HV Battery Died with 7 miles range left showing on Range display

[HMHM]

If you can't understand that 0 means 0, then you shouldn't be driving.

Sorry to point out that if you can't discuss properly, with an open mind and without personal attacks, you will be much better off keeping your GREAT suggestions to yourself.

We are not talking math and science here. We are simple talking about battery behavior and the BMS. As someone recently said, there is some "black-magic" (inadequacies) in the system due to numerous unaddressed variables and what-if scenarios, which have not been properly qualified by Tesla yet.

 

[wk057]

Sorry to point out that if you can't discuss properly, with an open mind and without personal attacks, you will be much better off keeping your GREAT suggestions to yourself.

We are not talking math and science here. We are simple talking about battery behavior and the BMS. As someone recently said, there is some "black-magic" (inadequacies) in the system due to numerous unaddressed variables and what-if scenarios, which have not been properly qualified by Tesla yet.

As the only person outside of Tesla with any real knowledge of the inner workings of the Tesla BMS: Respectfully, you have no idea what you're talking about and should probably stop. Also, I've made no personal attacks, so probably best to not try to lie about such things.

Tesla's BMS tracks more variables than any other BMS I've examined. It's even able to accurately calculate many things that it has virtually no way to directly measure, either because doing so is actually impossible or because the hardware has no way to do so because it'd be prohibitively expensive or complex.

It is by far the most accurate and well thought out system available to the public by far based on everything I've seen over the years. There's no "black magic" going on here. It's the result of well over a decade of real world research and development put into action in what's by far the best BMS hardware and software on on the market.

I'm quick to jump on Tesla's case when they do something shady, cut corners, or otherwise do something wrong. But I'm also going to give them credit where credit is due, and their BMS is something that's entirely under rated and deserves substantial kudos. So I'll definitely defend them on this particular point, because they've more than earned it.

 

[HMHM]

I'm quick to jump on Tesla's case when they do something shady, cut corners, or otherwise do something wrong. But I'm also going to give them credit where credit is due, and their BMS is something that's entirely under rated and deserves substantial kudos. So I'll definitely defend them on this particular point, because they've more than earned it.

That is exactly what I'm trying to do. Just trying to point out some seemingly clear shortcomings. Notwithstanding the degradation level, the BMS which takes a 7 mi and 4 mi remaining SoC on the display but calculates it as "0" SoC (triggers a shut-off sequence) is definitely lacking.

 

[dafish]

Wk057:

A pleasure to have such insight available. I’m sure we’d all appreciate a deeper understanding of how a BMS works. Would you be willing to take a crack at a primer for us?

Since I’m imposing, may I ask a few pointed questions?

1). I’ve understood capacity is difficult to extrapolate from voltage, yet voltage extremes are strong cell damage boundaries that must be honored/avoided.

a. Is this reasonably accurate?
b. is this part of the challenge a BMS has in predicting range?
c. I assume things like cell temp, driving behavior, and current drain averages are factors, but cell health seems like a real wild card. How does a BMS manage, even identify, cell health vs pack, module, etc?

2). What factors determine a car stopping, and is it conceivable such limit, let’s say minimum voltage, can be hit faster than a BMS predicts? if so, might a failing or degraded cell masked by overall pack health be indicated?

thanks for anything you care to share.

-d

 

[ewoodrick]

That is exactly what I'm trying to do. Just trying to point out some seemingly clear shortcomings. Notwithstanding the degradation level, the BMS which takes a 7 mi and 4 mi remaining SoC on the display but calculates it as "0" SoC (triggers a shut-off sequence) is definitely lacking.

No, that's not what you are trying to do.

What you are trying to do is to take a SINGLE CASE and extrapolate it into millions of vehicles.

Sure, there may have been one or two cases where a car didn't make it to 0. But that's happened in other cars as well.

The moral of the story is that you should NEVER drive your car near 0. That has nothing to do with a Tesla. In an ICE, first it can guess wrong as well, but even worse, you can suck in debris. Did you happen to know that gas tanks have a reserve in them? Specifically, to try to keep you from sucking in crap?

In the VAST MAJORITY of situations, the car will go to and beyond 0 miles remaining.

But in the VAST MAJORITY of cases, drivers will never know because they aren't going to take it anywhere near 0.

So yes, "once in a blue moon" the car can erroneously show a higher than actual remaining charge. Are you happy now? Does anything need to be done about it? No.

 

[wk057]

Seems someone doesn't like reality and decided to spam disagree unrelated posts of mine. lol.

Capacity is not determined from voltage alone. Tesla's BMS tracks a LOT of variables besides voltage to be able to accurately track capacity, including highly accurate usage/regen/charge data at high frequencies. It can pretty much count the mWh in real time, and use that along with all of the other data points available to come up with a very good estimate of capacity available at any given time.

Tesla's BMS can identify metrics about each individual cell group in the pack (usually 96 groups) with their cleverly designed hardware. They can detect a failing cell group before anyone would notice it from voltage alone.

There are hard limits the BMS will never allow the system to go past (which vary by pack), and can result in a shutdown at any time should some failure of that kind occur. For example, there is a low voltage limit that is allowed under load, which adjusts as SoC lowers. In fact, you can reach lower voltages under load at 50% SoC than you're safely allowed at say, 10% SoC, hence the power limits that start to get put in place at low SoCs.

There's nothing "masked by overall pack health"... the BMS knows the health of every single cell group and acts accordingly. You can't hide anything from it or get anything by it. Trust me, I've tried all sorts of ways to appease the Tesla BMS in failure situations without success. It's too good.

If the car shuts down with miles showing on the dash, it's either because of a calibration issue (BMS never acquired enough data in the first place to accurately estimate range, usually only an issue on very new cars with shallow cycling but have seen minor 1-2% errors creep in on some cars without any actual failures), or because of an actual issue that causes the BMS's estimates to be based on pre-failure data and not post-failure data.

Either way, you're not going to find a more accurate BMS out there. Tesla's BMS provides to most possible range out of the cells. And because of that, 0 means 0. You can't expect range below 0.

 

[MP3Mike]

Seems someone doesn't like reality and decided to spam disagree unrelated posts of mine. lol.

Yep, only the jokes on them as the "disagree" rating has no impact on the reaction score. (So the only thing it really accomplishes is to make them look silly/petty.)

 

[Rocky_H]

Seems someone doesn't like reality and decided to spam disagree unrelated posts of mine. lol.

He did it on mine too.

 

[HMHM]

Did you happen to know that gas tanks have a reserve in them?

Interesting! Only Teslas are not supposed to have a reserve capacity per some of the self described best minds (Bullies) on this forum. Elon says otherwise!

 

[ewoodrick]

Interesting! Only Teslas are not supposed to have a reserve capacity per some of the self described best minds (Bullies) on this forum. Elon says otherwise!
View attachment 951424

You have me totally confused. Yes, I know this. That's what I thought I was saying. Teslas have a reserve.

But yes, there can be times in which the BMS just guesses wrong.

 

[Rocky_H]

Teslas have a reserve.

That might be better phrased as "Sometimes some of them might." It's not really a given.

 

[dafish]

Thanks WK!

 

[wk057]

Interesting! Only Teslas are not supposed to have a reserve capacity per some of the self described best minds (Bullies) on this forum. Elon says otherwise!
View attachment 951424

Elon actually doesn't know what he's talking about here. (And I do believe I already debunked this back when that tweet was posted, too.)

He's correct that driving until the car stops moving won't hurt the pack (as long as you recharge relatively soon afterwards), but there is no "reserve". 0 still means 0.

There is an internal variable called "buffer" which is used as a fudge factor during discharge to help account for variations early in a pack's life, which is the time you're most likely to actually have usable capacity below zero (but not always and not guaranteed). This is probably what Elon is mistakenly referring to as a "reserve", but it's not actually handled that way. The rated range shown is adjusted on the fly during discharge and any amount of this buffer that turns out to be available for use is reflected in the energy shown to be available (rated miles) during the drive... again, targeting that 0 means 0. This is actually part of the reason for inconsistent usage metrics in the trip meter on some cars when driven differently to low SoCs on multiple trips.

There's also an internal variable (no clue a name, it's not exposed anywhere in English) which is the true bottom buffer which is the completely unusable portion of energy left at the very bottom of the capacity that is 100% untouchable by anything outside of the battery pack. It's there to give the BMS enough energy to run for a few months should the pack not be charged for some reason after reaching 0 miles.

But everything in the BMS targets roughly 500 Wh of energy below 0 miles showing on the dash as the actual usable capacity. It has that tiny bit of margin of error by default. It can increase that value when it's more and more uncertain about actual capacity, shifting some miles to below zero as calibrations go out of whack from less ideal driving/charging habits.... but it doesn't actively try to keep any kind of buffer or reserve below 0 miles. There's no reserve whatsoever designed into the system.

And that's exactly as it should be.

 

[jjrandorin]

Yep, only the jokes on them as the "disagree" rating has no impact on the reaction score. (So the only thing it really accomplishes is to make them look silly/petty.)

(moderator note)

We dont discuss this change much, but yes its true that the algorithm was changed by the site owners to remove disagree from reaction scores due to people weaponizing them.

With that being said (speaking in general terms here, I wont be commenting publicly on specific cases) when a moderator sees what they believe to be review score bombing, there will usually be some sort of action taken by the moderator.

 

[wk057]

(moderator note)

It looks like this particular person backtracked on a bunch of them (or you guys removed them?) anyway. And if they don't count anymore, then that's fine. I'm not sure that's the best approach (there are posts that need to be "Disagree"'d into oblivion for sure), but guess that's how things are going these days (looking at you, YouTube dislikes...)

Thanks

 

[ewoodrick]

That might be better phrased as "Sometimes some of them might." It's not really a given.

A Tesla has 4 wheels and tires.

Do I have to add a disclaimer that they may be stolen? And only have 2?

This is the thread where the original poster didn't just run out going to a charger, they ran out after numerous occasions of running the car to near zero (and beyond, if I'm not mistaken) Can I drive and charge in such a manner to completely confuse the BMS? Sure.

But is it common, no. Is it rare? It's probably even beyond rare, into the extremely rare stages. And it only happens to people who run to empty. If you were in an ICE, and it said 7 miles left, what would you expect that probabilities that you are going to make it to a gas station 7 or 8 mile away going to be?

It is a rare edge case that isn't worth mentioning.

I think I can easily say that there's a higher probability that you will have a crash getting to the charger at 7 miles remaining than running out of juice.

 

[dafish]

Ewood:

I mean to keep thing on balance, I can say:
* I've never had any ICE, and I've owned a lot, not run below 0 miles left on DTE.
* I don't know that running an EV car to very low, even zero shoud make a BMS inaccurate. In fact, I believe we're told full range excursions help them. Now that's not fact, it's more rumor I've acquired reading this forum, but it's a pretty commonly expressed,

Mind you my ONLY position on this thread is to learn from WK's coaching, so take anything I say with a grain of salt.

 

[dafish]

@wk057

Sir (I just assume), if we accept BMS's have, by their nature, so inaccuracies, (however minor that might be), what should I think of charge level accurance. I ask for this reason: I rarely supercharge, almost never exceed 55% SoC, and even traveling don't expect to charge beyond 80%. Given what little I know of range prediction I'm not giving an average BMS a lot of accuracy. But if I charge to say 75%, and I really on got 65%, well, that's not good.

So: How accurate can I generally assume the percent SoC level is post charge? Is say 75% shown within a few percent or reality, or could it be off by 10%? And how soon might I know that?

Thanks again for helping us!

-d

 

[flixden]

How do I get to this "battery health" screen?
Been trying for an hour and can't find it.
Any help would greatly be appreciated!

 

[wk057]

If we accept BMS's have, by their nature, so inaccuracies, (however minor that might be), what should I think of charge level accurance. I ask for this reason: I rarely supercharge, almost never exceed 55% SoC, and even traveling don't expect to charge beyond 80%. Given what little I know of range prediction I'm not giving an average BMS a lot of accuracy. But if I charge to say 75%, and I really on got 65%, well, that's not good.

So: How accurate can I generally assume the percent SoC level is post charge? Is say 75% shown within a few percent or reality, or could it be off by 10%? And how soon might I know that?

Thanks again for helping us!

-d

The SoC exposed to the user is the BMS's best estimate of usable capacity. It adjusts this on the fly, and under normal circumstances you can be pretty well assured that if it says you have x% available, you have x% available.

If it's inaccurate for some reason, there's generally no way to know until you run the pack down to a very low SoC and the BMS quickly realizes the issue.

It's also possible for things to make these calculations completely useless, but this is generally the result of a third party modifying the pack in a way that makes it impossible for it to make accurate estimates (replacing modules with mismatched ones, snipping cell fuses, etc)... in which case shutting down with even 100+ miles on the dash wouldn't be unexpected. (Yes, this is a PSA to not let anyone butcher your battery pack.)