One way to deal with volatility is dynamic minute-by-minute adjustment of prices, and appliances (like car chargers, water heaters, HVAC systems, etc.) that turn themselves on and off based on the price.

storage in hydrogen and allow a 80% devaluation of the energy that was used. That doesn't matter because the energy would be thown away otherwise like it is today.

Electricity->Hydrogen->Electricity loses about 60%, not 80%.

Electricity -> gasoline or kerosene has lower losses, easier transport, and safer storage.

Are you talking about renewable assisted oil extraction? There are some oil extraction methods that have a EROI below 1 which can ironically only be powered by renewables.

Easier transport than existing gas pipelines and existing/to-be-built gas caverns?

You need to upgrade those to be able to use hydrogen, but you don't lose a lot more efficiency if you upgrade the hydrogen to methane afaik.

Depends on how decentralized you can inject hydrogen from overflowing renewables, then. If you could do so at the city's connections to the main long-distance network, it might not be economical to have the infrastructure to create methane from it, and, on top, the hydrogen fraction would be CO2-free even if burnt in a normal boiler/stove. I wonder what upgrades would be needed, though? Are they not already using one of the steel alloys that is resistant to hydrogen embrittlement?

I doubt the efficiency, do you have sources? What are the chemicals you start with?

even better!

Yeah, hydrogen's the only storage medium that makes sense at that scale.

I'd expect some - actually probably quite a lot - of excess CAPEX on wind+solar, and some deficit (less CAPEX, but buying some electricity at retail) so you don't have electrolysers and thermal plants just lying around.

You're still stuck with the investment issue: either you pay (Hinckley-scale!) fees to providers to just lie around, or the public sector runs it itself, and the "private more efficient" bollocks is exactly that, because the private sector isn't providing the value. The people providing SLAs are.

The point of hydrogen is not mainly energy storage but hydrogen is the only viable way to clean up two of the most carbon intensive industrial processes: steel smelting and ammonia (fertilizer) synthesis. Smelting steel requires coal to reduce the oxygen from raw iron to carbon dioxide. The coal can be replaced by hydrogen, which will then react to water vapor. Synthesizing ammonia needs elemental hydrogen which is as of now extracted from methane (natural gas) using steam reformation. This leads to CO2 as waste product whereas electrolysis crates hydrogen and oxygen directly/cleanly from water.

You need huge amounts of "base load" hydrogen generation capacities for these industries alone. Flexible generation, i.e. produce more to soak up excess renewable energy which you can then export or turn back into electricity (at a huge loss) later, is just the cherry on top.

The initial funding will absolutely come from the steel and chemical industry (and also government subsidies to get the transformation started).

You can purify iron by electrolysis too. You don't need hydrogen for that.

But you still must get a way to put carbon in it. This is usually done by mixing the steel with CO, and the easiest way to get from renewables to CO is by reacting CO2 and H2.

The Salzgitter AG certainly thinks hydrogen is the way to go.

Well, hydrogen is a way to go.

I don't think anybody has enough information to know what is the way to go. But well, I'm an outsider with very little information.

Hydrogen is not at all required for steel melting. Look at Boston Metal. In fact I would argue its the worst solution of the two by a big margin.

For ammonia this will eventually be required.

Smelting isn't just melting, it chemically reduces Iron oxide and needs a reducing agent (hydrogen, CO etc.)

Aluminum is reduced through electrolysis. I think he means that method can be used for iron which means you don't have to use electrolysis on hydrogen.

As always the much better solution would be to build dedicated nuclear plants that operate at really high thermal energy and use that to make they hydrogen.

Nuclear neat is totally green, you actually get 100% utility out of the plant rather then having constant highly inefficient 'start-stop' operations.

And its actually a better less complex process then electrolysis.

But I guess its much better to absurdly gigantically over-provision renewables so one in a while we can run electrolysers and then by inefficent fossil fuels to drive inefficent electrolysers the rest of the time.

And its not at all clear that hydrogen as storage is actually a good idea, and its also not at all clear its the only thing that scales. It very much unproven at the kind of production scale AND storage scale you suggest. And even in the best case its incredibly inefficient and incredibly expensive make/store/produce.

Something like Iron based batteries will likely turn out to be a better solution.

Japan is going down this path.

> hydrogen's the only storage medium that makes sense at that scale.

There's more hydrogen in a liter of gasoline than in a liter of liquid hydrogen.

Electricity to gasoline produces a safe product that is easily storable for years and easily transportable requiring only minor safety precautions. Not something that can be said for hydrogen.

I mean that's a technology that could be used, but I don't see mention of it in the article. Is it in the law?

- mix of supply sources

- over provisioning

- geographic distribution

- biomethane

- international grid ties

A recent study on Germany based upon 35 years of weather data [0] shows that we would need about 24 days worth of storage to cope with this weather conditions.

[0] https://twitter.com/QvistStaffan/status/1427625795355349004?...

It's always interesting to see what ridiculous limitation they place on these to make them not work out.

Popular ones are:

Refusing to allow energy to be traded with neighbours.

Refusing to allow demand response.

Refusing to allow over building (so the installed capacity exactly matches the yearly required energy and storage needs multiply).

Refusing to target efficiency measures.

Refusing to use existing low carbon power sources (hydro, biogas).

Often they use low volume prices for a massive rollout but at least in this case they're estimating storage capacity rather than price so they won't have done that one.

edit: reading through it now, they cite other people who didn't do the above, and then state that they know some of their choices will overstate things:

> the fact that we model Germany as an island may lead to an overestimation

Guess that bit didn't fit in the tweet.

I don't fully understand this next one, but their headline number is based on running biogas as a constant baseload, which seems utterly ridiculous. They also claim that allowing that to flex with supply and demand has this impact:

> "Adding other sources of flexibility for the example of bioenergy, the duration of period that defines storage requirements lengthens to more than one year."

That's a very strangely worded sentence, with a very counterintuitive plain reading. Are they trying to intentionally confuse people? I'm not sure, can't figure out anything reasonable from the article. How can adding 8GW of flexible biogas generation increase storage requirements? Are they counting the storage of biogas?

I am not trying to defend the study, but I don't think any of your points actually mitigate the issue of volatility.

1. trades with neighbours: most european countries face similar weather conditions at the same time. The french will need their own nuclear capacity to make up for their own lack of wind.

2. demand response: switching off factories or heating when you need it? That's no solution

3. over building: the problem is that the volatility of wind is massive, if you look at the uk grid website [1], it can go to nearly zero for more than a week. If the volatility was smaller, over provisioning could be a solution (provided the economics work).

4. efficiency measures: you only make the size of the problem slightly smaller, but you still have a volatility problem. And with cars going electric and us not relying on russian gas for heating, I don't see the demand for electricity going down

5. hydro, biogas: there is only so much hydro you can build. And the places where you can build some (Sweden, Norway) leave you at the mercy of a russian submarine cutting the cable. Biogas: isn't that co2 emitting?

[1] https://gridwatch.co.uk/

Biogas is carbon negative, since you make it with manure and food waste that would otherwise give off methane.

EVs and heat pumps are over 4x the efficiency. 75% of your energy just not needed anymore to do the exact same work.

The real clincher is that most of the storage they predict is hydrogen and at the end they calculate that the crazy amount of storage required for Germany as an island is in fact, the same size as the existing gas storage facilities which can be reused for that purpose.

Which I think really rams home how utterly boring this allegedly insurmountable challenge is.

I am so tired of pretending things that aren't impossible aren't. Just do it and then reach the next level of civilization. Honestly the only reasoning I can come up with is that a handful of individuals would rather be kings than let everyone live a more advanced life.

>2. demand response: switching off factories or heating when you need it? That's no solution

It's a solution for charging the batteries inside your vehicles.

This excludes connecting further to Germany’s neighbours, e.g. using Norway’s hydro exports as a battery.

https://en.m.wikipedia.org/wiki/European_super_grid

Ireland and Britain has massive offshore wind potential that is only being slowly ramped up, for example. Macron has announced a new nuclear reactor building program, so potentially more carbon-free electricity to nuclear-adverse Germany. The massive Danish offshore wind park/energy island has yet to come online but will supply northern Germany.

And if you want to go even more exotic, you could get 24/7/365 renewable energy by piping geothermal to continental Europe from Iceland which has an oversupply. Or even building enormous solar parks in the Sahara (geopolitical risks notwithstanding).

There are plenty of options enabled by geographic distribution of supply without even looking at storage. The main problem is the lack of political will until now, which is changing with the realisation in Germany that Russia isn’t a reliable partner.

Edit: none of these rely on any technical breakthroughs either, merely ramping up the interconnection of grids and renewable sources of electricity.

> we would need about 24 days worth of storage

Couldn't you just burn fossil fuels for those 24 days, and then do atmospheric carbon capture for the other 341?

Even if pulling the carbon out of the air takes 10 times as much energy as was released during those 24 days, that could still be achieved over the course of the year, with a reasonable amount of over provisioning.

In principle, yes.

In practice, it's far more efficient to capture right when you burn, as each ton of CO2 capture is many times cheaper then.

Hum. Over provisioning doesn't help when you have a month without wind (or you have to overprovision by like a factor of 100), neither does geographic distribution (cf UK electricity grid last year, the whole country goes down to a small production [1]).

Is biomethane considered as green? Isn't it like natural gas?

International grid ties basically means you are using coal/nuclear/etc from your neighbours, who will likely have shortage at the same time given they also have their own share of sun and wind.

[1] https://gridwatch.co.uk/

International grid doesn't mean "import whatever is on offer". You can contract for renewable energy, and fund international production capacity. Depending on Russian oil was a geopolitical (and environmental) mistake. Depending on French nuclear or Spanish solar seems much safer.

International grid ties also means:

* When it's cloudy or calm, you can import from somewhere sunny or windy. When it's sunny or windy, export to somewhere cloudy and calm. Shortages a thousand miles southwest aren't going to correlate strongly with shortages where you are, and if you if you have multiple sources a thousand miles in each direction you should be able to balance demand far better. Weather conditions are reasonably local. Demand patterns will also vary a bit.

* You can import solar from somewhere sunnier, or sunny at different. You're not going to transmit your electricity 12,000 miles east/west, but Madrid has sunset an hour later than Berlin, and demand tends to be closer to 7pm than 2am.

Yeah, it's a hard engineering problem, and there will be downsides, but we can do it if we seriously commit to it.

if you look at my response to a sibling comment, I am not convinced there is that much diversification in europe [1]. And you can tell annecdoctically, when you have heatwave news stories in one european countries, you typically get similar stories in other european countries. Europe is not that big geographically.

[1] https://news.ycombinator.com/item?id=30934588

Wow, thanks for putting together that data!

I'm not surprised to see Germany, UK, Norway, Sweden, and Netherlands all be similar, since (I assume) they're all using the North Sea. I am surprised to see Germany's output is so similar to Italy, Portugal, and Greece. It does look like there's a decent separation between places I expect to use the Mediterranean vs. those I'd expect to use the North Sea, but yeah, we likely need a lot of excess capacity, a lot of storage, or a lot of not-wind-or-solar. (Or I guess a lot of flexibility on energy usage.)

There has never been a single hour in the last 30 years where the wind has not been blowing somewhere in Europe.

Well, the wind blowing on some hill somewhere in europe isn't of much help, it needs to be diverse enough that some countries can make up for the entire lack of production of other countries.

I am not convinced there is much diversification. I just did some back of the enveloppe analysis to convince myself. I am using hourly wind electricity production data from this source [1] which is referenced on the europa website [2] (first source I found, I don't necessarily vouch for it). I only looked at 2014-2015 (the last two years of this time series), took the daily production at 2pm (seems to be peak production and peak demand), and then calculated a weekly average of the 2pm production by country for every day. I took the low points for the UK (arbitrarily, that's where I am now), and looked at the percentile of each country for those UK low points.

So the way you read this is for the week ending 28 June 2014, when the UK was at its 1% worst wind production across 2014-2015, at that time Norway was at its 15%th worst production, Sweden 9%th worst, etc. "15%th worst" means that the norwegian production for that week was below that level only 15% of the weeks in 2014-2015.

                     2014-06-28    2014-09-09    2014-09-16
  United Kingdom             1%            1%            2%
  Norway                    15%           21%           20%
  Sweden                     9%            0%           11%
  Netherlands                1%           19%           18%
  France                     2%            9%           15%
  Germany                   23%            1%           41%
  Italy                     63%            9%           18%
  Portugal                  32%            5%           60%
  Greece                    55%            5%            8%

so you do get some outliers (Italy and Greece in the first column, Portugal in the last), but overall, all the major countries are down at the same time, so it's not Portgual that will produce the wind energy for all the other countries.

[1] https://zenodo.org/record/4803353 [2] https://data.jrc.ec.europa.eu/dataset/jrc-emhires-wind-gener...

I've looked at studies from researchers running simulations on weather data sets, and yes, you need some dispatchable sources in Europe. The question is how much you need.

Generally, the biggest issue seems to be heating in the winter. Heat is actually much cheaper to store than electricity. If you have a district heating system, you can even store it seasonally (storing solar heat in the summer) and be competitive with gas, even before the recent price increases.

I think the main reason we don't see a lot of cheapish long term electricity storage yet is that it's not needed. Once we start producing it, economies of scale will kick in. That's at least how it worked for PV, wind turbines, li-ion batteries.

Europe is windier in the winter, so if you choose a week in the summer, they're all going to be low. This is not unknown or unexpected. It's also neatly complemented by solar which has the opposite yearly cycle:

https://mobile.twitter.com/ntsafos/status/148056006004465254...

I have done all this research in like 2017 and I am so tired of all the idiots slowing down humanity's progress.

Which is great when everyone likes each other. Which traditionally is a dubious bet to place in Europe when losing it results in serious things like ‘will all the old people freeze to death tomorrow because we don’t have enough electrical capacity’.

It's ok...we're over provisioned on solar generation capacity. So it doesn't matter that it is night :P.

Do you work in the factory at night, bally0241?

I do sometimes work nights, but not in a factory. But your missing the point, there are lots of industrial processes that you don't just turn off for 12 hours because the electricity is not there or more expensive. Spooling them back up can take far in excess of that, so a 50% duty cycle is a non-starter.

I'm sorry we can't all be laptop-class software devs, someone has to make the 'stuff'.

If every country main argument is 'we just use the international grid for balance' non of this will scale.

Over provisioning one thing just means its not in another place for the first decades.

And batteries. This is only viable for a few days at most though, it doesn't work for seasonal differences.