Probabilistic Allocation of Electric Service Costs
Many methods have been proposed to allocate capacity related costs to the various customer classes in an electric power system. These methods are mainly deterministic. Deterministic techniques do not consider the stochastic nature of system components which actuary affect the cost behaviour of the system. An important consideration in the process of colt allocation is to ensure that the method used follows the cost behaviour of the system studied. The cost behaviour is basically dependent on the methods and criteria used in system planning and operation. The criterion and planning method used drives the development of the system which further drives the cost of the future investment in the system. It is therefore highly desirable, the case of cost of service allocation, to use a similar method and criterion and to include same variables which are considered in the system adequacy evaluation. All Canadian utilities now use probabilistic techniques in their generation system planning. The most commonly used planning techniques are the Loss of Load Expectation (LOLE) and th e Loss of Energy Expectation (LOEE) methods. It therefore follows that probabilistic techniques should also be used in the cost of service allocation process. There is increasingly greater emphasis on the use of energy based indices in system planning. A probabilistic technique using an energy index, Loss of Energy Expectation, is proposed in this thesis. The technique considers the important factors which actually affect the system cost behaviour. There is also increasingly utility interest in the use of composite system reliability evaluation techniques in system planning. It is expected that utilities will be using composite system reliability evaluation in the future as their basic system planning technique. It therefore follows that a comprehensive cost of service allocation technique which utilizes a similar approach should also be used. The Loss of Energy Expectation method proposed in this thesis is modified and extended for use in composite generation and transmission cost of service allocation. The response of the proposed techniques to different system parameters and configurations is presented in this thesis. These studies clearly demonstrate the superiority of the proposed probabilistic methods over the deterministic methods. The proposed techniques have been applied to two test systems using new load databases developed in this research project. The results demonstrate the sensitivities of these methods to various system parameters which are not now considered by the popular deterministic methods. The results clearly show that the composite system method is the best overall technique as it responds to many variables that are not considered by other methods. A strategy for the selection of an appropriate method for cost of service allocation is proposed in this thesis. This strategy is designed to ensure that the cost of service allocation method is compatible with the method used in system planning and operation which ensures that the cost behaviour of the system is followed.
Doctor of Philosophy (Ph.D.)
Electrical and Computer Engineering