Disputation David Lingfors: "Solar Variability Assessment in the Built Environment: Model Development and Application to Grid Integration"
- Location: Ångströmlaboratoriet, Lägerhyddsvägen 1 Häggsalen
- Contact person: David Lingfors
During the 21st century there has been a rapid increase in grid-connected photovoltaic (PV) capacity globally, due to falling system component prices and introduction of various economic incentives. To a large extent, PV systems are installed on buildings, which means they are widely distributed and located close to the power consumer, in contrast to conventional power plants. The intermittency of solar irradiance poses challenges to the integration of PV, which may be mitigated if properly assessing the solar resource. In this thesis, methods have been developed for solar variability and resource assessment in the built environment on both national and local level, and have been applied to grid integration studies. On national level, a method based on building statistics was developed that reproduces the hourly PV power generation in Sweden with high accuracy; correlation between simulated and real power generation for 2012 and 2013 were 0.97 and 0.99, respectively. The model was applied in scenarios of high penetration of intermittent renewable energy (IRE) in the Nordic synchronous power system, in combination with similar models for wind, wave and tidal power. A mix of the IRE resources was sought to minimise the variability in net load (i.e., load minus IRE, nuclear and thermal power). The study showed that a fully renewable Nordic power system is possible if hydropower operation is properly planned for. However, the contribution from PV power would only be 2-3% of the total power demand, due to strong diurnal and seasonal variability. On local level, a model-driven solar resource assessment method was developed based on low-resolution LiDAR (Light Detection and Ranging) data. It was shown to improve the representation of buildings, i.e., roof shape, tilt and azimuth, over raster-based methods, i.e., digital surface models (DSM), which use the same LiDAR data. Furthermore, the new method can provide time-resolved data in contrast to traditional solar maps, and can thus be used as a powerful tool when studying the integration of high penetrations of PV in the distribution grid. In conclusion, the developed methods fill important gaps in our ability to plan for a fully renewable power system.