Anyone running thirty buildings already does a rough version of this in their head: which sites feel neglected, which have old galvanised pipework, which never seem to flush properly. A pipe health index just asks whether that instinct could be written down as a number, and whether doing so would help you look in the right place first.
It is a tempting idea and a dangerous one in equal measure. Tempting because triage across a large estate is genuinely hard. Dangerous because a single tidy score invites people to treat it as a verdict, and there is no condition number on earth that tells you whether Legionella is present. So let me lay out what such an index could be, how you might build one, and the lines it must never cross.
Why a single score is even worth discussing
The prevailing approach to prioritisation on a big estate is the calendar. Every asset gets the same inspection interval regardless of its actual state, and remedial budgets go to whoever shouts loudest. That is defensible and auditable, but it is blunt. A two-year-old plant room on a well-used building gets the same scrutiny as a half-occupied 1970s block with a corroding tank, even though the second is plainly working harder against you.
A competent risk assessment already weighs the real factors, temperature behaviour, stagnation, materials, deposits, for each system, and that assessment, not a score, is what the duty holder must act on under ACoP L8 [1]. The honest question is narrower: could the factors an assessor already considers be combined into one comparative figure that helps decide where to send the assessor first? Not to replace judgement. To point it.
The framework: a four-signal condition view
If I were sketching a pipe health index, and I’d hold it loosely, I would build it from four groups of signals, each of which is something a person could already record on a survey. Call it the PHI for shorthand, but treat the label as a filing system, not a measuring instrument.
1. Thermal behaviour. How reliably does the system hold cold water cold and hot water hot at the sentinel and representative outlets over time? Persistent drift, slow heat-up, or cold outlets sitting warm are the signals here. Temperature is the backbone of control [2], so this group carries the most weight.
2. Turnover and water age. How long does water sit between fresh supply and use? Low-occupancy wings, oversized storage, dead legs and missed flushing all push water age up, and high water age is the condition that lets temperature drift, biofilm and disinfectant decay compound. It is a risk indicator, not a measurement of anything living, but it is one of the most useful things you can track proactively rather than discovering through a failed flush.
3. Physical condition. Pipe material and age, internal scaling, sediment, corrosion and known biofilm-prone fittings. Legacy galvanised steel, heavy limescale and gritty deposits all raise the surface area and shelter available to organisms, which is why physical state belongs in the picture even though it tells you nothing about contamination directly.
4. Maintenance reality. Not the maintenance schedule, the maintenance history. Overdue tank inspections, recurring out-of-range temperatures, repeated remedial actions that never quite close, gaps in the logbook. A system that keeps generating exceptions is telling you something.
You combine these into a comparative score, weighted, because thermal behaviour and water age should outrank a cosmetic scaling note. The output is not “safe” or “unsafe”. It is a queue: which assets warrant a closer look, sooner.
How you’d actually use it: a triage decision tree
The index earns its keep only as the front end of a decision, never as the decision. Here is the logic I would apply to each asset’s score, expressed as branches you could follow on a spreadsheet review.
- Is any single thermal or stagnation signal a hard failure (e.g. a sentinel outlet persistently out of range, a known dead leg in use)?
- Yes → That asset goes to the front regardless of its overall score. A composite figure must never average away a serious individual failure, escalate to the risk assessor now.
- No → Continue.
- Is the composite score high (poor condition) relative to the estate?
- Yes → Bring forward inspection and a competent review; treat the score as a prompt to verify on site, not a finding.
- No, the score is mid-range → Keep the scheduled interval, but flag the weakest single signal for attention at the next visit.
- No, the score is low (good condition) → Maintain routine control and monitoring. Do not stretch inspection intervals or skip sampling on the strength of a good score, a low number is not a clean bill of health.
- At every branch: does the data behind the score look stale or thin (old survey, missing logs)?
- Yes → Treat the score as unreliable and default to inspection, not to confidence.
The shape that matters is the asymmetry. A bad score can pull an asset forward; a good score can never push one back past the controls the law requires. Prioritisation is allowed to add scrutiny. It is never allowed to remove it.
The honest limits
This is where most “scoring tool” articles go quiet, so let me be blunt about the failure modes.
It detects nothing. No combination of temperature trends, water age, scaling and maintenance gaps detects Legionella, measures it, or proves a system safe. The index describes the conditions that tend to favour growth, habitat and opportunity, not the organism. Confirming presence still needs sampling, and sampling itself is a snapshot that sits alongside, not above, temperature, flow and cleanliness control [2]. Anyone who reads a green score as “this building is fine” has misused it.
Weighting is a judgement, dressed as arithmetic. Deciding that thermal behaviour is worth more than scaling is a defensible opinion, but it is an opinion. Two competent people would build two different indices. The number’s precision is false comfort; the inputs are what deserve scrutiny.
Garbage in, false confidence out. A score is only as honest as the survey and logbook behind it. A building with poor record-keeping can score better than a well-documented one simply because its problems were never written down. That is the most dangerous result of all, a tidy figure that launders an information gap into apparent safety.
A score that overrides a competent person’s site-specific risk assessment has stopped being a triage aid and become a liability. Used the other way, as one structured input that the assessor interrogates and can overrule. It is a reasonable way to reason across many assets at once [3]. The deciding factor is always whether the index serves the assessment or pretends to be it.
FAQ
Could a pipe health index ever replace Legionella sampling?
No. A condition score and a water sample answer different questions. The index estimates whether a system’s conditions favour growth, drawn from temperature, water age, materials and maintenance history; a sample tests for organisms at one point and time. Neither proves a system safe on its own, and the score has no microbiological content at all. Sampling and statutory temperature, stagnation and cleanliness control stay exactly where they are.
If the weightings are subjective, is the score worth anything?
It can be, provided you treat the inputs as the real product and the number as a sorting key. The value is in forcing a consistent question across every asset, how does each one behave on temperature, turnover, physical condition and maintenance, rather than in the false precision of the total. Publish your weighting logic so an assessor can challenge it, and never present the figure as objective fact.
Is this only useful for large estates?
Largely, yes. A single small building rarely needs a composite score, because a competent person can hold its whole picture in mind and inspect it directly. The concept earns its place when you are triaging dozens of assets and cannot give each the same immediate attention, it helps decide the order of the queue, not the standard of care any one site receives.
Sources
- HSE, ACoP L8 (2013), “Carrying out a risk assessment”, p.12. https://www.hse.gov.uk/pubns/books/l8.htm
- HSE, HSG274 Part 2 (2024), “Operation and inspection of hot and cold water systems”, p.70. https://www.hse.gov.uk/pubns/books/hsg274.htm
- BSI, BS 8580-1:2019, clause 8.2 (risk rating systems). https://www.bsigroup.com/