Fast Demand Response with Cooling Devices Semester

Share of generating units from renewable energy sources is constantly increasing and introducing stochastic nature into the system. This results in undesirable fluctuations of mains frequency and thus, a rising need for frequency regulation services. Providing sufficient power reserves presents a complicated task and usually requires special types of fast response power plants running part-loaded. It would be a great contribution to power system efficiency if existing generators would run at its full potential, i.e. were excluded from frequency regulation. One of possible solutions for this issue, which is analyzed here, is demand side management via dynamic demand response. Presented method is based on thermostatic control of domestic cooling devices, in this case refrigerators, and modification of their energy consumption according to frequency deviation signals. Some previous studies showed that deterministic approach to demand response results in a synchronizing effect of individual appliances which leads to unacceptable levels of energy demand after steady state recovery. Other papers have addressed the issue of communication network between devices and proved that centralized method is not justified from an economic point of view. After taking these conclusions into consideration, the focus of this paper is on introducing stochastic behavior of controlled frequency responsive devices by means of designing decentralized random algorithm. This approach should enable faster regulation of system frequency during normal periods of operation, as well as avoiding critical contingency states and instability phenomena.