Case Studies in Thermal Engineering (Aug 2024)
ANN and CFD driven research on main performance characteristics of solar chimney power plants: Impact of chimney and collector angle
Abstract
Solar energy systems operate directly connected to the sun. Solar chimney power plants are privileged systems that can provide power output even in cloudy weather and during hours when there is no sun. The design and sizing of this system, which researchers focused on after its first application in the 1980s, is very effective on its performance. In this study, the collector slope and chimney slope that give maximum power output for the Manzanares pilot plant are investigated with a 3D CFD model. Simulations made using the RNG k-e turbulence model and the DO (discrete ordinates) solar ray tracing algorithm provide results that are in high compatibility with experimental data and literature. It is understood that the system provides maximum power at 0.6° collector slope and 1.5° chimney divergence angle. It is seen that the system, which gives a power output of approximately 46 kW in the reference case, exceeds the power output by 4.5 times and reaches 216.853 kW in the design that includes the collector and chimney slope. The effects of the main elements of the system on the performance are also included by changing the collector radius and chimney height while preserving these inclination angles. More than the power output in the reference case, 49.233 kW, can be achieved with the inclined design, with a collector radius of 73.2 m and a chimney height of 155.68 m. Although the effect of increasing the chimney height on power output continues after 1.2 floors, its effect decreases. In the study, it is seen that increasing the chimney height and changing the collector radius provide a greater increase in power output. Furthermore, the scope extends to the incorporation of an Artificial Neural Network (ANN) model, presenting a novel approach to predicting SCPP system performance. The findings ascertain the utilisation of 9 neurons in the hidden layer of the ANN, demonstrating a precise alignment with the study data.