Reheat after a dehumidifying coil costs about 0.34 kW per 1,000 m³/h per kelvin, or 34 kW for 5 K on a 20,000 m³/h unit. Per kWh delivered, resistance and boilers cost most, heat pumps less, condenser recovery and wrap-around heat pipes least; tariffs and the site decide what fits.
Light office wear is about 0.5 clo, while full cleanroom gowning sits nearer 1.0 clo, and standing line work is about 1.6 met or more. Gowned operators are comfortable cooler than office staff, so set the room temperature in the URS with QA and EHS before the load calculation sizes the coil.
A full-size 592 × 592 mm bag filter is commonly rated at about 3,400 m³/h. A 20,000 m³/h AHU needs six frames; a casing with room for four runs each at 47% over, with higher pressure drop, possibly lower efficiency and fluttering pockets. The frame count belongs in the specification and the FAT.
Ten per cent on the load, on the airflow and on the coil multiplies to 1.33, because each margin sits on the one before. The oversized coil runs at part load most of the year, condenses less water, and RH climbs while temperature holds. One named margin and a part-load check avoid it.
An extra 100 Pa of filter resistance on a 10,000 m³/h supply costs about 460 W, roughly 4,000 kWh a year. The cheapest change point comes when the latest month's extra energy cost exceeds the average monthly cost since fitting, filter set included, provided it sits below the manufacturer's maximum final pressure drop.
A 0–10 V transmitter where the IO list says 4–20 mA passes a visual FAT and fails on site. That and four more defects (missing tags, gasket gaps, drain pans that hold water, missing documents) are worth checking at every AHU FAT against the approved specification and IO list.
A test at minimum fan speed with a door held open is OQ evidence; PQ shows normal operation, with production or a representative simulation, over a stated period. A step that forces a condition usually belongs in OQ, and separate protocols allow OQ to close before PQ starts.
On the suction side of the fan, any casing leak between the final filter and the fan draws unfiltered plant-room air into the supply. On the pressure side leaks blow outward and the filter catches what the fan sheds. The cost is a plenum, and the filter bypass class still needs specifying in either layout.
A soft starter cuts an AHU fan's starting current to about 3–4 times full load but still runs it at one speed. With a VFD a cleanroom supply fan can hold design airflow as filters load, at the cost of harmonics that need a DC choke and a study against IEEE 519-2022.
An AHU condensate line piped directly into a floor drain gives backed-up water and sewer gas a route to the unit. Each AHU needs its own trap, then a visible air break into a tundish or trapped gully, sized to the local plumbing code. The gap also shows anyone passing whether condensate is running.
A pressure gauge shows filter loading but not a collapsed bag, a slipped gasket or a wet drain pan. Viewing windows and internal lights in the filter, coil, humidifier and fan sections let technicians check with the unit running. Double-glaze windows on cold sections, and keep the doors for checks that need a stop.
Margins compound: four 10% margins give about 46%. A diversity factor works the other way, removing load that never occurs at once. Take it from the process schedule, apply it to shared plant such as the chiller or main switchboard, never to a single room's AHU, and record it in the design basis.