The operating cost per unit for evaporative cooling is low compared with air conditioning technology, but its benefits can be very application-dependent, says Reginald Brown
The use of evaporative cooling to reduce indoor temperatures is not new.
The ancient Egyptians used rudimentary evaporative cooling more than 4,000 years ago, while many other cultures, particularly in the Middle East, developed practical evaporative cooling strategies for comfort and to reduce food spoilage, which have continued in use until the present day.
Leonardo da Vinci, inventor of the first practical hygrometer in the 16th century, would have no difficulty in understanding the evaporative cooling systems of some 21st century datacentres.
The difference between then and now is the availability of efficient fans to drive air through the evaporation media and our thorough understanding of the thermodynamics of evaporation.
The advantage of evaporative cooling is that under ideal conditions we don’t have to input much energy to get a large cooling effect; usually just the fan power. The disadvantage is that the cooling effect is dependent on the existing humidity.
We can only get down to the wet bulb temperature. If the outside air is already saturated with moisture, then we won’t get any further cooling.
It might be assumed that evaporative cooling would only be successful in hot dry countries, but that is not true. In the UK climate, there is plenty of opportunity to use evaporative cooling provided we accept that it won’t always be available. Evaporative cooling is therefore considered for two main scenarios:
- To integrate with and offset mechanical cooling in high thermal load buildings when weather conditions permit;
- To provide a low cost comfort cooling option for non-air conditioned building for those few summer days that don’t justify the installation of a mechanical cooling system.
But there are circumstances where increasing the humidity of the ventilation air as a consequence of evaporative cooling is undesirable.
An alternative approach is indirect evaporative cooling, where the cooling effect is transferred to the ventilation air in a heat exchanger.
While compact air to air heat exchangers are readily available, the fundamentals of heat transfer mean we can’t transfer 100 per cent of the cooling effect. For example, a direct evaporative cooler might achieve a temperature drop of 5K, but only 3K of this is available on the dry side of the heat exchanger.
This reduction in temperature performance is less of a problem in data centres, where the priority is to offset mechanical cooling, but has been a significant barrier to the widespread adoption of indirect evaporative cooling for standalone comfort cooling.
But there is a relatively simple solution.
In 1987, Dr Valerij Maisotsenko and Alexandr Gershuni invented a thermodynamic cycle based on a combination indirect heat exchanger and evaporator where part of the cooled dry atmospheric air is passed back through the evaporator layers.
This works because cooled air with the same moisture content has a lower wet bulb temperature and can be subjected to further evaporative cooling.
In the product that was later developed by Maisotsenko in the US (“Coolerado”), the diversion of air occurs at many points within a complex multi-layer plastic heat exchanger.
The dry cooled air is used within the building while the wet air is discharged to the atmosphere.
The overall result is that the cooling effect is amplified. Cold dry air is available at a temperature that can be below the wet bulb temperature of the original atmospheric air; something that a direct evaporative humidifier cannot do.
Although individual Coolerado modules are relatively small in terms of cooling effect (5-10 kW in UK summer mid-day conditions with a peak COP of up to 20), multiple modules can be combined for larger scale applications.
While the technical challenges and performance issues of indirect evaporative cooling have largely been solved, it is not a cheap product, particularly when compared with direct evaporative cooling.
The operating cost per unit cooling is very low compared to an air conditioner but the number of hours per annum when the availability of cooling and demand coincide is very application-dependent.
Any proposed project should therefore undertake hourly modelling of performance over a full year and whole-life costing to assess its viability.
Reginald Brown is head of energy and environment at BSRIA