Rethinking numerical modelling and calibrations on net-zero energy buildings and districts in hot and humid climates

Numerical models are critical for long-term performance predictions with high efficiency and labour cost savings. However, numerical modelling and validation approaches for cross-scale integrative energy systems are diverse. In this study, cross-scale modelling and validation approaches for climate-adaptive zero-energy buildings and districts are comprehensively reviewed. Advanced materials, innovative design and integrative system operation are studied for building energy savings in terms of glazing, building envelopes, and onsite renewables. Considering both aleatory and epistemic parameter uncertainties, uncertainty calibrations on scenario parameters are studied, including dynamic model calibration, Bayesian calibration and approximate Bayesian computation. Furthermore, considering the complexity and challenges of integrative energy systems, such as time simulation mismatch for cross-scale and integrated components, semivirtual experiment-based modelling and validation approaches have also been reviewed. The results indicate that traditional modelling validation approaches for certain periods (such as one day or one week) fail to achieve long-term performance prediction and evaluation with high accuracy. Combining dynamic model calibration, Bayesian calibration, and sensitivity analysis, the validation period (such as one day or one week) can be extended to a full year to enhance the accuracy of long-term performance predictions in building energy performance simulation. Moreover, the uncertainty of measured data and property degradation of thermophysical parameters need to be integrated in modelling development. Building energy modelling calibration is essential for accurate performance prediction, and automated calibration methods, such as optimization with advanced algorithms, machine learning methods, and Bayesian methods, need to be developed. Finally, semivirtual experiment-based modelling and validation are widely used for cross-scale integrative energy systems, including TRNSYS-LabVIEW bidirectional connections, hardware-in-the-loop simulation on MATLAB/Simulink platforms, etc. Research results can provide guidelines for numerical modelling and validation approaches for performance prediction and technoeconomic-environmental feasibility analysis on cross-scale integrative energy systems.