IntroductionOperational decarbonization in prefabricated construction requires decision support that links factory production, transport logistics, energy prices, grid carbon intensity, and carbon-trading rules. This study develops an environmental systems engineering framework for carbon-management decision support in prefabricated component production and delivery.MethodsThe framework represents the supply chain as a coupled production–transport system in which steam-curing intensity, time-of-use electricity pricing, time-varying grid carbon factors, diesel transportation emissions, and stepped carbon trading jointly shape operational choices. A tri-objective model is formulated to minimize project completion time, energy and fuel costs, and net carbon-trading costs. The model is solved using the Bi-layer Cooperative Evolutionary Algorithm with Q-Learning (BCEA-QL), which jointly searches production and delivery decisions.ResultsComputational tests on nine synthetic test instances, including a recent reinforcement-learning-assisted baseline, show that BCEA-QL achieves the highest HV on all nine instances, with up to 6.88% higher hypervolume on large-scale instances. A 25-group metropolitan metro precast case further shows that, relative to a time-oriented schedule, a carbon-oriented schedule reduces energy and fuel costs by approximately 23% and yields only a small carbon-trading credit. A post-processing delay-cost analysis identifies manager-dependent switching thresholds near 57 and 212 CNY/h.DiscussionThe results indicate that coordinated production and delivery scheduling can help environmental managers interpret operational carbon-management trade-offs under asynchronous price–carbon signals, without implying universal carbon reduction or full field validation.