Temperature and derating: why the rating has a condition attached

A cell rated 35 A continuous is rated 35 A while its surface temperature stays at or below the cut-off the manufacturer used when measuring — commonly 80 °C or 83 °C for high-power cylindrical cells. The current and the temperature are one statement, not two. Remove the temperature and you have a figure that can be printed in larger type and applied to any enclosure.

What the cut-off is, and what it is not

The cut-off temperature is a test condition, not a safety limit and not a target. It defines the point at which the measurement stopped. Operating at the cut-off for a long time is not "within specification" in any useful sense — it is operating at the edge of the assumption the rating was measured under, with no margin for a hot day or a blocked vent.

Derating when the enclosure cannot hold the condition

If the housing cannot keep the cell at the specified temperature at the rated current, the usable current is lower. There is no universal formula — the right answer comes from a thermal measurement on your own enclosure — but the shape of the correction is predictable:

Qualitative direction only — the numbers come from your thermal measurement, not from a table.

The most common mistake is measuring air temperature instead of cell temperature. A pack with 35 °C air can have cells at 70 °C, because the thermal path from the cell to the air is the bottleneck. Put the thermocouple on the cell can, on the side facing the hottest part of the array.

Heat is not evenly distributed

In any array, the cells in the middle have less surface exposed to the outside and receive heat from neighbours on more sides. They run hotter, age faster, and their impedance rises sooner — which makes them hotter still. Over a few hundred cycles the middle of the pack becomes the limiting element even though every cell was identical at build.

This is a design problem with mechanical answers: spacing, thermally conductive potting or spacers, a metal cold plate, or simply fewer cells doing less work each. It is not solved by choosing a cell with a higher current rating, because the higher rating was measured under the same thermal assumption.

Low temperature: a different failure mode

Cold does not damage a cell the way heat does, but it changes what the pack can do. Electrolyte conductivity falls, so resistance rises and voltage sag increases; usable capacity falls well before 0 °C, and at −20 °C it can be a fraction of the room-temperature figure.

Charging below 0 °C is a separate and more serious issue: lithium plating on the anode during charge at low temperature can be irreversible and is a safety concern, which is why most cells specify no charge below 0 °C unless the pack implements a specific low-temperature strategy. A charger that does not know the pack temperature will violate this quietly, every cold morning.

A worked example of the decision

A tool pack needs 40 A continuous from a 5S2P arrangement, so each cell carries 20 A. The shortlisted cell is rated 25 A continuous at an 80 °C cut-off. On paper there is 25% margin.

The enclosure is sealed, the cells sit in a printed carrier with 1 mm gaps, and the measured surface temperature at 20 A in a 35 °C ambient is 74 °C. That is inside the cut-off, but the margin has closed from 25% to almost nothing — and the 80 °C condition assumed a thermal path the carrier does not provide.

Three legitimate responses, in order of cost: add spacing or a conductive spacer to improve the path and re-measure; accept a lower continuous current in the specification and state it; move to a higher-current cell so the same 20 A sits further from its cut-off. What is not legitimate is quoting 25 A because the data sheet did, because the number was measured in a condition the product does not reproduce.

What to write into the specification

• The ambient range the product must work in, stated as a range rather than a typical value.
• The cell surface temperature the design is expected to reach at maximum continuous current, with the measurement point named.
• The derated current at the top of that ambient range — not just the room-temperature figure.
• The charge temperature window, and what the product does outside it (block charge, reduce rate, or heat first).
• The test that demonstrates all of the above, so the claim can be re-checked after a design change.

A specification written this way survives a design review. A specification that quotes the data sheet current with no condition does not, because the first reviewer who asks "at what temperature" has no answer to check against.

Questions this guide does not answer in passing