Mitigating particle deposition on the glass window of a fluidised bed solar receiver

Abstract

Concentrated solar energy can be utilised for many thermal processes that require high temperatures. A solar receiver is one such device that receives concentrated solar radiation. Different types of solar receivers such as fluidised beds, packed beds and solid particle solar receivers are used in different thermal applications. The focus of this study is to investigate the flow behaviour in a Fluidised Bed Solar Receiver (FBSR). One of the major limitations in the use of an FBSR under direct irradiation is particle deposition on the receiver glass window. This has two consequences: it reduces solar radiation transmission into the receiver and results in the failure of the glass window. There have been a number of investigations on mitigating particle depositions onto the glass windows of solar receivers. However, the aerodynamics of solar receivers using particles is still not well understood. The aim of this project is to mitigate particle deposition on the glass window of an FBSR by developing a better understanding of the flow patterns under different operating conditions, which can assist in the development of an aerodynamic seal. Continuous operation of the FBSR can result in particle deposition on the glass window, which is directly related to the flow behaviour of the receiver. Therefore, it is essential that the flow pattern in an FBSR is investigated under single and two-phase flow conditions. Analysing the flow behaviours under various conditions, enables the mechanisms of particle deposition on a glass window to be understood. Due to the complexity of the actual FBSR, a scaled FBSR was selected for this study. A well-defined and uniform in-flow condition was introduced below the aperture of the receiver. Computational Fluid Dynamics (CFD) were utilised to model the flow under gaseous and particle-laden conditions. The Renormalised Group Theory (RNG)-based k-ε turbulence model was used to capture the flow pattern at steady conditions. The Discrete Particle Model (DPM) was used to investigate the two-phase flow behaviour. The single phase flow results were validated against experimental data collected inside a similar device operating under the same conditions. The turbulent flow velocity was measured using a Turbulent Flow Instrumentation (TFI) cobra probe and a Pitot tube. The three-dimensional velocity components were measured at different radial and axial positions in the receiver for different Reynolds numbers. The FBSR was oriented vertically; consisting of a cylindrical cavity, above which were located a converging-diverging secondary cavity and a window glass. The bottom of the FBSR was considered as an inlet, with two tangential outlets placed closer to the secondary cavity. The results of this investigation revealed that mass flow into the secondary concentrator of the receiver was reduced significantly when the ratio between the outlets and inlet areas was 0.5, and the ratio between the aperture and receiver diameter was 0.41. Since the glass window was located at the top of the secondary concentrator, the lower circulation of the inlet flow into the secondary concentrator resulted in lower particle deposition on the glass window. When using window shielding jets, the number of jets were found to be critical for preventing particle deposition. At a constant mass flow rate, increasing the number of window shielding jets reduced the suction pressure from the core to the aperture. Consequently, the outward axial velocity towards the glass window was reduced. It was found that the introduction of particles into the flow influenced the flow pattern inside the receiver and affected the flow velocity on the glass window. In a gas-particle flow analysis, gravity was found to be important for capturing the flow patterns in the receiver accurately. When assessing the effect of particles on flow patterns under the same operating conditions, it was found that the average outward axial velocity, the maximum velocity and the aperture area with outward axial velocity were higher than for a single-phase flow. Apart from the aperture section, the slip velocity was found to be negligible in the receiver cavity, as is evident from the comparison of the fluid and particles’ velocity profiles. The findings of this investigation could potentially provide insights into the industrial application of FBSR, where the particles damage the glass window of the receiver during long-term operation.