As competition in the electric industry increases, so does utility concern about the effect of demand-side management (DSM) programs on electricity prices. Because DSM programs often raise prices, several utilities have recently reduced the scope of their DSM programs or focused these programs more on customer service and less on improving energy efficiency (see sidebar). Whether all utilities should follow suit is, however, open to question. We contend that DSM programs do not always exert upward pressure on prices (em just sometimes. The actual effect will depend on many factors: the intensity of DSM programs, the underlying utility cost structure and retail tariffs, avoided costs, and regulatory treatment of DSM-program costs.
We used the Oak Ridge Financial Model (ORFIN) to examine the two factors that contribute to DSM's upward pressure on prices: 1) the cost of the programs themselves, and 2) the loss of revenue associated with fixed-cost recovery. (The second factor reflects the reduction in revenues caused by DSM-induced energy and demand savings that exceed the reduction in utility costs.) Our analysis examined DSM price impacts as functions of the factors shown in Table 1. (For details, please consult the Oak Ridge National Laboratory report, Price Impacts of Electric-Utility DSM Programs.)
Using data from the Energy Information Administration on the 180 largest investor-owned utilities, we created a "typical" U.S. utility: Avoided costs are very low until 1999, reflecting a regional market that has considerable excess capacity and low-cost energy. Beginning in 2000, avoided costs increase rapidly to their steady-state values in 2002. These higher values reflect the need to construct new facilities to meet increasing demands. Total avoided costs are based on the assumption that the DSM programs avoid 50 percent of the system-average demand-related transmission and distribution (T&D) costs.
We then constructed a reference DSM program that operates in 1995, 1996, and 1997 to yield a
1-percent reduction in peak demand as of January 1, 1998. The program's conservation load factor (CLF) of 40 percent means that electricity consumption is cut 0.67 percent in 1998, given a system load factor of 60 percent. (CLF is the ratio of the reduction in average demand versus peak demand, as induced by a DSM program.)
The initial cost of the program is $1192 per kilowatt (Kw) (3.6 cents per kilowatt-hour (›/Kwh)), of which the utility pays half. The measures are assumed to last 15 years on average; the utility costs are added to rate base and capitalized over a 10-year book life. The program's costs and effects are split 33/67 percent between the residential and
commercial/industrial (C/I) sectors, consistent with the sectors' shares of total sales. The program defers 50 percent of the demand-related T&D avoided cost. The initial cost is set to yield a TRC benefit-to-cost ratio of 1.5. So that the effects of DSM fall entirely on customers, not on utility shareholders, our analysis included annual rate cases based on a future test year.
Over the 15-year lifetime of the DSM investment, the program cuts total costs by 0.13 percent, and raises average electricity prices by 0.25 percent. In the initial years, price increases grow as the program costs are added to rates, and avoided costs are low (see Figure 1 on p. 28). The price impact peaks in 1998 at 0.7 percent, and though always positive, it declines through 2012 to 0.05 percent. In this case, program costs account for 55 percent of the price increase over the analysis period; fixed-cost recovery (FCR) accounts for the other 45 percent.
In the base case, monthly customer charges are low ($10 to $15 a month) for both the residential and C/I classes. This low charge includes only 5 percent of the utility's fixed costs. In the cases examined here, we assigned increasing fractions of the fixed costs to the customer charge. Fixed costs include all the operating costs associated with T&D and customer service not assigned on a per-kilowatt basis plus all the capital costs (depreciation, property and income taxes, interest payments, and returns to shareholders).
The effect of DSM programs on prices decreases as the percentage of fixed costs assigned to the customer charge increases (see Figure 2 on p. 28). This change occurs because increasing the customer charge reduces the demand and energy charges. Lowering these volumetric charges toward their short-term marginal-cost values reduces the FCR component of the DSM-induced price increase. Stated differently, the price impacts of DSM-program cost recovery are independent of the structure of retail tariffs, whereas the recovery of fixed costs depends strongly on the structure of these tariffs. Assigning 100 percent of fixed costs to the monthly customer charge renders the FCR component negative (meaning, electricity prices are lower) and cuts the price impact of DSM from 0.018 to -0.35›/Kwh (or 0.25 to
-0.04 percent) over the 15-year period.
The irony of these results is that with all fixed costs assigned to the monthly customer charge, customers face no adverse price effects. On the other hand, because the volumetric charges are lower, customers find little incentive to invest in efficiency measures on their own. And those that participate in the utility's DSM programs gain less. In the cases examined here, the residential energy charge declines from 8.9›/Kwh in the base case to 3.8›/Kwh in the current case. Correspondingly, the customer charge increases from $11 to $91/month, a level that many regulatory commissions and customers may find unacceptably high. However, these changes may be more consistent with a competitive electricity market, in which prices reflect more closely the time-varying short-term costs of production.
ORFIN results show that rate impacts can be minimized by reducing program costs (by using market transformation strategies, working closely with trade allies, or shifting more costs to participating customers) and focusing DSM programs on geographical areas where large T&D investments can be deferred. The FCR component of DSM price effects can be reduced by putting more of the utility fixed costs in the monthly customer charge.
Adjusting the timing of DSM programs to match avoided costs can also cut price impacts.
We combined these factors to gauge the net effect on electricity prices. Cutting DSM program costs in half (so that customers now pay 75 percent of total costs) cuts the 15-year price increase by one-fourth. Increasing the percentage of T&D costs that can be avoided by DSM programs from 50 to 150 percent cuts the 15-year price increase in half. Increasing the percentage of fixed costs assigned to the monthly customer charge from 5 to 20 percent cuts the price increase 15 percent. Shifting avoided costs four years earlier cuts the price increase 20 percent. Combining these four changes cuts the average price increase from 0.25 to -0.03 percent (see Figure 3).
This combination of factors leads to a DSM program that lowers electricity prices. Very small price increases occur while the program is in effect. Beginning in 1999, however, yearly prices are lower with DSM than without. Prices decline because avoided costs are higher and undepreciated program costs are lower.
Whether this combination of factors and its effect on electricity prices is reasonable depends on the specific utility and its DSM programs. We think it possible to run carefully designed and targeted DSM programs that lower electricity prices. But because such programs require participants to pay a large share of the DSM costs, participation is likely to be lower than in programs where the utility pays for most of the DSM. Similarly, because such programs focus on areas with high avoided T&D costs, the potential to reduce the need for generation (and its attendant pollution) is reduced relative to systemwide programs.
Utilities that run broadly based DSM programs, however, are likely to experience modest price increases. Only if natural gas prices increase or pollution-control requirements on power plants become stricter will DSM consistently offer the possibility of both cost and price decreases.
Ultimately, utility and PUC decisions on DSM programs will
hinge on much more than price impacts alone. As San Diego Gas & Electric noted in a June 1994 filing to the California Public Service Commission:
"Currently, SDG&E has a large and successful DSM program in place, continuing the direction that was established as a result of the California Collaborative Process in 1990. This program was implemented to address market barriers to cost-effective energy-
efficiency measures. At that time, it was determined that utility involvement in energy efficiency was necessary to overcome these barriers, so that cost-
effective energy efficiency could be a viable resource option in California.
SDG&E believes that the market barriers that necessitated utility DSM programs still exist and a strong utility role in DSM is still required if those programs are to continue to thrive."
DSM provides substantial economic and environmental benefits to utilities, to their customers, and to society at large. One important benefit is lower emissions of carbon dioxide, a major contributor to greenhouse warming. Alternative ways to reduce U.S. emissions of greenhouse gases might be much more expensive than DSM programs. A recent study from the U.S. Environmental Protection Agency estimated that a $120-per-metric-ton tax on the carbon content of oil, natural gas, and coal would roughly stabilize carbon dioxide emissions at 1990 levels over the period 1990 to 2030. Imposed on electricity consumption, such a tax would roughly double the price of electricity (that is, the tax would be equivalent to almost 8›/Kwh).
DSM programs often increase electricity prices, but the effects are quite small, both in absolute terms and relative to the many other factors that affect electricity prices. Thus, the threat of increasing competition, by itself, should not deter utilities and their regulators from acquiring cost-effective DSM resources. t
Eric Hirst and Stan Hadley are researchers at Oak Ridge National Laboratory. They specialize in electric industry issues, including resource planning, demand-side management, and industry restructuring. Hirst holds a PhD in mechanical engineering from Stanford University, and Hadley holds master's degrees in nuclear engineering and engineering management from the University of Missouri-Rolla.
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.