Quantification of Uncertainty in Predicting Building Energy Consumption: a stochastic approach

Traditional building energy consumption calculation methods are characterised by rough approaches providing approximate figures with high and unknown levels of uncertainty. Lack of reliable energy resources and increasing concerns about climate change call for improved predictive tools.

A new approach for the prediction of building energy consumption is presented. The approach quantifies the uncertainty of building energy consumption by means of stochastic differential equations. The approach is applied to a general heat balance for an arbitrary number of loads and zones in a building to determine the dynamic thermal response under random conditions. Two test cases are presented.

The approach is found to work well, although computation time may be rather high. The results indicate that the impact of a stochastic description compared with a deterministic description may be modest for the dynamic thermal behaviour of buildings. However, for air flow and energy consumption it is found to be much more significant due to less “damping”.

Probabilistic methods establish a new approach to the prediction of building energy consumption, enabling designers to include stochastic parameters like inhabitant behaviour, operation, and maintenance to predict the performance of the systems and the level of certainty for fulfilling design requirements under random conditions.