Sioshansi F.P. Smart Grid: Integrating Renewable, Distributed & Efficient Energy
Подождите немного. Документ загружается.
$30,000 to $39,999 13.6 12.2% 11,431 $1,143 3.3%
$40,000 to $49,999 11.0 9.9% 11,658 $1,166 2.6%
$50,000 to $74,999 19.8 17.8% 12,440 $1,244 2.0%
$75,000 to $99,999 10.6 9.5% 13,559 $1,356 1.5%
$100,000 or More 14.2 12.8% 15,382 $1,538 1.5%
Total 111.1 100%
Source: 2005 Residential Energy Consumption Survey, Energy Information
Agency
This chapter begins by describing and characterizing smart
grid, an elusive term that perhaps defies precise definition.
Some have approached this definitional challenge not by
describing the technologies that smart grid encompasses, but
by describing the desired attributes, as Hauser and Crandall
do in Chapter 1 of this volume. If smart grid means a
generation, transmission, and distribution system that is more
intelligent than today's because it takes advantage of the
leapfrog improvements in informational and computational
technologies, it is a useful term. If the term is used to set up a
false choice between “smart” and “dumb,” rather than to be
informative, it is pejorative and brings discussion to a halt.
Next, important questions related to smart grid underscore
that smart grid technologies and policies raise serious and
legitimate concerns. Some have argued that dynamic pricing,
which smart meters could enable, is more equitable than
current flat rates because low-income customers would have
lower electricity bills (e.g., as argued by Faruqui in Chapter
3). Others question whether residential ratepayers would
benefit from advance metering infrastructure [4], [5] and [6].
To what extent, if at all, equity analysis of dynamic pricing
bears on smart grid more generally is open for discussion.
There is nothing close to a majority view, let alone a near
consensus, that these technologies should be adopted. As one
251
report concluded: “The mindset of state regulators regarding
what constitutes prudent and value-adding smart grid
investments is currently in flux”[7]. Even EPRI [1, p. 4–1]
acknowledges that estimates of the benefits of smart grid are
questionable:
There have been a number of studies
which have estimated some of the benefits
of a Smartz Grid. Each varies somewhat
in their approach and the attributes of the
Smart Grid they include. None provides a
comprehensive and rigorous analysis of
the possible benefits of a fully functional
Smart Grid. EPRI intends to conduct such
a study, but it is outside the scope of the
effort presented in this report.
Note that EPRI's report, which does not contain a
“comprehensive and rigorous analysis” of smart grid, is a
report on the costs and benefits of smart grid. The claim that
smart grid would provide a net positive value to society is far
from certain, and society needs to honestly assess these costs
and benefits and not be “wowed” by the prospects of smart
grid [6]. Even if correct, this claim does not mean smart grid
would be equitable under any definitions of equity.
The chapter then turns to discussing different conceptions of
equity in the context of smart grid investments and policies.
There are multiple definitions of equity, and no one definition
is acceptable to all. A complete analysis of smart-grid-related
equity concerns, therefore, requires analyzing different equity
definitions and considerations. Understanding the conditions
under which multiple definitions may lead to the same
252
conclusion is extremely useful because it provides an opening
to move forward [2]. The chapter concludes with a discussion
of what types of questions and analyses regulators should ask
and require in order to evaluate the many issues that smart
grid raises, particularly those with equity implications.
Smart Grid Is Not Consistently Defined, Which
Leads to a Mismatch of Its Costs and Benefits
The most revealing quip about smart grid is this: “Smart grid
is the best thing ever, although we do not know what it is.”
Others have noted, perhaps wryly, that smart grid is
“variously defined”[8]. It seems as if every report has its own
definition. Beginning with the more specific and moving to
the more general, here are a few:
1. “The smart grid is a digital network that unites electrical
providers, power-delivery systems and customers, and
allows two-way communication between the utility and its
customers”[9].
2. “A smart grid is the electricity delivery system (from
point of generation to point of consumption) integrated
with communications and information technology for
enhanced grid operations, customer services, and
environmental benefits” (U.S. Department of Energy,
undated).
3. “Smart grids are electricity networks that can
intelligently integrate the behavior and actions of all users
connected to it—generators, consumers, and those that do
both—in order to efficiently deliver sustainable, economic
and secure electricity supplies”[10].
253
4. “Smart grid is an enhanced electric transmission or
distribution network that extensively utilizes internet-like
communications network technology, distributed
computing and associated sensors and software (including
equipment installed on the premises of an electric
customer) to provide
i. smart metering;
ii. demand response;
iii. distributed generation management;
iv. electrical storage management;
v. thermal storage management;
vi. transmission management;
vii. power outage and restoration detection;
viii. power quality management;
ix. preventive maintenance improves the reliability,
security and efficiency of the distribution grid;
x. distribution automation; or
xi. other facilities, equipment, or systems that operate in
conjunction with such communications network, or that
directly interface with the electric utility transmission or
distribution network, to provide the capabilities
described in clauses (i) through (x) in paragraph
(A)”[11].
The first definition above includes only the digital
communication network; the second adds in the electric
delivery network including generation; the third adds user
behaviors; and the fourth includes everything but the
254
proverbial kitchen sink. Figure 4.1 illustrates these
ever-expanding definitions.
Figure 4.1
Diagram of ever-broadening definitions of smart grid.
It is beyond the scope of this chapter to resolve the smart grid
definitional problem, and there are many more variations of
its definition not provided. The point here is that if one is not
extremely careful about what smart grid does and does not
include, it is impossible to assess its benefits, costs, and
equity implications. The real danger is that broader
definitions are used when discussing benefits, and narrower
ones are used when discussing costs. For instance, much is
made of smart grid's environmental and sustainability benefits
(see Chapter 5, Chapter 9, Chapter 10 and Chapter 18 in this
text), but much of the costs of achieving those benefits are not
included in smart grid cost-benefit analyses [1]. But at best,
smart grid enables, and in reality may only facilitate, the use
of renewable energy resources. Assuming that smart grid is an
255
enabler, the costs of renewable resources, electric vehicles,
and thermal and electric storage, which are very expensive,
must be considered in the cost-benefit analysis of smart grid.
If advocates of smart grid are going to claim the benefits of
smart grid from these additional technologies, they must
explicitly include their additional costs, which as noted, as
noted above, is not always done.
At worst, smart grid may not be necessary to increase the
penetration of cleaner energy technologies. For instance,
large-scale wind farms and large numbers of solar facilities
exist today without smart grid. A major expansion of nuclear
power or carbon capture and sequestration facilities does not
require smart grid at all. If this is the case, an analysis is
needed to determine the extent to which smart grid facilitates
cleaner energy technologies. This could be accomplished by
comparing the deployment of renewable resources, electric
vehicles, and grid storage with and without compulsory smart
grid so that the incremental benefits and costs of smart grid
technologies can be isolated.
This “but for” analysis must also incorporate the fact that
smart grid technologies will be naturally adopted by utilities
over time as existing equipment wears out and the costs of
smart grid technologies decline. For instance, the compulsory
purchase of smart meters by customers should be compared to
what would happen if smart meters were not made
compulsory. At some point in the future, smart meters would
be installed as replacement meters for existing meters that fail
or as new meters for new customers, and if and when the cost
savings to utilities of smart meters outweigh their installed
cost. In presenting the cost-benefit analysis of smart grid,
multiple deployment options must be considered, from
256
naturally occurring smart grid to accelerated deployment in
order to obtain a critical mass. EPRI raises the latter issue but
does not provide a comparison of costs and benefits under
multiple deployment scenarios [1, pp. 2–6]. In fact, it
acknowledges that the role of critical mass, tipping point, and
penetration of implementation in affecting smart grid costs
and benefits is unclear [1, pp. 2–6]. Assuming, which has yet
to be demonstrated, that a rapid and compulsory deployment
is preferable from a cost-benefit perspective to a naturally
occurring one, it is not clear that the additional savings of the
former should outweigh the non-compulsory advantages of
the latter. In fact, several authors have noted the importance
of obtaining customer buy-in, particularly with smart meters,
if smart grid is to be successful [6] and [30]. Rushing
deployment may undercut public support.
Furthermore, policy changes may also be required to achieve
many of the benefits of smart grid investments. For instance,
if regulators do not permit or restrict dynamic pricing or
customers do not respond as anticipated to dynamic pricing,
then many of the benefits of smart meters will not materialize
[12]. Some regulators are reluctant to approve dynamic
pricing because of equity concerns based upon positions taken
by consumer advocates [4] and [6]. If transmission expansion
to support large-scale wind does not occur, either due to
problems with siting or determination of how costs are to
allocated “both perennial problems in the United States” then
many of the renewable energy benefits that smart grid is
supposed to enable may not occur.
There are other fundamental questions related to smart grid
that must be addressed. The U.S. Department of Energy, for
example, has issued three requests for information listing
257
dozens of fundamental questions about smart grid for
stakeholders to address [13]. These questions cover
substantive and major issues such as whether it is preferable
to wait before adopting major smart grid investments to learn
more about the technologies and costs, how to calculate costs
and benefits, how to determine who pays what, what should
be done if there are cost overruns or benefits do not
materialize, how data privacy and cyber security will be
ensured (see [14]), and the list goes on. The point here is not
to answer these questions, but to suggest that regulators
should not consider the compulsory adoption of smart grid
paid for by ratepayers until these questions have been
coherently addressed.
Take the case of smart meters, a maturing technology, but
perhaps one of the most misunderstood in the suite of smart
grid technologies. With both costs and need uncertain, and
with the prospect of further cost reductions and technological
improvements, a compulsory approach to the adoption of
smart meters becomes questionable. Only about 6% of
electric meters are smart, so our experience is still limited. As
with most relatively new technologies, further cost reductions
are expected [7]. Before mandating the compulsory purchase
of smart meters by ratepayers, regulators should require
utilities to evaluate when, if at all, in the future would the
utility would be willing to pay for meter replacement because
smart meters save more money for the utility than they cost.
Recall that many of the benefits of smart meters—for
example, reduced maintenance, billing, and customer service
costs—benefit the utility. Compulsion has an equity cost.
Granted, while hard to quanify it is nonetheless a cost; but it
is assumed to be zero when analyzing smart meter and other
smart grid deployments. With approximately 60 million smart
258
meters planning to be installed in the United States [15],
regulators need to start addressing these questions before it is
too late.
But there is even a more fundamental question: some analysts
question whether the smart meter should be the gateway to
the home for smart grid technologies [7] and [13]. Perhaps the
meter can be directly bypassed and communication with the
grid can be done directly with smart appliances. Perhaps a
component can be added to existing meters to make them
functionally smart. If so, why (subject to obvious safety
issues) should that component be owned by the utility? Also,
why must a customer use a utility-owned communication
network versus the internet?
The above analysis also applies more generally to smart grid
investments. Assume that the above definitional and
uncertainty issues can be worked through—along with data
privacy and security issues, no small challenges—but
advocates of smart grid are misapplying cost-benefit analysis.
The decision rule employed by advocates of compulsory
smart grid investments is that so long as the benefit-cost ratio
is greater than 1, or the net present value is greater than zero,
then compulsory investments should proceed. This is
incorrect because it implicitly assumes that the status quo is
static and smart grid would not occur on its own in the future.
It may be the case that the natural penetration of smart grid
technologies, albeit at a slow pace, has a smaller benefit-cost
ratio than a compulsory and accelerated approach. Unless this
comparison is made, regulators will not be completely
informed. But even if the compulsory approach yields a
greater benefit-cost ratio than the naturally occurring
approach by saving on rollout costs, regulators may
259
reasonably conclude that the additional expected benefits of
the compulsory approach do not outweigh the equity costs of
compulsion, which are not incorporated into a benefit-cost
analysis. Or regulators may want to consider policies that
protect consumers, particularly low-income families, when
authorizing smart grid investments. Where exactly to draw
that line is not clear, but that should not mean that utilities
and regulators ignore the issue.
But there is still another major issue: even if a cost-benefit
analysis is performed and used correctly, it may still be
factually wrong, that is, in reality the investment may not turn
out as expected. At this point, some history is in order. The
U.S. electric utility industry does not have a perfect track
record with respect to cost effectiveness and completion of
long-term investments. Two examples come immediately to
mind: nuclear power and long-term power contracts under the
Public Utility Regulatory Policies Act of 1978 (PURPA). In
both cases, ratepayers ended up being stuck paying for much,
if not most, of the cost overruns and out-of-market or
stranded costs. Estimates of stranded costs vary, but most
range from $100 to $200 billion, in late 1990s dollars [16].
Of course, just because past investments did not always work
out does not automatically mean smart grid investments will
not do so. But the replacement of existing, working meters
with smart meters raises questions about stranded costs. In
fact, one utility concedes that its plan to install meters will
result in $42 million of stranded costs if 600,000 smart meters
are deployed [17]. It is also interesting to compare the cost
estimates of smart grid over time. In 2004, EPRI estimated
the cost of smart grid at $165 billion (pp. 1–2); seven years
later, the cost is between $338 to $476 billion [1, p. 1–4].
260