A calibrated thermometer in a competent person’s hand has carried Legionella temperature monitoring for decades, and it still does the job. The real question is narrower than the sales pitch suggests: where do fixed wireless sensors genuinely close a gap that manual readings leave open, and where are they an expensive way to record what a five-minute walk would have caught anyway?
Get that framing right and this stops being old versus new. It becomes a question about your actual blind spots, and which method covers them.
The real decision: depth or breadth
The two approaches are good at opposite things, and most arguments about them go wrong by ignoring that.
A manual reading is a snapshot with breadth. A competent person works round the building with a calibrated thermometer, hitting the sentinel outlets — usually the points nearest to and furthest from each calorifier or cold-water tank — and rotating through other outlets over the year [2]. One visit can touch dozens of points, and the person taking the readings sees what a number never shows: a weeping TMV, scale crusting a shower head, a stored-cold tank quietly warming in a hot plant room.
A wireless data logger is the mirror image: depth at a fixed point. Fix a sensor to a sentinel outlet or to a calorifier flow and return, and it records continuously, sending readings over a wireless link to a dashboard. It catches the 3am drift on the hot return, the bank-holiday weekend when the cold tank crept past 20°C, the slow decline no monthly visit would have seen in time [3]. But it only knows the point it is bolted to. A sensor on one outlet says nothing about the two hundred it is not on.
So the honest comparison is continuous-but-narrow against periodic-but-wide. Neither is simply “better”. They fail differently, and that is the whole game.
Where the comparison actually bites
Two things decide this more than continuous-versus-snapshot does, and both tend to get glossed over.
The first is calibration. A reading is only worth the accuracy of the instrument behind it, and that cuts both ways. A handheld thermometer needs periodic calibration; so does every fixed sensor, which drifts with age and rarely announces it. “Install and forget” is the quiet failure mode of remote temperature monitoring — a wall of green readings that are all two degrees optimistic. Whichever route you take, sensor calibration and thermometer calibration are a recurring line in the plan, not a one-off at purchase.
The second is who answers the alert. A data logger that emails a threshold breach to an inbox nobody owns is worse than a paper logbook somebody actually reads, because it manufactures confidence while changing nothing. The value of continuous monitoring lives entirely in the response process hung off it: a named person, a threshold that matches the written scheme, and a defined action when a reading falls outside it [1]. Without that chain, you have bought a very precise way to not notice a problem.
How the two stack up
| What you’re comparing | Manual readings | Wireless data loggers |
|---|---|---|
| Evidence pattern | Point-in-time snapshot at each visit | Continuous, including nights, weekends and voids |
| Coverage | Rotates across many outlets over time | Limited to the points where sensors are fitted |
| Physical faults | A person spots leaks, scale, damage, smell | Reports only the value it measures |
| Effort shape | Recurring labour at every visit | Front-loaded install, then review and alert handling |
| Calibration | Calibrate the thermometer periodically | Verify or recalibrate every sensor periodically |
| Typical failure | Gaps and unrecorded checks when staff are stretched | False confidence from a dashboard no one owns |
Which to use, and where
For a small, stable, well-supervised site — a single office, a low-rise with a caretaker who knows the building — manual readings are usually proportionate and often enough on their own. The risk assessment sets the frequency, not a supplier’s subscription model [2].
For large, complex or thinly-staffed estates the calculus shifts. Continuous logging earns its place on the points where drift is the real danger and a person cannot be present round the clock: calorifier flow and return, the hardest-to-reach sentinel outlets, a cold tank baking in a warm roof space. Let the sensors hold the line between visits, and send the competent person to do what sensors can’t — inspect, flush, descale, and use their eyes.
That blend is where most well-run buildings land. Loggers for continuity at the few critical points; manual rounds for breadth, physical inspection and the low-use outlets that need a human to run them off anyway. If you are weighing this across a portfolio, Integrating Legionella control with Building Management Systems covers feeding sensor data into systems you already run, and Smart thermometers — using IoT for Legionella control goes deeper on the hardware itself.
What neither method changes
This is temperature monitoring, not Legionella control, and the two are easy to blur. A flawless record — paper or live dashboard — proves a temperature was achieved at a point; it does not flush a dead leg, service a TMV, clean a tank, or replace the judgement of a competent person reading the trend [4]. The exact temperature targets, monitoring frequencies and the action you take on an exception all come from your site-specific risk assessment and written scheme, applied by someone competent — not from a comparison table, this one included. Treat any figure here as general guidance to confirm against current HSE references for your own system.
Start here
Before you price a single sensor, mark up a plan of your system with the points that matter: every sentinel outlet, each calorifier, every stored-water tank, and the outlets that already give you trouble. Against each, write what your current manual record actually proves and where it goes blind — the overnight hours, the long voids, the corner nobody visits. Wherever a real blind spot lines up with a high-risk point, that is where a logger pays for itself. Everywhere else, a calibrated thermometer and someone who reliably turns up is still the standard to beat.
FAQ
Do wireless data loggers replace the monthly manual checks?
Not by default. They can reduce how much manual reading you do at the points they cover, but your risk assessment sets what monitoring is required and how often [2]. Many sites keep manual rounds for breadth and physical inspection while loggers cover continuity at the critical points.
Will continuous logger data satisfy an HSE inspector or auditor?
Good continuous data is strong evidence, but the records are not the point — the management behind them is. An auditor will look for calibrated instruments, thresholds tied to your scheme, named ownership and a clear action trail when a reading goes out of limit [1]. A dashboard without that is weaker evidence than a well-kept paper log.
How often do the sensors themselves need calibrating?
Treat verification and recalibration as a scheduled, recurring task rather than a one-off, on the interval the manufacturer and your risk assessment support [4]. A sensor drifting unnoticed is one of the few ways automated monitoring can leave you worse off than the manual checks it replaced.
Sources
[1] HSE, “Legionnaires’ disease. The control of legionella bacteria in water systems — Approved Code of Practice and guidance (L8)”. https://www.hse.gov.uk/pubns/books/l8.htm [2] HSE, “Legionnaires’ disease: Technical guidance (HSG274)”. https://www.hse.gov.uk/pubns/books/hsg274.htm [3] HSE, “Hot and cold water systems”. https://www.hse.gov.uk/legionnaires/hot-and-cold.htm [4] HSE, “Testing and monitoring your water system for legionella”. https://www.hse.gov.uk/legionnaires/testing-monitoring-water-system.htm