Can residential energy systems withstand the heat? Quantifying solar photovoltaic and heat pump yields for future New Zealand climate conditions

Abstract We investigate how higher temperatures resulting from climate change impact the energy system. Specifically, we examine the cumulative effects of fluctuating solar photovoltaic (PV) generation performance, heating and cooling demand, and heat pump efficiency on such days. To achieve this, we used the climate analogue space, which maps a given city’s future climate to an existing one. By employing climate analogues, we can predict the impact of higher temperatures by 2050, transforming Auckland, New Zealand’s climate into one akin to Sydney, Australia. This approach avoids reliance on historical weather data, which many energy system models use. We used this future climate time series as an input to a residential energy system model for Auckland, New Zealand. The residential energy system model simulates solar PV generation output via mapping of experimental data, building thermal characteristics via grey-box resistance–capacitance ( RC ) modelling, and hourly coefficient of performance for air source heat pumps (ASHP) via linear regression. Our findings revealed that a future climate doubles the cooling demand and reduces the heating demand by one-third, with the heat pump demand peak load projected to be 40% higher than current demand. Although solar PV generation experiences a decrease in efficiency of 8%, there is a 40% increase in annual direct usage of ASHP. Despite the high cooling demand, the combined yearly electricity demand for heating and cooling decreased by 6.5% overall, and the system saw a 50% improvement in demand fulfilment. However, the system performance volatility at hotter-than-normal temperatures and the potential for significant energy shortfalls remain concerns. The shift from a predominantly heating to a cooling environment is a critical design condition that should be considered in energy expansion planning and future electrification. The framework and time series developed in this work can be expanded and applied to other energy system modelling exercises.