When large amount of gas turbine hot exhaust gas is required to be cooled for reuse in a closed cycle system, the conventional cooling methods will no longer be successful. The great temperature difference, the little space available for cooling, the large speed by which the exhaust gas is expelled, and finally the low exhaust gas pressure represent a great challenge for such a practice. Evaporative cooling is increasingly used as an efficient method to enhance thermal comfort in exhaust gas recirculation. In a water spray system, a cloud of very fine water droplets is produced using pressure nozzles. Analysis of spray penetration and evaporation requires a careful integration of the process conditions such as exhaust gas velocity, nozzle spray pattern, and water flow rate. The droplet size distribution is a critical input to calculations of droplet trajectory, evaporation time, and quenching rate. Computational Fluid Dynamics (CFD) is considered a valuable tool when assessing the potential and performance of evaporative cooling. In this paper, a systematic evaluation of the Lagrangian-Eulerian (LE) approach for evaporative cooling provided by the use of a water spray system with a hollow-cone configuration of two nozzles differs in rate and particle size distribution is presented. The evaluation is based on grid-sensitivity analysis and verified using experimental measurements. The effect of the hot exhaust gas temperature, the cooling water flow rate and nozzle Sauter Mean Diameter (SMD) is also presented. The results show that CFD simulation of evaporation by the LE approach, in spite of its limitations, can accurately predict the evaporation process. Finally, a good matching with the experimental measurements of the cooled exhaust gas temperature is achieved.