Investigating an adequate level of modelling for retrofit decision-making: A case study of a British semi-detached house
This paper investigates what level of modelling (zoning or internal load scheduling) is required to support heating related retrofit decision-making. First, this paper tests the effect of thermal zoning by incrementally reducing the number of thermal zones from modelling every room as a separate zone to modelling the house as a single zone. Second, this paper examines the influence of internal load schedules (occupancy, lighting and equipment schedules) on prediction accuracy. Actual internal load schedules were derived from the smart meter data of 666 households collected by the Customer-Led Network Revolution project. Cluster analysis was applied to extract a set of prototypical schedules to capture major variations across all households. Last, this paper evaluates the effects of the zoning and internal load scheduling modelling assumptions in the context of thermal retrofit decision-making. For the specific parameters studied and the specific building design, the use of different zoning strategies and different internal load schedules yielded the same ranking of top retrofit options. For the specific climate and the baseline assumptions for the retrofits, different cluster schedules resulted in different magnitudes of energy savings, but the ranking of top retrofit options was not impacted by the choice of household internal load schedules. However, the actual internal load schedules affected the energy-saving potentials achievable by the same set of retrofit options. The case study highlights that the optimal set of retrofit options selected given the specific physical characteristics of a house is the same regardless of differences in the input of internal load schedules. However, it was found that energy-saving potentials achievable by the same retrofit option substantially vary according to the actual internal load schedules. This finding implies that energy retrofit policies can be tailored to target certain groups of households selected by clustering their actual energy use profiles to cost-effectively maximise energy savings from the domestic sector.