Você está na página 1de 7

E E Transactions on Power Delivery, Vol. 8, No.

1, January 1993 E

EFFECT OF DISTRIBUTION SYSTEM GR0TJ"G Ramon T. Mancao James J. Burke Member Senior Member Power Technologies, Inc. Schenectady, NY
ABSTRACT
The recent industry change to MOV arresters has created an increased awareness of the importance of calculating accurate magnitudes of line-bneutral overvoltages on the unfaulted phases of a multigrounded neutral three-phase circuit during a fault-toground on one phase. Since most MOV arresters do not incorporate a gap in their design, as did their silicon carbide predecessors, the arrester is always in conduction and considered to be much more prone to damage due to this type of steady state overvoltage. This paper, based on work performed for the Empire State Electric Energy Research Corporation (ESEERCO), investigates many design parameters which affect this overvoltage using a calculation method that considers a detailed representation of neutral grounding electrodes and their resistance which are normally neglected when symmetrical components method are utilized.

ON MOV SELECTION

139

Andrew Myers a New York State Electric & G s Corp. Binghamton, NY

The Working Group summarized the relationship between per-unit overvoltage and conductor size with the results shown in Figure 1. These resultswere calculated using sequence components using both resistance and reactive components: i.e., impedance ratios. The results were considered conservative based upon such factors as:
0

Distribution transformer exciting current was neglected Smaller than normal neutral conductor was used Fault impedance was neglected Voltage drop due to load was neglected.

KEYWORD!? Grounding, MOV, Overvoltage.


INTRODUCTION
In 1972,a Working Group of the Surge ProtectiveDevices Committee proposed in their transaction paper "Voltage Rating Investigationfor Application of Lightning Arresters on Distribution Systems" C7l TP M ~ - P w Rthat the rating of lightning arresters selected for open)~~ wire multigroundedneutral systems be equal to or greater t a the hn product of the nominal (line-bneutral voltage)*(theRange A factor 1.05)*1.2which is the maximum voltage rise on the unfaulted phases of a loaded circuit. This is equivalent to 1.25 limes nominal line-toneutral system voltage.

The conclusions of the group derived f o a series of field tests and rm transient network analyzer studies was that a calculated value of the unfaulted phase voltage should be multiplied by a factor of approximately .92 to account for transformer saturation and other system conditions. Referring to Figure 1, the maximum voltage is approximately 1.30 per unit. This voltage, multiplied by .92, is the 1.2 fador still in common use today. The last 20 years, since the report, several new considerations have arisen that the original study did not address3. 4*5# *' ** Recent developments in more advanced analytical tools now allows the engineer to calculate overvoltages by considering into the analysis a detailed representation of the neutral grounding electrodes; pole ground footing resistances and substation ground mat resistance. The items addressed by this paper are as follows: Symmetrical component calculations Distribution of fault currents Effect of footing resistance and ground rod spacing Effect of neutral wire size Effect of open neutrals Effect of substation ground mat impedance Effect of soil resistivity and lateral wires

TECHNICAL APPROACH
The common approach of calculating overvoltages on the unfaulted phases durin ground faults is normally done using symmetrical components19 or simply referred to in this paper as U L a . A different calculation method, CALC2, was investigated which

Figure 1. Relationship between per-unit overvoltage and conductor sue.

92 Un 233-7 PWRD A paper recommended and approved by the IEEE Surge Protective Devices Committee of the IEEE Power Engineering Society for presentation at the IEEE/PES 1992 Winter Meeting, New York, New York. January 26 - 30, 1992. Manuscript August 21, 1991; made available for printing January 9. 1992.
0885-8977/92/S3."93

Yooting

reslatmm

Figure 2. Circuit model representation used for CALC2.

EE E

140

considers the physical layout of a distribution feeder. The analysis considersan electrical network that takes into account the substation ground mat resistance and the periodic bonding of ground footing resistances on the neutral wire. The complex layout of such a network is illustrated in Figure 2. Having the equivalentnetwork set-up, the node voltages and branch currents are calculated by utilizing a computer program which handles multi-phasecircuit configurations. The software developed, solves a power system by consideringeach wire as a separate entity. Sources, wireruns, transformers, capacitance, and load are connected either between pairs of nodes or between a node and ground. This allows the engineer to model distribution power systems with a broken neutral, various ground impedancevalues, transformerswith unequal winding impedances, multi-wire circuits and other special problems. The node reference of the program is the "far ground reference and all node potentials may be expressed in reference to this far ground or node to neutral values. The pole span spacing between each ground footing resistance defines a section of a long ladder network formed by the equivalent circuit. The parameters of each section consist of the self and mutual impedances of the phase and neutral wires. These impedances are calculated from a transmission line characteristic program using standard formulas with Carson's earth correction terms. The resulting values are used directly for each defined section with the corresponding type and length of the conductor wire. The intent is to make a stronger tie to the physical system problem and obtain a more accurate solution when determining the amount of overvoltage on the unfaulted phases produced by a single line to ground fault. Having all the nodes defined, the calculation method CALC2 builds and inverts a branch impedance matrix and a system admittance matrix. If the system impedance matrix is successfully prepared, CALC2 executes an iterative 'load flow" solution. In essence, the method provides an exact solution of determining the fault current flow in each section of 'the ladder network when applying a line-toground fault at the end of the line. The sharing of fault currents between the neutral wire and earth determines the overall voltage drop across each section. As a result, the maximum rise of the lineto-neutral voltage on the unfaulted phases is the vectorial sum of the voltage drops of each neutral wire section from the fault to the substation and the phase-to-neutral voltage. Solutions obtained by CALC2 are direct methods and exact and are obtained in a definite number of arithmetic operations. In addition, the circuit configuration in CALQ does not make the common assumption that the impedances coupling the positive, negative, and zero sequences are negligible nor that the three phase conductorsare regularly transposed. The circuit model used for the study not only provides a single solution but can be manipulated in a number of ways to determine a variety of results when investigating other values of zero sequence parameters. Typical approaches using symmetrical components, on the other hand, makes simplifying assumptions12. The method considers the neutral conductor and the ground path perfectly coupled, such that the equivalent zero sequence. impedance is basically a parallel combination of the impedances between the neutral and earth ground without the coupling of the ground footing resistances. As a result, this method would give a low zero sequence impedance value of a multi-grounded neutral.1

phase conductor is 336.4 MCM ACSR and the neutral conductor is sized at 3/0 ACSR. The substation two-winding transformer is connected delta-wye with a 21 MVA rating, 138 kV to 13.8 kV and a 10%impedance. The substationgroundmat impedance is assumed to be 1.ohm. The soil resistivity is assumed to be 100 ohm-meters, and the ground rods are placed every quarter mile with an impedance of 25 ohms.

Figure 3. Armless design. Calculation of overvoltages during ground faults is nokally done using the cdlculation method CALCI. When CALCl is used, the effect of ground footing resistances is basically ignored. Figure 4 illustrates the maximum increase of p.u. voltage across the line-toneutral when applying faults at various points on the entire length of the feeder system. The figure shows the differences between CALCl and the CALC2 ap$oach. CALC2 is run for ground footing resistances (Rf) of 25 ohms each and also at 100 ohms each. Method CALCl predicts the highest overvoltage when the fault is administered at the end of the feeder line reaching to as high as 1.290 p.u On the contrary, method CALC2 predicts the highest overvoltage when the fault is implemented closer to the substation. For Rf = 25 Q, the overvoltage is 1.313 p.u. for a fault distance 2 miles down the line and 1.290 p.u. for a fault applied at the end of the line. When R, is changed to 100 51 each, the overvoltage reached 1.339 p.u. for a fault 3 miles and 1.338 p.u. for a fault at the end of the line. The results clearly show the effects of ground footing resistance on overvoltages during ground faults when altered to a different value.

..

CALC2 (Rf=25ohm)

0.751

COMPARISON OF METHODS CALCl AND CALCZ


0
1 2

10

The base case model for the study consists of a distribution feeder with a total main feeder length of 10 miles. Initial line parameters were chosen to reflect some of the more typical configurations that are practiced today. The line construction uses an armless type with a triangular conductor configuration as shown'in Figure 3. The

Fault Distance From Substation (miles)


Figure 4. Comparison of voltage rise between methods CALCl and CALO when applying ground faults at various locations.

141

GROUND vs. NEUTRAL CURRENT FLOW

CALC2 has the capability of monitoring the current flow of individual grounds even if the footing resistances are of different values. This capability allows the program to plot the division of current between the neutral wire and the ground for the entire feeder. Figure 5 shows the division of fault current, on the base case, for a fault located 10 miles from the substation for various sizes of neutral wire. As can be seen, most of the current flows in the neutral at the point of the fault as well as at the source. At points near the middle of the feeder, where the effective earth resistivity stays relatively constant, the division of current remains stable and is approximately 50/50. Near the source, where the effective earth impedance increases, it can be seen that some of the fault current is pushed back into the neutral wire.

the maximum (1.30P.u.) voltage concluded by the Working Group. Field experiencewould also indicate that ground footing resistances above 100 ohms are commonplace.

I
~~

Ground Footingfk&bnce (ohms)


~ ~

Figure 6. Maximum overvoltages.seen at some fault distance of the feeder line for various values of ground footing resistances. (*gpm-groundingrods per mile.) The results a s show that the recommended ground footing lo resistance of at least 4 gpm is not Sufficient to keep the overvoltage down to the desired level. If the number of gpm is increased to 8 gpm, there will be a reduction on the overvoltage level of about 2% with Rf at 25 ohms. In effect, augmenting the gpm value does have an apparent effect in lowering the overvoltage level. Figure 5. Division of fault current for various sizes of neutral wires.

B.

Effect of Feeder Length

As is expected, the size of the neutral does affect the division. For example, use of a full rating neutral (336.4) instead of a 32% rated neutral (1/O) increases the neutral current flow at the substation from 50% to 60%.
FACTORS AFFECTING OVERVOLTAGES The base system used for this study was considered to be more representative of newer designs because it utilized larger wire sizes than .those shown in Figure 1 The expeded overvoltages on . unfaulted phases of a 3 phase system during a lineto-ground fault were expected to be considerably lower &.an those dculated by the Working Group. The ability of the CALC2 method to more accurately calculate these overvoltages, as well as investigate factors that might influence this voltage, are the main areas of interest of this study. Some of the more pertinent findings are as follows:
A.

Another interesting consideration is the effect of the feeder length on the overvoltage level during ground faults. Figure 7 shows that as the main feeder line is made shorter, the resulting overvoltage is higher which is contrary to the CALCl calculations. For 2 miles of .. feeder length, the overvoltage will reach a maximum of 1355 P u , 1.346p.u for 4 miles, 1.319 p.u for 6 miles and 1.313 p.u. for 8 miles as opposed to the 1313 p.u. for 10 m l s Again, this part of the ie. analysis shows how grounding can influence the voltage rise when the length of the feeder is varied. The reason shorter lines are exposed to higher overvoltage during ground faults is mainly attributed to the lesser number of ground footing resistances thereby increasing the zero sequence resistance of the circuit.
1. , h

Effect of Footing Resistance and Ground Rod Spacing

The circuit model used for the study is simulated with various levels of footing resistance while applying ground faults at different locations on the feeder line. The grounding span interval for the study is made to equally represent the industry practice of 4 grounding rods per mile (gpm). The highest overvoltage encountered on the unfaulted phase at some fault distance on the feeder is plotted in Figure 6. A major question for the industry has been, "What value of footing resistance is sufficient?". If an arbitrary voltage limit of 20% is desired, then it can be shown that without other system changes, such as a larger neutral wire, a ground footing resistance value of less than 1 Q is required to satisfy the condition. The normally recommended value of 25 ohms produces an overvoltage of approximately 1.31 p.u. which already goes beyond

Figure 7. Effect of distance of feeder lines on overvoltage with ground faults applied at various locations.

142
C.

Effect o Neutral Wire Size f

Figure 8 illustratesthat wire s e e has a considerable effect on system overvoltages during ground faults. The base case model considers a neutral conductor size of 3/0 ACSR where its cross sectional area is 50%smaller when compared to the 336.4 ACSR used for the phase wires. As shown in the figure, the overvoltage can be reduced by approximately 5%if the neutral wire size is selected to be of equal f rating with that o the phase conductors. This reduces the overvoltage, for a fault 1.5 miles from the substation, from 1.308 p.u. to 1.242 pa.. If the neutral wire size is grossly undersized relative to the phase wires, it would' not be surprising to see the level of overvoltage go up. For a fault 1.5 miles from the substation, the utilization of a 1/ O ACSR neutral conductor increases the overvoltage to approximately 1.354p a . or an increase of about 4% from the base case results.

the data base structure. The question raised is "What if the ground rod resistance near the fault is very high?". To illustrate this effect, three fault locations along the feeder are selected. The first fault point is located two miles from the substation, the second is about midway or 4 miles down and the third is at the end of the feeder line or at 10 miles. At the selected fault area, the ground footing resistance at that point is vaned from a low of 1 $2 to as high as 10,000 Q. The effects of this analysis are shown in Figure 9. As can be seen, the rise in voltage on the unfaulted phase is reduced considerably at the fault location when Rf at that location is vaned from a low of 1Q to as high as 10 a. Very important is the fact that a high ground rod impedance near the fault has very little effect if other system ground rod resistances are low.

E.

Effect of Broken Neutrals

pj
0.

, , , , , , , , , , , , , .......................
I

1
9

A major concern in the industry is the effect a broken (open) neutral conductor can have on these steady state overvoltages. In the past, this calculation was not possible using symmetrical components. Results of this study show that the overvoltage produced for this condition is very much related to the location of the fault. For example, if a fault is placed at the end of the 10 mile line, the worst overvoltage calculated is about 1.33per unit for any location of the broken neutral. If the fault were located 1.5 miles from the substation, some unusually high voltages are found as shown in Figure 10. As can be seen, if the break occurs between the fault location and the substation, voltages of greater than 1.5 per unit can . beexpected.
Case 1. Faun at end of line

10

Fault Distance F

m Substation (miles)

IO mi __l____e/

neutralwire (N)

Figure 8. Effect of neutral wire size on the voltage rise with ground faults applied at various locations.

1.0 p.u.

r
Maximum rawe d volage rise on the unfaultedphase (P.u.)

D.

Effect o Ground Footing Resistance at the Fault Location f


Case2. Faun applied at 1.5 miles frofnsubstaiion

In actual conditions, the ground footingresistance from one location


to another may not be of the same ohmic value. For practical purposes, the base case model for the study considers each ground footing resistances to be of equal value for the entire main feeder which refers to the industry recommended value of 25 Q each. The assumption makes the construction of the circuit model representation less difficult and helps facilitate the handling of
I

w*
1.o p.u.

+1.5rri

neutralwire (N)

I
1.34 p.u.

1.58 p.u.

0
Maximumrange of voltage rise on the unfaulted phase (PA.)

Figure 10. Maximum overvoltage seen for a ground fault applied at the end of the line and 1.5 miles down the line with the presence of a broken neutral.

F.

Effect of Substation Ground Mat Impedance

Ground Resistance at Fault Location

Figure 9. Effect of ground footing resistance at the location of the fault on the overvoltage during ground faults.

The substation modeled for this study was solidly grounded. This means that no intentional impedance has been purposely added to reduce the fault current. No grounding is perfect however, and for study purposes, the resistance to the substation ground is assumed at 1 R. It is interestingto note that the substation ground impedance can have an effect on the level of overvoltage during line-to-ground 1 f faults. Figure 1 showsthe effect o substation grounding resistance (Ro) when simulatedat R, = 0.5 Q, % = 1.0 Q, = 2.0 Q, and = 5.0 R. With 4 = 5 C, the overvoltage experienced for ground fault & applied at various locationson the feeder is predominantly above the 1.20 p.u. with the highest overvoltage magnitude of 1.345 p.u. Lowering ground impedance to = 0.5 still has most sections above the 1.20 p.u. value with 1.304 p.u. as the maximum voltage rise for a fault 2 miles down the l i e .

143

Previous work from 1971 by a Working Group of the Surge ProtectiveDevice Committeesuggested that an overvoltage factor of 1.25 per unit be utilized for arrester selection. This factor was based on a 1.2 per unit overvoltage component that included transformer saturation as well as a 5% regulation factor. This factor was considered conservative because system parameters utilized were thought to be pessimistic. This study recommends that a new overvoltage factor of at least 1.35 per unit be considered by all utilities utilizing MOVs for 4 wire multigrounded systems. This new factor is suggested based on the following:
1.

2 0.75
z

Saturation curves in newer transformers are higher. Overvoltages reaching 1.30 per unit are attainable on common system designs. Regulation above 1.05 per unit is common.

2.

3. Figure 11. Effect of substation ground mat impedance on the voltage rise for ground faults applied at various locations

*Membersof ESEERCO are: Central Hudson Gas & Electric Corporation Consolidated Edison Company of New York, Inc. New York Power Authority New York State Electric & Gas Corporation Niagara Mohawk Power Corporation Orange and Rockland Utilities, Inc. Rochester Gas and Electric Corporation

G.

Other Effects

f A couple o other areas studied but having little influence on overvoltage were:
1.

Soil Resistivity - When soil resistivity was changed from 100 ohm-m .to loo0 ohm-m, virtually no change in overvoltage occurred. While the study concluded that soil resistivity had little effect, it was noted that the attainable ground rod resistance is directly affected and ground rod resistance has a significant effect as noted earlier. Influence of Lateral Lines - Laterals were placed along the

REFERENCES
1.

2.

30 feeder to determine their effect in reducing overvoltages.


While the lateral grounds did decrease the overvoltage, this effect was generally less than 2% CONCLUSIONS AND RECOMMENDATIONS The study produced a number of results of considerable significance to the distribution engineer. Several of the more important findings are as follows:
1.

Working Group of Surge Protective Devices Committee, "Voltage Rating Investigation for Application of Lightning Arresters on Distribution Systems", IEEE Transactions on Power Apparatus and Systems, vol. 91, no. 3, pp. 1067-1074, May/June 1972.

2.

ANSI Standard C84.1-1970, Voltage Ratings for Electric Power Systems and Equipment (60 Hz).
E.C. Sakshaug, J.S. Kresge, S.A. Miske, Jr., A New Concept in Station Arrester Design", IEEE Transactions on Power Apparatus and Systems, vol. 96, p. 647, March/April1977. S.A. Seeker, "Protection Substation Equipment with MOV Surge Arresters", T&D Magazine, May 1981. E.C. Sakshaug, J.J. Burke, J.S. Kresge, "Metal Oxide Arresters on Distribution Systems Fundamental Considerations", IEEE Transactions on Power Delivery, Vo1.4, No.& pp. 2076-2089, October 1989.

3.

4.
5.

Obtaining a 25 ohm ground footing resistance does not insure low overvoltages during line-to-ground faults. Contrary to the findings of method CALCl calculations,the highest overvoltages during line-to-ground faults can be found on short feeders and for faults close to the substation. Small neutral wire can significantly increase overvoltages. Broken neutrals can result in overvoltagesexceeding 1.5per unit during line-to-ground fault conditions (neglecting transformer saturation).

2.

6.

3.
4.

J.J. Burke, E.C. Sakshaug, S.L. Smith, "The Application of Gapless Arresters on Underground Distribution Systems", IEEE Transactions on Power Apparatm and Systems, vol. 100, pp. 1234-1243, March 1981.
R.H. Hopkinson, "Better Surge Protection Extends URD Cable Life", IEEE Transactions on Power Apparatus and Systems, vol. 103, pp. 2827-2836, October 1984. S.S. Kershaw, G.L. Gaibrois, K.B. Stump, "Applying Metal Oxide Surge Arresters on Distribution Systems", IEEE Transaction, PD, Vol4, No. 1, pp. 301-3W, January 1989.

7.

As discussed in this ESEERCO" sponsored study, it is highly probable that voltages up to or exceeding 1.30 p.u. will occur on typical 4 wire, multigrounded circuits during line-to-ground fault Field conditions i f transformer saturation is neglected. measurements,13, l4 performed on a random basis, show per unit overvoltages in excess of this 25% value indicating that transformer saturation cannot always be relied on to limit temporary overvoltages to less than 25%. It is also welI known that voltage regulation on the distribution feeder often exceeds 1.05 per unit and has been measured in excess of 1.10 per unit."

8.

9.

J.J. Burke, V. Vameckas, E. Chebli, G. Hoskey, "Application of MOV & Gapped Arresters on Non Effectively Grounded Distribution Systems", IEEE Transactions on Power Delivery, Vol. 6, N0.2, pp. 794-800, April 1991.

144

10.

M.V. Lat, "DeterminingTemporary Overvoltage Levels for Application of Metal Oxide Surge Arresters on Multigrounded Distribution Systems", IEEE Transactionson Power Delivery, vo1.5, No.2, pp. 936-946, April 1990. J.J. Burke, D.A. Douglass, D.J. Lawrence, EPRI Report, "Distribution Fault Current Analysis", EL3085, May 1983. "ElectricalTransmission and Distribution Reference Book, Westinghouse Electric Corp., 1964. P.P. Barker, J.J. Burke, R.T. Mancao, C.W. Burns, J.J. Siewierskii "Power Quality Measurements on the Niagara Mohawk Power System ...Revisited",PCIM/Power Quality '90, October 21-26,1990. W.M. Edmunds, G.B. Roloson, "Rural Distribution Voltages", AIEE Transactions, Vol. 67-11,pp. 13241329,

11.

12. 13.

14.

1948.

RAMON T. MANCAO - Mr. Mancao received his BSEE degree from the University of San Carlos in the Philippines in 1978 and a MSEE degree from Syracuse University in 1980. He also earned his MEng in Electric Power Engineeringfrom Rensselaer Polytechnic Institute in 1982. He joined M n l aia Electric Company in the Philippines in 1982 and worked in the area o f distribution engineering.
Mr. Mancao joined Power Technologies, Inc. in January 1989 as an Analytical Engineer in the Distribution Engineering Unit. Among the areas he is involved with are overvoltage protection and distributioninstrumentation. Mr. Mancao is a member of the IEEE Distribution Subcommittee.

JAMES 1. BURKE - h4r. Burke received his BSEE from the University of Notre Dame in 1965 and his MSIA from Union College in 1970. In 1965 he joined the General Electric Company, and worked as district application engineer and later as a protection engineer.
Mr. Burke joined Power Technologies, Inc., September 1978 as a Senior Engineer, and is presently unit manager in the area of distribution engineering. He has authored and co-authored over 40 technical papers, is a senior member of IEEE, and is presently Chairman of the Distribution Subcommittee.

ANDREW MYERS - Mr. Myers received his BSEE (1972)and MSEE (1974)from Clarkson University. He has worked at a New York State Electric & G s Corp. for 16 years in electric distribution engineering. He has been involved with overhead distribution performance, design, and problem analysis. He presently supervises the evaluation, maintenance, quality assurZnceand failure investigations of primary overhead and underground distribution equipment.

145

DISCUSSION
The authors would like to thank Mr. Mackevich for his interest and well thought out questions. The questions will be addressed in the order in which they appear:

RA0 S THALLAM, Salt River Project, Phoenix, . Arisona: The authors have presented results of an exhaustive study of distribution system temporary overvoltages experienced during As the single phase-to-ground faults. authors have correctly pointed out metal oxide arresters are more sensitive to such overvoltages than silicon carbide arresters. The authors have used Igload flow" type solution for studying overvoltages. This 1s appropriate as the purpose of the study is t o calculate the long duration overvoltages. However, these overvoltages can also be studied by modeling the system with time domain programs like EMTP (Electromagnetic Transients Program). EMTP gives both a loadflow solution, and also the time domain values. Another advantage is transformer saturation can also be modeled which can have impact in some cases with high overvoltages. Have the authors considered verifying some cases with EMTP modeling?
M a n u s c r i p t r e c e i v e d F e b r u a r y 24, 1992.
J.P. Mackevich (Raychem Corporation, Menlo Park, CA) n e authors have presented very interesting findings in this paper. Such information is of great value to the industry, and the authors are to be " m e n d e d for their innovative work. Could the authors comment on the following:

.5 The 1 3 p.u. factor is recommended for effectively grounded systems i.e. %/xl 3 and ro/xl < 1 Many . four-wire, multigrounded systems do not meet this criteria. Based on study results, Figure 6 would suggest a 1.42 p.u. overvoltage factor as a result of the assumed base case model. Our interpretation on Figure 6 is that the effect of ground footing resistance cannot be considered minimal.
We would use the 1.42 factor for the unigrounded system as well. The IEEE "ApplicationGuide for Grounding Distribution Systems" (soon to be published) recommends a 1 4 factor. We would not treat this as a broken . conductor. We would classify a unigrounded system as a "non-effectively grounded system so the previous comments would apply.
No, since the present value used for an underground system is 1.82 p.u. and is considered conservative.

The major variable not considered is saturation. The computer programs available can simulate these effects but it is difficult to reach agreement on what family of saturation curves is valid.

1) 'Ihe authors appear to imply that ground footing resistance has a minimal effect on the overvoltage magnitude. Is it the conclusion of the authors that the suggested 1.35 per unit factor be applied for both effectively grounded as well as noneffectively grounded systems? If not, what value would the authors suggest be used for non-effectively grounded systems?
2 ) How applicable i the study performed by the authors for a unigrounded wye s

system that is considered effectively earthed? Would this be treated as a broken neutral case with a 1.5 per unit factor, or is the factor a larger number? Would the suggested factor change significantly for a non-effectively earthed unigrounded wye system? 3) Does the development of this new analytical tool suggest a different overvoltage factor for ungrounded systems or is that beyond the scope of this software program?
4) Clarification of the validation of the simulations by the data of references 13 and 14 would be appreciated. What variables of the simulation are not represented by

The authors would also like to thank Dr. Thallam for his interest. Many of the results shown in the paper were run on both CALC2 and EMTP for verification. The results are almost identical. Once again, the problem with saturation is not with the software capability but with the correct model. This will hopefully addressed in the future when more field test data on transformer design becomes available.
M a n u s c r i p t r e c e i v e d March 31, 1992.

field measurements and what further validation might be considered? M a n u s c r i p t r e c e i v e d F e b r u a r y 11, 1992.

Você também pode gostar