Most AHU failures appear in BMS data before they become alarms. Trend filter ΔP, fan speed or power at constant airflow, bearing vibration and temperature, and coil valve position. In a GMP plant, catching a failure as a trend turns it into planned work under change control.
The typical differential between adjacent grades is 10–15 Pa, and alarms catch only the failure. A room pair drifting from 14 to 11 Pa has not alarmed, yet something changed. Plot the daily median of each differential against its commissioned value and review it weekly.
When a chilled water plant designed for 5 K runs at 3 K, it needs about 67% more flow for the same load and stages chillers early. The usual causes are at the coils: three-way valves, fouling and low setpoints. Fix the coils and trend ΔT on the BMS.
Monthly route-based vibration surveys catch slow fan faults such as imbalance, misalignment and bearing wear at a low cost per point. Permanent online sensors catch faults that develop between visits. Put sensors on fans whose stop takes a Grade A or B room out of spec with no standby, and route the rest.
A maintenance contractor's remote view should be read-only, with setpoints changed only under the site's change control. Its alerts watch the plant behind the room, such as fan speed, filter pressure drop and chilled water temperature, and the escalation path to site is written before the contract starts.
A pre-filter changed at 150 Pa on a 0–1,000 Pa gauge sits in the bottom 15% of the dial with about 20 Pa of uncertainty. A 0–250 Pa gauge puts it at 60% with 5 Pa. Range each stage's gauge and transmitter to its own change point.
The P-F interval is the time between the point where a failure first becomes detectable and the point of functional failure. An inspection has to come round more often than that, with time left to act. The interval depends on the failure mode and on the detection method, so one calendar frequency cannot suit every fault.
A unit that ran for years can fail days after its annual strip-down. Intrusive tasks can leave faults behind, and many failure modes are not related to age, so a fixed overhaul cannot prevent them. Keep wear-out parts on the calendar and open everything else on condition evidence.
For decades, IT (Information Technology) and OT (Operational Technology) existed as two parallel worlds within an enterprise — IT managed data, emails, office software, and network security; OT managed machinery, production lines,…