating the extra cost involved
for a full console unit test in a manufacturers works. If a full unit test will be performed, it is still recommended that the console be tested at the point of manufacture prior to the unit test in the critical equipment vendors shop. This action
will determine console design problems prior to shipment to the vendor’s plant.
I. System requirements—see Fig. 7.5.8 for input data
1. Total pump flow
(Positive displacement pump)
Total pump flow
(Centrifugal pump)
2. Bypass flow
(Positive displacement pump)
= ______ × equipment flow (la)
= ______
= ______ (total flow in la)
3. Total heat load (from la)
= ______ BTU/h or kW
4. Pump discharge pressure
Viscosity [(SSU) or centistokes]
A. Lube oil pressure (at equipment)
B. Elevation ∆P
C. Pipe ∆P
D. Valve ∆P
E. Cooler ∆P
F. Filter (clean) ∆P
G. Miscellaneous ∆P
Pump discharge press
(Add A through G)
______ ______
______ ______
______ ______
______ ______
______ ______
______ ______
______ ______
______ ______
______ ______
______ ______
Pressures in kPa or PSI
= (1) – la total flow
= ______– _______
= ______________
Note: If system is combined with seal and/or control oil, add highest value to determine pump discharge
pressure.
II. Component requirements
Confirm sizing as stated in the following and check data sheet and specific
requirements for each component.
367
B.P. 7.5
More Best Practices for Rotating Equipment
FIGURE 7.5.8 Auxiliary system component sizing audit form.
368
Chapter | 7
Auxiliary Systems
A. Pump selection
1. Pump type
2. Make
3. Model
4. Speeda
5. Disch. Press@ 10 cSt [60 SSU (rated)]
6. Disch. Press@ Max. cSt (SSU)
7. Rated flow @ 10 cSt (60 SSU)
8. Flow @ Max. SSU
9. Flow @ relief valve press (positive
displacement pump only)
10. Rated kW (BHP)
11. Max. kW [BHP (@R.V. and max. viscosity)]
12. NPSH available—m (ft.)
13. NPSH required—m (ft.)
14. Suction lift [if pumps are mounted above
fluid level—m (ft.)]
Positive displacement
Main/aux
Centrifugal
main/aux
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
a
If steam turbine driver is used, it is recommended that speed should be 2 pole (3600/3000 RPM)
motor speed to minimize steam rate.
B. Coupling selection
1. Pump
2. Coupling model
3. Size
4. Driver max. power—kW (HP)
5. kW (HP)/100 RPM
6. Coupling kW (HP)/100 RPM
Main
Auxiliary
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
Note: Confirm appropriate coupling service factor is used. Rotary (P.D.) pumps require a higher service factor.
C. Driver selection
1. Service
2. Type
3. Speed
4. Pump max. power kW (HP)
5. Driver-rated power kW (HP) (= 1.1 × pump
max. kW or HP)
6. Driver normal power [@ 10 cSt (60 SSU)]
7. Turbine steam rate kg/kW-h [lb/HP-h (max./rated)]
8. Steam quantity @ minimum steam energy
condition kg/h (LB/h)
9. Driver starting timea
10. RPM-Rated Speed
Main
Auxiliary
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________
________________ ________________
________________ ________________
________________ ________________
________________ ________________
________________ ________________
Note: Confirm sufficient steam is available at minimum energy conditions.
a
Calculated for minimum energy conditions (minimum steam energy or motor minimum starting
voltage) and pump rated conditions. If greater than 3 s, accumulator(s) should be used.
369
B.P. 7.5
More Best Practices for Rotating Equipment
D. Relief valve selection (Positive displacement pumps only)
1. Pump max. discharge pressure at max. viscosity =
2. Relief valve pressure = 1.1 × D1 or 172 kPa
(25 PSI). Greater, whichever is higher.
Relief valve type (modulating preferred)
Model
Set pressure
Overpressure (pressure to pass full flow)—kPa (PSI)
Normal leakage (valve closed)
__________________
__________________
= ________________
= ________________
= ________________
= ________________
= ________________
E. Reservoir sizing (based on rectangular tank) Per API 614
1. Normal flow L/min (GPM)
2. Retention time (min)
2A. Capacity = (1) × (2) (Refer to Fig. 7.5.9)
Confirm size
3. Reservoir length × width mm (in.)
= ________________
= ________________
__________________
4. Capacity cm (in.) of height
(3)
231in.3 /gal
= _____ L/cm (gal/in.)
= _____ cm (in.) above grade
= (5) + level required to
= maintain prime
= ________________
= ________________
=
5. Level E (pump suction level)
6. Level D (suction loss level)
7. Level C = (6) + 5 (min) × (1)
Note: 5 min = Working time
Working capacity (volume between levels C&D) or level C =
8(min) × (1)
+ Tank bottom height above grade
(4)
Note: 8 min = retention time
Retention capacity = volume between bottom of tank and level C.
8. Level B
9. Level A
= Highest level of oil during operation
(approximately 1 min retention time)
= Highest level oil can reach
= Level B + capacity contained in all components
that drain back to the reservoir/(4)
Note: This quantity should also include allowance for interconnecting piping and any overhead tanks.
10. Minimum reservoir free surface area:
= 232 cm 2 /LPM of normal flow (0.25 ft.2 /GPM)
= 0.25 × (1)
= ft.2
Confirm reservoir internals, material, etc. meet data sheet and specifications
required. Review reservoir internal drawing.
370
Chapter | 7
Auxiliary Systems
F. Reservoir heating requirements
Type
Electric
Steam
Time to head oil from _______
1. Calculated heat load (minus reservoir
heat loss)
2. Heater size kJ/h (BTU/h)
°C (°F) to _______
= _______ kJs (BTUs)
°C (°F) = _______ h
3. Electric heater max. watt density
= _______
=
(F1)
Total time allowed (h)
Note: Confirm if heaters can be removed without draining reservoir.
G. Supply pipe velocity
Maximum velocity 1.2–1.8 m/s (4–6 ft./s)
Maximum console supply pipe velocity = _________ m/s (ft./s)
Maximum unit supply pipe velocity = _________ m/s (ft./s)
H. Control valve sizing
H1. Bypass (back pressure) valve
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
1.10
1.11
Type: self acting, pneumatic, or electric controller
Make
Model
Action—Direct or Reverse
Valve plug type
Failure mode
Actuator size
Actuator force available/force required
Maximum valve Cv
Operating Cv min. (one pump dirty system)
Operating Cv max. (two pumps clean system)
=
=
=
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
Note: Operating Cvs should be between 10% and 90% of valve max. Cv.
Sensing line pulsation snubber required? If so, confirm proper orientation.
Confirm fast response is to open or close valve.
H2. Transfer valve (s)
2.1
2.2
2.3
2.4
2.5
2.6
2.7
Make
Model
Size
Plus type—Taper, Straight, Globe
Lifting jack required?
Tight shut-off required?
Max. ∆P on changeover
_______________
_______________
_______________
_______________
_______________
_______________
_______________
371
B.P. 7.5
More Best Practices for Rotating Equipment
H3. Temperature control valve(s)
3.1 Make
3.2 Model
3.3 Size
3.4 Normal flow L/min (GPM)
3.5 Temperature range °C (°F)
3.6 Valve max. operating Cv (If two-way valve, Cv must be based on
clean cooler)
3.7 Valve maximum Cv
_______________
_______________
_______________
_______________
_______________
_______________
_______________
Note: Butterfly type valve often used for two-way applications.
H4. Pressure reducing valve
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
4.10
4.11
1.
2.
3.
4.
5.
6.
7.
8.
9.
Type: self acting, pneumatic, or electric
Make
Model
Action-direct or reverse
Valve plug type
Failure mode
Actuator size
Actuator force
Maximum valve Cv =
Normal valve operating Cv = (Unit at operating speed)
Minimum valve operating Cv = (Unit at rest—oil system on)
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
Type
Shell and Tube
Air (fin fan)
Twin or single
Make
Model
Size
Heat load kJ/HP (BTU/h)
Oil side ∆P clean kPa (PSI)
Fouling factor (total)
Oil flow LPM (GPM)
Water quantity LPM (GPM)
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
_______________
I. Filter sizing
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
372
Make
Model
Type—Surface Depth
Normal flow LPM (GPM)
Max. flow LPM (GPM)
Filtration (micrometers)
Clear filter ∆P max.—kPa (PSI)
Cartridge material
Type end seals
Cartridge-single or multiple
Cartridge center tube material
∆P at max. viscosity—kPa (PSI)
Collapse pressure—kPa (PSI)
Number of cartridges
LPM (GPM) per cartridge
______________
______________
______________
______________
______________
______________
______________
______________
______________
______________
______________
______________
______________
______________
______________
Auxiliary Systems
Chapter | 7
J. Switches or transmitters
Confirm proper range, type, materials, and maximum deadband (change in
actuation point) of each switch. Confirm proper selection of transmitters.
K. Gauges
Confirm proper range, type, material of each pressure, differential pressure,
temperature, and level gauges.
L. Accumulator sizing
1. Type: Bladder ________ or direct acting ________
2. System flow KPM (GPM) =________
3. System transient time (s) =________
(2) × (3)
4. Capacity of fluid required =
=________ L (Gallons)
60
5. System pressure below which accumulator begins to drain kPa (PSIA)
=________
6. Precharge pressure kPa (PSIA) =________
7. Proposed accumulator internal volume
(approximately 90% of normal size) =________
8. Actual fluid capacity per accumulator =________
  (6)  
(7) × 1−    =________
  (5)  
9. Precharge type: manual, self contained, automatic
M. Additional tank sizing and construction confirmation
Overhead rundown (lube)
Overhead (seal)
Degassing tank(s)
These tanks should be checked against specifications data sheets for proper
capacity, construction, and ancillaries.
N. Piping, vessel, flange, and component material
Confirmation
Confirm that all specified materials are supplied.
O. Console and unit connection orientation
Refer to Fig. 7.5.9 and finalize all connection locations (Fig. 7.5.10).
373
B.P. 7.5
More Best Practices for Rotating Equipment
FIGURE 7.5.9 Reservoir levels and oil level glass details.
FIGURE 7.5.10 Connection orientation drawing. (Courtesy of Elliott Co.)
374
Auxiliary Systems
Chapter | 7
B.P. 7.6: Install high point vents on direct acting valves
Direct Acting (Hydraulically Actuated) valves have been in use in oil systems for decades and have been proven to be very fast reacting, especially when
employed for the backpressure control valve. They are much faster reacting
than pneumatic valves, simply due to the fact that oil is incompressible and will
provide immediate actuation on the valve diaphragm.
While this is absolutely the case, there needs to be a provision to assure the
sensing lines are air free.
L.L. 7.6: Failure to vent direct acting control valve sensing lines has resulted in delayed response of control valves and unit trips
Throughout the life of an oil system, there will definitely be air bubbles that
will make their way through the system and many times will get trapped in the
sensing lines for the valves. This is especially important for the backpressure control valve and can delay the closing time of the valve in the event the auxiliary
pump auto start has been engaged. This has resulted in numerous trips of oil systems globally.
BENCHMARKS
This best practice has been used since the introduction of direct actuated control
valves in oil systems and has been an aid in maintaining total centrifugal compressor train reliability of greater than 99.7%.
SUPPORTING MATERIAL
See Fig. 7.6.1 which details the industry best practice for venting direct acting
control valves. Note that an anti-pulsation valve (Pneu-trol) is depicted in this
Fig. 7.6.1 which assures smooth valve operation.
375
B.P. 7.7
More Best Practices for Rotating Equipment
FIGURE 7.6.1 Control valve venting guidelines.
B.P. 7.7: Do not install time delays in oil system trip circuits
The oil console should have designed properly such that it reacts quickly
during transient conditions and the unit does not trip. Many times in the field,
after a unit trips due to low oil pressure, a user may initiate a time delay for a
certain instance so that the unit will not trip.
While this will not initially cause a trip of the unit, worse problems will most
likely arise since the effect is being taken care of and not the actual cause. The
reason the system will trip is because oil is not making its way to the components. Remember, a bearing installed on a compressor operating at 6000 RPM
376
Auxiliary Systems
Chapter | 7
will rotate 100 times/s!!! If we don’t have oil for even half a second, the bearing
just rotated 50 times. With that said the trip is there for a reason and there is
another reason why the trip is being initiated.
If a trip has been initiated, a full component condition check should be conducted to determine what is causing the oil system failure.
L.L. 7.7: Installing time delays on trip circuits do not go after the cause of
failure but put a “Band Aid” on it and will delay the time to troubleshoot
the actual problem, resulting in profit loss for the facility
BENCHMARKS
This best practice has been used since the mid-1990s to optimize the reliability
of oil systems and to achieve compressor train reliabilities exceeding 99.7%
SUPPORTING MATERIAL
See the supporting material in BP 7.4 for details on how all the components
within an oil system function.
See the following table for a general list of items to monitor when there is
an issue in a lube oil system.
Component/item
Specified value
Actual value
Comments
Oil Reservoir
Level
Oil Temp.(°C)
Air in Oil? (Y/N)
Gas in Oil?
Oil Sample?
Pumps
Aux. Pump Operating?
P2 (bar)
Suction Noise?
Suction Filter ∆P (bard)
Vibration (µm)
Brg. Bracket Temp. (°C)
Couplings
Noise?
Strobe Findings
Turbine Driver
Operating Speed (RPM)
Trip Speed Setpoint (RPM)
Vibration (µm)
Brg. Bracket Temp. (°C)
Gov. Hunting?
377
B.P. 7.7
More Best Practices for Rotating Equipment
Component/item
Trip Lever Condition
Gov. Oil Condition
Motor Driver
Operating?
Vibration (µm)
Brg. Bracket Temp. (°C)
Axial Shaft Movement (µm)
Fan Noise?
Relief Valves
Passing?
Set Pressure (bar)
Pump P2 Press. (bar)
Check Valves
Aux. Pump Turning Backwards?
Noise?
Back Pressure Valve
% Open
Stable?
Valve Noise?
Set Pressure (kPa)
Maintained Pressure (kPa)
Transfer Valves
One Bank Operating?
Noise?
Coolers
∆T Oil
CW Valve Pos.
Cooler Operating?
Vent Valves Open?
TCVs
% Open
Set Temp. (°C)
Stable?
Actual Temp. (°C)
Filters
∆P (bar)
Vent Valves Open?
Last Filter Change
Accumulators
Pre-charged Pressure (bar)
Last PM Date
Lube Oil PCV
% Open
Set Pressure (bar)
Actual Pressure (bar)
Stable?
Control Oil PCV
% Open
Set Pressure (bar)
Actual Pressure (bar)
Stable?
378
Specified value
Actual value
Comments
Auxiliary Systems
Component/item
Specified value
Actual value
Chapter | 7
Comments
Lube Oil Rundown Tank (or Emerg. Pump)
Pump or Tank?
Pump Operating?
Tank Overflow
Lube Oil Supply Lines
Leaks?
Noise?
Vibration (µm)
B.P. 7.8: Install a Differential Pressure gauge across seal oil drainers when
a balance line DP gauge or transmitter is not installed on the compressor
By doing this you will essentially be monitoring the balance line DP, since
the suction seal is sealing against suction pressure and the higher pressure seal
is sealing against the pressure behind the balance drum minus any losses across
the inner seal labyrinth, which is minimal. Therefore, an increase in the DP between the two seal drainers, indicates Balance Drum Labyrinth wear.
L.L. 7.8: Thrust bearing assemblies are frequently changed, without considering balance system differential pressure trends only to find that balance device deterioration is the root cause and compressor disassembly is
required forcing a 5–7 day loss of revenue
BENCHMARKS
This Best Practice has been in use since 1990 when it was found to be easier to
install these devices than a Balance Line DP transmitter or gauge on compressors that did not have them equipped prior. This action has resulted in success in
limiting balance device maintenance to turnarounds and compressor reliabilities
exceeding 99.7%.
SUPPORTING MATERIAL
Impeller Thrust Forces
Every reaction type compressor blade set or impeller produces an axial force
toward the suction of the blade or impeller. Refer to Fig. 7.8.1.
In this example, the net force toward the compressor suction is 8900 N
(2000 lbs) for the set of conditions noted. Note that the pressure behind the impeller is essentially constant 344.75 kPa (50 psi), but the pressure on the front
side of impeller varies from 344.75 to 275.8 kPa (50–40 psi) because of the
pressure drop across the eye labyrinth. Every impeller in a multistage compressor will produce a specific value of axial force toward its suction at a specific
flow rate, speed and gas composition. A change in any or all of these parameters
will produce a corresponding change in impeller thrust.
379
B.P. 7.8
More Best Practices for Rotating Equipment
FIGURE 7.8.1 Impeller thrust force.
Rotor Thrust Balance
Fig. 7.8.2 shows how a balance drum or opposed impeller design reduces thrust
force. The total impeller force is the sum of the forces from the individual impellers. If the suction side of the impellers is opposed, as noted in Fig. 7.8.2, the thrust
force will be significantly reduced and can approach zero. If the suction side of
all impellers are the same (in series), the total impeller thrust force can be very
high and may exceed the thrust bearing rating. If this is the case, a balance drum
must be mounted on the rotor as shown in Fig. 7.8.2. The balance drum face area
is varied such that the opposing force generated by the balance drum reduces the
thrust bearing load to an acceptable value. The opposing thrust force results from
the differential between compressor discharge pressure (PF) and compressor suction pressure (P1) since the area behind the balance drum is usually referenced to
the suction of the compressor. This is accomplished by a pipe that connects this
chamber to the compressor suction. This line is typically called the “balance line.”
It is very important to note that a balance drum is used only where the thrust
bearing does not have sufficient capacity to absorb the total compressor axial load.
And the effectiveness of the balance drum depends directly on the balance drum
seal. Fail the seal, (open clearance significantly) and thrust bearing failure can result.
A common misunderstanding associated with balance drum systems is that
a balance drum always reduces the rotor thrust to zero. Refer to Fig. 7.8.3 and
observe that this statement may or may not be true depending on the thrust
balance system design. And even if it is, the thrust is zero only at one set of
operating conditions.
380
Auxiliary Systems
Chapter | 7
FIGURE 7.8.2 Rotor thrust force.
FIGURE 7.8.3 Rotor system designed four different ways.
381
B.P. 7.9
More Best Practices for Rotating Equipment
Fig. 7.8.3 shows a rotor system designed in four different ways. Note how
the thrust always changes with the flow rate regardless of the design. Another
misconception regarding thrust balance systems is the normal or “active” direction of thrust. In many cases, the active thrust is assumed to always be toward
the suction of the compressor.
Observing Fig. 7.8.3, it is obvious that the “active” direction can change
when the turbo-compressor has a balance drum or is an opposed design. It is
recommended that the use of active thrust be avoided where possible and that
axial displacement monitors be labeled to allow determination of the thrust direction at all times.
B.P. 7.9: Always mark oil system control valves after a turnaround to give
a baseline condition and determine wear throughout a run (from turnaround to next scheduled shutdown)
Control valves are basically a flow meter when they are in a fixed position
(fixed orifice). Therefore, it is very critical to know the position of them at all
times. Especially after a turnaround, if you just mark the stem and yoke of the
control valve, you then have a baseline position of the control valve. Any change
in that position indicates a change in flow, which will in turn indicate wither
wear or a component failure within the system.
L.L. 7.9: Inability to know control valve position from beginning to end of
a run (from turnaround to next scheduled shutdown) has resulted in delays
in troubleshooting the root cause of oil system failure and loss of production
BENCHMARKS
This best practice has been used since 1990 to produce oil systems of highest
reliability which has resulted in unit reliabilities above 99.7%.
SUPPORTING MATERIAL
Referring to the general definition of an auxiliary system which is to “continuously supply cool, clean fluid to each specified point at the required pressure, temperature, and flow rate,” we can see that the controls and instruments play a major
role in the reliability of auxiliary systems. The function of the controls and instrumentation is to continuously supply fluid to each specified point at the required
pressure, temperature, and flow rate. While it is true that pumps and coolers must
be present, system controls modify the operational characteristics of these components to achieve the desired results. In addition, system instrumentation initiates transient system response, continuously monitors operation and shuts down
critical equipment in the event of an auxiliary system malfunction. In this section,
we will examine important concepts that are at the heart of auxiliary system reliability, define the function of major control and instrumentation components, and
discuss items that can significantly reduce auxiliary system reliability.
382
Auxiliary Systems
Chapter | 7
TABLE 7.9.1 Major Auxiliary System Controls and Instrumentation (By Function)
Controls
Instrumentation monitor and alarm
• Positive displacement pump system
flow control
• Dynamic pump system flow control
• Stand-by pump automatic start
• Cooler temperature control
• System differential supply pressure
control (constant reference pressure)
• System differential supply pressure
control (variable reference pressure)
• System reservoir level
•
•
•
•
Pump operation
System pressure
System temperature
Filter differential pressure
• System differential pressure (variable
reference pressure)
• Variable speed pump driver speed indicator
Types
Types of major auxiliary system controls and instrumentation are outlined in
Table 7.9.1. Note that types are defined by function. As an example, a positive
displacement pump system flow control consists of a pressure control valve that
bypasses excess flow from the pump back to the system reservoir to maintain a
set system pressure. The function of this component however is to continuously
supply the required flow of fluid to the system under varying system pressure
drops and critical equipment component conditions (worn bearing, seal, etc.).
All system controls and instrumentation must function perfectly under both
steady state and transient conditions. Under normal operation, a steady state
control mode is approached since flows, pressures, and temperatures change
very slowly if at all. While this mode of operation may appear to be ideal, it can
be dangerous since control valves and instrumentation can bind up due to debris
and lack of movement. In the transient mode, components must have response
times on the order of milliseconds. When one considers the function of an auxiliary system and the fact that the slowest of critical equipment units operate at
approximately 60 revolutions per second (3600 RPM), the necessity of rapid
system response time is appreciated. If the controls cannot respond to a transient response, the instrumentation and the critical equipment shutdown system
(circuit breaker, steam turbine trip valve system, etc.) must operate on demand
to stop equipment operation. If the system controls and instrumentation do
not have sufficient response times, a system liquid supply source (accumulator)
is required to provide flow during transient conditions. Using our system as an
example, 60 GPM are supplied to the unit or 1 gallon per second. Suppose the
main pump trips and the normal flow to the equipment is not reached for 3 s
(until the stand-by pump is at full speed and flow rate). An accumulator with
a liquid capacity of 3 gallons would enable the system to function normally
during the upset since it would supply the required flow of 1 gallon per second.
383
B.P. 7.9
More Best Practices for Rotating Equipment
Note an accumulator size greater than 3 gallons would be required. This will be
covered separately.
Concepts
The use of concepts can be helpful in understanding the function of auxiliary
system components and systems. In this section we will discuss:
l
l
l
l
l
An equivalent orifice
Sub-systems
An equivalent vessel
Control valve liquid coefficient—Cv
A flow meter in every system
The Concept of an Equivalent Orifice
Bearings, seals, etc. can be reduced to the concept of an equivalent orifice
(Fig. 7.9.1). The equation for orifice flow is:
Q = C × C f × D2
∆p
S.G.
From the previous equation it can be seen that flow to any component is
the function of the dimension “D2” and ∆P across that component. The system
components essentially experience two types of flow changes. The gradual flow
change due to component wear (i.e., D2 change as in the case of bearing wear) or
FIGURE 7.9.1 Reduce it to an equivalent orifice.
384
Auxiliary Systems
Chapter | 7
the sudden flow change due to a pressure change in the system. As can be seen
from the previous equation, a sudden change of pressure as in the case of a hunting control valve or a sudden pressure spike due to component starting or stopping will cause a corresponding sudden change in flow rate to the component.
Considering the speeds involved in critical equipment, one can appreciate that a
short-term transient flow change can lead to significant component damage of the
critical equipment (bearing, seals, etc.) The previous concept of reducing each individual critical equipment component (bearing, seals, orifices, etc.) to an equivalent orifice helps enormously in conceptualizing transient system reactions.
Sub-Systems
Both positive displacement and dynamic pumps alone do not contain the desired
characteristics for operation within an auxiliary system. To achieve the objectives of an auxiliary system, these components must be combined in a controlled
subsystem to achieve desired results. The sub-system is the combination of the
pump and a control valve which together produce the flow characteristic required.
Viewing components in control and instrumentation as being part of various subsystems also helps in understanding the total function of auxiliary systems.
Equivalent Vessel
Refer to Fig. 7.9.2. Systems and sub-systems can be reduced also to that of
equivalent vessels. As an example, the supply pipe from a lube oil console can
be reduced to an equivalent vessel as shown in Fig. 7.9.2.
FIGURE 7.9.2 Reduce it to an equivalent vessel.
385
B.P. 7.9
More Best Practices for Rotating Equipment
When supply flow equals exit flow, the pressure in any equivalent vessel remains constant. If supply flow is less than exit flow, the pressure reduces rapidly.
The function of an accumulator can be understood easily by using this equivalent vessel concept. If a vessel is installed downstream of the equivalent vessel
in Fig. 7.9.2, during the period of reduced inlet flow the vessel would supply
flow to the system. This is exactly the function of an accumulator.
Another example of using the equivalent vessel concept is as follows: Imagine again the equivalent vessel is a supply pipe from a lube oil console. Suppose
the main pump trips on overload and the auxiliary pump does not start immediately. Since the auxiliary pump did not immediately start, the supply flow to
the equivalent vessel is less than the exit flow. As a result, the pressure in the
equivalent vessel will drop. This is why pressure switches in auxiliary systems
are used as alarm, auxiliary pump start, or trip devices. Using our concept of an
equivalent vessel it can be seen that the pressure switch actually acts as a flow
indicator and will activate on low flow even though it is measuring pressure.
Control Valve Liquid Sizing Coefficient—Cv
“Cv” is an important concept that must be understood when dealing with any
type of control valve on liquid service. Cv “valve sizing coefficient” is defined
by the following equation:
Cv = Q (GPM)
S.G.
∆P
where: S.G. (specific gravity) = 0.85 (for oil), ∆P = value pressure drop (P.S.I.).
Solving this equation for GPM we see that:
Q (GPM) =
Cv
S.G.
∆P
We can see referring back to “The concept of an equivalent orifice” that this
equation is similar to that of an orifice. Naturally the only difference is that a valve
is a variable orifice. Valves are sized using this concept of Cv (valve coefficient).
Each valve has a maximum Cv. Depending on the type of internal valve design,
seats, plugs, and body, a valve will exhibit a certain characteristic. Refer to Fig. 7.9.3
which is a graph of valve characteristics. Plotted on the ordinate (Y axis) is valve
flow in percent of maximum flow and plotted on the abscissa (X axis) is travel of the
valve plug in percent of rated travel. As we cover specific valve application later in
this section, the characteristics of particular valves will be discussed. Referring back
to the relationship for valve coefficient, it can be seen that the valve coefficient is dependent on flow rate, differential pressure across the valve, and fluid characteristic.
As an example, suppose that a valve is sized to pass 20 GPM under normal
conditions of 150 PSI pressure drop. The fluid in this case is light turbine oil at
386
Chapter | 7
Auxiliary Systems
FIGURE 7.9.3 Control valve flow characteristics. (Courtesy of Fisher Controls Inc)
150°F (60 SSU). Solving for the valve Cv per the previous equation, we arrive
at a value of 1.51. If the valve pressure drop were to decrease to 100 lbs, and
we still required 20 GPM to pass the valve coefficient would be 1.84. This change
represents approximately a 22% change in the valve coefficient. Depending
on the characteristic curve of the valve in question, it would represent a given
amount of valve plug opening (increase of travel). In the same example, now let
us assume that the flow changes to 40 GPM with 100 lbs pressure drop across
the valve. The Cv now would be 3.69 or approximately 200% the previous value. Depending on the valve size, this coefficient may or may not be obtainable.
Refer to Table 7.9.2 which is a typical valve coefficient table showing valve
TABLE 7.9.2 Typical Liquid Valve Sizing Coefficient Table
Valve Cv’s for different sizes and valve travel
% Travel
12.5
25
50
75
100
Valve Travel (in.)
Body size (in.)
Port size (in.)
1/32
1/16
1/8
3/16
1/4
1
3/4
1.4
3.1
4.2
5.3
7
1
1
2.4
4.2
7
10
12
387
B.P. 7.9
More Best Practices for Rotating Equipment
coefficients for % travel of a particular valve. When sizing all control valves, Cv
maximum, Cv normal, and Cv minimum must be calculated. A general rule is
that all of the aforementioned values should fall between 10% and 90% of the
maximum Cv for a particular valve selected.
When dealing with viscous liquids as in the case of oil, valve coefficient viscosity corrections must be made. For the example case mentioned earlier, the
correction factor for 220 cSt (1000 SSU viscosity) would be approximately
1.5–2. Therefore the valve coefficient required would be 1.5–2 times that required
at normal viscosities (60 SSU for light turbine oil at normal operating temperatures). Viscosity correction nomographs are available from control valve manufacturers for determining valve sizes required under high viscosity conditions.
A Flow Meter in Every System
Considering the relationship discussed earlier it can be seen that every control
valve can be considered as a flow meter if the fluid differential pressure across
the valve, valve travel, and a valve characteristic chart is known. While not a
completely accurate flow measuring device, this concept can be extremely valuable while troubleshooting auxiliary systems. Obtaining the valve travel and
using the valve coefficient chart, the Cv can be obtained. Calculating for GPM
knowing the Cv, the pressure drop across the valve and the specific gravity of
the liquid can then yield the flow rate. It is important to note that with small
valve travels on the order of 1/4 in. maximum, an accurate means of measuring
valve travel must be obtained. It is the writers experience that many times travel
indicators are not furnished with the valve. It is strongly recommended that
valve travel indicators be supplied or retrofitted in the field.
Bypass Control
The first application to be discussed in this section will be that of a bypass
control valve. A bypass control valve and actuator pictured in Fig. 7.9.4 is used
with a positive displacement pump to alter the pump’s flow characteristic to that
of variable flow.
Refer to the schematic of a lube oil system typical of the example in
Fig. 7.9.5. This system incorporates positive displacement pumps. The control
valve’s function is to continuously control flow to the critical equipment such
that the required flow is supplied under normal and transient conditions. Since
a positive displacement pump essentially is a constant flow device, the control
valve in the bypass mode must allow for excess pump flow to be recirculated
back to the reservoir. Utilizing the concept of an equivalent orifice, as the bearings in the system wear the orifice diameter becomes larger, therefore the flow
required to the critical equipment will be greater. Since the downstream pressure
across the bearings is atmospheric pressure, the upstream pressure will initially
decrease when the bearing area becomes larger for the same flow. The bypass
valve will sense the upstream pressure reducing and will close to force the
388
Auxiliary Systems
Chapter | 7
FIGURE 7.9.4 Reverse acting actuator and valve body typically used as a back pressure
regulator (bypass control). (Courtesy of Fisher Controls Inc)
389
B.P. 7.9
More Best Practices for Rotating Equipment
FIGURE 7.9.5 Typical lube oil supply system.
additional required flow to the critical equipment. Even though the bypass valve
is a pressure device, it’s acting as a flow control device to divert bypass flow
to required system components. Therefore using the concept of a sub-system,
the bypass valve and the positive displacement pump form a variable flow subsystem that will supply variable flow to the critical equipment on demand.
In addition to accounting for small changes in system flow requirements,
the bypass control valve must also act under transient conditions. If the main
pump were to suddenly shut off, the system would immediately sense a pressure decrease. Referring to the equivalent orifice concept of bearings, the flow to
these components would drop proportionately to the square root of the pressure
drop across the component. At hundreds of revolutions per second, the bearing
390
Auxiliary Systems
Chapter | 7
shaft interface would not last long with the absence of system flow. In this transient mode, the control valve must close quickly to divert all bypass flow to the
system to account for the absence of flow from the pump. The control valve
characteristic, its actuator and supply to its actuator whether direct (hydraulic)
or indirect (pneumatic) must function instantaneously. If the valve system experiences instabilities or excessive friction, as in the case of valve stem binding,
the system will experience an instantaneous loss of flow and will (hopefully) be
shut down on this signal. Again referring to concepts discussed previously, the
concept of an equivalent vessel is useful in ascertaining how pressure and flow
are related and why pressure switches are used to determine loss of flow under
transient conditions. This concept also shows why time delays in auxiliary systems are not desired to be used with any trip devices. It’s true that a time delay
would preclude a trip of the unit under transient conditions but could also cause
severe and perhaps catastrophic damage to the critical equipment.
The bypass control valve also must exhibit rapid transient response in the
open direction. In the case of dual pump simultaneous operation, the amount
of flow to be recirculated to the reservoir will be equal to the normal bypass
flow of one pump plus the full flow of the stand-by pump. If the bypass valve
does not act as a variable orifice and opens at a slower rate than the flow
rate increase, referring to the orifice equation, the pressure drop across the valve
will simultaneously rise. This increase may exceed the setting of the relief valve in
the system. If this is the case, the system is exposed to the potential of the relief
valve not re-seating. If this were to occur, a new “orifice” would be introduced
into the system and the flow to the critical equipment would be reduced to the
point of requiring the stand-by pump to start and possibly causing critical equipment shutdown. In order to meet the above control and transient requirements,
the bypass control valve must be sized properly. An example of valve sizing
using the system shown in Fig. 7.9.5 is shown in Table 7.9.3. We wish to reemphasize that once the valve is sized properly, the actuator and the sensing
lines in the system that supply the force to operate the valve must be designed
for rapid response. In many systems, sensing line snubbers are used to dampen
impulse signals that can lead to valve instability. It must be noted that snubbers
are designed to provide quick response in one direction and retarded or slower
response in another. It is of extreme importance that these devices be installed
properly. Understanding the function of the particular valve in question and
examining the direction of the snubber device in a sensing line is essential to
correct system operation. Many times these snubber devices are installed improperly in the wrong direction.
Pressure Reducing Control
Pressure reducing control has two primary applications in auxiliary systems.
l
l
To control the flow from a dynamic pump.
To reduce the pressure in the system.
391
B.P. 7.9
More Best Practices for Rotating Equipment
TABLE 7.9.3 Valve Sizing Example—Back Pressure (Bypass) Control
Given:
1. Normal valve flow
= pump flow − normal system requirement
= 73 GPM − 60 GPM
= 13 GPM
2. Maximum valve flow
= main and auxiliary pump flow − minimum
system requirement
= 146 GPM − 60 GPM
= 86 GPM
3. Maximum valve P
= pump discharge pressure @ maximum supply
flow and component P
= 250 Psig
4. Minimum valve P
= pump discharge pressure @ minimum supply
flow and component P (clean system)
= 160 Psig
5. Oil specific gravity
= 0.85
Determine:
Cv Minimum
Cv Maximum
1. Cv Min.
= Q NORMAL
S.G.
∆P Max.
= 13 × 0.0583
= 0.758
2. Cv Max.
= Q MAX.
S.G.
∆P Min.
= 86 × 0.0729
= 6.268
Refer to Table 7.9.2 for 1-in. valve with 3/4 in. port and obtain:
Valve maximum
Cv = 7.0
Valve operating maximum
Cv = 6.268
Valve operating minimum
Cv = 0.758
Valve maximum travel (opening)
= 90%
Valve minimum travel (opening)
= 9%
Note: Valve minimum and maximum openings are at the limit for satisfactory operation.
A typical pressure reducing control valve and actuator are shown in Fig. 7.9.6.
For the first case, the flow characteristic is variable. The flow is therefore determined by the pressure at the discharge flange of a dynamic pump. A pressure reducing valve set to sense the pressure downstream of the valve will automatically regulate the discharge or the back pressure on the dynamic pump for the desired flow
392
Auxiliary Systems
Chapter | 7
FIGURE 7.9.6 Direct acting actuator and valve body used for PRV (pressure reducing)
control. (Courtesy of Fisher Controls Inc)
393
B.P. 7.9
More Best Practices for Rotating Equipment
of the system. Referring back to the equivalent orifice concept, if a bearing were
to wear, the equivalent diameter of the orifice would increase. Therefore, initially
for the same flow rate, the pressure in the system would decrease since the flow is
the same. If the bearing clearance increases (equivalent D), the ∆P must decrease.
Therefore the pressure control valve sensing decreasing system pressure
will open to increase the system pressure. This action will result in a decrease
of resistance on the dynamic pump discharge flange and allow the centrifugal
pump to operate at a greater capacity to provide the desired flow to the critical
equipment. It can be seen that in this case the pressure reducing valve and the
dynamic pump combine for a sub-system that meets the objective of providing
continuous flow to the critical equipment. The control valve essentially renders
the variable flow, constant head device a variable head device by compensating
for changes in system pressure. The aforementioned case represents the normal
control case. Lets now examine the transient case.
If the main dynamic pump were to suddenly trip, the system pressure will
suddenly fall as a result of greater flow exiting the system than the amount of
flow entering the system (equivalent vessel concept). In this case, the pressure
reducing valve sensing downstream pressure would instantaneously open allowing the dynamic pump to move out to a higher flow point on its curve while
the auxiliary or stand-by pump were to start. As soon as the auxiliary pump
starts the pressure reducing valve sensing additional flow into a fixed system
resistance would then close meeting the flow requirements. Dynamic systems
in general tend to be somewhat softer than positive displacement systems. That
is, they are more tolerable to transient system changes.
In the case of the auxiliary stand-by pump and the main pump operating
simultaneously, the pressure reducing valve would automatically compensate
for the increased flow by reducing its travel or increasing the system resistance at
the discharge flange of both pumps. That is, increasing the discharge pressure
to the level where the combined flow of both pumps would exactly equal the
critical equipment system required flow. Again referring to the concept of an
equivalent orifice, if excessive flow were forced through the orifice (the bearings) the pressure drop would increase. The pressure reducing valve sensing the
increased system pressure drop would tend to close to reduce the pressure at its
sense point. In doing so it will increase the discharge pressure on both dynamic
pumps and since their characteristic is reduced flow on increased pressure the
desired flow will be obtained. Therefore it can be seen again that the dynamic
pump pressure reducing valve sub-system has the function of flow control to the
critical equipment even though it is sensing pressure.
The other primary application for pressure reducing valves in auxiliary systems
is for reducing system pressure to other desired pressure levels. Refer to Fig. 7.9.5
and observe the pressure reducing valve at the discharge of the lube oil system.
Its function is to reduce pressure from control oil pressure to lube oil pressure.
Control oil pressure is controlled by the equivalent orifice in the control system
and the set point of the bypass control valve as shown in Fig. 7.9.5. The bypass
394
Auxiliary Systems
Chapter | 7
control valve senses pressure and controls flow to satisfy the requirement of the
equivalent orifice in the control system and the equivalent orifices in the lube system. The pressure reducing valve simply senses pressure downstream of the valve
and controls it to the preset value. It should be noted that in most auxiliary systems, the console (reservoir, pumps, etc.) is usually below the level of the critical
equipment, therefore the set point of any pressure reducing valves on the console
should compensate for the height or head difference between the console and the
critical equipment. Control valves used in pressure reducing service usually are not
exposed to system transient changes as in the case of bypass valves. Therefore their
sizing is relatively easy and their valve Cvs do not significantly change. A sizing
example for a direct acting pressure reducing valve is shown in Table 7.9.4.
TABLE 7.9.4 Valve Sizing Example—Pressure Reducing Control
Given:
1. Minimum and normal lube oil flow to unit
= 60 GPM
2. Maximum lube oil flow to unit
(Bypass valve failed closed)
= 73 GPM
3. Valve ∆P
= 120 PSIG − 25 PSIG
= 95 PSI (15 PSIG supply + 20 PSIG)
pressure drop for elevation
Note: This example is for a PRV located on the lube oil console at grade.
4. Oil specific gravity
= 0.85
Determine:
Cv Normal
Cv Maximum
1. Cv Normal
=Q ×
S.G.
Normal
∆P
= 60 × 0.0946
= 5.675
2. Cv Maximum
=Q ×
S.G.
Maximum
∆P
=73 × 0.0946
= 6.906
Refer to Table 7.9.2 for a 1-.inch valve with a 1-.inch port and obtain valve maximum
Cv = 12.0
Valve operating normal
Cv = 5.675
Valve operating maximum
Cv = 6.906
Valve normal travel (opening)
= 40%
Valve maximum travel (opening)
= 50%
395
B.P. 7.9
More Best Practices for Rotating Equipment
Temperature Control Valves
Temperature control valves are usually required in auxiliary systems to regulate the supply temperature to the critical equipment components. Especially in
systems where liquids have viscosity characteristics (oil systems), temperature
control is important to insure correct oil viscosity to components. Referring to
concepts previously discussed in this section, the temperature control valve plus
the system coolers make up a cooling sub-system whose function is to continuously supply the required fluid to critical equipment at a specified temperature.
Two types of control valves that are used are direct acting three way valves and
air operated two way valves. Both valves sense the mixed temperature downstream of the cooler.
A two way valve is a simple bypass around the cooler while a three way
valve is a true mixing valve. It should be noted that when sizing a two way
valve, the pressure drop across the cooler must be known to assure that the valve
coefficient is large enough to pass the required flow. Many systems using two
way valves are insufficiently sized. This can result in cooler oil constantly being
supplied to the system since the pressure drop through the cooler is less than the
minimum pressure drop through the control valve.
Instrumentation
The instrumentation in any auxiliary system is extremely important in assuring
quick system response, accurate monitoring of system condition and rapid system shutdown in the event of upsets. In this section we will examine stand-by
pump start-up operation, critical equipment shutdown and monitoring functions
of the instrumentation in the auxiliary system.
Stand-By Pump Automatic Start
As previously mentioned, interruption of pump flow to the critical equipment
results in a rapid deterioration of system flow and pressure. Referring again
to the concept of an equivalent vessel, the absence of inlet flow to the system
while exit flow is continuing will instantaneously produce a pressure drop. This
concept is utilized in using a pressure switch to signal the immediate start of the
stand-by pump. Practice has shown that locating the pressure switch takeoff as
close as possible to the pump discharge results in the quickest response time to
initiate stand-by pump start. Some systems incorporate dual pressure switches,
one close to the pump discharge and another up close to the critical equipment.
Both switches start the stand-by pump on signal. The pressure setting of the
switch is usually set just below the lowest discharge pressure that the pump will
produce. In order to insure the rapid start of the auxiliary pump when required,
many systems incorporate an on-off-automatic switch on the auxiliary pump or
on the stand-by pump motor starter for testing the system. It is extremely important that the position of the switch always be in the automatic mode during
396
Auxiliary Systems
Chapter | 7
critical equipment operation. It is recommended that an alarm be supplied and
annunciated in the event that the auxiliary or stand-by pump is not in the auto
position during critical equipment operation.
Critical Equipment Trip Instrumentation
A general critical equipment design philosophy is to avoid trips circuits as much
as possible. That is, to only install trip switches in those situations that are absolutely necessary. Typically, auxiliary systems incorporate only one-trip function. As an example, the low lubricating oil trip in lube systems and the low
seal oil differentials trip in seal systems. Sometimes a high temperature switch
is also installed to trip, but this is not usually the case. The setting of the trip
switch is very important. It must be selected such that the equipment will shut
down when actuated in order to prevent any long term damage.
It must also be selected to prevent spurious, unnecessary trips of the unit
since they are extremely costly in loss of revenue. In addition, the quality of
any trip switch is extremely important since this relatively low cost device
could cost millions of dollars of lost product revenue per day in the event of a
malfunction. Attention is drawn to correct selection of switch component materials to prevent corrosion or any abnormality that would cause drifting of
switch setting and unnecessary unit shutdowns. Again the concept of a system
is extremely important to consider. It must be remembered that the trip switch
and the shutdown system for the critical equipment together must function accurately in order to terminate equipment operation immediately upon signal
from the initiating trip switch. “Best practice” is to use smart (self-diagnostic)
triple redundant (two out of three voting) transmitters for all pump start and trip
services.
Auxiliary System Monitoring
Refer to Table 7.9.1 and observe the different monitoring and alarm functions normally used in an auxiliary system. In order to insure reliable auxiliary system operation, the personnel must continuously observe and record any
changes in instrument readings and promptly attend to alarms to insure that the
system continues to operate as required. Changes in any of the system instruments indicates a change in the operating condition of the system and must be
followed through to insure that components are operating as required. As an
example, slowly deteriorating lube oil supply pressure could indicate either a
valve malfunction, reduction in speed of a main turbine pump driver, excessive pressure drop in the system oil filter or many other types of problems. It
is extremely important to maintain a program of auxiliary system instrumentation calibration to assure all instruments are reading properly. This will aid
greatly in determining malfunctions of the system and assist in the site preventive maintenance program.
397
B.P. 7.10
More Best Practices for Rotating Equipment
B.P. 7.10: Install a bypass with a valve and orifice around accumulator
isolation valve
This will aid in checking the accumulator pre-charge pressure, such that the
bypass can be opened to fill the accumulator back up without having to throttle
the isolation valve and risk tripping the unit on low oil pressure.
Note that the orifice should be sized properly so the bypass valve (usually
ball valve) can be opened quickly and the flow is not enough to trip the unit.
L.L. 7.10: Many unit trips have been traced back to opening the isolation
valve to the accumulator after regular maintenance too quickly
While checking accumulator pre-charge pressure is critical in assuring the
optimal oil system reliability, it is just as critical to be certain it is done safely.
Many units have tripped unexpectedly on low system pressure, causing loss of
revenue due to filling up the accumulator too quickly
BENCHMARKS
This best practice has been in use since the late 1990s to assure plant personnel
checked accumulator pre-charge pressure in a safe manner. Since that time, this
advice when implemented has resulted in plant oil system and unit reliabilities
above 99.7%.
SUPPORTING MATERIAL
An accumulator is simply a vessel which compensates for rapid short-term
flow disturbances in the auxiliary system. Most accumulators contain bladders
(Fig. 7.10.1). It is important to remember that transient disturbances are on the
order of micro seconds and usually less than 5 s in duration.
A schematic for a pre-charged accumulator is shown in Fig. 7.10.2.
The pre-charge pressure is set at the pressure that the volume of the accumulator flow is required in the system. (This value is usually around 60–70% of the
normal header pressure in which the accumulator is installed.) The quantity of
oil available from a pre-charged accumulator is extremely low.
As an example, a system with a flow capacity of 120 GPM has a motor
driven auxiliary pump that requires 3 s to attain full speed when started by
a pressure switch or transmitter at 140 PSIG. Normal header pressure equals
160 PSIG. Determine the amount of oil required to prevent the pump header
pressure from falling below 100 PSIG and the number of pre-charged 10-gallon
accumulators required (Table 7.10.1).
Many times an accumulator is improperly sized because of the misconception that its stated size is in fact the capacity contained therein. Actual capacity in any accumulator is equal to the internal volume minus the gas volume
over the liquid volume. Typically these values are 50% of the stated capacity
or less.
398
Auxiliary Systems
Chapter | 7
FIGURE 7.10.1 Typical oil system accumulator. (Courtesy of Greer)
System Reliability Considerations
Concerning auxiliary system control and instrumentation, a number of reliability considerations are worthy of mention.
Control Valve Instability
Control valve instability can be the result of many factors. To name a few;
improper valve sizing, improper valve actuators, air in hydraulic lines, or water in pneumatic lines. Control valve sensing lines should always be supplied
with bleeders to assure that liquid in pneumatic lines or air in hydraulic lines
is not present. Presence of these fluids will usually cause instability in the
system. Control valve hunting is usually a result of improper controller setting
on systems with pneumatic actuators. Attention is drawn to instruction books
399
B.P. 7.10
More Best Practices for Rotating Equipment
FIGURE 7.10.2 Accumulator pre-charging arrangement. (Courtesy of Elliott Co.)
to insure that proper settings are maintained. Frequently direct acting control
valves exhibit instabilities (hunting on transient system changes). If checks
for air prove inconclusive, it is recommended that a snubber device mentioned
previously be incorporated in the system to prevent instabilities. Some manufacturers install orifices which sufficiently dampen the system. If systems suddenly act up where problems previously did not exist, any snubber device or
orifice installed in the sensor line should be checked immediately for plugging.
Excessive Valve Stem Friction
Control valves should be stroked as frequently as possible to assure minimum
valve stem friction. Excessive valve stem friction can cause control valve instabilities or unit trips.
Control Valve Excessive Noise or Unit Trips
Squealing noises suddenly produced from control valves may indicate valve operation at low travel (Cv) conditions. Valves installed in bypass functions that
400
Auxiliary Systems
Chapter | 7
TABLE 7.10.1 Accumulator Sizing
Given:
• System required flow = 120 GPM
• System pressure at accumulator (at which accumulator effect is desired) = 140 PSIG
− 154.7 PSIA (P2)
• Gas precharge pressure (pressure at which accumulator oil flow ceases, assuming
system pressure does not fall below this level) = 110 PSIG = 124.7 PSIA (P1)
• Volume of accumulator = 9 gallons (Va) (accounts for volume of internal parts)
Determine:
• Amount of oil required
• Number of 10 gallon accumulators required
Amount of oil required:
120 gal/min
60 s/min
= 2 gal/s
• Oil required = 3 s × 2 gal/s
= 6 gallons
• Volume of oil entering system for each 10 gallon accumulator.
• System flow per second =
  P 
Voil = (Va ) 1−  1  
  P2  
  124.7  
= (9 gal) 1− 

  154.7  
= 1.75 gal. per accumulator
• Number of 10 gallon accumulators required
Oil quantity required
Quantity available per accum.
6 gal.
=
1.75 gal.
= 3.42 accumulators required
= four 10 gal. accumulators
Number of 10 gallon accumulators =
This is a large number of accumulators and is caused by:
The conservative setting of P2 and the neglect of the effect of system control valves and
partial auxiliary pump flow during pump acceleration. Let’s set P2 just (1 PSIG) below
the normal header setting and recalculate the number of accumulators required.


Voil = (9 gal) 1−  124.7  
  175.7  
= 2.6 gal/accumulator
= 3 accumulators required
The above example demonstrates the importance of properly sizing an accumulator.
exhibit this characteristic may be signaling excessive flow to the unit. Remember the concept of control valves being crude flow meters. Observation of valve
travel periodically during operation of the unit will indicate any significant flow
changes.
401
B.P. 7.10
More Best Practices for Rotating Equipment
Control Valve Sensing Lines
Frequently, plugged or closed control valve sensing lines can be a root cause
of auxiliary system problems. If a sensing line that is dead ended (Fig. 7.9.5) is
plugged or closed at its source, a bypass valve will not respond to system flow
changes and could cause a unit shutdown. Conversely, if a valve sensing line
has a bleed orifice back to the reservoir (to assure proper oil viscosity in low
temperature regions), plugging or closing the supply line will cause a bypass
valve to fully close rendering it inoperable and may force open the relief valve
in a positive displacement pump system.
Valve Actuator Failure Modes
Auxiliary system control valve failure modes should be designed to prevent critical equipment shutdown in case of actuator failure. Operators should observe
valve stem travel and pressure gauges to confirm valve actuator condition. In the
event of actuator failure, the control valve should be designed for isolation and
bypass while on line.
This design will permit valve or actuator change out without shutting down
the critical equipment. During control valve on line maintenance, an operator
should be constantly present to monitor and modulate the control valve manual
bypass as required.
Accumulator Considerations
Concerning accumulators, checks should be made when unit is shut down for accumulator bladder condition if supplied with bladders. One area which can cause
significant problems in auxiliary systems are accumulators supplied with a continuous charge. That is, charge lines (nitrogen or air) that come directly from a plant
utility system. Any rupture of a diaphragm will provide a means for entrance of
charge gas directly into the lube system. Most plant utility lines contain pipe scale
that could easily plug systems and cause significant critical equipment damage.
In addition, the following reliability factors should be noted (refer to
Fig. 7.10.2):
l
l
l
l
Be sure to install a check valve upstream of the accumulators to assure all
accumulator oil is delivered to the desired components.
Accumulators should be checked periodically (monthly) for proper precharge and bladder condition by isolating and draining the accumulator. Note
that the accumulator pre-charge pressure cannot be determined while on line.
When refilling the accumulators, care must be taken not to suddenly open
the supply valve. Best practice is to install an orificed bypass valve to be
used for filling the accumulator.
Best practice is also to install two full size accumulators to assure that one
accumulator is always on line during monthly checks.
402
Chapter 8
Pump Mechanical Seals
B.P. 8.1: Use plant, company, and industry lessons learned to properly
select mechanical seal and flush system and document details on data
sheets in Pre-FEED stage.
It is most important to utilize the experience within your company and industry in order to specify exactly what type of seal and flush system should be
used and how it should be arranged as early as possible in the project.
Details to include should be at least the following:
l
l
l
l
l
l
Type of seal preferred (Single, Double, Pusher, Bellows, etc.)
Type of flush plan preferred
Tubing or piping size and length of run
Cooler type (if required)
Type of control (whether pressure, differential pressure, or flow for an external type flush)
Any other specifics taken from troubleshooting your bad actor seals in similar applications
L.L. 8.1: Failure to utilize previous plant, company and industry lessons
learned and include details on the mechanical seal data sheet will result in
lower than optimum seal MTBFs.
Many users have been provided with mechanical seal arrangements and
flush systems that are not ideal for specific applications. Whether it be the specific type of flush plan or the way the system is arranged, it has been seen over
and over again that seals have been bad actors from inception, which could
have been different if input from the plant was taken into account early on in
the project.
BENCHMARKS
The writer has used this best practice since 2009 when involved with selecting
critical pumps on projects to assure maximum mechanical seal MTBFs (Greater
than 100 months).
More Best Practices for Rotating Equipment. http://dx.doi.org/10.1016/B978-0-12-809277-4.00008-5
Copyright © 2017 Elsevier Inc. All rights reserved.
403
B.P. 8.1
More Best Practices for Rotating Equipment
FIGURE 8.1.1 Pusher versus Non-Pusher.
SUPPORTING MATERIAL
Pusher Versus Non-Pusher
Mechanical seals are typically categorized into two major types: Pusher and
Non-Pusher. Refer to Fig. 8.1.1 for a picture of these two seal types.
On the left side of Fig. 8.1.1 is a typical pusher type seal. This consists of
a primary sealing ring assembled with an o-ring and springs (can be one or
multiple). The purpose of this o-ring (dynamic o-ring) is to force the sealing
fluid across the face and keep it from leaking to the ID (atmospheric) side of the
seal. The dynamic o-ring is designed to move axially (Pushed) along the shaft
or sleeve (in a cartridge seal). Therefore, the surface underneath the dynamic
o-ring must be very smooth (<32 RMS) to allow for the axial movement. In addition, if solids are abundant in the sealing fluid, they can build up on the o-ring
and prevent this axial movement (hang up).
The right side of Fig. 8.1.1 shows a typical bellows (Non-Pusher) type seal.
The bellows is a component that acts as both the load element (like a spring in a
pusher type) and a secondary sealing element (like an o-ring in a pusher type).
Because the bellows prevent any leakage to the atmospheric side of the seal, and
has a large clearance between itself and the shaft or sleeve, it can move freely in
the axial direction, reducing the potential for hang up.
Refer to Table 8.1.1 for details on these two types of seals.
TABLE 8.1.1 Single Seal (Pusher vs. Non-Pusher)
• Pusher Seal
– Closing force supplied by spring(s)
– Used in low temp services
– O-Ring secondary seals
– Used in light end services
(ethylene, propane, methane,
butane, etc.)
404
• Non-Pusher Seal
– Closing force supplied by bellows
(no dynamic o-ring)
– Can be used in high temp services
(Metal Bellows)
– Metal Bellows use grafoil secondary
seals to handle high temperature
Pump Mechanical Seals
Chapter | 8
FIGURE 8.1.2 Balance Ratio = Closing Area/Opening Area.
As seen in Table 8.1.1, pusher type seals are used more commonly in low
S.G. (<0.7) services. Remember our previous discussion on Balance Ratio
(Fig. 8.1.2), where balance ratio is the ratio of closing area to opening area of
the seal.
Our depiction (Fig. 8.1.2) is showing the concept of balance ratio in a pusher
type seal. In a bellows seal, the secondary sealing element (bellows) is generally
at a larger diameter than a pusher seal, therefore the closing area is less. Since
the closing area is larger and the width of the primary ring face is limited (cannot be too large or it won’t fit in the bellows assembly), the balance ratio cannot
be varied as much as in a pusher seal. With light S.G. fluids it is important to be
able to have a range of balance ratios to control where the fluid will vaporize
across the faces. It is for this reason that a pusher type seal is desired in light
S.G. services. Note that some applications can have a S.G. of less than 0.7 and
contain solids. In these applications, it is still recommended to use a pusher type
seal for the low S.G., however provisions need to be made to assure the seal will
not hang up in operation. Take a look at Table 8.1.2.
TABLE 8.1.2 Considerations for Pusher Type Seal in a Low S.G. Dirty
Service
• Single Coil Spring—reduces potential for springs to hang up since one large
diameter spring is used
• Filtration—Can be high cost and needs to be maintained for high reliability
• External Flush—If an external flush plan can be used, it will provide optimal seal life
in this app.
405
B.P. 8.1
More Best Practices for Rotating Equipment
An advantage for using the bellows seal, besides being less likely to hang up,
is that they typically utilize grafoil packing rings as their secondary seals. Grafoil packing rings can withstand temperatures of approximately 800°F (425°C),
allowing metal bellows seals to be used in refinery bottoms applications with
great success.
Dual Un-Pressurized Versus Dual Pressurized Seals
Today, due to environmental restrictions, Dual seals are being selected for more
and more applications. There are two arrangements in which dual seals can be
used, they are Dual Un-pressurized (used to be called Tandem) and Dual Pressurized (used to be called Double).
Refer to the diagram in Fig. 8.1.3, which shows the Dual Un-pressurized
arrangement.
A dual un-pressurized seal uses a buffer fluid at or near atmospheric pressure
to lubricate the atmospheric side seal. The buffer fluid pressure is significantly
less than the seal chamber pressure, so any leakage occurring across the process
side seal will leak into the buffer fluid. This arrangement is very common in
applications containing VOCs as it can potentially reduce the leakage to the
environment greatly.
Refer to Fig. 8.1.4, showing a dual pressurized seal arrangement.
Dual Pressurized Seals use a barrier fluid (same fluid as dual un-pressurized)
at a pressure 25 psi (1.75 bar) above seal chamber pressure to lubricate the
seals. Since the barrier fluid pressure is higher than the seal chamber pressure,
any leakage occurring at the process side seal will enter the pump. This arrangement (when working properly) assures no leakage of the pumped fluid
to atmosphere.
Refer to Table 8.1.3, listing the facts of these two seal arrangements.
It has been our experience that if an inert gas (or external fluid) is available
at the required pressure and the pumped fluid can accept the barrier fluid (compatible), a dual pressurized seal is potentially more reliable in VOC services.
FIGURE 8.1.3 Dual Un-Pressurized.
406
Pump Mechanical Seals
Chapter | 8
FIGURE 8.1.4 Dual Pressurized.
TABLE 8.1.3 Dual Unpressurized Versus Dual Pressurized
• Dual Unpressurized
– Primary leakage is pumped
fluid, therefore the secondary
seal is essentially a backup seal
– Does not require the use of
nitrogen or other inert gases
– Seal Reservoir must be vented
to flare or vapor collection
system to be effective
• Dual Pressurized
– Primary Leakage is barrier fluid into the
process, therefore the process must be
able to accept a small amount of oil
(usually synthetic ISO VG 32)
– Requires use of nitrogen (or other
compatible inert gas) to pressure seal
reservoir approximately 25 psi (1.75 bar)
above seal chamber pressure in a plan 53
– A plan 54 uses an external fluid
(synthetic skid or another pump) to
lubricate the seals at a pressure 25 psi
(1.75 bar) above seal chamber pressure
Material Selection of Faces and Secondary Components
In the seal design module, we discussed the design and function of the different
components in a seal. One aspect of design we did not touch on, however, was
material selection. It is very important that all parts be corrosion resistant to the
sealing fluid and allow for optimal sealing at the operating conditions.
Metal Parts (Adaptive Hardware)
Adaptive hardware will typically be made of 316 SS, however certain applications can dictate different materials be selected. Refer to Table 8.1.4.
Secondary Sealing Elements (O-Rings)
O-ring materials are very sensitive to different applications, therefore caution
must be used in selecting an o-ring. Refer to Table 8.1.5 for o-ring materials and
guidelines for use.
407
B.P. 8.1
More Best Practices for Rotating Equipment
TABLE 8.1.4 Adaptive Hardware Materials
• Normally 316 SS for large components (Sleeve, Gland, Retainer, etc.)
• Normally Hastelloy C or other corrosion resistant alloy for springs
• Acid or Chloride services may require hardware be constructed of Hastelloy C,
Chrome Alloys, Monel, or other corrosion resistant material.
TABLE 8.1.5 O-Ring Material and Usage Guidelines
• Fluorocarbon (Viton)—Most common (relatively cheap) and highly recommended
for HC services under 350°F (175°C)
• Perfluoro-elastomer (Kalrez or AFLAS)—used in higher temperature services (350 to
500°F or 175–260°C) than Viton and generally highly chemically resistant.
• EPDM—Common in hot water (BFW) applications, as it is more resistant to thermal
attack in hot water than the two listed previously.
• Buna-N—not recommended over the aforementioned three materials for most (if not
all) applications
Teflon (PTFE) and Grafoil packing have the highest temperature resistance,
but are stiff and poor secondary sealing elements compared to o-rings (o-rings
should be used when the application allows).
Seal Faces
Seal face materials also need to be selected to be compatible with the sealing
fluid, however there is another reason. If you recall our discussion on Face Generated Heat, you will remember the term f (coefficient of friction). This term
describes the amount of friction between the two face (primary ring and mating
ring) materials. Refer to Table 8.1.6 on material selection for faces.
Note the last point on Table 8.1.6. If the sealing fluid is close to its vapor margin (potential to flash), the last thing you would want is to create more
heat by having more face friction. Hard face combinations (Silicon Carbide vs.
Tungsten Carbide) should never be used in a sealing fluid with a high vapor
TABLE 8.1.6 Seal Face Materials
• Two dissimilar face materials are typically used (one softer than the other)
– Carbon vs. Silicon Carbide—f = 0.1
– Carbon vs. Tungsten Carbide—f = 0.12
• Abrasive services may require the use of two hard faces (Tungsten Carbide vs.
Silicon Carbide—f = 0.15 or more)
– Note that if fluid has a potential to flash, two hard faces should never be used.
408
Pump Mechanical Seals
Chapter | 8
TABLE 8.1.7 Optimal Flush Plans for Various Applications
• Mechanical Seal Flush must possess the following qualities for optimal seal life:
– Cool
– Clean
– Approx. 50 psi above vapor pressure (psia)
– Most importantly it must be cost effective!!!
pressure. Carbon vs. Silicon Carbide has the lowest coefficient of friction value
and is recommended for these applications.
Optimal Seal Plans for Various Applications
We know the design aspects and how to monitor different flush systems, but
when would certain flush systems be ideal over others? Tables 8.1.7–8.1.12 list
the parameters required for a reliable flush source.
Therefore, if you can say that the flush system for an application can provide
all the qualities (most important is cost effectiveness) listed in Table 8.1.7–8.1.12,
then you have the optimal flush.
The following slides will outline different sealing scenarios listing the considerations for different flush plans. Note that these are general and should not
be taken as being practical for every application, however they can aid in flush
plan selection.
TABLE 8.1.8 Optimal Flush Plans for Various Applications (Cont.)
• HC service with no known solid particles and temperature under 300°F
– Assuming satisfactory vapor margin (approx. 50 psi above vapor pressure) in seal
chamber a Plan 11 would be the optimal choice
– A vertical pump application would require venting of the seal chamber back to
suction if possible (Plan 13 or 14)
– A plan 52, 53, or 54 (53 or 54 preferred) would be recommended in low S.G.
(<0.7) services.
TABLE 8.1.9 Optimal Flush Plans for Various Applications (Cont.)
• Clean HC Service (e.g. # 2 FO) between 300 and 450°F
– Plan 23 is most efficient in cooling, however proper installation and venting is
required.
– Plan 21 will be sufficient in most cases and will be easier to operate…Note that
the orifice sizing is critical here as this determines the velocity of the fluid through
the heat exchanger.
409
B.P. 8.2
More Best Practices for Rotating Equipment
TABLE 8.1.10 Optimal Flush Plans for Various Applications (Cont.)
• Hot oil service above 450°F (typically tower bottoms)
– A single seal with the use of a plan 32 is the most reliable/cost effective if a
reliable source is available nearby…note that if the source is a product that has to
be reprocessed it may not be cost effective.
– A dual pressurized seal would be the best option (plan 53 or 54) if a plan 32 is
deemed not feasible.
TABLE 8.1.11 Optimal Flush Plans for Various Applications (Cont.)
• Acid Service (e.g., H2SO4)
– Reliable plan 32 can be used, however it is very critical for it to be operating at
all times when pump is installed in field (standby and startup situations)
– A dual pressurized seal, whether contacting (Plan 53/54) or non-contacting (Plan
74) will give the optimal seal life with ease of operation…Note that process
side seal components will need to be constructed of materials that are corrosion
resistant to the particular acid.
TABLE 8.1.12 Optimal Flush Plans for Various Applications (Cont.)
• Dirty Service (Containing Suspended Solid Particles)
– A cyclone separator (Plan 31) can be effective, however it must be sized correctly
and particle size must not fluctuate greatly.
– A clean external flush source (Plan 32 for single, Plan 54 for dual) will isolate the
seal faces from solids that can cause premature failure due to abrasion.
B.P. 8.2: Do not use nitrogen bottles to pressurize plan 53A flush systems.
While the nitrogen usage is minimal in this pressurized seal system, there
still is a small amount and if left connected to a nitrogen bottle, there will eventually be a time when the bottle capacity is used up.
Therefore, the plant N2 header should be used and if required, an amplifier
(small booster compressor, usually air driven) can be used if header pressure is
inadequate.
Remember that a plan 53 that has lost its method of pressurization has just
turned into a plan 52 (unpressurized seal) and you obviously installed a plan 53
because you are dealing with a hazardous fluid that should have close to zero
emissions.
L.L. 8.2: The use of nitrogen bottles to pressurize plan 53 flush plans have
resulted in unexpected seal failures when capacity runs low.
410
Pump Mechanical Seals
Chapter | 8
BENCHMARKS
This best practice has been in use since the late 1990s and has aided in maintaining MTBFs of close to 100 months in pressurized seals.
SUPPORTING MATERIAL
API Flush Plan 53
A plan 53 is basically a combination of a 52 and 54. It uses the same reservoir
as a plan 52, however the reservoir is pressurized at 25 psi (1.75 bar) above the
seal chamber pressure, just like a plan 54. Refer to Figure 8.2.1.
The reservoir is pressurized, however the system holds a constant pressure
of 25 psi above seal chamber pressure, requiring a means of circulation (pumping ring). The pumping ring circulation can be monitored the same way as in a
plan 52.
Also, as in a plan 52, pressure and level switch are recommended to alarm on
excessive leakage. Since this system is at a higher pressure than the seal chamber, leakage will migrate into the pump. A low pressure or level in the reservoir
will indicate this excessive leakage.
It is highly recommended to use a constant nitrogen supply (header) to
maintain this pressure as there will be a slight amount of usage. If a nitrogen
FIGURE 8.2.1 Plan 53—Dual Pressurized seal using a pressurized barrier fluid (usually
pressurized by a nitrogen blanket) to lubricate the seals.
411
B.P. 8.3
More Best Practices for Rotating Equipment
header is unavailable in the location of the seal, then a nitrogen bottle can be
used but should have a backup tank with a quick switchover valve, when the first
tank is capacity is used up.
B.P. 8.3: API plan 52/53 fluid circulation guidelines.
Whether Dual Pressurized or Unpressurized, the secondary seal flush (API
53A/B/C or API 52 respectively) are closed loop systems that rely on a pumping ring and thermo-siphoning (some seal chambers do not have enough radial
room for a pumping ring so just rely on thermo-siphoning) to circulate the
buffer/barrier fluid. This circulation is critical in lubricating and taking heat
away from the outboard (atmospheric side seal) as well as the inner diameter of
the inboard (process side seal) seal.
Therefore, it is essential to have the setup of these systems such that we
promote the best circulation for optimal seal life. Following are guidelines to
promote maximum buffer/barrier fluid circulation:
l
l
l
Mount bottom of reservoir (not outlet to seal but bottom of reservoir) between 18 and 24 in. (approximately 0.5–0.66 m) above the centerline of the
secondary seal flush inlet to the seal. This is high enough to give enough
head to return the fluid to the seal and low enough to typically allow for the
pumping ring to return the fluid back to the reservoir.
Use at least ¾ in. SS tubing to minimize pressure drop in the circuit
Minimize total tubing run to approximately 10 ft. (slightly more than 3 m)
and make sure not to have any hard 90 degree bends anywhere. Also there
should be a gradual slope up to the reservoir from the seal and back down to
the seal inlet with no drops below the centerline of the seal inlet.
The mentioned guidelines are important to incorporate early on in the project in order to promote optimal seal reliability from day one (see B.P. 8.1).
If, however you are experiencing Atmospheric seal failures on these types
of systems to the tune of about a seal failure a year or more, then circulation of
buffer fluid should be checked. This can be checked simply by looking at the
temperature difference in the tubing (using a contact thermometer) from inlet to
the reservoir to outlet of the reservoir. If the temperature is minimal (less than
5°F) or at zero, circulation is minimal or reverse and the mentioned guidelines
should be followed as close as possible to optimize the seal reliability.
L.L. 8.3: Improper piping and reservoir setup for dual seals has resulted in
MTBFs that are 12 months or less.
While the seal and the system type may be ideal for the specific application,
many times the author has seen where the way the system is set up in the field is
not ideal. It has been observed on many bad actor seals that the circulation of the
secondary flush was minimal due to hard small diameter piping runs, multiple
hard 90 degree bends, improper height of reservoir, etc. resulting in premature
seal failures.
412
Pump Mechanical Seals
Chapter | 8
BENCHMARKS
The writer has used this best practice since 2005 for new projects and “Bad Actor”
dual seals in the field. This best practice has resulted in seal MTBFs of greater
than 48 months for seals that previously had seal MTBFs below 12 months.
SUPPORTING MATERIAL
Pumping Rings
All dual unpressurized (Tandem) or dual pressurized (Double) seals utilize a
barrier fluid (oil, mixture of ethylene glycol and water, etc.) to lubricate and
take the heat away from the Atmospheric side seal and the ID of the Process side
seal. Since this is a closed loop system (Fig. 8.3.1), the pressure is essentially
equal throughout, therefore a means of circulation is needed. This means of
circulation is a pumping ring.
Probably the most appropriately named component in a mechanical seal, a
pumping ring, produces exactly what its name implies. It works as a very small
pump impeller to produce enough head to supply the lubricant back to the seal reservoir or heat exchanger. As shown in Fig. 8.3.2, there are many different designs
of pumping rings, however they fall into two categories, Axial and Radial Flow.
FIGURE 8.3.1 Closed Loop Circulation via Pumping Ring. Plan 52—Dual Unpressurized seal
using synthetic buffer fluid to lubricate the atmospheric side seal. A pumping ring in the seal circulates the buffer fluid (pressure less than process) to the reservoir.
413
B.P. 8.3
More Best Practices for Rotating Equipment
FIGURE 8.3.2 Different types of pumping rings.
A radial flow pumping ring is more widely used in horizontal applications
as it is believed to produce more circulation in the same system. Radial flow
pumping rings work on the principal of centrifugal force, with the barrier fluid
entering the ID of the pumping ring and being thrown out to the OD directly
into the barrier outlet port.
Axial flow pumping rings are always preferred in vertical pumps. They
work similarly to a screw pump with vanes that look like threads. The clearance between the gland ID and OD of the pumping ring vanes is critical in the
performance of these pumping rings. Although they may not produce the same
circulation in horizontal pumps as a radial flow pumping ring, axial flow pumping rings have the advantage in vertical pumps since they positively displace
the barrier fluid. The fluid enters a vane and will be pushed up to the barrier out
port. Since a radial flow pumping ring draws the fluid into the ID and throws it
out into the barrier out port (the top port in a vertical pump), it produces limited
to zero circulation in a vertical pump.
Since the amount of head produced by a pumping ring is very limited due to
the space inside the seal chamber, the system friction is very crucial for efficient
circulation in a closed loop system. It is very important that maximum size tubing (with no hard bends) is used and the heat exchanger or reservoir be placed
close (10 ft. of total tubing) to the seal to minimize this friction.
In addition to being used in tandem or double seals, pumping rings are used
in an API Flush Plan 23 (Fig. 8.3.3), which is a closed loop circulation of the
pumped fluid through a heat exchanger. This plan is very commonly used in
BFW service, as it can provide more efficient cooling than a typical cooled (API
Plan 21) flush plan.
414
Pump Mechanical Seals
Chapter | 8
FIGURE 8.3.3 Plan 23— Closed loop circulation of process fluid through a heat exchanger
via a pumping ring.
API Flush Plan 52
Dual Un-pressurized Seals (Tandem) rely on a buffer fluid at or near atmospheric pressure to lubricate the atmospheric side seal. This buffer fluid is circulated
via a pumping ring from the seal to the seal reservoir and back to the seal (in a
closed loop). Take a look at Fig. 8.3.4 for a schematic.
This flush plan is very common in applications with VOCs (Volatile Organic
Compounds), however if not set up similar to the schematic in Figure 8.3.4 it
may not be effective.
The reservoir is at atmospheric pressure (less than seal chamber pressure),
so the leakage across the process side seal faces migrates into the seal reservoir
and will either increase pressure, level, or both in the reservoir. Since every seal
does leak a certain amount, it is essential to have the reservoir vented to a
flare or vapor collection system. If the reservoir is allowed to reach the seal
chamber pressure, the atmospheric side seal will most likely fail (if it hasn’t
already) as it is not typically designed to handle seal chamber pressure. If this is
a concern in the plant, you may want to consider requesting the seal vendor to
redesign the atmospheric seal to handle maximum seal chamber pressure. In addition, as the process side seal leaks in this flush plan, the atmospheric side seal
will essentially be sealing the pumped fluid, exposing the plant to the release of
flammable and/or toxic vapors.
Monitoring of seal leaks can be done by checking the level and pressure of
the reservoir, as one or both may increase in the event of excessive leakage. The
seal vendor (or support system vendor) may supply high level and/or pressure
switches which would alarm the operators of a seal leak. It is highly recommended to specify this instrumentation in new projects, as it will cut down on
the already high workload of operators (if alarm sounds then the level or pressure can be personally verified in the field).
415
B.P. 8.3
More Best Practices for Rotating Equipment
FIGURE 8.3.4 Plan 52—Dual Unpressurized seal using synthetic buffer fluid to lubricate
the atmospheric side seal. A pumping ring in the seal circulates the buffer fluid (pressure less than
seal chamber) to the reservoir.
In addition to checking for excessive leakage (pressure or level increase),
temperature in and out of the seal can help verify proper circulation via the
pumping ring. In a reservoir that is not cooled, there should be a temperature
drop of approximately 5–10°F (1–2°C) from the seal outlet to the seal inlet. If
there is no temperature drop (or if the temperature increases), this indicates zero
or possibly reverse circulation. The reservoir needs to be placed close to the seal
(10 ft. of total tubing and 18–24 in. above the seal with ¾ in. SS tubing minimum) to assure the pumping ring will be able to provide the proper circulation.
It has been our experience that plan 52 flushes may not be vented properly
(blocked in) and level may be low or at zero in the reservoir. For these reasons,
it is very important to have the operators trained to understand the necessity to
monitor this system.
API Flush Plan 53
A plan 53 is basically a combination of a 52 and 54. It uses the same reservoir
as a plan 52, however the reservoir is pressurized at 25 psi (1.75 bar) above the
seal chamber pressure, just like a plan 54. Refer to Fig. 8.3.5.
The reservoir is pressurized, however the system holds a constant pressure of 25 psi above seal chamber pressure, requiring a means of circulation
416
Pump Mechanical Seals
Chapter | 8
FIGURE 8.3.5 Plan 53—Dual Pressurized seal using a pressurized barrier fluid (usually
pressurized by a nitrogen blanket) to lubricate the seals.
(pumping ring). The pumping ring circulation can be monitored the same way
as in a plan 52 and reservoir height and tubing sizing/length is just as critical.
Also, as in a plan 52, pressure and level switch are recommended to alarm on
excessive leakage. Since this system is at a higher pressure than the seal chamber, leakage will migrate into the pump. A low pressure or level in the reservoir
will indicate this excessive leakage.
It is highly recommended to use a constant nitrogen supply (header) to
maintain this pressure as there will be a slight amount of usage. If a nitrogen
header is unavailable in the location of the seal, then a nitrogen bottle can be
used but should have a backup tank with a quick switchover valve, when the first
tank is capacity is used up.
B.P. 8.4: When replacing a mechanical seal, ALWAYS check throat bushing
clearance and replace if out of tolerance.
Throat bushings are found in most pumps and can serve one of two purposes. In most instances (API Plan 11,13,14, 21, and 32) it is sized to create an
optimal flow and pressure in the seal chamber. In the case of a plan 23 (usually
used on Boiler Feed Water service) the purpose of the throat bushing is to separate the cool seal chamber from the hot process. Regardless of the flush plan,
a worn throat bushing will result in non-ideal conditions for the seal (whether
decreased pressure or increased temperature).
417
B.P. 8.4
More Best Practices for Rotating Equipment
Therefore, it is essential to be sure the throat bushing clearance is within
tolerance every time the seal is replaced, as it may actually have been the reason
for the seal failure in the first place.
Note that it is very important that the throat bushing be installed into the
pump casing such that it can be removed relatively easily in order to promote
the maintenance team to want to check the clearance. NEVER WELD THE
THROAT BUSHING INTO THE PUMP CASING!!!
L.L. 8.4: Failure to check and replace throat bushing has resulted in poor
seal MTBF due to the seal chamber conditions not being ideal.
Many times in the field it has been experienced that MTBF has been limited
and poor due to a throat bushing that was in very poor condition and even times
where it was not replaced because it was welded to the pump casing.
BENCHMARKS
This best practice has been recommended since 2005 to assure maximum seal
MTBFs in the field.
SUPPORTING MATERIAL
Seal Vendors design the seal balance ratio and select face materials based on a
parameter known as the PV to be able to change the fluid to a vapor approximately ¾ of the way down the seal faces (Fig. 8.4.1).
FIGURE 8.4.1 Mechanical seal primary face vaporization point.
418
Pump Mechanical Seals
Chapter | 8
In Fig. 8.4.1, three distinct operating modes are shown for the primary ring
(on the left) and the mating ring (on the right).
l
l
l
The top mode shows a condition of early vaporization which can occur
on light fluids, or hot fluids or where seal chamber pressure is lower than
designed (Note: Seal chamber pressure is designed to be at least 345 kpa
(50 psi) or above the seal fluid vapor pressure).
The middle figure shows the desired design condition of changing the fluid
to a vapor approximately ¾ of the way down the faces. This is the design
basis since it is known that fluid characteristics, temperature and/or pressure
will change during operation a certain amount. Excessive changes in any or
all of these parameters will lead to seal failure.
The last mode shows the case of no vaporization and apparent seal failure.
In reality, this apparent failure is not a failure but only a seal fluid condition
change that does not allow the seal fluid to reach the fluid vapor pressure
between the seal faces before it exits the seal.
As can be seen from Fig. 8.4.1, the ability to achieve the objective of vaporization ¾ or 75% down the seal faces (See center case in Fig. 8.4.1) depends on
the following:
l
l
l
Seal fluid characteristics—Cleanliness, Specific Heat, Vapor pressure, &
Viscosity
Pressure of the seal fluid in the seal chamber
Temperature of the seal fluid entering the seal chamber
What determines the condition of items 1, 2, 3 above? ------------------The
Process. Therefore, effective Mechanical Seal Condition Monitoring requires
that all seal process conditions, as noted previously are considered.
API Plan 11
The most commonly used flush plan, an API Plan 11 flush utilizes the pumped fluid to lubricate the seal faces. The pumped fluid is taken from discharge and sent to
the seal chamber through an orifice. Refer to Fig. 8.4.2 for the flush plan schematic.
The orifice is used to control the flow of the pumped fluid above the minimum required flow rate. Seal vendors require an orifice to assure the flow is not
too great either, as high flow rates can cause erosion of the seal faces. Equally
important, however, is the fact that an orifice limits the amount of recirculation
through the seal chamber back to the pump (The pump pumps money). A
3 mm (1/8 in.) orifice is the most commonly used size orifice as it is the smallest
practical size and plan 11 flushes are normally used in services that are easy to
seal (good lubricating qualities). Refer to Table 8.4.1, outlining general guidelines for orifice sizes.
Remember that Table 8.4.1 just shows guidelines for orifice sizes and are not
always followed, but if followed they should not potentially harm the seal if the
419
B.P. 8.4
FIGURE 8.4.2
seal.
More Best Practices for Rotating Equipment
Plan 11—recirculation from discharge through orifice and to mechanical
TABLE 8.4.1 Orifice Size Guidelines
j
j
j
Single stage pumps—1/8 in.
Multistage pumps—gang orifices (series of orifices back to back, sized by seal vendor)
Low S.G. fluids may require higher flow rates to prevent flashing (early vaporization)
where a 3/16 in. orifice may be necessary
pump is not operating in a region of the centrifugal pump curve where vaporization can occur. Note that orifice sizing does not give an exact flow due to the
system friction (piping, coolers, etc.), therefore the more information about the
system the seal vendor has, the more accurately they can size the orifice.
A vendor may require a close clearance throat bushing be installed in certain
instances to increase the pressure in the seal chamber above the vapor pressure
of the pumped fluid. As the throat bushing wears over time, the seal chamber
pressure will drop.
In a Plan 11 flush system, the main thing to monitor is the temperature across
the orifice. If the temperature drops by more than 10%, the orifice is most likely
plugging up. If this is the first occurrence and the fluid does not normally contain solids, the best option would be to clean the orifice out ASAP and continue
to check the temperature drop after the pump is started back up. If the orifice
has plugged up more than once, another flush plan option should be considered.
Refer to Table 8.4.2.
Look at Fig. 8.4.3 showing a typical plan 11. This is a between bearing
double suction pump with an orifice to each seal.
420
Pump Mechanical Seals
Chapter | 8
TABLE 8.4.2 Orifice Monitoring
j
j
j
j
Monitor temperature across the orifice
If temperature decreases by more than 10% it is beginning to plug
If first occurrence, clean and continue to monitor
If problem reoccurs, consider using a different flush plan
FIGURE 8.4.3 API Plan 11 (Typical).
API Plan 13
An API Plan 13 is widely used in vertical pump applications or when seal chamber pressure is at or near discharge pressure of the pump. This flush plan basically vents the seal chamber at a high point back to the suction of the pump
(ideally a high point in the suction piping). Refer to Fig. 8.4.4 for a schematic
of a Plan 13 flush.
As with a plan 11, this flush plan also utilizes an orifice, however it is more
to create a back pressure on the seal chamber than to control flow. The orifice
in a plan 13 flush is typically 6 mm (¼ in.), which is usually large enough to
vent vapors accumulated in a vertical pump, while keeping the vapor margin at
345 kPa (50 psi).
Since a plan 13 uses a larger orifice, it is not typically monitored with the
frequency a plan 11 is. From time to time, however it is a good idea during
421
B.P. 8.4
More Best Practices for Rotating Equipment
FIGURE 8.4.4 Plan 13—recirculation from seal chamber through orifice and back suction.
rounds to check the temperature across the orifice like you would do in a plan
11, to assure flow out of the seal chamber and no plugging of the orifice.
Very commonly, a plan 13 will be used in conjunction with a plan 11, this is
defined by API as a plan 14 flush. The same monitoring rules apply to a plan 14
that do to 11 and 13 flush plans.
API Plan 21
A flush plan 21 is a plan 11 with the addition of a cooler to lower the temperature of the pumped fluid. This plan is generally used when the pumped fluid is
naturally a good lubricant, however the vapor margin is low at the current pump
operating temperature and therefore requires the addition of a cooler to increase
the vapor margin. Refer to the schematic of a plan 21 in Fig. 8.4.5.
Just like a plan 11, a plan 21 utilizes an orifice to control the flow of the
pumped fluid. In this plan, orifice sizing is more critical, since you prefer to
have a flow close to the minimum required flow to get maximum cooling of
the pumped fluid. It has also been our experience that it is beneficial to place
this orifice downstream of the cooler (especially in low S.G. fluids) to prevent
vaporization before the cooler.
With the addition of the cooler, it is essential to check the temperature differential across the cooler. This should be done at initial installation (or after
cleaning of the cooler) and trended on a time basis. Typical seal flush coolers
should provide above 38°C (100°F) temperature drop, however this depends on
the cooling media temperature, showing why trending is important. If a water
cooled heat exchanger is being used, Cooling Water (CW) temperatures should
also be checked regularly, as a decrease in cooling efficiency could be the result
422
Pump Mechanical Seals
Chapter | 8
FIGURE 8.4.5 Plan 21—recirculation from pump discharge through an orifice and heat
exchanger to the mechanical seal.
of CW temperature or flow changes. It is ideal to have a thermometer installed
after the cooler to allow easy check of the seal fluid temperature. A thermometer
is recommended by API as an option, and if possible, it should be installed on
a plan 21. The pump operating conditions need to be considered as they could
alter the cooler inlet temperature.
Refer to Fig. 8.4.6. This pump is utilizing a process flush with a cooler,
before making its way to the seal. Note that this installation also includes a
component called a cyclone separator. The cyclone separator is used in services
that contain suspended solids and it works by sending the heavier solids back to
the suction of the pump while the “clean” liquid goes to the seal chamber. Our
experience with cyclone separators has been one of frustration at times. It is
very essential that the cyclone separator be sized correctly and that the solids be
significantly heavier (and don’t change size) than the pumped fluid. If not, the
solids will carry over to the seal and may even potentially plug the cyclone separator, preventing seal face lubrication. In addition, the solids that cause heavy
seal face wear are smaller than 2 µm (typical gap between seal faces), which a
cyclone separator cannot separate from the pumped fluid. With that said, when
a cyclone separator is properly sized it can reduce the potential of seal hang up
due to solids building up on the springs or dynamic o-ring.
The picture shown above is an example of an excellent flush plan installation. As you can see, a pressure gauge and thermometer are installed in the
piping just before entering the seal. These instruments give you an idea of the
flush system performance (and aids in troubleshooting failures) by allowing
the seal chamber pressure and temperature to be gathered easily (VAPOR
MARGIN!!!).
423
B.P. 8.4
More Best Practices for Rotating Equipment
FIGURE 8.4.6 API Plan 41 (Cyclone separator).
Considerations for External Flush Plans
When the pumped fluid contains a significant amount of solids, is very corrosive, or is at a very high or low temperature, an external flush should be
considered. An external flush plan uses a fluid from another pump or it may
be a separate console with a process compatible liquid to lubricate the seal
faces. We will now discuss two types of external flush plans, a plan 32 and
a plan 54.
API Flush Plan 32
A flush plan 32 utilizes a fluid from another pump that has good lubricating
qualities (Cool, Clean and at an acceptable vapor margin) at a higher pressure
than the seal chamber [at least 66 kPa (10 psi)], to keep the pumped fluid out of
the seal chamber. Refer to Fig. 8.4.7.
Since the plan 32 is at a higher pressure than the seal chamber, the plan 32
fluid will leak into the pump, therefore it must be compatible with the pumped
fluid and at an acceptable vapor margin. The leakage into the pump is controlled
by a bushing located either in the seal itself or the pump casing known as the
throat bushing.
For optimal seal life with a plan 32 flush (very common to exceed 3 years), the
flow is very critical. A typical value used by seal vendors is 3 LPM (0.75 GPM)
424
Pump Mechanical Seals
Chapter | 8
FIGURE 8.4.7 Plan 32—clean seal flush from external source.
per inch of seal size, however it is essential to assure the minimum required flow is met, while minimizing the flow into the pump. Refer to
Table 8.4.3 describing the methods of controlling the flow to the seal in a
plan 32.
Although they can be somewhat expensive, a flow meter installed in conjunction with a throttle valve is a very accurate way of controlling the flow to
the seal. No matter which method is used, operator training on the method is
important in assuring a reliable system.
A flush plan 32 can provide very long seal life, however it always needs
to be justified. A compatible fluid (with the process and with vapor margin) is
required and it must be provided at all times from start-up through shutdown of
the pump. Also, if the selected fluid is a product of the plant, long-term costs
(reprocessing a product because the product is not compatible with the pumped
fluid) may need to be considered.
TABLE 8.4.3 Manual Flow Control Options for API Plan 32
j
j
j
Orifice—not very accurate, since all system losses need to be considered
Throttle valve and pressure gauge—can be effective if seal chamber pressure does
not fluctuate much
Throttle valve and flow meter—most effective manual flow control
425
B.P. 8.5
More Best Practices for Rotating Equipment
FIGURE 8.4.8 Plan 23—closed loop circulation of process fluid through a heat exchanger
via a pumping ring.
API Flush Plan 23
A flush plan 23 is used on single seals, and consists of piping (usually tubing)
connected to the seal and a cooler. The fluid (seal chamber is filled with the
pumped fluid) is circulated to the cooler and back to the seal (Fig. 8.4.8).
This flush plan offers more efficient cooling than a flush plan 21, since it
does not have to continually cool down the pumped fluid. This can be related
to your car air conditioning for example. Most cars today have a button for recirculation of air, this closes the valve to the atmosphere, so the A/C just has to
cool of the air in the car, not the air from outside. A plan 23 works in the same
fashion. This however creates some concerns, because it relies heavily on the
pumping ring to circulate the fluid to the cooler. High system friction or vapor
pockets (not vented properly) will result in limited to no circulation, due to the
limited pumping capability of the pumping ring.
Therefore, the cooler needs to be placed close to the seal [3 m of total
tubing and 7–10 cm above the seal (10 ft. of total tubing and 18–24 in. above
the seal)] to assure the pumping ring will be sufficient. The cooler needs
to be above the seal to allow for venting and flow back to the seal (through
gravity and thermosyphon). It is also essential in a plan 23 to have a high
point vent and block valve installed to allow for proper venting before pump
start-up. Operators should be trained to understand the necessity of venting
these systems.
B.P. 8.5: If using an air type cooler, assure a fan is used to aid in cooling
and promote fluid circulation.
A fan installed on a fin-type air cooler will assure that a proper temperature
drop is achieved, which is the point of having a cooler installed in the first
place.
L.L. 8.5: The use of fin type air coolers without a fan have resulted
in minimal to zero temperature reduction and minimal to zero fluid
circulation.
426
Pump Mechanical Seals
Chapter | 8
Many times it has been observed that a fin type cooler without a fan provided
less than expected temperature drop and actually experienced no circulation
in closed loop circuits, since these coolers are generally large and create more
pressure drop than the pumping ring can handle.
BENCHMARKS
The writer has used this best practice since 2008 when it was observed that a
dual pressurized seal (plan 53B) had an air cooler installed in the circuit (in
the middle east region) which was vertically oriented and a height of over 4 m.
Naturally the pumping ring could not circulate the barrier fluid and the seal
failure rate was multiple times per year. By incorporating this best practice,
MTBFs have been greatly increased to more than 48 months.
SUPPORTING MATERIAL
See B.P. 8.4 supporting material for details.
B.P. 8.6: API plan 23 configuration and operation guidelines.
With this flush plan being installed in most Boiler Feedwater services it can
achieve very high MTBFs if proper setup and operation procedures are followed. Following are guidelines to be followed for optimal reliability in these
systems:
l
l
l
l
Mount heat exchanger between 18–24 in. (approximately 0.5–0.66 m) above
seal inlet centerline to assure the pumping ring can circulate the fluid.
Use at least ¾ in. ss tubing with a maximum total run of 10 ft. (3 m) with
minimal 90 degree bends to minimize system friction
When starting up the pump, the system needs to be vented of any vapor. It
should be done during start-up, then performed again about an hour later to
assure that all air is out of the system. Note that if the exchanger is vertically mounted, a vent before and after the cooler is needed in order to allow
operations to vent out the actual heat exchanger. Horizontally mounted heat
exchangers only require one high point vent.
Temperature should be monitored across the cooler and should not see a
significant drop (more than 20°F). Temperature typically should not reach
greater than 160°F (70°C) in order to assure optimal seal reliability. If temperature out of the seal increases and is approaching process temperature,
this indicates wear of throat bushing. If there is Zero temperature drop, then
there is air in the circuit and most likely needs to be vented.
L.L. 8.6: Failure set up and operate/monitor API plan 23 seal flush systems have resulted in seal MTBFs far below expected values (Lower than
12 months).
427
B.P. 8.6
More Best Practices for Rotating Equipment
This is usually because of inadequate circulation, whether it be from friction
in the system or improper venting of the system.
BENCHMARKS
The writer has used this best practice since 2005 and has resulted in seals in API
Plan 23 flush systems with MTBF’s greater than 48 months.
SUPPORTING MATERIAL
API Flush Plan 23
A flush plan 23 is used on single seals, and consists of piping (usually tubing)
connected to the seal and a cooler. The fluid (seal chamber is filled with the
pumped fluid) is circulated to the cooler and back to the seal (Fig. 8.6.1).
This flush plan offers more efficient cooling than a flush plan 21, since it
does not have to continually cool down the pumped fluid. This can be related
to your car air conditioning for example. Most cars today have a button for recirculation of air, this closes the valve to the atmosphere, so the A/C just has to
cool of the air in the car, not the air from outside. A plan 23 works in the same
fashion. This however creates some concerns, because it relies heavily on the
pumping ring to circulate the fluid to the cooler. High system friction or vapor
FIGURE 8.6.1 Plan 23—closed loop circulation of process fluid through a heat exchanger
via a pumping ring.
428
Pump Mechanical Seals
Chapter | 8
pockets (not vented properly) will result in limited to no circulation, due to the
limited pumping capability of the pumping ring.
Therefore, the cooler needs to be placed close to the seal [3 m of total tubing
and 7–10 cm above the seal (10 ft. of total tubing and 18–24 in. above the seal)]
to assure the pumping ring will be sufficient. The cooler needs to be above the
seal to allow for venting and flow back to the seal (through gravity and thermosyphon). It is also essential in a plan 23 to have a high point vent and block valve
installed to allow for proper venting before pump start-up. Operators should be
trained to understand the necessity of venting these systems.
B.P. 8.7: Utilize a constant flow control (Kates or equal) for external flush
systems (API 32 or 54).
If a single seal is preferred and the fluid is not desired to leak to the atmosphere an API plan 32 is ideal. Of course, the fluid being introduced to the seal
will go through the pump typically at a few gallons per minute or more flow,
therefore it needs to compatible with the pumped fluid. If that is the case, the
most important to have installed on this system is a way of automatically controlling the flow to the seal.
A constant flow control valve can be installed (not very expensive) in the
line and set to a specific value and will maintain this flow at all times. These
valves have been used in the industry for decades in Compressor oil seal applications and are very reliable.
For Dual pressurized seals it is also critical to assure a constant positive flow,
therefore this type of valve can be advantageous for control in these systems as well.
L.L. 8.7: Manually controlling the flow or pressure of a plan 32 or utilizing
just pressure control on plan 54s have resulted in numerous seal failures
due to inadequate flow during process changes.
Most installations of a plan 32 include a throttle valve with either a pressure gauge or flow meter and the operator sets this to the vendor recommended
settings (either flow or pressure). The problem is that when there is a process
change, the operator cannot get out there quick enough to modulate the flow of
the plan 32 and a seal failure will usually occur shortly after.
Many plan 54s incorporate pressure control without knowing the process
pressure on the other side of the process side seal. If process conditions change
and pressure increases on the other side of the process seal, the plan 54 pressure
setpoint can be inadequate.
BENCHMARKS
The writer has used this best practice for decades in Compressor Seal oil
systems and in the past few years in API plan 32 for mechanical seals and has
resulted in greatly increased seal MTBF due to automation of control versus
manual control.
429
B.P. 8.7
More Best Practices for Rotating Equipment
SUPPORTING MATERIAL
See details on API plan 32 in B.P. 8.4 supporting material.
The following figure is an excerpt of a brochure of the type of valve discussed in this B.P. and is provided courtesy of Custom Valve Concepts.
430
Pump Mechanical Seals
Chapter | 8
API Plan 54
A flush plan 54 is an external flush used on Dual Pressurized (Double) Seals.
This flush plan is typically used in applications with very corrosive fluids, or
when a plan 32 may not be feasible. Refer to Fig. 8.7.1 for a schematic of a plan
54 flush.
A plan 54 can use a fluid from another pump in the plant (Process Plan 54),
like a plan 32, or it can be a separate console with process compatible liquid
with sufficient vapor margin (Synthetic Plan 54). The advantage with a separate
console is that a potential product of the plant is not used, therefore long-term
FIGURE 8.7.1 Plan 54—Dual pressurized seal using external source to lubricate both seals
(source can be a process fluid or synthetic oil which is preferred).
431
B.P. 8.7
More Best Practices for Rotating Equipment
costs for reprocessing need not be considered. The synthetic plan 54 can potentially produce the longest seal life, but the up front cost (capital expenditure) of
this flush plan needs to be justified.
As with a plan 32 flush, the seal vendor will calculate the minimum flow
required for the seal (using the heat generated from two sets of faces). This flow
is used in sizing the seal pump (value can be from 6–10 GPM or 24–40 LPM).
Since the flow is not going directly to the seal chamber (through the bushing
into the pump but is a “through” flow back to the seal reservoir ), the pressure
is controlled at 25 psi (1.75 bar) above the seal chamber pressure. This assures
the plan 54 fluid is pushed across the faces, providing adequate lubrication.
A Process Plan 54 will most likely use a throttle valve and pressure gauge to
set the pressure, while a synthetic plan 54 is usually provided with a pressure
control valve to automatically control the pressure. Note that pressure control is
adequate in most applications, however if the seal chamber pressure fluctuates
during operation, a differential pressure control should be considered.
Another consideration when making a decision on a synthetic plan 54 versus a process 54 or a plan 32 is the potential addition of extra components. A
synthetic plan 54 will contain pump(s), coolers, filters, and instrumentation. As
everybody knows, the most reliable equipment has a balance of quality components and is as simple as possible (fewest number of components).
432
Chapter 9
Dry Gas Seals
B.P. 9.1: Submit a seal gas system P&ID to vendor as early as possible in
the project based on plant, company and industry lessons learned.
The P&ID can be a sketch, but should be detailed enough to show Best Practices you would definitely like in the system and the number of instruments to
be in the system. If done early enough (preferably during the Pre-FEED stage)
in the project, this will assure no cost adders for any additional BPs you would
like in the system.
When submitting this P&ID to the vendor it should definitely be made
known that it is not to force them into components and philosophies they disagree with or are uncomfortable with. Rather, this P&ID shows what the user
would prefer, but if the vendor has a better idea it can be discussed during the
Pre-bid meetings and decided by all parties what the best path forward is to assure the highest reliability in the system.
L.L. 9.1: Failure to provide seal gas system P&ID to vendor based on lessons learned has resulted in unreliable systems that have caused unplanned
shutdowns and revenue loss.
The following examples highlight omitted details in dry gas seal specifications that have resulted in seal MTBFs less than 12 months:
l
l
l
l
l
l
Failure to identify saturated seal gas conditions at start-up, upset or operating conditions to determine if an external seal gas or gas conditioning unit
should be required
Failure to properly specify maximum flare header pressure and identify if a
high signal select in primary vent is required
Failure to prohibit the use of orifices in the secondary vent resulting in seal
pressure reversals
Failure to specify oil sampling devices in the secondary seal vent port (Sight
glasses, valves, or automatic drainers) leading to secondary seal oil contamination and eventual failure
Failure to include Secondary Seal Monitoring
Failure to include a backpressure control device in the primary vent
More Best Practices for Rotating Equipment. http://dx.doi.org/10.1016/B978-0-12-809277-4.00009-7
Copyright © 2017 Elsevier Inc. All rights reserved.
433
B.P. 9.1
More Best Practices for Rotating Equipment
BENCHMARKS
This best practice has been in use since the late 1990s to specify dry gas seal
system requirements during projects and for field modifications. This approach
has resulted in dry gas seal systems of the highest safety levels and reliability
(Seal MTBFs greater than 90 months).
SUPPORTING MATERIAL
Fig. 9.1.1 shows a typical P&ID sketch that would be provided to the vendor
for a low suction pressure compressor with a gas that could have condensables
and an external clean/dry seal gas was not available. The sketch contains a lot of
industry best practices that should be considered based on the application and
the vendors should be familiar and experienced with the methods presented in
this sketch or other methods they use to accomplish the same goal.
Dry Gas Seal (DGS) systems have been used for the past two decades, and
are specified by many end users as the seal of choice for most compressor applications. One would therefore think that seal and system designs are well-known
and proven. However, experience shows that failures are still quite common. For
instance, in 2007, FAI dealt with nearly 50 DGS failures.
These failures raise several questions. Are they all caused by “foreign material” contamination, or ingestion? Are they connected with improper seal selection or unreliable system hardware? Who is responsible: seal vendors, compressor vendors, or end-users?
In reviewing DGS failures experienced in 2006 and previous years, the conclusion is that in a majority of cases, the root cause is that the seal and system
configuration were not designed to handle all the actual site operating conditions, including startup, shut-down and upsets that should and could have been
anticipated.
The end-user has the most complete knowledge of the process and plant operating procedures. Therefore, he or she needs to be proactive in terms of project
DGS requirements, and specify the type of seal and system most suited to the
plant and application, based on his or her knowledge and experience. Seal and
compressor vendor input and experience are obviously required, but neglecting
to evaluate the proposed system in detail against all operating modes subjects
the user to the risk of unacceptable downtime and revenue losses, particularly in
the “mega plants” being built today.
Fig. 9.1.2 shows a recommended “Best Practice” P & ID for a tandem dry
gas seal system in a critical (unspared) application used in a large plant of high
daily revenue (greater than $1MM/day).
The reliability of critical equipment is dependent on the reliability of each
component in every auxiliary system connected with the critical equipment
unit. How do we maximize critical equipment reliability? The easiest way is
to eliminate the auxiliary systems. Imagine the opportunity to eliminate all of
the components; pumps, filters, reservoirs, etc. and thereby increase reliability
434
Dry Gas Seals
Chapter | 9
435
FIGURE 9.1.1 Dry gas seal system sketch.
B.P. 9.1
More Best Practices for Rotating Equipment
FIGURE 9.1.2 Best Practice Tandem Seal P & ID.
436
Dry Gas Seals
Chapter | 9
and hopefully, the safety of the equipment. The gas seal as used in compressor
applications affords the opportunity to achieve these objectives. However, the
gas seal is still part of a system and the entire gas seal system must be properly
specified, designed, maintained, and operated to achieve the objectives of optimum safety and reliability of the critical equipment.
In this section, the principles of gas seal design will be discussed and applied
to various gas seal system types. In addition, best practices will be discussed for
saturated gas systems as well as shutdown philosophies.
System Function
The function of a gas seal system is naturally the same as a liquid seal system.
The function of a fluid seal system, remembering that a fluid can be a liquid or
a gas is to continuously supply clean fluid to each specified seal interface point
at the required differential pressure, temperature, and flow rate. Therefore, one
would expect the design of a gas seal and a liquid seal to be very similar, which,
in fact, they are. Then why are their systems so different?
Comparison of a Liquid and Gas Sealing System
Fig. 9.1.3 shows a liquid sealing system. Compare this system to Fig. 9.1.4
which shows a gas seal system, if the same compressor were retrofitted for a
gas seal. WOW!! What a difference. Why are there such a small amount of
components for the gas seal system? As an aid, refer to Fig. 9.1.5 which shows a
typical pump mechanical seal in an API flush plan 11. This system incorporates
a liquid mechanical seal and utilizes pump discharge liquid as a flush for the
seal. Refer now to Fig. 9.1.4 and observe the similarities. It should be evident
that a gas seal system is simplified in compressor applications over a liquid seal
system merely because the gas seal utilizes the process fluid. This is exactly
the same case for a pump. By using the process fluid, and not a liquid, one can
eliminate the need to separate liquid from a gas, thereby totally eliminating the
need for a liquid supply system and the need for a contaminated liquid (sour
oil), drain system.
Referring back to Fig. 9.1.3, therefore, we can see that the following major
components are eliminated:
1.
2.
3.
4.
5.
6.
7.
8.
9.
The seal oil reservoir
The pumping units
The exchangers
The temperature control valves
The overhead tank
The drain pot
The degassing tank
All control valves
A significant amount of instrumentation
437
B.P. 9.1
438
More Best Practices for Rotating Equipment
FIGURE 9.1.3 Typical seal oil system for clearance bushing seal. (Courtesy of M.E. Crane Consultant)
Dry Gas Seals
Chapter | 9
FIGURE 9.1.4 Typical gas seal system for dry air or inert gas. (Courtesy of John Crane Co.)
Referring back to the function definition of the gas seal system, all requirements are met. “Continuously supplying fluid” is met by utilizing the discharge
pressure of the compressor. The requirements for “specified differential pressure, temperature and flow rate” are met by the design of the seal itself which
can accommodate high differential pressures, high temperatures, and is sized
to maintain a flow rate that will remove frictional heat necessary to maintain
seal reliability. The only requirement not met is that of supplying a clean dry
fluid, and this can be seen in Fig. 9.1.4. This requirement is met by using a dual
system coalescing filter.
439
B.P. 9.1
More Best Practices for Rotating Equipment
FIGURE 9.1.5 Pump mechanical seal in API flush plan 11.
When one considers all the advantages, the next question to ask is, okay,
what are the disadvantages? Naturally, there are disadvantages. However, proper design of the gas seal system can minimize and eliminate many of the disadvantages. Do not forget that the requirements for any system mandate proper
specification, design, manufacture, operation, and maintenance. One can never
eliminate these requirements in any critical equipment system.
Considerations for System Design
As mentioned earlier, there are disadvantages to a gas seal system which are not
insurmountable but must be considered in the design of such a system. These
considerations are as follows:
Sensitivity to dirt—since clearances between seal faces are usually less than
0.0005 in. and seal design is essential to proper operation, the fluid passing
between the faces must be clean (5–10 µm maximum particle size). If it is not,
the small grooves (indentations) necessary for seal face separation will become
plugged thus causing face contact and seal failure.
Sensitivity to saturated gas—saturated fluids increase the probability of
groove (indentation) blockage.
Lift-off speed—as will be explained later, a minimum speed is required for
operation. Care must be taken in variable speed operation to assure that operation is always above this speed. It is recommended that the seal test be conducted for a period at turning gear speed to confirm proper “lift off” followed
by seal face inspection.
Positive prevention of toxic gas leaks to atmosphere—since all seals leak,
the system must be designed to preclude the possibility of toxic of flammable
gas leaks out of the system. This will be discussed in detail subsequently.
440
Dry Gas Seals
Chapter | 9
Possible oil ingestion from the lube system—a suitable separation seal must
be provided to eliminate the possibility of oil ingestion from the bearings.
Whenever a gas seal system is utilized, the design of the critical equipment by
definition incorporates a separate lube oil and seal system. Consideration must
be given during the design or retrofit phases to the separation between the liquid
(lube) and gas seal system.
“O” ring (secondary seal components) design and maintenance—most seal
vendors state that “O” ring life is limited and should be changed every 5 years
for operating seals as well as spare seals. The writer’s experience has shown that
dry gas “O” ring seals can exceed this limit. It is recommended that seal vendors
be required to provide references for similar applications prior to making a decision to change out the seals after 5 years.
If all of the aforementioned considerations are incorporated in the design of
a gas seal system, its reliability has the potential to exceed that of a liquid seal
system and the operating costs can be reduced.
Before moving to the next section, however, one must consider that relative
reliability between gas and liquid seal systems is a function of proper specification, design, etc. as mentioned previously. A properly designed liquid seal
system that is operated and maintained can achieve reliabilities of a gas seal
system. Also, when one considers operating costs of the two systems, various
factors must be considered. While the loss of costly seal oil is positively eliminated, with a gas seal system (assuming oil ingestion from the lube system does
not occur) the loss of process gas, while minimal, can be expensive. It is argued
that the loss of process gas from a liquid seal system through drainer vents and
degassing tank vents is also significant. While this may be true in many cases,
a properly specified, designed, and operated liquid seal system can minimize
process gas leakage such that it is equal to or even less than that of a gas seal.
There is no question that gas seal systems contain far fewer components
and are easier to maintain than liquid seal systems. These systems will be used
extensively in the years ahead. The intention of this discussion is to point out
that existing liquid seal systems that cannot be justified for retrofit or cannot be
retrofitted easily can be modified to minimize outward gas leakage and optimize
safety and reliability.
Dry Gas Seal Design
Principles of Operation
The intention of this sub-section is to present a brief detail of the principles of
operation of a dry gas seal in a conceptual form. The reader is directed to any
of the good literature available on this subject for a detailed review of gas seal
design.
Refer to Fig. 9.1.6.
Fig. 9.1.6 shows a mechanical seal utilized for pump applications, while
Fig. 9.1.7 shows a dry gas mechanical seal utilized for compressor application.
441
B.P. 9.1
More Best Practices for Rotating Equipment
FIGURE 9.1.6 Typical pump single mechanical seal.
The seal designs appear to be almost identical. Close attention to Fig. 9.1.7,
however, will show reliefs of the rotating face of the seal.
Considering that both seals operate on a fluid may give some hint as to
why the designs are very similar. The objective of seal designs is to positively
minimize leakage while removing frictional heat to obtain reliable continuous
operation of the seal. In a liquid application, the heat is removed by the fluid
which passes between the rotating and stationary faces and the seal flush and
changes from a liquid to gaseous state (heat of vaporization). This is precisely
why all seals are said to leak and explains the recent movement in the industry
to sealless pumps in toxic or flammable service. If the fluid between the rotating
faces now becomes a gas, its capacity to absorb frictional heat is significantly
less than that of a liquid. Therefore an “equivalent orifice” must continuously
exist between the faces to reduce friction and allow a sufficient amount of fluid
to pass and thus take away the heat. The problem obviously is how to obtain this
“equivalent orifice.” There are many different designs of gas seals. However,
regardless of the design, the dynamic action of the rotating face must create a
FIGURE 9.1.7 Typical design for curved face—spiral groove non-contact seal; curvature may
alternately be on rotor. (Courtesy of John Crane Co.)
442
Dry Gas Seals
Chapter | 9
FIGURE 9.1.8 Dry gas seal. Top: typical design for curved face—spiral groove non-contact seal;
curvature may alternately be on rotor; bottom: typical spiral groove pattern on face of seal typical
non-contact gas seal. (Courtesy of John Crane Co.)
dynamic opening force that will overcome the static closing forces acting on the
seal to create an opening and hence “equivalent orifice.”
Refer to Fig. 9.1.8 which shows a typical gas dry seal face. Notice the spiral
grooves in this picture, they are typically machined at a depth of 100–400 µin.
When rotating, these vanes create a high head low flow impeller that pumps
gas into the area between the stationary and the rotating face, thereby increasing the pressure between the faces. When this pressure is greater than the
static pressure holding the faces together, the faces will separate thus forming
an equivalent orifice. In this specific seal design, the annulus below the vanes
forms a tight face such that under static (stationary) conditions, zero leakage
can be obtained if the seal is properly pressure balanced. Refer to Fig. 9.1.9 for
a force diagram that shows how this operation occurs.
In Fig. 9.1.9, the rotation of the face must be counter-clockwise to force the
gas into the passages and create an opening (Fo) force. This design is known
as a “uni-directional” design and requires that the faces always operate in this
direction. Alternative face designs are available that all rotate in either direction
and they are known as “bi-directional” designs.
Ranges of Operation
Essentially, gas seals can be designed to operate at speeds and pressure differentials equal to or greater than those of liquid seals. Present state-of-the-art
443
B.P. 9.1
More Best Practices for Rotating Equipment
FIGURE 9.1.9 Hydrostatic force balance on seal stator (FC = FO). (Courtesy of John Crane Co.)
(2010) limits seal face differentials to approximately 17,250 kPa (2,500 psi) and
rubbing speeds to approximately 122 m/s (400 ft/s). Temperatures of operation
can reach as high as 538°C (1000°F). Where seal face differential exceeds these
values, seals can be used in series (tandem) to meet specifications provided sufficient axial space is available in the seal housing.
Leakage Rates
Since the gas seal when operating forms an equivalent orifice, whose differential is equal to the supply pressure minus the seal reference pressure, there will
always be a certain amount of leakage. Refer to Fig. 9.1.10 for leakage graphs.
It can be stated in general that for most compressor applications with suction
pressures on the order of 3450 kPa (500 psi) and below, leakage can be maintained on the order of one standard cubic foot per minute per seal. For a high
pressure application 17,250 kPa (2,500 psi), differential leakage values can be
as high as 8.5 Nm3/h (5 SCFM) per seal. As in any seal design, the total leakage
is equal to the leakage across the seal faces and any leakage across secondary
seals (O-rings, etc.). There have been reported incidence of explosive O-ring
failure on rapid decompression of systems incorporating gas seals, thus resulting in excessive leakage. Consideration must be given to the system in order to
tailor system decompression times in order to meet the requirements of the secondary seals. As previously mentioned, all gas seals will leak, but not until the
face “lifts off.” This speed known, oddly enough, as “lift off speed” is usually
444
Dry Gas Seals
Chapter | 9
FIGURE 9.1.10 Dry gas seal leakage rates. (Courtesy of John Crane Co.)
less than 500 rpm. Caution must be exercised in variable speed applications to
assure the system prevents the operation of the variable speed driver below this
minimum lift off value. One recommendation concerning instrumentation is to
provide one or two thermocouples in the stationary face of each seal to measure
seal face temperature. This information is very valuable in determining lift off
speed and condition of the grooves in the rotating seal face. Any clogging of
these grooves will result in a higher face temperature and will be a good indication of requirement for seal maintenance.
Gas Seal System Types
As mentioned in this section, in order to assure the safety and reliability of gas
seals, the system must be properly specified and designed. Typical gas seal system applications in use today are listed as follows.
Low/Medium Pressure Applications—Dry Air or Inert Gas
Fig. 9.1.4 shows such a system. This system incorporating a single dry gas
seal is identical to that of a liquid pump flush system incorporating relatively
clean fluid that meets the requirements of the seal in terms of temperature and
445
B.P. 9.1
More Best Practices for Rotating Equipment
pressure. This system takes the motive fluid from the discharge of the compressor through dual filters (10 µm or less) incorporating a differential pressure gage
and proportions equal flow through flow meters to each seal on the compressor.
Compressors are usually pressure balanced such that the pressure on each end
is approximately equal to the suction pressure of the compressor. The clean gas
then enters the seal chamber and has two main paths:
1. Through the internal labyrinth back to the compressor. Note that the majority of supplied gas takes this path for cooling purposes (99%).
2. Across the seal face and back to either the suction of the compressor or to
vent.
Since the gas in this application is inert, it can be vented directly to the atmosphere or can be put back to the compressor suction. It must be noted, however,
that this port is next to the journal bearing. Therefore a means of positively
preventing entry of lube oil into this port must be provided in order to prevent
the loss of lube oil or prevent the ingestion of lube oil into the compressor if
this line is referenced back to the compressor suction. A suitable design must
be incorporated for this bushing. Typically called a disaster bushing, it serves a
dual purpose of isolating the lube system from the seal system and providing a
means to minimize leakage of process fluid into the lube system in the event of
a gas seal failure. In this system, a pressure switch upstream of an orifice in a
vent line is used as an alarm and a shutdown to monitor flow. This switch uses
the concept of an equivalent vessel in that increased seal leakage will increase
the rate of supply versus demand flow in the equivalent vessel (pipe) and result in a higher pressure. When a high flow is reached, the orifice and pressure
switch setting are thus sized and selected to alarm and shut down the unit if
necessary. As in any system, close attention to changes in operating parameters
is required. Flow meters must be properly sized and maintained clean such that
relative changes in the flows can be detected in order to adequately plan for seal
maintenance.
High Pressure Applications
In this application, for pressures in excess of 6895 kPa (1000 psi), a tandem seal
arrangement or series seal arrangement is usually used. Since failure of the inner seal would cause significant upset of the seal system, and large amounts of
gas escaping to the atmosphere, a backup seal is employed. Refer to Fig. 9.1.11
which shows a triple dry gas tandem seal. For present designs up to 17,250 kPa
(2,500 psi), double tandem seals are proven and used.
The arrangement is essentially the same as low/medium pressure applications except that a backup seal is used in place of the disaster bushing. Most
designs still incorporate a disaster bushing between the backup seal and the
bearing cavity known these days as the barrier seal. Attention in this design
must be given to control of the inter-stage pressure between the primary and
backup seal. Experience has shown that low differentials across the backup seal
446
Dry Gas Seals
Chapter | 9
FIGURE 9.1.11 Dry gas seal: a triple tandem dry gas seal arrangement. (Courtesy of DresserRand Corp.)
can significantly decrease its life. As in the case of liquid seals, a minimum
pressure in the cavity between the seals of 172–207 kPa (25–30 psi) is usually
specified. This is achieved by properly sizing the orifice in the vent or reference
line back to the suction to assure this pressure is maintained. All instrumentation and filtration are identical to that of the previous system.
Dual Seal and System Options for Toxic and/or Flammable Gas
Applications
There are many field proven options available today for use in toxic and/or flammable gas applications. In this section we will discuss the following systems:
j
j
j
j
Tandem seals for dry gas applications
Tandem seals for saturated gas applications
Tandem seals with inter-stage labyrinth and nitrogen separation gas
Double seal system for dry gas or saturated gas applications
Tandem Seals for Dry Gas Applications
The tandem seal arrangement for this application is shown in Fig. 9.1.12 and a
schematic of this seal in the compressor seal housing is shown in Fig. 9.1.13.
Gas from the compressor discharge enters the port closest to the compressor
labyrinth end and the majority of the gas enters the compressor through this
labyrinth. To assure that process gas, which is not treated by the dry gas system, does not enter the seal chamber, velocities across the labyrinth should be
447
B.P. 9.1
More Best Practices for Rotating Equipment
FIGURE 9.1.12 Tandem seal. (Courtesy of Flowserve Corp.)
FIGURE 9.1.13 Tandem seal and barrier seal typical housing arrangement.
maintained between 6 and 15 m/s (20–50 ft/s). It is the writer’s experience that
considering labyrinth wear, the design should be closer to 15 m/s (50 ft/s).
Approximately 1.7–3.4 Nm3/h (1–2 SCFM) flow (standard cubic feet per
minute) leak across the first tandem seal faces (primary seal) and exit through the
primary vent. Based on the backpressure of the primary vent system, 1.7 Nm3/h
(1 SCFM) or less will pass through the second tandem seal faces (secondary
448
Dry Gas Seals
Chapter | 9
seal) and exit through the secondary vent. To assure that oil mist from the bearing housing does not enter the dry gas seal chamber and that seal gas does not
escape to atmosphere, an additional barrier seal is used and provided with pressurized nitrogen at approximately 35 kPa (5 psi).
A typical seal system for this arrangement is shown in Fig. 9.1.14. As previously mentioned, dry gas seal reliability depends on the condition of the gas
entering the seal faces. The function of the seal gas supply system for any dry
gas seal option is to continuously supply clean, dry gas to the seal faces. During
start-up, when the compressor is not operating with sufficient pressure to supply the seals, an alternate source of gas or a gas pressure booster system should
be provided. These items are shown in Fig. 9.1.14 and are typical for any type
of dry gas seal application. Note that the following options exist regarding the
primary, secondary vent and barrier seal instrumentation and components:
j
j
j
j
j
Primary seal vent triple redundant (2 of 3 voting) flow or differential pressure alarm and shutdown
Primary seal vent rupture discs in parallel with vent line to rupture at a set
pressure and prevent excessive pressure to the secondary seal on primary
seal failure
Spring loaded exercise valves in the primary vent line to exert a backpressure
on the primary seal to close the faces in the event of dynamic “O” ring hang-up
Secondary vent line flow or differential pressure alarms and trips
Barrier seal supply pressure alarm and permissive not to start the lube oil
system until barrier seal minimum pressure is established.
Tandem Seals for Saturated Gas Applications
The tandem seal arrangement for this application can be exactly the same as
that shown in Figs. 9.1.12 and 9.1.13 for the dry gas application. The changes
required for a saturated gas are solely in the seal system. A typical system is
shown in Fig. 9.1.14 and incorporates a cooler, separator, and heater in addition
to the normal components used for a dry gas application to assure that saturated
gas does not enter the seal chamber. Typical values for the cooler are to reduce
the gas temperature to 30°F below the saturation temperature of the gas. The typical dimensions for the separator vessel, complete with a demister, are 460 mm
(18 in.) diameter and 1.8 m (6 ft) high. The typical requirements for the heater are
to reheat the gas to 15°C (30°F) above the saturation temperature. Temperature
transmitters are provided upstream and downstream of the cooler and downstream of the heater. As a precaution, in the event of cooler or heater malfunction,
a dual filter/coalescer, complete with a drain back to the suction is provided.
Tandem Seals With Interstage Labyrinth
The present (2010) industry “best practice” tandem seal arrangement for dry or
saturated gas applications is shown in Fig. 9.1.15. This arrangement features
a labyrinth between the primary and secondary seals. This action assures that
449
B.P. 9.1
450
More Best Practices for Rotating Equipment
FIGURE 9.1.14 Typical tandem seal system for saturated process gas.
Dry Gas Seals
Chapter | 9
451
FIGURE 9.1.15 Typical tandem seal system for dry process gas.
B.P. 9.1
More Best Practices for Rotating Equipment
FIGURE 9.1.16 Tandem seal with interstage labyrinth. (Courtesy of Flowserve Corp.)
gas vented from the secondary seal will always be nitrogen since the nitrogen
supplied between the primary and secondary seals is differential pressure controlled to always be at a higher pressure than the primary seal vent thus assuring
that only nitrogen will be in the chamber between the primary and secondary
seals. Figs. 9.1.16 and 9.1.17 shows a typical nitrogen upply system used with
this tandem seal configuration.
Double Seal System for Dry Gas or Saturated Gas Application
Fig. 9.1.18 depicts a double seal used in either dry gas or saturated gas applications where the process gas is not permitted to exit the compressor case. For this
application process gas can be used, after it is conditioned, or an external source
can be used if it is compatible with the process gas. If the gas used between the
seals is toxic or flammable, a suitable barrier seal, provided with nitrogen, as
shown in Fig. 9.1.13 must be used. The seal systems previously shown will be
used for the supply of conditioned gas to the seals as required by the condition
of the seal gas (dry or saturated).
Summary
Since there are significant advantages to the use of dry gas seals, many units
are being retrofitted in the field which incorporates this system. In many cases,
significant payouts can be realized.
If a unit is to be retrofitted, it is strongly recommended that the design of
the gas seal be thoroughly audited to assure safety and reliability. As mentioned in this section, retrofitting from a liquid to a gas seal system renders
the unit a separate system type unit, that is, a separate lube and gas seal
system. Naturally, loss of lube oil into the seal system will result in significant costs and could result in seal damage or failure by accumulating debris
between the seal rotating and the stationary faces. The adequate design of
the separation barriers between the lube and seal face must be thoroughly
examined and audited to assure reliable and safe operation of this system.
452
Dry Gas Seals
Chapter | 9
453
FIGURE 9.1.17 Typical tandem seal system with an interstage labyrinth-nitrogen supply.
B.P. 9.2
More Best Practices for Rotating Equipment
FIGURE 9.1.18 Double seal. (Courtesy of Flowserve Corp.)
Many unscheduled field shutdowns and safety situations have resulted from
the improper design of the lube system, seal system separation labyrinth.
In addition to the aforementioned considerations, a critical speed analysis,
rotor response and stability analysis [if the operating discharge pressure is
above 3450 kPa (500 psi)] should always be conducted when retrofitting
from liquid to dry gas seals.
B.P. 9.2: If sufficient Nitrogen pressure is not available for normal operation
of a double dry gas seal, utilize a nitrogen amplifier (booster compressor).
A double dry gas seal allows for the most reliable seal gas system since
much of the piping and instruments are eliminated that are contained in a tandem dry gas seal arrangement. That said, why doesn’t everyone install double
dry gas seals across the board? Well, one simple reason is the nitrogen header
pressure can be insufficient for many applications.
Therefore, following this best practice will allow for the use of double seals
in many applications (as long as a very small amount of N2 can be accepted into
the process). The amplifier or booster is a small reciprocating compressor that
is generally very reliable, however it is highly recommended to install two of
these items with one as an auxiliary since one is needed to provide sufficient N2
pressure at all times during operation.
L.L. 9.2: Double dry gas seal systems as compared to tandem systems eliminate the following items to reduce complexity and optimize reliability.
l
l
l
Primary vent hardware and instrumentation
Concerns with flare header pressures that can cause seal pressure reversals
The intermediate N2 gas system
454
Dry Gas Seals
Chapter | 9
It is the writer’s opinion that end user lack of participation in the system
specifications has resulted in the use of more complex tandem seal systems
where double seal systems could be employed. The N2 amplifier allows for these
seals to be used in many more applications where N2 pressure is insufficient.
BENCHMARKS
FAI has recommended double seals for new projects since 2000. Double gas
seals have been used in low pressure coker and wet gas compressor applications.
SUPPORTING MATERIAL
Double seals can help simplify the seal gas control system, minimize the quantity of seal gas, and optimize system reliability. Double seals are normally applied where an inert seal gas (usually N2), which is compatible with the process,
is available at a pressure exceeding the maximum process pressure at the seal
interface (to prevent a seal pressure reversal). If N2 from a regulated system is
used, the seal gas control valve can be eliminated (Fig. 9.2.1).
If the process gas is sour, a sweet buffer gas must be injected between the process labyrinth and DGS to prevent sour gas contact and potential DGS fouling.
FIGURE 9.2.1 Double dry gas seal.
455
B.P. 9.4
More Best Practices for Rotating Equipment
Differential pressure control is typically used. Flow control is also an acceptable
option, provided the flow is sufficient to maintain a velocity of 15 m/s (50 ft/s)
through the process labyrinth at twice the maximum design clearance.
B.P. 9.3: Use an amplifier (booster compressor) for start-up on tandem
seals when the primary seal gas supply is taken from discharge of the compressor.
It is highly recommended to have a filtered seal gas supply when starting up
the compressor to keep the unfiltered gas that is going through the compressor
away from the seal faces and other components that may be susceptible to hangup (o-rings and maybe even springs if a lot of contamination).
Therefore, an amplifier similar to the one mentioned in B.P. 9.2 is a very
reliable way to assure a filtered gas is supplied at the appropriate differential
pressure (or flow) over the reference gas.
Since the amplifier will not be needed during normal operation, you can use
just one (no auxiliary needed), but when not in use it should be checked and
PMd regularly to assure it will be available during the next train start-up.
L.L. 9.3: Failure to have a start-up gas has resulted in seal failures right
after a start-up and revenue loss.
It has been seen numerous times where seals that were in clean (Natural Gas
or Propane Refrigeration) services, had failures after a recent start-up (sometimes these were newly replaced seals and other times were the same good
condition seals that were in operation for some time) and immediately needed
to shut down and replace.
BENCHMARKS
This Best Practice has been in use for the last 10 years when it has been observed that many seal failures have occurred in very clean applications right
after a unit turnaround. Many times the seals were not even touched during the
turnaround and were in service for years without any issues, since the gas was
very clean during normal operation. Using this best practice has increased the
reliability of dry gas seals more than it already has been.
SUPPORTING MATERIAL
See supporting material for B.P. 9.1 for more details.
B.P. 9.4: Assure sensing lines for alarm and trip devices are as minimal as
possible.
Whether you are alarming/tripping on primary vent pressure, flow, or DP
the normal values of leakage are always very small. Therefore, when trying to
measure these values it is essential to minimize the pressure drop in the sensing
lines as this pressure drop can be significantly greater than the normal leakage.
456
Dry Gas Seals
Chapter | 9
Instrumentation should be mounted as close to the sensing point as possible,
with zero or minimal hard bends of tubing to the instrument.
L.L. 9.4: Improper sensing line setup for primary vent instrumentation
has lead to inability for the instruments to alarm and trip the machinery
when operating at unsafe leakage levels.
The writer has witnessed a case where a plant had always seen zero leakage
in the primary vent in a certain application. When a site visit was conducted, it
was observed that the sensing line for the pressure transmitter had much more
than 100 ft of equivalent piping!!! It was obvious, then why a value was never
readable in this system and it was really unknown if the seal was in good condition or not. This is a safety and potential environmental issue depending on the
process gas that should not be overlooked.
BENCHMARKS
The writer witnessed the aforementioned situation in 2010 and has been recommending this best practice ever since to assure optimum safety in dry gas seal
applications.
SUPPORTING MATERIAL
See B.P. 9.1 Supporting material for details on dry gas seal systems.
B.P. 9.5: Install a backpressure control valve in the primary vent with an
electronic position indicator.
In B.P. 9.10 of Forsthoffer’s Best Practice Handbook for Rotating Machinery it was explained that a backpressure device be installed in the primary vent
for two distinct reasons as follows:
l
l
By maintaining a pressure in the primary vent (vendors will recommend
generally from 1–2 barg) one can then monitor the condition of the secondary seal easily if the pressure decreases.
A back pressure in the primary vent also puts a decent differential pressure
across the secondary seal which promotes longer secondary seal life.
This Best Practice takes this concept one step further by using a typical
control valve to maintain this back pressure with an electronic position indicator. The position indicator will allow the user to trend the position of this valve.
Simply put, this should be trended from start-up with a new seal and if the valve
is opening then the leakage is increasing from the primary seal (or intermediate laby has worn). However, if the valve is trending closed, it is indicating the
secondary seal is wearing and leaking more.
If valve position is steady, the seal is in very good condition!!!
457
B.P. 9.5
More Best Practices for Rotating Equipment
L.L. 9.5: Inability to accurately monitor dry gas seals has lead to premature
seal replacement that could have been saved for a planned shutdown.
BENCHMARKS
This best practice has been used in the last few years when it was observed in
an application by the writer and provided feedback to the user that has not been
available before.
SUPPORTING MATERIAL
Refer to B.P. 9.1 for details on dry gas seals and their systems.
458
Chapter 10
Construction, Installation,
Commissioning, and
Turnarounds
B.P. 10.1: Conduct machinery pre-turnaround audits to determine scope
of work during the turnaround
The audit (condition assessment) should be conducted for all critical machinery within the unit(s) undergoing the turnaround. It must be based on
the condition of all of the major components within each piece of equipment
(Rotor, Journal Bearings, Thrust Bearings, Seals, and Auxiliaries) and should
contain several data points (ideally continuously trended) from an initial baseline condition.
A condition assessment should be started approximately 2 years before the
turnaround and a final assessment should be made just prior to the cutoff date in
order to properly plan for a full overhaul if needed.
Typically, this assessment will identify machinery that DOES NOT require
work and save a lot of time and manpower for the turnaround.
L.L. 10.1: Inability to properly define turnaround work scope for critical
machinery has often resulted in overhauls that were not required
Many times it has been seen where critical machines were opened up only
to find that the condition of the machine was good to begin with and time was
wasted. During a turnaround any wasted time can be equal to revenue if the unit
could have started up earlier.
BENCHMARKS
FAI has been conducting Site Machinery Pre-Turnaround audits since 1990 and
has saved many facilities from performing overhauls on equipment that was in
good condition.
More Best Practices for Rotating Equipment. http://dx.doi.org/10.1016/B978-0-12-809277-4.00010-3
Copyright © 2017 Elsevier Inc. All rights reserved.
459
B.P. 10.1
More Best Practices for Rotating Equipment
TABLE 10.1.1 Major Machinery Components and Systems
•
•
•
•
•
Rotor
Radial bearing
Thrust bearing
Seal
Auxiliary systems
SUPPORTING MATERIAL
The Major Machinery Components
Table 10.1.1 presents the major component classifications for any type of machinery:
l
l
l
l
l
l
Pumps
Steam Turbines
Compressors
Motors
Gas Turbines
Fans etc.
Regardless of the type of machinery, monitor these components and you will
know the total condition of the machine.
Component Condition Monitoring
As previously stated, component and system functions must first be defined
and the normal values for each component listed. These facts are presented in
Table 10.1.2.
Once the function of each component is defined, each major machinery
component can be monitored as shown in Fig. 10.1.1.
Baseline
Having defined all condition parameters that must be monitored, the next
step in a condition monitoring exercise is to obtain baseline information. It
is important to obtain baseline information as soon as physically possible
TABLE 10.1.2 Component and System Functions
• Define the function of each affected component
• Define the system in which each affected component operates
• List the normal parameters for each affected component and system component
460
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
FIGURE 10.1.1 Component condition monitoring.
TABLE 10.1.3 Base Line Condition
If you don’t know where you started, you don’t know
where you are going!
after start-up of equipment. However, operations should be consulted to
confirm when the unit is operating at rated or lined out conditions. Obtaining baseline information without conferring with operations is not suggested
since misinformation could be obtained and thus lead to erroneous conclusions in predictive maintenance (PDM). Table 10.1.3 states the basics of a
baseline condition.
It is amazing to us how many times baseline conditions are ignored. Please
remember Table 10.1.3 and make it a practice to obtain baseline conditions as
soon as possible after start-up. You can only trend if you have a start point!!!
Trending
Trending is simply the practice of monitoring parameter condition with time.
Trending begins with baseline condition and will continue until equipment
shutdown. In modern day thought, it is often conjectured that trending must be
performed by micro-processors and sophisticated control systems. This is not
necessary! Effective trending can be obtained by periodic manual observation
of equipment or using equipment available to us in the plant, which will include
DCS systems, etc. The important fact is to obtain the baseline and trends of
data on a periodic basis. When trending data, threshold points should also be
defined for each parameter that is trended. This means that when the parameter pre-established value is exceeded action must be taken regarding problem
analysis. Setting threshold values a standard percentage above normal value is
recommended. Typically, values are on the order of 25–50% above baseline
values. However, these values must be defined for each component based on experience. Fig. 10.1.2 presents trending data for a hydrodynamic journal bearing.
461
B.P. 10.1
More Best Practices for Rotating Equipment
FIGURE 10.1.2 Trending data.
All of the parameters noted in Fig. 10.1.2 should be monitored to define the
condition of this journal bearing.
Specific Machinery Component and System Monitoring
Parameters and Their Limits
Following is information concerning what parameters should be monitored
for each major machinery component to determine its condition. In addition,
typical limits are noted for each component. Note that these are typical alarm
limits and are not intended to be the point of concern. If component condition
is trended properly, the idea is to investigate the issue long before these typical alarm limits are reached as that is when it is usually too late and work is
required.
The Rotor
Rotor condition defines the performance condition (energy and efficiency) of
the machine. Table 10.1.4 presents this value for a pump.
Radial Bearings
Tables 10.1.5 and 10.1.6 present the facts concerning anti-friction and hydrodynamic (sleeve) radial or journal bearing condition monitoring.
462
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
TABLE 10.1.4 Pump Performance Monitoring
1. Take value at minimum flow (shut off discharge valve)
2. Measure:
• P1
• P2
• Driver BHP
• Specific gravity
3. Calculate:
ft.-lb f ∆P × 2.311
=
lbm
S.G.
A.
Head Produced
B.
Pump Efficiency (%) =
hd × gpm × S.G.
3960 × bhp
4. Compare to previous value if > −10% perform maintenance
TABLE 10.1.5 Condition Monitoring Parameters and Their Alarm Limits,
Journal Bearing (Anti-Friction)
Parameter
Limits
1. Bearing Housing Vibration (Peak)
0.4 in./s (10 mm/s)
2. Bearing Housing Temperature
185°F (85°C)
3. Lube Oil Viscosity
Off Spec 50%
4. Lube Oil Particle Size
• Non Metallic
• Metallic
25 µm
Any Magnetic Particle In The Sump
5. Lube Oil Water Content
Below 200 ppm
TABLE 10.1.6 Condition Monitoring Parameters and Their Alarm Limits,
Journal Bearing (Hydrodynamic)
Parameter
Limits
1. Radial Vibration (Peak To Peak)
2.5 Mils (60 µm)
2. Bearing Pad Temperature
220°F (108°C)
3. Radial Shaft Position
a
>30° Change and/or 30% Position Change
4. Lube Oil Supply Temperature
140°F (60°C)
5. Lube Oil Drain Temperature
190°F (90°C) Off
6. Lube Oil Viscosity
Spec 50% > 25
7. Lube Oil Particle Size
Micrometers Below
8. Lube Oil Water Content
200 ppm
a
Except for gearboxes where greater values are normal from unloaded to loaded.
463
B.P. 10.1
More Best Practices for Rotating Equipment
TABLE 10.1.7 Condition Monitoring Parameters and Their Alarm Limits,
Thrust Bearing (Anti-Friction)
Parameter
Limits
1. Bearing Housing Vibration (Peak)
• Radial
• Axial
0.4 in./s (10 mm/s)
0.3 in./s (1 mm/s)
2. Bearing Housing Temperature
185°F (85°C)
3. Lube Oil Viscosity
Off Spec 50%
4. Lube Oil Particle Size
• Non Metallic
• Metallic
>25 µm
Any Magnetic Particles With Sump
5. Lube Oil Water Content
Below 200 ppm
TABLE 10.1.8 Condition Monitoring Parameters and Their Alarm Limits,
Thrust Bearing (Hydrodynamic)
Parameter
Limits
a
1. Axial Displacement
>15–20 mils (0.4–0.5 mm)
2. Thrust Pad Temperature
220°F (105°C)
3. Lube Oil Supply Temperature
140°F (60°C)
4. Lube Oil Drain Temperature
190°F (90°C)
5. Lube Oil Viscosity
Off Spec 50%
6. Lube Oil Particle Size
>25 µm
7. Lube Oil Water Content
Below 200 ppm
a
And thrust pad temperatures >220°F (105°C).
Thrust Bearings
Tables 10.1.7 and 10.1.8 show condition parameters and their limits for antifriction and hydrodynamic thrust bearings.
Seals
Table 10.1.9 presents condition parameters and their limits for a pump liquid
mechanical seal.
Auxiliary Systems
Condition monitoring parameters and their alarm limits are defined in
Table 10.1.10 for Lube oil systems.
464
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
TABLE 10.1.9 Condition Monitoring Parameters and Their Alarm Limits,
Pump Liquid Mechanical Seal
Parameter
Limits
1. Stuffing Box Pressure
>50 psig (350 kPa) above the fluid vapor
pressure
2. Temperature drop across orifice
Strainer, or cyclone sep.
Should be negligible unless plugged
3. Temperature drop across cooler
<30°F for Plan 23, >30°F for Plan 21.
Temp. should be less than 160°F for water
4. Temperature drop across reservoir
If zero, no circulation is occurring
TABLE 10.1.10 Condition Monitoring Parameters and Their Alarm Limits,
Lube Oil Systems
Parameter
Limits
1. Oil Viscosity
Off Spec 50%
2. Lube Oil Water Content
Below 200 ppm
3. Auxiliary Oil Pump Operating Yes/No
Operating
4. Bypass Valve Position (P.D. Pumps)
Change > 20%
5. Temperature Control Valve Position
Closed, Supply Temperature > 130 (55°C)
6. Filter ∆P
>25 psid (170 kPag)
7. Lube Oil Supply Valve Position
Change > ±20%
Predictive Maintenance Techniques
Now that the component condition monitoring parameters and their limits
have been presented, PDM techniques must be used if typical condition limits
are exceeded. This action will assure that we minimize site-troubleshooting
exercises.
One final recommendation is presented in Table 10.1.11.
Tables 10.1.12–10.1.14 present condition monitoring parameters and limits
for dynamic compressor performance, liquid seals, and seal oil systems.
See Tables 10.1.15–10.1.20 which contain typical parameters to be monitored, with embedded equations, for calculating performance for a Steam Turbine Driven Compressor Train.
465
B.P. 10.1
More Best Practices for Rotating Equipment
TABLE 10.1.11 Obtain and Maintain Management Support by ...
1. Clearly Stating Impact Of Problem On Plant Profit
2. Prepare A Brief Statement Of:
• Problem
• Impact On Plant
• Action Plan
3. Be Confident!
4. Be Professional!
5. Provide Timely Update
TABLE 10.1.12 Compressor Performance Condition Monitoring
1. Calibrated: Pressure And Temperature Gauges And Flow Meter
2. Know Gas Analysis And Calculate k, Z, MW
3. Perform As Close To Rated Speed And Flow As Possible
4. Relationships:
A.
(T )
LN 2
N −1
(T1)
=
(P2 )
N
LN
(P1)
B.
EFFICIENCYpoly =
k −1
k
n −1
n
n −1


 Ft − lb f  1545
n
  P2  n − 1
=
×
×
×
×
T
Z
C. HEAD poly = 
1
avg
  P1 

 Lbm  MW
n −1


5. Compare To Previous Value. If Decreasing Trend Exists Greater Than 10%, Inspect At
First Opportunity.
TABLE 10.1.13 Condition Monitoring Parameters and Their Alarm Limits,
Compressor Liquid Seal
Parameter
Items
1. Gas Side Seal Oil/Gas ∆P
• Bushing
• Mechanical Contact
<12 ft. (3.5 m)
<20 Psi (140 kPa)
2. Atmospheric Bushing Oil Drain Temperature
200°F (95°C)
a
3. Seal Oil Valve Position
>25% Position Change
4. Gas Side Seal Oil Leakage
>20 gpd Per Seal
Return valve = −25%.
Note this assumes compressor reference gas pressure stays constant.
a
Supply valve = +25%.
466
Chapter | 10
Construction, Installation, Commissioning, and Turnarounds
TABLE 10.1.14 Condition Monitoring Parameters and Their Alarm Limits,
Compressor Liquid Seal Oil Systems
Parameter
Limits
1. Oil Viscosity
Off Spec 50%
2. Oil Flash Point
Below 200°F (100°C)
3. Auxiliary Oil Pump Operating Yes/No
Operating
4. Bypass Valve Position (P.D. Pumps)
Change > 20%
5. Temperature Control Valve Position
Closed, Supply Temperature 130°F (55°C)
6. Filter ∆P
25 psid (170 kPag)
7. Seal Oil Valve Position
Change > 20% Open (Supply)
> 20% Closed (Return)
8. Seal Oil Drainer Condition
(Proper Operation)
• Constant Level (Yes/No)
Level Should Be Observed
• Observed Level (Yes/No)
Level Should Not Be Constant
• Time Between Drains
Approximately 1 h (Depends On Drainer
Volume)
TABLE 10.1.15 Compressor Performance Monitoring
Item/Section #
Date/Time
Given
M.W.
P1 (PSIA)
T1 (°F)
P2 (PSIA)
T2 (°F)
K
Z
Inlet Flow (ACFM)
N (RPM)
Calculate
(K − 1)/K
—
—
—
—
(n − 1)/n
—
—
—
—
3
Gas Density (lbs/ft. )
—
—
—
—
Mass Flow (lbs/min)
—
—
—
—
(Continued)
467
B.P. 10.1
More Best Practices for Rotating Equipment
TABLE 10.1.15 Compressor Performance Monitoring (cont.)
GHP
—
—
—
—
Poly Hd (ft. − lbf/lbm)
—
—
—
—
Poly Eff’y
—
—
—
—
Does Compressor Need Maintenance?
TABLE 10.1.16 Steam Turbine Performance
Extraction/
Condensing
Backpressure
Theoretical Steam Rate (lb/HPh, TSR)
—
—
HP for HP Section (or HP of BP Turbine)
—
Type
Item #
Date/Time
Given
P1 (psi)
T1 (°F)
P2 (psi)
T2 (°F)
Speed (RPM)
Flow Rate (lb/h)
Total Power at Coupling (HP)a
Determine
b
h1 (BTU/h, From Mollier Diagram)
b
h2 Isentropic (BTU/lb, From Mollier Diagram)
HP for LP Section
—
—
—
b
h2 Actual (BTU/lb, From Mollier Diagram)
—
Actual Steam Rate (lb/HPh, ASR)
—
—
—
Steam Turbine Efficiency
—
—
—
% Moisture LP Section
Actions Required
Do not fill shaded cells.
a
Total Power at Coupling can be determined via torque meter or from Compressor Total Gas Power
plus mechanical losses for bearings and seals.
b
Note that values for h1, h2 isentropic, and h2 actual (HP Case only) are obtained from a Mollier
Diagram contained in this section or by using steam tables.
468
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
TABLE 10.1.17 Component Condition Monitoring Worksheet
Item #:
Date/Time:
Journ. Brgs.
Compressor DE Horiz. Vibes (mils)
Compressor DE Vert. Vibes (mils)
Compressor DE Pad Temp (°F)
Compressor DE Pad Temp (°F)
Compressor NDE Horiz. Vibes (mils)
Compressor NDE Vert. Vibes (mils)
Compressor NDE Pad Temp (°F)
Compressor NDE Pad Temp (°F)
Steam Turbine DE Horiz. Vibes (mils)
Steam Turbine DE Vert. Vibes (mils)
Steam Turbine DE Pad Temp. (°F)
Steam Turbine DE Pad Temp. (°F)
Steam Turbine NDE Horiz. Vibes (mils)
Steam Turbine NDE Vert. Vibes (mils)
Steam Turbine NDE Pad Temp. (°F)
Steam Turbine NDE Pad Temp. (°F)
Thrust Brgs.
Compressor displ.
Compressor displ.
Compressor Active Pad Temp. (°F)
Compressor Active Pad Temp. (°F)
Compressor Inactive Pad Temp. (°F)
Compressor Inactive Pad Temp. (°F)
Balance Line Diff. P (psid)
Steam Turbine displ.
Steam Turbine displ.
Steam Turbine Active Pad Temp. (°F)
Steam Turbine Active Pad Temp. (°F)
Steam Turbine Inactive Pad Temp. (°F)
Steam Turbine Inactive Pad Temp. (°F)
469
B.P. 10.1
More Best Practices for Rotating Equipment
TABLE 10.1.18 Component Condition Monitoring Worksheet
Item #:
Date/Time
Journ. Brgs.
Steam Turbine DE Horiz. Vibes (mils)
Steam Turbine DE Vert. Vibes (mils)
Steam Turbine DE Pad Temp. (°F)
Steam Turbine DE Pad Temp. (°F)
Steam Turbine NDE Horiz. Vibes (mils)
Steam Turbine NDE Vert. Vibes (mils)
Steam Turbine NDE Pad Temp. (°F)
Steam Turbine NDE Pad Temp. (°F)
Thrust Brgs.
Steam Turbine displ. (mils)
Steam Turbine displ. (mils)
Steam Turbine Active Pad Temp. (°F)
Steam Turbine Active Pad Temp. (°F)
Steam Turbine Inactive Pad Temp. (°F)
Steam Turbine Inactive Pad Temp. (°F)
DE Gland Condenser Pressure (in mmHg)
NDE Gland Condenser Pressure (in mmHg)
TABLE 10.1.19 L.O. and SO Syst.
Equip #:
System Name:
Component/Item
Oil Reservoir
Level
Oil Temp. (°F)
Air in Oil? (Y/N)
Gas in Oil?
Oil Sample?
Other
Other
470
Date:
Time:
Specified value
Actual value
Comments
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
TABLE 10.1.19 L.O. and SO Syst. (cont.)
Equip #:
System Name:
Component/Item
Date:
Time:
Specified value
Actual value
Comments
Pumps
Aux. Pump Operating?
P2 (psig)
Suction Noise?
Suction Filter ∆P (psid)
Vibration (in./s)
Brg. Bracket Temp. (°F)
Other
Other
Couplings
Noise?
Strobe Findings
Other
Other
Turbine Driver
Operating Speed (RPM)
Trip Speed Setpoint (RPM)
Vibration (in./s)
Brg. Bracket Temp. (°F)
Gov. Hunting?
Trip Lever Condition
Gov. Oil Condition
Other
Other
Motor Driver
Operating?
Vibration (in./s)
Brg. Bracket Temp. (°F)
Axial Shaft Movement
(in./s)
Fan Noise?
Other
(Continued)
471
B.P. 10.1
More Best Practices for Rotating Equipment
TABLE 10.1.19 L.O. and SO Syst. (cont.)
Equip #:
System Name:
Component/Item
Other
Relief Valves
Passing?
Set Pressure (psig)
Pump P2 Press. (psig)
Other
Other
Check Valves
Aux. Pump Turning Backwards?
Noise?
Other
Other
Back Pressure Valve
% Open
Stable?
Valve Noise?
Set Pressure (psig)
Maintained Pressure
(psig)
Other
Other
Transfer Valves
One Bank Operating?
Noise?
Other
Other
Coolers
∆T Oil
CW Valve Pos.
Cooler Operating?
Vent Valves Open?
Other
472
Date:
Time:
Specified value
Actual value
Comments
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
TABLE 10.1.19 L.O. and SO Syst. (cont.)
Equip #:
System Name:
Component/Item
Date:
Time:
Specified value
Actual value
Comments
Other
TCV’s
% Open
Set Temp. (°F)
Stable?
Actual Temp. (°F)
Other
Other
Filters
∆P (psid)
Vent Valves Open?
Last Filter Change
Other
Other
Accumulators
Pre-charged Pressure
(psig)
Last PM Date
Other
Other
Lube Oil PCV
% Open
Set Pressure (psig)
Actual Pressure (psig)
Stable?
Other
Other
Control Oil PCV
% Open
Set Pressure (psig)
Acual Pressure (psig)
Stable?
(Continued)
473
B.P. 10.1
More Best Practices for Rotating Equipment
TABLE 10.1.19 L.O. and SO Syst. (cont.)
Equip #:
System Name:
Component/Item
Date:
Time:
Specified value
Other
Other
Lube Oil Rundown Tank (or Emerg. Pump)
Pump or Tank?
Pump Operating?
Tank Overflow
Other
Other
Lube Oil Supply Lines
Leaks?
Noise?
Vibration (in./s)
Other
Other
Seal Oil Supply Valve
Position
Stable?
Seal Oil Supply Pressure
Seal Oil OH Tank
Level
S.O. Differential Pressure
Drainers
Level
Flow Through Vent
Orifice?
474
Actual value
Comments
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
TABLE 10.1.20 DGS System Monitoring
Date:
Time:
Equipment # :
Item
Observations
Comments
Primary Gas Filter DP (psid)
Primary Gas Supply DP
Prim. Vent Flow suct. (scfm)
Prim. Vent Flow Disch. (scfm)
Sec. Gas Filter DP (psid)
Sec. Gas Supply Press. (psig)
Sec. Supply Flow Suct. (scfm)
Sec. Supply Flow Disch. (scfm)
Sec. Vent Suct. Flow (scfm)
Sec. Vent Disch. Flow (scfm)
Oil in Sec. Drain Suct.?
Oil in Sec. Drain Disch.?
Seperation Gas Filter DP (psid)
Balance Line DP (psid)
B.P. 10.2: Conduct site specific training for all disciplines involved with
machinery to better understand how the major components are supposed
to work and the effect that the process and all other related systems have
on the reliability of these components
This training can be in-house or through an outside training company, but
must be specific to the actual equipment on site. By having all the major disciplines involved, it will promote comradery between everyone and less fingerpointing.
The class should be formed in a manner where topics are discussed, then
practiced in the field on actual equipment in order to immediately utilize and
hone the newly found skills.
If using an outside company to conduct this training, be sure to have an
experienced person from the sponsored department to review the agenda and to
make sure it includes the topics desired.
L.L. 10.2: Inability to properly train all disciplines on the importance of
the process and system’s effect on machinery and components will lead to
reoccurring machinery failures since the root cause may not be identified
the first time
475
B.P. 10.2
More Best Practices for Rotating Equipment
BENCHMARKS
This best practice has been used since the mid-1980s to provide the plant
personnel with the proper tools to understand the basics of how the major
components of their machinery are designed and supposed to operate. It has
been seen in plants that incorporate this kind of training that “Firefighting” is
minimal if not non-existent since all personnel understand the importance of
why certain parameters need to be monitored and usually identify issues prior
to failures.
SUPPORTING MATERIAL
Following is a typical agenda for a fundamentals type course on Rotating Machinery that incorporates all the aspects mentioned in this BP.
SITE ROTATING EQUIPMENT FUNCTION OVERVIEW WORKSHOP
SESSION MODULE
DAY 1
1
DESCRIPTION
COURSE INTRODUCTION
• Workshop Objectives
• Instructor Introductions
• Attendee Introductions
• Workshop Agenda
2
ROTATING EQUIPMENT OVERVIEW
• Definition of Rotating Equip.
• Classifications of types
• Component similarities
• The Equipment “Train”
• Important fundamentals
3–6
TYPES OF PUMPS ON SITE
• Single Stage overhung
• Between Bearing
• Horizontal Double Suction
• Horizontal Multistage
• Vertical
• Single stage low speed
• Single stage high speed
• Multistage
7, 8
EFFECT OF THE PROCESS ON POSITIVE
DISPLACEMENT AND DYNAMIC
EQUIPNMENT AND COMPONENT
CONDITION MONITORING
• Concepts (Head, Efficiency, Power)
• Head Required
• Head Produced
• Concept of Component Condition
Monitoring (CCM)
• CCM component parameters/limits
476
CHAPTER
TAB
Chapter | 10
Construction, Installation, Commissioning, and Turnarounds
SITE ROTATING EQUIPMENT FUNCTION OVERVIEW WORKSHOP
SESSION MODULE
DAY 2
9
DESCRIPTION
THE CONCEPT OF PUMP HEAD
• Head = Energy
• Head Required by the Process
• Head Produced by the Pump
11, 12
HYDRAULIC DISTURBANCES
• Maintaining a liquid
• Types and position on the curve
• EROE (Operating in the
good region)
• Detecting disturbances
• Preventing disturbances
13, 14
PUMP MECHANICAL DESIGN – VOLUTES,
WEAR RINGS, IMPELLERS BEARINGS AND
BALANCE DRUMS
PUMP MECHANICAL SEALS
SITE VISIT #1 PUMP CCM
• Two selected Pumps
• CCM – all 5 components
20, 21
CLASSROOM DISCUSSION- VISIT #1
• Review of CCM Results
• Abnormalities
• Corrective Action
22
COMPRESSOR TYPES & APPLICATIONS
• Barrel Compressors
• Horizontal Split Compressors
• Integral Gear Compressors
23, 24
THE CONCEPT OF COMPRESSOR HEAD &
PERF. CURVE EXAMPLES
• Definition and similarity
to Pumps
• Head Required by the Process
• Head Produced by the Impellers
• Never evaluate using pressures!
DAY 4
25
TAB
PUMP PERFORMANCE CURVES AND DATA
SHEETS
• Single Stage Overhung
• Horizontal Multistage
• Vertical Multistage
10
15, 16
DAY 3
17–19
CHAPTER
STALL, SURGE AND STONEWALL
• The Cause of Surge
• Surge Facts
• Limits of the Curve
• Stonewall (Choke)
477
B.P. 10.2
More Best Practices for Rotating Equipment
SITE ROTATING EQUIPMENT FUNCTION OVERVIEW WORKSHOP
SESSION MODULE
DESCRIPTION
26
DYNAMIC COMPRESSOR MECHANICAL
DESIGN OVERVIEW
• Review of component functions
• Gas wetted components
• Mechanical components
27
COMPRESSOR RADIAL BEARING DESIGN
• Hydrodynamic Bearing Function
• Types
• Condition Monitoring
• Vibration Principles
28
COMPRESSOR THRUST BEARING DESIGN
& THRUST BALANCE
• Thrust Bearing Function
• Impeller Thrust Forces
• Thrust Balance
• Condition Monitoring
29–30
DRY GAS SEAL PRINCIPLES
• Comparison to Pump Seal
• Function
• Tandem Seals
• DGS System
• Condition Monitoring
TYPES OF STEAM TURBINES ON SITE
• Function
• Single Stage/Single Valve
• Multistage/Multivalve
31
32
DAY 5
33, 34
STEAM TURBINE PERFORMANCE
CHARACTERISTICS
• The Mollier Diagram
• Theoretical Steam Rate
• Actual Steam Rate
• Efficiency
• First Stage Pressure
LUBE OIL SYSTEM OVERVIEW
• Function
• Types
• Component Functions and Monitoring
35–37
SITE VISIT #2 COMPRESSOR TRAIN CCM
• Mug Compressor CCM
• Mug Steam Turbine CCM
• DGS & Oil System CCM
38–40
SITE VISIT #2 CLASSROOM DISCUSSION
• Review of CCM Results
• Abnormalities
• Corrective Action
478
CHAPTER
TAB
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
B.P. 10.3: Review machinery instruction manuals prior to shipment from
vendor
Today this can be done electronically, and what this does allows the user
to have at least a draft of the instruction manual prior to the arrival of the
equipment.
Key items that need to be reviewed and included are:
l
l
l
l
Are the proper units on data sheets and curves?
Head versus Flow performance curves to be provided for centrifugal compressors.
All specific component maintenance and operating instructions to be
included.
Any specific vendor procedures to be included (e.g., wash water or oil operation, use of bundle removing tooling, and so on).
L.L. 10.3: Inadequate vendor instruction manuals have resulted in longer
mean time to repair equipment since communication with vendor for specific details is required
BENCHMARKS
This best practice has been used since 1990 and has resulted in the ability to better monitor and maintain the equipment when all required details are included
in the manual.
B.P. 10.4: Spare critical machinery rotor storage guidelines
The following guidelines should be followed at minimum to assure proper
storage of spare critical machinery rotors:
l
l
l
l
Stored in a container in vertical position to prevent rotor bow.
Stored inside in a controlled environment.
Should be under N2 pressure with a low pressure alarm. If nitrogen bottles
are used there should be a backup bottle with a changeover valve once the
first bottle is spent and alarm comes on.
Preservation fluid should be such that it is adequate to preserve but not too
time consuming to remove. If N2 pressure is maintained and monitored,
there will be no problems.
L.L. 10.4: Improper storage of spare critical machinery rotors have resulted in severe turnaround delays when the spare rotor was completely
corroded in critical areas and unable to use
BENCHMARKS
This has been used since the mid-1990s and has resulted in no surprises when it
is time to change out the rotor of a critical machine.
479
B.P. 10.4
More Best Practices for Rotating Equipment
FIGURE 10.4.1 General site considerations.
SUPPORTING MATERIAL
Regardless of the quality of design and manufacture, regardless of a successful
test and efficient shipment, installation will determine the amount of maintenance required and the resulting revenue of the process unit. Fig. 10.4.1 shows
the general site considerations that are required for a successful field installation. Each one of these items will be covered in following sections.
Site Procedures
The importance of site installation procedures cannot be overemphasized.
Fig. 10.4.2 shows the most commonly required site installation procedures.
It must be remembered that the objectives of the construction contractor
and of the end user are identical in terms of profit. However, they are dissimilar
in means to achieve their common objectives. The contractor’s objective is to
construct a safe and reliable process unit within the budget and on time. This objective is opposed to the end user’s objective, who must operate the process unit
for 30 years or more at maximum profit and thus requires maximum reliability
of the installed equipment. The only leverage that the end user has in meeting
his objectives is to require practical, proven site installation procedures that will
result in the most reliable, safe, and cost effective installation of his equipment.
Frequently the contactor will rely on the equipment vendor to provide most of
the site procedures. It is strongly recommended that the end user, early in the
project, require approved procedures for every major site installation milestone.
These procedures include, but are not limited to:
l
l
l
Equipment preservation
Equipment installation
Grouting
480
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
FIGURE 10.4.2 Site installation procedures.
l
l
l
l
Alignment
Flushing
Functional checks
Initial run in of equipment
It must be mentioned that preservation procedures are often ignored early in
a construction project and become written and implemented too late to effectively prevent equipment deterioration due to corrosion. Again, it is required that
procedures be written and approved well in advance at the start of construction.
A specific “initial run in procedure” for each major piece of rotating equipment should be reviewed by the end user well in advance. It is recommended
that the end user review these procedures during the shop test with vendor field
service engineers to assure that the Run In procedure is in accordance with vendor and plant best practices. An example of not performing a pre-Run In review
and its consequences concerned a large high pressure condensing steam turbine
cold start-up time versus speed curve. Since the turbine was designed for an automated sequenced start, the cold start-up curve was programmed into the PLC
without detailed review by the end user. The curve was simulated and did not take
the specific parameters of the application into account. The result was a severe
rub that damaged rotor and stationary internals and resulted in over a 40-day
plant start-up delay. The daily profit of this plant was approximately 0.25 $MM.
Construction Special Tools
Most of the equipment that is installed will be custom designed. Therefore,
special tools will be shipped with equipment that can only be used for that
481
B.P. 10.4
More Best Practices for Rotating Equipment
particular item. Consequently, these tools must be listed and stored in a proper
location that will allow maintenance personnel to easily locate these tools when
required. Many of the tools, such as hydraulic jacks, special mounting devices,
and so on also will require preservation during storage to prevent corrosion.
Many times this requirement is overlooked. Be aware!
An example of not properly storing special tools and spare parts is a high
humidity, tropical island installation where the spare rotor and coupling hydraulic mounting tools were to be used for a turnaround. They were not inspected prior to the turnaround and the result was that the spare rotor could
not be installed due to excessive corrosion. The coupling hydraulic mounting
adapter had to be replaced with a new adapter due to excessive corrosion. It
should be noted that this equipment was stored in a sealed container with a
nitrogen purge but unfortunately the seal was faulty and the nitrogen purge
pressure was not monitored.
Installation Manuals
Like site procedures, installation manuals frequently arrive after they are first
required. The installation manual will contain valuable information pertaining
to receipt of equipment, preservation, interim storage, and of course, installation. It is recommended that the end user again require that manuals be approved and received well in advance of the start of any installation activity.
In an effort to assure that the instruction manual contains information that
is accurate and specific to the job, we have written into the job specification
that the instruction book shall be completely reviewed at the time of the shop
test and shipped with the unit. Almost every such review has uncovered incorrect information and/or general information that is not specific to the particular
project that would have resulted in confusion during disassembly/assembly of
the equipment leading to possible equipment reliability problems.
Spare Parts
The end user must require the contractor to be responsible for the proper storage
of spare parts on site and most importantly, the receipt of all required start-up
spares, operating spares, and capital spare parts well in advance of the start of
construction activity. Many times spare parts are required prior to initial operation of equipment since components are broken during shipment.
It has been our experience that many of the spare rotors for major, unspared
equipment are not properly inspected and maintained upon receipt from the
vendor. A specific rotor container inspection procedure should be written by
the contractor and approved by the vendor and end user to assure that all rotor
preservation will be maintained from initial receipt date on site. There have
been many cases where the rotor containers have never been inspected prior to
the intended use date and could not be used and had to be returned to the vendor
for rework.
482
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
On Site Storage
Prior to the arrival of any equipment on site the contractor should review the
manufacturer’s requirements and provide extended on site storage facilities that
meet or exceed those requirements. Additionally, the contractor should provide
the required preservation compounds for all of the equipment as needed.
When selecting the preservative compounds, special care should be given
to the selection of compounds for the specific site environment and should
be based on local experience. In humid, seacoast, and offshore environments, special care should be given to assure the preservative compounds
can resist high moisture/salt environments. In addition, components should
be checked frequently to confirm that the compounds are providing the required protection.
B.P. 10.5: Assure vendor for epoxy grout is on site for initial pours and
provides training
Only epoxy grout, when properly applied, will assure a machinery baseplate
support that will last for the life of the installed equipment.
By having an experienced vendor rep. available for the initial pours of the
epoxy grout, they will assure that the proper procedures are being followed. At
that time it would be a good idea for them to conduct a site training for a day
or so on the procedure used and any special requirements for the specific type
of epoxy used.
L.L. 10.5: Failure to specify the use of epoxy grout for all machinery installations, have an approved grout procedure in place, and an experienced epoxy
grout contractor have caused significant project delays and foundations that
required re-grouting before or during the first scheduled plant turnaround
BENCHMARKS
This best practice has been used since the mid-1980s when significant epoxy grouting issues were experienced. The final solution, after a project delay was to require
that the epoxy grout vendor representative to come to site and conduct epoxy grout
training for the grout contractor who did not have any experience in epoxy grout
installation in the geographical area in which the plant was located. Since that time,
we have required that the site grout procedure be reviewed and approved prior to
start of construction and that the contractor’s experience be confirmed.
SUPPORTING MATERIAL
Grout (General)
The grouting plays an important role in the availability of equipment. Improper grout type and application of the grout has caused many an unscheduled
483
B.P. 10.5
More Best Practices for Rotating Equipment
TABLE 10.5.1 General Grout Considerations
1. Approved grouting procedure
2. Epoxy grout for:
j Greater than 75 kW (100 hp)
j All axial, centrifugal, or reciprocating compressor trains
3. Special environmental conditions
j Temperatures >50°C (140°F)
4. Proper surface preparation
j Clean
j Chipped
j Water free (for epoxy grout)
j Grease anchor bolts, jackbolts, chocks
shutdown in the field. Table 10.5.1 presents some general grout considerations
that have been proven through many long, hard construction projects.
Most important is an approved grouting procedure. Not a simple procedure
that states the type of grout and how much will be used, but a detailed procedure specifying the equipment used for proper grout pours, the forms, the form
preparation, the details concerning depth of pour, specifications for grout, etc. It
has been our experience that contractors are not experienced in proper grouting
procedures. Remember, the installation phase is only a short period in the life
of equipment. The decisions made during grouting will affect the equipment for
its lifetime.
Epoxy grout is usually required for equipment greater than 75 kW (100 hp)
and all reciprocating types of rotating equipment. Although epoxy grout is much
more expensive than conventional grouts, it certainly pays out in the long run
since it is impervious to oil and resists cracking. In the application of any grouts,
ambient conditions are very important. Be sure that the site grouting procedure
takes the local ambient conditions into account.
Like most jobs, proper preparation significantly affects the quality of the
finished product. Clean, chipped, water free (for epoxy grout) foundations are a
necessity. Also anchor bolts, jack bolts, and chocks should be greased for ease
of operation once the grout starts to cure.
Epoxy Grout
Epoxy grout is clearly the grout of choice for critical (unspared) equipment
installation since it lasts the longest and is impervious to most external sources.
Table 10.5.2 presents some epoxy grout considerations.
It is important that epoxy grout be poured in accordance with the grout manufacturer’s recommendations. Most contractors need experience in epoxy grout
installation. In fact, it has been our experience that an on site demonstration of
epoxy grout by the epoxy grout manufacturer is a worth while expenditure and
484
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
TABLE 10.5.2 Epoxy Grout Considerations
1. In accordance with grout manufacturers procedures
2. Wax or grease all forms
3. Limit thickness of pour to 10 cm (4 in.)
4. Fill bolt sleeves or pockets
5. Check for voids. Fill with epoxy pressure grout
6. Seal grout holes
saves countless repours and project delays. Some other epoxy grout considerations are:
l
l
l
l
l
l
Wax or grease all forms
Limit thickness of pour to approximately 10 cm (4 in.)
Assure that proper mixing and pouring tools are available
Fill bolt sleeves or pockets to assure that grout does not spill into these areas
Check for voids and fill with epoxy pressure grout when required
Seal grout holes in metal base plates
Non Shrink or Cementous Grout
Frequently for cost considerations, contractors will attempt to use non shrink
or cementous grout on large pieces of equipment. In certain instances, this is
acceptable as in the case of large oil console foundations, which are usually
installed with cementous grout for reasons of mass. This action solidifies the
console base and significantly minimizes pipe and component vibration.
It should be mentioned that some types of non shrink grout incorporate
metal filings. It has been found that in some instances these filings will corrode
with time and cause separation of grout from the foundation. Prior to application of any grout, a proper procedure and details of the grout must be defined.
In the event of any doubts ask the original equipment manufacturer regarding
his considerations.
Foundations
The installation of properly designed and constructed foundations play an important part in the long term availability of equipment. This section will cover
the major aspects of sound foundations.
General Considerations
After the proper civil work is done, and the foundation is designed in accordance with specifications, there are certain general considerations required.
These are shown in Fig. 10.5.1.
485
B.P. 10.5
More Best Practices for Rotating Equipment
FIGURE 10.5.1 Foundations.
Foundations must be rough enough to allow grout to adhere. The elevation
of the top surface should allow at least 1 in. of grout under the base plates or
sole plates. When machinery is mounted directly on the foundation, sole plates
must be provided. It is wise to epoxy grout sole plates to facilitate easy removal
and installation of equipment during maintenance. Sole plates must be leveled
within themselves and in all other planes.
Foundation Bolts
Each equipment manufacturer has foundation bolt requirements. It is a good idea
to review the contractor’s foundation bolt arrangements prior to the start of any
equipment installation and assure that the contractor’s procedure, types of bolts,
and bolt arrangement, meet or exceed the equipment vendor’s requirements. Many
a project has been delayed by not incorporating this requirement. Fig. 10.5.2 shows
three typical installation arrangements for anchor bolt installations.
A case history for the installation of a large reciprocating compressor in a
refinery shows how poor planning can cause a significant construction delay. The
foundation re-bar pattern was not coordinated with the foundation bolt pattern for
the crankcase and crossheads. After the foundation was set, with the foundation
bolts in place, it was discovered that the bolt locations had moved from the original positions and that the crankcase could not be positioned over the foundation
bolts. The re-bar had interfered with the foundation bolts causing them not to
be correctly positioned. The result was complete foundation rework to correctly
position the foundation bolts that resulted in a delay of 1 month to the construction schedule. At that time the lost revenue was approximately 1 $MM per day.
One last word regarding foundation bolts. It is easy to mislocate the bolt locations relative to the machinery base plate holes. Before bolt holes are randomly elongated to facilitate misplaced location bolts, all facts should be discussed
486
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
FIGURE 10.5.2 Anchor bolt installations. (Courtesy of ME Crane Consultant)
with both the contractor and the vendor of the equipment. Irresponsible action
regarding elongation of bolt holes has caused machinery problems. Fabricated
foundation bolts on which welding is used in the fabricated assembly must be
stress relieved after welding.
Leveling
Fig. 10.5.3 presents the basics of leveling of equipment.
Equipment must be leveled within a tolerance of 0.05 mm/m and confirmed
with a calibrated engineer’s level. Any special leveling instructions given by the
vendor must be followed. In the case of reciprocating equipment, it is important
that shims straddle hold down bolts. When jacking screws are used for leveling
FIGURE 10.5.3 The basics of leveling of equipment.
487
B.P. 10.7
More Best Practices for Rotating Equipment
equipment it is not necessary to remove them after grouting but they must be
backed off at least two turns and the hold down bolts must be retorqued to their
correct value after grout has adequately cured. There should be a minimum of
one jacking bolt for each hold down bolt.
B.P. 10.6: Bring key millwrights and operators to factory acceptance test
This is very important since they are the personnel who will be maintaining
and operating the equipment for many, many years to come.
It is important for the millwrights and operators to observe the following:
l
l
l
The layout of the equipment so they can be familiar in the field.
Opening of casing and rotor replacement. This is as important for the operators as it is for the millwrights since they will see the inside of the equipment
and will get an idea of how long a typical overhaul would take in a controlled
environment like the vendors shop.
Maintenance procedures used by the vendors and should note any variation
from what is used in the field. If any tools are used that look unfamiliar, ask
and find out how to acquire.
L.L. 10.6: Unfamiliarity with the equipment by plant personnel has resulted in delays during unit turnarounds
BENCHMARKS
This best practice has been in use since the mid-1990s and has helped in minimizing plant turnaround duration.
B.P. 10.7: Have vendor service representative available at factory acceptance test
This is a good opportunity to meet the service rep. and evaluate their capabilities. If there is any concern at this stage, it can be corresponded to the vendor
and another service representative can be assigned to you.
It is also very important to review the resume (CV) of the service representative prior to meeting them so you can see the background and experience they
have. This usually assures that the person has extensive experience since you
wouldn’t be asking the vendor for a CV to get an inexperienced rep.
L.L. 10.7: Inability to screen the service representative for your equipment can result in inadequate help and delays in start-up
The author had a very recent encounter with service representatives who
mentioned they were working in a plant for an issue on a particular tag #. By
knowing the equipment tag # and what the machine was, the author asked them
a few questions based on the reciprocating compressor they were working on
and it was then stated that neither of the servicemen had worked on that particular type of compressor before.
488
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
BENCHMARKS
This best practice has been used since the late 1980s and has resulted in the
vendor providing quality service representatives when required.
B.P. 10.8: Assure dry gas seal piping from source to the panel (and including the panel) is stainless steel
Dry gas seals are always going to be stainless steel, but if there is any carbon steel piping upstream or from the panel and any moisture content in the
gas, rust can form and cause potential issues in the system and possibly the
compressor.
L.L. 10.8: Failure to assure all dry gas seal piping is stainless steel has resulted in severe fouling of a compressor and delays in initial plant start-up
In brand new installation a gas plant was utilizing its sales gas as the primary
seal supply for a tandem dry gas seal. The seal gas system was specified to be
all of stainless steel piping, but it was observed after start-up of the train that the
efficiency was over 10% below what it was expected to be. The machine was
then opened up and found to be fouled with iron sulfide. It took some time to
finally pin point that the culprit was carbon steel piping in the sales gas supply
that carried over rust and apparently collapsed a filter cartridge.
BENCHMARKS
The writer has used this best practice since 2009 when the mentioned lesson
learned occurred. Even when you specify systems out properly you have to be
sure whatever is supplying that system is up to specs as well.
B.P. 10.9: Perform initial functional testing on auxiliary systems prior to
initial start-up of the train
When you are ready to start up the train it is very important to assure the
auxiliary systems you were provided with, work as specified in transient conditions in the field installation (interconnecting piping different than aux system
factory acceptance test). This should not delay start-up significantly, if at all, but
will reassure that the system will be able to keep the train running if there is a
sudden transient situation (i.e., Main pump turbine tripping).
L.L. 10.9: Failure to functional test auxiliary systems has resulted in unplanned shutdowns when the system did not recover quickly enough during a main pump driver trip
BENCHMARKS
This best practice has been recommended since the mid-1990s and it has resulted in centrifugal compressor train reliability above 99.7%.
489
B.P. 10.9
More Best Practices for Rotating Equipment
SUPPORTING MATERIAL
Functional Testing
Having satisfactorily installed and flushed the auxiliary system, all auxiliary equipment should be functionally tested and all instruments and controls
checked for proper setting prior to operation of equipment. A functional test
outline and procedure is included at the end of this section. We will highlight
the major considerations of the procedure at this time.
It is recommended that the console vendor and/or the equipment purchaser
prepare a detailed field functional test procedure and calibration check form.
The format of this procedure can follow the factory test procedure if it was
acceptable. As a minimum, the auxiliary system, bill of material, and schematic should be thoroughly checked in order to include the calibration and
functional test of each major component in the auxiliary system. That is, components on consoles and up at unit interfaces. A detailed record should be
kept of this functional test procedure. This will help significantly during the
operation of the unit. The functional test procedure should be accomplished
without the critical equipment running initially and then with the auxiliary
system at design operation conditions as closely as possible. The functional
test procedure should first require that all instrumentation is properly calibrated before proceeding. Each specific functional test requirement should
then be performed and results noted. If they do not meet specified limits as
noted, testing should stop and components should be corrected at this point.
Each step should be followed thoroughly to assure each component meets
all requirements. It is recommended that operators assigned to this particular
unit assist in functional testing to familiarize themselves with the operation
of the system. In addition, site training courses should be conducted prior to
functional test to familiarize operators with system’s basic functions. This
training, again, significantly increases understanding of the equipment and
assures unit reliability.
Satisfactory acceptance of a functional test then assures that the unit has
been designed, manufactured, and installed correctly such that all system design
objectives have been obtained and that equipment reliability is optimized.
One remaining factor to be proven is the successful operation of the system
with the critical equipment unit in operation. During initial start-up, it is recommended that the functional test be re-performed with the unit operating. While
this advice may seem dangerous, unless the unit operators are assured that the
subject system has the ability to totally protect critical equipment while operating, auxiliary equipment will never be tested while the unit is in operation.
Remember, critical equipment is designed for 30 years or greater life. The
components that comprise the auxiliary system are many and have characteristics that will change with time. Therefore, reliability of auxiliary systems
can only be maintained if the systems are totally capable of on-line calibration and functional checks. The functional pre-commission procedure should be
490
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
modified to include an on-line periodic functional checking procedure. Such a
procedure is included at the end of this section.
At this point, we can clearly see that the major determination of continued
equipment reliability rests with the operation, calibration, and maintenance of
the equipment. In order to assume maximum continued auxiliary equipment reliability, periodic on-line functional checks and calibrations must be performed.
How can this be done? The only way is by convincing plant operations of the
safety of performing these checks and the increased reliability produced. This
can be reinforced during pre-commissioning by including operators in functional testing checks and on-site training sessions to show the function of the
system. A site training course modified for the specific equipment would prove
immensely valuable in achieving those results.
Only by involving unit operators in the pre-start check ups can it be hoped to
establish a field functional checking procedure that will be utilized and followed
through. Remember, a pressure switch less than $400 in cost could cause equipment shutdown that could reduce on-site revenue on the order of 1–2 $MM per
day. The pressure switch selected could be the best, the highest quality in the
world, properly installed, and set. If its calibration is not periodically checked, it
could cause an unnecessary shutdown of equipment and result in this revenue loss.
Functional Lube/Seal System Test Procedure Outline
Objective:
To confirm proper functional operation of the entire system
prior to equipment start-up
Procedure format:
Detail each test requirement. Specifically note required
functions/set points of each component. Record actual
functions/set points and all modifications made.
Note:
All testing to be performed without the unit in operation.
I
Preparation
A. Confirm proper oil type and reservoir level
B. Confirm system flush is approved and all flushing screens are removed
C. Confirm all system utilities are operational (air, water, steam, electrical)
D. Any required temporary nitrogen supplies should be connected
E. All instrumentation must be calibrated and control valves properly set
F. Entire system must be properly vented
II Test procedure
A. Oil Reservoir
1. Confirm proper heater operation
2. Check reservoir level switch and any other components (TIs, vent
blowers, etc.)
B. Main pump unit
1. Acceptable pump and driver vibration
2. Absence of cavitation
491
B.P. 10.9
More Best Practices for Rotating Equipment
3. Pump and driver acceptable bearing temperature
4. Driver governor and safety checks (uncoupled) if driver is a steam
turbine
C. Auxiliary pump unit
Same procedure as item B.
D. Relief valve set point and non-chatter check
E. Operate main pump unit and confirm all pressures, differential pressures,
temperatures, and flows are as specified on the system schematic and/or
Bill of material
F. Confirm proper accumulator pre-charge (if applicable)
G. Confirm proper set point annunciation and/or action of all pressure, differential pressure, and temperature switches
H. Switch transfer valves from bank “A” to bank “B” and confirm pressure
fluctuation does not actuate any switches
I. Trip main pump and confirm auxiliary pump starts without actuation of
any trip valves or valve instability
Note: Pressure spike should be a minimum of 30% above any trip settings
J. Repeat step I but slowly reduce main pump speed (if steam turbine) and
confirm proper operation
K. Simulate maximum control oil transient flow requirement (if applicable)
and confirm auxiliary pump does not start
L. Start auxiliary pump, with main pump operating and confirm control
valve and/or relief valve stability
Note: Some systems are designed to not lift relief valves during two
pump operation
III Corrective action
A. Failure to meet any requirement in Section II requires corrective action
and retest
B. Specifically note corrective action
C. Sign off procedure as acceptable to operate
Electro-Hydraulic Governor Functional Test Procedure Outline
Objective:
To confirm proper system functional operation prior to equipment
start-up
Procedure format:
Detail each test requirement. Specifically note required
functions/set points. Record actual functions/set points and all
modifications made.
Note: All testing to be performed without the unit in operation.
I
Preparation
j Confirm all shut down contracts are in the normal condition
j Confirm all power supplies are on
492
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
Secure necessary test equipment
l Pressure sources (nitrogen bottles) for pressure simulation at transmitters
l Frequency generators for simulating speed signals
II Test procedure
j Take required action to put system in “run” mode
j Open trip and throttle valve only after insuring the main steam block
valve is closed
j Simulate turbine start, slow roll, and any start sequence “hold” points up
to minimum governor operating point
j Confirm proper operation of “raise” and “lower” speed buttons
j Connect external process signal inputs (one at a time) and confirm proper
governor action to input signal variation
j Check overspeed override feature
j Confirm automatic transfer to and from backup governor “position control” for each of the following cases:
l Loss of main governor power supply
l Zero external input signal
l Failure of “final driver” (internal governor component)
l Zero speed inputs
j Confirm manual transfer to and from backup governor and “emergency
override”
j Check raise and lower speed controls while in backup governor mode
j Confirm governor shutdown (trip) operation under the following
conditions:
l Overspeed setting
l Failure of both main and backup governor controls
III Corrective action
j Failure to meet any requirement in Section II requires corrective action
and retest
j Specifically note corrective action
j Sign off procedure as acceptable to operate
j
Steam Turbine Solo Run Functional Test Procedure Outline
Objective:
To confirm acceptable mechanical operation of the steam
turbine, governor system, and safety (trip) system
Procedure format:
Detail each test requirement. Specifically note required
test limits (Note: shop test data should be used to define
acceptable limits). Record actual test values using appropriate
instrumentation and note all modifications made.
I
Preparation
j Confirm all auxiliary system tests are complete (governor, lube system, etc.)
j Confirm all inlet steam lines have been cleaned and signed off
493
B.P. 10.9
More Best Practices for Rotating Equipment
Confirm all installed instrumentation is calibrated
Secure all required instruments
l Calibrated pressure and temperature gauges
l Oscilloscope(s)
l Vector filters
l Amplifier(s)
l Spectrum analyzer
l Tape recorder or information gathering module
j “Walk” all steam inlet, extraction, and exhaust lines. Confirm all spring
hangers are released (unlocked) and safety valves are installed
II Test procedure
j Confirm all auxiliary systems are operational and at proper conditions
(lined out)
l Lube/control oil system
l Governor/trip system
l Turning gear (if applicable)
l All warming lines drained and operational
l Condensing system including condensate pumps (if applicable)
l Extraction system (if applicable)
l Steam seal system
l Condition monitoring systems (vibration, temperature, etc.)
l Steam conditions within allowable vendor limits
j Slow roll and start unit as per vendor’s instructions (refer to cold start-up
speed vs. time chart)
j Demonstrate manual trip (panic button) at low speed (500 rpm)
j Reset trip, accelerate back to desired speed—listen for rubs, etc.
j Gradually increase speed to next speed step. Record the following data
for each vendor’s required speed step up to minimum governor speed
l Overall vibration at each vibration point (record frequency if specified limits are exceeded)
l Bearing oil temperature rise at each bearing
l Bearing pad temperature (axial and radial) at each point (if applicable)
l Turbine speed
l Axial shaft displacement
l Turbine exhaust temperature
Note: Use shop test data for comparison
j After confirming stable operation at minimum governor speed, accelerate
carefully to overspeed trip setting and trip the turbine 3 times. Each trip
speed should fall within the vendors’ trip speed set point allowable range
j Return to minimum governor speed and confirm satisfactory manual and
automatic speed control. Also confirm automatic transfer from main to
backup governor
j Connect vibration recording instruments, reduce turbine speed to
500 rpm, and record shaft vibration (at each vibration monitoring point)
j
j
494
Construction, Installation, Commissioning, and Turnarounds
Chapter | 10
and phase angle while gradually increasing speed to maximum continuous speed. Repeat step in reverse direction (maximum continuous speed
to 500 rpm)
Note: This data will be reduced to Bode, Nyquist, and Cascade plots and
should be compared to shop test data
j Increase turbine speed to maximum continuous speed and run for 4 h or
until bearing temperatures stabilize
j Finally trip the turbine using a system trip switch (simulate low oil pressure, etc.)
III Corrective action
j Failure to meet any requirement in Section II requires corrective action
and retest
j Specifically note any corrective action
j Sign off equipment as acceptable to operate
Lube/Seal System Test Procedure
Item:
Reference DWGS:
Turbine utility
P&ID _____
Compressor utility
P&ID _____
Purpose:
To fully prove functional operation of entire lube/seal system, including all permissive, alarm, and shutdown functions prior to initial
operation of the unit
Note:
All testing to be performed without the unit in operation. When
specified values are not satisfied, correct and retest
Preparation
Prior to testing of the system, confirm and sign off that the following
has been checked:
j
j
j
j
j
j
j
j
j
j
Oil reservoir at proper level
Specified oil is used
Oil heater in operation
Oil cooler water supply on
System clean (all test screens out)
Instrument air in operation at all instruments
Temporary N2 supply connected
All instrumentation noted on attached list has been calibrated to
specified values
Steam lines to console blown
Entire system vented
Lube System Check List (Example)
Item
Description
Specified value
Actual value
1.
Record reservoir temp. rise in
4 h with heater on.
Read on T.I.
Record actual ∆T
______
Witnessed
by
495
B.P. 10.9
More Best Practices for Rotating Equipment
Item
Description
Specified value
Actual value
2.
Check reservoir level switch
setting. Read on Annun.
High level 14 in.
from top. Low level
45 in. from top
High
______
Low
______
0.2 in./s
0.2 in./s
None
165°F
3130 kPa
______
______
______
______
______
690 kPa rising
_____
2262 kPa
_____
124–138 kPa
_____
883 kPa
_____
241 kPa
_____
241 kPa
_____
Approx. 2600
Less than 65°C
49°C
Less than 70 kPa
_____
_____
_____
_____
3.
Energize aux. lube pump
check:
3A. Pump vibration
3B.
Motor vibration
3C. Cavitation
3D. Pump/motor brg. Temp.
4. (see Block in aux. pump using
note 1) pump discharge valve-set
relief valve. Valve chatter
is not acceptable
5.
Confirm the “Aux. pump
running” annun. is actuated
by switch by shutting off
pump and restarting while
reading pressure on PI
6.
Allow system to heat up to
49°C downstream of coolers
and adjust the following items
(if required) to attain specified
values:
6A. Back press. regulator
Read value on PI.
6B.
Lube oil supply valve PCV
Read value on P.I.
6C. Control oil valve PCV
Read value on P.I.
6D. L.P. case seal oil differential
valve PDCV. Supply N2 press
of 5 PSIG at gas reference side
of PDT and read differential
press on PDI
6E.
HP case seal oil differential
valve. Supply N2 pressure of
30 PSIG at gas reference side
of PDT and read differential
press on PDI
7.
Record the following:
7A. Pump disch. press. on PI
7B.
Oil temp. upstream on TI
7C. Oil temp. downstream on TI
7D. Cooler/filter ∆P on PDI
Switch transfer valve and
record:
Bank “A” ∆P
Bank “B” ∆P
496
Witnessed
by
Construction, Installation, Commissioning, and Turnarounds
Item
7E.
7F.
7G.
7H.
7I.
7J.
7K.
Description
Lube press. at console on PI
Control press. at console on PI
LP case on ∆P PDI
HP case on ∆P PDI
Lube press. at unit on PI
Control press. at unit on PI
All “sight” glasses show oil
flow
7L.
Lube oil head tank is full
7M. Turbine accumulator press.
on PI
8.
Record seal oil drainer level
for each drainer for 1 h
Drainer A
Drainer B
Drainer C
Drainer D
9.
Switch transfer valve from “A”
to “B” bank and observe press.
fluctuation on PI and confirm
that PSL does not actuate
10.
Bleed low side of filter switch
PD SH and confirm PDAH
actuates at specified value
11.
Bleed pressure off PSL (low
oil press) and confirm @
unit annun. PAL actuates at
specified value. Read P.I.
12.
Increase temporary N2 press.
on PDSL (L.P. case seal low
∆P alarm switch) and confirm
Annun. PDAL actuates at
specified value. Read PDI
Repeat above for PDSL (H.P.
13.
(see
case seal low ∆P alarm).
note 2) Confirm Annun. PDAL
actuates read on PDI
14.
Bleed pressure off PSL
(low oil trip switch) and
observe:
A. Annun. PALL functions.
Read PI.
B. T&T valve closes
C. Valve rack closes
15.
Increase temporary N2 press.
(see
on PDSLL (L.P. case seal low ∆P
note 2) trip switch) and observe that:
A. Annun. PDALL functions.
Read on PDI
B. Action occurs as in 14
Chapter | 10
Specified value
Actual value
a
Approx. 283 kPa
1091 kPa
241 kPad
241 kPad
124–138 kPa
883 kPa
_____
_____
_____
_____
_____
_____
_____
—
a
Approx. 900 kPa
_____
_____
2 fills per hour
_____
_____
_____
_____
_____
241 kPa Rising
_____
a
Witnessed
by
90 kPa Falling
207 kPa Falling
_____
kPa Falling
_____
76 kPa Falling
______
138 kPa Falling
______
497
B.P. 10.9
More Best Practices for Rotating Equipment
Item
Description
Specified value
Actual value
16.
Repeat above for PDSLL
(H.P. case seal low ∆P trip)
and observe Annun. PDALL
functions at specified value.
Read PDI
Shut off aux. oil pump
and confirm all valves are
stable
Time rundown of oil tank.
138 kPa Falling
______
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
498
Observe level switch LSL
actuates Annun. LAL
A. Disconnect main lube oil
pump from turbine
B. Drain steam inlet line to
main lube pump turbine
C. Drain turbine
D. Confirm turbine bearing
cooling water is on
E. Confirm trip is reset
Open inlet valve, gradually
bring turbine up to speed and
confirm the following:
A. Rated speed (Strobe trac)
B. Inlet press. PI
C. Inlet temp. TI Temp
D. Exhaust temp. PI Temp
E. Vibration (using matrix
vibration instrument)
Disable governor and check
overspeed trip 3 times
Manually trip turbine using
hand trip
Check pump/turbine
alignment and couple up
main pump
Slowly bring pump up to
speed and check:
24A. Pump vibration
24B. Cavitation
24C. Pump brg. temp.
Adjust Governor so all valves
are as noted in step 7
Record speed RPM
Block in pump using pump
discharge valve. Set relief
valve PSV. Valve chatter is not
acceptable
______
Approx. 4 in.
above Tang. Line
______
2274 kPa
327°C
517 kPa
0.2 in./s
4140 RPM
_____
0.2 in./s
None
165°F
_____
_____
_____
_____
_____
Witnessed
by
Construction, Installation, Commissioning, and Turnarounds
Item
Description
27.
With turbine operating at speed
noted in step 25 and disch.
block valve open, manually
trip turbine and observe:
A. Aux. pump starts
B. Min. press. spike on P.I.
lube
P.I. Control
28.
29.
30.
31.
32.
P.D.I. LP Case ∆P seal
P.D.I. HP Case ∆P seal
C. Alarm and trip switches
connected with lube and seal
oil are not actuated
D. All valves are stable
Restart pump turbine and
dump control oil pressure
using hand valve at turbine.
Observe that no alarm or trip
lights are actuated and that all
valves remain stable
Start aux. oil pump with
turbine operating and observe:
A. RVs do not lift
B. All control valves are stable
C. Oil pump turbine does not
hunt (speed remains stable)
Stop aux. pump motor and
observe:
A. All control valves remain
stable
B. Min. press. spike on P.I.
lube
P.I. Control
P.D.I. LP Case ∆P
P.D.I. LP Case ∆P
With turbine operating. Bleed
press., from PSL and observe
aux. pump starts and that all
valves remain stable
Having satisfactorily
completed all previous items,
secure both aux. and main
pump and sign off as being
acceptable for operation.
Specified value
Chapter | 10
Actual value
Witnessed
by
_____
90 kPa
T&T valve does not
close
207 kPa
207 kPa
_____
_____
_____
_____
_____
_____
_____
_____
_____
90 kPa T&T valve
does not trip
207 kPa
207 kPa
_____
_____
_____
______
_____
Note: At this point, elect one driver and continue to operate the console 24 h a day. Note 1: RVs will
continuously pass a small stream of oil. Actual setting will be that pressure at which stream volume
increases. Observe by un-bolting FLGS at reservoir. Accumulation value is with pump discharge block
fully closed. Note 2: Block out gas signal to diff. control valve during this step.
a
Adjust as required to attain proper values at the unit.
499
B.P. 10.9
More Best Practices for Rotating Equipment
The site air or inert gas run procedure should be mutually developed with
the machinery vendor to assure that all operational differences (Power requirements, dry gas seal conditions, pressures, temperatures, and so on) are considered and planned for.
The test objective is to completely confirm the mechanical operation of the
machinery and the associated auxiliary systems prior to plant operation.
Process piping may have to be disconnected for these runs to prevent overheating. In some applications, temporary temperature measuring devices have
to be installed to curtail testing when casing temperatures exceed design limits
approximately 200°C (400°F).
500
Chapter 11
Predictive and Preventive
Maintenance
B.P. 11.1: Begin Root Cause Analysis (RCA) immediately when a change
in condition of one or more components has been observed
By using the philosophy of component condition monitoring (CCM) to
properly monitor and trend the condition of the major machinery components,
you will be able to see any significant change. A significant change is defined
generally as anything greater than 10% change. This indicates a potential issue
and it needs to be evaluated immediately so that it does not reach alarm or trip
values and cause an unplanned shutdown.
L.L. 11.1: Failure to identify component condition change early enough
has resulted in numerous unplanned shutdowns that could have been
avoided
BENCHMARKS
This best practice has been recommended for years and although it is followed
at times, it is not typically followed to the satisfaction of the author. Particularly
in the last few years it has become more prominent as staff at many plants is of
less experience and less quantity and are basically caught up with “firefighting”
failures.
SUPPORTING MATERIAL
The five machinery failure classifications are presented in Table 11.1.1.
The details concerning each of these failure classifications were discussed in
the previous chapter. How can these failure causes be prevented?
Re-examination of the details concerning each failure classification shows
that the solution to the prevention of each failure cause is identical. This fact is
presented in Table 11.1.2.
Let’s now examine each of the action items noted previously in detail.
More Best Practices for Rotating Equipment. http://dx.doi.org/10.1016/B978-0-12-809277-4.00011-5
Copyright © 2017 Elsevier Inc. All rights reserved.
501
B.P. 11.1
More Best Practices for Rotating Equipment
TABLE 11.1.1 Failure Classifications
• Process condition changes
• Improper assembly/maintenance/installation
• Improper operating procedures
• Design deficiencies
• Component wearout
TABLE 11.1.2 Prevent Machinery Failures by …
• Component function awareness (what should it do?)
• CCM (what is it doing?)
• Using PDM techniques
• Teamwork—reliability is everyone’s responsibility
CCM, Component condition monitoring; PDM, predictive maintenance.
TABLE 11.1.3 What Is It Supposed to Do?
Thoroughly understand the function of each component by:
• Reading the instruction book
• Asking questions of:
• Site reliability group
• Site technical group
• Machinists
• Operators
• Referring to reference books
• Organizing “mini” information sessions for operators and machinists
Component Function Awareness—“What Should It Do?”
Component (machinery part) function awareness allows you to determine what
the component is supposed to do. It is obvious that a certain amount of knowledge is required to accomplish this fact. Remember, you may not have all of the
knowledge required. OBTAIN IT! Table 11.1.3 presents sources of where the
information may be obtained.
Fig. 11.1.1 presents the important principle of knowledge base. The greater
this base, the more effective predictive maintenance (PDM) and RCA procedures will be.
502
Predictive and Preventive Maintenance
Chapter | 11
FIGURE 11.1.1 The more you know, the better you care!
One final word. Do not be afraid to admit to management that you do not
know certain aspects of a problem. But be sure to state that you will find out.
After all, management must understand that this is a learning process and does
require time.
To aid in the understanding of component function definition, we have included an example for an anti-friction bearing in Table 11.1.4.
Naturally, we cannot measure directly all of the items noted in Table 11.1.4.
However, based on the instruments and measuring devices available on site,
what can be measured to assure the component (bearing) is performing correctly.
Component Condition Monitoring—“What Is It Doing?”
In reference to the anti-friction bearing example, the CCM parameters are presented in Table 11.1.5.
TABLE 11.1.4 Component Function Example
An anti-friction bearing continuously supports all static and dynamic forces of a rotor
by providing sufficient bearing area and requires oil flow to remove the generated
frictional heat.
This statement then defines the items that must be monitored to determine the
bearing’s condition:
•
•
•
•
Static and dynamic forces
Bearing area
Oil flow
Frictional heat
503
B.P. 11.1
More Best Practices for Rotating Equipment
TABLE 11.1.5 CCM Parameters (For Anti-Friction Bearing)
• Bearing housing vibration
• Bearing housing temperature
• Lube oil condition
• Viscosity
• Water content
• Oil particle content
A similar exercise can be conducted for all of the major components and
systems in any piece of equipment.
What are the major components and systems of any piece of rotating equipment? How many are there? And are the same components contained in any
type of rotating equipment? The answers to these questions will be discussed in
a later chapter of this book and form the principle of CCM.
Preventive and Predictive Maintenance
At this point, the distinctions between preventive maintenance (PM), PDM, and
troubleshooting must be discussed.
Preventive Maintenance
PM requires that maintenance be performed at predetermined intervals. It is
time based. A most common PM step is an automotive oil change. The objective of this action is to remove the oil from the engine before oil contamination
and deterioration cause excessive wear to the engine components. Fig. 11.1.2
presents the components of a typical site PM program.
In our experience, a well-planned PM program can truly be effective. However, the question must be asked, “Is the maintenance performed always necessary?” Refer to Table 11.1.6.
What is the basis for replacing components? Unnecessary component
­replacement exposes the machinery unit to a failure classification (improper
assembly of components, improper installation, component malfunction, component improper storage procedures, etc.).
In addition, PM can cause a mindset that automatically determines maintenance at every turnaround regardless of component condition. This can be a
costly practice. A case history also demonstrates where PM can lead to, if not
properly monitored.
A centrifugal compressor in a large refinery was scheduled for maintenance
during the upcoming turnaround. Maintenance planning had scheduled bearing
inspection and change if necessary. During the turnaround when bearings were
inspected, excessive clearances and signs of deterioration were found. ­Naturally
504
Predictive and Preventive Maintenance
FIGURE 11.1.2
Consultant)
Chapter | 11
A typical preventive maintenance (PM) program. (Courtesy of M.E. Crane
TABLE 11.1.6 PM
• PM prevents but … takes time.
• Is it always necessary?
the bearings were replaced. However, because the bearings were replaced, it
was decided that the seals, which are more difficult to remove and inspect, be
observed. Upon seal removal the seals were also in a distressed condition and
needed to be replaced. Now the tough decisions had to be made. It was decided that the compressor would be dissembled to inspect interior condition for
possible causes of seal and bearing failure. Upon disassembly, no significant
abnormalities were found within the compressor and it was consequently reassembled.
This case history demonstrates how a standard PM approach can lead to unnecessary maintenance and significant loss of revenue to the operating unit. In
this case, the operating unit did not make use of site instrumentation. Nowhere
had people answered the question “What changed?”. This approach therefore
led to unnecessary disassembly of the compressor. If bearing parameters (temperature, vibration, etc.) and seal parameters (inner and outer seal leakage) had
been monitored for change, the conclusions of only bearing and seal change
would have been made without unnecessary disassembly. Remember, to disassemble a compressor, significant additional tools and materials are required.
Typical time for compressor disassembly can easily reach 1 week. It can be seen
505
B.P. 11.1
More Best Practices for Rotating Equipment
FIGURE 11.1.3 A predictive maintenance (PDM) program. (Courtesy of M.E. Crane C
­ onsultant)
therefore, that the effective way to perform any maintenance activity is to thoroughly plan that activity based on condition changes to equipment. This leads
us to the discussion of PDM.
Predictive Maintenance
PDM is based on CCM and trending. Fig. 11.1.3 presents the definition of PDM.
Troubleshooting
Wherever I travel, worldwide, Troubleshooting is the “keyword.” More recently, other “keywords” have emerged:
l
l
l
Failure analysis
RCA
Reliability centered maintenance (RCM)
Regardless of the “keyword,” it’s still troubleshooting. This term is defined
in Table 11.1.7.
What are the requirements to accomplish an effective troubleshooting exercise? These facts are presented in Table 11.1.8.
Do these requirements sound familiar? They certainly should. These are the
requirements for PDM! The differences between these two terms are presented
in Table 11.1.9.
Therefore, if we use site-wide PDM techniques, we can potentially detect a
change in condition before failure. Please refer to Table 11.1.10.
506
Predictive and Preventive Maintenance
Chapter | 11
TABLE 11.1.7 Definition
Troubleshoot—to discover and eliminate (root) causes of trouble
TABLE 11.1.8 An Effective Troubleshooting Exercise
• Troubleshooting requires that all abnormal conditions be defined
• However, to determine abnormal conditions, the normal conditions must be known
• Therefore baseline (normal) conditions must be known
TABLE 11.1.9 PDM and Troubleshooting
• PDM requires baseline and trend data to predict the root cause of the change in
condition
• Troubleshooting requires baseline and trend data to predict the root cause of failure
TABLE 11.1.10 Troubleshooting …
Is PDM after a failure!
Notice that in previous discussion, the word “potentially” was in italics.
Remember that the majority of rotating equipment in any plant is general
purpose or spared equipment that is not continuously monitored in the control
room DCS system. This equipment is also the source of most reliability problems (Bad Actors). How can this equipment be effectively monitored?
Let’s now discuss the final topic of B.P. 11.1.
Reliability: Everyone’s Responsibility
You’ll have fewer problems if you and the mechanic (operators and machinists)
know more and—work as a team!
Reliability must be everyone’s responsibility. The entire plant operations,
maintenance, and engineering departments must be aware of the reliability program philosophy and must be able to implement it. Having operators and machinists equipped and trained in the use of simple vibration instruments (vibration pens), oil condition monitors, and laser temperature guns will ­significantly
increase the reliability of general purpose (spared) equipment through the implementation of an effective PDM program.
507
B.P. 11.2
More Best Practices for Rotating Equipment
FIGURE 11.1.4 Article on proactive maintenance.
See Fig. 11.1.4 for an article published on the details of the subject of this
best practice in 2013.
B.P. 11.2: Try to postpone pump maintenance until turnaround to assure
that a spare pump is always available
By utilizing the concept of CCM for pumps you identify if the pump is not
producing the desired flow rates based on the head required by the process. If
508
Predictive and Preventive Maintenance
Chapter | 11
it is determined that the rates will soon have to be reduced due to deterioration
of performance for the main pump (A pump), rather than take it to the shop for
maintenance, switchover to the B pump and operate A as a spare in case of an
emergency where it would be needed. It’s better to have an underperforming
pump as a backup than have no pump available if there is a failure to the B pump
when the A pump is out for maintenance.
L.L. 11.2: Inability to have a spare pump available (because one pump is
in the shop for maintenance) has resulted in numerous instances of plants
that had to significantly reduce rates because of no pumps available for a
particular service
BENCHMARKS
The writer has recommended this best practice since the late 1990s when involved with a number of refinery machinery reliability audits. Since that time,
this best practice has improved unit reliability and the ability to maintain desired
unit rates until the next scheduled turnaround.
SUPPORTING MATERIAL
Table 11.2.1 presents the major component classifications for any type of machinery:
l
l
l
l
Pumps
Steam Turbines
Compressors
Motors
Gas Turbines
Fans etc.
l
l
Regardless of the type of machinery, monitor these components and you will
know the total condition of the machine.
Component Condition Monitoring
As previously stated, component and system functions must first be defined and the
normal values for each component listed. These facts are presented in Table 11.2.2.
TABLE 11.2.1 Major Machinery Components and Systems
•
•
•
•
•
Rotor
Radial bearing
Thrust bearing
Seal
Auxiliary systems
509
B.P. 11.2
More Best Practices for Rotating Equipment
TABLE 11.2.2 Component and System Functions
• Define the function of each affected component
• Define the system in which each affected component operates
• List the normal parameters for each affected component and system component
FIGURE 11.2.1 Component condition monitoring (CCM).
Once the function of each component is defined, each major machinery
component can be monitored as shown in Fig. 11.2.1.
Baseline
Having defined all condition parameters that must be monitored, the next step in
a condition monitoring exercise is to obtain baseline information. It is important
to obtain baseline information as soon as physically possible after start-up of
equipment. However, operations should be consulted to confirm when the unit
is operating at rated or lined out conditions. Obtaining baseline information
without conferring with operations is not suggested since misinformation could
be obtained and thus lead to erroneous conclusions in PDM. Table 11.2.3 states
the basics of a baseline condition.
It is amazing to us how many times baseline conditions are ignored. Please
remember Table 11.2.3 and make it a practice to obtain baseline conditions as
soon as possible after start-up. You can only trend if you have a start point!!!
Trending
Trending is simply the practice of monitoring parameter condition with time.
Trending begins with baseline condition and will continue until equipment
shutdown. In modern day thought, it is often conjectured that trending must be
TABLE 11.2.3 Base Line Condition
If you don’t know where you started, you don’t know where you are going!
510
Predictive and Preventive Maintenance
Chapter | 11
FIGURE 11.2.2 Trending data.
performed by micro-processors and sophisticated control systems. This is not
necessary! Effective trending can be obtained by periodic manual observation
of equipment or using equipment available to us in the plant, which will include
DCS systems, etc. The important fact is to obtain the baseline and trends of
data on a periodic basis. When trending data, threshold points should also be
defined for each parameter that is trended. This means that when the parameter pre-established value is exceeded action must be taken regarding problem
analysis. Setting threshold values a standard percentage above normal value is
recommended. Typically, values are on the order of 25–50% above baseline
values. However, these values must be defined for each component based on
experience. Fig. 11.2.2 presents trending data for a hydrodynamic journal bearing. All of the parameters noted in Fig. 11.2.2 should be monitored to define the
condition of this journal bearing.
Specific Machinery Component and System Monitoring
Parameters and Their Limits
Following is information concerning what parameters should be monitored for
each major machinery component to determine its condition. In addition, typical limits are noted for each component. Note that these are typical alarm limits
and are not intended to be the point of concern. If component condition is trended properly, the idea is to investigate the issue long before these typical alarm
limits are reached as that is when it is usually too late and work is required.
511
B.P. 11.2
More Best Practices for Rotating Equipment
TABLE 11.2.4 Pump Performance Monitoring
1. Take value at minimum flow (shut off discharge valve)
2. Measure:
• P1
• P2
•
•
Driver BHP
Specific gravity
3. Calculate:
ft. − lb f
∆P × 2.311
=
lbm
S.G.
A.
Head Produced
B.
Pump Efficiency (%) =
hd × gpm × S.G.
3960 × bhp
4. Compare to previous value if > −10% perform maintenance
The Rotor
Rotor condition defines the performance condition (energy and efficiency) of
the machine. Table 11.2.4 presents this value for a pump.
Radial Bearings
Tables 11.2.5 and 11.2.6 present the facts concerning anti-friction and hydrodynamic (sleeve) radial or journal bearing condition monitoring.
Thrust Bearings
Tables 11.2.7 and 11.2.8 show condition parameters and their limits for antifriction and hydrodynamic thrust bearings.
TABLE 11.2.5 Condition Monitoring Parameters and Their Alarm Limits,
Journal Bearing (Anti-Friction)
Parameter
Limits
1. Bearing Housing Vibration (Peak)
0.4 in./s (10 mm/s)
2. Bearing Housing Temperature
185°F (85°C)
3. Lube Oil Viscosity
Off Spec 50%
4. Lube Oil Particle Size
• Non Metallic
• Metallic
25 µm
Any Magnetic Particle In The Sump
5. Lube Oil Water Content
Below 200 ppm
512
Predictive and Preventive Maintenance
Chapter | 11
TABLE 11.2.6 Condition Monitoring Parameters and Their Alarm Limits,
Journal Bearing (Hydrodynamic)
Parameter
Limits
1. Radial Vibration (Peak To Peak)
2.5 Mils (60 µm)
2. Bearing Pad Temperature
220°F (108°C)
3. Radial Shaft Position
a
>30° Change and/or 30% Position Change
4. Lube Oil Supply Temperature
140°F (60°C)
5. Lube Oil Drain Temperature
190°F (90°C) Off
6. Lube Oil Viscosity
Spec 50% > 25
7. Lube Oil Particle Size
Micrometers Below
8. Lube Oil Water Content
200 ppm
a
Except for gearboxes where greater values are normal from unloaded to loaded.
TABLE 11.2.7 Condition Monitoring Parameters and Their Alarm Limits,
Thrust Bearing (Anti-Friction)
Parameter
Limits
1. Bearing Housing Vibration (Peak)
• Radial
• Axial
0.4 in./s (10 mm/s)
0.3 in./s (1 mm/s)
2. Bearing Housing Temperature
185°F (85°C)
3. Lube Oil Viscosity
Off Spec 50%
4. Lube Oil Particle Size
• Non Metallic
• Metallic
>25 µm
Any Magnetic Particles With Sump
5. Lube Oil Water Content
Below 200 ppm
Seals
Table 11.2.9 presents condition parameters and their limits for a pump liquid
mechanical seal.
Auxiliary Systems
Condition monitoring parameters and their alarm limits are defined in
Table 11.2.10 for Lube oil systems.
513
B.P. 11.2
More Best Practices for Rotating Equipment
TABLE 11.2.8 Condition Monitoring Parameters and Their Alarm Limits,
Thrust Bearing (Hydrodynamic)
Parameter
Limits
1. Axial Displacementa
>15–20 mils (0.4–0.5 mm)
2. Thrust Pad Temperature
220°F (105°C)
3. Lube Oil Supply Temperature
140°F (60°C)
4. Lube Oil Drain Temperature
190°F (90°C)
5. Lube Oil Viscosity
Off Spec 50%
6. Lube Oil Particle Size
>25 µm
7. Lube Oil Water Content
Below 200 ppm
a
And thrust pad temperatures >220°F (105°C).
TABLE 11.2.9 Condition Monitoring Parameters and Their Alarm Limits,
Pump Liquid Mechanical Seal
Parameter
Limits
1. Stuffing Box Pressure
>50 psig (350 kPa)
Above the fluid vapor pressure
2. Temperature drop across orifice
Strainer, or cyclone sep.
Should be negligible unless plugged
3. Temperature Drop across cooler
<30°F for Plan 23, >30°F for Plan 21.
Temp. should be less than 160°F for water
4. Temperature drop across reservoir
If zero, no circulation is occurring
TABLE 11.2.10 Condition Monitoring Parameters and Their Alarm Limits,
Lube Oil Systems
Parameter
Limits
1. Oil Viscosity
Off Spec 50%
2. Lube Oil Water Content
Below 200 ppm
3. Auxiliary Oil Pump Operating Yes/No
Operating
4. Bypass Valve Position (P.D. Pumps)
Change > 20%
5. Temperature Control Valve Position
Closed, Supply Temperature >130 (55°C)
6. Filter ∆P
>25 psid (170 kPag)
7. Lube Oil Supply Valve Position
Change > ±20%
514
Chapter | 11
Predictive and Preventive Maintenance
TABLE 11.2.11 Obtain and Maintain Management Support By ...
1. Clearly Stating Impact Of Problem On Plant Profit
2. Prepare A Brief Statement Of:
• Problem
• Impact On Plant
• Action Plan
3. Be Confident!
4. Be Professional!
5. Provide Timely Update
Predictive Maintenance Techniques
Now that the CCM parameters and their limits have been presented, PDM techniques must be used if typical condition limits are exceeded. This action will
assure that we minimize site-troubleshooting exercises.
One final recommendation is presented in Table 11.2.11.
Tables 11.2.12–11.2.14 present condition monitoring parameters and limits
for dynamic compressor performance, liquid seals, and seal oil systems.
TABLE 11.2.12 Compressor Performance Condition Monitoring
1. Calibrated: Pressure And Temperature Gauges And Flow Meter
2. Know Gas Analysis And Calculate k, Z, MW
3. Perform As Close To Rated Speed And Flow As Possible
4. Relationships:
(T )
LN 2
N −1
(T1)
A.
=
(P2 )
N
LN
(P1)
B. EFFICIENCYpoly
k −1
= k
n −1
n
n −1


 Ft − lb f  1545
n
  P2  n − 1
C. HEAD poly = 
=
×
×
×
Z
×
T
1
avg
  P1 

 Lbm  MW
n −1


5. Compare To Previous Value. If Decreasing Trend Exists Greater Than 10%. Inspect At
First Opportunity.
515
B.P. 11.3
More Best Practices for Rotating Equipment
TABLE 11.2.13 Condition Monitoring Parameters and Their Alarm Limits,
Compressor Liquid Seal
Parameter
Items
1. Gas Side Seal Oil/Gas ∆P
•
Bushing
<12 ft. (3.5 m)
•
Mechanical Contact
<20 Psi (140 kPa)
2. Atmospheric Bushing Oil Drain Temperature
200°F (95°C)
a
3. Seal Oil Valve Position
>25% Position Change
4. Gas Side Seal Oil Leakage
>20 gpd Per Seal
Return valve = −25%.
Note this assumes compressor reference gas pressure stays constant.
a
Supply valve = +25%.
TABLE 11.2.14 Condition Monitoring Parameters and Their Alarm Limits,
Compressor Liquid Seal Oil Systems
Parameter
Limits
1. Oil Viscosity
Off Spec 50%
2. Oil Flash Point
Below 200°F (100°C)
3. Auxiliary Oil Pump Operating Yes/No
Operating
4. Bypass Valve Position (P.D. Pumps)
Change > 20%
5. Temperature Control Valve Position
Closed, Supply Temperature 130°F (55°C)
6. Filter ∆P
25 psid (170 kPag)
7. Seal Oil Valve Position
Change > 20% Open (Supply)
8. Seal Oil Drainer Condition
(Proper Operation)
> 20% Closed (Return)
• Constant Level (Yes/No)
Level Should Be Observed
• Observed Level (Yes/No)
Level Should Not Be Constant
• Time Between Drains
Approximately 1 h (Depends On Drainer
Volume)
B.P. 11.3: Initiate site machinery instrumentation excellence program to
assure all installed instruments are calibrated and in working condition
All instruments are installed for a reason and that is to give a glimpse of
a condition of a certain component. If they are not functioning properly, we
­cannot accurately monitor or know if that component is operating properly and
are at risk of a component failure.
516
Predictive and Preventive Maintenance
Chapter | 11
Therefore, a program should be initiated at each site in order to make sure all
gauges, transmitters, and switches are in working order and calibrated p­ roperly.
If a gauge is not working properly or the pointer is missing, etc. replace it immediately!!!
It is also important to train operators on monitoring their equipment and
making sure they are aware of the importance of each instrument.
L.L. 11.3: Inaccurate or non-working instruments have resulted in unplanned shutdowns because a certain component was not accurately being
monitored, causing lost revenue
BENCHMARKS
This best practice has been recommended since the late 1980s to all plants visited. Plants following this best practice are in the first quartile in machinery reliability and face very little firefighting because they are accurately monitoring
and trending their equipment.
Supporting Material
This principle can be easily seen after looking at Fig. 11.3.1.
In Fig. 11.3.1, a typical instrument panel on passenger car is shown. Just
think what would happen if the needles were not working on the gauges, or
FIGURE 11.3.1 Instrument panel.
517
B.P. 11.4
More Best Practices for Rotating Equipment
TABLE 11.3.1 Machinery Instrumentation Key Facts
1. Know the Purpose of each Instrument on Your Panel
2. If you don’t know ask
3. Know the Alarm and Trip Values
4. Observe Change in Values—Don’t wait for Alarms
5. Know the Purpose of each Local Instrument
6. If instruments are broken or unreadable—Correct Now!
better yet somebody spray painted the gauges black so you couldn’t see!!! You
wouldn’t feel comfortable driving it, would you?
Now, just think, your instrument panel for the machinery that is being operated in the field are all of the gauges and instruments in the field. If one of them
is not reading properly, you should have the same feeling as if your gauges in
your car are not working or are unreadable.
Table 11.3.1 depicts the important facts that we all must remember about our
machinery instrumentation.
B.P. 11.4: Utilize a company machinery database for lessons learned in
order to improve machinery reliability
Obtain daily revenue value (based on nominal market prices) and calculate the time that specific issues (Lesson Learned) have resulted in a shutdown or reduced rates and what the total loss of revenue was for each Lesson
Learned.
This information should be tabulated in a company wide database along
with the Best Practice that would be focused on eliminating the root cause.
Then all facilities within the corporate umbrella can utilize these Best Practices
to resolve ongoing issues.
L.L. 11.4: Failure to utilize Corporate Wide LL/BP Database has resulted
in certain plants within the company that have ongoing issues that other
plants have already solved
Not having specific Lessons Learned in a corporate database segregates the
plants from each other and if one plant has solved a machinery issue, the others
may not know the solution and continue to lose daily revenue.
BENCHMARKS
This Best Practice has been used by Forsthoffer Associates, Inc. since 1990 and
has been incorporated into companies with the following types of Plants with
the benefits listed previously:
l
l
MEGA Ethylene Plants
MEGA Butyl Rubber Plants
518
Predictive and Preventive Maintenance
l
l
Chapter | 11
Methanol Plants
MEGA LNG Plants
B.P. 11.5: Conduct in-house training using supervisors within a unit to
instruct young personnel on the importance of and how to perform key
PM tasks
There are an unbelievable amount of tasks that need to be performed daily
and among them are some key items to optimize the reliability of the machinery.
While most plants have programs to perform these tasks, many times they do
not get done on a regular schedule because of all of the urgent items going on
within the plant.
By having key supervisors explaining and showing (actually performing
these tasks) these tasks to young personnel, they will see the importance of
continuing to perform these certain tasks regularly.
Following are key machinery tasks that need to be done on a regular basis:
l
l
l
l
l
Pump Changeover—To assure spare pump is ready to start and components
are all in good condition
Steam Turbine Trip Valve Exercising—To make sure the valve stem can
move when it is required to
Accumulator Pre-charge Check—To assure the accumulator is available to
provide oil to its supported components for at least 4 s during a transient
situation
Check Transient Functionality of Oil System—to assure the console can
withstand a quick process change or a pump shutoff and not trip the train
Stroke check critical control valves periodically (usually during turnaround)
L.L. 11.5: Inability to implement programs to carry out regular machinery
PM tasks has resulted into numerous unplanned shutdowns and revenue lost
Recent (2010) involvement with a site root cause failure analysis for a steam
turbine overspeed incident required a review of all Predictive (Condition Based)
and Preventive (Time Based) maintenance procedures with the affected train.
While a program was initiated to exercise the Trip valve for the steam turbine,
it had actually not been performed for a few years and when it was needed to
close, it did not.
BENCHMARKS
This best practice has been recommended in the last 5 years when it was really
noticed that many plants lacked the implementation of many important machinery PMs. It has been found that by spreading the knowledge to the personnel
who has to perform the tasks of how important the tasks are in order to maintain
the highest safety and reliability standards, the better chance you will have to
regularly implement these tasks.
519
B.P. 11.6
More Best Practices for Rotating Equipment
B.P. 11.6: Assure all oil system and seal gas control valves have a means of
position indication
Every valve in the plant is basically a flowmeter since for a fixed position in
a service with the pressures upstream and downstream the same, the flow is constant. Therefore, any change in position is a change in flow and can tell you a great
deal about the components that system is supporting (i.e., Dry gas seals, bearings,
oil seals, etc.). In order for this to be done, a position indicator is required.
If available for the type of valve you want to monitor, it is always recommended to have an electronic position indication installed, so the position can
be captured and trended from start-up of the equipment.
L.L. 11.6: The inability to monitor control valve position in auxiliary systems has led to many surprises and replacements soon after a turnaround.
Monitoring of valve stem position would have identified worn components
and allowed replacement during a turnaround
Remember that turnaround action does not affect product revenue, unplanned action does!
Replacement of an oil pump can take 2 days considering alignment.
Replacement of a bearing or seal can take 3–5 days!
BENCHMARKS
This best practice, since the late 1980s, has saved many millions of USDs by
moving component replacement to the “Turnaround Revenue Loss Free Zone!”
SUPPORTING MATERIAL
Referring to the general definition of an auxiliary system which is to continuously
supply cool, clean fluid to each specified point at the required pressure, temperature, and flow rate, we can see that the controls and instruments play a major role
in the reliability of auxiliary systems. The function of the controls and instrumentation is to continuously supply fluid to each specified point at the required
pressure, temperature, and flow rate. While it is true that pumps and coolers must
be present, system controls modify the operational characteristics of these components to achieve the desired results. In addition, system instrumentation initiates transient system response, continuously monitors operation, and shuts down
critical equipment in the event of an auxiliary system malfunction. In this section,
we will examine important concepts that are at the heart of auxiliary system reliability, define the function of major control and instrumentation components, and
discuss items that can significantly reduce auxiliary system reliability.
Types
Types of major auxiliary system controls and instrumentation are outlined in
Table 11.6.1. Note that types are defined by function. As an example, a positive
520
Predictive and Preventive Maintenance
Chapter | 11
TABLE 11.6.1 Major Auxiliary System Controls and Instrumentation (By
Function)
Controls
Instrumentation monitor and alarm
• Positive displacement pump system
flow control
• System reservoir level
• Dynamic pump system flow control
• Pump operation
• Stand-by pump automatic start
• System pressure
• Cooler temperature control
• System temperature
• System differential supply pressure
control (constant reference pressure)
• Filter differential pressure
• System differential supply pressure
control (variable reference pressure)
• System differential pressure (variable
reference pressure)
• Variable speed pump driver speed
indicator
displacement pump system flow control consists of a pressure control valve that
bypasses excess flow from the pump back to the system reservoir to maintain a
set system pressure. The function of this component however is to continuously
supply the required flow of fluid to the system under varying system pressure
drops and critical equipment component conditions (worn bearing, seal, etc.).
All system controls and instrumentation must function perfectly under both
steady state and transient conditions. Under normal operation, a steady state control mode is approached since flows, pressures, and temperatures change very
slowly if at all. While this mode of operation may appear to be ideal, it can be
dangerous since control valves and instrumentation can bind up due to debris and
lack of movement. In the transient mode, components must have response times
on the order of milliseconds. When one considers the function of an auxiliary
system and the fact that the slowest of critical equipment units operate at approximately 60 revolutions per second (3600 RPM), the necessity of rapid system response time is appreciated. If the controls cannot respond to a transient response,
the instrumentation and the critical equipment shutdown system (circuit breaker,
steam turbine trip valve system, etc.) must operate on demand to stop equipment
operation. If the system controls and instrumentation do not have sufficient response times, a system liquid supply source (accumulator) is required to provide
flow during transient conditions. Using our system as an example, 60 gallons per
minute (GPM) are supplied to the unit or 1 Ga/s. Suppose the main pump trips
and the normal flow to the equipment is not reached for 3 s (until the stand-by
pump is at full speed and flow rate). An accumulator with a liquid capacity of 3
Ga would enable the system to function normally during the upset since it would
supply the required flow of 1 Ga/s. Note an accumulator size greater than 3 Ga
would be required. This will be covered separately.
521
B.P. 11.6
More Best Practices for Rotating Equipment
Concepts
The use of concepts can be helpful in understanding the function of auxiliary
system components and systems. In this section we will discuss:
l
l
l
l
l
An equivalent orifice
Sub-systems
An equivalent vessel
Control valve liquid coefficient—Cv
A flow meter in every system
The Concept of an Equivalent Orifice
Bearings seals, etc. can be reduced to the concept of an equivalent orifice
(Fig. 11.6.1). The equation for orifice flow is:
Q = C × Cf × D 2
∆p
S.G.
From the previous equation it can be seen that flow to any component is
the function of the dimension “D2” and ∆P across that component. The system
components essentially experience two types of flow changes: the gradual flow
change due to component wear (i.e., D2 change as in the case of bearing wear)
or the sudden flow change due to a pressure change in the system. As can be
seen from the previous equation, a sudden change of pressure as in the case of
FIGURE 11.6.1 Reduce it to an equivalent orifice.
522
Predictive and Preventive Maintenance
Chapter | 11
a hunting control valve or a sudden pressure spike due to component starting or
stopping will cause a corresponding sudden change in flow rate to the component. Considering the speeds involved in critical equipment, one can appreciate
that a short term transient flow change can lead to significant component damage of the critical equipment (bearing, seals, etc.) The previously mentioned
concept of reducing each individual critical equipment component (bearing,
seals, orifices, etc.) to an equivalent orifice helps enormously in conceptualizing
transient system reactions.
Sub-Systems
Both positive displacement and dynamic pumps alone do not contain the
­desired characteristics for operation within an auxiliary system. To achieve
the ­objectives of an auxiliary system, these components must be combined
in a ­controlled subsystem to achieve desired results. The sub-system is the
­combination of the pump and a control valve which together produce the flow
­characteristic required. Viewing components in control and instrumentation as
being part of various sub-systems also helps in understanding the total function
of auxiliary systems.
Equivalent Vessel
Refer to Fig. 11.6.2. Systems and sub-systems can be reduced also to that of
equivalent vessels. As an example, the supply pipe from a lube oil console can
be reduced to an equivalent vessel as shown in Fig. 11.6.2.
FIGURE 11.6.2 Reduce it to an equivalent vessel.
523
B.P. 11.6
More Best Practices for Rotating Equipment
When supply flow equals exit flow, the pressure in any equivalent vessel
remains constant. If supply flow is less than exit flow, the pressure reduces
rapidly. The function of an accumulator can be understood easily by using this
equivalent vessel concept. If a vessel is installed downstream of the equivalent
vessel in Fig. 11.6.2, during the period of reduced inlet flow the vessel would
supply flow to the system. This is exactly the function of an accumulator.
Another example of using the equivalent vessel concept is as follows: Imagine again the equivalent vessel is a supply pipe from a lube oil console. Suppose
the main pump trips on overload and the auxiliary pump does not start immediately. Since the auxiliary pump did not immediately start, the supply flow to
the equivalent vessel is less than the exit flow. As a result, the pressure in the
equivalent vessel will drop. This is why pressure switches in auxiliary systems
are used as alarm, auxiliary pump start or trip devices. Using our concept of an
equivalent vessel it can be seen that the pressure switch actually acts as a flow
indicator and will activate on low flow even though it is measuring pressure.
Control Valve Liquid Sizing Coefficient—Cv
“Cv” is an important concept that must be understood when dealing with any
type of control valve on liquid service. Cv “valve sizing coefficient” is defined
by the following equation:
Cv = Q (GPM)
S.G.
∆P
where S.G. (specific gravity) = 0.85 (for oil); ∆P = value pressure drop (P.S.I.).
Solving this equation for GPM we see that:
Q (GPM) =
Cv
S.G.
∆P
We can see referring back to “The concept of an equivalent orifice” that this
equation is similar to that of an orifice. Naturally the only difference is that a
valve is a variable orifice. Valves are sized using this concept of Cv (valve coefficient). Each valve has a maximum Cv. Depending on the type of internal valve
design, seats, plugs, and body, a valve will exhibit a certain characteristic. Refer
to Fig. 11.6.3 which is a graph of valve characteristics. Plotted on the ordinate
(Y axis) is valve flow in percent of maximum flow and plotted on the abscissa
(X axis) is travel of the valve plug in percent of rated travel. As we cover specific valve application later in this section, the characteristics of particular valves
will be discussed. Referring back to the relationship for valve coefficient, it can
be seen that the valve coefficient is dependent on flow rate, differential pressure
across the valve, and fluid characteristic.
524
Chapter | 11
Predictive and Preventive Maintenance
FIGURE 11.6.3 Control valve flow characteristics. (Courtesy of Fisher Controls Inc.)
As an example, suppose that a valve is sized to pass 20 GPM under normal conditions of 150 PSI pressure drop. The fluid in this case is light turbine
oil at 150°F (60 SSU). Solving for the valve Cv as per the earlier equation,
we arrive at a figure of 1.51. If the valve pressure drop were to decrease to
100 lbs, and we still required 20 GPM to pass, the valve coefficient would
be 1.84. This change represents approximately a 22% change in the valve
coefficient. Depending on the characteristic curve of the valve in question, it
would represent a given amount of valve plug opening (increase of travel).
In the same example, now let us assume that the flow changes to 40 GPM
with 100 lbs pressure drop across the valve. The Cv now would be 3.69 or
approximately 200% the previous value. Depending on the valve size, this
coefficient may or may not be obtainable. Refer to Table 11.6.2 which is a
typical valve coefficient table showing valve coefficients for percent travel of
TABLE 11.6.2 Typical Liquid Valve Sizing Coefficient Table
% Travel
(12.5%)
(25.0%)
Valve Travel
Body size
Port size
(50%)
(75%)
(100%)
*
*
1/32 in.
1/16 in.
1/8 in.
3/16 in.
1/4 in.
1 in.
3/4
1.4
3.1
4.2
5.3
7.0
1 in.
1 in.
2.4
4.2
7.0
10.0
12
525
B.P. 11.6
More Best Practices for Rotating Equipment
a particular valve. When sizing all control valves, Cv maximum, Cv normal,
and Cv minimum must be calculated. A general rule is that all of the previous
values should fall between 10% and 90% of the maximum Cv for a particular
valve selected.
When dealing with viscous liquids as in the case of oil, valve coefficient
viscosity corrections must be made. For the example case mentioned earlier,
the correction factor for 220 cSt (1000 Sabolt Universal Seconds viscosity SSU)
would be approximately 1.5–2. Therefore the valve coefficient required would
be 1.5–2 times that required at normal viscosities (60 SSU for light turbine oil at
normal operating temperatures). Viscosity correction nomographs are available
from control valve manufacturers for determining valve sizes required under
high viscosity conditions.
A Flow Meter in Every System
Considering the relationship discussed previously it can be seen that every control valve can be considered as a flow meter if the fluid differential pressure
across the valve, valve travel, and a valve characteristic chart is known. While
not a completely accurate flow measuring device, this concept can be extremely
valuable while troubleshooting auxiliary systems. Obtaining the valve travel
and using the valve coefficient chart, the Cv can be obtained. Calculating for
GPM knowing the Cv, the pressure drop across the valve, and the specific gravity of the liquid can then yield the flow rate. It is important to note that with small
valve travels on the order of 1/4 in. maximum, an accurate means of measuring valve travel must be obtained. It is the writer’s experience that many times
travel indicators are not furnished with the valve. It is strongly recommended
that valve travel indicators be supplied or retrofitted in the field.
Bypass Control
The first application to be discussed in this section will be that of a bypass control valve. A bypass control valve and actuator pictured in Fig. 11.6.4 is used
with a positive displacement pump to alter the pump’s flow characteristic to that
of variable flow.
Refer to the schematic of a lube oil system typical of the example in
Fig. 11.6.5. This system incorporates positive displacement pumps. The control valve’s function is to continuously control flow to the critical equipment
such that the required flow is supplied under normal and transient conditions.
Since a positive displacement pump essentially is a constant flow device, the
control valve in the bypass mode must allow for excess pump flow to be recirculated back to the reservoir. Utilizing the concept of an equivalent orifice, as the bearings in the system wear, the orifice diameter becomes larger,
therefore the flow required to the critical equipment will be greater. Since the
downstream pressure across the bearings is atmospheric pressure, the upstream
526
Predictive and Preventive Maintenance
Chapter | 11
FIGURE 11.6.4 Reverse acting actuator and valve body typically used as a back pressure
regulator (bypass control). (Courtesy of Fisher Controls Inc.)
pressure will initially decrease when the bearing area becomes larger for the
same flow. The bypass valve will sense the upstream pressure reducing and
will close to force the additional required flow to the critical equipment. Even
though the bypass valve is a pressure device, it’s acting as a flow control device to divert bypass flow to required system components. Therefore using the
527
B.P. 11.6
More Best Practices for Rotating Equipment
FIGURE 11.6.5 Typical lube oil supply system.
concept of a sub-­system, the bypass valve and the positive displacement pump
form a variable flow sub-system that will supply variable flow to the critical
equipment on demand.
In addition to accounting for small changes in system flow requirements,
the bypass control valve must also act under transient conditions. If the main
pump were to suddenly shut off, the system would immediately sense a pressure decrease. Referring to the equivalent orifice concept of bearings, the flow
to these components would drop proportionately to the square root of the pressure drop across the component. At hundreds of revolutions per second, the
bearing shaft interface would not last long with the absence of system flow. In
this transient mode, the control valve must close quickly to divert all bypass
528
Predictive and Preventive Maintenance
Chapter | 11
flow to the system to account for the absence of flow from the pump. The control valve characteristic, its actuator and supply to its actuator whether direct
(hydraulic) or indirect (pneumatic) must function instantaneously. If the valve
system experiences instabilities or excessive friction, as in the case of valve
stem binding, the system will experience an instantaneous loss of flow and will
(hopefully) be shut down on this signal. Again referring to concepts discussed
earlier, the concept of an equivalent vessel is useful in ascertaining how pressure and flow are related and why pressure switches are used to determine loss
of flow under transient conditions. This concept also shows why time delays
in auxiliary systems are not desired to be used with any trip devices. It’s true
that a time delay would preclude a trip of the unit under transient conditions
but could also cause severe and perhaps catastrophic damage to the critical
equipment.
The bypass control valve also must exhibit rapid transient response in the
open direction. In the case of dual pump simultaneous operation, the amount
of flow to be recirculated to the reservoir will be equal to the normal bypass
flow of one pump plus the full flow of the stand-by pump. If the bypass valve
does not act as a variable orifice and opens at a slower rate than the flow rate
increase, referring to the orifice equation, the pressure drop across the valve
will simultaneously rise. This increase may exceed the setting of the relief
valve in the system. If this is the case, the system is exposed to the potential
of the relief valve not re-seating. If this were to occur, a new “orifice” would
be introduced into the system and the flow to the critical equipment would be
reduced to the point of requiring the stand-by pump to start and possibly causing critical equipment shutdown. In order to meet the previously mentioned
control and transient requirements, the bypass control valve must be sized
properly. An example of valve sizing using the system shown in Fig. 11.6.5
is shown in Table 11.6.3. We wish to reemphasize that once the valve is sized
properly, the actuator and the sensing lines in the system that supply the force
to operate the valve must be designed for rapid response. In many systems,
sensing line snubbers are used to dampen impulse signals that can lead to
valve instability. It must be noted that snubbers are designed to provide quick
response in one direction and retarded or slower response in another. It is of
extreme importance that these devices be installed properly. Understanding
the function of the particular valve in question and examining the direction
of the snubber device in a sensing line is essential to correct system operation. Many times these snubber devices are installed improperly in the wrong
direction.
Pressure Reducing Control
Pressure reducing control has two primary applications in auxiliary systems.
l
l
To control the flow from a dynamic pump.
To reduce the pressure in the system.
529
B.P. 11.6
More Best Practices for Rotating Equipment
TABLE 11.6.3 Valve Sizing Example—Back Pressure (Bypass) Control
Given:
1. Normal valve flow
= pump flow − normal system requirement
= 73 GPM − 60 GPM
= 13 GPM
2. Maximum valve flow
= main and auxiliary pump flow − minimum
system requirement
= 146 GPM − 60 GPM
= 86 GPM
3. Maximum valve P
= pump discharge pressure @ maximum supply
flow and component P
= 250 Psig
4. Minimum valve P
= pump discharge pressure @ minimum supply
flow and component P (clean system)
= 160 Psig
5. Oil specific gravity
= 0.85
Determine:
Cv Minimum
Cv Maximum
1. Cv Min.
= Q NORMAL
S.G.
∆P Max.
= 13 × 0.0583
= 0.758
2. Cv Max.
S.G.
∆P Min.
= 86 × 0.0729
= 6.268
= Q MAX.
Refer to Table 11.6.2 for 1-in. valve with ¾-in. port and obtain:
Valve maximum
Cv = 7.0
Valve operating maximum
Cv = 6.268
Valve operating minimum
Cv = 0.758
Valve maximum travel (opening)
= 90%
Valve minimum travel (opening)
= 9%
Note: Valve minimum and maximum openings are at the limit for satisfactory operation.
A typical pressure reducing control valve and actuator are shown in
Fig. 11.6.6.
For the first case, the flow characteristic is variable. The flow is therefore determined by the pressure at the discharge flange of a dynamic pump. A pressure
reducing valve set to sense the pressure downstream of the valve will automatically regulate the discharge or the back pressure on the dynamic pump for the
530
Predictive and Preventive Maintenance
Chapter | 11
FIGURE 11.6.6 Direct acting actuator and valve body used for PRV (pressure reducing)
control. (Courtesy of Fisher Controls Inc.)
desired flow of the system. Referring back to the equivalent orifice concept, if
a bearing were to wear, the equivalent diameter of the orifice would increase.
Therefore, initially for the same flow rate, the pressure in the system would decrease since the flow is the same. If the bearing clearance increases (equivalent
D), the ∆P must decrease.
531
B.P. 11.6
More Best Practices for Rotating Equipment
Therefore the pressure control valve sensing decreasing system pressure
will open to increase the system pressure. This action will result in a decrease of resistance on the dynamic pump discharge flange and allow the
centrifugal pump to operate at a greater capacity to provide the desired flow
to the critical equipment. It can be seen that in this case the pressure reducing valve and the dynamic pump combine for a sub-system that meets the
objective of providing continuous flow to the critical equipment. The control
valve essentially renders the variable flow, constant head device and variable
head device by compensating for changes in system pressure. The previously
mentioned case represents the normal control case. Lets now examine the
transient case.
If the main dynamic pump were to suddenly trip, the system pressure
will suddenly fall as a result of greater flow exiting the system than the
amount of flow entering the system (equivalent vessel concept). In this case,
the pressure reducing valve sensing downstream pressure would instantaneously open allowing the dynamic pump to move out to a higher flow point
on its curve while the auxiliary or stand-by pump were to start. As soon
as the auxiliary pump starts the pressure reducing valve sensing additional
flow into a fixed system resistance would then close meeting the flow requirements. Dynamic systems in general tend to be somewhat softer than
positive displacement systems. That is, they are more tolerable to transient
system changes.
In the case of the auxiliary stand-by pump and the main pump operating
simultaneously, the pressure reducing valve would automatically compensate
for the increased flow by reducing its travel or increasing the system resistance at the discharge flange of both pumps. That is, increasing the discharge
pressure to the level where the combined flow of both pumps would exactly
equal the critical equipment system required flow. Again referring to the concept of an equivalent orifice, if excessive flow were forced through the orifice
(the bearings) the pressure drop would increase. The pressure reducing valve
sensing the increased system pressure drop would tend to close to reduce the
pressure at its sense point. In doing so it will increase the discharge pressure on both dynamic pumps and since their characteristic is reduced flow
on increased pressure, the desired flow will be obtained. Therefore it can be
seen again that the dynamic pump pressure reducing valve sub-system has the
function of flow control to the critical equipment even though it is sensing
pressure.
The other primary application for pressure reducing valves in auxiliary
systems is for reducing system pressure to other desired pressure levels. Refer to Fig. 11.6.5 and observe the pressure reducing valve at the discharge of
the lube oil system. Its function is to reduce pressure from control oil pressure to lube oil pressure. Control oil pressure is controlled by the equivalent
orifice in the control system and the set point of the bypass control valve as
shown in Fig. 11.6.5. The bypass control valve senses pressure and controls
532
Predictive and Preventive Maintenance
Chapter | 11
flow to satisfy the requirement of the equivalent orifice in the control system and the equivalent orifices in the lube system. The pressure reducing
valve simply senses pressure downstream of the valve and controls it to the
preset value. It should be noted that in most auxiliary systems, the console
(reservoir, pumps, etc.) is usually below the level of the critical equipment,
therefore the set point of any pressure reducing valves on the console should
compensate for the height or head difference between the console and the
critical equipment. Control valves used in pressure reducing service usually
are not exposed to system transient changes as in the case of bypass valves.
Therefore their sizing is relatively easy and their valve Cvs do not significantly change. A sizing example for a direct acting pressure reducing valve
is shown in Table 11.6.4.
TABLE 11.6.4 Valve Sizing Example—Pressure Reducing Control
Given:
1. Minimum and normal lube oil flow to unit
= 60 GPM
2. Maximum lube oil flow to unit (bypass
valve failed closed)
= 73 GPM
3. Valve ∆P
= 120 PSIG − 25 PSIG
= 95 PSI (15 PSIG supply + 20 PSIG)
pressure drop for elevation
Note: This example is for a PRV located on the lube oil console at grade.
4. Oil specific gravity
Determine:
= 0.85
Cv Normal
Cv Maximum
1. Cv Normal
S.G.
Normal
∆P
= 60 × 0.0946
= 5.675
2. Cv Maximum
S.G.
Maximum
∆P
=73 × 0.0946
= 6.906
=Q ×
=Q ×
Refer to Fig. 11.6.2 for a 1.inch valve with a 1.inch port and obtain valve maximum
Cv = 12.0
Valve operating normal
Cv = 5.675
Valve operating maximum
Cv = 6.906
Valve normal travel (opening)
= 40%
Valve maximum travel (opening)
= 50%
533
B.P. 11.6
More Best Practices for Rotating Equipment
Temperature Control Valves
Temperature control valves are usually required in auxiliary systems to regulate the supply temperature to the critical equipment components. Especially in
systems where liquids have viscosity characteristics (oil systems), temperature
control is important to insure correct oil viscosity to components. Referring to
concepts previously discussed in this section, the temperature control valve plus
the system coolers make up a cooling sub-system whose function is to continuously supply the required fluid to critical equipment at a specified temperature.
Two types of control valves that are used are direct acting three way valves and
air operated two way valves. Both valves sense the mixed temperature downstream of the cooler.
A two way valve is a simple bypass around the cooler while a three way
valve is a true mixing valve. It should be noted that when sizing a two way
valve, the pressure drop across the cooler must be known to assure that the valve
coefficient is large enough to pass the required flow. Many systems using two
way valves are insufficiently sized. This can result in cooler oil constantly being
supplied to the system since the pressure drop through the cooler is less than the
minimum pressure drop through the control valve.
Instrumentation
The instrumentation in any auxiliary system is extremely important in assuring
quick system response, accurate monitoring of system condition, and rapid system shutdown in the event of upsets. In this section we will examine stand-by
pump start-up operation, critical equipment shutdown, and monitoring functions of the instrumentation in the auxiliary system.
Stand-By Pump Automatic Start
As previously mentioned, interruption of pump flow to the critical equipment results in a rapid deterioration of system flow and pressure. Referring
again to the concept of an equivalent vessel, the absence of inlet flow to the
system while exit flow is continuing will instantaneously produce a pressure
drop. This concept is utilized in using a pressure switch to signal the immediate start of the stand-by pump. Practice has shown that locating the pressure
switch takeoff as close as possible to the pump discharge results in the quickest response time to initiate stand-by pump start. Some systems incorporate
dual pressure switches, one close to the pump discharge and another up close
to the critical equipment. Both switches start the stand-by pump on signal.
The pressure setting of the switch is usually set just below the lowest discharge pressure that the pump will produce. In order to insure the rapid start
of the auxiliary pump when required, many systems incorporate an on–off
automatic switch on the auxiliary pump or on the stand-by pump motor starter
for testing the system. It is extremely important that the position of the switch
always be in the automatic mode during critical equipment operation. It is
534
Predictive and Preventive Maintenance
Chapter | 11
recommended that an alarm be supplied and annunciated in the event that the
auxiliary or stand-by pump is not in the auto position during critical equipment operation.
Critical Equipment Trip Instrumentation
A general critical equipment design philosophy is to avoid trips circuits as
much as possible. That is, to only install trip switches in those situations
that are absolutely necessary. Typically, auxiliary systems incorporate only
one-trip function. As an example, the low lubricating oil trip in lube systems
and the low seal oil differentials trip in seal systems. Sometimes a high temperature switch is also installed to trip, but this is not usually the case. The
setting of the trip switch is very important. It must be selected such that the
equipment will shut down when actuated in order to prevent any long term
damage.
It must also be selected to prevent spurious, unnecessary trips of the unit
since they are extremely costly in loss of revenue. In addition, the quality of
any trip switch is extremely important since this relatively low cost device
could cost millions of dollars of lost product revenue per day in the event of
a malfunction. Attention is drawn to correct selection of switch component
materials to prevent corrosion or any abnormality that would cause drifting of
switch setting and unnecessary unit shutdowns. Again the concept of a system
is extremely important to consider. It must be remembered that the trip switch
and the shutdown system for the critical equipment together must function accurately in order to terminate equipment operation immediately upon signal
from the initiating trip switch. “Best practice” is to use smart (self-diagnostic)
triple redundant (two out of three voting) transmitters for all pump start and
trip services.
Auxiliary System Monitoring
Refer to Table 11.6.1 in this section and observe the different monitoring and
alarm functions normally used in an auxiliary system. In order to insure reliable auxiliary system operation, the personnel must continuously observe and
record any changes in instrument readings and promptly attend to alarms to
insure that the system continues to operate as required. Changes in any of the
system instruments indicates a change in the operating condition of the system and must be followed through to insure that components are operating as
required. As an example, slowly deteriorating lube oil supply pressure could
indicate either a valve malfunction, reduction in speed of a main turbine pump
driver, excessive pressure drop in the system oil filter, or many other types of
problems. It is extremely important to maintain a program of auxiliary system
instrumentation calibration to assure all instruments are reading properly. This
will aid greatly in determining malfunctions of the system and assist in the site
PM program.
535
B.P. 11.7
More Best Practices for Rotating Equipment
B.P. 11.7: Check and confirm oil system relief valve settings on the console
during a turnaround
The relief valves in all oil systems have an accumulation buildup until they
are fully open and need to be set so they are fully open at the required pressure.
This is a task that can only truly be done on the console and is something that
most people do not want to do while in operation.
Therefore, utilize the time when you are down for a turnaround or unit shutdown to confirm these settings are correct.
L.L. 11.7: Many unit trips have been traced back to improper setting of relief
valves that caused them to open at lower than set pressures, which required the
auxiliary pump to start. Starting of the auxiliary pump was either too late or
caused control valve instability resulting in a low oil pressure trip and a unit trip
BENCHMARKS
This best practice has been used since the early 1980s and has resulted in plant
oil system and unit reliabilities above 99.7%.
SUPPORTING MATERIAL
Relief Valves for Positive Displacement Pumps
Since positive displacement pumps are not self-limiting, that is, they can produce increasing pressure if sufficient driver power is available, a device to limit
pump pressure and horsepower is required.
The function of a relief valve as a protection device is to limit pump discharge pressure and horsepower to a specified value without generating any
valve instabilities and to positively reseat. While the function of a relief valve
is simple enough, valve chatter (instability) and failure to positively reseat can
cause the shutdown of the critical equipment. Relief valve chatter can cause
high pressure pulses that will activate shutdown pressure switches and damage
valve seats and plugs.
The inability to reseat properly will introduce an “equivalent orifice” into
the system that will reduce or totally eliminate the system flow to the critical
system components.
Experience has shown that a sliding piston type relief valve, which is a modulating device, as opposed to a spring loaded poppet valve, which is an on–off
device, meets the requirements of stability and positive shutoff for liquid auxiliary system service. A typical relief valve used is shown in Fig. 11.7.1. A sizing
chart for this type of relief valve is shown in Fig. 11.7.2. Relief valve set pressure
is usually set 10% above the pump maximum discharge pressure. However, the
maximum pressure ratings of all system components must also be considered.
Given the maximum pump flow and the relief valve set pressure, the maximum system pressure can be determined as follows:
536
Predictive and Preventive Maintenance
Chapter | 11
FIGURE 11.7.1 Modulating relief valve. (Courtesy of Fulflow Specialties Co. Inc.)
Maximum system pressure = Relief valve set pressure + Relief valve overpressure.
Relief valve overpressure is the valve pressure drop necessary to pass full
pump flow.
For the present example using a 2 in. valve:
1. Maximum pump discharge pressure = 200 PSIG
2. Relief valve set pressure (cracking pressure) = 1.1 × 200 = 220 PSIG
3. Maximum system overpressure = 220 PSIG + 25 PSIG = 245 PSIG (from
Fig. 11.7.2 for Y spring and 86 GPM flow)
537
B.P. 11.7
More Best Practices for Rotating Equipment
FIGURE 11.7.2 Relief valve sizing chart. (Courtesy of Fulflow Specialties Co. Inc.)
Note that the overpressure values are viscosity sensitive and can be used up
to a viscosity of 500 SSU. Above this value, the overpressure can be estimated
to vary by the relationship:
overpressure@viscosity = overpressure@ 500 SSU
×
4
viscosity
500 SSU
1000
500
= 35 psi × 1.19
= 42 psi
= 35 psi
4
Therefore, the maximum pressure at 1000 SSU will be 262 psi or 19%.
Relief valve overpressure expressed as a percentage of relief valve set point
is defined as accumulation. Typical values of accumulation vary between 10%
and 20%.
538
Predictive and Preventive Maintenance
Chapter | 11
B.P. 11.8: Check the function of all main oil pump steam turbine (if you
have one) components during turnaround
Items that need to be checked are as follows:
l
l
l
Governor linkage can freely move (not seized)
Replace carbon ring seals and clean cups to assure no dirt can keep the rings
from “floating” freely
Change bearing oil and install oil condition monitoring bottle if not already
there
L.L. 11.8: Steam Turbines used for main oil pump drivers have the lowest
reliability of oil system components and have been responsible for many oil
system trips
BENCHMARKS
This best practice has been used since 1990 to produce oil systems of highest
reliability, which has resulted in unit reliabilities above 99.7%.
SUPPORTING MATERIAL
Single Stage Turbine Guidelines
The five common problems with single stage turbines are noted in
­Table 11.8.1.
We will now discuss each problem in detail. Please refer to Fig. 11.8.1 which
has each problem area circled.
Bearing Bracket Oil Contamination
Please refer to Item 1 in Fig. 11.8.1.
The most common reliability problem with single stage steam turbines is the
contamination of the oil in the bearing housing with water. The root cause of
the problem is the ineffectiveness of the floating carbon ring shaft seal system
to stop.
TABLE 11.8.1 Single Stage Steam Turbines Common Reliability Problems
• Bearing bracket oil contamination (inadequate carbon ring steam seal design)
• Slow governor system response (inadequate governor linkage maintenance and
governor power)
• Hand valve(s) closed on critical services
• Bearing bracket oil viscosity reduction and bearing wear (high pressure service)
• Use of sentinal valves on turbine cases
539
B.P. 11.8
More Best Practices for Rotating Equipment
TABLE 11.8.2 Bearing Bracket Oil Contamination (Root Cause)
• Shaft carbon ring seal cannot positively prevent steam leakage
TABLE 11.8.3 Steam Turbine Bearing Bracket Oil Contamination
Monitoring Action Plan
•
•
•
•
Install oil condition site glasses in bearing bracket drain connection
Inspect once per shift
Drain water as required
Sample oil monthly initially
TABLE 11.8.4 How to Correct Carbon Ring Seal Ineffectiveness?
• Install steam eductor on each seal chamber leak off drain (between 4th and 5th
carbon ring)
• Design eductor to pull 5–10 in. of H2O vacuum at this point
• Alternative approach—install bearing housing isolation seal (“Impro” or equal)
Unless present site systems are modified to eliminate the root cause, the
best action plan is to minimize the effect of the contamination so a bearing failure will not occur. Such an action plan is presented in Tables 11.8.2–
11.8.4.
Slow Governor System Response
Please refer to Item 2 in Fig. 11.8.1. Another very common reliability problem
is the slow or non-movement of the governor system linkage during start-up
and normal operation during steam condition changes. It will appear that the
governor is not responding because speed will not be controlled when it should.
Typical examples are:
l
l
Speed will continue to increase when throttle valve is opened, turbine will
trip on overspeed
Speed will increase or decrease when:
l Steam conditions change
l Driver equipment changes
These facts are presented in Table 11.8.5.
540
Predictive and Preventive Maintenance
Chapter | 11
541
FIGURE 11.8.1 Single stage steam turbines common reliability problems.
B.P. 11.8
More Best Practices for Rotating Equipment
TABLE 11.8.5 Slow Governor System Response
1. Rapid speed change and trip on start-up
2. Speed increase or decrease on steam condition or load condition change
3. Governor instability (hunting) around set point
Note: #1 usually occurs on “solo,” #2 occurs during steady state operation.
TABLE 11.8.6 Slow Governor System Response Condition Monitoring
Action Plan
• Install tachometer on all single stage steam turbines
• Always test speed control on “solo run” (1)
• Monitor turbine speed once per shift. Take corrective action if speed varies ±5%
(200 rpm)
Note: (1) since load is very low, test acceptance is the ability to stabilize speed and prevent
overspeed trip when throttle valve is slowly opened.
Since most single stage steam turbines are not supplied with tachometers,
it is difficult, if not impossible to condition monitor this problem. A condition
monitoring action plan is provided in Table 11.8.6.
The usual root cause of the problem is that the friction in the mechanical
linkage and/or valve stem packing exceeds the maximum torque force that the
governor output lever can deliver. The governor designations TG-10, TG-13,
and TG-17 simply mean “turbine governor with … FT-LB torque.” Therefore,
if a TG-10 governor is installed and the torque required to move the valve steam
exceeds the value of 10 FT-LBs, the governor system will not control speed.
Taking the governor to the shop will not solve the problem. Causes of excessive
friction are shown in Tables 11.8.7 and 11.8.8.
Hand Valve(s) Closed on Critical Services
Most single stage steam turbines are supplied with one or more hand valves
in the steam chest. Refer to Fig. 11.8.1, Item 3. The purpose of the hand
valves is to allow more or less inlet steam nozzles to be used during operation. Optimizing the steam nozzles used, maintains turbine efficiency during
load changes. However, the efficiency of single stage steam turbines is only
35% at best! Therefore, adjustment of hand valves, other than during start-up
or during slow roll, should not be required. Fig. 11.8.2 is a top view of a one
hand valve.
We have witnessed many unscheduled shutdowns of critical (unspared)
compressor units because the general purpose steam turbine that is the main
542
Predictive and Preventive Maintenance
Chapter | 11
TABLE 11.8.7 Causes of Excessive Governor Mechanical Linkage System
and Valve Friction
•
•
•
•
Linkage bushings not lubricated with high temp. grease
Valve steam packing too tight
Steam deposits in valve and/or packing after extended shut down (turbine cold)
Bent steam valve stem
TABLE 11.8.8 Slow Governor System Response Condition Monitoring
Action Plan
• If problems occur (Table 11.8.7), disconnect linkage and confirm ease of valve
movement
• Replace bushings and/or lubricate with “molycote” or equal
• Clean deposits from valve and packing as required
• If above action does not correct problem, replace governor (inspection and/or
­adjustment of governor droop is required)
FIGURE 11.8.2 Single valve turbine admission path.
543
B.P. 11.8
More Best Practices for Rotating Equipment
TABLE 11.8.9 Single Stage Steam Turbine Hand Valve Recommendations
• Never throttle hand valves
• Hand valves should be open on main oil pump and auto-start steam turbines
lube oil pump driver, had the hand valves closed. An upset in the steam system
reduced steam supply pressure and caused the turbine and lube pump to slow
down. This was because, hand valves were closed and the throttle valve, even
when full open, could not meet steam flow requirements. When the speed of the
steam turbine decreased, the lube oil pressure dropped and guess what? … The
auxiliary pump did not start in time and the unit tripped.
Table 11.8.9 presents the recommended action plan in the refinery for single
stage steam turbine hand valves.
Bearing Bracket Oil Viscosity Reduction and Bearing Wear on
High Pressure Single Stage Steam Turbines
Please refer to Fig. 11.8.1, Item 4. Observe the jacket in the bearing housings.
The purpose of this jacket is to cool the oil in the bearing bracket. When the
inlet steam pressure is high, the high temperature of the steam is transmitted to
the steam end inlet bearing through the shaft. Although the jacket in the bearing
housing does reduce the oil temperature in the bearing housing, it cannot effectively reduce the oil temperature at the shaft/bearing interface. Table 11.8.10
presents these facts.
This problem is a design issue. A small single stage turbine is not provided
with an effective oil system to remove the heat between the shaft and bearing
when the turbine is operating on high temperature (up to 750°F) steam. The
solution is to require pressure lubrication for this application.
Naturally, it is difficult and not cost effective to retrofit these turbines for
pressure lubrication. The field proven solutions to this problem are presented
in Table 11.8.11.
TABLE 11.8.10 High Pressure Single Stage Steam Turbine Bearing Problems
and Oil Viscosity Reduction
• Sleeve bearings (usually steam inlet end) wear out quickly
• Oil viscosity is reduced and difficult to maintain
544
Predictive and Preventive Maintenance
Chapter | 11
TABLE 11.8.11 Eliminate Bearing Wear and Oil Viscosity Reduction (High
Pressure Service)
• Assuring bearing housing jacket passages are open (flushed)
• Consulting with turbine vendor for bearing material change
• Using special high temperature service oil (synthetic based oil)
TABLE 11.8.12 Prevent Excessive Sentinel Valve Maintenance
• Removing sentinel valves
• Assuring that inlet and exhaust casings are protected by properly sized and set
­pressure relief valves
Continued Use of Sentinel Valves on Turbine Cases
Please refer to Fig. 11.8.1, Item 5. Sentinel valves were used, years ago, as
alarm devices indicating that the steam turbine case (low pressure part) was
under excessive pressure.
These devices are not pressure relief valves and will not protect the case
from failure during over pressure events.
It is a known fact that the sentinel valves wear, leak, and require steam turbine shutdown for repair. Most large company specifications prevent the use of
sentinel valves and require full relief valve protection on the inlet and exhaust of
all single stage turbines. These facts are presented in ­Table 11.8.12.
B.P. 11.9: Performance monitoring should be the responsibility of the machinery reliability department
The easiest way to implement this best practice is by including an experienced process engineer in the reliability group to conduct this task. This will do
the following:
l
l
l
l
Bring “process awareness” into the program to significantly increase machinery reliability.
Increase operator and process engineer awareness of machinery reliability
key factors.
Increase recommendation implementation by having process and operations
support.
Minimize “finger pointing” between maintenance and operations.
Remember that the rotor is in direct contact with the process and if the process changes the condition of the rotor can change. Also, remember what is
545
B.P. 11.9
More Best Practices for Rotating Equipment
connected to the rotor (bearings, seals, and auxiliaries that support these components), so changes in the rotor can affect all of the other major components
that the reliability group generally monitors.
L.L. 11.9: Reliability groups not incorporating operations and process engineering input produce lower machinery MTBFs and less implementation
of recommendations
Maintenance centered plant reliability programs produce lower machinery
MTBFs than reliability programs that have integrated maintenance, operations,
and process engineering functions.
BENCHMARKS
This best practice has been recommended since the mid-1990s when the company was involved with a number of site machinery audits in refineries and gas
plants. Implementation of this best practice led to significant improvement of
machinery reliability plant wide.
SUPPORTING MATERIAL
See B.P. 11.1 for supporting material.
546
Chapter 12
Reliability Optimization
B.P. 12.1: Establish a methodology for identifying plant bad actors.
It is important to have a plant wide methodology that is followed across the
board to identify which equipment or specific components are bad actors as you
can only fix what you know is not operating properly.
See supporting material for details on the following terminology:
l
l
l
l
Mean Time Between Failure (MTBF)
Mean Time To Repair (MTTR)
Reliability
Availability
L.L. 12.1: Inability to identify the plant bad actors accurately can result in
recurring failures.
BENCHMARKS
This best practice has been in use since 1990 and has resulted in a significantly
increased recommendation implementation rate (above 50%) and increased machinery safety and reliability.
SUPPORTING MATERIAL
Determining and Measuring Availability
Once information regarding failures and repair times is gathered and analyzed,
MTBF, failure rate, MTTR, and availability can be determined.
MTBF
Mean time between failure is determined by dividing the total operating time for
the period to be analyzed by the number of failures in that time period. MTBF
can be determined for a unit, a specific piece of equipment or a component. The
relationship is noted in Table 12.1.1.
More Best Practices for Rotating Equipment. http://dx.doi.org/10.1016/B978-0-12-809277-4.00012-7
Copyright © 2017 Elsevier Inc. All rights reserved.
547
B.P. 12.1
More Best Practices for Rotating Equipment
TABLE 12.1.1 Mean Time Between Failure (MTBF)
MTBF =
TOTAL OPERATING HOURS
NUMBER OF FAILURES
As an example, determine the MTBF for an LNG circulating pump given
the following data:
1.
Operating period 1990–92
2.
Year
Operating hours
Failures
1990
1991
1992
8600
8000
8500
2
4
1
MTBF =
25,100h
7 FAILURES
MTBF = 3586h
Failure Rate
Failure rate is the number of failures per machine year. In other words, it is the
reciprocal of MTBF. Table 12.1.2 presents failure rate.
For the same example, the failure rate for the LNG circulating pump is:
F.R. =
1
= 2.789 × 10 −4 per hour
3586
MTTR
Mean time to repair is the total time to repair a unit, equipment item or
component during a specific time period divided by the number of repairs
(Table 12.1.3).
TABLE 12.1.2 Failure Rate
THE NUMBER OF FAILURES PER MACHINE YEAR
OR FAILURE RATE =
548
1
MTBF
Chapter | 12
Reliability Optimization
TABLE 12.1.3 Mean Time to Repair (MTTR)
MTTR = TOTAL NUMBER OF REPAIR HOURS FOR A SPECIFIC
• UNIT
• EQUIPMENT ITEM
• COMPONENT
DIVIDED BY NUMBER OF REPAIRS
As an example, determine the M.T.T.R. for the following MS 5038 General
Electric Gas Turbines during a 24 month period as noted below.
Repair
Date
Repair description
Total hoursa
1
1/1/93
Replace H.P.T. nozzles
96
2
3
4
5
6
4/8/93
7/20/93
12/23/93
5/6/94
11/15/94
Replace fuel nozzles
Replace No. 1 & 2 bearings
Replace P.T. rotor and bearings
Replace H.P.T. nozzles
Replace compressor (L. P.)
rotor, stators and bearings
TOTAL MAINTENANCE HOURS =
NUMBER OF REPAIRS =
72
30
36
80
80
394
6
a
Includes cool down time.
394h
6 REPAIRS
= 65.67h
MTTR =
Availability
Availability is a more effective measurement of reliability since availability is
the percentage of time that a unit or equipment item operates compared to the
time it is available to operate. Like reliability, it is normally used as a measurement for critical (un-spared) equipment. Availability can be directly expressed as a function of time or as a function of MTBF and MTTR as shown in
Table 12.1.4.
TABLE 12.1.4 Availability
AVAILABILITY =
NO. OF OPERATING HOURS/YR
8760 − PLANNED DOWNTIMES (T & I's OR TUR NAROUNDS)
AVAILABILITY =
MTBF
MTBF and MTTR
549
B.P. 12.1
More Best Practices for Rotating Equipment
As an example, if the MTBF for the gas turbine in the previous example is
2836 h, what is this G.T.Ds availability?
Given,
MTBF = 2836 h
MTTR = 65.67 h
2836 h
2836 + 65.67 h
= 97.73%
AVAILABILITY =
Included on the next pages for your use are:
l
l
One availability factors
Table transparency
Two paper tables
Please have group members prepare charts for the last two years of operation
on your assigned units.
IDENTIFYING TARGETS FOR IMPROVEMENT
Once the site reliability audit data has been reduced, areas for reliability can be
identified. In the previous section, the area’s lowest availability were identified
progressively as shown in Table 12.1.5.
Normal Component Reliability Comparison
Once low availability machinery items or components are identified, they must
be compared to normal values to determine if a reliability improvement program
is warranted. A suggested course for comparison is Table 4.3 from Machinery
Reliability Assessment by Heinz P. Block and Fred K. Geitner, Copyright 1990
by Van Nortrand Reinhold. This table contains “Best” and “Worst” failure rates
for a variety of components as well as basic failure modes. At this point, list
the component failure rates for your groups’ lowest availability equipment item
(from availability factor worksheet) on the reliability comparison sheet. Then
TABLE 12.1.5 Identifying Targets for Reliability Improvement
TRAIN AVAILABILITY
↓
UNIT AVAILABILITY
↓
ITEM AVAILABILITY
↓
COMPONENT AVAILABILTY
550
Reliability Optimization
Chapter | 12
refer to the handout for normal reliability data and note the normal failure rate
for each affected component. Then compare the site actual to normal failure
rates and mark the action required column as appropriate.
Included on the next pages for your use are:
l
l
One worksheet transparency
Two paper worksheets
B.P. 12.2: Establish cost of unavailability for critical equipment.
By knowing the revenue lost daily for a particular machine train being down,
this will help implement an action plan to fix the problem.
The easiest way to get anybody’s attention is to speak in terms of cost, and
if it can be justified by saving money, management will definitely listen and get
the resolution implemented.
L.L. 12.2: Failure to establish cost of unavailability has resulted in continuing failures since resolution could not be implemented.
BENCHMARKS
This best practice has been in use since the early 1990s and has resulted in high
implementation rates (over 50%) of resolutions to the problem.
SUPPORTING MATERIAL
Cost of Unreliability
At this point, the “Bad Actors” or the “Hit List” has been identified and the
specific availability measurements quantified. What remains is perhaps the most
difficult task. See Fig. 12.2.1.
Regardless of how great your salesmanship is, you will not succeed unless
your plan is “Cost Effective” in management’s opinion.
FIGURE 12.2.1 The most difficult task.
551
B.P. 12.3
More Best Practices for Rotating Equipment
TABLE 12.2.1 The Cost of Unreliability Critical Rotating Equipment
(Per Year)
•
•
•
•
•
Lost product revenue × days forced outage
Maintenance costs
Replacement part cost
Labor cost
Unnecessary turnaround timea
a
Assumes process unit start-up is delayed by activity.
Therefore, a “Cost” must be assessed for each bad actor. We define this
cost as the, “COST OF UNRELIABILITY.” The cost factors are stated in
Table 12.2.1.
Simply add the costs of unreliability for each component that does not meet the
component reliability norms. Once these figures are obtained, the reliability assessment process is complete. We are now equipped with the data to proceed up the
reliability pyramid to prepare reliability improvement plans. At this point, please
refer to the “Cost of Unreliability” worksheets on the following pages and tabulate
the costs of unreliability for the components on the “Bad Actor” or Hit List.”
Included are:
l
l
One cost of unreliability transparency worksheet
Two paper worksheets
B.P. 12.3: Bring component condition monitoring (CCM) philosophy into
reliability centered maintenance (RCM).
It has been found by the writer that using component condition monitoring
principles can simplify RCM paperwork and allow for easier identification of
issues and problem resolution.
L.L. 12.3: Failure to simplify RCM has resulted in personnel being flooded with paperwork and machinery problems not being solved properly.
BENCHMARKS
The writer has and is recommending that this best practice be used globally to
simplify the sometimes daunting program of reliability centered maintenance.
SUPPORTING MATERIAL
Tables 12.3.1–12.3.10 are spreadsheets used to identify the proper parameters
to monitor for a Pump, Steam Turbine Driven Compressor Train, and a Gas
Turbine respectively.
552
TABLE 12.3.1 Pump Component Condition Monitoring
Performance
Item #
Date
Input
P1 (psig)
P2 (psig)
S.G.
Pump Speed (RPM)
Flow Rate (gpm)
Calculate
Head (ft.)
Flow Determinationa
Control Valve Position
CV for Valve and Trim Typeb
Fluid S.G.
Control Valve ∆P
Calculated Valve Flow
Motor Amps
Volts
Power Factor
Motor Eff’y
Calculated Power
Flow From Pump Curvec
T1 (Deg. F)
T2 (Deg. F)
CP (Specific Heat)
Calculated Head
Calculated Pump Eff’y
Flow From Pump Curvec
Pump Maintenance Requiredd
EROE Determination
BEP Flowe
EROE Min Flow
EROE Max Flow
Is Pump in EROE
EROE Targets for Operations
Flow
Amps
Pump ∆ T (Measured on Inlet
and Discharge Pipes)
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
a
If a flow meter is not available for the pump flow will be determined by one or more alternative
methods. These alternative methods are: (1) Portable Ultrasonic Flowmeter, (2) Control Valve
Position to calculate flow, (3) Motor Amps to calculate Power, and (4). Pump ∆T to calculate pump
efficiency.
b
The CV can be found on the Valve Manufacturers curve for specific valve type and trim. Note that
most of the major valve manufacturers post these curves on their websites.
c
Flow will be estimated using the original shop test curve. Note that if the pump is not in good
condition these estimates will not be very accurate.
d
Pump Maintenance should be considered if the head and flow (operating point) when plotted on
the test curve is approx. 10 % below test curve flow for calculated head.
e
Flow at Highest Efficiency taken from the shop test curve in the supplimentary manual.
553
B.P. 12.3
More Best Practices for Rotating Equipment
TABLE 12.3.2 Pump Mechanical Seal and Flush System Condition
Monitoring
Item #
Group Name
Seal Chamber Condition
Pressure (psig)
Temperature (deg F)
Fluid Sample ?
Sample Results
Is Pressure as Designed?
Is Temp. as designed?
Action Required
Seal Flush System Condition
Temp Difference across orifice (deg F)
Is orifice plugged?
Temp Difference across Cyclone
Is Cyclone Plugged?
Temp Difference across Cooler
Is Cooler Fouled?
External Flow (Plan 32) as Designed?
Plan 52/53 Res. Diff Temp. (deg F)
Rservoir Level
Reservoir Pressure (psig)
Vented to Flare?
Blocked In?
Barrier Fluid Sample?
Steam Condition if Steam Quench
installed?
Action Required
Overall Required Action
554
Reliability Optimization
Chapter | 12
TABLE 12.3.3 Pump Bearing System Condition Monitoring
Bearings
Item #
Group Name
Housing Temperature (°F)
Outboard Journal
Coupling End Journal
Thrust Bearing
Housing Vibration (in/sec)
Outboard Journal Horizontal
Outboard Journal Vertical
Coupling Journal Horizontal
Coupling Journal Vertical
Thrust Horizontal
Thrust Vertical
Thrust Axial
Oil Condition
Appearance
Water Content
Constant Level Oiler Condition
Bottom Bracket Oil Condition
Monitor? (Y/N)
Flow From Pump Curve
Action Required
555
B.P. 12.3
More Best Practices for Rotating Equipment
TABLE 12.3.4 Compressor Performance Monitoring
Item/Section #
Date/Time
Given
M.W.
P1 (PSIA)
T1 (F)
P2 (PSIA)
T2 (F)
K
Z
Inlet Flow (ACFM)
N (RPM)
Calculate
(K−1)/K
—
—
—
—
(n−1)/n
—
—
—
—
Gas Density (lbs/ft.3)
—
—
—
—
Mass Flow (lbs/min)
—
—
—
—
GHP
—
—
—
—
Poly Hd (ft.-lbf/lbm)
—
—
—
—
Poly Eff’y
—
—
—
—
Does Compressor Need
Maintenance?
556
Reliability Optimization
Chapter | 12
TABLE 12.3.5 Steam Turbine Performance
Type
Extraction/Condensing
Backpressure
Theoretical Steam Rate (lb/HPh, TSR)
—
—
HP for HP Section (or HP of BP Turbine)
—
Item #
Date/Time
Given
P1 (psi)
T1 (°F)
P2 (psi)
T2 (°F)
Speed (RPM)
Flow Rate (lb/h)
Total Power at Coupling (HP)a
Determine
b
h1 (BTU/h, From Mollier Diagram)
b
h2 Isentropic (BTU/lb, From Mollier
Diagram)
HP for LP Section
—
—
—
b
h2 Actual (BTU/lb, From Mollier
Diagram)
—
Actual Steam Rate (lb/HPh, ASR)
—
—
—
Steam Turbine Efficiency
—
—
—
%Moisture LP Section
Actions Required
Do not fill shaded cells.
a
Total Power at Coupling can be determined via torque meter or from Compressor Total Gas Power
plus mechanical losses for bearings and seals.
b
Note that values for h1, h2 isentropic and h2 actual (HP Case only) are obtained from a Mollier
Diagram contained in this section or by using steam tables.
557
B.P. 12.3
More Best Practices for Rotating Equipment
TABLE 12.3.6 Component Condition Monitoring Worksheet
Item #:
Date/Time:
Journ. Brgs.
Compressor DE Horiz. Vibes (mils)
Compressor DE Vert. Vibes (mils)
Compressor DE Pad Temp (°F)
Compressor DE Pad Temp (°F)
Compressor NDE Horiz. Vibes(mils)
Compressor NDE Vert. Vibes (mils)
Compressor NDE Pad Temp (°F)
Compressor NDE Pad Temp (°F)
Steam Turbine DE Horiz. Vibes (mils)
Steam Turbine DE Vert. Vibes (mils)
Steam Turbine DE Pad Temp. (°F)
Steam Turbine DE Pad Temp. (°F)
Steam Turbine NDE Horiz. Vibes (mils)
Steam Turbine NDE Vert. Vibes (mils)
Steam Turbine NDE Pad Temp. (°F)
Steam Turbine NDE Pad Temp. (°F)
Thrust Brgs.
Compressor displ.
Compressor displ.
Compressor Active Pad Temp. (°F)
Compressor Active Pad Temp. (°F)
Compressor Inactive Pad Temp. (°F)
Compressor Inactive Pad Temp. (°F)
Balance Line Diff. P (psid)
Steam Turbine displ.
Steam Turbine displ.
Steam Turbine Active Pad Temp. (°F)
Steam Turbine Active Pad Temp. (°F)
Steam Turbine Inactive Pad Temp. (°F)
Steam Turbine Inactive Pad Temp. (°F)
558
Reliability Optimization
Chapter | 12
TABLE 12.3.7 Component Condition Monitoring Worksheet
Item #:
Date/Time
Journ. Brgs.
Steam Turbine DE Horiz. Vibes (mils)
Steam Turbine DE Vert. Vibes (mils)
Steam Turbine DE Pad Temp. (°F)
Steam Turbine DE Pad Temp. (°F)
Steam Turbine NDE Horiz. Vibes (mils)
Steam Turbine NDE Vert. Vibes (mils)
Steam Turbine NDE Pad Temp. (°F)
Steam Turbine NDE Pad Temp. (°F)
Thrust Brgs.
Steam Turbine displ. (mils)
Steam Turbine displ. (mils)
Steam Turbine Active Pad Temp. (°F)
Steam Turbine Active Pad Temp. (°F)
Steam Turbine Inactive Pad Temp. (°F)
Steam Turbine Inactive Pad Temp. (°F)
DE Gland Condenser Pressure (in mmHg)
NDE Gland Condenser Pressure (in mmHg)
TABLE 12.3.8 L.O. and SO Syst.
Component/Item
Specified value
Actual value
Comments
Oil Reservoir
Level
Oil Temp. (°F)
Air in Oil? (Y/N)
Gas in Oil?
Oil Sample?
Other
Other
(Continued)
559
B.P. 12.3
More Best Practices for Rotating Equipment
TABLE 12.3.8 L.O. and SO Syst. (cont.)
Component/Item
Pumps
Aux. Pump Operating?
P2 (psig)
Suction Noise?
Suction Filter ∆P (psid)
Vibration (in./s)
Brg. Bracket Temp. (°F)
Other
Other
Couplings
Noise?
Strobe Findings
Other
Other
Turbine Driver
Operating Speed (RPM)
Trip Speed Setpoint (RPM)
Vibration (in./s)
Brg. Bracket Temp. (°F)
Gov. Hunting?
Trip Lever Condition
Gov. Oil Condition
Other
Other
Motor Driver
Operating?
Vibration (in./s)
Brg. Bracket Temp. (°F)
Axial Shaft Movement (in./s)
Fan Noise?
Other
Other
560
Specified value
Actual value
Comments
Reliability Optimization
Chapter | 12
TABLE 12.3.8 L.O. and SO Syst. (cont.)
Component/Item
Specified value
Actual value
Comments
Relief Valves
Passing?
Set Pressure (psig)
Pump P2 Press. (psig)
Other
Other
Check Valves
Aux. Pump Turning
Backwards?
Noise?
Other
Other
Back Pressure Valve
% Open
Stable?
Valve Noise?
Set Pressure (psig)
Maintained Pressure (psig)
Other
Other
Transfer Valves
One Bank Operating?
Noise?
Other
Other
Coolers
∆T Oil
CW Valve Pos.
Cooler Operating?
Vent Valves Open?
Other
Other
(Continued)
561
B.P. 12.3
More Best Practices for Rotating Equipment
TABLE 12.3.8 L.O. and SO Syst. (cont.)
Component/Item
TCV’s
% Open
Set Temp. (°F)
Stable?
Actual Temp. (°F)
Other
Other
Filters
∆P (psid)
Vent Valves Open?
Last Filter Change
Other
Other
Accumulators
Pre-charged Pressure (psig)
Last PM Date
Other
Other
Lube Oil PCV
% Open
Set Pressure (psig)
Actual Pressure (psig)
Stable?
Other
Other
Control Oil PCV
% Open
Set Pressure (psig)
Acual Pressure (psig)
Stable?
Other
Other
562
Specified value
Actual value
Comments
Reliability Optimization
Chapter | 12
TABLE 12.3.8 L.O. and SO Syst. (cont.)
Component/Item
Specified value
Actual value
Comments
Lube Oil Rundown Tank (or Emerg. Pump)
Pump or Tank?
Pump Operating?
Tank Overflow
Other
Other
Lube Oil Supply Lines
Leaks?
Noise?
Vibration (in./s)
Other
Other
Seal Oil Supply Valve
Position
Stable?
Seal Oil Supply Pressure
Seal Oil OH Tank
Level
S.O. Differential Pressure
Drainers
Level
Flow Through Vent Orifice?
563
B.P. 12.3
More Best Practices for Rotating Equipment
TABLE 12.3.9 DGS System Monitoring
Item
Observations
Comments
Primary Gas Filter DP (psid)
Primary Gas Supply DP
Prim. Vent Flow suct. (scfm)
Prim. Vent Flow Disch. (scfm)
Sec. Gas Filter DP (psid)
Sec. Gas Supply Press. (psig)
Sec. Supply Flow Suct. (scfm)
Sec. Supply Flow Disch. (scfm)
Sec. Vent Suct. Flow (scfm)
Sec. Vent Disch. Flow (scfm)
Oil in Sec. Drain Suct.?
Oil in Sec. Drain Disch.?
Seperation Gas Filter DP (psid)
Balance Line DP (psid)
TABLE 12.3.10 Gas Turbine Component Condition Monitoring
Compressor Performance
P atmosphere (psia)
T ambient (°F)
P Disch. (psia)
T Disch. (°F)
Speed (RPM)
K
(K−1)/K
-
-
-
-
(n−1)/n
-
-
-
-
Polytropic Efficiency
-
-
-
-
564
Chapter | 12
Reliability Optimization
TABLE 12.3.10 Gas Turbine Component Condition Monitoring (cont.)
What Action is Required based on %
difference from design efficiency?
Gas Turbine Performance
Fuel LHV (BTU/SCFH)
Flow (lb/h)
FG Pressure (psia)
FG Temperature (R)
FG Density (lb/ft.3)
-
-
-
-
SCFH FG Flow
-
-
-
-
Site Heat Rate (BTU/h)
-
-
-
-
GT Efficiency
-
-
-
-
Output HP
What Action is Required based on
Gas Turbine Performance info. above?
% PT Speed
Bearing Condition
# 1 Journal Brg. Vib X
# 1 Journal Brg. Vib Y
Major Frequency Observed
Radial Shaft Position
# 1 Brg. Pad Temp. 1 (°F)
# 1 Brg. Pad Temp. 2 (°F)
# 1 Brg. Drain Temp (°F)
GG Thrust Brg. Displacement
Direction of Thrust
Pad Temp. Inlet End 1 (°F)
Pad Temp. Inlet End 2 (°F)
Pad Temp. Exhaust End 1 (°F)
Pad Temp. Exhaust End 2 (°F)
Thrust Brg. Drain Temp. (°F)
# 2 Journal Brg. Vib X
# 2 Journal Brg. Vib Y
Major Frequency Observed
Radial Shaft Position
# 2 Brg. Pad Temp. 1 (°F)
# 2 Brg. Pad Temp. 2 (°F)
(Continued)
565
B.P. 12.3
More Best Practices for Rotating Equipment
TABLE 12.3.10 Gas Turbine Component Condition Monitoring (cont.)
# 2 Brg. Drain Temp (°F)
# 3 Journal Brg. Vib X
# 3 Journal Brg. Vib Y
Major Frequency Observed
Radial Shaft Position
# 3 Brg. Pad Temp. 1 (°F)
# 3 Brg. Pad Temp. 2 (°F)
# 3 Brg. Drain Temp (°F)
PT Thrust Brg. Displacement
Direction of Thrust
Pad Temp. Inlet End 1 (°F)
Pad Temp. Inlet End 2 (°F)
Pad Temp. Exhaust End 1 (°F)
Pad Temp. Exhaust End 2 (°F)
Thrust Brg. Drain Temp. (°F)
# 4 Journal Brg. Vib X
# 4 Journal Brg. Vib Y
Major Frequency Observed
Radial Shaft Position
# 4 Brg. Pad Temp. 1 (°F)
# 4 Brg. Pad Temp. 2 (°F)
# 4 Brg. Drain Temp (°F)
Brg. Inlet Oil Pressure (psia)
Brg. Inlet Oil Temperature (°F)
Viscosity (cst)
% Water in oil
Lube Oil Flashpoint (°F)
What Action is Required based on
Gas Turbine Bearing info. above?
Air Filtration Sys. Condition
Filter DP (psid)
Action of Self Cleaning Air System
(Active or Inactive)?
Action?
Variable Inlet Guide Vanes
Hyd. Supply Differential Press. (psid)
566
Chapter | 12
Reliability Optimization
TABLE 12.3.10 Gas Turbine Component Condition Monitoring (cont.)
Guide Vane Position
Air Flow Set Point
Measured Air Flow
IGV Exhaust Temp. Reference
Measured Exhaust Temp. (°F)
Action?
Cooling and Sealing Air Sys.
Cooling Air Pressure (psia)
Other Cooling Sys. Observations
Air Sealing System Supply Pressure
(psia)
Fuel Gas System
Ambient Temp. (°F)
-
-
-
-
-
-
-
-
Relative Humidity
Fuel Supply Pressure (psia)
Fuel Flow Rate (lb/h)
Fuel Pressure Upstream of Shut Off
Valve (psia)
Fuel Shut Off valve position
Fuel Pressure Upstream of Fuel
Control Valve (psia)
Fuel Control Valve position
Fuel Supply Pressure to combustors
(psia)
Action?
Combustion Monitoring System
Ambient Temp. (°F)
Relative Humidity
Exhaust Temp. 1 (°F)
Exhaust Temp. 2 (°F)
Exhaust Temp. 3 (°F)
Exhaust Temp. 4 (°F)
Exhaust Temp. 5 (°F)
Exhaust Temp. 6 (°F)
Exhaust Temp. 7 (°F)
Exhaust Temp. 8 (°F)
(Continued)
567
B.P. 12.4
More Best Practices for Rotating Equipment
TABLE 12.3.10 Gas Turbine Component Condition Monitoring (cont.)
Exhaust Temp. 9 (°F)
Exhaust Temp. 10 (°F)
Exhaust Temp. 11 (°F)
Exhaust Temp. 12 (°F)
Allowable Spread (°F)
Top Spread 1 (°F)
Top Spread 2 (°F)
Top Spread 3 (°F)
Exhaust Thermocouple Alarm on?
Combustion Alarm on?
High Exhaust Temp. Spread trip?
Monitor Enable Activated?
Action?
Exhaust Temperature Control System
Compressor Disch. P (psia)
Fuel Stroke Reference
Ambient Temperature (°F)
Relative Humidity
Calculated Firing Temp. (°F)
Action?
B.P. 12.4: Root Cause Analysis (RCA) guidelines.
As mentioned many times throughout this book, it is imperative to identify
issues with your machinery before a failure occurs. Therefore, we do not like to
practice Root Cause Failure Analysis (RCFA) but rather RCA, which is more
proactive.
Following are key guidelines to effectively conduct an RCA:
l
l
l
l
l
Clearly state the problem and estimate the loss of revenue this problem has
or can result in
Make sure all the facts that you gather are Objective and can be back by data
Define the function of the component that may fail
Be sure to define and list all of the systems that can affect the equipment and
what components they are comprised of
Action plan needs to be cost effective, timely, and actually solve the Root
Cause of the problem.
568
Reliability Optimization
Chapter | 12
L.L. 12.4: Ineffective root cause analyses have resulted in the inability to
identify a root cause of failure, which leads to repeat failures and revenue
lost.
BENCHMARKS
Forsthoffer Associates, Inc. utilizes this RCA procedure, which time and time
again identifies the Root Cause effectively and determines the proper action to
resolve the problem.
SUPPORTING MATERIAL
Troubleshooting is the action of discovering and eliminating causes of trouble.
Table 12.4.1 presents the definition.
Troubleshooting can be as simple as discovering why a light bulb does not
function to as complicated as debugging modern day computer software. In this
section we will deal with the important aspects of troubleshooting and practice
those aspects to develop troubleshooting skills. Do not expect to acquire these
skills in this workshop. Like any skills, they require practice. This module will
introduce you to all of the important aspects and practice exercises. Attendees
are encouraged to implement the principles used in this workshop immediately
and continue to use these principles in order to perfect your analysis skills.
Table 12.4.2 presents the basic requirements of troubleshooting.
We must first find abnormal conditions. To define abnormal conditions, the
normal condition must be known. Therefore the baseline conditions are required. The concept of normal condition and change of condition have been
previously discussed. We must be sure to define all abnormal conditions, consequently full condition monitoring must be practiced.
TABLE 12.4.1 Definition
TROUBLESHOOT
TO DISCOVER AND ELIMINATE (ROOT) CAUSES OF TROUBLE
TABLE 12.4.2 Establish a Baseline
• Troubleshooting Requires That All Abnormal Conditions Be Defined
• However, To Determine Abnormal Conditions, The Normal Conditions Must
Be Known
• Therefore Baseline (Normal) Conditions Must Be Known
569
B.P. 12.4
More Best Practices for Rotating Equipment
THE ROOT CAUSE ANALYSIS (TROUBLESHOOTING)
PROCEDURE OVERVIEW
The intent of this subsection is to provide an overview concerning the troubleshooting procedure. In subsequent subsections, specific details concerning each
major item will be covered. Fig. 12.4.1 presents the total concept of troubleshooting based on our experience.
It is important that each step be thoroughly completed before proceeding to
the next. Certainly there may be instances where it will be required to recycle
and go back to a preceding step to obtain more information or correct misinformation that exists. Remember this procedure is generic and may be slightly
refined for specific problems. However, we have found this procedure to be the
most consistent in effectively defining root causes and correcting those problems in a cost-effective manner.
I.
II.
Initial Fact Finding
Do not leave any stone unturned. Ask all affected groups (operations,
maintenance, engineering, contractor, vendor, etc.).
Thoroughly Understand the Affected Component and System Function
Do not be afraid to admit lack of knowledge. Obtain knowledge through
experienced sources, instruction book and other publications. Confirm
proper understanding of the facts before proceeding.
FIGURE 12.4.1 The troubleshooting process.
570
Reliability Optimization
Chapter | 12
III. Define Abnormal Conditions
Abnormal conditions may not mean that the condition is in alarm. It is
helpful to use percentage to define deviation from baseline. Any significant
deviation may define an abnormal condition.
IV. Listing All Possible Causes
Note all causes, even if they appear to be highly improbable, causes can be
eliminated when reviewing causes based on facts.
V. Eliminate Causes Based on Facts
Review each cause listed in light of all facts. Eliminate causes that are not
possible based on information.
VI. State the Root Cause(s)
In this area be sure to thoroughly investigate all systems and subsystems of
components. Identify any subcomponents that are operating in abnormal
conditions. Root causes are usually found in subsystems of components.
VII. Develop an Action Plan
Include a concise action plan that can be presented to management to obtain full management support and aid in the implementation of the action
plan.
Initial Fact Finding
As previously mentioned, this phase is the most important phase in the troubleshooting process.
1. State the Apparent Problem
Remember the apparent problem is usually an effect and not a cause.
Table 12.4.3 presents the major items involved in obtained the apparent
problem.
2. Define the Affected Components
Once the apparent problem is stated, all affected components must be defined. Table 12.4.4 presents the action required to define these components.
TABLE 12.4.3 State the Apparent Problem
• Ask the involved personnel
• Inspect affected parts
• Confirm the apparent problem
TABLE 12.4.4 Define All Affected Components
• Thoroughly inspect for all affected components
• List all affected components
571
B.P. 12.4
More Best Practices for Rotating Equipment
TABLE 12.4.5 Fact Finding Guide Lines
SUGGESTED QUESTION LIST
1.
2.
3.
4.
5.
6.
7.
8.
What Is The Problem?
What Components Failed?
What Are Facts Concerning Failed Components?
How Long Has Unit Been Operating Without This Component Failure
Has This Component Failed Before?
What Were Component System Parameters Prior To Failure
What Parameters Exceeded Normal Values
What Changed?
a. Process Conditions
b. Operating Procedure
c. New Components (Equipment And System)
d. Piping System
e. Foundation
9. Parts Out Of Tolerance
10. Has This Type Of Failure Occurred In Other Locations? (Network – User’s Groups)
3. Obtain All Important Facts
Once the problem and affected components are defined, all the important
facts surrounding these components must be known. Table 12.4.5 is a suggested question list that has proven to be effective in fact finding. Remember,
this is a generic list. Some questions may not be appropriate or additional
questions may be required in specific cases.
4. Baseline Conditions and Trends
During fact finding activity it is extremely important to define all baseline conditions concerning the parameters around each major component involved. Refer back to the information in this manual regarding condition-monitoring parameters and be sure that all these parameters are checked for baseline condition
and trends. This procedure may be very frustrating and may take a long time
based on the data that the affected plant has available. Remember, be patient.
Table 12.4.6 lists the requirements to obtain baseline conditions and trends.
5. Inspection
Each affected component must be thoroughly inspected. This inspection will either add facts or possibly eliminate a component from
TABLE 12.4.6 Baseline Conditions and Trends
• Establish conditions before failure (baseline)
• Utilize distributed control system, operator’s logs, and reliability data for baseline
and trend data
• Establish changes from baseline conditions prior to failure
• Express condition changes in percent
572
Reliability Optimization
Chapter | 12
TABLE 12.4.7 Failed Component Inspection
•
•
•
•
•
Thoroughly inspect all parts
List all facts
Utilize site experience
Obtain vendor and/or consultant option if required
Fully define inspection procedure and provide all facts if “outside” inspection source
is used.
consideration. Table 12.4.7 presents information concerning component
inspection.
It may be necessary in this activity to enlist the help of additional associates as a check to inspection thoroughness. In addition, it may be necessary
to call on outside sources (non-destructive testing companies, troubleshooting specialists, consultants, etc.). In this event, the effectiveness of their activity will significantly depend on how well their work effort is defined and
how complete the information given to them is. Keep this in mind. Take the
time required to thoroughly define what is required of outside sources and
provide of the information required.
At this point, refer to Classroom Exercise 1—fact findings. The instructors will ask each group for input concerning this exercise.
Thorough Knowledge of Equipment, Component,
and System Functions
This area requires careful consideration. It also requires a significant amount of
paperwork to first define the function of each component, define the system and
list all of the effective parameters. Table 12.4.8 presents these facts.
This activity will, most of the time, require confirmation of component function from instruction book sources, outside material, articles, etc. or discussion with equipment experts. Be sure that the function of each component is
properly defined in a simple manner before proceeding, but the definition must
be complete. Once the component function is defined, then list all systems in
which the affected component operates Next list each parameter for the affected
component that must be checked for condition. Again, consult the information
concerning component parameters in this manual, which will be helpful.
TABLE 12.4.8 Component and System Functions
• Define the function of each affected component
• Define the system in which each affected component operates
• List the normal parameters for each affected component and system component
573
B.P. 12.4
More Best Practices for Rotating Equipment
TABLE 12.4.9 Define Abnormal Conditions
• Note value of each abnormal condition based on facts
• Express abnormal conditions in percent above normal limits
• State when each abnormal condition appeared
Define Abnormal Conditions
Once all components conditions are listed, obtain information concerning baseline and trends. Table 12.4.9 presents these facts. Again, it is cautioned that an
abnormal condition may not necessarily be an alarmed condition. This could
occur for many reasons: improper instrument setting, improper instrumentation
functioning, high setting, etc. It is helpful to list in percentage the deviation
between normal (baseline) and abnormal conditions. It may be necessary to
consult other experienced sources regarding a certain deviation to determine if,
in fact, this abnormality is significant.
List All Possible Causes
Table 12.4.10 presents information concerning causes.
It has been our experience that in many instances people start with very
complicated causes and causes that are difficult to prove. Start with the most
obvious causes from the fact-finding phase and obtain more facts if necessary in
order to list a cause. Again, be thorough and take your time. Insufficient information is the cause of most troubleshooting activity failures.
Once the possible causes are noted, the sources for these causes must be
defined. Subdivide the causes into the possible failure cause classifications. The
five failure classifications are:
1.
2.
3.
4.
5.
Process condition change
Improper installation/maintenance/assembly
Improper operating procedures
Design deficiencies
Component wear-out
Use the facts obtained to determine if the causes qualify for the above
sources. As an example, a design problem usually shows up from the first day
TABLE 12.4.10 List All Possible Causes
•
•
•
•
574
Start with most obvious causes
Obtain additional information if necessary
List remaining causes
Sub divide into the possible failure cause classifications (the 5 Why’s)
Reliability Optimization
Chapter | 12
TABLE 12.4.11 Eliminate Causes
Eliminate Causes Not Completely
Supported by Facts
of operation. If the problem has recently occurred, design and manufacturing
can be eliminated (if a new part was not recently installed). In this case, the
operating procedures and process conditions should be checked.
Eliminate Causes Not Related to the Problem
Refer to Table 12.4.11. Review each cause noted and test this cause against all
of the facts. Again, in this section it may be necessary to recycle and obtain additional facts or confirm facts previously stated.
State Root Causes of the Problem
Table 12.4.12 presents the facts concerning root causes. It is important to mention in this section that the thorough definition of component systems and subsystems must be completed. If the affected component systems and subsystems
are not thoroughly defined, important information regarding root causes may be
neglected. Once the systems are defined, all of the components in those systems
must be investigated for abnormalities.
Develop an Action Plan
Table 12.4.13 presents guidelines in developing an action plan.
It is important to develop the plan in an outline fashion that can be effectively presented to management. The method in which the action plan is defined,
TABLE 12.4.12 State Root Cause(s)
• Consider cause sources
• Obtain additional input if necessary
• Be sure all component systems and sub systems are considered
TABLE 12.4.13 Develop an Action Plan
•
•
•
•
Clearly state root cause
Define revenue loss if problem is not corrected
Clearly state action plan and define responsibilities
Present plan to management
575
B.P. 12.5
More Best Practices for Rotating Equipment
written, and presented will have a significant effect on whether the action plan
is implemented. Be sure to emphasize the impact on profit to management in
presentation of the plan.
Before an action plan is presented, it may be necessary to hold meetings
with contractors, vendors and/or consultants to thoroughly define all action
required. The more complete an action plan, the better the chances for its
success.
This concludes this module on the Root Cause Analysis (Troubleshooting)
Procedure.
B.P. 12.5: Guidelines to gathering facts when conducting a RCA.
When trying to find out information about a failure or potential failure, as
mentioned in B.P. 12.4, you have to make sure that the facts are based on actual
data. If the data is unavailable, then an action item needs to be to implement a
way to gather the appropriate data.
Key items to remember when gathering facts to ensure everything is
covered:
l
l
Think about Component Condition Monitoring and be sure to check all
of the following 5 major components for each machine that is having an
issue
○ Rotor (Performance)
○ Journal Bearings (Vibration and Temperature)
○ Thrust Bearings (Displacement and Pad Temperature)
○ Seals
○ Auxiliary Systems
Remember the 5 Reasons for machinery failures and prioritize based on
likelihood of what causes the most failures first
○ Process and related systems
○ Improper Maintenance
○ Improper Operating Procedures
○ Design
○ Wearout
L.L. 12.5: Failure look at the 5 components and 5 causes of failure when
gathering facts has resulted in long drawn out RCA’s that have not determined the root cause of the problem.
BENCHMARKS
Forsthoffer Associates, Inc. utilizes this RCA procedure, which time and time
again identifies the Root Cause effectively and determines the proper action to
resolve the problem.
576
Reliability Optimization
Chapter | 12
SUPPORTING MATERIAL
Rotating Equipment Does not Fail Randomly
Regardless of the location, rotating equipment usually fails when we don’t want
it to … on the weekend! In the Middle East it fails late on Wednesday afternoon.
In other places, failure occurs late Friday afternoon! Are these events random
failures? Can we predict them?
There is always a root cause of failure and there are indications in the failed
component condition. However, general purpose equipment, because it is not
usually continuously monitored (directly in the control room), certainly can
appear to fail randomly.
Please refer to Table 12.5.1.
How can we minimize random failures and our “Bad Actor List?” By being
aware of the major reasons for failure and by observing the condition of the
machinery components.
Please refer to Table 12.5.2.
Will this involve more data collection, more work? Many times, workload
and meetings are reduced.
It all comes down to...
Awareness, knowing what to look for.
In the following sections of this module, the root causes of machinery failures will be discussed in detail. In the next module, the ways to prevent machinery failures will be discussed.
THE MAJOR CAUSES OF MACHINERY
FAILURE—FAILURE CLASSIFICATIONS
The causes of machinery failure can be grouped into the Failure Classifications noted in Table 12.5.3. Note that usually, failures are the result of more than one cause.
TABLE 12.5.1 Equipment Does Not Fail Randomly!
• There Are Root Causes
• The Condition Of The Failed Component Will Change.
TABLE 12.5.2 How to Stop... Firefighting (Random Failure)
• Know Why Failures Occur (five Why’s)
• Condition Monitor The Major Components
• Make It A “Team Effort”
577
B.P. 12.5
More Best Practices for Rotating Equipment
TABLE 12.5.3 Failure Classifications
•
•
•
•
•
Process Condition Changes
Assembly/Installation
Operating Procedures
Design Deficiencies
Component Wear-out
1. Process Condition Changes
This classification is the most overlooked in terms of troubleshooting. For this
discussion, the most common type of Driven Equipment—Pumps will be used.
There are two major classifications of pumps, positive displacement and kinetic, centrifugal types being the most common. A positive displacement pump
is shown in Fig. 12.5.1. A centrifugal pump is shown in Fig. 12.5.2.
In a typical refinery, greater than 95% of the installed pumps are the centrifugal type.
Positive displacement pumps increase the pressure of the liquid by operating
on a fixed volume in a fixed space. The most common types of positive displacement pumps are listed in Table 12.5.4.
The characteristics of positive displacement pumps are detailed in
Table 12.5.5.
FIGURE 12.5.1 Positive displacement plunger pump.
578
Reliability Optimization
Chapter | 12
FIGURE 12.5.2 Centrifugal pump.
TABLE 12.5.4 Types of Positive Displacement Pumps
I. Pulsating—Non-continuous Flow
• Plunger
• Diaphragm
• Piston
II. Rotary—Continuous Flow
• Screw
• Gear
TABLE 12.5.5 Positive Displacement Pump Characteristics
•
•
•
•
Constant flow
Variable pressure produced
Require a pressure limiting device (PSV)
Flow does not vary with specific gravity changes
579
B.P. 12.5
More Best Practices for Rotating Equipment
TABLE 12.5.6 Pump Performance
• Pumps produce the pressure required by the process
• The flow rate for the required pressure is dependent on the pump’s characteristics
It is most important to remember that all driven equipment (pumps, compressors, fans, etc.) react to the process system requirements. They do only what
the process requires. This fact is noted in Table 12.5.6 for pumps.
Based on the characteristics of positive displacement pumps noted in
Table 12.5.5, positive displacement pump flow rate is not significantly affected
by the process system. This fact is shown in Fig. 12.5.3.
Therefore, since the flow rate of a positive displacement pump is not affected by the system, it is easy to determine if a positive displacement pump has
worn internals. This fact is shown in Table 12.5.7.
Centrifugal (Dynamic) Pumps
Centrifugal pumps increase the pressure of the liquid by using rotating blades to
increase the velocity of a liquid and then reduce the velocity of the liquid in the
volute. Refer again to Fig. 12.5.2.
FIGURE 12.5.3 A positive displacement pump in a process system.
TABLE 12.5.7 Positive Displacement Pump Internal Wear
• Is identified by reduced flow rate
• Control valve closing
• Reduced amps
580
Reliability Optimization
Chapter | 12
TABLE 12.5.8 Centrifugal Pump Characteristics
•
•
•
•
Variable flow
Fixed pressure produced for a specific flow
Does not require a pressure limiting device
Flow varies with differential pressure (P2−Pl) and/or specific gravity
A good analogy to this procedure is a football (soccer) game. When the ball
(liquid molecule) is kicked, the leg (vane) increases its velocity. When the goal
tender (volute), hopefully, catches the ball, its velocity is significantly reduced
and the pressure in the ball (molecule) is increased. If an instant replay “freeze
shot” picture is taken of the ball at this instant, the volume of the ball is reduced
and the pressure is increased.
The characteristics of any centrifugal pump then are significantly different
from positive displacement pumps and are noted in Table 12.5.8.
Refer again to Table 12.5.6 and note that all pumps react to the process
requirements.
Based on the characteristics of centrifugal pumps noted in Table 12.5.8, the
flow rate of all types of centrifugal pumps is affected by the Process System.
This fact is shown in Fig. 12.5.4.
Therefore, the flow rate of any centrifugal pump is affected by the system.
Refer to Fig. 12.5.5 and it can be observed that all types of mechanical
failures can occur based on where the pump is operating based on the process
requirements.
Since greater than 95% of the pumps used in this refinery are centrifugal,
their operating flow will be affected by the process.
Important facts concerning this failure classification are noted in Table 12.5.9.
FIGURE 12.5.4 A centrifugal pump in a process system.
581
B.P. 12.5
More Best Practices for Rotating Equipment
FIGURE 12.5.5 Centrifugal pump component damage and causes as a function of operating
point.
TABLE 12.5.9 Centrifugal Pump Reliability
•
•
•
•
Is affected by process system changes (system resistance and S.G.)
It is not affected by the operators
Increased differential pressure (P2−PI) means reduced flow rate
Decreased differential pressure (P2−PI) means increased flow rate
At this point it should be easy to see how we can condition monitor the centrifugal pump operating point. Refer to Table 12.5.10.
The definitions and characteristics of positive displacement and dynamic
equipment are presented in Fig. 12.5.6.
Driver reliability (motors, steam turbine, and diesel engines) can also be affected by the process when centrifugal driven equipment (pumps, compressor,
and fans) are used.
TABLE 12.5.10 Centrifugal Pump Practical Condition Monitoring
• Monitor flow and check with reliability unit (RERU) for significant changes.
• Flow can also be monitored by:
• Control valve position
• Motor amps
• Steam turbine valve position
582
Reliability Optimization
Chapter | 12
FIGURE 12.5.6 Positive displacement—dynamic pump comparison.
Refer to Fig. 12.5.7 and observe typical centrifugal pump curve.
Since the flow rate will be determined by the process requirements, the power (BHP) required by the driver will also be affected. What would occur if an
8½ diameter impeller were used and the head (differential pressure) required by
the process was low? Answer: Since the pressure differential required is low, the
FIGURE 12.5.7 A typical centrifugal pump performance curve.
583
B.P. 12.5
More Best Practices for Rotating Equipment
TABLE 12.5.11 Effect of the Process on Drivers
• Motors can trip on overload
• Steam turbines can reduce speed
• Diesel engines can trip on high engine temperature
flow rate will increase and for the 8½ diameter impeller, the power required by
the driver (BHP) will increase.
Therefore, a motor can trip out on overload, a steam turbine’s speed can
reduce or a diesel engine can trip on high engine temperature. These facts are
shown in Table 12.5.11.
Auxiliary System Reliability is also affected by process changes. Auxiliary systems support the equipment and their components by providing …
clean, cool fluid to the components at the correct differential pressure, temperature, and flow rate.
Typical auxiliary systems are:
l
l
l
l
Lube Oil System
Seal Flush System
Seal Steam Quench by System
Cooling Water System
The reliability of machinery components (bearings, seals, etc.) is directly
related to the reliability of the auxiliary system. In many cases, the root cause of
the component failure is found in the supporting auxiliary system.
As an example, changes in auxiliary system supply temperature, resulting
from cooling water temperature or ambient air temperature changes, can be the
root cause of component failure. Table 12.5.12 presents these facts.
As a result, the condition of all the auxiliary systems supporting a piece of
equipment must be monitored. Please refer to Table 12.5.13.
TABLE 12.5.12 Component (Bearing and Seal) Reliability
• Is directly related to auxiliary system reliability.
• Auxiliary system reliability is affected by process condition changes.
• “Root causes” of component failure are often found in the auxiliary system.
TABLE 12.5.13 Always “Think System”
• Monitor auxiliary system condition
• Inspect auxiliary systems during component replacement
584
Reliability Optimization
Chapter | 12
2. Improper Assembly Maintenance/Installation.
In proper assembly, maintenance (lubrication) of components and/or improper
installation practices will shorten the life of components and cause eventual
failure because the anticipated design factors were not met.
Tolerances, maintenance requirements, and installation procedures are provided to assure maximum component and equipment life. As an example, refer
to Fig. 12.5.8, which shows anti-friction bearings commonly used in pumps.
The relationship that determines how long an anti-friction bearing will last
is shown in Table 12.5.14.
Note that the life of the bearing is directly dependent on the forces acting
on the bearing to the 3rd power. As an example, if the forces were twice the design value, the life of the bearing would be reduced eight times! The minimum
FIGURE 12.5.8 Anti-friction bearing.
TABLE 12.5.14 L-10 Life
“B” or “L”-10 Life is defined as the life in hours that 9 out of 10 randomly selected
bearings would exceed in a specific application
"B" OR "L"-10 LIFE = 16700  C 
N  F 
3
Where N, RPM; C, load in LBS that will result in a bearing element life of 1,000,000
revolutions; F, actual load in LBS.
585
B.P. 12.5
More Best Practices for Rotating Equipment
TABLE 12.5.15 Sources of Forces
•
•
•
•
•
•
•
•
Increased process pipe forces and moments
Foundation forces (“soft” foot, differential settlement)
Fouling of plugging of impeller
Misalignment
Unbalance
Rubs
Improper assembly clearances
Thermal expansion of components (loss of cooling medium, excessive operating
temperature)
• Radial forces (single volute—off design operation)
• Poor piping layouts (causing unequal flow distribution to the pump)
specified life for a bearing is three years or 36 months. In this example, the life
would be reduced to approximately five months!
What can cause excessive bearing forces in this case? Refer to Table 12.5.15.
When any bearing is designed, it has a maximum acceptable total force
which will allow it to operate trouble-free for a period equal to, or in excess of,
its specified life. If the total forces acting on the bearing exceed this value, there
will be a bearing failure.
Therefore, the bearing must be installed in accordance with the vendor’s
instructions as detailed in the instruction book. The use of general procedures
or “rules of thumb” (using typical values) should be avoided.
In addition to the component assembly procedure, the pump installation
procedure must be followed. The pump installation procedure assures that the
following items are checked as noted in Table 12.5.16.
Now refer back to Table 12.5.15 and observe what additional forces, not
anticipated by the vendor that can be added to the total bearing force!
Therefore, improper installation procedure values will cause bearing failure.
In summary, the proper steps to prevent assembly and/or installation errors
are presented in Table 12.5.17.
A final word of advice regarding procedures is given in Fig. 12.5.9.
The final step in any assembly and/or installation procedure is to confirm
that the procedure was performed properly. This is accomplished by condition
monitoring of the replaced components after the machine is operating at normal
conditions... “lined out.” This method is outlined in Table 12.5.18.
TABLE 12.5.16 Pump Installation Procedure Requirements
• Proper pump/driver alignment
• Minimum external pipe forces
• Minimum external foundation forces (soft foot)
586
Reliability Optimization
Chapter | 12
TABLE 12.5.17 Avoid Assembly/Installation Errors by...
•
•
•
•
Having the instruction book available
Completely following specified procedures
Using only specified parts
Using refinery procedures for:
• Alignment
• Pipe stress check
• Soft foot check
FIGURE 12.5.9
TABLE 12.5.18 Post Component Replacement Check Guidelines
• Obtain component condition data
• Compare data to:
• Data before component replacement
• Site guidelines
• Confirm all component condition data is satisfactory (sign-off)
A good question at this point would be… How long after “line out” should
this check be performed? The norm is during the first 4–8 after normal operation
“line out” is attained.
All rotating equipment, when manufactured, is only tested for approximately 4 h. In the case of a pump operating at 3600 RPM, this pump will have rotated 1,000,000 times during this period! If the assembly/installation procedure
requirements were not met, we will know from the guidelines at this time.
3. Improper Operating Procedures
Failure of machinery and/or components can occur because equipment will be
subjected to conditions that exceed the design values. Refer to Table 12.5.19.
Operating procedures can be the root cause of failure. Please refer to
Table 12.5.20.
587
B.P. 12.5
More Best Practices for Rotating Equipment
TABLE 12.5.19 Improper Operating Procedures
• Subject the equipment to conditions that exceed design value limits.
TABLE 12.5.20 Operating Procedures Can Cause Failure if...
• They are not complete
• They are not followed
• The actual operating conditions are different than specified
Most machinery and/or component damage and wear occur during start-up
or shutdown (transient) conditions. During this time, the equipment is subjected
to rapid temperature, pressure and speed changes.
Shown in Table 12.5.21 are some examples of operating procedure requirements, the reason for the requirement, the consequences, the checks and
corrective action.
How can failures associated with operating procedures and their implementation be avoided?
Table 12.5.22 presents some guidelines.
The importance of having operating procedures (RIM’s) that are accurate,
properly written and completely followed cannot be overemphasized. As previously stated, the transient (rapid change) conditions that equipment is exposed
to during start-up and shutdown can cause rapid component wear or failure.
These facts are presented in Table 12.5.23.
4. Design Problems
Possible design problems manifest themselves like all other causes of failure
… component condition values are exceeded. Before a design problem is confirmed, three previously discussed causes of failure should be checked as shown
in Table 12.5.24.
Design problems can fall into three categories as shown in Table 12.5.25.
Design problems usually show up shortly after the process unit is at normal
conditions—“line out.” However, there are cases when design problems manifest themselves after extended operation and even the warranty period.
The cause of design problems is that the machine and/or its components were
not designed for the specified field operating conditions. See Table 12.5.26.
Often, the vendor is accused of a design error when, in fact, the specified
conditions and/or operating procedures have changed.
After “line out” of the process, if problems exist (component condition
values exceeded), the equipment data sheet should be compared to the actual
588
TABLE 12.5.21 Operating Procedure Requirements
Affected
components
Check(s) to prevent
problem
Assure only liquid is present
to remove frictional heat from
close running parts, assure
continuous pumping and
prevent immediate seal failure
Casing,
Impeller(s), Shaft,
Wear rings, Seal(s)
and Couplings
That discharge pressure is
reached immediately and
does not fluctuate
Check process system to
determine cause.
If discharge pressure does not
build up, shut down pump and
investigate
2. Pump suction
valve wide open,
no suction line
restrictions
Same as item 1, prevent
Cavitation and assure
discharge pressure is reached
Same as item 1
Suction valve wide open,
discharge reached and
does not fluctuate, no
Cavitation noise
Confirm suction valve is wide open.
Shut down pump if problem
remains check suction strainer
(suction basket if sump pump).
Look for suction line obstructions
3. Discharge pump
valve pinched for
start-upa
Prevent pump from running
dry
Prevent high flow cavitation
Prevent driver overload and
high inrush motor current
(which will reduce insulation
life)
Same as item 1
Discharge valve is not
full open
Discharge pressure is
reached quickly, with a
steady rise in pressure
No cavitation noise
Confirm discharge valve is not full
open
Shutdown pump if problem remains
Re-start with discharge valve
partially closed
Fully open discharge after pump
has reached full speed
4. Steam Turbine
and inlet steam
line warmed
Prevent slugging the turbine
with condensate
Bearings, steam
seals and possibly
turbine blades
Check drains for presence of condensate and
check that steam is above
saturation temperature
Open drain lines until condensate
flow stops.
Start turbine slowly, using small
bypass valve if supplied and
confirm absence of condensate
Requirement
Reasons
1. Pump and Seal
vented and full of
liquid
Corrective action
Reliability Optimization
Chapter | 12
589
(Continued)
B.P. 12.5
590
TABLE 12.5.21 Operating Procedure Requirements (cont.)
Check(s) to prevent
problem
Corrective action
Reduces thermal
shock,assures proper liquid
viscosity and correct shaft
alignment
Pump casing,
seals, bearings
and coupling
Check temperature of
casing, seal area and
suction line
Confirm bypass valve around the
discharge check valve and suction
valve are open
6. Cold service
pump childown
Eliminates vapor in pump
case and seal
Wear Rings and
Seals
That discharge pressure is
reached immediately,
Vent casing high point
and seal chamber.
Shut down pump if discharge pressure is not reached. Open vent
lines of casing and seal chamber
and confirm fluid in unit is 100%
liquid.
7. Steam turbine
slow roll & startup sequence
Reduces thermal shock &
assures correct alignment
Casing, rotor,
internal seals,
shaft end seals,
coupling and
bearings
Check drains for
presence of condensate
and check that steam is
above saturation temperature
Prior to slow roll, drain casing,
throttle valve, and inlet line.
Do not commission steam seal
system (if supplied) until turbine
is turning. Completely follow
vendor’s start-up instructions.
Reasons
5. Standby hot
service pump
warmed
a
For process unit start-up when no other pumps are in operation.
More Best Practices for Rotating Equipment
Affected
components
Requirement
Reliability Optimization
Chapter | 12
TABLE 12.5.22 Operating Procedure Reliability Guidelines
• Confirm the operating conditions are as specified
• Confirm the refinery instruction manual “RIM” is in agreement with vendor’s instructions
• Understand the reason for the requirement
• Do not hesitate to ask … (TSU, RERU)
TABLE 12.5.23 Follow Procedures Completely Because...
• During start-up and shutdown, equipment components are subject to rapid
(transient)
• Temperature changes
• Pressure changes
• Speed changes
• Remember, most component wear occurs during transient conditions.
TABLE 12.5.24 Possible Design Problem?
• First confirm the following classifications are not the root cause:
• Process condition change
• Assembly/installation procedures
• Operating procedures
TABLE 12.5.25 Design Problem Categories
• Engineering errors
• Material problems
• Manufacturing problems
TABLE 12.5.26 A Design Problem Exists if:
• The machine and/or components are not designed for specified field operating
conditions.
data to confirm the equipment was designed to the actual field operating conditions. If not, field conditions should be corrected if possible. If field conditions
remain different than specified on the data sheet, it is not a design problem, it is an “application problem.” In this situation, the vendor is justified in
asking for redesign costs if necessary.
591
B.P. 12.5
More Best Practices for Rotating Equipment
FIGURE 12.5.10 Single component mechanical seal.
An example of a possible design problem is a leaking mechanical seal. Refer
to Fig. 12.5.10.
All mechanical pump seals are designed to convert the liquid to a vapor
across the seal face. If the actual operating liquid conditions (vapor pressure,
temperature, and pump pressures) are not as specified, a mechanical seal failure
can occur because either the liquid is not changed to a vapor or the liquid vaporizes in the seal chamber (stuffing box).
Fig. 12.5.11 shows the change of pressure and temperature across a seal face
in three cases:
l
l
l
Early vaporization
Proper design
No-vaporization
FIGURE 12.5.11 Seal face vaporization.
592
Reliability Optimization
Chapter | 12
In this example, a pump “bad actor”, with more than one seal failure per
year, should first be checked for proper liquid conditions at the seal face before
it is classified as a mechanical seal design problem. The cause of failure may be
a process related issue (improper liquid conditions or plugged flush line orifice).
If all conditions are as specified, then it is truly a mechanical design problem.
Another example of a possible design problem is oil contamination in the
bearing brackets of a single stage steam turbine shown in Fig. 12.5.12.
Assume continuous problems are experienced with water in the oil causing
bearing failures. It is also confirmed that the source of water is from the carbon
ring seal leakage of steam into the bearing bracket.
As was previously stated, first check the other failure causes:
l
l
l
Process condition changes
Assembly/Installation procedures
Operating procedures
A check confirms that the failure causes noted above did not occur.
Therefore, the carbon ring seal system (carbon ring seals and bearing bracket isolator) is not designed to prevent oil contamination in the bearing bracket.
This case is an example of a true design problem. For this example, there are
two possible modifications:
l
l
Install an eductor system to positively prevent steam leakage from the seal
assembly (presently a requirement of Saudi Aramco general purpose steam
turbine spec).
Install a bearing isolator to positively prevent steam condensate from entering the bearing bracket (“Impro” type of equal).
In summary, the factors concerning possible design problems are noted in
Table 12.5.27.
5. Component Wear Out
Like design problems, component wear out is often determined to be a root
cause of failure.
However, as shown in Table 12.5.28, apparent component wear out usually
is caused by other failure classifications.
Always investigate the other failure classifications first. In many cases, component wear out is the effect, not the cause of the problem. Table 12.5.29 presents this information.
Refer back to failure classification 1, process condition changes
Figure 12.5.5, what components could “wear out” if the process required high
differential pressure?
Also refer to failure classification 2, assembly, installation problems, Fig. 12.5.8,
Tables 12.5.14 and 12.5.15, why could a bearing “wear out” if the foundation
cracked? Why would the bearing “wear out” if a shim fell out of pipe support?
593
B.P. 12.5
594
More Best Practices for Rotating Equipment
FIGURE 12.5.12 Single stage steam turbine.
Reliability Optimization
Chapter | 12
TABLE 12.5.27 Design Problem Determination and Action Plan
• Confirm all other failure causes do not exist
• Confirm specified operating conditions exist in the field (check data sheet)
• Conduct a design audit meeting with the vendor (if necessary)
TABLE 12.5.28 Apparent Component Wear Out
•
•
•
•
•
Usually is caused by one or more of the following failure classifications:
Process condition changes affecting the equipment and/or its auxiliary systems
Assembly/maintenance/installation errors
Improper operating procedures
Design deficiencies
TABLE 12.5.29 Wear-out
Component Wear out Is Often
The Effect, Not The Root Cause
Of The Problem
Finally, refer to failure classification 3, improper operating procedures,
Table 12.5.21. Why could a seal with a flush from its discharge line wear out if
a loading pump were started with the discharge valve wide open?
These examples are presented in Table 12.5.30.
Therefore, when will equipment components wear out? If all the considerations to eliminate failure classifications are met, components will last a long
time.
Think of some site pumps and name the shortest and longest period between:
l
l
Bearing replacement
Seal replacement
TABLE 12.5.30 Apparent Component Wear Out Examples
COMPONENT
ROOT CAUSE
Bearing, seal, impeller wear ring
Process condition change
Bearing
Installation problems
Mechanical seal
Operating procedure pump run dry
595
B.P. 12.5
More Best Practices for Rotating Equipment
TABLE 12.5.31 Component Life (Failure Classifications 1–4 not Present)
Component
Life (years)
Anti-friction bearing
8–10
Sleeve bearings
15–20
Mechanical seal
7–10
Wear rings
10–12
Impellers
15–20
TABLE 12.5.32 Component Wear Out Guidelines
• Monitor component condition parameters
• Plan scheduled shutdown using Predictive Maintenance (PDM) Principles
• Confirm failure classifications 1–4 are not present.
Table 12.5.31 presents some components and their life if failure classifications 1–4 are not present.
In many cases, component wear out is a result of the wear out of the “secondary” parts in the component. An example is “o” rings in mechanical seals.
As in the case of the previous failure classifications, component wear out
does not randomly occur. The condition parameters associated with these components will change. Table 12.5.32 presents the guidelines to determine component wear out.
596
Index
A
Accumulator, 398
failure modes, 402
pre-charge pressure check, 398
reliability factors for, 402
sizing, 373, 401
Accuracy of data, 161
Actual steam rate (A.S.R.), 228, 231
Additional tank sizing, 373
Aero derivative types gas turbine, 291, 294
advantages of, 296
applications, 300
classification of, 297
disadvantages of, 296
vs. industrial facts, 297
Anchor bolt installations, 487
Anti-whirl bearings, 191
API flush plans
plan 11, 419–420, 437, 440
plan 13, 421–422
plan 21, 422–423
plan 23, 426–429
plan 32, 424–425
plan 52, 412, 415–417
plan 53, 411–412, 416–418
plan 54, 431–432
Applications as a function of flow
(ACFM), 74
A.S.R.. See Actual steam rate (A.S.R.)
Audits, 9
Automatic start, 396
Auto-transfer of lube pump, 348
Auxiliary oil pump (AOP), 341
Auxiliary pump, 335
auto start, 346
Auxiliary systems, 6, 37, 42, 155
component, sizing audit form, 368
controls
function of, 384
and instrumentation, 521
reliability considerations, 399
types of, 383
dynamic response, 343
function, 342
monitoring, 397
operational cases, 344
reservoir, schematic representation of, 352
Axial position, 202
Axial thrust monitor, 252
B
Babbitt, 190
Back pad cooling, 190
Back pressure, 392
control valve, 335
definition/selection, 241
turbines, 233, 240
advantages, 233
valve. See Bypass valve
Balance line, 251, 380
DP transmitter, 379
Balance ratio, 405
Band aid, 250
Barrier fluid, as lubrication, 413, 417
Barrier seal, 448, 452
Baseline condition, 153, 164, 460, 461, 510
Baseline establishment, 163
Baseline performance conditions, 164
Bearing application, guidelines, 60
Bearing brackets, 191, 197
monitoring action plan, 222
oil contamination, 220, 539, 540
root cause, 222
Bearing design, 129, 187
Bearing failure, 220
Bearing load point temperature, 190
Bearing support stiffness, 129
Bearing wear, 345
elimination, 225
and oil viscosity reduction, 545
BEP. See Best efficiency point
Best efficiency point (BEP), 42
Blade corrosion, 233
Blading types, 142
backward lean, 142
radial blading, 143
radial vanes, 142
Bode plot, 263
Brayton cycle, 288
597
Index
Breather vent, 95
Buffer fluid, 416
Bypass control, 91, 388, 526
rapid transient response in open
direction, 529
valve, 529
Bypass valve, 356, 371, 529
C
Calibration accessibility, 339, 361
Capacity control methods, 90, 91
Capital investment, 3
Carbon ring seal, 540
ineffectiveness, process to correct, 222
CCM. See Component condition
monitoring (CCM)
CDP. See Compressor discharge pressure
CDT. See Compressor discharge temperature
Cementous grout, 485
Centrifugal compressors, 73, 74, 107, 504
Centrifugal impeller, 244
Centrifugal multi-stage
barrel, 118
horizontal split, 118
with side loads, 118
Centrifugal pumps
characteristics, 45
control options, 46
and drivers, 43, 44
component damage, 47
performance curve, 48
practical condition monitoring, 47
process effects, 48
head vs. flow curve, 49
positive displacement plunger pump, 43
in process system, 45
reliability, 45
Closed loop circulation of fluid, 415
Coalescing filter, 439
Coefficient of friction, 408
Company profit optimization, 2
Component and system functions, 152,
460, 510
Component arrangement, 338, 360
Component condition monitoring (CCM),
152, 153, 460, 461, 501, 503,
509, 552
parameters for anti-friction bearing, 504
worksheet, 469, 470, 558, 559
Component disassembly, 339, 361
Component function, 503
awareness, 502
598
Component material, 373
preferences, 335
Component reliability, 39
Component sizing audit, 351
Compression head, 109
Compressor discharge pressure (CDP), 250, 302
Compressor discharge temperature (CDT), 301
Compressors, 23, 445
application range chart, 75
discharge pressure, 250
efficiency, 115
head, effect on, 179
impellers, 113, 120
performance monitoring, 467, 556
condition, 158, 466, 515
protection systems, 265
rotor critical speed map, no damping, 125
section-definition/performance curve, 148
stage, 136
and characteristic curve, 137
defined, 136
suction pressure, 250
torque curve vs. speed curve, 188, 189
train, 187
Computer-aided design (CAD), 145
Condensing turbines
operation at exhaust pressures, 243
steam turbines, definition/selection, 241
Condition monitoring, 93
components monitored, 93
parameters/alarm limits, 155, 156, 463,
465, 512, 516
compressor liquid seal, 158, 466, 516
oil systems, 159, 467, 516
journal bearing
anti-friction, 463, 512
hydrodynamic, 156, 157, 463, 513
lube oil systems, 157, 514
pump liquid mechanical seal, 157,
465, 514
thrust bearing
anti-friction, 156, 464, 513
hydrodynamic, 464, 514
Connection orientation drawing, 374
Console, 373
Console baseplate construction, 338, 360
Console layout, 338, 360
Construction, 351
confirmation, 373
contractor, 22
special tools, 481
Continuously lubricated gear type coupling
with spacer, 203
Index
Control oil systems, 337
Control system analogy, 267
Control system sudden demand, 345
Control valve
excessive noise, 400
flow characteristics, 525
instability, 399
liquid sizing coefficient (Cv), 40–41, 386, 524
flow characteristics, 387
selection, 356
flow characteristics, 357
sensing lines, 402
snubber devices (dampers), 358
sizing, 371
Cooler sizing, 359
Cooling system, 105
Cooling water (CW), 422
Corporate databases, 1
Cost of unavailability
reliability improvement, 551
Cost of unreliability
critical rotating equipment, 552
supporting material, 551
Coupling attachment
methods, 212
Coupling drive, 214
Coupling hub, 213
assemblies, 202
Coupling hydraulic shrink fit, 213
Coupling installation, 212–216
Coupling lubricant, 204
Coupling manufacturers, 204
Coupling mounting procedure, 215
Coupling removal, 212–216
Coupling selection, 369
Coupling system, 211–212
Crankcase condition monitoring, 95
and safety devices, 95
Crankcase low oil level switch (optional), 95
Crankcase oil level gauge, 95
Crankcase oil temperature gauge, 95
Crankcase vibration detector (optional), 95
Crankshaft bearings, 95
Critical equipment trip instrumentation, 397
Critical equipment vendor data, 333, 334
Critical frequency, 254
Critical speeds, 124
and rotor response, 123
Crosshead assembly lubrication, 96
Cruise control, 267
Curtis stage, 237
Cv. See Control valve liquid sizing coefficient
CW. See Cooling water (CW)
Cyclone separator, 423, 424
Cylinder and liner, 99
Cylinder and packing lubricators, 104
Cylinder distance piece, 98
Cylinder packing, 98
and intercooler cooling water system, 106
D
Daily revenue losses, 13
Data (factors) require, 161
Data sheet review, 350
Design and manufacturing audits, 30–33
manufacturing audit guidelines, 31
suggested design audit activity, 32
summary and follow-up action, 32
vendor audit requirements, 31
Design audit agenda, 350
Determining section performance, 148
DGS. See Dry gas seals (DGS)
DGS system monitoring, 475
Diaphragm couplings, 207
Differential pressure, 46
gauge, 379
Direct acting actuator, 393
Direct acting (Hydraulically Actuated) valves,
375
Direct cooling, 190
Disaster bushing, 446
Discharge pressure, 53
Double acting self-equalizing thrust bearing
assembly, 167
Double dry gas seal systems, 454
Double flow steam turbine, 242
Double suction pumps, 69
Driven equipment, 43
Driver selection, 369
Driver sizing, 354
mechanical requirements, 355
Dry couplings, 206
Dry gas seals (DGS), 433–458, 489. See also
Double dry gas seal systems
design, 441–455
double seal system, for dry gas/saturated gas
application, 452
gas seal system types, 445
high pressure applications, 446
leakage rates, 444, 445
low/medium pressure applications, dry air or
inert gas, 445–446
options for toxic and/or flammable gas
applications, 447
principles of operation, 441–443
599
Index
Dry gas seals (DGS) (cont.)
ranges of operation, 443
tandem seals, 434, 436
for dry gas applications, 447–449
with interstage labyrinth, 449
for saturated gas applications, 449
Dry packing, 99
Dry screw, 92
compressor facts, 73, 82
twin screw compressor, 77
Dry type couplings
guard dimensional design criteria, 217
Dual shaft, 298
Dynamic compressors, 74, 107
number of stages per section, limiting
factors, 148
E
Eddy current, 197
Elastomer insert flexible drive members
couplings with, 210–211
Electrohydraulic extraction control and
protection system, 274
Electrohydraulic governor, 271, 274
block diagram, 272
functional test procedure outline, 492
Empty cooler-filter bank, 347
Enclosed coupling guards, 216
Enclosed impellers, 113, 121, 122
Epoxy grouts, 96, 484
Equipment, 487
Equipment reliability, 36
Equipment reliability operating envelope
(EROE), 49, 51, 53
determination, 49–51
Equivalent orifice, 343, 344, 384
concept, 522
reduction, 522
Equivalent vessel, 345, 385, 523–524
EROE. See Equipment reliability operating
envelope
Excessive governor mechanical linkage system
and valve friction
causes of, 223, 543
Excessive valve stem friction, 400
Excitation forces, 197
Expansion turbine
functions, 235
governors, facts concerning, 267
operation principle, 237
Expansion vapors
types, 237
600
External flush plans, 429
API flush plan 32, 54, 424–425,
431–432
Extraction/condensing steam turbine
high-pressure section, after 1st stage
pressure curve, 248
F
Factors limiting compressor impeller stage
head, 147
Factory Acceptance Test (FAT), 176
Factory testing, 361
Fan Laws, 117
FEED. See Front End Engineering Design
Field operating speed, 23
Field performance checks, 87
Filter sizing, 359, 372
Fin type air coolers, 426
Flange, 373
Flexible coupling
disc, 206–209
features, 206
function, 201–205
types, 202
Flexible membrane, 206–209
Floating carbon ring shaft seal system, 220
Flooded screw facts, 84
Flooded screw reliability, 85
Flow measuring device, 42
Flow meter, 39, 42, 388
in every system, 526
Flow rate, 55
Fluid head, 178
Fluid vaporization, 54
Flushing, 362
arrangement of components, 366
procedure, 362
Flush plan, types of, 403
Flush system condition monitoring, 554
Fly ball governors, 269
Forsthoffer Associates, Inc., 7
Fouling
definition, 245
effect on
absolute velocity, 246
impeller stage curve, 247
operating point, 245–246
relative velocity, 246
tangential velocity of gas, 247
mechanism, 244–245
Foundations, 486
bolt arrangements, 486
Index
Frame and running gear, 95, 96
Front End Engineering Design (FEED), 21
Functional testing, 490–491
lube/seal system test procedure outline, 491
G
Gas analysis guidelines, 162
Gas characteristics, 109, 110
Gas composition
change on HEAD, effect, 180
turbo-compressor flow rate, effect, 182
Gas compressors
typical operating range of various
types of, 75
Gas flow delivered, 108, 109
Gas flow produced, 108, 109
Gas generator, 287, 288
Gas Horsepower, 116
Gas seal system, 439
applications
low/medium pressure applications, dry air
or inert gas, 445–446
options for toxic and/or flammable gas
applications, 447
principles of operation, 441–443
ranges of operation, 443
tandem seals
for dry gas applications, 447–449, 451
with interstage labyrinth, 449, 452
for saturated gas applications, 449, 450
considerations for design, 440–441
lift-off speed, 440
oil ingestion from lube system, 441
o-ring, design and maintenance, 441
sensitivity
to dirt, 440
to saturated gas, 440
toxic gas leaks to atmosphere,
prevention of, 440
failure, 446, 449, 456
Gas turbines, 287
applications, 300
automotive engine and, 289, 290
bearing load, 323
building of, 290
classifications, 288, 291
by design type, 291
component condition monitoring, 313
compressor efficiency, 302
compressor fouling, 317
configuration, 291
critical equipment vendor data, 319, 320
cycles, 298, 299
combined, 299
regenerative, 299
simple, 299
design audit activity, 301
drive configurations, 297, 299
effect of, firing temperature, on power and
efficiency, 305
effect of increased firing temperature,
on produced power and engine
efficiency, 313
equivalent orifice system, 324
expansion turbines, 290
gas turbine vs. steam turbine cycles, 289
history, of development, 291, 292
industry, 213
inlet air density effect, on produced power
and heat rate, 307
ISO conditions, 305
and site performance, 305
lube systems for, 318
control oil system requirement, 328
minimum system operating capacity, 326
number of shafts, 297, 298
advantages/disadvantages, 298
output power and heat rate vs. ambient
temperature, 307
performance, 302–304
effect of inlet conditions, 307
vs. steam performance, 303, 304
single, 287
site conditions data, 320
site rating correction factors, 305
site rating exercise, 308
figure for, 310, 311
supporting material for, 318
system flow rate, 325–326
system heat load, 324
system parameters determination, 321
system requirements, 319
system resistance, 326–329
turbo compressors, 290
vs. steam turbine, 288
washing procedures, 317
Gauges, 373
Gearbox reliability, 187
Gear couplings, 202–205
dry couplings, replacement with, 201
Gear foundation, 188
Gear mesh, 188
Gear no-load pressure, 187
Gear radial bearing force transmission
path, 190
601
Index
Gear reaction (bearing) forces, 188
at bearings, 189–201
Gear rotor torque transmission path, 188
Gear shaft, 190
Gear tooth coupling, 203
Gear type flexible couplings
disadvantage, 204
General site considerations, 480
Gland seals and drains
condensing turbine, 281
noncondensing automatic-extraction
turbine, 280
Grafoil packing rings, 406
Grease pack couplings, 203
Grouting, 483, 484
H
Hand push coupling, 215
Hand valve(s)
closed on critical services, 223–225, 542
HDS off-gas-lube oil system, 106
HDS off-gas-piston rod and piston, 103
HDS off-gas running gear, 97
HDS suction and discharge valves, 102
Heat recovery steam generator (HRSG), 298
Helical pinion tooth reactions
at pitch diameter, 191
High pressure single stage steam turbine
bearing problems, 225
and oil viscosity reduction, 544
oil viscosity reduction, 225
Holset coupling
non-spacer type, 211
HRSG. See Heat recovery steam generator (HRSG)
Hybrid turbine, 236
design, 236
Hybrid type industrial gas turbine, 296
Hydraulic control systems, 274
Hydraulic fit coupling, 214
Hydraulic oil, 215
Hydrodynamic bearings
load effectiveness, 190
types, 190–194
Hydrodynamic journal bearing, 195
condition determining parameters, 195
trending data, 196
Hydrodynamic thrust bearing, 166
I
Ideal gas head equations, 112
Impeller design pre-bid meeting guidelines, 176
Impeller discharge velocities, 139
602
Impeller fouling, 247
Impeller geometry vs. specific speed, 115, 122
Impeller performance, 244
Impeller thrust forces, 171, 379, 380
Impeller types, 110
and specific speed, 120
Impeller with side plate removed, 138–141,
143, 144
Impulse blading, 235
advantages, 236
disadvantages, 236
Industrial type gas turbine, 291, 292
advantages and disadvantages, 293
classification of, 297
vs. aero derivative, 297
Inert gas, 439, 445–446
Inert seal gas, 455
Inlet pressure, 53
Inspection, 361
Installation manuals, 482
Installation of properly designed and
constructed foundations, 485
Instrumentation, 360, 396
in auxiliary system, 534
auxiliary system monitoring, 535
critical equipment trip
instrumentation, 535
stand-by pump automatic start, 534
types of, 383
Instrument calibration facts, 162
Instrument panel, 517
International Standards Organization
(ISO), 303
Invitation to bid (ITB), 176
Isentropic (adiabatic) reversible expansion, 236
J
Jaw and spider coupling, 210
Journal bearing
condition monitoring, 195
selection curve, 194
vibration instabilities, 197–201
L
Lessons learned, database of, 1
Leveling of equipment, 487
basics of, 487
Life cycle cost, 4
Liquid/gas
factors define, 178
Liquid seal systems, 441
Index
Liquid valve sizing coefficient, 387
LM6000 gas turbine, 295
Load torque, 214
Load vector, 190
Loss of revenue, 17
Lube oil systems, 377, 446
check list, 495
and seal oil system, 470
and seal system test procedure, 495
supply system, 342, 356, 390, 528
Lubricated screw compressors, 73
Lubricated to dry couplings
field retrofits, role of, 216–217
M
Machine reliability, 3
Machinery assets, 6
Machinery failure, 43
anti-friction bearing, 585
assembly maintenance/installation, 587
avoid errors, 587
improper, 584
auxiliary system reliability, 584
centrifugal pump, 579, 580
characteristics, 581
component damage and causes, 582
performance curve, 583
practical condition monitoring, 582
process system, 581
reliability, 582
classifications, 501, 502
component life, 596
component reliability, 584
component wear out, 593
apparent, 595
guidelines, 596
design problems, 588, 591, 595
L-10 Life, 585
mechanical pump seals, 592
operating procedures
improper, 587, 588
reliability guidelines, 591
requirements, 589
positive displacement pumps, 579, 580
characteristics, 579
dynamic pump comparison, 583
internal wear, 580
post component replacement check
guidelines, 587
prevention, 502
process condition changes, 578
pump installation procedure requirements, 586
pump performance, 580
seal face vaporization, 592
sources of forces, 586
think system, 584
Machinery instrumentation key facts, 518
Machinery vibration, 205
Maintenance accessibility, 339, 361
Maintenance costs, 95
Major machinery components, 152, 460
and systems, 460
Mars gas turbine, 294
Maximum profits, 3
Mean time between failure (MTBF), 547–548
availability, 549–550
component for high speed
packing, 95
piston rings, 95
valves, 95
failure rate, 548
for LNG circulating pump, 548
Mean time to repair (MTTR), 548, 549
availability, 549–550
failure rate, 548
MS 5038 General Electric Gas Turbines, 549
Measured rotor response, 133, 263
Bode plots, 133
Mechanical components, 87
auxiliary systems, 90
journal bearings, 89
rotor, 87
sealing devices, 89
thrust bearings, 89
timing gears, 89
Mechanical governor system, 268
Mechanical/hydraulic extraction control, 273
Mechanical hydraulic governor, 269
Mechanical-hydraulic single extraction
governor system, 272
Mechanical seal, 442
life, 49
replacement, 417
Mega plants, 287
Minimum condition monitoring, 94
Modulating relief valve, 537
Mollier diagram, 228
Motor Control Center (MCC), 341
Multiple, convoluted diaphragm-spacer
coupling, 209
Multistage centrifugal pump, 66
Multistage turbines
assembly drawing, 238
efficiency of, 232
multi-valve, protection system, 276
603
Index
Multi-valve steam turbine, 239
mechanical-hydraulic governor system, 270
protection system, 276
N
Net positive suction head (NPSH), 56, 58
Nitrogen amplifier, 454
Nitrogen bottles
used to pressurize plan 53A flush
systems, 410
Non-contact displacement measuring
system, 198
Non-destructive testing companies, 573
Non shrink, 485
NPSH. See Net positive suction head
NSS. See Suction specific speed
O
Oil film seal damping effects, 129
Oil injected twin screw compressor, 77
Oil level glass, 374
Oil separation system, 85
Oil system accumulator, 399
Oil system backpressure control valve, 187
Oil system control valves, 382
Oil system trip circuits, 376
Oil system vendor, 334
Oil viscosity
reduction elimination, 225
selection, guidelines for, 333
Online testing, 339, 361
Open impellers, 110, 121
Operating conditions, 55
Operating cost, 3
Operation principles, 79
Optimal seal plans for applications, 409–410
Optimum design velocity
for inlet of impeller and discharge of
impeller, 137
Orifices, 419, 425, 433, 446
monitoring, 421
Overall vibration, 199
P
Parameters, 161
Performance instrumentation location
guidelines, 163
Performance monitoring, 545
responsibility of machinery reliability
department, 545
604
Performance relationships, 85, 86, 110
Periodic conferences organization, 5
PFD. See Process flow diagram
P&ID. See Piping and instrumentation
diagram/drawing
Pinion shaft, 190
Piping, 335, 373
considerations for, 70
and instrumentation diagram/drawing
(P&ID), 12
Piston, 102
assembly, 101
rod packing, 98
Position indicator, 457
Positive displacement
compressors, 74
and dynamic characteristics, 78
Power, effect, 182
Pre-bid meeting, 8
compressor train pre-bid meeting,
agenda, 19
pre-bid procedure, fact summary, 18
vendor pre-bid meeting, details, 17
Pre-charged accumulator, 398, 400
Predictive maintenance (PDM) program, 502,
504, 506, 507
Predictive maintenance techniques, 156,
465, 515
Pressure drop, 71
Pressure reducing control, 391–395, 529
direct acting actuator and valve body used
for, 531
dynamic pump, 532
pressure control valve sensing decreasing
system pressure, 532
primary applications
in auxiliary systems, 529
for pressure reducing valves, 532
pump operating simultaneously, 532
Pressure reducing valve, 357, 372
Preventive maintenance (PM) program,
504, 505
Primary components of support stiffness
in order of decreasing increasing
influence, 124
Prime mover, 235
Process control valve, 39
Process engineers, 12
Process flow diagram (PFD), 12
Project best practices, 1
Project budget estimate, 4
Project engineer, 10
Proximity probe system, 197
Index
PTFE materials, 99
Pulsation dampeners, 103, 135
Pump, 39
bearing system condition monitoring, 555
component condition monitoring, 553
flow calculation, 39
maintenance, 508
mechanical requirements, 353
performance
centrifugal pumps, 353
effect of oil viscosity on centrifugal
pump, 354
positive displacement, 352
reliability, 43
safe operating area, 68
seal. See Pump mechanical seals
unit couplings, 354
Pumping rings, 413–415
circulation, 411, 416
Pump mechanical seals, 403–432, 554
components of
bellows, 404
secondary components of
adaptive hardware, 407, 408
o-rings, 407, 408
seal faces, 408
mating ring, 408
primary ring, 408
types of, 404–406
dual pressurized, 406, 407, 411, 431
dual un-pressurized, 406, 407, 416
non-pusher, 404–406
pusher, 404–406
Pump parallel operation, 51–53
identical pumps in parallel operation, 52
non-identical pumps operating in parallel, 52
Pump performance monitoring, 155, 352,
463, 512
Pump selection, 369
Pusher type seal, considerations for
external flush, 405
filtration, 405
single coil spring, 405
R
Radial bearing loads, 154
determination of, 192
Radial (centrifugal) impellers, 110
Radio tuner/vibration filter analogy, 201
Rankine cycle, 288
RCFZ. See Root Cause Failure Analysis
Reaction blading, 235
advantages, 236
disadvantages, 236
Reaction forces
gear unit design basis, 189
Reciprocating compressor cylinder valves,
73, 100, 101
Reciprocating compressors
medium and high speed, 94
Reciprocating gas engine, 94
Reduced moment convoluted (wavy)
diaphragm spacer coupling, 209
Refrigeration applications, 121
Reliability centered maintenance (RCM),
506, 552
Reliability groups
not incorporating operations and process
engineering input, 546
Reliability improvement, 74, 507, 509
cost of unavailability, 551
identifying targets, 550–551
normal component reliability comparison,
550–551
Relief device, 95
Relief valve
for positive displacement pumps, 536
selection, 355, 370
sizing chart, 538
Required head equation, 56
Required process data, 93
Reservoir, 335
heating requirements, 371
levels, 374
sizing, 351, 370
Reverse acting actuator, 389
and valve body, 527
Reverse dial indicator procedure, 215
Review machinery instruction manuals, prior
to shipment, 479
Rod, 102
Rod loading, 103
Rolls Royce RB211 two-shaft gas turbine, 295
Root cause analysis (RCA), 501, 570–571
abnormal conditions, define, 574
causes, eliminate, 575
component condition monitoring, 576
cost-effective manner, 570
develop an action plan, 575
DGS system monitoring, 564
equipment/component/system functions,
thorough knowledge of, 573
failure classifications, 578
gas turbine component condition
monitoring, 564
605
Index
Root cause analysis (RCA) (cont.)
guidelines, 568
to gathering facts, 576
initial fact finding, 571
affected components, define, 571
apparent problem, 571
baseline conditions/trends, 572
failed component inspection, 573
guide lines, 572
important facts, 572
inspection, 572
list all possible causes, 574
L.O./SO Syst., 559
machinery failures, 576
state root causes of problem, 575
Root Cause Failure Analysis (RCFA), 13, 568
Rotary pump, 333
Rotating boiler, 288
Rotating equipment, 577
firefighting, 577
Rotor, 154
case thermal movement, 208
critical speeds, 254–256
instabilities, prevention of, 193
natural frequency mode shapes, 131, 261
response modeling-rotor, 126
response output. See Rotor response output
data-dimensions, masses, and
unbalances, 127
Rotor response output, 130
drive end bearing (D.E.), 132
at non-drive end bearing (NDE), 132
Rotor stiffness, 126, 257
Rotor system
critical speed map, 255
designing, 173
different way of, 252
input, 126, 256–260
response (output), 260–263
at drive end bearing (DE), 263
at non-drive end bearing (NDE), 262
Rotor thrust balance, 171, 250–253, 380, 381
Rotor thrust force, 251
Running gear, 95
S
Safety and reliability issue, 37
Schematic (P&ID) review, 350
Scope, confirmation of, 350
Screw compressors, 74
advantages, 78
types, 80
606
twin screw-oil flooded, 83
twin screw-oil free (non oil injected), 80
Screw pv diagram, 80
Seal chamber, 419, 428
Seal design, 129
Seal gas, 433
Sealing fluid, 407, 408
Seal leakage, 415
monitoring, 415
Seal oil
pumps, 348
Seal oil, 441
Seals, 155
Seal stator, 444
Seismic vibration, 197
Selection guidelines, 91
Self-equalizing tilt-pad thrust bearing, 168
Sentinel valve maintenance
excessive prevention, 545
prevention, 226
Sentinel valves, 545
Shaft displacement, 200
monitor, 199
Shaft end seals, 279–281
expansion turbine, 280
Shaft end separation, 206
Shaft misalignment, 202
Shaft movement analysis, 196
Shaft position, 195
Shaft system, components, 211–212
Shaft tapers, 213
Shaft vibration, 187, 200
alarm, 198
Single admission steam turbine, 241
Single and multi-stage turbines
power limitations, 239
Single diaphragm spacer coupling, 208
Single extraction steam turbine, 241
Single flow steam turbine, 242
Single shaft, 298
industrial gas turbine, 293
Single stage expansion turbine
limiting factors, 239
Single stage turbines
assembly drawing, 238
efficiency of, 233
guidelines, 220, 539
steam turbines
common reliability problems, 220,
221, 540
hand valve recommendations, 225, 544
Single valve and multi-valve steam turbines
choice determining factors, 240
Index
Single valve steam turbines, 219, 239
Single valve turbine admission path, 224, 543
Site conditions data, 334
Site equipment, examples, 14–17
extraction-condensing steam turbine, 15
high-pressure centrifugal compressor, 15
horizontal oil console arrangement, 16
multiple, convoluted diaphragm-spacer
coupling, 16
Site installation procedures, 481
Site machinery instrumentation excellence
program, 516
Site procedures, 480
Site rotating equipment function, 476
Site storage, 483
Slide valve, 91
control, 92
operation, 93
Slow governor system response, 220–223, 540
condition monitoring action plan, 223,
542, 543
Small Kingsbury six-shoe, two direction thrust
bearing, 167
Spare critical machinery rotor storage
guidelines, 479
Spare parts, 482
Specific gravity, 53, 55
Specific impeller pattern, 59
Specific machinery component, 154, 462, 511
Specific speed, 110
Spillback valve, 187
Spline fits, 213
Spur gears
external, 189
internal, 189
Stage and section performance, 149, 150, 183
Stage curve, 146
Steam conditions, 227
limits, 227
Steam cycle efficiency, 272
Steam rate
determination, 230
Steam turbines, 208, 288
applications, 242–243
bearing bracket oil contamination
monitoring action plan, 540
control, 268
driven compressor train, 465
driven pump, 337
efficiencies, 231
fouling
causes, 247–249
gen sets, 233
governor system application chart, 275
manually exercising, 286
performance, 468, 557
power, 226
protection system, 276, 282–286
internal protection, 285
overspeed protection, 284
protection systems, 265, 286
shaft sealing systems, 281
shut-off valves, 278, 285
solo run functional test procedure
outline, 493
types, 234–241
vs. gas turbine, 288
Straight sleeve bearing liner, 192
Sub-component details, 351
Sub-systems, 346, 385, 523
Sub-vendor, 349
Suction recirculation, flow pattern, 59
Suction specific speed (NSS), 70
flow separation probability, relation to, 71
Suction throttling, 91
Supply pipe velocity, 371
checks, 336, 358
Supporting material, 2–5, 13, 17
rotating equipment
classifications of, 13
life span, 4
practices for, 2
Switches/transmitters, 373
System monitoring parameters, 154
and limits, 462, 511
System requirements, 351
System resistance, effect on, 180
curve, process changes, 182
T
Tandem seals, 447, 448
for dry gas applications, 447–449
with interstage labyrinth, 449
for saturated gas applications, 449
Tangential force vector, 190
Teflon packing, 408
Temperature control valves, 358, 372, 396, 534
Temperature transmitters, 449
Temporary test setup, 366
component requirements, 367
functional testing, 366
system requirements, 367
Test agenda, 362
Theoretical enthalpy, 228
Theoretical steam rate (T.S.R.), 228
607
Index
Thermal growth, 23
Thermocouples, 190
Throttle valve position indicator, 219
Thrust analysis, 250
Thrust bearings, 154
Thrust condition monitoring, 174
Thrust loading, 250
Thrust pad temperature and thrust load,
relationship between, 170
Tilt pad journal bearing, 192
assembly, 193
Time constraints, 2
Torque load, 188
Total radial load, 189
Total train control and protection objectives,
264–266
Training, 475
Transfer valve, sizing, 359, 371
Transient case pump auto transfer, 348
Transient response, 349
Transient torsional excitation, 210
frequency vs. motor speed, 210
Trending, 153, 164, 461, 510
Trending data, 462, 511
Trending guidelines, 164
Trend parameters, useful, 165
Trimetal, 190
Troubleshooting, 506, 507, 569
definition, 569
process, 570
root cause, 570–571. See also Root cause
analysis
T.S.R.. See Theoretical steam rate (T.S.R.)
Turbine
efficiency, 230
internal protection, 277
manually exercising, 278
overspeed protection, 277
power, 227
protection methods, 275–279
sentinel valves, use of, 225
speed control, 266–275
Turbo-compressor, 209
condition monitoring, 162
flow rate, effect on, 182
head, effect of
gas composition and temperature change
on, 180
608
impeller head change and curve shape
summary, 181
performance condition
monitoring facts, 162
parameters, 161
pressure ratio, effect on, 178
turbines functions, 235
Two pump operation, 347
Two-stage, tandem screw oil flooded
compressor, 83
Typical axial displacement monitor, 175
Typical axial thrust monitor, 174
Typical compressor oil film bearing
parameters, 128
Typical multi-stage refrigeration compressor, 119
Typical pipe sizing chart, 336, 359
U
Undamped critical speeds, 256
Unit connection orientation, 373
Unit trips, 400
Utility supply arrangement, 339, 361
V
Valve sizing, 530
Vaporization point, 418
Vapor margin, 422
Vapor pressure, 55
Variable speed, 90
Vendor Coordination Meeting (VCM), 335
agenda checklist, 38
Vendor exceptions, to specifications, 351
Vendor service representative available, at
factory acceptance test, 488
Vertical pump, in condensate hot well service, 67
Vessel, 335, 373
Vibration, 197
frequency, 200
severity chart, 199
Viscosity, 41, 55
Visual detection of gas flow (vent), 99
Volatile organic compound (VOC), 406
W
Wet screw, 92