A quarter less gas in storage. The margin for error is shrinking.

48,20 TWh less than a year ago

143,49 rather than 191,69 TWh. That was the working gas stored on 30.09.2026 and 30.09.2025 respectively. Using unrounded figures: (191,6921 − 143,4933) ÷ 191,6921 × 100 = 25,14 percent less gas. The fill level fell from 76,68 to 57,97 percent — a decline of 18,71 percentage points. GIE AGSI+, calculation by OPTUM.

Demand did not fall alongside it. January to August 2026 consumption was around 552,6 TWh, against 543,1 TWh in the same period a year earlier — up 1,76 percent. September is not yet fully available in this series. Bundesnetzagentur/THE, our sum of monthly figures.

On 2 October, the Bundesnetzagentur rated supply as stable but expected a comparatively low stock at the end of winter to be likely. Our conclusion: with a smaller reserve, ongoing deliveries and demand must match more reliably. Official assessment.

Check current storage figures for yourself: the gas storage monitor lets readers compare reported stocks and historical trends, alongside individual facility reports and the European comparison. This article uses the fixed reference date 30.09.2026. Consumption, weather and prices are supported by their respective primary sources.

Reported capacity fell only from 249,98 to 247,53 TWh. The substantial loss of gas therefore cannot be explained by a comparably large reduction in reported storage space. Compared with the end of September 2024, storage holds 97,70 TWh less.

The 25,14 percent refers neither to total storage capacity nor to annual consumption. For the same gas day, the EU reports 71,50 percent and 809,1 TWh, with estimated status. The European comparison supplements the German balance; an average cannot replace a review of transport routes.

An earlier winter used 72 TWh more

602,6 TWh in winter 2025/26, 674,7 TWh in winter 2020/21. The earlier period was around 72,1 TWh or almost twelve percent higher. This is observed demand, not an invented stress assumption. It does not predict a repeat. Bundesnetzagentur: current series; historical SLP and RLM figures.

Cold weather affects a large share of consumption. Households and smaller commercial customers accounted for 50,1 percent in January 2026 and 46,5 percent, or 280,2 TWh, over the winter. This includes more than space heating alone. Around 56 percent of homes are heated with gas, including biomethane and LPG. Bundesnetzagentur; BDEW 2024.

Consumption alone does not explain storage withdrawals. Total consumption in 2024/25 was 596,9 TWh, slightly below 2025/26. Yet storage supplied a net 170,13 rather than 137,51 TWh. Ongoing imports, exports and other balance factors also matter. Looking only at weather understates this dependence.

Source: Bundesnetzagentur/THE. Our sum of published monthly averages in GWh/day × actual calendar days ÷ 1.000. Values are approximate because the source figures are rounded. SLP is calculated from network data and can be revised; RLM also includes gas-fired power generation. This is not added to total consumption a second time.

The annual view also challenges an easy extrapolation: gas consumption was around 1.029 TWh in 2021 and around 864 TWh in 2025. An earlier, higher level cannot simply be treated as today’s requirement — production, savings and technology have changed. It does show why a higher-consumption scenario deserves examination.

Cold weather can raise demand quickly. Total consumption in January 2026 was around 136,0 TWh, ten percent above January 2025. Heating can be managed more efficiently, but a cold building cannot wait months for a cheaper delivery date. We do not attribute the whole increase in consumption to weather alone.

Four historical winters would require more gas

We apply documented net withdrawals to the opening stock for 2026. Calculation: 143,4933 TWh minus historical withdrawals plus historical injections, in each case from October to March. The table shows the results of this conditional winter balance.

Even the last winter would leave little room. Net withdrawals in 2025/26 were 137,51 TWh: 158,70 TWh withdrawn minus 21,19 TWh injected. With today’s opening figure, 5,98 TWh would remain at the end of March, falling temporarily to just 2,40 TWh in the replayed February path. A positive balance does not prove adequate daily delivery capacity.

Withdrawals like 2024/25 exceed the opening stock by 26,63 TWh. This additional volume is not in storage at the reference date. In the model, it would have to be covered in time by higher net inflows or lower consumption. This says nothing about supply contracts already in place. Once the calculation falls below zero, a real adjustment is needed.

2017/18 would require as much as 29,62 TWh of additional cover by the end of March; 2016/17 and 2020/21 would also require more than is available. With the lower withdrawals of 2023/24, 66,25 TWh would remain. The historical paths show a range of outcomes, not probabilities.

Our calculation from daily GIE reports. Net = gross withdrawals minus injections. “Additional cover required” denotes the requirement assuming complete mathematical use of the opening stock; any desired reserve buffer would be extra. 2016/17, 2017/18 and 2020/21: entirely estimated historical series. 2023/24: 183 rather than 182 reference days. The lowest interim balance can fall below the closing balance. No network or price forecasting model.

The unrounded calculation for 2025/26 is: 143,4933 − 137,50965 = 5,98365 TWh, around 2,4 percent of today’s working gas capacity. Later injections lift the closing balance above the temporary minimum. 2024/25: 170,13 TWh net withdrawals. 2020/21: 23,37 TWh additional requirement at the season’s end, temporarily 25,82 TWh; 2017/18 temporarily 29,72 TWh; 2016/17 temporarily 28,42 TWh. The closing balance and the most critical point are therefore shown separately.

Reported net withdrawals are not Germany’s total gas consumption: ongoing deliveries meet a large part of demand directly. The historical conditions are already included in the historical net withdrawals. Additional injection must therefore not be credited a second time. An improvement of 10 TWh in the winter balance raises the calculated remainder by 10 TWh; higher demand or lower net inflows reduce it accordingly.

Limits of the calculation: in winters 2025/26 and 2024/25, 29 and 22 daily reports respectively are estimated; the available data for 2016/17, 2017/18 and 2020/21 are entirely estimated and subsequently updated. Stock changes and reported flows do not match exactly because of revisions and capacity changes. The calculation therefore uses flow totals. A positive seasonal remainder does not demonstrate sufficient daily withdrawal or network capacity. It cannot establish a specific shutdown date.

The 2023/24 reference includes its extra leap day. Every 10 TWh improvement in the balance raises the calculated remainder by 10 TWh. Any desired reserve buffer would be additional to the extra requirement shown.

A mild winter is no guarantee against cold spells

Even a winter that is mild on average can have critically cold weeks. The DWD describes 2017/18 as a mild winter with freezing temperatures in late February. Replaying its net withdrawals would temporarily require 29,72 TWh of additional cover. Weather was not the only cause. DWD: winter 2017/18.

Northern Europe in 2023/24 provides a documented forecast error. The ECMWF seasonal forecast expected warm anomalies in the area studied, but cold conditions dominated. A research study found three of 31 Fennoscandinavian periods that were colder than every ensemble run: 9,7 percent of this specific sample. This is not an error probability for Germany. ECMWF; FMI research paper.

We do not assign an unverified probability to Germany’s winter 2026/27. At the editorial cut-off on 3 October, we could not fully verify a current numerical German forecast; the regular ECMWF October run is only published on 5 October. Weather probabilities are not probabilities of a gas shortfall. A cold stress case therefore remains relevant to planning.

A clearly documented forecast error: Northern Europe, 2023/24. The ECMWF seasonal forecast of 1 October 2023 expected a positive temperature anomaly of over one degree for November to February in the studied area of Finland/northeastern Sweden. Cold anomalies dominated instead. ECMWF documents the deviation itself; shorter-lead forecasts captured some of the cold developments better. ECMWF assessment, April 2024.

How often was cold underestimated this severely? A study by the Finnish Meteorological Institute found three of 31 October–January periods in the studied ECMWF series for Fennoscandia when the observed temperature was below every ensemble run: 2002/03, 2010/11 and 2023/24. That is 9,7 percent of this specific sample. At the same time, 16 of the 31 observations fell within the central 50 percent of the model distribution — roughly as often as expected. FMI research paper, first published May 2024.

For the individual period October 2023 to January 2024, only three of 552 runs from eight forecasting systems were at least as cold as the subsequent observation: around 0,5 percent. This 0,5 percent is the model share for one case, not an error rate. Nor do three extreme deviations in 31 periods establish a general error probability for German winters. Region, lead time, ensemble size and the mix of historical hindcasts and contemporary real-time forecasts limit the comparison. Study dataset and methodology.

Forecasts are therefore not generally worthless: ECMWF rates its temperature forecast for Western and Central Europe in 2022/23 as good. Its current report from September 2026 nevertheless describes continental cold anomalies as a continuing challenge. ECMWF: winter 2022/23; ECMWF: current performance assessment. We could not verify a reliable blanket rate of “weather services are wrong X percent of the time” for Germany at this lead time.

What is supported for 2026/27? On 11.09.2026, the UK weather service described increased chances of a warm and wet British autumn/early winter. That is not a forecast for Germany’s December-to-February winter. We could not fully check a current numerical probability for Germany in the original DWD release, so no figure is used. The regular ECMWF October run is only published on 5 October, after our editorial cut-off. Met Office, September 2026; DWD climate forecasts; ECMWF publication schedule.

Where Germany gets its gas — and where dependencies remain

The latest full-year balance shows the structure: in 2025, 44 percent of German gas imports came from Norway, 24 percent via the Netherlands and 21 percent via Belgium. German LNG terminals contributed 10,3 percent. These are rounded shares of gross imports in 2025, not current daily shares for 2026. Bundesnetzagentur: 2025 annual review.

The country of entry does not always reveal the country of production. Some gas arriving via the Netherlands and Belgium is already regasified LNG. According to the current Bundesnetzagentur FAQ, Germany’s direct LNG imports are mainly from the USA; gas from the Persian Gulf plays no significant direct role. Bundesnetzagentur: supply routes.

A large supply share is not a reserve that can be called up at short notice. Imports from Norway and neighbouring countries already meet ongoing demand. Replacement requires additional available volumes and transport options. In the week from 24 to 30 September 2026, 15,1 TWh arrived by pipeline and 2,2 TWh via German LNG terminals. These actual deliveries are not an unbooked additional offer. Current weekly flows.

Our concern is several pressures occurring together: less gas in storage than a year ago, disrupted LNG deliveries, competition for replacement supply and cold weeks. Additional procurement could then become harder and more expensive at the same time. For an unhedged buyer, it is not enough that gas exists somewhere: they need a deliverable offer for their volume and date.

Supply routes and possible disruption mechanisms, not a report of simultaneous current outages. The Bundesnetzagentur rates German supply as stable at the editorial cut-off. LNG is liquefied natural gas; after regasification, it flows through the network as gas.

The Bundesnetzagentur points to additional LNG import options and currently rates supply as stable. A lower storage level alone therefore does not prove an overall more dangerous supply situation than in 2022. Our warning describes a conditional risk from several simultaneous pressures; it does not establish a probability of occurrence. Bundesnetzagentur: storage and the new import structure.

The 2025 annual balance reports 1.031 TWh of gross imports, 221 TWh of exports and 34 TWh of domestic production. Import shares are therefore not shares of German final consumption. Border flows also do not provide a complete origin record for individual gas volumes. Annual balance.

Predictable pipeline inflows and additional LNG infrastructure can cushion risks. They do not imply unlimited additional available volumes or a guaranteed price. National storage and import data do not show how much an individual buyer has already secured.

Storm risk has historical evidence: in July 2024, an LNG export plant in Texas halted operations as a precaution before Hurricane Beryl; all three liquefaction trains were back in service on 28 July. This illustrates a possible supply-chain interruption, not a current outage in October 2026. EIA, September 2024. The infrastructure risks in the other cards are checkpoints derived from the supply routes described, not outage forecasts.

Qatar and Hormuz: little direct supply, global price risk

Around 19 percent of global LNG trade passed through the Strait of Hormuz in 2025. This mainly refers to LNG from Qatar and the United Arab Emirates — not 19 percent of German gas imports. Almost 90 percent of the LNG passing through Hormuz went to Asia. If cargoes are lost, their buyers also seek replacements in Europe’s procurement markets. IEA: reference bases and trade routes.

Agreements with Gulf states matter — and so does their delivery status. During Olaf Scholz’s term, companies signed long-term contracts in 2022 for up to two million tonnes of Qatari LNG annually. The start originally planned for 2026 is now expected in 2027 according to the German government, because commissioning of the Brunsbüttel terminal was delayed. This contracted volume is therefore not evidence of deliveries already running in winter 2026/27. Contract announcement, 2022; Corrective government footnote, 2026.

An announcement of 28.09.2026 confirms extended outages of Qatari deliveries to Italy until early December: 35 cargoes totalling around 4,6 billion cubic metres. The importer also reports replacement procurement and its ability to fulfil customer contracts. This documents supply disruptions and replacement purchases; it does not establish a German supply gap. Primary announcement.

Price competition is documented. From March to June 2026, an Asian price premium attracted flexible LNG cargoes from Europe to Asia. If disruptions persist and winter demand rises, European buyers must compete for limited additional volumes. The smaller storage buffer leaves less time to wait out that competition. IEA, July 2026.

During Scholz’s visit to the United Arab Emirates in September 2022, an initial LNG cargo of 137.000 cubic metres was agreed; the importer reported its arrival in Brunsbüttel on 15.02.2023. There was also a memorandum of understanding for multi-year deliveries. Such an understanding does not establish a specific annual volume deliverable today. Agreement, 2022; Arrival, 2023. Cubic metres of LNG must not be equated with cubic metres of gaseous natural gas.

The price channel has already been observed: the IEA describes an Asian price premium from March to June 2026 that diverted flexible LNG cargoes from Europe to Asia. In the IEA’s dollar-based analysis, the European TTF price in the second quarter was on average 32 percent above the previous year. IEA, July 2026: observed volumes and prices. This does not establish either a complete blockade or a complete reopening of the strait on 3 October.

The Qatar contracts signed in 2022 provide for at least 15 years of supply. Political support, a commercial contract, terminal readiness and an actual cargo arrival are different steps. The current delay to the start of the German contract is attributed to the terminal in the government footnote; it must not be attributed to the Middle East conflict without further evidence. Continued disruption to production or sea routes could also make future deliveries more difficult.

26,63 TWh: replacement purchases need time and available volumes

The 2024/25 reference case requires 26,63 TWh more than was stored on 30 September. This additional cover would be needed across the winter, in each case before the threatened shortfall. Ongoing deliveries are already included in the historical net withdrawals. The balance therefore has to improve: more net inflows or less consumption. Storage data do not show which additional volumes are already secured by contract.

26,63 TWh is equivalent to around 12,1 weeks of Germany’s entire LNG inflow from 24.–30. September. Deliveries then amounted to 2,2 TWh per week. These deliveries already serve demand; they are not a free additional offer. The comparison shows scale, not a delivery deadline. Bundesnetzagentur: weekly flows.

According to the transmission system operators, alternative LNG deliveries require two to six weeks’ lead time, depending on the route. Cargo, landing capacity and onward transport must also be available. Terminal utilisation of 48,8 percent in that September week does not prove that a cargo can be purchased immediately. FNB Gas; Bundesnetzagentur.

Gas buyers and suppliers must procure these volumes, not individual households in the LNG market. Possible additional costs reach customers later, depending on their contract. Households face price risk when prices are reset or contracts change. Property managers must review heat budgets, instalments and possible balancing payments when contracted costs change. Businesses must finance open volumes or adjust consumption. If volumes do not arrive in time, a price risk becomes a volume problem. This does not imply a blanket reduction for every household.

Using the unrounded model requirement of 26,63417 TWh: spread over 182 winter days, around 146,34 GWh of additional cover would be needed each day. If it all came additionally through German LNG terminals, that would mean around 46,56 percent more than the latest measured weekly inflow, sustained across the entire winter. Within a hypothetical 30-day period, the requirement would be around 887,81 GWh a day. The later the response starts, the greater the daily additional capacity required. This is neither a predicted shortage duration nor evidence of freely available supply volumes.

Possible routes include additional LNG cargoes and higher net inflows through the European network. LNG can arrive at German terminals or in neighbouring countries; Norway, the Netherlands and Belgium are important pipeline routes. Additional gas requires available volumes, secured delivery and transport to the point of consumption together. The Bundesnetzagentur stresses LNG’s lead time compared with stored gas, while also considering sufficient gas available for procurement in the market. Bundesnetzagentur: supply routes and lead times.

LNG cargoes already at sea may be diverted more quickly in some circumstances. The cited two to six weeks guarantee neither a particular volume nor a delivery date. The public sources examined do not specify a binding, immediately available additional German volume of 26,63 TWh. Equally, there is no evidence that it cannot be procured. The FNB statement is dated 27.05.2026.

Six low storage reports deserve attention

The six lowest current GIE reporting units were below 20 percent on 30.09.2026. A reporting series can represent part of a capacity or a storage product; it does not always correspond to a whole site. GIE AGSI+.

Size matters too. The separately reported series for Rehden and Breitbrunn together contain only 5,70 TWh against 47,16 TWh of working gas capacity. The unfilled 41,455 TWh represents 39,85 percent of the 104,033 TWh of empty working gas space in the German aggregate. This distance from 100 percent is neither an established winter requirement nor a gap against a statutory target.

Frankenthal reports zero percent. For 2026/27, the operator shows zero of 100 fixed packages booked and no operational shutdowns when checked on 3 October. This describes the booking situation, not the reasons behind it. Operator information.

Rehden has a low fill level but increased in September: up 1,46 TWh, from 8,54 to 12,63 percent. The announced maintenance from 27 September to 16 October alone does not explain the low stock over preceding months. Decommissioning applications for Breitbrunn and Wolfersberg concern spring 2027; they do not establish a closure today. Site evidence and definitions are set out below.

Low stocks can also limit daily delivery capacity. Storage characteristics, pressure, maintenance and the network connection determine how much gas actually leaves. Three low series belong to sites in Upper Bavaria. This does not, however, establish a calculable local shutdown risk: gas is not automatically reserved for the nearest town. The facility comparison and European overview add context.

Unfilled = reported working gas capacity minus stock. Not a statutory target gap. Rehden here refers to its separately reported share; storage zones and other site shares are not assigned. Nüttermoor H-2 is only an H-gas subseries for the site. Source: GIE AGSI+, saved retrieval from 02.10.2026; calculation by OPTUM.

Rehden: a low reported share, not an unchecked whole-site figure. The separately listed series reports 12,63 percent and 4,51 TWh. Since April, SEFE has shown around 35,74 TWh of working gas capacity for it. Storage Zone North additionally combines technical options at Rehden and Jemgum; its separately reported 7,41 TWh cannot simply be allocated to either site without a physical attribution. SEFE: Rehden capacities and Storage Zone North. SEFE documents the product structure in its 2024 management report, Storage Zone North section.

The reported Rehden stock rose by 1,46 TWh from 31 August, from 8,54 to 12,63 percent with unchanged reported capacity. This is a traceable increase, still at a low level. SEFE announced scheduled maintenance from 27 September to 16 October. These time-limited works alone do not explain the low stock over preceding months; the announcement also does not quantify today’s withdrawal limit. SEFE announcement, 24.09.2026.

Breitbrunn and Wolfersberg: separate historical problems from future decisions. For Breitbrunn, Uniper documented unfavourable marketing conditions and a missed filling trajectory in storage year 2025/26. This establishes an earlier conflict between economics and filling, but not automatically the cause of the September 2026 stock. The proposed decommissioning on 31 March 2027 is likewise not evidence of a closure today. For Wolfersberg, the operator cites a decommissioning application for 1 April 2027; we do not claim approval or completed decommissioning. Uniper: filling trajectory, 2025, Breitbrunn decommissioning application and bayernugs: Wolfersberg.

What are the technical implications? Less working gas means less stored buffer in that unit. Falling storage pressure can also reduce daily withdrawal capacity. Porous storage typically serves seasonal volumes, while caverns generally respond faster. Assessing how much a specific facility can actually deliver requires its performance curve, available equipment, maintenance, gas quality and transport capacity. Declared maximum capacity is not capacity available at every moment. Even 0 percent working gas does not mean that the cushion gas required for operation has disappeared. Uniper: storage technology, EWE: fill-dependent performance curves and Bundesnetzagentur: working gas and cushion gas.

Where are the low reporting series? Three belong to sites in Upper Bavaria: Wolfersberg in Oberpframmern, Breitbrunn/Eggstätt near Gstadt am Chiemsee and Inzenham-West in Schechen near Rosenheim. Rehden and Nüttermoor are in Lower Saxony, Frankenthal in Rhineland-Palatinate. Nüttermoor H-2 denotes a separately reported H-gas unit, not a hydrogen store or the whole site. The locations are confirmed by site information for Breitbrunn and Inzenham, Wolfersberg’s operator and the Nüttermoor site overview.

What does this mean for the region? Low storage nearby warrants a closer look at network connections and ongoing inflows. Gas in a store is not automatically reserved for the nearest town: Rehden connects to several transmission networks, and Nüttermoor H-2 has access to the THE and TTF market areas. Geographic clustering alone therefore does not establish a calculable regional supply gap or shutdown risk. Likewise, a good EU average cannot replace a local supply balance. The Bundesnetzagentur continued to rate supply as stable in its report of 2 October. Rehden’s network connection and Bundesnetzagentur situation report. The current comparison of individual storage reports and open European overview should be read together.

Legal timing: the general requirement for 1 November is 80 percent per facility; six named facilities, including Frankenthal and Rehden, have a 45 percent requirement. A September figure below this does not, by itself, prove a breach of the November deadline. Reporting and precautionary mechanisms can nevertheless apply earlier. Virtual zones cannot readily be treated as individual legal facilities. GasSpFüllstV and § 35b EnWG.

What we do not know: stocks do not reveal complete booking motives, ownership positions or ongoing technical availability. Breitbrunn has no complete historical series through September 2026; the confirmed reference-date value is separate. A September 2026 trend cannot be calculated from it. Capacity changes also complicate some year-on-year comparisons. We therefore do not publish guessed regional shutdown dates or days of supply.

Gas volume is not the same as delivery capacity

A high price creates neither an immediate LNG cargo nor additional pipeline capacity. Gas must arrive in time, be regasified and transported onwards. Storage bridges delays; daily capacity can decline as fill levels fall. The ENTSOG methodology accounts for withdrawal curves and network limits. ENTSOG, pages 47–49.

Our volume calculation does not test this peak capacity. A positive winter closing balance therefore cannot rule out shortages on individual cold days. Regional assessments require actual flows, usable transport capacity and demand peaks.

A small, nearly full store is therefore not automatically safer than a large, half-full one, and gas available across Europe is not immediately usable everywhere. The European overview with percentages and TWh shows reported reserves. A reliable local shortage calculation would also require actual gas flows, usable capacities, demand peaks and the duration of any outage.

Gas was available in 2022. Its price still multiplied.

From 50,65 to 293,73 €/MWh for the same delivery year. Gas for 2023 at the German THE trading hub cost around 5,8 times its average price from 1 January to 23 February 2022 on 25 August 2022 at 18:45. That is a wholesale increase from 5,065 to 29,373 ct/kWh — 24,308 ct/kWh more. The Bundesnetzagentur documents both figures in the same situation report. Price table on page 7.

Storage was far from empty. The situation report of 26 August 2022 showed a fill level of 81,28 percent, referring to the storage date of 24 August, and supply that was stable at that moment. The authority also warned of sharply rising customer prices. Nord Stream was delivering only around one fifth of its maximum capacity. Bundesnetzagentur: contemporary situation report.

Having gas does not protect against an extreme price for the next volume. Real supply cuts, concern about the coming winter and additional procurement pressure came together. Buyers with volumes secured cheaply were in a different position from those needing fresh purchases at short notice. For parts of industry, energy costs meant production cuts; high collateral requirements also put suppliers under liquidity pressure. IEA; ECB.

What this meant for buyers: the same gas volume cost almost six times as much at the later wholesale price. For three million kWh, the comparison produces an additional 729.240 euros. The graphic shows the calculation for three consumption volumes.

Source: BNetzA situation report of 26.08.2022, page 7. An average over a period is compared with a later snapshot for the same delivery year. Wholesale, excluding retail components; not a forecast. Ratio 293,73 ÷ 50,65 = around 5,8; increase around 480 %.

339 €/MWh — equivalent to 33,9 ct/kWh of gas at wholesale. The ECB reports this record TTF level for late August 2022; it was around 18 times the level at the beginning of 2021. The low starting price helps explain the size of that multiple. This was a futures-market peak, not an annual average or a household tariff. ECB: price developments in 2021/22.

What 2022 demonstrates — and what it does not: according to the ECB, real Russian supply cuts were the main trigger of the energy shock. These sources would not support saying “fear was the biggest driver.” They do document strong reactions to uncertainty and liquidity pressure from higher collateral requirements in futures trading. ECB, November 2022.

In August 2022, ESMA observed, among other developments, electricity companies rapidly building buying positions. Required collateral was nevertheless provided; the study does not document a general collapse in market liquidity. Panic is therefore not a blanket explanation for every price movement. ESMA: August 2022.

The market does not wait until gas runs short. According to the IEA, European gas prices rose by 50 percent from the previous day on 24 February 2022. Flows through Ukraine into Slovakia were not yet affected that day. This shows how quickly fear of further disruptions can move prices. IEA: situation on 24.02.2022.

2026 starts from a different position. Import routes, demand, infrastructure and political risks have changed. The delivery year and starting price of a specific offer also differ. The historical factor of 18 is therefore not a forecast for a current contract. What matters is the absolute cost increase on the actual unhedged volume.

The same increase, a different multiple: with an assumed starting price of 5 ct/kWh, another 10 ct/kWh raises it to 15 ct/kWh — three times the original price. With an assumed starting price of 10 ct/kWh, it rises to 20 ct/kWh — twice the original price. Both cases increase procurement costs by 10 ct/kWh. The starting prices are calculation assumptions, not current market offers.

Storage need not be empty for a price shock

Fear can bring demand forward and intensify a rally. When many market participants buy additional physical volumes at the same time, procurement pressure rises. Futures hedging can also move prices. Newly traded or repriced gas from existing stocks can become more expensive too. Existing fixed-price contracts do not automatically increase as a result. Our concern focuses on open volumes and expiring price arrangements.

What does that mean for heat supply? Whether one site or several properties, the calculation depends on the gas volume. With three million kWh of gas used for heat, an assumed increase of 10 ct/kWh already costs 300.000 euros extra if the full volume has to be purchased at the higher price. A tenth of a cent price advantage on the same volume saves 3.000 euros. Both deserve attention — but they are very different orders of magnitude.

Even at 500.000 kWh, an extra 1 ct/kWh in procurement costs would mean 5.000 euros. For property managers with 20 to 30 million kWh of affected gas volume, the figure would be 200.000 to 300.000 euros; at an assumed 10 ct/kWh, two to three million euros. The whole annual volume is affected only if it is all purchased at the higher price. Allocation to individual buildings, owners and tenants depends on contracts, consumption and billing.

Credible additional deliveries, mild weather or lower demand can ease expectations and prices. A stock reduction of 25 percent does not imply a specific percentage price increase. Costs at an individual site depend on building size, use, weather and efficiency.

A gas shock can reach electricity and heating bills

Gas-fired plants can set the wholesale electricity price. When their generation is needed to meet demand, higher gas costs can drive the price. Wind, solar, other plants, storage and imports change this effect in each market period. The European Commission describes a weaker but continuing link. European Commission.

The cost lever: ten euros more per MWh of gas means, in the model, 20 euros more in fuel costs per MWh of electricity — two cents per kWh. The assumption is two MWh of gas per MWh of electricity on the same energy basis; CO₂ and other costs remain constant. ECB, 24.09.2026.

The contract determines the individual bill. Fixed prices do not automatically move with the exchange; open volumes, new contracts and renewals face repricing risk. Network charges and taxes also play a part. District heating depends on the local generation mix and price formula. Price components; § 24 AVBFernwärmeV.

Goods and services can also become more expensive if energy, transport and input costs remain higher. Competition, contracts and margins determine pass-through. Our gas balance implies neither a fixed retail surcharge nor a blackout.

Calculation examples, not a forecast of a gas price increase. CO₂ and other costs constant; no extrapolation to the annual wholesale price or a retail tariff. Price effects depend on the generation that actually sets the price.

20.01.2025 illustrates how tight individual periods can become: from 17 to 18, the German day-ahead price reached 583,40 euros per MWh. Expected consumption was 69,7 GWh, with only 3,4 GWh of wind and solar generation. This demonstrates a price spike with high remaining demand for other generation, not isolated proof of a gas-price effect. SMARD: 2025 annual assessment.

The plant’s additional cost affects the market price accordingly only if gas is actually price-setting and offers pass on those costs. The table is not an extrapolation to the annual wholesale price or a retail tariff.

Diesel +50 percent: cost chains are already under pressure

In September 2026, E10 cost 36,0 percent and diesel 50,3 percent more than a year earlier. This follows from the full ADAC monthly averages in the table. The fuel-tax relief from October is not yet included. ADAC 2025; ADAC 2026; Tax relief.

Oil and gas are both exposed to disruptions in international supply routes. The German government attributes the latest fuel cost pressure to the Middle East conflict; the IEA documents a temporary loss of almost one fifth of global LNG supply in March 2026. This demonstrates a shared shock channel, not a fixed price relationship. IEA, review of March.

Refineries, distribution and taxes also affect fuel prices. Businesses can nevertheless face a combined increase in transport, gas, electricity and input costs. A hedged gas price does not protect the entire cost chain.

Our percentage calculation from the ADAC monthly averages. These are not daily prices for 03.10.2026 or a predicted increase in gas or electricity tariffs. Tax changes from October are not yet included in these September figures.

The comparison describes September’s cost situation. Prices from October must account for the new tax relief: the Bundestag and Bundesrat approved a reduction effective 1 October. A subsequent fall at the pump would therefore not, by itself, prove that global supply conditions had eased. German government: tax relief from October.

Production can stop while gas is still flowing

Available does not mean affordable. If an order does not cover its additional energy costs, shifts and production come under pressure. The IEA estimates a 15 percent decline in industrial gas demand in OECD Europe from January to August 2022, associated with production cuts. IEA: the 2022 crisis.

Even a hedged customer can lose a supplier. Under stress, delivery dates shift, inputs go missing or costs rise. For your own operation, minimum consumption, safe shutdown time and restart effort therefore matter. The reduction that is technically possible must be checked for each plant.

When procurement and savings are not enough

If market measures remain insufficient and the German government declares the emergency stage, the Bundesnetzagentur takes over national gas allocation. It can limit gas consumption and order shutdowns if necessary. There is no fixed shutdown list by sector. Bundesnetzagentur.

Households and other protected customers have priority. § 53a EnWG includes standard load profiles, essential social services and certain heat supply cases, among others. This protection is not an unlimited physical guarantee, but today’s reserve does not establish an imminent household shutdown. § 53a EnWG.

There is no predetermined shutdown list by sector. Decisions depend on gas flows, network conditions and the consequences of a reduction, among other factors. Even a business considered important cannot obtain statutory protected status by application. Bundesnetzagentur: case-by-case decisions and protected customers.

Protection extends beyond private households: § 53a EnWG includes customers with standard load profiles, essential social services and certain heat supply cases. District heating plants must meet conditions, and protection covers the necessary share of heat deliveries. The specific connection and use must therefore be checked. § 53a Energy Industry Act.

Priority supply is still not an unlimited physical guarantee. If even essential demand could not be fully met despite all available measures, that would be the most severe escalation of this scenario. It cannot be inferred from today’s storage level. Nor does the emergency stage automatically end trading: according to the shared understanding of the Bundesnetzagentur, THE and EEX, the gas spot market should generally remain open. Bundesnetzagentur: trading at the virtual trading point in the emergency stage.

Contract expiry is a deadline. Not a buying strategy.

We are already receiving enquiries for gas and electricity in 2027 where buyers negotiate over 0,1 ct/kWh. That matters at large volumes. Equally relevant is the question: how robust are the price and delivery promise if the market comes under stress again? The last decimal place cannot replace a review of contract terms and the supplier’s financial capability.

A strikingly cheap offer deserves questions, not a premature verdict. Earlier procurement, lower costs or different services can explain price differences. A 2027 offer must be compared over the same delivery period, similar volumes and load profiles, and the same scope of guarantees. Today’s wholesale price alone proves neither inadequate hedging nor security.

Choice can disappear in a crisis too. In 2021/22, several suppliers suspended new customer intake. An announcement of 10 March 2022 explicitly cites rejected procurement enquiries and temporarily unavailable additional purchasing options — while supply to existing customers was secured. Buyers seeking replacement under time pressure can face higher prices and fewer offers at once. Case documented at the time.

Property managers are responsible for heat budgets for owners and tenants. With annual demand of 20 to 30 million kWh, just one cent of additional cost can amount to hundreds of thousands of euros. Delayed purchasing can strain liquidity planning, instalments and possible balancing payments. Businesses with one, three or five million kWh also have real money at stake.

Our recommendation: prepare procurement while room to act remains. Identify contractual and cancellation deadlines, open volumes and tolerable price risks early; review comparable offers and supply terms. Purchasing in tranches can also be considered if it suits consumption and the contract. Early planning does not guarantee the lowest price. It creates time to make a decision before the calendar forces one.

What happens if a supplier fails? Insolvency alone does not mean deliveries stop. If consumption is no longer assigned to a fulfilled supply contract, final consumers connected at low voltage or low pressure generally enter replacement supply — for no more than three months, at separate and often higher prices. Commercial customers can also qualify. Annual volume alone is not decisive. Bundesnetzagentur; § 38 EnWG.

At medium voltage or medium pressure, a locally agreed transitional supply under § 38a EnWG may exist. The connection and local rules must be checked. A safety net does not promise continuation of the previous price. Supplier failure must also be distinguished from a physical gas shortage. § 38a EnWG.

Offer comparisons should cover price guarantees and exceptions, volume flexibility, payment terms, price-change and termination clauses, and available information on creditworthiness and hedging. A low price alone does not prove an unreliable business model; a contract cannot exclude every future event either. § 41 EnWG; Bundesnetzagentur: prices and guarantees.

Further sales pauses at the time are documented for 27.10.2021 and 04.01.2022. These are reports by the individual providers about their own business, not evidence of widespread sales stops today.