What closed system sampling looks for
The suite of tests is chosen around the two failure modes that matter: corrosion and microbiology. Corrosion is checked through pH, conductivity, inhibitor concentration where an inhibitor is used, and the metals that indicate it — iron for mild steel circuits, copper for copper and brass components. Bacterial contamination is checked with total viable counts and, where specified, sulphate-reducing and nitrite-reducing bacteria, which are associated with microbially influenced corrosion.
Where glycol is used for freeze protection, its concentration and condition are tested too, because glycol degrades over time and its breakdown products make the water more corrosive.
Why a closed system should not corrode — and does
In theory a sealed system with the correct inhibitor should last decades. In practice systems lose water through leaks and automatic top-up, air enters through failed seals and pump glands, and inhibitor is consumed. Iron oxide sludge then settles in low points, coats heat exchanger surfaces and reduces heat transfer, while bacteria find a home in the deposits.
The consequences show up as cold radiators, noisy boilers, blocked plate heat exchangers and eventually plant failure. By the time the symptoms are obvious, the system usually needs cleaning rather than just dosing.
What to do with the results
If inhibitor is simply depleted and the water is otherwise clean, corrective dosing is usually enough. If corrosion metal concentrations are significantly elevated or sediment is present, dosing alone will not fix it — a dynamic flush is needed to remove the deposits, followed by re-dosing and verification sampling. If bacterial counts are high, cleaning and disinfection of the loop is appropriate, with attention to filters and strainers where biofilms establish.
Result reports should give a target range for each parameter, the current value, and a pass or fail statement. A page of numbers without a target range is not useful to a building services engineer.
How often to sample
Good practice is sampling after pre-commission cleaning at handover, then at least annually, and more frequently in the first year of a new system or where results have been unsatisfactory. Trend is more informative than a single result: two consecutive reports showing rising iron is a warning even if both are within range.
This article is general guidance, not a site-specific risk assessment. Control measures for your building should follow your written scheme of control and the Approved Code of Practice L8 with HSG274.