Pterra Consulting

Pterra Report R149-16

Assessment of Inverter-based
Distributed Generation Induced
Ground Fault Overvoltage on Delta-
Wye Substation Transformers

Submitted to

NYSERDA
January 3rd, 2016

4 Automation Ln, Albany, NY 12205, Tel: (518) 724-3832, www.pterra.com

 Contents

Executive Summary .............................................................................. 1

Introduction .......................................................................... 4
Background of 3V0 Requirements ............................................................. 4
Objective............................................................................................... 6
Methodology and Assumptions ................................................ 7
Methodology .......................................................................................... 7
Assumptions .......................................................................................... 8
Simulation Results and Discussion ........................................... 9
Scenario 1 through 15 (Table 3-1) ............................................................ 9
Scenario 16 through 23 (Table 3-2) ........................................................ 11
Scenario 24 through 28 (Table 3-3) ........................................................ 13
Conclusions and Future Work ................................................. 19
Conclusions ......................................................................................... 19
Future Work ........................................................................................ 20
Appendix A-Simulation Plots for Each Scenario ......................................... 1
4.2.1. Scenario#1....................................................................................... 1
4.2.2. Scenario#2....................................................................................... 2
4.2.3. Scenario#3....................................................................................... 4
4.2.4. Scenario#4....................................................................................... 5
4.2.5. Scenario#5....................................................................................... 7
4.2.6. Scenario#6....................................................................................... 8
4.2.7. Scenario#7..................................................................................... 10
4.2.8. Scenario#8..................................................................................... 11
4.2.9. Scenario#9..................................................................................... 13
4.2.10. Scenario#10 ................................................................................. 14
4.2.11. Scenario#11 ................................................................................. 16
4.2.12. Scenario#12 ................................................................................. 17
4.2.13. Scenario#13 ................................................................................. 19
4.2.14. Scenario#14 ................................................................................. 20
4.2.15. Scenario#15 ................................................................................. 22
4.2.16. Scenario#16 ................................................................................. 23
4.2.17. Scenario#17 ................................................................................. 25
4.2.18. Scenario#18 ................................................................................. 26
4.2.19. Scenario#19 ................................................................................. 28
4.2.20. Scenario#20 ................................................................................. 29
4.2.21. Scenario#21 ................................................................................. 31
4.2.22. Scenario#22 ................................................................................. 32
4.2.23. Scenario#23 ................................................................................. 34
4.2.24. Scenario#24 ................................................................................. 35
4.2.25. Scenario#25 ................................................................................. 36
4.2.26. Scenario#26 ................................................................................. 36
4.2.27. Scenario#27 ................................................................................. 37
4.2.28. Scenario#28 ................................................................................. 38
 Legal Notice

This report was prepared by Pterra, LLC as an account of work

undertaken for the New York State Energy Research and Development
Authority (NYSERDA), which individually and collectively do not:
1. Make any warranty or representation, expressed or implied,
with respect to the use of any information contained in this
report, or that the use of any information, apparatus,
method, or process disclosed in the report may not infringe
privately owned rights.
2. Assume any liabilities with respect to the use of or for
damage resulting from the use of any information, apparatus,
method, or process disclosed in this report.
The report is confidential and proprietary and may not be copied,
duplicated, or loaned or transferred to a non-designated party without
the express written permission of Pterra, LLC, and NYSERDA

Pterra Report R149-16 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Executive Summary
New York State Energy Research and Development Authority (NYSERDA)
contracted Pterra, LLC (“Pterra”) to conduct an engineering study on the
conditions when potential ground-fault-overvoltage (GFOV) and a consequent
3V01 requirement occurs involving inverter-based distributed generation
(DG).
Recent interconnection studies for proposed inverter-based photovoltaic DG
have identified a potential regarding GFOV where a portion of the combined
subtransmission and distribution system would not have a grounding source
after the opening of a recloser or a circuit breaker upstream of the fault
location. Such studies, conducted by utilities in New York State, have
identified the need for 3VO-based protection schemes to mitigate the
potential GFOV. The 3VO requirement could be cost prohibitive for DG
projects because it requires installation of new potential transformer (PT) on
the transmission side and additional detection and relaying equipment.
The objective of this study was to demonstrate the impact of GFOV with
inverter-based distributed generations on a distribution circuit considering
the following parameters:
 Penetration level (DG to load ratio) on the islanded system
 The load on distribution, transmission, and combination of both
 Inverter control designs from different manufacturers
 Surge arresters on the transmission side
 The load’s Quality Factor
Based on the study, the hope is that a better understanding of the
phenomena relating to GFOV and related 3V0 protection is achieved, and
alternative potential resolutions that are less cost prohibitive may be
identified.
This study is Phase 1 of an ongoing effort to provide technical clarity on the
issue of GFOV. This phase is focused on identifying, quantifying and
evaluating the potential impacts of the phenomenon of GFOV. Subsequent
phases are planned to consider alternative mitigation options and system
performance criteria.
This study used time-domain simulation of various operating scenarios to
evaluate and clarify the GFOV phenomenon. The software used was the

1
Normally, the three winding voltages of a transformer add vectorially to zero. When ground
faults occur, the voltage that appears corresponds to 3 times the zero-phase-sequence component of any
one of the three phase-to-ground voltages at the potential-device location. This voltage is referred to as
"3V0 " and is the basis for certain types of protective relaying schemes.

Pterra Report R149-16 1 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
PSCAD/EMTDC commercial package developed by the Manitoba HVDC
Research Center.
A test system comprising of a subtransmission circuit feeding a distribution
feeder through a delta-wye grounded transformer was modeled in software.
Proprietary manufacturer models for three different inverter designs were
integrated with the model. Twenty-eight scenarios were then developed to
help focus on various aspects of GFOV. These scenarios can be generally
grouped into the following:
 Scenarios to identify critical penetration levels for different inverters,
load locations, and DG to load ratio
 Scenarios illustrating the effect of surge arresters in the simulations
 Scenarios to evaluate the impact of multiple inverters of different
manufacture
 Scenarios to study the impact of load quality factor

FINDINGS:
The following is a summary of the findings of this study based on the
observed results of simulations.
 While inverters can potentially cause overvoltage on the delta side of
the substation transformer, some inverter designs can detect a single-
line-to-ground fault condition and trip instantaneously. Time domain
simulation is a potential tool for evaluating the fault detection
capability of inverters for this purpose.
 According to ANSI/IEEE C62.92, the GFOV for an effectively grounded
system is to be limited to 138%. This value can also be used to limit
the overvoltage for ungrounded systems. Simulation results indicate
that overvoltage on the delta side of the study substation transformer
peaks at 1.38 PU as the PV/load ratio approaches 65%. At
penetration levels below 65%, no overvoltage is observed. Two
important notes relate to this finding:
a. The calculation for the load should account for those connected
to the transmission side as well as the distribution side of the
isolated system.
b. Though this ratio seems close to the threshold proposed by
National Grid (i.e. 67%), there is possibility of under counting
the load if only the distribution side load is considered.
 The 65% penetration limit (based on 1.38 PU overvoltage threshold)
can be relaxed if:

Pterra Report R149-16 2 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 a. Damage to equipment connected to delta side of the substation
transformer is the reason for requiring 3VO protection; and
b. Surge arresters connected to delta side of the substation
transformer are taken into account
Simulations conducted in this study with station class surge arresters
indicate that arresters can safely operate for penetration levels of up
to 100%.
This report is not intended to impose pass-fail criteria for GFOV nor does it
provide basis for requiring 3VO protection due to GFOV. The intention is to
add to the body of knowledge on this topic and contribute to future pass-fail
criteria that may be developed through a consensus-based process including
various industry stakeholders and taking into account all circuit
configurations, a variety of inverter manufacturers and the best available
information on distribution system requirements.

FUTURE WORK:
As noted earlier, this study is phase 1 of an overall effort to provide technical
clarity on the issue of GFOV. A future aspect of study is:
 Investigation of less expensive countermeasures for 3V0 requirement
to alleviate a potential GFOV issue. Such countermeasures could
include, but not limited to, protective relaying detection on the
distribution side, detection by the inverter, application of a grounding
switch on distribution feeder, and addition of a grounding bank on the
transmission side.

Pterra Report R149-16 3 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Introduction

New York State Energy Research and Development Authority (NYSERDA)

contracted Pterra, LLC (“Pterra”) to conduct the following engineering study
in relation to ground-fault-overvoltage (GFOV) impact and mitigation via 3VO
protection.

Background of 3V0 Requirements

Recent interconnection studies for proposed inverter-based photovoltaic DG
have identified a potential regarding GFOV where a portion of the combined
subtransmission and distribution system would not have a grounding source
after the opening of a recloser or a circuit breaker upstream of the fault
location. Such studies, conducted by utilities in New York State, have
identified the need for 3VO-based protection schemes to mitigate the
potential GFOV. The present 3VO requirement could be cost prohibitive for
DG projects because it requires detection and installation of potential
transformers (PT) on the transmission side as shown in Figure 1-1,
highlighted in yellow.
To further illustrate the problem related to GFOV and the 3VO protection
requirement, the following sequence of events can be postulated:
1. A single-line-to-ground (SLG) fault occurs at any of the locations F1,
F2, or F3. For this example, we assume the fault location is at F1 or
Phase A as shown in Figure 1-1.
2. The fault is detected by the subtransmission protection which would
trip the circuit breaker. Prior to the opening of the circuit breaker,
there is no overvoltage because the ground source from sub-
transmission system is still intact. Figure 2-2(b) show VAN = 0, VBN =
VCN = 1 PU.
3. After several cycles of the SLG fault detection, the circuit breaker
trips; At this point the system loses its grounding source and causes
neutral point to shift. The overvoltage could occur on the un-faulted
phases as shown in Figure 2-2(c): VBN = VCN =1.732 PU or 173%
overvoltage, , assuming an initial voltage of 1.0 PU.
4. One option to protect against the overvoltage is to provide for 3V0
detection capability through the installation of PTs on the high-side of
the transformer and consequent tripping signals.
A key assumption that supports this sequence of events is:
 DG does not detect the fault and does not trip immediately and thus
continue to energize and feed the fault.

Pterra Report R149-16 4 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Sub-Transmission Distribution
Circuit Breaker Fault Location
Substation Transformer

PT
Distributed Generation(DG)

Figure 1-1. Illustration of the 3VO Requirement with Potential Transformer (PT) on the
Transmission Side

Figure 1-2. (a)Voltages Prior to Single-line-to-ground Fault (SLG), (b)During SLG with Breaker
Close, and (c)During SLG with Breaker Open

Pterra Report R149-16 5 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Objective
The objective of this study is to demonstrate the impact of GFOV with
inverter-based distributed generations on a distribution circuit considering
the following parameters:
 Penetration level (DG to load ratio) on the islanded system
 The load on distribution, transmission, and combination of both
 Inverter control designs from different manufacturers
 Surge arresters on the transmission side
 The load’s Quality Factor

Pterra Report R149-16 6 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Methodology and Assumptions

Methodology
Time domain simulation software (PSCAD™)2 is used for the assessment. Figure 2-1
shows a test circuit considered in this report for evaluation of ground fault
overvoltage in an islanded circuit. After DG (i.e. PV) reaches steady state condition,
a permanent single line to ground fault is initiated on delta side of substation
transformer. In five cycles, circuit breaker connected on faulted side of transformer
opens to clear the fault from the grid. The operation of the circuit breaker forms an
island with local loads and existing PVs. Single line to ground fault together with the
lack of grounding source makes delta side of transformer vulnerable to high
overvoltage.

Figure 2-1: Single Line Diagram of Studied Circuit with Single Inverter

A total of 28 Scenarios were identified, developed and tested for the study. The
scenarios are summarized in three tables, as follows:
1. Table 3-1 (Scenario 1 through 15). The objective of these scenarios is to
illustrate critical penetration level for different inverters, load locations, and
DG to load ratio.
Three inverters used for the study are commercially available; however,
manufacturer names and model numbers are replaced using arbitrary
designations for confidentiality.
 Inverter#1: 250 kW, three-phase, UL-1741 certified
 Inverter#2: 250 KW, three-phase, UL-1741 certified
 Inverter#3: 1 MW, three-phase, UL-1741 certified

2
PSCAD/EMTDC is a commercial software package developed by the Manitoba HVDC Research Center.

Pterra Report R149-16 7 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Three locations of the load were considered:
 All the load is on the transmission side/delta side of the substation
transformer. Load on delta side is assumed un-grounded.
 All the load is on the distribution side / wye-grounded side of the
substation transformer. Load on wye side is assumed grounded.
 Combination of the above
For each inverter, the load on the circuit was modified in such a way that
the DG to load ratio (penetration level) in the island varies from 50% to
105%.
Critical penetration level is then determined when the GFOV reaches the
magnitude of 1.38 PU. According to ANSI/IEEE C62.92, the GFOV for an
effectively grounded system is limited to 138%. In other words, even if
there is no DG on the circuit, the 138% GFOV still could be expected on an
effectively grounded system.
2. Table 3-2 (Scenario 16 through 23). The objective of these scenarios is
to study the impact of the surge arresters in relation to the GFOV as well as
to demonstrate the capability of the surge arresters in handling the GFOV.
Critical penetration level is determined by the capability of the surge arrester
in handling the GFOV. It is assumed that surge arresters are coordinated and
designed to protect the substation equipment.
3. Table 3-3 (Scenario 24 through 28) are intended to study the impact of
multiple inverter-types connected to the same feeder and modeling of high-
quality load factors.

Assumptions
The following assumptions apply to the modeling of scenarios and the simulations
conducted:
 To obtain conservative results, active anti-islanding protections in inverters
and non-linearity of transformers (magnetizing and saturation) were
deactivated in all scenarios.
 The external grid is modeled as an ideal source. This assumption also leads
to conservative results as infinite short circuit capability of the grid tends to
mask the fault for inverters.
 The reactive portion of the load is tuned to minimize reactive power
mismatch in the island. This is likewise a conservative assumption. Note
that Pterra has studied many inverter models developed by leading vendors.
A common pattern observed in all studied models indicates that even small
amount reactive power mismatch is sufficient for such inverters to detect
islanding conditions.
 The current chopping limit for the circuit breaker is at zero amps.

Pterra Report R149-16 8 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Simulation Results and Discussion

For clarity, simulation results are presented and discussed based on the group of
scenarios as summarized in Table 3-1, Table 3-2, and Table 3-3. All the simulation
plots are provided in Appendix 1.

Scenario 1 through 15 (Table 3-1)

The objective of these scenarios is to illustrate critical penetration level for different
inverters, load locations, and DG to load ratio.
Table 3-1 summarizes the first set of simulation results. This table shows the GFOV
magnitude on the delta side of the substation transformer. It is observed that
GFOV depends on penetration level and individual inverter’s control and protection
scheme. Table 3-1 also shows that the load on sub-transmission side or delta side
of substation transformer will affect the magnitude of the GFOV and should be
considered in the GFOV study.
The loads are modeled as follows:
 In Scenario #1 through #5, the load is connected on wye side or distribution
side of the substation transformer. The penetration level is varied from 50%
to 105%.
 Scenario# 6 through 10 are similar Scenario #1 through #5, except the load
is connected on delta side or sub-transmission side of the substation
transformer.
 Scenario 11 through 15 are similar to Scenario#1 through #5, except the
loads are connected to both distribution and sub-transmission side.
Each scenario was simulated with each of three different inverters mentioned in
Section 2. Since inverters have different initialization time, fault initialization time
was different among studied cases. The fault was activated at .6 s, 1.2 s and 1.6 s
for Inverter#1, Inverter #2, and Inverter #3, respectively. In all cases, circuit
breaker BRK1 in Figure 2-1 clears the fault from the grid in five cycles.
From the simulation plots, Inverter#1 did not detect the fault and maintained
energization of the island for all scenarios. Inverter#2 detects the fault and trips
within 3 cycles after the island is formed in all scenarios. The fault detection
performance for Inverter#3 falls between Inverter #2 and Inverter #1.
Among the three inverters, the inverter with the least fault detection capability (i.e.
inverter#1) first caused GFOV of 1.38 PU on the delta side of substation
transformer when penetration level increased to 65%. According to ANSI/IEEE
C62.92, the GFOV for an effectively grounded system is limited to 138%. Assuming
1.38 PU as the maximum permissible overvoltage for the GFOV, the critical
penetration limit is 65%. This penetration limit is slightly lower than the 67%
criteria used by a National Grid.

Pterra Report R149-16 9 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
It should be noted that loads connected to delta side or sub-transmission system
should be considered in the calculation of the penetration level. Comparison among
scenario#1 through scenario#5, scenario#6 through scenario#10 and scenario# 11
through scenario#15 show comparable GFOV magnitude for the cases with different
load locations.
Table 3-1: Scenarios without Surge Arresters
DG/LD_L1 DG/LD_H2 DG/(LD_L+LD_H) Over Voltage (Steady State)
Scenario
% % % Inverter#1 Inverter#2 (Inverter#3)
1 50 1.2 N/A3 N/A
2 65 1.38 N/A N/A
3 90 1.62 N/A N/A
4 100 1.71 N/A 1.73
5 105 1.75 N/A 1.78
6 50 1.2 N/A N/A
7 65 1.38 N/A N/A
8 90 1.62 N/A N/A
9 100 1.7 N/A N/A
10 105 1.75 N/A 1.77
11 50 1.2 N/A N/A
12 65 1.38 N/A N/A
13 90 1.62 N/A N/A
14 100 1.71 N/A 1.72
15 105 1.75 N/A 1.77
1) LD_L load on the distribution side / low voltage or wye-grounded side of the substation transformer
2) LD_H load on the sub-transmission side / high voltage or delta side of the substation transformer
3) N/A: No overvoltage was observed. Inverter tripped almost instantaneously following fault/islanding

Pterra Report R149-16 10 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Scenario 16 through 23 (Table 3-2)
Overvoltage magnitudes tabulated in Table 3-1 can be conservative because no
surge arresters were considered for the substation transformer. Potential
overvoltage can be further reduced by considering the effect of surge arresters.
The impact of surge arresters depends on the type and rating of the arresters. In
order to illustrate the impact of surge arresters, selected scenarios in Table 3-1
were repeated with the following surge arresters:
 Case A: One 22 KV MCOV station class surge arrester (i.e. EVP0 02200) on
transmission side
 Case B: One 24.4 KV MCOV station class surge arrester (i.e. EVP0 02400)
on transmission side
 Case C: One 29 KV MCOV station class surge arrester (i.e. EVP0 02400) on
transmission side
Simulation results are tabulated in Table 3-2. In scenario 16 through scenario#19,
the load is connected to distribution feeder; whereas in scenario#20 through
scenario#23, the load is connected to the transmission (delta side of the
transformer).
Simulation results tabulated in Table 3-2 demonstrate the importance of surge
arrester rating, as follows:
 In Case A, the 22 KV MCOV surge arrester limits the overvoltage to 1.45 PU
for all considered penetration levels as compared to 1.75 PU in the case
without surge arrester.
 In Case B, the 24.4 KV MCOV surge arrester limits the overvoltage to 1.6 PU
for all considered penetration levels as compared to 1.75 PU in the case
without surge arrester.
 In Case C, the 29 KV MCOV surge arrester does not have a noticeable impact
on the overvoltage level because the overvoltage as compared to its MCOV
rating is not high enough to activate the surge arrester to absorb the
temporary overvoltage.

Pterra Report R149-16 11 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Table 2: Scenarios with Surge Arresters-Inverter#1
Temporary Over Voltage
Without 22 KV MCOV 24.4 KV MCOV 29 KV MCOV
Scenario DG/LD_L DG/LD_H Surge
Arrester PU of PU of PU of
PU PU PU
PU1 MCOV MCOV MCOV

16 65 1.38 1.38 1.25 1.38 1.13 1.38 0.95

17 90 1.62 1.44 1.30 1.56 1.27 1.62 1.11
18 100 1.71 1.45 1.31 1.58 1.29 1.7 1.17
19 105 1.75 1.45 1.31 1.59 1.30 1.75 1.20
20 65 1.38 1.38 1.25 1.38 1.13 1.38 0.95
21 90 1.62 1.44 1.30 1.56 1.27 1.62 1.11
22 100 1.7 1.45 1.31 1.58 1.29 1.7 1.17
23 105 1.75 1.45 1.31 1.59 1.30 1.75 1.20
1) Results from Table 3-1, it is included in this table for comparison purpose

Table 3-1 considers 1.38 PU as the maximum permissible overvoltage on high

voltage side of the substation transformer. Using the same criteria, Table 2 shows
the GFOV of 1.38 PU occurs when penetration level reaches 65%. This is the same
PV penetration limit found in scenarios without surge arresters (See Table 3-1).
Table 2 shows the surge arresters start making a difference in reducing the
overvoltage when the overvoltage is higher than 1.38 PU.
Assuming surge arresters are designed to protect the substation equipment, they
should be the first component on the substation to get damaged due to overvoltage
conditions. Figure 3-1 shows temporary overvoltage capability of EVP surge
arresters used in Case A, B, and C. Comparing the overvoltage levels summarized
in Table 2 with the curve shown in Figure 3-1, the following can be observed:
 22 KV MCOV and 24.4 KV MCOV surge arresters can safely operate for up to
15 seconds even when penetration level is 105%.
 With the same penetration level, 29 KV MCOV arrester can operate for more
than 100 seconds.
In the absence of rotating generators, it is extremely unlikely to have run-on
islanding time in the order of 15 seconds. Thus, considering the surge arresters
used in this study, the critical penetration limit could be set at 100%.
Note that TOV capability curve of surge arresters connected to delta side of a
substation transformer is needed prior to determining penetration limit based on
equipment damage. Simulations conducted in this study with station class surge
arresters indicate that arresters can safely operate for penetration level as high as
100%.

Pterra Report R149-16 12 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Figure 3-1: Temporary Overvoltage Capability of EVP Surge Arresters with Prior Duty

Scenario 24 through 28 (Table 3-3)

These scenarios are intended to study multi-inverter cases as well as potential
impacts of having a load with a high-quality factor.
The quality factor of the load can be defined as:

QL  QC
Q
PR
Where:
Q= Load’s quality factor
QL: Inductive power of the load
QC: Capacitive power of the load
PR: Resistive power of the load

It is generally more difficult to detect islanding condition when islanded load has a
high-quality factor and resonates close to the fundamental frequency. IEEE P1547.1
and IEEE Std.929, respectively, recommended islanding test procedure based on
load quality factors of unity and 2.5. The test requirement of Q<2.5 has been
determined to cover all reasonable distribution line configurations.

Pterra Report R149-16 13 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
Table 3-3 summarizes the simulation results which demonstrate the following:
 Although the high value of the Q indicated the potential impact on islanding
duration, changing load’s quality factor in studied scenarios does not change
the conclusion of the analysis in terms of potential overvoltage magnitude at
high voltage side of substation transformer.
 Having multiple inverters on an island does not necessarily degrade
protection performance of inverters from GFOV point of view.
The scenarios in Table 3-3 consider PV penetration fixed at 100% and the load was
assumed to be connected to low voltage side of the substation transformer (wye
side).
Scenario #24 and 25 were considered to evaluate load’s quality factor in the single-
inverter case. Inverter#2 was chosen for these scenarios because it was the only
inverter which tripped for 100% penetration in scenario#4, 9 and 14 (i.e. cases
without high-quality factor load).
Scenario#26 considered a multi-inverter case composed of inverter#1 and
inverter#2. The impact of load’s quality factor in the multi-inverter case was
illustrated in scenario #27 and #28. Maximum overvoltage observed in
scenario#26, #27 and #28 at high voltage side of the substation transformer is
about 1.2 PU.
Figure 3-2(a),3-2(b), and 3-2(c), respectively, show voltage at high voltage side of
substation transformer in scenario#4 with Inverter#1, scenario#4 with Inverter#2
and scenario#26 (both Inverter#1 and 2). Figure 3-3 plots outputs of both
inverters power output of scenario#26. Figure 3-2 and Figure 3-3 illustrate that
having multiple inverters on an island does not necessarily degrade protection
performance of inverters.
For scenarios in this assessment, overvoltage observed in scenario#26 with both of
inverter#1 and inverter#2 is much lower than overvoltage in scenario#4 with
inverter#1 only. If one of the inverters in the island can detect abnormal situation
and trip, island’s operating point drifts from balanced power zone, which in turn,
help other inverters to detect abnormality/islanding condition and expedite collapse
of the island. In addition, trip of one of the inverters reduce the ratio of in service
DG over the load; therefore, reduce potential overvoltage on delta side of the
transformer. The worst case is expected when none of the inverters can detect the
fault and subsequent islanding condition. Such a case is essentially similar to have
an island composed of a single type of inverter with ineffective protection scheme.
In this case, penetration level may need to be limited to 65% or 100% depending
on consideration of the surge arrester capability in handling the overvoltage before
3V0 or other protection is considered.

Pterra Report R149-16 14 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Table 3-3: Scenarios with High-Quality Factor Load/Multi-Inverters
Load Over Voltage
SC DG/LD_L Quality Inverters
Factor (Steady State)

24 100 1 Inverter#2 12
25 100 2.5 Inverter#2 12
26 100 N/A1 Inverter#1 + Inverter#2 1.23
27 100 1 Inverter#1 + Inverter#2 1.183
28 100 2.5 Inverter#1 + Inverter#2 1.113
1) Load with high-quality factor is not in service in scenario#26. The load is modeled similar to scenarios in
Table 3-1 or Scenario 1-15
2) Inverter trip following islanding; No overvoltage was observed.
3) In scenario#26,27 and 28, inverter#2 trip almost instantaneously after islanding. Inverter#1 trip within .6s
after inverter#2.

Pterra Report R149-16 15 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
a) Scenario 4 with Inverter#1, Permanent SLG fault initiates at t=.6 s; circuit breaker opens after 5
cycles

b) Scenario 4 with Inverter#2, Permanent SLG fault initiates at t=1.2 s; circuit breaker opens after 5
cycles

c) Scenario 26 with Inverter 1 & 2, Permanent SLG fault initiates at t=1.2 s; circuit breaker opens after
5 cycles

Figure 3-2: Per Unit Voltage at Delta side of Substation Transformer

Pterra Report R149-16 16 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Figure 3-3: Inverter’s Output Power in Scenario#26

Figure 3-4 show voltage at high voltage side of substation transformer of

scenario#25 where load quality factor is equal to 2.5. Comparison of Figure 3-2(b)
(i.e. inverter#2 without high-quality factor load) and Figure 3-4 (i.e. inverter#2
with load quality factor of 2.5) illustrates how load’s quality factor can increase
islanding duration. In this case, it increases from about 3 cycles to about 12 cycles.
Although the high value of the Q indicated the potential impact on islanding
duration, changing load’s quality factor in studied scenarios did not change the
conclusion of the analysis in terms of potential overvoltage magnitude at high
voltage side of substation transformer. In other words, no incremental impact was
observed on the magnitude of overvoltage at high voltage side of substation
transformer due to load with the high-quality factor. Comparing scenario# 24 and
#25 with scenario#4 illustrates that even with the addition of high-quality factor
load, inverter#2 is still able to detect the single-line-to-ground fault and trip.
Similarly, comparing scenario#27 and #28 with scenario#26 indicates that loads
quality factor does not significantly affect the GFOV magnitude in the multiple
inverters cases.

Pterra Report R149-16 17 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Figure 3-4: Voltage at Delta side of Substation Transformer in Scenario#25

Pterra Report R149-16 18 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Conclusions and Future Work

Conclusions
Time domain simulations were performed for three commercial inverter models to
illustrate potential GFOV on delta side of delta-wye substation transformers due to
penetration of inverter based DG on wye side of transformer feeding distribution
feeders.
The following is a summary of the findings of this study based on the observed
results of simulations.
 While inverters can potentially cause overvoltage on the delta side of the
substation transformer, some inverter designs can detect a single-line-to-
ground fault condition and trip instantaneously. Time domain simulation is a
potential tool for evaluating the fault detection capability of inverters for this
purpose.
 According to ANSI/IEEE C62.92, the GFOV for an effectively grounded
system is to be limited to 138%. This value can also be used to limit the
overvoltage for ungrounded systems. Simulation results indicate that
overvoltage on the delta side of the study substation transformer peaks at
1.38 PU as the PV/load ratio approaches 65%. At penetration levels below
65%, no overvoltage is observed. Two important notes relate to this
finding:
a. The calculation for the load should account for those connected to the
transmission side as well as the distribution side of the isolated
system.
b. Though this ratio seems close to the threshold proposed by National
Grid (i.e. 67%), there is possibility of under counting the load if only
the distribution side load is considered.
 The 65% penetration limit (based on 1.38 PU overvoltage threshold) can be
relaxed if:
a. Damage to equipment connected to delta side of the substation
transformer is the reason for requiring 3VO protection; and
b. Surge arresters connected to delta side of the substation transformer
are taken into account
Simulations conducted in this study with station class surge arresters
indicate that arresters can safely operate for penetration levels of up to
100%.
This report is not intended to impose pass-fail criteria for GFOV nor does it provide
basis for requiring 3VO protection due to GFOV. The intention is to add to the body
of knowledge on this topic and contribute to future pass-fail criteria that may be

Pterra Report R149-16 19 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
developed through a consensus-based process including various industry
stakeholders and taking into account all circuit configurations, a variety of inverter
manufacturers and the best available information on distribution system
requirements.

Future Work
This study is phase 1 of an overall effort to provide technical clarity on the issue of
GFOV. A future aspect of study is:
 Investigation of less expensive countermeasures for 3V0 requirement to
alleviate a potential GFOV issue. Such countermeasures could include, but
not limited to, protective relaying detection on the distribution side, detection
by the inverter, application of a grounding switch on distribution feeder, and
addition of a grounding bank on the transmission side.

Pterra Report R149-16 20 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation
Transformers
 Appendix A-Simulation Plots for Each
Scenario

4.2.1. Scenario#1

Delta Side Voltage with INVERTER#1_250 KW

Delta Side Voltage with Inverter#2_250 KW

Pterra Report R149-16 A-1 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with INVERTER#3_1 MW

4.2.2. Scenario#2

Delta Side Voltage with INVERTER#1_250 KW

Pterra Report R149-16 A-2 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with Inverter#2_250 KW

Delta Side Voltage with INVERTER#3_1 MW

Pterra Report R149-16 A-3 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 4.2.3. Scenario#3

Delta Side Voltage with INVERTER#1_250 KW

Delta Side Voltage with Inverter#2_250 KW

Pterra Report R149-16 A-4 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with INVERTER#3_1 MW

4.2.4. Scenario#4

Delta Side Voltage with INVERTER#1_250 KW

Pterra Report R149-16 A-5 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with Inverter#2_250 KW

Delta Side Voltage with INVERTER#3_1 MW

Pterra Report R149-16 A-6 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 4.2.5. Scenario#5

Delta Side Voltage with INVERTER#1_250 KW

Delta Side Voltage with Inverter#2_250 KW

Pterra Report R149-16 A-7 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with INVERTER#3_1 MW

4.2.6. Scenario#6

Delta Side Voltage with INVERTER#1_250 KW

Pterra Report R149-16 A-8 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with Inverter#2_250 KW

Delta Side Voltage with INVERTER#3_1 MW

Pterra Report R149-16 A-9 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 4.2.7. Scenario#7

Delta Side Voltage with INVERTER#1_250 KW

Delta Side Voltage with Inverter#2_250 KW

Pterra Report R149-16 A-10 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with INVERTER#3_1 MW

4.2.8. Scenario#8

Delta Side Voltage with INVERTER#1_250 KW

Pterra Report R149-16 A-11 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with Inverter#2_250 KW

Delta Side Voltage with INVERTER#3_1 MW

Pterra Report R149-16 A-12 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 4.2.9. Scenario#9

Delta Side Voltage with INVERTER#1_250 KW

Delta Side Voltage with Inverter#2_250 KW

Pterra Report R149-16 A-13 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with INVERTER#3_1 MW

4.2.10. Scenario#10

Delta Side Voltage with INVERTER#1_250 KW

Pterra Report R149-16 A-14 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with Inverter#2_250 KW

Delta Side Voltage with INVERTER#3_1 MW

Pterra Report R149-16 A-15 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 4.2.11. Scenario#11

Delta Side Voltage with INVERTER#1_250 KW

Delta Side Voltage with Inverter#2_250 KW

Pterra Report R149-16 A-16 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with INVERTER#3_1 MW

4.2.12. Scenario#12

Delta Side Voltage with INVERTER#1_250 KW

Pterra Report R149-16 A-17 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with Inverter#2_250 KW

Delta Side Voltage with INVERTER#3_1 MW

Pterra Report R149-16 A-18 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 4.2.13. Scenario#13

Delta Side Voltage with INVERTER#1_250 KW

Delta Side Voltage with Inverter#2_250 KW

Pterra Report R149-16 A-19 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with INVERTER#3_1 MW

4.2.14. Scenario#14

Delta Side Voltage with INVERTER#1_250 KW

Pterra Report R149-16 A-20 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with Inverter#2_250 KW

Delta Side Voltage with INVERTER#3_1 MW

Pterra Report R149-16 A-21 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 4.2.15. Scenario#15

Delta Side Voltage with INVERTER#1_250 KW

Delta Side Voltage with Inverter#2_250 KW

Pterra Report R149-16 A-22 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with INVERTER#3_1 MW

4.2.16. Scenario#16

Delta Side Voltage with Delta Side Voltage with 22KV MCOV Surge Arrester -AE 250 KW

Pterra Report R149-16 A-23 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with 24.4KV MCOV Surge Arrester -AE 250 KW

Delta Side Voltage with 29KV MCOV Surge Arrester -AE 250 KW

Pterra Report R149-16 A-24 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 4.2.17. Scenario#17

Delta Side Voltage with 22KV MCOV Surge Arrester -AE 250 KW

Delta Side Voltage with 24.4KV MCOV Surge Arrester -AE 250 KW

Pterra Report R149-16 A-25 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with 29KV MCOV Surge Arrester -AE 250 KW

4.2.18. Scenario#18

Delta Side Voltage with 22KV MCOV Surge Arrester -AE 250 KW

Pterra Report R149-16 A-26 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with 24.4KV MCOV Surge Arrester -AE 250 KW

Delta Side Voltage with 29KV MCOV Surge Arrester -AE 250 KW

Pterra Report R149-16 A-27 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 4.2.19. Scenario#19

Delta Side Voltage with 22KV MCOV Surge Arrester -AE 250 KW

Delta Side Voltage with 24.4KV MCOV Surge Arrester -AE 250 KW

Pterra Report R149-16 A-28 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with 29KV MCOV Surge Arrester -AE 250 KW

4.2.20. Scenario#20

Delta Side Voltage with 22KV MCOV Surge Arrester -AE 250 KW

Pterra Report R149-16 A-29 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with 24.4KV MCOV Surge Arrester -AE 250 KW

Delta Side Voltage with 29KV MCOV Surge Arrester -AE 250 KW

Pterra Report R149-16 A-30 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 4.2.21. Scenario#21

Delta Side Voltage with 22KV MCOV Surge Arrester -AE 250 KW

Delta Side Voltage with 24.4KV MCOV Surge Arrester -AE 250 KW

Pterra Report R149-16 A-31 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with 29KV MCOV Surge Arrester -AE 250 KW

4.2.22. Scenario#22

Delta Side Voltage with 22KV MCOV Surge Arrester -AE 250 KW

Pterra Report R149-16 A-32 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with 24.4KV MCOV Surge Arrester -AE 250 KW

Delta Side Voltage with 29KV MCOV Surge Arrester -AE 250 KW

Pterra Report R149-16 A-33 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 4.2.23. Scenario#23

Delta Side Voltage with 22KV MCOV Surge Arrester -AE 250 KW

Delta Side Voltage with 24.4KV MCOV Surge Arrester -AE 250 KW

Pterra Report R149-16 A-34 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Delta Side Voltage with 29KV MCOV Surge Arrester -AE 250 KW

4.2.24. Scenario#24

Delta Side Voltage with Inverter#2; Load Quality Factor Equal to 1

Pterra Report R149-16 A-35 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 4.2.25. Scenario#25

Delta Side Voltage with Inverter#2; Load Quality Factor Equal to 2.5

4.2.26. Scenario#26

Delta Side Voltage with Combination of Inverter#1 and Inverter#2; Load with High
Quality Factor is Out of Service

Pterra Report R149-16 A-36 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
Output of Inverter#1 and Inverter#2; Load with High Quality Factor is Out of Service

4.2.27. Scenario#27

Delta Side Voltage with Combination of Inverter#1 and Inverter#2; Load Quality
Factor Equal to 1

Pterra Report R149-16 A-37 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Output of Inverter#1 and Inverter#2; Load Quality Factor Equal to 1

4.2.28. Scenario#28

Delta Side Voltage with Combination of Inverter#1 and Inverter#2; Load Quality
Factor Equal to 2.5

Pterra Report R149-16 A-38 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer
 Output of Inverter#1 and Inverter#2; Load Quality Factor Equal to 2.5

Pterra Report R149-16 A-39 Pterra Consulting

Assessment of Inverter Induced Ground Fault
Overvoltage on Delta-Wye Substation Transformer