No abstract

No abstract

SUMMARYThe nature and control of existing distribution networks limits the amount of distributed generation that can be connected. To increase the penetration of distributed generation a distribution management system controller (DMSC) can be used. The use of a DMSC requires a state estimator algorithm that provides an estimate of the network state in real time. In this paper, a state estimation algorithm based on the methods used on transmission networks is presented. The application of these methods into distribution networks requires existing measuring systems to be supplemented with the addition of new real-time measurements, and the use of load estimates.

The interaction between a power system and a wind-farm can be studied using either a simplified ‘aggregated’ or a complex ‘detailed’ model. The detailed-model results compare well with the actual, therefore, provided that both models have similar dynamic characteristics, the aggregated model provides a rapid and cost-effective way of representing large wind-farms in power system dynamic studies. A brief overview of wind-farm aggregation methods is given and their application in power systems analysis is discussed. Aggregation techniques are then applied to a wind-farm with thirty wind turbines. It is shown that the aggregation of the swing equation and equivalent representation of network and generator impedances can be used for aggregated representation of fixed speed wind turbines. A similar approach applied to the d-and q-axis controller of a doubly-fed induction generator (DFIG) failed initially to provide satisfactory dynamic characteristics with the aggregated model. Consequently, a simple method was used to scale the rotor currents of the controller, so the dynamic characteristics with DFIGs became acceptable.

High penetration of distributed generation is currently limited by passive operating methods of distribution networks. At distribution level, insufficient measurements are available to allow satisfactory control, so measurement is extended through the use of state estimation. Aspects that are described in this paper include optimal measurement location, the problems occurring with few measurements, and load models used for "pseudo-measurements". Two control philosophies are described. The first, applicable to smaller network segments, is a local control, working with existing control devices such as transformer automatic voltage control (AVC) relays. The second philosophy is optimal control with constraints.