What does a commercial EV use for remaining capacity?

SLS BATTERY AUCTIONS are now live at www.batteryhookup.com EV, Lithium, LifePo4, Inverters & More


Jan 5, 2022
This could go in the E-vehicle thread, but it could apply to battery usage in general.

I used this in my e-motorcycle build to have a heads-up display for remaining battery life. I know it was cheap, I didn't need it to be exact. Reviews stated that it would need a few charge-dischage-charge cycles before its capacity meter was any accurate. Problem is, it's still not even close. I'm using the BMS bluetooth app prior to long-ish rides to check actual capacity remaining. The good news it that the voltage display is quite accurate for its price point, within .5V, so I can use my big ol brain to guess SOC, and its helpful to me seeing voltage sag during a ride. It's just that the capacity function is incredibly unreliable, and it got me to thinking, how do the big companies do it?

If I had to guess, I would say that each manufacturer uses some sort of proprietary program that uses a combination of factors to estimate remaining capacity: current SOC, miles traveled since last charge, amps delivered on last charge, and estimated miles remaining based on this info, and then translate that info into a "miles remaining" display on your dashboard. Not simply SOC. Yeah, it would be nice to have a better, easy-to-read gauge on my motorcycle so that I could get accurate remaining capacity info without having to open an app on my phone and connect via bluetooth every time. But I can make do with current SOC readings and mileage based on range tests. I'd rather not buy another cheap product that's just gonna give me a similarly inaccurate reading. It just got me thinking about how other EV's do it.


Active member
Dec 15, 2018
It's an interesting topic.

I believe that an accurate shunt tracking coulombs in/out is one of the preferred methods for SoC reporting - and used by the highly regarded Victron Smart Shunt? I could be wrong.

My shunts/software (Batrium and Midnite Classic) that track SoC are not accurate. So I've been tracking discharge ah/v for a couple of years now. The powerwall is bounded to operate between 4.0v -> 3.54v / cell - which is in the middle/more-linear portion of the lithium-ion discharge curve. For example, my current average for Sep is 238ah/v. It varies because my average peak charge voltage per month varies and extreme load differences can pull down the voltage a bit - which changes the portion of the discharge curve and the voltage range used for the metric. But it's interesting to me.

The original reason I started recording this was some half-baked idea to detect battery bank capacity degradation from year to year as the cycles mount. However, it doesn't seem precise enough for small % changes.

The other use is operational capacity estimations. For example, last night at 10pm the current voltage was 51.3v and I have a fixed 49.5v cut-off. 51.3v - 49.5v = 1.8v difference. 1.8 * 238ah/v = est 428ah till 49.5v cut-off. 428ah * 50.5v = 21.6kwh remaining to reach 49.5v. At 3kw/hour load, that's 7.2hrs of remaining operation - e.g. 5:30am cut-off. In other words, it's not SoC that I focus on so much as how much operational capacity do I have left.

EVs also show estimated mileage (operational capacity) left - and like you, I wonder what they use to generate the info.
Last edited: