Вы находитесь на странице: 1из 71

IEEE Power & Energy Society TECHNICAL REPORT

July 2018 PES-TR68

Impact of Inverter Based


Generation on Bulk Power
System Dynamics and Short-
Circuit Performance

PREPARED BY THE
IEEE/NERC Task Force on Short-Circuit and System Performance
Impact of Inverter Based Generation

© IEEE 2013 The Institute of Electrical and Electronics Engineers, Inc.


No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written permission of the publisher.
THIS PAGE LEFT BLANK INTENTIONALLY

ii
TASK FORCE ON

Short-Circuit and System Performance Impact of Inverter Based


Generation

Co-Chairs: Kevin W. Jones and Pouyan Pourbeik

Vice-Chair/Secretary: Gary Kobet

Members

Aaron Berner
Normann Fischer
Fred Huang
Juergen Holbach
Michael Jensen
John O’Connor
Manish Patel
Michael Ropp
Jun Wen
Tao Yang

Other Contributors

Rich D. Bauer
Robert W. Cummings
Ryan D. Quint
Nicole Segal
Mohamed Osman
Michael Higginson
Randy Cunico
Jason Espinosa
Kamal Garg
Vahid Madani
Evangelos Farantatos
Mukesh Nagpal
Jim O’Brien
Pratap Mysore
Ritwik Chowdhury
Philip Baker

iii
Public

iv
KEYWORDS

Bulk power system

Directional element

Distance element

Fault

Frequency response

Inverter

Load shedding

Modeling

Negative sequence

Out-of-step

Power swing

Protective relay

Rate of change of frequency

Reactive power

Ride-through

Short-circuit

Synchronous machine

Underfrequency

Voltage control

v
CONTENTS

1. INTRODUCTION 1

2. Large System Impact Issues Related to Large Penetration of Inverter Based Resources 3
2.1 Voltage-Control/Reactive Support 3
2.2 Frequency Response and Control 6
2.3 Low-Short Circuit-Levels 11
2.4 Control and Grid Interactions 13
2.5 Planning Process 13
2.6 Modeling 14
2.7 Operations / Economic Issues 15
2.8 Other Issues 16
2.9 Recommendations 17

3. Protective Relay Issues Related to Large Penetration of Inverter Based Resources 19


3.1 Inverter-Based Fault Currents and Voltages 19
3.2 Relay Element Response to Inverter-Based Fault Currents and Voltages 22
3.3 Relay Scheme Selection Issues Caused by Inverter-Based Resources (IBR) 51
3.4 Short Circuit Study Issues Caused by IBR 52
3.5 NERC Protection and Control (PRC) Reliability Standards Issues Caused by IBR 59
3.6 Conclusions 60
3.7 Recommendations 61

vi
THIS PAGE LEFT BLANK INTENTIONALLY

Public

vii
1. INTRODUCTION
Electric power grids around the world are undergoing a significant change in generation mix,
from synchronous ac rotating machines to inverter-based resource technologies.
Conventional planning and operating practices are adapting to the benefits and challenges
these resources bring to reliability of the bulk power system (BPS). One issue that is
increasingly becoming apparent under higher penetrations of inverter-based resources such
as wind and solar photovoltaic (PV) generation, is a reduction in fault currents and short
circuit strength.

Synchronous machines are capable of supplying a significant amount of current to the BPS
during fault conditions. Many transmission system protection systems are designed with this
concept in mind, and rely on fault current to differentiate between fault conditions and
normal operating conditions. On the other hand, inverter-based resources consisting of
power electronics do not supply this type of fault current, since the power electronic
inverters will tightly control and limit current, and are often limited in their short circuit
contribution to between zero (e.g. the inverter is blocked for faults near or at its terminals)
to three times rated current output (which may be achievable on doubly-fed units and
inverters with significant short-term capability).

Naturally, as the penetration of inverter-based resources continues to increase on the BPS


and sub-transmission system, the levels of fault current and short circuit strength will
continue to drop. Low fault current and low short circuit strength conditions pose a number
of challenges to BPS reliability, including:

• Transmission protective relaying impacts due to a reduction in fault current


contribution; particularly, the identification and secure operation of protection
systems that rely on fault current magnitude for operation. Various protection
schemes and example systems are provided to illustrate the concepts and issues
identified. A review of fault study processes for updating and setting protective relays
and coordination of protection systems is also considered.
• Large system stability issues related to decreasing short circuit strength and the
issues that “weak grid” conditions can pose for larger systems with high penetration
of inverter-based resources. In particular, the decrease in short circuit ratio and grid
strength can impact inverter-based resource control system design as the lower
short-circuit strength at the point of interconnection of inverter base resources can
lead to potential control loop instability for the various control functions of the
inverter (e.g. closed-loop voltage control). These issues are discussed in this report.
• Inverter design philosophies under high penetration conditions also need to be
considered. As the grid pushes towards a 100% inverter-based resources

1
composition then under certain operating conditions, a review of the controls
interactions and requirements from a large system perspective is warranted.

In the end, the primary objectives of this report are:

1. lay out these various potential concerns and opportunities for inverter-based
resources bulk integration into the BPS and sub-transmission systems;
2. identify where current, or previous, broad industry activities have addressed or are
addressing the various issues, and;
3. identify where there may still exist gaps and thus the need to establish new IEEE
and/or North American Electric Reliability Corporation (NERC) technical task forces
and groups, with broad involvement from all stakeholders, including utilities,
reliability entities, vendors, researchers and consultants to investigate in more detail
the issues at hand and identify practical and achievable solutions for any challenges
identified.

NERC and IEEE signed a Memorandum of Understanding (MOU) in October 2016 1, seeking to
increase coordination and collaboration between NERC and IEEE. As part of this MOU,
technical leadership from NERC and IEEE have sought opportunities to engage in mutually
beneficial activities for the electric utility industry, and NERC and IEEE stakeholders. One of
the key topics prioritized by NERC and IEEE leadership is proactively exploring and better
understanding the implications of potential low fault current and short circuit strength
conditions on the BPS. The goal is to identify solutions to these potential conditions such that
when they are encountered by the electric utility industry in the future, utility engineers will
have the tools and capabilities to address these issues reliably. This report covers the various
aspects of low fault current conditions and how to accommodate a changing resource mix.

1 NERC and IEEE Memorandum of Understanding October 2016: https://www.ieee-


pes.org/images/files/pdf/technical-council/NERC_MoU_Signed_20161023.pdf

2
2. Large System Impact Issues Related to Large Penetration of
Inverter Based Resources
This section provides a high-level outline of the power system dynamic performance type
issues that would relate to massive integration of inverter-based resources (IBR). The focus
here is on inverter interfaced resources only, such as type 3 and 4 wind turbine generators,
photovoltaic (PV) generation, battery energy storage systems, and other similar technologies
(e.g. though not as prevalent presently, energy storage systems, and wave- and tidal-
generation systems also often employ inverter interfaced generators).

As indicated in the introduction, and given the scope of this short-lived task force effort, the
list provided here is intended to be a list of the key potential issues presently understood and
anticipated, with references to existing work and technical groups in this area and the
proposed or demonstrated solutions outlined by such existing or previous work. Any
perceived gaps that exist in investigating the issues discussed are outlined, with the
conclusions section identifying proposed new potential technical task forces and working
groups to address these gaps. The intent of this document is to identify such needs for further
in-depth technical work. Thus, this document is intentionally brief with a focus on providing
a quick overview of the key issues, where existing answers can be found for these issues, and
where further work is needed.

2.1 Voltage-Control/Reactive Support

As inverter-based resources displace synchronous conventional generation, there is likely to


be a degradation in the amount of reactive support and fast automatic-voltage control
available on the transmission system. There are two aspects to this issue:
1. Impact due to utility scale IBR displacing conventional generation
2. Impact due to massive amounts of distributed IBR (e.g. roof-top solar-PV etc.)
displacing conventional generation
Solutions: Modern IBR have voltage-control/reactive support capabilities similar to static
VAR compensators (STATCOMs) and thus can be designed to provide voltage-
control/reactive support. This can be further augmented by placing coordinated switched
shunt compensation or additional dynamic reactive support (e.g. SVC, STATCOM, or
synchronous condenser) at the point-of-interconnection, if deemed necessary. In the case of
massive amounts of utility scale or distributed IBR displacing conventional generation,
additional equipment may need to be implemented at the transmission level to supplement
the lost reactive support/voltage-control at the transmission level (e.g. SVC, STATCOM or
synchronous condensers).
Gaps:
1) Automatic voltage control: There is a potential gap in the uniform definition of
“automatic voltage control” and the steady-state and dynamic performance

3
characteristics that it entails. Considerations include determining the typical expected
performance from the perspective of utilities, and ascertaining the ability to have
automatic voltage response in a reasonably fast response time. For example, automatic
voltage control (AVR) systems on synchronous generation will have response times in
the hundreds of milliseconds to seconds time frame, and are tuned at each plant
specifically to the needs of the system for ensuring transient stability and other
consideration. However, in recent years with the advent of IBR, at least one utility in
North America has had the experience of seeing many utility scale PV plants coming
into its system that provide extremely slow voltage control (i.e. closed-loop response
that takes many tens of seconds to minutes). In this case, this is not due to any
limitations in the inverter technologies, but rather appears to have been a case where
the plant developers did not fully appreciate the need for fast closed-loop (i.e. seconds
time frame) voltage control and thus the plants were not built with a capability to
provide such control. Therefore, it may be useful, at the NERC level to provide some
high-level guide document on the expected performance of automatic voltage control
of generation resources 2. This may also be complemented by a survey of existing
practices and requirements by utilities across North America for plant level voltage
control.
2) Voltage control mode and Reactive Capability: IBR should be in automatic voltage
control mode (as currently required by the NERC Reliability Standard VAR-002-
4.1. 3standards, as opposed to controlling reactive output or in power factor control).
Automatic voltage control by IBR helps minimize voltage deviation/fluctuation issues
driven by the intermittent nature of these resources. Furthermore, FERC Order 827
requires a defined reactive power capability at the point of measurement (POM) within
a specified power factor range for newly interconnecting resources, as a minimum
requirement. That is, for the resource to be able to operate in a given range of power
factor from full-load to no-load, at the point of interconnection. This translates to a
triangular capability curve (see Figure 1). However, inverters are current limited
resources and are able to provide significantly more reactive current when active
current is not at maximum capability. Therefore, it is typically feasible to design the
actual inverters to provide a constant maximum/minimum reactive capability over the
operating range of the inverter for steady-state operation (green lines in Figure 1), and
to go to the maximum current rating of the device during faulted conditions
(“semicircular” capability curve in Figure 1). Therefore, IBR should not artificially limit
their reactive capability to meet only the FERC Order, but rather be limited by inverter
capability and plant-level performance requirements.
3) Defining “dynamic”: Dynamic reactive response generally means response that is
smooth and continuous within a given range, as opposed to large discrete steps effected

2 This is being presently worked on: NERC IRPTF Draft Reliability Guideline:
https://www.nerc.com/comm/OC_Reliability_Guidelines_DL/Inverter-
Based_Resource_Performance_Guideline_20180503.pdf
3 See NERC Reliability Standard VAR-002-4.1. Available:

http://www.nerc.net/standardsreports/standardssummary.aspx

4
by mechanically switched devices (e.g. mechanically switched capacitor banks). Note,
thyristor switched devices are certainly considered as dynamic, since they can switch
in and out in cycles. However, in general, the speed of response for IBR, as well as the
transition between different operating modes (e.g., local inverter control versus plant-
level control) should be explained in detail.
4) IBR performance during voltage sags: Inverters should not block or terminate output
for relatively small to medium dips in voltage, as seen at the terminals of the inverter
(i.e. momentary cessation).
5) Central vs local control: In some systems (e.g. ERCOT), regional/system wide voltage
control is currently being discussed. In these cases, voltage control is no longer a
localized task. Proper coordination across a wide area may become a concern.
6) Distributed IBR voltage control: There are still many concerns related to whether or
not distributed IBR (e.g. residential PV, IBR buried in the distribution system) should
have voltage control, and how that should be implemented and what the concerns are
in terms of impact both at the distribution system and at the BPS for large penetration
of DG. These issues are discussed within the current IEEE Std. 1547-2018 and the
California Rule 21 smart inverter working group, and are outside the scope of this
document.

Reactive Current Q
Actual Inverter Current Limit Inverter Capability

At 1 pu Voltage
P
Active Current

Constant Power Factor

Figure 1: Constant power factor (orange line), versus constant Qmax/Qmin (green line)
versus inverter current limit (red line).

In summary, several of the gaps/issues mentioned above are being discussed and addressed
within the NERC Inverter Based Resource Performance Task Force for BPS connected IBR 4.
The IEEE Std. 1547-2018 has addressed the issues related to IBR connected at the
distribution level. Nonetheless, NERC and IEEE are collaborating in the IEEE PES Wind and
Solar Power Plant Interconnection and Design Subcommittee and the Wind and Solar Plant

4NERC IRPTF Draft Reliability Guideline:


https://www.nerc.com/comm/OC_Reliability_Guidelines_DL/Inverter-
Based_Resource_Performance_Guideline_20180503.pdf
5
Interconnection Working Group, which report to the IEEE PES Energy Development & Power
Generation Committee. These ongoing efforts are expected to provide guidance on the
expectations for automatic voltage regulation dynamic performance and future
considerations such as centralized versus local voltage control and needed coordination
among such controls, for IBRs connected above distribution voltage but below BPS voltages. 5.
References and Further Reading:
1. V. Singhvi, J. C. Boemer, P. Pourbeik and A. Tuohy, “Impact of Fault-ride Through Capabilities
of Inverter-based Distributed PV on Voltage and Frequency Performance of the Bulk
System”, 5th Solar Integration Workshop: International Workshop on Integration of Solar
Power into Power Systems, October 2015.
2. V. Singhvi, P. Pourbeik, J. C. Boemer and A. Tuohy, “Impact of High Levels of Solar Generation
on Steady State and Dynamic Behavior of the Transmission System: Case Studies and
Lessons Learned”, 5th Solar Integration Workshop: International Workshop on Integration
of Solar Power into Power Systems, October 2015.
3. S. Achilles, N. Miller, E. Larsen and J. MacDowell, “Stable Renewable Plant Voltage and
Reactive Power Control”, NERC ERSTF, June 11-12, 2014
4. http://www.nerc.com/comm/Other/essntlrlbltysrvcstskfrcDL/VoltVarControl_Weaksys%2
0ERSTF%20JMM%20GE_0612.pdf
5. For distributed generation: IEEE Std 1547-2018, IEEE Standard for Interconnecting
Distributed Resources with Electric Power Systems
https://ieeexplore.ieee.org/document/8332112/
6. Variable Generation Interconnection Lessons Learned and Best Practices in the Western
Interconnection, WECC Variable Generation Interconnection Task Force, February 15, 2016
7. Essential Reliability Service Task Force: Measures Framework Report, NERC ERSTF, Atlanta,
GA, November 2015:
http://www.nerc.com/comm/Other/essntlrlbltysrvcstskfrcDL/ERSTF%20Framework%20
Report%20-%20Final.pdf
8. Distributed Energy Resources Connection Modeling and Reliability Considerations, NERC
DERTF, Atlanta, GA, February 2017:
http://www.nerc.com/comm/Other/essntlrlbltysrvcstskfrcDL/Distributed_Energy_Resourc
es_Report.pdf
9. NERC Inverter Based Resource Performance Task Force
https://www.nerc.com/comm/PC/InverterBased%20Resource%20Performance%20Task
%20Force%20IRPT/IRPTF_Scope_20170619.pdf
and https://www.nerc.com/comm/OC_Reliability_Guidelines_DL/Inverter-
Based_Resource_Performance_Guideline_20180503.pdf

2.2 Frequency Response and Control

As IBR displace synchronous conventional generation, there is likely to be a degradation in


both primary frequency response on the grid and inertial response. Thus, both the (i) nadir
of the frequency dip after a large loss of generation, and (ii) the rate of change of frequency
(RoCoF) at the inception of the event will increase. In islanded systems, small systems, and
parts of the system that are weakly connected to the rest of the bulk-electric system, these
issues may present reliability concerns. The RoCoF of a frequency event is determined by the
inertia of the system and a factor to account for the load damping characteristics of the

5 BPS voltage is considered to be 100 kV and above.

6
Interconnection. 6 NERC’s 2012 “Frequency Response Initiative Report –The Reliability Role
of Frequency Response” details the various aspects of frequency response [17]. The inertia of
the system is most influential during the arresting period of a frequency event, and the time
from pre-disturbance frequency (Value A) to the time of the frequency nadir (Point C) (See
Figure 2). It is during the arresting period that the combination of system inertia, load
damping, and the initial primary control response of resources act together to limit the
duration and magnitude of the frequency deviation from the pre-disturbance frequency to
the Point C nadir. It is essential that, for low frequency events, the decline in frequency be
arrested during this period to prevent activation of automatic under frequency load shedding
(UFLS) schemes in the Interconnection, which is the predominant reliability risk.
Deteriorating performance during the arresting period of an event increases that risk.

Figure 2: Frequency Response Periods during a Low Frequency Event 7

NERC performs an annual analysis of frequency response performance at the


Interconnection 8 level during both the arresting period and the stabilizing period of
frequency events.

Solutions: Modern IBR have the ability to provide frequency response. The ability of wind
turbine generation (type 3 and 4) to provide both primary frequency response (PFR) and
inertial based fast frequency response, is well established (see some of the references below,

6 For more detailed explanation of the relationships of inertia and load damping on RoCoF, see the
ERSTF Measures Framework report, November, 2015, pp. 28-29.
7 Source: NERC.

8 See the NERC 2017 “Frequency Response Annual Analysis:”

https://www.nerc.com/comm/oc/bal0031_supporting_documents_2017_dl/2017_fraa_final_20171113.pdf

7
and example in Figure 3). Equally, the ability of PV generation to provide primary frequency
response has also been demonstrated (see again some of the references below). Battery
energy storage, as another form of supplemental IBR, can also be effectively married with
wind/PV or independently deployed to provide both PFR and fast frequency response. Also,
in North America, as of the issuing of FERC Order 842 9 on February 15, 2018, all new
synchronous and non-synchronous resources (i.e. inverter-based generation) connecting to
the grid must maintain and operate a functioning governor or equivalent controls as a
precondition of interconnection. FERC Order No. 842 requires new generation to have the
capability, and to be able to respond when possible, to frequency excursion events when
frequency falls outside the deadband of +/- 0.036 Hz, and for the generation to adjust its
output in accordance to a 5% droop on the turbine MW base. This response must be timely
and sustained rather than injected for a short period and then withdrawn – i.e. it must be
essentially primary frequency response. Reserving generation headroom to provide
frequency response to under frequency events is not required. 10 Therefore, a response is only
expected for under-frequency events if the plant has been curtailed or dispatched below
maximum available power and has headroom to respond. However, resources should
respond to over frequency excursion events outside the deadband by reducing active power
output in accordance with the 5% droop provisions. A similar requirement has always been
in place in ERCOT 11. Furthermore, ROCOF is also monitored and a minimum level of inertial
response is maintained in the system in some islanded regions in North America. ERCOT
takes the approach of monitoring system inertia on-line and deploying additional resources
if inertia falls below the Critical Inertia System Operating Limit 12. An alternative approach is
taken by Hydro Québec for maintaining adequate inertia response in their system 13 (see also
pages 12-16 of reference [11]).

Another inter-national working group that is presently working on this subject is the CIGRE
JWG A1/C4.52 Wind generators and frequency-active power control of power systems 14.

https://elibrary.ferc.gov/IDMWS/common/downloadOpen.asp?downloadfile=20180215%2D3099%28326952
75%29%2Epdf&folder=15219837&fileid=14823757&trial=1
10 Wind and solar PV plants are typically operated at the maximum available incident wind/solar resource and

are therefore typically unable to respond to underfrequency events. However, it is possible that the plant may be
curtailed due to other reasons (e.g., transmission congestion), in which case it then can respond to an
underfrequency event – see Figure 2 for an example.
11 http://www.ercot.com/mktrules/guides/noperating/current
12

http://www.ercot.com/content/wcm/lists/144927/Inertia_Basic_Concepts_Impacts_On_ERCOT_v0.
pdf
13

http://www.hydroquebec.com/transenergie/fr/commerce/pdf/exigence_raccordement_fev_09_en.p
df
14 http://c4.cigre.org/WG-Area/JWG-A1-C4.52-Wind-generators-and-frequency-active-power-

control-of-power-systems

8
Gaps: The potential gaps in this area are commercial/economic and not
technological/technical. For example, as stated above, wind, PV, energy storage etc. can
easily provide primary frequency response and do so much more predictably than
conventional generation (see e.g. reference [1] or [2]). However, there are significant lost
opportunity costs in maintaining a curtailed output level on a wind or PV plant (or installing
energy storage) for the purpose of keeping MW in reserve for providing PFR. The economic
mechanism for this is not clear in many regions. Some regions (e.g. ERCOT and PJM) do have
a market mechanism for this. This gap was raised in reference [8] below and further
elaborated on, in significantly more detail, recently in reference [9]. It is likely an issue
outside of the scope of IEEE and NERC and to be addressed by ISOs and utilities.
Some of the other system specific issues to be considered and study on a case by case basis,
are:
1. The potential negative impact of low inertia in the system:
a. Will a high RoCoF affect generation resources or even trip generators if
experiencing high RoCoF?
b. Considerations to be given to the need for PFR and inertial-based fast-frequency
response (or other means of fast frequency response, e.g. from battery energy
storage systems) from IBR as IBR penetration increases thereby impacting
reliability, e.g. system frequency nadir, after the loss of the largest resource on
the system, approaching under-frequency load shedding (see section 3.2.9).
2. Considerations on the needed speed of response of the resource and how that truly
impacts frequency response.
3. Capability to respond versus the ability/market driven issues: To the extent that it is
technically feasible, IBR should have an active power-frequency control system with
the capability to provide primary frequency control if operating in a condition that
would allow for them to respond. Ensuring the capability to provide this response is
available and functional will help the balancing authority (either in a market or non-
market environment) in their procurement of sufficient levels of frequency responsive
reserves. Having the capability to respond is not the same as a requirement or
recommendation to actually respond. For example, most inverter-based resources
operate at the maximum power level possible (e.g., maximum output based on solar
irradiance or available wind speed at any given moment) and therefore do not have any
additional input power to generate extra electrical output power during an under-
frequency event. However, if the unit is dispatched at some level less than its maximum
available output (e.g., curtailed or market signal), it should have the capability to
respond in the upward direction for an under-frequency event (see Figure 3). Similarly,
all online generating units should be able to respond to an over-frequency event by
reducing output based on their control settings. This has now been addressed in North
America with the issuing of FERC Order 842, referred to above.

9
Figure 3: Example of a type 4 wind power plant responding to a system under-frequency
event (©IEEE 2017, from reference [16] below)

References and Further Reading:


1. S. Sharma, S-H. Huang and N. Sarma, “System Inertial Frequency Response estimation and
impact of renewable resources in ERCOT interconnection”, Proceedings of the IEEE PES GM,
July 2011
2. GE Report, Technology Capabilities for Fast Frequency Response, March 2017
https://www.aemo.com.au/-
/media/Files/Electricity/NEM/Security_and_Reliability/Reports/2017/2017-03-10-GE-
FFR-Advisory-Report-Final---2017-3-9.pdf

3. P. Pourbeik, S. Soni, A. Gaikwad and V. Chadliev, “Providing Primary Frequency Response


from Photovoltaic Power Plants”, CIGRE Symposium 2017, Dublin, Ireland, May 2017.
4. Y.C. Zhang, V. Gevorgian, E. Ela, V. Singhvi and P. Pourbeik, “Role of Wind Power in Primary
Frequency Response of an Interconnection”, Proceedings of the International Workshop on
Large-Scale Integration of Wind Power Into Power Systems as Well as on Transmission
Networks for Offshore Wind Power Plants, London, United Kingdom, October 22–24, 2013
5. N. W. Miller, M. Shao and S. Venkataraman, “Impact of Frequency Responsive Wind Plant
Controls on Grid Performance”, CIGRE Session 2012, paper C4-113
6. N. W. Miller, M. Shao, S. Pajic, R. D’Aquila and K. Clark, “Frequency Response of the US
Eastern Interconnection under Conditions of High Wind Generation”, CIGRE Session 2014,
paper C4-108
7. NREL/EPRI Report, “Active Power Controls from Wind Power: Bridging the Gaps”, January
2014, Technical Report, NREL/TP-5D00-60574,
http://www.nrel.gov/docs/fy14osti/60574.pdf
8. IEEE PES Technical Report, PES-13, Interconnected Power System Response to Generation
Governing: Present Practice and Outstanding Concerns, May 2007
http://resourcecenter.ieee-pes.org/pes/product/technical-publications/PESTR13
9. IEEE PES Technical Report, PES-TR-24, Measurement, Monitoring, and Reliability Issues
Related to Primary Governing Frequency Response, sponsored by the IEEE PSDP Committee.
http://resourcecenter.ieee-pes.org/pes/product/technical-
publications/PESTECRPTGS00001

10
10. NERC Reliability Guideline: Primary Frequency Control, NERC Resources Subcommittee,
Atlanta, GA, December 2015:
http://www.nerc.com/comm/OC_Reliability_Guidelines_DL/Primary_Frequency_Control_fin
al.pdf.
11. Technical Brief on ERS Framework Measures 1,2, and 4: Forward Looking Frequency
Analysis
https://www.nerc.com/comm/Other/essntlrlbltysrvcstskfrcDL/ERS_Forward_Measures_12
4_Tech_Brief_03292018_Final.pdf
12. ERS Framework Measures 1, 2 & 4 – Historical Frequency Analysis Historical Frequency
Trends Technical Brief
https://www.nerc.com/comm/Other/essntlrlbltysrvcstskfrcDL/Item_6b.ii._ERS_Historical_
%20Measures_124%20_Technical%20Brief_DRAFT_%2020171107.pdf
13. ERS Disposition of Measure 3 : Synchronous Inertial Response at the BA Level
https://www.nerc.com/comm/Other/essntlrlbltysrvcstskfrcDL/ERS_Disposition_of_Measur
e_3_03292018_Final.pdf
14. ERS White Paper on Sufficiency Guidelines -
https://www.nerc.com/comm/Other/essntlrlbltysrvcstskfrcDL/ERSWG_Sufficiency_Guideli
ne_Report.pdf
15. ERSTF Final Measures Framework Report -
https://www.nerc.com/comm/Other/essntlrlbltysrvcstskfrcDL/ERSTF%20Framework%20
Report%20-%20Final.pdf
16. P. Pourbeik, J. Sanchez-Gasca, J. Senthil, J. Weber, P. Zadehkhost, Y. Kazachkov, S. Tacke and J.
Wen, "Generic Dynamic Models for Modeling Wind Power Plants and other Renewable
Technologies in Large Scale Power System Studies", IEEE Trans. on Energy Conversion,
September 2017, https://ieeexplore.ieee.org/document/7782402/
17. NERC “Frequency Response Initiative Report - The Reliability Role of Frequency Response,”
October 2012:
https://www.nerc.com/comm/OC/BAL0031_Supporting_Documents_2017_DL/FRI_Report_
10-30-12_Master_w-appendices.pdf

2.3 Low-Short Circuit-Levels

As the amount of IBR increases and displaces synchronous generation, the amount of short-
circuit capability available on the grid decreases. Also, in many regions the wind and solar
resource is most attractive in remote regions which are electrically quite remote from load
and generation centers. Thus, the IBRs are being interconnected to the sub-transmission
systems and the BPS where the short-circuit strength is already quite low. This presents
several challenges:
a) Control loop stability of the inverters: Where the short-circuit strength of the system
is quite low, this can present a challenge for keeping the faster inner current control
loops of the inverter stable, as well as closed-loop voltage control and the phase-lock-
loop dynamics. This is documented in detail in references [1] and [2].
b) Ride-Through and Islanding: The ability of the IBR to ride-through low/high voltage
transients, and to properly disconnect under unintentional islanding scenarios may
be more challenging and require more careful design in the case of low-short circuit
conditions. See references [1] and [2].
c) Appropriate Modeling: There is a need for criteria to understand at what level such
issues may arise, since detailed 3-phase modeling, together with proprietary vendor
11
models, is typically needed at an initial phase for the study and tuning of controls in
these cases. Reference [3] offers some criteria for this.
d) Model and Tool limitations: Better clarity is needed on the applicability of different
tools and modeling methods for low-short circuit systems. For example, it is well
understood that “generic” positive-sequence stability models are not able to capture
the details of the inverter dynamics (e.g. phase-lock loop dynamics and inner-current
control loop dynamics) for a detailed look at local controls in weak-grid conditions.
However, with proper enhancements to such models it may still be feasible to use
them for system wide analysis, while reserving detailed local studies for more
detailed models and electromagnetic transient (EMT) type tools. Thus, more clarity
is needed on the boundary of applicability of different modeling tools and techniques.
A clear quantitative metric for identifying what constitutes a low-short circuit region
of the power system, as opposed to a single substation, does not presently exist, and
may be very difficult to achieve. Reference [3] discusses various options for
calculating the short-circuit ratio for not just one plant, but a group of plants in a
region. However, these issues may need some further investigation.
e) Control Interactions: See next issue Section 2.4, below.
Solutions: These types of issues are often site-specific. There are several possible solutions
in these cases, including (i) tuning the controls for the expected low-short circuit system
conditions, (ii) augmenting the controls with vendors specific weak grid options, and (iii)
providing other equipment in the vicinity of the IBR plants to improve system short-circuit
strength, such as synchronous condensers.
New control approaches, currently under research and development, may also help in low-
short circuit conditions, such as having the inverter-based generation develop its own
internal frequency-signal (i.e. grid-forming, instead of grid-following inverters), or other
innovative designs that may be developed in the future.
Gaps: As indicated above the main gaps in this area are around: (i) better criteria for
quantifying what is a low-short circuit region of the system, and (ii) greater clarity on the
limitations and boundaries of applicability of the various modeling tools and techniques.
References and Further Reading:
1. Reliability Guideline: Integrating Inverter-Based Resources into Low Short Circuit Strength
Systems, NERC System Analysis and Modeling Subcommittee, Atlanta, GA, December 2017:
http://www.nerc.com/comm/PC_Reliability_Guidelines_DL/Item_4a._Integrating%20_Inver
ter-Based_Resources_into_Low_Short_Circuit_Strength_Systems_-_2017-11-08-FINAL.pdf
2. CIGRE WG B4.62 Connection of Wind Farms to Weak AC Networks, 2016, https://e-
cigre.org/publication/671-connection-of-wind-farms-to-weak-ac-networks
3. Essential Reliability Service Task Force: Measures Framework Report, NERC ERSTF, Atlanta,
GA, November 2015:
http://www.nerc.com/comm/Other/essntlrlbltysrvcstskfrcDL/ERSTF%20Framework%20
Report%20-%20Final.pdf

12
2.4 Control and Grid Interactions

As the proliferation of power electronic based generation (IBR) and transmission (FACTS 15
and HVDC 16) increases, there is equally a greater potential likelihood of control interactions
between these devices both among the resources and between the resource and the active
transmission elements. Some examples include:

a) Control interactions between adjacent wind/PV power plants, such as the voltage
control loops fighting each other or not properly sharing vars. This is often solved by
introducing reactive current droop in the voltage control loop at each plant to ensure
that the plants in close electrical proximity to each other do not control the same
electrical point and properly share vars.
b) Control interactions between a large wind/PV plant and a nearby HVDC link. This
may require coordination and detailed studies between the two facilities.
c) Self-excitation, subsynchronous resonance between a type 3 wind power plant and
nearby series capacitor compensation on the transmission system (see references
below). This can be solved by controls modifications in the wind power plant and
protection on the series capacitor [2].
Solutions: Vendors have proposed and implemented solutions in some cases (e.g. reference
[2] below). These problems are case/site specific and require extensive simulation and
design work, and highly proprietary vendor specific 3-phase models. Thus, solutions have to
be developed on a case by case basis.
References and Further Reading:
1. J. Adams, “ERCOT Experience Screening for Sub-Synchronous Control Interaction in the
Vicinity of Series Capacitor Banks”, Proceedings of the IEEE PES GM 2012
2. E. Larsen, “Wind Generators and Series Compensated AC Transmission Lines”, Proceedings
of the IEEE PES GM 2012.
3. P. Pourbeik, R. J. Koessler, D. Dickmander and W. Wong, “Integration of Large Wind Farms
into Utility Grids (Part 2 - Performance Issues)”, Proceedings of IEEE PES General Meeting,
July 2003.

2.5 Planning Process

For the purpose of planning a power system with significant penetration of renewable
resources (wind and PV) there are challenges associated with the scenarios to be studied.
With conventional generation, the historic approach to planning has been to look at the
summer-peak and winter-peak cases (depending on whether the system is summer or winter
peaking, or perhaps both) and a light-load scenario. With increasing penetration of wind and
PV generation, several questions arise:

15 Flexible AC Transmission Systems


16 High-Voltage Direct-Current
13
a) What scenarios should be studied? The renewable resources are not likely to peak
with load and so it is more than likely that simply looking at just a peak and off-peak
scenario is not likely to capture all potential issues of concern.
b) Perhaps a range of scenarios need to be considered depending on the issues being
studied, e.g. if concerned with frequency regulation, a shoulder-case with maximum
wind/PV penetration is likely a more onerous case to study, then a peak-load
condition with minimal wind/PV on-line.
Gaps: There is perhaps not a single reference available to give concise guidance on these
issues and so NERC/IEEE should consider the need for such a technical task force group to
develop a guide document on this subject. The North American Transmission Forum (NATF)
document on Distributed Energy Resource Modeling and Study Practices, which is intended
to be made public soon, does offer some guidance on these issues.
References:
1. Integration of Variable Generation Task Force IVGTF , Summary and Recommendations of 12
Tasks, June 2015,
https://www.nerc.com/comm/PC/Integration%20of%20Variable%20Generation%20Task
%20Force%20I1/IVGTF%20Summary%20and%20Recommendation%20Report_Final.pdf
2. CAISO Study Plan for 2018-2019 Planning Cycle.
http://www.caiso.com/Documents/Final2018-2019StudyPlan.pdf
3. North American Transmission Forum document on Distributed Energy Resource Modeling
and Study Practices (to be made public soon on http://www.natf.net).

2.6 Modeling

One of the key issues with IBR is the ability to model these resources for large scale power
system stability and power flow analysis using public, standard models in commercially
available power system software tools.
Solution: A suite of 2nd generation generic models has been developed in North America,
primarily within the Western Electricity Coordinating Council (WECC), ) for this purpose,
which are being increasingly deployed and successfully utilized. All utilities and vendors are
encouraged to start using these models. The group 17 continues to make efforts to improve
the models by looking at adding additional modules for better modeling in low-short circuit-
strength conditions, fast frequency response options, weak-grid options and more.
Gaps: The industry is not yet unified on the use of these 2nd generation generic models for
interconnection-wide modeling purposes, and transmission planners and operations
engineers still handle many user-defined models. NERC may wish to continue advocating the
use of these models for interconnection-wide base cases, with user-defined models
augmenting detailed local studies, since these public models are transparent and easily
shared across various commercial software platforms.
References and Further Reading:
1. WECC 2nd Generation WTG Model Specifications: https://www.wecc.biz/Reliability/WECC-
Second-Generation-Wind-Turbine-Models-012314.pdf

17 WECC’s Renewable Energy Modeling Task Force

14
2. WECC PV Model Specifications: https://www.wecc.biz/Reliability/WECC-Solar-PV-Dynamic-
Model-Specification-September-2012.pdf
3. WECC PV Modeling Guide:
https://www.wecc.biz/Reliability/WECC%20Solar%20Plant%20Dynamic%20Modeling%2
0Guidelines.pdf
4. WECC WTG Modeling Guide:
https://www.wecc.biz/Reliability/WECC%20Wind%20Plant%20Dynamic%20Modeling%2
0Guidelines.pdf
5. Model User Guide for Generic Renewable Energy System Models:
http://www.epri.com/abstracts/Pages/ProductAbstract.aspx?ProductId=00000000300200
6525
6. IEC TC88 WG27 International Standards on Generic Models for Wind Power Plants (still in
progress) – this effort is similar to the models developed in North America, however, the IEC
models have some additional features (see reference [8] below).
7. Ö. Göksu, P. Sørensen, J. Fortmann, A. Morales, S. Weigel, P. Pourbeik, “Compatibility of IEC
61400-27-1 Ed 1 and WECC 2nd Generation Wind Turbine Models”, Conference: 15th
International Workshop on Large-Scale Integration of Wind Power into Power Systems as
well as on Transmission Networks for Offshore Wind Power Plants, November 2016.
8. P. Pourbeik, J. Sanchez-Gasca, J. Senthil, J. Weber, P. Zadehkhost, Y. Kazachkov, S. Tacke and J.
Wen, "Generic Dynamic Models for Modeling Wind Power Plants and other Renewable
Technologies in Large Scale Power System Studies", IEEE Trans. on Energy Conversion,
September 2017.

2.7 Operations / Economic Issues

From an operational and economic standpoint many factors need to be considered for the
purpose of operation planning studies and system operations. This is a complex area of
technical details, outside of the scope of this document. However, a brief mention will be
made here of some of the salient points. These may include:

a) Predicting hour ahead and day ahead availability of the inverter based renewable
resources (wind/PV)
b) Managing resource variability:
i. Reserve allocation
ii. Having enough system flexibility (e.g. fast startup units etc.) to manage the
maximum expected ramp rates from the variable resources, both up and
down.
iii. Preparedness and managing extreme weather events and other natural
events (e.g. (i) severe weather patterns that might cause sudden shutting
down and ramping up of large portions of wind generation, (ii) icing of wind
turbines, (iii) storms resulting in rapid cloud movement that affects PV
generation, (iv) solar eclipse effects on PV generation, etc.)
c) Properly accounting for capacity and capacity credit for wind and potential gaps and
references in this subject matter are not discussed here, as the operational and
economic issues are outside of the scope of this TF.
15
Consideration might be given to starting a separate technical working group, specifically on
these issues, if deemed necessary by IEEE PES and NERC.

2.8 Other Issues

There are likely to be other issues/challenges that will emerge as experience with the massive
integration of IBR continues in North America, and other systems around the world. Here we
list a few other potential issues that do not necessarily fall precisely in any of the categories
listed above:

a) Issues related to low-short circuit conditions, due to large influx of inverter-based


generation, which may lead to consequence for voltage flicker (power quality) in the
vicinity of fluctuating installations, e.g., arc furnaces, etc. Also, consideration should
be given to the potential for increasing harmonics.
b) Issues with energization of large power transformers under weak grid conditions.
When energized, large power transformers will draw large amounts of inrush current
for a short duration of time. In weak-grid conditions (e.g. parts of the network
dominated by IBR) this may result in difficulty in energizing such power
transformers.
c) National Policy/Political Issues (months to years time horizon) – Significant long-
term reduction of wind/solar generation due to expiring tax credits/incentives or
other economic or national policy factors. This is simply mentioned here to highlight
the fact that policy can have significant influence and impact on the deployment of
IBR. Nonetheless, such policy issues are clearly outside of the scope of any NERC or
IEEE task force and working groups.
d) Another issue is that of system restoration in light of large penetration of inverter-
based generation. This is presently being studied by an existing IEEE Task Force
under the Power System Dynamic Performance Committee – the Task Force on Power
System Restoration with Renewable Energy Sources 18.
Some work has been, and is being done, by CIGRE Study Committee C4 – System Technical
Performance on some of the issues related to power quality, harmonics and transformer
energization19. However, more work may be needed to further look at such issues within the
context of the North American system.
Glossary:
Bulk Power System (BPS) 20: (A) facilities and control systems necessary for operating an
interconnected electric energy transmission network (or any portion thereof); and (B)
electric energy from generation facilities needed to maintain transmission system reliability.
The term does not include facilities used in the local distribution of electric energy.

18 http://sites.ieee.org/pes-psdp/power-system-stability-controls-subcommittee/
19 See http://c4.cigre.org/WG-Area and https://e-cigre.org/
20 BPS voltage is considered to be 100 kV and above.

16
Inverter-based resource (IBR): An inverter-based resource is any source of electrical power
that is interface to the grid through the use of a power electronic converter interface. These
include doubly-fed asynchronous generators, full-converter connected generators, photo-
voltaic generation, battery-energy systems, and other similar technologies.

2.9 Recommendations

In brief, if we consider the various headings under section 2, the following summarizes how
the various gaps are presently being addressed, or need to be addressed by potentially
creating new technical task forces within IEEE/NERC. Links within IEEE on any new work
suggested below, should be established with the IEEE PES Power System Dynamic
Performance Committee.

1. Voltage-Control/Reactive Support: All the gaps identified under section 2.1 are
presently being addressed by a NERC Inverter-Based Resource Performance Task
Force(IRPTF). At the timing of writing this report the NERC IRPTF has drafted a
Reliability Guideline: BPS-Connected Inverter-Based Resource Performance, which is
currently out for industry comment. This guideline is expected to be completed by the
end of 2018.
2. Frequency Response and Control: Within North America, the gap identified in
section 2.2 with respect to having frequency response capability for IBR has been
addressed by FERC Order 842 issued on February 15, 2018. The other gaps
pertaining to addressing the economic impact of requiring frequency response
capability from IBR is outside of the scope of IEEE and NERC, and is something that
needs to be addressed by market mechanisms, or other means, within each region.
3. Low-Short Circuit Levels: As indicated in section 2.3, the main gaps in this area are
around: (i) better quantifying criteria around what is a low-short circuit region of the
system, and (ii) greater clarity on the limitations and boundaries of applicability of
the various modeling tools and techniques.
4. Control and Grid Interactions: These tend to be quite specific on a case-by-case
basis and required detailed vendor-specific proprietary 3-phase models. There is a
PSRC JTF2 (“Protection issues related to subsynchronous system oscillations”)
looking at these issues as they pertain to protection – that TF should perhaps also
coordinate its efforts with the IEEE PES PSDP Committee to see if further work is
needed on the large system issues related to control interactions.
5. Planning Process: There is perhaps not a single reference available to give concise
guidance on how to address the various issues such as what scenarios should be
studied for wind/PV etc. Presently each region, and utility, addressed the planning
process, in light of large penetration of IBR, in their own way. There is perhaps room
for NERC and IEEE to consider the need for such a technical task force group to
develop a guide document on this subject, or at least provide a comprehensive
summary of what is being done by various regions.

17
6. Modeling: The development of public and open (sometimes referred to as “generic”)
stability models for IBR has been undertaken quite extensively by many groups, most
notable WECC’s Renewable Energy Modeling Task Force and the International
Electrotechnical Commission (IEC). Many of these models have already been
implemented and tested in the major commercial software platforms used in North
America, and in some tools in Europe. The models are being used, and NERC in
encouraging greater use of these models throughout North America. Furthermore,
the groups involved continue to update and improve the models. Therefore, it is not
felt that any new groups need to be formed in this area.
7. Operational/Economic Issues: Again, it is felt that most such matters fall outside of
the scope of NERC and IEEE. Some of the technical aspects in this arena may already
be under consideration by the IEEE PES Power System Operations, Planning &
Economics Committee.

18
3. Protective Relay Issues Related to Large Penetration of
Inverter Based Resources
The addition of IBR to the BPS is presenting ever increasing challenges to the way protective
relays are applied on the BPS. IBR produce less fault current than conventional synchronous
generation. The fault currents produced are of low magnitude. The IBR fault current does
not contain sufficient levels of negative or zero sequence quantities for proper detection and
determination of directionality required by directional relaying. Lower fault current levels
(primary side) can result in lower relay sensitivity for fault or polarity detection at some
locations on the BPS. Lack of adequate negative and zero sequence quantities may result in
incorrect directional declaration of directional elements polarized by these quantities. Using
proper settings and providing additional supervision are critical for secure and dependable
protective relaying. The alternatives are for transmission owners to modify some existing
relay settings, where possible, or in some instances, protective relays may have to be
upgraded or replaced in order to take advantage of newer, more flexible technologies. Power
system relaying practices and applications are covered in many relay application guides and
text books or IEEE guides beyond the intent of this report. The details of protective relay
issues caused by IBR are discussed in the following sections.

Note the following working groups of the IEEE/PES Power System Relaying and Control
Committee (PSRC) are mentioned throughout this section:

• IEEE PSRC WG C24 (Modification of Commercial Fault Calculation Programs for Wind
Turbine Generators)
• IEEE PSRC WG C25 (Protection of Wind Electric Plants)
• IEEE PSRC WG C32 (Protection Challenges and Practices for interconnecting solar or
other inverter-based generation to utility transmission systems)
• IEEE PSRC WG D29 (Tutorial on Setting Impedance-Based Power Swing Blocking and
Out-Of-Step Tripping Functions on Transmission Lines)
• IEEE PSRC WG D38 (Impact of High SIR on Distance Relaying)

There are also other international working groups looking at these issues, such as the CIGRE
JWG B5/C4.61 Impact of Low Inertia Network on Protection and Control 21.

3.1 Inverter-Based Fault Currents and Voltages

3.1.1 Fault Current

21http://c4.cigre.org/WG-Area/JWG-B5-C4.61-Impact-of-Low-Inertia-Network-on-Protection-and-
Control

19
Unlike synchronous generators, the short circuit current contribution from the inverter-
based generation is usually limited to 100-120% of the rated load current 22.

The inverter phase current limitation is a cutoff value engineered mainly to protect the power
electronic equipment within the inverters, according to thermal limits, and varies by
manufacturer design. IBR fault currents typically consist of two components, an initial
inception current with a duration from a quarter to two cycles and a magnitude that can be
as high as 2.5 pu 23, and a longer duration low magnitude fault current. The low magnitude
fault current is limited by the inverter controls and a relatively short duration (100 ms) to
protect the IBR power electronics. This reduced output is limited to a value of 1.1-1.2 per
unit of nominal current. The magnitude of fault current contribution varies depending on the
fault-type, fault duration and pre-fault operating condition. The fault current power factor
can be capacitive, inductive, or resistive based on inverter control and system conditions. An
IBR typically provides negligible negative- or zero-sequence current. The angle of fault
current (with respect to voltage) would vary depending on regional or interconnected power
system requirements, such as ride-through during faults and reactive support requirements
during system disturbances for which the inverter controls are required to support.

Another factor in fault current contribution from the IBR is the control strategy implemented
in the inverter. Inverters could be programmed such that they operate as Current Regulated–
Voltage Source Inverters (CR-VSI) or Current Regulated–Current Source Inverters (CR-CSI).
It should be noted, however, that in the case of large utility scale inverter-based generation,
CR-VSIs are by far the dominant type. The mode of operation flexibility varies greatly from
one manufacturer to another. Regardless of the control strategy, it is widely observed that
the inverter acts as a constant current source during faults. A simplified positive, negative
and zero sequence equivalent of the IBR is shown in Figure 4. The model shown in this figure
is suitable for steady-state fault level calculations but may not meet all protection study
requirements.

22 Type 3 wind turbine generators (WTG), where the electrical machine’s stator is directly connected
to the grid, and depending on the type of controls (e.g. employing active-crow bar protection) may
contribute close to 2pu or so, while full-converter interfaced resources (i.e., Type 4 wind and PV) are
typically limited to zero (blocking) to 1.2 pu (converter current limit). See the IEEE Joint Working
Group Report, Fault Current Contributions from Wind Plants, 2013 for more details (http://www.pes-
psrc.org/kb/published/reports/Fault%20Current%20Contributions%20from%20Wind%20Plants.p
df).
23 Again, the higher values are more typical of type 3 wind turbines, that employ features such as

active-crowbar protection on the rotor circuit.

20
S2 S0

Status of Switch S2
depends on control

I1 I2

Positive Sequence Negative Sequence Zero Sequence

Figure 4 Positive, Negative and Zero Sequence equivalent of IBR

The lack of negative sequence current contribution poses another challenge to traditional
protection schemes.

3.1.2 Fault Voltages

As noted in Section 3.1.1, inverters are typically controlled to operate as positive-sequence


current sources regulated to inject a particular real and reactive power. When the inverter
terminal voltage drops due to a fault, the inverter will increase its output current to attempt
to regulate back to its P-Q setpoints until it reaches its manufacturer-specified maximum
allowed operating current. At that current level, the inverters will become more or less a
constant current source, and during the fault event they will remain in that mode until and
unless the voltage recovers to a level at which the inverters can produce their commanded P-
Q outputs at a current lower than the maximum.

Inverters programmed to maintain a fixed voltage schedule at the inverter terminals will
drop out of voltage regulation mode once the inverter terminal voltage drops sufficiently that
the inverters’ output current reaches its maximum allowed value. Thus, as long as the
inverters’ terminal voltage drops to a sufficiently low level that the current required to reach
the inverters’ commanded P-Q output is larger than the inverters’ maximum current, that
maximum current magnitude (but not necessarily the phase angle) will be the same
regardless of control mode.

Most inverter controllers are designed to regulate positive-sequence current and suppress
negative-sequence current. This design causes the inverters to have an effective or apparent
negative-sequence impedance that is significantly larger than the apparent positive-sequence
impedance.

21
3.2 Relay Element Response to Inverter-Based Fault Currents and
Voltages

This section addresses the issue of how different protective relay elements perform when the
bulk of the current during a power system disturbance or fault is supplied from an inverter-
based power source.

In order for a protective relay to correctly detect and isolate faults in the power system, the
characteristic of inverter driven fault currents are described. The electric industry is
continuing discussions surrounding accurate modeling alternatives of IBR for proper set
points with consideration for security and dependability. Some basic differences in the fault
current characteristics are observed compared to traditional non-inverter-based generation
(i.e., impact on the directional (polarizing) elements).

3.2.1 Relay Directional Elements

Directional elements have to determine the direction of the fault location relative to the
location of the relevant instrument transformers. The quantities used for the directional
determination may vary based on the fault location and fault type. Modern numerical relays
may have complex evaluation logic to find the best fit for a certain fault type and location. The
following subsections explain commonly used directional elements and how they are affected
by inverter-based resources.

3.2.1.1 Negative Sequence Directional

The arrival of numerical/digital relays has contributed to widespread applications of


negative sequence relaying to identify and determine the direction of unbalanced faults. The
negative sequence current (commonly referred to as the operating quantity) is compared
with the negative sequence voltage (the reference or polarizing quantity) to decide the fault
direction. Figure 5 shows the connection of the three sequence networks for a line-to-ground
fault in a networked line which is connected to conventional generators at both terminals. It
can be observed from the figure that the negative sequence voltage and the negative sequence
current I2C (or I2V) is determined by the negative sequence impedance of the source behind
the relay Z2C (or Z2V).

22
Z1C Z1TC Z1L*m Z1L*(1-m) Z1TV Z1V

Positive
Sequence I1C I1V
C V

Z2TC Z2L*m Z2L*(1-m) Z2TV


Negative
Sequence
I2C I2V Rfault
Z2C Z2V

Z0TC Z0L*m Z0L*(1-m) Z0TV


Zero
Sequence
Z0C I0C I0V Z0V

Inverter System Synchronous System

Figure 5 Sequence components for single line to ground fault

In a transmission system with conventional rotating machine sources, the negative sequence
impedances are highly inductive and thus the negative sequence current leads the
corresponding negative sequence polarized voltage by 180°- Z1ANG (where Z1ANG is the
positive sequence line angle) for a forward fault. Likewise, negative-sequence current lags
negative-sequence voltage by Z1ANG. Generally speaking, the protective relays use this
angular relationship between negative sequence quantities with some safety margin to
declare the unbalanced fault direction. However, reliable evaluation requires that the
magnitudes of the negative sequence currents and voltages are above a certain, normally
settable limit (i.e., if the magnitudes are too low, the results are not reliable).

The short circuit response of IBR, including the negative sequence current output, differs
significantly from the short circuit response of a conventional rotating generator. Unlike
conventional generators, the IBR does not have universal short-circuit response
characteristics, rather the inverter response is based on specific and often proprietary
inverter control system designs. To limit thermal overloads of power electronics, inverter
control systems not only limit the maximum short-circuit current but may also partially or
fully suppress negative sequence current output. Thus, the negative sequence current from
an inverter-based generating resource may have either a magnitude too low or have an
undefined phase angle relationship to the negative sequence voltage to use reliably in
relaying applications, including negative sequence directional relaying.

23
Reference [1] presents actual relay records showing examples of misoperation of the
negative-sequence directional relay in presence of inverter-based sources. Reference [2]
documents misoperation events of negative sequence directional relays within a wind park
with Type IV WTGs. Reference [3] analyzes the substantially different negative-sequence
fault current characteristics of inverter-based generating resources compared to
synchronous generators and shows examples of negative sequence directional relay
misoperations using simulation models and tests systems inspired by actual transmission
networks.

In many cases the negative sequence current contribution by inverter-based resources can
be programmed. This is an inverter controller design decision, and in the absence of grid
codes in several countries, it is usually dependent on the inverter manufacturer. In the future,
grid codes may be applied, similar to the one recently being proposed in Germany [4],
requiring the inverter-based generating resources to support the negative sequence system.
In that case, potential risks related to applicability of negative sequence directional relaying
may be mitigated.

References:
1. M. Nagpal and C. Henville, “Impact of Power-Electronic Sources on Transmission
Line Ground Fault Protection”, IEEE Transactions on Power Delivery, Vol. 33, No. 1,
February, 2018, pp. 62-70.
2. A. S. Chen, A. Shrestha, F. A. Ituzaro, and N. Fischer, “Addressing protection
challenges associated with type 3 and type 4 wind turbine generators,” in 68th
Annual Conference for Protective Relay Engineers, College Station, TX, March 2015,
pp. 335–344
3. System Protection Guidelines for Systems with High Levels of Renewables: Impact
of Wind & Solar Generation on Negative Sequence and Power Swing Protection, Palo
Alto, CA: 2017. 3002010937
4. VDE-AR-N 4120, “Technical Requirements for the Connection and Operation of
Customer Installations to the High Voltage Network (TCC High Voltage), Jan. 2015.

3.2.1.2 Zero Sequence Directional

Referring back to Figure 5, it can be seen that the zero and negative sequence networks are
similar, i.e. comprising of highly inductive impedances in a transmission system. Thus, the
zero sequence current and voltage exhibit a phase angle relationship similar to the negative
sequence system. Because of the simplicity of zero sequence filter design, requiring the
simple addition of three-phase currents or three-phase voltages, the zero sequence voltage
polarized directional relays are widely employed on the networked lines for directionality
decisions. However, this protective element suffers from one drawback that its directionality
is not always correct when there is strong mutual coupling from adjacent circuit(s) sharing
the right-of-way. This is one of the reasons that the negative sequence directionality principle
started to gain popularity when it became easily available to the relay application engineers
in numerical relays.

24
As discussed in Section 3.2.1.1, though the negative sequence directional protective element
offers some advantage over zero sequence directional protection, it typically cannot be
applied on a line terminal having significant IBR because of the undefined negative sequence
source characteristic.

Conventional synchronous generators are typically high impedance grounded-wye


connected machines supplying negligible zero sequence current at the generator bus.
Interconnection transformers at these sites are typically connected delta/wye-grounded and
supply zero sequence current to the transmission system. To clarify, this connection is delta-
connected on the generator side and wye-grounded on the transmission system side.

IBR do not supply zero sequence current at the inverter output terminals due to the inherent
design of inverters. However, the interconnection transformer connections used at IBR sites
may vary depending on site specific design considerations. IBR interconnection transformers
that are connected delta/wye-grounded will supply zero sequence current to the
transmission system (the same as typical conventional synchronous generators). The zero
sequence current and voltage phase angle relationship supplied to the transmission system
is largely independent of the generator characteristic with this transformer connection.

Delta/wye-grounded interconnection transformers lead to correct directionality decisions


by zero-sequence line terminal relays as long as there is enough zero sequence current and
the line does not have strong zero sequence mutual coupling from nearby circuits.

The protection system design should also be evaluated for times when the interconnection
transformer is not electrically connected to the transmission system [R1].

References:
1. M. Nagpal and C. Henville, “Impact of Power-Electronic Sources on Transmission
Line Ground Fault Protection”, IEEE Transactions on Power Delivery, Vol. 33, No. 1,
February, 2018, pp. 62-70.

3.2.1.3 Current Polarized Directional

The current polarizing method compares the zero sequence current measured in the neutral
of a wye-grounded transformer and the zero sequence current measured or calculated on a
line. The advantage of this method is that it does not require any PT connection. The IBR does
not negatively impact the zero sequence directional element as the zero-sequence network
is formed by the grounding of the neutral point of the wye winding of the transformer, as
discussed earlier. The zero sequence directional element will operate provided the fault
current contains sufficient zero-sequence current.

3.2.1.4 Polarizing Methods Used in Distance Relays

Distance relays normally measure the fault voltage of a phase used together with the
measured fault current of the same phase to determine the direction of the fault. This can be

25
done by a direct comparison of the angle between the voltage and the current or via the mho
characteristic that inherently processes the direction. Provided the measured fault voltage
and current are of sufficient magnitude for the assessment, the method will be able to
determine the direction correctly. An inverter driven fault current will result in a correct
directional decision as long as it produces enough fault current and fault voltage.

The challenge arises when the fault voltage is insufficient for use in a reliable directional
calculation. This can be the case on faults right in front of the relay or on lines with high SIR
ratios. As inverter driven generation increases, the SIR ratios will increase in the future due
to the higher source impedance of this type of generation.

It is the practice today that distance relays typically use one of the two following discussed
methods to solve the problem of not having enough fault voltage available for the direction
calculation.

3.2.1.5 Memory Voltage Polarized

Particular modern numerical protection relays have the ability to store data in memory for
future use. The memory voltage polarization method uses the assumption that the measured
voltage at the relay position will not change its phase angle during the fault but will change
its amplitude. This is to a certain degree correct in systems that consist mainly of
conventional synchronous generators. The load of the line causes an angle shift but this is
normally well below 60 degrees. For that reason, distance relays can use the memorized pre-
fault voltage and compare it with the fault current to determine if the fault is in forward or
reverse direction. See Figure 6.

Figure 6 Quantity used for memory voltage polarization

The assumption about the voltage angle may not be correct on mainly inverter driven fault
currents in the future. The angle of the fault current driving the voltage of inverters is not
determined by a magnetic field in the generator that can only change slowly with the
generator inertia, but by a control algorithm that can jump to any value based on the control
strategy. The assumption used for the memorized polarization method may no longer be
valid. This must be studied in more detail and considered in future applications.
26
3.2.1.6 Cross-Voltage Polarized

The second method used in distance relays to replace the missing fault voltage is using the
assumption that the phase relation between the fault voltage and the healthy phase voltages
does not change during the fault. This allows the calculation of a voltage vector from the
healthy voltages that would be in phase with the fault voltage as shown in Figure 7.

Figure 7 Quantity used for cross polarization

With inverter driven fault currents, the assumption used for the cross polarization may not
be correct. The inverter will support only a positive sequence system. The fault voltage also
in the healthy phases may see a phase shift caused by the fault. Detailed simulations are
required to investigate the behavior of the voltages during a fault.

3.2.2 Faulted Phase Selection

Faulted phase selection logic ensures that the correct fault loop (fault type) is identified and
so enables the correct directional, distance and tripping elements etc. Present faulted phase
selection logic is driven by either voltage, current or a combination of voltage and current
and is based on the principle that a power source can be represented as a voltage source
behind an impedance. Therefore, the relationship (in magnitude and phase) between the
voltages and currents (faulted and unfaulted) are determined by the fault type and the
impedance of the power system network. This means the laws of physics can be used to
determine the relationship between the phase voltages and currents and their derived
quantities (i.e. sequence voltages and currents) and using these, the faulted loop (fault type)
can be identified.

For example, consider a single-phase-to-ground (or single-line-to-ground SLG) fault on A-


phase in a power system where the power sources can be represented by a voltage behind an
impedance. The relationship between the phase and sequence voltages and currents is as
shown in Figure 8.

27
VA IA

IC IB
VC VB

a.)

V1

I1
I0
I2

V0

V2
b.)

Figure 8 a) Magnitude and phasor relationship for the phase voltages and currents for an A-
phase to ground fault; b) Phasor relationship for the sequence voltage and current for an A-
phase to ground fault, using the A-phase as reference. (For a power system where the power
source can be represented by a voltage behind an impedance)

The phasor relationship between the faulted and unfaulted phase voltages and currents (in
magnitude and phase) are typical for an A-phase to ground fault on a power system where
the power source can be represented as a voltage behind an impedance. Therefore, a fault
identification logic based on these principles can be used on any power system were the
power source can be represented as a voltage behind an impedance.

Should it not be possible to represent an inverter-based power source as a voltage source


behind an impedance, then the relationship between the voltages and currents (in magnitude
and angle) are no longer determined by the fault type and the impedance of the source.
Classical fault calculation methods may not apply and therefore, cannot be used to identify
the faulted phase.

3.2.3 Fault Location

Impedance based fault location algorithms calculate the impedance between the relaying
(measuring) point and the fault point using the measured voltage(s) and current(s) of the
fault loop. Fault resistance and remote terminal current infeed in a non-homogeneous power
system can corrupt the calculated impedance to the fault in a single ended impedance-based
method. Therefore, single ended impedance-based fault location algorithms are often
implemented so that the algorithm does account for the fault resistance and the remote
current infeed and negates their effect. Impedance based fault location algorithms require
that the relationship between the fault voltage and current be determined by the impedance
of the power system. Should the relationship between the voltage and current in the fault

28
loop not be determined by impedance between the relaying point and the fault point, any
impedance-based algorithm will calculate the incorrect fault location.

Travelling wave-based fault location methods use the arrival times of the surges generated
due to a fault. The traveling wave generated during a fault is only dependent on the pre-fault
voltage at the fault point, the characteristic impedance and the propagation velocity of the
transmission line and is independent of the power source(s). Therefore, traveling wave fault
locators are ideally suited for application on power systems with a large ingression of IBR.

The calculation of the fault location is normally conducted not as a high speed (sub-cycle)
task but as an offline task after the fault. The fault impedance calculation for protection
purposes will select speed above accuracy. The protection decision is a yes/no decision and
does not really need to know the accurate impedance value. The objective of the fault location
task on the other hand is to calculate the accurate impedance and distance to the fault as this
information is used to dispatch a maintenance crew.

Different algorithms are used to compensate for factors that contribute to errors when
calculating the fault impedance such as:

• Load current
• Fault resistance
• Mutual coupling
• Non-homogenous line configurations

The change in the zero and negative sequence network of IBR will change the underlying
assumptions of some of these algorithms. Many algorithms for example compensate for the
load (i.e., Tagaki, Novosel, etc.) based on the assumption that the voltage source and angle of
the load driving current does not change during the fault. This may not be true for IBR.

Also, many of the algorithms assume a homogenous power system behind each line terminal
of the faulted line. With the different characteristics in negative and zero sequence systems
this may not be the case when different generation mixes are present on different line
terminals.

An additional problem that is not only relevant for the fault location but even more so for the
actual distance protection impedance calculation is the fact that inverters may inject reactive
fault current into the system. With this, fault impedance may actually be seen as reactive from
one-line terminal when the remote line terminal injects reactive current. The line reactance
calculation will experience higher errors than with conventional synchronous generators.

3.2.4 ΔV/ΔI Fault Detection

Incremental quantities (∆V and ∆I) can be used not only to detect the presence of a
disturbance or fault in the power system but also can be used to determine the direction of
the fault or disturbance. An incremental quantity is obtained by calculating the difference
between the present value of a quantity and the value of the quantity 1 or 2 power system
cycles previously. For example, the incremental voltage with a cycle buffer is calculated as
follows:
29
∆𝑉𝑉𝑘𝑘 = 𝑉𝑉𝑘𝑘 − 𝑉𝑉(𝑘𝑘−1𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐) [F1]

Where k = present value

Therefore, incremental quantities only exist (live) for a short duration (i.e. 1 or 2 power
system cycles). Distinguishing a fault from a disturbance in the power system can be done by
requiring a minimum incremental change in either voltage or current or both or that the
product of these two quantities be greater than a set value. The relationship between the
incremental voltage (∆V) and incremental current (∆I) is determined by the impedances of
the power system. The direction of the fault or disturbance is determined by observing the
polarity (if instantaneous quantities are used) or phase (if vectors quantity are used) between
the ∆V and ∆I.

For a disturbance or fault in front of the relaying (measuring) point, ∆V measured at the
relaying point is equal to the product of the source impedance (ZS, the impedance behind the
relaying point) and ∆I. The incremental current is equal to the negative of the prefault voltage
(-VPRE) at the fault point divided by the impedance of the fault loop. Figure 9 shows the
relationship between ∆V and incremental replica current (∆IZ ) ,( incremental replica current
is the incremental current multiplied by 1 ohm of the line impedance) for a forward fault.

Fault
Point
mZLINE
∆I
ZS -VPRE
∆V

∆v

∆iZ

Figure 9 a) Incremental quantity circuit diagram for a fault/disturbance in front of the


relaying point; b) relationship between the instantaneous incremental voltage and
incremental replica current; c) relationship between phasor incremental voltage (ΔV),
incremental current (ΔI) and incremental replica current (ΔIZ) (For a power system where
the power source can be represented by a voltage behind an impedance)

30
For a disturbance or fault behind the relaying point (reverse direction) the ∆V measured at
the relaying point is equal to the product of the sum of the remote source impedance and the
protected transmission line (ZR+ZLINE , the impedance in front of the relaying point) and ∆I.
The incremental current is equal to the negative of the prefault voltage (-VPRE) at the fault
point divided by the impedance of the fault loop. Figure 10 shows the relationship between
∆V and incremental replica current (∆IZ ).

Fault
Point
(1-x)ZS xZS ZLINE
∆I
-VPRE
∆V ZR

a.)

∆v

∆iZ

Figure 10 a) Incremental quantity circuit diagram for a fault/disturbance behind the relaying
point; b) relationship between the instantaneous incremental voltage and incremental
replica current; c) relationship between phasor incremental voltage (ΔV), incremental
current (ΔI) and incremental replica current (ΔIZ) (For a power system were the power
source can be represented by a voltage behind and impedance)

From the above description, the relationship between the incremental ∆V and ∆I (in
magnitude and polarity/phase) is determined by the power system impedance and if the
relationship is no longer true, incremental quantities cannot be used to determine the
presence of a disturbance or the directionality of the disturbance. The feature of incremental
quantities however is that the physical system only needs to remain valid for a short period
of time (e.g. 1 cycle). This may be true for an IBR before the controller makes a control
decision and modifies the fault current contribution.

31
3.2.5 Phase and Ground Overcurrent Elements

Phase and ground overcurrent elements use the magnitude of the fault current applied to the
relay to detect faults. The overcurrent elements can be directional or non-directional and the
operating quantity could be based on phase or sequence components. In the case of phase
overcurrent elements, the setting of the phase element needs to be low enough to detect the
fault but high enough to not constrain the protected equipment from full output or load under
normal or emergency conditions. Setting overcurrent elements to detect ground and
unbalanced phase faults can be a challenge with IBR generation due to the lack of phase and
ground fault current contributions. As described in previous sections, existing practice has
demonstrated that the phase fault current contribution from IBR is typically limited to 1.1 to
1.2 per-unit of inverter rated current for phase faults (similar to a conventional synchronous
generator during steady state).

As an example, consider a phase overcurrent tripping element intended to detect phase-to-


phase and three-phase faults. As a starting point, the detection (pickup) may be set at 60% of
the minimum fault current. Doing so would result in a setting of 0.66 per-unit (1.1*0.6) of
source rating, which would limit the IBR output. This is because IBR generation above the
0.66 per-unit value would result in the phase overcurrent tripping element operating
erroneously on balanced three-phase power output, as opposed to operating for a fault
condition.

In the case of conventional generators, phase overcurrent elements are often supervised with
voltage (torque) control or restraint in order to achieve a reliable protection scheme. These
techniques may also be effective for three-phase faults when the fault current source is IBR.

IBR generation may have high negative sequence impedance as referenced in Figure 4. As
shown in figures 11 and 12 below for SLG and phase-to-phase (P-P, sometimes referred to
as line-to-line L-L) faults, the high negative sequence impedance Z2C results in a lack of
negative sequence current from the inverter. The inverter will provide positive sequence
current as long as it is paralleled with a system that has low negative sequence impedance,
but once separated from the system, the inverter high negative sequence impedance results
in greatly reduced fault current flow and non-detection of the fault from inverter-based
generation. This leads to the observation that as inverter-based generation is increased, this
could result in decreased detection or non-detection of SLG and L-L faults especially on
transmission systems very close to inverter-based generation. For double-phase-to-ground
(also known as double-line-to-ground L-L-G) faults, the resulting phase current sequence
components consist of positive and zero sequence with the resulting fault current reduced
due to the lack of negative sequence current. This result is not as severe as the reduction for
SLG and L-L faults.

32
Positive Negative
Sequence Inverter System Sequence

V
C

Z2V
Z2C
Z1V
Z1C
Z1TC

Z1TV

Z2TV
Z2TC

I2V
I2C=0
I1C

I1V
Z1L*(1-m)

Z2L*(1-m)
Z1L*m

Z2L*m
Rfault

Figure 11 Sequence components for a phase-to-phase fault

Positive Negative Zero


Sequence Inverter System Sequence Sequence

V
C C
Z0V
Z0C
Z2C

Z2V
Z1TV
Z1C

Z2TV
Z2TC

Z0TC
Z1TC

Z1TV

Z0TV
I2C=0

I0V
I2V
I1C

I0C
I1V

Z2L*(1-m)

Z0L*(1-m)
Z1L*(1-m)
Z1L*m

Z2L*m

Z0L*m

Rfault Rfault

Figure 12 Sequence components for a double-phase-to-ground fault

Ground overcurrent detection elements use zero sequence or ground current. During normal
balanced conditions zero sequence current is ideally zero, therefore the pickup value for this
33
element can be set under generator and/or equipment maximum loading and is not
influenced by the current limiting function of the IBR. The IBR interconnecting transformer
configuration may, however, provide a zero sequence current path during unbalanced
conditions or some types of faults. For example, a delta/wye-grounded transformer, with
delta connection tied to IBR, will provide a zero sequence current path on the wye-grounded
side for operation of the ground overcurrent elements when unbalanced conditions are
detected (see section 3.2.1.2).

It should be noted that for SLG faults, the amount of resulting 3I0 current is low (mainly due
to the high negative sequence impedance which limits source current), such that the ground
overcurrent element will have the same limitation for ground fault protection during SLG
faults. Similar low level current values are available for phase overcurrent current measuring
elements (refer to SLG Figure 5).

At the present time there is still a large amount of synchronous generation which supports
negative and zero sequence current components during unbalanced faults. It is a good
practice to verify the available negative and zero sequence current components for
contingency conditions, such as a synchronous generator out of service, or as new IBR
generation sources are interconnected to the system. In extreme cases of heavy penetration
of IBR generation sources interconnected in one area, the ground overcurrent fault levels
could be too low for adequate fault detection. Likewise, over time it is a good practice to
validate phase currents available for phase-to-phase and phase-to-phase-to-ground faults, to
determine whether phase overcurrent elements can be set appropriately to discriminate
between load current and fault current.

3.2.6 Phase and ground distance elements

Phase and ground distance relays measure the voltage and current for fault determination.
This measurement is placed on an R-X diagram and is also known as an impedance relay (see
Figure 13).

Figure 13 Mho Phase Distance element

34
For a two terminal line the relay is set based on the impedance of the line. To add a level of
security, most modern distance elements are supervised with phase current fault detectors
for phase distance elements and ground current fault detectors in the case of ground distance
elements. The setting criteria for the fault detectors is to set the detectors above load and low
enough to pick-up for all in section faults in the case of phase distance. The ground distance
current supervision can have the same pick-up as a ground overcurrent element. System
configurations that have either islanded IBR or dominant IBR penetration (relative to
synchronous generation) may experience a lack of relay dependability for phase and ground
distance protection. Notably, the high negative sequence impedance for SLG and LL faults
will result in low phase and ground currents. These low current quantities will result in a
higher apparent impedance and may prevent the scheme from providing adequate
protection.

Since phase distance elements use voltage and current as operating quantities, they are less
susceptible to load encroachment than phase overcurrent, therefore the effects of low
inverter fault current will not be as pronounced as for phase overcurrent protection.
However, the lack of current can result in having to reduce the pick-up current of the fault
detector to the minimum range of the relay. The minimum available fault current may also
be at the minimum detection range of the relay thereby preventing the distance element from
operating for an in-section fault.

Voltages and currents utilized in phase distance and ground distance elements are listed in
tables 1 and 2 below. Note all distance elements operate by comparing an operating quantity
(listed) against a polarizing quantity which is different depending on the design of the relay
(e.g., self-polarized, cross-polarized, etc.):

Table 1 Phase Distance voltages and currents applied to distance relays for multi-phase
faults
Relay Element Fault Type Operating Polarizing
Quantity Sop Quantity Spol
1 A phase to B phase IAB *Zr - VAB Design dependent
2 B phase to C phase IBC *Zr - VBC “
3 C phase to A phase ICA *Zr - VCA “
where Zr is the relay reach

Table 2 Ground Distance voltages and currents applied to distance relays for multi-phase
faults
Relay Element Fault Type Operating Quantity Polarizing
Sop Quantity Spol
1 A phase to ground (IA+3I0k) *Zr - VAG Design dependent
2 B phase to ground (IB+3I0k) *Zr - VBG “
3 C phase to ground (IC+3I0k)*Zr - VCG “
where k= (Z0-Z1)/3Z1

35
Figure 14 shows an example of a tapped interconnection of the inverter-based generation to
a very strong system. This is practically a multi-terminal line but the fault current
contribution for line faults from the inverter-based generation is very low compared to the
contribution from sources A and B. Therefore, any distance elements applied at the inverter-
based generation to protect the line is susceptible to large “infeed” effects from the stronger
sources resulting in a need for a higher impedance setting, unless sequential tripping is
permitted (the IBR terminal trips after the two strong source terminals have tripped). This
may limit the line loading on the tap line to the inverter-based generation.

Figure 14 Multi Terminal Line with Inverter based generation

Protection elements designed to date are based on the assumption that a power source can
be presented as a voltage behind an impedance. This may result in distance elements not
operating dependably in these power systems.

References:
1. IEEE Std C37.113-2015 IEEE Guide for Protective Relay Applications to
Transmission Lines.
2. C-24 Report: Modification of Commercial Fault Calculation Programs for Wind
Turbine Generators

3.2.7 Conventional Resistive and Reactive Blinder Power Swing Block and Out-of-
Step Trip Elements

When inverter-based resources (IBR) replace or displace synchronous generation on the BPS,
system source impedances will typically increase and system inertia will decrease 24, due to
the synchronous generators being taken offline. The change in system source impedances

24As discussed in the previous section, many modern IBR technologies have commercially available
functions for providing primary frequency response and fast frequency response. Here the statements
being made are based on the assumption that such supplemental controls are not being deployed with
the IBR.

36
(both at the local and remote ends of the line under study) may alter the power swing
impedance trajectory to be either closer to or further from the line terminal under study.
Decreasing system inertia may result in power swings with faster slip rates. The magnitude
of the effect of changing source impedances on power swing block (PSB) and out-of-step trip
(OST) relaying needs consideration.

Below are examples of issues that should be considered and studied in more detail. It should
be noted these examples all still assume that there is a percentage of remaining synchronous
generation in the system (i.e. not 100% IBR penetration), and detailed study will need to be
performed for each system on a case-by-case basis.

1) Line Terminal Without PSB that Requires PSB Under High IBR Penetration Conditions

Consider the example line terminal below shown in Figure 15 where a synchronous
generation station is connected to the BPS via a single transmission line. In this example, the
plotted power swing corridor is behind the zone 1 and zone 2 phase distance relays. At the
time the plant was put in service (before IBR was added to the system), it was determined
that PSB relaying was not required on the transmission line from Oak to Maple since the
power swing corridor was not near the zone 1 or zone 2 phase distance relays.

37
Figure 15 Generation Remote From Load Without Need for PSB

After time, significant amounts of IBR have been added to the system resulting in significant
amounts of synchronous generation being cycled offline due to the fluctuating nature of IBR.
Figure 16 shows the resulting power swing corridor with remote synchronous generation
offline. In this case, it is obvious that it is necessary to add PSB to the transmission line
terminal to keep it from tripping for a power swing.

38
Figure 16 Generation Remote From Load WITH Need for PSB Due to High IBR Penetration

2) Faster Swing Rate Caused By Lower System Inertia

Consider the test system shown in Figure 17 from the IEEE PSRC Working Group D29. This
test system uses a commercially available transient stability program, allowing dynamic
stability studies to be simulated. At t = 0.1 sec., a disturbance is simulated by opening the
breaker at Fir to Elm. When this breaker is opened, the maximum power transfer at the Maple
to Spruce line terminal is exceeded, resulting in an unstable power swing.

39
Figure 17 IEEE PSRC WG D29 Test System

The simulation is run for one second. The plot as seen in the R-X impedance plane is shown
in Figure 18. The calculated slip rate for this power swing is about 0.63 Hz. To show the effect
that lower system inertia has on a power swing, the same system is used and the same
disturbance is simulate out to t = 1 second, but the inertia constant (H) for each generator is
reduced by half. Figure 19 shows the plot of the power swing in the R-X impedance plane for
the low inertia test. As can be seen, the power swing travels significantly further out to t = 1
second and actually begins a second slip cycle through the relay elements. The calculated slip
rate for this power swing is about 0.77 Hz, which is an increase in rate of about 22%.

The increase in slip rate caused by a reduction of system inertia due to high penetrations of
IBR may impact both PSB and OST relay blinder settings. The inner and outer resistive
blinders may have to be set further apart and/or the PSB/OST timers may have to be set
lower to allow for correct operation at higher slip rates.

40
Figure 18 Maple to Spruce Unstable Power Swing at t = 1 sec. with Full System Inertia

41
Figure 19 Maple to Spruce Unstable Power Swing at t = 1 sec. with Half System Inertia

As the two examples above illustrate, PSB and OST settings may be impacted by high
penetrations of IBR. Power swing reactance, resistance and timer elements may have to be
set considering higher system source impedances and lower system inertia caused by high
penetrations of IBR. The issues described in this section should be considered and addressed
by IEEE PSRC Working Groups C32 and D29.

3.2.8 Settings-less power swing relay algorithms

Several settings-less PSB and OST algorithms are available from different relay
manufacturers. These relay algorithms dynamically track the impedance trajectory or
monitor the swing center voltage to determine if a power swing is present and to determine
if the power swing is stable or unstable. These algorithms are generally capable of correctly
operating to block distance elements from tripping for stable power swings and to allow
tripping for unstable power swings. They can operate correctly for a wide range of power
swing conditions and can operate for a wide range of power swing slip rates.

42
Settings-less power swing relays may be challenged by higher than normal swing rates due
to lower inertia levels in systems with high IBR penetration. IEEE PSRC Working Groups C32
and D29 should investigate the impact of lower system inertia and larger quantities of IBR
currents and voltages on settings-less power swing relaying to determine if it is significantly
impacted. If settings-less power swing relaying is impacted, these working groups should
work with relay manufacturers to develop solutions to the problems uncovered.

3.2.9 UFLS Frequency Time Delay Settings

Under Frequency Load Shed (UFLS) is necessary to keep load and generation in an electrical
island balanced and as close to the nominal system frequency (60 Hz for North America) as
possible after the loss of significant amounts of generation or after the loss of significant
power imports following a system separation event. The rapid influx of IBR has both offset
and replaced conventional fossil generation resources. Because of this, lower levels of system
inertia are available at any given time, which results in more rapid frequency decay following
a loss of generation or power imports.

Consider a system with 8,000 MW of load, 7,000 MW of fossil generation and 1,000 MW of
power imports from a synchronous interconnection as depicted in Figure 20.

Figure 20 Balanced Load-Generation Island with Fossil Generation and Imported Power

If an islanding event occurs resulting in the loss of 1,000 MW of imports, a deficit of 1,000
MW of generation results, which will cause the system frequency to decay. The frequency will
decay according to the following formulae [F2 - F4]:

𝑡𝑡
𝑓𝑓 = 𝑓𝑓𝑠𝑠𝑠𝑠𝑠𝑠 − ∆𝐿𝐿 ∙ (1 − 𝑒𝑒 − 𝑇𝑇 ) ∙ 𝐾𝐾 ∙ 𝑓𝑓𝑠𝑠𝑠𝑠𝑠𝑠 [F2]

43
𝑀𝑀 1
𝑇𝑇 = 𝐾𝐾 =
𝐷𝐷 𝐷𝐷
In the equations above, fsys is the base system frequency (60 Hz), ΔL is the change in load in
per unit, t is time in seconds, M is the inertia constant which equals 2H, and D is the load
damping constant. Rewriting the equation and substituting values for T and K yields:

𝐷𝐷∙𝑡𝑡 1
𝑓𝑓 = 𝑓𝑓𝑠𝑠𝑠𝑠𝑠𝑠 − ∆𝐿𝐿 ∙ (1 − 𝑒𝑒 − 2∙𝐻𝐻 ) ∙ ∙ 𝑓𝑓 [F3]
𝐷𝐷 𝑠𝑠𝑠𝑠𝑠𝑠
The load damping constant, D, represents the phenomenon that occurs where system power
(in MW) increases or decreases due to higher or lower system frequency. Motor loads that
are frequency dependent will operate at nominal power when the system frequency is 60 Hz.
If the system frequency increases, the motor will speed up, yielding a higher electrical power
level. If the system frequency decreases, the motor will slow down, yielding a lower electrical
power level. The damping constant (D) is expressed as a percent change in load for a one
percent change in system frequency. Typical ranges of (D) are 1 – 2 percent [Reference 1]. An
in-between value of 1.5 is assumed for use in this example, thus a one percent change in
system frequency will result in a 1.5% change in load. The equation for (D) is as follows:

𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝑆𝑆𝑆𝑆𝑆𝑆 ∙ 1.5
𝐷𝐷 = [F4]
100
In the equation above, LoadSys is the remaining system load after separation, 1.5 is the percent
change in load for a 1% change in system frequency and 100 is the system base in MVA.

For the system depicted in Figure 19, the total system inertia (H100) in per unit on a 100 MVA
base is calculated as follows:

𝐻𝐻 ∙ 𝐺𝐺𝐺𝐺𝐺𝐺𝑀𝑀𝑀𝑀𝑀𝑀 3.0 ∙ 7000


𝐻𝐻100 = = = 210𝑀𝑀𝑀𝑀 ∙ 𝑠𝑠/𝑀𝑀𝑀𝑀𝑀𝑀 [F5]
𝑀𝑀𝑀𝑀𝑀𝑀𝐵𝐵𝐵𝐵𝐵𝐵𝐵𝐵 100

Here it is assumed, for the sake of simplicity, that the inertia constant for all the synchronous
generators is H = 3 MWs/MVA. In a real power system, the calculation is more complex due
to the varying inertia of different units. Following the trip of the tie lines and the loss of 1,000
MW of imported power, the islanded system experiences a drop-in frequency until UFLS
occurs or until governor response occurs in 3 – 5 seconds [F3]. Typical UFLS programs will
shed around 10% of the system load per Under Frequency (UF) set point. Generally, three UF
set points are used (e.g., 59.3 Hz, 59.0 Hz, 58.7 Hz which are used in the following discussion).
Typical time delays for UF relays are from 6 – 30 cycles. Some time delay must be used to
prevent false tripping during fault conditions.

Occasionally, longer time delays (up to 30 cycles) are necessary to prevent false tripping of
UF relaying on tapped transmission lines with significant motor load. Most UF relaying is

44
supervised by a voltage element that only allows UF tripping if the voltage is above a
threshold (typically 70 – 85%). If UF relaying is applied at a tap off of a transmission line that
opens due to a fault condition, the voltage at the tap location will not drop quickly to zero if
heavy motor load exists at the tap location. Even though the transmission line is de-energized,
the heavy motor load will support the voltage while the large motors slow down. During this
condition, it is possible for the UF relay to false trip if the motor load holds the voltage above
the UF relay voltage threshold while the frequency of the de-energized line decays below the
UF relay frequency set point.

For the system in Figure 20, once the tie lines trip resulting in a loss of 1,000 MW of
generation, the frequency will decay per formula [F3]. Figures 21 and 22 show the theoretical
frequency decay assuming only system inertia is arresting the frequency decline.

Figure 21 Frequency Decay for Loss of 1,000 MW out to 15 Seconds

45
Figure 22 Frequency Decay for Loss of 1,000 MW out to 2 Seconds

As Figures 21 and 22 show, the frequency will decay to a low enough level to initiate UFLS.
Using formula [F3], the first UF set point of 59.3 Hz will be reached at t = 0.528 seconds.
Assuming a 6-cycle intentional relay time delay, 1-cycle relay contact operating time and 3-
cycle breaker operating time, the load will be shed at t = 0.695 seconds. At t = 0.695 seconds,
the frequency will have decayed to 59.10 Hz, which is above the second UF set point (59.0
Hz). Therefore, only the first level of UFLS trips, resulting in only 800 MW (10%) of load shed.

Figure 23 shows the same system with half of the internal generation being supplied by IBR.
For this example, we assume that half of the fossil generation has been displaced by the IBR
generation, resulting in a loss of half of the total system inertia.

46
Figure 23 Balanced Load-Generation Island with Wind, Solar and Fossil Generation and
Imported Power

If we assume the tie lines trip with the generation mix as shown in Figure 23, the theoretical
frequency decay, assuming only system inertia is arresting the frequency decline, will be
depicted as shown in Figures 24 and 25.

Figure 24 Frequency Decay for Loss of 1,000 MW out to 15 Seconds

47
Figure 25 Frequency Decay for Loss of 1,000 MW out to 15 Seconds

As Figures 24 and 25 show, the frequency will decay to a low enough level to initiate UFLS.
Using formula [F3], the first UF set point of 59.3 Hz will be reached at t = 0.264 seconds.
Assuming the same relay and breaker operating times from above, the load will be shed at t
= 0.431 seconds. At t = 0.431 seconds, the frequency will have decayed to 58.91 Hz, which is
below the second UF set point (59.0 Hz). Therefore, with half of the islanded load being served
by IBR, two levels of UFLS trip resulting in 1,600 MW of load shed.

By taking the first derivative of formula [F3], we obtain an equation that will allow the
calculation of the Rate-of-Change-of-Frequency (RoCoF) at any given point in time:

𝑑𝑑𝑑𝑑 −30 ∙ ∆𝐿𝐿 − 𝐷𝐷∙𝑡𝑡


= ∙ 𝑒𝑒 2∙𝐻𝐻 [F6]
𝑑𝑑𝑑𝑑 𝐻𝐻

Calculating the worst-case slope at t = 0 sec. yields the simplified equation:

𝑑𝑑𝑑𝑑 −30 ∙ ∆𝐿𝐿


= [F7]
𝑑𝑑𝑑𝑑 𝐻𝐻

Table 3 shows the RoCoF for several different IBR penetration levels for the system depicted
above after system separation occurs:

48
Table 3 RoCoF at t = 0 sec. for Varying Penetrations of IBR

After the South Australia blackout in September, 2016, the Australian Energy Market
Operator (AEMO) established RoCoF restrictions of 3 Hz/second to better assure that their
UFLS schemes would have a chance to operate to create a balanced load/generation island
[Reference .4]. As the examples above show, higher penetrations of IBR will require changes
in UFLS relay settings as well as UFLS practices.

The analysis of this section has identified several issues that need to be addressed related to
UFLS relay settings:

• High levels of RoCoF may require faster operating times of UFLS relays. This may
require lowering the time delay of UFLS relays that trip heavy motor loads to values
closer to 6 cycles. This may require adding complex communications schemes via
fiber optics or other communications medium to supervise and inhibit UFLS tripping
if the transmission source breakers are open. It may also require changing UFLS relay
trip locations to the transmission breakers or other non-industrial load locations. It
may further require using UF relays that monitor the RoCoF and trip in less than 6
cycles.
• High levels of RoCoF may require changes to NERC Standard PRC-006. Planning
Coordinators may have to change regional standards to allow more UFLS trip set
points, lower UF relay time delays, or require that more load be shed at each set point.

These issues should be addressed by NERC and IEEE PSRC Working Group C32. NERC and
the industry are actively investigating the effects of the changing resource mix on frequency
response, including RoCoF.

References:

49
1. P. Kundur, “Load Response to Frequency Deviation,” in Power System Stability and Control,
McGraw-Hill, Inc., 1994, pp. 584 – 587.
2. P. Kundur, “Factors Influencing Frequency Decay,” in Power System Stability and Control,
McGraw-Hill, Inc., 1994, p. 624.
3. P. Kundur, “Factors Influencing Frequency Decay,” in Power System Stability and Control,
McGraw-Hill, Inc., 1994, p. 625.
4. Australian Energy Market Operator (AEMO), “Black System South Australia 28 September
2016 – Final Report”. AEMO, Melbourne, Victoria. [Online]. Available: Black System South
Australia 28 September 2016 - FINAL REPORT

3.2.10 Frequency Tracking

A protective relay may use frequency to determine the periodicity of a signal by using a digital
band pass filter such as a Fourier filter or a derivative of the Fourier filter, so that the number
of samples (or data) consumed by the filter is always the same. For example, a one cycle
Fourier filter requires that the data consumed by the filter is always one cycle worth of data.
Should a one cycle filter consume data greater than or less than one cycle, the output of the
filter will contain errors. Similarly, the periodicity of the power system is important when
incremental quantities are used.
Therefore, it is essential that the relay accurately track the frequency of the power system.
Either voltage or current can be used to determine the frequency of the power system.
However, it is more common to use voltage for frequency tracking. Regardless of what
quantity is selected for frequency tracking the quantity must be representative of the
dominant frequency of the power system, irrespective of the power source. Should this not
be the case, the relay will acquire the incorrect frequency and this will lead to degraded
performance of the protection elements.
As the previous section illustrated, more IBR will result in a greater RoCoF, which could
impact frequency tracking of digital relays. This in turn could result in the mis-operation of
protective relays if the RoCoF exceeds the limits of the relays 25. This phenomenon was
observed in the September 28, 2016 South Australia blackout, where eight relays operated
due to rapidly declining system voltage and frequency [1].
IEEE PSRC WG C32 should study the effect of rapid RoCoF on protective relaying and
recommend solutions to prevent misoperation of protective relay elements during these
conditions.
References:
1. Australian Energy Market Operator (AEMO), “Black System South Australia 28 September
2016 – Final Report”. AEMO, Melbourne, Victoria. [Online]. Available: Black System South
Australia 28 September 2016 - FINAL REPORT

25This is distinct from the fact that in positive-sequence simulation platforms, frequency cannot be
easily and reliability calculated during disturbances that lead to a sudden and large change in voltage
phase angle (e.g. fault at a bus) and thus cannot be effectively used for emulating the performance of a
protection relay (element) during a frequency excursion (a WECC white-paper is presently being
developed on this subject, not yet available at the time of finalizing this report).

50
3.2.11 Synchronization

Two IEEE standards define the acceptable synchronization limits for synchronous
generators: IEEE Std C50.12 Standard for Salient-Pole 50 Hz and 60Hz Synchronous
Generators and Generator/Motors for Hydraulic Turbine Applications Rated 5MVA and
Above, and IEEE Std C50.13 Standard for Cylindrical-Rotor 50 Hz and 60 Hz Synchronous
Generators Rated 10 MVA and Above. Even though inverters are not rotating machines,
similar limits may be acceptable. Utilities interconnection guides also provides detailed
requirements for inverter synchronization. As inverter technology is ever-changing,
discussion with inverter manufacturer is recommended, a typical application may require
synchronization between the entities as below:

• Inverter/inverters to other inverter/inverters


• Group of inverters to the rotating machine/machines,
• Inverter/inverters to utility grid
• Island (Combination of inverters, rotating machine and loads) to the utility grid.
A centralized controller may help manage synchronization, isochronous/droop and load
sharing for a group of inverters/machines.

3.3 Relay Scheme Selection Issues Caused by Inverter-Based Resources


(IBR)

High levels of IBR can influence the surrounding power system so that it may be considered
weak, in that the system exhibits low fault currents (that may be even lower than load
currents), large voltage changes coincident with load changes and faults, and as noted above,
relatively low polarizing quantities.
Note that these effects may appear when the power system is operating in a contingency
condition. Power system conditions change based on maintenance or following an outage of
an element. During normal system operation with all elements in-service, there may be no
protection issue. In a contingency condition such as absence of a source or a line section, one
or more protection schemes may need to detect and operate with a weak source.
The following information is offered on protection schemes applied to transmission lines. For
all schemes described, low fault current levels due to high levels of IBR present a challenge
to scheme sensitivity, for both tripping and directional elements.
It should be noted that protection schemes for other elements such as bus protection,
transformer protection, etc. could experience the same difficulty with IBR if the available fault
current levels are near or below scheme sensitivity.
Common schemes applied to transmission lines include:

• Permissive Over-reaching Transfer Trip (POTT)


• Permissive Under-Reach Transfer Trip (PUTT)
• Directional Comparison Blocking (DCB)
• Directional Comparison Unblocking (DCUB)
51
• Line Current Differential (LCD)
• Phase Comparison
• Pilot Wire
• Direct Under-Reach Transfer Trip (DUTT)
• Step Distance/Directional Overcurrent
For systems with IBR, overreaching elements must be checked for adequate sensitivity for
minimum fault contribution. At a weak terminal, no trip may occur if fault contributions are
not above the trip settings. Both POTT and PUTT may require weak feed echo; weak feed
echo logic is ineffective (and unnecessary) in DCB schemes.
Due to the improved sensitivity, LCD and pilot wire schemes tend to be good choices for
systems with IBR. A significant advantage is that the LCD scheme can operate on internal
faults even with no infeed at one or more terminals provided the total supplied fault current
is above the scheme sensitivity.
Phase comparison schemes may not operate on internal faults if one terminal is connected to
a weak source while the other is connected to a strong source. If prefault load is flowing from
the strong terminal to the weak terminal, the phase angle of the current from the weak source
may not reverse during an internal fault, which can result in the scheme not detecting the
fault.
Regarding IBR, a single-phase comparison blocking scheme would be more dependable (as
compared to dual phase or segregated phase comparison) since at least one terminal would
trip if no signal is received. Single phase comparison permissive and dual phase comparison
schemes using the powerline for a path may result in no trip for internal faults due to the
potential loss of the communication signal.
Step distance/directional overcurrent schemes generally have the same sensitivity issues as
the elements used in pilot schemes, with the added inherent time delays associated with the
overreaching distance and ground time overcurrent elements.
See IEEE Standard C37.113 for additional information on each scheme.
References:
1. IEEE Std C37.113TM-2015 Guide for Protective Relay Applications to Transmission Lines, pp
73-86

3.4 Short Circuit Study Issues Caused by IBR

As noted in section 3.1.1, IBR resources present unique short-circuit fault response
characteristics that are different compared to traditional rotating machines in which the fault
response depends on the machine electrical and physical characteristics. Instead, for IBR the
fault response depends on the inverter control strategy implemented, CR-VSI or CR-CSI,
which may be proprietary.

3.4.1 Source-To-Line Impedance Ratio (SIR) Issues

One problem that can arise in weak systems is that of high source-to-line impedance ratio.
This can occur even with lines that may not be considered physically short. Such lines can
52
have protection problems such as transient overreach of distance elements due to poor
performance of instrument voltage transformers if located on the line side of the terminal,
specifically coupling capacitor voltage transformers (CCVT). This can and has resulted in
misoperation of underreaching distance elements on external faults.
SIR issues can arise or be made worse at or near generating stations that are being
replaced/displaced by IBR. Transmission Owners will have to be aware of this issue at these
locations and will have to study whether an increase in SIR will cause CCVT transient
overreach issues. When CCVT transient overreach caused by SIR is a concern, there are at
least four solutions that can be implemented:

• Shorten the zone 1 reach. This may or may not be effective based on the apparent
impedance trajectory and may result in pulling the zone 1 element back so far that it
would be useless.
• Add time delay to zone 1 to allow CCVT response to stabilize (if possible). This added
delay could increase the possibility of stability issues.
• Where available, enable CCVT transient detection logic.
• Use Line Current Differential (LCD) relaying.

It is recommended that IEEE PSRC Working Group C32 work with IEEE PSRC Working Group
D38 to study the potential increase of SIR when high penetrations of IBR are present on the
BPS. It is further recommended that these Working Groups propose solutions to high SIR
caused by high penetrations of IBR if deemed necessary.

References:
1. D. Hou and J. Roberts, “Capacitive Voltage Transformers: Transient Overreach
Concerns and Solutions for Distance Relaying”. Available at http://www.selinc.com
2. J.L. Blackburn, Protective Relaying - Principles and Applications, p. 179.

3.4.2 Short Circuit Study Process Issues

When inverter-based resources (IBR) replace or displace synchronous generation on the


Bulk Power System (BPS), short circuit current magnitudes decrease in the area where the
synchronous generators are replaced or cycled offline. This decrease in fault current can have
a significant impact on protective relaying if not considered. Prior to the influx of renewable
generation in the 2000’s, the most common practice for performing short circuit studies was
to use system intact (all generation resources online) for base studies, especially when setting
ground instantaneous relay elements. Some utilities would consider multiple generator
outages at a power plant or in a local geographic area when performing coordination studies
if the multiple contingencies were credible. Other utilities would consider off-peak, low load
seasons (i.e., spring/fall) when performing short circuit studies and would outage high-load
peaking units as a base case starting point.

The rapid growth of IBR over the last decade has changed the dynamics of short circuit
current on the BPS and has brought new issues to light that need to be considered when
53
setting current sensitive protective relay elements. It is no longer sufficient to use only
system intact (maximum generation) when setting relays and verifying coordination. It is
now necessary to consider the minimum generation conditions as well, such as maximum and
minimum IBR penetration. The maximum generation profile is still used to establish
instantaneous overcurrent (IOC) tripping element pickup values (especially ground IOC) and
should still be used when performing ground loop coordination studies. The minimum
generation profile should be used to establish pick-up values for low-set relay fault detectors
(i.e., phase and ground distance, phase overcurrent enabled to trip during loss-of-potential
(LOP), switch-onto-fault (SOTF), etc.), to establish ground time overcurrent (TOC) pickup
settings for line-end ground faults, to check for weak-feed terminal requirements and should
be used when performing ground loop coordination studies.

The determination of the minimum generation profile is not a trivial task. Consultation with
a utility’s Transmission Planning and/or Transmission Operations department is necessary
to determine which synchronous generation units need to be set offline during off-peak, high
IBR conditions. Given today’s modern control center technology, it should be possible to
query a utility’s Energy Management System (EMS) to identify the timeframe that results in
the most synchronous generators being offline at the same time. This timeframe can then be
used to establish the minimum generation profile in the short circuit model. Since generation
profiles change over time, it is prudent to check and adjust the minimum generation profile
on a periodic basis (e.g., annually, etc.).

Ignoring minimum generation conditions caused by IBR can result in a variety of issues:

• Phase/ground distance and quad relay element fault detectors may not be set
sensitive enough to detect remote faults within their reach, which could result in non-
operation of the supervised element and could result in delayed clearing by other
zone relay elements, which could cause stability issues.
• Phase overcurrent elements enabled to trip during LOP conditions and SOTF
elements may not be set sensitive enough and may result in non-operation when
required.
• Ground TOC element pickup settings may not be set sensitive enough to achieve
desired multiples of pickup for remote line end faults which could result in delayed
clearing that could cause mis-coordination with other ground relays or could cause
system stability issues.
• Weak feed terminals may not be identified which could result in non-operation of the
weak feed terminal.
• Electromechanical ground IOC elements may not reach the desired length down the
protected line, which could cause upstream coordination issues. If the coordination
issue is severe enough, it may be necessary to replace the electromechanical ground
overcurrent relay with a microprocessor relay that has ground distance relay
elements.
• Ground IOC elements across a system may need to be phased out entirely in favor of
more predictable reaching ground distance elements.

54
An example of the magnitude of change that can occur when considering the minimum
generation profile is detailed in the following example. A utility within the Eastern
Interconnection with a summer load of 6000 MW has a peak IBR penetration of 3000 MW.
When 3000 MW of IBR is online, 3000 MW of synchronous generation is cycled offline. A
comparison of the BPS bus fault currents during zero IBR penetration and 50% IBR
penetration was performed and the percent decrease in fault current at each BPS bus was
calculated. Table 4 below summarizes the decrease in fault current magnitude caused by IBR
penetration of 50% for single-phase-to-ground and three-phase (3PH or 3LG) faults.

Table 4 Fault Current Percent Decrease with IBR Penetration of 50%

As can be seen from Table 4, only 12% of the buses result in a fault current decrease of more
than 10% for SLG faults. Only 25% of the buses result in a fault current decrease of more than
10% for 3PH faults. The smaller decrease in fault currents for SLG faults can be attributed to
the large number of auto transformers with delta tertiaries which provide a significant
amount of zero sequence current. While the percentages are low, the number of buses
impacted that may require additional study is significant (42 buses for SLG faults and 87
buses for 3PH faults).

Table 5 below analyzes the 3PH fault buses with a percent decrease in fault current greater
than 10%. This table shows that the impact is only significant at the generating station where
the synchronous generation was cycled offline and up to two buses away from the generating
station. Only 17% of the buses three to four buses away from the generating station had fault
current decreases greater than 10%. No buses had fault current decreases greater than 10%
five or more buses away from the generating station. Table 5 also shows that the fault current

55
decreases at the generating stations can be significant (up to 50%). Even up to two buses
away, the percent decrease can be in excess of 20%.

Table 5 Detailed Breakdown of 3PH Fault Buses with Fault Current Decreases Greater Than
10%

Another example from the Western Grid is illustrated in Table 6 and Figure 26 demonstrating
the reduction of fault current. In this example 2660 MW of synchronous generation was
converted to IBR in this area. As can be seen in Table 6 the fault current reduction drops
rapidly as one progresses from the generation busses.

MAMMOTH F
230.kV

NORTH STAR 2 Converted to Current Limit


230.kV

MAMMOTH E
WHISTLER BE 230.kV
230.kV

Convertered to Current Limit

WHISTLER BF
NORTH STAR 3 230.kV
230.kV

Converted to Current Limit

MAMMOTH
SIERRA PP 2 MAMOTH D 115.kV
SIERRA P 1 230.kV
230.kV 230.kV

Figure 26 System in which 2660 MW of Synchronous Generation was converted to IBR

Table 6 3LG, L-L and SLG fault values for the system in Figure 25 before and after
Synchronous generation is converted to an IBR
Before After

56
1LG 1LG 3LG LL 1LG
Bus Name kV 3LG (A) LL (A) 3LG (A) LL (A)
(A) (A) %DIFF %DIFF %DIFF
MAMMOTH D 230 52,584 45,458 51,684 45,037 37,991 46,225 -14.4% -16.4% -10.6%
MAMMOTH E 230 52,603 45,475 51,703 45,046 37,999 46,237 -14.4% -16.4% -10.6%
MAMMOTH F 230 52,550 45,429 51,652 45,011 37,968 46,195 -14.3% -16.4% -10.6%
MAMMOTH D 115 45,500 39,368 47,990 46,424 36,950 45,812 2.0% -6.1% -4.5%
WHISTLER BE 230 13,519 11,675 11,042 13,804 11,242 10,802 2.1% -3.7% -2.2%
WHISTLER BF 230 13,626 11,767 10,976 13,913 11,332 10,740 2.1% -3.7% -2.2%
NORTH STAR 3 230 12,691 10,959 9,763 12,937 10,579 9,573 1.9% -3.5% -2.0%
NORTH STAR 2 230 12,631 10,907 10,165 12,874 10,530 9,960 1.9% -3.5% -2.0%

Table 7 provides another example of fault current reduction on a system from the Eastern
Interconnection. The peak load demand on this system is approximately 28 GWs. The typical
day time demand during the spring season is 13.5 GWs. In this example, 30% of the
generation in the spring weather case was converted to the IBR based generation.

Table 7 Fault Current Percent Decrease with IBR penetration of 30% on light load case
Percent Current Decrease for SLG Faults Total Bus
Voltage <5% 5% - 10% 10% - 15% - >20% Count
(kV) 15% 20%
500 45% 30% 15% 3% 6% 33
230 84% 9% 2% 1% 4% 310
115 98% 2% 0% 0% 0% 690
Percent Current Decrease for Three Phase Faults Total Bus
Voltage <5% 5% - 10% 10% - 15% - >20% Count
(kV) 15% 20%
500 36% 24% 33% 6% 0% 33
230 69% 21% 6% 2% 2% 310
115 99% 1% 0% 0% 0% 690

The synchronous machine based generation that is turned off is located on the 500kV system.
The IBR that replaced this generation are scattered throughout the system and are connected
to the 230kV and 115kV systems. As can be seen from Table 7, the fault current decrease on
the 115kV system is mostly less than 5%. The 115kV system is electrically far away from the
500kV system where the synchronous machine based generation is turned off, which explains
why the reduction in fault current at transmission buses away from the traditional
synchronous machine based generation that is turned off is very minimal. The most
significant reduction in fault current is observed on the 500kV system which is closest to
locations where synchronous machine based generation is turned off and replaced with IBR.

These three examples illustrate that the penetration of IBR is not expected to reduce fault
currents consistently throughout the system. Instead, the significant reduction is expected to
be in small pockets of the system and could be scattered around the system.

57
The issues described in this section should be considered and addressed by IEEE PSRC
Working Group C32. IEEE PSRC WG C32 should provide guidance on best practices in the
development and use of minimum and maximum generation profiles. They should also
provide detailed guidance and examples of the use of maximum/minimum generation
profiles to establish settings for overcurrent and fault detecting relay elements to mitigate
the bulleted list of issues above.

3.4.3 Short Circuit Program Model Issues

The industry is well-versed in modeling traditional conventional synchronous machines,


both in short circuit studies and in stability studies, using subtransient, transient, and
synchronous impedances and time constants.

However, until recently, commercial short circuit study programs did not have an accurate
model for inverter-based generators. In 2014 IEEE PSRC Working Group (WG) C24 was
formed to address this industry gap by preparing a report titled “Modification of Commercial
Fault Calculation Programs for Wind Turbine Generators”. The WG (which is still active) has
proposed that phasor domain programs can model inverter-based generation using an
iterative solution method to consider the nonlinear fault response of these devices due to the
converter controls. The inverter-based resource may be modeled in short circuit programs
as a voltage dependent current source, which is a model substantially different to the
traditional voltage source behind an impedance model used for synchronous generators. The
WG has proposed that the voltage dependency of the current can be indicated either by using
tabular data structures with data provided by inverter manufacturers, or by models with
algorithms that represent generic control modes/logic of the inverter-based resources.

Such generic models have been recently developed by EPRI [2]. EPRI is a main contributor to
the PSRC WG C24 and has presented these models to the industry at various conferences and
IEEE meetings [6]. EPRI is also providing technical support to two widely popular commercial
short circuit study program vendors to incorporate these newly developed models into their
respective tools [7, 8, 9].

Finally, in parallel to the ongoing work presented before, EPRI is working to validate the
models in collaboration with EPRI members, PSRC WG C24 members and inverter
manufacturers through the use of a) recorded fault response measurements and b) detailed
EMT models of inverter-based resources provided by manufacturers under confidentiality
terms.

References:
1. J. Keller, B. Kroposki, “Understanding fault characteristics of inverter-based distributed
energy resources”, Technical Report NREL/TP-550-46698, January 2010.
2. Short-Circuit Phasor Models of Converter Based Renewable Energy Resources for Fault
Studies, EPRI, Palo Alto, CA: 2017. 3002010936.
3. Investigation of Solar PV Inverters Current Contributions during Faults on Distribution and
Transmission Systems Interruption capacity – A Quanta Technology report (http://quanta-
technology.com/sites/default/files/doc-files/Solar%20PV%20Inverter%20formatted.pdf)

58
4. Fault Current Contribution from Wind Plants – A PSRC Report (http://www.pes-
psrc.org/Reports/Fault%20Current%20Contributions%20from%20Wind%20Plants.pdf)
5. Modification of Commercial Fault Calculation Programs for Wind Turbine Generators – A
PSRC Report (http://pes-psrc.org/c/c24/c24.html)
6. T. Kauffmann, U. Karaagac, I. Kocar, H. Gras, J. Mahseredjian and E. Farantatos, "Phasor
domain modeling of Type III wind turbine generator for protection studies," Proc. 2015 IEEE
PES General Meeting, Denver, CO, 26-30 July 2015.
7. U. Karaagac, T. Kauffmann, I. Kocar, H. Gras, J. Mahseredjian and E. Farantatos, "Phasor
domain modeling of Type IV wind turbine generator for protection studies," Proc. 2015 IEEE
PES General Meeting, Denver, CO, 26-30 July 2015.
8. T. Kauffmann, U. Karaagac, I. Kocar, S. Jensen, J. Mahseredjian and E. Farantatos “An Accurate
Type III Wind Turbine Generator Short Circuit Model for Protection Applications”, IEEE
Transactions on Power Delivery, 2017
9. “EPRI Wind/Solar Phasor Domain Short-Circuit Models and Implementation Status in
Commercial Tools” Webcast recording, available at
https://www.epri.com/#/pages/product/000000003002010940/

3.5 NERC Protection and Control (PRC) Reliability Standards Issues


Caused by IBR

Some NERC Standards could be influenced by a change in short circuit capacity. Reduced
short circuit capacity due to higher penetrations of inverter-based resources warrant a
review of these Standards to ensure the transmission or generator entities underlying
assumptions for the requirements are still valid. A brief description of the affected Standards
follows.

MOD-32: A requirement for entities to provide short circuit modeling data to their
Transmission Planners and Planning Coordinators. Entities should ensure their short circuit
models are current and an accurate model is provided to their Transmission Planner and
Planning Coordinator.

PRC-002: Determination of locations that require disturbance monitoring equipment.


Entities should review their criteria used for determining where disturbance monitoring
equipment should be located to determine if a change in short circuit capacity has changed
their determination of disturbance monitoring equipment locations.

PRC-006: NERC Standard for Automatic UFLS. Planning Coordinators may have to change
regional standards to allow more UFLS trip set points, lower UF relay time delays, or require
that more load be shed at each set point. This Standard should be reviewed by Planning
Coordinators to determine if high levels of RoCoF will impact regional UFLS programs.

PRC-026: Requirement for relays to not trip for stable power swings. Entities should review
the criteria to determine if the existing angular stability assumptions are still valid for
changing penetrations of inverter-based resources.

59
PRC-027: Requirement for entities to ensure relay coordination for changing short circuit
capacity. Entities should review their protection coordination and ensure coordination for
changing short circuit capacities due to changing penetrations of inverter-based resources.

TPL-001: Requirement for entities to perform an annual Planning Assessment based on short
circuit analysis. Entities should ensure their short circuit model reflects changing
penetrations of inverter-based resources and determine if those changes impact their
Planning Assessment.

3.6 Conclusions

Large penetrations of IBR have the potential to cause protective relay issues across the North
American BPS that could result in delayed operation, mis-operation or non-operation of
protective relay elements. Fault currents and voltages produced by IBR’s are not like the
currents and voltages produced by synchronous generation. Nearly all BPS protective
relaying in use today is designed to operate on currents and voltages provided by
synchronous generation.

It has been shown that fault direction can be misinterpreted by relay directional elements if
significant IBR fault currents and voltages are present. This is especially true for negative
sequence directional elements. Relay elements that misinterpret fault direction can result in
relay mis-operation. IBR currents and voltages that do not follow the classical fault
calculation methods present challenges to faulted phase selection and fault location, which
can add to event analysis time and outage duration. It has also been shown that the
intermittent nature of IBR can cause lower fault currents to be present at or near generating
plants where synchronous generation is cycled offline due to large penetrations of IBR. These
lower fault current magnitudes can desensitize relay element fault detectors, which can cause
mis-operation or non-operation of protective relay elements.

Large penetrations of IBR that replace and displace synchronous generation reduces the total
system inertia on the BPS. Reduced system inertia can result in faster power swing rates
which can affect PSB and OST settings. Reduced system inertia increases the RoCoF, which
can result in deeper frequency excursions that could result in more UFLS than necessary.
Higher RoCoF has also been shown to affect protective relay frequency tracking algorithms,
which can result in relay mis-operations.

Short circuit study processes used by most utilities over the years were simple. The strongest
system possible was modeled so that maximum fault current conditions were present. Known
weak parts of a system were simulated if it was credible to have multiple units at a power
plant offline during off-peak seasons. With the rapid influx of IBR, these weak parts of a
system are changing dynamically, requiring a more sophisticated method of determining
weak system conditions. It is now necessary to set relay fault detectors and to perform relay
coordination studies using both strong and weak system conditions to ensure relay
coordination and adequate relay sensitivity for low and high IBR penetrations.

60
Traditional short circuit programs that calculate fault currents based upon synchronous
generation and the laws of physics do not produce the correct currents and voltages for a
system with large penetrations of IBR. Strides are being made to incorporate IBR currents
and voltages into these short circuit programs so that more accurate fault studies can be
performed to give a more exact representation of the actual fault currents and voltages
present. Finally, several NERC Reliability Standards may be affected by reduced fault current
levels due to large penetrations of IBR which will warrant their review.

3.7 Recommendations

This Task Force has concluded that large penetrations of IBR will affect short circuit values
across the North American BPS. Throughout this section, several IEEE PSRC Working Groups
and NERC have been identified to further investigate the effects of low short circuit values
caused by large penetrations of IBR. Below is a summary of the recommendations of this Task
Force broken down by the specific IEEE PSRC Working Groups and other organizations
identified to perform the additional studies:

IEEE PSRC WG C24 (Modification of Commercial Fault Calculation Programs for Wind
Turbine Generators)

• Continue work with short circuit study program vendors and EPRI to get accurate
working models for IBR implemented in commercial short circuit study programs.
• Ensure that convergence issues mentioned in this document and any other
outstanding issues are resolved prior to the anticipated report delivery date in
January, 2019.

IEEE PSRC WG C25 (Protection of Wind Electric Plants)

• Continue work to report on the present practices on collector feeder, grounding


transformer, bus, main transformer, capacitor bank and transmission tie line
protection for systems near wind power plants.
• Coordinate with IEEE PSRC WG C24 to ensure consistency with results seen from the
various short circuit study programs.
• Consider presenting the practices from this report to the major United States relay
conferences in 2020.

IEEE PSRC WG C32 (Protection Challenges and Practices for interconnecting solar or other
inverter-based generation to utility transmission systems)

• Ensure that all the issues identified in Sections 3.2 (Relay Element Response), 3.3
(Relay Scheme Selection) and 3.4 (Short Circuit Study Issues) are addressed in the
output of WG C32. Detailed descriptions of the issues and proposed solutions to the
issues should be given.
61
• Coordinate with IEEE PSRC WG D29 (see below) on the output related to Sections
3.2.7 (Conventional resistive and reactive blinder power swing block and out-of-step
trip elements) and 3.2.8 (Settings-less power swing relay algorithms).
• Coordinate with IEEE PSRC WG D38 (see below) on the output related to Section 3.4.1
(Source-To-Line Impedance Ratio (SIR) Issues).
• Coordinate with IEEE PSRC WG C24 on the output related to Section 3.4.3 (Short
Circuit Program Model Issues).
• Consider producing a quick turn-around white paper report with presentations to the
major United States relay conferences proposing solutions to the issues in Sections
3.2.1 (Relay Directional Elements), 3.2.5 (Phase and Ground Overcurrent Elements)
and 3.4.2 (Short Circuit Study Process). It would be desired to have this white paper
report completed by the end of 2018 with presentations to the major relay
conferences in 2019.
• Consider changing the output of this WG from a Report to an IEEE PSRC Standard or
Guide.
• Consider studying the benefit of IBR contribution of negative sequence current during
unbalanced transmission system faults to aid protection systems in regards to
selectivity, dependability and reliability.

IEEE PSRC WG CTF34 (Inverter-Based Short Circuit Current Impacts)

• Continue formation as a Task Force or get approval from the PSRC C-Subcommittee
to form a Work Group to oversee and coordinate other IEEE PSRC WG’s identified in
this section’s recommendations to ensure issues identified in this document are
adequately and promptly addressed.
• This Task Force/WG should retain the existing leadership and membership to allow
for efficient resolution of issues.
• Coordinate with NERC to ensure that NERC Reliability Standard issues identified in
this document are adequately and promptly addressed.

IEEE PSRC WG D29 (Tutorial on Setting Impedance-Based Power Swing Blocking and Out-Of-
Step Tripping Functions on Transmission Lines)

• Add examples and guidance to the output of WG D29 on setting PSB and OST relay
elements with consideration to high penetrations of IBR.
• Coordinate with IEEE PSRC WG C32 on the output related to Sections 3.2.7
(Conventional resistive and reactive blinder power swing block and out-of-step trip
elements) and 3.2.8 (Settings-less power swing relay algorithms).

IEEE PSRC WG D38 (Impact of High SIR on Distance Relaying)

• Add examples and guidance to the output of WG D38 on SIR issues caused by high
penetrations of IBR.

62
• Coordinate with IEEE PSRC WG C32 on the output related to Section 3.4.1 (Source-
To-Line Impedance Ratio (SIR) Issues).

North American Electric Reliability Corporation (NERC)

• Ensure that all NERC Reliability Standards mentioned in Section 3.5 are reviewed and
evaluated to determine if high penetrations of IBR will affect them enough to warrant
modification.
• Coordinate with IEEE PSRC WG CTF34 to ensure that NERC Reliability Standard
issues identified in this document are adequately and promptly addressed.

63

Вам также может понравиться