Given how popular (and damn near impossible to discern) unsafe and unprotected lithium-ion batteries are, the idea of making a homegrown power wall terrifies me in ways I can't even begin to contemplate.
An 18650 battery, if combusted, can produce hydrofluoric acid with just water vapor in the air (or the water in your lungs) if they use lithium hexafluorophosphate as the electrolyte. Hydrofluoric acid can and will cause permanent damage to lung tissue if inhaled, and nasty, nasty chemical burns to exposed skin. 18650 batteries are supposed to have several failsafes to prevent thermal runaway, but counterfeit batteries can lack those failsafes. Not to mention, in some laptop battery packs, the protection was pack-wide, and not in individual cells. Not knowing the chemistry of the cells or the state of the protection circuitry of the battery would make me very nervous. Maybe someone else can provide more detail on how to test them safely.
You have to realize that 1kWh is a lot of energy. Imagine being hit (in a world without an atmosphere) by a 60 pound suitcase falling out of an airliner flying at 45,000 feet. Same amount of energy as that battery pack. (Alternatively: the most powerful air conditioner you can buy running full blast for 30 minutes. Incidentally, that battery pack has as much energy as 100mL of gasoline, which is why gasoline-powered cars are so popular.)
When you put a lot of energy in a small space and release at all at once, it's called a bomb. Li ion batteries are small bombs that can be used to power electrical devices.
Some chemistries are safer than others, like A123 cells. The market did not care about safety and they went bankrupt. Energy density is king.
If I had to guess, it's someone who's like "I wonder how big that AC unit is" because the idea tickled their imagination, googles it, then finds that it doesn't quite match with the scenario given.
Building a pack from random cells like this is pretty much a cardinal sin with Li-Ion pack design. Cell resistance and state of health are carefully matched during pack manufacture to ensure cells charge and discharge evenly. Li-ion cells are pretty sensitive, so on top of that, large packs have battery management systems (BMS) to balance cell voltages and monitor for faults and damaged cells.
Attempting to charge a pack like shown has a very real risk of causing one or more of the cells to overcharge and breach, or worse, violently catch fire.
Yeah, very much. Some basic dilligence with a somewhat even set of cells will probably be harmless. I would not be afraid of that kind of homebuilt power-store, if some decent practices are followed. (But you should do your dilligence & check to make sure all the cells remain ok, as you start using this pack!)
This... this is an hellish unbelievable nightmare, is asking for something to go wrong. It is certianly possible to invest the time & effort to balance the hell out of such a chaos-pack, but for one's own sanity, using normalized cells will save a lot of sanity & rework, and drastically decrease risks of catastrophic f-up. Don't do like this person! Do build weird electric systems though, unlike OP's way too broadscale doomsaying fear-mongering.
Even just drawing in parallel from batteries like this is a fire hazard.
I think you've confused parallel with series; in parallel, the batteries will naturally balance themselves to the same voltage and discharge proportionally to their capacity. It's the series configurations that are riskier and need extra balancing circuitry.
Also, I'm pretty sure the bulk of fires are from the much more fragile pouch "lipo" cells, not hard-cased 18650s like these. You may note the extremely large number of lipo fire videos on YouTube as evidence.
On the other hand, 18650s tend to be quite a lot more robust and difficult to ignite:
I think @gravypod actually meant parallel connection, it might be quite dangerous in fact - you don't even need to draw anything from the battery to mess things up. If you connect a 100 cells in parallel and one cell dies creating a short in the circuit, the rest of cells will discharge through the shorted one. When you short a single 18650 cell, it outputs about 6-7 A of current on average. Multiplying by 100 gives you 600-700 A of current flowing through a single cell. That's why people use fuses to create parallel modules.
Yes that was mentioned in one of the maaaany EEVBlog videos iv watch that these fail closed sometimes as well as the recommendation that "you should never buy lipo cells without a built-in controller" comeing from a guy far smarter then I about electronics design. If he is afraid of it, it's for a good reason.
> Yes that was mentioned in one of the maaaany EEVBlog videos iv watch that these fail closed sometimes as well as the recommendation that "you should never buy lipo cells without a built-in controller" comeing from a guy far smarter then I about electronics design. If he is afraid of it, it's for a good reason.
If you're building a large pack you don't want cells with a built-in protection. You connect all the cells to a pack-wide BMS that keeps things safe.
The important thing is to never use unprotected cells without some protection circuit in the pack.
Wow, as someone uneducated about anything related to electricity and power, seeing the picture in the article really made me want to emulate the setup. Not anymore.
You joke buy many many people don't know how volatile battery chemistry is and how different charging of Lithium cells works.
I'd not have known how much of a bad idea this was had I not been a long time fan of EEVBlog. It's an entirely different world when touching Lithium cells. Even drawing in parallel from cells from the same batch can be a bad idea.
It's pretty safe so long as you use a battery controller with appropriate monitoring to prevent overcharging.
Recycled 18650 cells that have worked fine in laptops for years are very unlikely to suddenly explode on you for no reason. I'd be more worried about the mystery Chinese ones you can get new on the web!
Here's some Materials Science students at Oxford who built packs using recycled laptop cells. They used controller software that monitors the health of each individual cell:
>very unlikely to suddenly explode on you for no reason
It really depends on their condition. If you get some cells from a laptop and they've all fallen down to reading 1 volt, then yeah repeatedly charging them (especially at a high amperage) is going to be pretty dangerous.
I think your concerns are somewhat alarmist and tend to be a bit scaremongering. You may want to read an actual study about just how much HF is present in the fumes of burning cells. In this one, near the end they burned a whole laptop pack and analysed the gases that were emitted:
"As seen the yields of HF is much lower for the lap top cells, in fact the HF detected online was below the determined detection limit."
The graphs on page 101 also show HF concentrations below 4ppm --- and keep in mind this is basically all of the gases collected from the burning pack. Looking at the safety data for HF, the permissible exposure limit is only slightly lower at 3ppm and it's immediately dangerous at 30ppm.
In other words, it would be a bad idea to deliberately concentrate and inhale the fumes of a burning cell, but the HF probably isn't going to kill you. Some good ventilation of the area should be sufficient.
I had to come back because the paper doesn't specify the chemistry of the laptop battery; you can't use this as evidence to support anything other than caution around Li-ion batteries and cells.
And, more alarmingly, the part you omitted or glossed over: their plausible explanation for the lower HF detection (which, again, would depend wholly on the chemistry of the cells):
> the laptop cells exploded with liquid splashed on the walls in the equipment
Whoops.
Don't mess with these things unless you know how they can be safely tested, safely charged, and safely monitored. Li-ion protection failures (of various modes) were responsible for the Boeing Dreamliner fleet being grounded, Tesla refitting their battery packs (road debris caused a fire when it pierced a Li-ion cell), hoverboards being recalled, millions of dell laptops recalled after battery fires, and more recently, the Samsung Galaxy Note 7.
So, yeah. Safely tested, charged and operated is the key, and I wouldn't want to take that responsibility for a pack with as many batteries (and who knows which ones have their own standalone protection circuit in the cap) as in TFA.
And, more alarmingly, the part you omitted or glossed over: their plausible explanation for the lower HF detection (which, again, would depend wholly on the chemistry of the cells):
> the laptop cells exploded with liquid splashed on the walls in the equipment
The less electrolyte that gets burnt, the less HF generated, which I hope you agree is a good thing.
and more recently, the Samsung Galaxy Note 7.
How many actually burned? It's a very tiny fraction. Manufacturers recall precisely to avoid the sort of hysteria that statements like yours generate, and it's not any indication of impending doom. The aviation industry is particularly risk-averse, so it's not surprising. But I don't think the risk is very high compared to a lot of other things in our lives, especially when you consider all the lion cells out there of questionable quality in use and yet behaving themselves. China has taken the lead in making 18650s and other lion cells almost as easily available as alkalines, but reports of fires do not appear to have increased anywhere near in a direct proportion to that. (Reports of cells failing to meet capacity specs, on the other hand...)
By all means handle with care (as you should with any energy source), but lion cells are not lethal weapons that will instantly kill you at the slightest provocation.
Are you serious, or trolling? Weird question to ask outright, but I wanted to save time.
Regarding the battery exploding. Yeah, I'd agree that ... no! No, I don't agree at all! First, it exploded, so there's that. A cell in the middle of a multi-cell pack has nowhere to go. Second, it invalidated the experiment; they don't know how much HF would've been released in a real world scenario where the compressed cylinder of reactive metals and gasses is not free to "rapidly disassemble" into the air.
Now, moving on to Samsung's recall. How many failed? Sure, a small fraction. Do you know if yours will fail? How could you possibly know? Wait, I had something for this... oh, yeah. Safely test each cell.
Samsung didn't have that capability, their battery vendor didn't have enough engineers for all those house calls. You don't have that capability (you seem quite cavalier about the whole endeavor, so I'm guessing there). So, yes, since they can't know without testing, they recalled them all. What do you suppose they'll do then?
Probably, get together with the vendor, and test them.
Please stop suggesting to other readers that these things are perfectly safe when you know they're not. They're not tiny little bombs waiting to go off if you sneeze near one, but anyone using untested lithium-ion cells/batteries in untested configurations without a controller on each cell should be discouraged. We don't need a new category of Darwin award winners.
The PF6 ions in the electrolyte of some Li-ion batteries (not all use the same chemistry) will readily combine with hydrogen in the air to form HF if the battery is combusting. Cylindrical Li-ion cells, like those part of an 18650 battery, literally hold their chemistry under pressure, and a thermal runaway will normally cause the vents in the cell to release the pressure.
If that's happening due to a short (which can and does happen when a cell is charged improperly, or even put under resistive load after the cell is discharged below a safe limit -- copper or other ions, depending again on battery chemistry, can literally bridge across the electrolyte to short the cathode and anode) the result of the release of pressure combined with the reactive nature of now exposed lithium is a small rocket that shoots a plume of gasses like sulfur dioxide or hydrofluoric acid. Again, this depends largely on battery chemistry, and without an MSDS on these particular cells, there's no way to know. But there's a reason why any chemist you care to phone up will say, if you tell them you're in possession of a cylinder of a compressed compound that contains fluorine (it won't matter what kind, fluorine really really likes to bond promiscuously and with reckless abandon), they'll tell you that you're braver they are. Tell them you're intend to string up a few hundred of them and pass an electric current into them, and they'll call a hazmat team for you.
A casual stroll through forums frequented by flashlight fans (yeah, it's a thing) and you'll see plenty of demonstrations of what happens when a Li-ion cell begins a thermal runaway. And plenty of memorials to those who gave up their passion because they're on ventilators after experiencing profound damage to their lung tissues. Hell, ask the firefighters that respond to a fire at a battery recycling plant; granted, that's a bigger scale than one cell, but even one cell in a small room can, depending on its chemistry and how violently it's outgassing, emit enough HF to cause lung damage.
LiPo "pouches" mentioned elsewhere in this thread are much less likely to explode spectacularly before bursting, and much less likely to exhaust a plume of corrosive, toxic chemicals in their wake.
"A casual stroll through forums frequented by flashlight fans (yeah, it's a thing) and you'll see plenty of demonstrations of what happens when a Li-ion cell begins a thermal runaway. And plenty of memorials to those who gave up their passion because they're on ventilators after experiencing profound damage to their lung tissues."
I was interested to read about flashlight enthusiasts on ventilators, but all I could come up with was this:
Cylindrical Li-ion cells, like those part of an 18650 battery, literally hold their chemistry under pressure, and a thermal runaway will normally cause the vents in the cell to release the pressure.
Unlike what you seem to be implying, lion cells are NOT "under pressure" in normal use. They are at atmospheric pressure. The vent is specifically so in case of abnormal conditions, it allows any pressure to escape instead of causing the cell to explode.
result of the release of pressure combined with the reactive nature of now exposed lithium
In a lion cell, there is no "exposed lithium" in the metallic, highly-reactive sense. It's a solution of lithium ions in an electrolyte, whose only real hazard is its flammability.
I believe I know the flashlight forums you refer to, and the incidents of HF poisoning too. Those were lithium primary cells, which actually contain significant quantities of metallic lithium and are definitely more reactive than lion. Even the MSDS you refer to says "in spite of their name, these batteries do not contain any lithium metal".
The MSDS is also the easiest way to scare yourself into thinking everything is highly dangerous. For example, look at the ones for sodium chloride. By the way, I would not trust that one so much as it gives a boiling point for ethyl acetate, a liquid at room temperature, of -84C. Similarly, dimethyl carbonate is listed as boiling at +4C, when that's its melting point. Perhaps that's where your impression that the cells are normally under pressure came from?
Likewise, I recommend looking at the data on the electrolyte solvents:
1 Wh is about 0.9 kilocalories. So 1kWh is about 100g of butter. The energy expenditure of a human in one day is about 2kWh. 1kWh is about the energy needed to heat 10L of water from 0 deg celsius to 100 deg celsius.
It actually takes some extra energy to move from ice to water, and also some extra energy from water to steam, so your numbers only work out if you are strictly inside the 0-100C range.
It's interesting to realize thought that we programmers are code producing machines that run on roughly 2kWh per day. If we produce 100 lines of code in a day, that's 20Wh per LOC. Not bad, actually.
Nope. Heat pumps have a COP > 1. It's quite literally a pump moving heat from a cold place to a warm place. If it takes 1 kWh to pump 3 kWh of heat from outside your house to inside (in winter), you've heated your house by 4 kWh. Note that the electric energy used to drive the pump is transformed into low grade heat, which in this case is useful.
Good point! It kinda threw me off too, since heat energy isn't usually measured in kWh. But yeah, heat is a kind of higher-entropy form of energy than electricity, so you can, in fact, turn one unit of the latter to more than one unit of the former.
The "exchange rate" depends on the temperature; the maximum ("Carnot") heat pump efficiency is T_high/(T_high - T_low), where T_high is the temperature of the room being heated and T_low is the surrounding environment the heat is being pumped from (all on the absolute scale).
When T_high = T_low, the max COP is infinity, which has the physical meaning that "You don't need to spend any energy to keep a room the same temperature as its environment; that happens automatically."
It's an air conditioner in reverse. You run your AC at 2kWh (or so) and you've heated the outside environment by more than 2kWh; otherwise your house wouldn't be cold, energy and heat don't magically appear or disappear.
The book recommended here, David MacKay's Without the Hot Air really is remarkable, and I highly recommend it to anyone with an interest in energy, renewables, or nuclear (which MacKay somewhat grudgingly endorses), and the energy-intensity of modern industrial life.
Of the book, I find a few things particularly illuminating. For starters, it is rather UK-centric, though the concepts are of course generally applicable. Beyond that:
1. It goes through the major uses of energy in modern life. Getting a feel for the comparative magnitudes is quite useful.
2. It compares the various options for renewable energy. It turns out that there's a lot less energy in renewables than would be convenient. Wind and Solar are much of the easy stuff.
3. It mapps out both energy consumption and use by area. Realising how many watts per square meter are used and are available is useful.
4. He really presses the point that solving the energy conundrum requires large changes. Unplugging charging devices won't cut it. Hitting major consumption, especially transport, heating, lighting, and refrigeration, help a lot.
If you're interested in pursuing the issues further, I strongly recommend Vaclav Smil, whose books I've been going through. For a historical view, Smil's Energy in History, and the more recent two-volume book, Sources of Power, by Manfred Weissenbacher, explore how human history has been shaped by access to energy, from gatherer-hunter days, agriculture, coal, oil, and whatever comes next.
In a Nissan Leaf, you can get around 5 miles per KWh average. You can plug in the 1.1 kWh charger and realize you will get 5 miles per hour. If you use a bigger 6.6 kWh charger, it's 30 miles per hour.
Better yet, you can fiddle with the dash controls and get a power meter: acceleration takes 30-60 kW, but regeneration can make up to 30kw.
Like the OP says, this kind of thinking really changes your perspective.
I think the poster means the EPA test cycle for miles, they have a pre-defined "cycle" of acceleration, braking, coasting etc for Highway ratings and City driving.
just guessing, it's a standardized way to declare the driven distence, given all the variables.. such as: up hill, down hill, low/high rolling resistance surface?
Drivetrains on bikes can be > 98% efficient, so I assume you mean the human power output efficiency. But 250W applied directly to the crank via an electric motor should get you a good 40km/h for 4 hours! :)
I do not know why (and I do not have a strong opinion on this), but somehow this (relatively recent) tradition of naming units after people feels weird to me. I know we are already used to these units, but imagine the unit of mass was called Einstein instead of kilogram.
The Newton unit has been around since 1948. But that was hardly the first term to be named after someone: Watt has been used since 1882, Farad since 1881, Fahrenheit since 1776...
Note that the name of the unit is 'degrees Fahrenheit', but I agree - this could be something that started the tradition, which became stronger with new discoveries in physics and chemistry in 1800s.
I am sure Newton didn't name the unit of force after himself, so I would roughly place the invention of such possibility around early/mid nineteenth century, that is, about 200 years ago, which is when the research in such areas as electricity and magnetism, thermodynamics etc. was in full swing.
It doesn't make much difference whether you use Watts or Joules, since 1 Watt = 1 Joule per second, so both are just as convenient for electricity: 1 Watt = 1 Volt Amp, so 1 Joule = 1 Volt Amp Second = 1 Volt Coulomb (although amps seems to be more "normal" than coulombs).
It's a bit redundant to multiply power by time, when that's just an energy, but going between Watts and Joules is trivial (by design), so it's not too bad.
However, I can't think of a reason why hours would be easier to work with than seconds (as kiloseconds, if you want the same order of magnitude), especially regarding electricity?
It's because power needs are reported as watts, and impactful devices are typically utilized for several hours. It's much easier to do your monthly utilization napkin math in multiples of kilowatt-hours than of 3.6 megajoules.
I'm personally not a fan of the kWh, because it keeps being mistaken for kW, but I recognize that lack of need for a conversion factor help a lot of people make sense of energy.
My personal favorite is the quad: one quadrillion British thermal units of energy. It's used for whole-planet and large-nation scales of energy per year.
"How does it make sense to have 2 different units of time in there?"
This way it's easier to calculate how much energy your appliances uses. For example, if your vacuum cleaner is 1000W (1kW), running it for an hour will use 1kWh of energy which cost you can clearly see on your electricity bill. Unless you also want to re-lable all appliances, electric engines etc. in joules.. Sure, joule makes much more sense in physical calculations.
Then you compare with the amount of energy in a liter of petrol (~10kW/h). And you realise how we're living literally awash in crazy amounts of energy, equivalent to having all hundreds of slaves working for us 24h a day.
It's even more terrifying considering the fact that it took millions of years of to accumulate such amounts of energy in fossil fuels (it's literally sun energy laying underground). Sooner or later, we'll run out of it…
BTW, it's not 10 kW/h, but 10 kWh in a litre of petrol.
It would have been nice if the article's author had provided a short technical explanation of kilowatt-hour. A kilowatt-hour is a unit of energy, and energy is the time integral of power. It follows that energy in kilowatt-hours is the time integral of power expressed in kilowatts, i.e. one watt for a thousand hours, a thousand watts for one hour, and so forth.
I think that's pretty accessible and could only add to the article's value by answering the question, "What does 'kilowatt-hour' mean?"
Riding an electric bike has made me very aware of how much work can be performed with a given amount of power. Watt-hours aren't abstract anymore. 1kWh can get me and my stuff 120km easily if I pedal, probably closer to 80km if I used battery power only. Also instantaneous power, 200W is enough to move my bicycle, me and my stuff at a low cycling speed (15-20km/h) on the flat.
It really puts things in perspective when I look at a 1KW space heater. That's a lot of power!
I like his battery haha, colorful. Also nuts to test all those cells. Make sure they're all roughly the same performance. wonder how hard it would be to find a bad cell.
Probably the most obvious indication would be temperature. Assuming they are in parallel, it'd be quite wise to measure their discharge curves to only select similar cells - that would also highlight any bad ones.
Hmm. It did appear that the cells were all oriented in the same direction.
What a great idea I think although I wonder if it's a pain, cutting through that hard, protective plastic that these cells are in for laptop batteries. Unless you buy just the cells themselves.
I was just commenting today how approachable this book makes understanding the requirements to transition to renewable energy. I would warn though that the little bit of economics and the discussions about battery storage and solar PV at the end are very dated. The exponential drop in pv prices wasn't anticipated by the author. I think if modern prices were used the 5 models at the end would look very different.
he methodically tests old 18650 laptop cells, sorts them, builds battery packs and so on. He's converted a classic VW bus to electric power using his home-made battery packs and packs salvaged from wrecked teslas.
California, mostly, but not exactly a common thing yet. Will increase as more cars are totaled in rollovers or other crashes that mess up all of the body work.
Lithium-ion batteries require an expensive materials and overly complex manufacturing (Gigafactory). They are also dangerous as evidenced by the Samsung Note 7 debacle. Hopefully, there will be advances in making lead-acid and hydrogen fuel cells cheaper and safer. That would allow viable energy storage for solar PVs.
A single 12V 100Ah VRLA Deep-Cycle battery might be a better idea that's more familiar to a much larger chunk of the population, and you have the added bonus of having enough capacity to make up for the typical conversion/transmission losses in most scenarios, thus delivering a true kilowatt-hour!
Batteries have different discharge curves, influenced by many factors such as chemistry, age, etc. Often "packs" of cells are balanced in an attempt to minimize variance. Something tells me your 10kW "wall" may not perform as you expect it to (providing you are actually loading it for real use).
I know it's not as simple as "connecting 10k batteries together". It's an interesting topic anyways, there are a few guys already doing it with promising results (check HBPowerwall on youtube for instance).
BTW: I haven't even connected these cells together yet - even that is not as simple as it may look like.
I've been very curious for a very long time about people doing indoor bike workouts while storing the pedaling energy in batteries. Any of you do such things ?
One the one hand, this would be good for people who overeat (more calories than they usually expend). On the other hand, in absolute terms, it is just turning agriculture into a source of power very inefficiently (a methane digester would be much better).
Efficiency is not the goal, it's more symbiosis. Humans benefit from "wasting" energy through exercise. Instead of lifting weight and riding bikes, why not try to leverage it.
You would get very little out of it. Enough to run a dim LED for a couple hours at most.
If you did it in addition to (as opposed to instead of) your exercise then the energy in the food you ate is massively greater than what you can get from the bicycle. If you eat more as a result then you wasted your money.
It wouldn't be quite that bad. You're probably putting out something like 100-200W. Even factoring in generating and storage losses, you're looking at running a reasonably bright LED bulb for several hours on the proceeds of a 30-minute session. Still not worth the effort of building the equipment, of course.
200w seems for pro cyclists. 100w for above average sports fan. Today lots of computing devices fall under 10w power usage. Hence my daydream. For light or electric appliances that's another story.
I've seen similar things in Africa were people just attached bikes and drums together. Didn't have hot water though. For cold water + soap washing it's clearly a win.
An 18650 battery, if combusted, can produce hydrofluoric acid with just water vapor in the air (or the water in your lungs) if they use lithium hexafluorophosphate as the electrolyte. Hydrofluoric acid can and will cause permanent damage to lung tissue if inhaled, and nasty, nasty chemical burns to exposed skin. 18650 batteries are supposed to have several failsafes to prevent thermal runaway, but counterfeit batteries can lack those failsafes. Not to mention, in some laptop battery packs, the protection was pack-wide, and not in individual cells. Not knowing the chemistry of the cells or the state of the protection circuitry of the battery would make me very nervous. Maybe someone else can provide more detail on how to test them safely.