Thermal optimization of a natural draft wet cooling tower

N. Williamson*, M. Behnia, S. W. Armfield

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

27 Citations (Scopus)


This study attempts to quantify the potential improvement in a natural draft wet cooling tower (NDWCT) performance that can be attained by optimizing the fill and water distribution profiles across the tower and to provide designers with the modelling tools for such an investigation. A simple two-dimensional (2D) model is described, which allows rapid evaluation of NDWCT performance for use with an optimization procedure. This model has been coupled with an evolutionary optimization algorithm to determine the optimal fill shape and water distribution profile to maximize the cooling range of a typical NDWCT. The results are compared against a 2D axisymmetric numerical model. The extended 1D model is found to significantly reduce computational time compared with the numerical model, allowing a wide range of parameters to be tested rapidly with reasonable accuracy. The results show that the optimal layout differs significantly from a uniform profile, with both the water flow rate and the fill depth decreasing towards the centre of the tower where the air is warmer with reduced cooling potential. The overall improvement in the tower cooling range is very low under the design conditions tested, due largely to the highly coupled nature of the airflow and heat transfer in the tower. It is concluded that any design modifications of the type considered would need to be carefully optimized to have any possibility of improving performance.

Original languageEnglish
Pages (from-to)1349-1361
Number of pages13
JournalInternational Journal of Energy Research
Issue number14
Publication statusPublished - 2008
Externally publishedYes


  • Cooling tower
  • Merkel method
  • Optimization
  • Poppe method


Dive into the research topics of 'Thermal optimization of a natural draft wet cooling tower'. Together they form a unique fingerprint.

Cite this