TransitionFiles.com
← Back to overview
Chris Author
System Thinker for a Sustainable Industry Transition in EU
· · 10 min read

From Moscow to Washington: We Swapped the Dependency, We Did Not End It

In mid-September, Germany's gas storage sites are just over 55 per cent full, roughly 136 terawatt hours. The Federal Network Agency points out that this is already slightly more than was withdrawn over the whole of last winter – just under 134 terawatt hours – and that injection can continue until the end of October. That is correct. It is also the friendliest reading the numbers allow. The gas price is around 134 per cent above last year's level, and according to the storage operators' scenarios a normal winter requires about 77 per cent by 1 November.

I wrote about the storage sites two weeks ago. Today it is about the question behind them: where does the gas come from that is supposed to fill the gap – and who, as a result, owns our winter?

The answer is the United States, and that should reassure no one

In 2025, 96 per cent of the liquefied gas arriving through Germany's terminals came from the United States. That is a good ten per cent of all German gas imports. Add the American LNG that is passed on via Belgium and the Netherlands, and the US share is around one fifth. At EU level, 29 per cent of all gas imports in the first quarter of 2026 came from the United States; for LNG alone it was 57 per cent. Norway remains Germany's most important supplier at 44 per cent – by pipeline, secured under long-term contracts, and at the limit of its capacity.

Then there is a contract. In the Turnberry agreement of July 2025, the EU committed to purchasing around 750 billion dollars' worth of US energy by the end of 2028, predominantly LNG and oil – in return for a tariff cap of 15 per cent. So we have not merely accepted the dependency; we have signed it.

Now, dependency on the United States is not the same as dependency on Russia, and anyone who equates the two is making it too easy for themselves. Russia could turn off a pipeline; the gas was physically gone. American LNG is a commodity on a world market. If Washington stopped exports to Europe, the gas would go to Asia and we would buy other gas away from there – dearer, but available. A market expert interviewed by ntv therefore regards the mere announcement of an export halt as the real lever: it drives the price, not the availability. That is the difference. Moscow had its hand on the valve. Washington has its hand on the price.

For a household heating its home in January, that is no consolation. And for the chemical industry, the largest industrial gas consumer in Germany, the price decides whether a site produces or not.

What can be done for this winter

The McKinsey analysis from March reaches a sober conclusion: supply is more robust today than in 2022, and a disruption can technically be compensated – at high cost and with tangible risks. The building blocks are known, and they are finite.

First: inject now. Six weeks remain. That the economics minister rejects state intervention is a decision, not a law of nature. A state-procured reserve of ten terawatt hours would buy the insurance that the market does not buy when summer prices exceed winter prices. It costs money that the Federal Network Agency would have to spread via a levy anyway – just before the winter rather than after it.

Second: max out the pipelines. Norway is delivering close to its capacity limit; not much more will come from there. From France, around seven billion cubic metres a year have been possible since this year, now that the deodorisation plant that had limited eastward flows is in place. That is the quietest and most robust reserve we have.

Third: buy LNG more broadly. Qatar's export infrastructure is damaged and will be out for the medium term. That leaves Canada, Nigeria, Algeria, Egypt – dearer, further away, but not tied to Washington. And there is a homework assignment that is embarrassing: the terminal at Stade could regasify 7.7 billion cubic metres a year, but can feed only 1.7 into the grid by year end. The grid connection was only approved in March. We built terminals and forgot the pipe.

Fourth: demand. Forty per cent of German gas consumption goes to households and commerce, sixty to industry. One degree less room temperature saves around six per cent. In power generation, gas plants can be replaced by coal and oil for a few weeks – a step backwards for climate policy, a buffer for security of supply. And industry showed in 2022 that fuel switching and process adjustments in the double-digit percentage range are possible when the price forces them.

Fifth: buy as Europeans rather than against each other. The EU platform for joint procurement has existed since 2023. If twenty countries bid for the same tankers in January, the tanker wins.

That is enough for a normal winter. For a cold January it is probably enough at prices we already saw once in 2022. For a cold January combined with a supply halt from Washington it is not enough without cuts. That is the situation, and it was avoidable.

The only way out: needing less gas

Everything so far manages the problem. It is solved at one point only: the quantity of natural gas we still have to import at all. Every terawatt hour that disappears from the system is one that neither Moscow nor Washington nor Doha can negotiate with us about.

The path there has three layers, and they interlock.

The first layer is electrification. Power from wind and sun is the only energy that is produced in Germany itself in relevant quantities. Heat pumps instead of gas boilers, electric process heat instead of gas burners in industry, battery-electric trucks instead of diesel – I wrote about that a week ago. The price is a power system that doubles in size: annual demand rises from around 500 to over 1,100 terawatt hours by 2038, peak load from 80 to a good 190 gigawatts, mainly due to heat pumps in winter. The grid development plan reckons with 630 gigawatts of grid capacity against 200 today. That is the largest infrastructure task since post-war reconstruction, and it is the prerequisite for everything else.

The second layer is batteries. In June 2026, around 30 gigawatt hours of battery storage were installed in Germany. They balance hours – midday against evening, wind against calm overnight. They do that very well, very fast and by now very cheaply. What they cannot do: bridge a week of dark, windless weather in January. No realistic battery build-out changes that. For hours, batteries are the answer. For days and weeks, a different medium is needed.

The third layer is hydrogen – as a bridge and as storage. Here it pays to be precise, because a great deal of nonsense has been talked about hydrogen. Hydrogen will not run a heating system; it is too expensive and too inefficient for that. Nor will hydrogen power trucks – the market has decided that. Hydrogen has two jobs, and both are indispensable.

One is the bridge. The power plant security act has been tendering 12 gigawatts of new capacity since this year: hydrogen-ready gas plants, storage and flexible loads. These plants will initially run on natural gas, a few hundred hours a year, precisely during the dark lulls – and will be converted as soon as hydrogen is available. Then there is the core network: around 9,000 kilometres by 2032, 60 per cent of it repurposed natural gas pipelines. We are literally building the hydrogen economy into the pipes through which the gas from Texas still flows today.

The other job is seasonal storage. Summer power that is surplus in July is turned into hydrogen by electrolysis, stored in caverns and converted back to electricity in January. That is inefficient – two thirds of the energy is lost along the way – and it is nevertheless the only way to carry energy from one season to another. Germany has a geological advantage here that almost no neighbour has: salt domes in which caverns can be created. RWE will commission Germany's first commercial hydrogen storage facility at Gronau-Epe in 2027, 28 million cubic metres. Today's natural gas storage holds 250 terawatt hours; as hydrogen storage, given the lower energy density, that would be only around 50. A new cavern takes eleven years from planning to operation.

Eleven years. Anyone who wants to store seasonally in 2037 has to start in 2026. That is not a metaphor; that is the construction time.

Where we actually stand

Honestly: the ramp-up is stalling. Of the ten gigawatts of electrolysis that were supposed to be in place by 2030, a fraction is in operation. The steel companies are postponing their direct reduction plants. Uniper is planning larger storage capacities only from the mid-2030s, on condition that they pay. And grid charges have doubled in five years.

One could conclude from this that the transformation has failed. I conclude the opposite. The winter of 2026/27 shows what the alternative costs: we are paying two and a half times the price for gas whose origin we cannot choose, to a supplier who understands the price as a political instrument, and we have written it into a contract until 2028. Every euro that goes into a heat pump, a substation or a cavern is a euro we do not have to transfer to Texas year after year.

The storage sites save this winter. The transformation saves the next twenty. Both together would be precaution. Either alone is hope.

My question to you: where would you put the first euro – into the strategic gas reserve for this winter, or into the cavern that will be ready in 2037?

Sources

0 comments
Like Comment

Comments (0)

No comments yet – be the first voice.