D018268-0019
D017414-8954
Removable lid
This half of an
unwelded diaphragm
illustrates the
aerodynamic flow
path and liquid
cooling passages.
Main structural section (diffuser side)
Controlling Greenhouse Gases
SwRI researchers
develop advanced
centrifugal compressor
technology for
carbon capture
and sequestration
by J. Jeffrey Moore, Ph.D., Andrew Lerche,
Hector Delgado and Tim Allison, Ph.D.
C
Engineers in SwRI’s Fluids and Machinery Engineering Department are working with both
government and industry to advance technologies for reducing the release of greenhouse gas
emissions from power plants. Pictured are (from left) Senior Research Engineer Dr. Tim Allison,
Senior Research Engineer Andrew Lerche, Manager Dr. J. Jeffrey Moore and Research Engineer
Hector Delgado. The team applies various analysis techniques to solve problems associated with
rotating machinery and to advance new designs for industry.
6
Technology Today • Winter 2011
oal is used to generate about half
of all electricity consumed in the
United States. Unfortunately, coal
has a high carbon content and
produces twice the carbon dioxide (CO2)
emissions when burned compared to
natural gas. Because of growing concern
over greenhouse gas emissions, the U.S.
government and utilities are developing
technologies to separate CO2 both preand post-combustion.
To reduce the release of CO2 greenhouse gases to the atmosphere, industry
is pursuing technologies for sequestering CO2 emitted from new power cycles,
including integrated gasification combined cycle (IGCC) and oxy-fuel, as well
as from traditional pulverized coal (PC)
power plants. Significant progress has
been made in sequestration of CO2 from
power plants and other major producers
of greenhouse gas emissions. Compressing the captured CO2 stream requires
significant power, which affects plant
availability, capital expenditures and
operational costs. The power penalty for
carbon capture can be as high as 27 to 37
percent for a traditional PC power plant
and 13 to 17 percent for a typical IGCC
plant. Compression represents a significant percentage of this total.
Diffuser side of bulb
Return channel side of bulb
Main structural section (return channel side)
Novel compression and pumping
techniques
SwRI researchers, under funding
from the U.S. Department of Energy
National Energy Technology Laboratory
and industry partners, are working to
reduce this penalty by developing novel
combined compression and pumping
processes. The research supports reducing the energy requirements for carbon
capture and sequestration in electrical
power production. These technologies
are also applicable to other gases, including natural gas and air compression. The
primary objective of this work is to boost
the pressure of CO2 from near-atmospheric
to pipeline pressures with the minimal
amount of energy required.
Previous thermodynamic analysis
identified optimum processes for pressure rise in both liquid and gaseous
states. Isothermal compression is well
known to reduce the power requirements by minimizing the temperature
of the gas entering downstream stages.
Furthermore, because CO2 is a heavy
gas, it may be liquefied at moderate temperature and pressures using industrial
refrigeration systems that permit this
use of a centrifugal pump for the highpressure stages. Combined compressionliquefaction-pumping schemes are
being pursued in this study. Isothermal
compression can be accomplished with
an integrally geared machine, where
multiple pinions are driven by a common
gear. However, the reliability of integrally
geared compressors cannot match the inline centrifugal machines widely used in
the oil and gas industry. To address these
issues, SwRI researchers designed an
internally cooled, in-line centrifugal compressor diaphragm (patent pending) that
removes the heat of compression without
the need for external intercoolers
using liquid cooling.
The diaphragm removes heat
of compression between each
impeller by passing cooling water
through the diaphragm. Temperature increases, due to the work
input of the impeller, are reduced
through the diaphragm flow path,
thereby reducing the temperature
into the downstream stage. To
validate their predictions for both
aerodynamic and heat transfer
performance, SwRI researchers
built and tested a single-stage prototype diaphragm.
Significant challenges exist in cooling a high-velocity gas within the
compressor due to the
limited surface area
and the need for minimum pressure drop
of the gas stream.
However, using
three-dimensional
computational fluid
dynamics, researchers
developed conjugate
heat transfer models
that combine the diaphragm
structure, cooling fluid and the diaphragm process gas flow path. Using
these models, researchers achieved an
optimal design that provides good heat
transfer with no additional pressure
drop.
The diaphragms were designed to
permit fabrication by first machining the
individual components, then welding
D018305
D018307
SwRI researchers developed
a conjugate heat transfer
computational fluid
dynamics (CFD) model to
predict the aerodynamic flow
field of the compressor and
calculate the heat transfer to
the cooling fluid through the
aluminum diaphragm.
This internally cooled compressor diaphragm
removes the heat of compression inside a
multi-stage centrifugal compressor used for
CO2 compression applications.
308
D018
Technology Today • Winter 2011
7
D018268-0019
D017414-8954
Removable lid
This half of an
unwelded diaphragm
illustrates the
aerodynamic flow
path and liquid
cooling passages.
Main structural section (diffuser side)
Controlling Greenhouse Gases
SwRI researchers
develop advanced
centrifugal compressor
technology for
carbon capture
and sequestration
by J. Jeffrey Moore, Ph.D., Andrew Lerche,
Hector Delgado and Tim Allison, Ph.D.
C
Engineers in SwRI’s Fluids and Machinery Engineering Department are working with both
government and industry to advance technologies for reducing the release of greenhouse gas
emissions from power plants. Pictured are (from left) Senior Research Engineer Dr. Tim Allison,
Senior Research Engineer Andrew Lerche, Manager Dr. J. Jeffrey Moore and Research Engineer
Hector Delgado. The team applies various analysis techniques to solve problems associated with
rotating machinery and to advance new designs for industry.
6
Technology Today • Winter 2011
oal is used to generate about half
of all electricity consumed in the
United States. Unfortunately, coal
has a high carbon content and
produces twice the carbon dioxide (CO2)
emissions when burned compared to
natural gas. Because of growing concern
over greenhouse gas emissions, the U.S.
government and utilities are developing
technologies to separate CO2 both preand post-combustion.
To reduce the release of CO2 greenhouse gases to the atmosphere, industry
is pursuing technologies for sequestering CO2 emitted from new power cycles,
including integrated gasification combined cycle (IGCC) and oxy-fuel, as well
as from traditional pulverized coal (PC)
power plants. Significant progress has
been made in sequestration of CO2 from
power plants and other major producers
of greenhouse gas emissions. Compressing the captured CO2 stream requires
significant power, which affects plant
availability, capital expenditures and
operational costs. The power penalty for
carbon capture can be as high as 27 to 37
percent for a traditional PC power plant
and 13 to 17 percent for a typical IGCC
plant. Compression represents a significant percentage of this total.
Diffuser side of bulb
Return channel side of bulb
Main structural section (return channel side)
Novel compression and pumping
techniques
SwRI researchers, under funding
from the U.S. Department of Energy
National Energy Technology Laboratory
and industry partners, are working to
reduce this penalty by developing novel
combined compression and pumping
processes. The research supports reducing the energy requirements for carbon
capture and sequestration in electrical
power production. These technologies
are also applicable to other gases, including natural gas and air compression. The
primary objective of this work is to boost
the pressure of CO2 from near-atmospheric
to pipeline pressures with the minimal
amount of energy required.
Previous thermodynamic analysis
identified optimum processes for pressure rise in both liquid and gaseous
states. Isothermal compression is well
known to reduce the power requirements by minimizing the temperature
of the gas entering downstream stages.
Furthermore, because CO2 is a heavy
gas, it may be liquefied at moderate temperature and pressures using industrial
refrigeration systems that permit this
use of a centrifugal pump for the highpressure stages. Combined compressionliquefaction-pumping schemes are
being pursued in this study. Isothermal
compression can be accomplished with
an integrally geared machine, where
multiple pinions are driven by a common
gear. However, the reliability of integrally
geared compressors cannot match the inline centrifugal machines widely used in
the oil and gas industry. To address these
issues, SwRI researchers designed an
internally cooled, in-line centrifugal compressor diaphragm (patent pending) that
removes the heat of compression without
the need for external intercoolers
using liquid cooling.
The diaphragm removes heat
of compression between each
impeller by passing cooling water
through the diaphragm. Temperature increases, due to the work
input of the impeller, are reduced
through the diaphragm flow path,
thereby reducing the temperature
into the downstream stage. To
validate their predictions for both
aerodynamic and heat transfer
performance, SwRI researchers
built and tested a single-stage prototype diaphragm.
Significant challenges exist in cooling a high-velocity gas within the
compressor due to the
limited surface area
and the need for minimum pressure drop
of the gas stream.
However, using
three-dimensional
computational fluid
dynamics, researchers
developed conjugate
heat transfer models
that combine the diaphragm
structure, cooling fluid and the diaphragm process gas flow path. Using
these models, researchers achieved an
optimal design that provides good heat
transfer with no additional pressure
drop.
The diaphragms were designed to
permit fabrication by first machining the
individual components, then welding
D018305
D018307
SwRI researchers developed
a conjugate heat transfer
computational fluid
dynamics (CFD) model to
predict the aerodynamic flow
field of the compressor and
calculate the heat transfer to
the cooling fluid through the
aluminum diaphragm.
This internally cooled compressor diaphragm
removes the heat of compression inside a
multi-stage centrifugal compressor used for
CO2 compression applications.
308
D018
Technology Today • Winter 2011
7
D018294
D018296
D018297
The SwRI-designed CO2 pump test loop, located
at the Institute, will be used to characterize the
performance and mechanical behavior of a multistage turbopump adapted for use with liquid CO2.
Valve
Knock-out Drum
Motor
CO2
Pump
Tank
D018306
Measured pump performance
showing head (pressure rise)
versus flow (in gallons per minute).
the pieces together and heat-treating and
machining the weldment.
Compressor loop test rig
Researchers retrofitted the compressor diaphragm into the SwRI closed-loop
centrifugal compressor casing. Instrumentation was installed internally in
the compressor diaphragm as well as
externally throughout the test loop to
measure aerodynamic and heat transfer
performance.
Multiple pressure and temperature
sensors were installed at various stations
throughout the flow path. An end-to-end
calibration of all temperature probes was
performed prior to installing the probes
inside the loop and diaphragm. Cables
for all of the internal instrumentation
were passed through the compressor
casing via sealing glands.
Researchers developed the data
acquisition and loop control software
for the CO2 compressor loop using
LabVIEW™ software. The program is
capable of sending and receiving data to
and from PSI NetScanner 9116 pressure
scanners and National Instruments
CompactDAQ™ systems to control and
monitor the test rig.
Temperature measurements at various locations inside the compressor were
converted to total temperatures. The flow
velocity at the suction and discharge measurement locations was used to calculate
8
the total temperature and pressure
at these locations
using procedures
based on those
in the ASME PTC10 specification.
This conversion
procedure was
also performed
for temperature
measurements at
the impeller exit, diffuser vane exit and
return-channel bend. Stage performance
was calculated at the bridgeover locations to eliminate the effect of the inlet
and exit collector. This approach better
represents a central stage in a multi-stage
compressor.
Performance tests were completed
at three speeds (10,280, 11,565 and 12,850
rpm), three cooling flow rates (0, 12 and
20 gpm), three gas suction pressures
(30, 60 and 90 psia) and two cooling water
temperatures over a range of compressor
flows. Adiabatic refers to no heat transfer
(no cooling water flow) while diabatic
is with cooling. Because this program
tested only a single-stage compressor, the aerodynamic work input (and
power) remained the same between the
adiabatic and diabatic tests. Note that
the benefit of the cooled diaphragm
is realized on downstream stages in a
multi-stage compressor. In some cases,
the gas flow temperature decreased by
Technology Today • Winter 2011
Schematic of the liquid CO2 pump loop
constructed to perform qualification tests of
a 12-stage centrifugal turbopump.
Conclusions
This plot shows the measured and predicted
total temperature profiles throughout the
compressor at different stations. Both adiabatic
and diabatic conditions are given. The results
show significant temperature reduction with the
cooled diaphragm and good correlation with
the CFD predictions.
more than 20 degrees F, demonstrating
the effectiveness of the cooled diaphragm
concept. Using this measured amount of
heat transfer, hypothetically a five-stage
centrifugal compressor would require
20 percent less power using the internal
cooling technology compared to a conventional compressor.
Pump loop test rig
Because a liquid CO2 pump of the
size required for this service was not
commercially available, SwRI engineers
worked with a pump manufacturer to
All of the goals set forth in the test
program were accomplished. Early in 2011,
SwRI was awarded a $9.9 million contract
for Phase 3 to develop a multi-stage version
of the cooled diaphragm and to construct a
demonstration facility at SwRI consisting of
an integrated compression-liquefactionpumping pilot plant. Tests will commence
in 2013. This pilot-scale testing will increase
to a technical readiness level, permitting
implementation in a full-scale power plant.
adapt an existing pump design for CO2
service. The pump test rig consists of a
newly constructed liquid CO2 test loop
and an industrial multi-stage turbopump
adapted for service on liquid CO2. The
pump is the smallest frame size for this
product line, while still retaining the same
configuration as large-scale pumps that
would be used for power plant applications. The smaller size was selected to
keep the cost of the facility to a minimum, yet still provide valid performance
and mechanical data.
The pump loop consists of a 12-stage
centrifugal pump driven by a variablespeed electric motor and is fed from a
1,000-gallon pressurized vessel that maintains liquid CO2 at its boiling temperature
of -12 degrees F at 250 psia. The discharge of the pump feeds an orifice
flow meter followed by a control valve
that drops the pressure from 2,215 psia
to 250 psia. The control valve discharges into a knock-out drum for
liquid-gas separation, since some flashing of the gas back to the vapor phase
will occur. Finally, the liquid CO2 will be
returned to the main vessel through a
drain line, and the remaining gaseous
CO2 will be vented to the atmosphere
through a back-pressure control valve.
The tests measured pump performance
to quantify the power requirements.
Also, mechanical performance was
quantified, including vibration, temperatures and seal flows.
Technology Today • Winter 2011
Questions about this article?
Contact Moore at (210) 522-5812 or
jeff.moore@swri.org.
Acknowledgments
The authors would like to thank the U.S.
Department of Energy National Energy Technology Laboratory for providing the majority
of the funding for this work, and Dresser-Rand
and BP for additional co-funding. Machining
of the compressor diaphragm components
by Supervisor Robert Rendon and Staff Craft
Technician Marc Johnson in SwRI’s Space Science and Engineering Division also is acknowledged. Thanks also to Principal Technicians
Brian Moreland and Richard Skinner, both of
the Mechanical Engineering Division, for the
detail design of the cooled diaphragms and
fabrication of the pump loop, respectively.
9
D018294
D018296
D018297
The SwRI-designed CO2 pump test loop, located
at the Institute, will be used to characterize the
performance and mechanical behavior of a multistage turbopump adapted for use with liquid CO2.
Valve
Knock-out Drum
Motor
CO2
Pump
Tank
D018306
Measured pump performance
showing head (pressure rise)
versus flow (in gallons per minute).
the pieces together and heat-treating and
machining the weldment.
Compressor loop test rig
Researchers retrofitted the compressor diaphragm into the SwRI closed-loop
centrifugal compressor casing. Instrumentation was installed internally in
the compressor diaphragm as well as
externally throughout the test loop to
measure aerodynamic and heat transfer
performance.
Multiple pressure and temperature
sensors were installed at various stations
throughout the flow path. An end-to-end
calibration of all temperature probes was
performed prior to installing the probes
inside the loop and diaphragm. Cables
for all of the internal instrumentation
were passed through the compressor
casing via sealing glands.
Researchers developed the data
acquisition and loop control software
for the CO2 compressor loop using
LabVIEW™ software. The program is
capable of sending and receiving data to
and from PSI NetScanner 9116 pressure
scanners and National Instruments
CompactDAQ™ systems to control and
monitor the test rig.
Temperature measurements at various locations inside the compressor were
converted to total temperatures. The flow
velocity at the suction and discharge measurement locations was used to calculate
8
the total temperature and pressure
at these locations
using procedures
based on those
in the ASME PTC10 specification.
This conversion
procedure was
also performed
for temperature
measurements at
the impeller exit, diffuser vane exit and
return-channel bend. Stage performance
was calculated at the bridgeover locations to eliminate the effect of the inlet
and exit collector. This approach better
represents a central stage in a multi-stage
compressor.
Performance tests were completed
at three speeds (10,280, 11,565 and 12,850
rpm), three cooling flow rates (0, 12 and
20 gpm), three gas suction pressures
(30, 60 and 90 psia) and two cooling water
temperatures over a range of compressor
flows. Adiabatic refers to no heat transfer
(no cooling water flow) while diabatic
is with cooling. Because this program
tested only a single-stage compressor, the aerodynamic work input (and
power) remained the same between the
adiabatic and diabatic tests. Note that
the benefit of the cooled diaphragm
is realized on downstream stages in a
multi-stage compressor. In some cases,
the gas flow temperature decreased by
Technology Today • Winter 2011
Schematic of the liquid CO2 pump loop
constructed to perform qualification tests of
a 12-stage centrifugal turbopump.
Conclusions
This plot shows the measured and predicted
total temperature profiles throughout the
compressor at different stations. Both adiabatic
and diabatic conditions are given. The results
show significant temperature reduction with the
cooled diaphragm and good correlation with
the CFD predictions.
more than 20 degrees F, demonstrating
the effectiveness of the cooled diaphragm
concept. Using this measured amount of
heat transfer, hypothetically a five-stage
centrifugal compressor would require
20 percent less power using the internal
cooling technology compared to a conventional compressor.
Pump loop test rig
Because a liquid CO2 pump of the
size required for this service was not
commercially available, SwRI engineers
worked with a pump manufacturer to
All of the goals set forth in the test
program were accomplished. Early in 2011,
SwRI was awarded a $9.9 million contract
for Phase 3 to develop a multi-stage version
of the cooled diaphragm and to construct a
demonstration facility at SwRI consisting of
an integrated compression-liquefactionpumping pilot plant. Tests will commence
in 2013. This pilot-scale testing will increase
to a technical readiness level, permitting
implementation in a full-scale power plant.
adapt an existing pump design for CO2
service. The pump test rig consists of a
newly constructed liquid CO2 test loop
and an industrial multi-stage turbopump
adapted for service on liquid CO2. The
pump is the smallest frame size for this
product line, while still retaining the same
configuration as large-scale pumps that
would be used for power plant applications. The smaller size was selected to
keep the cost of the facility to a minimum, yet still provide valid performance
and mechanical data.
The pump loop consists of a 12-stage
centrifugal pump driven by a variablespeed electric motor and is fed from a
1,000-gallon pressurized vessel that maintains liquid CO2 at its boiling temperature
of -12 degrees F at 250 psia. The discharge of the pump feeds an orifice
flow meter followed by a control valve
that drops the pressure from 2,215 psia
to 250 psia. The control valve discharges into a knock-out drum for
liquid-gas separation, since some flashing of the gas back to the vapor phase
will occur. Finally, the liquid CO2 will be
returned to the main vessel through a
drain line, and the remaining gaseous
CO2 will be vented to the atmosphere
through a back-pressure control valve.
The tests measured pump performance
to quantify the power requirements.
Also, mechanical performance was
quantified, including vibration, temperatures and seal flows.
Technology Today • Winter 2011
Questions about this article?
Contact Moore at (210) 522-5812 or
jeff.moore@swri.org.
Acknowledgments
The authors would like to thank the U.S.
Department of Energy National Energy Technology Laboratory for providing the majority
of the funding for this work, and Dresser-Rand
and BP for additional co-funding. Machining
of the compressor diaphragm components
by Supervisor Robert Rendon and Staff Craft
Technician Marc Johnson in SwRI’s Space Science and Engineering Division also is acknowledged. Thanks also to Principal Technicians
Brian Moreland and Richard Skinner, both of
the Mechanical Engineering Division, for the
detail design of the cooled diaphragms and
fabrication of the pump loop, respectively.
9