Wiki Sudden Loss Of Range With 2019.16.x Software

[Chaserr]

It's rare and you should see this

 

[Guy V]

Ok ok! We're getting somewhere! We're half way there! Now let's keep going!

The link above has someone saying Tesla asked them if it was parked on an incline. Got that. (I wasn't even disputing that the question was asked.)

Now, finish it up. Link to the part where Tesla said they asked this because they thought there might be a coolant leak.

LOLOL, are you suggesting it had something to do with window washer fluid? Seriously, what else fire-related could be affected by parking on an incline than fluid, and what other than battery fluids could possibly matter???

 

[lightningltd]

So.. What is X and Y and why would it be so important to mitigate them?
Since all I get from Tesla is "All is well" and "Nothing to see here, move along" (with a wave of the hand)...
This is simply MY speculation based on science and things I have learned here and some simple deduction.
I may be off on the specifics, since I am by no means an expert. I am just a dumb IT guy who is trying to make sense out of the situation and get Tesla to fix my damn battery. I have no idea what the overall voltages are supposed to be...

X (Critical, emergency action needed):
The only critical issue I can think of for X would be a shorted cell or cells (dendrites). Since BMS cannot see each individual cell and its temperature, during IDLE times (not charging, car off) I would have the BMS look for thermal variations between modules over a threshold along with a module discharging at a much faster rate when not in use. This extra heat and voltage drop would be abnormal, since the module should be cooling along with the other modules.
Apparently they did not detect X (that we know of) and that's a good thing. If it WAS detected, it is likely that the BMS would shut down the systems and notify the driver and Tesla that immediate service is needed.

Y (Not critical, but mitigation needed):
In my opinion, Y is the loss of a certain percent of individual cells in the string, either dead cells or very weak ones. Since the BMS cannot see individual cells, it would need to look at each module and the system as a whole.
If I were to try to detect this, I would look at the total voltage of each module, compared to the baseline after a load has been applied (after driving). If the voltage of certain modules is lower than the others, or if the voltage as a whole is a certain percent lower than expected, then I would assume some cells have dropped out or malfunctioned and are not holding or accepting a charge, or are at a weakened state.
Why is this important? Because when we charge and discharge, we charge and discharge the entire string as a whole. If the string is charged at 400 volts and all is well, then great. But what if 10-15% of the cells are not taking the charge? We are then giving 400 volts (or whatever it is) to less cells, resulting in potentially overcharging the remaining cells. To the system, it would still see 400 volts at the end (or whatever the voltage should be).
This would add stress to the remaining cells and the battery as a whole, possibly resulting in creating condition X. How would I remediate this? I would reduce the overall voltage during charging, and have BMS set to not exceed the new calculated voltage for the charging. This would make each cell appear to be getting less than full voltage at 100% charge because again, we cannot see the actual voltage of each cell, just each module.
I find it interesting that Tesla took away the ability to see what the charge amperage and voltage are on the screen and only show us the KW and miles/hr (or km) of charging. Its too bad, since that would help us see what is going on.

Just my speculation, but food for rational thought and educated FRIENDLY debate.
I will leave it up to those who have the tools to measure such things to validate or invalidate my THEORY.

(edited to say cells when I mean individual cells)

 

[Battpower]

Since BMS cannot see each individual battery and its temperature,

You mean cell? Like AA battery?

 

[Chaserr]

This is an excellent hypothesis. Dendrites being a worst case, overcharging being a lesser one... and slower charging would help because you don't want those broken "strings" heating up more than the rest et cetera.

@Battpower He is right, Tesla's BMS is fairly limited. There are thousands of cells and it can only look at them in large groups, so "dead cells" might not have been detected. It's a valid and well thought through hypothesis that fits the data.

There was a lot of talk about overcharging months ago and someone even had a battery that was at 4.2v well under 100%. Broken cells in those "strings" might make it overcharge some cells which is very bad.

 

[Battpower]

So.. What is X and Y and why would it be so important to mitigate them?
Since all I get from Tesla is "All is well" and "Nothing to see here, move along" (with a wave of the hand)...
This is simply MY speculation based on science and things I have learned here and some simple deduction.
I may be off on the specifics, since I am by no means an expert. I am just a dumb IT guy who is trying to make sense out of the situation and get Tesla to fix my damn battery. I have no idea what the overall voltages are supposed to be...

X (Critical, emergency action needed):
The only critical issue I can think of for X would be a shorted battery or batteries (dendrites). Since BMS cannot see each individual battery and its temperature, during IDLE times (not charging, car off) I would have the BMS look for thermal variations between modules over a threshold along with a module discharging at a much faster rate when not in use. This extra heat and voltage drop would be abnormal, since the module should be cooling along with the other modules.
Apparently they did not detect X (that we know of) and that's a good thing. If it WAS detected, it is likely that the BMS would shut down the systems and notify the driver and Tesla that immediate service is needed.

Y (Not critical, but mitigation needed):
In my opinion, Y is the loss of a certain percent of individual batteries in the string, either dead cells or very weak ones. Since the BMS cannot see individual cells, it would need to look at each module and the system as a whole.
If I were to try to detect this, I would look at the total voltage of each module, compared to the baseline after a load has been applied (after driving). If the voltage of certain modules is lower than the others, or if the voltage as a whole is a certain percent lower than expected, then I would assume some cells have dropped out or malfunctioned and are not holding or accepting a charge, or are at a weakened state.
Why is this important? Because when we charge and discharge, we charge and discharge the entire string as a whole. If the string is charged at 400 volts and all is well, then great. But what if 10-15% of the cells are not taking the charge? We are then giving 400 volts (or whatever it is) to less cells, resulting in potentially overcharging the remaining cells. To the system, it would still see 400 volts at the end (or whatever the voltage should be).
This would add stress to the remaining cells and the battery as a whole, possibly resulting in creating condition X. How would I remediate this? I would reduce the overall voltage during charging, and have BMS set to not exceed the new calculated voltage for the charging. This would make each cell appear to be getting less than full voltage at 100% charge because again, we cannot see the actual voltage of each cell, just each module.
I find it interesting that Tesla took away the ability to see what the charge amperage and voltage are on the screen and only show us the KW and miles/hr (or km) of charging. Its too bad, since that would help us see what is going on.

Just my speculation, but food for rational thought and educated FRIENDLY debate.
I will leave it up to those who have the tools to measure such things to validate or invalidate my THEORY.

Likewise, non-expert.

I can see that since individual cells in bricks are not monitored, and the bricks are in series, it must get harder to keep them balanced once a bunch of cells drop out. If the BMS can keep them within spec as muiltiple cells drop out I don't know, but I expect it pushes things to the limit.

 

[aerodyne]

This is an excellent hypothesis. Dendrites being a worst case, overcharging being a lesser one... and slower charging would help because you don't want those broken "strings" heating up more than the rest et cetera.

I might have this problem developing.

String one, module 2, is 40ma out of balance. Even after a moderate charge and drive

 

[Battpower]

This is an excellent hypothesis. Dendrites being a worst case, overcharging being a lesser one... and slower charging would help because you don't want those broken "strings" heating up more than the rest et cetera.

@Battpower He is right, Tesla's BMS is fairly limited. There are thousands of cells and it can only look at them in large groups, so "dead cells" might not have been detected. It's a valid and well thought through hypothesis that fits the data.

Yes.

I was just getting confused by his cells being called batteries in some places I think!

 

[Battpower]

is 40ma out of balance

40 mV?

So like 10x greater than ideal?

 

[lightningltd]

You mean cell? Like AA battery?

Correct. Edited the original to make it clearer

 

[aerodyne]

40 mV?

So like 10x greater than ideal?

From everyone I have asked, yes.

 

[Guy V]

This is the all-encompassing problem with Chaserr’s participation in this matter. Repeated, fanatical insistence on making factual claims that are impossible to substantiate.

Things he could not possibly know.
Claims of future actions and outcomes that are impossible to predict.
Things that are extreme exaggerations of available evidence and observation.
Things that are quite simply lies.

I admit, he seems intelligent enough to me to know that he’s doing this, so I’m left with the only conclusion that he’s being deliberately obtuse in an effort to bolster his legal claims and further his self interest.

No, that really is how science is conducted, extrapolate logical hypotheses from incomplete available data, attempt to prove or disprove them with more data and/or other insights. Chaserr may be no Einstein (sorry Chaserr ) but that's how Einstein did it.

 

[Chaserr]

Is String the official word for a sub group of cells (I assume within a single module)? I just learned it from you, so I'm trying to use it in the same context.

Thank you for this. I like your hypothesis best. It fits the data and doesn't make Tesla sound like they want to kill us all with dendrites constantly trying to burn our garage.

Why would constant cooling be needed all the way down to 70%? Does this help in balancing the imbalanced strings? Are overcharged cells hotter?

No need to apologize. I've been following the scientific method all along, and trying to help others learn how to reason scientifically when they don't understand what we're doing. It doesn't take any smarts, just deductive reasoning and the willpower to avoid rejecting data you don't like.

 

[Battpower]

I would look at the total voltage of each module

I assume your module is my brick? Bunch of parallel connected cells? If a bunch of cells in a brick drop out (their individual fuse blows) then the BMS will only see the difference when that brick is under load. Its internal resistance will increase as there are fewer cells in parallel but its no-load voltage clould look 'normal'. Also, the bricks are in series. The charging current is the same through all bricks wired in series, but is shared between more cells in bricks that have not lost any cells.

So, as a battery gets older and cells drop out - especially if all in one brick - that brick will get a beating under heavy charge or discharge. I am sure that one brick having excessive volt increase / decrease will be picked up by the BMS, but may there may be edge conditions where this happens too slowly / late?

 

[Battpower]

Are overcharged cells hotter

I believe internal resistance goes down with heat. Hence need to warm batteries to get highest output.

Cells could get hotter as they reach full charge.

 

[ucmndd]

No, that really is how science is conducted, extrapolate logical hypotheses from incomplete available data, attempt to prove or disprove them with more data and/or other insights. Chaserr may be no Einstein (sorry Chaserr ) but that's how Einstein did it.

As someone with two science degrees from accredited universities, I must admit I’ve never come across the step in the scientific method that says “make things up and use definitive factual language without supporting evidence to strengthen your hypothesis”.

 

[aerodyne]

As someone with two science degrees from accredited universities, I must admit I’ve never come across the step in the scientific method that says “make things up and use definitive factual language without supporting evidence to strengthen your hypothesis”.

One of your degrees was not in political science, I gather...

 

[ucmndd]

One of your degrees was not in political science, I gather...

Point taken, and no.

 

[Battpower]

Re-reading @lightningltd 's post, I woulld suggest it's more of a current issue than voltage - at least with no current, bricks with multiple dropped cells just sit there. I am struggling to visualize a situation that turns into thermal runaway (ie: fire) which is what this is all about. If it is not dendrites or some other internal condition within a cell, then any condition in which current flows would be external to the cell and the fuse would blow. Mechanical fuses are generally fairly blunt instruments so what if cells start blowing fuses in a cascade fashion? Remember we are looking for a runaway situation that involves currents not controlled or monitored by BMS which would be within a brick.

Reducing the max cell voltage is obviously reducing total energy stored, but also energy (and probably temperature) inside each brick.

 

[JRP3]

Since the BMS cannot see individual cells, it would need to look at each module and the system as a whole.

Cells in parallel are at the same voltage and are read as a single large "cell". If some cells in parallel were dead then that cell group would sag more under load than other cell groups and it would reach "full" faster than other groups during charging. Both conditions would be seen by the BMS.

If the string is charged at 400 volts and all is well, then great. But what if 10-15% of the cells are not taking the charge? We are then giving 400 volts (or whatever it is) to less cells, resulting in potentially overcharging the remaining cells.

The BMS would see the lower capacity group of cells reaching max voltage faster and should stop charging from continuing. If it didn't the lower capacity cell group would be pushed beyond max voltage. I would think that at some threshold the imbalance would be too great and the BMS would produce an error requiring pack replacement.