Reactive power analysis in a 10 kV distribution grid

Falbygdens Energi AB is a local grid owner in Västra Götaland county and have had an increase in the amount of cable in the grid during the last decades. This increase in cable length, combined with a lack of reactive power monitoring, has led to the reactive power situation in the grid not being fully investigated. This paper aims to examine where the reactive power in the grid is being consumed and produced, how the increase in capacitance affects the company's current arc supression coils, and how the reactive power affects contract limits towards the regional grid owner. Finally, appropriate reactive power compensation is suggested. By creating a network model that excludes minor contributions to reactive power, including line inductance and resistance, a simplified one phase-thevenin equivalent circuit is developed that only includes the capacitance of 10 kV-cables, leakage reactance of distribution transformers and load reactance. The results show that the contract limits are exceeded during summer when there is low or no load connected, but comparisons with contract data show that the developed model overestimates the reactive power consumed in 10/0.4 kV-distribution transformers. The arc supression coils being used today are determined to be under-dimensioned, with coils in stations T2 and T5 being nearly 50 % of the size required with regards to cable length. Furthermore, the simulation shows that a variable shunt reactor located at the furthest node in each line yields the lowest over voltage during low load conditions and can compensate for the reactive power exceeding the contract limit, with a reactor of 175 kVAr being appropriate for station T5. The method used to determine the production and consumption of reactive power in each outgoing line is shown to be accurate regarding the reactive power produced in cables, but inaccurate regarding the reactive power consumed in distribution transformers. An inquiry with more accurate transformer leakage reactance data, as well as an investigation concering the economical aspects of the shunt reactor compensation is recommended