Information gap decision theory for operation of combined cooling, heat and power microgrids with battery charging stations

The role of combined cooling heat and power (CCHP) microgrids (MGs) in sustainable cities, due to high penetration of renewables and coordinated supply of different energy demands, is significant. CCHP MGs have energy management systems that conduct unit commitment (UC). In this research, a risk-aware information gap decision theory (IGDT)-based energy management system is developed for CCHP MGs with battery charging stations, considering ON/OFF history of power/heat/cooling units, minimum up-time and minimum down-time constraints, start-up and shut-down ramp rate limits. The schedule of components in CCHP MG is determined from the view of risk-seeker and risk-averse decision maker and is compared with the case in which the uncertainties are ignored. In this research, the uncertainties of demands and wind power are taken into account. The operation of CCHP MG is modeled as a mixed-integer linear problem (MILP) and is solved with CPLEX solver. The case study is an isolated wind-integrated CCHP MG with power generation units, combined heat and power (CHP) units, boiler, heat recovery unit, absorption chiller, electric chiller, battery charging station, storage systems and adjustable electric, thermal and cooling loads. In risk-averse IGDT-based UC, a robust scheduling with respect to uncertainties of electric, thermal and cooling demand and wind power is achieved; in this way, the decision maker is sure that no realization of uncertain parameters in real-time may result in an operation cost above critical operation cost. For instance, in the case considering the uncertainty of electric demand, MG operation cost will not exceed $1568.8586 if real-time realization of electric demands fall within the 6.7% band around nominal electric demands. On the other hand, risk-seeking IGDT determines the least needed deviation of different uncertain parameters which make a target operation cost accessible; in this way, the best realization of uncertain parameters in real-time results in an operation cost equal to the target operation cost. The results for risk-seeking and risk-averse cases are compared and the impact of risk on the schedule of MG components is meticulously investigated.