Dimensioning of Heliostat Components under Wind and Gravity Load: the Map Approach

Abstract

Estimating annual optical performance of heliostats under realistic load is computationally very expensive as complex structural deformation and ray tracing calculations are necessary. An approach is presented that vastly accelerates these calculations by using pre-calculated multidimensional maps derived from a limited number of precisely computed grid points through linear interpolation. Structural maps represent mirror misalignment as a function of wind velocity, direction and gravity. Power maps contain the intercepted power of a heliostat given its position in the field, mirror misalignment and the current sun position. The maps are generated based on specific assumptions about the tower and receiver dimensions as well as heliostat dimensions and on simple experimental investigations as for the drive properties. An estimation of the annual energy yield is obtained by integrating current intercepted power values chronologically for a given heliostat field configuration. For this purpose, representative time series for weather data (direct normal insolation, wind velocity and direction) are generated artificially. The misalignment of mirrors is calculated from wind loads taking into account wind shading effects. By applying this method to different drive configurations, the impact on the annual yield under wind loads can be rapidly identified. Thus, the method allows easy assessment of drive and other component modifications and therefore can be helpful in reducing specific costs. Results of this method are presented, using a specific heliostat design as reference. A drive train with typical stiffness properties is compared to an ideally stiff one. It is found that the method works in that it provides plausible results within reasonable computing time. Future extensions of the tool to include backlash and other drive effects are briefly described.