Just want to bounce around an idea here.
It's based on these general "impressions" which I have gained from my 2 year peek into battery technology:
.
Impression A) People want more range than what the currently available batteries can provide.
Impression B) Batteries last much longer if they are being cycled in many shallow cycles, rather than frequent deep cycles.
Impression C) Batteries get damaged by heat caused by a variety of "mechanisms". The worst damage is probably done by heating up excessively from the inside out, particularly by exceeding maximal charge and/or discharge rates, and over-charging and over-discharging.
Impression D) You need high voltage batteries to keep cable sizes and resistive losses low in the overall system. More volts generally mean more "Ooomph" and more speed. Unfortunately, this also makes it very dangerous.
Impression E) To get more Ooomph you need many cells and that means a complex and expensive Battery Management System (BMS).
Impression F) Batteries generally like it to be used to no more of 50% of their capacity, for both peak current draw and total Ah drain.
Impression E) There is always one cell that bites the dust first - depending on circumstances, it may then take the others down with it...
.
Now to the conclusions from these impressions:
1) Combining A) + B) +F) ===> In order to get good battery-lifetime range, the available trip range needs to be reduced even further - by about 50% ! Not a popular idea, I'm quite certain...
2) Because it is always one cell that goes down first, you might as well make sure you know exactly which cell this will be. If you can be certain that this cell will be the one that gets over-charged or over-discharged before all others and that it suffers more heat damage than all other cells, then you only need to monitor this single cell! That's the idea. Quite simple.
The proposed solution (possibly hogwash!):
Place a sacrificial cell into the battery string, monitor only this cell. The operating conditions for all other cells should then always be within the parameters conducive to maximising the lifetime energy delivery of the battery.
.
Here is an example to illustrate the idea:
Imagine a 40s 40Ah battery.
Instead of monitoring every cell, place just one 25Ah cell into the series and monitor this cell constantly.
When it dies (and die it will!), replace it with another 25Ah cell, again and again and again, until the pack has deteriorated down to maybe 30Ah.
Once the pack capacity has deteriorated significantly, despite the very gentle treatment it is receiving, you could either get a new pack or use an even lower spec sacrificial cell to keep it going.
Needless to say, older cells of the same product line might be perfect sacrificial cells. Down-cycling at it's finest!
.
What say you?
It sounds like an idea, it's probably a way to preserve a good pack, but..
-wouldn't you have to limit the max amp/amph draw to what that 25ah cell can deliver? (loss of "ooomph"/range)
-wouldn't there be a possibility of reverse charging, just like with a dead cell?
-if another cell is dying due to any reason, you still wouldn't know I think.
that's why I was thinking.. isn't there a device that can do what a good BMS does (among other stuff): measure each cell, and report in any way?
basically a volt-meter with 40 probes (or whatever, i'd be happy to wire them to each cell) and a display or usb port for read-out.
we'd just have to check once in a while..
does anyone know about such a device?
"doing nothin = doing nothing wrong" is invalid when the subject is environment
Exactly, that is the price you pay for pack longevity! I'm trying to find an easy way of getting 100,000km or more out of a battery pack.
The simple rule to follow seems to be: Never drain them and never stuff them!
The Sacrificial-cell-BMS monitors this cell and shuts down the entire EV (or puts it in turtle-mode of some kind) to prevent this.
That is correct; but the failure of another cell would be very unlikely because they are never used under demanding conditions. Only the sacrificial cell works hard.
A good BMS for every cell might be complex and expensive. It remains to be seen what the component failure rate under real world conditions will be. The road is a very harsh environment, and the more parts your BMS has, the more can go wrong in the long run. Think of corrosion due to water and salt; vibration and extreme heat and cold; and problems like dry solder connections could be a nightmare to find in a complex device. And each single tap wire is live with potentially lethal voltage between it and some other tap wires.
And 40 cells is not enough for most EV's. 80 to 100 cells (Lithium) or 200 to 300 cells (Nickel) will probably be much more common numbers.
.
Taking the 102s 30Ah NiMH battery of the Vectrix as an example:
Put in one weak cell in a relatively easy to reach location, like the top layer of the rear battery. Then monitor that cell continuously for voltage and temperature.
The BMS could have all the bells and whistles (without getting very expensive), because it is only needed for one cell:
- Optical (or other galvanic) isolation from the battery system for complete safety;
- backup power on board to log continuously, even during prolonged breaks.
- USB connectivity to computers, or bluetooth or WiFi or something similar to tell you when there is a problem during charging.
- the ability to terminate charging by turning off the power to the charger when the monitored cell is full (Voltage > 1.45V). An occasional trickle charge can be applied to the entire battery to make absolutely certain that the 101 good cells always have a higher SOC than the sacrificial cell.
- the ability to light a warning symbol on the instrument panel whenever the voltage drops below 0.9V during riding. This will allow to tailor the riding perfectly to the demands of the weakest cell, but still allows almost full power for emergencies.
- the ability to give a temperature warning light for over-temperature conditions and to terminate charging when the sacrificial cell temperature reaches a cutoff value during trickle charging, something like 40degC. It might also turn on the impellers instead of turning off the trickle charger.
I think there are existing BMS's for Lithium cells which can do most or all of these functions, but I have not seen one that can be used for NiMH cells.
After maybe each 10,000km the sacrificial cell is thrown out and replaced with another one. You need an extra 9 cells over the 100,000km life of the pack.
This information may be used entirely at your own risk.
There is always a way if there is no other way!
battery surgery every 10,000km sounds like a pita.
I'm hoping my 40AHA Sky Energy cells have enough capacity to cover my normal commute at less than 80% DOD.
With 42 cells, the stock Vectrix controller should limit charge and discharge to acceptable levels, with the EVWorks BMS as back up.
At 60km per charge and the manufacturers claim of > 2000 cycles, I'm looking at 120,000 km. but I think that cycle life refers to 0.3C "standard discharge"... 30,000km wil pay for the cost of the pack compared to my alternative transport. I guess we are alpha testers here :)
I think you are right in the points in your first post.
Basically, current battery technology is not yet upto consumer demand/expectation.
For EV's to become mainstream, either they need to get better or ICE cars become less viable.
Sacrificial cell BMS might be the way to go for backyard tinkerers, but an EV consumer won't tolerate less range,less performance and higher maintenance costs for the sake of battery longevity.
It's no big deal when the "bad" cell is in the top layer. No heavy lifting of batteries, no tedious searching for the damaged cell(s). And because the BMS has monitored the cell extremely well, the replacement would not come as a surprise, but could be done as a planned service.
I am most certainly looking forward to your test results! Hopefully we will be able to figure out what works best, despite the enormous difficulties when trying to compare performance under different circumstances, like terrain, climate, driving habits etc.
I don't think the maintenance costs are necessarily higher. From a manufacturers point of view, knowing what to do, and when, is much cheaper and easier than diagnosing and repairing a suspected bad battery pack. The costs for replacing just that one cell in all bikes might well be below the costs caused by a standard battery failure in every 40th bike.
The bike could be $1000,- cheaper in initial cost without a per-every-cell-BMS. That would make a big difference to sales!
The range, if advertised correctly, would be about 1/4 of the inflated claims of the competitors, that's a very big marketing challenge!
But one of the first questions asked about EV's is usually: "How long will the battery last and how much does it cost to replace it?"
This information may be used entirely at your own risk.
There is always a way if there is no other way!
if the sacraficial cell is to be used to limit discharge in Ah (and possibly peak in A), there is another way to do it.
for the vectrix, changing the measured capacity to say 20Ah would achieve the same thing, although you wouldn't have to tear into the pack every 10'000km.
Matt
Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km
I think that would be different from true monitoring of a single sacrificial cell, because of the creeping error of the self-discharge rate.
It does not matter what the overall capacity is, when some cells reverse or overcharge while the system is assuming an incorrect SOC.
This information may be used entirely at your own risk.
There is always a way if there is no other way!