Opinion 1/20 and the modernisation effort of the Energy Charter Treaty

On 16 June the Court found Belgium’s request for an Opinion pursuant Article 218 (11) TFEU on the compatibility of Article 26 of the ‘modernised’ Energy Charter Treaty (ECT) inadmissible on grounds that it did not have ‘sufficient information’ on its envisaged provisions. Article 26 is the provision that allows foreign investors to have recourse to investor-state dispute settlement (ISDS) when their investments are negatively affected by government action in breach of the substantive provisions of the ECT. While the modernisation process did not foresee any changes to Article 26 ECT and at the time of the opinion there was no public information available that parties may do so, the Court speculated that the parties may change their positions and that it therefore did not have ‘sufficient information’. No less than 12 days later, an agreement in principle was reached by the contracting parties of the ECT to amend the ECT that inter alia introduced a so-called disconnection clause (a clause that would make part of the ECT inapplicable between EU Member States). This post will offer commentary on the Court’s interpretation of the Opinion procedure (Article 218 (11) TFEU) in the wake of ongoing negotiations to ‘modernise’ the ECT and offer some thoughts on the outcome of the negotiations themselves from the perspective of climate change mitigation efforts. It will start with a brief introduction to why Belgium requested this Opinion, outline the rationale of the Opinion procedure, and subsequently discuss the ruling of the Court. It will end with a brief discussion with the current efforts to ‘modernise’ the ECT and its relationship with efforts to mitigate climate change.  

Context: Belgium’s request for an Opinion

Under Article 218 (11) TFEU a small number of applicants (Member States, the Council, the European Parliament, and the Commission) may request an Opinion of the ECJ on the compatibility of an ‘agreement envisaged’ is compatible with the EU Treaties. The procedure’s purpose is to prevent legal complications that may arise if the EU were to enter into an agreement with third parties that would not be compatible with the Treaties (see also para. 25 of Opinion 1/20). For instance, if the ECJ would find the actual ECT incompatible with the EU Treaties, it would create difficulties for the EU institutions and the Member States to abide by their international obligations (such as awards issued by investment tribunals under the ECT) given the supremacy of EU law.   

Belgium had requested this Opinion in light of a disagreement between the Member States on the scope of the Achmea judgment. That judgment had found an investment agreement between Member States containing investor-state dispute settlement (ISDS) provisions (an intra-EU agreement) to adversely affect the autonomy of EU law. A few Member States had maintained that the ECT was not affected by Achmea even though the ECT was concluded by almost all Member States, did not contain a disconnection clause, and investors and the arbitration industry were very busy with using investment arbitration under the ECT for intra-EU disputes. The ECT was after all a most unusual agreement: not only all the Member States were party except Italy, but also 21 other states (such as Afghanistan, Turkmenistan, and Japan) and remarkably also the EU itself. In fact, the ECT remains the only international agreement in force to which the EU is party that contains ISDS.

This view from Member States such as Sweden and Luxemburg hampered efforts by the EU and the other Member States to put negotiation of a disconnection clause on the table during the ‘modernisation’ effort of the ECT. That modernisation effort was launched when the Energy Charter Conference adopted in November 2018 a list of topics for the modernisation of the ECT, kickstarting the negotiations for amending the ECT. Notably absent from the list of topics of negotiation was Article 26 of the ECT (the arbitration clause), yet the topic of ‘REIO’ (Regional Economic International Organisation) was. As the decision itself clarified, that topic does not refer ‘to the definition of REIO, but to consider which ECT provisions (if any) should not apply among members of a REIO’. In other words, the Energy Charter Conference was willing to take up the issue of a disconnection clause, but the EU itself did not subsequently make any textual proposals to address the rather clear conflict between the ECT and the EU Treaties. Rather, the EU’s proposals were focussed on bringing the ECT more in line with the Commission’s current approach on investment as seen in CETA. 

In the wake of this stalemate within the EU, Belgium requested an Opinion by asking the following (narrowly and carefully drafted) question to the ECJ:

‘Is the draft modernised Energy Charter Treaty compatible with the Treaties, and in particular Article 19 TEU and Article 344 TFEU:

so far as concerns Article 26 of that agreement, if that article may be interpreted as allowing the intra-EU application of the dispute settlement mechanism?

in so far as, if Article 26 of that agreement were to be interpreted as allowing the intra-EU application of the dispute settlement mechanism, that agreement does not lay down a specific, express rule or an explicit disconnection clause, in particular in the definitions of investment and investor in Article 1 of the envisaged agreement, providing for the non-applicability of the general mechanism of Article 26 between the Member States?’

The ECJ chose to already answer that question less than a year later by finding in Komstroy that the arbitration clause in the ECT (Article 26 ECT) was ‘not applicable’ to intra-EU disputes (para. 66). Yet, Belgium maintained its request for an Opinion. This in effect turned the Opinion procedure (for ‘agreements envisaged’) into a case about the compatibility of the current ECT (that was already in force), as the information submitted to the Court did not foresee a change to Article 26 ECT. 

 In the end, the Court sidestepped this issue by finding that the Opinion request was inadmissible because it did not have ‘sufficient information’. I will first spend a few words on the Opinion procedure itself, before turning to the Court’s reasoning. 

The Opinion procedure

The Opinion procedure is ex ante in nature and therefore are logical limits to when requests are admissible. On the one hand of the spectrum, the Court cannot give opinions when agreements already have entered into force and are no longer ‘envisaged’. Such an opinion would be devoid of purpose as it can no longer prevent the EU from contracting international obligations that it cannot undertake under EU law. Thus, in Opinion 3/94, the Court found Germany’s request for an Opinion on the Framework Agreement on Bananas inadmissible even though it had made the request before that agreement entered into force. The Opinion was inadmissible because at the time of the delivery of the Opinion, the agreement had entered into force and therefore the opinion procedure was devoid of purpose. 

On the other hand of the spectrum, the Court can also not give an opinion when it is simply to premature to speak of an ‘agreement envisaged’ and therefore there is not ‘sufficient information’ for the Court to give an Opinion on the such ‘agreement envisaged’. The Court has thereby distinguished between questions of compatibility and questions of competence. Questions of competence are met with a lower standard and are therefore more easily admissible than questions of compatibility. Thus, in Opinion 2/94 the Court refused to rule on the question of whether an envisaged accession agreement to the ECHR was compatible with the Treaties, but it did rule on the question of whether the EU had competence to conclude such an accession agreement. In that case, no negotiations had taken place and no detailed information for instance in the form of textual proposals existed. For questions of compatibility, therefore, the Court needs ‘sufficient information on the actual content of the agreement’ (Opinion 1/20, para. 39).

In Opinion 1/20 the Court was confronted with the question of compatibility. In its ruling the Court, first, made two observations: 1) the request concerns the compatibility of the ECT (not the modernised ECT) and 2) there was no text of the envisaged ‘modernised’ version of Article 26 ECT. Belgium had noted that this should not affect the admissibility of the Opinion as it was clear from the decision of the Energy Charter Conference that no change to Article 26 ECT was foreseen, and thus that Article 26 ECT would be maintained in its current form. 

Opinion 1/20

The Court, however, concluded that it could not be inferred ‘that Article 26 of the ECT will not be subject to amendments at the end of those negotiations’ (para. 43):

‘In the first place, although it is true that the Charter Conference identified a list of areas open to negotiation and that that list does not include the dispute resolution mechanism referred to in Article 26 of the ECT, the fact remains that, on the date on which the present request for an Opinion was submitted, the negotiations were at a very early stage and that the judgment of 2 September 2021, Republic of Moldova (C741/19, EU:C:2021:655) had not yet been delivered. It follows that a consensus could have, and might still, emerge, among the Contracting Parties, in favour of the inclusion in that list of the area to which that article applies. Consequently, the outcome of any negotiations concerning that area is not sufficiently foreseeable and it cannot be ruled out that Article 26 of the ECT may be amended.

In the second place, as the Kingdom of Belgium itself pointed out, negotiations were opened in the light of the definition of the concepts of ‘investment’ and ‘investor’, within the meaning of Article 1 of the ECT, such concepts affecting the scope of the dispute resolution mechanism provided for in Article 26 of the ECT. Not only has no amending text of Article 1 been adopted at this stage, but, in addition, the impact that any amendments to those concepts might have on that dispute resolution mechanism cannot be assessed in the absence of any element making it possible to ascertain, with a certain degree of precision, the rules governing that mechanism’ (paras. 44-45).

The Court therefore concluded it had not ‘sufficient information’ and ruled that the request was inadmissible. 


On the face of it, Belgium’s request was straightforward: Article 26 ECT would not be touched in the renegotiations and this is the article that allows for intra-EU arbitration. If it would be part of an agreement envisaged, would it be compatible with the Treaties? The text of the article was there. Moreover, Belgium had clarified in its request that the clause may be interpreted as allowing for intra-EU arbitration. Yet, it is also certainly understandable that the Court would need a text of other provisions of the agreement to be able to rule on compatibility given the interconnection between dispute settlement provisions and more substantive parts of an agreement in its case-law. At the same time, the text of the current ECT that would serve as the basis for any amendments was already there and a starting point for the negotiations. On this basis, it would have certainly been possible for the Court to give an Opinion. By choosing not to, the Court has limited the scope of the Opinion procedure itself. Where questions of compatibility are asked, the Court appears to need a consolidated text that spans significant parts of the envisaged agreement. The Court’s approach therefore is cautious and suggests that it is not willing to give opinions on questions of compatibility at early stages of negotiations. While understandable, this may in itself create complications internationally where the Court’s Opinion is needed but can only be requested at the very end, which may result in the negotiating parties being forced to go back to the drawing board or even abandon the effort.

More surprising perhaps is the first part of the reasoning of the Court. The Court is speculating that the parties to the ECT may actually reconsider their approach to Article 26 of the ECT despite the fact that the parties had explicitly not included that Article to be part of the negotiations. What is more, the Court speculated that this might be the case, because of its own judgment in Komstroy. There was no public information that the parties had changed their intentions until the announcement 12 days after the Opinion other than a brief mention that a disconnection clause had been discussed by the end of April (this information was not referred to in the Opinion). In fact, the ‘modernisation’ negotiations have been notoriously secretive, a far-cry from the more public discussions surrounding reform of ISDS under the auspices of UNCITRAL. Yet, as if the Court had been gifted the special powers of foresight, 12 days later an agreement in principle had been reached by the parties to the ECT over a revamped text that included a commitment to a disconnection clause that would render Article 26 ECT inapplicable for intra-EU disputes.  

The modernised ECT and climate change mitigation

The agreement in principle on the modernised ECT itself appears to address many concerns over compatibility of the ECT with EU law.  It addresses the issue of applicable law and the relationship between ‘domestic’ law and the ECT, introduces the ‘safeguards’ on regulatory autonomy of EU institutions by taking a similar approach to CETA in relation to the investment protection provisions, and of course, it contains a disconnection clause. The devil is in the detail, however, and it remains to be seen if the Commission was able to secure the same textual safeguards as in CETA, for instance by circumscribing in a sufficient manner the ‘fair and equitable treatment’ standard. It also remains to be seen if the disconnection clause itself is worded in a manner that is compatible with EU law.

On the other hand, it does not address the many concerns civil society in particular have over the ECT and reaching the goals of Paris or about ISDS in general. To start with the latter, the agreement in principle fully maintains ISDS, no major procedural innovations (fixing the asymmetry of the system) or limitations (such as exhaustion of domestic remedies) have been foreseen, not even the minor procedural changes in the appointment of arbitrators introduced with the Commission’s ‘Investment Court System’.

Moreover, the provisions that seek to bring the ECT in line with the parties’ commitments with Paris are rather disappointing. Last year, the International Energy Agency warned that new oil and gas projects must stop and no new coal-fired power plants can be built by December 2021 if the world is to stay within the zero-emissions target by 2050. The EU has committed to this goal. One would expect therefore that with a ‘modernised’ Energy Charter Treaty the EU and its Member States would have negotiated to ban and prohibit such projects. Unfortunately, the agreement in principle does quite the contrary. Not only does it not introduce such a ban, it encourages them by continuing to protect investments in new fossil fuel projects. The carve-out proposed by the EU and the Member States for new investments in fossil fuel projects has been watered down to a ‘flexibility mechanism’. This mechanism ‘allows Contracting Parties, based on a  Conference decision, to exclude investment protection for fossil fuels in their  territories’. 

For the EU, this would mean that new investments in fossil fuels would still be protected until 14 August 2023 (a year and a half too late) and there would be ‘limited exceptions’. Apparently, only the UK is the other country that has agreed to such a carve-out, meaning that the EU is effectively entering into an agreement with 20 parties that explicitly protect new fossil fuel investments under the ECT. Importantly, the agreement in principle states that the ‘envisaged exclusions will not, as a matter of principle, affect investment protection in the territory of other Contracting Parties, unless they opt to apply  them vis-à-vis investors from the aforementioned Contracting Parties reciprocally’. This suggests that, for instance, French investors in new fossil fuel projects in one of those 20 countries would still be protected. The EU would also agree to letting third states continue to protect new investments in fossil fuel extraction. What is more, the sunset clause for 10 years for existing investments in fossil fuels is too long, given the efforts of Member States to phase out the use of coal before 2030.

Thus, through this revamped agreement the EU is still encouraging EU investors to go abroad and invest in new fossil fuel projects. All in all therefore, the agreement in principle continues to undermine the goals of Paris and the EU’s ambitions to reach net zero in 2050. That may not come as a surprise to some, as the Commission continues to fund new fossil fuel projects in the EU. 

The agreement contains ISDS and offers protection of portfolio investment by foreign investors in the energy sector. This means that the agreement will likely be a mixed agreement and require ratification by the Member States as well as the EU. In some Member States opposition to the ECT is rising. Last week, the Dutch parliament called on the government to support Spain’s efforts to collectively withdraw from the ECT declaring that investor-state dispute settlement (ISDS) under the ECT should be ended ‘as soon as possible’. France and Poland have also expressed their intention to leave the ECT if modernisation of the ETC fails on addressing litigation risks from fossil fuel companies because of climate action. The European Parliament has urged the Commission to ensure the future ECT ‘immediately prohibits fossil fuel investors from suing EU countries’ (and thus not include a 10 year sunset clause for existing investments as negotiated) and to prepare for a coordinated withdraw from the ECT. Given this opposition, the proponents of the ‘modernised’ ECT  still have a long way to go if they want to preserve ISDS for Europe under the ECT.