CP Interference January 2008 ©NACE International, 2006 IMPORTANT NOTICE Neither the NACE International, its officers, directors, nor members thereof accept any responsibility for the use of the methods and materials discussed herein. No authorization is implied concerning the use of patented or copyrighted material. The information is advisory only and the use of the materials and methods is solely at the risk of the user. It is the responsibility of the each person to be aware of current local, state and federal regulations. This course is not intended to provide comprehensive coverage of regulations. Printed in the United States. All rights reserved. Reproduction of contents in whole or part or transfer into electronic or photographic storage without permission of copyright owner is expressly forbidden. Acknowledgements The scope, desired learning outcomes and performance criteria of this course were developed by the CP Task Group under the auspices of the NACE Education Administrative Committee. The time and expertise of several members of NACE International have gone into the development of this course—and its task analysis, course outline, student manual, classroom lab manual, presentation slides, and examinations. Their dedication and efforts are greatly appreciated. On behalf of NACE, we would like to thank the task group for its work. Their efforts were extraordinary and their goal was in the best interest of public service—to develop and provide a much needed training program that would help improve corrosion control efforts industry-wide. We also wish to thank their employers for being generously supportive of the substantial work and personal time that the members dedicated to this program. CP Interference Course Development Task Group Paul Nichols, Task Group Chairman Brian Holtsbaum Kevin Parker David A. Schramm Steven R. Zurbuchen Steven Nelson Donald R. Mayfield Shell Global Solutions, Houston, Texas CC Technologies Canada, Ltd., Calgary, Alberta CC Technologies, Mt. Pleasant, Michigan EN Engineering, Woodridge, Illinois EN Engineering, Topeka, Kansas Columbia Gas Transmission, Charleston, West Virginia Dominion Transmission, Delmont, Pennsylvania CP Interference Daily Course Outline DAY ONE Introduction, Welcome, Overview Chapter 1 Stray Current Interference DAY TWO Chapter 2 DC Interference (Includes Experiment 2-1) DAY THREE MORNING Chapter 2 DC Interference AFTERNOON Chapter 3 AC Interference (Includes experiments 3-1, 3-2, and 3-3) DAY FOUR Chapter 3 AC Interference DAY FIVE MORNING Chapter 3 AC Interference AFTERNOON Chapter 4 Telluric Current Interference DAY SIX MORNING Exam CP Interference Course Manual © NACE International, 2006 January 2007 Introduction Introduction The Cathodic Protection (CP) Interference course is a six-day course focusing on alternating current (AC) and direct current (DC) interference. The course includes in-depth coverage of both the theoretical concepts and the practical application of identifying interference and interference mitigation techniques. Students will learn to identify the causes and effects of interference as well as conduct tests to determine if an interference condition exists and perform calculations required to predict AC interference. The course is presented in a format of lecture, discussion and hands-on, in-class experiments, case studies and group exercises. There is a written examination at the conclusion of the course. Who Should Attend This course is designed for persons who have extensive CP field experience, a strong background in mathematics, and a strong technical background in CP. Prerequisites • CP 3–Cathodic Protection Technologist certification recommended • Minimum of 3 years CP work experience Length The course begins at 1 p.m. on Sunday and concludes Friday afternoon. Daily class hours: 8 a.m. to 6:30 p.m. Monday through Thursday and 8 a.m. to 3 p.m. Friday. Reference Book Students will receive the CP Interference Course Manual prior to the start of the course. A course manual on CD-ROM will be provided to students on-site. CP Interference Course Manual © NACE International, 2006 July 2007 1 Introduction Quizzes and Examinations There will be four (4) quizzes distributed during the week and reviewed in class by the instructors. This course has a written final examination. The final examinations will be given on Friday. The written final examination is open-book and students may bring reference materials and notes into the examination room. Non-communicating, battery-operated, silent, non-printing calculators, including calculators with alphanumeric keypads, are permitted for use during the examination. Calculating and computing devices having a QWERTY keypad arrangement similar to a typewriter or keyboard are not permitted. Such devices include but are not limited to palmtop, laptop, handheld, and desktop computers, calculators, databanks, data collectors, and organizers. Also excluded for use during the examination are communication devices such as pagers and cell phones along with cameras and recorders. A score of 70% or greater on the examination is required for successful completion of the course. All questions are from the concepts discussed in this training manual. You will receive written notification of your exam results as quickly as possible. Your results will not be available on Friday. Introductions We would like for each of you to stand, one at a time and introduce yourself to the class. Tell us: • Your name • Your company’s name and location • Your job function • Your experience related to CP Interference. CP Interference Course Manual © NACE International, 2006 July 2007 2 CP Interference Course Manual Table of Contents General Course Information Daily Course Outline Introduction Chapter 1–Stray Current Interference 1.1 Historical Background ........................................................... 1:1 1.2 Typical Stray Current Circuit Arising from a Transit System Operation ................................................................. 1:5 1.3 Stray Current Charge Transfer Reactions on a..................... Metallic Structure 1:6 1.4 Effects of Stray Current on Metallic Structures ..................... 1:9 1.4.1 1.4.2 1.4.3 At the Current Discharge Location...................................... At Area of Current Pick-Up ................................................. Along the Structure ............................................................. 1:9 1:15 1:19 1.5 Summary .............................................................................. 1:21 Summary of Equations.................................................................. 1:22 Figures Fig. 1-1 Fig. 1-2 Fig. 1-3 Fig. 1-4 Fig. 1-5 Fig. 1-6 Early Electric Trolley.............................................................. Pipe-to-soil Potential Changes due to Transit System Stray Current Activity were Recorded on Smoked Charts.. Co-efficient of Corrosion at Different Frequencies for Iron Electrode Denoted as Average Electrode Loss........... Typical Stray Current Paths Around a DC Transit System .... Typical Stray Current Interference on a Metallic Underground Structure ....................................................... Simplified pH Pourbaix Diagram For Iron in Water at 25ºC Showing Potential Shift Direction for Current Pick-up and Discharge at Low pH ................................................... CP Interference Course Manual © NACE International, 2006 June 2007 1:1 1:3 1:4 1:6 1:6 1:8 Fig. 1-7 Fig. 1-8 Fig. 1-9 Fig. 1-10 Fig. 1-11 Fig. 1-12 Fig. 1-13 Fig. 1-14 Fig. 1-15a Fig. 1-15b Fig. 1-16 Simplified pH Pourbaix Diagram For Iron in Water at 25ºC Showing Potential Shift Direction for Current Pick-up and Discharge at High pH................................................... Current Discharge from a Metal Structure to Earth via an Oxidation Reaction ........................................................ Superposition of a Stray Current and a Cathodic Protection Current at a Metal/Electrolyte Interface .............................. Randle’s Electrical Circuit Model of a Metal/Electrolyte Interface.............................................................................. Theoretical Conditions of Corrosion, Immunity and Passivation of (a) Aluminum at 25ºC and (b) Lead at 25ºC ................................................................. Comparison of Zn and Al Coatings for Corrosion Resistance as Functions of pH ........................................... Typical Section Through a Joint in Two Types of PCCP ....... Cathodic Blistering/Disbondment of Protective Coating ........ Stray Current Discharge and Pick-Up Around an Electrically Discontinuous Joint Though the Earth.............. Stray Current Discharge and Pick-Up Through the Internal Aqueous Medium Around an Electrically Discontinuous Bell and Spigot Joint on Cast Iron Piping.... Stray Current Circuit in an AC Electrical Distribution System................................................................................ 1:9 1:10 1:10 1:14 1:16 1:17 1:18 1:19 1:19 1:20 1:20 Tables Table 1-1 Theoretical Consumption Rates of Various Metals and Substances .................................................................. 1:12 Table 1-2 Electrochemical and Current Density Equivalence with Corrosion Rate.................................................................... 1:13 Chapter 2–DC Interference 2.1 Introduction ........................................................................... 2:1 2.2 Detecting Stray Current ........................................................ 2:23 2.2.1 Mitigation of Interference Effects from Impressed Current Cathodic Protection Systems ..................................... 2:24 a. Source Removal or Output Reduction .......................... 2:25 b. Installation of Isolating Fittings...................................... 2:26 c. Burying a Metallic Shield Next to the Interfered-with Structure .................................................................... 2:27 d. Installation of Galvanic Anodes on Interfered-with Structure at Point of Stray Current Discharge............ 2:28 e. Installation of an Impressed Current Distribution System on the Interfered-with Structure at Point of Stray Current Discharge................................................................... 2:33 f. i. Installing a Bond Between the Interfered-with and CP Interference Course Manual © NACE International, 2006 June 2007 Interfering Structures................................................ ii. Calculation of Bond Resistance ............................... g. Use of Coatings in the Mitigation of Interference Effects 2.2.2 2:33 2:35 2:40 Other Sources of DC Stray Current .................................... 2:41 a. DC Transit Systems ...................................................... 2:42 i. Analysis of Transit System Stray Currents ............... 2:44 ii. Mitigation of Transit System Stray Currents ............. 2:51 b. High Voltage Direct Current (HVDC) Electrical Transmission Systems ..................................................................... 2:55 c. DC Welding Operations ................................................ 2:57 Experiment 2-1: To Demonstrate DC Interference and Its Mitigation........................................................ 2:59 ………………………………………………………. ….. 2.64 Summary of Equations ............................................................................ 2:65 Case Study Figures Fig. 2-1 Fig. 2-2 Fig. 2-3 Fig. 2-4 Fig. 2-5 Fig. 2-6 Fig. 2-7a Fig. 2-7b Fig. 2-8 Fig. 2-9 Fig. 2-10 Fig. 2-11 Fig. 2-12 Fig. 2-13 Fig. 2-14 Fig. 2-15 Fig. 2-16 Parallel Current Paths in the Earth .................................... Parallel Current Paths in a Pipeline Cathodic Protection Section................................................................................ Parallel Current Paths in Vertically Stratified Soil Conditions Parallel Current Paths in Horizontally Stratified Soil Conditions........................................................................... Polarization Test Results....................................................... Stray Current in a Metallic Structure Parallel to a Cathodically Protected Structure ........................................ Voltage vs. Distance from a Vertically Oriented Anode ......... Multiple Vertical Anodes Connected to a Common Header Cable ..................................................................... Multiple Horizontal Anodes Connected to a Common Header Cable ..................................................................... Hemispherical Electrode........................................................ Cathodic Protection Circuit Model with Foreign Structure Intercepting the Anode Gradient......................................... Potential Profile along the Interfered-with Structure .............. Electrical Model for Interfered-with Pipe ................................ Attenuation Model.................................................................. Voltage Gradient in the Earth Around a Cathodically Protected Bare Pipeline ...................................................... Cathodic Protection Circuit Model ......................................... Cathodic Protection Circuit Model with Foreign Structure Intercepting the Anode Gradient......................................... Stray Current in a Foreign Metallic Structure that Intercepts both the Anodic and Cathodic Voltage Gradient................. CP Interference Course Manual © NACE International, 2006 June 2007 2:1 2:2 2:3 2:3 4:5 2:5 2:6 2:7 2:8 2:9 2:11 2:14 2:14 2:15 2:18 2:18 2:19 2:20 Fig. 2-17 Cathodic Protection Circuit Model with Foreign Structure Intercepting both Anodic and Cathodic Voltage Gradient... Fig. 2-18 Stray Current in a Foreign Metallic Structure that Intercepts the Cathodic Protection Gradient........................................ Fig. 2-19 Cathodic Protection Circuit Model for Foreign Structure Intercepting the Cathodic Voltage Gradient........................ Fig. 2:20 Typical Potential Profile on an Interfered-with Structure that Intersects both Anodic and Cathodic Voltage Gradient with the Current Source Interrupted..................... Fig. 2-21 Current Changes In and Near an Interfered-with Structure ... Fig. 2-22 Stray Current Arising from Installation of Isolating Fittings .... Fig. 2-23 Using a Buried Metallic Cable or Pipe as a Shield to Reduce Stray Current Interference..................................... Fig. 2-24 Cathodic Protection Current Model for a Buried Metallic Shield Connected to the Negative Terminal of the Transformer-Rectifier.......................................................... Fig. 2-25 Interference Mitigation using Galvanic Anodes at Stray Current Discharge Location ................................................ Fig. 2-26 Electrical Circuit Model for Mitigating Stray Current Interference at a Stray Current Discharge Site Using Galvanic Anodes................................................................. Fig. 2-27 Potential Profile Changes on a Pipeline where Stray Current is Discharging in an End-Wise Pattern .................. Fig. 2-28 Interference Mitigation Using a Resistance Bond.................. Fig. 2-29 Measurements Required to Determine Size of Resistance Bond Re .............................................................................. Fig. 2-30 Use of a Dielectric Coating to Mitigate Interference .............. Fig. 2-31 Typical Stray Current Paths Around a DC Transit System .... Fig. 2-32 Typical Structure-to-Soil Potential Recording with Time Caused by Interference from a DC Transit System ............ Fig. 2-33 Current Clamp Used to Measure Pipeline Currents .............. Fig. 2-34 Line Current Survey to Locate Source of Interference Using IR-Drop Test Stations ............................................... Fig. 2-35 Line Current Plots for Example in Figure 2-34 ...................... Fig. 2-36 Exposure Survey to Locate Point of Maximum Exposure...... Fig. 2-37 Exposure Survey Plots for Example in Figure 2-36 ............... Fig. 2-38 Mutual Survey to Confirm Source of Interference.................. Fig. 2-39 Pipe-to-Soil Potential Versus Pipe-to-Rail Potential for Example in Figure 2-38....................................................... Fig. 2-40 Exposure Survey Conducted Without the Measurement Of Pipeline Currents ........................................................... Fig. 2-41 Exposure Survey Plots for Example in Figure 2-40 ............... Fig. 2-42a Typical Embedded Track Installation..................................... Fig. 2-42b Typical Direct-Fixation Isolating Fastener ............................. Fig. 2-43 Typical Utilities Drainage System at a Transit Substation ..... Fig. 2-44 Schematic Showing Circulating Current between Transit Substations Through Direct Bonds to Utilities .................... Fig. 2-45 Forced Drainage Bonds Using a Potential Controlled Rectifier............................................................................... Fig. 2-46 Electrical Schematic for a HVDC System .............................. CP Interference Course Manual © NACE International, 2006 June 2007 2:20 2:21 2:22 2:23 2:24 2:26 2:27 2:28 2:29 2:30 2:33 2:34 2:36 2:41 2:42 2:43 2:44 2:45 2:46 2:47 2:48 2:48 2:49 2:50 2:50 2:52 2:52 2:52 2:53 2:54 2:55 Fig. 2-47 Potential-Time Plot for a Metallic Structure being Interfered-with by a HVDC System..................................... Fig. 2-48 Stray Current Caused by DC Welding Operations ................ 2:57 2:58 Experiment Schematic No. 1................................................................... Experiment Schematic No. 2................................................................... Experiment Schematic No. 3................................................................... 2:59 2:60 2:61 Table 2-1 Specific Leakage Resistances and Conductances in 1000 Ω-cm Soil or Water ....................................................... Table 2-2 Types of Reverse Current Switches ...................................... 2:13 2:54 Tables Chapter 3–AC Interference 3.1 Introduction ........................................................................... 3.1.1 3.1.2 3.1.3 Experiment 3-1: 3:1 Electrostatic (Capacitive) Coupling..................................... 3:2 Electromagnetic (Inductive) Coupling ................................. 3:11 Conductive Coupling (Resistive Coupling) During Powerline Fault Conditions........................................................................... 3:14 To Demonstrate the Effects of Electrostatic Induction ............................................................................. 3:16 3.2 Basic Theory of Electromagnetically Induced Voltages ........ 3:19 3.2.1 3.2.2 Experiment 3-2: AC Circuit Theory ............................................................... The Nature of Induced AC Pipeline Voltages ..................... 3:19 3:34 To Demonstrate the Effects of Electromagnetic Induction ............................................................................. 3:42 3.3 Induced AC Voltages ............................................................ 3:44 3.3.1 3.3.2 Experiment 3-3: Factors that Affect the Longitudinal Electric Field............... Factors that Affect the Pipeline Voltages............................ 3:44 3:48 To Further Investigate the Effects of Electromagnetic Induction ............................................................................. 3:57 3.4 Deleterious Effects of AC Interference.................................. 3:60 3.4.1 3.4.2 3.4.3 Electric Shock Hazards....................................................... AC Corrosion ...................................................................... .1 Theory........................................................................... .2 AC Corrosion Case Histories........................................ .3 AC Corrosion Field Test Procedures ............................ Fault Current Effects........................................................... CP Interference Course Manual © NACE International, 2006 June 2007 3:60 3:67 3:67 3:75 3:90 3:93 3.5 Induced AC Voltage Prediction and Mitigation Calculations . 3.5.1 3.5.2 3.5.3 3:95 Data Gathering ................................................................... Field Estimation of LEF....................................................... Measurement and Interpretation of Soil Resistivity Data.... 3:95 3:97 3:98 3.6 Prediction of Steady-State Induced AC Voltages.................. 3:102 3.6.1 3.6.2 3.6.3 3.6.4 3.6.5 Introduction ......................................................................... Calculation of Pipeline Electrical Characteristics................ Sectionalization of Pipeline-Powerline Route ..................... Determination of Longitudinal Electric Field (LEF) ............. Calculation of Induced Pipeline Voltages ........................... 3:102 3:102 3:106 3:107 3:110 3.7 Prediction of Fault Voltages .................................................. 3:115 3.7.1 3.7.2 3.7.3 3.7.4 Introduction ......................................................................... Conductive Coupling Due to Fault Currents ....................... Inductive Coupling Due to Fault Currents........................... Other Related Calculations................................................. (a) Ground Electrode Resistance .................................... (b) Step and Touch Potential .......................................... (c) Conductor Size .......................................................... 3:115 3:115 3:122 3:123 3:123 3:125 3:126 3.8 Equipment for AC Mitigation ................................................. 3:126 3.8.1 3.8.2 3.8.3 DC Decoupling Devices...................................................... Test Stations....................................................................... Sacrificial Anodes ............................................................... 3:126 3:138 3:139 Group Activity – AC Mitigation System Design ....................................... 3:142 Summary of Equations ............................................................................ 3:145 Figures Fig. 3-1a Single Horizontal 3φ Circuit with Shield Wires....................... Fig. 3-1b Distribution System (1φ 4kV Primary and 2φ 240V Secondary with Neutral)........................................................................ Fig. 3-2 AC Voltage Waveforms in a 3φ Circuit .................................. Fig. 3-3 Elements of a Capacitor ........................................................ Fig. 3-4 Electrostatic Coupling during Pipeline Construction .............. Fig. 3-5 Voltage Divider Circuits – Resistive (left) and Capacitive (right) .................................................................................. Fig. 3-6 Calculation of Typical Capacitance Values for a Pipe on Skids .............................................................................. Fig. 3-7 Calculation of Typical Electrostatically Induced Voltage for a Pipe on Skids.............................................................. Fig. 3-8 Calculation of Typical Shock Current Resulting from Electrostatic Coupling ......................................................... Fig. 3-9 Calculation of Typical Electrostatically Induced Voltage CP Interference Course Manual © NACE International, 2006 June 2007 3:2 3:2 3:2 3:3 3:4 3:5 3:6 3:7 3:8 for an Automobile................................................................ Fig. 3-10 Calculation of Typical Electrostatically Induced Voltage for a Buried Pipe ................................................................. Fig. 3-11 Electromagnetic Field Created by Current Flow in a Wire..... Fig. 3-12 Electromagnetic Induction in a Multiple-Turn, Iron-Core Transformer ........................................................................ Fig. 3-13 Electromagnetic Induction in a Single-Turn, Air-Core Transformer ........................................................................ Fig. 3-14 Electromagnetic Coupling Between a Pipeline and an Overhead AC Powerline ..................................................... Fig. 3-15 Conductive Coupling During Line-to-Ground Fault Conditions........................................................................... Fig. 3-16 Determination of Voltage on a Transformer Secondary ........ Fig. 3-17 Effect of Interconnecting the Secondary Windings ................ Fig. 3-18 Effect on Polarity on a Series Combination of DC Voltage Sources .............................................................................. Fig. 3-19 Effect of “Polarity” on a Series Combination of AC Voltage Sources .............................................................................. Fig. 3-20 In-Phase 60 Hz AC Waveform .............................................. Fig. 3-21 Typical Electrical Distribution Transformer ............................ Fig. 3-22 Typical Residential Electrical Service .................................... Fig. 3-23 AC Waveforms on a Residential Electrical Service ............... Fig. 3-24 Plot of General Equation for Sinusoidal AC Waveforms........ Fig. 3-25 Typical Phasor Diagram ........................................................ Fig. 3-26 Series Combination of AC Voltage Sources .......................... Fig. 3-27 Phasor Diagram for Problem in Figure 3-26 .......................... Fig. 3-28 Determination of Current through a Capacitor....................... Fig. 3-29 Voltage and Current Waveforms for a Purely Capacitive Circuit.................................................................................. Fig. 3-30 Determination of Current through an Inductor ....................... Fig. 3-31 Voltage and Current Waveforms for a Purely Inductive Circuit.................................................................................. Fig. 3-32 Phasor Representation of a Three-Phase Circuit .................. Fig. 3-33 Electric Model of Single Pipe Section .................................... Fig. 3-34 Simplified Electrical Model of Single Pipe Section................. Fig. 3-35 Simplified Electrical Model of Single Pipe Section................. Fig. 3-36 Series Combination of Multiple Pipe Sections ....................... Fig. 3-37 Series Combination of Two Pipe Sections ............................ Fig. 3-38 Series Combination of Two Pipe Sections (Simplified).......... Fig. 3-39 Circuit Analysis Using Kirchhoff’s Law................................... Fig. 3-40 Circuit Analysis Using Kirchhoff’s Law................................... Fig. 3-41 Induced AC Voltage Profile Along Two-Section Pipe Method of Figure 3-39 ........................................................ Fig. 3-42 Profile of Induced AC Voltages and their Phase Angles along any Pipeline having Uniform Electrical Characteristics .................................................................... Fig. 3-43 Effect of Electrical Length of Pipeline on AC Voltage Profile. Fig. 3-44 Double Vertical Circuit ........................................................... Fig. 3-45 Quadruple Vertical Circuit...................................................... Fig. 3-46 Single Delta Circuit ................................................................ CP Interference Course Manual © NACE International, 2006 June 2007 3:9 3:10 3:11 3:12 3:13 3:13 3:14 3:20 3:20 3:21 3:21 3:22 3:23 3:23 3:24 3:25 3:26 3:27 3:27 3:30 3:31 3:32 3:33 3:33 3:35 3:36 3:36 3:36 3:37 3:37 3:37 3:38 3:39 3:39 3:40 3:44 3:44 3:45 Fig. 3-47 Effect of Phase Conductor Separation .................................. Fig. 3-48 Phase Arrangements for a Double Vertical Circuit ................ Fig. 3-49 Effect of Phase Arrangement on LEF Magnitude for Variation of d/s Ratios (and for the specific case Where ρ/s2 = 1Ω/m, s/h=0.3, and I=1000A) ....................... Fig. 3-50 Simple Pipeline-Powerline Corridor (Plan View).................... Fig. 3-51 AC Voltage Profile Along an Electrically Short Pipeline (Uniform Conditions – No Grounding) ................................ Fig. 3-52 Electrical Service Analogy for Pipeline-Powerline Corridor In Figure 3-50 ..................................................................... Fig. 3-53 AC Voltage Profile Along an Electrically Short Pipeline (Non-Uniform Conditions – No Grounding)......................... Fig. 3-54 Effect of Grounding One End of Electrical Service Secondary........................................................................... Fig. 3-55 Effect of Grounding One End of Pipeline in Figure 3-50 ....... Fig. 3-56 Effect of Grounding Both Ends of Pipeline or Adding Distributed Grounds............................................................ Fig. 3-57 Effect of an Insulator at the Midpoint of the Pipeline ............. Fig. 3-58 AC Voltage Profile Along an Electrically Long or Lossy Pipeline (Uniform Conditions – No Grounding)................... Fig. 3-59 AC Voltage Profile Along an Electrically Long or Lossy Pipeline (Zero Resistance Ground at Distance = 0) ........... Fig. 3-60 Effect of an Insulator at the Midpoint of an Electrically Long Pipeline ...................................................................... Fig. 3-61 Fibrillating Current vs. Body Weight (Various animals – 3 second shock duration)....................................................... Fig. 3-62 Possible Body Current Paths................................................. Fig. 3-63 Example of Typical Touch and Step Potentials at an Energized Structure ............................................................ Fig. 3-64a Coefficient of Corrosion at Different Frequencies for Iron Electrode Denoted as Average Electrode Loss........... Fig. 3-64b Maximum Penetration Depth as a Function of Test Duration at Constant Cathode DC Current Density (2A/m2) and Differing AC Current Density .......................... Fig. 3-65a Effect of CP Potential on AC Corrosion Rate ........................ Fig. 3-65b Effect of CP Potential on AC Current Density........................ Fig. 3-65c Pit Cluster and Pinhole Perforation (Case History No. 1) ...... Fig. 3-65d Hemispherical Shell of Hardened Soil Surrounding Anomaly (Case History No. 3) ............................................ Fig. 3-65e Hemisphere of Hardened Soil and Corrosion Pit (Case History No. 3) ........................................................... Fig. 3-65f Pinhole Corrosion Failure Following Removal of Repair Clamp (Case History No. 4)................................................ Fig. 3-65g Pipeline-Powerline Route (Case History No. 4)..................... Fig. 3-65h Nodule of Corrosion Products Protruding Through Coating (Case History No. 4).............................................. Fig. 3-65i Corrosion Pit After Removal of Coating and Corrosion Products (Case History No. 4) ............................................ Fig. 3-65j Effects of Installing Ground Electrodes at Sites A and B (Case History No. 4) ........................................................... CP Interference Course Manual © NACE International, 2006 June 2007 3:46 3:46 3:47 3:49 3:49 3:50 3:50 3:51 3:52 3:52 3:53 3:54 3:55 3:56 3:61 3:63 3:64 3:68 3:71 3:73 3:74 3:76 3:80 3:81 3:82 3:83 3:85 3:85 3:88 Fig. 3-65k Effects of Installing Ground Electrodes on AC Current Densities (Case History No. 4) ........................................... Fig. 3-66 Fault Damage to CP Bond..................................................... Fig. 3-67 Field Estimation of LEF Magnitude Using Horizontal Wire Method ....................................................................... Fig. 3-68 Soil Resistivity Measurement Using the Wenner Four-Pin Method................................................................................ Fig. 3-69 Determination of Pipeline Coating Resistance ...................... Fig. 3-70 Determination of Pipeline Internal Impedance....................... Fig. 3-71 Sectionalization of Pipeline-Powerline Route ........................ Fig. 3-72 Pipeline-Powerline Geometry for Calculation of LEF............. Fig. 3-73 Typical Series of Curves for Determining LEF....................... Fig. 3-74 Simple Pipeline-Powerline Corridor (Plan View).................... Fig. 3-75 Simple Pipeline-Powerline Model .......................................... Fig. 3-76 Equivalent Circuit for Line-to-Ground Fault ........................... Fig. 3-77 Distribution of Fault Current Along Powerline........................ Fig. 3-78 Distribution of Fault Current Along Powerline........................ Fig. 3-79 Calculation of Earth Voltage at Pipe due to Faulted Tower ... Fig. 3-80 Approximate Length of Pipeline Affected by Faulted Tower.. Fig. 3-81 Resistance of Coating Holiday to Earth ................................. Fig. 3-82 Modified Resistance of Coating Holiday to Earth due to Localized Soil Ionization Effects ..................................... Fig. 3-83 AC Pipeline Voltages Induced by Overhead Faulted Powerline (Per 1000 A of Fault Current)............................. Fig. 3-84 Motor Operated Valve – Effects of Grounding on Induced AC and CP Currents ........................................................... Fig. 3-85 Electrical Isolation of Motor Operated Valve from Pipeline.... Fig. 3-86 Electrical Grounding Schematic of Motor Operated Valve Showing Two Alternative Locations for a DC Decoupling Device................................................................................. Fig. 3-87 Decoupling Device Installed by Electrical Utility Between Primary and Secondary Grounds ....................................... Fig. 3-88 Isolation-Surge Protector Installed across Isolating Flange... Fig. 3-89 Electrical Schematic of One Model of Solid-State DC Decoupling Device.............................................................. Fig. 3-90 DC Decoupling Device Installed Across Insulating Flange for Lightning Protection....................................................... Fig. 3-91 AC Current Being Measured Through a Polarization Cell ..... Fig. 3-92 Polarization Cell Construction ............................................... Fig. 3-93 Corrosion of Plates Within a Polarization Cell ....................... Fig. 3-94 Grounding Cell....................................................................... Fig. 3-95 Electrolytic Capacitor............................................................. Fig. 3-96 Failure of Electrolytic Capacitors in Stray Current Area ........ Fig. 3-97 Metal-Oxide Varistors (MOVs)............................................... Fig. 3-98 Explosion-Proof Surge Protection Device Installed Across Insulator .................................................................. Fig. 3-99 Test Station Varieties (left to right): a) Terminals Exposed To Public; b) Terminals Covered by a Plastic Cap (Locking or Non-Locking); c) Dead-Front Terminals; d) Aluminum Test Station with Padlocked Cover..................................... CP Interference Course Manual © NACE International, 2006 June 2007 3:89 3:95 3:98 3:99 3:103 3:104 3:107 3:109 3:110 3:111 3:111 3:115 3:116 3:117 3:118 3:119 3:120 3:121 3:123 3:127 3:128 3:129 3:130 3:131 3:131 3:132 3:133 3:133 3:134 3:135 3:135 3:136 3:137 3:138 3:139 Fig. 3-100 a) Zinc Ribbon Anode of Various Sizes; b) Zinc Ribbon Being Installed in Pipe Trench ............................................ Fig. 3-101 Effect of Gypsum on Restoration of Zinc Potential in Bicarbonate-Rich Soil ......................................................... Fig. 3-102 Potential of Magnesium Versus AC Current Density in a Fe-Mg Cell ................................................................... 3:140 3:140 3:141 Tables Table 3-1 Table 3-2 Table 3-3 Table 3-4 Table 3-5 Effects of 60 Hz AC Body Currents on Humans .................... Let-Go Currents from Dalziel’s Experiments ......................... Let-Go Currents from Dalziel’s Experiments ......................... Voltage Puncture Levels for Various Holiday-Free Coatings. Specific Leakage Resistances and Conductances................ CP Interference Course Manual © NACE International, 2006 June 2007 3:60 3:62 3:63 3:94 3:103 Chapter 4–Telluric Current Interference 4.1 Background Theory .............................................................. 4.1.1 4.1.2 4:1 Distributed Source Transmission Line Equations ............... Factors that Affect the Induced Electric Field ..................... (a) Solar Cycle Variations................................................ (b) Sun’s Rotational Frequency....................................... (c) Earth’s Rotation ......................................................... (d) Plasma Magnetic Field Direction ............................... (e) Proximity of Pipeline to a Sea Coast.......................... (f) Pipeline Latitude ........................................................ Factors that Affect the Pipeline Lineal Impedance (Z) and Shunt Admittance (Y).......................................................... (a) Effect of Coating Quality ............................................ (b) Effect of Isolating Fittings........................................... (c) Effect of Pipeline Directional Change ........................ 4:13 4:13 4:14 4:15 4.2 Measuring the Geomagnetic Intensity and Determining the Electric Field (E).............................................................. 4:16 4.3 Interference Effects of Telluric Current on Pipelines ............. 4:18 4.1.3 4.3.1 4.3.2 4.3.3 4.3.4 4.3.5 General Considerations ...................................................... Corrosion ............................................................................ (a) Theoretical Considerations ........................................ (b) Calculating the Corrosion Rate .................................. (c) Telluric Corrosion Case Studies on Cathodically Protected Piping...................................................... Impact on Accuracy of Current and Potential Measurements ................................................................................... Impact of Telluric Current on Pipeline Coatings ................. Impact on Output of a CP Rectifier ..................................... 4.4 Mitigating the Effects of Telluric Current ............................... 4.4.1 4:6 4:8 4:8 4:9 4:9 4:9 4:10 4:12 4:18 4:18 4:18 4:22 4:27 4:29 4:31 4:32 4:33 Mitigating Corrosion Impact ................................................ (a) Making the Pipeline Electrically Continuous and Grounded ................................................................ (b) Using CP.................................................................... (i) Sacrificial Anodes ................................................ (ii) Impressed Current Systems ................................ Compensating for Measurement Error Caused by ............ Telluric Current .......................................................... 4:33 4:34 4:35 4:39 4.5 Summary .............................................................................. 4:49 Summary of Equations.................................................................. 4:51 4.4.2 CP Interference Course Manual © NACE International, 2006 June 2007 4:33 4:42 Figures Fig. 4-1 Interaction of Solar Particles on the Earth’s Magnetic Field .. Fig. 4-2a Plasma Charge Distribution around the Earth during Quiescent Period ................................................................ Fig. 4-2b Plasma Charge Distribution around the Earth during a Magnetic Storm................................................................ Fig. 4-3 This Plot Shows the Current Extent and Position of the Auroral Oval in the Northern Hemisphere, Extrapolated From Measurements Taken During the Most Recent Polar Pass of the NOAA POES Satellite for September 16, 2004 at 14:22 UT .......................................................... Fig. 4-4 Schematic of Geomagnetic Induction Directly into a Pipeline and the Resulting Change in Pipeline Potential that is Produced ............................................................................ Fig. 4-5 Quiet Day Variation in the Geomagnetic Field and the Associated Change in the Electric Field and the Pipe-toSoil Potential....................................................................... Fig. 4-6 P/S Potential and Telluric Current in a Long Pipeline Exposed to an Induced Electric Field of 1 V/km, Having an Impedance of 0.1 Ω /km and an Admittance of 0.15 Ω /km................................................... Fig. 4-7 Equivalent Circuit for a Short Section of Pipeline .................. Fig. 4-8 History of Geomagnetic Effects on Ground Technology........ Fig. 4-9 Pipe-to-soil Potential Variations with Time ............................ Fig. 4-10 Charge Accumulation at the Coast Resulting from Larger Induced Currents in the Sea Compared to in the Land. The Charge Accumulation Increases the Electrical Potential of the Earth’s Surface Near the Coast ................. Fig. 4-11 Electric Field, E, Generated by Seawater Moving with Velocity, v, Through the Earth’s Magnetic Field, B ............. Fig. 4-12 Geomagnetic Hazard Percentage of Probability of Occurrence ......................................................................... Fig. 4-13 Telluric Induced Voltage Profile vs Distance for a Pipeline with Different Attenuation Constants..................... Fig. 4-14 Calculated Telluric Induced Voltage at the End of a Long Pipeline as a Function of Coating Conductance for an East-West Electric Field of 0.1V/km .............................. Fig. 4-15 Effect of Isolating Fittings on the Telluric Induced Voltage Profile on an Electrically Short Pipeline .............................. Fig. 4-16 Effect of Pipeline Directional Change on the Telluric Induced Voltage.................................................................. Fig. 4-17 Average Occurrence of 3-Hour Intervals with the Magnetic Activity Index Kp Equal to or Greater than a Specified Value. Kp=9 Corresponds to a Severe Magnetic Storm .... Fig. 4-18 Peak Electric Field Magnitudes as a Function of Kp ............. Fig. 4-19 Oxidation Reaction at Pipe Surface During Telluric Current Discharge in the Absence of CP............................ Fig. 4-20 Reduction Reactions During Negative Cycle Telluric CP Interference Course Manual © NACE International, 2006 June 2007 4:1 4:2 4:2 4:3 4:3 4:4 4:5 4:6 4:8 4:9 4:10 4:11 4:12 4:13 4:14 4:14 4:15 4:16 4:17 4:18 and CP Current Pick-up...................................................... 4:19 Fig. 4-21 Steel Surface pH versus Applied CP Current Density ........... 4:20 Fig. 4-22 Polarization Curves after Several Days of Potentiostatic Polarization ......................................................................... 4:20 Fig. 4-23 Experimental Anodic Polarization Curve of Steel in Hydroxide (pH 12.0)............................................................ 4:21 Fig. 4-24 Telluric Current Discharge from a Cathodically Protected Pipe .................................................................................... 4:22 Fig. 4-25 Coefficient of Corrosion at Different Frequencies for Iron Electrodes Denoted as Average Electrode Loss ......... 4:23 Fig. 4-26 Effect on Corrosion Rate of Reversing Direction of Current Compared to Steady State DC and Length of Time Between Reversals............................................................. 4:24 Fig. 4-27 Corrosion Current Density at a Coating Defect having an Applied Voltage of 1.0V in 1000 Ω-cm Soil for Various Coating Thicknesses ............................................. 4:25 Fig. 4-28 Chart Showing the Influence of Anodic Transient Time with Respect to Corrosion Experienced by Probe in Sandy and Clay Soil. Line (a) Represents the Corrosion Rate Expected from Faraday’s Law for the Clay Soil, and Line (B) for the Sandy Soil, Respectively ........................... 4:26 Fig. 4-29 Corrosion Pit at 112+307 (60 mils/497mils 07:30)................. 4:28 Fig. 4-30 Magnetic Field Intensity and Pipe-to-Soil Potential Superimposed..................................................................... 4:29 Fig. 4-31 Schematic of Potentially Controlled CP System Used to Mitigate Telluric Current Effects ....................................................... 4:30 Fig. 4-32 Current Flow and Calculated OFF Potentials during a GIC Incident........................................................................ 4:31 Fig. 4-33 Telluric Current Through a Bridge Rectifying Element During a Discharge Cycle ................................................... 4:32 Fig. 4-34 Schematic of a Telluric Bond Switch ..................................... 4:34 Fig. 4-35 Mitigation of Telluric Current Discharge Effects Using Galvanic Anodes................................................................. 4:35 Fig. 4-36 Effect of Connecting and Disconnecting Groups of Galvanic Anodes to a Pipeline Subjected to Telluric Current................................................................................ 4:36 Fig. 4-37 Maritimes DSTL Results Without Flanges ............................. 4:38 Fig. 4-38 Electrical Schematic at a Constant Voltage Transformer Rectifier During a Positive Telluric Voltage Fluctuation ...... 4:40 Fig. 4-39 Pipe Potential and Rectifier Current Output vs Time for An Impressed Current System Operating in Potential Control ................................................................................ 4:41 Fig. 4-40 Typical Pipe-to-Soil Potential Measurements at Test Station Having a Steel Coupon and Soil Tube ................... 4:42 Fig. 4-41 Typical Pipe-to-Soil Potential Recording at a Test Station Using a Coupon/Reference Probe.......................... 4:43 Fig. 4-42 Comparison Between Pipe/Coupon Potential with Time Recorded with Respect to a Copper-Copper Sulfate Reference on Grade and to a Coupon/Reference Probe Located at Pipe Depth ........................................................ 4:44 CP Interference Course Manual © NACE International, 2006 June 2007 Fig. 4-43 Pipe-to-Soil Potential Measurement Method to Compensate For Telluric Current Effects During a Close Interval CP Survey........................................................................... Fig. 4-44 CIPS Method Using One Moving and Two Stationary Data Loggers ...................................................................... Fig. 4-45 Pipe-to-Soil Potential Measurement Method to Compensate for Telluric Current Effects During a Close Interval CP Survey........................................................................... Fig. 4-46 Pipe Potential/Telluric Current Relationship at a Coupon Test Station......................................................................... Fig. 4-47 Four Wire Test Lead Arrangement for Measuring Pipe Current........................................................................ Appendices Appendix A – Curve Matching Appendix B – Pipe Data Table Appendix C – Anode Tables Appendix D – Wire Size Table Appendix E – Metric Conversion Table Appendix F – Dabkowski Paper NACE RP0177 NACE SP0169 NACE Glossary of Corrosion-related Terms Course Evaluation Instructor Evaluation CP Interference Course Manual © NACE International, 2006 June 2007 4:45 4:46 4:47 4:48 4:49 CHAPTER 1 STRAY CURRENT INTERFERENCE 1.1 Historical Background The term “interference” is understood in the pipeline industry as electrical interference and is defined as “any detectable electrical disturbance on a structure caused by a stray current where a ‘stray current’ is defined as a current in an unintended path”.1 This broad definition suggests that the structure, although often a pipeline, could be any metallic network such as electrical power grids and communication systems. Furthermore, although the interfering current is often a direct current (DC) from a cathodic protection (CP) impressed current source, the current can also originate from any electrical system that uses the earth either intentionally or inadvertently as a current path. Thus alternating current (AC) can also be included in the definition. Electrical interference concerns preceded the use of CP for corrosion control of pipelines. Telegraph systems were reported2 to interfere with the operation of the early telephone systems. Lighting systems, first introduced in about 1880, comprised arcs and incandescent lamps also interfered with the telephone systems, primarily because both the telephone system and the lighting systems used the earth as a current path. Then, in the late 1800s and early 1900s, street railways throughout North America were electrified.3 They ultimately led to the corrosion of cast iron watermains. Figure 1-1: Early Electric Trolley (courtesy of East Bay Municipal Utility District, Oakland, CA)4 1 CP3 – Cathodic Protection Technologist Course, NACE International, June 1, 2004, p.3-1. Anderson, John M., The Fight Over the Highways, IEEE Power Engineering Review, December 1997, p.45. 3 Anderson, John M., First Electric Street Car, IEEE Power Engineering Review, Oct. 1999, p.32. 4 Lewis, Mark, Once Vagrant Current, Now Impressed Current Cathodic Protection, MP, Vol 36, July1997. 2 CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:2 Corrosion on watermains as a result of interference from a DC transit system was first reported by Stone & Forbes in 18945, just 6 years after a New England transit system began operation. In 1901, damage to water and gas mains in Toronto, Ontario, was reported6 as being due “to railway currents.” The currents reportedly affected the watermains for two reasons: deterioration of the rail joint bonds and the practice of bonding the watermains to the rails at certain locations. The U.S. Bureau of Standards began studying the stray current traction problem in 1910. The bureau would issue 15 reports by 1921. Many of the investigations involved field studies, during which temporary electrolysis committees were formed consisting of interested utility representatives. The corrosion resulting from stray current was initially referred to as “electrolysis,” a term defined as “the decomposition of a substance by the application of a current”.7 The widespread corrosion of iron watermains by stray transit system currents led to the formation in 1913 of the American Committee on Electrolysis.8 Stray current activity on underground structures arising from transit system operation is not steady-state but dynamic in terms of current and potential amplitude. It often reverses direction. Typical structure potential activity was recorded on smoked charts. These charts collect data as a stylus moving in response to a changing potential input removes the smoke from the chart, which is rotated by a clock drive. The dynamic nature of the stray current effect on pipe potential is shown in Figure 1-2. 5 Stone, C.A. and Forbes, H.C., Electrolysis of Water Pipes, New England Water Works Association, Vol. 9, pp.1894-5. 6 Knudson, A.A., Report on the Joint Investigation and Survey for Electrolysis on the Water and Gas Mains in the City of Toronto, Ontario, July 1, 1906. 7 The Oxford Encyclopedic English Dictionary, Oxford University Press, 1991. 8 Meany, J.J., A History of Stray Traction Current Corrosion in the United States, NACE, Corrosion’74, Paper 152, p.3. CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:3 Figure 1-2: Pipe-to-Soil Potential Changes due to Transit System Stray Current Activity were Recorded on Smoked Charts Because of the variable nature of the stray current activity, it is difficult to predict how much corrosion would occur. The Bureau of Standards conducted a study9 in which iron samples where subjected to AC discharge and current pick-up for different periods of time. The resulting corrosion was compared to corrosion produced by a steady-state DC of the same current density and discharge period. The results of this study, reported in 1916, are summarized in Figure 1-3. 9 McCollum, B. and Ahlborn, G.H., Influence of Frequency of Alternating and Infrequently Reversed Current on Electrolytic Corrosion, Technologic Papers of the Bureau of Standards, U.S. Dept. of Commerce, No. 72, 1916. CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:4 100 90 80 LEGEND: Soil Soil + Na2CO3 70 60 50 40 30 20 10 0 -10 1/60S 1/15S 1S 5S 1M 5M 10M 1Hr. 2Days 2Weeks D.C. Logarithm of Length of Time of One Cycle Figure 1-3: Coefficient of Corrosion at Different Frequencies for Iron Electrode Denoted as Average Electrode Loss For short periods of reversals, the corrosion was only a small fraction of the corrosion at steady state. For equal periods of pick-up and discharge, the corrosion coefficient remained below 20% when the cycle remained below one hour. This meant that the corrosion occurring from dynamic stray currents was a function of the frequency. At 60hz the corrosion rate was less than approximately 2% of the steady state value. R.J. Kuhn, who investigated the effects of transit system stray current activity on iron water mains in New Orleans, Louisiana, is credited with the discovery of CP. It occurred to him in 1928 that “ordinary corrosion could be prevented by reversing these currents”.10 Sir Humphrey Davy11 was the first person on record to use CP by applying zinc castings to protect the copper sheathing on British warships in 1824. Although a technical success, Davy’s application was a 10 Kuhn, R.J., Cathodic Protection of Underground Pipe lines from Soil Corrosion, API Proceedings, Nov. 1933, Vol. 14, p.164. 11 Davy, H., Philosophical Transactions of the Royal Society, London, 1824-1825. CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:5 practical failure because the copper biofouled when the corrosion was stopped— thus reducing the speed of these sailing ships. It appears that neither Kuhn nor any other corrosion practitioner had knowledge of this. Hence, Kuhn is considered by one source12 as the “father” of CP (certainly as it applies to pipelines). Against this backdrop, stray current interference and its corrosion consequences for underground metallic structures were first evaluated. Today electrolysis committees exist throughout North America, and methods of mitigation that have subsequently been developed are commonly utilized. Sources of stray current interference are not confined to DC transit systems. They now include any electrical source that uses the earth either intentionally or inadvertently as a current path. This course addresses these sources and the mitigation methods that have been developed to mitigate not only the corrosion effects, but other deleterious consequences of stray current activity. 1.2 Typical Stray Current Circuit Arising from a Transit System Operation Figure 1-4 depicts stray current paths originating from the operation of an electric transit system. Although it is the intent that the DC operating current returns to the substation via the running rails (IR), some of the load current (IL) will pass through the earth (Ie) if the rail is in electrolytic contact with the earth. If there is a metallic structure in the earth, it, too, will carry some of the load current (IS). Therefore, the load current (IL)—after passing through the locomotive—divides into parallel paths. The amount of current in each path is inversely proportional to the resistance of each path relative to the total circuit resistance, as Equation 1-1 indicates. I path = where: 12 Ipath RT Rpath IL = = = = R T • IL R path current in a path total resistance of parallel paths resistance of current path load current von Baeckmann, W., Schwenk, W., and Prinz, W., Handbook of Cathodic Corrosion Protection, 3rd edition, Gulf Publishing Co., Houston, TX, 1997, p.16. CP Interference Course Manual © NACE International, 2006 January 2008 [1-1] Stray Current Interference 1:6 DC substation O/H power conductor IL IR ground Is running rails Is Is p ic k - u p metallic structure (e.g.,watermain) d is c h a r g e Ie Ie Figure 1-4: Typical Stray Current Paths Around a DC Transit System Hence, as the resistance of the rail path increases or the resistance of the alternative stray current path(s) decreases, a greater percentage of the load current will appear in the stray current path(s). 1.3 Stray Current Charge Transfer Reactions on a Metallic Structure Figure 1-5 illustrates the typical stray current situation on an underground metallic structure that is not electrically connected to the source of stray current. The stray current pattern consists of a pick-up of stray current from the earth at one or more locations and the subsequent discharge of stray current to the earth at one or more locations. Is Is stray current pick-up Is stray current discharge Is Figure 1-5: Typical Stray Current Interference on a Metallic Underground Structure CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:7 The principal charge carriers in the earth are ions. They are electrons in the metallic structure. For these reasons, electrochemical reactions must transfer the charge between the structure and earth at both the pick-up and discharge locations. At the pick-up location(s), it is through reduction reactions that the electrical charges are transferred. Depending on the nature of the electrolytic environment, the reduction reactions can be one or more of the following: H3O+ + e– Æ HO + H2O [a] O2 + 2H2O + 4 e– Æ 4OH– [b] 2H2O + 2e– Æ H2↑ + 2OH– [c] Reaction [b] is favored in well-aerated soils and waters; reduction reaction [a] is favored in acidic soils or waters. Reduction reaction [c], which involves the breakdown of water molecules to hydrogen gas and hydroxyl ions, can occur under all conditions if there is sufficient over-voltage applied. At the discharge location, one or more of the following oxidation reactions transfers the electrical charge. M0 Æ Mn+ + ne– [d] 4OH– Æ O2 + 2H2O + 4e– [e] 2H2O Æ O2 + 4H+ + [f] 4e– Reaction [d] tends to occur on most basic metals such as iron, copper, zinc, and aluminum when the electrolyte has an acid or neutral pH. Reaction [e] is more likely in electrolytes with a high pH. Reaction [f] is more likely to occur when the over-voltage reaches the oxygen line. The oxygen line is line “b” on the Pourbaix diagram for iron (Figure 1-6). CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference -2 1:8 0 2 4 6 8 10 12 14 2 2 1.6 1.6 1.2 b 1.2 0.8 0.4 0 current pick-up a 0.8 current discharge 0.4 3 passivation 0 corrosion -0.4 -0.4 -0.8 1 -1.2 immunity -0.8 2 corrosion -1.2 -1.6 -1.6 -2 16 0 2 4 6 8 10 12 14 16 pH (assuming passivation by a film of Fe2O3) Figure 1-6: Simplified pH Pourbaix Diagram for Iron in Water at 25ºC Showing Potential Shift Direction for Current Pick-up and Discharge at Low pH The Pourbaix diagram for iron in pure water represents three zones of thermodynamic stability: corrosion, immunity, and passivity based on a potential (SHE) vs pH relationship. Line (a) is the hydrogen line and line (b) is the oxygen line. Water is stable between these two lines. If the potential of iron is shifted to either of these lines, then oxygen is generated at line (b) and hydrogen gas at line (a). For an iron structure without CP that is exposed to a neutral or low-pH water, a current pick-up will cause the potential to shift in the negative direction toward the immunity zone and afford the structure some CP. Conversely, at the discharge location, the potential is shifted in the electropositive direction into the passive region if not at a low pH—where it would otherwise remain in the corrosion zone. On a cathodically protected structure as illustrated in Figure 1-7, where the electrolyte at the iron surface normally has a high pH, a current discharge resulting in a positive shift can produce a passive film given by the following reaction: Fe + 2H2O Æ Fe(OH)2 + 2H+ + 2e– CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference -2 1:9 0 2 4 6 8 10 12 14 2 2 1.6 1.6 1.2 b 1.2 0.8 0.8 0.4 0 0.4 3 a passivation 0 corrosion current discharge -0.4 -0.8 1 -1.2 immunity corrosion current pick-up 0 2 4 6 8 -0.4 -0.8 2 -1.2 -1.6 -1.6 -2 16 10 12 14 16 pH (assuming passivation by a film of Fe2O3) Figure 1-7: Simplified pH Pourbaix Diagram for Iron in Water at 25ºC Showing Potential Shift Direction for Current Pick-up and Discharge at High pH The ferrous hydroxide formed is relatively stable at high pH. Because this reaction also produces hydrogen ions, the pH will decrease with time. 1.4 Effects of Stray Current on Metallic Structures It is apparent that the effect of a stray current pick-up and a stray current discharge from an iron structure from a thermodynamic perspective can cause corrosion, passivation, or immunity, depending upon the direction of current and the pH of the aqueous electrolyte at the charge transfer location. 1.4.1 At the Current Discharge Location Identification of the current discharge site receives considerable attention in stray current investigations because it is the location where corrosion damage is most likely to occur on all metallic structures. When a current transfers from a metallic structure to earth (Figure 1-8), it must do so via an oxidation reaction that converts electronic current to ionic current. CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:10 metal structure (electrons) Is O X I D A T I O N Is earth (ions) Is Figure 1-8: Current Discharge from a Metal Structure to Earth via an Oxidation Reaction The generic oxidation reaction is the corrosion of the metal as in Equation 1-2. Mo Æ Mn+ + ne– [1-2] For steel, the oxidation reaction is: Feo Æ Fe++ + 2e– [1-3] A stray current discharge from a metallic structure may not cause corrosion attack if the structure is receiving CP (Figure 1-9). Whether the superposition of a stray current discharge and a CP current pick-up at a metal/electrolyte interface causes corrosion will depend on time and the relative magnitudes of these two currents. metal structure O X I D A T I O N R E D U C T I O N Is Is Is Icp Icp earth Icp Icp Figure 1-9: Superposition of a Stray Current and a Cathodic Protection Current at a Metal/Electrolyte Interface CP current transfers across the metal/earth interface via a reduction reaction, which produces hydroxyl ions in either of the three following reactions: H3O+ + e– Æ HO + H2O [1-4] O2 + 2H2O + 4e– Æ 4OH– [1-5] 2H2O + 2e– Æ H2Ç + 2OH– [1-6] In the presence of a high concentration of hydroxyl ions, a possible oxidation reaction is given in Equation 1-7. The reaction involves the oxidation of hydroxyl ions to oxygen and water. 4OH– Æ O2 + 2H2O + 4e– [1-7] CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:11 This latter reaction does not consume metal atoms; therefore, there is no corrosion damage. Hence, as long as the polarized potential at the structure electrolyte interface is not driven more electropositive than the CP criterion (e.g., –850mVcse for iron or steel), significant corrosion would not be expected. If the metal has a surface passive film or is a relatively inert material (such as some of the materials used for impressed current anodes), then not all of the stray current need transfer through a corrosion reaction. If the stray current polarizes the metal surface electropositively to the oxygen line on the Pourbaix diagram, then the hydrolysis[13] of water molecules by the following reaction 1-8 is likely. 2H2O Æ 4H+ + O2Ç + 4e– [1-8] This oxidation reaction does not result in the consumption of the metal surface, but it does produce an acidic pH from the generation of hydrogen ions. On an iron or steel structure without CP, the oxidation reaction is usually the dissolution of the metal according to Equation 1-9 Feo Æ Fe++ + 2e– [1-9] The severity of corrosion depends on the magnitude of the stray current and time as related by Faraday’s Law: Wt = M t I corr nF [1-10] where: Wt = total weight loss at anode or weight of material produced at the cathode (g) n = number of charges transferred in the oxidation or reduction reaction Icorr = the corrosion current (A) F = Faraday’s constant of approximately 96,500 coulombs per equivalent weight of material (where equivalent weight = M ) n M = the atomic weight of the metal that is corroding or the substance being produced at the cathode (g) t = the total time in which the corrosion cell has operated (s) 13 Hydrolysis is defined as a double decomposition reaction involving the splitting of water into its ions and the formation of a weak acid or base or both. CRC Handbook of Chemistry and Physics, CRC Press, 53rd Edition, 1972-1973, PF-83. CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:12 Given the atomic weight of pure iron as 55.85 g and assuming 100% efficiency and pure DC, the consumption rate of iron as illustrated in Table 1-1 is 9.13 kg/A-y. Table 1-1: Theoretical Consumption Rates of Various Metals and Substances Reduced Species Oxidized Species Al Cd Be Ca Cr Cu H2 Fe Pb Mg Ni OHZn Al+++ Cd++ Be++ Ca++ Cr+++ Cu++ H+ Fe++ Pb++ Mg++ Ni++ O2 Zn++ Molecular Weight, M (g) 26.98 112.4 9.01 40.08 52.00 63.54 2.00 55.85 207.19 24.31 58.71 32.00 65.37 Electrons Transferred (n) 3 2 2 2 3 2 2 2 2 2 2 4 2 Equivalent Weight, M/n (g) 8.99 56.2 4.51 20.04 17.3 31.77 1.00 27.93 103.6 12.16 29.36 8.00 32.69 Theoretical Consumption Rate (Kg/A-y) 2.94 18.4 1.47 6.55 5.65 10.38 0.33 9.13 33.9 3.97 9.59 2.61 10.7 On pipelines, the total weight loss is usually less important than the penetration rate. By re-arranging Faraday’s Law, the weight loss per unit time per unit area is shown to be directly proportional to current density (i = I/A) as in Equation 1-11. Wt A tt = M i nF [1-11] Dividing this equation by the density (d) of the metal or alloy produces the corrosion rate (rcorr), which can be expressed in mm/y (Equation 1-12). rcorr = k M is nF d where: M n i k d rcorr = = = = = = CP Interference Course Manual © NACE International, 2006 January 2008 atomic weight (g) number of charges transferred in corrosion reaction current density (μA/cm2) unit correction term ≈ 3.156 x 108 mm s/cm yr density (g/cm3) penetration rate in (mm/yr) [1-12] Stray Current Interference 1:13 Example: Using Equation 1-12 to calculate the penetration rate based on a current density of 1 A/m2 (10-4 A/cm2): where: i = 10-4 A/cm2 d = 7.87 g/cm3 M = 55.85 g n = 2 F = 96,500 coulombs then: rcorr = 3.156 × 10 8 mm s/ cm yr × 55.85g × 10 -4 A/cm 2 2 × 96,500 coulombs × 7.87 g/cm 3 = 1.16 mm/y Table 1-2 gives the penetration rate, in mpy and 10-3 mm/y, equivalent to a current density of 1μA/cm2 for a number of common pure metals. Table 1-2: Electrochemical and Current Density Equivalence with Corrosion Rate for Some Common Pure Metals Metal/Alloy Pure Metals Iron Nickel Copper Aluminum Lead Zinc Tin Titanium Zirconium Element/ Oxidation State Density (g/cm3) Equivalent Weight (g) Fe/2 Ni/2 Cu/2 Al/3 Pb/2 Zn/2 Sn/2 Ti/2 Zr/4 7.87 8.90 8.96 2.70 11.4 7.13 7.3 4.51 6.5 27.93 29.36 31.77 8.99 103.6 32.69 59.34 23.95 22.80 Penetration Rate Equivalent to 1 μA/cm2[1] (mpy) 10-3 mm/y[2] 0.46 0.43 0.46 0.43 1.17 0.59 1.05 0.69 0.45 11.6 10.8 11.6 10.9 29.7 15.0 26.6 17.4 11.5 Note: [1] A current density of 1 μA/cm2 is approximately = 1 mA/ft2 [2] 10-3 mm/y = 1 μm/y and 1 mpy = 25.4 μm/y The foregoing corrosion rates apply to stray current situations involving a continuous DC discharge. Corrosion rates decrease for periodic reversals of DC and are substantially less for 60Hz AC (Figure 1-3). CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:14 The low corrosion rate for a 60Hz current is attributed to the relatively low impedance of the interfacial capacitance. The structure/electrolyte interface can be modeled electrically by a Randle’s Circuit shown in Figure 1-10. where: Cdl = double layer capacitance (1-200 μF/cm2) Rp = polarization resistance (1-104 Ω-cm2) Eac Rp Iac,rp Cdl steel Re Re = resistance of steel surface to remote earth potential difference (volts) Eoc = Iac Iac = total AC crossing the interface soil (electrolyte) Ia,rp = total AC through polarization resistance Ia,dl = total AC through double-layer capacitance Iac,dl Figure 1-10: Randle’s Electrical Circuit Model of a Metal/Electrolyte Interface This circuit model illustrates that the interface is not simply a resistance but a parallel combination of the polarization resistance (Rp) and a capacitor (Cdl) called the double-layer capacitance. Unlike DC, AC can pass through the doublelayer capacitance. There is no mass transfer in this current path and hence no corrosion polarization results from current transfer in this path. The proportion of AC (Iac,dl) through the double-layer capacitor is a function of the relative impedance of this path compared to the polarization resistance. The reactance (Xcdl) of the double-layer path is given by the following equation: Xc dl = 1 2π f C dl [1-13] where: f = frequency (Hz) Cdl = capacitance (farads) Xcdl = reactance (ohms) Assuming a 1cm2 surface area and mid-range values of both the polarization resistance and the double-layer capacitance as follows, then the proportion of AC through the capacitor can be calculated. CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference Assume: 1:15 Cdl = 100 μf/cm2 Rp = 103 Ω−cm2 Using Equation 1-13: Xc dl = 1 2π 60 × 100 × 10 -6 = 10 4 376.8 = 10 4 120π = 26.54 Ω The total impedance Zt to 60Hz AC of the parallel combination of the polarization resistance (Rp) and the double-layer capacitance is therefore: therefore: 1 Zt = 1 + Rp 1 Zt = 10 -3 + 37.7 × 10 −3 = 38.7 × 10 −3 Zt = 10 3 38.7 1 Xc dl = 1 + 10 3 1 26.54 = 25.8 Ω Then the proportion of AC current through the double-layer capacitance is: I ac,dl I ac,dl = = Z t I ac, t Xc dl 25.8Ω × I ac, t = 0.974 I ac, t or 97.4% 26.5 Accordingly, only approximately 2.6% of the AC would pass through the polarization resistance and only the positive half-cycle of the current would be involved in the corrosion reaction. 1.4.2 At Area of Current Pick-Up At the area of current pick-up, a negative shift will result in cathodic polarization. If the foreign structure is mild steel, then there is a beneficial effect because the structure is receiving some measure of CP. If the structure is coated and has its CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:16 own CP system, the additional polarization from the stray current pick-up may result in cathodic blistering of the coating. If the foreign structure is not mild steel but is made of an amphoteric metal such as aluminum, lead, or zinc, then the high pH developed at the structure/earth interface caused by the reduction reaction can effect “cathodic” corrosion. Amphoteric metals such as aluminum are susceptible to corrosion at both high and low pH. Figure 1-11 shows this phenomenon for aluminum. (a) Aluminum (b) Lead Figure 1-11: Theoretical Conditions of Corrosion, Immunity, and Passivation of (a) Aluminum at 25ºC and (b) Lead at 25º C Source: Pourbaix, M., Atlas of Electrochemical Equilibria in Aqueous Solutions, National Association of Corrosion Engineers, Houston, TX, 1974, p.172 CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:17 70 Al 60 Zn 50 40 30 20 10 14 13 12 Alkaline 11 10 9 8 7 pH 6 5 4 3 2 1 Acid Figure 1-12: Comparison of Zn and Al Coatings for Corrosion Resistance as Functions of pH One can see that aluminum is particularly sensitive to high pH attack. Aluminum is often used underground for water irrigation systems, gas distribution piping in rural areas, AC secondary distribution conductors, and the sheathing on communication cables. Zinc and lead are also amphoteric metals. The corrosion rate of zinc, as indicated in Figure 1-12, is not as high as aluminum in alkaline conditions but is much greater in acid conditions. Lead sheathing was commonly used on belowground AC power cables. Not only are these amphoteric materials susceptible to corrosion according to Faraday’s Law at rates indicated in Table 1-2 at stray current discharge locations, but also at stray current pick-up locations. Prestressed concrete cylinder pipe (PCCP) used for both water and sewage transmission is composed of a mild steel inner cylinder, over which a highly stressed steel wire is wound to give the concrete/steel cylinder strength. Typical cross-sections of the two types of PCCP are shown in Figures 1-13a and 1-13b. CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:18 Prestressing Wire and Wire Fabric Around Bell or Thicker Bell Ring and Wire Fabric Cement - Mortar Coating Grout Joint After Installation Prestressed Wire Steel Cylinder Concrete Core Rubber Gasket Steel Bell Ring Cement Mortar Placed in Field or Other Protection Steel Spigot Ring a. Lined Cylinder Pipe Cement - Mortar Coating Grout Joint After Installation Concrete Core Steel Spigot Ring Prestressed Wire Rubber Gasket Cement Mortar Placed in Field or Other Protection Steel Cylinder Steel Bell Ring b. Embedded Cylinder Pipe Figure 1-13: Typical Section Through a Joint in Two Types of PCCP Source: Prestressed Concrete Pressure Pipe-Steel Cylinder Type for Water and Other Liquids, AWWA Standard C301, American Water Works Association, Denver, CO The prestressing wire in these pipes is normally cold drawn steel with a yield strength in the order of 200 ksi. The cold-worked hardened surface of the wire makes it susceptible to hydrogen embrittlement. It is recommended that the polarized potential be limited to –970 mVcse or less negative to minimize the production of atomic hydrogen. If a stray current causes excessive cathodic polarization, then a catastrophic failure could occur. CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:19 If the foreign structure is coated at the stray current pick-up site, then coating blistering or disbondment can occur. Coating blistering is caused by the pressure build-up beneath the coating due to the movement of water through the coating, due to electroendosmosis. Electroendosmosis is defined as “the inward flow of a fluid through a permeable membrane due to an electric field”. The high pH produced by the reduction reaction at the metal surface can attack the coating adhesion bonds or a surface oxide layer, resulting in coating disbondment. DC H2O _ _OH OH _ OH _ _ OH OH _ _ _ _ OH OH OH OH H2O soil metal substrate Figure 1-14: Cathodic Blistering/Disbondment of Protective Coating 1.4.3 Along the Structure Stray current in a metallic structure does not usually cause damaging effects between the stray current pick-up and discharge locations unless the current is very large or the structure is not electrically continuous. If the structure is electrically discontinuous (as is often the case with cast iron water distribution piping or PCCP transmission piping), the structure resistance (Rs) is greater than if it were electrically continuous, which reduces the magnitude of Is, but creates a current discharge/current pick-up pattern at each electrical discontinuity (Figures 1-15a and 1-15b). electrically discontinuous joints Is Is Is Figure 1-15a: Stray Current Discharge and Pick-Up Around an Electrically Discontinuous Joint Through the Earth In many of these structures not every joint is discontinuous, but localized corrosion will occur on the discharge side of the discontinuous joints. Furthermore, on water and sewer piping, there is not only a soil path for the stray CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:20 current but also an internal path through the aqueous medium as illustrated in Figure 1-15b. rubber seal aqueous medium Figure 1-15b: Stray Current Discharge and Pick-Up Through the Internal Aqueous Medium Around an Electrically Discontinuous Bell and Spigot Joint on Cast Iron Piping Current in an AC distribution system can also affect the transformation characteristics in distribution transformers. At the AC distribution transformer, which supplies the AC service for an impressed current transformer-rectifier, a ground cable is normally run from the AC neutral to a ground rod at the base of the service pole. The ground rod, being relatively close to the groundbed, will pick up stray current. The distribution neutral and the AC phase conductor will carry the stray current to ground at remote transformers because DC does not encounter a high resistance through the primary winding. This circuit is illustrated schematically in Figure 1-16. Remote Distribution Transformer CP AC Distribution Transformer L1 Is,2 Is,2 N N T/R N L2 L Is,1 Is CP groundbed Is Figure 1-16: Stray Current Circuit in an AC Electrical Distribution System A DC in the primary or secondary windings of a transformer will produce a magnetic flux in the transformer core that will tend to saturate the core and thus spoil its voltage transformation properties. This is a deleterious effect that is in CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:21 addition to the corrosion damage that results from the stray current discharging off the ground rod at the remote distribution transformer. 1.5 Summary Stray current is an irrevocable factor to which all metallic underground structures are exposed because so many electrical systems use the earth as a current path. The following list of possible stray current sources is extensive: • • • • • • • • • CP systems High-voltage AC transmission systems Low-voltage AC distribution systems High-voltage DC transmission systems AC and DC transit systems Welding operations Geomagnetically induced currents Low-frequency communication systems Land-line telephone systems. Pipeline corrosion control practitioners are often acutely aware of the various sources of stray current, yet impressed current CP systems remain among the most prevalent stray current sources. As public pressure mounts to force more stray current sources into joint-use corridors, stray current control becomes increasingly important and decidedly more complex. CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:22 Summary of Equations I path = [1-1] where: Ipath RT Rpath IL = = = = R T • IL R path page 1:5 current in a path total resistance of parallel paths resistance of current path load current [1-2] Mo Æ Mn+ + ne– page 1:10 [1-3] Feo Æ Fe++ + 2e– page 1:10 [1-4] H3O+ + e– Æ HO + H2O page 1:10 [1-5] O2 + 2H2O + 4e– Æ 4OH– page 1:10 [1-6] 2H2O + 2e– Æ H2Ç + 2OH– page 1:10 [1-7] 4OH– Æ O2 + 2H2O + 4e– page 1:10 [1-8] 2H2O Æ 4H+ + O2Ç + 4e– page 1:11 Feo [1-9] Æ Wt = [1-10] Fe++ + 2e– M t I corr nF page 1:11 page 1:11 where: Wt = total weight loss at anode or weight of material produced at the cathode (g) n = number of charges transferred in the oxidation or reduction reaction Icorr = the corrosion current (A) F = Faraday’s constant of approximately 96,500 coulombs per equivalent weight of material (where equivalent weight = M ) n M = the atomic weight of the metal that is corroding or the substance being produced at the cathode (g) t = the total time in which the corrosion cell has operated (s) CP Interference Course Manual © NACE International, 2006 January 2008 Stray Current Interference 1:23 Wt A tt [1-11] M i nF = page 1:12 rcorr [1-12] = kMi nF d page 1:12 where: M n i k d rcorr [1-13] Z dl = = = = = = = atomic weight (g) number of charges transferred in corrosion reaction corrosion current density (μA/cm2) unit correction factor ≈ 3.156 x 108 mm s/cm y density (g/cm3) penetration rate in (mm/yr) 1 2π f C dl where: f = frequency (Hz) Cdl = capacitance (farads) Zdl = impedance (ohms) CP Interference Course Manual © NACE International, 2006 January 2008 page 1:14 CHAPTER 2 DC INTERFERENCE 2.1 Introduction The term “interference” in cathodic protection (CP) parlance means electrical interference as opposed to physical or chemical interference. Hence interference can be defined as any detectable electrical disturbance on a structure caused by a stray current. In turn, a stray current is defined as a current in an unintended path. Many electrical systems rely on the earth as a conducting medium either for the main transmission of electrical energy (as with CP systems) or as an electrical ground. Still, other systems—such as electrified transit systems—may not be adequately isolated from ground. Regardless, any electrical system that is in contact with the earth is a possible source of stray currents. As illustrated in Figure 2-1, a current entering the earth at point “A” has many parallel paths available at point “B.” In I4 A I3 I2 I1 Rn R4 R3 B R2 R1 Figure 2-1: Parallel Current Paths in the Earth The amount of current in each path is inversely proportional to the resistance of each path. It can therefore be argued that current will take all available paths. If point “A” is considered an impressed current groundbed connected to the positive terminal of a transformer-rectifier and point “B” is a pipeline connected to the negative terminal, then the parallel current paths may all have similar resistances—in which case all the currents are the same. This is only possible in homogeneous soil where points “A” and “B” are a long distance apart and where the pipe has no lineal resistance. However, if the soil resistivity varies or the pipe has lineal resistance, the current paths will have unequal resistances as illustrated in Figure 2-2. CP Interference Course Manual © NACE International, 2006 January 2008 DC Interference 2:2 R4,e R1,e R3,e R2,e T/R I2 I1 I3 I4 R3,p R2,p R4,p R1,p drain point Figure 2-2: Parallel Current Paths in a Pipeline Cathodic Protection System It is apparent that each current path is composed of resistance through the earth (Re) plus a resistance through the pipe (Rp) from the point of current pick-up back to the drain point. Therefore, the total resistance (Rt,i) of each parallel path is different and given by Equation 2-1. Rt,i = Ri,e + Ri,p [2-1] Because the length of each current path is different both in the earth and in the pipe in any direction away from the drain point, the total resistance of each current path will increase with distance from the drain point. The amount of current in each path is given by Equation 2-2 Ii = where: R t ,n R t ,i It [2-2] Rt,n = the total resistance of n parallel paths 1 1 1 1 + + + R4 R3 R2 R1 1 = R t ,n ⋅⋅⋅ 1 Rn and: It = I1 + I2 + I3 + I4 … In In stratified soil conditions where the soil resistivity or cross-sectional area of each stratum is different, even current paths of equal length will not have equal resistances as illustrated in Figures 2-3 and 2-4. CP Interference © NACE International, 2006 January 2008 DC Interference 2:3 T/R ρmod till R1,e ρlow clay R2,e ρhigh R3,e rock Figure 2-3: Parallel Current Paths in Vertically Stratified Soil Conditions ρhigh ρlow R4,e R1,e R3,e R2,e T/R I2 I1 I3 R2,p I4 R3,p R4,p R1,p drain point Figure 2-4: Parallel Current Paths in Horizontally Stratified Soil Conditions It is more common than not for soil geology to be stratified both vertically and horizontally and for the current in the low-resistivity soils to be proportionately greater than the high- or moderate-resistivity soils. Furthermore, the stratification need not be caused by different soils but can be due to similar soils with different moisture content. In the vertically stratified soils, the resistance of the current paths is not only a function of soil resistivity but also dependent upon the cross-sectional area of the current path (Equation 2-3). L [2-3] R i,e = ρ s A x,s where: Ri,e = resistance of the current path (ohm [Ω]) ρs = resistivity of the soil CP Interference © NACE International, 2006 January 2008 DC Interference 2:4 L = length of current path Ax,s = cross-sectional area of soil path From a point source electrode like a CP groundbed, the cross-sectional area of the soil increases exponentially with distance from the electrode. Therefore, the resistance of each current path is not linear with distance from the source. Soil resistivities (ρs) are typically in the range of 103 to 106 Ω-cm whereas metal resistivities (ρm) are in the range of 10-5 to 10-6 Ω-cm. Hence the ratio of metal/soil resistivity can range from: ρm 10 -6 10 -5 = to ρs 10 6 10 3 ρm = 10 -8 to 10 -12 ρs Put in perspective, for high soil resistivity (e.g., 106 Ω-cm) a metal object in the earth having a cross-sectional area of 100 cm2 or 10-2 m2 is equivalent in resistance to a cross-section of soil that is given by Equation 2-4. ρm = ρs substituting: ρm = 10 -12 ρs then: A x,s = A x,s = A x,m A x,s [2-4] A x,m 10 -12 10 -2 10 -12 = 1010 m 2 That is, a metal conductor having a 0.01-m2 cross-sectional area is equal to a soil cross-sectional area of 1010 m2 if the soil resistivity is 106 Ω-cm. This means that when a metallic structure is present in the earth, it can be a very attractive current path—thus resulting in a stray current (Is) in the metallic structure as illustrated in Figure 2-5. CP Interference © NACE International, 2006 January 2008 DC Interference 2:5 metallic structure Is Is,4 Is,5 Is,3 Is,2 Is Is,1 R4,e R1,e R3,e R2,e T/R I2 I1 I3 R2,p I4 R3,p R4,p R1,p drain point Figure 2-5: Stray Current in a Metallic Structure Parallel to a Cathodically Protected Structure The stray current is picked up on the foreign metallic structure where it is being impacted by the groundbed anodic voltage gradient. If there is no direct electronic path between the foreign structure and the pipeline, then the current will discharge from the metallic structure remote from the pick-up area. The amount of stray current in the metallic structure is a function of the resistance of the stray current paths and the driving voltage left at the location where the foreign metallic structure intersects the anodic voltage gradient. Currents from a single electrode, placed vertically in the earth, produce a voltage drop in the soil near the electrode that forms equipotential surfaces perpendicular to the current paths (Figure 2-6). CP Interference © NACE International, 2006 January 2008 DC Interference 2:6 Va,re 9 8 Vx,re 7 6 Va,x 5 4 3 2 1 0 x CL distance Va,x Figure 2-6: Voltage vs. Distance from a Vertically Oriented Anode An equipotential surface has the same voltage difference between the anode and any place on its surface. Projection of each equipotential surface at grade and denoting its voltage and distance produces the voltage drop (Va,x) profile in the earth with distance from the anode, as illustrated. The voltage rise (Vx,re) in the earth with respect to remote earth can be calculated using Equation 2-5. Vx,re = I ρ s ⎡ ⎛⎜ L + ⎢ln 2πL ⎢ ⎜⎝ ⎣ L2 + x 2 x ⎞⎤ ⎟⎥ ⎟⎥ ⎠⎦ where: Vx,re = voltage rise in earth with respect to remote earth at a distance “x” from the anode CP Interference © NACE International, 2006 January 2008 [2-5] DC Interference 2:7 I = anode current output ρs = soil resistivity L = length of anode For example, for a 10-m-long anode in 3000- Ω-cm soil having an output of 10A, the voltage rise at 100 m is: Vx,re = 10A × 30 Ω - m ⎡⎢ ⎛⎜ 10m + ln 2π 10m ⎢ ⎜ ⎣ ⎝ (10m )2 + (100m )2 ⎞⎟⎤⎥ 100m ⎟⎥ ⎠⎦ ⎡ ⎛ 10m + 100.5m ⎞⎤ = 4.77 ⎢ln ⎜ ⎟⎥ 100m ⎠⎦ ⎣ ⎝ Vx,re = 4.77 [ln 1.105] = 4.77 [0.1] = 0.48V If a metallic structure was present 100m from this anode, it would be subjected to approximately 0.5V between that point and remote earth. This is the driving voltage that would produce a stray current in the structure. Most impressed current groundbeds, however, do not simply comprise a single electrode placed vertically in the earth. Rather, they typically consist of a number of electrodes placed either vertically or horizontally and interconnected by a common header cable ( Figures 2-7a and 2-7b). Rgb,v s L Ra,1 d Ra,2 Ra,3 Ra,n Figure 2-7a: Multiple Vertical Anodes Connected to a Common Header Cable CP Interference © NACE International, 2006 January 2008 DC Interference 2:8 Rgb,h t d CL s L Figure 2-7b: Multiple Horizontal Anodes Connected to a Common Header Cable Calculation of the voltage rise to remote earth becomes more complicated for multiple anode groundbeds. The following procedure, which equates the resistance of the multiple anode groundbed to the resistance of hemispherical electrode, is a method of estimating the resistance remaining to remote earth. The estimate is then multiplied by the current output to obtain the voltage rise to remote earth. The first step is to calculate the resistance to remote earth of the multiple anode array using Sunde’s equation. Rv = ⎧⎛ 8 L ⎞ ⎫ ρ 2L ln (0.656N) ⎬ ⎟ −1+ ⎨⎜ ln 2 π NL ⎩⎝ d ⎠ s ⎭ [2-6] where: Rv ρ L d s = = = = = N = resistance of multiple vertical anodes to remote earth (Ω) soil resistivity (Ω-cm) length of anode (cm) diameter of anode (cm) anode spacing (cm) number of anodes Note that this equation is simply Dwight’s equation divided by “N” with a “crowding” correction factor added. Example Calculation: As an example, calculate the resistance of 10 vertical anodes, each anode being a 1.5-m-long high silicon iron anode in a 30-cm. diameter by 2-m-long column of metallurgical coke. The anode spacing is 5 m and the soil resistivity is 6,000 Ω-cm. d = 0.3 m L = 2m S = 5m CP Interference © NACE International, 2006 January 2008 ρ = 60 Ω-m N = 10 DC Interference 2:9 Therefore: Rn = 60 Ω - m 2 π 2m × 10 anodes ⎧⎛ 8 × 2 m ⎞ ⎫ 2× 2 m ln (0.656 × 10 anodes)⎬ ⎟ −1+ ⎨⎜ ln 5m ⎩⎝ 0.3 m ⎠ ⎭ = 0.478 {(3.98) – 1) + 0.8 (1.88)} = 0.478 {4.484} = 2.14 Ω This resistance is then equated to an equivalent hemisphere in order to determine the hemisphere radius (r) where the equation for the resistance to remote earth of a hemispherical electrode is given in Equation 2-7 as follows: Rh where: = ρ 2πr [2-7] ρ = resistivity (Ω-m) r = radius of hemispherical electrode (m) Rh = resistance to remote earth (Ω) Icp r1 r Figure 2-8: Hemispherical Electrode The radius of a hemispherical electrode having an equivalent resistance as the multiple anode groundbed is calculated by rearranging the previous equation. CP Interference © NACE International, 2006 January 2008 r = ρ 2π Rh r = 60 Ω - m 6.28 × 2.14 Ω [2-8] = 4.46 m DC Interference 2:10 Then the equivalent hemisphere has a radius of 4.46 m and the resistance included in the earth to a distance r1 is given by the following equation: = R ρ 2π ⎛1 1⎞ ⎜⎜ − ⎟⎟ r1 ⎠ ⎝r [2-9] Therefore 100m from the impressed current groundbed will incorporate a resistance of: R 100 = 60 Ω - m ⎛ 1 1 ⎞ − ⎜ ⎟ 6.28 ⎝ 4.46 m 100 m ⎠ = 9.55 (0.2242 – 0.0100) = 9.55 (0.2142) = 2.04 Ω Then the resistance between a point in the earth 100 m from the center of the hemispherical electrode and remote earth is: R100 →∞ = 2.14 – 2.04 = 0.10 Ω and the voltage rise per ampere of current put out by the groundbed will be 100 mV/A. Hence, a pipeline located 100 m from the 10 anode groundbed operating at 10A output would be subjected to 1 V with respect to remote earth. How much interference current (Is) is picked up by the pipeline would be a function of the pipeline resistance to earth in the pick-up area (Rs,e), the longitudinal resistance (Rs) of the pipeline between the current pick-up and discharge locations, and the resistance to remote earth (Rs,re) at the discharge location (Figure 2-9). CP Interference © NACE International, 2006 January 2008 DC Interference 2:11 Ra,re Icp Rc,a Icp ' A Rs,e Is B Rs Rc,p Rs,re remote earth Icp Rp,re where: Icp = Icp ' + Is Rc,a & Rc,p = cable resistances Ra,re = anode resistance to remote earth Rp,re = pipe resistance to remote earth Rs,e = resistance of foreign pipe to earth in a stray current pick-up area Rs,re = foreign structure resistance to remote earth Rs = longitudinal resistance of foreign structure between pick-up and discharge sites Figure 2-9: Cathodic Protection Circuit Model with Foreign Structure Intercepting the Anode Gradient Calculation of the pipe-to-earth resistance can be carried out by a number of methods. For electrically short lengths of bare pipe (i.e., where attenuation is not significant) Equation 2-10 can be used. R s,e ρ L d t where: = = = = = ⎧ (L )2 ⎫ ρ ln ⎨ ⎬ 2πL ⎩ td ⎭ [2-10] soil resistivity length of pipe diameter of pipe depth below grade Example Calculation: Assuming a 100 m long, 0.25 m diameter pipeline at 1 m depth in 60 Ω-m soil, the pipe resistance would be: R s,re = ⎧ (100 )2 ⎫ 60 Ω - m ln ⎨ ⎬ 6.28 × 100 m ⎩1 × 0.25 ⎭ = 0.096 Ω × 10.6 = 1.02 Ω CP Interference © NACE International, 2006 January 2008 DC Interference 2:12 If the pipe is coated, then the voltage rise in the earth at the pipe will appear across the coating. In this case, the resistance of the pipe is the series combination of Equation 2-10 plus the resistance across the coating (Rs,c). To obtain Rs,c the specific coating resistance (r′e) is needed. Accordingly, given a good-quality coating having a specific coating resistance of 5 × 103 Ω-m 2 from Table 2-1 in 1000 Ω-cm soil, the specific coating resistance in 6000 Ω-cm soil (r′c @ 6000 Ω-cm) is then obtained by multiplying the specific coating resistance at 1000 Ω-cm (r′c @ 1000 Ω-cm) by the ratio of the actual soil resistivity divided by 1000 Ω-cm. rc′@ 6,000 Ω-cm = rc′@ 1,000 Ω-cm × 6,000 Ω - cm 1,000 Ω - cm = 5 × 103 Ω-m 2 × 6 = 3 × 104 Ω-m 2 CP Interference © NACE International, 2006 January 2008 DC Interference 2:13 Table 2-1: Specific Leakage Resistances and Conductances in 1000 Ω-cm Soil or Water Long Pipelines with Few Fittings Average Specific Coating Conductance g′c Average Specific Coating Resistance r′c Quality of Work Siemens/ft2 Siemens/m2 Ω-ft2 Ω-m2 Excellent <1 x 10-5 <1 x 10-4 >105 >104 Good 1 x 10-5 to 5 x 10-5 1 x 10-4 to 5 x 10-4 2 x 104 to 105 2 x 103 to 104 Fair 5 x 10-5 to 1 x 10-4 5 x 10-4 to 1 x 10-3 104 to 2 x 104 103 to 2 x 103 Poor Bare Pipe (2” to 12”) (5cm to 30cm) >1 x 10-4 >1 x 10-3 <104 <103 4 x 10-3 to 2 x 10-2 4 x 10-2 to 2 x 10-1 50 to 250 5 to 25 Gas or Water Distribution with Many Fittings Average Specific Coating Conductance g′ c Average Specific Coating Resistance r′c Quality of Work Siemens/ft2 Siemens/m2 Ω-ft2 Ω-m2 Excellent <5 x 10-5 <5 x 10-4 >2 x 104 >2 x 103 Good 5 x 10-5 to 1 x 10-4 5 x 10-4 to 1 x 10-3 104 to 2 x 104 103 to 2 x 103 Fair 1 x 10-4 to 5 x 10-4 1 x 10-3 to 5 x 10-3 2 x 103 to 104 2 x 102 to 103 Poor Bare Pipe (2” to 12”) (5cm to 30cm) >5 x 10-4 >5 x 10-3 <2 x 103 <2 x 102 4 x 10-3 to 2 x 10-2 4 x 10-2 to 2 x 10-1 50 to 250 5 to 25 Therefore the coating resistance (Rs,c) of the 100 m long pipe is: R s,c = where: therefore: rc′@ 6,000 Ω-cm As As = surface area of pipe = πdL R s,c 3 × 10 4 Ω − m 2 = π × 0.25m × 100 m = 382 Ω and the total resistance (Rs,re) of the 100m of coated pipe will be: CP Interference © NACE International, 2006 January 2008 DC Interference 2:14 Rs,e = Rs,c + Rs,re = 382 Ω + 1.02 Ω = 383 Ω [2-11] The resistance of the coated pipe in the pick-up area is primarily the resistance across the coating. Assuming the interference situation illustrated in Figure 2-5, the interference current will discharge in an endwire pattern as illustrated in Figure 2-10. discharge Ecorr pick-up Ep Figure 2-10: Potential Profile Along the Interfered-with Structure The resistance looking left (RS,L) and right (RS,R) from the pick-up location will be affected by attenuation if the length of pipe is relatively long. IS RS,L RS,R Figure 2-11: Electrical Model for Interfered-with Pipe The total resistance of the pipe to remote earth is the parallel combination of the two resistances. Therefore: CP Interference © NACE International, 2006 January 2008 DC Interference 2:15 = R s,re (discharge) R S, L × R S, R [2-12] R S, L + R S, R The resistance looking along a structure where attenuation is a factor is given by the following equation. RS,O = RG coth αx [2-13] where: RG = characteristic resistance α = attenuation constant x = length of pipe The attenuation constant α is a function of the leakage resistance (RL) and the lineal resistance (Rm). or Rm Rm RS,L Rm RS,R RL RL RL Figure 2-12: Attenuation Model The attenuation constant is calculated from the Equation 2-14: α = where: Rm RL [2-14] Rm = lineal resistance of the pipe RL = leakage resistance of pipe to earth R m = ρm × where: Rm ρm L Ax = = = = lineal resistance of the pipe resistivity of pipe material length of pipe section cross-sectional area of pipe CP Interference © NACE International, 2006 January 2008 L Ax [2-15] DC Interference 2:16 RL = and where: RL r′c L AS = = = = rC′ AS [2-16] leakage resistance of pipe to earth specific resistance of pipe coating length of pipe section surface area of section = πdL Example Calculation: For 100m long, 0.25m diameter (10-in dia.) schedule 40 pipe having an outside diameter (OD) of 27.3cm (Table B-1 from Appendix B) and an inside diameter (ID) of 25.4 cm. α = Rm RL Rm Rm = 18 × 10 −6 Ω − cm × 10,000 cm {(27.3 cm) 4 π 2 − (25.4 cm ) 2 } = 1.8 × 10 − 2 Ω − cm 2 0.785 745.3 cm 2 − 645.2 cm 2 = 1.8 × 10 − 2 78.6 { = } 2.29 × 10 -4 Ω/100m The leakage resistance was previously calculated (Eqn. 2-11) at 383 Ω for a 100 m length of pipe having a good coating in 6,000 Ω-cm soil. Therefore the attenuation constant is: α = = 2.29 × 10 -4 Ω 383 Ω 0.598 × 10 -6 = 0.773 × 10 -3 The resistance looking both ways from the center of the interfered-with pipe can be determined using Equations 2-11 and 2-13. But if the pipe is CP Interference © NACE International, 2006 January 2008 DC Interference 2:17 infinitely long, the resistance looking into it is just the characteristic resistance (i.e., RS,O = RG) and the characteristic resistance is given by Equation 2-16. R S,O = R G = Rm × RL = 2.29 × 10 -4 Ω × 383 Ω = 8.77 × 10 -2 Ω 2 [2-17] = 0.296 Ω Therefore the resistance of the pipe to remote earth (RS,re) from Equation 212 is: R S, re (discharge) = 0.296 Ω × 0.296 Ω 0.592 Ω = 0.148 Ω Thus the total circuit resistance (RT) related to the interfered with pipe circuit is the sum of the resistance in the pick-up area and the discharge area. In this case: RT = 383 Ω + 0.148 Ω The coating resistance dominates this circuit and limits the stray current magnitude. A similar voltage drop occurs in the earth around a bare pipeline as indicated in Figure 2-13. CP Interference © NACE International, 2006 January 2008 DC Interference 2:18 current path equipotential surface Figure 2-13: Voltage Gradient in the Earth Around a Cathodically Protected Bare Pipeline The typical CP circuit can then be modeled as a series circuit shown in Figure 214. Ra,re Rc,a remote earth Rc,p Rp,re where: Rc,a & Rc,p = cable resistances Ra,re = anode resistance to remote earth Rp,re = pipe resistance to remote earth Figure 2-14: Cathodic Protection Circuit Model If a metallic structure is located in the earth as shown in Figure 2-5 then it will intercept the anode voltage gradient such that there will be a parallel path inserted into the model as illustrated in Figure 2-15. CP Interference © NACE International, 2006 January 2008 DC Interference 2:19 Ra,re Icp Rc,a A Icp ' Rs,e Is B Rs Rc,p Rs,re remote earth Icp Rp,re where: Icp = Icp ' + Is Rc,a & Rc,p = cable resistances Ra,re = anode resistance to remote earth Rp,re = pipe resistance to remote earth Rs,e = resistance of foreign pipe to earth in a stray current pick-up area Rs,re = foreign structure resistance to remote earth Rs = longitudinal resistance of foreign structure between pick-up and discharge sites Figure 2-15: Cathodic Protection Circuit Model with Foreign Structure Intercepting the Anode Gradient The presence of the foreign structure has introduced a parallel circuit into the model where the voltage drop between point “A” and remote earth is applied to the foreign structure. This will lower the overall resistance of the anode to remote earth and diminish the CP current beyond point “A” to I′cp by the amount of Is. If the foreign structure also crosses the pipeline as shown in Figure 2-16, then the foreign structure resistance to the pipeline will be lowered because of the close proximity of the two pipelines at the crossing. This would result in a larger stray current because the driving voltage between “A” and “B” will be greater (Figure 2-17). CP Interference © NACE International, 2006 January 2008 DC Interference 2:20 metallic structure Is Is A Is,4 Is,5 Is,3 Is,2 Is,1 R4,e R1,e R3,e R2,e T/R I2 I1 I3 R2,p Is,1 I4 B B R3,p R4,p R1,p drain point Figure 2-16: Stray Current in a Foreign Metallic Structure that Intercepts both the Anodic and Cathodic Voltage Gradient Rc,a Is Icp Icp ' A Ra,re Rs,e remote earth Rs Rc,p Rs,p Icp Rp,re B Icp ' where: Icp = Icp ' + Is Rc,a & Rc,p = cable resistances Ra,re = anode resistance to remote earth Rp,re = pipe resistance to remote earth Rs,p = resistance of foreign pipe to cathodically protected pipe at discharge area Rs = longitudinal resistance of foreign structure between pick-up and discharge sites Rs,e = resistance of foreign pipe to earth in a stray current pick-up area Figure 2-17: Cathodic Protection Circuit Model with Foreign Structure Intercepting both Anodic and Cathodic Voltage Gradient A metallic foreign structure can also be subject to a stray current even if it only intersects the cathodic voltage gradient as illustrated in Figures 2-18 and 2-19. CP Interference © NACE International, 2006 January 2008 DC Interference 2:21 Rn,e A R4,e R1,e R3,e R2,e Is T/R I2 I1 I3 R2,p I4 B B R3,p R4,p R1,p Is drain point A Figure 2-18: Stray Current in a Foreign Metallic Structure that Intercepts the Cathodic Protection Gradient In this situation the interfered-with pipeline picks up stray current at remote earth “A” and transports it to the crossing, where it discharges back to the interfering structure. This means that any pipeline protected by impressed current systems can cause interference on crossing metallic structures that are otherwise remote from the impressed current groundbeds. Furthermore, the stray current discharge need not be to the interfering structure but rather to a third-party structure acting as an intermediate current path. CP Interference © NACE International, 2006 January 2008 DC Interference 2:22 Ra,re Rc,a Icp Rs Rc,p Icp Rp,re Rs,re A remote earth Is Rs,p B Icp ' where: Icp = Icp ' + Is Rc,a & Rc,p = cable resistances Ra,re = anode resistance to remote earth Rp,re = pipe resistance to remote earth Rs,re = foreign structure resistance to remote earth Rs,p = foreign structure resistance to cathodically protected structure at stray current discharge location Rs = longitudinal resistance of foreign pipe between remote earth and discharge location Figure 2-19: Cathodic Protection Circuit Model for Foreign Structure Intercepting the Cathodic Voltage Gradient As has been demonstrated, a stray current can occur in a foreign metallic structure if it is impacted by either the anodic or cathodic voltage gradient produced by a pipeline impressed current system. The magnitude of the stray current is directly proportional to the voltage between the current pick-up and discharge location and inversely proportional to the resistance of the interference current path. CP Interference © NACE International, 2006 January 2008 DC Interference 2.2 2:23 Detecting Stray Current There will be potential and current changes on and near a metallic structure due to any stray current. These electrical disturbances are as follows: • structure-to-soil potential changes at both stray current pick-up and discharge locations • current changes in the structure between the current pick-up and discharge locations • current changes in the earth near the structure at the current pickup and discharge locations. If the output of the transformer-rectifier shown in Figure 2-16 is cyclically interrupted and a close-interval potential survey is conducted over the foreign structure from left to right, the potential profile as illustrated in Figure 2-20 would be typical. + E E off E S/S - E E on B A Distance Figure 2-20: Typical Potential Profile on an Interfered-with Structure that Intersects both Anodic and Cathodic Voltage Gradient with the Current Source Interrupted Point “A” is the location on the structure immediately opposite the groundbed location, and point “B” is at the pipeline crossing. When the current source is on, there is a negative shift at the pick-up region (point “A”) and a positive shift at the discharge location (point ‘B’). Detection of current magnitude changes—involved in the stray current situation in Figures 2-5, 2-16, and 2-18—are illustrated in Figure 2-21. CP Interference © NACE International, 2006 January 2008 DC Interference 2:24 Vs t A Is Ve B Rs,1 Is Ve = where: Vs t = Is R e Is Rs,1 Is Ve Figure 2-21: Current Changes In and Near an Interfered-with Structure Current changes are detected by measuring the voltage drop in the earth adjacent to the interfered-with structure and the voltage drop (ΔVst) over a length of the structure with the stray current source cyclically interrupted. Therefore, there is a change in the earth voltage drop (ΔVe) at both “A” and “B” due to the stray current Is. At “A,” ΔVe will be positive because the stray current is toward the structure. At “B” it will be negative, indicating a current way from the structure. The change in structure voltage drop will be in the positive direction for the meter polarity shown. 2.2.1 Mitigation of Interference Effects from Impressed Current Cathodic Protection Systems There are a number of methods that can be used to lessen the deleterious effects of CP system stray currents, as listed below: • remove the source or reduce its output • install electrical isolating fittings in the interfered-with structure • bury a metallic shield parallel to the interfered-with structure at the stray current pick-up zone • install additional CP at current discharge locations on the interferedwith structure • install a bond between the interfered-with and interfering structures • apply a coating to the interfered-with structure in the area of stray current pick-up or to the interfering structure where it picks up the returning stray current. CP Interference © NACE International, 2006 January 2008 DC Interference 2:25 Before any mitigation activity can commence, it is necessary to conduct mutual interference tests where the output of the suspected interference source is cyclically interrupted and field measurements taken in the presence of representatives of the interfering and interfered-with companies. Interference cases are often reported through local electrolysis committees, especially where there may be more than one interfered-with party. Presuming that a need for mitigation is determined, the mutually acceptable mitigation technique(s) will depend on the location and severity of the interference, on the CP operational preferences of each party, and on the relative capital and maintenance costs of the mitigation options. 2.2.1(a) Source Removal or Output Reduction It is a difficult proposition to have a source removed if the interfering system was present before the interfered-with structure was installed. However, in the opposite situation, where the interfering source is newly installed, this method has greater appeal. If the interference is caused primarily by the proximity of the interfered-with structure to the interfering groundbed, it may not be necessary to remove the transformerrectifier but simply relocate the groundbed location or reduce the current output. Equation 2-5 or similar equations1 can be used to estimate how remote a particular groundbed needs to be from a foreign structure in order to minimize the interference effects. It should be noted, however, that the voltage rise at any point distance “x” from the groundbed is a percentage of the total voltage drop to remote earth (Vx,re/Vgb,re × 100). The voltage rise is a function only of the geometry of the groundbed (i.e., its length “L”) because the groundbed current output and soil resistivity would not change. Therefore, only the length parameter in the equation significantly affects the percentage. 1 Von Baekmann, Schwenk, and Prinz, Cathodic Corrosion Protection, 3rd Edition Gulf Publishing, 1997, pp.538-539. CP Interference © NACE International, 2006 January 2008 DC Interference 2:26 Reducing the current output of the source is also a viable option as long as there are safeguards to prevent the output from being raised inadvertently. 2.2.1(b) Installation of Isolating Fittings Installing isolating fittings as a stray current mitigation measure is an attempt to increase the path resistance (Rs) of the interfered-with structure, thus decreasing the stray current (Is). This is seldom adequate as a stand-alone method. The stray current will certainly be reduced, but the lesser amount of stray current will bypass each isolating fitting in the soil path. Hence, several points of interference will be created (as previously shown in Figure 1-15a). Consequently, additional CP may be needed at each isolating joint to compensate for the residual stray current. The installation of isolating fittings to electrically sectionalize piping systems, as illustrated in Figure 2-22, is a common practice. isolating fitting I''' s I's I''s isolating fitting T/R Figure 2-22: Stray Current Arising from Installation of Isolating Fittings Unfortunately, inserting electrical isolation often produces a stray current condition at the isolating fitting. Therefore on piping networks protected with impressed current systems, electrical isolation should be used sparingly. When electrical isolation is used, facilities to mitigate the expected interference should be provided at each point of electrical isolation. CP Interference © NACE International, 2006 January 2008 DC Interference 2.2.1(c) 2:27 Burying a Metallic Shield Next to the Interfered-with Structure The intent of a buried metallic conductor is to intercept the stray current and thus provide an alternative low-resistance path for the stray current compared to the metallic structure path. Connecting the metallic shield, which could be a bare cable or pipe, directly to the negative terminal of the offending transformer-rectifier—as shown in Figure 2-23 and modeled in Figure 2-24—would be more effective than connecting it to the interfered-with structure. Interfered-with Structure Bare Shield Is Icp T/R Cathodically Protected Structure Figure 2-23: Using a Buried Metallic Cable or Pipe as a Shield to Reduce Stray Current Interference CP Interference © NACE International, 2006 January 2008 DC Interference 2:28 A Ra,re Rc,a Icp Rsh,e Icp ' Rs,e I's B Rsh Rc,p I''s Rs Rs,re remote earth Icp Rp,re where: Icp = Icp ' + I's + I''s Rc,a & Rc,p = cable resistances Rsh,e = shield resistance to earth Rsh = shield cable longitudinal resistance I's I''s = stray current in shield wire path = residual stray current in foreign pipeline Figure 2-24: Cathodic Protection Current Model for a Buried Metallic Shield Connected to the Negative Terminal of the Transformer-Rectifier The alternative approach which would be to connect the buried metallic shield to the interfered-with structure, increasing the stray current discharge at point “B.” This buried metallic shield method has most merit either where the interfered-with structure is made of an amphoteric material or there is a concern about coating blistering or cathodic disbondment. For the interfering system, there is considerable disadvantage to this technique because it could seriously disrupt the current distribution pattern to the cathodically protected structure, perhaps even necessitating the installation of additional CP units to make up for the poorer current distribution. 2.2.1(d) Installation of Galvanic Anodes on the Interfered-with Structure at Point of Stray Current Discharge When the area of stray current discharge is very localized—such as at a crossing with the interfering structure—and where the total stray current (Is) is typically less than an ampere, the installation of galvanic anodes (Figure 2-25) has considerable benefit. CP Interference © NACE International, 2006 January 2008 DC Interference 2:29 Test Station Interfering Structure Is,t I's I's' I's Icp I's Icp I's Icp I's I's I's Icp Icp Icp Interfered-with Structu re Figure 2-25: Interference Mitigation using Galvanic Anodes at Stray Current Discharge Location If the interfered-with structure is coated at the crossing, then the path resistance (Rap) through the galvanic anodes will be substantially lower than the interfered-with structure resistance (Rs1,p). The electrical circuit model in Figure 2-26 depicts the structure resistance. Although there can still be a residual stray current (Is′′), it is expected that the total CP current (∑Icp) will be greater—thus assuring total remediation of the interference. CP Interference © NACE International, 2006 January 2008 DC Interference 2:30 Icp Is Ra,re Rc,a I'cp A Rs,e I's + Icp,g Ra,p Rc,p Rp,re Icp = Icp ' I''s Rs,p Icp,g Icp where: Eg remote earth B I'cp + I's + I''s Rc,a & Rc,p = cable resistances Ra,p = anode(s) resistance to the interfering pipe Eg = galvanic anode driving voltage Icp,g = I's = I''s = galvanic anode current stray current through galvanic anodes residual stray current discharging from foreign pipeline Figure 2-26: Electrical Circuit Model for Mitigating Stray Current Interference at a Stray Current Discharge Site Using Galvanic Anodes Ideally, the galvanic anodes are distributed alongside the interfering structure in order to minimize the path resistance (Ra,p); therefore, the stray current (I′s) is a large percentage of the total stray current (Is). The design life of the galvanic anodes must take into account the additional consumption by the stray current (I′s) component of its total output. Several advantages of this method are as follows: • the interfered-with structure can maintain CP independence • the galvanic anode CP current output boosts the level of protection at the crossing as an added buffer should the interference current (Is) increase • low maintenance requirements compared to a direct bond. The disadvantages are that it is relatively expensive compared to a direct bond and the interference current mitigation capacity is somewhat limited. To mitigate large interference currents, an impressed current system can be utilized having the drain point at the crossing but the groundbed remote from both piping systems. CP Interference © NACE International, 2006 January 2008 DC Interference 2:31 Example: Design Calculations for a Galvanic Anode Interference Mitigation System At a crossing of two coated and cathodically protected pipelines, a temporary resistance bond of 3 Ω passing 350 mA was required to mitigate the interference on Pipeline A caused by Pipeline B’s impressed current systems. It has been decided to mitigate that interference problem using magnesium anodes because the soil resistivity is low (3100 Ω-cm) and Pipeline A wishes to maintain its cathodic protection independence. Step 1: Choose a #17D2 magnesium anode from Table 1 in Appendix C. Step 2: Calculate the resistance of a single vertical anode from Dwight’s Equation: Ra = where: Ra ρ L d Ra = then: Step 3: = = = = ρ ⎧⎛ 8 L ⎞ ⎫ ⎨⎜ ln ⎟ − 1⎬ 2πL ⎩⎝ d ⎠ ⎭ [2-18] resistance of vertical anode to remote earth soil resistivity (Ω-m) = 31 Ω-m length of packaged anode (m) = 1.5 m diameter of packaged anode (m) = 0.15m ⎧⎛ 12 ⎞ ⎫ 31 ⎟ − 1⎬ = 3.3{3.38} = 11.2 Ω ⎨⎜ ln 6.28 × 1.5 ⎩⎝ 0.15 ⎠ ⎭ Calculate minimum number of anodes (N) to achieve a 3-Ω resistance. Assuming no mutual resistance effects between anodes then N= Step 4: Calculate anode CP current output assuming a pipeline polarized potential of –850 Vcse Ia = Step 5: Ra 11.2 = = 3.72 or 4 anodes 3.0 RT DrivingVol tage 1.700 V − 0.850 V = = 75.9 mA Ra 11.2 Ω Calculate total current output (It) per anode. It = Ia + Iint CP Interference © NACE International, 2006 January 2008 DC Interference 2:32 = 75.9 mA + Step 6: It = 163.4 mA Calculate anode life: L = where: Note: L Wt U E Ia Cr = = = = = = Wt × U × E I a × Cr [2-19] effective service life (y) total weight of anode alloy (kg) utilization factor efficiency current output (A) theoretical consumption rate (kg/A-y) Consumption rate for magnesium @ 50% efficiency is approximately 8 kg/A-y. An utilization factor of 0.85 is assumed. L = Step 7: 350 mA 4 7.7 kg × 0.85 = 5y 0.163 A × 8kg / A − y Calculate minimum anode weight to achieve a 20-year life using equation [2-19]. Wt = Wt = L × I a × Cr U×E 20 yr × 0.163 A × 8kg / A − yr = 30.7 kg 0.85 therefore: Total anode weight for the mitigation system would need to be 4 × 30.7 kg = 122.7 kg Step 8: Chose a larger weight anode from Table 1 in Appendix C. GROUP EXERCISE Determine an appropriate anode size and number to achieve a 20-y life. CP Interference © NACE International, 2006 January 2008 DC Interference 2.2.1(e) 2:33 Installation of an Impressed Current Distribution System on the Interfered-with Structure at Point of Stray Current Discharge The stray current situation, depicted in Figure 2-5, results in a typical potential profile along the interfered-with structure. In this case, the stray current discharge (+ΔE) occurs in an end-wise fashion (Figure 2-27). + E E S/S + E - E A B Distance Figure 2-27: Potential Profile Changes on a Pipeline Where Stray Current is Discharging in an End-Wise Pattern Although the positive potential shift may be modest, the length of the discharge can be extensive. Under these conditions, the installation of an impressed current system at the discharge locations (“A” and “B”) can be an effective means of compensating for the stray current interference. Care must be taken to ensure that the impressed current systems do not create interference on the original interfering structure. 2.2.1(f)(i) Installing a Bond Between the Interfered-with and Interfering Structures Perhaps the most common stray current mitigation method is the installation of a bond. The bond usually has some resistance between the two structures and usually is located at the point of maximum stray current discharge, such as at a crossing as shown in Figure 2-28. CP Interference © NACE International, 2006 January 2008 DC Interference 2:34 I's P1 Bond Cable Variable Resistor P2 Bond Cable Test Station (Rb) interfer ed-with structure P1 I's I's Is bond cable and test station I's' P2 interfering structure buried reference electrode Figure 2-28: Interference Mitigation Using a Resistance Bond The electrical circuit model is similar to Figure 2-26, except for the fact that the bond resistance (Rb) replaces the galvanic anode resistance (Ra,p) in the circuit. Typically, the bond resistance is determined by monitoring the potential of the interfered-with structure while adjusting the resistance until the interfered-with structure is returned to its CP criterion or native potential on a structure having no CP. A zinc reference installed between the two structures at the crossing is an optional, but nevertheless worthwhile, feature. A resistance bond will not eliminate all the current discharge at the crossing because there will still be a residual stray current discharge (Is′′), which must be countered by the interfered-with structure’s CP system. The major advantages of a resistance bond over other mitigation methods include: • relatively inexpensive to install • easy to adjust if stray current magnitude changes • has a high current capacity. Disadvantages of using a resistance bond are as follows: CP Interference © NACE International, 2006 January 2008 DC Interference 2:35 • resistance bonds are vulnerable to AC fault current transients that can burn out the resistor, unless protected with fault current devices. • connecting two structures through a resistance bond means that CP changes on either structure will affect protection levels on the other structure • surveys to measure true polarized potentials on either structure may require the synchronous interruption of the bond with impressed current systems on both structures • resistance bonds are considered critical components and, by regulation, require frequent inspection. • inadvertent removal or adjustment by unqualified people. 2.2.1(f)(ii) Calculation of Bond Resistance The value of the resistor needed for the interference bond can be determined in the field using trial-and error methods, or it can be calculated using the procedures originally proposed by Pearson2 and later described by Seifert3, McGary4, and others. Consider the pipeline system shown in Figure 1-1. Line 1, which is cathodically protected by an impressed current system, causes a stray current interference problem where it crosses Line 2. 2 J.M. Pearson, “Measurements of Cathodic Polarization and of Problems of Interference on Underground Structures” 3 R.L. Seifert, “The Use of A Programmable Electronic Calculator in Underground Corrosion Related Activity.” NACE Corrosion/79, Paper 190. 4 W.McGary, “Interference Effects on Underground Structures and the Criteria for Correction by Electrical Bonds.” NACE Northeast Regional Conference, Philadelphia PA, 1964. CP Interference © NACE International, 2006 January 2008 DC Interference 2:36 Rectifier IB V12 + RB Line 2 IR V1 + + V2 - Line 1 Figure 2-29: Measurements Required to Determine Size of Resistance Bond RB In order to mitigate this interference, a resistance bond must be installed in the joint test station at the crossing, which will restore Line 2 to its natural potential (i.e., the pipe-to-soil potential measured with the rectifier off). The first step in sizing this bond is to record the following measurements: IR = Rectifier current V1 (on, oc) = Pipe-to-soil potential of Line 1 at crossing, with rectifier on, and bond open-circuited V1 (off, oc) = Pipe-to-soil potential of Line 1 at crossing, with rectifier off, and bond open-circuited V2 (on, oc) = Pipe-to-soil potential of Line 2 at crossing, with rectifier on, and bond open-circuited V2 (off, oc) = Pipe-to-soil potential of Line 2 at crossing, with rectifier off, and bond open-circuited A temporary bond having an arbitrary resistance is then installed, and the following additional measurements are taken: IB = Bond current V1 (on, cc) = Pipe-to-soil potential of Line 1 at crossing, with rectifier on, and bond close-circuited V1 (on, oc) = Pipe-to-soil potential of Line 1 at crossing, with rectifier on, and bond open-circuited V2 (on, cc) = Pipe-to-soil potential of Line 2 at crossing, with rectifier on, and bond close-circuited V2 (on, oc) = Pipe-to-soil potential of Line 2 at crossing, with rectifier on, and bond open-circuited The following parameters are then calculated: RR1 = Rectifier Protective Coupling with Line 1 (i.e., the potential change on Line 1 per ampere of rectifier current) CP Interference © NACE International, 2006 January 2008 DC Interference 2:37 V1 (off , oc ) − V1 (on, oc ) IR RR1 = [2-20] RR2 = Rectifier Interference Coupling with Line 2 (i.e., the potential change on Line 2 per ampere of rectifier current) RR 2 = V2 (on, oc ) − V2 (off , oc ) IR [2-21] RR1,2 = Rectifier Mutual Coupling between Lines 1 and 2 (i.e., the potential change between Lines 1 and 2 per ampere of rectifier current) RR1, 2 = RR1 + RR 2 [2-22] RB1 = Bond Interference Coupling with Line 1 (i.e., the potential change on Line 1 per ampere of bond current) RB1 = V1 (on, cc ) − V1 (on, oc ) IB [2-23] RB2 = Bond Protective Coupling with Line 2 (i.e., the potential change on Line 2 per ampere of bond current) RB 2 = V2 (on, oc ) − V2 (on, cc ) IB [2-24] RB1,2 = Bond Mutual Coupling between Lines 1 and 2 (i.e., the potential change between Lines 1 and 2 per ampere of bond current) RB1, 2 = RB1 + RB 2 [2-25] V1,2 (off, oc) = Voltage between Lines 1 and 2 with rectifier off and bond open V1, 2 (off , oc ) = V2 (off , oc ) − V1 (off , oc ) [2-26] The required bond size, and the resulting bond currents with the rectifier on and off, can then be calculated as: CP Interference © NACE International, 2006 January 2008 DC Interference 2:38 RB = V1, 2 (off , oc ) + I R RR1, 2 − RB1, 2 RR 2 IR RB 2 I B (on ) = I B (off ) = I R RR 2 RB 2 V1, 2 (off , oc ) RB + RB1, 2 [2-27] [2-28] [2-29] The final pipe-to-soil potentials with the bond installed can be calculated for both lines as follows: V1 (on, cc) = V1 (on, oc) + I B (on) RB1 [2-30] V1 (off , cc) = V1 (off , oc) + I B (off ) RB1 [2-31] V2 (on, cc) = V2 (on, oc) − I B (on) RB 2 [2-32] V2 (off , cc) = V2 (off , oc) − I B (off ) RB 2 [2-33] Sample Calculation: In the pipeline system of Figure 2-29, conduct field testing and perform the necessary calculations to determine the required bond size to mitigate interference on Line 2. Solution: The following field measurements are made: Rectifier Current: IR = 10 A Pipe-to Soil Potentials: Protected Line (Line 1): Rectifier on: V1 (on, oc) = –1.45 V Rectifier off: V1 (off, oc) = –0.95 V Foreign Line (Line 2): Rectifier on: V2 (on, oc) = –0.75 V CP Interference © NACE International, 2006 January 2008 DC Interference 2:39 Rectifier off: V2 (off, oc) = –0.90 V A piece of resistor wire (of unknown resistance) is used to create a temporary bond between lines 1 and 2. The following additional measurements are then made while the rectifier remains on: Temporary Bond Current: IB = 0.75 A Pipe-to Soil Potentials (while interrupting bond current): Protected Line (Line 1): Bond closed: V1 (on, cc) = –1.30 V Bond open: V1 (on, oc) = –1.45 V Foreign Line (Line 2): Bond closed: V2 (on, cc) = –0.825 V Bond open: V2 (on, oc) = –0.75 V R R1 = RR 2 = V1 (off , oc ) − V1 (on, oc ) − 0.95 − (− 1.45) V = = 0.05Ω 10 IR A V2 (on, oc ) − V2 (off , oc ) − 0.75 − (− 0.90 ) V = = 0.015Ω IR A 10 RR1, 2 = RR1 + RR 2 = 0.05 + 0.015 = 0.065Ω RB1 = RB 2 = V1 (on, cc ) − V1 (on, oc ) − 1.30 − (− 1.45) V = = 0 .2 Ω 0.75 IB A V2 (on, oc ) − V2 (on, cc ) − 0.75 − (− 0.825) V = = 0.1Ω IB A 0.75 RB1, 2 = RB1 + RB 2 = 0.2 + 0.1 = 0.3Ω V1, 2 (off , oc ) = V2 (off , oc ) − V1 (off , oc ) = (− 0.90 ) − (− 0.95) = 0.05V CP Interference © NACE International, 2006 January 2008 DC Interference 2:40 The bond current required for mitigation is: I B (on ) = I R RR 2 (10 )(0.015 ) = = 1 .5 A RB 2 0 .1 The required bond resistance is: RB = V1, 2 (off , oc ) + I R RR1, 2 0.05 + (10 )(0.065) − RB1, 2 = − 0.3 = 0.167Ω RR 2 0.015 IR 10 RB 2 0.1 The current through the bond when the rectifier is off is: I B (off ) = V1, 2 (off , oc ) RB + RB1, 2 = 0.05 = 0.11A 0.167 + 0.3 The final potentials on the pipelines with the rectifier interrupted are: V1 (on, cc) = V1 (on, oc) + I B (on) RB1 = −1.45 + (1.5)(0.2) = −1.15V V1 (off , cc) = V1 (off , oc) + I B (off ) RB1 = −0.95 + (0.11)(0.2) = −0.93V V2 (on, cc) = V2 (on, oc) − I B (on) RB 2 = −0.75 − (1.5)(0.1) = −0.90V V2 (off , cc) = V2 (off , oc) − I B (off ) RB 2 = −0.9 − (0.11)(0.1) = −0.91V 2.2.1(g) Use of Coatings in the Mitigation of Interference Effects Application of a coating is an attempt to increase the resistance of the stray current path thus decreasing the stray current magnitude. As a stand-alone method, coating should only be applied at current pick-up locations. If the discharge area of a structure is coated, there is a risk of corrosion failure owing to a high discharge current density at a holiday in the coating. CP Interference © NACE International, 2006 January 2008 DC Interference 2:41 There are two current pick-up regions, one on the interfered-with structure and one on the interfering structure in the vicinity of the stray current discharge as shown in Figure 2-30. Is Is coated pipe sections T/R I1 Figure 2-30: Use of a Dielectric Coating to Mitigate Interference This technique is easy to implement on a new structure, where a high resistance coating can be used in areas where stray current pick-up is anticipated. It may be generally impractical for existing facilities. 2.2.2 Other Sources of DC Stray Current Besides impressed current CP systems, there are other sources of DC stray current: • • • • DC transit systems DC welding equipment high-voltage DC transmission systems (HVDC) DC rail systems in mines. Because these sources have a variable loading nature, the resulting stray current activity is dynamic (i.e., effects vary in magnitude and often location with time). Another source of dynamic stray current, telluric currents, is discussed in Chapter 4. CP Interference © NACE International, 2006 January 2008 DC Interference 2.2.2(a) 2:42 DC Transit Systems The electrification of transit systems in the late 1800s throughout North America resulted in considerable interference corrosion on gray cast iron watermains. Much of the early attempts to mitigate this interference eventually led to the development of CP technology.5 DC substation O/H power conductor IL IR Is running rails ground Is Is pick-up metallic structure (e.g.,watermain) Ie Ie discharge Figure 2-31: Typical Stray Current Paths Around a DC Transit System The load current (IL), after passing through the trolley motor, divides into a number of current paths depending on the resistance of each path. therefore: IL = IR + Is + Ie [2-34] Although the rails provide a relatively low-resistance path, the current leakage off the rails can be 5 to 10% of the load current. This may seem a small percentage, but the stray currents can be substantial because the start-up load current can be several hundred amperes for a single trolley and several thousand amperes for a subway train. 5 Kuhn, R.J., Cathodic Protection of Underground Pipelines from Soil Corrosion, API Proceedings, Nov. 1933, Vol. 14, p.164. CP Interference © NACE International, 2006 January 2008 DC Interference 2:43 Not only will the magnitude of the stray current vary with time of day and whether the vehicle is accelerating or decelerating, but the location of stray current pick-up on the metallic structure will change as the trolley moves along the rail. Thus, a structure-to-soil potential recording will have a dynamic appearance (Figure 2-32). 0 -1000 -1500 -2000 -2500 -3000 Time Figure 2-32: Typical Structure-to-Soil Potential Recording with Time Caused by Interference from a DC Transit System The potential-time recording of stray current effects from a DC transit system has a distinctive pattern. There are considerable potential fluctuations during the morning and evening rush-hour periods, light activity in the middle of the day and late evening, and virtually no changes during the early morning hours. Although the stray current pick-up locations change with time, the discharge sites are predominantly in proximity to the substation ground. In urban areas, localized stray current can discharge from water piping around electrically discontinuous joints and from crossings with other utilities remote from the substation ground. Determining the impact of transit-caused stray current on metallic facilities in urban areas requires considerable potential and current recording, starting in the vicinity of the substation grounds and along the transit system route. CP Interference © NACE International, 2006 January 2008 10:00 9:00 8:00 7:00 6:00 5:00 4:00 3:00 2:00 1:00 0:00 23:00 22:00 21:00 20:00 19:00 18:00 17:00 16:00 15:00 14:00 13:00 12:00 11:00 -3500 10:00 Potential wrt CSE (mV) -500 DC Interference 2.2.2(a)(i) 2:44 Analysis of Transit System Stray Currents A comprehensive method of analyzing dynamic stray current activity involves the construction of beta curves from current and potential measurements. A line current survey can be conducted to determine the magnitude and direction of the currents flowing along the pipeline, providing a way to locate the source of the interference. However, this requires that some means be available to measure pipeline currents at a number of locations throughout the area of interest (which is not always possible). In cases where the pipe is exposed or rises above grade, this can be done using a pipeline current clamp (Figure 2-33). Otherwise, as Figure 2- 4 shows, IR-drop test stations must be used. Figure 2-33: Current Clamp Used to Measure Pipeline Currents CP Interference © NACE International, 2006 January 2008 DC Interference 2:45 Substation + Positive Rail Load Negative Return Rail IA ID IB IC IE + V - + V - + V - + V - + V - A B C D E Figure 2-34: Line Current Survey to Locate Source of Interference Using IR-Drop Test Stations A line current survey is conducted by taking a series of simultaneous pipeline current measurements at adjacent locations on the pipeline, and plotting these measurements with respect to one another. The measurements may be conducted simultaneously using either global positioning system (GPS) synchronized data loggers or manually using a two-person crew who communicate by radio. The best way to illustrate the procedure is through the following example. In Figure 2-34, a set of pipeline current measurements are made at the IR-drop test station at location “A”; a second set of current measurements is simultaneously recorded at Location “B.” The measurements obtained at “B” are plotted against those obtained at “A” in Figure 2-35, and it is found that the relationship is linear and the data produce a line having a slope of greater than unity. Because the relationship is linear, the currents measured at “A” and “B” must emanate from the same source of interference. Also, because the slope is greater than 1, more line current exists at location “B” than at location “A”; the pipeline must therefore be picking up interference current in this area. Simultaneous measurements are also taken at locations “B” and “C,” producing the second chart in Figure 2-35. Here, the plot is once again linear. However, the slope of the line is unity. This indicates that there is no net pick-up or discharge of interference current between these two locations. CP Interference © NACE International, 2006 January 2008 DC Interference 2:46 In the third chart of Figure 2-35, the slope of the line is less than unity; this indicates that interference current is discharging from the pipeline and that mitigative measures must be taken in this area. Note that in all three of these charts, the plots do not pass through the origin. This is an indicator that there are other currents also flowing along the pipeline that are unrelated to the interference currents, such as CP current. In the fourth chart of Figure 2-35, there is no correlation between the currents measured at location “D” and those measured at location “E.” This indicates that the line currents at location “E” must emanate from some other source of interference. This other source could simply be another load somewhere else along the transit system, or it could be a source that is totally unrelated to the transit system. IC IB ΔIB ID IE ΔIC ΔIB ΔIA ΔIC ΔID IA IB IC ΔIB >1 ΔIA ΔIC =1 ΔIB ΔID <1 ΔIC ID Non-Linear Figure 2-35: Line Current Plots for Example in Figure 2-34 A second type of survey that can be conducted is the exposure survey, where pipeto-soil potential measurements are recorded simultaneously with pipeline currents (Figure 2-36). At each location, current is plotted versus potential to determine the point of maximum discharge. CP Interference © NACE International, 2006 January 2008 DC Interference 2:47 Substation Positive Rail + Load Negative Return Rail IA - V + A IB V - - V + V - B ID IC - V + C V - - V + D V - IE - V + V - E Figure 2-36: Exposure Survey to Locate Point of Maximum Exposure As an example, the first two charts in Figure 2-37 show that as the current flowing along the pipe increases, the potential of the earth becomes more positive with respect to the pipeline (i.e., pipe potential becomes more negative). This indicates areas of current pick-up. Because the slope of the plot is greater at “A” than at “B” (i.e., potential variations are greater per unit of interference current), location “A” is considered to be the point of maximum current pick-up. At location “C,” the pipe potential is unaffected by the interference current; consequently, there is neither pickup nor discharge in this area. This location should also correspond to the point of maximum current flow along the pipeline. At location “D,” the earth potential becomes more electronegative (i.e., the pipe potential becomes more electropositive) as the current flowing along the pipeline increases; this indicates a point of current discharge. Although it has a positive slope, Location “E” is also a point of current discharge because the direction of current flow along the pipeline is in the opposite direction to that at locations “A” through “D.” Because the absolute value of the slope of the plot at “D” is greater CP Interference © NACE International, 2006 January 2008 DC Interference 2:48 than that at “E,” Location “D” is the point of maximum exposure—where mitigative measures must be taken. Vp/s Vp/s ΔI ΔVp/s ΔI α Location A I I Location B Vp/s ΔVp/s Δ Vp/s =0 ΔI ΔVp/s α Vp/s Vp/s ΔI I α α Location D Location C I α Location E Figure 2-37: Exposure Survey Plots for Example in Figure 2-36 A third type of survey that can be conducted is the mutual survey, which does not involve the measurement of pipeline currents. Voltages are measured between the pipeline and the interfering system; simultaneously, pipe-to-soil potentials are measured at the point of maximum exposure (Figure 2-38). The pipe-to-soil potentials Ep/s are plotted versus the pipe-to-rail potentials Ep/r. If a correlation exists, as shown in Figure 2-39, then the source of interference has been positively identified. Substation Positive Rail + Load Negative Return Rail V + - V + Figure 2-38: Mutual Survey to Confirm Source of Interference CP Interference © NACE International, 2006 January 2008 ΔI I ΔVp/s DC Interference 2:49 Vp/s ΔVp/s ΔVp/r α Vp/r Figure 2-39: Pipe-to-Soil Potential Versus Pipe-to-Rail Potential for Example in Figure 2-38 Plots that relate pipe-to-soil potential to pipe-to-rail potential generally are called beta curves because the slope of the linear plot is called beta. The equation of the line is as follows: Vp/ s = α + ΔV p / s ΔV p / r V p / r = α + βV p / r [2-35] Because the pipe-to-soil potential in this example is a linear function of the pipeto-rail potential (Figure 2-39), and because the pipe-to-soil potential is also a linear function of the pipeline current (Figure 2-37), it follows that pipeline current is a linear function of the pipe-to-rail potential. Therefore, it has been argued that an exposure survey may be conducted without measuring pipeline currents at all. This line of reasoning calls for simply recording pipe-to-soil potentials at various locations along the pipeline (Figure 2-40) and plotting these versus pipe-to-rail potential (Figure 2-41). CP Interference © NACE International, 2006 January 2008 DC Interference 2:50 Substation Positive Rail + Load Negative Return Rail V + - V + - A V + - B V + C Figure 2-40: Exposure Survey Conducted Without the Measurement of Pipeline Currents Vp/s Vp/s Vp/s Δ Vp/s =0 Δ Vp/r ΔVp/s ΔVp/r Vp/r α α Vp/r ΔVp/r α ΔVp/s Location B Location A Location C Figure 2-41: Exposure Survey Plots for Example in Figure 2-40 In general, the steeper the slope of the beta curve, the greater the pick-up or discharge. However, the polarity of the slope will depend upon the point of connection for the voltmeter measuring the pipe-to-rail potential. CP Interference © NACE International, 2006 January 2008 Vp/r DC Interference 2.2.2(a)(ii) 2:51 Mitigation of Transit System Stray Currents Mitigation methods for minimizing the deleterious effects of DC transit system stray currents are similar to those used for ameliorating CP stray currents. They include: • • • • • electrical isolation of rails and substation electrical bonds reverse current switches forced drainage bonds CP. On existing transit systems, stray current has been reduced significantly by improving the isolation between the rail and ballast. This is accomplished by installing insulating pads between the rail and ties, between the hold-down plates and the rail, and ensuring that the ballast is well drained. These measures, coupled with disconnecting the negative rails from electrical grounds, have proved relatively successful in many instances. Disconnecting the DC substation from electrical ground allows the rails and transit vehicles to electrically float in a manner that requires the installation of switching devices that connect the rails to earth if a specific rail voltage-to-ground potential is exceeded. The effectiveness of substation isolation in minimizing stray current activity is therefore lost during the time that the safety switches are activated. For new transit systems, it has become common to electrically isolate the entire rail pocket if the rail is embedded in the road surface (Figure 2-42a) or isolate the rail from ties (Figure 2-42b). CP Interference © NACE International, 2006 January 2008 DC Interference 2:52 Polyurethane Sealant Flangeway Rail Elastomer Pad 36 Mils of Coal Tar Epoxy Rail Clip Concrete Tie Third Pour Fiber Concrete Concrete Invert Second Pour Fiber Concrete 115 RE Rail Preformed Rail Trough 30 Mil PVCl Sheet on Bottom of Trough 3-20 Mil Polyethylene Sheets Figure 2-42b: Typical Direct-Fixation Isolating Fastener First Pour Concrete Slab Structurally Reinforced Source: Fitzgerald J.H. and Lauber, M.D., Stray Current Control for the St. Louis Metrolink Rail System, MP, Vol. 34(1), Jan. 1995, p.22 Figure 2-42a: Typical Embedded Track Installation Source: Sidoriak, W., Rail Isolation on the Baltimore Central Light Rail Line, MP, Vol. 32(7), July 1993, p.36 Some transit systems use a separate isolated rail (so-called fourth rail) as a current return path, which negates the need to isolate the running rails. The earliest attempts to mitigate the corrosive effects of transit stray currents simply involved running bonds from the utilities to the negative bus at each substation. This provided an electronic path for the stray current to return, thus reducing the amount of stray current in the electrolytic path as shown in Figure 2-43. positive bus to 3rd rail rails DC substation negative bus shunt Is,2 Is,1 metallic structures Figure 2-43: Typical Utilities Drainage System at a Transit Substation Facilities such as lead-sheathed power cables, steel gas piping, telephone grounds, and iron water piping would be connected in series with a switch and a shunt to the CP Interference © NACE International, 2006 January 2008 DC Interference 2:53 negative bus. The shunt provides a means of recording the stray current magnitude and direction. One weakness of this drainage arrangement is that providing a direct low resistance path for the stray current causes the underground structures to pick up more current than they would otherwise. For structures with electrical discontinuities, such as iron watermains, this can result in more severe corrosion at the isolating joints. A second disadvantage of the direct bond drainage system becomes evident where there are multiple substations and many trains. The utilities represent an alternative path to the rails between substations, and the stray currents can actually reverse. This situation is depicted in Figure 2-44. bus bus SS 'A' I''L,B bus 3rd rail IL,A I'L,A + I''L,B I''L,B I''L,A + IS,A IL,B L IL,A '' + I'L,B running rails I''L,B + IS,B utilities bus SS 'B' I''L,A I''L,A Figure 2-44 Schematic Showing Circulating Current between Transit Substations Through Direct Bonds to Utilities With the transit load located between substations “A” and “B,” it will draw some of the load current from each station. Hence, each substation’s load current has an alternative path through the utility bonds back to its respective source. To prevent circulating currents, reverse current switches can be installed in each bond. These devices present a high resistance in one direction (the reverse direction) and a low resistance in the other (direction of intended drainage). There are several types of reverse current switches,[6] as listed in Table 2-2, each with differing operational characteristics. 6 Munro, J. I., Comparison and Optimization of Reverse Current Switches, NACE, Corrosion/80, Paper No. 142, March 1980. CP Interference © NACE International, 2006 January 2008 DC Interference 2:54 Table 2-2: Types of Reverse Current Switches Type Characteristics Electromagnetic (relay) Requires AC power to operate the relay, relay must conduct all current, may be slow to open Diodes (germanium, silicon) Requires a minimum of 0.4V to conduct, have resistance, subject to surge failures and reverse voltage breakdown Hybrid (relay in parallel with diodes) Smaller relay required because diodes carry most current and are subject to reverse voltage breakdown. Potential Controlled Rectifier (Figure 2-40) Can drain all the stray current but are relatively expensive. Although CP is beneficial in mitigating transit system stray current, the stray currents are often so large that they preclude mitigation with galvanic anodes. Moreover, large-capacity impressed current systems in an urban area will likely create interference on other facilities. CP thus has limited effectiveness. One the most successful measures is the use of a forced drainage bond. As shown in Figure 2-45, a forced drainage bond is a bond with a potential-controlled rectifier connected in series with the bond. Is Potential Controlled Rectifier Is structure buried reference electrode Figure 2-45: Forced Drainage Bond Using a Potential Controlled Rectifier The voltage output of the auto-potential rectifier varies depending on the potential measured between the structure and a buried reference electrode. If the measured potential is more positive than the potential set on the controller, the rectifier output CP Interference © NACE International, 2006 January 2008 DC Interference 2:55 voltage increases to force more current through the bond. With a DC voltage source in series with the bond, the bond resistance is negative. The negative resistance ensures that all the stray current is drained from the structure and there is no residual stray current in the soil path. Nevertheless, the controller must be adjusted so that there is no bond current during periods of no stray current activity. Otherwise, the transit system rails and grounding system will be corroded. To be completely effective, the forced drainage bond must be located at the point of maximum discharge. Just as with a resistance bond, if the structure is not electrically continuous then a forced drainage system will aggravate corrosion at any isolating joints. 2.2.2(b) High Voltage Direct Current (HVDC) Electrical Transmission Systems HVDC systems that transmit large blocks of electrical power over long distances have operating cost advantages over high voltage alternating current (HVAC) transmission. Unlike HVAC systems, there are no inductive or capacitive losses on HVDC. Moreover, for lengths greater than approximately 800 km, the power savings easily justify the extra capital costs to build the AC/DC converter stations and their extensive electrical grounding systems. HVDC systems are built to operate in bi-polar mode; that is, there is both a positive and negative circuit with large grounding electrodes at each terminus as illustrated in Figure 2-46. Idc positive cables load end AC / DC negative Converters supply end cables Idc L > 800 km Figure 2-46: Electrical Schematic for a HVDC System Under normal operating conditions, the DC line currents are typically in the 1000A range and imbalance currents are approximately 1 to 2% of the line currents. Such small currents do not pose a significant stray current risk on CP Interference © NACE International, 2006 January 2008 DC Interference 2:56 underground metallic structures because the electrodes are intentionally located remotely from other utilities. During emergency operating conditions, where either the positive or negative cable networks are faulted or de-energized for maintenance, the line current passes through the earth via the grounding electrodes. Under these circumstances, the system is operating in monopolar mode. HVDC grounding electrodes are large compared to impressed current groundbeds, although CP anode materials such as high-silicon iron and coke are often used. The electrode is typically in the shape of a ring having approximately 100m diameter and a depth of 1 to 2m. Despite the large size and relative remoteness, the voltage gradient around the electrode can be appreciable even a long distance away when the electrode is passing hundreds of amperes. For example, the voltage rise in earth at some distance “x” from such an electrode can be estimated using Equation 2-36. Vg, x = Ie ρs 2πx [2-36] where: Vg,x = voltage rise with respect to remote earth at a distance “x” from the electrode Ie = electrode current ρs = soil resistivity x = distance from the electrode given: then: Ie = 500A ρs = 50 Ω-m r = 1000m Vg, x @ 1 km = 500A × 50 Ω - m 6.28 × 1000 m ≈ 4V Hence a metallic structure located 1 km from the electrode would be exposed to 4V during monopolar operation under the foregoing conditions. It is claimed that the HVDC system will operate in monopolar mode a small percentage of time. Nevertheless, the rather large voltage gradients can present a serious corrosion risk on some structures on a cumulative basis. CP Interference © NACE International, 2006 January 2008 DC Interference 2:57 Also, the effect can be either a positive or negative potential shift on the structure (Figure 2-47) depending on which of the power circuits has the outage. +1.0 0.0 E s/s (Vcse ) + E -1.0 - E -2.0 -3.0 -4.0 t1 t3 t2 t4 Time Figure 2-47: Potential-Time Plot for a Metallic Structure being Interfered-with by a HVDC System Assume that the potential plot in this figure was from a structure located near the supply end groundbed. Under this scenario, the negative shift from t1 Æ t2 would result from a failure on the positive circuit and the positive shift from t3 Æ t4 would result from a failure on the negative circuit. Note that the potential shifts are not necessarily equal—even if the stray current is the same—because the cathodic and anodic polarization characteristic can be different. Most structures would not extend the full 800km, nor be close enough to the electrode to make it economical to install a bond. Because of the large voltage shifts, galvanic anodes many not adequately compensate. The most practical mitigation method is to use an impressed current system powered by a potential controlled rectifier. Not only would the CP power supply be able to counteract the large positive potential shifts, but during the negative shift periods it would shut down—thus minimizing the stress on the coating if the structure was a coated steel pipeline. 2.2.2(c) DC Welding Operations Welding operations on ships and barges have been known to create stray current interference, sometimes so severe that it has resulted in the sinking of the vessel. Interference arises where the negative of the welding generator is connected to CP Interference © NACE International, 2006 January 2008 DC Interference 2:58 electrical ground on the dock and there is no electrical bond between the dock and the vessel. Under these circumstances, the welding current (which can be hundreds of amperes) discharges from the vessel to the dock as illustrated in Figure 2-48. DC welding generator Is sheet steel piling Figure 2-48: Stray Current Caused by DC Welding Operations The interference is mitigated by bonding the vessel to the dock or by attaching the negative of the welding generator directly to the vessel. Experiment 2-1: To Demonstrate DC Interference and Its Mitigation CP Interference © NACE International, 2006 January 2008 DC Interference 2:59 Experiment 2-1 To Demonstrate DC Interference and Its Mitigation 3 1 cm steel rod 1 cm 2 mag anode 4 5 6 1 A 9V 10 ohm Experiment Schematic No. 1 Procedure Step: A. Place bare steel rod along one end of the tub in 5cm of water obtained from the cold water tap. Connect the 9V battery, 10 Ω resistor, ammeter, and switch in series between the steel rod and magnesium anode. Close the switch and allow the CP system to operate for a minimum of 5 minutes. B. Measure and record the potential on the steel rod with the reference positioned at locations 1, 2, and 3. Record the current. C. Open the switch and insert the second steel rod (foreign structure) perpendicular to the first steel rod at Location 2. D. Measure and record the foreign structure potential at reference locations 4, 5, and 6 with the switch remaining open. CP Interference © NACE International, 2006 January 2008 DC Interference 2:60 Experiment 2-1, cont’d E. Close the switch and allow the CP system to operate for a minimum of 5 minutes. F. Measure and record structure potentials at all reference locations on both structures and record the CP current. G. Calculate the shift in potential at locations 4, 5, and 6 on the foreign structure and the change in CP current. Discussion Break H. Mitigate interference using a resistance bond connected between the cathodically protected structure and the foreign structure as in Schematic No. 2. Adjust the resistance bond until the foreign structure potential at Location 4 is equal to or more negative than its native potential. resistance bond A 3 steel rod 2 mag anode 4 5 6 1 A 9V 10 ohm Experiment Schematic No. 2 I. Measure and record potentials on both structures, at all reference locations, measure CP and mitigation current, record bond resistance. CP Interference © NACE International, 2006 January 2008 DC Interference 2:61 Discussion Break Experiment 2-1, cont’d J. Disconnect the resistance bond and install a galvanic anode to mitigate the interference as shown in schematic No. 3. A mag anode 3 steel rod 2 mag anode 4 5 6 1 A 9V Experiment Schematic No. 3 K. Measure and record all structure potentials, CP current, and galvanic interference current (Igalv.). Discussion Break CP Interference © NACE International, 2006 January 2008 DC Interference 2:62 Results Icp B Structure Potentials (mVCSE) CP’d Structure Foreign Structure 1 2 3 4 5 6 X X X D X X STEP X X F G CP’d structure only Foreign structure only Both structures X X X Shift calculations Mitigation I Ib = _______ Rb = ______ ohm K Igalv. ______ CP Interference © NACE International, 2006 January 2008 DC Interference 2:63 Experiment 2-1, cont’d Conclusions 1. Foreign structure potential shifts electropositively at the stray current discharge location (#4). 2. Foreign structure potential shifts electronegatively at the stray current pick-up location (#6). 3. The CP current distribution on the cathodically protected structure is affected by the presence of the foreign structure. 4. The CP current increases when the foreign structure is present. 5. The resistance bond mitigates the stray current interference on the foreign structure. 6. The CP current increases with the resistance bond inserted, but the cathodically protected structure is less well-protected. 7. A galvanic anode mitigation system can mitigate the interference problem and maintain protection on the cathodically protected structure. 8. The stray current magnitude is greater for the resistance bond than for the galvanic mitigation system. CP Interference © NACE International, 2006 January 2008 DC Interference 2:64 2.2.1 Case Study Two coated, cathodically protected pipelines cross each other at right angles and are separated by 0.2 m. A single AWG #10 test lead is connected to each pipeline and is routed into a common test station. Pipe-to-soil potential data is measured independently on each pipeline while the other pipeline’s CP system is interrupted. The recorded data are as follows: A’s TR Pipeline A B ON -1100 -930 B’s TR OFF -905 -980 ON -1010 -980 OFF -1100 -870 Does this situation require mitigation, or is further investigation required? Please explain. Assume mitigation is required and described how you would mitigate this situation assuming a soil resistivity of 3500 Ω-cm. Describe how you would test the piping after mitigation to determine if the piping is adequately protected. CP Interference Course Manual © NACE International, 2006 January 2008 DC Interference 2:65 Summary of Equations [2-1] Rt,i = Ri,e + Ri,p [2-2] Ii = where: R t ,n R t ,i page 2:2 It page 2:2 Rt,n = the total resistance of n parallel paths 1 = R t ,n 1 1 1 1 + + + R1 R2 R3 R4 ⋅⋅⋅ 1 Rn and: It = I1 + I2 + I3 + I4 … In R i,e = ρ s [2-3] L A x,s page 2:3 where: Ri,e ρs L Ax,s = = = = resistance of the current path (ohm) resistivity of the soil length of current path cross-sectional area of soil path ρm = ρs [2-4] substituting: ρm = 10 -12 ρs then: A x,s = CP Interference Course Manual © NACE International, 2006 January 2008 A x,m 10 -12 A x,m A x,s page 2:4 DC Interference 2:66 A x,s = 10 -2 10 -12 = 1010 m 2 Vx,re = [2-5] I ρ s ⎡ ⎛⎜ L + ⎢ln 2πL ⎢ ⎜⎝ ⎣ L2 + x 2 x ⎞⎤ ⎟⎥ ⎟⎥ ⎠⎦ page 2:6 where: Vx,re = voltage rise in earth with respect to remote earth at a distance “x” from the anode I = anode current output ρs = soil resistivity L = length of anode Rv [2-6] = ⎧⎛ 8 L ⎞ ⎫ ρ 2L ln (0.656N) ⎬ ⎟ −1+ ⎨⎜ ln 2 π NL ⎩⎝ d ⎠ s ⎭ page 2:8 where: Rv ρ L d = = = = resistance of multiple vertical anodes to remote earth (Ω) soil resistivity (Ω-cm) length of anode (cm) diameter of anode (cm) s = anode spacing (cm) N = number of anodes Rh [2-7] where: [2-8] CP Interference © NACE International, 2006 January 2008 = ρ 2πr page 2:9 ρ = resistivity: Ω-m r = radius of hemispherical electrode (m) Rh = resistance to remote earth (Ω) r = ρ 2π Rh page 2:9 DC Interference 2:67 R [2-9] R s,e [2-10] ρ L d t where: = = = = ⎛1 1⎞ ⎜⎜ − ⎟⎟ r1 ⎠ ⎝r = ρ 2π = ⎧ (L )2 ⎫ ρ ln ⎨ ⎬ 2πL ⎩ td ⎭ R s,re (discharge) [2-12] = page 2:14 R S,L × R S,R R S,L + R S,R RS,O = RG coth αx [2-13] page 2:11 soil resistivity length of pipe diameter of pipe depth below grade Rs,e = Rs,c + Rs,re [2-11] page 2:10 page 2:15 page 2:15 where: RG = characteristic resistance α = attenuation constant x = length of pipe [2-14] α = where: Rm RL page 2:15 Rm = lineal resistance of the pipe RL = leakage resistance of pipe to earth [2-15] R m = ρm × where: Rm ρm L Ax = = = = lineal resistance of the pipe resistivity of pipe material length of pipe section cross-sectional area of pipe CP Interference © NACE International, 2006 January 2008 L Ax page 2:15 DC Interference [2-16] 2:68 RL = and where: RL r′c L AS = = = = rC′ AS page 2:16 leakage resistance of pipe to earth specific resistance of pipe coating length of pipe section surface area of section = πdL R S,O = R G = [2-17] Rm × RL = 2.29 × 10 -4 Ω × 383 Ω = 8.77 × 10 -2 Ω 2 page 2:17 = 0.296 Ω Ra = [2-18] where: Ra ρ L d = = = = ρ ⎧⎛ 8 L ⎞ ⎫ ⎨⎜ ln ⎟ − 1⎬ 2πL ⎩⎝ d ⎠ ⎭ resistance of anode to remote earth soil resistivity (Ω-m) = 31 Ω-m length of packaged anode (m) = 1.5 m diameter of packaged anode (m) = 0.15m L = [2-19] where: L Wt U E Ia Cr [2-20] CP Interference © NACE International, 2006 January 2008 = = = = = = page 2:31 Wt × U × E I a × Cr page 2:32 effective service life (y) total weight of anode alloy (kg) utilization factor efficiency current output (A) theoretical consumption rate (kg/A-y) RR1 = V1 (off , oc ) − V1 (on, oc ) IR page 2:37 DC Interference 2:69 RR 2 = [2-21] V2 (on, oc ) − V2 (off , oc ) IR page 2:37 RR1, 2 = RR1 + RR 2 page 2:37 [2-22] [2-23] RB1 = V1 (on, cc ) − V1 (on, oc ) IB page 2:37 [2-24] RB 2 = V2 (on, oc ) − V2 (on, cc ) IB page 2:37 [2-25] RB1, 2 = RB1 + RB 2 page 2:37 [2-26] V1, 2 (off , oc ) = V2 (off , oc ) − V1 (off , oc ) page 2:37 [2-27] RB = V1, 2 (off , oc ) + I R RR1, 2 − RB1, 2 RR 2 IR RB 2 I B (on ) = [2-28] I B (off ) = [2-29] I R RR 2 RB 2 V1, 2 (off , oc ) RB + RB1, 2 page 2:38 page 2:38 page 2:38 [2-30] V1 (on, cc) = V1 (on, oc) + I B (on) RB1 page 2:38 [2-31] V1 (off , cc) = V1 (off , oc) + I B (off ) RB1 page 2:38 [2-32] V2 (on, cc) = V2 (on, oc) − I B (on) RB 2 page 2:38 [2-33] V2 (off , cc) = V2 (off , oc) − I B (off ) RB 2 page 2:38 CP Interference © NACE International, 2006 January 2008 DC Interference 2:70 IL = IR + Is + Ie [2-34] [2-35] Vp/ s = α + [2-36] CP Interference © NACE International, 2006 January 2008 ΔV p / s ΔV p / r Vg, x = V p / r = α + βV p / r Ie ρs 2πx page 2:42 page 2:49 page 2:56 CHAPTER 3 AC INTERFERENCE 3.1 Introduction Electrical energy from an overhead powerline can be transferred to a pipeline by three possible mechanisms: electrostatic (capacitive) coupling, electromagnetic (inductive) coupling, and conductive (resistive) coupling. The latter occurs only during fault conditions. Each mechanism is discussed with respect to how it affects pipeline integrity, along with the safety of pipeline personnel and the general public. The methods that are available to predict both the effects of the interference and the required mitigative measures are also discussed, as are the methods for implementing the mitigation. Note that the prediction of alternating current (AC) interference effects is a complex matter requiring fairly sophisticated mathematics. This course discusses methods of estimating the effects for a few very simple cases, but most problems can only be solved using either complicated analytical techniques or specialized software. A pipeline can experience AC interference as a result of being in the proximity of any AC powerline. However, the vast majority of interference problems are created by three-phase (3φ) power transmission systems (Figure 3-1a), because these involve both high currents (during steady-state and fault conditions) and high voltages. Moreover, these system are more likely to run parallel to pipelines for long distances than—for instance—low-voltage distribution systems (Figure 31b). A 3φ power transmission system consists of three energized conductors. Each conductor has approximately the same voltage to ground, and each carries approximately the same amount of current. One or two additional conductors, known as shield wires, may also be present. Shield wires run between the tops of the powerline support structures (Figure 3-1). Although their purpose is to protect the powerline from lightning strikes rather than to transmit power, they nevertheless affect how electrical energy is transferred to a pipeline. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:2 Figure 3-1a: Single Horizontal 3φ Circuit with Shield Wires Figure 3-1b: Distribution System (1φ 4kV Primary and 2φ 240V Secondary with Neutral) In a three-phase circuit, the AC waveforms for each of the three phases are 120 degrees apart from one another (Figure 3-2). Waveforms that have the same frequency but start and end at different times are said to be out-of-phase with one another. When investigating the effects of AC interference on a pipeline, phase relationships between waveforms are just as important as the magnitudes of the waveforms; this will be discussed later in the chapter. 1 V 0 -1 0 90 180 270 360 Angle (Degrees) Figure 3-2: AC Voltage Waveforms in a 3φ Circuit 3.1.1 Electrostatic (Capacitive) Coupling With electrostatic coupling, energy is transferred through the electrical capacitance that exists between the powerline and the pipeline. Any two conductors that are separated by a dielectric material can be considered a capacitor. Capacitance is a measure of the ability to store electrical charge Q between two conductors, relative to the voltage V between the conductors; that is: CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:3 C= Q V coulombs/volt [3-1] The unit of coulombs/volt is more commonly referred to as a farad (f). Capacitance is proportional to the area A of the conductors but is inversely proportional to the separation d between the conductors (Figure 3-3). Furthermore, capacitance is directly dependent upon a physical property of the dielectric material known as permittivity (ε), having the units of f/m. Therefore, C=ε C∝ A d [3-2] A d Conducting Plate (having area A) d Dielectric Conducting Plate Figure 3-3: Elements of a Capacitor When a direct current (DC) voltage source is applied to a capacitor, current will flow and charges will accumulate on the plates of the capacitor. As time passes and charges continue to accumulate, the current flow decreases and eventually becomes zero when the voltage on the capacitor is equal to the applied voltage. This time period is very short; for all practical purposes, a capacitor appears as an open circuit to DC. When an AC voltage source is applied to a capacitor, current begins to flow and the conducting plates again begin to accumulate charges. As the polarity of the voltage source reverses during the second half of the AC cycle and current flows in the opposite direction, however, the plates of the capacitor discharge and begin charging with the opposite polarity. This process of charging, discharging, charging in the opposite direction, and discharging again repeats itself every cycle, and an AC continually flows through the capacitor. As the frequency of the voltage source increases, fewer charges can accumulate on the capacitor’s plates. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:4 As a result, there is less opposition to the flow of current. frequencies, a capacitor therefore appears as a short circuit to AC. At very high The opposition that a capacitor offers to the flow of AC is called capacitative reactance, Xc. Reactance has the units of ohms (Ω) and is similar to resistance— except that it not only controls the magnitude of the current flowing in the circuit, but also affects the phase relationship between the voltage and the current (Section 3.2.1 discusses this). Reactance is dependent upon both frequency f and capacitance and is determined by the following equation: XC = 1 2πfC [3-3] Consider the case in Figure 3-4 where a pipeline is under construction. Lengths of pipe have been strung out along the pipeline route and have been placed on wooden skids in preparation for welding. Although this may not look like a capacitor as previously discussed, the elements necessary for the construction of a capacitor are present; these elements include two conductive plates separated by a dielectric material. In this case, the powerline is one conductive plate and the pipe is another. They are separated by air, which serves as a dielectric. Similarly, a second capacitor is formed between the pipe and the earth because the earth (although nonmetallic) is also a conductive plate. A section of pipe sitting on skids beneath an AC powerline can therefore be represented as an electrical circuit consisting of two capacitors in series with an AC source, which forms a capacitive voltage divider. Conducting Plates Air Dielectric Figure 3-4: Electrostatic Coupling during Pipeline Construction CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:5 Recalling Kirchhoff’s Laws, the sum of the voltage drops across the resistors in a series circuit (Figure 3-5) will be equal to the sum of the voltage sources. Furthermore, these voltage drops are in direct proportion to the resistances that create them. Similarly, the voltage drops across the capacitors in an AC series circuit will be in direct proportion to the respective capacitive reactances; their sum will be equal to the sum of the voltage sources. I R1 I V1 V V1 C1 V R2 V1 V2 = V2 V2 C2 R1 V1 R2 V2 = XC1 XC2 = C2 C1 Figure 3-5: Voltage Divider Circuits – Resistive (left) and Capacitive (right) Therefore, in the pipeline construction case of Figure 3-4, the line-to-ground voltage of the powerline is divided between the two capacitors in inverse proportion to their capacitances. Depending upon the relative capacitance values and the powerline voltage, very large voltages can be electrostatically generated on a single pipe joint—assuming it is well insulated from earth. To provide a very rough estimate of the magnitude of the induced voltages, consider the case of a single pipe section raised upon on skids (Figure 3-6). Example Calculation: The pipe has a diameter of 0.3m and is 5m in length. It therefore has a surface area of approximately 5m2. The powerline conductor has a much smaller diameter than that of the pipe. Hence, it has a smaller surface area while the underlaying earth has a greater surface area than the pipe. Assume that these areas are 0.2m2 and 20m2, respectively. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:6 0.2 m2 100 kV 10 m C1 5 m2 C2 1m 20 m2 Figure 3-6: Calculation of Typical Capacitance Values for a Pipe on Skids The powerline conductor and the pipe, which form the two places of C1, have different areas. So, in order to use Equation 3-2, the geometric mean of the two areas is used. Similarly, the pipe and the earth lying beneath it have different areas. So, the geometric mean is calculated to determine the area of the plates for C2. Amean = A1 × A2 [3-4] AC1 = 0.2m 2 × 5m 2 = 1m 2 AC 2 = 5m 2 × 20m 2 = 10m 2 The separation distance between the powerline conductor and the pipe is typically much greater than between the pipe and the earth. These distances are given as being 10m and 1m, respectively. The values of C1 and C2 can now be calculated, given the permittivity of air (εair) has a value of 9×10-12 f/m. C = ε air C1 = 9 × 10 −12 C 2 = 9 × 10 CP Interference Course Manual © NACE International, 2006 January 2008 −12 A d 1m 2 f /m = 9 × 10 −13 f 10m 10m 2 f /m = 9 × 10 −11 f 1m AC Interference 3:7 Again, note that these capacitance values are very rough estimates only; although accurate capacitance values could be calculated, such a task would involve much more complicated equations. It is only important to understand that the pipe-to-earth capacitance should always be larger than the pipe-topowerline capacitance and that these capacitance values are typically very small. For instance, a capacitor used on an electronic circuit board (having a physical size similar to that of a pencil eraser) might have a capacitance of 10 × 10-6 farads (10 μf)—yet this capacitance would be roughly a million times larger than the pipe capacitances calculated above. In order to calculate the voltage that is electrostatically induced on the pipe in Figure 3-6, the values of C1 and C2 are substituted into the capacitive voltage divider circuit of Figure 3-5. In Figure 3-5, the voltage applied across the capacitors is the line-to-ground voltage of the powerline; it is given as 100 kV in this example. V pipe = V pipe = C1 V powerline C1 +C 2 [3-5] 0.9 × 10 −12 100 × 10 3 V = 1000V 0.9 × 10 −12 + 90 × 10 −12 V = 100 kV 0.9 pF V = 1 kV 90 pF Figure 3-7: Calculation of Typical Electrostatically Induced Voltage for a Pipe on Skids CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:8 The pipe voltage in the example seems unrealistically high. However, this is indeed the magnitude of voltage that might be typically induced electrostatically on a single pipe joint—provided that the pipe is well-insulated from earth and that the voltage is measured with a high-impedance voltmeter. In order to determine if this voltage may present an electrical safety hazard, it is necessary to calculate the current that could possibly be generated by this circuit. The current that can be produced is limited by the reactance of the powerline-topipe capacitance (C1). The reactance is calculated using Equation 3-3. X C1 = 1 1 = = 3 × 10 9 Ω −13 2πfC1 2π ⋅ 60 ⋅ 9 × 10 The current that can flow through a human body, assuming the worst-case of a zero-ohm body resistance, is then determined using a calculation that is essentially Ohm’s Law. I body = V powerline X C1 = 100 × 10 3 V ≈ 30 μ A 3 × 10 9 Ω Vpowerline = 100 kV XC1 = 3 GΩ I = 30 μA Figure 3-8: Calculation of Typical Shock Current Resulting from Electrostatic Coupling Such a low current is considered non-hazardous. It is in fact well below the 1-mA threshold at which the human body can sense electric current (this is discussed in Section 3.4.1). Therefore, even though electrostatic coupling can induce large voltages on sections of pipe that are well-insulated from ground, the circuit impedance is generally too high to produce a significant shock current. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:9 Consider the case where an automobile is parked beneath a high-voltage power transmission line (Figure 3-9). Because the car is well-insulated from earth by the rubber tires, the situation is very similar to the one illustrated in Figure 3-8. If the high voltages generated by electrostatic coupling were capable of presenting an electrical shock hazard, then consider the problems this scenario might cause to the public. V = 100 kV C1 = 0.01C 2 V = 1 kV C2 Figure 3-9: Calculation of Typical Electrostatically Induced Voltage for an Automobile The sample calculations for a pipe raised up on skids were for the case of a single pipe joint. As the pipe joints become welded together, the surface area of the pipe increases and the pipe-to-powerline capacitance increases accordingly. This results in a lower capacitive reactance between the pipeline and powerline, which will permit more current to flow through the body (Figure 3-8). However, as the pipeline increase in length, two other factors become important. Firstly, the amount of energy being electromagnetically induced in the pipeline becomes significant—more significant, in fact, than the electrostatically induced energy (see Section 3.1.2). Secondly, as the pipe increases in length, the total resistance between the pipe and earth through the increasing number of skids decreases. Therefore, the voltage generated across C2 decreases (a bar pipe on skids would have an even lower electrostatically induced voltage). Most importantly, however, is that as the pipe joints are welded together the construction crews begin to lower the pipe into the trench. This not only results in an even lower pipe-to-earth resistance, but also results in a much higher pipe-to-earth capacitance (Figure 310). CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:10 3 x 109 3 x 104 Figure 3-10: Calculation of Typical Electrostatically Induced Voltage for a Buried Pipe Considering only the effect of the increased pipe-to-earth capacitance, a decrease of the pipe-to-earth separation from 1m to 1mm (the thickness of the coating) would result in a 1000 times increase in the capacitance C2, a 1000 times reduction in the capacitive reactance XC2, and a 1000 times reduction in the pipe voltage Vpipe. Electrostatically induced voltages thus essentially disappear once the pipeline is laid into the trench. Note that in the example of Figure 3-10, a pipe-to-ground capacitive reactance of 30 kΩ is sufficient to reduce the pipeline voltage from 1000 V to 1 V. This suggests that when the pipe is raised up on skids, it should be very easy to ground the pipe to mitigate electrostatically induced voltages. In practice, it is found that nearly any type of ground connection—even one as insignificant as a test lead connected to the pipe and contacting the earth—is often sufficient to completely mitigate the induced voltages. In the capacitance calculations in the examples above, note two things: the capacitance calculations are very approximate and the effects of only one phase of the three-phase circuit have been considered. It should also be apparent from the sample calculations that powerline voltage—not powerline current—determines the magnitude of the electrostatically induced pipe voltages. Although electrostatic coupling generally cannot produce enough current to create an electrical safety hazard, it may result in nuisance voltages that produce a sensation similar to a shock from static electricity. This, in turn, could CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:11 conceivably create a secondary safety hazard if, for instance, someone on a pipeline construction project was to overreact to the sensation of an electrostatic voltage on a section of pipe. Because electrostatically induced voltages are typically not hazardous and are easily mitigated, the remainder of Chapter 3 will focus on the much more serious concerns of electromagnetic and conductive coupling effects. 3.1.2 Electromagnetic (Inductive) Coupling Voltages and currents are electromagnetically induced onto a pipeline in the same manner that an inductive pipe locator induces an audio signal onto a pipeline or the primary winding of a transformer induces current to flow through the secondary winding. First consider the flow of electric current in a simple conductor (Figure 3-11). The flow of current creates an electromagnetic field around the conductor, indicated by the lines of magnetic flux F. The intensity of the magnetic field is directly proportional to the current magnitude and is inversely proportional to the distance from the conductor. Using a convention known as the right-hand rule, if a person were to place their right hand around the wire—with the thumb pointing in the direction of current flow—the fingers would indicate the direction of the magnetic flux. Φ I Figure 3-11: Electromagnetic Field Created by Current Flow in a Wire Electromagnetic induction occurs whenever there is a relative motion between an electrical conductor and a magnetic field. This motion may result either from the physical movement of a conductor through a stationary magnetic field, or the CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:12 movement of a magnetic field through a stationary conductor. The most obvious example of the first case is an electrical generator, in which a rotating coil of wire passes through a stationary magnetic field to generate electric current. A less obvious example (discussed in the chapter dealing with telluric current interference) is where the tidal movement of seawater (a conductor) passing through the earth’s magnetic field creates geomagnetic earth currents. In the second case, where both the source of the magnetic field and the conductor are stationary, the magnetic field itself must be in motion in order to induce current in the conductor. This is done by using AC to create a time-varying magnetic field, which expands and collapses around the conductor, to create a relative motion. The best example of this is an electrical transformer (Figure 3-12). I1 Φ I2 Figure 3-12: Electromagnetic Induction in a Multiple-Turn, Iron-Core Transformer An AC I1 flows through the primary winding of the transformer. This creates a magnetic field around each turn of the winding, and these fields link together to create one large magnetic field. The magnetic field around this coil would normally tend to stray well outside the vicinity of the coil; but, by introducing a transformer core made of iron or some other magnetic material, the magnetic field becomes primarily confined to the core. A secondary winding is also wound onto the iron core. Also, the magnetic field created by the primary winding is now expanding and collapsing around the turns of the secondary winding; the secondary winding consequently induces a secondary current flow I2. In order to make transformers energy-efficient, the windings and cores are designed to transfer as much energy as possible from the primary winding to the secondary winding. A transformer can be formed, however, simply by placing a conductor within a time-varying magnetic field around another conductor (Figure CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:13 3-13); such a transformer would be highly inefficient, though. Note that in Figure 3-13, the induced current is not shown to be flowing in the same direction as the primary current. Lenz’s Law states that the induced current flows in a direction that creates a secondary magnetic field, which tends to oppose any change in the primary magnetic field. Because this is AC, the arrows only indicate the current direction at a particular instant of time. They are intended to show that the secondary current is out of phase with the primary current. Φ I1 I2 Figure 3-13: Electromagnetic Induction in a Single-Turn, Air-Core Transformer The case of the single-turn, air-core transformer in Figure 3-13 represents the electromagnetic coupling that occurs when a pipeline runs parallel to a powerline (Figure 3-14). Whereas the voltages that are generated electrostatically are proportional to powerline voltage, the voltages and currents that are electromagnetically induced are proportional to powerline current. As the length of parallelism between the pipeline and powerline increases, the electromagnetic coupling between them improves—just as increasing the number of turns on the primary and secondary windings of a transformer improves the efficiency of the transformer. I1 Φ I2 CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:14 Figure 3-14: Electromagnetic Coupling Between a Pipeline and an Overhead AC Powerline The AC that is induced in the pipeline results in pipe-to-ground voltages where the current discharges into the earth. These electromagnetically induced currents and voltages are a function of powerline current, not powerline voltage. The induced voltages can affect both the integrity of the pipeline and the safety of personnel and the general public. The prediction of the location and magnitude of the voltage peaks. The design of an appropriate mitigation system to limit these peaks to acceptable levels is the primary focus of this chapter. Section 3.5 will cover this in detail. 3.1.3 Conductive Coupling (Resistive Coupling) During Powerline Fault Conditions Conductive coupling can occur when there is a line-to-ground short-circuit or fault on the powerline (Figure 3-15). Under fault conditions, the current leaving the powerline will return to its source using all paths available to it—including powerline shield wires, the earth, and metallic structures in the earth such as pipelines. Figure 3-15: Conductive Coupling During Line-to-Ground Fault Conditions CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:15 The amount of current that is transferred to a pipeline is dependent upon the relative impedances† of all parallel paths available to the fault current. It is a function of the separation distance between the faulted structure and the pipeline, the available fault current, the impedance of the faulted structure to ground, and the pipe-to-earth impedance. On high-voltage powerlines faults are most likely to occur as the result of lightning, which can ionize the air in the vicinity of an insulator. Faults can also occur as the result of high winds, failure of the powerline structures or insulators, or accidental contacts between the powerline and other structures—such as cranes and other construction equipment. Fault current is transferred to the pipeline through the pipeline coating. The better the coating quality (i.e., the fewer the holidays) and the higher the coating’s dielectric strength (i.e., breakdown voltage), the lower the current transfer to the pipeline. Because fault currents are much greater in magnitude than steady-state powerline currents, conductive coupling can result in very high pipeline voltages. However, power system protection devices limit the length of time that these voltages are present on the pipeline to a fraction of a second (typically 0.1s or less on highvoltage systems but longer on low-voltage systems). Even over such a short time period, large amounts of energy can be transferred to the pipeline—resulting in coating damage or even pipeline failure caused by melting or cracking of the pipe wall. The high pipeline voltages resulting from conductive coupling represent a safety hazard to pipeline personnel and, perhaps, the general public in cases where test leads and pipeline appurtenances are accessible. Electric shocks can be painful and can result in the loss of muscular control at body currents of less than 50 mA, but the primary concern for short-duration shocks resulting from fault currents is ventricular fibrillation. Ventricular fibrillation is a condition that may occur at body currents greater than 50 mA and certainly occurs at body currents greater than 100 mA. It results in the total loss of coordination of the heart caused by the disruption of its electrical signals. It will lead to death without defibrillation (i.e., a strong electrical pulse to restore the heart to its normal beating pattern). † To this point, resistances and reactances have been discussed. When dealing with AC circuits, any combination of resistance, capacitive reactance, and inductive reactance (see Section 3.2.1), is in general called an impedance. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:16 In addition to conductive coupling, fault conditions also affect the voltages and currents that are electromagnetically induced on the pipeline. Fault conditions result in increased powerline currents and large imbalances between the phases, which can greatly increase the induced pipeline voltages—albeit for short periods of time. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:17 Experiment 3-1 To Demonstrate the Effects of Electrostatic Induction 3’ 2’ Copper Plate V AC Meter Test Probe 1’ Ground Pin or Screwdriver Procedure Note: This experiment is to be conducted beneath the powerlines at the NACE test site, or beneath any high-voltage AC powerline. The experiment may be conducted by one large group or by several smaller groups. Step: A. Install a ground pin (or screwdriver) into the soil, 2 to 3 inches deep. B. Connect a high-impedance AC voltmeter (or digital multimeter, with ACV selected) to the ground pin using a test lead with an alligator clip. C. Connect a test lead with a test probe to the other terminal of the voltmeter. D. Beginning at a height of 1 foot above grade, increase the height of the test probe at approximately 1-foot intervals until a reaching a height of 6 or 7 feet is reached (taking an AC voltage measurement each time). E. Disconnect the test probe from the meter and connect a test lead having an alligator clip. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:18 F. Connect the alligator clip to the copper plate and repeat each of the AC voltage measurements. If possible, support the plate by contacting only the insulated cover of the alligator clip. G. Repeat the measurements in Step F while holding directly onto the copper plate—and while one knee is contacting the ground. H. If there is a car parked near the powerline, measure the AC voltage between the steel frame of the car and the ground pin. Results Height (ft) AC Voltage Measurement (V) Copper Plate Copper Plate Probe (Standing) (Kneeling) 1 2 3 4 5 6 7 Questions for Discussion 1. What are the voltages being measured with the meter test probe suspended in the air? Is this just electrical noise? Why? 2. In what way do the voltage measurements change when the meter is connected to a copper plate? Why? 3. Does it seem that the measured voltages should be higher, lower, or similar to those recorded? CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:19 4. What happens to the voltages measured using the copper plate when one knee contacts the earth? Why? 5. How does the voltage measured on the car compare with those measured on the copper plate? Explain. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3.2 3.2.1 3:20 Basic Theory of Electromagnetically Induced Voltages AC Circult Theory Before starting a detailed discussion of how AC voltages are electromagnetically induced on a pipeline, it is important to understand some of the basic principles of AC circuits. Consider first the case of a transformer. The voltage induced in the secondary winding (VS) relative to the primary voltage (Vp) is equal to the ratio of the number of turns in the secondary winding to those in the primary winding (see Equation 34a). VS N = S Vp Np [3-6a] Alternatively, one may think of this relationship in the following terms: the volts/turns ratio for the primary winding is equal to the volts/turns ratio for the secondary winding, or windings, in the case of a transformer having multiple secondary windings. Vp Np = VS NS [3-6b] For example, 100V is applied to the primary winding of the transformer in Figure 3-16. Because this winding has 100 turns, then the volts/turns ratio is 1V/turn. This same volts/turns ratio also applies to the secondary winding having only five turns, so that it will develop a voltage of 5V. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:21 100 V 100 Turns 1 V/turn 5 turns 5 turns 5 turns 1 V/turn V ⎛N ⎞ ⎛ 5 ⎞ VS = VP ⎜ S ⎟ =100V⎝ = 5V ⎝ NP ⎠ 100 ⎠ Figure 3-16: Determination of Voltage on a Transformer Secondary Now consider the case where several secondary windings of the transformer are connected together in series (Figure 3-17). Assuming that all of the secondary windings have been wound onto the transformer core in the same direction, then the voltages are additive. 100 V 100 Turns 5 turns V1 5 turns V2 V 5 turns V3 1 V/turn 1 V/turn V = V1 + V2 + V3 = 5V + 5V + 5V = 15V Figure 3-17: Effect of Interconnecting the Secondary Windings This is analogous to adding the voltages of a number of batteries that are connected together in series, as shown in Figure 3-18a. If, however, one of the batteries was reversed (Figure 3-18b), then its voltage would be added to the voltages of the other batteries as a negative value. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference (a) 3:22 (b) 1.5V 1.5 V 1.5 V 1.5V 1.5 V 1.5 V V - V + - + V = 1.5V -1.5V +1.5V =1.5V V = 1.5V +1.5V+1.5V = 4.5V Figure 3-18: Effect of Polarity on a Series Combination of DC Voltage Sources Consider once again the case of the transformer in Figure 3-17. Although AC voltages and currents exhibit no polarity in the DC sense, reversing the connections to one of the secondary windings would be akin to reversing the polarity of one battery in Figure 3-18—as Figure 3-19 shows. 100 V 100 Turns 5 turns 5 turns 5 turns 1 V/turn 1 V/turn V V = 5V - 5V + 5V = 5V Figure 3-19: Effect of “Polarity” on a Series Combination of AC Voltage Sources Even though the polarity of an AC voltage continuously alternates from positive to negative, the voltages in the specific case shown can still be added and subtracted just as if they exhibited a constant polarity. When working with AC circuits, the term “polarity” is generally not used because it applies only to a particular instant in time. Instead, waveforms are compared in terms of their phase relationships. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:23 AC waveforms are said to be either in-phase or out-of-phase with one another. “In-phase” means that two waveforms have the same polarity at every instant in time (Figure 3-20). In the case of the transformer in Figure 3-17, the voltages in all three secondary windings must be in-phase with one another because they are all generated by the same magnetic field. When the connections are reversed on the middle winding, its voltage waveform becomes exactly out-to-phase with the waveforms of the other two windings. In other words: the waveforms start at the same instant in time but, when one goes positive, the other goes negative. When viewed on an oscilloscope, an AC waveform indicates the variation of the voltage or current versus time. When the frequency of the waveform is 60 Hz, the period of the waveform is 1/60 s or 16.67 mS. When solving interference problems related to AC power systems, the frequency of all waveforms will be the same (either 50 Hz or 60 Hz, depending upon the country). Therefore, frequency (and time) can be omitted from the analysis. Instead, it is easier to discuss waveforms in terms of phase angle (Figure 3-20): one full cycle of the sinusoidal waveform is comprised of 360 degrees or, alternatively, 2π radians. Voltage or Current 1 0 -1 0 90 0 π/2 0 4.2 180 270 Angle (Degrees) π 3π/2 Angle (Radians) 8.3 12.5 Time (ms) at 60 Hz 360 2π 16.6 Figure 3-20: In-Phase 60 Hz AC Waveforms To better understand the use of phase angles in AC circuits, consider the case of a typical distribution transformer that would be used to supply a residential electrical service (Figure 3-21). The secondary of the distribution transformer consists of CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:24 two windings, both of which supply approximately 120 V. Recognize that the secondary windings would exhibit no voltage with respect to ground in the configuration shown because all terminals of the secondary windings are floating. In order for there to be any relationship between the voltages across the secondary windings and ground, a connection must be made between one of the secondary terminals and earth. In a residential electrical service, this connection is made at the connection between the two windings (Terminal B)—as shown in Figure 3-22. 4140 V 4140 V 120 V 120 V Figure 3-21: Typical Electrical Distribution Transformer 4140 V 4140 V A +120 V 120 V 120 V B C -120 V Figure 3-22: Typical Residential Electrical Service By connecting Terminal B to ground, the voltage at this point is forced to become 0 V with respect to earth. Furthermore, because a voltage of 120 V exists across each of the windings, there is now a voltage of 120 V with respect to earth at both terminals A and C. It is important to realize, however, that the voltages at these two terminals are completely out-of-phase with one another. At a moment when the voltage across the windings is rising with the polarity indicated by the arrows, CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:25 Terminal A will have a positive polarity while Terminal B will have a negative polarity. The AC voltage waveforms with respect to earth would therefore appear as shown in Figure 3-23 1 Voltage Terminal A 0 Terminal C -1 0 90 0 π/2 180 Angle (Degrees) π Angle (Radians) 270 360 3π/2 2π Figure 3-23: AC Waveforms on a Residential Electrical Service The two waveforms in Figure 3-23 are said to be 180 degrees out-of-phase with one another because any point on one waveform is found on the other waveform to be shifted horizontally by 180 degrees. When waveforms are in-phase with one another, such as those in Figure 3-20, they have a phase shift of 0 degrees between them. In general, any sinusoidal AC voltage waveform v(t) can be described by the following equation: v(t ) = Vm cos (ωt + φ) [3-7] where Vm is the peak voltage, ω is the angular frequency, t is time, and φ is the phase angle of the waveform. Angular frequency simply expresses frequency in either degrees/second—or, more commonly, in radians/second—rather than in cycles per second and is therefore simply: ω = 2πf CP Interference Course Manual © NACE International, 2006 January 2008 [3-8] AC Interference 3:26 The phase angle is the angle in degrees (or radians) that the subject waveform is shifted from a pure cosine wave. Equation 3-7 is plotted in Figure 3-24 for the cases where the phase shift is equal to 0 degrees and –45 degrees. Vm1 φ=0° φ=-45° 0 -Vm -1 0 90 180 270 360 ωt (Degrees) Figure 3-24: Plot of General Equation for Sinusoidal AC Waveforms When performing analysis on AC waveforms having the same frequency, the only quantities required to distinguish a waveform are its amplitude and its phase angle. Therefore, instead of using Equation 3-7 to denote a waveform’s characteristics, phasor notation is used. A sinusoidal AC voltage waveform can be identified in phase notation as: V = V∠φ [3-9] where V is the voltage phasor, V is the magnitude of the voltage waveform, and φ is its phase angle with respect to a pure cosine wave. Therefore, recalling the example of a residential electrical service in Figure 3-22, the correct way to refer to the voltages at terminals A and C would be 120V /0° and 120V /180°, respectively, rather than +120V and –120V. Phasors can also be represented graphically using a phasor diagram. The graphical representation is essentially a vector, having its origin at (0,0), having a length equal to the voltage magnitude, and an angle with respect to the positive x-axis is equal to the phase angle. For example, two waveforms having the phasor notations 1V /0° and 1V /45° would be represented graphically as shown in Figure 3-25. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:27 Y-axis +1 1V /45° 1V /0° -1 +1 X-axis -1 Figure 3-25: Typical Phasor Diagram Phasor diagrams are important because they help illustrate how AC voltages and currents must be added and subtracted. In the case of determining the total voltage of a series of batteries (Figure 3-18), the voltages are simply added or subtracted based on their polarities. The same is true when dealing with AC voltages that are either in-phase, or 180 degrees out-of-phase—as was the case with the transformer windings in figures 3-17 and 3-18. However, when determining a voltage difference between two points in an AC circuit—where the phase difference between the waveforms is something other than 0 degrees or 180 degrees—the rules of vector algebra must be applied. Example Calculation: Two AC voltage sources are connected together in series, as shown in Figure 3-26. Both have a voltage output of 1 V; however, the waveforms are 45° out-of-phase with one another. Determine the total voltage of the system. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:28 1V /0° 1V /45° A B VAB Figure 3-26: Series Combination of AC Voltage Sources The voltages of the two AC sources can be added together using the phasor diagram in Figure 3-27 and the rules of vector algebra. In order to add the two phasors together, the vector for source B is shifted horizontally so that its tail is placed at the head of the vector for Source A—as shown in Figure 327. The total voltage for the two sources is then provided by the new vector running from the origin to the head of the vector for source B. Y-axis +1 VAB φΑΒ VA = 1V /0° -1 +1 V B= 1V /45° X-axis -1 Figure 3-27: Phasor Diagram for Problem in Figure 3-26 CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:29 If this diagram was drawn to scale, then VAB could be determined using a ruler to measure its magnitude, ⏐VAB⏐, and a protractor to measure its phase angle, /φAB. An easier and more accurate way, however, is to use trigonometry. The vectors for VA and VB are broken down into x- and y- components. The x- and y- components for VAB are then determined by adding the xcomponents together and then adding the y-components together, as follows: X-component of VAB: 1⋅ cos(0º) + 1⋅ cos(45º) = 1+ 0.71 = 1.71 Y-component of VAB: 1⋅ sin(0º) + 1⋅ sin(45º) = 0 + 0.71 = 0.71 The magnitude and phase angle for the new vector is then determined as follows: Magnitude of VAB: V AB = x 2 + y 2 = 1.712 + 0.712 = 1.85 Angle of VAB: ⎛ 0.71 ⎞ ⎛ y⎞ ∠V AB = tan −1 ⎜ ⎟ = tan −1 ⎜ ⎟ = 22.5° ⎝ 1.71 ⎠ ⎝x⎠ Note that most scientific calculators can now add and subtract vector quantities directly, without the need to break the vectors down into components and then reassemble them. Phasor multiplication and division are much simpler procedures than addition and subtraction. To determine the product of two phasor quantities, their magnitudes are multiplied together while their angles are added together. Division is done similarly, as shown below: Α∠φ × Β∠θ = Α ⋅ Β∠( φ + θ) [3-10] Α∠φ ÷ Β∠θ = Α ÷ Β∠( φ − θ) [3-11] In addition to vector algebra, the analysis of AC circuits requires the use of complex mathematics involving the use of both real and imaginary numbers. This is a difficult concept for some to understand, but it can be simplified as follows. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:30 In complex mathematics, the x- and y- axes in the phasor diagram of Figure 3-25 are renamed the real and imaginary axes, respectively. Any phasor A/φ can then be broken down into real and imaginary components as follows: A = x + jy [3-12] where the real part of A, Re(A) is determined as: x = ⏐A⏐cosφ [3-13] and where the imaginary part of A, Im(A) is determined as: y = ⏐A⏐sinφ [3-14] The significance of j is to denote the imaginary component of the phasor, and it is known as the complex operator. The mathematical equivalent of j is: j = −1 [3-15] hence it is referred to as an imaginary number. Note that in mathematics, an imaginary number is typically represented by the symbol I; but, so as not to confuse it with the symbol for current, electrical engineers have adopted the symbol j. Although the mathematical definition for j can be perplexing, its use in AC circuit calculations is not. In essence, j is simply a phasor having a magnitude of unity and a phase angle of 90 degrees; that is, j = 1/90º [3-16] Therefore, if a phasor is multiplied by j, the magnitude of the phasor is unaffected; however, the phase angle becomes increased by 90º. Similarly, if a phasor is divided by j, its magnitude is unaffected but its phase angle will be decreased by 90º. Multiplication by -j has the same effect as dividing by +j, as shown below: A/φ × j = A/φ × 1/90º = A/φ + 90º [3-17] A/φ × (-j) = A/φ × -1/90º = A/φ + 90º = A/φ – 90º [3-18] CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:31 A/φ ÷ j = A/φ ÷ 1/90º = A/φ – 90º [3-19] One of the uses for the complex operator is the calculation of impedances, along with determining what effect an impedance has on the phase angle of an AC waveform. Just as the two AC voltage sources could not be added together directly in Figure 3-26, a resistance and a capacitive reactance connected together in series cannot be simply added together to determine their combined impedance. Consider again the case of a capacitor in an AC circuit (Figure 3-28), as previously discussed. The magnitude of the reactance offered by the capacitor is given by Equation 3-3; therefore, the magnitude of the current I from the voltage source V can already be determined. However, to determine the effect that the capacitor has on phase angle, the following formula must be used: XC = IC 1 j 2πfC [3-20] C V Figure 3-28: Determination of Current through a Capacitor The current is then calculated using Ohm’s Law as: IC = V V = = j 2πfCV = 2πfCV∠90° XC ⎛ 1 ⎞ ⎜⎜ ⎟⎟ ⎝ j 2πfC ⎠ CP Interference Course Manual © NACE International, 2006 January 2008 [3-21] AC Interference 3:32 By studying Equation 3-21, the capacitor is found to shift the phase angle of the current by 90 degrees relative to the voltage waveform. This is illustrated graphically in Figure 3-29. 90° Shift V 1 m Applied Voltage 0 Resulting Current -Vm -1 0 90 180 270 360 ωt (Degrees) Figure 3-29: Voltage and Current Waveforms for a Purely Capacitive Circuit Because the current waveform is starting 90 degrees before the voltage waveform, it is said that the current leads the voltage in a capacitor. Now consider the case of an inductor, L (Figure 3-30). In the same way that reactance in a capacitor results from the accumulation of electrical charges, reactance in an inductor results from the generation of a magnetic field. An inductor is generally a coil of wire; however, any conductor (such as a pipeline) has an inductive component. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:33 IL L V Figure 3-30: Determination of Current through an Inductor The reactance of an inductor is calculated as: X L = j 2πfL [3-22] Because inductive reactance is proportional to frequency, an inductor has the opposite behavior to that of a capacitor; in effect, the inductor appears as a short circuit to DC and as an open-circuit to very high-frequency currents. Once again, Ohm’s Law can be used to determine the current through the inductor: IL = V V V = = ∠ − 90° XL j 2πfL 2πfL [3-23] The phase angle for the current in Equation 3-23 indicates that an inductor shifts the phase angle of the current by –90 degrees relative to the voltage waveform; it is therefore said that the current lags the voltage in an inductor. This is illustrated graphically in Figure 3-31. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:34 -90° Shift Vm1 Resulting Current Applied Voltage 0 -Vm -1 0 90 180 270 360 ωt (Degrees) Figure 3-31: Voltage and Current Waveforms for a Purely Inductive Circuit The effect of capacitors and inductors on the phase angle of the current can be remembered by the following simple pneumonic device: “ELI the ICE man.” This indicates that E leads I in an inductive (L) circuit, whereas I leads E in a capacitive (C) circuit. As Section 3.1 notes, this discussion of AC interference applies primarily to 3φ AC power lines having voltage waveforms that are 120 degrees apart—as shown in Figure 3-2. The phasor representation of such a 3φ circuit is shown in Figure 332. Phase C +1200 Phase A Phase B -1200 Figure 3-32: Phasor Representation of a Three-Phase Circuit CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:35 When dealing with 3φ powerlines, the system voltage is always specified in terms of the line-to-line voltage: that is, from phase A to phase B, phase B to phase C, or phase C to phase A. As the phasor diagram shows, the phase-to-phase voltage (from the tip of one phasor to the tip of another) is greater than the phase-toground voltage (the length of any one phasor). To determine the relationship between the phase-to-phase and phase-to-ground voltages, apply the rules of vector algebra as follows: Vφ–φ = VA – V B = Vφ–G /0º – Vφ–G /120º = (Vφ–G cos 0º + jVφ–G sin 0º) – (Vφ–G cos 120º + jVφ–G sin 120º) = (1.5V – j0.866V) VΦ − G ∴ Vφ − φ = 1.5 2 + (− .866 ) ⋅ Vφ−G = 2 3 Vφ − G [3-24] Example Calculation: Determine the line-to-ground voltage of a 500-kV transmission line. Vφ − G = 3.2.2 Vφ − φ 3 = 500 kV = 289 kV 3 The Nature of Induced AC Pipeline Voltages In the discussion of electromagnetic coupling in Section 3.1.2, it was shown that AC could be induced in a pipeline by a mechanism similar to that used in a transformer. It will now be discussed how these induced pipeline currents result in pipe-to-ground voltages. Consider the electrical model of the pipeline shown in Figure 3-33. An AC voltage source represents the electrical energy induced in the pipe. The magnitude and phase angle of the resulting current that travels down the pipe will depend upon the longitudinal impedance internal to the pipeline—as well as on the shunt impedance between the pipeline and the earth. The longitudinal impedance is made up of the series combination of the longitudinal resistance of the pipe steel (RL) and the longitudinal inductance of the pipeline (LL). The shunt impedance is CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:36 made up of the parallel combination of the coating resistance (RC) and the shunt capacitance across the coating (CS), which is discussed in Section 3.1.1. V RL CS LL RS Figure 3-33: Electrical Model of Single Pipe Section The shunt capacitance and the longitudinal inductance are very important when conducting induced AC pipeline calculations, but the model will be simplified (Figure 3-34) to illustrate how pipe voltages are generated along a pipeline. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:37 V RL RS Figure 3-34: Simplified Electrical Model of Single Pipe Section Alternatively, the pipe section can be modeled as shown in Figure 3-35, where the shunt resistance RS is split into two parallel shunt resistances of 2RS. These models can be combined together to model a long pipeline as shown in Figure 336; however, for a simple analysis, two pipe sections will be sufficient (Figure 337). Note that the parallel combination of two shunt resistances of 2RS in Figure 3-37 has been simplified to form a single shunt resistance of RS in Figure 3-38. RL 2RS 2RS Note: RS = 2RS // 2RS Figure 3-35: Simplified Electrical Model of Single Pipe Section Figure 3-36: Series Combination of Multiple Pipe Sections CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:38 RL 2RS RL 2RS 2RS 2RS Figure 3-37: Series Combination of Two Pipe Sections RL RL 2RS RS 2RS Figure 3-38: Series Combination of Two Pipe Sections (Simplified) Example Calculation: The simple electrical network in Figure 3-38 can be solved using Kirchhoff’s Law, which states that the sum of the voltage sources in any loop of a circuit is equal to the sum of the voltage drops. Assuming that a current of I1 flows in the first loop and a current of I2 flows in the second loop (Figure 3-39), the following formulae can be derived: V1 A RL I1 2RS B RS V2 RL I2 C 2RS Figure 3-39: Circuit Analysis Using Kirchhoff’s Law V1 = I1RL + I1RS – I2RS + 2I1RS V2 = I2RL + I2RS – I1RS + 2I2RS CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:39 In this simple example, it is assumed that the magnetic field to which the pipeline is exposed is uniform along the length of the pipeline. Therefore, the energy transferred to the first section of pipe will be equal to that transferred to the second section of pipe and V1 = V2. Therefore, the above equations must be equivalent, and I1RL + I1RS – I2RS + 2I1RS I2 (RL + 4RS ) ∴ I2 = I2RL + I2RS – I1RS + 2I2RS = I1 (RL + 4RS ) = I1 Finally, because I1 and I2 are equal, they can both be replaced in Figure 3-39 with I; also, Figure 3-39 can be simplified as shown in Figure 3-40. A V RL I 2RS B V RS RL I C 2RS Figure 3-40: Circuit Analysis Using Kirchhoff’s Law From Figure 3-40, it is a simple matter to calculate the pipe voltages at each end of the pipeline (points A and B) and at the middle of the pipeline (Point B). VA = 0 – 2RSI = –2RSI VB = IRS – IRS = 0 VC = 0 + 2RSI = +2RSI These voltages have been plotted in Figure 3-41. This plot indicates that for a simple induced AC problem—where the electrical characteristics of the pipeline and the magnetic field generated by the powerline are both constant along the length of the pipeline—that the induced AC voltage will be zero at the middle of the pipeline and will peak at the ends of the pipeline. Furthermore, the voltage CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:40 peaks at the ends of will be equal in magnitude; but, they will be of opposite polarity—or, more accurately, the phase angles for the voltages at each end of the pipeline will be 180 degrees apart. V +2RsI 0 A B C Distance -2RsI Figure 3-41: Induced AC Voltage Profile along Two-Section Pipe Method of Figure 3-39 If voltage measurements are taken at various points along a similar pipeline using an AC voltmeter, the voltage profile in Figure 3-42 would be obtained. The differences in the phase angles of the induced AC voltages could not be determined using a voltmeter but would be as shown in the figure. Vpeak Magnitude 0 Distance 180º Phase Angle 0º Distance Figure 3-42: Profile of Induced AC Voltages and their Phase Angles along any Pipeline having Uniform Electrical Characteristics The AC voltage profiles given above are based on a model consisting of only two pipe sections. A similar analysis could be conducted on a larger model, such as the CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:41 one in Figure 3-36; however, this would require much more rigorous mathematics. Nevertheless, a larger model would yield identical results. Again, this assumes that both the pipeline electrical characteristics and the magnetic field as seen by the pipeline are constant along the entire length of the pipeline. It also assumes that the pipeline is electrically short, or non-lossy. An electrically short pipeline is one where the longitudinal impedance looking down the pipeline is much lower than the shunt impedance between the pipeline and the earth. A well-coated pipeline has a high impedance to earth and would likely be considered electrically short. However, as the pipeline becomes physically long, the longitudinal impedance of the pipeline increases and its shunt impedance to earth decreases. When the longitudinal impedance can no longer be considered insignificant compared to the shunt impedance (i.e., when the pipeline starts to become lossy), the pipeline can no longer be considered electrically short and its AC voltage profile will start to exhibit non-linear behavior (Figure 3-43). V Pipeline Becoming Increasingly Lossy L/2 0 0 Distance L Figure 3-43: Effect of Electrical Length of Pipeline on AC Voltage Profile This exponential signal attenuation is similar to what one sees with cathodic protection (CP) currents, except that the AC attenuation constant that determines the shape of the attenuation curve differs from the DC attenuation constant. This variation exists because the AC attenuation constant is a function of pipeline inductance and shunt capacitance—not just the longitudinal and shunt resistances. Note that electrically long pipelines subjected to electromagnetic coupling can CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:42 exhibit zero voltages over much of their length, provided that the electromagnetic field and the electrical characteristics of the pipeline and the soil are uniform along this length. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:43 Experiment 3-2 To Demonstrate the Effects of Electromagnetic Induction AC Power Line VAC Pipe Test Lead 1 2 3 4 5 6 7 8 9 10 Coated Pipe Section Experiment Schematic No. 1 Procedure Note: This experiment is to be conducted on the buried section of coated pipeline at the NACE field test site or be omitted. The experiment may be conducted by one large group or several smaller groups. Step: A. Using a high-impedance AC voltmeter (or digital multimeter, with ACV selected), measure the AC voltage on each of the pipe test leads with respect to a ground pin placed near the base of the test station (TS). B. At TS #1 only, monitor the AC voltage for one or two minutes to determine if it varies with time. C. Return to the classroom and plot the results versus test station number. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:44 Results TS # 1 2 3 4 5 6 7 8 9 10 Voltage (AC mV) Questions for Discussion 1. Does the plot of induced voltages along the pipeline appear as expected? CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3.3 3:45 Induced AC Voltages 3.3.1 Factors that Affect the Longitudinal Electric Field The electromagnetic field produced by the powerline current generates an electric field running longitudinally along the pipeline—known as the Longitudinal Electric Field (LEF)—that has the units of volts per meter (V/m). The LEF is represented by the symbol E. It is a complex number, meaning that it has a magnitude and a phase angle. The voltages that are induced on the powerline are directly proportional to the magnitude of the LEF. The LEF is directly proportional to the electromagnetic field and is therefore directly proportional to the powerline phase currents. Furthermore, because the electromagnetic field strength varies inversely with distance from the powerline, so does the LEF. The LEF is also a function of how the conductors are arranged on the tower. In addition to the horizontal configuration shown in Figure 3-1, the phase conductors may be arranged vertically (figures 3-44 and 3-45), in a delta configuration (Figure 3-46), or in other, less orderly configurations. Note that even for the two vertical configurations shown, in one case the phase conductors are arranged truly vertical with respect to one another. In the other case, however, the middle conductor has been offset from the upper and lower conductors. Figure 3-44: Double Vertical Circuit CP Interference Course Manual © NACE International, 2006 January 2008 Figure 3-45: Quadruple Vertical Circuit AC Interference 3:46 Figure 3-46: Single Delta Circuit The separation distance between the phase conductors is also a primary factor. In general, the LEF increases linearly with increasing conductor separation. The reason for this is described below. Consider the case of an AC power cord for an electrical appliance. If one was to take a clamp-on AC ammeter and attempt to measure the current being drawn by the appliance, the ammeter would read zero. This is because the current entering the appliance is equal to the current leaving it, except that these currents are exactly 180 degrees out-of-phase with one another. Therefore the electromagnetic fields created by the two conductors cancel each other out completely and the net magnetic field sensed by the ammeter is zero. Similarly, if a clamp-on ammeter was placed around all three phase conductors supplying a piece of three-phase equipment, the current entering would equal the current leaving—even though this is less obvious than in a single-phase case. Therefore, the net magnetic field surrounding the conductors would again be zero. The clamp of a clamp-on ammeter completely surrounds a conductor. It therefore senses the entire magnetic field created by the current in the conductor. In the case of a pipeline-powerline corridor, the pipeline senses only a portion of the magnetic fields created by the currents in each of the phase conductors; also, it senses them in inverse proportion to the separation distances to each conductor. Consider the case of a single horizontal circuit (Figure 3-47). The conductor whose magnetic field will have the greatest effect on the LEF generated along the pipeline will be Phase C, whereas the conductor having the least effect will do so CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:47 in Phase A. The greater the separation distance ratios PC/PA, PC/PB, and PB/PA, the less cancellation will exist between their magnetic fields as seen by the pipeline and the greater the LEF. B A C PB PC PA P Figure 3-47: Effect of Phase Conductor Separation If one considers the extreme case of all three phase conductors being bundled together into a single three-conductor cable, then the pipe-to-conductor separation distances for all three phases would be essentially equal and the pipeline would see no net magnetic field. This is one of the reasons why pipelines are generally unaffected by buried three-phase cables that run parallel and in close proximity to the pipeline. Another primary factor affecting the LEF applies only to multiple-circuit powerlines (such as the one shown in Figure 3-44) is the physical arrangement of the phases for one circuit versus the other. Four possible phase arrangements for a double vertical circuit powerline are shown in Figure 3-48. A B A B A B C B A B C A A B B A C C C A C B C C Centre Line Symmetric Centre Point Symmetric Full Roll Partial Roll Figure 3-48: Phase Arrangements for a Double Vertical Circuit CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:48 In the same way that increased conductor separation reduces cancellation effects between the fields created by the three phases, the arrangement of phases on a double-circuit powerline can also have a large impact on these cancellation effects. For instance, in the case of a pipeline sitting to the left of the center-point symmetric powerline shown in Figure 3-48, the pipeline is closest to Phase C of the left circuit but is farthest from Phase C of the right circuit. Therefore, the field created by the two Phase C conductors (as seen by the pipeline) would be less than that in the case of the center-line symmetric powerline. An example of just how significant a factor phase arrangement can be is shown in Figure 3-49. Note that these curves have been calculated for a specific set of values for pipeline-topowerline separation distance d, conductor separation s, conductor height h, soil resistivity ρ, and phase current I. 70 60 Centre Line 50 |E| (V/km) Partial Roll 40 30 Full Roll 20 Centre Point 10 0 0 1 2 3 4 5 6 7 8 9 10 d/s Figure 3-49: Effect of Phase Arrangement on LEF Magnitude for Variation d/s Ratios (for the specific case where ρ/s2= 1Ω/m, s/h = 0.3, and I = 1000A) The curves in Figure 3-49 indicate that the center-line symmetric phase arrangement results in the highest LEF magnitude, whereas the center-point symmetric phase arrangement results in the lowest LEF magnitude. In some extreme cases, there may be an order of magnitude difference between the LEFs generated by these two different phase arrangements. This begs the question as to why would all double-circuit vertical powerlines not be constructed using a centerpoint symmetric phase arrangement, particularly when the majority of utilities appear to have adopted the center-line symmetric arrangement as their standard. There does not appear to be a definitive answer to this question, other than that “this is the way it has always been done”. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:49 Normally, a three-phase power transmission circuit is loaded so that the currents carried by each of the three phases are approximately equal. Any significant imbalance between these currents may reduce cancellation effects between the fields and cause an increase in the LEF. The most extreme case of an imbalance, of course, is a line-to-ground fault condition; however, this will be discussed in Section 3.7. Imbalances may also exist between two circuits of double-circuit powerline, which can reduce cancellation effects and increase the LEF. The most extreme case occurs under emergency conditions, when one circuit of a double-circuit powerline is temporarily deactivated and its load is added to the load of the second circuit. Two other factors having only a secondary impact on the magnitude of the LEF are soil resistivity and the presence of shield wires. In general, as soil resistivity increases the LEF generated along the pipeline will show a slight increase; however, a resistivity increase has a much more significant effect on the effectiveness of mitigation than on the LEF. When shield wires are introduced onto a powerline, currents are induced in the shield wires. The shield wires will then, according to Lenz’s Law (Section 3.1.2), have a magnetic field that opposes any change in the primary magnetic field. In other words, the field generated by the shield wire currents should lessen the overall field seen by the pipeline. This theoretically should also apply to currents that are generated in other paralleling structures, such as railway rails and foreign pipelines. Although the discussion above has involved only factors that affect the magnitude of the LEF, it is important to realize that all of these same factors also affect the phase angle of the LEF. Phase angle is often just as important as magnitude in determining a pipeline’s response to the field. The determination of the magnitude and the phase angle of the LEF for a particular pipeline-powerline corridor is a complex task that will be discussed in more detail in Section 3.6. 3.3.2 Factors that Affect the Pipeline Voltages In the case of a simple pipeline-powerline corridor, where the LEF is uniform along the length of the pipeline (Figure 3-50), the location and magnitude of the voltage peaks can be calculated once the LEF has been determined. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:50 Powerline d L Figure 3-50: Simple Pipeline-Powerline Corridor (Plan View) As Figure 3-41 shows, the voltage peaks for this system will be where the pipeline stops paralleling the powerline—in this case, at the insulators. It is also known that a zero voltage would occur at the midpoint of the pipeline. Assuming that this pipeline is electrically short, the voltages at the insulators can be calculated using Equation 3-25. The voltage profile for this ideal case is shown in Figure 3-51. Note that polarity is once again being used to denote that the voltage peaks have phase angles that are 180 degrees apart. VO, L = ± E ⋅L [3-25] 2 V E ⋅L 2 0 − 0 L/2 Distance L E ⋅L 2 Figure 3-51: AC Voltage Profile along an Electrically Short Pipeline (Uniform Conditions – No Grounding) The analogy of this simple pipeline-powerline problem, using the residential electrical service, is shown in Figure 3-52. The voltage peaks that exist at the ends CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:51 of the secondary winding are of equal magnitude but opposite polarity (i.e., phase) to one another, and the voltage at the midpoint is zero. In the case of the pipeline, there is no ground connection at the center of the pipe that forces this point to zero volts. However, because it is assumed that the pipeline has a uniform distributed shunt impedance along its length, this is equivalent to lumping these impedances into one shunt impedance at the midpoint. In contrast, if the electrical characteristics of the pipeline are not uniform along its length, then the voltage peak at one end of the pipeline may be higher than that predicted by Equation 325. The peak at the other end would be lower, and the zero crossing would have shifted from the center of the pipeline (Figure 3-53). 4140 V 4140 V 120 V 120 V -120 V +120 V Figure 3-52: Electrical Service Analogy for Pipeline-Powerline Corridor in Figure 3-50 V E ⋅L E ⋅L 2 0 0 − L/2 Distance L E ⋅L 2 −E ⋅L Figure 3-53: AC Voltage Profile Along an Electrically Short Pipeline (Non-Uniform Conditions – No Grounding) CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:52 Consider once again the case of a residential electrical service, where the ground connection is moved from the midpoint of the secondary winding to the end (Figure 3-54). The voltage at the midpoint now becomes 120 V, whereas the opposite end rises to 240 V because a voltage of 240 V must still exist from one end of the winding to the other. 4140 V 4140 V 120 V 120 V +120 V +240 V Figure 3-54: Effect of Grounding One End of Electrical Service Secondary The situation described in Figure 3-54 also applies to a well-coated pipeline if one end is connected to a low (near-zero)-resistance ground. The grounded end of the pipeline is forced to zero volts; however, the voltage that has been induced from one end of the pipeline to the other still exists. Hence the opposite end of the pipeline rises to a voltage that is twice what it was previously (Figure 3-55). Grounding is an effective means of mitigating induced AC interference, but this example shows that grounding may actually increase pipeline voltages if it is installed on the pipeline indiscriminately. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:53 V E ⋅L 0 0 L/2 Distance L Figure 3-55: Effect of Grounding One End of Pipeline in Figure 3-50 If grounds are applied to both ends of the pipeline, or if grounds are uniformly distributed along the pipeline, then all voltages can be mitigated to less than those predicted by Equation 3-25—as Figure 3-56 shows. V E ⋅L E ⋅L 2 0 − 0 L/2 Distance L E ⋅L 2 −E ⋅L Figure 3-56: Effect of Grounding Both Ends of Pipeline or Adding Distributed Grounds The installation of insulators will introduce additional voltage peaks on the pipeline. On an electrically short pipeline, the installation of an insulator essentially creates two electrically separate pipelines. Each pipeline has smaller voltage peaks than the original pipeline because the voltages are proportional to the physical lengths of each pipe section (Figure 3-57). An important CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:54 consideration in this case, is that even though the voltage peaks are now reduced, a new voltage peak appears at the midpoint of the pipeline where previously the voltage had been zero. Furthermore, because the voltage on one side of the insulator is opposite in polarity (i.e., phase) to the voltage on the other side, the voltage across the insulator is twice the voltage of the pipeline to ground. V E ⋅L 4 0 − E ⋅L 4 0 L/2 L Figure 3-57: Effect of an Insulator at the Midpoint of the Pipeline It is important to note that the value ⏐E⏐⋅ L represents an absolute limit of the maximum induced voltage that can appear on a pipeline, regardless of whether the pipeline is electrically long or short or where the grounds and insulators are installed. For example, if the field strength along the pipeline is 10 V/km under maximum loading conditions and the pipeline parallels the powerline for 5 km, then the maximum voltage that can appear on the pipeline is 50 V. However, it is more likely that this voltage would be more evenly distributed—with approximately 25 V appearing at each end and 0 V appearing near the pipeline’s midpoint. As the pipeline increases in length, voltages cannot continue to increase proportionately and without limit as predicted by Equation 3-25; this stems from the effects of attenuation. Furthermore, as the pipeline becomes more lossy, the voltage profile becomes less linear (Figure 3-43). In order to calculate voltages on an electrically long pipeline, the pipeline’s physical length in Equation 3-25 is replaced by a parameter that may be thought of as being the pipeline’s electrical length, l (2/Γ), resulting in Equation 3-26. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:55 VO, L = ± E [3-26] Γ The parameter Γ is known as the pipeline propagation constant. It is a constant related to the electrical characteristics of the pipeline and is closely connected to the pipeline’s AC attenuation constant. The value of Γ can be calculated, as shown in Section 3.6.2; however, it is a complicated function of pipe depth, soil resistivity, AC frequency, and coating resistance as well as pipe diameter, wall thickness, and material. Alternatively, the value of Γ can be obtained from tables or graphs. Figure 3-58 shows the AC voltage profile along a pipeline for the simple pipeline-powerline geometry shown in Figure 3-50, where the pipeline is electrically long or lossy. V E Γ 0 − 0 L/2 Distance L E Γ Figure 3-58: AC Voltage Profile along an Electrically Long or Lossy Pipeline (Uniform Conditions – No Grounding) It was found in the case of an electrically short pipeline that grounding one end could actually increase voltages at the opposite end. In the case of an electrically long pipeline, however, grounding one end of the pipeline would have no effect on the other end because the two points are electrically remote from one another (Figure 3-59). CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:56 V E Γ 0 − 0 L/2 Distance L E Γ Figure 3-59: AC Voltage Profile along an Electrically Long or Lossy Pipeline (Zero Resistance Ground at Distance = 0) As was previously mentioned, electrically long pipelines may exhibit zero voltages over much of their length—provided that the electromagnetic field and the electrical characteristics of the pipeline and the soil are uniform along this length. If there was a change in any one of these parameters at some point along the pipeline (referred to as an electrical discontinuity), an additional voltage peak would be introduced at that point. Although the magnitude of this voltage peak would depend on the nature of the discontinuity, it could possibly create an additional peak of V = ⏐E⏐/Γ. This is an important aspect of electrically long pipelines. To better illustrate the difference between electrically long and electrically short pipelines, consider the case of a pipeline several hundred kilometers long running west across North America from the Atlantic coast. The pipeline is paralleled by a powerline for the entire distance, and the electrical characteristics of the pipeline, powerline, and the earth are uniform along the entire route. Because the pipeline is electrically long, voltage peaks would be created at each end of the pipeline having a voltage of V = ⏐E⏐/Γ. The majority of the pipeline, however, would exhibit a zero voltage. Now consider what happens when the pipeline is extended to the Pacific coast, again assuming that all conditions remain uniform along the length of the pipeline. Even though the pipeline is now perhaps ten times longer, voltage peaks still exist only at the two ends of the pipeline and the magnitude of these voltage peaks remains limited to V = ⏐E⏐/Γ. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:57 The installation of insulators will also introduce additional voltage peaks on an electrically long pipeline. However, if each of the newly created pipe sections remain electrically long, then the insulators will have no effect on reducing the magnitude of these voltage peaks (Figure 3-60)—as was the case with the electrically short pipeline (Figure 3-57). V E Γ 0 − 0 L/2 L E Γ Figure 3-60: Effect of an Insulator at the Midpoint of an Electrically Long Pipeline Also note this fact when installing an insulator on an electrically long pipeline: not only are the voltage peaks not reduced, but the voltage that appears across the insulator will now be double the maximum pipe-to-ground voltage appearing anywhere else on the line. In other words, the voltage across the insulator is now 2⏐E⏐/Γ. Therefore, as was the case with electrical grounds, do not indiscriminately use electrical insulators on a pipeline affected by induced AC interference because this could introduce new voltage peaks. It could also create voltage differences twice as severe as those that could exist between the pipeline and ground. The determination of the LEF, and the calculation of pipeline voltages, will be discussed in detail in Section 3.6 CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:58 Experiment 3-3 To Further Investigate the Effects of Electromagnetic Induction AC Power Line VAC Guy Rod 1 2 3 4 5 6 7 8 9 10 Coated Pipe Section Experiment Schematic No. 1 Procedure Note: This experiment is to be conducted on the buried section of coated pipeline at the NACE field test site or be omitted. The experiment should be conducted by one large group of students. Step: A. Using a high-impedance AC voltmeter (or digital multimeter, with ACV selected), measure the AC voltage on each of the pipe test leads with respect to a ground pin placed near the base of the test station (i.e., repeat the measurements made in Step A of Experiment 3-2). B. Ground the south end of the pipe by connecting a bond cable between the electrical pole guy rod and a pipe test lead (see Schematic No. 1), and repeat the measurements in Step A. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:59 C. Remove the ground connection to the guy rod, and observe the effects of using other types of ground (e.g., pipe casing, sacrificial anode, foreign pipeline). D. Take two 50-foot spools of test lead wire and connect them together. Unspool the wire until there is enough to extend from the first test station to the last test station. E. Suspend the wire in the air by tying it near the tops of the plastic test posts. Ensure that the connection between the two spools is tied to the test post closest to the middle of the pipeline and that the exposed portions of the conductor are well-insulated from earth. Also suspend the spools from the tops of the test stations (See Schematic No. 2). F. Measure the AC voltage to each at both ends of the suspended wire and at the exposed midpoint. G. Ground one end of the wire using a ground pin (or screwdriver) and repeat the measurements in Step F. AC Power Line Excess Lead Wire Test Post Test Lead Wire Experiment Schematic No.2 CP Interference Course Manual © NACE International, 2006 January 2008 VAC AC Interference 3:60 Results AC Voltage (mV) at TS No. Grounded to 1 2 3 4 5 6 7 8 9 10 Guy Rod Anode Bare Pipe Condition AC Voltage Measurement (V) on Suspended Wire South End Midpoint North End No Ground Grounded Questions for Discussion 1. How are pipeline voltages affected by grounding one end of the pipe? Is this consistent with the theory discussed in class? 2. How do the voltages along the suspended wire compare with those induced along the pipeline? Explain any significant difference between the general appearance of the voltage profile or the magnitude of the voltages. 3. What effect does the installation of a small ground electrode have on the voltages along the suspended wire? Can you think of any possible value that these voltage measurements might have when conducting an induced AC voltage investigation? CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3.4 3:61 Deleterious Effects of AC Interference 3.4.1 Electric Shock Hazards There are two types of electrical shock hazards that may be encountered on a pipeline affected by AC interference. The first is a short-duration shock resulting from coincidental contact with the pipe at the same instant that a powerline fault occurs. The second is a sustained shock resulting from contact with a steady-state induced AC voltage. The severity of a shock is dependent both upon current magnitude and duration. Table 3-1 lists the effects of various shock current magnitudes (at 60 Hz) on the human body. Table 3-1: Effects of 60 Hz AC Body Currents on Humans1 Current (mA) Physiological Effect <1 .... No Sensation 1 to 8 .... Threshold of perception, painless 8 to 15 .... Painful, no loss of muscular control 15 to 20 .... Painful, loss of muscular control, can’t let go 20 to 50 .... Painful, severe muscular contractions, breathing difficulties 50 to 100 .... Ventricular fibrillation possible 100 to 200 .... Fibrillation certain, death results without defibrillation >200 .... Severe burns, severe muscular contractions Small electrical shocks can be a nuisance and can potentially cause an involuntary movement that may cause an accident. However, the smallest shock of significance is the current at which a person loses voluntary muscular control and cannot let go of an energized object. The maximum current at which a person can still let go is known as the let-go current. This is considered the maximum safe body current for sustained shocks because a person can withstand repeated exposures to their let-go current without serious after-effects. 1 NACE Recommended Practice RP0177-2000 – “Mitigation of Alternating Current and Lightning Effects on Metallic Structures and Corrosion Control Systems” as referenced in Accident Prevention Manual For Industrial Operations – National Safety Council. W.B. Kouwehoven, Ph.D., “Treatment of Electric Show” In Low Voltage Shock Hazards, The Johns Hopkins University, (June, 1962). CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:62 As current increases beyond the let-go current, shocks become increasingly painful and more dangerous to the body. The most serious shock hazard occurs when currents are capable of causing ventricular fibrillation. When ventricular fibrillation occurs, the electrical signals to the heart become disrupted and the heart stops pumping. Under these circumstances, the heart almost never starts pumping again on its own and death can only be prevented by administering cardiopulmonary resuscitation (CPR) until defibrillation can be applied. A defibrillator uses a strong pulse of DC to put the entire heart into contraction, after which the heart might be restored to its rhythmic pattern. Beginning in the 1930s, studies have been conducted to determine the physiological effects of shock current magnitude and duration. Because it would obviously not be possible to study the effects of excessive currents on human subjects, experiments were conducted on a variety of different animals. Attempts were made to project these results to humans. In the 1960s, Charles Dalziel analyzed this data, and correlated fibrillation currents with body weight (Figure 361). Among his conclusions were that the fibrillation current increases linearly with body weight, regardless of the animal species, and that it also increases with the inverse square root of shock duration. 400 11 CALVES 300 IAVG (DOGS) 25 SHEEP IAVG (4 SPECIES) 3 PIGS 200 0.5% MINIMUM FIBRILLATING CURRENT 107 mA 100 0 0 0.5% MAXIMUM 91 mA NONFIBRILLATING CURRENT 67 mA 20 40 60 80 100 120 BODY WEIGHT - kg Figure 3-61: Fibrillating Current versus Body Weight (Various animals – 3 second shock duration) These conclusions led to the following equations for the maximum tolerable body currents for humans—that is, the body currents above which would result in CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:63 ventricular fibrillation. These equations remain in use today as the basis for designing safe electrical grounding systems. IB = IB = 0.157 ts 0.116 ts (70 kg body) [3-27] (50 kg body) [3-28] Dalziel also conducted additional experiments at low currents on human volunteers. He observed the following information on let-go currents. Table 3-2: Let-Go Currents from Dalziel’s Experiments2 Let-Go Current (mA) Threshold Average Women 6 10.5 Men 9 16 Dalziel also examined the effect of AC frequency on let-go currents. He found that typical power frequencies (i.e., 25 Hz to 60 Hz) happen to be the most severe, as Table 3-3 shows. 2 Charles F. Dalziel and W.R. Lee, “Lethal Electric Currents”, IEEE Spectrum, Vol. 6, No. 2, Feb. 1969, p.45. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:64 Table 3-3: Let-Go Currents from Dalziel’s Experiments3 Median Let-Go Current (mA) Frequency (Hz) Adult Males Adult Females Children 0 76.1 50.3 38.0 5 25.5 16.9 12.7 10 17.3 11.4 8.6 25 15.9 10.5 7.9 60 15.9 10.5 7.9 180 18.3 12.1 9.1 500 19.3 12.8 9.6 1000 24.2 16.0 12.1 2500 35.2 23.3 17.6 5000 51.6 34.1 25.8 10000 74.8 49.5 37.4 A shock current may pass through the body using a number of different paths, the most common of which appear in Figure 3-62. Hand-to-Hand Hand-to-Feet Foot-to-Foot Figure 3-62: Possible Body Current Paths An electric shock can occur when a person touches an energized structure, or even when a person is simply standing in the vicinity of an energized structure that is in contact with the earth. As an example, the structure in Figure 3-63 has become energized to a voltage of 10 kV. The fault current IF passes from the structure to the earth, creating a voltage gradient. A person touching the structure will be exposed to a voltage of 2 kV because this is the potential difference between the structure and the point on the earth where the person is standing. A shock current 3 Charles F. Dalziel et al., “Effect of Frequency on Let-Go Currents”, AIEE Trans. Vol. 62, Dec. 1943, pp745-750. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:65 will pass from the hand, through the body, and to the two feet. The voltage to which the person is exposed in this case is known as the touch potential. It is defined as the potential difference between a grounded metallic structure and a point on the earth’s surface separated by a distance equal to the normal maximum horizontal reach (approximately 1 m). IF Touch Potential = 2 kV Step Potential = 1 kV 10 kV 9 kV 8 kV 7 kV Figure 3-63: Example of Typical Touch and Step Potentials at an Energized Structure A second person, who is not touching the structure, is exposed to a voltage of 1 kV because this is the potential difference between the two points on the earth where the person is standing. The shock current will pass from one foot, through the body, and to the other foot. This voltage is known as the step potential. It is defined as the potential difference between two points on the earth’s surface, separated by a distance of one pace (approximately 1 m), in the direction of maximum voltage gradient. The hand-to-hand current path can occur when two structures that can be simultaneously touched are energized to different potentials—such as across an aboveground insulator. It can also occur when one hand is touching an energized structure and the other hand is touching a structure such as a remotely grounded conductor, which is transferring the potential of remote earth (i.e., zero volts) to the site where the person is standing. This is generally referred to as transferred potential. The hand-to-hand path is typically regarded as the most serious situation because it places the heart directly in the current path. Although equations 3-27 and 3-28 provide the maximum body currents that a person can tolerate, they do not give the maximum tolerable touch and step CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:66 potentials. In order to derive this, one must know the resistance of the current path through the body. It is estimated that the resistance of the internal body tissues is 300 Ω and that the resistance of the body, including the skin, is between 500 Ω and 3000 Ω. A value of 1000 Ω is therefore taken as a body’s approximate resistance, regardless of the current path. In addition to this resistance, one must consider the contact resistance between the feet and the soil. Approximating the human foot as a circular plate having a diameter D, on the surface of soil having a resistivity ρ, the resistance of the foot is calculated as: R = ρ 2D [3-29] For a foot having a diameter of 6.5 in (16.7 cm), the foot would have a resistance of 3ρ, where ρ is given in Ω-m. Therefore, the resistance of two feet in parallel, R2Fp (in the case of a touch potential), or two feet in series, R2Fs (in the case of a step potential), can be calculated as: R2Fp = 1.5ρ [3-30] R2Fs = 6ρ [3-31] The tolerable limits for touch and step potential can now be calculated using Ohm’s Law: V = R×I [3-32] V = ( Rbody + Rfeet) × Ibody [3-33] Therefore, for a person weighing 50 kg Vstep50 = (1000 + 6ρ) 0.116 Vtouch50 = (1000 + 1.5ρ) CP Interference Course Manual © NACE International, 2006 January 2008 tS 0.116 tS [3-34] [3-35] AC Interference 3:67 And a person weighing 70 kg Vstep70 = (1000 + 6ρ) 0.157 tS Vtouch70 = (1000 + 1.5ρ) 0.157 tS [3-36] [3-37] Note that these calculations do not assume any additional resistance that might be contributed by footwear. Also, consider all possible body current paths to be of equal detriment. Example Calculation: Calculate the maximum tolerable touch and step potentials for a 70-kg man standing on 100 Ω-m soil, exposed to a fault having a 0.1-s duration. Vstep70 = (1000 + 6ρ) Vstep70 = (1000 + 6 ⋅ 100) 0.157 tS 0.157 = 794V 0 .1 Vtouch70 = (1000 + 1.5ρ) Vtouch70 = (1000 + 1.5 ⋅ 100) 0.157 tS 0.157 = 571V 0 .1 Because the probability of a fault is low and the duration of a fault is very short (typically <0.1s), the probability of contacting the pipeline or an appurtenance at the moment a fault occurs is very low. Therefore, systems that are designed to protect the safety of personnel in the event of a fault are designed with the goal of preserving life; designing with a higher goal, such as preventing a painful shock, is not easily justifiable. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:68 In the case of an induced steady-state voltage that is always present on the pipe, the mitigation system must be designed so that contact with the voltage will, hopefully, not prevent a person from letting go and will definitely not cause breathing difficulties. A value of 15 mA has been chosen as the maximum sustainable body current. Assuming a minimum body resistance of 1000 Ω, the maximum steady-state AC voltage is determined by Ohm’s Law as: Vss = Iss × Rbody [3-38a] = 15mA × 1000Ω = 15 V This 15-V limit for steady-state induced AC pipeline voltages has been adopted by various standards, including NACE RP0177 and the Canadian Standards Association’s CSA C22.3 No.6. Some sources note that the value of 15 V actually comes from a maximum sustainable body current of 10 mA and an assumed body resistance of 1500 Ω; however, the end result is the same. Also remember that, even though 15V is a well-recognized safe limit, there are cases where even 15 V would be considered excessive—such as where small children could reasonably be expected to come in contact with this voltage. 3.4.2 AC Corrosion 3.4.2.1 Theory Corrosion of steel by alternating current was investigated as far back as the early 1900s. A comprehensive study by the U.S. Bureau of Standards,[4] concludes that AC corrosion decreases with increasing frequency, does not occur beyond a limiting frequency between 15 and 60 Hz, and results from irreversibility—during the negative half cycle—of the corrosion that occurs during the positive half cycle. Their results, for iron electrodes exposed to normal soils at various frequencies, appear in Figure 3-64a. Hence, the amount of corrosion is expressed as a “coefficient” percentage of the amount of corrosion that would be caused by an equivalent amount of DC. All electrodes were operated at an AC current density of 5 A/m2. At 60 Hz, the coefficient was less than 1 percent under natural soil considerations. 4 B. McCollum and G.H. Ahlborn, Technological Papers of the Bureau of Standards, No. 72 Influence of Frequency of Alternating or Infrequently Reversed Current on Electrolytic Corrosion, Washington, DC, Aug. 1916. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:69 100 90 80 70 LEGEND: Soil Soil + Na2 CO3 60 50 40 30 20 10 0 -10 1/60S 1/15S 1S 5S 1M 5M 10M 1Hr. 2Days 2Weeks D.C. Logarithm of Length of Time of One Cycle Figure 3-64a: Coefficient of Corrosion at Different Frequencies for Iron Electrode Denoted as Average Electrode Loss (McCollum and Ahlborn, 1916) A number of investigators examined the effects of AC on cathodically protected steel. Nearly all of them agreed that AC corrosion could be overcome with CP. Bruckner (1964) observed that CP reduced AC corrosion to negligible values, but the DC current density of 0.42 to 0.53 A/m2 was considered much greater than appears necessary in practice. Hewes (1969) stated that in the corrosion rate, being in the order of 0.1% of an equivalent magnitude DC, is readily overcome by normal cathodic protection procedures. Even much more recently, Hamilin (1986) concluded that metals under the influence of AC can be cathodically protected, but usually at higher current densities. Then, in Germany, two corrosion perforations occurred on a polyethylene-coated gas pipeline that was installed in 1980 parallel to an AC (16-2/3 Hz)-powered rail transit system. A subsequent investigation, as reported by Prinz, attributed the corrosion to induced AC arising from the transit system operation. At the corrosion sites the polarized potential from the CP system was –1000 mVcse and the corrosion product pH was 10. These figures indicate that the CP system was operating adequately with respect to current industry standards. A CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:70 potential gradient survey indicated additional coating holidays. Upon excavation, the holidays revealed “crater-like” corrosion pits underneath corrosion product “bulges” that had not been observed before but whose appearance was apparently consistent with similar observations on other pipelines in Germany and Switzerland. The relatively low soil resistivity of 1900 Ω-cm was a result of deicing salt contamination. A steel rod coupon having a holiday surface area of 1 cm2 was installed and monitored for a period of 220 days before removal for examination. Despite a CP current density of 1.5 to 2 A/m2 and a resulting “ON” potential of –1800 to –2000 mVcse, the coupon exhibited pitting corrosion at a rate of 210 mpy caused by an AC current density that varied from 20 to 220 A/m2. Funk conducted laboratory tests using 10-cm2 coupons in synthetic soil solutions subjected to AC current densities of 100 and 50 A/m2 and field tests using coupons in both sandy and clay soils at AC current densities of 10 to 30 A/m2 and 300 to 1000 A/m2, respectively. A test coupon was perforated after 168 days at an AC current density of 100 A/m2, and corrosion rates greater than 42 mpy (1 mm/a) were observed. After these preliminary results, additional testing to better define the influence of current density was carried out. The testing indicated that AC current densities greater than 30A/m2 caused corrosion rates greater than 4 mpy (0.l mm/a) at a constant CP current density of 2 A/m2. The corrosion rates increased with increased AC current density but decreased with time. Helm conducted short-term tests (up to 1000 hours) and long-term tests (up to 1 year) in flowing and stagnant waters while varying the AC and DC current density in an attempt to establish an effective corrosion control criterion for pipelines exposed to AC. They concluded that, with up to 20 A/m2 of AC, there is “probably no risk” of accelerated corrosion using the conventional criteria. They found that corrosion is possible between 20 and 100 A/m2 because the conventional criteria are unreliable. Moreover, they noted that corrosion damage is to be expected when AC current densities exceed 100 A/m2. Gustav Peez reported corrosion rates of up to 55 mpy (1.3mm/a) at current densities from 100 to 200 A/m2. In addition, field inspections on the Erdgas Sudbayem (ESB) gas pipeline system indicated that corrosive attacks—starting at an AC current density of 15 A/m2—could not be ignored. Field inspections carried out by Hartmann at identified coating holidays on the 30.8-km Hunze-Hambom gas pipeline revealed corrosion pits after 2½ years in operation of 42 mils (1 mm) in 20,000 Ω-cm sandy soil at AC current densities of 74 to 165 A/m2, which is an average corrosion rate of approximately 17 mpy. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:71 Increasing the CP current density from 2A/m2 to 5A/m2, as determined by Funk, decreased the AC corrosion rate at an AC current density of 50A/m2 by at least one half. Using results on test specimens in flowing waters, Helm found that CP current densities up to 0.25 A/m2 had no mitigating effect. However, he did observe a demonstrable benefit at 4 A/m2. Helm also found no detectable difference between 16 2/3 Hz and 50 Hz in flowing water at an AC current density of 10 to 20 A/m2 and a DC current density of 0.2 A/m2. Helm’s conclusion is similar to McCollum and Ahlborn’s findings (Figure 3-64), which show no significant difference in corrosion coefficient between 15 and 60 Hz. AC corrosion rates appear to be dependent on the type of environment. Both Prinz and Helm indicate that the presence of sodium bicarbonate (NaHCO3) and calcium carbonate (CaCO3) increases corrosion whereas sodium chloride (NaCl)containing media seem to inhibit corrosion. This accelerating effect of carbonates was also apparent at 60Hz in the McCollum and Ahlborn study. Flowing water produced a higher corrosion rate than stagnant water of the same composition according to Helm. This was ascribed to the enhanced supply of Ca++ and HCO3– ions to the surface. Tests on low-alloy steel specimens in 0.1 N NaCl solutions by Jones indicated that the corrosion rate compared to the control, at an AC current density of 300A/m2, was unaffected in aerated conditions but increased by a factor of five in the deaerated conditions. Bertocci also demonstrated, based on polarization theory, that when the cathode is under diffusion control—such as what one might expect in aerated conditions—corrosion acceleration would be minimized. Bruckner deduced that the AC corrosion rate in deaerated conditions was greater than for aerated conditions, although he was unable to explain this result. Frazier and Barlo found that corrosion rates on steel coupons at AC current densities in the order of 1000A/m2 varied substantially in two different simulated groundwaters as well as when the groundwater was deaerated. AC had a greater corrosion accelerating effect in a clay soil as compared to mineral waters, according to Pookote and Chin. Figure 3-64b clearly indicates that the corrosion rate decreases with time regardless of the AC current density. Williams’ corrosion studies, conducted in the absence of CP, also verified that the AC corrosion rate decreases asymptotically with increased time. Prinz, however, reported that there was an “incubation” time of 30 and 120 days for AC current densities of 100 and 50A/m2, respectively, after CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:72 which the corrosion rate increased; but, this has not been reported elsewhere. Furthermore, these short time test periods would not have any significant relation to a pipeline in the long term. 0.50 iac = 100 A/m2 sheet metal specimens pipe specimens 0.40 iac = 50 A/m2 0.30 iac = 30 A/m2 0.20 0.10 0.00 0 1 2 3 4 5 6 7 8 9 10 11 12 Time / 103 (hr.) Figure 3-64b: Maximum Penetration Depth as a Function of Test Duration at Constant Cathode DC Current Density (2A/m2) and Differing AC Current Density (Funk et al., 1992) Another time factor is the general increase in resistance with time and a consequent decrease in AC current density as reported by two investigators (Williams; Bruckner) when a constant AC voltage is applied. Because this type of AC situation closely simulates actual field conditions, it implies that lower corrosion rates are to be expected in practice as time increases. The surface area of the pipe at a coating holiday should be important because the corrosion rate increases with increasing current density and, hence, large holidays would therefore have a lower current density than smaller holidays if both are exposed to the same soil conditions. In this regard Peez reported on observations made at a number of holiday sites on the Erdgas Sudbayem system. The observations indicated that the majority of the corrosion occurred at holidays having approximately l cm2 surface area. A second paper (Heim and Peez), based on the same investigation, reported that no corrosion activity was observed at small holidays of 0.01 cm2 and only minimum corrosion at one of two sites having CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:73 a holiday area of 0.03 cm2. The lack of corrosion attack at the smaller holidays was attributed to obstruction of these relatively small openings. Prinz recounted that, when coupons having surface areas ranging from 0.5 cm2 to 5 cm2 were buried next to an AC-affected pipeline, the highest corrosion rate occurred on the 1-cm2 coupon. Unfortunately, the existing literature does not definitively describe the actual AC corrosion mechanism. McCollum and Ahlborn generally reasoned that AC corrosion resulted from the irreversibility of the corrosion reaction such that metal ions created during the anodic half cycle were not re-plated during the negative half cycle. Although this was equated to a rectification effect, Williams concluded that the corrosion mechanism was not rectification but rather a sole result of the positive half cycle. Bruckner thought that the observed AC corrosion may have been partially a result of “thermal activation,” although Pookote and Chin—who attempted to investigate the influence of temperature on the rate of corrosion— were unable to draw a firm conclusion because of scattered data. Bertocci explained the relatively low corrosion efficiency of AC compared to DC by demonstrating that the majority of the sine function AC and higher-frequency harmonics are shunted by the double-layer capacitance “without causing material transport across the electrode interface.” He also showed this effect could be particularly pronounced under diffusion-controlled (i.e., aerated) conditions. Jones, in explaining why there was greater corrosion acceleration on steel in deaerated rather than aerated environments, demonstrated that superimposed AC caused depolarization of the anodic reaction. He inferred that this could be caused by anion desorption or surface film reduction during the cathodic half cycle. Similarly, Chin and Fu were able to show a breakdown in anodic passivity with increasing 60Hz current density by using anodic polarization tests on mild steel electrodes in a pH 7, 0.5M sodium sulfate (Na2SO4) solution. Hamlin concluded, however, that “AC does not have any significant effect on the polarization or depolarization of cathodically protected steel⎯”. In contrast, Lalvani and Lin were able to show that the corrosion characteristics can be classified in terms of the ratio of the anodic and cathodic Tafel slopes by generating a number of potentiodynamic polarization curves. The extreme complexity of determining all the variables influencing AC corrosion is apparent in the literature, and several investigators emphasized the need for additional research. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:74 More recently, it has been found that excessive amounts of CP can actually increase AC corrosion rates. In laboratory experiments conducted by Neilsen[5], it has been found that very electronegative CP potentials (e.g., -2200 mV on w.r.t. CSE) resulted in increased AC current densities and increased AC corrosion rates (figures 3-65a and 3-65b, respectively). This has been attributed to the lowering of the electrolyte resistivity immediately adjacent to the site of the holiday, which coincides with the high pH resulting from the increased level of CP. Figure 3-65a: Effect of CP Potential on AC Corrosion Rate 5 L.V. Neilsen, "Role of Alkalization in AC Induced Corrosion of Pipelines and Consequences Hereof in Relation to CP Requirements." Paper 05188, NACE 2005, Houston, TX. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:75 Figure 3-65b: Effect of CP Potential on AC Current Density It has also been recently reported that the effect of the ionic composition of the soil, as discussed earlier, may influence the AC resistance of the holiday—and, therefore, the AC current density[6]. If earth alkaline ions such as Ca2+ and Mg2+ are present in the soil, CP will result in the formation of calcareous deposits at the holiday site that can increase the holiday resistance by a factor of 100. In contrast, if alkaline ions such as Na+, K+, and Li+ are present, this may result in the formation of highly soluble hydroxides that can lower the resistance of the holiday by a factor of 60 times. The ratio of alkali ions to earth alkali ions can therefore result in a range of holiday resistances spanning three orders of magnitude. At the time of this writing, the one factor whose importance cannot be disputed is the effect of AC current density. AC current density can be calculated for a circular holiday, by combining the formulae for the resistance and surface area of a circular disk with Ohm’s Law, as follows: iAC = 6 8VAC ρπd [3-38b] Technical Specification - Corrigendum to EN12954, "Evaluation of AC Corrosion Likelihood of Buried Pipelines - Application to Cathodically Protected Pipelines." Comité Européen de Normalisation, TC219 - WG 1 - Ad Hoc Group 4, Dec. 2004. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:76 where: iAC VAC ρ d = = = = AC current density (A/m2) pipe AC voltage to remote earth (V) soil resistivity (ohm-m) diameter of a circular This equation can then be used in conjunction with the conclusions of the German investigators (Prinz; Funk, et al.; Helm, et al.; Peez; Hartmann), which were as follows: • At AC current densities of less than 20 A/m2 there is no AC induced corrosion, and • AC corrosion is unpredictable at AC current densities of between 20 A/m2 and 100A/ m2, and • At AC current densities of greater than 100A/ m2, corrosion is to be expected, and that • The highest corrosion rates are found at holidays having a surface area in the range of 1 to 3 cm2. The magnitude of the AC current densities may at first seem unusually high; but, in low-resistivity soils, relatively low AC voltages can produce high current densities when the surface area of the coating holiday is small. For example, the AC voltage required to produce a current density of 100 A/m2 in 1000 Ω-cm soil at a 1 cm2 holiday (d = 0.011 m), would be: VAC = iAC ρπd 100 ⋅10 ⋅ π ⋅ 0.011 = = 4.4V 8 8 It is clear from the foregoing calculation that cathodically protected pipelines subjected to AC voltages—that are below the maximum safe operating level of 15 V—can suffer from AC corrosion at holiday sites having a surface area of approximately 1 cm2 in a soil resistivity of 3000 Ω-cm or less. 3.4.2.2 AC Corrosion Case Histories Several selected AC corrosion case histories that have been reported in the literature follow.[7, 8] 7 R. Wakelin, R. Gummow, S. Segal, “AC Corrosion – Case Histories, Test Procedures, and Mitigation.” Paper 565, NACE 1998, San Diego, CA. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:77 Case History No. 1 In 1991, after only four years of operation, a corrosion failure occurred on a 300mm diameter high pressure gas pipeline in Ontario, Canada. The pipeline was coated with a double-layer of extruded polyethylene (PE), and its joints were fieldcoated with a hot-applied coal tar tape. The pipeline was cathodically protected using distributed magnesium anodes. It exhibited on potentials ranging from -1.45 VCSE to -1.50 VCSE. A high-voltage AC power line 14 m away from the pipeline paralleled the pipeline’s entire 4400-m length. Induced AC voltages had been mitigated by coupling the pipeline through capacitors to the station piping at each end of the pipeline and to ground rods installed at the test stations. Voltages typically ranged from 6 V to 10 V, but they often rose to 26 V when capacitors failed. At the time the failure was being investigated, the AC voltage was 28 V. The leak occurred at a joint under the center of a four-lane roadway. A 50-mmdiameter pit cluster was found immediately adjacent to the weld at the eighto’clock position, within which a pinhole perforation was found (Figure 3-65c). Figure 3-65c: Pit Cluster and Pinhole Perforation (Case History No. 1) 8 R. Wakelin, C. Sheldon, “Investigation and Mitigation of AC Corrosion on a 300 mm Diameter Natural Gas Pipeline.” Paper 4205, NACE 2004, New Orleans, LA. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:78 Based on a nominal pipe wall thickness of 5.56 mm, the penetration rate of this pit was calculated to be 1.4 mm/y (55 mils/y). A second pit cluster discovered adjacent to the weld at the four-o’clock position, was similar in appearance and slightly larger in size but did not penetrate the pipe wall. The section of pipeline crossing the roadway had been fabricated from two lengths of pipe that were welded together above grade. The joint was coated before the pipe was installed through an open cut in the road. Although the crew that repaired the leak had noted that the coating was intact except at the leak site itself, the location and symmetrical appearance of the two pit clusters suggest that the coating had indeed been damaged during installation—perhaps by the pipeline boom grip used to lower the pipe into the trench. The soil at the failure site was a dark brown sandy clay having a pH of 8.8, an electrical resistivity of 130 Ω-cm, and a chloride ion concentration of 3600 ppm. Because the pipeline crosses the roadway on a hill, the high-chloride ion concentration was no doubt caused by the frequent application of de-icing salts. The soil on either side of the roadway exhibited lower chloride ion concentrations (50 ppm to 500 ppm) and correspondingly higher resistivities (1000 Ω-cm to 4800 Ω-cm). An inspection of the pipe joints at each side of the roadway found no evidence of corrosion damage—even though the joint coating had been poorly applied, resulting in the ingress of moisture between the pipe and coating. The cause of the corrosion could not be identified because the failure site had been disturbed by the emergency repair crew prior to the investigation. It was originally speculated that the failure could have been the result of an occluded cell (because of the high chloride ion concentrations in the soil) or that bacterial corrosion could have been responsible (because sulfides were found in the corrosion products). In retrospect, it is more likely that the corrosion was ACinduced—a result of the extremely high AC current density calculated for the pit site. Using Equation 3-38b, the AC current density at the pit was calculated to be 1100 A/m2. This current density is based on a voltage of 28 V, a soil resistivity of 130 Ω-cm, and an assumed holiday diameter equal to the pit diameter of 50 mm. It is well in excess of the 100-A/m2 threshold above which AC corrosion is expected to occur. Case History No. 2 A 250-mm-diameter oil products pipeline, coated with extruded PE, was installed in 1976. It was cathodically protected by a jointly operated impressed current system serving a number of pipelines that shared the power line corridor across the CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:79 north end of Toronto, Ontario. An internal inspection tool passed through the piping in 1986 identified a minor anomaly (much less than 25% wall penetration). A subsequent internal inspection in 1994 indicated a significant anomaly at this same location. Upon excavating and examining the pipeline, a crater-like corrosion pit was found. The pit was 50 mm long by 45 mm wide by 6.9 mm deep. It penetrated through 88% of the pipe’s 7.8-mm-thick wall. Assuming that the pit grew from 25% to 88% penetration over the eight years between internal inspections, this translates to an average corrosion rate of 0.61 mm/y (24 mils/y). Pipe-to-soil potentials more electronegative than -1.27 VCSE on had been recorded in the vicinity of this corrosion site during previous surveys, thus indicating a satisfactory level of protection. Moreover, the potential with the reference electrode located at the pit was -1.18 VCSE compared to -1.47 VCSE on with the electrode located at grade level. The soil adjacent to the pit exhibited an electrical resistivity of 300 Ω-cm, a chloride ion concentration of 1920 ppm, and tested negative for sulfides. The high chloride concentration was attributed to the application of de-icing salts because this corrosion anomaly was located under the edge of the roadway. The reddish brown corrosion product tested negative for sulfate-reducing bacteria and had a pH of 10.7, compared to a pH of 8.0 for the bulk soil. The induced AC voltage at this location was 15 V at the time of this investigation; it was 12 V during the previous year’s CP survey. Using Equation 3-38b, the AC current density was calculated to be 200 A/ m2 at 12 V—well above the 100-A/ m2 threshold value. Except for the possibility that this was an occluded corrosion cell, the cause of corrosion was concluded to be induced AC current. Case History No. 3 A 500-mm-diameter, 74-km-long, coal tar-coated high-pressure natural gas pipeline was paralleled for 40 km by a high-voltage AC power line. In 1995, an internal inspection identified a number of anomalies—all of which were located along the power line right-of-way. Defects estimated to have a corrosion depth of greater than 40% of the 7.1-mm wall thickness were concentrated along two discrete pipe sections. Two of these anomalies located were estimated to penetrate 80% and 63% of the pipe wall, respectively. A review of pipe-to-soil potential data collected during test station surveys and close-interval surveys since the pipe’s installation in 1972 indicated no evidence of CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:80 subcriterion potentials along this pipeline. Rectifier data for the influencing impressed current systems indicated that there were only 10 days of downtime since 1983, all occurring in 1994 while an AC mitigation system consisting of banked magnesium anodes was being installed. At the first dig site, three corrosion anomalies were investigated. The deepest was a smooth and generally round corrosion pit, having a diameter of 5 cm and a maximum depth of 6.1 mm (86% of the wall thickness). The adjacent soil was moist clay having a resistivity of 2000 Ω-cm. A hard, tightly adhering tubercle protruding 5 cm above the pipe’s surface covered the pit. The coating around the pit was disbonded over a 20-cm radius. A pH test using litmus paper indicated that the pH was greater than 8.5 at this pit, as well as at two smaller pits examined in the vicinity. Furthermore, there was no evidence of bacterial corrosion and the local CP potential was -1.56 VCSE on. Accordingly, there was no apparent cause for the observed corrosion. The AC current density at this site was calculated to be 84 A/ m2 at 33 V, which was the average induced AC voltage prior to the installation of the AC mitigation system in 1994. At the second dig site, three additional anomalies were investigated. The deepest of these was found at the two-o’clock position beneath a large hemispherical shell of extremely hard soil, approximately 15 cm thick (Figure 3-65d). The pit was 56 mm in diameter by 6.34 mm in depth (89.3% penetration) and was smooth and dish-shaped. A 25-mm-diameter steel pipe was found to be wedged against the pipe at the pit. The pH of the soil immediately adjacent to the pit was 8.2, the soil resistivity of the moist clay soil was 1350 Ω-cm, and only trace amounts of chlorides and sulfides were found. The pipe-to-soil potential with the reference electrode located on top of the corrosion product was -1.05 VCSE, compared to -1.49 VCSE on with the reference at grade. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:81 Figure 3-65d: Hemispherical Shell of Hardened Soil Surrounding Anomaly (Case History No. 3) Although the small steel pipe, if in contact with the pipeline steel, would electronically shield the pit from receiving CP current, it is likely that the electrical contact was broken as corrosion progressed—thereby eliminating any shielding effect. In such a case, however, the small pipe would serve to focus both CP current and AC current at the pit location because of its low resistivity compared to that of the surrounding soil. The pipe-to-soil potential suggests that the pit did not lack for CP current. Furthermore, the possibility of bacterial corrosion was dismissed because the CP level was more electronegative than the -0.95 VCSE criterion generally considered sufficient to prevent corrosion from sulfate-reducing bacteria. The other anomalies investigated at the second dig site were both similar in appearance to the first. Both were smooth, round, and dish-shaped, and both were initially covered in a hemispherical shell of hardened soil. One of these actually consisted of two pits located immediately beside one another (Figure 3-65e). CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:82 Figure 3-65e: Hemisphere of Hardened Soil and Corrosion Pit (Case History No. 3) Before the distributed AC mitigation system was installed on this pipeline in 1994, the AC voltage at the second dig site was approximately 25 V, resulting in a calculated AC current density of 84 A/ m2. This is identical to the AC current density calculated for the deepest pit at the first dig site; coincidentally, the corrosion rates were nearly identical (0.27 and 0.29 mm/y). Accordingly, it was concluded that the observed corrosion at both dig sites was AC-induced. Case History No. 4 A 20-km-long, 300-mm-diameter pipeline coated with side-extruded PE and cathodically protected by banks of magnesium anodes was installed in New York State in 1991. In its eleventh year of service, the pipeline experienced a corrosion failure close to a weld at the seven-o’clock position on the pipe (corrosion rate of 24 mils/year). A pinhole perforation occurred near the center of a smooth, round, dish-shaped pit that was approximately 25 mm in diameter (3-65f). The failure was attributed to corrosion, although CP records for a test station located only a few meters from the failure site suggest that the pipeline had been well-protected from the time it was constructed (potentials more electronegative than -1.3 VCSE on). Furthermore, there was no evidence to suggest that the failure site had been shielded from receiving protective current. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:83 Figure 3-65f: Pinhole Corrosion Failure Following Removal of Repair Clamp (Case History No. 4) The failure occurred in a rural area on the south side of a paved two-lane roadway where the pipeline diverges from a major electric powerline . The pipeline is paralleled by four single-circuit 345-kV powerlines for 5300 m and a doublecircuit 115-kV power line for 9400 m (Figure 3-65g). A study conducted at the time of construction indicated that peak voltages could reach 300 V where the pipeline entered and exited the 345 kV corridor; as a result, magnesium anode beds were installed at critical locations. Despite these measures, AC voltages along the pipeline remained high—with peak pipeline voltages of 50 V and 80 V at the south and north ends of the shared 345-kV corridor, respectively. These locations are referred to throughout the paper as sites A and B, respectively, where Site A is also the site of the failure. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:84 4 - 345 kV Single Horizontal Circuits Oswego B 9400 m A 5300 m 5300 m 115kV Double Vertical Circuit North Volney Figure 3-65g: Pipeline-Powerline Route (Case History No. 4) One month after the failure occurred, a site visit was made to collect additional data and to attempt to find other corrosion sites where the corrosion mechanism might be identified. Although a large magnesium anode was located immediately adjacent to the failure, the anode bed was found to be disconnected from the pipeline at its test station. It was later discovered that the anode bed had been disconnected two years before the failure; this was evident because the open-circuit potential of the anodes (-1.30 VCSE) was slightly less electronegative than the pipe’s on potential (-1.35 VCSE) and the anodes were receiving rather than contributing CP current. The anode bed was temporarily reconnected to the pipe, resulting in a current discharge of 9.6 A AC to ground. This lowered the AC voltage of the pipe at the test station from 33 V to 22 V; but, when measured with respect to remote earth, the pipe voltage was only reduced from 53 V to 47 V. At the north end of the collocation between the pipeline and the 345-kV powerlines (Site B), the AC pipe voltage was 46 V—although voltages as high as 80 V had been recorded. A large magnesium anode was installed at this location when the pipeline was constructed, but it was found to have no AC or DC current output at the time of this investigation. It was considered unlikely that additional corrosion damage, similar to that which had resulted in the failure, could be found by conducting random excavations of the pipeline. The pipeline nevertheless was excavated at the first joint CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:85 immediately adjacent to where the failure had occurred, with the hope of finding and investigating additional corrosion sites. A small nodule was found at this site. Measuring approximately 20 mm in diameter by 20 mm in height, it was found on the top of the pipe beneath a 10mm-diameter break in the coating. Immediately after uncovering this anomaly, the potential of the pipe was measured by placing a copper sulfate reference electrode directly on top of the coating holiday. The pipe potential was found to be -1.08 VCSE. The pH inside the nodule was measured using a combination pH/reference microelectrode and a high-impedance meter; it was found to be 17.0. Because the pH meter was calibrated using buffer solutions of pH 4.0, 7.0, and 10.0, it is expected that this unrealistically high pH value was the result of not using a high-pH buffer solution during calibration. However, the pH of the nodule was later measured in the lab and found to be 13.1. The coating was removed from the vicinity of the anomaly, and the nodule was removed for analysis. A corrosion pit was found immediately beneath where the nodule and coating holiday had been. The pit was round, smooth, and dishshaped. It measured 30 mm in diameter by 1.35 mm deep. Upon further investigation, a second nodule was found on top of the pipe. Nearly identical in appearance to the first, it was located approximately 600 mm to the west of Anomaly Nº 1 (Figure 3-65h). In this case, the nodule was protruding through the pipe’s factory-applied PE coating. Once again, measurements were taken within the break in the coating. The pipe potential was found to be -1.17 VCSE and the pH inside the nodule was 12.2. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:86 Figure 3-65h: Nodule of Corrosion Products Protruding Through Coating (Case History No. 4) After removing the coating, the nodule of corrosion products was retrieved and the pipe surface was cleaned. The cleaning revealed another round, dish-shaped corrosion pit that measured 40 mm in diameter by 2.3 mm deep (Figure 3-65i). Figure 3-65i: Corrosion Pit after Removal of Coating and Corrosion Products (Case History No. 4) CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:87 The corrosion products from the two pits, as well as those obtained from the failure site, were analyzed for chloride ion content and pH. Two soil samples were also retrieved from the excavation close to the sites of the two anomalies. They were analyzed for the characteristics that are typically considered to influence corrosivity. Aside from the two coating holidays at which the corrosion pits were found, both the factory-applied and the field-applied coatings were found to be in good condition. They exhibited excellent adhesion to the pipe. A portion of the anode bed was also excavated. The intent was to remove one anode for analysis, to determine the cause of the low open-circuit potential of the magnesium. An anode was located and was removed from the excavation. Upon closer examination, however, no magnesium was found to remain. All that was left of the 14.5-kg anode was its steel strap core, a small block of white magnesium oxide, and the select backfill from the anode package. The CP potentials measured directly on top of the two anomalies (-1.08 VCSE and 1.17 VCSE) indicate that the pipe surface exposed at the coating holidays was wellprotected. This was further verified by the high pH values of the nodules as measured both in the field and afterward in the lab. There was no evidence that the pipe was ever shielded from receiving CP current, either by disbonded coating or by rocks in the soil. Furthermore, there was no evidence that DC stray current interference problems had ever affected the pipeline; a recent close-interval potential survey indicated that no such problems exist at this time. Bacterial corrosion was also ruled out as a possible cause because CP potentials more negative than -0.95 VCSE should be sufficient to prevent bacterial corrosion. At both anomalies, the opening in the coating was approximately 10 mm and the pipe voltage was approximately 50 V; however, a peak of 60 V was recorded during the site visit. The soil resistivity in the excavation varied from 12.5 Ω-m to 16.0 Ω-m. Using the more conservative values of 50 V and 16 Ω-m for the calculation, the AC current density at the anomalies is found to be 800 A/ m2— well above the threshold value of 100 A/ m2 at which AC corrosion can occur on a cathodically protected pipeline. Note that the soil resistivities measured in the excavation were significantly lower CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:88 than the bulk resistivities measured along the pipeline route using the Wenner Four-Pin Method (92 Ω-m to 206 Ω-m, 160 Ω-m average). This is attributed to the moderately high chloride ion content of the soil at the road crossing (190 ppm to 215 ppm), presumably caused by the application of de-icing salts to the roadway during the winter months. As is generally found to be the case, the chloride ion contents were even higher in the corrosion products than in the surrounding soil; they ranged from 275 to 470 ppm. This may be because the negatively charged chloride ions are attracted to the corrosion site by the surplus of positively charged iron ions. Alternatively, the chloride ions may become trapped in the corrosion product matrix and accumulate during subsequent wetting and drying of the corrosion products; a similar scenario occurs in the cement mortar of buried concrete piping. The highest chloride ion content was found in the corrosion products taken from the failure site, which is consistent with this site having a higher corrosion rate than either of the two anomalies. From the potentials and currents measured at the two primary anode beds on the pipeline (sites A and B), and from the examination of one magnesium anode, these anode beds appeared to have been totally consumed. Typically, a 14.5-kg anode installed on a well-coated pipeline in moderately high-resistivity soil should have a life in excess of 20 years; however, at high AC current densities the life of an anode can be significantly reduced. Considering that the anode bank at Site A was disconnected two years before the failure occurred, the life of this anode bed was only eight years or less. The most effective method of mitigating AC corrosion is to lower the pipeline voltages to acceptable levels. In the moderately high-resistivity soil (i.e., 150 Ωm) that exists along the majority of the pipeline route, 30 V would be required to produce AC current densities of 50 A/m2. Where chloride contamination of the soil exists, however, pipeline voltages must be reduced to much lower values. In order to determine the requirements for the AC mitigation system, the pipelinepowerline corridor was modeled using software developed as a research project by the Pipeline Research Council International (PRCI). With no mitigation in place, the modeling predicted voltage peaks of approximately 200 V where the pipeline enters and leaves the 345-kV powerline corridor (Figure 3-65j). AC voltages are most effectively mitigated by installing ground electrodes at major electrical discontinuities along the pipeline-powerline corridor because this CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:89 is where the peak voltages are generated. In order to reduce the peak voltages below 30 V (to reduce AC current densities below 50 A/m2), it was determined that an 800-m-long horizontal wire electrode would need to be installed at each of the two critical locations. In order to further reduce these voltage peaks to 15 V for electrical safety reasons, the ground electrodes would each need to have a resistance of 0.1 Ω, which could not be realistically achieved, given the local soil conditions. Induced AC Pipe Voltage (V) 1000 100 1 2 15 V 10 3 A B 1) No Mitigation 2) 0.5 Ω at A & B 3) 0.1 Ω at A & B 1 0 5 10 15 20 Distance from North Volney (km) Figure 3-65j: Effects of Installing Ground Electrodes at sites A and B (Case History No. 4) In addition to the ground electrodes, DC decouplers were to be installed across the insulators at each end of the line to further reduce the induced voltages. A detailed survey of soil resistivities was first conducted using an electromagnetic soil conductivity meter, which can identify pockets of low-resistivity soil that might be AC corrosion hot-spots. Using Equation 3-38b in conjunction with the close-interval soil resistivity data, the AC current densities were calculated assuming the worst case of a 1-cm2 holiday (Figure 3-65k). Note that detailed soil resistivity data are only available between the chainages of 4 km and 10.5 km. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:90 AC Current Density (A/sq.m) 1000 1 100 2 50 3 10 A B 1 0 5 10 1) No mitigation 2) 400 m wires at A & B 3) 800 m wires at A & B; DC decouplers at insulators 15 20 Distance from North Volney (km) Figure 3-65k: Effects of Installing Ground Electrodes on AC Current Densities (Case History No. 4) It is important to note that the mitigation wire also provides a secondary benefit in the mitigation of AC corrosion, just as it does in the mitigation of AC voltages. For a coating holiday located in the vicinity of the mitigation wire, the effective resistance of the holiday is increased because of the mutual resistance between the holiday and mitigation wire; the increase thereby reduces the AC current density at the holiday to a value less than that predicted by Equation 3-38b. The resistance of a 1-cm2 circular holiday can no longer be calculated using the equation for the resistance of a circular disk because its resistance is now closer to that of a 1-cm2 area on the surface of the mitigation wire. This mutual resistance effect can be thought of in terms of anode interference effects, where the installation of one anode close to another effectively raises the individual resistances of both anodes. In view of this, it is quite likely that the use of two 800-m-long mitigation wires would reduce AC current densities at all locations along the pipe to 50 A/ m2 or less. Recall that the corrosion failure occurred immediately adjacent to a large magnesium anode bed but that the prematurely consumed anode bed had been disconnected from the pipeline approximately two years before the failure occurred. Even though the anode bed was not contributing CP current and its effectiveness as an AC ground electrode was reduced, it should still have been effective in limiting the magnitude of the AC current density at nearby holidays. It is therefore suggested that the majority—if not all—of the pit growth may have actually occurred over a two-year period rather than a 10.5-year period and that CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:91 the AC corrosion rate may have been as high as 125 mils/y. A mitigation wire typically consists of a ribbon of sacrificial anode material installed parallel to the pipeline, and connected directly to the pipeline. In the case of zinc ribbon, the ribbon must be installed in select sacrificial anode backfill to ensure that the anode material does not passivate and that it maintains a low resistance to earth. Alternatively, bare copper wire may be used—provided that it is connected to the pipeline through a DC decoupling device to prevent it from draining CP current away from the pipeline. For large ground electrodes, the cost of this device is easily offset by the lower material costs for copper wire versus zinc ribbon in backfill. Furthermore, the copper wire is easier to install than the zinc because it can be installed using a cable plough rather than a ditching machine and, again, need not be installed in special backfill. Two ground electrodes were subsequently constructed using AWG 2/0 copper cable. They were center-connected to the pipe (through a DC decoupler) in order to reduce the effects of attenuation along the wire. It must be emphasized that when using copper wire as a ground electrode, maintaining the DC isolation of the copper wire from the pipe is critical; it requires a commitment to regular monitoring because, should the device fail, it would do so in short-circuit mode that would have serious corrosion consequences if undetected. 3.4.2.3 AC Corrosion Field Test Procedures At this writing, there is no specific test for the identification of AC corrosion— other than calculating the AC current density at the pit site and systematically eliminating all other possible causes. As described in the case histories above, corrosion investigations conducted on pipelines subject to AC interference must be conducted carefully using the procedures suggested below. Note that most of these steps would be conducted in the course of any corrosion investigation, regardless of whether AC corrosion was suspected. 1) Carefully excavate the anomaly, being careful not to disturb the soil immediately adjacent to the anomaly or the corrosion products. 2) Measure DC and AC potentials at several stages of the excavation. 3) Obtain soil samples from adjacent to the anomaly and from the side of the excavation at pipe depth, and determine: CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:92 a) b) c) d) e) f) g) Soil resistivity, both as-found and then saturated with distilled water Moisture content pH Chloride ion concentration Sulfide ion concentration. A quantitative test can be conducted by mixing the corrosion products with iodine and 3% sodium azide solution and checking for the evolution of nitrogen gas. Concentrations of other cations, specifically Ca++, Mg++, Na+, K+, Li+. Soil type, color, and any other special characteristics. 4) Photograph the anomaly after first exposing it. 5) Examine the condition of the coating and determine if the anomaly may have been shielded from receiving CP current. 6) Measure the potential at the anomaly by placing a reference electrode immediately on top of any corrosion products. 7) Using a combination pH/reference micro-electrode and a compatible meter, measure the pH and potential at the bottom of the pit. 8) Remove the corrosion products from the pit and conduct tests to determine: a) b) c) 9) pH Concentrations of ions as discussed in Step 3. Sulfate-reducing bacteria concentration. This can be determined using a kit such as Conoco’s RapidChek II SRB Detection System. Photograph the pit after cleaning it and measure its dimensions. 10) Conduct 24-hour recordings of AC and DC pipe potentials and review the history of these potentials over the life of the pipeline. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:93 After gathering the above data, conduct the following analysis to determine if AC corrosion was the primary cause of the pit. 1) Determine if the pit site was cathodically protected by considering the pH and potential measurements taken in the soil, on the corrosion products, and at the bottom of the pit. If measurements taken adjacent to the pit indicate protection, but those taken within the pit do not, determine if the pit could have been electrically shielded by the coating, the corrosion products, or something in the soil. Consider the CP history of the pipeline and determine if some prior CP deficiency or stray current interference problem could have caused the pit. 2) Determine if the pit could have been caused by bacterial corrosion by considering the pipe potential (relative to -950 mVCSE), the presence of sulfides, the degree of soil aeration (anaerobic), and the count of sulfate-reducing bacteria measured within the pit. 3) If the pit site appears to have received adequate CP over the life of the pipeline, and if bacterial corrosion can be ruled out, investigate the possibility of AC corrosion. Using the soil resistivity and the surface area of the coating holiday (or, if not known, the area of the pit), calculate the AC current density while considering any variations in AC voltage that may have occurred over time. Consider the appearance of the pit site compared to the appearance of the sites discussed in the case histories (i.e., hard hemisphere of soil surrounding the pit site; smooth, round dish-shaped pits having a minimum diameter of 1 cm; hard tubercles covering the pit; etc.). 4) If the CP potential at the pit site is very electronegative, causing the pH to be very high, determine the total ratio of earth alkaline ion to alkaline ion concentrations at the pit site and in the bulk soil to determine what effect this might have on AC current densities at the pit site. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:94 3.4.3 Fault Current Effects The greatest concern regarding the transfer of fault current between a faulted powerline structure and a pipeline is whether there is enough energy available to create an electric arc through the soil. If this should occur, the current path through the soil to the pipeline becomes ionized and results in much higher currents and current densities than would be the case during normal conditions of soil conduction. Such conditions lead to a greater risk of pipeline damage. There have been more than 20 reported instances of AC fault current causing the wall of a steel pipeline to melt and puncture. Partial melting of the pipe wall will create a heat-affected zone subject to embrittlement from the application of CP current. The most effective means of preventing arcs during fault conditions is to maintain a safe separation distance between the powerline structures and the pipeline. Minimum separation distances are usually specified by both the power company and the pipeline company; however, safe separation distances specifically to prevent arcing must either be calculated or determined from research reports. One such calculation is provided by Sunde, who gives the following equations for the distance r (m) over which an arc could occur—based on soil resistivity ρ (Ω-m) and fault magnitude If (kA). r = 0.08 I f ⋅ ρ ( ρ < 100 Ω - m) [3-39a] r = 0.047 I f ⋅ ρ ( ρ > 1000 Ω - m) [3-39b] If safe separation distances are unattainable, then screening electrodes can be used to intercept the fault current. These would typically consist of either lengths of zinc ribbon or banks of packaged sacrificial anodes connected directly to the pipeline, installed between the pipeline and the powerline structure. Screening electrodes may prevent damage to the pipe at the location of fault current pick-up, they lower the resistance between the pipeline and the powerline structure. Hence, they encourage fault current to use the pipeline as a current path. Because this could possibly increase the risk of pipeline damage at locations of fault current discharge, screening electrodes should be used with caution. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:95 Fault currents can also damage pipeline coatings, thereby increasing CP current requirements. Dawalibi noted that 2500 V might be sufficient to cause a slow degradation of the coating—if the coating’s dielectric strength was not high enough to withstand the stress and if the frequency of the faults is relatively high. Southey reported that damage to bitumen coatings can occur at voltages ranging from 1000 V to 2000 V. Damage to polyethylene and FBE coatings occurs at voltages of between 3000 V and 5000 V. Dabkowski provided voltage limits for a variety of coatings (Table 3-4). Table 3-4: Voltage Puncture Levels for Various Holiday-Free Coatings9 Coating Puncture Level (V) Coal Tar Epoxy.............................................. 3500 Coal Tar ......................................................... 4500 Coal Tar Enamel............................................ 5000 Asphalt........................................................... 7000 Fusion Bonded Epoxy ............................. 1000/mil The nature and severity of the damage that occurs to both the coating and the pipe wall are dependent upon the type of coating used. A fault occurring on a coal tarcoated pipe has been found to result in the least damage to the coating, but the greatest damage to the pipe wall, with the size of the damaged areas to both being approximately the same. PE and FBE coatings, however, experience much greater damage to the coating relative to the size of the damaged area on the pipe wall (sometimes an order of magnitude greater). High-current, short-duration faults have been found to cause greater damage to coatings than low-current, high-duration faults having the same energy (i.e., the same product of current and duration). Fault currents can also damage CP test facilities, rectifiers, electrical insulators, DC decoupling devices, and bonds (see Figure 3-66). 9 Pipeline Coating Impedance Effects on Powerline Fault Current Coupling”, PRCI/AGA Report #PR-200634, Dec. 1989, pp3-56. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:96 Figure 3-66: Fault Damage to CP Bond 3.5 Induced AC Voltage Prediction and Mitigation Calculations 3.5.1 Data Gathering The first step in calculating induced AC voltages, is the careful collection of pipeline, powerline, and route data. The pipeline characteristics that must be determined are those that affect the pipeline’s electrical properties. They include pipe wall thickness and diameter as well as coating resistance and thickness. If the pipeline is cathodically protected using sacrificial anodes, this will greatly affect its electrical characteristics. On the other hand, if the pipeline is new and sacrificial anodes are to provide both AC mitigation and CP, then the CP current requirements for the pipeline must be known. Any changes in the electrical characteristics along the pipeline should also be identified—the most important of these being the location of any electrical insulators. The information required for the powerlines is more extensive, and generally more difficult to obtain, than the pipeline data. For each powerline circuit, information on the current loading must be obtained. It is generally assumed that each phase of a three-phase circuit carries approximately the same current; however, the power company might provide information that indicates otherwise. Because loads increase with increased development, current loading may vary over the course of a day (e.g., daytime vs night-time), from day to day (weekday vs weekend), seasonally (winter vs summer), or annually. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:97 In addition to these types of load variations, loads may also change abruptly during emergency conditions. For example, if a powerline circuit is temporarily removed from service, its load must be transferred to one or more other circuits. The maximum current with which a circuit can be loaded depends not only upon the conductor size, but also upon the duration of the emergency condition (the shorter the duration, the higher the current capacity) and the ambient temperature (the colder the temperature, the less the conductors will sag as the loading increases). In the case of determining fault current effects, the maximum line-to-ground fault current must also be obtained for each paralleling high-voltage circuit. Where the pipeline parallels the powerline over a significant distance, the fault current that is available at one end of the collocation may be different from that available at the other end. Hence, fault current data for each circuit may be required at a number of different locations. With regard to the geometric configuration of the powerlines, the horizontal and vertical separation distances between each of the phase conductors and shield wires must be known—as well as the height of the conductors above ground. Because the conductors sag considerably between towers, this height is usually an average height based on a simple formula to be discussed later. The physical construction of the phase conductors is unimportant, but the construction of the shield wires is somewhat important because it affects the magnitude and phase of the currents that are induced in the shield wire. The information typically required for a shield wire is its geometric mean radius, or GMR, and its lineal resistance in Ω/km. Finally, the order in which the phases are arranged on the towers must be known because—as Section 3.3.1 shows—this is critical in determining the phase and magnitude of the LEF. Once the pipeline and powerline information is known, the only other information required is that pertaining to the route shared by the pipeline and powerlines. Route-related information includes the pipeline-to-powerline separation distances, soil resistivities along the route at pipe depth, and the locations of electrical discontinuities. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:98 In addition to pipeline insulators, electrical discontinuities include any location where the LEF as seen by the pipeline changes in either magnitude or phase—such as at the following locations: • A change in the separation distance between the pipeline and the powerline. • The powerline ends, such as at a station, or diverges from the pipeline right-of-way. Similarly, where a pipeline ends or diverges from the powerline right-of-way. • A change in the position of one powerline relative to the position of other powerlines, where multiple powerlines exist. • A change in the configuration of the conductors on a powerline. • A change or transposition in the conductor phase arrangement. • A tie-in to a substation. The electrical power company can provide much of this information, but important details such as phase transpositions quite often may go unreported. It is thus important to visually inspect the entire route of the pipeline-powerline co-location to identify discontinuities and to confirm information provided by the power company on details such as tower type and conductor configuration. Note that obtaining powerline information is often the slow step in the AC mitigation design process. 3.5.2 Field Estimation of LEF In some simple cases, the magnitude of the LEF can be estimated in the field using the horizontal wire method. Using this method, a well-insulated wire is placed along the ground, parallel to the powerlines and in the approximate positions that the pipeline will be constructed. One end of the wire is grounded using a small ground pin or screwdriver. The AC voltage at the other end of the wire is then measured to ground using an AC voltmeter and a second ground pin (Figure 3-67). CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:99 LEF ≅ V L L Figure 3-67: Field Estimation of LEF Magnitude Using Horizontal Wire Method The AC voltage induced in the wire, divided by the wire length, is approximately equal to the magnitude of the LEF. LEF = V AC L [3-40] This measured value can either be used to validate calculated values of LEF magnitude or, in simple problems involving only one powerline and a limited number of discontinuities, it might be used in lieu of obtaining any powerline information. In the latter case, however, exercise caution because current loading and the resulting LEF may change significantly over time. Also, the use of the measured value of LEF may result in the inadequate design during times of peak loading. 3.5.3 Measurement and Interpretation of Soil Resisitivity Data The accurate measurement and interpretation of soil resistivities along the pipeline route is critical in AC mitigation design work, even more so than for CP system design. Soil resistivities affect the magnitude of voltages induced on the pipeline, the effectiveness of the mitigation system, the AC current densities at holidays and the AC corrosion risk, the distance over which arcing may occur between the pipeline and a faulted structure, and the body currents that result from electrical shocks. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:100 In some cases, the resistivities required are those at pipe depth, such as for AC corrosion calculations and induced AC voltage predictions. However, when designing large electrodes to serve as grounds for AC mitigation purposes, examine the variation of soil resistivity versus depth so that a layer-soil model can be established. The model would be suitable for calculating the ground electrode’s resistance. When resistivities are measured to a constant depth at regular intervals along the pipeline route, this is referred to as a resistivity profile. When the variation of resistivity versus depth is examined at a single location, this is referred to as a resistivity sounding. Resistivities may also be measured from small samples of soil either in-situ or after removing the sample from the earth. These methods and their applications are discussed below. In North America, the simplest and most common method with which to conduct a soil resistivity sounding is the Wenner Four-Pin Method. In this technique, four equally spaced pins are placed in the earth in a co-linear array where the pin separation is “a.” A current is circulated between the outside pair of pins while the resulting potential is measured between the inside pair of pins (Figure 3-68). The apparent resistivity for the soil in the general vicinity of the pins is then determined by Equation 3-41. I V Figure 3-68: Soil Resistivity Measurement Using the Wenner Four-Pin Method ρ a = 2πa CP Interference Course Manual © NACE International, 2006 January 2008 V I [3-41] AC Interference 3:101 The test current used for measuring resistivity can be AC, DC, or reversing DC. The instrumentation used to produce and measure the test current, and measure the resulting voltage, may consist of either a current source in combination with an ammeter and a voltmeter or a combination of all these components into a single resistivity instrument. The apparent resistivity measured at a particular pin spacing is not the average resistivity of the soil to a depth equal to the pin spacing, nor is it the resistivity of the soil at a depth equal to the pin spacing. The apparent resistivity is actually a weighted average of the resistivities of all soil layers existing beneath the pin array. The soil having the greatest influence on the apparent resistivity measurement is that which exists at a depth equal to one-third the pin spacing (a/3). The investigation depth for the array (i.e., the depth at which one-half the contribution to the apparent resistivity comes from the soil above that depth, and the other half comes from the soil below that depth) is approximately equal to onehalf the pin spacing (a/2). Soil resistivity can be measured using any combination of two current pins and two potential pins. The geometric factor relating pin spacing, current, and voltage to resistivity can be established using a simple mathematical procedure. However, as the pin array becomes less and less orderly compared to the Wenner array, the interpretation of how soil resistivity varies with depth becomes increasingly difficult Soil resistivity is seldom uniform over large depths. When designing a large ground electrode for AC mitigation purposes, the resistivity of the surface soils alone cannot be used to accurately calculate electrode resistance. When conducting a soil resistivity sounding, apparent resistivities are measured over a wide range of pin spacings (e.g., 1 m to 100 m). This data can then be used to determine the thicknesses and resistivities of the various soil layers so that ground electrode resistance can be accurately calculated. Various empirical methods exist for interpreting apparent resistivity data—most notably, the Barnes layer method. However, the assumptions upon which these methods are based are flawed and can lead to inaccurate soil models. There are numerous software packages available for interpreting apparent resistivity data; these packages are known as inverse-modeling software. Nevertheless, in lieu of such software, curve-matching techniques are the most accurate method of determining a layered soil model. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:102 In a two-layer soil (e.g. soil over bedrock), the apparent resistivity as measured using a Wenner array is a function of pin spacing, upper layer height, and the resistivity of the upper and lower layers. A set of standard curves can be produced, which when fitted to the field data, will indicate the height of the upper layer, and the resistivities of the upper and lower layers. Even in cases where the soil consists of multiple layers, a two-layer model is generally an adequate representation for electrode resistance calculations, and provides better accuracy than if the earth was assumed to be uniform. The curve matching procedure is explained in Appendix A, along with an example of its use. Note that the use of this procedure will be demonstrated in class. Soil resistivity profiling is used extensively in agricultural and archaeological applications, but is seldom used for pipeline corrosion purposes. Some possible applications for conducting a soil resistivity profile along a pipeline route would be to determine anode resistances for a distributed anode cathodic protection system, to locate pockets of low resistivity soil for installing ground electrodes, and to locate pockets of low resistivity soil which may pose an AC corrosion concern. Profiles may be conducted using a four-pin array, either by relocating the entire array at regular intervals along the pipeline route, or by orienting the array parallel to the pipeline route, and leap-frogging the rear pin to the front of the array each time a measurement is made. Automated systems have been developed for rapidly conducting four-pin resistivity profiles, however electromagnetic methods also exist which require no physical connection to the earth and which can obtain continuous resistivity data along the pipeline route, at the same rate as which the surveyor can walk. These systems tend to be insensitive to resistivity variations in high resistivity soils, and are also very susceptible to errors arising from structural interference, and can therefore not be used in close proximity to existing pipelines, metallic fences, buried cables, etc. A resistivity profile will not be able to indicate the actual resistivity at pipe depth, however if the pin spacing is chosen to be three times the pipe depth, the soil at pipe depth will have the greatest influence on the measured resistivity. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3.6 3:103 Prediction of Steady-State Induced AC Voltages 3.6.1 Introduction In order to calculate the steady-state AC voltages which are electromagnetically induced on a pipeline, the following steps must be followed: • Calculate the electrical characteristics of the pipeline. • Divide the pipeline-powerline route into sections, where the pipeline along each section has uniform electrical characteristics, and the LEF along each section has a constant magnitude and phase angle. • Determine the LEF for each section. • Calculate the resulting pipe voltages. 3.6.2 Calculation of Pipeline Electrical Characteristics The following parameters must be calculated. Yi Zi Γ α Z0 Coating Admittance (Ω-1/m) Internal Impedance (Ω/m) Propagation Constant (m-1) Attenuation Constant (m-1) Characteristic Impedance (Ω) The admittance of the coating is essentially the coating’s conductance—in other words, the inverse of its electrical resistance. However, admittance is the general term used here because it is the inverse of the coating’s impedance. The resistance of the pipeline coating could be determined experimentally by taking a section of coating having an area, A, and placing it between two metallic plates to which an ohmmeter is connected (Figure 3-69). The specific coating resistance r′cwould then be determined as the resistance value measured by the ohmmeter and multiplied by the area of the coating. Specific coating resistance is constant regardless of the area of the section of coating being tested because the measured resistance is inversely proportional to coating area. r′c = R ⋅ A CP Interference Course Manual © NACE International, 2006 January 2008 [3-42] AC Interference 3:104 If the resistivity of the coating material is known, then the specific coating resistance can be calculated as r′c = ρ ⋅ L / A [3-43] Figure 3-69: Determination of Pipeline Coating Resistance Using Equation 3-43 in conjunction with values of resistivity for the base materials used in pipeline coatings may result in unrealistically high values for specific coating resistance. This possibility exists because, in practice, a coating contains holidays that lower its resistance. Table 3-5: Specific Leakage Resistances and Conductances in 1000 Ω-cm Soil or Water Long Pipelines with Few Fittings Quality of Work Average Specific Coating Conductance g′c Siemens/ft2 -5 Average Specific Coating Resistance r′c Siemens/m2 Ω-ft2 Ω-m2 5 >104 Excellent <1 x 10 Good 1 x 10-5 to 5 x 10-5 1 x 10-4 to 5 x 10-4 2 x 104 to 105 2 x 103 to 104 Fair 5 x 10-5 to 1 x 10-4 5 x 10-4 to 1 x 10-3 104 to 2 x 104 103 to 2 x 103 Poor Bare Pipe (2 to 12 in) (5 to 30 cm) >1 x 10-4 >1 x 10-3 <104 <103 4 x 10-3 to 2 x 10-2 4 x 10-2 to 2 x 10-1 50 to 250 5 to 25 CP Interference Course Manual © NACE International, 2006 January 2008 <1 x 10 -4 >10 AC Interference 3:105 A section of pipe having a length L and diameter D has a resistance to earth through its coating that is calculated as: rc′ (Ω ⋅ m 2 ) rc′ = Rc (Ω ) = πDL A pipe (m 2 ) [3-44] The coating’s conductance is therefore the inverse of Equation 3-44, or: Gc = πDL 1 = Rc rc′ [3-45] The coating conductance per unit length of pipeline, gc would therefore be: gc = GC πD = L rc′ [3-46] having the units Ω-1/m,or mhos/m, or siemens/m: The internal impedance of the pipe is that which would be measured, if an impedance meter was connected from one end of the pipe to the other, as shown in Figure 3-70. The pipe’s internal impedance is a function of pipe diameter and wall thickness, the electrical resistivity and magnetic permeability of the pipeline steel, and the frequency of the AC source. Ζ(Ω) Figure 3-70: Determination of Pipeline Internal Impedance The internal impedance of the pipe is calculated as: CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference Zi = 3:106 ⎧ [sinh (t n ) + sin (t n )] + j [sinh (t n ) − sin (t n )]⎫ ⎬ (2π)(0.0127 D ) ⎨⎩ cosh (t n ) − cos(t n ) ⎭ 0.5ωμ s ρ s [3-47] where: t n = 0.036 t ω μs ρs [3-48] and where: t = wall thickness (in) μs = magnetic permeability of steel = 3.78 × 10-4 H/m ρs = resistivity of steel = 1.7 × 10-7 Ω-m ω = frequency (radians/s) D = pipe diameter (in) j = complex operator The pipeline propagation and attenuation constants both describe how an AC signal attenuates along the length of a pipeline. The attenuation constant used in AC interference calculations differs from the attenuation constant used in CP calculations because the DC constant does not take into account capacitance across the coating and inductance along the pipeline. The pipeline propagation constant Γ is calculated as follows ⎡ 1.12 ⎤ ⎡ ⎤ 1n ⎢ ⎥ jωμ0 1.85 2 1 ′ ⎥ ⎢ Γ a Γ⎢ + •1n ⎥ = Zi + −1 ⎢ a′ Γ 2 + jωμ ρ −1 + jωε ⎥ 2π )⎦ ⎢Yi π (ρ + jωε)⎥ 0( ⎣ ⎣ ⎦ where: h = pipe depth (m) μ0 = permeability of free space = 1.26 × 10-6 H/m ε = soil permittivity = 2.66 × 10-11 f/m ρ = soil resistivity (Ω-m) j = complex operator CP Interference Course Manual © NACE International, 2006 January 2008 [3-49] AC Interference 3:107 Yi = coating admittance = GC and where: a′ = 0.25 D 2 + 4h 2 [3-50] and j, ω, D, Zi, and were previously defined. Note that in Equation 3-49, Γ exists on both sides of the equation. This type of equation, known as a transcendental equation, can therefore only be solved iteratively using computer methods. The pipeline attenuation constant is then simply the real part of the propagation constant: α = Re[Γ] [3-51] α = ⏐Γ⏐cos(∠Γ) [3-52] that can be calculated as: Finally, the characteristic impedance of the pipeline can also be calculated once the propagation constant is known: Z0 ⎡ 1.12 ⎤ 1n ⎢1 Γ a′ ⎥ = Γ⎢ + ⎥ −1 ⎢Yi π (ρ + jωε)⎥ ⎣ ⎦ [3-53] 3.6.3 Sectionalization of Pipeline-Powerline Route Once the electrical characteristics of the pipeline have been determined, the pipeline-powerline collocation must be divided into sections. The pipeline along each section exhibits uniform electrical characteristics, and the LEF along each section exhibits a constant magnitude and phase angle. A location where there is a change in either the electrical characteristics or the LEF is referred to as a node. Depending upon the method of calculation, it is generally at the nodes where the induced voltages are calculated. Any such change is known as an electrical discontinuity. These discontinuities create induced voltage peaks. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:108 The pipeline-powerline route may be further sectionalized to reflect the changes in soil resistivity (which may introduce small voltage peaks) or to simply increase the number of nodes where induced voltages may be calculated. Conversely, where the changes that occur between one section and another are relatively insignificant (such as where a small change in the separation distance between the pipe and powerline occurs), the number of sections might be reduced to simplify the analysis of the problem. As Section 3.5.1 discusses, electrical discontinuities include—but are not limited to—the following: 1) pipeline insulators, 2) locations where the pipeline and powerline diverge from one another, 3) locations where the powerline circuit configuration changes, and 4) locations where the powerline phase arrangement changes. An example of this procedure appears in Figure 3-71. Node N o 1 Section N o 3 2 1 2 4 3 Powerline Pipeline Figure 3-71: Sectionalization of Pipeline-Powerline Route 3.6.4 Determination of Longitudinal Electric Field (LEF) The most difficult step in the calculation of induced voltages is in the determination of the magnitude and phase angle of the LEF to which each section of the pipeline is exposed. In the simple case where a pipeline parallels a single three-phase circuit—where the pipeline electrical characteristics and the LEF are constant along the entire collocation—the magnitude of the LEF may be measured (this was discussed in Section 3.5.2). The magnitude of the field can then be correlated with the loading at the time of the measurement (if this information is available from the power company) to determine what the maximum field strength might be during times of peak loading. If two or more circuits exist along the pipeline route, it would be difficult to make this correlation. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:109 Where more complex geometries exist in the pipeline-powerline route, the magnitude LEF along each section of the pipeline can still be measured as described above; however, there is no simple way of measuring the phase angle of the LEF along each section. The phase angle of the LEF is not required to calculate pipeline voltages in the simple case where only one section exists; but, it is important when two or more sections exist. For complex geometries, the LEF must be calculated using either the basic equations or specialized software; it may also be determined graphically using published charts. The LEF resulting from a Iφ flowing in a powerline conductor is a function of the mutual impedance ZM between the pipeline and the powerline. E = I φ ⋅ ZM [3-54] In a three-phase system, the mutual impedances between the pipeline and each of the phase conductors (ZMA, ZMB, ZMC), as well as between the pipeline and each of the shield wires (ZMSI, ZMS2), must be calculated. Also, the currents induced in the shield wires (ISI, IS2) must be determined. E3φ = IA ⋅ ZMA + IB ⋅ ZMB + IC ⋅ ZMC + IS1 ⋅ ZMS1 + IS2 ⋅ ZMS2 [3-55] In general, the mutual impedance between two parallel conductors is calculated using Carson’s Equation as follows: ZM = where: and where: j ⋅ f ⋅ μ 0 ⋅ 1n (H − H '+ 2 ρ / j 2 π f μ 0 ) + d 2 2 (H [3-56] + H' ) + d 2 2 f = frequency (Hz) μ0 = permeability of free space = 1.26 × 10-6 H/m ρ = soil resistivity (Ω-m) j = complex operator H, H′, d, and D define the pipeline-powerline geometry, as shown in Figure 3-72 CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:110 Figure 3-72: Pipeline-Powerline Geometry for Calculation of LEF As an alternative to calculating the LEF, a series of standard curves have been published by the Electrical Research Power Institute (EPRI Report No. EL-3106). The report relates the magnitude and phase angle of the LEF to conductor height h, conductor spacing s, pipeline-powerline separation distance d, soil resistivity ρ, phase current I, as well as phase arrangement and conductor configuration. Although there is theoretically an infinite number of possible combinations of these factors to account for, the publication contains a manageable number of curves to cover most practical applications. The number of required curves has been minimized by adopting a parametric approach, where variables such as d and h are normalized with respect to s, thereby effectively reducing the number of variables. Figure 3-73 gives an example of one set of curves to determine LEF magnitude for a particular tower geometry and soil resistivity. Note that the use of these curves will be further demonstrated in class. In cases where the pipeline is paralleled by more than one powerline, the fields generated by each individual powerline are determined. They are then added together (see Equation 3-57), using the rules of vector algebra that Section 3.2.1 describes. E = E1 + E2 + E3 + … + EN CP Interference Course Manual © NACE International, 2006 January 2008 [3-57] AC Interference 3:111 s =1 h 0.056 B C 0.8 0.048 s 0.7 A s TOWER GEOMETRY h 0.040 0.6 d 0.5 0.032 0.4 Balanced Phase Currents ρ > 3 ohm-meters 0.024 0.3 0.016 s = 0.2 h 0.008 0 2 4 6 8 10 12 14 Normalized Distance From Tower Center - d / s Electric Field Horizontal Circuit Configuration Figure 3-73: Typical Series of Curves for Determining LEF (Reprinted from J. Dabkowski, NACE Corrosion /80) 3.6.5 Calculation of Induced Pipeline Voltages In the case of a simple pipeline-powerline geometry (Figure 3-74), the induced voltage at any point along the pipeline can be calculated by using a transmission line model (Figure 3-75). Powerline d L CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:112 Figure 3-74: Simple Pipeline-Powerline Corridor (Plan View) Powerline LEF = E0 x=0 Z0 x=L Pipeline Z1 Z2 Figure 3-75: Simple Pipeline-Powerline Model Vx = {[ ] [ ] E0 Z 2 (Z1 − Z 0 ) − Z1 (Z 2 + Z 0 )e ΓL e −Γx − Z1 (Z 2 − Z 0 ) − Z 2 (Z1 + Z 0 )eΓL e Γ( x−L ) Γ (Z1 + Z 0 )(Z 2 + Z 0 )eΓL − (Z1 − Z 0 )(Z 2 − Z 0 )e −ΓL [ ] } [3-58] where: E0 L x Γ Z0 Z 1, Z 2 e = = = = = = = magnitude of LEF pipeline length (m) distance to point of interest (m) pipeline propagation constant (m-1) pipeline characteristic impedance (Ω) pipeline terminating impedances (Ω) 2.718 In this model, the characteristic impedance determines how the induced AC signal attenuates along the pipeline; the terminating impedances can be used to represent whatever might be physically attached to the ends of the pipeline. The terminating impedances may include: Pipeline insulator................................ Z = ∞ Ground electrode................................ Z = ground electrode resistance Pipe section (electrically long)........... Z = Z0 Pipe section (electrically short).......... Z = ZG CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:113 where ZG is the grounding impedance of a short section of pipe having a length L, as determined by: ZG = Z0 coth (ΓL) [3-59] Hence this simple model enables one not only to predict the induced AC voltages, but also to determine the effects of various AC mitigation systems. With this model, the locations where discrete ground electrodes can be added are limited to the endpoints; however, these are where the peak voltages occur and are thus the optimal locations for ground electrode installations. Furthermore, this model allows for the use of distributed ground electrodes; but, this may not be readily apparent by looking at the model. Recall that the characteristic impedance Z0 for the pipeline is a function of the coating admittance. If uniformly distributed ground electrodes (such as sacrificial anodes) are installed along the pipeline, they can essentially be modeled as coating defects. If the resistance of a single anode is RA, then the parallel combination of N anodes uniformly distributed along the pipeline will be RA/N, assuming that the anodes are installed sufficiently far apart from one another so that interference effects can be ignored. The total admittance of the anodes to earth YAT is therefore: Y AT = N RA [3-60] and the admittance per unit length would be as shown in Equation 3-61, where L is the length of the pipeline and S is the anode spacing. Y AT N 1 = = L L ⋅ RA S ⋅ RA [3-61] To determine the total admittance of several admittances connected in parallel, the admittances need only be added together. Therefore, the effective coating admittance after mitigation, YM, is simply the original coating admittance plus the anode admittance per unit length. Y M = Yi + CP Interference Course Manual © NACE International, 2006 January 2008 1 S ⋅ RA [3-62] AC Interference 3:114 The value calculated for YM is then substituted back into the equations for the propagation constant and characteristic impedance, allowing one to determine the mitigative effect of the distributed anodes. Note: Because the equations for the propagation constant, characteristic impedance, and the resulting pipeline voltages are complex, example calculations will be conducted in class with the help of a spreadsheet program. For more complex pipeline-powerline geometries than that shown in Figure 3-75 (i.e., those with more than one section, such as that shown in Figure 3-71), the procedure for calculating induced pipeline voltages is more complicated. One such method, as described in EPRI Report No. EL-3106, is outlined below. The voltage at any node on a pipeline having N discontinuities is calculated as: N V = P0 ∑ E Si ni e − jαλ i [3-63] i =1 where P0, known as the parameter coefficient of voltage, is calculated as: P0 = 1 2Γ [3-64] The variable Esi is the difference in the LEF to the left and right of the ith discontinuity, which is responsible for generating the voltage at the discontinuities. ESi = ELi – ERi [3-65] However, the voltage at ith discontinuity is not only a result the change in the LEF that occurs at this discontinuity, but also dependent upon the magnitude of ESi that is generated at all other discontinuities along the pipeline. As the discontinuities become increasingly remote from the discontinuity in question, their effect on the magnitude of the voltage generated at the subject discontinuity diminishes; this is accounted for in Equation 3-63 by the scaling factor ni. ni = 10 kαλ i where: CP Interference Course Manual © NACE International, 2006 January 2008 [3-66] AC Interference 3:115 k = a constant ≈ -0.458 α = pipeline propagation constant (m-1) λι = distance from the ith discontinuity to the subject discontinuity (m) Similarly, as the discontinuities become increasingly remote from the subject discontinuity, there is a rotation that occurs in the phase angle of the ESi value of the ith discontinuity when its effect is accounted for at the subject discontinuity. This phase angle rotation is determined by the variable e −kαλ , which according to Euler’s Identity has the value: i e − jαλi = cos(−αλi ) + j sin(−αλi ) where: Note: [3-67] the value αλI represents a phase rotation in radians Because the calculations for this method are complex, particularly for pipeline-powerline systems having numerous nodes, this method of induced voltage calculation is provided only for reference. It will be demonstrated by the instructor in class using spreadsheet software. The procedure described above is now seldom used for solving complex problems now that AC mitigation software packages are available for personal computers. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3.7 3:116 Prediction of Fault Voltages 3.7.1 Introduction The effects of a powerline fault on a pipeline are more difficult to accurately determine than the steady-state induced voltages, particularly because the location and magnitude of a fault are impossible to predict. The following discussion provides only a rudimentary analysis of fault current effects. For a more detailed analysis, it is recommended that commercially available software be used. 3.7.2 Conductive Coupling Due to Fault Currents The electrical power company is generally able to provide the maximum line-toground fault current IF that is available at any point along the power transmission line. This fault current value will typically assume that the line-to-ground fault path has an impedance of 0 Ω, whereas the fault will in fact have a finite impedance (e.g., the footing resistance of a faulted tower ZT) that will further limit the magnitude of the fault current. Furthermore, in the case where the powerline has aerial shield wires or a counterpoise, not all the fault current will enter the earth because a significant portion will return to the source (e.g., generating station or substation) via these conductors. A simple equivalent circuit for a line-toground fault is shown in Figure 3-76. IF ZP Powerline VL-G ZS IFS Shield Wire ZT IFT Tower Figure 3-76: Equivalent Circuit for Line-to-Ground Fault As a rough approximation, the fault current entering the earth at the tower could be calculated by analyzing the following simple network: CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:117 I FT = V L −G ⎛ Z ⋅Z Z p + ⎜⎜ S T ⎝ Z S + ZT ⎛ ZS ⋅ ⎜⎜ ⎞ ⎝ Z S + ZT ⎟⎟ ⎠ ⎞ ⎟⎟ ⎠ [3-68] The impedance of the tower to earth can be calculated knowing the details of the tower footing and the local soil resistivity. The impedance of the power transmission system, Zp, may be provided by the power company; in cases where it is not, a rough approximation would be to assume that the impedance is purely reactive (i.e., phase angle = 90 degrees) and to estimate the magnitude as: ZP = V L −G IF [3-69] When the powerline includes a shield wire and the shield wire is electrically continuous between towers, the shield wire will return fault current to the source and distribute fault current to adjacent structures where it may be discharged to the earth (Figure 3-77). IFS1’’ IFS1’ IFS1 IFS2 IFS2’ IFS2’’ Fault IFT1’ IFT1 IFT IFT2 IFT2’ Figure 3-77: Distribution of Fault Current along Powerline The impedance of the shield wire ZS is difficult to determine, although it might be supplied by the power company. The assumption that all fault current will enter the earth at the faulted tower will result in worst-case pipe-to-earth voltages because it results in the highest fault current densities. As a conservative approximation, it might instead be assumed that 25 percent of the fault current CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:118 returns via the shield wires—assuming that they are electrically continuous back to the source. If the faulted structure is located in close proximity to the pipeline, arcing occurs, an ionized path (i.e., an extremely low-resistance path) is created in the soil between the two structures, and essentially all of the fault current entering the earth is transferred to the pipeline. This represents a worst case because, not only could the high current density at the pipe’s surface damage the pipe, the pipeline rises to nearly the same voltage as the faulted tower. Soil ionization can occur for a fault current If (kA) when the distance r (m) from the tower footing to the pipeline meets one of the following criteria: r = 0.08 I f ρ (for ρ < 100 Ω - m) [3-70] r = 0.047 I f ρ (for ρ > 1000 Ω - m) [3-71] If arcing does not occur, then the fault current flows radially away from the tower footing. This produces a voltage gradient in the soil (Figure 3-78). IFT VT Vr r Figure 3-78: Distribution of Fault Current along Powerline In this case, the voltage of the earth at the pipeline (Vr) can be estimated by approximating the tower footing as a hemisphere having a radius req and calculating the voltage at a distance r away from the tower (Figure 3-79). CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:119 req Vr r RC’ VP ZG Remote Earth Figure 3-79: Calculation of Earth Voltage at Pipe Due to Faulted Tower The equivalent radius of the tower footing is a function of the soil resistivity ρ and the tower-to-earth resistance RT. req = ρ 2πRT [3-72] The voltage of the tower footing is: VT = I FT ⋅ RT = I FT ⋅ ρ 2πreq [3-73] Therefore, the voltage at a distance r from the tower footing is: Vr = req ρ ⋅ I FT = VT 2π r r [3-74] Now, due to the resistance of the pipeline coating, the voltage of the pipeline Vp, will be less than the voltage of the earth immediately outside the coating. It will be determined by the following voltage divider (Figure 3-38). CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:120 V P = Vr ZG Z G + RC′ [3-75] where: Vr = voltage of the earth in the vicinity of the pipe ZG = grounding impedance of pipeline Rc′ = modified coating resistance The voltage across the coating is therefore: VC = Vr − V P = Vr − Vr ⎛ ⎞ ZG ZG ⎟⎟ = Vr ⎜⎜1 − + Z G + RC ' Z R G C' ⎠ ⎝ [3-76] The grounding impedances of a pipeline having a length L for the cases where the fault current is injected into the end of the pipeline, or into the middle of the pipeline, are as follows: ZG = Z0 ⎛ ΓL ⎞ coth⎜ ⎟ 2 ⎝ 2 ⎠ Z G = Z 0 coth (ΓL) (center injection) [3-77] (end injection) [3-78] The resistance presented to the fault current by the coating may be linear, depending on whether localized soil ionization occurs in the immediate vicinity of the coating holidays. The length of pipe affected by the fault current is typically equal to 2r, where r is the distance from the tower to the pipe (Figure 3-80). r = distance from tower to pipeline Affected Length of Pipe ≅ 2r Figure 3-80: Approximate Length of Pipeline Affected by Faulted Tower CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:121 For the case where soil ionization does not occur, the resistance offered by the coating along the affected length of pipe is: RC′ = rC′ 2π ⋅ r ⋅ D [3-79] The resistance of a single holiday (Figure 3-81) having a diameter d would be: RH = ρ 2d Coating Holiday [3-80] Soil Resistivity: ρ d Pipe Wall Figure 3-81: Resistance of Coating Holiday to Earth The number of these holidays that might exist along the affected length of pipe would therefore be estimated based on how many holidays would be required in parallel to account for the coating resistance R′C. N= RH RC′ [3-81] The high current densities that can arise at the holidays may cause large voltage gradients may occur at the holidays. These gradients may be of sufficiently high magnitude to result in localized soil ionization effects, thereby effectively increasing the diameter of the holiday and decreasing the holiday’s resistance to earth (Figure 3-82). CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:122 Ionized soil in vicinity of holiday as a result of AC fault current pick-up Breakdown Voltage of Soil: VBD ≅ 1 MV/m to 2 MV/m Soil Resistivity: ρ Voltage across coating: VC Pipe Wall d Figure 3-82: Modified Resistance of Coating Holiday to Earth Due to Localized Soil Ionization Effects The modified resistance of a coating holiday is given by: R H′ = ρ ⋅ V BD 2π ⋅ Vr [3-82] where: VBD is the breakdown voltage of soil, which typically may range from 1 MV/m to 2MV/m The resistance of the coating when considering soil ionization therefore becomes: RC′′ = R H′ N [3-83] and the pipeline voltage and the voltage across the coating therefore become: V P = Vr ZG Z G + RC′′ ⎛ ZG VC = Vr ⎜⎜1 − ⎝ Z G + RC′′ CP Interference Course Manual © NACE International, 2006 January 2008 [3-84] ⎞ ⎟⎟ ⎠ [3-85] AC Interference 3:123 3.7.3 Inductive Coupling Due to Fault Currents During fault conditions, the voltages induced on the pipeline will increase due to the increased current flow in the powerline—as well as to the current imbalance between the three phases. The induced voltages along the pipeline during fault conditions can be determined using the method in Section 3.6; however, the LEF occurring during fault conditions must be recalculated. This requires that the mutual impedance ZM between the faulted powerline and the pipe be determined. Z M = j ⋅ f ⋅ μ 0 ⋅ 1n (h − h′ + 2 ρ / j 2πfμ 0 ) 2 +d2 (h + h′)2 + d 2 [3-86] where: f μ0 ρ h h′ d = = = = = = frequency (Hz) permeability of free space average soil resistivity (Ω-m) average height of conductor average depth of pipeline mean distance from powerline to pipeline The mutual impedance is then substituted into the following equation, along with the powerline fault current If, to determine the average electric field strength during fault conditions: E 0 = I fZM [3-87] This field strength is then used to determine the peak voltages occurring at the major discontinuities under fault conditions. As an alternative to the method described above, the graph in Figure 3-83 can be used to estimate pipeline voltage rise caused by inductive coupling with a fault current. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:124 600 10 20 5.6 8 10 30 40 50 70 100 200 300 500 10 11.2 12.5 14.8 18 25 31 40 14 17 20 22.5 26.5 32 45 55 71 18 25 31 36 40 90 100 12.5 32 45 55 64 71 56 80 100 112 125 148 180 250 310 400 100 140 170 200 225 265 320 450 550 710 190 250 310 350 400 475 560 800 1000 1250 500 400 300 47.5 56 85 100 143 175 225 200 100 0 5 10 Pipeline-Powerline Separation (m) 15 Length of Parallelism (km) Figure 3-83: AC Pipeline Voltages Induced by Overhead Faulted Powerline (Per 1000 A of Fault Current)10 3.7.4 Other Related Calculations In addition to calculating the effects of AC interference on a pipeline, other calculations are required. They include the following: 3.7.4(a) Ground Electrode Resistance Resistance of a hemispherical electrode having a radius r installed at grade: R= ρ 2πr [3-88] Resistance of a circular plate electrode having a diameter D installed at grade: 10 Electricite de France (EDF) and Gaz De France (GDF). Recommendations for Protection of Steel Pipelines against Electrical Interference, 1967. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:125 R= ρ 2D [3-89] Resistance of a vertical electrode, having a diameter D and length L, installed at grade: R= ρ ⎡ ⎛ 8L ⎞ ⎤ ⎢ln⎜ ⎟ − 1⎥ 2πL ⎣ ⎝ D ⎠ ⎦ [3-90] ρ 4L ln 2πL D [3-91] or alternatively: R= Resistance of a vertical electrode, installed at a depth T: R= ⎛ 2 L 4T + 3L ⎞ ρ ⎟ ln⎜⎜ 2πL ⎝ D 4T + L ⎟⎠ T, L >> D [3-92] Resistance of a groundbed consisting of N electrodes separated by a uniform spacing S, each having a resistance R installed in a collinear array. RN = 1 N ⎡ ρ ⎛1 1 1 1 ⎞⎤ ⎢ R + πS ⎜ 2 + 3 + 4 + ... + N ⎟⎥ ⎝ ⎠⎦ ⎣ S≥L [3-93] or, when N is large, this simplifies to: RN = ρ 1⎛ ⎞ ln(0.66 N ) ⎟ ⎜R + N⎝ πS ⎠ S≥L [3-94] Resistance of a horizontal electrode, installed at a depth T: R= or alternatively: CP Interference Course Manual © NACE International, 2006 January 2008 ⎛ L2 ⎞ ρ ⎟ ln⎜⎜ 2πL ⎝ TD ⎟⎠ T, D << L [3-95] AC Interference 3:126 R= ρ ⎛ 2L ⎞ − 1⎟ ⎜ ln πL ⎝ T ⋅D ⎠ T << L [3-96] Resistance of a vertical electrode, installed in a two-layer soil where the electrode does not penetrate the lower layer: nh ⎤ ⎡ + 1⎥ ∞ ρ1 ⎢ 8 L kn L − + R= ln 1 ln ⎢ ⎥ ∑ nh ⎥ 2πL ⎢ D n =1 2 −1 ⎥⎦ L ⎣⎢ where: [3-97] ρ1 = upper layer resistivity ρ2 = lower layer resistivity h = upper layer height k = reflection factor k= ρ 2 − ρ1 ρ 2 + ρ1 [3-98] Resistance of a horizontal electrode installed in the upper layer of a two-layer soil: 2 ⎡ ⎤ ⎛ 2nh ⎞ ⎢ ⎥ + + 1 1 ⎜ ⎟ 2 ρ1 ∞ n ⎢ 8nh ⎥ ⎝ L ⎠ ⎛ 2nh ⎞ + −4 ⎜ R= k ⎢4 ln ⎟ + 1⎥ ∑ 2nh 2πL n =1 L ⎝ L ⎠ ⎢ ⎥ L ⎢ ⎥ ⎣ ⎦ 3.7.4(b) h,L >> D [3-99] Step and Touch Potential The voltage at a distance x from the outside of a loop of wire located at grade, having a radius r, discharging a current I into the earth: V ( x) = CP Interference Course Manual © NACE International, 2006 January 2008 Iρ ⎡ −1 r ⎤ sin x + r ⎥⎦ 2πr ⎢⎣ [3-100] AC Interference 3:127 Similarly, in a two-layer soil, this becomes: ∞ Iρ1 ⎡ −1 r ⎢sin V ( x) = + 2∑ k n sin −1 x+r 2πr ⎢ n =1 ⎣ 3.7.4(c) 2r (2nh )2 + x 2 + (2nh )2 + (x + 2r )2 ⎤ ⎥ ⎥⎦ [3-101] Conductor Size The minimum required conductor size (in circular mils) to prevent fusing during a fault: A = 197 ⋅ I tc α r ρr ⎛ ⎜ Tm − Ta T CAP 1n ⎜1 + ⎜ 1 − Tr + Ta ⎜ αr ⎝ ⎞ ⎟ ⎟ ⎟ ⎟ ⎠ [3-102] where: tc = fault duration (s) αr = thermal coefficient of resistivity ρr = conductor resistivity (μΩ-cm) Tm = maximum allowable temperature (ºC) Ta = ambient temperature (ºC) Tr = reference temperature (ºC) TCAP = thermal capacity (joules/cc/ºC) I = fault current (A) Note that for copper conductors at 20ºC, this formula simplifies to: A = 6.83I t c 3.8 [3-103] Equipment for AC Mitigation 3.8.1 DC Decoupling Devices An important component of most AC mitigation systems is the DC decoupling device, which permits the flow of AC but blocks the flow of DC. Consider the case of a motor-operated valve on a pipeline, which must be electrically grounded for operational reasons and to satisfy the local electrical codes. As Figure 3-84 CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:128 shows, electrical ground provides a low-resistance path to drain induced AC currents to earth—thereby lowering the induced AC pipeline voltages; however, it also needlessly picks up CP current intended for the pipeline itself. Therefore, CP current requirements increase, CP potentials decrease (both locally at the valve site and possibly elsewhere along the pipeline), and CP attenuation increases. M Induced AC Current Induced AC Current CP Current Figure 3-84: Motor Operated Valve – Effects of Grounding on Induced AC and CP Currents One solution to this problem is to electrically isolate the valve from the pipeline (Figure 3-85). When this is done, a bond must be installed across the valve to maintain electrical continuity along the pipeline for CP purposes and to prevent the generation of an induced AC voltage peak. This approach solves the CP problems, but the pipeline has perhaps lost an important AC mitigation facility and pipeline voltages may consequently increase. The solution also requires that the insulators be above-grade because buried insulators may not be as effective and would be subject to stray current interference. If the valve is below-grade, it would now require its own separate CP system because it is isolated from the pipeline’s CP system. Finally, whenever insulators are installed on a pipeline that is exposed to induced AC interference there exists a risk that the insulators may be damaged as a result of fault currents. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:129 M CP Current Figure 3-85: Electrical Isolation of Motor-operated Valve from Pipeline Rather than isolate the valve from the pipeline, a preferred solution would be to install a DC decoupling device between the valve and electrical ground. This would provide AC continuity but break the DC current path. Consider the electrical grounding schematic of a motor-operated valve (Figure 386). Installing a DC decoupling device in the ground connection between the electrical service entrance and the valve allows the valve to become isolated from the service entrance ground and secondary grounding system (perhaps a copper loop and some ground rods) and the extensive primary grounding system owned by the power company (consisting of pole grounds, substation grids, connections to watermains, etc). CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:130 Service Entrance Distribution Transformer Line Line Fuse Neutral Arrestor DC Decoupler Motor Operated Valve Line Primary Neutral Secondary DC Decoupler Primary Ground Secondary Ground Service Entrance Ground Figure 3-86: Electrical Grounding Schematic of Motor Operated Valve Showing Two Alternative Locations for a DC Decoupling Device As an alternative, the DC decoupling device may also be installed between the primary and secondary grounding systems (Figure 3-87). This has the advantage of being able to isolate several grounded pipeline components in a station with a single DC decoupling device, although the local secondary grounding system would still be connected to the piping and would pick up some CP current. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:131 Figure 3-87: Decoupling Device Installed by Electrical Utility Between Primary and Secondary Grounds The most reliable DC decoupling devices, which are also the most commonly used today, are solid-state devices such as the one shown in Figure 3-88. These devices have a very high DC impedance, a very low AC impedance, and can pass steadystate AC currents as well as lightning and fault currents. These devices are selfpowered. Although the internal construction of these devices may vary depending upon the particular manufacturer and the device requirements, the device may include the components shown in Figure 3-89. Steady-state AC current passes through the electrolytic capacitor. AC fault currents pass through the thyristors. The surge protector passes lightning currents. The inductor prevents the lightning currents from damaging the capacitor and thyristors. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:132 Figure 3-88: Isolation-Surge Protector Installed across Isolating Flange Electrolytic Capacitor – + Gate Thyristor Thyristor Inductor Gate Surge Protector – + Figure 3-89: Electrical Schematic of One Model of Solid-State DC Decoupling Device CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:133 DC decouplers may also be installed directly across insulating flanges to provide protection against lightning and fault current damage. Note, however, that the connections must be kept as short as possible (Figure 3-90) to provide protection against lightning. Otherwise, the voltage drop created across the inductance reactance of the lead wires alone may be enough to break down the insulator. Figure 3-90: DC Decoupling Device Installed across Insulating Flange for Lightning Protection Solid-state DC decouplers are reported to have a very low failure rate; should one fail, however, it will fail in the short-circuit mode. This is considered to be the fail-safe mode from an AC safety viewpoint, but it will be detrimental to the CP system. Prior to the advent of solid-state DC decouplers, the polarization cell (Figure 3-91) was used to pass steady-state and fault AC currents—as well as lightning currents—while maintaining DC isolation. The polarization cell consists of a series of nickel or stainless steel plates immersed in an alkaline hydroxide solution, such as the one that appears in Figure 3-92. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:134 Figure 3-91: AC Current Being Measured Through a Polarization Cell O Ho Ho O O Ho O Ho O Ho O KOH Ho Figure 3-92: Polarization Cell Construction Depending upon its construction (the size, number, and spacing of the plates), the polarization cell is capable of carrying fault currents of tens of thousands of amperes. It has a very low AC impedance, which is typically in the 0.1-mΩ range. Initially, the cell also has a low DC resistance; but, as CP current passes through the cell, anodic and cathodic polarization of the plates occurs and a DC backvoltage develops. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:135 Although the polarization cell and the solid-state DC decoupler perform similar functions, the polarization cell has several disadvantages. Under steady-state AC load the plates tend to depolarize, allowing a significant amount of DC to pass through the cell. This reduces the effectiveness of the DC isolation that the cell is attempting to maintain, thereby reducing the effectiveness of the CP system. Moreover, it results in corrosion of the anodic plates inside the cell. As the plates corrode, the cell’s AC impedance increases. In severe cases, the cell may fail entirely. Figure 3-93: Corrosion of Plates within a Polarization Cell Simpler and less costly alternatives exist to the solid-state DC decouplers and polarization cells discussed above; however. these also tend to be less effective. The zinc grounding cell is similar in appearance to a packaged sacrificial zinc anode, except that two zinc electrodes are installed side-by-side inside the anode package and are separated by insulating blocks (Figure 3-94). The lead wires from the zinc electrodes are installed on opposite sides of a pipeline insulator. The lowresistivity anode backfill, when saturated, provides a low-resistance path between the two zinc electrodes—on the order of 0.03 Ω, which provides a reasonably lowimpedance path for AC. As CP current flows from the unprotected side of the flange, through the grounding cell, and back to the protected piping, the zinc electrode polarizes cathodically and thereby limits the amount of DC that can flow through the cell. However, in order to develop a significant back voltage across the insulator of 0.5 V, a substantial amount of DC is required—on the order of 500 mA—that compromises the effectiveness of the insulator. Furthermore, a steadyCP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:136 state AC load allows a significant amount of DC to pass through the cell, leading to the premature consumption of the anodic electrode and the eventual open-circuit failure of the cell; this results in an AC safety hazard. Cathodically protected side of insulator This electrode polarizes cathodically and resists the flow of direct current Insulating flange Grounding cell consisting of two - 5 ft. long zinc anodes separated by insulating blocks and surrounded with low resistivity backfill Figure 3-94: Grounding Cell Another alternative for passing steady-state AC is the use of an electrolytic capacitor (a component of the solid-state DC decoupler). The capacitor can be connected between the pipeline and a ground electrode, such as a pipeline casing (Figure 3-95). Figure 3-95: Electrolytic Capacitor CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:137 Because the electrolytic capacitors are polarity-sensitive, the negative terminal must be connected to the more electro-negative structure (i.e., the pipeline). In cases where this polarity may reverse (such as in DC stray current areas) and when the capacitor is carrying AC, capacitors have been known to explode and/or catch fire (Figure 3-96). Figure 3-96: Failure of Electrolytic Capacitors in Stray Current Area Capacitors generally fail in short-circuit mode, which is the fail-safe condition from an AC safety viewpoint but which can be detrimental to the CP system. Capacitors are also susceptible to damage from electrical transients and benefit when they are paralleled with surge protectors, such as metal-oxide varistors (Figure 3-97). CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:138 Figure 3-97: Metal-Oxide Varistors (MOVs) Note that in cases where the pipeline casing is used as an AC ground electrode, the vent pipes (if applicable) will rise to the same AC voltage as the pipeline. Hence precautions must be taken to ensure that the public is protected from exposure to these voltages (e.g., cutting the vent pipes off below-grade). Solid-state devices are also available for installation across insulators to protect the insulator from transients. Such devices may be explosion-proof (Figure 3-98) and will conduct both AC and DC when a predetermined voltage limit is exceeded (e.g., +1V/-2V, +4V/-4V). Note, however, that such devices may not be appropriate for areas where steady-state induced AC voltages are present because any voltage higher than the voltage limit will cause the device to conduct— thereby compromising the effectiveness of the insulator. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:139 Figure 3-98: Explosion-Proof Surge Protection Device Installed across Insulator 3.8.2 Test Stations When a pipeline is exposed to induced AC voltages, the test leads at CP test stations can potentially expose pipeline personnel—as well as the public—to hazardous voltages. Test station selection is therefore an important consideration when designing an AC mitigation system. Figure 3-99 shows a number of different test station types. A test station in which the test lead terminals are exposed is obviously the poorest choice where induced AC voltages are present. Such a test station would also be a poor choice for CP purposes because the test leads could contact foreign metallic structures such as fences. The vast majority of commercially available test stations include a cover, thereby limiting the chance of contacting the test lead terminals. In some cases these covers can be easily removed, whereas a better choice in an area subject to AC interference would be a cover that incorporates a locking device. Such locking devices are not tamper-proof, however, and should not be relied upon to prevent public access to hazardous voltages. The safest test station choice, from a public safety point of view, is one in which the cover can be padlocked. The terminals inside such a test station should be of dead-front design to prevent accidental contact with AC voltages on the test leads by authorized personnel. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:140 Figure 3-99: Test Station Varieties (left to right): a) Terminals Exposed to Public; b) Terminals Covered by a Plastic Cap (Locking or Non-Locking); c) Dead-Front Terminals; d) Aluminum Test Station with Padlocked Cover 3.8.3 Sacrificial Anodes An important consideration when selecting anodes for CP and/or AC mitigation on a pipeline subject to steady-state induced AC interference is the effect that the AC current may have on the anode consumption rate. Consumption rates of both zinc and magnesium anodes increase with increasing current density. Sacrificial anodes are generally installed in wettable packages containing special backfill. In some cases, the anode may be installed as a continuous ribbon (Figure 3-100). This ribbon is often installed directly in the pipe trench without special backfill, but in some cases (such as where bicarbonates are present in the soil) the anode’s surface may passivate—causing its resistance to increase and rendering it ineffective as both a sacrificial anode and a ground electrode. Figure 3-101 illustrates the effect of a bicarbonate-rich soil on the potential of zinc and the subsequent reactivation of the zinc’s surface with the addition of calcium sulfate (CaSO4). CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:141 Figure 3-100: a) Zinc Ribbon Anodes of Various Sizes; b) Zinc Ribbon Being Installed in Pipe Trench -1.10 -1.00 -0.90 -0.80 Original Environment -0.70 600 ppm HCO2 73 ppm NO3 20 ppm CO3-2 -0.50 Room Temperature Saturated CaSO4 Added As Gypsum -0.50 -0.40 0 20 40 60 80 TIME - DAYS Figure 3-101: Effect of Gypsum on Restoration of Zinc Potential in Bicarbonate-Rich Soil CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:142 In the case of magnesium anodes, AC causes the potential of magnesium to shift in the electropositive direction (Figure 3-102) and could even become electropositive with respect to the pipeline. This can be prevented by ensuring that the AC current density at the anode’s surface is maintained below 10 A/m2 (1 A/ft2). -1400 1 day 5 days 9 days -1200 -1000 -800 -600 -400 -200 0 200 0 0 100 155 200 310 300 465 400 620 500 mA/in2 775 A/m2 AC Current Density Figure 3-102: Potential of Magnesium Versus AC Current Density in a Fe-Mg Cell CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:143 Group Activity – AC Mitigation System Design Introduction: The students shall break into groups, with each group preparing a design for a pipeline AC mitigation system. The students shall be given 60 minutes to prepare their designs, and 30 minutes will be available to present their designs to the class. Required Materials: 2-layer resistivity master curves Transparent log-log graph paper AC mitigation spreadsheet Laptop computer Course text Problem: A pipeline parallels a pipeline for a distance of L = 20 km at a constant separation distance of 15 m, as shown in Figure 1. The powerline has the geometry shown in Figure 2, where the spacing s between conductors is 5 m and the average height h of the conductors is 15 m. The powerline carries a maximum steady-state current of 1000A per phase and has a line-to-ground fault level of 20,000A. The pipeline has a diameter of 500 mm and is buried at a depth of 1.5 m. The pipe is coated with extruded polyethylene, which is considered to be an excellent coating, requiring a CP current density of 0.1 mA/m2. Assume that only 20-pound high-potential magnesium anodes are available for this project and that these are 1.5 m in length and 125 mm in diameter. A 50-m-long insulated wire was laid out along the pipeline route and was grounded at one end. The AC voltage to ground measured at the opposite end was found to be 0.6V. Soil resistivities along the pipeline route are typically 5000 Ω-cm and are generally uniform with depth—except at Location “C” (Figure 1), where a Wenner 4-Pin survey produced the following data: CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:144 Soil Resistivity Data Pin Spacing Resistivity (Ω-cm) 2m 88,000 4m 55,000 6m 29,000 8m 16,000 10 m 10,000 Pipeline A B C L d Powerline Figure 1 - Pipeline/Powerline Route s s h CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:145 Figure 2 - Powerlne Geometry Tasks: 1) Determine the steady-state induced voltages on the pipeline at locations A, B, and C. 2) Design an AC mitigation system to satisfy the objectives as discussed in this course. 3) Determine the results of this mitigation system on induced voltages. 4) Present the findings to the class. Note: In order to produce a variety of solutions, one group should be assigned the case where both insulators are intentionally shorted and one group should be assigned the case where only the insulator at “A” is shorted. CP Interference Course Manual © NACE International, 2006 January 2008 AC Interference 3:146 Summary of Equations C= [3-1] Q coulombs/volt V A d page 3-3 1 2πfC page 3-4 C=ε [3-2] XC = [3-3] Amean = [3-4] A1 × A2 V pipe page 3-6 AC1 = 0.2m 2 × 5m 2 = 1m 2 AC 2 = 5m 2 × 20m 2 = 10m 2 V pipe = [3-5] page 3-3 C1 V powerline C1 +C s page 3-7 0.9 × 10 −12 = 100 × 10 3 V = 1000V −12 −12 + 90 × 10 0.9 × 10 [3-6a] [3-6b] [3-7] CP Interference Course Manual © NACE International, 2006 January 2008 VS N = S Vp Np Vp Np = VS NS v(t ) = Vm cos (ωt + φ) page 3-19 page 3-19 page 3-24 AC Interference 3:147 [3-8] ω = 2πf [3-9] V = V∠φ page 3-25 [3-10] Α∠φ × Β∠θ = Α ⋅ Β∠( φ + θ) page 3-28 [3-11] Α∠φ ÷ Β∠θ = Α ÷ Β∠( φ − θ) page 3-28 [3-12] A = x + jy page 3-29 [3-13] x = ⏐A⏐cosφ page 3-29 [3-14] y = ⏐A⏐sinφ page 3-29 [3-15] j = −1 page 3-29 [3-16] j = 1/90º page 3-29 [3-17] A/φ × j = A/φ × 1/90º = A/φ + 90º page 3-29 [3-18] A/φ × (-j) = A/φ × -1/90º = -A/φ + 90º = A/φ – 90º page 3-29 [3-19] A/φ ÷ j = A/φ ÷ 1/90º = A/φ – 90º page 3-30 XC = [3-20] [3-21] [3-24] 1 j 2πfC page 3-30 V V = j 2πfCV = 2πfCV∠90° = XC ⎛ 1 ⎞ ⎟⎟ ⎜⎜ ⎝ j 2πfC ⎠ IC = X L = j 2πfL [3-22] [3-23] page 3-24 IL = Vφ −φ = CP Interference Course Manual © NACE International, 2006 January 2008 page 3-32 V V V = = ∠ − 90° XL j 2πfL 2πfL 1.5 2 + (−.866) ⋅ Vφ −G 2 page 3-30 = 3 Vφ −G page 3-32 page 3-34 AC Interference 3:148 VO, L = ± [3-25] [3-27] IB = [3-28] IB = 0.157 ts 0.116 ts R = page 3-49 2 VO, L = ± [3-26] [3-29] E ⋅L E page 3-54 Γ (70 kg body) page 3-62 (50 kg body) page 3-62 ρ 2D page 3-65 [3-30] R2Fp = 1.5ρ page 3-65 [3-31] R2Fss = 6ρ page 3-65 [3-32] V = R×I page 3-65 [3-33] V = ( Rbody + Rfeet) × Ibody page 3-65 0.116 [3-34] Vstep50 = (1000 + 6ρ) [3-35] Vtouch50 = (1000 + 1.5ρ) [3-36] Vstep70 = (1000 + 6ρ) [3-37] Vtouch70 = (1000 + 1.5ρ) [3-38a] CP Interference Course Manual © NACE International, 2006 January 2008 tS 0.116 tS 0.157 tS Vss = Iss × Rbody 0.157 tS page 3-65 page 3-65 page 3-66 page 3-66 page 3-67 AC Interference 3:149 iAC = [3-38b] 8VAC ρπd page 3-74 [3-39a] r = 0.08 I f ⋅ ρ ( ρ < 100 Ω - m) page 3-93 [3-39b] r = 0.047 I f ⋅ ρ ( ρ > 1000 Ω - m) page 3-93 V AC L page 3-98 V I page 3-99 [3-42] r′c= R ⋅ A page 3-102 [3-43] r′cc = ρ ⋅ L / A page 3-103 [3-40] LEF = [3-41] ρ a = 2πa Rc ( Ω ) = [3-44] rc′ (Ω ⋅ m 2 ) rc′ = 2 πDL A pipe (m ) page 3-104 [3-45] Gc = πDL 1 = Rc rc′ page 3-104 [3-46] gc = Gc πD = L rc′ page 3-104 [3-47] Zi = ⎧ [sinh (t n ) + sin (t n )] + j [sinh (t n ) − sin (t n )]⎫ ⎬ (2π)(0.0127 D ) ⎨⎩ cosh (t n ) − cos(t n ) ⎭ 0.5ωμ s ρ s [3-48] [3-49] CP Interference Course Manual © NACE International, 2006 January 2008 tn = 0.036t ωμ s ρs page 3-105 page 3-105 page 3-105 AC Interference 3:150 ⎡ 1.12 ⎤ ⎡ ⎤ 1n ⎢ ⎥ j 1.85 ωμ 2 1 ′ ⎥ ⎢ 0 Γa Γ⎢ + •1n ⎥ = Zi + −1 2 − 1 ⎢ a′ Γ + jωμ ρ + jωε ⎥ 2π )⎦ ⎢Yi π (ρ + jωε)⎥ 0( ⎣ ⎣ ⎦ a′ = [3-50] [3-52] Z0 [3-53] α = ⏐Γ⏐cos(∠Γ) ⎡ 1.12 ⎤ 1n ⎢1 Γ a′ ⎥ = Γ⎢ + ⎥ −1 ⎢Yi π (ρ + jωε)⎥ ⎣ ⎦ page 3-106 page 3-106 page 3-108 E3φ = IA ⋅ ZMA + IB ⋅ ZMB + IC ⋅ ZMC + IS1 ⋅ ZMS1 + IS2 ⋅ ZMS2 Z M = j ⋅ f ⋅ μ 0 ⋅ 1n [3-56] (h − h′ + 2 ρ / j 2πfμ 0 ) +d 2 page 3-108 2 page 3-108 (h + h′)2 + d 2 E = E1 + E2 + E3 + … + EN [3-54] Vx = page 3-106 E = I φ ⋅ ZM [3-54] [3-58] page 3-106 α = Re[Γ] [3-51] [3-55] 0.25 D 2 + 4h 2 {[ ] [ page 3-109 ] page 3-111 E0 Z 2 (Z1 − Z 0 ) − Z1 (Z 2 + Z 0 )e ΓL e −Γx − Z1 (Z 2 − Z 0 ) − Z 2 (Z1 + Z 0 )eΓL e Γ( x−L ) Γ (Z1 + Z 0 )(Z 2 + Z 0 )eΓL − (Z1 − Z 0 )(Z 2 − Z 0 )e −ΓL [ [3-59] [3-60] CP Interference Course Manual © NACE International, 2006 January 2008 ZG = Z0 coth (ΓL) Y AT = N RA ] } page 3-112 page 3-112 AC Interference 3:151 Y AT N 1 = = L L ⋅ RA S ⋅ RA [3-61] Y M = Yi + [3-62] page 3-113 1 S ⋅ RA page 3-113 N V = P0 ∑ E Si ni e − jαλ i [3-63] page 3-113 i =1 P0 = [3-64] 1 2Γ page 3-113 [3-65] ESi = ELi – ERi page 3-113 [3-66] ni = 10 kαλ i page 3-114 [3-67] e − jαλi = cos(−αλi ) + j sin( −αλi ) page 3-114 [3-68] I FT = V L −G ⎛ Z ⋅Z Z p + ⎜⎜ S T ⎝ Z S + ZT ZP = [3-69] [3-70] r = 0.08 I f ρ [3-71] r = 0.047 I f ρ ⎛ ZS ⋅ ⎜⎜ ⎞ ⎝ Z S + ZT ⎟⎟ ⎠ V L −G IF page 3-116 page 3-116 (for ρ < 100 Ω - m) page 3-117 (for ρ > 1000 Ω - m) page 3-117 ρ 2πRT [3-72] req = [3-73] VT = I FT ⋅ RT = I FT ⋅ CP Interference Course Manual © NACE International, 2006 January 2008 ⎞ ⎟⎟ ⎠ page 3-118 ρ 2πreq page 3-118 AC Interference 3:152 Vr = [3-74] V P = Vr [3-75] [3-76] [3-77] [3-78] req ρ ⋅ I FT = VT 2πr r VC = Vr − V P = Vr − Vr ZG = ZG Z G + RC′ [3-80] [3-81] [3-82] [3-83] [3-84] [3-85] CP Interference Course Manual © NACE International, 2006 January 2008 page 3-119 ⎛ ⎞ ZG ZG ⎟⎟ = Vr ⎜⎜1 − + Z G + RC ' Z R G C' ⎠ ⎝ Z0 ⎛ ΓL ⎞ coth⎜ ⎟ 2 ⎝ 2 ⎠ page 3-119 (end injection) page 3-119 rC′ 2π ⋅ r ⋅ D ρ RH = 2d RC′ = page 3-120 page 3-120 ρ ⋅ V BD 2π ⋅ Vr page 3-121 R H′ N page 3-121 RC′′ = V P = Vr page 3-120 RH RC′ N= R H′ = page 3-119 (centre injection) Z G = Z 0 coth (ΓL) [3-79] page 3-118 ZG Z G + RC′′ ⎛ ZG VC = Vr ⎜⎜1 − ⎝ Z G + RC′′ page 3-121 ⎞ ⎟⎟ ⎠ page 3-121 AC Interference [3-86] 3:153 Z M = j ⋅ f ⋅ μ 0 ⋅ 1n (h − h′ + 2 ρ / j 2πfμ 0 ) 2 +d2 (h + h′)2 + d 2 E 0 = I f ZM [3-87] page 3-122 page 3-122 [3-88] R= ρ 2πr page 3-123 [3-89] R= ρ 2D page 3-124 ρ ⎡ ⎛ 8L ⎞ ⎤ ⎢ln⎜ ⎟ −1⎥ 2πL ⎣ ⎝ D ⎠ ⎦ page 3-124 ρ 4L ln 2πL D page 3-124 R= [3-90] R= [3-91] [3-92] [3-93] [3-94] R= RN = ⎛ 2 L 4T + 3L ⎞ ρ ⎟ ln⎜⎜ 2πL ⎝ D 4T + L ⎟⎠ 1 ⎞⎤ ρ ⎛1 1 1 1 ⎡ R+ ⎜ + + + ... + ⎟⎥ ⎢ N⎣ πS ⎝ 2 3 4 N ⎠⎦ RN = [3-97] ρ 1⎛ ⎞ ln(0.66 N ) ⎟ ⎜R + N⎝ πS ⎠ ⎛ L2 ⎞ ρ ⎟ ln⎜ R= 2πL ⎜⎝ TD ⎟⎠ [3-95] [3-96] T, L >> D R= S≥L T, D << L 2L ρ ⎛ ⎞ − 1⎟ ⎜ ln πL ⎝ T ⋅D ⎠ T << L nh ⎤ ⎡ + 1⎥ n ∞ ⎢ ρ1 k 8L L − 1 + ∑ ln R= ⎢ln ⎥ nh ⎥ 2πL ⎢ D n =1 2 −1 ⎢⎣ ⎥⎦ L CP Interference Course Manual © NACE International, 2006 January 2008 S≥L page 3-124 page 3-124 page 3-124 page 3-124 page 3-125 page 3-125 AC Interference 3:154 k= [3-98] ρ 2 − ρ1 ρ 2 + ρ1 page 3-125 page 3-125 [3-99] ⎡ ⎤ ⎛ 2nh ⎞ ⎢ ⎥ 1 1 + + ⎜ ⎟ 2 ρ1 ∞ n ⎢ 8nh ⎥ ⎝ L ⎠ ⎛ 2nh ⎞ R= k ⎢4 ln + −4 ⎜ ⎟ + 1⎥ ∑ 2nh L 2πL n =1 ⎝ L ⎠ ⎢ ⎥ L ⎢ ⎥ ⎣ ⎦ 2 V ( x) = [3-100] Iρ ⎡ −1 r ⎤ sin 2πr ⎢⎣ x + r ⎥⎦ page 3-125 page 3-126 [3-101] V ( x) = [3-102] h,L >> D ∞ Iρ1 ⎡ −1 r ⎢sin + 2∑ k n sin −1 2πr ⎢ x+r n =1 ⎣ A = 197 ⋅ I [3-103] CP Interference Course Manual © NACE International, 2006 January 2008 2r (2nh )2 + x 2 + (2nh )2 + (x + 2r )2 tc α r ρr ⎛ ⎜ Tm − Ta T CAP 1n ⎜1 + ⎜ 1 − Tr + Ta ⎜ αr ⎝ A = 6.83I t c ⎞ ⎟ ⎟ ⎟ ⎟ ⎠ ⎤ ⎥ ⎥⎦ page 3-126 page 3-126 CHAPTER 4 TELLURIC CURRENT INTERFERENCE 4.1 Background Theory Telluric currents are currents that are geomagnetically induced in the earth and in metallic structures on the earth—such as powerlines and pipelines—as a result of the interaction of solar particles on the earth’s magnetic field (Figure 4-1). Here the earth’s magnetic filed is compressed on the sun side of the earth and stretched on the dusk side. The solar plasma arises from two solar phenomena: sun spot activity and corona mass ejections (CME), which are commonly referred to as solar flares. The geomagnetic storms that result from the interaction of the solar plasma with the earth’s magnetic field cause currents to be induced in the earth and metallic structures on the earth. Figure 4-1: Interaction of Solar Particles on the Earth’s Magnetic Field Source: Place, Trevor and Sneath, T. Owen, Practical Telluric Compensation for Pipeline Close-Interval Surveys, NACE Corrosion 2000, Paper No. 741, Orlando, Florida, March 2001 (Powerpoint Presentation) (MP, Vol. 40(9), 2001 p.22 Charged solar particles, composed mostly of electrons and protons that enter the earth’s atmosphere, are deflected by the earth’s magnetic field. This creates current rings in the ionosphere centered around the north and south poles as well as at the equator (figures 4-2a and 4-2b). Electrons are deflected in one direction and protons are deflected in the opposite direction around the earth. This creates a current as the earth’s magnetic field narrows on the dark side of the earth. CP Interference Course Manual © NACE International, 2006 January 2008 Telluric Current Interference Figure 4-2a: Plasma Charge Distribution around the Earth during Quiescent Period Source: Lerner, Eric J., Storms and Hurricanes Don’t Leave Off Where the Atmosphere Ends. Space, Discover, August 1995, p.60 Figure 4-2b: Plasma Charge Distribution around the Earth during a Magnetic Storm Source: Lerner, Eric J., Storms and Hurricanes Don’t Leave Off Where the Atmosphere Ends. Space, Discover, August 1995, p.60 CP Interference © NACE International, 2006 January 2008 4:2 Telluric Current Interference 4:3 The current ring in the auroral regions forms an oval as shown in Figure 4-3. This “electrojet,” as it is sometimes called, typically contains more electrical charges than are generated by man on earth. Figure 4-3: This plot shows the extent and position of the auroral oval in the northern hemisphere, extrapolated from measurements taken during the most recent polar pass of the NOAA POES satellite for September 16, 2004 at 14:22 UT. Source: http://www.sel.noaa.gov/pmap/pmapN.html - 9/16/2004 Because of the amplitude variation and directional changes in this electrojet current, a changing magnetic field is produced that induces an electric field in the earth and in any metallic conductor on or in the earth’s surface (Figure 4-4). Varying Magnetic Field Figure 4-4: Schematic of Geomagnetic Induction Directly into a Pipeline and the Resulting Change in Pipeline Potential that is Produced Source: Boteler, D.H., Gummow, R.A. and Rix, B.C., Evaluation of Telluric Current Effects on the Maritimes and Northeast Pipeline, NACE Northern Area Eastern Conference, Ottawa, October 1999, Paper No. 8A.3, p. 8 CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:4 Measured Pipe-to-Soil Potential Calculated Electric Field Measured Magnetic Field The effect of the changing magnetic field caused by the electrojet, which creates both a changing electric field and a corresponding change in the pipe-to-soil potential on a pipeline, is shown in Figure 4-5. Figure 4-5: Quiet Day Variation in the Geomagnetic Field and the Associated Change in the Electric Field and the Pipe-to-Soil Potential Source: Trichtchenko, L. et al, The Production of Telluric Current Effects in Norway, NACE Corrosion 2001, Houston, TX, March 2001, Paper No. 314 CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:5 The longer variations in these three parameters are a result of the night/day effect of the earth’s rotation. The short variations (i.e., minutes to hours) are a result of the variation of solar particles interacting with the earth’s magnetic field. It should be noted that this data was obtained during a quiet geomagnetic period on a pipeline in Norway located at approximately 60 degrees geogmagnetic latitude. For a long pipeline subjected to an induced electric field, the induced voltage and current profile is typically as shown in Figure 4-6. 10 15 E = 1 V/km 5 10 0 5 -5 -10 0 20 40 60 80 0 100 Distance (km) Figure 4-6: P/S Potential and Telluric Current in a Long Pipeline Exposed to an Induced Electric Field of 1 V/km, having an Impedance of 0.1 Ω /km and an Admittance of 0.15 Ω /km Source: Boteler, D.H. and Seager, W.H., Telluric Currents: A Meeting of Theory and Observation, NACE Canadian Region, Western Conference, Edmonton, Alberta, Feb. 1997. This figure shows that for a long coated pipeline ungrounded at the ends, and subjected to an electric field of 1 V/km, the induced voltage reaches a peak at the end points and decreases with distance from either end toward the center and the voltage reaches zero in the middle of the pipeline. Note that the voltage to earth (pipe-to-soil potential) at each end is out of phase (i.e., when one end is positive, the opposite end is negative). Conversely, the current induced into the pipe is near zero at each end of the ungrounded pipe but reaches a maximum in the middle. This produces the counter-intuitive result that where the voltage peaks are greatest the current in the pipe is the least. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:6 4.1.1 Distributed Source Transmission Line Equations The induced voltage profile can be calculated using distributed source transmission line (DSTL) equations, as has been shown by Boteler and Cookson[1] using an electrical model of the pipeline shown in Figure 4-7. E R L Vind G C Figure 4-7: Equivalent Circuit for a Short Section of Pipeline Each short section of pipeline is represented by a series impedance Z where Z = R + jwL, and a parallel admittance Y where Y = G + jwC and an induced electric field E represented by a voltage source. The response of the pipeline depends on the propagation constant γ and characteristic impedance Z0 given by: γ = Z0 = ZY [4-1a] Z Y [4-1b] The voltage and current along the pipeline are then given by: dV dx dI dx 1 = E − 1Z [4-2] = − VY [4-3] Boteler, D.H. and Cookson, M.J., Telluric Currents and Their Effects on Pipelines in the Cook Strait Region of New Zealand, Materials Performance, Vol. 25(3), March. 1986, p.27-32. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:7 Differentiation and substitution leads to the equations: d2V − γ 2V = dx 2 d 2I dx 2 dE dx [4-4] − γ 2 I = − YE [4-5] If the disturbance is uniform along the pipe, then the electric field does not vary with distance (i.e., dE = 0 ) and these equations then have solutions of the form: dx I = E γE 0 V = ( 1 + Ae E γ ( Ae - γ ( x − x1 ) -γ ( x − x1 ) + Be - γ ( x2 − x1 ) − Be -γ ( x2 − x1 ) ) ) [4-6] [4-7] where A and B are constants determined by the conditions at the ends of the pipeline. For a long pipeline, of length L, terminated at ends 1 and 2 by impedances to ground Z1 and Z2 respectively, this becomes I = V V E − 1 e - γx − 2 e - γ ( L − x ) Z0 Z0 Z V(x) = − V1e - γx + V2 e - γ ( L − x ) [4-8] [4-9] where: V1 = Z1 E × γ Z 0 + Z1 and V2 = Z2 E × γ Z0 + Z2 [4-10] Reviewing these calculations and Equation 4-10, it is apparent that the magnitude of the induced voltage that appears at each end of the pipeline (i.e., V1 and V2) is directly proportional to the induced electric field (E), inversely proportional to the propagation constant (γ) and the characteristic impedance (Z0), and dependent on the relative impedances to ground, Z1 and Z2 at each end of the pipe. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:8 In practical terms the variables can be listed as follows into two categories—those factors that affect the induced electric field (E) and those that affect the longitudinal impedance (Z) and the shunt admittance (Y). 4.1.2 Factors that Affect the Induced Electric Field (E) The value of the induced electric field (E) is a function of the following factors and events: 4.1.2(a) Solar Cycle Variations The solar cycle produces peaks of solar activity at approximately 11-year intervals. These periods correspond to a change in the location of the north and south magnetic poles of the sun. This periodicity of solar activity is illustrated in Figure 4-8, which is a history of geomagnetic effects over the last 150 years. The intensity of sunspot activity on average appears to be increasing with time. The next peak should be expected from approximately 2011 to 2013, whereas a general quiescent period should be expected from 2005 to 2007. Geomagnetic Effects 100 Sunspot Number 80 150 60 100 40 50 Magnetic Disturbances 200 20 0 0 1860 1880 1900 1920 Year 1940 1960 1980 2000 Figure 4-8: History of Geomagnetic Effects on Ground Technology Courtesy of D.H. Boteler, Geological Survey of Canada, Geomagnetic Laboratory, Ottawa, Ontario, Canada CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4.1.2(b) 4:9 Sun’s Rotational Frequency The sun’s rotational frequency of approximately 27 days will produce a variation in the solar plasma because sunspots and solar flares are not uniformly distributed over the sun’s surface. 4.1.2(c) Earth’s Rotation The earth’s rotation means that a metallic structure will experience the geomagnetic difference between the sun and night side of the earth. Voltage changes that have a repeating 24-hour variation are often called “diurnal” fluctuations. Diurnal fluctuations in pipe-to-soil potential are evident in Figure 49. Note that in this case the most electropositive potentials occur at midday. -2000 -1500 -1000 -500 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 0 12:00 Pipe Potential wrt CSE (mV) -2500 Time (Atlantic) Figure 4-9: Pipe-to-Soil Potential Variations with Time 4.1.2(d) Plasma Magnetic Field Direction The direction of the plasma magnetic field has a significant impact on the magnitude of the induced electric field. When the solar particles leave the sun, the sun’s magnetic field at the point of emission is frozen in the plasma blob. When the plasma magnetic field is directed southward (against the earth’s magnetic field), then geomagnetic substorms produce a larger electric field. However, when the plasma magnetic field is northward, there are no significant changes in the induced electric field. Hence the impact of a corona mass ejection CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:10 on the induced electric field is very much a function of the alignment of its magnetic field relative to the earth’s magnetic field. 4.1.2(e) Proximity of Pipeline to a Sea Coast The proximity of the pipeline to a sea coast also introduces a potential change on a pipeline. As illustrated in Figure 4-10, a voltage gradient exists between the low-resistivity seawater and the higher-resistivity land. This is due to charge accumulation because of the larger induced currents in the sea compared to the land, which increases the electrical potential of the earth near the coast. This effect is also true on land at sudden transitions between high- and low-resistivity soils. Land Sea Earth Surface Potential Figure 4-10: Charge accumulation at sea coast resulting from larger induced currents in the sea compared to in the land. The charge accumulation increases the electrical potential of the earth’s surface near the coast. Source: Boteler, D.H., Gummow, R.A. and Rix, B., Evaluation of Telluric Current Effects on the Maritimes and Northeast Pipeline, NACE Northern Area Eastern Conference and Exhibition, Ottawa, ON, October 1999, p.11. Tidal activity can also generate ocean currents due to the Hall effect as illustrated in Figure 4-11. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:11 Magnetic Field v Sea Land E Earth Surface Potential Figure 4-11: Electric Field, E, generated by seawater moving with velocity, v, through the earth’s magnetic field, B Source: Boteler, D.H., Gummow, R.A. and Rix, B., Evaluation of Telluric Current Effects on the Maritimes and Northeast Pipeline, NACE Northern Area Eastern Conference and Exhibition, Ottawa, ON, October 1999, p.12. As the water moves with a velocity (v) in a perpendicular magnetic field, positive and negative charges are forced in opposite directions perpendicular to the tidal direction. The potential difference created by this tidal dynamo can be approximated by the following equation. E = VBZW where: Ε v BZ W = = = = [4-11] the potential difference the water velocity the vertical component of the magnetic field the width of the water channel Using this equation and assuming that the vertical component of the magnetic field was approximately 50×10-6 Tesla, a potential difference (E) of 52V was calculated for the Bay of Fundy—where some of the largest tides in the world occur.[2] 2 Boteler, D.H., Gummow, R.A. and Rix, B., Evaluation of Telluric Current Effects on the Maritimes and Northeast Pipeline, NACE Northern Area Eastern Conference and Exhibition, Ottawa, Ontario, October 1999, p.11. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4.1.2(f) 4:12 Pipeline Latitude The location of the pipeline relative to the earth’s magnetic poles has a major impact on the magnitude of the induced electric field. Figure 4-12 illustrates the probability of a geomagnetic peak at about 0.2% at mid-latitude and decreases toward the north pole and the equator. Figure 4-12: Geomagnetic Hazard Percentage of Probability of Occurrence Source: Molinski, Tom, Geomagnetically Induced Currents: - Causes, Effect, and Mitigation, IEEE Canadian Review – Fall 1996, p.13 The peak probability coincides with the general location of the auroral electrojet shown in Figure 4-3. Furthermore Figure 4-2b illustrates an electrojet located at the equator. There have been reports of telluric activity on pipelines located near the equator in Panama.[3] 3 Soto, Gonzalo, Control de Corrosion en El Oleoducto de Panama, El VIII Seminario Latinamericano de Corrosion y Electroquimica, 1985, City of Panama. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:13 4.1.3 Factors that Affect the Pipeline Lineal Impedance (Z) and Shunt Admittance (Y) Besides the geophysical factors that can affect the electric field magnitude, pipeline factors such as the lineal impedance (Z) and the shunt admittance (Y) also affect the induced voltage. Both the propagation constant (γ) and the characteristic impedance (Z0) of a pipeline, as indicated in Equations 4-1a and 41b, are dependent on these parameters. Small values of lineal impedance or small shunt admittance result in a small propagation constant that produces a more linear relationship between induced voltage and distance (Figure 4-13). large γ (electrically long) 0 0 small γ (electrically short) Figure 4-13: Telluric Induced Voltage Profile vs Distance for a Pipeline with Different Attenuation Constants A pipeline with a large propagation constant is considered electrically lossy or long and a pipeline with a small propagation constant is considered to be electrically “short.” 4.1.3(a) Effect of Coating Quality The shunt admittance of a coated pipeline is primarily a function of the coating quality and, to a lesser extent, the soil resistivity. Figure 4-14 shows that as the coating conductance increases, the voltage induced on a long pipeline decreases. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:14 25 E = 0.1 V/km 20 Coating = 1 microS/sq.m. 15 10 Coating = 10 microS/sq.m. 5 Coating = 100 microS/sq.m. 0 200 400 600 800 1000 Pipeline Length (km) Figure 4-14: Calculated Telluric Induced Voltage at the End of a Long Pipeline as a Function of Coating Conductance for an East-West Electric Field of 0.1V/km Source: Boteler, D.H., Evaluation of Telluric Current Effects on the Maritimes and Northeast Pipeline, Geomagnetic Laboratory, Geological Survey of Canada, Ottawa, Aug. 1998, p.8. A high-quality coating, although important for cathodic protection (CP) effectiveness, results in higher induced voltages because the induced current cannot easily leak to earth. 4.1.3(b) Effect of Isolating Fittings Isolating fittings in a pipeline produce a voltage peak on each side of the electrical isolation that are 180 degrees out of phase. Multiple isolating fittings therefore create multiple peaks, albeit with lesser voltage differences across the isolation (Figure 4-15). 0 0 1 isolating fitting in middle 3 isolating fittings no isolation Figure 4-15: Effect of Isolating Fittings on the Telluric Induced Voltage Profile on an Electrically Short Pipeline CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:15 Note that telluric currents are alternating. Therefore, the polarity will change periodically and the voltage appearing across the isolator will be greater than the voltage-to-earth on either side of the isolator. 4.1.3(c) Effect of Pipeline Directional Change A change in pipe direction has the same impact on the telluric induced voltage (Vt) as it does with the induced AC voltage profile where the pipe crosses or leaves the powerline right-of-way. Because of the electromagnetic discontinuity created by the direction change, a voltage peak is created as illustrated in Figure 4-16 for an electrically long pipeline. Vt pipeline bend Figure 4-16: Effect of Pipeline Directional Change on the Telluric Induced Voltage Profile The same effect as illustrated in Figure 4-16 will also occur at a sudden change in earth conductivity (e.g., clay/rock). But these earth conductivity changes are more difficult to predetermine than directional changes. In both cases, however, the induced voltage is also dependent on the direction of the induced electric field. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4.2 4:16 Measuring the Geomagnetic Intensity and Determining the Electric Field (E) Geomagnetic activity is continuously monitored at observation posts around the world that record the geomagnetic disturbances caused by the solar wind. In North America, geomagnetic activity information can be obtained at www.geolab.nrcan.gc.ca in Canada and www.sel.noaa.gov in the U.S. The magnetic variations are recorded by magnetometers in units of nanoteslas (nT). There are a number of indexes that have been created to express the geomagnetic activity. For pipelines the Kp index is the most useful. The Kp index is an arithmetic average based on three-hour intervals. This index is logarithmic and spans from 0 (quiet) to 9 (severe), where activity greater than Kp 4 is considered a geomagnetic storm. The probability of a geomagnetic storm decreases logarithmically as the Kp index increases (Figure 4-17). 10- 0 2 10-1 2 10- 2 2 -3 10 2 -4 10 0 1 2 3 5 4 6 7 8 9 Kp Figure 4-17: Average Occurrence of 3-Hour Intervals with the Magnetic Activity Index Kp Equal to or Greater than a Specified Value. Kp=9 Corresponds to a Severe Magnetic Storm Source: Boteler, D.H. and Rix, B., Telluric Current Considerations in the CP Design for the Maritimes and Northeast Pipeline, NACE Corrosion 2001, Houston, TX, March 2001, Paper No. 317. This figure shows that small disturbances occur frequently but the most severe storms are very infrequent. A Kp 6 storm, which is likely to occur 2 percent of the time, is considered significant because it relates to an average electric field magnitude of 100mV/km as shown in Figure 4-18. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:17 1000 100 10 1 0 1 2 3 5 4 6 7 8 9 Kp Figure 4-18: Peak Electric Field Magnitudes as a Function of Kp Source: Boteler, D.H. and Rix, B., Telluric Current Considerations in the CP Design for the Maritimes and Northeast Pipeline, NACE Corrosion 2001, Houston, TX, March 2001, Paper No. 317. Note that within a three-hour period for a Kp 6 storm, the electric field ranged from approximately 30 to 300 mV/km. The foregoing data were used in estimating the effect on a 762 mm diameter 1000-km pipeline running from Goldboro, Nova Scotia, through to New England at subauroral latitudes. For pipelines in other geographical locations, a similar plot could be obtained from the appropriate geomagnetic laboratory. This is necessary to produce an accurate prediction of tellurically induced voltages using the DSTL model. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4.3 4:18 Interference Effects of Telluric Current on Pipelines 4.3.1 General Considerations The impact of these geomagnetically induced currents has historically been considered more of a nuisance when measuring CP parameters than a serious corrosion concern. However, there are three main areas of concern regarding the effects of geomagnetically induced currents on coated pipelines: 1. Corrosion during the positive half-cycles of the telluric waveform. 2. Accuracy of pipeline current and potential measurements when determining the level of CP for comparison with industry criteria. 3. Coating damaged caused by excessively negative potentials during the negative half-cycles of the telluric waveform. 4.3.2 4.3.2(a) Corrosion Theoretical Considerations During the time when telluric current transfers from the pipe to earth (positive portion of the telluric cycle), the charges must transfer through an oxidation reaction. For a steel pipe without CP, the primary oxidation reaction is corrosion of the steel (Figure 4-19) and as expressed by the following reaction: Fe° ⇒ Fe++ + 2e- (corrosion) Grade it Feo = Fe++ + 2eFigure 4-19: Oxidation Reaction at Pipe Surface During Telluric Current Discharge in the Absence of CP CP Interference © NACE International, 2006 January 2008 [4-12] Telluric Current Interference 4:19 Theoretically, approximately 10kg of steel will be lost in 1 year for every ampere of continuous direct current (DC) that discharges. When a pipeline is being cathodically protected or is receiving telluric current (Figure 4-20), the charge transfer reactions can be one or both of the following depending on the soil conditions; H 3 O + + e- or 2H2O + O2 + 4e- ⇒ H0 + H20 (in deaerated or acidic soils) [4-13] ⇒ 4OH- (in alkaline or neutral aerated soils) [4-14] Figure 4:20: Reduction Reactions During Negative Cycle Telluric and CP Current Pick-up Both these reduction reactions produce a high-pH environment, typically in the range of 10 to13, at coating flaws (holidays). The magnitude of the pH has been shown to be proportional to the logarithm of the current density[4] as shown in Figure 4-21. 4 Thompson, N.G., Barlo, T. J., Fundamental Process of Cathodically Protecting Steel Pipelines, 1983 International Gas Research Conference, p.279. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:20 14 13 12 11 10 9 8 10-9 10-7 10-8 10- 6 10- 5 10-4 C.P. Current Density, A/cm2 Figure 4-21: Steel Surface pH versus Applied CP Current Density When positive charges transfer from a surface that has been cathodically protected, the initial oxidation reaction is therefore likely to result in the formation of a passive film (Figure 4-22).[5] Here it can be seen that, as the steel becomes progressively more cathodically polarized, the anodic polarization curve exhibits progressively more passive behavior. -0.5 -0.6 -0.7 Before polarization anodic polarization -0.8 -200 mV mV -200 -0.9 -1 -400 mV -1.1 -1.2 -1.3 0.0001 0.001 0.01 0.1 1 Current Density (mA/cm2 ) Figure 4-22: Polarization Curves after Several Days of Potentiostatic Polarization Source: Hesjevik, S.M. and Birketveit, O., Telluric Current on Short Gas Pipelines in Norway – Risk of Corrosion on Buried Gas Pipelines, NACE Corrosion 2001, Houston, TX, Paper No. 313, p16. (Figure redrawn) 5 Hesjevik, S.M. and Birketveit, O., Telluric Current on Short Gas Pipelines in Norway – Risk of Corrosion on Buried Gas Pipelines, NACE Corrosion 2001, Houston, TX, Paper No. 313, p16. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:21 Where the pipeline has been cathodically protected for a long period of time and the pH at the pipe/soil interface is high, the initial potential/current density relationship should be similar to that shown in Figure 4-23. Here, for a pH of 12, one can see that, as the potential moves from approximately –950mVsce, the current increases up to the primary passive potential of approximately –850mVsce, after which it diminishes as the potential moves through the passive range to the start of the transpassive region at approximately +600mVsce. Because the corrosion reaction is one that produces a passive film, then the initial corrosion rate (i.e., the current density) resulting from this anodic excursion would be low. +700 +500 -200 -400 -600 pH 12.0 -800 -1000 -1200 1 10 100 1,000 10,000 Current Density, Microamps/cm2 Figure 4-23: Experimental Anodic Polarization Curve of Steel in Hydroxide (pH 12.0) Source: Thompson, N.G., Lawson, K.M., and Beavers, J.A., “Exploring the Complexity of the Mechanism of Cathodic Protection”, Corrosion ’94, Paper No. 580, NACE International, 1994, p.11. (Figure redrawn) If the telluric current discharge is sustained but the residual pH remains high, then the oxidation reaction could be expressed by Equation 4-15, the oxidation of hydroxyl ions, or by Equation 4-16, the hydrolysis of water (Figure 4-24); neither of these equations results in metal loss. 4OH- ⇒ 2H2O + O2↑ + 4e+ 2 H2O ⇒ O2↑ + 4H + 4e CP Interference © NACE International, 2006 January 2008 - [4-15] [4-16] Telluric Current Interference 4:22 Grade icp it 4OH2H2O H2O + O2 + 4eO2 + 4H+ + 4e- Figure 4-24: Telluric Current Discharge from a Cathodically Protected Pipe Accordingly, the total corrosion that occurs at a coating defect as a result of current discharge is not strictly proportional to the charge transferred as would be predicted by Faraday’s Law for a steady-state DC. 4.3.2(b) Calculating the Corrosion Rate Cyclic variations in telluric current of equal amplitude and period will corrode steel less than a steady state DC of the same magnitude applied for the same time period, as previously discovered in a National Bureau of Standards (NBS) investigation[6] and as illustrated in Figure 4-25. 6 McCollum, B., Ahlborn, G.H., Influence of Frequency of Alternating or Infrequently Reversed Current on Electrolytic Corrosion, National Bureau of Standards Tech Paper No. 72, 1916. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:23 100 90 LEGEND: Soil Soil + Na2CO3 80 70 60 50 40 30 20 10 0 -10 1/60S 1/15S 1S 5S 1M 5M 10M 1Hr. 2Days 2Weeks D.C. Logarithm of Length of Time of One Cycle Figure 4-25: Coefficient of Corrosion at Different Frequencies for Iron Electrodes Denoted as Average Electrode Loss This study, which was commissioned to determine the relative corrosivity of stray currents arising from DC transit systems, has some merit with respect to telluric stray currents because the periods of activity are somewhat similar. In fact Campbell[7] produced the following mathematical relationship using the NBS findings to estimate the corrosion as a function of the telluric current cycle for a fixed amplitude: C = ( 4.7 ± 1.3) T+0.186 where: [4-17] C is percent of DC corrosion that would occur at the same amplitude T is the period of the current cycle in seconds Peabody[8], as shown in Figure 4-26, also summarized the NBS findings in a different graphical representation that demonstrates a relationship between the logarithm of the period and the logarithm of the percentage of corrosion compared to an equal amount of DC. 7 Campbell, W.H., Induction of Auroral Zone Electric Currents Within the Alaska Pipeline, Pure and Applied Geophysics, Vol. 116, No.6, 1978, p.1167. 8 Peabody, A.W., Corrosion Aspects of Arctic Pipelines, Materials Performance, Vol. 18, No.5, May 1979, p.30. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:24 Time Interval in Hours Between Current Reversals 100 10 2 1 0.5 0.1 0.01 0.001 5 10 22 29 50 100 Percentage of Direct Current Corrosion Rate Figure 4-26: Effect on Corrosion Rate of Reversing Direction of Current Compared To Steady State DC and Length of Time Between Reversals Source: Peabody, A.W., Corrosion Aspects of Arctic Pipelines, Materials Performance, Vol. 18, No.5, May 1979, p.30. (Figure redrawn) CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:25 Although telluric frequencies cover a wide spectrum, the induced electric field peaks typically at periods between 30 minutes to 2 hours.[9] This corresponds to corrosion activity that would be approximately 22 to 29% of an equivalent DC. It should be noted, however, that diurnal telluric activity—which is typically less intense than the shorter telluric fluctuations—would produce a corrosion rate of approximately 50% of an equivalent DC because it would have a 12-hour period. The amount of stray telluric current produced during the positive period depends on the intensity of the telluric disturbances. On very well-coated modern pipelines, current transfer between the pipe and soil occurs primarily at small coating defects. Relatively small potential fluctuations in the order of 0.5 to 1.0V can produce a large current density as shown in Figure 4-27.[10] Here, for a 1-cmdiameter circular holiday in a 0.3-mm thick coating, which is a typical thickness for fusion bonded epoxy coatings, the current density for a soil resistivity of 1000 Ω-cm and a telluric voltage change of 1.0V, would be approximately 2500µA/cm2 and produce a corrosion rate of approximately 31.3mm/y. Corrosion Current Density (µA/cm2) 100,000 t t=0 t = 0.3mm 10,000 t = 1mm Pipe Wall t = 3mm 1µA/cm2 = 0.0125mm/a (Fe) d Soil 2,500 t = 10mm 1,000 100 0.1 1 10 100 1000 Defect diameter (mm) Figure 4-27: Corrosion Current Density at a Coating Defect having an Applied Voltage of 1.0V in 1000 ohm-cm Soil for Various Coating Thicknesses Source: Von Baeckmann, W., Schwenk, W., Handbook of Cathodic Protection, Portcullis Press, England, 1975, p.365. (Figure redrawn) 9 Campbell, W.H. and Zimmerman, J.E., Induced Electric Currents in the Alaska Oil Pipeline Resulting from Auroral Electrojet Current Sources, Geophysical Journal of the Royal Astronomical Society, Vol. 61, No.2, p.1164. 10 Von Baeckmann, W., Schwenk, W., Handbook of Cathodic Protection, Portcullis Press, England, 1975, p.365. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:26 The corrosion rate arising from Figure 4-26 for a 1-cm-diameter defect with a steady state voltage of +1V applied in 1,000 Ω-cm soil can be expressed as follows: CR = Ki y P where: Ki = corrosion current density factor (2.5 x 10-3 A/cm2 per volt) P = corrosion penetration factor (12.5 x 10-3 mm/y per 10-6 A/cm2) CR = corrosion rate (mm/yr) To calculate the theoretical corrosion rate caused by telluric voltage fluctuations, modify the corrosion rate formula to account for the cyclic variations in the telluric wave form (Fp), the duration of time that the activity is present (Ft), and the magnitude of the telluric voltage (ΔVt ) as follows: CRtelluric = Ki y P y ΔVt y F(p) y F(t) ΔVt = change in potential of the pipe caused by telluric activity F(p) = fraction of steady state corrosion due to alternating period of the telluric current F(t) = fraction of time that telluric activity is present As would be expected, the corrosion rate for a given potential change (ΔVt) varies with the soil resistivity and the anodic transient time as illustrated in Figure 4-28. (A) (B) 100 10 Clay soil 1 0.1 Sandy soil 0.1 1 10 Anodic Transient Time (min) 100 Figure 4-28: Chart Showing the Influence of Anodic Transient Time with Respect to Corrosion Experienced by Probe in Sandy and Clay Soil. Line (A) Represents the Corrosion Rate Expected from Faraday’s Law for the Clay Soil, and Line (B) for the Sandy Soil, Respectively. Source: Birbilis, N., Holloway, L.J. and Forsyth, M., Technical Note: Simulated Transient Loss of Cathodic Protection for Buried Pipelines, Corrosion, Vol. 61, No.5, May 2005, p.500. (Figure redrawn) CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:27 This figure[11] shows the results of laboratory tests conducted on resistance probes placed in clay (4,000 Ω-cm) and sand (50,000 Ω-cm) soil. The probes were cathodically protected to –1000mVssc and subjected to an anodic transient to 0mVssc for 20% of the time for 1-minute, 10-minute, and 60-minute periods. Up to a 10-minute anodic period, the corrosion rate is less than 1% of the theoretical value based on the anodic current. It must be expected, however, that if the CP potential was less negative than –1000mVssc (IR drop free), then the corrosion rate would be greater than 1% of the theoretical value. 4.3.2(c) Telluric Corrosion Case Studies on Cathodically Protected Piping Although corrosion of steel pipelines due to telluric current activity is theoretically probable, it has not been considered by the pipeline industry as a serious threat to the integrity of cathodically protected pipelines. This view was probably a result of the findings of a study conducted by the American Gas Association on four pipelines in the U.S. between 1966 and 1970. Their investigation concluded that the effects are insignificant, both for coated, protected lines and for bare lines.[12] This study, however, focused on pipelines that were relatively short, located at latitudes lower than 46 degrees N, on relatively poorly coated pipelines, and during a period of relatively quiet telluric activity. Subsequent findings by other investigators[13,14] on existing, cathodically protected pipelines located in auroral zones using coupons showed that corrosion was mild but not insignificant. From a study on Norwegian pipelines, Henricksen, et al concluded that telluric current corrosion in auroral zones is approximately the same magnitude as normal soil corrosion where telluric corrosion is lacking. 11 Birbilis, N., Holloway, L.J. and Forsyth, M., Technical Note: Simulated Transient Loss of Cathodic Protection for Buried Pipelines, Corrosion, Vol. 61, No.5, May 2005, p.500. 12 Gideon, D.N. et al., Earth Current Effects on Buried Pipelines – Analysis of Observations of Telluric Gradients and their Effects, AGA Project PR-3-41, April 1970, p.71. 13 Henriksen, J.F. et al., Telluric Current Corrosion on Buried Pipelines, Proceedings of the 8th Scandinavian Corrosion Congress, NKM8, Helsinki, Vol. II, 1978, p.167-176. 14 Martin, B.A., Telluric Effects on a Buried Pipeline, Corrosion, Vol. 49(4), 1993, p.349. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:28 Martin[15], following a telluric corrosion study on a 515-km gas pipeline in northeastern Australia, reported corrosion rates in excess of 0.01 mm/y (4 mpy). A more serious case of tellurically caused corrosion was discovered in 2001 on a 24-in Ø natural gas pipeline east of Montreal, Quebec.[16] This fusion-bonded epoxy (FBE)-coated pipeline installed in 1998 was found to have a 60-mil pit at a subcriterion location identified during a close-interval potential survey; the results of the survey appear in Figure 4-29. The pipe-to-soil potential fluctuations at this location were later correlated with the magnetic field variation (Figure 4-30). 1800 1600 1400 Potential (-mVcse) 1200 1000 OFF ON crit 800 600 400 200 0 107000 -200 108000 109000 110000 111000 112000 113000 114000 115000 Kilometers Figure 4-29: Potentials Measured with Rectifiers ‘ON’ and ‘OFF’ Source: Brochu, B., Telluric Current Effects on Underground Steel Pipelines, NACE Northern Area – Eastern Conference, Quebec City, August 2001 15 16 Ibid [14] p.349. Brochu, B., Telluric Current Effects on Underground Steel Pipelines, NACE Northern Area – Eastern Conference, Quebec City, August 2001. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:29 44000 Pipe-to-soil potentials 43900 43800 43700 43600 Ey Geomagnetic Field 43500 1 2 :0 0 :0 0 1 2 :2 8 :4 8 1 2 :5 7 :3 6 1 3 :2 6 :2 4 1 3 :5 5 :1 2 1 4 :2 4 :0 0 1 4 :5 2 :4 8 1 5 :2 1 :3 6 Figure 4-30: Magnetic Field Intensity and Pipe-to-Soil Potential Superimposed Source: Brochu, B., Telluric Current Effects on Underground Steel Pipelines, NACE Northern Area Eastern Conference, Quebec City, August 2001. The high corrosion rate of approximately 15 mils/y was unlikely because of the high soil resistivity. Further investigation indicated that the pipeline at the corrosion location was situated above a low-resistivity (100 Ω-cm) graphite schist that extended northeastward for more than 100 km toward the Gulf of St. Lawrence. This soil anomaly was thought to provide a relatively low-resistance path between the pipeline and the Atlantic Ocean. 4.3.3 Impact on Accuracy of Current and Potential Measurements In the absence of a stray current on a cathodically protected structure, a pipe-tosoil potential measured using a high-input resistance voltmeter will be the sum of the polarized potential (Ep) that appears across the pipe-to-earth interface and the voltage drop (Ve) in the earth due to the CP current (Icp) through the earth path resistance (Re) Equation 4-17. Vm = Ep + Ve where: [4-18] Ve = Icp • Re The polarized potential (Ep) must be equal to or more electronegative than –850 mVcse in order to satisfy the NACE –850 mV criterion. It is usual on impressed CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:30 current systems to momentarily interrupt the CP current (Icp = 0) so that an instant-off potential (Ei-off), measured the moment of interruption, is a reasonably accurate representation of the polarized potential (Ep). If a telluric current is present, the pipe-to-soil potential difference measurement will incorporate an additional voltage drop (Vt) owing to the telluric current in the soil path between the reference electrode and the pipe as represented by equation 4-19 and as illustrated in Figure 4-31. Vm = Ecp + Ve ± Vt Test Station [4-19] Voltmeter V Portable Reference Electrode Grade Icp Ve Pipe Test Lead Vt It Ep Figure 4-31: Schematic of Potentially Controlled CP System Used to Mitigate Telluric Current Effects Source: Gummow, R.A., Telluric Current Effects on Corrosion and Corrosion Control Systems on Pipelines in Cold Climates, NACE Northern Area Western Region Conference, Alaska, Feb. 2001, Paper CldCli01, p.12. Because telluric current is alternating, the error can make the pipe appear either better protected or more poorly protected depending on its direction and change the polarized potential if the telluric current is sustained with time. Unlike an impressed current, the telluric current cannot be arbitrarily interrupted, which then compromises the accuracy of a pipe-to-soil potential measurement. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:31 4.3.4 Impact of Telluric Current on Pipeline Coatings NACE SP0169[17] cautions that “the use of excessive polarized potentials on externally coated pipelines should be avoided to minimize cathodic disbondment of the coating.” This precaution is typically being interpreted as a maximum polarized potential equal to –1200mVcse. During geomagnetic storms as illustrated by the calculated instant-off potential in Figure 4-32, –1200 mVcse potentials can easily result on well-coated pipelines during periods of telluric current pick-up. 5.5 2.0 5.0 1.5 Eon 1.0 4.5 Calculated Eoff 0.5 Current Density 4.0 3.5 0.0 -0.5 3.0 -1.0 2.5 -1.5 2.0 -2.0 1.5 -2.5 1.0 -3.0 0.5 -3.5 0.0 -4.0 -0.5 -4.5 -1.0 -5.0 -1.5 -5.5 03:50 03:55 04:00 04:05 04:10 04:15 04:20 04:25 04:30 04:35 04:40 -2.0 04:45 Time Figure 4-32: Current Flow and Calculated OFF Potentials during a GIC Incident Source: Hesjevik, S.M. and Birketveit, O., Telluric Current on Short Gas Pipelines in Norway – Risk of Corrosion on Buried Gas Pipelines, NACE Corrosion 2001, Paper #01313. (Figure redrawn) Cathodic disbondment and cathodic blistering both result from water migration through the coating due to electro-osmosis. Generally, the thicker the coating is and the better its moisture transmission resistance, then the less susceptible it is to 17 NACE Standard SP0169 – Control of External Corrosion on Underground or Submerged Metallic Piping Systems, NACE International. NACE International publishes three classes of standards: standard practices, standard material requirements, and standard test methods. Until June 23, 2006, NACE published standard recommended practices, but the designation of this type of standard was changed to simply standard practice. New standards published after that date will carry the new designation (SP), and existing standards will be changed as they are revised or reaffirmed. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:32 these effects. Accordingly, FBE, being a thin film coating, is particularly vulnerable to moisture transmission and consequential cathodic blistering at locations of poor coating adhesion and cathodic disbondment as a result of the high pH developed in the blister. 4.3.5 Impact on Output of a CP Transformer/Rectifier An impressed current transformer/rectifier will pass a telluric current through the rectifier element to its groundbed if the telluric current is in a discharge cycle, as illustrated in Figure 4-33. It Icp It Icp groundbed pipeline Figure 4-33: Telluric Current Through a Bridge Rectifying Element During a Discharge Cycle This will also be true for a center-tapped transformer/rectifier. When operating in constant voltage mode, the total output current (Io) of the transformer/rectifier will increase: that is: Io = Icp + It [4-20] This is because the voltage difference (Vo) between the pipe and the groundbed is the sum of the superimposed telluric voltage and the transformer/rectifier output voltage (VTR): that is: Vo = VTR + Vt [4-21] CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:33 But if the transformer/rectifier is operating on constant current, the CP current (Icp) will drop when the telluric current (It) is present: Icp = Io – It that is: [4-22] This is undesirable because an increase rather than a decrease in the transformer/rectifier output would be preferred at the time of a telluric current discharge. Hence a transformer/rectifier should not be operated in constant current when the pipeline is subjected to telluric current activity. 4.4 Mitigating the Effects of Telluric Current 4.4.1 Mitigating Corrosion Impact 4.4.1(a) Making the Pipeline Electrically Continuous and Grounded Telluric voltages on pipelines arise from electromagnetic induction and are therefore analogous to induced alternating current (AC) voltages. Similarly, grounding the pipeline can be an effective method of mitigating telluric voltages just as it is with AC voltages. Telluric voltages, which appear across an insulated flange, can be reduced by electrically bonding around the isolating joint. As with AC mitigation, however, the bond must be designed to maintain the performance of the CP system. A telluric bond switch (Figure 4-34) has been used[18] to pass telluric current across an insulator separating onshore and offshore portions of a cathodically protected pipeline. 18 Boteler, D.H., Gummow, R.A., and Rix, B.C., Evaluation of Telluric Current Effects on the Maritimes and Northeast Pipeline, NACE International Northern Area Eastern Conference, Ottawa, Paper No. 8A, 3, October 24, 1999. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:34 MOV DUAL DIODE To Offshore Pipeline Auto Resetting Fuse Variable Resistor Shunt Shunt To Onshore Pipeline By-pass Switch Figure 4-34: Schematic of a Telluric Bond Switch Back-to-back diodes provide a fault path for the large telluric currents once the breakover voltage of the diodes (typically 0.8V) has been breached. These diodes are therefore rated to handle the largest expected telluric current typically arising from a once-per-year severe storm (i.e., Kp 9 on Figure 4-18). Adjustment of the variable resistor allows for a steady-state drain of current to balance the CP systems between the onshore and offshore sections of the pipeline. Lightning protection is provided by the metal oxide varistor. It is also possible to mitigate telluric effects by connecting the pipeline to electrical ground using AC coupling-DC isolating devices such as isolating surge protectors and polarization cells. Grouped galvanic anodes connected to each side of the isolating fitting can also be used but the anode capacity must be chosen to provide a reasonable life and with enough current output to compensate for any residual telluric current discharging from the pipeline. 4.4.1(b) Using CP CP systems can be designed and operated to mitigate telluric voltage fluctuations by a combination of two related mechanisms. Impressed current output can be increased to compensate for a telluric current discharge, or galvanic anodes can provide a grounding path for the telluric current to pass to earth. The capacity to perform these functions varies with the type and operating characteristics of the CP system relative to the operating characteristics of the pipeline system and the magnitude of the telluric activity. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4.4.1(b)(i) 4:35 Sacrificial Anodes Sacrificial CP systems have a limited voltage capacity to compensate for a telluric potential shift because they have a relatively small fixed output voltage. They do, however, offer an alternative path to earth for the telluric current (It) because of their low resistance to earth compared to a coated pipeline. Some proportion of the telluric current (Itl) will transfer to earth via the anode (Figure 4-35), depending primarily on the anode-to-earth resistance compared to the pipe-toearth resistance—both locally and looking down the pipe in the direction of the current. If the CP current (Icp) is equal to or greater than the residual telluric discharge current (Itll), then stray current corrosion will not occur on the pipe under the influence of the anode. I't'' It I't' Icp galvanic anode Icp + I't I't where: residual telluric current discharge telluric current discharge from galvanic anode It = I't + I''t + I''' t Figure 4-35: Mitigation of Telluric Current Discharge Effects using Galvanic Anodes This CP method, which makes the pipeline electrically lossy, has been used on the Trans-Alaska pipeline[19] in the form of a zinc ribbon anode that was placed at pipe invert elevation on each side of the pipe for the full extent of the underground portion of the pipeline. Grouping of zinc and magnesium sacrificial anodes at selected intervals has also been shown to be effective by Henriksen, et al.[20] when used on a pipeline in northern Norway, where the telluric potential fluctuations were reduced from ± 5 V to ± 0.1 V (Figure 4-36). Just as with induced AC mitigation, the more electrically lossy a pipeline is, the lesser the magnitude of the telluric voltage fluctuations. 19 20 Martin, B.A., Telluric Effects on a Buried Pipeline, Corrosion, Vol. 49 (4), 1993, p.343-350. Henriksen, J.F., Elvik, R. and Granasen, L., Telluric Current Corrosion on Buried Pipelines, Proceedings of 8th Scandinavian Corrosion Congress, Tehory andPraxis at Corroisons Prevention, Volume II, p.167-176, Helsinki, 1978. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:36 mV, Cu/CuSO4 Grounding -3000 OUT IN OUT IN OUT IN IN OUT -2000 -1000 0 +1000 +2000 2100 5/10-74 2300 0100 6/10-74 0300 0500 0700 0900 Time Figure 4-36: Effect of Connecting and Disconnecting Groups of Galvanic Anodes to a Pipeline Subjected to Telluric Current For instance, if a 0.5-m-diameter coated pipeline has a conductance of 10-6 S/m2 in 10,000 Ω-cm soil (a reasonable expectation for modern coatings) then it has a conductance per 100m of 0.157 x 10-3 S. Example Calculation: Consider a coating having a specific leakage conductance (G) of 10-6 S/m2 in 10,000 Ω-cm soil. For 100m of 0.5-m-diameter pipe, the leakage conductance (g) would be: g pipe = G 10,000 × A P = 10 S/m × 157 m -6 2 g pipe = 1.57 × 10 - 4 S AP = πdl 2 = 3.14 × 0.5 m × 100 m = 157 m 2 Assume a packaged magnesium anode (9.1 kg × 1.52m lg) is attached to the piping for each 100m length. The conductance (g) of the anode to earth in 10,000 Ω-cm soil is given by the following equation, which is the reciprocal of the anode resistance as calculated by Dwight’s equation for a vertical electrode: CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:37 g anode = where: L = 1.52 m d = 0.12 m ρ = 100 Ω-m 1 R anode = 2πL 1 × 8L ρ -1 ln d = 6.28 × 1.52 m × 100 Ω - m [4-23] 1 12.16 -1 ln 0.12 1 = 0.0263 S 3.61 = 26.3 × 10 -3 S = 0.095 × g anode The net conductance (gn) for a 100m of pipe with the anode attached is therefore: g n = g anode + g pipe = 26.3 × 10 -3 S + 0.157 × 10 -3 S = 26.5 × 10 -3 S This is an increase in conductance of 167 times, which is well over two orders of magnitude. As Figure 4-37 shows, an increase in conductance of this order can significantly reduce the magnitude of telluric induced voltage (i.e., 90% reduction). CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:38 Figure 4-37: Effect of Increased Coating Conductance on the Voltage on Each Side of Isolated Flanges There is some belief that the telluric current seen on pipelines results primarily from current transfer (conductance) between the pipe and earth rather than from inductance directly. If this were the case, then magnesium anodes would be preferred over zinc anodes because they would not pick-up current until the pipeto-earth potential exceeded their open-circuit potential (approximately – 1.750Vcse). In contrast, zinc would accept telluric current when a potential of – 1.100Vcse was exceeded. Magnesium anodes would therefore lessen the amount of current pick-up and provide more CP current compared to zinc. There may also be net CP benefit with the use of sacrificial anodes in the presence of a telluric current. Results from an experiment[21] that applied a signal simulating a telluric wave form to a combination of a steel pipe and a zinc ribbon found that there was a net pick-up of the alternating signal on the pipe. Conversely, there was a net increase in the amount of current discharged from the 21 Unpublished results from research to determine the potential and current effects on a steel pipe/zinc ribbon couple, CORRENG Consulting Service Inc., Downsview, ON, Canada 1993. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:39 anode. This may be due to the fact that the anode does not pick-up the AC until its open-circuit potential is exceeded and the pipe does not discharge current until the anode potential is polarized electropositively to the pipe polarized potential. 4.4.1(b)(ii) Impressed Current Systems Impressed current CP (ICCP) systems can theoretically be designed with unlimited voltage capacity, although it is inefficient to continuously operate the system at higher voltages just to provide a buffer for the anticipated telluric positive voltage shift. Moreover, the very high negative potentials produced, as a result of operating ICCP systems at high current outputs, can cause cathodic disbondment of the coating. Martin[22] found that operating rectifiers in constant voltage or constant current mode had “little mitigative effect” because they caused “overprotection during local negative transients and underprotection during local positive transients”. There have been reports[23,24] that telluric voltage fluctuations are more pronounced near rectifier locations than between them. There is no doubt that anodic telluric currents will pass to earth through the rectifying element in the transformer-rectifier as discussed in Section 4.3.5. When operating in constant current mode (where Io is kept constant), CP current will be reduced by the amount of the telluric current through the rectifier— thereby diminishing the amount of CP available to counteract the residual telluric current discharge from the pipe. Hence, it would seem that—from a telluric current mitigation point of view—impressed current systems should not be operated in constant current mode. 22 Martin, B.A., Telluric Effects on a Buried Pipeline, Corrosion, Vol. 49 (4), 1993, p.345. Proctor, T.G., Experience with Telluric Current Interference in the Cathodic Protection of a Buried Pipeline in New Zealand, NACE, Corrosion /74, Paper No. 57, p11. 24 Private communication with Ian Munro, Corrosion Service Co. Ltd., Feb. 2001. 23 CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:40 Martin[25] and other operators[26,27] have used the potential control mode to successfully ameliorate telluric currents even though Proctor[28] concluded that “the value of constant potential impressed current power sources in compensating for telluric current interference is questionable”. The voltage and current output of these units change automatically in response to the pipe potential as measured to a local reference electrode, as illustrated schematically in Figure 4-38. Potentially Controlled DC Power Supply - S R + Icp and It Icp and It Permanent Reference Electrode/Coupon Remote Groundbed Figure 4-38: Schematic of Potentially Controlled CP System used to Mitigate Telluric Current Effects Here the coupon potential is measured continuously with respect to the permanent reference electrode and compared to a pre-set potential in the controller of the DC power supply. When a telluric current attempts to discharge from the pipe/coupon, the reference senses the positive potential shift and the power supply immediately increases its output to maintain the set potential value. The impressed current system therefore presents a negative resistance path for the telluric current to earth and thus there is no residual discharge of telluric current from the pipe as long as the voltage or current output of the power supply is within its rating. A coupon is used to minimize IR drop between the reference electrode and the nearest holiday so that the rectifier can control to a potential that has minimal IR drop component. 25 Martin, B.A., Telluric Effects on a Buried Pipeline, Corrosion, Vol. 49 (4), 1993, p.345.. Peabody, A.W., Corrosion Aspects of Arctic Pipelines, Materials Performance, Vol. 18 (5), May 1979. 27 Degerstedt, R.M., Kennelley, K.J., Lara, P.F., and Moghissi, O.C., Acquiring “Telluric-nulled” Pipe-tosoil Potentials on the Trans Alaska Pipeline, Corrosion ’95, Paper No. 345, NACE International. 28 Proctor, T.G., Pipeline Telluric Current Difference as one Phase of a Wider Interdisciplinary Technological Problem, NACE, Corrosion /74, Paper No.60, p.16. 26 CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:41 600 6 500 5 400 4 300 3 200 2 100 1 0 Pipe Potential wrt ZRE (mV) -100 0 1 2 3 4 5 6 7 -1 -200 -2 -300 -3 -400 -4 -500 -5 -600 -6 -700 -7 -800 -8 -900 -9 -1000 -10 Day Figure 4-39: Pipe Potential and Rectifier Current Output vs Time for an Impressed Current System Operating in Potential Control Note that, in this example, the rectifier operates only when the pipe potential attempts to go more electropositive than the set potential of –100 mV/ZRE (approximately –1200 mV/cse). Telluric current is drained to earth during periods of telluric current discharge. During periods of telluric current pick-up, the current output goes to zero and thus limits the magnitude of the negative potential applied across the coating. This mode of operation effectively eliminates the positive telluric voltage fluctuations in the vicinity of the rectifier while minimizing excessively negative potentials and maximizing the life of the groundbed. This technique works most effectively when the transformer/rectifier and groundbed are located at peak locations of telluric current activity. CP Interference © NACE International, 2006 January 2008 Rectifier Current (A) The power supply voltage and current capacity must be sized to provide the needed CP current in addition to the amount of telluric current to be drained. This type of CP system functions as a telluric current “forced drainage” system. Its mitigating effect is illustrated in Figure 4-39, which compares typical rectifier current output and pipe potential over time. Telluric Current Interference 4:42 4.4.2 Compensating for Measurement Error Caused by Telluric Current Because geomagnetically induced current cannot be arbitrarily interrupted, an alternative pipe-to-soil potential measurement method has been employed by some companies[29,30]. The method uses a small steel coupon installed next to the pipe, which is interconnected with the pipe inside a test station. The coupon simulates the pipe/soil surface at a defect in the coating. When the coupon is temporarily disconnected and the reference electrode is placed in the soil tube (Figure 4-40), both the telluric and CP voltage drops in the earth are removed from the measured potential difference and the “instant off” potential (Ep) of the coupon is measured for comparison to the –850 mVcse criterion. Test Station Switch Voltmeter V Grade Pipe Test Lead Portable Reference Electrode Non-metallic Tube filled with Sand/ Bentonite Mixture Steel Coupon Figure 4-40: Typical Pipe-to-Soil Potential Measurement at Test Station having a Steel Coupon and Soil Tube 29 Stears, C.D., Moghissi, O.C., Degerstedt, R.M., and Bone, L., Field Program on the Use of Coupons to Monitor Cathodic Protection of an Underground Pipeline, Corrosion ’97, Paper No. 564, NACE International, Houston, TX, 1997. 30 Greenwood, R., The Effects of Transient Stray Current on Cathodically Protected Pipelines, British Gas Engineering Research Report, July 1986, p4-6. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:43 This test arrangement is not suitable for recording the polarized potential with time, however, because the coupon has to be disconnected for each measurement. The use of a reference/coupon combination, as illustrated in Figure 4-41, has proved to be an excellent method of recording a polarized potential with time. The coupon in this device does not require disconnection because the reference is located inside the pipe coupon, where there is neither CP nor telluric voltage gradient. Test Station VR Recording Voltmeter Grade Pipe test lead Coupon test lead Zinc Reference test lead Coupon/Reference Probe Figure 4-41: Typical Pipe-to-Soil Potential Recording at a Test Station Using a Coupon/Reference Probe Figure 4-42 compares the pipe/coupon potential recorded to a CSE reference placed on grade and the reference located inside the coupon. The difference between the potential values is the soil voltage gradient caused by both the telluric and CP currents. Note that, despite the significant potential fluctuations in the potential measurement using a surface copper-copper sulfate electrode, the actual potential at the coupon/soil interface is relatively stable with time. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:44 -3000 -2500 Potential (mVCSE) -2000 Potential wrt portable reference on Grade -1500 -1000 Potential wrt Coupon Reference -500 *For convenience, the readings were converted in mVCSE using a zinc potential of -1100 mVCSE 0 9:00 11:24 13:48 16:12 18:36 21:00 23:24 Time Figure 4-42: Comparison between Pipe/Coupon Potential with Time recorded with respect to a Copper-Copper Sulfate Reference on Grade and to a Coupon/Reference Probe Located at Pipe Depth Although the use of a coupon is a relatively simple solution at a test station, the measurement of telluric free potentials is more complex for close interval potential surveys (CIPS) where the reference is moved and placed over the pipe at intervals (typically < 3 m) along the route of the pipeline. Proctor[31] proposed a measurement method to compensate for the telluric induced voltage that involved the correction of the measured potential (Vm) with respect to the moving reference by the change in potential (ΔVf) measured with respect to a fixed reference located at a nearby test station such that where: 31 Vps = Vm ± ΔVf [4-24] ΔVf = Vfave ± Vf [4-25] Proctor, T.G., Experience with Telluric Current Interference in the Cathodic Protection of Buried Pipeline in New Zealand, Materials Performance, Vol. 13, No. 6, 1974, p29. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:45 This measurement technique is illustrated in Figure 4-43 in which two separate data loggers are used to record the potentials with respect to the fixed and moving electrodes. Roving Datalogger Synchronized Fixed or Moving Datalogger Vm Vf Moving Reference Fixed Reference Survey Length Figure 4-43: Pipe-to-Soil Potential Measurement Method to Compensate for Telluric Current Effects During a Close Interval CP Survey This technique can be used with synchronous interruption of the rectifiers such that a telluric compensated “instant off” potential can be calculated in software from the recorded data. The accuracy of this technique depends on whether the average potential (Vfave) truly represents an average potential unaffected by telluric current and on the proximity of the fixed location to the moving electrode because long separation distances can introduce errors caused by potential differences in the earth parallel to the pipe route and to telluric current voltage drop in the pipe. Place and Sneath[32] have used a variation of the foregoing technique in combination with CP current interruption to produce close interval survey data that is telluric-compensated. Their test arrangement (Figure 4-44) uses two stationary data loggers, one at the start of the CIPS (Vrs) and one at the end of the survey span (Vrf). 32 Place, T., and Sneath, O., Practical Telluric Compensation for Pipelines, Proceedings, NACE Northern Area Western Conference, Saskatoon, Feb. 2000. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:46 Vrf Vm y Vrs x Figure 4-44: CIPS Method using One Moving and Two Stationary Data Loggers All data loggers are synchronized by referencing the global positioning system (GPS). The telluric compensation is a linear extrapolation of the telluric shift at each data logger relative to the moving reference’s proximity to each stationary reference. This correction routine, done in software, is expressed as follows; Vps = Vm ± ΔVrs • Where: y x ± ΔVrf • x+y x+y [4-26] the ΔVrs and ΔVrf are the differences in potential compared to the average potential [Vrfave and Vrsave] recorded at each location over a period of time prior to the survey. This technique tends to minimize the error inherent in the previous method when the distance between the moving reference and the single stationary data logger increases significantly. Both techniques assume that the telluric voltage amplitude is linear over the relatively short distances surveyed and that pipeline voltage drop error created by the telluric current in the pipe between the start and finish test stations is negligible. Also, each method is dependent on the validity of the prerecorded data that establish the average potential with time at the start and finish test stations. The shorter this period is prior to the survey, the greater will be the influence of short duration telluric activity and the less will be the effect of any diurnal telluric activity. Figure 4-45 compares the typical before and after correction pipe-to-soil potential data. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:47 Figure 4-x22: Comparison of Raw Pipe-to-Soil Potential Data to Compensated Data Figure 4-45: Pipe-to-Soil Potential Measurement Method to Compensate for Telluric Current Effects During a Close Interval CP Survey Note that this compensation technique did not remove all the telluric voltage fluctuations because of its limitations. Degerstedt, et al[33] have used a “telluric null” technique for surveys on the Trans Alaska Pipeline System, which overcomes some of the limitations of the foregoing survey methods. They recorded the potential and current parameters at a test station with time to produce a fundamental characteristic for each test location, as illustrated in Figure 4-46. 33 Degerstedt, Ross, M., Kennelley, K.J., Lara, P.F., Moghissi, O.C., Acquiring Telluric-Nulled Pipe-toSoil Potentials on the Trans Alaska Pipeline, Corrosion ’95, Paper No. 345, NACE International, Houston, TX, 1995. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:48 +0.8 +0.6 I telluric (A) (downstream) Telluric Voltage Correction Factor +0.4 +0.2 I telluric (A) (upstream) -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 Telluric ‘Null’ Potential -1.6 Pipe Potential (VCSE) Figure 4-46: Pipe Potential/Telluric Current Relationship at a Coupon Test Station The telluric current was measured using magnetometers placed on grade on each side of the pipeline. It can be seen that there is a linear relationship between the telluric current and the pipe potential and that, through regression analysis, the “telluric null” potential is identified as the intercept with the pipe potential axis. With a historical characteristic established at each test station, the CIS is conducted using GPS time stamping to record both pipe current magnitude and potential with respect to a moving reference. This potential is corrected relative to the voltage at the fixed electrodes at the adjacent test stations by an appropriate correction factor. In lieu of magnetometers, the pipe current can also be determined by measuring the voltage drop along the pipe as illustrated in Figure 4-47, although this arrangement would require installation of pipe test leads at each test station location. Where telluric current activity is anticipated, the four wire test arrangement should be installed at each test station location so that the telluric null method can be utilized. In addition, each test station should also incorporate CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:49 a coupon/reference probe to facilitate the recording of pipe-to-soil polarized potentials with time. Ecal V Ip Ical Rp/l L where: Ip = Icp ± It Figure 4-47: Four Wire Test Lead Arrangement for Measuring Pipe Current 4.5 Summary In order to maintain effective corrosion control on relatively long coated pipelines that have high leakage resistance and that are located in latitudes close to the magnetic poles and therefore subjected to telluric currents, the following measures should be taken: • Maintain good electrical continuity throughout the system. • Integrate mitigation facilities with the CP system to reduce the magnitude of the telluric voltage fluctuations in both the positive and negative directions. • Install test station facilities incorporating coupons that can be used to measure “telluric free” pipe-to-soil potentials. • Install four wire test station facilities so that the pipe current can be recorded with time. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:50 • Use data loggers that are time synchronized and apply a correction factor to obtain accurate close interval pipe-to-soil data. CP Interference © NACE International, 2006 January 2008 Telluric Current Interference 4:51 Summary of Equations [4-1a] γ = [4-1b] Z0 = [4-2] dV dx page 4:6 Z Y page 4:6 = E − 1Z page 4:6 = − VY page 4:6 dI dx [4-3] ZY [4-4] d2V − γ 2V = 2 dx [4-5] d 2I dx 2 I = [4-6] E γE 0 V = [4-7] I = [4-8] page 4:7 − γ 2 I = − YE ( 1 + Ae E γ dE dx ( Ae - γ ( x − x1 ) -γ ( x − x1 ) page 4:7 + Be - γ ( x2 − x1 ) − Be -γ ( x2 − x1 ) ) ) V V E − 1 e - γx − 2 e - γ ( L − x ) Z Z0 Z0 V(x) = − V1e - γx + V2 e - γ ( L − x ) [4-9] page 4:7 page 4:7 page 4:7 page 4:7 where: [4-10] V1 = Z1 E × γ Z 0 + Z1 CP Interference © NACE International, 2006 January 2008 and V2 = Z2 E × γ Z0 + Z2 page 4:7 Telluric Current Interference 4:52 E = VBZW [4-11] where: Ε v BZ W = = = = the potential difference the water velocity the vertical component of the magnetic field the width of the water channel Fe° ⇒ Fe++ + 2e- (corrosion) [4-12] H 3 O + + e- [4-13] page 4:11 page 4:18 ⇒ H2 + OH- (in deaerated or acidic soils) page 4:19 ⇒ 4OH- (in alkaline or neutral aerated soils) page 4:19 or [4-14] 2H2O + O2 + 4e- [4-15] 4OH- ⇒ 2H2O + O2↑ + 4e- page 4:21 [4-16] 2 H2O ⇒ O2↑ + 4H+ + 4e- page 4:21 [4-17] C = ( 4.7 ± 1.3) T+0.186 page 4:23 Vm = Ep + Ve page 4:29 [4-18] where: Ve = Icp • Re Vm = Ecp + Ve ± Vt [4-19] Io = Icp + It [4-20] page 4:30 page 4:32 [4-21] Vo = VTR + Vt page 4:32 [4-22] Icp = Io – It page 4:33 g anode = [4-23] [4-24] where: [4-25] CP Interference © NACE International, 2006 January 2008 1 R anode = 2πL 1 × 8L ρ -1 ln d page 4:37 Vps = Vr ± ΔVf page 4:44 ΔVf = Vfave ± Vf page 4:44 Telluric Current Interference [4-26] Vps = Vm ± ΔVrs • 4:53 y x ± ΔVrf • x+y x+y page 4:46 Where: the ΔVrs and ΔVrf are the differences in potential compared to the average potential [Vrfave and Vrsave] recorded at each location over a period of time prior to the survey. CP Interference © NACE International, 2006 January 2008