--- title: "Legionella control in smart buildings" source_url: https://legionella.io/articles/legionella-control-in-smart-buildings/ canonical_url: https://legionella.io/articles/legionella-control-in-smart-buildings/ pillar: "Technology & Remote Monitoring" summary: "Smart sensors and BMS dashboards can hide a stagnant wing. How UK facilities teams decide where monitoring really cuts Legionella risk, not just looks busy." primary_keyword: "smart building Legionella" date_published: 2026-01-07 date_reviewed: 2026-06-26 author: "Legionella.io editorial team (REMOTE TECH LTD)" reviewed_against: "HSE L8 and HSG274 guidance" region: "United Kingdom" license: "(c) REMOTE TECH LTD. Quote freely with attribution and a link to source_url." --- # Legionella control in smart buildings A smart building can dashboard its way into a false sense of safety. Temperature sensors glow green, the building management system logs every reading, and whoever holds the water-safety duty assumes the problem is handled. The technology is doing something. Whether it is doing the right thing is a different question. That gap — between data collected and control proven — is where smart buildings come unstuck. The decision worth getting right is not "should we fit monitoring?" but "which parts of this water system does monitoring actually make safer, and which parts does it just decorate with green ticks?" The clearest way to see the difference is to follow a building that got the answer half right. ## A sensor-rich office that still missed the obvious What follows is an illustrative composite, built from patterns that recur across smart-building retrofits rather than any single real site. Picture a city-centre commercial office, recently fitted with a smart water platform: networked temperature sensors on the calorifier flow and return, a probe in the cold water storage tank, automated logging into the building management system (BMS), and email alerts when a reading drifts out of range. On paper, model coverage. For months the plant-room readings were clean. Hot water left the calorifier hot and came back warm enough; cold storage held below the figure the risk assessment had set. The dashboard reported near-perfect compliance, and the quarterly review took twenty minutes because the numbers "spoke for themselves". What no sensor watched was the third floor. Half of it had been mothballed when a tenant moved out, and the original fit-out had put probes at the plant and on the risers — not at the outlets. So a row of en-suite shower rooms on the vacated wing sat with warm pipework, no use, and no flushing task, because the BMS reported the building as in range. The dashboard was green because it was measuring the places that were always fine. The gap surfaced through a person, not the platform. A newly appointed responsible person cross-checked the digital task list against the floor plan and asked why an entire wing had no flushing record and no outlet readings. The sensors had never lied. They had simply never been pointed at the risk. ## The decisions that turned it around Three calls separated a near-miss from a managed system. The first was to map sensor coverage against the risk assessment instead of the plant layout. Once the low-use outlets, dead legs and the mothballed wing were marked up, the blind spots were obvious — and cheaper to close with a scheduled manual flush-and-temperature task than with more hardware. Coverage should follow the assessment, not the cable run. Where the assessment itself is thin, the monitoring inherits its blind spots; [Risk assessment errors that lead to Legionella growth](https://legionella.io/articles/risk-assessment-errors-that-lead-to-legionella-growth/) covers how those errors creep in. The second was to give every alert an owner and an escalation path. An email landing in a shared inbox is not a control. The team assigned named responsibility for each alert type, agreed what an acceptable reading was, and defined what happens when one falls outside it — including who acts out of hours. The third was to record the reasoning, not just the task. "Cold tank probe, reviewed weekly, escalates to the responsible person on two consecutive out-of-range readings" tells an auditor — and the next post-holder — why the control exists and what good looks like. A bare log of numbers does not. ## What the technology was genuinely good for None of this is an argument against smart monitoring. In the same building, remote temperature monitoring caught a calorifier return drifting low weeks before a monthly manual check would have, and flagged a failing thermostatic mixing valve from a single odd reading. It cut clipboard time and made the trend visible instead of buried in a folder. Used well, the platform improved the speed and consistency of the evidence. The principle underneath is straightforward. Technology strengthens Legionella control when it improves evidence, speed and consistency — but it supports the duty holder, responsible person and competent person; it does not replace them [1]. L8 expects records of the precautions, the monitoring and the management arrangements [1], and HSG274 frames temperature monitoring as part of a managed scheme rather than a substitute for one [2]. A sensor earns its place when it is calibrated, sited at a point that matters, reviewed by someone competent, and wired into a response. A dashboard that only displays data — no owner, no threshold, no action — is theatre. ## Lessons worth carrying to your own building - Sensor placement is a risk-assessment decision, not an IT one. Cover the outlets and dead legs that carry risk, not only the plant that is easy to reach. - Every monitored point needs a threshold, an owner and an action. Without all three, an alert is just a notification. - Stagnation hides where there are no users and no sensors. Low-use and mothballed areas are exactly where a green dashboard is least trustworthy. - Sensor calibration is part of the control. An uncalibrated probe reporting "in range" is worse than no probe, because it manufactures false confidence. - Write the decision, not just the reading. The note explaining why a control exists is what turns routine logging into managed compliance — and what an audit actually tests. ## Before you trust the dashboard A green status is only as honest as the risk assessment behind where the sensors were placed, and the limits any platform flags as "in" or "out" of range come from your assessment, not the vendor's defaults. Sensors and dashboards are evidence tools; they do not make engineering or clinical judgements, and a clean screen is no substitute for a competent, site-specific review of the whole system. Treat any figure a supplier presents as a fixed safe value as a prompt to check it against current HSE guidance and your own assessment. A practical next step: put your sensor map and your Legionella risk assessment side by side this week. Mark every monitored point in one colour and every low-use or seldom-occupied outlet in another. Anywhere the second colour has no first-colour dot near it is a blind spot — and that, not the plant room, is where to look first. ## FAQ ### Does remote monitoring meet our Legionella monitoring duties on its own? No. It can carry out and record parts of the monitoring efficiently, but the duty stays with the duty holder and responsible person, and the scheme still needs competent review, defined responses and the wider controls — temperature, flushing, cleanliness — that a probe does not perform [1]. Sampling, where used, follows the system and the risk assessment rather than a sensor's schedule [3]. ### Where should Legionella sensors go first in a smart building? Wherever the risk assessment says risk concentrates — typically calorifier flow and return, cold storage, sentinel outlets, and any low-use or intermittently occupied areas — rather than wherever cabling is simplest. Plant-room-only coverage is the most common and most misleading gap. ### Our BMS shows everything in range — is that enough for an audit? Not by itself. An assessor will ask what the sensors cover, who reviews the readings, what thresholds apply, what happens on an exception, and whether probes are calibrated. A green dashboard with no ownership or escalation behind it is weak evidence, however tidy it looks [2]. ## Related reading - [Risk assessment errors that lead to Legionella growth](https://legionella.io/articles/risk-assessment-errors-that-lead-to-legionella-growth/) - [Self-disinfecting fixtures: smart showers and taps](https://legionella.io/articles/self-disinfecting-fixtures-smart-showers-and-taps/) - [Rapid Legionella detection: what instant testing can and cannot prove](https://legionella.io/articles/rapid-legionella-detection-what-instant-testing-can-and-cannot-prove/) - [Case study: a model Legionella control programme](https://legionella.io/articles/case-study-a-model-legionella-control-programme/) ## 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, "Testing and monitoring your water system for legionella". https://www.hse.gov.uk/legionnaires/testing-monitoring-water-system.htm