Optimal dispatch of interruptible and curtailable service options

Service options for the electric utility industry are increasingly including provisions to dynamically dispatch prices or service interruption calls to customers. When the number of such calls is contractually limited, then the issue of optimally dispatching the calls must be addressed. In this paper, an algorithm for optimally dispatching interruptible and curtailable (I/C) service options is developed within a stochastic dynamic programming (SDP) framework. The solution yields thresholds for calling load curtailments as a function of: 1) the number of days remaining in the period, 2) the available number of interruptions, and 3) the level of uncertainty regarding future price and cost fluctuations. The SDP approach also provides a framework for optimally designing the I/C program parameters. An illustration of the dispatching routine is provided using data from Niagara Mohawk Power Corporation (NMPC). The algorithm developed in the paper is currently being employed by NMPC as part of its I/C program for industrial and commercial customers.