Aircon drain gradient: why the fall must be continuous

A standard split aircon has nothing that pushes water out. The drain line is open at the pan and open to the air at the far end. Water moves along it because each stretch of pipe sits lower than the stretch before. There is no pump anywhere in that path unless somebody fitted one, and most wall units in Singapore flats do not have one.

The line carries more water than most people picture. A unit cooling a bedroom through a humid night strips litres of moisture out of the air. All of it leaves through a pipe about the width of a finger, under no pressure at all. The only force behind it is the weight of the water already sitting in the pipe.

Installation instructions for split systems put a figure on the fall. The drain run should slope at least 1 in 100, which is one centimetre of drop for every metre of pipe. Steeper is better where the route allows it. The figure matters less than the word continuous. The fall has to hold across the whole run, from the pan outlet to the discharge point.

A rise anywhere cancels every bit of fall before it. Water arriving at a high point stops there and pools behind it. The pipe past that point can be laid perfectly and it will still sit dry, because nothing is reaching it. One badly supported stretch is enough to do this to an otherwise sound job.

Fall is easy to lose without anyone deciding to lose it. Flexible hose sags between clips when it is held at too few points, and a sag is a belly the water settles into. A long horizontal run with little headroom leaves nothing to spare, so one clip fixed slightly high flattens the stretch behind it. None of that stays visible once the trunking cover goes back on.

Most recurring leaks are made here. A stretch of pipe with no fall does not empty when the unit stops. It holds a standing pool, and that pool is still sitting there when the unit next runs.

Standing water is what biofilm needs. Slime forms on the pipe wall wherever water rests, and it thickens rather than clearing, because no flow scours it away. Each layer takes a little more off the bore. In Singapore's humidity, those conditions hold for most of the year rather than for a season.

A line laid with continuous fall rarely reaches that state. Water enters it and keeps moving, so the pipe runs empty between cooling cycles and dries out. Growth has little to hold on to, and the flow carries off most of what does settle. The difference between a drain that blocks over and over and one that never blocks is usually not how dusty the flat is. It is whether the pipe empties itself.

Blockage frequency is therefore a signal, not just a nuisance. A line that blocks once has probably taken in something it should not have: debris during works, an insect, a lump of growth dislodged during a clean. A line that blocks again and again, in the same way, is holding water somewhere along its route. The first is an event. The second is a shape.

A belly in the run traps air as well as water. The dip fills, the water seals it, and air behind that seal has to be shoved through in slugs. That is the gurgle heard at the indoor unit, and it reads as an early report rather than a quirk. It says water is not running freely, which on a line with proper fall it should be.

The leak appears where the line is open, not where it is wrong. Water backing up a restricted drain has one route available, and that route is backwards. It returns to the pan, fills it, reaches the lip and spills into the room.

The visible drip is the end of a queue, and the queue can be metres long. A defect near the discharge point, on the far side of a wall, still presents as a wet patch under the indoor unit. Where the water shows says almost nothing about where the pipe is at fault.

This accounts for a large share of repeat visits. Treating the leak as an indoor-unit fault leads to cleaning the pan and clearing the reachable section. The work is real, and the drip does stop. The stretch that caused the queue was never reached, so the queue forms again.

Push back if a leak is being explained entirely in terms of the indoor unit. The pan and the first stretch of pipe are the easy part of the route. On a line that has leaked before, they are also the part least likely to be the cause. Ask what was checked past the wall.

Most of a drain line cannot be seen. It leaves the pan, enters trunking within a short distance, and carries on behind casing, above a false ceiling, or inside a chase cut into the wall. It shares that trunking with the rest of the pipework.

Concealment removes the simplest check there is. On an exposed run, fall can be read with a spirit level and an eye. Once the pipe is boxed in, seeing the slope means opening the route. That is destructive on a chased wall and disruptive above a ceiling, so it is rarely the first move.

Diagnosis then works by inference. Behaviour stands in for inspection: how the outlet flows while the unit is working hard, whether the indoor unit gurgles, whether the line empties after shutdown, and what pattern the leak follows. None of that is as good as seeing the pipe. It is usually all there is before deciding whether opening the route is justified.

Two events are worth raising with a technician without being asked. Renovation that re-boxed a ceiling or moved trunking can flatten a route that used to fall correctly. A mounting bracket working loose does the same thing at the start of the run, because the pan tilts with the unit it is bolted to.

A flush restores the bore. It does not restore the fall.

Water returns to the same flat stretch as soon as the unit runs again. Growth restarts in the same place, under the same conditions, at roughly the same rate. Nothing about the geometry changed, so nothing about the outcome changes either.

That produces a pattern worth recognising. The leak comes back after each clearing, in the same spot, behaving the same way, with roughly even spacing between one clearing and the next. A homeowner who notes nothing more than the date of each clearing is holding the most useful evidence in the case.

Push back when a further clearing is offered as the fix for a leak that has already been cleared more than once. Clearing is the right first response, and often the only response needed. Offered again and again for the same returning leak, with nobody having checked the fall, it treats a fixed defect as a run of accidents.

Correcting fall is usually smaller work than it sounds. Often the run only needs re-clipping so the belly comes out, which means opening a length of trunking rather than replacing pipe. Where the route itself is wrong, the honest choices narrow to a shorter run toward a nearer discharge point, or a pump on a stretch that cannot be made to fall. Any of those is a one-time job. What matters is that somebody read the pipe before choosing, rather than reaching for the flush pump again.

Gravity is not always available, and a pump is the correct answer when it is not. A ceiling cassette or a ducted fan coil sits in a void with no downhill path out of it. Those units are built around a small pump, which is a design decision rather than a patch.

The same holds where an indoor unit sits below the only usable discharge point, or where renovation moved that point to the wrong side of the room. Recovering continuous fall across the whole run can mean opening finished surfaces. A properly specified pump is a better answer than a pipe that almost falls.

A pump changes the failure mode rather than removing it. It adds a float, a motor and a power feed, and each of those can stop working. A condensate pump fault then fills the tray exactly as a blocked line would, which is why unit type belongs in the first description of any leak.

What does not belong is a pump fitted to rescue a gravity run that could have been laid correctly. It turns a one-time routing decision into a part that has to keep working for the life of the system. Fall costs nothing to maintain once it is right. A pump never gets to make that claim.