February 03, 2020

Residential Demand Response in Wholesale Markets

Rujeko Muza

I. Introduction

In 2016, the Supreme Court of the United States issued a landmark decision in FERC v. Electric Power Supply Ass’n. The Supreme Court upheld the Federal Energy Regulatory Commission’s (FERC) policy regarding the participation of “demand response” programs in wholesale capacity markets. The decision to compensate demand response at the same rate as electricity generation was a monumental step toward decarbonizing the U.S. electricity grid. However, further policy making regarding demand response (DR) is paramount to achieving deep decarbonization. While the residential demand response (RDR) market is smaller than its commercial and industrial counterpart, aggregated RDR can provide a substantial amount of load reduction and environmental benefits. Accordingly, the following is a brief discussion regarding the regulatory opportunities and challenges associated with allowing RDR to play in wholesale markets. The first section of this paper will overview and explain the importance of DR programs and their existing participation in energy and capacity markets. The second section will overview the regulatory challenges associated with allowing RDR programs to participate in markets. The final section will suggest ways of better integrating RDR programs into wholesale markets.

II. Analysis

A. The Different Flavors of DR

FERC defines DR as “the reduction in consumption of electricity by customers from their expected consumption in response to either reliability or price triggers where the customer forgoes power use for short periods, shifts some high energy use activities to other times, or uses onsite generation.” In other words, DR pays consumers for reducing their consumption of electricity during peak demand. By reducing the need for “peaking” power plants, DR can reduce the overall carbon intensity of the grid. If DR is treated as equal to other energy resources when planning for future electricity supply, it can also help to defer or avoid the construction of new fossil fuel plants. Using DR to fill the voids caused by variable-output renewables, such as wind and solar, can also help to increase renewable energy integration onto the grid. DR also increases grid flexibility and can relieve transmission congestion.

DR programs are able to participate in both wholesale and retail markets. As explained in FERC Order No. 745, the Commission “has jurisdiction over demand response in organized wholesale energy markets, because it directly affects wholesale rates.” FERC began approving wholesale DR programs developed by regional transmission operators (RTOs) in 1999. DR resources can participate as voluntary reliability resources and can also bid into RTO day-ahead markets specifying the amount and price at which they are willing to curtail. DR resources can also participate as capacity resources, receiving advance reservation payments in exchange for their commitment to participate when needed. For example, DR is able to bid into the PJM Interconnection LLC’s (PJM) “Base Residual Market,” an annual auction based on projections for what electricity demand will be in three years. DR is also able to participate in PJM’s “Incremental Auctions,” smaller balancing auctions held every year leading up to the delivery date whereby bidders can buy or sell their commitments. DR serves an important function in capacity markets. It is often the lowest cost resource, which suppresses the price of all of resources in the market by lowering the market clearing price.

States have jurisdiction over the sale of electricity to end-users—i.e., retail markets. Although commercial and industrial end-users may closely monitor electricity prices, residential consumers are often apathetic to the changing price of electricity. DR engages residential consumers by paying them to reduce their consumption of electricity at times of peak demand. Accordingly, many state public utility commissions have embraced residential DR programs.

The retail DR programs once managed by utilities are increasingly bid into wholesale markets. DR is accumulated through third-party aggregators or curtailment service providers (CSPs) who recruit customers that are too small to participate on their own, such as schools, commercial chains, or groups of residential customers. By aggregating small customers, CSPs have increased customer participation in many wholesale reliability and emergency programs.

Traditionally, commercial and industrial DR programs have been the dominant participant in wholesale markets. High electricity consuming factories that have an individual ability to affect load were the natural first target for DR programs. Particularly because large commercial and industrial customers were more likely to meet RTO minimum thresholds for market participation. Now, there is a shift toward increased residential DR. Indeed, global RDR capacity is expected to increase from 13.8 gigawatts (GW) in 2019 to 47.4 GW in 2028. Key drivers of this growth are new grid-connected technologies and advanced data analytics as further explained below. 

B. Regulatory Implications of Incorporating Residential DR into System Planning and Operations
1. Technological Innovations in Residential DR
In theory, the technological ability to increase the adoption of residential DR programs into wholesale markets exists. As a result of technological innovations and policy directives, new types and applications of DR are emerging. New technologies encompass the use of smart appliances that respond in near real-time to price or other signals. These customer-facing technologies may allow consumers to respond more easily to price signals as they require little customer monitoring or interaction. For example, real-time, price-based DR management software for residential customers could be programmed in smart meters for determining the optimal operation of residential appliances while considering variation in load. These nonintrusive programs automatically determine the optimal operation of residential appliances within five-minute time slots while considering uncertainties in real-time electricity prices. Widespread deployment of the technology is feasible given that nearly half of all homes in the United States now have smart meters. Further, a “learning algorithm,” referred to as Consumer Automated Energy Management System (CAES), can model energy prices and residential device usage. CAES can adapt to individual residents’ preferences, life patterns, and pricing changes over time to modify device usage and save residents money. This is important because maintaining a balance between energy consumption cost and users’ comfort satisfaction is paramount to residential user adoption and retention of RDR programs.

2.. Jurisdictional Barriers to Incorporating Residential DR
While the technological ability to incorporate RDR into system planning and operation is on the cusp of reality, the regulatory implications are unclear. A state may have viable legal arguments to oppose FERC’s acceptance of a tariff from an RTO that provides for residential participation. As explained in Order No. 745, “jurisdiction over demand response is a complex matter that lies at the confluence of state and federal jurisdiction.” A state could argue that consumers who reduce their electricity consumption in order to secure RDR payments would otherwise purchase that electricity in the retail market, which is under state jurisdiction. By approving the tariff, FERC is effectively regulating retail rates and usurping or impeding state regulatory efforts concerning demand response. Indeed, the Commission expressly declined to regulate retail rates in Order No. 745 for the same reason.

FERC could respond by arguing that the logic of EPSA v. FERC should be extended to permit federal jurisdiction over residential customers’ participation in programs that directly impact sales and wholesale and transmission of power in interstate commerce. In EPSA v. FERC, the Court approved the Commission’s finding that the Federal Power Act (FPA) required it to “ensure that the rates charged for energy in wholesale energy markets are just, reasonable, and not unduly discriminatory or preferential.” Accordingly, it would not “refrain from acting on demand response compensation in the organized wholesale energy markets because of the potential actions that state retail regulatory authorities may or may not take.” Similarly, in the context of residential demand response, a commitment to reduce residential electricity consumption lowers the wholesale rate and helps maintain a reliable supply of electricity—two considerations that are crucial to FERC’s obligation to ensure that wholesale prices are “just and reasonable.”

The state could again argue that the opportunity to receive a payout for reducing electricity consumption is a critical factor in a sophisticated consumer’s decision to purchase retail electricity––as important as the retail rate itself. However, the Commission could respond by arguing that its approval of the tariff does not run afoul of section 824(b) of the FPA’s proscription against FERC regulation of retail sales just because it affects the quantity or terms of retail sales. “Transactions occurring on the wholesale market have natural consequences at the retail level, and so too, of necessity, will FERC’s regulation of those wholesale matters. That is of no legal consequence. When FERC regulates what takes place on the wholesale market, as part of carrying out its charge to improve how that market runs, then no matter the effect on retail rates, §824(b) imposes no bar.” Based on the foregoing, FERC would have strong justifications for approving an RTO tariff with RDR provisions.

3. Practical Barriers to Incorporating Residential DR
There are also practical challenges surrounding the integration of RDR programs into wholesale markets. Utilities and residential DR aggregators must be able to reasonably predict load to accurately determine load reductions. DR rewards consumers for reducing load during certain market conditions and at specific times. However, it is difficult to measure and quantify this reduction. Measuring and verifying the reduction requires measuring a consumer’s baseline usage against their actual usage. The difference is measured to determine the reduction in the event they are called upon to lessen their load.

An accurate measure of typical usage is important to prevent or detect gaming by participants and to reduce the likelihood of outages. However, there are many opportunities for load forecasting error as well as issues with evaluating, measuring, and verifying DR load reductions in wholesale markets. A utility may provide “false positives” when they estimate a load reduction but customers do not in fact reduce load. A utility may also provide “false negatives” when they estimate no load reduction but customers do actually reduce load. Forecasting errors could jeopardize the reliability of the grid and could result in outages. Outages based on poor forecasting may tread into FERC jurisdiction because FERC is responsible for ensuring the reliability of the grid. FERC approval of an RDR program that posed a significant threat to the reliability of the grid could thus be vulnerable to challenges.

A way to mitigate load uncertainty would be to use probability distributions to estimate the trigger conditions for RDR programs. There may also be a role for state public utility commissions in terms of reviewing a CSP’s ability to deliver aggregated RDR. In regulating third-party aggregators, states will need to resolve several key questions including: (1) what types of RDR can be aggregated; (2) who may aggregate RDR; (3) and whether third-party aggregators should be allowed to bid directly into wholesale markets or whether they should only offer services to the local utility. If CSPs could bid directly into the market, states would have to consider whether state regulators of retail electricity services would cede oversight and authority to federal regulators of wholesale markets. Although, there may be a strong need for a federal backstop in terms of reliability, it would likely lead to contentious questions regarding jurisdictional authority that may require adjudication.

Determining the amount of compensation for residential participants is an additional challenge. “Baselines” are currently used to compensate participants in retail DR programs. Baselines are calculated using historical interval meter data for each DR or statistical sampling to estimate electricity consumption of an aggregated demand resource. Utilities or third-party aggregators could adopt a regime similar to critical peak pricing whereby a customer is refunded at a single, predetermined value for any reduction in consumption relative to what the utility deemed the customer was expected to consume. Another suggestion is using rebates similar to net metering programs, a billing mechanism that credits solar energy system owners for the electricity they add to the grid so that customers are only billed for their “net” energy use. Similarly, utilities could devise a billing mechanism that credits residential customers for a fixed amount or percentage of the benefit the utility received from bidding RDR into wholesale markets.

Admittedly, devising a program that identifies, controls, and compensates individual unaggregated RDR would be costly for a utility. However, the aforementioned technological improvements help make tracking and compensating unaggregated RDR a more worthwhile investment. Moreover, utilities can opt to shift the cost and risks of deploying an RDR program to third-party aggregators by contracting with CSPs.

C. Suggestions for Increasing Residential DR Program Participation in Wholesale Markets
Updating market rules will allow RDR to better penetrate wholesale markets. However, careful market design is key for mitigating the above challenges and reducing the threat of market failure. First, there would need to be a strong penalty policy for utilities or third-party aggregators who bid into the market yet do not curtail load as promised. False positive forecasting creates free-rider issues and jeopardizes the reliability of the grid. Particularly during weather emergencies like the 2014 polar vortex, it is important that if RDR is called upon to curb demand, it has the ability to do so.

In terms of bidding into the market, it would be prudent to require utilities and third-party aggregators to maintain a “buffer” around their bid amount so that they are not overcommitting their load reduction. Perhaps this rule could be structured similarly to existing RTO policies regarding the capacity percentages wind and solar are allowed to bid into the market due to their variable nature. Given that buffer calculations are already prevalent in energy capacity estimations, it seems viable for RTOs or FERC to devise buffer bid rules for RDR market participants. There may even be a role for the North American Electric Reliability Corporation to play in developing rules and conducting regulatory oversight.

The policies adopted by retail regulatory authorities like state utility commissions and the governing bodies of municipal and cooperative electric utilities, along with the market rules adopted by RTOs, will strongly influence the viability of RDR programs and the attraction and retention of residential consumers. For example, utility reform can help increase the adoption of RDR programs. Under the traditional cost of service model, a utility may not have an incentive to encourage residential consumers to reduce their energy consumption. However, utilities can be incentivized by mechanisms that decouple profits from the amount of electricity sold, or can adopt performance based regulations which reward utilities that meet or exceed reduction targets. Utility buy-in as well as customer buy-in and retention are key drivers for the adoption of RDR programs.

Simplifying ISO/RTO market rules and procedures for wholesale electricity market participation may also be a way of increasing RDR participation in the market. Rules for wholesale market participation are complex and any company or customer wishing to individually participate in the market must invest significant time and money. The transaction costs of market participation are substantial and thus only practical for large industrial and commercial customers who can afford to dedicate professional staff to this activity. Simplifying the rules to allow increased participation of RDR aggregators could transform wholesale and retail electricity markets. As part of the revision process, regulators should also seek to remove any implicit bias in service and participation definitions.

Another idea is to reduce the amount of operating reserves in wholesale markets. When markets have operating reserves far in excess of resource adequacy requirements, load curtailments are rarely needed, meaning RDR resources can expect very little market value. If RTOs and ISOs cut back reserves, market revenues for RDR have a greater likelihood of exceeding the transaction costs for market participation. A greater profit margin for utilities and CSPs could encourage more aggregators to invest in RDR programs and deployment.

III. Conclusion

Markets are important. They dictate the electricity system of the future because they direct and encourage investments in different kinds of assets including DR. Although there are regulatory challenges associated with incorporating DR into system planning and operations, if we want to encourage the future use of DR, we must continue to find meaningful ways for DR programs to play in electricity markets. Residential DR is a non-negligible contributor to the deep decarbonization of our grid.


    Rujeko Muza

    Rujeko Muza’s essay “Residential Demand Response in Wholesale Markets” received third place in the 2019 Energy Law Writing Competition sponsored the Section of Environment, Energy, and Resources. Rujeko is a student at Georgetown Law.