Rapid Prototyping of the Small Scale CSP Fan

Case Studies: Aerodynamics

The MinWaterCSP project aims to reduce the water consumption of concentrating solar power (CSP) plants by 75 to 95% relative to wet cooling systems and to improve plant efficiency by 2 to 3 % relative to current dry-cooled systems by introducing novel dry/wet cooling technology. This hybrid cooling system will make use of large axial flow fans to condense the process fluid.
As part of the project a new high efficiency, reduced noise axial flow fan was developed and manufactured for a 24 ft. full scale test facility at Stellenbosch University (constructed as part of the MinWaterCSP project). This fan is referred to as the M-fan. The M-fan was designed with an optimised hub-tip ratio and minimised exist kinetic energy velocity profile. At its design flowrate of 333 m³/s and a blade setting angle of 34°, full scale 3-D CFD simulations predicted a pressure rise of 116.7 Pa at a fan static efficiency of just below 60%.
A 1.5 m diameter scaled model of the M-fan was manufactured and tested in the BS848 fan test facility, also at Stellenbosch University. The test results showed a fan static efficiency at the design point of 59%, for a tip clearance of 2 mm, at a blade setting angle of 35°. Measurements performed on the 1.5 m fan do however indicate a small discrepancy in its effective, as-manufactured designed blade angle.
To confirm the design process for the M-fan, its outlet velocity profile was measured on a 630 mm diameter ducted fan test facility. The measurements included measuring the fan performance characteristics, as well as the inlet and outlet velocity profiles. The 630 mm diameter fan blades were 3D printed in ABS into two parts, glued together, and in the end mounted to an aluminium hub. In this case, this procedure guaranteed the maximum prototype cost reduction. The results showed that the small scale fan can replicate the operating point of the 24 ft fan albeit at a slightly lower efficiency due to the effects of a different installation. All fan tests were fans performed at a tip speed of 58 m/s.
A comparison of the 1.5 m and 630 mm diameter fan test results to performance values scaled from the 24 ft. CFD simulation data, indicate that accurate correlation of data is dependent on the accuracy of the fan manufacturing process, the accuracy of measurements performed and the CFD process used for the simulation of the fan.