Toward net-zero building: an optimisation methodology for passive and active retrofits with renewable integration – a UK office building case study

Existing commercial buildings account for a large share of energy use and carbon emissions, highlighting the need for net-zero retrofit strategies. This study develops an integrated framework for building retrofits, combining calibrated building energy simulation with multi-source renewable energy system design and optimisation to reduce energy demand, maximise on-site renewable generation, and enable cost-effective decarbonisation. A calibrated energy model was created in DesignBuilder to evaluate retrofit measures, while renewable energy systems and storage were optimised for maximum renewable fractions using HOMER Pro. A two-stage MATLAB-based Pareto front and TOPSIS multi-criteria optimisation evaluated and ranked retrofit configurations. The framework was applied to a UK office building case study (hot-summer, cold-winter climate), identifying the optimal configuration, which combines lighting upgrades, office equipment improvements, a water-to-water ground-source heat pump, carport-mounted solar PV, and rooftop horizontal-axis wind turbines. The retrofit reduces annual energy demand by 38.4% and achieves an average renewable fraction of 66.3%, reducing net grid energy import by 91.53%, with on-site generation exceeding demand during peak summer months. Annual CO 2 emissions decrease by 39.0% after retrofitting and by 86.1% after the integration of renewables compared to the baseline model of the case-study building. Economic analysis shows strong financial viability, with a levelized cost of £0.172/kWh and a payback period of 6.48 years. Sensitivity analyses confirm the solution remains optimal under varying energy prices and stakeholder priorities, demonstrating robustness for commercial deployment. The results highlight the framework's ability to guide transferable, multi-criteria retrofit planning and support deep operational decarbonisation across diverse building types.