This study supposes an energy service company (ESCO) that serves as an electricity mamagement aggregator for a community. The ESCO, here, procures electricity from various types of source via a two-settlement system that includes a day-ahead (ex-ante) and a real time (ex-post) transactions. The paper forcuses on the demand and supply management for a single time slot (30 minute trading interval) out of for example, 48 possible slots in a day. In the two-settlement system, each generator submits a two-part offer, to the ESCO, that includes the prices for day-ahead and real-tme offers as well as its generation limit. The community, here, also owns a wind farm supplying a free electricity that, however, presents a major uncertainty in addition to the demand uncertainty. The ESCO, as the service operator, has multiple choices for electricity procurement, such as forward contracts, real-time trading, and call options for capacity as well as for load-shedding, i.e., demand-response arrangements with customers. The stochastic programming formulation is shown to well accommodate the strategic decision as well as the tactical decision involed in the two-settlement market. A numerical example demonstrates the significance of having those derivative-type participants for strategic decisions under uncertainty and risk.