A time series probabilistic synthetic load curve model for residential customers

For the planning and operation of low-voltage systems models of the loads are needed. Little is known about residential loads besides the fact that they have a diverse behavior. This means that each residential load has seldom a maximum demand and the maximum demands of several customers do rarely coincide. As the number of residential customers increases the load curve of the group is becoming smoother. The utilities have information about the customers´ yearly energy consumption and can estimate the maximum load of a group of customers. So far this information was sufficient since little attention was turned on the low-voltage networks. With the development of smart grids and the integration of distributed generation a better understanding of residential customers has to be obtained. With the proposed bottom-up approach residential loads can be composed by looking at each possible and relevant appliance, respective power consumption, frequency of use, turn-on time, operating time as well as potential correlation between appliances. Measurements were carried out on different types of appliances to determine individual load curves and to analyze the sequence of operation. The uniqueness of the presented model is the computational step time of 30 seconds for a single residential load.