On most sites, a low-use tap has an easy answer: send someone to run it for a few minutes each week. In a custodial setting that one sentence hides a security escort, a regime that might be in lockdown, and a cell you cannot enter on demand. That gap, between the control the rules expect and the access the building allows, is the real subject here.
The controls themselves are not exotic. Keep hot water hot, cold water cold, keep it moving, keep tanks and fittings clean, and hold records that prove it. UK practice rests on the Approved Code of Practice (L8) and the technical guidance in HSG274, and the duty holder’s responsibilities do not change because the site has a perimeter fence [1][2][7]. What changes is delivery: how you actually reach the water, and how you evidence control across hundreds of intermittently occupied outlets you are not free to walk up to.
So the decision worth getting right is not what to control. It is how to keep water moving and prove it where access is governed by the regime, not by your maintenance calendar.
What makes a prison water system its own problem
Four features push prisons and detention centres out of the standard playbook.
Access is controlled, not assumed. Reaching an in-cell basin can mean an officer escort, an unlock, and a slot in a regime that may be suspended at short notice. A flush that takes thirty seconds of water can take an hour of someone’s day. When staffing is tight, that is the first task to slip, and it slips exactly where the risk concentrates.
Occupancy swings hard. Cells empty between prisoners, wings get mothballed and recommissioned, segregation and healthcare units run to their own rhythm, and immigration removal centres can turn population over quickly. Each void cell is a small reservoir of warm, still water. The link between stagnation and risk is the recurring theme of neglected water systems, and a houseblock standing empty before its first intake is that problem at scale.
The fittings work against you. Anti-ligature and anti-vandal design favours push-button timed taps, integrated basin-and-WC units, and spray-pattern outlets. Spray taps and showers are among the fittings HSE flags as creating foreseeable risk, because they turn water into the fine aerosol that carries the bacteria into the lungs [4][6]. Many also run at low flow, so they shift very little water even in use, which is poor for turnover and awkward to clean.
The people are more susceptible, and the water is deliberately tepid. Severe Legionnaires’ disease is more likely in older people, smokers, and those with weakened immune systems or existing lung conditions [5], a profile well represented in custody. At the same time, scald prevention pushes hot water to be blended down at the outlet, which can leave fittings sitting in the warm band where Legionella multiplies. Balancing scald risk against Legionella risk is a genuine design tension, not a box-tick, and the figures belong in your risk assessment rather than a rule of thumb [3].
Comparing the ways to keep restricted outlets safe
There is no single right control for a low-use, hard-to-reach outlet. There are four broad approaches, and most estates end up blending them. The honest comparison is by what each actually reduces, what evidence it leaves behind, and what it costs in officer time and security disruption.
| Approach | Best suited to | Evidence it leaves | The catch behind the wing door |
|---|---|---|---|
| Manual escorted flushing | Stable, supervised areas with reliable access: staff facilities, workshops, routinely worked healthcare wings | Task sheets and temperature readings, only as good as the person doing them | Every flush costs an escort and an unlock; first task dropped when staff are short; void cells and seg get missed |
| Automated flushing devices | Outlets that are genuinely inaccessible or idle for long spells: mothballed wings, voids, far ends of runs | Device logs of flush events, stronger when tied to monitoring | A timer firing into a dead leg proves activity, not turnover; needs install, power and maintenance; can mask a design fault |
| Outlet and pipework rationalisation | Old estates with decades of alterations, redundant outlets and dead legs | As-built drawings, fewer assets, a smaller surface to manage forever | A capital job inside a secure perimeter; needs survey against schematics; biggest long-term payoff, slowest to deliver |
| Remote temperature/flow monitoring | Large estates needing assurance across many points without repeated physical entry | Continuous data and exception alerts, a strong audit trail | Sensors need calibration and someone acting on alerts; in-cell installs face anti-ligature constraints; data alone changes nothing |
Which to reach for, and when
Design out the problem first where you can. Pulling a redundant pipe or capping a dead leg once beats flushing it for the next decade, and it permanently lowers the access burden that defeats every other method. Where outlets must stay but access is the binding constraint, automated flushing earns its place, provided you verify it moves real water to temperature and is not just clicking a valve. Keep manual flushing for the areas you can genuinely reach on schedule, and use monitoring as the layer that tells you, without an unlock, whether any of it is actually working.
The trap to avoid is funding only the visible part of the programme and treating flushing as free labour. It isn’t, and under-resourcing it is one of the quieter ways control drifts, the pattern behind a lot of cost-driven failures.
A caveat specific to custody
None of the above sets your numbers. Temperatures, flushing intervals, sampling and remedial actions depend on your system, your population and your control scheme, decided through a competent, site-specific risk assessment, and in a custodial estate the security regime is part of that assessment rather than an excuse to leave gaps in it. Where a duty is shared between an operator and a facilities provider, write down who owns each outlet and each record before an incident forces the question. Sampling can support verification or investigation, but its frequency follows the risk assessment, and a clean result from one accessible outlet says nothing about the cell you could not open that week [2][8].
Where to start this week
Take your asset list and mark every outlet by how easily you can actually reach it under the current regime, then overlay which cells and wings are standing empty. The outlets that are both hard to access and rarely used are your priority. That short list, not the full schedule, is where to decide between flushing, automating, or designing the outlet out. Keeping those decisions and their evidence in one place, rather than scattered task sheets, makes the next audit far shorter; integrating records with your facilities-management system is the practical way to get there.
FAQ
Who is the duty holder in a prison run under a contract?
The organisation in control of the premises holds the duty, and in a contracted or private estate that responsibility, including oversight and records, has to be pinned to a named role rather than assumed to sit “with the contractor”. A provider carries out tasks under the duty; it does not absorb accountability for them [7].
How do you flush cells that are occupied or in lockdown?
You plan around the regime rather than against it: build flushing of accessible outlets into routine unlocks, schedule void and segregation cells with the wing, and use automated flushing or monitoring where physical access is unreliable. The risk assessment should record how access constraints are handled, not pretend they do not exist.
Do anti-ligature spray taps increase Legionella risk?
They can, on two counts: they produce aerosol, and they often run at low flow so they move little water. That does not rule them out, because the safety of the occupant comes first, but it does mean they need deliberate turnover, cleaning and descaling, and they belong on the higher-attention end of your outlet list [4][6].
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, “Systems most likely to create legionella risk”. https://www.hse.gov.uk/legionnaires/risk-systems.htm [5] NHS, “Legionnaires’ disease”. https://www.nhs.uk/conditions/legionnaires-disease/ [6] CDC, “How Legionella Spreads”. https://www.cdc.gov/legionella/causes/index.html [7] HSE, “Legionnaires’ disease - what you must do”. https://www.hse.gov.uk/legionnaires/what-you-must-do/index.htm [8] HSE, “Testing and monitoring your water system for legionella”. https://www.hse.gov.uk/legionnaires/testing-monitoring-water-system.htm