Aircon Height Difference: Which Unit Sits Higher Matters

Gravity gets hold of two things inside a refrigerant circuit. The first is the column of liquid refrigerant standing in the liquid line, which has weight and must be raised or dropped. The second is the film of lubricant the compressor sheds, which has to arrive back. A climb makes the trip harder and a fall makes it easier.

The two allowances also lean on each other, which most quotes treat as settled. Johnson Controls, in its piping application data sheet for split systems, states that the elevation allowed falls as the total equivalent line length rises. A three ton unit routed to 175 equivalent feet is permitted 75 feet of rise on one liquid line size, and 95 feet on the next size up. The taller the climb, the less distance is left to spend.

Put the condenser above the room and the lubricant is on a climb. Vapour leaving the indoor unit has to carry it up to the compressor. Johnson Controls names that arrangement as the one where a minimum riser velocity applies. The requirement bites in that direction only, because that is where oil has to beat gravity to get home.

Liquid in the same arrangement gains instead of losing. The document puts the static pressure change at half a pound per square inch for every foot of vertical drop. Coming down, it notes, that gain can cancel the friction loss outright, and a pressure gain in the liquid line does the system no harm.

With the indoor unit above the condenser, liquid has to be lifted, and that half a pound per foot works against the circuit, eating into subcooling. Once subcooling is spent, refrigerant begins to flash inside the vertical section, and the loss rate climbs as more gas forms; here the liquid line caps the elevation allowed. Shutdown also runs against the direction: the Daikin Sky Air manual covering RZQG and RZQSG units warns about oil sitting in riser piping and draining back into the compressor once the machine stops. Both hazards stay silent while the machine is running well.

Published allowances sit far enough apart that no general figure is safe to carry into a quote; two wall splits from two makers differ by three metres on the same measurement. Daikin's manual for the FTXC25 to 60A indoor units with RXC25 to 60A outdoor units gives a maximum allowable elevation of 15 m for a Singapore flat. Pipe length runs to 20 m on the 25 and 35 classes and 30 m on the 50 and 60. Mitsubishi Electric's manual for the MSZ-AP15VGD and MSZ-AP20VGD allows 12 m against 20 m of pipe, with no mention of direction.

One residential manual does split by direction, and it is worth reading twice. Mitsubishi Electric publishes 15 m for the MXZ-2F42VF3 and MXZ-2F53VF(H)3 multi-split outdoor units, then footnotes it: if the outdoor unit sits higher than the indoor unit, the allowance drops to 10 m. The smaller MXZ-2F33VF3 starts at 10 m and has nothing left to give back.

Daikin's residential multi declines to make that distinction on equipment of the same class. Its manual for the 2MXM, 2MXT, 3MXM and 4MXM outdoor units draws the arrangement twice, once with the outdoor unit above and once below, and both diagrams are marked 15 m. Two makers, two multi-splits, an open disagreement about whether direction changes the answer.

Larger equipment climbs without changing the argument. Daikin's Sky Air RZQG71 to 140L manual gives 30 m between indoor and outdoor, while LG publishes 49.2 feet for its Multi F and Multi F MAX outdoor units, landing back at 15 m. Daikin's TP-110 long line publication goes furthest, covering R-32 splits between 18,000 and 60,000 Btu/h. With the outdoor unit above, single stage equipment has no elevation restriction beyond the equivalent length ceiling and two stage is held to 25 feet; below, both are capped at 80 feet.

A trap is a low point formed on purpose in a vertical line. There, oil the gas cannot carry as a film pools, then leaves as a slug once deep enough. Gravity beats a thin film almost every time, but loses to a body of oil with the whole flow behind it.

Daikin's Sky Air manual for the RZQG and RZQSG range asks for one at each 10 m of height difference in the riser gas piping. It gives two reasons. Oil held in the riser flows back into the compressor when the machine stops, and oil return deteriorates. Then comes the line that makes the geometry matter: a trap is not necessary when the outdoor unit is installed higher than the indoor unit.

The same maker's long line publication asks for one in the opposite arrangement. Where the outdoor unit sits above and the elevation passes 80 feet, it calls for an oil trap at the indoor unit. Where the outdoor unit sits below, it calls for an inverted loop before the indoor unit instead. Copeland's refrigeration manual takes the general trade position of a P-type oil trap at the base of each suction riser, with any suction riser over five feet trapped at the bottom.

Two documents take the other side entirely. Johnson Controls states flatly that traps are not required where the piping has been sized correctly, and that a trap only adds pressure drop and cuts capacity further. LG is stricter again on its Multi F and Multi F MAX systems, listing field oil traps as not permitted anywhere on the refrigerant piping. Its stated reason is that those systems already carry their own arrangements for getting oil back.

Both departures cost the same. A trap fitted on general principle deviates from the specification as much as a trap left out where the maker asked for one. An installer answering from habit has answered about some other machine. The document for the model on the nameplate is the only authority that settles it.

There is also the trap nobody drew. LG's manual notes that sagging pipes turn into oil traps and names that as a route to equipment malfunction. Johnson Controls warns that sound piping practice can still create oil traps where the two units sit at different levels. The difference is what happens next: a designed trap releases what it gathers, and a dip in a poorly supported run holds on to it.

Most flats never come close to testing the figure. A wall unit sits high in a bedroom, the condenser sits on the ledge immediately outside, and the two are a metre or two apart vertically. That is nowhere near 12 m, let alone 30, and counting storeys gets an owner most of the way without a tape, since every floor of separation is worth several metres.

The question turns live wherever the condenser could not follow the room. Plant on a ground-level slab can serve flats several storeys above. A shophouse may have equipment on the roof and the served space below, and a landed house may have the plant deck beside the ground floor with bedrooms two storeys up. Each puts a real climb into the pipe, and its direction decides which hazard applies.

There is a commercial edge to the figure that rarely gets raised at quotation. Johnson Controls states in the same data sheet that warranty coverage excludes failures caused by improper application. A route built outside the published allowance therefore moves risk off the maker and onto whoever signed the job off, and knowing that before the copper goes into the wall is worth something.

Two things can be established without tools: which unit ends up higher, and roughly by how much. The model number is on the nameplate outside, and installation manuals are published against model numbers. Most of this is checkable before anyone is called.