Hacker News Reader

Comment by rapsey | Hacker News Reader

Efficiency is not the only factor. There is also cost and availability. You can't build pumped hydro just anywhere. I have no idea what the cost is compared to lithium batteries though.

View on HN

> You can't build pumped hydro just anywhere.

No, but electrical grids can efficiently transport energy across continent-scale distances, so it doesn't really matter. It's true, that stored hydrogen only needs to beat "electricity from the nearest dam" and not local generator numbers, but that's still a tall order.

Sure but how many of these locations are there? You don't just need a hill, you need large amounts of water available. Not to mention the construction is definitely not cheap or fast.

I mean there is a shit ton of land available in say Spain for solar+hydro plants. No matter the lower efficiency, you could build enough to power Europe.

What If We Covered the Sahara With Solar Panels? - https://youtu.be/62ASvupr8Zg

The Problem with Solar Energy in Africa - https://youtu.be/7OpM_zKGE4o

One of the difficult problems with extremely large solar projects is the infrastructure to move the power. In a lot of these cases, the space to build solar is large areas of land that no one uses... which means that there's not much infrastructure to build the solar power plants, and then not much infrastructure to move that power back to places where people live.

Yes, there's a lot of land that could be used for solar. Moving the enough power to power all of Spain (or all of Europe) based on solar requires some very impressive transmission lines.

Well there's also this thing called "roofs" and there is collectively a lot of them.

While the economic efficiency of massively distributing your solar collection across residential and commercial buildings is less than a grid scale, at the same time you get far more resilience with all the buildings having local energy generating capacity in disaster situations. It also alleviates the total amount of energy the grid needs to transport (especially for home/business charging of EV vehicles), so grid development/maintenance costs will stay sane.

Not that grid scale isn't important. Heat pumps, home geothermal, residential solar, grid solar, wind, battery storage, pumped hydro storage, and whatever else works will be necessary. Hopefully synthetic fuels / algae fuels / aluminum air batteries / next gen nuclear / grid geothermal can also all contribute.

> No, but electrical grids can efficiently transport energy across continent-scale distances

That doesn't matter. E.g. in Europe most natural location for pump storage are already used as such. I guess you could build some in Asia, I am not sure. We would just have to trust russia to transport our electricty back and forth. Good idea /sarcasm

Also the cost of lithium may rise as more vehicles start using Li-ion batteries, but price of hydrogen storage will likely remain the same

That's contrary to conventional economics -- costs fall as production increases.

I think the reason hydrogen storage costs won't fall much is because the cheapest technology (metal tanks) have already benefited from economies of scale. The parts that make them suitable for hydrogen storage specifically will get cheaper, but it's unlikely that there's a lot of low hanging fruit for manufacturing the tanks themselves. There could be a breakthrough in metal hydride storage or cryogenic storage that could reduce costs, but I'm not too optimistic. I think the most likely scenario is that most electrolyzed hydrogen is converted to methane for storage and use. Methane is much easier to convert to liquid and much more energy dense, which helps with storage costs.

Lithium prices should fall dramatically in the future as production increases.

And the rest of metals? Nickel, copper?

Storage batteries are moving rapidly toward a LiFePO4 chemistry that doesn't require rare resources. You can buy cars (even Teslas) and home batteries with them already.

Only a matter of time before phosphorus becomes as expensive as nickel or copper and you are back on the beginning with lack of materials for batteries.

> Earth's commercial and affordable phosphorus reserves are expected to be depleted in 50–100 years and peak phosphorus to be reached in approximately 2030.

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

No, phosphate is vastly cheaper than battery metals. I mean, yes, we're using way too much of it (for agriculture, in a manner that ends up unrecoverably flushed into the oceans). And we're going to hit a wall, and its price is going to skyrocket. But to matter to a battery producer it would have to be so expensive that we'd have all starved anyway. The price levels between batteries and fertilizers are just too different.

> No, phosphate is vastly cheaper than battery metals

Lithium used to be cheap. Is not anymore. Phosphate will go same route. That's the whole point.

I repeat: if phosphate was so expensive as to make a significant portion of the material cost of a LiFePO4 battery, then we would have long since starved for lack of crops. Relative to food, batteries are an extraordinarily expensive luxury good.

How is this contrary to conventional economics? Using the most basic model of supply and demand, EVs becoming more popular would shift the demand curve to the right, since more people would want to buy Li-ion batteries at any given price. We would expect the market equilibrium to shift both upward and rightward, increasing both quantity and price. It might only shift only one way or the other, if supply is particularly elastic or inelastic. But it definitely wouldn't cause a decrease in price by itself.

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

The effects of experience curves vary widely between different industries. For semiconductors, it's huge. For mining, not so much. Battery costs are only 25% manufacturing. So at this point, battery supply is rapidly becoming a resource extraction problem.

The scale effects in mining are very significant. Big mines are a lot cheaper per tonne than little mines.

This is no longer a question of economic axioms, but of the facts on the ground. How do we know that the scale effects relative to lithium mines as they stand today will be sufficient to offset the price increase inherently caused by the increased demand? Do you have a source for this?

The increased demand only causes a temporary increase in price. Lithium is a commodity, and for commodities price == marginal cost.

The marginal cost for lithium is about 1/10th of the current price, so the price of lithium will eventually fall.

Of course "eventually" could be a long time away...

Then why is cost up 10x over the last two years?

Price is up 10x. Cost hasn't changed.

Cost is literally another word for price - I'm not sure what you're trying to convey by trying to distinguish the two. When a battery manufacturer goes to buy a ton to lithium, the cost they have to pay has indeed increased 10x.

The price of extracting a ton of lithium from the ground may be the same, but it's not enough to keep up with demand. Which is why the cost of lithium on the market is skyrocketing.

Price is up because it takes 18-24 months to produce lithium on the margin. The primary source of lithium is solar concentrated brine, so today’s supply is peak Covid planning decisions.

The people that had the guts to invest in lithium production in the middle of Covid when the price was in the tank, are currently harvesting that lithium and getting a massive payday.

Price will come back down.

This is predicated on the assumption that Lithium is a scarce resource with new new sources available. In fact the opposite is true: Lithium is pervasively available almost everywhere[1], the problem is that it's expensive to extract and requires a bunch of processing facilities be built.

So you'd absolutely expect Li supply to grow along with demand and push prices down due to economies of scale. And that's exactly what we're seeing.

[1] Basically, go find a salt deposit -- that's a dried up ocean, which is the best concentrator we can find for Lithium compounds.

I think causality works better the other way - production increases as costs fall.

Edit: LegionMammal has the right words, demand curve.

That is not necessarily true for minerals, which are mined not produced.

https://tradingeconomics.com/commodity/lithium

Reply on news.ycombinator.com