Strategic Demand Management with Performance Incentives

The sections below describe the trends emerging from our research and recommend best practices for structuring SDR PIMs.


SDR potential remains largely untapped, reaching nowhere near the Brattle-estimated potential of 200 GW of cost-effective load flexibility and consumer benefits of $15 billion per year. Nonetheless, our review of several case studies demonstrates that PIMs can be an effective strategy for incentivizing SDR. Of the five cases studied that have available results, four administrators (in the Massachusetts, New York, Texas, and Vermont cases) met or exceeded their targets. Administrators in the seven states reviewed earned incentives ranging from 2% to 25% of their program spending. However this range is meant to illustrate only a possible range as the baseline spending for each utility was calculated in a slightly different manner, as the table describes. Utilities saved about 0.69–6.5% of total peak demand, and each program created consumer benefits, with benefit–cost ratios of 2 and higher.

Although these incentives are generally successful, we found two key design features that require careful consideration: sizing incentives to be both financially meaningful and aligned with the scale of benefits, and maintaining a focus on outcomes relative to process or activities in incentive design.
Regulators face a challenge when sizing an incentive to encourage utility uptake and performance beyond business as usual while also attempting to maximize performance benefits for customers. One consideration in that analysis is goal stringency. If the goals are easily reached, a low incentive could be sufficient. Conversely, if a performance target represents a big departure from business as usual, a larger incentive may be necessary to compensate utilities for the additional risks involved in making major changes to their operations and business model.
Two case studies from Hawaii and Texas, and an additional example from Missouri, illustrate these tradeoffs and the difficulty of assessing whether SDR incentives are sized appropriately. Hawaii’s SDR PIM failed to elicit utility performance, which implies that the incentive was not well aligned with the target. Multiple factors were at play, however, including that HECO’s SDR PIM lacked recurring annual revenue and faced a difficult timeline. Experience with Ameren Missouri’s energy efficiency PIM similarly demonstrated that the incentive alone was not sufficient to motivate the utility to consistently increase its efficiency savings levels (Nowak et al. 2015).
According to our analysis, administrators (mostly utilities) are earning $50,000–62,500/MW of saved peak demand from SDR PIMs. A recent LBNL study evaluated costs borne by program administrators for first-year peak demand savings through energy efficiency programs; these costs ranged from $568,000–2,353,000/MW across the nine states it evaluated. It is important to note that these programs also produce energy savings, which are additional benefits against which these costs can be compared. The states are in different climate zones and have different peak demand periods, which may also affect the cost of saved peak demand (Frick et al. 2019). Despite the difficulty of comparing heterogeneous SDR programs, the cost figures do provide some helpful context. For example, in Texas, the incentive earned is equal to about 6.8–13% of the cost per peak MW saved. In Massachusetts, the incentives are higher, but the costs are also higher—equal to about 6.5% of the cost per peak MW saved. These figures are similar to conventional energy efficiency performance incentives, which are typically 3–15% of program spending (Relf and Nowak 2018).
Additionally, an incentive structure with a cap or ceiling can encourage performance up to that cap but not beyond it. For example, ConEd achieved its full peak demand reduction PIM by acquiring about 85 MW of peak demand reduction, just over its approved maximum target of about 82 MW. In contrast, Texas utilities regularly exceed their performance targets and have no incentive cap. However the Texas program’s high achievement level indicates that the demand reduction performance thresholds could be increased; they have held constant for nearly two decades.
Table 6 above attempts to show each incentive’s relative size by evaluating the percentage of SDR program spending in the “Results for utilities” column. For the newer PIMs, these may need to be recalibrated over time, as it may take one to two cycles to calibrate and size an
incentive (Whited, Woolf, and Napolean 2015). While it is not ideal to change investment expectations for targets or incentives set in the first performance period, the second performance period should certainly learn from the first. Additionally, states wary of misappropriating ratepayer dollars can create a measurement-only period to find a baseline from which to work—an approach taken by the Minnesota PUC in its proceeding to develop metrics that support future PIMs (MN PUC 2019).
With the exception of the performance incentives for VEIC in Vermont, the SDR PIMs studied here are upside-only for the utility.17 PIMs can be designed as upside-only, downside-only, or symmetrical, with penalties for poor performance and rewards for excellent performance. This range of design choices is not theoretical; many reliability or safety PIMs are negative-only, while many cost risk-sharing PIMs are symmetrical (Whited, Woolf, and Napolean 2015). Upside-only PIMs provide benefits for new SDR programs without much downside for poor performance, and they introduce little new risk to utility shareholder returns. These upside-only PIMs are useful incentives in an environment such as SDR, where significant technology risk might lead to worse than expected performance. They also aim to compensate the utility for missed capex earnings opportunities.
To make SDR a core part of the utility business model, incentives and other policies can continue to strengthen the link between utility performance on SDR and investor returns. Doing so may fundamentally shift investor attitudes about utility risk, depending on how the PIMs are structured and sized. As confidence builds around utility performance on SDR, regulators will have the opportunity to iterate and expand compensation mechanisms, try new compensation structures that introduce additional upside and downside risks, increase the stringency of performance targets, and try innovative new metrics. Creating adaptive processes that promote continuous improvement is key to growing SDR as utilities gain experience promoting this important resource.
Traditional energy efficiency and demand response are the most common resources currently providing SDR as part of PIMs, but other resources, such as behind-the-meter battery and thermal storage, fuel switching, and solar PV, are emerging in several states.18
Incentives that focus on outcomes (the effect of the utility’s activities) rather than process (which activities were required of the utility) often enable a technology-neutral approach, while also providing more clarity and transparency to stakeholders on the benefits that were actually achieved (O’Boyle and Aggarwal 2015). Rhode Island’s peak demand reduction PIM exemplifies outcome-based measurement and compensation. The PIM measures annual peak capacity savings, and allows for diverse approaches to meet this goal including distributed solar PV and storage. Utilities and customers split the benefits, measured as
17 VEIC’s performance compensation is based on the achievement of a broad set of QPIs that includes some minimum performance requirements that assess financial penalties if targets are not achieved.
18 As described above, utilities may have different inherent incentives to pursue fuel switching or utility-owned storage; we have yet to see SDR PIM designs account for these nuances.
avoided-capacity market and transmission tariff costs. The active demand component of Massachusetts’ PIM is also technology neutral.
Likewise, New York’s SDR EAM encourages utilities to partner with third parties and customers to use whichever technology can deliver the results in a cost-effective manner. ConEd included a wide range of resources in its system peak reduction program; thermal storage and fuel switching provided the greatest peak reductions, with additional contributions from demand response and EVs. In Michigan, the MI PSC has indicated that it may move to an outcome-based metric that includes demand response in NWAs, rather than rewarding utilities for consideration of specific resources, whether or not they were actually deployed. These outcome-based metrics include potential challenges, however, including the administrative costs in setting up and accurately defining the metrics and ongoing challenges with measuring success with uncertain baselines. Regulators will also need to avoid double-counting with any existing activity- or program-based metrics.
Current SDR PIMs focus primarily on long-term adaptation of customer demand in response to prices and efficiency measures (shape) and traditional utility and wholesale market demand response programs (shed). Most SDR PIMs have yet to focus on enabling daily changes to consumption by moving demand from one time of day to another and on grid-balancing measures targeting ramping services (shift) that can better support renewables integration. Over time, these resources will require PIMs to account for time and locational value, and they may also require new metrics for assessing success.
An exception to this focus on shape or shed services is Massachusetts, which has created a PIM to deliver active demand management, which would provide shed and shift demand reductions to support renewables integration. New York’s load-factor reduction PIM also begins to get at valuing shift as well as shed and shape, although it is an inherently approximate measurement of SDR. Some states, including New York, Vermont, and Rhode Island, are also implementing financial incentives for NWAs that reward utility procurement of locational SDR as well, though only a handful of demonstration projects exist.19
Most of these states provide greater incentives for the shape services provided by some energy efficiency investments, which face both throughput and capex bias, than SDR focused on shed and shift, which primarily faces the capex bias. In designing PIMs, regulators are analyzing the costs and benefits of different SDR services and resources, and balancing those costs and benefits against the underlying incentives utilities have to deliver SDR with different resources. This follows the principle that PIMs should help influence the utility to do what it might not otherwise be inclined to do under traditional regulation. Utilities are more inclined to shift and reduce demand when sales remain constant.
19 See generally, page 25: “A number of states—including California, Maine, New Hampshire, New York, Rhode Island, and Vermont—require utilities to consider distribution level non-wires solution projects that meet defined screening criteria. “
In the three states we examined with demand incentives focused on shed or shift services, the available incentives for energy efficiency and shape services are higher than for active demand reduction from shed and sometimes shift. In Michigan, utilities can earn up to 20% of their spending for energy efficiency, but a maximum of 15% for demand response, specifically for shedding services. In the proceeding to determine Consumer Energy’s incentive, staff and NRDC recommended a lower incentive for demand response than the one proposed by Consumers Energy, which would have included the opportunity to earn up to 20% of spending for demand response. The administrative law judge agreed, noting that energy efficiency results in significantly more lost revenue for the utility since it operates year-round and not simply in the few annual on-peak hours; the commission ultimately adopted the lower incentive (15% of spending) for demand response (MI PSC 2019). In Massachusetts, the difference is more stark: more than $65 million is available for what it calls passive demand reductions (that is, shape demand reduction), while only $5 million is available for active demand reductions (shed and possibly shift).
In our case study review and analysis, the incentives that combined a consistent policy signal over a long period (at least a decade) and included regular, sequential program cycle updates (such as those in Massachusetts, Texas, and Vermont) tended to elicit the most SDR. Most of the examples we reviewed were relatively long term. However Hawaii’s one-time PIM stood out as a short-term, nonrecurring PIM; in it, the utility was not successful in receiving the incentive. Utilities, which plan and invest over long time horizons, likely need a multiyear program to properly incent them to integrate SDR technology and policy.
We also found that the best-performing PIMs were revisited through interim policy updates (anywhere from annual to triennial). Without an opportunity for iterative updates, commissions and utilities cannot fine-tune the program over time or adjust to new technology. Synapse’s Utility PIMs report notes that incentives “may need to be adjusted over time”; targets and metrics may also require adjustment. For example, many of our case studies evolved from energy efficiency programs and mechanisms and incorporated elements of demand response (called active demand management in some states), and they were able to do this in cycle updates.
Reporting and data collection should be available to the public for accountability and transparency in the regulatory process. Synapse’s report suggests that, to report and track key data, commissions should create dashboards that are accessible, contextualized, clear and concise, comprehensive, and up-to-date (Whited, Woolf, and Napoleon 2015). Incentives that allow for regulatory certainty, adaptation, and transparency prove to be the most effective in producing results.
Durable, long-term incentives also relate to the contract terms and details of the incentive itself. If contract terms are reduced to short time frames, utilities may not have the incentive to acquire new resources. Until 2006, load-management programs in Texas were required to have a standard minimum measure life of 10 years like other efficiency measures, with annual incentive payments for a 10-year contract term. As a result, new load resources would have to be added each year to those already under contract in order for utilities to
capture additional demand reduction credit. With changes to allowable measure lives from these programs and no change to the PIM, utilities were effectively able to use the same loads every year to apply to their demand savings, increasing their opportunity to earn the incentive without increasing demand reduction investment (Beville and Howell 2017).
Although PIMs can encourage SDR, the cases illustrated here show that successful states also have complementary policies in place. These include energy efficiency and other clean energy targets; business model reforms, such as decoupling and energy efficiency PIMs; independent EM&V; and valuation mechanisms in wholesale markets, rate design, and distribution resource planning. Many of these reforms fall under grid modernization proceedings underway across the country.
Research shows that utilities require program cost recovery, decoupling of revenue from sales to remove the throughput incentive, and performance incentives to achieve robust energy savings (Molina and Kushler 2015). Additionally, EERS, renewable portfolio standards (RPS), and other clean energy requirements mandate procurement of certain resources and stimulate market growth in those sectors, driving down costs and increasing availability. Such business model reforms can help to facilitate a cultural shift toward incorporating clean energy as a core part of business operations. Of the seven states studied, all have an energy efficiency PIM, an RPS, and an electric EERS in place. Five have electric decoupling.
Beyond PIMs, rate design can further incentivize investment in SDR. Currently, few customers receive price signals through their rates that reflect SDR’s value. TVR, such as time-of-use and critical peak pricing rates, can provide price signals to customers that encourage energy efficiency and SDR on a granular time and locational basis. Of the states studied in this paper, at least one utility in five of the states has a residential time-of-use rate in place.
Wholesale markets provide another opportunity to realize additional SDR value and can also help to inform demand reduction PIM design. This is the case in Rhode Island, New York, and Massachusetts, where peak demand reductions are defined within the context of bulk power system peak. The wholesale market also provides a market-based price signal for SDR resources’ value; this price signal is used to compensate resources in a value-stacking reward mechanism, like that in New York. Further evolutions in wholesale market design, including participation models for aggregated distributed energy resources,20 will provide new pathways for customers to provide and be paid for SDR.
Value streams beyond PIMs are important drivers for investment in SDR resources. Many states do not have a specific PIM in place for SDR resources, but utilities can be rewarded for their performance by bidding SDR resources into a wholesale market and sharing some of the savings. For example, demand resources are growing as a part of the PJM capacity
20 As of this writing, FERC is assessing DER participation models in Docket No. RM18-9. See generally
market, which does not have specific SDR PIMs in place (Relf and Baatz 2017). Capacity payments can be considered a kind of PIM, although capacity costs are typically passed through to retail customers. In the most recent PJM auction, payments to demand-side resources were more than $820 million. In contrast, states in ISO-NE are stacking wholesale market payments with SDR PIMs and have been successful at achieving SDR. Additional research may be necessary to determine the interactive effects of wholesale market payments with PIMs. This is an ongoing question in Maryland, where the commission has held off on setting SDR targets for utilities due to questions about changes to PJM’s market structures and uncertainty about the current market saturation of direct load-control programs.21
Wholesale markets capture locational value on a subregional scale; including SDR in distribution system planning can capture locational value on an even more granular level. As discussed above, PIMs can help to incentivize use of SDR in NWAs (as in Michigan and New York). States can also require the consideration of distribution system planning to evaluate SDR resources and NWAs as another way to capture this value and ensure that all resources are being evaluated on a level playing field. Coordinated system planning helps to improve overall system efficiency, which is often a desired outcome of SDR and SDR PIMs.
Utilities and grid operators are also paying increased attention to distribution system planning due to new loads coming online on the customer side of the meter, including from electrification. Electrification, that is, fully or partially switching from technologies that directly use fossil fuel to those that use electricity, is critical to achieving long-term GHG reduction goals, and it will require more grid capacity although it also provides more load flexibility (Williams et al. 2015; Jadun et al. 2017). Electrification measures create grid flexibility by providing an opportunity to shift large aggregate loads, such as vehicles and space conditioning, to provide SDR that complements renewable energy. However utilities and customers may not take advantage of that inherent flexibility absent requirements or motivation to do so. As utilities and states pursue electrification, they should incorporate strategies such as smart charging to strategically manage new loads in a way that does not exacerbate peak demand and that reduces system costs. Policy should require that utilities optimize new load growth and strategically deploy energy efficiency and SDR when and where it is needed most alongside electrification. However utilities are likely to pursue electrification to increase volumetric sales, and thus may not require the same scale of performance incentives to pursue SDR for new electric end uses.
All of the complementary strategies discussed above require robust EM&V that creates confidence in achievements and provides both information on possible improvements and data for additional analysis. This is important for SDR PIMs as well.
State commissioners, regulators, and utilities are experimenting with newer forms of SDR PIMs to achieve what traditional cost-of-service regulation has not: load flexibility, load shaping, and load responsiveness. To achieve the estimated economic benefits of more than $15 billion annually from load flexibility, legislators and regulators need to use the many tools at their disposal to encourage utilities to integrate SDR as a core part of their business operations (Hledik et al. 2019). The states studied in this report show a growing interest in and experience with SDR to achieve a variety of goals, including cost savings for consumers and reducing GHG emissions.
Within the next wave of states considering changes to the utility business model, many are considering incentivizing SDR. New Hampshire is currently considering updates to its performance incentives, and its commission staff members have put forth a proposal that includes incentives for demand reductions (New Hampshire Performance Incentive Working Group 2019). Recent legislation in Washington opens the door for demand reduction PIMs, as it explicitly calls for demand response targets and establishes authority to create performance-based rates (Washington Legislature 2019). Minnesota’s performance-based metrics proceeding highlights the “cost-effective alignment of generation and load” as a key policy outcome (MN PUC 2019). Hawaii selected DER asset effectiveness as a key outcome in its PBR docket (HIPUC 2019d). Michigan’s DTE has proposed a demand response PIM that has not yet been approved by the commission (DTE 2019). California utilities are piloting “pay-for-performance” efficiency programs that target SDR applications and place requirements on customers for achieving greater value through energy efficiency (St. John 2019).
This research highlights successful elements of SDR PIMs for consideration as more jurisdictions move toward performance-based regulation. Regulatory reform in conjunction with robust programs can reduce demand strategically to transition to an affordable, resilient, and clean electricity system. Complementary policies can help to reduce SDR conflicts with the cost-focused utility business model. It is also important to continuously evaluate what each PIM is incentivizing and how effectively it is delivering benefits to customers. Well-designed SDR PIMs can help move utility behavior to align with desired policy outcomes including reduced customer costs, improved reliability, and reduced environmental impacts during a time of rapid change in the energy sector.