Thermal Modeling of Solar Central Receiver Cavities

Abstract

Results are presented from a numerical model of the steady-state energy transfer in molten-salt-in-tube solar cavity receivers that includes convective energy transfer at a local (spatially resolved) level. Molten salt energy absorption and gray radiative transfer between all cavity surfaces are also included. This model is applied to the Molten Salt Subsystem Component Test Experiment (MSS/CTE) cavity receiver. Results for this receiver indicate the global (entire cavity) receiver thermal efficiency is invariant within a few percent to most parameters investigated, although front surface temperatures of the nonabsorbing walls vary considerably, and are particularly sensitive to the type of convective submodel used. Absorption efficiencies indicate the effects of the cavity enclosure environment. For all conditions investigated, tube inner wall temperatures remain under 855 K, ensuring that the salt remains chemically stable. Global results for the receiver indicate thermal conditions in the receiver are temporally constant within an hour of solar noon, and solar panel temperatures are governed by the temperature of the flowing salt (the outlet temperature is maintained at 839 K by varying the salt mass flow rate). The dominant loss mechanism is radiative transfer, although convective loss predictions are of the same order of magnitude. The absorbing panel front surface temperatures and the panel thermal losses are somewhat invariant with incident flux. Losses from the nonabsorbing surfaces (comprising over 60 percent of the cavity surface area), however, do vary with incident flux levels. These results suggest that a correction for nonconstant losses in the Barron flux-on loss method is necessary. Global predictions compare well with data obtained in the MSS/CTE experiment. Predicted salt flow rates and absorbed powers were within one percent of values measured during steady-state tests. Predicted global loss values compare well with current loss estimates from measured data.