|
It's
no secret that distributed generation technologies have become
more efficient and less costly. As a result, the
potential of distributed generation (DG) to provide cost-effective
alternatives to central-station generating facilities and
traditional "poles and wires" investments for electric distribution
has increased. But despite the loudest voices of DG's greatest
proponents-manufacturers and, in some cases, utility regulators-DG
isn't viable in all situations, and rumors of the death of
traditional utility investments are greatly exaggerated.
DG technologies, especially
fuel cells, are receiving increasing attention in the popular
press as the "Holy Grail" of distributed generating technology.
For example, in "Dreams of the New Power Grid," which appeared
in the March 2002 issue of Popular Science, the goal is fuel
cells in every home. That article quotes one fuel cell proponent
who compares price projections for fuel cells to the rapid
drops in prices for VCRs. Alas, images of microturbines and
fuel cells humming away quietly and reliably in millions of
basements remain just that: images. While DG has filled certain
niche applications well, some of the promised technological
breakthroughs have not materialized as quickly as expected.
Furthermore, like the proverbial tortoise, generation manufacturers
have continued to improve the efficiency and cost-effectiveness
of traditional generation technologies.
So what role can
DG play in the future? Answering that question is critical
for utilities, regulators, and especially consumers, who will
continue to demand greater quantities of electricity. Utilities,
developers, regulators, and consumers all have an interest
in ensuring that the best applications of distributed resources
are identified, lest DG be over-sold and have its real promise
squandered.
A number of regulatory
initiatives establishing DG programs have begun in states
as diverse as California and Vermont. In the fall of 2001,
the New York State Public Service Commission (NYPSC) issued
an order requiring the electric utilities it regulates to
participate in a pilot program designed to test the applicability
of DG alternatives.1 The Vermont Department of Public
Service is currently engaged in a collaborative program with
the state's electric utilities to develop standards for evaluating
and installing DG. Indiana's regulators also have begun a
similar process to investigate the role of DG.
All of these efforts
will need to address a variety of environmental, reliability,
and safety issues. The fundamental test of DG, however, ought
to be its impact on the bottom line: will it increase utility
costs and lead to higher customer rates?
Through case studies
and the development of an economic model that evaluates DG
alternatives using advanced investment analysis techniques,
we have discovered several critical issues surrounding the
economics of DG. Our studies have specifically incorporated
future uncertainties concerning market prices, operating costs,
and load growth, as well as reliability consequences. We believe
that successful DG applications can be identified only after
analyzing these critical issues. Otherwise, regulators and
utilities risk unpleasant surprises.
The Danger of Hyperbole
There has been no
shortage of mythology developed about the applicability of
DG technologies. As has been observed with many previous emerging
technologies, the claims almost surely exceed what DG applications
can possibly provide. The danger of such "irrational exuberance"
is, of course, unmet expectations: if DG is too broadly promoted
as an alternative to central-station generating supplies and
traditional transmission and distribution (T&D) capacity investments,
the entire concept cannot help but fail. DG will be installed
where it ought not to be, and fail to provide the hoped for
advantages to utilities and their customers.
We have found four
common myths pervading DG and its application:
- DG is cheaper
than traditional system power;
- DG makes sense
because traditional electric utilities will be obsolete
in the not-too-distant future;
- DG can defer
traditional T&D system investments; and
- Even if it
is more expensive today, DG investments should be emphasized
because T&D investments are likely to become stranded
in the (not-too-distant) future.
Myth 1: DG is cheap
Cheap. Modular. Who
could ask for anything more? Unfortunately, it's not quite
true. The most common commercially available DG technologies
are either simple-cycle turbines or diesel generators. These
tend to have higher capital costs and higher operating costs
than central-station alternatives, owing to diseconomies of
scale and higher heat rates. And, while the fuel cell Holy
Grail continues to improve, it has yet to become commercially
viable.
It is true that DG
applications may be able to avoid some T&D costs at the margin,
including system losses. But installing DG can just as easily
require additional T&D costs, such as more sophisticated monitoring,
switching, and safety systems. Furthermore, the environmental
impacts of fossil-fueled DG technologies may be higher than
those associated with central-station generation, precisely
because DG is designed to be installed near customer loads,
where more individuals can be affected by pollutant emissions.2
That is why we sometimes observe local environmental regulators
adamantly opposing the very DG installations that utility
regulators are promoting-an untenable situation for any electric
utility.
Myth 2: Traditional
electric utilities are soon to be obsolete
This myth may have
been divined by the same pundits who predicted that Enron
would become the world's dominant electric, gas, water, and
broadband company. At the very least, many utilities would
likely take exception to such pronouncements of their imminent
demise, as they continue to provide safe and highly reliable
electric supplies. DG technologies may play an increasingly
important role in helping utilities meet customers' differing
needs for power quality and reliability, but it is not at
all clear that many customers will want to be in the electric
generation business. Furthermore, the "generator-in-a-box"
technology for the basement of every house, which will provide
electricity cost-effectively and with high reliability like
today's water heaters and furnaces, does not yet exist.
Myth 3: DG can
defer traditional T&D investments
Of all of the myths
claimed for DG, its ability to defer traditional "poles and
wires" investments is probably the most cited. The argument
goes as follows: by installing DG, utilities can avoid the
need to upgrade substations and circuits, thereby saving themselves
and their customers millions of dollars. But while DG may
defer the need for some system upgrades, such deferral should
be seen as a consequence of installing DG, not a goal.
The goal should be
to reduce utility and ratepayer costs without sacrificing
reliability and power quality. A moment's thought shows that,
from a strict economic standpoint, introducing DG applications
to further a goal of deferring T&D investments as long as
possible will not result in any savings for utilities or their
customers. This is precisely the same flawed economic reasoning
that many regulators used to promote demand-side management
(DSM). By using the same "avoided cost" methods as was done
for DSM, DG investments are selected as long as their average
cost is no greater than the T&D alternatives and, ultimately,
no savings in average costs are realized by utilities or consumers
(not to mention the additional regulatory oversight costs.)
Instead, as we discuss in the next section, applying a more
sophisticated economic analysis can identify the most beneficial
applications for DG.
Myth 4: DG should
be emphasized today to reduce the potential for future "stranded"
T&D costs
The logic of this
myth would require utilities to install DG even if such investments
were uneconomic, in order to lessen the likelihood of stranding
future assets. In doing so, however, utilities easily could
be found to be violating their obligation to serve or, in
the case of unbundled utilities providing only distribution
service, their obligation to connect. Of course, it may be
perfectly sensible for utilities to invest in distributed
resources even when not strictly economic in order to better
understand the ramifications of DG on T&D systems. But neither
utilities nor regulators should have to cloak such legitimate
research and experimentation behind poorly conceived economic
concepts.
If these myths are
not dispelled, it is far more likely that utilities will install
distributed resources in situations where they are uneconomic,
thus raising costs for all, and hindering beneficial implementation
of DR/DG in the long run. Fortunately, using a more sophisticated
economic approach, we can identify the circumstances under
which DR/DG has the greatest economic value. Thus, applications
of distributed resources can be more appropriately targeted.
In that way, they will be more likely to produce "win-win"
situations that reduce overall costs and improve reliability.
The Appropriate
Economic Framework
So what makes a DG
application economic? Fundamentally, most DG applications
require a tradeoff between higher costs and greater flexibility.
One of the great advantages of DG is its modularity and flexibility:
a utility can install several megawatts of new DG capacity
as needed, rather than building large, central-station generating
plants or signing long-term purchase-power agreements. This
is particularly beneficial because load growth at the local
level is more uncertain than at the overall utility level.
At the local level, we often see load growth occurring in
fits and starts (what economists call a "lumpy," rather than
a smooth process). By providing additional flexibility, DG
can allow utilities to improve their use of scarce capital,
while continuing to meet customer needs. Capturing the true
economic value of this flexibility is important if DG is to
be installed where it can provide the greatest economic benefits.
To capture that economic
value, we developed with the Electric Power Research Institute
(EPRI) an economic framework to evaluate DG applications based
on advanced investment analysis techniques. Our approach incorporates
future uncertainties concerning market prices, operating costs,
and load growth, and shares common aspects with the financial
and real options models that are increasingly popular, in
that they can determine the most valuable investment strategies
today and in the future.3 We have found that these
uncertainties can make or break DG economics. To evaluate
DG opportunities, we consider other options that can meet
necessary engineering specifications for a local distribution
planning area. Then, we compare all of those options on an
equal basis, much as a financial analyst would compare alternative
investments.
For example, in one
study we looked at a local distribution area that encompassed
an area with a growing residential/commercial presence, with
a large new shopping mall, commercial office space, and new
residential housing developments. However, the timing of that
new load growth is uncertain: although several large chain
stores committed to opening in the new mall, strict land use
regulations and continuing court challenges made the timing
uncertain. The same was true of anticipated residential housing
developments.
From the local utility's
standpoint, the problem was vexing. Without the necessary
infrastructure, the new growth cannot be accommodated. Yet
the utility has an obligation to serve. One option for the
utility was to install a large new substation and several
new feeder lines. That would provide sufficient capacity,
but would require a large cash outflow. Additionally, if the
development process were delayed or halted-a real possibility
for this shopping development, which had been first proposed
two decades earlier-the utility could find itself with much
unused new distribution capacity, which would raise unpleasant
regulatory issues. Alternatively, the utility could prepare
the sites for modular DG installation, and bring in trailer-mounted
combustion turbines. Although more expensive on a per-kW basis
than the substation and feeders, the DG option would provide
far greater flexibility. The DG could also defer the need
for a new substation or it might allow a smaller, less expensive
substation to be constructed. Although the mathematics is
somewhat complicated, the question is straightforward: is
the "insurance" value provided by DG worth the additional
cost to avoid the large cost commitment required to build
the substation and feeder lines?
To see the tradeoff
more clearly, it's easiest to show a hypothetical example
where there is no uncertainty about costs or load growth.
Figure 1 represents the utility's cost tradeoff. In the figure,
we show the present value of the cost of installing 3 MW increments
of DG every 12 months to match local area load growth, and
the present value cost of installing a 100 MVa substation
at various future times. As the substation is deferred longer
into the future, its present value cost steadily declines.
(We're assuming that the substation's purchase cost doesn't
change.) The present value of the cost of the annual sequence
of DG investments is a step function, such that the height
of each step is the present value of the incremental investment.
Therefore, the height of the steps continually decreases.
In this example,
the lowest present value cost occurs after 11 months. Thus,
the utility would be better off installing DG today and deferring
the new substation for 12 months. Of course, in the interim,
new information may come to light that would allow the utility
to refine its planning further. If several industries decided
not to build at the site, then the utility will not have devoted
scarce capital to a large, and mostly unused, substation.
Alternatively, if development is seen to be accelerating,
the utility might be able to install additional DG and then
build the substation. In general, the optimal timing will
depend on the relative differences between per-kW installation
costs, capacities of the DG and substation alternatives, fuel
cost, and the utility's discount rate.
Incorporating uncertainty
about DG fuel costs, electric market price, and load growth
complicates the analysis, but the underlying logic remains.
The difference is that we search for the lowest expected cost
solution. In the approach we developed with EPRI, called the
Area Investment Planning Model,4 we simultaneously
evaluate all possible distribution alternatives, installation
constraints (such as timing and compatibility with previous
installations), and uncertainty.5 While this approach
is more complex than a typical deterministic "avoided cost"
approach, it is far more likely to capture the true value
of DG investments and therefore identify greater benefits
for utilities and customers.6
Identifying the Conditions
Most Conducive for DG Applications
Fundamentally, the
benefits of DG applications stem from the modularity and the
planning flexibility they can provide. Utilities and customers
benefit when DG investments defer traditional large T&D capital
investments, while lowering the overall present value of distribution
and generation costs. In our past case studies, we have shown
that the value of DG is greatest when there is a high degree
of future load growth uncertainty.7 If a utility knows
precisely when local area loads will develop, then the value
of flexibility is eliminated. Thus, DG applications can be
thought of as providing "real option" value, whose value increases
with greater uncertainty, and vanishes without uncertainty.
It also turns out
that the value of DG is greatest when load growth is not too
rapid. As load growth rates increase, the value of the distributed
resources decrease. While this may seem paradoxical-after
all, DG can be brought in rapidly, such as with truck-mounted
generators-DG's value decreases because the deferral benefits
it provides decrease as load growth rates increase. Hence,
the greater unit capital cost ($/kW) of distributed resources
becomes more difficult to justify economically as the amount
of deferral benefit decreases.
What Does the Future
Hold?
No doubt, DG technologies
will continue to improve. Perhaps the futuristic vision described
in Popular Science and others will even be realized, with
DG replacing many new central-station generation and local
area T&D investments. Today, however, DG investments are not
universally beneficial, and so it is critical to identify
the conditions under which DG likely will provide the greatest
possible benefits to utilities and their customers. But over-selling
DG as a universal alternative is dangerous, since the greater
the hyperbole, the more likely that expectations for DG will
be unmet. Such a situation could result in wholesale rejection
of DG, even when its benefits are clear.
We believe the best
approach for utilities, regulators, and customers will be
to identify those situations where DG applications can have
the most value and focus development efforts there. Our work
has shown that, when load uncertainty is great but the overall
trend in load growth is relatively small, DG is likely to
have the greatest benefits. In this era of increasing competition
and greater energy market volatility, we expect those conditions
increasingly to be found.
Jonathan Lesser
is senior managing economist with Navigant Consulting Inc.
Charles Feinstein is President of VMN Consulting LLC.
- Opinion
and Order Approving Pilot Program for Use of Distributed
Generation in the Utility Distribution Planning Process,
Opinion No. 01-5, October 26, 2001.(N.Y.P.S.C.)
- Oddly enough,
in the 2002 legislative session, a bill was introduced
that would defined some fossil-fueled DG applications
as "renewable energy" sources.
- The approach
we use is called "dynamic programming." It is a mathematical
optimization technique that forms the basis for the solution
of virtually all dynamic investment problems, and which
is often used in "real options" analysis.
- For a description
of this model, see S. Chapel, C. Feinstein, P. Morris,
and M. Thapa, User's Manual: Area Investment Strategy
Model. 1999. EPRI. See also, C. Feinstein, P. Morris,
and S. Chapel. "Capacity Planning Under Uncertainty:
Developing Local Area Strategies for Integrating Distributed
Resources." Energy Journal, Special Issue on Distributed
Resources. 1998, 85-110.
- A complete
discussion can be found in C. Feinstein and J. Lesser.
"Defining Distributed Resource Planning." Energy Journal,
Special Issue on Distributed Resources. 1998, 41-62.
- An analysis
of the difference between this approach and the avoided
cost method is given in J. Lesser and C. Feinstein. "Electric
Utility Restructuring, Regulation of Distribution Utilities,
and the Fallacy of 'Avoided Cost' Rules." J. Regulatory
Economics 15:93-110 (1999).
- See also C.
Feinstein, Strategic Role of Distributed Resources in
Distribution Systems, EPRI, Palo Alto, CA: 1999, TR-114095.
Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.
|
|
 |
Public
Utilities Fortnightly & Spark
