Synthetic Load Profiles of Various Customer Types for Smart Grid Simulations

Grid planners often confront situations where only few to even any measured data from loads is available for their grid simulations. This paper presents a methodology that allows the generation of synthetic load profiles for commercial, industrial and agricultural customers. These profiles show characteristics that are more realistic and variable than traditional standard load profiles, providing grid planners with an alternative to simulate grids when few to none measured data from loads is available. The proposed method is based on a stochastic linear regression model, which is applied to several measured load profiles of real commercial, industrial and agricultural customers of German distribution system operators (DSO). A set of parameters is obtained from the datasets for each type of customer, facilitating the production of new synthetic load profiles representing them. The application of a power-energy-denormalization (PED) method allows maintaining the annual amount of energy as well as the maximal power value of each profile intact, based on measurements of metering devices. The obtained synthetic profiles were validated against real measured data. Additionally, the aggregation of individual synthetic load profiles at a low voltage substation was researched within this paper. The performed validations show a good correlation between modelled and measured load profiles, at both the individual and aggregated levels.