Stochastic Technical Losses Analysis of Smart Grids under Uncertain Demand

Technical losses of smart grids can be computed using the customer's smart meter measurements (active and reactive energy) and the energy measurement registered by the Low Voltage (LV) supervisor deployed at secondary substations. However, in some LV networks, some customers do not provide information regarding the energy consumed and produced in real time. This fact complicates the calculation of technical losses because this information is necessary for estimating the load demand for this subset of customers. In this paper, a stochastic approach is proposed for the estimation of technical losses in smart grids under uncertain load demands (e.g., non-telemetered customers and uncertain smart meters readings). Load demand estimation of non-metered customers was performed by means of a top-down approach. Intra-hour load demand profiles of customers were synthetically generated by applying a Markov process. The data and network used in this process corresponded to the roll-out deployed by the Spanish Research and Development (R&D) demonstration project OSIRIS.