Fault resistances in medium voltage distribution networks.
Faults in electrical networks can occur on one phase, on two phases, or on all three phases. Earth faults are the most frequent faults in medium voltage distribution networks. Faults are either temporary or permanent. Majority of earth faults on overhead lines are temporary. They are cleared either by self-extinguishment or by automatic reclosing (high speed auto-reclosing or delayed auto-reclosing). Permanent faults require utility patrol action and repair.
Fault resistance of a typical earth fault is composed of three parts: arc resistance, resistance of intermediate fault path and the ground resistance.
Arc resistance is non-linear. Arc characteristics are nonlinear voltage-current dependence, random behaviour, fluctuations, and harmonics.
Intermediate fault path could consist of a tree, the surface contact between the conductor and the ground or anything else that the conductor could rest on (Carpenter et al., 2005).
Failing insulators, tree branches touching phase conductors, and downed conductors may cause high impedance earth fault (HIF) on overhead lines. The levels of currents of these faults are often much smaller than detection thresholds of traditional earth fault detection devices. Therefore, it is difficult to detect these faults (Zamanan et al., 2007).
Many techniques have been proposed to improve the detection of HIFs. Some of these techniques have been implemented, either at the prototype level or at the production level, others have only been suggested (Nengling & JiaJia, 2007).
A series of staged-fault tests were performed at the 12.5 kV voltage level (in the USA), in order to collect data on typical fault currents on various surfaces. Some of results for fault resistances [R.sub.F] were (Goodfellow, 2007):
--concrete (reinforced): [R.sub.F] = 100 [OMEGA]
--wet grass, wet sod: [R.sub.F] = 150-180 [OMEGA]
--dry grass, dry sod: [R.sub.F] = 280-360 [OMEGA]
--wet sand: [R.sub.F] > 500 [OMEGA]
--dry sand, dry asphalt: [R.sub.F] >> 500 [OMEGA]
--trees (fault pathway): [R.sub.F] = 150-250 k[OMEGA]
Results may considerably vary depending on season, precipitations, and other climatic changes.
2. MV NETWORK MODEL
The developed model of a typical real 20 kV medium voltage distribution network is reported under then References (Cucic, 2005). The model is verified by field tests with artificial earth faults in real network. The network is of mixed-type, i.e. it consists of both overhead lines and cables. The total capacitive current of the network is 137 A (actual coil step 100 A). Neutral earthing system has recently changed from unearthed to partially compensated. Partially compensated system includes fixed coil tunable off-load and resistor in parallel. Coil is tunable by means of fixed steps of 100-150-200-250-300 A. The resistance value of a parallel resistor is 80 [OMEGA] (150 A). The same value is usually used in low-resistance neutral earthing systems.
3. CALCULATION RESULTS
The effect of fault resistance ([R.sub.F]) is investigated for earth faults at a substation 20/0.4 kV, 713 m away from (i.e. relatively near) supply substation 110/20 kV. Grounding resistance 0.34 [OMEGA] of a substation 20/0.4 kV is considered. Calculations are performed for different fault resistances, varying from 0.34 [OMEGA] up to 20000 [OMEGA].
Fig. 1 shows the effect of [R.sub.F] on voltages (RMS values) at 20 kV supply busbars (at substation 110/20 kV), during an earth fault in an unearthed network. Abscissa is in a logarithmic scale, in order to obtain more visible dependency also for low resistance values.
It can be seen that, in case of a bolted earth fault (R.sub.F] = 0 [OMEGA], i.e. 0.34 [OMEGA]) in an unearthed network, phase-to-ground voltages on the healthy phases (L2 and L3) can reach [square root of 3] times higher values than in normal operation, i.e. the phase-to-phase nominal voltage (20 kV). Simultaneously, phase-to-ground voltage on the faulty phase (L1) has a very low value. Therefore, system must have a phase-to-phase insulation level.
[FIGURE 1 OMITTED]
Nevertheless, maximum temporary (50 Hz) overvoltage does not appear with bolted earth faults, but with fault with approximately [R.sub.F] = 20 [OMEGA] (overvoltage factor 1.87).
Generally (for the phase L3, above [R.sub.F] = 20 [OMEGA]), as [R.sub.F] increases, voltages on healthy phases decrease and the voltage on faulty phase increases. It is observed that the voltage of the healthy phase L2 and the voltage of faulty phase L1 significantly change in the fault resistance range between 20 [OMEGA] and 200 [OMEGA]. On the other hand, voltage of healthy phase L3 as well as zero-sequence voltage (residual voltage) 3[U.sub.0], significantly change in a range between 50 [OMEGA] and 500 [OMEGA]. However as [R.sub.F] increases above 500 [OMEGA] toward 20 k[OMEGA], the aforementioned voltages do not change much.
Earth fault initial transients are mostly caused by the voltage rise of the two healthy phases. Maximum transient overvoltage on healthy phases occur if the earth fault happens at the instantaneous voltage maximum in faulty phase. Calculated overvoltage factors are given in Table 1.
Considering different neutral earthing systems, it can be observed that the maximum overvoltage is lower in partially compensated network (maximum overvoltage factor 2.39) than in unearthed network (maximum overvoltage factor 2.58). Such overvoltages can cause an insulation failure at weak points of the healthy phases thus leading to a double fault to earth.
In a real system, any resistance may damp transient phenomena. Thus, the overvoltage factors decrease with the increase of fault resistances. Notice that fault resistance, at which initial transient becomes overdamped, is lower in partially compensated network than in unearthed network. Results have shown, that "overdamping" fault resistance ranges between 20 [OMEGA] and 50 [OMEGA].
The effect of a wide range of variation of fault resistance [R.sub.F] on zero-sequence current (residual current) 3[I.sub.0] in a faulty feeder, as well as on fault current [I.sub.k1] at the fault point, is depicted in Fig. 2.
[FIGURE 2 OMITTED]
In case of a bolted earth fault ([R.sub.F] = 0 [OMEGA], i.e. 0.34 [OMEGA]) in unearthed network, fault current [I.sub.k1] is equal to the total capacitive current of the network (if conductance (G) of feeders is neglected). We can notice, that there is a difference between earth fault current at the fault point and residual current at sending end of faulty line, because of the capacitive current of the faulty feeder. However, the difference for higher fault resistances is smaller.
As is known, the fault resistance at the fault point reduces the earth fault current. We can see from Fig. 2 that the fault resistance of some tens or hundreds of ohms drastically reduces the fault current. For a very high resistive earth fault, magnitude of residual current (and residual voltage) is small, so it is unlikely that a conventional protection will operate.
Most directional overcurrent relays for sensitive earth fault protection use fundamental-frequency methods based on measured phasors (and angles) of residual current and residual voltage. Among other parameters, fault detection sensitivity depends also on normal system unbalance, on type of used current transformer (sensitivity is increased if we use torodial current transformer instead of Holmgreen connected current transformers) and on current transformer ratio (sensitivity increases if we use current transformer with lower ratio).
Current transformers with ratio 300/5 are used on feeders in real network. Directional overcurrent relays are adjusted to 2.6% [I.sub.n] and 1.3% [I.sub.n], which yield 8 A and 4 A respectively. Calculation results show that these values correspond to fault resistances 1130 [OMEGA] and 2280 [OMEGA]. We can conclude that the earth faults with fault resistances higher than those mentioned above and with existing protection scheme will not be detected (located) and cleared.
This paper has shown the effects the fault resistances at the fault location have on voltage and current magnitudes. Fault resistances of some tens or hundreds of ohms considerably reduce both currents and voltages. Maximum calculated overvoltage factors (transient overvoltages during an earth fault) in unearthed and partially compensated network were 2.58 and 2.39 respectively. Transient and temporary overvoltages could cause an insulation failure. Future work should further investigate how many and how often faults with fault resistances higher than few kohms occur and depending on the results, if it is necessary to reconsider possible change in the existing relay protection scheme, or to upgrade it with special algorithm(s) used for detection and location of HIFs. That is particularly important for resonant earthed networks, due to lack of selectivity for HIF detection in those systems.
Carpenter, M; Hoad, R. F. & Bruton, T. D. (2005). Staged-Fault Testing For High Impedance Fault Data Collection, Proceedings of the 58th Annual Conference for Protective Relay Engineers, April 5-7, 2005
Cucic, R. (2005). Modelling of the medium voltage distribution network for purpose of simulating single-phase-to-earth faults, UL-FE, Ljubljana (in Slovenian)
Goodfellow, J. W. (2007). Understanding how trees cause interruptions, IEEE Distribution Subcommittee Meeting, 8-10 January 2007, Orlando, Florida, USA
Nengling, T & JiaJia, C. (2007). Wavelet-based approach for high impedance fault detection of high voltage transmission line, European Transactions on Electrical Power (ETEP)
Zamanan, N.; Sykulski, J. & Al-Othman, A. K. (2007). Arcing high impedance fault detection using real coded genetic algorithm, Power and Energy Systems, April 2007, Thailand
Table 1. Overvoltage factors in unearthed and partially compensated network Fault Overvoltage Partially Resistance Factor Unearthed Compensated [[OMEGA]] Overvoltage Network Network 0.34 Transient 2.585 2.394 0.5 2.573 2.383 1 2.541 2.354 2 2.493 2.296 5 2.350 2.136 10 2.153 1.932 15 1.997 1.782 20 1.873 1.667 0.34 Temporary 1.766 1.666
|Printer friendly Cite/link Email Feedback|
|Author:||Komen, Vitomir; Cucic, Renato|
|Publication:||Annals of DAAAM & Proceedings|
|Article Type:||Technical report|
|Date:||Jan 1, 2007|
|Previous Article:||Towards accurate, low-cost vision system based on time-of-flight (TOF) laser scanner for mobile robot applications.|
|Next Article:||Modern sensor systems of intelligent robots based on the slip displacement signal detection.|