Traditional methods for estimating the solar energy potential in buildings determine energy yields on an annual base and make use of highly aggregated geo-spatial data. This work proposes a method for detailed assessment of the potential solar energy yield in the temporal and spatial domain. Solar irradiance is evaluated using numerical methods based on hourly variation of solar irradiance and on actual building geometry. Results of our initial studies allow exploration of the variation patterns in solar yield depending on local and time-varying factors, which cannot be seen in coarse level solar planning tools. This helps identifying surfaces with good solar yield that are deemed unfavorable according to traditional planning practices.