"My battery is dying" is one of the most common claims in e-bike support, and one of the most frequently wrong.

Three different things produce the same symptom — shorter range — and only one of them is capacity loss: State of charge (SoC) — how full the pack is right now. Trivial, but a partially charged pack explains a surprising number of "my range collapsed" reports. Capacity loss (SoH) — how much energy the pack can still hold compared with when it was new. This is what people mean by degradation. Voltage sag — the pack delivering less energy under load because its internal resistance has risen, or because it is cold. The capacity may be fine; the delivery is not.

You cannot fix the third by replacing the first, and you cannot diagnose any of them by guessing. Here is how to measure properly, without lab equipment.

What "state of health" actually means

State of health is a ratio:SoH = current usable capacity ÷ rated capacity

If your pack is rated at 720 Wh and it now delivers 590 Wh from full to cutoff, its SoH is about 82%.

Two caveats that matter in practice: Rated capacity is not always usable capacity. Manufacturers often quote nominal Wh; the BMS may reserve a buffer at the top and bottom to protect the cells. Your measurement reflects the usable portion, which is the number that affects your day. SoH is temperature-dependent. A pack measured at 5°C will look worse than the same pack measured at 22°C. Any comparison is only valid if the conditions match.

Method 1 — The controlled range test (free, repeatable)

This is what most riders should do first, because it measures the thing they actually care about: how far the bike goes.

The protocol: Pick one route and ride it the same way every time. Same direction, same assist level, same rider weight, same tyre pressure. Start from a full charge, on a pack at room temperature. Not straight off a cold night in a garage. Ride to the same cutoff point — either the same low-battery warning or the same voltage under load, whichever your bike reports consistently. Record distance and, if your display shows it, watt-hours consumed. Repeat three times and average. A single run is weather, not data. Write the number down with the date. Without a baseline from when the bike was new, you have a number with nothing to compare it to.

The output is Wh per km (or per mile). If that figure has risen by 15% since your baseline, under the same conditions, your usable capacity has probably fallen by a similar amount.

What this method is good at: real-world impact, zero equipment. What it is bad at: precision. Wind, temperature, and route changes swamp small differences.

Method 2 — Measure what goes in (charger-side energy meter)

A plug-in energy meter between the wall socket and the charger records the watt-hours consumed to refill the pack.

Procedure: discharge to a consistent cutoff point, then charge to full, and read the meter.

The catch: the meter measures energy at the wall, and the charger is not perfectly efficient — some energy becomes heat. That means the wall figure will always be higher than what entered the cells. Charger efficiency varies between models, so you cannot simply convert one to the other with a fixed number.

What you can do is compare against your own baseline. If the same charger, on the same pack, drew 780 Wh when the bike was new and now draws 640 Wh to cover the same state-of-charge range, that is a meaningful signal — the charger's inefficiency is a constant in both measurements.

Good for: a rough trend with cheap equipment. Not good for: an absolute SoH percentage.

Method 3 — Measure what goes out (inline watt meter)

The most accurate consumer method, and the one worth doing if you have a warranty claim to support.

Fit a watt meter between the battery and the controller — many are plug-and-play with the right connector — and record the watt-hours drawn on a full-to-cutoff run.

This measures energy actually delivered to the bike, which is closer to true usable capacity than the wall reading. Do the run at a steady temperature and a steady assist level, and repeat it.

Good for: a defensible number. Not good for: anyone who is not comfortable working with battery connectors. If that is you, use Method 1 and let a shop do Method 3.

Method 4 — What the BMS or app reports

Many packs report cycle count and, sometimes, an estimated capacity. Useful context, but treat these numbers as estimates: Cycle count is usually a count of charge events, not full equivalent cycles, and definitions vary by manufacturer. Reported capacity is often calculated from voltage and coulomb counting and can drift. A BMS that reports 100% health on an eight-year-old pack is telling you about its algorithm, not about your cells.

Use BMS data as corroboration, not as the primary measurement.

What does not work: voltage alone

A multimeter reading of pack voltage tells you almost nothing about health, because lithium-ion cells hold a nearly flat voltage across most of their usable range. A worn pack and a healthy pack can read almost identically at rest.

Voltage is still useful for two things: confirming a pack is not deeply discharged, and spotting a dead cell group through an abnormal reading. It is not a capacity test.

Accuracy, honestly compared

表格

Reading your result

Those bands are conventions, not physics. What matters is your warranty's stated threshold — read it, because "80% after X cycles" and "70% within two years" are very different promises.

What actually causes the fade

Heat. Charging or discharging a hot pack accelerates ageing more than any other single factor. Sitting at 100%. A pack stored full, especially warm, ages faster than one stored around half. Sitting at 0%. Deep discharge damages cells and can leave a pack unrecoverable. High current. Sustained high assist on a small pack, or repeated fast charging, adds stress. Time. Cells age chemically whether used or not.

The practical routine: charge to full before a long ride, store around half for anything longer than a week, never charge a frozen or hot pack, and avoid leaving it plugged in for days.

FAQ

How do I know if my battery is degraded or just cold?Test at room temperature. Cold reduces available capacity and increases voltage sag temporarily; a pack that recovers its range indoors is not degraded, it was cold.

Can I test capacity without riding?Not accurately with household tools. A bench discharge test with a programmable load works, but it is a shop procedure, and repeatedly deep-discharging a pack to test it does the pack no favours.

How often should I test?Once a year is enough for a commuter, or any time you notice a step change in range.

Is a 20% loss normal after two years?Depends on how the pack was treated and what the manufacturer rates it for. Heat and full-charge storage are the usual accelerants; a well-treated pack often does better.

Does a second battery fix degradation?It fixes range, not health. Two packs sharing the load also age more slowly than one pack doing all the work, which is why delivery riders often run a pair.