Towards optimal load management with day ahead pricing

Demand Response is under implementation throughout the globe by many utilities to incorporate the end user as an active player in reducing supply-demand imbalances. Day-ahead pricing is provided as an option to schedule electric loads so as to take advantage of time-varying prices. However, user convenience is also a factor that must be taken into account, as users may be willing to forego some savings to reduce inconvenience. We formulate an optimal scheduling problem considering both aspects. As the search space is exponentially large, we propose two greedy algorithms to find good schedules. To assess performance, we obtain the optimal schedule via Markov Chain Monte Carlo (MCMC) based simulations. We apply the framework to two case studies; one study uses appliance energy profiles obtained by actual measurements using the Joule Jotter, a device designed and developed in-house. Results indicate that the proposed algorithms perform very well, achieving performance within 10% of the optimal.