Short Term Load Forecast for Small Scale Power System Using Fuzzy Logic M.F.I. Khamis*, Z. Baharudin, N. H. Hamid, M. F. Abdullah

Load forecast has an important role in planning and operations in power generation plant. Many new forecast systems have been presented and proposed in recent years. Gas District Cooling (GDC) plant is built to supply electricity and chilled water for Universiti Teknologi PETRONAS (UTP). It operates on island operation during normal mode with utility company supply on hot standby mode. The plant has two generation units rated at 4.2 MW each, with a maximum generation capacity of 8.4 MW. In this paper, a practical short term load forecasting (STLF) method for UTP is presented. In the proposed method, a fuzzy logic approach is used and the STLF model is designed based on UTP semester January 2009 electricity demand data. The proposed model is tested with actual load data for January till June 2009 period. The test results show that the mean absolute percentage error of 4.76%.