MODELING GUIDELINES FOR VERY FAST TRANSIENTS IN GAS INSULATED SUBSTATIONS

Report Prepared by the Very Fast Transients Task Force of the IEEE Working Group on Modeling and Analysis of System Transients J.A. Martinez (Chairman), P. Chowdhuri, R. Iravani, A. Keri, D. Povh occurred at system voltages above 420 kV, they have been correlated with disconnect switch and circuit breaker operation. The generation and propagation of VFT from their original location throughout a GIS can produce internal and external overvoltages. The main concern are internal overvoltages between the center conductor and the enclosure. However, external VFT can be dangerous for secondary and adjacent equipment. These external transients include transient voltages between the enclosure and ground at GIS-air interfaces, voltages across insulating spacers in the vicinity of GIS current transformers, when they do not have a metallic screen on the outside surface, voltages on the secondary terminals of GIS instrument transformers, radiated electromagnetic fields (EMF) which can be dangerous to adjacent control or relay equipment. VFT can also occur during switching of vacuum breakers and with certain lightning conditions. The objective of this document is to present an explanation of the VFT phenomena that can occur in GIS and provide guidelines for representing GIS components in digital simulations. Some examples with detailed input data are presented. A discussion about the accuracy of the simulations and their verification with field measurements is also included.

Abstract- This document is aimed at providing modeling guidelines for digital simulation of very fast transients (VFT) in gas insulated substations (GIS). A short explanation about the origin of VFT overvoltages, their propagation and effects on GIS equipment is included. The document presents modeling guidelines of GIS components proposed in some previous works. The accuracy of digital computations using these guidelines is also discussed. Several examples corresponding to actual cases with detailed data input and validated simulation results are presented.

Keywords : Gas Insulated Substations, Very Fast Transients, Modeling, Digital Simulation.

1. INTRODUCTION An accurate representation of each component of a system is essential for a reliable simulation of its transient performance. This representation must be done taking into account the frequency range of the transients to be simulated. Very Fast Transients (VFT) belong to the highest frequency range of transients in power systems. Component models to be used in VFT simulations must be suitable for frequencies varying from 100 kHz up to 50 MHz [21]. VFT arise within a gas-insulated substation (GIS) any time there is an instantaneous change in voltage. Most often this change occurs as the result of the opening or closing of a disconnect switch, but other events, such as the operation of a circuit breaker, the closing of a grounding switch, or the occurrence of a fault, can also cause VFT. These transients generally have a very short rise time, in the range of 4 to 100 ns, and are normally followed by oscillations having frequencies in the range of 1 to 50 MHz. Their magnitude is in the range of 1.5 to 2.0 per unit of the line-to-neutral voltage crest, but they can also reach values as high as 2.5 per unit. These values are generally below the BIL of the GIS and connected equipment of lower voltage classes. VFT in GIS are of greater concern at the highest voltages, for which the ratio of the BIL to the system voltage is lower. Some equipment failures and arcing problems between grounded parts have

2. BACKGROUND 2.1 ORIGIN OF VFT IN GIS VFT overvoltages are generated in a GIS during disconnector or breaker operations, or by line-to-ground faults. During a disconnector operation a number of pre- or restrikes occur due to the relatively slow speed of the moving contact. Figure 1 shows the simplified configuration used to explain the general switching behavior and the pattern of voltages on closing and opening of a disconnector at a capacitive load [4], [17]. During closing, as the contacts approach, the electric field between them will rise until sparking occurs. The first strike will almost inevitably occur at the crest of the power frequency voltage, due to the slow operating speed. Thereafter current

6-1

will flow through the spark and charge the capacitive load to the source voltage. As it does so, the potential difference across the contacts falls and the spark will eventually extinguish. The behavior on opening is very nearly a complete reversal of the above description. In case of a line-to-ground fault, the voltage collapse at the fault location occurs in a similar way as in the disconnector gap during striking. Step-shaped travelling surges are generated and injected to GIS lines connected to the collapse location. The rise time of these surges depend on the voltage preceding the collapse.

outcoupling at the transition to the overhead line. In case of power transformers feeding the GIS, overall transients with frequencies in the range of 20 to 100 kHz can be observed caused by the oscillation of the whole system consisting of the GIS and the transformer. Due to the travelling wave behavior of the VFT, the overvoltages caused by disconnector switches show a spatial distribution. Normally the highest overvoltage stress is reached at the open end of the load side. The maximum value of the local VFT overvoltages is dependent on the voltage drop at the disconnector just before striking and on the location considered. For the calculation of VFT stresses, the trapped charge remaining on the load side of the disconnector must be taken into consideration. For a normal disconnector with a slow speed, the maximum trapped charge reaches 0.5 pu resulting in a most unfavorable voltage collapse of 1.5 pu. For these cases, the resulting overvoltages are in the range of 1.7 pu and reach 2.0 pu for very specific cases. For a high speed disconnector, the maximum trapped charge could be 1.0 pu and the highest overvoltages reach values up to 2.5 pu. Extremely high values of more than 3.0 pu have been reported. It can be shown, however, that these values have been derived by calculation using unrealistic simplified simulation models. An example of these transient phenomena measured in an actual GIS is given in Figure 2, where one prestrike of a disconnector switching is depicted showing the steep voltage transients at the supply and load sides. The basic frequency component of the VFT in the MHz range, the overall transient and the steady state waveform are also shown. The two following examples will be useful to illustrate the generation of VFT in GIS, and the influence of some parameters on the frequency and maximum values of these transients. Figure 3 shows a very simple case, a GIS bus duct represented as a lossless distributed parameter transmission line is fed from a step-shaped source. The reflections of the travelling wave at both terminals of the duct will produce at the open terminal a pulse-shaped transient of constant magnitude 2 pu - and constant frequency. The frequency of this pulse can be calculated from the following expression

2.2 PROPAGATION OF VFT IN GIS VFT in GIS can be divided into internal and external. Internal transients can produce overvoltages between inner conductors and the encapsulation, external transients can cause stress on secondary and adjacent equipment. A summary about the propagation and main characteristics of both types of phenomena follows.

2.2.1 Internal transients Breakdown phenomena across the contacts of a disconnector during a switch operation or line-to-ground faults generate very short rise time travelling waves which propagate in either direction from the breakdown or fault location. Surges travelling throughout GIS and to other connected equipment are reflected and refracted at every transition point. As a consequence of multiple reflections and refractions, travelling voltages can increase above the original values and very high frequency oscillations occur. The propagation of VFT throughout GIS can be analyzed by representing GIS sections as low-loss distributed parameter transmission lines. Each section may be characterized by a surge impedance and a transit time. Travelling waves are reflected and refracted at every point where they encounter a change in the surge impedance. The generated transients depend on the GIS configuration and on the superposition of the surges reflected and refracted on line discontinuities like breakers, "T" junctions or bushings. The main frequencies depend on the length of the GIS sections affected by the disconnector operation and are in the range of 1 to 50 MHz. The internal damping of the VFT influencing the highest frequency components is determined by the spark resistance. Skin effects due to the aluminum enclosure can be neglected. The main portion of the damping of the VFT occurs by

1 4 -

(1)

being - the transit time of the line. If the propagation velocity

6-2

is close to that of light, the frequency, in MHz, of the voltage generated at the open terminal will be

f M

75 d

evaluated as it will have an important influence on the maximum generated overvoltage; the second plot of Figure 4.c was obtained assuming that the trapped charge was 0.5 pu.

(2) 2.2.2 External transients

where d is the duct length, in meters. In this case, d = 12 m, then the frequency is 6.25 MHz, which corresponds to a period of 160 ns, as shown in Figure 3.b. Therefore, the main reason why VFT are generated in GIS is due to the short length of ducts. Each GIS section, represented as a lossless distributed parameter transmission line, may be characterized by a surge impedance and a transit time. In the previous case, the surge impedance of the bus was 50 6; however, this value was unimportant because the supplying source was assumed ideal. Figure 3.b shows the simulation result obtained with a more realistic representation of the source, R = 10 6. One can observe than now the frequency is still the same, but the maximum overvoltage at the open terminal is lower than 2 pu, and the transient is damped. At the end, the voltage value at this terminal is equal to that of the source. The equivalent impedance at the source side of an actual GIS disconnector will be different from a pure resistance; this representation has be considered here to facilitate the introduction to VFT generation. As mentioned above, trapped charge remaining on the load side of a disconnector must be taken into consideration. Figure 3.c shows that the voltage at the open terminal can reach 3 pu if the transient is started with a 1 pu trapped charge on the transmission line. This increase is due to the fact that now the travelling wave is a 2 pu step, which duplicates at the open terminal; the final 3 pu value is the result of subtracting the remaining trapped charge. A more accurate simulation should assume that the equivalent impedance at the source side is not negligible and a lower trapped charge. The second plot of Figure 3.c depicts the new results with a non-zero source resistance and a lower trapped charge. Maximum voltages can reach higher values in more complex GIS configurations. Figure 3 shows a "T" junction GIS network. The simulations performed with the previous case were repeated with this new configuration. All bus ducts had the same surge impedance - 50 6 - and the same propagation velocity, that of light. The simulations show that node 4 in Figure 4.a is the open terminal where overvoltages are higher. From the new plots one can deduce that VFT as higher as 4.5 pu can be generated; however, as with the previous case, a realistic simulation cannot neglect the source impedance. In addition, the value of the trapped charge has to be accurately

An internally generated VFT propagates throughout the GIS and reaches the bushing where it causes a transient enclosure voltage and a travelling wave that propagates along the overhead transmission line. An explanation about the generation of external transients and some comments on their main characteristics follow. a) Transient enclosure voltages Transient enclosure voltages (TEV), also known as transient ground potential rise (TGPR), are short duration high voltage transients which appear on the enclosure of the GIS through the coupling of internal transients to enclosure at enclosure discontinuities. The usual location for these voltages is the transition GIS-overhead line at an air bushing, although they can also emerge at other points such as visual inspection ports, insulated spacers for CTs or insulated flanges at GIS/cables interfaces. The simplified circuit shown in Figure 5 will be used to explain the generation of TEV [20]. At the GIS-air interface three transmission lines can be distinguished : the coaxial GIS transmission line, the transmission line formed by the bushing conductor and the overhead line, the GIS enclosure-to-ground transmission line. Each of them has a different surge impedance. When an internal wave propagates to the gas-to-air bushing, a portion of the transient is coupled onto the overhead line-to-ground transmission line, and a portion is coupled onto the GIS enclosure-to-ground transmission line. The latter constitutes the TEV. In general, TEV waveforms have at least two components, the first one has a short initial rise time and is followed by high frequency oscillations determined by the lengths of various sections of the GIS, they are concentrated in the range of 5 to 10 MHz. The second component is of lower frequency, hundreds of kHz, and is often associated with the discharge of capacitive devices with the earthing system.

6-3

a) Diagram of the capacitive circuit

b) Opening operation

c) Closing operation Figure 1. Variation of load and source side voltages during disconnector switching [4].

6-4

a) steep voltage transients b) basic frequency component of the VFT in the MHZ range c) overall transients in the kHZ range d) low frequency transient and steady state condition Figure 2. Transients on the source and load side of a GIS due to disconnector switching.

6-5

a) Scheme of the network

v [mV]

v [mV]

1800

1800

1600

1600

1400

1400

1200

1200

1000

1000

800

800

600

600

400

400

200

200

0 0 100 200 300 400

0 0 100 200 300 400

( 1) NUDO1

( 2) NUDO2

t [ns]

( 1) NUDO1

( 2) NUDO2

t [ns]

R=0 b) Simulation results without trapped charge

v [mV]
3000

R = 10 6

v [mV]
3000

2500

2500

2000

2000

1500

1500

1000

1000

500

500

-500

-500

-1000

-1000

-1500 0 100 200 300 400

-1500 0 100 200 300 400

( 1) NUDO1

( 2) NUDO2

t [ns]

( 1) NUDO1

( 2) NUDO2

t [ns]

R=0 c) Simulation results with trapped charge Figure 3. Generation of VFTs in a GIS bus duct.

R = 10 6

6-6

a) Scheme of the network

[mv]
3000

[mv]
3000

2500

2500

2000

2000

1500

1500

1000

1000

500

500

-500 0 200 400 600 800

-500 0 200 400 600 800

( 1) NUDO1

( 4) NUDO4

t [ns]
( 1) NUDO1 ( 4) NUDO4

t [ns]

R=0 b) Simulation results without trapped charge

[v]
5

R = 10 6

v [V]
5

-1

-1

-2 0 200 400 600 800

-2 0 200 400 600 800

( 1) NUDO1

( 4) NUDO4

t [ns]

( 1) NUDO1

( 4) NUDO4

t [ns]

R=0 c) Simulation results with trapped charge

R = 10 6

Figure 4. Generation of VFTs in a GIS with a bifurcation point.

6-7

Both components are damped quickly as a result of the lossy nature of the enclosure-to-ground plane transmission mode. TEV generally persists for a few microseconds. The magnitude varies along the enclosure; it can be in the range of 0.1 to 0.3 pu of the system voltage, and reaches the highest magnitude near the GIS-air interface. The TEV wave which couples onto the enclosure encounters earthing connections which form transmission line discontinuities and attenuate TEV. Mitigation methods include grounding using low surge impedance, short length leads and the installation of metal-oxide arresters across any insulating spacers. b) Transients on overhead connections a) Propagation of travelling waves caused by a disconnector operation 1 inside coaxial bus duct 2 on overhead line 3 between ground and encapsulation A portion of the VFT travelling wave incident at a gas-air transition is coupled onto the overhead connection and propagates to other components. This propagation is lossy and results in some increase of the waveform rise time, although transients can have rise times in the range of 10 to 20 ns if the air connection is relatively short. In general, external waveforms have two different characteristics * the overall waveshape is dictated by lumped circuit parameters, such as the capacitance of voltage transformers or line and earthing inductance; the rise time is in the range of a few hundred nanoseconds * a fast front portion which is dictated by transmission line effects; it has a rise time in the range of 20 ns and is usually reduced in magnitude due to discontinuities in the transmission path. The fast rise time of the initial portion is possible as capacitive components, such as bushings, are physically long and distributed, and they cannot be treated as lumped elements. The magnitude of the rise time portion of external transients is generally lower than that of internal VFT, the voltage rate-ofrise can be in the range of 10-30 kV/s. However, as VFT occur during normal conditions in GIS and each disconnector operation can generate tens to hundreds of individual transients, possible aging on the insulation of external components must be considered. These overvoltages can cause stress on adjacent equipment, and resonance phenomena can occur in exposed transformers. c) Transient electromagnetic fields c) Single-line diagram Figure 5. Diagram to explain the generation of TEV [20]. EMF are radiated from the enclosure and can cause some stress on secondary equipment, especially when sophisticated computer-controlled equipment may be affected. Their frequency depends on the GIS arrangement, but is typically in

b) Equivalent circuit

6-8

the range of 10 to 20 MHz.

2.3 EFFECTS ON EQUIPMENT The level reached by VFT overvoltages originated by disconnector switching or line-to-ground faults inside a GIS are below the BIL of substation and external equipment. However, aging of the insulation of external equipment due to frequent VFT must be considered. TEV is a low energy phenomenon, and it is not considered dangerous to humans. The main concern is in the danger of the surprise-shock effect. TEV can also cause interference with or even damage to the substation control, protection and other secondary equipment, and radiate EMF which may induce voltages and currents within electric circuits. The main effects caused by VFT to equipment and the techniques which can be used to mitigate these effects are summarized in Table 1.

assuming the external enclosure to be perfectly earthed. If TEV have to be considered, it is necessary to add one more mode (enclosure-ground) since at these high frequencies, the earth connections assume significant impedance values. A short explanation about the representation of the most important GIS components follows. a) Bus ducts For a range of frequencies lower than 100 MHz, a bus duct can be represented as a lossless transmission line. The surge impedance and the travel time can be calculated from the physical dimensions of the duct. Empirical corrections are usually needed to adjust the propagation velocity. Experimental results show that the propagation velocity in GIS ducts is close to 0.95 - 0.96% of the speed of light [14], [24]. The error committed by ignoring skin effect losses is usually negligible. Other devices such as elbows or closed disconnectors can also be modeled as lossless transmission lines. b) Surge arresters A surge arrester model should take into account the steep front wave effect : the voltage developed across the arrester for a given discharge current increases as the time to crest of the current increases, and reaches crest prior to the crest of the discharge current. A detailed model must represent each internal shield and block individually, and include the travel times along shield sections, as well as capacitances between these sections, capacitances between blocks and shields, and the blocks themselves. The model shown in Table 2 considers two sections, represented as lossless transmission lines, and a capacitance paralleled by a resistance between sections. Experimental results show that switching operations do not produce voltages high enough to cause MOVs to conduct. Although sophisticated models have been developed to represent an MOV, only its capacitance needs to be taken into account. c) Circuit breakers The representation of a circuit breaker is very complicated due to internal irregularities. In addition, circuit breakers with several chambers contain grading capacitors. As these components are not arranged symmetrically, a circuit breaker has a different transient response depending upon which terminal is connected to the surge source.

3. MODELING GUIDELINES Due to the travelling nature of VFT, modeling of GIS components makes use of electrical equivalent circuits composed of lumped elements and distributed parameter lines. At very high frequencies, the skin losses can produce a noticeable attenuation. However, due to the geometrical structure of GIS and the enclosure material, skin losses are usually neglected, which gives conservative results. Only the dielectric losses in some components, e.g. capacitively graded bushing, need be taken into account. The next two subsections present modeling guidelines to represent GIS equipment in computation of internal transients and TEV.

3.1 COMPUTATION OF INTERNAL TRANSIENTS All the distributed parameter lines take into account the internal mode (conductor-enclosure) only, assuming that the external enclosure is perfectly grounded. If TEV is of concern, then a second mode (enclosure-ground) is to be considered. Table 2 shows the equivalent circuits proposed to represent main components of a typical GIS [17]. More accurate models were presented in [27]. Distributed parameter models shown in Tables 2 take into account the internal mode (conductor-enclosure) only,

6-9

TABLE 1 - EFFECTS OF VERY FAST TRANSIENTS [17] EQUIPMENT VFT EFFECTS

Direct connected transformers can experience - an extremely nonlinear voltage distribution along the high voltage winding, connected to the oil-SF6 bushings, due to steep fronted wave impulses - extremely high part-winding resonance voltages due to transient oscillations generated within the GIS. The insulation system of breakers and loadbreak switches is not endangered by VFT overvoltages generated in adjacent GIS equipment. Ground faults induced by VFTO have been observed in disconnectors operations, as residual leader branches can be activated by enhanced field gradient to ground and by feeding them with GIS-generated VFTO. Breakdown caused by VFT overvoltages is improbable in a well-designed GIS insulation system during normal operations. However, breakdown values can be reduced by insulation irregularities like edges and fissures. The breakdown probability is very low for low VFTO amplitudes, it increases with the frequency of the oscillations and the degree of the field homogeneity. TEV can cause sparking across insulated flanges and to insulated busbars of CTs, and puncture of insulation which is intended to limit the spread of circulating currents within the enclosure. ENCLOSURE TEVs can be minimized - by a proper design and arrangement of substation mats - by keeping ground leads as short and straight as possible in order to minimize the inductance - by increasing the number of connections to ground - by introducing shielding to prevent internally generated VFT from reaching the outside of the enclosure - by installing voltage limiting varistors where spacers must be employed. Avoid high impedance in the connection of the last graded layer to the enclosure. Grounding connections must be modified to eliminate troubles. Correct cable connection procedures may minimize interference. The coupling of radiated energy may be reduced - by mounting control cables closely along the enclosure supports and other grounded structures - by grounding cable shields at both ends by leads as short as possible - by using optical coupling services. Voltage limiting devices may have to be installed.

MITIGATION TECHNIQUES
Transformers can withstand the stress built up by steep front waves. In critical cases, it might be necessary to install varistors to protect tap changers against very high frequency transient oscillations.

TRANSFORMERS

The development of a ground fault by branching of the leader discharge during a disconnector operation can be avoided by a proper disconnector design.

DISCONNECTORS & BREAKERS

SF6 INSULATION

BUSHINGS CABLES

SF6 insulated bushings can be affected as other SF6 equipment. Very few problems have been reported with capacitively graded bushings. No problems have been experienced on the main insulation. VFT effects appear always in grounded circuits and are originated by TEVs. TEV may interfere with secondary equipment or damage sensitive circuits - by raising the housing potential if they are directly connected - or via cable shields to GIS enclosure by emitting free radiation which may induce currents and voltages in adjacent equipment.

SECONDARY EQUIPMENT

TABLE 2 - GIS COMPONENT MODELS [17]

COMPONENT Bus duct Spacer Elbow Spherical shield Surge arrester Closed switch Open switch Closed disconnector Open disconnector Disconnector during sparking Bushing (capacitive type)

EQUIVALENT CIRCUIT

NOTES Loss-free distributed parameter transmission line ( C  20 to 30 pF)

(C  few pF)

(n = number of breaking chambers)

r = r(t) ; R = a few $ C = a few tens pF n = number of equivalent shields (5 to 8) simulated C = a few tens pF Zs 250 $ parameters evaluated from the frequency response of the transformer sometimes negligible

Bushing (gas filled) Power transformer (termination) Current transformer Capacitive voltage transformer Earth connection Aerial line or long cable (termination)

r = surge impedance

6-11

A closed breaker can be represented as a lossless transmission line. The surge impedance is calculated from the diameters of the conductor and enclosure. The electrical length is equal to the physical length. The propagation velocity is also reduced to 0.95 - 0.96% of the speed of light. The effect of grading capacitors can be ignored. The representation of a closed circuit breaker is more complicated because the electrical length is increased and the speed of progression is decreased due to the effects of the higher dielectric constant of the grading capacitors [24]. If the intermediate voltages are needed, the breaker is divided into as many sections as there are interrupters, all connected by the grading capacitors. A simpler model consists of two equal lengths of bus connected by a capacitor with a value equivalent to the series combination of all the grading capacitors. Parameters of the two bus sections are calculated from the physical dimensions of the breaker. A different representation has been proposed for circuit breakers which contain pre-insertion resistors [24]. d) Gas to air bushings A bushing gradually changes the surge impedance from that of the GIS to that of the line. A detailed model of the bushing must consider the coupling between the conductor and shielding electrodes, and include the representation of the grounding system connected to the bushing. A simplified model consists of several transmission lines in series with a lumped resistor representing losses. The surge impedance of each line section increases as the location goes up the bushing. If the bushing is distant from the point of interest, the resistor can be neglected and a single line section can be used [24]. More advanced models for capacitively graded bushings were proposed in [9], [16] and [28]. e) Power transformers A common practice is to model a power transformer as a capacitor representing the capacitance of the winding to ground. When voltage transfer has to be calculated, interwinding capacitances and secondary capacitance to ground must also be represented. At very high frequencies a winding of a transformer behaves like a capacitive network consisting of series capacitances between turns and coils, and shunt capacitances between turns and coils to the grounded core and transformer tank. The equivalent capacitance, Ce, is given by

ground must be added to (3) to obtain the total capacitance of the winding. Most of this terminal capacitance comes from the capacitance of the terminal bushing to ground. If voltage transfer is not of concern, an accurate representation can be obtained by developing a circuit that matches the frequency response of the transformer at its terminals. At very high frequencies, the saturation of the magnetic core can be neglected, as well as leakage impedances. f) Current transformers Insulating gaps are usually installed in the vicinity of current transformers. During high voltage switching operations, these gaps flash over, establishing a continuous path. Travelling waves propagate with little distortion. Current transformers can often be neglected. The parameters needed to represent these models can be determined either from manufacturer's data or by calculation based on the physical sizes of the equipment. If neither of these is possible, the capacitance values can be estimated from those shown in Table 2, while surge impedances can be estimated around 50-80 ohms.

3.2 COMPUTATION OF TEV At the frequencies of the VFT caused by dielectric breakdown within the GIS (breakdown across disconnect switch contacts, line-to-ground faults), currents are constrained to flow along the surface of the conductors and do not penetrate through them. The inside and the outside of the enclosure are distinct, so that transients generated within the GIS do not appear onto the outside surface of the enclosure until discontinuities in the sheath are encountered. These discontinuities occur at gasto-air terminations (the most frequent case), GIS-cable transitions, or external core current transformers. The modeling of the GIS for computation of TEV must include the effects of the enclosure, the representation of ground straps and the earthing grid. A GIS-air termination can be modeled as a junction of three transmission lines each with its own surge impedance, see Figure 5. This equivalent network can be analyzed using lossless transmission line models to determine reflected and transmitted waves. The basic mechanism of TEV is defined by the refraction of waves from the internal coaxial bus duct to the enclosure sheath-to-ground system. The travelling wave incident onto the GIS-air termination is reflected at this termination being the magnitude of the transmitted wave onto

Ce
Cg Cs

(3)

where Cs and Cg are the equivalent series and ground capacitances of the winding. The equivalent series capacitance, Cs, is more difficult to compute. The details of computation have been discussed in [37]. The terminal capacitance to 6-12

the outside of the enclosure sheath given by the coefficient

(Z1Z2Z3)

2Z3

Zs
60 ln

2 h r

(6)

(4) where r is the strap radius and h the average height of the section. The representation of the earthing grid at TEV frequencies is a very complex task. Furthermore, this grid may not be designed to carry very high frequency currents, as no standards for very high frequency earthing systems are currently available. A simplified modeling may be used by representing the earthing grid as a low value constant resistance. Advanced models for GIS components in computation of TEV might consider a frequency-dependent impedance for ground straps, a frequency-dependent model for the enclosureto-ground line (which could take into account earth losses) and the propagation of phase- to-phase modes on the three enclosures [6]. Distributed parameter models shown in Tables 2 take into account the internal mode (conductor-enclosure) only, assuming the external enclosure to be perfectly earthed. If TEV have to be considered, it is necessary to add one more mode (enclosure-ground) since at these high frequencies, the earth connections assume significant impedance values. A more detailed model is then required, see Section 5.3.

where Z1, Z2 and Z3 are the surge impedance of the coaxial bus duct, the overhead line and the sheath-to-ground system, respectively [5]. The negative sign means that there is an inversion of the waveform with respect to the internal transient. TEV propagates back from the gas-to-air termination into the substation on the transmission line defined by the enclosure and the ground plane. The first discontinuity in the propagation is generally a ground strap. For TEV rise times, most ground straps are too long and too inductive for effective grounding. However, ground leads may have a significant effect on the magnitude and waveshape of TEV. This effect can be explained by considering two mechanisms [6] : * the ground lead may be seen as a vertical transmission line whose surge impedance varies with height; when the transient reaches the ground strap, a reflected wave is originated which reduces the magnitude of the transmitted wave, being the reduction expressed by the coefficient

2Zg 2Zg

 Z3

(5)

where Zg is the surge impedance of the ground strap and Z3 the surge impedance of the enclosure-to-ground transmission line; as Zg is usually much larger than Z3, the attenuation produced by the ground strap will be usually small * the portion of the wave which propagates down the ground strap meets the low impedance of the ground grid, then a reflected wave will be produced at this point which propagates back to the enclosure where it will tend to reduce the original wave. The representation of a ground lead as a constant surge impedance is not strictly correct. In reality, the ground strap has a continuously varying surge impedance, so that a continuous reflection occurs as a wave propagates down the lead. An analysis of the performance of different models for the ground lead was presented in [6]; simulation results did show that a constant inductor model may be adequate for straps with travel time less than the surge rise time, while a nonuniform impedance model may be necessary for much larger straps. Reference [6] proposes to divide a ground strap into sections, each one represented by a surge impedance calculated from the following expression 6-13

4. VALIDATION The accuracy of a simulation depends on the quality of the model of each individual GIS component. In order to achieve reasonable results even for time periods of some micro-seconds or for very complex GIS structures, highly accurate models for each internal component and also for external components, connected to the GIS, are necessary. Figure 6 shows an example of how a 420 kV disconnector with an arrangement of spacers, shielding electrodes and varying diameters can be simulated by the equivalent circuit derived from its geometry. Figure 6.c shows the measured step response and the simulation result using the equivalent circuit shown in Figure 6.b. An excellent correlation between measurements and calculations can be observed.

a) Arrangement of GIS equipment

Figure 7. Comparison of simulation and measurement of disconnect switch induced overvoltages in a 420 kV GIS.

b) Equivalent circuit

c) Field measurement and simulation result Figure 6. Validation of a 420 kV disconnector operation. An accurate modeling of each individual GIS component makes it possible to reproduce VFT waveforms with a relatively high precision, especially in short GIS structures or test equipment. Figure 7 demonstrates the accuracy of such techniques by comparing a computer simulation with a direct measurement of a transient waveform in an actual GIS. The simulation performed neglects the presence of propagation losses which result in somewhat less damping of the high frequency part of the waveform. The effects of spacers, flanges, elbows, corona shields, and other connection hardware, were included in the model. 6-14

a) Diagram of the substation

b) Comparison of simulation results and measurements at measuring point M2 Figure 8. Measurement and simulation of overvoltages in a

420 kV GIS at closing of disconnect switch Q12. Amplitude and frequency of VFT can be reproduced with good accuracy for several periods and many details in the waveform can be explained. Figure 8 shows a part of a 420 kV GIS on which measurements have been made at point M2. By closing disconnect switch Q12, one part of the switchyard has been connected. The measurements have been reproduced by a detailed simulation. Certain differences in the range of higher frequencies did occur because the simulation was performed with a low damping equivalent circuit and measuring instruments did not capture very high frequencies. However, the main waveform has been reproduced with sufficient accuracy.

5.1 CLOSING OPERATION IN A 420 kV GIS [35] Detailed calculations have been made for a 420 kV GIS with 2 line feeder bays, a transformer and a bus coupler, see Figure 9. The double busbar system also included bus sectionalizers. Single-phase enclosure is applied to the bays and three-phase enclosure for the busbars. According to their internal design, all GIS components have been represented thoroughly by line sections with the corresponding surge impedance and transit time, and by lumped capacitances for spacers and additional capacitances caused by internal shielding devices, see Figures 11 to 14. The three-phase encapsulated busbars are represented by surge impedances in positive and zero sequence system. Detailed data are given in Appendix A, Tables A1 to A4. Capacitive grading of the bushings has been simulated assuming two representative screens for each. The behavior of the transformer winding under high-frequency transients has been simulated by an equivalent circuit, see Figure 13, proposed by the manufacturer, according to the high-frequency measurements performed in the factory. The behavior of the spark in the disconnecting switch during closing was represented by a fixed resistance of 0.5 ohms in series with an exponentially decreasing resistance, R = R0 exp(t/T), with R0 = 1012 ohms and T = 1 ns, resulting in a time duration of voltage breakdown of about 10 ns. Calculations were performed for a closing operation of the by-pass disconnecting switch in the line feeder bay 1, see Figure 10, connecting the busbar 1 to the voltage source. Time-step size was 0.15 ns. Simulation results are shown in Figure 15 for the measuring points M1 to M5 : * point M1 - node 35 in Figure 13 -, voltage stress at the connection inside the transformer between bushing and transformer winding * point M2 - node 34 in Figure 13 -, high-frequency oscillations at the GIS-side of the transformer bushing are caused mainly by the capacitive grading system of the bushing * point M3 - node 10 in Figure 11 -, the voltage oscillation near the end of the switched busbar SS2 starts from nonzero conditions, since part of the feeding voltage on busbar SS1 is capacitively coupled to busbar SS2 via the capacitance of the open circuit breaker in the coupling bay * point M4 and M5 correspond to nodes 3 in Figure 11 and node 18 in Figure 13, respectively. The investigation clearly shows that very detailed information of the internal design not only of the GIS but also of the external equipment, like bushings and transformer windings, is necessary to achieve reliable results.

5. CASE STUDIES As it has been explained above, VFT in GIS are caused by dielectric breakdowns. The collapsing electric field during a breakdown produces travelling waves which propagate in both directions from the disturbance location. This propagation can be analyzed and simulated using transmission line theory, and assuming that propagation losses are negligible. Travelling waves appear externally at enclosure transitions, e.g. gas-to-air terminations. At these transitions, reflected waves travelling back onto the station and transmitted waves coupled onto the outside of the enclosure sheaths are generated. The magnitude of the travelling waves will depend on their source (disconnect switch operation, fault) and the GIS configuration. Depending on the transient of concern, a different modeling is to be considered. Guidelines for simulation of internal and external transients were discussed in Section 3. Due to the very high frequencies generated by a dielectric breakdown within the GIS, a digital simulation is restricted to calculations during the VFT waveform period, usually 1 or 2 s. If the simulation is performed with an EMTP-like program, which uses a constant time step size, then the value of this step size will depend on the shorter transit time in the GIS. This step size must be equal or smaller than one-half the shorter transit time. Three case studies are included in this section. The first one presents the simulation of VFT in a 420 kV substation generated by a closing operation. The next two cases are related to low voltage tests in a 765 kV GIS. Low voltage tests are a very useful tool for development and validation of GIS models. The first of these two cases presents the simulation of internal transients, while the second one is aimed at calculating TEV. 6-15

Figure 9. Basic arrangement of the 420 kV GIS.

Figure 10. Configuration of the 420 kV GIS used for the simulation. 6-16

Figure 11. Arrangement of busbars, including bus sectionalizer.

Figure 13. Line feeder.

Figure 12. Transformer feeder. Figure 14. Bus coupler.

6-17

5.2 LOW VOLTAGE TEST OF A 765 kV GIS [24] Figures 16 and 17 show the one-line and the connectivity diagram of a 765 kV test bay. Models used to represent components of this case are presented in Table 2. The procedure followed to develop these models is detailed in [24]. A summary of this procedure follows : 1) Low voltage tests on individual components were performed using waves with fronts of 4 and 20 ns. 2) Models based on physical dimensions were developed, assuming a propagation velocity equal to that of light. 3) Digital models were adjusted so simulation results were matched to measurements. The main adjustment was to decrease propagation velocity to 0.96 that of light. Two transients have been reproduced in the first one, a ramp voltage is applied at t = 0 in the second case, the ramp voltage source is also used but the transient starts after closing a switch at the instant the ramp reaches its maximum value.

Waveforms obtained for each case at two nodes are shown in Figures 18 and 19. It can be observed that waveforms for both cases are essentially the same, except for the first nanoseconds in the vicinity of the input node UC1. These simulation results were validated by comparison with low voltage measurements. For normal studies, the input wave will be one of three forms * a ramp voltage with a magnitude determined by the voltage across the switch * two ramp currents on opposite sides of the switch such that the voltage across the switch is equal to zero at the crest of the inputs * charge both sides of the switch to the desired value and close the switch.

Figure 15. Simulation results. 6-18

Figure 16. One-line diagram of a 765 kV GIS.

Figure 17. Connectivity diagram of the 765 kV GIS. 6-19

5.3 CALCULATION OF TEV IN A 765 kV GIS [24] Modeling of GIS components to simulate TEVs must include the effects of * an enclosure, which cannot be assumed to be continuously grounded * the surge impedances and lengths of the grounding structures * the resistance of the earth ground. Each GIS component can be represented as a two-phase ideal transmission line defined by two modal parameters, Z0 and Z1, whose values can be approached as follows [24] Z = 60 ln(D/d) Z1 = Z/2 Z0 = 20000 - Z1

a) Voltage at location UC1

where D is the inside diameter of the enclosure and d is the outside diameter of the center conductor. At the bushing, the two modes split with one going up to the bushing and the other connected to the grounding surge impedance, see Figure 20.

b) Voltage at location UK Figure 18. Simulation results with 4 ns ramp.

Figure 20. Connection of GIS to an air bushing. Figures 21 through 24 show simulation results at different locations produced by a 4 ns ramp as input voltage.

6. CONCLUSIONS a) Voltage at location UC1 A description of the origin and main characteristics of VFT in GIS, as well as their effects on substation and adjacent equipment, has been summarized in this document. Modeling guidelines for digital simulation of GIS networks in VFT studies have been discussed. Their application was illustrated with three case studies. Although guidelines proposed in this document neglect propagation losses for many GIS components and very simple models are proposed for most components, validation tests have shown that an excellent correlation between simulation results and field measurements can be achieved. More accurate models may be needed in some cases for which propagation losses at very high frequencies should not be neglected.

b) Voltage at location UK Figure 19. Simulation results from closing a switch.

6-20

7. REFERENCES A. Ecklin, D. Schlicht and A. Plessl, "Overvoltages in GIS caused by the operation of isolators", Surges in high-voltage networks, K. Ragaller (Ed.), pp. 115-129, Plenum Press, 1980. [2] S. Narimatsu et al., "Interrupting performance of capacitive current by disconnecting switch for gas insulated switchgear", IEEE Trans. on Power Apparatus and Systems, vol. 100, no. 6, pp. 2726-2732, June 1981. [3] S. Matsumara and T. Nitta, "Surge propagation in gas insulated substation", IEEE Trans. on Power Apparatus and Systems, vol. 100, no. 6, pp. 3047-3054, June 1981. [4] S.A. Boggs et al., "Disconnect switch induced transients and trapped charge in gas-insulated substations", IEEE Trans. on Power Apparatus and Systems, vol. 101, no. 6, pp. 3593-3602, October 1982. [5] N. Fujimoto, E.P. Dick, S.A. Boggs and G.L. Ford, "Transient ground potential rise in gas-insulated substations - Experimental studies", IEEE Trans. on Power Apparatus and Systems, vol. 101, no. 6, pp. 3603-3609, October 1982. [6] E.P. Dick, N. Fujimoto, G.L. Ford and S. Harvey, "Transient ground potential rise in gas-insulated substations - Problem identification and mitigation", IEEE Trans. on Power Apparatus and Systems, vol. 101, no. 6, pp. 3610-3619, October 1982. [7] L. Blahous and T. Gysel, "Mathematical investigation of the transient overvoltages during disconnector switching in GIS", IEEE Trans. on Power Apparatus and Systems, vol. 102, no. 9, pp. 3088-3097, September 1983. [8] G. Bernard, J. Massat, G. Ebersohl and G. Voisin, "Study of electromagnetic transients due to disconnector switching in metal enclosed substations", Revue Gnrale de l'Electricit, no. 11, pp. 667-694, November 1983. [9] R.J. Harrington and M.M. El-Faham, "Proposed methods to reduce transient sheath voltage rise in gas insulated substations", IEEE Trans. on Power Apparatus and Systems, vol. 104, no. 5, pp. 1199-1206, May 1985. [10] J. Lalot, A. Sabot, J. Kieffer and S.W. Rowe, "Preventing earth faulting during switching of disconnectors in GIS including voltage transformers", IEEE Trans. on Power Delivery, vol. 1, no. 1, pp. 203211, January 1986. [11] S. Ogawa et al., "Estimation of restriking transient overvoltage on disconnecting switch for GIS", IEEE [1]

Figure 21. Voltage between the center conductor and the enclosure at location UC1.

Figure 22. Voltage between the enclosure and ground at the base of the bushing.

Figure 23. Voltage between the center conductor and ground at location UK.

Figure 24. Voltage between the center conductor and the enclosure at location UK. 6-21

Trans. on Power Delivery, vol. 1, no. 2, pp. 95-102, April 1986. [12] T. Yoshida et al., "Distribution of induced grounding current in large-capacity GIS using multipoint grounding system", IEEE Trans. on Power Delivery, vol. 1, no. 4, pp. 120-127, October 1986. [13] J. Ozawa et al., "Suppression of fast transient overvoltage during gas disconnector switching in GIS", IEEE Trans. on Power Delivery, vol. 1, no. 4, pp. 194-201, October 1986. [14] N. Fujimoto, H.A. Stuckless and S.A. Boggs, "Calculation of disconnector induced overvoltages in gas-insulated substations", Gaseous Dielectrics IV, Pergamon Press, 1986. [15] R. Boersma, "Transient ground potential rises in gasinsulated substations with respect to earthing systems", Electra, no. 110, pp. 47-54, January 1987. [16] R. Witzmann, "Fast transients in gas insulated substations. Modelling of different GIS components", 5th Int. Symposium on HV Engineering, Paper no. 12.06, Braunschweig, 23-28 August, 1987. [17] CIGRE Working Group 33/13-09, "Very fast transient phenomena associated with gas insulated substations", CIGRE Paper No. 33-13, 1988. [18] CIGRE WG 33/13-09, Monograph on GIS Very Fast Transients, 1988. [19] N. Fujimoto and S.A. Boggs, "Characteristics of GIS disconnector-induced short risetime transients incident on externally connected power system components", IEEE Trans. on Power Delivery, vol. 3, no. 3, pp. 961-970, July 1988. [20] J. Meppelink, K. Diederich, K. Feser and D.W. Pfaff, "Very fast transients in GIS", IEEE Trans. on Power Delivery, vol. 4, no. 1, pp. 223-233, January 1989. [21] CIGRE Working Group 33.02, Guidelines for representation of networks elements when calculating transients, 1990. [22] S. Yanabu et al., "Estimation of fast transient overvoltage in gas-insulated substation", IEEE Trans. on Power Delivery, vol. 5, no. 4, pp. 1875-1882, October 1990. [23] A.M. Miri and M. Schelker, "ATP simulation of transient ground potential rise in gas-insulated substations", EMTP News, vol. 4, no. 3, pp. 14-21, September 1991. [24] D.L. Nickel, "Very fast transients in Gas-insulated substations", EPRI Report, 1991. [25] S. Okabe, M. Kan and T. Kouno, "Analysis of surges measured at 550 kV substations", IEEE Trans. on Power Delivery, vol. 6, no. 4, pp. 1462-1468, October 1991. [26] H. Toda et al., "Development of 800 kV gas-insulated switchgear", IEEE Trans. on Power Delivery, vol. 7, 6-22

no. 1, pp. 316-323, January 1992. [27] Z. Haznadar, C. Carsimamovic and R. Mahmutcehajic, "More accurate modeling of gas insulated substation components in digital simulations of very fast electromagnetic transients", IEEE Trans. on Power Delivery, vol. 7, no. 1, pp. 434-441, January 1992. [28] A. Ardito et al., "Accurate modeling of capacitively graded bushings for calculation of fast transient overvoltages in GIS", IEEE Trans. on Power Delivery, vol. 7, no. 3, pp. 1316-1327, July 1992. [29] A.M. Miri and M.A. Nothaft, "Simulation of the effects of ZnO-varistors in reducing the transient ground potential rise between a GIS enclosure and the cable outlet flange", Proc. of the First European Conference on Power Systems Transients, pp. 53-60, Lisbon, June 17-18, 1993. [30] H.W. Dommel, "Simulating travelling waves inside and outside GIS enclosures with the EMTP", presented at the Canadian Electrical Association, Toronto, March 1994. [31] A.M. Miri and C. Binder, "Investigation of transient phenomena in inner- and outer systems of GIS due to disconnector operation", Proc. of the Int. Conference on Power Systems Transients, pp. 71-76, Lisbon, September 3-7, 1995. [32] Y. Yamagata et al., "Suppression of VFT in 1100 kV GIS by adopting resistor-fitted disconnector", IEEE Trans. on Power Delivery, vol. 11, no. 2, pp. 872-880, April 1996. [33] M. Ianoz, L. Dellera, C.A. Nucci and L. Quinchon, "Modeling of fast transient effects in power networks and substations", CIGRE Paper 36-204, 1996. [34] K. Mizuno et al., "Investigation of PD pulse propagation characteristics in GIS", IEEE Transmission and Distribution Conference Proceedings, pp. 204-212, Los Angeles, September 15-20, 1996. [35] IEEE TF on Very Fast Transients (D. Povh, Chairman), "Modelling and analysis guidelines for very fast transients", IEEE Trans. on Power Delivery, vol. 11, no. 4, October 1996. [36] IEC 71-2, Insulation Co-ordination - Part 2 : Application Guide, 1996. [37] P. Chowdhuri, Electromagnetic Transients in Power Systems, RSP-John Wiley, 1996. [38] CIGRE Joint WG 33/23.12, "Insulation co-ordination of GIS : Return of experience, on site tests and diagnostic techniques", LECTRA, no. 176, pp. 66-97, February 1998.

APPENDIX A - DATA OF THE 420 kV GIS TABLE A1 - LINE FEEDER DATA

Branch 1 2 3 4 5 6 7 7 9 10 11 12 13 13 15 16 17 18 19 20 21 22 23 24 8 17 26 27 28 27 30 31 32 2 5 6 8 11 13 17 21 22 23 24 25 26 27 29 30 31 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1) 17 18 19 20 21 22 23 24 25 25 26 27 28 29 30 31 32 33 2) E E E E E E E E E E E E E E E E E Length (m) 1.20 0.75 0.15 0.30 0.45 0.45 0.15 0.75 -----0.30 1.10 1.15 1.20 0.75 -----0.30 0.65 0.40 1.10 0.30 0.70 1.10 1.15 0.70 2.30 0.80 0.90 0.60 0.60 4.00 0.15 1.50 4.90 -------------------------------------------------------------------------------------Z ( $) 95 42 42 42 95 95 42 42 --42 95 95 95 42 --42 95 42 42 95 95 66 66 95 95 95 72 95 25 95 95 95 95 ----------------------------------C (pF) ----------------4 ----------4 ------------------------------------10 14 10 10 24 10 10 10 32 32 32 10 10 10 10 16 16

TABLE A2 - TRANSFORMER FEEDER DATA

Branch 1 2 3 4 5 6 7 7 9 10 11 12 13 13 15 16 17 18 19 20 21 22 23 24 8 17 26 27 28 29 30 31 32 31 34 2 5 6 8 11 13 17 22 23 24 25 26 27 28 29 30 31 33 34 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 25 26 27 28 29 30 31 32 33 34 35 1) E E E E E E E E E E E E E E E E E E E Length (m) 1.20 0.75 -----0.30 0.45 0.45 0.15 0.75 0.15 0.30 1.10 1.15 1.20 0.75 -----0.30 0.65 0.40 1.10 0.30 0.70 1.10 1.15 0.70 2.30 0.80 0.90 0.80 0.90 1.50 0.80 0.60 0.60 11.0 2.30 -----------------------------------------------------------------------------------------------Z ( $) 95 42 --42 95 95 42 42 42 42 95 95 95 42 --42 95 42 42 95 95 66 66 95 95 95 72 95 95 95 95 95 25 25 126 --------------------------------------C (pF) ----4 ----------------------4 ----------------------------------------10 14 10 10 24 10 10 10 32 32 32 10 10 16 10 10 10 10 14

1) by-pass disconnect switch

2) special bushing representation 1) special bushing representation

6-23

TABLE A3 - BUSBAR DATA

Branch 1 2 3 4 5 6 7 8 9 10 1 2 3 4 8 9 10 11 2 3 4 5 6 7 8 9 10 11 E E E E E E E E Length (m) 1.80 3.70 5.60 0.90 0.85 0.15 0.20 1.80 7.40 1.80 ----------------------------------------Z0 ( $) 157 157 157 157 104 104 104 157 157 157 --------------------------------Z1 ($) 113 113 113 113 60 60 60 113 113 113 --------------------------------C (pF) --------------------12 3 3 3 3 3 3 12

APPENDIX B - DATA OF THE 765 kV GIS

Branch

Z ( $) J3 J4 T22 T23 D9 D88 D66 D22 D1 T21 T20 T19 T18 T17 T16 J7 T24 T25 T26 T27 T14 T13 T11 T12 J2 T9 T10 T28 J6 UK T29 J5 T30 T32 J1 T4 T3 T5 T6 75 75 75 51 78 68 59 33 330 75 51 160 65 75 65 75 75 51 75 51 160 51 75 65 75 75 51 160 75 75 65 75 75 51 75 75 51 75 51

Travel time (ns) 6.40 48.0 2.20 1.90 2.20 1.80 4.20 5.80 9.10 2.20 1.90 0.67 1.70 6.80 1.70 8.50 2.20 1.90 2.20 1.90 0.67 1.90 9.90 1.70 7.50 2.20 1.90 0.67 7.10 6.40 1.70 8.80 2.20 1.90 6.70 2.20 1.90 2.20 1.90

TABLE A4 - BUS COUPLER DATA

Branch C 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 8 2 6 7 8 11 14 15 16 17 18 22 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 14 E E E E E E E E E E E Length (m) 1.20 0.45 0.15 0.90 0.60 0.90 1.00 0.30 -----0.80 0.85 -----0.30 0.70 1.70 0.45 0.45 0.45 0.15 0.75 0.15 1.20 ------------------------------------------------------------Z ( $) 95 42 42 42 95 72 95 66 --66 66 --66 95 95 95 95 42 42 42 42 95 ------------------------(pF) ----------------4 ----4 --------------------900 10 10 10 32 40 32 10 24 10 14 10

UC1 J3 J4 T22 J4 D9 D88 D44 D22 J3 T21 T20 T19 T19 T17 T17 J7 T24 J7 T26 T17 T14 T13 T11 T11 J2 T9 T10 T28 J6 T28 T28 J5 T30 J2 J1 T4 J1 T5

6-24