Performance of a sequential versus holistic building design approach using multi-objective optimization

Integrated design processes are currently pushed forward in order to achieve net-zero energy building designs at affordable cost. Through a case study of a residential building, this paper compares a sequential versus a holistic design approach based on multi-objective optimization. In the holistic approach, 39 design variables related to the architecture and HVAC systems are simultaneously optimized. In the sequential approach, the architecture variables are first optimized; several optimal solutions are then selected for the second phase optimization of the heating system parameters. Carbon footprint, life cycle cost and thermal comfort are optimized by the algorithm NSGA-II. With only 100 computational hours, the holistic approach found 59% of the optimal solutions, whereas it took 765 h to find 41% of the optimal solutions with the sequential approach. This comparison shows the negative effects of making irreversible variable selections in the early phase of a design process, as it reduces the ability to find optimal solutions in the end.