Containment does one job: it keeps hot exhaust air away from cold supply air. That sounds almost too simple to need a name, but without it a cooling system spends a surprising share of its output cooling air it has already cooled once, and racks at the top of a row run hot for no good reason.
What Goes Wrong Without It
Two things, and they are opposites. Recirculation is hot exhaust finding its way back round to a rack intake — over the top of a row, around the end of it, or straight through an empty rack slot. The rack then draws in air that is already warm, so its inlet temperature climbs regardless of how cold the supply air is. This is why the top third of an uncontained row tends to run hottest.
Bypass is the reverse: cold supply air that returns to the cooling unit without passing through any equipment at all. It has done no useful work, but it has been chilled and moved, so it has cost real energy. It also arrives back at the cooling unit cool, which makes the unit behave as though the room needs less cooling than it does.
Choosing Between Them
Cold aisle containment encloses the cold aisle: racks face each other across it, a roof goes over the top and doors close each end. Everything outside that enclosure is at exhaust temperature. Hot aisle containment does the opposite — it encloses the hot aisle between two back-to-back rows and usually ducts it into the ceiling void, so the room itself sits at supply temperature.
In a new build, hot aisle containment is usually the better answer. The room stays comfortable for anyone working in it, and the air returning to the cooling units is hotter, which makes those units more efficient and widens the window in which free cooling is viable. In a retrofit the decision is often made for you: cold aisle containment needs no ceiling ductwork, so it is cheaper and far less disruptive to add to a room that is already running.
Both work. A well-sealed cold aisle beats a leaky hot aisle every time, which is why the sealing detail matters more than the choice between the two.
The Details That Undo It
Containment is only as good as its seal, and the leaks are rarely dramatic. Blanking panels are the usual culprit: any unfilled U in a rack is a direct path from the hot side to the cold side, and a row with a dozen empty slots has a substantial hole in it. Unsealed cable cutouts in a raised floor do the same job in the other direction, letting supply air escape where no equipment will ever use it.
Then there are the edges — the gap between the top of a rack and the containment roof, doors that no longer close properly, and the space left when a rack is removed and nothing is fitted in its place. None of these will show up as a fault anywhere. They just quietly move the operating point until someone notices a warm rack.
What to Measure
- Differential pressure across the containment barrier — the single best indicator that the enclosure is still doing its job
- Rack inlet temperature at the top, middle and bottom of each rack, not one sensor per row
- The difference between supply and return temperature, which tells you how much air is bypassing equipment
- Door state on end-of-row doors, because a propped-open door is a common and invisible cause of drift
Pressure is the measurement people most often skip, and it is the one that answers the question directly. Temperature tells you that something has gone wrong; differential pressure tends to tell you before it does, because a containment seal degrades gradually while inlet temperatures stay within tolerance right up until they do not.
Common Questions
What is the difference between hot aisle and cold aisle containment?
Both stop hot exhaust air mixing with cold supply air; they differ in which aisle you enclose. Cold aisle containment encloses the cold aisle, so the room itself sits at hot-exhaust temperature. Hot aisle containment encloses the hot aisle and usually ducts it to the ceiling return, leaving the room at cool supply temperature.
Which is better, hot aisle or cold aisle containment?
Hot aisle containment is generally preferred in new builds because the room stays comfortable to work in and the return air reaching the cooling units is hotter, which improves their efficiency. Cold aisle containment is often cheaper and less disruptive to retrofit, since it does not require ceiling ductwork. The right answer usually depends on the building you already have rather than on the theory.
What should you monitor in a contained aisle?
Differential pressure across the containment barrier, rack inlet temperatures at the top, middle and bottom of each rack, and the temperature difference between supply and return. Pressure tells you whether the containment is doing its job; inlet temperatures tell you whether any rack is being starved of air.
Containment is one of the cheapest efficiency measures available in an existing room, and one of the easiest to let slip. The barrier goes in once; the blanking panels, the floor seals and the doors need someone to keep an eye on them — which is really an argument for measuring the aisle rather than trusting it.