I first met Brian Noto in 1995, shortly before our disastrous BBSCON in the Tampa Convention Center in August. The advisability of holding a convention in August is on my mind these days for obvious reasons as I may have made the same mistake again with EVCCON 2013 scheduled for August 6-11. But in 1995 I was younger and prettier and a little more naive. The basic premise was that the Tampa Convention Center would GIVE us the facility for five days if we would just hold it there. They couldn’t get anyone else in the country to schedule an event at that time and the hotels are just withering into the ground that month as well. They didn’t tell me of course that they DID charge $800 a day to run the escalator. Or that there was another charge for the air conditioning. Some aspects of business are simply bitter and negative.
In any event, it was a massive and gorgeous facility. The vendors could drive their trailer trucks right out onto the exhibit floor to unload. Once all of them had, we closed the doors and started the air conditioners. Millions of dollars worth of electronics equipment of course and all the vendors working furiously to erect their displays.
The cold air conditioned air hit the very warm, very moist air that was in this huge hall from having the doors open earlier. And of course, the humidity immediately condensed. I’d never actually PERSONALLY been responsible for a rainstorm before and didn’t know quite how to react as the vendors formed a posse to drag my remains through the public square. Luckily, the rain continued for 15 minutes and everyone stopped to marvel that such a thing could even occur.
As it naturally started to trickle off, I stepped out on a balcony, held out my hands and commanded the winds and the rain to cease. This caused howls – I think they were laughing but they may have just as well been howls of rage looking back on it.
[jwplayer file=”news030813-iPhone.m4v” hd.file=”news030813-1280.mov” image=”http://media3.ev-tv.me/news030813.jpg” streamer=”rtmp://s3einxnpkaij93.cloudfront.net/cfx/st/” provider=”rtmp” width=”950″ height=”584″ html5_file=”http://media3.ev-tv.me/news030813-1280.mov”]
In the interim, El Braino has worked for me five times making this his sixth run at employment, not counting my silent partnership in his Krav Mahga/Brazillian Ju Jitsu venture in Southern California. A complex relationship, you have to understand I’ve never hired the man even once.
He came into the Boardwatch Offices on Bowles Avenue in Littleton Colorado one morning and indicated he was looking for work as an advertising salesman. He learned of our existence because we printed with a company that also printed a “Thrifty Nickel” type neighborhood advertising newsprint he’d been involved with that didn’t work out so well. Harried and disorganized, even then, I was searching for my sunglasses (on my head as it turns out) and trying to do something at the time. I told him we didn’t have any current openings but that I was sorry the rag didn’t work out and good luck.
Looking back, I guess he didn’t want to go home and talk to the wife, Debbie I think that one was, so he just hung out around Boardwatch chatting with different employees. He came in the next day as well and somehow talked somebody out of temporary use of an office, desk and phone. By the end of the week he was a fixture. At some point I guess Kathy thought we had hired him and had him fill out a withholding form and got his Social Security Number. Seemed like he was always around after that.
He did go into a room where I had had a guy from some Boulder magazine about Mercenary Soldiers who worked for us briefly on retail distribution. Brain found the printouts of all the stores in the country where we appeared on the newsstand and started going through them. Within a few months, he had trimmed our drop to these stores and increased our sell through probably 30% turning this money loser segment into a slight profit. I started to warm up to him about then.
Ever since, he comes and he goes. I never hire him. I’ve never fired him. When I get REALLY dependent on him, he goes away. When I get used to doing it on my own, kind of badly, he shows up. As he’s on his fifth wife, I suppose it could all be even more emotional if I was a woman.
In any event, while I am totally disorganized, near hysteria most of the time, and trying of course to do 12 things at once, the Brain is calm, methodical, organized, and instead of telling everyone they are f***ing morons he has this cunning way of talking to them calmly and getting them to do whatever he needs done. I don’t really know how that works. But its very effective.
So Yin and Yang, we are entirely different but quite complimentary in an odd way. We don’t really “share” or have “feelings”. Brain was an only child and doesn’t really do that. His mother says when he was a kid she had to tie a pork chop around his neck to get the dog to play with him. So after 17 years and five rodeos, I guess I don’t really know him very well. I know he’s a couple years younger than me, and likes weddings. Actually he was best man at my last wedding. As the preacher was in his cups during rehearsal, Brain calmly showed him how to do the ceremony and where everyone stood and didn’t even call the guy a f***ing anything.
His last wedding got rained out, so we moved it to my wine cellar and about 100 people got REALLY messed up on some pretty good wine.
He’s particularly good at what I call fitment. When I put something on a car, it is usually hanging by some zip ties, crookedly and not very securely, and at an odd angle. Somehow he has a way of removing the zip ties, and carefully mounting the item on the car as if it were done on purpose, in the exact position I WOULD have put it if I was good at fitment. That’s kind of handy.
But I suppose life with Jack has it’s stresses and strains. And at some point, he goes away. The question is, why does he come back? He’s very capable at anything he does.
I kind of think he is addicted to drama. At Boardwatch, and at EVTV, every day features a new plan on how to take over the world, from the Steven Spielberg cartoon Pinky and the Brain – ergo the Brain. If you get used to that, it kind of feels like you are on a mission. It is of course a relief to escape it to a normal world of magazine publishing and ad sales. But at some point it occurs to you they are there to make money and have a life with their wives and kids and dogs and sell magazines. No real mission. No real drama. Commerce. Once you get addicted to taking over the world, it’s hard to do that for any length of time. It’s restful. But not very satisfying. You get to where you kind of LIKE taking over the world.
Of course everything at EVTV is almost the opposite of what it appears. In Speilberg’s cartoon, the BRAIN is a megalomaniac bent on grandiose but implausible/impossible plans to take over the world, and of course from there – the universe. Pinky is the pragmatic who bails him out of each episode just in the nick of time as Brain gets a wee bit overextended. And so now you know “the rest of the story” as Paul Harvey puts it.
Technically, this week we continue our bizarre quest to heat the Escalade. Preview of next week. The SSR I installed in this episode has already failed catastrophically and I was forced into deep appreciation of the forethought of a maintenance switch. ALWAYS have a heavy duty maintenance switch somewhere handy to break your battery pack. I’m not talking about a little slap switch to disengage a contactor. I’m talking a heavy duty switch so you don’t have to use the insulated cable cutters to cut a cable.
Oh, and the 4000 watt heater melted the heater hoses. Other than that, great ideas.
Tesla did file their 10K finally on Thursday. We talk about that a bit. But notably is their admission that on the Tesla Roadster, overall battery capacity, it looks like now, will come in at 60-65% of the original capacity of the car within 7 years or 100,000 miles. This kind of wrecks the resale value of these vehicles as it is a very different matter from 80% in 10 years.
The heart of this problem is their choice of batteries. This choice is actually kind of cunning in some interesting ways. They can design a battery module really pretty much ONCE as the 18650 form
factor is forever. These cells are used in millions of small electronic devices – the lithium version of the AA battery. Not only do they gain the economies of scale as millions of these batteries are manufactured, but as the chemistry improves to provide more power, energy, and thus range, their upgrade development cost is limited to a few scraggly lines of software code.
The downside is they are using a flashlight battery for a car. And flashlights and cars don’t have quite the same needs. In a flashlight, energy density and power density are important as well as cost. All that is true for a car. But nobody cares how long a rechargeable flashlight battery lasts really. And no one making them is going to spend much on getting them to last LONGER than they do now. But for me, the MAIN advantage of lithium over lead is NOT that they are smaller and lighter, however welcome and enabling those features are. It is that the cycle life is MUCH longer. 300 cycles for lead, 3000 cycles for Lithium ionic cells. Panasonic, when they talk about it at all, talk 300-800 cycles for their cells. And so two years ago I talked about the cycle life of these cells and what Tesla could have done to address this problem. Thursday’s 10K indicates that the answer is “not very much.” There’s just not much to be done. The new NNP NCR18650B from Panasonic doesn’t even LIST a cycle life rating.
Most Roadsters don’t actually get a lot of miles. They are kind of toy cars for the weekend and for the most part not employed in a daily driver role. But I think this darker side of Tesla ownership is going to become an issue with the Model S. It IS likely to be the daily driver. Model X as well. The Tesla battery technology is kind of central to their game. Maybe at some point, they will be buying enough cells to have them custom made with more attention to life. But I think you will hear more on this topic and not from me. Replacement modules are NOT going to be cheap. But if you are nearing the end of warranty, measure carefully and it will be interesting to see what their policy is on battery replacement if you are under 80%.
That takes us to a discussion of battery costs generally. As you know, I’m a big fan of the new CALB cells. But there should be choices in life and the CALB cells have not gone down in price AT ALL. In fact, when the China Aviation Missle Academy first invested in Sky Energy, the prices went UP slightly. When they gained a U.S. warehouse they went UP again. It is true we are offering at about $450 per kWh in a world that is totally convinced throughout America that the going rate for lithium is $600 per kWh. But it isn’t going down.
So I’ve been talking to other battery vendors in China about alternatives. There aren’t many. But there are a few. They don’t have the power output or cold weather performance of the CALB, so
surely they would be less expensive.
Oddly no. Well they are, but like a nickel an AH. Not enough to fool with really. I keep chasing 90 cent an AH batteries only to learn later that they are padding the shipping or adding on extras or something.
What I think is that the Chinese have FOUND the bottom on battery prices. They are not going to come down from here, and all that are waiting for the big crash in prices to start their project are going to be waiting sometime.
The reasons for this are actually pretty understandable. There is just a certain amount of copper and aluminum and polyethylene and plastic that go into a 12.5 lb 180Ah cell. The process to put it there requires some capital intensive investment in precision equipment. It has to be in a pretty controlled environment in a nice facility. Worse, once it is there, nobody in the building wants to buy a battery. So the cells have to be MOVED. ANd every time someone picks one up, it costs. And every time that 12.5lbs gets shipped, there is shipping. And moving a lot of them around is very expensive. So trying to negotiate a lower price eventually we just reach a point where nobody makes enough per cell to fool with them. The little dark secret in all this is that
the Chinese government actively subsidizes the export of these cells. YOU are buying them here for less than the Chinese can buy them in CHina – even after shipping several times.
So why do I keep looking forward to cheaper batteries? Well I do. I just didn’t realize until recently what YOU expected with respect to cheaper batteries.
If we say that a CA180AH 12.5 lb cell is $259, note that the previous SE version was really about the same. While the price went up when they took over, and the price went up when they opened the US. warehouse, it did NOT go up when they went from SE to CA version. Even though the batteries were much better.
We have reached the “floor” price on a 12.5lb battery cell largely made of copper and aluminum. But the CHEMISTRY can be improved. And those improvements could mean higher energy density, higher power output, and longer cycle life, all “GOOD THINGS” in the World of Martha Stewart EV’s.
Let’s say a way was found to account for the 400% volumetric expansion of silicon on the anode. And some combination of carbon and silicon was developed that would DOUBLE the capacity of the cell.
Well that would be great. Using the same physical cell format, we coudl replace our cells and go from a nominal 80 mile car to a 160 mile range in our car immediately. On introduction, this new chemistry would undoubtedly command a premium, but as the numbers ramped up and other competitors also produced higher capacity cells, prices would sink to about the floor we are at now.
This isnt’ going to save you anything. The pack is still $10,000. But now you can go twice as far.
Let’s say then that a solution is found for sulphur reacting with electrolyte to form crippling byproducts and suddenly we could use sulphur on the cathode side and the battery capacity doubled AGAIN. Now our 12.5 lb cell, which had been 180 AH and then 360 Ah is suddenly 720Ah. Still $259. Well the truth is there is limited utility for me personally in going from 160 to 320 miles on a charge. I would be sorely tempted to cut my pack SIZE in half and WEIGHT in half and thus COST in half and just settle for 160 mile range. Where am I going anyway?
And so I DO see the cost of “batteries” going down generally. But I don’t see the cost of manufacturing, distributing, selling, and shipping a 12.5 lb metal cell going down. We’ll get an ever improving cell. And I do NOT buy into the 6-8% per year gig. These things go in fits and starts. We’ll be stalled out for awhile, then big breakthrough and trickle down period. Then stalled out, then big breakthrough and trickle down period. It might ultimately be 8% per year on average, but it will go in fits and starts.
I continue to be amazed at the almost total misunderstanding of these cell chemistries by some very talented, bright and educated engineers. They seem UNABLE to let go of myth and leftovers from previous chemistries and the misinformation seems to feed on itself.
Also this week, the NTSB issued an interim fact finding report on the Boeing 787 APU battery. Boeing and the battery designer are working FURIOUSLY to put together a package of engineering improvements to satisfy the FAA that it is no longer a problem. Kind of difficult since NO examination has shown a smoking gun – a failed part – a failed procedure, or anything other than that the cells failed and shorted as if by magic.
My fear is that the conclusion that will be reached is simply that these cells are simply dangerous, and should not be allowed on airplanes. In fact, since we don’t know why they failed, perhaps we should ban their SHIPMENT on aircraft.
I take some joy in reading this report, though I was doing it most carefully yesterday while editing the show. Puzzles are fun for me. And I am pretty confident I have decoded EXACTLY what is going on and exactly what caused these cells to fail and an almost ZERO cost solution to keep them from ever failing again. I’m most delighted by this because they use Lithium Cobalt cells, not the hybrids but the LiCo cells that I detest. And I think this explanation covers ALL the fires including the laptops and cell phones and all of them.
I could be delusional. I cannot believe that over all this time, I’m the only one that understands these cells this way. I by vote and by all that’s holy HAVE to be totally wrong on this. AND I happen to know I’m not. Very peculiar situation. Kind of gives me a chill. And I’ll tell you why. YES. I COULD have automated our testing with Labview. I can also automate our bottom balancing. The only way I can account for deep knowledge of these cells not elsewhere available:
1. I didn’t automate it. And so I was FORCED to STAND and watch these cells for hundreds of hours at a time.
2. I learned quickly that most of what was printed about these cell chemistries was utter nonsense.
3. Best of all, I bought them from the Chinese. Their idea of documentation on my first purchase was a badly laser printed black text on a half sheet of not very white paper. It read in total:
Your glad acceptance is our warmest happiness. Best quality. Best Service.
I had NO instructions or specifications. So little in the way of procedure to deal with them. And the mistakes cost me direct ducats in a very painful way.
I never did get a spec sheet on the Seidens’ But I WAS able to get one with my first purchase of 90AH Thundersky yellow cells. And I found the charge procedure, to 4.2vdc very strange.
Would you believe I went to no small effort originally to develop a BATTERY MANAGEMENT SYSTEM to measure and to balance those cells? To laugh now. But I REALLY did.
But the best thing I did was spend time with them. I woudl charge them. And not only watch them charge, I watched them for hours AFTER I charged them. By voltage, our easiest measurement.
I also discharged them. I’ve bottom balanced THOUSANDS of these at this point. You discharge to a voltage, and then it BOUNCES back up. So you have to do it again. If one is to low, you have to add a little charge. And then the voltage FALLS in the corresponding move to the discharge. They are kind of “spongy” this way.
And after awhile you start to give them names. And cells sometimes don’t “feel” right. They are perfectly operational. But they don’t “feel” right. I marked one when putting together the Cadillac Escalade battery pack. It actually had a kind of a high capacity. But something wasn’t right with it. It didn’t react like the others when I was manually balancing. We drove the car a few times and had a problem with the pack. I told the kid working in the shop at the time to change out cell 38. I didn’t even go look or put a meter on it. Sure enough. I just KNEW that cell had a bad attitude and would never grow up to amount to anything. Bad seed.
I recall explaining all this to young Hauber. He thought I was truly mentally disturbed. So I had him bottom balance a pack he had kind of mixed up badly for the Green Van. At the end of three days, he had given them names. And if I got to near the pack he would step between us protectively. I wan’t allowed near his young. HE would take a look.
Ok, ok so I spend too much time alone in a dark shop. My point is that I have viscerally come to learn some things about these cells that putting them on a cycle machine and reading the graphs doesn’t quite cover. Couple this with having read almost every journaled paper on the topic for three years, and I have some deep knowledge.
I guess the most intimate thing I have learned has been the one I have failed most generally to communicate. That is, that everything we measure on these batteries is kind of a once removed one off that does NOT reflect the battery itself. Voltage DOES indicate SOC. But only static voltage. If you put energy in or take energy out, you set off a massive rearrangement of the chairs within the cell that I loosely term DIFFUSION. And you will not be able to ACCURATELY measure a voltage for at least 12 hours and in reality, better 24 hours.
Even the discharge curves all batteries now come with are approximations. They will show you one curve at 0.5C discharge and another curve at 1.0C discharge and another curve at 2C and at 3C. All will be different at different temperatures. They are INDICATIVE in that they are using a CONSTANT CURRENT load and so we can graph the VOLTAGE of teh curve, either charging, or discharging, and after awhile you think that IS something. It IS something. Just not anything real or that the cell cares about.
If you leave it alone for 24 hours you can measure the voltage and tell pretty much what you have. But then if you do anything about it you change it and have to wait 24 hours.
And so we operate on these cells in really a one -off fashion. We can develop charge PROCEDURES and bottom balancing PROCEDURES and if you didn’t know better, you would actually after awhile think they become real. They are not.
Let’s do an analogy. I’m going to give you a recipe for a chocolate cake. I want you to mix some Hershey’s cocoa, some flour, a couple of eggs, some milk and some baking powder (pinch of salt) in a large bowl. Pour it into a pan. Put the pan in the oven. Set the oven to 350F. Cook for 30 minutes. Remove the pan. You have a cake.
Now if you put it in over at 350F and take it out after five minutes, will you have cake? No. It won’t be done. How about if you get some precision equipment and set it to 350.00001 degrees and build $2000 worth of temperature control to make sure that never varies more than 0.001 degree F? Won’t help. How about if we make 350F a religion and actually put the cake in a pan that is stamped 350 in the bottom backwards so that if we DID ever get it cooked it would come out saying 350F across the top?
Actually there is no magic in 350F. We could cook it at 300F and extend teh time a bit and cook it for 40 minutes. We’d probably get a cooked cake. Let’s say we went to 400 degrees at 20 minutes. Yes, it would be slightly different but broadly a cooked cake.
But let’s say we have our best success at 350F. So we’ll call it a 350F cake. Test question. What’s its temperature when we eat it? 350F?
It’s a procedure. We now if you do it this way you’ll get a cake.
So let’s assume we have a battery, and the only way we can really measure it is to wait 24 hours with no load and measure it. How do we charge it?
Well, how about this. Could we add energy to it at 1C until we have added 2 AH. And then wait 24 hours. And measure the voltage. Then add 2AH more, and wait another 24 hours, and check the voltage.
After about a month we would have a fully charged cell. Now there ARE some absolutes. We actually KNOW the electrical potential of a carbon anode. And we actually KNOW the electrical potential of a lithium iron phosphate cathode. And if we add them together we KNOW they algebraically sum to 3.38v. If we REACH those maximum potentials, we know FOR A FACT that we have a fully charged cell. But we can only measure this accurately after a days rest.
So let’s try something. Lets charge more quickly in one continous charge at a much higher voltage and get this over with. We know, that our electrolytes, even though they are organic solvents and do not hydrolize above 2 v like aqueous solutions do, STILL have some problems starting at about 4.5v. So we don’t want to go there. Let’s pick a value safely under that, say 4.2v.
We will charge at full current until we reach this arbitrary 4.2 volts and then let’s hold it there until our current required to hold it there diminishes. As the energy in the cell fills, the difference in potential will diminish and so will the current. And at some point then let’s terminate the charge. Then let it set for a day, and see how close we come.
For 4.2v, we can by doing this, adjust the point where the current has diminished to get a result we like- pretty close to 3.38 after a days rest. Oh, let’s add a little pad and say 3.36.
Could we develop such a charge curve at 4.0V? Sure we could. Using the same procedure. How about 3.8v? Or 3.6v? Actually, I have seen charging instructions for the SAME chemistry cell, from the SAME manufacturer, at each and every one of those voltages.
That’s because they aren’t real. They are a PROCEDURE to get to an open circuit voltage of 3.36 indicating fully charged with a little pad. We need a little pad because the penalties for overcharging are severe. Swelling. Which dislocates the proximity of the thin foil plates. Shorts. Fires. So 3.36 instead of 3.38.
So, the wise and learned learn NOT to exceed the CHARGE voltage because they DON’T want to overcharge the cells. And the corollary will then be if they DON’T exceed the charge voltage, they WON’T overcharge the cells. Make sense?
Sure it does. In your mind. Did you check with the cells on their take on this?
Since it is a contrived procedure, you have to follow the WHOLE procedure. And the WHOLE procedure is to TERMINATE ALL CHARGING at a specific current. What happens if you charge to the charge voltage and terminate IMMEDIATELY. You don’t fully charge the cells. What happens if you charge to the charge voltage, whatever it is or whatever you like, and then continue to add energy comfortable in the knowledge that it will decrease to a very low value and probably won’t do any harm.
You are either ADDING energy to the cell, or you ARE NOT. How quickly only matters over time.
Let’s go back to the cake. Our instructions are 350F for 30 minutes. How about if we put it in a preheated oven at 350F and after 30 minutes we turn the oven off. But we leave the cake in the oven. The temperature gradually decreases to 120 degrees over the next hour and a half. Care for a piece of cake? You didn’t FOLLOW the procedure. 350F WAS part of it. But it wasnt’ the entire procedure.
The FULLY charged voltage of a LiFePo4 cell is 3.38. If you connect it to a 3.39v source for long enough, you will overcharge the cells. You can tell this because they will swell. How could that possibly be since you did not exceed the 350F? And I mistake that for 3.8v on purpose because EACH value has about the same amount of matter to the cell. IT was a PROCEDURE. There was no magic to 3.8volts.
You can’t trickle charge lithium ion cells because you will inevitably and without fail overcharge them. And if you overcharge them they will swell. And if you overcharge them enough and they swell enough they will short. Actually not. The SEI layer breaks down allowing electrolyte access to lithium ions which causes a thermal event. This thermal event is sufficient often to cause the release of oxygen on the cathodes giving rise to a fire of copper, aluminum and plastic that provides its own oxygen. It gets hot. And you can’t put it out.
As it so happens, LiFePo4 has the hIGHEST temperature for the release of free oxygen from the cathode of any currently available Lithium ionic chemistry. And Lithium Cobalt, at 150C has the lowest. By a lot.
If you examine the CT Scan on page 22 of the NTSB Interim Factual Report you can clearly see several cells are dramatically swollen and the rest are somewhat swollen. The NTSB and Boeing are curiously in agreement that the cells have NOT been overcharged but largely from logs of the charging which does indicate any unusual charging activity or failure in the Battery Charging Unit or the TWO Battery Management Units. As the BMU were pretty much destroyed by fire, they can’t tell if they failed or not.
I rather gather they did not. You will find it alarming and out of character that I am NOT going to blame this fire on the BMS. I do not believe it caused it. Further, I think if they COMPLETELY redesign the BMU AND the BCU, using the very latest components available and the brightest engineers, add separators, and a really really CONTAINED can to store it all in, they event will happen at about the same life of the battery and with adequate containment, could very possibly convert a smoke and flame event into an explosive event sufficient to blow the plane in half.
Let’s talk briefly about what an APU is. The acronym stands for Auxiliary Power Unit and they are a necessary blessing on every aircraft and not for emergency backup or use. The plane has multiple engines and multiple generators on each engine and there is no shortage of electricity in this class of plane as long as it is flying. As soon as it lands and the engines shut off they have no electricity. With the sun beating down, the plane turning into parking, 300 people inside, if you don’t have air conditioning you will WISH YOU DID!!!.
Enter the APU. It’s actually a small jet engine that turns a turbine. It is started by a battery. And in this case has a battery that works with it. Aircraft systems are supposed to be 28v but this generator operates at 32v. Kind of like our 12v SLI batteries are actually 12.6. So the generator supplies the power, the batteries kind of buffer that from spikes etc, and if the APU generator fails you do have a few minutes of battery as well.
Table 2 on page 11 provides the timeline. The plane lands at 10:00 and 24 seconds. About a half minute late. What’s that about? Does anybody in this country know how to run a railroad.
Four minutes later they start the APU and it starts fine. Two minutes later they shut down the main engines and we are on APU power at 32 volts. Fourteen minutes later we have a decrease in voltage, followed by a jump to 45 amps of current INTO the batteries and from there all hell breaks loose.
The open circuit voltage of a LiCoO2 cell is considered to be 3.66volts, midway between two reactive peaks at 3.72 and 3.61 volts. It is normally charged to 4.4 volts and discharge is normally cutoff at 2.5v. And so at any voltage above 3.72, these cells are taking on energy. With 8 of them, 29.76 volts.
In operating at 32volts, the bus voltage of the APU, these cells are on trickle charge. So if we use them to start the APU, we do take energy out of them. But as we use them on the bus, we are charging them. Not with the BCU. And not as part of a charge procedure to 4.4v. But just trickle charging them.
In this case, the failure was after 14 minutes of this. Not terribly important. EVERY time this plane landed, these cells were trickle charged at 32 volts for some period of time that is operationally a variable. The time on ground from when they started the APU until they shut down the aircraft entirely and hooked up to ground power.
The planes ostensibly each had less than 100,000 hours on them. This is a cumulative effect where they are simply overcharging the cells A LITTLE BIT every time they land. Like being nibbled to death by ducks, it doesn’t show up until they burn. And now we know that happens somewhere within the first 100,000 hours.
If true, and they keep flying, they would have had more fires. They are grounded, at enormous expense. And their solution, if approved, could provide an even greater threat to life and limb by trying to contain it in a stronger can. A bomb if you like.
They basically are operating these batteries on a 32v bus when they have a fully charged open circuit voltage of 3.72. And so they charge WHENEVER they are in operation, in addition to when they are actually CHARGED.
That it is a “little bit of an overcharge” is no relief. The charge, and the damage, is cumulative.
So the fix is almost free. Go to 9 cells. And proving it is almost free as well.
Let’s connect one of the cells to a regulated 4.0v source and leave it. Preferably leave it outdoors. On a patio or concrete slab. With no flammables nearby.
I don’t need to perform the experiment actually. I left an Apple Laptop plugged into a wall socket for 4 months without ever turning on the laptop. The battery looked like Jiffy Pop and I found the laptop cocked over on this “hump” of swollen battery pack. Fortunately caught it before it burned down the house.
If you cook a 350F cake at 160 degrees for 4 months. It’s just as cooked. In fact it is overcooked. And NOBODY is accounting for this in laptops, powertools, or anything else except US. Our chargers terminate, and by terminate I mean they turn OFF. You have to disconnect them, and then reconnect them to the mains to get them to come up charging again. Good practice.
It is my belief that these cells are failing because they are being slowly and cumulatively overcharged repeatedly. At some point, they are sufficiently mechanically swollen and the SEI is sufficiently damaged to allow recombination and a thermal event. LiCoO2 cathodes are peculiarly susceptible to this as an ignition source.
The BMS is not the problem. The Battery CHarging Unit is not the problem. The cells are not the problem. The container is not the problem. The problem is they are overcharging my cells BY DESIGN in routine operation, and then they are going to blame the cells. I am deeply offended.
Shall they not cease, I will have Brain take care of them. And trust me, they do NOT want his undivided attention.