This paper presents Genetic Algorithm (GA) optimization technique as a tool for tuning PID Load Frequency Control LFC. This is to ensure the sustainability of the formed islands after uncontrollable system separation. A two-area single-reheat tandem-compound thermal system is considered. It is focused on an island with excess generation to be equipped with local PID-LFC. The PID controller uses Genetic Algorithm (GA) to obtain the values of PID controller parameters. PID controller parameters are tuned off-line to minimize integral square error over a wide-range of interrupted power. The PID controller is tuned by Ziegler-Nichols method in order to demonstrate the superior efficiency of the proposed (GA) in tuning the PID controller. Therefore, the proposed technique could tune the PID controller for optimal response at any value of the interrupted power. Testing of the developed technique shows that, the proposed PID-LFC could present optimal performance and it capable of improving the transient frequency response of thermal power plants. So it contributes in improving the overall area response. Results are provided in the form of time domain simulations via MATLAB /SIMULINK.