A method for profiling lighting and small power demands in buildings

A method is proposed for generating daily power demand profiles for lighting and small power in commercial buildings. The method is based on second-order regression model fitting to monitored half-day data sets. Results for both maximum and minimum daily patterns of demand are presented thus forming an envelope of certainty within which a 'typical' daily profile might be expected to lie. A comparison between the predicted envelopes and randomly selected monitored data shows that the model describes actual power utilization with a good degree of accuracy. A method is proposed for generating 'typical' power demand profiles using a random number sequence with its limits defined by the bounds of the power demand envelope and results are presented for summer, winter and mid-season conditions. The results are applicable to power network design, building energy simulations and for predicting casual heat gain profiles for building thermal analyses.