You can have a compliance dashboard that is entirely green and still lose sleep over it. With one building, you walk the plant room and you know. With thirty buildings, you are managing through other people’s paperwork — and the site that hurts you is almost never the one filing the tidiest returns.
So across a portfolio the useful question is rarely “is this reading bad?” It is “why did control drift here and not at the eleven sites running the same scheme, and where else is the same gap hiding?” Treat the estate as one system with local faults, and multi-site Legionella turns from a filing exercise into fault-finding.
Why a portfolio fails differently
The control measures do not change with scale. Keep hot water hot, cold water cold, water moving, tanks and fittings clean, and prove it with records — that is true for one site or fifty, and HSG274 sets out the technical detail behind each [2]. What changes is the failure mode. A single building fails because of its pipes. A portfolio fails because of its variance: one site with a vacant wing nobody flushes, one with a contractor who quietly skips the awkward roof tank, one newly acquired site with no schematic and a calorifier nobody has assessed.
That variance is the thing you are actually managing. The systems most likely to cause harm are the same everywhere [4]; your job is to find the sites where they are managed worst, before someone else finds them for you.
From central red flag to root cause
When a symptom reaches you at portfolio level, resist the urge to log it and move on. Trace it backwards from what you can see centrally to what is actually happening on site. The table below maps the common symptoms to the cause they usually point at, the one check that confirms it, and the action that follows.
| Symptom you see centrally | Most likely cause | The check that confirms it | The action |
|---|---|---|---|
| One site fails the same temperature target month after month | Either a recurring missed task or a genuine system deficiency (long dead leg, weak return) — not “bad luck” | Pull that site’s task history; if the task was done and the result is still out, inspect the outlet physically | Done-but-failing means remedial design work; skipped-and-failing means a scheduling and ownership fix |
| Dashboard all green, but a site audit finds problems | Data entered to satisfy the form, not to record reality | Compare raw readings against the asset on site; look for implausibly identical values week to week | Re-task and retrain the recorder; verify independently before trusting that site’s returns again |
| Records have gaps, but nothing is flagged as failing | No competent person actually owns the site locally; tasks fall between contractor and facilities team | Check who is named as responsible person and whether they can explain why each control exists | Name a competent local owner and close the accountability gap before adding any more monitoring |
| A newly acquired or transferred site has no usable history | The inherited system was never assessed under your scheme | Look for a current, site-specific risk assessment and an asset register; if absent, assume neither exists | Commission a fresh assessment and rebuild the schematics before reporting the site as compliant |
| The same defect type appears at several sites | A portfolio-wide design or procurement pattern, not isolated faults | Cross-tabulate defects by type across every site, not site by site | Fix at policy level — specification, contractor scope, standard fittings — rather than raising fifty separate jobs |
The last row is where multi-site management earns its keep, and it leads straight to the most important call you make.
Local fault, or system fault?
The single most useful habit across an estate is asking, every time, whether a finding is local or systemic. One warm cold-water tank in one plant room is a defect: fix it, record it, move on. The same warm tank in five plant rooms is not five defects — it is one decision about where you site tanks and how you specify plant rooms, repeated. Treat it as a system fault and fix it once at the top.
This is also how you stop drowning. If you cannot tell a local fault from a system one, every site generates its own list and you firefight forever. Group the findings by type first, decide which are systemic, and you convert dozens of jobs into a handful of policy changes plus a short list of genuine one-offs. Comparing sites fairly needs a consistent scoring method, so a care home and a quiet office rank on the same scale rather than on whoever shouted loudest — see Legionella risk scoring for a way to do that.
A word on the records themselves: write down the decision, not just the task. “This outlet is flushed weekly because use is intermittent; a missed flush escalates to the responsible person; repeated misses trigger a use-pattern review” tells the next person why the control exists. A bare tick does not. Across a portfolio, where the person reading the log is rarely the person who set it up, that difference is what makes a written scheme actually usable [2].
When to stop firefighting and escalate
Some findings should jump straight past the site team to a portfolio-level review. Escalate when a defect type recurs across multiple sites, when you stop trusting a contractor’s returns, or when a single site keeps failing despite the tasks being done — that last one usually means the system needs redesign, not more flushing. A suspected or confirmed case of illness linked to any site is no longer a site problem at all; it triggers the duty holder’s responsibility to act and to bring in competent help across the estate, and HSE sets out what that involves [3].
Outsourcing the legwork to a national contractor does not move that responsibility. Choosing a provider who signs up to a recognised standard, such as the Legionella Control Association’s code of conduct, is sensible — but oversight, decisions and the records that prove control stay with you [5].
A note on the limits of this
None of the above is a control scheme you can drop onto your estate. Every site sets its own temperatures, monitoring intervals and remedial triggers through its own risk assessment, carried out by someone competent — the value of managing a portfolio is spotting patterns across those assessments, not overriding them with a single house rule. Where a specific figure matters, take it from the site’s scheme and current HSE guidance, not a template. Risk assessment and the review of controls carry Approved Code of Practice status under L8, which is exactly why a portfolio summary sits on top of the individual assessments rather than replacing them [1].
FAQ
Can one risk assessment cover several sites at once?
No. Each site needs its own site-specific assessment, because the systems, occupancy and exposure differ building to building. What you can — and should — standardise is the methodology, so every assessment is built and scored the same way and the results compare. A consistent code of practice for risk assessment helps keep that methodology aligned across the estate [6].
If we use one contractor everywhere, does that mean we are covered?
Not on its own. A competent service provider doing the same work to the same standard across your sites is a real advantage, but accountability does not transfer with the work. The duty holder keeps oversight, has to be able to challenge the returns, and owns the records — picking a provider against a recognised code of conduct is part of that, not a substitute for it [3][5].
How often should each site be sampled for Legionella?
There is no fixed estate-wide interval to copy across. Sampling supports verification or investigation, and HSE is clear that its frequency follows the system and the risk assessment rather than a calendar set centrally [4]. A portfolio-wide “everyone tests quarterly” rule is a sign the schedule is being driven by procurement, not by risk.
Next step
Pull every site onto one sheet this week, with three columns each: who is the named responsible person, when was the risk assessment last reviewed, and what is the oldest open remedial action. The sites with a blank in any column are your starting point — not the ones with the worst readings, but the ones where you cannot even see whether the readings are true. Book a verification visit to the three you trust least.
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, “Legionnaires’ disease - what you must do”. https://www.hse.gov.uk/legionnaires/what-you-must-do/index.htm [4] HSE, “Testing and monitoring your water system for legionella”. https://www.hse.gov.uk/legionnaires/testing-monitoring-water-system.htm [5] Legionella Control Association, “Code of Conduct for Service Providers”. https://www.legionellacontrol.org.uk/ [6] BSI, “BS 8580-1:2019 - Risk assessments for Legionella control. Code of practice”. https://knowledge.bsigroup.com/products/water-quality-risk-assessments-for-legionella-control-code-of-practice-1