17th May 2019
On 29 November 2018, the Court of Justice of the European Union (CJEU) published the Opinion of Advocate General Kokott concerning a preliminary reference from the Belgian Constitutional Court concerning a life-time extension of two nuclear power plant reactors without prior environmental impact assessment (EIA). The Opinion is of particular relevance because Kokott is recommending to overturn an earlier CJEU judgement that excluded life-time extensions from EIA, basing herself heavily on the Aarhus and Espoo Convention.
According to a Belgian law of 2003, Belgium’s seven existing nuclear power plants were supposed to be phased out after 40 years of operation (between 2015-2025), unless security of electricity supply would be threatened. However, with a law of 2013 Belgium postponed the end date for the “Tihange” reactor by 10 years and by law of 2015 the end-date for the “Doel 1” and “Doel 2” reactors by 10 years as well. Doel 1 had in 2015 already seized to produce energy but was to be taken back online following a permit procedure. Two Belgian NGOs challenged the 2015 law extending operation by 10 years for Doel 1 and 2 before the Constitutional Court.
The Belgian Constitutional Court has referred nine questions to the CJEU (para. 37) but the central three issues are the following:
Kokott first addresses the question whether the obligation to conduct EIA could be circumvented because the life-time extension was essentially permitted by way of a legislative act. She refers to existing case law (such as Cases C-128/09 Boxus and C-348/15 Stadt Wiener Neustadt) but adds the requirement that public participation needs to have been conducted before the relevant decision (see para. 60). This will likely rule out such a circumvention in the present case. However, the latest amendment of the EIA Directive removed the provision exempting EIA for activities approved by legislative acts (article 1(4)), so this case law will not apply any longer for projects with EIAs initiated after 16 May 2017.
Kokott then turns to the question whether there has been a failure to carry out an EIA. She addresses head on the crucial precedent which had interpreted a “project” under the EIA Directive as works or interventions involving alterations to the physical aspects of the side (Case C‑275/09 Brussels Hoofdstedelijk Gewest and Others). Kokott finds that this case law is not consistent with both the Aarhus and Espoo Convention.
For the Espoo Convention, Kokott refers to the findings of the Espoo Implementation Committee that held that the continued operation of an NPP is an activity for the purposes of the Convention (paras. 77-78), that it does not matter that the permit was unlimited in duration (paras. 79-84) and that the continued operation of 10 years also necessarily creates a significant risk of a transboundary impact as per article 3(7) of the Espoo Convention (paras. 85-59).
For the Aarhus Convention, Kokott considers that the permission to produce electricity with a nuclear reactor essentially the same as a permit for a new reactor (para. 93). She thereby goes beyond recent concerning the Dutch nuclear power plant at Borssele, which considered that a life-time extension was a reconsideration or update (under article 6(10) of the Convention) but did not consider it necessary to also consider whether a life-time extension was also effectively equivalent to a new project.
Kokott accordingly advises to interpret the EIA Directive in line with the Aarhus Convention and Espoo Convention to categorize a 10 year extension of production of electricity from an NPP as a project and thereby deviate from the exiting case law (para. 111). Alternatively, Kokott advises to give direct effect to the Aarhus and Espoo Conventions to the same effect (para. 121). As a second alternative, Kokott finds that the lifetime extension would also be an upgrade in the sense of article 1(2) of the EIA Directive (para. 125) in conjunction with point 24 of Annex I to the EIA Directive (paras. 129-34).
Kokott then holds that EIA had to take place when all options were open (Article 6(4) of the EIA Directive) which means it had to take place before adoption of the legislative act to extend the lifetime and not only at the permitting stage (paras. 138-142). She also considers that there is no possibility to derogate from the obligation to conduct a transboundary EIA under Article 7 of the Directive (paras. 145-146) and that a derogation from conducting a national EIA would need to be based on a “grave and imminent peril” (paras. 157-159 and 161-162).
As regards the Habitats Directive, Kokott first highlights that all EIA “projects” are also projects under the Habitats Directive but that the Directive goes beyond that (paras. 166-172). Based on that logic, if the CJEU changes its interpretation of “project” under EIA, obligations under the Habitats Directive will automatically apply. If the CJEU does not, she considers that the extension nonetheless gives rise to Habitats obligations because of the increased risks of impact on the protected sites (paras. 172-173). She also finds that there is no exclusion for legislative acts (para. 176).
AS regards the consequences of these violations, Kokott firstly states that based on the principle of effectiveness and equivalence, a decision should in principle be suspended or annulled for failure to comply with EIA Directive (para. 197-199) and “as a rule”, it would be appropriate to “halt the operations of a plant if its consent is revoked or suspended” based on such an infringement (para. 200). However, Kokott further considers that there could be exceptional cases in which the effect of the permit could nonetheless be maintained despite the breach, arguing by analogy to cases under the SEA Directive (Case C-41/11 Inter-Environment Wallonie and C-379/15 Association France Nature Environment, see para. 212). She emphasizes, however, that strict conditions must be fulfilled:
This is a strong Advocate General’s Opinion with many useful statements on applying the EIA Directive in light of Espoo and Aarhus obligations and on how to interpret the Habitats Directive. Most importantly, the Opinion recognizes the need to carry out EIA for life-time extensions without physical alterations to the project. This is not only important for ageing nuclear power plants, which have often been constructed in the 1970s and 1980s and are currently reaching the end of their design lifetime, it is also crucial for heavily polluting infrastructure such as coal-fired power plants. For instance, Rovinari coal-fired power plant in Romania has initially been permitted without EIA (before entry into force of the EIA Directive) and also did not undergo EIA covering the entire plant at a later stage. Recently, the authorities again approved continuous operation of the plant without an EIA. Clear direction from the CJEU could prevent such cases in the future.
Another important clarification concerns the compatibility of article 2(4) of the EIA Directive with the Aarhus and Espoo Conventions. Applying Kokott’s rationale would mean that application of this provision would be excluded for projects with significant transboundary impact and could only be applied for national EIAs where there is a risk of “grave and imminent” peril. The latter is the wording of the “necessity defence” as defined by the International Law Commission in article 25 of the Articles on the Responsibility of States for Internationally Wrongful Acts. Since a Member State would accordingly need to demonstrate that there was a grave and imminent peril to the State as such which could only be avoided by not conducting an EIA, this is an almost impossible threshold to meet. This is appropriate because there are indeed few good reasons not to assess the environmental impacts of massively polluting infrastructure.
As regards the consequences of a breach of the EIA requirements, Kokott does well in first emphasizing that there is a need to halt the project in principle. The fact that she then permits regularization at all may be criticized. However, construing regularization under the SEA logic sets a quite high threshold for justification (see 4 steps above). Most interesting is perhaps the last requirement (point (d) above): The burden is shifted explicitly on the government to justify continued operation as an “overriding public interest” and the national court is particularly required to consider “minimum supply of energy” as opposed to “security of supply”. This will require to show why stopping operation would indeed impact basic needs and not only be economically the most viable solution.
We can be expect the CJEU judgement to be published in the following months. In this case, the Advocate General’s Opinion squarely places the EU legislation in its international law context. It is due time for the CJEU to recognize that this international context has evolved and that the CJEU’s jurisprudence should do so as well. It would be absurd to maintain an interpretation of the Directive that effectively exempts the oldest and most polluting facilities indefinitely from EIA and public participation as long as the responsible authorities do not propose any physical improvements or other alterations.