Any gap in an AHU casing between the final filter and a downstream fan draws plant-room air straight into the supply.
Pressure across the casing sets the direction of a leak. Upstream of the fan the unit runs below plant-room pressure, so a gap at a panel joint, a door gasket or a cable gland draws air in. If that gap sits between the final filter and the fan, the air it admits has never seen a filter. Downstream of the fan the casing is above plant-room pressure, and the same gap leaks conditioned air out to the plant room, where the loss shows up in fan power and cooling load.
The fan is the second argument. Motor, bearings and, on older units, a belt all shed particles, and with the filter upstream whatever the fan sheds goes straight to the ductwork.
The pressure side has its own costs. Fan discharge is uneven, so the filter needs a plenum or a diffuser plate in front of it for an even face velocity, and that adds length. Bypass around the filter is a separate matter: leakage across the frame seal is driven by the filter's own pressure drop, whichever side of the fan it sits on, so the EN 1886 filter bypass class matters in both layouts.
For a GMP supply unit we would still choose the pressure side, and write the casing leakage class (L1 or L2) and the filter bypass class into the specification instead of leaving them to the catalogue.
Which way do your supply units run: draw-through or blow-through?
Topics



