C-rate is a normalisation, nothing more. It expresses current as a multiple of the cell's rated capacity so that engineers can compare a 3 Ah cylindrical cell with a 300 Ah prismatic one without converting units in their heads every time.
The definition: 1C is the current, in amperes, numerically equal to the rated capacity in ampere hours. A 5 Ah cell at 1C draws 5 A. The same cell at 2C draws 10 A, and at C/5 or 0.2C it draws 1 A.
Worked examples
| Case | Rated capacity | Current or power | C-rate | Nominal duration |
|---|---|---|---|---|
| Cylindrical cell, moderate discharge | 5 Ah | 10 A | 2C | 30 min |
| Prismatic LFP cell, grid duty | 280 Ah | 70 A | 0.25C | 4 h |
| Power tool pack | 2.5 Ah | 50 A | 20C | 3 min |
| EV pack, motorway cruise | 77 kWh | 20 kW | ~0.26C | ~3.9 h |
| EV pack, hard acceleration | 77 kWh | 300 kW | ~3.9C | ~15 min |
| EV pack, 350 kW DC charge peak | 77 kWh | 350 kW | ~4.5C | n/a, peak only |
| Grid BESS, two-hour system | 200 MWh | 100 MW | 0.5C | 2 h |
| Grid BESS, frequency response | 20 MWh | 20 MW | 1C | 1 h |
Two things to take from the table.
For packs quoted in kWh rather than Ah, dividing power in kW by energy in kWh gives the C-rate directly. A 300 kW draw from a 77 kWh pack is 300 divided by 77, which is roughly 3.9C. This works because both numerator and denominator carry the same voltage term, and it saves converting to ampere hours.
The nominal duration column is theoretical. It assumes the full rated capacity is available at that current, which it is not.
Why the duration never quite matches
Delivered capacity falls as current rises. Concentration gradients build in the electrolyte and in the solid particles, ohmic and charge-transfer losses drop the terminal voltage sooner, and the cell reaches its cut-off voltage with lithium still in the anode. In lead acid this rate-capacity effect is described by Peukert's equation. In lithium-ion the effect is much smaller but still present, and the mechanism is better described by porous-electrode models of the kind Doyle, Fuller and Newman set out.
Practical consequence: a cell rated at 5 Ah when tested at 0.2C might deliver 4.6 Ah at 3C. If a datasheet quotes capacity without quoting the test current, the number is decorative.
Part of the missing capacity returns on rest, because concentration gradients relax. That is why a device that shuts down under load will often run again briefly after a pause.
The connection to grid storage that usually gets missed
In the stationary storage market, systems are sold by duration: a two-hour battery, a four-hour battery, an eight-hour battery. Duration and C-rate are reciprocals of one another, so:
- One-hour system, 1C
- Two-hour system, 0.5C
- Four-hour system, 0.25C
- Eight-hour system, 0.125C
This matters commercially as much as technically. A four-hour system running at 0.25C keeps its cells in a gentle thermal and electrochemical regime, which is part of why phosphate cells achieve high cycle counts in that duty. A one-hour frequency-response system running at 1C, with frequent direction changes, sits in a harder regime and will show more resistance rise for the same energy throughput. When a developer compares warranty terms across durations, the C-rate difference is doing much of the work behind the numbers.
Continuous, peak and pulse ratings
A datasheet usually gives three current ratings and they are not variations of the same thing.
Continuous rating is what the cell can sustain indefinitely within its temperature limits, on a defined cooling assumption. Change the cooling and the number changes.
Peak or pulse rating applies for a stated duration, often ten or thirty seconds, from a stated starting temperature and state of charge. A 10C pulse rating means nothing without the duration and the conditions.
Charge ratings are almost always lower than discharge ratings, and asymmetric for good physical reasons. On discharge, lithium leaves the graphite. On charge, it has to intercalate into it, and if it cannot do so fast enough it plates as metal. That asymmetry is why fast charging is limited by the anode and why charge C-rate limits tighten sharply as temperature falls.
Typical figures by application: energy-optimised EV cells, whether NMC or LFP, sit around 1C continuous discharge, with fast-charge designs accepting 3C to 4C charge over part of the state-of-charge window and the most aggressive production cells reaching 6C or more. Grid-storage LFP cells run at 0.25C to 0.5C by design. Lithium titanate (LTO) cells charge and discharge at 5C to 10C continuous, because the anode operates near 1.55 V versus lithium and cannot plate. Power-tool and drone cells discharge at 15C to 30C, at a substantial cost in energy density.
Heat is the real constraint
Ohmic heat generation is proportional to the square of current multiplied by internal resistance. Doubling the C-rate roughly quadruples the ohmic heat. Add the entropic and reaction-heat terms and the total is more complex, but the square-law term dominates at high rate.
That single relationship explains most C-rate engineering:
Why high-rate cells have thinner electrodes, more tabs and larger current-collector cross-sections, and therefore lower specific energy.
Why cell format matters: a large-format prismatic cell has a longer thermal path from its centre to its cooled surface than a 21700 cylindrical cell, so it develops larger internal gradients at the same C-rate, and gradients cause uneven ageing within a single cell.
Why the C-rate a pack can sustain is almost always below what its cells could sustain in isolation. The binding constraint moves to the cooling system, the busbars, the contactors and the fuse.
Three mistakes worth avoiding
Confusing C-rate with power. C-rate is dimensionless, power is in watts. Two systems can both run at 0.5C and differ by three orders of magnitude in power.
Assuming C-rate capability at cell level transfers to system level. It does not, and the gap is where most integration disappointment lives.
Comparing cycle life figures without checking the C-rate they were measured at. A cell quoted at 4,000 cycles at 0.5C and a cell quoted at 4,000 cycles at 1C are not equivalent products, and the second one is the better cell.
Informational and educational content only. Not professional, financial, legal, or engineering advice.