Comment by manfredo | Hacker News Reader

Comment by manfredo | Hacker News Reader

Parent Full thread

You seem to be under the impression that the cost of hydrogen storage comes almost exclusively from the cost of electricity used for electrolysis of water. This is not the case. Even if we assume zero electricity costs, th cost of storing hydrogen amounts to over a thousand dollars per KWh.

Current estimates place this at $1,400 per KWh per year for the average case assuming that the electricity provided is free [1]. That works out to $3.80 per day per kilowatt hour of storage. Even for the optimistic estimated provided by this study, it's still $800/KWh/year - still over $2.00 per day. And again, this is assuming that free excess wind generation is used to get generate hydrogen.

Us retail energy costs are about $.13/KWh. Industrial costs are $.07/KWh. For diurnal storage, hydrogen energy storage represents a 20x cost increase even in the best case estimates for the storage cost and assuming that excess wind generation will pick up th slack.

And for non-diurnal storage the costs are even higher per KWh because a smaller amount of delivered electricity needs to pay off the same amount of capacity cost.

Also, before you point at figure 16, know that this graph includes "capacity credits" - it's not the actual cost.

1. https://www.google.com/url?sa=t&source=web&rct=j&url=https:/...

> You seem to be under the impression that the cost of hydrogen storage comes almost exclusively from the cost of electricity used for electrolysis of water.

Nowhere did I say or imply that. My argument is actually the complete opposite.

What hydrogen does is get a much lower cost per unit of energy storage capacity, at the cost of lower efficiency . The cost per energy capacity of batteries is two orders of magnitude higher than for hydrogen storage. So, hydrogen exploits that for long term storage, it's ok to waste more input energy, if you can get the capital cost down.

> Us retail energy costs are about $.13/KWh. Industrial costs are $.07/KWh.

The correct figure to look at is wholesale price, which is often below $.03/kWh in US power markets, and lower at off peak times when one would make hydrogen.

https://www.eia.gov/todayinenergy/detail.php?id=42456

> For diurnal storage, hydrogen energy storage

I explicitly wasn't talking about diurnal storage. That's not the use case for hydrogen. The use case is long term storage.

> And for non-diurnal storage the costs are even higher per KWh because a smaller amount of delivered electricity needs to pay off the same amount of capacity cost.

Hydrogen has a huge advantage in long term storage, because the per-energy capacity cost is so damned low, compared to batteries. Yes, the cost per kWh stored in long term storage is higher than the cost per kWh in short term storage. But they are different use cases and one cannot be substituted for the other.

If you optimize the problem of providing a steady output from renewables over the days and years, you will find in most places that a mix of short term and long term storage is optimal. Neither substitutes for the other.

> The correct figure to look at is wholesale price, which is often below $.03/kWh in US power markets, and lower at off peak times when one would make hydrogen.

Again, the estimates shared in the study are made with the assumption that electricity used for hydrogen production is free. This is why I am left with the impression that you think the energy used for electrolysis amounts to a large portion of the cost of hydrogen storage. Because you seem to be fixated on the cost of electricity, even though hydrogen storage is still incredibly expensive even if electricity is 100% free.

No, hydrogen does not have low cost of energy storage. Average estimates place the cost at $1,400 per KWh per year for average estimates and $800 per KWh per year for optimistic estimates. And for the third time, just to make sure that this fact gets across, these estimates are made with the assumption that the electricity supplied to generate hydrogen is at zero cost from surplus renewable generation.

This cost needs to be recovered from people drawing the stored energy. If all this energy is being drawn in one day of the year, it's going to cost $1,400 or $800 per kilowatt hour depending of whether you use the average or optimistic estimates. If you draw the energy on 10 days out of the year then it's going to cost $140 or $80 per KWh. How does this make it more suitable for long term storage? It doesn't. Even if the cost were spread out over the entirety of the year it would cost anywhere from 60 to 120 times as much as the wholesale price of $0.03 per KWh. If this energy is only going to be draw on for part of the year this cost increase is going to be even larger because the same cost of storage needs to be recuperated from a shorter period of sales.

It's the most suitable solution for long term storage by virtue of the fact that it's one of the only mechanisms we have for long term storage with our current technology. It's nowhere near cost effective. The costs are the total opposite of "so damn low".

Sigh. You are just not connecting the dots.

Yes, the cost to store hydrogen for LONG TERM STORAGE is higher than the cost to store electricity in batteries for SHORT TERM STORAGE. But the cost to store hydrogen for long term storage is LOWER than the cost to store electricity in batteries for the long term.

If you are comparing batteries short term with hydrogen long term, you are comparing apples and oranges.

Your study does NOT say batteries are cheaper than hydrogen for long term storage. They could not possibly be cheaper for that, since there are too few charge/discharge cycles to amortize the batteries' very high cost per kWh of capacity.

You seem to be saying "just use short term storage!". But that's not a legitimate approach, since there are long term variations in supply and demand. Short term storage only applies to short term variations.

My reply was refuting your claim that hydrogen storage was cheap. Now you're going off on a tangent about short term vs. long term storage and ignoring the central point of my previous comment: the cost estimates for hydrogen storage are nowhere near cheap.

> Yes, the cost to store hydrogen for LONG TERM STORAGE is higher than the cost to store electricity in batteries for SHORT TERM STORAGE. But the cost to store hydrogen for long term storage is LOWER than the cost to store electricity in batteries for the long term.

> If you are comparing batteries short term with hydrogen long term, you are comparing apples and oranges.

I'm not comparing anything. I'm providing a source on the cost of hydrogen storage, to refute your unsubstantiated claim that hydrogen storage is cheap.

And as it turns out this cost is extremely expensive, regardless of whether it's used for short term or long term storage.

> Your study does NOT say batteries are cheaper than hydrogen for long term storage. They could not possibly be cheaper for that, since there are too few charge/discharge cycles to amortize the batteries' very high cost per kWh of capacity.

Correct, this study is about the cost of hydrogen not a comparison versus batteries. And for the second time, nowhere did I compare hydrogen against batteries. My reply doesn't even mention batteries once.

All I used this study for was tho find an estimate for the cost of hydrogen storage. And that cost is $1,400 per kilowatt hour per year for normal estimates and $800 per kilowatt hour per year for optimistic estimates. This is incredibly expensive, contrary to your claim that the cost of hydrogen storage is "so damn low".

> You seem to be saying "just use short term storage!". But that's not a legitimate approach, since there are long term variations in supply and demand. Short term storage only applies to short term variations.

The point is that it's extremely expensive, and not a viable solution for any type of storage. Short term or long term. The cost per kilowatt hour of storage is just way too high to be viable. $800 per kilowatt hour per year for the optimistic estimates, and $1,400 for realistic estimates is insanely expensive. And again this is assuming that the electricity provided to convert water to hydrogen is free.

In what world is $800 to $1,400 per kilowatt hour per year, "so damn cheap"? For hydrogen to achieve even just 1 day's worth of energy storage for the USA (11.5 TWh daily usage) at a cost of $800 per kilowatt hour per year would cost $9 trillion dollars per year. And remember, that's the optimistic estimate and assuming that the energy used to produce hydrogen is free.

This isn't about short term vs long term storage. Hydrogen storage is incredibly expensive, it's not "so damn cheap".

You misrepresent what I wrote. I didn't say hydrogen was cheap. I said the PER ENERGY CAPACITY part of hydrogen storage was cheap. And it is! It's like 200x times cheaper than the same aspect of battery storage. So if we get into a situation where the per-energy capacity costs are really import (such as, long term storage), hydrogen beats batteries.

There are, of course, OTHER parts of a hydrogen storage system, whose costs scale with the input and output power, not the stored energy. And these other parts (and the lower efficiency) make hydrogen uncompetitive for diurnal load leveling.

But hydrogen is superior to batteries if the goal is to store the energy for considerably longer periods. It is TOTALLY about short vs. long term.

As for whether hydrogen is "incredibly expensive": for the things it is suitable for, it is cheaper than the alternatives. Try to optimize a CO2-free power system using just wind, solar, and batteries in Europe, vs. one that also includes hydrogen storage. The latter is cheaper! It's also cheaper than a system that includes new nuclear power plants

> In what world is $800 to $1,400 per kilowatt hour per year, "so damn cheap"?

Your units there don't even make sense. BTW, I hope you aren't taking ratio of energy capacity to power-related costs suitable for diurnal storage and applying that to a seasonal storage system. The latter charges up and discharges over months, so the ratio is very different.

> I said the PER ENERGY CAPACITY part of hydrogen storage was cheap. And it is! It's like 200x times cheaper than the same aspect of battery storage. So if we get into a situation where the per-energy capacity costs are really import (such as, long term storage), hydrogen beats batteries.

Right, and this claim is totally false. It's not cheaper than batteries, let alone 200x cheaper.

Where are you getting this figure that hydrogen storage is 200 times cheaper than battery storage?

> But hydrogen is superior to batteries if the goal is to store the energy for considerably longer periods. It is TOTALLY about short vs. long term.

No, it's not. Again, are you just completely ignoring the costs presented in the study?

> As for whether hydrogen is "incredibly expensive": for the things it is suitable for, it is cheaper than the alternatives. Try to optimize a CO2-free power system using just wind, solar, and batteries in Europe, vs. one that also includes hydrogen storage. The latter is cheaper! It's also cheaper than a system that includes new nuclear power plants

No, it's not. Again, even just providing 1 day's worth of nuclear power costs over 9 trillion dollars every year using the optimistic estimates. Using the mid range estimates this figure is $16T per year.

> Your units there don't even make sense.

Cost of storage is measured in both capacity in duration. The unit answers the question, "how much does it cost to provide X amount of storage over Y duration"? Hence, kilowatt hour per year. Or $ KWh/yr as shown in the study. I'm not sure what is hard to understand about cost per kilowatt hour per year. $800 KWh/yr means it costs $800 dollars to provide 1 kilowatt hour of storage capacity for one year.

> Right, and this claim is totally false. It's not cheaper than batteries, let alone 200x cheaper.

Let's look at a source of numbers, shall we?

Go to https://model.energy/ and click on "Show advanced assumption settings".

Now scroll down a bit:

Battery energy capital cost: 142 Euro/kWh

Hydrogen energy capital cost: 0.7 Euro/kWh

What's the ratio of those? About 200!

It should be really clear why these numbers are what they are, if you understand what they mean. The hydrogen energy capital cost is the cost of creating the underground storage cavern where the compressed hydrogen will be stored. It does not include electrolyzers, compressors, turbines, or generators -- those are all POWER related capital costs.

The NREL hydrogen cavern cost numbers are even lower.

What this means is that if you want to expand the energy storage capacity of your storage system, but keep the input/output power the same, hydrogen scales FAR better than batteries do (with the partial exception of flow batteries, but the energy capacity related costs of those will still be higher than that of hydrogen storage, due to the need for tanks and expensive metals like vanadium for the anolyte/catholyte). And THIS is why hydrogen is far superior to batteries for long term storage.

> Go to https://model.energy/ and click on "Show advanced assumption settings".

> Now scroll down a bit:

> Battery energy capital cost: 142 Euro/kWh Hydrogen energy capital cost: 0.7 Euro/kWh

> What's the ratio of those? About 200!

You realize that this site is a calculator, right? It's a calculator used to calculate how much energy would cost for the cost values supplied. You can put whatever you want for that field. This site isn't a study. The value for this figure comes from an external pay walled link. And it looks like that link isn't even talking about storing hydrogen, but storing methane produced from hydrogen gas (power to gas). Methane is much easier to store and is more energy dense by unit of volume, as well as makes use of existing natural gas transportation infrastructure. But power to gas needs an external source of carbon dioxide.

And this isn't the actual cost of hydrogen storage, only the capital cost of constructing a storage container. Which is kind of a ridiculous figure to provide for storing gas in a cavern: the storage container is already built for you. But this goes back to the problem of geographic limitation: you can only store hydrogen in a limited set of places. This is like saying we should just power all our energy needs with hydroelectricity and geothermal power.

It's a site that gives references to where it got the data. And it's more than a calculator -- it's an optimizer, that gives the most economical mix of generation and storage to supply a constant power output, using real weather data. Those optimization results show that hydrogen beats batteries for long term storage.

> And this isn't the actual cost of hydrogen storage, only the capital cost of constructing a storage container.

Of course. If you had been reading what I wrote, you would know I know that. The total cost in the model is obtained by adding the energy related cost and the power related cost (as well as the cost of the input energy). And no, the cavern is not already built for you, it's produced by solution mining of an existing salt formation.

Follow the NREL link if you want data that's not behind a paywall. Their numbers are even lower.

But anyway, the basic point I've been making is extremely elementary, and it's remarkable and more than a little appalling that you are unable to understand it. You need to step back and stop embarrassing yourself.

> And no, the cavern is not already built for you, it's produced by solution mining of an existing salt formation.

This plan is to repurpose existing mines. Yes the cavern is already built for you as far as the cost estimates are concerned.

> But anyway, the basic point I've been making is extremely elementary, and it's remarkable and more than a little appalling that you are unable to understand it. You need to step back and stop embarrassing yourself.

And by resorting to insults you've demonstrated that you are not interested in participating in good faith. This is the exact opposite of what you want to do if you want to convince someone that your claims are correct.

I think that you are misreading units somewhere. I don't see anywhere in that report that looks like $1400/kWh for hydrogen systems. Are you referring to Figure 14? The "Spilled Wind - Current Technology" bar there looks like it could be about $1400. But that graph is dollars per kilowatt, not per kilowatt hour. It refers to a system with 6 hours of storage.

Reply on news.ycombinator.com