A distribution automation study

Material Information

A distribution automation study remote control of feeder capacitor banks at electric utilities
Befus, Craig J
Publication Date:
Physical Description:
ix, 199 leaves : illustrations, map ; 30 cm


Subjects / Keywords:
Capacitor banks ( lcsh )
Capacitor banks ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 196-199).
General Note:
Leaves 163-199 in pocket.
General Note:
Map on folded leaf in pocket.
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Science, Electrical Engineering.
General Note:
Department of Electrical Engineering
Statement of Responsibility:
by Craig J. Befus.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
31250692 ( OCLC )
LD1190.E54 1994m .B43 ( lcc )

Full Text
B.S.E.E., University of Colorado at Denver, 1986
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Electrical Engineering
b y
Craig J. Befus

This thesis for the Master of Science degree by
Craig J. Befus
has been approved for the
Department of
Electrical Engineering
Pankaj K. Sen
William R. Roemish
,9- 9^94

Befus, Craig J. (M.S., Electrical Engineering)
A Distribution Automation Study: Remote Control of Feeder
Capacitor Banks at Electric Utilities
Thesis directed by Professor Pankaj K. Sen
This thesis describes feeder capacitor bank application and
control, and how recent distribution automation technology is
being applied to control conventional switched capacitor banks on
electric utility distribution systems. The fundamental concepts of
distribution capacitor bank application and control are reviewed
in detail. The control of switched capacitors has evolved notably
over the years, and the basic control types are discussed including
conventional electronic controls and newer microprocessor
The significance of distribution automation is discussed
along with the major system components that comprise a
distribution automation system. With this background, the reader
can better understand the various implementation approaches to
remote capacitor control and automation at various U.S. utilities.
An overview will be given on how capacitor control automation
has been implemented at a several major U.S. utilities including

Tampa Electric, Union Electric, Virginia Power and Southern
California Edison. The Public Service Company of Colorado (PSCo)
implementation of capacitor control automation will then be
addressed in greater detail, including the economic benefits,
software development and functionality, capacitor control radio
design, and project management approach. Finally, a case study is
presented to demonstrate how a typical feeder is analyzed and set
up for automated remote control of switched capacitor banks.
This abstract accurately represents the content of the candidate's
thesis. I recommend its publication.
Pankaj K. Sen

I have finally come to see the light at the end of a long
tunnel as this thesis completes my pursuit of a graduate degree in
electrical engineering. It has been an interesting and worthwhile
effort, and I owe many thanks to those who have encouraged and
helped me along the way.
I wish to first thank my graduate advisor, Dr. P. K. Sen, for
all his time and dedication in helping me to complete this thesis.
His guidance from the very beginning of my graduate studies has
been invaluable. I also wish to thank Dr. Bill Roemish for sharing
his knowledge and choosing to teach at the University of Colorado.
Three power systems courses with Dr. Roemish have taught me a
great deal and have made a very positive contribution to my
engineering career. My thanks also goes to Dr. Shelly Goggin for
being a reader and graciously serving on my graduate committee.
I owe thanks to my supervisor, Don Clark, for his
encouragement and support of the work that went into this thesis,
The interesting and challenging work assignments Don has given
me over the years directly relate to the subject matter presented
in this thesis. I also wish to thank my friend and co-worker Ken

Moss for offering to proofread and edit this work. Both the
content and format of this thesis have greatly benefited from
Don's and Ken's input. I must also give thanks to the many people
at PSCo who helped make the capacitor automation project a
reality: Lynn Zakrzewski, John Armenta, Richard Brocaw, Ray
Osterwald, Norm Palm, Dick Lucas, Max Huffman, Rob Webb, Scott
Gomer, Jim Johnson, Brian Laws, Dave Corkill, Jim Fox, Gene Roach,
Bob Belecky, Mark Felling, and Dwight Coppock to name a few.
Lastly, I am truly grateful to my wife, Denise, for being
patient and helping me reach this goal. My household
responsibilities are sure to elevate after this is over.

1. INTRODUCTION.............................................. 1
SYSTEMS ................................................. 6
2.1 Introduction....................................... 6
2.2 Shunt Capacitors................................... 6
2.3 Power Factor Correction........................... 10
2.4 Voltage Correction.............................. 14
2.5 Loss Reduction.................................... 18
2.6 Controls for Switched Capacitors.................. 21
2.6.1 VAR Controls.............................. 2 3
2.6.2 Current Controls ......................... 2 5
2.6.3 Voltage Controls.......................... 2 5
2.6.4 Time and Temperature Controls............. 2 6
2.6.5 Radio Controls............................ 2 7
2.6.6 Combination Controls...................... 2 8
2.6.7 Comparison of Capacitor Controls.......... 2 9
CONTROL............................................... 3 1
3.1 Overview and Background......................... 3 1
3.2 DA Capabilities and System Components........... 3 3
3.3 Supervisory Control and Data Acquisition........ 3 6
3.4 DA Telecommunication Systems.................... 4 4
3.4.1 Radio Frequency Communication............. 4 5
3.4.2 Distribution Line Carrier Communication ... 4 8
3.4.3 Wireline Communication.................... 4 9
3.5 Feeder Capacitor Automation at Other Utilities ... 5 1
3.5.1 Tampa Electric Company.................... 5 2 DA Experiments at TEC............. 5 4 TEC's Energy Control Systems...... 5 5 System Wide Implementation........ 5 7
3.5.2 Union Electric Company.................... 5 8
vii Background of UE System.............. 5 9 Control System Used at UE............ 6 0
3.5.3 Southern California Edison Company........... 6 1 Conservation Voltage Regulation.... 6 3 Integrated Volt/VAR Control.......... 6 4 Estimated Project Benefits........... 6 6
3.5.4 Virginia Power.............................. 6 8
3.5.5 Other Utilities............................. 6 9
3.6 Summary............................................ 7 1
OF COLORADO ....................................... 7 2
4.1 Overview and Background............................ 7 2
4.2 Switched Capacitor Banks........................... 7 3
4.3 Feeder Capacitor Bank Utilization.................. 7 4
4.4 System VAR Requirements ........................... 7 5
4.5 Distribution Automation Committee.................. 7 7
4.6 First Capacitor Automation Pilot Project........... 7 7
4.6.1 Pilot Project Host Computer.................. 7 8
4.6.2 Closed-Loop Control Algorithm............... 7 9
4.6.3 Radio Paging Terminal....................... 8 1
4.6.5 Capacitor Control Radio Receivers........... 8 2
4.6.6 Pilot Project Results....................... 8 3
4.7 Energy Management System Replacement............... 8 4
4.7.1 Substation RTU Upgrades...................... 8 6
4.7.2 New Feeder Capacitor Control Application .. 8 6
4.7.3 CAPCON Functionality........................ 8 7
4.7.4 CAPCON Closed Loop Control Algorithm........ 8 8
4.7.5 Control Operation Modes..................... 9 1
4.7.6 Operator Displays........................... 9 2
4.7.7 CAPCON Application Interfaces............... 9 2
4.7.8 CAPCON Verification Scheme.................. 9 3
4.8 New Capacitor Control Radio Equipment.............. 9 8
4.8.1 New Capacitor Control Radio Design........... 9 9
4.8.2 Paging Protocols Evaluated................... 105
4.9 Economic Benefits .................................. 106
4.10 Future Enhancements................................. 108
4.11 CAPCON Operational Results.......................... 109

4.12 Project Management Approach..................... 113
4.12.1 Project Definition and Planning.......... 114
4.12.2 Procurement.............................. 116
4.12.3 Implementation........................... 119
FEEDER......................................... 122
5.1 Objectives and Assumptions ..................... 122
5.2 Radial Distribution Analysis Package............ 125
5.2.1 Radial Power Flow Analysis................ 126
5.2.2 Impedance and Admittance Calculations. . 129
5.3 Modeling Feeder 1974 ........................... 133
5.4 Collecting and Entering Feeder Load Data........ 136
5.5 RDAP Output..................................... 139
5.6 CAPCON Database Parameters...................... 144
5.7 Some Rules of Thumb............................. 146
6. CONCLUSION............................................ 150
A. Sample Displays....................................... 154
B. Map of Feeder 1974.................................... 161
C RDAP Output........................................... 163
REFERENCES............................................... 196

Capacitors have been used extensively on utility power
distribution systems since the early 1940s. Capacitors are
installed on distribution feeders as pole-top banks on overhead
line sections or as padmount banks on underground line sections.
A power capacitor is a system of conductors and dielectrics
arranged so that a large electrical charge can be stored. Early
power capacitors were constructed of pure aluminum foil sheets
separated by layers of oil impregnated kraft paper.
Improvements in dielectric materials have allowed capacitor sizes
and efficiency to increase significantly over the last 40 years,
resulting in a lower cost per kVAR to electric utilities. Common
dielectric materials used today in the electric utility industry
include paper, polyester, acetate and polypropylene films. The
primary function of capacitors in power systems is to regulate
voltage and reactive power flow. [1], [2]
A capacitor can be thought of as a VAR generator that
supplies the magnetizing requirements of inductive loads. The

generation of reactive power at a power plant and its transmission
to a distant load is possible, but not economically optimal.
However, reactive power can be economically provided by
capacitors located close to the load centers. On power systems,
capacitors are installed primarily in shunt at various locations
including substations, distribution feeders, within industrial plant
electrical systems, and directly at motors. For the electric utility
industry as a whole, approximately 60 percent of capacitors are
applied to feeders, 30 percent to substation busses, and the
remaining 10 percent to transmission systems [3]. As illustrated
in Figure 1.1, individual capacitor units are connected in parallel
for VAR rating and then in series for voltage rating. For economic
reasons, capacitor banks are generally applied at or near rated
voltage. [1], [2]
Figure 1.1: Connection of capacitor units [1].

The internal conductors and dielectric provide the capacitance and
voltage withstand capability to obtain the desired per phase kVAR
rating of the capacitor unit:
kVAR = V22jifC/1000 (1-1)
V = capacitor voltage rating, V
f = system frequency, Hz
C = capacitance, microfarads.
Shunt capacitors affect the power factor of the load (as seen
by the source), whereas the series capacitors cancel the inductive
reactance of the circuit where they are applied. Series capacitors
have been used to a very limited extent on distribution systems
because special problems associated with their application and
complex engineering analysis required. Hence, further discussion
of series capacitors is excluded from this thesis. Shunt capacitors
can be installed as fixed banks or as switched banks. Switched
capacitor banks can be turned on as the reactive load becomes
greater during a daily load cycle or turned off during light-load
conditions. The methods of controlling switched capacitors have
evolved from more basic local control shemes to the use of more
sophisticated automated remote control schemes. Supervisory
control and data acquisition (SCADA) and advances in distribution

automation (DA) have revolutionized the methods for controlling
switched capacitor banks on utility distribution systems.
Distribution automation refers to the ability to remotely
monitor, coordinate and operate distribution components in a
real-time mode from remote locations [4]. Distribution automation
provides a tool to more effectively manage the continuous,
minute-by-minute operation of a distribution system. Maximum
utilization of a utility's physical plant and higher quality of service
to it's customers is achievable through distribution automation
technology available today. Like any automation process,
automating the distribution system involves collection of data,
analysis of information, decision making, and then verification of
the desired results.
Distribution automation functions are a set of discrete, yet
interconnected, system processes. Each function is discrete since
it involves a clearly definable objective, although one function
might depend on another. Automated remote control systems for
feeder capacitor bank control have been fairly straight forward
and relatively easy for utilities to implement, as compared to
other functions like switch automation and remote meter reading.
This can be attributed to the fact that a major change out of
equipment is not required to automate and remotely control

switched capacitor banks. The equipment required for
implementing automated remote feeder capacitor control includes:
microprocessor controls with radio interface,
line sensors (if not already installed),
radio transmitter and remote radio devices,
computer and software. [4]
The purpose of this thesis will be to lay the foundation for
capacitor bank control on distribution feeders in general, and then
detail how distribution automation can improve the level of
control and utilization of existing capacitor banks. An overview of
how capacitor control automation systems have been
implemented at a few major U.S. utilities including Tampa Electric,
Union Electric, Virginia Power and Southern California Edison will
be discussed. The Public Service Company of Colorado (PSCo)
implementation of capacitor control automation, including the
economic benefits, software development, capacitor control radio
design, and project management approach will also be addressed.
Finally, a case study will be presented to demonstrate how a
typical PSCo feeder is analyzed and set up for automated remote
control of switched capacitor banks.

2.1 Introduction
This chapter covers the fundamental aspects of capacitor
bank application and control of switched capacitor banks. The
theory of capacitors presented in this chapter are well known, but
are reviewed here to lay the foundation for later discussion of
how distribution automation is now being applied to feeder
capacitor switching. The control of switched capacitor banks has
evolved notably over the years, and the basic control types are
discussed toward the end of this chapter.
2.2 Shunt Capacitors
A capacitor connected in shunt is in parallel with the load.
The function of a shunt capacitor is to supply lagging reactive

power to the system at the point where it is connected. Capacitors
located as close to the reactive load as practical reduce the path of
reactive power flow and thus releases thermal capacity of lines
and equipment for handling real power. A shunt capacitor has the
same effect as an overexcited synchronous condenser, generator,
or motor. [ 1 ]
Shunt capacitors are used extensively on power distribution
systems to release thermal capacity, reduce kVAR generation
requirements, reduce losses, and regulate or improve voltage.
Shunt capacitors were first used for power factor correction
around 1914. Practically all capacitors used prior to 1940 were
installed indoors within industrial plants. Outdoor pole-top units
were later developed and extensive utility use on primary
distribution systems continues to the present. [3]
A fixed capacitor bank raises the voltage all along the
inductive circuit from the source to the capacitor bank to the load.
A switched capacitor bank can be turned on or off when needed
according to local or system conditions. The further out on a
feeder a capacitor bank is installed the greater the voltage rise it
will provide when energized. Figure 2.1 is a picture of a typical
switched capacitor bank pole-mounted unit.

The major benefits of using shunt capacitor banks on
distribution feeders are summarized as follows:
1. Reduce reactive load current flow along a feeder, and
therefore reduce power losses.
2. Reduce kVA demand, thus releasing system capacity.
3. Improve voltage regulation.
4. Increase revenue or decrease customer energy
consumption. [5]
Most electrical equipment requires two components of
current to operate. The reactive current is the component
required to produce the magnetic flux necessary to operate
induction devices. Without reactive current, energy could not

flow through a transformer core or across the air gap of an
induction motor. The power-producing or real current is the
component which is converted by the equipment into useful work,
usually in the form of heat, light, or mechanical power. The two
components of power loss, or I2R losses, on a distribution feeder
are caused by both the real current and reactive current
components. Losses due to real current are not significantly
changed by application of capacitors. Losses due to reactive
current are reduced by the application of capacitors and are
considered to be avoidable losses.
By reducing the reactive load current supplied by the
substation, the total current is reduced and thus the kVA demand
on substation transformers and feeder circuits is reduced. Less
total current results in less voltage drop and lower losses. Feeder
voltage raised by capacitor banks and/or substation transformer
taps will increase customer energy consumption since load, and
therefore energy consumption, is proportional to the voltage
squared. Lower losses and increased consumption mean more
revenue to the utility. Conversely, feeder voltage can be reduced
to decrease customer energy consumption by switching capacitor
banks off and/or by lowering substation transformer taps. [6], [7]

2.3 Power Factor Correction
Power factor correction on primary distribution systems is
often required due to loads with low power factor. Low power
factor can be attributed to several factors and equipment types
such as partially loaded induction motors, transformers,
fluorescent lighting, rectifiers, variable speed drives, air
conditioning systems, and electronic equipment. The most
common method for improving power factor on distribution
systems is to apply shunt capacitors along the primary feeders.
The meaning of power factor is better understood after
describing the current, voltage, and power relationships in a
power system as follows. As stated previously, the total current
in an a-c circuit consists of the reactive and real current
components. Real current is the component that is exactly in-
phase with the voltage waveform on a given phase. Reactive
current is the component of current that is 90 out-of-phase with
the voltage waveform. Both current and voltage are measured on
a per-phase basis in a three phase system.
In typical distribution loads, the load current lags the
voltage as shown in Figure 2.2a. The cosine of the angle between
current and voltage measured at a common point is known as the

power factor. The power factor is also defined as the ratio of real
power to apparent power.
Real current
I R = I COS 0
P, kW
Figure 2.2: (a) Phasor diagram of current and voltage and (b)
power triangle for a typical distribution load [8].
Figure 2.2b shows the triangular relationship between real power
(kW), apparent power (kVA), and reactive power (kVAR). Power
is defined as voltage times current. The reactive power unit of
measurement is the volt-amp-reactive, or VAR, and is computed
by multiplying the reactive current component by the voltage.
The unit of measurement for real power is the watt, or W, and is
computed by multiplying the real current component by the
voltage. The total current multiplied by voltage is known as
apparent power, or VA.
The power factor may be leading or lagging, depending on
the relative direction of both the real and reactive power flows. If
both the real and reactive power components flow in the same

direction, then the power factor is lagging. If the power
components flow in opposite direction of each other, then the
power factor is leading. Stated differently, if the total current lags
the voltage then the power factor is lagging. Conversely, if the
total current leads the voltage then the power factor is leading.
Looking at the power factor of a particular load, for example,
would show that an induction motor has a lagging power factor
since its reactive power requirement has to be supplied
externally. In other words, the induction motor requires both real
and reactive power to flow into the motor (same direction). On
the other hand, an overexcited synchronous motor can supply its
own reactive power requirements (from the motor d-c field
action) which may result in a leading power factor depending on
the d-c excitation current.
For all practical purposes, the only way to improve power
factor on a feeder is to reduce the reactive power component. To
illustrate how a capacitor at a load can reduce the source kVA and
kVAR, Figure 2.3 shows the power factor correction for a given
1 2


Figure 2.3: Illustration of power-factor correction [1].
As Figure 2.3 shows, the capacitor draws leading reactive power
from the source; stated differently, the capacitor supplies lagging
reactive power to the load. Assume that a load with real power P,
lagging reactive power Q and apparent power 5^ at a lagging
power factor of
With a shunt capacitor of Qc kVAR installed at the load, the load
power factor is improved to
cos h = -
1 3

and the apparent power and reactive power are decreased from 5^
kVA to S2 kVA and from Ql kVAR to Q2 kVAR. This results in a
reduction of total current, which in turn reduces the I2R power
losses in the circuit. Shunt capacitors do not significantly affect
the total load current or power factor beyond their point of
application. [8]
2.4 Voltage Correction
Since shunt capacitors raise a feeder's voltage, they can be
used to supplement or replace feeder voltage regulators, or
substation transformer load tap changers (LTC). However,
switched capacitors are not usually used for fine voltage control
on distribution feeders. When voltage is kept to the rated value
at the load, better load performance (e.g., of motors and
fluorescent lighting) can be expected. The most important voltage
benefit of capacitors installed on a primary feeder is the reduction
of the voltage spread between the first and last customer.
ANSI Standard C84.1-1989 establishes the nominal voltage
ratings and operating tolerances for 60-Hz electric power systems
ranging from above 100 V through 230 kV. The purpose of this
standard is, in part, to promote standardization of nominal system

voltages and equipment voltage ratings and tolerances. The
standard defines the acceptable system service voltagethe
voltage at the point where the electrical systems of the supplier
and user are connectedas ranging from 114 V to 126 V for a
two-wire 120 V nominal system. [9]
Figure 2Aa-d shows the single phase diagram and the
respective voltage-phasor diagram before and after a shunt
capacitor is switched on. This figure shows the difference
between the sending and receiving end voltage before and after a
capacitor is switched on, with the assumption that the sending end
voltage is held constant. The difference between the sending and
receiving end voltages is the total voltage drop, or rise, depending
on whether the receiving voltage is lower, or higher, than the
sending end voltage.

Z = R + jXL
Z = R+jXL
(C) (d)
Figure 2.4: Voltage-phasor diagrams for a feeder with lagging
power factor: (a) and (c) shunt capacitor switched off and (b) and
(d) with shunt capacitor switched on. [1], [10]
The voltage drop along a feeder (or short transmission line)
with lagging power factor load can be approximated as

V-V + 'A v <24>
voltage drop per phase, V
total resistance of feeder, ohms per phase
total inductive reactance of feeder, ohms per
real component of current, A
reactive component (lagging) of current drawn by
the load, A
1 6

The real and reactive components of current in Equation 2-4 are
derived from the relationships
IR= I cos 0
Ix = I sin 0
I = total phasor current magnitude, A
0 = power factor angle, degrees.
When the capacitor at the receiving end of the line in Figure 2.4 is
switched on, the voltage drop is now approximated as
Ic = the reactive component of leading current drawn by the
capacitor in Amps.
Since capacitors reduce the reactive current supplied by the
source they reduce the voltage drop by an amount equal to the
capacitor current multiplied by the line reactance. The
approximate voltage rise is the difference between Equations 2-4
and 2-6 and is expressed as
1 7

2.5 Loss Reduction
One of the more important economic benefits of capacitors
installed on primary feeders is loss reduction. Capacitors are
effective in reducing the portion of losses due to the reactive load
current. The power loss caused by a current / flowing through a
resistance R is equal to I2R. If the current has a lagging power
factor angle 0, then I2R = Ir2R + I2R without capacitors, or I2R =
1r2R + (IX-1C)2R with capacitors. The losses due to the real
component of current are not appreciably changed by the
application of capacitors. Since substation capacitors cannot
reduce feeder losses, it is important to locate power capacitors on
the feeders as close to the loads as practical.
Figure 2.5 shows the reduction in source current to /' with
the addition of a capacitor.
Figure 2.5: Current is reduced upon applying a capacitor.

The reduction in current results in a reduction in I2R loss in a
feeder due to adding capacitors. The loss reduction (LR) formula
is derived as follows.
LR = I2R (I')2R
= [(/H2) 2)J*
= (21 where
LR = loss reduction per phase, W.
Equation 2-8 shows that the avoidable losses depend on the
reactive component of the load current, but not on the real
Another way to determine loss reduction by the addition of
capacitors is to note that system conductor losses are proportional
to the square of the current. Since current is reduced in direct
proportion to power factor improvement, the losses are inversely
proportional to the square of the power factor. This leads to
another derivation of loss reduction as follows.
, p 1 I = (2-9)
T ( Original PF Y Losses o' (2-10)
^ Improved PF J

Percent LR = 100 x
f Original PF ^
v Improved PF,
Neagle and Sampson [11] have shown that the maximum
loss reduction on a feeder with distributed load (e.g., a uniformly
or tapered distributed load) is obtained by locating a capacitor
bank at the point on the feeder where its kVAR rating is equal to
twice the peak load kVAR. This principle also holds true when
more than one capacitor bank is applied to the feeder. For
example, a 1,200 kVAR capacitor bank should be located at the
point on the feeder where the peak kVAR flow is 600 for
maximum loss reduction. The exact location of the capacitor bank
can deviate by as much as 10% of the total feeder length from the
point of maximum loss reduction without appreciably affecting
the loss benefits. In practice, the greatest loss reduction and
voltage benefits are obtained by locating the capacitor bank just
beyond the optimum loss reduction location. [11]
The annual energy losses are reduced as a result of
decreasing I2R losses by the installation of capacitors. The energy
saved equals the average power loss reduction multiplied by the
amount of time the capacitor bank is on the line (assuming an
efficient switching method), and the cost savings equals the

energy saved multiplied by the value of avoided energy [5]. This
can be expressed as
Annual Savings = kWSAvED x h/yr x 0/kWh, $ (2-12)
To give a very simple example, assume that a capacitor bank is
on-line 1,500 hours per year, the average loss reduction is 80.5
kW, and the energy is worth 50 per kWh. The annual savings is
then = 80.5 kW x 1,500 h/yr x 50/kWh = $6,038 per year.
2.6 Controls for Switched Capacitors
Feeder kVAR loading can vary widely during the daily load
cycle from day to day and from season to season. This makes it
necessary to switch capacitors to meet, but usually not exceed, the
VAR requirements of the load. Also, voltage along a feeder can be
varied by varying the amount of connected shunt capacitor banks.
Capacitor controls and sensing equipment provide a means of
determining the need for capacitance and adding or removing
capacitance as required to maintain optimum performance of the
system. [12]
Switched capacitor bank controls receive intelligence from
an external source and initiate the switching of the capacitors into

or out of the circuit. The components of the control circuit are the
receiver of the intelligence, relay coils to operate oil or vacuum
insulated mechanical switches, and the control power source. The
functional elements of a capacitor control system are shown in
Figure 2.6. In its simplest form, the control circuitry consists of a
sensor, sensing circuit, and relay contacts which are actuated to
operate a capacitor switch on each phase. [12]
120 V
Figure 2.6: Capacitor control system [12].
All capacitor controls are essentially SPDT (single pole, double
throw) switches activated by one or more conditions which reflect
VAR or voltage requirements. Theoretically, any intelligence
which responds to a change in VAR supply required by the load
can be utilized for automatic switching of capacitors. [13], [14]
The types of capacitor bank controls that are available today
include the following:

(1) Temperature control
(2) Time clock control
(3) Voltage sensing control
(4) Current sensing control
(5) VAR sensing control
(6) Radio control
(7) Combination control (two or more of the above types
combined into one unit)
Each of the above control types available today can be separated
into two categories: (a) conventional electronic controls employing
discrete logic circuitry, and (b) electronic microprocessor-based
controls. Electronic microprocessor-based controls are relatively
new and provide superior functionality and flexibility over
conventional electronic controls. Capacitor controls can also be
classified one of three ways: as a local control device that
operates autonomously according to local conditions (e.g., current,
voltage, VARs, temperature, etc.); as a remote control device that
operates via a remote signal; or as a combination of both a local
and remote control.
2.6.1 VAR Controls
A VAR control responds to reactive current and will switch
the capacitor bank on-line at a setting that corresponds to some

point in the reactive load cycle. Current and voltage sensors or
instrument transformers are needed to provide input signals to
the VAR control unit. Sensors may be installed as three phase
devices or on a single phase, even though capacitor banks are
generally three phase. Measurement of VARs must be made in
the circuit on the load side of where the capacitors are connected
and requires a voltage phase reference. With capacitor banks
installed on primary feeders, the cost of sensors may be a major
A logical application of a VAR control is where a large load is
connected to a feeder and a capacitor bank is located near the load
to supply the reactive power of that specific load. The VAR
control also finds use on switched capacitor banks at the source
end of feeders. These capacitor banks will maintain the power
factor near the substation end of the feeder under all loading
conditions. Another common application of a VAR control is to
switch a capacitor bank in a substation where instrument
transformers are readily available. Reactive current at any point
along a feeder is affected by downstream capacitor banks, hence
VAR controls are highly susceptible to interaction from
downstream banks.

2.6.2 Current Controls
Line current is a common quantity used for sensing in
feeder capacitor bank control. Line current is an indication of
reactive current only if the load power factor is known and is not
erratic. Line current does not change significantly when a
capacitor bank is switched on, and therefore current controls are
only moderately sensitive to interaction from downstream
capacitor banks. Also, since feeder voltage drop is directly
proportional to the line current, current controls provide a fairly
good combination of both loss reduction and voltage control. The
current control application requires a current transformer to
isolate the control from the primary feeder. The current
measuring element must be connected on the line side of the
capacitor bank to avoid the vector addition of capacitor current
with load current.
2.6.3 Voltage Controls
Voltage is also a very common quantity used for sensing in
feeder capacitor bank control. Voltage controls are the logical
choice when maintaining voltage profile along the feeder is the

primary purpose of the capacitor system. Voltage controlled
capacitor banks have substantially reduced the use of regulators.
Voltage controls are used at a point where the circuit
voltage decreases as the load increases, generally where the
voltage spreadthe difference between maximum and minimum
voltagesis more than 2 V (on a 120 V reference base). Hence,
the voltage control is useful when the capacitor bank is located a
significant distance out on a feeder. Voltage controlled capacitor
banks are highly susceptible to interaction from both upstream
and downstream banks and voltage regulators. Since a voltage
transformer is needed in any switched capacitor to provide power
for the control circuit and the oil or vacuum switches, a voltage
signal is readily available for voltage control.
2.6.4 Time and Temperature Controls
Both time and temperature controls are immune from
interaction between multiple capacitor banks. However, since
they do not respond to any line parameter, the effectiveness of
their settings depends upon the degree of correlation with the
actual parameter of interest such as voltage or reactive power. If

the correlation is weak, capacitor bank application can become
inefficient or even problematic.
Temperature controls, which respond to ambient
temperature, find application on feeders where there is large
heating or air conditioning load in residential and commercial
areas. Time controls are used where the utility can predict the
load curve based on time of day. Both these types of controls are
used mostly in residential and commercial areas. Settings on
conventional time or temperature controls cannot readily
anticipate unusual conditions or lighter feeder loading that may
occur on holidays or weekends. Microprocessor based time or
temperature controls can be programmed to adapt for these
2.6.5 Radio Controls
Radio controls have been in wide use since the early 1980's
and their popularity is growing as advances in distribution
automation technology make their application more practical.
Radio capacitor controls can be used to achieve effective loss
reduction and maintain a flat voltage profile, depending on the
sophistication of the control system employed. Most radio

controls do not sense local line parameters and require an
external data acquisition system which is tied to various
monitoring points on a distribution system.
A supervisory control and data acquisition (SCADA) system
is commonly utilized in conjunction with a centralized capacitor
control system to control switched capacitors on feeders via a
radio transmitter. Since SCADA systems acquire and process data
in real-time, radio controlled capacitors are well suited to
variations in the load pattern. The investment in this type of
system can be large, but the economic benefits of radio controlled
capacitors can far outweigh the investment costs.
2.6.6 Combination Controls
Combination controls respond to two or more composite
signals to determine when to switch a capacitor bank. The
objective of using a combination control is to permit more than
one signal to connect the capacitor bank as a function of local and
system requirements. State-of-the-art combination controls are
microprocessor based and offer a high degree of flexibility and
programmability. Microprocessor based capacitor controls can be

programmed to respond to multiple quantities including voltage,
current, temperature, time of day, VARs, and radio signals.
A two-way radio interface to microprocessor based capacitor
controls is particularly useful when the utility needs to remotely
monitor the status or override local control of capacitor banks.
Two-way radio communication to combination controls can also
permit remote re-programming and diagnostic testing. These
types of controls can eliminate routine maintenance since
improper operation of capacitors can be detected and reported
immediately to the central computer that is monitoring the
system. This is sometimes termed as "just-in-time" maintenance.
2.6.7 Comparison of Capacitor Controls
The controls available today for switching capacitor banks
include VAR, current, voltage, time, temperature, radio and
combination controls. Table 2.1 summarizes the attributes of the
various capacitor control types. Utilities throughout the United
States are replacing older local controls at switched capacitor
banks with new microprocessor based controls that have remote
monitoring and control capabilities when tied to a distribution
automation system. Depending on the system control scheme and

capacitor control features, line sensors may be wired as inputs the
control, or sensors may be installed at the substation and tied into
a SCADA system. Since electric utilities typically have extensive
communication infrastructures in place, it may be possible to use
existing radio transmitters for remote control and monitoring. A
separate communication system can be cost justified if several
functions on the distribution system are automated concurrently.
Control Type Current Signal Required Loss Reduction Efficiency Voltage Profile Improvement Interaction Between Banks
VAR Yes Highest Moderate High
Current Yes Moderate High Moderate
Voltage No Moderate Highest High
Temperature No Problematic Problematic None
Time No Problematic Problematic None
Power Factor Yes Problematic Problematic High
Radio SCADA needed High High None
Combination Yes High High None to moderate
Table 2.1: Summary of Capacitor Control Attributes [5].

3.1 Overview and Background
This chapter provides a broader definition of distribution
automation and the major system components that comprise a
distribution automation system. With this background the reader
will better understand the various implementation approaches to
remote capacitor control at various U. S. utilities described later in
this chapter. There are several complex and challenging
information management issues that will become more apparent
regarding long term implementation, integration, and interaction
of distribution automation functions. Particular attention will be
given to automatic remote feeder capacitor control as the
significance of distribution automation is expanded in this chapter.
Distribution automation (DA) refers to the ability of an
electric utility to remotely monitor, coordinate and operate

distribution components in a real-time mode from remote
locations. This definition is broad and includes several systems
often discussed separately such as supervisory control and data
acquisition (SCADA), feeder and substation automation, load
management, and automatic meter reading (AMR). With current
technology distribution automation is a combination of automatic
(closed-loop) and manual (open-loop) monitoring and control
functions. Manual monitoring and control is accomplished by
interacting with the distribution automation system through a
user interface (UI). The UI allows utility personnel to remotely
monitor and control points on the distribution system. [4]
A variety of distribution feeder equipment can be remotely
monitored and controlled with available hardware and software.
The equipment that can be remotely monitored and/or controlled
includes load break switches, switched capacitor banks, line
reclosers, voltage regulators, transformers, circuit breakers,
automatic throw-overs (ATOs), sectionalizers, voltage and current
sensors, fault indicators, smart meters, and customer loads [15].
Electronic and electromechanical control devices (e.g., capacitor
controls, recloser controls, and ATO controls) have local
intelligence so that they can operate autonomously in addition to
the ability to respond to remote signals. Automatic or manual

remote monitoring and control of the aforementioned equipment
is typically done from one or more regional computer systems.
3.2 DA Capabilities and System Components [4]
The three basic capabilities of distribution automation are
monitoring, control, and protection of the distribution system.
Monitoring is the ability to determine the state of the distribution
system, as indicated by significant information retrieved remotely
including: status (contacts closed/open, etc.), analog values
(voltage, current, temperature, etc.), and pulse accumulator values
(energy consumption, etc.). Control is the ability to alter the state
of the distribution system for the purposes of efficiency, plant
utilization, outage restoration, and quality of service. Protection is
the ability to detect, identify, locate and isolate distribution
system faults or equipment malfunction.
Adequate performance in executing the various distribution
automation functions is dependent on the computer facilities and
field equipment provided. The computer facilities may consist of
one central computer or several distributed computers. The
computers may be located in one central facility or in regional
control centers in a system where computers communicate to one

another over a wide-area, high speed data communications
The distribution automation computer system software
generally consists of an operating system, a database management
system, various software applications to model and control the
distribution system, graphical user interface (GUI) software, and
software to acquire data from and control field equipment over a
wide area communications system. The software may be
categorized as real-time (for time critical functions), interactive
(for human intervention), batch (for routine or scheduled events),
and expert systems (i.e., artificial intelligence to reduce the
possibility of human error). Relational and object oriented
database management systems are being applied to distribution
automation functions as computer speed and processing power
has dramatically improved to support these powerful and popular
software tools.
The field devices used in a distribution automation system
include remote terminal units (RTUs), programmable controllers,
intelligent electronic devices (IEDs), metering equipment, current
and voltage sensors, communications equipment, motor operated
switches, transducers, relays, batteries, etc. Remote terminal
units, programmable controllers, or IEDs may operate distribution

components autonomously in response to local conditions, or in
response to a remote signal that might be predicated on system
conditions. Remote terminal units are the more commonly used
data collection and control devices in a distribution automation
system today, but will become obsolete as specialized IEDs (that
incorporate RTU functionality) become more economical.
Computers, software, workstation terminals, communications
equipment, and electronic field devices are all needed to carry out
the many distribution automation functions. The three leading
feeder automation functions being implemented throughout the
electric utility industry in the U. S. are remote feeder switching,
recloser automation and remote capacitor control [16]. Due to
limitations with available software, remote feeder switching is
primarily implemented as open-loop control requiring human
intervention before control is executed. On the other hand, feeder
capacitor switching is typically implemented as closed-loop
control with minimal human intervention required. In contrast to
feeder switching, equipment or system malfunctions with remote
capacitor control is generally not detrimental to energy supply
The major benefits of distribution automation include
improved reliability, control, and customer service and lower

operating and maintenance costs. More specifically, the main
objectives stated by various electric utilities for implementing
automated remote capacitor control include the following:
1. Conservation voltage regulation to reduce customer
energy consumption
2. Flatter voltage profile and thus better power quality
3. Improved VAR management and the ability to bring
feeder power factors closer to unity
4. Fast emergency response to transmission level voltage
and stability problems
5. Greater distribution system efficiency
6. Improved equipment utilization and reliability
particularly during peak load conditions
7. Just-in-time maintenance
8. Immediate notification of equipment malfunction
9. Better coordination of multiple capacitor bank
10. Remote programming of capacitor control parameters
11. More flexible control of capacitor banks
These objectives are further discussed later in this chapter where
several electric utility capacitor automation projects are described.
3.3 Supervisory Control and Data Acquisition [17]
One of the principal applications of a distribution automation
system today is supervisory control and data acquisition. A

SCADA system provides the ability to exercise control over a
device that is usually remote, and to confirm its performance in
accordance with the directed action. Simply put, SCADA is a
system for performing remote monitoring and control from a
central location. Traditional utility SCADA systems have
databases and software that were originally designed for bulk
power transmission operation and generation dispatch. More
recently, SCADA systems have been developed specifically for
supporting distribution system monitoring and control.
The functions that a SCADA system performs include data
acquisition and supervisory (manual) control of remote devices,
alarm processing, data processing and conversion, data storage
and archival, error detection, system diagnostics, user interface to
display information, and report generation. Remote terminal units
or IEDs installed in the field monitor local conditions including
analog and digital parameters. Digital parameters are obtained by
monitoring external contacts of circuit breakers, protective
devices, relays, meters, etc. Since electronic devices, such as an
RTU, can handle only very small voltage and current signal inputs,
analog parameters are obtained through the use of transducers or
sensors. For example, transducers convert current and voltage
transformer secondary signals to milliamp level signals which an
RTU can then convert to digital values to be transmitted to the

host computer system. Remote terminal units also carry out
control by means of a digital or analog signal. Control is
accomplished by energizing an interposing relay or outputting a
milliamp signal to operate an external device such as a circuit
breaker or load-tap changer (LTC). Most control functions that are
initiated by human operators require a select-before-operate
(SBO) control sequence. The SBO feature is essentially an end-to-
end system security check where the host computer sends a point
selection message, receives a checkback message from the RTU,
then sends the operate message, and finally receives a control
verification message from the RTU.
Most of SCADA control is open-loop, which implies that
control has to be initiated by a person rather than automatically.
Automatic closed-loop control, for functions such as capacitor
control or generation control, requires additional software
applications that are typically interfaced to the SCADA system
database. The SCADA system performs most, if not all, of the
electric system data acquisition required of these other software
applications. Consequently, this is why SCADA is considered to be
one of the principal software applications of a distribution
automation system today.

In the 1970's, the term "energy management system" (EMS)
was coined to describe a more comprehensive computer control
system that included SCADA, automatic generation control (AGC),
and other network analysis and advanced applications for bulk
power system operations. The EMS software usually includes
extensive and computer intensive application programs for the UI,
dynamic mapboard control, AGC, economic dispatch of generation,
contingency analysis, security enhancement, transmission load
flow analysis, and data exchange programs for inter-utility
transaction analysis.
The number of RTUs in an EMS can amount to several
hundred, where the RTUs are located in transmission and
distribution substations, as well as utility interchange and
switching stations. Since an EMS is generally considered to
comprise the main control center for an electric utility, many
utilities began initial distribution automation development on
their EMS. In some electric utilities, distribution operations are
closely aligned with transmission operations so the EMS was
expanded to provide limited functionality to distribution
operators. In these cases, substation-oriented SCADA functions
were provided where it was cost effective to do so.

A more recent development is the concept of a distribution
management system (DMS) as a complement to the EMS. The
intent of a DMS is to employ all of the applicable EMS capabilities
and offer new functions in a manner that is better suited to
distribution system operations. With advancements in technology
and development of individual distribution related computer
systems such as load management; automated mapping, facilities
management, and geographic information systems (AM/FM/GIS);
feeder automation; and automatic meter reading (AMR), utilities
are finding that they need to migrate toward a DMS to coordinate
these proliferating systems that all support distribution operation
and engineering functions. [18]
A DMS and EMS are similar in that both: 1) collect and
coordinate information from RTUs, IEDs, and meters, 2) provide a
graphical user interface to present information, 3) contain
analytical functions to interpret or predict system events, 4) store
information for later analysis and reporting of events, and 5) are
typically connected to other computer systems to exchange
pertinent data. However, there are some fundamental differences
between distribution and transmission operations making EMS
technology inappropriate for any extensive distribution

The key differences between a DMS and EMS stem from the
following facts: 1) distribution systems are typically radial,
whereas transmission systems are network connected, 2)
distribution system devices are dispersed along the length of
feeders, whereas transmission system devices are generally
located only at substations, 3) the number of locations requiring
RTUs and/or IEDs for remote control in a distribution system is at
least an order of magnitude greater than the locations in the
associated transmission system, 4) the resulting distribution
database is also an order of magnitude larger than that of the
transmission database, 5) a distribution system configuration is
much more dynamic and exposed to hazards, whereas topology
changes rarely occur in transmission systems, and 6) distribution
system operators require circuit diagrams that are geographically
(map) oriented as opposed to the one-line schematics that bulk
power system operators utilize.
The DMS functions can be grouped into the following
1. Substation and feeder SCADA
2. Substation automation
3. Feeder automation (e.g., capacitor automation, switch
automation, etc.)
4. Automatic meter reading (AMR)

5. Distribution system analysis routines (e.g., power flow,
voltage profile, load forecasting, cold load pickup,
system loss, trouble call and outage restoration)
6. Interfaces to other computer systems (e.g., EMS,
AM/FM/GIS system, customer information system,
AMR system, load management system, etc.)
7. Distribution automation wide-area communications
The technologies of DA wide-area communications deserve special
attention and are treated in the next section. Further description
of other DMS functions are beyond the scope of this thesis, but can
be found in [4] and [18].
While the DMS concept is emerging, however, it is not yet a
fully developed commercial reality for extensive distribution
automation purposes. This is due to several factors including the
persisting high cost of DA technology, expensive field equipment
installation, technology limitations (mostly software related),
substantial capital investment requirements, lack of industry
standards for data access and multiple vendor support, relatively
few competitive and regulatory incentives, and utility corporate
cultures that make for very slow progress. The latter issue of
utility corporate cultures is most significant because
implementation of distribution automation crosses almost every
organizational boundary in a utility.

Distribution Customers

f Energy Management System Distribution Management System

Computer and
Figure 3.1: The utility system with distribution automation [4].
Figure 3.1 graphically depicts the elaborate interrelationship
of a DMS with other corporate computing systems and functional
organizations within an electric utility. These interfaces can be far
more complicated for a combined gas and electric utility since the
DMS and communication infrastructure may have to serve both
electric and gas operational requirements.
4 3

3.4 DA Telecommunication Systems
Any extensive automation of an electric distribution system
requires the use of a sophisticated and specialized wide-area
communication network to transmit data and control signals
between the host computer(s) and the significantly large number
of remote devices. A communications network for distribution
automation is considerably different from that serving bulk power
generation and transmission. Conventional EMS SCADA
communication technologies that use poll-response data transfer
techniques between the host computer and substation RTUs are
not appropriate for DA. Communication technology is the key
element in the development of DA [16]. It cannot be
overemphasized that the communications system selected
ultimately dictates the functionality of DA.
In general, a wide-area communication system to support
multiple distribution automation functions should have the
following characteristics:
Reliability of data communications and equipment
Cost effectiveness
Sufficient capacity to meet present and future data
Two-way communications capability where needed
Ability to support real-time SCADA functions

Ability to communicate into power outages or faulted
areas of the distribution system
Ease of installation, operation, and maintenance
Ability to interface a variety of devices from different
manufacturers by employing industry standards
and/or non-proprietary interfaces
Intelligence of the communications network to be able
to route information to different computer systems
and facilities, provide remote diagnostics and fault
tolerance, establish peer-to-peer communications,
error recovery, etc.
It is likely that no single communication system technology will
conform to every one of these characteristics. Furthermore, a
wide range of media alternatives are available where the
advantages and disadvantages of each must be considered by the
utility. A hybrid communication system utilizing two or more of
the telecommunication options, to be discussed, may be needed to
meet all of the requirements of a particular utility. [4]
3.4.1 Radio Frequency Communication
The use of radio frequencies (RF) for the control and
operation of power utility systems has steadily increased over the
last 40 years [17]. Radio has also proven itself to be a viable
communications medium for many DA functions. Radio can

provide broad coverage over a geographic area and can be
implemented as a one-way and/or two-way system. With battery
backup power, radio can be used to communicate into areas with
faulted distribution lines. Radio systems commonly employed
today specifically for SCADA and DA include the frequency bands
shown in Table 3.1.
FCC Allocated Frequency Bands for Utilities Electromagnetic Spectrum Designation Propagation Characteristics Typical Utility Uses Related to SCADA and DA
154 MHz 150-170 MHz Very high frequency (VHF) band (30-300 MHz) Nearly line-of-sight, scattering due to temperature inversions, cosmic noise, multipath One-way radio paging systems, load control Two-way land/mobile radio, SCADA applications
450-470 MHz 860-902 MHz Ultra high frequency (UHF) band (300 MHz to 3 GHz) Line-of-sight, cosmic noise, atmospheric absorption, multipath Single channel two-way, full duplex voice/data communications Mobile radio systems used for fixed and mobile data applications
902-928 MHz Unlicensed spread spectrum and packet radio, DA applications, automatic meter reading
928-952 MHz Multiple address systems (MAS) and private cellular radio for point-to-multipoint SCADA/DA applications
>1 GHz SCADA and protective relaving applications
3-30 GHz Super high frequency (SHF) Line-of-sight, atmospheric attenuation Satellite communications for mostly geographically isolated SCADA applications
Table 3.1: Frequency bands used for SCADA and DA [17].

Radio systems have several advantages that include the
Isolation from power line faults as compared to power
line carrier or telephone cables attached to the same
Smaller initial investment as compared to trenching
and cabling costs of wireline
Long term savings over ongoing leased telephone
Ease of relocation of remote sites
Operation and maintenance are under the control of
Higher reliability and expansion capability over
wireline or power line carrier
Utilities generally have existing radio systems and
transmitter sites that can be utilized for DA
There are also some disadvantages of radio systems that should
be considered:
Spectrum congestion and difficult FCC licensing process
Unexpected interference, multipath, and path
Security against jamming or spoofing
Cost of installing new transmitter or repeater sites
Expensive and complicated remote radio site
installation if directional antennas are required

Despite these disadvantages, radio is the most widely used
communication medium for DA today because of its lower cost and
increased flexibility as compared to other media. Significant
research and development is being conducted by several vendors
to develop radio technologies specifically for DA purposes.
Many utilities that have implemented automated remote
capacitor control have found that one-way VHF radio is a very
cost effective telecommunication system that meets most of their
requirements. Other utilities have implemented two-way radio
systems for multiple DA applications including capacitor
automation. Two-way radio is definitely the system of choice if it
can be economically justified.
3.4.2 Distribution Line Carrier Communication
Power line carrier (PLC), first introduced in the 1920's, has
been applied as a communication medium on transmission lines
and more recently on distribution lines. Like PLC, distribution
line carrier (DLC) utilizes a carrier frequency to transmit data over
existing distribution lines. A few utilities have demonstrated DLC
as a suitable communications method for DA applications.
Typically, DLC uses frequencies from 2 to 20 kHz where data is

encoded on the carrier using a modulation technique such as
amplitude modulation (AM), frequency modulation (FM),
frequency shift keying (FSK), single side band (SSB), zero crossing,
or ripple control. For transmission line applications, PLC uses a
carrier frequency between 50 and 500 kHz. [17]
Unlike transmission lines, distribution lines are electrically
more complex due to the existence of numerous junctions,
transformers, and shunt capacitors. DLC offers two-way
communication capability but at slow data rates (less than 300
baud). DLC is unreliable for communications when power lines
are damaged. Power system harmonics can cause interference
with PLC and more significantly so with DLC because of the lower
carrier frequencies used. Hence, harmonics can further reduce
data transmission rates. Consequently, utilities have looked
favorably upon PLC use on transmission lines, but less so on
primary distribution lines. FCC licensing is not required for either
PLC or DLC operation. [4], [17]
3.4.3 Wireline Communication
Wireline media for distribution automation includes public
telephone networks, utility-owned metallic cable pairs, cable

television (CATV), and fiber optic cable. The integrity of a
wireline system is dependent on the host structures such as poles
and conduits. To date, wireline media has been typically
employed for distribution automation only in special applications
in limited areas of a utility's service territory. For example, data
communication between large commercial and industrial
customers and the utility can be economically accomplished using
leased telephone lines. However, most possibilities of wireline
medium are too expensive for utility telemetry applications at
residential customer sites.
Cable TV has been used to a far lesser degree than leased
telephone lines for telemetering or telecontrol because CATV
networks are, in general, not yet capable two-way communication.
Changes now occurring in the CATV industry is likely to make
cable TV a more attractive option in the future for utilities to
implement residential customer DA functions like automatic meter
reading, remote disconnect, meter tamper detection, load profile
data acquisition, and time-of-use rate structures.
Fiber optic cable has been used for distribution automation,
for example, in downtown underground networks, airports, and
commercial and industrial areas. It has been economically
installed in existing feeder cable ducts, under-built with

distribution or transmission lines, or imbedded in neutral or
shield wires. The relatively high cost of splicing, repeaters,
transceivers, special connectors, and installation of fiber optic
systems have prevented it from being used more extensively for
distribution automation applications.
3.5 Feeder Capacitor Automation at Other Utilities
The remainder of this chapter considers DA as applied
specifically to remote feeder capacitor control at several U.S.
electric utilities. Throughout the electric utility industry emphasis
is being placed on gaining maximum utilization of existing
distribution system facilities. This is fueled by increasing
regulatory pressures, public opposition to building new power
plants and power lines (both distribution and transmission), and
less available capital for replacing aging power system
components. Electric demand continues to grow while utility
budgets are shrinking. Distribution automation technology
advances are a response to utilities needs for improving efficiency
and equipment utilization in a cost effective manner.
The extent of information obtained on capacitor automation
projects varied between utilities depending on availability of

published materials and tightening policies with regard to sharing
proprietary information. The information gathering process was
very time consuming requiring several phone call interviews,
personal interviews, attending industry conferences, travel to
make site visits in person, and collecting related articles and
papers that have been published. Specific costs, quantifiable
benefits and results of the utility capacitor automation projects
presented were generally considered classified information which
was not made available.
3.5.1 Tampa Electric Company [19]
Tampa Electric Company (TEC) is an investor owned electric
utility in west-central Florida that serves a population of 1 million
customers covering 2,000 square miles. Tampa Electric has
actively pursued DA to increase system reliability, defer new
generating plants, and reduce operating costs. Capacitor control
was one of the first areas of DA that the utility began
investigating starting in 1983.
Tampa Electric's distribution system currently has over 500
fixed capacitor banks and around 600 switched capacitor banks
installed. Prior to having remote control capability, mechanical

time controls were used to switch feeder capacitors. The time
controls had to be set by operating engineers in the various
districts of TEC's service territory. During the winter months,
around 170 of all capacitor banks had to be removed from service
due to seasonal light loading conditions. This was accomplished
by sending crews to disconnect capacitor banks from the system
by pulling fuses connected to the primary.
Several other problems relating to the conventional fixed
and time switched capacitor banks at TEC led them to implement
DA. Capacitor banks became very maintenance intensive and the
mechanical time controls were both expensive and unreliable.
Every quarter, routine inspections were made of all capacitor
bank installations which revealed common problems such as
blown fuses, damaged lightning arrestors, malfunctioning time
controls, failed capacitor units, and failed oil switches. Rapid load
growth in their service territory required the installation of more
capacitor banks, thus increasing routine maintenance. An
alternative to the mechanical time controls was needed to reduce
routine maintenance and increase reliability of switched capacitor
53 DA Experiments at TEC
In 1983, the first DA pilot project was launched to remotely
control switched capacitors on the TEC 13 kV distribution system.
The communications system attempted first was distribution line
carrier (DLC). The use of DLC at TEC proved to be impractical due
to the dynamic configuration of their distribution feeders, and the
reliability of the DLC system. However, it was still apparent that
remote control of feeder capacitor banks was important enough to
pursue further in order to improve system power factor.
In 1990, a second pilot project was initiated to remotely
control feeder capacitor banks. The second pilot project was
conducted using two existing computer systems which included
TECs load management system (LMS) and EMS. The goal of this
pilot project was to first demonstrate the feasibility of operating
capacitor banks with a modified version of the load management
VHF radio receivers and the LMS and EMS host computers.
Initially, the system remotely controlled capacitor banks
according to time of day, which was similar to the load
management strategy to control customer loads.
54 TEC's Energy Control Systems
The LMS, used primarily to curtail residential and
commercial loads (air conditioners, electric heat, water heaters,
pumps, etc.) during system peaks, controlled distribution
capacitors using VHF radio and ran on Hewlett Packard 1000
computers. The LMS was implemented in 1985 and was supplied
by RELM Communications. The LMS was used only during peak
load periods and was idle approximately 90 percent of the time.
The EMS, implemented in 1989, is a Control Data Corporation
computer system. The EMS gathers data from approximately 200
RTUs at substations. The RTUs monitor the MVARs, MWs, voltage,
and current of substation transformers and 13 kV feeders. The
telemetered VAR data on the EMS is used to monitor the operation
of the capacitor banks when the LMS issues radio signals to
capacitor controls. The change in VARs detected by the EMS
verified proper operation of capacitor banks as well as indicated
when capacitors did not function properly. The pilot system used
for automated capacitor control at TEC is shown in Figure 3.2.

VHF Radio
Energy Management Load Management
System-performs System-encodes
SCADA functions capacitor addresses







Figure 3.2: Automated capacitor control pilot system at TEC.
The experiment showed that capacitor operation was
reliable and capacitor bank malfunction could be detected by the
EMS by observing the change in VARs. However, TEC found that
the load management receivers modified for capacitor control
were not of an entirely appropriate design for this application.
Hence, before a system wide implementation, TEC developed
specifications for a new capacitor control radio receiver (CCR). The
new CCRs were designed to operate on a VHF frequency of
173.20375 MHz using multi-tone frequency shift keying (FSK)
modulation techniques [20].
56 System Wide Implementation
To implement remote control of all switched capacitors at
TEC, it was also decided to develop the control algorithms to run
on a personal computer (PC) linked to the Control Data EMS.
Software developed for an IBM 386 PC runs the capacitor control
algorithm shown in Figure 3.3.
Test leading circuits
(most negative VARs)
> '
Test lagging circuits (most positive VARs)
Select capacitors to
"OPEN" and write
addresses to record
Select capacitors to
"CLOSE" and write
addresses to record

Figure 3.3: Capacitor control algorithm at TEC [19].

The system was enhanced to control capacitors based on feeder
VARs and power factor rather than time of day. Engineers at TEC
determined that this would give them more judicious control of
distribution capacitors and the ability to maintain distribution
system power factor around 0.98 lagging.
Development and testing of TEC's software was completed in
1992. As of November 1993, TEC had installed over 150 CCRs in
the field. Some enhancements to the system were identified and
implemented as part of the full scale effort. These include control
of capacitors based on changing circuit VARs monitored at the
substation, modifying CCR to incorporate separate trip/close
relays, and further developing the calculation routines on the IBM
PC for optimum VAR control.
3.5.2 Union Electric Company [21]
Union Electric Company (UE) of St. Louis Missouri
implemented an automated capacitor control system back in 1987.
They did this by using their existing SCADA system, adding a
capacitor control master station computer, and using two FM radio
transmitters. In their automated capacitor control scheme the
SCADA system monitors over 200 distribution substations to
gather feeder amps on each phase, bus voltage, and MW and

MVAR on each transformer every 15 seconds. Feeder capacitor
banks are remotely controlled using feeder amps, time or
temperature data. Background of UE System
The automated capacitor control system at UE was cost
justified on the basis of system loss reduction and elimination of
routine capacitor bank maintenance. Conventional capacitor
control methods using local controls did not make the most
effective use of capacitor banks. Time clock controls switched too
many capacitor banks on during weekends and holidays.
Temperature controls sometimes overcompensated, since their
settings assumed air conditioning loads, for example, would come
on at set ambient temperatures. Current controls allowed for
better control of capacitors by measuring load current, but were
relatively expensive and needed frequent field adjustments for
load growth and feeder reconfiguration. Using local controls, UE
found at least 5% of their capacitor banks were off-line during
peak load periods.
59 Control System Used at UE
The first generation automated capacitor control system
used at UE included a PC based master station with a special radio
communications interface. The capacitor control logic developed
by UE is depicted in Figure 3.4.
To Be
To Be
Avg Amps
Avg Amps
> Amps
j No
| No
* (Time
on 2 hrs)
on -10)
| NO
Figure 3.4: Capacitor ON sequence logic in master station [21].
The SCADA system downloads feeder amperes to the capacitor
control master station every 20 minutes. Ambient temperature
readings from four substations are also downloaded and used as
backup control parameters to the switching scheme. With these
control parameters plus time of day and date, the computer

determines if a capacitor bank should be switched on or off. The
database is organized by substation and feeder. Each capacitor
record in the database has summer and winter control settings
based on current, time and temperature. If feeder amperes are
not available, due to some problem with the SCADA system, then
the system defaults to time or temperature control as a backup
scheme. Temperature is also used to bias the current control of
Each capacitor control radio receiver is individually
addressable. The receivers operate on a 154.46375 MHz carrier
frequency. The radio receiver replaced former local controls,
which were plugged into a four-jaw meter socket. Each radio
receiver has a 10 A, 120 V latching relay to switch the capacitor
bank, an internal antenna, and an operations counter.
3.5.3 Southern California Edison Company [22], [23]
Southern California Edison Company (SCE) is the second
largest utility in California serving over 10 million customers. SCE
is known as a pace setter utility in distribution automation
throughout the world and their automated capacitor control
system implementation is by far the most aggressive of any

utility. The Electric Distribution Division of SCE began a program
in 1992 to improve system operations and electrical efficiency.
The program, Distribution System Efficiency Enhancement
Program (DSEEP), has five principal projects: feeder switch
automation, feeder lock-out alarming, substation automation,
outage management, and distribution capacitor automation project
(D-CAP). The largest and most intensive of the distribution
automation projects being implemented at SCE is D-CAP.
Southern California Edison developed and is implementing a
capacitor control system that fine tunes customer voltage levels to
reduce energy consumption. The system uses two-way, spread
spectrum packet radio technology along with solid state electronic
meters that provide real-time customer voltage and energy
consumption. The radio system transmits customer meter voltage
and capacitor status at all telemetered locations to a computer
system. The computer system runs a very sophisticated control
algorithm to determine which capacitors should be turned on or
off. The objective for their capacitor control system is to reduce
overall net energy transfer from the substation to the customer
and still meet system VAR requirements. As of November 1993,
SCE had tested the distribution capacitor automation system on 66
feeder capacitors located on 18 feeders, and 3 substation capacitor

banks. Efforts are now underway to roll-out the D-CAP system to
include over 7,600 switched capacitor banks by the year 1996. Conservation Voltage Regulation
In the late 1970's formal studies sponsored by the Electric
Power Research Institute (EPRI) were conducted to determine the
effect of voltage reduction on energy consumption. The EPRI
Project 1419-1 set out to quantify energy consumption of
electrical load components as a function of operating voltage, and
to develop a computer program model to predict energy
consumption as a function of operating voltage levels [24]. The
California Public Utilities Commissions (CPUC) Energy Division
staff requested that investor owned utilities in California quantify
the annual energy conservation potential of voltage reduction.
Southern California Edison's own studies reported in 1978
supported an overall energy savings of 1% for each 1% reduction
in voltage on feeders having a mixture of customer-class loads
(i.e., residential, commercial, industrial, agricultural, and other
power authorities).
The CPUC and other California utilities determined that
voltage reduction was technically feasible and that a significant

annual energy savings could be achieved with the implementation
of Conservation Voltage Regulation (CVR). However, it was not
until the late 1980's that SCE was able to acquire the
communications and metering technologies to enable them to
implement CVR on a large scale. CVR is now the main driver
behind the distribution capacitor automation project at SCE.
Again, the objective of their D-CAP effort is to reduce the average
service voltage to customers through feeder capacitor switching
and LTC control, thus "obtaining demand side energy conservation
by implementing a supply side solution." Southern California
Edison has shown that voltage reduction can be achieved while
improving system VAR support. Integrated Volt/VAR Control
The CVR demonstration program implemented at SCE is
intended to reduce the ANSI standard C84.1-1989 service voltage
bandwidth from 120 V (5%) down to 120 V (-5%). The key to
their successful capacitor automation demonstration was to
monitor customer voltage at key points and measure substation
VARs as primary control parameters. Their D-CAP closed-loop
algorithm is designed to minimize line losses and maintain VAR

flows within optimum limits while keeping service voltage near
117 V or a minimum of 114 V at the customer's meter panel.
The D-CAP system controls and maintains customer voltage
along each segment of the feeder as close as possible to 114 V.
Solid state electronic revenue meters that provide real-time
voltage readings are strategically placed to provide a statistically
consistent sample of lowest voltage customers. The central
computer collects customer voltage data and capacitor status
through a 902-928 MHz spread spectrum packet radio network.
Control commands to capacitor controllers are issued from the
computer and require no operator involvement. Microprocessor-
based controls interfaced to the packet radios receive control
commands and report back any change in status.
Substation VARs are minimized by using control bandwidth
set points in the central computer database. The capacitor control
algorithm chooses the capacitor combination that minimizes
customer voltage and keeps substation VARs within set point
limits. In some cases where there are LTC transformers at
substations, the control algorithm calculates optimal bus voltage to
maintain minimum customer voltage and unity power factor. The
computer then controls the LTC according to calculated

The results of a demonstration project conducted at two
major substations in Orange County in January of 1993 showed
that the 15 minute voltage profiles with and without the
automation system differ by 4.2 V (120 V scale) on average.
Table 3.2 shows monthly results of CVR since January of 1993.
Month '93 Voltage Reduction Results Volts Percentage
Jan 4.50 3.70
Feb 2.10 1.80
Mar 2.90 2.30
Apr 4.80 4.00
May 4.40 3.60
Jun 4.50 3.70
Jul 5.30 4.30
Aug 4.70 3.80
Sep 4.60 3.70
Average 4.20 3.43
Table 3.2: SCE's D-CAP Voltage Reduction Results for 1993 [23]. Estimated Project Benefits
Three years of CVR testing at 13 substations provided an
estimate for system-wide energy savings within SCE's service
territory. The tests also indicated that the energy reduction
achieved by CVR varies by customer class. Table 3.3 shows the
CVR energy savings by customer class.

Customer Class Energy Savings for Every 1% Voltage Reduction
Residential 1.30%
Commercial 1.20%
Industrial/Agricultural 0.50%
Other Power Authority 1.20%
Average 1.05%
Table 3.3: SCE's Customer Class CVR Energy Savings [23].
The conclusion made from the tests is that system-wide energy
savings would be approximately 1% for every 1% voltage
reduction. By the year 2000 SCE estimates that over 1,000 GWH
in energy can be saved annually which translates to $44 million
per year, mostly in fuel cost savings, by implementing D-CAP
system wide.
Two-way communications to capacitor controllers allows SCE
to perform remote diagnostics. This capability has resulted in
significant operation and maintenance savings since routine
capacitor checks could be eliminated. With D-CAP, repair crews
are dispatched only when capacitor banks are found to be
inoperable through remote testing. Capacitor availability has also
increased since failures are quickly identifiable. SCE has found
that with D-CAP, many capacitor banks on feeders can be
removed or relocated. Improved utilization has allowed SCE to
defer purchases of new feeder capacitor banks.

The D-CAP system has also helped SCE to identify
overloaded transformers by keeping customer voltages in a
historical database for later analysis. Prior to D-CAP, SCE had to
rely on customer complaints or routine maintenance to discover
voltage problems. Other real-time information now available
include load power factor, status of switched capacitors (on or off-
line), and primary voltage at switched capacitor banks. Bulk
power operators have remote control of all switched capacitor
banks during system emergencies such as voltage collapse or
3.5.4 Virginia Power [25], [26]
Virginia Power serves over 1.7 million electric customers
within a two state area including Virginia and North Carolina.
Virginia Power's annual peak load is more than 12,000 MW. Their
service territory is divided up into five divisions, each responsible
for operation of the distribution system within the respective
divisions. During the 1980's, Virginia Power's SCADA system
evolved into an integrated component of both their transmission
and distribution systems. As their SCADA system matured, the
need arose to extend monitoring and control to the distribution

feeders. The result of this effort included developing an
automated capacitor control system.
Virginia Power has replaced switched capacitor controls
with radio controls. The communications path is a 153.485 MHz
radio paging system, where control commands are issued through
each division's SCADA master computer. The SCADA computers
are interfaced to the paging system, and signals are received by
remote receivers installed at each switched capacitor bank. Since
1987, a global Volt/VAR computer program called CAPCON has
been in service in each division's SCADA master computer.
CAPCON issues commands to remotely switch capacitor banks to
maintain unity power factor on the feeders while keeping voltage
within acceptable limits. Communications is one-way, so no direct
feedback of capacitor switch status is available. This was
determined to be acceptable given the non-critical nature of
switching relatively small rated capacitor banks.
3.5.5 Other Utilities
Other major U. S. utilities that are remotely controlling
feeder capacitor banks include Commonwealth Edison, Northern
States Power, Pacific Gas & Electric, Boston Edison, Baltimore Gas &

Electric, Public Service Gas & Electric, and Georgia Power. Table
3.4 summarizes the computer systems and communication
methods used by these utilities for remote capacitor control and
monitoring; this information was obtained primarily through
telephone interviews and on-site visits.
Electric Utility Computer System Parameters Used in Control Algorithm Communication Scheme Used
Commonwealth Edison (Chicago, IL) PC based interfaced to EMS Remote control based on total electric system load thresholds. VHF radio (1 way)
Northern States Power (Minneapolis, MN) EMS computer Remote control based on VAR flow at substation transformers, and time of day. Radio controls have local voltage override. 900 MHz radio (1 way)
Pacific Gas & Electric (Northern California) Regional distribution SCADA systems Local control based on time, temperature, and voltage with remote monitoring. Remote control for system emergency conditions only. 950 MHz MAS radio and 902-928 MHz spread spectrum packet radio (both types are 2 way)
Baltimore Gas & Electric PC controller at each substation Substation VAR flow and bus voltage VHF radio (1 way)
Boston Edison PC based Remote control based on time of day VHF radio (1 way)
Public Service Gas & Electric (New Jersey) DA master station Local control based on time, temperature, and voltage with remote monitoring. 902-928 MHz spread spectrum packet radio (2 way)
Georgia Power (Atlanta, GA) DA master station Local control based on VARs with voltage bias. Remote diagnostics from PC program. 950 MHz MAS radio and 902-928 MHz spread spectrum packet radio (both types are 2 way)
Table 3.4: Comparison of remote feeder capacitor control
systems at other major U. S. utilities.

3.6 Summary
Automated remote capacitor control for improved voltage
and VAR control in distribution systems has become economical
and relatively easy to implement as demonstrated by numerous
utilities across the U.S. Conventional local controls at switched
capacitors have proven to be more problematic and less optimum
than centralized control through distribution automation.
Capacitor control using conventional local controls does not
completely consider demand diversity, changes to the feeder
connectivity, remote variables, and coordination with other
capacitors. Conventional control methods also do not readily
detect and report equipment failures to utility personnel. Since
the complexity of voltage and VAR control in modern distribution
systems fundamentally exceeds the capability of local controls or
operator control, closed-loop remote control and/or monitoring is
becoming the preferred capacitor switching method for many
7 1

4.1 Overview and Background
Public Service Company of Colorado (PSCo) serves the
electric needs of the Denver Metropolitan area as well as other
areas in the State of Colorado. In the Denver area, PSCo has 50
distribution substations with over 300 feeders. The primary
distribution voltage is 13.2 kV in the Denver area. Excluding the
downtown network system, PSCo feeders are radially configured
and range from one to twenty miles in length. PSCo has over 800
shunt capacitor banks installed on overhead feeder segments in
the Denver metropolitan area totaling to around 1,000 MVAR of
compensation installed. The capacitor bank sizes used are 600,
1,200, and 1,800 kVAR, where 1,200 kVAR banks are the most
common. Typically, 2 to 6 capacitor banks are installed on an
overhead feeder. [27]

4.2 Switched Capacitor Banks
Of the 800 capacitor banks installed on PSCo's system in the
Denver area, approximately 600 of them are switched for a total
of 720 MVAR; the remaining banks are "fixed." Switched
capacitor banks are controlled by automatic local control or
remotely by one-way radio control. Local controls that sense
voltage, current, temperature, or VARs locally are set to
automatically switch capacitor banks on and off at certain input
values. Local controls operate switched capacitor banks
independently and there is no direct coordination of operation
between capacitor banks. Typically, voltage controls were applied
toward the end of a feeder, where there is the largest voltage
change from a switched capacitor, and current controls were
applied toward the substation end. Where one-way radio controls
are used, MVAR and MW are measured at the substation-end of
each feeder to determine when capacitors should be switched on
or off and in what order.
Historically, visual inspections of capacitor banks were
conducted every summer in the Denver metropolitan area. The
purpose of these inspections was to determine the operational
status of each capacitor bank, get a counter reading from each
control, and look for any obvious problems (open fused-cutout,

bad oil switch, etc.). The status and counter information were
entered into a computer program to generate a report that
compared current year to previous year counter readings of the
switched capacitor banks to calculate the number of close-
operations of each capacitor bank.
While the annual capacitor report was a way to track
capacitor bank operation, it sometimes gave misleading
information to distribution planning engineers. A classic example
of this is with voltage controlled capacitor banks. Often times
voltage controlled banks would close, thus incrementing the
counter, but due to high voltage would immediately open back up.
This situation led distribution planning engineers to install more
capacitor banks on feeders to improve feeder power factors
during system peak, not realizing that the voltage controlled
capacitor banks were not being properly utilized since the counter
readings indicated otherwise.
4.3 Feeder Capacitor Bank Utilization
Based on a five year review of annual feeder capacitor
reports conducted in 1991, PSCo found that the availability of
switched capacitors at the time of system peak varied from 70 to

85 percent each year in the Denver area. Attempts to improve
the availability by recalibrating the control settings proved to be
somewhat ineffective, partly due to the poor reliability of the
capacitor controls themselves and partly due to voltage controlled
capacitor banks. Frequent recalibrating of controls was very
costly and time consuming since it required a two-person line
crew and, often times, a distribution engineer to test and change
the control settings in the field. The procedure to recalibrate
controls involved determining the present control settings using
test equipment, recalibrating the on/off settings as necessary,
setting the time delays, and measuring the actual voltage rise
caused by switching the capacitor bank on. The labor time
involved per capacitor bank to recalibrate a local control is
approximately 3 work hours including travel time. Although the
overhead distribution system had a sufficient number of capacitor
banks installed, PSCo was not able to achieve enough power factor
correction from a system standpoint using the local controls,
particularly during system peak loads.
4.4 System VAR Requirements
A VAR task force was established at PSCo a decade ago to
recommend a long term strategy for locating shunt capacitors on

the company's electric transmission and distribution systems. In
the early 1980's Denver experienced significant population
growth, and it was necessary to ensure on a system wide basis
that there were adequate reactive power resources to meet
growing load requirements and maintain adequate power factor
and voltage on both the electric distribution and transmission
systems. The committee's study and report in 1985 indicated that
345 MVAR of additional capacitors were needed on the electric
system to achieve unity power factor during the summer peaks
[28]. Early studies also determined that the most economical
locations to install capacitors were on 115 kV transmission busses
and on overhead distribution feeders. The challenge to the
Electric Distribution Division was to maintain unity power factor
on the overhead feeders at the time of system peaks. It was not
until 1987 that automated remote control of switched capacitor
banks appeared to be the most feasible and cost effective way to
improve capacitor bank availability during system peaks and
maintain unity power factor at substations that had mostly
overhead distribution feeders. Feeders with all underground
conductors typically do not have power factor correction installed
due to the high cost of padmount switched capacitors.

4.5 Distribution Automation Committee
In 1987, executive management at Public Service
established the Distribution Automation Committee with the
mission of identifying and recommending gas and electric
distribution automation opportunities which would reduce cost
and/or provide better service reliability to customers.
The committee prepared a report [29] for executive
management that recommended pursuing several distribution
automation functions. The committee's recommendations were
predicated on the assumption that "no automation option would
be installed on a large scale basis until it had successfully
completed a small scale trial installation." From a list of eight
automation functions recommended, management selected
automated remote capacitor control as one of the first options to
further evaluate by conducting a pilot project. [29]
4.6 First Capacitor Automation Pilot Project
In 1988, a small scale pilot project was conducted to
demonstrate the concept and feasibility of automated feeder
capacitor control through a trial on 4 feeders out of East
substation in Aurora, Colorado. The pilot project period ran for

just over a year. As shown in Figure 4.1, the system comprised a
PC based host, 17 radio capacitor controls, a remote terminal unit
(RTU) installed at the substation, and the corporate paging
Transmitter xl '
Figure 4.1: Diagram of system used for the feeder capacitor
automation pilot project at PSCo.
4.6.1 Pilot Project Host Computer
A vendor supplied PC-based capacitor control system was
selected to host the pilot project. The system consisted of partial

SCADA and closed-loop capacitor control proprietary software
programs. The SCADA program polled the RTU to acquire feeder
Watt/VAR data at East substation once every minute. The PC host
communicated to the RTU using a 928/952 MHz multiple address
radio system. The Watt/VAR data acquired was used by the
capacitor control software algorithm to determine when to switch
capacitor banks OPEN or CLOSED on each of the 4 feeders. If
feeder MVAR measured at the substation were outside of the
valid operating limits defined in the database (e.g., -0.4 MVAR
leading to 1.0 MVAR lagging could be valid operating limits for a
feeder), then the closed-loop capacitor control program would
send a command to the paging terminal. The paging system
would then transmit a page to turn a capacitor bank on or off,
depending on the function executed. Since radio communication
from the paging system was one-way, proper operation of a
capacitor bank was determined by monitoring the change in
feeder VARs after a control signal was issued.
4.6.2 Closed-Loop Control Algorithm
To give an example of how the closed-loop capacitor control
algorithm worked, suppose there is one 1,200 kVAR capacitor
bank on a feeder. Assume that the RTU reports in real-time that

the feeder load has 1.8 MVAR lagging. If the operating limits
were defined as 1.4 MVAR lagging and 0.2 MVAR leading, the
capacitor control algorithm would send a signal to turn on the
1,200 kVAR bank. Once the capacitor bank is energized, an
approximate 1.2 MVAR change in the feeder load data would be
reported by the RTU, giving a reactive power flow out of the
substation of around 1.8 1.2 = 0.6 MVAR after the control
It can be shown that a 1,200 kVAR capacitor bank out on a
feeder may not give an exact 1,200 kVAR change when switched
on or off. This can be illustrated by noting that the per phase
VAR correction capacity of a capacitor is dependent on voltage
according to the relation:
VAR = V2/XC (4-1)
VAR = volt-amp reactive power
V = line-to-neutral voltage
Xc = single phase capacitive reactance
As shown in equation 4-1, the reactive power output of the
capacitor is proportional to voltage squared. For example,

consider a 3 phase 1,200 kVAR capacitor bank with capacitive
reactance of 145.2 ohms per phase and a nominal line-to-neutral
voltage rating of 7,620 Volts. At 1.05 per unit voltage, the
reactive power output of the capacitor is
kVAR(30) = 3(8,000)2/l 45.2
= (1.05)2 x 1,200 = 1,322
At 0.95 per unit voltage, the reactive power output of the
capacitor is
kVAR(30) = 3(7,239)2/l 45.2
= (0.95)2 x 1,200 = 1,083
This means that the capacitor control algorithm could see a three-
phase VAR change within +10% of the capacitor rating due to
voltage variations of 5%. Hence, in setting the capacitor control
parameters for each feeder the valid operating limits or
"bandwidth" should be set to at least 110% of the largest capacitor
bank on the feeder.
4.6.3 Radio Paging Terminal
The capacitor control PC host was interfaced to PSCo's
corporate radio paging terminal, located at another facility,

through a dedicated telephone line. Previously, the paging
terminal had been used exclusively for sending digital messages
to "belt" pagers of employees. However, PSCo's telecommunication
department recognized the paging terminal had excess capacity
and recommended to the DA Committee that it be utilized for
capacitor control, thus taking advantage of an existing
telecommunications resource and company asset.
The paging terminal accommodated an external computer
interface through a standard RS-232 serial interface and an ASCII
protocol. Each radio capacitor control address was mapped in the
paging terminal database. To control a capacitor control radio
receiver (CCR), the PC would send an ASCII character string to the
paging terminal, representing the CCR "subscriber" account and
function code. Paging subscribers are typically identified by a 7
digit phone number and "cap code." The paging terminal would
queue pages and transmit them in batch. The paging terminal
consisted of four transmitters in the Denver-Boulder region which
provided overlapping signal coverage.
4.6.5 Capacitor Control Radio Receivers
Capacitor control radio receivers (CCRs), commercially
available from only one vendor at the time, were chosen for the

pilot project. The CCRs operated according to a function code
contained in the paging protocol message structure. The CCR was
capable of four functions: OPEN, CLOSE, RESET TIMER, and TEST.
The OPEN function would energize a 10 Amp, magnetically
latching relay, which opened a contact to de-energize the capacitor
bank by operating 3 oil switches. The CLOSE function did just the
opposite of the OPEN function. The RESET TIMER function would
override a close inhibit time delay, which was pre-set by the
factory. The close inhibit timer would prevent a close control
signal from re-energizing a capacitor bank within 7.5 minutes of
an OPEN control function. Lastly, the TEST function would
illuminate a yellow LED for radio signal testing purposes only and
performed no control operation of the capacitor bank.
4.6.6 Pilot Project Results
The pilot project was successful from the standpoint of
proving the concept of automated remote control of feeder
capacitors. However, the pilot system was not expanded using the
same CCR equipment or computer system for a few reasons. First,
the equipment used in the pilot turned out to be highly unreliable.
Secondly, there was an effort to work towards integrating certain

distribution automation functions within a new energy
management system as described further in Section 4.7.
Fifty percent of the CCRs failed by the end of the pilot
project term due to a variety of problems. The causes of CCR
problems that could be determined included cold-solder joints on
the circuit board, weak mounting provisions for some of the larger
components on the circuit board, power supply failures, and
generally poor radio receiver performance. The 50% failure rate
of the CCRs was attributed to physical defects or mechanical
failures. PSCo later determined that the CCR manufacturer's
receiver specifications were substandard, and that the CCRs would
not work reliably from an RF standpoint. Furthermore, the PC-
based capacitor control software license was too expensive and
limited in functionality to warrant expansion beyond the pilot
project. In order for capacitor automation to be continued a more
suitable computer platform and field hardware had to be selected.
4.7 Energy Management System Replacement
In 1990, PSCo began the planning and procurement process
for replacing the corporate EMS. At the same time, another
project team was in the process of evaluating vendors for a

separate distribution automation and SCADA platform. When the
DA team came up with the disappointing conclusion that the
established SCADA vendors had very little to offer in the way of
developed and proven distribution automation technology,
management from several divisions agreed to combine part of the
DA efforts with the EMS replacement project.
Subsequently, the scope of PSCo's EMS replacement project
included SCADA, automatic generation control (AGC), advanced
network applications, and some distribution automation
functionality. The DA functions included feeder capacitor control,
extended SCADA to the feeder level, and automated report
generation of substation and feeder peak demands. The new EMS
computer system, based on Digital Equipment Corporation's VAX
computer platform, was commissioned in September of 1992.
The major economic justification presented to management
for replacing the EMS included $3.5 million present worth in
maintenance savings and an additional $2.0 million savings by
implementing automated remote feeder capacitor control. The
savings from automated capacitor control was based on deferred
capital construction costs of substation capacitor banks planned to
be installed over the next two to three years. Improving the
availability of feeder capacitors during system peaks could, in

part, make up the difference in reactive power supply of the
deferred substation capacitor installations.
4.7.1 Substation RTU Upgrades
To implement remote feeder capacitor control on a full scale
basis, additional transducers had to be installed in the substation
to monitor three phase Watts and VARs on every feeder. A
project was already underway to replace or upgrade all substation
RTUs in the Denver area and incorporate additional feeder analog
point monitoring. The RTU replacements were scheduled to span
a five year period, from 1991 through 1995. Hence, automated
feeder capacitor control would be implemented to correspond
with the substation RTU upgrades and new installations over this
time period.
4.7.2 New Feeder Capacitor Control Application
The EMS replacement project included feeder capacitor
control as one of the DA applications specified. During contract
negotiations with the vendor, it was decided that PSCo would
develop this software application in-house, with consulting from

the vendor as needed. The development project was assigned to a
very experienced energy control systems programmer analyst at
PSCo. The automated remote capacitor control software
application at PSCo was named CAPCON (not to be confused with
Virginia Power's capacitor control program with the same name).
4.7.3 CAPCON Functionality
By developing the CAPCON application in-house, PSCo was
able to incorporate more functionality in the application than the
EMS vendor could provide at the time, and at a more reasonable
cost. In fact, minor changes and features were added throughout
the development cycle. By having direct interaction with the
programmer, the design team was able to better communicate and
work out important application design issues. In developing the
application, several factors in the software's functional design
were considered. These include the following:
database design
closed-loop control algorithm
automatic and manual control modes
control parameters for individual and group operation
capacitor tracking and operation verification
interfaces to other applications
alarming and reporting

maintenance of the system
operator interface
paging terminal interface
ability to add features or make future modifications.
4.7.4 CAPCON Closed Loop Control Algorithm
The primary telemetered data used in the CAPCON algorithm
is MW and MVAR flow measured at each feeder breaker and
monitored by the substation RTU. During each automatic control
cycle, which is every 15 minutes, all feeders are examined to
determine if conditions warrant the operation of a capacitor bank.
If outside the normal operating range, the algorithm then looks at
feeder MW to determine the load relative to the feeder's peak
demand defined in the data base. The sign of the MVAR flow
determines whether to open or close a switched capacitor. For
example, if the MVARs are negative, implying a leading power
factor, then a capacitor will be opened; if positive, implying a
lagging power factor, then a capacitor will be closed. A simplified
flow diagram of the algorithm is shown in Figure 4.2.
Three switching priority schemes are defined in the
database for every feeder and are based on three ranges of feeder
load. The switching scheme determines the order in which the

capacitors on a given feeder will be switched OPEN or CLOSED at a
given load level; the default scheme is that capacitors closest to
the substation will be switched OPEN first and CLOSED last, and
the furthest capacitor switched CLOSED first and OPEN last. This
switching priority scheme is intended maintain an acceptable
voltage profile and minimize losses at various load levels while
trying to achieve near unity power factor at the substation at all
times. The switching priority feature can also be used to prevent
CAPCON from closing certain capacitor banks during light load
conditions to avoid causing a high voltage condition (i.e., greater
than 126 V service voltage to consumers).

Figure 4.2: Capacitor control algorithm at PSCo [27].

4.7.5 Control Operation Modes
There are five defined capacitor bank operation modes. The
five capacitor modes and their definitions are as follows [30].
1. Full Automatic or Closed Loop (AUTO) In this mode, when
CAPCON determines the need for a capacitor operation, and
an eligible capacitor is found, the control will be sent
immediately. No operator intervention is required.
2. Permissive (PERM) In this mode, when CAPCON determines
the need for a capacitor operation and an eligible capacitor
is found that is in PERM mode, an alarm will be issued to
inform the operator that the application has determined that
the capacitor should be controlled, but the function must be
done manually by the operator. It is then the operators
decision whether or not to control the capacitor.
3. Manual (MANL) When a capacitor is in MANL mode, the
application will not attempt to operate it. The operator can
still manually operate the capacitor, however. Note that if
field switching (i.e. feeder reconfiguration) has taken place
that causes a capacitor to be temporarily connected to a
different feeder than the one assigned in the database, the
operator would place the capacitor in MANL mode. This will
keep the automatic process from operating a capacitor based
on data from an incorrect feeder.
4. Unrestricted manual (UNRES) When a capacitor is in
unrestricted mode, the operator may manually control it at
will. A capacitor in this mode is not constrained by
parameters such as maximum operations per day, a time