Forecasting flexibility in electricity demand with price/consumption volume signals

The introduction of renewable energy sources, particularly wind power, is limited by their dependence on weather conditions and by the difficulty of storing surplus energy for use at times when production is low. One effective way of tackling the energy storage problem is to minimise the need for storage, i.e. to switch from a system based on producing electricity in response to the unpredictable whims of demand to one in which consumption adapts to supply. Demand can be managed indirectly via the sending of price/consumption volume signals. This paper presents a mathematical model for forecasting the aggregated electricity demand of a group of domestic consumers signed up to an incentive-based demand management programme. Under this programme consumers receive signals that offer financial incentives for limiting their volume of consumption at time intervals when system peak demand is forecast. The resulting optimisation model is a mixed-integer linear programming problem implemented in JAVA and solved using free software. This model is applied to a case study in which the objective is to limit consumption by a population of 15932 consumers from 15:00 to 17:45 on a specific summer day. The responses to two different incentive amounts are shown.