To optimize the heat transfer performance of the thermal-system heater, based on the particle swarm optimization algorithm and entransy dissipation theory, in this paper, a minimum entransy dissipation approach for optimization design was developed, taking entransy dissipation number as the objective function. The optimum size of the heat exchanger was determined without the pre-set structure of the heat exchanger and the phase change of working fluid of the heat exchanger was considered in the optimization design progress, which is different from the traditional design calculation. The convection heat transfer coefficient of steam condensation on the shell side was corrected by the latent heat value. For a specific thermal-system heater, the optimized design results showed that the exchanger effectiveness was increased by about 7.8%, while the pumping power reduced by about 19.6%, indicating that it achieved minimum power consumption when the thermal performance of thermal-system heater was optimized.