Hacker News Reader

Comment by drizzle87 | Hacker News Reader

I was quite excited by compressed air storage when Lightsail came into the picture almost 5-7 years ago. Sadly, they could not make the economics work and ended up pivoting into a company that sells carbon fibre tanks for high-pressure storage.

As someone who follows the energy space closely, Lithium ion batteries now have so much momentum and $$$ pumped into them via R&D work, it's going to be hard for competing energy storage technologies to catch up. Flywheel and hydrogen fuel-cell also often come up but I think we're at a point now where batteries are going to take off simply because of their wide area of applicability. With that economy of scale costs will drive down and even more research and $$$ will flow.

View on HN

>As someone who follows the energy space closely, Lithium ion batteries now have so much momentum and $$$ pumped into them via R&D work, it's going to be hard for competing energy storage technologies to catch up. Flywheel and hydrogen fuel-cell also often come up but I think we're at a point now where batteries are going to take off simply because of their wide area of applicability. With that economy of scale costs will drive down and even more research and $$$ will flow.

I've been surprised by the development honestly. 15 years ago I assumed heavy industry and automotive would make the big decisions about what energy storage tech we go with and the rest of the economy would follow suit and start working with either hydrogen fuels cells, supercapacitors, kinetic storage, or chemical batteries based on the R&D they did.

Instead it's been consumer electronics that led the way and heavy industry is stuck with chemical batteries because that's what worked best on an MP3 player.

Don't rule out hydrogen just yet - at least not for cars (and then using your car as a generator to power your home):

http://www.thedrive.com/tech/14431/are-hydrogen-cars-the-nex...

I'd be very surprised if hydrogen vehicles work out. Hydrogen is not a source but rather a store of energy like a battery is. There's a few practical and technical issues:

1 it would require building out entirely new infrastructure, running up against severe NIMBYism at this point. It's incredibly hard to even build new power lines right now. Try a highly combustible, very leaky gas.

2 it's far cheaper to just strip H2 from natural gas, which releases carbon (unless you capture it, increasing cost), so not carbon free unless you force using electrolysis from renewable power sources

3 Electrolysis efficiency is ~70%, fuel cell efficiency is ~60%, so round trip efficiency is <50%. Much more waste than a battery (~80%). You're both using essentially the same electricity to create the power, but hydrogen is wasting more of it, so will necessarily be more expensive

4 in vehicles, fuel cells end up charging the battery toi power the motor anyway, so the energy loss from the battery is a wash

The one thing they have is range and speed to refuel, but as batteries get cheaper and better that will be minimized before H2 gets big enough to take over.

There's lots of big companies with smart people who are still pushing for this so hopefully they know what they are doing (mostly in Japan it seems)

Possibly a bit off topic,but why do we charge the batteries at stations rather than switching them with full ones?

probably because we park our cars, instead of giving them away when we reach our destination, and getting one when we need to go again.

Have we figured out a good way to store Hydrogen that avoids leakage? I've heard that even thick-walled stainless steel vessels can leak ~1% a day.

Yeah, you coat the cell interior with platinum. Adds significantly to the cost and complexity of production though.

and use platinum in the fuel cells. the real-world problems are pretty obvious though - Pt is worth stealing and it's expensive 'cause there aint that much of it so you can't expect to use 100x as much as now and not expect the price to change. FC's are niche. Batteries won.

I guess storage tech is optional for auto and heavy industry - you can use diesel engines or whatever. For smartphones and the like though there is not much option apart from developing batteries.

It would really be something if that miniature nuclear reactor NASA tested ended up being viable for cars, realizing one of the dreams from the early nuclear age.

Given that even a small reactor still requires considerable shielding, generates waste that could be used for nuclear proliferation (if ammassed), and emits harmful radiation, I would say chances of a nuclear car for the every man are slim to none.

There's that, but on the other hand, small gasoline vapor explosions are quite harmful and require considerable shielding, and generate waste that is destroying our ecosystem.

The nuclear waste problem seems to be not so severe in the age where everything is tracked every second; and it doesn't need to be solved completely - just raise the cost of (illegal) waste extraction to be considerably higher than getting it from alternative sources, and you're set.

Yeah, but we are talking about thousands of small reactors driving along highways... that will never scale.

There's a video somewhere of a failing flywheel, it's a fairly energetic explosion. I follow the energy space, also, and ... while I didn't expect Li batteries to have grid scale roles, I didn't think flywheels were going to succeed there. They have properties more like ultracapacitors (discharge rate suitable for continuous backup, clean up messy grid sine waves). I thought pumped air/pumped water was more likely because it scaled up so well and was old tech. Fuel cells, too, seems to fall more into the backup/spot generation/remote generation+NG access niches.

Looking on the bright side, anything that drives Li efficiency/density research is probably a very good thing.

> it's a fairly energetic explosion

So is a lithium battery explosion. So is a gasoline explosion. So is a flour mill explosion. So is a wind turbine explosion.

Energy is energy, and losing control of energy is never great, no matter where it comes from. So while I catch your meaning, it might be better to phrase it with respect to how controllable that energy is.

From what I know of flywheel storage, the problem mostly comes down to keeping the wheel from coming apart, and containing it when it does. The nice thing about using flywheels for grid storage is that you can bury them and make them large. The earth contains your explosion risk and the lack of jostling means that your bearings don't need to take as much stress and limits that failure risk.

I would disagree with you slightly. Energy is not totally fungible. I can punch you in the face or shoot you with a gamma ray (edit: though of course, I would never do either), and it would have very different effects, potentially. Same for a fastball vs a small caliber weapon. Heat, light, and kinetic energy are important distinctions - particularly when you have a bunch of 30k flywheels in close proximity and one of them catastrophically fails. Now, I know that's the first thing someone would think about deploying a bunch of them and they would take precautions. Li batteries are less kinetic and more thermal and that's a bit easier to manage/less likely to cascade.

I thought the trend in flywheels was magnetic suspension and removing mechanical linkages? Admittedly I haven't kept up.

> that's a bit easier to manage/less likely to cascade.

What gives you that impression? Seems to me that it explodes if you contain it, and if you don't contain it, it can spout jets of thermal energy at virtually any angle. With flywheels you need to arrest it in bulk heavy objects that don't tend to sustain fire. That seems a lot simpler to me.

The bigger problems with flywheels are cost of manufacture and (depending on the technology used) efficiency for overnight storage.

> I thought the trend in flywheels was magnetic suspension and removing mechanical linkages? Admittedly I haven't kept up.

IIRC flywheels with limited motion gimbals (to reduce the tolerances on the wheel) are becoming more popular, still magnetic bearings.

What gives you that impression?

Insulation is cheap, effective, and very compact. And it's easy to transfer heat quickly, also (either via injecting lots of cold extinguisher or flush lots of hot oxidizing gas)?

Cool, thanks for the update on flywheels. I ... am not trying to create a false dilemna, here. Fuel cells for stranded methane deposits are great. Flywheels have outstanding responsiveness and energy density. Li / compressed air / pumped water et al scale well. They all fit into a more resilient grid storage strategy that permits a transition to periodic sources of input from non-renewable base load.

Except we've had a carbon free baseload tech for decades now.

It's kind of bizarre how nuclear had achieved a sort of counterculture renaissance. It has many, many drawbacks too numerous to go into here that are technical and sociological in nature.

And anyway, you tried to be too clever. I said "non-renewable" and not "carbon free" just to avoid this conversation. Unless you can start synthesizing utility grade quantities of well-behaved fissile material at a net energy surplus then it's not renewable even if we have decades/centuries of supply.

> Unless you can start synthesizing utility grade quantities of well-behaved fissile material at a net energy surplus then it's not renewable even if we have decades/centuries of supply.

Someday, even the Sun will run out of fuel. In the long run, we are all dead - unless someone figures out how to reverse entropy.

http://www.multivax.com/last_question.html

We still have 5 billion years to figure that out. Until then we should probably stick to worrying about energy problems of today.

edit: Love reading that short story, it somehow never gets old.

We don't need billions of years though, "hundreds of years" of fuel is hopefully long enough for us to figure out the tech that will get us through the next couple hundred years after that. And nuclear is fine for that. We're not talking kicking the can down the road by 20 years, we're talking hundreds or perhaps thousands of years of fuel.

Humans are plenty smart and the people of tomorrow will be better equipped to solve tomorrow's problems.

It's not even a matter of "ripping off a band aid" and paneling up the planet - solar panels aren't going to last hundreds of years either, we will be lucky to get 30 years out of them. Can we make more, sure, but we could also do nuclear and then make solar panels in 500 years when we're running out of fuel.

The biggest problems are that we need to come up with the political will to reprocess waste (extracting additional usable fuel and compacting the amount of true waste that needs to be disposed of) and then dispose of it in a proper repository rather than just letting it sit around on-site indefinitely Fukushima-style.

If you listen very hard, you can hear my eyes rolling through your computer. But I enjoyed reading the story and begrudgingly admire your pedantry on the matter!

Some animes have theorized a different method for reversing entropy. Like nuclear fission, we may be able to release a tremendous amount of energy by smashing the dreams of young girls and harnessing the power of their tears.

https://wiki.puella-magi.net/Thermodynamics

It's not renewable, but it is closer to carbon-neutral than fossil fuels. Advocates generally push it as a stopgap measure for climate change, after we transition away from baseload coal and before we transition to baseload renewables.

Batteries don't like insulation, in fact preferring active cooling.

Which is itself a good thermal-runaway damper, to speak to the second sentence of your second paragraph.

The other kind of firewall: https://en.wikipedia.org/wiki/Firewall_(construction)

>So is a flour mill explosion

Having witnesses a corn silo explosion, I was unprepared for the ferocity of that ignition.

>So is a wind turbine explosion

Off to YouTube...

The Hornslet collapse[1] is the one I was thinking about in particular. It is especially daunting now that I've toured a site for myself. Those things are huge!

[1] https://en.wikipedia.org/wiki/Hornslet_wind-turbine_collapse

Ever seen a tire come off of a semi?

Assuming the flywheel keeps its integrity it's much harder to predict the "blast radius" of where that thing is going to go.

Are buried flywheels used in practice? Is the flywheel axis aligned with the Earth rotation axis? Do they pull a vacuum around it?

Yes, in cases, to the first two. Containment is frequently buried concrete vaults.

Not AFAIK for the last, though precessional torque bearing load is a nontrivial consideration.

Thanks for this response. My untrained guess is that precessional torque would cause more loss than air friction, assuming the wheel is cleanly symmetric (ex. no bolt heads sticking out). It's fascinating how many paths there are to push something up various kinds of potential energy gradient in a way that it mostly stays there by itself but is still available to us to access in a controllable way.

Wikipedia claims flywheel loss rate circa 2013 of 5% per day. https://en.wikipedia.org/wiki/Flywheel_storage_power_system

That compares to recent estimates of Tesla li-ion loss rates at under 5% per month -- 0.16% per day.

Amber Kinetics is one company building fixed flywheel storage products. http://amberkinetics.com/

They have one 32 kWh, 5-ton, 98% steel flywheel installation on Oahu; pictures here: http://amberkinetics.com/hawaiian-electric-and-amber-kinetic...

Frictionaal losses are part of the consideration, but just plain bearing wear is a bigger one AFAIU.

Angular momentum in 100 kWh - multi MWh rotational systems is large.

Also flywheels don't give off noxious fumes during failure.

For now. If you want to use one for long term energy storage, though, you're going to need to invent some kind of room temperature superconductor and frictionless surfaces. Who knows whether those would be noxious or not.

Why? Magnetic bearings and copper work just fine...

Are you sure? Metal fumes and vaporized epoxies are usually pretty toxic.

...or during production (relatively speaking).

> So is a lithium battery explosion. So is a gasoline explosion. So is a flour mill explosion. So is a wind turbine explosion.

No, they're not all the same.

Compressed air, flywheel, flour mill - very dramatic events.

Lithium battery - much more mild, typically.

Gasoline - it depends.

The core insight for me was when Musk talked about how little the cost of Li mattered in the unit economics of Li batteries. It was something like 3% of the cost. The rest of it was mostly manufacturing costs. With so much room for improvements at scale and so many potential applications, it seemed like a sure winner.

That's a very useful insight. My "aha" moment came during a lecture given by DARPA at PARC about 5 years ago, outlining how wasteful current battery technology is in terms of density entirely because of safety requirements. They had at the time over a dozen different research projects on ways to improve density by changing how safety is achieved. It is only a matter of time until some of them get to market, and in fact I believe one of them has. They were all 2x-10x improvements.

There’s a kind of Moore’s law in batteries, but the multiplier looks more like inflation than 2x.

Any tech that falls outside of the growth curve seems to run into issues with production or cost that delay it until it fits under the curve. Something cheaper and easier gets picked first.

The first modern EVs had lead acid batteries. More sophisticated than your starter battery, sure, but lead acid all the same. Which is why Tesla was a big deal. We talked about LiPo for something like fifteen years before it showed up in consumer electronics, and then they started catching on fire.

All of this stuff is painfully slow. The big story in EVs is how crazy efficient the motors can get. A company I used to follow (whose name is escaping me now) had a motor that was 95% efficient in its sweet spot. They had scaled up the design to 100 HP.

Yeah electric motor efficiency is awesome. Back in 2005 I had a film-canister sized motor capable of 300W at 90% efficiency, and magnets and motor design have just improved since then.

Not only does materials overall matter little, but Lithium is also super cheap compared to the other materials in the batteries. There's a lot more of Nickel in a Li-Ion battery and it's much more expensive per volume and weight as well.

The batteries would be called Nickel-Carbon or something like that rather than Lithium-anything, if they were named by the amounts or costs of materials in them.

Why can’t we bury flywheels to protect from any explosions?

I've seen utility installations done like this. More expensive but safer.

Well yeah you can bury anything to keep it safe, but at what cost?

The utility flywheel installations I'm aware of are underground. More cost, but safer.

I suspect large-scale fixed installations are more likely to end up using some kind of flow battery:

https://spectrum.ieee.org/green-tech/fuel-cells/its-big-and-...

The big advantage there is independent scaling: peak discharge is tied to the total membrane area / pumping capability while capacity is tied to the amount of reactants that can be stored.

I'm pretty excited about the potential benefits of flow batteries. Hopefully they can get the tech to work economically for grid storage.

In a sense then you could say it's the Javascript of energy storage.

Not the best on technical merits alone, but good enough along with a massive force of investment and wide adoption.

I think the surname of John B. Goodenough, one of the early Li-ion pioneers, is apt for the status of the technology.

I'll be damned, this wasn't a troll. Today I learned...

https://en.wikipedia.org/wiki/John_B._Goodenough

Just yesterday I was talking to my friend regarding this. The key idea is, humans are really good at driving the price down to bare material cost via agressive production line automation. For e.g. without all the automation including subpart a single phone with all the technology would easily cost upwards of billion. So comparing ANY technology regarding their future prospects, we can easily deduct that which ever can be made fully automatically in production line will thrive in the end. ( Compare solar cell with all other renewable technology. Sooner or later energy produced by solar cell will bypass all other combined by tenfold or more.)

> are really good at driving the price down to bare material cost via aggressive production line automation

I've noticed this with steel products and plastics.

Brings up one advantage batteries have, they are small. That you need to make a trillion of them works directly towards the cost being a small percentage over material and energy inputs.

Aren't flywheels used in datacenters to cover the split second between when the power goes out and the generators kick on?

The various datacenters I've had such in-depth knowledge of over the years all used a big battery for that.

I toured a datacenter 5 years ago or so - and yes, they had a big flywheel for temporary power loses (until they could get the generators running)

That's only if they don't /cant run the entire DC of battery -whilst the gensets come online this is how telco exchanges / central offices work and important ones may have power supplied via different routes.

Some DCs do, but the need to keep them spinning 24/7/365 is a big maintenance issue, so given the advances in battery tech, I'd assume batteries would supplant flywheels for cost and safety reasons.

> but the need to keep them spinning 24/7/365 is a big maintenance issue

So is keeping batteries up to date. Many a failure has been attributed to stuff like:

- batteries gone bad without anyone noticing over the years (lack of acid, crystallization, loose contacts, dust in cooling components, ...)

- switch-over between grid and battery inverters fails somewhere

- some part of the inverters fail when they're idling for years and then have to go into full load suddenly

A huge flywheel only needs bearing lubrication, that's it.

For the most part, but every few years the bearings will need replacement. This means shutting the entire thing off and dismantling it for several hours. For redundancy, you therefore need multiple flywheels to cover for the maintenance periods.

Batteries can and indeed do go bad without warning, but current sophisticated UPS monitoring systems are able to detect failures, track battery ages and raise alerts (whether anyone actually acts on those alerts or not is another matter...). I've also found that lead-acid batteries last longest as long as there is some amount of power continuously trickling through them - I have an APC UPS at home with its original batteries from 2010, and they still provide ~1h of runtime at 20% load. And with most DC-grade UPSen, batteries can be replaced in situ without powering the device or any connected equipment off.

As of about 10 years ago our data centers always had 1 out of 4 generators configured with a big M/G flywheel on the output shaft. The flywheel momentum guaranteed generator start, I don't know the latency but it was on the order of a handful of cycles.

Some of them do, but the overhead is ridiculous. I don't have the data in mind, but I think that the NoBreak brand consumes 20kW+ just idling. It's ridiculous

Only some of them have flywheels. The large telco datacenter I visited had huge pallets of batteries to cover the time needed for the diesel generator to start.

Might be more a telco thing as they tend to regard serious outrages as a one every one or two generations as just about acceptable.

There is still one more possible alternative, if only we can get the physics to work: room temperature super conduction. That - if it ever comes to fruit - is going to cause a whole pile of revolutions and unlike most other ideas that rest upon unobtanium actually has a chance of success.

How would room temperature super conduction be used for energy storage?

Or would it just make all of the other technologies much more efficient?

Energy storage in superconducting loops. This is done on a small scale already for ride-through and power stabilization purposes but it is expensive due to the non-room temperature super conducting technology used which requires cooling down the superconductor.

Flow batteries are looking pretty good. Costs are coming down, and the very high (some say unlimited) cycle life could give an advantage over Li-ion for stationary applications.

On the other hand, used Li-ion from electric vehicles could be so widely available that they also win for stationary storage.

Interesting times.

My concern with batteries is we're so gung ho on producing these, we haven't stopped to consider the environmental impact from the manufacturing process or the incredible task of recycling these once they've outlived their usefulness.

It reminds me a lot of when people pushed plastic for everything in order to save the trees and reduce paper consumption. Some 40 years later, we're now trying to get off plastic and dealing with massive environmental issues surrounding nearly every kind of plastic.

>we haven't stopped to consider the environmental impact

We have. We've painstakingly analysed the energy and material inputs and the pollution outputs of the entire lifecycle of a lithium battery pack. After considering all these factors, lithium batteries remain an attractive energy storage option.

https://link.springer.com/chapter/10.1007/978-3-319-48768-7_...

>the incredible task of recycling these once they've outlived their usefulness

Recycling lithium battery packs is complex and hazardous, but no more so than a multitude of other industrial processes. There is already substantial recycling capacity, thanks in part to the EU's WEEE directive. Using existing technologies, only 1% of the pack mass goes to landfill. We can't yet recover all of the lithium in a useful form, but we can recover most of the cobalt.

https://onlinelibrary.wiley.com/doi/abs/10.1002/978111899197...

Back in the late 1800s, city planners met to try to come up with some plan for dealing with "the manure problem". Large cities couldn't handle the amount of manure produced by the horses in the city. They ended the meeting early, because they couldn't come up with any sort of plan.

Then cars came along and they were all like "Yay! The pollution problem is solved!"

I too wonder what the impact of so many lithium batteries is going to be. I mean, I know they are saying that they are extremely recyclable, and I hope that's true. But I wonder what the unintended consequences of this switch is going to be.

We can look at the current (used) battery market, the disposable Duracell thingies and such. I think there's a good recycling program for them, but even so, part of used batteries end up getting lost and in landfills and such.

https://www.youtube.com/watch?v=b9gnvZLkewg

(That's the silicon valley clip explaining the London manure problem being solved by the replacement of horses with cars.)

Maybe it's just me, but not a fan of bare links unless the URL is descriptive.

Not economical to recycle Lithium right now.

Recycled lithium is as much as five times the cost of lithium produced from the least costly brine based process. It is not competitive for recycling companies to extract lithium from slag, or competitive for the OEMs to buy at higher price points from recycling companies. Though lithium is 100% recyclable, currently, recycled lithium reports to the slag and is currently used for non-automotive purposes, such as construction, or sold in the open-markets. However, with the increasing number of EVs entering the market in the future and with a significant supply crunch, recycling is expected to be an important factor for consideration in effective material supply for battery production.

Closed loop recycling, where the recycled materials are sold back to OEMs, is likely to help against potential price fluctuation of metals or compounds. EV battery recycling is expected to play a significant part of the value chain by 2016 when large quantities of EV batteries will come through the waste stream for recycling.

https://waste-management-world.com/a/1-the-lithium-battery-r...

This guy actually goes to a lithium "mine" in the middle of the desert...

https://www.youtube.com/watch?v=0KX2qw79qpk

Do you think it’s better or worse than the initial mining?

Reply on news.ycombinator.com