Statistical evaluation of voltages in distribution systems with embedded wind generation

Abstract

The commercial viability of wind-driven embedded renewable generation (ERG) projects is sensitive to connection costs. These costs vary with the voltage level at which the ERG is connected; generally the higher the voltage, the higher the cost. Windfarm developers of ERG therefore prefer to connect at lower voltages. However, because distribution networks are normally limited by permissible voltage fluctuations rather than by thermal considerations, network operators prefer to connect ERG at higher voltages where their overall impact on voltage profiles is minimal. This conflict of objectives between ERG developers and network operators is usually settled through simple deterministic load flow studies, usually based on one critical case representing conditions of minimum load and maximum ERG output. Such studies only consider extreme events and do not take into account the likelihood of these events occurring. However, new European standards describe voltage characteristics in statistical terms. Although these standards are not intended for use in assessing the level of penetration of ERG, they point to statistical and probabilistic approaches as the way forward in such assessments. This is particularly important for wind generation because of its stochastic output. Two simple wind speed models based on the Markov modelling technique are presented, which may be used to evaluate the impact of windfarms. The models give satisfactory results when applied to the evaluation of voltage profiles in distribution networks. The probability density function of voltage can be obtained and used in assessing the level of ERG that might be accepted on an existing network. A 281-node generic distribution system model based on a real network is used to illustrate the application of the proposed models.