Loss minimization in distribution networks with multiple load scenarios

In distribution networks, the problem of finding the radial configuration that minimizes losses is usually solved for the peak load scenario, since this is the worst case situation regarding both losses and network loading. In general, the solution will remain optimal for other load levels, provided that there is an homothetic load variation. However, in some situations, the load pattern in the entire network may change during the day, or in seasonal periods, leading to situations where the solution is not optimal outside the peak period, or outside a specific period. On the other hand, the presence of dispersed generation directly connected to the distribution network may also contribute to these situations. This paper addresses this issue by considering a set of load scenarios and finding single or multiple configuration solutions that minimize both energy losses and switching actions, in a bicriteria framework. The optimization engine uses a simulated annealing algorithm previously reported to give good results in this class of problem. The methodology is illustrated with an example.