Inter-monsoon weeks: when a working aircon looks faulty
Between the two monsoons sit weeks that ask more of a cooling system than either season does. Nothing inside the machine has changed, but the complaints start. Telling a harder ask from a genuine decline is the whole question.
By Team Snowflake | Updated 7 Aug 2026
What are the inter-monsoon periods?
Singapore's year gets described as two monsoons, which leaves four months unaccounted for. The Meteorological Service Singapore names two inter-monsoon periods in those gaps. The first runs from late March to May. The second runs from October to November.
The monsoons themselves are the parts most households can name. The met service gives the Northeast Monsoon as December to early March, and the Southwest Monsoon as June to September. Each carries a settled wind direction and a fairly predictable character. The weeks between them carry neither.
The met service describes inter-monsoon winds as light and variable, interacting with land and sea breezes. It records thunderstorms, at times severe, in the afternoon and early evening. For the first period it adds that hot afternoons are common, with a maximum above 32°C. That combination is why these weeks are the hardest ones a cooling system meets.
A season and a single storm are different questions, and both already have pages here. Rain-linked symptoms, and which of them signal a real problem, are handled under aircon during monsoon. A line of thunderstorms that crosses the island before noon is narrower again, and the sumatra squall page takes that hour on its own. Neither is the subject here. The difficulty in the transitional weeks is not water getting in. It is heat and moisture arriving on the same afternoon.
Two periods, and not the same weather twice
The two periods behave differently from each other. The met service describes the second, from October to November, as generally wetter than the first. The first is the one that carries the hot afternoons.
Lightning splits the same way. The met service reports the highest incidence of lightning activity in November, followed by April and May, and puts the inter-monsoon months at more than half the year's lightning strokes. An owner who finds one of the two periods harder than the other is not imagining the difference.
Why still air and a hard sun load a system differently
Two peaks land on the same day in these weeks, and neither monsoon does that. Strong daytime heating drives the first. The afternoon storm drives the second. A system that coped all year meets both of them between lunch and dinner.
Solar heating is the first half, and it is the same process that builds the storm. The met service sets it out plainly: solar radiation warms the surface, the lowest layer of air warms with it, and that warm humid air rises and condenses into cloud. Long before any of that becomes a storm, it has spent the morning heating walls, windows and roof slabs. What reaches the room from all that is set out under heat load.
Light winds are the second half, and they work on the outdoor unit. A condenser sheds heat into the air around it, and moving air carries that heat away. Still air does not. Warm air lingers near the coil instead of being replaced, so every cycle rejects its heat a little more slowly. A unit boxed into a sheltered ledge feels this first, because it was already living on whatever breeze reached it.
Then the storm arrives and the demand changes shape rather than size. Outside air saturates. More of the machine's capacity now goes on drying that air, and less on bringing the temperature down. A room can sit at its set point and still feel wrong. The moisture side of that has a reading of its own, and wet-bulb temperature is where it gets explained.
Order matters as much as size. Peak heat comes first, the storm follows it, and the saturated evening comes last. There is no easy stretch in the middle for the system to make up ground. In a settled season the load builds and falls once. In these weeks it builds, changes character, and then refuses to fall away.
The load does not end when the sun goes down
Evening does not reset the day in these weeks. Concrete, tiles and sun-facing walls give back the heat they took on, long after the sun has dropped. Air that saturated during the storm stays close to saturated.
So a machine that struggled at four does not always get an easy nine. The overnight stretch that normally lets a room settle runs shorter and warmer than it would in a settled month. A system with no margin left starts the next hot day already behind.
Which complaints cluster in these weeks
Four complaints arrive together in the transitional weeks, and they are reporting different things. Each has its own page, because each has causes that owe nothing to the season. What the season adds is a reason for all four to appear at once in a system that was fine last month.
Read the table as a translation, not a diagnosis. The left column is what gets described on the phone. The centre column names what the weather is doing to produce it. The right column points at where the fault-side causes get dealt with properly.
| What gets reported | What the season is doing | Where the causes are covered |
|---|---|---|
| What gets reportedThe room needs much longer to get comfortable | What the season is doingPeak heat load meeting a condenser in still air | Where the causes are coveredtakes too long to cool |
| What gets reportedCold overnight, weak by mid-afternoon | What the season is doingLoad swings hard with the sun, the machine does not | Where the causes are coveredcools at night not afternoon |
| What gets reportedCool enough, but the air still feels damp | What the season is doingMore capacity going to moisture, less to temperature | Where the causes are coveredroom humid even when cooling |
| What gets reportedWater showing up where it never used to | What the season is doingHigher condensate volume, and surfaces below the dew point | Where the causes are covereddew point |
The complaint that arrives with a date attached
All four of those are load reports rather than fault reports. Same machine, harder week. Whether any one of them is also a fault stays a separate question, and the answer comes from what happens once the week eases.
A complaint that started on a particular week is more useful than one that has always been there. It puts a boundary on the search. Whatever changed was sitting outside the machine, and the calendar names the first candidate.
That is worth saying out loud to whoever attends. A system that never cooled that room well and a system that stopped cooling it in April point at two different searches. The first is about sizing and installation. The second is about load, and load is seasonal until something rules the season out.
Seasonal ask, or a system in decline?
Recovery is the test, and watching for it costs nothing. A system meeting a seasonal load recovers as soon as the load drops. A system that has lost capability does not.
Recovery looks specific enough to check. The room that lagged through the afternoon is comfortable by late evening. The morning after a storm, it pulls down at its usual pace. A cooler, cloudier day passes with no complaint at all. Capability is intact, and the week was simply asking more of it than the room usually asks.
A genuine decline holds its shortfall after conditions ease. Evenings stop fixing it. Mild days stop fixing it. The gap widens instead of tracking the sky. A unit that held the room through last year's hot spell and cannot hold it on a mild evening now has lost something real.
The second signal is direction of travel between seasons. A seasonal ask repeats without deepening. A decline compounds. Setting this inter-monsoon period against the last one says far more than setting this week against last week. Where a complaint comes and goes, record it as it happens. The intermittent aircon fault log lists which details survive being remembered.
The trap these weeks set
If a gas top-up gets proposed because cooling weakened in these weeks, ask what was measured first. A sealed system does not use refrigerant up. Weak cooling on a hot, still, saturated afternoon is what a correctly charged unit does under load. Airflow, coil condition and condenser clearance cost less to check and are far more often the answer.
A top-up that appears to fix a seasonal complaint has fixed nothing. The complaint returns with the next hot spell, and the charge in the system is now a guess rather than a known quantity. Anyone reaching for refrigerant before naming the load has skipped the step that would have settled it.
The larger version of the same trap is a capacity conversation. A unit that lags through the hardest weeks and holds the room for the rest of the year is not undersized in the way a replacement quote implies. It is being asked for more, for a stretch of the year. What a machine is rated for and what it delivers in a real room are two different numbers, and actual cooling capacity is where that gap gets explained.
Why the machine gets blamed
Nothing inside the unit changed between the settled month and the hard one. That is precisely why the complaint reads as a fault. A person notices a change in outcome and goes looking for a change in cause. Here the change sat outside the window.
None of this is a defence of every unit. Plenty of systems genuinely decline, and these weeks are when a marginal one finally shows it. The season has to be ruled in or out before the machine gets blamed, and ruling it out is a free step that almost nobody takes.
What is worth setting up before the period starts
Anything that helps has to be in place before the load arrives, not during it. None of the following is a repair. It is the difference between meeting a hard week at full capability and meeting it already short.
Timing is the only genuinely seasonal part of this. Every item below is worth attention in any month of the year. What the calendar changes is the cost of getting to them late, because a marginal filter or a crowded condenser stays invisible until the week that needs the margin.
The list is short. Only a few things move the margin, and all of them are about capability rather than symptoms.
- Filter condition. A partly loaded filter costs airflow, and the coil needs airflow most on the days the load peaks
- Coil surface, indoors and outdoors. Fouling slows heat transfer in both directions, and still air is already slowing one of them
- Condenser clearance. Boxing in an outdoor unit removes the small help a light breeze was giving it
- Drain path. Condensate volume rises with humidity, and a drain that copes on an average day can back up on a saturated one
- Room-side heat. Blinds and curtains on the sun-facing side cut the load before the machine has to remove it
What the season does not justify
A turning calendar is no reason on its own to book work. Knowing the season does the opposite job. It removes a class of false alarms, so attention lands on the complaints that survive the weather changing.
The season also does not explain everything blamed on it. A room that feels different from the room beside it, at the same setting, is not a seasonal effect. Where that gap comes from is covered under how cool a room feels.
The reading that settles everything is the one taken after the period passes. A complaint that faded with the weather described a hard few weeks and nothing more. A complaint still sitting there in a settled month is describing the machine, and that one is worth acting on.
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.