What Is a Cooling Tower? How It Works

Walk onto almost any commercial roof in the Gulf and you will find them: large boxes with fans on top, a steady plume of vapour, and a sound like heavy rain. Ask the people who work in the building what they are and you will usually get "the air conditioning". That is roughly true and completely unhelpful. A cooling tower does not cool the building. It cools water, and the water carries the building's unwanted heat out to the sky. Once you have that sentence, and understand that the cooling comes from evaporating a small part of the water rather than from blowing air over it, everything else about tower design, water treatment and hygiene follows logically.

The message up front: a cooling tower is a heat rejection device that uses evaporation to dump heat from a water circuit into the atmosphere. Because it evaporates water, it concentrates whatever was dissolved in that water, which is why blowdown and chemical treatment exist. Because it puts warm water in intimate contact with moving air, it can release fine droplets, which is why cooling towers are a legally regulated area in many jurisdictions and are treated as a hygiene asset, not just a mechanical one.

1. What a cooling tower is, in one paragraph

A cooling tower is a heat rejection device. Its job is to take water that has absorbed unwanted heat somewhere else in the system, bring that water into contact with a stream of atmospheric air, and send it back out cooler than it arrived. Everything inside the tower serves that one transaction: exposing the maximum water surface to the maximum air for the longest practical time, then separating the water from the air again before the air leaves.

It is worth being precise about the word "cooling", because the tower does not produce cooling the way a chiller does. There is no refrigerant, no compression cycle, no lift. A cooling tower moves heat that has already been collected, which makes it closer in spirit to a radiator than to an air conditioner, except that a radiator relies purely on temperature difference and a tower has a far more powerful mechanism available. Within the family of building services equipment described in the HVAC systems and components overview, the tower sits on the heat rejection side, at the very end of the chain.

2. Where the tower sits in a water cooled chiller plant

The clearest way to understand a cooling tower is to trace the heat. In a water cooled plant there are three separate circuits, and the heat passes from one to the next like a relay baton:

The cooling tower owns the last leg. Condenser water leaves the chiller carrying both the heat picked up from the building and the extra heat the compressor added doing the work of moving it. That warm water is pumped up to the tower, spread across the tower's internals, cooled, collected in the basin at the bottom, and pumped back to the chiller condenser to do it again. The loop is continuous and closed apart from the water the tower deliberately loses.

This is why a tower problem shows up as a chiller problem first. If the tower cannot deliver condenser water at the temperature the chiller was selected for, condensing pressure rises, the compressor works harder for the same cooling output, and energy consumption climbs quietly every hour the plant runs. The operator sees a chiller symptom and investigates the chiller, when the cause is a fouled fill pack or a fan running backwards on the roof. Read the chiller types and how chillers work explainer alongside this one; the two pieces of equipment only make sense together.

Not every building has a cooling tower. Air cooled chillers reject heat directly to the air through finned coils and do away with the tower, the condenser water pumps and the entire water treatment regime. The trade is thermodynamic: an air cooled machine can only reject heat down towards the ambient air temperature, and a tower can do better. The next section explains why.

3. How evaporative cooling actually works

This is the part most explanations skip or get subtly wrong, and it is the single most useful thing to understand about a cooling tower. The common assumption is that a tower cools water the way a car radiator does: air at one temperature flows past water at a higher temperature, and heat crosses from the hot side to the cold side. That mechanism is present, and it is called sensible heat transfer, but on a well designed tower it is the minor contributor.

The dominant mechanism is evaporation. When water evaporates it must absorb a substantial amount of energy to change from liquid to vapour. That energy is called latent heat, and it does not come from nowhere: it comes out of the water left behind. So when a small fraction of the circulating water evaporates into the air stream, it carries a disproportionately large amount of heat with it, and the remaining water, the great majority of it, drops in temperature as a direct consequence.

That is the whole trick: a cooling tower deliberately sacrifices a small fraction of its water in order to cool the rest. You can feel the same physics on your own skin. Step out of a swimming pool into a breeze and you feel cold, not because the air is cold but because water is evaporating off you and taking your body heat with it.

4. Wet bulb temperature, explained without the psychrometrics

If evaporation is doing the work, then the limit on a cooling tower's performance is not how hot the air is. It is how much more water vapour the air is capable of absorbing. Dry air has a large appetite for water and supports vigorous evaporation. Air that is already close to saturated cannot take much more, so evaporation slows and the cooling effect weakens. Two sites at the same air temperature can therefore get very different results out of identical towers, purely because of humidity.

The single number that captures this is the wet bulb temperature, and the original measurement was beautifully literal: take a thermometer, wrap the bulb in a wet cloth, and move air past it. Water evaporates off the cloth, that evaporation cools the bulb, and the thermometer settles below the true air temperature. How far below depends entirely on how readily the air accepted the moisture. In very dry air the drop is large; in very humid air it is small; in fully saturated air there is no drop at all, because no net evaporation can occur. So wet bulb is best understood as the lowest temperature that evaporation alone can achieve in that particular air, right now. The ordinary air temperature, measured by a normal thermometer, is called the dry bulb temperature to distinguish it.

That gives us the theoretical limit for any cooling tower: it can approach the wet bulb temperature of the entering air but it can never reach it, and it certainly cannot go below it. Wet bulb is the floor. Everything a tower designer does, more fill surface, better water distribution, more airflow, longer contact time, is an attempt to get closer to that floor for an acceptable amount of fan power and capital cost.

The implication for anyone operating a plant in a humid coastal climate is worth stating plainly. In the Gulf the summer wet bulb is high, and highest on humid coastal days, so the floor rises and every cooling tower on every roof loses capability simultaneously at exactly the moment the building most needs cooling. That is not a fault and no amount of maintenance will fix it. It belongs in the plant design and in the operator's expectations rather than being treated as a breakdown.

5. The components and what each one does

Open a cooling tower access panel and the internals are simpler than they look. Every part does one of four jobs: spreading the water out, getting air through, keeping water and air together long enough to work, or separating them again at the end.

Two observations from walking a lot of tower decks. First, the fill and the water distribution system between them account for most of the performance a tower has, and they are also the two components least likely to be inspected properly, because doing so means opening the tower and getting wet. Second, drift eliminators are routinely treated as optional trim during a fill replacement, and they are not.

6. The type distinctions and what each one changes

Cooling tower terminology sounds like jargon, but each distinction corresponds to a real engineering decision. There are four worth knowing, and they are independent of each other, so a given tower is described by a combination of them.

The practical point for a facilities manager is that these labels tell you in advance what your maintenance access will be like and where your risks sit. A forced draught counterflow tower wedged into an enclosure with poor inlet clearance is telling you that recirculation and blind internals will be your recurring problems. A crossflow induced draught tower on an open roof is telling you to expect easier inspection, and more sunlight and debris in the hot water basins.

7. The water chemistry cycle: why blowdown and treatment exist

Here is the consequence of evaporative cooling that catches people out. When water evaporates, only the water molecules leave. Everything dissolved in that water, the calcium, the magnesium, the chlorides, the silica, stays behind in the circulating volume. The tower tops up with fresh make-up water, which brings its own load of dissolved solids, and evaporates again.

The arithmetic is relentless: pure water leaves, mineral-bearing water comes in, and the circulating water gets progressively more concentrated. Left alone it would keep concentrating until minerals precipitated out and deposited on every surface in the system. That is not a theoretical risk, it is the normal behaviour of any evaporative system, and it is why a cooling tower needs a water treatment regime while a sealed chilled water circuit largely does not.

There are two tools for managing it. The first is blowdown, also called bleed: deliberately discharging some of the concentrated circulating water so that the make-up replacing it dilutes what remains. This is usually automated, with a conductivity probe measuring dissolved solids and opening a bleed valve when the reading passes its setpoint. The ratio between the concentration in the circulating water and that in the make-up water is called the cycles of concentration, and it is the main dial a water treatment specialist turns. Higher cycles save water and chemicals but push the system closer to scaling; lower cycles are safer chemically but waste water, chemicals and the energy used to produce that water. The correct setting depends on the make-up water quality at that specific site, which is why a figure copied from another building is meaningless.

The second tool is chemical treatment, dosed to a regime designed for that water, and it is aimed at four distinct problems that people tend to blur together:

• Scale. Dissolved minerals, principally calcium carbonate, coming out of solution and depositing as a hard adherent layer on hot surfaces. Scale is a thermal insulator and it forms most readily exactly where you least want it, on the condenser tubes and the fill, progressively destroying heat transfer while every gauge still reads normally.
• Corrosion. Electrochemical attack on the metal components, accelerated by oxygen, which an open tower supplies in abundance, and by chlorides, low pH and certain microbial activity. Corrosion is more serious than scale because it removes metal permanently. Scale can be cleaned off; a pinholed tube or a perforated basin has to be replaced.
• Fouling. Suspended solids settling out where flow is slow: airborne dust and sand drawn in through the tower, corrosion products and process debris. Fouling blankets heat transfer surfaces, restricts flow, blocks nozzles and creates sheltered sediment beds where organisms live protected from the circulating biocide.
• Biological growth. Microbiologically, a cooling tower is an excellent habitat: warm water, plenty of oxygen, sunlight in places, and a steady supply of airborne nutrients and organisms. Growth appears as algae in the lit areas, slime films on the fill and basin, and biofilm on pipe walls. Biofilm matters for three reasons: it insulates thermally, it drives localised corrosion underneath itself, and it shelters organisms from treatment chemicals.

8. Approach and range: the two performance terms

Two terms come up in every cooling tower specification and commissioning report, and they are constantly confused. Both are temperature differences.

Range is the difference between the water temperature entering the tower and the water temperature leaving it: how much the tower actually cooled the water on that pass. Range is set by the load and the water flow rate rather than by the tower design, since the same heat carried by a smaller flow produces a larger range. Range tells you how hard the system is working.

Approach is the difference between the water temperature leaving the tower and the wet bulb temperature of the air entering it. It measures how close the tower got to the thermodynamic floor described earlier, which makes it the real measure of tower capability and condition: it compares the outcome against the best that was physically available in those conditions. A smaller approach means a more effective tower, and driving it down costs money in fill surface, tower size and fan power, with diminishing returns as you near the wet bulb.

The value of understanding approach is diagnostic. Leaving water temperature alone tells you very little, because it moves with the weather: warmer leaving water on a humid day may be a tower performing perfectly against a higher wet bulb. But if the approach has widened compared with the same conditions six months ago, something has genuinely degraded, whether fouled fill, poor water distribution, reduced airflow or air recirculating back to the inlet. That comparison is the single most useful trend to hold on a cooling tower, and it is a derived metric that belongs in a plant analytics layer rather than a paper log, which is the subject of the chiller plant analytics guide.

9. Drift versus plume: two different things

This distinction is worth getting right because the two are routinely conflated, and only one of them is a problem.

Plume is the visible white cloud above a running tower on a cool or humid day. It is water vapour that left the tower as invisible gas within saturated air and then condensed into tiny visible droplets on meeting colder outside air, the same effect as your breath becoming visible in cold weather. Plume is pure water: no dissolved solids, no treatment chemicals, nothing from the basin, because it formed by condensation after leaving the tower. It is not a fault, it carries no hygiene significance, and its presence or absence tells you about the weather rather than the tower.

Drift is entirely different: liquid water droplets mechanically entrained in the air stream and carried out of the tower without ever evaporating. Because those droplets were part of the circulating water, they contain everything the circulating water contains, including dissolved solids, treatment chemicals and whatever biological content the system is carrying. Drift is why drift eliminators exist, and it is the mechanism by which a tower can release an aerosol into its surroundings.

So: plume is visible, harmless and weather-driven; drift is largely invisible and is what the hygiene controls are aimed at. Complaints about "the towers putting chemicals over the car park" are almost always about plume and almost always unfounded, but the correct response is to verify the eliminators rather than dismiss the question, because drift is real and the eliminators are the control that manages it.

10. Water hygiene: why cooling towers are treated differently

Every other component in an HVAC system is a mechanical and energy asset. A cooling tower is a mechanical asset, an energy asset and a public health asset, and the third category is what changes how it must be managed.

The reason sits in the physics already described. A cooling tower holds a substantial volume of warm water, open to the atmosphere, in a nutrient-rich oxygenated environment favourable to microbial growth, and it moves a large volume of air through that water at high velocity. That combination can generate fine water droplets which leave the tower and disperse into the surrounding air, where they can be inhaled by people nearby. No other common building services component does all of those things at once, and that single capability, aerosol generation from a warm open water system, is why cooling towers attract specific attention from health authorities and are a legally regulated area in many jurisdictions.

The control principles are consistent, and they explain why the maintenance regime looks the way it does:

• Keep the system physically clean. Sediment, scale, slime and debris are not only efficiency problems. They provide the sheltered surfaces where organisms establish themselves out of reach of treatment chemicals. Cleanliness of the fill, basin and distribution system is a hygiene control in its own right, not housekeeping.
• Manage water temperature where it is practicable. Microbial growth rates depend strongly on temperature. Design and operational choices that avoid unnecessarily warm stagnant water, dead legs and infrequently used branches remove favourable conditions rather than trying to treat them.
• Maintain an effective water treatment regime. Not simply dosing chemicals, but a regime designed for the specific water and system, with biocide strategy, scale and corrosion control, and verification that it is achieving its intended result rather than just running.
• Keep drift eliminators complete and in good condition. They are the engineered control limiting how much aerosol can leave the tower. Missing, broken or badly refitted eliminators defeat a primary control, and that most often happens immediately after other maintenance work.
• Monitor, and keep records. A control regime that is not measured is an assumption. Monitoring closes the loop between the intended regime and reality, and the records are how the organisation demonstrates the system was actually managed.
• Put competent people in named roles. A written control scheme, a person accountable for it, and specialist input are the norm here. Cooling tower hygiene is not a task to hand to a general maintenance contractor without a defined scheme and competent oversight.

11. What actually goes wrong in service

Cooling towers fail in a small number of predictable ways, and the striking thing is how few of them announce themselves. A chiller that trips gets attention within minutes; a tower losing capacity degrades silently for months while the energy bill absorbs the cost. The recurring patterns, roughly in order of how often I see them:

• Fouled or degraded fill. Scale, slime and silt coating the fill, or fill that has partially collapsed and closed off its own air paths. The approach widens, the chiller works harder, and nothing appears broken.
• Poor water distribution. Blocked nozzles or silted basin orifices leaving parts of the fill dry. Cheap to correct and frequently missed, because finding it means opening the tower while it is wet.
• Air recirculation and inlet restriction. Saturated discharge air finding its way back to the inlet, or the tower boxed in by later construction, adjacent plant or acoustic screening never assessed for airflow. The tower then works permanently against air far worse than ambient, and no maintenance activity fixes a clearance problem.
• Water treatment drift. A conductivity probe out of calibration, a dosing pump airlocked for weeks, a partially blocked bleed line, or a treatment contract delivering site visits and paperwork but not monitored outcomes. This failure mode has the longest tail, because the damage accumulates invisibly in the condenser tubes and the basin metalwork.
• Drift eliminators disturbed by other work. Removed for access during fill replacement or cleaning and either refitted badly or not refitted at all. A hygiene control quietly defeated by a maintenance activity.
• Fan and drive neglect. Gearbox oil past its service life, belts slack, bearings dry, blade pitch drifted, or a motor reconnected in the wrong rotation after a repair so the fan turns backwards and moves a fraction of its design airflow.
• Make-up float misadjustment. A tower quietly running its make-up to overflow around the clock, wasting water and dumping treatment chemicals to drain, with nothing on any dashboard to show it.

I am deliberately not turning that list into a maintenance schedule here. The task list, the frequencies, the shutdown and clean sequence and the records belong in the cooling tower maintenance checklist. For how the tower fits into the wider plant room routine alongside the chiller, the condenser water pumps and the log readings, see boiler and chiller preventive maintenance, and for the broader building services PM frame, preventive maintenance for HVAC systems.

One structural point, since it decides whether any of that maintenance is ever tracked properly. A cooling tower should be its own asset record, with its fan, gearbox and basin as child records, and it should not be modelled as an attribute of the chiller. Towers, condenser water pumps and chillers have different failure histories, cost profiles and regulatory obligations, and collapsing them into one record is how a site loses the ability to see that its tower has been degrading for two years.

12. When a tower is not the right answer

Evaporative cooling is thermodynamically excellent and it is not free of consequences, so it is worth stating where a cooling tower is the wrong choice. A tower consumes water continuously, by design, through evaporation and blowdown. In water-scarce regions, and in the Gulf in particular where much of the supply is energy-intensive to produce, that is a real cost and increasingly a real constraint. A tower also carries an unavoidable management burden: a treatment regime, a written control scheme, monitoring, records and competent oversight. For a small plant on a site with no in-house technical capability that burden can outweigh the efficiency benefit, and an air cooled chiller which trades some energy performance for the removal of the entire water system is often the better engineering decision overall.

There is a middle option. Closed circuit coolers and adiabatic or hybrid designs reduce or seasonally eliminate the evaporative water consumption, and closed circuit arrangements keep the process fluid protected. They cost more in capital and, in the closed circuit case, accept a thermal penalty from the extra heat transfer surface, which is the same trade-off present in any indirect arrangement and is covered generally in the heat exchanger explainer. What they do not do is eliminate water hygiene management, because wherever there is spray water and an air stream there is still an aerosol to control.

The idea to walk away with

A cooling tower is not an air conditioner and it is not a radiator. It is an evaporator that sacrifices a small fraction of its water to cool the rest, which is why it can push water below the temperature of the air around it, why its capability is governed by humidity rather than air temperature alone, and why its performance should be judged by approach rather than by leaving water temperature.

Everything else follows from the evaporation. Evaporation concentrates dissolved solids, so blowdown and chemical treatment are structural requirements rather than optional extras. Evaporation demands intimate contact between warm water and moving air, so the tower can generate an aerosol, which is why it is a hygiene asset governed by a written control scheme and competent oversight rather than a simple mechanical one. And because its failures are gradual and quiet, it will degrade for a long time without declaring itself, which is why trending approach against the commissioning baseline is worth more than any amount of visual inspection.

Final thoughts

Cooling towers sit at an awkward intersection. They are mechanical plant, so the maintenance team owns them. They are the largest single lever on a water cooled plant's energy performance, so the energy manager cares about them. And they are a regulated water safety asset, so the compliance function has a stake. In practice those three interests are held by three different people, none of whom has the complete picture, and the tower ends up managed in fragments.

If you take one operational habit from this article, make it the approach trend: record leaving water temperature and entering wet bulb together and compare the gap against previous seasons and against commissioning. You will see a tower degrading long before anyone notices it on a walk round. If you take one governance habit, make it this: establish who your competent person is for cooling tower water safety, confirm a written control scheme exists for your site under your own jurisdiction's requirements, and check that the drift eliminators went back in properly the last time anybody opened the tower.

Related reading: What is HVAC: systems and components, What is a chiller: types and how chillers work, Cooling tower maintenance checklist, What is a heat exchanger, Boiler and chiller preventive maintenance, Chiller plant analytics, Preventive maintenance for HVAC systems. Reference bodies: ASHRAE , ISO .