Aircon plenum box: the one part that gets made on site

A plenum box is a chamber wider than the pipes leaving it. Air arrives through one opening and leaves through several, at a lower speed than it had on the way in. Ducted systems carry them in two positions. One sits at the discharge of the indoor unit and divides that flow into branches. The other sits above a ceiling outlet, taking a single duct and spreading it across the face of the fitting.

None of it comes out of a carton. The Daikin FDXM installer reference guide states the position in one line. The ducting is to be field supplied. Its diagram then labels the flange, the insulation and the aluminium tape the same way, as field supply.

Mitsubishi Electric writes the job as a short list of instructions. Its PEAD ducted range asks for a canvas duct between the unit and the duct, incombustible material for duct parts, and full insulation on the inlet duct flange and outlet duct. Not one of those arrives finished. They get cut, fitted and wrapped by hand, above a ceiling.

So the box is the one component in the path whose quality is settled entirely on site. The machine was tested before it shipped. The diffuser was tested before it shipped. What joins them was folded and taped by whoever was working above that ceiling on the day.

Every throw, pressure and noise figure in a diffuser catalogue was taken under one inlet condition, and the catalogues say so. Titus states that its data is based on ASHRAE Standard 70, which requires several diameters of straight duct at the approach to the diffuser. The sentence after it carries the weight. Field installations seldom, if ever, allow these ideal conditions.

The measuring point is written down as well. Titus takes an outlet's total pressure as the sum of static and velocity pressures in the supply duct, one and a half duct diameters ahead of the outlet, measured at that position under the same standard. A number taken at a stated position only describes an install that reproduces the position.

Price Industries arrives at the same place from the design side. Its guidance for keeping an air distribution system quiet asks for smooth air flow at all duct elements, including branches, elbows, transitions and air outlets. It asks for straight ductwork between those elements, preferably five to ten duct diameters. A ceiling void in a flat has neither the depth nor the length for that.

Both makers then name the same remedy, which is a fair measure of how common the problem is. Price asks for equalizing grids when non ideal inlets cannot be avoided. Titus lists the same item in its own guidelines, as equalizing grids on direct diffuser connections. That is a bought part, fitted at the connection, and it costs both money and the ceiling depth a tight void does not have.

The penalties are published, and they are not small. Titus puts the effect of joining a diffuser straight onto a duct at as much as 12 NC above catalogue levels, caused by regions of localized high velocity inside the fitting. That gap separates a bedroom somebody sleeps in from one they can hear.

Flexible duct arriving at the inlet was tested as its own case. Titus reports that flexible ducts at a diffuser inlet increase pressure drop, increase sound levels and produce nonuniform air distribution from the unit. An ideal gentle 90 degree connection adds about 1 NC. A kinked one adds 7 to 9.

Air pattern moves with all of it, and that part carries no number. The same list records only that the air distribution pattern can be greatly affected, and that results were not the same for all diffuser types. A warm corner under a correctly selected diffuser can come from the shape of the metal above it.

The equalizing grid figures show what an approach is worth in decibels. Titus reproduces two cases from the ASHRAE Handbook. With an equalizing grid, sound levels match the maker's rating. Without one they run up to 12 dB higher. The second case compares eight duct diameters of straight run against two, and the short version measures 12 to 15 dB above the published rating.

The proportions of the box itself decide where the air points. Titus tabulates the discharge angle of an end-fed linear diffuser against the ratio of slot area to duct area. At 0.4 the airstream leaves at 15 degrees. At 0.75 it leaves at 25, at 1.0 at 35, and at 1.5 at 40. Nothing changed there but the relative size of two openings.

Getting a flat discharge back takes a part. Titus assures a 0 degree angle along the whole face by including a set of straightening blades behind the face vanes or bars. Leave those out and the face still looks finished, while the air leaves at an angle nobody chose.

The machine end has a geometry rule of its own. The PEAD guidance warns that noise from the intake will increase dramatically if the intake is fitted directly beneath the main body, and asks for it to sit as far from the body as possible. It also asks that an inlet duct of 850 mm or more be constructed.

Insulation on the box answers condensation before it answers comfort, and the manuals give that as the reason. Daikin's FDXM guidance asks that the duct be insulated to prevent condensation from forming, and names glass wool or polyethylene foam 25 mm thick for that range. Mitsubishi asks for full insulation on the inlet duct flange and outlet duct, and gives the same reason.

Climate is what those instructions were written against. The PEAD manual sets a threshold for it. Where air above the ceiling runs at high temperature and high humidity, with a dew point above 26 degrees C, condensation may form in the indoor unit. In that case the manual asks for 10 to 20 mm of insulation across the whole unit surface. A closed void above a flat here is warm and still, and nobody measures the dew point inside it after handover.

An uninsulated box in that void behaves the way a cold glass does on a table. Moisture in the surrounding air condenses on the metal and runs to the lowest point it can find. What it usually reaches first is the ceiling board, so the fault presents as a stain or a drip in a room, with nothing wrong inside the machine at all.

Sealing is the second job, and a standard sits behind that one. The 1995 edition of SMACNA's duct construction standard, the one published openly, defines sealing as closing openings in the surface of the ductwork and field-erected plenums and casings through which air leakage would occur. The words field-erected sit inside the standard's own definition. A box made on site is squarely within its scope.

Whatever escapes goes into the ceiling, where nobody sees it. Cooled air leaving an unsealed seam still crossed the coil and was still paid for, and no room ever received it. Daikin is direct about the expectation on its own connections. Wind aluminium tape around the intake side flange and duct connection, and make sure there are no air leaks at any other connection.

No code reaches into a flat to govern this. SS 553 sets out good practice for air-conditioned and mechanically ventilated buildings in Singapore, and its opening clause fixes the scope at all commercial, office and institutional buildings except hospitals. Inside it sit a clause titled ductwork and other air passages, a ductwork seal requirements table, and a table of minimum duct R-values. A home is outside the building types it applies to.

The construction standard says as much about itself. SMACNA records that duct systems for residences are not ordinarily subject to the provisions in this document. Where its own pages are silent, it hands the choice over. The contractor shall select configurations suitable for the service.

So the answer to who decided the shape of the box is always the same. It was not a maker, an engineer or a code. It was the installer, working to whatever depth the ceiling left and whatever the job allowed that afternoon.

One request tests a quotation without any argument about it. Ask for the outlet selection in writing, along with the approach condition it assumed and how the box and its joints will be insulated and sealed. A contractor who picked the outlet off a selection table already holds both on paper. Where the answer comes back as a machine size, the path between that machine and the room was never priced.