Wind And Solar Energy Lulls: Energy Storage in Germany

German emissions rose in both 2015 and 2016, despite addition of 30 TWh of renewable electricity. What happens when the rest of the nuke plants, worth almost 90 TWh of clean electricity, close?

Long ago Germany’s scientists had discussions and studied (Agora) about longer winter lulls (2 months).
As usual a mix of alternatives came up as best & cheapest solution.

The two most prominent:
– More interconnection capacity with more countries +
– Substantial over-capacity of wind & solar combined with Power-to-Gas.
are not stated in the post.
Only the unsuitable battery alternative.

While the problem won’t be urgent until 2035, Germany:
– is expanding interconnection capacities (e.g. to Norway)*)
– is developing Power-to-Gas fast.
The first major pilot plant (Falkenhagen) operates since 2013. Since then major ongoing improvements (efficiency & lower costs also due to smaller volume).
MIT research contributes at a fundamental level (efficiency, stable membranes, etc)

They now have ~30 major P2G pilot plants & research facilities. ***)
The produced syn gas is partly H2 injected in the grid**), some are situated at H2 car refill stations, etc. Partly the gas is methanized, e.g. the Audi e-gas project.
They target to have a capacity of 1GW PtG running in 2022, and to start with regular roll-out in 2025. So their plan allows for years of delays.

Note that:
– in order to cover 5 weeks of a complete lull in a grid with 100GW consumption, the av. PtG production only need to be ~10GW.
– it’s cheapest to use unmanned ‘peaker’ gas turbines, as those need lowest investment and maintenance. Especially since those will run less than 10% of the time. Siemens has H2 gas turbines in development.
– nuclear is a non-option as nuclear is widely considered being extremely dangerous (their scientists confirmed some effects of Chernobyl fall-out, despite being 1000mile away; Germans have relations with German speaking farmers in the Chernobyl area). Furthermore nuclear is n-times more expensive
_____
*) Note that there was no such lull in the wind during December in the Nordic.

**) For the climate it’s better that the renewable syn gas replaces natural gas burning immediately. First storage in earth cavities as may imply some losses.
Storage in earth cavities is extremely cheap. Germany use such stores to cover long supply interruptions of Russian gas.
In NL those allow to size the max. capacity of the gas processing plants on the average consumption, which is substantial less than the winter peak.

***) You can find out more about each through clicking on the dot.

What type of energy storage was being suggested with that magnitude of capacity? Or was the article simply pointing out the magnitude of the problem?

I know of various ways to store that much energy, but they’re not well known. Aside from large scale pumped hydro, none have been proven in any pilot demonstrations.

Gas storage in earth cavities has proven itself during many years for very cheap seasonal storage in NL, and for huge strategic reserve in Germany.
Such storage can handle more than 900TWh.

– in order to cover 5 weeks of a complete lull in a grid with 100GW consumption, the av. PtG production only need to be ~10GW.

That assumes 100% round trip efficiency. P2G2P is hard-pressed to get 35%, and that’s using costly hydrogen fuel cells for generation. Using cheap hydrogen-fueled peaker turbines, round trip efficiency would be closer to 15%. 20% would be generous. So you’re talking about dedicating additional power equal to more than 50% of average demand to making hydrogen for your 5 weeks of lull. That’s a complete non-starter.

Not any and all “gas”. Large scale methane storage is proven, and was similarly stored eons before the advent of humans.

I’d be happy to have a 5-day lull.  My area has gone through a full week of next to no wind and still has a couple more days forecast.

I saw an interesting article about P2F2P a while back, I think from MIT. Their idea was to store hydrogen for peaking power plants underground in salt caverns, but also store the co-produced oxygen in nearby caverns. This combination burns at a hotter temperature, thus allowing more efficient conversion to electricity (70% vs. 60%), as well as cutting NOx emissions.

The trick is, this was suggested for use with nuclear plants, with waste heat being used to keep the O2 warm. That way, if there is a large leak, the O2 will be lighter than air, and float high enough to disperse. Otherwise, a large O2 leak will produce a heavier-than-air holocaust plume, that drifts to the nearest city, finds an ignition source, and accelerates combustion of everything in its path.

Neither the battery nor the PHS storage appear to be even technically feasible to build given finite resources for the 100 hr, 20 TWh storage scenario.

For a battery solution, note that Tesla’s new battery factory will produce 53 GWh/yr and is doubling global Li-ion production capacity. With battery replacement every 10 years, 100 hr storage is never achieved even with dozens of such factories dedicated to German storage alone. BTW, the mass of 20 TWh of Li-ion battery (using Tesla’s 170 kWh/ton) would at least 100 million tons, and several times that for NAS.

Large PHS has a depth of a dozen hours or so, and the largest power output is 3 GW, with typical capacity much lower.

http://spectrum.ieee.org/energywise/green-tech/fuel-cells/new-flow-batte...

The late Swedish Social Democratic Prime Minister, Olof Palme, is claimed to have said that ” If the facts and the Social Democratic policy platform are in conflict, too bad for the facts.”

When sufficient capital and political capital have been invested into a project, it is usually declared a success, regardless of the problems or the facts. That’s the nature of politics.

Nathan,
I believe you are referring to this article by Charles Forsberg, discussing various options of utilizing Nuclear in conjunction with fossil extraction or pairing with different storage techniques. Short read and full of interesting options. HIPES is the particular P2F2P you describe.

“…100% round trip efficiency…“??
No.
The expert expectancy of (at least) 40% round trip efficiency will do.

‘normal’ years have say 3 weeks with wind&solar lull during which still a lot is produced by wind & solar, easily 20%. Furthermore:
– other renewable continue to produce ~10GW (~10%)
– import will deliver ~10GW (improved interconnections)
So the produced syn gas of those years is too much. The surplus will be kept in storage (earth cavities are cheap storage!).
So after a number of years there will be enough H2 in storage to cover a few real bad years.*)

Regarding economics, note that German wind & solar is now already 2 times cheaper than nuclear. That difference will be a factor >2 bigger in 2050.
So even unnecessary 25% or 50% overcapacity will then be much cheaper than nuclear!
____
*) note that the situation migrates during >20 towards that 100% renewable. So time enough to build up enough H2 storage.

Willem,
How come you think that it’s adequate to compare your solution for the winter lulls with a battery fantasy solution which was never taken serious in Germany as far better & cheaper solution range is available?
Which the Germans are developing actively further since roughly 2007. But you ignore those developments, while you lived in the EU and seem to know German.

Considering also your misrepresentation of the German electricity prices by not adding the remark that taxes play a major role (little tax in France, lot tax in Germany),

I get the impression you want to show the Energiewende is / will be a failure even if that story / argument doesn’t fit with reality….

Hmm, you’re reframing the issue on the fly, Bas. You had written:

– it’s cheapest to use unmanned ‘peaker’ gas turbines, as those need lowest investment and maintenance. Especially since those will run less than 10% of the time.

You’re right to be concerned about the cost of the G2P leg of the system, since at a 10% capacity factor, the cost of power will be dominated by the cost of capital. But a cheap peaker unit will only give about 27% thermal efficiency. And that’s from gas that was produced from low voltage DC to the electrolysis cells at an optimistic efficiency of 80%. (It’s possible to split water with somewhat higher efficiency than that, but it requires operating the cells at very low current. Totally demolishing their capital productivity.)

Factor in’ perhaps 92% conversion efficiency for grid AC to LVDC distributed to the electrolysis cells, and your “cheap” P2G2P system is operating at .27 x .80 x .92 = 19.8% round-trip efficiency. Oh, and that’s ignoring the considerable energy required to compress the hydrogen for storage in “earth cavities”.

(BTW, did you really mean “cavities” — i.e., salt caverns of the type used for oil and gas storage in the US Gulf Coast region — or did you mean depleted gas fields of porous rock below an impermeable cap? I believe the latter are much more common for gas storage than the former, and they certainly have more capacity, but I don’t know that they’ve ever been used for hydrogen storage. They likely have issues with the fraction of stored gas than can be recovered, and they certainly would have issues with purity of the gas recovered from storage. It would contain substantial amounts of methane and other light hydrocarbons. Wouldn’t be a problem for burning the gas in a turbine, but would be for a PEM fuel cell. Salt caverns might save a bit on compression energy, but I’m not sure there’s enough storage capacity there for the magnitude of hydrogen storage one would need for P2G2P over one-year timeframes.)

Siemens has H2 gas turbines in development.

I don’t know what Siemens may be developing. I’m pretty sure, however, that if it’s cheap, it won’t be efficient, and if it’s efficient, it won’t be cheap.

The expert expectancy of (at least) 40% round trip efficiency will do.

As to that, 40% is indeed do-able, but it means you’re back to using expensive fuel cells, or at least high efficiency CCGT units, for the G2P leg. It will be a lot more expensive than firing up a gas-fueled peaker to supply power during lulls. If it’s mandated by German law, it will be done (in Germany). But it will push the overall cost of electricity still higher. And you’re already getting pushback from German businesses over the high cost of electricity.

And now I’ll do a little reframing of my own.

We’ve been talking about the impact of extended lulls in wind and solar power, of the sort that hit Germany this winter. An implicit assumption has been that production from wind and solar vary stochastically around an overall yearly average. I don’t know to what extent that holds for wind in Europe, but I know it does not hold even remotely for solar.

At the latitude of Frankfurt, there are 16 hours of daylight in summer and 8 in winter. But that 2:1 ratio understates the summer to winter variance by a wide margin. The winter sun has a low elevation angle even at noon, and delivers only about half the wattage per square meter as the summer sun, even in clear air with a tracking array. If there’s any haze or cloud cover, it’s worse. Of course Frankfurt is in the north of Germany, but things aren’t very much better even at Germany’s southern border. 5:1 seems a conservative estimate of the solar power variance between the mid summer and mid winter.

So it’s not just 5 weeks of non-production every few years that your stored energy system needs to cover. It’s more like four months of average consumption every year. Good luck maintaining a competitive economy with the cost of electricity that that order of energy storage is going to mean.

Link to “this article” failed to come through. Googling for “Charles Forsberg HIPES” will get one immediately to the abstract, but the contents have to be purchased. I found this presentation to be adequately informative, with the advantage of covering more options.

One point that the presentation brings out is that the pure oxygen co-product of any P2G scheme does carry a high value. I don’t know about storing large quantities of it; that’s a little scary. But using it for oxy-fueled combustion of hydrocarbons with sequestration of the CO2 output stream makes sense.

For stationary applications, flow batteries seem more attractive than numerous small cells, especially in their apparent ease of recycling the active materials.

That said, I think the growing BEV market will push solid batteries down in cost so rapidly that flow batteries will fall far behind.

Another problem is that batteries are entering the grid storage market at the 0.2-1 hour range (spinning reserves), and slowly shifting to 5 hours, to serve peak evening hours. Flow batteries have better economics at longer run-time (which give more daily energy for the same pumps and membranes), but longer run-time storage will compete mostly with cheap nighttime power, and only be relevant at high solar penetration. And at high solar penetration, solar loses it’s capacity value, even with storage, so a storable-fuel backup is needed.

In other words, after the 5 hour market segment, the next category is not 12 hours, it’s weeks as Willem suggests (the exception is thermal energy storage at CSP and nuclear plants, where that 12 hour thermal storage is backed up with weeks of combustion fuel). Note that li-ion batteries (and therefore the improved flow battery in the article) have an energy density an order of magnitude worse than liquid ammonia (a storable fuel).

So flow batteries get stuck between cell batteries and methane/ammonia/hydrogen storage, and can never gain a foothold.

Willem,

Having read the post and the comments, it seems obvious that the Germans are going to have capacity markets, in other words, fossil fuel plants are going to kept in reserve and their owners compensated for keeping the plants in running order.

The German goal of 80% renewable power in 2050 does not mean that there are natural gas and coal plants providing a steady 20% stream of the power 24/7 and steady 80% of renewable generation.

In practice, Germany will need 200-300 GW of solar and wind capacity to provide that 80%. That means massive oversupply at times. Then there are those lulls when wind and solar generation is close to zero the fossil capacity is needed. And lots of it.

Short-term storage will be developed but if lulls continue for over 12 hours, storage becomes prohibitively expensive. The German system will probably become some sort of a hybrid. The fossil fuel plants already exist so there is no need for new investment.

Finally, I must say that one leg of Energiewende, namely increased efficiency gains that supposedly decrease electricity demand, is unlikely. Power supply is just one source of emissions. Germany needs to clean up transportation emissions (oil) and heating emissions (natural gas). France and Finland use a lot of electricity for heating and I think that is also the way Germany will choose. Add millions of BEVs and the equation is clear.

We talk about a fictitious situation which won’t become an issue in next decade. So we won’t know who is right.

Since I feel we won’t agree on the numbers, I don’t want to spend more time on this imagined issue.

Btw.
We in NL store the processed gas (to cover higher demand in winter) in a few salt domes (we have more).

Jarmo,
Seems to me Germany is more prone to follow Denmark where only energy neutral new buildings are allowed, etc.

I have a particular gripe about energy efficient and energy neutral buildings. In Finland, they started building those after the oil crisis in the ´70s and 80s.

It’s like living in a plastic bag. The buildings don’t “breathe”, condensation builds up easily in the structures. Worst are the indoor air problems. Any problems in the mechanical air circulation and that’s it.

The is even an organisation in Finland advocating for the so-called “mold refugees”:

http://www.homepakolaiset.fi/in_english.html

Willem,

Germans talk about “capacity reserve” which is different from capacity market as in the UK. At least that is my understanding.

We had similar in NL. But those designs improved during >30years, Now those issues are very rate, if they occur at all.

Jarmo, agree.
German authorities rejected capacity market ideas as being unnecessary high subsidies for incumbent utilities with old plants, keeping those plants longer open than necessary.*)

But they do create temporal reserves if necessary.
E.g.
When Merkel closed 8 NPP’s after Fukushima (spring 2011), they arranged cold standby with 2 already closed plants during next winter (2011/’12). During that winter one was asked to become hot standby during a week. But none had to deliver electricity.
_______
*) The utilities then united with more utilities from other countries (Spain, UK, etc), arranged a study which showed such capacity markets would be necessary, mobilized publicity and political support and then requested the EU to implement those.

So German authorities had to explain in Brussels that those are a waste of money. As the issue was not solved, Brussels decided that countries can decide themselves.

Now you see only capacity markets in countries such as UK and Spain, where incumbent utilities have prime influence on politics.
Btw. those also have high whole sale prices…

Willem,

Just making sure apples are apples. Personally, I think Merkel & co are playing with semantics. Market or reserve, the result is identical.

@Bentvels – Nuclear power generation has the smallest deathprint, rarely discussed. The deathprint is the number of people killed by one kind of energy per kWhr produced, coal is the worst and nuclear is the best. Nuclear is NOT n-times more expensive, when excluding the first built Westinghouse AP1000’s and the first built Areva EPR’s. South Korea and China have already proven this. Furthermore, GEN III and GEN IV reactors are no longer considered to be dangerous. Nuclear power generation in Germany (when taking the capacity factors into account) is still well over twice that of Wind and Solar combined. GLOBAL LIVE POWER USEAGE: http://data.reneweconomy.com/LiveGen

Harry, Death print
Sorry, but the idea that nuclear would have a small deathprint is imagination. It’s only true when one accepts the unrealistic estimations of pro-nuclear.
A typical example is the publication by (in nuclear circles extolled) climate guru James Hansen & Pushker Kharecha: 43 deaths for Chernobyl.
While other estimates are up to 4-8 million until the end of this century.

Read the Annals of the New York Academy of Sciences when you want to have a more realistic view regarding the huge death print of nuclear.

Jarmo,
Except that the costs for capacity markets are factor 2 -4 higher…

Harry, Nuclear is expensive
FOAK?
Check the huge cost increase in France in last century, when nuclear had to become more safe. So FOAK hardly seem to play a role, as confirmed when one checks the new reactor generation:
– Hinkley C are 5 and 6 of the EPR. Still the price is 2-5 times that of wind & solar.
– The AP1000’s in USA are number 3-6. Not clear what the costs are but calculating with the official costs deliver already <10cnt/KWh (not including the costs of the loan guarantee subsidy. neither the $15B loss of Westinghouse.
So in the end the AP100) will costs similar as the EPR or even more.

China, S.Korea
We don’t know the real costs in communist China, neither those in S.Korea as there the utility is building himself, which delivers excellent options for invisible cross-subsidies. Considering their wide-spread fraud with nuclear parts, we can assume that they use those.

Anyway when you want to compare, you have to compensate for the really low wages and much higher productivity (longer working hours, more dedicated) of workers in China (have experience with that). Assume similar with S.Korea.

Those are also the cause Chinese (& Korean) products are so cheap that they compete us (in the west) off the market.
For solar panels they compete even while:
– we have import tariffs of ~45%, and
– those are far less labor intensive than NPP’s.

Not dangerous. GEN III and GEN IV no longer???
Same was promoted before Chernobyl.
Same was promoted for the Fukushima generation after Chernobyl…
Meanwhile even GEN III+ will end in similar disaster when attacked with a big freight plane loaded with steel, etc. Nowadays attackers, e.g. IS, may use more advanced weapons incl. drones and remote controlled planes.

Furthermore: didn’t see measures with GEN III+ against the regular spreading of radio-activity, which nuclear facilities incl nuclear power plants, do.
Which brings highly significant genetic, hence also health, damage to newborn up to 40km away. Especial important as genetic damage is transmitted to next generations.

Pro-nuclear seem to care little about health and intelligence of their next generations.
German govt does.
So they closed their prime dry cask nuclear waste store, Gorleben, prematurely (while the huge building was still half empty) because of that damage

Click “Jahresbericht Umgebungsüberwachung 2014” at the link and you can see pictures of the Gorleben site with its huge building with 0.5m thick walls in which the dry casks are stored and the high dike around it, as well as the excellent radiation measurement network, incl. neutrons, in and around it.

“flow batteries get stuck between cell batteries and methane/ammonia/hydrogen storage”
Sorry, don’t share your opinion. Even German grid installs flow batteries, while they also have ~35 pumped storage facilities…
Together with pumped storage, they fill the gap between ‘normal’ batteries and Power-to-Gas which supports seasonal storage. Their benefit is that they can be installed everywhere and have higher round trip efficiency than pumped storage.

Btw.
With the increasing penetration level of solar panels, cheap night time rates are ending. Here the difference between the day-time and night-time rates decreased already a lot in past decade. So much that I was questioning myself whether it was worth the extra costs of the double tariff meter. But we get a smart meter before 2020, so more flexibility. Hence we may see lower tariffs during sunny days (so I will then adjust the hours my dish washer, washing & dry machines run).

Well the quality of the article is shown in the size of existing pumped storage being too low by factor 4 – the main pumped storage is missing, maybe because it’s not a short time storage but a weeks-storage. (Schluchseewerke).
Beisde this, main balancing of lulls etc is done by grid, not by storage. Storages are just needed when grids are forbidden in the model.