A probabilistic approach to observability of distribution networks

This paper presents a novel probabilistic approach to distribution network observability. The observability analysis is an important part of the state estimation process. Traditionally it determines whether the state of a network can be estimated based on the available set of measurements. This works well in transmission networks, where numerous metering devices are normally available. In distribution networks, however, only a few metering devices are usually installed. Pseudo measurements with large margins of error are often used in the absence of real measurements to perform state estimation. This implies that if a large number of pseudo measurements is used to make a network observable, the estimated state can be significantly different from the actual state even if the network is classified as observable. To overcome this limitation a new approach to observability is proposed in this paper. The proposed method takes the uncertainty of the state estimation into account, and therefore, assesses the network observability depending on the accuracy of the estimated network state. This paper also demonstrates how a meter placement method based on the proposed observability assessment can be implemented. The presented methods have been tested on a modified 34-bus IEEE test feeder. Results are compared with existing methods.