Prefabricated steel construction projects require an integrated supply chain due to the strong interdependence among off-site production, component transportation, temporary storage, installation sequencing, and site constraints. Delays in delivering critical components such as columns, main beams, and connection plates can interrupt installation activities, increase project costs, and generate cascading delays. This study develops a risk-averse mixed-integer linear programming model for optimizing the supply chain and installation planning of prefabricated steel components with fourth-party logistics collaboration. The proposed scenario-based model simultaneously determines supplier selection, 4PL hub activation, direct or 4PL-based transportation routes, fleet allocation, shipment quantities, hub inventory levels, and installation scheduling decisions. Uncertainty in transportation costs, delivery times, fleet capacity, and route risks is incorporated through a set of operational scenarios. To control severe losses under adverse scenarios, the conditional value-at-risk criterion is embedded in the objective function. In addition, a carbon emission constraint is imposed to ensure environmental feasibility. The model is evaluated using a case study based on a steel construction project in Tehran. The results show that selective 4PL collaboration outperforms both direct transportation and full 4PL collaboration by reducing expected cost, installation delay, logistics risk, and CVaR while satisfying the carbon emission limit. The findings indicate that 4PL creates the greatest value when applied selectively, data-driven, and aligned with the installation schedule.