A Legionella logbook earns its keep on the worst day, not the routine one. The day an inspector turns up unannounced. The day you hand the site to a new responsible person who has never seen the plant room. The day a case of Legionnaires’ disease is being traced back through every building in the area and yours is on the list. In those three moments your records either speak for you or sink you, and plenty of logs that look fine on a calm Tuesday fall apart under exactly that pressure.

The fix is rarely “record more”. It is “organise what you record so a stranger can follow it”. An auditor, a colleague inheriting your role, or an investigator building a timeline should be able to open the log cold and reconstruct four things: what was meant to happen, what actually happened, what slipped, and what you did about the slip. If your records answer those four questions without you in the room narrating, they are organised. If they only make sense with your commentary, they aren’t.

L8 puts record-keeping at the centre of competent control: duty holders are expected to keep records of the risk assessment, the written scheme, the monitoring carried out, and the management arrangements behind it [1]. That is the floor, not the finish line. The finish line is a trail anyone can audit.

What an auditor should find in two minutes

Think of the log in four layers. The routine temperature readings are only one of them, and on their own they prove very little. Group your records so each layer is obvious, complete, and quick to navigate.

The foundation — the “why” behind every task

  • Hold the current risk assessment so it is the first thing anyone finds, clearly dated and clearly the live version.
  • Keep the written scheme of control beside it, naming each control measure, the acceptable result, and who owns it.
  • Maintain an asset register and a simple schematic so every reading maps to a real, locatable outlet or plant item.
  • Record who the responsible person is, who deputises, and the date each appointment was made.

The routine evidence — the “what happened”

  • Log temperature checks at sentinel and rotational outlets with the date, the reading, and whether it met target.
  • Record flushing of low-use outlets by naming the outlet and the person, not just “flushing done”.
  • Capture tank, calorifier and TMV inspections and servicing with the actual findings, not a bare tick.
  • File cleaning and descaling of shower heads and spray fittings, plus any sampling reports, against the specific asset.

The exception trail — the part audits actually test

  • Flag every out-of-range result and every missed task at the point it happened.
  • Record the remedial action, who carried it out, when, and the re-check that confirmed it was resolved.
  • Note where a task could not be done and why (room occupied, outlet isolated), so a gap reads as a decision rather than a failure to look.

The continuity layer — surviving change

  • Keep superseded versions of the risk assessment and the scheme, so you can show what was known and when.
  • Record review dates and what triggered each review.
  • Log training and competence for everyone who touches the system.
  • Keep a handover record so the next responsible person inherits the trail, not a mystery.

How to organise it so the trail holds

Good Legionella log organisation comes down to one habit more than any tool: record the decision, not just the task. “Outlet flushed” is a task. “Outlet G-12 flushed weekly because the room sits void between lets; two missed flushes escalate to the responsible person; a third triggers a use-pattern review” is a control. The second version tells an auditor you are managing the risk, not feeding a sheet.

Make assets unambiguous. A reading of 48°C logged against “shower 3” is close to useless if nobody can find shower 3 next year. Tag outlets with consistent IDs that match the schematic, and use the same ID everywhere — the temperature log, the flushing record, the sample report. This is where a digital logbook tends to pull ahead of paper: not because the screen is nicer, but because it forces every reading to attach to a named asset, a timestamp and a user, and it makes the exception trail searchable instead of buried in a ring binder. If you are weighing that switch, Is a digital logbook worth it? Cost-benefit analysis works through the cost-benefit honestly.

Then keep the structure still. Re-organising the log every time someone new takes over quietly destroys the timeline that an investigator most wants to read. Agree a layout once — foundation, routine, exceptions, continuity — and hold it across the years and across contractor changes.

The bits people skip (and inspectors look for)

Most logs fail an audit on the same handful of omissions.

The first is the missing close-out. A team records the out-of-range reading and stops there. An open exception with nothing after it is worse than no record at all, because it proves you saw the problem and shows no response. Close every loop with the re-check that brought it back into range.

The second is the blank that looks like neglect. An empty week reads as “nobody checked”. The same blank, annotated “outlet isolated for refurbishment, removed from schedule, see RA review 14/03”, reads as control.

The third is version amnesia. People overwrite the risk assessment and bin the old one. After an incident the live document shows what you think now; the superseded versions show what you knew then — and that is usually the question that decides whether you were reasonable.

The fourth is the orphaned record. A sample report or a contractor service sheet that lives in someone’s inbox or a desk drawer, never filed against the asset it describes. If it is not in the trail, then in an audit it did not happen.

A fair caveat

This is general guidance on how to structure records, not a ruling on what your particular system must contain. What you monitor, how often, and what counts as an acceptable result are set by a competent, site-specific risk assessment and your written scheme — and they shift as the building and its use change. Treat retention periods, sampling frequency and remedial limits as things to confirm against current HSE guidance and your own risk assessment, never as fixed numbers lifted from an article. On retention specifically, HSG274 sets out how long monitoring records should be kept; confirm the period that applies to each record type before you cull anything [2]. And where you do sample, frequency follows the risk assessment rather than a calendar [3].

Where to start this week

Don’t audit your own log — you already know where everything is, which is exactly the blind spot. Hand it to someone who was not involved and give them one task: find the most recent out-of-range result and prove, from the record alone, that it was resolved. If they manage it inside five minutes, your trail holds. If they can’t, you have just found the first thing to fix, and probably the layer — foundation, routine, exception or continuity — that needs the work.

FAQ

How long do we need to keep Legionella monitoring records?

Long enough to show a continuous history of control, which in practice means years rather than months. HSG274 sets retention expectations for monitoring records [2]; confirm the exact period for each record type — routine logs, inspections, sampling — and don’t dispose of anything that still forms part of your live evidence trail.

Is a digital logbook required, or is paper still acceptable?

Neither is mandated. What matters is that records are accurate, complete, attributable to a named person, and retrievable on demand. Paper can clear that bar on a small, stable site. A digital logbook usually wins on larger or multi-site estates because it ties each reading to a named asset, a time and a user, and surfaces exceptions instead of letting them hide.

What is the difference between a complete log and an audit-ready one?

A complete log has all the readings. An audit-ready log lets a stranger reconstruct the story without you — what was planned, what happened, what slipped, and how each slip was closed. The gap between the two is almost always the exception trail and consistent asset naming.

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