Power Quality Blog

VCB Current Chopping Evaluation

Published by Electrotek Concepts, Inc., PQSoft Case Study: VCB Current Chopping Evaluation, Document ID: PQS1207, Date: June 1, 2012.

Abstract: This case study presents a customer vacuum circuit breaker (VCB) current chopping transient overvoltage evaluation. A high-frequency transient model was created to simulate vacuum circuit breaker opening and closing and the resulting transient overvoltages and arrester energy duties. Vacuum circuit breaker operations are one of the causes of high rate-of-rise (dv/dt) transients. The simulation results show that properly – designed R-C snubbers will reduce the initial rate-of-rise of the transient voltages, which may be beneficial because severe dv/dt transient voltages can damage the first few turns of insulation of dry-type transformers and motors or excite internal transformer resonances producing severe overvoltages.

VCB CURRENT CHOPPING TRANSIENT CASE STUDY

A customer vacuum circuit breaker (VCB) current chopping transient overvoltage evaluation was completed for the system shown in Figure 1. The principal objectives of the case study were to determine transient overvoltages and evaluate mitigation alternatives during vacuum circuit breaker current chopping events. The power conditioning mitigation alternatives of MOV surge arresters and R-C snubbers were also evaluated.

The simulations for the case study were completed using the PSCAD® program. A high frequency transient model was created to simulate the vacuum circuit breaker current chopping transients and resulting overvoltages and arrester energy duties. A high frequency model was required to accurately represent the very high current chopping transient frequencies.

A high-frequency transient model of a portion of the customer facility and the adjacent power system was created using the PSCAD simulation program. The transient simulation model consisted of an equivalent source impedance at the 21kV service entrance, two 2,000 kVA step-down transformers, 21kV vacuum circuit breakers (BKR #1 and BKR #2), cable segments between the main 21kV substation bus and transformers, and equivalent secondary load representations.

Traditional inductive transformer models generally look like an open circuit to very high frequency transients. Therefore, the 60 Hz transformer model can be improved by adding capacitances between windings and from the windings to ground. This type of model will act as a capacitive voltage divider to transfer a portion of the surge from the primary to the secondary windings. Bushing and winding capacitance values for the substation and customer step-down transformers were assumed based on typical data. Other substation equipment, such as circuit breakers and instrument transformers, are represented by their stray capacitances to ground. Typical stray capacitance values of substation equipment are provided in Annex B of IEEE Std. C37.011.

Figure 1 – Illustration of Oneline Diagram for Current Chopping Overvoltage Evaluation

Power system apparatus, such as transformers, switchgear, and cables may be exposed to various types of transients. IEEE Std. 1159 (Recommended Practice for Monitoring Electric Power Quality) defines the various transient power quality categories. Some of the categories include additional subcategories for a more accurate description of a particular power quality variation. High-frequency oscillatory transients have a principle frequency range of 0.5 – 5.0MHz, typical durations of 5μs, and typical voltage magnitudes of 0 – 4.0 per-unit.

High-frequency transients and very steep overvoltages may cause problems for electrical equipment because they can create local overstressing of the insulation system. Vacuum circuit breaker opening and closing operations are one of the causes of these high rate-of – rise (dv/dt) transients. Dry-type transformers and motors are often more vulnerable due to their lower insulation level (BIL) ratings.

Vacuum circuit breakers are understood to be capable of initiating a phenomena described as current chopping. The physics of the vacuum circuit breaker allow for a smaller space to be utilized in the interruption of current in a vacuum. It is well well-known that these devices can interrupt (chop) current. This is a different behavior than typical air circuit breakers, which normally allow current arcing following contact separation until a natural zero crossing occurs. Usually, the current chopping phenomenon is not troublesome. However, there are specific circuit configurations that can cause problems. The most common concern results from the use of vacuum interrupters to de-energize unloaded transformers or other highly inductive circuits. In this case, the inductive current to the transformer is interrupted, causing a transient overvoltage. The equivalent circuit for this condition is shown in Figure 2.

Figure 2 – Equivalent Circuit for Current Chopping for an Unloaded Transformer

The transient overvoltage may be approximated by evaluating the energy transfer between the inductor and the stray system capacitance in the circuit:

Energy = 1/2*L_mI_c² = 1/2*CV²

V = √(L_m/C)*I_c = Z_s*I_c

Z_s = √(L_m/C)

where:
I_c= chopping current level (A)
L_m = transformer magnetizing inductance (H)
C = stray system capacitance (transformer side of switch) (F)
R = transformer losses (Ω)
Z_s = surge impedance (Ω)
V = transient switching surge voltage (V)

The factor Z_s in the equation is termed surge impedance. The equation shows that the transient overvoltage reached is the product of the current chopped (amps) and the surge impedance (ohms). It is an interesting relationship because it is independent of the actual system operating voltage. Current chopping has the capability to cause overvoltages that are many multiples (per-unit) of the system voltage. The expressions also highlight the importance of the stray system capacitance. In other words, more capacitance results in lower transient overvoltages. Most vacuum circuit interrupters are only capable of chopping 2-10 amps of current, which means the current will arc across separating contacts until the value of instantaneous current is below that range.

Because of a vacuum’s dielectric characteristics, vacuum circuit breakers are somewhat more likely to interrupt high-frequency current components during pre-striking and current chopping events than other types of switches (e.g., oil, air, SF6). This is due to the high di/dt at the moment that the current passes through zero. Pre-striking during vacuum circuit breaker closing occurs relatively often. However, the resulting transient overvoltages are generally relatively low compared to de-energizing current chopping.

There are several mitigation alternatives for controlling the high-frequency transients and very steep overvoltages that can overstress the insulation system of the electrical equipment. The most popular protection method is MOV surge arresters connected at the terminals of transformers and switchgear. Surge arresters provide overvoltage protection; however, they may not adequately limit very high rate-of-rise (dv/dt) transient voltages. Surge arresters do not filter the high-frequency oscillations and they do not eliminate reflected waves.

In addition to surge arresters, there are several mitigation alternatives that can control the rate-of-rise of the transient voltages. This is beneficial because severe dv/dt transient voltages can damage the first few turns of insulation of dry-type transformers and motors. Additional mitigation options include surge capacitors, snubbers, and series inductances. Snubbers are R-C filter networks that include fuses, capacitors and resistors.

A high-frequency transient model of a portion of the customer facility was developed using the PSCAD simulation program. This model was used for the vacuum circuit breaker current chopping analysis. The model was based on oneline drawings and other information that is summarized in this section. The model was designed so transient voltages could be determined for various vacuum circuit breakers operating conditions. A oneline diagram of the transient model was previously shown in Figure 1. The accuracy of the simulation model at 60 Hz was determined using simulated short-circuit fault current magnitudes and other steady-state quantities.

The representation of the 60 Hz source equivalent at the service entrance under assumed normal system conditions included:

Three-phase (I_3φ) fault current: 5,500 A, X/R = 11.4 (200.1 MVA)
Single-line-to-ground (I_φG) fault current: 5,000 A, X/R = 11.4 (181.9 MVA)

Table 1 summarizes the results for the initial steady-state fault cases. The simulation model was also verified using a number of other steady-state quantities, such as bus voltages and transformer currents. The short-circuit equivalent at the service entrance bus represents a reduction of the entire utility system.

Table 1 – Steady-State 60Hz Short-Circuit Fault Comparison

The step-down transformers were modeled in PSCAD using the classical three-phase, two winding transformer model. The two transformers had winding voltage ratings of 21kV/480V and were connected delta/wye-ground. The transformer BIL ratings included 110kV for the high-side windings and 10kV for the low-side windings.

A summary of the transformer data included:

Name	Rating	Z_tx(%)	No Load Loss	Loss Load	I_e _{@ 100%V}
TX #1	2,000kVA	7.50%	4,500W	18,000W	0.75%
TX #2	2,000kVA	7.50%	4,800W	17,525W	1.10%

The nonlinear portion (saturation) of the transformer model was represented by specifying three parameters of the core saturation characteristic. The air core reactance of the transformers was assumed to be twice (2.0x) the leakage reactance (in per-unit), the knee voltage was assumed to be 1.2 per-unit, and the magnetizing currents were determined from the transformer test reports.

A traditional inductive transformer model generally looks like an open circuit to very high frequency transients. The 60 Hz transformer model can be improved by adding capacitances between windings and from the windings to ground. This type of model will act as a capacitive voltage divider to transfer a portion of the surge from the primary to the secondary windings. Capacitance values for the step-down transformers were based test report data. The capacitance values included in the transient model were C_H = 200ρF, C_L = 1,800ρF, and C_HL = 1,200ρF.

Other substation equipment, such as circuit breakers and instrument transformers, were represented by their stray capacitances to ground. Typical stray capacitance values of substation equipment are provided in Annex B of IEEE Std. C37.011. When a range of values is given, the middle value was used. Based on the facility drawings, the relevant equipment was summarized and the equipment capacitances were totaled to determined effective values on the different segments of the 21kV apparatus. The values used in the simulation model included:

Effective Capacitance (segment between source and 200’ cable): 300ρF
Effective Capacitance (other elements on main switchgear bus): 1,200ρF
Effective Capacitance (between breaker #1 and transformer #1): 500ρF
Effective Capacitance (between breaker #2 and transformer #2): 500ρF

The MOV surge arrester was modeled in PSCAD using the built-in metal oxide surge arrestor model. This component models a gap-less metal oxide surge arrester, where the user may specify the I-V characteristic. The arrestor evaluated during the study included a Hubbell DynaVar PDV-100 for the step-down transformer primary winding.

The surge arrester ratings included:

Hubbell DynaVar PDV-100 (214222) Heavy Duty:
Rated Voltage (Duty Cycle): 27 kV
Maximum Continuous Operating Voltage (MCOV): 22 kV
Maximum Energy Discharge Capability: 2.2 kJ/kV_rated MCOV
Maximum Energy Discharge Capability: 48.4 kJ
Maximum Switching Surge Protective Level (MSSPL): 58.0 kV (@ 500 A)
Maximum 0.5μs Discharge Voltage: 86.0 kV (@ 10kA)
Protective Characteristic (peak voltage – 8×20μsec discharge):
[1.5kA – 63.7kV, 3kA – 68.6kV, 5kA – 72.4kV, 10kA – 80.0kV, 20kA – 91.8kV, 40kA – 108.3kV]

The 21kV cable sections shown in Figure 1 were included in the transient model. Impedance data for the 60 Hz cables included:

Conductor: 500 kcmil
Positive sequence impedance (Z₁): 0.1340 +j0.0970 Ω/1000’
Zero sequence impedance (Z₀): 0.4420 +j0.3160 Ω/1000’
Capacitance (C₁): 84.60ρF/ft
Conductor: 1/0 awg
Positive sequence impedance (Z₁): 0.0350 +j0.0790 Ω/1000’
Zero sequence impedance (Z₀): 0.3170 +j0.2170 Ω/1000’
Capacitance (C₁): 51.35ρF/ft

A traveling wave model in PSCAD was used to represent each feeder segment for the high frequency vacuum circuit breaker switching analysis. The traveling wave model, which is based on the Bergeron method, is based on a distributed L-C parameter traveling wave line models, with lumped resistances. It represents the L and C elements of a PI section in a distributed manner. The program calculates the line constants for the cable segments before each simulation. The calculated surge impedance of the 500 kcmil conductor was approximately 55Ω and the calculated surge impedance of the 1/0 awg conductor was approximately 64Ω.

The initial simulation case (Case 1) included all of the components in Figure 1. Case 1 did not include any faults, vacuum circuit breaker operations, arresters, or other surge suppression devices. It was completed to assure that the desired steady-state voltages and power flow quantities were achieved before the current chopping cases were completed. The simulated steady-state 21kV bus voltage is shown in Figure 3. The simulation duration for Case 1 was 0.025 seconds (1.5 cycles) and the solution time step was 0.25μsec. It should be noted that a 1.0 per-unit peak line-to-ground voltage is 17.146kV (21kV*√2/√3). The step-down transformer secondary reactive load values were assumed to be 400kVAr, which results in a lightly loaded step-down transformer with approximately 10 amps rms current on the 21kV primary windings.

Figure 3 – Three-Phase 21kV Bus Voltages for Case 1

Opening a vacuum circuit breaker with no (0 amps) chopping current (ideal operation) produces relatively small transient voltages. Case 2 was completed to show the opening of the vacuum circuit breaker #1 (BRK1) with an ideal operation (no chopping current). The resulting transient voltages at the transformer 1 primary windings are shown in Figure 4.

Figure 4 – Transformer #1 Primary Voltages (Ideal Circuit Breaker Opening) for Case 2

Most vacuum circuit interrupters are only capable of chopping 2-10 amps of current, which means the current will arc across separating contacts until the value of instantaneous current is below that range. The current chopping value for the 21kV vacuum circuit breakers was assumed to be 8 amps.

Case 2 involved opening the vacuum circuit breaker #1 (BKR1) with an assumed chopping current of 8 amps. It should be noted that Case 2 did not include any arresters or other surge suppression devices. The timing for the circuit breaker contacts to open was 8.50msec.

Figure 5 shows the resulting three-phase circuit breaker #1 currents for Case 2. The figure highlights the phase currents being chopped at an 8 amp value. The current chopping produces the transient voltages at the transformer #1 primary winding that are shown in Figure 6. The magnitude of the transient overvoltage is the product of the current chopped (amps) and the surge impedance (ohms) of the circuit.

The peak transient voltage in Figure 6 was 118.601kV. This compares with a high-side transformer winding BIL rating of 110kV. The simulation case did not have any surge arresters included in the model. This case is used to show the worst-case voltages that would be present without any surge protection in-service. The time for the transient voltage to reach the 118.601kV value was approximately 110μsec.

Figure 5 – Circuit Breaker #1 Current for Case 3

Figure 6 – Transformer #1 Primary Voltages for Case 3

Case 4 involved opening circuit breaker #1 with an assumed chopping current of 8 amps. Case 4 included a Hubbell DynaVar PDV-100 (27kV Rating, 22kV MCOV) arrester (three – phase set) connected to the transformer #1 primary windings.

Figure 7 shows the resulting three-phase transformer #1 primary winding voltages for Case 4. The peak transient voltage in Figure 7 was 44.551kV. This compares with a high-side winding BIL rating of 110kV. For comparison, the peak transient voltage without the surge arrester (Case 3) was 118.601kV for the 8 amp current chopping value.

The maximum simulated arrester energy duty was 0.12 kJ, which is approximately 0.25% of the assumed rated energy capability 48.4 kJ. The time for the transient voltage to reach the 44.551kV magnitude was approximately 70μsec. Case 4 shows how the surge arrester reduces the transient voltage magnitudes at the transformer primary terminals during the current chopping event.

Case 5 investigated opening circuit breaker #1 with a 8 amp chopping current and with an R-C snubber connected at the switchgear terminals. The snubber’s resistor value was 30Ω and the capacitor value was 0.25μF. Figure 8 shows the resulting three-phase transformer #1 primary winding voltages for Case

The peak transient voltage was 47.789kV. Case 5 shows how an R-C snubber affects the transient voltages at the transformer primary terminals during the current chopping event.

Figure 7 – Transformer #1 Primary Voltages for Case 4

Figure 8 – Transformer #1 Primary Voltages for Case 5

Figure 9 shows a comparison of the transformer #1 primary (Phase B) voltages for Cases 4 and 5. Case 4 included the arrester connected to the transformer #1 primary windings, while Case 5 included the R-C snubber connected at the switchgear terminals. Figure 9 emphasizes the fact that the rate-of-rise (dv/dt) for the case with the surge arresters is more severe. The comparison of the results shows that surge arresters provide overvoltage protection, but they may not limit high rate-of-rise (dv/dt) transient voltages. Properly – designed R-C snubbers should reduce the initial rate-of-rise of the transient voltages, which may be beneficial because severe dv/dt transient voltages can damage the first few turns of insulation of dry-type transformers or excite internal transformer resonances producing severe overvoltages.

Figure 9 – Transformer #1 Primary Voltages for Case 4 and Case 5

SUMMARY

This case study summarized a customer vacuum circuit breaker (VCB) current chopping transient overvoltage evaluation. A high-frequency transient model was created to simulate the current chopping transients and resulting overvoltages and arrester energy duties. A high-frequency model was required to accurately represent the current chopping phenomena. The simulation results show that properly designed R-C snubbers will reduce the initial rate-of-rise of the transient voltages, which may be beneficial because severe dv/dt transient voltages can damage the first few turns of insulation of dry-type transformers and motors or excite internal transformer resonances producing severe overvoltages.

REFERENCES

IEEE Guide for the Application of Insulation Coordination, IEEE Std. 1313.2-1999, IEEE, October 1999, ISBN: 0-7381-1761-7.
IEEE Application Guide for Transient Recovery Voltage for AC High Voltage Circuit Breakers Rated on a Symmetrical Current Basis, IEEE Std. C37.011-1994, IEEE, ISBN: 1-55937-467-5.
Electrical Transients in Power Systems, Allan Greenwood, Wiley-Interscience; Second Edition, April 18, 1991, ISBN: 0471620580.
R.C. Dugan, M.F. McGranaghan, S. Santoso, H.W. Beaty, Electrical Power Systems Quality, McGraw-Hill Companies, Inc., November 2002, ISBN 0-07-138622-X.
“IEEE Recommended Practice for Monitoring Electric Power Quality,” IEEE Std. 1159-2009, IEEE, June 2009, ISBN: 978-0-7381-5940-9.

RELATED STANDARDS
IEEE Std. 1313.2, IEEE Std. C37.011

GLOSSARY AND ACRONYMS
BIL: Basic Impulse Level
MCOV: Maximum Continuous Operating Voltage
MOV: Metal Oxide Varistor
PSCAD: Power Systems Computer Aided Design
TVSS: Transient Voltage Surge Suppressor
VCB: Vacuum Circuit Breaker

Electric Shock Hazard Limitation in Water During Lightning Strike

Published by Jarosław WIATER, Białystok University of Technology

Abstract: Swimming during a thunderstorm is one of the most dangerous things that can be done. Lightning regularly strikes water, and since water conducts electricity, a nearby lightning strike could kill or injure human being. This paper will present simulation results of scalar potential distribution in water during lightning strike with respect to water conductivity. Lightning limitation buoy will be used for electric shock hazard reduction. All calculations results were obtained by CDEGS software.

Streszczenie. Kąpiel na otwartych akwenach w trakcie burzy jest jedną z najbardziej ryzykownych czynności, która może być wykonywana. Występujące wyładowania piorunowe w powierzchnie wody mogą spowodować obrażenia a nawet śmierć osób w niej się znajdujących. W tym artykule zaprezentowano wyniki obliczeń rozkładu potencjału w wodzie podczas bezpośredniego wyładowania piorunowego w wodę, w zależności od jej rezystywności. Zaproponowano wykorzystanie „boi piorunowej” dzięki, której uzyskano znaczące zmniejszenie zagrożenia porażenia prądem elektrycznym. (Minimalizacja zagrożenia porażeniem prądem elektrycznym w wodzie podczas wyładowań piorunowych).

Keywords: lightning, swimming, water, electric shock hazard, lightning buoy.
Słowa kluczowe: wyładowanie piorunowe, kąpiel, woda, porażenie prądem elektrycznym, boja piorunowa.

Introduction

Swimmers sometimes get struck by lightning. For example in 2005 three people were struck while swimming in the ocean near Tampa, and four more were hit in waters off Chiba Prefecture, Japan. Two of them were seriously injured [1]. Swimming pools aren’t necessarily safer too. In July 2006 a 50-year-old men was dangling his feet in the pool at a rented villa in Italy when lightning struck the water, killing him and injuring a friend [1]. Even showers and tubs are dangerous during thunderstorm because current can be transferred through plumbing.

Looking at US government data collected between 1959 and 2005, we see that incidents involving boats and water account for 13 percent of all lightning fatalities nationwide (among cases where circumstances are known), coming in behind instances where victims were out in the open (28 percent) or under a tree (17 percent). In Florida, which ranks first among the states in lightning casualties, boating and other water-related incidents make up 25 percent of lightning deaths [1, 2].

The chances that someone is going to be struck by lightning while swimming are strictly correlated with it height above water level. A lightning strike certainly can cause a high surge current to pass through water. The lightning current may spread out in all directions and dissipate within few meters or so. It is crucial to minimize this distance. Electric shock hazard bet on how close the strike will be. Distance of influence depends also on water type – salt or fresh [1].

This paper will present simulation results of scalar potential distribution in water during lightning strike with respect to water conductivity. Lightning limitation buoy will be used for electric shock hazard reduction. All calculations results were obtained by CDEGS software.

Water conductivity

Electrical resistivity is a measure of how strongly a material opposes the flow of electric current [3]. Electrical conductivity is the reciprocal quantity, and measures a material’s ability to conduct an electric current [3]. When we describe water propriety conductivity is used instead of resistivity. During the lightning strike conductivity of water is major factor which corresponds to water potential rise.

The conductivity of a solution of water is highly dependent on its concentration of dissolved salts, and other chemical species that ionize in the solution. Electrical conductivity of water samples is used as an indicator of how salt-free, ion-free, or impurity – free the sample is [3]. The purer the water corresponds to the lower the conductivity.

Conductivity measurements in water are often reported as specific conductance, relative to the conductivity of pure water at 25 °C [3]. Table 1 presents water conductivity for different type of it.

Table 1. Water resistivity and conductivity at 25 °C [3, 4]

Watering place model

This paper will consider electric shock hazard during lightning strike to water. Calculations were made for two different configurations. First one is typical case when swimmer is in the water alone. Second case is when swimmer is nearby “grounded” buoy. “Grounded” buoy is authors proposition to reduce lightning current influence on electric field in water. Scalar potential level is assumed as electric shock factor. Buoy were connected to 2m long copper wire with steel truss (1x1m) on it end. Proposed buoy could be a some kind of lightning protection system. It also protect against direct lightning strike to swimmer.

Fig. 1. Arrangement of watering place without and with grounded buoy

All calculations were performed for all described in table 1 river water types. It was assumed that water have got constant resistivity with respect to its depth. On figure 1 letter A reflect to assumed lightning strike point.

Numerical simulations were performed by MultiFields software package, which is a part of CDEGS package [5].

The numerical model includes an lightning channel (22m long), simple human body model (simple conductor with constant resistance equal 1kΩ) as well as simplified models of aboveground elements such as metallic buoy structure and simple steel truss on it end.

The computation methodology assumes frequency decomposition of the time domain current surge [5], frequency domain computations for a single harmonic unit current energization and superposition of the frequency domain computations modulated by the amplitude of the lightning current – shape 10/350μs, peak value 100kA [5].

(1) i(t) = I / η ( e^-αt – e^-βt )

where: t – time, a – reciprocal of time constant, b – reciprocal of time constant, I – peak current, η – correcting factor

Fig. 2. Arrangement of the observation points

Fig. 3. Scalar potential distribution for river water without buoy

Fig. 4. Scalar potential distribution for river water with buoy

Fig. 5. Step voltage distribution for river water with buoy

Computation results

Calculations were made for six different profiles. Distance between them were equal 20 meters. Figure 2 presents arrangement of the observations points. Calculated scalar potential were along X-Y axis on constant depth equal 10cm (average distance from human neck to heart level). Figure 3 and 4 presents scalar potential distribution along for two cases – without and with lightning protection buoy. All presented results were for one selected moment in time – t=10μs. In this specific time scalar potential reaches its maximal value. Without buoy voltage magnitude reaches up to 440kV. With lightning protection buoy voltage rise up to 146kV. It is three times lower scalar potential value with respect to case without it. According to calculation results safe distance for a human being from lightning strike point is about 30 meters (see figure 5). As a safe voltage level 985V were assumed according the IEEE Std 80-2002 (fault clearing time 0,1s and step voltage 985 V) [6,7,8,9].

Conclusion

In order to ensure the safety of people at a open area such a watering place during the lightning, it is necessary to ensure protection against fulguration. Statistical data shows that direct lighting strike causes majority deadly accidents. Many accidents happen when storm and rain don’t even started. The simulations allowed an evaluation of surface potential in open water area giving information about magnitude of crest value of scalar potential and the graphical distribution of it during lighting strike.

Proposed lighting protection buoy reduces probability of direct lightning strike. It also reduces scalar voltage level tree times. It is no difference in author opinion if buoy reach ground or not. The most important is scalar potential shaping and reduction hazard level nearby a water surface. Range of protection against direct lightning strike depends strictly to buoy height.

This work was co-funded by the European Union under the European Social Fund.

REFERENCES

[1] http://www.straightdope.com/columns/read/2263/is-lightningreally-that-dangerous-to-swimmers
[2] National Climatic Data Center, U.S. Department of Commerce
[3] http://en.wikipedia.org/wiki/Electrical_resistivity
[4] http://corrosion-doctors.org/Water-Glossary/Glossary.htm
[5] “CDEGS Current Distribution, Electromagnetic Interference, Grounding and Soil Structure Analysis” Safe Engineering Services & Technologies Ltd., Montreal Canada.
[6] IEEE Std 80-2002: IEEE Guide for Safety in AC Substation Grounding.
[7] Augustyniak L.: Surge voltage portable generator generating 1.2/50 mu s test waveshape of peak value up to 4 kV. Przeglad Elektotechniczny. V: 83, Issue: 9, pp. 37-38, 2007.
[8] Wiater J.: Remote earth localization for lightning surge condition on the high voltage substation. Przeglad Elektrotechniczny, V:86, Issue: 3, pp. 96-97, 2010.
[9] Markowska R., Sowa A., Wiater J.:Simulation measurements of lightning risk of electronic systems. Przeglad Elektrotechniczny, V:86, Issue: 3, pp. 146-149, 2010.

Authors: dr inż. Jarosław Wiater, Bialystok University of Technology, Department of Telecommunications and Electronic Equipment, ul. Wiejska 45d, 15-351 Białystok, Poland. E-mail: jaroslawwiater@we.pb.edu.pl

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097, R. 88 NR 8/2012

Utility Capacitor Switching Trips Electronic Voltage Regulator

Published by Electrotek Concepts, Inc., PQSoft Case Study: Utility Capacitor Switching Trips Electronic Voltage Regulator, Document ID: PQS0301, Date: January 10, 2003.

Abstract: The application of utility capacitor banks has long been accepted as a necessary step in the efficient design of utility power systems. Also, capacitor switching is generally considered a normal operation for a utility system and the transients associated with these operations are generally not a problem for utility equipment. These low frequency transients, however, can cause problems for low voltage power electronic-based loads.

This case presents the results of measurements associated with capacitor switching on the utility system and the resulting problems for an industrial process facility. An electronic tap switching voltage regulator was affected by the transient voltages caused by capacitor switching on the utility system.

PROBLEM STATEMENT

A semiconductor chip manufacturer was having problems with an electronic tap switching voltage regulator. The voltage regulator was used to supply “conditioned power” to sensitive electronic chip testers. Periodically, the regulator would trip, dropping the tester loads.

The monetary losses per event were calculated as follows:

5.7 production units lost x $3,228 per unit = $18,319 per event

The manufacturer replaced the internal boards of the power conditioner only to have the regulator trip again. The facility engineer stated during the initial site survey that most of the regulator trips occurred early in the morning.

DEVELOPING A MONITORING PLAN

Figure 1 – shows a one-line of the plant and the monitoring locations. The plant is supplied by a 12kV distribution feeder directly across the street from the substation. The utility has a 2100 kVAr capacitor bank at the substation.

Figure 1 – Plant Oneline and Monitoring Locations

The 480 volt bus supplying the sensitive equipment was monitored to determine if any events were originating on the utility system. The input and output of the voltage regulator were also monitored to characterize its performance.

Monitoring Results

Monitoring results revealed that a capacitor switching transient occurred every day at 6:00 am. The 2100 kVAr bank at the substation was time-switched every morning to provide voltage support on the feeder. Figure 2 show an example of a capacitor switching transient voltage (phase-to-phase) measured at the service entrance.

Figure 2 – Capacitor Switching Transient

CAPACITOR BANK SWITCHING TRIPS REGULATOR

The early morning capacitor switching transient passed through the voltage regulator input filters and arresters and caused the microprocessor control to trip itself, thereby dropping the load. Figure 3 shows the transient waveforms that were recorded on the input and output of the voltage regulator. Notice how, ½ cycle after the transient occurs, the output voltage collapses to zero, but voltage remains at the regulator input.

Figure 3 – Voltage Regulator Input and Output Monitoring Results

ELECTRONIC VOLTAGE REGULATOR

The voltage regulator was an electronic tap switching type (SCR gate driven) with a microprocessor control (shown in Figure 4). Tap switching regulators have very fast response time of approximately ½ cycle and are designed to filter input voltage variations. However, the regulators can trip when the output voltage exceeds 110% of nominal. This is generally done to protect the load from excessive overvoltage conditions.

This regulator was adversely affected by the transient caused by capacitor switching on the utility system.

Figure 4 – Schematic of an Electronic Tap Switching Voltage Regulator

SOLUTION

Power conditioning devices should never be more sensitive than the load that they are protecting. In this case, the transient was not severe enough to cause any damage to the chip testers. The regulator was the weak link.

One solution would be to replace the voltage regulator with a regulator with better filtering or to take it out completely. Regulation of this type may not be warranted.

Another solution would be to contact the manufacturer to see if the output overvoltage trip setting could be increased.

REFERENCES

G. Hensley, T. Singh, M. Samotyj, M. McGranaghan, and T. Grebe, Impact of Utility Switched Capacitors on Customer Systems Part II – Adjustable Speed Drive Concerns, IEEE Transactions PWRD, pp. 1623-1628, October, 1991.

G. Hensley, T. Singh, M. Samotyj, M. McGranaghan, and R. Zavadil, Impact of Utility Switched Capacitors on Customer Systems – Magnification at Low Voltage Capacitors, IEEE Transactions PWRD, pp. 862-868, April, 1992.

Electrotek Concepts, Inc., Evaluation of Distribution Capacitor Switching Concerns, Final Report, EPRI TR-107332, October 1997.

RELATED STANDARDS
IEEE Standard 1036
IEEE Standard 1159

GLOSSARY AND ACRONYMS
ASD: Adjustable-Speed Drive
PWM: Pulse Width Modulation
MOV: Metal Oxide Varistor
SCR: Silicon Controlled Rectifier
TVSS: Transient Voltage Surge Suppressors

Hazardous Step Voltage Nearby Pole Exited by High Voltage Surge

Published by Jarosław WIATER, Politechnika Białostocka, Wydział Elektryczny, Katedra Telekomunikacji i Aparatury Elektronicznej

Abstract. This paper presents a ground potential rise (GPR) and voltage difference measurement results. Measurements were made for a model of a pole installed in homogeneous ground. The test stand was built in high voltage lab. It allows safe high voltage testing of different types of ground systems. During measurements current and voltage surges were produced by the UCS 500M impulse generator. The prepared physical model makes it possible in the future to investigate different methods of reducing the value of touch and step voltages in the vicinity of earth systems.

Streszczenie. W artykule przestawiono wyniki pomiarów wzrostu potencjału ziemi i różnicy napięć dla modelu słupa zainstalowanego w jednorodnej ziemi. Specjalnie zbudowane stanowisko badawcze umieszczono w laboratorium wysokich napięć. Konstrukcja i lokalizacji umożliwiła prowadzenie badań różnych typów systemów uziomowych w sposób bezpieczny z wykorzystaniem generatorów prądów udarowych wysokiego napięcia. Podczas badań udary prądowo-napięciowej wytwarzał generator udarowy UCS 500M. Zbudowane stanowisko badawcze umożliwi w przyszłości badanie różnych metod ograniczania poziomów napięć krokowych i dotykowych w pobliżu elementów systemu uziomowego. Wyniki pomiarów wzrostu potencjału ziemi i różnicy napięć dla modelu słupa zainstalowanego w jednorodnej ziemi

Słowa kluczowe: wyładowanie piorunowe, napięcie krokowe, wzrost potencjału ziemi, wysokonapięciowe badania.
Keywords: lightning, step voltage, Ground Potential Rise (GPR), high voltage tests.

Introduction

The basic requirements and features of an earthing system can be summarized as follows:

• Provides personnel safety and reduces fire hazard during fault conditions by maintaining low or zero potential difference between all conductive elements of a structure.
• Provides low impedance path for lightning current to earth and improves system tolerance to electrostatic energy discharge;
• Minimizes service interruptions and equipment damage under fault conditions;
• Facilitates equipment operation i.e. signaling with earth return by ensuring low impedance ground reference;
• Reduces radiated and conducted electromagnetic emissions and susceptibility of equipment.

Personnel safety in various objects under power fault conditions has been studied extensively [1] and is well defined in international standards [2]. A lot of technical publications related to transient lightning behavior of various ground grids are also available [3, 4, 5, 6]. However, they usually consider only an overall scalar potential distribution or Ground Potential Rise (GPR). Still not much information concerning the actual values of step and touch voltages that people can be exposed to during lightning strokes is provided [7]. Moreover, the analyses often relate to a single frequency (usually a low frequency), which does not give complete in-formation because of strong dependence of the behavior of ground grids on frequency. Furthermore, aboveground structures are also often neglected. These structures however, are very important be-cause on the one hand, a current distribution in aboveground structure is determined by the location and specific behavior of earth electrodes and on the other hand, a current distribution in buried electrodes depends on the geometry of the aerial part of the structure.

The main purpose of the article is to present the results of potential measurements in the closest vicinity of the pole to the earth system. Conducting measurements using high voltage generators is very risky in real conditions. Hence, the physical model of the pole in homogeneous soil was built in the laboratory. Such a test stand will allow for safe conduct of measurements using high-voltage surge generators.

Ground Potential Rise, Step And Touch Voltages

Dissipation of the lightning current into the earth means that a good electrical connection to earth at zero potential reference i.e. remote ground should be provided. The impedance of this connection is not ideal due to the soil resistivity within which the grounding system is buried. Hence, the lightning current that flows through the earthing network to earth results in the local ground potential rise (GPR) with respect to remote ground. The GPR is a source of potential gradients within and around the earthing network area, which determine the values of step and touch voltages. An illustration of GPR, step and touch voltages is presented in figure 1.

The step voltage is defined as the potential difference between one’s outstretched feet, usually 1 m apart. The touch voltage is the potential difference between one’s outstretched hand touching an earthed structure and one’s feet. The maximum hand-reached distance of 1 m is usually assumed.

The figure presents also some special cases of touch voltages. The worst case of the touch voltage called a mesh voltage is defined as a potential difference between the centre of a given mesh and an earthed structure. The potential transferred for some distance via reference metallic conductor produces the transferred voltage.

Fig. 1. Illustration of GPR, step and touch voltages [9]

Research Stand And Measurement Results

Determining the level of shocks is essential for the safety of people near structures that can be struck by lightning. In order to be able to carry out safe measurements in the conditions closest to reality, a metal bathtub of 2.5 m wide, 2.5 m long and 0.5 m high was built (fig. 2). The metal bath was filled with a quartz decorative sand with a resistivity of 1500 Ωm (fig. 3). The sand was sifted and the grains were no larger than 1 mm. The voltage-current surge was brought to the top of the column structure. The column itself is made of metal truss 40x40x30 cm (width, depth, height). To the corners of the metal truss, a metal structure imitating the base earth was connected (fig. 4). The built-in structure is symmetrical.

Voltage-current surges were produced by the high-voltage impulse generator – UCS 500M6B. The UCS 500M6B cover transient and power fail requirement according to international standards with voltage capability of up to 6,6kV.

• voltage (open circuit) 250-6600V,
• pulse front time 1,2μs +/- 30%,
• pulse time to half value 50μs +/- 20%,
• current (short circuit) 125-3300A,
• direct output Via HV-coaxial connector, Z_i=2Ω.

Fig. 2. Research stand made for safe high voltage tests

To measure output current of the generator TCP0150 Tektronix AC/DC current probe was used (DC to 20 MHz bandwidth, 500 A peak pulse current). Voltage and current waveforms were registered by Tektronix DPO 7254 digital oscilloscope.

Arrangements during step voltage measurements presents figure 4 and 5. The generator output terminal was connected to the column support structure. The generator ground terminal was connected to a metal tub. Additional electrodes for measuring the voltage distribution near the pole are pushed every 5 cm. The column is centrally located in a metal tub.

Fig. 4. Physical pole model designed to bury in homogeneous ground.

Fig. 5. Arrangement of step voltage measurements – front view.

Fig. 6. Arrangement of step voltage measurements – top view.

The depth of foundation of the additional voltage electrodes was set at 8 cm. For the purpose of analysis, the human foot is usually represented as a conducting metallic disc and the contact resistance of shoes, socks, etc., is neglected [4]. Traditionally, the metallic disc representing the foot is taken as a circular plate with a radius of 0,08 m. A value of 1000Ω were used as a resistance of a human body from one foot to the other foot [10]. During the measurements voltage electrodes dug on 0,08 m depth represents human foot.

Measurement results are shown in figures 7-10. Figure 7 shows generator output current for different capacitor charging voltages. Figure 8 and 9 shows the step and touch voltage waveforms depending on the capacitor charging voltage for point 1. The step voltage changes according to the distance from the pole are shown in figure 10. The results of the measurements are summarized in table 1 and 2.

Table 1. Step and touch voltage results

Fig. 7. High voltage surge generator output current for 2,3,4,5,6 kV

Fig. 8. Step voltage in point 1 for 2,3,4,5,6 kV charging voltage level.

Fig. 9. Touch voltage in point 1 for 2,3,4,5,6 kV charging voltage level.

Fig. 10. Step voltage in point 1,2,3,4,5 for 6 kV charging voltage level.

Table 2. Step voltage results depending on the distance

Discussion

The measurements performed show a linear relationship between the generator’s charge voltage and the value of the step and touch voltages. Debatable issue from the point of view of the threat of living beings is the value of the spacing between the legs when measuring step and touch voltage. From the anatomical point of view, the distance of 1 meter between the legs is only possible during a quick walk. It is difficult to imagine a situation when someone is going quickly and simultaneously touching an element that may potentially be energized during an earth lightning strike. The author’s experience and press reports clearly indicate that an electric shock caused by a lightning strike occurs when the victim is standing under an object such as a tree. According to the author, in case of electric shock caused by lightning discharge, a new stand voltage definition should be introduced. Stand voltage it is potential difference on the ground surface at anatomical distance between the legs equal 5 cm.

Conclusions

Generally lightning strikes have got a sudden and unexpected nature. In present time when mass-media delivers information about electric shock caused by lightning strike without any delay de-tail knowledge about lighting safety is necessary. This also creates life fear in society. Unfortunately very often happens that those information’s are not true, not complete and not compliant with science knowledge. Medical aspects of news are simple. Victim survived or not. Unreliable information’s causes periodically and rapid increase attraction about lightning electric shock hazard. Most of questions concentrates on one subject. How to correctly behave during thunderstorm. This problem is not so easy as it seems to be. In our climate conditions there are up to several current strokes during cloud-to-earth lightning discharge [1,2,3]. It is difficult to estimate the actual number of consecutive components of lightning discharge. It is, however, possible to establish in safe way possible values of step, touch, stand voltages during lightning surge current excitation.

Electrical accidents caused by electric shocks occur during work, leisure and day-to-day operations. They always involve certain economic, human and social losses as well as the appearance of fear.

In order to reduce losses, it is necessary to use appropriate technical solutions supported by appropriate normative acts, which will reduce the number of catastrophic events and, in most cases, limit their effects.

Preventive measures should be based on the development and implementation of such technical solutions, which, in extreme situations, protect people from the effects of lightning strikes, irrespective of where they are located. One cannot forget about indirect prevention, which should include, first of all, various information measures promoting the use of lightning-resistant technical solutions. The carelessness and misbehavior of man largely leads to accidental injuries, so it is appropriate to educate young people from an early age. Another important but also important aspect is the minimization of the effects of electric shocks by spreading the rules of first aid at the scene of an accident. There is also a clear progress in the lightning protection of building constructions, resulting in changes in the approach to many safety issues during lightning discharges. It is therefore important to consider the introduction of legal regulations recommending participation in periodic lightning protection training for designers and lightning protection workers on new buildings: both public utilities, residential homes and industrial facilities.

The presented results of the measurements clearly indicate a high level of danger of step voltages caused by lightning discharges. The number of traumas causing injuries in people is growing, so it is important to continue research in this area.

Acknowledgment: The research was conducted within the project S/WE/1/2015, financially supported by Polish Ministry of Science and Higher Education.

REFERENCES

[1] J. B. M. van Waes, A. P. J. van Deuersen, M. J. M. van Riet, F. Provoost; Safety Aspects of GSM Systems on High-Voltage Towers: An Experimental Analysis; IEEE Transactions on Power Delivery, vol. 17, no. 2, April 2002; pp. 550–554.
[2] IEEE Std 80-2000: IEEE Guide for Safety in AC Substation Grounding.
[3] Grcev L. D.; Computer Analysis of Transient Voltages in Large Grounding Systems; IEEE Transactions on Power Delivery, vol. 11, no. 2, pp. 815–823, April 1996.
[4] Geri A.; Practical Design Criteria of Grounding Systems under Surge Conditions; 25th International Conference on Lightning Protection; Rhodes, Greece, 2000; Proc. 5.18.
[5] Lorenzou M. I., Hatziargyriou N. D.; Effective Dimensioning of Extended Grounding Systems for Lightning Protection; 25th International Conference on Lightning Protection; Rhodes, Greece, 2000; Proc. 5.9.
[6] Ma J., Dawalibi F. P.; Analysis of Grounding Systems in Soils with Cylindrical Soil Volumes; IEEE Transactions on Power Delivery, vol. 15, no. 3, July 2000; pp. 913–918.
[7] Ala G., Di Silvestre M. L.; A Simulation Model for Electromagnetic Transients in Lightning Protection Systems; IEEE Transactions on Electromagnetic Compatibility, vol. 44, no. 4, November 2002.
[8] Markowska R.; Rozkłady napięć na terenie stacji elektroenergetycznych przy przepływie prądów piorunowych w systemach uziomów; Urządzenia piorunochronne w projektowaniu i budowie; Kraków 26–27 October 2000, pp. 115–122.
[9] AC substation earthing tutorial–ERA Technology Ltd. [10] Electricity Association Technical Specification 41-24: Guidelines for the Design, Installation, Testing and Maintenance of Main Earthing Systems in Substations.

Auhtor: dr inż. Jarosław Wiater, Białystok Technical University, Department of Telecommunications and Electronic Equipment, ul. Wiejska 45d, 15-351 Białystok, Poland E-mail: jaroslawwiater@we.pb.edu.pl

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 93 NR 12/2017

Utility Capacitor Switching Fails VAX Disk Drive

Published by Electrotek Concepts, Inc., PQSoft Case Study: Utility Capacitor Switching Fails VAX Disk Drive, Document ID: PQS0408, Date: September 30, 2004.

This case illustrates a situation where a power conditioning device was the weak link in an overall equipment protection scheme. The power conditioner, which was located near the sensitive equipment, was magnifying utility capacitor switching transients that were not very severe in magnitude.

PROBLEM STATEMENT

A data processing company had a critical VAX (a computer-family of Digital Equipment Corporation) computer that had a disk drive failure about once a month. All data not backed up was lost, and the downtime associated with each failure was several hours.

The computer was supplied by a low impedance power conditioner (LIPC) that was designed to filter high frequency transients and to make a local neutral-to-ground bond. These types of power conditioners are specifically designed to interface with electronic equipment, especially computers.

DEVELOPING MONITORING PLAN

Disturbance analyzers were brought in to monitor facility power quality. One monitor was installed at the service entrance 480 volt bus supplying the sensitive equipment. Another monitor was installed on the input and output of the low impedance power conditioner supplying the VAX computer to characterize its performance.

Initial monitoring results revealed that a capacitor switching transient occurred every morning at 8:00 am An example of this transient voltage is shown in Figure 1.

Figure 1 – Capacitor Switching Transient Recorded at the Service Entrance

TRANSIENT OVERVOLTAGE MAGNIFIED BY LOW IMPEDANCE POWER CONDITIONER

The monitor on the input and output of the low impedance power conditioner recorded some interesting waveforms as illustrated in Figure 2. A disturbance was triggered on the input, but as the waveform shows, it was not severe enough to cause any problems. However, the disturbance that was recorded on the output of the power conditioner shows that the transient voltage was magnified considerably.

LOW IMPEDANCE POWER CONDITIONER

Low impedance power conditioners are used primarily to interface with the switch-mode power supplies (SMPSs) commonly found in power-electronic equipment. These power conditioners have lower impedance than isolation transformers, and a filter as part of their design (shown in Figure 3). The filter is on the output side and protects against high-frequency transients. However, low-to-medium frequency transients (including utility capacitor switching transients) have been know to cause problems for these devices.

The VAX disk drive never seemed to fail as a direct result of the magnified transient. However, since the disk drives on older VAX machines such as this one are connected directly across the line with no internal protection, it was felt that over time this daily transient caused the disk drive to fail approximately once per month.

SOLUTION

Power conditioning devices should not be the weak link in the overall equipment protection scheme. In this case, the power conditioner was magnifying a transient overvoltage, which was not very severe, near the location of the sensitive equipment.

Replacing the low impedance power conditioner with a standard isolation transformer provided enough impedance to sufficiently reduce the transient overvoltage while maintaining a neutral-to-ground bond.

REFERENCES

Electrotek Concepts, Inc., Evaluation of Distribution Capacitor Switching Concerns, Final Report, EPRI TR-107332, October 1997.

RELATED STANDARDS
IEEE Std. 1036
IEEE Std. 1159

GLOSSARY AND ACRONYMS
ASD: Adjustable-Speed Drive
LIPC: Low Impedance Power Conditioner
MOV: Metal Oxide Varistor
PWM: Pulse Width Modulation
SMPS: Switch Mode Power Supply
TVSS: Transient Voltage Surge Suppressors
VAX: Virtual Address eXtension

Harmonic Current Impact on Transient Overvoltages during Filter Switching-Off

Published by Yuriy VARETSKY¹, Roman PAVLYSHYN², Michał GAJDZICA¹,
AGH University of Science and Technology (1), Lviv Polytechnic National University(2)

Abstract. This paper focuses on the features of power filter switching-off in industrial power supply systems. The filter switching-off behavior under a large amount of a harmonic current component has been analyzed. The effect of harmonic current content in interrupting current, filter order tuning and switching condition are considered in the analysis. Finally, several oscillograms of simulated cases are included to show main points of the investigation. (Harmonic current impact on transient overvoltages during filter switching-off)

Streszczenie. W artykule przedstawia sie charakterystyczne cechy wyłączenia filtrów mocy w sieciach zasilających zakłady przemysłowe. Przeprowadzono analizę przebiegu wyłączenia przy dużym udziale harmonicznych w prądach filtrów. W badaniach uwzględniono wpływ wartości prądów harmonicznych w przerywanym prądzie, rzędu strojenia filtru oraz warunków wyłączenia. Przedstawiono niektóre przebiegi symulacyjne charakteryzujące problem. (Wpływ harmonicznych prądu na generowanie przepięć w chwili wyłączania filtrów)

Słowa kluczowe: system zasilający, wyłącznik, filtr harmonicznych, przepięcie przejściowe, modelowanie.
Keywords: power supply system, circuit breaker, harmonic filter, transient overvoltage, simulation.

Introduction

Rapid growth of nonlinear loads such as static power converters, welders, arc furnaces, voltage controllers and frequency converters has led to many harmonic problems in power systems. So their solutions have become a major concern for present day engineers. The harmonic filtering is one of the solutions to prevent the troublesome harmonics from entering the supply system. Single tuned filter is the most commonly used filter. It supplies some or all of the fundamental frequency reactive power required for power factor correction. The filter components may be tuned to provide a low impedance shunt path to a specific frequency. The quality factor of the inductor determines the sharpness of tuning.

In many applications the filters are often switched (on a daily basis or performance requirements basis). As a consequence, the filter circuit switching causes transient overvoltage across the individual components of the filter that exceed the voltage at the bus.

Transient oscillation between the lightly damped filters can likewise be much higher then expected. Traditional surge arresters connected to the bus, may be inadequate overvoltage protection because of the filter components are subjected to extra stresses. Certain switching operation can also present potentially hazardous overvoltage conditions, not only to the filters, but to other substation equipment such as circuit breakers and transformers. As known from experience [1, 2, 3], switching transients in some industrial power systems inclusive filter circuits can result in damages of its components and circuit breakers. It has been noted in [3] that transient overvoltages caused by system faults or normal switching operations are well documented and accounted for in the design of adequate surge protection devices, but application of SVCs may increase the potential for excessive overvoltages. As it has been registered from field tests and simulation [4], the greater a harmonic content in the switching-off filter current, the higher residual filter voltage after interrupting will be.

The paper presents studies on switching-off the filter containing large portion of a harmonic current. The study has been carried out with typical arc furnace power supply system containing a set of single tuned harmonic filters. The investigated power supply system example includes many common features of other industrial power systems. Electromagnetic Transient Program [5] was used to simulating transient behaviors under filter circuit switching-off.

The studies have shown that harmonic content in the filter circuit current causes to grow recovery voltage between the circuit breaker contacts and residual voltage at the filter circuit. Increase in harmonic content of interrupting current tends to the higher overvoltage magnitudes and effects the possibility of circuit breaker restrike.

The authors are hoping that the results of these studies will be useful for harmonic filters application planning and improve its operation reliability.

Examined power supply system

The examined power supply system shown in the Fig.1 involves 220 kV bus supplying 35 kV bus by means of step down wye-delta connected transformer TS of 160 MVA with the primary neutral solidly grounded. Couple of 50 MVA electric arc furnace (EAF) units is connected to the 35 kV bus. The SVC circuit is assembled from four single-tuned filters and thyristor controlled reactor unit and connected to the bus through the appropriate circuit breakers. The individual filters are sized to supply 25, 30, 17 and 20 MVAr for the 2^nd, 3^rd, 4^th, and 5^th harmonic filters respectively. The filters are connected to the 35 kV bus through cables by air blast circuit breakers Q₂ – Q₄, which have become commonly used to switch filter circuits and capacitor banks. Damping circuits are connected to the switched cable ends for limiting fast transient overvoltages.

Fig. 1. Single-phase diagram of the EAF supply system

As it have been shown in [6], switching-on ungrounded wye filters in the investigated power system results in transient overvoltage magnitudes approaching 1.5…1.7 p.u. on the substation bus.

In general, however, the overvoltages, associated with normal filters energizing in the presented system, are do not dangerous for the filter equipment and do not usually endanger substation equipment at the bus location. The peak currents in the filters are a few times higher than steady-state levels. The energizing of a filter generates steep fronted voltage waves on filter reactor which can result in high local overvoltages along reactor winding length. As a consequence of this phenomenon the adequate measures to prevent the reactor insulation dielectric failure must be provided.

If under filter switching-off there is successful interruption of the capacitor current at zero crossing and the switching device withstands the transient recovery voltage there are no significant transients on clearing a filter. The occurrence of reignitions during circuit breaker poles opening tends to cause adverse transient recovery voltage conditions.

To investigate switching-off transient overvoltages in the presented scheme the equivalent circuit shown in Fig.2 was constructed for the Electromagnetic Transient Program. The every source voltage u_A, u_B, u_C was adjusted as a set of the system frequency and a harmonic frequency voltages to modeling required harmonic content in interrupting current. Circuit breaker Q in the equivalent circuit was modeled by typical air blast breaker voltage-second characteristic for switching-off load currents. Current distorted wave for each phase was interrupted by crossing zero, so after first interrupted phase current interruption in second and third phase occurs simultaneously in the presented system.

For example, Fig.3 shows voltages and current when circuit breaker interrupts the 3^rd filter currents (without harmonic presence) following one and two restrikings. As it was noted, restriking of switching breaker under interrupting currents produces sufficiently higher overvoltage magnitudes in comparison with no breaker restriking. The transient voltages include oscillation in accordance with system natural frequency. Application of damping circuit C_D , R_D allows to limit magnitude and rate of rise of the transient recovery voltage across the opening breaker contacts.

Fig. 2. Equivalent circuit of the supply system: u_A, u_B, u_C – distorted voltage waves; R_S , L_S – power supply system; F , C_F , L_F – filter circuit; R_D , C_D – damping circuit; C_ll, C_lg – cable capacitances; C_B – equivalent bus capacitance; C_S – equivalent stray capacitance.

As it can be observed from the oscillograms, current reigniting between circuit breaker contacts will produce transient voltages and currents significantly high in magnitude than those occurring during closing. Since restrikes can occur when there is a charge remaining on the filter capacitor bank it is possible for restrikes to generate transient overvoltages that are much higher in magnitude than on closing. The transient voltages on a filter and recovery voltages across a switching device can be reduced during restrikes by installing arresters on the filter side of the switching device.

Fig. 3. Voltages and current for phase A (interrupted as the first) under one (a) and two (b) 3rd filter breaker restrikings

Arresters connected phase-to-ground will limit the recovery voltage but do not necessarily limit the voltage trapped on filter capacitors during restrikes. Arresters are sometimes connected from phase to neutral to limit the trapped voltages to lower levels, thus reducing switch recovery voltage and minimizing the possibility of multiply restrikes [8].

Harmonics affect overvoltages

If the circuit breaker is applied to filter circuit loaded a large amount of a harmonic, the behavior of the transient voltages will be different from described above. As it has been noted above, harmonic content in the interrupting current will influence on the transient behavior. Such a case can be observed in the power supply system described in the Fig.1. EAF generate significant harmonic currents that flow in SVC filter circuits. The harmonic currents vary randomly during EAF operating cycle. Magnitudes of individual harmonic currents may reach sufficient values. During EAF operating it is possible that a specific filter branch may be out of service. In the circumstances the filters in service will be overloaded due to possible resonances in supply system. It provokes activation of the filter overload protection and switching-off the filter.

Energizing arc furnace transformer in EAF supply system is a powerful harmonic disturbance. In the examined supply system arc furnace transformer (T) energizing occurs several times a day. When the transformer is energized, inrush current can be high in magnitude. The transformer inrush current consisting high harmonic content can be long duration (lasting several seconds). The harmonic content causes resonance in the filter circuit that extends the duration of the inrush transient and resonance. The resonance in the filter may cause the filter relay undesirable operation. So, the filter circuit breaker will be switched-off under high harmonic presence, increasing possibility of overvoltages.

As example let us consider for comparison the oscillograms shown in the Fig.4.

Fig. 4. Voltages for 2nd filter breaker switching-off without restrikes: a – no second harmonic presence; b – second harmonic is equal to fundamental.

If the interrupted current contains harmonic component, the maximum recovery voltage between filter circuit breaker contacts increases giving rise to the possibility of restriking. When compared with the interruption of the current having no harmonic component the presence of harmonic current will also result in greater overvoltage magnitudes. Fig. 5 shows residual voltages on 2^nd filter phases versus harmonic content in interrupted currents (phase A is first interrupted). Base voltage is crest value of nominal phase-to-ground voltage.

Fig. 5. Residual voltages on 2^nd filter phases vs 2^nd harmonic content.

Fig. 6. Voltages for 2^nd filter breaker switching off during restrike: a – no second harmonic presence; b – second harmonic is equal to fundamental.

A close examination was conducted on the overvoltage magnitudes during reigniting circuit breaker current by Electromagnetic Transient Program simulating. The overvoltage magnitude depends on the order of the switching filter and harmonic current phase shift. As it has been observed from the experiments the most dangerous rise of the overvoltage magnitude versus harmonic takes place for the 2^nd filter.

Let us consider transient behavior under restriking between a filter circuit breaker contacts. Fig.6 shows transient voltages during 2^nd filter circuit breaker current reigniting.

The next investigation have been carried out is examination of the conditions where the maximum overvoltage magnitudes occur. If restriking occurs at the maximum recovery voltage, the great peaks are produced in other phases. In the certain circumstances the peak transient overvoltages may exceed withstand impulse voltage for substation insulation. The maximum values of transient voltage magnitudes on filters under restriking are shown in Table 1.

Table 1. Maximum filter overvoltages under restriking

**Note:** Base is the crest value of phase-to-ground voltage

As it can be observed from the Table 1 the restriking surges depend on the order of the filter and the value of harmonic current. When compared with the switching-off filter having no harmonic component it has been noted that presence of harmonic results not only higher recovery voltage in the first interrupted phase but also higher residual voltages of the second and third interrupted phases. Since overvoltage magnitudes under circuit breaker restriking is great it is necessary to evaluate its relationship to withstand impulse voltages. Therefore, if the restriking is observed for the filter circuit breaker, it should be necessary to evaluate possibility installing the surge protective devices.

A special case of transients which must be considered in the examined power supply system is analysis of the phenomenon under arc furnace transformer energizing. As it noted above inrush currents contain full range of harmonics beside their fundamental and dc components. Furthermore the inrush phenomenon can last many cycles and activate overvoltage and overload relays of SVC filter circuit causing trip a filter breaker under inrush condition. So, protection coordination studies should be implemented to remain SVC in service over the range of normal and transient conditions.

Conclusions

The paper presents the results of filter breaker switching-off phenomenon at an example of industrial power supply system consisting on EAF and SVC.

A general analysis shows that presence of harmonic content in the interrupting filter current increases both recovery voltages across the circuit breaker contacts and filter residual voltage. The greater harmonic content, the higher the voltage magnitudes, the more possibility of arc restrikes between the circuit breaker contacts.

Special considerations must be carried out for arc furnace power supply systems because of possible presence of great harmonic magnitudes in filter circuits and necessity of prevention of the filter protection misoperation.

Acknowledgments
The present work was supported by the Polish Ministry of Science (Grant AGH No. 11.11.210.198)

REFERENCES

[1] Bonner J.A. et al.: Selecting ratings for capacitors and reactors in applications involving multiple single-tuned filters. IEEE Trans. on Power Delivery, vol.10, no.1, 1995, pp. 547-555.
[2] Harder T.E. AC filter arrester application. IEEE Trans. on Power Delivery, vol.11, no.3, 1996, pp. 1355-1360.
[3] A Working Group of the Substation Protection Subcommittee of the IEEE Power System Relaying Committee: Static VAR compensator protection. IEEE Trans. on Power Delivery, vol.10, no.3, 1995, pp. 1224-1233.
[4] Nishikawa H., Yokokura K., Masuda S. et al. Harmonic current interruption phenomena in arc furnace filter circuits: IEEE Trans. on PAS, vol.103, no.10, 1984, pp. 3000-3006.
[5] Ravlyk A., Gretchyn T. Digital complex for modelling transient processes in electric circuits. Proc. of III Int. Symp. “Metody matematyczne w elektroenergetyce”, Zakopane, 1993, pp. 17-20.
[6] Varetsky Y. Exploitative characteristics of SVC filter circuits, Proc. of 6-th Int. Conf. Electrical power quality and utilization, Cracow, 2001, pp. 297-302.
[7] Varetsky Y. Transient overvoltages during filter circuit switching-off. Proc. of Int. Conf. on Modern Electric Power Systems, Wrocław, 2010, pp. 1-4.
[8] Working Group 3.4.17 of the IEEE Surge Protective Devices Committee: Impact of shunt capacitor banks on substation surge environment and surge arrester applications. IEEE Trans. on Power Delivery, vol.11, no.4, 1996, pp. 1798-1807.

Authors: prof. dr hab. inż. Yuriy VARETSKY, AGH – University of Science and Technology, Faculty of Energy and Fuels, Department of Fundamental Research in Energy in Energy Engineering, 30 Mickiewicza Ave,30-059 Krakow, Poland, E-mail: jwarecki@agh.edu.pl; mgr inż. Roman PAVLYSHYN, Lviv Polytechnic National University, Institute of Energy Engineering and Control Systems, Department of Power Systems and Grids, E-mail: pavlyshyn@gmail.com; mgr inż. Michał GAJDZICA, AGH – University of Science and Technology, Faculty of Energy and Fuels, Department of Fundamental Research in Energy in Energy Engineering,30 Mickiewicza Ave,30-059 Krakow, Poland, E-mail: michal.gajdzica@wp.pl.

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 89 NR 4/2013

An Analysis of the Inverter Overvoltage Generated by the Motor

Published by Algirdas BASKYS¹, ², Vytautas BLEIZGYS¹, ², Tadas LIPINSKIS¹, ²
Center for Physical Sciences and Technology (1), Vilnius Gediminas Technical University (2)

Abstract. The overvoltage in the inverter that supplies the AC induction motor, which during the deceleration operates as a generator delivering current back into the inverter DC bus, has been investigated. The investigation was performed experimentally using a special test bench. The impact of the motor deceleration rate, motor load and initial rotation velocity, at which the deceleration starts, on the overvoltage was investigated and analyzed. The obtained results were employed for the development of the overvoltage fault protection of the inverter.

Streszczenie. W przekształtniku zasilającym silniki AC powstają przepięcia w czasie zwalniania silnika. Przeprowadzona badania tego zjawiska dla różnych: prędkości zwalniania, obciążeń silnika i prędkości obrotowej (Analiza przepięć w przekształtniku zasilającym silnik)

Keywords: inverter, overvoltage, AC induction motor, motor deceleration.
Słowa kluczowe: przekształtnik, przepięcia, sinlik AC.

Introduction

The AC induction motor used in the variable speed drive based on the frequency converter can act as a generator under certain operating conditions. The inverter, which is the main block of the frequency converter, supplies the motor with the variable frequency variable amplitude three phase AC voltage. The motor rotation velocity is determined by the AC voltage frequency. If the frequency increases the motor accelerates, if it decreases – the motor decelerates. If the AC induction motor rotation velocity during the deceleration exceeds the synchronous velocity, it starts to operate as a generator delivering current back into the DC bus of the inverter through the transistors, which operate as switches of the inverter. Therefore, the capacitors of the DC bus are charged and voltage of the DC bus increases [1-3]. The maximal voltage value (overvoltage) depends on the motor deceleration rate, motor load and its inertness, capacitance of the DC bus capacitors and initial rotation velocity of the motor (rotation velocity, at which the deceleration starts). If the overvoltage of the DC bus exceeds the safe operation limits, the transistors of inverter switches, DC bus capacitors and other components used in the inverter can be damaged [3, 4]. Therefore, the problem of the overvoltage in the inverter, is topical [5]. There are lot of works, e.g. [6-9], dedicated to the investigation of the overvoltage in the inverter using simulation. However, during the frequency converter development process it is important to have accurate data, which can be obtained only experimentally. They are needed for the development of the overvoltage fault protection of the inverter.

The investigation technique

The investigation of the overvoltage in the inverter caused by the AC induction motor deceleration was performed using a special test bench. The block diagram and picture of the test bench are given in Figs.1 and 2. It includes the 4 kW AC induction motor fed from the inverter of the experimental example of the developed frequency converter. The motor drives the 5.5 kW DC generator, which acts as the motor load and is characterized by the relatively high inertness. The test bench includes the motor load torque and rotation velocity sensors and appropriate circuits for conversion of sensor signals to standard signals, which are used for measurement.

Fig. 1. The block diagram of the inverter overvoltage investigation test bench

The motor load torque is controlled by the variation of the DC generator rotor current and the generator electrical load. The transients of the inverter DC bus voltage (U_DC), motor load torque (M) and motor rotation velocity (V_r) have been investigated. The measurements of transients were provided using the Tektronix digital oscilloscope TPS2024.

The following investigation technique has been used. Firstly, the appropriate frequency (f_p) of the inverter output voltage and deceleration rate are preset and the motor is started. Secondly, the motor load torque is fixed by variation of the rotor current and electrical load of the DC generator. After this, the motor is stopped at the assigned deceleration rate and U_DC, M, V_r transients are measured. The motor deceleration rate (D) is expressed by the decrease rate of the frequency of the inverter output voltage, i.e. D has the Hz/s dimension.

The two situations can be observed during the motor deceleration when the motor acts as a generator and the DC bus capacitors are charged. The first situation is when the DC bus voltage spike does not reach the overvoltage fault protection trigger level, the second one – when the voltage spike reaches this level. In the first case the motor is stopped at the preset deceleration rate. However, if the overvoltage fault protection is triggered, the transistors of inverter switches are closed and the motor does not provide the energy to the DC bus. Therefore, the rise of the DC bus voltage is stopped, the motor is not decelerated by the inverter and, as a consequence, the motor deceleration becomes uncontrolled.

The examples of the transients of the U_DC, M and V_r for the case when the overvoltage fault protection is not triggered are presented in Fig. 3. It is seen that the motor deceleration causes the DC bus voltage spike and the motor load torque decreases and becomes negative during the deceleration. Additionally, the transient of the motor load torque has oscillations.

The DC bus voltage transient example for the case when the voltage spike reaches the overvoltage fault protection trigger level is given in Fig. 4. The rising edge steepness of the voltage spike in the analyzed case is about 1.5 V/ms. When the voltage reaches the fault protection trigger level, the transistors of inverter switches are closed by the protection circuit. Since the steepness of voltage spike is relatively low, the DC bus voltage practically is fixed at the value, which corresponds to the overvoltage fault protection trigger level even in the case when overvoltage fault protection circuit with the response time up to several hundreds of microseconds is employed. This fact allows us to use slow overvoltage protection circuit, i.e. circuit, witch has low sensitivity to electromagnetic disturbances produced by the inverter.

Fig.3. The *U_DC* (upper curve, it has been measured using the 1/3 voltage divider, therefore 1div ~ 100V), *V_r* (middle curve, 1div ~ 1000rpm) and M (bottom curve, 1div ~ 6Nm) transients caused by the motor deceleration at D=16.7Hz/s and capacitance of the DC bus capacitor C=470µF. *U_DCn*≈ 540V

The shape of the falling edge of the voltage spike (Fig. 4) is determined by the slow discharge of the DC bus capacitors by the frequency converter circuitry, which is fed from the DC bus voltage.

The inverter overvoltage investigation results

The investigation was accomplished at various motor load torques for the different motor deceleration rate and initial motor rotation velocity, at which the deceleration starts. The results were obtained for DC bus capacitor capacitances C = 470 and 880 µF. During the investigation the duration of U_DC spike (τ) and the maximal U_DC value U_DCm (overvoltage) (Fig. 3) were estimated. The results are presented in Figs. 5–7. The overvoltage depends on the motor load – it decreases if the motor load increases. This can be explained by the fact that the motor even during the deceleration supplies the energy to the load and only the excess energy of the motor is supplied to the DC bus. It is seen that the overvoltage increases when the motor deceleration rate and initial rotation velocity (initial inverter output voltage frequency) increase (Figs. 5 and 6). The increment of the capacitance of the DC bus capacitors allows us to reduce the overvoltage. However, the overvoltage decrement is slight even imperceptible (compare the dependences given in Fig. 5 with the corresponding dependences presented in Fig. 6). This can be explained by the fact that the capacitor voltage is proportional to the square root of energy (E) used for the capacitor charge. Knowing the nominal DC bus voltage U_DCn (in the analyzed case U_DCn≈540V), the DC bus capacitance C and the amount of energy, which is supplied by the motor to the DC bus during the motor deceleration, the DC bus voltage can be calculated using a well known equation U_DC=[(2E/C) + U²_DCn]^1/2, where E is expressed in Joules, C – in Farads and the voltage − in Volts.

Fig.5. The dependences of the inverter DC bus overvoltage on the deceleration rate of the motor for various motor load torques and motor initial rotation velocities (initial inverter output voltage frequencies) for the case when C=470 µF. Initial inverter output voltage frequency *f_p*=50 Hz corresponds to *V_r* ≈2800rpm, *f_p*=40 Hz – to *V_r* ≈2200rpm, *f_p*=30 Hz – to *V_r* ≈1700rpm and *f_p*=20 Hz – to *V_r* ≈1100rpm)

Fig.6. The dependences of the inverter DC bus overvoltage on the deceleration rate of the motor for various motor load torques and motor initial rotation velocities (initial inverter output voltage frequencies) for the case when C=880 µF

For example, if U_DCn = 540V and the energy of 27.7 Joules is supplied to the 470µF DC bus capacitor, it is charged according to the given equation up to U_DC = 650V. If the capacitance is increased up to 880 µF (by the 87%), the calculated voltage at the same amount of energy U_DC = 595V, i.e. theoretically the voltage in the analyzed situation should decrease by 7.5% only.

The investigation of the U_DC spike duration (Fig. 7) shows that it decreases when the motor initial rotation velocity (initial inverter output voltage frequency) decreases. However, the dependence of the spike duration on the deceleration rate is not monotonic. It has a peak, at which the spike duration reaches the maximal value. The location of the peak depends on the motor load. It is seen (Fig. 7) that in the analyzed case the dependences have the peak at D = 10Hz/s if the motor load is 3.25Nm, and at D = 17Hz/s when the motor load is 6.5Nm.

Fig.7. The dependences of the inverter DC bus voltage spike duration on the deceleration rate of the motor for various motor load torques and motor initial rotation velocities for the case when C = 470 µF

Conclusions

1.The rising edge steepness of the voltage spike caused by the motor deceleration is relatively low (about 1.5 V/ms). Therefore, the slow and, as a consequence, insensitive to electromagnetic disturbances overvoltage protection circuit with the response time up to several hundreds of microseconds can be employed.

2.The duration of the inverter DC bus voltage spike caused by the motor deceleration decreases when the motor initial rotation velocity decreases.

3.The dependence of the spike duration on the deceleration rate is not monotonic. It has a peak, at which the spike duration reaches the maximal value. The location of the peak depends on the motor load.

4.The increment of the capacitance of the inverter DC bus capacitors allows us to reduce the overvoltage slightly.

This work was supported by the Lithuanian State Science and Studies Foundation under High–tech development program project B-13/2007–2009 and by the Company “Ventmatika” under project U-2007/8.

REFERENCES

[1] Swamy M. M., Kume T.J., Fujii S., Yukihira Y., Sawamura M., A Novel Stopping Method for Induction Motors Operating from Variable Frequency Drives, IEEE Transactions on Power Electronics, 19 (2004), No. 4, 1100-1107
[2] Inoue K., Minamiyama M., Kato T., A design methodology of an optimal torque minimizing energy loss under torque limit for an induction motor, Proc of Energy Conversion Congress and Exposition, ECCE 2009, San Jose, USA, September 20–24, (2009), 163-167
[3] Li J., Tang T., Wang T., Yao G., Modeling and simulation for common dc bus multi-motor drive systems based on activity cycle diagrams, Proc of IEEE International Symposium on Industrial Electronics, ISIE 2010, Bari, Italy, July 4–7, (2010), 250–255
[4] Swiątek H., Michalski A., Flisowski Z., Pytlak A., Insulation co-ordination in power electronic devices at voltage stresses of external origin, Przegląd Elektrotechniczny, 79 (2003), nr 11, 798-805
[5] Hinkkanen, M., Luomi J., Braking Scheme for VectorControlled Induction Motor Drives Equipped with Diode Rectifier without Braking Resistor, IEEE Transactions on Industry Applications, 42 (2006), No. 5, 1257-1263
[6] Wang Y., Yang G., Hong T., Analysis and Implementation of AC Motor Braking Method without Energy Returning or Braking Unit, Proc 8-th Int. Conference on Electrical Machines and Systems, ICEMS 2005, Nanjing, China, September 27–29, (2005), 1447–1451
[7] Lehtla M., Laugis J., Computer models for Simulation and Control of a Traction Supply System, Proc 12-th Int. Conference on Power Electronics and Motion Control EPEPEMC 2006, Slovenia, August 30 Sepember 1, (2006), 1372-1377
[8] Hairik H.A., Thejel R.H., Kadhem W.A., Proposed scheme for plugging three-phase induction motor, Proc 15-th IEEE Mediterranean Electrotechnical Conference MELECON 2010, April 26–28, (2010), Valletta, Malta, 1–5
[9] Baskys A., Rinkeviciene R., Petrovas A., Overvoltage Limitation in Variable Speed Drive with Inverter, Proc 17-th Int. Conference on Electromagnetic Disturbances, EMD 2007, Bialystok, Poland, September 19–21, (2007), 2.3-1–2.3-4

Authors: prof. dr Algirdas Baskys, M. S. Vytautas Bleizgys, M. S. Tadas Lipinskis, Center for Physical Sciences and Technology, A. Gostauto str. 11, 01108 Vilnius, Lithuania, Email: mel@pfi.lt and Vilnius Gediminas Technical University, Naugarduko str. 41, 03227 Vilnius, Lithuania, E-mail: algirdas.baskys@vgtu.lt.

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097, R. 87 NR 5/2011

Utility Capacitor Switching Causes Nuisance Tripping of ASD

Published by Electrotek Concepts, Inc., PQSoft Case Study: Utility Capacitor Switching Causes Nuisance Tripping of ASD, Document ID: PQS0319, Date: July 18, 2003.

PWM ASDs are susceptible to dc link overvoltage trips caused by utility capacitor switching. In general, an increase in input inductance (choke or isolation transformer) will reduce the possibility of nuisance tripping. However, if the customer has power factor correction capacitors on the same bus, it may be necessary to take additional remedial actions.

PROBLEM STATEMENT

A food processing manufacturer was experiencing a problem with a small PWM adjustable-speed drive tripping off-line each morning at the same time. Drive diagnostics indicated an overvoltage trip condition.

The short term solution was to have a technician available to reset the drive each morning.

SYSTEM MODEL

The oneline diagram shown in Figure 1 illustrates the system model used to investigate the nuisance tripping problem. The following parameters describe the system:

System Source Strength: 200 MVA
Switched Capacitor Size: 1200 kVAr
Step-down Transformer Size: 1500 kVA
Substation Load: 5 MW
480 Volt Load: 300 kW

Figure 1 – Oneline Diagram for Nuisance Tripping Case

Capacitor energizing operations, that cause a transient oscillation on the utility system, can cause an overvoltage trip condition on overly sensitive drives.

PWM ASD Characteristics

Pulse-width modulation (PWM) adjustable-speed drives (ASDs) typically have a voltage source inverter (VSI) type of design and use PWM inverters to supply the motor. Drives that use a voltage source inverter design are sensitive to utility capacitor switching transients due to the capacitor used in the dc link between the rectifier and inverter. The protection on the dc link capacitor is very sensitive to transient voltages on the ac power side. It is not uncommon for the dc overvoltage control to cause tripping of the drive whenever the dc voltage exceeds 1.2 per-unit (typical value). Figure 2 illustrates the ASD configuration.

The potential for nuisance tripping is dependent primarily on the switched capacitor bank size (utility), the dc link capacitor size (ASD), and the inductance between the two capacitors. Figure 4.10 illustrates the transient overvoltages associated with energizing the 1200kVAr substation capacitor bank. Energizing transient overvoltages are generally not a concern for utilities because their magnitudes are usually below the protective level at which surge protective devices operate.

However, these transients can be magnified at the customer facility if the customer has power factor correction capacitors. Even if the customer does not have capacitors, the transient on the low voltage bus can cause a surge of current, at a relatively low frequency, into the dc link capacitor. The current surge charges the dc link capacitor and causes a overvoltage to occur. If this voltage exceeds the overvoltage trip setting, the drive will shut down. This is often refereed to as nuisance tripping because the situation can occur day after day, often at the same time. Figure 3 illustrates the dc current and voltage surge for this case.

Figure 3 – Overvoltages Associated with Utility Capacitor Switching

The problem can be solved by adding an isolating inductance in series with the individual drives. This inductance can be in the form of a simple ac choke or an isolation transformer. Typically, a choke size of 3-5% is chosen. If the choke has too much impedance harmonic voltage distortion levels will increase (increased voltage drop due to harmonic current flowing through higher impedance). Figure 4 summarizes the impact of choke size on the dc overvoltage for a case where the switched capacitor bank (utility) was increased to 3000kVAr.

Figure 4 – Effect of Choke Size on dc Link Overvoltage

SUMMARY

Additional actions may include configuring the power factor correction capacitors as harmonic filters. Power factor correction capacitors can create the potential for voltage magnification. This increased voltage transient can often be controlled using harmonic filters, thereby reducing the transient voltage to a level where the 3-5% choke is again effective.

An important point in this case is that the customer should not arbitrarily increase the choke size (above 5%) assuming it will continue to decrease the likelihood of nuisance tripping.

REFERENCES

Electrotek Concepts, Inc., Evaluation of Distribution Capacitor Switching Concerns, Final Report, EPRI TR-107332, October 1997.

RELATED STANDARDS
IEEE Standard 1036-1992

GLOSSARY AND ACRONYMS
ASD: Adjustable-Speed Drive
PWM: Pulse Width Modulation
MOV: Metal Oxide Varistor
TVSS: Transient Voltage Surge Suppressors

Disturbances in Industrial Power Networks

Published by Marek KAŁUSKI, Marek MICHALAK, Mirosław PIETRANIK, Karolina SKRZYPEK, Monika SZAFRAŃSKA, National Institute of Telecommunications

Abstract. The paper presents some preliminary results of the in-situ measurements of disturbances that were observed in mine underground power network. The purpose of the measurements was to identify potential interfering signals that may occur in such and similar power networks and may be dangerous for electronic equipment powered from, or working in close proximity of such network.

Streszczenie. W artykule przedstawiono wstępne wyniki pomiarów in-situ zaburzeń obserwowanych w podziemnej sieci zasilającej kopalni. Celem pomiarów było zidentyfikowanie potencjalnych sygnałów zakłócających pracę innych urządzeń, które mogą wystąpić w takich i podobnych sieciach zasilających i mogą być niebezpieczne dla urządzeń elektronicznych zasilanych z takiej sieci lub pracujących w jej bezpośredniej bliskości (Zakłócenia w przemysłowych sieciach zasilających).

Keywords: power lines, disturbances, industrial environment, mines
Słowa kluczowe: sieci zasilania, zaburzenia, środowisko przemysłowe, kopalnie

Comparison between disturbance and immunity levels

The level of disturbances observed in industrial environments and the required immunity of the devices expected to work in these environments are very often contradictory. Although it is believed that the fulfillment of the requirements according to standards concerning industrial environment (EN 61000-6-2: required levels of immunity and EN 61000-6-4: emission limits) should ensure compatible operation of all devices within a given environment, there are cases when this is not true. In some environments (for example underground mines) the problem of EMC was neglected for many years – there are many old machines that do not comply with standard industrial environment requirements and produce much bigger disturbances. Therefore there is significant risk, that devices introduced in the same environment will not be immune to these disturbances, even if they comply with EN 61000-6-2.

In current practice, many device manufacturers and future users of these devices present the opinion that meeting the aforementioned standard requirements is sufficient “to a peaceful sleep”. Well, it is not so!

It should be noted that the device expected to work in the industrial electromagnetic environment is tested in conditions much different from the environmental conditions in the destination of their workplace. In addition, tests are carried out “here and now” (in a short period of time), recording “continuous” disturbances, which in fact are present in the moment of measurements. In addition, one should note that in the real environment, there may be situations of simultaneous incidence of different disturbances, which present different characters, different ways of coupling etc., resulting in complex disturbance environment.

The EN 61000-6-2 [1] and EN 61000-6-4 [2] standards do not require observing the equipment in the long term, while there is a possibility of higher level disturbance appearance over a longer period of time. This particularly applies to immunity testing.

There are also significant problems resulting from the functional safety of devices and people safety in the environment where even rare events resulting from the disturbance (causing incorrect work of the device) may cause disastrous consequences. Occurrence of such situations when there is a possibility of high level disturbances appearance (occurring from time to time) must be related to functional safety and unanticipated consequences in real environment.

These considerations led the IEC to start work on finding new, more stringent requirements for the immunity of equipment working in special industrial environments. As a result, the new standard is being developed, EN 61000-6-7: “Electromagnetic compatibility (EMC) – Generic standards – immunity requirements for safety related systems and for equipment intended to perform safety-related functions (functional safety) in industrial environments”, for testing the immunity of the devices expected to work in the special industrial environments.

In-situ conducted interference – pulse disturbances

Power network is medium in which many electromagnetic events occur, for example switching on/off of high power devices can produce substantial problems and can provoke unwanted (and potentially dangerous) behaviour of equipment connected to the same power network. Most often these unwanted events are high-amplitude pulses of various duration. They can directly interfere with operations of various electronic control devices, or interfere with machine equipment, resulting in undesired reactions, that can even lead to very serious accidents.

Also sudden on/off operations of large inductive loads, which often occur in case of engines operating heavy lifts, turbines etc. can result in large pulses superimposed on the power waveform, with amplitudes exceeding significantly the peak value of nominal voltage in the power network. The fast transients visible on the slopes of voltage waveform are also observed with amplitudes of the order of few tens of Volts. In medium power and high-power networks the switching can result in the series of pulses of very short rise times (several ns), which can provoke substantial danger for microprocessor driven equipment.

The team of National Institute of Telecommunications, EMC Laboratory in Wroclaw conducted a series of power quality measurements in one of copper mines, both on and underground [4]. Measurements were done with power quality analyzer connected, for different periods of time, to power line connections of different mine machines (for example lifts or conveyor belt engines). Particular attention was paid to the on/off operation incidents and what phenomena connected with these operations can be recorded in terms of power quality.

Fig. 1 shows voltage changes in power line cables of ventilator engine of high effectiveness. Presented is max hold diagram of 40 sample measurements performed in 200 ms intervals. The amplitude of fast changes (in time period of about 100 μs) of voltage waveform exceeds 800 V.

The other phenomena typical for industrial power networks are strong distortions of power waveforms, and consequently the occurrence of harmonic distortions, see Fig. 2. Such disturbances are caused particularly by nonlinear equipment, such as for instance thyristor converters, often used in the lift machine circuits. The associated disturbances can particularly be seen in the lower frequency ranges.

Fig. 1. Voltage changes in power line cables.

Fig.2. Example of current waveform measurement for one of the phase lines of power inverters (1450 rotations/minute) and harmonic distortions. Symbols shown are according to standard EN 61000-4-30

During the tests in underground mines, the team of National Institute of Telecommunications collects information about the electromagnetic environment in the real operating conditions of the mining plant. These measurements are not performed on separated plants or mines for experimental studies, these tests have been done during copper exploitation in real life active mine. The most relevant research concerning the disturbances in electromagnetic environment in the mine turned out to be conducted disturbances – it is the result of the fact that, in most mines, power network is of the “soft” type to which devices, that generate significant transient conditions, are connected.

The measurement of conducted interference in LV industrial power network can only be performed using RF current probe. The use of Artificial Mains Network (typical for laboratory measurements of conducted interference) is impossible for industrial plant (with work currents of several hundreds of Amps with the voltages of 500/1000 Volts, for some industries even higher). There are no AMNs commercially available that can be used in such conditions. Furthermore, in most cases it is impossible to insert the AMN into the power line of a working mine or at least it causes great difficulty. Therefore it is another reason for current probe usage – the simplicity of measuring set-up and ease of use in real world industrial plant environment; there is no need to break the power line, which is necessary in case of AMN. The alternative could be the use of voltage probes, but it can also be difficult in real world industrial plant applications, because of the impossibility of definitive description of reference ground plane in in-situ conditions (especially in mine tunnels) and usually the mains cords are isolated, so there is a problem of how to connect the probe tip to the tested line.

APD measurements of conducted disturbances

Because of disturbance characteristics that can appear in industrial plants power networks, it is justified and worth to use APD (Amplitude Probability Distribution) measurements. As of today, the APD is usually recommended to be used for frequencies above 1 GHz, for wideband digital systems measurements [3]. However, it is worth to draw attention to the fact that nowadays digital systems appear also quite often at the frequencies below 1 GHz, for example BPLC systems in frequency band below 30 MHz. These systems are often used to control big machines or as industrial plants telecommunication systems.

APD can be very useful for disturbance characteristics description in cases, when those disturbances are random and short in time. In those cases one can, in realistic amount of time, evaluate parameters of the disturbances and compare them with current limits, set for measurements with a quasi-peak detector.

APD measurement technique is basically the evaluation of the probability of disturbance level being higher than allowed limit (for radiated or conducted emissions). The measurements are performed after determining the most critical frequency for the evaluated situation.

It is worth mentioning, that long time data gathering using APD allows to determine given disturbance occurrence percentage and its strength (level). Tests of this type are especially important for disturbance connected risks evaluation when safety is concerned.

Measurements in real industrial environment – mine power network

Presented below are some real world measurement results of conducted disturbances measured in real industrial power network. The figures show the electromagnetic disturbances in power networks in relation to current standards. Such relation shows how important it is to provide adequate protection against disturbances of devices in unique environments such as underground mines.

On figures 3 and 4 some real world measurements of conducted emissions in mine power lines are presented in comparison with existing limits for disturbances in industrial environments. As can be seen on those figures, those limits are significantly exceeded, therefore a simple conclusion that there is a need for further tests and analysis of disturbances occurring in those environments. One can ask the question if the immunity requirements are enough to give required protection for equipment introduced to such special environments, especially in case of equipment connected with safety (for example gas sensors and meters).

As it was mentioned before, the moments of on/off operations of high power machines in industrial environment can produce big currents in power lines (see figures 5 and 6). Those currents can present some serious interference for other equipment connected to the same power network. The biggest threat can exist for signal lines as well as control and safety equipment.

Fig.3. Comparison of some real world disturbances with existing limits. Measured in underground conveyor control room in a working copper mine.

Fig.4. Comparison of some real world disturbances with existing limits. Measured in underground conveyor control room in a working copper mine.

Fig.5. The current changer in power line. In 3rd minute – start of the transportation engine.

In existing mines it is also common to arrange different line (power, control and signal) parallel and there is some serious crosstalk between lines. The result of this can be the occurrence in one line of disturbances whose source is in a different line – when disturbances from power line occur in signal or control line they can potentially present significant danger, especially when some safety equipment is connected to these lines (for example gas sensors or meters).

Fig.6. The current changes in the moment of switching the conveyor on, power network analyzer probe connected before inverter.

Conclusion

The description of electromagnetic events given in this paper does not form the complete characterization of all phenomena. Its objective is to draw attention to some disturbing events, observed in the real world, that should be taken into account when evaluating their potential impact on the equipment working in underground mines. Further works are needed to describe some special industrial electromagnetic environments such as the one of underground mines. In 2010 the scientific consortium was formed by AGH University of Science and Technology, Institute of Innovative Technologies EMAG, National Institute of Telecommunications, Institute of Power Systems Automation Ltd. The purpose of this consortium is to prepare such a description for the electromagnetic environment in mines.

REFERENCES

[1] EN 61000-6-2 “Electromagnetic compatibility (EMC) – Part 6-2: Generic standards – Immunity for industrial environments”
[2] EN 61000-6-4 “Electromagnetic compatibility (EMC) – Part 6-4: Generic standards – Emission standard for industrial environments”
[3] CISPR/A/WG2/(Yamanaka-Shinozuka)2003-01, June 2003 Guidance for applying APD method to the compliance test and developing APD limits
[4] Kałuski M., Michalak M., Pietranik M., Skrzypek K. i inni, prace statutowe Instytutu Łączności nr Z21/21300026/1009/06, Z21/21300018/1184/08, Instytut Łączności, 2006-2008

Authors: mgr inż. Marek Kałuski, E-mail: m.kaluski@itl.waw.pl; mgr inż. Marek Michalak, E-mail: m.michalak@itl.waw.pl; dr inż. Mirosław Pietranik, E-mail: m.pietranik@itl.waw.pl; mgr inż. Karolina Skrzypek, E-mail: k.skrzypek@itl.waw.pl; mgr inż. Monika Szafrańska, E-mail: m.szafranska@itl.waw.pl; National Institute of Telecommunications, EMC Department, ul. Swojczycka 38, 51-501 Wrocław.

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097, R. 88 NR 9b/2012

Transient Voltages at an Industrial Facility Capacitor Switching

Published by Electrotek Concepts, Inc., PQSoft Case Study: Transient Voltages at an Industrial Facility Capacitor Switching, Document ID: PQS0505, Date: June 30, 2005.

Abstract: This case study describes concerns for capacitor switching transients within customer facilities. Magnification of transients initiated at higher voltage capacitors and tripping of sensitive variable frequency drive (VFD) loads are evaluated in particular.

The results of measurements and simulations associated with capacitor switching on the utility system and the resulting impacts at an industrial process plant are presented.

These results illustrate the important effect of customer load characteristics on the transients experienced within the facilities.

INTRODUCTION

Capacitor switching is considered a normal system event on the utility system and the transients associated with these switching operations are generally not a problem for utility system equipment. However, the transients can be magnified within customer facilities or can cause misoperation of sensitive electronic equipment, such as variable frequency drives (VFDs).

These concerns were evaluated in detail for a manufacturing facility. The transient overvoltages were simulated using the Electromagnetic Transients Program (EMTP) and field tests were performed to verify the modeling and evaluate the impact on actual plant loads.

Magnification of capacitor switching transients is perhaps the most important concern because the transient voltages can be very high and the energy levels associated with these transients can cause failure of electronic equipment and protective devices. Even without magnification, normal capacitor switching transients can cause problems with sensitive electronic loads. VFDs, in particular, can experience nuisance tripping for transient voltage magnitudes as low as 1.2 per unit.

This case focuses on field tests and simulations performed to evaluate the impact of switching capacitor banks on the transmission system supplying a facility and switching new capacitor banks installed at the distribution substation for the facility.

SYSTEM DESCRIPTION

The facility is supplied from a 115kV transmission system as shown in the representative online diagram provided in Figure 1. The main station that supplies the industrial customer includes two switched 48 MVAr shunt capacitor banks.

The service to the facility is through two 22.4 MVA, 115/12.47 kV substation transformers located on the site. The plant loads are then supplied from service entrance stations that step down to 480 Volts. The electrical loads at the facility consist primarily of motors (pumps and fans) and electronic loads. Most of the manufacturing equipment requires 120/208 V power which is provided through local step-down transformers serving individual loads or combinations of loads through a distribution panel. The most sensitive production equipment is supplied from large scale centralized Uninterruptible Power Supply (UPS) systems. These systems are connected in an on-line full capacity, parallel redundant configuration.

Power factor correction capacitors are not used within the facility. Therefore, magnification at low voltage capacitors is not a problem. In order to maximize capacity of the substation, the utility installed power factor correction capacitors at the 12.47 kV buses. The proposed configuration consisted of 7200 kVAr (6×1200 kVAr) at bus #2 and 3600 kVAr (3×1200 kVAr) at bus # 3. The addition of the substation capacitors prompted the concern for switching transient voltages within the plant. Transients during switching of either the 48 MVAr, 115kV banks or during switching of the individual 1200 kVAr steps were evaluated.

A preliminary model for use with the EMTP was developed to evaluate the concerns for capacitor switching transients. The model was built to accurately represent the system source equivalents at the transmission buses, transformers stepping down to the distribution substations, capacitor banks, and equivalents for loads at the facility and nearby utility buses.

Figure 1 – Representative Oneline for Utility/Customer Capacitor Switching Interactions

A VFD was modeled in detail (Figure 2) to evaluate the impacts of the capacitor switching transients on a typical VFD load. The main concern for the VFDs is that they will experience nuisance tripping caused by an increase in the dc bus voltage within the drive during a capacitor switching transient. The particular drives being evaluated had relatively high dc overvoltage trip settings of 850 Volts. This setting can be as low as 760 Volts for some drives. Previous publications have shown that the nuisance tripping caused by dc overvoltages can be solved by installing an isolating reactor (choke) in series with the drive. The drives being evaluated at the facility did not have chokes but were separated from the main 480 Volt supply by relatively long cable lengths.

The two 48 MVAr, 115kV capacitor banks have 0.3 mH series reactors to limit back-to-back switching inrush currents. The banks are switched with conventional SF6 breakers.

The new 12.47 kV capacitor consist of individual 1200 kVAr steps and series reactors to limit the inrush current during the back-to-back switching. The individual 1200 kVAr steps are switched with motorized air-break switches.

SIMULATION RESULTS

Simulations were performed using the model described to evaluate the expected transient voltages at the facility loads during capacitor switching at either 115 kV or 12.47 kV locations.

Energizing the 48 MVAr, 115kV Capacitor Banks

Switching of the 115 kV capacitor banks was being performed on a regular basis for some time prior to the addition of the 12.47 kV capacitors. This case results in a transient voltage at the facility that has been recorded by monitoring equipment on a number of occasions. Figure 3 is a typical case showing the transient voltage at a 480 Volt bus. These transients were not causing any problems at the facility.

The waveform in Figure 3 was used to estimate the damping being provided by the 115 kV transmission system and load equivalents. Figure 4 illustrates simulation results for the same case. The damping in the simulation case is somewhat less than the damping in the measured waveform that should result in conservative results from the simulations

Figure 3 – Measured and Simulated 480 Volt Bus Voltage during 115kV Capacitor Energizing

Energizing the 12.47 kV Capacitor Banks

This case involves the addition of power factor correction capacitor banks at the 12.47 kV buses. These capacitor banks are configured in 1200 kVAr steps. The first step to be energized results in the highest transient. The transients associated with energizing of subsequent steps are lower because the transient is shared with the other capacitors already in service. Figure 4-Figure 6 give the simulated transients at the 60 HP VFD load for energizing the first 1200 kVAr step. The transients are illustrated in per unit (pu). This is the actual voltage divided by the normal peak voltage magnitude. In Figure 4, the peak transient is approximately 1.3 pu, or about 880 Volts (L-L). The transient magnitude in the dc bus of the drive should be somewhat less than this due to current limiting impedances in the drive. Therefore, it is not likely that this transient would cause drive tripping.

Figure 4 – Simulated VFD Location – ac Input Current

Figure 5 – Simulated VFD Location – ac Input Current

Figure 6 – Simulated VFD Location – dc Bus Voltage

FIELD TEST RESULTS

It was clear from the initial simulations that the potential for nuisance tripping of VFDs or other electronic loads existed during capacitor switching operations. Field tests were performed to evaluate the actual transient voltages at the facility during switching of the new 1200 kVAr capacitor banks and during switching of the 48 MVAr, 115kV capacitor banks. These tests were scheduled during maintenance shut down at the facility to minimize the impacts if any nuisance tripping was encountered.

The test cases performed are summarized in Table 1 below. Basically, energizing operations for the three individual 1200 kVAr capacitor steps at the 12.47kV bus were evaluated first and then one of the 48 MVAr, 115kV banks was energized to evaluate the potential for magnification. Finally, the individual 1200 kVAr steps were de-energized and then put back into service.

For each of the test cases, the transient voltages were recorded at a number of locations. Two of the locations were selected to evaluate the impact of the transients at electronic loads. One site was a 60 HP VFD and the other load site was one of the main UPS installations. The 12.47kV bus was supplying these locations.

Table 1 – Summary of Field Measurements

SUMMARY

The simulations and measurements demonstrated the importance of plant loads in reducing capacitor switching transient voltages. Electronic loads and UPS systems can provide significant damping due to the capacitors and batteries that are effectively connected to the 480 volt bus through diodes or SCRs. These loads make up a significant portion of the total load at the facility.
Nuisance tripping of VFDs and other sensitive electronic loads can be a very important concern in industrial plants with continuous processes. Careful evaluation of the capacitor switching transients using digital simulations can indicate potential problems. However, it may be necessary to perform actual field measurements in cases where the expected transients are near the device limits. For the system studied, damping provided by system loads prevents nuisance tripping of the VFDs.

REFERENCES

S. S. Mikhail and M. F. McGranaghan, “Evaluation of Switching Concerns Associated with 345 kV Shunt Capacitor Applications,” IEEE Transactions PAS, Vol. 106, No. 4, pp. 221-230, April 1986.

G. Hensley, T. Singh, M. Samotyj, M. McGranaghan, and R. Zavadil, “Impact of Utility Switched Capacitors on Customer Systems – Magnification at Low Voltage Capacitors,” IEEE Transactions PWRD, pp. 862-868, April, 1992.

G. Hensley, T. Singh, M. Samotyj, M. McGranaghan, and R. Zavadil, “Impact of Utility Switched Capacitors on Customer Systems, Part II – Adjustable Speed Drive Concerns,” IEEE Transactions PWRD, pp. 1623-1628, October 1991.

RELATED STANDARDS
IEEE Std. 1036-1992

GLOSSARY AND ACRONYMS
ASD: Adjustable-Speed Drive
EMTP: Electromagnetic Transients Program
MOV: Metal Oxide Varistor
PWM: Pulse Width Modulation
TVSS: Transient Voltage Surge Suppressor
UPS: Uninterruptible Power Supply