CFD Modeling of the Solar Two Central Tower System Receiver
thesisposted on 06.08.2018, 00:00 by Simone Pastura
Nowadays, CSP technologies cannot compete on price, for example, with photovoltaics (solar panels), which have been subjected to a huge growth in these last years because of the falling prices of the panels and a much smaller operating cost, as well as exceptionally subsidized, especially in our Country. Concentrated Solar Power system requires cost reduction and increased efficiency to make a breakthrough in the energy market. Paying attention to central tower systems, these ones allow higher concentrations than parabolic troughs and therefore show high potential in that perspective for future deployment. CFD modelling is a strong work tool to study geometrically complex engineering problems and, in this case, it can help in the optimization of the design of the receiver. The development and validation of a computational fluid dynamic model of the ‘Solar Two’ central tower system receiver at Sandia National Laboratory (SNL), in Albuquerque (NM) USA, by means of the general purpose CFD software STAR-CCM+, is considered. The verification, benchmark and validation of a purely hydraulic, external flow (air) sub-problem have been approached first (both in 2- and in 3-dimensional analysis), as well as the verification of 3-dimensional, thermal-hydraulic, internal flow (molten salts within the receiver) sub-problem, separately in two different scenarios experimentally studied at SNL. Once the separated analyses have been completed, the coupled external/internal flow simulations of both cited scenarios have been performed: the CFD results are in agreement with those available from employed correlations for smooth cylinders (by Churchill and Achenbach) and from on-site measurements, within the corresponding error bars. The computed convective heat loss (equal to the surface integral of the heat flux from the receiver to the air, subtracted of the radiative part) at statistically stationary state are ≈ 1.5 % (2.2 %) of the incident power on the receiver surface, for scenario #1 (#2).