[DIYguy]

If you have some in a heated battery box and others not... .that's a different thermal environment. It wasn't really quanitified ...however... temperature effects just about everything If you push the extremes, you can change state of matter. lol

My take was that the useful capacities are affected. This is well established and posted by all the battery companies.

[JRP3]

Originally Posted by Qer

I guess this was an attempt at being sarcastic. Since you apparently misunderstood me and thus your attack were uncalled for I'll try to be civil and explain again what I mean rather than bitch back (except for this sentence which I, quite frankly, think you deserved ).

Really? You're going to play sarcasm police, after this statement?

Originally Posted by Qer
Also, this means you run the cells repeatedly into a very deep DoD and if it's true that it's deep DoDs that really make the cells age fast you'll effectively shorten the life span of the whole pack instead of individual cells.

But it sure beats running without a LVC-BMS on a top balanced pack.

I detected sarcasm and responded in kind, it didn't seem uncalled for and I reject your attempt to chastise me. So there.

[Jimdear2]

It seems to me there is some real reluctance in the DIY EV community to size a battery for real world conditions. All I am seeing are a lot of the people that look at buying a battery just a bit wrong.

Kind of like buying a ICE vehicle with an X gallon gas tank, since they normally only drive Y miles a day that all they need.

The very first thing is they have to realize the pack needs to be oversized just to protect itself (untouchable excess). Then they have to add a bit more for the time in slow traffic with the wipers and headlights on.

From what I've been reading, for prismatic cells, it seems to me that the best Battery Management System is actually a Battery Monitoring System that controls what you put into and take out of the pack as a whole, not adjusting the in and out of individual batteries.

The BMS has to become transparent to the operator.

That is it needs some way of keeping your cells from terminal low or high voltage and damage. This is the desired result and is a real world requirement. By the KISS principle the best method is to never go where the damage can happen.

Normally, people who drive ICE never even come close to using the maximum fuel available. I.e. they don't fill the tank until it runs over or drive around until they run out of fuel, this id the ICE KISS principle.

Now we get to my questions about how and if.

Mostly this is addressed to Tesseract and QER as the only EV hardware and software experts that I know of and who know each other and have an understanding of what they each can do. I know there are others out their please jump in.

I'm thinking that for prismatic cells to come into there own, they need two things, just like the cylinders of a modern engine.
One) NEVER be taken into the area that potential damage could occur.
Two) Each cell needs a positive identifier so if damage does happen, you know where to look.

Tess, is it possible to make a really reliable voltage sensing device that would have a (this is where QER comes in) unique code assigned and then be potted to the point of indestructibility? These would need to be attached to each battery as it was installed. Then, can each unit be polled over the existing traction battery cables? I know that you can buy intercoms that use the existing wiring in a house to transmit and receive, I would think this would be somewhat similar.

Then, (TESS and QER) could a central processing unit be made that would look at these signals from each battery and could watch during charging and shut things down at chosen percentage, say 80 or 90% charge for the highest voltage battery in the pack. Then identify that battery if it were consistently more then X% out of balance. Thus preventing any single battery from ever being driven too high, and allowing quick positive location of a bad or damaged cell.

This CPU would give a full FUEL gauge reading if the all batteries in the pack successfully reached the 80 or 90% chosen to be a full pack. It would also give a less then full FUEL reading at the end of charging, if not all batteries reached a balanced full charge. The system would identify any battery that was consistently out of balance by a set percentage (by the unique number of the sensing unit).

The converse of the above would protect the pack at the low end. The system would show a empty gauge and to protect the battery actually shut down the vehicle (with a couple of shakes and judders first, just like running out of gas) at a safe 10 or 20% of pack capacity. Again if one or more batteries were out of balance above a set percent consistently, the CPU would identify that battery.

So we would have a monitoring system that manages the pack as a whole. Never allowing an individual battery to become damaged, and giving each individual battery a bit of wiggle room to come back into line without ever actually adding or subtracting from a specific battery. Finally it gives the operator, by monitoring the voltage of th pack as a whole a reasonable full to empty fuel gauge.

There would need to be some pretty fancy programming in there to say that if 1 or if 10 out of twenty batteries reached full charge voltage before the rest, it would affect the overall percentage or the rate the gauge went from full to empty.

I don't have the skills to know if this is possible I await your comments

[JRP3]

Originally Posted by Jan

I wonder if that's the case. Why advise to store a battery at or below 50% SOC. If going low is bad. It all seems to contradict to me.

Manufacturer data clearly shows higher cycle life by staying away from 100%DOD, progressively longer to 70%DOD usually showing more than 3000cycles. After that additional cycle life gains are basically meaningless so presumably any possible benefits from storing at a higher SOC than 50% are probably outweighed by the benefits of staying around 50%. If you mostly discharge to 70-80% on a regular basis projected cycle life should allow the pack to outlast the vehicle, even with the occasional 100% DOD event.

[JRP3]

Originally Posted by Jimdear2

Finally it gives the operator, by monitoring the voltage of th pack as a whole a reasonable full to empty fuel gauge.

Voltage is not a great indicator of SOC through the middle discharge range so ah counting is better. I've used voltage as a monitor but it does fluctuate a lot under load so you really have to get a feel for how your pack behaves. Ah counting takes the guesswork out.

[Jan]

Originally Posted by DIYguy

If you have some in a heated battery box and others not... .that's a different thermal environment. It wasn't really quanitified ...however... temperature effects just about everything If you push the extremes, you can change state of matter. lol

My take was that the useful capacities are affected. This is well established and posted by all the battery companies.

Grm. Why is it all so unclear?

The temperature affects the capacity in two ways:

Low temperatures decreases the capacity temporarily. Simply because the chemical processes in the battery need a certain minimal temperature, and operate better at higher temperatures

But higher temperatures, speed up the aging process. What decreases the capacity in the long term.

The cells heat up, according to their IR under load, and cool down according to their position in the car/pack. Both are different per cell.

Which effect causes permanent drift? And how serious?

Anyway. So far I know now, deap DoD ansich doesn’t do anything bad to a cell. On the contrary, higher SOC (high voltage) is the bad guy. But most of all: temperature.

The only correlation between DoD and aging is the temperature: The low SOC mean higher IR, which mean higher temperature under the same load. Combined with the fact a deap DoD defacto means a lot of amps for a long time. Which in its turn means higher temperatures for a longer time.

Deap DoD => high temps for a relative long time. It doesn’t much matter if you come down from 100% to 20% SOC, or from 80% to 0% SOC. The same time period, the same amps, only a little higher IR in the last one. I doubt that will be measurable.

At no load, the higher voltage has a much better documented and noticeable aging effect on the cells.

Conclusion so far: Bottom balance and keep them cool. Keep them all just as cool. But not to cool.

Oh, and CMS?
Inconclusive.

[Jan]

Originally Posted by JRP3

Manufacturer data clearly shows higher cycle life by staying away from 100%DOD, progressively longer to 70%DOD usually showing more than 3000cycles. After that additional cycle life gains are basically meaningless so presumably any possible benefits from storing at a higher SOC than 50% are probably outweighed by the benefits of staying around 50%. If you mostly discharge to 70-80% on a regular basis projected cycle life should allow the pack to outlast the vehicle, even with the occasional 100% DOD event.

I disagree. 0% is better than 100% SOC for a li-ion. All books, handbooks, and other online literature I read confirms this.

It’s just HOW you go from 100% to 0%. Or from 80% to 20%.

What is the temperature of your cells and how long do you punish them with this? Of course going to 100% DoD or to 50% DoD, under the same circumstances, the first one will be worse for your cells.

But I believe (at the moment) that going slowly, with proper cooling from 80% SOC to 0% SOC is better for the pack then from 100% to 20% SOC without proper cooling and with a higher load.

Temperature is what will kill them finally.

[ElectriCar]

I wish the good doctor would get on this thread. He seriously could answer some questions and put other notions to bed! JRP3 maybe you or DIYguy could invite him, whichever one of you has been conversing with him. He would only need to subscribe to a few threads and answer anything contrary to the truth, which on this site is quite a bit.

[JRP3]

Originally Posted by Jan

It’s just HOW you go from 100% to 0%. Or from 80% to 20%.

None of the manufacturers show this. They use the same C rates in their test at 100%, 90%, 80%, 70%, and clearly show longer cycle life with shallower discharge. At constant C rates the cells reach a temperature during use but if it's within the capacity of the cell to shed that temperature a cell shouldn't get any hotter discharging to 100% than 70%, so I think you are jumping to conclusions that may not exist. Manufacturers do allow higher C rates with more aggressive cooling.

[ewert]

You know, I was thinking about the whole CMan/CMon/BMan/BMon thing. Has anyone done intentional high/low cell monitoring setup?

Say 40 cells. Bottom balance. Charge one cell say 3% of capacity. Charge whole pack to knee (with the high charged as marker). Discharge one cell say 3% of capacity. Do cell level monitors on just these two cells (now that's only 5% of the cost and much-o-less wiring of something like dmitri's CMon's instead of using one on all cells).

Now, during charging HVC trips on high cell and leaves rest mostly un-kneed. During discharge, LVC trips on low cell and leaves rest again mostly un-kneed. Beware though, you really REALLLLLLY do not want to go past that LVC, or the designated marker cell is gonna buy the farm.

This same system could very easily be made into a cheap and effective middle SoC running system by increasing the top and bottom differences. Maybe even do split pack voltage monitoring of the "base cells", so that cell #1 LVC, cell #40 HVC, and cells 2-39 are split into 2-20 and 21-39 with voltage monitors for splitpack monitoring, so you will see if your main pack ever goes into imbalance. Add a switch to bypass the #1 LVC cell for emergencies, so you can go to super deep DoD if a true emergency?

Sounds good to me ...