A cooling tower works by throwing water through moving air. Most of what leaves as vapour is harmless evaporation. The problem is the fraction that leaves as actual droplets of tower water, drift, because those droplets carry whatever the tower carries, including Legionella, and they are small enough to be breathed in by people who never go near the plant. The drift eliminator is the single engineering barrier standing between the recirculating water and the public’s lungs, and it is one of the least inspected parts of the whole system.
That neglect is the gap this piece addresses. Temperatures, biocide and conductivity get logged; the eliminator gets a glance, if that. Yet a degraded or missing eliminator quietly raises the breathable emission of a tower regardless of how clean the water test looked that week.
Drift is not evaporation, and not the visible plume
Three things leave a tower through the air, and they are routinely confused.
Evaporative loss is pure water vapour. It leaves no dissolved solids and carries no bacteria. It is how the tower cools.
The visible white plume on a cold morning is condensed water vapour, fog, and is also, in itself, not the hazard. A tower can plume heavily and drift very little, or plume invisibly on a warm day and still throw droplets. Judging risk by how much plume you can see is unreliable.
Drift is the genuine concern: liquid droplets of recirculating tower water entrained in the discharge air. Unlike vapour, drift carries the full chemistry and microbiology of the basin water at its current concentration. Because the tower concentrates dissolved solids and supports biofilm, drift is not dilute, it is a sample of the worst water in the system, aerosolised and released at height.
How a drift eliminator works
A drift eliminator is a bank of closely spaced, profiled blades or a moulded cellular pack fitted in the air path on the discharge side of the tower, after the heat-exchange fill and before the air leaves the unit.
It works by inertia, not filtration. The air leaving the fill is forced through a series of sharp directional changes, the blades zig-zag the flow. Air, being light, follows the bends. The heavier water droplets cannot turn fast enough, so they strike the blade surfaces, coalesce into larger drops, and drain back down into the tower rather than escaping. Air passes; water is intercepted and returned.
The performance number to know is drift loss, expressed as a percentage of the recirculating water flow rate. The lower the figure, the less water, and less aerosol, escapes. Modern, well-fitted eliminators are commonly specified to extremely low drift, with figures of around 0.001% of recirculating flow or better often cited, whereas older designs or degraded units can lose very much more. Treat any single percentage cautiously: it is a design figure for a clean, correctly installed pack, and it degrades as the component does.
Picture the airflow path
If you cannot see inside the tower, build the model in your head from bottom to top. Warm return water is sprayed or distributed across the top of the heat-exchange fill, the open honeycomb-like media that spreads it into thin films. Air is drawn or blown across or up through that fill, picking up heat, and, inevitably, fine droplets. Immediately downstream of the fill, in the path the air must take to exit, sits the drift eliminator pack: tight, angled channels the air threads through while the droplets are flung against the walls and run back down. Beyond the eliminator is the fan and discharge to atmosphere.
So the order the air meets components is: fill, then eliminator, then the outside world. Anything the eliminator misses goes straight out. If the pack is cracked, sagging, displaced or simply absent, the air takes a clear run from contaminated fill to open air with nothing in the way. That is the whole risk in one sentence.
What good and degraded condition looks like
Good condition is a complete pack: every section present, seated tightly in its frame with no gaps at the edges, blades or cells intact and aligned, no large mineral scaling bridging the channels, no biological slime, and no daylight or bypass routes around the sides where air can shortcut past it.
Degraded condition takes several forms, and each defeats the eliminator differently. Sections can be missing entirely, removed for a clean and never refitted is a classic. They can be displaced, lifted or shifted so air bypasses around the edge. The media can be broken or distorted by age, UV embrittlement on exposed units, or freeze damage. Channels can be blocked or fouled with scale and biofilm, which sounds protective but distorts the airflow and can tear water back off the surfaces. And the surrounding frame or seals can fail, opening a bypass path that no amount of intact media in the middle will fix.
Crucially, none of this shows up in a water sample. You can hold a satisfactory Legionella result and a clean dip slide and still be emitting far more aerosol than designed, because the water test measures what is in the basin, not what is leaving through the air.
Why a failed eliminator is a public-health problem
Cooling towers sit on roofs and plant decks, discharging at height into open air, often near intakes, walkways, windows and streets. Aerosol fine enough to escape an eliminator can travel on the wind well beyond the building footprint, and documented outbreaks have affected people who simply passed nearby [4]. This is why evaporative cooling devices are notifiable to the local authority and are treated as a distinct, higher-consequence hazard class within the guidance [1][3].
The eliminator is where engineering control meets that public exposure. Water treatment reduces how much Legionella is in the droplets; the eliminator reduces how many droplets get out. Lose the second barrier and the first is doing the work alone, with the consequence pushed outside your boundary where you cannot see who is affected. The aerosol route is the same one that makes cooling towers such efficient spreaders of infection in general [1].
What to inspect, and what to record
Eliminators should be examined as part of the periodic physical inspection of the tower’s internals, typically alongside the routine clean and disinfection rather than as a separate visit, and the inspection should be evidenced, not just performed [1].
When the tower is safely isolated, locked off and accessible, check and record: that all eliminator sections are present and correctly seated; that media is intact with no cracks, sagging, distortion or UV embrittlement; that channels are clear of heavy scale and biofilm; that there is no bypass at edges, seals or frame; and that any sections removed for cleaning have been refitted before return to service. Photograph defects. Note the make-up of the pack if known, so a replacement can be matched.
The two failures worth flagging hardest are the missing-after-clean section and the silent edge bypass, because both leave the water records looking perfectly healthy while the aerosol barrier is gone. If you find either, the corrective action is a component repair or replacement, not a re-dose of biocide, they fix different things.
This is general guidance on what drift eliminators do and how to assess them, not a maintenance procedure or a substitute for the manufacturer’s data and a competent, site-specific risk assessment. Specific drift-loss figures, inspection intervals and replacement decisions belong to that assessment and to the equipment’s own documentation. Nothing here is legal, medical or design advice.
FAQ
Is a big visible plume from the cooling tower a sign of a drift problem?
Not reliably. The visible plume is condensed water vapour and depends mostly on the weather, cold, humid air produces fog from a perfectly healthy tower. Drift is liquid droplets of tower water, which can be released with little or no visible plume on a warm day. Judge drift by the condition of the eliminator and the inspection record, not by how much you can see leaving the stack.
Can a drift eliminator be retrofitted to an older tower?
Often, yes, eliminator packs are a recognised upgrade where an older unit has poor or damaged media, and improving drift performance is one of the more direct ways to cut a tower’s aerosol release. The pack has to suit the tower’s airflow path and frame, so it is a design and selection question for the manufacturer or a competent water-treatment engineer, matched to the specific unit rather than chosen generically.
If my Legionella samples are satisfactory, does the eliminator condition still matter?
Yes, and this is the key point. A water sample tells you what is in the basin; the eliminator governs how much of that water leaves as breathable aerosol. The two are independent. A satisfactory sample with a missing or bypassed eliminator still means more aerosol is escaping than the design intends, so the component must be inspected on its own merits, not waved through on the strength of a water result.
How far can aerosol from a cooling tower travel?
Far enough to affect people with no connection to the building. Fine aerosol that escapes the eliminator can be carried on the wind, and outbreaks have involved cases located some distance from the source tower. The exact reach depends on droplet size, height of discharge, weather and surroundings, which is precisely why the eliminator, the thing limiting how much aerosol gets out, is treated as a public-health control rather than a maintenance detail.
What to do next
Pull your last tower inspection record and look for one specific thing: an explicit, dated note on the condition of the drift eliminator, ideally with a photograph. If the eliminator is only ever covered by the general phrase “tower cleaned”, or not mentioned at all, add a discrete eliminator condition check to the next physical inspection, with present/seated, intact, clear and no-bypass as recorded line items. That single change closes the most overlooked gap in most cooling-tower regimes.
Sources
- HSE, “Legionnaires’ disease. Part 1: The control of legionella bacteria in evaporative cooling systems (HSG274 Part 1)”. https://www.hse.gov.uk/pubns/priced/hsg274part1.pdf
- HSE, ACoP L8 (2013), “Carrying out a risk assessment”, p.12. https://www.hse.gov.uk/pubns/books/l8.htm
- HSE, “Cooling towers and evaporative condensers”. https://www.hse.gov.uk/legionnaires/cooling-towers.htm
- World Health Organization, “Legionella and the prevention of legionellosis”. https://www.who.int/publications/i/item/9241562978