LIVE-LINE WORKING AND. EVALUATION OF RISK ON 400kV TRANSMISSION LINE

Transcription

1 LIVE-LINE WORKING AND EVALUATION OF RISK ON 400kV TRANSMISSION LINE A THESIS SUBMITTED TO THE UNIVERSITY OF MANCHESTER FOR THE DEGREE OF DOCTOR OF PHILOSOPHY FACULTY OF SCIENCES AND ENGINEERING 2017 Pietro Martini School of Electrical and Electronic Engineering

2 2 Table of Contents ABSTRACT...19 DECLERATION...20 COPY RIGHT STATEMENT...21 AKNOWLEDGMENT.22 TERMS, DEFINATION 23 CHAPTER Introduction Introduction to Live-line Working Live-line Working Tools and Methods Hot stick Bare Hand (Potential Method) Helicopter Techniques Ground-Based Robots Live-line Working Risk and Challenges Minimum Approach Distance (MAD) Objectives and Conclusion CHAPTER IEC 61472, Live-line Working Safety Standards IEEE Standard IEC Description of Calculation Procedure... 41

3 Correction Factors Impact of Correction Factors on MAD At Tower At Mid-Span Discussion of Standards CHAPTER Introduction Travelling Waves Wave Velocity on Overhead Lines Wave Velocity on Cables Wave Reflection and Line Characteristics Impedance Transient Classification Lightning Overvoltage Review of Main Sources of Switching Overvoltages Line Energisation, re-energisation and Disconnection: Switching Impulse Strength Effect of Wave shape The U-Curve Wave Polarity Effect of Atmospheric Conditions Discussion and Conclusion CHAPTER

4 Introduction Simulation Methodology PSCAD Goodness-of-Fit Testing for Weibull Distribution Parameters Influencing the Overvoltage on Transmission Line Transmission Line Effect Type and Length of cable Section Cable Section Position on transmission Line Capacitor bank Network for Overvoltage Studies Overvoltage Simulation Results Calculation of Minimum Approach Distance Influence of Atmospheric Conditions Influence of Floating object on Minimum approach distance Discussion CHAPTER Introduction Live-line Working Risk Evaluation Risk Assessment Methodology for Risk Assessment (Standard Switching Transient) Stress on the gap Strength of the gap Intersection area

5 Methodology for Risk Assessment (Non-standard Switching Transient) Evaluation of Risk Based on Simulation Results Discussion: CHAPTER Conclusion Impact of different Parameters on Minimum Approach Distance Future Work References Appendices Word Count: 43474

6 6 Table of Figures Figure 1-1: Live-Line Work Using Hot sticks, A: Fibre Glass Ladder, B: Hot Stick, C: Bare Hand [1.12] Figure 1-2: Live-Line Work Bare Hand or Potential Method, Where the Linesmen Are at Same Potential as the Live Part and Isolated From the Earth [1.13] Figure 1-3: Live Men on 400kV Using the Live-lines Helicopter Method (Pictures Provided by National Grid) Figure 1-4: Single Pick Robotic Arm which captures the Energized Conductor above the H-Frame Structure [ ] Figure 1-5: Typical Live-Line Working Task [IEC 624/13] Figure 2-1: Flow Chart Illustrating the Calculation Procedure for the Minimum Approach Distance Figure 2-2: Flow Chart Illustrating the Calculation Procedure for the Minimum Approach Distance Figure 2-3: Electrical Distance for m altitude at L6 tower, With and Without Floating Object Figure 2-4: Electrical Distance for m Altitude at Mid-span L6 Tower, With and Without Floating Object Figure 3-1: Pi-section Presentation of Overhead Line and Cable Figure 3-2: Small Section of Transmission Line Figure 3-3: Simple PSCAD Power System Model Figure 3-4: Surge travelling time: Top: E_sending; The Voltage at the Sending and Bottom: E_receiving; The Voltage at the Receiving End of the Line... 65

7 7 Figure 3-5: National Grid direct buried cable diagram Figure 3-6: Impulse Generator Used in PSCAD Figure 3-7: Voltage at Sending Point (Blue Curve) Due to Current Impulse where Ea and Eb are the sending and receiving voltages respectively Figure 3-8: PSCAD Simulation Travelling Wave; Top: Voltage at Beginning, Bottom: Voltage at the End of Transmission Line Figure 3-9: Behaviour of Voltage Travelling Wave at Transition Point Figure 3-10: Sum of reflected voltage and current and sending waves Figure 3-11: Voltage at The Sending and Receiving End Due to Energisation of 60km Line on 400kV System Figure 3-12: PSCAD Simulation Model of Trapped Charge Figure 3-13: Energising of a Line, Top; Without Trapped Charge, Bottom; With Trapped Charge Figure 3-14: Voltage Due to Top; Energisation, Middle; Re-energisation, Bottom; Disconnection Figure 3-15: Oscillatory Transient Due to Interruption of Fault Current on PSCAD Model- ES: Voltage Sending Point, EL: Voltage along the Line, Earc: Circuit Breaker Arc Voltage Figure 3-16: Standards Switching Impulse Where V 50 is a half the time to crest of Crest of a Transient Wave [4.1] Figure 3-17: U-Curves Obtained with Impulse Voltages of Various Time-to-Crests (T cr µs) Applied to Rod-Plane Gaps. Atmospheric Humidity in These Experiments Was Varied [3.22, 3.23] Figure 3-18: A; Switching Impulse Flashover Voltage of Rod-Plane Gap, B; Estimation of CRIEPI s Equation... 86

8 8 Figure 3-19: Rod-Plane Gap; 1- Minute Critical Withstand AC and DC Voltages; 50% Percent Spark Over Voltage with Standard and Long Front Impulses [3.26] Figure 4-1: Model of Event Occurrence in Simulation Figure 4-2: Switching Overvoltage Distribution (pu) Figure 4-3: Overvoltage Weibull Distribution Plot Figure 4-4: Sample PSCAD Model of Transmission Line Figure 4-5: Overhead Model Figure 4-6: PSCAD Model of Line-Cable Combination Figure 4-7: Change of Overvoltage at Beginning and End of Cable Section Due to Changing the Length Figure 4-8: Overvoltage at Beginning and End of Cable Section vs. Cable Type Figure 4-9: Schematic Model of Transmission Line Figure 4-10: Time Required for Wave to Travel along the Cable Figure 4-11: Overvoltage, Sending (Blue Curve) And Receiving (Green Curve) With Cable Section at Beginning of the Line Figure 4-12: Schematic Model of Transmission Line with Cable Section Place in the Middle of the Line Figure 4-13: Overvoltage, Sending (Blue Curve) and Receiving (Green Curve) With Cable Section at Middle of the Line Figure 4-14: Maximum Overvoltage, Sending (Blue Curve) and Receiving (Green Curve) with Cable-Line at End of Transmission Line Figure 4-15: Series Capacitor Bank Modelling with a 41.91µF series Capacitor Figure 4-16: Overvoltage with 20% Series Compensation Figure 4-17: Overvoltage with 50% Series Compensation Figure 4-18: Overvoltage with 80% Series Compensation

9 9 Figure 4-19: A; PSCAD Model of Transmission Line, B; Schematic diagram of the network Figure 4-20: Top; P-E, Bottom; P-P. Influence of Length of Transmission Line on the Minimum Approach Distance Figure 4-21: Top; P-E, Bottom; P-P - Minimum Approach Distance Influenced by Altitude and Fault Levels Figure 5-1: Risk and Hazard Explanation [5.1] Figure 5-2: Risk Management Process Figure 5-3: Live-Line Working Risk Evaluation Process Figure 5-4: Switching Overvoltage Distribution Figure 5-5: Flowchart Illustrating the Steps Undertaken for Calculation of Gap Strength Figure 5-6: Air Gap Voltage Breakdown Probability Figure 5-7: Combination of Air Gap Voltage Breakdown Probability and Switching Overvoltage Distribution Figure 5-8: Risk as the Function of Time to Crest Figure 5-9: Risk of Failure as a Function of Time to Crest on Different Towers for Top: P-E and Bottom: P-P Voltage Figure 5-10: Risk of Failure for P-E Voltage as the Function of Changing the Gap Size, Bottom: The Zoom in Graph of the Top Graph Figure 7-1: Conductor Coordinates of Overhead Line- Refer to Table Figure 7-2: PSCAD Fault Type and Time Selection Modules Figure 7-3: PSCAD Overhead Line Model Figure 7-4: P-E Calculation Design Figure 7-5: P-P Calculation Modules

10 10 Figure 7-6: Rod to Plane Sparkover versus Gap Length D, CRIEPI_ Figure 5-2 [2.23] Figure 7-7: Switching Impulse Flashover Voltage of Rod-Plane Gap, Estimation of CRIEPI s Equation

11 11 List of Tables Table 2-1: Minimum Approach Distances D A for Several Countries [2.2] Table 2-2: Comparisons of Minimum Approach Distance, IEC Correction Factors: Gap Factor=1.2, Altitude Factor=0.94, Insulation Factor=0.95 and Floating Factor = Table 2-3: Gap factors for some actual phase to earth configurations [2.6]. The gap factor (k g ) in Table 2-3 is presented by "k" Table 2-4: Values of Exponents, m of Air Density Correction and w For Humidity Correction as the Function of Parameter g (IEC 60060) - [2.7] Table 2-5: Average k a Value [2.8] Table 2-6: Set of P-E and P-P Overvoltages Table 2-7: Effect of Humidity of the Minimum Approach Distances at a temperature of 20 o C and a pressure of 101.3kPA Table 2-8: Comparison of the Calculation Results for The Minimum Clearances Based on IEEE and IEC Method [2.21] Table 3-1: Surge Impedance and Propagation Constant for Normal and Lossless Line [3.3] Table 3-2: Generator Parameters Table 3-3: Overhead Line and Circuit Breakers Parameters Table 3-4: Sample Cable Data for 400kV Single Core Cable, 1200mm2 ABB XLPE Cable [3.6] Table 3-5 CIGRE Classification of Overvoltage Based on Frequency [2.6] Table 3-6 IEC Classification of Overvoltage Based on Time Duration [3.8] Table 3-7: Shapes and Classes of Overvoltages Standards Voltage [3.29] Table 3-8: U 50 of Rod-Plane for Fast and Slow Wave Shape [3.21]... 83

12 12 Table 3-9: U 50 of Rod-Plane as the Function of Wave Shape, Non-Standard Switching Wave Form [3.21] Table 3-10: Effect of Polarity on Rod-Plane Gap [3.15], [3.19] Table 4-1: U 2 Value Comparison Achieved by PSCAD and Excel Table 4-2: Simulation Result of MATLAB Output File Table 4-3: Magnitude of Switching Overvoltage Due to Various Length of Transmission Line Table 4-4: Three Types of Cable Specification Used by National Grid Table 4-5: Cable and Overhead Line Specification Table 4-6: Series Capacitor Size Table 4-7: Overvoltage Results for Line Energisation Table 4-8: Overvoltage Results for Line Re-Energisation Table 4-9: Overvoltage Results for Line Dis-Connection Table 4-10: Overvoltage Results for Fault & Clearance Table 4-11: Overvoltage Results for Fault & Clearance Due to Simulation Setting Table 4-12: Overvoltage Results for Fault & Clearance Due to 80% LG Faults, 17% LL Faults,2% LLG Faults and 1% LLL Faults Table 4-13: Overvoltage Results Due to Fault & Clearances with Inductive Compensation Table 4-14: Overvoltage Results Due to Fault & Clearances with Capacitive Compensation Table 4-15: Overvoltage Results Due to 80% LG Faults, 17% LL Faults, 2% LLG Faults And 1% LLL Faults with Inductive Compensation Table 4-16: Overvoltage Results Due to 80% LG Faults, 17% LL Faults, 2% LLG Faults and 1% LLL Faults with Capacitive Compensation Table 4-17: Example Selection Table for ka

13 13 Table 4-18: Electrical Distances (Du) for Fault & Clearance Simulation Scenarios (With No Ergonomic Distance D A ) Table 4-19: Electrical Distances (Du) for Fault & Clearance Simulation Scenarios with Inductive Compensation (With No Ergonomic Distance D A ) Table 4-20: Electrical Distances (Du) for Fault & Clearance Simulation Scenarios with Capacitive Compensation (With No Ergonomic Distance D A ) Table 4-21: Electrical Distances (Du) for Fault & Clearance Simulation Scenarios with 80% LG Faults, 17% LL Faults, 2% LLG Faults And 1% LLL Fault Probability (With No Ergonomic Distance D A ) Table 4-22: Electrical Distances (Du) for Fault & Clearance Simulation Scenarios with Inductive Compensation With 80% LG Faults, 17% LL Faults, 2% LLG Faults And 1% LLL Fault Probability (With No Ergonomic Distance D A ) 123 Table 4-23: Electrical Distances (Du) for Fault & Clearance Simulation Scenarios with Capacitive Compensation With 80% LG Faults, 17% LL Faults, 2% LLG Faults And 1% LLL Fault Probability (With No Ergonomic Distance D A ) 123 Table 4-24: Influence of Altitude on Electrical Distances (Du) Due to Fault and Clearances (Without Compensation) - With No Ergonomic Distance D A. 124 Table 4-25: Influence of Altitude on Electrical Distances (Du) Due to Fault and Clearances (Inductive Compensation) - With No Ergonomic Distance D A 124 Table 4-26: Influence of Altitude on Electrical Distances (Du) Due to Fault and Clearances (Capacitive Compensation) - With No Ergonomic Distance D A Table 4-27: Electrical Distances for Fault & Clearance Simulation Scenarios at 500m Altitude With Floating Object With 2m Length in Direction of Phases (With No Ergonomic Distance D A )

14 14 Table 4-28: Electrical Distances for Fault & Clearance Simulation Scenarios with Inductive Compensation at 500m Altitude with Floating Object with 2m Length in Direction of Phases (With No Ergonomic Distance D A ) Table 4-29: Electrical Distances for Fault & Clearance Simulation Scenarios with Capacitive Compensation at 500m Altitude with Floating Object With 2m Length in Direction of Phases (With No Ergonomic Distance D A ) Table 5-1: Calculation Extracted from Simulation Results in Figure Table 5-2: Minimum Approach Distance s Risk of Failure Obtained from Probability of Air Gap Breakdown and Switching Overvoltage Distribution Table 5-3: Estimation of Risk Based on Transient Time-to-Crest Table 5-4: Calculated Risk for Fault & Clearance Simulation Scenarios Table 5-5: Calculated Risk for Fault & Clearance Simulation Scenarios with Inductive Compensation Table 5-6: Calculated Risk for Fault & Clearance Simulation Scenarios with Capacitive Compensation Table 5-7: Calculated Risk for Fault & Clearance Simulation Scenarios with 80% LG Faults, 17% LL Faults, 2% LLG Faults And 1% LLL Fault Probability Table 5-8: Calculated Risk for Fault & Clearance Simulation Scenarios with Inductive Compensation with 80% LG Faults, 17% LL Faults, 2% LLG Faults and 1% LLL Fault Probability Table 5-9: Calculated Risk for Fault & Clearance Simulation Scenarios with Capacitive Compensation with 80% LG Faults, 17% LL Faults, 2% LLG Faults and 1% LLL Fault Probability Table Rate of Change of the Risk Due to Change of Wave Time to Crest Table 7-1: Atmospheric Factor k a for Different Reference Altitudes and Values of U 90 _ (IEC 61472)

15 15 Table 7-2: Average k a Values IEC Table 7-3: Floating Conductive Object Factor k f Table 7-4: Conductor Coordinates (Including Sag) for Overhead Line Designs [2.1] Table 7-5: PSCAD Configuration of L2 Tower Table 7-6: PSCAD Configuration of L6 Tower Table 7-7: PSCAD Configuration of L8 Tower Table 7-8: PSCAD Configuration of L9 Tower Table 7-9: PSCAD Configuration of L12 Tower Table 7-10: Overvoltage Simulation Results for Fault and Clearance Table 7-11: Overvoltage Simulation Results for Fault and Clearance, Inductive Compensation Table 7-12: Overvoltage Simulation Results for Fault and Clearance, Capacitive Compensation Table 7-13: Minimum Approach Distance for Fault and Clearance at Sea Level Table 7-14: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at Sea Level Table 7-15: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at Sea Level Table 7-16: Minimum Approach Distance for Fault and Clearance at 500m Altitude. 175 Table 7-17: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at 500m Altitude Table 7-18: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 500m Altitude Table 7-19: Minimum Approach Distance for Fault and Clearance at 1000m Altitude 176 Table 7-20: Minimum Approach Distance for Fault and Clearance, Inductive compensation at 1000m altitude

16 16 Table 7-21: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 1000m Altitude Table 7-22: Overvoltage Simulation Results for Fault and Clearance & Weighted Fault Type Table 7-23: Overvoltage Simulation Results for Fault and Clearance, Inductive Compensation & Weighted Fault Type Table 7-24: Overvoltage Simulation Results for Fault and Clearance, Capacitive Compensation & Weighted Fault Type Table 7-25: Minimum Approach Distance for Fault and Clearance at Sea Level & Weighted Fault Type Table 7-26: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at Sea Level & Weighted Fault Type Table 7-27: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at Sea Level & Weighted Fault Type Table 7-28: Minimum Approach Distance for Fault and Clearance at 500m Altitude & Weighted Fault Type Table 7-29: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at 500m Altitude & Weighted Fault Type Table 7-30: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 500m Altitude & Weighted Fault Type Table 7-31: Minimum Approach Distance for Fault and Clearance at 1000m Altitude & Weighted Fault Type Table 7-32: Minimum Approach Distance for Fault and Clearance, Inductive compensation at 1000m altitude & Weighted Fault Type Table 7-33: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 1000m Altitude & Weighted Fault Type

17 17 Table 7-34: Minimum Approach Distance for Fault and Clearance at Sea Level with Floating Object of 2m Table 7-35: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at Sea Level with Floating Object of 2m Table 7-36: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at Sea Level with Floating Object of 2m Table 7-37: Minimum Approach Distance for Fault and Clearance at 500m Altitude with Floating Object of 2m Table 7-38: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at 500m Altitude with Floating Object of 2m Table 7-39: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 500m Altitude with floating object of 2m Table 7-40: Minimum Approach Distance for Fault and Clearance at 1000m Altitude with Floating Object of 2m Table 7-41: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at 1000m Altitude with Floating Object of 2m Table 7-42: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 1000m Altitude with Floating Object of 2m Table 7-43: Minimum Approach Distance for Fault and Clearance at Sea Level with Floating Object of 2m (Weighted Fault Type) Table 7-44: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at Sea Level With Floating Object Of 2m (Weighted Fault Type) Table 7-45: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at Sea Level with Floating Object of 2m (Weighted Fault Type)

18 18 Table 7-46: Minimum Approach Distance for Fault and Clearance at 500m Altitude with Floating Object of 2m (Weighted Fault Type) Table 7-47: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at 500m Altitude with Floating Object of 2m (Weighted Fault Type) Table 7-48: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 500m Altitude with Floating Object of 2m (Weighted Fault Type) Table 7-49: Minimum Approach Distance for Fault and Clearance at 1000m Altitude with Floating Object of 2m (Weighted Fault Type) Table 7-50: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at 1000m Altitude with Floating Object of 2m (Weighted Fault Type) Table 7-51: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 1000m Altitude with Floating Object of 2m (Weighted Fault Type) Table 7-52: Rod to Plane Gap Experimental Sparkover Data, Positive polarity (Continue), CRIEPI_ Table 5-1 [2.23] Table 7-53: 50% Flashover Voltage (kv) as the Function of Gap Size and Time to Crest Based on Table Table 7-54: Estimated formulae for Calculation of U 50 Voltage as the Function of Gap Size for Each Transient Time to Crest

19 19 ABSTRACT Power industries in transmission and distribution level are obligated to maintain and replace their electrical equipment. Maintaining the quality and continuity of supply is their priority to avoid customers' complaints and financial penalisation. Live-line working as one of the most important methods of maintenance has been used since the 1900s where the new methods in 1960s made the live-line workers enabled to work on the higher voltage levels up to 800kV. Various industries adopt different techniques to calculate the minimum approach distance (MAD) during the live-line work. A suitable method reduces the risk to live-line workers and provides adequate safety distances between the live parts and linesmen. Therefore, setting an appropriate safety distance between the linesmen and live parts ensures the safety of the workers and minimise the risk of flashover. In this thesis, different methods of calculation of the minimum approach distance are described, and results from overvoltage simulations are used as an input to the methodology outlined in IEC Also, this thesis highlights and investigates the impact of a range of factors within 400kV transmission line on the minimum approach distance (MAD). Factors examined include the time to crest of the overvoltage (wave shape), the fault type, the probability of occurrence of each type of fault, fault level and the type of overhead line and towers. Furthermore, the minimum approach distances and also associated risk due to each factor and scenario have been calculated. The calculated risk in this thesis presents the risk of failure of a gap against the switching overvoltages due to the simulation of sources of overvoltage. A new set of estimated equations is developed to consider the influence of wave shape in the calculation of the minimum approach distance (MAD). This thesis does not propose a method to replace the international standards, but it could be used in many situations including where utility companies wish to develop a complete understanding of the risk associated with live-line working. Calculation of the minimum approach distance (MAD) within the National Grid UK is based on the methodology described in the IEC 61472, whereas EDF Energy uses the IEEE method to calculate the minimum approach distance. The choice of a smaller / larger minimum approach distance (MAD) using different methods will have an impact on the risk associated with live-line working. Previous works intend to investigate the magnitude of switching overvoltages on one part of a network and calculate the appropriate minimum approach distance for the work in that section. This work is based on the examination of the switching overvoltages under the worst case scenarios. As a result, the simulated overvoltages in this work are higher than expected overvoltages in National Grid network. Also as in practice, the magnitude of switching overvoltages in National Grid network is controlled by different protections equipment therefore, the simulated results and the calculated minimum approach distances in this work are very conservative.

20 20 DECLARATION No portion of the work referred to in this thesis has been submitted in support of an application for another degree or qualification of this or any other university or other institute of learning.

21 21 COPYRIGHT STATEMENT I. The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the Copyright ) and s/he has given The University of Manchester certain rights to use such Copyright, including for administrative purposes. II. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made. III. The ownership of certain Copyright, patents, designs, trademarks and other intellectual property (the Intellectual Property ) and any reproductions of copyright works in the thesis, for example graphs and tables ( Reproductions ), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property and Reproduction scan not and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property and/or Reproductions. IV. Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property and/or Reproductions described in it may take place is available in the University IP Policy (see 0), in any relevant Thesis restriction declarations deposited in the University Library, The University Library s regulations (see and in The University s policy on Presentation of Theses.

22 22 ACKNOWLEDGMENTS First and foremost, I would like to express my sincere gratitude to my supervisor, Prof. Ian Cotton, for his continuous guidance and support throughout these years with his very positive attitude and openness. I appreciate his helpful comments and discussions which have contributed a lot to this achievement. A special acknowledgment goes to the Engineering and Physical Sciences Research Council (EPSRC) and National Grid UK who have sponsored this project. I would particularly like to thank all the members in National Grid and Mr. Chris Land, live Working Operation Engineer. I would also like to thank all my friends and colleagues in the Electrical Energy and Power Systems (EEPS) research department at the University of Manchester. Their good companionship and the excellent opportunities they provided to me to develop ideas and exchange knowledge have made this journey very enjoyable. I warmly thank and appreciate my parents (Nasser & Sima) continues help and support, to whom I wish to dedicate this Thesis to. Words cannot express how grateful I am to them for their love, patience, kindness and support. I owe them my life, career, future and who I am and I will be. Special thanks to my mom s Dampayees which pushed me to where I am standing now. Finally, I would like to thank my wife Mobina for being always next to me and unconditionally supporting me throughout all these years. Pietro Martini February 2017 Manchester

23 23 Terms, definitions and symbols For the purpose of this thesis, the following terms, definitions and symbols apply. Damaged insulator Insulator having any type of manufacturing defect or in-service deterioration which affects its insulating performance. Electrical distance (D U ) Distance in air required to prevent a disruptive discharge between energized parts or between energised parts and earthed parts during live working [IEC : ]. Ergonomic component of distance (D E ) Distance in air added to the electrical distance, to take into account inadvertent movement and errors in judgement of distances while performing work [IEC : ]. Fifty per cent disruptive discharge voltage (U 50 ) Peak value of an impulse test voltage having a fifty per cent probability of initiating a disruptive discharge each time the dielectric testing is performed [IEC :1987, ]. Highest voltage of a system (U s ) Highest value of operating voltage which occurs under normal operating conditions at any time and any point in the system (phase to phase voltage) [IEC :1985, ].

24 24 Minimum working distance (D A ) Minimum distance in air to be maintained between any part of the body of a worker, including any object (except tools appropriate for live working) being handled directly, and any part(s) at different electric potential(s). Minimum approach distance (MAD) The minimum approach distance is the sum of the electrical distance appropriate for the maximum nominal voltage and of the selected ergonomic distance [IEC : ]. Ninety per cent statistical impulse withstand voltage (U 90 ) Peak value of an impulse test voltage at which insulation exhibits, under specified conditions, a ninety per cent probability of withstand [IEC :1987, ]. Part Any element present in the work location, other than workers, live working tools and system insulation. Gap The gap refers to the space between the phases of overhead line or phase to ground/ tower s body where the live-line working takes place. Per unit value (pu) Expression of the per unit value of the amplitude of an overvoltage (or of a voltage).

25 25 Transient overvoltage Short duration overvoltage of few milliseconds or less, oscillatory or non-oscillatory, usually highly damped [IEC :1987, ]. Two per cent statistical overvoltage (U 2 ) Peak value of a transient overvoltage having a 2 per cent statistical probability of being exceeded. Work location Any site, place or area where a work activity is to be, is being, or has been carried out [IEC : ]. A d : Length of damaged insulator or number of damaged units in an insulator of length A o, not shunted by long arcing horn or grading ring A o: Length of undamaged insulator or number of undamaged insulator units not shunted by long arcing horn or grading ring β: Ratio of the total length in the direction of the gap axis of the floating conductive objects (s) to the original air gap length D: Length of the remaining air gap phase to earth D A : Minimum Approach Distance D E : Ergonomic distance D U : Electrical distance necessary to obtain U 90 D Lins : Minimum residual insulation length (d1, d2, d3, d4): Distances between the worker(s) and parts of the installation at different electric potentials F: Sum of all lengths, in the direction of the gap axis, of all floating conductive objects in the air gap (in metres) K s : Statistical safety factor

26 26 K t : Factor combining different considerations influencing the strength of the gap k a : Atmospheric factor k d : Coefficient characterizing the average state of the damaged insulators k f : Floating conductive object factor k g : gap factor k i : Damaged insulator factor k ic : Damaged composite insulator factor k is : Damaged insulator strings factor k s : Standard statistical deviation factor L f : Original air gap length P: Length of the remaining gap phase to phase r: Distance of a conductive object from the axis of the gap s e : Normalized value of the standard deviation of U 50 expressed in per cent U e2 : Two per cent statistical overvoltage between phase and earth U e90 : Ninety per cent statistical impulse withstand voltage phase to earth U p2 : Two per cent statistical overvoltage between two phases U p90 : Ninety per cent statistical impulse withstand between two phases u e2 : Per unit value of the two per cent statistical overvoltage phase to earth u p2 : Per unit value of the two per cent statistical overvoltage between two phases U s : Highest voltage of a system between two phases PE: Phase to earth voltage PP: Phase to phase voltage a.c.: Alternative current Dc: Direct current

27 CHAPTER 1. Introduction to Live Line Working 27 CHAPTER 1 Introduction to Live Line Working 1.1. Introduction Transmission and distribution companies spend millions of pounds to maintain and replace the power transmission lines, cables and equipment to ensure the continuity and reliability of power supply to their customers. Where the continuity and quality of power supply are necessary, disconnecting the power consumers from the power supply can be very costly. Quality and continuity of supplying electricity have always been a primary consideration for all transmission companies and this fact becomes indispensable where supplying the public and industry with power is concerned. To guarantee the quality and continuity of the service, continual and efficient plant maintenance without taking the plant out of service can be necessary. On the another hand, working on energised equipment reduces the need for a spare line and increases the utilisation and operational continuity of existing lines and also it has more financial and environmental advantages [1.1]. In the past, the maintenance on the transmission lines and substations required disconnection of some parts of the network. Therefore, live-line working began operation in 1975 with the purpose of the maintenance of electrical component and

28 CHAPTER 1. Introduction to Live Line Working 28 transmission lines operating at medium or high voltage, while the whole system is in service Introduction to Live-line Working In the early years of the 20th century, live-line working techniques were developed to prevent the power shortage and blackout at the time of plant maintenance. In the 1960s, some methods were tested in the laboratory to establish a safe process of working in much closer contact with high voltage lines. Nowadays, these methods are still in use by transmission companies where live-line working takes place. However, forward movement of technology provides the staff and their companies with new tools, techniques and equipment but, in general there are four primary methods of performing the live-line work as explained in the next sections [1.2] Live-line Working Tools and Methods The fibreglass ladder, hot sticks, conducting suit and insulating rubber gloves are some of the common equipment in live-line working whereas some methods, such as bare hand, aerial service or ground-based robots, are used to ensure that the live-line working is as safe as possible Hot stick Hot stick (live-line tool) was invented towards the end of the 20 th century. Linesmen could do some limited jobs such as screwing, moving poles and switching by standing at a safe distance from live equipment in the voltage range of 0.05kV- 800kV [1.12]. The invention of fibreglass has improved usage of the hot stick. Unlike the wooden stick, the

29 CHAPTER 1. Introduction to Live Line Working 29 fibreglass stick did not get any damp or moist, and this is one of the most important advantages of the fibreglass stick. Figure 1-1 shows the application of the hot stick during live-line working away from live parts. Unlike the hot stick, the linesmen are energised by working on a live line while they are standing on a fibreglass ladder. A B C Figure 1-1: Live-Line Work Using Hot sticks, A: Fibre Glass Ladder, B: Hot Stick, C: Bare Hand [1.12] At lower voltages up 36kV, linesmen wear insulating rubber gloves to be able to work in direct mechanical contact with live parts Bare Hand (Potential Method) The first procedures for bare hand working were initiated in In the bare hand (potential method) shown in Figure 1-2, the linesman body s potential needs to be raised to the same electrical potential (voltage) as the live equipment. Wearing the insulating rubber gloves enables the linesmen to work in direct contact with live parts. In this method, the linesman and the live part or line are at the same electrical potential, and they are isolated from the surrounding. However, there is only a small flow of current through the linesman s body [1.3]. Before establishing a contact between a linesman and a live part, the linesman s body needs to reach the same

30 CHAPTER 1. Introduction to Live Line Working 30 electrical potential as the live part. This process is initiated by using a conducting tool which gets hooked on the live part refer to Figure 1-2. Figure 1-2: Live-Line Work Bare Hand or Potential Method, Where the Linesmen Are at Same Potential as the Live Part and Isolated From the Earth [1.13] After finishing the live working task, the process needs to be reversed to disengage the linesman from the live part. The advantage of this method compared to the hot sticks method is that the linesman can do more varieties of tasks such as line splicing, vibration damper or conductors spacers replacement, etc., refer to Figure Helicopter Techniques Linesmen can also work on high voltage live-lines while they are isolated from the ground potential by standing in a basket which is attached via special insulated rope(s) to a helicopter or a crane. Figure 1-3: Live Men on 400kV Using the Live-lines Helicopter Method (Pictures Provided by National Grid) In this method, linesmen are wearing Faraday suits, overalls made from conducting fibres, conducting gloves and socks. The linesmen can work of a platform fixed to the

31 CHAPTER 1. Introduction to Live Line Working 31 side of the helicopter where, the body of the helicopter, linesmen and hanging basket are at the same potential as the transmission line while they are isolated from the earth - refer to Figure 1-3. Linesmen can be lowered by a line attached to the helicopter or crane while standing inside the basket. Similar to the bare hand method, as the linesman approaches the wire, an arc will form between the live part and linesman s body. The worker must immediately bond to the line to prevent further arcing. The linesman may use a conducting band during the approach to make the connection Ground-Based Robots In this method, a ground-based insulated long robotic arms can capture, remove or lift the heavy live conductors and parts. This method is used when there is difficulty in excitation of a project by use of other methods. These robots are remotely controlled from the ground by use of a radio controller device. This method is used up to 500kV [1.14] and [1.15], during some live-line projects, such as replacement of rotten poles, reconductoring of existing transmission lines, substation repairs including nuclear plants, and replacement and re-insulating existing structures. Figure 1-4: Single Pick Robotic Arm which captures the Energized Conductor above the H-Frame Structure [1.14] and [1.15].

32 CHAPTER 1. Introduction to Live Line Working Live-line Working Risk and Challenges In the UK, many electricity companies such as National Grid, Scottish Power, EDF Energy, etc., have applied live-line working as part of their maintenance scheme. Different methods and tools have been used to mitigate the considerable hazards involved with working on the live lines. Being in contact with a live-line, exposes the linesmen to high electromagnetic fields (EMFs). To prevent receiving the exposures above the relevant limits specified by Table 2 of ICNIRP guidelines [1.11], the linesmen usually wear the conducting suits which screen them against the electric field. Apart from live-line working tools and techniques explained in the previous sections, different organisation such as IEC, IEEE, OSHA, etc., developed a guideline for the calculation of the minimum approach distance (MAD). Each guideline introduces a method for the calculation of the minimum approach distance for phase to ground (P-E) or phase to phase (P-P) cause which must be maintained by the linesmen when they are exposed to energised parts. These guidelines are set to mitigate the risk involved in liveline working. Determination of maximum electrical stress due to transient overvoltages is the first step of calculation of the minimum approach distances. By knowing the magnitude of electrical stress at the work site during the live-line working, the minimum approach distance can be calculated. The magnitude of stress and also the strength of the gap are influenced by many factors such as system parameters, gap geometry, atmospheric conditions and altitude, the presence of the insulation in the air gap, surge wave shape and presence of tools or floating object in the air gap. Therefore, working standards are developed to minimise the risk of flashover in the air gap where live-line working takes place. EC/TC78 was initially created to standardise

33 CHAPTER 1. Introduction to Live Line Working 33 the tools and equipment used in the live-line working in North American and European countries [1.4]. Later, IEC as one of the TC78 projects provided the minimum approach distance required for live-line working. Recently National Grid as the UK s transmission company uses this method for the calculation of the minimum approach distances for the purpose of live-line working. This project initiated by the National Grid as part of their requirement to ensure the safety of their live-line staffs. Currently, there are not many pieces of research or investigations concerning the minimum approach distances, influencing parameters or risk involved with live-line working. Although these methods and equipment intend to minimise the risk facing the linesmen, yet working on live system involves risks which can have fatal results. To address the importance of the safety, some world- known associations such as IEEE, EPRI, CIER, CIGRE, LWA, etc., actively work on live working safety, but still according to UNIPEDE survey, there were 171 accidents and five fatalities due to live-line working [1.5]. Some academic studies highlighted the importance of influencing factors such as altitude, humidity or broken insulator on the flashover voltage of the air gap [1.6]- [1.8]. At the same time, some studies suggested a new insulation coordination approach to address the effect of wave shape on the voltage breakdown of the gap [1.9], [1.10]. It has been noticed by the author of this thesis that there is a missing link between the influencing parameters affecting the air gap flashover and risk involved with the minimum approach distance. Also, as it is shown further in Chapter 2 of this thesis, the inconsistency between the calculated minimum approach distances using different available methods features the very first objective of this project.

34 CHAPTER 1. Introduction to Live Line Working Minimum Approach Distance (MAD) As this project considers the Barehand and helicopter methods explained in section 1.4 of this thesis, it is necessary to examine the safety factors concerning the live-line workers. When a linesman climbs on a tower or uses a ladder to hang from any part of a tower, the minimum phase-earth (PE) or phase-phase (PP) safety clearance have been interrupted. This is because the linesman s body will provide a conductive shortcut between the phases or phase to ground. As a result, his/her body can conduct any possible flashover from closer phase to another phase or tower s body/ earth. Therefore, to prevent any flashover due to the presence of the linesman, a safe distance needs to be defined which can be referred to the minimum approach distance (MAD). According to IEC 61472, Ed 3.0, the minimum approach distance (MAD) is the minimum distance in air to be maintained between any part of the body of a worker, including any object (except tools appropriate for live working) being handled directly, and any part(s) at different electric potential(s) [2.8]. Therefore, to process the live-line work, a required withstand voltage and a minimum approach distance (MAD) need to be calculated. Based on IEC 61472, in the calculation of the minimum approach distance (MAD) between phase-phase or phase-ground, the presence of the hanging basket or the linesman s body on the transmission line can be considered to reduce any possible risk of flashover that can cause severe or fatal injuries. Figure 1-5 presents various live working tasks and MAD configurations in which can occur. Each individual scenario in Figure 1-5 shows the minimum approach distances involved with live line working, i.e. d 1 in A, d 1 and d 2 in B, d 1 +d 3 and d 3 in C and d 1, d 2, d 3 and d 4 in D are the minimum approach distances in each case. These tasks can also be done with the presence of a hanging basket from a helicopter where the

35 CHAPTER 1. Introduction to Live Line Working 35 minimum approach distance (MAD) is interrupted due to the presence of a floating conductive object between the phases or phase to earth. Figure 1-5: Typical Live-Line Working Task [IEC 624/13] In Figure 1-5, the minimum approach distance (MAD) varies based on the linesmen s position on the tower. In towers A and B, the minimum approach distance has to be bigger than d1. In Tower C, the minimum approach distance has to be bigger than d1+d3 and d2+d3 whereas, in Tower D, the minimum approach distance has to be larger than d1, d2, d3 or d4. 'd' is the minimum approach distance between the linesmen and live or tower structure Objectives and Conclusion The aims of the research described in this thesis were to investigate different available methodologies that can be used to determine the safety of live-line workers, all of which carry a range of assumptions for calculation of the minimum approach distance. This thesis considered 400kV transmission lines and different type of towers used in the UK s HV transmission network to investigate the factors influencing the minimum approach distances during the live-line working. Due to inconsistency in the calculated minimum approach distances using different methods and also due to the lack of

36 CHAPTER 1. Introduction to Live Line Working 36 understanding the risk involved with live-line working, the aims of these project are as follow; Review of existing insulation co-ordination methodology and the method used by National Grid, Review of the historical background behind the definition of existing safety clearances and the guidance of standardisation bodies such as IEC and IEEE in this area, To carry out a literature review to highlight the existing work or method, Investigate the parameters influencing the minimum approach distance and magnitude of switching overvoltages, such as tower and overhead lines type, the length of transmission lines, cable, fault level, the effect of fault type, time-tocrest, etc. Propose a fundamental model of power network and investigate the effect of different component of transmission line on the minimum approach distances, Propose a new set of minimum approach distances, Evaluate the risk involved with live-line working due to standard switching transients. Investigate the risk involved with live line working due to non-standard switching transients and propose a method for investigation of effect of wave shape on the risk.

37 CHAPTER 2. Analysis of International Standards 37 CHAPTER 2 Analysis of International Standards 2.1. IEC 61472, Live-line Working Safety Standards In 1975, live working standards was developed to address standardisation needs of North American and European countries in the field of live-line working [1.4]. TC78 standard committee was concerned with tools and equipment used for live working. IEC (The International Electro Technical Commission) is an organisation that sets, prepares and publishes international standards for all electrical, electronic and related technologies. After many years and drafts, IEC standard [2.8] was published, and method of calculation of the minimum approach distance was set. Since then, the method introduced by IEC has been used by transmission companies. IEC defines the calculation method of the minimum approach distance for live working for a voltage range between 1kV up to 800kV. National Grid applies the IEC TC78 method as a fundamental approach for calculation of safety clearances. Based on National Grid technical guidance note (TGN (T) 54) [2.1], it is assumed that the deployed method used by National Grid may provide slightly larger minimum approach distances (phase-to-earth) compared to the method used by EDF Energy. Based on their assumption, the IEC results are 10% greater than those produced using the ANSI/IEEE Standard 516, 1987 [2.1].

38 CHAPTER 2. Analysis of International Standards 38 In some countries, like Canada and within some other power companies, different minimum approach distances values are in use [2.2]. These approach distances are based on various experiences, empirical and analytical methods. The phase- to - ground minimum approach distances in live-line working, implemented in various countries are shown in Table 2-1. Unfortunately the actual test conditions and the method used within each individual country in Table 2-1 are not available. However, as shown further in this thesis, the weather conditions can have a large impact on the minimum approach distances. Table 2-1: Minimum Approach Distances D A for Several Countries [2.2] Voltage of Design Us (kv) Country D A (m) D A (m) US France Sweden China (dry air) Finland (low humidity) Voltage of Design Us (kv) D A (m) D A (m) Mexico As an example, Table 2-2 shows some calculation results to illustrate the difference between the IEC method and other standards accepted by the United States Department of Labour (OSHA). Two different switching transient magnitudes of 2.3pu and 3.5pu are used for these calculations. The altitude used by OSHA method is considered to be any altitude less than 900m, whereas 500m altitude is set for IEC method. The results in both Table 2-1 and Table 2-2 indicate a large difference in results calculated by different standards and countries. Unfortunately, the test condition and deployed methods are not available for further investigations.

39 CHAPTER 2. Analysis of International Standards 39 Table 2-2: Comparisons of Minimum Approach Distance, IEC Correction Factors: Gap Factor=1.2, Altitude Factor=0.94, Insulation Factor=0.95 and Floating Factor =0.85. System Voltage (kv) Maximum Transient pu Phase- Ground Phase- Phase Floating object included IEC Minimum Approach Distance (m) Phase- Phase- Ground Phase Accepted OSHA Minimum Approach Distance (m) Phase- Phase- Ground Phase Table 2-2 shows that the minimum approach distances deployed by IEC and OSHA [2.3] are different. As a result, there is a need for further investigation and reviewing of the calculated minimum approach distances by standards. Apart from IEC standard, IEEE Std [2.4] also provides a calculation method for the minimum approach distance which is also based on experimental results of U 50 (fifty percent disruptive discharge voltage) on a particular length of the air gap. In HV and EHV systems, the voltages that cause the highest risk of flashover are those associated with lightning and switching operations. These voltages determine the external insulation design under their large magnitudes. In the next sections, two different standards available for calculation of the minimum approach distance have been examined in more details IEEE Standard In 1987, after the publication of several papers regarding the safety aspects of the liveline maintenance, the IEEE Transmission and Distribution Committee published a full used ANSI/IEEE standard for the purpose of the live-line working. In 1990, the ESMOL Subcommittee (Engineering in Safety, Maintenance, and Operation of Lines) revised the standards to update the guidelines to conformance with other international

40 CHAPTER 2. Analysis of International Standards 40 standards. The flow charts below show the full methodology used to calculate the minimum approach distance based on the IEEE method. Step 1 Step 2 Step 3 Step 4 Step 5 Step 6 Step 7 Measuring system voltage based on simulation or experimental test Calculation of rms phase-ground voltage of the system, kv LG Calculation of saturation factor Calculation of MAD d= (C 1.C 2 +a).t. kvlg Calculation of truncation value Estimation of U 2 voltage Multiplication of MAD by altitude factor Figure 2-1: Flow Chart Illustrating the Calculation Procedure for the Minimum Approach Distance In this method, calculation of withstand voltage for selection of air gap was done based on 13 laboratories experiments over 30 years. According to the IEEE method, the minimum approach distance (d) depends on two factors; 1. rms phase to ground voltage of the system 2. The maximum per-unit switching overvoltage factor (T) The minimum approach distance (d) is calculated by use of Equation (2.1) which is designed to fit the experimental curve obtained from withstand voltage of different size air gaps [2.4]. d= (C 1.C 2 +a).t. kv LG (m) (2.1) Where: d: Insulation distance, (m); C 1 : 1% phase-ground system voltage (kv)

41 CHAPTER 2. Analysis of International Standards 41 C 2 : 1% for presence of no tools in air gap a: Saturation factor of the crest 2.. kv of voltage 630kV and above, this factor can be approximated to within 2% by use of Equation (2.2); kv LG : rms system Phase-ground voltage- actual. a= ( kv LG 4.75) (2.2) T: maximum per-unit switching overvoltage factor, i.e. truncation value of distribution of overvoltage which no other overvoltage occurs after that point IEC Description of Calculation Procedure IEC describes a method for calculation of the minimum approach distances at maximum voltages between 72.5 kv and 800 kv for the purpose of live-line working. The required withstand voltage and also the minimum approach distances described in the IEC standard are evaluated taking into consideration the followings: Workers are trained for, and skilled in working in the live working zone; The anticipated overvoltages do not exceed the value selected for the determination of the required minimum approach distance; Transient overvoltages are the determining overvoltages; Tool insulation has no continuous film of moisture or measurable contamination present on the surface; No lightning is seen or heard within 10 km of the work site; Allowance is made for the effect of conducting components of tools;

42 CHAPTER 2. Analysis of International Standards 42 The effect of altitude, insulators in the gap, etc., on the electric strength, is taken into consideration. The flow chart of Figure 2-2 shows the full methodology used to calculate the minimum approach distance based on IEC Step 1 Step 2 2% Statistical Overvoltage obtained from Network or Simulation (U e2 & U p2 ) Calculation of U 90 : Multiplying the U e2 & U p2 by Statistical Correction Factor Step 3 Consideration of correction factors (K t ) Step 4 Calculation of Electrical Clearance Step 5 Calculation of the Minimum Approach Distance (MAD) Figure 2-2: Flow Chart Illustrating the Calculation Procedure for the Minimum Approach Distance The statistical analysis assumes that switching overvoltages are distributed according to a given probability law, i.e. a normal distribution. In the IEC method, the U 2 (2% Statistical Switching Overvoltage) voltage is used for calculation of the minimum approach distance [2.5]. This value can be obtained by Monte Carlo procedure and usually performed by use of a digital computer. However, 2% statistical switching overvoltages in this project are obtained from simulation results which are explained in Chapter 4. The required withstand voltage for live-line working is taken to be equal to U 90 which refers to a ninety percent probability of withstand voltage [2.6].

43 CHAPTER 2. Analysis of International Standards 43 The U e2 and U P2 voltages (U 2 voltage for Phase-Earth and Phase-Phase respectively, expressed in kv) are extracted from simulation results and multiplied by K S which is a statistical safety factor. The electrical stress at the work place during the live-line working is described as the statistical overvoltage that may be presented at the work location. In a three-phase a.c. power system, the statistical overvoltage U e2 (phase to earth_ kv), U p2 (phase to phase_ kv), (phase to earth_ pu) and (phase to phase_ pu) are calculated by Equations (2.3) (2.5) extracted from IEC The U s is the system nominal voltages. = (kv) (2.3) = (kv) (2.4) U e90 = K S U e2 & U P90 = K S U p2 (kv) (2.5) As the minimum approach distance (MAD) consists of electrical and ergonomic distance, when no ergonomic distance is used, the value of 1.1 is recommended for K S to reduce the overall risk of breakdown of the insulation to a level that correlates with other electrical work operations. If the per unit phase to phase data are not available, an approximate value can be derived from u e2 by the Equation (2.6). = (2.6) In order to examine the validity of the modelling used in this work, Equation (2.6) was applied to a set of simulations. The results from calculations show a small difference less than 5% in all cases. As a result, the U 90 (ninety percent probability of withstand voltage) and the minimum electrical distance (D U ) can be calculated by Equations (2.7) and (2.8).

44 CHAPTER 2. Analysis of International Standards 44 U 90 =K S x U 2 (kv) (2.7) D = 2.17 e ( ) 1 + F (m) (2.8) As it will be explained later, the strength of the gap is influenced by different factors which can be combined in a correction factor (K t ) to produce Equation (2.9). Equation (2.9) presents the K t which is combination of various factors influencing the strength of the air gap. K t =k S.k a.k i.k g.k f (2.9) Also, the factor F in Equation (2.8) is the sum of all lengths of any floating conductive objects (in meters) in the direction of the air gap axis. Moreover, the minimum approach distance (D A ) is calculated by adding the electrical distance D U and the ergonomic distance D E. These distances are further defined in IEC as: D U Distance in air required to prevent a disruptive discharge between energised parts or between energised parts and earthed parts during live-line working. D E Distance in air to take into account inadvertent movement and errors in judgement of distances while performing work. Therefore, the minimum approach distance is given by: D A = D U +D E (m) (2.10) Correction Factors Standard Statistical Deviation Factor k S k S factor is the statistical nature of the breakdown voltage, and its value is calculated by IEC based on the relationship between the statistical withstands voltage, U 90 and the 50% disruptive discharge voltage, U 50 as below; U 90 = U s e.u 50 (kv) (2.11)

45 CHAPTER 2. Analysis of International Standards 45 Where 'Se' is the normalised value of the standard deviation of U 50 expressed in percent. Therefore, the k S can be defined by Equation (2.12). k S = s e (2.12) Unless the value of S e is known from the tests representative of the gap configuration and distance concerned, a value of S e = 5 % should be assumed. Equation (2.13) then becomes: k S = (2.13) Gap Factor k g The gap factor (k g ) accounts for the varying electric field distribution in the gaps of varying shapes. The gap factor (k g ) depends on the gap configuration. The gap factor is used to adjust the strength of a gap of a specific geometry to the reference rod-plane case. The fundamental gap factor equal to one is calculated for the rod-plan gap, whereas, typical gap factor values for standard configurations of gap factor and other parameters are shown in Table 2-3 which is reproduced from CIGRÉ 72 and also presented by IEC The gap factor k g in Table 2-3 is presented as "k" and it is permitting the calculation for different gap configurations. Table 2-3: Gap factors for some actual phase to earth configurations [2.6]. The gap factor (k g ) in Table 2-3 is presented by "k". Configuration Formula Typical value k= 1.45

46 CHAPTER 2. Analysis of International Standards 46 k=1.25 k= 1.15 for conductor plane to 1.5 or more k=1.45 k1=1.3 k2= H Equation (2.14) extracted from Table 2-3, and it is used to calculate k g for National Grid s towers. Where; k g H d = S e d Łd1 ł Ł ł Ł d1 ł (2.14) H = Height of overhead line conductor from the ground (m) d 1 = Distance from the conductor up to the point where it is connected to the cross-arm (m) d 2 = Horizontal distance between the conductor and the tower structure (m) S = Thickness of the tower along the distance d 2 (m)

47 CHAPTER 2. Analysis of International Standards 47 It should be noted that this equation is valid only for the conditions where: 2m d 1 10m 1 d 2 / d S/d H/d 1 1 This project uses the National Grid towers and transmissions lines specifications for the purpose of overvoltage studies. For different type of towers used by National Grid, the lowest values of k g were chosen to be used in the calculations of the minimum approach distances as these values will give the most conservative results. Therefore, to comply with National Grid calculations, the k g values used in this thesis are taken to be equal to 1.2 and 1.45 for the phase to earth and phase to phase respectively as stated in TGN (T) 54 Technical Guidance Note [2.1]. Atmospheric Factor k a In the calculation of 50% voltage breakdownof the gap, the atmospheric factor takes into account the effect of air density influenced by temperature, humidity and altitude. The effect of temperature and humidity is negligible in comparison with altitude. For an instant, U 50 decreases at a location higher than the reference altitude [2.23] and, as a result the required distance increases and this can be determined by multiplying the electrical distance by an altitude correction factor. The atmospheric factor can be calculated as below: k a =K 1 K 2 (2.15)

48 CHAPTER 2. Analysis of International Standards 48 Air Density Correction Factor, K 1 The air density correction factor k 1 depends on the relative air density δ where temperatures t and t 0 are expressed in degrees Celsius, and the atmospheric pressure P and P 0 are expressed in the same units. K 1 = δ m (2.16) δ = p p t t (2.17) Calculation of Exponents m and w: To calculate the exponents m and w, the following formula is being used and values of m and w are extracted from Table 2-4 as below: g = U 500 L δ k (2.18) Where L is the minimum discharge path in meter, δ is the relative air density, and k is the dimension less parameter defined by Equation (2.19). Table 2-4: Values of Exponents, m of Air Density Correction and w For Humidity Correction as the Function of Parameter g (IEC 60060) - [2.7] ǥ m w < to 1.0 ǥ (ǥ 0.2) /0.8 ǥ (ǥ 0.2) / to to (2.2 ǥ ) (2.0 ǥ)/ Humidity Correction Factor, K 2: The effect of humidity on the voltage breakdownis more complex compared with previous cases. It is usually accounted for using a factor defined as k in IEC , the value of which is empirically related to humidity, and an exponent w, which

49 CHAPTER 2. Analysis of International Standards 49 depends on the gap length and its configuration and the wave shape. Thus, the humidity correction factor K 2 is: K 2 =k w (2.19) The value of w factor, as well as the exponent m for the relative air density can be determined by the methods given in IEC and explained in more details in next section. In Equation (2.19), k is a parameter that depends on the type of test voltage and it may be obtained as a function of the ratio of absolute humidity (h) to the relative air density (δ) and can be calculated using Equation (2.20) [2.7]. (2.20) The appropriate value (average) of k a can be selected from the Table 2-5 for the average value of k a or can be calculated for specific altitudes according to the calculation method explained. Table 2-5: Average k a Value [2.8] Altitude (m) k a average Table 2-7 presents the effect of humidity on a set of P-E and P-P overvoltages respectively which are shown in Table 2-6. The calculation of the minimum approach

50 CHAPTER 2. Analysis of International Standards 50 distances is done using the IEC method refer to section 2.3. During this examination, the pressure and temperature remain at 20 o C and 101.3kPA respectively. Increasing the absolute humidity decreases the minimum approach distances due to the increase of the voltage breakdown of the gap. The examination results shown in Table 2-7 comply with IEC standard section [2.7]. Table 2-6: Set of P-E and P-P Overvoltages Overvoltage Samples (kv) P-E P-E Min Overvoltage (kv) Max Overvoltage (kv) Mean Overvoltage (kv) U 2 Overvoltage (kv) Table 2-7: Effect of Humidity of the Minimum Approach Distances at a temperature of 20 o C and a pressure of 101.3kPA Minimum Approach Distance (m) P-E P-P Relative Humidity 5% Relative Humidity 10% Relative Humidity 15% Relative Humidity 20% Relative Humidity 40% Relative Humidity 60% Relative Humidity 80% Relative Humidity 90% Relative Humidity 100% Floating Object Factor k f The k f takes into account the presence of floating objects within the gap. In the absence of a floating object in the air gap, the value of k f will be equal to 1; otherwise, k f must be calculated. Based on IEC 61472, for long or flat shaped conductive objects situated perpendicular to the air gap or where no specific experimental data is available, a conservative value of k f equal to 0.75 may be assumed. The value of k f can be selected from Tables provided in Appendix 1. However, in this project, the value of k f

51 CHAPTER 2. Analysis of International Standards 51 was taken to be equal to 0.85 to match the value used by National Grid s assumption in [2.1]. Damaged Insulator Factor k i A damage insulation can have a significant impact on the withstand voltage of an air gap at a live-line work location. As a result, the number and location of the damaged units and also the degree of their damage can have a significant effect on the strength of the gap and as a result, on the minimum approach distance during the live-line working. The strength of an air gap can be reduced significantly in the case of glass insulators as the glass insulators are made of pre-stressed toughened glass, and they always shatter completely in the event of any incident. Regarding composite insulators, the strength reduction is significantly larger with conductive or semi-conductive defects. The strength of composite insulators becomes null when a conductive damage involves the whole insulation length. k i is the insulation string factor concerning the insulators damage and allowing for the system or tool in the gap. k is for cap or pin insulators can be calculated based on IEC using Equation (2.21). Where; k is =1-0.8 k d (A d /A o ) (2.201) A d is the number of damaged insulator units in the string; A o is the number of insulator units in the string; k d is assumed 1.0 for glass and 0.75 for porcelain; k is is the damaged insulation string factor. k i also can be calculated for composite insulators. The insulation factor for composite insulator will be used for consideration of damaged insulators and allowing for the system or tool insulation in the gap. k ic for composite insulators can be calculated based on IEC using Equation (2.22).

52 CHAPTER 2. Analysis of International Standards 52 Where; k ic =1- (l d /l o ) (2.22) l d is the damaged length along insulator axial direction; l o is the insulating length of the insulator; k ic is the damaged composite insulator factor. There have been many suggestions to detect the faulty composite insulators on HV power lines i.e. the electric field measurement [2.9], [2.10] which can be beneficial for calculation of the gap strength for the purpose of live-line working and method used for the insulator replacement task. In this thesis, it has been assumed that k i is equal to 0.95, which is a recommended value in IEC In summary, throughout this thesis correction factors with values of 1.2, 1.45, 0.95 and 0.85 have been used for P-E gap factor, P-P gap factor, insulation factor and floating factor respectively Impact of Correction Factors on MAD Unlike the IEEE method, the method introduced by IEC is very simple to apply in different circumstances by changing some correction factors. Below, the effects of atmospheric conditions on the minimum approach distances are presented for both at tower and at mid-span of an overhead line. Two overvoltages with a magnitude equal to 1.46 and 3.64 were assumed for this study where the altitude ranging from the reference altitude (sea level) to 1000m above the sea level was considered. The gap factor (k g ) was set to for the tower and 1.36 for mid-span.

53 CHAPTER 2. Analysis of International Standards At Tower The results from calculations based on Equation (2.8) show that increasing altitude increases the minimum approach distances. The calculation results show 10% differences in minimum electrical distances between sea level and 1000m for the overvoltage levels about 500kV (1.46pu), whereas, by increasing the magnitude of overvoltage, this difference reduced to 6% for the voltage levels around 1248 (3.64pu). Therefore, increasing the altitude has a smaller effect on a system with higher voltage. This statement also can be used for a gap with a floating object. The results of calculations can be found in Figure 2-3. Minimum Electrical Distance (m) Without Floating Object 1.46 pu 3.64 pu Altitude (m) Minimum Electrical Distance (m) With a Floating Object 1.46 pu 3.64 pu Altitude (m) Figure 2-3: Electrical Distance for m altitude at L6 tower, With and Without Floating Object

54 CHAPTER 2. Analysis of International Standards At Mid-Span The results from calculations based on Equation (2.8) show that increasing altitude increases the minimum approach distances. The calculation results show 9% differences in minimum electrical distances between sea level and 1000m for the overvoltage levels about 500kV (1.46pu), whereas, by increasing the magnitude of overvoltage, this difference reduced to 6% for the voltage levels around 1248 (3.64pu). Therefore, increasing the altitude has a smaller effect on a system with higher voltage. This statement also can be used for a gap with a floating object. The results of calculations can be found in Figure 2-4. Minimum Electrical Distance (m) Without Floating Object 1.46 pu 3.64 pu Altitude (m) 4.5 With a Floating Object Minimum Electrical Distance (m) pu 3.64 pu Altitude (m) Figure 2-4: Electrical Distance for m Altitude at Mid-span L6 Tower, With and Without Floating Object

55 CHAPTER 2. Analysis of International Standards 55 The dielectric strength of a gap is also proportional with the air density in gaps less than 2m. However, in a gap with larger distance, the air breakdown is less proportional with air density [2.6]. This means that air density has a small effect on the strength of a gap, and as a result, there is a negligible effect on the minimum approach distance at a live-line working location where the air gap is limited to the gap sizes larger than 2m. However, pressure is the main influencing factor of atmospheric condition on the flashover voltage of a gap during the live-line working [2.2]. Decreasing the pressure due to increasing altitude reduces the voltage breakdown of a gap, and as a result, a smaller magnitude of switching overvoltage is required to cause a flashover within the gap. Therefore, the minimum approach distance will increase as a result of increasing the altitude or decreasing the pressure. As shown in Figure 2-4, the magnitude of switching overvoltages has a greater influence than atmospheric factors on the minimum approach distance of the towers with a line spacing larger than 2m. In the calculation of minimum approach distance, it is important to consider that most of the UK s lands with low plains and downs with the major hill regions situated in the north (mostly Scotland and Wales), and some places in the west and south-east of the country. The elevations of these lands do not rise above 305 metres (1,000 feet) at any point.

56 CHAPTER 2. Analysis of International Standards Discussion of Standards Table 2-6 presents some example calculations of the electrical distances based on both IEC and IEEE methods. These values are obtained based on the assumption of existence of no broken insulators and no floating object when K s =1.1, k s =0.936, k g =1.2, k i =1.0, k f =1 and F=0. The K t value calculated based on Equation (2.8), and it is equal to In Table 2-6, the altitude assumed to be 900m to adjust the results from IEC method with equivalent IEEE method. Table 2-8: Comparison of the Calculation Results for the Minimum Clearances Based on IEEE and IEC Method [2.21] U S (kv) kv LG T a u U 2 (kv) U 90 (kv) IEC D U (m) IEEE D(m) The correction factor for altitude in IEEE method does not take into account the altitude below 900m whereas, in the IEC method, the effects of different parameters such as altitude, weather conditions (temperature, humidity and pressure) and also the effects of a broken insulator and floating objects have been considered. Based on Table 2-6, the IEEE provides a method that recommends a smaller electrical distance in comparison with IEC method, but both approaches agree on the higher values of U 2 overvoltage. In the IEEE method, the factor T is interpreted as the maximum anticipated overvoltage (truncation value of the overvoltage which no other overvoltage occurs after that point) is different from U 2 values used by IEC method. Neither of the two methods contains the exact nature of the tested gaps and test conditions, but considering the safety matter, using the IEC method provides a larger

57 CHAPTER 2. Analysis of International Standards 57 distance as the IEC standard takes into account different correction factors. However, the IEEE method could be adequate. In the calculation method deployed by IEEE, details of the exact nature of the tested gaps and test conditions have been lost, and more work needs to be done on altitude correction factors. Also, the effect of floating objects within the phases has been ignored, whereas live-line working can be carried out by use of a hot-stick or a basket hanging from a helicopter. Both methods have considered the system maximum operational voltage whereas in reality, this might not be possible. The electrical system normally operates at a voltage that the system components are designed. This voltage usually is 5 to 10 percent below the maximum system voltage [2.22]. Neither of the two methods, consider the live-line working duration and the probability of occurrence of the maximum overvoltage at liveline working location as the location of the live-line working site might not coincide with the maximum overvoltage due to the switching. At the same time, half of switching overvoltages are not severe as they might have negative polarity. Previous experimental results proved that the sparkover strength of an air gap and, as a result, the minimum safety clearances are varied according to the wavefront (time to crest) of transient switching overvoltage [2.11]-[2.13] and transient wave shape [2.14] - [2.19]. As shown later in Chapters 4 and 5, different line length/ source inductance influence the time to crest of the transient wave and, hence the probability of flashover can be affected. Therefore, there is a missing link between the calculated minimum approach distance using the IEC and IEEE methods and the wave shape. Although the existing minimum approach distances set by the IEC standard are more conservative than the IEEE method, due to the importance of human safety factor, there is a need for further investigation on the competency of these clearances according to the network specifications.

58 CHAPTER 2. Analysis of International Standards 58 Unlike IEEE, in the calculation of the minimum approach distance developed by IEC, the effects of altitude, floating objects, weather conditions and broken insulators are taken into account. Therefore, as the IEEE method does not directly account for some details, and at the same time the IEC method is more general and flexible and provides larger and more conservative safety distances, IEC method is more applicable in a calculation of the minimum safety approach. Therefore, this research used the IEC method as it is also confirmed by the British Standards Institution (BSI) for calculation of the minimum approach distances. This project intended to use the following standards in its calculations; IEC 61472:2004, Live working Minimum approach distances for A.C systems in the voltage range 72.5 kv to 800 kv method of calculation. IEC Standard :2010, High-voltage test techniques, Part 1: General definitions and test requirements. IEC Standard :2006, Insulation co-ordination Part 1: Definitions, principles and rules. IEC/TC78 Live Working : Background, Structure, Program of Work, and Market, Relevance. PD IEC/TR :2004, Insulation coordination. Computational guide to insulation coordination and modelling of electrical networks.

59 CHAPTER 3. Transient and Air Breakdown in Power System 59 CHAPTER 3 Transients and Air Breakdown in Power Systems 3.1. Introduction When the voltage in whole or part of the system exceeds the nominal or design voltage limit, this phenomenon called overvoltage. In HV and EHV systems, the voltages that cause the most risk of flashover within air gaps are those associated with lightning and switching operations. Circuit breaker opening/closing due to the fault and clearances, maintenance or network requirement, changing in load demand and power generation, etc., can affect the power system steady state, which needs to be settled down and reverted to the initial steady state situation. Thus, exchanging electromagnetic and electromechanical energy between the system components takes some time to push the power system back to the initial steady state which causes a short burst of energy in a very short time which is defined as transient [2.22]. On the other hand, the most important transient overvoltages are switching surges [3.1]. Whether, these overvoltages caused by energisation, disconnection, re-closing of the circuit breakers or by nature, i.e. lightning, fault due to unpredicted accident; the design, structure and performance of the network will be set according to the system s nominal voltage, magnitude of fault level and overvoltages.

60 CHAPTER 3. Transient and Air Breakdown in Power System 60 In Chapter 2, the strength of the gap and the methods used for calculation of the minimum approach distance have been investigated. However, as calculation of the minimum safety distance is based on both stress (switching overvoltage) and strength of the gap, in this Chapter, different sources of overvoltages (stress) have been investigated. These overvoltages have been studied further by use of PSCAD (Power System Computer Aided Design) simulation tool to illustrate the switching transient s behaviour along the transmission line. In the first part of this Chapter, different types of switching transients have been studied. In the second part of this Chapter, factors influencing the magnitude of switching transients have been reviewed. These factors are directly accounted for modification of the minimum safety distance for live-line working. Throughout this project, PSCAD [3.28] is used as an Electromagnetic Transient Simulation Program (EMTDC). Before its release in 1992, PSCAD has been extensively tested in North America, Japan, Australia and Europe. PSCAD is a graphical user interface program which represents and solves differential equations in the time domain. Users are enabled to run a simulation, analyse the results and manage the data in a graphical environment Travelling Waves Overhead lines and cables are presented by a pi-section to demonstrate a switching transient's characteristics- refer to Figure 3-1. In the pi-section models, electric and magnetic field properties are shown by the capacitance (C) and inductance (L). In Figure 3-1, by closing the switch, the current flows through the first inductor (L1) and charges the first capacitor C1. A gradual gathering of charge on the first capacitor (C1) creates a voltage that causes a current to flow through the second inductor (L2). Once

61 CHAPTER 3. Transient and Air Breakdown in Power System 61 again, this current charges the second capacitor (C2) and accumulation of the charges on the second capacitor causes a current flow to the third inductor (L3) and so on. Figure 3-1: Pi-section Presentation of Overhead Line and Cable This travelling wave propagates along the overhead lines and cables near to the speed of light due to disturbance of the steady state in a power system. They reflect back when reaching the open end of the line or where the impedance of the system is changing due to different component s connection. They could cause very high overvoltages which can cause insulation failure in the power system components. Also, they can cause a flashover between air insulated conductors. The high-speed travelling waves are known as Transverse Waves that are oscillating perpendicular to the direction of propagation. Although, these waves are explained by Maxwell s equations, however in a power system, analysing the overvoltage caused by travelling wave is done by travelling wave equations Wave Velocity on Overhead Lines Electric and magnetic Transverse waves that exist on the transmission lines appear on two or more separate conductors [3.2]. Therefore, by dividing the line into smaller sections as shown in Figure 3-2, the wave equation can be presented by the use of Table 3-1 [3.3]. Figure 3-2: Small Section of Transmission Line

62 CHAPTER 3. Transient and Air Breakdown in Power System 62 Table 3-1: Surge Impedance and Propagation Constant for Normal and Lossless Line [3.3] Propagation Constant Shunt Admittance Series Impedance of the line Surge Impedance Propagation Constant Lossless line (R=G=0) Surge Impedance Lossless line (R=G=0) γ(ω) = Y(ω). Z(ω) Y(ω) = G + jωc Z(ω) = R + jωl Zc = Z(ω)/Y(ω) jω LC L/C Calculation of the line capacitance (C) and the line inductance (L) per unit length (m) of the overhead line are shown in (3.1) and (3.2) for a single phase line where the inductance and capacitance depend on conductor radius (r) and conductors spacing (D ab ). L = 4 10 ln ( D r ) (H/m) (3.1) C = πε ln D (F/m) (3.2) r To illustrate transmission line characteristics, Figure 3-3 presents a single-phase transmission line where the line fed from an ideal generator via a circuit breaker (BRK1) at the beginning of the line. The transmission line is 100km of L6 tower used within National Grid network. The line and tower specifications are shown in Tables 3-2 and 3-3. The line contains a bundle of two Zebra overhead lines with 0.5m spacing. To simplify the model and alsoto better understanding the wave propigation, in this model the circuit breaker closes at the start point of the simulation. Figure 3-3: Simple PSCAD Power System Model

63 CHAPTER 3. Transient and Air Breakdown in Power System 63 The circuit breaker at the beginning of the L6_1 line is set to be closed at the peak voltage while the BRK2 is always open. The model contains an ideal generator to illustrate the line inductance and capacitances. All tower configurations used in this thesis are presented in Appendix 2. Table 3-2: Generator Parameters Source Voltage (Line Voltage) 400 kv Frequency 50 Hz Phase Angle 0 Inductance (Series) 0 [H] Resistance (Series) 0 [ohm] Resistance (Parallel) 0 [ohm] Table 3-3: Overhead Line and Circuit Breakers Parameters For All Overhead line Cable Short Line Case Tower Type L6 Single Circuit Steady State Frequency [Hz] 50 Number of Conductor 1 Total Overhead Line Length [km] 100 Shunt Conductance [mho/m] 1.0e-011 Conductor Radius [m] DC Resistance [Ω /m] e-3 Height from ground [m] 30 Ground resistivity [Ω *m] 100 Number of Sub-conductor 2 Sub-conductor space: D ab [m] 0.5 Breakers Open Resistance [Ω] 1.0e6 Breakers Closed Resistance [Ω] 0.1 Based on Equations (3.1) and (3.2), the capacitance and inductance of each phase will be calculated as follows; = 4 10 ( ) = μ = / π = ln ( ) = pf m = nf/km

64 CHAPTER 3. Transient and Air Breakdown in Power System 64 Therefore, the surge impedance (Z 0 ) and the time required for the wave to travel from the beginning to the end of the line were calculated using Equation (3.3) where R and G losses have been ignored in the analysis of the surge phenomena. + = + Therefore, the surge impedance becomes as follow; [Ω] (3.3) = [Ω] = = Ω As shown above, the surge impedance was found to be equal to Ω which is in the range of typical overhead line surge impedance of 200 to 500Ω [3.4]. By using the line inductance and capacitance calculated previously, the wave velocity is calculated as below; (Travelling wave velocity ) = 1 = = This calculation gives a wave velocity equal to x 10 6 m/s, which is very close to the speed of light ( x 10 6 m/s). Therefore, the time required for the wave to travel along 100km of the mentioned line is about 333µs which is very close to PSCAD simulation plot in Figure 3-4 with a value equal to 310µs. It needs to be highlighted that either both the calculation and the simulation may consist of many sources of error, such as plotting error, rounding up the values and also curve examination error due to human vision error. Therefore, these errors cause the calculation value to be slightly higher than the speed of light or the reading to be differing by 23µs.

65 CHAPTER 3. Transient and Air Breakdown in Power System 65 Figure 3-4: Surge travelling time: Top: E_sending; The Voltage at the Sending and Bottom: E_receiving; The Voltage at the Receiving End of the Line Wave Velocity on Cables In order to illustrate the wave velocity on cables, the cable used at this part of the thesis contains a single circuit three-phase 400kV cable used by National Grid, which is buried in a trench in the ground [3.5]. This cable is equivalent to 400kV, 1200mm 2 XLPE, ABB cable specification [3.6] presented by Table 3-4. The picture in Figure 3-5 shows a typical position of cables in the ground used by National Grid. Figure 3-5: National Grid direct buried cable diagram

66 CHAPTER 3. Transient and Air Breakdown in Power System 66 The cables are spaced horizontally with 400mm between each phase and buried approximately 0.9m-1.1m deep, depending on the location. The specification of the cable core used in this model is shown in Table 3-4. All binder, semi-conducting screen, insulation and conductor screen are merged into the insulator layer in PSCAD model. Table 3-4: Sample Cable Data for 400kV Single Core Cable, 1200mm2 ABB XLPE Cable [3.6] Cross section of conductor Diameter of conductor Insulation thickness Diameter Over insulation Cross section of screen Outer diameter of cable Capacitance Charging current Inductance Surge Impedance mm 2 mm mm mm mm 2 mm µf/km A/km mh/km mh/km Ω In order to measure the propagation speed and the surge impedance of the travelling wave, the model in Figure 3-6 is used. The impulse generator injects 10kA, 1.2/50µs current into 1 km of the cable section at the sending point and the time for the wave to travel along the cable and the surge impedance are calculated by monitoring the open end of the cable before any reflection. Figure 3-6: Impulse Generator Used in PSCAD In Figure 3-7, the PSCAD voltmeter at the sending end (Ea) was recording 318kV that is indicating a surge impedance of 31.80Ω which is very close to real data on Table 3-4. The propagation speed was calculated as v=102382km/s.

67 CHAPTER 3. Transient and Air Breakdown in Power System 67 Figure 3-7: Voltage at Sending Point (Blue Curve) Due to Current Impulse where, Ea and Eb are the sending and receiving voltages respectively Wave Reflection and Line Characteristics Impedance When a travelling wave reaches either an end of a line with a higher or a lower impedance of the current path, some portion or even the whole wave reflects back toward the original propagation source. The polarity and magnitude of reflected waves depend on the transmission line s impedance and reflection coefficient of the transmission line s discontinuity [3.7]. This can be explained with the help of Figure 3-8 and PSCAD model of transmission line shown in Figure 3-3. After the circuit breaker closure (BRK1) at the beginning of the line (point A on Figure 3-8), the voltage wave (Blue Curve) travels along the line and reaches the open end of the line (point B on Figure 3-8) at a time equal to. After the reflection, it takes 2 from the start time, for the travelling wave to reach the beginning of the line and causes a peak voltage (point C on Figure 3-8). The magnitude of this peak depends on the line inductance and capacitance. The period of the travelling surge will be equal to 4 and, in this case, it is equal to 1.048ms. Therefore, the following formulae confirm the results from the PSCAD.

68 CHAPTER 3. Transient and Air Breakdown in Power System 68 = (Hz) (3.4) ( ) = (line length (km))/(wave velocity(km/s)) (3.5) Figure 3-8: PSCAD Simulation Travelling Wave; Top: Voltage at Beginning, Bottom: Voltage at the End of Transmission Line This phenomenon can also be explained by the travelling wave theory by simplifying the PSCAD model to a schematic diagram shown in Figure 3-9. By considering the source impedance (Z S ) and transmission line surge impedance, (Z L ), the following equations can be produced to demonstrate the impacts of the source on reflected travelling waves. Figure 3-9: Behaviour of Voltage Travelling Wave at Transition Point In Figure 3-9, point A is a transition point where there is a change of circuit constant due to a junction between the transmission line and the generator. The impinging travelling wave/ incident wave (e) from the generator faces the reflected wave (e r ) from the transition point of the line (Point A). At this instance, the rising wave due to the

69 CHAPTER 3. Transient and Air Breakdown in Power System 69 conflict of the incident wave and reflected wave at the transition point would be reflected back into the transmission line which is known as transmitted wave (e t ). Incident Wave: Z = [Ω] (3.6) Reflected Wave: Z = [Ω] (3.7) Transmitted wave: e + e r = e t... Z = (3.8) Therefore, Equations (3.9) and (3.10) can be derived from Equations (3.6)-(3.8) as follows; e = Z Z e Z + Z (3.9) e = 2Z e Z + Z (3.9) Where (e r ) and (I r ) are reflected voltage and current waves respectively, and (e t ) and (I t ) are transmitted voltage and current waves respectively. Therefore, by considering Equations (3.9) and (3.10), based on the simulation results shown in Figure 3-8, after 262µs, the travelling wave reaches the open end of the line (point B), it takes almost 524µs (point C) for the wave to reach the transition junction again (between the transmission line and generator). By considering the direction of the surge from the end of the line toward the transition junction, Equations (3.11) and (3.12) can be derived from Equations (3.9) and (3.10) to calculate the magnitude of reflected wave toward the open end of overhead line. = (V) (3.10) = (V) (3.11) Therefore, the characteristic impedance of the transmission line and also load impedance can influence the polarity and magnitude of reflected waves. When a travelling wave is not facing a higher or lower impedance, there would be no reflection, however, when the wave meets the line-cable junction with higher impedance or open end of a transmission

70 CHAPTER 3. Transient and Air Breakdown in Power System 70 line, the reflected wave can be double in magnitude and reflects back toward to the source [3.7] Transient Classification Transients are defined and classified based on their origin into atmospheric or switching, or based on their transient generation mode into electromagnetic or electromechanical transients. In general, transients can be classified based on frequency and rate of voltage rise. According to CIGRE Classification of Overvoltage Based on Frequency [2.6]and IEC classification [3.8], transients are categorised into five groups based on their frequency ranges (Table 3-5), whereas, according to IEC 60071, the magnitude and duration of overvoltages classify the transient overvoltages as shown in Table 3-6. Table 3-5 CIGRE Classification of Overvoltage Based on Frequency [2.6] Frequency Classification Abbrev The Origin Magnitude Range Temporary Earth fault & Load Up to TOV Seconds Overvoltages Rejection 1.5pu Low-frequency Load rejection & Fault ---- Oscillation clearing 0.1Hz-3kHz Slow-front surges SFO Line switching 50 Hz-20 khz Fast-front surges FFO Reignition & prestrike/ lightning 10kHZ-3MHz Very-fast-front Disconnection 100kHz- VFFO surges switching in GIS 50MHz Up to 4pu Up to 7pu Table 3-6 IEC Classification of Overvoltage Based on Time Duration [3.8] Nature of the Transient Phenomena Lightning Switching Sub-synchronous resonance Transient stability Dynamic stability, long-term dynamics Tie line regulation Daily load management, operator actions Time Duration 0.1 μs 1.0 ms 10 μs to less than a second 0.1 ms 5 s 1 ms 10 s s s Up to 24 h

71 CHAPTER 3. Transient and Air Breakdown in Power System 71 Therefore, transients can be classified based on their causes or nature. In other words, transients in the power system can be due to external sources such as lightning or internal sources as a result of switching or temporary overvoltages Lightning Overvoltage Fast Front Overvoltages (FFO) are mostly caused by a lightning strike with a magnitude up to 7pu of nominal system voltage [2.6]. The first step of a lightning discharge is the formation of leader stroke due to the potential difference between the positively charged ionosphere and the negatively charged earth. At the earthing point, a large impulse current equal to tens of kilo amperes occurs which causes damage to the power system, and as a result, it causes a large magnitude of a transient wave at the point of strike. The transient waves caused by the lightning strike move along the transmission line and tower body and can cause an overvoltage up to 7pu [CIGRE Classification of Overvoltage Based on Frequency] in some parts of the network. The stroke damage with a speed very close to half the speed of light and a temperature up to 20,000 o C, time to crest of few seconds and decay time of microseconds is devastating. However, as live-line working takes place in good weather condition, the lightning overvoltage is not considered in this project Review of Main Sources of Switching Overvoltages In theory, the switching action on electrical circuit occurs by a single break action (opening) and a single making action (closing), and as a result, the magnitude of thr switching overvoltage can exceed even more than twice the system voltage. However, in reality, due to the interaction of the system and switching factors, the switching

72 CHAPTER 3. Transient and Air Breakdown in Power System 72 operation differs from the ideal. These factors can be one or combination of many factors such as; line and source impedance, transformer excitation characteristics, the existence of system compensation, Ferranti effect, circuit breaker characteristics, etc. The switching surge is a voltage transient/spike with a high amplitude and a different waveform from the system nominal voltage at any point or part of the system. The switching transient can take any shape depending on system configuration and transient source. The switching surge wavefront duration and its rate of rise determine the magnitude of the switching transient (stress), and, as a result, it has a significant impact on the voltage breakdown of a gap. Therefore, as switching transients (stress) are the most common sources of overvoltages, this project intends to investigate the switching transients to illustrate their influences on the minimum approach distances. The switching transient can be initiated by various events such as switching on/off the transmission or distribution network or a circuit with inductance and capacitance. The sources of switching transients are classified as below; Line energisation, Line re-energisation, Line disconnection, Fault initiation and fault clearance, Switching off small capacitive or inductive currents Line Energisation, re-energisation and Disconnection: The predominant switching overvoltage in the majority of HV and EHV systems are those caused by energisation or re-energisation of the unloaded line. Due to closing between the poles of circuit breakers and also electrical coupling between the phases, the maximum switching overvoltage on transmission lines can be severe.

73 CHAPTER 3. Transient and Air Breakdown in Power System 73 Furthermore, the magnitude of this overvoltage can be higher and even more severe if the circuit breaker re-striking or circuit breaker re-closure happened on a transmission line with a trapped charge. By closing the circuit breaker at the sending point of an open-end transmission line, voltage and current travelling waves rush into the transmission lines. The electrical circuit configuration influences the magnitude and waveform of switching surge (travelling wave) on both sides of the circuit breaker. In practice, before a circuit breaker mechanical closure, the electrical contact can be made due to the formation of circuit breaker prestrike. The time for prestrike occurrence (arc flash) depends on the voltage at the terminal of the circuit breaker and the withstand voltage across the breaker's terminals. In this part of Chapter 3, the source of switching transients such as energisation, reenergisation, disconnection and fault and clearance will be analysed by use of travelling wave theory and PSCAD simulation tool. I. Energisation: In Figure 3-10, energisation of an open-end transmission line by closing the circuit breaker produces a transient wave (e), which reflects back (e ) after reaching the open end of the line. In theory, switching surges due to energisation of a line with no trapped charge can create a value of overvoltage not exceeding twice of the system voltage. Figure 3-10: Sum of reflected voltage and current and sending waves

74 CHAPTER 3. Transient and Air Breakdown in Power System 74 By applying an open end resistance (Z k ) and a line resistance (Z) in the travelling wave theory, the following calculations can be produced: e= iz I = i-i e = i Z e = e+e e = i Z Total voltage due to reflection: e = 2Z k / (Z+Z k ) Therefore, the total reflected voltage can be calculated by use of Equation (3.12). e = 2Z k / (Z+Z k ) (3.12) In theory, switching surges due to energisation of a line with no trapped charge can reach a value up to twice of the system voltage, however, in reality switching surge can rise to 3pu-3.5pu [3.9]. Some of the system parameters influencing the magnitude of switching transients are line length and impedance, effect of series compensation or shunt reactors, source X/R ratio, transformer excitation, the behaviour of circuit breakers at the time of opening/closing and effect of surge arresters [3.12]. Figure 3-11 presents a transient simulation due to energisation of the 400kV transmission line with 120km of L6 overhead line shown in Figure 3-3. Point A presents the time when the circuit breaker closed at the beginning of the line, whereas point B shows the time when the travelling wave reaches the open end of the line.

75 CHAPTER 3. Transient and Air Breakdown in Power System 75 Figure 3-11: Voltage at the Sending and Receiving End Due to Energisation of 60km Line on 400kV System The simulation step time ( t) was set to 100µs. The time required for the wave to travel 120km at a speed of light is about 400µs which is very close to results from the simulation which is about 390µs. The voltage at the sending point (ESending) is about 327kV, and this voltage ramps up to the maximum value of 641kV (EReceiving) after reflecting back and forward along the line. The oscillating section (C) in Figure 3-11 shows the effect of travelling wave reflection on the transmission line. These oscillations wade away, depending on the characteristic impedance of the line, and the system returns to steady state after some period of time. II. Re-energisation: In 245kV systems, applying voltage to a no-load line without any trapped charge or open-end circuit can create a travelling wave as big as 2pu once it reflects back from the end of the line. At the same time, re-closing the circuit breaker at the beginning of a line with a trapped charge of -1.0pu can cause a total overvoltage up to 3pu in case of an ideal circuit [3.10].

76 CHAPTER 3. Transient and Air Breakdown in Power System 76 The magnitude of a switching surge depends on the size of the trapped charge and the point of voltage wave at which circuit breaker closure happens. The magnitude of switching transient for the case of re-energisation of a line for a single phase transmission line is different from 3-phase lines. This is due to interphase coupling and sequential pole closure of the circuit breaker on each line. The trapped charge stays for s on transmission lines if no wound voltage transformers (VTs), power transformer and reactors are connected to the transmission line. The only losses will be due to corona and leakages and, therefore, the losses and decay depend on the weather conditions [3.12]. The magnitude of the overvoltage at the end of the transmission line will rise to the highest value if the circuit breakers close at the opposite polarity voltage to the residual voltage on the line. Based on travelling wave theory, the maximum overvoltage occurs when the supply voltage is at its peak and the residual voltage is at its peak of opposite polarity. Under this condition, the voltage at the sending point has a magnitude of up to 2pu and when it reaches the open end of the line, it would rise to a value up to 3pu (phase to earth). The PSCAD simulation model of a transmission line with a trapped charge is shown in Figure In order to simulate the re-energisation of the transmission line, the circuit breaker (BRK1) in the model stayed closed at the beginning of the simulation while the circuit breaker (BRK2) at the end of the line stayed open throughout the simulation. Then, seconds after starting the simulation, the circuit breaker (BRK1) opened and re-closed at a random time again. The simulation repeated and the circuit breaker reclosing occurred 500 times within the voltage full cycle to achieve the highest overvoltage produced by circuit breaker closure.

77 CHAPTER 3. Transient and Air Breakdown in Power System 77 Figure 3-12: PSCAD Simulation Model of Trapped Charge The highest overvoltage was observed when there was a trapped charge of ~(-1)pu on the line. In Figure 3-13, the circuit breaker was opened 170ms after starting the simulation and after further 8ms the re-closure occurred. The maximum transient overvoltage at the end of transmission line reached 2.1pu whereas, in the case of energisation, the maximum overvoltage that appeared on the line was only 1.5pu. Figure 3-13: Energising of a Line, Top; Without Trapped Charge, Bottom; With Trapped Charge

78 CHAPTER 3. Transient and Air Breakdown in Power System 78 III. Disconnection Switching overvoltage due to disconnection events are generated when a system in a steady state is disconnected by a circuit breaker. Disconnection overvoltages could happen due to disconnection of an open-end line or a capacitor, open circuit transformer or disconnection of the line due to clearing a fault in the system. Before disconnection of a circuit even at current zero, the overhead lines, cables and even transformers contain some magnetic energy. The sudden interruption of current or the steady state of the system destabilising the changes to the system. Therefore, disconnection of a line can produce an additional transient that superimposes the instantaneous condition of the system. Figure 3-14 compares the simulation results due to energisation, re-energisation and disconnection of a simple 60km open-end of transmission line.

79 CHAPTER 3. Transient and Air Breakdown in Power System 79 Figure 3-14: Voltage Due to Top; Energisation, Middle; Re-energisation, Bottom; Disconnection The green curve in Figure 3-14 presents the voltage at the receiving end of the line. As shown in all switching configurations, the voltage at the end of the line is higher than the voltage at the sending point of the line. The computed maximum overvoltages due to energisation, re-energisation and disconnection are equal to 561kV, 775kV and 379kV respectively where the nominal voltage set to 400kV.

80 CHAPTER 3. Transient and Air Breakdown in Power System 80 IV. Fault and Clearance At the time of a fault, circuit breakers interrupt the current at the zero crossing. The fault could appear either after the breaker terminal (bolted terminal fault) or somewhere further on the transmission line. At the time of a fault, the line will be left with a charge at the instant of current interruption. This charge is at its maximum value on the breaker side, and it is equal to zero at the fault side. Therefore, the network tries to balance itself, and as this balancing of the voltage potential cannot take place instantaneously, an overshoot of voltage occurs and produces a travelling wave on the transmission line part. At the same time, the charge on the breaker contacts changes from zero to the instantaneous value of power frequency and creates a Transient Recovery Voltage (TRV) at the circuit breaker terminals which can also generate a travelling wave along the line. This oscillatory transient is due to the sudden change of the voltage or current in the steady state condition with polarity influenced by the polarity of the system nominal voltage wave. The rate of change of voltages and its magnitude depend on the length of the line, characteristics of the line and distance from the fault location with a frequency determined by inductance and capacitance of the line. However, the magnitude of the TRV on a circuit breaker depends on the rms value and the interrupted current (load current, fault current, etc.) [3-12]. As in practice, power systems are inductive under a fault condition, the power factor of the circuit from the circuit breaker s side is zero and lagging, and the power frequency of TRV is at its peak value at the instant of current zero when the interruption occurs.

81 CHAPTER 3. Transient and Air Breakdown in Power System 81 Figure 3-15: Oscillatory Transient Due to Interruption of Fault Current on PSCAD Model- ES: Voltage Sending Point, EL: Voltage along the Line, Earc: Circuit Breaker Arc Voltage. Figure 3-15 shows oscillatory transients due to the circuit breaker opening after 50ms from the fault time. At the point A, when the system voltage (phase 1: green curve) is at its maximum value, the circuit breaker terminals will be disconnected at current zero, and the green wave would bounce back and forward between the fault location and open terminal of the transmission line. The time to crest of each tooth shape travelling wave depends on transmission line s surge impedance. The red curve shows the TRV imposed on the circuit breakers' open terminals where its frequency is determined by the inductance and capacitance seen from the breakers looking upstream into the network. This TRV could be worse if a fault happens a few hundred meters up to a couple of kilometres away from the circuit breaker. This phenomenon is because the travelling wave on the line side has very high frequency and superposition of TRV and travelling wave creates a very high overvoltage on circuit breaker terminal and features a transient wave on the line side.

82 CHAPTER 3. Transient and Air Breakdown in Power System Switching Impulse Strength In HV and EHV systems, the voltages that cause the highest risk of flashover are those associated with lightning and switching operations. These overvoltages determine the external insulation design due to their high magnitudes. There are many factors which are influencing the breakdown voltages of uniform and non-uniform air gap. Therefore, the strength of the switching surge is very dependent on the maximum overvoltage and condition of the surrounding where the live-line working takes place. These factors have an impact on the strength under switching surge and the minimum voltage breakdown of the gap, and they are used in the determination of electrical distance and insulation coordination within a power system. As the gap flashover and its strength depend on a number of parameters, in the next section, the effect of these parameters and their influence on the strength under switching impulses and also their impact on the minimum approach distance are briefly explained Effect of Wave shape The switching wave shape is described by its time to crest and time to half value on their tail- refer to Figure Figure 3-16: Standards Switching Impulse Where V 50 is a half the time to crest of a Transient Wave [4.1]

83 CHAPTER 3. Transient and Air Breakdown in Power System 83 Based on IEC , the shapes and classes of overvoltages with a standard voltage shape are shown in Table 3-7. Table 3-7: Shapes and Classes of Overvoltages Standards Voltage [3.29] The time to crest (t cr ) is a primary factor that influences the formation of flashover due to transient overvoltages. The wave that produces the lowest value of U 50 (the voltage that has fifty percent probability of flashover) is called a critical wave of the gap where the air breakdown happens at or near the peak of the transient wave. At the same time, if the critical wave is shorter than time-to-crest, the voltage breakdownoccurs after the peak of transient, and it has a higher value of U 50 [3.19], [3.20]. This is shown in Tables 3-8 and 3-9. Table 3-8: U 50 of Rod-Plane for Fast and Slow Wave Shape [3.21] D(m) U 50 (kv) Fast TOV (1.2/50 µs) Slow TOV (t cr = 52µs) (t cr =112 µs) (t cr =375 µs)

84 CHAPTER 3. Transient and Air Breakdown in Power System 84 As shown in Table 3-7, the slow front transients are those with time to peak between 20µs and 5000µs and time to half value equal or less than 20ms. Switching overvoltages are slow front transients whereas lightning is a fast front transient. As shown in Table 3-8, a larger gap has a higher value of U 50 voltage. Also, as the critical wave of a fast front transient is shorter than its time-to-crest, the voltage breakdown occurs after the peak of the transient wave and it causes a higher value of U 50 in comparison to the slow front transient. As shown in Table 3-9, some wave shape are classified based on their time to crest (t cr ), whereas some of the wavshape are presented by their time to crest (t cr ) x the time to half value of the tail after the peak of the wave. Table 3-9: U 50 of Rod-Plane as the Function of Wave Shape, Non-Standard Switching Wave Form [3.21] D(m) x (t cr = 100µs) 412 (t cr = 72µs) 700 (t cr = 1052µs) U 50 (kv) and Wave shape 400 (t cr = 70µs) x x x x x x x4000 Further in Chapter 4, the influence of the line length on time to crest if transient wave has been investigated The U-Curve As explained earlier, the shape of the impulse has an impact on the strength of a gap. As a result, the insulation strength of a gap is influenced by the wave shape as the function of time to crest and time to the half value. By plotting U 50 values against the time to crest of transient overvoltages for a gap size, a U-shaped curve will be formed. This curve shows the voltage breakdown of the gap as a function of the time to crest of the transient overvoltage. Also, it indicates the minimum

85 CHAPTER 3. Transient and Air Breakdown in Power System 85 value that is corresponding to the critical wave and lowest U 50. The voltage data points along the curve are the voltages at which the strength of the gap is a minimum and the stress due to transient overvoltages causes the flashover within the gap. The U 50 (50% possibility of voltage breakdown) spark overvoltages as a function of time to crest with different spacing under different atmospheric conditions are shown in Figures 3-17 and The curve is called U-curve which is widely used in the calculation of insulation coordination [3.22, 3.23] Figure 3-17: U-Curves Obtained with Impulse Voltages of Various Time-to-Crests (T cr µs) Applied to Rod-Plane Gaps. Atmospheric Humidity in These Experiments Was Varied [3.22, 3.23] As shown in both Figures 3-17 and 3-18, small changes of time to crest of switching transients do not have a significant effect on U 50 of the gap. That means, breakdown voltages of the small gaps are not massively influenced by transient s time to crest. However, the bottom values of the U-curves (red arrow in Figure 3-17) show the minimum voltage required to form a flashover within the gaps. In other words, these points are where a gap has its minimum strength against the different transient times to crest.

86 CHAPTER 3. Transient and Air Breakdown in Power System 86 Therefore, based on experimental results shown in Figures 3-17 and 3-18, the minimum required voltage to form a flashover within a gap is very close to the bottom of the U- curve. At the same time, each point along the U-curve presents the voltage breakdowndue to the different times to crest of each air gap. A B Figure 3-18: A; Switching Impulse Flashover Voltage of Rod-Plane Gap, the picture on right corner of Figure A, indicates the rod-plan gap, B; Estimation of CRIEPI s Equation The results from experiments led to the assumption of Equation (3.13), where according to IEC 60060, the standard switching overvoltages is assumed to have the time to crest of 250µs. Equation (3.13) used by CRIPEI [3.21] calculates the U 50 of an air gap, and it

87 CHAPTER 3. Transient and Air Breakdown in Power System 87 is more complicated than the previous equations introduced by Paris, Gallet or Cortina and Herbec formulae [3.15], [3.19] and [3.21]. U 50RP = 1080 ln(0.46d + 1) (kv) (3.133) Equation (3.13) has been achieved by plotting an estimated curve connecting the critical points of experimental results from other researchers, and it has the advantage of being adjusted for the larger air gaps. It is also closer to experimental results when smaller gaps are in used. The formula has been adopted by IEC standards and used and developed by many utility companies as the fundamental formula in the calculation of the minimum safety distance. In live-line working, Equation (3.13) is used to calculate the 50% sparkover of a rodplane gap with a length of d (meter) which is estimated from the lowest part of the Ucurve of different gaps where the voltage breakdown is at its lowest value. The lowest values of voltage breakdown of the gaps give the highest possibility of flashover over where the smallest stress due to transient overvoltage overcome the strength of the air gap. By considering the minimum value of U 50 at the bottom of a U- curve, the risk of flashover will be at the minimum value Wave Polarity The switching surge flashover and also the strength of the gap depend on the polarity of the surge. As the gap between the electrodes is non-uniform [3.24], the positive polarity switching surge strength is lower than that under negative polarity [2.6]. A negative discharge applied to a field has less ramification and shorter length in comparison to positive surge [2.6]. Table 3-10 shows the effect of polarity on rod-plane gap for standard switching transients [3.15], [3.19].

88 CHAPTER 3. Transient and Air Breakdown in Power System 88 Table 3-10: Effect of Polarity on Rod-Plane Gap [3.15], [3.19] D(m) U 50 (kv) Positive Negative Wave shape (µs) x x x x4000 As shown in Table 3-10 and report in [3.25], the positive polarity voltage breakdowns are lower than corresponding negative polarities. Therefore, in the case of negative switching transients, higher voltage breakdowns will be required to form a flashover in a gap. As a result, for the purpose of live-line working, the positive polarity flashovers are considered for calculation of the strength of a gap. In very rare cases, due to different atmospheric conditions, gaps under negative polarity surges have a lower voltage breakdown compares with positive polarity switching surges [3.26]. Figure 3-19 shows experimental results of the rod-plane gap spark over voltages for both positive and negative polarities of DC and AC voltages. As shown in Figure 3-19, the positive flashover voltages have a lower magnitude of voltage breakdown than negative flashovers. Figure 3-19: Rod-Plane Gap; 1- Minute Critical Withstand AC and DC Voltages; 50% Percent Spark Over Voltage with Standard and Long Front Impulses [3.26].

89 CHAPTER 3. Transient and Air Breakdown in Power System Effect of Atmospheric Conditions The voltage breakdown of an air gap depends on atmospheric conditions at which the air breakdown or flashover occurs and it is influenced by three factors; pressure, humidity and temperature. In order to adjust the test results in any weather condition according to standard weather condition with a temperature equal to t 0 =20, a pressure of P 0 =101.3 kpa and humidity of 11g/m 3, the correction factor (k a ) is used. The voltage breakdown of the gap increases with the air density and humidity whereas, rain and its combination with a large variety of agents such as; coal and cement dust, fly ashes, salt spray, etc., can reduce the voltage breakdown of the gap and porcelain insulations [3.27]. Reducing the voltage breakdown of the gap influences the gap strength, and as a result, the minimum approach distances need to be increased for the purpose of live-line working.

90 CHAPTER 3. Transient and Air Breakdown in Power System Discussion and Conclusion PSCAD is world known transient simulation tool with slightly slower computation speed in comparison with other simulation software such as ATP- EMTP. However, due to having more type of controllers, power generators, recording components and also easier graphical interface, this project used PSCAD for the purpose of transient studies. In the calculation of the minimum approach distance, the influence of lightning overvoltage has been ignored, and only switching overvoltage needs to be considered. Switching transient due to energisation of an open-end line can create a reflected wave with the same polarity and a magnitude as twice as original wave. Switching transient due to re-energisation of an open-end line can create a reflected wave with the same polarity and a magnitude bigger than original wave. Large reflected wave with the same polarity means at the open-end of the circuit; the voltage can be varied while the current is zero. In the case of short circuit line, the reflected wave has an opposite polarity in comparison with the original voltage wave. Increasing the temperature or decreasing the air pressure will decrease the air voltage breakdown while increasing the humidity can increase the air voltage breakdown. Although IEC provides a guideline for calculation of the minimum approach distances, the influence of some factors such as wave shape, time to crest, the worst case atmospheric conditions, polarity, etc., has not been clearly investigated. These factors clearly introduce a considerable safety margin into the calculation of minimum safety distances.

91 CHAPTER 3. Transient and Air Breakdown in Power System 91 The waveshape can have a massive impact on voltage voltage breakdown of a gap whereas in IEC 61472, calculation of minimum approach distance has been done only by introducing the standard wave shape transient. Simulation scenarios in the next Chapters show that wave shape can be influenced by many factors within the power system and the risk of gap breakdown due to nonstandard switching transient is affected by transient wave shape.

92 CHAPTER 4. Network Studies and Calculation of MAD 92 CHAPTER 4 Network Studies, Overvoltage Levels and Resulting MAD 4.1. Introduction Calculation of the minimum approach distance (MAD) for live-line working can be done using the methodology explained by IEC in Chapter 2. This project uses U2 voltage (2% statistical overvoltage) from simulations as an input into the method described in IEC This methodology delivers a minimum approach distance based on the simulation results of the modelled network refer to Figure This chapter as the main body of this thesis intends to investigate the parameters influencing the magnitude of switching transients on a transmission line. A simple PSCAD model (Figure 4-19) presents each switching transient scenario to demonstrate the relationship between overvoltage level and source of transient i.e. energisation, reenergisation, fault /clearance, etc. Moreover, a fundamental transmission line model is produced to establish a new suggested set of minimum safety approach distances for the 400kV transmission line. In the simulation of switching overvoltage in this section, PSCAD software is used as a transient simulation tool.

93 CHAPTER 4. Network Studies and Calculation of MAD Simulation Methodology Throughout this project, each simulation has been repeated up to a maximum value of 2400 runs to achieve the highest accuracy of the results. Multiple run function in PSCAD has been used to simulate each scenario within a full voltage cycle of 20ms (50Hz). The incident time selections have been achieved in two ways: 1. Random 2. Sequential For example, Figure 4-1 presents the 27 points along the voltage wave when the switching or fault could occur (it assume that the time distances between points are equal). During the random selection, each simulation event such as switching or fault takes place at any random time within 20ms window. These event times do not necessary need to be at each exact incident point (green point) along the wave refer to Figure 4-1. In the sequential method, 20ms time window is divided by the number of simulation runs with equal interval time, i.e. 27 simulation events- refer to Figure 4-1 where a full cycle is given as an example of 20ms time window. The simulation starts with 1 st simulation run at first green point, and it continues to occur at each event point (green points) - refer to Figure 4-1. PSCAD tool produces a U 2 voltage at the end of each set of simulation. The simulation step times ( t) throughout this study were set to 100µs. Smaller steps could have also been used, however due to memory restriction and time consumption of smaller steps <100µs, the 100µs found to be adequate.

94 CHAPTER 4. Network Studies and Calculation of MAD 94 Figure 4-1: Model of Event Occurrence in Simulation In order to calculate the minimum approach distance for live-line working, the U 2 voltage (2% statistical overvoltage) is used. This is because the low-voltage tail of overvoltage distribution does not cause a flashover or it has very small or almost negligible probability of flashover. The U 2 overvoltage can be obtained by use of PSCAD simulation tool. It also can be obtained by use of PERCENTILE formula in Excel. In order to demonstrate the method for calculation of U 2 voltages, the distribution of switching overvoltages in one set of simulation has been plotted against the frequency of occurrence of each overvoltage magnitude- refer to Figure 4-2. The highest probability of switching transient in Figure 4-2, features the first peak of the transient distribution graph. Figure 4-2: Switching Overvoltage Distribution (pu)

95 CHAPTER 4. Network Studies and Calculation of MAD 95 To include the impact of the smallest magnitude of overvoltages on the minimum approach distance, the switching overvoltage distribution presented in Figure 4-2 is set based on 0.01pu scaling value (1pu = 343kV). Table 4-1 compares various values of U 2 voltage achieved by 2400 simulation runs produced by PSCAD and PERCENTILE formula from Excel. Table 4-1: U 2 Value Comparison Achieved by PSCAD and Excel Method Data from 2400runs in PSCAD PSCAD PERCENTILE EXCEL Switching Overvoltage (kv) Case 1 Case 2 Case 3 Case 4 Minimum (kv) Maximum (kv) Mean (kv) Std Deviation % Level % Level (U 2 ) % Level % Level (U 2 ) As shown in Table 4-1, the magnitudes of U 2 overvoltages from PSCAD multiple runs analyser are lower than Excel calculation. The Excel calculation was simply obtained by use of the PERCENTILE formula for 0.98 of switching overvoltages. By plotting the simulation values into a histogram probability plot, such as in Figure 4-2, the U 2 value will be slightly higher than PSCAD value and very close to Excel calculation. Therefore, in this project, analysing the data and achieving the U 2 values are performed by using the PERCENTILE formula in excel as higher values of overvoltage achieved by PERCENTILE formula in excel can feature more conservative results in the calculation of the minimum approach distance.

96 CHAPTER 4. Network Studies and Calculation of MAD PSCAD Goodness-of-Fit Testing for Weibull Distribution As shown in the Figure 4-2, there are some infrequent gaps along the distribution of switching overvoltages especially within the range of 1.15pu to 1.19pu and 1.21pu to 1.22pu. As Weibull distribution is a continuous distribution, the probability of dropping the variables to zero is almost zero. Considering of an existence of many falls to zero value along the switching overvoltage distribution could lead to the suggestion of using other methods such as moment estimators [4.2] which are not affected by zeros. However, there are lots of ways [4.3] to deal with zero data and estimation of the Weibull parameters which is beyond this project focus. As the distribution of overvoltages is not always a normal Gaussian distribution, IEC [4.1] suggests to use a Weibull distribution which is calculated based on U 50 and U 16 of Gaussian distribution, and it is truncated at three standard deviations (3σ) from U 50. U 50 and U 16 are values of 50% and 16% discharge voltage of self-restoring insulations. In PSCAD, the Weibull function is used to present the transient simulation results and the two percent statistical overvoltage. Figure 4-3 shows the Weibull overvoltage distribution plot on MATLAB, from simulation results of the model in the previous section.

97 CHAPTER 4. Network Studies and Calculation of MAD Overvoltages (kv) Figure 4-3: Overvoltage Weibull Distribution Plot The two percent overvoltage (U 2 ) is found to be 421.2kV which is the very close to PSCAD U 2 value. The result of MATLAB output file is shown in Table 4-2. Table 4-2: Simulation Result of MATLAB Output File Cases Overvoltage (kv) Maximum Minimum Mean % level % level Standard Deviation Table 4-2 illustrates the results from MATLAB which are very close to the result in Table 4-1. However, in this project, to obtain the U 2 voltage the PERCENTILE formula in Excel is used, as this approach provides more conservative estimation of U 2 voltage. The value of U 2 voltage obtained from this method produces a more conservative value for the minimum approach distances.

98 CHAPTER 4. Network Studies and Calculation of MAD Parameters Influencing the Overvoltage on Transmission Line The magnitude of transients on transmission line depends on the circuit parameters, performance characteristics of circuit breakers, system voltage profile and the time when the switching or fault occurs on the voltage wave cycle [4.4]. At the same time, many parameters within the power system are influencing the magnitude of switching transient on a transmission line. These parameters are length and type of transmission line, tower types, the presence of cable section, the probability of occurrence of different fault type and compensations. In this section of the thesis, a very simple model is produced to illustrate the impacts of different parameters on the magnitude of switching overvoltage on a transmission line. As a result, later in this Chapter, the minimum approach distance will only be considered for the worst case scenario for each study. In order to investigate the effect of different parameters on the magnitude of switching overvoltages, the model of transmission line system in Figure 4-4 is presented. Figure 4-4: Sample PSCAD Model of Transmission Line The complete list of towers and transmission lines data are provided in Appendix 2. The model in Figure 4-4 comprises a single circuit three-phase overhead lines which consist of four towers with an equal span distance. The 400kV model also consists of two circuit breakers at the both ends of the transmission line where they are connected to

99 CHAPTER 4. Network Studies and Calculation of MAD 99 generators. The type, length of transmission line and towers are changed based on nature of each study Transmission Line Effect The magnitude of switching overvoltages can be influenced by changing the length of transmission line. This phenomenon is explained by setting a different length of a transmission line on the PSCAD simulation model in Figure 4-4. The model consists of L6 towers used by National Grid network. For this study, the same values of X/R ratio were used in each set of simulation to investigate the impact of transmission line length on travelling wave transient. The generators in this section were set to have resistances and inductances to observe the effect of travelling wave on the impinging wave from generators. For the following study, the length of the transmission line was set to be 10km, 40km, 80km and 90km where the fault level was set to be 10kA. Furthermore, different switching scenarios have been investigated using a various lengths of transmission line. For each length of transmission line, the test was repeated 200 times sequentially to obtain the maximum overvoltage along the transmission line. As shown in Table 4-3, the maximum U 2 switching overvoltages due to energisation were observed on the longest line due to higher line inductance. Table 4-3: Magnitude of Switching Overvoltage Due to Various Length of Transmission Line Length of Transmission Line (km) Study Case P-E P-P P-E P-P P-E P-P P-E P-P Energisation Re-energisation Disconnection Fault & Clearances

100 CHAPTER 4. Network Studies and Calculation of MAD 100 As explained in Chapter 2, by increasing the length of a line, the time required for the wave to travel along the line will be increased. For clarification, the line can be considered as the model shown in Figure 4-5, where both line capacitances and inductances are responsible for the rise of voltage at the end of a transmission line. If the time required for the wave to bounce back from the open end of the line and reach the sending point is smaller than the voltage rise time at the sending point of the line, the peak of overvoltage will have a maximum positive value. Figure 4-5 was explained by use of Figure 3-1 in section 3.3 of Chapter 3. Figure 4-5: Overhead Model Type and Length of cable Section In order to investigate the effect of type and length of cable section, the following model in Figure 4-6 has been used. Length and type of the cable used in the cable section were individually set in PSCAD model according to Table 4-4. The PSCAD model consists of a 30km overhead line connected to a cable section- refer to Figure 4-6. The simulation was repeated 200 times for each case, and the circuit breaker (Breaker one) closed at a random time. Figure 4-6: PSCAD Model of Line-Cable Combination

101 CHAPTER 4. Network Studies and Calculation of MAD 101 By increasing the length of the cable section, the magnitude of overvoltage observed at both ends of the cable section was increased, whereas the overvoltage observed at the beginning of transmission line (Sending point) was decreased. The rate of change of overvoltage due to the different length of cable section is shown in Figure 4-7. Figure 4-7: Change of Overvoltage at Beginning and End of Cable Section Due to Changing the Length In order to illustrate the effect of cable type used by power networks on transient overvoltage, the data from Table 4-4 are used, and cable specifications in PSCAD model were set. Cable data in Table 4-4 are extracted from National Grid databases for the cable used on their network around the UK. Table 4-4: Three Types of Cable Specification Used by National Grid Corrugated Aluminium sheath 800mm 2 XLPE 1600mm 2 XLPE 2500mm 2 XLPE Diameter (mm) Diameter (mm) Diameter (mm) Conductor Binder Conductor screen Insulation Insulation screen Bedding Copper wire Screen Equalizing tape Screen binder Clearance Sheath Over sheath

102 CHAPTER 4. Network Studies and Calculation of MAD 102 Dc resistance ohm/km 16 ohm/km ohm/km Dielectric losses 2.43 W/m 3.31 W/m 3.90 W/m Conductor temperature 90 0 C 90 0 C 90 0 C As shown in Figure 4-8, increasing the size of a cable used within a transmission line increases the magnitude of switching overvoltage at both ends of the cable section. This phenomenon is due to the decreasing of the cable surge impedance. In larger cables, the inductance is lower whereas the capacitance due to thicker insulation layer is higher and As a result, the surge impedance Z 0 will be smaller in a cable such as the 2500mm 2 XLPE cable. Figure 4-8: Overvoltage at Beginning and End of Cable Section vs. Cable Type Cable Section Position on transmission Line Including the cable line section can increase the overvoltage that appears at the open end of the line. This is due to having different surge impedances on the overhead line and the cable section which increases the reflection coefficient that leads to increasing the voltage magnitude at cable-line junction. In order to investigate the overvoltage that appears on a transmission line, a cable section has been placed in 3 different positions along the line as shown in Figure 4-9 where the line and cable specification are shown in Table 4-5.

103 CHAPTER 4. Network Studies and Calculation of MAD 103 Figure 4-9: Schematic Model of Transmission Line Table 4-5: Cable and Overhead Line Specification For All Overhead line Cable Cable Overhead line sections Steady state Frequency 100 x 10 3 Hz 100 x 10 3 Hz Number of Conductor 3 3 Segment Length 15.5[km] 72 [km] Total Length 15.5[km] [km] Conductor Radius [m] x 10-3 [m] Total Impedance (Ω) 2.8 ohm 13.7 ohm 1 Section Inductance (µh) x Section Capacitance (µf) 3.23 x x 10-4 Surge Impedance (Ω) The time required (τ) for the wave to travel along the cable due to energisation is 98.7µs as shown in Figure Therefore, the propagation speed of the wave on a cable found to be 1.56 x 10 8 m/s which is almost half the value of wave propagation speed on an overhead line. Figure 4-10: Time Required for Wave to Travel along the Cable

104 CHAPTER 4. Network Studies and Calculation of MAD 104 By placing the cable section at the beginning of a transmission line, and due to the lower surge impedance of the cable, the reflected wave toward the generators will be smaller than other cases. At the same time, compared to other two cases that will be explained later, it takes shorter time for the voltage wave to reach its peak value (5.6ms and 8.1ms if the cable section placed at the middle and end of the line respectively). When the travelling wave on the cable reaches the cable-line junction (cable-overhead line joining point), and due to the high transmission coefficient (β), a significant portion of the travelling wave goes through the overhead line. At the same time, only a very small portion of the travelling waves reflected back into the cable section, and that is due to smaller reflection coefficient (ϒ). The transmission and reflection coefficients can be calculated by (4.1) and (4.2) where Zoh and Zcable are overhead line and cable surge characteristic impedances. β=2z oh / (Z oh +Z cable ) (4.1) ϒ= (Z oh -Z cable ) / (Z oh +Z cable ) (4.2) In Figure 4-11, the blue curve is the voltage at the beginning of the cable section, and as it was explained earlier, due to a small reflection at the cable-line junction, the rate of rise of the voltage was small. When the wave reaches the open end of the line, it bounces back toward the cable section. Due to a small transmitted coefficient (δ) of the line-cable section, only a very small portion of the wave would travel into the cable section, whereas due to high reflection coefficient (Φ) of the line-cable section, most of the travelling wave returns to the open end of the transmission line. The transmission and reflection coefficient can be calculated as follow; δ = 2Z cable / (Z cable +Z cb ) (4.3) Φ = (Z cable -Z oh ) / (Z cable +Z cb ) (4.4)

105 CHAPTER 4. Network Studies and Calculation of MAD 105 As shown in Figure 4-11, it takes 2ms for the surge to reach the maximum value of 484kV (at sending end). Due to smaller reflection coefficient at the cable-line junction in comparison to the cable section, it takes 98.7 µs for the reflected surge to reach the open end of the line. As the voltage reaches its highest peak value, the maximum overvoltage at the end of the line occurred 2.9ms after circuit breaker closing time with the highest value of 978kV. Figure 4-11: Overvoltage, Sending (Blue Curve) And Receiving (Green Curve) With Cable Section at Beginning of the Line By placing the cable section at the middle of the transmission line, the maximum overvoltage and the time to reach this value differ from the previous case. In this case shown in Figure 4-12, the first transmitted travelling wave sent from the circuit breaker travels through the overhead line, and it reaches the cable-line junction. After reaching the cable junction, due to lower transmission coefficient of the cable in comparison with an overhead line, a small portion of travelling wave goes through the cable section and a large part of the travelling wave reflects back toward the generator and conflicts with the voltage wave supplied by the generator.

106 CHAPTER 4. Network Studies and Calculation of MAD 106 In this process, the beginning end of the transmission line experiences its maximum voltage peak around half cycle after the circuit breaker closing time (point D in Figure 4-13). Due to the higher overhead line impedance compared to that of the cable section, the voltage peak is higher than the previous case. In this case, the overvoltage at the open end of the line reaches the maximum value of -757kV after 8.5ms from the circuit breaker closing time which is longer than the previous case. Figure 4-12: Schematic Model of Transmission Line with Cable Section Place in the Middle of the Line Figure 4-13: Overvoltage, Sending (Blue Curve) and Receiving (Green Curve) With Cable Section at Middle of the Line By placing the cable section just before the open end of the line, the maximum overvoltage of kV will be observed just 11ms after the circuit breaker closing time. It takes longer for the wave to reach the maximum value as the most of the wave facing the cable section (line-cable junction) reflects back toward the sending point.

107 CHAPTER 4. Network Studies and Calculation of MAD 107 However, as the wave bounces back from the cable section, a reflection with a negative magnitude reaches the overhead line junction. Due to the higher impedance of the line, its magnitude increases and a higher magnitude of reflected wave will be penetrating back through the cable section and reaches the open end of the line- refer to Figure Figure 4-14: Maximum Overvoltage, Sending (Blue Curve) and Receiving (Green Curve) with Cable-Line at End of Transmission Line Capacitor bank Capacitor banks are used to increase the power system's efficiency, maintain and control the transmission voltage profile within the limit, power factor correction and mainly regulation of reactive power. Therefore, insertion of reactive power elements into transmission line can provide the following: Reducing line voltage drops Limiting load-dependent voltage drops Influencing load flow in parallel transmission lines Increasing sign transfer capability Reducing transmission angle Increasing system stability

108 CHAPTER 4. Network Studies and Calculation of MAD 108 Energisation of a capacitor bank can give rise to transient in a network [4.7]. In general, if the system and overhead line resistance are ignored, the following equations express the inrush current into the capacitors. Where: ( ) = sin (4.5) = (4.6) = 1 (4.7) Increasing the size of a capacitor increases the inrush current and decreases its frequency. Therefore, in the case of a larger capacitor, a higher magnitude of switching transient will be observed. Therefore, the magnitude of switching surge and, as a result, the minimum approach distance for a transmission line connected to a capacitor bank is also influenced by the size of the capacitor bank. The series capacitor banks are generally used in a long transmission line or large generating power plant. Their main aim is to increase the efficiency of a transmission line. The series capacitor bank shall be capable of withstanding the rated continuous current, system swing currents, emergency loading, and power system faults and, in some applications, harmonic currents; these quantities normally are specified by the purchaser [4.5]. The series compensation on the transmission line is selected based on system power flow, system stability, short circuit and synchronous resonance and cost of the equipment. The main issue with series compensated line is that DC component of current associated with most faults would not decay, and instead, it generates an AC transient component of current on fault inception with a frequency equal to Equation (4.8).

109 CHAPTER 4. Network Studies and Calculation of MAD 109 ƒ = (4.8) Also, the degree of series compensation is calculated by the ratio of capacitive reactance of series compensation over inductive reactance of the line and shown in Equation (4.9); = 100% (4.9) The percentage selected for series compensation can be in the range of 20% to 80% of transmission line s impedance. If the degree to be set at 100%, a large current will flow into the system in the presence of small fault or disturbance and also it can cause the series resonant at the fundamental frequency. To investigate the effects during the line energisation, a single series capacitor bank was inserted in the middle of 300km L6 overhead line refer to Figure Zoom Figure 4-15: Series Capacitor Bank Modelling with a 41.91µF series Capacitor The properties of the line under investigation were set to be the same as in previous sections. The compensation value was changed in a range of 20 to 80 percent of the total line inductive reactance refer to Table 4-6.

110 CHAPTER 4. Network Studies and Calculation of MAD 110 Table 4-6: Series Capacitor Size Inductive compensation of the line 20% 50% 80% Total per phase line inductance [mh] Inductive reactance of 300km line [Ω] Capacitive reactance [Ω] Size of capacitors [µf] 104.7e e e-6 Adding a series compensation increases the transmission capacity. However, the voltage characteristics of a transmission line will also be changed. By increasing the degree of series compensation on the transmission line, the maximum overvoltages due to energisation of a line will be reduced. This phenomenon is due to the reduction of inductive reactance on a transmission line. According to Figures 4-16 to 4-18, the overvoltage due to energisation of the open end or lightly loaded line with 20%, 50% and 80% of the line inductive reactance are 1.92pu (660kV), 1.87pu(644kV) and 1.08pu(358kV) respectively for a 400kV network where the base value assumed to be 343kV. Therefore, increasing the compensation level can decrease the maximum overvoltage on the line. Figure 4-16: Overvoltage with 20% Series Compensation

111 CHAPTER 4. Network Studies and Calculation of MAD 111 Figure 4-17: Overvoltage with 50% Series Compensation Figure 4-18: Overvoltage with 80% Series Compensation In all above cases, a capacitor bank was inserted in the middle of the line as there is no set of standards concerning the location and number of series capacitors installed on a transmission line. Typically series capacitor banks are installed as a set of two, i.e. one at each end of the line. Series capacitors increase the voltage within the power system whereas, the inductive reactance of a transmission line can cause a voltage drop. The current through the capacitive reactance of the bank terminals cancel out the voltage drop and maintain the acceptable voltage profile. Therefore, by splitting the series capacitors and inserting them at two different locations, excessive voltage rise will be divided into two parts of the line. This case and also compensation at both ends of the line has been investigated in section 4.4.

112 CHAPTER 4. Network Studies and Calculation of MAD Network for Overvoltage Studies In order to illustrate the results from different switching events and calculate the minimum approach distances, a fundamental model of a power system transmission line has been created. The power network model in Figure 4-19 represents a section of the UK transmission system. A double circuit overhead line with a length varying between 10 and 120km (a range of tower models representing the different versions found in the UK 400kV system) is connected to a 400kV substation at both ends. Live-line work takes place on one of the two circuits on this line. The two substations (substations A & B) have two further 120km double circuit connections to remote substations. A basic generator model consisting of a voltage source and an impedance appropriate to represent the fault level of X/R ratio is connected at these remote substations each substation having the same generator type. Overvoltages on the system are monitored at five locations; each end of the double circuit line and at 25%, 50% and 75% distances along the line. Both phase to earth (PE) and phase to phase voltages (PP) on the live circuit (the one on which workers are active) and the coupled circuit are monitored. The PSCAD model and also schematic diagram of the network is presented by Figure 4-19.

113 CHAPTER 4. Network Studies and Calculation of MAD 113 A Figure 4-19: A; PSCAD Model of Transmission Line, B; Schematic diagram of the network The maximum P-E and P-P voltages are recorded in a range of overvoltage scenarios as below. The complete results from simulation will be presented in Appendix 3, whereas only the highest values of switching overvoltages will be considered for calculation of the minimum approach distances. The scenarios are as below: B

114 CHAPTER 4. Network Studies and Calculation of MAD 114 Energisation of the coupled circuit (with and without trapped charge) while liveline work takes place on the live circuit. Energisation takes place at a random time on the AC cycle with no significant scatter assumed between the closing times of the three phases. Disconnection of the coupled circuit while live-line work takes place on the live circuit. Disconnection takes place at a random time on the AC cycle with no significant scatter assumed between the closure times of the three phases. Faults and resulting clearance through operation of both lines end circuit breakers on both the live and coupled circuits (a range of fault types and locations being simulated). The fault is applied at a random time on the AC cycle with the circuit breaker being commanded to operate after a fixed time delay representing the operating time of the projection. Faults are cleared at current zero on each phase with no current chopping being assumed. These overvoltage studies have been carried out in one of two ways below; runs of the overvoltage simulations have been carried out with each type of fault being applied equally. 2. Another set of simulations has separated the runs for line-ground (LG), line-line (LL), line-line-ground (LLG) and line-line-line (LLL) faults. In this case, it is possible to generate an overvoltage profile based on an uneven mix of fault types. The use of an 80% LG, 17% LL, 2% LLG and 1% LLL ratio is an example of a more realistic distribution of fault type. The distribution of the fault type is obtained from test results in [4.8], and it could differ from one network to another.

115 CHAPTER 4. Network Studies and Calculation of MAD 115 Each generator at a feeding substation provides one-quarter of the fault level contribution to set the fault level at the main substation around 10kA to 40kA. The overhead lines were modelled using the Frequency Dependent (Phase) model. This is the most accurate overhead line model available in PSCAD. The model takes the geometric arrangement of overhead lines into the simulation along with the size of conductors used. From this information, PSCAD automatically computes the line parameters that are used in the model. L2, L6, L8, L9 and L12 towers were modelled in this work, these representing a small and large tower size used on the 400kV network. Information about the tower geometry and conductor types were taken from TGN (E) 166 and entered into PSCAD model. The geometric data was entered into PSCAD. Phase sequencing was rotated on the two circuits (abc and cba). Conductors co-ordinations and data are shown in Appendix Overvoltage Simulation Results The following tables present the U 2 value of switching overvoltages (which exceeded in only 2% of all cases) obtained from the various simulation scenarios. In the majority of simulations, the U 2 voltage was extremely close to the maximum overvoltage observed and, in some simulations, was higher (particularly when a high standard deviation existed in the overvoltage levels). The U 2 overvoltages shown in the following tables are given in per-unit on a base of 343kV. The peak phase to earth voltage on a 400kV system is taken to be operating above nominal at a voltage of 420kV. In all cases, the highest value of overvoltage achieved on a longer length of the transmission line when the fault level was set to 40kA during the fault and clearance scenario. The simulations in which the circuit breaker energised the coupled line (with the other circuit breaker remains open) yielded the following results.

116 CHAPTER 4. Network Studies and Calculation of MAD 116 Energisation of the coupled circuit (without -1pu trapped charge) occurred while liveline work takes place on the live circuit. Tables 4-7 and 4-8 present the results of energisation and re-energisation of 120km of different tower types. Table 4-7: Overvoltage Results for Line Energisation Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Overvoltage (pu) Table 4-8: Overvoltage Results for Line Re-Energisation Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Overvoltage (pu) In the event of line disconnection, the live circuit remains live at all times. The parallel circuit on the double circuit tower is then energised at one end by closing the circuit breaker at the substation. Disconnection of the coupled circuit occurred while live-line work took place on the live circuit. Disconnection took place at a random time on the AC cycle with no significant scatter assumed between the closing times of the three phases. Table 4-9: Overvoltage Results for Line Dis-Connection Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Overvoltage (pu) As mentioned earlier, the overvoltage studies due to the fault and clearance scenario in this thesis have been carried out in one of two ways;

117 CHAPTER 4. Network Studies and Calculation of MAD runs of the overvoltage simulations have been carried out with each type of fault being applied equally. 2. Another set of simulations has separated the runs for line-ground (LG), line-line (LL), line-line-ground (LLG) and line-line-line (LLL) faults. In this case, it is possible to generate an overvoltage profile based on an uneven mix of fault types. The use of an 80% LG, 17% LL, 2% LLG and 1% LLL ratio is an example of a more realistic distribution of fault type. In the first method, an equal number of fault types at a random time were set. The circuit breakers at each end of the overhead line were instructed to open 50ms after the application of the fault with clearance taking place at the current zero following the open instruction - refer to Table Table 4-10: Overvoltage Results for Fault & Clearance Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Overvoltage (pu) In the second method, simulation of different fault types were separated, and Table 4-11 produced to compare the influence of each type of fault with respect to the likelihood of that fault type occurring within the simulation for both P-E and P-P voltages. The results are based on an analysis of 120km L6 towers with a fault level at 40kA. Table 4-11: Overvoltage Results for Fault & Clearance Due to Simulation Setting Fault Probability: Individual fault type Fault Type 25% of each fault type (LG, LL, LLG AND LLL Total no of runs U 2 (pu) P-E P-P LG LL LLG LLL

118 CHAPTER 4. Network Studies and Calculation of MAD % LG faults, 17% LL faults, 2% LLG faults and 1% LLL faults As shown in Table 4-11, the U 2 voltages for each fault type are different from the values of the U 2 voltages in the other two scenarios with a combination of entire overvoltage distributions. P-E and P-P of U 2 voltages are equal to 2.48pu and 3.40pu when an equal number of fault type occurs in one complete set of simulations. These values are reduced even further to 2.37pu and 3.21pu when the fault types are weighted according to their likelihood of occurrence. This fact compares with values of 2.60pu and 3.48pu if the U 2 voltage is simulated for each individual fault type and the worst case is selected. Table 4-12: Overvoltage Results for Fault & Clearance Due to 80% LG Faults, 17% LL Faults,2% LLG Faults and 1% LLL Faults Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Overvoltage (pu) In the simulation of a line with compensation, reactive compensation was included through the addition of either an inductance or capacitance on the busbar of the main substation (substation A). For a 420kV line, the capacitive charging current of the line is approximately 1A per km (0.25 MVAR per phase), and,for instance, a 200km line would typically require 40MVAr of shunt reactive compensation per phase depending on system's operational requirement [3.10]. The reactive compensation had a value of 225MVAR (lagging or leading) which has been taken as a maximum representative value from the National Grid Seven Year Statement [3.30]. 75MVAR per phase at a phase voltage of 231kV gives a load impedance of 711Ω which is equivalent to an inductance of 2.26H or a capacitance of 4.48µF at the compensation bank refer to Tables 4-13 and 4-14.

119 CHAPTER 4. Network Studies and Calculation of MAD 119 Table 4-13: Overvoltage Results Due to Fault & Clearances with Inductive Compensation Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Overvoltage (pu) Table 4-14: Overvoltage Results Due to Fault & Clearances with Capacitive Compensation Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Overvoltage (pu) The results from fault and clearance simulation of a transmission line with compensation can also be modified for consideration of the uneven probability of fault type occurrence as shown in Tables 4-15 and Table 4-15: Overvoltage Results Due to 80% LG Faults, 17% LL Faults, 2% LLG Faults And 1% LLL Faults with Inductive Compensation Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Overvoltage (pu) Table 4-16: Overvoltage Results Due to 80% LG Faults, 17% LL Faults, 2% LLG Faults and 1% LLL Faults with Capacitive Compensation Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Overvoltage (pu)

120 CHAPTER 4. Network Studies and Calculation of MAD Calculation of Minimum Approach Distance As an example, the transient overvoltage results listed in Table 4-10, section 4.5 (Fault & Clearance) for the L6, 120km and 40kA fault current, are used to illustrate the calculation procedure. It must be noted that, in the calculations presented here, the ergonomic distance D E, the presence of a floating object and, thus, the floating object distance F are excluded. This implies that the factor k f =1.0. a. Determination of U 90 U e2 = 3.40 pu Using Equation (2.6): U e90 = 1.1 x 3.40 pu = 3.74 pu (= kv) Known values: k f = 1, k i = 1, k s = and k g = b. Determination of k a The gap factor for L6 Tower was calculated, and it is equal to k g = Using Equations (2.14) to (2.19), Table 4-17 was constructed which presents the values of k a at different altitudes and different voltage levels. Altitude (m) Table 4-17: Example Selection Table for ka Range of U 90 (kv) < >

121 CHAPTER 4. Network Studies and Calculation of MAD 121 For the voltage U e90 = kv, this falls in the range of >1200 kv. Given the selection of a sea level altitude, this gives a value of k a = c. Determination of D A After obtaining all the factors (k f = 1, k a = 0.991, k i = 0.95, k g = 1.346, k s = 0.936), Equation (4.10) was used to calculate K t : K t = k f k a k i k g k s (4.10) Therefore, the correction factor is calculated and found to be equal to Finally, Knowing that U e90 = 3.74 p.u (= kv) and by letting the floating object distance F and the ergonomic distance D E to be zero, Equation (4.11) explained in Chapter 2, was used to calculate the D A : D = 2.17 e ( ) 1 + F (4.11) D = 2.17 e. (. ) 1 = 3.67 m Based on the above method, the minimum approach distances can be calculated to correspond with the switching overvoltages in section 4.5. Tables 4-18 to 4-20 present the electrical distances (Du) calculated based on the fault and clearance scenarios. The clearances in Tables 4-18 to 4-20 are calculated based the selection of a 500m altitude, a conservative value for the most locations in the UK, where k f = 1, k i = 0.95, k g = and k s = The values in the below tables are only concerned with the fault and clearances scenarios as their overvoltages yield the highest values. The full calculated electrical distances (Du) for all scenarios can be found in Appendix 3. The minimum approach distances can be calculated by adding 0.4m (ergonomic distance) to the calculated electrical distances (Du).

122 CHAPTER 4. Network Studies and Calculation of MAD 122 Table 4-18: Electrical Distances (Du) for Fault & Clearance Simulation Scenarios (With No Ergonomic Distance D A ) Minimum Approach Distances (m) Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 4-19: Electrical Distances (Du) for Fault & Clearance Simulation Scenarios with Inductive Compensation (With No Ergonomic Distance D A ) Minimum Approach Distances (m) Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 4-20: Electrical Distances (Du) for Fault & Clearance Simulation Scenarios with Capacitive Compensation (With No Ergonomic Distance D A ) Minimum Approach Distances (m) Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P As mentioned in section 4.5, U 2 overvoltages can be affected by different fault type configurations and as a result, the required minimum approach distance will be varied. Tables present the electrical distances (Du) due to an uneven mix of fault types by use of 80% LG, 17% LL, 2% LLG and 1% LLL fault. Table 4-21: Electrical Distances (Du) for Fault & Clearance Simulation Scenarios with 80% LG Faults, 17% LL Faults, 2% LLG Faults And 1% LLL Fault Probability (With No Ergonomic Distance D A ) Minimum Approach Distances (m) Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P

123 CHAPTER 4. Network Studies and Calculation of MAD 123 Table 4-22: Electrical Distances (Du) for Fault & Clearance Simulation Scenarios with Inductive Compensation With 80% LG Faults, 17% LL Faults, 2% LLG Faults And 1% LLL Fault Probability (With No Ergonomic Distance D A ) Minimum Approach Distances (m) Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 4-23: Electrical Distances (Du) for Fault & Clearance Simulation Scenarios with Capacitive Compensation With 80% LG Faults, 17% LL Faults, 2% LLG Faults And 1% LLL Fault Probability (With No Ergonomic Distance D A ) Minimum Approach Distances (m) Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Influence of Atmospheric Conditions As explained in Chapter 2, the atmospheric conditions such as pressure, temperature and air density have an impact on the calculation of the minimum approach distance. The tables in Appendix 3 present the influence of weather condition on the electrical distances calculated in this investigation. As explained earlier, changing air density and pressure has a larger effect on the minimum approach distance than temperature. The results shown in Tables are produced to present the effect of altitude on the electrical distances. However, for the purpose of the investigation, the tables only present the electrical distances due to the simulation of fault and clearance as it yields the higher magnitude of switching overvoltages in comparison to the energisation, disconnection, etc.

124 CHAPTER 4. Network Studies and Calculation of MAD 124 Table 4-24: Influence of Altitude on Electrical Distances (Du) Due to Fault and Clearances (Without Compensation) - With No Ergonomic Distance D A Altitude (m) Tower Type Minimum Approach Distances (m) L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 4-25: Influence of Altitude on Electrical Distances (Du) Due to Fault and Clearances (Inductive Compensation) - With No Ergonomic Distance D A Altitude (m) Tower Type Minimum Approach Distances (m) L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 4-26: Influence of Altitude on Electrical Distances (Du) Due to Fault and Clearances (Capacitive Compensation) - With No Ergonomic Distance D A Altitude (m) Tower Type Minimum Approach Distances (m) L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P

125 CHAPTER 4. Network Studies and Calculation of MAD Figures 4-20 presents the influence of transmission line length and fault level for L6 tower at 500m altitude on the minimum approach distance. Figure 4-20: Top; P-E, Bottom; P-P. Influence of Length of Transmission Line on the Minimum Approach Distance The drop at the highlighted area in Figure 4-20, is due to the tooth effect of bouncing traveling wave on a short circuit between the circuit breaker and fault location. As this bouncing current has a high voltage potential and a very short distance to travel, the magnitude of the switching transient at this type of fault rapidly increases which causes a need to larger minimum approach distances.

126 CHAPTER 4. Network Studies and Calculation of MAD 126 It is clear that increasing the fault level and the transmission line length increases the magnitude of switching overvoltage. Figure 4-21 presents the influence of altitude on the minimum approach distances when the 120km line is under consideration. Figure 4-21: Top; P-E, Bottom; P-P - Minimum Approach Distance Influenced by Altitude and Fault Levels 4.8. Influence of Floating object on Minimum approach distance The minimum approach distance can be reduced by an introduction of tools, hanging basket or equipment required for live-line working, into the air gap. Nevertheless, the presence of the floating conductive object(s) reduces the net electrical length of the air gap [IEC 61472]. According to IEC 61472, the minimum strength of the gap in the presence of a floating conductive object can be estimated by use of Equation (4.12). In Equations (4.12) (4.14), the floating object correction factor is k f where L f and F are the

127 CHAPTER 4. Network Studies and Calculation of MAD 127 overall length of an air gap and the maximum dimension of floating object along the gap axis respectively. U = ln(0.46(l F) + 1) kv (4.52) Therefore, the ninety percent withstand voltage (U 90 ) of the gap, and minimum electrical distance (D U ) can be calculated as follow; U = k. ln(0.46(l F) + 1) (kv) (4.13) D = 2.17 e. 1 + F (kv) (4.14) If the length of floating in the axis of an air gap between the overhead lines to be assumed as 2m, based on Table 1 of IEC 61472, the floating object correction factor will be equal to Therefore, the switching overvoltages obtained from simulation of fault and clearance can be used to calculate the minimum approach distances for different towers with a presence of floating object. These clearances are shown in Tables 4-27 to 4-29 when the switching overvoltage for fault and clearances yields the highest values. The electrical distances are calculated based on the assumption that the floating object has a width of 2 meters between the air gaps. These calculations also have been done base on 500m altitude. Table 4-27: Electrical Distances for Fault & Clearance Simulation Scenarios at 500m Altitude With Floating Object With 2m Length in Direction of Phases (With No Ergonomic Distance D A ) Tower Type Minimum Approach Distances (m) L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P

128 CHAPTER 4. Network Studies and Calculation of MAD 128 Table 4-28: Electrical Distances for Fault & Clearance Simulation Scenarios with Inductive Compensation at 500m Altitude with Floating Object with 2m Length in Direction of Phases (With No Ergonomic Distance D A ) Tower Type Minimum Approach Distances (m) L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 4-29: Electrical Distances for Fault & Clearance Simulation Scenarios with Capacitive Compensation at 500m Altitude with Floating Object With 2m Length in Direction of Phases (With No Ergonomic Distance D A ) Tower Type Minimum Approach Distances (m) L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P As shown in the Tables , the highest value of P-E electrical distance for L2, L6, L8, L9 and L12 towers with a presence of floating object are 5.03m, 5.05m, 5.02m, 5.04m and 4.99m. These values for the P-P clearances are 6.83m, 6.79m, 6.90m, 6.93m and 6.88m for each tower respectively. Based on Equation (4.12), these values can be reduced or increased based on the geometry of floating object i.e. 5.03m when the floating object is 2 meters wide or 4.03m and 6.03 when the floating object has a width of 1 and 3 meters respectively.

129 CHAPTER 4. Network Studies and Calculation of MAD Discussion The results from simulations in this section illustrate the effect of different parameters on U 2 overvoltage and as a result, on the minimum approach distances. By examining the simulation results obtained from PSCAD, it is clear that some factors, such as trapped charge on a transmission line, fault level and source configuration, transmission line length and compensation, have a significant impact on the electrical clearances while other factors such as altitude have less impact. These results help the live-line operators to estimate the expected changes on the minimum approach distances due to different parameters. Also, it can assist the power system operator to estimate the magnitude of switching transients due to various events. This fact can help them to process the required preparation before a live-line working task. Also, it can be used to investigate the feasibility of live-line operation at different sections of a transmission line. As shown in the next section, the risk can also be estimated from the minimum approach distances of an air gap, and this calculation can provide the operator with an extra safety assurance. These studies illustrated that a longer length of transmission line causes a higher magnitude of switching overvoltage, however, if the length of a line is very long, the magnitude of switching transient can be decreased, and this phenomenon is due to the high impedance of extra-long transmission line. Simulation results indicate that there is little difference between the overvoltages seen on L6 and L9 towers, whereas the magnitude of switching overvoltages for these two towers are higher than those for the case of L2, L8 and L12 towers. A higher magnitude of switching overvoltages in L6 and L9 could be due to higher line inductance as these towers consist of a bundle of four overhead conductors.

130 CHAPTER 4. Network Studies and Calculation of MAD 130 Based on the simulation conditions in this work, fault and clearance scenarios yield the highest value of switching overvoltages and, as a result, calculation of the minimum approach distance was only base on transient simulation of fault and clearances. By considering the results from all simulations, fault and clearance scenarios with a presence of a capacitor bank(s) yield the highest magnitude of switching overvoltage. At the same time, increasing the number or size of capacitor bank on transmission line increases the maximum inrush current and reduces the overvoltage at the capacitor bank station. In practice, the system protection, load losses, transmission capacitance and many other factors can reduce the overvoltages. However, these overvoltages can be controlled by using the pre-insertion resistors and inductors, surge arresters or synchronous closing and at the same time any of these methods can fail due to human error, under-sizing the equipment, equipment contaminations, etc. [4.6]. While system parameters have a significant influence on the magnitude of transients and, consequently, on the electrical clearances, atmospheric conditions have an impact on the strength of the air gap. For instance, increasing the pressure due to higher altitude decreases the voltage breakdown of the air gap and, as a result, larger minimum approach distances will be required. As shown in Chapter 4, increasing the pressure due to increasing altitude has more influence on the minimum approach distance than other atmospheric factors. And finally, this work is based on the examination of the switching overvoltages under the worst case scenarios. As a result, the simulated overvoltages in this work are higher than expected overvoltages in National Grid network. Also as in practice, the magnitude of switching overvoltages in National Grid network is controlled by different protections equipment therefore, the simulated results and the calculated minimum approach distances in this work are very conservative.

131 CHAPTER 5. Live-line Working Risk Evaluation 131 CHAPTER 5 Live-line Working Risk Evaluation 5.1. Introduction In general, a workplace hazard is any possible potential damage or harm whether the cause is the work materials, work method, the condition of work or equipment that can affect someone s health. They can be originated from different sources such as; knife, benzene, electricity, wet floor, etc., whereas, a risk is a probability or chance, high or low, that any source of hazards harms someone refer to Figure 5-1. Figure 5-1: Risk and Hazard Explanation [5.1] Risk consists of two factors of severity and probability. Throughout this section, the probability of risk of failure of an air gap's safety distance will be assessed. This assessment is beneficial for the live-line operator to provide the lowest possible likelihood of risk of failure. Also, the calculation in this section provides a better understanding of the risk concern to the live-line workers. If the hazard and the risk are

132 CHAPTER 5. Live-line Working Risk Evaluation 132 measurable, therefore, by using a correct risk management process, they can be controlled. Figure 5-2 illustrates the risk management cycle. IDENTIFY EVALUATE REVIEW CONTROL Figure 5-2: Risk Management Process 5.2. Live-line Working Risk Evaluation As mentioned before, the risk in live-line working consists of two factors: severity and probability. The probability factor takes into account the probability or likelihood of occurrence of flashover in the gap where the live-line work takes place. This likelihood or probability is compared against the estimated probability of failure of the minimum approach distance s calculation method. As mentioned in previous Chapters, weakening the air gap strength by a higher magnitude of stress due to transient overvoltage or certain atmospheric conditions can cause a flashover. Therefore, the air gap insulation failure can be the result of either higher switching overvoltage or lower air gap insulation strength. Therefore, a wrong estimation of switching overvoltage or an air gap insulation strength can cause failure of an air gap and, as a result, it can cause an accident during the liveline working. Furthermore, calculation of risk of failure of a transmission line is necessary as any accident due to live-line working can result in a severe or fatal injuries. Although some

133 CHAPTER 5. Live-line Working Risk Evaluation 133 world-known association such as IEC, CIER, CIGRE, OSHA, LWA, etc., actively work on live-line working safety, but still according to UNIPEDE survey [5.2], there were 171 accidents and five fatalities due to live-line working. Unfortunately, the statistics of the scope and type of accident in individual countries are not unified, but live-line working accident and its fatal outcomes suggest the need for an expert benchmarking research to examine the existing minimum approach distance and associated risk with the method of calculation of these clearances. The IEC method only produces a value of MAD, and it contains no way of assessing the risk to a live-line worker. However, based on the IEC standards 61472, following statements had a minimising effect on the overall risk of gap breakdown. The actual system voltage is not always at a maximum value; The location of the work is not likely to correspond to the place where a transient overvoltage is at maximum value; The stress of the actual transient overvoltage wavefront is less than the critical front; Approximately, half of the transient overvoltages will be of negative polarity, and are less severe; The frequency and amplitude of transient overvoltages are reduced by restricting re-closing of circuit breakers Risk Assessment To investigate and evaluate the risk of failure at the time of live-line working, the following assumptions are set; Transient waves are divided into two categories: standards and non-standards transient waves.

134 CHAPTER 5. Live-line Working Risk Evaluation 134 The standard switching transients under investigation are those with time to crest of 250µs and time to half value of 2500µs as stated in Table 3-7 of Chapter 3 of this thesis. Any transient wave with a shape outside the above time classification (250/2500 µs) assumed to be non-standard waves. As this project used the IEC method for calculation of the minimum approach distance, all the assumptions set by IEC and explained in the previous section should be complied with Methodology for Risk Assessment (Standard Switching Transient) In this method, simulation results from Chapter 4 are used and the assumption is made that all switching transients are standard switching transients where the minimum voltage breakdown of the gap has a time to crest equal/close to 250µs with a time to half value of wave equal to 2500µs. Figure 5-3 presents the risk calculation methodology applied in this project. Risk Identification Risk Review Control Evaluation Switching overvoltage measurement Calculation of voltage breakdown of each particular gap Time to crest, live-line working time, lineman position, fault type, tower type and etc. Calculation of risk involved with particular live-line work Figure 5-3: Live-Line Working Risk Evaluation Process The IEC only produces a value of MAD, and it contains no way of assessing the risk to a live-line worker. The probabilistic method used for calculation of the risk to a

135 CHAPTER 5. Live-line Working Risk Evaluation 135 live-line worker in this Chapter is based on stress strength analysis. The probability of a specific value of switching overvoltage is combined with the probability of gap failure when that overvoltage is applied. This method is illustrated in Equation (5.1) where R is the risk per event, P b is the probability of a specific value of overvoltage and P 0 is the probability of the breakdown of a gap of a particular size for that particular voltage. = ( ). ( ) (5.1) Stress on the gap Switching overvoltage values from each set of the simulation are used to obtain the probability of switching overvoltage distribution. Microsoft Excel was used to process the PSCAD results. An example of overvoltage distribution produced by the analysis of the network model shown in Figure 4-19, is presented by the use of Figure 5-4. Figure 5-4 presents the overvoltage distribution caused by faults and clearance where the faults took place on the coupled circuit of the overhead line. The overvoltage magnitude ranges found to be from 2.11pu to 3.11pu (1pu being 343kV, the peak phase voltage of the 400kV system). Probability 2.00% 1.80% 1.60% 1.40% 1.20% 1.00% 0.80% 0.60% 0.40% 0.20% 0.00% SWITCHING OVERVOLTAGE DISTRIBUTION Switching Overvoltage (kv) Figure 5-4: Switching Overvoltage Distribution

136 CHAPTER 5. Live-line Working Risk Evaluation 136 This overvoltage distribution is used to calculate the voltage that has a 2% probability of being exceeded (U 2 ). Then, the U 2 value is converted to the required U 90 that can be applied to a gap with a ninety percent probability of withstand voltage. This voltage is then used to calculate the MAD. The results from the distribution of switching overvoltage and also associated minimum approach distance based on IEC are presented in Table 5-1. Table 5-1: Calculation Extracted from Simulation Results in Figure 5-4 Maximum U Voltage Type Minimum (kv) 2 D Std Deviation u (kv) (kv) (m) P-P Strength of the gap In order to calculate the risk, this project uses the IEC method for calculation of the minimum approach distances. The flow chart in Figure 5-5 presents the steps and influencing parameters in the calculation of gap strength. Figure 5-5: Flowchart Illustrating the Steps Undertaken for Calculation of Gap Strength

137 CHAPTER 5. Live-line Working Risk Evaluation 137 As shown in Figure 5-5, the atmospheric condition affects the strength characteristics of the gap (U 50 ) and, as a result, U 90 of the gap will be affected. As explained in Chapter 2 and in [5.3], the standard deviation (σ) decreases when the absolute humidity (h) and relative air density (δ) increase and, as a result, the nature of standard deviation can be expressed by Equation [5.2]. σ = [ (h 11)]/(1 + δ) (5.2) CFO At this stage, the minimum approach distance value calculated in Table 5.1 will be used to calculate the strength of the gap. The minimum approach distance of 2.98m is used to obtain the strength of the gap shown in Figure 5-6. In Figure 5-6, the U 50 voltage is equal to 1175kV where the probability of flashover is 50%. This value is calculated based on IEC equation as shown below: U 50 = 1080 ln (0.46d +1) (5.3) Equation (5.3), used by IEC, is based on the CRIPEI s formula [5.5], which is more complex than other equations used previously. This Equation is extracted from the relationships between the gap length and U 50 voltage (possibility of 50% sparkover voltage) during various experimental tests. Some of these results are shown in Appendix 4. Compared to other existing formulae, the CRIPEI s formula is being adjusted for larger air gaps, and it is closer to experimental results when smaller gaps are used. The formula has been adopted by IEC standards and used and developed by many utility companies as the fundamental formula for calculation of the minimum safety distance. According to the IEC and [5.4], only the upper tail of switching overvoltage distribution and lower tail of strength distribution up to a maximum of three standard deviations are required for calculation of the risk. Therefore, the calculated U 50

138 CHAPTER 5. Live-line Working Risk Evaluation 138 influenced by correction factors will be considered only within the range of ±3σ. Figure 5-6 presents the probability of voltage breakdown of a particular gap. Probability U Air Gap Voltage Breakdown (kv) Figure 5-6: Air Gap Voltage Breakdown Probability Intersection area The probability of each specific set of switching overvoltages obtained in section is then combined with the probability of a gap failure for a particular gap size in section when that overvoltage is applied. Figure 5-7 illustrates the application of Equation (5.1) and the orange curve presents the strength of the gap and the blue columns show the distribution of switching overvoltage along the x-axis.

139 CHAPTER 5. Live-line Working Risk Evaluation 139 Figure 5-7: Combination of Air Gap Voltage Breakdown Probability and Switching Overvoltage Distribution The probability of failure of the gap during a live-line work can be obtained from a product of the probability of switching overvoltages at and beyond the U 2 voltage and the probability of the gap failure at those corresponding voltages. Table 5-2 below presents the probability of switching overvoltage and gap failure for each corresponding voltage during the above simulation. Table 5-2: Minimum Approach Distance s Risk of Failure Obtained from Probability of Air Gap Breakdown and Switching Overvoltage Distribution Switching Overvoltage (kv) No of Occurrence Probability of Switching Overvoltage Probability of Gap Strength Failure Risk of Failure of LLW Gap E

140 CHAPTER 5. Live-line Working Risk Evaluation 140 As shown in Table 5-2, the risk as a product of two probabilities; gap breakdown and switching overvoltage, has a value equal to at 1071kV. This value is obtained as a result of multiplying the switching overvoltage probability and air gap failure probability ( ). Although, at some voltages below 1071kV, the probabilities for both switching and gap failure might have a value > 0 but, the product of these probabilities causing no risk to the gap where the live-line takes place. This is because the minimum approach distance for live-line working is calculated based on U 2 voltage and, as a result, the risk of failure is only considered for the switching overvoltages that are U 2 voltage. In Table 5-2, the value of correspondence to the risk of failure of the gap, needs to be divided by two as this value contains the risk for both positive and negative switching overvoltages. As explained in Chapter 1, in the case of live-line working, only positive switching transients are considered because a lower positive polarity needs to cause a flashover within a gap in comparison to the negative switching transient. As a result, the actual value of risk for the air gap found to be equal to Methodology for Risk Assessment (Nonstandard Switching Transient) The steps undertaken to calculate the risk of non-standard switching transient are similar to the procedure explained in the previous section. However, the probability of the breakdown of a gap needs to be estimated based on the transient wave shape. Based on the calculated MAD and assuming the worker is operating at this minimum approach distance, the risk associated with the full overvoltage distribution can be estimated. This estimation is carried out using equations that relate the gap sizes and

141 CHAPTER 5. Live-line Working Risk Evaluation 141 voltage breakdowns of the gaps as a function of time to crest. These equations are based on data for rod-plane gap sparkover with positive polarity extracted from Table 5-1 of [5.5]. Table 7-54 in Appendix 5 presents these Equations for different gap size as the function of time-to-crest. As an example, Equation (5.4) is applied for calculation of U 50 of any gap size within the range of 1m-10m with a time to crest of 50µs. U = d d d (5.4) The U 90 of the gap is then obtained by multiplying the U 50 by a correction factor K t to form the ninety percent statistical withstand voltage of the gap as explained in Chapter 2. The U 90 and U 50 are then used to obtain the probability of voltage breakdownof the gap. Table 5-3 shows the risk involved with live-line working at the calculated MAD of 2.98m shown in Table 5-1 for both standard and non-standard switching overvoltages. These values are for a worker within a phase to earth gap. Table 5-3: Estimation of Risk Based on Transient Time-to-Crest Case Risk of flashover per overvoltage event 50% positive voltages / standard time-tocrest % positive voltages / 50 µs rise time crest % positive voltages / 100 µs rise time crest % positive voltages / 200 µs rise time crest % positive voltages / 250µs rise time crest % positive voltages / 450 µs rise time crest Figure 5-8 presents, the calculated risk as the function of time to crest for the same gap size under switching overvoltage distribution shown in Figure 5-4. As shown in the Figure 5-8, the surge with a time to crest equal to 250 µs has the highest probability of failure with a value equal to As explained previously, this is because the

142 CHAPTER 5. Live-line Working Risk Evaluation 142 gap has the lowest voltage breakdown at the bottom of U-curve where the surges time to crest is around µs, at the bottom of U-curve. 1.00E E E E-06 Risk of Failure 6.00E E E E E E E Time-to-crest (µs) Figure 5-8: Risk as the Function of Time to Crest The risk calculated based on time to crest can also be influenced based on length of transmission line, tower type and also other system influencing factors. In the next section, risks are calculated for each simulation result presented in Chapter Evaluation of Risk Based on Simulation Results Anticipated risks involved with live-line working in this section are based on simulation results shown in Tables 4-18 to 4-23 of Chapter 4 where the maximum switching overvoltages yield the highest value due to fault and clearance of a model with and without compensation. The risk is purely estimated based on simulation results for each particular network where positive overvoltages with different time to crest have participated in risk calculation.

143 CHAPTER 5. Live-line Working Risk Evaluation 143 As shown in Tables , the risks are calculated for the worst case scenarios when the fault and clearances happen on the 120km overhead line with 40kA fault level. These risks can be even further reduced if the probability of each fault type is considered in simulation method. The results suggest that the risk is higher when the worker is performing a task in proximity to the shield wire (and hence vulnerable to phase to earth overvoltages). The risk remains very close to 1 in 100,000 per overvoltage event. However, if the ergonomic distance to be added to the calculated electrical distance in previous sections, the total risk value would be lower than 1 in 100,000 per overvoltage result. Table 5-4: Calculated Risk for Fault & Clearance Simulation Scenarios Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P RISK x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 Table 5-5: Calculated Risk for Fault & Clearance Simulation Scenarios with Inductive Compensation Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P RISK x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 Table 5-6: Calculated Risk for Fault & Clearance Simulation Scenarios with Capacitive Compensation Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P RISK x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 Tables 5-7 to 5-10 present the impact of changing the simulation method and distribution of switching overvoltage on the calculated risk. The Tables

144 CHAPTER 5. Live-line Working Risk Evaluation 144 present the risk involved with live-line working when the weighted type of fault are considered. Table 5-7: Calculated Risk for Fault & Clearance Simulation Scenarios with 80% LG Faults, 17% LL Faults, 2% LLG Faults And 1% LLL Fault Probability Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P RISK x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 Table 5-8: Calculated Risk for Fault & Clearance Simulation Scenarios with Inductive Compensation with 80% LG Faults, 17% LL Faults, 2% LLG Faults and 1% LLL Fault Probability Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P RISK E x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 Table 5-9: Calculated Risk for Fault & Clearance Simulation Scenarios with Capacitive Compensation with 80% LG Faults, 17% LL Faults, 2% LLG Faults and 1% LLL Fault Probability Tower Type L2 L6 L8 L9 L12 P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P RISK x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 x10-6 Figure 5-9 compares the risk of failure for the calculated electrical distance presented in Table The figure compares the risk for both P-E and P-P voltages where the time to crest of the surges are varied. As it shown, the highest value of the risk for all type of towers is when the surge has a time to crest value around µs.

145 CHAPTER 5. Live-line Working Risk Evaluation 145 Risk of Failure 8.00E E E E E E E E E+00 L2 L6 L8 L9 L Time-to-crest (us) 1.40E E-05 Risk of Failure 1.00E E E E-06 L2 L6 L8 L9 L E E Time-to-crest (us) Figure 5-9: Risk of Failure as a Function of Time to Crest on Different Towers for Top: P-E and Bottom: P-P Voltage The risk of failure caused by changing the length of the gap for one set of switching overvoltage distribution is shown in Figure The calculated length of the gap based on IEC method found to be ~1.63m for the set of the simulation model.

146 CHAPTER 5. Live-line Working Risk Evaluation E-02 P-E Risk of Failure 1.74E E-02 Risk of Failure 1.24E E E E E E Air Gap (m) Risk of Failure 1.00E E E E E E E E E E-05 Risk of Failure: 2.63E-05 at m gap 0.00E Air Gap (m) Figure 5-10: Risk of Failure for P-E Voltage as the Function of Changing the Gap Size, Bottom: The Zoom in Graph of the Top Graph

147 CHAPTER 5. Live-line Working Risk Evaluation Discussion: The risk to the live line workers in this project, was evaluated based on the presumption of occurrence of the highest magnitude of overvoltage at the same location as line workers. This assumption provides pessimistic results in the calculation of the risk involved with live-line working. This risk can be reduced even further, as the position of the linesmen and a fault at each part/section of a network might be different. The calculated risks show a lower value of risk involved with live-line working when there is a consideration of fault type probability in the simulation method. Consideration of uneven fault distribution is a more realistic assessment of the risk in live-line working and as it shown, the risk clearly influenced by the probability distribution of fault types. As the risk of a fault and clearance event is higher than the risk during an energisation, re-energisation or disconnection events, the values shown in all tables within this chapter are based on the fault and clearances scenario. These results yield a conservative value of risk for the purpose of live-line working. The results in Table 5-3 and Figure 5-9 present the influence time to crest on the calculated risk. The highest risk remains in all towers when the time to crest is at the critical value while lower risks exist for waveforms with time to crests outside of this value. Table 5-10 presents the rate of change of the risk for different wave shapes for both P-E and P-P clearances at L6 Tower. For example, in the case of P-E risk, by changing the time to crest of the transient 50µs to 100µs, 100µs to 200µs and 200µs to 250µs, the risk values increase and equal to 6.88 x 10-9, 5.45 x and 4.29 x 10-6 respectively. Also for the same P-E risk values, if time to crest of the transient changes from 250µs to 300µs or 300µs to 450µs, the values of risk will be reduced to x 10-6 and x Table 5-10 clearly illustrates the influence of the change of wave shape on the

148 CHAPTER 5. Live-line Working Risk Evaluation 148 magnitude of the risk involved with live-line working. Throughout the study, the risk has been predicted to change dramatically, due to various transient times to crest. Risk Time to Crest (µs) 50 to to to to to 450 P-E 6.88 x x x x x 10-7 P-P 1.40 x x x x x 10-6 Table Rate of Change of the Risk Due to Change of Wave Time to Crest

149 REFERENCES 149 CHAPTER 6 Conclusion and Further Works 6.1. Conclusion This thesis assessed the impact of the various variables within a particular section of a transmission line on the magnitude of switching overvoltages and relative minimum approach distances. These assessments were based on the simulation of switching overvoltages, using 400kV transmission line model in PSCAD. These simulations only considered switching overvoltages, as live-line working only takes place in good weather conditions and, as a result, lightning overvoltages are irrelevant. This thesis also takes into account the method used by other standards, and it has been found that the IEC method provides a more comprehensive approach compared with the IEEE method. The IEEE method does not take into account the altitude below 900m whereas, in the IEC method, the effects of different parameters such as altitude, weather conditions (temperature, humidity and pressure) and also the effects of a broken insulator, floating objects have been considered. According to the conducted calculations in this thesis and also based on the IEC approach, it has been assumed that the highest magnitude of overvoltage occurs at the same location as the line workers. This value provides pessimistic results in the calculation of the risk involved with live-line working. The magnitude of switching transient can be reduced due to travelling surge along the transmission line. Therefore,

150 REFERENCES 150 the position of linesmen and also an average number of occurrence of a fault within one part or section of a network can change the risk value. Consideration of existing surge arresters along the transmission lines, also the existence of protection devices, circuit breakers and switches within transmission network can reduce/ control the magnitude of switching overvoltages, and, as a result, they lessen the value of the risk involved with live-line working. However, due to a very high importance of safety, the primary consideration of this research was to ignore all the possible limiting conditions and consider the worst case scenario that an incident could happen. Any incident could happen due to, human error, equipment failure, atmospheric contamination (i.e. surge arrester failure), equipment ageing and deploying the wrong equipment (under sizing), etc., which can alter or even abandon the performance of protection devices. This work is based on the examination of the switching overvoltages under the worst case scenarios. As a result, the simulated overvoltages in this work are higher than expected overvoltages in National Grid network. As in practice, the magnitude of switching overvoltages in National Grid network is controlled by different protections equipment therefore, the simulated results and the calculated minimum approach distances in this work are very conservative. This work is based on the examination of the switching overvoltages under the worst case scenarios. As a result, the simulated overvoltages in this work are higher than expected overvoltages in National Grid network. Also as in practice, the magnitude of switching overvoltages in National Grid network is controlled by different protections equipment therefore, the simulated results and the calculated minimum approach distances in this work are very conservative. Therefore, the minimum approach distances calculated by IEC method and used by National Grid are adequate and can be applied. The minimum approach distances

151 REFERENCES 151 used by National Grid have a very small risk below one out of 100,000 events which is even less than accepted risk of flashover within an airgap clarified by the IEC standards Impact of different Parameters on Minimum Approach Distance During simulation of different sources of overvoltage, the fault and clearance scenario was found to have the highest magnitude of switching overvoltages compared to energisation, re-energisation and disconnection events. It also has been found that the magnitude of switching overvoltage is higher when the transmission line is connected to a capacitive compensation bank(s). Therefore, the results of simulations shown in all tables within the main body of the thesis are based on simulation of fault and clearance scenarios on transmission lines. Results from the simulation of events such as energisation, re-energisation and disconnection could be very pessimistic as in this project, the circuit breaker opening and closing occur at a random/ sequential time within a time window of 20ms. However, in reality, circuit breakers operation are manually controlled, and they are operated (opening/ closing) at a particular time along the system voltage wave. This method of operation can reduce the magnitude of switching overvoltage due to circuit breaker closuring time. The results show that overvoltages are more likely to yield higher values in the case of L6 and L9 towers. The L6 and L9 towers conductors have bundles of four ACSR Zebra whereas L2, L8 and L12 towers have bundles of two conductors refer to Appendix 2. Therefore, the line inductance and capacitance of L6 and L9 transmission lines are higher than other towers under investigation in this thesis. Therefore, the magnitude of

152 REFERENCES 152 switching overvoltages was found to be higher and, as a result, the minimum approach distance required for these types of overhead lines were larger. Increasing the length of transmission line increases the magnitude of switching overvoltage along the line and, as a result, the required minimum approach distance would be larger. It has also been demonstrated that increasing both fault level and transmission line length increases the magnitude of switching overvoltages in all simulation cases and, as a result, higher fault level and longer transmission line yield the highest switching overvoltage. External influencing parameters found to have a direct impact on the strength of the gap and, as a result, they have a bearing on the minimum approach distance. For an instance, the humidity will increase the voltage breakdown of the air gap, whereas increasing the altitude decreases the strength of the gap and its voltage breakdown. It has been found in this thesis work and other literature reviews that altitude has more influence on the minimum approach distance than other atmospheric conditions and more likely dictates the voltage breakdown of a gap where live-line working takes place at various altitudes. This is due to the changing of the pressure at different altitudes. Therefore, decreasing the pressure due to increasing the altitude decreases the voltage breakdown of the gap and, as a result, a smaller magnitude of switching overvoltage is required to form a flashover within the gap. Therefore, the minimum approach distance will be increased as a result of increasing the altitude or decreasing the pressure. The results from calculations performed in this work shows 12% difference in the minimum electrical distances when the altitude changes from the sea level to 1000m for voltages lower than 500kV. At the same time, by increasing the voltage, this difference reduced to 4% for voltages above 900kV. A broken/ contaminated insulator can reduce the strength of the gap and the result; a smaller voltage breakdown will be required to form a flashover within a gap. Therefore,

153 REFERENCES 153 a larger minimum approach distance will be required for such scenarios. This fact can be true when the live-line working takes place at the tower. The simulation of transients on transmission line illustrates that when the time to crest of transient overvoltages are around 200µs to 300µs, the voltage breakdown of the gap is at the lowest value. The IEC method for calculation of U 50 of the gap calculates the electrical distance where the gap is at the venerable time. In other words, the method used by IEC calculates the worst case scenario for all gap sizes. The suggested method used for calculation of the risk of failure in this thesis confirms that the IEC method is adequate for all switching transient wave shape and illustrates the validity of the method for all the gap sizes. This thesis has illustrated a framework that could be used to assess the risk to a live-line worker at the time of a switching event. It is not proposed that this method replaces other international standards, but it could be of use in many situations including where utility companies wish to develop a complete understanding of the risk associated with live-line working. However, the new proposed method in Chapter 5 can be applied to calculate the risk based on different wave shapes and peak voltages. Finally, based on the simulation of a particular section of the network and a suggested method for calculation of the risk, the minimum approach distances using the IEC approach founded to be very conservative and adequate with a low-risk value Future Work Although the work presented within this thesis has fulfilled all of the research aims, nevertheless, due to the vast application of live-line working, there are some areas where this research could be extended.

154 REFERENCES 154 This thesis has explained the method structure and also the calculation of the risk involved with the minimum approach distance obtained from the IEC method. However, both calculation methods can be applied for voltage range between the 72.5kV and 800kV. Therefore, the same method of calculation can be implemented for 275kV, 500kV and 750kV (transmission level). Same calculation method can be applied to investigate the minimum approach distances at substations. However, the gap factor (kg) values are different for each scenario. The impacts of climate change can be considered and need to be studied. It would be valuable to extend the research to examine the effects of climate change on the current thermal capacity of cables and overhead lines, performances of transformers, surge arrester, insulators, circuit breakers, etc. As the magnitudes of switching overvoltages and also the probability of insulator failure can be directly affected by the climate change/ atmospheric conditions, further investigation may be possible. The work can also be extended to renewable sections as some renewable sources such as Batteries can have various export capacities. For instance, the output of a battery farm could change from a maximum export to a maximum import with massive voltage step change in just a few millisecond and, as a result of changing power flow direction, the minimum approach distances can be affected. Throughout this project, the position of the linesmen assumed to correspond to the location of the maximum switching overvoltage/ U2 voltage whereas, in practice, the maximum switching overvoltage could appear hundreds of kilometres away. Therefore, the overvoltage to be seen at the linesmen's location might be lower than the actual maximum overvoltage and, as a result, a smaller minimum approach distance would be required. Therefore, the risk and the minimum approach distance might need to be calculated by considering the location of linesmen and the fault within the system.

155 REFERENCES 155 Distribution of fault type used by this project is based on [4.8] whereas, further investigation can be conducted to estimate the fault type probabilities at different transmission levels. This information needs to be obtained from power system operator for each part of the network. Furthermore, the value of the risk can vary if the statistics of the fault events, the number of faults occurrence and key parameters of the network under study is available.

156 REFERENCES 156 REFERENCES 1.1 S. Machi, C. Malaguti and M. Ricca, "Live Line Working in Italy," in IEEE Power Engineering Review, vol. PER-1, no. 4, pp , April Live Work Guide for Substations". EPRI (Electric Power Research Institute), Palo Alto, CA: Technical Update October Retrieved 8 December G. Stix, "Working hot: life at 765 kv," in IEEE Spectrum, vol. 25, no. 9, pp , Sept doi: / G. Gela, C. Vincent and M. Charest, "IEC/TC78 live working : structure, strategy and program of work," Transmission and Distribution Construction, Operation and Live-Line Maintenance Proceedings IEEE ESMO IEEE 9th International Conference on, pp. 41-5, Montreal, Que., V. Lovrencic and B. Gomišcek, "Live working as an example of electrical installation maintenance with the zero accidents philosophy," Live Maintenance (ICOLIM), th International Conference on, pp. 1-8, Budapest, T. Harada, Y. Aihara and Y. Aoshima, "Influence of Humidity on Lightning and Switching Impulse Flashover Voltages," in IEEE Transactions on Power Apparatus and Systems, vol. PAS-90, no. 4, pp , July Q. Xu, "The study of humidity on flashover voltage characteristics for positive SI," Properties and Applications of Dielectric Materials, Proceedings. Second International Conference on Properties and Applications, pp vol.1, Beijing, C. S. Venkataraman, B. R. Prabhakar and M. R. Nandagopal, "Effect of humidity on the power frequency flashover voltage of insulators and gaps at varying temperature conditions," India, IEE-IERE Proceedings -, vol. 14, no. 6, pp , November-December Doi: / iipi

157 REFERENCES Y. Li, J. He, J. Yuan, C. Li, J. Hu and R. Zeng, "Failure Risk of UHV AC Transmission Line Considering the Statistical Characteristics of Switching Overvoltage Waveshape," in IEEE Transactions on Power Delivery, vol. 28, no. 3, pp , July R. R. Nunes and W. d. C. Boaventura, "Insulation Coordination Considering the Switching Overvoltage Waveshape Part I: Methodology," IEEE Transactions on Power Delivery, vol. 24, no. 4, pp , Oct R. R. Nunes and W. d. C. Boaventura, "Insulation Coordination Considering the Switching Overvoltage Waveshape Part II: Application and Results," IEEE Transactions on Power Delivery, vol. 24, no. 4, pp , Oct ICNIRP GUIDELINES for limiting exposure to time - varying electric, magnetic and electromagnetic fields (up to 300GHz), HEALTH PHYSICS 74 (4): ; Live-Line Work Using Hot sticks s, Available at: Live-Line Work Bare Hand or Potential Method, Available at: Single pick robotic arm, Available at: Single pick robotic arm, Available at: PowerPoints/2015/DSI, Robots for Energised Transmission LineWork.pdf 2.1 Live Line Working, Minimum Safe Working Distances on 275 and 400kV NGC System. Technical Guidance Note TGN (T) 54, Issue 1, September A. Villalobos and P. A. Calva, "Live line working distances in reduced air densities," Electrical Insulation and Dielectric Phenomena, 2000 Annual Report Conference on, pp vol.1, Victoria, BC, 2000.

158 REFERENCES The minimum approach distances deployed by OSHA Available at: IEEE Guide for Maintenance Methods on Energized Power Lines," IEEE Std , vol. no. pp.0_1, Martinez, J. A., Natarajan, R., & Camm, E. (2000). Comparison of statistical switching results using Gaussian, uniform and systematic switching approaches. Power Engineering Society Summer Meeting. IEEE (Vol. 2, pp ). IEEE, CIGRE Working Group 07(Dielectric strength of external insulation under transient voltage) of Study Committee 33. (Overvoltages and Insulation Coordination), IEC :2010. High-voltage test techniques - Part 1: General definitions and test requirements. 2.8 IEC 61472:2013- Edition 3.0. Live working - Minimum approach distances for a.c. systems in the voltage range 72.5 kv to 800 kv - A method of calculation. 2.9 G. H. Vaillancourt, S. Carignan and C. Jean, "Experience with the detection of faulty composite insulators on high-voltage power lines by the electric field measurement method," IEEE Transactions on Power Delivery, vol. 13, no. 2, pp , Apr G. Göcsei, B. Németh and I. Berta, "Brittle fracture of high voltage composite insulators-a new method in inspection and detection," Electrical Insulation Conference (EIC), 2014, pp , Philadelphia, PA, Y. Watanabe, "Switching Surge Flashover Characteristics of Extremely Long Air Gaps," in IEEE Transactions on Power Apparatus and Systems, vol. PAS- 86, no. 8, pp , Aug doi: /TPAS

159 REFERENCES L. Paris, "Influence of Air Gap Characteristics on Line-to-Ground Switching Surge Strength," IEEE Transactions on Power Apparatus and Systems, vol. PAS-86, no. 8, pp , Aug doi: /TPAS K. J. Lloyd and L. E. Zaffanella, Insulation for switching surges, in Transmission Line Reference Book 345 kv and Above, J. J. LaForest, Ed., 2nd ed. Palo Alto, CA: EPRI, pp , G. Carrara, L. Dellera and G. Sartorio, "Switching Surges with Very Long Fronts (above 1500 μs): Effect of Front Shape on Discharge Voltage," IEEE Transactions on Power Apparatus and Systems, vol. PAS-89, no. 3, pp , March doi: /TPAS T. Harada, Y. Aihara and Y. Aoshima, "Influence of Switching Impulse Wave Shape on Flashover Voltages of Air Gaps," IEEE Transactions on Power Apparatus and Systems, vol. PAS-92, no. 3, pp , May doi: /TPAS C. Menemenlis and K. Isaksson, "The Front Shape of Switching Impulses and its Effect on Breakdown Parameters," IEEE Transactions on Power Apparatus and Systems, vol. PAS-93, no. 5, pp , Sept doi: /TPAS C. Menemenlis and K. Isaksson, "Influence of the various parts of the switching impulse front on discharge development," IEEE Transactions on Power Apparatus and Systems, vol. 94, no. 5, pp , Sept doi: /T-PAS B. Hutzler and J. Lalot, "Influence of Non-Standard Switching Impulses on the Flashover Mechanisms of Air Gap," IEEE Transactions on Power Apparatus and Systems, vol. PAS-97, no. 3, pp , May doi: /TPAS

160 REFERENCES C. Menemenlis, G. Harbec and J. F. Grenon, "Behavior of Air Insulating GAPS Stressed by Switching Overvoltages with a Double Peak," IEEE Transactions on Power Apparatus and Systems, vol. PAS-97, no. 6, pp , Nov doi: /TPAS IEEE Guide for Maintenance Methods on Energized Power Lines," IEEE Std , vol., no., pp.1-144, June doi: /IEEESTD D. A. Gillies, D. M. Wallis, G. Gela, P. W. Hotte and M. Charest, "Discussion of "IEC method of calculation of minimum approach distances for live working" and closure," IEEE Transactions on Power Delivery, vol. 16, no. 4, pp , Oct doi: / J.C. Das. Book: Transients in Electrical Systems: Analysis, Recognition, and Mitigation. McGraw Hill Professional, ISBN: , 2010, 736 pages General Electric Company CRIEPI, Lenox, Massachusetts, Air gap Sparkover and Gap Factors, EPRI, Tech. Rep TR , Project 3787, Final Report, December V. D. Sluis, Book: Transients in Power Systems. The University of Michigan: John Wiley & Sons Ltd, ISBN: , , pp, U.A. Bakshi, A.V. Bakshi. Book: "Electromagnetic Waves and Transmission Lines. Technical Publications, ISBN: , Pp, C. M. Philip, A. Weisshaar, V. K. Tripathi, G. C. Alexander. Book: Transmission Lines and Wave Propagation., Fourth Edition. CRC Press, ISBN: , Pp,

161 REFERENCES Y. Hase. Book: Handbook of Power System Engineering. P John Wiley & Sons, 2 nd edition (December 26, 2012), ISBN: , pages Cable trench Available at: Access time: 20/11/2014 at Cable trench Available at: A. Greenwood. Book: Electrical Transients in Power Systems. The University of Michigan. Wiley, ISBN: , nd ed J. A. Martinez, J. Mahseredjian and R. A. Walling, "Parameter determination: procedures for modelling system transients," in IEEE Power and Energy Magazine, vol. 3, no. 5, pp , Sept.-Oct.. doi: /MPAE Juan A. Martinez-Velasco., ed. Book: Transient Analysis of Power Systems: Solution Techniques, Tools and Applications. John Wiley & Sons, ISBN: , Ryan, H. M. High Voltage Engineering and Testing. IET (Institution of Electrical and Electronic Engineering), London, ISBN: , rd edition J. P. Bickford and P. S. Doepel, "Calculation of switching transients with particular reference to line energisation," in Electrical Engineers, Proceedings of the Institution of, vol. 114, no. 4, pp , April doi: /piee

162 REFERENCES M. Haddad and D. Warne (Eds.). Book: Advanced in High Voltage Engineering. IET (Institution of Electrical Engineers), ISBN: , Pp, Juan A. Martinez-Velasco. Book: Power System Transients: Parameter Determination. Boca Raton: CRC/Taylor & Francis Group, ISBN: P G. Gallet, G. Leroy, R. Lacey and I. Kromer, "General expression for positive switching impulse strength valid up to extra-long air gaps," in IEEE Transactions on Power Apparatus and Systems, vol. 94, no. 6, pp , Nov doi: /T-PAS G. Gallet, G. Leroy. Expression for switching impulse Strength Suggestion a Highest Permissible Voltgae for A.C System, IEEE Conference paperc73-408, IEEE Conference paper C73-408, IEEE Transactions on power apparatus and systems, Vl. PAS-92 No. 6,No./Dec P General Electric Company CRIEPI, Lenox, Massachusetts, Air gap Sparkover and Gap Factors, EPRI, Tech. Rep TR , Project 3787, Final Report, December Binnes, J.K. Lawrernson, P.J. Book: Analysis and computation of electric and magnetic field problems. Published by Elsevier Ltd. ISBN: X, nd edition Electrical Performance of a Portable Protective Gap (PPG) in a Compact 550- kv tower. Final Report, TR , Electric Power Research Institute, Palo Alto, California, February G. Gallet, G. Leroy, R. Lacey and I. Kromer, "General expression for positive switching impulse strength valid up to extra-long air gaps," IEEE Transactions

163 REFERENCES 163 on Power Apparatus and Systems, vol. 94, no. 6, pp , Nov doi: /T-PAS Electrical Performance of a Portable Protective Gap (PPG) in a Compact 550- kv tower. Final Report, TR , Electric Power Research Institute, Palo Alto, California, February General Electric Company CRIEPI, Lenox, Massachusetts, Air gap Sparkover and Gap Factors, EPRI, Tech. Rep TR , Project 3787, Final Report, December W. Busch, "Air Humidity: An Important Factor for UHV Design," in IEEE Transactions on Power Apparatus and Systems, vol. PAS-97, no. 6, pp , Nov doi: /TPAS Cotton, I. Allen, N. Crossley, P. Wang, Z. March An introduction to insulation Co-ordination. Technical review report. 3.24Juan A. Martinez-Velasco. Book: Power System Transients: Parameter Determination. Boca Raton: CRC/Taylor & Francis Group, ISBN: pp Page N. H. Malik, A. A. Al-Arainy, and M. I. Qureshi. Book: Electrical insulation in power systems. CRC Press.ISBN: , F. A. M. Rizk, "A model for switching impulse leader inception and breakdown of long air-gaps," IEEE Transactions on Power Delivery, vol. 4, no. 1, pp , Jan doi: / W. Diesendorf. Book: "Insulation Coordination in High Voltage Power Systems." Elsevier. ISBN: , , pages PSCAD v4.2.1 (Aug ) Insulation co-ordination Part 1, Definitions, principles and rules, IEC ; Third edition , Brussels, CENELEC, 1997.

164 REFERENCES National Grid Seven years Statement Available at: Access time: 8/10/2012 at 11: Insulation co-ordination Part 1, Definitions, principles and rules, IEC ; Third edition , Brussels, CENELEC, H. Rinne. Book: The Weibull Distribution: A Handbook. November 20, Chapman and Hall/CRC, Reference - ISBN Pages B/W Illustrations. Page Johnson, N. L., Kotz, S., and Balakrishnan, N. Continuous Univariate Distributions. New York: Wiley, Page J. A. Martinez, et al., "Parameter determination: procedures for modelling system transients, "Power and Energy Magazine, IEEE, vol. 3, pp , BS EN :2004, Series capacitors for power systems. General, Status: Revised, Withdrawn, Published: September 2005 Replaced By: BS EN : IEEE Draft Guide for the Application of Shunt Power Capacitors Prepared by the Shunt Capacitor Application Guide Working Group of the Capacitor Subcommittee of the IEEE Transmission & Distribution Committee," IEEE Std P1036/D13a, vol 142., pp 18., G. Gopakumar, H. Yan, Dr. A. Bruce. Mork, & Kalyan K. Mustaphi. Shunt Capacitor Bank Switching Transients: A Tutorial and Case Study. Retrieved from http: mip99. Pdf 4.8 J. A. Martinez and J. Martin-Arnedo, "Voltage sag analysis using an Electromagnetic Transients Program," Power Engineering Society Winter Meeting, IEEE, 2002, pp vol.2. doi: /PESW

165 REFERENCES Risk & Hazard Explanation available at: Visited (03/04/2016). 5.2 V. Lovrencic and B. Gomišcek, "Live working as an example of electrical installation maintenance with the zero accidents philosophy," Live Maintenance (ICOLIM), th International Conference on, pp doi: /ICOLIM , Budapest, T. Suzuki, I. Kishijima, Y. Ohuch and K. Anjo, "Parallel Multigap Flashover Probability," IEEE Transactions on Power Apparatus and Systems, vol. PAS- 88, no. 12, pp , Dec doi: /TPAS A. R. Hileman. Insulation coordination for power systems. Technology and Engineering, CRC Press, 15 June General Electric Company CRIEPI, Lenox, Massachusetts, Air gap Sparkover and Gap Factors, EPRI, Tech. Rep TR , Project 3787, Final Report, December 1994.

166 Appendix Appendices Appendix 1: The standard proposes a range of voltages to construct a Table in which there is a range of values for k a which varies with altitude 0 m m. With the use of the Equations in Chapter 2 and shown below, Table 56 can be constructed in which for each value of altitude and range of voltages a value of k a exists: g O UO = Ø kg -1ø 500d Œ1+ œ º 3 ß U O = Voltage breakdownunder standard conditions (kv, at a temperature of 20 o C, at a pressure of 101.3kPA and at a humidity of 11g/m 3 ) g O = Undefined factor used in the calculation of k a 1. Calculate: go T = g O ( k ) 1.6 g T = Undefined factor used in the calculation of k a 2. Calculate: ( H p ) d = = e Ł po ł δ = Relative air density p = Actual pressure (kpa) H = Height above sea level (m) This makes an assumption that δ is affected by mean pressure only and not by temperature.

167 Appendix Finally, calculate k a by: ( ) ( d ) ø( d go ) ( 1-0.2g ) 0.8 Ø k = U / U = º ß a O U 90 Voltage (kv) Altitude (m) < > Table 7-1: Atmospheric Factor k a for Different Reference Altitudes and Values of U 90 _ (IEC 61472) The average of k a can be assumed to be according to Table 7-2 as below: O Altitude (m) k a average Table 7-2: Average k a Values IEC Table 7-3 presents a simplified criterion for the k f determination independence of β and L f. The k f values are derived from the interpolation of the data shown in Annex F of IEC Table 7-3 contains the values of β in function of the original gap length L f rather than in function of the remaining air gap length D because the original gap length L f is one of the important quantities that characterise the constructed a.c. system.

168 Appendix 168 Table 7-3: Floating Conductive Object Factor k f

169 Appendix 169 Appendix 2: Below Table presents the conductor coordinates for L2, L6, L8, L9 and L12 used by National Grid UK. These coordinates are used for simulation and construction of models throughout this thesis. Table 7-4: Conductor Coordinates (Including Sag) for Overhead Line Designs [2.1] Figure 7-1: Conductor Coordinates of Overhead Line- Refer to Table 59

170 Appendix 170 Circuit # Cond. # Connection X (from Y GW. # Connection X (from Phasing # tower centre) (at tower) Phasing # tower centre) Eliminated 0.0 [m] Tower: L2 Tower Conductors: Quad Zebra Tower Centre 0.0 [m] Table 7-5: PSCAD Configuration of L2 Tower Ground_Wires: Zebra Circuit # Cond. # Connection X (from Y GW. # Connection X (from Phasing # tower centre) (at tower) Phasing # tower centre) Eliminated 0.0 [m] [m] Tower: L6Tower Conductors: Quad Zebra Tower Centre 0.0 [m] Ground_Wires: Zebra Y (at tower) [m] Y (at tower) [m] Table 7-6: PSCAD Configuration of L6 Tower Circuit # Cond. # Connection X (from Y GW. # Connection X (from Phasing # tower centre) (at tower) Phasing # tower centre) Eliminated 0.0 [m] Tower: L2 Tower Conductors: Quad Zebra Tower Centre 0.0 [m] Ground_Wires: Zebra Y (at tower) [m] Table 7-7: PSCAD Configuration of L8 Tower Circuit # Cond. # Connection X (from Y GW. # Connection X (from Phasing # tower centre) (at tower) Phasing # tower centre) Eliminated 0.0 [m] Tower: L9 -Tower Conductors: Quad Zebra Tower Centre 0.0 [m] Ground_Wires: Zebra Y (at tower) [m] Table 7-8: PSCAD Configuration of L9 Tower

171 Appendix 171 Circuit # Cond. # Connection X (from Y GW. # Connection X (from Phasing # tower centre) (at tower) Phasing # tower centre) Eliminated 0.0 [m] Tower: L12 Tower Conductors: Quad Zebra Tower Centre 0.0 [m] Table 7-9: PSCAD Configuration of L12 Tower Ground_Wires: Zebra Y (at tower) [m] BTIME 0.05 F + + D Fault Timed Fault Logic F1 F2 F3 F4 F5 F6 L2PE1 L2PP1 1 Ch. 1 Ch. 2 Meas-Enab.. V1. V2 FLTTime FLTLoc Channel Decoder Select Data Fault L2PE2 L2PP2 Ch. 3 Ch. 4 Multiple Run V3 FType FType Select Data 6 Channel Decoder FT1 FT2 FT3 FT4 FT5 FT6 Figure 7-2: PSCAD Fault Type and Time Selection Modules L21A L22A L23A L24A L25A L21B 1 L12_1 1 L22B 1 L12_2 L23B L12_3 L24B L12_ L25B 1 L21C L21D L12_1 L12_1 L22D L22C L12_2 L12_2 L23C L23D L23E L12_3 L12_3 L24C L24D L24E L12_4 L12_4 L25C L25E L25D L21E L22E L21F L22F L23F L24F L25F Figure 7-3: PSCAD Overhead Line Model

172 Appendix 172 L22A L22B L22C E F G L21C D Max C B G E F L25B L25C L21B L21A Max L23B L23A B C C B A L23C L25A D Max G F E L24B L24A L2PE1 L24C L22F L22D L22E E F G L21E D Max C B G E F L25B L25C L21D L21F Max L23E L23F B C C B A L23D L25A D Max G F E L24E L24F L2PE2 L24D Figure 7-4: P-E Calculation Design L21A L21B L21C D + - F L21B D + - F L21C D + - F L21A C D E Max F L21PP1 L22A L22B L22C D + - F L22B D + - F L22C D + - F L22A C D E Max F L22PP1 L23A L23B L23C D + - F L23B D + - F L23C D + - F L23A C D E Max F L23PP1 L24A L24B L24C D + - F L24B D + - F L24C D + - F L24A C D E Max F L24PP1 L25A L25B L25C D + - F L25B D + - F L25C D + - F L25A C D Max E F L25PP1 L25PP1 L21PP1 L22PP1 B C Max D E F L24PP1L23PP1 L2PP1 L22PP2 L24PP2 L23PP2 L25PP2 B C Max D E F L21PP2 L2PP2 Figure 7-5: P-P Calculation Modules

173 Appendix 173 Appendix 3: The study result of the fundamental model of a transmission line network in Chapter 4 is shown by following Tables. These Tables are presenting the maximum U 2 Overvoltages due to the simulation of fault and clearance at sea level, 500m and 1000m altitudes. The following Tables also present the result of simulation with and without floating object with a length of 2m. Case 3A: Without Floating Object Tower Type L2 L6 L8 L9 L12 Fault Voltage (kv) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-10: Overvoltage Simulation Results for Fault and Clearance Tower Type L2 L6 L8 L9 L12 Fault Voltage (kv) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-11: Overvoltage Simulation Results for Fault and Clearance, Inductive Compensation

174 Appendix 174 Tower Type L2 L6 L8 L9 L12 Fault Voltage (kv) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-12: Overvoltage Simulation Results for Fault and Clearance, Capacitive Compensation Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-13: Minimum Approach Distance for Fault and Clearance at Sea Level Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-14: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at Sea Level

175 Appendix 175 Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-15: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at Sea Level Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-16: Minimum Approach Distance for Fault and Clearance at 500m Altitude Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-17: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at 500m Altitude

176 Appendix 176 Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-18: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 500m Altitude Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-19: Minimum Approach Distance for Fault and Clearance at 1000m Altitude Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-20: Minimum Approach Distance for Fault and Clearance, Inductive compensation at 1000m altitude

177 Appendix 177 Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-21: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 1000m Altitude Case 3B: Without Floating Object & Weighted Distribution of Fault Type Tower Type L2 L6 L8 L9 L12 Fault Voltage (kv) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-22: Overvoltage Simulation Results for Fault and Clearance & Weighted Fault Type

178 Appendix 178 Tower Type L2 L6 L8 L9 L12 Fault Voltage (kv) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-23: Overvoltage Simulation Results for Fault and Clearance, Inductive Compensation & Weighted Fault Type Tower Type L2 L6 L8 L9 L12 Fault Voltage (kv) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-24: Overvoltage Simulation Results for Fault and Clearance, Capacitive Compensation & Weighted Fault Type Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-25: Minimum Approach Distance for Fault and Clearance at Sea Level & Weighted Fault Type

179 Appendix 179 Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-26: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at Sea Level & Weighted Fault Type Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-27: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at Sea Level & Weighted Fault Type Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-28: Minimum Approach Distance for Fault and Clearance at 500m Altitude & Weighted Fault Type

180 Appendix 180 Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-29: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at 500m Altitude & Weighted Fault Type Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-30: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 500m Altitude & Weighted Fault Type Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-31: Minimum Approach Distance for Fault and Clearance at 1000m Altitude & Weighted Fault Type

181 Appendix 181 Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-32: Minimum Approach Distance for Fault and Clearance, Inductive compensation at 1000m altitude & Weighted Fault Type Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-33: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 1000m Altitude & Weighted Fault Type

182 Appendix 182 Case 3C: With Floating Object Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-34: Minimum Approach Distance for Fault and Clearance at Sea Level with Floating Object of 2m Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-35: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at Sea Level with Floating Object of 2m

183 Appendix 183 Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-36: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at Sea Level with Floating Object of 2m Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-37: Minimum Approach Distance for Fault and Clearance at 500m Altitude with Floating Object of 2m Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-38: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at 500m Altitude with Floating Object of 2m

184 Appendix 184 Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-39: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 500m Altitude with floating object of 2m Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-40: Minimum Approach Distance for Fault and Clearance at 1000m Altitude with Floating Object of 2m Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-41: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at 1000m Altitude with Floating Object of 2m

185 Appendix 185 Tower Type L2 L6 L8 L9 L12 Fault Current (ka) Minimum Approach Distance (m) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-42: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 1000m Altitude with Floating Object of 2m With Floating Object and Weighted Fault Type Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-43: Minimum Approach Distance for Fault and Clearance at Sea Level with Floating Object of 2m (Weighted Fault Type)

186 Appendix 186 Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (M) Current (Ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-44: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at Sea Level With Floating Object Of 2m (Weighted Fault Type) Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-45: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at Sea Level with Floating Object of 2m (Weighted Fault Type) Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-46: Minimum Approach Distance for Fault and Clearance at 500m Altitude with Floating Object of 2m (Weighted Fault Type)

187 Appendix 187 Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-47: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at 500m Altitude with Floating Object of 2m (Weighted Fault Type) Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-48: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 500m Altitude with Floating Object of 2m (Weighted Fault Type) Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-49: Minimum Approach Distance for Fault and Clearance at 1000m Altitude with Floating Object of 2m (Weighted Fault Type)

188 Appendix 188 Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-50: Minimum Approach Distance for Fault and Clearance, Inductive Compensation at 1000m Altitude with Floating Object of 2m (Weighted Fault Type) Tower Type L2 L6 L8 L9 L12 Fault Minimum Approach Distance (m) Current (ka) P-E P-P P-E P-P P-E P-P P-E P-P P-E P-P Table 7-51: Minimum Approach Distance for Fault and Clearance, Capacitive Compensation at 1000m Altitude with Floating Object of 2m (Weighted Fault Type) Appendix 4 Table 7-53 is extracted from Table 5-1 and Figure 5-2 of CRIEPI [2.23] and it contains U 50 values for Rod to Plane gap with respect to their time-to-crest. Most of the Table data are switching type impulses (time to crest greater than 100µsec). The Equation for calculation of U 50 suggested by CRIPEI is extracted from Table 5-1 of [2.23], and it has been presented in Table 7-53.

189 Appendix 189

190 Appendix 190

191 Appendix 191

192 Appendix 192 Table 7-52: Rod to Plane Gap Experimental Sparkover Data, Positive polarity (Continue), CRIEPI_ Table 5-1 [2.23] Y: Critical Wave - : Not a critical Wave T: Data from Table TF: Data from Figure Figure 7-6 presents the data from Table 7-53 where the U 50 voltages have been plotted against the gap sizes.

193 Appendix 193 Figure 7-6: Rod to Plane Sparkover versus Gap Length D, CRIEPI_ Figure 5-2 [2.23] Appendix 5 The extracted data to produce the suggested Equations in Chapter 5 used the data shown in Figure 7-7. The U 50 of the gap has been extracted from the gap size and time to crest of the transient wave. As a result, Table 7-54 has been produced where the presented U 50 s are the product of the various time to crest and gap sizes.

194 Appendix 194 Figure 7-7: Switching Impulse Flashover Voltage of Rod-Plane Gap, Estimation of CRIEPI s Equation