A Disaggregated Nonhomothetic Modeling of Responsiveness to Residential Time-of-Use Electricity Rates

In addition to quantifying the load impacts of time-differentiated rates on aggregate peak, off-peak groupings, a very finely disaggregated Rotterdam demand system is described and estimated for explaining changes in detailed features of customers' electricity load patterns. This linear first difference formulation allows for nonhomotheticity, permits an examination of load impacts on critical hours of the week and makes use of both control and time-of-use data. Moreover, the proposed specification is parameter parsimonious. An illustrative use of this model is portrayed through recent empirical evidence from a residential experiment of a northern winter-peaking utility in Ontario, Canada