Cooling Tower Monthly Legionella Risk & Water Treatment Inspection

ContextStructural deterioration is more than a maintenance concern: cracks and gaps allow debris ingress, create dead-water accumulation zones, and produce anaerobic micro-environments in crevices where biofilm establishes beyond the reach of any biocide programme. Pay particular attention to the junction between the basin and the outer casing — this is where standing water, corrosion, and biological growth concentrate. Any unsealed penetration, such as a conduit sleeve, pipe entry, or access hatch with a missing grommet, should be sealed immediately to prevent bird and insect ingress. Photograph all findings for the inspection report.

ContextFill media massively increases the surface area for water-air contact — which is precisely what makes it the primary Legionella amplification site when it deteriorates. Healthy fill should be structurally intact, smooth, and free of visible deposits. White or grey scale (calcium carbonate) restricts airflow and traps sediment that biocide cannot penetrate. A brown or grey gelatinous slime is biofilm — the protective matrix in which Legionella thrives, largely shielded from oxidising biocides by the extracellular polymeric substance layer. Collapsed or misaligned fill creates channelling: entire vertical sections carry no flow, leaving stagnant water sitting warm. Any zone showing more than approximately 20% blockage or visible biofilm warrants a deep-clean before the next monthly cycle.

ContextDrift eliminators are the last physical barrier between contaminated water aerosols and the surrounding atmosphere. Even a partially dislodged panel creates an uncontrolled exit for droplets that completely bypass all chemical treatment. Check every panel is seated flush with no visible daylight at the edges or frame joints. Scale or biofilm accumulation on eliminator surfaces reduces separation efficiency and increases the static pressure drop across the unit. Towers specified to meet a drift loss rate of 0.0005% or less will almost certainly exceed this with damaged or significantly fouled eliminators. Individual replacement panels are relatively low cost; the risk profile of a compromised eliminator is not.

ContextThe basin is where suspended material settles and organic matter accumulates — creating a nutrient-rich sediment layer that acts as a persistent Legionella reservoir even when bulk water biocide levels read correctly. Run a gloved hand along the bottom corners and edges: anything more than a thin, light film warrants a scheduled drain-and-clean. Green algae confirms UV light is reaching surfaces it shouldn't — inspect all covers and side panels for gaps. Dead leaves, insects, and bird matter are direct organic nutrient sources. A basin with an earthy or sulphurous odour is a significant warning sign of anaerobic microbial activity in the sediment layer, which standard aerobic biocide programmes do not address.

ContextBlocked distribution nozzles create dry zones on the fill media below — areas receiving no water flow where biofilm establishes and grows unchallenged between inspection visits. During normal operation, walk the full distribution deck and observe each nozzle: every one should produce a consistent, evenly distributed spray. Any nozzle showing a reduced, distorted, or completely absent pattern must be cleared or replaced. Scale-blocked nozzles can often be cleared with dilute acid descalant during planned maintenance; physically damaged nozzles need replacement. Record the total count of blocked or malfunctioning nozzles; more than approximately 5% non-functional in a single distribution zone should trigger a full deck inspection rather than targeted spot repairs.

ContextSample integrity is the foundation of meaningful microbiological results. Use sterile 500 ml bottles pre-dosed with sodium thiosulfate (to neutralise residual biocide and preserve viable organisms accurately), collected at approximately 30 cm depth — not from the surface film or directly adjacent to the outlet. Avoid sampling during or within 4 hours of an oxidising biocide dose, as artificially low counts from a recent dose will not reflect the system's true microbiological status. Culture testing takes 7–10 days but yields colony-forming unit counts that directly map to regulatory action levels. PCR testing returns results in 24–48 hours but detects both live and dead cell DNA; always interpret a PCR positive alongside current culture data, particularly in the weeks following a disinfection event. Use only a UKAS-accredited laboratory (UK) or AIHA-LAP equivalent (US).

ContextTemperature is the single most critical environmental parameter in Legionella risk assessment. Legionella pneumophila proliferates optimally between 25°C and 45°C, with peak reproductive activity around 35–37°C — effectively the natural operating range of a cooling tower's recirculating water in warm climates. The cold water leaving the basin should ideally remain below 20°C; this is often unachievable during summer without process changes, which is precisely why chemical control must be intensified in warmer months. Hot water returning from the condenser should ideally exceed 55°C, above which Legionella cannot survive. A basin reading above 25°C combined with a condenser return below 50°C is a high-risk temperature profile that requires immediate review of biocide dosing frequency and blowdown rate.

ContextFree residual is the active biocide immediately available to kill pathogens — analogous to blood pressure: it must sit within a specific working range, not simply be detectable. For chlorination programmes, the typical operational target is 0.5–1.0 mg/L free chlorine at the basin; a reading below 0.2 mg/L provides inadequate protection, while a reading above 3.0 mg/L accelerates corrosion of steelwork and elastomers and may indicate a dosing pump fault. Use a DPD colorimetric photometer rather than visual test strips, which lack the accuracy required for regulatory-compliant monitoring. Always test at a remote distribution point as well as the basin: if the basin reads 0.8 mg/L but a far-end sample reads 0.1 mg/L, there is excessive biocide demand somewhere in the circuit — a warning of significant organic loading that no amount of increased dosing will sustainably resolve.

ContextConductivity measures the total dissolved mineral content of the water. As water evaporates from the tower, minerals that cannot evaporate remain and concentrate — this is cycles of concentration (CoC). Divide the basin conductivity by the makeup water conductivity to determine CoC (for example, 2,400 µS/cm basin reading ÷ 480 µS/cm makeup = 5 cycles). Most treatment programmes target 3–5 cycles; running higher accelerates scale deposition and corrosion inhibitor depletion, and creates conditions in which some scale/corrosion protection products lose efficacy entirely. If CoC is trending upward across consecutive monthly readings with no change in cooling load, suspect a blowdown controller fault or a clogged blowdown valve rather than simply adjusting the setpoint.

ContextpH influences virtually every aspect of cooling water treatment performance simultaneously. Oxidising biocides are significantly more active at lower pH — chlorine at pH 6.5 is approximately 10 times more effective than at pH 8.5, meaning a system running at high pH needs proportionally more biocide to achieve equivalent kill. Scale inhibitors and corrosion inhibitors also have optimal pH windows; outside these ranges they may precipitate or become ineffective without any visible indication. A sudden unexplained pH rise can indicate alkaline cleaning product contamination or an acid feed failure; a sudden drop may indicate ongoing anaerobic bacterial activity (organic acids from deep biofilm) or an acid dosing pump running continuously. Always re-test biocide residual after any pH correction, since readings outside the correct pH range are unreliable for programme compliance purposes.

ContextVisually clear water is not necessarily clean water, but visually turbid water is almost certainly microbiologically compromised. A turbidity reading above 5 NTU indicates suspended solids — biological material, corrosion products, or scale particulates — that consume biocide, protect Legionella within particle surfaces, and block UV disinfection if this is part of your programme. Foam on the basin surface typically indicates one of three things: organic contamination from decomposing biofilm, detergent carry-over from a cleaning operation, or overdosing of polymer-based scale inhibitors. Persistent foam despite correct inhibitor dosing warrants a full water chemistry review and a laboratory total organic carbon (TOC) analysis to identify the source. An earthy smell from the recirculating water is a particularly reliable indicator of microbial activity that pure chemistry tests may not yet have captured.

ContextInhibitor depletion is silent: the water may look clear, temperatures appear acceptable, and biocide residuals read correctly — but unprotected metal surfaces are being attacked and scale is forming in low-velocity zones. Most contemporary treatment programmes incorporate a tracer dye or fluorescent marker proportionally linked to inhibitor concentration, enabling a simple fluorometer reading to confirm accurate dosing without time-consuming chemical assays. Where molybdate-based inhibitors are used, test directly for molybdate at the programme target (commonly 5–15 ppm). Phosphonate inhibitors require a specific colorimetric assay. Inhibitor levels consistently 20% or more below target indicate a dosing pump fault, an exhausted chemical drum, or an undetected system bypass — investigate all three systematically before assuming it is a calibration issue.

ContextDosing pump failure is the single most common cause of inadequate biocide residuals in cooling water systems, and it is frequently undetected for days or weeks because the pump appears to be operating. Confirm: the pump is powered and actually stroking (listen for the solenoid click or check the stroke counter), the suction tube is fully submerged and air-lock free, and the output matches the programme sheet. If there is any doubt about output rate, use a measuring cylinder at the outlet to verify mL per minute directly. A pump that has been manually adjusted to a lower setting since the last inspection — perhaps by a contractor who found the residual 'too high' during a parallel piece of work — should be reset to the programme rate, and the unauthorised change recorded as a non-conformance.

ContextNon-oxidising biocides — typically isothiazolinone blends, glutaraldehyde, or DBNPA formulations — are essential for managing biofilm communities and organisms that oxidising biocides alone cannot reliably penetrate. They must be applied at the correct frequency (most commonly every 14 days) and at the correct dose volume. The critical and frequently missed control is the blowdown restriction: blowdown must be halted or significantly reduced during the contact period, commonly 4–6 hours, to maintain adequate biocide concentration throughout the system water. Dosing a non-oxidising biocide then immediately running a standard blowdown cycle dilutes it below effective concentration within the first hour. Check the blowdown controller log or written manual records to confirm the contact window was protected.

ContextRunning out of biocide mid-cycle is an entirely preventable Legionella risk event that happens with some regularity on sites where chemical inventory is not tracked against consumption. Check each tank or drum, estimate remaining days of supply at the current dose rate, and reorder if supply falls below approximately 14 days. Verify that chemical storage temperatures are within the manufacturer's specification — many biocide formulations degrade when stored above 30°C or exposed to direct sunlight, and a degraded product may pass a basic residual colorimetric test while no longer achieving the kill kinetics specified in the programme. Bund condition is a separate but significant compliance issue: a biocide spill reaching a surface water drain is reportable to the Environment Agency or equivalent authority in most jurisdictions.

ContextA blocked injection quill delivers zero chemical to the system despite the dosing pump running at full output — you may see the pump operating normally while system residuals remain undetectable. Test each quill by briefly disconnecting the feed line (with appropriate PPE) to confirm flow, or use a chemical indicator downstream of the injection point if line disconnection is impractical. Non-return valves on injection lines prevent back-siphoning of system water into the chemical tank, and also prevent drainage of the chemical supply line when the pump stops — a failed non-return valve can drain a significant volume of chemical through gravity between dose cycles. Injection points should be positioned at least 300 mm upstream of any bend, dead zone, or flow restriction to ensure thorough mixing with the bulk water before it reaches the distribution system.

ContextThe blowdown controller is the primary mechanism for maintaining water quality within the designed cycles of concentration. A controller whose probe has accumulated a scale or biofilm coating will underread conductivity — allowing the system to drift to 8 or 10 cycles of concentration while reporting 4, silently stressing every aspect of corrosion and scale inhibition. Test the controller against an independent portable conductivity meter: if readings diverge by more than 10%, clean the probe with dilute acid or a proprietary descalant and recheck. A controller that has failed 'open' — continuously blowing down — wastes substantial volumes of treated water and chemicals; one that has failed 'closed' will allow conductivity to rise to damaging levels within days, particularly in hot weather with high evaporation.

ContextMakeup water consumption is a sensitive indicator of system health. A sudden unexplained increase without a corresponding rise in cooling load typically points to one of three causes: a float valve stuck in the open position, an abnormal evaporation rate from increased heat rejection, or an undetected water loss (leaking heat exchanger, cracked basin, or third-party connection). A sudden decrease is equally concerning: a failed or restricted makeup valve may be starving the basin, allowing water level to fall and causing pump cavitation while concentrating minerals to damaging levels. Maintaining a month-by-month consumption log creates a clear baseline from which anomalies are immediately obvious, and provides the data required for mandatory water efficiency or sustainability reporting on many larger estates.

ContextWater utilities periodically modify source blending ratios — particularly during drought conditions — which can significantly alter the hardness and conductivity of the incoming supply. If makeup water conductivity has increased by more than 15% from the programme baseline, the blowdown setpoint and inhibitor dose rates should be recalculated; the previous settings will no longer maintain the design cycles of concentration or the required inhibitor reserve. Sites fed by private boreholes face additional variability: groundwater quality can shift seasonally, and relying on an annual source water analysis for programme design is insufficient on these sites. Consider quarterly source water testing if your borehole supply shows historical variability.

ContextMechanical degradation of the fan system has a direct and underappreciated effect on Legionella risk: a fan running at reduced capacity cannot shed heat efficiently, and basin water temperature rises toward the Legionella proliferation band without any change in chemical dosing or water chemistry. Check fan blades with the unit safely isolated — look for cracks, erosion pitting on leading edges, or any blade visibly displaced from its design angle. Belt-driven fans should have approximately 12–20 mm of deflection under firm thumb pressure on the longest belt run; more indicates slippage and progressive belt overheating. Gearbox oil level and colour should match the manufacturer's specification; milky or discoloured oil indicates water ingress, which accelerates internal corrosion and bearing failure.

ContextReduced pump flow translates directly to reduced water velocity through the fill media, creating the extended dwell times and low-velocity zones where biofilm establishes more rapidly. Compare pump discharge pressure against the baseline value recorded in the WMP — a drop of more than 10% without any corresponding system change indicates either internal pump wear or partial blockage in the suction strainer. Cavitation produces a characteristic crackling or gravel sound from the pump casing and rapidly destroys impellers; it is usually caused by low basin water level or a partially restricted suction valve. Mechanical seal leaks should be addressed promptly: beyond the water loss, drips on hot surfaces near electrical components and motor windings create secondary hazards.

ContextStrainers are the system's immune system against debris recirculation, but a fouled strainer rapidly becomes the opposite — a warm, stagnant mass of biological material sitting in the flow path, consuming biocide and slowly releasing organic compounds into the water. Monthly inspection and cleaning of Y-strainers and basket filters is appropriate for most systems; sites near construction zones, tree canopy, or high-traffic areas may require fortnightly cleaning during peak fouling seasons. If a differential pressure gauge is installed across the strainer, use it: a ΔP that has doubled compared to last month's reading indicates significant fouling even before flow degradation becomes visible elsewhere in the system.

ContextAn unsecured hatch or displaced access panel is an open invitation for birds, insects, and wind-carried debris to enter the cooling tower — each of which adds biological loading and organic material that biocide programmes are not sized to manage. This final check is easily overlooked when an inspector is packing equipment at the end of a long inspection, particularly in warm weather. Make it a physical act: walk the perimeter of the tower and check each access point against a simple hatch list. Some sites also require that the access log — recording who entered, when, and for what purpose — is updated as the last administrative step. This log is occasionally requested during incident investigations to establish whether unauthorised access may have introduced a contamination source.

ContextIn any Legionella outbreak investigation, the monthly inspection log is among the first documents regulators request — and the first document that can demonstrate either a competent programme or a systemic management failure. Incomplete or retrospective records are treated as evidence of inadequate management, not as a minor administrative lapse. The log must include: exact date and time, inspector name and qualification reference number, every test result with its acceptable range stated alongside it, all visual observations including 'no abnormalities noted' for areas checked and found acceptable, and every corrective action raised with a reference number, assigned responsibility, and required completion date. Entries written at the time of inspection carry significantly more legal weight than those completed later from field notes.

ContextMost water management programmes define three response tiers: an operational target range for normal operation, an Action Level 1 trigger (investigate and take corrective action within 24–72 hours), and an Action Level 2 trigger (immediate action, escalation to the responsible person, possible system isolation or emergency disinfection). For Legionella culture results, HSG274 Part 1 guidance sets Action Level 1 at greater than 100 cfu/100 mL and Action Level 2 at greater than 1,000 cfu/100 mL — the latter potentially requiring notification to the local authority or HSE. Critically, do not wait for laboratory results before beginning corrective action if on-site parameters such as temperature, residual biocide, or turbidity already indicate the system is outside acceptable control. These parameters are available immediately; laboratory results are not.

ContextA single Legionella positive result at low count may fall within the acceptable operational band defined in your WMP, but three consecutive monthly positives — even at apparently low individual levels — constitute a trend that demands systematic root cause investigation, not simply a biocide boost and a note in the file. Document each result against the sampling date rather than the receipt date, as laboratory culture turnaround is typically 7–14 days and can create apparent date confusion in trend analysis. If PCR testing is used as a rapid screening method alongside culture, note that PCR cannot distinguish between viable and non-viable cells: a positive PCR result in the weeks following a chlorine shock dose may reflect dead cell DNA rather than an active Legionella population, and should always be interpreted in the context of concurrent culture data.

ContextThe responsible person is the named individual with legal accountability for Legionella risk management under ACoP L8 or equivalent legislation. Their countersignature is not a formality — it is the documented evidence that a senior, accountable individual has reviewed the system's status and either accepted the current risk level or authorised specific corrective action. An RP who countersigns reports without reading them is exposed to significant personal liability if an outbreak occurs. If the designated RP is unavailable, escalate immediately to the nominated deputy named in the WMP — do not leave the report unsigned to avoid a difficult conversation. In larger organisations, the RP countersignature also triggers required reporting to the property owner, insurer, or corporate risk register.

ContextFive years is the standard minimum retention period under UK ACoP L8 guidance; in the US, OSHA's general records retention rules and some state-level Legionella regulations effectively mandate indefinite retention for records related to any Legionella incident. Digital filing is acceptable and preferable from a disaster recovery standpoint, provided backups are maintained off-site and file naming conventions are consistent — include the building reference, cooling tower identifier, and inspection date in each filename to enable rapid retrieval during an investigation. Some institutional property managers, insurers, and public sector frameworks require quarterly summary reporting to a central register in addition to the statutory site-level retention; check contractual obligations as well as statutory minimums before assuming local filing alone is sufficient.