Optimal recirculation in radiant and convective furnaces is chosen based on numerical simulations and optimal system organization principles. In optimization, the focus is on multifunctional solutions that not only increase the energy and resource conservation and decrease the environmental impact but also improve the thermodynamic efficiency. Computational experiments for radiant furnaces-specifically, coke ovens-have been used to determine the optimal gas recirculation ratio (more than twice that used in practice), in combination with cogeneration and the reduction of nitrogen oxides emissions in horizontal coke ovens with indirect heat transfer from the bottom. A multifunctional engineering solution is developed for convective furnaces with recirculation for heating a heat-sensitive material. Optimal recirculation ratio is chosen in combination with waste heat recovery in the Rankine thermodynamic cycle, ensuring cogeneration while maintaining flue gas nitrogen oxide concentrations at daily-averaged threshold limit values. The exergy efficiency of a system with the furnace is practically quadrupled.