Hacker News Reader

Comment by epistasis | Hacker News Reader

People also severely overestimate the cost of battery storage today, and underestimate the cost curve of how it's decreasing.

For example, for batteries charged directly from a paired solar installation, the cost is estimated to be $85-$158/MWh. [0] The rest of the solar instal costs $28-$37/MWh, unsubsidized. If 3/4 of energy consumption is from the primary solar install, and 1/4 from storage, then you have a firm energy source at a cost of $42-$67/MWh. The national average for generation cost of electricity is something like $50-$60/MWh, (with the rest of the average price of $130/MWh going to transmission and distribution).

So today, solar plus storage offers cost competitive, zero carbon energy in the average US location. And costs are going to drop by 50% over the next decade or two.

We have a really bright energy future, where energy is cheap and plentiful and carbon free. And the faster we build new renewable resources, the faster the technology gets better, and the faster the prices drop.

Slowing down the deployment of renewables is a huge risk that makes us all poorer, and also happens to make our climate challenge worse. We should be doing everything we can to massively expand solar wind and battery production capacity and deployment, as fast as possible.

The tech is ready, it just needs to be deployed.

[0] https://www.lazard.com/perspective/levelized-cost-of-energy-...

View on HN

Diversification already has lingo for this in the 21st century:

1) Blue sky innovation - most intermittent renewables. They have toxic downsides too.

2) Gray sky innovation - Oil, gas, nuclear, etc fit perfectly here. Much of it is tried and true technology... And yes, it has downsides too

3) No-sky innovation - With space exploration, energy on the moon, mars, spacecraft, etc may be better served with laminar flow rotors using light/lasers... Just another option with diversification.

It looks like stats on 50% of the plastic on the planet have been created after 2006 seem legit, with the trash heaps in the oceans (not to mention plastic breaking down into water system contaminates). Battery technology recycling just looks like another source of toxic groundwater for future generations, since recycling has been gibberish with pretty words thus far.

The reuseable rocket technologies mean that battery technology, nuclear waste, etc can be jettisoned towards the sun after their useful life and that landfills may be a thing of the past.

Know your innovations and how well diversification copes with the farce of "one size fits all". Four seasons still exist in many places.

> for batteries charged directly from a paired solar installation, the cost is estimated to be $85-$158/MWh

This is theoretical. When I got home solar quotes, the difference before incentives between an install with and without 27 kWh Tesla backup was over $1,300/kWh.

$85-$158/MWh is not theoretical, this is real life utility scale installations.

Tiny tiny residential scale battery installations have that 5x markup that you are observing.

If you were to build your own, it's not hard to get to $300/kWh-capacity, all in, these days, even for small installs of a dozen kWh, using Chinese LFP batteries. But you won't get any of the support that your local installer gives you, and you had better know what you are doing really well...

> $85-$158/MWh is not theoretical, this is real life utility scale installations

Do you have an example? The actual cost in 2019 in the U.S. was $300 to $500/kWh [1]. The cost for the batteries alone was $190, but that’s akin to counting the cost of nuclear power by the fuel alone.

[1] https://atb.nrel.gov/electricity/2021/utility-scale_battery_...

I think you are comparing two different things. The CapEx and installation costs are typically measured in $/kWh for capacity of the battery (though some costs scale by the $/kW, for example the inverters...).

The cost of storing and delivering the energy from the battery over its lifetime, is refered to the level used cost of storage. This includes operations costs, debt service, etc. is measured by the cost of delivering energy to the grid, and grid-scale costs are typically measured in $/MWh (as opposed to residential costs typically quotes in $/kWh).

Yikes, autocorrect really jumbled my second paragraph, which wasn’t very clear to begin with.

I’m contrasting the caped with the *levelized cost of storage, which though it can be measured in $/MWh or $/kWh, is measuring something completely different than the capex. And the levelized cost of storage is the relevant figure for comparing to electricity sold on the grid.

> CapEx and installation costs are typically measured in $/kWh for capacity of the battery (though some costs scale by the $/kW, for example the inverters...)

How would you describe Table 1?

Exactly as it is labeled and exactly as quoted, capex measured in $/kWh and $/kW.... what is the confusion?

Where are the terawatts of batreries coming from? It will be decades before we have the manufacturing capacity to make that many batteries, assuming we have the natural resources to make that possible.

Batteries are reasonably priced per unit, but they do not scale reasonably. Batteries will not provide grid scale storage in our lifetime, IMO.

I think the concept of grid scale storage is misguided Fission (and hopefully fusion) can provide base load at all times, with as much wind & solar as needed to provide for variable load.

> Where are the terawatts of batreries coming from?

The usual place: factories.

Which are getting built as fast as people can get the investment money and the planning permits.

> It will be decades before we have the manufacturing capacity to make that many batteries,

Even if that was correct (from what I've seen, half of what we need is already being planned and is scheduled for completion before 2030), even then it would be fine, as we only have a critical need of them when renewables become a dominant part of the power grid.

> assuming we have the natural resources to make that possible.

We do, despite the memes.

> I think the concept of grid scale storage is misguided Fission (and hopefully fusion) can provide base load at all times

If the whole world did that with fission, we'd run out of fissile materials pretty fast.

> > assuming we have the natural resources to make that possible.

> We do, despite the memes.

Markets disagree: https://tradingeconomics.com/commodity/lithium

No, that's not what markets are saying. They are saying "build more extraction capacity" but they certainly are not saying "there are fundamental limitations on lithium supply"

Right, and unless we are able to build more extraction capacity there is indeed going to be a fundamental limitation on lithium supply. "Build more mines" is a lot easier said than done. The nature of mining is that the most accessible reserves are exploited first, and more and more difficult to access sites get developed as scarcity grows.

The common discourse around lithium ion batteries doesn't comprehend the idea of input limitations. If they were taken back to 1920, they'd look at how a car cost $100,000 in 1910 and $10,000 in 1920 and conclude that cars will cost $1,000 in 1930, $100 in 1940 and $10 in 1950. They'll just build more iron mines and steel mills. That's not how it works: mining isn't as amenable to order of magnitude increases: A car cannot cost less than it's constituent input materials, and mining and metallurgy can't deliver at that scale.

> The common discourse around lithium ion batteries doesn't comprehend the idea of input limitations.

I'm not sure I agree with that at all. Even Fox News reported on the IEA's "we need more mineral extraction than fossil fuel extraction" report from a few years ago, because my anti-wind-power father got on the phone to tell me as soon as he heard it, as if it would surprise me at all. Everybody knows we need this stuff, and I have never encountered a person that disagrees in the least.

Cost drops for lithium ion batteries are not coming from lower input costs, and in fact input costs are going up.

Batteries are getting cheaper because of process improvements, using less materials, less labor, etc. The same thing happened in solar, cost drops come from improved tech that use less material and manufacturing improvements, not from cheaper inputs.

The story of industrialization is a story of innovating around materials shortages, finding alternatives, and continually outcompeting the older less innovative companies. I don't know why problem suddenly forget this when it comes to renewables.

For the third time, battery prices aren't going down anymore. They've actually started to rise. You're right that improvements in chemistry and manufacturing means that battery costs of 2018 were nearly two orders of magnitude lower since 1990. But after that the cost history has been mostly flat. And in recent years, following large increases of input material costs they've actually started to creep upwards. Improvements in manufacturing aren't keeping up with the current costs of inputs.

Again, were going down years ago but aren't going down now. So the entire premise of this comment is on a broken foundation.

Wait, are you seriously asserting that because they have risen in price now, that's evidence that batteries won't continue their march downwards over the next decade? Would you actually be willing to put money behind that assertion?

As new markets open up and new applications become feasibly, prices will experience temporary spikes due to supply constraints, but it would be foolish to think that these somehow stop the future drop in prices. You see it all over the drop in solar prices; every temporary price increase results in a huge influx of capital, new production capacity, and improvements to manufacturing tech that comes with that.

Yes, because the bottleneck of lithium battery production has shifted to extraction. You are correct that the higher prices might justify more expensive extraction options, but said options are only profitable if the price remains high. This is why cars don't cost $10: the price of steel would have to be pitifully low, and no mines or steel mills can profitably operate at such a low cost.

Solar, similarly, isn't seeing the same order-of-magnitude improvements anymore. It was, back when it was new technology and big improvements could be made. But now the cells themselves are near the thermodynamic maximums: there's fundamentally only so much energy you can store with a gram of lithium. Consequently, the price drops are miniscule compared to what they used to be.

At present the grid/stationary storage battery market is essentially a sideline for vehicle battery makers.[1] Vehicle batteries have to balance between energy density, power performance, tolerance for temperature extremes, structural factors (vibration), safety, cycle life, and cost.

As stationary batteries become a market of their own, some of these requirements can be relaxed; engineers can focus on capital cost and cycle life. (This is already happening with early-stage technologies like carbon polymer, redox flow, and molten salt, which are already competitive with lithium chemistries.)

This allows, and will almost certainly result in, speeding up the rate at which battery costs fall.

1. Similar to the situation for PV cells up to about 2004 - 2005: they were a sideline business for wafer makers focused on logic semiconductors. After crossing over to having their own silicon production (by 2009) the fall in PV cell costs accelerated.

On PV price drops, they are pretty much on trend, using Wright's Law.

There's about a 16% reduction in cost per doubling of cumulative production.[1]

If announced capacity expansions and renewables commitments come true, we can expect four cumulative doublings of installed PV modules in the next decade, which drops their price by half.

However perovskites could gazump that. If their calendar lifetime can be made acceptable (currently around 10 years), they offer capital PV module costs about a tenth of today's.

Also a version of Amdahl's law applies. Modules are now under 20% of total system cost: cost like the inverters, the connection to the grid, earthworks, structures and wires, installation labor, and so on, add up.

Putting batteries on the PV farm means most of these costs are only incurred once, for much more electricity produced.

1. https://www.cell.com/joule/fulltext/S2542-4351(22)00410-X

> For the third time, battery prices aren't going down anymore. They've actually started to rise.

What's your source?

This was a surprising claim, so I tried googling, but none of my first 50 or so results were more recent than by 2020.

Google seems to have a massive inability to deliver lithium ion battery information. It's nearly impossible to find recent articles I know about unless I bookmark them and skip Google entirely.

That said, I don't have the numbers whining battery price increases, but I would expect that when Lazard releases 2022 data in the coming weeks it will show an increase in cost over 2022. There is absolutely massive demand, and supple chain shortages for all sorts of parts on battery installs which is increasing costs.

And all the car manufacturers are about 3-5 years behind where they should be in securing their own production capacity, leading to massive shortages of batteries for cars. And since cars are a higher margin product than utility installs, the car market is going to get served before the utility scale storage market when it comes to allocating limited production capacity. (That said most utility storage is LFP, which is not usually the preferred chemistry for US cars where customers still want massive batteries).

There was a story about this three or four weeks ago.

There is a temporary shortage of lithium refining capacity, taking in lithium ore from the mines and producing lithium carbonate /lithium oxide in the purity required for battery manufacturing.

That drove up the price of refined lithium carbonate and therefore finished lithium cells and battery packs.

Note the "temporary". It's fairly quick and easy to expand refining capacity, but it had been in oversupply for a long time, so people were caught napping.

There may soon (2 years?) be a shortage of lithium mining capacity, but miners are jumping on the bonanza, so I don't expect that to happen.

There is conversation about whether we have sufficient extractable lithium reserves to replace 2 billion ICE cars. Do you we have enough lithium to do that and back up the entire global grid?

For currently identified reserves of lithium perhaps, but:

1) we stopped putting much effort into looking when we had more than we could mine — there's not much point looking inside mountains when there are known huge salt lakes where evaporation has conveniently concentrated it on the surface.

2) the rate of discovery (in the "you're not literally tripping over it") category implies there is much more that is still easy to find. The estimated total yet to be located is far more than we need.

And 3), batteries aren't reliant on lithium, lithium is merely convenient and cheap. Sodium-ion batteries are available as an alternative.

Lithium can be obtained from plain old seawater. The issue is that the concentration per liter is low so it takes a lot of seawater to obtain useful amounts of lithium -- it concentrates out just like any other salt, and as we know seawater is saltwater.

But these processes are getting more efficient. And you can also recover other salts from seawater that may also be commercially useful.

This is a planetary scale resource that can be "mined" from just about any coastline anywhere in the world.

> despite the memes.

As far as I can determine the typical commenter believes that a lithium-ion battery is just solid lithium. I believe this is a result of a sustained campaign of disinformation driven by fossil fuel interests.

In reality, a standard 18650 cell contains less than a gram of lithium.

Who cares about breakdown by mass. Break a battery cell down by cost: https://www.visualcapitalist.com/breaking-down-the-cost-of-a...

The cathode alone contributes to half of the cost of a lithium ion battery. Shortages of lithium are indeed driving up the prices of batteries. If doesn't matter if a small amount of it is necessary, if obtaining even that small amount is expensive.

And yet, all battery makers are continuously reducing the need for expensive materials in their cathodes, because humans respond to incentives. Never bet against human ingenuity.

And yet, battery costs are still rising: https://asia.nikkei.com/Spotlight/Market-Spotlight/Battery-c...

Ingenuity can't change thermodynamics. There's a minimum amount of lithium necessary to hold a charge. Realistically the only reliable path of grid scale battery storage is a different batter chemistry, but none has proven as effective as lithium ion so far.

CATL seem to disagree with that last point https://www.caixinglobal.com/2022-10-25/catl-aims-to-mass-pr...

This is an entirely different battery chemistry, it's not even using lithium.

Which is enough to invalidate what you wrote:

> There's a minimum amount of lithium necessary to hold a charge. Realistically the only reliable path of grid scale battery storage is a different batter chemistry, but none has proven as effective as lithium ion so far.

The point is that the big order-of-magnitude gains in lithium ion batteries during the 1990s is not possible today, since they're already pushed near their limit.

Given that the cement talks about a minimum amounts of lithium to hold a charge, good faith readers are capable of comprehending that this is talking about lithium based battery chemistries.

So what scales better? Off river PHES is cheaper, but limited in scope. Electrolysis works but only for long duration. PWRs have no fuel supply if expanded. They also cost more and the historic highest construction is not to the same scale as the current renewable + lithium battery supply chain.

The minimum amount of lithium necessary for utility-scale batteries is zero because many of them are made of lead-acid batteries, because nobody gives a damn how bulky and massive a battery is in fixed service.

This has nothing to do with thermodynamics and everything to do with you parroting anti-renewable propaganda.

Lead acid batteries have much lower life cycles. Often around 300 cycles at 80% depth of discharge. True, bulk does not matter. But replacing your batteries every year absolutely does. That's why most battery storage projects use lithium batteries.

So then it isn't a law of thermodynamics that a certain amount of lithium is required to hold a charge? The problem with your post that kicked off this tangent is that obviously human ingenuity includes the ability to look at other materials if necessary...

Sure, you can look at completely different battery chemistries, like lead acid or iron. But so far they've all got detriments that make them less desirable than lithium based chemistries.

Heck, you can zoom out and propose storage mechanisms other than electrochemical. That true, but again most of the talk is focused on lithium ion batteries because that's presently the most effective option.

> That true, but again most of the talk is focused on lithium ion batteries because that's presently the most effective option.

"Most effective" varies between use cases. LiIon is fantastic for drones and personal electronics, but the reason it's used for cars is that Elon Musk is a very capable salesman and he bet on it — ten years ago, when Tesla was a cute startup but not a manufacturing behemoth, most of the talk I remember was still about hydrogen fuel cells.

I think used car batteries are likely to be a significant part of the global power grid because there will be enough of them, and it's a good reuse option before they need to be fully recycled into new car batteries.

If they're cheap enough to use directly in this way (and they seem to be already in some specific circumstances), great. If not, doesn't matter too much, there are too many alternatives for me to worry about it.

Recycling batteries means there's lower usable lifetime. The recycled batteries would be cheaper to source, but the more frequent replacement drives up labor costs.

Recycling also doesn't alter the fundamental disparities in scale. Production needs to increase by close to two orders of magnitude to make any significant amount of grid storage. We're struggling to keep pace with just EV demand for batteries, adding just one hour of grid storage for the USA (about 500GWh) would eat up an entire year's worth of batter production.

What alternatives are you talking about? The existing alternatives aren't presently being used because lithium based chemistries beats them in cost. Other storage systems like compressed air are in their infancy and it's unclear if they'll become cost effective enough.

Let me re-use your logic: we don't need solar, or wind, or storage. We'll just use fusion. Fusion is making big strides each year. There's no limit to the number of ways we can squeeze together two hydrogen atoms, so if the tokamak doesn't work out there's too many alternatives for me to think about.

> What alternatives are you talking about? The existing alternatives aren't presently being used because lithium based chemistries beats them in cost.

This just isn't the case. There are many large-scale lead-acid energy storage installations.

Where? There doesn't seem to be any listed here: https://en.m.wikipedia.org/wiki/List_of_energy_storage_power...

> Recycling batteries means there's lower usable lifetime.

Irrelevant, it's a cost reduction.

> The recycled batteries would be cheaper to source, but the more frequent replacement drives up labor costs.

Insignificant, and automatable.

> Recycling also doesn't alter the fundamental disparities in scale.

Unimportant, the scaling up is already occurring. Currently planned infrastructure gets us half way all by itself by 2030. There's no need to hypothesise about beyond that, that's a problem for investors and governments.

> Production needs to increase by close to two orders of magnitude to make any significant amount of grid storage.

Yes, and?

> We're struggling to keep pace with just EV demand for batteries,

Which is why Tesla is investing in mining and alternative chemistries.

> adding just one hour of grid storage for the USA (about 500GWh) would eat up an entire year's worth of batter production.

Unimportant, more factories and more mines are being built as fast as people can find investors. Assuming today's output is a constant going forward is only useful for extremely short time horizons; in this context, extremely short is in the order of 6-12 months, given how fast global capacity is changing.

> What alternatives are you talking about? The existing alternatives aren't presently being used because lithium based chemistries beats them in cost.

False: https://insideevs.com/news/575956/tesla-battery-chemistries-...

> Other storage systems like compressed air are in their infancy and it's unclear if they'll become cost effective enough.

Hydroelectric and hydrogen are both large scale, easy to make.

China recently finished a big hydroelectric plant. We have not run out of places we can fill will water, not even places that fill themselves with water.

Hydrogen electrolysis is so easy that it predates alternators and dynamos, and I did it myself when I was single-digit-years-old. There hasn't been much point to it until recently as it's a store rather than a source, but now we have solar cheaper than natural gas, it's trivial and obvious.

> Let me re-use your logic: we don't need solar, or wind, or storage. We'll just use fusion. Fusion is making big strides each year.

Logic without a connection to reality is not useful.

Fusion, to my personal frustration, is not making "big" strides each year, certainly not if you are defining the current growth rate of batteries (let alone renewables) as anything less than "extremely big".

Logic is also not useful when you ignore some of the consequences of your assumptions, e.g., the best current estimates for the quantity of superconductors needed to make 2 TW of fusion reactors (ITER magnet is ~1000 tons and the reactor is 500 MW-thermal) is also enough for a lossless global power grid.

> There's no limit to the number of ways we can squeeze together two hydrogen atoms, so if the tokamak doesn't work out there's too many alternatives for me to think about.

You are currently using a lithium ion or lithium polymer battery.

Demonstrate a working fusion reactor that makes more electricity than it consumes. You only get to count the sun if you're willing to use PV.

This is the crux of your mistake: you're treating all futures as equals, when the environmental goals need us to act as quickly as possible.

Fission is too slow to build, and we have working designs; fusion is too slow to build even the experimental reactors whose useful descendants are currently scheduled to arrive after it's too late. The new fusion startups? Well, good luck to them, I wish them the best, but any policy decisions that assume they will be ready to start building their own reactor factories before the already-planned-just-not-yet-built battery factories are finished, is wishful thinking.

Or do you think it's unfair to count "merely" planned battery factories and mines, even though there's nothing fundamentally novel about them?

This amounts to a long winded way of hand-waving away the shortcomings of battery storage without actually proposing a plan to resolve them. We do have lithium ion and lithium iron phosphate batteries, but not in sufficient quantities. Your proposal amounts to crossing our fingers and hoping that extraction industries can accommodate a 100x increase in output over the next decade, without significant increase in price. There are indeed alternative battery chemistries but they aren't being used because various shortcomings make lithium based batteries the most effective.

Nobody has ever commercially deployed hydrogen storage, and most hydrogen is produced via steam reformation. The fact that you performed electrolysis as a child does not somehow mitigate the difficulties of using it as a form of grid storage. If you do have a plan to make it viable, go talk to VCs and you'll be a billionaire. But alas, actually delivering grid storage is harder than making promises.

> Nobody has ever commercially deployed hydrogen storage,

Because it wasn't previously useful.

Now it is starting to get useful.

Do you understand the concept of "change"? That the future is not always like the present?

> and most hydrogen is produced via steam reformation.

Why do you think that's relevant?

> If you do have a plan to make it viable, go talk to VCs and you'll be a billionaire.

The billionaires are already doing this. Literally. That specific thing is being done on the orders of the richest person on the planet, as part of a plan to get even richer.

He's hardly the only one.

This is why a lot of new infrastructure in this field is getting funding and being built.

This is why the infrastructure that has been planned and whose construction is in progress is so much larger than what currently exists.

This isn't "crossing my fingers and hoping", it's the default outcome unless something else makes it all redundant.

All of the issues you're using to doubt this are "throw money at them and they go away" type of problems, and enough money being thrown at them to be interesting.

Fusion, sadly, is still in the "throw money at the problems and we expect to find new ones" category.

> I believe this is a result of a sustained campaign of disinformation

Don't waste your energy. The ignorance of the 75th percentile journalist is stunningly broad and deep; it's no wonder that ordinary people haven't a clue.

While you are mostly correct, I think "don't waste your energy" is a memetic hazard. We all do that, we're back to superstition and magical thinking, not civilisation.

> Batteries are reasonably priced per unit, but they do not scale reasonably.

What is your basis for the second part, that batteries can not scale? This seems to be an article of faith among many, but there's no data that can back it up, and it doesn't pass a basic sniff test, using common comparisons.

For example, a car that drives 300,000 miles in its lifetime, at say 30mpg, would consume 10,000 gallons of gasoline, weighing approximately 100,000 pounds. This is an order of magnitude more weight than the battery, and the battery can be recycled into new batteries afterwards, it's not even consumed. Why, with this fantastic reduction in the amount of input materials needed, would batteries of all things run out of materials? State your reasoning, please!

We are currently producing ~300GWh lithium ion batteries per year, and will likely 10x our production every five years, according to the industry reference I heard a year or two ago.

The idea that we won't have grid scale batteries within our lifetimes is contradicted by actual reality, today.

Competing batteries to fission is a sever blind spot that seems endemic. We have a shipping off the shelf tech that is scaling at an amazing rate, yet it's compared to science fiction that has been decades away for half a century.

This is exactly the blindness people have about batteries, and it's why all the traditional car manufacturers have been caught completely flat footed in the transition to EVs

All other areas of energy are going to be caught completely flat footed because Pepe for some reason believe fantasies about batteries rather than the real operating truth right in front of their eyes. What I can't figure out is, why?

> yet it's compared to science fiction that has been decades away for half a century.

That's fusion, not fission.

Though certainly it's possible fission has missed the window and, by the time terawatts of it could be designed, approved and built, solar, wind and others will have taken over.

Fission at scale is science fiction. There are no closed loop fuel cycles and 6TWe of PWR reactors would exhaust world reserves un under 5 years. Reprocessing all of the waste adds maybe a year, and enriching the tailings adds maybe another year.

Sea mining and breeders are significantly less real than 46% efficient solar or 1200Wh/kg AlS batteries.

I think people need to realize that not all batteries are chemical. There are many ways to store/recover electric energy: a motor/generator with weights, a pump/turbine with water, motor/generator with a spinning kinetic mass in a vacuum, etc.

And most of all, thermal storage, which we can achieve today with heat pumps and water heaters by simply insulating houses well, and then incentivizing a market protocol to let people take advantage of better pricing automatically, if they wish to.

But really, chemical batteries will be at the TW scale by 2030 and we will have lots of chemistries for lots of varied applications and demands.

Sure, but none of those are even remotely competitive with lithium batteries - except for hydroelectricity, which requires the right geography for a dam or pumped storage.

And iron flow batteries exist at the same price point (or significantly lower if we pretend your $500/kWh comparison was an accurate representation of medium term price trends) for 6-24 hr storage in spite of very immature production chains.

The fact that lithium is used but iron flow is less popular is a fairly strong indicator that first and second response (which NPP cannot do) is more important than accounting for daily variability right now.

Similarly sodium ion will hit the market next year and hit similar scale to current Li ion is now in about 2025. Being a smidge heavier is hardly a deal breaker for utility uses, and the reduced fire danger should lower integration costs significantly.

> And iron flow batteries exist at the same price point (or significantly lower if we pretend your $500/kWh comparison was an accurate representation of medium term price trends

This was the net cost of storage, including construction, power transformers, etc. Batteries are around $150 of that cost. Note that even zero cost batteries won't eliminate more than about a third of the total cost.

What are the production figures for iron flow batteries? Also, can you point me to a market that sells them? I can see places that sell lithium ion and lithium iron phosphate batteries[1], but no such results for iron redox.

And we'll see how far sodium ion batteries go. Scaling from literally zero to 400 GWh in the span of two years is rather optimistic to put it lightly.

1. https://www.google.com/aclk?sa=l&ai=DChcSEwioxYW9y4j7AhW9MK0...

> This was the net cost of storage, including construction, power transformers, etc. Batteries are around $150 of that cost. Note that even zero cost batteries won't eliminate more than about a third of the total cost.

And if you put a chemistry that isn't a fire hazard directly onto an MPPT that you have already paid for, how much does it cost?

Or what if you're simply building storage for daily variability rather than a first response peaking station picked as a straw man?

> And we'll see how far sodium ion batteries go. Scaling from literally zero to 400 GWh in the span of two years is rather optimistic to put it lightly.

The manufacturing process is designed to fit existing lithium supply chains. And has had billions spent on the parts that are not drop-in. This is like claiming someone expects Olkiluoto to go from zero nuclear power to gigawatts overnight some time early next year.

Also what's your alternative proposal? Lets examine it on the same basis.

First of all, can you or can you not tell me where I can order some iron redox batteries? You write that they exist at the same price point as lithium ion, but I'm not even seeing them for sale at all let alone for the same price.

The manufacturing chain for lithium batteries is indeed designed to fit existing lithium supply chains. Which is why we're only producing 400GWh per year. Because the lithium supply can't accommodate more production.

The alternative is to do what France has already done: serial production of the same designs of nuclear plants. American nuclear construction similarly experienced much lower costs when plants were built at scale [1]. Unlike lithium mining, which as never been done at even 2% the scale required for battery grid storage, countries have indeed succeeded in converting a mostly fossil fuel grid to nuclear power in a short amount of time [2]. Again, it's been done before, with even worse technology than we have now. You're willing to assume that batter production will increase by 50-80x when there is no precedence for that scale. Meanwhile the is precedence for nuclear power being deployed more cheaply when the same designs are built repeatedly. We only need to build 3.5 nuclear plants for each one that exists in the US to get to 100% hydro and nuclear. That's a lot more feasible than increasing battery output by 100x.

1. https://www.sciencedirect.com/science/article/pii/S030142151...

2. https://en.m.wikipedia.org/wiki/Nuclear_power_in_France#Mess...

> The alternative is to do what France has already done

Have a fleet with around 65% capacity factor that has correlated shutdowns right in the middle of an energy crisis? Offshore wind with no storage would be a better choice. At least you'll get _some_ energy during the lulls.

> The manufacturing chain for lithium batteries is indeed designed to fit existing lithium supply chains. Which is why we're only producing 400GWh per year. Because the lithium supply can't accommodate more production.

The supply chain for sodium ion is being built to be compatible existing lithium ion factories. Expansion of those 100s of GWh of production is already being done, and even if it wasn't, building the factory, the supply chain, and then the product takes around the same time as building NPP.

> Unlike lithium mining, which as never been done at even 2% the scale required for battery grid storage, countries have indeed succeeded in converting a mostly fossil fuel grid to nuclear power in a short amount of time

Now do the level of expansion of Uranium mining.

Don't forget you need to overbuild by a factor of 3 from average to provide peaking (otherwise you'll need those exact same 4hr batteries).

You'll need about 400,000 tonnes of enriched fuel or 3 million tonnes of natural Uranium for your first fuel load. This is over 40x the existing annual supply chain. Don't forget to build 10-40x as many centrifuges as exist. You'll probably also need to massively expand sulfuric acid production. Then (assuming you can only burn what you need..which no reactors can do on a scale of minutes) you'll burn the other 75% of known reserves in about two decades. Reprocessing will give you another five years. If you want power anywhere else there's under 6 years total (or rather you just can't because you can't load the reactors even once).

Now do the same for cadmium and silver and indium for control rods. Latest generation copper metallized solar cells use about an eightth of the silver for the same net power as a comparable NPP. Also the Zirconium for fuel rods.

> First of all, can you or can you not tell me where I can order some iron redox batteries? You write that they exist at the same price point as lithium ion, but I'm not even seeing them for sale at all let alone for the same price.

Go talk to ESS, a non-retail technology not being available at retail isn't an indicator of anything. Or Natron for some aqueous sodium ion currently being sold at pilot project prices. Or of you have enough money to jump the queue I'm sure CATL will let you put in an order for a few GWh in 2025. If you ordered a few GWh I'm sure form energy would sign a contract too -- although I'm less certain they can deliver (it seems about as probable as something like Vogtle).

> Have a fleet with around 65% capacity factor that has correlated shutdowns right in the middle of an energy crisis?

Given that most renewables average around 25-40% capacity factor when working well, 65% during a maintenance period is a pretty good thing!

> Now do the level of expansion of Uranium mining

The USA already generates 20% of it's electricity from nuclear power. 10% of global electricity generation is through nuclear power. Recycling alone would reclaim enough fissile material to offset the increase in generation. Where are you getting the figure for 40x increases in uranium production? The reality is 5-10x at most - probably less than that because nuclear electric power isn't the only application of uranium.

> Go talk to ESS, a non-retail technology not being available at retail isn't an indicator of anything.

It means the technology is immature and doesn't have a real cost history. If you can't buy meaningful quantities of it, the price could skyrocket the moment anyone tries to provision a gigawatt hour of storage. These new types of batteries aren't being sold in any significant number, that's the reality.

> Given that most renewables average around 25-40% capacity factor when working well, 65% during a maintenance period is a pretty good thing!

Not if it's correlated, takes months, and is unplanned. And each reactor takes 5-10x as much money and resources as the same gross power. Noone builds a utility solar plant claiming it will produce nameplate wattage at night. Nuclear is always sold as if it has 90-100% availability.

> The USA already generates 20% of it's electricity from nuclear power. 10% of global electricity generation is through nuclear power. Recycling alone would reclaim enough fissile material to offset the increase in generation. Where are you getting the figure for 40x increases in uranium production? The reality is 5-10x at most - probably less than that because nuclear electric power isn't the only application of uranium.

You need to load your reactors. Bringing them online takes around 6 years of fuel -- an AP 1000 takes 100t of enriched Uranium. Your 800GW-1.2TW (minimum required to meet peak electric loads without storage, but does not touch other energy) of reactors in the US will require a 700, tonnes of natural uranium. A single fuel load for enough PWRs to make all electricity 100% nuclear world wide will require all known reserves. This is including reprocessing (MOX only gives you 15% or so more). Then you still have the other 60% of primary energy to cover.

The nuclear industry has never been at the same scale as the current lithium battery and renewable industry. And it cannot be at the same scale because it is limited by critical resource reserves -- not just a temporary limitation on extraction of a critical resource for one possible chemistry. Just matching the scale of the existing industry by installing 50GW/yr (which you've asserted is 2% of what's needed) requires doubling uranium mining.

> US will require a 700,

Should read 700,000 to 1,000,000

> Go talk to ESS, a non-retail technology not being available at retail isn't an indicator of anything.

Well, it is. Anyone can produce lithium batteries. Whereas with the magical batteries it's "ESS Inc is the only manufacturer and holder of patents on its flow batteries".[1]

Which is not how you want to quickly ramp up production and solve storage.

The targeted production is 750MWh per year which is nothing, really.

Their biggest planned project is to have 400MWh in Australia by 2026. I'll let you do the math on how laughable that is by yourself: https://www.energy.gov.au/data/renewables

Iron redox batteries as they currently are don't really exist: the production is low, and locked behind the patent lock of a single company.

[1] https://www.energy-storage.news/ess-inc-ramps-iron-flow-batt...

Then order some sodium batteries from CATL. They're building around a TWh/yr of production which uses the same infrastructure for both sodium and lithium.

Compare to the state of FNR reactors, where there is a single reactor (yet to close the fuel cycle even experimentally) that doesn't catch fire constantly (if you believe the russian government).

> Then order some sodium batteries from CATL. They're building around a TWh/yr of production

I'll believe it when I see it.

> Compare to the state of FNR reactors

Let's see how many people in this thread were talking about FNR reactors. Oh, look: only you. Compare this to discussions where people argue in good faith.

FNRs came up as a direct response to suggesting decarbonising via nuclear. There are no other nuclear technologies that even come close to the criteria of both existing and being scalable (and FNRs are borderline on the first and have huge problems for the second). PWRs exist, but there is barely enough fuel for even the current generation at 10% or so of electricity. Let alone enough to provide primary energy.

So what are you proposing if not one of those two options?

> FNRs came up as a direct response to suggesting decarbonising via nuclear.

You were literally the only one mentioning it in the whole thread. In the response to me talking about storage technologies and batteries. Do not pretend otherwise.

> There are no other nuclear technologies that even come close to the criteria of both existing and being scalable

Thank you for derailing the conversation away from * checks notes * discussion on batteries, but you can do it on your own, I'm not going to engage further.

You replied to a conversation where the direct subject was nuclear vs. renewables & battery storage.

The only viable option for expanding nuclear is much further from reality than existing battery technology that is in the process of commercialisation.

The peak of historic new nuclear production wasn't even at the scale of existing lithium ion production.

As such there is no better option other than the other main renewable storage technologies of electrolysers and PHES.

People also need to realize that none of these solutions have been proven to work at scale.

So, like nuclear? Which also has not been proven to work at the scale needed to decarbonize the world. In particular, current commercial burner reactors are unsuitable for that, as the uranium runs out very quickly at that scale.

Not sure which world you've been living in, but nuclear has been proven to work at scale, while requiring significantly less area than solar or wind, and being able to be load bearing and load following.

If the world's entire current primary energy demand were supplied with burner reactors -- about 6000 3MW(th) reactors -- the world's estimated uranium resource would last about five years.

Your "at scale" is a tiny fraction of the scale that would be needed, and at that larger, true scale, current commercially demonstrated nuclear technology would quickly fail. Breeders would be needed to enable use of much more expensive uranium ores (or thorium), and those are not demonstrated at even your smaller scale.

It's funny how you pulled the conversation entirely away from discussing tech that often is strictly theoretical, and most of which hasn't been proven to work at scale on the level of even one nuclear plant to... «but what about nuclear».

> Your "at scale" is a tiny fraction of the scale that would be needed, and at that larger, true scale,

Which of the tech listed here has this "true scale": https://news.ycombinator.com/item?id=33387011 ?

I'm simply pointing out that if one objects to renewables because scaling has not been demonstrated, and also advocates nuclear, then one is being a hypocrite -- because nuclear also lacks demonstration of the necessary scaling.

I don't think the argument that scaling has not been demonstrated is a strong argument, unless there is a good reason to think the scaling won't work. There isn't a good reason to think that for renewables overall. In particular: limits on area are not sufficient, and limitations on particular storage technologies (say, from materials requirements) also fail, since there are many different storage technologies, some of which require no rare materials at all.

> I'm simply pointing out that if one objects to renewables because scaling has not been demonstrated, and also advocates nuclear, then one is being a hypocrite -- because nuclear also lacks demonstration of the necessary scaling.

See, there's a difference between scaling and scaling.

It is a well-established fact that to provide the same stable production on the same scale as existing nuclear reneables need to be either vastly overprovisioned or have extremely high-capacity storage available. Possibly, the combination of both.

And yet, somehow, when we talk about scaling, it's suddenly "oh no, nuclear cannot scale to support the entire world, it's proof it cannot scale" when renewables can barely sustainably reach the existing scales.

> I don't think the argument that scaling has not been demonstrated is a strong argument, unless there is a good reason to think the scaling won't work.

I will remind you that the original comment I was replying to was this: https://news.ycombinator.com/item?id=33386636

It was talking about storage solutions which are all but required for renewables. And literally none of those solutions have been proven to work at scale. Moreover, some of those solutions are at best theoretical.

But sure, do tell me how it's not a strong argument?

> limitations on particular storage technologies (say, from materials requirements) also fail, since there are many different storage technologies, some of which require no rare materials at all.

Yes. There are "many storage solutions". Go ahead and show me those that work at scale.

And literally no one was talking about the need or lack thereof of rare materials for that storage. We'll cross that bridge when we come to it.

There were roughly 30GW of net capacity in nuclear added in 1985. Compare to around 500GWh/yr of already existing battery production and around 50GW net of new renewable production each year.

The nuclear industry was briefly almost at the current a scale of renewable + storage production. It is now limited by availabity of fuel and cannot grow without new technologies.

> There were roughly 30GW of net capacity in nuclear added in 1985.

Didn't know we are living in 1985.

> Compare to around 500GWh/yr of already existing battery production

1. Are these batteries all used in grid energy storage, or are you taking the full output including things like batteries for remotes?

2. How many of those batteries are actually scalable solutions, deployed, and working. It's kinda funny how my opponents continuously try to steer away the conversation away from this.

3. Compare that to "nuclear power plants generate about a tenth of the world's electricity" which amounts to about 2 653 344 GWh [1]

> The nuclear industry was briefly almost at the current a scale of renewable + storage production.

Storage solutions are not even remotely near nuclear. Without storage renewables have to be significantly overprovisioned to be used reliably.

To repeat again.

> Your "at scale" is a tiny fraction of the scale that would be needed, and at that larger, true scale,

Which of the tech listed here has this "true scale": https://news.ycombinator.com/item?id=33387011 ?

----

Edit: don't bother. This discussion isn't in good faith, and being willingly derailed into discussing everything and anything other then the original statements I replied to.

I'm not going to engage in this conversation further.

----

[1] https://en.wikipedia.org/wiki/Nuclear_power_by_country

> Didn't know we are living in 1985.

Well we can compare to new capacity added in 2021 if you like? There were 5GW added. There were enough batteries made in 2021 for a 50 hour storage for every nuclear plant that came online with plenty to spare.

The point is that if renewables + batteries are not of sufficient scale, then nothing is. If the bar for pursuing a solution to a problem is that the problem already be solved, then noone will ever solve anything.

Lithium battery production is much closer than the scale of nuclear construction ever was. Having a tantrum when this is pointed out doesn't change it.

> It is a well-established fact that to provide the same stable production on the same scale as existing nuclear reneables need to be either vastly overprovisioned or have extremely high-capacity storage available. Possibly, the combination of both.

Do tell. Then how can it be that globally, in 2021, wind + solar delivered more energy to the grid than nuclear did? 10.31% of total world production vs. 9.94% for nuclear.

Anyone with even a passable knowledge of how grids work, and who isn't stuck arguing in bad faith, would recognize that providing 10% of the grid requires no overprovisioning or storage whatsoever.

It seems that this milestone that you claim is so impossible was already achieved.

I love how consistently pro-nuclear talking points are actually arguments for renewables and against fission once you look at the numbers.

The nuclear shills here have given me a lot more hope for a renewable future.

Another fun scaling fact. Loading the 50 or so advanced nuclear reactors required to meet the scale of last year's renewables would require doubling world uranium output.

Er, 6000 3GW(th) reactors.

I’m going to guess lots of factories that we mass produce, aka gigafactories.

> I think the concept of grid scale storage is misguided Fission (and hopefully fusion) can provide base load at all times, with as much wind & solar as needed to provide for variable load.

Where is the uranium coming from? Commercial fast breeders are as mythical as a 10000Wh/kg AlS battery. All nuclear reactors that aren't fiction are just a small multiplier on the U235 supply including DT fusion.

Quite the contrary, people vastly underestimate the cost of battery storage and neglect the fact that it's cost is increasing as input materials become more scarce. Actual real-world cost of deployment is mugh higher than estimates [1] frequently exceeding $500/kWh. Estimates often exclude the cost of installation, maintenance, transformers to handle voltage conversion, etc. This is how $130/KWh battery costs turn into over $500/KWh of real world cost.

Furthermore, the existing global production figures are somewhere around 400GWh per year. The world uses 60 TWh of electricity per day. Even just one day's worth of storage would mean dedicating 100% of battery production to grid storage for over a century. Actually attempting to providing a significant amount of battery storage would spike the cost of batteries as demand exceeds supply.

The input materials for batteries have seen skyrocketing costs in the last couple years [2], and only 24% of battery cost is manufacturing. The rest is input materials. It is very dubious that batter production can keep up with the massive scale required by grid storage.

Batteries can (and are) being deployed. But not at a scale that makes any difference, and it's unlikely they ever will be.

1. https://www.utilitydive.com/news/new-york-battery-storage-co...

2. https://tradingeconomics.com/commodity/lithium

3. https://www.visualcapitalist.com/breaking-down-the-cost-of-a...

> Actual real-world cost of deployment is mugh higher than estimates [1] frequently exceeding $500/kWh.

Can you point why you think this number, for installed battery cost, is somehow a contradiction of the numbers I posted? Additionally, temporary supply shortages due to massive new markers opening up as the cost drops is completely expected. All the temporary higher prices finance massive production capacity expansion that drops prices in the future. This is a well understood dynamic that we see again and again, even in fossil due production!

> Even just one day's worth of storage would mean dedicating 100% of battery production to grid storage for over a century.

There are two extreme errors with writing something like this: 1) assuming that battery production stays constant, when it has been growing 10x/5 years, and we are seeing unprecedented demand for new applications. 2) Using primary energy production for comparison to electrified electricity use, when electrified versions of the fossil fuel versions require 3x-4x less electricity. This sort of basic error is as bad as comparing 1GW of solar to 1GW of nuclear without correcting for capacity factor.

With reasonable battery production scaling, we will be producing 20-30TWh/year of batteries in a decade. For comparison, it takes more than a decade for us to build a nuclear reactor in the West, so if we are in course to have a fully scaled battery production supply chain for global use, whereas we can't even scale nuclear construction in the US to match the rate at which we will be retiring our aging fleet over the next 20 years.

It's time to take a serious look at what carbon-free energy tech is on the table right now and start deploying it as fast as possible, right now, today. At the same time, we should continue research into fanciful ideas like small modular nuclear reactors, in case they pan out. But we need to go to war with the army we have right now.

> It's time to take a serious look at what carbon-free energy tech is on the table right now and start deploying it as fast as possible, right now, today

Correct. And with the battery production capacity we have "on the table right now" would take more than a century to produce 24 hours of storage capacity. Assuming that there's going to be massive multiple order-of-magnitude capacity increase is not taking a serious look at the tech we have today, it's a hugely optimistic take on what we might have decades from now.

> With reasonable battery production scaling, we will be producing 20-30TWh/year of batteries in a decade.

Again, you're literally assuming that battery production figures will increase by 50-80 times within a decade. This is the complete opposite of taking "what we have on the table now", it's assuming that we'll have an industry 50-80 times larger than what we presently have. Moore's law rarely applied to produces outside of semiconductors, because most products hit bottlenecks with inputs.

> But we need to go to war with the army we have right now.

And again, "the army we have now" would take a century to produce only 16 hours worth for storage for the world via batteries. Actually, it'd take even more than that because after a century global electricity use will have increased substantially as the global south develops and wants amenities like air conditioning. I'm really puzzled by why you chose to close with this statement given that your proposal is to assume that our metaphorical army will be 50-80 times larger by the end of the decade.

Reply on news.ycombinator.com