Software calibration

The software is an analytical package that employs many semi-empirical equations that calculates the airflow loss coefficients (pressure drops) over the different components of the cooling tower. These air stream losses include:
  • Tower support losses
  • Drift eliminator losses
  • Spray zone losses
  • Fill losses
  • Tower inlet losses
  • Rain zone losses
  • Expansion losses
  • Inlet louvre loses
  • Fill support losses
  • Water distribution system losses
  • Contraction losses
and for mechanical draft towers, these losses as well:
  • Fan upstream losses
  • Fan downstream losses
  • Plenum losses
  • Fan diffuser losses
It is highly unlikely that the geometry of the actual tower that you want to analyze is 100% consistent with the geometry used when the empirical equations were determined. It is therefore unlikely that the software will 100% predict the actual cooling tower performance at the first attempt. Having said that, the semi-empirical equations employed in the software are very robust and are obtained from leading published research. The software will therefore give results that are pretty close to the actual cooling tower, if the software is used correctly.

How do you match the software solution to an actual tower?
For some of the losses mentioned above, only the loss coefficient can be specified. Some recommended loss coefficients for these are given in the literature. If the cooling tower performance does not match actual cooling tower performance closely then it is advisable to play around with these loss coefficients until the cooling tower performance matches closely. The cooling tower is then calibrated to the actual cooling tower.

Parametric studies
Even if the performance predicted by the software is not 100% matched to the actual cooling tower performance, the software can still be used with confidence to do parametric studies, i.e. to answer all those what-if questions. For example, what will the percentage heat load change be if the water flow rate is increased by 10%?

Blog Information