Australia first to test new lithium-sulphur batteries | Hacker News Reader

Australia first to test new lithium-sulphur batteries

newscientist.com

43 pointsntalbott54 comments

This is a much better article with more technical details yet accessibly written: https://www.newscientist.com/article/2228681-a-new-battery-c...

Publication link: https://advances.sciencemag.org/content/6/1/eaay2757

That first article title is so bad, mainly because those who have been following the smartphone industry for years know that if given access to such battery technology smartphone manufacturers would just add 8k screens, Threadripper-level CPUs, and hardware accelerators for mind-reading AI (so they can better target ads at you, not for anything remotely useful, obviously...) - all likely resulting in slightly less battery life than your phone has today, or at best a little more.

Changed from https://www.monash.edu/news/articles/supercharging-tomorrow-.... Thanks!

However, lithium-sulphur batteries may face similar ethical problems to lithium-ion batteries. The metal oxides in lithium-ion batteries are typically nickel, cobalt or manganese, which are expensive and diminishing in natural stores. They also have associated ethical problems: a significant proportion of cobalt is sourced by child miners in the Democratic Republic of the Congo, for example.

“In order to have much cheaper energy and more ethical batteries, we need a radically new energy storage system,” says Shaibani. The researchers will further test battery prototypes with a view to manufacturing them commercially in Australia in coming years.

It appears that Shaibani is saying that their new battery chemistry is an example of a radically improved battery that removes ethical problems while it improves energy density. The way the New Scientist article is written, that preceding paragraph makes it sound like Shaibani's new chemistry still needs improvements to remove cobalt.

There is already no nickel, manganese, or cobalt in this new lithium-sulfur cathode (nor in most lithium-sulfur cathodes). See Table S1 in the supplementary table for elemental analysis:

https://advances.sciencemag.org/content/advances/suppl/2019/...

I've read dozens of papers on Li-S over the years because of how much incredible potential the technology seems to have. The "dirty little secret" seems to be not the charge cycles (which are less than, but similar to, other batteries), but the large amount of electrolyte required, which reduces the effective energy density of the battery. There are all sorts of papers reporting high-capacity and durable cathodes but fewer that address the electrolyte problem, so I'm curious to see what their plan for that is.

A few papers about the electrolyte problem are e.g.:

https://pubs.acs.org/doi/abs/10.1021/acscentsci.7b00123

https://onlinelibrary.wiley.com/doi/abs/10.1002/adma.2017059...

For utilities I don't imagine that electrolyte is a big issue as there is plenty of space and you aren't lugging them around like you are in an EV. Maybe their plan is to use it mostly for grid power?

Sounds too good to be true! Did I read right? A fourfold increase in capacity with no downside? Going to production this year? How did I miss this until now?

Yes, definitely fourfold, shipping in 5 years!!1

Meanwhile Li-ion batteries were invented in seventies, it took 20 years to commercialization. First cells had 100W h/kg, and 30 years later we are slowly approaching 300W h/kg.

Sion Power has been working on this technology "since 1989".

https://sionpower.com/about/

It still only lasts a couple of hundred charge cycles before wearing out:

https://advances.sciencemag.org/content/6/1/eaay2757

"The cells are stable for more than 200 cycles..."

200 charge cycles at 4x the storage = 800 lithium ion charge cycles. That’s easily competitive. Put another way if your getting 300 miles of range a current EV that’s 300 x4 x200 = 240,000 miles.

Only if you change your habits to maximize the lifetime of the batteries. For example avoid charging the car every night and instead charge it for a few nights after it's completely empty.

Most batteries handle partial charging better than full charging.

But, assuming it’s an issue a hybrid design with say 150 miles of daily driving lithium ion and 150 x4 = 600 miles of extended range is another option.

That's actually a good idea. I was thinking supercapacitors for fast acceleration/recuperation and batteries only rated for sustained driving at cruising speed. Should last much longer.

As I understand it, this is a terrible thing to do to lithium ion batteries. NiCad batteries were like that, but li-ion shouldn't be allowed to go below 20% if you're interested in longevity.

I thought this was basically a non-issue as the battery circuitry would kill power before it got past that point. Or does that just stop it from emptying completely?

I think it depends on the application where the circuitry cuts off. Almost all applications will have circuitry that cuts power before the cell dies completely. You even have individual 18650 cells that have such a circuit built in (for flash-lights and so on)

But increased battery degradation starts way before you hit that limit. At least with cell phones, that tend to push battery cells pretty hard, you'll have pretty bad degradation when discharging to 0%. I think most EVs have a higher cut-off, and most people don't discharge EVs to near 0% anyway.

To add to what others have said with some numbers:

https://batteryuniversity.com/learn/article/how_to_prolong_l...

To distill it down to a single figure: If you drain the battery down to 40% capacity between recharges, you can expect it to take ~600 charge cycles for total capacity to drop by 30%. If you only drain it down to 90% between recharges, the number of cycles increases to ~6,000.

The article doesn't specify, but I assume that's using a slow charge. I am guessing that with a fast charge, which is what is implemented in most consumer electronics, the difference would be even more stark.

Then divide the battery pack into two (or more) packs. The battery controller then uses one pack to exhaustion, then switches over to the other pack. Normal charging will only charge fully depleted packs. This would reduce maximum range. So add in a charging override for an occasional max distance drive.

What? No.. if you go from 80% to 60% back to 80%, that's not considered a full cycle.

You seem to carry a misconception from an older battery chemistry (I don't remember which, but it was common in early cell phones), where it was supposedly better to discharge the battery completely before charging again.

Most chemistries are not like that, as far as I know. In fact, with li-ion it's better to charge every night, if your EV battery has a good buffer, or you can configure it to charge to 80% except for days where you'll actually need 100%.

Maintaining 99% for 200 cycles seem pretty good to me. Possibly better than Li-ion? It depends on how fast the battery degrades after that. But I'm pretty sure my EV lost its first 1% way before 200 cycles.

You misread that, it doesn't wear out after 200 cycles, it maintains 99% efficiency after 200 cycles.

Even if it did wear out after 200 cycles, you could still use it as part of a hybrid battery pack. Imagine 50% lithium ion, 50% lithium sulphur. If LiS has 4x the energy capacity of LiIon, you'd have 2.5 times the maximum range of a pure LiIon pack. But most of the time, you don't drive anywhere near the full range, so you discharge the LiIon pack first, and only dischange the LiS pack on those rare days when you need the extra range (or if you only very very rarely need the extra range, dischange the LiS half occasionally to avoid the LiIon half wearing out first).

Am I correct in assuming that a four-fold capacity in electric vehicles is coming? That will likely decimate oil dependency.

> That will likely decimate oil dependency

Until subsidies are increased...

Pet peeve of mine is how big business like to champion capitalism, but when they start failing they no longer like those rules and want government help to stay relevant and afloat.

Socialism for the rich, "the (not so) free market" for the peasantry.

Reduce it by a tenth?

I know you're referring to the original meaning of the term, but a 10% decrease in global oil consumption in the transportation sector would be a great accomplishment. Not nearly enough of course, but so far we haven't even been able to flatline.

Best not to assume anything about battery breakthroughs - there's one reported every week.

Many companies are on the fore-front of batteries...if this chemistry is legit and available, you will know when Tesla or LG chem or one of the big players buys this groups' IP.

Two thoughts of mine.

A twofold improvement in capacity would kill gasoline cars dead. That's 300 miles of range out of a 600lb battery. Adding up the weights of various components the electric car would weight about the same as a gasoline one.

While the failure modes of metal lithium anode batteries are terrifying that's probably okay for grid applications. Difference between a cell phone stuffed under a pillow and a battery in a concrete vault at at substation.

Lithium-Sulphur has high energy per kilogram which makes it good for transportation. Also high energy per dollar to manufacture which makes it good for grid storage (where weight and size don't matter too much, but cost does).

At the nominal rate of 750 amp hours per kilogram for lithium-Sulphur is well above normal lithium-ion batteries. But compared to gasoline, it raises the bar from 1% vs gas, to 2%. Do I have that right?

What makes electric cars viable isn't that batteries have anywhere near the energy density of gasoline, it's that if you use them, you get to replace half a ton of engine, transmission, alternator, fuel pump, emissions and exhaust with a <100 pound electric motor, and have that cost and weight budget to use for batteries instead.

This naturally makes battery improvements a huge win. If you double power density you can cut the weight of the battery by more than half for the same range, since not only do you get the same power from a lighter battery, now the car is lighter and requires less energy to accelerate.

That's a good point, I wonder what the adjusted energy density calculation compared to gas is when considering the amount of weight removed. Something like this:

new weight for energy density calculation = (total battery weight) - (ICE component weight) + (electric motor weight)

Nowhere near. The Tesla battery pack is 1000 lbs. There are entire cars that don't weigh that. Race car engine is 200lbs.

And battery packs have cooling systems too. So no savings there.

Specific energy (watts-hours per kilogram) is the entire ballgame with batteries and transportation.

Can you name a modern production car sold in the US that weighs less than 1000 lbs? Lightest available appears to be the Smart Fortwo at ~2000 lbs.

The curb weight of a Model 3 is in the same ballpark as a BMW 3 Series or Ford Taurus. Lighter cars exist but mostly because they're significantly smaller or slower or both.

> And battery packs have cooling systems too. So no savings there.

That's not quite true. At the very least, EVs require a much smaller radiator, if it has one at all. Some EVs don't have cooling at all (Nissan Leaf), although that increases degradation in hotter climates.

I think his point stands.. An ICE engine weight at least 200lb. With transmission it could be up to 600lb. The Model S engine is 70lbs.

You "only" need to halve the weight of a Model S battery for the drivetrain+battery to be in the same ballpark as an ICE drivetrain as far as I can tell.

You can look at actual numbers. Instead of a Model 3. Lets compare a Chevy Bolt with a Prius.

Chevy Bolt is about 3500lbs Prius is about 3000lbs.

Chevy Bolts battery weighs 440 kg or about 968lbs.

Reduce the battery weight by 1/2 and save 484 lbs. So the weight drops to 3016.

So point stands if the battery power density improves 2X then the weight penalty disappears.

Electric cars are about 3 times as energy-efficient as ICE cars. So it's not 1% to 2% but 3% to 6%. And with other savings (much lighter engine, no need for transmission) it's even better.

The problem is with degradation - it seems the new batteries only last 200 cycles.

Its all problems, so far. Weight, heat, size, charge cycles, range. Its only recently that its been even possible to make a practical electric car.

Don't fool yourself - there's a transmission element in an electric car too. The power has to get to the wheels, regardless of the power plant.

An electric car transmission is a usually a simple reduction (two gears). An ICE transmission is usually a six speed shifter, often with hydraulic controls. A world of difference in terms of weight and maintainability.

That's just talking gearbox. Then there's casing, shafts, mounts, cooling. The difference in number of gears is the smaller part.

This is Tesla mechanically: https://qph.fs.quoracdn.net/main-qimg-fa63245953aaf5f393e851...

Please don't pretend it's anywhere close in complexity and weight to the ICE. Of course if you include batteries then it's heavier. But that's why people are excited for any small change in battery energy density.

That’s really not that much worse than a lithium ion battery (300-500 charge cycles). Also, higher capacity means less frequent charging.

Gasoline has about 33 kWh/kg (118.8 MJ/kg):

https://en.wikipedia.org/wiki/Gasoline_gallon_equivalent

Unfortunately, internal combustion engines have a pathetic fuel economy since they run at low temperatures (around the boiling point of water). All heat engines are limited by the Carnot efficiency, which improves with higher temperature differential. In practice, other cycles like Otto, Diesel, Rankine and Brayton are lower than Carnot and improve with things like higher compression ratio:

  Carnot efficiency = (T.hot - T.cold)/T.hot
  where T is in Kelvin

https://en.wikipedia.org/wiki/Thermal_efficiency#Carnot_effi...

A low compression, naturally aspirated engine running at room temperature with nothing done to improve fuel economy runs at (373.15 - 298)/373.15 = 20% efficiency. I've heard figures as low as 8% for rubber meets the road efficiency in older passenger cars, which I believe, since we drove a ’68 Cadillac that got 5 mpg back in the 90s when gas was under $1 per gallon.

The best modern high compression engines typically achieve 25-30% efficiency at best. So I figure there are about 8-10 kWh/kg (28.8-36 MJ/kg) available in gasoline with modern vehicles. Cars built before ‘70s efficiency standards would be more like 2.5-3 kWh/kg (9-10.8 MJ/kg).

Unfortunately, it's not just that people don't care how ridiculously inefficient their vehicles are, it's that politicians corrupted by the fossil fuel industry and vehicle manufacturing lobbies never stop conspiring to lower efficiency standards:

https://www.vox.com/2019/4/6/18295544/epa-california-fuel-ec...

But I digress.

Electric motors typically run at about 95% efficiency, so we can probably assume 90% efficiency to the road. That’s over 10 times more efficient than classic cars!

Looks like Tesla lithium ion batteries are 0.254 kWh/kg (0.914 MJ/kg):

http://theconversation.com/teslas-batteries-have-reached-the...

Which is very close to the theoretical ideal for lithium ion of 0.294 kWh/kg (1.058 MJ/kg):

https://en.wikipedia.org/wiki/Energy_density#Tables_of_energ...

I'm having trouble finding energy densities for the new lithium sulfur batteries:

https://advances.sciencemag.org/content/6/1/eaay2757

https://advances.sciencemag.org/content/advances/6/1/eaay275...

I'm going to use their low number of 1200 mAh/kg, working between 1.7 and 2.5 V, so averaging 2.1 V (which is very inaccurate without integration), we can call it about 2.520 kWh/kg (9.072 MJ/kg). That would be about 10 times denser than Tesla batteries. Maybe they are estimating half the density in the real world due to packaging or something, in order to arrive at their "5 times longer battery life" headline.

So anyway, the real numbers are:

  Gasoline       33 kWh/kg    118.8 MJ/kg   (ideal)
  Gasoline       8-10 kWh/kg  28.8-36 MJ/kg (actual for modern vehicle)
  Gasoline       2.5-3 kWh/kg 9-10.8 MJ/kg  (actual for pre-70s vehicle
  Lithium sulfur 2.520 kWh/kg 9.072 MJ/kg   (ideal)
  Lithium sulfur 1.260 kWh/kg 4.536 MJ/kg   (actual)
  Lithium ion    0.294 kWh/kg 1.058 MJ/kg   (ideal)
  Lithium ion    0.254 kWh/kg 0.914 MJ/kg   (actual for Tesla)

My numbers might be off by a fair amount, but the important thing here is to think in orders of magnitude. Lithium sulfur is halfway to the energy density of classic cars and aircraft, with all the positives, like electric motors having 10 times the power as gas engines by weight, much higher torque, and substantially higher endurance/simplicity.

Gas engines with 40-45% efficiency are possible [1].

[1] https://phys.org/news/2019-06-efficiency-gas.html

Hey thanks for that! I didn't know that methane has low knock so could be thought of as a high-octane fuel (which allows for higher compression which translates to higher thermodynamic efficiency). That has promising ramifications for a future methane fuel economy, where hydrogen from high-temperature solar thermal electrolysis is combined with a carbon source (perhaps CO2 from the air) to create methane. Propane would be ideal due to its low pressure storage if its conversion from methane can be scaled up. But compressed natural gas (CNG) is fine as well with methane and the gasses have similar properties.

I was thinking about what I said about the Carnot cycle and maybe it wasn't quite accurate. I tend to think about the world through a first-order effects lens. So the easiest way to explain why a turbine is usually more efficient than an internal combustion engine is that the turbine runs at a much higher temperate.

But the gasses in an internal combustion engine can reach a fairly high temperature as long as it's beneath the sag temperature of the metal block (otherwise you get warped valves). There was a lot of nonsense in engines before fuel injection attempting to prevent preignition when the mixture passed by the valves (in order to run as lean as possible, which caused excess heat) that I always thought was pretty silly.

Also there was a lot of work in the 80s and 90s to make ceramic engines in order to run at a higher temperature that never went anywhere as far as I can tell. They would have been lubricated by graphite and basically last forever. I think they were abandoned due to brittleness, but they would be great today with a continuously variable transmission or as a generator running at constant RPM like with locomotives.

Ceramics never made it into internal combustion engines, but they did make it into jet engines. In particular the whole 737-MAX debacle is because of the fabulous new engine built by GE, which Boeing simply thought they cannnot not install it on their 737 (unfortunately, they did a patch work at that).

Here's a quote from [1].

"These “super ceramics” are as tough as metals, but they are also one-third as heavy and can operate at 2,400 degrees Fahrenheit—500 degrees higher than the most advanced alloys. This combination allows engineers to design lighter components for engines that don’t need as much cooling air, generate more power and burn less fuel."

2400F is about 1600K. If you plug this into the Carnot efficiency formula and use a T_cold of about 220, you get something like 86%. If they ever find a way to replace the silicon carbide they currently use with hafnium carbide, they can reach 90%

[1] https://www.ge.com/reports/space-age-cmcs-aviations-new-cup-...

Four decisions to make on cookies before being able to read the article? Are you trying to break the internet?

I'd recommend using using "I don't care about cookies" and "Cookiebro" to restore your sanity. Basically you want to automatically accept all cookies, say yes to all cookie banners, then remove all cookies when you close the browser except the ones you explicitly want to keep (for staying logged into certain sites). Those two firefox extensions facilitate that.

I'm using just uBlock Origin and since I haven't seen any cookie dialog I assume it did the job just fine. YMMV.

Unless you're on mobile and then you're helpless and it's even worse because the dialog takes up 30% of the screen, and the top sticky ads take up the top 15%, and then you have autoplay video ads at the top of the page filling up the rest of the viewable area, then you go to try and scroll the page and then a new well engineered ad has a delayed load to pop up right where you're about to click....

> Unless you're on mobile and then you're helpless

uBlock Origin is very reliable on my Firefox on Android.

You know, Safari and Chrome are not your only options on mobile.

> Basically you want to automatically accept all cookies

I don't.

funny how I saw many posts about Amazon fires and while Australia is burning to the ground(which is showing to be much worse than Amazon fires), I can only see a post about some battery; this makes me question this website..

It’s Monash University, not a news website, so doesn’t seem odd to me.

First of all, it's hacker news, not a general news site. And while the width of breadth of the type of news you'd find on here is larger than just coding and other hackery, it's still at least usually based in technology. Like batteries. Still, the official guidelines say anything "a hacker might find interesting" and we seem to be a environmentally-minded bunch, so maybe it would fly.

Second, it's based on use submissions. You think we should know something interesting about the AU fires? Submit a good article about it I guess. If I and others learn something interesting, it'll get voted up. That's how it works.

Reply on news.ycombinator.com