This paper presents a comparative investigation of different FA based optimization technique to tune the PID parameters for LFC. The investigation is carried out on a two area interconnected reheat thermal system with generation rate constraint (GRC). The efficiency and robustness of three different FA (standard FA, improved inertia weight FA (IWFA) and chaotic FA (CFA)) as controller optimization technique is investigated. Three different loading scenarios were analyzed in which the frequency deviation step response and tie line power deviation of all three different optimization techniques were observed. The average and minimum value of the system's error criterion based on integral time weighted squared error (ITSE) was also analyzed. Simulation results show that the CFA based PID outperforms the standard FA and IWFA. The CFA is able to successfully find the optimal gains and tune the LFC effectively under different loading scenarios.