Precaution Is Not Optional: Gas Storage and a Short Memory

Germany's gas storage sites were around 54 per cent full at the beginning of September. A year earlier the figure was 71 per cent. The level has never been this low at this point in the year since fill levels began to be recorded fifteen years ago. At the same time, gas is trading in Amsterdam at around 79 euros per megawatt hour – before the blockade of the Strait of Hormuz in March it was below 32. Each figure on its own is uncomfortable. Together they make a problem that could be seen coming. Let me start with the obvious: storage is insurance. You take it out before the fire, and you pay the premium during the years when nothing happens. That is precisely its purpose. Anyone who starts filling once the scarcity has arrived is buying insurance while the building burns – and pays accordingly. The calculation behind this is not a forecast but arithmetic. The Initiative Energien Speichern has set it out: a normal winter requires about 77 per cent by 1 November, and the buffer then lasts into spring. In extreme cold it does not last – on individual days in January a gap of more than 25 per cent would open up. So we are starting with a cushion that is barely sufficient in average weather and insufficient in a cold January. Why has so little been put into storage? This is the part worth arguing about. It was not carelessness on the part of individual companies. It was the predictable consequence of how the incentives were set. Until the end of 2025 there was a storage levy that spread the cost of filling across all consumers. It was abolished on 1 January 2026 – for understandable reasons, it weighed on prices. At the same time the filling requirements were relaxed: Germany requires 80 per cent per facility, while the EU target has been softened with leeway of ten to fifteen percentage points. And then the market said: not worth it. When short-term gas costs more than delivery in winter, everyone who fills storage now loses money. For an individual company it is commercially correct to refrain. For a country the result is wrong. This is the classic case of market failure: security of supply is a common good that no one wants to pay for alone. Anyone who scraps the levy, loosens the requirements and then trusts the market does not end up with full storage. They end up with exactly what we are seeing now. That was not unfortunate coincidence but a decision with a foreseeable consequence. What the war has to do with it – and what it does not Since March, hardly a tanker has passed through the Strait of Hormuz. Around one fifth of global LNG trade normally runs through that channel, and Qatar alone supplies some 19 per cent of global LNG exports. QatarEnergy has extended its force majeure declaration into mid-October. On top of that comes a prolonged outage at the Norwegian Ormen Lange field. That is a genuine shock, and nobody had a war in their storage planning. But: the shock began in March. The injection season runs from April to October. So there were six months in which the situation was known. Storage is not empty because something took us by surprise, but because we carried on after the surprise as before. That Covestro, a chemical company, now speaks openly of a "certain risk" to supply over the winter is notable. Companies of that size do not put such sentences on the record lightly. What I take from this Three things, none of them new. First: security of supply costs money even when nothing happens. Those costs are the price of precaution, not evidence of its failure. Abolishing a levy without organising the contribution to filling in some other way merely shifts the cost – from summer into winter, and usually with a mark-up. Second: when the market systematically rewards the wrong behaviour on a question of security, you have to change the framework rather than hope for insight. Tenders for strategic volumes, hedging of the price risk, binding interim levels instead of a single target on 1 November – the instruments are on the table and have been discussed since 2022. Third, and this is the uncomfortable part: every kilowatt hour of heat that does not come from gas makes this debate smaller. Almost every second home in Germany is heated with gas. As long as that holds, we will have this conversation every August. For this winter that changes nothing. For the ones after it, it decides everything. The winter will probably turn out fine. A mild January, an agreement in the Gulf, a few additional LNG cargoes – and by spring nobody will be talking about it. That is precisely the pattern. We get through, draw no lesson, and next time we start again with the same question. My question to you: do you consider binding fill requirements with state hedging of the price risk to be the right approach – or should the state stay out of storage management and instead cut gas demand in earnest? Sources • ZDF: Historically low fill level – will gas run short this winter? (8 September 2026) https://www.zdfheute.de/wirtschaft/gasspeicher-fuellstand… • Initiative Energien Speichern (INES), gas scenarios September 2026, as cited in: https://www.drweb.de/gasspeicher-zu-53-prozent-reicht-das… • Cleanthinking: gas storage levels and gas reserve 2026, with references on the German storage ordinance, the EU target and the storage levy https://www.cleanthinking.de/gasreserve-gasspeicher-2026/ • Technik + Einkauf: consequences of the Iran war for LNG supply and gas prices (March 2026) https://www.technik-einkauf.de/energiebeschaffung/explodi… • Handelsblatt, Catiana Krapp and Bert Fröndhoff: research on gas supply over the winter, including the quoted assessment from Covestro