SERIES I M P E D A N C E A N D SHUNT A D M I T T A N C E MATRICES OF A N UNDERGROUND CABLE SYSTEM by Navaratnam Srivallipuranandan B.E.(Hons.), U n i v e r s i t y of M a d r a s , India, 1983 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F T H E REQUIREMENTS FOR T H E DEGREE OF M A S T E R OF APPLIED SCIENCE in T H E F A C U L T Y O F G R A D U A T E STUDIES (Department of Electrical Engineering) We accept this thesis as conforming to the required standard T H E UNIVERSITY O F BRITISH C O L U M B I A , C Navaratnam Srivallipuranandan, 1986 November 1986 1986 In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of requirements f o r an advanced degree a t the the University o f B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make it f r e e l y a v a i l a b l e f o r reference and study. I further agree t h a t p e r m i s s i o n f o r e x t e n s i v e copying o f t h i s t h e s i s f o r s c h o l a r l y purposes may department o r by h i s o r her be granted by the head o f representatives. my It is understood t h a t copying or p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l gain s h a l l not be allowed without my permission. Department o f The U n i v e r s i t y o f B r i t i s h Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 Date 6 n/8'i} written SERIES I M P E D A N C E A N D SHUNT ADMITTANCE OF A N UNDERGROUND MATRICES CABLE ABSTRACT This thesis describes numerical impedance matrix methods for the: evaluation of the series and shunt admittance matrix of underground cable systems. In the series impedance matrix, the terms most difficult to compute are the internal impedances of tubular conductors and the earth return impedance. The various form u hit- for the interim!' impedance of tubular conductors and for th.: earth return impedance are, therefore, investigated in detail. evaluating the elements of the Also, a more accurate admittance matrix with way of frequency dependence of the complex permittivity is proposed. Various impedance standard formulae of buried have cables. been developed Using for the the Polhiczek's earth return formulae as the for comparison, the formula of Ametani and approximations proposed by other authors are studied. Mutual impedance between an underground cable and an overhead conductor is studied as well. The internal impedance of a laminated tubular conductor is different from that conductor. of a homogeneous tubular Equations have derived to evaluate the internal impedances of such laminated conductors. (ii) been tubular Table of Contents Abstract — - - Table of Contents l i - List of Table iij •• - List of Figures - - V VI -:— List of Symbols Acknowledgement - - Viii , - 1. INTRODUCTION 2. SERIES I M P E D A N C E A N D S H U N T A D M I T T A N C E M A T R I C E S •- .—.:. ix 1 2.1 2.2 Basic Assumptions Series Impedance matrix [Z] for N Cables in Parallel 2.2.1 Submatrix [Z„] 2.2.2 Skin Effect 2.2.3 Internal Impedance of Solid and Tubular Conductors 2.2.4 Submatrix [Z,-yJ 2.2.4.1 Proximity Effect 2.2.4.2 Proximity Effect of a Single-Phase Circuit of Two Identical Conductors 2.2.4.3 Shielding Effect of the Sheath 2.2.4.4 Elements of Submatrix [Z„] 2.3 Shunt Admittance Matrix [K]; for N Cables in Parallel 2.3.1 Leakage Conductance and Capacitive Suceptance 2.3.2 Frequency Dependence of the Complex Permittivity 2.3.3 Submatrix 2.4 Conclusion •• 6 7 9 13 14 15 16 17 17 19 22 22 23 27 29 3. C O M P A R I S O N O F I N T E R N A L I M P E D A N C E F O R M U L A E 3.1 3.2 3.3 3.4 Exact Formulae for Tubular Conductors Internal Impedance of a Solid Conductor Internal impedance of a Tubular Conductor Conclusion — (iii) -•..— ; -•- 30 31 36 47 4. E A R T H 4.1 RETURN IMPEDANCE Earth Return Impedance of Insulated Conductor 50 4.2 Earth Return Impedance in a Homogeneous Infinite Earth 50 4.3 Earth Return Impedance in a Homogeneous Semi-Infinite Earth 52 4.4 Formulae used by Ametani, Wedepohl and Semlyen 57 4.5 Effect of Displacement Current Term and Numerical Results 61 4.6 Cable Buried at Depth Greater than Depth of Penetration 64 4.7 Mutual Impedance between a Cable Buried in the Earth and an Overhead Line or vice versa 67 4.8 Conclusion 69 5. L A M I N A T E D 5.1 5.2 CONDUCTORS 70 5.1:1 Internal Impedance with External Return 70 5.1.2 Internal Impedance with Internal Return 72 Application to Gas-Insulated Substations Case i: Core and Sheath not Coated 73 74 5.2.2 Case ii: Only Sheath Coated 74 5.2.3 Case iii: Only Core Coated 76 5.2.4 Case iv: Both Core and Sheath Coated 5.2.5 Stainless Steel Coating 79 5.2.6 Supermalloy Coating 82 5.2.7 Comparison between Stainless Steel and Supermalloy Coatings Conclusion 6. T E S T 6.1 6.2 6.3 TUBULAR Internal Impedance of a Laminated Tubular Conductor 5.2.1 5.3 - 77 82 85 CASES Single-Core Cable Three-Phase Cable Shunt Admittance Matrix 86 91 93 7. C O N C L U S I O N 94 APPENDIX A 96 APPENDIX B : 98 APPENDIX C 103 APPENDIX D 106 REFERENCES : 114 (Iv); List of Tables 3.1 3.2 3.3 3.4 Internal Impedance Internal Impedance Mutual Impedance ing Inside Internal Impedance Outside 4.1 4.2 4.3 5.1 5.2 5.3 6.1 6.2 of a Solid Conductor Z of a Tubular Conductor (Z (,) of a Tubular Conductor with Current Return- 33 38 a a 43 Zy of a Tubular Conductor with Current Returning -. , Solution of PoIIaczek's Equation by Numerical Integration and Using Infinite Series Earth Return Self Impedance with and without Displacement Current Term .' 1 Earth Return Self Impedance as a Function of Frequency Resistivity and Relative Permeability of Coating Materials Skin Depth of Stainless Steel Skin Depth of Supermalloy Impedances of Single Core Underground Cable Mutual Impedance between Two Cables with Burial Depth of 0.75m and Separation of 0.30m (V) 44 58 61 64 '• 79 79 82 87 91 List of Figures 1.1 2.1 2.2 2.3 2.4 2.5 2.6 2.~! 2.8 2.9 2.10 2.11 Potential Difference V, between Core and Sheath and F between 2 Sheath and Earth Basic Single Core Cable Construction Loop Currents in a Single Core Cable Potential Difference between Two Concentric Conductors Three Conductor Representation of a Single Core Cable Sheath with Loop Currents I and I ... •Two Cable System , -. Circuit Arrangement of Primary, Secondary and Shielding Conductors, with Shielding Conductor Grounded at Both Ends".. Transmission System Consisting of a Single Conductor and a Cable ; x 2 Cross-Section of a Coaxial Cable (a),(b) - Measurements of e'(<o) and ("(<*) of an OiMmpregnated Test Cable at 20°cC Values of e'(o>) and «"((•)) Obtained from the Empirical Formula : 3 7 9 10 10 15 16 18 21 23 24 26 2.12 Polarization-Time Curve of a Dielectric Material 27 3.0 3.1 Loop Currents in a Tubular Conductor (a),(b) - Impedance of a Solid Conductor as a Function of Frequency 30 3.2 (a),(b) - Errors in Wedepohl's and Semlyen's Formulae for a Solid Conductor 3.3 3.4 3.5 35 Cross-Section of a Tubular Conductor (a),(b) - Impedance Z of a Tubular Conductor (with Internal Return): as a Function of Freqency Errors in Wedepohl's, Schelkunoff's and Bianchi's Formulae for Z 30 a 39 a 40 3.6 3.7 3.8 Errors in Wedepohl's and Schelkunoff's Formulae for Z (a),(b) - Impedance Z of Tubular Conductor (with External Return) as a Function of Frequency ..' Errors in Wedepohl's, Schelkunoff's and Bianchi's Formulae for Z 4.3 4.4 45 b „ :.. 4.1 4.2 42 ab b , Electric Field Strength at Point P Error in Replacing a Conductor of Finite Radius by a Filament Conductor —-Solution of Real and Imaginary Part of Equation (4.9), for a Freqency of 1MHz. > - «... Relative Error in the Evaluation of Carson's Formulae with an Asymptotic Expansion (vi) 46 51 53 56 59 4.5 Error in the Earth Return Self-Impedance if the Displacement Current is Ignored 4.6 (a),(b) - Earth Return Self Impedance as a Function of Freqency 4.7 Errors in Earth Return Self Impedance 60 4.8 68 5.1 Differences in Resistance Values of Semi-Infinite and Infinite Earth Return Formulae : (a),(b) - Numbering of Conductor Layers to Find the Internal Impedances of a Laminated Tubular Conductor 71 5.2 Representation of the nt th Layer 71 5.3 Core and Sheath not coated 74 5.4 Inner Surface of the Sheath only, Coated 75 5.5 Core Alone Coated 76 5.6 Core as well as Inner Surface of the Sheath Coated 5.7 Dimensions of the Bus Duct in a Gas-Insulated Substation 5.8 (a),(b) - Variation of Resistance; and Inductance with Frequency for the Four Cases; Stainless Steel Coating, Thickness, 0.1mm (a),(b) - Variation of Resistance and Inductance with Freqency for the Four Cases; Stainless Steel Coating, Thickness 0.5mm (a),(b) - Variation of Resistance and Inductance with Freqency for the Four Cases; Supermalloy Coating, Thickness 0.01mm , 5.9 5.10 5.11 65 77 the Four Cases; Supermalloy Coating, Thickness 0.05mm 6.2 Errors in Ametani's and Wedepohl's Approximations in Z B. l 80 81 83 84 Errors in Ametani's and Wedepohl's Approximations in Z A.l A. 2 < • 78 (a),(b) - Variation of Resistance and Inductance with Frequency for 6.1 6.3 6.4 62 cc gc Errors in Ametani's and Wedepohl's Approximations in Z Errors in Ametani's and Wedepohl's Approximations in the Mutual Impedance between Two Cables Three-Phase Cable Set-up for the Study Basic Construction of Each Single Core Cable ef 88 89 90 92 96 96 The Relative Directions of the Field Components in a Coaxial Transmission Line 98 B. 2 Loop Currents in a Tubular Conductor 101 C. 1 Representation of a Buried Conductor in an Infinite Earth 104 D. l Current Carrying Filament in the A i r 106 D.2 Current Carrying Filament Buried in the Earth (Vii) I l l LIST O F E — electric field strength, H = magnetic field strength frequency, SYMBOLS / = IT = 3.1415926 (o = 27rX /, angular frequency fi = 4JTX 1 0 , absolute permeability of free space /x, — relative permeability of the medium i t*i ~ / o / r i > total permeability of the medium « 7 = p, = resistivity of a particular medium t — 1, complex operator 0 r - 7 i €,((o) = x i Euler's constant €, — complex dielectric constant or permittivity of a particular medium » <j> = / = current J — current density flux density c,exp = exponential In = natural logarithm m, — ( joy/ty |:— 1 o> fi^t- | , known as intrinsic propagation constant of a particu- lar medium t. If the displacement currents are ignored, then the ( joi / i , I value of m is equal to I I . Displacement currents are ignored unless li Pi J otherwise specified /„ = modified Bessel function of the 1st kind and of the n t h order K — modified Bessel function of the 2nd kind and of the n t h order K — characteristic impedance F = propagation constant T = relaxation time of a dielectric R = resistance per unit length L = inductance per unit length n G = conductance per unit length C — capacitance per unit length Y = shunt admittance Z = series impedance matrix matrix (viii) Acknowledgements I would like to acknowledge my appreciation to Dr. H. W. Dommel for his encouragement and supervision throughout the course of this research. I would also like to thank Mrs. Guangqi L i , Mr. Luis Marti and Mr. C. E. Sudhakar for their many valuable suggestions. Thanks are due to Mrs. Nancy Simpson for typesetting the manuscript. The financial support of Bonneville Power Administration, U.S.A., and from my brother "Anna" is gratefully acknowledged. (ix) Portland, Oregon, -1- 1. INTRODUCTION Underground cables are used extensively for the transmission and distribution of electric power. Although expensive when compared with overhead transmission, laying cables underground is often the only choice in urban areas. As the urban areas expand, the cable circuits tend to increase in length. At present, cable circuits are being employed which have lengths of the order of 100km. With an increase in system lengths and higher system voltages, induction effects on nearby communication circuits are becoming more important. the Also, for the power system itself, the steady-state and transient behaviour of underground cables must be known. For interference studies as well as for power system studies, methods for finding cable parameters over a range of frequencies are, therefore, needed. The transmission characteristics of an underground cable circuit or submarine cable circuit are determined by their propagation constant T, and characteristic impedance /<", which may be calculated for an angular frequency w from the following equations F = V(K K Where R,L,G + j<aL)(G + j u C ) (1.1a) = V{R + joiL)/(G + j<aC) (1.1b) and C are the four fundamental line parameters, i.e., resistance, inductance, conductance and capacitance per unit length. These cable parameters are therefore the basic data for all interference and power system studies. Cables are principally classified based on i) their location, i.e., aerial, submarine and underground ii) their protective finish, i.e., metallic (lead, aluminium) or non-metallic (braid) iii) the type of insulation, i.e., oil-impregnated paper, cross linked polyethylene ( X L P E ) etc. iv) the number of conductors, i.e., single conductor, two conductors, three conductors and so on. In this thesis, a single conductor aluminium cable with a concentric lead sheath and with insulation of either oil impregnated paper or X L P E is studied in detail (refer to Appendix A). The series impedance and shunt admittance matrices of a cable system made up of N cables, can be written as - 2 - (Z | 21 (zd 1*8*1 • • • Z = (1.2a) " l*/w] and (nil py 1*2*1 Y = (1.2b) PAWI Submatrix [Z„\ a l o n g t h e d i a g o n a l o f m a t r i x [Z\ is t h e s e l f i m p e d a n c e o f c a b l e i w h i c h c a n be w r i t t e n a s w Z.. sic z.. Z Z„ = z c { s s (1.3) s where = Z. c c s e l f i m p e d a n c e o f t h e c o r e o f c a b l e i" Z . . = Z .= c .7j s s;c = s mutual i m p e d a n c e b e t w e e n the core a n d the s h e a t h of cable » self i m p e d a n c e o f the s h e a t h o f c a b l e i T h e off-diagonal submatrix 2, ; is t h e mutual impedance between the cable i a n d cable j which c a n be w r i t t e n a s (1.4) where Zee = mutual impedance between the core of the cable i a n d core of the cable j Z - mutual impedance between the core of the cable i a n d the sheath of the cable j z.. . = mutual impedance between the sheath of cable i a n d the core of the cable j Z mutual impedance between t h e s h e a t h o f the cable t a n d the s h e a t h of the cable j ci t Si = E v a l u a t i o n o f a l l t h e s e e l e m e n t s ; o f s u b m a t r i c e s [Z„\ a n d \Z \ i s , i n g e n e r a l , n o t e a s y . The tj b e s t a p p r o a c h s e e m s t o be t h e o n e p r o p o s e d b y W e d e p o h l [22] b a s e d o n t h e e a r l i e r w o r k b y S c h e l k u n o f f [6]. B o t h a u t h o r s f i n d t h e s e i m p e d a n c e s f r o m the the drops can core potential and sheath, difference [Z = V, - V/J). These between the longitudinal core and the voltage sheath and longitudinal the done voltage drops be o b t a i n e d potential from difference in the - 3 between the sheath and the earth, as shown in Figure 1.1. The potential differences V 1 and V 2 can be expressed as a function of the loop currents /j and J with the help of Schelkunoff's theorems. (Appendix B) 2 For finding the elements of the shunt admittance matrix, it is a usual practice (22,23,27] to assume the permittivity of the insulation to be a real constant. In reality, the value of the permittivity is a frequency-dependent complex value, the real part of which accounts for the susceptance term and the imaginary part accounts for the conductance term. Therefore, it is necessary to find a general expression for the permittivity as a function of frequency. Chapter 2 discusses, in detail, the following topics: i) Formation of series impedance and shunt admittance matrices, ii) Wedepohl's approach for finding the elements of the series impedance matrix, iii) Frequency-dependence of the permittivity constant iv) Proximity effect and shielding effect in the evaluation of the mutual impedance submatrix Z, r As explained in Appendix B, the potential differences V, and V 2 of loop currents /, and I can be written as 2 can be evaluated in terms using SchelkunofTs theorem. For example, the potential difference V l V, = (Z ctt +Z tHl + Z, ,)/, - Z A s i k m 7 4 - 2 (1.5) where Z c r f = internal impedance of the core with current return outside, Z, — impedance of the insulation between the core and sheath Z tkl = internal impedance of the sheath with current return inside Z sKm — mutual impedance between the loops 1 and 2. ns The formulae developed by Schelkunbff to evaluate these internal impedances, and mutual impedance, Z , tKm Z ,Z , crt sh which take the skin effect into account, are exact and given in terms of modified Bessel functions. These exact expressions are not suitable for hand calculation purposes. There have been several attempts to obtain approximations to these classical formulae in order to make them suitable for hand calculations, (22, 23, 24]. Some of these approximate formulae are compared wjjth the exact formulae of (6] in Chapter 3, in terms of accuracy and computer time. The errors which are caused by neglecting the displacement current are discussed in Chapter 3 as well. Generally, the earth acts as a return path for part of the current in the underground or submarine cable system. The cable parameters are very much influenced by the earth return impedance. These impedances are obtained from the axial electric field strength in the earth due to the return current in the ground. In Chapter 4 the following topics are discussed: (i) E a r t h return impedances of cables buried in an infinite earth, where the depth of penetration of the return current in the ground is smaller than the depth of burial, or in iother words, where the distribution of return current in the ground is circularly symmetrical. (ii) Earth return impedances of cables buried in a semi-infinite earth, where the depth of penetration of the return current in the ground is larger than the depth of burial. (iii) Error introduced in the answers if the displacement current is neglected in the computation. (iv) Approximations proposed by Wedepohl and Semlyen and a comparison of their equations with the classical formula of Pollaczek [lj in terms of accuracy and computer time. "I (v) (vi) Ametani's (27] cable constant routine in the E M T P program. Mutual impedances between a buried conductor and an overhead conductor. In Chapter 5, we turn our attention from cables made up of homogeneous conductors to cables whose core and sheath are made up of laminated conductors of different materials. A practical application of this type of conductor is proposed by Harrington [32]. He suggested - 5 - that transient sheath voltages in gas-insulated substations can be reduced by coating the conductor and sheath surfaces with high-permeability and high-resistivity materials. Formulae for the internal impedances of such laminated conductors are derived and used to show the damping effect as a function of frequency. Chapter 6 concludes this thesis by comparing the values of the cable parameters obtained for a particular three-phase cable system shown in Appendix A, by using the exact formulae of Schelkunoff [6] and Pollaczek [l], as well as Ametani's approach [27] and Wedepohl's approximation [22]. - 6 - 2. SERIES I M P E D A N C E A N D S H U N T A D M I T T A N C E M A T R I C E S 2.1 Basic Assumptions The transmission characteristics of a conducting system such as an underground cable circuit or a submarine cable circuit are determined by its propagation constant T and characteristic impedance K, which can be calculated for the angular frequency a) from the formula T = V(R + {juiL)(G + jcoC) K =; V(/? + jo>L)/(G + jmC) (2.1) where R,L,G and C are the four fundamental line parameters - resistance, inductance, conductance and capacitance, all per unit length. Determination of these parameters in a cable is not easy, but involves rather difficult analysis. Pioneering work in the calculation of underground cable parameters has been done by Wedepohl and Wilcox (22], based on the earlier work of Schelkunoff (6). The first step in defining the electrical parameters of an underground or submarine cable system is to set up the equations which describe the electric and magnetic fields. A complete set of such equations would constitute a perfect mathematical model. If these equations could be solved without any approximations then the response of the model would be indistinguishable from that of the real system, which it represents. In practice, however, this ideal situation cannot be realized. For example, it is not possible to perfectly represent the electrical properties such as resistivity, permeability and permittivity of the earth which form the return path for the currents flowing in the cable. Rigorous representation of such factors would lead to a set of very complex equations which may be very difficult to solve. In practice, therefore, some simplifying assumptions are made. The assumptions made in this thesis are: 1. The cables are of circularly symmetric type. The longitudinal axes of cables which form the transmission system are mutually parallel and also parallel to the surface of the earth. It is implied in this assumption, that the cable has longtitudinal homogeneity. In other words, the electrical constants do not vary along the longtitudinal axes. 2. The change in electric field strength along the bngitodinal axes of the cables are negligible compared to the change in radial electric field strength. This assumption permits the solution of the field equations in two dimensions only. 3 The electric field strength at any point in the earth due to the carrents flowing in a cable is not significantly different from the field that would result if the net current were concentrated in an insulated filament placed at the centre of the cable and the volume of the cable were replaced by the soil. - 74. Displacement currents in the air, conductor and earth can be ignored. Assumptions 3 and 4 are justified up to high frequencies (1MHz], as will be demonstrated later in Chapter 3 and Chapter 4. 2.2 Series I m p e d a n c e M a t r i x [Z\ f o r N C a b l e s in P a r a l l e l Let us assume that the transmission system consists of N cables. Each cable has a cross section of the type shown in Figure 2.1, representative of a typical high voltage (H.V.) underground cable. Figure 2.1 Basic Single Core Cable Construction The core consists of a tubular conductor C with the duct being filled with oil. In the case of solid conductors, the inner radius r would be zero. The insulation between the core and the 0 sheath is usually oil-impregnated paper, surrounded by a metallic tubular sheath 5, and insulation around the sheath. In such systems, there are n = 2N metallic conductors. T h e soil in which the cables arc ; buried constitutes the ( n - f l)th conductor which is chosen as the reference for the conductor voltages. Such a transmission system can be described by the two matrix equations 4^= dx -ZI (2.2a) (2.2b) - 8 d]_ = dx - -IV where'V and / are n-dimensional vectors of voltages and currents, respectively, at a distance z along the longitudinal axis of the cable system. A l l voltages and currents are phasor values at a particular angular frequency a>. The series impedance matrix Z is given by \Zn\ (*IN1 l*«l 1*2=1 \Z \ W Z = (2.3) \ZNN] Each submatrix, \Z \ assembled along the leading diagonal is a square matrix of dimension 2 U representing the self impedances of cable i , by itself, *£.s; *c,c, (2.4) where Z.. = self impedance of the core of cable 1 Z,. . = Z.. = c c s t Z. s = St mutual impedance between core and sheath of cable » e self impedance of the sheath of cable » The off-diagonal submatrix |Z ] represents the mutual impedances between cable »" and cable j. 1; This submatrix is also a square matrix of dimensions 2, l*,l Z - . Z'c - ? • Z' - . Z * . . r r r c s inhere mutual impedance between core of cable t and core of cable j Z.. = £ c Z Ciij mutual impedance between core of cable 1 and sheath of cable j — mutual impedance between sheath of cable 1 and core of cable j Z,.. = t Z. t $i = mutual impedance between sheath of cable 1 and sheath of cable j Similarly, the shunt admittance matrix Y"can be defined as: (2.5) - 9 - Y = (2.6) •where the submatrices jV„j and |Y" ) can be defined in a similar way, as described in section 2.3. y 2.2.1 S u b m a t r i x |Z„J The elements of the submatrix \Z„] can be determined by considering a single cable whose longitudinal cross section is as shown in Figure 2.2. The longitudinal voltage drops in such a cable are best described by two loop equations, with loop 1 formed by the core and sheath (as return) and loop 2 formed by the sheath and earth (as return). insulation 2 sheath insulation 1 core Figure 2.2 Loop Currents in a Single Core Cable. It has been shown by Carson [4] that the change in the potential difference between j and (/ + 1) of a concentric cylindrical system as shown in Figure 2.3 is given by ——• + E — Ej + x } —ju>n4>, where Ej = longitudinal electric field strength of the outer surface of the conductor j E *j = longitudinal electric field strength of the inner surface of conductor (j'+l) ; (27) - 10 - Axis Conductor j Conductor (j+l) Figure 2.S Potential Difference between Two Concentric V. <t>, = Conductors. potential difference between the j and the (j+l) conductor magnetic flux through the area described by the contour ABCD. Since part of the current in the cable can return through the earth, the cable must be represented by 3 conductors (core, sheath, earth), as shown in Figure 2.4. Axis Sheaih Insulation 2 Earth Figure 2.4 Three Conductor Representation of a Single Core Cable. - 11 - The values of longitudinal electric field strengths E ,E ,,E cre sk ihe (i.e., on the external sur- face of the core, internal surface of the sheaih and the external surface of the sheath respectively) can be expressed as E E = Z cre ikt E /„ [rt (2.8a) = -*,»./, + Z I 2 i(2.8b) = Z /, (2.8c) skm ikc ikc I -Z 2 tkm The electric field strength along the surface of the earth can be written as E = -Z I c ei (2.9) 2 where Z crc = internal impedance per unit length of the core's external surface with current returning through a conductor outside the core. Z skt = internal impedance per unit length of the sheath's internal surface with current returning through a conductor inside the tubular sheath. Z skm = mutual impedance per unit length of the sheath which gives the voltage drop on the external surface of the sheath, when current passes through the internal surface or vice versa. Z ike — internal impedance per unit length of the sheath's external surface with current returning through a conductor outside the tubular sheath. Z es = self impedance of the earth's return path. Equation (2.7) can be then be written for the contours A B G D and E F G H a i follows: = E , - E , - joL„h ik dx d\ c (2.10a) c r ~dx 2 = E -E t ikt - jo>L„/ (2.10b) 2 In equation (2.10a), the total magnetic flux through the area described by the contour A B C D is L cs /j where » it *( ^) L = l r (211) and r and r, being the outer and inner radii of the insulation. The term L„ can be defined 0 similarly. The parameters jtaL , c insulations. and j i * L l t are the impedances Z nl and Z n2 of the respective - 12 - Substituting the values for the electric field strengths from equation (2.8) and (2.9) into equation (2.10), we have dz dV~ ~ Zikm Z tht + Z m2 + Z (2-12) ti dx The matrix equation (2.12) relating the potential differences between the concentric cylindrical conductors and the loop currents can also be obtained from Figure 2.2. directly, i.e.. d\\ dx z Z Z Z x d\' 2 m m '/.' 2 (2.13) dx where the self impedance of loop 1 consists of 3 parts Z\ = Z cre + Z i n l + Z , itix and similarly for loop 2 Z2 ~ Z . sh 4- Z, 2 + n Z, ti while the mutual impedance between loop 1 and loop 2 is Z m Zh = S m Equations (2.12) or (2.13) are not yet in the usual form in which the voltages and currents of the core and the sheath are related to each other. They can be brought ijjto such a form by. considering the appropriate terminal conditions namely v 2 = V tk I = I 2 ik (2-14) + /„ where V cr V = voltage from the core to the local ground, sll = voltage from the sheath to the local ground, I = total current flowing in the core, I = total current flowing in the sheath. C1 sk Substituting the values for voltages V,,V and currents 7,,/ from equation (2.14) into equation 2 (2.12), and adding rows 1 and 2, we obtain 2 - 13 - dV„ ' Z re Z, Z, 2 dx dx + C n Z ,+ Z + nl sh Z es tkc — 1Z + lkm Icr Z $ ~~ z e tkm (2-15) Z kt s ^««2 ^ « — ' ^«ifn - The impedance matrix given in equation (2.15) is the self impedance submatrix [Z„\ for cable t. It can be seen that the elements of the impedance submatrix \Z„] are obtained from the internal impedances of tubular conductors • and from the earth return impedance. These impedances are frequency dependent because of the skin effect, which is discussed in the next section. 2.2.2 S k i n E f f e c t In the derivation of formulae for resistance and inductance of conductors, it is often assumed that the current density is constant over the cross section of the conductor. This assumption is justified only if (i) the resistivity is uniform over the cable cross section, and if (ii) the conductor radius is small compared to the depth of penetration However, as the size or permeability of the conductor increases or as the frequency increases (resistivity still being uniform), the current density varies with the distance from the axis of the conductor, current density being maximum at the surface of the conductor and minimum at the centre. The reason for skin effect is as follows: In a long conductor of uniform resistivity the direction of current is everywhere parallel to the axis, and the voltage drop per unit length is the same for all the parallel filaments into which the conductor may be imagined to be subdivided, since these filaments are electrically in parallel. The voltage drop in each filament consists of a resistive component proportional to and in phase with the current density in the filament, and an inductive component, equal to joi times the magnetic flux linking the filament. There is more flux linking the central filament of a round conductor than linking the filaments at the surface, because the latter are surrounded only by the external flux, whereas the former is surrounded also by all the internal flux. The greater the flux linkage and the inductive drop, the smaller must be the current density and the resistive drop in order for the total drop per unit length to be the same. Hence, the current density is least at the centre of the conductor and greatest at the surface [7]. - 14 - The ac resistance, which is defined as the power lost as heat, divided by the square of the current, is increased by the skin effect, because the increase in loss caused by the increase in current in the outer parts of the conductor is greater than the decrease in loss caused by the decrease in current in the inner parts. The inductance, defined as flux linkage divided by current, is decreased by skin effect because of the decrease in internal flux. 2.2.3 Internal Impedance of Solid and Tubular Conductors As mentioned in the previous section, the voltage drop per unit length is the same in all the parallel filaments into which the conductor can be subdivided because all filaments are The ratio between this voltage drop and the sum of al! filament electrically in parallel. currents is the internal impedance. For a solid conductor of radius a and resistivity p. the internal impedance is given by [7] pm I {ma) 0 2ira/,(m(i) where /„./, = modified Bessel functions of the first kind and of zero and first order, respectively. m = the intrinsic propagation constant of the conductor of equation (2.16). The derivation of the internal impedance formula for tubular conductors is more complex due to the boundary conditions. For example, if we consider the sheath in Figure 2.2. the loop current /,. passes through the inner surface of the sheath and returns internally and the loop current I passes through the outer surface of the sheath and returns externally. This is illus2 trated in Figure 2.5. Therefore, we have to consider the magnetic field strengths on both surfaces (which are then the boundary conditions) while solving the Maxwell's field equations to determine the formulae for the internal impedances. A detailed analysis of this problem had been done by Schelkunoff [6]: his formulae which are relevant to this thesis are summarized in Appendix B. As shown in Figure 2.5, the return path for the current flowing in a tubular conductor may be provided either inside or outside the tube, or partly inside and partly outside. We designate Z„ as the internal impedance of the inner surface of the tube with internal return, and Z as the internal impedance of the outer surface of the tube with external return, and Z b ab as the mutual impedance between one surface of the conductor to the other. The values of Z ,Z a b and Z are given as follows: ai Z = ^ • [ / ( m a ) K , ( m J ) + AT (ma )/,(*.&)] a 0 0 - 15 - Figure 2.5 • Sheath with Loop Currents /, and /j. Z st = P 2xabD (2.17:.,b,c) where D = /j(m6)A',. (ma) - /,(ma) AT,(m6), p = resistivity o f the conductor, m = i n t r i n s i c p r o p a g a t i o n c o n s t a n t o f t h e c o n d u c t o r o f e q u a t i o n (2.17), /<>,/, «= m o d i f i e d B e s s e l f u n c t i o n s o f t h e f i r s t k i n d a n d o f z e r o a n d f i r s t o r d e r , r e s p e c t i v e l y . /f ,W, 0 = the modified Bessel functions o f the second kind a n d of rero and first ordf-r. respectively. U s i n g t h e s e f o r m v . ' a e t h e e l e m e n t s o f t h e s u b m a t r i x [Z.,\ c a n b e f o u n d . 2.2.4 Submatrix The and |ZJ off-diagonal submatrix [Z \ w h i c h r e p r e s e n t s t h e m u t u a l i m p e d a n c e s b e t w e e n c a b l e i tJ cable j c a n be best explained if we consider a transmission s y s t e m consisting o f only two c a b l e s « a n d j a s s h o w n i n F i g u r e 2.6. Before we analyze the elements of the submatrix, we - 16 - Earth Cable j Figure 2.6 Tvuo Cable System will briefly discuss the influence of proximity effects and shielding effects on these elements. 2.2.4.1 P r o x i m i t y E f f e c t Skin effect is caused by the non-uniformity o f current density in a conductor. This current, density is a function o f distance from the axis, but not o f direction from the axis. However, in parallel conductor transmission, in addition to the self-magnetic field (field generated by the current flowing through the conductor), there will be magnetic fields generalcd by currents iD adjacent conductors. These fields interact and result in distortion in the ovr;:.l! symmetric field distribution. The effects of the distortion of symmetry are known-j\|>r«.^ii:ii: \ effects, which in most cases affect the distributed parameters of the transmission system {3.3]. A.H.M. Arnold [13], has given a comprehensive treatment on proximity effect resistance ratios for single-phase and three-phase circuits. He has given equations and tubulated functions of i (defined below) for determining the proximity effect resistance ratios R'zr' in a single-phase circuit of two identical tubular conductors, with solid conductors being a special case. The ratio R'/R' is defined as the ratio of the effective ac resistance with proximity effect taken into account to the effective oc resistance wbcu the conductors are far apart, such that the proximity efTect is negligible. Further, factors to be applied to the ratio /?'//?'while considering a three-phase circuit with symmetrical given in the same reference. triangular spacing or with flat spacing arc also - 17 2.2.4.2 P r o x i m i t y Effect of a Single-Phase C i r c u i t of T w o Identical Conductors The ratio R'/R', defined earlier, for a tubular conductor with the solid conductor being a special case, depends upon three variables, i.e., t/d,d/a and x defined as t/d = ratio of thickness t of the tube to its outside diameter d.(t/d = 0.5 for solid conductor). d/s = ratio of outside diameter d of a conductor to distance s between the axes of the conductors. x = 2nV2ft{d-t)/p x can be further simplified to [13], x — 1.52-\/f/R dc where = the dc resistance of the conductor in Dim. The proximity effect resistance ratio is then given by /?'//?'=- (2.18) where A. B and C are functions of x and t/d which can be determined from tables given in [13]. Similarly, proximity effect inductance ratios of a single-phase cable can be obtained as well. Also, both proximity resistance and inductance ratios for a 3-phase system can be obtained from the single-phase proximity resistance and inductance ratios. For the example chosen in this thesis d/s is less than 0.35. ity effect can be ignored up to frequencies of 1MHz, For such a value, the proxim- [13, 33]. Hence, proximity effects are ignored here. 2.2.4.3 S h i e l d i n g E f f e c t of the S h e a t h Another factor of importance in determining the mutual impedance, is the shielding effect of the cable sheath, which is normally grounded at both ends. Consider a primary circuit 0. a parallel exposed secondary circuit x and a shielding conductor * whose ends arc grounded as d\-'° shown in Figure 2.7- Let netic coupling without dx be the induced voltage in the exposed circuit duc to the any shielding conductor and let — — dx be the induced voltage in the i dV? shielding circuit. The current in the ground shielding conductor is then is the self impedance of the shielding conductor with earth return, Now, exposed circuit by the current in the shielding conductor is mag- — , where Z., Z dx ss voltage induced in the — , where Z. z is the mutual Figure 2.7 • Circuit Arrangement of Primary, Secondary and Shielding Conductors, with Shielding Conductor Grounded at Both Ends impedance between the shielding and the exposed circuit. Therefore, the net voltage induced in the exposed circuit is d\ r x dV? dVf dx dx dx dVf Z Z S ;J (2.19) 5 d\r? If the voltages — — and — — are expressed in terms of the current in the primary circuit as dx dV? dx dx d\ ° r = IZ 0 anc l ~.— = IoZoi> dx 0s £ 1 1 - dx z»z Zn then equation (2.19) simplifies to Os ZQJ (2.20) ZQZIQ and the shielding factor of the grounded shielding conductor is then given by n= l Z,i Z 0i z„z ss (2.21) Ox If the shielding and the secondary conductor are exposed to the same field, which is the case for a shielding wire very close to a telephone line, and for the cable sheath around the core conrO ductor, then dX dx dx ' Therefore Z 0l = Z , which makes the shielding factor to be equal 0s to (2.22) - 2.2.4.4 E l e m e n t s of Submatrix 19 - (Z ). t ; K e e p i n g in m i n d the shielding effect d e s c r i b e d a b o v e , the submatrix \Z, ]. A g a i n , loop currrents are } the elements df the submatrix j Z „ j . we will now derive the elements u s e d , as has been d o n e before for of determining C o n s i d e r i n g t h e c a b l e s y s t e m s h o w n in F i g u r e 2.3, we c a n define the following loop currents for the i t h cable i) loop current /',, whose path consists of the core's external surface a n d sheath's internal surface ii) loop current 7 , whose path consists of the sheath's external surface and earth. 3 S i m i l a r l y , t h e l o o p c u r r e n t s l[ If we consider the a n d I' 2 loop currents F cable i due to the loop current side the cable j there [6], c a n be d e f i n e d for cable I\ 2 a n d I{, there j. will be n o i n d u c e d voltage in loop 2 of a s t h e n e t f i e l d p r o d u c e d b y t h e l o o p c u r r e n t 1\ is z e r o o u t - U s i n g t h e law o f r e c i p r o c i t y o f m u t u a l i m p e d a n c e s , it c a n be d e d u c e d t h a t w i l l be n o i n d u c e d v o l t a g e in the loop 1 of cable j due to the loop current Hence, V. 2 r e l a t i n g the loop c u r r e n t s w i t h the p o t e n t i a l differences between the c o n d u c t o r s we o b t a i n : 'dV\/dx z\ z Z' Z' 0 0 0 0 diydx 0 d\" /dj 2 Zj. , impedance terms in e q u a t i o n Z (2.23), S I 0 r\ Z ft z\ Am n Zin Zk 2 m d\'\/dx Most m x si have already been (2.23) defined except the w h i c h is t h e m u t u a l i m p e d a n c e b e t w e e n t h e e a r t h r e t u r n l o o p s 2 o f c a b l e s »' a n d j . Sy If cables are b u r i e d in a n h o m o g e n e o u s infinite e a r t h , where the p e n e t r a t i o n d e p t h of the c u r r e n t i n t h e e a r t h is l e s s t h a n t h e d e p t h o f b u r i a l , t h e n t h e v a l u e o f y _ pm"lK (ms) \_ 0 Z S i S > is g i v e n b y the return [9]: ' ' term (2 24) where a = r,,Tj = external radii of the c a M c s i a n d j , m = For distance between the centres of the cables, and intrinsic propagation constant of the earth. a h o m o g e n e o u s s e m i - i n f i n i t e e a r t h , w h e r e t h e p e n e t r a t i o n d e p t h i n t h e e a r t h is m o r e t h a n the d e p t h of burial, the m u t u a l i m p e d a n c e Z. t Sf is g i v e n b y (22]: - _ jap. 20 - K (m/?) - /C (mZ) + J 0 (2.25) 0 where d,,d = depth of burial of cables » and j , } m = intrinsic propagation constant of the earth, z = V«' + (d, + d,y s = horizontal separation between cables »' and j. If we measure the voltages with respect to ground, then we can write v = 2 V\ h n = vu and V = I\ = /' /{ = I' n= ih + Hr (2.26) Substituting the values given by equation (2.26), into equation (2.23), and adding rows 1,2 and rows 3,4, we obtain the series impedance matrix for two cable system, as dV\,/dx dVJdx z\ + 2z; + r Z' + Z' z' + z\ z\ 2 m m dV{,ldx Zss dV{ ldx Zss s,s Z Zss h 2 Z{ .+ 2ZL } s i Z' 2 + Zss Hr Zss Ilk + z>Zln 2 Z' + z zi 2 2 (2.27) Hr 7 «*. From equation 2.27, the impedance submatrix [Z, ] defined earlier in equation (2.5) is } given by - 21 - |2,l = It is i n t e r e s t i n g core of cable j the same. equations. to note, that the the c o r e o f cable i a n d the a n d the m u t u a l i m p e d a n c e botween the core of cable i a n d s h e a t h of cable j are T h i s raises the It be i l l u s t r a t e d (2.28) mutual question whether impedance between the shielding effect is i n d e e d i m p l i c i t l y t a k e n c a r e o f i n t h e f o r m u l a t i o n with the help of a c o n d u c t o r w e a r t h , as s h o w n in F i g u r e is p r o p e r l y r e p r e s e n t e d witb loop currents. p l a c e d in close p r o x i m i t y in the This can to a cable buried in the 2.8 Earth Conductor W Figure 2.8 - Transmission System Consisting of a Single Conductor and a Cable. Fov t h e s y s t e m s h o w n i n F i g u r e 2 . 8 , t h e v o l t a g e s a n d c u r r e n t s o i l t h e c a b l e c a n be w r i t t e n as: — %cc Ic dx dV dx sh ~ Z ci Z I CS S CU VI I + Z, /, + Z c t (2.29a) + Z ,I lV) (2.29b) /„ where Z ,Z et tt — t h e s e l f i m p e d a n c e o f t h e c o r e a n d s h e a t h o f the c a b l e , r e s p e c t i v e l y , — the m u t u a l Z ..Z . ca ia = the mutual i m p e d a n c e b e t w e e n the c o r e a n d the s h e a t h , a n d . impedances between a n d c o n d u c t o r w, respectively. the core and conductor w and between the sheath - 22 - Suppose that the cable sheath is not grounded at the ends, but used as the return pa;b for the current flowing in the core. Then there is no magnetic field outside the sheath, and no voltage will therefore be induced in conductor u>. Since this induced voltage is Z^L. with I sk — ~ / T . it follows that Z^ — Z £ + Z . /,.. s must be true. On the other hand, if the sheath is sw grounded at the ends, then there will be a circulating current through the sheath and earth, and V sk becomes zero. Hence, the value of 7 can be found from equation (2.29b) as: $ Z SUI (2.30) ^ss Substituting the value of 7 into equation (2.29a) we obtain S dV, Z-^ dx z " z.. z„ - 7 7 1 - c z s ~z~ 1 - Za ' Z Z ls Si The term Z $' Z . L + sw IwZ u ' Zcv, C (2.31) is the shielding factor of the sheath for the field produced by conduc- Za Z , cu tor w. With Z = Z , cw sw it can be simplified to 1 - Zss which is the same as equation (2.22). Hence, the shielding effect is implicitly taken care of in the equations. 2.3 S h u n t A d m i t t a n c e M a t r i x Y for N C a b l e s in P a r a l l e l In a manner similar to the series impedance matrix Z, the shunt admittance matrix Y can be expressed in terms of two submatrices [Y„] and [Y,J. Since the soil acts as an electrostatic shield between the cables, the off-diagonal submatrix [>',,] will be a null matrix. Hence, we only have to derive the submatrix [Yj,|. Before we obtain the elements for the submatrix [}'„], the admittance of insulation will be discussed first. 2.3.1 L e a k a g e C o n d u c t a n c e and Capacitive Susceptance Figure 2.9 shows a cross-section of a coaxial cable, with insulation between core and sheath, and between sheath and earth. Let us assume that the insulation has a relative permittivity of The admittance Y per unit length of the insulation is defined as j<i)27reo€* y = G + jB = — In r /r 2 1 - 23 - Figure 2.9 - Croaa-Section of a Coaxial Cable. 77^7 V ~ J jti)27rf„( In TJT , (2.32) In rjr, The first term in the left hand side of equation (2.32), is the leakage conductance of the insulation. It is the result of the combined effects of leakage current through the insulation and of the dielectric loss [7]. The second term is known as the capacitive susceptance of the insulation between the two conductors (core and sheath, or sheath and earth). 2.3.2 Frequency Dependency of the Complex Permittivity Generally, the dielectric constant i is assumed to be a real constant with the imaginary part of t neglected due to its relatively small value, (of the order of 10" compared to the real 4 value [7,13,22.27]). However, the complex dielectric constant i is not a constant as its name implies. It depends on a number of factors such as the frequency of the applied field, the temperature and the molecular structure of the dielectric substance [15]. Let us consider two commonly used, insulating materials for power cables, namely crosslinked polyethylene (XLPE) and oil-impregnated paper. The values of i and i for XLPE are approximately constant for at least up to frequencies of \00MHz [16]. Typically, they have values of i = 2.33 t *=4.66 1 ( r 4 (2.33) - 24 - Unfortunate!}', little is mentioned in the literature about the frequency dependency of the permittivity in the case of oil-impregnated paper (18]. Recently Johanscn and Breien [17] published the measured values of i and t for the oil-impregnated paper for a frequency range of 1Hz to 10Q.4/7/;. The value of i was found to vary by 20%, whereas, the value of ' varies by 2 0 0 % for the same frequency range, at a temperature of 20°c. Figure 2.10(a),(b) Measurements of t ( t o ) and t*(co) j * on an Oil-Impregnated Test Cable at 2(f c. Figure 2.10(a) and (b) show the experimental data obtained for I and t for the frequency range 10* to 10 Hz only. Based on this experimental data, the authors [17] developed an empirical 8 formula for t'(oi). - " + (, + y J x . o - T " 1 2 3 1 1 Figures 2.11 (a) and (b) show the plot for i and ('obtained from the empirical formula, which closely match the experimental data of Figure 2.10(a) and (b). - 26 - Frequency [Hz] Figure 2.11 Values of t\ui) a n d t"(o>) Obtained from the Empirical Formula. - 26 - A general formula for tbe complex permittivity of any material as a function of frequency is given by Bartnikas [15]. According to Bartnikas, when a dielectric is subjected to an ac field, at low electric field gradients, its electrical response will depend upon a number of parameters such as the frequency of the applied field, the temperature and the molecular structure of the •dielectric substance. Under some conditions, no measurable phase difference between the dielectric displacement; D, and voltage gradient E will occur, and consequently the ratio DIE will be defined by a constant equal to the real value of the permittivity, c'. When a dc field E is suddenly applied across a dielectric, tbe dielectric will almost instantaneously, or in a very short time, attain a finite polarization value. This polarization value will be almost instantaneous, since it will be determined by the electronic and atomic polarizability effects. The limiting value of the real dielectric constant <'for this polarization is defined as <«, so that the resulting dielectric displacement is D or i E. m m The slower processes, due to the dipolc oriontatiou or ionic migration, will give rise to a polarization which will attain its saturation value considerably more gradually because of such effects as the inertia of the permanent dipoles. The static dielectric displacement vector, D it in this case is equal to i E, where t, is the static value of s the real dielectric constant, e. Figure 2.12 - Polarization-Time Curve of a Dielectric Afaterial In the idealized polarization time curve depicted in Figure 2.12, P is the achieved saturaf tion value of tbe polarization resulting from permanent dipoles or from any other displacement of free charge carriers. Depending upon the temperature and the chemical and physical structure of the material, the saturation value, P may be achieved in a time that may vary anyt where from a few seconds to several days. If we denote the time-dependent portion of P as % P(t), the equation of the curve in Figure 2.12 can be represented by a form characteristic of the - 27 - charging of a capacitor P(t) = P [l - exp(-f/r)] (2.35) s where T is the time constant of the charging process. The time constant, T is a measure of the time lag and is referred to as the relaxation time of the polarization process. Now, the real and imaginary parts of the permittivity of an insulating material can be expressed as a function of frequency in terms of tm.e, and T as (2.36a) t = Rc(e') = £«, + 1 + O) ^ 2 I = Im( ') = £ (t -e„)oiT s (2.36b) 1 + o) ^ 2 In summary, frequency dependence of the permittivity t is complicated, although, for some insulating materials, such as (XLPE), i is practically constant. nificant, i.e., of the order of 10 , for oil-impregnated paper. 2 The changes are very sig- Typical values for the real and imaginary parts of t for X L P E are given in equation (2.33). The real and imaginary values of t for oil-impregnated paper can be obtained from equation (2.34), based on the reference [17]. A general formula for ''-he complex permittivity of any material is given by equation (2.37a). 2.3.3 Submatrix \Y \ U We The shall now determine the self admittances of the cable system shown in Figure 2.2. loop equations for the current changes along loops 1 and 2 will be: di, 0 dx 0 n dl 2 v (2.37a) 2 dx ?here Yi = G x Y 2 2 = G + jB — admittance of insulation between core and sheath, A + jB 2 — admittance of insulation between sheath and earth, Vj = voltage between the core and sheath, V 2 — volt age between the sheath and earth. Substituting the values for currents II lt 2 and voltages VltV2 from equation (2.14) into (2.36) and subtracting row 2 from row 1, we obtain: - dl , dx 28 - c Yx -Yr -Y, Y + Y x (2.37b) . »*. v 2 dx Hence, the submatrix \Y„] is given by: Y ~Yx -Yx Yx + Y (2.38) t 2 Recently, Dommel and Sawada [19,20] suggested that the admittance matrix should include the effect of the grounding resistance as well if the insulation between the sheath and earth is electrically poor, as in oil or gas pipelines. In such cases, the leakage current flows through the series connection of the insulation resistance and the finite grounding resistance. For conduction effects in pipelines they, therefore, use 1 ' insulation (2.39) R,earth where the grounding resistance /? h i given by 3 eirt R earth 2J_ fa V(2ff) + (t/2) + 1/2 g Pearth 2 (2.40) + 4;r D lnV(2H) 2 + (1/2) 2 - 1/2 with Pearth = earth resistivity, H = depth at which the cable is buried, / — length of the cable. Strictly speaking, G in equation (2.39) is no longer an evenly distributed parameter 2 because /? eirtt in equation (2.40) is a function of length. In [20], it is shown that the change in the value of C .with the length is practically negligible, and treating G as an evenly distri2 2 buted parameter is therefore a reasonable assumption. 2.4 Conclusion First, the series impedance and shunt admittance matrices were defined. These matrices are made up of self and mutual impedance (or admittance) submatrices. Elements of the self impedance submatrices were obtained from the internal impedance formulae for tubular conductor and from the earth return self impedance formula. The elements of (he mutual impedance submatrices were obtained from the earth return mutual impedance formulae. The shielding effect of the sheath and the proximity effects between the conductors were studied next to assess their influence on the elements of the mutual impedance submatrix. Finally, the - 29 - self admittance submatrix and mutual admittance submatrix were defined. Since the earth acts as an electrostatic shield, the elements of the mutual admittance submatrix are zero. The permittivity c* of the insulation which is needed to evaluate the elements of the self admittance submatrix, is frequency dependent and complex. An empirical formula for finding the real and imaginary parts of the permittivity t * as a function of frequency was shown, which can then be used to find the elements of the self admittance submatrix. - 30 - 3. C O M P A R I S O N OF INTERNAL IMPEDANCE FORMULAE lu the previous-chapter the scries impedance matrix was assembled from the internal impedances of tubular conductors and from the earth return self and mutual impedances of tubular conductors. The formulae for the internal impedances and earth return impedances are given in terms of modified BesscI functions, which can be expressed as an infinite scries f o r : small arguments and as an asymptotic series for large arguments. Before computers became available, exact calculations were almost impossible and approximate formulae were therefore developed. Such approximations had been proposed by several authors (G,22,2'5',"2 l]. chapter, approximate In this formulae for the internal impedances are compared with i lie exact for- mulae in terms of cpu time and in terms of accuracy. The earth return formulae are discussed in the next chanter. 3.1 E x a c t F o r m u l a e f o r T u b u l a r Conductors Axis I 17/ Figure 3.0 - Ix>op Currents in a Tubular Conductor The internal impedances of a tubular conductor as given in equation (2.17) are as follows: f/otmoJK-.M) + Z. = Z t = -^L. \ Z.» = I o { r n b ) K l (m ) a KjmaVAmb)] (3.1a) + K^mbVJma)] (3.1b) (3-lc) 2ncbD where D - /j(ml»)K',(ina) - I (ma)K {mb) l l - 31 - The argumeDts for the modified Bessel functions 7 , / i and K ,Ki 0 0 of the first kind and second kind are complex because the intrinsic propagation m of the conductor is ^/w/y/p) . For more exact- calculations, m should be {j<a/i/p— <>>/i<) , where the first term under the 2 square root accounts for the conduction current and the second term accounts for the displacement current. The displacement current can be ignored in good conductors as it is negligible compared to the conduction current up to frequencies of 10MHz. For example, copper conductors with p = 1.7 10~ /?m resistivity would have a displacement current at 10MHz which is 11 8 orders of magnitude smaller than the conduction current, and even smaller than that below 10MHz. However, in the earth, the displacement current has an influence on the impedance as the earth's conductivity is of the cider I O such as copper. - 1 0 smaller than the conductivity of a good conductor Displacement currents are therefore taken into account in the earth return impedances discussed in the next chapter. Subroutine T U B E originally developed by H. W. and I. I. Dommel [30] (and later modified by L. Marti [29]) and Amctani's Cable Constants program implemented in BPA's E M T P , assume that the displacement currents can be neglected. This fact enables us to express the modified Bessel functions of the 1st and 2nd kind in terms of Kelvin functions. For example, the complex functions K {mr) 0 and I (mr) can be expressed as real and imaginary parts as fol0 lows: A'„(mr)= K (VJ\m]r) 0 7 (mr)= 0 = Kj\m\r)+ /o(V7T^lT)= B„(\m\r)+ jA' (|m|r) (3.2a) tl jB (|m|r) (3.2b) ei where B„ and £?,, are Kelvin functions of the first kind, and K~ and K ¥ cl are those of the second kind. T o evaluate the Kelvin functions, infinite series and asymptotic series can be used for small real arguments and large real arguments, respectively. Subroutine T U B E and Ametani's routine use such series with a sufficient number of terms to guarantee high accuracy, 3.2 Internal Impedance of a Solid Conductor The exact, formula for the internal impedance of a solid conductor of radius r follows from Z of equation (3.1) by setting o = 0 and 6 = r: h ^1^1 2jir/,(mr) where r = radius of the conductor 3 -32 - m = intrinsic propagation constant of the conductor / , / m o d i f i e d Bessel functions of the first kind and of zero and first order, respectively. 0 1= Wedepohl [22] suggested an approximation to this exact formula, given by z _ pmcoth(fcmr) p(l-l/2k) + This approximation was developed by first considering the equation = pmcoth(fnr) ( 3 5 ) 27rr This equation is known to exhibit similar properties as the exact equation given by Equation (3.3). For example, at high frequencies, the impedance term Z tends to be pm/2xr which is a x well known skin effect formula [22], while at lower frequencies it represents pure resistance although not, in fact, equal to the required value p/xr . Equation (3.5) can be improved to take 2 account of the dc resistance more precisely by writing Z = -^-coth(W)+ ^ - f * ) 2nr nr (3.6) > where k is an arbitrary constant. The second term on the right hand side of this equation corrects the impedance at direct current. The value of A; chosen to give the correct resistive component is 0.777. There is another interesting formula derived by Semlyen in the discussion of reference 24, where the internal impedance of a solid conductor is given as Z = y/R? + Z x (3.7) a where R is the dc resistance given by p/xr and, Z is the impedance at very high frequencies 2 c given by pm/2-r. a Table 3.1 shows the values of resistance R = R {Z } and inductance x e x L = — 7m{Z,} obtained from subroutine T U B E and from Wedepohl's and Semlyen's approxix (i) mation formulae. Figure 3.1(a) and (b) show the resistance and inductance as a function of frequency. The errors in the values of resistance and inductance in Wedepohl's formula and Semlyen's formula are plotted in Figure 3.2(a) and (b). From the table and figures we can see that Wedepohl's formula has an error of 1-3% in the frequency range 100Hz to 1kHz for the resistive part, and 4% error up to frequency of 1kHz in the inductive part. Semlyen's formula has an error of 4-7% for the frequency range 60Hz to 20kHz, in the resistive part, and an error of 4-11% for the frequency range 2 0 H i to 300Hz in the inductive part. - 33 - T a b l e 3.1 Internal Impedance of a Solid Conductor (p = 1 7 10~'Om FREQUENCY (Hz) and r = 0.0234m) WEDEPOHL TUBE SEMLYEN RESISTANC;E (Q/km) .01 . 1 1 10 100 1,000 10,000 100,000 1 ,000,000 10,000,000 0.0098825 0.0098825 0.0098858 0.0102067 0.0203380 0.0582719 0.1786977 0.5596756 1.7644840 5.5744390 i 0.0098825 ; 0.0098825 0.0098874 0.0103294 0.0190556 0.0561595 0. 1763397 0.5572406 1.7620250 5.5719720 0.0098775 0.0098776 0.0098809 0.0102034 0.0211463 0.0592378 0.1797193 0.5607150 1.7655290 5.5754860 INDUCTANC3E (jiH/km) .01 .1 1 10 100 1,000 10,000 100,000 1,000,000 10,000,000 50.000000 49.999920 49.991580 49.181730 27.567070 8.853760 2.803902 0.886793 0.280432 0.088680 51.800000 51.799920 51.792250 !i 51.042710 i 28.354910 ; 8.868063 2.804328 0.886806 0.280433 0.088680 The cpu times were found to be: TUBE 0.042 ms Wedepohl 0.038 ms Semlyen 0.022 ms ; 50.000000 49.999750 49.974780 47.836480 25.930690 8.798590 2.802123 0.886736 0.280430 0.088680 - 34 - figure 3.1(a) and (b) Impedance of a Solid Conductor ae a Function of Frequency. - 35 - 5- WEDEPOHL SEMLYEN '- 2 f Ui w c ro i ----.: x -1-1 - *—« i CO 0) / / / \ / cc -5-7-4 11 10~ 2 10 _1 1 10' 10 2 10 5 10 4 10 10 s 6 10' 12 WEDEPOHL 8 SEMLYEN ^ 4 o L— uj o o c —4 o T3 1/ -8 10~* 10" 1 1 10' 10 2 10 3 10 4 10 5 10 Frequency [Hz] Figure 3.2(a) and (b) Errors in Wedepohl's and Stmlycn'» Formulae for a Solid Conductor. 6 10 7 - 36 - To verify that the displacement current is indeed negligible, a modified subroutine TUBEC was written which takes displacement currents into account. Within the accuracy given in Table 3.1. TUBE and TUBEC produced identical mults. 3.3 Internal Impedance of a T u b u l a r Conductor; Figure S.S • Cross Section of a Tubular Conductor There are three impedances associated with a tubular conductor of the type shown in Figure 3.3: 1) Internal impedance Z , which gives the voltage drop on the inner surface, when a unit a current returns through a conductor inside the tube. 2) Internal impedance Z , which gives the voltage drop on the outer surface when a unit t current returns through a conductor outside the tube. 3) Mutual impedance Z ai of the tubular conductor which gives the voltage drop on the outer surface when a unit current returns through a conductor inside the tube or vice versa. The formulae for Z^,Z and Z, , originally developed by Schelkunoff, are given in equation (3.11. b t These formulae are given in terms of modified Bessel functions, and are obviously not suitable for hand calculations. Schelkunoff therefore approximated these exact formulae by replacing the modified Bessel functions J ,I ,K 0 X 0 and K", by their asymptotic expressions and performing the necessary division as far as the second term. Schelkunoff's approximations are as follows [6]: Z = -^cotb(m(6-o)) - -^—{^ t + Z = | ^ c o t h ( m ( 6 - a ) ) + ^ ( 3 / a + l/6} i ( 3 8 a ) (3.8b) - Zah = 37 - ^|LcoBeft(m(6--c)) (3. ) 8c Wedepohl and Wilcox, give a similar approximation [22] but with a different approarh. The magnetic intensity H and the electric current density / in a tubular conductor can be related by the equations: f f-:' + = m 2 / <»;»») (3.9b) / If the tube is thin compared with its mean radius, equation (3.9a) can be written as: dH 2H , Equation (3.9b) and (3.10) lead to a second-order differential equation in H. Now, solving for H from this second-order differential equation and following the same procedure as given in Appendix A for the exact formulae, we obtain the following equations: Z t = ^ c o i H m { b . a ) Z t = ^ c o t H m ( b - a ) <» Z = ) ) - - ^ + ( ^ ^ {3 -7~TCoseA(m(6-a)) 7t(a + 0) 3, l a .1, ) b ) (3.11c) There is another approximation derived by Bianchi [23]. If the difference between the radii b and a is very much less than either of the radius o and (6 — a)«a,b then the impedances can be expressed as Z Z The ° = > °> Z = /™ = coth(m(6-a)) (3.12a) -7^«>seA(m(6-a)) impedance Z ti 6, or in other words, if • (3.12b) obtained by Bianchi in equation (3.12b) is the same as that obtained by Wedepohl in equation (3.11c). Using the exact formula (subroutine TUBE) Wedepohl's approximate formula, Schelkunoff's approximate formula and Bianchi's approximate formula, the values of resistance and inductance terms of the impedances Z Z t t and Z 4t were obtained for a typical tubular con- ductor. Table 3.2 shows the resistance and inductance values as a function of frequency for the internal impedance Z . a Figures 3.4(a) and (b) show the resistance and inductance for the fre- quency range 0.01 Hz to 10MHz. T o highlight the differences in the results, the errors in the - 38 - Table S.2 Internal Impedance Z of i Tubular Conductor with t Current Returning Inside (p = FREQUENCY (Hz) TUBE 2.1 \0~ Dm,a 7 = 0.0385m,6 = SCHELKUNOFF WEDEPOHL 0.0413m) BIANCHI RESs I STANCE ($2/k.m) .01 . 1 1 10 100 1000 10000 100000 1000000 10000000 0. 299163 0. 299163 0. 299163 0. 299163 0. 299169 0. 299761 0. 354424 1. 18036 3. 75275 11 .8915 0.288640 0.288640 0.288640: 0.2886401 0.288646 0.289224 0.343956 1.17015 3.74259 11.8813 0.299163 0.299163 0.299163 0.299163 0.299169 0.299762 0.354480 1.18068 3.75312 11.8919 f 0.299163 0.299163 0.299163 0.299163 0.299169 0.299741 0.352539 1 .14975 3.63192 11.4852 INI 5UCTANCE ( H/'km) M .01 . 1 1 10 100 1000 10000 100000 1000000 10000000 4. 84605 4. 84605 4. 84605 4. 84605 4. 84603 4. 84339 4. 60048 1 .89284 0. 59905 0. 18944 4.84848 4.84848 4.84848 4.84848 : 4.84846 4.84581 4.60256 1.89296 0.59906 0.18944 1 4.84848 4.84848 4.84848 4.84848 4.84846 4.84581 4.60256 1.89296 0.59906 0.18944 4.67836 4.67836 4.67836 4.67836 4.67834 4.67578 4.44106 1.82654 0.57804 0.18279 approximate formulae are plotted in Figure 3.5. From these figures and the table it can be seen that 1. Wedepohrs formula has almost no error up to a frequency of 1MHz for both the resistance and inductance. 2. Schelkunoff's approximation has an error of 2-4% up to a frequency of 10kHz in the resistive part. The inductance value obtained by SchelkunofTs approximation is the same as that obtained by Wedepohl's approximation. - 39 10- TUBE WEpEPOHL SCHELKUNOFF BIANCHI v o c _p or O.H 10" 10" 10' 10 2 10 3 10' 10 10' 10 10* 10' 10' 10 5 10- TUBE E WEDEPOHL \ x jt aj o c _o "o SCHELKUNOFF BIANCHI i- T> C 0.1+- 10" 10" 10' 10 J 10 3 Frequency [Hz] Figure S.4(a),(b) • Impedance Z, of a Tubular Conductor (with Internal Return) as a Function of Frequency. - 40 - WEDEPOHL ^ SCHELKUNOFF 2 BIANCHI o " tt> 0 o c \ . - ' " \ (0 \ /' A co "to tu cc _-> z 10~ 2 10"' 1 10 10 1 2 10 10 J 10 4 10 s 10 6 7 WEDEPOHL e£ SCHELKUNOFF 2 BIANCHI o k_ k_ Ul v t> 0 o c <0 u "D c -2 I .Mill,, 10" 2 IHj 10" 1 1 I 1,1,11V ! I llllll, 10' I .IIIItT, 10 2 I I I Mill, 10 5 I I llllll, 10" I I II.Ill, 10 5 I 10 s llllllf 10 7 Frequency [Hz] Figure S.5 -Errors in Wedepohl'*, Sehelkunoff'e forZ . t and Bianchi'a Formulae 3. 41 - Bianchi's approximation formulae is good for frequencies less that 10 kHz in the case of resistance, where the error is almost zero, but beyond that frequency the error increases. In the inductance the error is around 3% for the whole frequency range. The cpu time for the routines were found to be TUBE 0.059 ms Wedepohl 0.037 ms Schelkunoff 0.036 ms Bianchi 0.033 ms Similar comparisons were made for the mutual impedance Z ab and the impedance Z . t Table 3.3 gives the values of resistance and inductance of the impedance term Z..^ obtained from routine TUBE, from Wedepohl's approximation formula and from Schelkunoff's approximation formula at different frequencies. Since the approximations proposed by Wedepohl and Bianchi are identical, only Wedepohl's approximation was considered. The errors in the approximate formulae are shown in Figure 3.6. It can be seen from Figure 3.6 that the errors in both Wedepohl's and Schelkunoff's approximations are less than 0.5^ up to a frequency of 1MHz. Wedepohl's approximation is closer to the exact formula. The cpu times were found to be: TUBE 0.056 ms Wedepohl/Bianchi 0.038 ms Schelkunoff 0.037 ms In a similar manner, the resistance and inductance components of Z were calculated. t Table 3.4 shows the values of these parameters using the exact formula and various approximations at different frequencies. Figures 3.7(a) and (b) show the resistance and inductance as a function of frequency for the range 0.01 Hz to 10MHz. The errors in the approximations are plotted in Figure 3.8. We note the following for the impedance Z 4 - 42 - 0.4 WEDEPOHL SCHELKUNOFF 0.1- \ ui CP o c CD - -o.H TJ C — 0.3 f 111 10~ 2 (Mill I 10" I I lllll, 1 i i inn: ' i i mill i linn - TT!!I, 10' 1 1 lllTHt ' T 1 MTIir, 10 \2 2 10 r>3 I I 1 llllll' s 4/ 1 10" 10* 10 6 Frequency [Hz] Figure S.6 Errors in Wedepohl's and Sehetkunoff'e Formulae for Z tl 10 7 - 43 - Table S.S Mutual Impedance (Z , of a Tubular Conductor e6 (p = 2.1 10"/?m,o = 0.0385m ,6 = 0.0413m) 7 FREQUENCY (Hz) TUBE WEDEPOHL SCHELKUNOFF RESISTANC:E (8/km) .01 .1 1 10 100 1 ,000 10,000 100,000 1,000,000 0.29916343 0.29916343 0.29916343 0.29916338 0.29915838 0.29865873 0.25299441 -0.06964902 0.00001930 INDUCTANC3E .01 .1 1 10 100 1,000 10,000 100,000 1 ,000,000 0.29934775 0.29934776 0.29934776 0.29934771 0.29934270 0.29884247 0.25312757 -0.06966688 0.00001930 0.29916343 0.29916343 0.29916343 0.29916338 0.29915838 0.29865846 0.25297170 -0.06962398 0.00001929 -2.3386052 -2.3386052 -2.3386052 -2.3386052 -2.3385803 -2.3361094 -2.1095191 -0.0097380 0.0000082 (uH/km) -2.3391813 -2.3391813 -2.3391813 -2.3391810 -2.3391563 -2.3366832 -2.1099033 -0.0097146 0.0000082 -2.3406226 -2.3406226 -2.3406226 -2.3406223 -2.3405975 -2.3381230 -2.1112033 -0.0097206 0.0000082 T h e errors in W e d e p o h l ' s a p p r o x i m a t i o n are practically negligible for b o t h resistance a n d inductance up to a frequency of The 10MHz. errors in Schelkunoff's a p p r o x i m a t i o n are around a n d decreases for higher frequencies. 3.5% u p to 10kHz a frequency of Schelkunoff's a p p r o x i m a t i o n gives the s a m e values as W e d e p o h l ' s a p p r o x i m a t i o n f o r the i n d u c t a n c e t e r m . The errors in Bianchi's approximation increase thereafter for the whole frequency range. are resistance term. negligible for frequencies up to T h e e r r o r i n t h e i n d u c t a n c e i s 3.8% 1kHz for but the - 44 - Table S-4 Internal Impedance Z of a Tubular Conductor k tvith Current Returning Ouleide (p = 2.1 10~ /?m,a = 0.0385m ,6 = 0.0413m) 7 FREQUENCY (Hr) TUBE RE<51 STANCE .01 .1 1 10 100 1000 10000 100000 1000000 10000000 0. 299163 0. 299163 0. 299163 0. 299163 0. 299169 0. 299720 0. 350679 1. 120643 3. 518632 11 .10564 SCHELKUNOFF WEDEPOHL BIANCHI ( j B / J;m) 0.299163 0.299163 j 0.299163 0.299163 0.299169 0.299721 0.350729 1.120912 3.518953 11 .10605 0. 309686 0. 309686 0. 309686 0. 309686 0. 309691 0. 310243 0. 361252 ; i . 131444 3. 529473 11 .11652 0. 299163 0. 299163 0. 299163 0. 299163 0. 299169 0. 299741 0. 352539 149755 • 3.631926 1 1.48527 ;T. INIHJCTANCE (uH/ 'km) .01 .1 1 10 100 1000 10000 100000 1000000 10000000 : 51759: 51759 4. 4 . 51759 4 . 51759 4 . 51756 4 . 51510 4 . 28865 76453 1 . 0 . 55844 0 . 17660 4.51977 j 4.51977 4.51977 4.51977 4.51975 4.51728 4.29052 1 .76463 0.55844 0.17660 1 i *• 51977 4.51977 ! 4.51977 4 . 51977 4. 51975 4. 51728 4 . 29052 1. 76463 0 . 55844 0. 17660 i I 4.67836 i ! 4.67836 4.67836 4. 67836 4. 67834 4. 67578 4. 441 06 82655 1 . 0. 57804 0. 18280 - °icr 2 io"' 1 io' 10 45 1 - 10 io 3 4 io 5 10 s Frequency [Hz] Figure S.7(a),(b) - Impedance Conductor (with External Return) Z of a b as a Function Tubular of Frequency. io 7 - 46 - / 2-- s V /\ O kUJ <o 0 o c to WEDEPOHL SCHELKUNOFF _•> to BIANCHI 10~ 2 10"' 1 10' 10 2 10 S 10" 10* 10* 10 7 o fc— fc— UJ a> u 0 cto WEDEPOHL u 3 TJ c SCHELKUNOFF -2 BIANCHI — A ] I I . i i i . i II,i| 10~ 10" 1 2 1 i i i . nn, • J • i in., .ii.i.ii, i . , i mi, III nm, .MI mil i i 11 in 10' 10 10 10 10 10* 10 2 S 4 S Frequency [Hz] Figure S.8 - Errors in Wedepohl's and Bianchi's Schelkunoff's Formulae for Z t 7 - 47 - The cpu times were found to be: TUBE 0.058 ms Wedepohl 0.038 ms Schelkunoff 0.037 ms Bianchi 0.033 ms 3.4 Conclusion The exact or classical formulae for finding the internal impedances for a tubular conductor were given by Schelkunoff. Since these formulae were not suitable for hand calculation purposes, approximations for these classical formulae were developed by many authors, including Schelkunoff himself. In this chapter, the accuracy and the cpu time taken by these approximations were compared with the classical formulae for the tubular conductor of a typical cable. The displacement current term is neglected in subroutine T U B E and in Ametani's cable constant routine, which use the exact formulae. Though not discussed in detail, a modified subroutine T U B E C was developed which takes the displacement current into account. In all cases, the results from T U B E C and T U B E were identical within the accuracy shown in the tables. Subroutines were then written for the approximate formulae of Wedepohl, Schelkunoff and Bianchi, and the values obtained from these approximations were compared with the values obtained from the classical formula. Wedepohl's approximation formulae were indeed very good if the conductor thickness is small compared to its mean radius. The approximation proposed by Schelkunoff is similar to that proposed by Wedepohl except for the 2nd term. Schelkunoff approximated the modified Bessel functions in the exact formulae by the asymptotic series and retained only two terms, which produces reasonably accurate results as long as the argument is larger than 8 [34]. In the example, the argument terms | ma | and | mb | did not reach the value 8 up to frequencies of 2kHz. Hence, approximations based on the asymptotic series would obviously produce errors at low frequencies. Bianchi's approximation is only good at low frequencies less than 1kHz, for the resistance and has an acceptable error of 3-4% in the inductance up to frequencies of 10MHz. It should also be noted that all the above approximations are valid only if the thickness of the conductor is smaller than the mean radius of the tubular conductor. - 48The routines for Wedepohl's approximation - W E D A P , Semlyen's approximation - SEMAP, Schelkunoff's approximation - SCHAP, Bianchi's approximation - B N C A P and T U B E C were written by the author. If these routines were written by a more experienced programmer, they might consume less cpu time than shown earlier. Even then, the cpu time for the exact formula ( T U B E ) would not be much more than that of the approximations. Therefore, the exact formula with subroutine T U B E is recommended for computer solution. For hand calculation or for calculations with electronic calculators, Wedepohl's formulae are recommended. - 49 - 4. EARTH RETURN IMPEDANCE The self and mutual impedance of conductors with earth return are of importance in studies of inductive interference in communication circuits from nearby overhead lines or underground cables. Also they are important in the calculation of voltages in power lines or communication circuits due to lightning surges or other transients [14]. Generally, the earth acts as a potential return path for currents in the underground, aerial or submarine cables. The values of cable constants therefore depend on the earth return impedances. In practical situations, the earth's electrical characteristics such as resistivity, permeability and permittivity are not constant. However, simulation results came reasonably close to fied test results if a homogeneous earth is assumed. The equations for self and mutual impedances are therefore developed with that assumption. The impedances are obtained from the axial electric field strength in the earth due to the return current in the ground, which in turn, can be obtained from Maxwell's equations. If a cable is assumed buried in an infinite earth, (where the depth of penetration of the return current in the ground is smaller than the depth of burial, or, in other words, the distribution of return current is circularly symmetrical) the electric field strength can be easily derived, because only the earth medium must be considered in Maxwell's equations (Appendix C). On the other hand, if the earth is treated as semi-infinite, (where the depth of penetration of the return current is larger than the depth of burial so that the depth of penetration of the return current is not circularly symmetrical) the problem of finding the axial electric field strength in the earth is quite complex, because both air and earth media must be considered in Maxwell's equations (Appendix D). The solutions for the electric field strengths, for both infinite and semi-infinite earth are derived, assuming first that the conductor is a filament with negligible radius. In the case where the return current in the ground is circularly symmetric, exact equations are still easy to derive for cables of finite radius. It is quite difficult, however, to extend the equations for the filament conductor to a conductor of finite radius in the case where the earth return current distribution is not circularly symmetric. In this chapter, we discuss the conditions under which these equations derived for the filament conductor can be extended to a conductor of finite radius. Furthermore, the effect of neglecting the displacement current term is discussed as well. Ametani's Cable Constant routine in E M T P uses different formulae for these impedances. His approach, as well as the approximations proposed by Wedepohl and Semylen are discussed and compared with the exact equations. Another impedance of interest is the mutual impedance between a buried conductor and an overhead conductor. This topic is also covered here. - 50 - 4.1 E a r t h Return Impedance of Insulated Conductor The simplest underground cable consists of a conductor laid at depth d below the surface of the ground with insulation around it which forms a concentric dielectric cylinder of external radius a. The earth then forms the return path. The ground return self impedance Z is ti defined as the ratio of the axial electric field strength at the external surface of the insulation to the current flowing in the cable. T h e earth return mutual impedance between the loops of two buried, insulated conductors is defined as the ratio of the axial electric field strength at the external surface of the insulation to the current flowing in the other conductor and vice versa. As a preliminary step, the self and mutual impedances will be first found on the assumption that the cables are buried in an earth which is homogeneous and infinite in extent. Clearly, this situation does not arise in practical applications, although it is a reasonable approximation if the cable is buried at great depth, or if it is at modest depth but the frequencies are so high that the return current will flow very close to the cable. Furthermore, this treatment will be found useful in justifying the simplifying assumption 2 in section 2.1 of Chapter 2, and will also be helpful in interpreting the results for the more general case of earth return impedance in a homogeneous semi-infinite earth. 4.2 E a r t h R e t u r n Impedance in a Homogeneous Infinite E a r t h The calculation of earth return impedance in an infinite earth is relatively easy since there is no surface discontinuity as in the semi-infinite case (earth and air). With the assumptions mentioned]in Chapter 2 (except for assumption 3), it is shown in Appendix C that the electric field strength at a radial distance r from a cable of insulation outer radius a carrying current / which returns through the earth can be written as: E = . _ pml Ko(mr) ( 4 ,) ITXQ. A'j(mo) where p = resistivity of the earth m = intrinsic propagation constant of the earth. The earth return self impedance per unit length of the cable is obtained from equation (4.1) by substituting r = a together with the general relation E = —ZI. It also follows from Z b of equation (3.1b), if we regard the earth as a tubular conductor with inside radius a whose outside radius b goes to infinity. - 51 - The mutual impedance between the cable and a filamentary insulated conductor at a radial separation R is obtained by substituting r = R. The mutual impedance between two cables with finite radii over their insulation is different from the case of a cable of finite radius and a filament conductor. It can, however, be deduced from the mutual impedance between the cable and a filamentary type conductor by invoking the law of reciprocity of mutual impedance [9]. We have already shown in equation (4.1) that the mutual impedance between a cable of radius o and a filament conductor separated radially by a distance R is given by: pm K (mR) 2na A'j(ma) 0 Z -- (4.2) Hence, from the law of reciprocity of mutual impedance, the electric field strength on the surface of the cable due to a current / in the filament can be written as: E = - pmIK (mR) 0 (1.3) 2za K ( m a ) t Now consider a cable of radius 6 carrying a current / as shown in Figure 4.1. The electric field strength at a point P at a radial distance R is given by: Earth Figure 4-1 - Electric Field Strength al point P pm I'K (mR) 0 27rbK (mb) l The same field strength will be experienced at point P, due to a filament conductor placed along the axis of the cable and carrying a current f, but now the equation will be: pmfK^mR) 2nb(\/mb) (as 6 - 0 , Ki(mb)-\/mb) - 52 - From equations (4.4) and (4.5) we obtain the value of f to be equal to f/mb K (mb). x Therefore, the electric field strength in the soil outside a cable of radius b carrying current / is indistinguishable from that of a current filament placed along the axis of the cable and carry- : ing a current f/mbK^mb). Hence, the electric field strength at the surface of the cable of radius o, which is at a radial distance R from a cable of radius b is found from equation (4.3) by substituting for I the value of f. Therefore, we obtain the field to be: _ P m*fK (mR) . 0 27imaK (ma)mbK (mb) ' ' ' 1 1 Hence, the mutual impedance between two cables of radii o and 6 respectively, buried in a homogeneous infinite earth is given by: pm K (mR) 2 0 27tma Ki(ma)mb K (mb) } - • (''•") Now we can deduce an interesting result. The series expansion of A'j(r) shows that as z-0, K (x)~l/x. 1 Therefore, for small values of \ma\ and |mi|, or in the limiting case of o,6-0 (filament conductors), the self and mutual impedances obtained from equation (4.2) and (4.7) will be given by: Z = -^Koima) ... (4.8a) Z (4.8b) s m = ^~K (mR) 0 It is interesting to know the error if the cable of finite radius is replaced by a filament conductor placed along its axis. For p = 10O— m (low earth resistivity), a = b — 7.5cm (large radii), and a separation of 30 cm, values which perhaps represent a worst case, the errors in the resistance and inductance from equation (4.8b), as compared with equation (4.7), is plotted in Figure 4.2. From Figure 4.2 we see that the approximate formulae have an error of less than 29o, up to \ma\ =? |m6| = 0.1. This happens at a frequency of IMHz. For much lower values, the error is practically negligible. This result is important as it will be used in justifying the extension of formulae for filament conductors to cables of finite radii. 4 . 3 E a r t h R e t u r n Impedance in a Homogeneous Semi-Infinite E a r t h The self and mutual impedances of cables buried in semi-infinite homogenous earth are deduced from the electric field strength in the ground due to a buried filament conductor. - 53 - / / H w 'Y 0) o c 1 1 0 to to cu —I- az -3 I I I! I 10" 10 I 10" 10~ 10" s ! I 1 T T1!| •-' ! '1 - T TTTT i 10"' 2 3 10' 25 15 cu o c \ CO o T 3 C -15 -25 -f 10" 1 1 I 1 Mll| 10" I 1 I I I Mil 10" 1 U I I I I I Hl| 10" 10" 1 1 I 1 I Ml| 1 1 llllll, 10- ma Figure 4-2 - Error in Replacing a Conductor of Finite Radius by a Filament Conductor 10' - 54 The electric field strength in the ground due to a buried insulated filament carrying a current which returns through the soil was first deduced by Pollaciek [l). In fact, he derived formulae for four cases: (1) The electric field strength in the air due to a current-carrying conductor in the air. (2) The electric field strength in the earth due to a current-carrying conductor in the air. (3) The electric field strength in the air due to a current-carrying conductor in the earth. (4) The electric field strength in the earth due to a current-carrying conductor in the earth. The mathematical derivation in all four cases become complicated by the plane of discon- tinuity at the earth's surface. Pollaczek [l] does not discuss the derivations and only mentions that they have been obtained through the reciprocity of Green's functions. Recently Mullineux [10,11,12] obtained expressions for the fields produced in air and earth due to an overhead conductor by using double integral transformation. This technique is equally applicable to buried cable systems [22], and is used in Appendix D to obtain the four types of fields. Comparisons between filamentary type conductors and conductors of finite radii for the infinite earth give every reason to expect that these formulae in Appendix D for filament conductors will be accurate enough for cables of finite radii provided that the condition | ma | <0.1 is satisfied. Hence, for the case of a buried cable at depth h, the electric field strength in the soil resulting from a net current I flowing in the cable is given by equation (D.32(c)) in Appendix D as: E__ — exp[-(o | x m )\h-y\] 2 + - e X P [-(^ + m )\h 2 + y\\ 2(p- + m ~ ) + where g . r exp(-|ftlfyl ' ; VV+ttr) " exp(jd>x )d<t> (4.9) . = horizontal distance between the filament and the point at which the field is being determined, y = depth at which the field is being determined, p = resistivity of the earth, and m = intrinsic propagation constant of the earth. The first integral term is identified as Jfc: (mr?) - rTo(mZ)j [1,22]; where R = V i " + (h-yf 0 and Z = V i s + (/i + y) . 2 The second integral part can be numerically evaluated [28] or - 55 - expanded into an infinite series [22]. The series expansion is in the form of modified Bessel functions. Therefore, equation (4.9) can be written in the series expansion form as follows: = -^~\KdmR) - K (mZ) -f —K^mZ) + ^ "*^Ki\mZ) In \ Z mZ 0 - } (* •+ ml) - ~ ^F(mZ,\x |,/)) mt P (4.10) where ffcVl^T F(mZ,\x\,l)= 2 ^==r\e- dt (4.11) mZt Vl-t 1/2 I 2 J Now the earth return self and mutual impedance terms may be extracted from equation (4.10) or from numerically integrating equation (4.9). The self impedance term is obtained by choosing the coordinates z and y to correspond to the location of the external surface of the cable and the mutual impedance by simply inserting the coordinates of the second cable axis . The numerical integration of equation (4.9) is quite difficult as the solution is highly oscillatory. For example. Figure 4.3 shows the solution for the resistance and inductance of the earth return self impedance at a frequency of 1MHz, for the cable discussed in Appendix A. For such cases, special numerical integration techniques have to be applied [ 2 8 ] . The numerical integration routine available in the U B C system library D C A D R E [ 2 6 ] , which uses a cautious adaptive Romberg extrapolation technique, has been used in obtaining the solution for the equation ( 4 . 9 ) . The solution converges fast for the case when the two cables are buried at d i f ferent depths below the earth, but in the case when the cables are at the same depth or in the case of finding the earth return self impedance, the convergence is rather slow. T o explain this phenomena, consider the first integral part on the right hand side of equation ( 4 . 9 ) , i.e.. exp [-(* 2 + m ) | A - j , | ] - exp[-(cA + m )\h + y\] 2 2 2 ?(c* + m ) 2 z e x p O ^ x )</<•> (4.12) If h = y then the first exponential term within the closed brackets { } will become 1 and the integration now becomes jl- xp[-( (<p c r +m + m)\h+y\ ) | / i + y | jUxp{j<t>x)d<t> 2 e 2 2 - 56 - 10 6- T A o> o c JO V -2 V) o> -6 -10 10' 1 "' '^f 1 I T 10 50 1 I 1 I 1111| 10 2 I i l l I 3 10 4 10 4 30- £ 10 o o c o o D ~ 1 0 \ \ \ / -30 -50 7— ••T-"T "I , , 10' m I I I 1 I I I I i 1CV 10* i i i i i i Interval Figure 4-8 - Solution of Real and Imaginary Part of Equation (4-9), for a Frequency of 1MHz. - 57 - Even though the exponential term within the closed brackets { } approaches zero very fast, we are left with which causes the slow convergence. On the other hand, if h^y then both the first and second exponential terms within the brackets { } of equation (4.12) will approach zero and hence the convergence is faster. Due to the slower convergence, the cpu time taken to compute the self impedance or the mutual impedance in case of two cables buried at the same depth is relatively high. For the self impedance, the computer cost varied between $.50 and $1.00, for one particular frequency. However, the results obtained by applying the D C A D R E integration routine to equation (4.9) and the results obtained from equation (4.10) are almost identical, as shown in Table 4.1 for the mutual impedance between two cables with the following data: depth of cable 1, y = 0.75m depth of cable 2, h = 0.76m radial distance between the two cables = 0.5m earth resistivity p eyAh = 100/7— m. Hence, from Table 4.1 we can see that the equation (4.10), which is the series expansion for the classical equation (4.9) is accurate enough for practical purposes. Therefore, equation (4.10) is taken as the standard equation for finding the self and mutual impedances of buried cables in a semi-infinite earth, and the results obtained by the other formulae proposed by Semlyen [24], Ametani [27], and Wedepohl [22] are compared with respect to it. 4.4 F o r m u l a e U s e d by A m e t a n i , W e d e p o h l a n d Semlyen. Ametani's approach is implemented in BPA's Cable Constant routine of EMTP, and is based on Carson's formulae for overhead conductors. The self and mutual impedances of overhead conductors with earth return effects can be derived from equation ((D.32(b)) Appendix C) together with the relation E — — ZI. In the case of overhead conductor at a height h from the ground, the electric field strength in air at a point (whose height is y and which is at horizontal distance x from the conductor) due to current / flowing in the conductor is given by - 58 - Table 4.1 Solution of Pollaczek'o Equation by Numerical Integration and Using In finite Series FREQUENCY (Hz) NUMERICAL INTEGRATION EQUATION 4.10 RESISTANCE (8/krn) .01 .1 1 10 100 1 ,000 10,000 100,000 1,000,000 10,000,000 0.00000986985 0.00009870394 0.00098721243 0.00987746990 0.09894307000 0.99462669000 10.0999510000 105.026550000 1119.21050000 10416.7130000 0.00000986986 0.00009870399 0.00098721145 0.00987752230 0.09894471000 0.99467682000 10.1014090000 105.062940000 1119.73660000 10410.0480000 REACTANCE (fi/km 0.00014814024 0.00133672137 0.01192043120 0.10477298900 0.90245276590 7.57239400000 61.8788630000 461.037952000 3018.40132100 13049.3294800 .01 .1 1 10 100 1 ,000 10,000 100,000 1,000,000 10,000,000 = — 2x fhere Z - Vi + 2 (A+y) , z ln(Z/R) + J 0.00014814024 0.00133672130 0.01192028400 0.10472980000 I 0.90245110000 ! 7.57234170000 . 61.8625200000 460.988500000 3017.08610000 13031.2620000 (4.15) - 59 - m = intrinsic propagation constant of the earth. The integral part of equation (4.15) can be further simplified to o U | + \ U +• m 2 2 (4.16) Equation (4.16) is widely known as Carson's formula. Strictly speaking, this formula is only valid for the case of overhead conductors. Ametani used this correction term in finding the earth return self and mutual impedances of buried conductors, instead of the second integral term used in equation (4.9). Carson's formula given by equation (4.16) can be numerically integrated [28] or can be expanded into an infinite series [5], in terms of r = | mZ \. Ametani chooses the latter approach. His procedure uses 2 different series, one when r S 5 and the other one when r > 5 . Recently, Shirmohamadi of Ontario Hydro [28] and L. Marti at U B C discovered that the error between the numerical evaluation and the asymptotic expansion is as high as 5-8% as shown in Figure 4.4, for the values of r between 5-10, depending on the value of 8. )000 4000 r«CQUCNCT ( H i ) Figure 4-4 - Relative Error in the Evaluation of Carson's Formula with an Asymptotic Expansion. - 60 - Shirmohamdi avoids this error by using Gauss-Legendre quadrature technique for the direct evaluation of equation (4.13) in this region of r, while L. Marti [24] avoids these errors by extending both the asymptotic and infinite series, and by using a switchover criterion which depends on the geometry of the line (ie. on the value of 0). The use of Carson's formula for underground systems will be reasonably accurate at low frequencies because the value of m with the exponent term exp(—| h+y\ Vc4 +m ) in equation 2 2 2 (4.9) is very small compared to the value of 4> at low frequencies, and, therefore, can be z ignored. A t high frequencies, however, that term becomes quite significant and, therefore, cannot be ignored. For this reason, the resistance and inductance obtained by Ametani's method has an error in the order of 1 0 % or more for frequencies above l K H z when compared with Pollaczek's equation, i.e., equation (4.9). Wedepohl and Wilcox [22], who proposed the infinite series expansion form of equation (4.9) gave an approximation to the infinite series expansion (equation (4.10)) which is valid only if the condition |mZ| <0.25 is satisfied. Their closed-form approximation for the self and mutual earth return impedances are given by: Z, = pm . hHEl + ln 0 2 2n pm •In— 2ir L 5 _ ± + 0.5 2 m h 3 ml (4.17a) (1.17b) 3 where 7 = Euler's constant, h = depth of burial of the conductor, I = sum of the depths of burial of the conductors, R = Vi m = intrinsic propagation constant of the earth. 2 + (h-yf and Semlyen and Wedepohl [24] developed another interesting formula for the self impedance of a cable of radius r in terms of complex depth which is nothing but 1/m, defined here as p. Accordingly, the self impedance term is given by Zs = " ^ M r + V'r) (4.18) - 61 - 4.5 Effect of Displacement Current and Numerical Results So far. the effect of displacement currents has been ignored. As shown in Chapter 2 . there is no noticeable error in the internal impedances of tubular conductor if displacement currents are ignored. Hut unlike in good conductors, the displacement currents in the earth arc noticeable, at least at high frequencies. The displacement current term can be easily incorporated in equation (4-10) by snbsti :ing Vm"-t> /i( for in. Table 4.2 shows the values of J resistance and inclu;'. :nce with and without the displacement currents. Tabic 4.2 Earth Return Self Impedance with and without Displacement Current Term WITH DISPLACEMENT CURRENT FREQUENCY (Hz) WITHOUT DISPLACEMENT CURRENT R E S I S T A N C E (8/ki .01 . 1 1 1 10 100 1 oob 10 0 0 0 10 1 00 000 000 000 ooo 000 0.00000986985 0.00009870393 0.00098720986 0.00987747671 0.09894371230 0.99468700004 10.1056633504 105.572646568 1 172.74925588 15170.9605539 0.00000986986 0.00009870393 0.00098720982 0.00987747340 0.09894339595 0.99465537187 10.1024602926 105.239909564 1136.35180546 11593.2569450 • O O O O O O — O O O O O - ooo- oo- o — — O O O O O O O — — o I N D U C T A N C E ("H/ nm) 2.82478692844 2.59451983737 2. 36423415099 2. 13388974446 1 .90335982756 1.67224525374 1 .43930244333 1 . 2 0 0 8 0 3 8 4 1 19 0.94712614357 0.64980341214 2.82478689712 2.59451979297 2.36423411661 2.13388971332 1.90335979916 1.67224520020 1.43930156514 1.20078329139 0.94694242963 0.66822941718 - 62 The errors in the answers obtained by neglecting the displacment current arc shown in Figure 4.5. The error in the resistance is less than 3% up to a frequency of 1MHz and increases to 209o in the frequency range 1MHz- lOMHr. oi_ k. UJ CD o c -5- to CO CD CC -15 -25 i 10~ 2 i i IIIMI— 10" ni]—i 1 i i inn,—i 10' i i M i n i — i i i inn, 10 2 10 1 i i mii| J 10 nil 4 10 5 1 i i mill 10 6 Frequency [Hz] Figure 4-5 - Error in the Earth Return Self Impedance if the Displacement Current is Ignored ii i mil! io 7 - 63 - Hence, we can neglect the effect of displacement current terms up to a frequency of 1MHz which is well within the limits of practical interest. The earth return self impedance is compared next for the following approaches, with the displacement current term neglected: 1. Pollaczck's original formula, 2. Wedepohl's approximations, - 3. Ametani's approach, 4. Semlyen's approximation. The value of resistance and inductance at different frequencies are tabulated in Table 4.3. Figures 4.6(a) and 4.6(b) illustrate the variation of resistance and inductance in the frequency range 0.001 Hz to 10MHz. The errors are plotted in Figure 4.7 for the same frequency range. Wedepohl's approximation gives an error of less than 1 % up to a frequency of 100kHz, for the resistive part, thereafter it increases steadily. It is around 2 5 % at a frequency 1MHz. error in the inductive part is almost zero up to a frequency of 1MHz. The The reason for the noticeable error in the resistance, beyond a frequency of 100kHz, is that the condition | mZ | <0.25 is violated. Semlyen's approximation is good at low frequencies for the resistive part but at high frequencies the error is higher. It has an error of around 4 % in the case of inductive part over the whole frequency range. As mentioned earlier, the error i i i Ametani's procedure is not significant at low frequencies, but increases from 2 % to 2 0 % iu (.-be frequency range 10kHz to 1MHz Three as shown in Figure 4.7. routines were developed for the calculation of earth return self and mutual impedances with Pollaczek's original formula. The part which is .difficult to evaluate is the integral term F(mZ,\x\,l) (equation (4.11)) in equation (4.10).. This part can either be expanded into a series with a suitable number of terms and each term can then be integrated, or a suitable library subroutine for numerical integration can be used. Routine SEARTH developed by the author uses the first approach by considering 15 terms. Routine L E A R T H developed by Luis Marti uses the UBC library subroutine D C A D R E for the evaluation of the integral. One more routine, namely C E A R T H was developed by the author, which can take the displacement current into account. This routine also uses the UBC DRE library subroutine DCA- to evaluate the function discussed earlier. Routines S E A R T H and L E A R T H give identical answers but differ in the cpu time. The former one takes 3.4 ms while L E A R T H takes 2.9 ms for the evaluation of resistance and inductance at a particular frequency. Routine C E A R T H takes 23.0 ms for the same evaluation. The routine for Ametani's approach takes 5.30 ms, while the routines for Wedepohl's and Semlyen's approximations take 0.30 ms and 0.24 ms of cpu time, respectively. Table 4.S Earth Return Self Impedance as a Function Frequency FREQUENCY P O L L A C Z E K WEDEPOHL AMETANI SEMLYEN (Hz) R E ; > I S T A N C E ( £ 2 / 1cm) .01 0 . 0 0 9 8 8 2 5 0 . 0 0 9 8 8 2 5 0 . 0 0 9 8 7 7 5 0 . 0 0 9 8 8 2 5 .01 0 . 0 0 0 0 0 9 9 0 . 0 0 0 0 0 9 9 0 . 0 0 0 0 0 9 9 0 . 0 0 0 0 0 9 9 .1 0 . 0 0 0 0 9 8 7 0 . 0 0 0 0 9 8 7 0 . 0 0 0 0 9 8 7 0 . 0 0 0 0 9 8 7 1 0 . 0 0 0 9 8 7 2 0 . 0 0 0 9 8 7 2 0 . 0 0 0 9 8 6 7 0 . 0 0 0 9 8 7 0 10 0 . 0 0 9 8 7 7 5 0 . 0 0 9 8 7 7 5 0 . 0 0 9 8 6 1 8 0 . 0 0 9 8 6 9 2 100 0 . 0 9 8 9 4 3 4 0 . 0 9 8 9 4 5 7 0 . 0 9 8 4 5 1 9 0 . 0 9 8 6 8 4 0 0 . 9 7 9 5 2 6 3 0 . 9 8 6 5 7 8 4 1000 0 . 9 9 4 6 5 5 4 0 . 9 9 4 8 5 6 1 10000 1 0 . 1 0 2 4 6 0 1 0 . 1 1 9 2 8 8 9 . 6 5 6 4 6 8 0 9 . 8 5 7 5 3 1 3 1 0 0 0 0 0 105.23991 1 0 6 . 5 9 1 7 2 9 3 . 4 9 8 2 7 0 9 8 . 3 1 5 0 5 3 1 0 0 0 0 0 0 1 1 3 6 . 3 5 1 8 1 2 3 6 . 6 4 3 8 9 1 3 . 0 3 2 1 4 9 7 4 . 9 9 1 2 3 1 0 0 0 0 0 0 0 1 1 5 9 3 . 2 5 7 1 7 7 6 5 . 2 8 8 1 0 8 2 6 . 0 5 8 9 4 9 8 . 8 3 9 4 I N IX J C T A N C E (mHy'km) 2 . 8 2 4 7 8 6 9 2 . 8 2 4 7 8 3 8 2 . 8 2 4 7 9 1 7 2 . 7 0 1 6 0 4 8 .1 2 . 5 9 4 5 1 9 8 2 . 5 9 4 5 1 6 7 2 . 5 9 4 5 4 1 8 2 . 4 7 1 3 4 6 7 1 2.3642341 2 . 3 6 4 2 3 1 0 2 . 3 6 4 3 1 0 5 2 . 2 4 1 0 8 9 5 10 2 . 1 3 3 8 8 9 7 2 . 1 3 3 8 8 6 6 2 . 1 3 4 1 3 7 8 2 . 0 1 0 8 3 5 2 1 . 7 8 0 5 8 9 8 .01 , 100 1 . 9 0 3 3 5 9 8 1 . 9 0 3 3 5 6 4 1.9041501 1000 1.6722452 1 . 6 7 2 2 3 8 6 1 . 6 7 4 7 4 1 4 1 . 5 5 0 3 7 2 9 10000 1 . 4 3 9 3 0 1 6 1 . 4 3 9 2 6 2 9 1 . 4 4 7 1 0 6 4 1 . 3 2 0 2 4 5 9 1 . 2 0 0 7 8 3 3 1 . 2 0 0 4 1 1 8 1 . 2 2 4 5 1 5 7 1 . 0 9 0 4 0 3 2 1 0 0 0 0 0 0 0 . 9 4 6 9 4 2 4 0 . 9 4 2 9 8 1 3 1 . 0 1 2 6 2 2 8 0 . 8 6 1 4 5 9 7 1 0 0 0 0 0 0 0 0 . 6 6 8 2 2 9 4 0 . 6 2 6 7 9 7 5 0 . 7 9 6 2 6 8 9 0 . 6 3 5 3 5 6 5 1 0 0 0 0 0 4.6 C a b l e s B u r l e d at Depth Greater than Depth of Penetration If the depth of burial is greater than the earth return current's depth of penetration, or other words, if the distribution of return current is circularly symmetrical, then the cable c be considered to be buried in an infinite earth. In practice, this can arise in two situations: 1. Cables arc buried at large depths below the ground, \ - 05 - POLLACZEK WEDEPOHL AMETANI SEMLYEN 0 _,j " 10 J •• , , , , „ „ 10" 1 .... r-^>..nr 10' io 2 10 Frequency [Hz] Figure 4.6(a),(b) t 10' Earth Return Self Impedance as a Function of Frequency 10 5 10 s 10 7 - 66 - 25 WEDEPOHL 15- / / AMETANI SEMLYEN 5- ui o <o CO -C -5. in CD or -15 — 25 "1—1 10" i 1111111—1 1 10"' 2 11111 n 1 1 1 11 IIIII 1 1 1111111—1 1 10' 10 1 . .1.11 10 2 11 10 J 11 . null—1 10 4 5 mii| 10 111 6 iiui| 10 1 7 25 WEDEPOHL 15- AMETANI SEMLYEN 5LU CD o c ra > -5 o T3 C — -15 ~f 10~ 11 —25 2 T mrn 10"' r-i TTTTTTI 1. 1 i 11 ini[ 10' 1 i i itui] 10 r-rri-TTTTi 2 10 r—r-rrrmj r iitini] r i i rim\ 10 10 10 J 4 Frequency [Hz] Figure 4- 7 Errors in Earth Return Self Impedance S 6 r-rrmrt} 10 7 11 mir - 2. 67 - C a b l e s are buried at normal depths ( l - 2 m ) b u t are used at high frequencies (100kHz a n d above). I n s u c h s i t u a t i o n s , t h e d e p t h o f p e n e t r a t i o n i n t h e e a r t h is g i v e n b y : d = 503.3\/pe«u// (4.19) ( w h e r e d is i n m , / is i n fi—m a n d / is i n H z ) and becomes smaller than the depth of burial. The s e c o n d possibility does n o t arise normally underground transmission system, with an earth lm, the frequency than l m is 3 M H z at which the penetration o r higher. cases d o arise, however, in power system studies. For a typical r e s i s t i v i t y o f 10 J ? — m , a n d a b u r i a l d e p t h o f of the return current in the earth b e c o m e s less In p o w e r s y s t e m s , o n e r a r e l y e n c o u n t e r s s u c h f r e q u e n c i e s . the infinite earth return impedance formulae If s u c h given by equation (4.2) a n d ( 4 . 7 ) c o u l d b e u s e d t o find t h e e a r t h r e t u r n i m p e d a n c e s . It is i n t e r e s t i n g t o k n o w t h e b u r i a l d e p t h is l a r g e . If t h e e a r t h equation cable e q u a t i o n (4.10) for the semi-infinite ( 4 . 1 9 ) w i l l be less t h a n at c a s e is s t i l l v a l i d if T h i s c a n be c h e c k e d as follows: r e s i s t a n c e is a s s u m e d t o b e 1 0 17— m, is b u r i e d mutual whether a depth impedances, using then the depth of penetration 5.5 m , f o r f r e q u e n c i e s 0.1 M H z a n d higher. of 5.5m, then the values obtained the (4.2) equations o r (4.7), for earth should o b t a i n e d u s i n g e q u a t i o n (4.10) f o r f r e q u e n c i e s a b o v e 0 . 1 M H z . given by H e n c e , if a return be t h e s a m e self a n d as those F i g u r e 4.8 s h o w s t h e d i f f e r - ence i n t h e v a l u e o f r e s i s t a n c e o b t a i n e d b y P o l l a c z e k ' s f o r m u l a a n d e q u a t i o n (4.2) f o r t h e earth return self decreases from to 1.59in. impedance in 1 9 % t o less t h a n the frequency range 10kHz to 1 % while t h e depth of penetration H e n c e , it a p p e a r s t h a t Pollaczek's formula 1MHz. The difference decreases from 15.9m is c o r r e c t e v e n a t l a r g e d e p t h s o f b u r i a l e v e n t h o u g h it is b e t t e r t o u s e t h e e q u a t i o n s ( 4 . 2 ) a n d ( 4 . 7 ) f o r t h e c a s e o f i n f i n i t e earth [9,22]. Subroutine T U B E cannot be u s e d author has options has an option for for finding for the finding mutual both finding t h e infinite e a r t h r e t u r n self i m p e d a n c e , b u t it impedance. self a n d m u t u a l T h e routine T U B E C developed impedances, a n d c a n take by the displacement c u r r e n t s i n t o a c c o u n t as w e l l . 4.7 Mutual Impedance or Vice Between a C a b l e B u r i e d in the E a r t h and an Overhead Line Versa A n o t h e r impedance of interest is t h e m u t u a l T h e electric impedance field strengths between t o p o w e r e n g i n e e r s as well a s t o c o m m u n i c a t i o n e n g i n e e r s an u n d e r g r o u n d cable a n d a n overhead line o r vice v e r s a . in air d u e t o a c u r r e n t c a r r y i n g c o n d u c t o r buried in the earth - 68 - 2 0 1 2 - cu u c CD "* 4- CU u c -4 CO to CO cc -12 -i—i—• I + - 2 0 10 10 4 s I I 10 6 F r q u e n c y [Hz] Figure 4.8 Differences in Resistance Values of Semi-Infinite and In finite Earth Return Formulae or the field Zs_+ in earth due to a current-carrying conductor in the air is given by Equations (D.32(b)) and (D.32(d)), respectively, in Appendix D. In both cases the mutual impedance is given by: 0 e x p j - / i | c*| -dV<t>2+ mj 2 exp(j>| x | )d<f> | c6| + V(t> +m 2 2 •where A = height of the conductor in air, d = depth of burial of the buried conductor, |i| = the horizontal distance between the conductors, m = intrinsic propagation constant of the earth. This integral can be evaluated in terms of infinite series in somewhat the same way as was done for E + + in [4,27] and for E _ _ in [22]. - 69 - 4.8 C o n c l u s i o n T o summarize, the self and mutual impedances of conductors with earth return were derived for two situations, namely for 1. Cables buried in infinite ' • earth, and for 2. Cables buried in semi-infinite earth. The impedances were obtained from the axial electric field strengths in the earth due to return currents in the ground. These electric field strengths were derived from Maxwell's equations, for filamentary type conductors of negligible radius. Since we were interested in cables of finite radius, the solutions for filamentary type conductors were extended to cables of finite radius. The solutions for the earth return impedance with semi-infinite earth is in infinite integral form. Wedepohl [22] transformed this infinite integral equation into an equation consisting of Bessei functions. It was found that the values obtained from the numerical integration of the infinite integral and from Wedepohl's transformation were very close. Ametani's approach for finding earth return impedances which is implemented in Cables Constants routine in the BPA's E M T P and other approximations suitable for hand calculations were compared for typical cable data. Ametani's approach gave erroneous results at high frequencies due to an erroneous assumption. Wedepohl's approximation was found to give reasonably accurate answers and is well suited for hand calculations. A t the end of the chapter, the evaluation of mutual impedance between a buried conductor and overhead conductor, and vice versa, is briefly discussed. - 70 - 5. Laminated Tubular Conductors In Chapter 3, formulae for internal impedances of homogeneous tubular conductors were derived. These formulae are used in this chapter to obtain the impedances of cables whose core and sheath are made up of laminated conductors of different materials. A practical application of this type of conductor was recently proposed by Harrington [32]. He suggested that the transient sheath voltage rise in a gas-insulated substation can be reduced by coating the conductor and sheath surfaces with high-permeability materials, thereby increasing the impedance of the surfaces for surge propagation, which in turn will damp out high frequency transients. 5.1 I n t e r n a l Impedances of a L a m i n a t e d T u b u l a r Conductor The internal impedances needed for laminated conductors are the same as those needed for homogeneous conductors, namely: The 1. internal impedance z 0(J of the laminated tubular conductor which gives the voltage drop on the inner surface when unit current returns through a conductor inside the tube. 2. The internal impedance z bb of the laminated tubular conductor which gives the voltage drop on the outer surface when unit current returns through a conductor outside the tube. 5 . 1 . 1 I n t e r n a l Impedance with E x t e r n a l R e t u r n Let us first number the layers consecutively with the inner most layer being number 1 as shown in Figure 5.1. For the analysis, we start with the mth outer most layer shown in Figure 5.2. Let = internal impedance of the mth layer with current returning inside, = internal impedance of the mth layer with current returning outside, Z^ z b = mutual impedance between the two surfaces, m h — internal impedance of all m layers when the current return is external For the very first layer, we note that Z [^ = z$\ b Chapter 3, then the loop current 7 _, m surface of the mth If we use concentric loop currents as before in of the first m —1 layer, while loop current I m layers combined, returns on the inner flows on the outer surface. Using Schelkunoff's theorem 2 from Appendix B, the electric field strength along the inner surface of the mth layer becomes - 71 - Axis Layer (m-l) Layer m Figure 5.1 Numbering of Conductor Layers to Find the Internal Impedances of a Laminated Tubular Conductor r m Figure 5.2 Representation of the mth Layer dV dx — (ZTb lm %aa 'm-l) (5.1) But the inner surface of the mth layer is the outer surface of the first m - l layers combined for which the electric field strength is given by . ~ z bi 'm-l (5.2) dx Therefore we can find a relationship between I„ and /„_, from equations (5.1) and (5.2), (5.3) lm + Zli ' - 72 - Now let us consider the electric Held strength on the outer surface of the mth layer. O n one hand it is — z^I , and on the other hand it is —(Z I — Z ^ / _ i ) using Schclkunoff's m m b t m m theorem 2. Thus we have the following identity: m Zbb m %bb ~ = m ^ %ab m ~~j ~ * m Substituting for 7 _i// m . ' (5.4) ' from equation (5.3), we obtain m • m _ m J^i! '/<:«;» 7 Zbb ~ *bb , m •^aa m - (o-5) x f»6 + which gives the internal impedance of all m layers of the laminated tubular conductor, with Starting with the first layer where z$ = Z y, we add the current return on the outside. b remaining layers one by one until we obtain the impedance of the complete laminated conductor made up of m layers. Ab = ZU - „ (Zlb? Z'at + , (5.6) , • = 2, • • • m Ab 5.1.2 I n t e r n a l I m p e d a n c e w i t h I n t e r n a l Return Similarly we can find the internal impedance of a laminated current returning inside. Let Z™, Z m b h tubular conductor with and Z™ related to the same internal impedances defined b in the previous section. Let z* be the internal impedance of all m layers when the current a Also, we note that for the very last layer, i.e. layer m in Figure 5.1, return is internal. ZTa = z Tn- Using Schelkunoff's theorem 2, we find the electric field strength along the outer surface of the 1st layer as = ~(Z \h ~ Z,\lo) b (5-7) But the outer surface of the 1st layer is the inner surface of the rn — 1 outer layers combined, for which the electric field strength can be written as 1 = (5-8) Therefore, we can find a relationship between I and I from equations (5.7) and (5.8), 0 /, z\ h Z \ + z- x a (5.9) t Now, consider the electric field strength on the inner surface of the first layer. O n one hand it is —{ — z \lo)t a a n d o n t n e other hand it is — (Z^/i Therefore we have the following identity, — Z}J ) using Schellkunoffs theorem 2. 0 - = ZL 73 - ~ Z^-j'o (5.10) Substituting for / j / / from equation (5.9) we obtain 0 i _ ._ aa ~ Z z aa (^)2 _] ^bb ,5.11) 2 + z at which gives the internal impedance of all m layers of the laminated tubular conductor, with current return on the inside. Starting with the last layer, i.e. layer m, where z™ = Z™, v.v add the remaining layers one by one until the impedance of the complete conductor made up of m layers is obtained, ) IZ' = -Zla ~ _, ^lb 2 '* + • » = m-l,m-2 -l. > (5.12) aa z 5.2 A p p l i c a t i o n t o G a s - I n s u l a t e d S u b s t a t i o n s The equations for the internal impedance of laminated conductors will now obtain the surge propagation be used to characteristics in a gas-insulated substation with conductor coatings. Gas-insulated substations are subjected to transient sheath voltage rises whenever switchings or fault surges occur. These surges propagate along the outer surface of the inner conductor and the inner surface of the sheath, as if the two surfaces were cylindrical wave guides, as well as along the outer surface of the sheath and the ground. The impedances of these surfaces play an important role in attenuating the surges, and thereby the transient sheath voltage rise. Since these surface impedances depend on the resistivity and the magnetic permeability of the material, it has been proposed by Harrington [32], to coat these surfaces with material of high resistivity and high permeability for surge suppression purposes. The coating should be such that its thickness is less than its current penetration depth at power frequency (60Hz or 50Hz), so that the resistance is not changed during steady-state operation. In addition to the base case without coatings, three different coating configurations are examined. The four cases considered are as follows: Core and sheath not coated. i. Only the inner surface of the sheath coated. ii. Only the outer surface of the inner core coated. v. Both the outer surface of the inner core as well as the inner surface of the sheath coated. For each of these cases, the formulae for the impedances for surge propagation arc derived. 5.2.1 C A S E i : C o r e and Sheath not Coated Axis Core S h c a l h Earth v/ssssss//////////////////////J^ \ J , % Figure 5.3 - Core and Sheath not J Coated This is the simplest of all the cases where the impedance for surge propagation in loop 1 is given by Z = Z :rc + Z, + Z ns (5.13) iht where Zcre ( core - with external return) can be obtaind from equation (3.3) if the core conductor is solid, or from equation (3.1b) if it is tubular. Z M S can be obtained from equation (2.11), and Z , (sheath - with internal return) from equation (3;.la). ih 5.2.2 C A S E ii: O n l y Sheath C o a t e d In this case the surface for the surge propagation consists of the outer surface of the core conductor and the inner surface, of the laminated conductor made up of coating paint layer and sheath. Hence, the impedance for surge propagation between core and sheath is given by Z = Z cre + Z, + zll ns where Zcre and Z,ni are the same as explained for case i. (5.14) zll is the internal impedance of the laminated conductor with internal return. This is obtained from equation (5.12} where layer 1 is the sheath and layer 2 is the coating material (superscipt "sp" denotes the paint layer on sheath and superscript "sh" denotes the sheath). Hence - 75 - Axis Core Paint Sheath Earth Figure 5.4 Inner Surface of the Sheath only Coated. sp\2 y*V T V - _ (5.15) + z,sk Zll The total impedance Z can then be written as Z —Z :rc +Z ins + Z , ip (Z spm ) 2 (5.16) z pc Z , s ik where Z , (sheath coated with paint layer - with internal return) and Z , (sheath - with intersp sK nal return) can be obtained from equation (3.1a). Z ipm paint layer) is found from equation (3.1c), and Z spe (mutual between sheath conductor and (sheath coated with paint layer - with exter- nal return) from equation (3.1b). Equation (5.16) can also be derived from the loop equations of the loops 1,2 ( f i g u r e 5.4), = -(zj, dx oT's = ~(Z I + n ~di~ i: + ZM ZI) 2 (5.17a,b) 2 where V, = potential difference between the core and paint v 2 = potential difference between the paint layer and sheath Z s "** Z , Zi m z — Zi Z — Z m 2 2 ipt sp Z m sp + z , $k Since the paint layer and the sheath are at the same potential we have V = 0. Hence from 2 equation (5.17b), - I 2 = (-Z m /Z )/ 2 Substituting the value of I 2 76 - (5.18) 1 into equation (5.16a) gives dV, (5.19) Therefore the impedance for surge propagation between core and sheath is given by z = z, - — or _ J- 7 _L 7 7 - 7 ( ^ s y m w h i c h is i d e n t i c a l w i t h e q u a t i o n 5 . 1 6 . 5.2.3 C A S E iii: O n l y C o r e Coated Axis 1, Sheath / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / Earth | Figure 5.5 Core Alone Coated The surface for surge propagation in this case consists of the surface of the conductor (made u p of core and the coating) a n d the inner surface of the laminated sheath. The i m p e d a n c e f o r s u r g e p r o p a g a t i o n b e t w e e n c o r e a n d s h e a t h is t h e r e f o r e g i v e n b y Z = where Z , , a n d A z$ + Z nt Z int + Z a (5.20) , are the s a m e i m p e d a n c e s explained earlier. the laminated conductor with external return. z$ is t h e i n t e r n a l i m p e d a n c e o f T h i s c a n be o b t a i n e d f r o m e q u a t i o n (5.6), w h e r e l a y e r 1 is t h e c o r e a n d l a y e r 2 is t h e p a i n t o n c o r e , Substituting ztf into equation (5.20) gives (Z Z = Z where Z cre cre - ) 2 7 + Z, + Z, (5.22) tk ns is the same as explained earlier. Z^ (core coated with paint layer - with external return) can be obtained from equation (3.1b). Z cpm ductor) from equation (3.1c), and Z Cfi (mutual between paint layer and core con- (core coated with paint layer - with internal return) from equation (3.1 a) 5.2.4 C A S E iv: B o t h Core and Sheath Coated Axis Paint Sheath Earth Figure 5.6 Core aa well as Inner Surface of the Sheath Coated In this case, the outside of the core and the inside of the sheath arc coated. The surfaces for surge propagation consist of.the surface of the laminated conductor 1 (made up of the core and coating paint layer on the core conductor) and the inner surface of the laminated conductor 2 (made up of the sheath and the coating paint layer on the sheath conductor). The impedance for surge propagation between core and sheath is given by Z = z% + Z ml + z?. Using the values of z$ and z\\ from equations (5.15) and (5.21) Z can be written as z z cpc + z, RS + Z ipt where all the impedance values have been explained earlier. (5.24) - 78 - Results Two coating materials arc considered, i.e., stainless steel and supermalloy. These routing materials would be applied iu the form of paints. The bus duct of the gus-"insuiatccl substation is assumed to have the dimensions given in Figure 5.7. Axis -ORCR . IRCR -IRSH C ORSH Sheath w/////r7/777777777777777777> \/77//7777777777777777//777777/, Earth j Figure 5,7 Dimensions of the Bus Duct in a Gas-htsulatcd Substation Inner radius of the core, IRCR = 10.0mm Outer radius of the core, O R C R = 65.0mm Inner radius of the sheath, IRSH = Outer radius of .the sheath, O R S H = 350mm 380mm The values of relative permeability and resistivity of the coating materials (stainless steel and supermalloy) and of the core and sheath material (aluminium) are given in Table 5.1. 5.2.5 Stainless Steel Coating The skin depth for a particular material is given by 6 = V2p/M/i (5.25) -79 - Table 6.1 Resistivity and Relative Permeability of Coating Materials RELATIVE TYPE MATERIAL Core and Sheath Aluminum PERMEABILITY RESISTIVITY 1.0 2.62E-08 1500.0 4.70E-07 Paint (a) Stainless Steel (b) Supermalloy 100 000.00 0.00E-07 Using the values of relative permeability and resistivity for "stainless steel, the skin depth at various frequencies was calculated and tabulated in Table 5.2. Table 5.2 Skin Depth of Stainless Steel FREQUENCY (Hz) SKIN D E P T H (mm) 10.0 2.81723 60.0 1.15013 100.0 ; 0.89089 1 000.0 0.28172 10 000.0 0.08909 Since the thickness of the coating should be very much smaller than the skin depth cf the material at normal operating frequency, coating thickness of 0.1mm and 0.5mm were assumed to be practical values. Figures 5.8(a) and (b) show the variation in resistance and inductance for a coating thickness of 0.1mm and Figures 5.9(a) and (b) for a coating thickness of 0.5mm. — 6U " 100 CASE 1 CASC 2 10 E CASE_3 CASE 4 c in a) rr 0.1 0.01 0.001 1CT 10' 10" 10 10 3 10' 3 10 10' s 10' Frequency [Hz] \ O.B- \ \ w x CAST 1 CASE.. 2 0.6 C_ASE_3 CASE ^ c o o ~o c o.* 0-2. 0-+ 10" 10' • 10- 10 J 10 5 10' •10 s 10* Frequency [Hz] Figure 5.8(a),(b) Variation of Resistance •with Frequency for the Four Stainless Steel Coaling, Thickness and Inductance Cases; 0.1mm. 10' - 82 - 5.2.6 S u p e r m a l l o y C o a t i n g The high resistivity and high permeability of supermalloy make its skin depth very small even at low frequencies, as shown in Table 5.3. T a b l e 5.3 Skin Depth of Supermalloy FREQUENCY (Hz) SKIN D E P T H (mm) 1.0 1.23281 10.0 0.38985 60.0 0.15915 100.0 0.12328 * 1 000.0 0.03898 10 000.0 0.01233 Since the coating thickness should be smaller than the skin depth at normal operating frequency, it would be necessary to keep the coating thickness to less than 0.1mm. Figures 5.10(a) and (b) show the variation of resistance and inductance with frequency for a coating thickness of 0.01 mm and Figures 5.11(a) and (b) for a coating thickness of 0.05mm respec- tively. 5.2.7 C o m p a r i s o n between Stainless Steel a n d S u p e r m a l l o y Coatings For the case of stainless steel, we note from the figures 5.10(a) that there is no noticeable difference in the resistance up to a frequency of 100Hz for all four cases if the coating thickness is 0.1mm. Beyond that it increases sharply for cases 2, 3 and 4 as compared to the base case. When the coating thickness is increased-to 0.5mm, the differences are pronounced at frequencies as low as 1Hz, as shown in Figure 5.11(a). This indicates that the coating thickness should not be increased beyond 0.1mm, since it would change the resistance at steady state operating frequency (50Hz or 60Hz) too much, and thereby increase the losses as well as the operating temperature. Due to the high permeability of stainless steel, the inductance is very high for cases 2, 3 and 4 as compared to case 1, as shown in Figures 5.10(b) and 5.11(b). However, the increase in inductance should not cause any problems in bus ducts which are very short compared to the length of transmission lines. - 83 - \ <2 £ ^ \ i- ~ \ CASE 1 \ V 3 CAse \ c o y 5 •o c X 2 2 L IL . A \ 3 CASE i \ 10" 2 10" 1 10' 10 J 10 5 10." 10 5 10' Frequency [Hz] Figure 5.10 (a) and (b) Variation of Resistance and Inductance with Frequency for the Four Cases; Supermalloy Coating, Thickness 0.01mm. 10 - 84 - 1000CH 1000- 100- E CASE 1 10- CASE 2 u c D tn ca 0.1. 0.01. 0.001 1CT J 1 10"' 10' 10 10 2 3 10' 10 10 5 s 10' Frequency [Hz] 20- \ X 15- \ E CASE 1 CASE 2 <u u \ 10- c o "o 3 "D C 0 CASE_3 CASE * + 10" 10" 10' .10* 10 J 10" 10* 10 Frequency [Hz] Figure 5.1l(a),(b) Variation of Resistance and Inductance with Frequency for the Four Cases; Supermalloy Coating, Thickness 0.05mm. e 10' - 85 In the case of supermalloy, due to its higher resistivity and very lur^e permeability the coating thickness should not be increased beyond 0.01mm for the same reasons explained earlier for stainless steel. The practicality of stainless steel or supermalloy coatings for surge suppression has been questioned by Boggs and Fujimoto [32]. Such coatings may, be cost effective. Using highly resistive materials such as steel for the entire sheath has been considered as well. This would be feasible with single-point ground, which would prevent currents from circulating through the sheath, thereby avoiding sheath losses. However, single-point grounding has adverse implications for transient ground rise, however. If switching surges are produced, transient overvoltages would appear at many points within the gas-insulated substation. 5.3 C o n c l u s i o n s The internal impedances of tubular laminated conductors have been derived. These equations are used to find the internal impedances of bus ducts in gas-insulated substations whose core and/or sheath are coated with high-resistivity paints for the suppression of surges. - 86 - 6. TEST CASES The internal formulae up the are impedance formulae were discussed in detail elements obtained for tubular i n C h a p t e r s 3 a n d 4, r e s p e c t i v e l y . [Z \. of submatrices [Z„] a n d for a specific conductors and the e a r t h return In i} underground cable These impedances this chapter the values of these s y s t e m , using the impedance exact make submatrices formulae as well as approximations. The approximate formulae tion formulae proposed by Wedepohl a n d t a k e v e r y little c p u t i m e . pruposes. Therefore sidered in this chapter. only the The agree are formulae proposed by W e d e p o h l also c o m p a r e d w i t h values o b t a i n e d C o n s t a n t s routine in the E M T P , w h i c h was d e v e l o p e d by A m e t a n i 6.1 Single C o r e m = w h e r e all e l e m e n t s Appendix Wedepohl's A, were defined earlier the values approximation in T a b l e 6.1. F i g u r e s 6.1 l a e f o r t h e i m p e d a n c e Z. . c Z. s and Z ss The cc and Z Z Ci S! return formulae errors of these formulae This stant m, from are the conCable [27]. were W i t h the d a t a of the test case d e s c r i b e d obtained becomes larger, from the exact formulae, as a n d 6.3, respectively, show the errors for the and Wedepohl's formulae are We not significant from tabulated errors in A m e t a n i ' s a n d W e d e p o h l ' s a p p r o x i m a t e all have s i m i l a r v a l u e s . is o n l y t r u e a t form and from Ametani's Cable Constant routine, F i g u r e s 6.2 used by A m e t a n i m a t r i x of the (6.1) in C h a p t e r 3. elements depicts the in A m e t a n i ' s Wedepohl's approximate fied. [22] z. q u e n c i e s , b u t for h i g h e r f r e q u e n c i e s t h e y c a n n o t be n e g l e c t e d . Z , exact Cable T h e i m p e d a n c e s o f a s i n g l e c o r e c a b l e a r e g i v e n b y a 2X2 in very closely with the T h e y also provide simple expressions for h a n d calcula- approximate results [22] formu- impedances at low also notice t h a t the errors T h e y are essentially c r e a t e d b y the e r r o r s in the a n d W e d e p o h l , as s h o w n i n F i g u r e 4.7. A s mentioned e a r t h r e t u r n f o r m u l a is v a l i d o n l y i f t h e c o n d i t i o n low f r e q u e n c i e s . At high frequencies, the a n d this c a u s e s the e r r o r s in the r e s u l t s . the separation between the cables becomes intrinsic in earth earlier, \ mZ \ <0.25 is propagation satiscon- T h e errors also increase if larger. T h e reasons for the errors in A m e t a n i ' s e a r t h r e t u r n i m p e d a n c e f o r m u l a e cies has a l r e a d y been d i s c u s s e d in C h a p t e r fre- 4. at h i g h frequen- - 87 T a b l e 8.1 Impedances of a Single Core Underground Cable. Zee (fi/km) F R E Q (Hz) 1 10 100 1000 10000 100000 R .010873 .020084 .119303 1.05509 10.4803 108.240 WEDE: P O H L AME1' ' A N I EX; R X .016082 .146299 1 .30456 1 1 .4759 99.6843 839.848 X .010873 .020069 .118805 1.03996 10.0343 96.4973 .016083 .146315 I . 30504 I I . 4916 100.175 854.761 R X .010868 .016093 .020081 .146416 . 120114 I . 30506 1.05626 I I . 4759 10.4983 99.6822 109.593 839.614 Zcs (£2/ km) 1 10 100 iooo 10000 100000 .000987 .009878 .098954 .995717 10.2001 106.430 .015097 .136501 1.22016 10.7494 92.8295 775.524 i 1 10 100 1000 10000 100000 .300151 .309041 .398112 1.29438 10.4531 106.361 .015083 .136354 1.21869 10.7347 92.6969 775.518 .000987 .009862 .098465 .980590 9.75414 94.6876 Zss .015098 .136516 1.22066 10.7651 93.3201 730.438 .000987 .009878 .098956 .995919 10.2170 107.782 .015097 .136501 1.22016 10.7494 92.8272 775.291 .300151 .309041 .398115 1.29458 10.4700 107.713 .015083 .136354 1 .21869 10.7347 92.6946 775.285 Wlim) .300150 .309025 .397612 1 .27925 10.0071 94.6180 .015083 .136369 1.21919 10.7504 93.1875 790.432 - 88 - 50 30- / 7 10- u c « -10 cn io cu rr AMETANI -30 —50 WEDEPOHL i \ 10~ i i Mini 10" 2 i i i nun —n TTTTTTTJ—i 1 nr T I itiiTf 10' 10 2 TTHTTT; 10 T-TTTTTTT] r 10 3 i i mit| 10 4 "i T T mn| " i 10 5 i ritin 10 6 7 20 12 4- 01 o c \ -4 to u ca cu CC AMETANI -12 WEDEPOHL -20 1 10" 2 III Mini 10"' i i i mill 1 i i i i i i mm IIIIII 10 i i i iiuii—i 10 1 2 10 i 5 11mil 10 i i 4 11iui| 10 s i i 11mil 10 i 6 i 11mi 10 7 Frquency [Hz] Figure 6.1 - Errors in Ametani's and Wedepohl's Approximations in Z cc - 89 - 50 30- / / 100) u c tncu -10- tc AMETANI -30- WEDEPOHL 1 T ITTIfH ' I '1 M l l T i r — 50 10" I 'T I 1 i m ] ' ~ T ' T T T T m r 1 10"' 2 10' T"T T^TIITf 10 2 I'TlTnil}' 10 1 I 1 11'ltT) 10 3 4 1 "TTTTnTJ 10 s I 10 I T'H'ffl 10 6 7 20 12 cu o c CO *-> \ - 4 - o ra CD CC AMETANI -12- — 20 WEDEPOHL | i 10" 2 i 11 — i i i IIIIII 10"' 1 I i i Mini 1 i . IIIIII—i 1 i i mill 10' 10 2 1 i I M i n i — i i i Mini 10 J 10 4 10 5 i i utii| i 10 6 l l Mini 10 7 Frquency [Hz] Figure 6.2 - Errors in Ametani 'a and Wedepohl's Approximations in Z a - 90 - 50 30- / / 10 LU co u « +—< -10 m cc AMETANI - 3 0 - WEDEPOHL —50 | III TO" 2 IIIIII i i i 10"' IIIIII—i 1 i i i i 11uii| IIIUI 10' 10 2 i i 11mil 10 I i 11uii|—> i i 10" 3 10 IIIIII 5 I I I IIIIII—i 10 i i inn 10 6 1 7 20 CD u c to o co CD at AMETANI -12 WEDEPOHL —20 I 10" I I I niM| 2 10"' i i i iiiii, 1 rrm] 10' 1 iiiinij 10 i i I I I I I I ; I 2 10 I I I I I M I 3 10" I I I I I I I I ) I 10 S I I I I I I I I 10 6 I i i m i q 10 7 Frquency [Hz] Figure 6.3 - Errors in Ametani's and Wedepohl's Approximations in Z, - 91 6.2 T h r e e - P h a s e C a b l e In the case of a three-phase cable system, the series impedance matrix is given by \Zn\ |Z ] \Z ) 12 K (0.2) 1*1\2A \ *\ z where [Z„] is the self impedance submatrix of cable i as given by equation (0.1). The mutual impedances between cable i and cable j are represented by submatrix |Z ] of the form: t; Z- . Z Zj.ry Z . . .f r r • (0.3) s s As shown in Chapter 2, all four.elements of submatrix are equal to each other. The values from the exact formula (4.10) and from Ametani's and Wedepohl's approximate formulae are tabulated in Table 6.2 for the three-phase cable system described in Appendix A. T a b l e 6.2 Mutual Impedance between Two Cables with Burial Depth of 0.7om and Separation of 0.80m. Zij EX;^CT (fi/kin) AME*]rANI WEDI3POHL FREQ RES REA RES REA RES REA 1 10 100 1000 10000 100000 .000987 .009877 .098943 .994644 10.1015 105.154 .012562 .111152 .966670 8.21457 67.5095 525.238 .000987 .009862 .098452 .979517 9.65569 93.4318 .012563 . 111167 .967169 8.23027 68.0000 540.150 .000987 .009878 .098946 .994856 10.1193 106.592 .012562 .111151 .966668 8.21452 67.5070 525.000 - 92 - 50 30 / O / )0 UJ cu o li -tocn cn AMETANI tn H - 3 0 WEDEPOHL -50 10~ — i — n : TTTT. 2 10"' i—r^rrrrrrr- 1 ni 10' 1 i i IIIIII 11 1 10 2 10 MM; 1 10 3 4 \ i i mi, 10 1—i t mil, 10 S 1—t 11 nn' io 6 7 50 3 0 - 10- O UJ cu o c \ -10 AMETANI 0> CC WEDEPOHL - 3 0 5 0 -f 10 - 2 1 i i mii| 1 i i inn, 10" 1 1 i i mii| 1 i i iiiiii 10' 10 2 1 i ' mii; 10 S 1 i mm, 10 4 1 i i i nn; 10 ""1 10 1 , 1 S 1 6 ' "" 10 11 1 F r e q u e n c y [Hz] Figure 6.4 - Errors in Ametani's and Wedepohl's Approximations in the Mutual Impedance between Two Cables. 7 - 93 - The errors in Ametani's and Wedepohl's approximate formulae are plotted in Figure 6.4. The reasons for the errors are essentially the same as those discussed in Section 6.1. 6.3 Shunt A d m i t t a n c e M a t r i x The elements of the shunt admittance matrix obtained from Ametani's Cable Constant routine in the EMTP shows that the relative permittivity t is assumed to be real and constant. As explained earlier in Chapter 2, the relative permittivity is complex as well as frequency-dependent, but this data is usually difficult to obtain. A real, constant permittivity should give reasonable answers in many cases. - 94 - 7. CONCLUSION Various formulae proposed in the literature for the series impedance and shunt admittance matrices of underground cable systems have been compared in this thesis. The elements of the series impedance matrix are evaluated from formulae for the internal impedance of tubular conductors and from formulae for the earth return impedance. Exact equations for the internal impedance of tubular conductors were first derived by Schelkunoff [6j. They are given in terms of modified Bessel functions, and are therefore not suitable for hand calculations. Since then closed-form approximations suitable for hand calculations have been proposed by many authors, including Schelkunoff. A comparison of these approximate formulae shows that the formulae proposed by Wedepohl [22] give answers which are usually accurate enough for engineering purposes. With computers being almost universally available nowadays, approxi- mate formulae are no longer that important, however, and programming the exact formulae may therefore be the best approach. The displacement current term is usually neglected impedances of conductors. in the formulae for the internal It is shown that it can indeed be neglected for frequencies up to 10MHz. The shielding effect of grounded sheaths is explained as well, and it is shown that it is implicitly accounted for in the mutual impedances. The permittivity of the insulating material is needed for the elements of the shunt admittance matrix. Its value is frequency dependent as well as complex. In some cases, (e.g., cross- linked polyethylene), the permittivity can be assumed to be constant and real up to very high frequencies, while in other cases (e.g., oil-impregnated paper) the changes with frequency are quite significant. Two insulating materials, namely cross-linked polyethylene and oil- impregnated paper, are discussed in detail because they are the materials most often used in power cables. A general formula for the complex permittivity of insulation materials is given by Bartnikas [15], based on the relaxation time of the dielectric material. Ametani's Cable Constants routine in the E M T P [27] assumes that the permittivity is real and constant which may not always be accurate enough. The earth return impedance formula derived by Pollaczek [l] for the case of a semi- infinite earth is valid only for filamentary type conductors of negligible radius. This formula can be used for a conductor of finite radius a, if the condition | mo | <0.1 holds. This condition is satisfied up to a frequency of 1 M H z even for a worst case low earth resistivity of 10 fl— m. Hence Pollaczek's formula is recommended as the accurate formula. Values obtained from various approximate formulae and from Ametani's Cable Constants routine in the E M T P were compared against Pollaczek's formula. The results agree closely at low and medium frequencies but significant differences arise at high frequencies. - 95 - Equations for the internal impedances of a laminated tubular conductor have been derived from the equations for homogeneous tubular conductors. They are used to study the increase in the surface impedances of bus ducts in gas-insulated substations if the conductors are coated with high-resistivity magnetic material. This coating technique has been proposed by Harrington [32] for reducing the transient sheath voltage rise during switching operations, although others have criticized it as impractical, [discussion 32] - 96 - APPENDIX A Test Examples for Buried Cables Earth d, G x A '- » x Figure A.l - Three-Phaae Cable Setup for the Study Each cable is of a single core type with dimensions as given below conducting sheath central conductor insulation Figure A.2 - Basic Construction of each tingle core cable di,d ,dt = 0.75m, depth of burial of each cable 2 X12 = 0.30m, horizontal distance between cables 1 and 2 X ja = 0.30m, horizontal distance between cables 2 and 3 0.0234m, radius of the core 0.0385m, inner radius of the sheath *2 0.0413m, outer radius of the sheath 0.0484m, outside radius of the cable 100 fl—m, resistivity of the earth 1.7X10" J?-m resistivity of the core material 8 Peon 2.1xi0" /?-m resistivity of the sheath material 7 P- r core? r shea.tb» /'fexrtb'/'rilr = 1.0, relative permeability of the core, sheath, earth, and air respectively. - 98 - APPENDIX B Internal Impedances of a Tubular Conductor Based on tbc work of Schelkunoff [6], the derivation of the internal impdance formulae for tubular conductors is summarized here. B . l Circularly Symmetric Magnetic Fields In polar coordinates, Maxwell's equations assume the following form: dllz rd<*> 3// dz dHz r dz ifdlrHJ = (\lp + »"toe)E,, = (l/> + i<ai)E , t dE 3Ez 3E, dE, dz — — iliifiH, = —ioifill^ 1 (9[rE ) 3H,\ — 1— r ( dr t rd(j> 4 — ' = d<p ) ^ + i«e)E . r r { dr dE \ t TT f = ~ 3<t> ) (B-1) where H and E are electric and magnetic field strengths, respectively. Here we are interested in the circular magnetic field around conductors, with its lines of force forming a system of coaxial circles. Such circular magnetic fields are associated with currents flowing in isolated wires, as for example in a single vertical antenna, or between the conductors of a coaxial cable, as shown in Figure B.l. Figure B.l - The relative directions of the field components in a coaxial transmission line. - 99 From equation (B.l) we see that when the quantities are independent of the angle <f>, one of the independent subsets composed of the 1st and 3rd equation on the left of equation (B.l), together with the 2nd equation on the right, define the circular magnetic field strengths as follows: dlrllA . or a, - (1/p + i<at)rE, (B.2a) -(l/p (B.2b) dE dE dr dz 2 + itoe)E r iattH . (B.2c) 4 It has been shown by Schelkunoff that H^,E and E r z have components which vary exponentially along the longitudinal axis of the cable, i.e., along the z axis in Figure B.l. If we express the exponential variation of the quantities E,,E and Z as E,eC *, E e~ r Tz z and H^,e~ ', then the r quantities E,,E and H# are functions of r only. Substituting these values into equation (B.2) 2 we obtain E = r iuuHt ^—A. dr f • H (B.3a) dE, = —— + TE, dr (B.3b) = (Up + iu>f)rE, (B.3c) 4 where the quantity T is called the longitudinal propagation constant. Now solving for H# from Equation (B.3), we obtain where m = 2 I' I co /ze|. 2 This quantity m is called the intrinsic propagation constant of P the conductor material. For solid conductors, the term u> ut which accounts for the displace2 ment current is negligibly small compared to the conduction current. Hence we can neglect it up to quite high frequencies. The intrinsic propagation constants of metals are relatively large quantities even at low frequencies as shown in Table B.l for copper. - 100 - Table B.l Propagation Constant of Commerical Copper p = 1.7 X 10" tt-m 8 y/a){i/p = | m | iH,) 0 0.0 1 21.40 10 87.67 100 214.00 10,000 2140.00 1,000,000 21400.00 100,000,000 214000.00 On the other hand, the longitudinal propagation constant T is relatively very small, even at high frequencies. For example, if air is the dielectric between the conductors T will be of the order of (l/3)ia>10~ . Hence, even at high frequencies T is negligibly small by comparison with 10 m. 2 2 Therefore, we can write equation (B.4) as d H* i dH, dr 2 dr' The solution for H sT = X* m H (B.5) of equation (B.5) is in the form of Bessel functions given by: t Al {mr) x + BK^mr) (B.6) Since we are interested in longitudinal voltage drops, we must find the longitudinal electric field stength first. This can be obtained from equation (B.3) and (B.6) along with the following rules of differentiation for modified Bessel functions of any order n, (B.7a) dx _d_(x*K )= dx n -x*K . n x The solution for the longitudinal electric field strength then becomes (B.7b) - 101 - E, = pm\AI (mr) - BK {mr)\ 0 (B.8) 0 In a tubular conductor whose inner and outer radii are a and b, respectively, coaxial return path for the current may be either outside or inside the tube or partly inside and partly outside. We designate Z return and Z b t as the internal impendance of the tubular conductor with internal as the internal impedance with external return. If the return path is partly internal and partly external, we have in effect a two-phase transmission line with a distributed transfer impeduace Z between the two loops of internal and external return. ab In order to determine these impedances, let us assume that a total current (/, + /.) is flowing in the tubular conductor, with part /„ returning inside and part I returning outside. b Figure B.2 - Loop Currents in a Tubular Conductor Since the total current enclosed by the inner surface of the conductor is — I and that enclosed a by the outer surface is I (7 + I — /„), the magnetic field strengths at these two surfaces take b 0 b the values ( — I /2na) and (I /2nb) respectively. Hence from equation (B.6) we have a b A/,(7»io)+ M , ( m a ) = -lJ2na (B.9a) A li(mb) + BK (mb) = I /2r.b (B.9b) x b From these two equations the values of A and B can be evaluated as 2-naD _ /»/»("») = 2naD where 2nbD 2xbD ( B 1 0 b ) - 102 - D = r (mb)k (ma) - I^majk^mb) 1 (B.II) 1 Substituting these values in (B.7) and using the identity /„ (z)/C"i(z) + K (x)I (x) = 1/x,, 0 x we obtain the longitudinal electric field strength at any point on the conductor. However, we are interested in its' values at the surfaces as they constitute the surfaces of propagation. Hence, equating r successively to a and 6 we obtain E,(a) = ZJ a + Z I ab E,(b) = Z I cb (B.12a) b + ZI a b (B.12b) b where 2naD > ~ 7 Z 7 = [irimajK^mb) + /v" (mo)/,(m6) 0 P ^/ (m6)K'i(r7ja) + AT (m6)/ (ma) j 2xbD m 0 0 1 P 2nabD (B.13) Schelkunoff stated these results in the following two theorems. Theorem 1 If the return path is wholly external (I = 0) or wholly internal (/;, = 0), the longitudinal a electric field strength on that surface of a tubular conductor which is nearest to the return path equals to the corresponding surface impedance per unit length multiplied by the total current flowing in the conductor and the field strength on the other surface equals to the transfer impedance per unit length multiplied by the total current. Theorem 2 If the return path is partly external and partly internal, the separate components of the field strength due to the two parts of the total current are calculated by the above theorem and added to obtain the total field strength. - 103 - APPENDIX C Calculation of Earth Return Impedances in an Infinite Homogeneous E a r t h If the return current distribution in the ground is circularly symmetrical, then we refer to such a case as infinite earth. This happens in practice when the cables are either buried at large depth or when the frequency is very high. In both cases, the penetration depth d given by 503 •It) m, becomes smaller than the depth of burial. Then only the earth medium must be considered, which simplifies the solution. If the cables are buried close to the earth's surface on the other hand, which is usually the case, then the distribution of current in the ground is no longer symmetrical (at least at low frequencies), and the magnetic field both in air and earth must then be considered which makes the solution more complicated. Consider a cable lying along the Z-axis of the cartesian coordinate reference frames as shown in Figure C . l . Let the positive direction be along the Z-axis, and let the conductor carry a current I flowing in the positive direction returning through the ground. Let the radius over the outer insualtion be a. From Ampere's Law (neglecting the displacement current term) the magnetic field strength H at a radius r & o is given by 2-nrH = I + J 2nrJdr i.e. T (C.l) H = —+-fjrdr 2nr r J v where J is the current density in the ground. Suppose that the earth is subdivided into concentric cylindrical shells of radius r and thickness dr in which the current density / and magnetic field strength / / a r e constant. Then the magnetic flux per unit length of such a shell is given by d<f> = BdA = ulldr (C.2) Substituting for H from equation (C.l) yields (C.3) -!-+±Jjrdr dtp = udr 2xr r •* Now let us write Kirchhoff's voltage law around the rectangle ABCD of unit length and width dr. The net resistive voltage drop is ' |~^ ~J^ , — : r a n c * t n e induced voltage is jmd<l> or jtaiilldr. - 104 - Figure C.1 - Representation of a Buried Conductor in an In finite Earth. Since the sum of these two voltages must be zero, we obtain dJ —p dr + jtauH dr = 0 (C.4) dr Substituting f o r / / f r o m equation (C.1), we have pdJ dr •dr + jtAfidr Multiplying this equation by — + -fjrdr = 0 2itr ~- and differentiating with respect to r we have par dJ 2 dr 2 If we substitute m* for Jfa>/i P (C.5) + dJ_ _ jvuJ rdr p = 0 , then equation (C.4) can be written as (C.6) - £ 105 - «•»««*. (C7) and equation (C:6) can be written as d J . dJ 2 dr' mJ = 0 l r dr (C.8) Equation (C.8) is immedately recognizied as a Bessel equation whose solution is of the form / - AI {mr) + BK {mr) 0 (C.9) 0 We note that I (x) approaches infinity as z approaches infinite. However, we cannot permit a 0 solution of J to increase indefinitely as r approaches infinity and we must conclude that , 4 = 0 PIHence, equation (C.9) becomes / = BK (mr) (C.10) 0 Using equation (C.7) we find a solution for the magnetic field strength H as m H = -BKAmr)m (Cll) 2 Now applying the boundary condition that H = I/2ira in the ground immediately adjacent to the cable, we obtain the value for the constant B from equation ( C l l ) as B = ~ 2nal<!(ma) ( C ' 1 2 ) Using the equation E = pJ, the solution for the electric field strength at any point in the soil is found to be pml K {mr) ~~2*a K^ma) . 0 E ( C 1 3 ) The earth return self impedance as well as the mutual impedance between two buried cables can be deduced from this equation (see Chapter 4). - 106 - APPENDIX D Calculation of E a r t h Return Impedances in a Semi-Infinite Homogeneous E a r t h The limited conductivity of the ground path for the return currents as well as conductor skin effects result in the frequency dependence of the line parameters. The parameters of a transmission line over a ground of perfect conductivity are given by textbook formulae, but the earth return effects and skin effects need special treatment. While a complete solution of the actual problem is impossible, on account of the uneven surface under the line and the lack of conductive homogeneity in the earth, a solution of the problem, where the actual earth is replaced by a plane homogeneous semi-infinite solid, gives reasonably accurate answers. The same applied to the underground case, too. The first step in finding the earth return impedances is to derive the respective longitudinal electric field strengths. Let us first consider au overhead line and derive the electric field strengths in air and in earth. A y p . (cc.h) X Earth Q* (x.y) Figure D.l - Current-Carrying Filament in the Air Let medium 1, denoted by subscript 1, correspond to air and medium 2, denoted by subscript 2, correspond to earth. Let point P{a,h) correspond to the current-carrying filament lying - 107 - along the Z-axis of the Cartesian coordinate system. Let E+ + (x,y;a,h) be the electric field strength in air at a point Q(x,y) and E- + (z,y';a,h) be the electric field strength in the earth at a point Q'(x,y'). Note that the y axis is positive in the air and that the y' axis is positive in the earth, as shown in Figure D.l. From Maxwell's theory, the general equation for electromeganetic wave propagation is given by V E - V(V£) = - tovjtf 2 (D.l) where p,fi and e correspond to the respective medium to which this equation is applied. assumption 4 (Chapter 2), we can say that VE Using — 0 in both air and earth. Hence, equation (D.l) can be written as V E - 2 CO Now let us define the fields which we would like to derive as follows E++ = E + z E-+ — E- + z + (D.2) [It = Electric field strength in the air due to the current-carrying filament in the air = Electric field strength in the earth due to the current-carrying filament in the. air If we assume that a sinusoidal current I of angular frequency' o> is passing through a filament concentrated at the point (a,h) in the x—y plane as illustrated in Figure D.l, then the current density is zero everywhere in the air except at the point (a,h) where it is infinite. Such an idealized situation can be represented by the Dirac delta function 5(x-^a) defined as in such a way that / 5(x-a)dx = 1 — ! (D.4) CB which implies that if f(x) is continuous at x = 0 and bounded elsewhere [12] J f(x)5(x-a)dx = f(a) Hence in the air, the current density can be expressed in the form (D.5) - 108 - I6(x-a)S(y-h) Now (D.6) keeping this result in mind and noting that we are interested only in the electric field strength along the Z-axis and 9E ' Z - = 0 (using assumption 2 from Chapter 2) equation (D.2) can be written for the case of air as d E+ + d E.+ 2 2 2 mE+ + p mfIS(x — a)6(y — h) 2 3z By' a + l (D.7) where Pi For the earth, equation (D.2) can be written as a £_ a E_ 2 2 + + m|E_. = dx 2 (D.8) • here m 2 The solutions for E++ and E_ = — I P2 CO 2 p2i 2 should be obtained in such a way that they satisfy the follow- + ing boundary conditions: 1. Continuity of E at the surface. Lim £++ E_+ = Lim V-*0 y— — Lim 0 £_ + = /? (say) 0 jr'-+o Vertical component of B is continuous at the surface dx (D.10) dx Horizontal component of H is continuous at the surface dE++ Pidy The solutions for E+ + and E_ + a£_ + P2&y aE_ + p-z^y' (D.il) can be found by using integral transform techniques. Taking the Fourier complex transformation of equations (D.7), (D.8), (D.9) and (D.ll) with respect to x, with 6 as the parameter, we obtain the following equations: - d E 109 - 2 -6 E^ + 2 + + <f £.. = mfE + /j,m /exp(-^a)%-A) (D.12) 2 + + 2 -6 E_ + + 2 £+ + | y-0 1 dE+* Hi Now E-+ — - mf E.+ ^ j | 'Y •j 0 — (D.13) (D:14) £Q i1 OD d £ __ + + dy - I fi dy' 2 taking the Fourier sine transformation of equation (D.12) with respect to y with 0 as the parameter, we have: -6 E^+ 2 — 0 E+ + + 0E = m E 2 2 O + + + p^rn? Iexp{-j6a)sin{0h) (D.16) i.e., (0 + C f ) E 2 + + where Cf = 0 + = 0E m 2 - m lexp(-jea)sin(0h) (D.17) 2 O Pl 2 Similarly, taking the Fourier sine transformation of equation (D.13) with respect to y' with 0' as parameter, we have -0 E_ 2 + + fl' E. + /J*E = m £ _ 2 (D.18) 2 + 0 + Hence + Cf)£_ where C'| = t? + m = £'E (D.19) 0 . : 2 + 2 Now, taking the inverse Fourier sine transformation of equation (D.17) with respect, to 0 we have = E exp( —C,y) 0 -•^ /exp(-^a)|exp(-C |/i-j/|)-exp(-C,|/ +y|)j L 1 l (D.20) Similarly, taking the inverse Fourier sine transformation of eqution (D.19) with respect tofl'we have F_ + = E exp(-C y) 0 By taking the derivative of Z? * with respect to y and the derivative of E_ + (D.20) 2 + with respect to y. - no - and substituting in equation (D.l 5), we obtain the value of E as follows: 0 P\fn f/exp( — j6a)exp{—C h) — x u i (D.22) (C, + :—<7 ) 2 a u Substituting E 0 in equation (D.20) and taking the inverse complex transformation with respect to 0, we obtain the value of E .2/ + as follows: + [expf-C,| h-y - | )-exp(-C | h +y |)] 2 2C, exp{-C \h+y\) 1 C. + — Similarly, substituting E 0 C exp(j6\ i - a | )d0 (D.23) 2 in equation (D.21) and taking the inverse transformation with respect to 6, we obtain the value of E. + as follows Pi fl m E-+ = ~ 2n " exp{— C^y — C h}exp(j6\ x — a\)d6 J r 1 Ci + —C (D.25) 2 U2 Now that we have derived the equations for the electric field strengths in the air and in the earth due to an overhead conductor, we will turn our attetion to the case of an underground conductor. Electric Field Strength in the A i r and in the Earth due to Current Carrying Filament Buried in the Earth. Let = E-- ~ Z Electric field strength in the earth due to the current-carrying filament buried in the earth. E+. = is _/= + Electric field strength in the air due to the current-carrying filament buried in the earth. Similar to equations (D.7) and (D.8), Maxwell's equations for electromagnetic wave propagation in air and earth respectively, for this case are given by, d £ _ 2 + dx 5 dE 2 + 2 By = m,E,. y^O, (D.26) - H Ill - y ? (x.y) x _i p.- Earth • (x,y') . W) Figure D.2 - Current Carrying Filament Buried in the Earth. 3 E__ ax 1 , , . 8 E__ 2 2 + ay — = m | £ _ _ + p m fS(x 2 2 (D.27) — a)5(y — h ) Similar to the procedure used in the derivation of fields and we solve for and such that they satisfy the boundary conditions given by equation (D.9) through (D.l 1). Hence we have E. = pmI 2 + 2TT pmI 2 £•_ _ .= - 2 2 " e x p { — C y — C h) l 2 -»- — C ?2 •xp{j6\x-d\ (D.28) )d0 2 [ e x p ( - C | h'-y'\ ) - e x p ( - C | h'+y'\ ] 2 2 2C, 2n exp(-C |/i'-rV|) 2 (D.29) exp(jO \ x-a'\ )d8 C, + —C 2 »2 If we assume that the relative permeability of air and earth are the same, i.e., u Tl = u, then 2 - i r l 112 - = fi and we can show that the equations derived for E+ ,E_ 2 + + ,E^_ and E+_ are the same as those derived by Pollaczek [lj. Now using the standard results i \ K (mr)= c exp{—aV^-+m }exp . . . , J \ (jis)d 2 y x 0 where r = and where K 9 y 2Vr+m • ' _ • . (D.30) 3 2 is the modified Bessel function of the 2nd kind and of the 0 zeroth order, we can write E+ + and E _ _ as follows: pm,/ ( 2 E „ = + / ; = (jtf|x-o»| \A + rnf + \ / V + )rfg 2 where /?, ~ V {T - af+(h - y?, Z x = V(z-a) 2 + (h+yf pm I ( 2 — —jA'o( 2^2) m K (m Z ) — G 2 2 exp{-|j/+/i V^+m22}exP( + / . (j*l z - a |)«« r — \A + m? + \A + ™ 2 where /? 2 .=• V ( z - o ] ' + (A'7P, Z = 2 2 = V(z-a') + (A'+^j 2 (D.31(a,b)) 2 2 Further using assumption 4 in Chapter 2, we can neglect the displacment current up to a fairly high frequency, and also noting the fact that the resistivity of air, i.e., p, is very large, we can conclude that the term mf ~ 0 . This produces the final equations: E+ + = 7(011aI ( —UniZM + / ^ - ^ ^ " ^ ^ e x n l ^ l z - a l ^ — \ e\ + y/e E_+ = - joidol 2n J- exp{-A I 0\ -y'Ve |*| + 2 w \/e +m 2 2 + 2 + m 2 rn }exp(;6l| z - a ] )d8 2 - 1 ( 2TT-\l<0(m2R2)-K{ 113 - ,{m Z ) 2 2 e x p l - b ' + A ' l V<? +m|} exp(j'tl| i - a ' | </0 2 \e\ + V^+^I jco/ip/ ; exp{-y | E _ + = / * — 2 6\-h\/e |*|. + 2 + m } 2 (D.32(a.b,o,d)) 2 exp^|x-a|)d* V^+m| - 114 - REFERENCES V. F. PoIIaczek, Ubrr das Fcld ciner un endlich langon Wechsel = Stromdurch flossenen Einfachleitung, E.N.T., Band 3, Heft 9, 1926, pp. 339-360. V. F. PoIIaczek, Uber die Induktionswirkungen ciner Wech selstromeinfachleitung, E.N.T., Band 4. Heft 1, 1927, pp. 18-30. J. R. Carson, The Rigorous and Approximate Theories of Electrical Transmission along Wires, Bell System Tech., Journal, January 1928, pp.11. J. R. Carson, Wave Propagation in Overhead Wires with Ground Return, Bell System Technical Journal, Vol. 5, 1926, pp. 539-554. Carson, J. R., Ground Return Impedance - Underground Dire with Earth Return. Bell Sys- tem Technical Journal, Vol. 8, 1929, pp. 94-98. [6 S. A. Schelkunoff, 77ie Electromagnetic Theory of Coaxial Transmission Line and Cylindrical Shells Bell System Technical Journal, Vol. 13, 1934, pp. 522-579. [7 E. W. Kimbark, Electrical Transmission of Power and Signal, John Wiley Publication, 1949. [8 J. R. Carson and J. J. Gilbert, Transmission Characteristics of the Submarine Cable, Bell System Technical Journal, July 1922, pp. 88-115. [9 D. J. Wilcox, Transient Analysis of Underground Cable System, Ph.D. Thesis, Manchester Insititute of Science and Technology, 1974. [10] N. Mullineux and J. Rl Reed, Images of Line charges and Currents, Proc. IEE, Vol. Ill, No. 7, July 1964, pp. 1343-1346. [ll] N. Mullineux. J. R. Reed and S. A. Soppitt, Present Mathematics Sylabuses and the Analytical Solution of Field Problems - I and II, Electrical Engineering Education, Vol. 8, 1970, pp. 221-233, and pp. 289-298. [12] J. P. Bickford, N. Mullineux and J. R. Reed, Computation of Power System Tratisients, IEE Monograph Series 18, 1980. [13] E. A. Clarke, Circuit Analysis of A.C. Power Systems, G. E. Advanced Engineering Pro- gram Series, 1949. [14] E. D. Sunde, 1968. Earth Conduction Effects in Transmission Systems, Dover Publication, - 115 - [15] R. Bartnikas and R. M. Eichchorn, Engineering Dielectrics - Vol. II A, A S T M publication, S T P 783, 1983. [16] G . S. Stone and S. A. Boggs, Propagation of Potential Discharge Pulses in Shielded Power Cable, IEE Conference on Electrical Insulation and Dielectric Phenomena, October 1982, pp. 275-280. [17] O. Breien and I Johansen, Attenuation of Travelling Waves in Single-Phase High Voltage, Cables, Proc. IEE, Vol. 118, No. 6, June 1971, pp. 787. [18] Hylten - Cavallius and S. Annestrand, Distortion of Travelling Waves in Power Cables and on Power Lines, CIGRE, paper 325, 1962. [19] H. W. Dommel, Line Constants of Overhead Lines and Underground Cables, E E 553 Notes, U.B.C. [20] H. W. Dommel and J. H. Sawada, The Calculation of Induced Voltages and Currents on Pipelines adjacent to A. C. Power Lines, Report submitted to B.C. Hydro, August 1976. [21] H. Bocker and D. Olding, Induced Voltages in Pipeline close to High-Voltage Lines, (in German) Elektrizitatsurintschaft, Vol. 65, 1966, pp. 157-170. [22] L. M. Wedepohl and D. J. Wilcox, Transient Analysis of Underground Power Transmission Systems, Proc. IEE, Vol. 120, No. 2, Feb. 1973, pp. 253-260. [23] L. Bianchi and G. Luoni, Induced Currents and Losses in Single-Core Submasive Cables, I E E E Trans., PAS-95, Jan. 1976, pp. 49-57. [24] H. W. Dommel, Overhead Line Parameters from Handbook Formulas and Computer Programs, IEEE Trans., PAS-104, Feb. 1985, pp. 366-372. [25] Mathematical (or Special) Functions, Users Manual Computing Centre, U.B.C, April 1983. [26] Integration (quadrature) Routines, Users Manual, Computing Centre, U.B.C, January 1984. [27] A. Ametani, Performance of a Theoretical Investigation and Experimental Studies Related to the Analysis of Electro-Magnetic Transients in Multiphase Power System Cables, Report No. 1, submitted to B.P.A., Oregon, July 1976. [28] D. Shirmohammadi and A. S. Morched, Improved Evaluation of Carson Correction Terms for Line Impedance Calculations, Trans. Canadian Electrical Association (Engineering & Operating Division), Vol. 24, 1985.. [29] Luis Marti, Improved Evaluation of Carson's Correction Terms for UBC's version of the Line Constants Program, Department of Electrical Engineering, U.B.C, December 1984. - 116 [30] I. I. Dommel, Program Documentation for Subroutine TUBE, May 1978. [31] M . Nakergawa, A. Ametani and K . Iwamoto, Further Studies on Wave Propagation in Overhead Lines with Earth Return: Impendance of Stratified Earth, Proc. IEEE, Vol. 120, No. 12, Dec. 1973, pp. 1521-1528. [32] R. J. Harrington and M. M. El-Faham, Proposed Methods to Reduce Transient Sheath Voltage Rise in Gas Insulated Substations, IEEE Trans. PAS-lOi, May 1985, pp. 1199-1206. [33] R. A. Chipman, Theory and Problems of Transmission Lines, Shaum's Outline Series, Shaum Publishing Company, New York. [34] M . Abramowitz, I. A. Stegun, Handbook of Mathematical Functions, Dover Publications, New York.
- Library Home /
- Search Collections /
- Open Collections /
- Browse Collections /
- UBC Theses and Dissertations /
- Series impedance and shunt admittance matrices of an...
Open Collections
UBC Theses and Dissertations
Featured Collection
UBC Theses and Dissertations
Series impedance and shunt admittance matrices of an underground cable system Navaratnam, Srivallipuranandan 1986
pdf
Page Metadata
Item Metadata
Title | Series impedance and shunt admittance matrices of an underground cable system |
Creator |
Navaratnam, Srivallipuranandan |
Publisher | University of British Columbia |
Date Issued | 1986 |
Description | This thesis describes numerical methods for the evaluation of the series impedance matrix and shunt admittance matrix of underground cable systems. In the series impedance matrix, the terms most difficult to compute are the internal impedances of tubular conductors and the earth return impedance. The various formulae for the internal impedance of tubular conductors and for the earth return impedance are, therefore, investigated in detail. Also, a more accurate way of evaluating the elements of the admittance matrix with frequency dependence of the complex permittivity is proposed. Various formulae have been developed for the earth return impedance of buried cables. Using the Pollaczek's formulae as the standard for comparison, the formula of Ametani and approximations proposed by other authors are studied. Mutual impedance between an underground cable and an overhead conductor is studied as well. The internal impedance of a laminated tubular conductor is different from that of a homogeneous tubular conductor. Equations have been derived to evaluate the internal impedances of such laminated tubular conductors. |
Subject |
Underground electric lines |
Genre |
Thesis/Dissertation |
Type |
Text |
Language | eng |
Date Available | 2010-07-21 |
Provider | Vancouver : University of British Columbia Library |
Rights | For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use. |
DOI | 10.14288/1.0097157 |
URI | http://hdl.handle.net/2429/26726 |
Degree |
Master of Applied Science - MASc |
Program |
Electrical and Computer Engineering |
Affiliation |
Applied Science, Faculty of Electrical and Computer Engineering, Department of |
Degree Grantor | University of British Columbia |
Campus |
UBCV |
Scholarly Level | Graduate |
Aggregated Source Repository | DSpace |
Download
- Media
- 831-UBC_1987_A7 N38.pdf [ 6.34MB ]
- Metadata
- JSON: 831-1.0097157.json
- JSON-LD: 831-1.0097157-ld.json
- RDF/XML (Pretty): 831-1.0097157-rdf.xml
- RDF/JSON: 831-1.0097157-rdf.json
- Turtle: 831-1.0097157-turtle.txt
- N-Triples: 831-1.0097157-rdf-ntriples.txt
- Original Record: 831-1.0097157-source.json
- Full Text
- 831-1.0097157-fulltext.txt
- Citation
- 831-1.0097157.ris
Full Text
Cite
Citation Scheme:
Usage Statistics
Share
Embed
Customize your widget with the following options, then copy and paste the code below into the HTML
of your page to embed this item in your website.
<div id="ubcOpenCollectionsWidgetDisplay">
<script id="ubcOpenCollectionsWidget"
src="{[{embed.src}]}"
data-item="{[{embed.item}]}"
data-collection="{[{embed.collection}]}"
data-metadata="{[{embed.showMetadata}]}"
data-width="{[{embed.width}]}"
async >
</script>
</div>
Our image viewer uses the IIIF 2.0 standard.
To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0097157/manifest