Evaluation of economic impact of external eauivalent models used in reactive OPF studies for interconnected systems

Reactive-optimisation procedures are responsible for the minimisation of online power losses in interconnected systems. These procedures are performed separately at each control centre and involve external network representations. If total losses can be minimised by the implementation of calculated local control actions, the entire system benefits economically, but such control actions generally result in a certain degree of inaccuracy, owing to errors in the modelling of the external system. Since these errors are inevitable, they must at least be maintained within tolerable limits by external-modelling approaches. Care must be taken to avoid unrealistic loss minimisation, as the local-control actions adopted can lead the system to points of operation which will be less economical for the interconnected system as a whole. The evaluation of the economic impact of the external modelling during reactive-optimisation procedures in interconnected systems, in terms of both the amount of losses and constraint violations, becomes important in this context. In the paper, an analytical approach is proposed for such an evaluation. Case studies using data from the Brazilian South-Southeast system (810 buses) have been carried out to compare two different external-modelling approaches, both derived from the equivalent-optimal-power-flow (EOPF) model. Results obtained show that, depending on the external-model representation adopted, the loss representation can be flawed. Results also suggest some modelling features that should be adopted in the EOPF model to enhance the economy of the overall system.