Thermal comfort in residential buildings by the millions: Early design support from stochastic simulations

In Danish building code and many design briefings, criteria regarding thermal comfort are defined for “critical” rooms in residential buildings. Identifying the critical room is both difficult and time-consuming for large, multistory buildings. To reduce costs and time, such requirement often causes other less critical rooms to be designed with the same constraints as the critical one. In this paper, we propose a method to overcome the difficulty of identifying critical rooms and exploit the design potential of other rooms. First we have defined a set of typical room variations present in most residential buildings. For each room variation, we perform 100.000 simulations while varying important design inputs such as window-floor-ratio, ventilation rates, glazing properties, and shading properties. Prior to this, the Morris method was used to identify and fixate insignificant inputs. A simulation engine based on the hourly version of ISO 13790 is used to calculate the number of hours with unacceptable operative temperature. As a result, the design team can assess a large number of room variations and input configurations by filtering millions of pre-calculated simulations accessible through a web service. An interactive parallel coordinate plot helps the design team to filter the many simulations. Such analysis reveals favorable input spans and assists the design team to quickly assess various design choices