Sumatra squalls: the aircon faults nobody can reproduce
A squall crosses from the west before dawn, drives rain sideways, and passes. Faults it triggers appear once, then clear before anyone can look. What you noticed while the front was overhead ends up being the only evidence there is.
By Team Snowflake | Updated 6 Aug 2026
What is a Sumatra squall?
A Sumatra squall is one named weather event, not a general word for a bad storm. The Meteorological Service Singapore describes it as an organised line of thunderstorms. The line forms over Sumatra or the Strait of Malacca, then travels eastward across the island. It passes over. It does not settle in.
Two published details decide how a squall works on an aircon. The met service gives the arrival as somewhere between the predawn hours and midday, and describes a sudden onset of gusty surface winds of 40 to 80 km/h with heavy rain lasting one to two hours. Gusts up to 93 km/h have been recorded during a passage. The bearing is consistent, because the movement is west to east on southwesterly or westerly winds.
The season is published too. The met service ties squalls to the Southwest Monsoon, which it gives as June to September, and to the inter-monsoon periods. That is not the same window as the wettest stretch of the Northeast Monsoon, which is what most households picture when they say rainy season.
Wind is what separates a squall from a heavy downpour. Vertical rain falls onto horizontal surfaces, and a building is built to shed it. Driven rain lands on vertical surfaces instead, with pressure behind it, arriving from one side and staying there until the line has moved through. Every lid, drip lip and seal on the west face is being loaded in the way it was least designed for.
One event, not a season
A squall is a single passage lasting part of a morning. A monsoon season is months of higher humidity, heavier rainfall and a heavier cooling load on the unit. The two produce different fault lists, and reading one as the other points the search at the wrong hardware.
Seasonal behaviour already has its own page. Which wet-weather symptoms are normal and which point at a real fault is covered in aircon faults in the rain. What follows here is narrower: the hour the front is overhead, and the things that only break then.
Which aircon faults track a passing squall?
Three fault families follow a squall, and an owner can separate them without opening anything. Water arriving where ordinary rain never puts it. A breaker going out at a specific point in the passage. An outdoor unit behaving oddly while the gusts last, then settling once they stop.
Driven rain reaches upward and sideways, so it finds gaps that face the weather rather than the sky. It travels under a drip lip, behind a wall sleeve, past a cover that copes with vertical rain perfectly well. The route is almost always on the windward face, and for a squall that means the west side of the building.
Where the trip falls in the passage narrows it further. A breaker that goes out in the first minutes, before much rain has landed, points at something the wind reached or shifted. A breaker that goes out well into the downpour points at water that has had time to track along a path and bridge something live.
An outdoor fan is built to push air one way, and a gust across the discharge grille pushes back. Air the unit has already heated gets driven straight back into the coil. That is the same hot air recirculation a boxed-in ledge causes, arriving all at once instead of gradually. Head pressure climbs, and a system that cannot shed heat quickly enough stops itself rather than carry on. That is high pressure protection doing its job, not a breakdown.
Some of what the wind does is mechanical, and it shows up late. A fan guard screw that was already slightly loose, a bracket never brought up to full tension, leaf litter packed hard against the coil. The squall finishes the job, the rain covers the noise, and the household first hears a new rattle the following evening.
| What you noticed | When in the passage | What it narrows to |
|---|---|---|
| What you noticedWater indoors near the pipe entry | When in the passageAt the heaviest rain, on the side taking the wind | What it narrows toThe wall sleeve or its seal, rather than the drain |
| What you noticedBreaker out as the wind arrived | When in the passageFirst minutes, before much rain had landed | What it narrows toSomething exposed or shifted at the outdoor unit |
| What you noticedBreaker out deep into the downpour | When in the passageAfter steady rain, with the unit running | What it narrows toWater reaching a connection that should be sealed |
| What you noticedUnit stopped, then restarted by itself | When in the passageWhile the gusts lasted | What it narrows toA protection cut-out, usually heat fighting the wind |
| What you noticedNew rattle that outlasted the storm | When in the passageHeard once the rain noise had gone | What it narrows toA guard, bracket or fixing the wind worked loose |
Why the fault disappears before anyone can test it
The difficulty is not the fault. It is that the conditions which produced it left with the weather, and they are not available on request.
Nothing on a calm afternoon recreates driven rain. A hose test is still worth doing and it does find gross faults, but it aims water inward and downward from close range. A squall aims it upward under a lip, at pressure, from one bearing, for as long as the line takes to cross. A seal can pass the first and fail the second.
Everything else resets itself as well. A breaker gets switched back on almost immediately. A damp patch dries before anyone thinks to photograph the wall. The unit that cut out has restarted and is cooling normally again. What arrives is a working system and an account of something that happened once.
That gap is where the wrong repair gets sold. A trip nobody can reproduce invites a part swap, because a part is a tangible thing to hand over and a seal is not. An isolator lid, a sleeve seal or a guard at the drain outlet is cheaper and far more likely. None of them gets examined unless somebody can say which way the rain was coming from.
Timing carries the whole difference. Told that it trips sometimes, a technician tests everything and finds nothing. Told that it went out as the wind hit, from the west, before anyone was awake, and has run normally since, the same technician has one face of one wall to start on.
What to note while the squall is still passing
Nothing here requires going outside. Everything worth recording can be seen or heard from inside the flat, and the ledge keeps until daylight and dry weather.
The general method already exists elsewhere. What to write down for a fault that comes and goes, and in what order, is set out in the intermittent aircon fault log. Two things are specific to a squall, and both are usually the parts left out: direction, and sequence.
Direction is the one nobody records. Every squall tracks the same way, west to east, so the windward face is predictable, and knowing which windows were taking the rain turns a search of the whole installation into a search of one wall. Sequence is the second: what came first, the wind or the water, the noise or the trip.
- Which side of the flat the rain was striking, and which windows were taking it
- Whether the aircon was running when it happened, and in which rooms
- Whether the fault came with the first gusts, or later once rain had been falling steadily
- Whether the unit stopped by itself, or the breaker went out
- Whether it recovered on its own after the line passed, or stayed down
- Where water first appeared, and whether it was still arriving after the rain had stopped
- A photo taken the same morning, while the wall is still marked
A note that survives until the next one
A single squall settles almost nothing. Watch the same symptom through the next two and it settles a great deal, because a repeated response to a repeated event is a pattern rather than an accident.
That is the honest reason to write it down at the time. Memory of a predawn storm is thin by the evening, and the detail that goes first is usually the one that mattered.
Was it the weather, or was it the installation?
A squall is a stress test, and the useful question afterwards is what it found. Part of what happens is the weather being the weather, and the right response to that is nothing at all.
Weather alone looks like this. A room that feels damp for the rest of the morning, because the outside air is saturated. A brief noise from the ledge while the gusts last. Water running off the ledge and drying. A single trip that never repeats. Those cost nothing to ignore, and chasing them wastes a visit.
A defect in the installation looks different, and the giveaway is that it repeats on schedule. Five show up often enough to be worth naming.
- Water inside a switch box or isolator on the windward face. A lid that sheds vertical rain and admits driven rain is doing half of its job
- Water at a wall penetration. The sleeve should fall towards the outside and be sealed at both faces, and a squall finds the ones that are not, from the same bearing every time
- A drain outlet that backs up under wind pressure and overflows indoors. That one has its own page, because an aircon leaks when it rains for several separate reasons
- A condenser sited to take the line head on, with no wall or screen between it and the weather. Outdoor unit placement is settled once and lived with afterwards
- A ledge that ponds, or a fixing that has worked loose under load. HDB aircon ledge problems covers what to look at, in dry weather and from a safe position
When the breaker is the part that repeats
A unit that goes out on every squall is not asking to be switched back on again. It is reporting a route between wet and live that only opens when rain arrives sideways, and each reset puts the same path back under load.
Trip timing narrows the electrical test on its own, and the aircon breaker-tripping patterns page sorts the common timings. Bring the weather detail to it. A trip tied to a bearing and a point in the passage is a much smaller search than a trip tied to nothing.
The reason any of this repays the trouble is that the trigger is predictable. A fault that answers to a squall will answer to the next one, and the next one is not a question of whether. An owner who can say what the unit did, from which direction, and at which point in the passage has handed over most of the diagnosis before anyone arrives.
Ready to get started?
Tell us what’s going on. Symptoms, setup, photos, anything we should know. We’ll assess and come back with the right next step.