Solar Variability Assessment and Grid Integration Methodology Development and Case Studies

Sammanfattning: During the 21st century there has been a tremendous increase in grid-connected photovoltaic (PV) capacity globally, due to falling prices and introduction of economic incentives. PV systems are in most cases small-scale, installed on residential dwellings, which means that the power production is widely distributed and close to the end-user of electricity. In this licentiate thesis the distributed PV in the built environment is studied. A methodology for assessing short-term (sub-minute) solar variability was developed, which in the continuation of this PhD project could be used to study the aggregated impact on the local distribution grid from dispersed PV systems. In order to identify potential locations for PV systems in a future scenario, methodology was developed to assess the rooftop topography on both local level using LiDAR data and nationally through building statistics. Impacts on the distribution grid were investigated through a case study on a rural municipality in Sweden. It was found that the hosting capacity, i.e. the amount of PV power generation that can be integrated in the grid without exceeding certain power quality measures, is high, at least 30%. However, the hosting capacity on transmission level needs further investigation. As a first step a methodology was developed in order to model scenarios for hourly solar power generation, aggregated over wide areas, here applied to the whole Swedish power system. The model showed high correlation compared to PV power production reported to the Swedish transmission system operator (TSO). Furthermore, it was used to model scenarios of high PV penetration in Sweden, which give some indications on the impact on the power system, in terms of higher frequency of extreme ramps.