Académique Documents
Professionnel Documents
Culture Documents
Future of
the Electric
Grid
AN INTERDISCIPLINARY MIT STUDY
Other Reports in This Series
ISBN 978-0-9828008-6-7
JERROLD M. GROCHOW
Research Affiliate, MIT Energy Initiative
Former Vice President for Information Services
and Technology, MIT
For well over a century, electricity has made predecessors focused on implications of
vital contributions to the growth of the U.S. national policies limiting carbon emissions,
economy and the quality of American life. The while we do not make assumptions regarding
U.S. electric grid is a remarkable achievement, future carbon policy initiatives. Instead, we
linking electric generation units reliably and mainly consider the implications of a set of
efficiently to millions of residential, commer- ongoing trends and existing policies.
cial, and industrial users of electricity through
more than six million miles of lines and We anticipate this report will be of value to a
associated equipment that are designed and wide range of decision makers in industry and
managed by more than 3,000 organizations, government as they guide the grids continuing
many of which are in turn regulated by both evolution. We have attempted to provide
federal and state agencies. While this remark- thorough discussions of key topics to serve as
able system of systems will continue to serve us references, to support our findings and recom-
well, it will face serious challenges in the next mendations, and to meet the needs of what we
two decades that will demand the intelligent expect will be a diverse audience in terms of
use of new technologies and the adoption of interest and expertise. Also, for those less
more appropriate regulatory policies. familiar with the industry, we include appen-
dices on the grids history and technology.
This report aims to provide a comprehensive, Chapter 1 provides an overview of the status
objective portrait of the U.S. electric grid and of the grid, the challenges and opportunities it
the challenges and opportunities it is likely to will face, and our major recommendations. To
face over the next two decades. It also highlights facilitate selective reading, detailed descriptions
a number of areas in which policy changes, of the contents of each section in Chapters 29
focused research and demonstration, and the are provided in each chapters introduction,
collection and sharing of important data can and recommendations are collected and briefly
facilitate meeting the challenges and seizing the discussed in each chapters final section.
opportunities that the grid will face.
The MIT Future of the Electric Grid Study
This study is the sixth in the MIT Energy gratefully acknowledges the sponsors of this
Initiatives Future of series. Its predecessors study: ABB Group, American Electric Power,
have shed light on a range of complex and Bechtel Foundation, Larry Birenbaum, Cisco
important issues involving energy and the Systems, Exelon Corporation, General
environment. While the previous studies have Electric Company, Iberdrola S.A., Microsoft
focused on particular technologies and energy Corporation, National Institute of Standards
supply, our study of the grid necessarily and Technology, and Southern California
considers many technologies and multiple Edison. In addition to providing financial
overlapping physical and regulatory systems. support, many of our corporate and government
Because of this breadth, our efforts were sponsors gave us access to staff members who
focused on integrating and evaluating existing provided frequent and detailed information
knowledge rather than performing original about technical and policy issues. We are very
research and analysis. In addition, this studys grateful for this cooperation.
Appendices
235 Appendix A: A Brief History of the U.S. Grid
235 A.1 Beyond Municipal Boundaries
236 A.2 A Federal Role Emerges
238 A.3 Interconnection and Competition
261 Glossary
The U.S. electric grid is a vast physical and expansion of the transmission system, often via
human network connecting thousands of unusually long transmission lines. Current
electricity generators to millions of consumers planning processes, cost-allocation procedures,
a linked system of public and private enterprises and siting regimes will need to be changed to
operating within a web of government institu- facilitate this expansion. In addition, increased
tions: federal, regional, state, and municipal. penetration of renewable distributed genera-
The grid will face a number of serious challenges tion will pose challenges for the design and
over the next two decades, while new technolo- operation of distribution systems, and may
gies also present valuable opportunities for raise costs for many consumers.
meeting these challenges. A failure to realize
these opportunities or meet these challenges Increased penetration of electric vehicles and
could result in degraded reliability, significantly other ongoing changes in electricity demand
increased costs, and a failure to achieve several will, if measures are not taken, increase the
public policy goals. ratio of peak to average demand and thus
further reduce capacity utilization and raise
This report, the fifth in the MIT Energy rates. Changes in retail pricing policies, enabled
Initiatives Future of series, aims to provide by new metering technology, could help to
a comprehensive, objective portrait of the U.S. mitigate this problem. Increased penetration
electric grid and the identification and analysis of distributed generation will pose challenges
of areas in which intelligent policy changes, for the design and operation of distribution
focused research, and data development and systems. New regulatory approaches may be
sharing can contribute to meeting the chal- required to encourage the adoption of innova-
lenges the grid is facing. It reflects a focus on tive network technologies.
integrating and evaluating existing knowledge
rather than performing original research. We Opportunities for improving the functioning
hope it will be of value to decision makers in and reliability of the grid arise from techno-
industry and in all levels of government as they logical developments in sensing, communica-
guide the grids necessary evolution. tions, control, and power electronics. These
technologies can enhance efficiency and
One of the most important emerging challenges reliability, increase capacity utilization, enable
facing the grid is the need to incorporate more more rapid response to remediate contingencies,
renewable generation in response to policy and increase flexibility in controlling power flows
initiatives at both state and federal levels. Much on transmission lines. If properly deployed and
of this capacity will rely on either solar or wind accompanied by appropriate policies, they can
power and will accordingly produce output that deal effectively with some of the challenges
is variable over time and imperfectly predictable, described above. They can facilitate the integra-
making it harder for system operators to match tion of large volumes of renewable and distrib-
generation and load at every instant. Utilizing uted generation, provide greater visibility of the
the best resource locations will require many instantaneous state of the grid, and make possible
renewable generators to be located far from the engagement of demand as a resource.
existing load centers and will thus necessitate
Hailed as the supreme engineering achieve- regulatory and governance structures that
ment of the 20th century by the National shape the systems evolution. The term grid
Academy of Engineering,1 the U.S. electric as used in this report refers not only to the
power grid serves more than 143 million physical transmission and distribution systems
residential, commercial, and industrial that link generators to ultimate loads but also
customers 2 through more than 6 million miles the associated operational, regulatory, and
of transmission and distribution lines owned governance structures. Appendix A briefly
by more than 3,000 highly diverse investor- summarizes the history of the U.S. grid.
owned, government-owned, and cooperative
enterprises.3 In 2009, electric generation This study considers the evolution of the
consumed 41% of the nations primary U.S. electric grid over the next two decades
energyup from 14% in 1949underscoring a period long enough to permit significant
the great and growing national importance of change but short enough to make it unlikely
the grids efficiency and reliability.4 that unforeseen
technologies will This study considers the evolution
The electric power system is composed of have significant of the U.S. electric grid over the
four interacting physical elements: energy impacts on the next two decades.
generation, high-voltage transmission, lower- system.i Even though
voltage distribution, and energy consumption, this is likely to be a period of slow growth in
or load. Two less tangible elements are also the U.S. demand for electricity by historical
important: the operational systems that protect standards, public policies and a variety of
and control the physical elements, and the technological and economic changes will alter
i
An excellent brief overview of many of the issues considered in this report is provided by P. L. Joskow,
Creating a Smarter U.S. Electricity Grid, Journal of Economic Perspectives, forthcoming.
ii
As we discuss in later chapters, this last requirement has implications for many policies, including the role
and design of competitive electricity markets.5
Source: ISO/RTO Council, http://www.isorto.org. Copyright ISO/RTO Council, all rights reserved.
non-discriminatory access to transmission which are responsible for balancing the supply
systems (see Figure 1.2). Nonetheless, this and demand for power in real time in specified
policy has major implications for the transmis- areas. A glance at this map makes plain the
sion grid. Where it has been implemented, it strong influence of history. New York, New
has led to organized markets with important England, and Texas are each tightly integrated
common elements. and have one balancing authority each, while
Arkansas and Arizona each have eight and
Figure 1.2 shows the geographic scope of Florida has eleven. These differences plainly
organized wholesale electricity markets, which do not reflect differences in levels of supply or
are operated by Independent System Operators demand or in system complexity.
(ISOs) or Regional Transmission Organizations
(RTOs) that do not own generators or serve Physically, the U.S. electric grid currently
retail customers. These markets now cover consists of approximately 170,000 miles of
two-thirds of the U.S. population and meet high-voltage (above 200 kilovolts or kV) electric
about two-thirds of U.S. demand. Box 1.1 transmission lines and associated equipment,7
provides an overview of the operation of these and almost 6 million miles of lower-voltage
markets. In the Southeast, the traditional distribution lines.8 These include approximately
vertically integrated utility model is dominant, 2,400 miles of 765 kV alternating current (ac)
while in the West, particularly the Pacific lines, the highest voltage lines in operation in
Northwest, federal, municipally owned, and the U.S., and more than 3,000 miles of 500 kV
cooperative enterprises play an important role dc lines.9 Several hundred entities currently
in the industry. own parts of the transmission or bulk power
system.10 Investor-owned utilities own about
Within these broad areas are 107 so-called 66% of the system, and federal enterprises own
balancing authorities, shown in Figure 1.3, 14%. The rest is divided among other publicly
.OTE &2## &LORIDA 2ELIABILITY #OORDINATING #OUNCIL -2/ -IDWEST 2ELIABILITY /RGANIZATION .0## .ORTHEAST 0OWER #OORDINATING #OUNCIL
2&# 2ELIABILITY&IRST #ORPORATION 3%2# 3%2# 2ELIABILITY #ORPORATION 300 3OUTHWEST 0OWER 0OOL 42% 4EXAS 2EGIONAL %NTITY
WECC = Western Electricity Coordinating Council.
At the distribution level, about 3,200 organiza- particularly the Pacific Northwest, municipally
tions provide electricity to retail customers.12 owned and cooperative utilities benefit
Nearly 2,200 are publicly ownedsix by the substantially from preferred access to low-cost
federal government and the rest by states and power from federal projects. Only 242 distribu-
municipalitiesbut they account for only 16% tion entities are investor owned, but they
account for 66% of electricity sales. Their retail
At the distribution level, about 3,200 organizations rates are regulated by state public utility
commissions (PUCs). Finally, about 7.5% of
provide electricity to retail customers. retail sales are accounted for by retail power
marketers that do not provide distribution
of electricity sold. Another 818 are coopera- services.
tives, which in aggregate account for 10.5%
of kWh sales. In some areas of the country,
iii
Losses are measured as the difference between energy generated and energy delivered to customers
and thus in practice include losses due to theft. Theft is not considered to be important in the U.S. today,
but it is significant in some other nations.
14
(% of Total Generation)
12
10
0
1926
1930
1934
1938
1942
1946
1950
1954
1958
1962
1966
1970
1974
1978
1982
1986
1990
1994
1998
2002
2006
Source: Data for all years prior to 1943 were reported by the Edison Electric Institute. Losses for the years 1943 to 1951 were
calculated using Edison Electric Institute data on generation, net imports, company use, producer use, and sales to customers.
Company and producer use data were not reported during this time period, so the average of these quantities from 1941 and 1942
was used. Data for all years from 1951 to the present are from the U.S. Energy Information Administration Annual Energy Reviews.
8
Canada
United Kingdom
7
(% of Total Generation)
Australia
Transmission and Distribution Losses Italy
France
6(% of Total Generation)
United
States China
5 Spain
Germany Japan South
Korea
4
3
4.5
2
Israel
4.0
1
3.5
0
3.0
0 100 200 300 400 500
1.5
1.0
0.5 2010 Cents Nominal Cents 0 100 200 300 400 500
0.0
1926
1930
1934
1938
1942
1946
1950
1954
1958
1962
1966
1970
1974
1978
1982
1986
1990
1994
1998
2002
2006
100
United
Kingdom
80
France
60
Italy
40
20
Germany
0
0 50 100 150 200 250 300
A final dimension of performance involves the beginning in the 1990s, spending on average
use of new technology to increase productivity. less than 1% of their revenues on R&D.
1926
1930
1934
1938
1942
1946
1950
1954
1958
1962
1966
1970
1974
1978
1982
1986
1990
1994
1998
2002
2006
The U.S. electric utility industry has historically Figure 1.7 shows a steady decline in collabora-
devoted a very small fraction of its revenues to tive R&D spending through EPRI. The decrease
R&D, instead relying
Transmission primarily onLosses
and Distribution its suppliers in utility R&D funding reflects, in part, reluc-
(% of Total
for innovation. Generation)
U.S. utilities have sometimes tance among utilities to incurand regulators
to approveR&D expenditures as federal and
The U.S. electric utility industry has historically state policies pursued more industry competi-
tion, particularly during the 1990s.
devoted a very small fraction of its revenues to
R&D, instead relying primarily
4.5 on its suppliers Productivity improvement in the electric power
for innovation. 4.0 industry has historically been rapid relative to
most other industries. This is reflected in a
collaborated3.5with vendors on R&D activities decline in real retail prices until the 1970s, as
and have participated
3.0 in collaborative research shown in Appendix A. The lack of a long-term
through the Electric Power Research Institute decline in later years suggests that the difference
2.5
(EPRI), a nonprofit consortium founded in may have decreased, but comparisons to other
2.0 years, however, utilities have
1973. In recent industries remain relatively favorable: the
shifted away1.5from longer-term, collaborative Bureau of Labor Statistics shows that over the
projects and toward shorter-term proprietary 1987 to 2008 period, output per labor hour rose
1.0
efforts.22 Moreover, investor-owned utilities, at an annual rate of 2.4% in power generation
which account
0.5 for almost all nonfederal utility and supply, as compared to 2.1% in the overall
R&D spending, reduced their R&D budgets private nonfarm business sector.23
0.0
600
$U.S. Million (2003 Prices)
500
400
300
200
100
0
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Source: T. Jamasb and M. Pollitt, Liberalisation and R&D in Network Industries: the Case of the Electricity Industry,
Research Policy 37 (2008): 9951008.
50 50 50 50
Data on broader measures of productivity do findings presented below are developed and
not seem to exist, and quantitative international supported in detail in later chapters.
comparisons do not seem possible. Anecdotal
1926
1930
1934
1938
1942
1946
1950
1954
1958
1962
1966
1970
1974
1978
1982
1986
1990
1994
1998
2002
2006
iv
Figure 1.9 relates to concentrated solar power generation, in which direct sunlight heats a working fluid
(oil, historically), which is then used to generate steam to power a turbine. Photovoltaic systems are more
tolerant of clouds (diffuse light) and hence perform well across a broader swath of the U.S. south.
Source: This information was prepared by the National Renewable Energy Laboratory for the U.S. Department of Energy.
This image has been reprinted from the National Renewable Energy Laboratorys GIS website
http://www.nrel.gov/gis/pdfs/windsmodel4pub1-1-9base200904enh.pdf, accessed November 16, 2011.
Cost allocation and siting have been particu- In some regions of the country, this process is
larly contentious for transmission facilities that facilitated by the historic cooperation of states
cross state borders or the 30% of U.S. land within an ISOs territory. But as a general
managed by federal agencies. The Federal matter, the special
Power Act of 1935 made siting of all transmis- difficulties of siting The special difficulties of siting
sion lines a matter for the states rather than boundary-crossing boundary-crossing transmission
FERC, and lines that cross land managed by transmission facilities facilities will pose an obstacle
federal agencies need the approval of those will pose an obstacle
to the efficient integration of
agencies.v Consequently, the construction of to the efficient
interstate transmission facilities requires the integration of renew- renewable generation.
consent of multiple state regulators and, able generation.31
sometimes, one or more federal agencies.
50 50 50
v
In contrast to the Federal Power Act of 1935, the Natural Gas Act of 1938 as amended in 1947, gave the
predecessor of FERC the right of eminent domain to site interstate natural gas pipelines. Interstate pipelines were
already important in the natural gas industry in the 1930s, while interstate transmission of electricity was much
less important than it is now or likely will be in the future.
Source: This information was prepared by the National Renewable Energy Laboratory for the U.S. Department of Energy.
This image has been reprinted from the National Renewable Energy Laboratorys GIS website
http://www.nrel.gov/gis/images/map_csp_national_lo-res.jpg, accessed November 16, 2011.
Figure 1.10 Normalized Load Duration Curves for New England and New York
1.0
0.9
0.8
0.7
Normalized Load
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1 1,001 2,001 3,001 4,001 5,001 6,001 7,001 8,001
Hours of Year
New England Average 19801984 NY Average 19801984
New England Average 20052009 NY Average 20052009
50 50 50 50
Chapter 1: Challenges, Opportunities, and Major Recommendations 15
New York and New England in the period encouraged overnight charging could increase
shown in Figure 1.10. demand when it would otherwise be low, thus
tending to flatten load duration curves.
A second factor that may have been more
important in other regions is the relative FINDING
decline in industrial use of electricity. On Ongoing changes in the character
average during 1950 to 1959, industrial
of electricity demand and the future
customers accounted for half of retail sales of
penetration of electric vehicles will, in
electricity.34 Industrial plants often run around
the clock all year, so the more important the absence of other changes, tend to
industrial load is, the flatter the load duration accelerate the decline in capacity utilization
curve tends to be. The relative importance of in the electric power system. This, in turn
industrial customers has declined steadily since will increase electricity costs.
the 1950s, however, and on average in the 2000
to 2009 period, they accounted for only 28%
of retail sales. This has made load duration Distributed Generation
curves steeper.
Existing policies at state and federal levels
Looking ahead, these trends are likely to favor distributed generation from low-carbon
continue, and their adverse effects on capacity sources, and these policies seem likely to
utilization may be exacerbated by the spread of continue. At the federal level, personal and
electric vehicles (EVs), which include plug-in corporate tax incentives encourage distributed
hybrids and battery electric vehicles. About 94% generators, which are small-scale systems,
of the energy used in the U.S. transportation generally connected to distribution networks.
system comes from petroleum, and more than Most states have programs that subsidize
half of crude oil and petroleum products are distributed generation.37 The California feed-in
imported.35 Increasing the penetration of EVs is tariff program for small-scale renewable
often viewed as an attractive means of reducing generation, particularly rooftop solar units, is
dependence on imported oil. As Chapter 5 perhaps the most visible of these.38 In addition,
discusses, although their penetration generally is 46 states and the District of Columbia have
projected to be slow at the national level under what are called net metering programs, which
current policies,36 these vehicles cannot be compensate end users for generating their own
ignored, even in the near term. EVs are expected energy at the retail electricity rate rather than
to achieve high levels of penetration quickly in the wholesale cost of energy.39 The difference
some high-income areas with environmentally between these rates is mainly the fixed cost of
conscious consumers. Whenever and wherever distribution (and, sometimes, transmission),
they are deployed in large numbers, their impact which is typically recovered by per-kWh
on the grid will depend on when they are charges. When an end user increases genera-
charged. If they are charged when commuters tion, the system saves only the wholesale cost
return home, as seems most likely under current of energy. Under net metering, however, the
policies, they could add significantly to system end user saves both this wholesale cost and
peak loads, worsening the problem depicted in the per-kWh charge used to recover fixed
Figure 1.10. On the other hand, measures that network costs. Thus net metering provides
Finally, as discussed earlier, the industry should One promising recent initiative has been
use the first round of AMI deployments to undertaken to enhance high-value data sharing.
learn how best to employ the capabilities of In February 2010, NERC created two nondis-
these systems and response automation tech- closure agreements to facilitate sharing PMU
nologies to make electricity demand more data. Unfortunately, a year and a half later,
responsive to system conditions. Among other many utilities with significant PMU deploy-
things, further research on consumer reactions ment activity had not yet signed these agree-
to dynamic pricing is needed, and effective ments. If this initiative fails to achieve the
consumer engagement and education strategies hoped-for results, the benefits of PMUs may
must be designed and tested in the field. not be realized. The responsible federal agencies
should take steps to ensure that the critical data
R E CO M M E N D AT I O N at issue are shared appropriately.
The electric power industry should fund
We have identified three additional areas in
additional research and demonstration
which ensuring the appropriate availability
projects to develop: computational tools
of useful data would be particularly valuable.
that will exploit the potential of new hard- While network data on the Western Intercon-
ware to improve monitoring and control nection are available at a level of detail adequate
of the bulk power system (Chapter 2); to support analysis, such data are not available
methods for wide-area transmission plan- for the more complex Eastern Interconnection.
ning (Chapter 4); processes for response to This inhibits both wide-area planning and the
improvement of wide-area planning methods.
and recovery from cyberattacks (Chapter 9);
Obviously, the general availability of detailed
and understanding of consumer response
data on the U.S. bulk power system would raise
to alternative pricing/response automation serious security concerns, but federal agencies
systems (Chapter 7). have a good deal of experience making
2
U.S. Energy Information Administration, Electric Power Industry Overview 2007, http://www.eia.
Sales, Revenue, and Average Price 2009 (Washington, doe.gov/cneaf/electricity/page/prim2/toc2.html.
DC: U.S. Department of Energy, 2010), Table 5A, S. McNulty, Texas Declares New Energy
16
(Washington, DC, 2011), http://www.epa.gov/ Reliability Indices Tracking within the United States
climatechange/emissions/usinventoryreport.html. (Palo Alto, CA, 2003).
5
W. W. Hogan, Electricity Wholesale Market Design Council of European Energy Regulators, 4th
18
in a Low Carbon Future, in Harnessing Renewable Benchmarking Report on the Quality of Electricity
Energy in Electric Power Systems, eds. B. Moselle, Supply 2008 (Brussels, Belgium, 2008).
J. Padilla, and R. Schmalensee (Washington, DC: Electric Power Research Institute, see note 17
19
11
S. M. Kaplan, Electric Power Transmission: in Network Industries: The Case of the Electricity
Background and Policy Issues (Washington, DC: Industry, Research Policy n
Congressional Research Service, 2009): 27. G. Nemet and D. Kammen, U.S. Energy Research
and Development: Declining Investment,
12
U.S. Energy Information Administration, State Increasing Need and the Feasibility of Expansion,
Electricity Profiles 2008 (Washington, DC: U.S. Energy Policy n AND 6 2EZENDES
Department of Energy, 2010), http://www.eia.doe. Electric Utility Restructuring: Implications for Utility
gov/cneaf/electricity/st_profiles/sep2008.pdf. R&D (Washington, DC: U.S. General Accounting
Office, 1998).
13
U.S. Energy Information Administration,
Electricity Retail Choice Is Mandated in Texas and U.S. Bureau of Labor Statistics Productivity
23
Growing in Three States, Today in Energy, May 18, Statistics (February 25, 2009), http://www.bls.gov/
2011, http://www.eia.gov/todayinenergy/detail. lpc/iprprodydata.htm.
CFMID AND 53 %NERGY )NFORMATION
Administration, Participation Lags in Most R. Schmalensee, Renewable Electricity Generation
24
Electricity Retail Choice States, Today in Energy, in the United States, in Harnessing Renewable
May 19, 2011, http://www.eia.gov/todayinenergy/ Energy in Electric Power Systems, ed. B. Moselle,
detail.cfm?id=1450. J. Padilla, and R. Schmalensee (Washington, DC:
RFF Press, 2010), 209232.
Renewable Energy Council, Database (IREC) of Tariffs Available for the Purchase of Eligible Small
State Incentives for Renewables & Efficiency, Renewable Generation, http://www.cpuc.ca.gov/
http://www.dsireusa.org/. PUC/energy/Renewables/hot/feedintariffs.htm.
26
U.S. Energy Information Administration, Electric North Carolina Solar Center and the Interstate
39
Power Monthly September 2011 (Washington, DC, Renewable Energy Council (IREC), see note 25
2011), http://www.eia.gov/cneaf/electricity/epm/ above.
epm_sum.html.
W. Reder, et al., Engineering the Future:
40
27
California Energy Commission, California A Collaborative Effort to Strengthen the U.S.
Renewable Energy Statistics & Data, http://www. Power and Energy Workforce, IEEE Power &
energyalmanac.ca.gov/renewables/index.html. Energy Magazine 8 (July/August 2010): 2735.
28
Ibid. U.S. Power and Energy Engineering Workforce
41
The transmission network is the first link which can have serious economic and social
between large power generation facilities and consequences. More efficient or lower-impact
electricity customers. It supplies energy at high technologies may help solve problems associ-
voltages to substations, where the energy is ated with network expansion, including
distributed to loads at lower voltages via the difficulty in siting new transmission lines to
distribution network. The transmission meet growing demand amid increasing pres-
network today operates reliably and efficiently, sure to limit environmental impact. And
but a variety of technologies offers the potential changes in system operation will help incorpo-
to improve system performance. Sophisticated rate growing penetrations of variable energy
new monitoring systems may reduce the resources, like wind and solar generation.
likelihood of rare cascading system failures,
i
Almost 5,000 generating units with at least 50 megawatts of expected on-peak summer capacity were
registered with the North American Electric Reliability Corporation in 2010.
ii
Though it is energy, not power, that flows through transmission lines, it is common industry practice
to speak of power flows rather than energy flows. With some exceptions, we use the common industry
vernacular in this chapter.
iii
VAR is the unit used to measure reactive power, which is present in an ac system when current and voltage
are not in phase.
Generator
Primary Control Economic Unit Mid-Term Expansion
Protection and AGC Dispatch Commitment Planning Planning
Source: I. J. Perez-Arriaga, H. Rudnick, and M. Rivier, Electric Energy Systems: An Overview, in Electric Energy Systems: Analysis
and Operation, eds. A. Gomez-Exposito, A. Conejo, and C. Canizares (Boca Raton, FL: CRC Press, 2008), 60.
iv
For example, any loss of generation greater than 2,000 megawatts in the Eastern or Western
Interconnection and 1,000 megawatts in ERCOT must be reported, regardless of whether the loss
of generation affects customers.
Generators and
Substations:
t 1SPUFDUJWF3FMBZT
t $JSDVJU#SFBLFST $POUSPM$FOUFS
t 5SBOTGPSNFST
t 4$"%"3FNPUF5FSNJOBM6OJUT
t 4VSHF"SSFTUFST &OFSHZ.BOBHFNFOU
t 1IBTF4IJGUFST 4UTUFN'VODUJPOT
t 'MFYJCMF"$5SBOTNJTTJPO t 4UBUF&TUJNBUJPO
4ZTUFNT t &DPOPNJD%JTQBUDI
t 1IBTPS.FBTVSFNFOU6OJUT t 6OJU$PNNJUNFOU
t 4FOTPST t $POUJOHFODZ"OBMZTJT
-JOFTBOE$BCMFT t "VUPNBUJD(FOFSBUPS
$POUSPM
t &YUSBIJHIWPMUBHF"$-JOFT
t )JHIWPMUBHF%$-JOFT t 1IBTF
.POJUPSJOH"MBSNT
t 4VQFSDPOEVDUJOH-JOFT
t *NQSPWFE4ZTUFN
t 'BVMU$VSSFOU-JNJUFST .PEFMT
t %ZOBNJD-JOF3BUJOH4ZTUFNT t 0TDJMMBUJPO%FUFDUJPO
Note: Existing technologies and functions are listed in black; new and emerging elements are shown in red.
SCADA = Supervisory control and data acquisition.
BOX 2.1 ESTIMATING THE STATE As with any model, the result of the state
OF THE POWER SYSTEM estimator is only an approximation of the actual
The voltages, currents, and power on all the system state. One reason is that sensor mea-
lines in transmission systems are under continu- surements from the supervisory control and
ous monitoring by system operators. These data data acquisition system are not sent at the same
are used in models of power systems that instant; data may be spread over a period of
include the lines, generators, and loads. These several seconds, and thus phase angle data
models are known as state estimators, and their cannot be observed. Another is that these data
are not always precise. State estimators address
output is the estimated system state. The state
these issues by exploiting the redundancy of
of a power system is a snapshot of the system
measurements throughout the system.
voltages and currents at one time that operators
use to assess the condition of the system and, State estimators use an iterative algorithm, and
if needed, take action. For example, operators the estimated system state is obtained after
may use the model results to identify anoma- several attempts to converge on a solution. The
lous system conditions, dispatch generation, algorithm is not perfect, and state estimators
and avoid stability and thermal limits. have trouble estimating a system state during
unusual or emergency conditionsunfortu-
nately, when they are most needed.
120
100
80
Degrees
60
40
Cleveland Michigan Angle
20
0
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
38
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
Minutes
Source: North American Electric Reliability Corporation Real-Time Application of PMUs to Improve Reliability Task Force,
Real-Time Application of Synchrophasors for Improving Reliability (Princeton, NJ, 2010), http://www.nerc.com/filez/rapirtf.html.
v
One other specific benefit of synchronized phasor measurements is mentioned in the next section
on increasing transmission capacity.9
further to ensure that even if any single major Long-Distance Transmission Technologies
system component (such as a large generator or
transmission line) is unexpectedly lost, the Challenging these physical constraints is
power flows on network lines still do not violate increasing political interest in integrating
these limits. Due to the large number of possible variable energy resources, such as wind and
situations under which contingencies might solar, into the grid. These energy resources have
occur, this analysis requires expert judgment, gained favor at the federal and state levels, but
time, and computational power. they are often distant from load centers. The
technical capability exists to provide long-
Figure 2.4 illustrates the typical limiting distance transmission, but the benefit of
factors on the power transfer of short, medium, access to distant renewable resources must be
and long lines. The thermal limit has to do with balanced with the higher cost and difficulty of
material properties of the line and is constant siting such lines. Striking the proper balance in
no matter the length of a line. Since stability transmission for distant renewable generation
is a system property rather than a material sources is a difficult problem to be addressed
property, stability limits change depending on through transmission planning policy and
the length of a line and other system condi- regulation, discussed in Chapter 4. Here we
tions. Thermal considerations generally limit discuss the two technologies appropriate for
power transfer on short lines, while longer lines long-distance transmissionextra-high-
tend to be stability limited. In particular, power voltage ac and high-voltage dc, two relatively
transfer on medium-length lines is usually mature technologies.
constrained by voltage stability, while the
longest lines are limited by transient stability.
Length of Line
Limits:
Thermal Constraint
Long
(More than Voltage Stability
150 miles)
Transient Stability
Medium
(50150 miles)
Short
(050 miles)
EHV ac transmission systems have voltages considerationsthe longer the line, the lower
greater than 242 but less than 1,000 kV. the capacity limit, though the effective length
The highest in commercial operation in the of lines may be extended by installing voltage
U.S. is 765 kV, while 345 kV and 500 kV are support equipment.
standard voltage levels.11 Higher voltage trans- Dc lines can be valuable
mission lines have been installed in China, The transmission system
Russia, and Japan, but only China operates its consists mainly of ac lines
additions to ac transmission
1,000 kV system at its rated voltage.vi, 12 due to their many desirable networks.
Compared to their lower-voltage counterparts, characteristics, such as the
such lines are capable of transmitting more ease of voltage transformation. However, dc
power over longer distances but require larger, lines can be valuable additions to ac transmis-
more expensive transformers, insulators, and sion networks. High-voltage dc lines are not
towers, as well as wider rights-of-way. As a limited by stability considerations and therefore
result, the highest-voltage ac lines are most theoretically are not limited in length.
economical for large-capacity, long-distance Conductor coststhe cost of the metal which
electricity transmission. Installing high-voltage, conducts electricityfor dc transmission lines
large-capacity links can improve reliability by are lower than for ac lines of the same voltage
allowing neighboring areas to support one because fewer conductors are necessary and
another and improving stability characteristics conductor utilization is better. But the cost for
of the network. The length and capacity of a dc substations is significantly higher because
long EHV ac line is typically limited by stability transformers only work for ac, so more
vi
Russia and Japan now operate their 1,000 kV lines at 550 kV.
vii
The transistor and thyristor are both semiconductor devices that function as a switch, but they have
different characteristics. Thyristor switches can be turned on easily but may only turn off under certain
conditions. Transistor switches can be turned on and off easily.
H Capacity Cost L
L Transmission Cost H
Source: Midwest Independent System Operator, Regional Generation Outlet Study (Carmel, IN, 2010).
viii
This is also true of FCLs constructed by other means, such as using power electronic devices.
ix
An island in this context is a self-contained section of the network, within which load and generation are
balanced. Determining where such islands might be created to aid system stability requires careful study
of system contingencies, and creating such an island involves switching many circuit breakers and also
likely shedding some amount of load within the island to balance load and generation.
x
Chapter 9 is devoted to issues surrounding communications, data security, and privacy.
Greater control of system voltages and power Although industry has been and continues to
flow can be achieved through the more exten- be engaged in the development of the technolo-
sive deployment of FACTS devices. The rapid gies discussed in this chapter, their full benefit
control capabilities of these devices can will occur only if there is greater cooperation
contribute significantly to network control if among utilities in their deployment.
integrated with PMUs and wide-area measure-
ment systems. However, the current high costs
of the most versatile of these devices is inhib-
iting their widespread deployment. The inte-
gration of FACTS devices with emerging
wide-area measurement systems can allow their
control capabilities to be leveraged to provide
even greater benefits and could make their costs
more readily justified.
R E CO M M E N D AT I O N
Research and development efforts should
be undertaken to develop 1) the analysis
tools necessary to generate actionable
information from data acquired from PMUs
and 2) the control schemes necessary to
make use of this information by realizing
the complementary potential of PMUs,
FACTS, and other hardware devices.
Risk Assessment of Reliability Performance, Longest and Most Powerful Transmission Link,
http://www.nerc.com/page.php?cid=4|37. press release, July 19, 2010; and V. F. Lescale,
5
Electric Power Research Institute, Estimating the U. strm, W. Ma, and Z. Liu, The Xiangjiaba-
Costs and Benefits of the Smart Grid, technical Shanghai 800kV UHVDC Project Status and
report 1022519 (Palo Alto, CA, 2011). Special Aspects (Paris: CIGR, 2010).
6
A. Vafas, S. Galant, and T. Pagano, Final WP1 J. A. Fleeman et al., EHV AC and HVDC
15
7
North American Electric Reliability Corporation Rudervall, Power System Stability Benefits with
Real-Time Application of PMUs to Improve VSC-HVDC Transmission Systems (Paris: CIGR,
Reliability Task Force, Real-Time Application 2004).
of Synchrophasors for Improving Reliability E. Koldby and M. Hyttinen, Challenges on the
17
(Washington, DC, 2010), http://www.nerc.com/ Road to an Offshore HVDC Grid, presented at the
filez/rapirtf.html. Nordic Wind Power Conference, Bornholm,
8
U.S.Canada Power System Outage Task Force, Denmark, September 1011, 2009.
Final Report on the August 14, 2003 Blackout in the ABB, HVDC Projects by Type and Power, http://
18
above; and Electric Power Research Institute, Phasor see note 8 above, page 75.
Measurement Unit Implementation and Applications, Midwest Independent System Operator, Regional
21
technical report 1015511 (Palo Alto, CA, 2007). Generation Outlet Study (Carmel, IN, 2010).
10
North American Electric Reliability Corporation, EnerNex, see note 19 above; and GE Energy,
22
NASPI Data Sharing Agreement Signatories, Western Wind and Solar Integration Study, report
http://www.nerc.com/page.php?cid=6|319|345. prepared for the National Renewable Energy
11
IEEE Standard Preferred Voltage Ratings for Laboratory (Golden, CO, 2011).
Alternating-Current Electrical Systems and Edison Electric Institute, Out of Sight, Out of Mind
23
Equipment Operating at Voltages Above 230 kV Revisited: An Updated Study on the Undergrounding
Nominal, IEEE Standard 1312-1993, i-2. of Overhead Power Lines (Washington, DC, 2009).
of Long Island Transmission Level HTS Cable, Monitoring and Decision Systems Maximize
IEEE Transactions on Applied Superconductivity 19 Energy Resources, Public Utilities Fortnightly,
(2009): 16921697. December 2009, 18; and A. Bose, Smart
Transmission Grid Applications and Their
25
J. F. Maguire et al., Progress and Status of a 2G Supporting Infrastructure, IEEE Transactions on
HTS Power Cable to Be Installed in the Long Smart Grid 1 (2010): 1119.
Island Power Authority (LIPA) Grid, IEEE
Transactions on Applied Superconductivity 21 D. Streiffert, R. Philbrick, and A. Ott, A Mixed
38
Most projections of U.S. electricity production uncertain because they are subject to only
show a substantial increase from wind and solar limited control and the energy they produce is
generation, which are receiving a strong push less predictable compared to energy from
from state and federal policy through subsidies conventional technologies. Variability and
and renewable portfolio standards. For example, uncertainty are familiar concepts in power
the U.S. Energy Information Administration systems. Through decades of experience, system
(EIA) expects renewable sources to constitute operators have developed approaches to cope
25% of the increase in total generating capacity with variability and uncertainty that stem, for
across the electric power sector between 2010 instance, from changing demand levels and
and 2030.1 failures of generation units. But wind and solar
generation as new sources of variability and
Commonly referred to as variable energy uncertainty present challenges to the operation
resources (VER), wind and solar power of the power system.
generators are known to be variable and
Average Load
Average Wind
Wind (MW)
Load (MW)
Hour
Note: This figure compares daily wind generation and load profiles in California in January 2002. The
colored lines in the upper portion of the plot illustrate each individual days load profile while the lines
in the lower half of the plot illustrate each individual days wind generation profile. The thicker lines
indicate the monthly average profiles for load and wind.
Source: GE Energy Consulting, Intermittency Analysis Project: Appendix B, Impact of Intermittent
Generation on Operation of California Power Grid (Sacramento, CA: California Energy Commission,
2007). 2010 Cents
Figure 3.2 Nevada Solar Photovoltaic (PV) Plant Output on a Sunny Day (Left)
and a Partly Cloudy Day (Right) in 2008
Source: North American Electric Reliability Corporation, Accommodating High Levels of Variable
Generation (Princeton, NJ, 2009).
Average Load
3.1 CHARACTERISTICS OF VARIABLE The power system must have enough response capacity,
ENERGY RESOURCES from interconnections, demand response, storage, and
Figure 3.1, a plot of Californias wind generation
backup supply to maintain reliability standards.
and its relation to demand, illustrates some of the
with low penetrations, which drives an increase
challenges associated with wind generation. As
in costs. Exacerbating these concerns, large-
seen here, wind output can change more rapidly
scale weather events, such as wind storms and
and over a wider range than demand, and it does
wide-area cloud cover, create challenging
not conform to daily cyclic patterns. An inverse
operational situations known as ramp events
correlation is sometimes observed in which wind
that are characterized by a significant portion
resources on average become most abundant
of the VER fleet ramping up or down in unison
during hours of limited demand at night.2
over a period of minutes to hours. Because
ramp events are often difficult to forecast ahead
Compared to wind, solar outputs are generally
of time, additional reserve generation (or
considerably more cyclic and better correlated
demand response) must be ready to increase
with load, typically achieving maximum output a
production (or decrease load) to meet the load
few hours before peak demand. However, espe-
if the VERs ramp down or decrease production
cially at the individual plant level, solar generation
if the VERs ramp up. Finally, the absence or very
without storage can also produce outputs that are
limited presence of the VERs during extended
considerably more variable and less predictable
periods of time (e.g. several consecutive days)
than wind. As illustrated in Figure 3.2, individual
also can result in operation challenges. The power
solar photovoltaic (PV) plants under cloud cover
system must have enough response capacity, from
have been observed to vary their outputs by as
interconnections, demand response, storage, and
much as 90% over the course of seconds.
backup supply to maintain reliability standards
under these worst-case conditions.
When wind variations or passing clouds result
in changes in output, dispatchable generators
All the operational adaptations to accommodate
or loads must make corresponding changes in
VERs require more flexible power systems,
an effort to balance generation with load over a
incurring operational costs. Fossil fuel plants
specific geographic area known as a balancing
must bear the expense of additional start-ups and
region. Larger balancing regions have a greater
shut-downs, as well as operation at output levels
diversity in VER resource availability and load
distant from points of optimal fuel efficiency and
over wide regions, smoothing out the effects of
air pollution control. In addition, more frequent
minute-by-minute variability. Adjacent smaller
start-ups, shut-downs, and ramping can increase
balancing areas can reap the full benefits of
mechanical stress on generation plants, poten-
resource diversity by consolidating or cooper-
tially resulting in higher maintenance costs
ating with each other.
and reduced life.4 Providing these services at
minimum cost while maintaining reliability and
Beyond variability, the high uncertainty
system stability requires careful planning of
associated with VER outputs also can compli-
investments, such as the timely addition of
cate existing balancing processes and is often
dispatchable plants with fast ramping capability.
cited as the main source of costs of integrating
VER generators into the grid.3 Since day-to-day
Beyond the variability and uncertainty of
wind speeds and cloud cover are significantly
their outputs, wind and solar generation have
less certain than the load forecast, systems with
control and electrical characteristics that are
high penetrations of VERs typically must
different from those of conventional
commit more reserve generation than those
synchronous generators. Due to their very low in system operating costs that results from high
penetrations, to date VERs have been required penetration of VERs is also a subject of current
to meet few performance standards in the U.S. debate, as described in Box 3.1.
As VER penetrations grow, it will become
increasingly important that interconnection All the adaptations described in this chapter
standards require all generators, including have been identified and studied in the wind
VERs, to play an active role in helping to integration literature, and many have been
maintain system stability and reliability under implemented in other countries and in parts
the new anticipated conditions. Because retrofit- of the U.S.9 Our focus is on wind farms and
ting generation is typically expensive, it is large-scale PV or concentrated solar thermal
particularly important that interconnection installations without storage. We address the
requirements for all generation technologies, implications of large-scale VERs for planning
both VER and conventional, be designed for and building transmission systems in Chapter 4,
anticipated rather than existing conditions. and we discuss the distribution system challenges
of distributed generation, usually residential
VERs historically have affected the U.S. bulk and commercial rooftop PV installations, in
power system very little because they have Chapter 5.
accounted for only a small fraction of energy
supply. However, with significant growth, these 3.2 VARIABLE ENERGY RESOURCES
policy-backed technologies will require changes AND THE COST OF RESERVES
to how systems are planned, operated, and
controlled. This chapter explores these chal- As explained in Chapter 2, generators are
lenges of system adaptation and the ways that initially dispatched according to unit commit-
grid management and operation can help ment schedules made one day in advance. The
minimize the potential grid-related costs of unit commitment or day-ahead market process
VER expansion. The allocation of the increase determines which generators will come online
i
A variety of terms are commonly used to describe each type of reserve in different systems and different
countries.11
Hours
Note: The red bars mark the typical time when generation schedules become final and binding for
different markets.
Source: Eurelectric, Integrating Intermittent Renewables Sources into the EU Electricity System by 2020:
Challenges and Solutions (Brussels, Belgium, 2010).
increases from the integration of VERs in the many of these plants simultaneously would
U.S.12 Understanding the full impact of this stop producing.13 Even when VERs do simulta-
increase in operating reserves on costs will neously ramp down, they typically take minutes
require a closer examination of how VERs to hours, which is much slower than traditional
affect each category of reserves. contingencies, such as a large generation plant
tripping off-line, that occur in seconds.iii
Regulating Reserves. The uncertainty and
the variability of VERs can create fluctuations Ramping and Load-following Reserves.
in production on the order of minutes. Ramping and load-following reserves are
Accommodating these fluctuations may require primarily used to counteract VER forecast
a modest increase in fast-responding regulating errors in the day-ahead scheduling of plants.
reserves required for normal operations. At this timescale, wind forecasts have an
average error of 15%30% mean absolute error
Frequency Response and Supplemental of production, despite the significant improve-
Reserves. All major international studies have ments in wind forecasting over the past decade,
concluded that large VER penetrations do not as illustrated in Figure 3.3.14 For comparison,
significantly increase the risks for traditional day-ahead load forecast errors are typically
contingencies.ii Large penetrations are expected below 1% mean average error of production.15
to be comprised of many small plants on the As a result, operators must conservatively
order of 15 megawatts (MW) over wide operate the system assuming that the actual
geographic areas, and it is highly unlikely that VER production could be at least 20% lower
ii
These studies assume that requirements for fault tolerance are present in grid interconnection standards.
This issue is described further in Section 3.4.
iii
An exception is the occurrence of wind exceeding maximum allowable speed and causing multiple
turbines to shut down simultaneously.
Note: The red line indicates the final scheduled wind generation (economic dispatch base point).
The blue line indicates actual wind generation.
Source: North American Energy Reliability Corporation, NERC IVGTF Task 2.4 Report: Operation
Practices, Procedures, and Tools (Princeton, NJ, 2011).
than forecasted, requiring large quantities of December 7 to nearly 2,000 MWa substantial
reserves, which are then dispatched throughout portion of the agencys total wind generation
the day if forecast errors become apparent. capacity (around 3,000 MW at the time). BPA
deployed more than 90 percent of its reserves
Load-following and ramping reserves also are as the wind ramped up and BPA had to curtail
used to hedge the risk of VER ramp events. some wind generation by issuing generation
Though not a traditional contingency, ramp limits to wind generators, in order to avoid
events can present a significant operating violating its reserve requirements. Wind output
concern. The risk of ramp events in most cases then ramped down to zero through the night
comes from the uncertainty of when they will hours of December 7 and the morning hours
occur and how long they will last. Indeed, of December 8. This down ramp resulted in
predictions of ramp event occurrence and BPA depleting 100 percent of its reserves and
timing are often difficult and can result in large prompted BPA to curtail transmission schedules
and sudden forecast errors that deplete reserves for wind generators. Uncertainty in both the
that are on-line in a time too short to activate size and duration of ramp events complicates
reserves that are fully off-line.16 operator responses.
Figure 3.4 shows two ramp events that depleted In the remainder of this section, we discuss a
reserves procured by the Bonneville Power set of tools that can serve to reduce the cost of
Administration in December 2009. Wind reserves associated with VERs. Many of these
output ramped up sharply in the late hours tools will allow the electricity market to better
of December 6 and the early morning of address the impending need for flexible
Source: Copyright Fraunhofer IWES, Germany; Institute for Solar Energy Technology, Wind Energy
Report Germany 2005 (Kassel, Germany, 2005); H. Holtinnen et al., Design and Operation of Power
Systems with Large Amounts of Wind Power: Final Report, Phase One 20062008, research note 2493
(Espoo, Finland: VTT, 2009).
iv
It is technically possible to operate nuclear generation facilities flexibly. Achieving flexible operations
requires relatively minor technical modifications to plant designs and fuel content. lectricit de France
regularly cycles their reactors to provide a range of grid services, including primary frequency regulation
and daily load following. Some of their units experience up to 250 power variations between 10% and
80% of their nominal rated capacity per year.47
v
In this report, hydro generation refers to impoundment, or dam-based, hydropower unless otherwise
noted. This is in contrast to run-of-river hydropower installations with little or no water storage. By
avoiding the high dams and extensive land inundation that characterize impoundment hydropower,
pure run-of-river installations are typically smaller and have less impact on the environment. However,
these facilities produce variable outputs and are often categorized as a type of VER.54
vi
Some modern wind turbines can employ blade-pitch control to affect some control of input.
vii
This provision is in the Institute of Electrical and Electronics Engineers Standard 1547.
and Data Assimilation to Improve Guidance for Preventing Undue Discrimination and Preference
Energy, presentation at the RE Prediction- in Transmission Services, submitted to the Federal
Research WorkshopNCAR, Boulder, Colorado, Energy Regulatory Commission, Docket No.
May 11, 2010; and J. Wilczak, Comments on RM05-25-000 (2006); and J. D. Chandley and W.
Docket Number RM10-11-000, submitted to the W. Hogan, Reply Comments on Preventing
Federal Energy Regulatory Commission, Docket Undue Discrimination and Preference in
No. RM10-11-000 (2010). Transmission Services, submitted to the Federal
Energy Regulatory Commission, Docket No.
19
ISO/RTO Council, Variable Energy Resources, RM05-25-000 (2006).
System Operations and Wholesale Markets (2011).
Federal Energy Regulatory Commission, see
33
20
Ibid. note 5 above.
21
Federal Energy Regulatory Commission, see note 5 Holtinnen, see note 13 above.
34
above.
M. Laughton, Variable Renewables and the Grid:
35
22
Wilczak, see note 18 above. An Overview, in Renewable Energy and the Grid:
23
National Oceanic and Atmospheric The Challenges of Variability, ed. G. Boyle (Oxford,
Administration, Comment to Change Intent of UK: Earthscan, 2007), 31.
Notice of Proposed Rulemaking to Include NOAA Holtinnen et al., see note 13 above; and A. Mills
36
as a Mandatory Recipient of VER Atmospheric and R. Wiser, Implications of Wide-Area Geographic
Observations on Docket RM10-11, submitted Diversity for Short-Term Variability of Solar Power
to the Federal Energy Regulatory Commission, (Berkeley, CA: Lawrence Berkeley National
Docket No. RM10-11-000 (2011). Laboratory, U.S. Department of Energy, 2010).
24
James Wilczak, Team Lead, Boundary Layer GE Energy and National Renewable Energy
37
Processes and Applications, Physical Sciences Laboratory, see note 3 above.
Division, National Oceanic and Atmospheric
Administration, conversation with authors, G. Giebel, P. Srensen, and H. Holttinen, Forecast
38
Laboratory, U.S. Department of Energy, 2010). Electric Institute: FERC Notice of Inquiry on the
Integration of Variable Energy Resources,
27
Ibid. submitted to the Federal Energy Regulatory
28
North American Electric Reliability Corporation, Commission, Docket No. RM10-11-000 (2010).
NERC IVGTF Task 2.4 Report: Operating Practices, International Energy Agency, see note 17 above;
41
Procedures, and Tools (Princeton, NJ, 2010). North American Electric Reliability Corporation,
29
International Energy Agency, see note 17 above; NERC IVGTF Task 1.4 Report: Flexibility Require-
and GE Energy and National Renewable Energy ments and Metrics for Variable Generation
Laboratory, see note 3 above. (Princeton, NJ, 2010); and North American
Electric Reliability Corporation, see note 28 above.
30
American Wind Energy Association, Comments
of the American Wind Energy Association on VER Holttinen et al., see note 10 above.
42
NOI, submitted to the Federal Energy Regulatory GE Energy and Independent System Operator
43
Commission, Docket No. RM10-11-000 (2010). New England, New England Wind Integration Study
31
ISO/RTO Council, Increasing Renewable Resources: Final Report (Schenectady, NY, 2010); GE Energy,
How ISOs and RTOs Are Helping Meet This Public Ontario Wind Integration Study, report prepared
Policy Objective (2007). for Ontario Power Authority, Independent
Electricity System Operator and Canadian Wind
Energy Association (Schenectady, NY, 2006); and
New York Independent System Operator, see
note 2 above.
Power Inflexible? Nuclear Future 5, no. 6 (2009): Compressed Air Energy Storage, SmartGrid.gov,
333341; and M. Debes, EDF Nuclear Sustainable http://sgdev.nrel.gov/project/new_york_state_
Production: Challenges and Adaptation to Market electric_gas_advanced_compressed_air_energy_
Needs, presentation at MIT-IEEJ Energy and storage; and Pacific Gas & Electric Company:
Global Change Workshop, Tokyo, Japan, Advanced Underground Compressed Air Energy
September 30October 1, 2010. Storage, SmartGrid.gov, http://sgdev.nrel.gov/
project/pacific_gas_electric_company_advanced_
D. Lew et al., How Does Wind Affect Coal?
48
underground_compressed_air_energy_storage.
Cycling, Emissions, and Costs, presentation at
WindPower 2011, Anaheim, CA, May 25, 2011. Denholm et al., see note 57 above.
62
U.S. Department of Energy, 2006); and B. T. Smith, Corporation, Wind Integration Study for Public
U.S. Hydropower Fleet and Resource Assess- Service of Colorado, Addendum: Detailed Analysis
ments, presentation at the National Hydropower of 20% Wind Penetration, prepared for Xcel
Association Annual Conference, Washington, DC, Energy (Knoxville, TN, 2008), http://www.uwig.
April 5, 2011. org/CRPWindIntegrationStudy.pdf.
Electric Advisory Committee, see note 51 above.
54 I. M. Dudurych, Statistical Analysis of Frequency
66
Responsive Wind Plant Controls Impacts on Grid Interconnection Standards Review Initiative: Draft
Performance, paper presented at the IEEE PES Straw Proposal (Folsom, CA, 2010).
General Meeting 2011, Detroit, MI, July 2428,
D. Brooks and M. Patel, Overview of NERC
72
2011; and RenewableUK, RenewableUK Position
Integrating Variable Generation Task Force Efforts
Paper on Inertia, Version 3.0 (London, 2011),
on High Penetrations of DERs, paper presented at
http://www.bwea.com/pdf/publications/
the IEEE PES General Meeting 2011, Detroit, MI,
RenewableUK_Inertia_Position_Paper.pdf.
July 2428, 2011.
68
J. Brisebois and N. Aubut, Wind Farm Inertia
North American Electric Reliability Corporation,
73
Emulation to Fulfill Hydro-Qubecs Specific
see note 10 above.
Need, paper presented at the IEEE PES General
Meeting 2011, Detroit, MI, July 2428, 2011. North American Electric Reliability Corporation,
74
Several factors beyond normal growth in renewable generation, such as large wind farms,
electricity demand will require new investment that will be connected to the high-voltage
in transmission capacity in the next two transmission system.
decades. Perhaps the most important is the
need to integrate large-scale renewable genera- Also driving expansion, the recession-induced
tion. The U.S. federal government and all state fall in electricity demand and the shale gas
governments provide financial support for the revolution have lowered electricity prices and
use of renewable energy to generate electricity, adversely affected the economics of operating
and 29 states and the District of Columbia have many older coal-fired units. A suite of new
enacted quantitative requirements.1 These rules from the U.S. Environmental Protection
programs will spur growth in large-scale Agency may provide further incentives for early
i
As a general matter, it is important to recognize that non-transmission investments can sometimes serve the
same purpose as transmission investments and that transmission investments do not always take the form
of new towers and wires on new rights-of-way.13
ii
A discussion of the reliability standards that are used to govern generation and transmission investment
decisions is beyond the scope of this study. For the discussion in this chapter, we take them as given.
However, the criteria underlying current reliability standards do not necessarily reflect rigorous costbenefit
analyses, an issue FERC is addressing.14
iii
This point is well illustrated by Midwest ISOs 2009 transmission expansion plan, which included $4 billion
worth of exclusively reliability lines that were expected to provide nearly $3.4 billion in economic benefits.20
Midwest ISOs systematic assessment of these investment benefits is unusual, though. The frameworks for
assessing the economic value of these reliability upgrades tend to be either very weak or nonexistent in most
regional transmission organizations.
processes that allow for public policy benefits. Some projects to deliver renewable power will
FERC Order No. 1000 will require that all local not increase reliability, nor will they necessarily
and regional transmission planning processes bring economic benefits until state or federal
include policy objectives. It also mandates that policies sufficiently reduce the cost of renew-
public utility transmission providers establish ables or increase the cost of fossil-powered
procedures to identify transmission needs generators. However, the policy dimension can
driven by public policy requirements and be made commensurable with economic and
evaluate alternative ways of meeting those reliability benefits for purposes of evaluating
needs. transmission investments.
iv
Alaska, Hawaii, and the ERCOT region of Texas are exceptions. In addition, about one-third of the
high-voltage transmission system in the 48 contiguous states is owned by government enterprises,
cooperatives, and other entities not subject to FERC regulation.21 These entities have much higher shares
in some regions. The legal and policy problem of harmonizing their behavior with that of the entities
regulated by FERC is complex and important, but it is beyond the scope of our study.
v
The corresponding contract rates may be similar to those that result from standard regulatory processes,
although they also may be higher because merchant investors are not prevented from seeking a high rate
of return when bargaining with beneficiaries.
vi
For example, to plan transmission for the state of Colorado, the Public Service Company of Colorado
participates in TEPPC, the regional transmission planning group WestConnect, the Colorado Coordinated
Planning Group within WestConnect, and state planning procedures.
Two recent developments may serve to expand While existing planning arrangements have
the effective scope of hierarchical wide-area enabled construction and ongoing expansion
planning in the U.S. Title IV of the Recovery of a reliable and efficient transmission grid,
Act appropriated $80 million to establish extensive use of renewable resources distant
interconnection-wide planning collaboratives from major load centers will require stronger,
with the goal of facilitating the development permanent interconnection-wide planning
of regional transmission plans and providing procedures, particularly in the East. The West
assistance in the form of modeling, support has demonstrated that this can be accomplished
to regions and States for the development of collaboratively. It may not require another layer
coordinated State electricity policies, programs, of authority or bureaucracy, though some
laws, and regulations. The Office of Electricity central staff and modeling capability will
Delivery and Energy Reliability at the DOE presumably be required to perform the
issued awards to five separate organizations, two top-down analysis necessary to complement
each in the Eastern and Western Interconnections existing bottom-up processes, as well as an
and one in ERCOT. These organizations are agreed-upon definition of planning criteria, a
made up of regional planning authorities from prescribed periodicity of the plan, transparency,
across each interconnection, but their effective- and adequate stakeholder participation. In
ness and impact on what actually gets built addition, problems of planning methods and
remain to be determined. Moreover, because of data availability will need to be addressed.
the nature of their funding, the new collabora- We discuss these next.
tives are only supported for a single round of
analysis and may well disappear afterward.
Even so, some generalizations are possible. Rather than proposing a specific cost-allocation
In most ISO regions, regional cost allocation method, we next lay out some fundamental
is performed differently for projects that serve principles that complement and extend the
primarily reliability, economic efficiency, and FERC order and then explore their application
generator interconnection purposes. Costs of to the transmission planning process.
projects justified solely on reliability grounds
are usually recovered on a uniform basis from Principles of Cost Allocation
electricity users. This general approach is
commonly referred to as socialization of An important policy challenge is to develop
costs. Individual users shares of total costs principles and procedures for mandatory cost
typically depend on total or peak consumption. allocation. Participant funding for various
forms of merchant lines or sale of capacity
rights on large dc lines provide other cost
vii
Large, discrete transmission investments can have a material effect on prospective market prices and the
distribution of benefits, and the associated costs are largely fixed ex post. Rather than recover those costs
through ordinary transmission tariffs, it is better to use a multipart charging mechanism that couples
ex post efficient usage pricing (reflecting congestion and marginal losses) with fixed access charges, assigned
to members of the coalition of putative beneficiaries in a way that preserves net benefits for being a member
of the coalition.
viii
In principle, if the project has positive net benefits, it is possible to compensate any losers for their losses
and make all affected entities better off. In practice, this is complicated and seldom, if ever, done. It is
generally argued that compensation is not deserved for the loss of economic benefits (high prices to
generators, low prices to loads) that exist only because of network congestion, but major environmental
impacts may raise more serious issues in the future. Such impacts might be claimed, for instance, if a
proposed line would cross a particular state but confer no benefits on its residents.
An inferior but commonly used alternative to the could cause a beneficial project not be built
beneficiary-pays principle is the socialization of cost, because not enough of the benefits have been
captured to cover the costs.
which spreads it uniformly throughout a region.
allocate costs in such a way as to make all An inferior but commonly used alternative
entities better off. Thus adopting the benefi- to the beneficiary-pays principle is the social-
ciary-pays principle helps with decisions about ization of cost, which spreads it uniformly
what should be built, as well as determining throughout a region. Socialization eliminates
who should pay for what is built. Fairness is locational signals, reducing the systems ability
important, but support of consistent incentives to promote investment in the best locations.
for investments is the key reason for embracing For instance, all else equal, socialization would
this principle. Of course, failure to recognize all always favor the best wind or solar resources,
beneficiariesthe generators, in particular regardless of their location and impact on
ix
This danger is illustrated clearly by PJM tariff submissions, which show a divergence as large as $1.2 billion
in cost-allocation outcomes of a specific project for socialization compared to the PJM DFAX flow-based
method.48 While DFAX is not a perfect application of the beneficiary-pays principle, it does employ one
commonly employed proxy for benefits. The Florence School of regulation has provided a useful
comparative analysis of cost allocation methods.49
Transmission Line
Total Benefit = 120
Generator Benefit = 60 Load Benefit = 60
Line Cost = 100
All this stands in contrast to some current generation and load. A consistent parsing of
procedures, in which, for instance, generators the benefits, with careful consideration of the
are at least initially responsible for the entire geographic scope of benefits from increased
cost of radial interconnection lines, while load reliability, allows for estimation of cost shares
entirely bears the cost of other network rein- that make the beneficiaries better off while
forcements. Regulators should be aware of the respecting the principle that those in regions
link that exists between the economic value of who do not benefit do not pay. This approach
any subsidy to renewables and the decision of can yield a workable approximation to a
how much to charge these generators as a beneficiary-pays allocation and make cost
function of their economic benefits. socialization a last resort.50
r 5IFBMMPDBUJPOPGDPTUTTIPVMECF
Applying this principle sometimes requires
established ex ante, before the project
judgment, as when significant uncertainty
about future benefits is expected to be resolved is built.
in the relatively near term. For example, the
imminent location of a large generating plant
might have a material impact on the expected Cost Allocation in Practice
distribution of benefits. In such cases, it may
In practice, the process of determining
who benefits and how much they benefit is
Once costs have been allocated on the basis of analytically complex, especially because of
anticipated benefits, that allocation should be left in uncertainty regarding future benefits.53 In
place for the life of the projector at least a long addition, the lumpiness and economies of scale
period on the order of a decade. of transmission lines means that it is often
sensible to build facilities with excess capacity
be most sensible to allocate cost after the in anticipation of future needs. This may
uncertainty is resolved rather than risk provide economic justification for at least
imposing costs well in excess of ex post benefits partial socialization of many projects during
on some parties. Current transmission cost- the early years of their operational lives, when
allocation methods in some interregional they are not operating at full capacity. If a
markets, such as in Central America and the portion of capacity is put in place to serve
EU, have adopted this approach. potential future users, it may be sensible to
socialize the costs for the early years of opera-
Applying these principles perfectly is difficult tion until those users appear.
in practice. But unless one begins with sound
principles and departs from them only to the Cost-allocation procedures are only workable
extent required by practical considerations, if they are roughly compatible with voluntary
the final system of cost allocation will lack agreements by the parties involved. The three
coherence, and the resulting pattern of invest- principles of cost allocation form a sound
ments is likely to be inefficient. foundation on which mutually beneficial
agreements can be constructed. Recognizing
that the first of themthe beneficiaries-pay
principlecan have somewhat different
operational meanings in different settings,
s -AINTAIN CURRENT TRANSMISSION NETWORK The former might emerge from discussions
charges for existing lines, and apply the among planning regions or be imposed by
cost-allocation method only to new or FERC. An interesting example of such a proce-
recently built lines.x dure is the European Inter-TSO Compensation
Mechanism.55 This mechanism uses network
s &OR ANY NEW POTENTIAL GENERATOR ESTIMATE flows as a proxy for benefits and deals only with
the benefits of recently built lines and lines the allocation of costs between regional system
that are scheduled to enter in service shortly. operators (TSOs in European terminology).
Any existing transmission capacity expansion Planners first employ flow-based methodologies
planning procedure should be helpful to determine how much external agents use the
in the evaluation of the benefits of new network of each region. Then they calculate
transmission facilities, and generators that the costs associated with that usage and allocate
have been built recently or are planned to be them to the corresponding external regions.
built soon have an impact on the network The net balance is credited or charged to each
expansion plans. region, and its network users pay based on that
regions chosen tariff method. This hierarchical
s 5SE THIS INFORMATION TO COMPUTE THE scheme provides a workable (if imperfect)
transmission charges for recently installed interregional cost-allocation system anchored
and prospective generators and loads, as well in beneficiary-pays logic and leaves each
as for the remaining network users. regional operator free to define its own system
for intraregional cost allocation.
Even gradual implementation of these guide-
lines should move transmission tariffs toward On the other hand, it may be difficult as a
a more cost-reflectiveand thus more practical matter for the many regional authori-
economically efficientstructure.54 ties involved in both the Eastern and Western
Interconnections to reach agreement on this or
Cost-allocation procedures for interregional, any other common interregional approach. The
interconnection-wide, and renewable genera- alternative is a set of bilateral or multilateral
tion projects are even less developed than those cost-allocation agreements. This outcome
for intraregional projects. But they will become would obviously lead to processes that would
more important in the future if large-scale be more sensitive to regional differences, but
renewable generation is substantially expanded. such a system would be less adept at dealing
with multiregional problems involving, for
Interregional or Interconnection-Wide Cost instance, loop flows. In this case, FERC might
Allocation. FERCs Order No. 1000 calls for the consider developing default procedures for
development of standardized interregional multiregional issues in the absence of prior
planning and cost-allocation procedures. agreements among all affected parties.
The result could be either common
x
Our major concern is the allocation of the costs of new transmission investments. Transmission
charges for the existing lines to new network users are of lesser importance to this study, but some
practical implementation guidelines can be found in the literature.56
s !PPLY COST
ALLOCATION METHOD ONLY TO NEW Two other issues associated with renewables
transmission projects. development deserve mention. First, wind and
solar power plants are often built in relatively
s 7HENEVER POSSIBLE USE AN ESTIMATE OF THE small increments of several tens to a few
projects benefits to allocate costs among the hundred megawatts, typically substantially less
regions involved; if this is not possible, some than the standard sizes of conventional thermal
measure of network usage might be used as a power plants. This is true even in areas with
proxy for benefits. good resources that may ultimately support
many such plants. High-voltage transmission
s 5SE THIS INFORMATION TO COMPUTE THE FRACTION lines, however, are often most efficiently
of the cost of the project to be covered by constructed at scales designed to serve a
each involved region. gigawatt of capacity or more. The effect of this
mismatch is that large amounts of transmission
s !LLOW EACH REGION TO ALLOCATE THE COST IT MUST capacity may not be used until more generation
cover to its network users according to its own comes online in an area, which could take years.
internal procedures, which should be built on In the meantime, a relatively small generator
the three basic principles presented here. could face the cost burden of oversized, under-
utilized transmission system upgrades.
Remote Renewable Generators. When all
generators were built by vertically integrated A second problem emerges because many of
utilities, lines connecting the generator to the the best wind and solar resources are far from
transmission grid were treated as (a generally the existing transmission system. Transmission
small) part of the cost of the generator. utilities have little interest in building capacity
Transmission cost differences were thus to remote areas far from the existing infra-
automatically taken into account in making structure, since it is generally unclear under
locational decisions, but only in unusual cases, current tariffs who would pay for such lines.
like mine-mouth coal plants, did they have As a result, transmission utilities would like
much of an effect on those decisions. to wait for generators to build first so that they
can then finance the transmission upgrades
In contrast, different entities may now build necessary to accommodate these resources.
and own generation and transmission facilities, Of course, generators will not build if their
and the cost of interconnection lines and other plants have to sit idle for years before they can
required transmission network upgrades may interconnect to the grid and start to sell their
represent a significant fraction of the cost of the power.57 This is the classic chicken-and-egg
remote, large-scale wind and solar plants that problem with transmission.
may be developed in the future. Under the
traditional approach, the generator would pay There are creative approaches to address both
the full project cost. But in the case of remote problems. Initially, regions could allocate the
renewables, this is likely to represent a signifi- cost of new transmission projects in remote
cant departure from the beneficiary-pays areas where wind or solar development is
principle. That principle should be applied to anticipated to load. Then, as generators come
xi
The U.S. Chamber of Commerce maintains a useful resource that describes many of these challenges and
includes a list of active transmission projects facing major hurdles (http://www.projectnoproject.com/
category/project/transmission/).
asp?PublicationID=1243.
Midwest ISO, Midwest ISO Regional Generation
17
3
Energy Information Administration, Annual Energy Outlet Study (Carmel, IN, 2010), https://www.
Outlook 2011 (Washington, DC, 2011). midwestiso.org/_layouts/miso/ecm/redirect.
4
Federal Energy Regulatory Commission, aspx?id=18974; and W. Hogan, Transmission
Transmission Planning and Cost Allocation by Benefits and Cost Allocation, white paper
Transmission Owning and Operating Public Utilities, (Cambridge, MA: Harvard University, 2011),
Order No. 1000 (July 21, 2011), http://www.ferc. http://www.hks.harvard.edu/hepg/Papers/2011/
gov/industries/electric/indus-act/trans-plan/ Hogan_Trans_Cost_053111.pdf.
fr-notice.pdf. Joskow, see note 13 above.
18
5
U.S. Department of Energy, 20% Wind Energy by Kwok, see note 10 above.
19
pdf; and Southwest Power Pool, The Benefits of 2009 (Carmel, IN, 2009), https://www.misoenergy.
a Transmission Superhighway (Little Rock, AR: org/Library/Repository/Study/MTEP/MTEP09/
Southwest Power Pool Communications MTEP09%20Report.pdf.
Department, 2011), http://www.spp.org/. S. M. Kaplan, Electric Power Transmission:
21
6
Midwest ISO, PJM Interconnection, Southwest Background and Policy Issues (Washington, DC:
Power Pool, Tennessee Valley Authority, Congressional Research Service, 2009).
Mid-Continent Area Power Pool, Joint Federal Energy Regulatory Commission, see note 4
22
Coordinated System Plan (2008). above.
7
Coalition for Fair Transmission Policy, http:// J. I. Prez-Arriaga, F. J. Rubio, J. F. Puerta, J. Arceluz,
23
www.thecftp.org/; and D. Springer and G. Dierkers, and J. Marn, Marginal Pricing of Transmission
An Infrastructure Vision for the 21st Century. Services: An Analysis of Cost Recovery, IEEE
(Washington, DC: National Governors Association, Transactions on Power Systems 10, no. 1 (1995):
2009), 1113. 546553; and Joskow and Tirole, see note 14 above.
8
Federal Energy Regulatory Commission, see note 4 J. Pfeifenberger, Barriers to Transmission
24
above. Investments and Implications for Competition in
9
E. Hirst, Transmission Crisis Looming, Public Wholesale Power Markets, presentation to the
Utilities Fortnightly 138, no. 14 (2000): 54. American Antitrust Institute, Washington, DC,
April 12, 2011; and R. Coxe and L. Meeus, Survey
10
P. J. Kwok, Electricity Transmission Investment in of Non-Traditional Transmission Development
the United States: An Investigation of Adequacy presentation to IEEE Power and Energy Society
(Masters thesis, MIT, 2010). 2010 General Meeting, Minneapolis, MN, July
11
Edison Electric Institute, Statistical Yearbook of the 2010, Paper 978-1.
Electric Power Industry (Washington, DC, 2009). Blue Ribbon Panel on Cost Allocation, A National
25
12
R. Gutman and E. R. Wilcox, 21st Century Perspective: On Allocating the Costs of New
Transmission Planning: The Intersection of Transmission Investment: Practice and Principles,
Engineering, Economics, and Environment, white paper prepared for the Working group for
presented at Integration of Wide-Scale Renewable Investment in Reliable and Economic electric
Resources into the Power Delivery System, CIGRE/ Systems (Washington, DC: WIRES, 2007).
IEEE PES Joint Symposium, Calgary, Canada, July R. J. Thomas, J. T. Whitehead, H. Outhred, and T.
26
2931, 2009; and Kwok, see note 10 above. D. Mount, Transmission System PlanningThe
13
P. L. Joskow, Patterns of Transmission Old World Meets the New, Proceedings of the IEEE
Investment, working paper 2005-004 (Cambridge, 93, no. 11 (2005): 20262035.
MA: MIT Center for Energy and Environmental
Policy Research, 2005).
and Transmission Study (Golden, CO: National American Electric Reliability Corporation, see
Renewable Energy Laboratory, U.S. Department note 33 above; and REALISEGRID, see note 34
of Energy, 2010). above.
28
Western Electricity Coordinating Council, U.S. Department of Homeland Security, National
40
Transmission Expansion Planning Policy Power Grid Simulation Capability: Needs and Issues
Committee, http://www.wecc.biz/committees/ (Argonne, IL: U.S. Department of Homeland
BOD/TEPPC/default.aspx. Security, Science and Technology Directorate,
2008).
29
PJM Interconnection, PJM Regional Transmission
Expansion Planning Process (Valley Forge, PA, Renewable Energy Transmission Initiative
41
2008); and Southwest Power Pool, Integrated Coordinating Committee, Renewable Energy
Transmission Planning: Process Document Transmission Initiative: Phase 1B, Final Report
(Little Rock, AR, 2009). (Sacramento, California, January 2009), http://
www.energy.ca.gov/2008publications/RETI-1000-
30
ENTSO-E, Ten-Year Network Development Plan 2008-003/RETI-1000-2008-003-F.PDF; and Public
20102020 (Brussels, Belgium, 2010). Utility Commission of Texas, CREZ Transmission
31
ENTSO-E, Study Roadmap towards Modular Program Information Center, http://www.
Development Plan on Pan-European Electricity texascrezprojects.com.
Highways System 2050: Way to 2050 Pan-European PJM Interconnection, A Survey of Transmission
42
Power System (Brussels, Belgium, 2011), https:// Cost Allocation Issues, Methods and Practices
www.entsoe.eu/system-development/2050- (Valley Forge, PA, 2010).
electricity-highways/.
Federal Energy Regulatory Commission, see note 4
43
32
Federal Energy Regulatory Commission, see note 4 above.
above.
Hogan, see note 16 above; and F. J. Rubio and J. I.
44
33
North American Electricity Reliability Prez-Arriaga, Marginal Pricing of Transmission
Corporation, Accommodating High Levels of Services: A Comparative Analysis of Network Cost
Variable Generation (Princeton, NJ, 2009), http:// Allocation Methods, IEEE Transactions on Power
www.nerc.com/filez/ivgtf.html. Systems 15, 1 (2000): 448454.
34
ENTSO-E, Research and Development Plan: F. Schweppe, M. C. Caramanis, R. D. Tabors, and R.
45
European Grid Towards 2020 Challenges and E.Bohn, Spot Pricing of Electricity (Dordrecht, the
Beyond (Brussels, Belgium, 2010); G. Latorre, Netherlands: Kluwer Academic Publishers, 1988).
R. D. Cruz, J. M. Areiza, and A.Villegas,
Classification of Publications and Models W. Hogan, Contract Networks for Electric Power
46
Framework Program Research Project (2009), Corporation on Remand, FERC Docket No.
http://realisegrid.rse-web.it/. EL05-121-006, May 28, 2010; and PJM Inter-
connection, Response of PJM Interconnection,
35
V. Rious, J. Glachant, and P. Dessante,
L.L.C. to Information Requests by FERC, FERC
Transmission Network Investment as an
Docket No. EL05-121-006, April 13, 2010).
Anticipation Problem, working paper EUI RSCAS,
2010/04 (Florence, Italy: Loyola De Palacio Florence School of Regulation, A Study on the
49
Regulatory Approach to Transmission Charging, Siting in the Western United States: Overview and
Office Journal of the European Union 250 Recommendations Prepared as Information to the
(September 24, 2010), 511. Western Interstate Energy Board, white paper
(Salt Lake City, UT, and Denver, CO: Holland &
P. Fox-Penner, J. Pfeifenberger, and D. Hou,
53
Hart LLP, 2009).
Comments of Peter Fox-Penner, Johannes
Pfeifenberger, and Delphine Hou, FERC Docket U.S. Department of Energy et al., MOU on
66
on the Inter-TSO Compensation Mechanism and Public Service Electric & Gas Group,
68
i
It is important to note that distributed generation is distinct from dispersed generation, which is not
connected to the grid. Dispersed generation is typified by standby diesel generators that provide backup
power in the event of a grid failure. Because these units typically do not impact utility operation or planning
activities, we do not discuss them. Though not connected to the grid, dispersed generators can participate in
demand response programs (see Chapter 7).
Source: U.S. Department of Energy, The Potential Benefits of Distributed Generation and Rate-Related Issues that May Impede
Their Expansion: A Study Pursuant to Section 1817 of the Energy Policy Act of 2005 (Washington, DC, 2007); and P. Chiradeja and
R. Ramakumar, An Approach to Quantify the Technical Benefits of Distributed Generation, IEEE Transactions on Energy Conversion 19,
no. 4 (2004): 764773.
generation can be owned and operated by engines, combustion turbines, and steam
utilities or their customers and can provide a turbines comprised more than 4 GW each of
variety of theoretical benefits to their owners installed capacity, while hydroelectric, wind, and
and the broader power system. Large DG units other generator technologies totaled 3 GW.4 In
are typically dispatchable and communicate the same year, 93,000 residential PV installations
with system operators like central station totaled about 450 MW of capacity.5 While 90%
generation facilities do. However, neither of solar PV installations between 1998 and 2007
utilities nor system operators typically monitor were smaller than 10 kW, the largest installations
or control the operation of small DG units, generated more than 14 MW.6
especially those in residential applications.
Renewable DG from wind and solar power also Federal and state policies are expected to drive
typically is not dispatchable or easily control- growth in DG in the coming decades. Sixteen
lable. These units present the greatest challenge states and the District of Columbia currently
and are the primary focus of this chapter. have renewable portfolio standards with specific
DG provisions.7 For example, some states have
Distributed generation can be owned and operated provisions in their renewable portfolio standards
by utilities or their customers and can provide that require some fraction of retail electricity
a variety of theoretical benefits to their owners and sales to come from renewable DG by 2020.
the broader power system. Distributed generation advocates cite a litany of
good things DG can do. Distributed generation
In 2009, about 13,000 commercial and indus- installations theoretically can improve reliability,
trial DG units with a combined capacity of reduce costs, reduce emissions, and improve
about 16 gigawatts (GW) were connected power quality (see Table 5.1).8 However, the
to utility systems in the U.S.2 Of these units, benefits of DG are highly dependent on the
10,800 (83%) were smaller than 1 MW, characteristics of each installation and the
averaging 100 kW each.3 Internal combustion characteristics of the local power system.
1.05 1.05
1.04 1.04
Fedder Voltage (pu)
Fedder Voltage (pu)
1.03 1.03
1.02 1.02
1.01 1.01
1.00 1.00
7:00 AM 11:00 AM 3:00 PM 7:00 PM 7:00 AM 11:00 AM 3:00 PM 7:00 PM
Hour Hour
(a) Without Voltage Regulation Capability (b) With Voltage Regulation Capability
Note: The voltage scales on these plots are in a normalized measure called per-unit (pu). The normalizing constant is the
nominal voltage of the line, 13.8 kV in this case. The line is operating at approximately 1.026 pu, which is 14.2 kV.
Since the original creation of IEEE Standard the required capabilities of DG and necessary
1547, IEEE has supported efforts to create eight operating procedures that can be used to create
additional supplemental standards documents intentional islands,
50 thereby
50 50 partially
50 filling the
intended to extend and/or clarify the provisions previously mentioned gap in IEEE Standard
in the main standard text.25 For example, IEEE 1547. While five of these documents have been
Standard 1547.4, completed in 2011, specifies completed, three additional documents are
Table 5.2 Representative Electric Vehicles Available in the United States by 2012
Source: Tesla Motors Inc., Roadster Features and Specifications, http://www.teslamotors.com/roadster/specs; Nissan Motors Company
Ltd., Nissan Electric Leaf Car: 100% Electric. Zero Gas. Zero Tailpipe, http://www.nissanusa.com/leaf-electric-car/; J. Wiesenfelder,
Cars.Com Field Trial: Mobile EV Quick-Charging, Kicking Tires, July 26, 2011, http://blogs.cars.com/kickingtires/2011/07/carscom-
field-trial-mobile-ev-quick-charging-.html; General Motors Company, 2011 Chevrolet Volt, http://www.gm.com/content/gmcom/
home/vehicles/browseByBrand/baseball_cards/chevrolet/volt.html; General Motors Company, Chevrolet Volts 240V Home Charging
Unit Priced at $490, press release, October 6, 2010, Detroit, MI, http://gm-volt.com/2010/10/06/gm-announces-chevrolet-volt-240v-
charger-pricing-and-installation-service-provider; Toyota Motor Sales, USA, Inc., Toyota Introduces 2012 Prius Plug-in Hybrid, press
release, September 16, 2011, Richmond, CA, http://pressroom.toyota.com/releases/toyota+introduces+2012+prius+plug-in+hybrid.htm.
Note: Battery Size gives the energy storage capacity of the battery in kilowatt-hours. This parameter also provides a relative physical size
of the batteryfor a given battery chemistry, e.g., lithium-ion, the battery size is directly proportional to capacity. Onboard Charger is
the power capability of the charger which is integral to the car. This is the rate at which the battery can be charged by the internal charger.
Quick Charger is the power capability of an external (optional) charger. The quick charger can provide a more rapid charge than the
internal charger, as shown by the Charging Time data.
The impact of EVs on the electrical system EVs will not initially be a concern for every
depends on their market penetration. Several utility. Varying geographic density of electric
past and ongoing studies attempt to estimate vehicles will mean that some utilities or regions
45
PHEVs/EVs on the Road (million)
40
35
30
25
20
15
10
0
2010 2015 2020 2025 2030
Source: Projection data from Committee on Assessment of Resource Needs for Fuel Cell and Hydrogen Technologies and National
Research Council, Transitions to Alternative Transportation TechnologiesPlug-in Hybrid Electric Vehicles (Washington, DC: National
Academies Press, 2010); Daily Compilation of Presidential Documents 2011 DCPD No. 00047, p. 3 (January 25, 2011); and U.S.
Energy Information Administration, Annual Energy Outlook 2011 (Washington, DC:50 50 50of Energy,
U.S. Department 50 2011).
Figure 5.3 Power Requirement of a Single Home in the San Francisco Bay Area with
and without Electric Vehicle Charging
16
Peak Power Requirements (kW)
14
12
10
0
Home Home + Level I Home + Level II
Charging Charging
Source: Data from D. Bowermaster, Plug-in Electric Vehicles and Their Impact: An Integrated, Multi-Stakeholder Approach,
presentation at Environmental Quality Policy Committee Meeting, League of California Cities, Sacramento, CA, January 21,
2011, http://www.cacities.org/resource_files/29491.PGEPEVIntro(2011-01-18).pdf.
EIA
NRC Probable
Obama Target 50 50 50 50 thereby
within utilities may be more severely impacted cluster in particular neighborhoods,
by the presence of EVs than others. Even within increasing concern for the local distribution
regions or utility footprints, only certain system. Early integration problems are likely
hotspots will need significant focus even in to arise most often when local demand rapidly
the medium term. increases because of uneven distribution of
vehicles. Importantly, promotional policies,
If the current geographic distribution of HEVs incentives, and the deployment of the necessary
is a good indicator of demand for PHEVs and infrastructure will strongly influence the
BEVs, distribution systems in California, geographic distribution of EVs.38
2010
Oregon, and Washington likely will 2015
experience 2020 2025 2030
considerably higher penetrations than FINDING
average.3650For example, Southern California Projections of EV penetration nationally are
Edison has projected a mid-case of 5% pene-
45 highly varied. However, some local regions
tration, or 0.5 million PHEVs and BEVs in its
40 are likely to experience penetrations much
service territory by 2020.37 Even citing average
penetration higher than the national average.
35 across a service area may under-
state the challenge, as PHEVs and BEVs may
30
25
20
15
118 MIT STUDY ON THE FUTURE OF THE ELECTRIC GRID
10
BOX 5.4 THE EFFECT OF UNCOORDINATED charging at two rates, 1.4 kilowatts (kW)
CHARGING ON TRANSFORMERS: and 3.3 kW, on a warm summer day. These
transformers are loaded beyond their design
DTE Energy recently conducted a study on the
capacity for both charging rates, and the 25 kVA
impact of plug-in hybrid electric vehicles (PHEV)
transformer exceeds even its peak (short-term)
on distribution system components in its service
capacity rating in both cases. This could lead to
territory. Figure 5.4 shows the impact on both
voltage dips, service interruption, and trans-
a 25 kilovolt-ampere (kVA) transformer and
former failure.
a 50 kVA transformer of three PHEVs
kW
20 30
20
10
10
0 0
12:00 AM 6:00 AM 12:00 PM 6:00 PM 12:00 AM 6:00 AM 12:00 AM 6:00 AM 12:00 PM 6:00 PM 12:00 AM 6:00 AM
Electrical Vehicle Charging PHEVs. On the other hand, PHEVs are expected
NRC Maximum EIA
to comprise a majority of EVs, and a few PHEVs
EVs are expectedNRC Probable
to charge at one of three Obama Target
charging simultaneously could have an impact
power levels. The Society of Automotive similar to one BEV. Therefore the potential
Engineers has established charging standards impact of these two types of vehicles likely
(in Standard J1772) that cover the following will be similar.
two charging power levels: 50 50 50 50
A study in 2008 estimated that if each North
s ,EVEL ) UP TO K7 American Electric Reliability Corporation
(NERC) region were to have a 25% penetration
s ,EVEL )) UP TO K7 of PHEVs in the year 2030, each area would
require less than a 5.5% increase in genera-
Level III has not yet been standardized in the U.S., tion.39 Aggregate power requirements are also
but will enable full BEV charging within minutes. unlikely to require significant upgrades to the
bulk power system. If 25% of the national fleet
At the residential level, the majority of PHEVs were PHEVs, the power requirements could be
are expected to charge using Level I chargers, up to 30% of generation capacity if simultane-
while BEVs are expected to charge at Level II. ously charged at 6 kW.40 However, this increase
Charging BEVs will have more impact on the is unlikely to materialize as there will be
distribution system due to their higher-power temporal diversity in the time of arrival at
charging and higher energy capacity than home and most charging will probably be at
BOX 5.5 VEHICLE-TO-GRID OPERATIONS sufficiently charged for driving the carand are
Some observers have suggested that the flow unlikely to achieve widespread deployment in
of energy between the power system and EVs the short term.
could be bidirectional.45 This concept is most Beyond the technical challenges, the economic
often referred to as vehicle-to-grid operation. incentives for V2G operation also appear weak.
While most often discussed in the context of In those markets with a regulation product, the
vehicles providing frequency regulation price paid to participants for regulation services
services, in theory, energy stored in vehicles' has historically been relatively low. The partici-
batteries could provide various types of operat- pation of EVs in these markets would likely
ing reserves. In regions with organized whole- cause the prices to decline further.
sale markets, it has been envisioned that
vehicles could participate in frequency regula- A more cost-effective alternative use of EVs
tion or other reserve markets by supplying would be to provide regulation or operating
reserves only through control of their (unidirec-
energy to the grid.
tional) charging ratefor example, decreasing
V2G operations would require substantial and their rate to provide up regulation, and increas-
expensive modifications to conventional ing it for down regulation. While still requiring
unidirectional vehicle chargers and controls. communication between vehicles and the
V2G concepts also face other substantial utility, the charger requirements would be
technical challengesdegradation of battery much simplified. This mode of operation would
MJGF
0&.XBSSBOUZJTTVFT
DPNQMFYJUZBOE also have a much more limited impact on
expense of added controls and communication vehicle battery life. Particularly attractive is the
with the utility, and the relatively small amount use for this purpose of commercial EV fleets
of energy involved if the battery is always to be which have deterministic charging patterns.
7
Database of State Incentives for Renewables the Impact of Distributed Generation on Power
& Efficiency, http://www.dsireusa.org/. Systems, presented at IEEE Power Engineering
Society Summer Meeting, Seattle, WA, July 1620,
8
U.S. Department of Energy, The Potential Benefits 2000.
of Distributed Generation and Rate-Related Issues
that May Impede Their Expansion: A Study Pursuant U. N. Khan, Distributed Generation and Power
22
to Section 1817 of the Energy Policy Act of 2005 Quality presented at the International Conference
(Washington, DC, 2007); and N Hatziargyriou, on Environment and Electrical Engineering,
H. Asano, R. Iravani, and C. Marnay, Microgrids: Karpacz, Poland, May 1013, 2009.
An Overview of Ongoing Research, Development, Barker and De Mello, see note 21 above.
23
Energy Alternatives to Electrical Distribution Grid Cells, Photovoltaics, Dispersed Generation, and
Expansion in Southern California Edison Service Energy Storage (SCC21), 1547 Series of
Territory (Oak Ridge, TN: Oak Ridge National Interconnection Standards, http://grouper.ieee.
Laboratory, 2005). org/groups/scc21/dr_shared/.
11
U.S. Department of Energy, see note 8 above. R. Lasseter et al., Integration of Distributed Energy:
26
12
J. Bluestein, Environmental Benefits of Distributed The CERTS MicroGrid Concept, LBNL-50829
Generation (Fairfax, VA: Energy and Environ- (Berkeley, CA: Lawrence Berkeley National
mental Analysis, Inc., 2000). Laboratory, 2002); and C. Marnay, F. J. Robio, and
A. S. Siddiqui, Shape of the Microgrid, presented
13
U.S. Department of Energy, see note 8 above. at IEEE Power Engineering Society Winter Meeting,
14
W. P. Poore et al., Connecting Distributed Energy Columbus, OH, January 28February 1, 2001.
Resources to the Grid: Their Benefits to the DER R. Lasseter and J. Eto, Value and Technology
27
Owner/Customer, Other Customers, the Utility, Assessment to Enhance the Business Case for the
and Society (Oak Ridge, TN: Oak Ridge National CERTS Microgrid (Madison, WI: University of
Laboratory, 2002), http://www.ornl. WisconsinMadison, 2010).
gov/~webworks/cppr/y2002/rpt/112701.pdf.
Microgrid Deployment Tracker (Boulder, CO: Annual Energy Outlook 2011 (Washington, DC:
Pike Research, 2011). U.S. Department of Energy, 2011.
R. Lasseter and J. Eto, see note 27 above.
29 36
November 2008 Dashboard: Sales Go From Bad
to Worse, HybridCars.com, December 10, 2008,
Barker and De Mello, see note 21 above; and
30
http://hybridcars.com/hybrid-market-dashboard/
N. Hadjsaid, J. F. Canard, and F. Dumas, Dispersed november-2008-dashboard-25328.html; and
Generation Impact on Distribution Networks, December 2008 Dashboard: The Key Is
IEEE Computer Applications in Power 12, no. 2 Production Numbers, HybridCars.com,
(1999): 2228. January 13, 2009, http://hybridcars.com/hybrid-
M. Scheepers et al., Regulatory Improvements for
31 sales-dashboard/december-2008-dashboard-
Effective Integration of Distributed Generation into focus-production-numbers-25416.html.
Electricity Distribution Networks: Summary of the D. Kim, Director of Plug-in Electric Vehicle
37
DG-GRID Project Results ECN-E--07-083 (DG- Readiness, Southern California Edison, conversa-
GRID Consortium, 2007). tion with the authors, March 22, 2010.
Electric Power Research Institute, National
32
U.S. Department of Energy, Alternative Fuels &
38
Resources Defense Council, and Charles Clark Advanced Vehicles Data Center, http://www.afdc.
Group, Environmental Assessment of Plug-in Hybrid energy.gov/afdc/laws/2010; and T. Woody and
Electric Vehicles Volume 1: Nationwide Greenhouse C. Krauss, Cities Prepare for Life with the Electric
Gas Emissions (Palo Alto, CA: Electric Power Car, The New York Times, February 15, 2010.
Research Institute, 2007); and Electrification
Coalition, Electrification Roadmap: Revolutionizing Hadley and Tsvetkova, see note 33 above.
39
2009); Deutsche Bank Securities Inc., Electric Cars: Required Solutions presented at The Networked
Plugged In (2008); Committee on Assessment of EV Conference, San Francisco, CA, November 9,
Resource Needs for Fuel Cell and Hydrogen 2010.
Technologies, National Research Council,
Transitions to Alternative Transportation D. Bowermaster, Plug-in Electric Vehicles and
43
To V2G, Or Not to V2G? That is the Question! Ilic, The Impact of Controlled EV Charging on
Intelligent Utility, February 17, 2010, http://www. Residential Low Voltage Networks presented at
intelligentutility.com/article/10/02/v2g-or-not- IEEE International Conference on Networking,
v2g-question; P. Carson, V2G: Weve Got the Sensing and Control, Delft, the Netherlands,
Whole Package, Intelligent Utility, February 24, April 1113, 2011.
2010, http://www.intelligentutility.com/
article/10/02/v2g-we-ve-got-whole-package; and
W. Kempton and J. Tomic, Vehicle-to-Grid Power
Fundamentals: Calculating Capacity and Net
Revenue, Journal of Power Sources 144, no. 1
(2005): 268-279.
46
M. C. W. Kintner-Meyer, K. P. Schneider and
R. G. Pratt, Impacts Assessment of Plug-in Hybrid
Vehicles on Electric Utilities and Regional U.S. Power
Grids: Part 1: Technical Analysis (Richland, WA:
Pacific Northwest National Laboratory, 2007).
The distribution system is the portion of the s DETECTS FAULTS AND OTHER ABNORMAL SITUATIONS
electric power system that carries power the on the distribution system and takes action to
few miles remaining between transmission protect people and system components; and
substations and consumers. In addition to
transporting electricity, the distribution system s RESTORES SERVICE FOLLOWING INTERRUPTIONS
s TRANSFORMS VOLTAGE TO THE APPROPRIATE LEVEL Today, less than 7% of total electrical energy
for customer use; in the U.S. is lost between generators and
customers, and most of this loss occurs in the
s REGULATES VOLTAGES WITHIN A LIMITED RANGE1 distribution system.2 2ELIABILITY FOR INDIVIDUAL
CUSTOMERS VARIES WIDELY BUT AVERAGES IN
i
0ERCENT OF INTERRUPTIONS ARE CALCULATED USING 3!)$) SYSTEM AVERAGE INTERRUPTION DURATION INDEX VALUES
3!)$) A RELIABILITY INDICATOR COMMONLY USED BY ELECTRIC UTILITIES IS THE AVERAGE OUTAGE DURATION FOR EACH
CUSTOMER SERVED USUALLY MEASURED IN MINUTES PER YEAR 3EE THE GLOSSARY FOR DETAILS
$ISTRIBUTION SYSTEMS TODAY ARE A PRODUCT OF with work crews. Electronic maps are only
this history. In many areas, todays distribution one example of the changes that upgrades
companies still employ essentially the same IN DISTRIBUTION SOFTWARE CAN BRING (OWEVER IN
process for locating and managing outages an industry that prizes reliability and depends
AS DECADES AGO )N OTHER AREAS SYSTEMS HAVE on tested operating procedures to maintain
EVOLVEDBUT NOT ALWAYS IN THE SAME DIRECTION reliability, such fundamental changes come at
resulting in heterogeneous distribution systems THE COST OF EXTENSIVE TRAINING PROGRAMS AND A
around the country. This heterogeneity reflects period of heightened risk as personnel become
the general lack of standardization as well as accustomed to new practices.
differences in geography and regulatory prac-
tices across state boundaries. 4HE TIME MONEY AND EXPERTISE REQUIRED TO
IMPLEMENT A NEW $-3 CAN VARY GREATLY
2ECENT ADVANCES IN INFORMATION AND COMMUNI- BETWEEN UTILITIES BECAUSE A SIGNIlCANT LEVEL OF
CATION TECHNOLOGIES HAVE ENABLED THE DEVELOP- CUSTOMIZATION IS NECESSARY ! VARIETY OF EQUIP-
ment of significantly more sophisticated ment throughout the distribution system can
DISTRIBUTION MANAGEMENT SYSTEMS $-3 THAT BE ENABLED TO INTERACT WITH $-3S -ANY NEW
CAN HANDLE AND FULLY INTEGRATE A WIDE VARIETY OF SENSORS PROTECTION EQUIPMENT AND CONTROL
SYSTEM MANAGEMENT TASKS "EYOND PROVIDING DEVICES HAVE COMMUNICATION AND COMPUTATION
REAL
TIME VISIBILITY INTO THE STATE OF ASSETS capabilities. The number of sensors that a
throughout the distribution system, state-of- UTILITY DEPLOYS AND INTEGRATES WITH ITS $-3
THE
ART $-3S OFTEN CAN SIMULATE DISTRIBUTION IMPACTS THE COST AND TIME REQUIRED FOR IMPLE-
system power flows in near-real time, helping mentation and the usefulness of the resulting
operators anticipate or respond to potential $-3 4O UPGRADE ITS $-3 FROM A SOFTWARE
PROBLEMS &OR EXAMPLE THESE SYSTEMS CAN ALLOW SYSTEM DEVELOPED IN
HOUSE TO A STATE
OF
THE
ART
operators to determine whether changing the VENDOR SOLUTION ONE #ALIFORNIA UTILITY ATTRIB-
configuration of the distribution network in UTED OF THE ESTIMATED IMPLEMENTATION COST
response to faults will result in a system with TO LABOR )4 AND EQUIPMENT UPGRADES AND
APPROPRIATE VOLTAGE LEVELS AND CURRENT mOWS OF THE COST TO SOFTWARE VENDOR SERVICES9
In conjunction with other technologies,
MODERN $-3S PROMISE TO MAKE DISTRIBUTION
SYSTEM OPERATORS MORE AGILE AND RESPONSIVE
to real-time system conditions.
To date, self-healing technology has been Operating at the upper end of the allowable
installed on only a small percentage of distribu- VOLTAGE RANGE IS A COMMON STRATEGY BECAUSE
tion feeders in the U.S.14 Typically, the least VOLTAGE IS MEASURED AT THE SUBSTATION BUT NOT AT
reliable circuits in distribution systems are first THE END OF THE LINE TO ENSURE THE VOLTAGE AT THE
TO BE UPGRADED BECAUSE IMPROVING THEM WILL BE end of the line is within limits, utilities often set
of the most benefit to customers and the cost of THE SUBSTATION VOLTAGE AT THE UPPER END OF THE
deployment discourages system-wide upgrades. specification. Because load will draw more
Many distribution utilities are now demon- POWER WITH A HIGHER VOLTAGE TIGHTER CONTROL OF
strating this technology through projects VOLTAGE CAN RESULT IN REDUCED POWER CONSUMP-
PARTIALLY FUNDED BY !MERICAN 2ECOVERY AND tion.iii 4IGHTER VOLTAGE LIMITS AND ITS BENElTS CAN
2EINVESTMENT !CT GRANTSii be effected by a more sophisticated approach to
VOLTAGE REGULATION CALLED hVOLT6!2 CONTROLv
Voltage and Power Flow Optimization WHICH EMPLOYS VOLTAGE SENSORS ON THE LINE AND
particularly at its end, which feed back the
As customers draw power from the distribution MEASURED VOLTAGES TO THE SUBSTATION #ONTROL
SYSTEM VOLTAGES DECLINE ALONG DISTRIBUTION EQUIPMENT AT THE SUBSTATION THEN ADJUSTS THE
LINES 7ITHOUT INTERVENTION THE VOLTAGE LEVEL AT SUBSTATION VOLTAGE AND THE VOLTAGE REGULATING
the end of a long distribution line will be lower EQUIPMENT ON THE LINE TO MAINTAIN THE LINE
than at the substation. With some exceptions, VOLTAGE AT THE LOW END OF ITS LIMITS
THE SIZE OF THE VOLTAGE DROP INCREASES WHEN
LOADS INCREASE 4O CONTROL VOLTAGE AT THE &IGURE ILLUSTRATES THE POTENTIAL EFFECT OF VOLT
CUSTOMER LOAD POINT UTILITIES HAVE LONG USED 6!2 CONTROL ON THE VOLTAGE ALONG A DISTRIBUTION
ADJUSTABLE TRANSFORMERS KNOWN AS hVOLTAGE FEEDER )N THE lGURE THE VERTICAL AXIS IS VOLTAGE
REGULATORSv AND hCAPACITOR BANKSvA COLLECTION LEVEL AND THE HORIZONTAL AXIS FROM LEFT TO RIGHT
of capacitors that can be switched in and out of represents the distance of a distribution line
THE CIRCUITTO KEEP VOLTAGES WITHIN A SPECIlED from the substation to the end of the line.
range along the entire length of distribution 4HE BLUE LINES SHOW THE VOLTAGE DISTRIBUTION
LINES 4HESE DEVICES ARE CONTROLLED FROM THE UNDER NORMAL OPERATION WHERE VOLTAGE
distribution substation, which ensures that measurements are not fed back to the substa-
THE VOLTAGE FOR EACH CUSTOMER IS BETWEEN TION 4HE VOLTAGE VARIATION ALONG THE LINE IS SEEN
AND OF RATED VOLTAGE15 &OR RESIDENTIAL TO BE 4HE RED LINES REPRESENT THE VOLTAGE
CUSTOMERS WHERE THE RATED VOLTAGE IS VOLTS DISTRIBUTION USING VOLT6!2 CONTROL 4HE EFFECT
THIS GIVES AN ACCEPTABLE RANGE OF VOLTS of sensing and feedback control is not only to
ii
4RACK THESE PROJECTS AT HTTPWWWSMARTGRIDGOV
iii
4HIS RELATIONSHIP DEPENDS ON THE NATURE OF THE LOAD 4HE CORRELATION IS VERY STRONG FOR RESISTIVE LOADS
SUCH AS INCANDESCENT LAMPS BUT WEAK FOR MANY ELECTRONIC DEVICES WITH MODERN POWER SUPPLIES THAT
COMPENSATE FOR VOLTAGE VARIATIONS
Voltage (% rated)
104%
98%
Substation ------------- Voltage Regulator 1 ------------- Voltage Regulator 2 --------------> End of Line
REDUCE THE AVERAGE VALUE OF THE VOLTAGE WHILE THEIR ADVANCED METERING INFRASTRUCTURE AND
maintaining it within its acceptable limits, but IS ESTIMATED TO SAVE THEIR CUSTOMERS BETWEEN
ALSO TO TIGHTEN ITS VARIATION ALONG THE LINE TO LESS 1% and 4% of energy consumption without
than 4%. ANY BEHAVIOR CHANGES18
*These projects include upgrades to natural gas metering systems in addition to electricity metering upgrades. The number of meters for these projects corresponds
to both electricity and gas meters.
.OTE 4HE DATA IN 4ABLE HAS BEEN EXTRACTED FROM STATE REGULATORY PROCEEDINGS 0ROSPECTIVE CALCULATIONS OF !-) BENElTS AND COSTS REQUIRE A RANGE OF ASSUMPTIONS
EG PROJECT LIFETIMES DEPRECIATION SCHEDULES !LSO AS DESCRIBED IN THE TEXT THESE RESULTS ARE SUBJECT TO SIGNIlCANT UNCERTAINTY EG TECHNOLOGY COSTS DEPLOYMENT
SCHEDULES DYNAMIC PRICING ADOPTION RATES ETC 4HE COST OF !-) PROJECTS ALSO DEPEND ON THE CHARACTERISTICS OF EACH UTILITYS SERVICE TERRITORY AND THE NATURE OF THEIR
existing technology and business processes related to customer metering.
Outlook 2009 7ASHINGTON $# 53 $EPARTMENT 0ERKINS *R ON "EHALF OF /KLAHOMA 'AS AND
OF %NERGY SECTION HTTPWWWEIADOE Electric Company, testimony before the
GOVEMEUAERELECTHTML Corporation Commission of the State of
/KLAHOMA -ARCH
3
* ( %TO AND + ( ,A#OMMARE Tracking the
Reliability of the U.S. Electric Power System: 0ERSONAL COMMUNICATION WITH $ -ALKIN
14
An Assessment of Publicly Available Information * -C$ONALD AND " 7OJSZCZYK 'ENERAL %LECTRIC
Reported to State Public Utility Commissions 3EPTEMBER
"ERKELEY #! ,AWRENCE "ERKELEY .ATIONAL .ATIONAL %LECTRICAL -ANUFACTURERS !SSOCIATION AND
15
(OBOKEN .* *OHN 7ILEY AND 3ONS -EASURES THROUGH $ISTRIBUTION %FlCIENCYv
5
0ERSONAL COMMUNICATION WITH 2OGER ,EE -ANAGER Metering International n
OF !SSET -ANAGEMENT 3YSTEM 2ELIABILITY 0ERKINS SEE NOTE ABOVE
17
Reliability Indices Tracking within the United States $IGITAL -ETERSv WHITE PAPER 0ALO !LTO #!
0ALO !LTO #! HTTPWWWSDGECOMDOCUMENTSSMARTMETER
%02)-ETER!CCURACYPDF
7
%TO AND ,A#OMMARE SEE NOTE ABOVE
Ibid.
8
Council of European Energy Regulators, 4th
Benchmarking Report on Quality of Electricity U.S. Energy Information Administration, Annual
21
$-3/-3 )NCREASING $ISTRIBUTION /PERATIONS %LECTRIC 0OWER )NDUSTRY 2EPORTv &ORM %)!
Efficiency, Electric Energy T&D Magazine 13, no. 2 7ASHINGTON $# 53 $EPARTMENT OF %NERGY
n HTTPWWWELECTRICENERGYONLINE
COMPAGESHOW?ARTICLEMAGARTICLE &EDERAL %NERGY 2EGULATORY #OMMISSION SEE
24
11
- * 3ULLIVAN - -ERCURIO AND * 3CHELLENBERG NOTE ABOVE
%STIMATED 6ALUE OF 3ERVICE 2ELIABILITY FOR %LECTRIC 53 $EPARTMENT OF %NERGY h2ECOVERY !CT
25
Utility Customers in the United States (Berkeley, 3ELECTIONS FOR 3MART 'RID )NVESTMENT 'RANT
#! ,AWRENCE "ERKELEY .ATIONAL ,ABORATORY !WARDSv HTTPENERGYGOVOEDOWNLOADS
RECOVERY
ACT
SELECTIONS
SMART
GRID
INVESTMENT
grant-awards-category.
Assessment of Demand Response and Advanced Investment in the Current World of Competitive
Metering Staff Report 7ASHINGTON $# Regulation 7ASHINGTON $# %DISON %LECTRIC
)NSTITUTE HTTPWWWEEIORGWHATWEDO
! &ARUQUI ET AL h4HE #OSTS AND "ENElTS OF 3MART
27
0UBLIC0OLICY!DVOCACY3TATE2EGULATION AND
-ETERS &OR 2ESIDENTIAL #USTOMERSv WHITE PAPER 4 ,YON h2EGULATORY (INDSIGHT 2EVIEW AND
7ASHINGTON $# %DISON &OUNDATION )NSTITUTE )NNOVATION BY %LECTRIC 5TILITIESv Journal of
FOR %LECTRIC %FlCIENCY *ULY AND $ * ,EEDS Regulatory Economics n
U.S. Smart Grid Market Forecast n '4-
2ESEARCH
! &ARUQUI AND , 7OOD h)%% 2ELEASES 4HE
28
since electricity on average accounts for only 7.1 WHY ENGAGE DEMAND?
2% to 3% of household expenditures.3 It also
reflects concerns about customer reactions As we have discussed in earlier chapters,
to AMI technology, volatile electric bills electricity systems must precisely balance
(particularly unexpectedly high bills), and supply and demand at each moment in time.
the differential impact of dynamic pricing on System operators historically have ensured the
low-income customers. This chapter discusses equality of supply and demand largely through
the nature and impact of current demand supply-side tools: operators adjust the dispatch
response programs and assesses the evidence schedule for generators to meet forecast
on the potential for expanded demand engage- demand, adjust dispatch as forecasts are
ment, particularly by residential customers. updated, and use ancillary services to adjust to
real-time deviations in demand from forecast.
The discussion highlights an important ques- Predictable variation in demand over time, day,
tion for regulators and utilities: what are the and season is associated with changes in the
trade-offs between rapid universal rollout of marginal cost of supplying electricity that
AMI technology and a staged deployment that regularly vary by a factor of two or three.
i
The 2015 projections may be aggressive given that AMI deployments frequently have been delayed or
deferred.4
Figure 7.1 2010 Average Daily versus Selected Hourly Average Real-Time Locational
Marginal Price (RT LMP) in the PJM Interconnection
Hourly RT LMP Jul 7 and Aug 11
400
200 350
300
250
180 200
150
160 100
50
0
140 1 3 5 7 9 11 13 15 17 19 21 23 24
Hour
LMP ($ / MWh)
120
100
80
60
40
20
0
1Jan
15Jan
29Jan
12Feb
26Feb
12Mar
26Mar
9Apr
23Apr
7May
21May
4Jun
18Jun
2Jul
16Jul
30Jul
13Aug
27Aug
10Sep
24Sep
8Oct
22Oct
5Nov
19Nov
3Dec
17Dec
31Dec
Date
Data Source: PJM, Monthly Locational Marginal Pricing, http://pjm.com/markets-and-operations/
energy/real-time/monthlylmp.aspx.
50 50 50 50
1 2 3 4 5 6
Standard Billing Enhanced Billing Estimated Daily/Weekly Real-time Real-time Plus
(e.g., monthly, (e.g., info and Feedback Feedback Feedback (e.g., home-area
bi-monthly) advice, household (e.g., web-based (e.g., based on (e.g., in-home networks, appliance
specific energy audits and consumption displays, pricing, disaggregation
or otherwise) billing analysis, measurements, signal capability) and/or control)
est. appliance by mail, email,
disaggregation) self-meter)
Information Availability
Low High
Cost to Implement
Source: Electric Power Research Institute, Residential Electricity Use Feedback: A Research Synthesis and
Economic Framework (Palo Alto, CA, 2009).
ii
Southern California Edison recently entered a contract with the energy management company EnerNOC,
for example, to support AMI-enabled commercial and institutional customer access to EfficiencySMART
Commissioning software. The utility promises participating sites 10% or more savings on their total
addressable energy expenditures. 12
20,000
15,000
10,000
5,000
0
Commercial Residential Wholesale Other
& Industrial
2006 FERC Survey 2008 FERC Survey 2010 FERC Survey
Source: Federal Energy Regulatory Commission, 2006 Assessment of Demand Response and Advanced
Metering Staff Report (Washington, DC, 2006); Federal Energy Regulatory Commission, 2008
Assessment of Demand Response and Advanced Metering Staff Report (Washington, DC, 2008); Federal
Energy Regulatory Commission, 2010 Assessment of Demand Response and Advanced Metering Staff
Report (Washington, DC, 2011). 50 50 50 50
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Measuring demand response potential is predictability of Measuring demand response
challenging. FERCs potential peak load demand response potential is challenging.
1926
1930
1934
1938
1942
1946
1950
1954
1958
1962
1966
1970
1974
1978
1982
1986
1990
1994
1998
2002
2006
reduction is a better indicator of aggregate resources.13 At present,
load enrolled in a demand response program NERC is focused in
than of expected reductions in system demand Phases I and II only on dispatchable,
at peak periods. Nevertheless, it currently controllable resources that system operators
provides the greatest comparability over time can direct to respond to reliability-driven
and across systems.iii There is some promise of events. In Phases III and IV, this will be
progress on this front. The North American extended to nondispatchable, reactive (defined
4.5
Electric Reliability Council (NERC) is to include price-mediated) demand response
developing 4.0
a Demand Response Availability resources. NERCs goal is to develop
Data System to codify reporting of demand performance-based data that will enhance
3.5
response enrollment and collect information industry confidence in the measurement and
on their actual
3.0 responses to system events in use of demand response resources to meet one
an effort to improve data accuracy and or more core objectives.
2.5
2.0
iii
Enrolled potential peak load reduction as measured by the FERC survey does not necessarily translate
1.5
to actual, expected, or available load reduction at any given time for a number of reasons. For example,
customers may not be drawing their full enrolled load at the time of a demand response event; local
transmission1.0congestion may render demand response from customers in uncongested areas irrelevant;
or restrictions in the program may explicitly or implicitly limit the frequency or duration of system operator
0.5 customer to reduce their load. Finally, voluntary demand response programs typically
calls on a given
provide less predictable and lower overall responses per MW of enrolled load than do direct load control or
0.0
mandatory response programs.14 FERC reports that aggregate actual demand response averages less than
one-third of potential peak reduction, with substantial variation in that figure across regions and years.15
More recent programs have begun to focus on The designs of demand response programs
reducing consumption in non-crisis peak used to accomplish these objectives are varied.
periods and smoothing demand over time. As They can be broadly categorized into load
noted in Chapter 1, in the capital-intensive control, wholesale marketadministered
electricity industry with no inventory and programs, and customer pricemediated
limited storage capability, generation, transmis- programs, as shown in Figure 7.4.
sion, and distribution capacity that is sized to
meet the maximum expected demand may go Load Control Programs
underutilized or unused in most periods, raising
the long-run average cost of supplying a given Load management or control programs
amount of energy. Solutions to this peak load comprise the largest segment of demand
problem focus on economic incentives, price response programs in place today. Used since
policies, and regulations that allow for an the late 1960s by load-serving entities, vertically
efficient investment in and use of capacity. While integrated utilities, and system operators to
these may complement reliability objectives, maintain reliability, they have more recently
their focus tends to be on long-run efficiencies begun to incorporate broader peak-shaving
rather than short-run operational exigencies. objectives. The immediate and often predict-
The greatest growth in programs targeted at able load reductions many of these programs
peak load management has been in regions that provide are especially valuable in responding to
allow demand-side resources to participate in system emergencies. They can also reduce costs
long-term wholesale capacity markets. by shaving the increasingly sharp peak loads
described in Chapter 1 that are associated with
Finally, there may be opportunities going increased penetration of air conditioning,
forward to use demand-side resources to help falling industrial load shares, and retail prices
balance energy supply and demand in real time, that do not vary with system conditions.17
especially for regions with significant variable While programs differ in operational details
energy resources, such as wind or solar genera- from entity to entity, they generally fall into
tors. As noted in Chapter 3, if loads could be three categories: emergency payment, inter-
quickly and reliably adjusted, demand response ruptible tariffs, and direct load control. Enrolled
might substitute for flexible supply-side regula- customers typically receive a payment or rate
tion services or storage. The value of this respon- credit as an incentive for participation. Load
siveness is likely to increase with penetration of management programs contributed
Interruptible Load
Non-spinning Reserves
Market Administered
Regulation
Other
Source: Federal Energy Regulatory Commission, Assessment of Demand Response and Advanced Metering
Staff Report (Washington, DC, 2011).
iv
These are in the range of maximum wholesale prices for generation in most U.S. markets. For example,
2010 maximum wholesale prices were $396/MWh, $343/MWh, and $1,102/MWh in the Southwest Power
Pool, Midwest, and Electric Reliability Council of Texas regions, respectively.20 Wholesale prices provide
an appropriate benchmark for incentive payments because they reflect the cost of energy inclusive of grid
conditions.
In light of these challenges, at least three s 2EAL
TIME PRICING SCHEMES WITH PRICES THAT
sources of information may help to narrow vary hourly or more frequently could reduce
the uncertainty in the benefits calculus: peak-period electricity consumption substan-
tially under plausible demand elasticity
s 3IMULATION METHODS WHICH ARE USEFUL TO assumptions. In the long run, this reduces
assess the general plausibility of a stylized necessary investment in peaking generating
economic case and determine which param- plants, leading to lower average electricity
eters or assumptions are most critical. costs than would otherwise occur.39
v
Some have expressed concern that this challenge may be so great as to threaten power system stability.
For example, recent work illustrates the potentially destabilizing effect of price-responsive demand in a
model of an ISO that uses only data on loads in past periods to forecast future load and sets price so that
generation equals forecast load, ignoring past prices, the impact of price changes, and its own past forecast
errors.41 This analysis is closely related to the classic cobweb model in economics, which has little
explanatory power in most settings because actors are more sophisticated forecasters than assumed in
the model.42 Forecasting load with reactive demand is difficult but soluble, as suggested by considerable
experience internationally as well as in the U.S. with dispatch under customer-directed responses to
emergency load-management, as well as wholesale price-mediated, customer time-of-use rates, and
dynamic pricing programs.
vi
A recent Institute for Electric Efficiency white paper simulates lifetime costs and benefits of AMI-
facilitated demand response for four prototype utilities.43 This paper probably is best viewed as
providing a framework for analysis rather than simulation results for net returns to AMI investment, as a
number of parameters are varied in ways that do not nest across prototypes, and the focus is on high-level
aggregate returns rather than analysis of sensitivity to parameter values or modeling assumptions.
TOU
TOU w/Tech
PTR
PTR w/Tech
CPP
CPP w/Tech
RTP
RTP w/Tech
50%
40%
Peak Reduction
30%
20%
10%
0%
1 109
Pricing Test
s 4HE QUALITY OF INFORMATION PROVIDED BY PILOTS s &EW PILOTS HAVE ANALYZED THE DISTRIBUTION OF
varies widely. Many pilots enrolled relatively responses or bill impacts across customers,
few customers, limiting the strength of though this may be an important input to
statistical inferences and making it dangerous regulators considering dynamic pricing plans
to generalize the results to the broader (discussed in greater detail in this chapter).
population. Others do not have adequate There is some evidence to show that most
control groups or baseline data that allow reduction in demand comes from a relatively
analysts to measure treatment group responses small number of customers, but we do not
against an untreated control group. know whether those could be identified
in advance and targeted or how persistent
s -OST PILOT STUDIES REPORT RELATIVELY SIMPLE their identity is over time.51
measures of responses, usually mean reduc-
tion in peak-period consumption relative System-wide deployments of AMI
to a baseline or control groups behavior, that combined with optional dynamic
mask a wide variance in the number of hours
pricing tariffs may provide the data
for which data are collected. In addition, there
has been little analysis of the predictability needed to fill in at least some of
of the mean response or the temporal distri- these gaps.
bution of responses by the group facing
dynamic prices, though this information may System-wide Deployments
be important to operate the system and assess
the value of demand response. For example, a System-wide deployments of AMI combined
5% peak reduction on every critical peak day with optional dynamic pricing tariffs may
may have dramatically different implications provide the data needed to fill in at least some
for system operation than a 10% reduction of these gaps. A number of utility systems have
on half the critical peak days and no reduc- recently completed or will soon complete
tion on the remaining (perhaps hotter) days, rollouts of AMI technology to all or most of
though both scenarios imply a 5% mean their customer base as a result of utility initia-
reduction. tives, statewide legislative or regulatory
mandates, or Smart Grid Investment Grant
Early AMI Adopters For the many utilities that have already installed
AMI technology across their systems or are
Many utilities have already made commitments committed to doing so, the largest cost of
to universal AMI deployment throughout their enabling greater demand engagement via
systems; in fact, some have already accom- dynamic pricing has been sunk, particularly
plished that objective. For other utilities, AMI if their information technology systems have
investments may pass costbenefit tests without simultaneously been upgraded to accommo-
requiring any significant demand-side benefit, date the data flow from AMI systems. If
and we would expect those commitments to be dynamic pricing tariffs are offered but not
made. For example, systems that operate mandated, as seems likely at least initially,
manually read, traditional electromechanical default service tariffs should be designed to
meters and those with low customer density, reduce subsidies to flat-rate customers and
high customer turnover or high billing losses encourage efficient evolution toward universal
may find that the operational cost savings from dynamic pricing defaults over time.
reduced meter-reading labor, outage detection,
remote connection and disconnection, and Structuring demand response program imple-
theft reduction largely offset the capital costs of mentation using best-practice pilot design will
universal AMI deployment across their service enhance the ability of researchers to measure
area over an estimated 20-year useful meter programmatic impact. For example, utilities
life.77 These systems may find the economics should collect baseline consumption data
of near-term AMI investments attractive and before exposing consumers to new rate designs
join the utilities that have already committed and may stagger the rollout of new rate options
to AMI. across service areas to provide a synthetic
control group. This structure also can improve
Early adopters confront a number of challenges a utilitys ability to allocate customer service
and responsibilities. Home energy management resources and respond effectively to any
systems, automated residential appliances, and unforeseen problems that develop in
other complementary technologies are relatively implementation.
Management (Boston, MA, 2005). Contract with Southern California Edison, press
2
For example, see A. Faruqui and S. Sergici, release, March 23, 2011.
Household Response to Dynamic Pricing of 13
North American Electric Reliability Corporation,
Electricity: A Survey of 15 Pricing Experiments, Demand Response Availability Data System
Journal of Regulatory Economics, 38, no. 2 (2010): (DADS): Phase I and II Final Report (Washington,
193225. DC, 2011).
3
U.S. Bureau of Labor Statistics, 2009 Consumer 14
S. Neumann, F. Sioshansi, A. Vojdani, and G. Yee,
Expenditure Survey (Washington, DC, 2009). The Missing Link, Public Utilities Fortnightly,
4
D. Leeds, US Smart Grid Forecast 20102015 March 2007, 52.
(San Francisco, CA: GTM Research, 2010), Edison Federal Energy Regulatory Commission, see note 6
15
no. 1 (2010): 720; W. W. Hogan, Providing Holland and Mansur, see note 40 above;
44
Incentives for Efficient Demand Response, Borenstein, see note 39 above.
prepared for the Electric Power Supply Association,
Comments on PJM Demand Response Proposals, S. Borenstein, Time-Varying Retail Electricity
45
Federal Energy Regulatory Commission Docket Prices: Theory and Practice, in Electricity
No. EL09-68-000 (2009). Deregulation: Choices and Challenges, eds. J. Griffin
and S. Puller (Chicago: University of Chicago
33
J. Bushnell, B. F. Hobbs, and F. A. Wolak, When Press, 2005); and S. Borenstein, Customer Risk
It Comes to Demand Response, Is FERC Its Own from Real-Time Retail Electricity Pricing: Bill
Worst Enemy? The Electricity Journal 22, no. 8 Volatility and Hedgability, The Energy Journal 28,
(2009), 918. no. 2 (2007): 111130.
34
Federal Energy Regulatory Commission, see note 6 A. Faruqui, The Tao of the Smart Grid,
46
above. presentation to the Michigan Smart Grid
35
Charles River Associates, see note 1 above. Collaborative, Lansing, MI, August 24, 2011, http://
www.brattle.com/_documents/UploadLibrary/
36
G. Neichin and D. Cheng, 2010 U.S. Smart Grid Upload973.pdf.
Vendor Ecosystem (The Cleantech Group, 2010);
D. Leeds, U.S. Smart Grid Market Forecast 2010 A. Faruqui, R. Hledik, and S. Sergici, Rethinking
47
2015 (San Francisco, CA: GTM Research, 2010); Prices, Public Utilities Fortnightly, January 2010;
and Electric Power Research Institute, Estimating A. Faruqui and S. Sergici, Household Response
the Costs and Benefits of the Smart Grid (Palo Alto, to Dynamic Pricing of Electricity: A Survey of
CA, 2011). Seventeen Pricing Experiments (San Francisco, CA:
The Brattle Group, 2009); and Faruqui and Sergici,
37
Federal Energy Regulatory Commission, see note 6 see note 2 above.
above; and P. Fox-Penner, Surprising Deployment
Developments, IEEE Smart Grid Newsletter, Faruqui, Hledik, and Sergici, see note 47 above,
48
and Benefits of Smart Meters for Residential Evaluation for Pacific Gas & Electric Companys
Customers, white paper (Washington, DC: Edison Residential SmartRate: Ex Post Load Impacts (San
Foundation Institute for Electric Efficiency, 2011); Francisco, CA: 2010).
and S. P. Holland and E. T. Mansur, The Short- A. Cooper and L. Wood, Summary of Rate-Payer
52
Run Effects of Time-Varying Prices in Competitive Funded Electric Efficiency Impacts, Budgets and
Electricity Markets, Energy Journal 27, no. 4 Expenditures, brief (Washington, DC: Edison
(2006), 127155. Foundation Institute for Electric Efficiency, 2011).
The U.S. electric power industry must invest quality; and accommodate new types of
significant amounts of capital over the coming generation, load, and demand response.1 To
decades to replace aging assets and expand the deliver on these promises in the most efficient
network to meet incremental load growth. That and cost-effective manner, the regulatory
investment easily could double if utilities deploy systems and policies that oversee the U.S. grid
new transmission and distribution technologies also must be modernized.
to improve system operation; enhance service
First, distribution services are universally Third, utility oversight generally reflects
treated as a natural monopoly. Even in areas a common set of objectives:
that have restructured their electricity markets,
delivery of electricity from the high-voltage s Operational Efficiency: Utilities should
transmission network to end-use customers deliver the quantity of electricity that
generally is assigned as an exclusive franchise to consumers wish to purchase at the lowest
a government-owned or cooperative enterprise reasonable cost while providing acceptable
or to an IOU subject to some form of price reliability and other aspects of performance.
i
Municipal, cooperative, and other non-IOU systems typically determine rates through different processes,
but they lead to similar outcomes as cost-of-service regulation and involve similar efficiency problems.4
ii
In competitive retail markets, the retailer determines the contractual energy price, which is combined
with the regulated transmission and distribution charges into a single customer bill.
Another challenge posed by the changing nature little or not at all over the hour or the day, leading
of new investments is that some significant to excessive consumption during system peaks
benefits of grid modernization may not be fully and inefficiently discouraging off-peak consump-
internalized by a single utility or its customers.11 tion. Real-time prices that reflect current system
For instance, if a particular distribution utility costs are rare, even among larger customers.
increases the responsiveness of its customers
demand or the efficiency of their energy use and If delivered prices are below incremental costs,
thereby reduces regional generation costs, some as may occur during peak demand periods for
of the benefits may accrue to customers in the customers facing time-invariant rates based on
same region who are not served by that distri- average costs, customers will use inefficiently
bution utility. Similarly, early adopters of new high quantities of electricity at high-cost times.
technologies and systems may end up incurring This can increase energy acquisition costs and
costs that are avoided in later adoptions through require additional investments in both genera-
learning spillovers from early experiences. tion and grid capacity to meet peak demand,
reducing capacity utilization and increasing
These considerations may bias investment average costs (and therefore average rates)
toward the mature technologies and assets that above efficient levels.
are familiar to utilities and regulators.12 Such
iii
Some have suggested that distortions induced by volumetric charges to recover fixed costs and the potentially
greater distortions caused by increasing block pricing may be justified by the presence of imperfections in the
market for electricityfor example, because cost-based prices do not reflect the incremental social cost of
electricity consumption, particularly from dirty sources, or because consumers need higher prices to
overcome their inertia and make desirable efficiency investments.15
While the economic theory of second-best pricing recognizes that marginal cost prices may not be optimal
in the presence of market and decision-making imperfections, there is no theorem suggesting that one
potential distortion merits another. Such decisions must be based on careful analysis and situation-specific
modeling and measurement to ensure improvements to welfare, with a healthy skepticism toward the
benefits of adding layers of distortions.16
iv
Regulators cannot ignore, however, that the main business of retailers is to sell electricity. When energy
efficiency and savings becomes a regulatory priority, energy service companies, whose only purpose is to
reduce consumption or to shift it to low cost periods of time, are best suited for this job.
No Requirements
QoS Penalty
QoS Penalty & Incentive
QoS Target
ROE-based PBR
Reporting Only
ROE-based PBR: 2
No Requirements: 12 QoS Penalty & Incentive: 3
QoS Penalty: 11
Definitions
Return-on-equity-based Performance-based Regulation (ROE-based PBR)
rate of return is set with a dead band (range the utility and shareholders assume
all benefits and cost) and a live band (range above and below the dead band
that would have a sharing mechanism assigned)
QoS Target: 11
Quality of Service (QoS) rate of return is set by using the conventional cost
of service methodology and the utility has reliability and/or quality of service Reporting Only: 12
targets set by the commission with either a penalty & reward incentive,
penalty only, or target only (neither penalty nor reward)
Reporting Only utility is required to report performance with no penalty, Number of States in Each Category
reward or target
Source: Davies Consulting, Inc., State of Distribution Reliability Regulation in the United States (Washington, DC: Edison Electric
Institute, 2005), Table 4-1.
v
One suggestive piece of evidence is that cooperatives have deployed AMI to almost 25% of their customers,
and political subdivisions (public power districts, public utility districts, and the like) have deployed it
to more than 20%. In contrast, only 6.6% of IOU customers have AMI, and penetration is even lower
for municipal entities (3.6%) and federal and state utilities (0.7%).36 Clearly these dramatic differences
reflect more than governance. In particular, cooperatives and political subdivisions tend to have more
geographically dispersed customers, which implies larger benefits from AMI communication capabilities.
But the differences are so large that it is hard to believe that governance is not part of the story.
Utility rate structures will assume increased In the past, recovering costs through volumetric
importance going forward, particularly as charges for smaller customers may have been
potentially large investments in transmission expedient: network charges were not a large
and distribution systems cause rates to increase fraction of the total electricity bill, and
and threaten the current political equilibrium. metering costs to compute appropriate demand
As noted earlier in the chapter, transmission charges were relatively high. Less costly alterna-
and distribution costs are largely independent tives used outside the U.S., such as limiters that
of delivered energy in the short term, but the prevent consumption levels above a contracted
recovery of those costs from residential and peak amount, had little appeal to U.S. regula-
small commercial customers generally is heavily tors. As network charges now approach parity
dependent on energy sales. Whenever prices are with energy costs for many customers, this
out of line with costs, behavior is distorted and structure demands closer attention. And as
economic efficiency is reduced.vi AMI diffuses, it becomes easier and less expen-
sive to vary fixed charges across customers in
FINDING ways that reflect their differential impacts on
Recovery of largely fixed network costs the need for network development.
through volumetric ($/kWh) charges distorts
AMI meters can record consumption that is
the behavior of utilities, their customers,
coincident with system peaks or measure
and investors in distributed generation average consumption over a set of designated
and may be politically unstable in an era peak hours. These facilitate a transition to
of rising costs. fixed monthly delivery charges that vary with
either usage in local peak periods or another
measure of impact on the need for network
Reducing reliance on energy sales for transmis- capacity. Because peak usage tends to be
sion and distribution network cost recovery correlated with total usage across households,
addresses these problems directly. Regulators these charges will lead to larger customers
can accomplish this by increasing fixed generally paying a greater fraction of system
customer charges or demand charges (charges costs than do smaller customers, as they do
that depend on peak kW or capacity use, not under volumetric pricing, but without
vi
In principle this problem also arises with regard to transmission costs in most areas, and both
transmission and distribution charges should be addressed similarly. However, because transmission costs
are considerably lower than distribution costs, the distortions caused by recovering fixed transmission
costs via volumetric charges may be less severe.
vii
An ideal approach would start with energy charges based on locational marginal prices at each
distribution node and point in time, and use customer fixed charges to recover more of the remaining
network costs. But current systems are far from measuring, let alone implementing that ideal.
Revenue Decoupling
Mechanism
Lost Revenue
Adjustment Mechanism
Pending
Source: Institute for Electric Efficiency, State Electric Efficiency Regulatory Frameworks (Washington, DC, 2011).
Reliance on volumetric cost recovery may be distortions may arise for investments in energy
particularly problematic when installations efficiency or conservation under volumetric
of distributed generation are subject to net rate structures.
metering. Under net metering, a kWh of
generated electricity reduces a customers bill Reliance on volumetric cost recovery may be
not only by the avoided energy cost but also by particularly problematic when installations of
the amount of the delivery service charge. The
distributed generation are subject to net metering.
greater the quantity of electricity generated
onsite, the less a customer contributes to cover
the grids cost. As discussed earlier, this implicit 8.4 CONCLUSIONS AND
subsidy to distributed generation may reduce RECOMMENDATIONS
utility revenue in the short run if rates are not
decoupled and will shift the burden of network Upward pressure on electricity rates will
costs to customers without (clean or dirty) increase the value of more efficient transmis-
distributed generation. This problem is more sion and distribution operations, and customer
serious, all else equal, the greater the fraction concerns with reliability and other dimensions
of network costs that are recovered through of performance are likely to increase over time.
volumetric charges and the less correlated a Collecting and publishing comparable data
customers peak net demand is with their total on utilities costs and service quality can help
(gross) consumption of electricity. Similar regulators evaluate and reward good, efficient
performance.
Traditional utility regulation has focused more Fixed charges should vary with the extent to
on curbing monopoly power and avoiding which customers contribute to the need for
excessive costs than on encouraging innovation. network development. This need might be
This emphasis is becoming increasingly expen- approximated by past demand in peak periods
sive in an environment with rapid technological or estimated by demand profiles. In systems
change and consequent potential for significant that continue to rely significantly on volumetric
efficiency gains. charges for the recovery of network costs,
utility incentives to increase sales can be
R E CO M M E N D AT I O N blunted by decoupling utility revenues from
State regulators and others in charge short-run changes in quantities sold.
of setting utility rates should design
mechanisms for risk allocation and
compensation to balance incentives
for innovative, risky investment with
efficiency gains and ensure that the
results of innovative investments are
shared with customers.
Estimate of the Investment Requirements and the Python: Is Lumpy Capacity Investment Used and
Resultant Benefits of a Fully Functioning Smart Grid, Useful? Energy Law Journal 23, no. 2 (2002): 383;
technical report 1022519 (Palo Alto, CA, 2011). and F. Graves and A. Baker, Disincentives to Utility
2
J.-J. Laffont and J. Tirole, A Theory of Incentives in Investment in the Current World of Competitive
Regulation and Procurement (Cambridge, MA: Regulation (Washington, DC: Edison Electric
MIT Press, 1993); and P. L. Joskow, Regulation Institute, 2004).
of Natural Monopoly, in Handbook of Law and 10
Colorado Public Utility Commission, In the Matter
Economics, ed. A. M. Polinsky and S. Shavell, vol. 2 of the Application of Public Service Company of
(Amsterdam: North-Holland, 2007), 12271348. Colorado for an Order Approving a SmartGridCity
3
V. Sakhrani and J. E. Parsons, Electricity Network CPCN: Order on Exceptions, Decision No.
Tariff Architectures: A Comparison of Four OECD C11-0139, Docket No. 10A-124E, January 5, 2011.
Countries, working paper WP-2010-008 11
U.S. National Research Council, Americas Energy
(Cambridge, MA: MIT Center for Energy and Future: Technology and Transformation
Environmental Policy Research 2010); National (Washington, DC: National Academies Press,
Regulatory Research Institute, The Electric Industry 2009).
at a Glance (Washington, DC, 2008); and J. C.
Bonbright, A. L. Danielsen, and D. R. Kamerschen,
12
Electricity Advisory Committee, Smart Grid:
Principles of Public Utility Rates, 2nd ed. (Arlington, Enabler of the New Energy Economy (Washington,
VA: Public Utilities Reports, 1988), 700. DC: U.S Department of Energy, 2008).
4
Joskow, see note 2 above; and W. L. Megginson
13
P. L. Joskow, The Difficult Transition to
and J. M. Netter, From State to Market: A Survey Competitive Electricity Markets in the United
of Empirical Studies on Privatization, Journal of States, in Electricity Deregulation: Choices and
Economic Literature 39, no. 2 (2001): 321389. Challenges, ed. J. Griffin and S. Puller (Chicago, IL:
University of Chicago Press, 2005).
5
U.S. Bureau of Labor Statistics, Consumer Price
Index, http://www.bls.gov/cpi/; U.S. Energy
14
Database of States Incentives for Renewables &
Information Administration, Annual Energy Review Efficiency, http://www.dsireusa.org.
2010 (Washington, DC: U.S. Department of Energy, 15
J. Lazar, L. Schwartz, and R. Allen, Pricing Dos
2010). and Donts: Designing Retail Rates as if Efficiency
6
K. R. Fabrizio, N. L. Rose, and C. D. Wolfram, Counts (Montpelier, VT: Regulatory Assistance
Do Markets Reduce Costs? Assessing the Impact Project, 2011), http://www.raponline.org/
of Regulatory Restructuring on US Electric document/download/id/939; and A. Faruqui
Generation Efficiency, American Economic Review and P. Fox-Penner, Energy Efficiency and Utility
97, no. 4 (2007): 1250-1277; L. W. Davis and C. Demand-Side Management Programs, presented
Wolfram, Deregulation, Consolidation, and to The World Bank, Washington, DC, July 14, 2011,
Efficiency: Evidence from U.S. Nuclear Power, http://www.brattle.com/Experts/ExpertDetail.
working paper 217 (Berkeley, CA: Energy Institute asp?ExpertID=164&page=1.
at Haas, 2011); and C. R. Knittel, Alternative 16
A. E. Kahn, The Economics of Regulation: Principles
Regulatory Methods and Firm Efficiency: Stochastic and Institutions (Cambridge, MA: MIT Press,
Frontier Evidence from the U.S. Electricity 1988), 6970.
Industry, The Review of Economics and Statistics 84,
no. 3 (2002): 530540. 17
Joskow, see note 7 above; T. Zeller Jr., Smart
Meters Draw Complaints of Inaccuracy, New York
7
P. L. Joskow, Inflation and Environmental Times, November 12, 2010.
Concern: Change in the Process of Public Utility
Price Regulation, Journal of Law and Economics 17, 18
P. L. Joskow, Regulatory Failure, Regulatory
no. 2 (1974): 291327. Reform, and Structural Change in the Electrical
Power Industry, in Brookings Papers on Economic
8
U.S. Department of Energy, Smart Grid System Activity, Microeconomics 1989, ed. C. Winston and
Report (Washington, DC: 2009); and M. Scheepers M. N. Baily (Washington, DC: Brookings
et al., Regulatory Improvements for Effective Institution Press, 1989), 125208.
Integration of Distributed Generation into Electricity
of Electricity Distribution Networks: Lessons of Megginson and Netter, see note 4 above.
35
Experience from Britain, Energy Policy 35, no. 12
(2007): 61636187. Federal Energy Regulatory Commission, 2010
36
The electric grid is a system of systems, electricity flowing through the grid. Many
managed by thousands of people, computers advances discussed in this studyfrom integra-
and manual controls, with data supplied by tens tion of variable energy resources to wide-area
of thousands of sensors connected by a wide situational awareness and real-time control to
variety of communications networks. Over the demand responseresult from or depend
next 20 years, the growth in percentage terms on this increase in data collection and
of data flowing through grid communications communications.
networks will far exceed the growth of
Operations Service
Markets Provider
Bulk
Generation
Transmission Distribution Customer
Source: National Institute for Standards and Technology, NIST Framework and Roadmap for Smart Grid Interoperability Standards,
Release 1.0, special publication 1108 (Washington, DC: U.S. Department of Commerce, 2010), 33, http://www.nist.gov/public_affairs/
releases/upload/smartgrid_interoperability_final.pdf.
s (OW MUCH ARE CONSUMERS OR SOCIETY AT s Commercial wide-area, field-area, and local
large) willing to pay to reduce this risk? (neighborhood) networks are used for
similar purposes to utility-owned networks as
s 7HO MAKES THE DETERMINATION OF SOCIETYS well as for communications among corporate
tolerance for risk and the cost of employing data centers. They rely on wired, wireless,
protections? radio-frequency or microwave, and power
line carrier communications, provided under
s (OW SHOULD THE COSTS OF EMPLOYING PROTEC- contract or operating arrangements from
tions be paid for? common public telecommunications service
providers.
s (OW IS DAMAGE MEASURED COST TO REPLACE
damaged equipment, number of people- s Public communications networks, such as
hours without power, number of other the telephone network and the Internet,
critical infrastructure nodes affected? transmit information, such as pricing signals
and daily generation schedules, and commu-
s 7HERE ARE INTERDEPENDENCIES MOST CRITICAL nicate with home energy networks.
To contribute to this discussion, this chapter s Satellite communications networks are used
examines critical topics and strategies to where microwave communication is prohibi-
increase awareness and resolution of cybersecu- tively expensive; phasor measurement units
rity and information privacy issues in the (PMUs) also use the GPS satellite navigation
future electric grid. system to synchronize timing.
Source: V. C. Gungor and F. C. Lambert, A Survey on Communication Networks for Electric System Automation, Computer Networks 50, 7 (2006): 87797.
Aggregator
RTO Transmission .FUFSJOH Distribution
*40350 SCADA SCADA System SCADA
Participant Internet/ Others
Internet/ e-Business
e-Business
Electric
.BSLFU Vehicle
Wide-area Energy
Services Field-area Distributed
Interface Networks Services
Networks Interface (FOFSBUJPO
Data
Plant Control Collector
Electric
System .FUFS Storage
Substation Substation Field Premises
LANs Controller Device Networks
(FOFSBUPST
Customer Appliances
Bulk Generation Substation Equipment
Device Distributed
Domain (FOFSBUJPO
Electric
Network Storage Customer Thermostat
Actor &.4
Gateway Actor Transmission Customer
Distribution
Comms Path
Comms Path Changes Owner/Domain
Source: National Institute for Standards and Technology, NIST Framework and Roadmap for Smart Grid Interoperability Standards,
Release 1.0, special publication 1108 (Washington, DC: U.S. Department of Commerce, 2010), 35, http://www.nist.gov/public_
affairs/releases/upload/smartgrid_interoperability_final.pdf.
Note: ISO/RTO = independent system operator/regional transmission organization; EMS = energy management sytstem SCADA =
supervisory control and data acquisition; WAMS = wide-area management system; DMS = distribution management system;
MDMS = meter data management system; CIS = customer information system; LAN = local-area network.
An important issue in the management and communications systems will need to enable
regulation of grid data communications customers to perform these multiple roles. As
systems will be the blurring of distinctions the grid evolves, the existing point-to-point and
one-way communications networks will need
Existing point-to-point and one-way to be expanded or replaced with networks
designed for two-way communication.5
communications networks will need to
be expanded or replaced with networks Data Communications Technologies
designed for two-way communication. and Applications
Network Requirements
Application Media Standard/ Expected Data Acceptable Frequency Reliability Security Backup
Protocol Rate/Bandwidtha Latencya of Useb Needa Needa Powera
Home-area Power line HomePlug,
Network communications;c ;JH#FF
*1
wireless
Advanced Power line For backhaul: m LJMPCZUFT 215 seconds 515 9999.99% High Not
.FUFSJOH communications,c,d 8J."9
-5& TFDPOE LCQT
NJOVUFT necessary
Infrastructure wireless radio For appliance OPEF
LCQT node
".*
frequency,e,f to meter: for backhaul
T1, microwave, *&&&
h
broadband (via ;JH#FFg
fiber, cable, digital
subscriber line),
commercial wirelessg
Demand 4BNFBT".* 4BNFBT".* LCQTmLCQT EBZT 9999.99% High Not
Response node or device milliseconds year necessary
1BSUPG".*
(ms)
several
minutes
Electric Power line ;JH#FF
m LCQT
2 seconds Daily 9999.99% Relatively Not
Transportation communicationsi *&&& h LCQT JT 5 minutes high necessary
wirelessh a good target
%JTUSJCVUJPO(SJE Fiber, wireless,j DNP3 (IEEE 1815), mLCQT NTm Continuous 9999.999% High m
.BOBHFNFOU satellite, cellularg *&$ 2 seconds hours
(004&
k8J."9
-5&
j IP,g
*&&&h
Distributed Fiber, wireless,j DNP3, m LCQT NTm Continuous 9999.99% High 1 hour
Energy microwave, *&$ 15 seconds
Resources and satelliteg (004&k 8J."9
-5&
j
Storage ;JH#FF
g
*&&& h
Wide-area 40/&5
"5.
$
m
LCQT NTmNT Continuous 99.999 High 24-hour
Situational 'SBNF3FMBZ
.1-4
f,g *&$ 99.9999% supply
Awareness fiber, microwave, (004&
k
(synchro- broadband over IPh,l
phasors#) power lineg
Interutility Fiber, microwave, ICCPk NFHBCZUFT < NT %4
Continuous 99.999 High 24-hour
communications wired second (mbps) 99.9999% supply
(Southern
California
Edison)
Interregional Standard telco IP LCQT m NT Continuous 99.999% High 24-hour
data T1 circuits with supply
communications copper endpoints
(ISO New (NERCNet)
England)
.BSLFUEBUB Wired IP 18 mbps + 45 mbps mNT Continuous 99.999% Relatively 24-hour
communications connections high supply
(ISO New
England)
*Communications between the utility and smart meters have different requirements than those between smart meters and appliances,
although these are sometimes lumped under the category advanced metering infrastructure. While the former necessitate reliable
communications over long distances, the latter necessitates low latency over short distances.
#A significant synchrophasor initiative is the North American SynchroPhasor Initiative. A communications network called NASPINet
to support these technologies is under construction. More information may be found at http://www.naspi.org.
a
Indicated column in source table from: U.S. Department of Energy, Communications Requirements of Smart Grid Technologies,
Appendix A (Washington, DC, 2010), http://www.doe.gov/sites/prod/files/gcprod/documents/Smart_Grid_Communications_
Requirements_Report_10-05-2010.pdf.
b
Frequency developed from: U.S. Department of Energy, Communications Requirements of Smart Grid Technologies, Appendix A
(Washington, DC, 2010), http://www.doe.gov/sites/prod/files/gcprod/documents/Smart_Grid_Communications_Requirements_
Report_10-05-2010.pdf.
c
. 0AVLIDOU ! * (AN 6INCK * 9AZDANI " (ONARY h0OWER LINE COMMUNICATIONS 3TATE OF THE !RT AND &UTURE 4RENDSv IEEE
Communications Magazine 41, 4 (April 2003): 3440.
d
EDN Europe, Maxim and Sagem to Develop Power-Line Comms for EDF, press release, December, 12, 2008, http://www.edn-europe.
com/maximsagemtodeveloppowerlinecommsforedf+article+2679+Europe.html.
e
V. C. Gungor and F. C. Lambert, A Survey on Communication Networks for Electric System Automation, Computer Networks 50, 7
(2006): 87797.
f
M. McGranaghan, D. Von Dollen, P. Myrda, and E. Gunther, Utility Experience with Developing a Smart Grid Roadmap, presentation
at IEEE Power and Energy Society General Meeting, Pittsburgh, PA, July 2024, 2008.
g
U.S. Department of Energy, Communications Requirements of Smart Grid Technologies: Department of Energy (Washington, DC, 2010),
http://energy.gov/sites/prod/files/gcprod/documents/Smart_Grid_Communications_Requirements_Report_10-05-2010.pdf.
h
Personal communication with Exelon Staff, April 25, 2011.
i
Renault Nissan, Renault and EDF Strengthen Collaboration on Zero-Emission Electric Vehicle, press release, June 22, 2009, http://
www.media.renault.com/download/media/specialfile/9210_1_5.aspx.
j
V. K. Sood, D. Fischer, J. M. Eklund, and T. Brown, Developing a Communication Infrastructure for the Smart Grid, presentation
at IEEE Electrical Power & Energy Conference, Montreal, QC, Canada, October 2223, 2009.
k
Personal communication with Southern California Edison Staff, March 15, 2011.
l
QualityLogic, IEEE C37.118 PMU Communications, http://www.qualitylogic.com/Contents/Smart-Grid/Technology/IEEE-C37_118.
aspx.
i
It is important to note that the use of IP for grid data communications is not the same as using the public
Internet. In most cases, the discussion of IP networks for grid communications envisions fully separate
networks that are not connected to the public Internet, although some data communications applications
do envision using the public Internet. These are separate debates. For the electric grid, it will be vital to keep
critical grid communications systems from talking to the public Internet and becoming infected as a result.
ii
Even this should be viewed as an evolving situation, although changes may be a decade or more away.
For example, the National Science Foundation supports a Global Environment for Network Innovations
program, which aims to design protocols that can run on the Internet in parallel with IP to reduce latency
and improve security for future applications (see http://www.geni.net/).
iii
In its National Broadband Plan, the FCC makes the following recommendations on the issue of grid data
communications network ownership: The country should pursue three parallel paths. First, existing
commercial mobile networks should be hardened to support mission-critical Smart Grid applications.
Second, utilities should be able to share the public safety mobile broadband network for mission-critical
communications. Third, utilities should be empowered to construct and operate their own mission-
critical broadband networks. Each approach has significant benefits and trade-offs, and what works in one
geographic area or regulatory regime may not work as well in another. Rather than force a single solution,
these recommendations will accelerate all three approaches.22
s Components: The electric grid will be In 2006, a Roadmap to Secure Control Systems
composed of components from multiple in the Energy Sector was prepared for the DOE
suppliers, with multiple interfaces and AND $(3 WHICH HAVE BEEN COLLABORATING ON ITS
protocols, and relying on multiple standards. implementation since 2007.iv, 29 In 2010, the
National Broadband Plan recommended (and
s Continuous transition: The information and the FCC is following up on) creating a more
communications technologies (ICT) used in far-reaching Cybersecurity Roadmap for
the grid will continue to change at a faster communications.30 That same year, the GAO
rate than utilities can change components in issued a report on challenges to cybersecurity
the grid, resulting in incompatibilities and research and development.31 In 2011, the DOE
security vulnerabilities between existing and announced a publicprivate collaboration
new ICT. including NIST and NERC to develop guide-
lines for cybersecurity risk management in the
electric sector.32
iv
The Energy Sector Control Systems Working Group, a publicprivate partnership that includes
REPRESENTATIVES FROM THE $/% AND $(3 IS TASKED WITH IMPLEMENTING THE ROADMAP 3EE HTTPWWW
controlsystemsroadmap.net/workinggroup.shtml. The 2011 update to the 2006 roadmap is now
available at http://www.controlsystemsroadmap.net/pdfs/roadmap.pdf.
Assess
vulnerabilities,
threats,
impacts
Reduce
Recover and
vulnerabilities,
restore
threats, impacts
MITIGATION
Prevent attacks,
Respond during
incidents, other
attack
outages
r*OTJEFST r1VSDIBTJOH
r)BDLFST"UUBDLFST r)JSJOH
r4PGUXBSF
%FWFMPQNFOU
r0QFSBUJPOT
1FPQMF 1SPDFTTFT
1IZTJDBM
&OWJSPONFOU 5FDIOPMPHZ
r%BUB$FOUFST r)BSEXBSF
r0UIFS*OGPSNBUJPOBOE r"QQMJDBUJPOT
$PNNVOJDBUJPOT-PDBUJPOT r'JSNXBSF
r$PNNVOJDBUJPO-JOFT r$PNNVOJDBUJPOT
*OUFSGBDFT
Process security ensures that all operational communication protocols, and communications
processes include measures to protect the interfaces. The future grid also will have millions
enterprise, its equipment, and its products. In of programmable devicesmost notably smart
the case of grid cybersecurity, examples include meters, but also electric vehicles, PMUs, devices
running or validating the results of various in electric grid substations, and other equip-
security checks on equipment before certifying mentthat all present software application and
them for purchase, performing outside security firmware security vulnerabilities.
checks on potential IT and communications
hires, implementing software development Communications security includes mitigating
processes that include security checklists, and protocol vulnerabilities that can impact the
doing physical security checks of computer and ability of communications network protocols
communications equipment areas. to transmit their data securely. In this case,
some security issues and solutions may be
Technology security involves the design, dependent on the protocol in use. Communi-
implementation, and interoperability of cations interfaces within and between grid
communications and IT hardware, application systems introduce critical vulnerability points
software, device-embedded software (firmware into the electric grid network. For example,
typically provided by the manufacturer), customer demand response might involve an
v
Case studies on the cybersecurity risks associated with not just AMI but also distribution grid manage-
ment and electric vehicles are available from the Advanced Security Acceleration Project for the Smart
Grid, a collaborative funded by DOE and various utility companies to accelerate the development of
security requirements for the grid. See http://www.smartgridipedia.org/index.php/ASAP-SG.
Sources:
a
S. McLaughlin, D. Podkuiko, and P. McDaniel, Energy Theft in the Advanced Metering Infrastructure, in Proceedings of the
4th International Workshop on Critical Information Infrastructure Security .EW 9ORK .9 )%%% 0RESS
b
R. Shein, Security Measures for Advanced Metering Infrastructure Components, in 2010 Asia-Pacific Power and Energy
Engineering Conference .EW 9ORK .9 !00%%# AND )%%%
c
InGuardians, Advanced Metering Infrastructure Attack Methodology, Vol 1.0. (Washington, DC, 2009).
d
F. M. Cleveland, Cyber Security Issues for Advanced Metering Infrastructure (AMI), presented at the IEEE Power and Energy
Society General Meeting, Pittsburgh, PA, July 2024, 2008.
e
# "ENNETT AND $ (IGHlLL h.ETWORKING !-) 3MART -ETERSv PRESENTED AT %NERGY )%%% #ONFERENCE ON 3USTAINABLE %NERGY
Infrastructure, Atlanta, GA, November 1718, 2008.
vi
A 2011 GAO report criticizes the NIST guidelines for their lack of information on combined
cyberphysical attacks and the absence of a final schedule for updating the guidelines.67
vii
The FCC identified the potential for conflicts between the existing CIP requirements and other standards
as an area of concern and opined that the resulting ambiguity was slowing utility decision-making and
deployment of some new technologies.68
viii
The study of information privacy deals with policy issues ranging from identification and collection
to storage, access, and use of information. The study of information security deals with protecting
information from unauthorized access and use as defined by information privacy rules or otherwise.
ix
While information and data are used somewhat interchangeably, they are related but different terms;
information is processed data.
x
Utilities also deal with personally identifiable information (PII) in the normal course of their operations,
as do all companies in consumer-facing businesses. PII is any data or information that identifies an
individual person or organization. For example, name, address, and phone number taken together are
considered PII. The privacy issues related to PII have been addressed in many forums and are not
discussed further in this report.
xi
See, for example, NERC Regulation CIP-003-4, which specifies implementation of a cyber security policy
that represents managements commitment and ability to secure its Critical Cyber Assets.
Private Public
network
Utility Back
Office network
Private or Systems
commercial
broadband
Ethernet on microwave Utility Public Govt.
power line, Central network
Agencies
private or Office
commercial
cellular or
broadband Local Data Third-party Public
Collector Services network
WiFi, Ethernet
on home wiring
Smart
Meter
Third-party
Home
Services
Public network Possible state border
xii
The Federal Trade Commission enforces the Fair Credit Reporting Act (see http://www.ftc.gov/os/
statutes/031224fcra.pdf), which includes the ability of consumers to sue for damages if data are
improperly disclosed by a credit reporting agency. The Payment Card Industry Security Standards Council
was formed by major credit card companies to develop standards in an effort to reduce public disclosure
of credit card information (see https://www.pcisecuritystandards.org/index.shtml).
xiii
As will be discussed in this chapter, there are several such guidelines from major national and
international organizations. This particular stricture does not appear in the Fair Information Practice
Principles published by the Federal Trade Commission and is often recommended for adoption related
to regulation of the future grid.
xiv
Note that electricity supply restructuring does not resolve this issue. Electricity customers may be able to
choose the generator of their electricity, but they almost always have to deal with a single distributor
the utility that brings the electricity into the home or business and sends them the bill. The one exception
is large industrial customer that may deal with more than one distributor.
Communications Commission on NBP Public Proposal, fact sheet, May 12, 2011, http://www.
Notice #2 (2009); and AT&T, Comments to the whitehouse.gov/the-press-office/2011/05/12/
Federal Communications Commission on NBP fact-sheet-cybersecurity-legislative-proposal.
Public Notice #2 (2009).
North American Electric Reliability Corporation
34
Plan Recommendation to Create a Cybersecurity Terror Report Warns of Insider Threat to Utilities,
Roadmap, public notice, August 9, 2010. ABCNews.com, July 20, 2011, http://abcnews.go.
U.S. Government Accountability Office, Key
31 com/Blotter/terror-alert-warns-insider-threat-
Challenges Need to Be Addressed to Improve infrastructure/story?id=14118119.
Research and Development (Washington, DC, National Institute for Standards and Technology,
48
Electric Institute to the Department of Energy Secretary, U.S. Department of Energy, November 2,
Request for Information on Addressing Policy and 2010, 2, www.doe.gov/sites/prod/files/oeprod/
Logistical Challenges to Smart Grid DocumentsandMedia/ABB_Comments.pdf.
Implementation (2010), http://energy.gov/sites/
Daniel Thanos, Chief Cybersecurity Architect,
61
prod/files/oeprod/DocumentsandMedia/EEI_-_
GE Digital Energy, personal communication with
DOE_SG_RFI.PDF; and North American Electric
the authors, October 25, 2010.
Reliability Corporation and U.S. Department of
Energy, see note 3 above. North American Electric Reliability Corporation,
62
No. 83410, In the Matter of the Application of Guidelines for Smart Grid Cyber Security, see note
Baltimore Gas & Electric to Deploy a Smart Grid 4 above.
Initiative and to Establish a Surcharge for the
U.S. Government Accountability Office, see
67
Recovery of Cost, June 22, 2010.
note 17 above.
56
Edison Electric Institute, see note 49 above,
Federal Communications Commission, see
68
page 60.
note 21 above.
57
Federal Energy Regulatory Commission, 2011
D. Whiteley, Testimony of David A. Whiteley,
69
Assessment of Demand Response and Advanced
Executive Vice President, NERC before the
Metering Staff Report (Washington, DC, 2011).
Subcommittee on Emerging Threats, Cybersecurity,
58
Guidelines for Smart Grid Cyber Security, see note 4 and Science and Technology Committee on
above; and Advanced Security Acceleration Project Homeland Security U.S. House of Representatives on
for the Smart Grid, Security Profile for Advanced The Cyber Threat to Control Systems: Stronger
Metering Infrastructure, Version 2.0 (2010), http:// Regulations are Necessary to Secure the Electric
www.smartgridipedia.org/images/9/90/AMI_ Grid, October 25, 2007.
Security_Profile_-_v2_0.pdf.
U.S. Energy Information Administration, Electric
70
59
Edison Electric Institute, note 49 above, page 54; Sales, Revenue, and Average Price 2009
American Public Power Association, Comments (Washington, DC: U.S. Department of Energy,
of the American Public Power Association in the 2010), Table 5A, http://www.eia.gov/cneaf/
Matter of Addressing Policy and Logistical electricity/esr/table5_a.html.
Challenges to Smart Grid Implementation,
Electric Power Research Institute, see note 16
71
(2010), 2.
above, pages 25.
Privacy Issues Related to Smart Grid Technologies U.S. Department of Energy, see note 74,
88
Commission, see note 21 above; and Guidelines for 111th Congress (2010), http://www.govtrack.us/
Smart Grid Cyber Security, see note 4 above. congress/bill.xpd?bill=h111-4860.
Commissioners, Resolution Urging the Adoption Comments of the Center for Democracy &
of General Privacy Principles for State Commission Technology and the Electronic Frontier
Use in Considering the Privacy Implications of the Foundation on Proposed Policies and Findings
use of Utility Customer Information (Washington, Pertaining to the Smart Grid, March 9, 2010,
DC, 2000). http://docs.cpuc.ca.gov/efile/CM/114696.pdf.
Grid and Consumer Data, presentation at the American Institute of Certified Public Accountants
92
Smart Grid Summit, Washington, DC, April 8, and Canadian Institute of Chartered Accountants
2010, http://epic.org/privacy/smartgrid/EPIC_ Privacy Task Force, Generally Accepted Privacy
Statement_Smart_Grid_Summit_Cybersecurity_ Principles .EW 9ORK .9 HTTPWWWAICPA
and_Privacy.pdf. org/InterestAreas/InformationTechnology/
Resources/Privacy/GenerallyAcceptedPrivacy
$ .ARCISO h0OLICE 3EEK 5TILITY $ATA FOR (OMES OF
78
Principles/Pages/default.aspx.
Marijuana-Growing Suspects, Columbus Dispatch,
February 28, 2011, http://www.dispatch.com/live/ Canadian Standards Association Technical
93
smartmetertexas.com/CAP/public/index.html.
Ontario Ministry of Energy, Smart Meters and
84
Thomas Edison introduced the first electric of carrying alternating current (ac) power over
power system in New York City in 1882. This long distances with relatively low losses. In 1896,
direct current (dc) system initially served 59 George Westinghouse began the hydroelectric
customers in the Wall Street area at a price of development of Niagara Falls, transmitting
about $5 per kilowatt hour (kWh).ii It operated significant power to Buffalo, New York, more
at 100 volts and mainly supplied electric lights. than 20 miles away. This inaugurated the practice
By the end of the 1880s, many cities had similar of locating generators at some distance from
small central stations that each served only a few load centers and linking them by high-voltage
city blocks. transmission, then using transformers to lower
the voltage delivered to ultimate customers.
To the extent that the industry was regulated,
city governments performed this function. City Since then, engineering research and the devel-
governments also became major customers opment of new materials have enabled the use
for street lighting and trolley servicesand of ever-higher voltages. In the U.S., ac lines with
could extract various concessions in exchange for voltages of up to 150 kilovolts (kV) were in place
the right to string wires. Soon, they also became by 1910, and the first 245 kV line was commis-
owners. By 1900, municipally owned utilities sioned in 1922. The invention of the transformer
accounted for about 8% of total U.S. generation. and high-voltage lines allowed private utilities to
Vigorous debates about the relative merits of expand beyond municipal boundaries and take
government- and investor-owned utilities better advantage of economies of scale. Such
continued in the U.S. through the 1930s, when expansion compounded problems with munici-
federal policies were established that today still pal regulation and led to state regulation of
favor government-owned and cooperative investor-owned electric utilities, generally with
utilities. the utilities active support. This trend began
with the establishment of regulatory commissions
A.1 BEYOND MUNICIPAL BOUNDARIES in Wisconsin, Georgia, and New York in 1907.
By 1914, 30 states had regulatory commissions,
The transformer was first demonstrated at scale and today all states and the District of Columbia
in Germany in 1891. This innovation enabled the have them.iii
use of relatively high-voltage transmission capable
i
Significant portions of Appendix A are taken from U.S. Energy Information Administration, The Changing
Structure of the Electric Power Industry 2000: An Update (Washington, DC: U.S. Department of Energy, 2000),
http://www.eia.gov/cneaf/electricity/chg_stru_update/update2000.html; and Ignacio J. Perez-Arriaga, Hugh
Rudnick, and Michel Rivier, Electric Energy SystemsAn Overview, in Electric Energy Systems: Analysis and
Operation, ed. Antonio Gomez-Exposito, Antonio J. Conejo, and Claudio Canizares (Boca Raton: CRC Press,
2008), 150. See also David E. Nye, Electrifying America (Cambridge: MIT Press, 1990).
ii
Except where otherwise noted, this price and others in Appendix A are expressed in 2010 dollars, using the
Consumer Price Index from 1913 forward and the extension to earlier years available at http://www.
measuringworth.org/datasets/uscpi/result.php/.
iii
Nebraskas regulatory commission does not oversee electricity rates because no investor-owned utilities sell
electricity in that state.
Appendix A 235
Vertically integrated, investor-owned firms As Figure A.1 shows, average electricity prices
performing generation, transmission, and across all customer classes continued to decline
distribution as the sole provider within desig- in real terms until fuel prices began increasing
nated service areasemerged as the dominant sharply following the first oil crisis in 1973.
model. States enabled these firms to charge Between 1949 and 1973, electricity use grew at
prices that allowed them to cover their costs. an annual average rate of 8.3%; it slowed to an
annual average growth rate of 2.5% between
Among those costs was a fair rate of return, 1973 and 2006 before declining slightly in
determined by the commission, on the capital 20072009.4 Today, virtually all homes have
the utility had invested; this regime came to be grid-provided electricity, and the average retail
known as rate-of-return regulation of electric price of electricity is approximately $0.10 per
and other utilities. kWh over all classes of consumers.
This evolution was politically rather than A.2 A FEDERAL ROLE EMERGES
technologically determined. Because the U.S.
political system was highly decentralized until The federal role in the electric utility industry
at least the 1930s and most electric utilities began in 1906, when legislation authorized the
operated within a single state, state regulation sale of surplus power from federal irrigation
was the politically natural successor to super- projects, giving sales preference to municipali-
vision by municipal governments. In a different ties. Navigable waterways had been under
political context, in England local utilities were federal jurisdiction since the early 19th century,
unable to expand for political reasons, and the and the Federal Water Power Act of 1920 both
system remained highly fragmented until a codified federal powers over navigable water-
1926 law mandated the establishment of an ways and established the Federal Power Com-
integrated nationwide grid.1 In the U.S., due mission, later the Federal Energy Regulatory
in part to strong faith in private enterprise over Commission (FERC), to issue hydroelectric
the first three decades of the 20th century, the power licenses.
relative importance of publicly owned utilities
declined during this period.2 During the Depression of the 1930s,
confidence in unregulated markets waned,
Between the turn of the century and the 1930s, and after significant political struggles, the
electric utility capacity and generation grew at federal role in electric power was expanded
average rates of about 12% per year, doubling dramatically under the New Deal. The Public
roughly every six years, despite a 14% drop Utility Holding Company Act of 1935 gave the
in generation between 1929 and 1932. The Securities and Exchange Commission regula-
efficiency of generators and transformers tory authority over utility holding companies
improved, and, as noted earlier, transmission and led to the dismantling of the largest
voltages increased. Residential prices fell companies, which had been constructed during
dramatically, from an average of about $4.30 the 1920s and associated with various stock
per kWh at the turn of the century to $0.88 market abuses. The Federal Power Act of 1935
per kWh in 1932 (residential customers only).3 empowered the Federal Power Commission to
Electricity was not yet universally available regulate the wholesale transmission and sale of
in 1932, but two-thirds of all homes already electric power. The Rural Electrification Act of
used it. 1936 established the Rural Electrification
45
40
35
2010 Cents / kWh
30
25
20
15
10
0
1926
1930
1934
1938
1942
1946
1950
1954
1958
1962
1966
1970
1974
1978
1982
1986
1990
1994
1998
2002
2006
2010 Cents
Source: Data compiled from periodic statistical bulletins published by the Edison Electric Institute (formerly the National
Electric Light Association). Historical prices expressed in 2010 cents using the Consumer Price Index.
50 50 50 50
Administration to provide loans and assistance the largest electricity generator in the nation at
to organizations (mainly rural electric coopera- the end of World War II. It now serves 9 million
tives) that would provide electricity to rural customers in most of Tennessee and parts of
areas instead of investor-owned utilities. six adjacent states, and only about 10% of its
production today comes from hydropower.5
Also beginning in the 1930s, the federal govern-
ment invested heavily in hydroelectric facilities Next, the federal government established the
on the waterways under its control and used Bonneville Power Administration in 1937 to
these facilities to provide less expensive energy market low-cost power from the Bonneville Dam
to preferred customers: mainly municipal and on the Columbia River. The Bonneville Power
cooperative utilities. Large Bureau of Reclama- Administration now provides about 35% of
tion damssuch as the Hoover Dam, which the power in the Pacific Northwest and exports
finished construction in 1936, and the Grand significant power to neighboring regions.
Coulee Dam, which finished construction in It also controls about 75% of the high-voltage
1942served the western states. Under the transmission lines in the Pacific Northwest.6
Tennessee Valley Authority Act of 1933, the Similarly, the Southwestern Power Administration
federal government supplied hydroelectric and the Southeastern Power Administration
power to states, counties, municipalities, and were established in 1944 and 1950, respectively,
nonprofit cooperatives through its own corpo- to market low-cost federal power to preferred
ration, the Tennessee Valley Authority. It was customers in designated regions.
Appendix A 237
By 1950, federal generation accounted for more encouraged, it was ultimately voluntary
than 12% of total U.S. generation.7 Enthusiasm throughout most of NERCs history. The
for expanding the federal role in the electric Energy Policy Act of 2005 and a subsequent
power sector declined after World War II, FERC order paved the way for these procedures
however, and this percentage has since fallen. to be formalized into mandatory, enforceable
standards. The current high degree of inter-
A.3 INTERCONNECTION AND connection and the possibility of large-scale
COMPETITION disruptions were dramatically demonstrated by
the 2003 blackout, which spanned five indepen-
In the very early years of the industry, trans- dent system operator (ISO) regions (Midwest
mission was planned and built by vertically ISO, PJM, New York ISO, ISO New England,
integrated utilities that met their native loads and Ontarios Independent Electricity Market
from their own generation. Even though Operator), as well as multiple independent
interconnection promised to improve reliabil- control areas from Michigan to New Jersey.9
ity, U.S. utilities were reluctant to interconnect
and thus lose some control over their systems Despite these concerns, and to some extent in
until pressed to do so during World War I.8 response to the Arab oil embargo, the federal
Interconnection accelerated after the war, and government began opening the door to non-
in 1927 three major utilities established the utility generation in the late 1970s in order to
PNJ Interconnection as a power pool linking reduce dependence on foreign oil, promote
transmission facilities in Pennsylvania, New alternative energy sources, and diversify the
Jersey, and Maryland that grew into the PJM electric power supply. In 1978, Congress passed
regional transmission organization (RTO). the Public Utility Regulatory Policies Act
Because different utilities often had standard- (PURPA), which required regulated electric
ized on different transmission voltages, mergers utilities to buy power from non-utility genera-
and interconnections between adjacent utilities tors using cogeneration, renewable, or other
often requiredand often still requiretrans- designated technologies at the utilitys avoided
formers to link lines with different voltages. cost of generation, which it left to state regula-
These transformers produce losses. Nonethe- tors to determine. California and some other
less, interconnection continued, and the states determined that avoided cost was
Eastern Interconnection in its present form substantially above current costs and thereby
was established in 1962. encouraged the construction of substantial gen-
eration capacity that proved to be uneconomic.
A major northeastern power blackout in 1965 Subsequent legislation, particularly the Energy
raised concerns about the reliability of inter- Policy Act of 1992, lowered barriers to the
connected power networks. In response, the market entry of independent power producers,
electric utility industry formed the North generators that have no legal obligation to
American Electric Reliability Council (NERC, provide power to ultimate customers, and
later renamed the North American Electric utilities that could generate electricity for
Reliability Corporation) and its regional customers outside their service territories.
reliability councils to address the reliability
and adequacy of the bulk power supply. NERC In the 1980s, a new model for power system
led the development and revision of reliability organization began to emerge in the academic
operating procedures for the grid. While and policy literature.10 In this model, organized
compliance with these procedures was strongly competitive markets would set the price of
wholesale electricity. Ownership of generation
Appendix A 239
necessary at all times in order to minimize This episode and the continued opposition
system cost, satisfy reliability constraints, and of municipal and cooperative utilities, which
respond to unexpected changes in load and benefit from a variety of subsidies and prefer-
other events. But now the cost to be minimized ences put in place in the 1930s, have reduced
is the cost of buying power from independent political support for the new industry structure.
generators in a wholesale market, at prices at Though the California ISO continues to
which the generators are willing to supply. operate, the state is no longer a leader in retail
Competitive suppliers of new generating competition.17 No new ISOs have been formed
capacity play an important role in deciding since this episode, though some existing ISOs
what gets built and where. The detailed imple- have added affiliates. Movements toward retail
mentations of this new bulk power system competition have been suspended in seven
structure have varied considerably over time states, and electricity price increases in some
and space, but they also have converged in areas have led to calls to roll back prior
important respects over time, and most reforms.18
observers agree that ISO/RTO systems have
generally performed well. In summary, the growth and evolution of
the electric power industry since the dawn of
At the retail level, most states were not eager electrification was determined by a multitude
to embrace competition. Nonetheless, 22 states of different interacting factors. The industrys
and the District of Columbia did begin a history is, in part, a story of continuous techni-
variety of processes aimed at allowing compet- cal innovation. However, just as importantly,
ing vendors of electric power at the retail level, the pace and direction of change in the industry
while continuing to treat the distribution of was also heavily influenced by shifting regula-
that power as a regulated monopoly.iv tory designs and policy priorities and the pace
of economic growth. Finally, the evolution of
California was a leader in the movement the grid was also strongly shaped by unforeseen
toward greater reliance on markets during emergencies like the 1965 and 2003 blackouts
the 1990s, with a wholesale market that began and the California energy market crisis. Factors
operation in 1998 and competing retailers of all these sorts will continue to shape the
allowed to enter in that same year. During 2000 evolution of the grid in the future.
and 2001, however, the California wholesale
market experienced dramatic price increases
and blackouts. The causes of this crisis involved
a flawed market design, unusual shortage
conditions in the western grid, illegal market
manipulation on the part of some market
participants, and inappropriate regulatory
responses to emerging problems.16
iv
State-by-state information on the history and status of vertical disintegration and retail competition
efforts can be found at http://www.eia.doe.gov/cneaf/electricity/page/restructuring/restructure_elect.html.
Electric Utility Industry through 1970 (New York, Reforms of Reforms, Journal of Regulatory
1973). Economics 21 no. 1 (2002): 103132, http://www.
springerlink.com/index/M488771227KN3R02.pdf.
3
U.S. Energy Information Administration, The
Changing Structure of the Electric Power Industry W. W. Hogan, Electricity Wholesale Market
14
2000: An Update (Washington, DC: U.S. Design in a Low Carbon Future, in Harnessing
Department of Energy, 2000), http://www.eia.gov/ Renewable Energy in Electric Power Systems, eds.
cneaf/electricity/chg_stru_update/update2000. B. Moselle, J. Padilla, and R. Schmalensee
html. (Washington, DC: RFF Press, 2010).
4
U.S. Energy Information Agency, Annual Energy International Energy Agency, Tackling Investment
15
Review 2009 (Washington, DC: U.S. Department of Challenges in Power Generation in IEA Countries:
Energy, 2010), http://www.eia.doe.gov/totalenergy/ Energy Market Experience (Paris, 2007): 116.
data/annual/index.cfm. F. A. Wolak, Lessons from the California
16
5
Tennessee Valley Authority, TVA Fact Book Electricity Crisis, working paper 110R (Berkeley,
(Knoxville, TN, 2011), http://www.tva.com/ CA: University of California Energy Institute,
abouttva/factbook.htm. 2003), http://www.ucei.berkeley.edu/pubs-csemwp.
html; and J. L. Sweeney, The California Electricity
6
Bonneville Power Administration, BPA Facts Crisis (Stanford, CA: Hoover Institution Press,
(Portland, OR, 2011), http://www.bpa.gov/ Stanford University, 2002).
corporate/about_BPA/Facts/.
KEMA, Innovation in Competitive Electricity
17
7
Edison Electric Institute, see note 2 above. Markets, white paper (Burlington, MA, 2011).
8
Hughes, see note 1 above. Ohio Lawmakers Will Take Another Look at
18
9
U.S.Canada Power System Outage Task Force, Re-regulating States Electric Industry, Electric
Final Report on the August 14, 2003 Blackout in Utility Week, October 25, 2004.
the United States and Canada: Causes and
Recommendations (Washington, DC: U.S.
Department of Energy, 2004), https://reports.
energy.gov/.
10
P. Joskow and R. Schmalensee, Markets for Power:
An Analysis of Electric Utility Deregulation
(Cambridge, MA: MIT Press, 1983).
11
U.S. General Accounting Office, Lessons Learned
from Electricity Restructuring, GAO-03-271
(Washington, DC, 2002), http://www.gao.gov/new.
items/d03271.pdf; and J. T. Kelliher, Statement of
Chairman Joseph T. Kelliher, presented at Federal
Energy Regulatory Commission Conference on
Competition on Wholesale Power Markets
AD07-7-000, February 27, 2007, Washington, DC.
Appendix A 241
242 MIT STUDY ON THE FUTURE OF THE ELECTRIC GRID
Appendix B: Electric Power System Basics
Electricity is critical to our daily lives, yet most Energy is the ability to perform work. Energy
people have little understanding of the complex cannot be created or destroyed but can be
process that brings electric power to our homes, converted from one form to another.ii For
offices, and factories whenever we demand it. example, chemical energy in fossil fuels can
This appendix is a tutorial on how the electric be converted into electrical energy, and electrical
power system works. We assume no prior energy in turn can be converted into useful work
knowledge in the area and start by providing in the form of heat, light, and motion. While the
a description of the physical foundations of scientific community measures energy in watt-
electricity. We next discuss the structure and seconds or joules, traditionally in the electric
components of the electric power system. We power industry, energy is measured in watt-hours
follow with an explanation of how the system is (Wh) and for larger values is expressed in kilowatt
operated and how wholesale electricity markets (thousand watt, kW), megawatt (million watt,
work. In the final section, we provide a brief MW), gigawatt (billion watt, GW), or terawatt
overview of system planning. Because there are (trillion watt, TW) hours.iii A 100 watt lightbulb
slight differences in the structure, operation, consumes 2,400 Wh (or 2.4 kWh) of energy in
and planning of the electric power system from 24 hours, and the total annual electrical energy
country to country and region to region, we consumption of the U.S. in 2010 was about
focus mostly on fundamental aspects that remain 3,900 TWh.1 One kilowatt hour is equivalent to
unchanged; however, where appropriate we 3.6 megajoules.
provide U.S.-centric details and highlight
important variations in practice. Voltage
i
Those who only desire a high-level understanding of electric power systems can skip this section.
ii
If mass is not considered a form of energy, an exception is in nuclear reactions, where mass and energy can be
transformed into one another.
iii
Watt is the unit of power, or the rate of flow (or consumption) of energy, as discussed later in this section.
Appendix B 243
Current Ac voltage and current waveforms are defined
by three parameters: amplitude, frequency, and
Current is a measure of the rate of flow of phase, as shown in Figure B.1. The maximum
charge through a conductor. It is measured in value of the waveform is referred to as its
amperes. Current can be considered analogous amplitude. The amplitude of the ac voltage in
to the rate of flow of water through a pipe. a standard 120 V outlet is 170 V. The 120 V in
this case refers to the root-mean-square (rms)
Dc and Ac value of the voltage and is the equivalent dc
voltage with the capacity to perform the same
Current can be unidirectional, referred to as amount of work. In the case of ac, the ampli-
direct current, or it can periodically reverse tude is equal to the rms value multiplied by the
directions with time, in which case it is called square root of two. In the case of dc, the
alternating current. Voltage also can be amplitude and rms values are the same.
unipolarin which one point is always at a
higher voltage than the otheror alternating in Frequency is the rate at which current and
polarity with time. Unipolar voltage is referred voltage in the system oscillate, or reverse direc-
to as dc voltage. Voltage that reverses polarity tion and return. Frequency is measured in cycles
in a periodic fashion is referred to as ac per second, also called hertz (Hz). In the U.S.,
voltage. Alternating currents and voltages in as well as the rest of North America and parts
power systems have nearly sinusoidal profiles. of South America and Japan, the ac system
frequency is 60 Hz, while in the rest of the world
it is 50 Hz.2 Dc can be considered a special case
of ac, one with frequency equal to zero.
Figure B.1 Amplitude, Frequency, Period, and Phase of an Alternating Current or Voltage
Waveform
root-mean-square
value
amplitude
phase
time
0
period
1
=
frequency
iv
Synchronized systems are at the same frequency and have a specific phase difference between their
voltages.
Appendix B 245
reactance of the transmission line and forces Real Power
the transmission voltage and current to be in Real power, also called active power or
phase (a good thing). This process is called average power, is the average value of instan-
line compensation. taneous power, as shown in Figure B.2, and is
power that actually does work. It is measured
The inductive reactance of a transmission line in watts. Although instantaneous power can
is proportional to both frequency and line be flowing in both directions, real power
length, and for long ac lines the inductive only flows in one direction, as shown in
reactance limits the amount of power the line Figure B.2(a)(c). Real power is zero if the
can carry. At zero frequency (dc) the reactance phase difference between voltage and current
is zero, making dc attractive for long-distance is 90 degrees, as shown in Figure B.2(d).
transmission.
Reactive Power
Power If the voltage and current waveforms are in
phasethat is, they cross zero at the same
Power is the rate at which energy is flowing or timethen instantaneous power, although
work is being done.v Since voltage is the amount varying, is always positive or flowing in one
of work done for each unit of charge that flows direction (Figure B.2(b)). In this case, all the
and current is the rate of flow of charge, the power is real power. However, if one waveform
product of voltage and current is the rate of is shifted in time relative to the other, a condi-
workpower, or more precisely instantaneous tion called out of phase, then power takes on
power. Since power loss is equal to the resistance both positive and negative values, as shown in
of a conductor times the square of the current, Figure B.2(a), (c), and (d). This phase differ-
loss in a transmission line can be reduced by ence can arise, for example, because of the
increasing the transmission voltage, which allows reactance of the transmission line. Here, in
the current to be reduced for the same amount of addition to the real power that is flowing in one
power transmitted. As a result, long transmission direction, there is back and forth movement of
lines employ high voltage. However, as discussed power called reactive power. While it does no
later, high-voltage lines also have drawbacks, useful work, reactive power flow still causes
including the need to maintain larger clearances power losses in the system because current is
to maintain safety. flowing through components, such as trans-
formers and transmission lines, which have
In ac systems, where voltage and current oscillate resistance. Reactive power is measured in
many times a second, the instantaneous power volt-amperes reactive (VAR).
they produce is also rapidly varying, as shown in
Figure B.2. In the figure, negative instantaneous Reactive power can be positive or negative. But
power is equivalent to power flowing in the unlike instantaneous power, its sign does not
backwards direction. In electric power systems, it indicate the direction of reactive power flow.
is more valuable to have measures of power that Instead, the sign simply indicates the relative
are averages over many cycles. These measures phase shift between current and voltage. When
are real power, reactive power, and apparent current lags voltage due to the presence of
power. Only two of these three measures are inductive reactance, reactive power is positive,
independent; apparent power can be determined as shown in Figure B.2(a); when current leads
from real power and reactive power. voltage due to the presence of capacitive
v
It is energy that actually flows in a power system, power being the rate of this energy flow. However,
though technically incorrect, common usage is to speak of power flow.
apparent power
apparent power
real power reactive power
0 0 real power
time time
reactive power
current voltage
voltage current
(c) Current Leads Voltage (d) Current and Voltage
Out-of-Phase by 90 Degrees
reactance, reactive power is negative, as shown B.3 STRUCTURE OF THE ELECTRIC POWER
in Figure B.2(c). Equipment that draws nega- SYSTEM
tive reactive power is often said to be supply-
ing reactive power. In power systems, capacitors The electric power system consists of generat-
are often connected near large inductive loads ing units where primary energy is converted
to compensate for their positive reactive power. into electric power, transmission and distribu-
tion networks that transport this power, and
Apparent Power consumers equipment (also called loads)
Apparent power is the product of rms voltage where power is used. While originally genera-
and rms current, and is always greater than or tion, transport, and consumption of electric
equal to real and reactive power. Electrical power were local to relatively small geographic
equipment, such as transformers and transmis- regions, today these regional systems are
sion lines, must be thermally rated for the connected together by high-voltage transmis-
apparent power they process. Apparent power sion lines to form highly interconnected and
is measured in volt amperes. The ratio of real complex systems that span wide areas. This
power to apparent power is called power interconnection allows economies of scale,
factor. Utilities like to maintain a unity power better utilization of the most economical
factor as it implies that all of the power that is
flowing is doing useful work.
Appendix B 247
Figure B.3 Structure of the Electric Power System
Generating Substation
Unit
Distribution Network
Industrial Load
Residential Load
Commercial Load
Nuclear generating units use an energy conver- and peaking units. Baseload units are used to
sion process similar to thermal units, except the meet the constant, or base, power needs of the
thermal energy needed to produce steam comes system. They run continuously throughout the
from nuclear reactions. Hydro and wind year except when they have to be shut down for
generating units convert the kinetic energy of repair and maintenance. Therefore, they must
water and wind, respectively, directly into be reliable and economical to operate. Because
rotation of the electric generators mechanical of their low fuel costs, nuclear and coal plants
shaft. Solar-thermal and geothermal generating are generally used as baseload units, as are
units use the suns radiation and the Earths run-of-the-river hydroelectric plants. However,
warmth, respectively, to heat a fluid and then nuclear and coal baseload units are expensive to
follow a conversion process similar to thermal build and have slow ramp ratesthat is, their
units. Solar photovoltaic generating units are output power can be changed only slowly (on
quite different and convert the energy in solar the order of hours).
radiation directly into electrical energy. Another
common type of generating unit is the gas, or Intermediate units, also called cycling units,
combustion, turbine. These burn a pressurized operate for extended periods of time but,
mixture of natural gas and air in a jet engine unlike baseload units, not at one power con-
that drives the electric generator. Combined- tinuously. They have the ability to vary their
cycle gas turbine plants have a gas turbine and output more quickly than baseload units.
a steam turbine. They reuse the waste heat from Combined-cycle gas turbine plants and older
the gas turbine to generate steam for the steam thermal generating units generally are used as
turbine and hence achieve higher energy intermediate units.
conversion efficiencies.vi
Peaking units operate only when the system
From the operational perspective of the electric power demand is close to its peak. They have to
power system, generating units are classified be able to start and stop quickly, but they run
into three categories: baseload, intermediate, only for a small number of hours in a year. Gas
vi
Combined-cycle plants can have efficiencies in the 55%60% range, compared to about 40% for
conventional thermal plants.
Appendix B 249
turbine and hydroelectric plants with reservoirs Measurement instrumentation collects voltage,
are generally used as peaking units. Gas tur- current, and power data for monitoring, control,
bines are the least expensive to build but have and metering purposes. Communication
high operating costs. equipment transmits these data to control
centers and also allows switchgear to be con-
Large generating units generally are located trolled remotely.
outside densely populated areas, and the power
they produce has to be transported to load Since transmission networks carry power over
centers. They produce three-phase ac voltage at long distances, the voltage at which they
the level of a few to a few tens of kV. To reduce transmit power is high to reduce transmission
power losses during onward transmission, losses, limit conductor cross-sectional area, and
this voltage is immediately converted to require narrower rights-of-way for a given
a few hundred kV using a transformer. All power. However, to maintain safety, high
the generators on a single ac system are transmission voltages require good insulation
synchronized. and large clearance from the ground, trees, and
any structures. Transmission voltages vary from
In addition to the main large generating units, region to region and country to country. The
the system typically also has some distributed transmission voltages commonly (but not
generation, including combined heat and power exclusively) used in the U.S. are 138 kV, 230 kV,
units. These and other small generating units, 345 kV, 500 kV, and 765 kV.3 A voltage of 1,000
such as small hydroelectric plants, generally kV has been used on a transmission line in
operate at lower voltages and are connected at China. Although most transmission is three-
the distribution system level. Small generating phase ac, for very-long-distance transmission,
units, such as solar photovoltaic arrays, may be HVDC can be beneficial because transmission
single-phase. lines present no reactive impedance to dc.
HVDC also only requires two conductors
Transmission instead of three. However, HVDC transmission
lines require expensive converter stations
The transmission system carries electric power (utilizing power electronics technology) at
over long distances from the generating units to either end of the line to connect to the rest of
the distribution system. The transmission the ac system.
network is composed of power lines and
stations/substations. Transmission system Transformers at transmission substations
power lines, with rare exceptions, are attached convert transmission voltages down to lower
to high towers. However, in cities, where real levels to connect to the subtransmission
estate is valuable, transmission lines are some- network or directly to the distribution network.
times made up of insulated cables buried Subtransmission carries power over shorter
underground. Stations and substations house distances than transmission and is typically
transformers, switchgear, measurement instru- used to connect the transmission network to
mentation, and communication equipment. multiple nearby relatively small distribution
Transformers are used to change the level of the networks. In the U.S., the commonly used
transmission voltage. Switchgear includes subtransmission voltages are 69 kV and 115 kV.
circuit breakers and other types of switches
used to disconnect parts of the transmission
network for system protection or maintenance.
vii
The power flow through a transmission line is roughly proportional to the phase difference between the
voltages at its ends and inversely proportional to its impedance.
viii
Power system stability limits also are discussed in Box 2.3 in Chapter 2.
Appendix B 251
BOX B.1 CONTROLLING POWER FLOW As a function of time, the voltages at the ends
Two factors determine power flow: the imped- of a transmission line are sinusoidal in shape.
ance of a line and the difference in the instanta- In the figure below, the two sinusoids represent
neous voltages at its two ends. Impedance is voltages at opposite ends of a line. When there
the combination of resistance and reactance. is power flow, the instantaneous values of
Resistance accounts for energy that is lost as voltage at the two ends of the line are different,
heat in the line. It is analogous to the physical as shown by the difference in voltages (V1 and
resistance exerted by water on a swimmer or V2) at time (t) in the figure. This instantaneous
wind on a cyclist. Energy lost in this way can difference is a function of the difference in
never be recovered. Reactance accounts for phase angle between the two sinusoids. The
energy associated with the electric and mag- phase angle difference is shown in the figure
netic fields around the line. This energy is as !. If the two voltages are in phase, that is,
analogous to the potential energy stored when if ! = 0, then there will be no difference in their
riding a bicycle up a hill. It is recovered (in the instantaneous values.
ideal case) when going down the other side. In The power flow on a line varies directly with the
an alternating current (ac) line in the U.S., this phase angle difference (or more precisely the
energy is stored and recovered 120 times per sine of the phase angle difference) and inversely
second, and thus is quite different from the with the lines impedance. Except in very special
behavior of energy stored in devices such as cases in which devices are used to control
batteries. The resistance of a line is determined power flow on individual lines, the flow of power
by the material properties, length, and cross- in a line is difficult to control when the line is
section of the conductor, while reactance is part of an interconnected network since the
determined by geometric properties (the position characteristics of the entire network collectively
of conductors relative to each other and ground). determine power flows. When special devices
In practical transmission lines, resistance is small are used to control power flow, they do so by
compared to reactance, and thus reactance has modifying impedance and phase angle.
more influence on power flow than resistance.
Phase Angle Difference (!) of Voltage Sinusoids at the Ends of a Transmission Line
V1
V2
Voltage
t
Time
Appendix B 253
Figure B.5 A Load Duration Curve
10
Power (Gigawatts)
0
0 8,760
Time (hours)
largest load and continuing in a monotonically must be balanced at all times. Since the load is
decreasing fashion, as shown in Figure B.5. For changing all the time in ways that cannot be
each point on this curve, the horizontal coordi- perfectly predicted, generation must follow the
nate is the number of hours in the year for load in real time. The balance between supply
which the load is above the power given by the and demand is maintained using a hierarchical
vertical coordinate. The load duration curve control scheme, with crude matching at the
provides a good picture of how widely the load longer timescale and finer matching at the
varies and for how many hours in a year it is shortest timescale (see Figure 2.1 in Chapter 2).5
above a particular level. It is more expensive to
meet the needs of a spiked load duration curve Protection
than a flat one, as generation capacity to meet
the peak load is needed, while the generations An important aspect of the operation of the
utilization is related to the average load. One electric power system is protection. This means
useful metric of power consumption is the load ensuring the safety of the system, including
factor, which is the ratio of average to peak load. generating units and other grid assets, and the
people who may come in contact with the
B.4 OPERATION OF THE ELECTRIC system. Protective action must be taken in
POWER SYSTEM fractions of a second to avoid equipment
damage and human injury. Protection is
The electric power system is operated through a achieved using sensing equipment as well as
combination of automated control and actions circuit breakers and other types of switches that
that require direct human (system operator) can disconnect and de-energize parts of the
intervention. The main challenge in operating system in the case of a fault, such as a damaged
the electric power system is that there is negli- transmission line or a short circuit. Once the
gible electrical storage in the system.ix Hence, fault is repaired, that segment of the system can
supply and consumption of electrical power be brought back online.
ix
Note that pumped storage, which uses electricity to pump water into an elevated reservoir and stores energy
in the form of potential energy, is not a form of electrical storage. A hydroelectric generating unit must be
run to convert this energy back into electrical form. Energy storage technologies are discussed in Chapter 3.
x
The generators output frequency is proportional to its rotational speed, and traditionally governors have
been designed to sense this speed.
Appendix B 255
without a commensurate increase in reactive system operator can call the generation opera-
power supply, the output voltage of the genera- tor over the phone and ask for an increase or
tor will decrease. Conversely, the output voltage decrease in output.
of the generator will increase if the generator is
supplying more reactive power than is being Reserves
drawn. The voltage can be restored to its Beyond a certain level of power imbalance,
original level by either adjusting the generators system operators need to call in generation
rotor current (which controls the amount of reserves. These may be additional generating
reactive power produced by the generator), or units that are on standby or generators that are
by using ancillary voltage support equipment, already producing power but can ramp up their
such as static VAR compensators that employ output on request. Having adequate reserves on
inductors and capacitors in conjunction with the system is essential to deal with load uncer-
semiconductor switches to absorb or supply the tainties and contingencies, such as the failure of
imbalance in reactive power. Voltage control is a generating unit.
also extremely fast.
Reserves are categorized based on the time it
Automatic Generation Control takes them to start delivering the requested
While governor control brings supply of and power; typical categories are 10-minute and
demand for real power in balance, it results in a 30-minute reserves. Reserves can be either
small change in system frequency. Furthermore, spinning or non-spinning. Spinning reserves
governor-based reaction of generators located are generating units with turbines spinning in
outside a control area to load changes inside the synchronicity with the grids frequency without
control area (or vice versa) can alter power supplying power. They can deliver the requested
flows between control areas from their sched- power within a few minutes. Non-spinning
uled levels.xi The errors in frequency and flows reserves are units that are offline but also can be
between control areas are corrected by the synchronized with the grid quickly. In systems
relatively slower AGC. AGC aims to eliminate with organized markets, reserves are paid not
the area control error (ACE). ACE is a measure only for the energy they produce but also for
of both the difference between actual and being available on short notice to deliver
scheduled net power flows to or from a control reserve power.
area and the error in system frequency. Ignor-
ing the effect of system frequency, a positive Other Power Balancing Options
ACE means that generation within the area Large customers in some regions often face
exceeds load by more than the scheduled net real-time pricing, which induces them to cut
power flow from the control area. In this case, loads when the system is under stress and the
the generation in the control area needs to be real-time incremental cost of supplying power
reduced. Conversely, negative ACE requires is accordingly high. However, when all other
local generation to be increased. The area options for balancing power have been
control center automatically sends signals to exhausted, the system operator must resort to
generators equipped with AGC to increase or proactively reducing the load, generally referred
decrease their output. In exceptional circum- to as load shedding. Load shedding can be
stances, when the required change in output is accomplished in a number ways. At first the
greater than the defined limit of AGC, the system operator can interrupt power to those
xi
From an operational perspective, a large electric power system is divided into multiple control areas, also
called balancing authority areas. These control areas are connected together via transmission lines that
are called tie-lines.
Appendix B 257
B.5 WHOLESALE ELECTRICITY MARKETS energy in the market.) All the dispatched
generators receive the same compensation,
The organizational structure of the electric called the clearing pricethe offer of the
power industry has changed significantly over last generator dispatched. The actual process is
the last 15 years, as discussed in Chapter 1. more complicated than this simple explanation,
Until the mid-1990s, the electric power indus- incorporating such parameters as the time
try in the U.S. mostly was vertically integrated: required to start generators, out-of-economic-
a single entity, a regulated monopoly, owned order dispatch due to congestion or reliability
and operated generation, transmission, and concerns, and security constraints. The goal of
distribution in each region.xii However, in 1996 the system operator is to determine the dispatch
the Federal Energy Regulatory Commission that minimizes total cost, as measured by
issued Order No. 888, which required that the generators bids, subject to security constraints.
transmission network be made available for use
by any generator. Since then independent This process determines the marginal cost of
system operators (ISOs) and regional transmis- meeting an increment of load at each location
sion organizations (RTOs) have been created (called a node) in the transmission system to
in certain parts of the US. In many regions, which load or generation is connected. These
ownership of generation and transmission have costs are termed locational marginal prices
been separated. In regions where they exist, (LMPs) and are the prices at which transactions
ISOs and RTOs coordinate organized wholesale for purchasing or selling energy in the market
electricity markets in which independent take place. Distribution companies or large
decisions of market participants (those who customers pay the applicable LMP for energy
buy and sell energy or other electricity market consumed. Similarly, generation is paid the
products, such as spinning reserves) set the LMP at the point at which it is located.
price of energy generation, respecting the
requirements of central coordination provided The LMP pricing structure used in modern
by the ISO or RTO. markets ensures that the profitable choice for
generators and loads is to follow the instructions
The theory of spot pricing provides the of the economic dispatch. Generators are only
foundations for successful market design.6 dispatched when their offer to sell is at a price no
In a framework known as bid-based, security- greater than the market-clearing price at their
constrained, economic dispatch, central location. Likewise, generators are not dispatched
coordination by the system operator is when the market price is less than their offer to
integrated with decentralized decisions by sell. The use of LMPs allows for the preservation
market participants. The process of selling of the traditional industry approach of security-
wholesale energy begins with a bidding process constrained, economic dispatch in the presence
whereby generators offer an amount of energy of independent system operators and organized
for sale during specific periods of the day next wholesale markets. The use of LMPs exploits
day at a specific price. These offers are arranged the natural definition of an efficient equilibrium
by the ISO/RTO in ascending order, called the for a market, utilizes the unavoidable central
bid stack, and the generators are dispatched coordination, and avoids the need for market
(told to generate) in this order until generation participants to track transmission flows or
matches expected load. (Large loads also understand the many constraints and require-
sometimes submit bids for the purchase of ments of the power system.
xii
The exceptions were small municipal and cooperative entities that were distribution-only operations and,
particularly from the 1930s on, federal systems such as the Tennessee Valley Authority.
In regions served by vertically integrated utilities, An added complexity in planning future trans-
generation and transmission expansion planning mission expansion in areas with organized
is carried out centrally by system planners at the markets is the uncertainty associated with future
utilities. Planners evaluate various options for generation investments. Transmission expansion
meeting future load demand in terms of capital decisions are more challenging for ISOs/RTOs
and operating costs. They select projects based because development of generating plants is
on minimizing system cost while providing based on individual company decisions. As a
adequately reliable service. Decisions also may be result, the ISOs/RTOs cannot know the location
influenced by government incentives, regula- and size of these future plants with certainty.
tions, and environmental impact restrictions.
Planning has to allow for the risk associated with
the significant uncertainty in long-term load
forecasts, future operating costs (directly related
to fuel prices), and technological changes.
Appendix B 259
REFERENCES I. J. Prez-Arriaga, H. Rudnick, and M. Rivier,
5
2
W. Steinhurst, The Electric Industry at a Glance R. E. Bohn, Spot Pricing of Electricity (Boston, MA:
(Silver Spring, MD: National Regulatory Research Kluwer Academic Publishers, 1988).
Institute, 2008).
3
S. W. Blume, Electric Power System Basics: For the
Nonelectrical Professional (Hoboken, NJ: Wiley
IEEE Press, 2007).
4
A. V. Meier, Electric Power Systems: A Conceptual
Introduction (Hoboken, NJ: WileyIEEE Press,
2006).
admittance The ratio of current to voltage, including the effects of both resistance and
reactance; the inverse of impedance.
advanced metering A system for measuring individual customers electricity consumption at intervals
infrastructure (AMI) of an hour or less and communicating that information at frequent intervals to
the distribution utility.
alternating current An electric current that reverses direction at regular intervals and is the dominant
(ac) form of electric power in transmission and distribution systems worldwide.
ampere A measure of the amount of electric charge passing a point in an electric circuit
per unit time.
ancillary services Services, such as spinning reserves, non-spinning reserves, and regulation, that
support the transmission of energy from generating resources to loads while
maintaining reliable operation of the network.
attack vector A path or means by which an attack can be or is made on critical infrastructure.
automatic generation An automatic system to vary mechanical input to a generator to match small
control (AGC) variations in system load.
balancing authority An entity responsible for balancing generation and load (with specified imports
and exports) within a specified geographic region.
bandwidth Broadly, the amount of information that can be communicated through a given
communications channel per unit time. Alternatively, the range of radio
frequencies in a given radio channel (spectrum).
battery electric vehicle A vehicle that operates solely with electric power provided by batteries.
(BEV)
bulk power system That part of the electric grid comprised of generators and high-voltage
transmission lines.
capacitance A parameter relating the charge stored in an electric field to the voltage producing
the field. Transmission lines have capacitance because their voltage creates electric
fields between conductors and between conductors and the ground.
capacitor An element exhibiting capacitance.
capacity market A wholesale forward market for resources to supply energy. These capacity
resources are usually, but not always, generators. See capacity market demand
response programs.
capacity market Wholesale forward market programs in which customers bid future load
demand response reductions as system capacity to replace procurement of conventional generation
programs or delivery resources, usually in exchange for upfront capacity payments.
congestion A condition that occurs when lack of transmission capacity prevents the least-cost
set of generators from serving load, causing an increase in the wholesale price of
electricity or cost of service at one or more locations in the system.
contingency An abnormal event in the power system, such as the tripping of a generator or
a transmission line.
converter A generic term referring to a system employing power electronics to convert
electrical energy from one form to another, e.g., from direct current at one voltage
to direct current at another voltage or alternating current at one frequency to
direct current or to alternating current at another frequency.
Glossary 261
critical peak pricing A dynamic pricing plan that combines peak/off-peak time-of-use rates with
substantially higher super-peak rates that apply only to peak hours on a limited
number of critical days during the year. Critical days typically are announced the
day before, on the basis of forecast market conditions.
current The amount of electric charge flowing past a specified circuit point per unit
of time.
demand response Customer loads that are responsive to conditions in the electric power system,
particularly at peak times.
direct current (dc) An electric current that flows in one direction and is used selectively in electric
power systems, primarily for point-to-point applications.
distributed generation Small-scale, on-site generation systems owned by entities that are primarily
(DG) consumers of electricity.
distribution The application of advanced technology to automate the maintenance, control,
automation and operation of the distribution network.
distribution primary The voltage at which power is distributed before the final step-down transformer
voltage to customer delivery voltage (typically 13.8 kilovolts, but can range from as low as
2 kilovolts to as high as 34.5 kilovolts).
distribution system The part of the power system that delivers electricity to customers, operating at
lower voltages than the transmission system.
dynamic line rating Line rating determined by the current ambient conditions, such as temperature
(DLR) and wind speed.
dynamic pricing A regime in which retail customers face energy prices that vary with the
contemporaneous cost of generation or state of supply-and-demand conditions
in the electric power system. Prices may be based on day-ahead or hour-ahead
forecasts of conditions, and may change for as few as 60 critical peak hours per
year, or may change hourly or more often in real-time pricing plans.
Eastern One of the two major synchronized alternating current power grids in North
Interconnection America, reaching from Central Canada eastward to the Atlantic coast (excluding
Qubec), south to Florida, and back west to the foot of the Rockies (excluding
most of Texas).
economic dispatch The assignment of generating units production in order to minimize overall
costs.
electric vehicle (EV) A vehicle that operates with electric power provided by batteries. EVs include
both plug-in hybrid electric vehicles and battery electric vehicles but do not
include hybrid electric vehicles, which are self-powered and never connected to
the electric grid.
Electric Reliability Synchronized alternating current power grid that occupies nearly all the state of
Council of Texas Texas.
(ERCOT)
extra-high voltage Transmission voltages between about 345 kilovolts and 765 kilovolts.
fault On a transmission or distribution line, an abnormal flow of electric current, e.g.,
an open circuit (an interruption in the flow) or a short circuit (a flow that bypasses
the normal load).
fault current limiter A device that limits line current from faults to some pre-determined level.
(FCL)
Federal Energy U.S. independent agency that: regulates the interstate transmission of electricity,
Regulatory natural gas, and oil; reviews proposals to build liquefied natural gas terminals and
Commission (FERC) interstate natural gas pipelines; licenses hydropower projects; and performs some
other related activities.
feed-in tariff A fixed price paid for electricity generated from specified renewable technologies.
Glossary 263
losses The difference between generated power and power delivered to the load,
typically caused by resistance in transmission lines and transformers and
converted to waste heat.
low-voltage ride The ability to maintain system operations and integrity despite a low-voltage
through event, principally due to a short-circuit fault.
microgrid A part of an electric power system consisting of distributed generators, loads, and
specialized controls that is capable of operating either in parallel with a utility
system or as a stand-alone system.
N-1 contingency Evaluation of the transmission line and transformer power flows and bus voltages
analysis in case of the loss of a single component, such as a particular generator.
phase angle The time, expressed as an angle, by which a voltage and current waveform, or two
voltage or two current waveforms, are shifted relative to each other.
phasor A mathematical concept used to represent a sinusoidal wave as a magnitude and
phase angle, where frequency is implicit. Voltage and current waves on the power
system are sometimes expressed as voltage and current phasors since their
frequency is constant (60 hertz in North America).
phasor measurement A device used to measure current, voltage, and frequency every 1/30th of a second
unit (PMU) or faster in synchronicity with other such measurements across a wide area based
on a Global Positioning System time signal.
plug-in hybrid electric A vehicle with an internal combustion engine as well as batteries that can be
vehicle (PHEV) charged using an external power source.
power The rate at which energy is flowing.
power electronics Electronic circuits, employing switching electronic semiconductor devices, whose
function is to control electrical energy and convert it from one form to another,
e.g., from alternating current to direct current, or alternating current at one
frequency to alternating current at another frequency.
power factor The ratio of real power to apparent power. Reflects the degree to which a given
amount of current is producing useful work.
power quality The extent to which the voltage waveform at a load conforms to the ideal
sinusoidal shape and nominal value. Poor power quality is generally the result
of loads that draw current that is not sinusoidal (a particular problem with
electronically controlled loads) or weak distribution networks producing frequent
outages or voltage sags.
price responsive Load that responds to prices that vary with system supply-and-demand
demand conditions.
public utility A state agency typically responsible for regulating retail electric rates and other
commission utility prices.
reactance The property of a conducting device that introduces a phase shift between voltage
and current and introduces an impediment to the flow of alternating current.
reactive power Power that exists in ac power systems when reactance is present. Reactive
power charges and discharges the energy stored in reactive elements. It does no
time-average work, but its presence still contributes to electrical losses and
voltage drops.
real-time pricing See dynamic pricing.
regional transmission An independent system operator (ISO) that the Federal Energy Regulatory
organization (RTO) Commission has certified to have satisfied a specified set of requirements and that
has slightly greater responsibilities for system reliability than ISOs that have not
been so certified.
Glossary 265
volt (V) Unit of electric potential and electromotive force, equal to the difference of
electric potential between two points on a conducting wire carrying a constant
current of one ampere when the power dissipated between the points is one watt;
roughly analogous to water pressure in a pipe.
volt ampere (VA) A measure of apparent power that defines the capacity of equipment, such as
transformers or generators, that is limited in voltage and current. It combines
both real (time average) and reactive power components.
volt-ampere reactive The unit used to measure reactive power, which is present in an ac system when
(VAR) current and voltage are out of phase.
voltage The value of electromotive force or potential difference, expressed in units of volts.
voltage source A power electronic device for converting a direct current voltage to an alternating
convertor (VSC) current voltage.
watt (W) The standard unit of electric power, the rate at which work is done when one
ampere of current flows through an electrical potential difference of one volt.
watt-hour A unit of electrical energy equal to 3,600 joules.
Western One of the two major synchronized alternating current power grids in North
Interconnection America. It stretches from Western Canada south to Baja California in Mexico,
reaching eastward to just over the Rockies into the Great Plains.
wide-area A network of devices, usually consisting of phasor measurement units, that
measurement systems measures quantities of interest on the transmission network across a large
(WAMS) geographic area in real time.