Development of a fault detection and location protection scheme based on the application of smart meters

Abstract

Global weariness of rolling blackouts and increasing concerns for the climate has spurred the demand for reliable renewable energy. The result is the microgrid which is changing the landscape of electricity supply. Local power generation, storage and consumption has increased electricity reliability to the consumer. However, various protection issues associated with the dynamic behaviour of the microgrid arise. The objective of the proposed research is to formulate a fault location algorithm based on voltage measurements from smart meters in an AC microgrid. Fault detection and location are complex tasks because of the varying network operating conditions. The proposed approach seeks to reduce that complexity by formulating a voltage-based location algorithm. This approach makes use of the global adoption of smart meters by increasing their functionality to include fault location. The benefit of using smart meters is that they are an existing network unit that can provide dispersed voltage measurements in the network. Dynamic Time Warping (DTW) is used to provide a comparison between pre-fault and faulted voltage measurements to detect the presence of a fault and the subsequent location. The novel application of this algorithm in this field allows for a robust, low data-intensive and simple solution to fault location. The study is approached by performing a manual design of a microgrid. A comparison is performed on the impact of different modelling techniques on the accuracy of the fault location algorithm. A review of the different protection strategies in the literature is presented and a case for the proposed approach is made. The influencing dynamic behaviour of the microgrid as well as the available protection devices are also discussed. The proposed algorithm proved accurate in the detection, classification and location of various fault types with differing fault resistances. The comparison method of the (DTW)algorithm allowed for accuracy to be maintained regardless of changes in the operating conditions of the network. However, the algorithm proved more suitable for an unbalanced network as opposed to a balanced network due to the increased available measurements from the different phases. It was noted that although the impedance modelling technique applied will affect any voltage-based protection solution, the algorithm developed in this study is robust enough to maintain accuracy. Furthermore, it was found that high-resistance faults were not the most determining factor in the accuracy of the algorithm, but it was the quality of the input data as well as the amount of data. Too little and low accuracy voltage measurements led to the algorithm showing high inaccuracies for the fault location estimation. Recommendations are then made for the measurement accuracy required of the smart meters as well as the minimum number in a given network