Determination of available transfer capability for a deregulated power system using fuzzy logic

Abstract

The emerging trend of deregulating power industries requires the operator to keep the transmission network congestion free when the generation companies compete to sell electricity to the distributors or wholesale purchasers. This requires real-time predetermination of an index termed Available Transfer Capability (ATC), between every possible pair of source (selling or injection point) and sink (purchase or extraction point) buses. ATC between a given pair of source and sink buses is quantified by the allowable highest magnitude of active power (MW) that can be transferred from the source to the sink over and above the already committed uses (base case) of the whole network without exceeding any line thermal loading and bus voltage limits. The works reported to date on ATC determination have the drawbacks of either oversimplification of the problems or too high computational burden to be implemented in on-line environment. Moreover, the methods that consider both line thermal loading (MA) and bus voltage constraints lack effective exploitation of the correlation between reactive power (MVAR) support and ATC. The present research has developed a real time compatible novel method based on fuzzy logic for concurrent determination of ATC with and without MAR supports. This method provides outputs through trivial computations in one shot from a given base case involving only three input variables in any power system. The architecture of the fuzzy logic model appropriate for ATC determination has been developed in the research. The developed method has been extensively tested and compared against a conventional full scale AC load flow based ATC determination method in terms of accuracy and CPU time using the same array of transactions, base cases, and generationlline outage scenarios in the standard IEEE 24 bus Reliability Test System and IEEE 11 8 bus test system. The method has yielded satisfactory results with much favourable computational time compared to the load flow method.