Sam Newell and Johannes Pfeifenberger are principals at The Brattle Group. Frank A. Felder is the director of the Center for Energy, Economic & Environmental Policy at Rutgers University and a senior advisor of The Brattle Group.
When summer heat waves cause electric demand to peak, they also often cause wholesale electricity prices to rise substantially above their average levels. However, since most electricity customers face retail rates that do not reflect this movement in wholesale market prices, they do not modify their consumption patterns, causing a significant drop in economic efficiency. The Energy Policy Act of 2005 calls upon states and utilities to evaluate and implement DR programs to mitigate this problem. California has initiated comprehensive regulatory proceedings about demand response, advanced metering, and dynamic pricing. Other states, including Hawaii, Idaho, Illinois, Missouri, and New Jersey, are conducting pilot programs with a variety of innovative demand response rates and technologies.
The Brattle Group has just completed a study of the value of demand response for the PJM Interconnection LLC (PJM) and the Mid-Atlantic Distributed Resources Initiative (MADRI). The objective of the study was to quantify market impacts and customer benefits that would result form a modest 3-percent reduction of peak loads during the top 20 five-hour blocks in the service areas of BGE, Delmarva, PECO, PEPCO, and PSEG. We used a simulation-based approach to quantify the market impact that the specified 3 percent load reductions would have brought about in 2005 and under a variety of alternative market conditions. We performed the analysis using the Dayzer market simulation model, which we calibrated to represent the PJM market using data provided by PJM and public sources. By comparing market simulations with and without curtailments, we obtained the following results:
• Curtailing 3 percent of each selected zone’s super-peak load, which reduces PJM’s peak load by 0.9 percent, yields an average energy market price reduction of $8 to $25 per megawatt-hour, or 5 to 8 percent on average, during the 133 to 152 hours in which curtailment occurs in at least one service area. The range in average impacts depends on market conditions. As shown in Figure 1, even these modest 3 percent load reductions can reduce market prices by $50 to $75 per megawatt-hour in some of the hours.
• Assuming all loads (either customers or their retail providers) are exposed to spot prices, the estimated price reductions could benefit non-curtailed loads in MADRI states by $57 million to $182 million per year. The potential energy market benefits to the entire PJM system amount to $65 million to $203 million per year. Producer surplus would decrease by a nearly equal amount.
• The market impact in each zone would be substantially smaller if it curtailed its load in isolation from the other zones. By the same token, the market impact would be larger if more than five zones implemented DR programs or if greater amounts of DR participation were achieved.
Our study also estimated benefits to DR program participants:
• The energy market benefit that resulted from curtailing load of much lesser value than the price of energy on the spot market was estimated to be $85 to $234 per megawatt-hour or $9 million to $26 million per year, based on the results of the Dayzer simulations and simplifying assumptions about the economic value customers place on their curtailable load.
• The reduction in capacity needed to meet reserve adequacy requirements for a load shape that has been modified by reducing the peaks was estimated to be $73 million per year (again for curtailment of 3 percent of load in the five zones), based on an assumed capacity price of $58/kW-yr.
More rigorous analyses of these participant benefits would be needed, along with an assessment of the costs of equipment and administration of demand-response programs, to evaluate fully the net benefits to participants.
Our conclusions are summarized in Table 1. As indicated, we did not quantify several additional categories of DR benefits. These include enhanced competitiveness of energy and capacity markets, reduced price volatility, the provision of insurance against extreme events that have not been captured in the scenarios considered, reduced capacity market prices, and deferred T&D costs. In addition, because we focused on curtailments to day-ahead schedules, our estimates do not capture the additional benefits that would accrue from real-time demand response, which can mitigate the effects of unexpected increases in load, generation outages, and transmission disruptions.
It is equally important to note that our study did not quantify secondary effects that could offset the benefits to non-curtailed loads. Consumers may shift load to other hours, which could casue prices to rise in those hours. Our estimates of price effects also would be offset partially by a more muted response of customers on real-time pricing, as a consequence of the lower market prices. Moreover, reduced energy prices and reductions in the demand for capacity could accelerate the retirement of old capacity or delay the construction of new capacity, leading to an eventual increase in energy prices relative to our estimated price reductions. In addition, assuming that energy and capacity markets reach competitive equilibrium, a reduction in energy market prices and hence energy margins likely would trigger an increase in capacity prices as suppliers raise capacity bids to recover their going-forward fixed costs. Ultimately, the long-term benefits will be determined by the extent to which adding DR to the resource mix lowers total resource costs.
The study is available at: http://www.energetics.com/madri/pdfs/BrattleGroupReport.pdf.