sources respectively; qc; and q, are the added capacitive and inductive VARs; and di is the fixed VAR installation cost for bus i. a, is a 0 1 decision variable. a, = 1 if the ith bus is selected for VAR installation otherwise a, = 0. C is the set of candidate buses consisting of buses to be considered for VAR installation. The constraint set (1) involves the load flow equations for the given operating condition including the new VAR sources as variables. Bounds on the system voltage magnitudes, the transformer taps, and generator production, are also included in the constraint set. The constraints set (2) include the inequality constraints that enforce a prescribed margin to collapse for the given operating condition. The security margin requirement is stated as a relative margin, defined in terms of actual and critical loads at the respective buses [2]. The resulting optimization problem is a mixed-integer nonlinear programming (MINP) problem for which no efficient solution method is available. A two-stage algorithm to obtain a near-optimal solution is therefore developed. The first stage subproblem is established by fixing the values of the decision variables a;, i E C, at 1. The resulting nonlinear programming problem installs VARs such that the variable component of the cost function is minimized, and the voltage level and SM requirements are satisfied. The second stage subproblem is established by linearizing the MINP problem around the solution point, obtained in the first stage, with some modifications. The resulting mixedinteger linear program optimizes the fixed components of the cost function by minimizing the number of buses where VARs were installed according to the solution of the first stage such that the goal regarding voltage level and SM is still realized. A procedure is subsequently described for planning VAR support such that the system voltage levels and SM requirements remain satisfied in the event of any one of a collection of contingencies occurring. The procedure is summarized in the following steps: