Load modeling, estimation and forecasting

Load modeling, estimation and forecasting are vitally important for power utilities both in regulated and deregulated economy. The price of power that power utilities have to procure for meeting their customers´ demand on long term basis versus the day ahead or real time basis under unscheduled interchange mechanism has fluctuated wildly. Sometimes the variation has been even ten times the long term average. Therefore, for the financial performance of the utilities it is important to make an accurate forecast of the demand for the long term, medium term as well as the short term and real time needs in order to efficiently deal with the shortages or surpluses. Some of the factors that influence the demand are (a) Weather parameters (b) Climate parameters (c) Events like holidays (d) Derived parameters like heat index, cool index, etc. and (e) Economic development activities like the master plan of the city, the population growth etc. With so many predictable and unpredictable parameters, human intelligence is used along with the combination of various statistical techniques for load forecasting. This paper aims to discuss the experiences of BSES, Delhi in load forecasting.