A methodology for simulation and assessment of Concentrated Solar Power plants
Allan Ricardo Starke1; José Miguel Cardemil2; Leonardo Lacerda Lemos1; Rodrigo Escobar3; Rafaela Frota Reinaldo4; Sergio Colle1
1 Federal University of Santa Catarina/LEPTEN; 2 bSchool of Industrial Engineering, Universidad Diego Portales - Chile; 3 Mechanical and Metallurgical Engineering Department, Pontificia Universidad Católica de Chile, Santi; 4 PETROBRAS/CENPES/PDEDS/GEGQ, Av. Horácio Macedo, 950 - Cidade Universitária, Rio de Janeiro, Brazil
doi:10.20906/CPS/COB-2015-1739
Resumo
A thermal analysis of Concentrated Solar Power plants is conducted considering a the following technologies: parabolic trough collectors (PTC), linear Fresnel collectors with direct steam generation (LFC-DSG), central receiver system with molten nitrate salts (CRS-MNS) and central receiver system with direct steam generation (CRS-DSG). The plant capacities were ranged from 50 to 800 MWth. The analysis focuses on the environmental conditions of selected Brazilian locations, such as Bom Jesus da Lapa and Petrolina. Thus, the study considers a parametric analysis of the main design parameter for different plant scales, in terms of the thermal performance indicators as solar field aperture area, power block rating capacity and plant annual efficiencies. The annual production of the plants is calculated by using the Transient System Simulation program (TRNSYS), which consider a new component library developed for that purpose. This library is based in the open access models developed by the U.S National Renewable Energy Laboratory and currently employed by the System Advisor Model (SAM) program. In addition, a new fluid properties subroutine compatible with TRNSYS codes standards was developed, which uses the freeware CoolProp library. These approaches allowed us to modify or create new configurations for CSP plants, e.g. thermal storage for the direct steam generation technologies and detailed Rankine cycle models, which are not available at the conventional configuration of SAM. The results obtained show good agreement with other software packages as well as with actual data from currently operating CSP plants. The annual comparison between SAM and TRNSYS results shows deviation lower than 1.3%, while the difference between TRNSYS results and the reference data are roughly 4%. Given the lack of information available, these results are considered a good approximation. Furthermore, hourly comparisons between SAM and TRNSYS results showed good agreement, with relative root mean squared errors ranging from 5 to 0.3%. Co
Palavras-chave: Concentrated Solar Power; Brazil; Simulation; TRNSYS