Generation of synthetic hot water demand profiles with dynamic Markov chains models for end-user-friendly demand response

• Main Authors: Binini, Giscard M., Munda, Josiah L., Popoola, Olawale M.
• Format: Conference Proceeding
• Language: English
• Subjects: Demand Response; Hot water demand modellin g; Markov Chains; Power engineering; Renewable energy sources; Stochastic processes; Temperature distribution; Temperature measurement; User comfort; Water; Water heating
• ISSN: 2771-7011
• DOI: 10.1109/REPE59476.2023.10511873

Demand response (DR) programmes have been shown to cost-effectively support the safe operation of the electric grid where a significant portion of the energy is generated from renewable sources. These programmes have been implemented via direct or indirect control of the end-user electrical loads. While financial incentives have shown to generate interest in the industrial and the commercial end-user to participate in the programme, this alone fails to entice residential end-users to voluntarily participate in DR programmes. To generate interest in the residential sector, a DR with a low impact on user comfort is needed. When electrical water heaters are targeted for a direct DR scheme, the user comfort is measured by the water temperature of the hot water when needed. To assess this impact without the need to implement costly real-time measurements, a generalised hot water demand that could capture the characteristics of hot water consumption is required. In this paper, multidimensional dynamic Markov chain models are used to generate synthetic hot water demand profiles that take into consideration the stochastic nature of water demand over time without neglecting additional factors that would influence the water demand profile. The dynamic models developed in this study have the propensity to provide generalised hot water demand profiles that consider seasonal and temperature influences in the model's development. Although the seasonal model performs slightly better than the temperature-based model, both approaches were able to capture the underlying variables determining the hot water demand for domestic electrical water heaters.