Modelling customer demand response to dynamic price signals using artificial intelligence

The marketing efforts in Eskom have shifted to a broader perspective, embracing such targets as load shape optimisation and energy efficiency. There have also been organisational shifts, including the introduction of key customer focus groups to improve customer services to key customers who are energy-intensive end users of electricity. One way of building good relationships is to ensure that an optimal range of product packages (tariff schemes) is on offer. Customised electricity pricing agreements are usually arranged between an electricity supply industry and its key customers. These special agreements will not only provide sufficient financial incentives for the key customers to participate in the utility's demand-side management (DSM) programs, but also provide the utility with sufficient revenue. However, the design of these customised pricing agreements can be suboptimal unless a well-formulated methodology is followed. Such a methodology is proposed in this paper, and a knowledge-based end user demand response modelling tool to assist in this design methodology is also be discussed. A case study is included for illustration purposes.