Beyond the Hour Meter: A Practical Audit of Ultraviolet Disinfection in UK Buildings

Ultraviolet disinfection has an appealing simplicity in the plant room: water passes a lamp, and the lamp is either on or off. That simplicity can encourage a false assurance. A unit may show 4,000 hours on its counter while delivering a fraction of the intended dose, because lamp output has fallen, the quartz sleeve is fouled, the UV transmittance of the water is lower than assumed, or the sensor is reading against an ageing reference. For facilities managers responsible for potable, process or reclaimed water, the useful question is not how long the lamp has run, but what evidence exists that the required dose was delivered to the water under the conditions actually encountered.

Why hours are a proxy, not a proof

UV disinfection works by exposing microorganisms to ultraviolet light, principally at around 254 nanometres. The delivered dose is a product of intensity and exposure time, and both are affected by real plant-room conditions. Lamp ageing reduces output well before the lamp fails. Sleeve fouling — from hardness, iron, manganese or general deposits — absorbs UV before it reaches the water. Flow rate determines contact time, and a booster set or valve change can alter that rate without anyone touching the UV unit. The transmittance of the water itself, expressed as UVT at 254 nm, may be lower than the design assumption if upstream treatment has drifted. A lamp-hour counter captures none of these variables.

Manufacturers typically publish a lamp life at which output falls to a declared percentage of initial. That figure is a replacement trigger under stated conditions, not a guarantee that dose has been maintained throughout the interval. In a well-run plant room, hours are one line in a maintenance file. They are not the conclusion of a performance assessment.

Start with the validation and commissioning file

Before inspecting hardware, establish what the unit was purchased to do. Look for the validation basis: the dose claimed, the flow rate at which it was validated, the UVT assumed, the lamp and sleeve type, and the sensor arrangement. UK facilities may encounter units validated to recognised protocols such as DVGW W294 or the USEPA Ultraviolet Disinfection Guidance Manual, or supported by manufacturer-specific validation. The point is not to debate the protocol but to check that site conditions still sit inside its envelope.

Commissioning records should show the as-installed flow, the measured UVT, sensor calibration at handover and the alarm set-points. If the file is missing, the audit should say so plainly. A missing validation basis is a finding, not a paperwork inconvenience.

Inspect the lamp and sleeve as a pair

Lamp hours are usually recorded, but sleeve condition is often not. During a plant-room visit, isolate the unit according to site procedure and inspect the quartz sleeves for clouding, scale, staining or physical damage. Check lamp ends for blackening or moisture ingress. Confirm that the lamp part number matches the validated type; a higher-output lamp is not automatically an improvement if the ballast and sensor were not designed for it.

Record the date of the last sleeve clean and the method used. Mild acid cleaning may be appropriate for scale, but abrasive pads and improvised chemicals can scratch or etch the quartz. A scratched sleeve is a permanent optical loss. Photographs at each service interval make it easier to distinguish normal wear from a developing problem.

Challenge the sensor and its calibration

The UV intensity sensor is the unit’s own witness, so its health matters. Check the calibration certificate, the reference method used and the interval since the last calibration. A sensor that has drifted low will overstate dose; one that has drifted high may trigger nuisance alarms and encourage operators to ignore them. Where a unit relies on a single sensor, confirm that its position and window are clean and that the manufacturer’s recommended calibration frequency is being met.

Ask whether the sensor reading is treated as a relative trend or an absolute dose indication. Many sensors are not reference-grade instruments, and the displayed percentage is not a direct measure of delivered dose in the water. The audit should record what the signal is used for: alarm, duty control, or trend monitoring. Each use carries a different tolerance for drift.

Check transmittance assumptions against reality

UVT is the quiet variable in many underperforming installations. If the design assumed 95% transmittance and the actual water sits at 85%, the delivered dose falls even with a healthy lamp. Upstream softening, filtration, coagulation or blending changes can shift UVT. So can seasonal raw-water quality on private supplies.

Where the unit serves a process or reclaimed system, look for recent UVT measurements. Where it serves potable water, the water supplier’s data may be available, but site-specific measurement is better if the UV unit is downstream of site treatment. If no UVT measurement exists, the audit should flag the assumption as unverified rather than assuming the design figure still holds.

Read the alarm history as a record of drift

Alarm logs are often cleared after an event and forgotten. Download or photograph the event history before it is reset. Look for repeated low-intensity alarms, lamp failure alarms, high-temperature trips and flow interlocks. A pattern of alarms that are routinely acknowledged without investigation is a stronger warning than a single event.

Check that alarm set-points have not been altered to silence nuisance trips. Confirm that the unit fails safe: if the intensity falls below the set-point or flow exceeds the validated maximum, does the system divert, alarm or shut down as the risk assessment requires? The answer should be documented, not assumed.

A simple proving routine

A practical monthly routine can be completed in twenty to thirty minutes. Record lamp hours and compare them with the replacement schedule. Note the sensor reading and any recent calibration date. Inspect the sleeve visually through the sight glass if fitted, and confirm that the alarm log has been reviewed. Check flow against the validated maximum and record any changes to upstream treatment. Every quarter, add a sleeve inspection and a UVT check where the site has the means to measure it. Annually, review the validation basis, sensor calibration and alarm set-points against the current duty.

The output should be a short evidence sheet: date, unit identifier, hours, sensor reading, flow, UVT if available, alarm status, sleeve condition and actions. That sheet does not prove absolute disinfection. It does show that the unit is being operated within its stated envelope, and it gives the next person a defensible record.

What the audit can and cannot say

A plant-room audit can confirm that a UV unit is installed, maintained and monitored in line with its validation basis. It cannot, by itself, certify microbiological performance. That requires validation, commissioning data and, where appropriate, sampling under a documented regime. The honest position is that UV is a barrier whose performance depends on conditions that change. Lamp hours are a useful prompt to act, but they are not the dose. Facilities teams that record the variables around the lamp — sleeve condition, sensor calibration, transmittance, flow and alarms — will be better placed to defend the barrier when it matters.