Forecasting UK Electricity and Gas Demand Under RCP Scenarios for Net-Zero Energy Security Using ML

Climate change is altering energy-demand patterns through changing temperatures and heating and cooling requirements. Long-term energy-demand projections are essential for energy security, infrastructure planning, and preparing net-zero energy systems. However, integrated assessments of climate-sensitive electricity and gas demand trajectories in the UK under long-term climate-forcing pathways and net-zero transition assumptions remains limited. To address this gap, this study develops a scenario-based machine-learning framework to jointly project electricity and gas demand in the UK up to 2050 under climate-forcing pathways. CMIP6 daily temperature projections at 0.25° resolution are used to calculate Heating Degree Days and Cooling Degree Days under low-, intermediate-, and high-forcing pathways, labelled RCP2.6, RCP4.5, and RCP8.5. These indicators are used as inputs to Random Forest models for electricity and gas demand. By 2050, electricity demand under RCP8.5 is 4.1% higher than under RCP2.6, while gas demand is 6.8% lower. Under the net-zero adjustment, gas demand declines because of the assumed 75% reduction in gas use, while electricity demand rises as part of displaced gas demand shifts to electricity. Adjusted electricity demand differs by 2–3 TWh between the highest- and lowest-warming pathways, while adjusted gas demand differs by 7–8 TWh. The framework jointly assesses climate-sensitive electricity and gas demand and links these projections to net-zero gas-reduction and electrification assumptions. The results support electricity-capacity and storage planning, electrification strategies, hydrogen infrastructure investment, and decisions on the future role of gas networks in UK energy-security planning under changing climate and transition conditions across Britain.