The potential benefits of enabling power consumers to know and respond to the cost of producing the power they consume have long been recognized. Indeed, demand response programs, including time-of-use (TOU) pricing and metering, were credited in part for helping California in 2001 avoid a repeat of the Stage 3 emergencies and rolling blackouts that roiled Golden State energy markets during the preceding year. Yet, notwithstanding these known benefits and a federal directive dating from 1978 in the Public Utility Regulatory Policies Act (PURPA) that state regulators consider implementing TOU and seasonally differentiated rates, regulated retail power markets have little availed themselves of real-time pricing and demand response, clinging instead to fixed rates that do not track changing production costs. According to an August 2004 study of the Government Accountability Office, (GAO, formerly the General Accounting Office), three barriers have blocked progress on demand response:
1. retail rates that shield customers from fluctuating production costs,
2 lack of metering and communications equipment on customer premises, and
3. ignorance of demand response opportunities.
But that could be about to change.
Twenty-seven years after Congress recommended in the original PURPA that state regulators consider adopting rates that track fluctuating costs, federal legislators have given it another shot in the Domenici-Barton Energy Policy Act of 2005. More explicit than the original PURPA directive, the new Energy Policy Act declares it to be national policy to promote demand response and to remunerate demand reduction/conservation just as much as supply resources, and more specifically directs state regulatory authorities within 18 months to consider imposing two obligations on retail electric utilities. The first is that each utility offer each of its customer classes a “time-based rate schedule” under which the price of power tracks “the utility’s costs of generating and purchasing electricity at the wholesale level.” These rate schedules range from relatively stable TOU tariffs that fix prices for blocks of hours on a seasonal basis, to more variable TOU tariffs that are allowed to vary hourly during periods of critical peak, and to potentially volatile real-time pricing tariffs under which rates may change hourly throughout the year. The Act also includes traditional peak-shaving type programs for large consumers in its definition of a time-based rate schedule.
To remove the second barrier the GAO identified, the Act also directs state regulators to evaluate requiring electric utilities to provide each customer that opts for a time-based rate schedule with a “time-based meter” capable of communicating to the customer what the cost of power production or purchase and corresponding rate are at any point in time. The time-based rate schedule and meter, if adopted by the state, would apply to both traditional franchise retail service providers as well as competitive providers in jurisdictions that allow for competition.
In a May 2005 report on California’s Statewide Pricing Pilot, Charles River analysts Stephen George and Ahmad Faruqui reported on the benefits of variously designed demand response programs implemented in that pilot, and importantly articulated the overarching question for regulatory policy: “Are the benefits from reductions in energy use and coincidental peak demand from more economically efficient pricing sufficiently large to off-set the net metering, billing and other infrastructure costs required to implement [demand-response] rate reform?” How state authorities conceptualize and evaluate time-based rate schedules and meters as directed in the new energy bill will surely influence their answer to this central question.
Comparing the Charles River study of California to a June 2003 survey of demand response literature prepared for New England by Resources Insight analysts Brian Tracey and Jonathan Wallach illustrates how the assumptions that go into a demand response program can determine whether the benefits in reliability and avoided costs justify the infrastructure investments required to implement demand response. The carefully designed California pilot, which evaluated TOU as well as various critical peak pricing models in tandem with TOU, and differentiated customer response by weather, location, level of consumption and other distinguishing factors, produced generally favorable results. Notably, the Charles River analysts reported, “Most customers liked the rates and, when given an opportunity to continue on a time-varying rate (even in the absence of incentives and with the requirement to pay for the metering), most have chosen to stay” with TOU pricing and demand response. In contrast, the resource assessment prepared for New England relied largely on a survey of earlier and less sophisticated studies of customer demand elasticity under TOU pricing and projected that customers, if given the chance, would shift only 2.44 percent of peak demand to off-peak. The resulting energy and capacity savings would not justify the infrastructure investment, the New England assessment concluded.
One can only hope that state regulators respond to the bill with a systematic analysis, grounded in demand response pilot programs such as California’s, instead of surveys of past demand response and TOU pricing experiments. Too often those past programs and experiments were “quickly and poorly crafted in reaction to an energy crisis” and abandoned once the crisis abated, according to a recent abstract Rates and Technologies for Mass-Market Demand Response from analysts at the Lawrence Berkeley Laboratory, Levy Associates and the California Energy Commission. These analysts contend that a more systematic and accurate measure of whether the benefits exceed the cost of implementing demand response as contemplated in the Energy Policy Act of 2005 requires approaching demand response as part of a utility’s obligation to serve rather than simply an optional utility program.
Programmatic treatment of demand response is likely to marginalize the benefits and yet concentrate the technology and communications costs to specific customers, with the result that a state’s evaluation will almost surely be negative and demand response rejected. But if approached as part and parcel of a utility’s service obligation-an approach arguably more in tune with the spirit of the energy bill’s intentions-then, as the Mass-Market Demand Response analysts contend, some form of TOU pricing together with demand response will be offered to all customers as a condition of service and the technology cost can rightly be recovered as a cost of service and not a component of demand response.
By way of example, they suggest two demand response services, one mandatory and the other voluntary, for the California market in which customer demand elasticity has been demonstrated and measured to be greatest and most consistent in connection with central air conditioning HVAC systems. The mandatory service would combine an inexpensive thermostat capable of receiving (but not communicating) price information with a demand-response rate. Additional voluntary demand response could be offered beyond HVAC systems to allow thermostats, water heaters or other technologies to respond to the same demand-response pricing.
While different mandatory systems are likely to be indicated in different markets with different weather and patterns of consumption, states would do well to analyze demand response under the Energy Policy Act as part of a utility’s going-forward service obligation: an obligation that marries meters and communications with a customer’s ability to respond to critical system conditions. By making demand response a core component of utility service, state policy makers implementing the energy bill will be able to gain a thorough and accurate understanding of the true benefits and costs.
From the perspective of state regulators charged with implementing the Act and utility investors, the Western energy crisis of 2000-2001 should provide a further measure of the benefit of demand response beyond the ability of customers to respond to supply shortages and high prices.
When wholesale market prices for power swamped fixed retail prices, Pacific Gas & Electric became insolvent and was forced to seek bankruptcy protection; Southern California Edison courted the same fate. The retail prices of San Diego Gas & Electric (SDG&E) alone among the California investor-owned utilities were not comparably capped and more than doubled in the summer of 2000, causing average demand to fall by 1.6 percent and by 6 percent during periods of peak consumption. As analysts from the Pepperdine School of Public Policy point out, were it not for widespread anticipation that rate caps would soon be reinstated for SDG&E customers, the demand response to high prices would have been even greater. The ability of SDG&E customers to see real system costs and respond to them not only allowed those customers to respond to critical system shortages, but it allowed the utility to cover its costs and remain solvent. In contrast, the demand of PG&E and Edison customers, shielded as they were from soaring wholesale prices, went unabated with disastrous consequences for both utilities.
Dan Watkiss (Dan.Watkiss@bracewellgiuliani.com) is a partner with Bracewell & Giuliani in Washington, D.C. Focusing on litigation and arbitration, his clients include utilities, banks and other lenders and energy project developers.