Cooling tower Legionella limits: what result requires action?
Summary
Legionella action starts at any detection in New York City, above 100 CFU/L under UK HSG274 Part 1, and at 10,000 CFU/L in Quebec. New York City counts all Legionella species under a binding rule, Quebec counts only Legionella pneumophila under its Safety Code, and HSG274 is guidance that UK dutyholders may meet another way (HSE, 2024).
New York City requires monthly Legionella sampling, with no more than 31 days between samples, beginning May 8, 2026 (NYC Health Department, 2026).
HSG274 Part 1 recommends routine Legionella monitoring of cooling systems at least quarterly, by culture to BS EN ISO 11731:2017 (HSE, 2024).
The Quebec Safety Code amendment on cooling tower maintenance published on 2014-05-28 came into force on 2014-07-12 (RBQ, accessed 2026).
Culture plates for Legionella are incubated for 7 to 10 days under BS EN ISO 11731:2017, as summarized in HSG274 (HSE, 2024).
These bands apply to Legionella culture results from cooling towers and evaporative condensers in New York City, Quebec and the UK. They do not cover hot and cold water systems, spas, qPCR results or other jurisdictions. In New York City, the maintenance program and plan can add stricter corrective actions, because Table 8-1 sets the minimum (NYC DOHMH, 2026).
What Legionella levels require action in a cooling tower, by jurisdiction?
Legionella levels that require action in a cooling tower start at any detection in New York City, above 100 CFU/L in the UK and at 10,000 CFU/L of Legionella pneumophila in Quebec. The table below decides what a culture result requires under each of the three rule sets. New York City bands come from Table 8-1 of 24 RCNY Chapter 8 as amended in 2026, UK bands from Table 1.11 of HSG274 Part 1, second edition, and Quebec bands from the Safety Code as explained by the Régie du bâtiment du Québec (RBQ).
Legionella result (CFU/mL and CFU/L) | Jurisdiction (source) | What to do |
|---|---|---|
Not detected | New York City, level 1 (24 RCNY Table 8-1, NYC DOHMH, 2026) | Maintain water chemistry and biocide levels |
Detected but below 10 CFU/mL (below 10,000 CFU/L), all species | New York City, level 1 (24 RCNY Table 8-1, NYC DOHMH, 2026) | Review the water treatment program; adjust water chemistry and biocide levels if needed |
10 to below 100 CFU/mL (10,000 to below 100,000 CFU/L), all species | New York City, level 2 (24 RCNY Table 8-1, NYC DOHMH, 2026) | Disinfect within 24 hours by raising the biocide concentration or using a different biocide; review the program; retest within 3 to 7 days |
100 to below 1,000 CFU/mL (100,000 to below 1,000,000 CFU/L), all species | New York City, level 3 (24 RCNY Table 8-1, NYC DOHMH, 2026) | Level 2 actions plus a visual inspection to decide on cleaning and further disinfection; retest within 3 to 7 days |
1,000 CFU/mL or more (1,000,000 CFU/L or more), all species | New York City, level 4 (24 RCNY Table 8-1, NYC DOHMH, 2026) | Increase biocides within 24 hours; within 48 hours hyperhalogenate, drain, clean and flush; retest within 3 to 7 days; notify the Health Department within 24 hours |
Not detected or up to 100 CFU/L (up to 0.1 CFU/mL) | UK (HSG274 Part 1 Table 1.11, HSE, 2024) | Keep control measures; keep the general aerobic count below 1×10^4 cfu/ml |
Above 100 and up to 1,000 CFU/L (above 0.1 and up to 1 CFU/mL) | UK (HSG274 Part 1 Table 1.11, HSE, 2024) | Reassess the control programme and aerobic count; adjust biocide dosage if control is poor; resample to verify and to check remedial actions |
Above 1,000 CFU/L (above 1 CFU/mL), or persistent low-level results | UK (HSG274 Part 1 Table 1.11, HSE, 2024) | Immediate action: resample, shot-dose biocide or raise continuous dosage, reassess the programme, resample again within 48 hours |
Legionella pneumophila below 10,000 CFU/L (below 10 CFU/mL) | Quebec (Safety Code, RBQ, 2014) | Continue the treatment and maintenance program |
Legionella pneumophila 10,000 to below 1,000,000 CFU/L (10 to below 1,000 CFU/mL) | Quebec (Safety Code, RBQ, 2014) | Find the cause, apply corrective measures and verify that they worked |
Legionella pneumophila 1,000,000 CFU/L or more (1,000 CFU/mL or more) | Quebec (Safety Code, RBQ, 2014) | Stop aerosol dispersal, apply the decontamination procedure at once, resample within 2 to 7 days; the laboratory reports to the RBQ and public health |
Count impossible because of interfering flora | Quebec (Safety Code, RBQ, 2014) | Find the cause of the interfering flora and apply corrective measures |
The UK row with the widest effect is the third: above 1,000 CFU/L, or with persistent low-level results, HSG274 Part 1 calls for immediate action, a shot dose or higher continuous biocide dosage, and a resample within 48 hours (HSE, 2024). In New York City, owners must report level 4 results to the Health Department within 24 hours of receiving them (NYC DOHMH, 2026). New York City defines hyperhalogenation as dosing the system with 5 to 10 ppm free halogen residual for at least 1 hour at an effective pH, with the disinfectant and pH recorded in the corrective action procedures (24 RCNY 8-05(f)(5), NYC DOHMH, 2026); handle the halogen biocide under its safety data sheet (SDS).
How do you use the cooling tower Legionella limits table?
You use the cooling tower Legionella limits table by reading what the laboratory counted, converting the count to both units, and starting the action for your jurisdiction within its deadline.
Check the laboratory report for the unit (CFU/mL or CFU/L), the species reported (all Legionella or Legionella pneumophila only) and any note on interfering flora.
Convert the count to the unit of your rule: multiply CFU/mL by 1,000 to get CFU/L.
Find the row for the jurisdiction where the tower operates and start the listed action.
Record the action and the retest date in the maintenance plan records.
Worked example (illustrative values): a laboratory reports Legionella species at 2,500 CFU/L, of which Legionella pneumophila is 2,000 CFU/L. In CFU/mL these are 2.5 and 2.0 (illustrative).
New York City: the all-species count of 2.5 CFU/mL is detected but below 10 CFU/mL, so level 1 applies and the owner reviews the water treatment program.
UK: 2,500 CFU/L (illustrative) is above 1,000 CFU/L, so HSG274 Part 1 calls for immediate action and a resample within 48 hours.
Quebec: Legionella pneumophila at 2,000 CFU/L is below 10,000 CFU/L, so treatment and maintenance continue unchanged.
Why do Legionella limits differ between countries?
Legionella limits for cooling towers differ between countries because the rules count different organisms, start their bands at different points and carry different legal weight. Quebec measures only Legionella pneumophila; the RBQ notes that most legionellosis cases in Quebec and elsewhere are caused by that species (RBQ, 2014). New York City requires test results of all Legionella species to be evaluated (NYC DOHMH, 2026). HSG274 Part 1 refers culture analysis to BS EN ISO 11731:2017, and its Table 1.11 names no single species (HSE, 2024).
New York City's rule states that "shall" and "must" indicate mandatory actions (NYC DOHMH, 2026), and the city strengthened its cooling tower law and rules after Legionnaires' disease clusters in the South Bronx in 2015 and Central Harlem in 2025 (NYC Health Department, 2026). Quebec's Safety Code requires the maintenance program to keep Legionella pneumophila below 10,000 CFU/L, and the RBQ says its values follow proposals from experts and some international standards (RBQ, 2014). HSG274 is guidance: following it is not compulsory unless stated, though following it will normally be enough to comply with the law (HSE, 2024).
In numbers, the UK bands act earliest: HSG274 Part 1 calls for immediate action above 1,000 CFU/L, while New York City requires disinfection from 10 CFU/mL (10,000 CFU/L). New York City's level 1 review applies to any detection, so how early it acts depends on the laboratory's detection limit. This suggests the New York City rule relies on monthly testing and review rather than low numeric bands; the adopted rule states no rationale for the band edges.
How do you convert Legionella CFU/mL to CFU/L?
To convert a Legionella result from CFU/mL to CFU/L, multiply by 1,000, because one litre holds 1,000 mL; to go from CFU/L to CFU/mL, divide by 1,000. New York City's 10 CFU/mL boundary therefore equals 10,000 CFU/L, the same number as Quebec's intervention threshold, although New York City counts all species and Quebec counts only Legionella pneumophila. The UK immediate-action level of 1,000 CFU/L equals 1 CFU/mL.
Units matter most at the low end because of the laboratory's limit of detection. HSG274 Part 1 states that to detect levels below 100 cfu/l reliably, the portion inoculated onto the culture medium should equal at least 30 ml of the original water sample (HSE, 2024). It also notes that different laboratory methods may lead to detection limits above 100 cfu/litre (HSE, 2024). A "not detected" result from a method with a detection limit of 1 CFU/mL, or 1,000 CFU/L (illustrative), cannot show whether a UK sample sits in the 100 to 1,000 CFU/L band. Ask the laboratory for its detection limit in the unit your rule uses before you read the table.
Do not convert qPCR results into these bands. HSG274 Part 1 states that qPCR units, genomic units per litre, are not directly comparable with the culture action levels in cfu/l, because the assay detects DNA from live and dead cells (HSE, 2024).
Does a Legionella result below the limit mean the cooling tower is safe?
A Legionella result below the limit does not mean the cooling tower is safe; it means one sample, taken days before the report, was below the band. HSG274 Part 1 states that "'Not detected' does not mean 'not present' or that there is no risk" (HSE, 2024), and that failure to detect legionella should not lead to relaxation of control measures and monitoring (HSE, 2024). Culture plates are incubated for 7 to 10 days (HSE, 2024), and one Quebec laboratory reports that results can take up to 12 working days (Bureau Veritas, 2017).
The CDC recommends maintaining measurable oxidizing disinfectant residuals throughout each day and operating at the lowest possible water temperature, below the 25 to 45 °C range favorable to Legionella growth where possible (CDC, accessed 2026). New York City requires a weekly bacteriological indicator, and a heterotrophic plate count or dip slide at 10,000 CFU/ml or more triggers disinfectant residual monitoring three times per day (NYC DOHMH, 2026).
The reverse also holds. The RBQ notes that exceeding its values does not always lead to an outbreak of legionellosis (RBQ, 2014).
Where the cooling tower Legionella limits do not apply
The cooling tower Legionella limits in this post do not apply outside New York City, Quebec and the UK, to qPCR or other non-culture methods, or to water systems other than cooling towers. HSG274 Part 1 covers evaporative cooling systems; Parts 2 and 3 cover hot and cold water and other risk systems (HSE, 2024). New York City defines a cooling tower broadly, as any device using recirculating water exposed to air for evaporative cooling, radiant cooling, or air cleaning or scrubbing (NYC DOHMH, 2026).
Bring in a specialist when a result reaches New York City level 4, Quebec's 1,000,000 CFU/L health risk threshold, or the UK immediate-action band, or when low-level positives keep returning. HSG274 Part 1 says results should be interpreted by experienced microbiologists (HSE, 2024). In New York City, the qualified person and the management team must evaluate all Legionella results (NYC DOHMH, 2026). In Quebec, the causes of a health risk result are identified with the professionals who prepared the maintenance program (RBQ, 2014). Lubinpla can organize your laboratory reports, treatment records and site history into one cited document for your review.
Frequently asked questions
How often must a cooling tower be tested for Legionella?
New York City requires Legionella culture testing at least every month during operation, with no more than 31 days between samples (NYC DOHMH, 2026). New York City also requires a hyperhalogenation between July 1 and August 31 each year, with a Legionella test 3 to 31 days after it (NYC Health Department, 2026). Quebec requires a monthly analysis in normal service, one at restart after wintering, and one 2 to 7 days after decontamination (RBQ, 2014). HSG274 Part 1 recommends at least quarterly monitoring, with weekly sampling at first when commissioning or changing a treatment programme (HSE, 2024).
Who must be told about a high Legionella result?
In New York City, results at level 4 must be reported to the Health Department within 24 hours of receiving them (NYC DOHMH, 2026). In Quebec, a Legionella pneumophila result of 1,000,000 CFU/L or more goes from the accredited laboratory to the RBQ and the regional public health director without delay (RBQ, 2014). HSG274 Table 1.11 sets no notification step.
What should you do first after a cooling tower result above the limit?
Start the action in your jurisdiction's row within its deadline, then check the disinfectant residual and pH, because the CDC ties chlorine-based disinfection to a pH below 8.0 and bromine-based to below 8.5 during remediation (CDC, accessed 2026). Handle each biocide under its safety data sheet (SDS) and, in the UK, a completed COSHH risk assessment (HSE, 2024). Drain disinfected water to the sanitary sewer only under applicable rules and permits (CDC, accessed 2026).