Measurement based analysis of active and reactive power losses in a distribution network with wind farms and CHPs

In Denmark, a large part of the electricity is produced by wind turbines and combined heat and power plants (CHPs). Most of them are connected to the network through distribution systems. Because the production units, the distribution systems and the transmission system are owned by different companies, rules regulating the conditions and fees for transmission of power apply. In Western Denmark, the TSO has made a regulation specifying bands for the exchange of reactive power between the 60 kV and the 150 kV networks. Further, the TSO is obligated to compensate the distribution network operator for the losses on the 10 kV radials fed by wind turbines. In order to make fair regulations, to optimize the operation of the distribution systems and to be able to utilize wind forecasts for planning, it is important to know, how the decentralized production affects the active and reactive power losses in the system. In a liberalized market, this knowledge can be used to avoid cross subsidizing in the transmission and distribution fees of consumers and producers [1], to generate incentives of the participants to change the consumption or production in periods with congestion [2;3] or to estimate the value of distributed generation in an area [4]. In systems where the investments and operation are partially or fully centrally controlled, the allocation of losses can be used to optimize the operation and investments and to minimize the losses. In periods with high load demand, the production from the decentralized production units will cover consumption in the vicinity. This can contribute to the reduction of the system losses. On the other hand, in periods where the decentralized production exceeds the local demand, it will cause the losses to increase. It is therefore clear that the influence of distributed generation on the system losses is determined by the location of the production units, the amount of distributed generation relative to the load demand and the simultaneity between demand and production. The aim of this paper is to describe these three effects in a simple way.