Sustainability quantification using multi-attribute decision-making techniques for construction portfolio optimization model: the effect of linearization in different solution methods

Integrating sustainability into construction portfolio decisions remains challenging due to the difficulty of quantifying heterogeneous criteria. This study develops a framework that integrates sustainability assessment with bi-objective portfolio optimization. Sustainability criteria are identified from the literature and weighted using the Bayesian Best–Worst Method (BBWM), while project-level sustainability scores are calculated using Weighted Aggregated Sum Product Assessment (WASPAS). The optimization model maximizes discounted after-tax profit and aggregate sustainability while accounting for cash-flow reinvestment, workforce adjustment costs, material capacity, and bracket-dependent taxation. An integer linear programming (ILP) reformulation is developed and compared with the original integer nonlinear programming (INLP) formulation using solver-based and metaheuristic approaches on small- and large-scale instances. In solver-based comparisons, the ILP formulation achieves identical objective values with lower computation times. For metaheuristic methods, it generally produces solutions closer to the solver-based benchmark, although runtime improvements vary across algorithms. Pareto-front analysis illustrates the trade-off between financial performance and sustainability, providing a quantitative basis for construction portfolio decisions.