This paper presents a relative comparison of the impact of cruciform screens, perimeter screens and walkways on 3 × 6 cell forced draft air-cooled condenser's (ACC) thermal performance and dynamic fan blade loading under windy conditions. Numerical simulations were carried out for the three mitigation measures at two fan platform heights, four wind speeds and three wind directions. The results indicate that walkways are a robust solution to ACC wind effects and offer benefits in terms of thermal performance and dynamic blade loading under all wind conditions considered. Cruciform screens offered the most effective mitigation of wind-related thermal performance deterioration under certain wind conditions, but the impact of these screens is sensitive to the wind direction. The dynamic blade loading impact of cruciform screens is variable, and these screens are not recommended for dynamic blade loading mitigation. Perimeter screens offered the most effective mitigation of dynamic blade loading and were particularly effective at high wind speeds but often exacted a penalty in terms of thermal performance at moderate to low wind speeds. The results of this study indicate that a correctly configured wind mitigation system, potentially consisting of more than one individual mechanism, could help improve thermal performance and simultaneously reduce dynamic blade loading under windy conditions resulting in a robust, wind resistant condenser.

Fans operating at the edges of large-scale air-cooled steam condensers often do so under distorted inlet air flow conditions. These conditions create variations in the aerodynamic loads exerted on a fan blade during rotation which causes it to vibrate. In order to isolate the sources of the unsteady aerodynamic loads as well as their effects on blade vibration, a potential flow fluid dynamics code was written to determine the aerodynamic loads exerted on a fan blade as a function of its rotation. The lift and drag forces were exported to a finite element code approximating the fan blade as a cantilever beam. With these two sets of code the response of the blade when subjected to varying aerodynamic loads could be determined. Furthermore, the effect of changing certain parameters such as blade stiffness or damping can be investigated. It was found that the blade’s response closely resembles that which was measured at the full-scale facility and that slight changes to the blade’s stiffness can potentially reduce the vibrational amplitude but may also lead to resonance.

No abstract

Large-scale cooling system fans often operate under distorted inlet air flow conditions due to the presence of other fans and the prevalent wind conditions. Strain gauge measurements have been used to determine the blade loading as a result of the unsteady aerodynamic forces. However, these measurements are of the blade’s response to the aerodynamic forces and include the deformation as a result of the first natural frequency being excited. When considering the dominant first natural frequency and bending mode of the fan blade, one can approximate the fan blade as a cantilever beam that acts as a single degree-of-freedom system. The response of a single degree-of-freedom system can be calculated analytically for any excitation if the system’s properties are known. The current investigation focuses on using these equations to create an algorithm that can be applied to the measured response of a fan blade to then extract the aerodynamic forces exciting it. This is performed by using a simple non-linear, least-squares optimization algorithm to fit a complex Fourier series to the response and using the coefficients of each harmonic term to determine the Fourier series representing the excitation function. The algorithm was first tested by applying it to the response of a finite element cantilever beam representing a simplified model of the fan blade. Good results were obtained for a variety of excitation forces and as such the algorithm was then applied to the measured response of a full-scale fan blade. The full-scale blade was excited with a shaker where the forcing function could be accurately controlled. Once validated, the algorithm was applied to a set of strain gauge measurements that were recorded at a full-scale fan while in operation. The reconstructed aerodynamic loading showed increased forces when the blade passed beneath the fan bridge as well as when it approached the windward side of the casing.

South Africa’s coal-fired power stations use super heated steam to drive generator turbines. In arid regions, air-cooled condensers (ACCs) are used to condense the process steam. These ACCs consists of an array of over 200 axial flow fans, each driven by a motor via a reduction gearbox. Distorted fan inlet air flow conditions cause transient blade loading, which results in variations in output shaft bending and torque. A measurement project was conducted where the input and output shaft of such a gearbox were instrumented with strain gauges and wireless bridge amplifiers. Gearbox shaft speed and vibration were also measured. Torsional and bending strains were measured for a variety of operational conditions, where correlations were seen between gearbox loading and wind conditions. The input side experienced no unexpected loads from the motor or changing wind conditions, whereas output shaft loading was influenced by the latter. Digital filters were applied to identify specific bending components, such as the influence of fan hub misalignment and dynamic blade loading. Reverse loading of the gearbox was measured during the fan stop period, and vibration analysis revealed torsional and gearbox vibrations. This investigation documented reliable full scale ACC gearbox loads.