Reduced models for operation management of distributed generation

Distribution grids evolve more and more to a decentralised structure with a high amount of distributed generation. With a high penetration of DER and RES it will be necessary to control the generators to avoid instabilities in the grids. Decentralised generation also offers the possibility to manage the load flow of the voltage level they are connected. Experience has shown that the complexity of a detailed distribution grid model is far too extensive for the operation management. For an efficient calculation, model reduction techniques may be employed for reducing the model complexity and thereby the computational effort. The reduction method presented in this paper focuses on the aggregation of grid segments, which are reduced to their dominant properties. Using the mathematical method of symbolic model reduction, the model of a low voltage grid will be analysed. Depending on the load and generation within the grid, the reduced model will reproduce the behaviour of the voltages and currents at the points of connection in comparison to the original detailed model. The reduced model shows a small but controllable deviation resulting from the application of the model reduction techniques.