Innovative Approaches to Verifying Demand Response of Water Heater Load Control

This study describes a pilot effort to measure load reductions from a residential electric water heater (EWH) load control program using low-cost statistically based measurement and verification (M&V) approaches. This field experiment is described within the larger framework of overcoming barriers to participation of noninterval metered customers in Demand Response (DR) Programs. We worked with PJM Interconnection and a Curtailment Service Provider (CSP) to collect hourly load data for two substations and several hundred households over six weeks of load control testing. The experimental design reflected constraints imposed by limited funding, manpower, equipment, and the routine operation of the load control system by the CSP. We analyzed substation- and premise-level data from these tests in an attempt to verify several "point estimates" taken from the hourly diversified demand curves used by the CSP to establish aggregate load reductions from their control program. Analysis of premise-level data allowed for provisional verification that the actual electric water heater load control impacts were within a -60 to +10% band of the estimated values. For sub-station level data, measured values of per-unit load impacts were generally lower than the CSP estimated values for Electric Cooperative #2, after accounting for confounding influences and operational test problems. Based on this experience we offer recommendations to ISO and utility DR program managers to consider before undertaking further development of alternatives to the conventional but costly program-wide load research approach.