Modelling solar access at the window level of buildings – a case study for urban densification simulations in Sweden

Abstract: Buildings are a are a major consumer of energy in urban areas. The use of solar energy in buildings can help reduce CO2 emitting energy consumption in these urban areas. One way to use solar energy is through passive solar energy systems. Passive solar energy systems allow a building to use daylight as an energy source. The use of daylight as passive solar energy in Sweden could reduce the energy consumption in buildings by 23% to 42% which would help reduce CO2 emissions. Architects can design these passive solar energy systems for individual buildings in building modelling software suites, but to model them and study how they interact within a built environment, a 3D city modelling approach is needed. In Sweden, a project called 3CIM (3 city information model) is currently working on a 3D city modelling national standard in CityGML. The aim of this case study is to propose methods to study the effects of urban densification on buildings access to daylight in the Bellevue area in Malmö, Sweden. These methods enable estimation on how much daylight will reach the windows of a building’s façade. To do this 3D city models, data collected in the field, meteorological data, and building data for future planned buildings are used to simulate buildings daylight access. Since window information for buildings is not readily available in the current 3D city model, data collection methods for windows are also designed and implemented in support of this aim. A handheld camera was used for data collection and photoshop and OpenCV was used for photo corrections. From there Blender, a 3D modelling software, was used for 3D modelling of the windows. Next, the urban multi-scale environmental predictor (UMEP) was the tool used to model the estimated solar irradiance or “daylight” on the building roofs and façades. Finally, the daylight estimation tool (DET) was used to model the estimated solar irradiance at window level. The estimation methods developed are evaluated in two cycles. The first cycle includes the buildings that are present in physical space at the time of writing the thesis. The second cycle also includes proposed buildings in the Bellevue area. The output of this will be two reports of the results that includes the windows for each building façade, their geometry, and their solar irradiance values. The results showed a measurable change in the amount of solar energy that reaches the windows of the building facades. These results show that the data collection methods were successful in a scenario where window information isn’t readily available and needs to be collected. Next, the methods for 3D modelling the windows were also successful, though the process is time intensive. Future automation could streamline this process. Additionally, the methods for estimating the solar irradiance at façade level and window level were successful, with some limitations being data type and data availability. Finally, the overall results when comparing the results of the two cycles show that some buildings were affected more than others. The highest change to a building was ~20% and the lowest change was ~0% with the total buildings show an estimated ~4% change.