--- title: "Cooling tower winterisation, shutdown and safe restart: managing a tower that goes idle" source_url: https://legionella.io/articles/cooling-tower-winterisation-shutdown-restart/ canonical_url: https://legionella.io/articles/cooling-tower-winterisation-shutdown-restart/ pillar: "Building Types & Use Cases" summary: "How to handle a cooling tower that goes idle: the three layup choices, why a stagnant tower breeds Legionella, and the clean-and-disinfect before restart" primary_keyword: "cooling tower winterisation" date_published: 2026-06-27 date_reviewed: 2026-06-27 author: "Legionella.io editorial team (REMOTE TECH LTD)" reviewed_against: "HSE L8 and HSG274 guidance" region: "United Kingdom" license: "CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote, adapt or republish with attribution to REMOTE TECH LTD and a link to source_url." license_url: https://creativecommons.org/licenses/by/4.0/ --- # Cooling tower winterisation, shutdown and safe restart: managing a tower that goes idle A cooling tower that stops working does not stop being a risk. The moment the fans and pumps go off, you have swapped a treated, circulating system for a warm, nutrient-rich pond of standing water, and stagnation is exactly what Legionella waits for. The danger is not the idle period itself. It is bringing the tower back into service as though nothing happened. This is the part of the year most cooling-tower failures are seeded. A tower drops off the cooling load in autumn, the water treatment visits taper off because "it's not running", and in spring someone fires it up to meet the first warm day. Whatever grew over winter is now being driven out as aerosol the same afternoon. The job has three stages, and each has a decision in it: how you shut down, how you hold the tower while it sits, and how you bring it back. Get the first two right and the third is routine. Get them wrong and the third is the one that hurts someone. ## The core principle: you never stop being in control Under the L8 Approved Code of Practice, the duty holder controls the risk for the whole period the system is in their hands, running or not [1]. An idle tower is still your tower. "We turned it off" is not a control measure; it is the start of a new, different hazard that needs managing on its own terms. That reframes the whole question. Winterising a tower is not switching it off and walking away. It is choosing, deliberately, how the system will be held in a controlled state until it runs again, and recording that choice so the next person knows what state the tower is in before they touch it. ## Three ways to hold an idle tower There are really only three honest options when a tower comes off load. Which one fits depends on how long the tower will sit, the frost risk on your site, and whether anyone will keep visiting. ### Keep it treated and recirculating The tower stays wet, the biocide programme continues, and the water is kept moving rather than left to stand. This suits short or intermittent breaks where the tower may be called back at any time. The trade-off is that you carry the full cost and visit frequency of an operating system for a tower doing no useful cooling, and circulating water in freezing weather brings its own frost-protection burden. ### Drain down and lay up dry For a longer planned stop, draining the system removes the standing water that drives the risk. A dry, clean tower is a far safer thing to leave for weeks than a stagnant wet one. The catch is that "drain down" only helps if it is a genuine drain: low points, the pond, the pack, the pipework runs and dead legs all need to actually empty, because a few litres left in a sump or a flat pipe section is enough to seed regrowth. A half-drained tower is arguably worse than one kept treated, because everyone assumes it is empty. ### Full winterisation In hard-frost conditions a partial measure is not enough. Full winterisation combines a proper drain-down with frost protection of anything that cannot be emptied, isolation of supplies, and protection of the tower internals over the cold months. This is the right call for towers idle for a whole season in an exposed location, but it is also the one that most needs documenting, because the tower is being left in a deliberately non-standard state. Whichever route you take, the principle from the system's water chemistry still applies: a tower carries scale, corrosion products and biofilm built up over its running life, and laying it up does not remove any of that. The restart, not the shutdown, is where that debt gets paid. ## The restart is where it goes wrong Here is the rule that matters more than any other on this topic. A cooling tower that has been through a prolonged shutdown should be inspected, disinfected and, if required, cleaned before it is returned to service, and the same applies after disruptive work on the system [2]. HSG274 Part 1 uses a month or longer as the main prolonged-shutdown point, but your risk assessment may set a shorter trigger, especially in summer or healthcare settings. Check the current guidance and your written scheme before restart. The reason is simple. However you held the tower over winter, you cannot assume what is living in it. Biofilm in the pack, sediment in the pond, residue in dead legs, any of it can harbour Legionella, and the first run drives that contamination straight out through the drift as breathable aerosol. Starting an idle tower without a clean and disinfect is one of the clearest ways to turn a dormant system into a public-health incident. Treat the return to service as a controlled event, not a flick of a switch. ### Restart clean-and-disinfect checklist Use this as a recordable pre-start audit. The procedure detail, chemicals and dwell times belong to your method statement and HSG274 Part 1, not to numbers chosen on the day. - **Confirm the layup state first.** Check the logbook for how the tower was left, treated, drained, or winterised, before you assume anything about its condition. - **Inspect before you wet it.** Open up and look: sediment in the pond, scale and fouling on the pack, the condition of the drift eliminators, any debris or nesting. Photograph what you find. - **Physically clean the system.** Remove sludge, scale and biofilm by cleaning, not by relying on biocide alone to fix a dirty tower. - **Disinfect to the procedure.** Carry out a pre-clean and disinfection per HSG274 Part 1 and the written scheme, with the operatives, PPE and COSHH controls the chemicals demand. - **Reinstate the treatment programme before load.** Get biocide dosing, bleed-off and inhibitor running and proven so the tower is under control the moment it starts circulating, not a fortnight later. - **Take a baseline check.** Confirm the treatment is dosing and within range, and follow your control scheme on confirmatory Legionella sampling before the tower carries normal load. - **Record and certificate.** Issue the clean-and-disinfect record, update the logbook to "in service", and note the date of return. ## Don't lose the tower in the records The quiet failure on seasonal towers is administrative. A tower comes off load and falls out of the maintenance rhythm, the visits stop, the logbook goes blank, and nobody can later say what state it was actually left in. When restart time comes, the team is guessing. Two record points close that gap. First, log the shutdown decision: which of the three routes you chose, when, and what was drained or isolated. Second, log the restart event with the clean-and-disinfect certificate attached. Between those two entries the logbook should show the tower was held in a known state, not abandoned. There is also a notification angle: a cooling tower is a notifiable device, and taking it permanently out of use, or bringing a new one in, is something the local authority is meant to know about, so check whether your shutdown changes your notification status [3]. This is general guidance on managing an idle cooling tower, not a method statement. The cleaning and disinfection procedure, chemical choices, frost-protection measures and sampling regime are specific to your tower and belong to a competent person working to your site's written scheme of control. Use a water-treatment provider working to a recognised service-provider code of conduct [4]. Nothing here is legal, medical or design advice. ## FAQ ### Which layup option should I choose if I'm not sure how long the tower will be off? Default to draining and laying up dry rather than leaving it standing wet. A tower kept wet but no longer properly monitored is the worst of both worlds, and "we might need it next week" is how that happens. If genuine on-and-off use is likely, keep it treated and recirculating so it stays under control, but resource the visits that decision implies. Either way, the restart clean-and-disinfect rule still applies once it has sat beyond the guidance period. ### Doesn't draining for frost protection conflict with Legionella control? No, they pull the same way. Draining down for frost protection also removes the standing water that drives Legionella risk, so a properly drained, dry tower serves both aims. The conflict only appears if the drain is incomplete: trapped water in a sump or dead leg both freezes and stagnates. The fix is a thorough drain of every low point, not a choice between the two risks. ### Do I need to tell the local authority when I shut a tower down for winter? A seasonal shutdown where you intend to run the tower again is not normally the same as taking the device out of use, but a cooling tower is a notifiable device and changes to its status can matter. If a tower is being decommissioned, removed, or a replacement installed, that is reportable. When in doubt, treat your notification as a live record and confirm your obligations under the cooling tower notification rules. ## Related reading - [Cycles of concentration explained](https://legionella.io/articles/cooling-tower-cycles-of-concentration-water-chemistry/) - [Pre-commission cleaning and disinfection of cooling towers](https://legionella.io/articles/cooling-tower-pre-commission-cleaning-disinfection/) - [A worked cooling tower control scheme](https://legionella.io/articles/cooling-tower-control-scheme-worked-example/) - [Cooling tower Legionella sampling and action levels](https://legionella.io/articles/cooling-tower-legionella-sampling-action-levels/) ## Sources [1] HSE, ACoP L8 (2013), “Carrying out a risk assessment”, p.12. https://www.hse.gov.uk/pubns/books/l8.htm [2] HSE, HSG274 Part 1 (2024), “Evaporative cooling systems: types, design and operation”, p.9. https://www.hse.gov.uk/pubns/books/hsg274.htm [3] HSE, "Cooling towers and evaporative condensers". https://www.hse.gov.uk/legionnaires/cooling-towers.htm [4] Legionella Control Association, "Code of Conduct for Service Providers". https://www.legionellacontrolassociation.co.uk/