Influence of modeling uncertainties on the simulation of building thermal comfort performance

In  this  paper,  the  parameters  in  a  building  thermal simulation model are tracked, which are subject to modeling uncertainty, i.e. uncertainty arising from commonly applied physical assumptions and simplifications. As an example, the simulation of the thermal comfort performance of a naturally ventilated office building without cooling plant is analyzed. The sensitivity of the simulated performance to the uncertainty in the parameters is investigated with a factorial sampling technique, which not only estimates the mean parameter effects, but also reveals the relative importance   of   non-linear   effects   and   parameter interactions.