Comparison to reference studies

The results of the set-up concerning buildings in their existing state is illustrated in Figure 8. Figure 8a shows the results obtained by simulating the set-ups with the energy carriers gas and oil. The reference studies combine gas and oil because they only considered boiler systems and the results differ marginally. The simulations constantly underestimated the average values of energy demand presented in the reference studies. The reference studies likewise display great differences between each other. The reason for the underestimation might be the simplification in the modelling of occupant behaviour. Wasteful behaviour was neglected, and thus also contributed to the narrower spread. On the other hand, the simulations yielded extremer results because the surface reconstruction exaggerated the living area. For example, after visual inspection of few buildings it became evident that unconditioned balconies were often unnoticed. Unconditioned balconies became living area, thereby expanded the conditioned area. In this regard, the number of occupants increased. In addition, the results for SFH/TH suffer from misclassification of actual bigger buildings. The distinction of the size classes occasionally infringed the definition of building types by classifying MFH as SFH. The results for MFH/AB differ less from Fisch et al. 2012, but more from Schröder et al. 2014. The lower medians and means of EUI comparing to the smaller class are associated with the additional thermal insulation provided by denser habitation. The conditioned volume of MFH is more limited than in SFH and their heating system might also benefit from some scale effect. Although the minimum EUI in SFH is lower, the narrower spread and lower average for MFH might be linked to less wasteful occupant behaviour and higher efficiencies.
The simulated results for the district heating in Figure 8b are proportional to those of gas and oil heating because there are only slight differences in their models. In general, the results follow the downwards tendency in the reference studies because district heating is more efficient. Less heat losses can partially explain this, but also the fact that more MFH are connected to district heating than SFH. The modelling of district heating in EPlus is also simple, and thus might contribute to the better results. However, the number of SFH with district heating in Fisch et al. 2012 is just 12.

Figure 8: Summary of annual energy use intensities for buildings in their existing states in comparison with Loga et al. 2017, which harmonizes the results of Fisch et al. 2012 and Schröder et al. 2014. Arithmetic means and standard deviations are symbolised by dashed lines, while medians and quartiles are symbolised by solid lines.

(a) Gas and Oil.

(b) District heating.

Figure 9 only presents Fisch et al. 2012 because Schröder et al. 2014 does not include refurbished buildings. In the study of Fisch et al. 2012, there is no distinction between SFH and TH. The energy performance of refurbished buildings, both simulated and measured, is better than of buildings in their existing state. The averages of simulated EUI tends to become lower with building size but still remain below the reference values. The standard deviations also become narrower. Although similar behaviour is displayed by the reference values, the spread of simulated values is still narrower. The simulations with refurbished buildings represent better the reference values than the simulations with buildings in their existing state. The greatest difference in averages and standard deviations between simulated and reference values can be observed in MFH because MFH might also include some luxurious SFH.

Figure 9: Summary of annual energy use intensities for buildings with usual refurbishment and gas/oil heating in comparison with Fisch et al. 2012. Arithmetic means and standard deviations are symbolised by dashed lines, while medians and quartiles are symbolised by solid lines.

The higher efficiency of district heating is supported by Figure 10 showing refurbished buildings with district heating. According to the results from the simulations and reference studies, district heating in refurbished buildings is unsurprisingly the least energy intensive way to provide viable conditions. The exception in Fisch et al. 2012 is SFH with higher average EUI and wider spread of values than the same class with gas and oil heating. The cause is a much lower number of buildings in this class with only 25. The performance in simulating MFH and AB improved comparing to gas and oil heating. However, the averages of simulated EUI is still below the reference values, while the spread of values is narrower.

Figure 10: Summary of annual energy use intensities for buildings with usual refurbishment and district heating in comparison with Fisch et al. 2012. Arithmetic means and standard deviations are symbolised by dashed lines, while medians and quartiles are symbolised by solid lines.