Hacker News Reader

Comment by mabbo | Hacker News Reader

Hydrogen production is a fascinating energy case study because of the many possibilities for production.

Consider solar and wind. With variable production, there will always be time where you have more energy produced than you need. Why not use the excess to create hydrogen and store it? Now the energy doesn't go to waste.

Consider nuclear power. Nuclear doesn't scale down well- it's most efficient at a specific rate of production. But demand is quite variable- 2am on a not-so-warm-or-cold day, who needs all that extra energy? Why not use the excess to create hydrogen and store it? Now the energy doesn't go to waste.

In short, the energy used to create hydrogen can be very low, provided you have a cheap means to store and move it to where it needs to be, and provided you can deal with variable production rates.

If this provides an income for the energy producer, then it becomes more financially viable to use wind, solar, and nuclear- all non-CO2 emitters.

View on HN

> Why not use the excess to create hydrogen and store it?

This is a very popular idea, but it doesn't work out as well as it sounds. The reason is that electrolyseurs aren't free, the capital costs are high. Not saying this is impossible, but it's challenging to make this economically feasible.

One more thought is that you can basically adapt all industry production to excess energy. I think the reason people want to do this with hydrogen electrolysis is because a) it's new and b) they associate both with energy transition. But neither means it's the best option.

This is a good point. Renewable energy is getting so cheap that throwing kWh away can be cheaper than the cost of idling other capital infrastructure.

We will need to rethink a lot of our economic assumptions in energy very soon. Solar farms are already designed to produce 30% more DC energy than their inverters can handle, because it's the lowest cost balance between PV panel capacity and inverter capacity. If somebody has an application for DC energy that can be co-located at solar farms behind inverters, and can be economical running only a few hours a day, there's huge amounts of nearly free energy already available.

>> Renewable energy is getting so cheap that throwing kWh away can be cheaper than the cost of idling other capital infrastructure.

That's missing the point.The problem isn't "what to do with excess of energy at peak times?", it's "where do we get energy from on quiet, wind-less nights?" Solar and wind are highly unreliable, so storing energy for is the necessity. Or we can use nuclear.

Who was making this point that I supposedly missed? Where we are in the conversation, your problem has already been addressed, and part of the solution has been to overprovision renewable generation such that we have excess electricity at many times, amounting to 50% to 200% of current electricity usage.

The problem you discuss has been addressed with tons of predictions and models. It's solvable with excess generation capacity, storage, and transmission capacity. Which path we choose will depend on the relative dollar costs and political costs of each. (For example HVDC transmission would likely save us lots of money, but is so difficult politically that it is unlikely to happen.)

But in any case, nearly all models show that under nearly all cost scenarios, we end up with lots of excess generation capacity, with curtailed renewables generation for a large chunk of time.

This is the future we are talking about, a future where we are meeting our energy demands with all renewables, and we have lots of excess generation capacity with zero marginal cost.

>> It's solvable with excess generation capacity, storage, and transmission capacity

So now you're arguing FOR energy storage? Your previous post was basically: "let's overprovision and let excessive energy go to waste, instead of building costly storage infrastructure.".

We will have a mixture of storage, over-provisioned generation, and transmission, the only question is the relative proportion. And that relative proportion will be determined by how quickly costs fall relative to each other. If a kWh storage costs 10x the amortized capital costs of a kWh solar generation, we will have a lot less storage than if storage only costs 2x the amortized capital costs of a solar kWh.

If hydrogen electrolysis can become cheap in 2030, maybe we will even use hydrogen for some storage, but we will have to balance the capital costs of hydrogen generation and conversion back to electricity with everything else, which will determine how much of it gets deployed.

It's not only about cost, it's also about practicality: there is some baseline demand on electricity even at night. Solar obviously cannot fulfill that demand. Wind also can't because it is too unreliable. So we do need some storage technology, and hydrogen is one of the options. It seems way cheaper than using Lithium-Ion batteries, but might be still more expensive than, say, pumping water uphill and using that to power a hydroelectric plant.

Energy transmission can be some solution here, but we might be forced to transmit energy across the continents (so, for example solar plant in Sahara delivers power to New York City at night) which probably is too costly.

I actually think "curtailment" is not throwing away electricity any more than not running a natural gas power plant 24/7 is.

The only difference is the natural gas plant has fuel costs and the solar doesn't. But both are big capital investments, so you could argue you're "throwing away" electricity you could've generated with that natural gas plant by not running it all the time.

Anyway, just my little contribution to how we think about curtailment. If you have a good grid connection and optimal storage, then you don't end up needing that much curtailment (say 30%), although some is basically always optimal. See: https://model.energy

Oh, also, under-sizing the inverter relative to the solar array DC capacity is not even close to a one-to-one curtailment example. The solar rarely makes its peak power rating, so undersizing the inverter by 25% means almost no curtailment in the real world. Effectively, you're increasing the AC capacity factor and it almost perfectly compensates (until you get very high ratios, but even then the amount actually curtailed is small). This has a bigger impact in places that have more cloudy or overcast weather. If you install more solar panels (which are a much smaller part of the total cost of the project than they used to be), you can compensate for not having as much sun without much increase in cost.

Bitcoin mining?

Might as well, good way to burn excess energy imo. CPU mining for other cryptos would work as well (XMR comes to mind)

crypto mining, or stacking bricks with a crane.

"Surplus" capture/use is a bigger problem than most people realize, not just for solar, but throughout the entire energy economy.

There's a reason you still see so many flare-offs in the petroleum industries, even though it's always been "burning dollars" and is now usually discouraged via imposed regulatory costs. Yet it still makes economic sense in many situations.

Why would people invest capital in capture technology that only gets used a fraction of the time? Whether it be at a solar plant, or a refinary, people will choose to invest that capital where it's use can be maximized. There are certainly situations where it makes sense economically, but many where it doesn't, and more situations where the capital can be better spent elsewhere.

Similarly, if a hypothetical town gets 15kw solar power during the day, and 15kw of hydro or wind during night, and needs 20kw backup/surge capacity from gas/nuclear/other, you've designed a system that requires a surplus of capacity, not just in generation, but also in transmission and distribution. In that context, surplus==inefficiencies.

This can be somewhat minimized by so-called smart-grid technology, but people tend to focus on technical capabilities and not what makes economical sense, which is much more limiting than people realize.

It's part of the puzzle.

It's true that high efficiency catalytic electrolysis requires expensive materials. But at significantly lower efficiencies you can just do the same thing we all did with a battery in grade school. That's not quite free to build out, but it's close. And relying on low-efficiency energy storage to fill the gaps in production is still better than firing up a gas plant to supply the power.

It seems not unlikely to me that a lot of wind and solar plants might see building out a local hydrogen electrolyzer and generator (or just sell the hydrogen) as a useful choice.

But yes: there's no silver bullet, and there are a lot of very reasonable competitors in the energy storage market.

> It's true that high efficiency catalytic electrolysis requires expensive materials.

And "high" here is only 70%.

And this way, you only get around 15 grams per kwh.

how many mols is that? Gas is not very dense.

I'd expect that as in most other cases, it would come down to payback time, wouldn't it? With the floor being "pays for itself before it reaches end of life".

If it doesn't pay for its own existence, it's likely a waste of effort. But if it does, you're adding capabilities & flexibility for zero cost. I don't really see a downside, other than tying up the capital.

What I am saying that hydrogen from natural gas has an incomparable economic advantage over electrolyzed hydrogen, and denying it is as silly as denying thermodynamics.

Electricity has to get tens of times cheaper for it to beat steam reforming, and you absolutely can't produce hydrogen cheaper than the energy you put into its production, most of which comes from, surprise, hydrocarbons.

And obviously, you can't produce more hydrogen than the energy equivalent in hydrogen you consume. This completely crushes whatever ignoramuses thinking that you can "use hydrogen to produce more hydrogen"

Hi slfnflctd! You replied to my comment on a thread on LSD trips some 42 odd days ago, and I kind of missed it when you posted it but I just stumbled upon it today and found your comment to be really insightful and I could relate to it really well. I think I could learn more from you if you were open to a discussion. I was wondering if we could connect through email or any other platform of your choice?

> The reason is that electrolyseurs aren't free, the capital costs are high.

While this is definitely the reason, there is also a bigger one - that the process of turning electricity into Hydrogen is fundamentally wasteful(Energy losses), particularly when the route is electricity->hydrogen->back to electricity.

Where it actually is the best option: and where most of the “Green Hydrogen” gas will likely end up being broadly used to me is looking like industrial applications like this article mentions (Steel production), along with ammonia for fertilizer production.

Capital costs as you mention (combined with utilization rates) - electrolyzers to be most economical need to run 100% of the time, meaning you need “excess energy” 100% of the time. With the implication that the energy will be from variable, renewable sources - there is a fundamental disconnect between these. A main reason why batteries and other techs are likely to win out in many “utilizing excess energy” over the long run.

No one cares about efficiency intrinsically. As long as it's cost effective, this is not an issue.

One of the major improvements in modern electrolyzers is that they don't need to run 100% of the time to function. As costs come down, it's likely we can tap excess energy whenever it is available.

No one care about efficiency, sure. But efficiency in this case is the “tax” that must be paid in all cases.

And you are missing my point: That as far as the options go for “tapping excess energy” hydrogen is overblown and overstated as a solution. Things like demand response and heating applications don’t have anywhere near the same kind of losses. Price and value is what actually matters.

Then it's merely reducing the "tax" until it is no longer an issue. A combination of efficiency improvements and lower source energy cost can solve this.

The alternatives have their own downsides. Usually in the form of weight or raw material usage. Since the raw material of hydrogen is water, this could easily be much cheaper than the competition.

Electrolyzer costs are dropping in a similar to way to solar power. This is likely not going to be a limiting factor pretty soon.

> Nuclear doesn't scale down well- it's most efficient at a specific rate of production. But demand is quite variable- 2am on a not-so-warm-or-cold day, who needs all that extra energy? Why not use the excess to create hydrogen and store it? Now the energy doesn't go to waste.

The problem is not of the energy going to waste or the plant being unable to throttle down. The problem is that fuel cost is immaterial; a plant running at 1% output power still costs 99% as much per day as a plant running at 100%.

That means that building 10% oversupply is tremendously expensive. If you're using it 10% of the time, it costs 10x more per MWh than the baseload power. If hydrogen is 11% as profitable as baseload, you are only gathering another 10% revenue, and the power is still 5x as expensive as normal electricity. That would be far more expensive than any other source.

You could easily estimate hydrogen economy would be worth >>10% of electricity... but only at the margin. Right now it's reasonable to say hydrogen is worth 50% of the value of electricity used to make it. For all regional grids in the US, daily peak is 30-50% higher than the daily minimum[1]. That's a huge amount of overcapacity to consume, far more than can be utilized by base industry.

Having >70% nuclear and/or relying on hydrogen economy to make nuclear economical is pretty highly dependent on hydrogen vehicles. Fermi math: 142 billion gallons annual[2] @25 mpg fleet efficiency[3] translates to .9 trillion kWh @4 mi/kWh, or ~25% of annual electricity. The bottom line is that it takes a relatively minor (~15-20%) overcapacity would tap out basically all possible uses for hydrogen.

[1]: https://www.eia.gov/todayinenergy/detail.php?id=42915

[2]: https://www.eia.gov/energyexplained/gasoline/use-of-gasoline...

[3]: https://www.reuters.com/article/us-usa-autos-emissions/u-s-v...

Nuclear does scale up and down very fast. Reference : https://hal-edf.archives-ouvertes.fr/hal-01977209/document

However because upfront construction cost is high, you want to use nuclear reactor as much as possible.

Also storing hydrogen is not as easy and cheap as other gasses due to hydrogen embrittlement, high pressures and very low boiling point (-250C). But as it often is with fuel it's expansive now but soon it might be proffitable (as it was with tar sands).

Reply on news.ycombinator.com