New Strategies for Control of the Electrical Maximum Demand

Abstract

The object of this investigation is to develop new strategies that minimise the maximum demand on the electric power system, and maximise plant utilisation, for ensembles of electroheat processes. Two classes of ensembles of similar on-off electroheat processes are thoroughly examined. The first ensemble is of magnitude ten or less, as when the processes are installed in a factory. The second is of large magnitude, as in domestic space and water heating. New strategies and important design criteria are presented for demand control of loads that are switched (a) deterministically, and (b) by thermostat. In the latter case, control decisions are based upon prediction of load by monitoring of past energy consumption and of process temperature states. New models are established to determine the statistical properties of the demand due to ensembles of similar two~position space heating processes. Results are applied to determine the penalty incurred, as a probability function, due to the demand exceeding a set level. These results are of particular value in assessing the economic viability of proposed space heating installations. The models are verified by digital simulation using pseudo-random numbers, and by sampling from known probability distributions; computer programs are presented. Switching characteristics are investigated experimentally for a number of similar thermostats, operating in turn within an environmental test chamber. Two criteria of demand optimisation are thoroughly analysed: 1. The maintenance of a "minimum-comfort" level for space heating. 2. Minimisation of the cost per unit product for a specific industrial electroheat process. A case is presented for control of the maximum demand by digital computer. A proportional plus integral controller is designed for continuous demand control of an electrothermal process. The system performance is simulated on an analogue computer.