Evaluation of the relative differences in building energy simulation results

Building energy modeling, also known as building energy simulation, has developed rapidly in recent years and plays a crucial role in building life-cycle analysis. It can be employed in the design phase to predict the energy consumption of different design schemes and evaluate various control and retrofitting measures at the operation stage. In such simulations, it is commonly understood and accepted that the simulated relative differences are more reliable than the predictions of absolute energy results. However, whether this common understanding is true is yet to be thoroughly investigated. In this study, we investigate the simulated relative differences and the extent to which they are affected by the degree of model input deviation. Simulation and Monte Carlo approaches are adopted for the analysis. The results indicate that the simulated relative differences are not as reliable as expected, and the outputs strongly depend on the degree of the model input deviation. When the degree of deviation is less than 15% or the model inputs are within reasonable ranges, the simulated relative differences match the baseline obtained using Monte Carlo simulations. Moreover, the model's error indicators meet the requirements of the ASHRAE Guideline 14-2014 when the degree of input deviation is below 15%.

Automated summary

Building energy modeling is crucial for building life-cycle analysis. This study investigates the impact of model input deviation on simulated relative differences, indicating the need to control input deviation for reliable results.