Mini-Split Standby Power: Why Nobody Agrees What It Costs

Short answer: calculate standby cost from measured energy over a known period, or from average real power and the hours spent in that state. There is no universal standby wattage for every mini-split. A reading from another model, season or operating mode is not a reliable estimate for yours.

Updated 1 October 2026. Calculation guide; examples below are hypothetical, not product measurements.

Define what you are measuring

Record whether the system is off at its controller but still energized, holding a setpoint, using fan-only mode or actively heating or cooling. Mixing these states makes comparisons misleading. Also note whether your monitor covers the whole system or only part of it.

Some equipment uses compressor warming while it is not providing room heating or cooling. The control strategy depends on the model. As one specific example, the Daikin FCAHG-H installation and operation manual requires power to be on six hours before operation for crankcase heating. That instruction is for the covered system; it is not a universal waiting time or a standby-power specification for other mini-splits.

Where the standby watts actually go

A mini-split that is “off” at its controller is not off. It is a networked appliance waiting to be told what to do, and several things in it stay alive. Knowing what they are explains both why the draw is not zero and why no two published figures agree.

• Control boards, indoor and outdoor. Each unit has a board that stays powered so the system can respond to the remote, run its clock and schedule, and keep the two ends talking to each other. This is small, steady, and present on every system.
• The infrared receiver and display. Always listening, by definition. Small.
• A Wi-Fi or cloud module, if one is fitted. Factory-fitted on some models, an accessory on others, and absent entirely on cheaper equipment. A radio that maintains a connection to a server draws more than one that does not exist, which is one reason two units of the same nominal capacity can measure differently.
• A crankcase or compressor heater, where the model has one and the conditions call for it. This is the item that matters.
• A condensate pump or float switch, on installations that have one. Fitted at installation rather than supplied with the equipment, so it is easy to forget it is on the same circuit you are measuring.

The compressor heater deserves its own explanation, because it is usually the largest term and the one that moves. When a compressor sits idle in cold weather, refrigerant tends to migrate towards the coldest part of the system and dissolve into the oil in the compressor. Start the compressor in that state and the oil can foam, losing its ability to lubricate at exactly the moment it is needed. The defence is to keep the compressor slightly warmer than its surroundings, with a resistance heater, so the refrigerant stays where it belongs.

Two consequences follow. First, a resistance heater is not subtle: where one is fitted and energised, it will generally dominate every other item on the list above. Second, and this is what makes the whole subject messy, it is weather-dependent and control-dependent. It may be on continuously, it may be thermostatically switched, it may only come on below some outdoor temperature, and on some inverter equipment the function is performed by passing current through the compressor windings rather than by a separate heater at all. The manufacturer’s documentation for your model is the only authority on which of these applies, and that documentation is also where you will find any instruction to energise the system for a period before operating it.

Why published figures disagree so much

If you search for a standby figure, you will find numbers that differ by an order of magnitude, all stated with confidence. They are not necessarily wrong. They are mostly answers to different questions. The usual reasons, worth checking before you trust any figure including your own:

• A different state was measured. Off-but-energised, idle while holding a setpoint, and fan-only are three different things, and only the first is standby in the sense most people mean.
• A different ambient temperature. If a compressor heater is in the picture, a reading taken in July and a reading taken in January are measuring different machines.
• A different boundary. A plug-in meter on a single indoor unit, a clamp on the outdoor feed, and a whole-circuit monitor do not see the same equipment. Multi-zone systems make this worse: one indoor head is not one-quarter of a four-zone system.
• A different method. Instantaneous watts at one moment, average watts over a day, and kWh accumulated over a week will give three different impressions of the same system, especially where anything switches on and off.
• Apparent power reported as real power. Amps multiplied by nominal voltage, as the section below explains, is not a watt figure, and at low loads with a poor power factor the gap is large.
• Different hardware. Wi-Fi module or not, condensate pump or not, separate heater or winding-current warming, single-zone or multi-zone. The badge on the front does not tell you which.

None of this makes measurement pointless. It makes your measurement the only one worth acting on, and it means a figure is only meaningful when it is published with its state, its season, its boundary and its method attached — which is what the log below is for.

Use watts or kWh, not just amps

Fluke distinguishes real power from apparent power: watts describe real power, while volts multiplied by amps gives apparent power in VA. Power factor is their ratio. Reading current alone and multiplying by the nominal supply voltage does not establish the equipment’s real energy use.

Do not judge a meter only by its price or a “true RMS” label. Check that its specifications cover real power or energy, the expected low-load range and the installation involved. A true-RMS current measurement alone is not a watt measurement. A correctly installed energy monitor recording kWh over a representative interval can be more useful than a single instantaneous reading.

Calculate your cost in three steps

1. Energy: average watts ÷ 1,000 × hours = kWh.
2. Cost: kWh × your applicable price per kWh.
3. Scope: label the time period, rate and operating state. Exclude hours spent in other states from a standby-only estimate.

Worked example: suppose an appropriately installed monitor records 0.36 kWh during 24 hours in a documented idle state. That is an average of 15 W. If the same conditions lasted for 30 days, energy would be 10.8 kWh. At an assumed $0.30/kWh, the cost would be $3.24. This is arithmetic, not evidence that your unit draws 15 W.

Replace the example tariff with your own. For time-of-use tariffs, calculate each rate period separately. Fixed account charges are not avoided by reducing standby consumption. These power values are calculation inputs, not a claimed normal range.

A measurement log you can reuse

• Indoor and outdoor model numbers; number of indoor units.
• Date, start and end times, and outdoor conditions.
• Controller mode and whether the system ran during the interval.
• Monitor model, circuit coverage and energy at the start and end.
• Elapsed hours, difference in kWh and tariff used.

Keep several comparable records before drawing a seasonal conclusion. If heating or cooling occurred during the interval, label it total system energy rather than standby energy. For multi-zone equipment, measure the covered system; do not multiply one indoor unit’s reading by the number of rooms and assume it represents the outdoor unit.

Is it worth doing anything about?

Work the arithmetic before the anxiety. The table above is deliberately wide because the honest answer depends on which row you land on, and the two ends of it are different situations. At the low end, a steady few watts is a rounding error against the cost of actually heating or cooling the house, and is not worth a change of habit. At the high end, a continuously energised resistance heater through a cold season is a real line on a bill, and is worth asking an installer about — not to disable it, but to find out whether the control strategy on your model is working as intended.

Two comparisons help keep it in proportion. First, compare the standby figure with what the same system uses on a day it actually runs: if standby is a small fraction of that, your attention is better spent on the running, which is where auxiliary heat and switchover settings live. Second, compare it with the fixed standing charge on your own bill, which you pay whatever you do. Standby cost that is small next to a charge you cannot avoid is not where savings are.

The one thing worth checking regardless of the number is whether anything on the circuit should not be drawing at all — a condensate pump cycling when there is no condensate, for instance, or a reading that does not fall when the weather warms on a system you expected to be heater-driven. An unexplained draw is more interesting than a large one.

Should you turn the power off?

Follow your model’s shutdown and restart instructions. Do not adopt a nightly breaker-switching routine from an online cost table. Ask the installer or manufacturer how to handle a long seasonal shutdown and any pre-start energizing requirement.

Measurement boundary: read an existing user-facing energy display if available. Do not open a panel, disconnect or live equipment enclosure, or alter hardwired wiring to install a monitor. Arrange that work with a qualified electrician. This article gives no live electrical measurement procedure.

Next reading

For a different source of running cost, see auxiliary heat or dual-fuel switchover. If you are comparing equipment rather than investigating a bill, start with the mini-split buying trade-offs.

Basis and limitations: the table is a reproducible calculation, not a laboratory test, measured installation or current tariff survey. Manufacturer requirements are model-specific. See our editorial standards or report an error.