Many recent posts on this blog have dealt with the theme of the infeasibility of a fully wind/solar/storage electricity system. Today I will deal with another study of the subject, this one from German authors Oliver Ruhnau and Staffan Qvist, titled “Storage requirements in a 100% renewable electricity system: Extreme events and inter-annual variability.” The Ruhnau/Qvist study does not have a date other than “2021,” although it appears to have come out toward the end of that year.

Although Ruhnau and Qvist do not say it explicitly, my conclusion from their paper is that it is a further demonstration of the complete infeasibility — indeed the complete absurdity — of attempting in the short term to replace all fossil fuel electricity generation in a modern economy with only wind, solar and storage.

The background of this issue is that large numbers of green activists, up to and including the current President of the United States, make regular statements indicating that they believe that fossil fuels can be eliminated from the modern economy by simply building sufficient capacity of wind and solar electricity generation. Such statements rarely consider or mention the necessity of energy storage, or the feasibility or cost of same. And yet any serious consideration of the intermittency of wind and solar inevitably leads to the conclusion that without dispatchable backup (fossil fuel or nuclear) they require vast amounts of energy storage to cover the periods of intermittency. Understanding the amount of storage required, its physical characteristics, and its cost, is completely essential to answering the question of whether a fully wind/solar/storage system is feasible.

And yet our governments are currently marching ahead with religious zeal with plans for “net zero” electricity generation, based almost entirely on wind and sun, without any serious consideration of the amount of storage required or of the cost or feasibility of the project. Nor has there ever been a demonstration of a workable prototype system that could achieve net zero emissions with only wind, sun and storage, even for a small town or an island.

Previous posts at Manhattan Contrarian on this subject have reviewed detailed work by Roger Andrews and by Ken Gregory. In this post from November 2018, I reviewed work by Andrews dealing with actual wind and solar generation data from the two cases of California and Germany. Andrews concluded that due to seasonal patterns of wind and solar generation, either California or Germany would require approximately 30 full days of energy storage to back up a fully wind/solar generation system. Based on current costs of lithium-ion batteries, Andrews calculated that building sufficient wind and solar generation plus sufficient batteries would lead to a multiplication of the cost of electricity by approximately a factor of between 14 and 22. In this post from January 2022, I reviewed work by Gregory dealing with actual wind/solar generation for the case of the entire United States. Gregory considered how much storage would suffice as the sole back up where the U.S. had fully electrified all currently non-electrified sectors (e.g., transport, home heat, industry, agriculture), thus essentially tripling electricity demand from the current level. His conclusion was that the batteries alone would cost about $400 trillion — about 20 times the full GDP of the United States.

Clearly, if either Andrews or Gregory is anywhere near right, converting a modern economy to fully wind, solar and storage is not remotely feasible.

Into this mix now come Ruhnau and Qvist. The focus of R&Q is once again the amount of storage needed to back up a fully wind/solar generation system, once fossil fuels have been eliminated as a back up option. The R&Q study deals only with the case of Germany, and only with supplying its current level of electricity demand, rather than demand that may be tripled or more by economy-wide electrification of transport, heating, and so forth.

The bottom line is that the result of the R&Q study is approximately in line with the findings of Andrews and Gregory. Where Andrews and Gregory had calculated that about 30 days of storage would be required to back up a fully wind/solar system, R&Q come up with 24 days. However, to get to the 24 day result, R&Q require massive overbuilding of the wind/solar system, to the point where its nameplate “capacity” is about triple Germany’s peak electricity demand, and five times average demand. The result is a system where vast amounts of surplus electricity on sunny/windy days must be discarded or “curtailed.” However, R&Q say that their model is based on cost minimization, because building vast excess capacity and discarding electricity by the terawatt hour is actually cheaper than adding additional storage.

The starting point of the R&Q study is a critique of prior authors who have calculated relatively low storage requirements by only looking at a supposed worst case multi-day wind/solar “drought” of calm and cloudy days. Some such studies cited by R&Q have derived storage requirements in the range of 4 - 8 days as supposedly sufficient to back up a fully wind/solar system. (Even those levels of storage requirements would likely make the cost infeasible.). But R&Q use available hourly wind and solar generation data over the course of entire years for Germany to show that much longer periods of relative calm and dark can occur, causing the storage requirement needed to avoid blackouts to be much higher.

While our time series analysis supports previous findings that periods with persistently scarce supply last no longer than two weeks, we find that the maximum energy deficit occurs over a much longer period of nine weeks. This is because multiple scarce periods can closely follow each other. When considering storage losses and charging limitations, the period defining storage requirements extends over as much as 12 weeks. For this longer period, the cost-optimal storage capacity is about three times larger compared to the energy deficit of the scarcest two weeks.

At pages 5-6 of their paper, R&Q lay out the generation (installed capacity) and storage requirements for their view of an optimized system.

First there will be a vastly over-built system of wind and solar facilities:

On the supply side, almost 300 GW of variable renewable generators are installed: 92 GW solar PV, 94 GW onshore wind, and 98 GW offshore wind . . . . For solar PV and onshore wind power, this is nearly twice as much as the installed capacity in 2020; for offshore wind power, this means more than a tenfold increase.

For comparison, Germany’s current peak demand is in the range of 100 GW, and average demand is in the range of 60 GW.

Then there will be some 56 TWh of storage, equivalent as discussed to about 24 days of full electricity consumption for the entire country of Germany at near-peak usage levels. To get a handle on how much that is, consider that a Tesla battery is in the range of about 100 KWh, and sells for about $13,500, or $135/KWh. So, if you were trying to cover the 56 TWh of storage with Tesla-type batteries, it would run you around 56,000,000,000 x $135, or about $7.56 trillion — which is about double the GDP of Germany.

But R&Q think they have a better idea than batteries, namely hydrogen as a vehicle for the storage. In their model, almost all (54.8 TWh out of the 56 TWh) of the storage comes from hydrogen. In the first instance, this requires adding yet another massive new cost element to the system, namely an entire network of some 62 GW of hydrogen-fired CCGT power plants, almost sufficient on their own to supply Germany’s grid at average levels of demand.

Add together the cost of three-times overbuilding of wind turbines and solar panels, 56 TWh of storage, and a network of new hydrogen-fired power plants almost as extensive as Germany’s entire current generation system, and you have a collection of costs that can’t possibly be feasible in any rational world.

And yet somehow, when R&Q get to their conclusions with respect to feasibility, they wave their hands and say there is no problem. Although they concede that there exists no utility-scale hydrogen storage, distribution and combustion system anywhere in the world as a basis to calculate costs, they somehow come up with a figure of 30 euros per MWH of load for the cost of the storage — less than the cost of Tesla-style batteries by a factor of over one thousand. Is there any basis? The closest they come is this:

As underground hydrogen storage is currently limited to pilot systems in Germany, the currently 250 TWh of German natural gas storage, which is mostly underground storage in salt caverns, may serve as a reference.

Unfortunately I don’t think that underground storage of natural gas is at all a valid reference. Natural gas can effectively be stored in non-airtight things like salt caverns because it does not ignite when it goes above about a 15% concentration in the air. Sadly, not so for hydrogen. Hydrogen also rapidly corrodes and leaks from pipelines and containers, causing potentially extreme hazards. I don’t claim to know all the engineering challenges of making a safe hydrogen-based electricity system, but they are clearly huge. If dealing with hydrogen in massive quantities were safe and easy, plenty would be doing it already. There is a reason that no massive hydrogen storage facilities or hydrogen pipelines exist.

The simple answer to all of this is that we must demand from our politicians a demonstration of feasibility of any replacement energy system before we embark on these multi-trillion fantasy building projects. Show us a fully wind/solar/battery or wind/solar/hydrogen system that works at reasonable cost for 5000 or 10,000 people over the course of a few years, before requiring entire countries of tens or hundreds of millions of people to be the guinea pigs. The idea that we would embark on replacement of our entire energy systems without demonstration of feasibility of the replacement is pure madness.