Here's a fun bit: in the article they say that lithium-sulfur is hard to measure charge level for due to the voltage properties of charging and discharging.
"The upshot is that voltage is not a good proxy for the state of charge and, to make things even more complicated, the voltage curve is asymmetrical for charge and for discharge."
Since it would be bad if your battery suddenly died and you dropped out of the sky, they had to develop complex statistical and neural network algorithms to accurately determine state of charge to within a few percent. One black box for staying in the sky and another in case you end up on the ground!
Or you could do what they did on Apollo (forget if it was the CM or LM). They had the problem of measuring how much was in a tank, but the tank was in 0G, so a float is no good. The proposed solution was some sophisticated radiation based thing where they measured the attenuation of some radioactive source through the tank. This wound up being highly complex, and the solution was simply to have a reserve tank. When the main tank ran out, you knew you had exactly the amount in the reserve tank.
This is the solution used on motorcycles (which don't usually have fuel gauges). There's only one tank, but in the normal position you get fuel only to a certain depth.
As you're riding along, if you notice the engine running out of fuel, you reach down and flip the fuel switch and you have access to a little more fuel (and you know to head to a gas station soonish).
For motorcycles people often just use the trip odometer to estimate fuel used. Interestingly, this is similar to coulomb counting which is one of the more common methods of state of charge tracking. Technically coulomb counting would be more analogous to measuring the amount of fuel being pumped out of the tank.
Just for correctness: A vast majority of new motorcycles for quite a while have had fuel gauges. There are only a few brands who are obtuse about it (e.g. Aprilia).
More like sports bikes don't have fuel gauges (saves weight, amirite?) and make do with low fuel warnings instead, which serves the same purpose as a reserve tank without the need for a switch.
Neither my MV Agusta nor my BMW S1000R have fuel gauges. The BMW does the odometer trick itself though to estimate remaining mileage. It's often way out of course, depending on throttle usage.
(It's possible the BMW has a meter but it doesn't present a level, only miles remaining. I still think it's a style thing - something you don't put on sport bikes.)
Pretty sure all the S1000R years have fuel, but yeah it's as miles remaining. S1000RR didn't have anything until 2019.
It never really made any sense, but now that race bikes have tire pressure sensors and full color LCD instruments it is really spectacularly dumb to not have gauges.
Any of the popular crotch rockets have had fuel gauges forever. I think it's mostly just crazy race/rep bikes like BMW and Aprilla (so i guess the guys that only make sport bikes in 300 or 1000+) that don't have gauges
On fuel injected bikes, the fuel switch (on/off only) is on the fuel pump, generally bolted to the underneath of the tank, and not reachable while riding.
Another possible solution is to measure how much fuel (or energy) got in and how much is getting out. You know the nominal capacity and you know the flow rate.
That technique is called Coulomb counting and is the standard way of giving a precise battery charge percentage for lithium based batteries. The battery capacity estimate is improved over time based on the previous observed amounts of energy to enter or leave the battery.
You'd have to account for charge losses (Some energy ends up just being heat while charging) but that could probably be simply guestimated on battery temp while charging and some constant factor.
For example, assume a 90% charge efficiency for a battery at 20C... or whatever makes sense.
Same trick is used on some external tanks for outboard engines: the tank has a raised separator in the bottom so that reserve fuel will be caught away from the engine intake. To access the reserve fuel you need to briefly tilt the tank.
State of charge tracking is tricky even for lithium ion batteries. I worked on it for a bit as part of a solar car competition. Usually it involves a mix of coulomb counting and voltage measurements. IIRC, a handful of laptops and consumer electronics have trained neural networks to assist in estimating SOC.
As far as aircraft goes, it's sufficient to put a lower bound on the remaining charge. Realistically, the poor (abysmal, really) energy density of batteries pretty much precludes their usage in any serious aircraft. For sustainable air flight I'm more optimistic about syngas or hydrogen.
Furthermore batteries don't get lighter as their charge is expended, as opposed to fuel that gets burned off. This means a plane powered by batteries has to land with the same mass of fuel as it does when it takes off. Many aircraft are worse at decelerating themselves than accelerating. This means they can actually take off with more mass than they can land, thus extending their range. This can't happen for battery powered planes.
Cryogenic storage of hydrogen is less difficult than it seems. Because the engines are constantly drawing fuel from the fuel tank, the tank's pressure is constantly dropping which cools it down. Most of the difficulty around hydrogen storage is in long term storage where hydrogen permitting through the vessel is a concern. The main challenge for applying it to aircraft is being able to make a vessel light enough and in a form factor suitable for planes, namely fitting it in the wing.
Let's just use RTGs instead! A gram of Polonium-210 produces 140W of heat. It's also an almost entirely alpha emitter, which means that little shielding is required. At about a 10% efficiency of an RTG you could effectively get 14W out of a single gram of it. A 737-300 seems to require about 7-10 MW of energy at level flight. Put 1 ton of Pu-238 onto a 737-300 and you can run the aircraft (probably) for 6-7 months without refueling. Okay, you probably need a smaller aircraft, because takeoff takes more energy, but still!
While Polonium-210 is incredibly toxic, it does have a half-life of 138 days. If you do have a crash then in <8 years you'd only have about a gram of it left.
The main tricky part is that we produce about 100 grams of it a year. But what's a little scaling up for a startup anyway?
(I'm sure I made a calculation error here somewhere.)
I could make the argument that fuels that burn are unsuitable for automobiles which crash into each other frequently.
But yeah, I think it would need lots of development to get battery technology to do something like a ny-tokyo flight. It might only work for short hop service.
I guess that's how it has worked for battery electric trucks. The first one basically did short, known routes around town with lots of stop/start. That controlled the parameters and made it feasible. I don't know if battery powered trucks will ever replace long-haul trucks unless batteries get lots better or charging gets really fast.
> I could make the argument that fuels that burn are unsuitable for automobiles which crash into each other frequently.
Lithium ion batteries don't like crashing into each other either. The article mentioned that lithium sulfide batteries degrade differently than lithium ion, so fires as a result of heavy use may be reduced, but fires as a result of unscheduled disassembly seem likely, and could be worse than fuel --- you can dump fuel or circle to use it up, but you can't (probably) dump batteries, so you're always going to have lithium available. Not sure if the lithium sulfur compounds are less flamible than the compounds found in lithium ion.
You're correct that batteries are much more suitable for cars, especially commuter vehicle and things like garbage trucks that have relatively limited routes and sit idle for much of the day. There's why were seeing electrification take hold in cars, way before it takes hold in planes (if ever).
Ah, simply have separable battery modules that you drop from the plane as their charge is depleted. Put pop-out wings on them and use gps guidence so that they glide back to recovery centers on the ground, which you've had the foresight (and funding) to build in sufficient density along any likely route your plane might operate. Problem solved.
It could work, and it'd certainly give you a good 'ground truth', but there's a couple of reasons currently why it's probably not the best option. Batteries tend to wear out faster when they're deeply discharged, and when they're discharged rapidly. This approach would rapidly, deeply discharge some of your batteries instead of slightly discharging all of them. You could work around it with a sort of 'wear leveling' type battery management system but you'll still be reducing the life of the batteries compared with using the cells all at once.
In RAID 4 you had a dedicated parity disk. What if you allocated one of your 10 batteries as the "parity". It would not be used until the others had ran out of charge (as originally expressed above)
And to prevent uneven wear levels, each time you plug in to charge, the parity battery changes to one of the other 10 randomly (or on a pattern but the end result is the same).
So over time, assuming your prng was decent, you would have an even wear level.
You'd probably want maybe two batteries as your parity in case of failure but it would still work.
I'm on my first coffee and it hasn't properly kicked in but it sounds plausible in my head :)
This idea certainly has more legs than the first one. It's like a battery equivalent of the reserve tank that motorcycles (used to?) have. You run until the engine cuts out, then you flip the valve to switch to the reserve tank and you know you have at least 100km or whatever left to find a petrol station.
Reserving, say, 10% of your overall capacity would add much less strain on the system and still give you the most important part of the results. Conventional techniques such as Coulomb counting and estimation from voltage are still probably more practical but it's fun to consider alternatives. :)
Also, lithium batteries which have a higher total capacity can deliver more consistent current with less strain and voltage sag, so it's advantageous both for performance and longevity to have bigger cells.
If you have the ability to tell when you get somewhere in the region of 10-30% charge, you could always switch there and not deep discharge unless you really have to.
AKA joule counting- for normal li-ion it gets you to ~5% most of the time, up to 20% off at the start and end. That's assuming the voltage stays constant the entire time- in reality the first bit is at 4.2+ volts, and the last bit is down to ~3 volts. That's a 40% energy difference per electron that leaves the battery.
It's also one of the reasons you get electronics that die suddenly at 5%- current gas gauges usually account for it, but older stuff wasn't always good at knowing when the voltage would drop off. Nowadays (and always, for the most part) the sudden shutoff is because electronics often pull very brief power spikes that drop the battery voltage below the minimum voltage temporarily. The chemistry takes a moment to recover after that.
The problem with Li-S batteries isn't just that they have a goofy curve- that can be charted and saved, even as the battery degrades (Note- I'm mostly up on conventional chemistry. Don't know much about Li-S). It's more have a couple phases they go through during discharge. Impedance and other properties of the battery change, which changes the discharge characteristics of the battery, which changes the voltage. Proportionally, the swing in voltage is also larger (although this kind of thing is always changing, so I may be out of date).
There's also a small amount of self-discharge and parasitic reactions that will consume electrons, but that number is necessarily fairly small and predictable. The main thing is that 50% of the energy variance is in the voltage, and you need to know a lot about the current chemistry inside the battery (as well as the future load profile) to be able to predict the voltage that all the remaining electrons will have as they leave the battery.
Many battery monitoring circuits do this, sampling both voltage and current to compute power over time. With a suitable inductor to limit the rate of current change to be within the nyquist sampling interval of the monitor, you can pretty accurately measure charge going in or coming out of a battery. Combined with a model for the battery and you've got a modern battery monitor circuit.
You could use this technique to estimate the remaining capacity starting from a full charge / known charge, but not to arbitrarily measure the remaining capacity of the battery
It sounds like the voltage curve is all over the place as the battery phases through its chain of different chemical reactions, unlike a normal battery. I take that to mean there's no way to just measure the voltage at a given point of time to estimate capacity, hence why the statistical method was required.
It's a good thing to measure the instant capacity, but the voltage and current (and, hopefully, thermal output) are being measured during the whole flight.
This makes me wonder how accurate float bulb based fuel level sensors really are. This sounds like a major problem and the solutions are interesting but potentially even better than what we are used to.
My 1990 Toyota never read full. The buffer on the fuel sender was so extreme that by the time the needle made it to the "F" I had already burned 1/8 of a tank! My current car warns me when I have about 50 miles of fuel left, I wonder how much historical data it uses in that calculation.
None of my motorcycles even have fuel gauges. I just keep an eye on the odometer and when I stop to stretch my legs I give the tank a shake or a peek.
They're pretty inaccurate in light aircraft where you're often flying at an angle / slightly asymmetrically; in commercial airliners, under most flight conditions, they're generally within a couple percent. Aircraft will generally use fuel flow sensors and use those to calculate remaining fuel (by integrating the fuel flow over time) and float/capacitance sensors in tanks are used as verification / a sanity check (which can sometimes only end up being noticed inflight, eg https://www.flightglobal.com/safety/boeing-modifying-777-fue...)
> This makes me wonder how accurate float bulb based fuel level sensors really are.
They aren't. On light aircraft, the only thing they're good for is as a double check on a manual fuel reading (using a dipstick) or a time-based calculation, and to confirm during flight that the fuel cap wasn't left off. Beyond that, the needles bounce around so much during flight the only thing you can really verify is "yes there's some liquid there; somewhere between empty and full".
Many light aircraft owners have since retrofitted "fuel totalizers" which measure fuel consumption, and are manually reset by the pilot to a dipstick value when fuel is added. My group aircraft's fuel totalizer seems accurate to within at least about 10%. It can be calibrated better, but manual dipstick readings are only accurate to a couple of gallons anyway.
However, one key difference is that I do know, to within about half an hour, my fuel endurance before departure. I'd want the same from a battery.
I went on a ride in a jet trainer. The instructions I got were "We only bail if there's an engine fire. If the engine goes out an I can't restart it, we glide to one of these airports; this plane is an excellent glider." On bailing out "it came with ejection seats, but the government doesn't like us having explosives, so we use parachutes."
"The upshot is that voltage is not a good proxy for the state of charge and, to make things even more complicated, the voltage curve is asymmetrical for charge and for discharge."
Since it would be bad if your battery suddenly died and you dropped out of the sky, they had to develop complex statistical and neural network algorithms to accurately determine state of charge to within a few percent. One black box for staying in the sky and another in case you end up on the ground!