Вы находитесь на странице: 1из 192

Application Guide for the Automation of

Distribution Feeder Capacitors

Technical Report

Effective December 6, 2006, this report has been made publicly available in accordance
with Section 734.3(b)(3) and published in accordance with Section 734.7 of the U.S. Export
Administration Regulations. As a result of this publication, this report is subject to only
copyright protection and does not require any license agreement from EPRI. This notice
supersedes the export control restrictions and any proprietary licensed material notices
embedded in the document prior to publication.
Application Guide for the
Automation of Distribution Feeder
Capacitors

1010655

Final Report, December 2005

EPRI Project Manager


A. Sundaram

ELECTRIC POWER RESEARCH INSTITUTE


3420 Hillview Avenue, Palo Alto, California 94304-1395 • PO Box 10412, Palo Alto, California 94303-0813 • USA
800.313.3774 • 650.855.2121 • askepri@epri.com • www.epri.com
DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES
THIS DOCUMENT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN
ACCOUNT OF WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH
INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE
ORGANIZATION(S) BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM:

(A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I)


WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR
SIMILAR ITEM DISCLOSED IN THIS DOCUMENT, INCLUDING MERCHANTABILITY AND FITNESS
FOR A PARTICULAR PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR
INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL
PROPERTY, OR (III) THAT THIS DOCUMENT IS SUITABLE TO ANY PARTICULAR USER'S
CIRCUMSTANCE; OR

(B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER


(INCLUDING ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE
HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR
SELECTION OR USE OF THIS DOCUMENT OR ANY INFORMATION, APPARATUS, METHOD,
PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT.

ORGANIZATION THAT PREPARED THIS DOCUMENT

EPRI Solutions, Inc.

NOTE
For further information about EPRI, call the EPRI Customer Assistance Center at 800.313.3774 or
e-mail askepri@epri.com.

Electric Power Research Institute and EPRI are registered service marks of the Electric Power
Research Institute, Inc.

Copyright © 2005 Electric Power Research Institute, Inc. All rights reserved.
CITATIONS

This report was prepared by

EPRI Solutions, Inc.


942 Corridor Park Blvd.
Knoxville, TN 37932

Principal Investigators
D. Crudele
T. Short

This report describes research sponsored by the Electric Power Research Institute (EPRI).

The report is a corporate document that should be cited in the literature in the following manner:

Application Guide for the Automation of Distribution Feeder Capacitors, EPRI, Palo Alto, CA:
2005. 1010655.

iii
PRODUCT DESCRIPTION

This is the fourth and final report in the Electrical Power Research Institute’s (EPRI’s) capacitor
reliability study, and it deals with automating distribution capacitors. Prior reports dealt with
nuisance fuse operations, operating and construction practices, and lighting protection and
grounding of capacitor controllers. This guide is concerned with applying automated switched
capacitors to distribution systems. Consideration is given to applications involving locally
controlled capacitor banks and to systems utilizing centrally controlled, switched capacitor
banks. The guide is designed for the distribution engineer considering capacitor automation for
his or her system.

Results and Findings


The Application Guide for the Automation of Distribution Feeder Capacitors attempts to provide
the utility engineer with the background needed to sufficiently understand automated capacitor
control and the ways it might be applied to his or her distribution system. This guide discusses
commonly applied capacitor control schemes, including both locally applied and centralized
control schemes. The reader is presented with resources for locating a variety of capacitor
control equipment currently available from several prominent manufacturers in this area. This
guide also discusses the issues of system integration, capacitor protection, control schemes, and
capacitor-related power quality issues.

Challenges and Objectives


This guide is intended to provide the necessary background for a distribution engineer to quickly
acquire a working knowledge of the issues associated with capacitor automation, including:
• Types of capacitor automation schemes (local control versus centralized control)
• Ways capacitor automation is employed
• Advantages and drawbacks of different types of capacitor controls
• Supervisory control and data acquisition (SCADA) systems for capacitor control
• Communication systems used for capacitor control
• Capacitor bank sizing and protection issues
• Capacitor power quality issues
Due to the potential variability of the capacitor control system from one utility to the next, it is
difficult to assign costing figures that will cover all capacitor automation systems. Therefore, this
Guide attempts to describe the various payback streams that come from implementing

v
sophisticated capacitor automation schemes. This will allow readers to assign their own dollar
savings to each category and determine their own potential payback.

Applications, Values, and Use


Distribution automation has emerged as a tremendous resource for increasing efficiency and
decreasing operating costs for the modern electric utility. Advancements in communication and
control technologies have made many automation programs—never-before available—a part of
the daily operation of utilities around the United States. Among the array of attractive
distribution automation technologies are automated capacitor controls, which lead the way as
perhaps the most desirable control technology in terms of increasing operating efficiency and
providing a quick return on utility investments. This guide provides a detailed look at many of
the aspects of distribution capacitor automation in order to help the distribution engineer quickly
gain the background needed to seriously examine capacitor automation applications.

EPRI Perspective
Capacitor automation technology has advanced greatly in recent years. Utilities now have access
to intelligent, automated capacitor controllers from numerous manufacturers. Many controllers
on the market also have advanced communication capabilities allowing them to be easily
integrated into SCADA systems. These advances in capacitor automation technology, coupled
with the modern utility’s need to operate ever more efficiently, have utilities taking a closer
examination of how capacitor automation can benefit their distribution systems. This guide is
intended to aid the distribution engineer or planner in determining how capacitor automation can
be a benefit to their distribution system as well as provide the background information and
automation fundamentals needed to seriously examine how to automate the capacitors on their
system.

Approach
The project team began by researching all available information on state-of-the-art capacitor
automation systems currently in use by utilities. From this research, sections have been added to
discuss the various types of control schemes used for capacitor automation and local control
verses centralized control topologies. The project team also researched SCADA systems used for
modern capacitor automation and have attempted to provide a detailed overview of SCADA
systems so readers may better understand how these systems can be utilized in distribution
automation.
No discussion of utility SCADA is complete without examining the many communication
channels available to transfer data from the central station to field units and back. Therefore, one
chapter of this Guide is dedicated to examining SCADA communication media, with particular
attention paid to which companies currently offer commercial communication services for each
medium. Finally, basic capacitor application information is presented in chapters dedicated to
capacitor installation sizing, location, protection, and power quality issues.

Keywords
Capacitor automation Capacitor control
Switched capacitor Distribution automation
SCADA Volt/VAR management
VARs

vi
EXECUTIVE SUMMARY

The EPRI Capacitor Reliability Study

Utilities have a substantial investment in distribution line capacitors. These investments are
justified, based on certain derived benefits to the power delivery system, the utilities, and the
end-users. When capacitors are not available due to some failure or operating error (or are
otherwise off-line), the anticipated benefits will not be achieved. Experience at utilities reveals
that capacitors are unavailable for operation too frequently. This project series was established,
therefore, to improve capacitor reliability. Initial scoping helped identify and prioritize several
issues affecting the overall reliability of capacitors. The EPRI capacitor reliability study spans
several years, from 2002 through the present. Each year a report is prepared dealing with a
different aspect of capacitor reliability. Reports from previous years have covered:
• Utility Survey and Literature Search (2002): This study was a utility survey and literature
search to assess the issues related to the reliability of switched capacitor banks used in
distribution systems (EPRI 1001691).
• Fusing and Transmission Support (2003): This study investigated causes of nuisance fuse
operations on capacitor banks. Additionally, utility practices for providing transmission level
VAR support with distribution capacitors were reviewed, and additional utility needs were
assessed (EPRI 1002154).
• Grounding and Lightning Protection of Capacitor Controllers (2004): Investigate the two
primary factors influencing the magnitude of surges reaching capacitor controllers and
provide controller mounting and wiring configurations for minimizing surge magnitude. The
first recommendation involves the physical location at which the capacitor controller should
be mounted with regard to the control power transformer (CPT) from which it draws power.
The second recommendation involves grounding considerations for the controller supply
power (EPRI 1008573).

This year’s report, 2005, examines automating switched capacitors at the distribution level. This
guide attempts to provide the utility engineer with the background needed to sufficiently
understand automated capacitor control and the ways it could be applied to their distribution
system. This guide discusses commonly applied capacitor control schemes, including locally
applied control and centralized control schemes. The reader is presented with a variety of
resources for locating capacitor control equipment from several prominent manufacturers in this
area. This guide also discusses the issues of system integration, capacitor protection, control
schemes, and capacitor-related power quality issues.

vii
Project Objectives

The primary focus of this guide is to provide distribution engineers with the necessary
information to examine options for applying a switched capacitor automation scheme on their
distribution system. This guide provides a detailed discussion of the all the key aspects of
distribution capacitor automation, including:
• Control Schemes: VAR, voltage, current, time, temperature, date, and combination control
programs
• Control Intelligence Location: Local control, central coordinated control, local control with
central station override
• Supervisory Control and Data Acquisition (SCADA) Systems: Components commonly found
in SCADA-based capacitor control systems, with examples cited from prominent
manufacturers
• Voltage and Current Measurements: Information on line parameters typically measured and
the potential for modern capacitor controllers to gather and report a wide array of line data to
aid distribution engineers in investigations beyond VAR management
• Capacitor Sizing and Placement: Detailed information size and placement of capacitor
banks on the distribution system
• Capacitor Installation Protection: Detailed information on proper application of fuses to
protect capacitor banks, with additional information regarding protecting capacitor
controllers from line surges and lighting strikes

Background

There is considerable industry activity in applying distribution feeder capacitors. Automated line
capacitors are being added by many utilities. Automation and communication technologies are
more advanced, more readily available, and more reasonably priced than even before. These
advancements in automation control and communication allow utilities to operate switched
distribution capacitors in a manner that has never before been possible. Utilities are using
capacitors in a variety of ways—to supplement transmission VARs, as substitutes for substation
capacitors, to manage distribution voltage profiles, and to reduce line losses. Communication
technology allows centralized control of distribution capacitors as if they were substation banks.
This adds the benefit of having the capacitors located closer to the loads they service, thereby
further improving their operating efficiency.

A typical switched capacitor bank installation is shown in Figure ES-1. Although Figure ES-1
only shows the capacitor assembly near the pole top, the capacitor controller is mounted lower
on the pole, approximately 10 ft (3 m) above the ground. There are many types of controllers on
the market, with many different configurations.

viii
Figure ES-1
Example of a Switched Capacitor Bank Configuration

Distribution line capacitors provide tremendous benefits to distribution system performance by


providing volt-ampere-reactives (VARs) at or near the VAR-consuming loads—and they do this
at a low cost. The main benefits that capacitors provide are:
• Reduced Losses and Increased Capacity: By canceling the reactive power to motors and
other loads with low-power factor, capacitors decrease the line current. Reduced current frees
up capacity. Reduced current also significantly lowers I2R line losses.
• Reduce Voltage Drop: Capacitors provide a voltage boost that cancels part of the drop
caused by system loads. Switched capacitors serve to regulate voltage on a circuit, having an
ancillary benefit of reducing the number of operations on voltage regulators, both line (and to
a lesser to degree) substation regulators and load-tap-changers (LTCs). This reduces
maintenance costs on regulators and LTCs.
• Reduced Cost of Production or Cost of Purchased Power: Because line capacitors provide
VARs, generators no longer have to produce VARs, thus capacity is freed up to produce
more real power. (In addition, transmission and distribution lines no longer have to transport
those VARs.)

If applied and controlled properly, capacitors can significantly improve the performance of
distribution circuits. But if not properly applied or controlled, the reactive power from capacitor
banks can create losses and can also create high voltages. The most danger of overvoltage is
under light loads. Good planning helps ensure that capacitors are sited properly. Compared to

ix
simple controllers (like a time clock), more sophisticated controllers (such as a two-way radio
with monitoring) reduce the risk of improperly controlling capacitors.

Capacitors work their magic by storing energy. Capacitors are simple devices—two metal plates
sandwiched around an insulating dielectric. When charged to a given voltage, opposing charges
fill the plates on either side of the dielectric. The strong attraction of the charges across the very
short distance that separates them creates a tank of energy. Capacitors oppose changes in voltage.
It takes time to fill up the plates with charge; and once charged, it takes time to discharge the
voltage.

On ac power systems, capacitors don’t store their energy very long—just one-half cycle. Each
half cycle, a capacitor charges up and then discharges its stored energy back into the system. The
net real power is zero. Capacitors provide power just when reactive loads need it. At the time a
motor with low-power factor needs power from the system, the capacitor is there to provide it.
Then in the next half cycle, the motor releases its excess energy, and the capacitor is there to
absorb it. Capacitors and reactive loads continue to exchange this reactive power. This benefits
the system because that reactive power (and extra current) does not have to be transmitted from
the generators all the way through many transformers and many miles of lines; therefore, the
capacitors can provide the reactive power locally. This frees up the lines to carry real power that
actually performs work.

Control Strategies

Local Control

Switched capacitor banks are controlled either locally or through centralized system controls. As
the name implies, local controls sit on or near the same pole as the capacitor bank and govern the
switching operations of only one local bank. There are several local control strategies available
for switched capacitor banks, as shown below (Marx 2003); (Short 2004b):
• VAR Control: The capacitor is switched on and off at an optimum point in the load cycle
based on VAR measurements on the line. VAR control is the most efficient control strategy
for maximizing the reduction of loss and demand on feeders having only one capacitor bank
installed. However, VAR control is susceptible to interaction from downstream capacitor
banks (downstream banks affect the reactive current flow upstream of their location).
Therefore, when applying multiple capacitor banks using VAR control on a single feeder, the
controls should be set such that the bank furthest downstream comes on-line first, followed
by the next upstream bank, and so on. Furthermore, the banks should then trip in the opposite
order by which they switched in (that is, the last to switch in should be the first to trip out).
• Current Control: The capacitor is switched on and off based on the line current measured
downstream of the capacitor. Reactive current can be determined from line current when the
power factor of the line is known. Current control engages the capacitor during periods of
heavy loads which generally have the greatest VAR requirements. Although not as effective
as VAR control schemes, current control provides a fairly good combination of loss
reduction and voltage control.

x
• Voltage Control: The capacitor is switched on and off based upon the voltage. To prevent
excessive operations, threshold minimum and maximum voltages are programmed into the
controller, as well as time delays and bandwidths. Voltage control is best suited for
applications in which the capacitor mainly provides voltage profile control and regulation.
Voltage controls can be influenced by both upstream and downstream capacitors, since they
affect the voltage along the whole line. Voltage regulators can also cause capacitor control
pumping problems. In general, capacitor controllers using voltage control schemes should be
configured to operate prior to the local voltage regulators. In this manner, the voltage
regulators operate only when the capacitors cannot maintain the desired voltage profile.

It should also be noted that voltage control schemes provide the greatest value on feeder
sections further from the substation. The capacitor should have a minimum effect of 2 V (on
120 V reference), and the cap on-to-off difference should be approximately 1.5 times the
expected voltage rise when the bank is switched on (Marx 2003).
• Time-Clock Control: The controller switches the capacitor, based upon the time of day.
Time-clock control represents the most basic approach for switching a capacitor on and off.
Most time-clock controllers allow for programmable on and off time settings, as well as
settings for weekends and holidays. While this is the least expensive control option, it is also
the most susceptible to energizing the capacitor at the wrong time; because switching is
based on expected line conditions rather than on measured conditions. Loads can be different
than those anticipated at any time, but holidays and weekends are particularly challenging.
Time-clock controllers, which are susceptible to mistaken time settings and inaccurate
clocks, can switch the capacitor at times other than those planned. Since time control is not
based on line measurements, time-clock controls are not susceptible to interaction with other
banks.
• Temperature Control: The capacitor is switched based upon the temperature. Like
time-clock controls, temperature controllers also provide a very basic level of capacitor
control. Typically, temperature controls are set to turn the bank on at 85-90º F (29.4–32.2º C)
and turn the bank off again at 75-80º F (23.8–26.7º C). Since temperature control is not based
on line measurements, they are not susceptible to interaction with other banks.
• Power Factor Control: The capacitor is switched based upon the power factor measured on
the line. This method of control is rarely used by utilities, mostly owing to the fact that power
factor is not a suitable parameter for controlling capacitor switching. Since power factor is
not necessarily an indication of load, power factor controls may fail to switch in the capacitor
during high loads, if the power factor is also high. To compensate for this shortcoming,
power factor controls may also incorporate voltage and current overrides, both of which
make the system more complicated. Due to these reasons, VAR control is typically used
rather than power factor control.

Many controllers offer some or all of these control strategies. Many are usable in combination;
for example, they will turn capacitors on for either low voltage or high temperature.

xi
Centralized Control

Advances in wireless communication technology have made remote capacitor control more
achievable and more economical than ever before. Cellular phones, pagers, and other wireless
technologies have become ubiquitous in modern life, turning up in new applications, such as
remote capacitor control. There are several control schemes available for remotely controlled
capacitor installations, including:
• Operator Dispatch: Most schemes allow operators to dispatch distribution capacitors. This
feature is one of the key reasons utilities automate capacitor banks. Operators can dispatch
distribution capacitors just like large station banks. If VARs are needed for transmission
support, large numbers of distribution banks can be switched on. This control scheme is
usually used in conjunction with other controls.
• Time Scheduling: Capacitors can be remotely switched, based on the time of day and
possibly the season or temperature. While this may seem like an expensive time control, it
still allows operators to override the schedule and dispatch VARs as needed.
• Substation VAR Measurements: A common way to control feeder capacitors is to dispatch
based on VAR/power factor measurements in the substation. If a feeder has three capacitor
banks, they are switched on or off in some specified order, based on the power factor on the
feeder measured in the substation.
• Capacitor Location VAR Measurements: The continuing advancement or capacitor
controller capabilities, coupled with increasing capability for two-way data transfer, are now
making it possible for capacitor controllers to measure line parameters at their location and
report that data back to a central station controller. The central station controller examines
the data from each capacitor location (and possible the substation as well) and makes
decisions for switching each capacitor individually. A major detractor of this type of
operation is that current transformers (CTs) need to be installed at each site in order to make
VAR measurements; and this carries a significant equipment cost—much higher than just
measuring at the substation.
• Other Methods: More advanced (and complicated) algorithms can be used to dispatch
capacitors, based on a combination of local VAR measurements and voltage measurements,
along with substation VAR measurements.

All of the control strategies mentioned above will typically utilize a local voltage override
feature, especially if the controller has only one-way communication capabilities. Local voltage
override prevents the capacitor from switching if doing so will push the voltage beyond limits set
by the user. Additionally, most controllers used for centralized control will have fail-safe modes
in which they will revert to a type of local control (voltage, current, VAR, time, temperature,
combination, and so on.) if communication with the central station is lost.

Capacitor Controllers

The capacitor controller is really the backbone of the automated switched capacitor system. Both
local control schemes and centralized control schemes utilize a local capacitor controller. At the

xii
most basic level, the controller provides the interface to the capacitor switch, telling it when to
open and close. In local control schemes, the controller provides the switching logic. In central
control schemes, the controller 1) houses and interprets the signals provided by the data radio, 2)
provides switching override functions based on local conditions, and 3) provides switching logic
in the event that communication with the central station is lost.

There are many models of capacitor controllers available from numerous manufacturers. The
controllers are typically packaged in weatherproof enclosures and are intended to be mounted on
the same pole as the capacitor bank and switch. Some examples of capacitor controllers from
various manufacturers are shown in Figure ES-2.

Figure ES-2
Examples of Capacitor Controllers from Various Manufacturers

Since there is wide variability in capacitor control needs from one utility to the next, there is also
a corresponding wide variety of features among currently produced capacitor controllers. Most
manufacturers try to cover most, if not all, of the possible features that a utility may require,
including:
• Communication: None (local control only), one-way, two-way
• Communication Channel: Radio, cellular, fiber optic, paging, copper line, and so on.
• Control Type: Volt, current, VAR, time, temperature, combination control
• Monitoring: Some controllers with two-way communication ability to also report data on a
variety of parameters: voltage, current, watts, power factor, temperature
• Data Storage: Some controllers can store operational data locally for retrieval by utility field
personnel via laptop computer
• Reverse Power Detection: As part of their monitoring capability, some controllers can detect
reverse power conditions on the feeder. Additionally, some controllers have the functionality
to calculate proper set points and compensate for atypical line measurements during reverse
power flow conditions.
• Neutral Current Monitoring: Monitoring the capacitor bank neutral current can help
diagnose problems, such as blown fuses, failing capacitor units, and high harmonic currents.
Further information on neutral current monitoring is available in Chapter 5, “Voltage and
Current Measurements.”

xiii
Most controllers have functionality for all local control types (volt, current, VAR, time,
temperature, and so on.); and they can often run a combination program incorporating two or
more of these parameters in a hierarchical manner. Most manufacturers also cover both local
control and centralized control with one- or two-way communication capabilities, frequently by
providing different models, each with distinct communication capabilities.

SCADA Systems

Basic SCADA systems, also referred to as telecontrol systems, consist of a master station(s)
communicating with one or more remote terminal units (RTUs) to provide data acquisition and
control functionality between a central location and dispersed field units. A very simplistic
diagram of a SCADA system is provided in Figure ES-3 to illustrate the concept of centralized
control of dispersed field units. The communication channel between the master controller and
remote units can be any one of a number of technologies, including radio, cellular, modem, or
hard-wired networks. There are numerous protocols available that define how communications
between the master station and remote units should be structured over the communication
channel, although the DNP3 protocol tends to dominate new capacitor control systems. The
master station runs application software that provides the human-machine interface and also
provides the functionality to perform the specific tasks for which the SCADA system is used
(that is, capacitor control, process control, data acquisition). Alternatively, in larger multi-
function SCADA systems, the master station may provide overall coordination and data archival,
while dedicated servers run individual function programs, such as the DCC system illustrated in
Figure ES-3.

Figure ES-3
Components of a Basic SCADA System

Communication Technology

There are several technologies currently in use for communicating with the capacitor controllers.
Some offer one-way communication while others offer two-way communication. With one-way
communication, commands can be dispatched to the capacitor controllers in the field, but there is

xiv
no communication from the field back to the control center. Two-way communications offer data
flow both from the command center to the field units and from the field units back to the
command center. The technologies used for centralized capacitor control communications
include:
• 900-MHz Radio: These systems are very common and widely applied for centralized
capacitor control. There are several spread-spectrum radios available that cover 902-928
MHz applications. Implementing 900-MHz radio control on a private network requires
infrastructure, including towers.
• Pager Systems: Pager systems offer inexpensive options, especially for systems with
infrequent switching. These systems are mostly one-way, but there are some two-way pager
systems available. Most commercial paging systems can be utilized, however that means that
while one-way coverage is rather wide-spread, two-way systems tend to be limited to clusters
around major cities.
• Cellular Phone Systems: These systems use commercial cellular networks to provide two-
way communications. Many vendors offer modems that are compatible with several cellular
networks, and coverage is typically very good.
• Cellular Telemetric Systems: These use the unused data component of cellular signals that
are licensed on existing cellular networks. They allow only very small messages to be sent to
perform basic capacitor automation needs. Coverage is typically very good, the same as
regular cellular coverage.
• Very High Frequency (VHF) Radio: Inexpensive, one-way communications are possible
with VHF radio communication. VHF radio bands are available for telemetry uses such as
this. Another option is a simulcast frequency modulation (FM) signal that uses extra
bandwidth available in the commercial FM band.

Economics

Utilizing automated, intelligent capacitor bank switching controls provides several channels of
payback that generally yield a very fast return on investment. In fact, there are few capital
projects that a utility can undertake that provide a faster return. Automated capacitor control
generates three main areas of cost savings as follows:
• Energy Savings: In this project, energy savings (also termed loss reduction) refers to
reducing line and transformer losses by using intelligent capacitor control to effectively
reduce the amount of reactive current flowing in the line. Since energy wasted in heating
conductors cannot be delivered to a customer, it generates no revenue. It also contributes to
fatigue on line conductors and apparatuses through heating.
• Capacity Savings: Improving the line power factor through proper application of capacitors
reduces the total line current, thus reducing kVA demand. The benefits provided by released
capacity are twofold. First, releasing line capacity allows more billable energy to be
transferred to customers, thus increasing the revenue that the line can generate. The second
benefit of releasing line capacity is that it can enable the deferral of equipment upgrades.
Improving the power factor releases transmission and generation capacity as well as
distribution capacity.

xv
• Operation and Maintenance Savings – Required labor hours can be greatly decreased when
upgrading to intelligent centralized capacitor controls via supervisory control and data
acquisition (SCADA) systems. SCADA control greatly reduces labor costs by allowing for
centralized switching control and monitoring of all capacitor banks. This dramatically
reduces travel time as well as time spent adjusting capacitor bank controls. Additional cost
savings come from the ability to remotely monitor capacitor bank status to determine when
capacitors fail. This also eliminates the need to have technicians travel to capacitor
installations to annually inspect bank functioning, which amounts to a considerable savings
in work-hours. The ability to quickly identify and fix failed capacitors also means that fewer
capacitors would need to be installed in the system, since a very high percentage would be
operational all the time. Over time, then, some capacitor banks could be taken out of service
and used for future installations, providing a capital cost savings.

Capital costs for capacitor control systems can vary greatly, depending on the level of
sophistication being employed and what, if any, existing utility infrastructure can be utilized for
the system. However, the level of existing hardware also plays a role in determining the design
of the capacitor control system. For example, if a utility already has an extensive 900-MHz radio
system in place, then they will likely utilize that system for communication in their capacitor
control system. If a utility does not have any communication system is place, they may opt for a
commercially provided communication system (such as a cellular control channel) rather than
building their own communication network. Even utilities that have a communication network in
place may opt for commercially provided communications, since commercial systems require no
infrastructure maintenance from the utility.

xvi
CONTENTS

1 PROJECT OVERVIEW...........................................................................................................1-1
Control Strategies..................................................................................................................1-3
SCADA and Communications ...............................................................................................1-4
Project Objectives .................................................................................................................1-5

2 CAPACITOR SIZING AND PLACEMENT .............................................................................2-1


Introduction ...........................................................................................................................2-1
Capacitor Ratings..................................................................................................................2-5
Released Capacity ................................................................................................................2-9
Voltage Support ..................................................................................................................2-11
Reducing Line Losses .........................................................................................................2-13
Energy Losses ....................................................................................................................2-16
Grounded versus Ungrounded ............................................................................................2-17
Impact of Switching on Capacitor Sizing and Placement ....................................................2-19
Switched Capacitor Bank Equipment Mounting Considerations .........................................2-19
Optimal Capacitor Placement Computer Programs ............................................................2-23

3 AUTOMATION STRATEGIES................................................................................................3-1
Best Use of Distribution VARs...............................................................................................3-1
Conservation Voltage Reduction...........................................................................................3-1
Optimizing Power Factor at the Substation ...........................................................................3-3
Distribution Capacitors for Transmission VAR Support.........................................................3-3
KCPL ................................................................................................................................3-3
Idaho Power .....................................................................................................................3-4
Cinergy Corp ....................................................................................................................3-5
Georgia Power..................................................................................................................3-6
Summary of Utility Practices ........................................................................................3-7
Transmission versus Distribution Optimization ............................................................3-8

xvii
Switching Control .........................................................................................................3-8
Station versus Feeder Evaluation ..............................................................................3-11
Automation and Other Infrastructure Requirements............................................................3-12

4 CONTROL STRATEGIES ......................................................................................................4-1


Control Strategies..................................................................................................................4-1
Local Control ....................................................................................................................4-1
Centralized Control...........................................................................................................4-4
Coordination of Switched Capacitors and Voltage Regulators..............................................4-6
Coordination of Switched Capacitors and Distributed Generation ........................................4-7

5 VOLTAGE AND CURRENT MEASUREMENTS....................................................................5-1


Basic Measurements.............................................................................................................5-1
Neutral Monitoring .................................................................................................................5-4

6 COMMUNICATION TECHNOLOGIES ...................................................................................6-1


Communications Technologies .............................................................................................6-1
Spread Spectrum 900-MHz Radio Systems..........................................................................6-4
Pager Systems ......................................................................................................................6-5
FLEX™ Paging Protocol...................................................................................................6-6
Cellular Systems ...................................................................................................................6-7
Cellular Data Channel Systems........................................................................................6-7
Cellular Digital Packet Data..............................................................................................6-9
Cellular Antennas .............................................................................................................6-9
Commercial Support for Communication Planning and Analysis ........................................6-11

7 CAPACITOR CONTROLLERS AND SCADA SYSTEMS......................................................7-1


Capacitor Controllers.............................................................................................................7-1
SCADA Overview ..................................................................................................................7-8
Master Stations ...................................................................................................................7-10
Protocols .............................................................................................................................7-10
Distributed Network Protocol (DNP3) .............................................................................7-11
IEC 60870.......................................................................................................................7-13
Utility Communications Architecture...............................................................................7-15
MODBUS........................................................................................................................7-16
RTUs, IEDs, and PLCs........................................................................................................7-16

xviii
SCADA Security ..................................................................................................................7-17

8 SOFTWARE AND DATA APPLICATIONS ............................................................................8-1


Capacitor Control Software for SCADA Systems..................................................................8-1
Device and Data Management Software...............................................................................8-3
Human-Machine Interface Issues..........................................................................................8-4

9 CAPACITOR AND CONTROLLER SURGE PROTECTION..................................................9-1


Primary Arrester Lead Length and Coordination with Fuses.................................................9-1
Lead Length Considerations.............................................................................................9-1
Arrester Installation Clearance Considerations ................................................................9-5
Capacitor Controller Surge Protection...................................................................................9-6
Modeling of Lightning Surges Originating on the Primary Conductors .............................9-7
Preliminary Recommendations.......................................................................................9-10
Key Considerations ........................................................................................................9-12
Controller Mounting Location .....................................................................................9-12
Ground Loops and Shielding .....................................................................................9-12
Arrester Lead Length .................................................................................................9-15
Auxiliary Surge Suppression......................................................................................9-15
Pole Ground Resistance ............................................................................................9-19
Consult the Manufacturer...........................................................................................9-19
Installation Guidelines................................................................................................9-19

10 CAPACITOR FUSING ........................................................................................................10-1


Fusing Guidelines................................................................................................................10-1
Reasons for Relaxing Fusing ..............................................................................................10-4
Maximum Fuse Sizes ..........................................................................................................10-6
Nuisance Fuse Operation....................................................................................................10-8
Outrush and Inrush..............................................................................................................10-9
Fuse Installation Issues.....................................................................................................10-16
Proposed Fusing Guidelines .............................................................................................10-18

11 CAPACITOR BANK POWER QUALITY AND RELIABILITY ISSUES..............................11-1


Harmonics ...........................................................................................................................11-2
Solutions to Harmonics...................................................................................................11-5
Switching Surges.................................................................................................................11-6

xix
Adjustable Speed Drive (ADS) Tripping ..............................................................................11-9
Solutions to Switching Transients .....................................................................................11-10
Telephone Interference .....................................................................................................11-12
Voltage Flicker ..................................................................................................................11-13

12 ECONOMICS......................................................................................................................12-1
Energy Savings ...................................................................................................................12-2
Capacity Savings.................................................................................................................12-2
Operation and Maintenance Savings ..................................................................................12-4
Estimated Cost Breakdown .................................................................................................12-4

13 REFERENCES ...................................................................................................................13-1

xx
LIST OF FIGURES

Figure 1-1 Example of Switched Capacitor Bank Configuration ................................................1-3


Figure 2-1 Capacitor Components.............................................................................................2-2
Figure 2-2 Overhead Line Capacitor Installation .......................................................................2-3
Figure 2-3 Released Capacity with Improved Power Factor....................................................2-10
Figure 2-4 Extra Capacity as a Function of Capacitor Size .....................................................2-10
Figure 2-5 Voltage Profiles After Addition of a Capacitor Bank ...............................................2-12
Figure 2-6 Optimal Capacitor Placement Using the “2/3’s” Rule .............................................2-14
Figure 2-7 Placement of 1200-kVAR Banks Using the ½-kVAR Method.................................2-15
Figure 2-8 Sensitivity to Losses of Placing One Capacitor on a Circuit with a Uniform
Load .................................................................................................................................2-16
Figure 2-9 Example of Real and Reactive Power Profiles on a Residential Feeder on a
Peak Summer Day with 95% Air Conditioning (Data from East Central Oklahoma
Electric Cooperative, Inc.) ................................................................................................2-17
Figure 2-10 Comparison of Grounded-wye and Ungrounded-wye Banks During a Failure
of One Unit .......................................................................................................................2-18
Figure 2-11 Example Switched Capacitor Bank Installation Courtesy of Donald M. Parker
at Alabama Power............................................................................................................2-20
Figure 2-12 Generic Switched Capacitor Bank Equipment Configuration ...............................2-21
Figure 2-13 Typical Layout of Pole-Top Equipment in a Switched Capacitor Installation
Courtesy of Donald M. Parker at Alabama Power............................................................2-22
Figure 2-14 Example Location of Fuses and Lightning Arresters in a Switched Capacitor
Installation Courtesy of Donald M. Parker at Alabama Power .........................................2-22
Figure 3-1 Three Steps for Applying Capacitors for Peak Shaving............................................3-7
Figure 3-2 Optimal Capacitor Location for Loss Reduction as the VAR Profile Changes........3-10
Figure 4-1 Example Feeder with a Switched Capacitor Located Just Upstream of a
Distributed Energy Resource .............................................................................................4-8
Figure 5-1 Typical Capacitor Controller Mounting Configuration with a Meter Socket
Courtesy of S&C Electric Company ...................................................................................5-1
Figure 5-2 Example of Connections in a 6-Jaw Meter Socket Used for Capacitor
Controller Installations........................................................................................................5-2
Figure 5-3 Generic Example of Pole-Top Connections for Input Signals to a Capacitor
Controller Note: Protection devices and other apparatuses have purposely been
omitted from this drawing for clarity. Actual installations would also utilize hardware,
such as surge arresters, cutouts, and fuses. .....................................................................5-3

xxi
Figure 5-4 Series 1301 PowerFlex® Current Sensors From Joslyn Hi-Voltage Courtesy
of Joslyn Hi-Voltage ...........................................................................................................5-4
Figure 5-5 S&C Electric Company’s CSV Line Post Current and Voltage Sensor
Courtesy of S&C Electric Company ...................................................................................5-4
Figure 5-6 Neutral Monitoring of a Capacitor.............................................................................5-5
Figure 5-7 Neutral Current Drawn by Failing, Grounded-Wye Bank, Depending on the
Portion of Bank Failed........................................................................................................5-5
Figure 6-1 Reflection of Radio Signals ......................................................................................6-3
Figure 6-2 SkyTel Telemetry Services Advanced Messaging Network Courtesy of SkyTel ......6-7
Figure 6-3 Example of an Omni-Directional Antenna and Resulting Coverage Pattern ..........6-10
Figure 6-4 Example of a Yagi Directional Antenna and Resulting Coverage Pattern ..............6-10
Figure 7-1 Examples of Capacitor Controllers from Several Manufacturers..............................7-1
Figure 7-2 Beckwith Electric’s M-2501B Autodaptive® Capacitor Control (left), M-2937
CAMP™ Remote Communication Module (middle) and M-2980 CAMP™ Utilinet®
Remote Communication Module (right) Courtesy of Beckwith Electric ..............................7-3
Figure 7-3 Cannon Technologies’ CBC-5000 (left) and CBC-7000 (right) Remote Power
Factor Control Courtesy of Cannon Technologies .........................................................7-4
Figure 7-4 S&C Electric’s Intellicap® Automatic Capacitor Control Courtesy of S & C
Electric ...............................................................................................................................7-4
Figure 7-5 S&C Electric’s Intellicap PLUS® Automatic Capacitor Control Courtesy of S &
C Electric............................................................................................................................7-5
Figure 7-6 A Fisher Pierce AutoCap™ Series 4400 Capacitor Control Courtesy of Fisher
Pierce / Joslyn Hi-Voltage ..................................................................................................7-6
Figure 7-7 A Fisher Pierce AutoCap™ Series 4500 Capacitor Control Courtesy of Fisher
Pierce / Joslyn Hi-Voltage ..................................................................................................7-6
Figure 7-8 ProCap™ 150T Capacitor Controller by Maysteel LLC Courtesy of Maysteel
LLC.....................................................................................................................................7-7
Figure 7-9 MicroCap (left) and MiniCap (right) Capacitor Switching Controllers from QEI,
Inc. Courtesy of QEI Inc. ....................................................................................................7-8
Figure 7-10 Capacitor Switching Controller eCAP-9040, QEI Inc. Courtesy of QEI Inc. ...........7-8
Figure 7-11 Components of a Basic SCADA System ................................................................7-9
Figure 7-12 Example of Basic SCADA Based Centralized Capacitor Control Using a
Master Station and a Dedicated Capacitor Control Server ................................................7-9
Figure 7-13 The ISO Seven-Layer, Open Systems, Interconnection Model ............................7-12
Figure 7-14 DNP3 Implementation Using the Enhanced Performance Architecture (EPA)
Model ...............................................................................................................................7-13
Figure 8-1 Example of Multiple Interfaces to Single Capacitor Control System.........................8-1
Figure 8-2 Example Screen from WinMon® Graphical User Interface Courtesy of S&C
Electric Company ...............................................................................................................8-4
Figure 9-1 Arrester Lead Length................................................................................................9-2
Figure 9-2 Example of Considerable Lead Length on a Riser Pole ...........................................9-3
Figure 9-3 Example of Almost Zero Lead Length on a Riser Pole.............................................9-4

xxii
Figure 9-4 Simulation of Protection Provided by Arresters at Adjacent Poles Only...................9-5
Figure 9-5 Blown Arrester with a Dangling Ground Lead ..........................................................9-6
Figure 9-6 Example of Switched Capacitor Bank Configuration ................................................9-7
Figure 9-7 CPT Secondary Voltage for Scenarios with and Without Secondary Arrester
and Ground Loop ...............................................................................................................9-9
Figure 9-8 Voltage at the Controller Terminals for Scenarios with and Without Secondary
Arrester and Ground Loop ...............................................................................................9-10
Figure 9-9 Example Configuration Using Shielded Control Cable ...........................................9-13
Figure 9-10 Ground Loop Created by Grounding the CPT Output and the Capacitor
Controller Neutral Terminal ..............................................................................................9-14
Figure 9-11 Cooper Power Systems Storm Trapper H.E. Secondary Surge Arrester
Courtesy of Cooper Power Systems ................................................................................9-16
Figure 9-12 Axiomatic 120Vac Surge Protector Courtesy of Advanced Surge Suppressor ....9-16
Figure 9-13 Example Configuration for Surge Protection Covering Incoming Lines for All
Surge Modes....................................................................................................................9-17
Figure 9-14 Approximate Size Relationship of Meter Socket and Typical Auxiliary Low-
Side Surge Protection (Note: Actual sizes will vary depending on what equipment is
used) ................................................................................................................................9-18
Figure 10-1 Capacitor Bank with a Blown Fuse (EPRI 1001691 2002) ...................................10-1
Figure 10-2 Capacitor Unit with a Failed Element ...................................................................10-5
Figure 10-3 Fuse Curves with Capacitor Rupture Curves .......................................................10-7
Figure 10-4 Comparison of Grounded-wye and Ungrounded-wye Banks During a Failure
of One Unit .......................................................................................................................10-8
Figure 10-5 Outrush from a Capacitor to a Nearby Fault.......................................................10-10
Figure 10-6 Outrush as a Function of the Resistance to the Fault for Various Size
Capacitor Banks (The sizes given are 3-phase kVAR; the resistance is the
resistance around the loop, out and back; the distances are to the fault) ......................10-13
Figure 10-7 Damaged Fuse Tubes from Loose Connections Courtesy of C. W. (Charlie)
Williams at Progress Florida ..........................................................................................10-17
Figure 10-8 Infrared Thermovision Scan of Cutouts Tested with 83 Amps of Current
Courtesy of C. W. (Charlie) Williams at Progress Florida ..............................................10-17
Figure 11-1 Waveform and Harmonic Spectrum of Typical 6-Pulse ac Motor Drives..............11-3
Figure 11-2 Harmonic Resonance ...........................................................................................11-4
Figure 11-3 Tuned Harmonic Filter ..........................................................................................11-6
Figure 11-4 Example Capacitor Switching Transient ...............................................................11-7
Figure 11-5 Scenario for Magnified Transients........................................................................11-8
Figure 11-6 Example of a Transient Magnified to Individual Customers .................................11-8
Figure 11-7 Effect of Capacitor-Switching Transient on the Direct Current Bus of an
Adjustable Speed Drive..................................................................................................11-10
Figure 11-8 Transient Caused by Synchronous Switching of a Capacitor.............................11-12
Figure 11-9 Telephone Influence Factor (TIF) Curve ............................................................11-13
Figure 11-10 GE Flicker Curve ..............................................................................................11-14

xxiii
LIST OF TABLES

Table 2-1 Substation versus Feeder Capacitors........................................................................2-4


Table 2-2 Common Capacitor Unit Ratings ...............................................................................2-5
Table 2-3 Maximum Permissible Power-Frequency Voltages ...................................................2-7
Table 2-4 Expected Transient Overcurrent and Overvoltage Capability....................................2-7
Table 2-5 Maximum Ambient Air Temperatures for Capacitor Application ................................2-9
Table 2-6 Percent Voltage Rise for Various Conductors and Voltage Levels (Impedance
is for all-Aluminum Conductors with GMD=4.8 feet) ........................................................2-12
Table 3-1 Substation Versus Feeder Capacitors .....................................................................3-12
Table 6-1 Frequency Bands for Typical Applications (Young 1999) ..........................................6-2
Table 6-2 Pro’s and Con’s of Radio Network Ownership...........................................................6-4
Table 10-1 Fusing Recommendations for ANSI Tin Links From One Manufacturer
[Cooper Power Systems, 1990] .......................................................................................10-3
2
Table 10-2 I t Comparisons on a 3-phase, 1,200-kVAR Bank at 12.47 kV (Iload=55.6 A)........10-11
2
Table 10-3 I t Comparisons on a 3-phase, 600-kVAR Bank at 12.47 kV (Iload=27.8 A) ......10-12
Table 10-4 Example Fuse Application Guidelines for a 12.47/7.2-kV System.......................10-19
Table 12-1 Estimated Benefits from Instituting Automated Capacitor Control on the
Kansas City Power & Light Distribution Systems .............................................................12-5
Table 12-2 Estimated Cost of Instituting Automated Capacitor Control on the Kansas
City Power & Light Distribution Systems ..........................................................................12-5

xxv
1
PROJECT OVERVIEW

The EPRI Capacitor Reliability Study

Utilities have a substantial investment in distribution line capacitors. These investments are
justified based on certain derived benefits to the power delivery system, the utilities, and the
end-users. When capacitors are not available due to some failure or operating error (or are
otherwise off-line), the anticipated benefits will not be achieved. Experience at utilities reveals
that capacitors are unavailable for operation too frequently. This project series was established,
therefore, to improve capacitor reliability. Initial scoping helped researchers identify and
prioritize several issues affecting the overall reliability of capacitors. EPRI’s capacitor reliability
study spans several years, from 2002 through the present. Each year a report is prepared dealing
with a different aspect of capacitor reliability. Reports from previous years have covered:
• Utility Survey and Literature Search (2002): A utility survey and literature search to assess
the issues related to the reliability of switched capacitor banks used in distribution systems
(EPRI 1001691).
• Fusing and Transmission Support (2003): An investigation of the causes of nuisance fuse
operations on capacitor banks, utility practices for providing transmission-level VAR support
with distribution capacitors, and assessments of additional utility needs (EPRI 1002154).
• Grounding and Lightning Protection of Capacitor Controllers (2004): Investigate the two
primary factors influencing the magnitude of surges reaching capacitor controllers and
provided controller mounting and wiring configurations for minimizing surge magnitude.
The first recommendation involved the physical location at which the capacitor controller
should be mounted with regard to the control power transformer (CPT) from which it draws
power. The second recommendation involved grounding considerations for the controller
supply power (EPRI 1008573).

This year’s report, 2005, examines automating switched capacitors at the distribution level. This
guide attempts to provide utility engineers with the background needed to sufficiently understand
automated capacitor controls and ways they could be applied to his or her distribution system.
This guide discusses commonly applied capacitor control schemes, including locally applied
control and centralized control schemes. The reader is presented with a variety of capacitor
control equipment from several prominent manufacturers in this area. This guide also discusses
the issues of system integration, capacitor protection, control schemes, and capacitor-related
power quality issues.

1-1
Project Overview

Background

There is considerable industry activity in applying distribution feeder capacitors. Automated line
capacitors are being added and operated by many utilities. Automation and communication
technologies are more advanced, more readily available, and more reasonably priced than ever
before. These advancements in automation control and communication allow utilities to operate
switched distribution capacitors in a manner that has never before been possible. Utilities are
using capacitors in a variety of ways—to supplement transmission VARs, as substitutes for
substation capacitors, to manage distribution voltage profiles, and to reduce line losses.
Communication technology allows centralized control of distribution capacitors as if they were
substation banks. This provides the added benefit of having the capacitors located closer to the
loads they service, thereby further improving their operating efficiency.

Distribution line capacitors provide tremendous benefits to distribution system performance by


providing VARs at or near the VAR-consuming loads, and they do this at a low cost. The main
benefits that capacitors provide are:
• Reduced Losses and Increased Capacity: By canceling the reactive power to motors and
other loads with low-power factor, capacitors decrease the line current. Reduced current frees
up capacity. Reduced current also significantly lowers I2R line losses.
• Reduce Voltage Drop: Capacitors provide a voltage boost that cancels part of the drop
caused by system loads. Switched capacitors serve to regulate voltage on a circuit, having an
ancillary benefit of reducing the number of operations on voltage regulators, both line (and to
a lesser to degree) substation regulators and load-tap-changers (LTCs). This reduces
maintenance costs on regulators and LTCs.
• Reduced Cost of Production or Cost of Purchased Power: Because line capacitors provide
VARs, generators no longer have to produce VARs, thus capacity is freed up to produce
more real power. (In addition, transmission and distribution lines no longer have to transport
those VARs.)

A typical switched capacitor bank installation is shown in Figure 1-1. Although Figure 1-1 only
shows the capacitor assembly near the pole top, the capacitor controller is mounted lower on the
pole, approximately 10 ft (3 m) above the ground. There are many types of controllers on the
market, with many different configurations.

1-2
Project Overview

Figure 1-1
Example of Switched Capacitor Bank Configuration

The capacitor controller is really the backbone of the automated switched capacitor system. Both
local control schemes and centralized control schemes utilize a local capacitor controller. At the
most basic level, the controller provides the interface to the capacitor switch, telling it when to
open and close. In local control schemes, the controller provides the switching logic. In central
control schemes, the controller 1) houses and interprets the signals provided by the data radio, 2)
provides switching override functions based on local conditions, and 3) provides switching logic
in the event that communication with the central station is lost.

Properly applied and controlled capacitors offer many benefits to the distribution system.
Capacitors provide both energy savings and capacity savings, which can increase revenue and
defer system upgrades. By serving distribution-level VAR needs on the feeder close to the load,
capacitors help reduce VAR flow at all levels of the utility system—generation, transmission,
and distribution. And by reducing VAR flow, properly applied capacitors will reduce wear and
tear on equipment at each level of the utility system.

Control Strategies

Automated control of switched capacitor banks takes two primary forms—locally controlled
automation and centrally controlled automation. In local control schemes, the control logic
resides at the capacitor location via an intelligent capacitor controller. The controller measures
line parameters, such as voltage or VARs, and makes switching decisions based on these

1-3
Project Overview

parameters. There are also less-sophisticated control schemes based on time, date, or
temperature. More information on the various control schemes is available in Chapter 4, Control
Strategies. Centrally controlled automation systems also use local capacitor controllers but in a
somewhat different role. In centrally controlled systems, switching decisions are made by a
master controller and then sent to the local capacitor controller, which tells the local switch to
either open or close. The capacitor controller may measure line parameters and transmit that
information back to the central station master (although not all central control schemes
incorporate local measurements). Some centralized control schemes make use of feeder data
gathered at the substation rather than at the capacitor location. In centralized control schemes, the
local capacitor controller provides local overrides to prevent the capacitor from switching, if
doing so would cause the line voltage to move out of ANSI C84.1 specifications. The local
controller also provides switching control intelligence in the event that communication is lost
between the central station and the local controller.

SCADA and Communications

Capacitor automation via centralized control utilizes supervisory control and data acquisition
(SCADA) technology to provide both overall data collection and control of the switched
capacitor banks. Most utilities have at least some type of SCADA system as part of their
operational infrastructure, although the extent to which utilities make use of SCADA varies
greatly. When trying to integrate new capacitor automation functionality into existing SCADA
systems, the challenge is to provide a method for the new system to interface with the existing
system. Depending on the system in place, newly added capacitor control functionality may
reside on a separate server in parallel with the existing SCADA master station. Other
installations may add capacitor control software directly into the master station. For some
utilities, this will be their first foray into SCADA; so they are unencumbered by compatibility
issues.

There are several technologies currently in use for communicating with the capacitor controllers.
Some offer one-way communication while others offer two-way communication. With one-way
communication, commands can be dispatched to the capacitor controllers in the field, but there is
no communication from the field back to the control center. Two-way communications offer data
flow, both from the command center to the field units and from the field units back to the
command center. The technologies used for centralized capacitor control communications
include:
• 900-MHz Radio
• Pager systems
• Cellular phone systems
• Cellular telemetric systems
• VHF radio

1-4
Project Overview

Project Objectives

The primary focus of this guide is to provide the distribution engineer with the necessary
information to evaluate applying a switched capacitor automation scheme to his or her
distribution system. This guide provides a detailed discussion of the all the key aspects of
distribution capacitor automation, including:
• Control Schemes: VAR, voltage, current, time, temperature, date, and combination control
programs
• Control Intelligence Location: Local control, central coordinated control, local control with
central station override
• Supervisory Control and Data Acquisition (SCADA) Systems: Components commonly found
in SCADA-based capacitor control systems, with examples cited from prominent
manufacturers
• Voltage and Current Measurements: Information on line parameters typically measured and
the potential for modern capacitor controllers to gather and report a wide array of line data to
aid distribution engineers in investigations beyond VAR management
• Capacitor Sizing and Placement: Detailed information size and placement of capacitor
banks on the distribution system
• Capacitor Installation Protection: Detailed information on proper application of fuses to
protect capacitor banks, with additional information regarding protecting capacitor
controllers from line surges and lighting strikes

1-5
2
CAPACITOR SIZING AND PLACEMENT

Introduction

Capacitors provide tremendous benefits to distribution system performance. Most noticeably,


capacitors are used to achieve the following benefits:
• Reduce Avoidable Losses and Free-up Capacity: By canceling the reactive power to motors
and other loads with low-power factor, capacitors decrease the line current. Reduced current
frees up capacity, thus the same circuit can serve more load. Reduced current also
significantly lowers the I2R line losses.
• Improved Voltage Profile: Capacitors provide a voltage boost that cancels part of the drop
caused by system loads. Switched capacitors can regulate voltage on a circuit.

If applied properly and controlled, capacitors can significantly improve the performance of
distribution circuits. But if not properly applied or controlled, the reactive power from capacitor
banks can create losses and can also create high voltages. The most danger of overvoltage is
under light load. Good planning helps ensure that capacitors are sited properly. Compared to
simple controllers (like a time clock), more sophisticated controllers (such as two-way radio with
monitoring) reduce the risk of improperly controlling capacitors.

Capacitors work their magic by storing energy. Capacitors are simple devices—two metal plates
sandwiched around an insulating dielectric. When charged to a given voltage, opposing charges
fill the plates on either side of the dielectric. The strong attraction of the charges across the very
short distance that separates them creates a tank of energy. Capacitors oppose changes in voltage.
It takes time to fill up the plates with charge; and once charged, it takes time to discharge the
voltage.

On ac power systems, capacitors do not store their energy for very long—just for one half-cycle.
Each half-cycle, a capacitor charges up and then discharges its stored energy back into the
system. The net real power is zero. Capacitors provide power just when reactive loads need it.
When a motor with low-power factor needs power from the system, the capacitor is there to
provide it. Then in the next half-cycle, the motor releases its excess energy, and the capacitor is
there to absorb it. Capacitors and reactive loads repeatedly exchange this reactive power. This
benefits the system because that reactive power (and extra current) does not have to be
transmitted from the generators all the way through many transformers and many miles of lines,
as the capacitors can provide the reactive power locally. This frees up the lines to carry real
power—power that actually performs work.

2-1
Capacitor Sizing and Placement

Capacitor units are made of series and parallel combinations of capacitor packs or elements put
together as shown in Figure 2-1. Capacitor elements have sheets of polypropylene film, less than
1-mil thick, sandwiched between aluminum-foil sheets. Capacitor dielectrics must withstand
voltage on the order of 2000 V/mil (78 kV/mm). No other medium-voltage equipment has such
high-voltage stress. An underground cable for a 12.47-kV system has insulation that is at least
0.175 inches (4.4 mm) thick. A capacitor on the same system has an insulation separation of only
0.004 inches (0.1 mm).

Courtesy of General Electric

Figure 2-1
Capacitor Components

2-2
Capacitor Sizing and Placement

Utilities often install substation capacitors, as well as capacitors at points on the distribution
feeders. Most feeder capacitor banks are pole-mounted, which is the least expensive way to
install distribution capacitors. Pole-mounted capacitors normally provide 300-3600 kVAR at
each installation. Many capacitors are switched, based either on a local controller or from a
centralized controller through a communication medium. A line capacitor installation has the
capacitor units in addition to other components, possibly including arresters, fuses, a CPT,
switches, and a controller (see Figure 2-2 for an example).

Figure 2-2
Overhead Line Capacitor Installation

While most capacitors are pole-mounted, some manufacturers provide pad-mounted capacitors.
As more circuits are put underground, the need for pad-mounted capacitors will grow.
Pad-mounted capacitors contain capacitor cans, switches, and fusing in a dead-front package that
follows standard, pad-mounted enclosure integrity requirements (ANSI C57.12.28-1998). These
units are much larger than pad-mounted transformers, so they must be sited more carefully to
avoid complaints. The biggest obstacles are cost and aesthetics. The main aesthetic-related
complaint is that pad-mounted capacitors are large. Customers complain about the intrusion and
the appearance of such a large structure.

Substation capacitors are typically offered as open-air racks. Normally elevated to reduce the
hazard, individual capacitor units are stacked in rows to provide large quantities of reactive
power. All equipment is exposed. Stack racks require a large substation footprint and are
routinely engineered for the given substation. Manufacturers also offer metal-enclosed

2-3
Capacitor Sizing and Placement

capacitors, where capacitors, switches, and fuses (normally current-limiting) are all enclosed in a
metal housing.

Substation capacitors and feeder capacitors both have their uses. Feeder capacitors are closer to
the loads—capacitors closer to loads more effectively release capacity, improve voltage profiles,
and reduce line losses. This is especially true on long feeders that have considerable line losses
and voltage drop. Table 2-1 highlights some of the differences between feeder and station
capacitors. Substation capacitors are better when more precise control is needed. System
operators can easily control substation capacitors wired into a SCADA system to dispatch VARs
as needed. Modern communication and control technologies applied to feeder capacitors have
reduced this advantage. Operators can control feeder banks with communications just like station
banks, although some utilities have found the reliability of switched feeder banks to be less than
desired. Further, the best times for switching in VARs, as needed by the system, may not
correspond to the best time to switch in the capacitor for the circuit on which it is located.

Table 2-1
Substation versus Feeder Capacitors

Feeder Capacitors

Advantages Disadvantages
• Reduces line losses • More difficult to control reliably
• Reduces voltage drop along the • Size and placement important
feeder
• Frees up feeder capacity
• Lower cost

Substation Capacitors

Advantages Disadvantages
• Better control • No reduction in line losses
• Best placement if leading VARs • No reduction in feeder voltage
are needed for system voltage drop
support
• Higher cost

Substation capacitors may also be desirable if a leading power factor is needed for voltage
support. With a leading power factor, moving this capacitor out on the feeder increases losses.
Substation capacitors cost more than feeder capacitors. This may seem surprising, but station
capacitors must be individually engineered; and the space they take up in a station is often
valuable real estate. Pole-mounted capacitors installations are more standardized.

Utilities normally apply capacitors on 3-phase sections. Capacitors are applied on single-phase
lines as well, but this is less common. Application of 3-phase banks downstream of single-phase
protectors is also uncommon because of ferroresonance concerns. Most 3-phase banks are
connected, grounded wye on 4-wire, multi-grounded circuits. Some are connected in floating
wye. On three-wire circuits, banks are normally connected as a floating wye.

2-4
Capacitor Sizing and Placement

Most utilities also include arresters and fuses on capacitor installations. Arresters protect
capacitor banks from lightning overvoltages. Fuses isolate failed capacitor units from the system
and clear the fault before the capacitor fails violently. In high-fault-current areas, utilities may
use current-limiting fuses. Switched capacitor units normally have oil or vacuum switches in
addition to a controller. Depending on the type of control, the installation may include a CPT for
power and voltage sensing and, possibly, a current sensor. Because a capacitor bank has a
number of components, capacitors normally are not applied on poles with other equipment.

Properly applied capacitors provide a return on investment very quickly. Capacitors save
significant amounts of money in reduced losses. In some cases, reduced loadings and extra
capacity can also delay building more distribution infrastructure.

Capacitor Ratings

Capacitor units rated from 50-500 kVAR are available. Table 2-2 shows common capacitor unit
ratings. A capacitor’s rated kVAR is the kVAR at rated voltage. 3-phase capacitor banks are
normally referred to by the total kVAR on all three phases. Distribution feeder banks normally
have one, two, or (more rarely) three units per phase. Many common-size banks have only one
capacitor unit per phase.

Table 2-2
Common Capacitor Unit Ratings

Volts, rms Number

(terminal-to- kVAR of BIL, kV


terminal) phases
**
216 5, 7 1/2, 13 1/3, 20, and 25 1 and 3 30

**
240 2.5, 5, 7 1/2, 10, 15, 20, 25, and 50 1 and 3 30

**
480, 600 5, 10, 15, 20, 25, 35, 50, 60, and 100 1 and 3 30

2400 50, 100, 150, 200, 300, and 400 1 and 3 75, 95, 125, 150, and 200

2770 50, 100, 150, 200, 300, 400, and 500 1 and 3 75, 95, 125, 150, and 200

4160, 4800 50, 100, 150, 200, 300, 400, 500, 600, 1 and 3 75, 95, 125, 150, and 200

700, and 800

6640,7200,7620,7960, 50, 100, 150, 200, 300, 400, 500, 600, 700, 1 95, 125, 150, and 200
8320, 9540, 9960, and 800

2-5
Capacitor Sizing and Placement

Table 2-2 (continued)


Common Capacitor Unit Ratings

Volts, rms Number

(terminal-to- kVAR of BIL, kV


terminal) phases

11 400, 12 470, 13 280,


13 800, 14 400

15 125 50, 100, 150, 200, 300, 400, 500, 600, 700, 1 125, 150, and 200
and 800

19 920 100, 150, 200, 300, 400, 500, 600, 700, and 1 125, 150, and 200
800

20 800, 21 600, 22 800, 100, 150, 200, 300, 400, 500, 600, 700, and 1 150 and 200
23 800, 24 940 800

Source: IEEE Std. 18-2002

IEEE Std. 18 defines standards for capacitors and provides application guidelines. Capacitors
should not be applied when any of the following limits are exceeded (IEEE Std. 18-2002):
• 135% of nameplate kVAR
• 110% of rated rms voltage and crest voltage not exceeding 1.2 2 of rated rms voltage,
including harmonics but excluding transients
• 135% of nominal rms current, based on rated kVAR and rated voltage

Capacitor dielectrics must withstand high-voltage stresses during normal operation—on the order
of 2000 V/mil. Capacitors are designed to withstand overvoltage for short periods of time.
IEEE Std. 18-1992 allows up to 300 power-frequency overvoltages within the time durations in
Table 2-3 (without transients or harmonic content). New capacitors are tested with at least a
10-second overvoltage, either a dc-test voltage of 4.3 times the rated rms or an ac voltage of
twice the rated rms voltage (IEEE Std. 18-2002).

2-6
Capacitor Sizing and Placement

Table 2-3
Maximum Permissible Power-Frequency Voltages

Duration Maximum permissible


voltage (multiplying
factor to be applied to
rated voltage rms)

6 cycles 2.20
15 cycles 2.00
1s 1.70
15 s 1.40
1 min 1.30
30 min 1.25
continuous 1.10

Source: IEEE Std. 18-1992

Capacitors should withstand various peak voltage and current transients; the allowable peak
depends on the number of transients expected per year (see Table 2-4).

Table 2-4
Expected Transient Overcurrent and Overvoltage Capability

Probable Permissible peak transient Permissible peak transient


number of current (multiplying factor voltage (multiplying factor to
transients to be applied to rated rms be applied to rated rms
per year current) voltage)

4 1500 5.0
40 1150 4.0
400 800 3.4
4000 400 2.9
Source: IEEE Std. 18-1992

2-7
Capacitor Sizing and Placement

The capacitance of a unit in microfarads is:

2.65Q kvar
C uF = Eq. 2-1
VkV 2

where,

VkV = capacitor voltage rating, kV

QkVAR = unit reactive power rating, kVAR

Capacitors are made within a given tolerance. The IEEE standard allows reactive power to range
between 100-110% when applied at rated sinusoidal voltage and frequency (at 25°C case and
internal temperature) (IEEE Std. 18-2002). Older units were allowed to range up to 115%,
(ANSI/IEEE Std. 18-1992). Therefore, the capacitance also must be between 100-110% of the
value calculated at rated kVAR and voltage. In practice, most units are from +0.5 to +4.0%, and
a given batch is normally very uniform.

Capacitor losses are typically on the order of 0.07 to 0.15 watts/kVAR at nominal frequency.
Losses include resistive losses in the foil, dielectric losses, and losses in the internal discharge
resistor.

Capacitors must have an internal resistor that discharges a capacitor to 50 V or less within five
minutes of when the capacitor is charged to the peak of its rated voltage ( 2Vrms ). This resistor is
the major component of losses within a capacitor. The resistor must be low enough such that the
RC time constant causes it to decay in 300 seconds as:

300
50 −
≤e RC Eq. 2-2
2V

where,

V = capacitor voltage rating, volts


R = discharge resistance, ohms
C = capacitance, farads.
So, the discharge resistor must continually dissipate at least the following power in watts:

Q kvar ⎛ 35 .36 ⎞
Pwatts = − ln ⎜ ⎟ Eq. 2-3
113 .2 ⎝ V ⎠

where,

2-8
Capacitor Sizing and Placement

QkVAR is the capacitor rating (single- or 3-phase).


For 7.2-kV capacitors, the lower bound on losses is 0.047 W/kVAR.

Some utilities use a shorting bar across the terminals of capacitors during shipping and in
storage. The standard recommends waiting for five minutes to allow the capacitor to discharge
through the internal resistor.

Capacitors have very low losses, so they run very cool. But, capacitors are very sensitive to
temperature and are rated for temperatures lower than other power system equipment, such as
cables or transformers. Capacitors do not have load cycles like transformers—they are always at
full load. Also, capacitors are designed to operate at high dielectric stresses, so they have less
margin for degraded insulation. Standards specify an upper limit for application of 40 or 46°C,
depending on arrangement (see Table 2-5). These limits assume unrestricted ventilation and
direct sunlight. At the lower end, IEEE standard 18 specifies that capacitors shall be able to
operate continuously in a –40°C ambient.

Table 2-5
Maximum Ambient Air Temperatures for Capacitor Application

Ambient air
temperature—
°C

Mounting arrangement 4-h average*


Isolated capacitor 46
Single row of capacitors 46
Multiple rows and tiers of capacitors 40
Metal-enclosed or -housed 40
equipment

*The mathematical average of the four


consecutive highest hourly readings during the
hottest day expected at that location.
Source: IEEE Std. 18-2002

Released Capacity

In addition to reducing losses and improving voltage, capacitors release capacity. Improving the
power factor increases the amount of real power load the circuit can supply. Using capacitors to
supply reactive power reduces the amount of current in the line, so a line of a given ampacity can
carry more load. Figure 2-3 shows that capacitors release significant capacity, especially if the
original power factor is low. Figure 2-4 shows another way to view the extra capacity, as a
function of the size of capacitor added.

2-9
Capacitor Sizing and Placement

Figure 2-3
Released Capacity with Improved Power Factor

Figure 2-4
Extra Capacity as a Function of Capacitor Size

2-10
Capacitor Sizing and Placement

Voltage Support

Capacitors are constant-impedance devices. At higher voltages, capacitors draw more current
and produce more reactive power as:

I = IratedVpu, and QkVAR = QratedVpu2 Eq. 2-4

where Vpu is the voltage in per-unit of the capacitor’s voltage rating. Capacitors applied at
voltages below their rating provide VARs in proportion to the per-unit voltage squared.

Capacitors provide almost a fixed voltage rise. The reactive current through the system
impedance causes a voltage rise in percent of:

Qkvar X L
Vrise = 2
Eq. 2-5
10 VkV ,l -l

where,

XL = positive-sequence system impedance from the source to the capacitor, ohms

VkV, l-l = line-to-line system voltage, kV

QkVAR = 3-phase bank rating, kVAR

While this equation is very good for most applications, it is not exactly right; because the
capacitive current changes in proportion to voltage. At a higher operating voltage, a capacitor
creates more voltage rise than the equation predicts.

Since the amount of voltage rise is dependent on the impedance upstream of the bank, putting the
capacitor at the end of the circuit will provide the voltage boost along the entire circuit. The best
location for voltage support depends on where the voltage support is needed. Figure 2-5 shows
how a capacitor changes the voltage profile along a circuit. Unlike a regulator, a capacitor
changes the voltage profile upstream of the bank.

2-11
Capacitor Sizing and Placement

104
With no load

Voltage rise from the capacitor


102

100
Voltage profile without the capacitor

100 With load

Voltage profile with the capacitor


98

96

Voltage profile without the capacitor

Figure 2-5
Voltage Profiles After Addition of a Capacitor Bank

Table 2-6 shows the percentage voltage rise from capacitors for common conductors at different
voltages. This table excludes the station transformer impedance but still provides a useful
approximation. Inductance does not change much with conductor size; the voltage change stays
the same over a wide range of conductor sizes. For 15-kV class systems, capacitors increase the
voltage by about 0.12% per mile per 100 kVAR per phase.

Table 2-6
Percent Voltage Rise for Various Conductors and Voltage Levels (Impedance is for all-
Aluminum Conductors with GMD=4.8 feet)

Percent voltage rise per mile with 100 kVAR per


phase

Line-to-line system voltage, kV


Conduct XL
or size Ω/mile 4.8 12.47 24.9 34.5

4 0.792 1.031 0.153 0.038 0.020


2 0.764 0.995 0.147 0.037 0.019
1/0 0.736 0.958 0.142 0.036 0.019
4/0 0.694 0.903 0.134 0.034 0.017
350 0.656 0.854 0.127 0.032 0.017
500 0.635 0.826 0.122 0.031 0.016

2-12
Capacitor Sizing and Placement

750 0.608 0.791 0.117 0.029 0.015

On switched capacitor banks, the voltage change constrains the size of banks at some locations.
Normally, utilities limit the voltage change to 3-4%. On a 12.47-kV circuit, a 3-phase
1200-kVAR bank boosts the voltage 4% at about 8 miles from the substation. To keep within a
4% limit, 1200-kVAR banks must only be used within the first eight miles of the station.

Reducing Line Losses

One of the main benefits of applying capacitors is that they can reduce distribution line losses.
Losses come from current flow against the resistance of conductors. Some of that current
transmits real power, but some flows to supply reactive power. Reactive power provides
magnetizing for motors and other inductive loads. Reactive power does not spin kWh meters and
performs no useful work, but it must be supplied. Using capacitors to supply reactive power
reduces the amount of current in the line. Since line losses are a function of the current squared,
I2R, reducing reactive power flow on lines significantly reduces losses.

Engineers employ the widely used the “2/3’s” rule for sizing and placing capacitors to optimally
reduce losses. Neagle and Samson (1956) developed a capacitor placement approach developed
for uniformly distributed lines and showed that the optimal capacitor location is the point on the
circuit where the reactive power flow equals half of the capacitor VAR rating. From this, they
developed the “2/3’s rule” for selecting and placing capacitors. For a uniformly distributed load,
the optimal size capacitor is 2/3 of the VAR requirements of the circuit. The optimal placement
of this capacitor is 2/3 of the distance from the substation to the end of the line. For this optimal
placement for a uniformly distributed load, the substation source provides VARs for the first 1/3
of the circuit, and the capacitor provides VARs for the last 2/3 of the circuit (see Figure 2-6).

2-13
Capacitor Sizing and Placement

Figure 2-6
Optimal Capacitor Placement Using the “2/3’s” Rule

A generalization of the 2/3’s rule for applying n capacitors to a circuit is to size each one to
1/(2n+1) of the circuit VAR requirement. They are applied, equally spaced, starting at a distance
of 2/(2n+1) of the total line length from the substation and added to the rest of the units at
intervals of 2/(2n+1) of the total line length. The total VARs supplied by the capacitors is
2n/(2n+1) of the circuit’s VAR requirements. So to apply three capacitors, each should be spaced
to 2/7 of the total VARs needed, and located at per-unit distances of 2/7, 4/7, and 6/7 of the line
length from the substation.

Grainger and Lee (1981) provide an optimal, yet simple, method for placing fixed capacitors on
a circuit with any load profile, not just a uniformly distributed load. With the Grainger/Lee
method, the reactive load profile of a circuit is used to place capacitors. The basic idea is, again,
to locate banks at points on the circuit where the reactive power equals one half of the capacitor
VAR rating. With this ½-kVAR rule, the capacitor supplies half of its VARs downstream, and
half are sent upstream. The basic steps of this approach are:
1. Select a Size: Choose a standard size capacitor. Common sizes range from 300-1200 kVAR,
with some sized up to 2400 kVAR. If the bank size is 2/3 of the feeder requirement, only one
bank is needed. If the size is 1/6 of the feeder requirement, five capacitor banks are needed.
2. Locate the First Bank: Start from the end of the circuit. Locate the first bank at the point on
the circuit where VAR flows on the line are equal to half of the capacitor VAR rating.
3. Locate Subsequent Banks: After a bank is placed, reevaluate the VAR profile. Move
upstream until the next point where the VAR flow equals half of the capacitor rating.
Continue placing banks in this manner until no more locations meet the criteria.

There is no reason to have to stick with the same size banks. A 300-kVAR bank could be placed
where the VAR flow equals 150 kVAR, then a 600-kVAR bank could be applied where the VAR
flow equals 300 kVAR. Finally, a 450-kVAR bank could be applied where the VAR flow equals
225 kVAR. Normally, it is more efficient to use standardized bank sizes, but different size banks
at different portions of the feeder might help with voltage profiles.

The ½-kVAR method works for any section of line. If a line has major branches, capacitors can
be applied along the branches using the same method. One may start at the end, move upstream,
and apply capacitors at points where the line’s kVAR flow equals half of the kVAR rating of the
capacitor. This will also work for lines that already have capacitors (it doesn’t optimize the
placement of all of the banks, but it optimizes placement of new banks). For large industrial
loads, the best location is often going to be right at the load.

Figure 2-7 shows the optimal placement of 1200-kVAR banks on an example circuit. Since the
end of the circuit has a reactive load above the 600-kVAR threshold for sizing 1200-kVAR
banks, the first capacitor would be applied at the end of the circuit. (The circuit at the end of the
line could be one large customer or could be branches off of the main line.) The second bank
would go near the middle. The circuit has an express feeder near the start. Another 1200-kVAR
bank could go in just after the express feeder, but that would not gain anything. The two

2-14
Capacitor Sizing and Placement

capacitors total 2400 kVAR and the feeder load is 3000 kVAR. Another 600-kVAR capacitor
would really be needed to zero out the VAR flow before it gets to the express feeder.

Figure 2-7
Placement of 1200-kVAR Banks Using the ½-kVAR Method

Fortunately, capacitor placement and sizing does not have to be exact. Quite good loss reduction
occurs, even if optimum sizing and placement cannot be achieved. Figure 2-8 shows the loss
reduction for one fixed capacitor on a circuit with a uniform load. The 2/3’s rule specifies that
the optimum distance is 2/3 of the distance from the substation and 2/3 of the circuit’s VAR
requirement. As long as the size and location are somewhat close (within 10%), the not-quite-
optimal capacitor placement provides almost as much loss reduction as the optimal placement.

2-15
Capacitor Sizing and Placement

Figure 2-8
Sensitivity to Losses of Placing One Capacitor on a Circuit with a Uniform Load

The voltage impacts of capacitors must be considered. Under light load, the capacitors should be
checked to ensure they have not raised the voltages above allowable standards. If voltage limits
are exceeded, the size or number of the capacitor banks should be reduced until voltage limits are
no longer exceeded. If additional loss reduction is desired, switching banks should be considered
as discussed below.

Energy Losses

The average reactive loading profile should be used to optimally size and place capacitors for
energy losses. Using the peak-load case, the ½-kVAR method optimizes losses during the peak
load. On a low-flow case with the average reactive load, the ½-kVAR method or the 2/3’s rule
optimizes energy losses. This leads to more separation between banks and less kVARs applied
than if we optimize for peak losses.

If an average system case is not available, it can be estimated by scaling the peak load case by
the reactive load factor, (RLF):

Average kvar Demand


RLF = Eq. 2-6
Peak kvar Demand

The RLF is similar to the traditional-load factor, except that it only considers the reactive portion
of the load. If no information is available on the RLF, the total-load factor can be used.
Normally, the RLF is higher than the total-load factor. Figure 2-9 shows an example of power
profiles. The real power (kW) fluctuates significantly more than the reactive power (kVAR).

2-16
Capacitor Sizing and Placement

Figure 2-9
Example of Real and Reactive Power Profiles on a Residential Feeder on a Peak Summer
Day with 95% Air Conditioning (Data from East Central Oklahoma Electric Cooperative,
Inc.)

Grounded versus Ungrounded

Capacitor banks can be grounded or ungrounded. The advantages and disadvantages of a


grounded capacitor bank include:
• Unit Failure and Fault Current: If a unit fails, the faulted phase draws full-fault current.
This allows the fuse to blow quickly but requires fuses to be rated for the full-fault current.
• Harmonics: The grounded-wye bank can attract zero-sequence harmonics (balanced 3rd, 9th,
15th, …). This problem is often found in cases of telephone interference.
Advantages and disadvantages of the floating-wye, ungrounded banks include:
• Unit Failure: If a unit fails, the floating neutral shifts to the phase voltage of the failed unit.
Then, the neutral shift stresses the remaining capacitors with 173% voltage.
• Fault Current: When one unit fails, the circuit does not draw full fault current—it is a high-
impedance fault. This is an advantage in some capacitor applications.
• Harmonics: Because the bank is ungrounded, zero-sequence harmonics (balanced 3rd, 9th,
15th, …) cannot flow to ground through the capacitor.

The response of the floating-wye configuration deserves more analysis. During a progressive
failure, when one series group shorts out, the shift of the neutral relieves the voltage stress on the
remaining series groups. In the example in Figure 2-10, for a floating-wye bank with half of the

2-17
Capacitor Sizing and Placement

series sections shorted, the line-to-neutral voltage becomes 0.75 per unit. The remaining
elements normally see 50% of the line-to-neutral voltage, but now they see 75% (1.5 per unit, so
the current is also 1.5 time normal). The reduction in voltage stress due to the neutral shift. The
excess heating at the failure point increases the risk of gas generation and case rupture. When
one pack fails, the fuse (or other protection) should trip quickly.

Floating-wye configurations are best applied with neutral detection. A potential transformer
measuring voltage between the floating neutral and ground can detect a failure of one unit. When
one unit fails, a relay monitoring the neutral potential transformer (PT) should trip the
capacitor’s oil or vacuum switch (obviously, this only works on switched banks).

Figure 2-10
Comparison of Grounded-wye and Ungrounded-wye Banks During a Failure of One Unit

Standard utility practice is to ground banks. On three-wire systems, utilities use both ungrounded
wye and delta configurations. Capacitor banks on industrial systems are normally ungrounded.

Utilities often use two-bushing capacitors, even though most also use a grounded neutral. Having
two bushings allows crews to convert capacitor banks to a floating neutral configuration if
telephone interference is a problem.

Utilities universally ground capacitor cases on pole-mounted capacitors (even though it is not
strictly required by the National Electrical Safety Code, IEEE C2-1997). In rare cases, banks
with single-bushing capacitors are floated when it becomes necessary to convert a bank to a
floating wye.

2-18
Capacitor Sizing and Placement

Impact of Switching on Capacitor Sizing and Placement

Utilizing automated switching schemes allows capacitor banks to be switched both frequently
and economically. Frequent economic switching enables the use of the new capacitor layouts that
were not previously economical when the capacitors had to be manually switched by a technician
in the field. For example, consider a feeder with a heavy but uniform load. Using a conventional
fixed bank, the capacitor would be sized at 2/3 the heavy kVAR load and located 2/3 the length
of the feeder from the substation. However, the greatest savings are achieved by using several
smaller banks distributed along the feeder to reduce, to the greatest extent possible. the reactive
current flow in the conductor. This configuration allows the total capacitor kVAR to more
closely equal the total load kVAR. Typically, no more than four switched capacitor banks are
used on a single feeder (Marx 2003).

Migrating to advanced capacitor control schemes, particularly centralized control with two-way
communication capability, can reduce the size and/or number of capacitor bank installations
required to efficiently operate the distribution system. These benefits stem from two key
operational improvements realized through centralized control: 1) installed capacitors are used
more efficiently, and 2) capacitor failures are discovered and repaired in a timely fashion,
resulting in a greater percentage of installed capacitors being available for use (Goeckeler 1999;
Girotti, Tweed et al. February 1990). If more installed capacitors are in service, then the total
number of installed capacitors can be reduced or maintained without growth, while still meeting
the VAR needs of the distribution system. As Goeckeler points out, during a four-year rollout of
capacitor automation, Kansas City Power & Light (KCPL) was able to remove 5 MVAR of
capacitors from a section of their system. The section, which comprised approximately 36% of
KCPL’s metropolitan load, experienced a reactive power demand growth of 32.5 MVAR over
that period. KCPL was still able to remove the 5 MVAR of capacitors while improving the
system power factor from 98.3% to 98.7% due to more efficient VAR use.

Switched Capacitor Bank Equipment Mounting Considerations

Switched capacitor banks tend to be pretty busy installations, due to the amount of equipment
that needs to be mounted on the pole. The majority of the equipment is located near the pole top,
with just the controller located lower on the pole, as shown in Figure 2-11.

2-19
Capacitor Sizing and Placement

Figure 2-11
Example Switched Capacitor Bank Installation
Courtesy of Donald M. Parker at Alabama Power

The height above the ground at which the controller is mounted varies by utility, and it is
generally in the range of 5-15 ft. (1.5-4.5 m) as shown in Figure 2-12. The example installation
shown in Figure 2-11 has the controller mounted at the low end of this range. The controller is
more easily accessible to utility personnel when it is located in the low end of this range, but that
also makes it more susceptible to vandalism, snow plows, and other hazards. Because of these
issues, some utilities are adopting standards calling for capacitor controllers to be located 10- or
15-ft. up the pole. Vandalism is a particular concern for controllers with external antennas for
radio communication. The antennas tend to be particularly tempting for vandals, and small
increases in the mounting height can help dramatically lower the instances of damaged antennas.

Most capacitor controllers can be specified for mounting in a 4- or 6-jaw meter socket or
rack/case, mounted with connections made to the terminal strip. Utility installations tend to make
use of the meter socket mounting option. Chapter 5 contains more information on mounting the
capacitor controller.

2-20
Capacitor Sizing and Placement

Figure 2-12
Generic Switched Capacitor Bank Equipment Configuration

Near the pole top, things become a bit more crowded as the capacitors, switches, CPT, and
junction box are all mounted at roughly the same level. Often, all this equipment will be
supported by the capacitor rack, as shown in Figure 2-13. Protective devices, such as fused
cutouts and lightning arresters, are also located near the pole top, often just above the capacitor
rack, as shown in Figure 2-14.

2-21
Capacitor Sizing and Placement

Figure 2-13
Typical Layout of Pole-Top Equipment in a Switched Capacitor Installation
Courtesy of Donald M. Parker at Alabama Power

Figure 2-14
Example Location of Fuses and Lightning Arresters in a Switched Capacitor Installation
Courtesy of Donald M. Parker at Alabama Power

2-22
Capacitor Sizing and Placement

Optimal Capacitor Placement Computer Programs

Optimal capacitor placement (OCP) programs can aid the distribution engineer in finding the
best sites in a network to place capacitors. In general, OCP programs follow the standard
methodology of optimizing the size and location of capacitor banks such that capacitor cost and
energy losses are minimized while still meeting system constraints for all load conditions. In
determining the locations with the highest economic return, OCP programs consider several
economic factors, including the initial cost of the capacitor, annual maintenance cost of the
capacitor, and cost of real and reactive power losses. The OCP program then 1) suggests the set
of locations where capacitors should be placed, 2) recommends the size of the capacitor bank
that should be placed at each site; and 3) recommends whether or not a switched capacitor is
needed. Many OCP programs can also produce a switching schedule or recommend control set
points for installations where switched capacitors are recommended.

There are several software packages available that provide assistance in optimizing the size and
location of capacitor bank installations on distribution circuits. Optimal capacitor placement
programs are typically available as additional modules in load-flow programs. As such, they
require input from load-flow simulations to operate. Capabilities, data requirements, and analysis
methods vary among the software packages, although the final criterion for comparing different
capacitor sizes and placements tends to be economic in nature. The final evaluation comes down
to which capacitor configuration offers the best opportunity for cost savings.

There is a wide variety of software available for conducting capacitor placement studies,
including:
• Distribution Workstation from EPRI
• PSS/ADEPT from Power Technologies, Inc.
• CYMDIST from CYME International
• Technical 2000 from EDSA Micro Corporation
• ETAP from Operation Technology, Inc.

Optimal capacitor placement programs typically rely on load-flow algorithms as part of their
evaluative process. Loss reduction is the main criteria when evaluating different combinations of
capacitor size and placement. Therefore, optimal capacitor placement programs use the results of
various load flows to study the effects of different combinations of capacitor size and placement.
The monetary savings through loss reduction is then optimized versus the capital and operating
costs of installing the capacitor installation(s) to determine the optimal number, size, and
location of capacitor installations (ORNL/TM-2001/215 2002).

The primary input for optimal capacitor placement software tends to be a network model for the
circuit(s) under consideration. Once the model is built for use in the load flow, only a few other
data elements are required to perform the optimized capacitor placement. Typically, the
following additional data are required:

2-23
Capacitor Sizing and Placement

• Cost of Real Energy: The cost of real energy, usually stated in $/kWh. Note that the
currency type is not important as long as consistency is maintained across the variables
• Cost of Reactive Energy: Similar to the cost of real energy, except is accounts for reactive
energy
• Cost of Electrical Demand: The cost of electrical generation of capacity that would have to
be purchased to replace the system’s losses
• Cost of Reactive Demand: Similar to the cost of electrical demand, except it accounts for
reactive demand
• Discount Rate: Used to equate future savings and costs back to present value in the financial
analysis
• Inflation Rate: The yearly increase in the cost of money
• Evaluation Period: The time period that the capacitors have to produce monetary savings
that cancel their investment and maintenance costs
• Cost of Installation for Fixed and Switched Capacitors: The monetary investment needed to
install a fixed or switched bank
• Maintenance Cost for Fixed and Switched Capacitors: The monetary investment needed to
maintain operation of a fixed or switched bank

Although this list of data requirements appears to be extensive, many of the items listed can be
easily obtained or reasonably estimated. Nearly all the data required should already be available
for the network under consideration; therefore, performing an optimal capacitor placement study
often only requires a modest time investment, if the load-flow model is already complete.

2-24
3
AUTOMATION STRATEGIES

Best Use of Distribution VARs

The two primary roles of distribution capacitors are VAR support for the feeder or substation and
voltage support for the feeder. Ideally, the utility will have a flexible, global-control system at its
disposal. This would allow control of the distribution capacitors to be optimized for both VAR
support and voltage support at the feeder level, while still having the override capability to
switch to support for the substation or transmission system. This adaptability is usually available
through centralized capacitor control systems, but it is not usually found on systems that rely
solely on local controls.

The economic benefit of utilizing switched capacitors has three main components: 1) energy cost
savings, 2) demand reduction, and 3) increased revenue. Of these three components, demand
reduction usually makes up a portion of the cost savings and tends to comprise the majority of
the total cost benefits. Therefore, optimizing demand reduction should provide the quickest
return on investment for the system and make the best use of the distribution VARs available.
Once VARs are introduced on the line, they continue to load the circuit's conductors until they
are neutralized. Optimum VAR use is best achieved when VARs are eliminated early by locating
capacitors down line, at or near inductive loads (Marx 2003). However, line voltage regulation
cannot be ignored. Further, real-world constraints often dictate that effective capacitor control
means managing VAR reduction and voltage regulation on a near real-time basis. An ideal
capacitor control system provides control that can switch between voltage control and VAR
reduction to optimize performance of the line.

A study report from Florida Power & Light, "Effective VAR Control" by Fred Walker and
Michael Keightly, indicated that even diligently maintained time-clock schedules did not
perform as well as real-time VAR control for efficiently operating distribution circuits (RCCS
2005). While local capacitor control is highly effective at managing VARs at specific locations,
centralized capacitor control can effectively managing VARs across the whole distribution
system.

Conservation Voltage Reduction

Reducing the feeder voltage can reduce energy consumption and/or reduce peak demand on the
circuit. Many loads, such as light bulbs and other electrical devices, operate satisfactorily at
reduced voltages. Therefore, by reducing feeder voltage, utilities can still adequately serve their
customers while saving energy and increasing capacity (Milosevic and Begovic 2004). Voltage
reduction also has the added benefit of increasing the system voltage stability margin (Begovic,

3-1
Automation Strategies

Novosel et al. 2000). Voltage reduction is most effective on resistive loads, because the power
drawn by a resistive load decreases with the voltage squared. Circuits that are primarily
residential or commercial in nature tend to have a greater percentage of resistive loads than
industrial circuits. Lighting and resistive heating comprise a great deal of the resistive load found
on most distribution feeders. The power on incandescent lights varies with the voltage to the
power of about 1.6. For heating devices, which regulate to a temperature, the results are not as
good. Reducing the voltage simply increases the unit’s duty cycle, causing it to run longer to
reach the desired heat level.

Analyses of conservation voltage reduction (CVR) trials at several utilities have shown potential
energy savings between 0.5-1% for each 1% voltage reduction (Kirshner and Giorsetto 1984).
Through regression analysis of the feeders, Kirshner and Giorsetto found that residential energy
savings were 0.76% for each 1% reduction in voltage; commercial and industrial loads had
reductions of 0.99% and 0.41%, respectively (but the correlations between load class and energy
reduction were fairly small). Voltage reduction also tends to have the greatest demand reduction
in the first 15 minutes. However, voltage reduction often shows a declining impact over the next
several hours, as many loads begin running increased duty cycles (McCarthy and Josken). One
study found that during a 5% voltage reduction over a 4-hour time period, the demand on one
typical residential circuit initially dropped by 4% but diminished to a 3% drop by the end of the
time period (Priess and Warnock 1978).

CVR is often accomplished by altering the set points or control scheme of the LTC and/or
voltage regulators governing the voltage on the feeder. However, feeder voltage can only be
lowered so far without violating the ANSI C84.1-1989 voltage limit of 120V ± 5% (114-126V
on a 120 V scale). Thus, it may not be possible to lower the substation voltage without violating
ANSI 84.1-1989 at the far end of a feeder when the circuit exhibits a high voltage drop along its
length. Therefore, using capacitors for voltage support is one solution to this problem, as it
permits greater voltage reduction at the substation while still maintaining the necessary
tolerances at the end of the feeder. Using capacitors to flatten the voltage profile along the feeder
can aid voltage reduction schemes, even when maintaining minimum ANSI 84.1-1989 voltage is
not a concern. In these cases, the flattened voltage profile allows for greater voltage reduction,
leading to better demand reduction.

Implementing a voltage reduction scheme requires the use of switched capacitor banks, since
more capacitors are required during CVR dispatch than at other times during operation. Either
local control or centralized SCADA control can be used for this purpose. Local control for CVR
tends to make use of voltage-based control, with all of the controllers on the feeder coordinated
so that they turn on in the proper sequence, once the substation voltage is lowered. Centralized
capacitor control makes CVR easier to accomplish, as it tends to provide some level of global
capacitor control and coordination. Additionally, as centralized capacitor control becomes more
and more common, the industry is starting to see CVR moving from a contingency operation
scheme to one that is used more readily to provide certain operations benefits.

3-2
Automation Strategies

Optimizing Power Factor at the Substation

Capacitors can be installed on the low voltage bus in the substation rather than out on the
distribution circuit. Capacitors installed in the substation do not affect the voltage drop along the
feeder, but they do raise the voltage profile of the entire feeder. Since the entire feeder voltage is
raised by the substation capacitor bank, it is often necessary to switch at least a portion of the
bank in order to prevent an overvoltage condition during light-load situations. Substation
capacitor banks also do not affect the feeder power factor, since the reactive load on the feeder is
not influenced by the substation bank. (The same amount of reactive current needs to flow out to
the load, regardless of whether or not there is a capacitor in the substation). Therefore, the
primary function of substation-based capacitor banks is to supply reactive power to the feeder,
thus releasing substation and transmission-line capacity (Westinghouse Electric Corporation
1959).

Distribution Capacitors for Transmission VAR Support

Although most DCC strategies only consider benefits provided to the distribution system, it is
possible to provide VAR support to the transmission system via distribution level capacitors. The
continuing advancement of capacitor automation technology now provides an unprecedented
level of capacitor control. Real-time monitoring capabilities, coupled with the ability to rapidly
and automatically switch distribution capacitor banks, allow for applications in which the banks
can be switched on at a moment’s notice to provide reactive power support to the transmission
system.

The following presents several brief examples of the systems that some utilities are using to
enable transmission VAR support from distribution capacitors:

KCPL

Sources: “WinMon Interface and Capacitor Controls Help KCPL Stabilize System,”
EnergyLine Case Studies, December 2000. http://www.elsi.com/case2-1200.pdf
C. R. Goeckeler, ”KCPL Enables DNP by Finding Missing Communication Link,”
Utility Automation, C. R. Goeckeler, “Wireless Remote Monitoring and Control,”
DistribuTECH 2002, March 1, 2002.

Kansas City Power and Light (KCPL) began a capacitor automation program in 1994 which
heavily utilizes EnergyLine’s WinMon’s Graphical User Interface (GUI), coupled with
EnergyLine’s automatic capacitor controls and Telemetric MicroRTUs. The Telemetric RTUs
provide a cellular network communications link that can be integrated with existing capacitor
controllers. As such, the combination of capacitor controller and RTU allows for remote
reporting and control of the capacitor installation.

This system was put to the test when KCPL experienced an outage at their metropolitan area
generation station, which they relied heavily upon for voltage control. The generation station was
off-line during both the summer peaks of 1999 and 2000, and transmission system operators had

3-3
Automation Strategies

been concerned about predicted low-transmission voltage levels during those times of peak
system loading. To remedy the situation, the transmission system operators used the WinMon
interface to issue a request to turn on all available distribution capacitors to support system VAR
and voltage needs during the peak loading periods. Using EnergyLine’s WinMon query tool, the
system operators were able to identify 160 capacitor banks that were open and switch them
on-line. The WinMon query tool lead the operators through the selection process and allowed
them to poll the capacitors based on criteria such as whether or not the bank was switched in.
The WinMon interface also allowed the operators to switch capacitor control algorithms from
more traditional control (volt, VAR, volt/VAR, and so on) to a forced on setting with automatic
overvoltage protection. Furthermore, these changes could be made globally to all of the
automated distribution capacitors at once over a radio communication network. The net effect of
the high-level capacitor control capability was that KCPL was able to proactively meet the
operating requirements of their system while maintaining a high level of power quality for their
customers.

Idaho Power

Source: M. R. Trammel and K. D. McCarthy, “Capacitor Control Gives Voltage a Lift,”


Transmission & Distribution World, August 1, 1999. http://www.tdworld.com

In 1996, Idaho Power began a program to address deficiencies in their local area reactive power
consumption that resulted in low voltage on their distribution system. Additionally, they had a
problem maintaining adequate system reactive margins during peak demand periods. Since
distribution capacitors are not normally utilized for supplying reactive power to the transmission
system, Idaho Power felt that many of their distribution capacitors were underutilized during
periods when the transmission system needed VAR support. Therefore, Idaho Power focused on
integrating a new distribution capacitor control (DCC) system into their existing energy
management system to maximize VAR support for both the distribution and transmission
systems.

Idaho Power partnered with Stellar Dynamics, Inc., to develop a DCC algorithm that was built
around the concept of measuring real and reactive values at the distribution substation
transformer and then controlling the distribution capacitors, based on those measurements. The
obvious benefits were 1) improving system power factor and reducing line losses, 2) improving
real current capacity, and 3) improving real power generation. In addition, DCC also allowed for
individual capacitor monitoring and alarm generation for improper capacitor operation or
abnormal power system conditions at specific capacitors. Finally, the DCC also provided VAR
support to the transmission system by over compensating each transformer VAR setting.

For the 1996 pilot program, the DCC was used on 14 capacitor banks on 3 feeders supported by
a 12.7-kV substation in a high-growth area west of Boise, Idaho. The station had serious VAR
deficiencies that resulted in low line voltage on three feeders. The 14 capacitor banks controlled
with the DCC were comprised of a mix of new and existing installations. As a result of the DCC
use, the total compensation for the substation could be held at a tight tolerance over a wider
range of loading. Additionally, inoperable capacitor banks were quickly identified, which

3-4
Automation Strategies

effectively increased the number of banks available at any one time and helped alleviate the need
for routine capacitor inspections.

Due to the great success of the pilot program, Idaho Power decided to move forward with the
DCC program. They installed DCCs on 16 substations in 1997 and on 14 substations in 1998. As
a result of the DCC implementation, the reactive power load curve at a station was shown to
become the inverse of the real power load curve (instead of following it). This effect reduced
transmission losses and improved voltage regulation. Improved feeder voltage profiles have
reduced or delayed rebuild requirements in many areas of the Idaho Power system, particularly
rural areas with high irrigation loads. The DCC program has also greatly reduced the work-hours
spent on manually operating capacitor banks. Idaho Power is continuing to equip its substations
with DCCs, and it has become a standard feature in all new substations. The cost savings realized
from reduced transmission and distribution losses are expected to cover the cost of conversion
and DCC installation in less than two years.

One unanticipated outcome of the DCC implementation was that many of the power
transformers were subjected to an increased transmission voltage, causing the automatic LTCs to
constantly buck the voltage. The solution to this problem required the Idaho dispatchers to raise
or lower voltage or adjust the bias control in the DCC to add or remove one or two capacitor
banks at each substation at the appropriate time.

Cinergy Corp

Source: B. Koch, “The new SCADA: Reaching Out to Parts (Formerly) Unknown,” Electrical
World T&D, Nov/Dec 2001. http://www.platts.com/engineering/ew_back_issues.shtml

Cinergy Corp uses SCADA-controlled distribution capacitors to provide transmission system


VAR support. This scheme allows the system operator to override the “normal” voltage-based
switching algorithm and control line capacitors located near substations through the
1993-vintage EMS/SCADA system used by Cinergy. Additionally, the capacitors can be
switched as a group or individually.

To implement the system, Cinergy chose the Fisher Pierce SmartLink Utility Application
Interface, with Series 2700 Cellular Communications for use as the RTUs. The SmartLink unit
provides alarm, maintenance, status reporting, and switch control through communications over
existing cellular and public switched telephone networks (PSTN). Referred to as cellemetry, this
type of communication process utilizes the digital control channel of the cellular phone network.
Furthermore, by using the existing cellular network, Cinergy avoids the cost associated with
constructing and maintaining a private communication network. One drawback of this approach
was the complexity of getting the different systems to communicate together as one. To facilitate
the communications, Fisher Pierce gave the SmartLink a Distributed Network Protocol (DNP)
3.0 interface and Applied Systems Engineering of San Jose, CA, supplied a remote
communications/front-end processor to enable communications between the existing
EMS/SCADA system and the new SmartLink units. Since there are several data conversions that
need to be made along the communication path, the system is relatively slow by SCADA
standards, with the round-trip communications path requiring 3-4 minutes to complete. However,

3-5
Automation Strategies

the operators do not have to wait for confirmation from the system to verify that the capacitors
have switched. They can also tell by watching the change in VAR readings on the operator
console.

Cinergy currently operates 150-200 SmartLink units attached to more than 130 MVAR of
capacitors and plans to implement 50-60 more SmartLink controllers in the near future.

Georgia Power

Sources: McCarthy, C. A. and Josken, J., "Applying Capacitors to Maximize Benefits of


Conservation Voltage Reduction," IEEE Rural Electric Power Conference, 2003.
McCarthy, C. and Roberts, S., “The CAPS Program: Capacitor Application for Peak
Shaving,” The LINE, Cooper Power Systems, May 2000.
McCarthy, C., "CAPS -- Choosing the Feeders, Part I," Systems Engineering
Technical Update: Cooper Power Systems, 2000.

Georgia Power takes quite a different approach to using capacitors for transmission support.
Their primary aim is to be able to implement voltage reduction during system-wide peaks to
reduce the system demand. They use capacitors to level out the voltage profile on feeders to
allow a broader voltage reduction. Georgia Power uses both fixed and switched capacitors to
level out the voltage profile. Both are placed to flatten the voltage drop on the circuit. Switched
capacitors are switched based on voltage control.

Capacitors help make it easier to implement voltage reduction. The VARs injected by capacitors
help flatten the voltage profile and allow a lower set voltage on the station LTC transformer. On
many circuits, just-fixed capacitors can flatten the profile enough to reduce the station set
voltage. McCarthy reported how Georgia Power used this strategy to reduce peak loads by 500
kW on circuits averaging approximately 18 MW.

By flattening the voltage profile, customer voltages are more easily controlled just by
communicating adjustments to the station LTC. No communication is needed to feeder mounted
regulating equipment.

The steps to applying this voltage reduction program are listed below (also see Figure 3-1):
1. Add capacitors to correct power factor and flatten voltage profile.
2. Lower substation voltage to the new operating point.
3. During peak times, further lower substation voltage to reduce demand.

3-6
Automation Strategies

Figure 3-1
Three Steps for Applying Capacitors for Peak Shaving

Where fixed banks alone cannot provide a flat-enough voltage profile, voltage-controlled
switched capacitors can help keep the voltage profile narrow enough to successfully implement a
voltage reduction program. (Feeder voltage regulators are another way to attack the problem.)

The Georgia Power approach to applying capacitors for system benefit is quite different from
other approaches. It does not require automated capacitors. Switched capacitors only need local
control.

Summary of Utility Practices

The ways utilities use distribution capacitors for system-wide support varies considerably, and so
does the technological infrastructure. Some approaches are:
• Control During Peaks: One of the main selling points of automation technology is to allow
utilities to dispatch their distribution capacitors for VAR support during system peaks.
• Control at the Substation Level: Distribution feeder capacitors are controlled to improve the
power factor at the substation.
• Peak-Demand Voltage Reduction: This novel method implemented by Georgia Power uses
VAR capabilities differently. Feeder capacitors are used to allow a more optimal voltage
reduction program during peak loadings to reduce demand.

In many cases, automation can provide the flexibility to either switch performance goals or to
meet several performance goals at once. In some cases, the driving force behind automation is
transmission needs. Part of this trend is political. If the distribution utility is separate from the
transmission utility, then the distribution company will likely have to pay for poor power factor
at the interface point. Traditional distribution capacitor application (including the 2/3 rule of
capacitor sizing and placing) has assumed that the transmission system would provide some of

3-7
Automation Strategies

the VARs on the distribution circuit. With power factor penalties, a distribution utility may thus
try to provide all VARs locally.

Transmission versus Distribution Optimization

Most of the time, transmission and distribution goals do not conflict. Where VARs are needed on
the transmission system is where VARs are needed on the distribution system. Reactive power
support is normally best applied as close to the load as possible. For many reasons, there are
situations where a utility wants VARs at a given substation, even though the power factor is
already good. In that case, distribution line capacitors can simply be placed on feeders just
outside of the substation. The key is finding free poles suitable for a capacitor bank installation.
It is okay to operate at a leading power factor. Some checks should be made, including:
• Equipment Loadings: Check that the additional VARs do not push equipment over its
ratings. Devices to check include cable ampacities on substation exits, reclosers, circuit
breakers, CTs, and the station transformer. If ampacities are exceeded, then capacitors need
to be moved to another bus in the substation or to a different substation.
• Voltage: Check that the VARs through the substation transformer do not cause overvoltage
and are compatible with the LTC or regulator line-drop compensator settings.

Problems like this should be uncommon.

Switching Control

For remotely automated capacitors, utilities use several strategies to switch capacitors:
• Operators: This is the most common method. Operators dispatch VARs as needed. The
capacitors are not optimally switched, either for distribution or transmission goals.
Capacitors may have local switching control to switch based on time of day, VARs, or other
means.
• Time Clock: A central control switches capacitors based on a time clock schedule for each
capacitor. Switching commands could be manually overridden.
• Substation VARs: Measurements at the substation are used to send switching commands to
feeder capacitors. Control can be centrally located or in the substation. Order of switching of
the line capacitors is normally preprogrammed.

One application option is to apply and install capacitors in two categories—remote feeder and
near substation. The remote-feeder capacitors are distribution feeder capacitors installed as
necessary for distribution support. If additional VARs are needed for transmission support, they
should be added near the station. Control of the near-substation feeders should be based on VAR
needs at the substation. Different options are available for control of remote units:
• Local: Control the remote feeder capacitors locally using VAR control. These do not need
remote communication, but they do need voltage and current sensing capability.

3-8
Automation Strategies

• Remote: Control the remote feeder capacitors based on substation measurements. The
remote feeder capacitors do not need voltage or current sensing, but they do need
communications ability.

Voltage control provides extra safety to prevent feeder capacitors from causing overvoltages.
The controller trips the bank if it detects excessive voltage, and the controller may also restrain
from switching in if the extra voltage rise from the bank would push the voltage above a given
limit.

Assuming that all capacitors must be controlled remotely, without knowledge of local VARs or
voltage, it is safest to switch in the units closest to the substation first. While this strategy is
safest from the point of view of causing overvoltages and/or losses during light load, it is not
optimal for distribution loss reduction or voltage support. The optimal point for loss reduction
for a capacitor moves out on the feeder as the VAR loading increases on the feeder, as shown in
Figure 3-2. Grainger and Lee (1981) showed that the optimal placement for capacitors is at
points on the circuit where the reactive power equals one half of the capacitor VAR rating. With
this ½-kVAR rule, the capacitor supplies half of its VARs downstream and the other half are sent
upstream.

3-9
Automation Strategies

Figure 3-2
Optimal Capacitor Location for Loss Reduction as the VAR Profile Changes

If VAR measurements are available on a per-feeder basis, more fine-tuning is possible. Then,
capacitors on the feeders with the most VAR load can be switched on first, again from a
closest-to-farthest switching order. Using measurements on each feeder will help reduce
distribution losses the most, even if the switching goal is the overall VAR flow on the
transmission side.

Further refinements are possible, depending on the loading profile on each feeder. If a circuit has
a relatively predictable load pattern, then switching patterns can deviate from switching closest
first. For example, if a circuit has a large single customer with poor power factor, a capacitor
near that facility could be placed early in the scheduled switching order to improve losses and
voltage drop all along the circuit, from the substation to the large customer.

3-10
Automation Strategies

Another consideration is whether to switch capacitors on a per-phase basis or switch them as a


unit. Virtually all utilities switch capacitors as a unit (all three phases together). Most controllers
are set up to control all three phases together. Alabama Power is distinctive in switching units
individually to control VARs on a per-phase basis (Clark 2001). Their capacitor controllers
switch based on local VAR control, but one could extend that philosophy to a centralized- or
substation-level control. Since distribution systems often have significant load imbalance
between phases, switching phases independently helps reduce losses and improve voltage
balance between phases and can also more precisely fine-tune substation VARs. The main
disadvantage of independent phase control is increased cost if voltages and currents are all
monitored locally on all three phases (there is little difference if switching is done remotely
based on substation VAR measurements). Independent control also increases the complexity of
controllers and/or control algorithms.

Most control algorithms for remote control of capacitors are simple—a time schedule or control,
based on VARs measured at the substation. More advanced algorithms could make better use of
the available VARs. This could optimize losses and improve voltage profiles on the distribution
circuit while still meeting transmission support goals.

Station versus Feeder Evaluation

Substation capacitors and feeder capacitors both have advantages. Feeder capacitors are closer to
the loads, and capacitors closer to the loads more effectively release capacity, improve voltage
profiles, and reduce line losses. This is especially true on long feeders that have considerable line
losses and voltage drop.

Table 3-1 highlights some of the differences between feeder and station capacitors. Substation
capacitors are better when more precise control is needed. System operators can easily control
substation capacitors wired into a SCADA system to dispatch VARs as needed. Modern
communication and control technologies applied to feeder capacitors have reduced this
advantage. Operators can control feeder banks with communications just like station banks.
However, some utilities have found the reliability of switched feeder banks to be less than
desired. Further, the best times for switching in VARs needed by the system may not correspond
to the best time to switch the capacitor in for the circuit on which it is located.

3-11
Automation Strategies

Table 3-1
Substation Versus Feeder Capacitors

Feeder Capacitors
Advantages Disadvantages
• Reduces line losses • More difficult to control reliably
• Reduces voltage drop along the • Size and placement important
feeder
• Frees up feeder capacity
• Lower cost

Substation Capacitors
Advantages Disadvantages
• Better control • No reduction in line losses
• Best placement if leading VARs • No reduction in feeder voltage
are needed for system voltage drop
support
• Higher cost

Substation capacitors may also be desirable if a leading power factor is needed for voltage
support. If the power factor is leading, then moving this capacitor out on the feeder will increase
losses. Substation capacitors cost more than feeder capacitors. This may seem surprising, but
station capacitors must be individually engineered; and the space they take up in a station is often
valuable real estate. Pole-mounted capacitors installations are more standardized.

Automation and Other Infrastructure Requirements

While automation provides great flexibility, it can take work to get the best use out of systems in
place. Several possibilities for advances in the technology would make implementation easier as
well as making it easier for utilities to get more benefit out of their systems.

Distribution capacitors offer another data recording point on the system. A utility can use this
data to improve power flow and load-profile models. Likewise, data can be used from other
sources (SCADA, metering, reclosers, and so on) to more efficiently apply and control
capacitors. With extra data, however, comes extra demands—demands for storage, network
bandwidth, communication between databases, and efficient retrieval. Improved information
infrastructure and data protocols will help ease integration and make the data more useful.
Protocols and information formats for sharing data are especially important. A widely used
concept for interoperable data exchange is XML (Extensible Markup Language). XML is a
simple text format that is easy for computers to understand and is very flexible. XML lays the
foundation for data interchange, but it is only a foundation. Industry standard definitions for data
description and exchange are still needed. Within the power system community, XML is used in
a variety of ways. EPRI has developed the Common Information Model (CIM), which the North
American Electric Reliability Council (NERC) has adopted as the standard for exchanging
models between transmission system operators (de Vos, Widergren et al. 2001; EPRI 1001801

3-12
Automation Strategies

2002). This has also been extended to cover distribution system models (Xiaofeng and Schulz
2000; Wang and Schulz 2003). Still, more definitions and standardization are needed, especially
for capacitors, controllers, and related equipment. With standardization, the data from a VAR
measurement recorded within one monitoring system can be sent to a capacitor switching system
with minimal programming to tie the systems together.

More advanced optimization algorithms would help improve transmission and distribution
performance. Centralized controls for systems we have researched of are fairly simple and use
only simple data inputs. Operators manually switch capacitors, capacitors are automatically
switched based on a time clock, or capacitors are automatically switched based on substation
VARs. With better knowledge of distribution voltage and VAR flows, switches can be done
more optimally. The problem formulation depends on the data inputs available and the goals.
Some possible data inputs, from most likely to least likely, are:
• Substation VARs
• VAR flow at each capacitor
• Voltage at each capacitor
• VARs and/or voltage at other locations from other data sources (such as meters or reclosers)
• Utility goals may be different; some possibilities are:
• Maintain unit power factor to the transmission system
• Minimize distribution losses
• Minimize voltage crop

Some of these goals may be obtained simultaneously. An optimization algorithm can also weight
optimization goals as appropriate to meet a utility’s objective. Any real-world implementation
must also be robust enough to deal with missing information or incorrect readings. Also, operator
control during emergency situations is beneficial. In addition, capacitor placement and control
are coupled. Optimal capacitor placement depends on the type of control implemented, and
optimal control depends on the placement.

More sophisticated communication architectures are needed to make advanced capacitor


controllers more “plug and play.” A promising platform is the IEC 61850 architecture that is an
open industry standard developed from EPRI’s Utility Communications Architecture (UCA).
Object models could also be developed for capacitor banks to make controllers more
interchangeable with each other and more interoperable with the overall distribution system
control system.

3-13
4
CONTROL STRATEGIES

Control Strategies

There are several different strategies that can be employed to control switched capacitor banks.
Each strategy has different advantages, although some strategies are more sophisticated than
others allowing for better optimization of the applied capacitors. The more sophisticated control
strategies, however, usually require more input signals to be measured and processed when
compared to the basic control strategies.

Local Control

Switched capacitor banks are controlled either locally or through a centralized system control. As
the name implies, local controls sit on or near the same pole as the capacitor bank and govern the
switching operations of only one local bank. There are several local control strategies available
for switched capacitor banks as shown below (Marx 2003); (Short 2004b):
• VAR Control: The capacitor is switched on and off at the optimum point in the load cycle,
based on VAR measurements on the line. On feeders with only one capacitor bank
installation, VAR control is the most efficient control strategy for maximizing loss and
demand reduction on feeders. However, VAR control is susceptible to interaction from
downstream capacitor banks (downstream banks affect the reactive current flow upstream of
their location). Therefore, when applying multiple capacitor banks using VAR control on a
single feeder, the controls should be set such that the furthest downstream bank comes on-
line first, followed by the next upstream bank and so on. Furthermore, the banks should then
trip in the opposite order by which they switched in (i.e., the last to switch-in should be the
first to trip-out).
• Current Control: The capacitor is switched on and off based on the line current measured
downstream of the capacitor. Reactive current can be determined from line current when the
power factor of the line is known. Current control engages the capacitor during periods of
heavy load which generally have the greatest VAR requirements. Although not as effective
as VAR control schemes, current control provides a fairly good combination of loss
reduction and voltage control.
• Voltage Control: The capacitor is switched on and off based upon the voltage. Threshold
minimum and maximum voltages, as well as time delays and bandwidths, are programmed
into the controller to prevent excessive operations. Voltage control is best suited for
applications in which the capacitor mainly provides voltage profile control and regulation.
Voltage controls can be influenced by both upstream and downstream capacitors, since they

4-1
Control Strategies

affect the voltage along the whole line. Voltage regulators can also cause capacitor control
“pumping” problems. In general, capacitor controllers using voltage control schemes should
be configured to operate before the local voltage regulators. In this manner, the voltage
regulators operate only when the capacitors cannot maintain the desired voltage profile.

It should also be noted that voltage control schemes provide the greatest value on feeder
sections further from the substation. The capacitor should have a minimum effect of 2V (on
120V reference) and the cap on-to-off difference should be approximately 1.5 times the
expected voltage rise when the bank is switched on (Marx 2003).
• Time-Clock Control: The controller switches the capacitor based upon the time of day.
Time-clock control represents the most basic approach for switching a capacitor on and off.
Most time-clock controllers allow for programmable on and off time settings, as well as
settings for weekends and holidays. While this is the least expensive control option, it is also
the most susceptible to energizing the capacitor at the wrong time; because switching is
based on expected rather than measured line conditions. Loads can be different from
anticipated at any time, with holidays and weekends being particularly challenging.
Time-clock controllers are also susceptible to mistakenly set or inaccurate clocks that may
switch the capacitor at times other those planned. Since time control is not based on line
measurements, time-clock controls are not susceptible to interaction with other banks.
• Temperature Control: The capacitor is switched based upon the temperature. Like
time-clock controls, temperature controllers also provide a very basic level of capacitor
control. Typically, temperature controls are set to turn the bank on at 85-90º F (29.4–32.2º C)
and turn the bank off again at 75-80º F (23.8–26.7º C). Since temperature control is not based
on line measurements, they are not susceptible to interaction with other banks.
• Power Factor Control: The capacitor is switched based upon the power factor measured on
the line. This method of control is rarely used by utilities, mostly owing to the fact that power
factor is not a suitable parameter for controlling capacitor switching. Since power factor is
not necessarily an indication of load, power factor based controls may fail to switch in the
capacitor during high loads if the power factor is also high. To compensate for this
shortcoming, power factor controls may also incorporate voltage and current overrides, both
of which make the system more complicated. Due to these reasons, VAR control is typically
used rather than power factor control.

Many controllers offer some or all of these control strategies. Many are usable in combination;
for example, they may turn capacitors on for low voltage or for high temperature.

Some of the control strategies listed above require sensing of one or more line parameters. VAR,
power factor, voltage, and current control require voltage sensing, cutter sensing, or both. To
minimize cost and complexity, controllers often switch 3-phase banks, based on measurements
from a single phase. Although it is not common, a utility may choose to switch each phase of a
3-phase bank independently, based on the VAR requirements of each phase, in order to optimize
loss reduction and reduce imbalance (Clark 2001).

4-2
Control Strategies

Pole-mounted switched capacitor installations can involve a good deal of equipment, which leads
to very busy or congested installations. Therefore, utilities try to reduce the amount of required
equipment for switched capacitor installations whenever possible. As previously mentioned, all
phases of 3-phase capacitor banks are commonly switched based on measurements from just one
phase (thus reducing by 2/3 the amount of hardware needed compared to separately measuring
all three phases). Capacitor controllers require a local power source to operate; and they are
usually served by a small, local control power transfer which also provides the required voltage
sensing. Furthermore, some utilities prefer to use voltage and/or current sensing insulators.
Meter-grade accuracy is not needed for controlling capacitors.

There are some additional considerations for VAR controller installations that are worth
reviewing at this point. As previously mentioned, VAR control is susceptible to interaction from
downstream capacitor banks, since downstream banks affect the reactive current flow upstream
of their location. Therefore, when applying multiple capacitor banks using VAR control on a
single feeder, the controls should be set such that the bank furthest downstream comes on-line
first, followed by the next upstream bank and so on. The banks should then trip in the opposite
order by which they switched on (the last to switch in should be the first to trip out).
Furthermore, the current sensor(s) should be located on the substation side of the bank so that the
controller can detect the reactive power change when the capacitor switches. To properly
calculate VARs, the wiring for the current transformer and potential transformer must provide
correct polarities to the controller.

One manufacturer provides the following rule of thumb for setting VAR control trip and close
settings (Fisher Pierce 2000):
• Close set point: 2/3 • capacitor bank size (in kVAR), lagging
• Trip set point: Close set point – 1.25 • bank size, will be leading
(This assumes that the CY is on the source side of the bank)

For a 600-kVAR bank application, this yields:


• Close set point: 2/3 • 600 kVAR = +400 kVAR (lagging)
• Trip set point: 400 – 1.25 • 600 = -350 kVAR (leading)

In this example, the capacitor trips when the load kVAR drops below +250 kVAR (lagging) with
a bandwidth of +400 to +250 kVAR. This bandwidth is wide enough to prevent excessive
capacitor switching operations or “cycling.” It is typically desirable to have a bandwidth of at
least 1.5 times the expected voltage change due to the capacitor bank. This means that on a 120
V scale, the bandwidth should be a minimum of 3 to 4 V. Additionally, the trip setting must be
set below the normal light-load voltage or the bank will not switch off.

Capacitor controller installations on circuits that can be operated in either direction must utilize
controllers that are capable of accurately operating with power flow in either direction. Only
control strategies that rely on measuring current will be affected by reverse power flow.
Therefore, time clock, temperature, and voltage-control schemes are not affected by reverse

4-3
Control Strategies

power flow, while VAR, current, and power factor control methodologies are affected by reverse
power flow. Some controllers can sense reverse power and shift control modes. One model
provides several options if it detects reverse power flow; namely, switch to voltage mode,
calculates VAR control while accounting for the effect of the capacitor bank, inhibits switching,
trip and lock out of the bank, or closes and holds the bank in. If the circuit has distributed energy
resource (generation or energy storage), then it is undesirable to shift modes based on reverse
power flow. The controller should shift modes only for a change in direction to the system
source.

Most capacitor controllers have built-in counters to record the number of switching operations
performed. Counter information can be very helpful in determining when to perform
maintenance and in diagnosing control setting problems. Control setting problems most often
result in excessive switching, which will be reflected in the counter data. Modifying the control
settings, time delays, or bandwidths to reduce the number of switching operations will correct
excessive switching. Some controllers can limit the number of switching operations within a
given time period to reduce wear on the capacitor switches.

Voltage control provides additional safety to prevent capacitors from causing overvoltages.
Voltage override can accompany any type of control strategy (time clock, current, VAR control,
and so on.) such that the controller trips the bank if it detects excessive voltage on the line.
Voltage override may also be used to prevent the controller from switching a bank on if it will
push the line voltage above a certain point.

Centralized Control

Advances in wireless communication technologies have made remote capacitor control more
easily achieved and more economical than ever before. Cellular phones, pagers, and other
wireless technologies have become ubiquitous in modern life, turning up in new applications
such as remote capacitor control. There are several control schemes available for remotely
controlled capacitor installations as follows:
• Operator Dispatch – Most schemes allow operators to dispatch distribution capacitors. This
feature is one of the key reasons utilities automate capacitor banks. Operators can dispatch
distribution capacitors just like large station banks. If VARs are needed for transmission
support, large numbers of distribution banks can be switched on. This control scheme is
usually used in conjunction with other controls.
• Time Scheduling – Capacitors can be remotely switched, based on the time of day and
possibly the season or temperature. While this may seem like an expensive time control, it
still allows operators to override the schedule and dispatch VARs as needed.
• Substation VAR Measurements – A common way to control feeder capacitors is to dispatch
based on VAR/power factor measurements in the substation. If a feeder has three capacitor
banks, they are switched on or off in some specified order, based on the power factor on the
feeder measured in the substation.
• Capacitor Location VAR Measurements – The continuing advancement of capacitor
controller capabilities, coupled with increasing capability for two-way data transfer, are now

4-4
Control Strategies

making it possible for capacitor controllers to measure line parameters at their location and
report that data back to a central station controller. The central station controller examines
the data from each capacitor location, and possibly the substation as well, and makes
decisions for switching each capacitor individually. A major detractor of this type of
operation is that CTs need to be installed at each site in order to make VAR measurements,
and this carries a significant equipment cost—much higher than just measuring at the
substation.
• Other Methods – More advanced (and complicated) algorithms can be used to dispatch
capacitors, based on a combination of local VAR measurements and voltage measurements,
along with substation VAR measurements.

All of the control strategies mentioned above will typically utilize a local voltage override
feature, especially if the controller only has one-way communication capabilities. Local voltage
override prevents the capacitor from switching, if doing so will push the voltage beyond limits
set by the user. Additionally, most controllers used for centralized control will have fail-safe
modes in which they will revert to a type of local control (voltage, current, VAR, time,
temperature, combination, and so on.) if communication with the central station is lost.

Centralized control with one-way communication sometimes relies on local control intelligence
to make daily capacitor switching decisions while maintaining the ability to globally control
capacitors via the central system. In this type of system, the local control would still utilize
voltage override at the local level to avoid high/low voltage situations during global dispatch. An
alternative control methodology uses the SCADA master capacitor controller to govern
switching for all capacitors based on data feedback from substations. In this manner there is still
only one-way communication with the capacitor controller, but the feedback loop is completed
by data transmission from the substation. The master controller processes the substation data and
dictates capacitor switching on a per-bank basis, thus optimizing the operation of each feeder.

Capacitor control systems utilizing two-way communication operate in a similar manner to those
with one-way communication but have more options for control, feedback, and monitoring. With
two-way communication, it is possible for the capacitor controller to report the line’s operating
conditions at the capacitor location back to the SCADA master controller. Operated in this
manner, the control system can make capacitor switching decisions based on actual line
parameters at the capacitor location. Of course, two-way communication also provides global
control options for addressing all capacitors at once. Global control provides flexibility to group
feeders in several different manners: by substation, by district, system wide, and so on.

The ease and flexibility with which capacitor banks can be switched off and on is one of the
main benefits of centralized capacitor control. While local controls may offer many
combinations of operating parameters for fine-tuned control, they still cannot compare to the
flexibility provided by centralized control. Centralized control allows any bank to be switched by
simply entering commands in the control room. Although it typically involves higher costs, two-
way communications also offer some advantages over one-way communication:
• Feedback: A local controller can confirm that a capacitor has switched on or off
successfully. Utilities can use the feedback from two-way communications to dispatch crews

4-5
Control Strategies

to fix capacitors banks with blown fuses, stuck switches, misoperating controllers, or other
problems.
• Voltage/VAR Information: Local information on-line VAR flows and line voltages allows
the control to more optimally switch capacitor banks to reduce losses and keep voltages
within limits.
• Load Flows: Voltage, current, and power flow information from pole-mounted capacitor
banks can be used to update and verify load-flow models of a system. The information can
also help when tracking down customer voltage, stray voltage, or other power quality
problems. Loading data helps utilities monitor load growth and plan for future upgrades. One
utility even uses capacitor controllers to capture fault location information, helping crews to
locate faults.

Another benefit of centralized control is the enhanced ability to detect misoperations and bank
failures. Neutral monitoring can detect capacitor failures while they are happening, mitigating
the need for periodic field visits by technicians to manually check each bank. Kansas City Power
& Light used their automated capacitor bank system to perform a detailed analysis of switched
capacitor bank reliability (Goeckeler 1999). Two-way communication provides the added benefit
of positive confirmation that the switching operation was successful. Alternatively, if only one-
way communication is available, then substation VAR flows can be monitored to verify
capacitor switching.

Coordination of Switched Capacitors and Voltage Regulators

Capacitor banks can present coordination issues with voltage regulators. Capacitors, either fixed
or switched, affect the voltage profile and reactive current flow on the feeder. Capacitors
upstream of a voltage regulator do not interfere with the regulators control action. However,
capacitors downstream of voltage regulators will affect the reactive current flow and voltage
profile as seen by the regulator.

Capacitors just downstream of a regulator affect the current that the regulator sees, but do not
measurably change the shape of the voltage profile beyond the regulator. In this case, one would
like the line-drop compensation to ignore the capacitor. The voltage-spread compensation1 with a
low compensator X/R or the zero-reactance compensator settings work well, because they ignore
or almost ignore the reactive current. This approach works with fixed or switched capacitor
banks downstream of the regulator. The load center approach is more difficult to get to work
with capacitors.

Capacitors closer to the end of the line have a pronounced effect on the voltage profile of the line
and thus will also effect the regulator control action. In this case, we do not want zero-reactance
compensation because some X is needed to compensate for the capacitive current.

1
With voltage-spread compensation, compensator settings are found by specifying the band over which the load-
side voltage should operate. Zero-reactance compensation is a form of voltage-spread compensation with the X/R
ratio equal to zero. See (Short 2004b) for more information.

4-6
Control Strategies

Although rare, switched capacitors utilizing voltage control schemes can affect the tap-changing
controls on upstream regulators. A regulator may respond to an upstream or downstream
capacitor switching, but that does not add up to many extra tap changes; since capacitor
switching tends to be infrequent. Normally, the capacitor will not cycle back and forth against
the regulator. The only case might be if the regulator has a negative setting for the reactive line-
drop compensation (Short 2004b).

Capacitor switching tends not to be affected by voltage regulator operation unless the capacitors
utilize a voltage control scheme. Cycling of the switching capacitor and voltage regulator are
very rare. Voltage regulators do not usually affect capacitor switching for several reasons
(Westinghouse Electric Corporation 1959):
• The voltage control bandwidth of the capacitor controller tends to be wider than the voltage
regulator bandwidth.
• The voltage change due to capacitor switching is less at the regulator location than it is at the
capacitor location.
• The capacitor controls usually incorporate a greater time delay then that regulator control
uses.

If regulator operation does affect capacitor switching, it can usually be remedied by small
adjustments to the bandwidth and time-delay settings of the capacitor controller.

Coordination of Switched Capacitors and Distributed Generation

The addition of distributed energy resources (DER), either generation or storage, to a feeder can
cause faulty operations of switched capacitor banks. If the DER alters the power flow, line
current, voltage profile, or other operational characteristics of the feeder, then the capacitor
controls may no longer be set appropriately, possibly resulting in high- or low-voltage on the
line, excessive losses, and poor power factor (Brady, Dai; et al. 2003).

As an example, consider the simple case shown in Figure 4-1 where there is a DER unit installed
just downstream of a switched capacitor bank. In reality, most feeders will have more than one
capacitor bank located along their length; and several, if not all of them, will be switched banks.
However, this simple example is sufficient for illustrating the need to re-evaluate the capacitor
control logic for the post-DER line conditions.

4-7
Control Strategies

Substation

Load Load Load Load DER Load Load Load Load

Capacitor Controller

Figure 4-1
Example Feeder with a Switched Capacitor Located Just Upstream of a Distributed Energy
Resource

Prior to installation of the DER unit, the capacitor control scheme was configured to provide
optimum benefit to the distribution line. It may have been setup to provide optimal VAR support
or voltage support, but for this example let’s assume that it was providing voltage support for the
line. If the capacitor is switched by time-clock control, then a distribution engineer has reviewed
historical line data to determine the optimal time to switch the capacitor on and off in order to
provide the proper voltage support to the line. However, the DER now serves some of the
demand on the feeder thus helping to prop up the voltage. It is possible that the capacitor will
now cause a high voltage condition at and around its location. Therefore it is necessary to review
the new line conditions in order to redefine the operating schedule (or possible remove the
capacitor bank altogether). Other types of capacitor controls may be affected in the following
manner:
• Current Control: This control methodology will exhibit a delayed turn on and shortened on
time, since the DER will lower power demand resulting in less current flow at the capacitor
metering location. The shortened duty cycle of the capacitor bank will cause increased kVAR
demand from the transmission system.
• Voltage Control: This control methodology will exhibit a delayed turn on and shortened on
time since the DER will lower power demand resulting in higher voltage at the capacitor
metering location. The shortened duty cycle of the capacitor bank will cause increased kVAR
demand from the transmission system.
• VAR Control: This control methodology will continue to function as it did under pre-DER
conditions, provided that the DER operates at unity power factor. If the DER does not
operate at unity power factor, then the control may not operate properly and could lead to an
overvoltage condition at and around its location.
• Controls With Voltage Override: These controls may not turn on at all, due to the elevated
voltage at the capacitor location stemming from the DER unit serving part of the feeder
demand. If the control does switch on, it may quickly switch off again due to the voltage
override, leading to capacitor cycling problems. Capacitor cycling is particularly grievous as

4-8
Control Strategies

it can lead to failure of the capacitor unit and expose the line to numerous capacitor switching
transients.
• SCADA-Based Centralized Controls: Depending on the sophistication of the capacitor
switching control used, the system will react in different ways. It is clear that the feeder
parameters will change with the addition of the DER, so the capacitor switching scheme may
need to be changed. Very sophisticated programs may automatically adjust to the new
operating conditions and not require any intervention from the operator. Less sophisticated
controls may require an engineer to evaluate the new line conditions and set a new switching
program. However, this should be easily accomplished, since all the required data should be
available from the SCADA system.

Adding DER to a feeder does not guarantee that the capacitor control will no longer function as
intended. However, it does pose enough of a concern that a review of capacitor switching
controls should be a part of any engineering work done when considering adding DER to a
feeder. The best-case scenario is one in which the capacitor is switched via SCADA-based
controls, since those are likely the easiest to alter and offer the most flexibility in configuration.
Of course, this should be looked at in relative terms. A home owner adding a 2-kW photovoltaic
system to his house is not likely to impact the distribution feeder, whereas a manufacturing
facility adding a 500-kW internal combustion plant may have a profound effect on the feeder. It
is impossible to define what determines a “large” DER facility, as it really depends on the
properties of the feeder, to which it will be connected. To be safe, a review of the impact of the
DER on capacitor switching should be completed for any proposed DER plant that warrants
study by the utility personnel.

4-9
5
VOLTAGE AND CURRENT MEASUREMENTS

Basic Measurements

At a minimum, switched capacitor installations measure line voltage as an input to the


controller’s switching algorithm. However, many controllers also include line current as a
measurement, as it is needed to operate in the VAR or current-control modes. Even installations
that switch solely on a time or calendar basis may measure line voltage in order to prevent
creating an overvoltage condition on the line.

Most capacitor controllers are designed to plug into a 4- or 6-jaw meter socket, as shown in
Figure 5-1. The type of meter socket needed (4- or 6-jaw) will depend on how many parameters
are being measured. Controllers mounting to 4-jaw sockets typically do not measure current, thus
the 4-socket tabs are configured for ac line voltage and neutral, as well as for the open and trip
command signals. Controllers utilizing 6-jaw sockets add current sensor hot and neutral tabs, as
shown in Figure 5-2. These can even be configured to include input from a neutral current
sensor. Socket configuration varies by manufacturer; so when ordering the controller, a variety
of socket pin-outs can often be specified.

Figure 5-1
Typical Capacitor Controller Mounting Configuration with a Meter Socket
Courtesy of S&C Electric Company

5-1
Voltage and Current Measurements

Capacitor controllers can also be specified with terminal strip inputs rather than meter socket
connections. In this case, the same connections are made to the capacitor controller, but they are
landed on a terminal strip on the controller rather than at a meter socket.

Capacitor controllers need to be powered at the pole; and for this purpose, they often make use of
a small (typically 1 kVA) CPT. The transformer is usually mounted high on the pole near the
capacitor cans and switches. As is evident in Figure 5-2, there is only one set of voltage inputs at
the meter socket. This is because for most capacitor controller installations, the CPT’s secondary
voltage is accurate enough to serve as a voltage input to the switching algorithm. However,
separate voltage transducers are sometimes used as well. Since voltage measurements are
typically required on only one phase for a 3-phase switched capacitor bank, using the output of
the CPT serves this need while negating the need for additional, metering-dedicated voltage
transformers. If dedicated voltage transformers are required, there is a wealth of models available
from many major manufacturers. However, it should be noted that some models of capacitor
controllers do not have additional inputs for voltage sensing beyond the combination power and
voltage sensing input mentioned above. A typical wiring diagram for a capacitor controller
installation is shown in Figure 5-3.

Figure 5-2
Example of Connections in a 6-Jaw Meter Socket Used for Capacitor Controller
Installations

5-2
Voltage and Current Measurements

Figure 5-3
Generic Example of Pole-Top Connections for Input Signals to a Capacitor Controller
Note: Protection devices and other apparatuses have purposely been omitted from this drawing for clarity. Actual
installations would also utilize hardware, such as surge arresters, cutouts, and fuses.

If current measurements are required, there are a numerous models of CTs available for use in
capacitor-controller installations. Several manufacturers make line-post current sensors that can
be used in place of traditional line-post insulators on the crossarm to support the phase
conductor. Some examples of line-post current sensors are shown in Figure 5-4 and Figure 5-5.
Some line-post sensors (such as the CSV line-post sensor from S&C Electric shown in Figure
5-5) perform both voltage-sensing and current-sensing duties; and they are available in 15-, 25-,
or 35-kV classes.

5-3
Voltage and Current Measurements

Figure 5-4
Series 1301 PowerFlex® Current Sensors From Joslyn Hi-Voltage
Courtesy of Joslyn Hi-Voltage

Figure 5-5
S&C Electric Company’s CSV Line Post Current and Voltage Sensor
Courtesy of S&C Electric Company

Neutral Monitoring

Neutral monitoring (Figure 5-6) is another protection feature that some capacitor controllers
offer. Neutral monitoring can detect several problems:
• Blown Fuse: When one capacitor fuse blows, the neutral current jumps to a value equal to the
phase current.
• Failing Capacitor Unit: As a capacitor fails, it shorts out internal groups of series packs.
Prior to complete failure, the unit will draw more current than normal. Figure 5-7 shows how
the neutral current changes when a certain portion of the capacitor shorts out. Capacitors
rated from 7.2 to 7.96 kV normally have three- or four-series sections, so failure of one
element causes neutral currents of 25% (for four-series) or 34% (for three-series) of the phase
current. If more than half of the capacitor’s series sections fail, the neutral current flow will
exceed the capacitor’s rating.
• High Harmonic Current: Excessive neutral current may also indicate high harmonic currents.

5-4
Voltage and Current Measurements

Figure 5-6
Neutral Monitoring of a Capacitor

Figure 5-7
Neutral Current Drawn by Failing, Grounded-Wye Bank, Depending on the Portion of Bank
Failed
(Note: The neutral current is in per-unit of the nominal capacitor current)

5-5
Voltage and Current Measurements

Neutral monitoring is common in substation banks, and many controllers for switched
pole-mounted banks have neutral-monitoring capability. Neutral current monitors for fixed banks
are also available, with either a local warning light or a wireless link to a centralized location.

Neutral monitoring can help reduce operations and maintenance needs by eliminating regular
capacitor patrols and field checks. Quicker replacement of blown fuses also reduces the time that
excessive unbalance is present (as well as reduces extra losses and the possibility of stray
voltage). This can not only lead to more reliable VAR regulation, but can also reduce the number
of capacitor banks needed.

Some capacitor controllers have several programmable levels of neutral current alarm. For
example, once the current gets to a certain level, the controller may be set to page the operator.
Then if the current exceeds a given amount beyond the initial alarm level, the controller may trip
the bank off-line.

5-6
6
COMMUNICATION TECHNOLOGIES

Communications Technologies

There are several technologies currently in use for communicating with the capacitor controllers.
Some offer one-way communication while others offer two-way communication. With one-way
communication, commands can be dispatched to the capacitor controllers in the field, but there is
no communication from the field back to the control center. Two-way communications offer data
flow, both from the command center to the field units and from the field units back to the
command center. The technologies used for centralized capacitor control communications
include:
• 900-MHz Radio: These systems are very common and are widely applied for centralized
capacitor control. There are several spread-spectrum radios available that cover 902-928
MHz applications. Implementing 900-MHz radio control on a private network requires
infrastructure, including towers.
• Pager Systems: Pager systems offer inexpensive options, especially for systems with
infrequent switching. These systems are mostly one-way, but there are some two-way pager
systems available. Most commercial paging systems can be utilized; however, that means
that while one-way coverage is rather wide-spread, two-way systems tend to be limited to
clusters around major cities.
• Cellular Phone Systems: These systems use commercial cellular networks to provide
two-way communications. Many vendors offer modems that are compatible with several
cellular networks, and coverage is typically very good.
• Cellular Telemetric Systems: These use the unused data component of cellular signals that
are licensed on existing cellular networks. They allow only very small messages to be sent to
perform basic capacitor automation needs. Coverage is typically very good—the same as
regular cellular coverage.
• VHF Radio: Inexpensive, one-way communications are possible with VHF radio
communication, and VHF radio bands are available for telemetry uses such as this. Another
option is a simulcast FM signal that uses extra bandwidth available in the commercial FM
band.

The communication technologies described above utilize different frequencies as dictated by the
Federal Communications Commission (FCC). (For more information on the rules governing
these and other communication technologies, please visit the FCC website at
http://www.fcc.gov ). Table 6-1 outlines the frequency bands typically used for certain
applications.

6-1
Communication Technologies

Table 6-1
Frequency Bands for Typical Applications (Young 1999)

Communication Technology Frequency Range

Paging Systems 30-300 MHz

Cellular, Air/Ground 300 MHz – 3 GHz

Microwave Transmission 3-30 GHz


Satellite Systems 30-300 GHz

Radio wave signal strength diminishes rapidly as the signal moves away from the transmitting
antennae. In perfectly flat terrain without obstructions the signal will fade according to the
inverse-squared law; that is the signal strength decreases by the square of the distance. In this
manner the signal strength decreases by a factor of 4 for each doubling of the distance from the
transmitter. For example, a signal that is 20 mW one mile from the transmitter will be 5 mW
two miles from the transmitter. This however only dictates what the minimum propagation loss
will be. Real-world networks also have a variety of obstacles in the transmission pathway which
further reduce the signal strength. In practice, signal loss tends to be closer to inverse of the 5th
or 6th power of the distance (Young 1999).

Just about everything an RF signal encounters during transmission from the antenna to the
receiver changes it in one of two ways: (1) it makes the signal smaller, or (2) it changes the
signal direction to a varying extent. Most materials that a signal encounters, including, air, rain,
concrete, wood, brick, glass, and foliage all make the signal smaller. Most obstructions in the
signals path can be viewed as passive devices with a small amount of insertion loss, often
referred to as absorption (Weisman 1999). Many objects also change the direction, or reflect, the
radio signal as well. Reflections can be total or partial. In a partial reflection some of the signal
is absorbed and some is reflected. Water often results in total absorption while concrete exhibits
partial absorption and metal provides total reflection. Reflections occur at the angle of incidence
of the wave. Therefore, a radio wave striking a building at 90º (perpendicular) will be reflected
back at 90º. Correspondingly, a wave striking a building at 45º will be reflected back at 45º as
shown in Figure 6-1.

6-2
Communication Technologies

Figure 6-1
Reflection of Radio Signals

Radio frequency and microwave signals are susceptible to reflection and absorption along their
transmission path; therefore, a line-of-site or path study must be part of the design process for
any radio or microwave communication network being considered. The study should take into
account any reflection or absorption that may occur from lakes, rivers, drainage fields, sandy
areas, marshland, or large, flat roofs along the transmission pathway. These, as well as other
common items, affect how the signal will transmit across the pathway. Urban pathways tend to
be shorter in length than rural transmission pathways, and they often use antennae placed on tall
buildings rather than on towers. In an urban setting, the ability to see from one end of the
pathway to the other with a pair of binoculars or telescope is a good indication that the pathway
is viable. Rural pathways, which are generally longer than urban pathways, often exceed 9 miles
(14.5 km) in length and make use of tower mounted antennas. It can be difficult to verify the
line-of-site pathway for a rural system, particularly if the towers are yet to be built. Obviously,
leasing space on towers or otherwise utilizing existing towers provides a pretty good indication
that the line-of-site is viable. When designing rural systems “from scratch,” it is often necessary
to plot the transmission pathway on a map and use specialized software tools to determine the
required antenna locations and heights, as determined by geography, obstacles, and reflection
criteria. It is also important to evaluate potential future obstructions in the line-of-site survey
path. Potential future obstructions include growing tree foliage, future building construction, and
evolving flight plans of nearby airports.

Utilities run the gambit of existing communication infrastructure from those that have their own
dedicated 900-MHz, spread-spectrum radio networks to those that do not have any
communication network in place. It makes sense, then, that utilities will try to work with the
equipment in place, as much as possible, when migrating to a centralized capacitor control
system. Because of this, there tends to be a high degree of variability from one utility to the next
in terms of the communications media they are using for centralized capacitor control. One trend
that is becoming apparent is that utility communications are increasingly being provided by
third-party sources that specialize in data communication systems. These companies span the
spectrum from paging service providers to companies that use a mix of cellular, microwave, and
satellite transmission services. As the price of these services continues to decrease, it becomes
more attractive for a utility to rely on an outside party to maintain a communication
infrastructure. Additionally, utilities are no longer the sole users of many of the technologies now
being employed for data transmission. Data services use existing cellular or paging

6-3
Communication Technologies

infrastructures, thus spreading the cost of building and maintaining those networks over a larger
base (thereby reducing the cost to the utility to utilize the network). Table 6-2 shows some of the
positive and negative aspects of owning communication infrastructure.

Table 6-2
Pro’s and Con’s of Radio Network Ownership

Public / Third Party Utility Ownership

Someone else builds and operates Security


Pro the network
Absolute control over network usage

Monthly / reoccurring cost for Network construction requires a


service major investment
Con
Someone else sets the network Operation and maintenance costs to
rules – lack of control maintain system integrity

Utility ownership of the communication network brings with it certain operation and
maintenance tasks that are needed to ensure the system is efficient and reliable. Communication
towers and antennas need periodic inspection and maintenance. Cellular carriers operate some of
the oldest towers in the communication industry, and they indicate that best way to ensure long,
reliable tower life is to implement a strong inspection and maintenance program (Bedell 2001).
Towers need regular inspections to identify corrosion or weakened members and to look for
metal fatigue and fractures. Guyed towers require even more frequent inspection to examine and
repair stretching of the guy wires. Tower foundations should also be examined during the
inspection, assessing the condition of the footing bolts and foundation members (such as
concrete footers). Towers and antennas may also need to be inspected after severe storms to look
for damage from ice accumulation, wind stress, or hail. Sticking to a proactive inspection and
maintenance schedule will help reduce storm-related outages by ensuring the equipment is in the
best possible condition prior to the storm.

There are many different service providers for the communication technologies mentioned
above. Several providers are outlined in this section to provide the reader with additional
background on what these services entail. This listing is for information only and neither EPRI
nor EPRI Solutions, Inc. endorses any particular communication provider.

Spread Spectrum 900-MHz Radio Systems

A 900-MHz signal inherently penetrates objects better than signals in the 2.4-GHz or 5-GHz
range, thus 900-MHz radio systems are less restricted by line-of-sight issues than 2.4-GHz and 5-
GHz systems. However, the bandwidth delivered is less for an unlicensed-band 900-MHz, fixed,
wireless system than for 2.4-GHz or 5-GHz systems.

The 900-MHz band is used for millions of low-powered, FCC Part 15 devices. The FCC Part 15
regulation exempts devices from licensing requirements if they 1) create no interference and 2)
6-4
Communication Technologies

tolerate any interference (Harris 2001). These devices include utility meters, cordless phones,
baby monitors, home security systems and more. The Part 15 rules provide a “safe harbor”
protection for unlicensed devices but mandate a primary role for licensed services. The safe
harbor requires, principally, power limitations and a height restriction on a wireless system’s
antenna. If these rules are followed, the unlicensed product is considered to be within its safe
harbor (Sanders 2002).

A utility can choose to operate its own 900-MHz radio system or work with commercial
providers to achieve connectivity with this technology. As previously discussed, there are pros
and cons on each side of system ownership. Private radio systems make the most sense when
construction costs can be held down due to existing infrastructure, good site geography, or other
considerations. Applications that require immediate and frequent data transmission provide the
most justification for using a private network. Spread-spectrum radio networks can provide data
transmission of 4,800 baud for up to 2 miles in transmission pathways with good line-of-sight
characteristics.

Spread-spectrum radio services are provided by several commercial companies including


Microwave Data Systems (www.microwavedata.com), DataRadio (www.dataradio.com), and
Cellnet (www.cellnet.com).

UtiliNet® Wireless Communication from Cellnet (www.cellnet.com)

The UtiliNet® system uses spread-spectrum radios in the license-free 902-928 MHz area of the
radio spectrum. The UtiliNet® system utilizes a mesh structure in which every data radio in the
field becomes a node in the mesh that can communicate with any neighboring node. This means
that every controllable device with a data radio is a node in the control mesh and can
communicate with any neighboring device. Since there is not a single point of failure, this results
in increased communication reliability.

Each end-device radio generates a packet of data that is transmitted into the network mesh. The
packet is addressed to the destination and traverses the network by hopping from radio to radio in
the direction of the destination-addressed radio. The route chosen for traversing the network is
dynamic, automatically routing the package over the shortest path; and it employs automating
network rerouting in the event that a particular data path is not clear. In addition to simple
polling schemes, UtiliNet® mesh networks can accommodate peer-to-peer and unsolicited
reporting schemes.

Pager Systems

Paging systems use FM transmitters; and more than one transmitter can be utilized, depending on
the geographical coverage requirements of the system. There are several companies that offer
paging services for SCADA and telemetry systems, or a utility can opt to operate its own paging
system.

Paging services often utilize mature networks and can be rather cost-effective in
capacitor-control operations. However, there is the potential for substantial message delays of up

6-5
Communication Technologies

to several minutes due to the store-and-forward nature of paging technology. Another potential
drawback of paging systems is that while one-way paging coverage is fairly widespread,
two-way service is still growing and is often limited to metropolitan areas. Paging providers
include Mobitex (www.mobitex.org) and SkyTel (www.skytel.com).

FLEX™ Paging Protocol

FLEX™ (one-way) and ReFLEX™ (two-way) are Motorola, Inc. paging protocols that are
widely used in current paging systems. These protocols were designed to give carriers more
capacity on their networks and faster transmission times. The FLEX family of protocols includes
FLEX™, ReFLEX™, InFLEXion™.

The FLEX™ protocol provides one-way communication which can be set at rates of 1600, 3200,
4800, or 6400 bps. Motorola worked with SkyTel to develop the ReFLEX™ protocol to provide
two-way paging communications. ReFLEX™ uses the FLEX™ protocol but provides a reverse
channel that operates at 9600 bps to provide message acknowledgement. ReFLEX™ saves
transmission time by first broadcasting to locate the intended target unit and then, once the target
responds, sending the actual message through the most efficient path to the target. When the
target unit has received the message, an acknowledgement is sent back to the message originator.
ReFLEX™ is actually comprised of two versions, ReFLEX-25™ and ReFLEX-50™, which
differ in speed and usage. Only SkyTel uses ReFLEX-50™, which is fixed at 6400 bps and
requires more bandwidth than the ReFLEX-25™ due to its higher transmission rate.

SkyTel ReFLEX Two-Way Data Telemetry Service (www.skytel.com)

SkyTel provides wireless data and messaging services, wireless email, interactive two-way
messaging, wireless telemetry services, and traditional text and numeric paging to customers
throughout the United States. SkyTel offers two-way FLEX paging services for utility
automation operations. The SkyTel ReFLEX two-way data telemetry network is a licensed
900-Mhz frequency band that is available in all U.S. major metropolitan areas. The SkyTel
network, illustrated in Figure 6-2, operates as follows:
1. Commands/data are received by the SkyTel Network Operating Center (NOC).
2. The commands are then transmitted to the redundant satellite system (land-based frame relay
networks are used as backup for communications).
3. From the satellite system, the command signals are sent to the land-based, 900-MHz
transmitter system.
4. The land-based transmitters send the command signals to the data radio in the capacitor
controllers by way of the 900-MHz network.

The data return path starts with a signal broadcast from the data radio in the capacitor controller
that is received by a base receiver. The data is transmitted back to the NOC by means of an MCI
frame relay operated as a digital telephone wire-line data circuit.

6-6
Communication Technologies

Figure 6-2
SkyTel Telemetry Services Advanced Messaging Network
Courtesy of SkyTel

Cellular Systems

The basic analog cellular service in the United States is the Advanced Mobile Phone System
(AMPS). In a cellular field communications system, an area is divided into a number of adjacent
cells, each containing a base-station tower. The tower is connected to a mobile-telephone
switching office, which serves partly as a conventional telephone switching office, including
connections to the regular land-based telephone service. The switching office is also connected
to all the other base stations in the system, and it circulates a “paging signal” from one base
station to another until it finds the mobile unit whose number has been dialed. The called mobile
unit responds by transmitting an acknowledgement to its local base station so that the switching
office knows that both parties are ready to communicate as well as knows which cell each party
is in. The switching center assigns a channel pair to each of the mobile units and the units use
these channels to communicate.

Cellular Data Channel Systems

In a wireless cellular telephone system, control channels, which do not carry voice
communications, are used to initiate calls, confirm subscriber status, and track roaming phones
as they move across the country. Significant excess control channel capacity exists in the range
of 90% at normal voice traffic times to 50% at peak times.

Cellular data channel systems work by mimicking a roaming cellular telephone to access the
cellular control channel. When the capacitor controller, through its cellular radio, initiates a
“call,” it begins by registering with the local cellular network. The registration information from
the capacitor controller contains the data being reported back to the master station. Since the data
6-7
Communication Technologies

packet is completely passed during the registration process, sending the message does not incur
any long-distance fees, and it is sent within seconds.

Cellular control channel communication offers very broad coverage reaching nearly 98% of the
U.S. population (Harris 2001). Cellular data channel communication is available anywhere
AMPS is available. Another benefit of this service is that individual devices do not require
individual cellular accounts. Cellular data channel communication can be very low-cost. The
technology does not require a substantial infrastructure investment from the utility, since it
leverages the unused portion of the existing cellular radio network. One drawback to using
cellular control channel communication is that data packets are limited to either 8 digits or 25
digits, depending on the provider.

Aeris has teamed up with Telemetric Corporation to provide cellular data channel
communication for centralized capacitor control systems. Rather than purchasing airtime from
the major carriers throughout North America, Telemetric purchases its communications from
Aeris and Aeris negotiates excess capacity from the two major carriers in each cellular region.
The local switch recognizes that the MicroRTU’s radio is not one of its own cell phones and
looks up its home market, Aeris, which looks like another cellular carrier to the local switch.
Once the data packet arrives at Aeris, the information is decoded and routing information is
deleted. Data is then passed from Aeris to the Telemetric NOC via a secure transmission control
protocol/Internet protocol (TCP/IP) connection and from Telemetric on to the utility’s SCADA
system. (For more information, please see www.telemetric.net and www.aeris.net.) Other
providers include Cellemetry® Data Service, who has partnered with Fisher Pierce to provide
integrated capacitor control via cellular data channel communications (www.cellemetry.com and
www.fisherpierce.com).

Data channel technology can have several advantages over conventional cellular services,
particularly for applications whose data size requirements are small. The costs for data channel
services tend to be lower than the alternatives without sacrifices to high reliability and
widespread coverage. Additionally, the small size of the data packets provides a robustness that
extends the range of the service considerably beyond conventional cellular networks.
For example, these data transmissions can penetrate multiple building walls without difficulty
and can typically communicate with a cellular transceiver at a distance far beyond that of cellular
voice capability, making it possible to easily cover remote facilities.

Cellular data channel communications also benefit from utilizing the existing cellular
communication network. Since the system relies on the existing network, there is very little
infrastructure that needs to be built and maintained by the utility. Build-out, operation, and
maintenance of the cellular telephone system are essentially paid for by the users of voice
services. The downside to this is that the utility is now relying on infrastructure that is beyond its
control. However, using the cellular communications network also means that coverage tends to
be very good, typically quoted at 98% of the populated area of North America.

6-8
Communication Technologies

Cellular Digital Packet Data

There is a significant amount of dead time on vacant channels and between voice transmissions
on wireless cellular systems. Cellular digital packet data (CDPD) uses channel-hopping
technology to capitalize on the dead time to transmit short packets of data. The dead time exists
both on vacant channels and on channels currently in use for voice communications where
conversation is silent, such as a pause between sentences or even syllables. Some estimates have
indicated that cellular systems only use 60% of their capacity at any one time and CDPD appears
as an attractive way for cellular companies to increase the efficiency of their networks and
generate more income (Young 1999).

CDPD systems use both planned and forced hops. In a planned hop, the system determines that
there is more idle bandwidth on another channel, so the data packet is inserted (“hopped”) to the
new channel in a continuous flow without buffering. A forced hop occurs when the data packet is
must hop channels because of a bandwidth shortage on the channel in use. Voice traffic is
considered higher priority than data, so if there are no channels available for an incoming voice
call, the data packet will be kicked off the channel to accommodate the voice transmission.
CDPD provides near real-time transmission of large quantities of data. However, even though
the system can support large data transmissions, capacity can be limited during times of high
voice volume, since voice transmission gets priority over data transmission.

Rather than charging for airtime, CDPD charges by the volume of data transmitted. Different
carriers offer different rates plans and fees can vary widely between carriers and rate plans.
Unfortunately, CDPD provides limited coverage and is typically only available to major
metropolitan areas. The technology also tends to be more costly than cellular data channel
service. AT&T (www.att.com) and Verizon (www.verizon.com) both provide CDPD service.

Although CDPD systems have met a fair amount of success in the marketplace, CDPD is likely
to be replaced rather quickly by other technology. The most likely successors to CDPD are web
access technologies and protocols, such as wireless application technology running over digital
wireless systems. CDPD technology is an ancillary value stream in analog cellular systems; and
as such, it probably won’t stand on its own from an economic standpoint.

Cellular Antennas

There are two main types of antennas used for cellular radio transmission: 1) omni-directional
antennas, and 2) directional antennas. As their name implies, omni-directional antennas send
radio waves of equal strength in all directions (as shown in Figure 6-3). Since the signal is
transmitted out in all directions with equal strength, the signal reaching a receiver from an omni-
directional antenna is only a small percentage of the original signal strength. However, omni-
directional antennas are less expensive than are their directional counterparts; therefore, they are
the preferred choice when signal strength does not pose a problem. Omni-directional antennas
are also widely applied because they do not suffer from orientation issues in the field. Directional
antennas must be oriented to broadcast toward their intended target, making them susceptible to
being moved out of alignment by debris, vandals, improper installation, and so on.

6-9
Communication Technologies

Figure 6-3
Example of an Omni-Directional Antenna and Resulting Coverage Pattern

Directional antennas use additional elements to focus their signal transmission or reception in
one direction—typically over a range of 35-100 degrees (whereas omni-directional antennas
operate over 360 degrees). Although there are many types of directional antennas, the one most
often used for cellular communication in capacitor automation systems is the yagi. Yagi antennas
are used for broadcast or reception of weak signals or to cover longer distances. However,
because they are directional, the antenna must be oriented in the direction of its intended target.

Figure 6-4
Example of a Yagi Directional Antenna and Resulting Coverage Pattern

6-10
Communication Technologies

Commercial Support for Communication Planning and Analysis

There are several commercial companies that specialize in wireless communication planning and
support. There services may be particularly helpful for those considering building and operating
a private communication network. These companies typically provide services in:
• Requirement modeling
• Communication system design
• Transmission pathway assessment
• Turnkey planning
• License application
• Spectrum management and auctions

Some companies also offer software packages that can aid the utility engineer in communication
system scoping, design, and administration. ATDI (www.atdi.com) offers both consulting
services and software packages. Their software is designed for networks operating in the 10-kHz
to 450 GHz range and provides radio network design, spectrum administration, and cartographic
data management.

6-11
7
CAPACITOR CONTROLLERS AND SCADA SYSTEMS

Capacitor Controllers

The capacitor controller is really the backbone of the automated switched capacitor system. Both
local control schemes and centralized control schemes utilize a local capacitor controller. At the
most basic level, the controller provides the interface to the capacitor switch telling it when to
open and close. In local control schemes, the controller provides the switching logic. In central
control schemes, the controller 1) houses and interprets the signals provided by the data radio,
2) provides switching override functions based on local conditions, and 3) provides switching
logic in the event that communication with the central station is lost.

There are many models of capacitor controllers available from numerous manufacturers. The
controllers are typically packaged in weatherproof enclosures and are intended to be mounted on
the same pole as the capacitor bank and switch. Some examples of capacitor controllers from
several manufacturers are shown in Figure 7-1.

Figure 7-1
Examples of Capacitor Controllers from Several Manufacturers

Since there is a high level of variability from one utility’s capacitor control needs to the next,
there is also a correspondingly high level or variety in the features that are available among the
currently produced capacitor controllers. Most manufacturers try to cover most, if not all, of the
possible features that a utility may require, including:
• Communication: None (local control only), one-way, two-way
• Communication Channel: Radio, cellular, fiber optic, paging, copper line, and so on.
• Control Type: Volt, current, VAR, time, temperature, and combination control
• Monitoring: Some controllers with two-way communication ability also report data on a
variety of parameters—voltage, current, watts, power factor, and temperature

7-1
Capacitor Controllers and SCADA Systems

• Data Storage: Some controllers can store operational data locally for retrieval by utility field
personnel via laptop computer.
• Reverse Power Detection: As part of their monitoring capability, some controllers can detect
reverse power conditions on feeder. Additionally, some controllers have the functionality to
calculate proper set points and compensate for atypical line measurements during reverse
power flow conditions.
• Neutral Current Monitoring: Monitoring the capacitor bank’s neutral current can help
diagnose problems such as blown fuses, failing capacitor units, and high harmonic currents.
Further information on neutral current monitoring is available in Chapter 5, “Voltage and
Current Measurements.”

Most controllers have functionality for all “local” control types (volt, current, VAR, time,
temperature, and so on); and they can often run a combination program incorporating two or
more of these parameters in a hierarchical manner. Most manufacturers also cover both local
control and centralized control with one- or two-way communication capabilities, often by
providing different models, each with distinct communication capabilities. A sampling of the
currently available capacitor controls is shown below. This list does not represent all controller
manufacturers, and it is not meant to be all-inclusive for the offerings from any individual
manufacturer. Rather this list is intended to provide an overview of the products currently
offered in the market place at the time of writing and to give the reader a brief summary of the
capabilities of some of the product features being offered. Neither EPRI nor EPRI Solutions, Inc.
endorses a particular controller manufacturer.

Controllers that have either one-way or two-way communications require a radio to


communicate over the chosen medium (900 MHz, modem, paging, and so on.). Several
manufacturers offer the radio as a factory-installed option that can be specified during ordering.
It is also common for the manufacturer to leave a space inside the controller enclosure for
mounting the radio, thus enabling field installations at a later date or custom radio applications to
be made by a third party. Communication technologies are discussed in more detail later in this
chapter.

Beckwith Electric (www.beckwithelectric.com)


Model: M2501-B

Communication: None standard. Optional communication modules are available as shown below:

M-2937 CAMP™ Remote Communication Module (provides two-way SCADA


communications via SkyTel®’s Data Telemetry Network)

M-2980 CAMP™ Utilinet® Remote Communication Module (provides two-way


SCADA communication via UtiliNet® Series II WanGate Radio)

Control: The control makes its switching decisions based on algorithms, using time and voltage
measurements. Also, upper and lower voltage limits may be set and fixed voltage set-point levels
may be implemented. The control will not call for a switch operation, which would cause the

7-2
Capacitor Controllers and SCADA Systems

voltage to violate these voltage levels. If the site voltage drifts outside these settings, the control
will act through the adaptable nonlinear timers to switch the bank if necessary. The control
adapts to system source impedances by responding to small voltage changes near a substation
and larger voltage changes near the end of the line.

Figure 7-2
Beckwith Electric’s M-2501B Autodaptive® Capacitor Control (left), M-2937 CAMP™
Remote Communication Module (middle) and M-2980 CAMP™ Utilinet® Remote
Communication Module (right)
Courtesy of Beckwith Electric

Cannon Technologies (www.cannontech.com)


Model: CBC-5000 Series shown in Figure 7-3.

Communication: 900 MHz FLEX® paging

Control: The Cannon Technologies approach uses local controllers guided by centralized
capacitor switching software running on the SCADA master station. The program utilizes Power
Factor feedback to provide optimum power factor for all feeders, during all load conditions. If
SCADA communication is not desired, then the CBC series of controllers providing local control
is based on configurable overvoltage/under-voltage override thresholds.

Note: Cannon Technologies also offers the Universal CBC-7000 Series shown in Figure 7-3. The
CBC-7000 utilizes a modular design to accommodate a wider variety of communication formats,
including two-way communication systems.

7-3
Capacitor Controllers and SCADA Systems

Figure 7-3
Cannon Technologies’ CBC-5000 (left) and CBC-7000 (right) Remote Power Factor Control
Courtesy of Cannon Technologies

S&C Electric Company (formerly EnergyLine Systems) (www.sandc.com)


Model: Intellicap® Automatic Capacitor Control shown in Figure 7-4.

Communication: None (local control only)

Control: VAR and current control strategies, voltage, time, temperature, and time-biased voltage.
Also makes use of voltage/temperature override.

Figure 7-4
S&C Electric’s Intellicap® Automatic Capacitor Control
Courtesy of S & C Electric

Model: Intellicap PLUS® Automatic Capacitor Control shown in Figure 7-5.

Communication: Compatible with a wide range of communication types, including:


• Schlumberger's UtiliNet®
• Schlumberger's CellNet®
• Locus Radio
• MDS Adaptive Broadband

7-4
Capacitor Controllers and SCADA Systems

• Modems (Bell 202, CDPD, or Hayes™-compatible)


• Cellular telephones
• Fiber optic transceivers
• Telemetric™ DNP Remote Telemetry Module™
• Motorola® Darcom™

The communication module (radio, modem, and so on) can be factory mounted or installed by
utility personnel in the field. There is space reserved inside the controller enclosure to
accommodate many communication modules.

Control: SCADA central station control, VAR and current control strategies, voltage, time,
temperature, time-biased voltage, time-biased temperature. Also makes use of
voltage/temperature override. Further, there is a neutral current/voltage option of detection of
blown fuses and stuck switches. The neutral current/voltage capability is standard in all software,
which makes retrofitting as easy as adding a sensor.

Figure 7-5
S&C Electric’s Intellicap PLUS® Automatic Capacitor Control
Courtesy of S & C Electric

Fisher Pierce (now part of Joslyn Hi-Voltage_ (www.joslynhv.com)

Model: AutoCap™ Series 4400 shown in Figure 7-6

Communication: None (local control only)

Control: VAR, voltage, current, temperature, time, and combination programs. The AutoCap™
Series 4400 utilizes Adaptive VAr™ control by measuring the size of the bank, and establishing
VAR set points for maximum energy loss reduction. The controller also has Adaptive Voltage
Guard™ technology, which identifies voltage change from bank switching and inhibits “Close
Operation” if Voltage Guard plus present line voltage exceeds high-voltage limit. The controller
also includes complete load data and event recording; and it automatically corrects for
installation errors, such as sensing voltage and current from different phases, reversed current
signal wiring, and reversed Trip/Close wiring.

7-5
Capacitor Controllers and SCADA Systems

Figure 7-6
A Fisher Pierce AutoCap™ Series 4400 Capacitor Control
Courtesy of Fisher Pierce / Joslyn Hi-Voltage

Model: AutoCap™ Series 4500 shown in Figure 7-7.

Communication: Cellular, modem, or radio-based, two-way communications

Control: The AutoCap™ Series 4500 provides much the same control as the Series 440 shown in
Figure 7-6 but adds two-way SCADA communication capabilities. In addition to VAR, voltage,
current, temperature, time, combination programs, and Adaptive VAr™ control, the Series 4500
can also be configured to operate under SCADA control or as a local control with SCADA
override.

Figure 7-7
A Fisher Pierce AutoCap™ Series 4500 Capacitor Control
Courtesy of Fisher Pierce / Joslyn Hi-Voltage

Note: These controllers represent two of the many models of capacitor controllers offered by
Fisher Pierce / Joslyn Hi-Voltage. Among other products, they also offer other one-way and two-
way controllers, as well as special-purpose devices, such as neutral current alarm indicators.

7-6
Capacitor Controllers and SCADA Systems

Maysteel Electric LLC (www.maysteel.com)

Model: ProCap™ 140T/150T shown in Figure 7-8.

Communication: None (local control only)

Control: Time, voltage, temperature, or combination program—all with voltage/temperature


override. All of Maysteel’s capacitor controllers feature Adaptive Trip technology in which the
control anticipates voltage changes (which are caused by the opening and closing of the capacitor
bank) by measuring the line voltage before and after each switching operation. The stored
information is used to prevent or defer unnecessary capacitor bank switching operations,
extending the life of the switches and the capacitors.

Figure 7-8
ProCap™ 150T Capacitor Controller by Maysteel LLC
Courtesy of Maysteel LLC

Note: Maysteel LLC offers several other models of capacitor controllers featuring different
control schemes, such as time control with voltage override, time control with
voltage/temperature override, and VAR control with voltage override.

QEI Inc. (www.qeiinc.com)

Model: MicroCap or MiniCap Capacitor Switching Controllers

Communication: None (local control only)

Control: Voltage, corrected voltage, current, power factor, VAR, temperature, date, time, day of
the week, or a combination program of these parameters. Both of these controllers can also
detect reverse power conditions on the line.

Note: The MiniCap controller is the same as the MicroCap, except it adds a two-line LCD
Display and an additional set of switches that can be used to set or edit most configuration
parameters in the field, without the use of a laptop computer.

7-7
Capacitor Controllers and SCADA Systems

Figure 7-9
MicroCap (left) and MiniCap (right) Capacitor Switching Controllers from QEI, Inc.
Courtesy of QEI Inc.

Model: eCAP-9040 shown in Figure 7-10

Communication: Two-way communication by way of radio, cellular, or fiber optic. The


communication radio is not included, but there is space reserved within the closure for mounting
a radio.

Control: Voltage, corrected voltage, current, power factor, VAR, temperature, date, time, day of
the week, or a combination program of these parameters. The eCAP-9040 also allows for remote
control through SCADA using the DNP3 protocol. This controller also offers reverse power
detection and can monitor and report on numerous line parameters, including voltage, current,
VAR flow, and power factor.

Figure 7-10
Capacitor Switching Controller eCAP-9040, QEI Inc.
Courtesy of QEI Inc.

SCADA Overview

Basic SCADA systems, also referred to as telecontrol systems, consist of a master station(s)
communicating with one or more RTUs to provide data acquisition and control functionality
between a central location and dispersed field units. A very simplistic diagram of a SCADA
system is shown in Figure 7-11 to illustrate the concept of centralized control of dispersed field
units. The communication channel between the master controller and remote units can be one of
many technologies, including radio, cellular, modem, or hardwired networks. There are

7-8
Capacitor Controllers and SCADA Systems

numerous protocols available that define how communications between the master station and
remote units should be structured over the communication channel; however, the DNP3 protocol
tends to dominate new capacitor control systems. The master station runs application software
which provides the human-machine interface and also provides the functionality to perform the
specific tasks for which the SCADA system is used (that is, capacitor control, process control,
data acquisition). Alternatively, in larger, multi-function SCADA systems, the master station
may provide overall coordination and data archival while dedicated servers run individual
function programs, such as distribution capacitor control as illustrated in Figure 7-12.

Figure 7-11
Components of a Basic SCADA System

Capacitor Controller

Capacitor Controller

Master Station Capacitor Control Server

Capacitor Controller

Figure 7-12
Example of Basic SCADA Based Centralized Capacitor Control Using a Master Station and
a Dedicated Capacitor Control Server

7-9
Capacitor Controllers and SCADA Systems

Master Stations

In the past, a SCADA master station was part of a turn-key SCADA installation delivered by one
of many manufacturers and vendors working in the utility automation arena. However,
technology advancements in the personal computer and its subsequent proliferation into the
utility industry has altered the nature of the SCADA master station, while maintaining or even
enhancing its functionality. Today’s master stations tend to be one or more personal computers
or dedicated servers running master-station software specific to the system it is overseeing. The
master station can be configured such that it utilizes several parallel servers, all interfacing with
each other and running dedicated control programs, such as capacitor control, substation bus
voltage control, and automated meter reading. The system can also be configured such that the
master station software, while running on a single server, is accessible from many workstations
within the utility’s LAN, WAN, or intranet.

One example of a master station capacitor control product that operates as a parallel unit to the
primary SCADA master station is the RCC-2000W capacitor control central station from Radio
Controlled Central Stations (www.rccscontrols.com). The RCC-2000W augments a utility's
primary SCADA/EMS facilities by off-loading down-line VAR management to a Windows NT
processor. The RCC-2000W obtains near-real-time substation load and voltage data directly
from a utility's existing SCADA facilities, and it determines switching requirements using
closed-loop, VAR- and voltage-control algorithms. The device dispatches switching commands
down-line using a variety of private or commercial wireless means. It monitors circuit and
transformer VARs and immediately reports switches or capacitors that fail to operate.

Kansas City Power & Light’s (KCPL) engineers needed a system that could be utilized from
numerous computers within the company. To make the capacitor control system accessible to
individuals throughout the company, KCPL chose a client/server approach to connect the data
from their CellNet communication provider to KCPL's corporate network. To achieve this level
of functionality, data coming form the CellNet monitoring network is converted to
manufacturing message specifications (MMS) which can then be transmitted throughout KCPL's
corporate TCP/IP network. Users anywhere on the network have access to live distribution
automation data via PC-based workstations with Cycle LiveData Client Software. WinMon, a
GUI developed as a collaboration between EnergyLine and Kansas City Power & Light, is used
to view and organize the data.

Protocols

Protocols provide the rules for remotely located computers and master station computers to
communicate data and control commands to each other. While DNP3 is the most widely used
protocol for capacitor automation, there are other protocols in use, such as UCA, IEC 61850-5,
and MODBUS.

Traditionally, each equipment manufacturer had its own proprietary communications protocol,
which resulted in the industry being flooded with hundreds of proprietary protocols. Since each
protocol was proprietary, it was very difficult to obtain interoperability between devices,
especially devices from different manufacturers. This situation meant that utilities were often

7-10
Capacitor Controllers and SCADA Systems

forced to decide on equipment based on the protocols it supported rather than on its features and
performance. Then in the late 1980s and early 1990s, efforts to migrate toward widespread
interoperability began to take hold. These efforts included the development of the MMS and IEC
870. Further outgrowth of these early endeavors include UCA, DNP3, and IEC 60870.

DNP3 and IEC 60870-5 were designed for transmitting data-acquisition and control commands
from one computer to another. They are intended for transmitting relatively small packets of data
in a reliable manner, with the messages involved arriving in a determined sequence. These are
not designed as general purpose protocol for hypertext, multimedia, or huge files, such as FTP,
but rather are intended for SCADA applications (Clarke, Reynders et al. 2004; Short 2004b)

The following discussion is intended only to provide some background on the different protocols
available for use in capacitor automation systems. A good working handbook for any of these
protocols would be extensive and beyond the scope of this work. However, references provided
at the end of this section will guide the reader to further information about these protocols.
Before selecting a system, it also may prove beneficial to seek the guidance of a SCADA or
protocol professional.

Distributed Network Protocol (DNP3)

DNP3 is a protocol for transmission of data from point A to point B using serial and IP
communications. DNP3 is intended to provide standards, based interoperability between
substation computers, RTUs, Intelligent Electronic Devices (IEDs) and master stations. The
protocol of DNP3 has also become widely utilized by electric utilities, water and wastewater
industries, transportation systems, and oil and gas industries. From its inception, DNP3 has
become (and remains) the most widely used protocol for capacitor controllers (DNP Users Group
2005); and it is supported by a plethora of vendors and users in various utility and other
industries. Currently, DNP3 holds the lion share of SCADA applications in North and South
America, South Africa, Asia, Australia, and New Zealand. The IEC 60870-5 protocol tends, of
course, to dominate the European market.

DNP3 originated in the American electric industry with Harris Distributed Automation Products.
The protocol was based on the International Electrotechnical Commission’s (IEC’s), IEC 870-5
draft standard for SCADA protocol (now known as IEC 60870-5). DNP3 is an open and public
protocol with complete documentation available to the public. Since it is a public domain
protocol, ownership of DNP3 falls to the DNP3 Users Group. The group consists of utilities and
vendors who utilize the protocol and are responsible for its evolution. The Users Group’s
Technical Committee evaluates suggested modifications and amends the protocol accordingly.
Complete documentation for DNP3 is available from the DNP Users Group for a nominal fee
(see Internet link at http://www.dnp.org/).

Since it is based on the IEC 60870 requirements, DNP3 is suitable for application in the entire
SCADA environment. DNP3 supports numerous communications, including master-slave, peer-
to-peer, and network communication architectures. DNP3 also supports multiple master
configurations. The protocol allows extensions, while still providing interoperability between

7-11
Capacitor Controllers and SCADA Systems

multi-vendor devices. Data objects can be added to the protocol without affecting the way that
devices interoperate (Short 2004b). Some of the other benefits of DNP3 are that it is:
• An open standard,
• Supported by a large, active users group,
• Supported by a large and increasing number of equipment manufacturers, and
• Optimized for use in SCADA systems.

Figure 7-13
The ISO Seven-Layer, Open Systems, Interconnection Model

DNP3 is based on the International Standards Organization’s (ISO’s) seven-layer, open systems,
interconnection model, shown in Figure 7-13. Rather than incorporate all seven layers defined in
the model, DNP3 trims this down to a three-layer construction with limited support for the
transport and network layers. It is important to note that the seven-layer interconnection model is
just that—a model and not a protocol specification. It provides definitions and a framework for
defining communication protocols but does not explicitly detail a protocol itself. While the
seven-layer model is robust, it is too cumbersome for SCADA systems; since it provides some
high-level functionality that is not necessary for SCADA operations. To better meet the needs of
SCADA users, the IEC addressed this issue with the creation of the enhanced performance
architecture (EPA) model. The EPA three-layer model uses the application, data, and link layers
shown in Figure 7-13. The DNP3 protocol is based on the EPA three-layer model but adds
limited network and transport layers, as shown in Figure 7-14.

7-12
Capacitor Controllers and SCADA Systems

DNP3 Implementation Using Enhanced


Performance Architecture Model
Layers

Application Layer 3

Limited Transport Layer -

Limited Network Layer -

Data Link Layer 2

Physical Layer 1

Figure 7-14
DNP3 Implementation Using the Enhanced Performance Architecture (EPA) Model

The following provides a very brief overview of the roles of the various layers used in DNP3.
For a more detailed discussion, the reader is referred to (Clarke, Reynders et al. 2004; Short
2004b):
• Physical Layer: Represents the physical communication infrastructure over which the
protocol signals are transmitted. This layer is defined in the models by its physical
characteristics for information transfer, such as electrical specifications, timing, and pinouts.
The actual specifications of the physical layer are usually provided by a separate standard
such as RS-232.
• Data Link Layer: The data link layer provides the framework for the transmission of data on
the physical network. The data link layer typically provides flow-control and error-checking
functionality.
• Limited Network and Transport Layers: These layers are used to facilitate transmission of
larger blocks of data. Network functions manage routing and flow-control of data. Transport
functions provide transparent, end-to-end delivery of messages, including disassembly,
reassembly, and error correction.
• Application Layer: The application level is the highest level in the protocol and is provided
as the functionality for requesting and sending data.

Although originally intended for low bandwidth bit-serial communications, additional


functionality has been developed into DNP3 to provide communication over networks using the
TCP/IP protocol suite.

IEC 60870

IEC 60870 is comprised of a set of standards written (between 1988 and 2000) by the IEC to
govern the transmission of SCADA telemetry control and information. Like DNP3, IEC 60870 is
also an open standard which strives for interoperability between equipment from many vendors.
The standard was originally referenced as IEC 870 with the prefix “60” added at a later date. IEC

7-13
Capacitor Controllers and SCADA Systems

60870 is widely used in Europe, but has found only limited success outside of Europe, mostly
due to competition with DNP3. IEC 60870 was written with electric utility industry applications
in mind, although it can be used in SCADA systems for other industries as well.

IEC 60870 utilizes a hierarchical structure of six main parts. Each part is comprised of multiple
sections that have been published separately in a progressive manner. Part 5, “Transmission
Protocols,” also makes use of four companion parts, each of which offer detailed information for
applying the standard to particular applications. The four companion parts focus the functionality
provided by Part 5 by adding specific information data objects for individual applications. The
overall structure of IEC 60870 is shown below, with detail given for Part 5, “Transmission
Protocols.”

IEC 60870 – Telecontrol Equipment and Systems


• Part 1: General Consideration
• Part 2: Operating Conditions
• Part 3: Interfaces
• Part 4: Performance Requirement
• Part 5: Transmission Protocols
– Section 1 – Transmission Frame Formats
– Section 2 – Link Transmission Procedures
– Section 3 – General Structure of Application Data
– Section 4 – Definition of Coding and Application Information Elements
– Section 5 – Basic Application Functions
• Part 5 Companion Standards
– Part 5-101: Basic Telecontrol Tasks
– Part 5-102: Transmission of Integrated Totals in Electric Power Systems
– Part 5-103: Informative Interface of Protection Equipment
– Part 5-104: Network Access for IEC 60870-5-101 Using Standard Transport Protocols
• Part 6: Telecontrol Protocols Compatible with ISO Standards and ITU-T Recommendation

The sections of Part 5, “Transmission Protocols,” provide the following functionality:


• Section 1 - Transmission Frame Formats: This section describes the services provided to the
higher layers by the physical and data link layers. There is a choice of four data link frame
types—FT1.1, FT1.2, FT2, and FT3—each of which provides a different level of security
against data errors.
• Section 2 - Link Transmission Procedures: This section represents the four-frame formats of
Section 1 and describes the transmission procedures.

7-14
Capacitor Controllers and SCADA Systems

• Section 3 - General Structure of Application Data: This section provides two models of the
structure of data at the application level.
• Section 4 - Definition of Coding and Application Information Elements: This section
provides the rules for the definition of information elements. It also defines a common set of
information elements that may be used for transmission of information in Telecontrol
applications.
• Section 5 - Basic Application Functions: This section describes the highest level protocol
functions. The highest level functions are those application functions above layer seven of
the ISO Seven-Layer Model shown in Figure 7-13.

Although it is often referenced as IEC 60870 or IEC 60870-5, most discussion regarding
SCADA applications of this standard actually speak to the companion standard IEC 60870-5-
101. This companion provides the application-level data objects, as well as background
information necessary to sufficiently define the SCADA transmission protocol. However,
Part 5-101 does not stand on its own, as it contains many references to the sections (5-1 through
5-4) of Part 5.

The functionality for communication of the TCP/IP protocol suite is found in companion
standard Part 5-104, “Network Access for IEC 60870-5-101 Using Standard Transport Protocol.”
As is the case with DNP3, this functionality was not originally part of the standard but was added
later, in this case, through a companion standard to Part 5.

IEC 60870, through Part 5-101, supports point-to-point and multi-drop communication links
carrying bit-serial, low-bandwidth serial communications. It also supports balanced or
unbalanced communications (in balanced systems, any station can act as a sending or receiving
station at the same time).

Utility Communications Architecture

UCA began as the first in a series of projects under EPRI’s Integrated Utility Communication
(IUC) program as an effort to provide interoperability between computer systems supplied to the
utility industry. UCA is a standards-based approach to utility communications aimed at
providing wide-scale integration of utility automation equipment from a variety of vendors. It is
meant for use in any type of utility, including electric, gas, and water/wastewater. UCA is
referred to as architecture rather than protocol; because it incorporates a family of protocols in
order to meet the wide-ranging needs of the various utilities it serves. UCA is not likely to
replace DNP3 anytime soon, due to DNP3’s widespread usage and functionality, but it is more
likely that the two will compliment each other as SCADA technology continues to advance.

UCA is intended to allow utilities to purchase off-the-shelf hardware that is UCA compliant and
that will automatically integrate into their SCADA systems. Once the UCA-compliant device is
connected, the data relevant to that device will be automatically transferred to the SCADA and
IT systems that identify themselves as requiring it.

7-15
Capacitor Controllers and SCADA Systems

Although initiated by EPRI, the development of UCA is now managed by the Institute of
Electrical and Electronic Engineers (IEEE). In 1999, the IEEE published UCA Version 2.0 as
standard IEEE-SA TR 1550-1999. Version 2 addresses issues identified during field testing of
the original version and adds functionality for the internet suite of protocols (IEEE-SA TR 1550-
1999; Clarke, Reynders et al. 2004).

UCA consists of three main building blocks:


• A uniform communication infrastructure
• A uniform data model
• A uniform application interface

UCA allows access with all current LAN and WAN technology and also specifies asynchronous
data link control (ADLC) technology for multi-drop, serial phone and radio links. ADLC should
allow for future interoperability between digital radios from different manufacturers.
Manufacturing message specification (MMS) is used by UCA to provide the message structure,
message syntax, and message dialog procedures for monitoring and control communication.
There has also been discussion within the industry of providing functionality to allow UCA to
use DNP3 rather than MMS. While the idea of using UCA with protocols other than MMS is a
basic part of the UCA philosophy, it is likely to take several years to develop this functionality, if
it is added at all (Clarke, Reynders et al. 2004).

MODBUS

The MODBUS protocol was originally developed by Gould Modicon, now Schneider Electric,
for use with their programmable logic controllers (PLCs). It differs from many other protocols in
that MODBUS does not provide an interface definition. Since MODBUS does not incorporate an
interface definition, the MODBUS user is free to choose between EIA-232, EIA-422, EIA-485,
and 20-mA-current loop. MODBUS is a relatively slow transmission protocol but that slowness
carries the benefit of making it compatible with all of these communication interfaces. Another
advantage of the MODBUS protocol is that it has found wide acceptance in industrial settings.
Since DNP3 and IEC 60870-5 (to a lesser extent in North America) have become near ubiquitous
for distribution automation applications, however, they make questionable (at best) the future of
MODBUS in utility applications.

RTUs, IEDs, and PLCs

RTUs are ruggedly constructed, specialized computers used for data acquisition and control in
SCADA systems. The RTU may contain analog-to-digital converters, digital-to-analog
converters, digital inputs, and digital outputs. The RTU tends to be specially built for field use,
meaning that they are hardened against electrical transients, typically comply with ANSI IEEE
C37.90a, and also operate over a wide temperature range. RTUs are often used to poll other
electrical equipment (such as meters, relays, and reclosers) in order to collect system information
and send it back to the master computer. Large substation RTUs are often floor standing units
with their own enclosures, and they may accommodate thousands of input signals.

7-16
Capacitor Controllers and SCADA Systems

The term intelligent electronic device does not have a clear definition in the electric power
industry. Over time, it has evolved to encompass any device that incorporates one or more of the
following characteristics:
• Some degree of advanced, local-control intelligence
• Versatile electrical protection functions
• Data acquisition and reporting capabilities
• Ability to communicate directly with SCADA systems

Among other uses, IEDs can take the form of meters, relays, reclosers, voltage and VAR
regulators, and tap-position indicators. Generally, IEDs can interface directly into the SCADA
system and do not need to communicate through an external RTU.

PLCs have been used to implement relay and control systems for many years. As PLC
technologies have advanced, the distinctions between PLC and RTU have become blurred. PLCs
were initially less expensive than RTUs; but as PLC functionality (and complexity) increased,
PLC costs increased as well.

Modern switched-capacitor controllers are sometimes referred to as intelligent electronic devices


(although so are modern, multi-function relays and other equipment, so the term IED is rather
ambiguous). In most of today’s centralized capacitor control systems, the switched-capacitor
controller provides the functions of the RTU (or PLC), sensors, and data radio, all rolled into one
small and cost-efficient package.

SCADA Security

Security is an issue for all SCADA systems, both at the data-communications level (in part due
to the widespread use of open source protocols) and at the master-system level. When the
industry was dominated by proprietary systems, they tended to rely on “security by obscurity,”
meaning that there were numerous systems, each of which had restricted access to information
about them. This obscurity made it difficult for would-be attackers to gain insight into their
function. However, information regarding today’s open standards, including most operating
manuals and background documentation, are readily available through the internet. The readily
available nature of this information, coupled with “plug and play” devices, provides many
potential avenues for attacking today’s SCADA systems. Additionally, modern SCADA systems
are increasingly interconnected to other information networks, including the Internet, which
further exposes them to new vulnerabilities. SCADA systems can be compromised using similar
hacking methods to those employed on corporate networking systems, and a connection to the
Internet often serves as a main route for hackers to infiltrate the SCADA system.

Given the mission-critical nature of most SCADA systems (capacitor automation systems
included), there are several steps that should be taken to help bolster security and minimize the
chances for the system being hacked (Pollet 2002):

7-17
Capacitor Controllers and SCADA Systems

• Install Firewalls: Use a firewall between the Internet and your corporate networking systems
AND install good firewall protection to wall off your SCADA system from the Internet and
corporate networking system. The idea here is to put at least two quality firewalls between
the SCADA systems and the Internet.
• Company-Wide Policies and Procedures: This applies to how personnel access the
company’s computer systems. Use stringent password requirements: no dictionary words, at
least one capital and one number, appropriate length (8 characters is recommended),
regularly changed, and absolutely no names. Do not affix passwords directly to computer
monitors. Require each operator to have a unique password rather than allowing several
people to share a password. Ensure that personnel log out rather than leaving workstations
unattended or constantly logged-in. Automatic log-out due to inactivity can be handy in these
cases. It is also important to swiftly remove the USER IDs of staff who leave the company,
regardless of the reasons for their departure. Most importantly, these procedures must be
strictly enforced on a continual basis!

Major Barry C. Ezell of the United States Army conducted research into the vulnerability of
SCADA systems utilized in the water industry. By surveying a great number of water
facilities, Ezell determined that the most significant risk to the surveyed systems came from
directly attacking the SCADA master system. Furthermore, Ezell found that the most likely
source of the threat was a previous employee—someone with inside information (Ezell 1998;
Ezell 2005).
• Operating Systems: Maintain proper security on the operating systems running the SCADA
applications.
• Internal SCADA Design: Segment SCADA systems onto their own IP segment. Use smart-
switches instead of hubs and use proper sub-masking techniques.
• SCADA Applications: Adequately address security within the SCADA applications.

This list is by no means comprehensive. Every system has its own unique operation parameters;
and not all of the suggestions mentioned above are necessary in every situation. When in doubt, a
professional specializing in SCADA security should be consulted. Additionally, a quick scan of
the above precautions shows the need to assemble an interdisciplinary team of operations
personnel and information technology (IT) personnel. Although we are dealing with SCADA
system security, many aspects of providing that security are more suited for personnel with IT or
computer science experience.

Up to this point we have only considered so called cyber-threats where the system’s operating
methods are attacked. However, a very real (and arguably more likely) threat is that of a physical
attack on hardware and operations centers. Physical attacks on hardware can range from simple
vandalism (broken antennas and gunshot damage are rather common problems for utility
equipment) to the outright assault by terrorist elements. While physical protection of SCADA
infrastructure is beyond the scope of this report, it is important to note that it is a real concern for
practical SCADA implementations; and specialists in the protection of physical assets should be
consulted to ensure the highest possible level of security is achieved for critical infrastructure.

7-18
8
SOFTWARE AND DATA APPLICATIONS

Capacitor Control Software for SCADA Systems

As previously mentioned in Chapter 7, master controllers used in modern SCADA systems are
usually comprised of dedicated master controller software running on one or more personal
computers or servers. The master station can be configured such that it utilizes several parallel
servers, all interfacing with each other and all running dedicated control programs, such as
capacitor control, substation bus voltage control, and automated meter reading. The system can
also be configured such that the master station software, while running on a single server, is
accessible from many workstations within the utility’s LAN, WAN, or Intranet, as illustrated in
Figure 8-1.

Engineer

Engineer

Capacitor Controller

Operator Communication System

Capacitor Control Server


Capacitor Controller
Operator

Line Crew
Capacitor Controller

Technician

Figure 8-1
Example of Multiple Interfaces to Single Capacitor Control System

8-1
Software and Data Applications

Some master station software packages allow role-based log-in to the system to determine the
level of control given to the current user. For example, an engineer role at login may be granted
full access to able to force controllers to switch, to reconfigure set points, and to review historical
data. In contrast, a line-crew role at login may be limited to read-only access of one-line
diagrams and circuit maps. In this manner, role-defined login schemes allow the flexibility to
tailor available functionality to the user’s specific needs. This helps limit unintentional parameter
changes and also helps to deter malicious attacks on the system.

SCADA master controller software (or dedicated master station computers) and dedicated
capacitor control programs are available from a variety of manufacturers, including:
• NTMC® 2005 from the BTE Corporation (www.btecentral.com)
• RCC-2000W Capacitor Control Central Station from RCCS, Inc. (www.rccscontrols.com)
• Yukon™ Advanced Energy Services Platform and Capacitor Control Server from Cannon
Technologies (www.cannontech.com)

Utilities currently have a tremendous range of technologies and equipment to choose from when
building a capacitor control automation system. Because of this, there tends to be a great variety
of technologies and system components in use from one utility’s central capacitor control system
to the next. As would be expected, utilities try to work with their existing infrastructure as much
as possible when designing a new centralized capacitor control system. This may mean
attempting to interface with an existing SCADA system or installing a new SCADA system if
none existed before. Some utilities may opt for a dedicated capacitor control server that operates
as a part of an overall SCADA system. Others may look for dedicated master control hardware
and software if this is their first foray into distribution automation.

In a centralized control scheme with routine switching decisions made by the master controller,
different line parameters, such as voltage, current, or VAR flow, are transmitted to the master
controller as inputs to the switching algorithm. The master controller analyzes these inputs on a
regular basis, every half-hour for example, and decides if the capacitor should be switched from
its current state. If it needs to be switched, the master controller sends an open or close command
to the controller. With one-way communication, the master controller will then examine the
VAR flow on the feeder to verify that the capacitor operated as intended. If two-way
communication is enabled, the capacitor controller may send positive verification back to the
master controller to verify switching. By examining this chain of command, it becomes evident
which information needs to be displayed to the operator. The most basic information really
consists of just a few parameters:
• Capacitor Switch State: Is the capacitor switch open or closed?
• Switching Criteria Data: For example, VARs if operating under VAR control or voltage if
operating under voltage control.

Most control systems go far beyond this basic information. In fact, many modern capacitor
controllers, when equipped with two-way communication, can report back on a variety of circuit
parameters to provide the distribution engineer great insight into what is happening on the line.
There is, therefore, the potential for a great deal of information to be made available for display:

8-2
Software and Data Applications

• Line Parameters: Reports several line parameters, including voltage, current, VARs, power
factor, temperature, and so on
• Control Location: Tells if the capacitor is being controller locally or remotely (centralized
control)
• Control Type: Describes what control method is currently being used and/or if switching is
currently running autonomously or is it being forced by the operator?
• Time, Date, and Reason of Last Operation: Provides information on the time and the reason
for the last time the capacitor switched
• Cycle Counter: Counts the number of times the capacitor has cycled (today, this month, this
year, and so on)
• Neutral Sensor: Identifies the neutral current?

Device and Data Management Software

A host of manufacturers provide software for device or data management. Many of the device
management products are proprietary to a specific device or family of devices. For example,
most capacitor controller manufacturers provide proprietary software for configuring their
devices. Modern relay manufacturers take a similar approach. Some of this software is intended
to be run on a laptop computer to interface with devices in the field, while others are intended for
networked operation in centralized capacitor control systems.

S&C Electric Company (www.sandc.com) produces a device management software package


called WinMon® that is capable of use in centralized capacitor control systems. WinMon® is a
GUI that provides on-line access to real-time and historical system data, device configuration,
and device operating reports. An example screen shot from the WinMon® software is shown in
Figure 8-2. KCPL uses WinMon® to provide the interface between their SCADA system,
communication network, and the KCPL operators (Goeckeler 1997). Using the WinMon®
interface, KCPL staff can access all of the hundreds of automated capacitors on the system,
either individually or in various groupings. KCPL operators can examine near real-time line
parameters, historical information, and controller set points for every automated controller on the
system. The ability to address groups of capacitors has also proven useful for KCPL. On August
19, 1998, KCPL experienced an equipment failure at its Hawthorn generating station, which
brought the plant off-line. The next day, the local 161-kV transmission system experienced
sagging voltage, causing the transmission operators to issue a request for VAR support to the
distribution operators. Using the WinMon® interface, KCPL was able to quickly query and
switch in all available distribution capacitors, thus providing VAR support for the transmission
system. The result was an increase in the transmission voltage of approximately 1.5-kV and a
much flatter voltage profile (Goeckeler 1999).

8-3
Software and Data Applications

Figure 8-2
Example Screen from WinMon® Graphical User Interface
Courtesy of S&C Electric Company

Human-Machine Interface Issues

The human-machine interface is a major challenge with any informational system, but especially
with systems that produce vast amounts of data. The data must be presented in a manner that
enables the operator to process the information, make decisions based on the data, and not
become unnecessarily fatigued in the process. Modern SCADA systems tend to utilize good on-
screen information layout with hierarchical structures that can be navigated intuitively. This
helps to reduce operator fatigue as well as ease the learning curve for system operators. Since
people adapt more easily to the familiar, it is beneficial for new software to present information
in a manner that resembles what users are accustomed to seeing. For instance, distribution
operators who are already accustomed to DOS-based control software will adapt more easily to
new software if it has a similar appearance and layout. This also helps to make training easier
and gets operators comfortable with software more quickly.

User customizable data layout and location control can also make software easier to learn and
more efficient to operate on a daily basis. Not all users have the same responsibilities or tasks.
While it makes sense for distribution engineers to be able to manipulate a controller’s set points,
distribution dispatchers do not need to have ready access to those features, unless it is part of

8-4
Software and Data Applications

their daily tasks. And any on-screen information that isn’t useful is simply taking up space that
could be used more efficiently.

8-5
9
CAPACITOR AND CONTROLLER SURGE PROTECTION

Primary Arrester Lead Length and Coordination with Fuses

Arrester versus fuse placement is an ongoing industry debate. An arrester on the load side of the
fuse, right at the equipment, protects the equipment best; but since the equipment is downstream
of the fuse, a lightning surge current will pass through the fuse. Lightning may blow a small fuse
unnecessarily, and many utilities have histories of nuisance fuse operations. Applying the
arrester upstream of the fuse keeps the surge current out of the fuse but usually results in long
lead lengths. Overall, the tank-mounted approach is best, along with using larger fuses or
surge-resistant fuses to limit unnecessary fuse operations.

Arresters upstream of fuses and exposed on the primary mainline have several disadvantages:
• An arrester failure forces an operation by a mainline protective device, interrupting many
more customers than if a fuse had operated.
• If the arrester isolator fails to operate (which can happen), the failure may be extremely hard
for the crews to find. A fuse helps localize the failure.
• If the arrester fails and the isolator operates, crews may re-close the circuit successfully if
they do not find the failed arrester. This leaves the equipment unprotected. Worse yet, the
failed arrester body may start to track across the bracket. Eventually the arrester bracket will
flash over, causing a hard-to-find permanent fault.
• With arresters upstream of a fuse, arrester lead lengths will be longer; and the equipment is
not protected as well against lightning surges. This can lead to more lightning-caused
equipment failures.
• If an arrester fails internally, the failure will be less violent if it is cleared by a fuse. The
arrester will absorb much more energy if a mainline protective device has to clear the fault,
because clearing the fault takes considerably longer.

It is not just failures of the arrester itself that can cause problems. Arresters also provide a prime
location for squirrels and other animals to cause faults. Arresters can also track externally.

Lead Length Considerations

The lead length component is very important; the lead voltage can contribute as much as the
arrester protective level for long lengths. The arrester lead inductance is approximately 0.4
µH/foot (1.3 µH/m). Commonly, a rate of current change is assumed to be 20 kA/µs. Together,

9-1
Capacitor and Controller Surge Protection

this is 8 kV per foot of lead length (26 kV/m). This is not an unreasonable rate of rise to use in
the calculation, as 20 kA/µs is about the median value for subsequent strokes during the rise
from 30% to 90% of the surge magnitude.

Lead lengths less than three feet (one meter) are often necessary to achieve a 50% margin for
protecting overhead equipment on 13.8-kV distribution circuits. The easiest approach is to
tank-mount arresters. Pole or crossarm mounting makes it harder to keep reasonable lead lengths.
It is important to remember that lead length includes the ground lead as well as the phase wire
lead (see Figure 9-1). The easiest way to trace the lead-length path is to follow the path that
lightning current would take as it flows from the phase wire to ground.

Figure 9-1
Arrester Lead Length

Some obvious but important directions for arrester application are:


• Don’t Coil Leads: While this may look tidy, the inductance is very high.
• Tie Ground Lead to the Tank: The NESC (IEEE C2-1997) requires arrester ground leads to
be tied to an appropriate ground. To achieve any protection, the ground lead must be tied to
the tank of the equipment being protected. Without attaching the ground lead to the tank, the
transformer or other equipment is left completely unprotected.
• Optimize Arrester Performance: Frame-mount the arresters on the capacitor support
structure to reduce the arrester lead length, and connect the arresters on the load side of the
fuse.

9-2
Capacitor and Controller Surge Protection

Ground Lead Length (G.L.) Line Lead Length (L.L.)

Unjacketed Concentric or
Semiconducting Jacketed
Direct Buried Cable Used
as a Pole Ground

From IEEE Std. 1299/C62.22.1-1996. Copyright 1997 IEEE. All rights reserved.

Figure 9-2
Example of Considerable Lead Length on a Riser Pole

Figure 9-2 shows an example of excessive lead length while Figure 9-3 shows how to obtain
almost zero lead length. To obtain minimum lead length on the phase side, jump the lead to the
arrester first, then jump it to the equipment. For the ground-side arrester lead (the most neglected
portion), jump the lead from the arrester directly to the tank or other equipment bonding point.
For riser-pole applications, jump both leads to the concentric neutral and to the pole ground. In
Figure 9-3, if jumper 1 were left off, the ground lead length would be very high. If jumper 2 were
instead connected to the cable sheath at the termination, then the current would flow through
jumper 1. This would increase the lead length by the length of jumper 1. While this may still be
acceptable, it is not as good as the configuration in Figure 9-3.

9-3
Capacitor and Controller Surge Protection

1
2

From IEEE Std. 1299/C62.22.1-1996. Copyright 1997 IEEE. All rights reserved.

Figure 9-3
Example of Almost Zero Lead Length on a Riser Pole

To ensure that equipment is protected properly, arresters must be located at the same pole as the
equipment. Moving arresters to adjacent poles will leave the equipment totally unprotected.
Figure 9-4 shows a simulation using EPRI’s Lightning Protection Design Workstation (LPDW)
where an equipment pole (could be a capacitor, a re-closer, a riser pole, or a transformer bank) is
left unprotected. The protection is moved to adjacent poles. For a strike right at the arrester
location, the arrester helps hold down the voltage at the strike point; but at the equipment pole,
the voltage across the equipment insulation reaches several hundred kilovolts. This voltage is
well above the basic insulation level of 95 kV.

9-4
Capacitor and Controller Surge Protection

Figure 9-4
Simulation of Protection Provided by Arresters at Adjacent Poles Only

Arrester Installation Clearance Considerations

Distribution arresters have isolators that remove failed arresters from the circuit. The isolator has
an explosive cartridge that blows the end off of a failed arrester. This provides an external
indication of failure. The isolator itself is not designed to clear the fault; thus, an upstream
protective device normally must clear the fault (although, in a few cases, the isolator may clear
the fault on its own, depending on the available short-circuit current and other parameters).

Crews should take care with the end lead that attaches to the bottom of the arrester. It should be
mounted in a manner that prevents the isolator from being able to swing the lead into an

9-5
Capacitor and Controller Surge Protection

energized conductor if the isolator were to operate. The proper lead size should also be used
(making sure it is not too stiff, which might prevent the lead from dropping). Figure 9-5 shows
an example of an arrester lead that has dropped dangerously close to the energized bushing of a
transformer. An animal contact could easily cause a line-to-ground fault in this situation.
Consideration should also be given to unexpected failure modes, such as the phase end of the
arrester breaking free of the mounting bracket. An attempt should be made to minimize the risk
of the phase breaking free and causing the same types of power quality (PQ) problems that were
discussed for the isolator lead. However, this failure mode is less common and is not as high a
priority as a swinging isolator lead issue when considering potential arrester mounting
configurations.

Figure 9-5
Blown Arrester with a Dangling Ground Lead

Capacitor Controller Surge Protection

Utilities have a substantial investment in distribution line capacitors. These investments are
justified based on certain derived benefits to the power delivery system, utilities, and end-users.
When capacitors are off line or are not available for operation (due to some failure or operating
error), those benefits will not be achieved. Experience at utilities revealed that capacitors were
unavailable for operation too frequently; therefore; this project series was established to improve
capacitor reliability. Initial scoping identified and prioritized several issues affecting the overall
reliability of capacitors, and work prior to this year addressed some of those issues. The purpose
of this work is to analyze and evaluate the potential impact of a lightning surge on the controller
for a switched, medium-voltage distribution capacitor bank.

9-6
Capacitor and Controller Surge Protection

A typical switched capacitor bank installation is shown in Figure 9-6. Although Figure 9-6 only
shows the capacitor assembly near the pole top, the capacitor controller is mounted lower on the
pole—approximately 10 ft (3 m) above the ground. While there are many types of controllers on
the market, with many different configurations, an attempt was made to choose a fairly typical
arrangement for this analysis. Most capacitor controllers are designed to plug into a standard
meter socket, and are typically located about ten ft above the ground on the utility pole. The
controller data inputs consist of a voltage signal and possibly a current signal as well. Often the
CPT secondary voltage is sufficiently accurate to serve as a voltage input. In some cases,
separate voltage transducers are used. If current input is required, it will be obtained from one or
more current transducers in series with the phase conductors, mounted with the insulators at the
top of the pole. If separate voltage transducers are used, they could be mounted here also. The
output of the controller consists of two 120 Vac signals, which operate the open and close
contactors of the capacitor switches that are located on the capacitor rack.

Figure 9-6
Example of Switched Capacitor Bank Configuration

Modeling of Lightning Surges Originating on the Primary Conductors

A simplified circuit model was constructed using power systems computer aided design
(PSCAD), with the objective of evaluating the surge voltage reaching the controller through its
voltage input terminals. It was assumed for this initial work that this would be the worst case of
surge transmission. The lightning surge was modeled as a 1.2 x 50 µs impulse wave with a peak

9-7
Capacitor and Controller Surge Protection

of 50 kA. This magnitude represents the 95th percentile of the cumulative probability distribution
for first-stroke peak current.2 For comparison, 50% of all natural first strokes exceed 31 kA.

Lightning-caused overvoltages on both the CPT secondary and the controller input were
investigated. All cases examined utilized primary surge arresters. As stated earlier, most of the
surge coupling for the CPT is capacitive. and it is believed that this also holds true for the
transducers. Therefore, it is estimated that the nature of the surges reaching the signal ports will
be similar to those reaching the power input terminals. PSCAD simulation results are shown in
Figure 9-7 and Figure 9-8 for the following scenarios:
1. No Secondary Surge Arrester (No Arrester): Here, the surge voltages at both the CPT
secondary and controller input are in the range of 40 kV. If the controller is rated to
withstand a 5 kV surge, additional protection may be necessary. This protection should
coordinate with the internal motor-operated valves (MOVs) of the controller.
2. Surge Arrester Near or on the Secondary Terminals of CPT (Arrester 1): In this case,
voltages were limited on both the CPT secondary and controller input, and they were less
than 5 kV—within the ratings of typical controllers. The surge arrester should coordinate
with the surge withstand rating and internal MOV protection of the controller.
3. Surge Arrester Near or on the Input to Capacitor Controller (Arrester 2): In this case the
controller input voltage was limited to approximately 1 kV, within the ratings of typical
controllers. The surge arrester should coordinate with the surge withstand rating and internal
MOV protection of the controller.
4. Neutral of 120-V Input to Controller Grounded Near the Controller Without Secondary
Surge Arrester (Loop): This condition results in what is called a ground loop. The parallel
path allows the surge current to travel down the secondary neutral conductor and (to a lesser
extent) to the hot conductor. A resulting drop in voltage in the range of 150 kV appears
across the controller input. Since such a voltage is not sustainable in this system, it would
likely result in equipment failure. It is recommended, therefore, that the secondary circuit be
grounded at only one point.
5. Neutral of 120-V Input to Controller Grounded Near the Controller, With Surge Arrester
Near or On the Secondary Output of CPT (Arrester 1 and Loop): This is similar to the
ground loop in the preceding case 4 without an arrester. The resulting peak of nearly 150 kV
is similar as well. It is unlikely that such a surge could be sustained without arcing.
6. Neutral of 120-V Input to Controller Grounded With Surge Arrester Near or On Input to
Controller (Arrester 2 and Loop): This is similar to the previous case 5, except that the
voltage peak appears across the CPT secondary instead of the controller. It is unlikely that
such a surge could be sustained without arcing.

2
A. Greenwood, Electrical Transients in Power Systems, New York, John Wiley, 1991.

9-8
Capacitor and Controller Surge Protection

Figure 9-7
CPT Secondary Voltage for Scenarios with and Without Secondary Arrester and Ground
Loop

9-9
Capacitor and Controller Surge Protection

Figure 9-8
Voltage at the Controller Terminals for Scenarios with and Without Secondary Arrester
and Ground Loop

Preliminary Recommendations

The surge exposure of capacitor controllers was examined under a variety of surge origins and
circuit configurations. Surges due to lightning striking the primary and secondary conductors
were investigated, as well as surges due to single line-to-ground faults occurring past the
capacitor controller. Additionally, multiple construction configurations were considered,
including locating the capacitor controller and the CPT on the same pole and mounting them on
separate poles, one or more pole spans apart.

9-10
Capacitor and Controller Surge Protection

Due to the many different capacitor controllers currently available and in use, it is extremely
difficult to provide one set of recommendations that apply to all controller installations.
However, the body of work contained in this report makes it possible to offer preliminary
guidelines that can help the utility engineer increase reliability in installing capacitor controllers.
The following recommendations are based on the transient modeling performed for this project:

Mounting Location: Whenever possible, locate the capacitor controller and CPT on the
same pole.

Lead Length: It is imperative that the length of all surge arrester leads be kept as short as
possible.

Grounding and Shielding: Shielded control cables can greatly reduce the induced surge
magnitude on the control wires, from current flow through the ground loop created by the
controller neutral. Therefore, shielded control cables should be used whenever possible.
If a shielded cable is not used, then the control cable and pole-down ground cable should
be run as far apart as possible. Also the grounding options that follow should be
considered for breaking the ground loop.

If the controller and CPT are on the same pole and there are no other loads on the secondary:
• Ground the secondary neutral through a surge arrester at the controller input
terminal.
• Do not ground the secondary neutral at the transformer output terminals.
• If possible, do not ground the secondary neutral in the capacitor junction box at the pole top.
However, grounding the neutral through a surge arrester in the capacitor junction box may
also offer satisfactory results. Note: the case ground connection should still be maintained!
Not grounding the controller case may present a safety issue.

If the controller and CPT are not on the same pole:


• Even optimal surge protection may not prevent a controller failure. It will be very difficult to
avoid ground loops.
• Apply auxiliary secondary surge protection.
• Investigate leaving the controller power input ungrounded (although this may result in a
flashover). Another option is to ground the neutral at the controller through a high-discharge
surge arrester (possibly a 600-V arrester). Note: the case ground connection should still be
maintained! Not grounding the controller case may present a safety issue.

Primary Surge Protectors: Install primary arrestors on the same pole as the capacitor
assembly in order to limit the primary voltage rise. However, a structure flashover is
often unavoidable for direct lightning strikes.

Secondary Surge Protectors: Give serious consideration to adding auxiliary,


secondary-surge protection for all installations.

9-11
Capacitor and Controller Surge Protection

• Add auxiliary, secondary-surge protection when the controller and CPT are not on the same
pole.
• Add auxiliary, secondary-surge protection for installations in high lightning areas or areas
with a history of controller failures.

When additional secondary arrestors are installed, do the following:


• Limit lead length by locating surge arresters at the input terminals to the controller (this is
often accomplished by making the arrester connections at the jaws in the socket base).
• Use care to select arrestors that will coordinate with the impulse withstand and internal
MOVs in the controller unit. The reader is strongly encouraged to consult with the capacitor
controller manufacturer prior to adding additional surge protection.

Key Considerations

There are several key issues to consider when evaluating a capacitor controller installation, with
regard to surge exposure and protection. Some considerations, such as pole grounding resistance,
may not be completely within the utility’s control. However, the utility will have control of the
two considerations with the greatest impact; namely, the controller mounting location and the
grounding configuration.

Controller Mounting Location

The placement of the capacitor controller (in relation to the CPT from which it is supplied) plays
a prominent role in the magnitude of transients reaching the controller. Locating the capacitor
controller and the CPT on the same pole provides the best surge performance. If possible,
therefore, the two units should be located on the same pole for optimum surge performance.

The capacitor control cable that runs between the controller and pole-top junction box should be
located as far as possible from the pole-down ground lead. In most cases, this means that the
down ground lead and control cable should be located on opposite sides of the pole. Maintaining
the greatest possible separation of the down ground and control cable 1) minimizes the chances
of flashover from the down ground to the control cable and 2) should allow for the least surge
induction in the control cable from surge current flowing in the ground lead.

Ground Loops and Shielding

Grounding at both the secondary-side of the CPT and the power input terminal to the capacitor
controller creates a ground loop (Figure 9-10) which causes the controller to be exposed to
higher magnitude transients than when the ground loop is not in place. Using a shielded
controller cable is one method for protecting the controller from the ground loop effects. A
shielded control cable will eliminate inductive coupling from current flow through the controller
neutral; and it will also block capacitively-coupled voltages. The shield of the cable should serve
as the controller neutral conductor. Therefore, a separate insulated neutral should not be run

9-12
Capacitor and Controller Surge Protection

inside the shielded cable. Both ends of the shield should be bonded to ground. The shield should
also be of sufficient construction to conduct the lightning surge current and possible power-
follow-through current without being damaged (consult the cable manufacturer to verify the
shield’s current-carrying capability). Figure 9-9 shows an example configuration using a shielded
control cable.

Figure 9-9
Example Configuration Using Shielded Control Cable

Whether or not shielded cable is used, consideration should also be given to avoiding ground
loops. In order to break a ground loop, the circuit should be grounded in only one place. Many
capacitor controller manufacturers provide separate secondary-neutral and case-ground lugs on
the controller to allow for not grounding the controller power input at the controller. That is not

9-13
Capacitor and Controller Surge Protection

to say that the capacitor controller case should be left ungrounded! When recommended by the
manufacturer, the case ground connection should always be maintained, as not grounding the
case could present a life-threatening safety issue.

The main criteria for avoiding a ground loop is to only make one connection to ground in the
circuit, be it at the CPT secondary, the capacitor controller power input, or somewhere else.
When possible, the ground connection should be provided at the controller power input without
grounding the neutral at the CPT secondary or pole-top junction box. However, this
configuration is not always feasible. Not all capacitor controllers are powered from a dedicated
transformer, thus there may be other customers sharing the secondary wires with the capacitor
controller. If this is the case, there will be ground connections at each customer’s service
entrance. The use of a common conductor for both the system neutral and the secondary neutral
also makes it difficult to avoid including a ground loop in the system wiring, especially when the
controller is not on the same pole as the transformer. Furthermore, when the controller and
transformer are not on the same pole, each pole will require a down ground; thus, a common
neutral will be grounded in at least two places.

Figure 9-10
Ground Loop Created by Grounding the CPT Output and the Capacitor Controller Neutral
Terminal

9-14
Capacitor and Controller Surge Protection

Arrester Lead Length

Arrester lead length includes both the primary-lead length and the ground-lead length. The total
lead length is measured from the point at which the arrester line connection is made to the point
where interconnection is made between the arrester ground lead and the protected equipment
ground lead, excluding the arrester length.
• Lead lengths must be kept as short as possible.
• Arresters should be mounted as close as possible to the objects they are protecting. When
possible, arresters should be mounted directly on or next to the equipment ports they are
protecting.

Short lead lengths are required because the leads contain inductance; therefore, they can cause
L(di/dt) voltage drops across the leads when they conduct lightning surge current. The voltage
drop across the arrester lead adds to the peak voltage appearing across the equipment that the
arrester is protecting. The lead voltage adds to the arrester discharge voltage only during the rise
of the discharge current. For standard lightning surge current test waves, this can add to the
voltage appearing across the protected equipment 1.6 kV per foot of lead length. For actual
lightning currents, which posses even greater rise times, the L(di/dt) voltage drop on the leads
can be even greater. For this reason, it is absolutely essential to use the shortest lead lengths
possible in arrester applications.

The internal connections of all capacitor controllers should be checked prior to installing the
controller in the field. The following should be verified when inspecting the controller’s internal
connections:
• Protection is provided for all surge modes. This means that the controller needs to have
protection covering line-to-neutral and line-to-line surge modes.
• All surge arrester leads are as short as possible.

As with the built-in surge protection, all auxiliary surge protection added to the controller must
use the shortest lead lengths possible. Minimizing arrester lead lengths is of the utmost
importance.

Auxiliary Surge Suppression

Depending on other installation considerations (mounting location, ground loops, and so on.), it
may be desirable to add auxiliary surge suppression to protect the capacitor controller. Most
likely, the secondary-side surge suppression will be provided by an MOV-technology-based unit,
such as the Cooper Power Systems Storm Trapper® or Storm Trapper® H.E. (high energy)
secondary surge suppressor, as shown in Figure 9-11. There are also other surge suppression
products available the make use of several technologies incorporated into two or more “stages”
of surge suppression. One example of this type of secondary surge suppressor is the Axiomatic
120Vac surge protector shown in Figure 9-12. It should be noted that EPRI and EPRI Solutions
do not endorse a particular surge suppression product. The surge suppression products shown
here are only for illustrative purposes. Currently in early 2005, most secondary-side surge

9-15
Capacitor and Controller Surge Protection

suppressors of this variety cost less than $100. However, depending on the number of controller
ports being protected, two or more protection units may be required.

Figure 9-11
Cooper Power Systems Storm Trapper H.E. Secondary Surge Arrester
Courtesy of Cooper Power Systems

Figure 9-12
Axiomatic 120Vac Surge Protector
Courtesy of Advanced Surge Suppressor

All surge modes for all controller ports should be covered for optimal surge protection. This
means that the incoming lines need to be protected for line-to-line and line-to-neutral surge
modes. Furthermore, the surge protection is best added directly across the capacitor controller
ports rather than at points further away. One possible surge-protection scheme is shown in
Figure 9-13. Note that Figure 9-13 is not drawn to scale. The size of the surge arrester enclosure
relative to the meter socket depends on the particular surge arresters used in the application. The
auxiliary surge protection in Figure 9-13 is not drawn to scale in order to provide a detailed view
of how the protection would be configured. In reality, the surge protection would probably look
more like what is shown in Figure 9-14, although that depends on exactly which capacitor
controller and surge protector(s) is utilized. It is essential to keep lead lengths as short as
possible. Depending on the equipment, locating the additional surge protection near the top of
the meter socket may offer the shortest lead lengths but will probably also make for a
troublesome installation, due to the number of wires that would need to fit into a small space.
Another possibility is to mount the surge protection on the lower portion of the meter socket.
This location also keeps lead lengths to a minimum while offering somewhat easier installation.
Of course, these are just two of the many possible configurations. The optimal mounting location

9-16
Capacitor and Controller Surge Protection

for additional surge protection will depend on the equipment being used in each individual
installation.

Figure 9-13
Example Configuration for Surge Protection Covering Incoming Lines for All Surge Modes

9-17
Capacitor and Controller Surge Protection

Control Cable to
Pole Top

Auxiliary Surge
Protection

Socket

120 V Neutral

Current Current
Lead Lead

Trip Close

Ground

Not Drawn to
Scale

Ground Wire to Pole


Ground

Figure 9-14
Approximate Size Relationship of Meter Socket and Typical Auxiliary Low-Side Surge
Protection (Note: Actual sizes will vary depending on what equipment is used)

It is not uncommon for utilities to also use arresters protection at the pole top junction box. When
used at the junction box, arresters are often connected between the phase conductor and ground
and as well as between the neutral conductor and ground. However, using a neutral-to-ground
arrester can create a ground loop if another neutral-to-ground arrester is used at the controller
input terminals, as suggested in this report. Therefore, the neutral and ground conductors would
be better be left isolated at the pole-top junction box and connected through a surge arrester at
the controller input.

At this point is should be mentioned that even the best possible capacitor controller surge
protection can not eliminate 100% of capacitor controller failures caused by lightning strikes.
Some lightning-caused surges are simply too damaging to be protected against. Furthermore,
there are many factors that may lead to less-than-optimal surge protection in certain
circumstances. Depending on the protective margins and the configuration of the built-in surge
protection, it may not be possible to coordinate the built-in surge protection with auxiliary surge
protection. Surge current will follow through the least resistant path to ground. If the built-in
surge protection is tightly applied (i.e., it begins to conduct near the rated line voltage), it may

9-18
Capacitor and Controller Surge Protection

not be possible to add additional protection that conducts at a lower voltage. Therefore, surge
currents will flow though the built-in protection, even if additional surge protection is added. In
general, the capacitor controller manufacturer should be consulted before adding additional surge
protection.

Pole Ground Resistance

Although it does not play as prominent a role as the mounting location or grounding
configuration, low pole-ground resistance can help reduce the magnitude of the transients
reaching the controller. This is particularly true for transients caused by faults beyond the
controller, as low pole-ground resistance can help provide an alternative return path to the
system neutral.

Consult the Manufacturer

It is recommended that the capacitor controller manufacturer be consulted before enacting the
recommendations given in this report. This includes using non-standard control power
configurations and adding auxiliary surge protection. There are a variety of unique capacitor
controllers currently available, and the recommendations given in this report may not be suited
for every available controller. When consulting the manufacturer, specific inquiries should
include:
• How would additional surge protection be properly coordinated.
• Is it suitable to not ground the power neutral at the input to the controller (if considering one
of the ungrounded connection options)?

Installation Guidelines

Due to both the many different capacitor controllers currently available (and in use) and the wide
array of possible installation configurations, it is extremely difficult to provide one set of
recommendations that were applicable to all controller installations. However, the body of work
contained in this report makes it possible to offer preliminary guidelines that will help the utility
engineer increase reliability with capacitor controller installation.

When planning for new capacitor controller installations (or reviewing existing installations), it
is important to review the key issues outlined above. The main point to consider is whether or
not the capacitor controller (and the CPT supplying it) can be mounted on the same pole.
• Ground loops occur when the neutral for the control power is grounded in multiple locations.
Where possible, ground loops in the controller power wiring should be avoided. To prevent
coupling to the controller conductors, a shielded cable can be used from the junction box at
the top of the pole to the controller. Alternatively, grounding only the neutral conductor at
one location is another option to consider.
• If the capacitor controller and CPT are on different poles, preventing a ground loop may not
be possible.

9-19
Capacitor and Controller Surge Protection

• If using unshielded controller conductors, the pole ground down lead and capacitor controller
cable should be located on opposite sides of the pole to maintain as large a separation as
possible.

Auxiliary secondary-side, surge protection should be considered for all capacitor installations,
but such protection is particularly important for installations with the following characteristics:
• Installations with the CPT and capacitor controller on different poles
• Installations in high lightning areas
• Installations with poor grounding
• Installations that have chronic controller failures

For maximum protection, all controller ports should be protected from all surge modes. Utility
engineers are reminded, however, that even optimal surge protection cannot prevent 100% of
surge-related controller failures. An installation with the CPT on one pole and the controller on a
different pole will be particularly troublesome, as it is challenging to avoid ground loops with
this configuration.

It is extremely important to verify that all arresters, built-in and auxiliary, have the shortest
possible lead lengths. Most capacitor controllers come equipped with some built-in surge
suppression with which auxiliary surge protection should be coordinated. The auxiliary surge
protection should have a lower voltage rating than the built-in surge protection (the lowest rated
arrester is the one that will absorb the energy). Manufacturers will be able to provide information
on the voltage rating of built-in surge arresters, so auxiliary arresters with equal or lower voltage
ratings can be selected.

9-20
10
CAPACITOR FUSING

Fusing Guidelines

The main purpose of the fuse on a capacitor bank is to clear a fault if the capacitor unit or any of
the accessories fail. The fuse must clear the fault quickly to prevent any of the equipment from
failing violently. Capacitor ruptures have historically been problematic, so fusing is normally
tight. Fuses must be sized to withstand normal currents, including harmonics.

A significant number of utilities have experienced problems with nuisance fuse operations on
capacitor banks. A fuse may be blown, but the capacitors themselves may remain functional.
These blown fuses may stay on the system for quite some time before they are noticed (see
Figure 10-1). Capacitors with blown fuses increase voltage imbalance. They can also increase
stray voltages and increase losses. Even if the capacitor controller is able to identify blown fuses,
the replacement requirements add to the workload of maintenance crews.

Figure 10-1
Capacitor Bank with a Blown Fuse (EPRI 1001691 2002)

10-1
Capacitor Fusing

IEEE guidelines suggest selecting a fuse capable of handling 1.25 to 1.35 times the nominal
capacitor current (IEEE Std. C37.48-1997); a 1.35 factor is most common. Three factors can
contribute to higher-than-expected current:
• Overvoltage: Capacitive current increases linearly with voltage, and the reactive VARs
increase as the square of the voltage. When estimating maximum currents, an upper voltage
limit of 110% is normally assumed.
• Harmonics: Capacitors can act as a sink for harmonics. This can increase the peak and the
rms of the current through the capacitor. Additionally, grounded, 3-phase banks absorb zero-
sequence harmonics from the system.
• Capacitor Tolerance: Capacitors are allowed to have a tolerance to 15% above their rating
(which would increase the current by 15%).

Most fusing practices are based on fusing as tightly as possible to prevent case rupture. So, the
overload capability of fuse links is included in fuse sizing. This effectively allows a tighter
fusing ratio. K and T tin links can be overloaded to 150%; so for these links with a 1.35 safety
factor, the smallest size fuse that can be used is:

1.35 I1
I min = = 0 .9 I 1 Eq. 10-1
1 .5

where,

Imin = minimum fuse rating in amperes

I1 = capacitor bank current in amperes

10-2
Capacitor Fusing

Table 10-1 shows one manufacturer’s recommendations based on this tight-fusing approach.

Table 10-1
Fusing Recommendations for ANSI Tin Links From One Manufacturer [Cooper Power
Systems, 1990]

3-Phase Bank System Line-to-Line Voltage, kV


kVAR
4.2 4.8 12.5 13.2 13.8 22.9 24.9 34.5

Recommended Fuse Link


150 20T 20T 8T 6T 6T
300 40K 40K 15T 12T 12T 8T 8T 5T
450 65K 50K 20T 20T 20T 10T 10T 8T
600 80K 65K 25T 25T 25T 15T 15T 10T
900 100K 40K 40K 40K 20T 20T 15T
1200 50K 50K 50K 30T 25T 20T
1800 80K 80K 80K 40K 40K 30K
2400 100K 100K 100K 65K 50K 40K

Fusing ratio for the recommended link (link rating/nominal current)


150 0.96 1.11 1.15 0.91 0.96
300 0.96 1.11 1.08 0.91 0.96 1.06 1.15 1.00
450 1.04 0.92 0.96 1.02 1.06 0.88 0.96 1.06
600 0.96 0.90 0.90 0.95 1.00 0.99 1.08 1.00
900 0.92 0.96 1.02 1.06 0.88 0.96 1.00
1200 0.90 0.95 1.00 0.99 0.90 1.00
1800 0.96 1.02 1.06 0.88 0.96 1.00
2400 0.90 0.95 1.00 1.07 0.90 1.00

Note: this is not the manufacturer’s most up-to-date fusing recommendation. It is provided mainly as an example of
a commonly applied fusing criteria for capacitors

With this tight-fusing strategy, fuses must be used consistently. If silver links are used instead of
tin links, the silver fuses can blow even from expected levels of current; because silver links
have no overload capability.

Prior to the 1970’s, a fusing factor of 1.65 was more common. Due to concerns about case
ruptures and PCB’s, the industry went to tighter fusing factors, 1.35 being the most common.

10-3
Capacitor Fusing

Because of the good performance of all-film capacitors (see the next section) and problems with
nuisance fuse operations, a looser fusing factor should be considered, possibly returning to the
1.65 factor. Slower fuses should also have fewer nuisance fuse operations.

Capacitors are rated to withstand 180% of rated rms current, including fundamental and
harmonic currents. Fusing is normally not based on this limit, and is typically much tighter than
this, usually from 125 to 165% of rated rms current. Occasionally, fuses in excess of 180% are
used. In severe harmonic environments (usually in commercial or industrial applications), fuses
normally blow before capacitors fail. Sometimes, however, capacitors fail before the fuse
operates, depending on the fusing strategy.

If a capacitor bank has a blown fuse, crews should test the capacitors before re-fusing. A
handheld digital capacitance meter is the most common approach, and it is an accurate measure.
Good multimeters also can measure a capacitance high enough to determine the capacitance on
medium-voltage units. There is a chance that capacitance-testers may miss some internal failures,
and thus require high voltage to break down the insulation at the failure. Measuring the
capacitance on all three phases helps identify units that may have partial failures. Partial failures
show up as a change in capacitance. In a partial failure, a series capacitor pack (one of several)
shorts out. The remaining packs appear as a lower impedance (higher capacitance). As with any
equipment about to be energized, crews should visually check the condition of the capacitor unit
and make sure there are no bulges, burn marks, or other signs that the unit may have suffered
damage.

Reasons for Relaxing Fusing

The industry move to all-film capacitors has dramatically reduced the tendency of capacitors to
fail violently. Capacitors can fail in two modes:
• Low current, progressive failure—The dielectric fails in one of the elements within the
capacitor (see Figure 10-2). With one element shorted, the remaining elements in the series
string have increased voltage and higher current (because the total capacitive impedance is
lower). With more stress, another element may short out. Failures can cascade until the
whole string shorts out. In this scenario, the current builds up slowly as elements
successively fail.
• High current—A low-impedance failure develops across the capacitor terminals or from a
phase terminal to ground. A broken connector could cause such a fault.

10-4
Capacitor Fusing

Series section

Individual
element

Failed element

Figure 10-2
Capacitor Unit with a Failed Element

Most failures are progressive. Sudden jumps to high current are rare. To detect progressive
failures quickly, fusing must be very sensitive.

Paper and paper-film capacitors have an insulating layer of paper between sheets of foil. When a
breakdown in a pack occurs, the arc burns the paper and generates gas. In progressive failures,
even though the current is only somewhat higher than normal load current, the sustained arcing
can create enough gas to rupture the enclosure. Before 1975, capacitors predominantly used
polychlorinated biphenyls (PCB’s) as the insulating liquid. Environmental regulations on PCB’s
greatly increased the costs of cleanup if these units ruptured (US Environmental Protection
Agency 40 CFR Part 761 Polychlorinated Biphenyls (PCBs) Manufacturing, Processing,
Distribution in Commerce, and Use Prohibitions). The environmental issues and safety concerns
led utilities to tighten up capacitor fusing.

In modern film-foil capacitors, sheets of polypropylene film dielectric separate layers of


aluminum foil. When the dielectric breaks down, the heat from the arc melts the film; the film
draws back; and the aluminum sheets weld together. With a solid weld, a single element can fail
and not create any gas (the current is still relatively low). In film-foil capacitors, the progressive
failure mode is much less likely to rupture the case. When all of the packs in series fail, high
current flows through the capacitor—this can generate enough heat and gas to rupture the
capacitor if it is not cleared quickly.

A 2002 EPRI survey found that case ruptures on modern film-foil capacitors are uncommon
(EPRI 1001691 2002). This gives us confidence that we can loosen fusing practices without
having rupture problems. This contrasts sharply with paper capacitors, where Newcomb reported
that film/paper capacitors ruptured in 25% of failures (Newcomb 1980).

In areas of high fault current, current-limiting fuses provide extra safety. Either a backup current-
limiting fuse in series with an expulsion link or a full-range current-limiting fuse is an
appropriate protection scheme in high fault-current areas. While it may seem that expulsion fuses
provide adequate protection even to 8 kA (depending on which rupture curve we use), current-
limiting fuses provide protection for those less frequent faults with longer internal arcs. They
also provide protection against failures in the capacitor switches and other capacitor-bank

10-5
Capacitor Fusing

accessories. Utilities that apply current-limiting fuses on capacitors normally do so for areas with
fault currents above 3 to 5 kA.

Maximum Fuse Sizes

The industry-recognized 1.35 factor is a recommended minimum fuse size. Industry standards do
not provide a recommended maximum; they do provide some application guidelines to address
the issue. Some of the main checks on going too large are:
• Capacitor rupture curves—Fuses should protect against rupture within capacitors.
• Ungrounded capacitors—On ungrounded capacitors, even a complete unit failure will not
only draw three times normal current through that unit’s fuse.
• Upstream coordination—If there is an upstream line device (normally a recloser), it may be
desirable to coordinate clearing, so that a fault downstream of the capacitor fuse does not trip
the upstream device.

Preventing case ruptures is a primary goal of fusing—the fuse should clear before capacitor cases
fail. Figure 10-3 shows capacitor rupture curves compared against fuse clearing curves. The
graph shows that there is considerable margin between fuse curves and rupture curves. Consider
a 12.47-kV, 900-kvar bank of three 300-kvar units, which has a nominal current of 41.7 A.
Utilities commonly use a 40 or 50-K fuse for this bank. Larger fuses for this bank are possible,
while still maintaining levels below case rupture curves.

10-6
Capacitor Fusing

40 65
100.0 50 80

10.0

Capacitor rupture curves


Time, seconds

Cooper

GE, 300 kvar and above


1.0

0.1
Fuse total clear curves
T links
K links

0.01
10+2 10+3 10+4

Current, amperes

Figure 10-3
Fuse Curves with Capacitor Rupture Curves

Ungrounded capacitors require tighter fusing than grounded capacitors because when one unit
fails, the floating neutral point shifts voltage—the failed unit does not draw full fault current, and
the unfailed units see 1.73 times normal voltage. During a progressive failure, when one series
group shorts out, the shift of the neutral relieves the voltage stress on the remaining series
groups. When one pack fails, we really want the fuse (or other protection) to trip quickly. Figure
10-4 compares the response of grounded and ungrounded capacitor banks. With a unit fully
failed, the failed unit draws 3 per unit. With half of the series groups failed, the unit draws 1.5
per unit. To ensure that an ANSI K or T fuse operates for a current of 3 per unit, the fuse must be
sized less than 3/2.2 = 1.36 times the capacitor rated current. The factor of 2.2 comes from the
fact that fuse total clearing curves start at a current of about 2.2 times their ampere rating. A
clearing time of 1 second is required to protect the capacitors on adjacent failures according to
IEEE Std. 1036. This is normally impossible to achieve (to achieve a 1-sec clearing time, the
fuse rating would have to be about 6 times lower than the desired clearing current of 3 per unit).
The IEEE Std. 1036 reports that good performance has been achieved by selecting the fastest
fusing (meeting the 1.35 criteria), which results in clearing times of 15 sec to 2 min.

10-7
Capacitor Fusing

Also consider the fusing differences if a normally grounded bank is converted to an ungrounded
bank because of harmonics or stray voltage problems.

Grounded wye Ungrounded wye

Partial failure:
Half of a unit’s V = 1 pu V = 0.75 pu
series groups I = 2 pu I = 1.5 pu
short out.

V = 1 pu V = 1 pu V = 1.15 pu V = 1.15 pu
I = 1 pu I = 1 pu I = 1.15 pu I = 1.15 pu

V = 0 pu
I = 3 pu
V = 0 pu
Full failure
I =bolted fault current

V = 1 pu V = 1 pu V = 1.73 pu V = 1.73 pu
I = 1 pu I = 1 pu I = 1.73 pu I = 1.73 pu

Figure 10-4
Comparison of Grounded-wye and Ungrounded-wye Banks During a Failure of One Unit

As an alternative to tight fuse protection on floating-wye configurations, we can use neutral


detection. A potential transformer measuring voltage between the floating neutral and ground
can detect a failure of one unit. When one unit fails, a relay monitoring the neutral potential
transformer (PT) should trip the capacitor’s oil or vacuum switch. Obviously, this only works on
switched banks.

Regarding coordination, larger capacitor fuses may cause miscoordination issues. One option is
to live with the miscoordination. After that, some adjustment of the capacitor fusing and/or
protective device setting can usually accomplish reasonable coordination.

Do not loosen fusing on older banks with paper or paper-film capacitors. Their much higher tank
rupture rate during failures requires tight fusing.

Nuisance Fuse Operation

Some utilities have problems with nuisance fuse operations on distribution transformers. Some
of the causes of capacitor fuse operations could be the same as transformer fuse operations, but
certain differences are apparent, such as:
• Capacitor fuses see almost continuous full load (when the capacitor is switched in).

10-8
Capacitor Fusing

• Capacitor fuses tend to be bigger than transformer fuses. Typical capacitor sizes are 300 to
1200 kVAR with 15- to 65-A fuses. The most common transformer size is 25 and 50 kVA,
typically with less than a 15-A fuse.
• Both capacitors and transformers have inrush, but a capacitor’s inrush is quicker.
• Transformers have secondary faults and core saturation that can contribute to nuisance fuse
operations while capacitors have neither.

Some possible causes of nuisance fuse operations on distribution capacitors include:


• Lightning: Capacitors serve as a low impedance to high-frequency lightning surges, so they
naturally attract lightning current, which can blow fuses. Smaller, faster fuses are most prone
to lightning. When using the standard guidelines that a fuse at least as big as 20K or 15T
should prevent nuisance operations, it appears that lightning itself should not cause a
significant number of fuse operations (as most capacitor bank fuses are larger than this).
• Outrush to Nearby Faults: If a capacitor dumps its stored charge into a nearby fault, the fuse
can blow. Capacitor banks also have inrush every time they are switched in, but this is well
below the melt point of the fuse.
• Severe Harmonics: Harmonics increase the current through the fuse.
• Animal or Other Bushing Faults: A fault across a bushing can blow a fuse, whether due to
an animal, contamination on the bushing, or tree contact. By the time anyone notices the
blown fuse, the squirrel or branch has disappeared. Animal guards and covered jumpers
should be used to reduce bushing fault incidents.
• Mechanical Damage and Deterioration: Corrosion and vibration can weaken fuse links.
Ontario Hydro collected fuse links from the field on transformers; and they found that 3%
had broken strain wires (CEA 288 D 747 1998). Another 15% had braids that were brittle
and had broken strands. Larger fuses used in capacitors should not have as much of a
problem.
• Installation Errors: Fuses are more likely to blow if crews put in the wrong size or wrong
type of fuse or if they do not properly tighten the braid on the fuse.

Outrush and Inrush

Outrush is highlighted as a possible failure mode that has been neglected by the power industry.
Outrush is sometimes considered for station banks in calculating the probability of a fuse
operation from a failure of an adjacent parallel unit. But the effects of nearby faults on
distribution fuse operations have not been considered.

The energy input into the fuse during outrush depends on the line resistance between the
capacitor and the fault (see Figure 10-5). The capacitor has stored energy; and when the fault
occurs, the capacitor discharges its energy into the resistance between the capacitor and the fault.
Closer faults discharge more energy into the fuse. The I2t that the fuse suffers during outrush to a
line-to-ground fault is:

10-9
Capacitor Fusing

1 2
CV pk 2.65QkVAR
I t= =
2 2 2
V pu Eq. 10-2
R R

where,

C = capacitance of one unit, µF

Vpk = peak voltage on the capacitor at the instant of the fault, kV

R = resistance between the capacitor and the fault, ohms

QkVAR = single-phase reactive power, kVAR

Vpu = voltage at the instant of the fault in per-unit of the capacitor’s rated voltage

Figure 10-5
Outrush from a Capacitor to a Nearby Fault

10-10
Capacitor Fusing

Table 10-2
I2t Comparisons on a 3-phase, 1,200-kVAR Bank at 12.47 kV (Iload=55.6 A)

Source I2t, amperes2-seconds

Lightning, median 1st stroke 57,000


Lightning, median subsequent stroke 5,500
Inrush at nominal voltage (ISC=5 kA, X/R=8) 5,940
Inrush at 105% voltage 6,550
Outrush to a fault 500-feet away 35,000
Outrush to a fault 250-feet away 70,000
Outrush to a fault 250-feet away with an arc restrike* 280,000
65K fuse, minimum-melt I2t 90,000
65T fuse, minimum-melt I2t 271,000

*Assumes that the arc transient leaves a voltage of 2 per unit on the capacitor before the arc restrikes

10-11
Capacitor Fusing

Table 10-3
I2t Comparisons on a 3-phase, 600-kVAR Bank at 12.47 kV (Iload=27.8 A)

Source I2t, amperes2-seconds

Lightning, median 1st stroke 57,000


Lightning, median subsequent stroke 5,500
Inrush at nominal voltage (ISC=5 kA, X/R=8) 2,970
Inrush at 105% voltage 3,270
Outrush to a fault 500-feet away 17,500
Outrush to a fault 250-feet away 35,000
Outrush to a fault 250-feet away with an arc restrike* 152,000
30K fuse, minimum-melt I2t 21,200
30T fuse, minimum-melt I2t 65,500

*Assumes that the arc transient leaves a voltage of 2 per unit on the capacitor before the arc restrikes

Table 10-2 and Table 10-3 show several sources of fuse operations and the I2t that they generate
for different bank sizes at 12.47 kV. For the 1200-kVAR bank, the nominal load current is
55.6 A. Utilities commonly use 65-A fuses for this bank. The table shows the minimum melt I2t
of common fuses. Outrush to nearby faults produces energy high enough to blow common fuses,
especially ANSI K links, which are fast fuses. Of the other possible causes of fuse operation,
none are particularly high except for a lightning first stroke. The lightning data is misleading
because much of the first stroke will go elsewhere. Usually, the line flashes over, and much of
the lightning current diverts to the fault.

10-12
Capacitor Fusing

Figure 10-6
Outrush as a Function of the Resistance to the Fault for Various Size Capacitor Banks
(The sizes given are 3-phase kVAR; the resistance is the resistance around the loop, out
and back; the distances are to the fault)

Figure 10-6 can be used to find outrush I2t for other cases. Two factors make outrush worse:
• Higher System Voltages: The outrush I2t stays the same with increases in voltage for the
same size capacitor bank. The line impedance also stays the same for different voltages. But,
higher-voltage capacitor banks use smaller fuses with less I2t capability. A 25-kV capacitor
installation, therefore, is more likely to have nuisance fuse operations than a 12.5-kV system.
• Larger Conductors: Larger conductors have lower resistance.

Consider a 1200-kVAR bank with 500-kcmil conductors. At 12.47-kV (Iload=55.6 A) with a 65K
fuse, the fuse may exceed its minimum-melt I2t for faults up to 150 feet away. At 24.94 kV
(Iload=27.8 A) with a 30K fuse, the fuse may melt for faults up to 650 feet away. At 34.5 kV
(Iload=20.1 A) with a 25 K fuse, the location is off the chart (it’s about 950 feet). Note that the
distance scales in Figure 10-6 do not include two important resistances: the capacitor’s internal
resistance and the fuse’s resistance. Both will help reduce the I2t. Also, the minimum-melt I2t
values of the fuses in Figure 10-6 are the 60-Hz values. For high-frequency currents, like an

10-13
Capacitor Fusing

outrush discharge, the minimum-melt I2t of expulsion fuses is 30 to 70% of the 60-Hz I2t
(Burrage 1981).

As an estimate to how much outrush contributes to nuisance fuse operations, consider a 900-
kVAR bank as 12.47 kV with 40-K fuses. We’ll estimate that the fuse may blow or be severely
damaged for faults within 250 feet (76 m). Using a typical fault rate on distribution lines of 90
faults/100 miles per year (56 faults/100 km/year), faults within 250 feet (75 m) of a capacitor
occur at the rate of 0.085 per year. This translates into 8.5% fuse operations per capacitor bank
per year—a substantial number.

The stored energy on the fault depends on the timing of the fault, relative to the point on the
voltage wave. Unfortunately, most faults occur at or near the peak of the sinusoid.

Several system scenarios could make individual instances worse; and most are situations that
leave greater than normal voltage on the capacitor before it discharges into the fault. These
scenarios include:
• Regulation Overvoltages: Voltages above nominal will increase the outrush energy by the
voltage squared.
• Voltage Swells: If a line-to-ground fault on one phase causes a voltage swell on another and
the fault jumps to the “swelled” phase, higher-than-normal outrush will flow through the
fuse.
• Arc Restrikes: If a nearby arc is not solid, but sputters, arc restrikes (much like restrikes of
switches) can impress more voltage on the capacitor and subject the fuse to more energy,
possibly much more, depending on the severity. (No evidence suggests that this occurs
regularly. Most arcs are solid, and the system stays faulted once the arc bridges the gap.)
• Lightning: A nearby lightning strike to the line can charge up the capacitor (and start the
fuse heating). In most cases, the lightning will cause a nearby flashover, and the capacitor’s
charge will dump right back through the fuse.
• Multiple-Phase Faults: Line-to-line and 3-phase faults are more severe for two reasons—the
voltage is higher and the resistance is lower. On a line-to-line fault, for example, the voltage
is the line-to-line voltage, and the resistance is the resistance of the phase wires (rather than
the resistance of a phase wire and the neutral in series).

These estimates are conservative in that they don’t consider skin effects, which can have
considerable impact at high frequencies. Skin effects increase the conductor’s resistance. The
transients oscillate in the single-digit kilohertz range. At these frequencies, conductor resistance
increases by a factor of two to three. On the negative side, the fuse element is influenced by skin
effects too, as higher frequency transients will cause the fuse to melt more quickly.

10-14
Capacitor Fusing

Capacitors also have inrush every time they are energized. Inrush into grounded banks has a peak
current of:

I pk = 1 . 41 I SC I 1 Eq. 10-3

where,

Ipk = peak value of inrush current, A

ISC = available 3-phase fault current, A

I1 = capacitor bank current, A (IEEE Std. 1036-1992)

The energy into a fuse from inrush is normally very small—it subjects the capacitor fuse to an I2t
(in A2-s) of:

I 2 t = 2.65 1 + k 2 I SC I 1 / 1000 Eq. 10-4

where,

k = X/R ratio at the bank location (Brown 1979)

Inrush is much worse if a capacitor is switching into a system with a nearby capacitor. The
outrush from the already-energized bank dumps into the capacitor coming on-line. Fuses at both
banks see this transient. In substation applications, this back-to-back switching is a major design
consideration, often requiring insertion of reactors between banks. For distribution feeder
capacitors, the design constraints are not as large. A few hundred feet of separation is enough to
prevent inrush/outrush problems. For back-to-back switching, the I2t is almost the same as that
for outrush:

1 2
CV 2 . 65 Q kVAR
I t = =
2 2 pk 2
V pu Eq. 10-5
R R

The only difference is that the capacitance is the series combination of the two capacitances:
C=C1C2/(C1+C2), and QkVAR=Q1Q2/(Q1+Q2). For the same size banks, C=C1/2, and QkVAR=Q1/2.
Figure 10-6 applies if we double the kVAR values on the curves. In most situations, maintaining
a separation of 500 feet between capacitor banks prevents fuse operations from this
inrush/outrush. Capacitor banks should be separated by 500 feet (150 m) on 15-kV class circuits,
therefore, to avoid inrush problems. Large capacitor banks on higher voltage distribution systems
may require modestly larger separations.

10-15
Capacitor Fusing

Fuse Installation Issues

C. W. (Charlie) Williams of Progress Florida (formerly the Florida Power Corporation) reported
significant problems with capacitor fuses at the 2003 Southeastern Electric Exchange working
group on power quality and reliability (Dallas, TX, 9/11/2003). Williams reported that Progress
Florida has had 20 feeder lockouts in one year due to capacitor fuses failing to clear. Progress
Florida’s tests and investigations found two main problems:
• Water entry—Fuses left open in cutouts can fill with water. Water entering the fuseholder
can damage the fuse tube lining. With a damaged lining, a fuse can have trouble clearing
low-current faults.
• Loose fuse connections—If crews do not tighten the fuse link enough, the barrel of the cutout
will heat excessively because the threads of the barrel make a high-resistance connection.

Many utilities switch capacitors seasonally. The most common way is for crews to open cutouts.
The best way to do this is to remove each fuseholder from its cutout and hang the fuseholders by
their pull rings. If left open in the cutout, water can easily enter. Water can damage the fuse tube
liner and the cardboard tube surrounding the fuse element. These two elements both help when
clearing low-current faults. Expulsion fuses have the most difficult time clearing high fault
currents and low fault currents. For low fault currents, the fuse liner and the cardboard tube melt
to release non-ionized gas that helps quench the arc.

On capacitors, a partial unit failure causes a low-current fault. If the water-degraded cutout
cannot clear the fault, the fuse assembly may melt down, causing a fault on the system.

Heat from loose fuse connections can also facilitate cutout failure. Heat can weld the cutout’s
contacts, so it cannot drop clear. If the cutout cannot drop clear, the whole assembly can fail. If
the fuse clears, but the cutout sticks, the fuse tube will be supporting line-to-neutral voltage.
Since the fuse tube is not rated to support this voltage, tracking will start and eventually lead to
flashover.

Heat can also burn away the cardboard tube surrounding the fuse element. The cardboard tube
provides the fusing assembly with extra ability to clear low-current faults. Extra heating is more
likely in a capacitor cutout because a capacitor is always running at nearly full load.

Loose fuse connections may also increase the chance of nuisance fuse operations. The extra heat
could force the system into thermal runaway and operate the fuse, especially if the fuse is subject
to another stress—overvoltage, harmonics, or a transient. Figure 10-7 shows damage to fuse
tubes caused during tests of cutouts operated with loose fuse connections.

10-16
Capacitor Fusing

Figure 10-7
Damaged Fuse Tubes from Loose Connections
Courtesy of C. W. (Charlie) Williams at Progress Florida

Figure 10-8 shows an infrared scan taken during tests. The cutout with the loose fuse connection
was running at 350°F. Progress Florida tested the cutouts with 83 A of current (600 kvar at 7.2
kV). The test setup had been operating for two weeks when the infrared scan was taken.

Figure 10-8
Infrared Thermovision Scan of Cutouts Tested with 83 Amps of Current
Courtesy of C. W. (Charlie) Williams at Progress Florida

10-17
Capacitor Fusing

The two ways to reduce problems with cutouts are:


• Tighten fuses—Make sure crews tighten the fuse cap with a wrench. The button and cap need
to be tight to make a good contact.
• Do not leave fuses open in cutouts—Have crews hang fuseholders by the pull ring.

In addition, crews should inspect the fuse liner whenever they replace the fuse element. A water-
resistant synthetic fuse liner is another option for reducing the possibility of water damage to
fuse liners. Finally, utilities can use infrared thermography to identify the extent of heating from
loose connections in cutouts on capacitor banks (Short 2004a).

Proposed Fusing Guidelines

We recommend using both a larger fuse and a slower fuse than is normally used. Larger fuses are
more resistant to transients and to steady-state heating. With these guidelines, pick the next
largest fuse current rating meeting the following criteria:

K links
• 1.65 × capacitor rated current

T links
• 1.35 × capacitor rated current

Do not use the overload capability of the fuse in any case. Do not use less than a 15T or 20K
fuse. For candidate fuses larger than a 65T or 80K, make sure that the fuse total clearing curve
coordinates with capacitor rupture curves. Table 10-4 shows an example application of these
guidelines on a 12.47-kV system.

Do not use the looser fusing guidelines under the following conditions:
• Ungrounded banks
• Paper or film/paper capacitors

In addition, these guidelines are only for distribution line capacitors with one capacitor unit per
phase. For group fusing with one fuse protecting more than one capacitor unit, further analysis is
required, mainly to ensure that the capacitor units are protected from rupture.

10-18
Capacitor Fusing

Table 10-4
Example Fuse Application Guidelines for a 12.47/7.2-kV System

Rated New fuse guidelines


Three-phase Traditional
capacitor current,
kvar fuse K T
A

150
6.9 8T 20K 15T

300
13.9 15T 25K 20T

450
20.8 20T 40K 30T

600
27.8 25T 50K 40T

900
41.7 40K 80K 65T

1200
55.6 50K 80K or 100K* 80T*

*check capacitor rupture curves

Also, consider the following additional recommendations:


• Fuse installation — Make sure crews tighten the button with a wrench. Loose fittings cause
excessive heat because the threads make a poor connection. The extra heat could contribute
to nuisance fuse operations and could prevent the fuse from being able to clear low-current
faults from partial capacitor failures. Very important.
• Open fuses — Don’t leave fuses hanging open in the cutout. Have crews remove them and
hang them up, open end down. If fuses are hanging open in the cutout, water will enter and
degrade the fuse tube lining and the cardboard sheath around the cutout.
• Animal guards — Use animal guards on capacitors and other equipment on a capacitor pole
to reduce the chance of bushing faults.

10-19
11
CAPACITOR BANK POWER QUALITY AND
RELIABILITY ISSUES

Several problems contribute to the overall reliability (or unreliability) of capacitor banks. Kansas
City Power & Light (KCPL) used their automated capacitor bank system to perform a detailed
analysis of switched capacitor bank reliability. KCPL reported that blown fuses were their
biggest problem (Goeckeler 1999), but their capacitor bank maintenance needs were not limited
to fuses, as shown in Table 11-1. The failure rates shown in Table 11-1 are high—much higher
than most distribution equipment. Switched capacitor banks are complicated, and they have a lot
of components that can fail. However, failure rates should be better than those shown in
Table 11-1.

Table 11-1
Maintenance Needs Identified by KCPS’s Capacitor Automation System*

Annual
percent
Problem failures

Primary fuse blown to capacitor (nuisance fuse


operation) 9.1
Failed oil switches 8.1

Hardware accidentally set to "Local" or "Manual" 4.2


Defective capacitor unit 3.5

Miscellaneous 2.4

CPT 1.5
TOTAL 28.8

* Based on two years of data (Goeckeler 1999)

An EPRI survey on capacitor reliability found that utilities had widely differing experiences with
capacitors (EPRI 1001691 2002). Roughly one-third of the survey responses rated feeder
capacitors very good; another one-third rated them typical of line equipment, and the final third
rated them problematic. The survey results, along with feedback from follow-up contacts,
highlighted several issues. These include:
• Misoperation of Capacitor Fuses: Many utilities have unnecessary fuse operations -
operations when the capacitor bank was not in danger of being damaged. This can unbalance

11-1
Capacitor Bank Power Quality and Reliability Issues

circuit voltages and reduce the number of capacitors available for VAR support. Fusing
practices should be reviewed to reduce this problem.
• Controllers: Controllers were rated problematic by a significant number of utilities. Some
utilities had problems with switches and with the controllers themselves.
• Lightning and Faults: Controllers can fail from lightning and they are especially vulnerable
in high-lightning areas. Controllers are quite exposed to lightning and to power-supply
overvoltages during faults. Surge protection practices as well as power and grounding of
controllers should be reviewed.
• Human Element: Many controllers are set up incorrectly. Some controllers are hard to
program; and field crews often don’t have the skills (or proper attitudes) to set up capacitors
and their controls. At some utilities, crews manually switch off nearby capacitors and often
forget to turn them back on after finishing their work. To reduce these problems, crews must
be trained on the proper use and setup of capacitors as well as taught (with greater emphasis)
the importance of having capacitors available.

Harmonics

Harmonics are sinusoidal voltages or currents having frequencies that are integer multiples of the
frequency at which the supply system is designed to operate (termed the fundamental frequency,
which is usually 50 Hz or 60 Hz). Harmonic distortion exists due to nonlinear characteristics of
the power system and its connected loads.

Most harmonic distortion originates with nonlinear devices on the power system. Nonlinear
devices produce nonsinusoidal current waveforms when energized with a sinusoidal voltage.
Examples of these devices are adjustable-speed drives, switch-mode power supplies (including
computers and other office equipment), fluorescent lighting, battery chargers, transformers,
generators, and arc furnaces. Although power system components (such as transformers and
generators) contribute to some distortion, the bulk of the distortion in most power systems comes
from nonlinear, end-use devices.

These devices can usually be modeled as current sources that inject harmonic currents into the
power system. Voltage distortion results because these currents cause nonlinear voltage drops
across the system impedance. Because of the increasing application of power electronics in
end-use equipment, harmonic distortion has become a growing concern for many customers and
for overall power systems as well.

Harmonic distortion levels can be characterized by the complete harmonic spectrum with
magnitudes and phase angles of individual harmonic components. It is also common to use a
single quantity—the total harmonic distortion—as a measure of the magnitude of harmonic
distortion. For example, the total current harmonic distortion is:

11-2
Capacitor Bank Power Quality and Reliability Issues

∑I
2
h
h=2
I THD
= Eq. 11-1
I1

where Ih is the RMS current of the hth harmonic current, and I1 is the RMS value of the
fundamental current. A typical voltage waveform does not exceed 5% THD. However, the power
supply input current THD could exceed 100%.

Harmonic currents result from the normal operation of nonlinear devices on the power system.
Figure 11-1 illustrates the waveform and harmonic spectrum for a typical adjustable-speed drive
input current. Current distortion levels can be characterized by a total harmonic distortion, as
described above; but this can often be misleading. For instance, many adjustable-speed drives
will exhibit high total-harmonic-distortion values for the input current when they are operating at
very light loads. This is not a significant concern; because the magnitude of harmonic current is
low, even though its relative distortion is high.

Figure 11-1
Waveform and Harmonic Spectrum of Typical 6-Pulse ac Motor Drives

To account for the relative harmonic current levels in a consistent fashion, IEEE Std. 519-1992
defines another term, total demand distortion [5]. Total demand distortion is the same as the total
harmonic distortion, except that the distortion is expressed as a percent of some rated load
current rather than as a percent of the fundamental current magnitude. Recommended practice
for acceptable harmonic current and voltage distortion levels on distribution and transmission
circuits are provided in IEEE Std. 519.

11-3
Capacitor Bank Power Quality and Reliability Issues

The increased voltage distortion resulting from a relatively weak source, such as a generator, can
cause potential problems for customer loads. The main issues are as follows [8]:
• The voltage-sensing circuit in an uninterruptible power supply (UPS) that determines the
transfer from utility line to battery during an undervoltage condition may be susceptible to a
distorted voltage waveform, causing the UPS to go to battery power. Its batteries may be
depleted after only a short time during the operation of the DG.
• Increased voltage distortion will reduce the life of induction motors and other customer
loads.
• A building power-line carrier signal, clock-synchronizing signal, or other communication
signal within the customer’s premises may be susceptible to increased voltage distortion.
• Increased voltage distortion may damage equipment that uses a front-end filter with
capacitors, such as lighting ballasts.
• Increased voltage distortion may cause excessive harmonic current, damaging power-factor-
correction capacitors within the customer premises.

When harmonics cause problems, capacitors are often a contributing factor. Capacitors can cause
resonances that amplify harmonics. This often shows up first at the capacitor bank, where
harmonics cause fuse operations or even capacitor failure. Most problems with severe harmonics
are found in industrial facilities where capacitors resonate against the system impedance. While
utility resonances are rare, they do sometimes cause problems.

One capacitor on a distribution circuit will resonate against the inductance back to the system
source (including the line impedances and transformer impedances), as shown in Figure 11-2.

Figure 11-2
Harmonic Resonance

The resonance point between the capacitor and the system (which is the same frequency at which
the system will ring during a switching surge) is shown as:

XC MVAsc
n= = Eq. 11-2
XS MVAR

where

11-4
Capacitor Bank Power Quality and Reliability Issues

n = order of the harmonic,

XC = line-to-ground impedance of one phase of the capacitor bank at nominal


frequency,

XL = system impedance at nominal frequency,

MVASC = 3-phase short-circuit MVA at the point where the capacitor is applied, and

MVAR = 3-phase MVAR rating of the capacitor.

If a nonlinear load is injecting a harmonic frequency equal to the system’s resonance point, the
circuit can have overvoltages. Common danger frequencies are n=5, 7, 11, and 13. Larger
capacitors lower the resonance point to where it increases the likelihood of causing problems.
Multiple capacitors on a circuit create multiple resonant points that can require more
sophisticated analysis. The worst conditions are when the harmonic source is right at the
capacitor or downstream of the capacitor. Harmonics injected further upstream are less
amplified.

Solutions to Harmonics

Identifying harmonics is easy to do with an oscilloscope-type meter that measures voltage and
current. Tracing the source of the harmonics is more difficult, however, especially if a circuit has
capacitors. Before chasing after the harmonics, therefore, the utility engineer must make sure that
the problem is really caused by the harmonics and not by capacitor fuse operations, equipment
operations, and so on. If harmonics is the problem source, the utility-side options are harmonic
filters or distribution system changes.

Before resorting to filters, other utility-side options could be explored. Many harmonic problems
originate because of resonances with capacitors, thus problems can be greatly reduced by
disrupting the resonance. To disrupt the resonance, capacitors may be moved, their size changed,
or they may be switched off. Usually, it does not require significant changes to move a resonant
point enough to reduce the harmonic amplification. The adjustments are typically easy but may
require some trial and error.

While not a mainstream or regular application, tuned filter banks that can be used to filter
objectionable harmonics on distribution circuits are available through several manufacturers.
These are available in pole-mounted or pad-mounted configurations. Most of the applications are
for industrial use, but some are used on utility distribution circuits. Harmonic filter banks consist
of one or more Series L-C tuned filters, as shown in Figure 11-3. Filters are normally tuned to
just below the offending harmonic.

11-5
Capacitor Bank Power Quality and Reliability Issues

Figure 11-3
Tuned Harmonic Filter

Filters accomplish two objectives—to correct the power-factor and to shunt one or more
harmonic currents to ground. A series-tuned filter is constructed on each phase by placing a
choke in series with a shunt capacitor and then tuning the choke so that the inductive and
capacitive reactance are equal but opposite at the desired harmonic. Normally, a filter bank is
custom-engineered for a given problem. Harmonic filters, therefore, must be properly designed.
Some key issues are:
• Tuning: Tuning a filter slightly below the desired harmonic, for example at the 4.7th instead
of the 5th harmonic, helps to reduce capacitor voltage without significantly degrading filter
performance.
• Sizing: Care must be taken to dedicate enough kVAR to the filter. In most cases, the filter
kVAR should be approximately the amount needed for power-factor correction. Filters with
smaller kVAR will have sharp tuning curves and will be easily overloaded by stray
harmonics that are present in the network.
• Voltage Rating: Because a filter capacitor usually experiences 1.2 to 1.3 p.u. rms voltage,
plus significant harmonics, care must be taken that the capacitor voltage rating is adequate.
The fact that kVARs decrease by the square of voltage must also be taken into consideration.
• Interaction: Filters improperly applied to existing capacitor banks can cause problems. A
nontuned capacitor adjacent to a filter will resonate with the filter at some frequency. If that
frequency corresponds to a key frequency (like the 7th harmonic), harmonics may be worse
than without the filter. Proper application of harmonic filters with nearby capacitors requires
a careful engineering study to ensure that there are no improper interactions between filters
and capacitors. Separately designed filters also can interact with each other. This is a key
constraint for distribution applications, especially since over the years, adjacent capacitors
could have been installed on the same circuit without realizing that there are tuned filters on
the circuit.

Switching Surges

Because capacitor voltage cannot change instantaneously, energizing a capacitor bank results in
an immediate drop in system voltage toward zero. This is followed by a fast voltage recovery

11-6
Capacitor Bank Power Quality and Reliability Issues

(overshoot) and finally an oscillating transient voltage superimposed on the 60-Hz fundamental
waveform. The magnitude of the transient is based largely on the point on the sine wave at which
it is energized and the characteristics of the circuit on which the capacitor is located.

Figure 11-4
Example Capacitor Switching Transient

Transient frequencies due to utility capacitor switching usually fall in the range 300-1000 Hz.

1 MVAsc
f = = 60 × Eq. 11-3
2π Ls C MVAR

Typically, switching surges range in magnitude from 1.1 to1.6 pu.

Switching surges can be magnified inside customer facilities. These transients can cause
malfunctions in some types of sensitive equipment. Switching surges can also be magnified on
certain distribution feeders. The capacitor’s energizing transient can be magnified when the
series combination of a step-down transformer and a lower-voltage capacitor bank cause the
energizing transient frequency to be magnified at the lower-voltage capacitor. The conditions for
magnification are:
• C1 >> C2.
• Natural frequency of high-voltage and low-voltage circuit is close
• No significant resistive load in the low-voltage circuit

Transient overvoltages resulting from capacitor switching are usually not a significant concern to
the utility, since peak magnitudes are below the level at which utility surge protection (such as

11-7
Capacitor Bank Power Quality and Reliability Issues

arresters) begins to operate. Because of the relatively low frequency, these transients will pass
through step-down transformers to customer loads.

One power quality complaint related to distribution capacitor switching is the shutdown of any
adjustable-speed drive (ASD) or other process equipment. The front end of an adjustable-speed
drive will rectify the incoming ac to dc voltage. The rectifier peak-tracks the incoming ac so that
a switching surge will charge up the dc bus. Because the electronics on the dc bus are sensitive to
overvoltages, drives normally have sensitive dc bus overvoltage relays. Therefore, capacitor
switching surges can cause drives to trip out.

Figure 11-5
Scenario for Magnified Transients

Figure 11-6
Example of a Transient Magnified to Individual Customers

11-8
Capacitor Bank Power Quality and Reliability Issues

The best local solution is the addition of a series inductor (line reactor) connected to the input
terminals on the adjustable-speed drive.

Utility-side solutions to prevent the transient include zero-closing switches or preinsertion


resistors or reactors (these close the impedance element between the capacitor and the circuit and
then short out the impedance, which results in a softer close). For distribution feeder banks, the
zero-crossing switch controllers are the only real option for reducing transients. These controllers
time the closing of each switch contact to engage the capacitor at the instant where the system
voltage is zero, or very close to zero. An uncharged capacitor closing in at zero volts causes no
transient.

Adjustable Speed Drive (ADS) Tripping

Capacitor banks, both at the distribution level and transmission level, can impose voltage
transients on the line during switching. The ASD is sensitive to transient overvoltages on its
internal dc bus; thus exposure to overvoltages often results in the automatic shutdown of the
drive. The problem is further complicated by the fact that many automated capacitor banks
switch frequently, creating the potential to trip nearby ASD on a regular basis. In this manner,
customer equipment, especially the ASD, acts as a peak detector. Even though this section
focuses on ASD tripping, all devices that use a dc link bus are susceptible to transients.

Figure 11-7 shows a typical ac transient caused by capacitor switching; the typical ASD
topology; and, finally, the effect on the dc bus voltage. The dc bus voltage is shown for an ASD
with and without a reactor (or choke) on the input. The improvement with the use of a relatively
inexpensive and easily applied reactor is obvious. However, ASD manufacturers also express
concern regarding the cumulative effect of electrical component overstresses that may not
manifest themselves until an outright failure occurs.

Standard choke sizes are 1.5, 3, and 5% impedance on the kVA rating of the ASD. A parametric
simulation study of transmission-switched capacitors found that a 3% inductor greatly reduced
the number of drive trips, but a number of transients still can trip some drives (depending on the
size of capacitor being switched and many other variables) (Bellei T. A., O'Leary R. P. et al.).

11-9
Capacitor Bank Power Quality and Reliability Issues

Figure 11-7
Effect of Capacitor-Switching Transient on the Direct Current Bus of an Adjustable Speed
Drive

Solutions to Switching Transients

Tracking down malfunctions of adjustable speed drives or other electronic equipment can be
tricky. Drives and other sensitive equipment can trip for a variety of reasons. On the power
systems side, voltage sags or swells could trip a drive. The drive could also trip for load-side
reasons or because of control problems. Power quality monitors installed at the drive are the best

11-10
Capacitor Bank Power Quality and Reliability Issues

way to verify the power-quality issues. If drive trips occur regularly and those times tend to
coincide with typical closing times of capacitors, 8 a.m. for example, that recurring event is
likely a sign of capacitor switching problems. Once a problem is identified as a transient, finding
the offending capacitor is usually straightforward.

Two utility-side means of limiting the transient have been successfully employed:
• Synchronous Closing: Synchronous closing refers to closing the contacts at a point on the
sine wave of extremely little voltage, preferably zero. This option can only be applied when
each phase can be switched individually.
• Pre-Insertion Impedance: The transient may also be reduced by closing the system through
an impedance and limiting the current until the main contacts are closed. The capacitor is
thus pre-charged. This is termed a pre-insertion resistor or pre-insertion inductor. This device
has a limited capacity and is incorporated into the mechanism that drives the main contacts. It
is, in effect, shorted out by main contacts during normal operation.

For line capacitors, synchronous closing is the only option. Station capacitor banks have options
for either solution. For line capacitors, a 2002 EPRI survey found that only 20% of utilities used
synchronous switches; and of those, most were either on a trial basis or were for targeted areas
(EPRI 1001691 2002).

The newest synchronous switching schemes employ active feedback to compensate for physical
changes; and they maintain precise position control by the use of voice-coil actuators. The
operation in the synchronous closing mode requires that voltage from the source be available.
The actual transient produced upon the synchronous closing of a capacitor bank at a typical
installation is shown in Figure 11-8.

The best local solution is the addition of a series inductor (line reactor) connected to the input
terminals of the ASD. A second inexpensive, utility-side option is to use clock controls and
switch the capacitor on during off-process times, such as before 5 a.m. (For most capacitor
switches, the on switch is most likely to cause a transient.) A third option is to disable switching
on a capacitor unit. This may increase distribution line losses; but as long as circuit voltages are
still okay, service to other customers will not be degraded. One could also investigate whether
the problems are caused by resonances that magnify transients. If so, disconnecting one of the
capacitors or changing its size will reduce the severity of the transient, possibly enough to stop
malfunction of the drives.

11-11
Capacitor Bank Power Quality and Reliability Issues

Figure 11-8
Transient Caused by Synchronous Switching of a Capacitor

Telephone Interference

Telephone interference has been a harmonics-related concern for many decades, but the gradual
phasing out of open-wire telephone circuits has reduced the number of interference problems.
When interference problems do occur, however, they can be difficult to trace and to fix. While
the frequency response of the telephone circuit and human ear is largely immune to 60-Hz
interference, higher harmonics fall into the low-audio range.

When harmonic currents on power lines inductively couple into nearby phone lines, they can
cause significant interference. Typically, the problem harmonics are either characteristic
six-pulse harmonics caused by large converters or 9th and higher multiples of three (i.e., zero
sequence) caused by transformer saturation. All things being equal, zero-sequence harmonics are
more problematic than positive- and negative-sequence harmonics. This is because a-b-c zero-
sequence fields are additive; therefore, they do not decay as rapidly with distance.

The telephone influence factor (TIF) curve shown in Figure 11-9 gives the relative interference
weighting that applies to inductively-coupled harmonic currents flowing in power lines. The
frequencies from 1,000 to 3,800 Hz are the most critical; that is, they are the frequencies most
likely to cause interference.

11-12
Capacitor Bank Power Quality and Reliability Issues

Figure 11-9
Telephone Influence Factor (TIF) Curve

Telephone interference problems are usually solved by the telephone company in cooperation
with the electric utility involved. Solutions are often trial-and-error. The normal solutions are:
• Change Size: For problems involving a resonance, increasing or decreasing the size of a
bank can shift the resonance point enough to ease interference problems. The easiest solution
is to disconnect the bank contributing to the problem. This is also a good first step to
quantifying the role of the capacitor bank in the interference.
• Move Bank: Moving a capacitor can change a resonance point enough to stop interference
problems. On some circuits, one can also move the capacitor away from the telephone
circuits having problems.
• Unground the Bank: A floating-wye connection has no connection to ground, so the
connection blocks zero-sequence harmonics. Two-bushing capacitor units are necessary for
floating the wye point (unless the utility floats the capacitor tanks and deals with the safety
issues that accompany that practice).
• Add a Neutral Filter: While not a common solution, a tuned reactor in the ground path of a
wye-connected capacitor bank is invisible to positive and negative sequences. However, the
reactor can change the zero-sequence resonant frequency of a distribution feeder, often
eliminating the resonance problem.

Voltage Flicker

Light flicker is due to rapidly changing loads and generation that cause fluctuation in the
secondary voltage. Even a small change in voltage can cause noticeable lamp flicker. The degree
of irritation depends on both the frequency and the magnitude of the fluctuations, as indicated in
the classic flicker curves shown in Figure 11-10. These curves depict the threshold of human
annoyance. Most people will notice 1% and smaller voltage changes that occur in the range of 1
to 33 changes per second (60 to 2000 changes per minute). The greatest sensitivity occurs around
900 changes per minute (or 15 changes per second), which is equivalent to a frequency of 7 to
8 Hz. Both incandescent lamps and fluorescent lamps are susceptible to voltage fluctuations. The

11-13
Capacitor Bank Power Quality and Reliability Issues

amount of flicker also depends on the type of lamp. Voltage fluctuations can also cause
televisions and computer monitors to waver.

Figure 11-10
GE Flicker Curve

Annoying lamp flicker can occur when rapid changes in load current cause the power system
voltage to fluctuate. Both incandescent and fluorescent lamps can flicker during voltage
fluctuations. The standards for measuring and limiting lamp flicker are based on the 60-W
incandescent lamp.

Flicker is a difficult problem to quantify and to solve. Flicker becomes a problem when some
deviation in voltage supplying lighting circuits combines with the presence of a person viewing a
change in light intensity resulting from the voltage deviation. Because the human factor
significantly complicates the issue, flicker has historically been deemed a problem of perception.
The voltage deviations involved are often much less than the thresholds of susceptibility, even
for the most sensitive electronic equipment. Only in rare cases does flicker actually cause system
operating problems. To observers in homes and offices, however, voltage deviations on the order
of 1 to 2% could produce extremely annoying changes in light output, especially if the range of
repetitive deviation is 5 to 15 Hz.

Due to the clear coupling of voltage fluctuations and lamp output changes, the term flicker often
means different things to different people, with the interpretations primarily governed by point of
view. In each case, the deviation, which may or may not be strictly periodic, is usually expressed
as a change relative to the steady-state level. For voltage variations, the change is usually
expressed as ∆V/V. There is a similar expression for light intensity variations as well.

In contrast to the behavior of incandescent, filament-type lamps, fluorescent gaseous discharge


lamps have very little thermal inertia—and they respond even faster. The time constant for a
60-W, 120-V, incandescent lamp is about 28 ms and the time constant for a 230-V lamp of the

11-14
Capacitor Bank Power Quality and Reliability Issues

same wattage is about 19 ms. A typical fluorescent lamp, on the other hand, has a time constant
of less than 5 ms. Also, fluorescent lamps are usually not amplifying the voltage changes in their
corresponding change in light output. This is why incandescent lamps are historically the most
sensitive and are used to set standards for allowable voltage fluctuations.

Flicker prediction and measurement can be challenging. The IEEE flicker curve, which was first
developed by General Electric in the 1920s, provides a good calibration of the threshold of where
voltage fluctuations at a specific frequency will cause visible lamp flicker. However, there is no
IEEE standard on how to measure the flicker level in a complex varying voltage. Because the
IEC does have an effective measurement standard, the IEEE working group on flicker recently
voted to adopt the IEC standards. The number of this standard is IEC 61000-4-15.

11-15
12
ECONOMICS

Utilizing automated, intelligent capacitor bank switching controls provides several channels of
return on investment (ROI), which generally yields a very fast payback. In fact, a utility can
undertake few capital projects that will provide a faster return. Automated capacitor control
generates three main areas of cost saving. These include:
• Energy Savings: Energy savings, also referred to as loss reduction, involves reducing line
and transformer losses by using intelligent capacitor control to effectively reduce the amount
of reactive current flowing in the line. Since energy is wasted in heating conductors, it cannot
be delivered to the customer nor can its use be billed (thus it generates no revenue). Line and
transformer losses also contribute to fatigue on line conductors and apparatuses through
heating.
• Capacity Savings: Also referred to as demand reduction. Capacity savings (also referred to as
demand reduction) involves improving the line power factor through proper application of
capacitors. This reduces the total line current thus reduces kVA demand. The benefits
provided by released capacity are twofold. First, releasing line capacity allows more billable
energy to be transferred to the customers, increasing the revenue that the line can generate.
Second, releasing line capacity enables the deferral of equipment upgrades. Improving the
power factor releases transmission and generation capacity as well as distribution capacity.
• Operation and Maintenance Savings: Required labor hours can be greatly decreased when
upgrading to intelligent, centralized capacitor control via SCADA. SCADA control greatly
reduces labor costs by enabling the centralization of switching control and monitoring of all
capacitor banks, thus dramatically reducing time required in traveling to location and
adjusting capacitor-control banks onsite. Additional cost savings come from the ability to
remotely monitor capacitor bank status to determine when capacitors fail. By eliminating the
need for technicians to travel to capacitor installations to inspect bank functioning every year,
a utility can save considerable man hours. The ability to quickly identify and fix failed
capacitors also means that fewer capacitors will need to be installed in the system; because a
very high percentage will be operational all the time. This further means that over time, some
capacitor banks can be taken out of service and used for future installations, providing
savings in capital costs.

Capital costs for capacitor control systems can vary greatly, depending on the level of
sophistication being employed and what, if any, existing utility infrastructure can be utilized for
the system. However, the level of existing hardware also plays a role in determining the design
of the capacitor control system. For example, a utility with an extensive 900-MHz radio system
in place will likely utilize that system for their capacitor control communications. A utility with
no communication system is place, however, may opt for a commercially provided

12-1
Economics

communication system, such as a cellular-control channel, rather than building its own
communication network. Even utilities that have a communication network in place may opt for
commercially provided communications, since it requires no infrastructure maintenance from the
utility.

Since capital costs depend on the type of system being installed and the utility’s existing
infrastructure, it is difficult to provide economic examples that cover the range of variability.
Furthermore, there are few technical documents within the industry that outline the costs
associated with individual, real-world capacitor control systems.

Energy Savings

One manner of calculating the energy loss savings from capacitor automation is to first estimate
the yearly kWh that each new automated capacitor would save. The saved kWh multiplied by the
cost per kWh determines the yearly savings as shown below (Marx 2003):
• $saved/year = (kWsave)*(hours/year)*($/kWh)
• Where, kWsave = kW saved when the capacitor is switched on
• hours/year = the number of hours per year the capacitor operates each year
• $/kWh = value of avoided energy

The yearly loss reduction used in the above equation can by calculated as follows:

W = (3*Rline)*[2*IIND*ICAP- ICAP2 ] Eq. 12-1

Where

Rline = single-phase line resistance in ohms

IIND = single-phase load inductive current in amps

ICAP = single-phase capacitor current in amps

W = 3-phase line real power loss reduction in watts

Capacity Savings

Capacity savings result from reduction in line current when the capacitor serves the reactive
portion of the load current. Therefore, the first step in determining the capacity savings is to
determine the reduction in line current due to the capacitor. The line current with the capacitor
bank switched onto the line can be determined from the following equation:

ILine = •(Iresistive2 +( ICAP- ICAP2)) Eq. 12-2

12-2
Economics

Where

Iresistive = single-phase load resistive current in amps

IIND = single-phase load inductive current in amps

ICAP = single-phase capacitor current in amps

ILine = single-phase line current in amps

The reduction in line demand can be found by comparing the pre-capacitor demand to the post-
capacitor demand. Line demand is calculated as follows:

Demand = 3*ILine*(kV/•3) Eq. 12-3

Where

ILine = single-phase line current in amps: For pre-capacitor demand use pre- capacitor
line current. For post-capacitor demand use post-capacitor line current.

kV = the line-to-line voltage

Demand = the line demand in kVA

A 30% reduction in peak line demand is not uncommon from properly applied switched
capacitor banks. Marx suggests a typical demand reduction value to be $80/year/kVA, based on a
5-year write-off cycle for capital equipment. The yearly demand savings from the capacitor
installation would be, then, the demand savings multiplied by the dollar value, per year, per
kVA. Demand reduction is affected by the magnitude of the resistive current; therefore, it is
more difficult to reduce the kVA demand when the power factor is high. It should also be noted
that capacitors that cause a line to go leading, either intentionally or unintentionally, increase the
line kVA demand, thereby reducing optimum demand savings.

When a capacitor switches onto the line, it causes the line voltage to increase, which creates a
corresponding increase in line demand. This increase in line demand must also be factored into
the final evaluation of cost savings from demand reduction. Equation 12-3 does not account for
the slight increase in demand from increased line voltage, so it must be calculated separately and
subtracted from the overall reduction. In a typical model, the peak demand for a diversified load
will increase by 1.6 times for each 1% increase of the voltage, and the long-term average demand
will increase 1/3 to 1% for each 1% voltage increase. Therefore, the demand increase resulting
from the capacitor-induced rise in voltage would be calculated as follows:

Demand Increase (%) = (% Voltage Increase)*(% Long-Term Load Change) Eq. 12-4

For example, if the voltage increases by 1.5% and the long-term load increases at 0.5% per 1%
voltage increase, then the demand increase would be:

12-3
Economics

Demand Increase = (1.5)*(0.5) = 0.75% Eq. 12-5

The demand increase due to higher line voltage must be accounted for when calculating the total
demand reduction. Therefore, the total demand savings are found by subtracting the demand
increase from the yearly demand savings Equation 12-3.

Operation and Maintenance Savings

In addition to demand and energy savings, automated capacitor control systems also provide
payback through reduced operation and maintenance costs. A major component of these savings
is achieved through elimination of routine capacitor patrols. Automated capacitors do not require
manual switching by a field technician on a periodic basis. Centralized control also provides
positive feedback of capacitor operation which allows failing capacitors to be identified from the
control center rather than via field inspection. The potential for man-hour reduction via capacitor
automation is huge. One utility reported that before enacting a capacitor automation system,
technicians were driving 24,000 (38,600 km) or more per year to adjust and maintain timed
capacitor banks. This workload was reduced by approximately 3 man-months in the first year
that capacitor automation was implemented (Goodrich 2004).

Estimated Cost Breakdown

Kansas City Power & Light (KCPL) has implemented an extensive capacitor automation project
than began in the early 1990s. As part of their capacitor automation project, KCPL developed an
economic value added model to help evaluate the costs and benefits associated with upgrading
capacitor-control technologies. As part of their model development, KCPL made several
assumptions, including, 1) they would retrofit 600 capacitor banks with programmable controls,
2) they would improve average power factor on distribution circuits, 3) they would defer the
purchase and installation of 110 MVAR of capacitors in the first five years of the project, and 4)
they could eliminate capacitor patrols on automated banks (Goeckeler 1999). Using these
assumptions, the KCPL engineers estimated the costs and benefits of applying capacitor
automation, as shown in Table 12-1 and Table 12-2.

12-4
Economics

Table 12-1
Estimated Benefits from Instituting Automated Capacitor Control on the Kansas City
Power & Light Distribution Systems

Estimated Benefits of Capacitor Automation

Benefit Percent of Total Benefit


Avoided Capacity Cost – Generation 28.9%
Marginal Energy Savings 24.5%
Avoided Capacity Cost – Substation and Distribution 13.3%
Capacitor Banks Avoided Costs 11.5%
Avoided Capacity Cost – Transmission 9.5%

Reduced Operations and Maintenance 8.0%

Controls Salvage Value 4.3%

Source: (Goeckeler 1999)

Table 12-2
Estimated Cost of Instituting Automated Capacitor Control on the Kansas City Power &
Light Distribution Systems

Estimated Cost of Capacitor Automation

Cost Percent of Total Cost


Equipment and Installation 58.0%
Equipment Lease Charges 26.3%
Engineering 8.1%
Software Purchase & Development 7.6%

Source: (Goeckeler 1999)

According to KCPL’s estimates in Table 12-1, the greatest percentage of savings would come
from avoided generation capacity. This savings was followed very closely by the marginal
energy savings associated with moving to the automated capacitor control system.

As Table 12-2 indicates, the greatest portion of the costs associated with implementing capacitor
automation tends to be the purchase and installation of the necessary equipment. Capacitor
controller prices generally range from $400 to $800 per controller. Special communication
requirements or other non-standard features can push this cost considerably higher, particularly
for very sophisticated control systems. Switched capacitor banks cost approximately $8/kVAR
installed; however, that figure doesn’t include radio or SCADA infrastructure that may be
needed to communicate effectively.

12-5
13
REFERENCES

1. ANSI C57.12.28-1998 Pad-Mounted Equipment Enclosure Integrity.


2. ANSI/IEEE Std. 18-1992 IEEE Standard for Shunt Power Capacitors.
3. Bedell, P. (2001). Wireless Crash Course. New York, NY, McGraw-Hill.
4. Begovic, M., D. Novosel, et al. (2000). Impact of Distribution Efficiency on Generation and
Voltage Stability. 33rd Annual Hawaii Int. Conference Systems Sciences, Maui, HI.
5. Bellei T. A., O'Leary R. P., et al. "Evaluating Capacitor-Switching Devices for Preventing
Nuisance Tripping of Adjustable-Speed Drives Due to Voltage Magnification." IEEE
Transactions on Power Delivery 11(3): 1373-8.
6. Brady, P., C. Dai, et al. (2003). Need to Revise Switched Capacitor Controls on Feeders with
Distributed Generation. IEEE Transmission and Distribution Conference Proceedings.
7. Brown, R. A. (1979). Capacitor Fusing. IEEE/PES Transmission and Distribution
Conference and Expo.
8. Burrage, L. M. (1981). High Frequency Characteristics of Capacitors and Fuses--Applied in
High-Voltage Shunt Banks. IEEE/PES Transmission and Distribution Conference.
9. CEA 288 D 747 (1998). Application Guide for Distribution Fusing, Canadian Electrical
Association.
10. Clark, G. L. (2001). Development of the Switched Capacitor Bank Controller for
Independent Phase Switching on the Electric Distribution System. Distributech 2001, San
Diego, CA.
11. Clarke, G., D. Reynders, et al. (2004). Practical Modern SCADA Protocols: DNP3, IEC
60870.5 and Related Systems. London, Elsevier.
12. de Vos, A., S. E. Widergren, et al. (2001). "XML for CIM Model Exchange." Power Industry
Computer Applications.
13. DNP Users Group. (2005). "Distributed Network Protocol Users Group." from
http://www.dnp.org.
14. EPRI 1001691 (2002). Improved Reliability of Switched Capacitor Banks and Capacitor
Technology. Palo Alto, CA, Electric Power Research Institute, Palo Alto, California.
15. EPRI 1001801 (2002). EPRI Common Information Model (CIM) Status and Guidelines for
Transmission and Distribution Applications. Palo Alto, Ca, EPRI.
16. EPRI 1002154 (2004). Improved Reliability of Switched Capacitor Banks and Capacitor
Technology - Fusing Recommendations and Using Distribution Capacitors for
Transmission Support. Palo Alto, Ca, EPRI.
17. EPRI 1008573 (2005). Grounding and Lightning Protection of Capacitor Banks. Palo Alto,
CA, EPRI.
18. Ezell, B. C. (1998). Risks of Cyber Attacks to Supervisory Control and Data Acquisiton for
Water Supply, University of Virginia. Master of Science.
19. Ezell, B. C. (2005). Quantifying Vulnerability to Critical Infrastructure, Old Dominion
University. Doctor of Philosophy.

13-1
References

20. Fisher Pierce (2000). POWERFLEX 4400 and 4500 Series Instruction Manual.
21. Girotti, T. B., N. B. Tweed, et al. (February 1990). "Real-Time VAR Control By SCADA."
IEEE Transactions on Power Systems 5(1).
22. Goeckeler, C. (1997). Beyond SCADA: Capacitor Automation at KCPL. Transmission &
Distribution World.
23. Goeckeler, C. (1999). "Progressive Capacitor Automation Yields Economic and Practical
Benefits at KCPL." Utility Automation.
24. Goodrich, D. (2004). Reaping the Rewards of Automated Capacitor Control. Transmission
and Distribution World.
25. Grainger, J. J. and S. H. Lee (1981). "Optimum Size and Location of Shunt Capacitors for
Reduction of Losses on Distribution Feeders." IEEE Transactions on Power Apparatus
and Systems PAS-100(3): 1005-18.
26. Harris, S. (2001). Evaluating Wireless Technologies for Distribution Automation. Utility
Automation.
27. IEEE-SA TR 1550-1999 "Utility Communications Architecture (UCA) Version 2.0." 1.
28. IEEE C2-1997 National Electrical Safety Code.
29. IEEE Std. 18-2002 IEEE Standard for Shunt Power Capacitors.
30. IEEE Std. 1036-1992 IEEE guide for application of shunt power capacitors.
31. IEEE Std. C37.48-1997 IEEE guide for the application, operation, and maintenance of high-
voltage fuses, distribution enclosed single-pole air switches, fuse disconnecting switches,
and accessories.
32. Kirshner, D. and P. Giorsetto (1984). "Statistical Tests of Energy Savings Due to Voltage
Reduction." IEEE Transactions on Power Apparatus and Systems PAS-103(6): 1205-10.
33. Marx, T. (2003). The Why and How of Power Capacitor Switching, Fisher Pierce.
34. McCarthy, C. A. and J. Josken "Applying Capacitors to Maximize Benefits of Conservation
Voltage Reduction."
35. Milosevic, B. and M. Begovic (2004). "Capacitor Placement for Conservative Voltage
Reduction on Distribution Feeders." IEEE Transactions on Power Delivery 19(3).
36. Neagle, N. M. and D. R. Samson (1956). "Loss Reduction from Capacitors Installed on
Primary Feeders." AIEE Transactions 75: 950-9.
37. Newcomb, G. R. (1980). Film/Foil Power Capacitor. IEEE International Symposium on
Electrical Insulation, Boston, MA.
38. ORNL/TM-2001/215 (2002). Evaluation of Distribution Analysis Software for DER
Applications. D. T. Rizy and R. H. Staunton, Oak Ridge National Laboratory.
39. Pollet, J. (2002). Developing a Solid SCADA Security Strategy. Sensors for Industry
Conference, Houston, TX.
40. Priess, R. F. and V. J. Warnock (1978). "Impact of Voltage Reduction on Energy and
Demand." IEEE Transactions on Power Apparatus and Systems PAS - 97(5): 1665-71.
41. RCCS. (2005). "Radio Control Central Stations." from
http://www.rccscontrols.com/Overview/consult.htm.
42. Sanders, T. (2002). "Threats to 900 MHz Broadband Wireless." Broadband Wireless Online,
from Shorecliff Communications LLC - http://www.shorecliffcommunications.com.
43. Short, T. A. (2004a). Distribution Capacitor Reliability and Changes in Fusing Practices.
T&D World Expo, Indianapolis, IN.
44. Short, T. A. (2004b). Electric Power Distribution Handbook. New York, CRC Press.

13-2
References

45. Wang, X. and N. N. Schulz (2003). "CIM Extensions to Electrical Distribution and CIM
XML for IEEE Radial Test Feeders." Power Delivery, IEEE Transactions on 18(3):
1021-1028.
46. Weisman, C. J. (1999). The Essential Guide to RF and Wireless. Upper Saddle River, NJ,
Prentice Hall PTR.
47. Westinghouse Electric Corporation (1959). Electric Utility Engineering Reference Book,
Volume 3: Distribution Systems.
48. Xiaofeng, W. and N. N. Schulz (2000). "The Extension of EPRI's CIM to Distribution Power
Systems." Proceedings of 32nd Annual North American Power Symposium.
49. Young, H. E. (1999). Wireless Basics. Overland Park, KS, Intertec Publishing Corporation.

13-3
Export Control Restrictions The Electric Power Research Institute (EPRI)
Access to and use of EPRI Intellectual Property is granted with The Electric Power Research Institute (EPRI), with major locations in
the specific understanding and requirement that responsibility Palo Alto, California, and Charlotte, North Carolina, was established
for ensuring full compliance with all applicable U.S. and in 1973 as an independent, nonprofit center for public interest
foreign export laws and regulations is being undertaken by energy and environmental research. EPRI brings together members,
you and your company. This includes an obligation to ensure participants, the Institute’s scientists and engineers, and other leading
that any individual receiving access hereunder who is not a experts to work collaboratively on solutions to the challenges of electric
U.S. citizen or permanent U.S. resident is permitted access power. These solutions span nearly every area of electricity generation,
under applicable U.S. and foreign export laws and delivery, and use, including health, safety, and environment. EPRI’s
regulations. In the event you are uncertain whether you or members represent over 90% of the electricity generated in the
your company may lawfully obtain access to this EPRI United States. International participation represents nearly 15% of
Intellectual Property, you acknowledge that it is your EPRI’s total research, development, and demonstration program.
obligation to consult with your company’s legal counsel to
determine whether this access is lawful. Although EPRI may Together...Shaping the Future of Electricity
make available on a case-by-case basis an informal
assessment of the applicable U.S. export classification for
specific EPRI Intellectual Property, you and your company
acknowledge that this assessment is solely for informational
purposes and not for reliance purposes. You and your
company acknowledge that it is still the obligation of you and
your company to make your own assessment of the applicable
U.S. export classification and ensure compliance accordingly.
You and your company understand and acknowledge your
obligations to make a prompt report to EPRI and the
appropriate authorities regarding any access to or use of EPRI
Intellectual Property hereunder that may be in violation of
applicable U.S. or foreign export laws or regulations.

Program:
© 2005 Electric Power Research Institute (EPRI), Inc. All rights reserved.
Power Delivery
Electric Power Research Institute and EPRI are registered service marks of
the Electric Power Research Institute, Inc.
1010655
Printed on recycled paper in the United States of America

ELECTRIC POWER RESEARCH INSTITUTE


3420 Hillview Avenue, Palo Alto, California 94304-1395 • PO Box 10412, Palo Alto, California 94303-0813 USA
800.313.3774 • 650.855.2121 • askepri@epri.com • www.epri.com

Вам также может понравиться