Power Quality Impact of Distributed Generation:
Effect on Steady State Voltage Regulation
Reginald Comfort
Manuel Gonzalez
Arshad Mansoor
Phil Barker & Tom Short
Ashok Sundaram
Reliant Energy
EPRI PEAC
EPRI
942 Corridor Park Blvd.
Knoxville, TN 37932
3412 Hillview Avenue
Palo Alto, California 94304
th
1111 Louisiana, 24 Floor
Room 2452, Houston, TX
77002-5231
ABSTRACT
With the advent of deregulation, distributed generation will play an increasing role in electric distribution
systems. Various new types of Distributed Generation (DG), such as micro-turbines and fuel cells, are
now being developed in addition to the more traditional solar and wind power. A common belief among
developers is that DR will improve power quality, and this potential for better quality is cited as one of
the value attributes of installing distributed generators. In some cases distributed generation and storage
are being promoted as an answer to the premium-quality power requirements of high technology or
sensitive end-use customers. Whether or not this value-attribute of DG is valid will depend on the
specific technologies, site conditions and potential interaction with the existing electric power system.
The objective of this paper is to provide a technical assessment of the impact of distributed generation
technologies on the power quality of the power distribution system. Power quality is a broad term
covering a wide range of operating parameters including both steady state and dynamic conditions. The
full range of power quality conditions are described in IEEE Std. 1159-1995 Recommended Practice for
Monitoring Electric Power Quality. This paper focuses on steady-state voltage regulation impacts of DG
and is the first of several papers covering the various power quality impacts of DG. The guidelines
provided in this paper will help utility engineers evaluate the impact of distributed generation on voltage
regulation and identify methods to mitigate problems that arise. The paper also makes recommendations
for voltage trip thresholds to be used for DG interconnection that will help reduce the susceptibility of DG
to nuisance trips but still provide utility system protection against sustained overvoltage.
Introduction
Distributed generation will affect steady-state voltage for end users. Voltage regulation can be improved
or degraded by the addition of DG. When DG improves voltage regulation, this is generally referred to as
“voltage support,” commonly cited as one of the benefits of DG. However, voltage support is by no
means guaranteed, and in some cases DG will actually degrade voltage regulation. This paper discusses
the impacts of DG on distribution system voltage regulation and the conditions under which voltage is
degraded.
Voltage regulation variations are defined as:
•
Long-Duration Variations: Rms deviations at power frequencies for longer than one minute. [IEEE
Std. 1159-1995] [1]. Long-duration variations can be caused by:
-
Improper voltage regulation
-
Broken neutral
Utility-initiated voltage reduction
Sudden change of load or unusual load on a distribution circuit
Figure 1 shows an example of a long-duration undervoltage scenario.
24-Hour Minimum/Average/Maximum Voltage
100
Voltage (%)
97
94
Undervoltage
91
88
85
7:12 PM
12:12 AM
5:12 AM
10:12 AM
3:12 PM
Figure 1. Example of a Long-Duration Undervoltage [2]
ANSI/IEEE standard C84.1 specifies the preferred voltage levels for electric power systems. Most
utilities have adopted the C84.1 standard as a minimum requirement, and some have more stringent
requirements imposed by their respective public utilities commission. The C84.1 standard has two ranges:
range A and range B service voltages [3]. These are defined as follows:
•
Range A – Proper range considered to be within limits.
•
Range B – Infrequent excursion – “The occurrence of service voltages outside of these limits should
be infrequent” and “When they occur, corrective measures shall be undertaken within a reasonable
time to improve voltages to meet Range A requirements.”
The standard gives ranges at two locations, the service voltage (“voltage at the point where the electrical
system of the supplier and the electrical system of the user are connected”) and utilization voltage
(voltage at utilization equipment). The C84 ranges are shown in Figure 2.
Voltage (120-V Base)
104
108
112
116
120
124
128
Utilization Voltage
Service Voltage:
120- to 600-V Systems
Range A
Service Voltage:
Systems 600 V or Greater
Utilization Voltage
Service Voltage: 120- to 600-V Systems
Range B
Service Voltage:
Systems 600 V or Greater
Nominal
System
Voltage
Figure 2
ANSI C84.1 Voltage Ranges. [3] The Shaded Portions on the Lower End of the Utilization
Voltages do not Apply for Circuits Supplying Lighting Load. The Shaded Portion at the Top of
Range A Utilization Voltage Does Not Apply to 120 to 600-V Systems
Long-duration under and overvoltages measured during a national power quality study [4] are shown in
Table 1. Long-duration variations are less common than many other power quality disturbances including
voltage sags and momentary interruptions. Long-duration undervoltages are more common than
overvoltages.
Table 1
Average Annual Number of Long-Duration Overvoltages and Undervoltages Per Site Recorded
in the NPL Power Quality Study [4]
Sustained overvoltage
V>110% for t>10 sec
0.3
Sustained
undervoltage
V<87% for t>10 sec
1.0
Regulation in Distribution Systems
Maintaining proper voltage regulation of the distribution system is a key operating objective and
responsibility of the electric utility company. Voltage conditions that are out of the normal operating
range can lead to poor operation of the power system and customer loads, damage to utility equipment
and customer loads, and in extreme cases can even be a safety hazard.
Generally speaking, utilities strive to meet ANSI C84.1 Range A service voltage standards, while
realizing that some customers some of the time won’t quite be in this range. On a 120-volt base, the
standard requires that the service entrance voltage be between 114 volts and 126 volts. To those
unfamiliar with power systems this may seem like a broad range that should easily be achievable.
However, given the lengths in distribution circuits, and the significant variations in load that occur, this is
not easy to achieve for all customers. It requires careful coordination and placement of voltage regulating
equipment, diligence in monitoring load growth on the system and in performing periodic load flow
studies to assess voltage conditions. Figure 3 shows the distribution of percentage range of daily voltage
changes measured in the EPRI DPQ study. Most undergo less than a 3% change, but a sizeable portion
has swings of more then 7%.
Distribution of Daily Range of RMS Voltage
18%
100%
16%
Frequency
12%
60%
10%
8%
40%
6%
4%
Cumulative Frequency
80%
14%
20%
2%
19%
18%
17%
16%
15%
14%
13%
12%
11%
10%
9%
8%
7%
6%
5%
4%
3%
2%
1%
0%
0%
0%
Daily RMS Volatge Range (% of Site Average Voltage)
Figure 3
DPQ Results for Regulation Voltage Ranges. The Plot Shows the Percent Daily Change in
Voltage Occurring Due to Demand Cycles on the Power System [2]
For radial power systems, voltage regulation practices are based on a single source of power (the
substation) and the power taking only one path from the substation to all loads on the system. This
condition leads to the assumption that the voltage will always drop on the primary feeder as the distance
from the substation increases. The only exception to this assumption is when there is too much reactive
compensation (this will cause a rise in voltage as one moves towards a capacitor bank). Utilities are
careful to avoid this condition, so it is reasonable to assume that the voltage drops on the primary feeder
for most applications. The condition of radial flow also implies that the voltage (on a per-unit basis) will
drop across each distribution transformer and secondary service (see Figure 4). DG introduction onto the
radial distribution system will impact both of these basic assumptions used for voltage regulation.
130
128
Voltage, V with base=120 V
126
ANSI Range A Upper Limit
124
Primary
feeder drop
122
120
118
Transformer
drop
116
Secondary
drop
114
ANSI Range A Lower Limit
112
110
0
1
2
3
4
5
6
7
8
Distance from the substation, miles
Figure 4
Example Voltage Profile on a Distribution Circuit
For network systems, voltage regulation is usually done entirely with load-tap changing (LTC)
transformers at substations, and there is no supplementary voltage regulation equipment on the primary
feeders or on the LV network side. Such equipment is not usually necessary because networks, being in
an urban environment, have relatively short primary feeder lengths, and the secondary grid is very stiff
electrically and tightly interconnected in a manner that allows little voltage drop. The amount of voltage
drop from the substation to the service entrance of a typical network customer is far less on a per-unit
basis than on a typical suburban or rural radial distribution circuit. The LTC device in the substation alone
is usually enough to compensate for any voltage drops.
In general, urban networks are good examples of systems that are thermally limited – this means that as
the load is increased, the transformers and cables reach their thermal limits long before voltage drop
becomes a major issue. In contrast, longer radial distribution systems are usually voltage drop limited –
meaning that the voltage drop becomes too excessive well before any thermal limits are reached. In
general, urban low-voltage networks, since they are thermally limited, will be less sensitive than radial
distribution circuits with respect to the impact of DG on voltage regulation. This means it takes a lot of
injected power to significantly alter the voltage on a network. It is important to recognize that while low
voltage urban networks are less sensitive with respect to DG’s impact on voltage regulation, that they are
actually more complex to deal with and more sensitive than radial distribution systems in regards to other
issues such as protection and fault coordination. This has to do with the directional relaying employed to
protect network systems. The relaying and protection issue is not discussed in this paper, but it should be
recognized as a serious complication and will be discussed in follow-up papers.
Distribution utilities have several ways to control steady-state voltage. The most popular regulation
methods include:
•
Substation load tap-changing transformers (LTCs)
•
Substation feeder regulators
•
Line regulators
•
Switched capacitors
As stated before, most networks use only LTCs. However radial circuits in suburban and rural areas may
use any of the above devices. For radial systems, most utilities use LTCs to regulate the substation bus
and supplementary feeder regulators and/or switched capacitor banks where needed. On relatively short
feeders, utilities may employ only an LTC transformer and switched capacitors. A few utilities use only
switched capacitors for regulation.
Voltage regulators are autotransformers with automatically adjusting taps. A common regulator range is
±10% with 32 steps (5/8% each). The tap changer controls are often adjusted to control the voltage with
line-drop compensation. This compensates for downstream voltage drop due to the impedance of the line.
Using line-drop compensation, it is possible to hold the voltage nearly constant at a point on the feeder
several miles downstream from the regulator. During heavy load, the voltage at the regulator output
terminals is boosted the most, and during light load it is boosted the least. The amount of boost is directly
proportional to the current on the line and offsets the voltage drop between the terminals of the regulator
and the point where the voltage is to be regulated (this point is called the constant voltage point or load
center). The line drop compensator uses an internal model of the impedance of the distribution line to
determine how much voltage boost is required for a given current level (see Figure 5). The current and
voltage on the line are measured with a CT and PT at the regulator. The user can set the R and X values in
the line drop compensator to adjust the compensation based on the type and size of wire (its impedance
characteristics) and the presumed load distribution on the feeder. The set voltage on the regulator is the
value it will hold at the constant voltage point (load center).
I
CT
R
PT
V/pt
I/ct
X
(R+jX)(I/ct)
R
Regulation poin
X
Voltage
Regulating
Relay
Figure 5
Line Drop Compensator Circuit
In selecting the line-drop compensator settings, one practice is to attempt to hold the voltage constant at
the physical center of the line (or center of the regulated region). For example, on a 5-mile line regulated
entirely from the substation, the voltage could be held constant at about the 2.5-mile point. In
determining the settings it is often assumed that the load is uniformly distributed along the line and
uniform wire size is used throughout the feeder section being regulated. To satisfy the objective of
holding the voltage constant at the center of the section being regulated under these conditions, analysis
shows that the R and X settings need to be set to 3/8th of the total line resistance and reactance of the
section being regulated. Some utilities add a bit of additional compensation to account for distribution
transformer and secondary voltage drops. The set voltage and compensation levels are also evaluated
from the perspective of the “first” and “last” customer. The first customer (closest to the source from a
voltage-drop perspective) is evaluated such that they will not see high voltage at any time, and the last
customer (farthest from the source from a voltage-drop perspective) is checked for low voltage. If the
settings can meet these criteria and provide nearly constant voltage at the center of the regulated region,
then they should work effectively and be considered good practice.
Load tap changing transformers (LTC) use the same type of controller with line-drop compensation as do
auto-transformer step regulators, but they often regulate an entire substation bus that serves many
distribution feeders. The LTC must therefore regulate several lines at once, so the characteristics of each
line need to be considered in the settings. It is not usually possible to pick the ideal settings for all feeders
served from a common bus. For this reason, many utilities apply compromise settings for the line drop
compensation that amount to using far less compensation than would be the case for any single feeder
application. In fact, some utilities use none at all (R=0 and X=0). When no line drop compensation is
applied it means that the constant voltage point is at the regulator location – which in this case would be
the substation bus. The use of no line-drop compensation leads to greater fluctuations in voltage for
customers at the end of the regulation zone.
Another complication in adjusting regulators is that the load may not be uniformly distributed on the
feeder. This means that regulator settings when based on uniformly distributed loads may cause a
mismatch between desired and actual voltage profile results. Load may be biased towards the front or rear
of the feeder depending on the situation. When load is at the rear, more line-drop compensation is needed.
When it is towards the front, less is needed. Capacitor banks and distributed generation are current
sources that confuse voltage regulators by making the load less uniformly distributed and perhaps even
causing reverse flows. Regulator operation should be coordinated with these devices when they are
significant in size relative to the system loads.
Regulator controls also have two other settings that help prevent an excessive number of tap changes:
Bandwidth: Voltage regulator controls monitor the difference between the measured voltage and the
voltage setpoint. When the difference exceeds the bandwidth, a tap change is initiated. As a rule of
thumb, the minimum bandwidth should be 2 times the step size (1.25% for 5/8% steps).
Time delay: After a tap-change operation, further changes are permitted until the time delay has elapsed.
Typical time delays are 30 to 60 seconds.
Summary of DG Impacts
DG can improve regulation or cause problems with regulation. The main ways that DG can cause
regulation problems are:
•
Low voltage due to DG just downstream of a regulator with line-drop compensation: Linedrop compensation is the technique commonly applied by LTC transformer controllers and linevoltage regulators to control the voltage on the distribution system based on the line current.
Under heavy load, a generator just downstream of the generator will reduce the observed load on
the feeder (so the regulator will not boost the voltage as much). This leads to lower voltage
downstream of the regulator.
•
High voltage due to DG: High voltages may be caused by reverse power flow. Under light load
for a location where the primary voltage is already high, the voltage rise can be enough to push
the voltage above ANSI limits. This can even happen for a small DG located on the secondary
because of the voltage drop along the service drop, the secondary wiring, and the distribution
transformer.
•
Interaction with regulating equipment: Another area of concern is with interaction of
regulation equipment and DG. If the DG has varying output, it may change the system voltage or
current flows enough to cause a regulator tap change or an operation of a switched capacitor.
Likewise, a distributed generator that has feedback to control voltage may interact negatively to
the utility regulation equipment. There may be undesirable cycling of regulation devices and
noticeable power quality impacts under such conditions.
DG Operating Ranges per Standards
Table 2 shows the normal operating range of photovoltaics per the IEEE 929-2000 interface standard [5].
Note that the high-end limit of 132 V is well above the ANSI range B upper limit of 127 V for the service
and utilization voltage. Since DG can cause a voltage rise because it injects real power, this is
inconsistent with the ANSI range B upper limits and could lead to long-duration overvoltages caused by
DG. The IEEE P1547 interconnection standard that is currently in draft form that applies to all
distributed generation was initially based on IEEE 929 but has evolved slightly to have different trip
levels and tripping times as shown in Table 3. Qualitatively, the proposed settings of IEEE P1547 are still
quite similar to IEEE 929 and both standards allow sustained DG operation outside of the ANSI C84.1
limits. Within the industry, this topic has been an area of active discussion as to the pro’s and con’s of this
approach. In subsequent sections of this paper, the voltage regulation issues related to DG are discussed,
and recommended alternative settings for IEEE 929 and 1547 are proposed.
Table 2
Standard Trip Thresholds (120-V nominal) for DG Operations per the IEEE 929 Photovoltaic
Interface Guideline
RMS Voltage
Trip Time*
V<60
6 cycles
60≤V<106
120 cycles
106≤V≤132
Normal Operation
132<V<165
120 cycles
165≤V
2 cycles
Table 3
Proposed voltage trip settings per IEEE P1547-D7. Note the IEEE P1547 document is a
proposed standard under development and these settings could change pending
committee review
Voltage Range (% of base
Note 1
voltage)
Clearing Time
(seconds)
Note 2
V<50
0.16
50≤V<88
2
88≤V<110
Operation Allowed (no trip)
110≤V<120
1
V• 120
0.16
Note 1: Base voltages are the nominal voltages stated in ANSI C84.1
Note 2: DR≤30 kW Maximum Clearing Time. DR>30 kW, Default Clearing Time or area EPS
operator may specify different voltage settings or trip times to accommodate area EPS system
requirements
Voltage “Following” versus Voltage Regulating DG Units
DG is usually operated in a voltage-following mode. This means that the DG makes no intentional
attempt to regulate the voltage on the power system. With this mode DG simply supplies real power at
nearly constant power factor and the voltage on the feeder changes according to the effects of the DG
power injection – in other words, the DG does not attempt to force a given voltage by using reactive
power control. The concept of voltage following is a source of confusion within the industry since many
people have incorrectly interpreted its meaning to be that DG will “follow” the existing utility system
voltage and will not change the voltage on the feeder in any way. However, the basic physics of injecting
power into any power system guarentee that the voltage will change as a result of current flow through the
system impedance. These changes will be superimposed (add to or subtract from depending on phase
angle of the injected currents) on the existing utility system voltage that is controlled by utility equipment
such as LTC’s and feeder regulators.
In a voltage following mode, as the real power output of the DG increases, the voltage will increase! Too
much power injected into the power system can cause higher than normal voltage. If relays are set to
allow the unit to continue to operate into the grid with voltage outside of normal voltage limits (as is the
case in the proposed P1547), then sustained overvoltage can occur.
When DG is operated in a voltage-regulating mode it attempts to hold the voltage at a constant set point.
Voltage regulation can be accomplished by adjusting the reactive component of the generator output
either to raise or lower the voltage to offset any rise due to the real power component. This is done with
rotating synchronous generators by adjusting the field excitation level. Leading reactive current (high
excitation) will increase the voltage and lagging current (low excitation) will decrease the voltage. As
long as the generator has sufficient capability to support the reactive and real power requirements
imposed by this type of duty, it can help regulate the voltage on the power system. This type of operation,
while done in some cases, is mostly avoided in grid parallel applications because it is usually not
desirable to give this responsibility to independently owned generation as it requires considerable
coordination with upstream equipment and the utility company. Also, many DG are too small (do not
have sufficient reactive capability) to regulate the distribution system effectively. Finally, it can be
difficult to coordinate operation of the independently owned DG with utility feeder regulation devices in a
suitable manner that may not severely penalize the generator real power capacity. Note that the voltageregulation mode of operation is preferred in stand-alone grid independent DG applications where the
generator supplies power to all of the load and is needed to perform the voltage regulation function.
Voltage Regulator Interaction
There can be considerable interaction between DG units and utility system voltage regulation equipment.
Figure 6 shows a scenario where a DG just downstream of a regulator can cause low voltage on the end of
the circuit due to the use of line-drop compensation on the regulator. The power injected by the generator
will cause the regulator to not raise voltage as much as it should. To determine if the DG will cause a
significant impact on the feeder voltage, the size and location of the DG, the voltage regulator settings,
and impedance characteristics of the line must be considered. The low voltage is exacerbated if there is
reverse power flow. Under this condition, the regulator can ratchet all the way to the lowest tap, which
would cause very low voltage downstream.
Regulator
FEEDER
(QGRI)HHGHU
Injected Power
G
Peak Load (no DG)
Peak Load (with DG)
9ROWDJH
CVP
ANSI Range A Lower Limit
Regulator
End of Feeder
Figure 6
DG Just Downstream of a Regulator that Leads to Low Voltage at the End of the Feeder
A simple screening criteria to avoid low voltage on circuits with line-drop compensation is that there may
be problems if the size of the generator exceeds 10% of the load at the regulator and the DG is located
closer to the regulator than the load center. When these conditions are met, then further study may be
warranted to investigate the decrease in voltage on the feeder and determine if mitigation is required.
If problems are found, there are several options. One option for minor undervoltages is to reduce the linedrop compensator settings and raise the voltage setpoint slightly. This moves the regulator constant
voltage point closer to the regulator and reduces the impact of the DG. The key is the location of the
constant voltage point (also called the load center – the fictitious point on the feeder where voltage is held
constant). DGs upstream of this point will lower the voltage downstream of this point, while DGs
downstream of the constant voltage point will support rather than lower the voltage. A more extreme
option for larger DGs is to remove (turn-off) the line-drop compensation on the regulator. Generally, the
voltage setpoint will have to be significantly raised to provide adequate voltage on the circuit at the end of
the regulation zone. If this is done, a check would have to be made to make sure high voltages don’t occur
on the feeder at light load due to the higher set voltage.
If the problems are due to reverse power flow, advanced regulator controllers are available to change
operating mode during reverse power flow to prevent the regulator from dropping to the lowest tap. For
example, the tap change can be blocked or programmed to revert back to the neutral position if power
reverses.
Other options include moving either the generator or the regulator so that the regulator is downstream of
the generator. Sometimes there is no way to avoid installation of new regulation equipment if the DG is
especially large relative to feeder capacity.
Another consideration with regulators (and switched capacitor banks) is unwanted interaction with DG
(especially fluctuating sources). The main way to avoid this is by widening the regulator bandwidth
setting and, in some situations, reducing line drop compensation settings. Increasing the time delay can
also help prevent excessive regulator tap changes depending on the rate of change of the DG energy
source. Wind and photovoltaic energy sources that impact the feeder load by 10% or more at voltage
regulators or switched capacitors are the most likely causes of these types of interactions. Wind and PV
energy sources experience wide fluctuations in output on a repeating basis due to partial cloud cover or
wind variations. These fluctuations may occur over many seconds or minutes depending on conditions.
DG-Caused High Voltages
There are also concerns with DG causing high voltages on distribution circuits because of reverse power
flow. It is possible to estimate the effect of a generator by using the standard voltage drop equations with
reverse power flow. The voltage drop along a feeder due to a load is approximately equal to:
Vdrop ≈ IR·R+IX·X
where,
Vdrop, voltage drop along the feeder
R, line resistance, ohms
X, line reactance, ohms
IR, line current due to real power flow, amps (negative for a generator injecting power)
IX, line current due to reactive power flow, amps (negative for a capacitor)
The voltage at the generator can be estimated by taking the highest prefault voltage and adding the
voltage rise due to the generator from the equation above. Note that this approximation is no substitute for
a proper load flow. It does not fully model the response of the load to the change in voltage, and it does
not consider regulator response. It is useful for a first attempt at estimating whether the voltage rise due
to the generator might be a problem.
The DG which is exporting mainly real power will cause voltage to go up most where X/R ratios are low.
The real power portion will cause the largest voltage rise when the line resistance is high. If the DG is
injecting vars like a capacitor or there are fixed capacitors nearby, the voltage rise is even larger. Under
the right circumstances, this voltage rise is beneficial (voltage support), but if too much rise occurs or it
occurs on a section of feeder where the voltage was already near the upper ANSI limit before the DG
started, then a high voltage problem may be created (see Figure 7).
SUBSTATION
(QGRI)HHGHU
FEEDER
Injected Power
G
9ROWDJH
ANSI Range A
Upper Limit
ANSI Range A
Lower Limit
Substation
After DG
Before DG
End of Feeder
'LVWDQFH
Figure 7
Voltage profile on feeder before and after addition of large DG. High voltage may occur at
the end due to the voltage rise caused by the DG injected power
If voltage rise can be a problem, there are several options. One would be to limit the size of the generator
to below the level necessary to cause problems. Another would be to relocate the DG to a more suitable
location on the distribution circuit or build an express feeder to the generator.
On the DG side, the generator could be operated to absorb more reactive power (by removing local
capacitors or operating a synchronous generator or line-commutated inverter at reduced power factor).
This is the opposite of what is normally done and reduces the T&D support value of DG. Reducing the
power factor of the generator causes voltage drop due to the reactive component of the generator. This
approach may increase losses on the feeder, however, and result in an effective reduction in available
capacity on the feeder.
A utility company option would be to reduce the resistance of the lines and transformers from the
substation bus to the DG site – a costly scenario. This would be done by using larger conductors on lines
and cables and specifying lower copper losses on transformers. Another utility-side option is to add
regulation equipment (capacitors or regulators) to counteract the voltage rise from the distributed
generator. Of course, the least costly utility option could be reducing the voltage level setting on the
existing regulation equipment – that is, if this can be done without risking low voltage on the feeder
during periods when the DG is off and/or there is heavy load. While these are all technically feasible
option the question remains who is going to pay for the technical study and implementation for changing
distribution system practices to accommodate customer owned DG.
Selecting DG Trip Thresholds That Work in the Real World
Sometimes DG units will unneccessarily trip offline due to short-term excursions of voltage that are
slightly outside of the steady state operating limits defined by ANSI C84.1. Temporary high voltages on
the distribution system can occur for a variety of reasons including the loss of a single large load or block
of loads when a fuse, switch or breaker interrupts power flow. Also, the switching of a large power factor
correction capacitor can result in a sudden increase in voltage. These events frequently lead to mild
overvoltages up to about 10% above nominal that persist until utility voltage regulation equipment can
respond to correct the situation. Utility voltage regulators usually employ time delays of anywhere from
10-90 seconds. Short term mild overvoltages of this nature are not a threat to loads, but they will cause
nuisance trips of DG that strictly use the ANSI C84.1 limits as the acceptable operating window and trip
offline very fast for anything outside the range.
To prevent nuisance DG trips, the IEEE 929 PV standard and various utility DG interface requirements
have employed voltage trip settings that allow DG to continue operating indefinitely even when the
voltage is slightly outside of the ANSI limits. For example, IEEE 929 allows PV inverters to operate at
up to 132 volts (+10% above nominal) before they are required to trip. Draft-standard IEEE P1547-D7
that is still in development also allows for a +10% limit. The upper ANSI range B limit is 127 volts (+6%
nominal) so both IEEE 929 and IEEE P1547-D7 allow the DG to function considerably higher than the
ANSI range B limit.
This approach certainly will reduce the incidence of nuisance trips, however, it also creates the potential
of sustained high voltage above ANSI range B limits being allowed to occur as a result of DG power
injection. In some cases this could be a threat to loads and power system equipment - potentially
causing damage and perhaps even posing safety issues. It is recommended that the industry consider an
alternative approach that provides suitable protection against sustained overvoltage but will still allow DG
to ride-through mild overvoltages without nuisance trips. This approach essentially means that DG should
have an acceptable continuous operating window that is consistent with ANSI C84.1 and should use a
long time delay for mild overvoltages. Table 4 shows recommended voltage trip settings for DG that
should offer good protection against nuisance trips and sustained overvoltages. This table essentially
follows the settings in the IEEE P1547-D7 draft standard, but includes an extra row of settings for
voltages ranging from 106% to 110% of nominal with a time delay of 180 seconds. It also allows a
continuous (no trip) operating range of only 88% to 106% that is a bit narrower than the current draft
standard.
Table 4
Alternative Voltage Trip Settings Recommended for IEEE P1547
Voltage Range (% of base
Note 1
voltage)
Clearing Time
(seconds)
Note 2
V<50
0.16
50≤V<88
2
88≤V<106
Operation Allowed (no trip)
106≤V<110
180
110≤V<120
1
V• 120
0.16
Note 1: Base voltages are the nominal voltages stated in ANSI C84.1
Note 2: DR≤30 kW Maximum Clearing Time. DR>30 kW, Default Clearing Time or area EPS
operator may specify different voltage settings or trip times to accommodate area EPS system
requirements
An instructive example of an overvoltage event and the DG nuisance trip response to it recently occurred
at a utility distribution system. In this case a 5 MW combustion turbine was operating in parallel with the
utility system at an industrial load that is supplied by a 12 kV distribution circuit. The DG unit is located
3 miles from the substation. The DG trip settings were in accordance with Texas PUC Rule 25.212 where
the first trip threshold is 126 volts with a 30 second time delay and the second trip threshold is 132 volts
with a 10 cycle time delay. In this case the generator unit was operating normally until the voltage at the
DG site suddenly increased to approximately 126 volts (on a 120 volt base) in the first stage. The first
stage of the voltage rise shown in Figure 8 was a result of the switching-on of a large bank of substation
capacitors used for transmission support and was not directly related to the DG since it was not exporting
net power to the feeder at that time (although by offsetting load on the feeder due to this customer it did
help raise the voltage).
Figure 8
Voltage rise; 7.281 kV, then 7.559 kV for 28 sec., then 7.809 kV for 14 sec.
The DG interconnection breaker tripped after the 30 second time delay had elapsed and since this also
caused the loss of customer load the voltage rose to 130V (on a 120V base) before the voltage on the
feeder was corrected by the substation LTC. This event is very instructive as it shows that the time delay
setting of the DG overvoltage relay had not been coordinated with the feeder voltage regulation
equipment. If the time delay had been set to 15 seconds longer than the LTC’s time delay then that would
have provided sufficient time for the LTC to lower feeder voltage and the generator would not have
tripped. Note that the present draft 7 IEEE P1547-D7 settings would have worked well for this case,
however, as stated before, they would also allow continuous operation at an elevated voltage should that
condition arise. The preferred and technically better approach is the settings recommended in Table 4.
These provide good mitigation of false trips due to temporary overvoltages up to a few minutes
duration but will protect against a sustained overvoltage outside the ANSI C84.1 standard.
References
[1] IEEE Std. 1159-1995, IEEE Recommended Practice for Monitoring Electric Power Quality.
[2] An Assessment of Distribution System Power Quality: Volume 2: Statistical Summary Report, EPRI
TR-106294-V2, Palo Alto, California, May 1996.
[3] ANSI C84.1-1989, American National Standard for Electric Power Systems and Equipment - Voltage
Ratings (60Hz).
[4] D. S. Dorr, T. M. Gruzs, M. B. Hughes, R. E. Jurewicz, G. Dang, and J. McClaine, “Interpreting
Recent Power Quality Surveys to Define the Electrical Environment,” IEEE Transactions on Industry
Applications, vol. IA-33, No. 6, Nov/Dec 1997.
[5] IEEE Std. 929-2000, Recommended Practice for Utility Interface of Photovoltaic (PV) Systems.