23 rd International Conference on Electricity Distribution Lyon, 15-18 June 2015 Paper 1510 CIRED 2015 1/5 APPLICATION OF DISTRIBUTION SYSTEM STATE ESTIMATION ON ENGINEERING INSTRUMENTATION ZONES OF LOW CARBON LONDON Jelena DRAGOVIC Danny PUDJIANTO Predrag DJAPIC Imperial College London – UK Imperial College London – UK Imperial College London –UK j.dragovic@imperial.ac.uk d.pudjianto@imperial.ac.uk p.djapic@imperial.ac.uk Mark J. BILTON Goran STRBAC Imperial College London – UK Imperial College London – UK mark.bilton04@imperial.ac.uk g.strbac@imperial.ac.uk ABSTRACT This paper discusses the performance and presents potential benefits of the application of State Estimation on real distribution network. The application is tested using real network measurements obtained during Low Carbon London project. It demonstrates that the deployed distribution system state estimation, through a limited number of optimally placed sensors with adequate accuracy, can robustly estimate voltage and power flows in high voltage distribution networks. INTRODUCTION Low Carbon London (LCL) [1] is a £28.3 million project funded by the Low Carbon Networks Fund (LNCF) [2] to identify innovative solutions to maximise the benefits of a wide range of Low Carbon Technologies (LCT) on London’s electricity distribution network (DN). The deployment of LCT controlled by active network management requires voltages and power flows across the network to be monitored closely in order to enable optimized control actions of LCTs and network control devices, in a coordinated manner in real time. However, in contrast to the national transmission system, where measurements are widely deployed [3], available measurement infrastructure in High Voltage (HV) DNs, is not sufficient to facilitate real time control, essential for the evolution to the smart-grid paradigm. Thus, additional measurements will need to be established to support the implementation of innovative real time active DN management practices necessary to facilitate cost effective integration of low carbon demand and generation technologies. Due to vast number of nodes in a DN, fully instrumenting the network would have a prohibitive cost. A special tool is required for Distribution System State Estimation (DSSE) which could be used to reduce the number of meters needed in a way to ensure that adequate network observability and the accuracy of state estimation (SE) can be achieved over a range of different operating scenarios. Although transmission SE algorithms have a central role in the operation of the transmission system for some decades [3], they cannot be applied directly to the DNs because of their different design and operating practices. There are a number of methodologies for DSSE proposed in the literature. A good overview of transmission and distribution SE can be found in [3]-[5]. The majority of the existing research on DSSE is performed on generic networks which have ideal network parameters data and measurements. However, when applied on the real network, the user of DSSE faces multiple practical issues. This work presents an effort in examining the role and possible application of DSSE in the UK DNs. Measurements carried out within LCL Engineering Instrumentation Zones (EIZs) demonstrate that the developed prototype DSSE, through a limited number of optimally placed sensors could robustly estimate voltage and power flows in HV DNs. This paper firstly introduces the methodology applied. Secondly, it discusses the results of DSSE in terms of voltage, consumption and power flow and the accuracy of estimation for the peak demand condition. Thirdly, the results of rigorous testing of DSSE model and analysis on the accuracy and robustness of voltage and power flow estimates using half-hourly data across one year are presented. This is followed with a study demonstrating that the way the pseudo measurements, as alternative for the missing real measurement data, are modelled has an important impact on estimation quality. Impact of sensors availability and quality to SE is discussed afterwards. Finally, it demonstrates the application of DSSE along EIZ feeders to determine the optimal number and locations of new sensors and recommends rules of thumb for their location. METHODOLOGY Having in mind the limited number of real measurements, the basic task of a DSSE is to estimate as accurately as possible the true state of a system (i.e. voltage and power flow profiles), using any relevant available information. The DSSE model deployed in this work is based on the maximum-likelihood estimation of state variables which employs Weighted Least Square (WLS) formulation, and is recognised as one of the most suitable techniques for state estimation in DNs [5]. WLS methodology works on the principle that the difference between measured and true values is minimised taking in consideration weight of each measurement, which is inversely proportional to the accuracy of the measurement. The Newton iterative