State estimation in MV distribution networks: experience in the Spanish smart grid project PRICE-GDI

Abstract

This paper shows a practical implementation of a state estimator for distribution systems in the context of the Spanish pilot project PRICE-GDI. The objective of this pilot project has been to develop a set of key technologies enabling the massive integration of renewable generation in distribution systems. Undoubtedly, dispatch centres involving advanced algorithm such as distribution state estimators and volt/var controllers are necessary for this challenge. This paper has analysed the implementation of a state estimator taken into account all the specific characteristics of the distribution systems: large proportion of injections in the measurement set, long and short lines coming to the same bus, presence of current flow measurements to the detriment of power flows, high R/X ratios and, mainly, reduced number of on-line measurements. As a matter of fact, this last issue is the responsible of the lack of observability of the system being necessary to incorporate to the classical estimation algorithm a load allocation procedure providing a number of pseudomeasurements. This paper proposes to use a methodology using Daily Load Curves (DLCs) to allocate the load of the Distribution Transformers (DTs) which obtain better results that the ones obtained using a classical procedure resorting on the contracted power of DTs. The paper is structured as follows. First, the introduction gives some information of the Spanish pilot project PRICE-GDI and the motivation of the paper. Then, a brief explanation of the state estimation algorithm specially intended for MV distribution networks follows. After that,the DMS architecture and how the algorithm has been integrated are described followed by a description of the feeder that has been used to test the algorithm and the main obtained results. Finally, the paper closes with the main conclusions.