Aircon standby power: what stays on when you switch off

The remote never reaches the supply. It sends a signal to a controller, and that controller has to be listening when the next signal arrives. Staying ready to hear the next one costs power. A machine parked by the remote is therefore holding a state, rather than sitting dead the way a switched-off kettle does.

What holds that state is short and specific. There is the receiver, the board reading it, the panel light or display, and the clock behind any timer setting. A network module counts too, where one is fitted, because a module that has dropped off the home network cannot be reached by an app. None of this moves air or refrigerant. All of it waits.

The draw is buying something, which is the part a warning about phantom load usually leaves out. A system without supply cannot be reached at all. A timer set for tonight does not count down. An app cannot wake the machine before anybody gets home. Those are small conveniences, and the indoor side of the fixed draw is what pays for them.

Standby and off are separate conditions in the measurement world, not two words for one thing. IEC 62301, the standard NEA names for standby power testing on regulated aircon, treats standby mode and off mode as distinct states with distinct readings. A split system parked by its own remote sits in the first of those. Reaching the second means the supply itself has gone.

Two neighbouring questions get settled elsewhere, and naming them keeps them from tangling into this one. Whether a timer or a hand on the remote is the better way to end a cooling session is a runtime question, and the timer versus manual off guide answers it. What the isolator is, and why the supply feeding it is kept clear of other loads, belongs with the dedicated power point. Standby is what is left over after both of those have been decided.

The outdoor unit runs a board of its own and has its own reason to keep it alive. It has to hear the indoor side and be ready to act on what it hears. Proximity to the remote is not what makes a system responsive. Being powered is, and the machine on the ledge is powered whether or not anybody indoors is thinking about it.

Some designs go further and keep the compressor conditioned while the machine sits idle. Refrigerant in a stopped system migrates towards the coldest point it can find, and it can settle into the compressor oil. Oil carrying dissolved refrigerant does not lubricate properly at the next start. A small warming load holds the oil above the point where that settling happens, and where a system carries one, it draws while the machine looks entirely inert.

That load is why an outdoor figure can be larger than anything the indoor unit accounts for. It is also why one number describes it badly. A warming load is usually governed by temperature rather than left permanently energised, so what a machine holds while idle shifts with the weather and with how the designer chose to handle the problem. The crankcase heater is its own subject and gets its own page.

Manufacturer literature is the clearest evidence that the load is doing real work rather than wasting power by oversight. Installation instructions commonly direct that the outdoor unit be left energised ahead of the first compressor start, so the warming load has finished its job before anything turns. Guidance of that kind exists because starting a compressor on refrigerant-loaded oil damages it.

The standby figure attached to a registered model is measured, not estimated. NEA names IEC 62301 as the standby power test standard for regulated aircon, in both its 2005 and 2011 editions. A supplier's test report has to be produced against it before the model can be registered for sale here. The figure is the outcome of a defined procedure. That makes it a different sort of number from anything a retailer improvises.

The standard defines how to measure and nothing beyond that. IEC states plainly that it specifies no maximum on power or energy consumption. Any ceiling is a separate instrument, and it belongs to the supply standard rather than to the test method. Where that ceiling sits today, and how it scales once a system has several indoor units, is set out in the MEPS guide and is not repeated here.

Measurement happens with the machine parked in a controlled room, doing nothing, under conditions written to be repeatable across laboratories. Repeatability is what makes the figure useful for setting two models beside each other. That same repeatability is what prevents the figure describing a real ledge on a real night. Out there the ambient is not the laboratory's, and a temperature-governed protective load may or may not be drawing at that moment.

Read the number as a comparison and then stop. It answers which of two models holds less while idle, tested on a single basis. What the machine outside your window is holding tonight is a separate question. No arithmetic on a declared figure converts the first answer into the second.

Finding the figure for a specific model is a documentation question rather than a shop-floor one. It sits behind the registration as part of the test report. A printed specification sheet may or may not carry it forward. Ask for the model's own paperwork rather than a number quoted from memory. A standby figure nobody can trace back to a test report is an assertion, and assertions are what a declared figure exists to replace.

Standby is a floor under the bill rather than a movement in it. It reads the same in a month of heavy cooling as in a month with none, because nothing about it responds to how the system gets used. That single property explains both why households miss it entirely and why it makes a poor suspect when a statement looks wrong.

Small and continuous is the shape people estimate worst. Any one hour of it is beneath noticing, so the draw gets discounted to zero without a second thought. The hours themselves never stop, which is what makes zero the wrong figure. Running consumption has the opposite shape, large and intermittent, and it collects all the attention because a compressor labouring on a hot afternoon is easy to picture.

The practical consequence of the shape is diagnostic. A bill that moved was not moved by standby, because standby did not move. Anything identical in every hour cannot account for a difference between two statements. Bill spikes have a page that works through the causes that do move, and the tariff and rebate effects that imitate a spike have pages of their own.

Where standby matters most proportionally is where the aircon runs least. A household cooling one bedroom at night spreads that fixed draw across few running hours. It therefore forms a larger share of what the system costs than it would in a flat cooled hard across several rooms. The share is largest precisely where the total is smallest. That is why chasing it so rarely repays the attention it gets.

The case for removing supply is real and narrow. It applies where a system will not be run for a long stretch, and it thins to nothing where the machine is going to be used again before the day is out. Length of disuse is the only variable that genuinely moves the answer, and every other argument tends to be a reformulation of it.

Set against nightly use, the trade is poor on both sides at once. The saving is a fixed draw across hours the machine was going to spend idle regardless. The cost is losing whatever protection that idle load provides, repeated every night, in a household that must then remember to restore supply before the next start. A protective margin converted into a nightly chore is a margin somebody will eventually skip.

The isolator itself belongs in the arithmetic and usually gets left out of it. It sits outdoors, takes weather, and corrosion goes to work on its contacts. A switch worked rarely and a switch worked every night do not age alike. A failing one produces symptoms that read like a fault in the machine behind it. The system stays silent when cooling is called for, or it wakes on one attempt and ignores the next. What that switch is for, and why it lives where it does, is covered alongside the circuit it belongs to.

None of this is an instruction to go and operate anything. Whether a specific system should sit without supply is a judgement about that machine, and it needs somebody who knows what is inside it. The manual, the installer, or the technician who services it can answer it in one exchange. Deciding is the homeowner's part. Improvising at the ledge is not.

The honest summary is that standby is real, small, and mostly not where the money is. It earns understanding because the machine does not behave the way the remote implies, and because the supply decision gets made on poor reasoning in both directions. Cutting power to a flat standing empty is sensible. Cutting it nightly to save a fixed draw buys a small saving with the protection that draw was there to provide.

• •Does this system hold a load that conditions the compressor while idle? The manual answers it, and so does the installer.
• •How long is the machine genuinely going to sit? Anything counted in nights is not the situation this applies to.
• •Who restores supply before the system is next asked to run, and will they remember?
• •Is the isolator in a fit state to be worked often, or has the ledge already had its way with the contacts?
• •Has anything larger been ruled out first? A fixed idle draw is the smallest lever available on an aircon bill.