CSP systems based on Parabolic-Trough Collectors (PTCs) account for a representative share of the global solar thermal capacity installed. Most of the existing PTC power plants use synthetic oil as heat transfer fluid. However, water can be alternatively used, where both single and two-phase flow occur. This technology is called Direct Steam Generation (DSG) and it presents some advantages in terms of system overall efficiency and plant cost reduction. Nevertheless, some complexities inherent to two-phase flow have to be addressed to succeed in the commercial deployment of DSG. The development of numerical codes to model such systems is of key importance to put new insights into them. Both, the thermal-hydraulic and optical modeling are of relevance.
Conventional parabolic-trough reflectors concentrate solar radiation on absorber tubes following a characteristic pattern with a non-homogeneous distribution in the angular direction. It means that the bottom half of the absorber tube receives much more flux than the upper half. Under specific two-phase flow patterns (e.g. stratified flow) the refrigeration of the absorber tube is not homogeneous and thermal bending may take place. This problem can also be reported in single-phase flow conditions (e.g. superheated steam), where a significant thermal gradient at the cross-sectional plane of the absorber tube can be critical. A new technique called Inverse Monte Carlo Ray Tracing (IMCRT) method is proposed in this thesis. It aims to design new reflector geometries so that a more homogeneous distribution of concentrated solar radiation on the absorber tube can be accomplished. In line with this, new reflector geometries have been proposed with the same aperture width as the LS-3 design for the same absorber design, where quasi-constant flux distribution is achieved within a specific angular range. Not only the round absorber tube has been studied, but as a matter of example, the potential of the IMCRT method has also been applied to flat absorbers.
A thermal model of a linear receiver has also been performed to describe the heat transfer mechanisms between all the elements on it. Heat diffusion in the absorber tube wall as well as the glass cover has been modeled in 3D, whereas the fluid domain (superheated steam) has been simplified according to a 1D approach. Results have been validated with experimental data from the DISS facility located at Plataforma Solar de Almería. Besides, a detailed thermal-hydraulic study has also been carried out by means of the commercial code RELAP5. This code is based on a transient two-fluid model where both, single-phase and two-phase flow regions can be simulated. Thermal behavior of the absorber tube can also be approximated by means of the standard capabilities of RELAP5. A full and detailed model of the DISS facility has been performed where the connection pipes between adjacent collectors have been included. This model has been validated with experimental data obtained from full day tests getting a good agreement. According to the results, the 1D approach followed by RELAP5 is able to reproduce transient phenomena which usually take place in common operation conditions. In addition, heat losses have been experimentally characterized in the same facility. A new correlation, which takes into account heat losses of receivers after two years and a half of operation, is proposed and included in the model. Thermal-hydraulic codes also make it possible to study severe slugging in connection pipes. For that purpose, a particular connection pipe alongside their two adjacent collectors have been simulated to figure out in which range of operation conditions severe slugging occurs at a pressure of 0.5 MPa.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados