Large-Scale Optimal Power Flow: Effects of Initialization, Decoupling & Discretization

This paper presents the results of extensive numerical testing of a second-order OPF solution method. The testing was conducted using a 1500 bus network under various loading conditions. Three issues were studied: ¿ Sensitivity of the solution with respect to the initial conditions ¿ Accuracy of decoupling the original OPF problem into separate active power and the reactive power OPF problems ¿ Effects of the discretization of transformer taps on solutions. The first issue is particularly important in an EMS environment. An OPF must produce consistent solutions if it is to be used to guide the decision-making of power system operators. This requires that the OPF solution not be sensitive to the (arbitrarily or randomly) selected starting point used by the OPF program, and that changes in the OPF solution point be consistent with the changes in the power system operating constraints. Such changes include the evolution of bus loads over time, control variables which move to their limits over time, and changes in topology due to disturbances. Hence the first objective of the OPF testing was to study the sensitivity of OPF solutions to the choice of initial points. The second issue is of importance because most of the OPF approaches proposed by EMS vendors rely on decoupling of the active and reactive OPF problems in order to achieve acceptable computational performance.