Authors: F. Boshoff
Year: 2025
Industry: SCO2
Keywords: Mention FSE, Dynamic, Validation, Comparison
ABSTRACT
An axial flow fan was designed for an air-cooled heat exchanger (ACHE) unit of a concentrated solar power plant, which makes use of a supercritical carbon dioxide (sCO2) Brayton cycle. The ACHE’s geometry and cooling requirements are used to derive the fan specifications. This results in an unusual axial flow fan design, with a hub-to-tip ratio of 0.51 and low blade twist. The fan was simulated using computational fluid dynamics (CFD), by means of a periodic three-dimensional model (P3DM). Fan performance is evaluated in two different domains: an annulus upstream and downstream of the fan blade, and a domain representing an ISO 5801 Type A fan test facility. The P3DM is used to demonstrate the effect of tip gap size, hub configuration and scaling on performance. A 1:4.78 scale fan was manufactured and experimentally tested in an ISO 5801 Type A facility. The experimental and simulated performance show good agreement at all tip gap sizes and flow rates simulated. After correcting for blade modifications during manufacturing, the experimental design fan static pressure, shaft power and static efficiency values are 1.8 Pa (1.8%) lower, 0.013 kW (0.9%) higher, and 1.7% (3.2%) lower than the corresponding P3DM results. The fan and ACHE interaction was investigated using an isothermal CFD simulation, which models the bundle using a Porous Media Model. The operating point agrees well with analytical results, with a difference of only 0.9 m3/s (0.5%), being 6.3 m3/s (3.6%) below the design flow rate. A range of plenum heights and an extended annulus configuration are tested. The results indicate that plenum heights of 0.3 to 0.5 times the fan casing diameter facilitates the highest flow rates, whereas longer plenums improve the velocity profile’s uniformity. The extended annulus configuration is shown to improve both the flow rate and velocity profile uniformity. To investigate the integrated ACHE’s thermal behaviour, a co-simulation model was developed which simultaneously solves the air flow field using three-dimensional CFD, and the sCO2 flow field using a one-dimensional network method. The air side model is validated through comparison with the isothermal model, while the sCO2 model is validated through comparison with results from literature. The co-simulation predicts a higher heat exchanger pressure drop than the isothermal model, resulting in the flow rate being 18.5 m3/s (10.5%) below the design value. The overall heat transfer rate is also found to be 21.5 MW (20.5%) below the intended value. The results indicate that the sCO2 in tubes directly downstream of the fan blades are cooled to a slightly lower temperature than the sCO2 in the other tubes. However, due to the overall heat transfer rate being lower than intended, the outlet conditions are still well above the critical point