VOLTAGE IMPACTS OF DG ON DISTRIBUTION GRID WITH VOLTAGE
REGULATOR AND SVC
A Project
Presented to the faculty of the Department of Electrical and Electronic Engineering
California State University, Sacramento
Submitted in partial satisfaction of
the requirements for the degree of
MASTER OF SCIENCE
in
Electrical and Electronic Engineering
by
Masood Afzal
SPRING
2012
© 2012
Masood Afzal
ALL RIGHTS RESERVED
ii
VOLTAGE IMPACTS OF DG ON DISTRIBUTION GRID WITH VOLTAGE
REGULATOR AND SVC
A Project
by
Masood Afzal
Approved by:
__________________________________, Committee Chair
Mohammad Vaziri
____________________________
Date
__________________________________, Second Reader
Fethi Belkhouche
____________________________
Date
iii
Student: Masood Afzal
I certify that this student has met the requirements for format contained in the University
format manual, and that this project is suitable for shelving in the Library and credit is to
be awarded for the project.
__________________________, Graduate Coordinator
Preetham B. Kumar
Department of Electrical and Electronic Engineering
iv
___________________
Date
Abstract
of
VOLTAGE IMPACTS OF DG ON DISTRIBUTION GRID WITH VOLTAGE
REGULATORS AND SVCS
by
Masood Afzal
Interconnection of the Distributed Generation (DG) at higher penetration levels to the
distribution grid is causing different problems with voltage profile of a typical
distribution feeder. In this paper, some of the voltage problems caused by increasing
penetration of DG along with the role of the Volt-VAr controlling devices to mitigate
these problems are investigated and discussed. A real distribution circuit containing a
synchronous generator, a line voltage regulator, and a Static VAr Compensator (SVC) it
was used for computer modeling. Simulations were conducted considering full load and
light load conditions, without, then with DG at various penetration levels. Simulations’
results verifying the high voltage conditions and unacceptable voltage flickers due to
increased penetration of DG have been presented. It was also concluded that voltage
regulator with certain settings and SVC can mitigate some of these problems.
_______________________, Committee Chair
Mohammad Vaziri
_______________________
Date
v
TABLE OF CONTENTS
Page
List of Tables ................................................................................................................. viii
List of Figures .................................................................................................................. ix
Chapter
1. INTRODUCTION .........................................................................................................1
2. VOLTAGE REGULATION STANDARD AND DEVICES........................................4
2.1. Voltage Regulation Standards in the U.S ...............................................................4
2.2. Voltage Regulator ...................................................................................................4
2.3. Static VAr Compensator (SVC) .............................................................................6
3. SIMULATION PARAMETERS ..................................................................................8
3.1. System Description ................................................................................................8
3.2. Scenarios ................................................................................................................9
4. RESULTS ...................................................................................................................11
4.1. Full Load with No DG .........................................................................................11
4.1.1
Full Load with Max DG .............................................................................13
4.2. Light Load with No DG .......................................................................................15
4.2.1. Light Load with Max Generation ..............................................................15
5. POSSIBLE SOLUTIONS ...........................................................................................17
5.1. Addition of Voltage Regulator .............................................................................17
5.1.1. Full/ Light Load with No DG ...................................................................17
5.1.2. Full Load with DG and Voltage Regulator ...............................................18
5.1.3. Light Load with DG and Voltage Regulator .............................................19
5.1.4. Interaction of DG with Voltage Regulator ................................................20
5.2. Solution by Installing SVC .................................................................................22
5.2.1. Full/ Light Load with No DG ...................................................................23
5.2.2. Full Load with DG and SVC .....................................................................23
5.2.3. Light Load with DG and SVC ..................................................................24
vi
6. CONCLUDING REMARKS ......................................................................................27
References .........................................................................................................................29
vii
LIST OF TABLES
Tables
Page
1.
Voltages at different generator output under full load ..........................................13
2.
Voltages at different generator output under light load ........................................16
3.
Voltages at different generator output under full load with voltage regulator .....18
4.
Voltages at different generator output under light load with voltage regulator ....20
5.
Voltages flicker at PCC for different generator output .........................................22
6.
Voltages under full load at different generator output with SVC ........................24
7.
Voltages under light load at different generator output with SVC .......................26
viii
LIST OF FIGURES
Figures
Page
1.
Cutout of the feeder, generated by CYME .............................................................9
2.
Simplified diagram of the feeder ..........................................................................10
3.
Feeder at full and light load with no DG ..............................................................12
4.
Voltage profile at 10 MW of DG output, full and light load ................................14
5.
Voltage graph under full and light load, at variable DG output ...........................16
6.
Voltage profile with voltage regulator at full and light load ................................17
7.
Voltage profile with voltage regulator at full and light load ................................19
8.
Full and light load profile at 10 MW of generation, with SVC ............................25
9.
Voltages at Point B with voltage regulator and SVC ...........................................25
ix
1
1. INTRODUCTION
The demand for electric power in the world is growing rapidly due to the population
growth and advent of technology. Considering the high growth, the existing power
system infrastructures will soon be inadequate to meet the future demands, thus requiring
alternative sources of energy. Over the last few years, a number of considerations like
environmental issues have introduced constraints on the construction of new transmission
lines. In addition, concerns about depletion of the world’s fossil fuel reserves have led to
interconnection of small scale Distributed Generation (DG) to the local distribution
systems. Some of the major benefits that DG can offer are, eliminating the need of new
transmission lines construction, reduction of the fossil fuel usage, and electricity market
liberalization [1]-[2]-[3].
Any renewable or conventional energy resources can be
classified as DG. In general, DG can be considered as Internal Combustion (IC) engines,
small wind turbines, micro turbines, geothermal plants, solar Photo-Voltaic (PV), or Fuel
Cells (FC) [4].
DG is usually installed at various locations including consumer sites and
interconnected to the distribution systems. Almost in all cases, the power distribution
feeders are “radial” in nature. A redial feeder is designed for unidirectional power flow
from the source to the load, contains no closed loops flow from the source to the load,
and contains no closed loops [5], [6]. The idea of DG was not under consideration in the
past, therefore the design of radial systems is not fully compatible and accommodating to
the interconnection of DG [5]. Despite the mentioned fact that the majority of distribution
feeders are radial, DGs are still being connected to these feeders. Addition of DG to a
2
radial distribution system usually will affect the feeder in several ways. For example, the
DG can alter the direction of power flow in the distribution system, causing protection
and control related issues [6], [16-18]. Adverse effects include fluctuation in voltage
levels, voltage flicker, power quality issues, degradation in system reliability, system
protection and harmonic distortion [5-6], [16-18]. The presence of DG in distribution
systems effects voltages and reactive power at locations close to DG installation. This
change in voltages and reactive power can be controlled locally. The operating voltage
profile of the feeder can be optimized by considering the load, DG power output, and its
mode of operation. The changes in loading condition or DG penetration level will directly
influence the operating voltage. The voltage and reactive power control in distribution
systems is typically done using; the On Load Tap Changers (OLTC) operations, shunt
capacitors, and voltage regulators. Today, electronic devices are also playing an
important role in maintenance of distribution systems voltages. Some of these devices are
Distribution Static VAr Compensators (D-SVC) and a Static synchronous Compensator
(STATCOM). High voltage conditions associated with DG at higher penetration levels
may also interfere with the objectives and control of other energy conservation programs
such as, Conservation Voltage Reduction (CVR). The main objective of CVR is to reduce
the utilization voltage to end-users. Purpose of using CVR is to reduce demands, and
energy consumption while maintaining proper voltage support to all of its customers.
CVR can be implemented using various strategies such as line drop compensation,
voltage control; voltage spread reduction, and Volt VAr Control. As the penetration
levels of DG increases on the feeder, higher voltages are expected in the vicinity of DG
3
installations. This increase in voltage is against the idea of lowering voltages to conserve
energy by CVR. Thus, increased penetration of DG negates the purpose of CVR. Detailed
analyses related to and the effects of DG on CVR will not be considered in this paper.
This paper focuses on voltage behavior in a typical radial system with the incremental
penetration of DG in the circuit. In this research, a real radial circuit is being considered
and simulated by using utility grade software, known as CYME®. For computer
simulations DG with variable output will be connected to the circuit, and its effects will
be analyzed considering the following four different scenarios;
1- Full load with no DG,
2- Full load with max DG,
3- Light load with min generation, and,
4- Light load with max generation.
Via simulation, the effects of DG will be demonstrated in each case, and the role of some
voltage control devices such as voltage regulators and D-SVC to mitigate these effects
will be studied. A literature survey of voltage regulation standard, voltage regulator and
SVC will be presented in chapter 2. Chapter 3 describes details of the feeder used for
simulation. Detailed simulation results, and possible solutions will be discussed in
chapters 4 and 5 respectively. The concluding remarks are presented in chapter 6.
4
2. VOLTAGE REGULATION STANDARD AND DEVICES
2.1.
Voltage Regulation Standards in the U.S
The process of keeping feeder voltage within a desired band is called voltage
regulation. The standard provided the American National Standards Institute (ANSI)
known as ANSI-C84.1 specifies the upper and lower limits of the utility voltages for the
distribution feeders [15], [17]

Under normal operating conditions, the regulation
requirement is +/- 5% on a 120-volt base.

Under unusual conditions, the allowable range for
these conditions is -8.3% to +5.8%.
Desired voltage level could be achieved by two ways; 1. By directly controlling the
voltages; that is, by lowering or raising the number of turns on the secondary side of
regulating device. Step voltage regulator and OLTC Transformers are examples of the
devices used for direct voltage control [7] [8]. 2. Voltages could also be regulated
indirectly by controlling the flow of reactive power in the feeder. Devices used for
indirect control of reactive power are shunt capacitors and SVCs [9].
2.2.
Voltage Regulator
A voltage regulator is an autotransformer with many steps also known as taps a
mechanism responsible for tap changing, and, a voltage sensing and control unit. By
changing its tap, voltage regulator can increase or decrease its output voltages, hence
controlling voltages on the sections downstream from its locations on the feeder. A
5
typical voltage regulator can increase (boost) or decrease (buck) the voltages up to 10%
providing total of 20% of regulation range. This boost or buck process is carried out in 32
steps with 0.62% voltage change per step [7]. Voltage control component has three basic
settings
 Base voltage and Line Drop Compensator (LDC) settings.
Base voltage set point, is the desired output voltage of the regulator at zero
current flow

Band Width, is tolerance/ difference at which the voltage regulator starts
regulating voltages

Time delay, is the time difference when regulator first detects the difference
between set point and measured voltages and corrective action starts.
Voltage sensing control receives input from current and voltage transformer to
constantly measure the system voltage. These measured system voltages are compared to
the “set point”. If the difference between measured and set voltage is more than one half
of bandwidth and it maintains itself for a period equal to time delay setting, the control
sends a command to lower or rise to the tap changing mechanism for proper action. If the
measured voltage is less than the set voltage, boost will be activated, if the measured
voltage is higher than the set point, a lowering process will be initiated. Line-Drop
Compensation (LDC) setting is used by line-voltage regulators to additional compensate
for the voltage drop on the line when the current increases. [6]. LDC in a voltage
regulator can be controlled by changing X and R setting on the unit. In our simulations,
the Line Drop Compensation functionality of the voltage regulator was disabled by
6
setting the resistance and reactance elements to zero. The voltage regulator was set to
regulate the output by the base voltage setting alone. This is known as “Regulator
Terminal” mode in the computer program. Other available operating mods include,
“Fixed Tap” and “Load Center”. Each of these setting for voltage regulator reacts
differently with DG. Here we used the “Regulator Terminal” mode for all simulations.
2.3.
Static VAr Compensator (SVC)
An SVC is a power electronic device, which can provide fast acting reactive power on
electrical networks in order to control voltages [9]. Typically, it is made up of coupling
transformer, thyristor valves, reactors and capacitors. By using high speed thyristor
switches and thyristor controlled devices, SVC provides reactive shunt compensation
which could be controlled quickly for dynamic voltage control [10]. Capacitors or
inductors can be added to or removed from the circuit on a per-cycle basis, making
control of system voltage a very fast process [9].Overall it acts as a shunt susceptance
which is controllable by predetermined settings of controls, which injects reactive power
into the system based on its terminal voltages. The SVC will inject reactive power into
system if the voltages at the node it is connected to begin to fall below its predefined
limits, hence increasing the bus voltage back to its net desired voltage level. On the other
hand, the SVC will inject less reactive power if node voltage increases, resulting in the
desired output voltage [9]. Because of their large size and high cost, so far SVCs are
predominantly used for voltage control in the transmission systems and very large
industrial loads. Distribution-SVC which is a compact version of full size SVC will be
considered for study in this paper. Since their compact size allows them to be installed
7
anywhere in the feeder they could be another viable option to solve high/low steady state
or transient voltage issues [11], [18]. Their effects on the distribution system with and
without DG penetration have been analyzed and compared in this paper.
8
3. SIMULATION PARAMETERS
3.1. System Description
The distribution feeder selected for this project is an actual 12kV distribution circuit
with several single phase as well as three phase laterals feeding multiple loads. Only a
portion of the feeder up to several end points and DG interconnections have been
considered for the study as illustrated in figures1and 2. The simulation is modeled using a
utility grade computer program known as “CYMEDIST®”. The load allocations are all
set to 99% with the following power factors: Domestic=0.95, Commercial=0.87,
Industrial= 0.85 and Agricultural= 0.85. Figures 1 and 2 provide visual representations of
the parameters described. Major items of interest for voltage study are; a synchronous
generator as a variable DG connected at 30000 feet from substation, a voltage regulator,
and a SVC connected at the distances of 27500 feet and 28800 feet respectively, and a
switched capacitor rated 100k VAr /phase installed at a distance of 18500 feet from the
substation. Other locations of interest are the end of feeder at 36000 feet, and an
automatic reclosing breaker switch located at 21837 feet from the substation. All of these
points are clearly marked in the voltage profile plot in Figure 3.
Synchronous
generator was run in voltage-controlled mod. And, the load flow calculation method used
was ‘voltage drop- unbalanced’ which allowed us to plot voltage profile of the system for
each of the 3 phases independently
9
To
Substation
Point F,
Capacitor at
18500 ft.
Point E,
Switch at
21837 ft.
Point D,
Voltage Regulator at
27500 ft.
Point C,
SVC at
28800 ft.
Point B,
Synchronous Generator at
30000 ft.
Point A,
Feeder End
36000 ft.
Figure 1: Cutout of the feeder, generated by CYME
3.2. Scenarios
In this paper, an actual peak load of 11.5 MW has been considered as full load and
30% of this load will be referred to as light load for the simulations. A synchronous
generator capable of producing a maximum of 10 MW of real power was used as the
variable DG source. 1 MW of generator output was considered as minimum generation,
while 10 MW output is referred to as maximum generation. The generator output was
10
varied from minimum to maximum in the intervals of 0.5 MW and results were recorded
at level of DG output.
Legend
Substation
12KV
Loads
Capacitor
Switch
Voltage
Regulator
SVC
Point F
18500 ft.
Generator
Point E
21837 ft.
Point D
27500 ft.
Point C
28800 ft.
DG
Point A
Feeder End
36000 ft.
Point B
30000 ft.
Figure 2: Simplified diagram of the feeder
11
4. RESULTS
The effects of increasing penetration of DG in the circuit were studied. The feeder
under consideration is evaluated under full load and light load conditions and the results
are being presented in the following sub-sections
4.1. Full Load with No DG
With no DG connected to the system the Power flow will be in one direction which
implies that the voltages will drop, with the increase in distance from the substation and
as the load increases [6]. The voltage drop at any point in the feeder can be calculated by
using following mathematical expressions [7].
VD  s  K  S3
where:
s  effective feeder length
S3  kVA rating of the feeder
K 
(1)
 1  1000 

3

(rcosθ  xsinθ) 
Vr VB
Or equivalently by;
VoltageDrop = K×d× (R× P + X L ×Q)
Where:
P = Realpower
Q = Reactivepower
K = Systemvoltagefactor
d = lengthoflinesection
R = Resistance(ohm)
X L = Indictivereactance
(2)
12
Figure 3 shows the recreated voltage profile of the circuit under consideration at full
load and light load condition without any DG connected to it. This profile will serve as a
base case for comparison studies between different scenarios. Locations for the generator,
capacitor, voltage regulator and SVC along with other points of interest are marked on
the base case graph on Figures1 and 2. These points will be used for comparison study.
From the simulation results shown in table I it can be noted that the voltages on all the
phases are sagging for this loading condition and as the distance is getting farther from
the distribution substation. Towards the end of feeder the voltages tend to go lower, and
they can drop below the lower limit of ANSI standard. The switched capacitor installed at
“point F” is injecting reactive power into the circuit, hence keeping the voltages within
acceptable range. So without DG connection no voltage problem was noted at any of the
locations along the feeder for this peak loading condition.
Light Load Plot
Point F
Capacitor
18500 ft.
Point D
Voltage
Regulator
27500 ft.
Point A
Feeder End
35500 ft.
Voltages (V)
Point B
Generator
30000 ft.
Full Load Plot
Point E
Switch
21837 ft.
Distance from
substation (feet)
Point C, SVC
28800 ft.
Figure 3: Feeder at full and light load with no DG
13
4.1.1. Full Load with Max DG
Next a synchronous generator is connected to the same circuit. Location of this
generator corresponding to the graph is marked as “Point B” in Figure3. Starting form 1
MW, the generators output is being increased to max of 10 MW with 0.5 MW
increments. Voltages at all the points of interest were recorded at each level of DG
output, and the results are shown in table I. Figure 4 shows the regenerated profile of the
circuit under full load and light load condition when DG at maximum of 10 MW
generation level. From table 1 it can be noted that voltages at all points of observation are
gradually increasing as output of the DG increases. During the generation level of 1 MW
through 4 MW, the flow of power is still unidirectional “from substation to end of
feeder”, which implies that the consumed by the load in the vicinity of the generator
power produced by the generator is balancing out the power.
Generator
output
(MW)
1
2
3
4
5
6
7
8
9
10
Table 1
Voltages at different generator output under full load
Voltages (V)
Point A
Point B
Point C
Point D
Point E
Point F
122.2
118.7
120.57
122.47
123.8
125.3
126.6
127.9
129.2
130.4
122.56
119.12
121.07
122.6
124.3
125.7
127.12
128.5
129.5
130.8
122.68
119.12
120.85
122.47
124.13
125.37
126.8
127.9
129
130.3
122.7
119.12
120.9
122.3
123.87
125
126.3
127.6
128.56
129.7
123.7
119.7
120.77
121.85
122.8
123.7
124.46
125.1
125.7
126.4
123.7
119.7
120.76
122
122.7
123.2
123.8
124.1
124.6
124.9
14
133.5 V
Voltages (V)
Light Load Plot
130.8 V
Full Load Plot
Distance from the source (feet)
Figure 4: Voltage profile at 10 MW of DG output, full and light load
When the generation output increases beyond 4 MW, the DG starts to back feed the
circuit. At this point, the generator penetration is higher than the power consumed by the
load, which in turn causes the voltages to increase higher than the power consumed by
the load. This condition will cause the voltages to increase at a higher rate on the line
sections where the power flow has change direction. Even though the voltages keep
rising, they remain within the acceptable range until the generation output level is about
6MW. At about 7 MW of DG output, voltages at points A, B and C rise to the extent that
they begin to violate the upper limit of the allowable voltage band. The Point of Common
Coupling (PCC) between DG and the feeder which is designated by “B” on the Figure 3
is being affected the most. Customers serviced in proximity of such locations will also
experience high voltage at this generation level. Minimum voltage increase was noted on
Points E and F since these locations are farthest from the DG.
At maximum generation
of 10 MW, flow of power is completely reversed, and the voltages in the vicinity of the
generator are much higher than the acceptable limits.
15
4.2. Light Load with No DG
As expected, simulations showed that for the light load condition the voltages also
drop along the feeder, as the distance from the substation increases, similar to the base
case scenario. But the voltage drop in this case was not as much as it was observed during
full load condition. Minimum voltages around point A are in the neighborhood of 123 V
as compared to 117 V at the same location during full load condition. This is an expected
behavior because light load draws less current and hence cause less voltage drop.
4.2.1. Light Load with Max Generation
In this part of study, again, a DG was connected to the system and its output was
varied from 1MW to 10 MW in the intervals of 0.5 MW similar to the full load scenario.
Resulting voltage levels on all the observation points at various DG output levels are
shown in table 2. Again, it was noted from simulations that the voltages on all check
points of the feeder increase as the DG penetration increases. From figure 4 and
comparison of tables 1 and 2 it is obvious that increasing DG penetration has more
adverse effects on the feeder voltage levels during light loading condition as compared to
full loading condition. Voltages quickly increase beyond the acceptable limit on all the
check points. Point B, the location of the generator interconnection and its surrounding
areas are being affected the most. The voltage levels near the PCC are unacceptable even
at the DG output of 6 MW.
16
Generator
output
(MW)
1
2
3
4
5
6
7
8
9
10
Table 2
Voltages at different generator output under light load
Voltages (V)
Point A
Point B
Point C
Point D
Point E
Point F
121.09
122.8
124.4
125.99
127.3
128.6
129.9
131.16
132.3
133.3
121.1
122.9
124.6
126.12
127.4
128.7
130.01
131.29
132.4
133.5
121.1
122.85
124.4
125.8
127.1
128.4
129.63
130.7
131.7
132.7
121.15
122.8
124.2
125.02
126.8
127.8
129.07
130.1
131.1
132.01
121.05
122.7
123.08
123.99
124.7
125.5
126.19
126.8
127.4
127.9
121.1
122.1
122.6
123.27
123.7
124.15
124.7
125.2
125.4
125.78
Figure 5 is the graphical representation of the voltages at point B and Point C at different
level of generator output. From graph is clear that the voltages are rising above 126 volts
of maximum limit
Figure 5: Voltage graph under full and light load, at variable DG output
17
5. POSSIBLE SOLUTIONS
5.1. Addition of Voltage Regulator
Adding a voltage regulator at point D could be a solution to low or high voltages with
or without the presence of DG. In this section, effects of the voltage regulator will be
studied under full load and light load conditions with and without DG. A reproduced
voltage profile in Figure 7 shows the comparison of the voltages under full and light load
conditions with voltage regulator in the absence of DG. In these simulations the voltage
regulator was used without LDC by setting R and X values to zero
5.1.1. Full/ Light Load with No DG
From the Figure 6 it is obvious that the voltage regulator is boosting the declining
voltages beyond the point of its installation under full load condition, bringing them
closer to the set point. In both cases the regulated voltages level on all three phases are
within the acceptable limit.
123.8 V
123.6 V
Voltages (V)
Light Load Plot
Full Load Plot
123.6 V 122.8 V
Distance from the source (feet)
Figure 6: Voltage profile with voltage regulator at full and light load
18
5.1.2. Full Load with DG and Voltage Regulator
With the voltage regulator connected to the feeder, and DG connected at point B, the
output was increased as before and the voltages at all the points of interest were recorded.
The simulations’ results are shown in table 3. The voltage profile for this case is
presented in the Figure 7; for DG at 10MW generation output level. Resulting voltages at
points of observation for the full load scenario with DG and the voltage regulator in
contrast with the case without the regulator could be compared by review of tables 3 and
2.
Table 3
Voltages at different generator output under full load with voltage regulator
Generator
Voltages (V)
Output
Point A Point B Point C
Point D
Point E
Point F
(MW)
1
122.98
123.4
123.4
123.6
123.8
123.8
2
123.2
123.6
123.5
123.7
119.6
120.2
3
123.4
123.6
123.8
123.7
120.73
121.09
4
123.88
124.3
124.2
124
121.9
122.1
5
123.9
124.32
124.1
123.9
122.8
122.7
6
124.4
125
124.5
124.2
123.6
123.2
7
125
125.5
125.1
124.6
124.3
123.7
8
124.7
125.2
124.7
124.1
125.1
124.2
9
125.2
125.6
125.1
124.4
125.8
124.6
10
125.5
126
125.3
124.6
126.01
124.9
Figures 4 and 7 show the voltage profiles at the same level of generation with and
without the voltage regulator. It is clear from the comparison that voltage regulator is
bucking the voltages down to reference voltage level, partly solving the high voltages
problem caused by the DG. It can be concluded that with the installation of a voltage
19
regulator, the DG output could be much higher than the case for the same feeder without
the voltage regulator.
5.1.3.
Light Load with DG and Voltage Regulator
Simulation process was repeated for the same circuit under light load condition as
previous section, the results are shown in the table 4. Light load condition with and
without the voltage regulator can be compared, by review of tables 2 and 4. Figures 4 and
6 show the voltage profile of the circuit at 10 MW of generation with and without voltage
regulator. From the point by point comparison of these tables and figures, it can be
concluded that the installation of voltage regulator partially solves the high voltage
problem caused by DG, by bucking the voltages down to acceptable limits. For example
in the vicinity of point D, the voltages are around 132.02 V without the regulator.
Voltages (V)
Light Load Plot
132.05 V
Full Load
Plot
124.6 V
Distance from the source (feet)
Figure 7: Voltage profile with voltage regulator at full and light load
However with the regulator installed at that point, the regulated voltages are in the
neighborhood of 124.9 V. On the sections between point D and point E still the voltages
20
are out of limit. To solve this problem another layer of voltage regulator could be
considered around point E. However, multiple stages of regulators on a feeder are known
to have certain problems such as hunting or coordination related issues and therefore
generally not recommended.
Table 4
Voltages at different generator output under light load with voltage regulator
Voltages (V)
Generator output
(MW)
Point A Point B
Point C
Point D Point E Point F
1
124.1
124.1
124.1
121.08
121.06
121.1
2
124.4
124.49
124.4
124.3
122.1
121.9
3
124.4
124.6
124.5
124.2
123.13
122.68
4
125.2
125.3
125.1
124.7
124
123.28
5
124.5
124.6
125.6
125.2
124.7
123.7
6
125.6
125.7
125.26
125.5
124.2
124.33
7
126.07
126.2
125.75
125.18
126.23
124.7
8
126.4
127
126.05
125.4
127
125.39
9
125.95
126.09
125.49
124.75
127.7
126
10
126.3
126.6
125.8
124.9
128.1
125.78
5.1.4. Interaction of DG with Voltage Regulator
Simulations’ results seem to clearly indicate that installation of a voltage regulator near
the PCC solves the high voltage problem caused by DG, and low voltage problems in the
absence of DG. However, this solution comes with its own set of problems. One of the
most important problems is voltage flicker which can be defined as a rapid voltage
magnitude variation between two consecutive levels [6], [17]. This sudden change may
be caused by the switching operations of the DG output levels, or an abrupt complete
disconnection of DG at a relatively high output level. If DG is disconnected abruptly the
voltages suddenly drop at the PCC to the values pre DG interconnection level. The low
21
voltage is exacerbated because the voltage regulator at this instant is usually in the
bucking position to lower the high voltages caused by DG prior to disconnection. Since
voltage regulator does not regulate voltage instantly this bucking process continues;
causing a steady state low voltage condition until the regulator has fully responded. This
can take up to several minutes depending on the number of steps needed to move from
the current buck position to the boost position requited by the setting. . The resulting
voltage flicker difference between the two voltage levels may be beyond acceptable
limits. This flicker can be calculated as follows;
Voltage Flicker = voltages at PCC with DG and regulator connected - voltage at PCC
without regulator and DG connected + voltage buck caused by the regulator.
Mathematically it can be described as
Vflicker = VMax - VMin + VReg
Where:
Vflicker = flicker voltage
(3)
VMax = Maximum voltage at PCC
VMin = Voltages without DG, Regulator
VReg = Voltages bucked by Regulator
For example; from Figures 3 and 7 consider point B at 10 MW generation level for full
load scenario, the voltage flicker is found as Vflicker = 126.1- 120.2 + 5 = 10.9 Table 5 below
shows the voltage flicker calculated at different level of generation for full load and light
scenario. Minimum voltages (V Min) for full load are in the neighborhood of120.2 V and
for light load they are around 122.3 V.
22
Table 5
Voltages flicker at PCC for different generator output
DG
Output
(MW)
5
6
7
8
9
10
V(Max)
Full
Load
124.3
125
125.5
125.2
125.6
126.1
Light
Load
124.6
125.7
126.2
126.9
126.09
126
V(Reg)
Full
Load
1
1
1.7
3.2
4
5
V(flicker)
Light
Load
1.5
3.3
4
6
6.5
7.3
Full
Load
5
5.8
7
8.2
9.4
10.9
Light
Load
3.8
6.2
7.9
10.7
10.3
11
IEEE Standard 519-1992 specifies acceptable values of voltage flicker versus
frequency of its occurrence. According to this standard as the frequency of flicker
increases, the max value of flicker magnitude decreases. For example during one hour
period if flicker occurs twice, 5% flicker is acceptable. On the other hand if this
occurrence is 20 times the maximum allowable flicker level decreases to only 1.8%.
Similarly on per minute basis if voltage fluctuation is 2 per minute, the flicker can be
0.5% of the base value. But, if there are 60 fluctuations per minute, then the flicker must
be below 0.12% of the base value.
5.2. Solution by Installing SVC
Because of their high speed response capability to any voltage changes in the system
addition of a SVC at one or multiple locations of the feeder can help maintain a smooth
voltage profile under different network conditions [9] [12]. SVC can also be a very
effective tool to overcome the problem of voltage flicker. Simulations’ results presented
23
below for the various scenarios verified SVC as viable solutions for both steady state and
transient voltage problems.
5.2.1. Full/ Light Load with No DG
From the simulations for both light load and full load scenario, it was observed that the
SVC is helping to maintain a smooth voltage on all phases and all points of observation
in the circuit. SVC is smoothly boosting the voltages in the area where voltages were
declining before, and bucking where they were previously rising.
5.2.2. Full Load with DG and SVC
A composite (light load and full load) voltage profile for the circuit with the SVC and
the DG at its maximum output level of 10 MW output is shown in Figure 8. Table 6
shows the voltages recorded at the different DG output levels for the full load scenario. If
we compare the results shown in table 2 and table 6, we can conclude that installation of
SVC has efficiently solved the high voltage problem caused by high penetration of DG.
With no DG when the voltages tend to go low toward the end of feeder, SVC boosts the
voltages to bring them to desired limit, by injecting reactive power into the feeder. In the
presence of DG, with its increasing penetration level voltages tend to go high, the SVC
bring the voltages down to acceptable limits, by absorbing reactive power from the
circuit.
24
Table 6
Voltages under full load at different generator output with SVC
Generator
Voltages (V)
output
Point A Point B
Point C
Point D
Point E
(MW)
1
124.2
124.6
124.7
124.6
123.8
2
124.4
124.5
124.6
124.6
123.6
3
123.4
123.6
123.8
123.7
120.73
4
124.3
124.6
124.5
124.3
123.4
5
124.4
124.8
124.7
124.5
123.3
6
124.4
124.9
124.7
124.4
123.1
7
124.5
125.04
124.6
124.4
122.9
8
124.7
125.2
124.7
124.1
125.1
9
124.7
125.25
124.7
124.3
122.54
10
124.8
125.3
124.7
124.3
122.28
Point F
123.7
123.4
121.09
123.1
123.05
122.7
122.5
124.2
122.02
121.7
5.2.3. Light Load with DG and SVC
Figure 8 shows the voltage profile of the circuit under light load condition, with SVC
connected and DG at 10 MW of output. Table 7 below shows the voltage on 1 to 10 MW
of DG output. SVC is equally effective in full load and light load conditions. This point
is clear by the voltages shown in tables 2, 7 and the profiles displayed by Figure 8. From
comparison of the table 7 and table 2, it can be concluded that SVC is a viable solution
for the high voltage problem caused by DG. At 10 MW of DG output penetration level
for example, without SVC the voltages at all the points are much higher than the voltages
recorded with the SVC in place. Use of SVC is also a superior solution than voltage
regulator. Comparison of tables 7 and 4 clearly shows that with the voltage regulator, the
maximum DG output must be below 6MW to avoid high voltage conditions. With the
SVC in place on the other hand, the DG output penetration level can be as high as 10
MW without any high voltages. Though in the presence of voltage regulator, high voltage
25
problems on the most of the points of interest were solved; feeder still sees high voltage
problems on the section right before the voltage regulator, as depicted from Figure 7.
SVC controls voltages smoothly as compared to voltage regulator, so the over-voltages
seen on the line sections between pint E and point D do not exist in the feeder with SVC.
126 V
Voltages (V)
Light Load Plot
Full Load
Plot
Distance from the source (feet)
Figure 8: Full and light load profile at 10 MW of generation, with SVC
Voltages (V)
Max Allowed Voltages
Generator Output (MW)
Figure 9: Voltages at Point B with voltage regulator and SVC
26
Table 7
Voltages under light load at different generator output with SVC
Generator
Voltages (V)
output
Point A
Point B
Point C
Point D
Point E
Point F
(MW)
1
124.9
125.04
125
124.9
124.3
123.9
2
125.07
125.1
125.03
124.9
124.2
123.8
3
125.01
125.07
124.9
124.7
123.9
123.57
4
125.1
125.18
124.9
124.8
123.8
123.37
5
125.2
125.27
124.9
123.6
123.1
122.7
6
125.39
125.45
125.07
124.8
123.5
122.9
7
125.4
125.47
125.02
124.8
123.3
122.64
8
125.46
125.55
125
124.7
123
122.33
9
125.63
125.7
125.09
124.8
122.8
122.07
10
126.08
126.1
126.13
125.5
123
122.14
Other advantage of using a SVC over a voltage regulator is that SVC responds almost
instantly to any changes in the circuit voltages as compared to a regulator that takes much
longer. SVC recovers voltages back to the desired levels within a few cycles as compared
to several minutes taken by the voltage regulator. The fast response time of SVC also
reduces the possible problem of voltage flicker. The chart in Figure 9 shows the
comparison of the voltages at point B, in the presence of voltage regulator and SVC at
increasing level of DG penetration.
27
6. CONCLUDING REMARKS
Interconnection of DG on the distribution feeders is causing high voltage problems
under both steady state and transient conditions which are known as voltage flickers.
Proper measures need to be taken into consideration to overcome these issues. As the DG
output penetration level increases, the voltages on various points in the circuit start to
rise. Under light load condition this rise is worse than the full load condition. This rise in
voltage could be so high that it can violate the maximum high voltage limit of 126 V. To
solve this problem a voltage regulator was connected closer to the DG. The LDC
functionality of the regulator was disabled, and its reverse operation mode was also
turned off. It was concluded from the simulations that, though voltage regulator can solve
the high voltage problem on most of the locations, it does not provide a comprehensive
solution. Line sections just before the voltage regulator still face high voltage problem.
So even in the presence of voltage regulator the DG penetration cannot go beyond a
certain level. In addition it amplifies the problem of voltage flickers, due to its nature of
functionality. Additionally increased penetration of DG may also interfere with the
objectives of CVR and its control schemes. Idea behind CVR is to reduce the feeder
voltage in order to reduce losses and conserve energy. Feeder voltages are lowered
primarily by using voltage regulators.
Since increasing DG penetration increases
voltages, it tends to nullify the purpose of CVR. Consequently, the CVR control scheme
may need to be revised. SVC was introduced as an alternative solution. From simulations,
its role was studied under light and full load conditions. It was concluded that SVC
controls the voltages throughout the circuit very smoothly. Without the presence of DG it
28
can boost the voltages and in the presence of DG it can lower the voltage levels down to
the desirable limits in a smooth and continuous manner. With the SVC connected, the DG
output penetration level can be much higher as compared to the output level for the
voltage regulator before any voltage problems become noticeable. Other characteristic of
SVC that makes it a batter solution than voltage regulator is its high speed response to
any change in the system voltage. This implies that it can raise or lower the voltages as
needed within a few power system cycles, hence reducing the magnitudes of possible
voltage flickers. Downside of SVC is its high cost; as compared to voltage regulator. So
there are some tradeoffs when it comes to selection of any one of these solutions. Voltage
regulator offers results which are not so satisfactory but it is cheap, SVC provides a much
better solution at higher costs. There are other solutions such as “re-conductor” to larger
wire sizes or higher voltages, installations of synchronous condensers, or voltage controls
at the DG units. These viable solutions have not considered in this paper.
29
REFERENCES
[1] A.R. Bergen and V. Vittal, Power System Analysis, Prentice Hall, 2000.
[2] “Kyoto Protocol to the United Nations Framework Convention on Climate Change”.
Available at http://unfccc.int/resource/docs/convkp/kpeng.html, last accessed 26
March 2012.
[3] F.A. Wivan, “Voltage Control and Voltage Stability of Power Distribution Systems
in the Presence of Distributed Generation” , Ph.D. Thesis Chalmers University of
Technology Goteborg, Sweden, 2008.
[4] CIRED preliminary report of CIRED Working Group 04, “Dispersed Generation”.
Issued at the CIRED Conference in Nice, June 1999.
[5] U. Naseer, “Impact of Distributed Generation on Electrical Power Network”
available at http://zet10.ipee.pwr.wroc.pl
[6] K. Tran, M. Vaziri, “Effects of Dispersed Generation (DG) on Distribution
Systems”, in IEEE Power Engineering Society General Meeting, June 2005
[7] T. Gönen, “Electric Power Distribution System”, McGraw-Hill Book Company,
1986.
[8] T.A. Short, “Electric Power Distribution Handbook”, CRC Press LLC, 2004.
[9] “Voltage Level Improving by Using Static VAR Compensator (SVC)” available at
http://globaljournals.org/GJRE_Volume11/3-Voltage-Level-Improving-by-UsingStatic-VAR.pdf.
[10] P. Lips, “Semiconductor power devices for use in HVDC and FACTS
controllers,” CIGRE Technical Brochure 112, Paris, France, April 1997.
30
[11] W.K. Wong, D.L Osborn, J.L. McAvoy, “Application of compact static VAR
compensators to distribution systems”, IEEE Transactions on Power Delivery,
Vol. 5, No. 2, April 1990
[12] “ABB Static Var Compensation for Utility”, available at
http://www.abb.com/industries/db0003db004333/c12573e7003305cbc125700400
3c963a.aspx
[13] M.N. Nwohu, “Voltage Stability Improvement using Static Var Compensator in
Power Systems”, federal university of technology, minna, Niger state, Nigeria.
[14] O. O. Stephen, Z. Guanzhou, S. Hui “Voltage Improvement By Integrated
Distribution Static VAR Compensator on the Nigerian Rural
Electric Network”,
in Journal of Economics and Engineering ISSN: 2078-0346, №4, December, 2010
[15] “Voltage Tolerance Boundary”. Available at
http://pge.com/includes/docs/pdfs/mybusiness/customerservice/energystatus/powe
rquality/voltage_tolerance.pdf
[16] M. Vaziri, S. Vadhva, T. Oneal, M. Johnson, “Smart Grid, Distributed Generation,
and Standards” Proceedings of the IEEE PES General Meeting – 978-1-45771002-5/11 Summer 2011, July 24-28 2011. Detroit, Michigan.
[17] M. Vaziri, S. Vadhva, C. Hoffman, K. Yagnic, S. Ghadiri “Standard, Rules, and
Issues for Integration of Renewable Resources” Proceedings of the IEEE PES
General Meeting – 978-1-4244-8357-0/10 Summer 2010, July 25-29 2010.
Minneapolis, Minnesota