Consumer Baseline Estimation Analysis for Implementation of Demand Response Programs

Abstract

ABSTRACT Demand response (DR) is the short-term change of consumers energy usage patterns in response to power utility needs. The utility benefits from DR adoption in terms of better load management, reduction of peak power deficits, energy deficits, enhanced reliability, and improved economic performance. Consumers gain from energy bill savings and incentive payments. Accurately estimating the consumer baseline load (CBL) is critical for evaluating the success of a DR program implementation. Estimation of CBL helps to decide the volume of load curtailment, decide the incentives to be offered, and assess the impacts of consumer participation. The real-time domestic 11 kV feeder data of the 6.3 MVA substation of Goa electricity board, Goa, India is used to analyze the performance of the DR program. For CBL analysis time series average, adjustment, exponential moving average, decomposition, and forecasting techniques are applied. For performance evaluation of DR programs, the Artificial neural network (ANN) based CBL estimation technique is utilized. The ANN-based technique of estimation has better prediction accuracy, lower bias, and reduced variability as compared to other CBL estimation techniques. Consumer response to demand changes is computed using elasticity and consumer benefit function. The ANN-based CBL estimation and demand variation obtained from elasticity values are further used to analyze the technical and financial performance factors, for dynamic price and incentive-based DR program implementation. The research can assist a state power utility that is implementing smart grids in categorizing consumers, establishing a baseline, identifying potential demand reductions, and evaluating the feasibility DR programs.