The €4 billion question

Part 4 of 5 in The Broken Ballot. This article applies the series framework to a live case study and can be read as a standalone piece on climate governance and litigation risk. Read article 3 here.

 

 

Background: INEOS, Project One, and ClientEarth

INEOS is one of the world's largest private chemicals companies, founded by Sir Jim Ratcliffe. It is privately held, with revenues of approximately $50 billion, and operates across petrochemicals, oil and gas, and speciality chemicals across approximately 148 manufacturing sites worldwide. Project One is INEOS's €4 billion ethane cracker at the Port of Antwerp, Belgium, designed to crack US-imported ethane into ethylene, the primary feedstock for plastics manufacturing. Ethane is a specific component of natural gas extracted from US shale fields, distinct from LNG (liquefied natural gas). At a designed capacity of 1.45 million metric tons of ethylene per year, it would be the largest new plastics production facility in Europe.

ClientEarth is a non-profit environmental law organisation with offices across Europe and beyond, known for its strategic use of litigation to enforce environmental law. It has brought four successive legal challenges against Project One's permitting, together with 14 partner NGOs including Greenpeace Belgium, WWF Belgium, and Bond Beter Leefmilieu. The first three challenges focused on the adequacy of environmental impact assessments, resulting in the annulment of INEOS's permit in July 2023. INEOS obtained a replacement permit in January 2024 and construction resumed. The fourth challenge, brought and renewed during 2024, argues that the Flemish authorities' approval of the replacement permit was illegal under both EU and national law because INEOS failed to provide a full assessment of the project's impacts across its entire value chain, including upstream gas extraction in the US and downstream plastics disposal. A secondary strand of the fourth challenge raises procedural concerns about the rapid administrative withdrawal and re-granting of the permit by reference to new regional nitrogen rules. Simultaneously, the nitrogen decree itself faces a constitutional challenge before the Belgian Constitutional Court, which, if successful, would remove the regulatory foundation on which the replacement permit rests. As of May 2026, mechanical completion is expected by the end of 2026, with start-up targeted for early 2027, and Project One remains subject to active fourth-round litigation whose outcome is materially contested.

A €4 billion investment was approved despite unresolved regulatory challenges, incomplete value-chain emissions analysis, and no structured assessment of community or intergenerational impacts. That is the governance failure this article examines.

The argument does not depend on INEOS. The same governance failures arise wherever a board must approve a major capital commitment under conditions of contested climate science, multi-party litigation, and stakeholder welfare trade-offs. Shell's 2021 climate strategy resolutions involved precisely these tensions, as did ExxonMobil's 2021 board elections driven by climate activist investors. Boeing's safety governance failures demonstrated identical patterns of deliberative collapse under commercial pressure. Meta's content moderation decisions and OpenAI's governance crisis of 2023 represent the same structural problem applied to platform governance: consequential institutional decisions made without the deliberative architecture that their stakes required. Project One is used here because it is live, because the litigation record makes the deliberative failures unusually visible, and because the €4 billion capital commitment concentrates governance complexity into a single analysable decision. It is illustrative, not foundational. The framework applies wherever consequential institutional decisions are made without adequate deliberative process.

Why this case study

The INEOS Project One case concentrates, in a single capital vote, precisely the governance failures this series diagnoses: contested climate risk, multi-party litigation, intergenerational welfare trade-offs, and regulatory uncertainty suppressed by governance architecture not designed to accommodate them.

Stage 1: what stakeholder mapping would have revealed

An automated stakeholder mapping exercise, conducted as a precondition for the vote, would have required the board to respond to a structured matrix of affected parties that extends well beyond the shareholder register. Direct parties, shareholders, management, and Port of Antwerp, represent only the first tier. Proximate parties include the residents of Antwerp's northern districts facing nitrogen deposition and air quality impacts, and the Belgian agricultural sector facing discriminatorily tighter nitrogen thresholds under the same regulatory decree that enabled the new permit.

Legally cognisable parties, those with active or potential legal standing, include ClientEarth and its 14 NGO partners, future generations bearing intergenerational climate harm from the project's full value chain emissions, and Gulf Coast communities in the United States whose environments are affected by the upstream gas extraction that feeds the facility. Systemic parties include the EU single market, which is actively evolving its plastics regulation framework, and insurance markets that are beginning to price petrochemical stranded asset risk in ways not yet reflected in most capital allocation models.

Stage 2: the scenarios the base case was not telling

The scenario modelling stage would have surfaced three probability-weighted pathways that the management base case, permit upheld, project completes, ethylene volumes delivered, systematically underweights.

The litigation scenario, given the fourth legal challenge's novel abuse-of-process grounds and the simultaneous constitutional challenge to the nitrogen decree's regulatory foundation, carries a materially different risk profile from the prior three challenges. The AI system would have flagged that the board's probability weight for this scenario diverged from the independent analyst consensus, requiring written justification before the vote proceeded.

The transition scenario captures a dimension that conventional capital allocation models handle poorly: the possibility that the EU extended producer responsibility framework and the global plastics treaty compress the addressable market for virgin ethylene before the facility's designed capacity is commercially reached.

Stage 3: the ethical trade-offs the binary vote erased

The ethical trade-off display would have presented four normative lenses simultaneously, not to resolve the competition between them, but to make it visible.

The utilitarian analysis is genuinely contested. INEOS would present employment creation, European industrial sovereignty, and shareholder value generation. Alongside this, the system would have presented quantified health costs from nitrogen deposition on Antwerp's population, long-term ecosystem damage to the Scheldt estuary, and the welfare costs of climate change attributable to the project's full value chain emissions. Both calculations, with their methodological assumptions explicit, would have appeared side by side.

The rights-based analysis presents a particularly strong case. The Flemish authorities' withdrawal and immediate re-granting of the permit, characterised by the NGO coalition as an abuse of process, raises a procedural rights concern that is independent of the substantive environmental merits.

Stages 4–6: bias, logic gates, and the audit trail

The Stage 4 bias detection module would have flagged concentration of voting power, herding around a single proxy recommendation, and patterns of institutional voting that appeared to reflect algorithmic proxy processing rather than independent analysis.

The Stage 5 logic gates would have generated three mandatory triggers before the vote opened: an independent legal risk assessment on the fourth legal challenge; a Scope 3 emissions estimate covering the project's full value chain; and an affirmative acknowledgement from each institutional investor that the abuse-of-process grounds for the fourth challenge had been reviewed. These gates would not have prevented any investor from voting in favour. They would have ensured that such a vote constituted a deliberate, documented fiduciary decision.

The Stage 6 audit trail would have generated a governance report that is, in the current climate litigation environment, a material asset: documented evidence that each institutional investor engaged with the full complexity of the decision before casting their vote.

What the framework could not have done

The INEOS case study demonstrates the power of AI-assisted deliberation. It also illustrates its limits. The framework does not guarantee that Project One would have been voted down, nor should it. Reasonable shareholders, having engaged with the full deliberative process, might reasonably conclude that the project's economic and industrial merits outweigh its legal, regulatory, and climate risks. That is a legitimate conclusion. What the framework guarantees is that the conclusion was reached through a process that engaged with the actual complexity of the decision.

 

The final article moves from diagnosis to pathway. We address seven categories of implementation constraint, cost, legal enforceability, data governance, liability, institutional resistance, proxy advisor integration, and governance capture, and propose a three-phase implementation pathway from a responsible business pilot through to ICGN global standards. We also address what responsible governance requires beyond the legal dimension: the corporate strategy, ESG, and institutional dimensions that a purely process-focused reform agenda risks overlooking.

illuminem Voices is a democratic space presenting the thoughts and opinions of leading Sustainability & Energy Thought Leaders, their opinions do not necessarily represent those of illuminem.

Track the real-world impact behind the sustainability headlines. illuminem's Data Hub™ offers transparent performance data and climate targets of companies driving the transition.