Effect of a part-hour shading methodology on the sensitivity of shading calculations to horizon uncertainty

•We extend the results of a previous study to include partial-hour shading in beam shade factors.•The 90% confidence sensitivity of irradiance calculations showed essentially no change from prior results.•Allowing for fractional shading is able to reduce the maximum possible sensitivity encountered.•The combined 90% confidence uncertainty for practical shading measurements remains around 3%. A previous study investigated how horizon measurement uncertainty leads to errors in annual irradiance calculations. Those results were based upon determining shading for whole hour periods by testing whether the mid-hour point fell below the horizon. In this study, we test an alternate approach that considers fractional shading during each hourly window. We find only very slight changes in the P90 sensitivity of irradiance to horizon measurement errors. In the previous study, we observed that approaches using whole-hour shading resulted in a maximum sensitivity that increased as smaller position increments were considered for the horizon measurement error. Here, we observe that part-hour shading is less sensitive to the magnitude of these errors, especially with regard to horizon azimuthal measurement. Additionally, we find that the maximum sensitivity is reduced when considering these smaller position increments. The decision to adopt a part-hour shading methodology requires balancing the relatively minor benefits against the additional computational complexity required for the approach.