PJM recently released an updated 2019 Peak Load Forecast, the primary change being the inclusion of approved Peak Shaving Load Adjustments for summer-only demand response programs (report and supporting data available at: http://www.pjm.com/library/reports-notices.aspx).
Demand Side Analytics has prepared a report for the Consumer Advocate of PJM States which provides a high-level overview of the PJM change and explores implications for program administrators. We focus on three primary areas: 1) understanding load forecast adjustments and the implications for participation and timing, 2) Offer Strategy and Considerations, and 3) Price Suppression Effects. The full report is available here.
Peak Shaving Adjustments
Historically demand resources such as demand response and energy efficiency have entered the market as supply and been eligible to compete alongside traditional supply side resources (power plants) in a competitive auction to fulfill the resource requirements for the region. Demand response resources such as utility direct load control of central air conditioners have recently encountered difficulty participating in the market due to PJM’s “capacity performance” definition of generation capacity. A Peak Shaving Adjustment (PSA) offers a fundamentally different means for demand response to participate in the Reliability Pricing Model (RPM). Instead of being treated as supply that is capable of fulfilling resource requirements, a Peak Shaving Adjustment enters the market on the demand side. The characteristics of the shaving are used to create modified peak load forecasts.
In the report we discuss the factors that affect how a Peak Shaving resource will affect the Variable Resource Requirement, key design components as adopted by PJM, and the implications of barring dual participation.
The peak shaving “pledge” happens before the auction so there’s some uncertainty with regard to the value of a commitment it is made. If you have a state program/resource, or are contemplating developing one, how do you balance maximizing the load forecast adjustments while maintaining cost effectiveness? For example, would it be better to shave 100 MW for three hours on all days hotter than 95 degrees, or shave 50 MW for 5 hours on all days hotter than 90 degrees? In the report we explore the effects of:
- System Load characteristics – how the amount of summer vs. winter peaking risk affects compensation, and considerations of event frequency vs. duration.
- Weather – varies from year to year, but commitments are based on THI thresholds. If predicting performance based on median weather what is the risk in extreme weather years and the cost/benefit calculus of underperforming.
- Customer rotation – how frequently can customers reasonably be called without fatigue?
The resource clearing prices in the PJM BRA are a function of zonal demand and the cost of resources available to meet those demands. Reducing peak capacity requirements generates value both by avoiding the costs associated with the load being shaved, and potentially by lowering the price for the remaining capacity that still must be procured. This second component is the price suppression effect. In reality the VRR is not a curve, but a staircase with tread width the size of power plants. Thus there is no guarantee that reducing peak will reduce the clearing price. Using PJM BRA sensitivity analyses from prior years, we estimate the slope of the supply curve for different market segments and provide bounds on the potential price suppression effect.