Optimum Luminescent Solar Concentrators

In this paper is presented a numerical optimization analysis of a luminescent solar concentrator (LSC). This optimization was made by implementing a genetic algorithms subroutine in the numerical ray-tracing Monte Carlo model, SIMSOLAR-P. The results from this study provide a theoretical upper limit to the performance of any LSC, and give guidance for the luminescent properties required of quantum nano-crystals to be used in a LSC. The computational time among Monte Carlo calculations has been efficiently reduced in the LSC simulator, by a master-slave parallel implementation featuring the Asynchronous Dynamic Load-Balancing library (ADLB) FORTRAN interface to the MPI parallel programming core. The initial use of the optimization is to determine the optimal parameters of a hypothetical ``perfect dye'' that obeys the Kennard-Stepanov (K-S) thermodynamic relationship between emission and absorption. Results will be presented for three LSC configurations: a single layer with a perfect dye, a tandem system of two optimized layers with perfect dyes (analog of a tandem solar-cell) and a two layer system featuring an organic dye in the first layer and an optimized perfect dye in the bottom layer.