Energy was the second-largest sector for arbitration at the International Chamber of Commerce last year, trailing only construction, and just over half of the 881 new cases the ICC registered in 2025 were seated somewhere in Europe. That’s not a coincidence. It’s the paper trail of a continent building interconnectors, hydrogen corridors, offshore wind farms and battery parks across borders faster than its legal protections have kept pace, and of a treaty regime that, for cross-border energy investors, has quietly stopped being reliable.
For two decades, the Energy Charter Treaty was the default shield. If a government changed the rules on a power plant or a wind farm mid-investment, Article 26 let the investor take the state to arbitration, treaty to treaty, regardless of what any underlying contract said. That shield started coming apart in 2021, when the Court of Justice of the EU ruled in Komstroy that the treaty couldn’t be used for disputes between an EU investor and an EU member state. The EU formally withdrew in June 2025; the UK left the treaty in April 2025, citing its incompatibility with net-zero commitments.
Even before the fracture, though, the treaty was never the sure thing it’s remembered as. When a German investor challenged the Czech Republic’s retroactive cut to solar feed-in tariffs under the ECT, in Antaris Solar v Czech Republic, the tribunal threw the claim out entirely in 2018, ruling that the investors had gone into an already volatile market with their eyes open and that “not every legislative change gives rise to a claim based on breach of legitimate expectation.” One arbitrator, Gary Born, Co-Chair of the International Arbitration Group at King & Spalding, dissented, arguing regulatory frameworks can create binding expectations of their own. Even at full strength, in other words, the treaty’s protection was contested and unpredictable.
What’s replaced that certainty since Komstroy is something closer to a postcode lottery. ICSID tribunals initially waved the objection away — in Infracapital v Spain, the panel called it “entirely irrelevant” to the case. But a Stockholm Chamber of Commerce tribunal took the opposite view in Green Power v Spain, accepting the objection specifically because the arbitration sat inside EU territory. By October 2024, even ICSID had shifted: two tribunals, in claims brought by Sapec and European Solar Farms against Spain, upheld the intra-EU objection for the first time. The SCC has since said publicly it will favour seating intra-EU energy disputes outside the EU altogether, purely to protect the resulting award’s enforceability. Whether an investor gets a hearing on the merits now turns less on what happened to their project and more on which arbitral institution, in which city, ends up hearing it.
The Dutch coal phase-out shows what that inconsistency costs in real money. RWE filed a €1.4 billion ICSID claim against the Netherlands in February 2021 over the law forcing its coal plants shut; Uniper filed a parallel claim weeks later. Both withdrew in 2025 — Uniper first, at the German state’s request, then RWE after Germany’s own Federal Court of Justice ruled that the treaty’s arbitration clause isn’t a valid basis for an intra-EU arbitration agreement under EU law. Yet the Netherlands had already paid Vattenfall €52.5 million to close its oldest coal plant under the same law, through a different legal route. Same policy, three utilities, three outcomes — decided by structure and timing rather than the underlying merits.
And winning an award is not the same as getting paid. Micula v Romania wasn’t an energy dispute either, but it has become the case every energy investor’s enforcement lawyer now studies. Two Swedish investors won an ICSID award of roughly €83 million against Romania in 2013 over withdrawn investment incentives; when they tried to enforce it in English courts, the European Commission stepped in, calling the payment illegal state aid and blocking it. Romania secured a stay of enforcement, and the fight ran for six years before the UK Supreme Court unanimously lifted it in February 2020, ruling that the UK’s obligations under the ICSID Convention took precedence over EU law. For any investor relying on an arbitral award against an EU government, that six-year gap between winning and collecting is the risk that matters just as much as the award itself.
That’s forced a rewrite of where protection actually has to come from: not the treaty, but the contract. Lawyers advising renewables investors, from Norton Rose Fulbright to Clifford Chance, are now pointing clients back to stabilisation clauses that lock in the regulatory terms an investment was made under, because a state changing course can no longer be assumed to trigger treaty compensation. Seat selection has stopped being an afterthought buried in a boilerplate schedule; as dispute lawyers at the Romanian firm Sora & Associates put it, the seat is “the legal home of the arbitration,” and choosing one with a genuinely reliable, enforcement-friendly track record is now a commercial decision, not a formality.
That fragmentation risk isn’t hypothetical, and wind projects have already lived through it. Braes of Doune, a 36-turbine wind farm near Stirling, split its build between a civil and electrical works contractor under an EPC agreement and a separate turbine supplier under an Interface Agreement — a structure most large wind and offshore projects still use today. When the project ran late, the arbitrator, John Uff QC, found the EPC contractor’s liquidated-damages clause unenforceable as a penalty, partly because it could have charged that contractor for delays actually caused by the turbine supplier under the other agreement entirely. Multi-contract wind projects carry exactly that risk: a single dispute — a delayed foundation, a defective cable joint — can fragment into parallel proceedings that don’t recognise each other’s findings unless the contracts are drafted to interlock from the outset.
Getting the wording right matters as much as picking the forum, and the same Braes of Doune case shows why. Its contract named Glasgow as “the seat of the arbitration,” but almost every other clause pointed to English law and the English courts. When the dispute reached the High Court, Mr Justice Akenhead ruled that Glasgow was only ever the venue, not the seat, because the parties’ substantive intent — repeated references to English law and the 1996 Arbitration Act — made England the true juridical seat. Confusing seat and venue in the drafting is a common, avoidable mistake, and nearly cost the winning side its ability to enforce the award. In Cable & Wireless v IBM, an English court went the other way and held a tiered dispute clause enforceable, precisely because it pointed to the Centre for Dispute Resolution’s published mediation procedure rather than a vague promise to talk things through; Mr Justice Colman ruled that referencing a concrete, established process gave the clause the certainty courts need to compel compliance. The lesson both cases teach is the same one: dispute resolution wording should never be dismissed as boilerplate.
Offshore wind has its own cautionary tale, and few cases illustrate how much contract wording alone can decide an outcome better than MT Højgaard v E.ON, still cited by offshore wind lawyers across Europe nearly a decade on. MT Højgaard built the foundations for the Robin Rigg wind farm exactly to the industry’s own design standard — and still lost at the UK Supreme Court, because a “fitness for purpose” warranty buried in a technical schedule set a stricter bar than the code-compliance clause sitting elsewhere in the same contract.
Storage is generating its own wave of disputes, for a simple reason: batteries don’t fit neatly into rules written for generators or consumers. In Germany, a battery storage operator argued it shouldn’t be billed the standard grid-connection subsidy charge, since storage both consumes and feeds power back; the Federal Court of Justice disagreed, reversing a lower court and ruling the charge applies regardless of the operator’s commercial strategy — a ruling German battery developers are now building into their financial models. Greece’s Council of State faced a related but different question this year, over whether stand-alone battery storage needed the same environmental permitting as the renewable projects it is typically paired with; the court upheld the streamlined regime, giving developers certainty they lacked before. Neither dispute involved a treaty or an arbitration clause, and that is rather the point: as storage scales up across Europe, it is regulators and domestic courts, not international tribunals, testing where the rules actually land, project by project and country by country.
None of this means arbitration has become less useful for Europe’s energy build-out; the ICC’s caseload says the opposite. It means the protection investors used to get from the treaty now has to be built into the contract itself: the seat chosen deliberately and correctly, the clauses drafted to survive a challenge, and every agreement on a project pointing toward the same forum rather than three different ones. Solar, wind and storage projects are each finding this out through their own disputes, and the lesson across all three is identical. The treaty isn’t gone. It’s just no longer something anyone can afford to assume — and, as Romania’s investors learned, winning it isn’t the last step either.
