A cell chosen purely on capacity will meet its runtime target on the bench and run hot, sag under load and age early in the product. Capacity answers "how much energy"; C-rate answers "how fast can I take it out without hurting the cell". Most specifications answer only the first.

What C actually means

1C is the current that would discharge the rated capacity in one hour. For a 1000 mAh cell, 1C is 1000 mA; 0.5C is 500 mA; 2C is 2000 mA. It is a normalised unit, which is what makes it useful — a 2C load means the same thing to a 200 mAh wearable cell and a 5000 Ah pack.

The reason engineers get caught is that rated capacity is measured at a low rate, typically 0.2C. If you then run the cell at 2C, you will not get the rated capacity out of it, because more of the energy leaves as heat in the cell's own resistance.

Continuous versus pulse: two different ratings

A datasheet may quote a maximum continuous discharge of 1C and a maximum pulse of 3C. These are not interchangeable, and "pulse" is meaningless without a duration and a duty cycle. A 3C pulse rating usually means something like a few seconds at low duty — not a 3C burst every 100 ms forever.

If your load is bursty — a cellular modem, a motor, a radio — characterise it properly: peak amplitude, pulse width, duty cycle, and the worst-case sequence (a modem registering on a cold network is not the same as one idling).

Where the heat comes from

Resistive heating in the cell goes as I²R. Doubling the current quadruples the heat. This is why a design that is comfortable at 1C can be thermally impossible at 2C even though the cell is "rated" for it — the rating is about the cell surviving, not about your enclosure staying below a skin-contact limit.

Two compounding effects make the worst case worse than the arithmetic suggests:

  • Resistance rises as the cell ages. The end-of-life cell generates more heat at the same current than the new one you prototyped with.
  • Resistance rises as the cell gets cold. Cold plus aged plus peak load is the corner your thermal and voltage budgets both have to survive.

Our piece on skin-contact temperature covers the enclosure side of this; thermal runaway covers what happens when the margin is gone entirely.

C-rate and cycle life trade against each other

Higher sustained rates shorten cycle life, and they do it in a way that does not show up in a short bench test. If you are quoting a cycle-life number from a datasheet measured at 0.5C and running the product at 1.5C, the field number will be lower — see reading a cycle-life curve honestly for why the conditions matter more than the headline.

Three ways to lower the C-rate a cell sees

  1. Use a bigger cell. The same absolute current is a lower C-rate on a larger cell. Often the cheapest fix if the enclosure allows it.
  2. Put cells in parallel. Two matched cells share the current, halving the rate each sees. Bring the matching and protection questions in series and parallel configuration with you.
  3. Flatten the load. A bulk capacitor across a short high-amplitude burst can cut the peak the cell sees substantially, and it is far cheaper than a bigger battery.

A working method

  1. Capture the real current trace, including worst-case bursts.
  2. Compute peak and continuous C-rate against the candidate cell's rated capacity.
  3. Check both against the datasheet's continuous and pulse ratings, with the pulse conditions stated.
  4. Re-check at end-of-life resistance and at your minimum operating temperature.
  5. If either check fails, apply one of the three fixes above rather than hoping the rating is conservative.

Steps 4 and 5 are the ones usually skipped, and they are where field failures come from.