Protection of Wind Electric Plants: PSRC Working Group C25

Protection of Wind Electric Plants
PSRC Working Group C25
Power System Relaying and Control Committee

Report of Working Group C25
of the
System Protection Subcommittee
i
Members of the Working Group
Chair: Martin Best Vice Chair: Amin Zamani
Members and Contributors
Galina Antonova
Brian Boysen
Sukumar Brahma
Duane Buchanan
Jason Buneo
Ritwik Chowdhury
Evangelos Farantatos
Juan Gers
Frank Gotte
Charles Henville
Keith Houser
Mital Kanabar
Raluca Lascu
Shuhui Li
Yuan Liao
Jacob Lien
Rene Midence
Dean Miller
Mukesh Nagpal
James Niemira
Kevin Ridley
Lynn Schroeder
John Seuss
Steve Turner
Arman Vakili
Jim van de Ligt
Jakov Vico
Mohammad Dadash Zadeh
ii
ACKNOWLEDGMENTS
The Working Group is truly grateful for the support of our sponsoring committee, the
Power System Relaying and Control Committee, and System Protection
Subcommittee C.
KEYWORDS
Collector
Fault
Feeder
Generator
Grounding
Harmonic
Protection
Substation
Voltage
Wind
iii
ABBREVIATIONS AND ACRONYMS

DFAG Doubly Fed Asynchronous Generator (also known as Doubly Fed
Induction Generator, DFIG)
DFIG Doubly Fed Induction Generator (also known as Doubly Fed
Asynchronous Generator, DFAG)
EMT Electro-Magnetic Transient
GSU Generator Step-up Unit transformer
HV High Voltage
HVDC High-Voltage DC
IGE Induction Generator Effect
LVRT Low Voltage Ride-Through
MV Medium Voltage
NERC North American Electric Reliability Corporation
PCC Point of Common Coupling
PFCC Power Factor Correction Capacitors
POI Point of Interconnection
POTT Permissive Overreaching Transfer Trip
Pwind wind generating plant rated Power (MW)
SCIM Squirrel Cage Induction Machine
SCR Short-Circuit Ratio
SSCI Sub-Synchronous Control Interaction
Ssc System available Short-Circuit MVA
SSR Sub-Synchronous Resonance
STATCOM Static Compensator
TOV Transient Overvoltage
VAR Volt-Ampere, Reactive
VRT Voltage Ride Through
VT Voltage Transformer
WEP Wind-powered Electricity generating Plant
WTG Wind Turbine Generator
iv
TABLE OF CONTENTS
KEYWORDS .............................................................................................................. iii

ABBREVIATIONS AND ACRONYMS............................................................................. iv
1 INTRODUCTION ....................................................................................................... 1
1.1 Scope ............................................................................................................... 2
1.2 Purpose ............................................................................................................ 2
2 Difference Between Wind Electric Plant Substations and Conventional Distribution
Substations ............................................................................................................... 3
2.1 Collector Feeder Design and Characteristics .................................................... 3
2.2 Wind Electric Plant Substation Arrangements ................................................... 5
2.3 Wind Electric Generator Characteristics............................................................ 9
2.3.1 Type 1 ........................................................................................................ 9
2.3.2 Type 2 ...................................................................................................... 10
2.3.3 Type 3 ...................................................................................................... 10
2.3.4 Type 4 ...................................................................................................... 11
2.3.5 Type 5 ...................................................................................................... 11
2.4 Fault Currents and Equipment Ratings ........................................................... 12
2.5 System Grounding .......................................................................................... 13
2.6 Transformer Connections and Characteristics ................................................ 15
2.6.1 Wind Turbine Generator Transformers ..................................................... 15
2.6.2 Main Substation Transformers ................................................................. 15
2.7 Harmonics and Sub-harmonics ....................................................................... 17
2.8 Voltage and Frequency Control Requirements................................................ 19
3 Typical Protective Relay Schemes at Wind Electric Power Plant Substations .......... 20
3.1 Collector Feeder Protection ............................................................................ 20
3.1.1 Overcurrent Protection and Coordination with WTG Transformer Protective
Devices................................................................................................................ 21
3.1.2 Voltage and Frequency Protection and Coordination ............................... 38
3.1.3 Arc Flash Protection ................................................................................. 40
3.1.4 Removal of WTGs from Collector Feeders Under Fault............................ 41
3.2 Grounding Transformer Protection .................................................................. 41
3.3 Bus Protection ................................................................................................ 42
v
3.3.1 Zone-Interlocked Scheme ........................................................................ 43

3.3.2 Percentage-Restrained Differential .......................................................... 45
3.3.3 High Impedance Bus Differential .............................................................. 46
3.4 Main Transformer Protection........................................................................... 48
3.4.1 Transformer Differential Protection ........................................................... 49
3.4.2 Overcurrent Protection and Coordination with Collector Feeders ............. 54
3.4.3 Mechanical Detection of Faults ................................................................ 60
3.5 Capacitor and Harmonic Filter Protection........................................................ 61
3.5.1 Voltage Protection .................................................................................... 61
3.5.2 Overcurrent Protection ............................................................................. 62
3.5.3 Harmonic Current and Voltage Considerations for Protection Scheme
Operation............................................................................................................. 62
3.6 Transmission Tie Line Protection .................................................................... 63
3.6.1 Typical Communications Assisted Protection Schemes ........................... 64
3.6.2 Back-up Protection Schemes ................................................................... 65
3.6.3 Voltage and Frequency Protection Requirements .................................... 66
3.6.4 Supervision Requirements for Transmission Line Breaker Closing .......... 66
4 Conclusion ............................................................................................................. 67
5. Bibliography............................................................................................................. 68
APPENDIX A : Directional Phase Overcurrent Setting Considerations for WTG
Operation ............................................................................................................. 71
vi
THIS PAGE LEFT BLANK INTENTIONALLY
vii
1 INTRODUCTION
Working group C25 was given the assignment to write a report to provide guidance
on present relay protection and coordination practices at Wind-powered Electricity
generating Plants (WEP). This report covers the engineering considerations for the
design of the protection systems intended to protect all the elements that form WEPs.
It covers the following protection systems:
• Generator step up transformers

• Collector system feeders
• Grounding transformers
• Collector buses
• Reactors
• Capacitors
• Main station transformers
• Tie lines
• Points of interconnection
• Associated arc flash issues
A WEP is made of many small generators spread over a large area and includes many
subsystems that need to be protected. It is important to make sure that all the
subsystems are well protected and coordinated to maximize the reliability, security,
and dependability of the overall protection and control system.
For those not familiar with the different elements that form a WEP, commonly known
as a Wind Farm, this report introduces a description of the different elements
comprising a wind farm and how their unique characteristics may be considered to
provide a proper design.
This report provides engineering details covering:
• Possible Wind Farm electrical layouts

• Equipment ratings
• System grounding
• Wind electric generator characteristics
• Transformer connections and characteristics
o Wind turbine generator (WTG)
o Main substation transformers
• Harmonics and sub-harmonics
1
• Voltage and frequency control requirements

• Protective relay schemes
For successful application of the information provided in this report, the working
group recommends that the reader also become familiar with the guides and
publications pertaining to the protection systems described in this report, some of
which are provided as references at the end of this document.
1.1 Scope
This report covers protection of generator step up transformers, collector system
feeders, grounding transformers, collector substation buses, reactors, capacitors,
main station transformers, tie lines, points of interconnection and associated arc flash
issues. Although the report addresses coordination with wind turbine generator
protective devices and static VAR sources, protection of the wind turbine generators
and static VAR sources themselves is not included.
1.2 Purpose
Large WEPs are becoming more prevalent as generation sources on the power
system. Construction of these plants is significantly different from traditional large
generation stations. A traditional plant may have a relatively small number of large
machines – perhaps 2 to 6 generators each with a rating of 100 MW to 500 MW. By
comparison, a large modern wind powered generation plant will have a large number
of small generators – perhaps 80 to 100 turbines each with a typical rating of 1.5 MW
to 3.0 MW. Larger units have become available mostly in offshore wind farms. In a
wind powered generation plant, the turbines may be spread over an area as large as
100 square miles (260 square kilometers) or more, where power is collected at
medium voltage (usually 34.5 kV) from the individual turbines to a collector
substation where the voltage is stepped up to the transmission level for integration
into the transmission system.
The performance of the WEP during a fault condition is different from that of a
traditional generating station [1]. Therefore, the protection considerations for a large
WEP will be different than those of a traditional central station generation plant. This
report is intended to provide guidance on relay protection and coordination practices
that have been commonly used at terrestrial WEPs to serve as a reference for
practitioners working in the design of these plants. Typical design of WEPs is
discussed for background information and the relaying practices that have been used
with success are presented. Much of the equipment found in a wind powered plant is
common to many electric distribution systems – busbars, cables, transformers, and
capacitor banks, for example – so references are made to existing standards and
guides for protection of that equipment. Any special considerations or cautions
2
particularly related to the application of the equipment in wind electric plants are
highlighted in this report. This report does not cover the protection schemes used
within the wind turbine generators themselves, or that of other specialty equipment
such as STATCOMs. Protection schemes within these types of equipment are designed
by their manufacturers and are integral to the equipment.
2 Difference Between Wind Electric Plant Substations and

Conventional Distribution Substations
Wind Electric Plants are composed of many wind turbine generators (WTGs) which
are connected to a collector substation through a collector system. The collector
system is connected to the main substation or Point of Interconnection (POI) where
the voltage is stepped up to the transmission voltage level. While WEPs share several
common features with conventional utility electric systems, they also have several
unique characteristics that are typically considered during the design. Utility
substation designs mainly focus on maintaining continuous reliability, whereas
economics and unit availability play a more important role in the design of medium-
voltage collector systems, collector substations and switchyards, and high-voltage
transmission lines/cables of a WEP [2]. In the following subsections, some of the main
design characteristics of WEP substations and collector systems are described.
2.1 Collector Feeder Design and Characteristics

A collector system consists of a series of underground cables or overhead lines that
deliver power from WTGs to the collector substation. The majority of WEPs are
designed with multiple collector feeders/circuits. A collector system may also have a
single collector feeder depending on the number of turbines, size of cables, and
required system reliability [3]. The design and layout of the WEP collector system
depends on several factors such as turbine placement, terrain, reliability, and soil
thermal resistivity. The thermal resistivity of the soil has a direct impact on the
current carrying capacity of the cables. Additional factors include owner
requirements, economics, and expected climatic conditions. However, considerations
resulting from the location of turbines and the POI are the more critical factors in the
design of the collector system.
The collector substation may be connected directly to the POI. If the POI is located far
from the WEP, a transmission line and switchyard1 may be required in addition to the
collector substation. In any case, the collector feeders are connected to one or more
substation transformers at the collector substation, where voltage is stepped up to
the transmission system voltage. Although collector systems are normally operated
1
The switchyard may also be referred to as the interconnect substation.
3
at the highest medium voltage (34.5 kV in North America or 33 kV in Europe), they

are sometimes designed to operate at a voltage as low as 12 kV depending on the
available substation voltages. The reason is that the existing low-voltage
equipment/infrastructure may be too costly to change or reconfigure.
The collector feeder is commonly of underground type but could be of overhead
construction depending on the thermal resistivity of the soil, grounding
requirements, economic aspects, operation and maintenance requirements, local
landowner requirements, and system efficiency. Collector systems may consist of a
combination of overhead transmission lines (for connecting the collector substation
to transmission lines) and trenched underground or submarine cable systems (for
the collector system). The collector cable and/or conductors are evaluated to
optimize losses, voltage drop, system reliability, and maintenance criteria [9][3], [4].
It is also noted that the number of WTGs connected to a collector feeder is determined
based on the conductor ampacity. The rating of the step-up transformer at the
collector substation is based on the number of WTGs on the system of collector
feeders.
Collector systems of large WEPs normally have a radial configuration where turbines
are interconnected in a daisy-chain style, moving from the collector substation to the
farthest turbine. Figure 1 shows a typical collector feeder. As seen in this figure,
collector feeders may also have branch strings connected by junction boxes. The
junction boxes, in turn, have separable connectors (or elbows) that allow isolation of
a feeder string to enable continuous operation of the remaining connected turbines
while maintenance or repair work is being performed.
4
Figure 1: A typical collector feeder [10]
Wind turbine generators are often interfaced with the collector feeder through a
generator step-up (GSU) transformer that increases the generator voltage (typically
690 V or lower for WTGs smaller than 3 MW, and 3.3 kV or 6 kV for larger generators)
to the collector system voltage (up to 34.5 kV). Several studies are normally
performed on collector systems to identify the effects of the WTGs on the power
system. Industrial and commercial power systems standards are sometimes used for
the execution of such interconnection and collector system studies [9].
2.2 Wind Electric Plant Substation Arrangements

The planning and design of the WEP substation and switchyard can have some
flexibility based on the required reliability, operability, and maintainability. Wind
electric plant substation designs typically consist of a collector (medium-voltage bus)
and interconnection (high-voltage bus) configuration. These two systems are
connected through wye-wye (with buried delta), delta-wye, or even wye-delta
transformers as shown in Figure 2. The wye-wye (with buried delta) transformer
connection shown in Figure 2(b) is commonly used [10]. Figure 2(c) indicates that a
substation transformer with the delta connection on the collector side requires some
other means to provide adequate grounding (e.g., a grounding transformer connected
to the medium-voltage side).
5
Utility grid Utility grid Utility grid
52
52 52
HV
HV HV
MV Grounding
MV
MV Transformer
52-1 52-2 52-3 52-1 52-2 52-3

52-1 52-2 52-3
Collector 1 Collector 2 Collector 3 Collector 1 Collector 2 Collector 3 Collector 1 Collector 2 Collector 3

(a) (b) (c)
Figure 2: Common collector substation transformer configurations

(a) delta–wye grounded, (b) wye grounded–wye grounded with buried delta,
and (c) wye grounded–delta winding configurations
Collector substations for WEPs typically use an open-air bus design with single bus
or sectionalized bus arrangements [2]. Figure 3 illustrates common medium-voltage
(MV) bus substation configurations in WEPs. The single-bus design (Figure 3(a)) is
simple and very cost effective, but is a single point of failure and may be less reliable
than some other (more costly) alternatives. The sectionalized bus configuration
(Figure 3(b)) improves system reliability and availability when one of the station
transformers is out of service, albeit at a higher price. One may also need to evaluate
the potential improvement in energy production against the increased costs to
determine the economic benefits offered by the second transformer in a sectionalized
bus configuration [11]. In large plants, the use of multiple smaller transformers and
sectionalized buses instead of a single larger transformer will result in lower short-
circuit currents, provided the transformers are not operated in parallel. Reduced
short circuit current may be desirable for reduced stress on equipment and reduced
arc flash hazard in case of a fault. The ideal collector substation location is within a
central area of the WEP to optimize the cost of collector lines and the efficiency of the
plant. In practice the final location of the collector substation may depend on other
factors, such as soil conditions, excavation requirements, land constraints, and the
location of the POI.
6
Utility grid Utility grid Utility grid
52 52 52
HV HV HV
Zig-Zag Zig-Zag
TX T1 T2 TX
Tie Breaker
MV Bus MV Bus
52-T
52-1 52-2 52-3 52-1 52-2 52-3 52-4 52-5 52-6
Collector 1 Collector 2 Collector 3 Collector 1 Collector 2 Collector 3 Collector 4 Collector 5 Collector 6

(a) (b)
Figure 3: Common collector substation arrangements (MV bus configurations)

for WEPs: (a) single-bus configuration and (b) sectionalized-bus configuration
While the MV buses of WEPs are primarily of the single and sectionalized-bus type,
the high-voltage (HV) buses could be of five common configurations: single,
sectionalized, ring, breaker-and-a-half, and double-breaker double-bus. Figure 4
shows these five configurations. More details about these five arrangements can be
found in IEEE Std. 666 [11]which discusses various bus configurations for generating
stations. Table 1 compares the cost and reliability of each bus configuration. The table
shows that the double breaker-double bus and breaker-and-a-half configurations
have the highest reliability, albeit at the highest cost. The bus configurations used in
the MV bus design have relatively low reliability as compared to the more
complicated HV bus configurations.
7
Utility Utility Utility Utility Utility
52 52 52 52 52
52 52 52 52 52
WEP WEP 1 WEP 2 WEP 1 WEP 2

Transformer HV Transformer HV Transformer HV Transformer HV Transformer HV
(a) (b) (c)
Utility Utility Utility Utility
52 52
52 52 52 52
52 52
52 52 52 52
52 52
WEP 1 WEP 2
Transformer HV Transformer HV
WEP 1 WEP 2
Transformer HV Transformer HV
(d) (e)
Figure 4: Different HV bus configurations: (a) single bus configuration, (b)

sectionalize bus configuration, (c) ring configuration, (d) Breaker-and-a-Half
configuration, and Double Bus–Double Breaker configuration
Table 1: Cost/Reliability Comparison for Different Bus Configurations [9]
Reference Bus Configuration Relative Relative Application

Cost Reliability HV MV
Fig. 4(a) Single 46.7% Low X X
Fig. 4(b) Sectionalized 57.0% Low X X
8
Fig. 4(c) Ring 53.3% Medium X

Fig. 4(d) Breaker-and-a-Half 73.8% High X
Fig. 4(e) Double Bus–Double 100% High X
Breaker
2.3 Wind Electric Generator Characteristics

Wind turbines can be classified by their mechanical power control as stall regulated
or pitched regulated. According to speed control, wind turbines can be divided into
fixed speed (Type 1), limited variable speed (Type 2), variable speed with partial
power electronic conversion (Type 3), variable speed with full power electronic
conversion (Type 4), and variable speed with mechanical speed/torque converter
(Type 5). These different types are shown in Figures 5 to 9. [9] IEEE PES PSRC
Working Group C17 produced a report [1] about WTGs and their respective response
to faults with respect to each Type of WTG. A brief summary of the characteristics of
the five WTG types is provided below, based in part on the C17 Working Group report
[1] and [16].
2.3.1 Type 1
A typical Type 1 WTG is a Squirrel Cage Induction Machine (SCIM). This induction
machine is typically connected to the collector system through a step-up transformer
and soft starter2. Power factor correction capacitors (PFCC) are typically included and
divided into different steps or stages that are switched in or out during differing
operating speeds of the turbine shaft. The speed of the turbine shaft is controlled to a
near constant value (typically 2-3% faster than the grid’s synchronous frequency).
Therefore, the WTG generates real power when the turbine shaft rotates faster than
the electrical grid frequency.
2
The main function of the soft starter is to enable synchronization to the grid without massive inrush
currents on the generator.
9
Figure 5: Typical configuration of a Type 1 WTG
2.3.2 Type 2
A typical Type 2 WTG is an induction machine having an external resistance inserted
into the rotor circuit to provide operation over a wider range of slips when compared
to a Type 1 induction machine. This induction machine is typically connected to the
collector system through a step-up transformer and soft starter. Power factor
correction capacitors (PFCC) are, again, typically included and divided into different
steps or stages that are switched in or out during differing operating speeds of the
turbine shaft. Like the Type 1 WTG, the speed of the turbine shaft of the Type 2 WTG
is controlled to a near constant value (typically up to 10% faster than the grid’s
synchronous frequency).
2.3.3 Type 3
A typical Type 3 WTG is also called the Doubly Fed Induction Generator (DFIG) or
Doubly Fed Asynchronous Generator (DFAG). An AC-DC-AC converter provides
variable frequency ac excitation to the rotor circuit to enable the WTG to operate at
variable speeds (typically ± 30% of the synchronous speed), and to provide reactive
power control and ac voltage regulation capabilities. The power factor of a Type 3
WTG typically ranges from 0.90 (lagging capacitive) to 1.0 (unity) to 0.9 (leading
inductive) at rated active power generation.
10
2.3.4 Type 4
A typical Type 4 WTG is composed of an electrical machine interconnected to the
collector system through a full-scale back-to-back (AC-DC-AC) frequency converter.
The electrical machine of this wind turbine type may use a synchronous machine
excited either by permanent magnets or by an asynchronous machine. In contrast to
Types 1-3, the generator of a Type 4 WTG is completely decoupled from the grid, so a
gearbox may not be required. If an asynchronous generator is used, a gearbox is often
included in the design. The electrical output of a Type 4 WTG is completely defined
by power electronics, that is, the full-scale converter, and not the inherent behavior
of the generator. This design allows Type 4 WTGs to rotate at an optimal aerodynamic
speed providing extreme flexibility in generation in combination with excellent grid
integration characteristics such as flexible reactive power capabilities and a wide
voltage and frequency operating range. Like the Type 3, the power factor of a Type 4
WTG typically ranges from 0.90 (lagging capacitive) to 1.0 (unity) to - 0.9 (leading
inductive) at rated active power generation.
2.3.5 Type 5
A typical Type 5 WTG consists of a WTG variable-speed drive train connected to a
speed/torque converter coupled with a synchronous generator. A Type 5 WTG
exhibits typical synchronous generator characteristics and behavior during faults.
11
Therefore, generator contributions to faults can be calculated from the generator

machine constants provided from the respective generator manufacturers.
2.4 Fault Currents and Equipment Ratings

To choose the ratings of equipment in the substation of a generating plant, maximum
fault currents need to be calculated. For selection and settings of protective devices
both maximum and minimum fault currents need to be calculated. For conventional
generating stations, short circuit models of synchronous generators are well known,
and the currents can be found using phasor domain short circuit programs that model
the generators as Thevenin models with appropriate impedances for sub-transient,
transient, or steady state time frames. The worst-case dc offset is also known and can
be worked into the fault current calculations.
For wind electric plants, the fault currents for a given fault on the collector circuit or
at the point of interconnection will depend on the type of WTGs employed in the plant.
While Type 1 and 2 WTGs can usually be represented by a Thevenin model, Type 3
and 4 exhibit unconventional behavior. The fault contribution from these machines
depends heavily on the proprietary controls implemented in their converters. A Type
3 machine can switch back and forth between crowbarred mode and controlled mode,
whereas a Type 4 is fully controlled, limiting fault currents to values comparable with
load currents (110% of load current, for example) within approximately two cycles.
Phasor domain short-circuit programs can model these machines as a voltage-
dependent current source. An iterative method of solution can then be used to
account for the nonlinear fault response caused by the converter controls. Additional
12
information on WTG modeling is available in the PSRC Working Group C24 report
titled, “Modification of Commercial Fault Calculation Programs for Wind Turbine
Generators” [33].
In the absence of precise fault models, the maximum fault current from a Type 3
machine can be calculated assuming it is crowbarred, which enables it to be modeled
as a Type 2 machine. For a Type 4, the maximum fault current will have to be provided
by the manufacturer for uncontrolled as well as controlled modes. Conversely, the
minimum fault contribution for Type 3 and 4 machines will have to be obtained from
the manufacturer. The minimum fault current will depend on the number of WTG
units in operation at the time of fault, and the fault location.
An important feature of the fault responses of Type 1, 2 and 3 (crowbarred) WTGs is

the amount of dc offset in the fault current. Studies show that the dc offset in
combination with ac decrement can cause the resulting ac current to be without a
zero-crossing for several cycles [1]. However, studies also show that a very small
value of resistance in the fault current path can result in very quick decay of the dc
component [5]. Due to the resistance of cables, pad mounted transformers, and the
resistance of the arc between circuit breaker contacts, in practice the dc offset does
not pose a serious challenge to the operation of circuit breakers.
2.5 System Grounding

Wind Electric Plant system grounding is an important consideration in plant design
because it directly leads to the determination of protection schemes that can be
applied to maximize the safe and reliable operation of this generating asset.
Moreover, the different types of wind turbines (Type 1, Type 2, Type 3, Type 4, or
Type 5) have different characteristics that cause individual WTG types to contribute
differently under various fault conditions. For example, for many short-circuit
studies, Type 1 and Type 23 generators can be modeled as an equivalent Thevenin
impedance such that they can contribute positive, negative, and zero sequence
current to a fault. The sequence current contributions from Type 3 and Type 4
generators are a function of the electrical characteristics of their converters. Both can
contribute positive and negative sequence current, but a Type 4 generator does not
produce zero sequence current. Type 5 generators have the same characteristics as a
synchronous generator. Therefore, the WTG characteristics are an important
consideration when applying the protection scheme design for both the generator
and its associated WTG step-up transformer.
3Depending on the study and the WTG’s crowbar system, replacing Type 2 WTGs with a Thevenin
equivalent may not be possible.
13
The secondary (low voltage or generator) side of the WTG step-up transformer is
typically connected in grounded wye to provide a stable reference point for the
system phase to neutral voltages to enhance equipment and personnel safety. The
grounded wye connection also causes the majority (or all) of the current for a ground
fault on the WTG and low voltage bus to come from the GSU transformer where the
ground fault current magnitude is usually large enough to operate simple phase and
ground overcurrent or fuse protection equipment. A primary reference for the
application of grounding techniques in North America is IEEE Standard 142-2007:
IEEE Recommended Practice for Grounding of Industrial and Commercial Power
Systems (also known as the “IEEE Green Book”), and subsequent approved
updates/revisions of this standard4. According to this standard, solid grounding is
generally recommended for the following:
a) Low-voltage systems (600 V and below) where automatic isolation of a faulted
circuit can be tolerated or where capability is lacking to isolate a ground fault
in a high-resistance grounded system.
b) Medium- or high-voltage systems (above 15 kV) to permit the use of
equipment with insulation levels to ground rated for less than line to line
voltage.
c) Medium- or high-voltage applications where higher ground fault current
magnitudes are required to provide selective ground-fault detection on
lengthy distribution feeders.
Bullet b) from the IEEE 142 excerpt above refers to the use of equipment with phase
to ground insulation levels on systems that are “effectively grounded”. Once this
determination is made, the designer typically evaluates the system to verify that for
all foreseeable conditions, including fault conditions, the system remains “effectively
grounded”. For example, the zero-sequence network impedance of a Type 4 WTG is
effectively infinite. Therefore, utilization of a wye-ground to wye-ground GSU for
interconnection of a Type 4 unit to the collector system will not maintain an
effectively grounded collector system when a ground fault is present, and the main
substation feeder breaker has tripped to clear the fault. After the breaker opens, the
WTGs may continue to energize the collector feeder until it shuts down from loss of
load. Such a condition can lead to the development of transient overvoltage (TOV) on
the collector circuit, if other methods of remediation are not undertaken.
Many WTG step-up transformers supplying underground collector feeders are
connected in grounded wye on the WTG side and delta on the collector feeder side of
the transformer. In that case, a permanent ground fault on the collector feeder may
4 At the
time of this report, IEEE Standard 3003.1-2019 is expected to replace IEEE Standard 142-2007
Grounding of Industrial and Commercial Power Systems.
14
also cause a TOV condition after the main collector feeder trips and the WTGs are
shutting down. Utilization of a properly sized ground bank on the collector feeder
circuit, 3E0/3V0 protection, grounding breakers, or direct transfer trip of the
generation generally provides effective means of such control.
2.6 Transformer Connections and Characteristics

The section discusses the considerations associated with various interconnection
transformer configurations.
2.6.1 Wind Turbine Generator Transformers
Most WTGs produce AC power at 690 V or lower for generators smaller than 3 MW,
and 3.3 or 6 kV for larger generators. A collector system network consolidates the
power produced by the individual WTGs in the wind electric plant. This collector
system network is typically operated at 33 - 34.5 kV, although for smaller plants, the
collector system network might be operated at 12 kV. In either case, transformers will
be required between the WTG and the collector system. The WTG is typically wye
connected but ungrounded. For this reason, the WTG transformers are normally
configured to provide a ground reference for the low voltage system. It is very
common to use delta - wye transformers with the wye on the low voltage side. In these
cases, the neutral on the wye may be connected to earth. Although this is a common
configuration for the WTG transformers, wye - wye transformers have also been used,
with the neutral on one or both sides of the transformers connected to earth.
The WTG transformers are typically in one of two locations: The WTG nacelle or on
the ground next to the tower. Many of the larger WTGs will have the transformers
located in the nacelle. The nacelle is the housing on the top of the tower which
contains the generator and inverters. This design will reduce the losses and the size
of the cable needed for the run down the tower. In this design, a medium voltage
breaker is typically installed to provide protection for the transformer. The other
configuration, which is typically used for WTGs smaller than 3 MW, has the
transformer on the ground next to the tower. In this design, fuses are typically
installed to provide the transformer protection.
2.6.2 Main Substation Transformers

For most transmission network interconnections to wind electric plants, the collector
network voltage is not the voltage of the transmission network, so another
transformation is required. Since the WTGs and their transformers typically do not
provide ground references to the MV network, to be able to use equipment that is
15
designed for an effectively grounded system it is necessary for the main substation
transformer to provide a ground reference to the MV system. Depending on the
transmission network connected to the plant, the transmission provider may require
that the main substation transformer also be a ground reference to the transmission
network. If a ground reference is not required by the transmission provider, then a
delta - wye grounded transformer with the wye on the MV side will be an adequate
transformer. If it is a requirement that the main substation transformer also be a
ground reference to the transmission network, then there are a couple of options. The
most commonly used transformer configuration is a wye - wye transformer with a
delta tertiary winding. Both HV and MV wye neutrals are connected to the earth. This
type of configuration is shown in Figure 10. The other option is to use a delta - wye
transformer in conjunction with a grounding transformer. The grounding
transformer would be connected on the transmission network side where the main
transformer delta winding is connected. The grounding transformer can be a two-
winding wye - delta transformer or a zig zag grounding transformer.
Figure 10: Simplified One Line Diagram of Wind Electric Plant
16
2.7 Harmonics and Sub-harmonics

The harmonics due to wind plants will depend on converter topology, applied
harmonic filters, and short-circuit current at the Point of Common Coupling (PCC).
Even harmonics in wind generation can arise due to unsymmetrical half waves and
may appear at fast load changes. Sub-harmonics are spectral components whose
frequencies are lower than the fundamental power frequency. Sub-harmonics can be
produced due to periodical switching with variable frequency or arise due to
interaction between wind generation and series compensated transmission lines.
Sub-harmonics can also be created due to various phenomena such as induction
generator effects, control interactions, torsional interactions, and torque
amplification.
When a series capacitor offsets partial system reactance, the system will have a
natural frequency less than the system frequency, which is referred to as sub-
harmonic. The generator armature sub-harmonic currents produce magnetic fields
with the natural frequency, which induce currents in the generator rotor. This rotor
current frequency will be the difference between the fundamental frequency and the
natural frequency. This new added rotor current frequency will cause sub-
synchronous armature voltages that may enhance the sub-synchronous currents and
may cause generator self-excitation. Because the rotor turns faster than the sub-
harmonic armature currents, a slip is created to simulate an induction machine. This
is called the Induction Generator Effect (IGE), where the rotor resistance of the
machine presents itself as a negative resistance with respect to sub-synchronous
frequencies. The system presents itself as a positive resistance at the system natural
frequencies, but if the negative resistance is greater than the positive resistance, then
sub-synchronous currents will be sustained and possibly amplified, thus causing
serious voltage and current oscillations. Generally, the greater the number of
turbines, the lower the magnitude of the harmonics and sub-harmonics, especially of
the lower order.
The energy portfolios of energy providers around the world are increasingly
comprised of renewable energy sources like wind and solar. As the size of these
individual energy facilities increases, the rated capacity of the connected
transmission systems will also increase. The transmission expansions to wind electric
plants are often radial in nature, and to maximize return on investment, they are
usually sized to carry no more power than is required for a wind power facility [8].
Series capacitor compensation in AC transmission systems is an economical means to
increase the load carrying capability of a line, to control load sharing among parallel
lines, and to enhance transient stability [14]. However, capacitors in series with
transmission lines may cause sub-synchronous resonance (SSR) with generators near
17
a series capacitor. Sub-Synchronous Resonance occurs when the mechanical mass of

the generator shaft resonates with the effective impedance of the system, which can
lead to turbine-generator shaft failure and electrical instability at oscillation
frequencies lower than the normal system frequency. Therefore, it is important to
fully understand the effects of SSR and to analyze them when planning series
capacitor compensation in power systems.
It is important that interconnection studies for any generator or power electronic
equipment in the vicinity of a series capacitor can account for system configurations
that may give rise to sub‐synchronous oscillations. Any generation or power
electronic equipment connected in a radial series‐compensated configuration is
particularly susceptible to the phenomena of Sub-Synchronous Control Interaction
(SSCI). It is also important to study these phenomena whenever series compensation
is being added to a transmission system. Sub-Synchronous Control Interaction is a
relatively new problem which has been observed between power electronic devices,
such as an HVDC link, a static VAR compensator, or a wind turbine and a series
compensated system [13]. A typical SSCI event is a condition where the wind
turbines/controls interact with nearby series capacitors, resulting in un-damped/fast
growing oscillations. This is a widespread and serious problem, affecting turbines
from most major manufacturers. Furthermore, standard transient stability studies
performed as per the interconnection standards do not show this problem because
they use fundamental frequency phasor solutions. The negative damping observed in
the SSCI phenomena is largely due to the wind turbine rotor side current controller
in the Type 3 topology. In all cases of potential SSCI, it is highly recommended that
detailed electromagnetic transient models of the turbine be obtained directly from
the manufacturer. Modern power system study tools can integrate code from the
actual WTG controls into simulation models such that excellent predictions of SSCI
phenomena may be produced. Increased awareness of SSCI is critical, along with
backup protection schemes capable of preventing equipment damage in the event of
SSCI.
To evaluate wind and transmission projects for possible sub-synchronous interaction
issues when used in conjunction with series compensated transmission systems, it is
helpful to consider the following recommendations:
• When possible, use detailed EMT models of the WTG unit obtained directly
from the manufacturer during the project planning stages to provide a
complete understanding of WTG performance, and to evaluate potential
problems. The best models directly use the real code from the actual WTG
controller hardware.
• Wind turbine manufacturers are encouraged to aggressively research SSCI
and see if improvements can be made in their turbine controls.
18
• It is important for wind developers to be aware of potential problems of WTGs

operating near series compensated lines and to be aware of the potential
complexities this introduces.
• Transmission providers are encouraged to update their interconnection
standards to require EMT/SSCI analysis to be performed when WTGs are
connected near series compensated lines, and to plan for proper protection
systems to be in place.
• Transmission planners are encouraged to include the additional costs for SSCI
related problems when comparing lines with series capacitors to higher
voltage AC lines, or DC solutions.
• It is important to perform sufficient studies that identify locations that are
susceptible to isolation of generation or power electronic equipment with
series compensated transmission lines. In addition, integration of wind farms
into systems where the short circuit ratio (SCR) is low (<5) may be identified
as potential locations for SSR (SCR = Ssc / Pwind).
Appropriate transmission system design enhancements need to be considered when
studying integration of large-scale wind farms. Some measures that may be
considered include:
• Limiting series compensation to safe levels.
• Installing thyristor-controlled series capacitors.
• Installing SSR blocking filters.
• Installing additional protection systems to detect SSR and take corrective
action.
• Installing additional protection systems to avoid SSR based on system
topology.
• Exploring alternatives to increase system strength.
• Exploring modifications to turbine control technology.
2.8 Voltage and Frequency Control Requirements

The voltage ride-through (VRT) capabilities of WTGs vary based on requirements
specified by various regulating authorities. For example, regulating authorities in the
United States require wind plants to ride-through a three-phase fault on the high side
of the substation transformer for up to 9 cycles, depending on the primary fault
clearing time of the circuit breakers at the location (Federal Energy Regulatory
Commission Order 661-A [19]). Additional regulating authority requirements may
define voltage and frequency protective relay settings to maximize a generating unit’s
ability to remain connected during a voltage or frequency excursion. It is important
19
for the protection engineer to verify that all proposed relay settings comply with any
standards established by area regulating authorities.
Some of the Type 1 WTGs have limited VRT capability and may require a central
reactive power compensation system to meet wind power plant VRT capability. Many
of the Types 2, 3, and 4 WTGs have certain VRT capabilities. None of WTG Types 1-4
have frequency control capability, but some WTG vendors can provide data on the
frequency ride-through capability of their machines. The Type 5 WTG is similar to
that of a standard grid-connected synchronous generator and therefore has both
voltage and frequency control and ride-through capability.
3 Typical Protective Relay Schemes at Wind Electric Power

Plant Substations
3.1 Collector Feeder Protection
Collector feeder circuit operation exhibits both radial and network characteristics.
When the WTGs are off-line, they draw only relatively low load current required by
their station auxiliary equipment (lights, heaters, turbine gear motors, etc.). This
station service current for the WTGs flows radially from the Collector Substation to
the WTGs on the line, and system fault current flows exclusively from the Collector
Substation to the fault location on the line or at a WTG. When the WTGs are on-line,
their collective output current flows from the WTGs to the Collector Substation, and
with multiple generation sources, the collector feeder operates as a network. If a fault
occurs on the line or at a WTG, most of the fault current comes from the Collector
Substation, but a small component of the total fault current does come from any WTGs
that are online at the time of the fault. Therefore, relay protection schemes that are
commonly applied for either radial or network distribution circuits may be
considered. Non-directional overcurrent protection schemes have traditionally been
applied for radial distribution line protection.
Directional over-current protection schemes have traditionally been applied for
protection of network or looped circuits, as well as directional distance relay
protection schemes. Ideally, the protection scheme applied at the main feeder breaker
at a WEP station provides primary protection for the collector feeder and backup
protection for the step-up transformer at each WTG location. The inverse time-
current characteristic of a time overcurrent relay can be selected to coordinate with
the thermal damage curve of a line conductor or transformer as well as with
overcurrent protective devices such as transformer fuses. Directional distance relays
can be set to provide excellent primary feeder protection by utilizing two or more
“zones” of collector feeder protection using a fixed impedance reach and definite
delay setting for each zone. However, to also provide backup protection for the WTG
20
transformers at more than two or three values of fault current, a multiplicity of

distance zones and time delays would be required. Therefore, non-directional and
directional overcurrent schemes are typically applied for collector feeder protection.
The response of the collector feeder relays is coordinated with other transmission-
side overcurrent relays and feeder side fuses or relayed fault Interrupters protecting
the generator step-up transformers. In some cases, it may be preferable to coordinate
the collector feeder relays with the protective device on the low voltage side of the
WTG GSU transformer. This approach can provide faster clearing times or higher
sensitivity on the feeder, which may reduce cable costs, or reduce the hazard
associated with arc flash. In this case, directional overcurrent relays can usually be
applied and set with enough sensitivity to detect a fault on the low voltage side of the
WTG GSU transformer. Additionally, a mix of directional and non-directional
overcurrent relays may be used to discriminate between faults on the transmission
system and faults on the feeder circuit.
Normal current flow at the wind plant is from the feeder towards the transmission
grid, so any significant current towards the generators on the feeder (that is, current
greater than auxiliary loads when the generators are not running or inrush current
when energizing the feeder) would be indicative of a fault. Fault currents with
direction toward the feeder are normally cleared promptly to minimize equipment
damage and personnel hazards. For faults with direction toward the transmission
system, the collector feeder protection may operate in a backup mode. In this case,
the feeder protection relay’s response may be delayed to allow the primary
protection schemes for the transformer or the transmission line to operate first, but
still be fast enough to prevent damage to the collector feeder and transmission system
equipment if those primary protection schemes do not operate as intended.
3.1.1 Overcurrent Protection and Coordination with WTG Transformer
Protective Devices
Overcurrent protection of collector feeder circuits typically include the application of
both instantaneous and time delay phase and ground overcurrent elements. Negative
sequence overcurrent elements can also be applied for additional protection and to
satisfy specialized device coordination requirements. In many cases, simple non-
directional overcurrent protection can be used to provide collector feeder protection.
However, collector feeders consisting of long cable runs between many high-output
wind turbine generators (WTGs) may require directional overcurrent elements to
provide adequate protection.
3.1.1.1 Non-directional Overcurrent Protection
Non-directional overcurrent protection can be applied to collector feeders for which
the fault current contribution from the WEP substation bus is much higher than the
combined fault or load current contribution from the WTGs on the collector feeder
21
circuit to be protected. In such cases, the current pickup level of the phase time
overcurrent (51P) element or relay at the substation collector feeder breaker can be
set at some factor times the combined output current capability of the WTGs. The
selection of the 51P current pickup level may be governed by requirements
established by various regulating authorities. The curve type and time dial are then
selected to coordinate with the expulsion fuse on the medium voltage side of the WTG
step-up transformer.
For example, on a collector feeder in the United States consisting of eight 3.06 MW at
0.9 power factor wind turbine generators, the full load current IFLA of all 8 generators
at the 34.5 kV main feeder breaker is:
8 × 3.06𝑀𝑊 × 1000
𝐼𝐹𝐿𝐴 = = 455 𝐴𝑚𝑝 (34.5𝑘𝑉 𝑠𝑖𝑑𝑒)
0.9 × 1.732 × 34.5𝑘𝑉
Since the 51P relay is non-directional, the area regulating authority requires its trip
current setting to be greater than or equal to 1.3 x IFLA for asynchronous generators
and inverters (North American Electric Reliability Corporation Standard PRC-025:
Generator Relay Loadability [20]). Moreover, since the 51P and its feeder breaker are
on the 34.5 kV side of the 138/34.5 kV Main Substation GSU transformer, PRC-025
Option 5a requires the setting to be calculated for a generator bus voltage
corresponding to 1.0 per-unit of the high-side nominal voltage times the turns ratio
of the GSU transformer. If the turns ratio of the GSU is 34.5 kV/139.725 kV, then the
corresponding generator bus voltage Vgen for a 1.0 per-unit nominal voltage of 138 kV
is:
1.0𝑝𝑢 × 138𝑘𝑉 × 34.5𝑘𝑉
𝑉𝑔𝑒𝑛 = = 34.07𝑘𝑉
139.75𝑘𝑉
Therefore, the 51P pickup setting (51PP) can be made be greater than or equal to:
1.3 × 8 × 3.06𝑀𝑊 × 1000
51𝑃𝑃 = = 599𝐴𝑚𝑝 (𝑝𝑟𝑖𝑚𝑎𝑟𝑦)
0.9 × 1.732 × 34.07𝑘𝑉
If the CT ratio on the feeder is 1200:5, then the current setting for the 51P tripping
element can be 2.5 A, secondary = 600 A, primary.
Wind turbine generator step-up transformers may be equipped with an expulsion
fuse in series with a current limiting fuse on each phase of the high voltage side of the
transformer. A low-voltage breaker equipped with an overcurrent trip device may be
applied to protect the low voltage cables and auxiliary equipment. The WTG vendor
usually sizes both the transformer high-side fuses and the low voltage breaker and
applies trip device settings for the breaker that will coordinate with the high-side
22
fuses. Therefore, the time overcurrent relays at the main collector feeder breaker can
be coordinated with high-side fuses of the WTG step-up transformer. The 3.5 MVA
WTG step-up transformers in the above example are protected by a 100 A current
limiting fuse in series with a 71 A expulsion type fuse. The coordination between the
phase TOC relay at the Main Feeder Breaker and the high-side fuses for a three-phase
fault at the first WTG on the circuit appears in Figure 11.
The chief disadvantage of the non-directional phase time overcurrent setting is that
its current pickup setting may be 8 – 12 times the full load capability of any individual
WTG on the circuit, resulting in loss of sensitivity for low-current faults. For example,
51P relay curve (11F4-50/51P) in Figure 11 provides good collector feeder
protection for the smallest conductor size on the feeder, but marginal protection for
the WTG step-up transformer because the relay trip current setting is approximately
ten times the full load current capacity of a single WTG. However, the high trip current
setting of a 51P relay enables it to operate securely, regardless of the direction of
current flow on the collector feeder. For this reason, non-directional phase time
overcurrent protection schemes may be applied on collector feeders for which the
system fault current is much larger than the combined output current of the WTGs on
the circuit.
Non-directional instantaneous phase overcurrent (50P) elements can be set to
provide instantaneous tripping for faults on the main collector feeder circuit. Ideally,
the current pickup setting of the 50P element is set above the combined inrush
current of the WTG step-up transformers on the circuit but low enough to see a 3-
phase feeder fault at the farthest WTG on the circuit. Transformer inrush currents can
be on the order of 8 – 12 times the full load current, based on the nameplate rating of
the transformer. For example, the combined full load current of the 8 WTG
transformers in the previous example is (8)(3.5 MVA) / (1.732)(34.5 kV) = 469 A. If
the combined transformer inrush current is considered to be 10 times the full load
current, then the combined transformer inrush current is 4690 A. If the fault current
seen at the main feeder breaker for a 3-phase fault at the end of the collector feeder
circuit is 10,520 A, then the 50P element may be set below 10,520 A but above 4690
A. If a 50P setting of approximately 80% of the 3-phase fault current at the end of the
feeder is applied, then the 50P element could be set to trip at approximately
0.8(10,520) = 8416 A. If the CT ratio is 1200/5, then the 50P element could be set just
below 8416/240 = 35.07 A, secondary. Therefore, a 50P setting of 35 A, secondary (=
8400 A, primary) could be applied. In cases where the transformer inrush current
exceeds the instantaneous trip current setting required to detect a fault at the end of
the collector feeder circuit, harmonic current blocking can be used to inhibit the
phase instantaneous overcurrent element from tripping for transformer inrush [4].
Alternatively, since the highest transformer inrush current is usually within the first
3 - 5 cycles upon energization, the trip delay of the phase instantaneous overcurrent
23
element can be increased from instantaneous to a delay of 10 - 12 cycles to bypass

the initial transformer inrush current time duration.
It is important to note that a 50P setting based on a percentage of the fault current
available at the end of the feeder may cause the main collector feeder breaker to trip
as the current limiting fuse blows for a fault between any WTG transformer and its
associated fuse. This can be remedied by setting a combination of instantaneous and
definite-time overcurrent elements. The instantaneous 50P element can be set to
provide instantaneous tripping for feeder faults between the main collector feeder
breaker and the first WTG on the circuit. The additional definite-time phase
overcurrent elements would be set at some percentage below the maximum fault
current available at the last WTG on the circuit. Then, an appropriate time delay to
provide coordination with the thermal damage curve of the conductor and to ride
through the combined transformer inrush for the circuit would be applied.
24
Figure 11: Non-Directional Phase Relay Time Overcurrent Coordination with

WTG Fuses
The relatively poor sensitivity of the phase time overcurrent protection scheme can
be mitigated for unbalanced faults by applying a non-directional negative sequence
time overcurrent element (51Q). Because a 51Q relay operates on I2 or 3I2, it can be
set below balanced 3-phase load current. This enables the user to set the minimum
25
operating current level of the 51Q relay more sensitively than that of the 51P, thus
providing a better match with the minimum current at which the WTG transformer
high voltage side fuse begins to melt. For coordination purposes, the 51Q element
may be considered as an “equivalent” phase overcurrent element. Pickup, curve type,
and time dial settings can then be derived for the “equivalent” phase overcurrent
element and coordinated with downstream phase overcurrent devices. After the
coordination is complete, the “equivalent” phase overcurrent pickup setting can then
be multiplied by the appropriate factor to convert it to a negative sequence pickup
setting for the 51Q in terms of 3I2 or I2. [27]. With the lower trip current setting, a
curve type can then be selected whose slope better matches the slope of the fuse
characteristic curve, and a time dial can be selected to provide a closer margin of time
between the 51Q and the high-side transformer fuse curve over a broader range of
unbalanced fault current.
In the previous example the full load capacity of one of the eight WTGs is 57 A.
Therefore, the equivalent minimum trip current setting of the 51Q can be set below
the 455 A of all 8 WTGs but somewhere above the 57 A of a single WTG. Therefore, a
setting of 130% of the full load capacity of one WTG can be established as the basis
for the minimum equivalent phase current sensitivity (Im) of the 51Q, where:
Im = 1.3(57) = 74 A.
However, the 71 Amp expulsion fuse will not blow at currents lower than about 140
A (approximately 2 times the continuous current rating of the fuse). Therefore, the
minimum equivalent phase current pickup of the 51Q can be no lower than 140 A.
Given the 1200/5 CT ratio, an equivalent phase pickup setting of 140 A primary would
be 0.58 A, secondary. Therefore, the equivalent phase pickup setting for the 51Q can
be 0.6 A, secondary (= 144 A, primary). If the 51Q operates on 3I2, then the 3I2 setting
for the 51Q relay would be 1.732(144 A) = 249 A, primary [27]. Given the 1200/5 CT
ratio, the 3I2 setting for the 51Q relay can be rounded up slightly to 1.1 A, secondary
= 264 A, primary to account for the protective relay precision.
The coordination plot of Figure 11 with a 51Q time-current characteristic (11F4-
50/51Q) set with the trip current pickup calculated above appears in Figure 12. The
coordination between the 51Q and the fuse curves is based on the equivalent phase
current pickup for the 51Q. After coordination is complete, the actual 3I2 pickup
settings can be applied to the 11F4-50/51Q in the field. The time curve and time dial
settings obtained from the equivalent phase current coordination can be applied
directly to the 51Q element. The 11F4-51Q curve provides better coordination with
the expulsion fuse curve for unbalanced faults that are too low for the 51P (11F4-
50/51P) relay to detect. A large proportion of faults on a power system are
unbalanced faults, and the 51Q element (11F4-50/51Q) provides good collector
feeder protection as well as backup protection for the individual WTG step-up
26
transformer during unbalanced faults. However, a 51Q relay will not operate for
balanced 3-phase faults. Faults on underground collector feeder circuits that start as
an unbalanced fault can quickly involve the un-faulted phases and become a 3-phase
fault.
Non-directional instantaneous negative-sequence overcurrent (50Q) relays or
elements can also be set at the user’s discretion to provide instantaneous tripping for
faults on the main collector feeder circuit. As with the 50P, the current pickup setting
of a 50Q element may be set above the combined inrush current of the WTG step-up
transformers on the circuit, but low enough to see a phase-phase feeder fault at the
farthest WTG on the circuit. For example, if the fault current seen at the main feeder
breaker for a phase-phase fault at the end of the collector feeder circuit is 8985 A, and
the combined WTG transformer phase inrush current is 4690 A, then the equivalent
phase pickup of the 50Q element may be set below 8985 A but above 4690 A. Given
the 1200/5 CT ratio, a pickup setting of 25 A, secondary (= 6000 A, primary) could be
selected for the equivalent phase pickup setting of the 50Q. The final setting based on
3I2 would therefore be 1.732(25 A) = 43.3 A, secondary (=10,393 A, primary). It is
important to note that transient negative sequence currents may appear during
breaker switching operations. Therefore, the 50Q element may need to be delayed 1.5
– 2 cycles to allow the transient currents to subside.
27
Figure 12: Non-Directional Phase and Negative Sequence Relay Coordination

with WTG Fuses
Since the WTG output to the substation feeder breaker is essentially balanced, the
current pickup of the neutral or residual ground time overcurrent (51N or 51G)
element can be set relatively low, typically 10 – 30% of the phase time overcurrent
setting. The curve type and time dial are then selected to coordinate with the
expulsion fuse on the medium voltage side of the WTG step-up transformer. An
example of the coordination between a 51G relay (11F4-50/51G) and the high-side
28
fuses of a WTG transformer appears in Figure 13. In Figure 13 the trip current setting
of the 51G relay is greater than the approximately 140 A current level at which the 71
A expulsion fuse begins to blow. A 0.6 A, secondary = 144 A, primary setting for the
51G places the 51G trip current setting at approximately 24% of the 51P trip current
setting.
Figure 13: Non-Directional Ground Time Overcurrent Relay Coordination with

WTG Fuses
29
Non-directional instantaneous or short time neutral or residual ground overcurrent

(50N or 50G) elements can be set in much the same way as the phase elements. To
provide instantaneous tripping for single phase to ground faults over the full length
of the collector feeder, the 50G element may be set low enough to see a phase-to-
ground feeder fault at the farthest WTG on the circuit, typically some percentage of
the fault current value, but high enough to avoid potential mis-operation from any CT
or system impedance unbalances during initial energization of the collector feeder
circuit. If, for example, the fault current seen at the main feeder breaker for a single
phase to ground fault at the end of the collector feeder circuit is 7200 A, then a 50G
setting of approximately 80% of 7200 A would be 24 A, secondary = 5760 A, primary.
This setting can be expected to be well above any zero-sequence current that would
occur during collector feeder transformer inrush. Alternatively, a high-set
instantaneous ground element can be set to detect a feeder fault up to the first WTG
transformer on the feeder. A second ground overcurrent element can then be set to
detect a phase-ground feeder fault at the farthest WTG on the circuit with a time delay
to avoid thermal damage to the cable and to allow initial energization of the GSU
transformers on the circuit.
3.1.1.2 Directional Overcurrent Protection
Directional phase time overcurrent (67P) elements can be set to provide increased
sensitivity and speed. Directionally supervising the overcurrent element may provide
the following features:
• Provide better coordination margins in systems where the fault current
contribution from the WEP substation bus is similar order of magnitude to the
full load current with all WTGs online
• Allow detection of faults on the low-voltage side of the WTG transformers
• Faster fault clearing times on the collector feeder, potentially reducing cable
damage and cost
• Assist mitigation of arc-flash hazards
When the WTGs are online, the direction of power flow is from the WTGs to the
substation main bus. The 67P relays operating at the substation feeder breaker can
be set such that the forward direction is towards the WTGs on the collector circuit.
The minimum trip setting of the 67P can therefore be set below the combined WTG
output capacity of the collector feeder. Since the only current drawn by the WTGs in
the forward direction will be station service current drawn by the WTGs while they
are off line, the pickup current level for the directional phase time overcurrent
element can be set above the aggregate station service load of the off-line WTGs and
above the minimum melting current of the high voltage fuse at an individual WTG
transformer. The curve type and time dial are then selected to coordinate with the
30
fuse on the high voltage side of the WTG transformer, or with any low voltage trip
devices such as breakers installed on the secondary side of the WTG transformer.
In the 8-WTG collector feeder example, the output capacity of a single WTG is 57 A,
and the WTG step-up transformer is protected by a 71 A expulsion fuse. Since the
minimum melt current of the fuse is approximately 140 A, the 67P trip setting may be
set at or just above 140 A. Given the 1200/5 CT ratio, the minimum 67P trip setting
would be 0.6 A secondary = 144 A primary. A plot of the resulting coordination
appears in Figure 14, where the 67P relay is denoted as “11F4-67P”. In Figure 14 the
67P (11F4-67P) provides both collector feeder protection and backup phase
overcurrent protection for the individual WTG step-up transformers for both three-
phase and unbalanced faults.
31
Figure 14: Directional Phase Time Overcurrent Relay Coordination with WTG
Fuses
A 67P element usually determines the trip direction by comparing the line current
with the system voltage. As such, it is important to manage the protection system’s
response to an inadvertent loss of voltage on one or more phases, such as the loss of
a VT secondary fuse. Most electronic relays have loss of potential logic which either
blocks the 67P element or allows it to operate non-directionally. Non-directional
32
operation of the 67P during a high generation condition could result in a needless trip
of the main collector feeder breaker. Blocking the 67P element for loss of voltage and
relying on a backup 51P element with a higher trip current setting would be a secure
solution.
Additionally, the directional characteristic angle and sensitivity settings are also
selected to keep the 67P element from declaring a fault under certain operating
conditions. Type 3 and 4 wind turbines can supply or absorb large amounts of
reactive power at low real power outputs. This can result in an extremely wide range
of power factor outputs of the feeders. In general, when the turbines are being
operated in a voltage control mode, they will absorb reactive power when they are
lightly loaded to counter-act the effects of the capacitance of the collector system.
Conversely, the WTGs may be required to supply reactive power to the system to raise
the voltage.
With the forward trip direction of the 67P element set towards the WTGs on the
collector feeder, the characteristic angle of a conventional directional relay will be
such that the current into the feeder lags the polarizing voltage approximately by the
angle of the positive sequence line impedance. Such a characteristic may be
superimposed on the capability diagram of the wind turbine generators as shown in
Figure 15. In Figure 15, the operating area of a Type 4 WTG is shown with a maximum
current magnitude of 1.0 per unit. With zero real power production, the power
electronics could still export or import up to 1.0 per unit of reactive power covering
all of the second and third quadrants. In the case of zero real power production, the
collector bus might export a small amount of real power into the feeder to supply
auxiliary power to the WTG and possible losses in the feeder cable itself.
33
Figure 15: Conventional Phase Directional Element 67P Voltage and Current
Phasor Diagram and Operating Region of Wind Generation
It can be seen from Figure 15 that there is an area in the third quadrant and a small
area in the fourth quadrant where a possible overlap of the normal operating
conditions and the region being declared as forward by a conventional 67P relay. In
this case, one could modify the relay characteristic angle to be less than the line so
that it will have a more limited operating region in quadrant 3. The minimum
operating current could be set higher than the maximum real power supplied to the
feeder (for auxiliary power with no wind, or to cover losses in the feeder) under any
condition. However, the conventional 67P relay with fixed operating region is
generally not optimum to provide sensitive feeder protection.
Some 67P relays have independently adjustable directional characteristics that can
modify the region for forward faults to be clear of the operating region of the wind
turbine generator. An example of this characteristic is shown in Figure 16. In Figure
16, angles  and  are independently adjustable. This characteristic prevents any
operation in quadrant 3. The minimum operating current can be set to prevent
operation in small areas of quadrants 1 and 2 due to small amounts of real power that
might be supplied to the wind turbine generators during no wind periods.
34
Figure 16: Phase Directional Element 67P With Adjustable Directional

Blinders
Another type of 67P relay may have a load blinding feature that could be shaped to
restrain for any power flow into the collector bus from the feeder, including low real
power and high reactive power. One type of relay uses a positive sequence impedance
element to define the load area and block the overcurrent function if the impedance
is in the load area. Figure 17 shows an example of a load blinding function that could
be applied. The overcurrent function will be enabled only when the measured
impedance is in the first quadrant. Note that a minimum operating current setting
will prevent the relay from operating on the very high impedance which will be
measured on the right side of the RX diagram when there is no wind and a small
amount of real power is supplied to the WTG for auxiliaries and line losses.
35
Figure 17: RX diagram showing normal operating regions of WTG and example
positive sequence impedance measuring load blinding characteristic
Additional discussion of directional phase overcurrent setting considerations for

wind turbine generator operation and simulation results of directional element
operation for various modes of generator operation can be found in Appendix A.
In some electronic multi-function relays, negative sequence polarization techniques
are used to control the operation of the 67P element. The negative sequence element
settings used for directional control in typical protection applications may not
provide reliable operation for directional protection of collector feeder circuits when
Type 4 turbines installed. This is because Type 4 turbines may produce very little
negative sequence current during faults. Type 3 turbines may also provide little
negative sequence current during faults when their response is being controlled by
their electronic controls. Chen, Shrestha, Ituzaro, and Fischer [21] described an event
in which a collector feeder breaker tripped correctly on a 67P for a BCG line fault, but
the 67P elements for both un-faulted collector feeders tripped incorrectly because
the WTGs on the un-faulted feeders were not able to produce enough negative
36
sequence current to enable the feeder relay to correctly determine the direction of
the fault. The authors’ solution was to change the line impedance angle setting in all
the relays to about 10ᵒ. Lowering the line impedance angle to 10ᵒ would move the
boundary angles of the positive sequence forward directional trip zone to 100ᵒ and
280ᵒ, respectively on an impedance diagram, or 80ᵒ and 260ᵒ, respectively on a
voltage/current diagram, such as that shown in Figure 15.
It is important to understand that the solution noted in [21] may not be generally
applicable in all cases. That solution was determined for a control system with a
particular type of WTG. In general, because of the uncertain characteristics of the
negative sequence current contribution from WTGs negative sequence directional
protection is difficult to apply. If it is possible to determine the maximum level of
negative sequence current that could be contributed to an external fault (outside the
protected feeder) from the WTG, then setting the pickup of the negative sequence
directional overcurrent function higher than its minimum pickup setting would keep
it secure in the event of such an external fault. In many cases, the magnitude of
negative sequence current is intentionally minimized by the WTG controls. In such
cases, the negative sequence current from the collector bus to an internal fault would
be significantly larger than the contribution to an external fault. Therefore, a simple
non-directional negative sequence overcurrent function could be secure and
dependable. Note that it is still important for security that non-directional negative
sequence time and instantaneous overcurrent protection coordinate with the high-
side fuse of the WTG step-up transformer in a similar fashion to the 50/51G elements
as shown in Figure 13.
The directional phase instantaneous overcurrent relay or element can be given the
same pickup setting as the 50P because it will detect the same collective inrush of the
WTG transformers upon energization as the 50P. Some electronic multifunction
relays have harmonic current blocking functions. In cases where the transformer
inrush current exceeds the instantaneous trip current setting required to detect a
fault at the end of the collector feeder circuit, harmonic current blocking can be used
to inhibit the directional phase instantaneous overcurrent element from tripping for
transformer inrush [4]. Similarly, the considerations discussed above for directional
phase instantaneous overcurrent elements apply to directional neutral or ground
instantaneous overcurrent element settings.
Directional neutral or residual ground time overcurrent relays (67N or 67G) can also
be applied for collector feeder protection. However, because they are normally set to
coordinate with the high-side fuse of the WTG step-up transformer, the minimum trip
current setting of a 67N or 67G will be the same as that of a 51N or 51G, and the
coordination would be the same as that shown in Figure 13. Similarly, the setting
criteria and coordination considerations for directional neutral or residual ground
37
instantaneous overcurrent elements are essentially the same as those for non-
directional 50G or 50N elements.
3.1.2 Voltage and Frequency Protection and Coordination
3.1.2.1 Continuous Voltage and Frequency Range
Wind electric plants may operate within a limited voltage range at the point of
interconnection due to the intermittent nature of these generators, and the variation
in real and reactive power flow across the entire range of plant output. In addition,
Transmission Providers are normally required by regulating authorities to operate
their systems at a normal voltage range of +/-5 percent from the nominal system
voltage. The collector circuits and substation may not require specific
under/overvoltage relaying since the wind WTG relaying and control functions are
designed to keep voltages within acceptable limits. The wind turbines might include
voltage relaying for generator protection as well as voltage control equipment
settings to enable them to recover from some types of system voltage excursions, per
regulatory standards such as North American Electric Reliability Corporation
Standard PRC-024 [23]. Any under/overvoltage protection schemes installed on the
collector circuits or in the collector substation would also have over/undervoltage
settings that would provide protection while not interfering with the generator’s
ability to recover from temporary system voltage excursions.
Frequency response requirements for WTGs in North America are not presently
addressed by regulating authorities. However, it is important to note that some
standards define certain frequency protective relay settings to maximize the
generating unit’s ability to remain connected during a frequency excursion.
3.1.2.2 Temporary Overvoltages

Wind plants can experience higher transient overvoltages (TOV) than typical
distribution systems. Collector circuit ground faults are a primary cause of TOV. For
ground faults, the collector circuit has its primary source of zero sequence current
supplied from the main transformer. Ground faults on the collector circuit will be
cleared by the collector relaying. When the breaker is tripped following the fault, this
opens a window of risk for TOV. The severity depends on the degree of system
grounding on the remaining circuit. Where effective grounding is not provided, the
voltages on the un-faulted phases will be higher, up to 1.73 times nominal phase to
ground voltage. This is an important consideration in the selection of surge arresters.
When collector circuits have underground cables, the cable capacitance further
increases the un-faulted phase voltages. A common way to reduce TOV is to use
grounding transformers.
A main grounding transformer may be needed on the low side of the main
transformer, if the winding connections do not provide a system ground allowing for
38
a zero-sequence source of fault current. A properly rated grounding transformer will

reduce the overvoltage seen by the un-faulted phases during line to ground faults.
Grounding transformers are also typically installed on the collector circuits at the
substation to provide a ground reference when the collector circuit breaker is opened
following a fault. The protection of these units is normally by the collector relaying.
Absent these grounding transformers, the TOV can exceed the 173% value due to the
circuit capacitance. Therefore, the application of grounding transformers can be an
effective means to limit TOV during the time period between the fault and the
shutdown of the wind turbines on the faulted collector circuit.
Other ways to reduce TOV include the application of high-speed grounding switches
on the collector circuits and the use of transfer trip schemes. Note that similar
overvoltage conditions may occur when the collector circuit breaker opens
inadvertently in the absence of a fault. This may occur due to self-excitation,
saturation, or ferro-resonance conditions. In many cases TOV cannot be avoided
completely, and this problem is normally addressed by proper insulation
coordination.
3.1.2.3 Voltage and Frequency Ride-through
The wind turbine protection scheme may include voltage and frequency relaying to
protect the generator. For example, some wind turbines may trip for frequencies at
or below 95 percent of nominal, or above 103 percent of nominal, with an appropriate
time delay. The WTG protection scheme may also trip the unit at voltages below 90%
of nominal or above 110% of nominal, with an appropriate time delay. Specific
voltage and frequency trip settings may be obtained from the WTG vendor. The
vendor’s voltage and frequency protection settings are typically outside of the
minimum limit settings defining the “no trip zones” for voltage and frequency as
specified by area regulatory authority standards to enable the WTG to remain
connected to the power system during a voltage or frequency excursion.
For example, in North America, NERC Standard PRC-024 does not specifically require
under/overvoltage and under/overfrequency relay protection for WTGs, collector
feeders, or the POI. However, if multifunction electronic relays are installed for
protection, either at the collector feeders or the POI, the NERC standard requires
voltage and frequency elements to be set to allow the generation to remain on-line
for voltage and frequency disturbances. . If the specific voltage and frequency
protection settings for the WTGs are known, then the voltage and frequency settings
in the collector feeder relay can be made to allow sufficient time for the WTG voltage
and frequency protection to operate first. In the absence of specific WTG settings, the
voltage and frequency settings in the collector feeder relay may be set per PRC-024
with enough margin to enable them to be outside of the “no trip zone” described by
the standard.
39
3.1.3 Arc Flash Protection

The arc flash levels on the low side of the wind turbine generator (WTG) step-up
transformers can be very high. To reduce the arc flash hazard in this area the
substation feeder relays may be set to trip high speed for faults on the low voltage
side of the WTG GSU transformer, forgoing coordination with the GSU protection. In
addition to the normal directional overcurrent elements described in Section 3.1.1.2,
an additional set of directional overcurrent elements can be enabled specifically for
arc flash protection during maintenance periods (typically called a “maintenance
mode”). The fast and sensitive “arc flash” elements discussed in this section are
normally enabled only in maintenance mode and are disabled for normal WTG
operation so that the normal directional overcurrent elements can maintain proper
overcurrent coordination.
Typically, there are two elements used. The first element is a directional phase
overcurrent element (67P) which is used to detect three phase faults. The second
element is a directional negative sequence overcurrent element (67Q), which will be
used to detect single-line-to-ground, line-to-line, and double-line-to-ground faults.
Generally, the high side of the WTG GSU transformer is delta connected. This
configuration is an open circuit in the zero-sequence network, so it is not possible to
use ground overcurrent elements to detect faults on the low side of the WTG GSU
transformer.
The directional phase overcurrent element can usually be set to pick up for the
minimum expected three phase arcing fault current that the feeder relay will see for
a fault on the low side of any WTG GSU transformer. It is important to account for the
arcing impedance when setting this element, as the ultimate goal is to protect against
an arcing fault, and the arc flash study will be using the arcing current in its
calculations. Similarly, the directional negative sequence overcurrent element can
usually be set to pick up for the minimum expected negative sequence arcing fault
current for a single-line-to-ground, line-to-line, or double-line-to-ground fault
current that the feeder relay will see for a fault on the low side of any WTG GSU
transformer. The directional negative sequence overcurrent element can typically be
set quite sensitive.
The arc flash/maintenance mode elements are often set to trip instantaneously.
However, they may require enabling supervision via SCADA or via the breaker status
and a short delay to avoid unwanted operation on inrush during energization of the
feeder circuit. They may also be set with a delay to allow the feeder to energize in the
maintenance mode. The sensitive phase directional element is also checked to verify
40
that it will not operate on generation current when the WTGs are absorbing reactive
power from the system, as discussed in Section 3.1.1.2.
3.1.4 Removal of WTGs from Collector Feeders Under Fault

The generator protection system for the WTGs is provided by the WTG vendor as part
of the overall generator control system. WTG protection and control systems are
equipped with over/undervoltage protection, over/underfrequency protection,
overcurrent protection, generator winding and bearing temperature, metering,
disturbance recording, power flow control, and communications. A fault on the
collector feeder will cause the main feeder breaker at the substation to trip. When the
WTG protection and control system detect the resulting loss of voltage and load, it
typically trips the WTG breaker to take the unit off-line and changes the angle of the
blades to allow them to stop. The WTGs cannot be brought back on-line until the
collector feeder has been successfully reenergized by the main feeder breaker at the
substation.
3.2 Grounding Transformer Protection

The protection of each component of an electrical power system as well as the overall
system is an essential aspect of plant design. The grounding transformer is an
important element of the system that provides a system ground reference and aids in
the control of Temporary Overvoltage conditions.
A
a
c
C B
b
51N
Figure 18: Delta-Wye Grounding Transformer
There are basically two types of transformer configurations [connections] that might
be selected for the application of a Grounding transformer. The first is a two-winding
transformer whose primary winding connected to the system that requires the
ground, is connected in grounded-wye and whose secondary winding is connected
41
delta [Figure 18]. The second two-winding transformer is the zigzag transformer
[Figure 19] which is usually more cost effective to apply than the typical delta-wye
transformer mentioned previously.
c
A
C
b B a
51N
Figure 19: Zig-Zag Grounding Transformer
The grounding transformer supplies zero sequence fault current to allow detection of
the fault and tripping by the appropriate relays. Though either of the transformer
configurations may be protected by overcurrent relay elements (51) or differential
relay elements (87), the application of overcurrent relay elements is typical. The
protective relaying and current transformers on the grounding transformer itself can
filter out zero sequence current for internal transformer faults by use of delta
connected CTs or by the internal compensation in the relay. The application of
overcurrent relay elements would generally include a 51 element on each phase and
a 51N element connected in the neutral-ground path. Internal faults are typically
detected by overcurrent relays which can respond to the positive and negative
sequence fault currents, allowing the transformer relaying to restrain for ground
faults external to the transformer. A separate neutral CT and ground overcurrent
relay can be used to provide backup ground fault protection, selectively coordinated
with bus and/or collector circuit ground protection. Differential relay elements, by
design, only respond to faults internal to the ground bank.
3.3 Bus Protection

The low-voltage bus in a collector substation of a wind electric plant can be a wide
variety of voltage levels with 34.5 kV being the most common. This is a voltage that is
also commonly used for distribution systems. Although the bus voltage is similar,
there are two significant differences between a collector substation and a distribution
substation. First, there are multiple sources of power connecting to the circuits out of
the collector substations. Second, in many cases the power transformers are
42
significantly larger. It is common to have transformers larger than 100 MVA in

collector substations. The larger transformers push the bus fault duties higher, which
drives the need to clear the bus faults faster than the time delayed overcurrent
relaying would permit. Due to these differences, it is common to apply bus protection
in collector substations that are similar to the protective relay systems applied to
transmission voltage buses.
The IEEE Guide for Protective Relay Applications to Power System Buses, C37.234
[29], is a very complete guide for the protection of substation buses. It is not the
intention of this report to include the full content of that Guide. The following is a brief
description of three different types of bus protection that would be suitable for a
collector substation.
3.3.1 Zone-Interlocked Scheme

The zone-interlocked scheme is by far the least expensive of the bus differential
schemes. It is most often applied on distribution substation buses where there is only
one source of fault current and multiple feeders servicing load. Although this is not
the case for the collector substation in most cases, the power transmission network
fed through the substation step-up transformer is the most significant source of
current to a bus fault. Figure 20 shows the typical collector substation simplified one-
line diagram with a zone-interlocked scheme applied for the bus protection. The
substation step-up transformer is connected to the bus through the M or Main
breaker.
50
Collector Transmission
Circuits M Network
block
50 OR 50
50
Figure 20: Zone Interlocked Bus Differential
Non-directional instantaneous overcurrent elements monitor the current through

the line breakers and send a blocking signal to the bus protection relay for a line fault.
The bus protection relay is another instantaneous overcurrent element that monitors
43
the current from the step-up transformer. The collector line and the bus protection
relay elements may be elements in multifunction relays which are used for other
functions. The overcurrent functions for the line could be part of the line relay, and
the overcurrent function for the step-up transformer could be part of the transformer
relay. The bus protection overcurrent function pickup will need to be set above the
maximum current output of the plant, with a margin, so that the relay will not pick up
under load or for the fault current contributed from the WTGs.
The bus protection overcurrent element is unable on its own to distinguish between
a bus fault and a collector line fault. For a collector line fault, the line overcurrent relay
will operate and send a blocking signal to the bus overcurrent relay. For a bus fault,
no blocking signals will be received so that the bus relay will trip after only a short
coordination delay. The pickup level of the line overcurrent relays is typically more
sensitive than the main breaker overcurrent relay, and the accuracy performance
class of the current transformers (CTs) used by the line relays is usually equal to the
accuracy performance class of the CTs used by the bus relay to enhance the security
of the scheme for line faults. However, it is important to set the line phase overcurrent
relays above the maximum fault current contribution from the WTGs for a bus fault.
Considering there may be little or no negative or zero sequence fault current
contribution from the WTGs to a fault, more sensitive negative and/or zero sequence
non-directional overcurrent functions may also be used to provide the blocking signal
to the bus protection.
The presence or absence of zero sequence current contribution from the WTGs is
easily determined by knowing whether any grounding transformers or neutral
grounding connections are applied on the lines. The presence or absence of negative
sequence current is not obvious and will need to be determined from the WTG
supplier, or whoever is responsible for setting the control systems on the WTG
interface. If there are no sources of negative and zero sequence current on the WTGs
or lines, the limitation on sensitivity of the line unbalance overcurrent functions will
be the capacitive current coming from the line shunt capacitance during unbalanced
faults.
The bus blocking signal circuit may be implemented using relay contact outputs from
the line relays and an input on the bus relay. The line relay contacts are wired in
parallel to operate the input on the bus relay for line faults. This type of scheme can
also be implemented over point to point communication circuits or over a local area
network. The advantage of this type of bus protection is that functions in relays that
will be required for line and transformer protection can be used so that no additional
relays and CTs will be needed. The disadvantages are that the bus protection will not
be as fast as other schemes, sensitivity to high impedance bus faults is limited,
44
complex networking between devices is required, and coordination for both

sensitivity and speed is required between the multiple elements of the scheme. Note
that sensitivity limitations for unbalanced faults can be mitigated significantly by the
possible application of sensitive negative and/or zero sequence overcurrent
detectors in the lines.
3.3.2 Percentage-Restrained Differential

The basic differential principle is to sum each phase of all the current flowing into and
out of the bus and feed the resultant of the combination into an overcurrent relay
element. The restrained or percentage differential protection is an enhanced
variation of the basic differential principle. To insure security against mis-operation
for line faults, even when the CTs are not performing perfectly due to high fault
currents, the current being supplied by each of the circuits connected to the bus is
monitored. The currents in the branches are individually measured and used to
develop a restraint current that is indicative of the current flowing through the zone
of protection. Typical methods are to sum the magnitudes of the branch currents or
take the maximum of the branch currents.. The operate current is the vector sum of
all the currents flowing into and out of the bus. The restraint value is compared with
the operate current to determine if there is a line or bus fault. The higher the fault
current, the greater the restraint value, so more differential current is required to
operate the relay. At high current levels, the error current produced by poor
performing CTs is greater, but this type of differential relay requires more operating
current in this situation so that the operate current caused by the CT performance
error will not result in an operation of the relay for an external fault. For bus faults
with fault impedance, the fault currents are low, restraint is low, so the sensitivity for
these low current magnitude internal faults are very good. Figure 21 shows the basic
current circuit design for this type of relay scheme.
45
Circuits M Network
87B
R R R R R – Restraint Coil/Input
O – Operate Coil/Input
Figure 21: Percentage-Restrained Bus Differential
The impedance in the secondary current circuits for each of the power elements
connected to the bus needs to be kept to a minimum, which is normally not a problem
for the compact collector substation. The maximum magnitude of a bus fault needs to
be taken into consideration when specifying the CT accuracy. Since the internal
impedance of modern relays is low, the CTs need not be dedicated to just the bus
differential relay. A line relay or a transformer relay can share common CTs with the
bus relay. For some relays, the CTs do not have to be on the same ratio in that the
ratios of all the in-feeds are normalized by the relay settings.
The advantage of this type of bus protection is that the relays are sensitive, fast and
secure. The CTs for the different circuits do not need to have matching accuracy
classes, and multi-ratio CTs do not need to be on their full-turns ratio. The CTs used
by the bus differential relays can also be used for other functions if the impedance of
the secondary circuit is kept low. The disadvantages are that the secondary currents
from the set of CTs associated with each of the circuits connected to the bus are
connected directly to the relay, and that may require the impedance of these circuits
to be kept to a minimum to provide acceptable CT performance during a fault.
3.3.3 High Impedance Bus Differential

On the surface, the high impedance bus differential appears to be configured like that
of the most basic of bus differentials. The outputs of all the CTs for the circuits
connected to the bus are summed per phase and fed into an operating circuit of the
relay. That is the only characteristic a high impedance bus differential has in common
with other types of bus differential relay schemes. The high impedance bus
46
differential system avoids the problem of poor performance of CTs for high current
through fault conditions by making the impedance of the operating circuit very high.
This configuration forces the error differential current through the CTs rather than
through the relay operate circuit. For an internal bus fault, the CTs of all contributing
circuits to the bus will be driven into saturation. The voltage produced by the CTs in
saturation is impressed across the operate circuit of the bus differential relay causing
it to operate. The voltage produced by the CTs is based on the accuracy class of the
CTs. The relay is set to operate based on a voltage level, not a current level. To obtain
the greatest sensitivity to high impedance bus faults, the impedance between the CTs
and the common point of all the CTs is normally kept low and matched. The voltage
setting level is based on the voltage drop across the highest impedance circuit
between the CTs and the common point. For this reason, the common point for most
applications is not at the relay but at a cabinet located in the substation yard at equal
distance from all the CT sets. A four-wire cable is then connected between the junction
box and the relay carrying the resultant differential signal.
The CTs used for a bus differential circuit can only be used for this application due the
high impedance and high voltage of the circuit. For the best performance of the
differential circuit, all the CTs applied to this circuit normally have the same accuracy
class and are connected on the same ratio. To avoid potential damage to multi-ratio
CTs, it is important for the CTs to be tapped at full ratio. If the CTs are tapped at a
lower ratio, a higher voltage than what is applied to the operation circuit of the relay
can be produced on the unused windings. This voltage could be higher than the rating
of the insulation of the CT. For best performance of the bus differential system,
auxiliary CTs are not normally used to match the CT ratios. However, there are ways
to apply high accuracy auxiliary CTs in the circuit and still have a secure and
functional system. Figure 22 shows the application of a high impedance bus
differential relay.
Circuits M Network
Junction
Box
87B
Operate Coil
Figure 22: High Impedance Bus Differential

47
Advantages of high impedance bus protection include:

• The relays are sensitive, fast, simple, and secure.
• The CT secondary circuits are not wired all the way to the relay.
• The CT wiring at the relay is very simple, with only four terminations.
The following are disadvantages of high impedance bus differential protection:

• A junction box in the substation yard is needed to terminate the CTs. The only
way to circumvent this requirement and still have the maximum sensitivity for
the scheme is to locate the relay at equal distance to the CTs.
• The CTs for this scheme are dedicated to the scheme.
• All CTs are connected for the same ratio, preferably using the full turns ratio
of the CTs.
• If some CTs in the scheme have a different accuracy class rating, the
differential setting may be dependent on the CTs having the lower accuracy
class rating.
3.4 Main Transformer Protection

The Main Power Transformer is likely the most expensive piece of equipment in the
collector substation. The high cost of repair or replacement and the possibility of
violent failure or fire involving adjacent equipment makes limiting the damage a
major objective. It is important to carefully consider the protection characteristics of
relays that will be applied for transformer protection, particularly for transformers
of larger sizes. It is common to have transformers greater than or equal to 100 MVA
in collector substations.
There are multiple ways to protect main power transformers that offer varying
degrees of sensitivity, speed, and selectivity. The selected protection scheme may
balance the best combination of these factors plus overall economics, while
minimizing:
• Cost of repairing damage
• Cost of lost production
• Adverse effects on the balance of the system
• Spread of damage to the adjacent equipment
• Duration of the unavailability of the damaged equipment
Figure 2 shows common types of collector substation transformer configurations.
48
The IEEE Guide for Protecting Power Transformers, C37.91-2008 [30] provides a
complete guide for the protection of substation power transformers. It is not the
intention of this report to include the full content of that Guide. The following is a brief
description of the different types of transformer protection that would be suitable for
a collector substation.
3.4.1 Transformer Differential Protection

Current differential schemes are commonly used for protecting power transformers.
Current differential protection has the following advantages:
• Provides faster protection and can limit equipment damage
• The location of the fault can be determined more precisely
• Accurate fault location allows for auto restoration techniques
• High speed clearing can significantly lower the arc flash incident energy
Figure 23 below illustrates the basic concept of the current differential scheme.
Figure 23: Basic concept of the current differential protection scheme [30]
If the ratios of CT1 and CT2 are 1:1, the operating current, I 0 , which is the difference
between the current entering one winding and the current leaving the other winding,
can be obtained by the following equation:
I 0 = I1 − I 2
where
I 0 is the operating current
I 1 is the current entering the transformer
49
I 2 is the current leaving the transformer
Figure 24 shows the transformer differential protection scheme connections for a

wye grounded-wye grounded transformer with a delta-connected tertiary winding
that is also brought out to bushings.
Figure 24: Typical schematic connections for differential protection of a three-

winding transformer [30]
Percentage differential relays that are commonly applied for transformer protection
may classified as:
• Percent differential relays with restraint actuated by currents going into and
out of the protection zone.
50
• Percent differential relays with restraint actuated by one or more harmonics

along with currents going into and out of the protection zone.
Transformer protection schemes at wind electric plants typically utilize percent
differential with harmonic restraint and/or blocking. The basis of the percentage
differential relay is that the difference current (as measured at the ends of the
protected zones) is more than a predetermined percentage of the restraint current.
The percentage difference can be fixed or variable (adaptive), based on the relay’s
design. There is also a minimum differential current threshold before tripping
without regard to the restraint current. Details of minimum pickup, restraint current,
and characteristic slope vary among manufacturers. Slope may not be a straight line
but may curve up depending on the design of the percentage restraint system. This
curve allows even larger percentage mismatches during heavy through-currents.
These options are shown in Figure 25.
Figure 25: Typical options for the characteristics of percentage differential

relays [30]
Modern microprocessor relays employ harmonic restraint and blocking functions to

make the scheme more secure during transformer energization. Second and 4th
Harmonics restraint or blocking functions can be set as a percentage of the
fundamental. These schemes can block the percent differential either independently
or utilize cross-blocking. A 5th Harmonic element can be set to alarm for
overexcitation conditions. Harmonic blocking and alarm settings may require a more
in-depth review at wind farm plants due to additional sources of Harmonics. Wind
farm plants could have inverter-based generation (Type 4 Turbines), Capacitor Banks
51
for VAR support, Capacitor Banks with Limiting Current Reactors (Harmonic Filters),
and STATCOMs for dynamic VAR support. Harmonic currents can cause overheating
and extra losses in many components including transformers. Overheating shortens
the useful lifetime of transformers, and in an extreme case, can lead to their
destruction. It is important to consider all of these issues when determining the best
scheme for harmonic restraint, harmonic blocking, and overexcitation protection.
To successfully detect faults in grounded wye-connected transformer windings, the

relay system is designed to discriminate between faults internal and external to the
protected zone. The ground differential relay, device 87G in Figure 26, which is
typically an overcurrent relay, or the directional ground relay, device 67G connected
as in Figure 27, is satisfactory. Both relay schemes will operate correctly for any
internal ground faults with the circuit breaker in the circuit to the grounded wye
winding being open or closed. They will operate correctly with an external zero-
sequence current source, and they will not operate for external ground faults. The
auxiliary CT is necessary if the phase and neutral CTs are of different ratio, and the
relay is of the electromagnetic or solid-state type. If a numerical relay is used, the
auxiliary CT is not needed because the ratio mismatch is taken care of during the
computations in the relay. Both schemes are particularly applicable where the
ground-fault current is limited, and phase differential relays may not respond. The
operating current in device 67G is zero for an external fault with CT ratios matched.
Therefore, it is advisable to select the auxiliary CT ratio to give definite non-trip bias
to device 67G for an external ground fault (auxiliary CT secondary current slightly
greater than the transformer neutral CT secondary current).
Unequal CT currents can produce residual error current during external phase faults.
No transformer neutral current is produced, and sensitive relays could operate
unnecessarily. Some modern relays allow relay operation only if the current in the
transformer neutral exceeds a threshold.
52
Figure 26: Complete ground-fault protection of a delta-wye transformer using

a residual overcurrent and differentially connected ground relay [30]
Figure 27: Complete ground-fault protection of a delta-wye transformer using

a residual overcurrent and directional relay [30]
53
3.4.2 Overcurrent Protection and Coordination with Collector Feeders

Instantaneous overcurrent elements can be used for fast clearing of high magnitude
internal faults. These elements are normally set above the maximum through fault
current for a fault outside the transformer zone of protection. The pickup range is
generally 125% - 200% of a three-phase fault on the low side of the transformer. The
elements are also set to be secure against transformer inrush. Some wind farm plants
may have a weak source. In this case, the instantaneous settings are typically checked
to see that they are not set above the maximum available fault current at the station.
In these cases, harmonic restrained instantaneous elements may be considered for
providing the desired protection.
When setting ground instantaneous overcurrent elements on autotransformers and
on three-winding transformers connected wye grounded-wye grounded with a delta
tertiary winding, the ground fault current seen by the relay for asymmetrical faults
on either side of the transformer will need to be considered. The ground
instantaneous element will see need to be set above the higher of the two ground fault
current contributions.
Time overcurrent elements are used to protect transformers due to a failure of the
protection system to clear an external fault. High through fault current can cause
internal, thermal, or mechanical damage to the main transformer. Annex A in IEEE
C37.91-2008 Guide for Protecting Power Transformers provides a detailed
description to develop the transformer through-fault-duration guide to protect
power transformers. Figure 28 shows the category IV transformer damage curves of
Figure A.4 from the C37.91-2008 guide.
54
Figure 28: Thermal Damage Curve for Category IV Transformers from IEEE
Standard C37.91-2008
The time overcurrent curve, tap, and time dial are selected to coordinate with the
transformer damage curve. The tap setting is also selected to carry the transformer
load plus a margin for overload. A coordination check will then need to be performed
with the collector circuit overcurrent protection. Sufficient margin is typically
included to provide adequate coordination between the collector circuit and the main
power transformer. Traditional distribution substation feeder circuits supply load,
55
and a check may need to be performed on high impedance faults to account for the
additional load current flowing through the transformer in addition to the fault on
the feeder circuit. Because collector feeder circuits supply generation back to the
station, the situation of the transformer seeing excessive load current in addition to
the fault current seen on the collector circuit during a high impedance fault is
minimized.
It is important to verify that the current pickup settings of the phase time overcurrent
elements also meet any asynchronous generator requirements defined by regulating
authorities. For example, PRC-025-1, Table 1 “Relay Loadability Evaluation Criteria
[20] requires that overcurrent elements shall be set greater than 130% of the
calculated current derived from the maximum aggregate nameplate MVA output at
rated power factor (including the MVAR output of any static or dynamic reactive
power devices).
The tertiary winding of an autotransformer, or three-winding transformer, is usually

of much smaller kVA rating than the main windings. Therefore, overcurrent relays set
to protect the main windings offer almost no protection to these tertiary windings.
During external ground faults on the high-voltage or low-voltage side of the
transformer, the tertiary windings may carry very heavy currents. Hence, in the event
of failure of the primary protection for external ground faults, separate tertiary
overcurrent protection may be desirable. The method selected for protecting the
tertiary winding generally depends on whether the tertiary is used to carry load. If
the tertiary does not carry load, protection can be provided by a single overcurrent
relay connected to a CT in series with one winding of the delta-connected winding.
This relay will sense system grounds as well as phase faults in the tertiary or in its
leads.
If the tertiary winding is used to carry load, partial protection can be provided by a
single overcurrent relay supplied by three CTs, one in each winding of the delta and
connected in parallel to the relay. This connection provides only zero-sequence
overload protection and does not protect for positive- and negative-sequence
overload currents. In this case, the relay will operate for system ground faults but will
not operate for phase faults in the tertiary or its leads. Where deemed necessary,
separate relays, such as differential type, may be provided for protecting against
phase-to-phase faults in the tertiary windings or its leads.
The setting of the overcurrent relays, which are provided for protecting the tertiary
windings, can normally be based on considerations similar to the other overcurrent
elements. However, three CTs (one in each phase) can be connected in parallel to
provide zero-sequence currents to an overcurrent relay; this relay can be set below
56
the rating of the tertiary winding. Be aware that this relay is still required to be set to
coordinate with other relays on the system.
When tertiary windings are connected by cables, the overcurrent protection provided
for the tertiary winding is typically set to account for the thermal withstand capability
of the cables. Alarming and tripping resulting from a prolonged unbalance condition
or LTC malfunction will therefore prevent damage to cables.
Below are example calculations and coordination curves for one of the two main
power transformers of a theoretical wind farm plant. The theoretical wind farm plant
has the following specifications:
• Total Size = 300 MVA

• Number of Transformers = 2. Each transformer:
o 3 Winding
▪ Wye, Wye, Delta
o 138:34.5kV
o 134/178/222 MVA
o Z = 11.2% @ 134 MVA
• Voltage = 138:34.5kV
• Total Number of Wind Turbine Generators = 88 (44 WTGs supply each main
low-side bus)
• Size of Type 3 Wind Turbine Generators = 3.4 MVA
• Studies indicated that no additional VAR support was required for the
installation.
Example Phase Instantaneous Overcurrent – 50P. 50P is generally set for the higher
of the following conditions: 1.3 – 1.5 times a low side three phase fault or 8 – 12 times
the transformer nameplate rating to account of inrush current. In our example, a
three-phase fault on the 34.5 kV bus produces 3125 A on the 138 kV side of the
transformer.
134MVA  1000
50 PHS = 1.5  3125 A = 4687.5 APRI or 50PHS = 12  = 6727 APRI
3  138kV
150MVA  1000
51PHS = 1.5  I FLA = 1.5  = 1.5  627 A = 941APRI
3  138kV
150MVA  1000
51PLS = 1.5  I FLA = 1.5  = 1.5  2510 A = 3765 APRI
3  34.5kV
HS CTR = 1200:5 or 240:1

57
LS CTR = 4000:5 or 800:1
6727 APRI 6727 APRI

50PLS = = = 28.2 ASEC  28 ASEC
CTR 240
941APRI 941APRI
51PHS = = = 3.92 ASEC  4 ASEC
CTR 240
Full load current of 44 WTGs is 626 A. Per regulatory standards (PRC-025 [20]), the
phase time overcurrent pickup can be set greater than or equal to 130% of the
maximum aggregate WTG nameplate output. 1.3 x 626 A = 814 A. Therefore, a 51PHS
current pickup setting of 4 = 960 A, primary will satisfy PRC-025. The same will be
true for the 51PLS current pickup setting.
3765 APRI 3765 APRI

51PLS = = = 4.7 ASEC  4.7 ASEC
CTR 800
The resulting coordination between the collector feeder and collector station main
power transformer overcurrent relays appears in Figure 29.
58
Figure 29: Main Transformer Phase Overcurrent Coordination
59
3.4.3 Mechanical Detection of Faults

Some transformer faults go undetected when differential and overcurrent schemes
are used. A turn-to-turn fault can cause considerable current to flow in the shorted
turn, while current in the remaining winding remains relatively unchanged. Since
there is little or no change in the current monitored by the CTs, there is no differential
current for the relays to operate. If the turn-to-turn fault either evolves to include
more turns or evolves into a ground fault, then the differential elements may be able
to operate for the change in current.
There are two methods of detecting transformer faults other than by electric
measurements. These methods are as follows:
a) Accumulation of gases due to slow decomposition of the transformer

insulation or oil. These relays can detect heating due to high-resistance joints
or due to high eddy currents between laminations.
b) Rapid increases in tank oil or gas pressures caused by internal transformer
faults.
The mechanical pressure relays come in a variety of types. Depending on the

configuration of the transformer, the application will vary. A sudden-oil-pressure
relay can be applied to any oil-immersed transformer. It would be mounted on the
transformer tank wall below the minimum liquid level. If an internal fault develops,
the rapid rise in oil pressure or pressure pulse is transmitted to the silicone oil by way
of the transformer oil and the bellows. This then acts against the piston, which closes
the air gap and operates the switch. A similar type of mechanical pressure relay can
be applied to monitor the gas pressure above the oil. There is also a type of relay,
commonly known as the Buchholz relay, which is applicable only to transformers
equipped with conservator tanks and with no gas space inside the transformer tank.
It will operate for small faults by accumulating the gas over a period of time or for
large faults that force the oil through the relay at a high velocity.
These devices can detect a small volume of gas produced by low-energy arcs,
overheating, and insulation decomposition. For example, when high current passes
through a shorted turn, a great deal of heat is generated. This heat, along with the
accompanying arcing, breaks down the oil into combustible gases. Gas generation
increases pressure within the tank. A sudden increase in gas pressure can be detected
by a sudden-pressure relay either located in the gas space or under the oil. The
sudden-pressure relay usually operates before other relays sensing electrical
quantities thus limiting damage to the transformer. The accumulator portion of the
relay is frequently used for alarming only; it may detect gas that is not the result of a
fault, but that can be evolved by gassing of the oil during sudden reduction of
60
pressure. This relay may also detect heating due to increased power transfer,
increased ambient temperature, full or partial failure of the cooling system, high-
resistance joints, high eddy current between laminations, low- and high-energy
arcing, or accelerated aging due to overloading.
One drawback to using a sudden-pressure relay is its tendency to operate on high-

current through-faults. Various methods to prevent undesired operation have been
developed. The most common method uses an instantaneous overcurrent relay to
supervise the sudden-pressure relay, whereby taking advantage of the fact that a
close-in through-fault creates a high current in the transformer. Another method,
used less often, is to place sudden-pressure relays on opposite corners of the
transformer tank. Any pressure wave due to through-faults will not be detected by
both sudden-pressure relays. The contacts of the sudden-pressure relays are
connected in series so that the operation of both relays will provide a genuine sudden
pressure trip output.
3.5 Capacitor and Harmonic Filter Protection

Protection schemes applied to capacitors and harmonic filters at the main collector
substation are essentially the same as those applied at conventional substations. The
primary purpose of the protection scheme is to take the protected equipment out of
service for faults. IEEE Standard C37.99: IEEE Guide for the Protection of Shunt
Capacitors [31] provides detailed information on the application and setting of
protective relays for shunt capacitors. Similarly, IEEE Standard C37.109: IEEE Guide
for the Protection of Shunt Reactors [32] provides detailed information on the
application and setting of protective relays for shunt reactors. While C37.109 does
not directly address the protection of the reactors used in harmonic filters, the
protection schemes presented are also applicable to the reactors used in harmonic
filters as well.
3.5.1 Voltage Protection
Two levels of undervoltage and overvoltage settings are typically applied for
capacitor protection. For example, a Level 1 undervoltage setting of approximately
90% of nominal voltage and a Level 2 undervoltage setting less than or equal to 45%
of nominal voltage may be applied. Similarly, a Level 1 overvoltage setting of
approximately 110% of nominal voltage and a Level 2 undervoltage setting of
approximately 120% of nominal may be applied. Appropriate undervoltage delay
settings are selected based on the desired “ride-through” time for a system
undervoltage condition. Appropriate overvoltage delay settings are chosen either on
the desired ride-through time for a system overvoltage condition or on the maximum
overvoltage withstand capability of the capacitor units.
61
For shunt capacitors at collector feeder substations, the voltage protection settings
are also coordinated with the voltage protection settings that are applied to the
WTGs. Ideally the capacitor voltage protection settings are delayed long enough to
allow the WTG voltage protection equipment to take the WTG off-line before the
capacitor breaker or circuit switcher at the collector feeder substation trips.
3.5.2 Overcurrent Protection
Overcurrent protection schemes are typically applied to trip shunt capacitors or
harmonic filters for both phase and ground faults in the equipment. Instantaneous
overcurrent elements are set based on the maximum expected inrush current, and
time overcurrent elements are set based on the maximum current rating of the
capacitor or reactor. The instantaneous element may require a short delay to provide
proper coordination with the expulsion fuses on the capacitor cans. Inverse time
curves are normally used for time overcurrent applications, and the time dial is
selected to coordinate with both the capacitor can fuses and any “upstream” time
overcurrent relays such as the backup time overcurrent relays associated with the
main power transformer at the collector substation. When overcurrent relays are
used for capacitor or harmonic filter protection, the presence of harmonics may
require the use of relays equipped with RMS based current detectors capable of
measuring both fundamental and harmonic currents.
3.5.3 Harmonic Current and Voltage Considerations for Protection Scheme
Operation
It is important to assess the effect of connecting WTGs to the electric utility’s power
system prior to their installation to verify that the system harmonic and voltage limits
required by the utility can be achieved. The electrical characteristics of the wind
turbines are required to assess the effect of the WTGs on the system. Given the
manufacturer’s typical electrical characteristics for the WTGs installed at the wind
electric plant site, the combined impact of the WTGs on voltage quality can be
calculated.
Inverters connected to WTGs can cause harmonics [4]. The order and magnitude of
the harmonic currents depend on the technology of the converter and the mode of
operation [6]. The voltage waveform can be distorted by the injection of harmonic
currents, which can propagate throughout the distribution grid. Furthermore, small
voltage distortions can cause large harmonic currents from series resonance
conditions between the cable capacitance and the supply inductance (transformer
leakage and cable inductance). Harmonic currents arising from the interaction
between the WTGs and the grid can become worse over time due to several factors:
• The wind farms become bigger, and therefore the source impedance may increase.
62
• The equivalent capacitance increases due to use of long extra high voltage cables
where the operators cannot obtain the necessary right of way for an overhead line.
• The use of large wind electric plant transformers on the grid.
• The use of switched capacitors and reactors on the grid.
• Increased harmonic voltage distortion over the years.
• Very low resistive losses in the site design to optimize power production.
All these factors together represent a general reduction of the resonant frequency
into areas where network and WTG harmonic currents exist to excite it (from 5 th to
15th harmonic order). In addition, a small harmonic current can create high levels of
voltage distortion, and a low series resistance can produce a very high magnification
factor. Finally, the source impedance will vary because the wind electric plant is a
combination of many individual WTGs, which shifts these resonant frequencies over
a range as the WTGs connect and disconnect. Consequently, the identification and
elimination of possible resonance conditions during the initial design of the wind
electric plant has become very important.
The spectrum of harmonic currents is a problem because it can increase the voltage
harmonic levels over many harmonic and inter-harmonic frequencies and phase
angles. Sometimes the difference between the calculated harmonic values used
during planning and design and the actual values observed in the field do not become
apparent until unforeseen harmonic issues crop up after the wind electric plant goes
into service. Harmonic levels may be normal most of the time but on some occasions,
they may exceed agreed standards. In these cases, some of or all the WTGs may
automatically shut down from the operation of their internal harmonic protection
schemes. It is desirable to avoid such operation if possible because it requires manual
intervention to restart the WTGs, leading to high levels of lost production.
Voltage flicker may also be seen on WTG collector feeder system. For example, the
tower effect of a fixed speed WTG can cause voltage flickers due to the wind shielding
effect of each blade of a three-blade turbine as it passes the tower. The injected
electrical power of the WTGs is reduced when the blade passes the tower, which
affects the grid voltage [6]. As the blade passes by the tower it causes a power
oscillation and voltage flicker at a frequency of three times the blade turning speed.
Alternatively, the presence of several WTGs distributed on the collector feeder may
reduce the network impedance of the system. As a result, the changing load current
will lead to a smaller voltage variation due to the changing load and hence improved
power quality [8].
3.6 Transmission Tie Line Protection

Transmission tie line protection faces two challenges that are different from
conventional transmission interconnections to generating stations.
63
a) Variability of source strength due to variability of wind. This means that

variable zero, positive and negative sequence source strength may be
considered a normal operating condition. Additional contingencies may need
to be taken into consideration when applying or setting transmission line
protection systems.
b) Short circuit current contribution will likely be controlled as to magnitude and
power factor.
With respect to the variability of source strength, the various sequence networks may
be considered separately. Due to lack of wind, all machines may be at standstill under
normal conditions, and all machines may be off-line. Reactive power compensation
devices (if applied) may be on or off-line. Thus, positive and negative source
impedances may be so high as to cause the source strength to be considered
negligible. Zero sequence source impedance however will be provided by ground
paths through the interconnecting power transformer(s). If there is only one
interconnecting transformer, a single contingency may result in it being out of service.
In effect then the wind plant will be off-line. This contingency may be considered as
the wind plant terminal being open. If there is more than one interconnecting
transformer, a single contingency may result in one of the transformers being out of
service. The line protection then would be designed to be suitable for infinitely high
positive and negative source impedance and higher than normal zero sequence
source impedance.
With respect to the controlled nature of the short circuit contribution additional
considerations become necessary. Most controlled sources will limit the magnitude
of short circuit current contribution to not more than 150% of rated current. But for
most WTGs to meet the low voltage ride through required by the grid codes, the limit
on the magnitude of short circuit current contribution does not take place for a few
cycles into the fault. Depending on the speed of operation for the line relays, the short
circuit current contribution from the wind plant for a line fault maybe significantly
higher than the limited value [24]. This means that instantaneous phase overcurrent
protection devices may be useful, but that phase time overcurrent protection devices
will be of little use. The controlled nature of the short circuit contributions also means
that the negative sequence current contribution to unbalanced faults may be
intentionally limited to negligible levels. This is true for the current design for the
Type 4 WTG [1]. This means that negative sequence current supervision of directional
functions may also be of little value [24]. Further, the reactive reach of quadrilateral
distance elements that depend on conventional inductive phase or negative sequence
current sources may also perform unreliably and may over or underreach [24],[25].
3.6.1 Typical Communications Assisted Protection Schemes
64
Typical communications assisted protection schemes may be divided into three

classes: directional comparison, phase comparison and line current differential. They
will be discussed separately.
Directional comparison schemes often used distance relays and may depend on
underreaching or overreaching distance functions as well as directional overcurrent
functions. As noted above, low fault current contributions may reduce the
dependability of distance functions due to current supervision functions failing to
assert. Also, as noted in [25], the reach accuracy of quadrilateral functions is highly
dependent on the control settings in the converter equipment. If control settings are
set to minimize or reduce negative sequence current contributions to unbalanced
faults, then negative sequence directional elements may also be unreliable. Since
interconnection transformers usually have a path for ground current flow, and since
this path will be highly inductive, zero sequence directional functions will often be
more reliable than negative sequence directional functions. To cover for cases with
little or no wind, weak source echo functions would often be helpful if permissive
overreaching transfer tripping (POTT) logic is used. In all cases of directional
comparison logic, direct transfer tripping from the (normally strong) utility terminal
of the interconnection would be an important supplement.
Phase comparison schemes need a minimum amount of current from both terminals
to have sufficient current to provide reliable phase angles. In view of the possible
minimal current from a wind plant, especially in faults not involving ground, phase
comparison schemes are usually not applicable.
Line current differential protection schemes do not need strong sources from both
terminals, and typically will provide the most reliable interconnection transmission
line protection. Of course, these schemes require high bandwidth communications
facilities.
3.6.2 Back-up Protection Schemes
As noted in the previous section, direct transfer trip is an important supplement to
directional comparison schemes. This direct transfer trip may indeed be the primary
protection for the wind plant terminal and could be applicable whether or not
directional comparison line protection is provided.
As backup protection, time delayed undervoltage protection on all three phases will
provide dependable tripping of the wind plant terminal even under weak or no source
conditions. Undervoltage protection is not very selective, so time delays on the order
of several seconds may be required to coordinate with slowly cleared faults on other
parts of the system. However, high speed tripping of a weak source like a wind plant
is not normally required unless there could be a power quality problem for loads
islanded with the wind plant.
65
3.6.3 Voltage and Frequency Protection Requirements

Voltage and frequency protection may be required under certain circumstances. At
the point of interconnection to the transmission network, relay elements to monitor
and force the disconnection of the wind plant for out-of-normal range of operation
may be required by the transmission provider. Typically, multiple steps of level
detection for both voltage magnitude and frequency for under and over conditions
are applied. The more out-of-tolerance the conditions are, the shorter the time delay
to the tripping [26]. The elements that pick up closer to the normal range are more
for power quality protection and operate with enough delay to permit the controls at
the wind plant to return system conditions back to normal before the trip. Experience
has shown that the most likely reason for the operation of these relay elements are
for failure of voltage control elements at the wind plant; however, the isolation of
wind plants with blocks of load has also occurred due to transmission system outages.
For these events the wind plants are not able to maintain acceptable voltage and
frequency, which can result in the damage of load equipment if the wind plants are
not forced to disconnect after the normal fault ride-through requirement period. The
elements that pick up for the more extreme conditions are set to protect the
transmission provider’s and other customers’ equipment from damage. Direct
transfer tripping of the wind plant for isolation of the transmission interconnection
from the plant to the to the integrated system may also be considered for fast
protection of other tapped loads from unacceptable voltage or frequency excursions.
An important consideration with respect to overvoltage is whether the wind plant is
an effectively grounded source. In some cases, wind plants may be connected without
an effective neutral ground connection [26]. In those cases, temporary overvoltage on
un-faulted phases during ground faults may be a concern. In such cases delayed
tripping of the utility terminal (until after the wind plant has tripped) may be
required for ground faults and sacrificial surge arresters may be used as backup
protection.
As noted in Section 3.6.2, time delayed undervoltage protection may provide effective
backup protection for the wind plant terminal of the interconnected line.
3.6.4 Supervision Requirements for Transmission Line Breaker Closing
The wind plant, like any other generation plant cannot be connected asynchronously
to another system. Therefore, three phase undervoltage supervision would be
employed to verify that all three phases of the line are dead before automatically
reclosing a transmission line breaker from the utility side. However automatic
reclosing of an overhead transmission line from the utility side (with the wind plant
side breaker open) could be a useful test of the line. Once the line was tested
satisfactorily from the utility, the wind plant could be reconnected in the normal
manner.
66
4 Conclusion
This report provides an overview of the protective and system requirements for wind
power plants, which can be unique and challenging due to the following:
• Wind power plants are different from conventional generation in that there
are typically numerous relatively small wind turbine generators (WTG)
distributed geographically over a wide area.
• The WTGs predominantly in service have some degree of inverter interface.
This affects the fault current levels and characteristics which protection
engineers have come to expect from sequence analysis.
• Some WTGs operate with a slip. When near transmission systems employing
series compensation, this can result in sub-synchronous interactions.
• There are various types of electrical layouts and grounding options depending
on the grid the wind power plant is connected to.
• It is important to consider applicable regulatory requirements, such as low-
voltage ride through (LVRT), to enable the wind power plant to assist the grid
during contingencies. Maintaining adequate power quality is another such
consideration.
Various protection elements have been used in practice to address many of the afore-
mentioned issues successfully. These elements are discussed in this report with
setting guidance to aid the protection engineer meet the general protective
requirements of wind power plants. Additional system studies may be required to
address complex conditions such as sub-synchronous interactions. As the
penetration of wind power plants in our system increases, the challenges and
protective considerations continue to evolve. The protection engineer will benefit
from further reading to understand the present best practices.
67
5. Bibliography
[1] “Fault Current Contributions from Wind Plants”, IEEE PSRC WG C17 report,
online: http://www.pes-
psrc.org/Reports/Fault%20Current%20Contributions%20from%20Wind%20Pl
ants.pdf”
[2] IEC Wind turbines - Part 21: Measurement and assessment of power quality
characteristics of grid connected wind turbines, IEC 61400-21:2008.
[3] IEC Electromagnetic Compatibility (EMC) Part 3: Limits – Section 6: Assessment
of emission limits for distorting loads in MV and HV power systems, IEC 61000-3-
6:1996.
[4] Coster, E.J.; Myrzik, J.M.A.; Kruimer, B.; Kling, W.L., "Integration Issues of
Distributed Generation in Distribution Grids," in Proceedings of the IEEE, vol.99,
no.1, pp.28-39, Jan. 2011. DOI: 10.1109/JPROC.2010.2052776
[5] Malati Chaudhary, Sukumar Brahma, and Satish Ranade, "Circuit Breaker
Selection in a Wind Farm with Type-2 Wind Turbine Generators," Proc. IEEE PES
General Meeting 2013, Vancouver BC, Canada, July 2013.
[6] T. Ackermann and V. Knyazkin, “Interaction of distributed generation and the
distribution network: Operation aspects”, Proc. PES T&D Conf., Asia Pacific, 2002,
vol. 2, pp. 1357–1362.
[7] B. Fox, D. Flynn, L. Bryans, N. Jenkins, D. Milborrow, M. O’Malley, R. Watson, and
O. Anaya-Lara, Wind Power integration, Connection and System Operational
Aspects, vol. 50, 1st ed. London, U.K.: IET, 2007, ser. Power & Energy.
[8] N. Jenkins, R. Allen, P. Crossley, D. Kirchen, and G. Strbac, Embedded Generation,
vol. 31, 1st ed. London, U.K.: IET, 2000, ser. Power & Energy.
[9] IEEE PES Wind Plant Collector System Design Working Group, “Wind power plant
substation and collector system redundancy, reliability, and economics”, IEEE
Power & Energy Society General Meeting, pp. 1-6, Jul. 2009.
[10] IEEE PES Wind Plant Collector System Design Working Group, “Wind power
plant collection system design considerations”, IEEE Power & Energy Society
General Meeting, pp. 1-7, Jul. 2009.
[11] R. A. Walling and T. Ruddy, “Economic optimization of offshore wind farm
substations and collection systems,” in Proc. 5 th Intl. Workshop Large-Scale
Integration of Wind Power and Transmission Networks for Offshore Wind
Farms, pp. 1-7, Apr. 2005.
[12] IEEE Design Guide for Electrical Power Systems for Generating Stations, IEEE
Std. 666TM-2007.
68
[13] “Sub-Synchronous Control Interactions between Type 3 Wind Turbines and

Series Compensated AC Transmission Systems”, Technical paper presented at
IEEE EPEC Conference 2011 Andrew L. Isaacs, Garth D. Irwin, and Amit K. Jindal
[14] “Reader's Guide to Subsynchronous Resonance”, IEEE Committee Report by
Subsynchronous Resonance Working Group of the System Dynamic Performance
Subcommittee, Transactions on Power Systems. Vol. 7, No. 1, February 1992
[15] “Sub-harmonic protection application for interconnections of series
compensated lines and wind farms”, Technical paper presented at Western
Protection Relay Conference, October 2012, René Midence, Joe Perez, P.E., Adi
Mulawarman, P.E.
[16] IEEE PES Wind Plant Collector System Design Working Group, “Characteristics of
wind turbine generators for wind power plants,” PES General Meeting, Calgary,
AB, Canada, July 26-30, 2009.
[17] J. C. Das, Power System Harmonics and Passive Filter Designs, Wiley-IEEE Press,
March 2015.
[18] Joe Perez, “Understanding sub-harmonics,” www.erlphase.com.
[19] FERC Order no. 661-A, "Interconnection for Wind Energy,” Docket No. RM05-4-
001, December 2005.
[20] NERC Reliability Standard PRC-025-1: Generator Relay Loadability, July 17,
2014.
[21] B. Chen, A. Shrestha, F. Ituzaro, and N. Fischer, “Addressing Protection Challenges
Associated with Type 3 and Type 4 Wind Turbine Generators,” Published in
Proceedings of the 68th Annual Conference for Protective Relay Engineers April
2015 (Texas A&M, College Station, TX).
[22] M. Zadeh, “Directional Overcurrent Protection of Wind Power Plant Collector
Challenges, Analysis and Solutions”, CARILEC 2018 Engineering Conference &
Exhibition, July 2018.
[23] NERC Reliability Standard PRC-024-2: Generator Frequency and Voltage
Protective Relay Settings, May 29, 2015.
[24] M. Nagpal, C. Henville, “Impact of Power Electronic Sources on Transmission Line
Ground Fault Protection”, IEEE Transactions on Power Delivery, February 2018,
Vol 33, p.p. 62-70
[25] Ali Hooshyar, Maher A. Azzouz, Ehab F. El-Saadany, “Distance Protection of Lines
Emanating from Full-Scale Converter-Interfaced Renewable Energy Power
Plants—Part I: Problem Statement”, IEEE Transactions on Power Delivery, Vol.
30, No. 4, August 2015.
69
[26] Mukesh Nagpal, “Nontraditional Protection Solutions Permit Tap Transmission

Connections of Nonutility Generators”, IEEE Transactions on Power Delivery,
Vol. 31, No. 5, October 2016.
[27] A. F. Elneweihi, E. O. Schweitzer III, M. W. Feltis, “Negative-Sequence Overcurrent
Element Application and Coordination in Distribution Protection”, IEEE
Transactions on Power Delivery, Volume 8, Issue 3, July 1993.
[28] IEEE Guide for Protection Systems of Transmission to Generation
Interconnections, C37.246-2017.
[29] IEEE Guide for Protective relay Applications to Power System Buses, C37.234-
2009.
[30] IEEE Guide for Protecting Power Transformers, C37.91-2008.
[31] IEEE Guide for the Protection of Shunt Capacitor Banks, C37.99-2012.
[32] IEEE Guide for the Protection of Shunt Reactors, C37.109-2006.
[33] “Modification of Commercial Fault Calculation Programs for Wind Turbine
Generators:, IEEE PSRC WG C24 Report, online: https://www.pes-
psrc.org/kb/published/reports/C24_WG_Report_Jun_2020_Final.pdf
70
APPENDIX A : Directional Phase Overcurrent Setting

Considerations for WTG Operation
Figure A1 shows the current characteristic of WTG Types 3 and 4. As shown in this
figure, when the WTG current lags its terminal voltage (Figure A1 left, blue region) or
the feeder current as measured by the 67P leads its measured voltage (Figure A1
right, blue region), the WTG supplies reactive power to the system. This occurs when
the collector system experiences low voltage, and most probably the collector system
is heavily loaded. Hence, the WTG power control system will operate to make the
WTG behave like a capacitor to raise the voltage. Figure A2 shows the directional
element (67P) characteristic on an R-X plot with the forward trip direction of the 67P
set towards the WTGs on the collector circuit. For the afore-mentioned condition, the
impedance seen by the 67P element falls within the blue region i.e. third and fourth
quadrants in Figure A2. If the power generated by the aggregate WTGs exceeds the
aggregate station service and power loss (PAux) in the collector system, then the
impedance seen by the 67P falls within the third quadrant, which is completely
outside the tripping region. Conversely, if the power generated by the aggregate
WTGs is less than PAux, the impedance seen by the 67P falls within the fourth
quadrant, which encompasses part of the tripping region. For the seen impedance to
fall within the tripping region in the fourth quadrant, the WTG will supply a very small
reactive power as well as an active power less than PAux. As it is confirmed through
simulation later in this section, this results in a current supplied by the aggregate
WTGs that is less than the 67P pickup setting; consequently, it is of no concern.
As shown in Figure 15 A1, when the WTG supplying current leads its terminal voltage
(Figure A1 left, red region) or the current measured by the 67P lags its measured
voltage (Figure A1 right, red region), the WTG absorbs reactive power. This occurs
when the collector system experiences high voltage, and as mentioned earlier, most
probably the collector system is lightly loaded. In that case, the WTG power control
system will operate to make the WTG behave like an inductor to lower the voltage.
For the afore-mentioned condition, as shown in Figure 16A2, the impedance seen by
the 67P element falls within the red region i.e. first and second quadrants. If the
power generated by the WTGs exceeds PAux, then the impedance seen by the 67P falls
within the second quadrant, which partially covers the tripping region. On the other
hand, if the power generated by the aggregate WTGs is less than P Aux,, the seen
impedance falls within the first quadrant, which completely encompasses the
tripping region. In the tripping region, there are certain areas where the measured
current will not exceed the pickup setting, while there are cases where the current
may exceed the 67P pickup setting, resulting in a false trip. To better learn and
investigate these cases, a simulation study was performed, and the results are
presented in Figure A3.
71
Figure A1: Fault Current Characteristic of WTG Types 3 and 4
Figure A2: 67P Characteristic RX-Plot
Figure A3 shows the seen impedances as measured by the 67P element protecting a
wind farm collector for 11 simulation cases [22]. In the first eight cases, the voltage
was varied from 0.95 pu to 1.1 pu while the WTG active power generation was almost
zero, and PAux was about 10% of the WTG rated power with a power factor of 0.85
lagging. In addition, the 67P pickup setting was set to 120% of the entire collector
WTG rated auxiliary power. As shown in this figure, when the collector main bus
voltage was forced to voltages such as 0.95 pu or 0.97 pu, the impedance seen by the
67P fell into the fourth quadrant outside of the typical directional characteristic
region. Hence, it was not of any concern. For voltages between 1 pu and 1.06 pu, the
impedance seen by the 67P fell within the directional characteristic in the first
72
quadrant. However, as the current value was below the pickup setting, the 67P
element did not pick up. For voltages equal or above 1.07 pu, the absorbed reactive
current by WTGs exceeded the pickup setting, and the 67P element falsely tripped.
Four more cases were also simulated where the active power generated by the WTG
exceeded PAux. As shown in Figure A3, the 67P falsely tripped for the cases of V=1.1
pu, Pg=25% and V=1.08 pu, Pg=50%. For the case of V=1.03 pu, Pg=25%, the seen
impedance fell within the forward region, although the 67P did not trip because the
current measured by the relay was less than the pickup setting. As illustrated in the
same figure, the 67P seen impedance angle exceeded 120 ◦ and fell outside of the 67P
characteristic for V=1.04 pu and Pg=60% while the measured current was above the
pickup setting. In conclusion, the combination of very high overvoltage and medium
to very low active power generation will increase the risk of false trips by the 67P
element.
Figure A3: 67P Seen Impedance and Trip Status for 11 Simulated Cases
One solution to minimize the potential of false tripping of the 67P element is to
contract the directional characteristic, as shown in Figure A4. This solution eliminates
any false trip for cases in which the impedance seen by the 67P is within the second
73
quadrant where the collector wind power generation exceeds PAux. However, the 67P
element may still trip in the extremely rare case of very small power generation close
to zero combined with a very large overvoltage of more than 6% on the collector bus.
Figure A4: Secure directional element by contracting the directional

characteristic
Similar to the first solution, as illustrated in Figure A5, the use of load encroachment
eliminates any false trip for cases in which the impedance seen by the 67P are in the
second quadrant (if the reverse load encroachment angles are set to 90 ◦ and 270◦) and
where the aggregate WTG power generation exceeds PAux. However, the 67P element
may still trip in the extremely rare case of very small power generation close to zero
combined with a very large overvoltage of more than 6% for the simulated system.
74
Figure A5: Secure directional element by a load encroachment element

(Simulation Results)
75

Protection of Wind Electric Plants: PSRC Working Group C25

Uploaded by

Protection of Wind Electric Plants: PSRC Working Group C25

Uploaded by

Protection of Wind Electric Plants

PSRC Working Group C25

Protection of Wind Electric Plants

Power System Relaying and Control Committee

Members of the Working Group

Chair: Martin Best Vice Chair: Amin Zamani

Members and Contributors

ABBREVIATIONS AND ACRONYMS

KEYWORDS .............................................................................................................. iii

3.3.1 Zone-Interlocked Scheme ........................................................................ 43

THIS PAGE LEFT BLANK INTENTIONALLY

• Generator step up transformers

• Possible Wind Farm electrical layouts

• Voltage and frequency control requirements

2 Difference Between Wind Electric Plant Substations and

2.1 Collector Feeder Design and Characteristics

at the highest medium voltage (34.5 kV in North America or 33 kV in Europe), they

Figure 1: A typical collector feeder [10]

2.2 Wind Electric Plant Substation Arrangements

Utility grid Utility grid Utility grid

52-1 52-2 52-3 52-1 52-2 52-3

Collector 1 Collector 2 Collector 3 Collector 1 Collector 2 Collector 3 Collector 1 Collector 2 Collector 3

Figure 2: Common collector substation transformer configurations

Utility grid Utility grid Utility grid

52-1 52-2 52-3 52-1 52-2 52-3 52-4 52-5 52-6

Collector 1 Collector 2 Collector 3 Collector 1 Collector 2 Collector 3 Collector 4 Collector 5 Collector 6

Figure 3: Common collector substation arrangements (MV bus configurations)

Utility Utility Utility Utility Utility

WEP WEP 1 WEP 2 WEP 1 WEP 2

(a) (b) (c)

Utility Utility Utility Utility

Figure 4: Different HV bus configurations: (a) single bus configuration, (b)

Table 1: Cost/Reliability Comparison for Different Bus Configurations [9]

Reference Bus Configuration Relative Relative Application

Fig. 4(c) Ring 53.3% Medium X

2.3 Wind Electric Generator Characteristics

Figure 5: Typical configuration of a Type 1 WTG

Figure 6: Typical configuration of a Type 2 WTG

Figure 7: Typical configuration of a Type 3 WTG

Figure 8: Typical configuration of a Type 4 WTG

Therefore, generator contributions to faults can be calculated from the generator

Figure 9: Typical configuration of a Type 5 WTG

2.4 Fault Currents and Equipment Ratings

An important feature of the fault responses of Type 1, 2 and 3 (crowbarred) WTGs is

2.5 System Grounding

2.6 Transformer Connections and Characteristics

2.6.2 Main Substation Transformers

Figure 10: Simplified One Line Diagram of Wind Electric Plant

2.7 Harmonics and Sub-harmonics

a series capacitor. Sub-Synchronous Resonance occurs when the mechanical mass of

• It is important for wind developers to be aware of potential problems of WTGs

2.8 Voltage and Frequency Control Requirements

3 Typical Protective Relay Schemes at Wind Electric Power

transformers at more than two or three values of fault current, a multiplicity of

element can be increased from instantaneous to a delay of 10 - 12 cycles to bypass

Figure 11: Non-Directional Phase Relay Time Overcurrent Coordination with

Figure 12: Non-Directional Phase and Negative Sequence Relay Coordination

Figure 13: Non-Directional Ground Time Overcurrent Relay Coordination with

Non-directional instantaneous or short time neutral or residual ground overcurrent

Figure 16: Phase Directional Element 67P With Adjustable Directional

Additional discussion of directional phase overcurrent setting considerations for

3.1.2.2 Temporary Overvoltages

a zero-sequence source of fault current. A properly rated grounding transformer will

3.1.3 Arc Flash Protection

3.1.4 Removal of WTGs from Collector Feeders Under Fault

3.2 Grounding Transformer Protection

Figure 18: Delta-Wye Grounding Transformer

Figure 19: Zig-Zag Grounding Transformer

3.3 Bus Protection

significantly larger. It is common to have transformers larger than 100 MVA in

3.3.1 Zone-Interlocked Scheme

Figure 20: Zone Interlocked Bus Differential

Non-directional instantaneous overcurrent elements monitor the current through

complex networking between devices is required, and coordination for both

3.3.2 Percentage-Restrained Differential

Figure 21: Percentage-Restrained Bus Differential