Paper Conference

Proceedings of BSA Conference 2015: Second Conference of IBPSA-Italy


Energy retrofit and conservation of built heritage using multi-objective optimization: demonstration on a medieval building

Francesca Roberti, Ulrich Filippi Oberegger, Elena Lucchi, Andrea Gasparella

Abstract: Energy retrofit of historic buildings is a complex activity, which requires a multidisciplinary approach. Interventions should limit energy consumption, consider users’ comfort and preserve cultural and aesthetic values. While the impacts of interventions on energy performance and comfort can be quantified in advance using simulation software, conservation aspects are less tangible. In this paper, we propose a method to identify retrofit strategies that are optimal from an energy, comfort and conservation point of view. The first step is to choose a set of interventions and conservation aspects to consider. This requires a multidisciplinary team of experts. Next, quantitative metrics for assessing energy performance, comfort and conservation are defined. Through a multi-objective optimization, the combinations of interventions that yield the best tradeoffs among these objectives are found. We demonstrate the method on a calibrated EnergyPlus model of the “Waaghaus” (weigh house), a medieval building in Bolzano located in the north of Italy. The aim is to transform this currently vacant building into a cultural center. We considered the following interventions: external and internal envelope insulation with varying materials and thicknesses, airtightness improvements, replacement of windows and summer ventilation availability. Conservation aspects taken into account were visual, physical and spatial impact of the interventions on the building’s heritage significance. We selected the hourly sum of all sensible and latent ideal loads for heating and cooling over a year as energy performance metric. All internal loads were modelled according to the planned future use of the building. We assigned a score to each intervention equal to the number of conservation aspects met. The yearly average of the absolute values of the predicted mean vote was used as a proxy for comfort. We performed the multi-objective optimization with the C code NSGA-II, which implements a genetic algorithm based on nondominated sorting. As a result, we obtained solutions with an absolute mean PMV of 0.5, an annual ideal load for heating and cooling of 20 kWh/m2 and a good level of conservation.
Pages: 189 - 197