Locational Marginal Price Decomposition Based On The Distributed-slack Power-flow Formulation

The practice of electricity market is the inevitable trend of electric industry in developing, which results in the change of the conventional power system in organization, and causes a series of new problems such as determining the nodal price component. To better understand the component of marginal cost, the decomposition of locational marginal price (LMP) is required and can be used in transmission cost calculation, congestion management and related ancillary services.This paper firstly introduces the status of application of LMP and research of the decomposition of LMP and it summarizes methods of transmission loss allocation which is mutually related to the decomposition of LMP. Then it gives the calculation formula of LMP based on optimal power flow and analyzes the dependence of the decomposition of LMP on the slack bus. Finally LMPs is decomposed using the distributed-slack power-flow formulation.When a single slack bus is used in the power flow calculation, the total power mismatch is balanced by generators of this node, and the incremental transmission losses and the power transmission distribution factors are all slack bus dependent. Hence, the decomposition of LMP depends upon the selection of slack bus. In a real system, the total power mismatch prefers to be balanced by generators of different nodes, so the distributed-slack power-flow formulation is more accordant with the actual conditions in power flow calculation. The formulation consists of a vector of distribution factors to determine the portion of the total real power mismatch. This paper points out that distribution factors used in power flow calculation are unsuitable for the decomposition of LMP, and then presents a new vector of distribution factors based on the analysis of the important role played by marginal units and the marginal effect of the marginal units'output variation with respect to load variation, which is applied to the decomposition of LMP. This method considers the system constraints and can reflect the system running case under the optimal conditions.