This paper deals with security constraied optimal power flow (SCOPF) in which FACTS devices are employed to meet system constraints under both normal and contingency states. The considered constraints are power generation limit, voltage limit, transmision limit and FACTS devices operation limit. In normal state, the objective function is to minimize operation cost while satisfying system constraints. If contingency occurs, FACTS devices are optimally controlled to eliminate violation of generator ramp rate as well as to meet system constraints. The iterative process is applied to ensure that there will be no generator ramp rate violation. Initially, normal state is simulated to obtain optimal power dispatch as a basecase. Using this basecase, contingency state is simulated in order to minimize generation ramp rate violation. If ramp rate violation is failed to be eliminated, the violation will be fed back to normal state as a basis to re-arrange the output of generators that will be the next basecase. By this basecase, contingency state is again simulated. This iterative process involving normal and contingency states will stop if ramp rate violation is no longer exist. To decompose main problem into normal and contingency state, Bender decomposition technique is used with relation between power generation under normal and contingency states as a coupling equation. The power generation deviation of particular unit should be less than the corresponding generator ramp rate. Each optimization problem is solved by sequential quadratic programming (SQP). IEEE 14 bus will be used to show the ability of the proposed approach to solve the SCOPF.