| Summary: | The objective of this thesis is to create a model, which provides a detailed description of the electrical load on a low voltage distribution network in the context of a typical UK urban feeder from a primary transformer. The aim of the work, when used in association with a power flow analysis package, is to help to provide a risk assessmenfto r over-voltagee vents and over-heatingi n the network when different levels of embedded generation are applied. A fine-grained domestic load model has been constructed with a three layered approach-to provide per-consumer, 1-minute averaged loads on an end-use basis. Datasets from the Load Research Group have been used as the starting point and form the basis for layer 1 which represents group-averaged demands on a halfhourly basis. Layer 2 of the model introduces diversity in terms of number of occupants, living space, ownership and soci?- economic factors. Layer -3 uses appliance duty cycles to create wider variations by random triggering to derive 1- minute loads from assigned half-hourly values. The domestic model has been adaptedf or use with smaller( sub MOW) non-domesticc onsumers. The research question for this study is whether or not the models provide an adequate representation of the electricity demand for a typical urban LV network, judged in terms of a variety of parameters. The output from the domestic model comparesw ell with measuredd ata giving realisticd emandc haracteristicsin terms of mean, peak, load factor and distribution. Compared against diversified peak demands currently in use within the industry, the model estimates values within 10% for groups fewer than 25 and 5% for groups of 100 or more. When used together with a power flow analysis package, the predicted voltage variation agrees with measured results in terms of mean value and distribution. The investigation of time and group averaging of demand, power factor surveys and, with a matching model for PV and solar thermal output, studies into electrical demand reduction within mixed communities are all possible additional applications for the model.
|
|---|
