Demand Response Evaluation

We have evaluated several demand response programs offered by FirstEnergy to residential, commercial, and industrial customers. During the summer of 2012, West Penn Power offered a smart-meter based critical peak pricing program to residential customers. We have evaluated the program by running hourly regressions on whole house interval meter data for eight separate participant groups and corresponding comparison groups. In 2012, Met-Ed ran two-way thermostat based direct load control program, while Penn Power and Penelec operated direct load control using one-way thermostats and switch programs were evaluated by installing data loggers on customers’ air conditioners. All companies also offered load curtailment programs to commercial and industrial customers.  We evaluated the commercial and industrial programs through application of various customer base load (CBL) algorithms and custom regressions. The evaluation effort for 2012 culminated with careful accounting of all portfolio impacts during the top 100 hours of load for each utility company, along with a customized calculation of marginal line losses during peak hours. The companies offer demand response programs in 2017-2020.  We evaluated the 2017 nonresidential demand response programs through application of various CBL algorithms and customer-specific regressions, and then performing a meta-analysis to combine results of the top three performing baseline load constructs.  In 2017, Penn Power also offered a behavioral demand response program for residential customers, administered by Oracle (formerly OPower).  The program was set up as a randomized control trial, and we used a variant of a “lagged seasonal” regression model to measure program impacts.