Large amount of low-grade waste heat is discharged into the environment during industrial processes. This part of waste heat can be collected to serve district heating systems as an important heat source. In most studies of industrial waste heat recovery, the proposed system simulations were unsophisticated in terms of analyzing the real processes. For this reason, the tangency analysis has recently been proposed, and it has been found effective in conducting optimization analysis for direct-heat-exchange systems with multi-heat sources. However, in this study, it has been found that the tangency method has limitations in designing systems with heat pumps, and therefore the disadvantages of tangency analysis are suggested and discussed. Exergy analysis reveals that without considering additional exergy generated by heat pumps, the systems designed by tangency technology tend not be the optimal configuration when heat pumps are employed. In this study, the process optimization principles have been developed from the exergy analysis of heat recovery systems with heat pumps. The optimization principles and mean-heat-transfer-times index are proposed as the key point of process design. Based on the principles, two specific optimization methods in graphic expression are suggested. In the case studies, energy input decreased by more than 70%, which compares favorably with that of tangency analysis.