Erosion of solar reflectors in desert environments

In order to reduce the investment risk of CSP plants, the performance over time of all components involved in the power generation process has to be known as precisely as possible. Since the expected plant lifetime is up to 25 years, accelerated laboratory aging experiments are used to estimate the performance loss of particular components for distinct sites. Those experiments have to be specially tailored to meet the natural conditions to which the components shall be exposed. So far, no CSP specific sandstorm erosion test is designed to predict the performance loss of opticalcomponents. This thesis starts with the identification of erosion determining physical parameters and a literature review about worldwide sand- and dust storm (SDS) activities with the focus on the identified parameters. Furthermore, a review about the state of the art erosion testing is given, which identifies insufficient knowledge about the field parameters to reproduce realistic testing conditions. Meteorological and geological parameters were analyzed for the two Moroccan sites Zagora and Missour. It was shown, that the wind velocity, the relative humidity and the particle size distribution of the soil in Zagora favors SDS occurrence. On the one hand, this finding was verified by augmented aeolian dust concentrations measured by an optical particle counter (EDM 164 from Grimm), and on the other hand by the presence of saltation events detected by a light barrier sensor (YH03PCT8 from Wenglor). Additionally, the microscope analysis of state of the art solar reflectors after an 26 months outdoor exposure showed clear evidence of intensive erosion taking place in Zagora while the glass surface did not show mentionable erosion defects in Missour. Three laboratory setups were investigated experimentally in this thesis in order to determine the most adequate methodology to perform accelerated erosion testing. The accelerated erosion tube (Acetube), which was designed and constructed for this thesis, allowed for a comprehensive parameter study with the following conclusions: wind velocity exhibits a power law relationship on the reflectance loss ∆ρ’. The threshold velocity to cause erosion damage with representative particles is around 8 m s−1 atan impact angle of 45°. The influence of the impact angle can be well described by asin2-behavior. The minimum particle size to cause erosion damage on glass showed a threshold value in the range of 50 µm in diameter. Due to its high hardness, quartz is the main responsible mineral to cause erosion damage. Experiments conducted with gypsum and calcite particles did not show effects on the reflector surface. Accelerated erosion tests with natural erodent sand confirmed that soils with higher quartz content produce a more severe erosion pattern. Two modeling approaches of the ongoing erosion in the field were discussed. The first approach was based on the description of aeolian particle movement by a novel parameter, the single particle impact energy distribution (SPIED). Because of instrumental limitations and uncertainties, this approach did not lead to realistic results.The second approach takes the defect size distribution (DSD) of the outdoor exposed reflectors as the basis for modeling. Input parameters for the Acetube experiments were found to accurately reproduce the DSD of the outdoor exposed reflectors. The tailored parameters to simulate erosion in Zagora can be used to predict the expected erosion behavior of different materials on site. The work concludes with the definition of a testing guideline for CSP components to be applied in desert environments. Recommendations to assess the site specific erosion risk are given. The suggested field setup with hardware costs of around 5800 €, allows for a thorough analysis of the erosion processes and to derive tailored parameters for accelerated erosion testing