Abstract: In view of the uncertainties of the quality and structure of parts and components of waste products such as rust and deformation, we introduce the destructive disassembly method into the disassembly line. We establish a multi-objective disassembly-line balancing model that considers destructive disassembly mode to optimize the number of workstations, workstation load, disassembly profit, and disassembly energy consumption simultaneously. To obtain high-quality disassembly schemes, we propose a new multi-objective genetic simulated annealing algorithm. In the process of encoding and decoding, we consider disassembly task sequence and disassembly mode sequence. Based on the precedence relationships of disassembly tasks, we redesign the two-point mapping crossover, single-point insertion mutation operation, and double-point exchange. We apply the proposed algorithm to two classical disassembly-line cases and compare it with known algorithm results, thereby verifying the effectiveness and superiority of the proposed algorithm. Finally, we apply the proposed algorithm to a waste television disassembly line. Results show that the proposed model and algorithm can provide a variety of high-quality partial destructive disassembly schemes.