A market is secure not merely when its supply is diversified, but when it can still act as one purposeful whole under pressure — when the buying side keeps pace with the supply side. Europe's supply side has been rebuilt. Its buying side has not. That gap is the whole story, and it is being tested in public right now.
The natural dial is the supply mix, and on that dial the story is genuinely reassuring. Russian pipeline gas has fallen from roughly 45% of EU gas imports (2021) to about 12–13% (2025), the residual now moving only via TurkStream since Ukraine transit ended on 1 January 2025; a binding EU regulation, in force since 3 February 2026, phases out the remainder on a fixed legal timeline. Norway is now the largest single supplier (about 54% of EU pipeline gas, close to a third of all EU gas), and US LNG has nearly tripled since 2021 to 99.5 bcm — 58% of EU LNG imports in 2025. No supplier holds a majority. Gas storage sat near 50% full in early July 2026, roughly on its normal summer refill path. None of this is hollow: it is the honest, sound side of the ledger.
Our reading is that supply diversification was never the dial that would move first under stress. The dial that moves first is what happens the instant a real, uneven shock lands: does the Union meet it with one buyer and one voice, or with capitals bidding against each other and against Asia for the same marginal cargo. That is not a hypothetical test. Since 28 February 2026 it has been running in public.
The Strait of Hormuz has been effectively closed to normal shipping since the US–Israel–Iran war opened on 28 February 2026. A mid-June ceasefire/MOU did not restore normal flow cleanly: Iran's military re-declared the strait closed on 20 June, contradicted hours later by its own foreign ministry, and the reopening has stayed contested and partial rather than resolving on the MOU date. On IMF PortWatch's tracked count — the closest to an authoritative daily figure — transit ran near 27 vessels a day in late June against a pre-crisis baseline of roughly 90–95, about a third of normal. A missile strike on Qatar's Ras Laffan complex, the world's largest LNG facility, knocked out an estimated 17% of Qatar's export capacity for as long as five years, forcing QatarEnergy to declare force majeure on long-term contracts including Italy, Belgium, Poland (Orlen/PGNiG, via the Świnoujście terminal), South Korea and China — a disruption still open in its fourth month as this is written, and one that removed close to a fifth of global LNG supply from the market. European gas prices moved sharply in response: an early-March data point put the spike at roughly 76% within the first week; later reporting shows both higher and lower readings at different points across a still-unresolved, four-month-and-counting crisis, so 76% should be read as one early observation, not a settled peak.
Exposure is sharply uneven, and this is where the reading turns structural rather than physical: reported figures put Italy's 2025 LNG imports from Qatar at around a third (with Qatar cited at roughly 30% of Italy's total gas), Poland in the mid-twenties percent, and Belgium in the mid-teens (a smaller share of the Zeebrugge terminal specifically) — these are the order of magnitude in the reporting available, not independently re-derived figures, and should be treated as approximate. The response that followed the shock is the evidence this report leans on. A UK–France-led naval coalition grew over the crisis — from an early diplomatic summit of dozens of countries to a formal 12 May 2026 joint statement of political support from 38 nations, with heavy NATO participation — to help reopen the strait: a security response, not a buying one. The IEA coordinated a 400-million-barrel strategic reserve release, reported as the largest joint release on record — but of oil, a different commodity and a different institution. For gas specifically, reporting around the crisis described EU industrial demand-management coordination as still developing while member states individually assessed curtailment. The Commission's operational message was to call on states to coordinate storage-filling ahead of winter 2026–27; the operational response visible in reporting was national and bilateral — Italy's prime minister flew to Algeria to replace her own country's shortfall.
On AggregateEU specifically, the EU's one demand-aggregation platform: we did not find public evidence that it shaped allocation or price during this shock, and we could not confirm from public reporting whether a round was even convened in the crisis window — that is a real gap in this report's evidence, not a settled finding, and it is narrower than it may first read. What is on firmer ground is the platform's design: 189 registered companies and roughly 100 bcm matched to demand across seven rounds since 2023 are real numbers, the Commission has called its aggregation rounds a “very good result,” and independent modelling has separately flagged the platform's non-binding, non-commitment structure as producing overbidding and unrealised matches (participants themselves call it the “Tinder of the gas markets”). Put together: the design was already known to be non-binding before this shock: what this report can defensibly claim is that nothing in the public record shows that design being overridden or supplemented under real pressure — not that the shock proved it useless.
Azerbaijan is the second, independent data point, and it triangulates rather than merely illustrates. The July 2022 MoU pledged Southern Gas Corridor expansion to >20 bcm/yr to the EU by 2027; actual deliveries have run 11.4 → 11.8 → ~12.9 → 12.8 bcm (2022–2025) — roughly 60–65% of the pledge, and down 1% in 2025, the very year security of supply mattered most.
The European Parliament's response has been anything but quiet. In October 2023, following Azerbaijan's military takeover of Nagorno-Karabakh and the exodus of roughly 100,000 ethnic Armenians, Parliament explicitly called on the Commission to suspend the 2022 MoU itself. It repeated condemnation in October 2024 (453–31–89) and December 2025, calling for EU Global Human Rights Sanctions on named officials and a freeze on military-and-security exports. A further resolution at the end of April 2026 criticised the treatment of Armenian detainees and the destruction of Armenian heritage sites. On 1 May 2026, Azerbaijan's parliament suspended institutional cooperation with the European Parliament specifically (in response to an EP resolution adopted 30 April 2026), including its Euronest channel — and, in the same reporting, explicitly left the relationship with the European Commission untouched, which continues, in the words of one contemporaneous analysis, to be “underpinned by Europe's demand for non-Russian gas.” That same analysis concluded Parliament's position was “likely to remain largely declaratory, with limited prospects of being translated into concrete EU policy.” Days apart in the same window, Italy's prime minister visited Baku on 4 May 2026 calling Azerbaijani gas and oil (reported at roughly 16% and 17% of Italy's imports respectively) “crucial,” and discussed expanding TAP toward the full 20 bcm. On 12 May, Azerbaijan's state oil company SOCAR finalised its acquisition of 99.82% of Italiana Petroli (two refineries, more than 4,600 filling stations) — a deal originally signed in September 2025 and, notably, actively cleared by the European Commission on 20 February 2026 under a simplified merger-review procedure finding no competition concerns. The Commission was not merely a passive non-objector here; it was a participating regulator in exactly the kind of hydrocarbon-sector deepening the Parliament's resolutions were condemning in the same period.
Read descriptively, not morally (a national government securing supply, a Parliament passing resolutions, and a Commission clearing a merger under its ordinary competition rules are all structurally ordinary acts): the Commission's channel and the Parliament's channel are simply not the same channel, and Azerbaijan's own conduct — formally severing the non-binding one while continuing to deal with the binding one — is consistent with the interpretation that Baku itself weights the Commission relationship as materially more consequential than the parliamentary one. That is a real, checkable data point about how a counterparty reads EU institutions, though it is one actor's revealed judgment, not an independent proof of which channel objectively binds. The gap between the two channels is not hidden; it sits in the treaties, and an outside party has visibly priced it into its own diplomacy.
Both cases above are independent surface expressions of one underlying structural gap, and naming that gap precisely — not asserting that “EU energy policy is incoherent” — is the actual claim on the table. The EU built real supply-side diversification (a portfolio problem, solved) but left demand-side aggregation and external conditionality-enforcement thin and voluntary (a collective-action problem, largely unsolved). Diversifying suppliers lowers the odds that any one of them holds leverage over the whole; it does not, by itself, make 27 buyers act as one buyer, or make a Parliament resolution bind a bilateral energy contract.
Why the gap exists, stated plainly: this is not primarily a policy choice someone declined to make. Energy supply and mix remain a member-state competence under the treaties (Article 194 TFEU makes security-of-supply measures subject to unanimity where a member state's energy mix is affected), and external conditionality of the kind Parliament asks for sits under the EU's common foreign and security policy, which itself requires unanimity (Article 24 TEU). AggregateEU is voluntary because a binding EU-wide purchasing mandate is not something the Commission can simply legislate into existence; Parliament's resolutions on Azerbaijan are non-binding for the same structural reason. So the more precise claim is not that the Union “never built” a layer it could otherwise have built at will — it is that the layer is kept at the member-state level by treaty design, while the Union's own public self-description (“energy union,” a single external voice) implies a binding layer that the treaties do not in fact provide. The gap is real either way; naming its source correctly changes what would fix it.
That reframing invites an obvious objection, worth answering directly rather than leaving for a reader to raise: if the gap is just the treaties, what has this report actually added beyond restating known EU architecture? Two things, neither of which is simply “the treaties say so.” First, Azerbaijan's own conduct — formally severing the non-binding parliamentary channel while nursing the binding Commission one — is evidence that a counterparty has read the architecture and priced it into its own diplomacy; that a gap exists in a treaty is one claim, that an outside actor visibly acts on it is a different, independently checkable one. Second, the claim that the same gap reproduces the same fragmentation pattern (national, bilateral, uncoordinated) across two structurally different kinds of pressure — a physical supply shock and a values-versus-conduct tension — is a generalisable, falsifiable pattern claim, not a restatement of Article 194. A single case would be an anecdote; two independent cases behaving the same way under different trigger types is what the falsifiers below are built to test.
This is a sharper claim than a generic complaint of incoherence, and a more precise one than “a layer nobody built”: it names which layer is missing and why it is structurally thin, specifies the type of condition under which the thinness becomes visible (real shock or real values-conduct tension, not calm conditions) rather than any calendar date, and is stated with falsifiers below so it can be checked and potentially overturned, not merely asserted.
We read any system on two clocks: a solvency clock (the stores — what has been built and earned) and a coherence clock (whether the system still works as one purposeful whole). They can run years apart, and the gap between them is where masked strain hides.
Supply diversification is real; storage capacity is real and roughly on its seasonal path; import infrastructure is real — Germany alone built multiple LNG regasification terminals since 2022. Nothing here would trouble anyone on its own terms, which is exactly why it proves little about the layer this report is reading.
Here the reading echoes the earlier trade report's tonne-mile paradox almost exactly, in the same plain language rather than as a newly named metric. Pre-2022 Russian pipeline gas priced roughly €11–27/MWh; 2026 TTF has averaged roughly €43–52/MWh across the year, with an early-March data point showing a further spike of roughly 76% at the start of the Hormuz shock — one observation early in a crisis that has since moved both higher and lower, not a settled peak. Some of that pipeline-to-LNG gap is liquefaction, shipping and regasification cost, not geopolitics; we have not decomposed how much of the remainder is a genuine security premium versus ordinary market volatility, and a reader should treat the gap as directionally real rather than precisely apportioned. The larger, harder-to-quantify part is what the trade report called distance substituting for trust: the system is paying more for a comparable molecule specifically to avoid a hostile counterparty and a chokepoint it no longer relies on. That is a real, defensible security purchase, not a waste — the same caution the trade report applied to “effort evaporating” applies here: a system deliberately buying supply security is choosing a different purposeful whole than one optimising for cost, and this report does not adjudicate between those two valid purposes. It only notes that the bill is being paid at the molecule level while the buying-power gap that would let the Union negotiate that bill down collectively stays thin.
Per the project's canonical three-reserves framing — nature, other people and social groupings, and the future borrowed as debt — applied here without adding a fourth: (i) cheap Russian pipeline gas (the other-people/geopolitical reserve) is closing on a fixed legal timeline — short-term contracts banned from April–June 2026, long-term from January–September 2027 — and is not being replaced at the same unit cost, as the price comparison above shows. (ii) The deferred future (the debt/subsidy reserve): publicly funded LNG terminal capacity in Germany sits substantially idle, a stranded-asset profile rather than a used one, and the Union's mandatory storage target was itself quietly relaxed from 90% to 80% for winter 2026–27 after the 2025–26 season drained reserves to their lowest level since 2018 — adjusting the target rather than meeting it is worth naming plainly. (iii) Fossil overshoot (the nature reserve): the professed direction is a secure and decarbonising energy union; the conduct record is new LNG import infrastructure and deepening hydrocarbon relationships (Azerbaijan, Gulf LNG). This report takes no position on whether that direction is right; it only notes, neutrally, that the professed direction and the allocation record diverge, which is the definition of a say-do gap rather than an endorsement of either side of the climate debate.
Phantom-anchor risk, stated as a condition, not a conclusion: the canon definition is a professed shared good whose gap from conduct widens until it implodes on itself. The evidence above shows a widening, measurable gap — pledge versus delivery on Azerbaijan, “energy union” versus national scramble on Hormuz/Qatar, “green transition” versus continued fossil build-out. It does not show implosion, and this report does not predict one — only that the gap is real, moving in one direction on current evidence, and worth watching.
Be clear about the evidence. This is a structural, directional read — the same lens used on institutions, states and the earlier trade-fragility case, applied to EU energy security, built entirely from public data (the figures above and their sources below). The canvas modelled is deliberately bounded to the energy and geopolitical layer — supply mix, routes, contracts, the Commission/Council/Parliament governance split, national bilateral deals — not the whole of EU politics or the merits of any climate policy. This report is a hand read on public data — an analyst applying the lens, not the output of a model run; no engine reading has been done for this canvas, and nothing here rests on a simulation. The Hormuz/Qatar shock is a genuinely live, still-unresolved event as of writing (July 2026); the Azerbaijan case draws on four years of public record (2022–2026). Two specific gaps flagged in the sandbox pass that preceded this report were chased down directly rather than left as design-level assumptions: whether the EU's joint-purchasing platform (AggregateEU) actually bound outcomes during the live shock (no public evidence found that it did; whether it even convened in this window could not be confirmed either way; the coordinated actions that clearly did occur — an oil reserve release, a naval coalition — sit in different institutions and a different commodity), and whether a genuine counter-case exists of the EU or a member state actually declining a deal on conditionality grounds (none was found for Azerbaijan; the nearest candidate, Germany's 2022 suspension of Nord Stream 2 certification, was aligned with rather than in tension with the diversification strategy already under way, so it does not cleanly test the claim under a real supply-security cost). A further round of external review also surfaced that the demand-aggregation/conditionality gap is substantially a treaty-competence feature (Article 194 TFEU, Article 24 TEU) rather than a discretionary policy omission; that distinction is now built into the mechanism section above rather than left implicit.
This is a directional, structural read: supply diversification succeeded and is not the exposed flank; the flank is a demand-side aggregation and conditionality-enforcement layer that is kept thin by treaty design; that same gap explains both the Hormuz/Qatar national scramble and the Azerbaijan say-do gap. It carries its own falsifiers, stated so they can be checked later:
A standing section of every report: what we looked for and did not fully find, and what this method cannot read. It is not a disclaimer; it is where the reading earns its trust.
| # | Open item |
|---|---|
| 1 | Snapshot of one live, unresolved shock plus one four-year case. This is a state read, not a validated trajectory (the Hormuz/Qatar shock had not resolved as of writing). |
| 2 | Window selection. The Hormuz/Qatar shock is an extreme, war-driven event; a milder shock might not reproduce the same national-scramble pattern — flagged above as a falsifier, not smoothed over. |
| 3 | AggregateEU's non-role is argued from its design plus the absence of any claimed success during this shock, not from a direct before/after measurement — we could not confirm from public reporting whether the platform even convened a round during the Feb–Jul 2026 window; the Commission's own claimed "very good results" refer to earlier, calmer aggregation rounds, not this shock. The claim in this report is narrowed accordingly to "no public evidence it was used," not "it was tested and failed." |
| 4 | The counter-case search returned no clean hit. No EU member or Commission decision was found declining or delaying an energy deal specifically on conditionality grounds at a real cost to supply security; Nord Stream 2's 2022 suspension is the nearest case but was aligned with, not costly to, the diversification strategy, so it does not bound the claim as a genuine counter-example. |
| 5 | Conduct over words. Where the report reads institutional intent, it is anchored in observed contracts, votes, and allocation — not stated policy. |
| 6 | Framework-derived, not engine-run. The mechanism (aggregation solved for supply, unsolved for demand/conditionality) is derived from the project's framework and an analyst's reading of public reporting; it has not been run in an engine for this canvas. |
| 7 | Neutral on decarbonisation. The say-do gap on "secure and decarbonising" is read structurally, not as a verdict that either hydrocarbons or the green transition is the correct policy. |
| 8 | The gap is substantially treaty-structural, not a discretionary omission. Energy supply (Article 194 TFEU) and external conditionality (Article 24 TEU, CFSP unanimity) are member-state-anchored by design; this report's contribution is the claim that the same gap reproduces the same fragmentation pattern across two different trigger types, not the discovery of the gap itself, which is a known feature of EU architecture. |
| 9 | Several precise-sounding figures are order-of-magnitude, not independently re-derived — national Qatar-LNG exposure shares, the naval coalition's exact headcount at any given date, and the German LNG utilisation figures (a 2025 snapshot, not re-checked for 2026) should all be read as approximate and dated, per the citations above. |
Sources. Storage & supply mix: Consilium (consilium.europa.eu) infographics and press releases on gas storage and the Russian-gas phase-out; European Commission DG ENER (energy.ec.europa.eu) on REPowerEU, AggregateEU and gas storage; Eurostat/Statista on LNG and pipeline supplier shares. Russian gas phase-out timeline: Council of the EU press releases (Oct–Dec 2025, Jan 2026); Jones Day and Harneys regulatory notes. Ukraine transit end: Reuters/AP/Carnegie Endowment/Bruegel/Wilson Center (Jan 2025). Azerbaijan MoU and deliveries: IEA policy database; Baku Research Institute; Caspian News; OilPrice.com. European Parliament resolutions and the Azerbaijan/EP suspension: europarl.europa.eu press releases and plenary texts (Oct 2023, Oct 2024, Apr 2026) and joint statement (5 May 2026); Anadolu Agency, JAM News, OC Media (May 2026); Brussels Signal (4 May 2026, "Azerbaijan cuts ties with European Parliament as Brussels' energy realism holds"). Meloni–Baku visit and SOCAR–Italiana Petroli: Bloomberg, Euronews, Caucasus Watch, MLex, IEU Monitoring, Caliber.Az (Sept 2025–Jun 2026, full deal timeline). Hormuz/Qatar crisis: Wikipedia "2026 Strait of Hormuz crisis" and "2026 Strait of Hormuz campaign"; House of Commons Library briefing; IMF PortWatch (transit counts); Hormuz Strait Monitor; CNN (traffic visualisation); UK Government joint statement on the Multinational Military Mission for the Strait of Hormuz (12 May 2026); Al Jazeera (coalition size); Finnotia and CNBC (TTF price move). Qatar force majeure by country: Al Jazeera, TRT World, Anadolu Agency (Italy, Belgium, South Korea, China); gasworld, Pravda EN, Bloomberg/Rigzone (Poland/Orlen/PGNiG). German LNG terminal utilisation: IEEFA, NaturalGasIntel, gCaptain (2025 data). AggregateEU design and criticism: European Commission Q&A; arXiv model-based analysis of overbidding and non-commitment. CSDDD/Omnibus: European Parliament Legislative Train; Clifford Chance; Rigzone. Figures are public and as-reported; periods are stated rather than smoothed; several exposure/headcount figures are order-of-magnitude approximations per the misses table above. Proof of concept — consistency ≠ validation. © The Great Homecoming Project.