Watt-hours, milliamp-hours, and what actually drains a laptop battery

Laptop batteries are quoted two ways. One of them can be compared across machines and one of them cannot, and manufacturers are not consistent about which they print. Getting this right is the difference between comparing two laptops and comparing two marketing numbers.

Watt-hours can be compared. Milliamp-hours cannot.

Energy stored in a battery is capacity multiplied by voltage. Watt-hours already include the voltage, so a watt-hour figure is directly comparable between any two batteries. Milliamp-hours do not: the same milliamp-hour figure at a different cell arrangement is a different amount of energy. Laptop batteries are built in different series and parallel arrangements, so their voltages differ, so their milliamp-hour figures are not comparable.

If a listing gives only milliamp-hours, convert: multiply by the nominal voltage, which is printed on the battery and in the service manual, and divide by a thousand. If the listing gives neither the voltage nor the watt-hours, it is not stating the battery capacity, and you should treat the omission as information.

Why airlines care, and why it caps the figure

The watt-hour figure is the one used for transport rules on lithium batteries, which is a large part of why laptop batteries cluster below certain values rather than growing indefinitely. A chassis that could physically hold more battery often does not, because crossing a threshold changes how the machine can be carried. This is a real design constraint and it explains why battery capacity has grown much more slowly than the rest of the specification.

Where the charge goes

In rough order, on a typical thin laptop doing ordinary work:

• The display. Usually the single largest consumer, and it scales directly with brightness. Dropping brightness from maximum to a comfortable indoor level is the largest single change available to you, and it costs nothing.
• The processor package, including integrated graphics. This is dominated by short bursts rather than steady load for most work: a web page that runs a lot of script wakes the processor constantly, which is why one badly behaved tab measurably shortens battery life.
• Radios. Wireless networking under poor signal costs substantially more than under good signal, because the radio transmits harder and retries more.
• Everything else — storage, ports, keyboard backlight — which together are smaller than people assume, with one exception: anything drawing power out of a port to charge another device comes straight off the battery.

Reading the runtime claim

Manufacturers quote runtime from standardised test scripts, at a defined brightness, usually with radios in a defined state, playing back video or running a light office loop. Those conditions are published alongside the claim in the footnotes. A claim without stated conditions is not a measurement.

The useful way to use a runtime claim is as a rank between machines tested the same way, and the useful way to predict your own runtime is arithmetic: take the watt-hour figure, divide by the power draw you actually observe. Every operating system will report current battery discharge rate in watts — on Windows through the built-in battery report or a hardware monitor, on Linux through the power supply class in the system filesystem, on macOS through the system information panel. Watch that figure for five minutes while you do your normal work; the division is your real runtime.

Battery health, and how to read it

A lithium battery loses capacity with charge cycles and with time, and both the design capacity and the current full-charge capacity are stored in the battery's own controller. Every platform exposes them. On Windows, the built-in battery report generates an HTML file that lists design capacity, full charge capacity and cycle count. On macOS, the system information panel reports cycle count and condition. On Linux, the power supply entries in the system filesystem carry design and full charge energy.

The ratio of full-charge capacity to design capacity is the honest health figure. Cycle count on its own is not: two machines with the same cycle count can differ substantially depending on how they were charged and how hot they ran.

This is also the single most important check on a used machine, and the easiest to ask a seller for — it is one screenshot. See the used-laptop checklist.

Charging habits that matter, and ones that do not

Heat and time spent at a very high state of charge are what age a lithium cell. Many manufacturers now ship a charge limit setting that holds the battery below full when the machine lives on a desk, and using it is the one habit with a measurable effect. Full discharges are not required by modern batteries and are mildly harmful. Leaving a machine plugged in is fine on any laptop built in the last decade, because the charging circuit stops; what is not fine is leaving it plugged in, hot, and at full charge in a warm room for years.

Replacement

Check, before buying any laptop you intend to keep for a long time, whether the battery is a serviceable part with a published part number, and whether reaching it requires removing the keyboard. Both facts are in the service manual. A battery is a consumable; a machine that treats it as structural is a machine with a shorter life than its specification suggests.