Optimal system security under capacity constrained conditions

In a deregulated power system relying on an energy spot market only, it is not possible to guarantee system adequacy. Thus, as long as demand is de facto inelastic in the short run, it is not possible to maintain traditional system security under all circumstances. The normal way to maintain power system security is by way of reserve requirements. This paper addresses the assessment of system security under capacity shortage conditions, where strict adherence to reserve requirements would lead to load shedding. A mixed integer model is used to optimize the use of generation resources with a one-hour time horizon, with the objective to minimize the sum of operational costs and the costs of load disruption. The model is demonstrated on the IEEE RTS-79 test system. It is shown how optimal security can be obtained without the use of fixed reserve requirements. It is also shown that different measures of system security like the loss of load probability, expected energy not served or the probability of system collapse, result in solutions with different security characteristics. The paper concludes that the use of fixed reserve requirements, practical and economical under normal circumstances, leads to high social costs under extreme demand conditions because of the necessity to shed load. Instead, reserve levels should be chosen such that an optimal balance between total costs and security is obtained