In this talk, we discuss recent efforts in developing analytics for estimating the potential short-term (i.e. intra-hour) reserve capacity that could be offered by a large number of controllable thermostatic loads (TCLs) through demand-side flexibility aggregators. TCLs include all types of electrical appliances with inherent thermal storage/inertia; for example, electric water heaters, space heating/air conditioning in buildings, fridges and freezers. The analytics proposed serve to estimate the equivalent power consumption that could be potentially dropped by turning off some portions of a TCL population as well as the uncertainties related to this estimate. We are showing how local appliance controls can be implemented and set to enhance the technical performance of the potentially freed-up power capacity at the power system level. Either through factory settings or possibly through online communications, a demand-side flexibility aggregator could influence (1) the response volume, (2) its uncertainty, (3) the sustainability of the TCL population response over time and (4) its post-response (i.e. reconnection) behaviour. The goal here for the aggregator is to have good tools to be able to offer commercially-interresting capacity volumes with technical characteristics which would make demand-based reserve an appropriate substitute to reserve provided by generation. We illustrate the validity and applicability of the analytics and control philosophies with the aid of a comprehensive case study based on a population of 10,000 electric water heaters in a Canadian setting.
Group for Research in Decision Analysis