Município de Mariana v BHP Group (UK) Ltd & Anor.

TABLE OF CONTENTS:

On 5 November 2015 Brazil suffered its worst ever environmental disaster when the Fundão Dam in South East Brazil collapsed, releasing around 40 million cubic metres of tailings from iron ore mining. The collapse and flood killed 19 people, destroyed entire villages, and had a widespread impact on numerous individuals and communities, not just locally but as a result of the damage to the River Doce system over its entire course to the sea some 400 miles away. The Brazilian public prosecutor has estimated the cost of remediation and compensation at a minimum of R$155 billion, about £25 billion at today’s exchange rates.

In this action some 202,600 claimants seek compensation for losses caused by the disaster from the first respondent (now BHP Group (UK) Ltd by a further recent name change), which is incorporated in England and Wales (“BHP England”); and from the second respondent which is incorporated in Australia (“BHP Australia”). BHP England and BHP Australia are now the only two defendants in this action. The claimants are all Brazilian and comprise (a) over 200,000 individuals, including some members of the indigenous Krenak community who have particular community rights, and for whom the river plays a unique part in their spiritual traditions; (b) 530 businesses, ranging from large companies to sole traders; (c) 15 churches and faith based institutions; (d) 25 municipalities; and (e) 5 utility companies.

The area affected by the dam collapse fell within two states, Minas Gerais, where the dam was, and Espírito Santo, in which the River Doce reaches the Atlantic ocean. The local government authority with responsibility for the area which included the dam itself, and the nearby villages which were destroyed, is the municipality of Mariana. At Appendix 1 to this judgment is a map identifying the affected places and the geographical distribution of the claimants.

The dam was owned and operated by Samarco Mineração SA (“Samarco”) a Brazilian company jointly owned, in 50% shares pursuant to a joint venture agreement, by two other Brazilian companies, Vale SA (“Vale”) and BHP Billiton Brasil Ltda (“BHP Brazil”). BHP Brazil is a subsidiary within the BHP group. Vale and BHP are two of the world’s largest mining concerns. BHP England and BHP Australia sit at the head of the BHP group. At all material times they have operated together as a single economic entity under a dual listed company structure, with boards of directors comprising the same individuals, a unified senior executive management structure and joint objectives. Although the corporate structure is such that it is BHP Australia which is the indirect parent of BHP Brazil, the claim is brought jointly and severally against BHP England and BHP Australia. The claim against each is materially identical, based on the same factual and legal allegations.

Claim forms were initially issued in November 2018, but were superseded by a claim form covering all surviving claims, which was issued and served on the two defendants in May 2019. Both defendants were served with the claim form here as of right: jurisdiction over BHP England arises by virtue of its domicile here under Regulation (EU) No 1215/2012 (“Brussels Recast”); jurisdiction over BHP Australia is established by it carrying on business at offices here, where the proceedings were served.

By that time the disaster had already given rise to a vast number of claims against other defendants in the Brazilian courts; and to an extensive compensation and remediation programme by the Renova Foundation (“Renova”), a Brazilian private foundation established by Samarco, Vale and BHP Brazil. About three quarters of the claimants had been involved in such litigation and/or the Renova programme.

On 7 August 2019 the defendants applied to strike out or stay the claims:

Following an eight-day hearing in July 2020 and further written submissions in September 2020, Mr Justice Turner (“the Judge”) handed down a judgment on 9 November 2020 (“the Judgment”), determining all four applications in favour of the defendants. In summary his main findings were as follows.

The claimants’ application to the Judge for permission to appeal was refused on paper in January 2021, as was its renewal to the single Lord Justice. The claimants then made a successful application under CPR 52.30 to set aside the latter’s refusal of permission. The judgment of this court granting permission to appeal in respect of all four applications has the neutral citation number [2021] EWCA Civ 1156 (“the PTA Judgment”).

Before considering the reasoning in the Judgment more closely, and identifying and addressing the rival arguments on the appeal, it is convenient to set out the nature of the claims in these actions, and the scope and course of the various processes which have taken place in Brazil. The latter, in particular, requires some little detail because the reasoning in the Judgment, and the rival arguments in the appeal, are dependent on such detail.

The claims are set out in a Master Particulars of Claim (“the MPOC”). In addition there are separate schedules providing particulars for the individual claims of each claimant. The claims are all advanced under Brazilian law. Four causes of action are pleaded.

The fourth ground was abandoned in March 2020 prior to the hearing of the applications before the Judge, in circumstances which are addressed more fully below in the context of a dispute about the admissibility of evidence about Samarco’s solvency and/or judicial reorganisation.

It is broadly speaking common ground for the purposes of the appeal, as it was for the hearing before the Judge, that there is an arguable claim against the defendants on each of the three remaining bases, although the defendants reserve the right to challenge the position of some claimants on grounds specific to their individual circumstances.

The claims are all for monetary compensation. There is no claim for injunctive relief. Appendix 2 to this judgment reproduces paragraphs 301 to 307 of the MPOC which list the types of loss by reference to each category of claimant. They involve a wide range of different types of losses, greater detail of which is to be found in the individual particulars of claim. The defendants have been able to conduct a detailed analysis of the make-up of the claimants and their claims from data provided in the answers to questionnaires which were completed by each of the claimants. A selection from those statistical analyses will give a flavour of the type and volume of claims involved:

The claims would need to be proved individually by the claimants as a matter of causation and quantum (and legal recoverability if in issue). We would make two observations. First, some causation and/or quantum issues would likely require an investigation unique to an individual claimant, as for example health issues in a claim for ill health from the polluted water or environment, or a loss of earnings claim by a particular individual or business. However, other claims raise causation/quantum issues which are likely to be common between particular cohorts of claimants - by way of example, the water quality at Governador Valadares, where over 132,000 of the claimants are to be found, as an ingredient of a claim for personal injury and/or loss of earnings. Since proceedings have not yet reached a stage where a defence has been pleaded, or where these questions have even been addressed in the defendants’ evidence, it is difficult to approach arguments about the scope and scale of such issues on anything other than a very provisional, broad and experience-based assessment. Secondly, although the vast majority of the claimants have brought claims for loss of supply or contamination of water, dubbed “water claims”, the losses claimed by individual claimants in these proceedings cover many additional heads of loss. It is not possible to tell from the evidence on these applications whether the loss claimed by any individual claimant is confined to water losses, and if so how many claimants fall into that category. In argument, Mr Choo Choy QC asserted there were none, so far as he was aware. The arguments must therefore be approached on the assumption that the claims are not confined to water losses, although this may prove not to be so for an unidentified number of individual claimants. This may be important, because a number of the processes in Brazil upon which the defendants rely as offering adequate full redress are concerned only with water losses.

There was a huge volume of expert evidence on Brazilian law and practice before the Judge, as well as very substantial factual evidence of the course of events there. It has been supplemented in relation to events since the Judgment by further evidence on the appeal. Some of the expert evidence reveals common ground but there are a large number of substantial and significant disputes. On the defendants’ side, the evidence comes principally from (i) Mr Vivan, a practising lawyer in the firm advising and acting for BHP Brazil; (ii) Professor Didier, an academic lawyer with a particular interest and expertise in civil procedure; (iii) Justice Rezek, a retired Justice of the Brazilian Supreme Court, a former Judge of the International Court of Justice in the Hague and a law professor; and (iv) Mr Calluf Filho, the senior in house lawyer at BHP Brazil. On the claimants’ side there is evidence from (i) Professor Rosa, a professor of law and practising attorney who was formerly the Prosecutor General for the State of Sao Paolo; (ii) Dr Janot, the Prosecutor General of Brazil at the time of the collapse of the dam and for two years thereafter; (iii) Ms Dodge, the subsequent Prosecutor General of Brazil and a current federal sub-prosecutor general; (iv) Mr Deluiggi, a Brazilian lawyer and international arbitrator; and (v) a number of other qualified Brazilian lawyers involved in various proceedings relating to the collapse in Brazil.

Generally speaking, all the witnesses appear well qualified to express the opinions in their evidence. Sometimes the areas of disagreement can be seen to be more apparent than real, or of relatively minor significance, but on some issues there is a sharp divergence of opinion. Where that is the case, as the Judge recognised, it is neither possible nor appropriate on these applications, without oral evidence or cross-examination, to seek to determine who is right. There simply remain areas of uncertainty on those aspects of Brazilian law, with whatever consequences such uncertainty has for the arguments we have to consider. We will seek to highlight what we consider to be the most important areas of uncertainty in what follows.

The Brazilian court system for dealing with civil claims is split into federal courts, in 5 federal regions, and state courts within each of the 26 states. Broadly speaking the jurisdiction of the federal courts is limited to cases in which the Federal Government or related bodies have an interest, international cases, cases involving indigenous populations, and state court cases connected to an existing federal court case which can be consolidated. Issues of whether jurisdiction lies with a state court or federal court are common, and have occurred to a significant extent in the dam-related claims. At each level there are “ordinary” courts of first instance, and second instance courts of appeal. At federal level the second instance appeal courts are Regional Federal Courts (“TRFs”) (which also have some original jurisdiction). At state level they are Justice Courts (“TJs”). Appeals from the first instance courts are as of right and involve rehearing de novo. Appeals at this level are not only available against final judgments, but may be made by one or more of a wide range of interlocutory challenges, including motions for clarification and various kinds of interlocutory appeal. The evidence about some of the dam-related litigation in Brazil suggests that such interlocutory challenges can be productive of considerable delay.

An appeal from those second instance courts lies in each case to the Superior Court of Justice (“STJ”), unless it raises a constitutional issue, in which case the appeal lies to the Federal Supreme Court (“STF”). An appeal to the STJ does not require permission, but lies only on points of law.

In addition to the “ordinary” courts, there are “special” courts, which are small claims courts at both state and federal level for cases of low value and minor complexity. These small claims courts cannot hear cases worth more than about R$40,000 and will not generally hear cases requiring expert evidence. Appeals lie to tribunals made up of first instance judges.

Apart from individual civil claims, there is available, at both federal and state level, a form of class action called Ação Civil Pública (“CPA”). It is available in three categories of case, the relevant one of which is to vindicate “homogenous individual rights”. Those are interests or rights deriving from a single event or common origin, and apply in the circumstances of this case where numerous individuals have suffered loss as the result of a single environmental disaster.

An individual may not initiate a CPA, nor (with immaterial exceptions) become a party to it. Only one of the entities specified by article 5 of the CPA Law may do so. They comprise (1) the Public Prosecutor’s Office, which is (a) the Federal Public Prosecutor (“MPF”) for federal CPAs and (b) the State Prosecutor’s offices for state CPAs; (2) the Public Defender’s Office; (3) the federal government, the states, the federal districts, and municipalities; (4) government agencies, public companies, foundations or government controlled companies; and (5) associations with a common interest which have been established for at least one year. The plaintiff must have an interest in the suit and the court will consider whether the plaintiff is appropriate. The MPF has a special role in relation to bringing CPAs, deriving from its constitutional role as the guardian of the “primary public interest” in Brazil. Whenever the case has social relevance it will be an appropriate plaintiff, whereas other bodies with standing will be appropriate plaintiffs in more limited circumstances.

In homogenous individual rights cases, a CPA may grant injunctive relief, for example ordering a defendant to conduct clean up or other remedial work in the case of an environmental incident; and may hold that a defendant is liable to compensate affected persons in respect of the losses which they have suffered. It does not, however, result in a money judgment in favour of anyone, but rather in a “generic sentence” which addresses liability. If an individual is within the class of affected person whom the generic sentence is intended to protect, the victim must then bring “liquidation proceedings” in the ordinary courts to recover their individual losses, where causation and quantum are required to be established. These liquidation proceedings are not part of the CPA, which comes to an end with the generic sentence; they are separate proceedings governed by the Civil Procedure Code. There is an issue between the experts as to whether a generic sentence in the 155bn CPA will address and determine factual issues of causation as well as liability, to which we will return. There is a provision in the Civil Procedure Code enabling a stay of other proceedings to be granted pending a CPA; but that has not occurred in relation to the main CPAs in these proceedings. The CPA can therefore be treated for present purposes as essentially an opt-out model of collective action: a claimant may commence and pursue their civil claim in the ordinary courts whilst a CPA is in progress, notwithstanding that they are within the class which may benefit from it, and their claim will not be stayed to await the outcome of the CPA.

The res judicata effect of a generic sentence depends upon whether it finds in favour of the defendants. If the defendants are found liable, such liability is res judicata in liquidation claims brought by individual claimants against them: the claimants need only establish causation and quantum of their losses. If, however, the generic sentence finds in favour of the defendants on liability, it has no res judicata effect in respect of individual claims: a claimant is free to seek to establish such liability by way of a claim in the ordinary courts.

There are special provisions relating to the settlement of CPAs. The CPA Law provides for a special form of instrument known as a conduct adjustment agreement (“TAC”). A TAC may be entered into before a CPA, whilst it is in progress, or by way of a final settlement. Where it is part of a settlement of a CPA, it is subject to “homologation” (i.e. ratification) by the court, which is designed to ensure court approval of the appropriateness of the TAC terms.

On 30 November 2015, less than four weeks after the collapse of the dam, a federal CPA (“the 20bn CPA”) was filed against Samarco, Vale and BHP Brazil (together “the Brazilian Companies”) by the Federal Government (represented by the Federal Attorney-General’s Office), the states of Minas Gerais and Espírito Santo, and nine government entities. In substance it sought orders that the Brazilian Companies present plans to deal with the environmental and economic consequences of the dam collapse; take measures to ensure that the specific matters in those plans were dealt with; and fund the implementation of those plans through a private foundation in a minimum amount of R$20bn. Professor Rosa points out that it was unusual in not being brought by the MPF, but rather by bodies who might themselves be defendants to such a CPA; and because it did not seek the usual generic sentence of liability of the defendants to pay compensation.

On 2 March 2016, the parties to the 20bn CPA agreed the Transaction and Conduct Adjustment Term (“TTAC”). TTAC was judicially homologated, but in due course that ratification was suspended and annulled. It provided for a foundation to be the vehicle through which the Brazilian Companies were to carry out the measures to remedy the environmental effects of the collapse and make compensation. Renova was established for this purpose on 5 July 2016. It was a private foundation which the claimants allege is under the control of the Brazilian Companies. Its governance provisions have changed since its first establishment and the defendants point to the involvement of independent bodies in such governance.

The MPF was not a party to TTAC. It was highly critical of TTAC’s terms and of Renova’s governance structure. On 2 May 2016 it filed another CPA in the 12^th Federal District Court against a wider range of defendants, challenging the adequacy of the relief provided for in TTAC and making its own claims for greater relief on behalf of the victims of the collapse, estimating the damage caused at a minimum of R$155 bn (“the 155bn CPA”). The defendants included not only the Brazilian Companies, but also the Federal Government and other plaintiffs in the 20bn CPA.

Despite the annulment of the homologation of TTAC, the parties to it proceeded to implement its proposed plans through Renova.

For its part, the MPF commenced negotiations under the 155bn CPA and on 18 January 2017, the 155bn CPA parties reached a Preliminary Terms of Adjustment Agreement (“the TAP”) seeking to establish new heads of terms with a view to reaching final settlement. Negotiations continued with various court-imposed deadlines being extended without any final settlement being achieved.

Some 18 months later, the parties reached a further and more detailed interim settlement agreement in the form of a Governance and Conduct Adjustment Agreement (“GTAC”) entered into on 25 June 2018. This provided a further and detailed framework by which the parties would progress in negotiations towards a final settlement of the 155bn CPA. Under GTAC the MPF accepted the basic structure of TTAC providing for remediation programmes, including compensation, to be administered through Renova, with expert advisers to assist in formulating and supervising the plans. The governance structure and supervision of Renova was to be improved, with participation of representatives of victims of the disaster. GTAC established detailed procedures for the process of renegotiation of TTAC, which was to take place on the basis of “full redress”. The renegotiation period was initially for two years but was extended.

GTAC brought the 20bn CPA to an end. The 155bn CPA was stayed as a result of the TAP and has remained stayed pursuant to GTAC at least until December 2021; there is a dispute whether the stay is still formally in place, but nothing turns on that because it is accepted that nothing will happen in the proceedings until conclusion of the negotiations, which are still continuing (leaving aside the Priority Axes proceedings, as to which see below, in respect of which there is a dispute as top whether they are part of the 155bn CPA). The negotiations are now taking place under the auspices of the National Council of Justice (“NJC”), which is an administrative body in the Brazilian justice system, presided over by the President of the STF.

Under TTAC, Samarco is obliged to fund Renova. If and to the extent it does not do so, BHP Brazil and Vale are each separately liable to fund 50% of it. That is not an unlimited liability. The funding obligation is to be in accordance with Renova’s budgets, which (the claimants allege) are ultimately within the control of the Brazilian Companies, and subject to minimum and maximum amounts until 2021. GTAC required the Brazilian Companies to maintain assets of R$2.2bn in the Brazilian Courts as security for their funding obligations. The evidence of Mr de Freitas, the current CEO of Renova, is that there has not hitherto been any deficit in Renova’s funding to make the compensation payments (or carry out the remedial works) which it regards as appropriate. Nevertheless Renova’s ability to make full compensation payments to all victims is ultimately contingent on funding by BHP Brazil and Vale, which do not currently have an unlimited funding obligation in relation to claims whose ultimate value is a matter of uncertainty.

There are a number of issues between the parties, and the experts, as to the entitlements to which TTAC and GTAC give rise. Before identifying them, we would make six general observations about GTAC and the Renova redress which it envisaged.

First, “full redress” is a beguilingly simple phrase but it is not always used in the same sense. It is sometimes used as shorthand for compensation which would be recoverable as a matter of right under Brazilian law, which is the sense in which it has generally been used in the parties’ submissions. On other occasions it is used to mean compensation irrespective of such legal liability, which is the sense in which it is sometimes used in the course of some of the Brazilian proceedings. We understand it to be used in the former sense in GTAC.

Secondly, even in that sense, the concept will often mean different things to different people in practice when applied to the multifaceted consequences of the dam collapse. That is inherent in the GTAC provisions themselves, which reflected a need to reorganise the structures and governance of Renova, and in the very fact of its provision for a further negotiation process to identify how “full redress” is to be provided for in practice, a process which has already been going on for approaching four years. Differences of opinion about what constitutes full redress will obviously arise between payors and payees, but they are not necessarily so confined. There may be differences of opinion between the MPF and individual claimants; if the MPF agrees with the Brazilian Companies that a particular loss need not be compensated, that does not mean that a claimant must accept his position, or that it cannot be proved as recoverable under Brazilian law. The MPF may be satisfied that it has got all it realistically can out of negotiations so as to provide the greatest good for the greatest number in the most practical way, without being able to ensure full compensation for everyone. The sheer variety of different circumstances thrown up by the consequences of the dam collapse seems to us to make that a real possibility.

Thirdly, it seems to us to be inherent in the nature and structure of GTAC that it recognised that full redress was not then available to victims of the disaster; and would remain unavailable until the negotiations were concluded and/or the 155bn CPA resumed and decided, which are yet to occur. The essence of the agreement was an acceptance that TTAC was inadequate and required renegotiation if full redress were to be achieved, but that entitlements under TTAC should continue to be provided, for the time being, whilst further renegotiation took place to seek an agreement on further and better redress for victims. The negotiations have not yet produced a settlement which takes the position beyond TTAC.

The potential inadequacy of the redress which has been achieved in the meantime is supported by the sums which have been spent to date. Mr Gibson QC for the defendants relied on the fact that by December 2021, according to the BHP Group accounts, the Brazilian Companies had spent a total of about R$19.6bn, through Renova, on making redress. That is the total, comprising both compensation and sums spent on remediation works. This is a large sum, but must be compared with the estimate of losses made by the MPF in the 155bn CPA that full redress will require at least R$155bn. That is merely an estimate, but if it is even approximately accurate, the extent of redress achieved to date, over 6 years after the catastrophe, is a relatively small fraction of what is required.

Fourthly, and relatedly, there is, at the lowest, a real prospect of the negotiations resulting in a further TAC, whether or not conclusory of the 155bn CPA, which will make provision for further redress to be available to victims beyond the remedies available through litigation in Brazil. What they might amount to is speculative, but there must be a real prospect that the processes to which they give rise will (a) improve the position of the claimants in this action, or at least some of them, beyond what is currently available under GTAC; but (b) take a considerable further period of time to implement, with it being difficult before the end of that period to make an assessment of whether the manner in which they have or have not been implemented provides adequate redress in practice.

Fifthly, as this last point foreshadows, an assessment of the adequacy of redress available from Renova under TTAC/GTAC depends not only on what Renova was supposed to be doing under the terms of those agreements, but also on what it has in fact been doing. Both sides pray this in aid. The defendants say, for example, that the dispute between the experts as to whether GTAC or TTAC imposes any legal obligations on Renova enforceable in the local courts is a sterile one because in practice when Renova has been sued in court proceedings it has never resisted a claim on the basis that it owes no legal obligations. The claimants say that the evidence shows that even where Renova avows that redress is available, in practice it has often not been made available or accessible to victims in a timely manner or at all. As will be seen, two Brazilian judgments lend support to the claimants in this respect.

Sixthly, participation in the redress schemes to be provided by Renova was to be entirely optional. GTAC recognises that victims would be entitled to pursue their claims in the courts instead of, or in parallel with, participation in Renova schemes. It could not, therefore, have been said in Brazil that the current claims, if pursued in the local courts, involved an abuse of the process of those courts on the grounds that adequate redress was available from Renova.

It is common ground that the terms of TTAC exclude from its scope monetary redress for 58 of the claimants in this action. They are 13 large businesses, 25 municipalities, 15 churches and faith based institutions and 5 utilities (“the 58”).

There are two potentially important matters on which there is material disagreement between the experts, namely:

TTAC defines the impacted parties who are to benefit from it as those “directly affected” followed by a list of specific categories of recoverable loss, which are drawn widely. Professor Didier says that he cannot imagine any type of damage which is recoverable under Brazilian law which would be excluded by the list, and accordingly TTAC provides for redress for all losses recoverable under Brazilian law. Professor Rosa disagrees. He points to the following limitations.

Professor Didier disagrees in response.

We note that in his first report at para. [220] Professor Didier appeared to recognise that there could be cases of recoverable losses which were outside the categories of defined loss provided for in TTAC, despite what the report says elsewhere.

This is a summary of some of the headline points in a large volume of material which we have considered. Our conclusion is that we are unable to determine whose views are to be preferred on these interlocutory applications. We therefore proceed on the basis that Professor Rosa’s views may be right, and that there may be a significant category (or categories) of losses for which TTAC and GTAC provide no redress.

Professor Rosa’s view is that TTAC imposes on Renova no legal liability to suit by victims; and that that remains the position under GTAC. As for Samarco, its only obligations under TTAC/GTAC are to fund Renova, not provide redress to victims. Professor Didier disagrees in vigorous terms. We do not consider it necessary to go into greater detail on the rival arguments advanced on this issue by the Brazilian experts. They are such that it is impossible to assess their weight or choose between them on an interlocutory basis. We therefore proceed on the basis that that GTAC and TTAC may or may not impose any legal liability on Renova or Samarco to compensate victims.

The defendants seek to sidestep this dispute on the expert evidence with an argument, accepted by the Judge, that this debate is sterile in the light of (a) the evidence of Mr de Freitas that Renova acknowledges its obligation to make redress in accordance with TTAC/GTAC and (b) the absence of evidence that Renova or Samarco has ever taken a point when sued (which has occurred in Renova’s case about one thousand times and in Samarco’s case tens of thousands of times) that either is an entity upon whom TTAC/GTAC does not impose legal liabilities.

It is not quite accurate to say that Samarco has never denied liability for the consequences of the collapse of the dam. On the contrary, Samarco has denied liability, and maintains a denial of liability, to pay the claims of the 13 large businesses at all. Moreover, Samarco has disputed liability in its defence to the claim by the water utility in Baixo Guandu for, amongst other things, the costs involved in water treatment and providing alternative water supply, on the grounds, amongst others, that strict liability does not apply because “the claim is not based on environmental damage per se but rather pecuniary losses … specific to [the utility]”; and in a claim by three individuals for mud damage to their furniture and moral damages, on the grounds that there was “no assumption of civil responsibility”. Nevertheless the defendants’ argument proceeds for the most part from a sound factual premise firmly grounded in the evidence. Samarco has been sued in some 67,000 separate lawsuits by these claimants, and the evidence of Mr Vivan, BHP Brazil’s lawyer, is that in none of these, or the many other cases in which it has been sued by other victims, has Samarco denied an obligation to make full redress. The claimants’ evidence does not contradict this, with the exceptions we have identified. Other suggested exceptions set out in Appendix 20 to their skeleton argument before the Judge seem to us, on analysis, to amount to denials of liability specific to particular forms of relief, not denials of any obligation to make redress.

If it could be said with reasonable certainty that the position of Renova or Samarco would not change, there would be force in the defendants’ point, leaving aside for the moment the position of the 13 large businesses. However the future course of events in Brazil is a matter of speculation and uncertainty. Renova’s and Samarco’s current stance is no doubt underpinned by the spectre of more onerous burdens which may be imposed in a final TAC or the pursuit of the 155bn CPA, for so long as negotiations are continuing. If there is a final TAC which brings negotiations and the 155bn CPA to an end, it does not seem to us to be by any means inconceivable that Renova will change its stance, if the TTAC model remains in place, or even that it will be replaced in a different implementation model. Samarco’s attitude, and therefore that of Renova too, is dependent on funding from Vale and BHP Brazil, which seems to us again a matter of uncertainty in the future, given the size of the sums potentially involved. In this context we attach significance to clause 256 of TTAC which expressly preserves its non-acceptance of legal liability; and to the fact that although Samarco has (with the limited exceptions identified) not sought to challenge liability for redress, it has carefully avoided explicitly accepting legal liability for the consequences of the collapse throughout the litigation in which it has been involved, with only one exception (a settlement agreement in the CPA brought by the State Prosecutor in Mariana). We asked Mr Gibson in the course of the hearing whether there was any evidence explaining why Samarco had not formally admitted liability, but he was unable to point to any, although he said on instructions that Samarco considers that the strict liability provisions under Brazilian Environmental law would not apply to the larger commercial claims.

There is a separate question whether Samarco could realistically deny liability in ordinary civil claims not reliant on GTAC/TTAC. Mr Gibson pointed to the submissions of Mr Hollander QC on behalf of the claimants before the Judge, which averred that Samarco could have no ground to dispute liability under the Environmental Law as the direct polluter. However this was merely the claimants’ submission, and is not reflected in any agreement by these English defendants. We cannot proceed on the basis that that is beyond dispute.

We would not accept, therefore, that this court can safely proceed on the assumption that Samarco’s liability to claimants is incontestable, or that there is no real risk of it ever being contested, in suits brought in Brazil. That remains a possibility which cannot be discounted as fanciful for the purposes of the present applications.

Mr Gibson argued that if this was a concern, it could be met by a stay which could be lifted if and when it arose. We would not find such a course attractive. If this action is to proceed at all, it is in the interests of the parties and the proper administration of justice that it should do so now, for reasons which we expand upon when dealing in particular with the article 34 arguments below.

The position of the 13 large businesses is more clear cut: Samarco continues to maintain that there is no legal liability towards them for the consequences of the dam collapse. The defendants argued that nevertheless Renova would not dispute liability, relying on Mr de Freitas’ third witness statement dated 10 December 2021, served after the Judgment. There he stated, amongst other things, that:

“17.

Further, if any of the English Claimants brought a claim in Brazil against Renova for compensation for losses suffered as a result of the Dam collapse, Renova would not argue that it is not obliged to make full redress to them. However, challenges could be made on whether or not that specific plaintiff suffered loss or damage, whether loss or damage has been caused by the dam break and is recoverable under Brazilian law and on the quantum of the compensation sought.”

Leaving aside the question as to whether any weight can safely be placed on this statement, which we address further below, Mr Gibson’s answer on instructions as to the grounds on which Samarco is disputing liability to the 13 would bring them within Mr De Freitas’ stated exception (i.e. excluding cases where it was contended that there was no recoverability under Brazilian law). He is not, therefore, to be taken as asserting that Renova would accept liability towards the 13.

Renova has provided compensation to a large number of victims, in addition to the remediation works it has undertaken. This has primarily been achieved through two programmes known as PIM Water for water interruption claims and PIM General for other heads of loss. There was also a programme for emergency financial assistance to which we do not need to refer.

The PIM Water programme first opened in Governador Valadares in October 2016 and was extended to other municipalities. It closed on 31 December 2017. It was intended to be a quick and simple mediation process. It offered a fixed compensation amount calculated on the basis of the period the impacted communities were left without water and the average amount of local water bills. There was an enhancement of 10% for the elderly, children and those in particular categories of vulnerability. It was thus limited to interruption of supply loss calculated by reference to the bills for that period, plus some uplift for the old, young and vulnerable. It did not address, for example, claims for loss of earnings as a result of interruption of supply/contamination, nor claims for personal injury. A total of R$270.7m was paid under this scheme. Those who claimed and received payments were required to sign waivers in favour of the Brazilian Companies covering “damage related to water supply and distribution” save in respect of future damage. This appears to exclude claims in rather wider terms than the kinds of losses for which compensation was made i.e. based on the average water bills for the interruption period. The lawfulness of the releases was challenged in a Water Settlements CPA, on grounds amongst others that victims were unlawfully coerced into signing. There were decisions declaring them to be unlawful and void, but there were a number of successful appeals by Renova. The Water Settlements CPA has not been finally resolved: it was transferred to the 12^th Federal Court and has been stayed since November 2018 pending the determination of the 155bn CPA.

The PIM General programme was to indemnify individuals and micro and small businesses for material damage, loss of income and moral damages (insofar as it did not arise from interruption of water supply). The process involves an application which is “formalised” by the applicant, followed by registration by Renova. Renova investigates the claims, usually on an individual basis but in some respects in accordance with a damage matrix. It is apparent that this was a slow process. In his judgment dated 30 October 2021 in the Priority Axis 7 proceedings (as to which see below), Judge Mario Franco Junior (“Judge Mario”) recorded that Renova’s registration programme had been paralysed since January 2018, over 3 ½ years earlier, with the result that “a huge number” of affected parties who had formalised their applications since that date had not had their registrations completed. The scale of Renova’s inability to handle the applications efficiently is apparent from the fact that whilst Judge Mario closed the scheme to new applicants with effect from 31 December 2021, he gave Renova until 1 January 2024 to complete the registrations. As we understand it, the registration process involves further clarification by Renova of individual claims and requests for supportive evidence, but it is only following registration that Renova starts its own assessment of the merits of the claim.

Recipients of payments under the PIM General programme are also obliged to sign waivers. Mr de Freitas suggests that the scheme will be available to all the claimants in this action by virtue of their having commenced claims in England prior to 30 April 2020. The reference in the judgment to claimants who have claimed anywhere prior to 30 April 2020 is, however, in respect of eligibility for the Novel System (see below), not PIM General. It is therefore possible, at the lowest, that those claimants in this action who have not already formalised applications under PIM General are no longer eligible for the scheme as a means of redress.

The claimant questionnaires reveal that 77,947 of the claimants were participating in the PIM General and/or Water programmes (at the time that the questionnaires were completed), of whom 7,956 had additionally brought civil claims. 97,890 of the individual claimants in this action stated that they have received some form of payment from Renova. The amounts, and the extent of their overlap with what is being claimed by the claimants in this action, are not apparent from the evidence. The claimants have said that they will give credit in this action for sums received from Renova.

Pursuant to TAP and GTAC, three experts were appointed: an environmental expert, Lactec, to report on environmental aspects of the consequences of the dam collapse; FGV, to consider socio-economic aspects of the consequences; and Ramboll to address the Renova programmes. The current position is that Lactec has produced its final report, and FGV and Ramboll have produced a number of interim/thematic reports. We have not seen any of those reports, although some of the claimants’ evidence refers to Ramboll being critical of Renova’s response in some specific respects.

These may be split, broadly, into those which have taken place pursuant to GTAC/TTAC on the one hand, and civil litigation on the other. The two have been running in parallel. We will deal with each in turn. There have also been some 70 other CPAs and quite a number of different kinds of civil proceedings other than individual civil suits, but those are not of central significance to the current dispute, and we do not propose to address them.

Clauses 255 and 258 of TTAC and clause 103 of GTAC provide for disputes between the parties to be submitted to the 12^th Federal District Court which sits in Belo Horizonte. The assigned federal judge who has dealt with the resultant litigation in that court is Judge Mario.

In December 2019, Judge Mario established a series of Priority Axis proceedings to address different aspects of the disputes, following complaints made by federal and state agencies in two petitions in May 2019 to the effect that the Renova programme was still leaving affected families without adequate redress. Priority Axis 7 is concerned with the compensation scheme operated by Renova under TTAC. In his later judgment in October 2021 Judge Mario said that the position under the Renova programme when the Priority Axis system was established (over four years after the dam collapse) was that practically no category of claimant had been recognised and no one had been compensated.

The jurisdictional basis for the Priority Axis proceedings is controversial. The defendants characterised them as essentially resolving disputes arising out of the Renova programmes and arising conceptually out of GTAC/TTAC, pursuant to provisions in those agreements which enabled matters in dispute to be referred to the 12^th District Court. However the MPF has challenged the legitimacy of the Priority Axis process under an appeal which seeks to have it declared a nullity, on the basis that it has “mutilated the participatory model” which was established under GTAC/TTAC, involving amongst others the Interfederative Committee. In that challenge, interim relief was refused but the appeal itself remains outstanding in the federal TRF, during the pendency of which Judge Mario has continued to conduct the Priority Axis process.

Professor Rosa explains that the Priority Axis proceedings are a novelty whose jurisdictional basis is uncertain and whose existence and validity is precarious in the light of the appeal by MPF. Professor Didier challenges what Professor Rosa says, and Mr Calluf Filho, in-house counsel at BHP Brazil, in his recent evidence, expresses confidence that the appeals will not succeed. He may well be right, but these conflicts cannot be resolved by us on these applications. There therefore remains an element of uncertainty over whether the Priority Axis process, and what has been done pursuant to it, will be nullified. This is significant because one of the major avenues of redress upon which the defendants place particular reliance in these applications, is the so-called Novel System, established by Judge Mario by a judgment dated 2 July 2020, and extended in a number of subsequent judgments. These decisions were made in the Priority Axis 7 proceedings which were concerned with Renova’s indemnification programmes.

The judgment of Judge Mario of 1 July 2020 establishing the Novel System was given in a claim brought by a local commission representing the 30,000 odd inhabitants of Baixo Guandu. We will refer to it as “the Baixo Guandu Judgment”. The important aspects are as follows.

Contrary to the defendants’ position, the criticisms of the availability of redress in the legal system were not confined to difficulties of evidential proof. They extended to the inability of the legal system to cope with a multiplicity of claims of this nature and the risk of conflicting judgments to which they give rise.

Since the Baixo Guandu Judgment, 72 local commissions have brought similar claims and Judge Mario has given judgments extending the application of the Novel System to new areas, amending the damages matrix and providing for a right of appeal, first to an expert and thereafter to Judge Mario himself.

There is also a significant recent judgment of Judge Mario dated 30 October 2021 (“the October 2021 Judgment”), which deals with the following.

The October 2021 Judgment has also been the subject matter of appeals. The Brazilian Companies challenge, amongst other things (1) the extension to new areas outside the TTAC definition; (2) the extension of its availability to all applicants until 30 April 2022, contending that it should only be open to those who complied with the earlier local commission deadlines in respect of already accepted areas; (3) the PIM General rulings; (4) certain aspects specific to two areas, Barra Longa and Gesteira; and (5) the recognition of water damages under the Novel System on the grounds that these are the subject matter of the IRDRs (described below). This last seems to us a significant challenge to the whole Novel System itself, given that the overlap between the water claims being dealt with in the IRDRs and those in the Novel System appears substantial even if not total. This is reflected in the fact that the Brazilian Companies sought a stay of the Novel System as regards water damages (pending their appeal), which was refused and unsuccessfully appealed to the STJ. The substantive appeal against the October 2021 Judgment by the Brazilian Companies remains pending.

As at 9 March 2022 some 119,000 claims had been made under the Novel System, including by about 19,400 of claimants in this action. Of these, 9,800 of the claimants in this action had reached agreements, and therefore presumably signed waivers. As we have mentioned, the claimants have said that they will give credit in this action for sums received from Renova, which would include any received under the Novel System.

The features of the Novel System of particular relevance to the issues in this appeal are, in summary, the following:

As well as about a thousand claims against Renova, many tens of thousands of individual civil claims have been filed against the Brazilian Companies in the small claims and ordinary state courts in Minas Gerais and Espírito Santo. There are 67,316 of the claimants in this action who have stated that they have commenced such claims, of whom 19,542 stated that such proceedings have “concluded”, resulting in a total payment of R$ 21.3m. It is not clear from the evidence to what extent these involved any court decision. The claimants’ position is that a large number of them will have been settled, in particular through an early conciliation scheme in Governador Valadares. Nor does the evidence reveal whether the 19,542 “concluded” claims included all the heads of loss claimed in this action. Mr Choo Choy submitted that the relatively small recovery per claimant (calculated as an average from the total recovered) suggested that these would primarily have been water claims. These figures mean that there were at the time the questionnaires were completed some 48,000 of the claimants who had claims pending in the civil courts in Brazil. Professor Rosa opines that it is likely that the vast majority of these claims were and remain stayed as a result of the Minas Gerais Damages IRDR (which we explain next), a view supported by a letter of 3 December 2019 from the President of the Court of Justice of Minas Gerais to the President of the NJC (“the December 2019 Letter”) which suggests that there were then 43,742 claims in Minas Gerais which were stayed pending determination of an IRDR because they included water claims.

A procedure known as IRDR was introduced in 2015 in both federal and state courts to meet a general perception that there were too many inconsistent rulings. It is a procedure by which courts can deal in a single set of proceedings with an issue of law which arises in more than one claim. It is confined to issues of law. Appeals from IRDR decisions lie to the STJ or STF.

There have been three IRDRs of potential relevance:

The Espírito Santo IRDR was commenced as a result of conflicting judgments in the small claims courts in that state as to the amount to be awarded as moral damages for water supply interruption. It led to a decision on 10 March 2017 of the 13 judge panel of the small claims court fixing the level of moral damages for water supply interruption at R$1,000 for interruptions lasting for between 3 and 5 days. This was a majority judgment, four of the minority considering that the amount of moral damages was not an issue of law to which the IRDR procedure could apply. This decision has not been appealed and is final and binding in state courts in Espírito Santo.

The Minas Gerais Jurisdiction IRDR was commenced by Samarco and resulted in a decision on 28 May 2018 that the small claims courts had no jurisdiction to determine claims for moral damages because such cases involved expert evidence about water quality. That decision was appealed and the appeal remains pending.

The Minas Gerais Damages IRDR was also commenced by Samarco. On 27 July 2017 a stay was imposed on all claims before any state or federal courts in Minas Gerais, including small claims courts, which included a claim for moral damages arising from the interruption of the water supply or doubts about water quality, pending the final determination of the IRDR. Five issues were raised, the last being what the quantum of such damages should be. This resulted in a decision of the state court of appeal on 24 October 2019 which fixed moral damages in a minimum amount of R$2,000 in what was described as “ordinary circumstances” but subject to review or adjustment in the light of the victim’s “particular circumstances” or “singular aspect” (two different translations of the relevant wording of the judgment). The decision on whether to admit the claims was a majority judgment splitting 6/5, with the minority holding that the fixing of damages was not a matter of law. There has been an appeal from the decision which remains pending. Meanwhile the stay on all claims in Minas Gerais which include water losses remains in place.

There is a dispute between the Brazilian law experts as to whether there is an effective procedure available in the state courts or federal courts to try issues of fact which arise in common in individual proceedings in a collective way (save by CPAs). Professor Didier points to mechanisms for collective liquidation proceedings under article 97 of the CDC (which is described as available in liquidation proceedings, which would mean they were only available following a favourable generic sentence in a CPA); and judicial cooperation in article 69 which would permit, for example, common production of evidence (which is identified to be available in relation to ordinary civil claims). Professor Rosa disagrees that collective liquidation proceedings would be feasible in practice and further disagrees that judicial cooperation would be available in ordinary civil claims, opining that it does not permit a shift of jurisdiction in relation to decisions on substantive issues. He further points to the fact that article 69 has not been invoked for that purpose in any case before. We would observe that what Judge Mario said in the Baixo Guandu Judgment seems to bear out Professor Rosa’s view that there is no satisfactory ability to try common generic issues together, or at least one which has been deployed in relation to victims’ claims in this case.

Again we cannot resolve the differences of opinion between the experts, and therefore proceed on the basis that there may or may not be a satisfactory procedure available for trying common issues of fact which arise in individual claims in the civil courts in Brazil.

Justice Rezek explains that legal aid is available to claimants in the small claims and ordinary local courts. Dr Janot advances a number of points on the adequacy of such availability, which to our thinking are not, on analysis, very powerful. The claimants relied also on an example given by Mr Deluiggi of a case brought by AHOMAR, an association which describes itself as one of the leading organisations for the protection of traditional fishing and the environment in South East Brazil, in which legal aid was refused on the basis of insufficient evidence of AHOMAR’s impecuniosity. However there is nothing to suggest that there was any error in that decision, which does not establish a lack of access to justice.

On the other hand, the evidence of Mr de Moura Baptista, a Federal Deputy representing the state of Minas Gerais, strikes us as more significant. He explains that obtaining a legal aid lawyer in that state is in practice very difficult because since September 2017 the Minas Gerais Bar Association has been recommending that its members do not undertake legal aid work, as a result of the state’s failure to pay thousands of legal aid lawyers. He also gives cogent reasons why individual claimants could not have their claims brought by Public Defenders or Public Prosecutors.

There is also clear evidence on the claimants’ side that third party litigation funding is not available in Brazil. The defendants’ evidence is that it is uncommon but not that it is unavailable.

The waivers in the Novel System suggest that where attorneys are used in that streamlined process, they take a fee of 10%.

The appellants served 19 factual witness statements dated between 5 and 13 March 2020 attesting to the particular problems encountered in seeking to obtain redress. Some are from individual claimant businesses or utilities. Others are from representatives of municipalities who attest not only to the situation of the municipality, but also to that of the individual claimants they represent, including one from someone representing the 132,000 odd claimants from Governador Valadares, which is both a city and a municipality. There is one from the president of an area fishing association who has a representative role in respect of fishers affected by the disaster more generally.

They give detailed examples of (1) the small proportion of individual claimants in some categories who had received any compensation to date; (2) Renova payments being non-existent or inadequate for (deserving) fishers, other individuals and businesses; (3) Renova acting arbitrarily and in an unjustified and unprincipled way in categorising losses as “indirectly affected” so as to refuse to compensate them; (4) Renova failing to spend more than a fraction of the budget allocated to programmes pursuant to TTAC; (5) Renova being obstructive, dragging its feet or making a volte face even where it was agreed in principle that there should be redress; (6) Renova imposing complexity, bureaucracy, and unfair qualification criteria for applicants, so as to impose substantial obstacles to the availability of redress in practice even where it was recognised as an entitlement in principle; (7) extensive delay in the court processes, including as a result of Samarco and Renova using the interlocutory appeals system and other procedural tactics as an obstacle put in the way of court proceedings, which have precluded recovery by business claimants and prevented the municipalities from recovering more than a small proportion of the very substantial losses which they have suffered; (8) Renova imposing settlements on individuals and small businesses, with attendant waivers, through threats and unfair economic duress; (9) serious criticisms of Renova by Ramboll, one of the experts set up under TAP to supervise the establishment and implementation of programmes.

These cannot be dismissed, as the defendants sought to do, on the basis that they are a small number of outliers which can be discounted because they would have been selected as the most extreme examples. They paint a picture of the difficulties faced by a wide range of claimants. Nor can they be dismissed on the basis that there will always be some difficulties in a programme of redress of this magnitude, and some dissatisfied claimants. They evidence serious deficiencies in Renova providing redress across a broad range of problems, not just by way of general assertion but backed by specific and detailed examples. They may not be wholly representative, and they may be inaccurate or misleading; that is not a matter which can be determined on these applications. But taken at face value they paint a coherent picture of widespread inadequacies in the redress available from Renova, within a reasonable time or at all. We can only proceed on the basis that they may (or may not) be accurate and representative of deficiencies in the redress available to victims in Brazil.

We have already referred to Judge Mario’s view expressed in the Baixo Guandu Judgment that the system was administratively unable to cope with the volume of claims resulting from the dam collapse.

The claimants also served a statement from Ms Dodge who was the Prosecutor General of Brazil between 2017-2019, in succession to Dr Janot, and remains in the service of the MPF as a sub-Prosecutor General. In the light of her current position she avowedly wished to remain neutral. Nevertheless she recognised that the vast scale of a single environmental event, specifically the nature and extent of the damages and the number of people directly affected, and the limited resources available to the courts, gave rise to undue delay and grave obstacles regarding effective access to justice on the part of victims.

The defendants relied upon statistics collated by Justice Rezek that claims were dealt with relatively swiftly in the small claims courts and ordinary local courts. These were not, however, tailored to the particular circumstances of these dam- related claims. The defendants (and the Judge) relied upon the December 2019 Letter as showing that such dam related claims would be dealt with in an average of 414 days. We do not think that the letter provides quite the degree of comfort which the defendants seek to draw from it. It refers to 82,383 lawsuits in various parts of Minas Gerais of which 27,050 were judged within an average time of 414 days for reaching a decision, with 43,742 stayed by an IRDR decision (no doubt the Minas Gerais Damages IRDR). Mr Soares, a Brazilian lawyer coordinating the claims of the largest cohort of individual claimants, the 132,000 from Governador Valadares, made a statement doubting the reliability of the figures; and suggesting that of the 27,050, half or more are merely ratification of settlements so that the average for contested claims being decided will be significantly higher. Justice Rezek gives apparently cogent and convincing reasons for doubting Mr Soares’ evidence. But even assuming that the 414 day average is representative, it must be taken together with the fact that over half the claims remain stayed to this day, over 6 years after the dam collapse.

The initial two year period for negotiations under GTAC expired in August 2020 and was not formally extended. Instead negotiations have continued since March 2021 under the auspices of the NJC, whose President is the President of the STF. The NJC negotiations involve the Federal and State Public Prosecutors, the Public Defenders, the Federal and State governments and the Brazilian Companies, together with representatives of the courts of Espírito Santo and Minas Gerais and the municipalities. There is a Letter of Premises which identifies the topics for the negotiations, which are at a fairly high level of generality, but provides that the negotiations are to be are to be conducted on the basis of full compensation. The view expressed by Mr Calluf Filho, who is involved in the negotiations, is that it is possible that they will be completed in mid-2022, and that if no agreement is reached by then it is more likely that they would not continue thereafter; and that the 155bn CPA will resume in respect of those matters not agreed or otherwise determined in the Priority Axes.

At para. [136] of the Judgment, the Judge proceeded on the basis that the 58 large businesses, municipalities, churches and faith-based institutions and utility companies could not benefit from a generic sentence in the 155bn CPA. Mr Toledano QC accepted that that was so, although the evidence is perhaps slightly more nuanced.

Mr Vivan’s evidence on behalf of the defendants was that the 155bn CPA did not include all losses sought by the 58 in the English proceedings:

Professor Rosa provides a cogent explanation why claims could not, as a matter of law and principle, be granted in a CPA in respect of homogenous rights in favour of the 58. Professor Didier disagrees for equally coherent reasons. We cannot resolve the conflict between the experts. Professor Didier does, however, accept that large businesses cannot benefit from a generic sentence in a particular CPA which makes clear it does not extend to them; and it does not seem to be disputed that the claims actually made in the 155bn CPA are limited in the way described above in the defendants’ own evidence.

The position for the 58 therefore appears to us to be that (1) it is uncertain whether any of them could in principle benefit from a generic sentence in the 155bn CPA; (2) it is tolerably clear that there will be no generic sentence upon which the 13 large businesses can rely; and (3) the MPF is not seeking a generic sentence which would cover large parts of the losses which the remainder of the 58 are seeking in this action, and consequently they could not benefit from a generic sentence in the 155bn CPA in respect of most of their losses.

The Brazilian Companies, the federal government, and the state government of Minas Gerais are now the only remaining defendants to the 155bn CPA. There are no municipalities which are parties to it. There is a sharp divergence in the evidence on the prospects of the 155bn CPA proceeding to a generic sentence at all. It is common ground that it will not do so if there is a final settlement reached, which is one possibility. In the absence of a final settlement, the claimants’ evidence is that there is in practice no prospect of the MPF pursuing it. The Judge said at para. [185] of the Judgment that he found the reasoning behind this evidence “elusive”. We do not think it can be dismissed on that basis. The evidence includes the following.

Professor Didier disagrees with the views of Dr Janot and Professor Rosa. His view is that it would make sense for the MPF to pursue the 155bn CPA if negotiations do not result in a satisfactory settlement, and that the effect of articles 99 and 103 of GTAC is not as contended for by Professor Rosa, but rather such as to leave all extant claims in the proceedings to be resolved, absent a decision of the MPF to withdraw any. Mr Vivan supports those views.

Again we cannot determine who is right on these issues. We must accordingly proceed on the basis that there is at least a very real possibility that, even in the absence of a final settlement of all issues, the 155bn CPA will never resume.

The defendants contend, as supported by Professor Didier’s evidence, that a resumed 155bn CPA will decide everything which is in the complaint document initiating the process, including the allegation of joint and several liability of the Brazilian Companies, and liability for all the remediation work then sought. However, if Professor Rosa’s views are to be preferred, it will not include any determination of the liability of the Brazilian Companies, even if it is pursued. Nor, if he is right about clauses 99 and 103, will any pursued claims for in kind remediation relief necessarily be all those in the original complaint. We must therefore proceed on the basis that it is uncertain whether the liability of the Brazilian Companies will be the subject matter of any decision in the 155bn CPA, even if it is resumed following the conclusion of negotiations, and that the scope of any decision on in kind relief is equally uncertain.

There was also considerable further dispute between the parties, and in the evidence, which the Judge found it unnecessary to resolve, over whether a resumed 155bn CPA would decide issues of causation or some factual questions relevant to quantum. Whilst Professor Didier’s views are powerfully reasoned, we cannot reject Professor Rosa’s views on an untested interlocutory basis.

The defendants seek to circumvent this dispute between the experts by resorting to logic. They submit that it inexorably follows that issues of causation will have to be determined from the very fact and nature of the in kind relief for remediation works which is being sought in the 155bn CPA. They argue that if the court is, for example, being asked to order a clean-up programme in a particular area, that will necessitate a finding of causation that the dam collapse has adversely affected the area. Mr Vivan states that this will necessarily, for example, involve an assessment of water quality before and after the collapse, which will be an important element of causation in the claimants’ claims in this action.

The argument has some attraction but its weight is lessened by three considerations. First, if Professor Rosa is right about the effect of clauses 99 and 103 of GTAC, the extent to which there will be a decision on remediation works does not depend on the content of the original complaint and is a matter of speculation. Secondly, a 155bn CPA decision on remediation will be based on the situation at the time it is given, which will be some 9 to 11 years after the collapse on the defendants’ most optimistic estimate, and quite possibly much later (see below). The principal exemplar of an overlapping causation issue given in the defendants’ evidence and argument before the Judge was a finding as to water quality in a particular place before and after the dam collapse. But the inquiry for the purposes of remediation works related to water quality will be concerned with the water quality when the 155bn CPA is decided. On the other hand it may well be the water quality in the immediate aftermath of the collapse which is primarily of relevance to the particular claims in this action, which may have changed after a decade or more of rain by the time the 155bn CPA addresses it. Thirdly, the argument does not transcend one of Professor Rosa’s points, which is that what the CPA court addresses as a matter of narrative in reaching its conclusions is not part of the generic sentence, and it is only the generic sentence itself, not the narrative, which gives rise to res judicata effect in Brazil.

As with other issues, therefore, we regard it as a matter of uncertainty whether a resumed 155bn CPA, if it takes place, will decide issues of causation or factual issues relating to quantum which overlap with issues in this action.

There is another question of potential relevance to the overlap between the 155bn CPA and the current action, which is the basis on which the liability of BHP Brazil as an indirect polluter will be addressed, assuming it will be addressed at all. Mr Choo Choy directed us to passages in the original complaint and BHP Brazil’s defence, in support of a submission that liability based on BHP Brazil control of Samarco was limited to an argument that such control existed merely by virtue of its shareholding. It is not clear to us that the case which is to be advanced is necessarily so limited, although the documents are consistent with such an approach. It is early days in the development of the case on each side, which would become clearer were the 155bn CPA to resume and address BHP Brazil’s liability. For present purposes all that can be said is that BHP Brazil’s liability as an indirect polluter based on control may or may not be limited to whether control is established merely by virtue of its shareholding.

Mr Vivan states that he would expect the 155bn CPA to be decided at first instance within two to four years of its resumption following the conclusion of negotiations. He explains that that is based on specific factors relevant to this CPA which would accelerate the process which might be expected in CPAs more generally. These include that the parties have already procured some expert reports; Judge Mario is already familiar with issues relating to the collapse; he has shown a determination to deal with matters speedily hitherto in the Priority Axis proceedings; and the high profile nature of the case would create pressure to resolve the case quickly.

The claimants estimate that it is likely to be at least a decade or more, based on evidence which includes the following:

The length which appeals might take is also addressed by Judge Rezek. The median length of an appeal from the first instance to second instance courts in his sample of CPAs was around 4 years three months. The average length of appeals to the STJ according to its 2019 Report was about one year.

Again, we cannot determine whose evidence is to be preferred where there are conflicts, and this is in any event an uncertain exercise in predicting the future: it is not inevitable, for example, that Judge Mario will be available as the assigned Judge for however long the further proceedings take; he was reassigned to another Federal Court and the matter left in the hands of another judge for a while, although he has subsequently been reassigned to the 12^th Federal Court. We must proceed on the basis that the most optimistic evidence of when the CPA will be resolved at first instance is mid-2024 to mid- 2026; but that it may well be significantly later, possibly by many years; and that if, as must be a real possibility, there are further appeals through two levels up to the STJ, that will probably add a further 5 years before the 155bn CPA is finally resolved. There is, in other words, a real possibility that final resolution of the 155bn CPA, if it resumes at all, is well over a decade away.

The defendants’ primary case is that they should be sued in Brazil, if at all, in a new CPA which could then be consolidated with, and heard as part of, the 155bn CPA. Alternatively they contend that they are amenable to suit in individual civil claims brought in local courts.

It is common ground that the Brazilian courts would have territorial jurisdiction over the claims brought in this action if brought against BHP England and BHP Australia in Brazil, in either procedural way, such jurisdiction being founded on (1) the Brazilian domicile of the claimants and (2) the location of the events in question. Moreover BHP England and BHP Australia have agreed to submit to the jurisdiction in a new CPA or in civil liability actions, whether liquidation proceedings or other ordinary civil actions.

We have already referred to Dr Janot’s evidence that the suggestion that the MPF would bring a new CPA against BHP England and BHP Australia is entirely implausible. Professor Rosa is of the same view. Their reasoning is persuasive; in particular a new CPA would involve new parties with different factual bases of claim from those previously alleged, and additional complications by virtue of their being foreign defendants. The MPF has already committed an enormous amount of resources to pursuing the Brazilian Companies in the 155bn CPA, and a new CPA would take a great deal of time to bring to a conclusion in Brazil and would further delay the 155bn CPA (if the latter were to be resumed and the cases heard together), with the prize merely being additional defendants, rather than a greater extent of liability. A generic sentence in a new CPA would only enable liquidation proceedings, which themselves would have to be pursued against foreign defendants with all the attendant complications of process and the difficulties involved of enforcement abroad.

Professor Didier’s evidence in his first report (which predated the claimants’ evidence) was that he had “no doubt” that the MPF would commence a new CPA against the defendants if asked because of “the social interest and the nature of the impacts that has been caused by the Dam collapse”. In his second report at para. [371] he addressed the evidence of Professor Rosa and Dr Janot on this point by saying that the point was “a very strange one to me”. He went on to recognise that the fact that the Prosecutors had not previously joined the defendants indicated that they did not think it necessary to do so, and that establishing liability against the defendants also seemed unnecessary in view of the fact that full redress was already established in Brazil under TTAC/GTAC. He said that “if, contrary to the above” the MPF thought it was necessary, or was persuaded by the claimants that it was necessary, he had no doubt that it would bring a new CPA against the defendants. He did not address other aspects of the reasoning of Dr Janot and Professor Rosa. This seems to us to be a recognition by him that the MPF is unlikely to be interested in bringing a new CPA against the defendants.

Justice Rezek’s response to this aspect of the claimants’ expert evidence is at para. [120] of his second report. It is that there was an illogicality in the claimants’ experts’ position because they were complaining of the inadequacies of the processes under the umbrella of the 155bn CPA, but at the same time asserting that one of the reasons why the MPF would not bring a new CPA was that he was committed to these processes; and that if the claimants’ complaints were accurate and supported by a claim being brought against the defendants, the MPF would be bound to act. We do not see the illogicality being suggested. The implausibility of the MPF commencing and pursuing a new CPA against the defendants was said to arise out of the balance between disadvantage and reward, where the reward was only the establishment of the additional liability of foreign defendants. There is nothing in this which is inconsistent with the claimants’ criticisms of the current situation.

However, we are not prepared to reject Justice Rezek’s conclusions on a summary basis. We would, therefore, accept that there is a prospect that the MPF might bring a new CPA. However the claimants have established by cogent evidence that this is unlikely.

It is recognised by Professor Didier that the MPF does not generally bring CPA proceedings on behalf of large businesses, or on behalf of municipalities and utilities. However he suggests that municipalities could join as co-plaintiffs, as could the remainder of the 58 by forming an association or associations. This is disputed, but must be treated in these summary applications as a real possibility.

What is still to be addressed is the defendants’ contention that in any event a new CPA or CPAs could be commenced by the 25 claimant municipalities and/or by the individual claimants forming an association or associations, irrespective of a CPA commenced by the MPF.

As to suit by the 25 claimant municipalities, Professor Rosa’s evidence is as follows.

As to the individual claimants forming an association, Professor Rosa explains that:

All this is disputed by Professor Didier, who contends that new CPAs are readily available to all claimants by one or other or both of these routes, and that they could be brought in a single new CPA by claimants with standing joining as co-plaintiffs in a CPA brought by another claimant with standing. Again we cannot resolve these disputes in the evidence.

For all these reasons we must proceed on the basis that there is considerable uncertainty as to whether a new CPA is an available route to these claimants pursuing BHP England or BHP Australia in Brazil. There is a real prospect that it is, but equally there is cogent evidence that there is a real risk that a new CPA is not an available route for any of the claimants. We should add that if it is available in principle by reason of municipality commenced CPAs, such availability depends upon the willingness of the municipalities to do so, and a number of municipalities have made clear that they would not be willing to commence CPAs against the defendants, for a number of the reasons which Professor Rosa and Dr Janot identify as making such a course unattractive for the MPF.

This is to some extent bound up with the dispute about the resumption of the 155bn CPA. Clearly there can be no consolidation if the 155bn CPA will not resume (whether because settlement discussions are successfully concluded and result in a final TAC, or because the claimants’ witnesses are right that there is in practice no realistic prospect of it resuming). Equally, if the claimants’ evidence is right that even if resumed it will not address the legal liability of the Brazilian Companies, there seems little basis for any consolidation. Moreover it is Professor Rosa’s opinion that if there were any new CPAs, and if they were consolidated in the 12^th Federal Court, it is likely that they would not be heard with the 155bn CPA but simply stayed to await the outcome of the 155 bn CPA. Professor Didier’s view, on the other hand, is that a new CPA or CPAs can and would be consolidated with the resumed 155bn CPA in the sense that they would be heard and determined at the same time by Judge Mario.

Professor Rosa expresses the view that liquidation and enforcement proceedings in Brazil following a successful new CPA would face additional complications because they would involve foreign defendants. Although territorial jurisdiction would exist, there would be delays in service and it is extremely unlikely that it would be practicable for individuals to enforce judgments against foreign defendants. This is, in Professor Rosa’s opinion, a further reason why it is unlikely that a new CPA would be brought on behalf of individual claimants, whether by a municipality or an association.

Towards the end of the hearing before us, the defendants produced a closely typed four page document replete with references to material in the trial and (in part) appeal bundles headed “D’s submissions on the New CPA”, to which the claimants responded in writing with points supported by yet further detailed references to documents. We are grateful to the parties for seeking to assist us with the mass of evidential material relevant to the points we have just been addressing. However a consideration of that material simply serves to reinforce our conclusion that there are divisions in the evidence on these points which it is both inappropriate and impossible to seek to resolve on these applications.

There have been a very few, perhaps a dozen, claims brought in the state courts against BHP England and BHP Australia, but not by any of the claimants in this action. The evidence suggests that they were all brought by individuals and for the purposes of protecting limitation.

Both Justice Rezek and Professor Didier state that if claims were to be brought against the defendants in Brazil the “obvious” or “most rational” way to do so would be through a CPA. Justice Rezek does not suggest the bringing of individual claims as an alternative. Professor Didier mentions it as an alternative possibility, but does not say more than that. In our view they are both implicitly recognising the disadvantages and limitations in seeking to advance the claims through individual lawsuits in a case of this kind. Indeed the problems in doing so are the very reason for a CPA procedure being established in Brazil in the first place, and the limitations of the Brazilian system in handling such a multiplicity of claims is eloquently attested to in the Baixo Guandu Judgment.

Professor Rosa’s opinion is that it would be impractical for the individual claimants to pursue claims in this way due to the high cost and complication of the evidential phase, which would be exacerbated by having foreign defendants, and addressing in part events abroad, which would all result in extensive delay. The small claims courts are not equipped to deal with such cases, and pursuing them in the ordinary courts would be impractical.

A number of Professor Rosa’s opinions about the difficulties involved in suits against foreign defendants (and in a new CPA) are supported in the witness statements of the Attorneys-General or mayors of Mariana, Ouro Preto, Rio Doce and Barra Longa.

Professor Didier disagrees with some of Professor Rosa’s views set out above, particularly on the significance of having foreign defendants. Again we cannot resolve those differences on these applications and must proceed on the basis that there is a real possibility that Professor Rosa’s views are to be preferred and equally a possibility that Professor Didier is right. However, even taking the views of Professor Didier and Justice Rezek at their highest, they do not advocate pursuit of the defendants in the local courts as an attractive alternative to a new CPA.

In paras. [63] and [213] of the Judgment, the Judge referred to the fact that the claimants had made clear that they had no intention of commencing proceedings in Brazil against BHP England or BHP Brazil, whatever the outcome of the applications before him. This was the position adopted before him, Mr Hollander having said that that was “not going to happen”. The claimants have not resiled from that position on the appeal.

The claimants have not sought to rely on issues of limitation as affecting the outcome of these applications, and limitation issues have not therefore been directly addressed in the evidence. It appears that the primary Brazilian limitation period for these claims is three years. We understand that the Claimants’ case will be that this primary limitation period does not apply to claims in respect of environmental damage. The evidence does suggest that there will be limitation issues raised by the defendants in this action, and Mr Gibson confirmed that to be so. As to the position in Brazil, there is no evidence that any litigant has faced a limitation defence hitherto, but the stance which might be taken by Samarco or Renova towards claimants in this action who were to commence a claim in Brazil if this action were stayed or struck out is not known.

The defendants have offered an undertaking that they will not take any limitation defence which is not already available in this action, in any proceedings brought in Brazil against BHP England or BHP Australia, provided that they are commenced within one year of the final disposal of these applications.

The court always has at its disposal the general case management powers under the overriding objective in CPR 1.1 and in CPR 3.1. It is to deal with cases justly and at proportionate cost, which includes ensuring that cases are dealt with expeditiously and fairly with an allotment of an appropriate share of the court’s resources; it is also to manage the proceedings by encouraging the co-operation of the parties. It has available to it a very wide and flexible array of case management tools, including for example the trial of preliminary issues, and the use of lead cases. These already existed under the old Rules of the Supreme Court, and have been deployed to manage litigation of considerable complexity involving very large numbers of parties and issues, as, for example, in the litigation arising out of the losses at Lloyd’s in the late 1980s and 1990s: see Deeny v Gooda Walker Ltd [1994] CLC 1224 and the many related cases.

These general case management powers apply equally in the Technology and Construction Court (“TCC”) (see CPR 60.3). The TCC is well-known for its ability robustly and actively to case manage complex litigation, including group litigation. The TCC Guide emphasises the importance of identifying the real issues, a realistic timetable, proper disclosure, costs management and party co-operation (see generally paras. [5.1.1] to [5.1.6]). The court will consider proactively the need for trials of preliminary issues or a staged trial (see paras. [5.4.1], [8.1.3] and [8.1.6]).

Group litigation orders (“GLOs”) are governed specifically by CPR 19.10 to 19.15 (supplemented by Practice Direction 19B) which grant the court even wider powers. These rules came into force on 2 May 2000. They reflect the recommendations of the Final Access to Justice Report (July 1996) and were designed to meet the following objectives:

The rules establish a framework for the case management of claims “which give rise to common or related issues of fact or law” (see CPR 19.10 and 19.11). In those circumstances, the court has power to make a GLO. This enables the court to manage the claims in a co-ordinated way, including via the establishment of a group register (with or without cut-off dates for entry). Judgments, orders and court directions are binding on all claims within the GLO (see CPR 19.12). The court again has extensive and flexible case management powers, including (under CPR 19.13) the use of test claims.

The parties have a duty to co-operate at all stages of the process. Co-operation is an integral part of CPR 1.4(2)(a) and the parties have an express duty under CPR 1.3 to assist the court in furthering the overriding objective. As was stated by Fraser J in Lungowe v Vedanta Resources plc [2016] EWHC 975 (TCC), [2016] BCC 774 (see para. [42]), co-operation in group litigation is of the utmost importance, given the particular challenges that group litigation presents both to the court and the parties.

The courts have developed a wide range of case management tools in group litigation including, importantly, the selection of lead cases, the trial of preliminary issues and the adoption of a staged approach, either in parallel with other progress in the litigation or as a stand-alone procedure. These operate in what is now a digitalised environment which includes sophisticated e-disclosure, data sampling and algorithm mechanisms.

As Coulson LJ commented in Jalla v Shell International Trading and Shipping Co Ltd [2021] EWCA Civ 1389 (at para. [49]), GLOs are used regularly in environmental cases. He also identified (at para. [50]) another methodology, also common in TCC cases, namely for all claimants to be parties to the action in their own name, with a representative sample then being taken as lead claimants for the purpose of deciding the substantive issues. Careful sampling by both sides ensures that the claimants as a whole can take advantage of those matters on which they have a genuine common interest, whilst also allowing decisions on the range of individual points and defences that may arise, some affecting one group of claimants, some another. The results are then extrapolated across all the claimants.

The claimants placed before the Judge the following illustrations of GLO deployment.

This list is by no means exhaustive. Another useful recent working example of a GLO can be found in Crossley v Volkswagen Aktiengesellschaft [2019] EWHC 698 (QB). In that litigation 117,000 claimant (present or past) car owners claim damages from car manufacturers arising out of diesel emissions. There are some 1.2million affected vehicles in this jurisdiction. At the first case management conference, Waksman J had to decide, amongst other things, on the broad shape of the litigation going forward. At para. [26] he identified that there was no reason in principle not to order trial of a preliminary issue at an early stage, as well as providing for the ultimate trial of selected cases. That is what in due course he ordered, namely that there should be a staged approach, involving preliminary issues on two issues (essentially of law), but also that the selection of lead cases should progress at the same time.

The Judge’s finding that the claims amounted to an abuse of process was not the exercise of a discretion. Rather it was an assessment in respect of which there could only be one correct answer (as to whether there was or was not an abuse of process). Whilst an appellate court will be reluctant to interfere with the decision of a judge where the decision rests upon balancing a large number of factors, the question for this court is whether or not the Judge reached the right answer (see Aldi Stores Ltd v WSP Group plc [2007] EWCA Civ 1260, [2008] 1 WLR 748 at para. [16]). The court can interfere if it considers the decision to be wrong by reason of some identifiable flaw in the treatment of the question to be decided, such as a gap in logic, a lack of consistency, or a failure to take account of some material factor which undermines the cogency of the conclusion (see for example Re Sprintroom [2019] EWCA Civ 932, [2019] BCC 1031 at para. [76]).

It is common ground that, in the event that the Judge’s decision to strike out (or stay) for abuse of process is set aside, we should proceed to determine the question afresh (and not remit the matter). It is also common ground that, for the purpose of assessing the correctness of that decision, the evidential position is to be judged as at the time of the Judgment. However, in so far as the decision falls to be made afresh, the court can (as a matter of principle) take into account post-judgment events (although there is a dispute between the parties as to whether the claimants should be permitted to rely on evidence concerning post-judgment events relating to the judicial re-organisation of Samarco (“the Samarco JR”)).

Given the nature of the issues arising on the appeal against the Judge’s finding of abuse of process, it is necessary to rehearse the structure and contents of the relevant parts of the Judgment in a little detail.

The Judge first considered the law relevant to abuse of process. Amongst other authorities, he cited Hunter v Chief Constable of the West Midlands Police [1981] UKHL 13, [1982] AC 529 (“Hunter”); Her Majesty’s Attorney General v Barker [2000] EWHC 453 (Admin), [2000] 1 FLR 759 (“Barker”); Johnson v Gore Wood & Co [2000] UKHL 65, [2002] 2 AC 1 (“Johnson”) (citing Henderson v Henderson [1843] 3 Hare 100 (“Henderson”)); Dexter Ltd v Vlieland-Boddy [2003] EWCA Civ 14 (“Dexter”); and AB v John Wyeth & Brother (no 4) [1994] PIQR 109 (“Wyeth”). He stated that claims involving very considerable numbers of parties and issues inevitably placed a very much greater burden on the court than would be assumed in a unitary action. It was his view that in such cases “considerations of the allocation of court resources and the procedural practicability of accommodating the ambitions of the parties [were] liable to come more strongly into play”. In his judgement, the claimants here were in a similar position to that of the plaintiffs in Wyeth (at para. [63]):

“…Where, as in [Wyeth], the choice of defendant brings no benefit to a claimant but the pursuance of a claim against such a defendant would result in the oppression of that defendant and/or would take a disproportionate toll on the court’s resources, the court is entitled to intervene…”

He then proceeded to examine the various factors that he considered relevant to abuse, recognising that ultimately it would be necessary to stand back and take a broad view. Under “Practicability of managing the claims in England”, the Judge drew attention to the difficulty caused by the progress of parallel proceedings in Brazil and observed that “how the English court would be able to cope, if at all, with the problems likely to be generated by the simultaneous progress of its Brazilian counterpart is an issue which warrants particular scrutiny”.

Under “Irreconcilable Judgments, Collateral Attack and Cross-Contamination of Issues”, the Judge carried out that “particular scrutiny”. He started with an examination of what he described, referring to Henderson, as “multiplicity of litigation”. Referring to Owusu v Jackson (Case C-281/02) [2005] QB 801 (“Owusu”), he accepted that it would be impermissible to deploy an abuse of process argument in order to achieve through the back door that which Brussels Recast barred through the front. However, that did not, in his judgement, mean that the risk of irreconcilable judgments, or considerations under the doctrine of forum non conveniens, should be ignored. He relied on Lungowe v Vedanta Resources plc at first instance and on appeal in the Supreme Court [2019] UKSC 20, [2020] AC 1045 (“Vedanta”).

At para. [86] he stated:

“Having closely considered the evidence relating to the issues likely to be adjudicated upon in both Brazil and England, I am satisfied that the risk of inconsistent judgments would be acute in this case in the event that these proceedings were permitted to go ahead in England.”

He identified the alleged status of BHP Brazil as an indirect polluter as an important example of this. As already indicated above, we do not agree with his conclusions in this paragraph. Further, he described the prospect of attempting to manage the claims of over 200,000 claimants, where such a high proportion of them were taking (or had taken, or reserved the right to take) steps to achieve compensation in Brazil for the same losses as claimed in England, as “nothing short of alarming”. His self-coined phrase of “cross-contamination” referred to the risk of decisions taken in the Brazilian litigation, both procedural and substantive, undermining or otherwise affecting decisions taken in the English litigation. He predicted that the task facing the managing judge in England would be “akin to trying to build a house of cards in a wind tunnel”. He rejected the suggestion that there could be imposed a condition that claimants relinquish proceedings in Brazil in order to be allowed to proceed in England; many had already recovered some level of damages in Brazil, including for the same claims, and they had been assured by their solicitors that they would remain entitled to pursue their remedies in Brazil alongside the English proceedings. He did not find it necessary to rule on the submission of abuse by reason of collateral attack; rather it was the practical implications of parallel proceedings that mattered.

The Judge went on to address what were in his view “further challenges”, “if any such were needed”. The level and rate of turnover of claimants would be likely to be unmanageable. The rule entitling a claimant to choose whom to sue in the event that more than one party may be liable in respect of the same losses was not absolute. The court was concerned to ensure that judicial and court resources are appropriately and proportionately used in accordance with the requirements of justice (per Wyeth at para. [54]). The claimants had understated the extent to which the impact on the court was a relevant factor. He remarked on “how thin” were the claimants’ suggestions for further case management. As for settlement, it would be wrong to permit an “abusively unmanageable claim to proceed in the hope that its abusive unmanageability would catalyse a compromise”. Even if the claimants had provided further procedural assistance, he was satisfied that “such detail would have fared badly under scrutiny”.

He concluded at para. [104]:

“In all the circumstances, I am entirely satisfied that these claims would be not merely challenging but irredeemably unmanageable if allowed to proceed any further in this jurisdiction.”

The Judge continued that even if, contrary to his findings, that were not the case, the proceedings would place a “very significantly deleterious impact indeed upon the scarce resources of the English courts”. His principal reason for that conclusion was that there would be “an immense pool of claimants with grossly disparate interests”, requiring a huge number of lead cases. Continuing developments in Brazil would mean that there was a real prospect of “almost interminable transatlantic iteration”. Mitigatory steps would achieve very little if anything. He further identified various other practical problems, including of language, translation, claimant and witness location and the need for expert evidence on Brazilian law. The Judge stated that, moreover, the claimants would not make full recovery in the English proceedings because of their agreement to pay a success fee of up to 30% to their solicitors.

He concluded at para. [120]:

“It follows that I am satisfied that it has been clearly proved that these claims amount to an abuse of the process of the court. In the words of Lord Bingham in Barker, they amount to ‘a use of the court process for a purpose or in a way which is significantly different from the ordinary and proper use of the court process’.”

Under “Full Redress”, the Judge addressed the question of whether or not the claimants would or would not get full redress in Brazil, an issue hotly in dispute between the parties. However, he stated on more than one occasion that it would be inappropriate for him to adjudicate on the detail of the matters contested, and he did not reach any conclusion as to the likelihood or otherwise of full redress for the claimants in Brazil. Rather, he had two points of “overarching significance”:

Nevertheless, the Judge did consider it permissible for him to make “a number of observations” based on what he described as the “undue pessimism” on the part of the claimants, or their advisers, as to the prospects of achieving full redress in Brazil. Among other things, he noted that by the end of 2019, 27,000 claims (of victims generally, not just the claimants here) had been adjudicated on in the state courts of Minas Gerais and that just under half of the claimants here had received payments from Renova. A considerable number of individual claims in Brazil had been stayed, but that was because they related to compensation for interruption to water supply and there was a pending appeal on a point of principle. No fewer than 192,000 of the claimants were bringing claims in respect of water interruption in this action.

Under “Renova”, the Judge acknowledged Judge Mario’s concerns over the delay and bureaucracy in the Renova processes, and the heavy criticism made by various witnesses. He went on, “nevertheless”, to make a series of counter-points, including a recognition of the vast nature of the task of compensation and the need to strike a balance between providing compensation which accurately reflects losses without overburdening the claimants with bureaucratic demands for documentary proof and disproportionate form-filling. He dismissed the dispute as to whether the categories of claim falling within the scope of the Renova programme covered every potential category of legitimate claim as being “more apparent than real”. He then observed, consistently with his earlier findings, that permitting the claimants to proceed in England “would not provide a panacea” and “would … generate even greater challenges”. Even if liability were established, the claimants would be embroiled in what would be a GLO “the management of which would almost certainly be fatally impracticable…and which would foul the progress of parallel proceedings in Brazil”.

The Judge was not attracted by the claimants’ “fallback” submission that the proceedings in England should be permitted to proceed at least for the 58 institutional claimants unable to benefit from Renova or the two CPAs. He acknowledged the very likely higher average potential value of these claims, but to allow them to proceed would give rise to the acute risk of irreconcilable judgments and conflicting developments in the parallel jurisdictions. In any event, the 58 claimants were not precluded from bringing proceedings on their own account in Brazil outside the CPAs.

Under “Stepping Back”, the Judge recognised the challenges facing claimants in Brazil and the genuineness of the witnesses’ concerns. However, he was “entirely satisfied that their confidence that anything of value is to be achieved in England is illusory”.

He went on to conclude that it was difficult to conceive of any way to exercise his discretion other than to strike out the proceedings: they were a clear abuse of process, in particular the claimants’ “tactical decision to progress closely related damages claims in the Brazilian and English jurisdictions simultaneously [was] an initiative the consequences of which, if unchecked, would foist upon the English courts the largest white elephant in the history of group actions”. It would also be “manifestly unfair” to the defendants to be required to engage in massively expensive and protracted litigation devoid of any realistic promise of substantial advantage to the claimants”. He was not attracted to the option of a stay pending resolution of the 155bn CPA.

What follows is a summary only of the parties’ respective positions as they crystallised in the hearing before us.

The claimants challenge the Judge’s conclusion on abuse of process as follows.

The defendants’ position is that the Judge’s evaluative assessment of the prospect of the claimants achieving a tangible benefit in the English courts which outweighed the disadvantages to the court and the parties was unimpeachable. It was clearly and obviously correct. The Judge was entitled to find the English proceedings to be abusive in circumstances where the claimants had no realistic prospect of achieving better redress against the defendants than was already available in Brazil. Thus, for example, individualised quantum assessments are available now in Brazil, the outcome of which will be full redress under Brazilian law. Around 67,000 of the claimants had exercised the right to bring claims in the small claims courts and state courts. Of those whose claims had concluded (c. 20,000), there was no evidence that any of those claimants had experienced difficulty in obtaining legal representation or a fair hearing, or that they did not recover full redress. Against that, the claimants in the English proceedings face major group action in order to establish the liability necessary - then, and only then, to progress to an individualised assessment.

The defendants emphasise the fact that in Brazil Renova and Samarco “accept and do not deny their obligation to make full redress in accordance with Brazilian law”, both through the Renova programme and in the courts (save in respect of the 13 large businesses, in respect of which liability is denied). In terms of assessing the separate positions of individual claimants, the defendants’ position is that the Judge was entitled to take a “group” approach and make a “group” strike-out, just as the court did in Wyeth. In any event, for the purpose of the appeal, Mr de Freitas, Renova’s chief executive officer, has provided a witness statement stating, amongst other things, that if any of the claimants in England brought a claim in Brazil against Renova for compensation for losses suffered as a result of the dam collapse, Renova would not argue that it is not obliged to make full redress to them.

The defendants submit that there are effective mechanisms through which the obligation to make full redress is discharged in practice. The Brazilian justice system has spent huge judicial and stakeholder resources ensuring that there are effective and accessible routes to redress, both on a group and individualised level, and working in an iterative process. It is said that a CPA in Brazil is as, if not more, sophisticated than group litigation in the English courts. The 155bn CPA is being managed by a dynamic and innovative judge in the person of Judge Mario. The defendants point to the scale of any liability trial in the English proceedings and the similarities between the operation of a GLO and the 155bn CPA. It is said that the extent of overlap in the pleaded issues and benefits arising is significant.

The defendants point to the Renova structure, involving programmes which are the product of “collaboration, consensus and ongoing improvement”. Prosecution experts have been appointed within it. The defendants rely on the innovative establishment of Priority Axis 7, as well as the institution of local commission proceedings and the Novel System. Importantly, if claimants are dissatisfied with these procedures or proceedings, they can always go to the small claims court, for which purpose there is free legal aid. It is notable, submit the defendants, that the claimants wish to preserve the ability to participate in these procedures or proceedings. In so far as there is evidence of dissatisfaction with Brazilian processes, there will always be scope for complaint in the inevitable complexity of a scheme dealing with damage on this scale. In so far as necessary, they invite us to elevate what the Judge described as only his “observations” in relation to the compensation opportunities for the claimants in Brazil into formal findings.

The defendants resist the claimants’ application to adduce fresh evidence relating to the Samarco JR on essentially two grounds. First, the claimants elected before the Judge not to pursue any reliance on any alleged inability to pay on the part of Samarco. It is an abuse to seek to go behind that election (see Khetani v Kanbi [2006] EWCA Civ 1621 at para. [40]). Secondly, in any event, the fresh evidence is irrelevant. Mr de Freitas has stated that in the event that a Brazilian court finally determines (in proceedings against Samarco and/or Renova) that any plaintiff has suffered recoverable loss from the dam collapse and is entitled to indemnification, Renova will make full payment in satisfaction of the judgment. That statement is consistent with what has happened in practice to date.

The defendants withdrew their reliance on Henderson principles or collateral attack as self-standing grounds of abuse of process. Such considerations, however, were said to remain relevant to an analysis of the burdens and disadvantages of litigation in the English courts. The underlying principles confirmed the strong public interest in the finality of litigation and the avoidance of multiplicity of litigation arising out of the same subject-matter and preventing attacks on previous judicial decisions.

In summary, the defendants contend that the English proceedings are pointless, which is in itself enough to render them abusive; they are also wasteful, in the sense of being disproportionately burdensome. In the event that there were concerns as to the future (for example in terms of commitment on the part of Renova to stand by Mr de Freitas’ assurances; or on the part of Samarco to continue to meet its obligations; or in respect of the possibility of limitation defences being raised in Brazil) the defendants press for a stay in the alternative.

The court has power to strike out a valid claim where there is an abuse of process under CPR 3.4(2)(b).

The classic exposition of an abuse of process remains that of Lord Diplock in Hunter (at p. 536C). An abuse involves:

“… misuse of [the court’s] procedure in a way which, although not inconsistent with the literal application of its procedural rules, would nevertheless be manifestly unfair to a party to litigation before it, or would otherwise bring the administration of justice into disrepute among right-thinking people.”

The categories of abuse of process are varied and not closed; the courts have declined to define or circumscribe the circumstances in which an abuse may be established. There are nevertheless now certain well-established categories of abuse, of which three are material for our purposes.

First, attempts to relitigate issues which were raised, or which could and should have been raised, in previous proceedings may be abusive even if they are not strictly res judicata. Although this form of abuse was first clearly recognised in Henderson, it is now most authoritatively analysed in Johnson: see also Dexter at paras. [49]-[53]. In Johnson (at p. 31D) Lord Bingham stated that whether or not a claim is abusive in the Henderson sense involves a broad, merits-based judgment which takes account of the public and private interests involved and also takes account of all the facts of the case, focussing attention on the crucial question whether, in all the circumstances, a party is misusing or abusing the process of the court by seeking to raise before it the issues which could have been raised before.

Secondly, collateral attacks on earlier decisions may be abusive (see Hunter at p. 541B).

Thirdly, proceedings may also be abusive if, even though they raise an arguable cause of action, they are (objectively) pointless and wasteful, in the sense that the benefits to the claimants from success were likely to be extremely modest and the costs to the defendants in defending the claims wholly disproportionate to that benefit (see Wyeth at pp. 114 and 115; and Jameel (Yousef) v Dow Jones Co Inc [2005] EWCA Civ 75, [2005] QB 946 (“Jameel”) at para. [69]). In Jameel it was held that the benefit attainable by a claimant was of small value and the costs of the litigation would be out of all proportion to what could be achieved, such that “the game [was] not worth the candle” (see para. [69]). There, at para. [57], Lord Phillips MR cited with approval the formulation of Eady J in Schellenberg v British Broadcasting Corporation [1999] EWHC 851 (QB) (at p. 319). The question in each case was whether:

“… there is any realistic prospect of a trial yielding any tangible or legitimate advantage such as to outweigh the disadvantages for the parties in terms of expense, and the wider public in terms of court resources.”

The point being captured was that, while the court must provide a remedy in a case that requires one, the process of the court should not be used in a case where the need has gone away (see Cammish v Hughes [2012] EWCA Civ 1655, [2013] EMLR 13 (CA) at paras. [55] and [56]). We would add that although in [54] Lord Phillips referred to the concern of the court to “ensure that judicial and court resources are appropriately and proportionately used”, the fact that proceedings may place a very heavy burden on the court’s resources cannot constitute a ground of abuse by itself.

Where multiple claims are brought by different claimants who do not stand in materially the same position, it is necessary to consider the question of abuse by reference to claims individually (or by relevant claimant category). Abusive factors applicable only to one claimant do not render another co-claimant’s claim abusive. We treat it as axiomatic that a claim brought by one claimant, which is not itself abusive, cannot become abusive merely because other claimants have chosen to bring abusive claims. The claimants should be in no different position, so far as an abuse argument is concerned, from that if each had brought separate proceedings, whether or not other claimants also brought proceedings. An individual approach is required. The court must be satisfied in relation to every claim, having regard to any differences between claimants or categories of claimant, that it is abusive and a strike-out or stay appropriate.

A finding of abuse of process does not lead automatically to a striking out of the claim. The court then retains a discretion as to the appropriate response, which must always be proportionate (see for example Cable v Liverpool Victoria Insurance Co Ltd [2020] EWCA Civ 1015 at paras. [63] and [64]).

Finally, but importantly for present purposes, litigants should not be deprived of their claims without scrupulous examination of all the circumstances and unless the abuse has been sufficiently clearly established: “the court cannot be affronted if the case has not been satisfactorily proved” (see Alpha Rocks Solicitors v Alade [2015] EWCA Civ 685, [2015] 1 WLR 4535 at para. [24]; and also Summers v Fairclough Homes Ltd [2012] UKSC 26, [2012] 1 WLR 2004 at para. [48]). Thus it has been stated repeatedly that it is only in “clear and obvious” cases that it will be appropriate to strike out proceedings as an abuse of process so as to prevent a claimant from bringing an apparently proper cause of action to trial (see for example Wallis v Valentine [2002] EWCA Civ 1034 at para. [31], approving the dicta of Simon Brown LJ in Broxton v McCelland [1995] EMLR 485 at pp. 497-498); JSC BTA Bank v Ablyazov [2011] EWHC 1136 at para. [10]; Optaglio Ltd v Thomas Tethal [2015] EWCA Civ 1002 at para. [63]).

For the reasons discussed below, we have reached the conclusion that the Judge’s decision to strike out, alternatively stay, the proceedings for abuse of process was flawed in a number of respects and wrong. In particular:

Accordingly, it will fall to us to decide afresh whether the claims should be struck out (or stayed) for abuse of process, and in particular whether the defendants have discharged the burden of establishing that the proceedings are clearly and obviously pointless and wasteful.

Unmanageability was not the basis of the defendants’ application before the Judge. Their application was based on the contention that the litigation was abusive for being pointless and wasteful (and/or by reference to Henderson principles and/or collateral attack). The evidence relied on in support of the application asserted that the claimants were already entitled to full redress for damage suffered as a result of the dam collapse under TTAC/GTAC. It was said that the most that the claimants could achieve after many years of litigation in England was exactly the same as that to which they were entitled in Brazil. The litigation in England would impose a massive burden on the parties and the court system, and there was no benefit to the claimants in bringing their claims in England, let alone one which could justify the cost.

However, it is clear that the Judge found the proceedings to be abusive, first and foremost, for “irredeemable unmanageability”. This was his key conclusion; whilst he later went on to “step … back”, unmanageability was, in his judgement, a sufficient basis for a finding of abuse without more. Thus, at para. [102], he referred to the proceedings as being “abusively” unmanageable, stating at paras. [104] and [105], that he was “entirely satisfied that these claims would be not merely challenging but irredeemably unmanageable if allowed to proceed”. The proceedings were “not practically workable”, and even if they were, would still have “a very significantly deleterious impact indeed upon the scarce resources of the English courts”. His conclusion at para. [120], namely that it had clearly been proved that the claims amounted to an abuse of the process of the court, was reached before and independently of any consideration of the question of the availability of full redress in Brazil. As he stated in terms when refusing permission to appeal, his finding of irredeemable unmanageability was “the point of central importance to which all other considerations [were] of secondary significance”.

Against that background, two questions arise:

As to the first question, Mr Gibson, at least orally, did not seek to defend the Judge’s finding of abuse by reference to unmanageability. He was right not to do so. We have considerable doubts as to whether proceedings can ever truly be said to be “unmanageable”. But putting that question to one side, unmanageability does not fall within any of the abusive mischiefs identified in the authorities. It does not amount to a misuse of the court process in a manner that would be manifestly unfair to the parties, nor would it bring the administration of justice into disrepute among right-thinking people. The Judge (at para. [120]) quoted Lord Bingham’s description of abuse in Barker (at para. [19]): “a use of the court process for a purpose or in a way which is significantly different from the ordinary and proper use of the court process”. As noted at para. [33] of the PTA Judgment, this was not an apt description for the question of unmanageability upon which he was relying. Equally, as identified above, the mere fact that the litigation would place a significant burden on the courts could not be an independent basis for a finding of abuse.

Here the claimants are bringing properly arguable claims against the defendants in this jurisdiction as of right. Even if the proceedings were unmanageable due to complications arising out of parallel proceedings in Brazil, or because of other procedural complexities, that would not mean that the court process was being misused, whether vexatiously, oppressively or otherwise. It would mean only that the court processes were not capable of meeting the challenge posed by the proceedings. The remedy might be a procedural stay as part of an exercise of case management powers; but the proceedings would not be abusive.

Such an approach is entirely consistent with the decision in Mastercard v Merricks [2020] UKSC 51, [2021] 3 All ER 285, [2021] 1 Bus LR 25, in the context of collective proceedings. At para. [74] Lord Briggs stated:

“The incompleteness of data and the difficulties of interpreting what survives are frequent problems with which the civil courts and tribunals wrestle on a daily basis. The likely cost and burden of disclosure may well require skilled case-management. But neither justifies the denial of practicable access to justice to a litigant or class of litigants who have a triable cause or action, merely because it will make quantification of their loss very difficult and expensive…”

For the avoidance of doubt, we would not rule out a finding of abuse of process as a matter of principle by reference to unmanageability if, for example, a claimant had deliberately made the litigation unmanageable with vexatious consequences for the defendant, thus misusing the court process. But there is no suggestion of that here, nor could there be. The MPOC is well-structured, coherent and entirely digestible; it clearly set out the facts relied upon, the causes of action and the relief claimed.

As to the second question, the Judge’s conclusion that the proceedings here were (clearly and obviously) “irredeemably unmanageable” is not sustainable. In the first place, no such conclusion could be reached safely at such an early stage of the proceedings. The Judge’s criticisms of the claimants’ position (at paras. [101]-[103]) are unfair. The claimants provided the court with clear illustrations of case management options. It was unreasonable to expect more definitive proposals at such an early stage of proceedings. The precise nature and scope of the issues between the parties had yet to be identified. The proper time for considering how to manage would be at a case management conference before the assigned judge. Significantly, by that stage the parties, represented by sophisticated litigators experienced in the field, would have been obliged fully to co-operate in putting forward case management proposals. Those could have included, for example, a proposal that the issue of indirect polluter strict liability be tried preliminarily, or that the control basis of such liability be determined first. A trial of such a preliminary issue would not appear to require the involvement of any evidence from any individual claimants; rather it would turn on the defendants’ knowledge, control and conduct, and the scope of the dispute on these aspects of the claim would be apparent from the defendants’ responsive pleading to the MPOC. Were liability to be established, generic causation and quantum issues could be identified and tried. The outcomes of quantum trials on individual selected test cases could be used to inform settlement of individual cases across the board.

Lloyd v Google LLC [2021] UKSC 50; [2021] 3 WLR 1268, a case involving a class representative action, does not assist the defendants. There (at paras. [24]-[83]) Lord Leggatt performed a high level overview of the different methods of redress available in group actions, collective proceedings and representative actions. At para. [28] he identified that GLO proceedings, which are “opt-in”, required quantum of loss to be proved in each individual case. Where the loss in each case was small, eligible individuals might be less likely to opt in, and the litigation might be rendered uneconomic. But he did not suggest that the use of group litigation for claimants with low-value claims who had opted in was in some way unmanageable, let alone abusively so. And of course, a number of the claimants in these proceedings have high-value claims.

The Judge was influenced significantly by what he considered to be the complications arising out of the existence of parallel proceedings in Brazil, and what he described to be an “acute” risk of “unremitting cross-contamination” of proceedings. There is a question, addressed below, as to whether the risk of inconsistent judgments was a legitimate consideration at all. Beyond that, the Judge identified the complications and the risk of cross-contamination in general terms. He did not, at least on the face of the Judgment, carry out the scrupulous analysis necessary to found a conclusion that such potential complications clearly and obviously rendered the proceedings unmanageable.

There is force in the claimants’ submission that a proper analysis, including by reference to the positions of the different classes of claimants, does not bear out his findings.

The position of the 58 is particularly striking when considering unmanageability. The fact that they are only a very small proportion of the total number of claimants – 0.03% – is not to the point; their claims are financially significant, and in any event, as we expand upon below, an individual approach to claimants is required. The terms of TTAC exclude their redress from its scope. It is clear that the 13 large businesses included within the 58 could not benefit from any generic sentence in the 155bn CPA. It is uncertain whether any of the remaining 58 could in principle benefit from any such sentence, and clear that for the most part they could not do so because the MPF is not seeking a generic sentence which would cover large parts of their losses. Broadly speaking, therefore, there is no risk of a conflicting 155bn CPA judgment so far as they are concerned.

The Judge nevertheless found that to allow the claims of the 58 to proceed in England would give rise to an acute risk of irreconcilable judgments and “in a broader sense, conflicting developments in the parallel jurisdictions”. He stated that “many of the municipalities and utility companies stand to benefit from the Renova programmes of infrastructure and environmental works”. It is difficult to see how this potential benefit of remedial works would create an abusive risk of irreconcilable judgments or conflicting developments. He also stated that the 58 could bring claims outside the scope of the CPAs. But the fact remains that whilst a few have done so, most have not, and the English court could not require them to do so, as we discuss further below.

In summary, the risk of unmanageability, or as the Judge put it “utter chaos”, due to the existence of proceedings in Brazil is not clear and obvious, whether by reference to the claimants as a whole, or, as is in fact required, considering their position on an individual basis. As the claimants rightly recognise, it may be that “down the line” some individual claims may need to be reviewed in the light of any releases signed by particular claimants, or possibly by reference to Henderson principles. It is not possible on the evidence before us to identify with any particularity whether such issues will arise, and if they do, by reference to which claimants. If they do, they can be addressed individually, or by category of claimant. The potential for such issues arising in due course cannot be said to make them abusive on the grounds that the claims are unmanageable.

For these reasons, there was no proper basis for the Judge’s finding that the proceedings were abusive on the basis of irredeemable manageability, both as a matter of principle and on the facts.

Before considering the question of the availability of full redress in Brazil, the Judge took into account both the risk of inconsistent judgments (as set out above) and also what he described as “the challenge of language”, with the claimants and many of the potential witnesses for both sides speaking Portuguese as their first or only language; translation costs and difficulties; the need to apply Brazilian law; and the unlikelihood of claimants or witnesses being permitted to give evidence remotely from Brazil (as a matter of Brazilian law).

As already indicated, the burden placed by the litigation on the English courts is insufficient in itself to establish abuse. There may also be cases, of which this is not one, where forum non conveniens factors may provide some evidential support for an argument that the proceedings have been brought for the improper collateral purpose of unfair harassment. However, save to that extent, the risk of inconsistent judgments and the other difficulties identified are matters to be confined to jurisdictional challenges, here under Brussels Recast (so far as BHP England is concerned) or forum non conveniens principles (so far as BHP Australia is concerned).

Under Brussels Recast the courts of a member state have no power to decline jurisdiction over a defendant domiciled and sued in that member state by reference to foreign proceedings, save in the limited circumstances of lis pendens identified in Section 9. Article 34 permits a stay of proceedings in favour of specific related pending proceedings in a non-member state in order to avoid the risk of irreconcilable judgments, in circumstances circumscribed by the conditions it imposes. Where those conditions are not fulfilled, article 4 must be given effect to. To strike out a claim against an English-domiciled defendant as abusive on the ground that the existence of parallel proceedings in a third state would give rise to a risk of irreconcilable judgments infringes the obligation of effectiveness in relation to article 4 and undermines the limited derogation from article 4 for which article 34 provides.

This follows from the decision in Owusu, in which the CJEU held that it was not open to the United Kingdom (or any other contracting state) to bypass the regime imposed by Brussels Recast on the basis that, although the defendant was domiciled here, jurisdiction could be declined by the application of the English common law principle of forum non conveniens. It relied on the mandatory nature of article 4, the principle of legal certainty and the uniform application of the rules of jurisdiction (see paras. [37]-[46]).

The Judge recognised (at para. [81]) that it was “impermissible to deploy an abuse of process argument in order to achieve through the back door that which the Recast Regulation bars through the front” door. But he went on to say that, “nevertheless, in cases in which the risk of irreconcilable judgments is just one of a number of factors relevant to the exercise of the abuse jurisdiction it should not be ignored”.

In reaching this conclusion he relied on Vedanta at first instance at para. [84] in which Coulson J (as he then was) stated that “in an appropriate case”, and notwithstanding Owusu, the court had to be able to exercise its case-management powers to grant a stay; and on Vedanta in the Supreme Court at para. [17], where Lord Briggs, having referred to Owusu, stated:

“This does not, of course, prevent any defendant from seeking to have a claim struck out as an abuse of process or as disclosing no reasonable cause of action, or from seeking reverse summary judgment on the basis that the claim discloses no triable issue against that defendant.”

There are a number of problems with the Judge’s approach. First, it is difficult to see how, as a matter of logic, the question of whether a risk of irreconcilable judgments is the only factor or but one of a number of factors should make a difference to the approach in principle. Secondly, Coulson J’s comments in Vedanta, upon which the Judge relied, concerned the power to impose a case management stay in “an appropriate case” as a matter of “good and sensible case management”, not the power to strike out for abuse of process. Even then, rare and compelling circumstances would be required to justify a stay in favour of a domiciled defendant (see para. [85]).

Two other decisions confirm that forum non conveniens factors cannot amount to such rare and compelling circumstances.

As for Lord Briggs’ comments in Vedanta at para. [17], there is nothing to suggest that he had in mind an application to strike out for abuse of process by reference to jurisdictional/forum non conveniens factors. On the contrary, given that he had just referred to Owusu in the preceding paragraph (and again at para. [39]), it seems clear that he had in mind applications to strike out by legitimate reference to other matters. Had he been departing from Owusu or decisions such as Mazur, he would have said so in terms and with reasons.

Thus, so far as BHP England was concerned, the risk of irreconcilable judgments and other forum non conveniens factors should not have played any part in the Judge’s finding of abuse. Such matters do not provide a proper basis for precluding the claimants from pursuing otherwise arguable claims against an English-domiciled defendant.

Similar considerations apply so far as BHP Australia was concerned. The appropriateness of BHP Australia being sued in the English courts must be judged by reference to the well-established principles laid down by Lord Goff in Spiliada Maritime Corp v Cansulex Ltd (The Spiliada) [1987] AC 460 (“Spiliada”). If the jurisdiction to strike out for abuse of process is to be engaged by reference to forum non conveniens factors, then the operation of those common law principles is impermissibly circumvented. As it was put for the claimants, the question of forum is the logically anterior question: in the event that England is established as the appropriate forum, it would be nonsensical for the proceedings to be struck out as being abusive by reference to the same forum considerations.

It is common ground that proceedings can be abusive on the basis that, objectively assessed, they are (clearly and obviously) pointless and wasteful. It was the defendants’ central contention that the English proceedings were indeed pointless and wasteful – on the grounds that the claimants could and would obtain “full redress” in Brazil. That was the platform for the defendants’ contention that the English proceedings were abusive.

The argument proceeded on the basis that it was a legitimate exercise for the Judge to compare the benefits of litigation in England with the benefits of litigation (or other compensation schemes) in Brazil. His conclusion on unmanageability was key in this context. Having determined that the English proceedings were unmanageable, he was then able to conclude that they were pointless and wasteful adopting the comparative approach: nothing sensible could ever be achieved here. On his analysis, whatever the difficulties facing the claimants in Brazil, the claimants would not be better off in England. Thus, he was not obliged to engage in any detail with those difficulties, and he did not do so.

In the light of our finding on manageability, however, the Judge’s route to a finding of abuse, and his resulting conclusion, must fall away. We must therefore consider the issue afresh.

However before considering the defendants’ argument that adequate redress is available in Brazil, we should make three general points about the context in which it arises.

First, the claimants bring viable claims against the defendants in this jurisdiction as of right. They have brought no claims against the defendants in Brazil. The basis for liability is factually distinct from any claim by any claimant against others in Brazil. The need for particular caution in striking out the English proceedings for being pointless and wasteful is self-evident. In principle, claimants are entitled to choose whom to sue. There may be diverse and legitimate reasons why a claimant may choose to sue a particular defendant or defendants and it is not part of the court’s function to interfere with that process. We would accept Mr Choo Choy’s submission that the claimants in this case could properly see a legitimate advantage in pursuing the action here for the purposes of improving the chance of a settlement: it would involve a public trial of the liability of the parent companies and raise for the defendants’ consideration whether a settlement would be thought advantageous to avoid any adverse reputational damage from findings of liability. A claimant’s unhindered right of access to justice in respect of properly arguable claims is a core constitutional right inherent in the rule of law (see for example R (on the application of Unison) v Lord Chancellor [2017] UKSC 51, [2020] AC 869, at paras. [61]-[85]), as well as being enshrined in article 6 (see for example Summers v Fairclough Homes Ltd [2012] UKSC 26, [2012] 1 WLR 2004 at paras. [46]-[48]). We do not go so far as to say that a claimant has an unfettered right to pursue an arguable claim against their chosen defendant: the Wyeth and Jameel abuse jurisdiction provides an exception to that general principle. Nevertheless, where the Henderson principle is not in play, it will be a rare case in which the court can say that there is no legitimate advantage in pursuing a defendant merely because there exists a claim for the same loss against another person, and especially so when it is advanced on a different basis of liability.

Secondly, the potential redress relied on in this case is in large part to be found in judicial or extrajudicial schemes which are optional, or are concurrent alternative remedies. The existence of the 155bn CPA does not preclude ordinary civil proceedings being commenced and pursued by claimants at the same time; nor do the Renova initiatives; nor does involvement in the Novel System. If the current claims had been commenced and were being pursued in the civil courts in Brazil, against the Brazilian Companies or these defendants, they could not be dismissed as abusive on the grounds that adequate redress was or would be available through the existing processes, whether from Renova pursuant to TTAC/GTAC, or from the Novel System, or from liquidation proceedings following conclusion of the 155bn CPA.

In those circumstances we have considerable doubts whether potentially adequate redress from those sources (assuming it to be such) could make the proceedings in England abusive, although it might be relevant to case management (cf Andrew v Barclays Bank [2012] EWHC B13 (Mercantile), [2012] CTLC 115). However, since this point was not argued before us, we have assumed in the defendants’ favour when addressing this issue that all redress which is potentially available in Brazil is properly to be taken into account.

Thirdly, the nature of the exercise, namely a close comparative assessment of the relative benefits and disadvantages of litigation/compensation schemes in Brazil compared with what is on offer in the English courts, arises in the context of a summary strike out application. Even after eight days of hearing below and five days of hearing on appeal, the English court has only been able to acquire a superficial understanding of the intricacies and hinterland of the legal landscape in Brazil. The exercise has given rise to very extensive factual and expert evidence, as the Judge was at pains to point out, much of which was in dispute. It has been far removed from a debate suitable for summary determination and very different from that carried out, for example, in Wyeth and Jameel. Thus in Wyeth it could be seen “at a glance” that the prescriber defendants would be put to “astronomical” expense in defending what were only alternative, contingent claims in respect of which the relevant claimants stood to gain at best only “extremely modest” benefit and most likely nothing (see p. 116). In Jameel it was disproportionate and an abuse of process for the claimant to proceed with a defamation claim when publication within the jurisdiction was minimal and did not amount to a real and significant tort, the damage to the claimant’s reputation was insignificant and the facts did not justify the grant of an injunction prohibiting further publication. Those cases involved a very different type of futility and wastefulness to that alleged to exist here. As already highlighted, it is only appropriate to strike out for abuse on the grounds that the claims are pointless and wasteful where it is clear and obvious that that is the case.

We have discussed above the position of the 58 in relation to supposed unmanageability. We here address the question whether their claims can be characterised as abusive for being pointless and wasteful.

The Judge’s analysis is, in our view, undermined by his adoption of a global approach, and in particular his failure to recognise the material respects in which the position of the 58 was different from that of the other claimants. Mr Gibson sought to justify this global approach by reliance on Wyeth, in which the relevant claimants were treated as a single cohort. But in Wyeth, unlike the position here, there were no material distinctions between the positions of the claimants in question.

So far as the 13 large businesses are concerned, it is clear on the evidence that they cannot benefit from the 155bn CPA, nor from the Renova or other schemes pursuant to TTAC/GTAC. The same is, broadly speaking, true of the other institutions making up the 58, although some but not all of the municipalities’ claims are arguably within the scope of a 155bn CPA generic sentence. Their ability to benefit from in-kind remedial works, in circumstances in which it is not suggested that they will thereby be fully compensated, has no bearing on whether it is abusive to pursue their financial claims.

Mr De Freitas’ suggestion that Renova will meet such claims affords them no comfort. We have referred earlier to paragraph 17 of Mr De Freitas’ third witness statement which is relied on in relation to the position of the 58 claimants, and explained why the caveat means that it involves no acceptance of liability towards the 13 large businesses. Nor, in our view, does it provide any adequate comfort for the remainder of the 58. The source of Mr de Freitas’ authority to give the assurance is unclear. He is Renova’s Chief Executive Officer, but Renova is subject to complex governance arrangements. Its ultimate decision-making body is a board of (nine) trustees, with decisions requiring at least a majority vote. Samarco, Vale and BHP Brazil have the right to appoint six members. There is also an Advisory Board, an Executive Board and a Fiscal Council. There is no evidence of any formal decision having been taken on the part of Renova (or any of the Brazilian Companies) to support Mr de Freitas’ statement. This absence is significant when it is remembered that Samarco has throughout (as confirmed in the defendants’ first instance skeleton argument) continued to deny any obligation to compensate the 13 large businesses, whose claims are amongst the most substantial in monetary terms. In any event, Mr de Freitas’ statement does not create any legally binding commitment for the future on the part of Renova (or anyone else). As set out above, the future course of events in Brazil is a matter of speculation and uncertainty.

For the 58, therefore, the comparison is, in substance, simply between redress through a civil claim in Brazil against these defendants, which has not occurred, and redress through a civil claim against the defendants here. Had the 58 been the only claimants, we do not consider that the defendants could have begun to mount an application that it was pointless and wasteful to sue them here, where they can be sued as of right, rather than in Brazil. The only possible basis for such an argument would have involved the illegitimate reliance on forum non conveniens factors. It follows that the strike out application against the 58 must fail.

The consequence for the other claimants is that liability issues, and any common issues of causation or quantum, will be decided here in any event (subject of course to the outcome of the article 34 and forum non conveniens applications). There may be an additional liability issue for the 13 large businesses, which does not arise for other claimants, as to whether their losses are recoverable under Brazilian law (based on what Mr Gibson has said on instructions), but there will need to be decided the logically anterior question of whether the defendants are liable as indirect polluters or on the other bases of liability set out in the MPOC, both for the 13 and for the remainder of the 58. This will be determinative of liability to all the claimants. There can therefore be no advantage to the other claimants in any acceptance of liability by Samarco or Renova in Brazil, or in any establishment of liability on the part of the Brazilian Companies in the 155bn CPA. The valid comparison for the other claimants is between the redress they may have in Brazil in respect of determining the quantum of their claims, and the determination of that quantum here. In Brazil that involves individual actions in the civil courts, whether in liquidation proceedings following a generic sentence or ordinary civil claims to the extent that redress is not afforded by the Renova programmes or the Novel System.

Against that background, in our judgement there is and was no safe or proper basis for concluding that the English proceedings are (clearly and obviously) pointless and/or wasteful such as to amount to an abuse of the process of the court for any of the claimants.

On the basis of the material that we have been asked to digest it is apparent, as foreshadowed above, that there are a number of significant and unresolved uncertainties as to the claimants’ ability to achieve full redress in Brazil. This is inherent in GTAC itself, which envisaged ongoing (and currently unconcluded) negotiations as a result of the inadequacy of redress under TTAC. We have identified above, in some detail, areas of dispute or uncertainty including, by way of example only, the following:

The following further matters need to be taken into account:

Further, there is the contentious question of the Samarco JR. The claimants seek permission to rely on fresh evidence relating to Samarco’s petition for JR under Brazilian legislation filed on 9 April 2021. It comes primarily in the form of a short witness statement dated 14 October 2021 (with exhibits) from Mr Marco Deluiggi, one of the claimants’ Brazilian lawyers, together with updating material since then which the parties have agreed (subject to the defendants’ objections as to relevance and admissibility). The petition was accepted by the 2^nd Business and Commercial Court of Belo Horizonte on 12 April 2021. An order was made which, amongst other things, granted a stay of all legal and enforcement actions against Samarco for 180 days, and ordered Samarco to present a JR plan within 60 days. The stay has subsequently been extended.

The purpose of the JR is to restructure Samarco’s indebtedness resulting from the challenging financial circumstances in which it finds itself following the dam’s collapse. The JR process has been far from smooth, including, for example, challenges to the election of the Creditors Committee, appeals by lenders and objections from bondholders. There have been not only internal appeals but also numerous court rulings. There has been a series of evolving JR plans proposed by Samarco as part of its bankruptcy protection application. Creditors are being asked to accept a 95% “haircut” on their claims against Samarco in order to secure its survival as a going concern. The latest JR plan (presented on 10 March 2022) contained a limit on Samarco’s obligation to provide further financial contributions to Renova according to its cash availability of up to a maximum sum of US$2.4billion. Clause 5.8 provides that this did not represent any limitation on the amounts to be made available to Renova or its budget for the implementation of programmes fully to repair the damage caused by the dam collapse, which would continue to comply with the terms of TTAC. The defendants’ position is that the obligations of BHP Brazil and Vale to fund Renova in accordance with TTAC are not limited or restricted by the JR plan.

The evidence is said by the claimants clearly to satisfy the requirements of Ladd v Marshall [1954] 1 WLR 1489: it could not have been obtained with reasonable diligence for use at the hearing below; it would probably have an important influence on the result of the case, since it goes directly to the question of full redress in Brazil, and is such as is presumably to be believed.

The defendants object to the application:

It is right that the claimants made a clear and unequivocal election before the Judge and in their grounds of appeal as advanced on their application under CPR 52.30 not to pursue any solvency argument so far as Samarco was concerned. In those circumstances, it would be wrong to permit them to adduce any evidence that relates solely to solvency. The question is whether the fresh evidence goes beyond mere solvency arguments.

Some of the evidence can be said to be essentially only solvency-related. We refer, for example, to the question of how Samarco’s dam-related liabilities are to be treated within the bankruptcy. In particular there is an issue whether they will be treated as unsecured pre-bankruptcy credits in competition with all other unsecured pre-bankruptcy credits, in which case they will be worth only 5% of true value.

However, we consider that there is also fresh evidence that goes beyond evidence of mere insolvency, and rather to the ability to enforce any judgment and possible routes of redress following the JR. Amongst other things, it reveals doubts within the JR (and disputes with creditors) as to whether Samarco should be treated as being liable for no more than a third of the liabilities arising in relation to the collapse, the other two-thirds being borne by the shareholders. The defendants disagree with this characterisation, but if it is correct and pursuant to the Samarco JR, or in the context of bankruptcy proceedings, Samarco is to be treated as a matter of law as being responsible for only a third, then any ability to sue Samarco alone is not a full remedy.

Nor can the fresh evidence be dismissed as irrelevant, as the defendants submit:

Standing back, we bear in mind that the fresh evidence is not, and is not said to be, critical to the arguments on abuse. At the same time, it would be artificial and wrong to shut our eyes to the admitted facts in the context of a strike out application where both sides have sought to update the court on material developments in Brazil since the Judgment. Notably, the defendants have sought for the purpose of the appeal to plug significant gaps in their evidential positions, for example through evidence from Mr de Freitas which could have been adduced before the Judge, but which was not. The Samarco JR is a significant new event which has only arisen post-judgment. There has been no suggestion of any evidential prejudice to the defendants.

In all the circumstances, we will admit the evidence. Although our conclusion does not depend on it, it underscores the conclusion that the English proceedings cannot be said to be obviously pointless and wasteful because the claimants are clearly able to obtain full redress in Brazil. On any view, it adds another layer of complexity and uncertainty to the overall picture.

For these reasons, the English claims are not clearly and obviously pointless and wasteful and are not to be struck out. There is a realistic prospect of a trial yielding a real and legitimate advantage for the claimants such as to outweigh the disadvantages for the parties in terms of expense and the wider public interest in terms of court resources. The appeal against the Judge’s order to strike out, alternatively stay, the proceedings for abuse of process must be allowed.

For the sake of completeness, the English proceedings cannot be said to be oppressive. The defendants are not involved as parties to any of the proceedings in Brazil (save in a few limited instances), and they are not sued there by any of the claimants. Nor can the burden on the English courts be said to be disproportionate in circumstances where these are arguable claims for significant sums.

We should make it clear that we recognise fully the efforts being made in the Brazilian courts to resolve the question of compensation for those who have suffered as a result of the catastrophe, and in particular the commitment of Judge Mario. We do not in any way seek to discourage claimants from engaging with whatever opportunities properly exist for them in Brazil: that is something on which they must make their own decisions, with the benefit of legal advice where available. Our conclusion is simply that the remedies available in Brazil are not so obviously adequate that it can be said to be pointless and wasteful to pursue proceedings in this country.

Article 34 is one of six articles in Section 9 of Brussels Recast which is headed “Lis pendens - related actions”. Articles 29 and 30 cover cases where proceedings are pending in the courts of other member states. They reflect similar provisions which were to be found in the predecessors to Brussels Recast, namely the Brussels Convention (Convention on Jurisdiction and Enforcement of Judgments in Civil and Commercial Matters 1968 (OJ 1978 L304 p36)) and the Judgments Regulation (Council Regulation (EC) 44/2001 (OJ 2001 L12 p1); and equivalent provisions in the Lugano Convention (Convention on jurisdiction and recognition of judgments in civil and commercial matters 21.12.2007 (OJ 2007 L339 p1)). The numbering is different but the substance is the same, so that authorities on the equivalent provisions in the Brussels Convention, the Judgments Regulation and the Lugano Convention are to be taken as authoritative in respect of articles 29 and 30 of Brussels Recast.

Articles 33 and 34 were new in Brussels Recast, and cover cases where proceedings are pending in the courts of a “third state”, that is to say a non-member state. Article 33 mirrors article 29 in being concerned with cases involving the same parties and the same causes of action. Article 34 mirrors article 30, being concerned with “related actions”.

Article 34 provides:

“Article 34

1.

Where jurisdiction is based on Article 4 or on Articles 7, 8 or 9 and an action is pending before a court of a third State at the time when a court in a Member State is seised of an action which is related to the action in the court of the third State, the court of the Member State may stay the proceedings if:

(a)

it is expedient to hear and determine the related actions together to avoid the risk of irreconcilable judgments resulting from separate proceedings;

(b)

it is expected that the court of the third State will give a judgment capable of recognition and, where applicable, of enforcement in that Member State; and

(c)

the court of the Member State is satisfied that a stay is necessary for the proper administration of justice.

2.

The court of the Member State may continue the proceedings at any time if:

(a)

it appears to the court of the Member State that there is no longer a risk of irreconcilable judgments;

(b)

the proceedings in the court of the third State are themselves stayed or discontinued;

(c)

it appears to the court of the Member State that the proceedings in the court of the third State are unlikely to be concluded within a reasonable time; or

(d)

the continuation of the proceedings is required for the proper administration of justice.

3.

The court of the Member State may dismiss the proceedings if the proceedings in the court of the third State are concluded and have resulted in a judgment capable of recognition and, where applicable, of enforcement in that Member State.

4.

The court of the Member State shall apply this Article on the application of one of the parties or, where possible under national law, of its own motion.”

The Recitals which are specific to articles 33 and 34 are Recitals (23) and (24) in the following terms:

“(23)

This Regulation should provide for a flexible mechanism allowing the courts of the Member States to take into account proceedings pending before the courts of third States, considering in particular whether a judgment of a third State will be capable of recognition and enforcement in the Member State concerned under the law of that Member State and the proper administration of justice.

(24)

When taking into account the proper administration of justice, the court of the Member State concerned should assess all the circumstances of the case before it. Such circumstances may include connections between the facts of the case and the parties and the third State concerned, the stage to which the proceedings in the third State have progressed by the time proceedings are initiated in the court of the Member State and whether or not the court of the third State can be expected to give a judgment within a reasonable time.

That assessment may also include consideration of the question whether the court of the third State has exclusive jurisdiction in the particular case in circumstances where a court of a Member State would have exclusive jurisdiction.”

The grant of a stay under article 34.1 is discretionary (“the court….may stay proceedings”) but the discretion only arises upon fulfilment of five threshold conditions.

The Judge started by recognising that jurisdiction over BHP England was established under article 4, and that article 4 was concerned with promoting certainty over flexibility (para. [153]). He observed that the broad object of article 34 was identified in Recital (23) (para. [154]). He identified the above five questions which all needed to be resolved in the defendant’s favour if a stay were to be granted, the sixth being the exercise of the discretion.

As to the threshold condition in 34.1(a), the Judge identified the decision of the House of Lords in Sarrio SA v Kuwait Investment Office [1991] AC 32 as the starting point for the proper approach to assessing the risk of irreconcilable judgments. That was a case concerned with article 22 of the Brussels Convention, the precursor to article 30 of Brussels Recast. The Judicial Committee rejected the approach of the Court of Appeal that in order for there to be a risk of irreconcilable judgments the inquiry was limited to “primary” issues which are those necessary to establish the cause of action, and did not include issues which the court might or might not decide and which would not be essential to its conclusion. Such an approach was inconsistent with Owners of Cargo Lately Laden on Board the Ship Tatry v Owners of the Ship Maciej Rataj (Case C-406/92) [1999] QB 515, [1994] ECR I-5439, and the opinion of Advocate General Tesauro in that case at [28] that

“The rationale of the provision is therefore to encourage harmonious judicial decisions and thereby obviate the danger of judgments which conflict with each other, albeit only as regards their reasoning. The court second seised should therefore be able to have recourse to the machinery envisaged by [article 22] whenever it considers that the reasoning adopted by the court hearing the earlier proceedings may concern issues likely to be relevant to its own decision.”

The Judge quoted Lord Saville’s conclusion at p. 41F:

“… I am of the view that there should be a broad common sense approach to the question whether the actions in question are related, bearing in mind the objective of the article and applying the simple wide test set out in article 22 and refraining from an over-sophisticated analysis of the matter.”

The Judge then observed that a narrower approach to relatedness for the purposes of article 34.1(a) had been taken by Mr David Donaldson QC sitting as a deputy High Court Judge in In re Zavarco Plc [2015] EWHC 1898 (Ch) , [2016] Ch 128 at para. [35] and by Stuart-Smith J, in Jalla v Royal Dutch Shell Plc [2020] EWHC 459 (TCC), to whom Sarrio did not appear to have been cited. The Judge adopted the wider Sarrio approach as applicable to article 34.1(a). He observed that even on the impermissibly narrower approach he would have concluded that it was fulfilled because of the overlap he identified in the issue of BHP Brazil’s control of Samarco, by reference to paragraph 275 of MPOC. His conclusion at para [181] was that:

“On a broader interpretation, the list of areas in which potentially irreconcilable judgments are liable to arise is almost endless. By way of example only:

(i)

What health consequences can and cannot be attributed to the pollution?

(ii)

What heads of damages are permissible as a matter of Brazilian Law?

(iii)

What geographical areas were affected by the pollution?

(iv)

What is the appropriate quantum of damages in any individual case?”

He then observed that it was irrelevant that there was an issue between the Brazilian experts whether factual findings in the 155bn CPA would be binding in Brazil, on the basis that they could be taken into account in England and need not be binding on an English Court. He rejected the relevance of the claimants’ evidence that it was “all but inevitable” that there would be no judgment in the 155bn CPA on the basis that he found the reasoning behind the approach “elusive”; and basing himself on an acceptance by Mr Hollander at the hearing that the prospect of Judge Mario determining issues in the absence of a negotiated settlement was not fanciful. This was sufficient to support there being a risk of irreconcilable judgments, the test being one of risk, not likelihood.

At para. [188] he adopted an alternative approach, which had not been advanced by the defendants:

“I further took the view, although it is not necessary for my determination of this case and indeed was not a position advanced by the defendants, that the process of homologation (or ratification) by the Brazilian court of any agreement under the 155bn CPA would be, in itself, a judgment falling within the scope of Article 34. In this regard, the entirety of the issues arising under the 155bn CPA, whether or not subject to agreement between the parties, would fall within the range of potentially irreconcilable judgments.”

At para. [189] he stated that two schools of thought had emerged as to what was meant by “expedient to hear and determine the related actions together”. That articulated by the Court of Appeal in PJSC Commercial Bank Privatbank v Kolomoisky [2019] EWCA Civ 1708, [2020] Ch 783 (“Privatbank”) at paras. [183]-[192] was that it meant no more than that it should be desirable that they be heard together, without any requirement that consolidation should be a practical possibility. The alternative view, based on the judgment of Bean LJ in EuroEco Fuels (Poland) Ltd v Szczecin and Swinoujscie Seaports Authority [2019] EWCA Civ 1932, [2019] 4 WLR 156, (“EuroEco Fuels”) at para. [48], was that there must be a practical possibility that the actions will at least be heard at the same time and by the same judge. The Judge analysed the tension between these decisions, noting that (1) in Privatbank the Court of Appeal had focused on, and analysed, what was meant by “expedient”, whereas Bean LJ had been focusing on what was meant by “heard and determined together” as if it were not qualified by “expedient”; (2) Bean LJ cited Privatbank without any indication that he intended to depart from it; (3) the citation of Privatbank involved Bean LJ observing that, in that case, had the English proceedings been heard in Ukraine, it would have involved separate proceedings, in support of a conclusion that the judge had no discretion to decline jurisdiction or impose a stay; whereas in Privatbank, that had not been a ground for holding there was no jurisdiction, but on the contrary that there was jurisdiction to grant a stay and this was merely a merely a factor in the exercise of the discretion; (4) in Privatbank the court expressly approved the decision in Nomura International Plc v Banca Monte Dei Paschi Di Siena SpA [2013] EWHC 3187 (Comm), [2014] 1 WLR 1584, in which the EuroEco approach was not applied by Eder J, it being a case where it would have been impossible for the related actions to be heard or determined together; and (5) that the Privatbank approach had been followed and treated as binding by Butcher J in The Federal Republic of Nigeria v Royal Dutch Shell Plc [2020] EWHC 1315 (Comm) at para [77].

He accordingly treated Privatbank as representing binding authority on the point, and held that any practical difficulties with consolidation in Brazil were irrelevant to the question of expediency.

As to the threshold condition in article 34.1(b), the Judge determined that the issue was at this stage a conceptual one which did not require the court to determine that the Brazilian court will, in fact, give such a judgment in the future, citing and adopting the observations of Fancourt J at first instance in Privatbank [2018] EWHC 3308 (Ch); [2019] 1 All ER (Comm) 971 at para. [150]:

“Under art 34, the next question is whether it is expected that the Ukrainian courts will give a judgment capable of recognition and – where applicable – enforcement in England and Wales. This criterion relates to the recognition and enforceability of a judgment of the third state in principle. The court of the member state cannot be expected to decide one way or the other whether the court in the third state will in fact give a judgment in future, though the apparent likelihood of its doing so or not doing so would be relevant to the exercise of discretion or the question of whether it was necessary in the interests of the proper administration of justice to grant a stay. At this stage of analysis, however, the question of recognition and enforcement is one of principle.”

The Judge then addressed the claimants’ argument that if the 155bn CPA judgment were unfavourable to the claimants, it would have no res judicata effect under Brazilian law and the claimants would be free to assert their claims in full in civil proceedings, however irreconcilable with the terms of the 155bn CPA judgment; and that this prevented it from being a judgment capable of recognition. The Judge rejected the argument on the basis that a judgment in favour of the claimants in the 155bn CPA would be final and binding in their favour and so capable of recognition here.

As to the article 34.1(c) threshold condition, the Judge started by referring to Recital (24) and rejected the claimants’ argument that the court could not consider factors which would overlap with a hypothetical application of the doctrine of forum non conveniens.

He then addressed the claimants’ reliance on the evidence that there could be no consolidation, and noted that whilst an ability to consolidate was not of itself a prerequisite to the application of article 34, its absence would normally weigh heavily in the balance against a stay. In doing so he cited the passage in the Court’s judgment in Privatbank at paras. [209]-[210] which includes the statement that “… absent some strong countervailing factor, the fact that proceedings cannot be consolidated and heard together will be a compelling reason for refusing a stay”. The Judge proceeded on the assumption that consolidation of the present claim with the 155bn CPA could not or would not take place, but concluded that “the availability or otherwise of consolidation, does not, of itself, provide the answer in the circumstances of this case”. His reasoning was that the significance of the availability or unavailability of consolidation was that it usually affected the risk of inconsistent judgments, but that the unavailability of consolidation did not increase that risk in this case because:

In the light of the first of these factors he later said that it would be idle to categorise any stay imposed as a “consolidation stay”: there would be no consolidation, so that a stay to permit it to occur would be pointless. He concluded that “[t]he real purpose of any stay would be to allow proceedings in Brazil to take their course until the risk of irreconcilable judgments had sufficiently attenuated or there had been some other development which justified lifting the stay” (para. [226]).

He identified the factors which satisfied him that a stay was necessary for the proper administration of justice as being that:

As to exercising the discretion, the Judge observed that once article 34.1(c) was fulfilled, such that the court was satisfied that it was necessary to grant a stay for the proper administration of justice, it was difficult to conceive of circumstances in which the discretion would be exercised other than by granting one; but that in any event there were no such circumstances in the present case.

We agree with the Judge that the correct approach to the test of expediency in article 34.1(a) is that set out in the Privatbank decision, by which we are bound, and which is not undermined by what Bean LJ said in EuroEco Fuels, for the reasons given by the Judge. What is required to fulfil article 34.1(a) is that it must be desirable for the two actions to be heard and determined together in order to avoid the risk of irreconcilable judgments, irrespective of whether that is a practical possibility. Mr Choo Choy did not argue the contrary, although he reserved the right to do so should the matter go further.

We also agree that the test of relatedness for the purposes of article 34 is the broad test identified in Sarrio and the European jurisprudence in, and following, The Tatry, which is applicable to article 30 and its predecessors. Again, Mr Choo Choy did not argue the contrary. He submitted that he had a number of points on relatedness, but that they were better addressed in the context of article 34.1(c), and he therefore asked us to assume, without deciding, that article 34.1(a) was fulfilled. We will do so without at this stage addressing the degree of overlap between this action and the 155bn CPA, to which we will return in the context of the arguments on article 34.1(c).

Mr Choo Choy submitted that the Judge fell into error in two ways, each of which falsified his conclusion that the article 34.1(b) threshold was met in this case. First, it was argued, no judgment was “expected” in the 155bn CPA, based on the claimants’ expert evidence. He submitted that the paragraph imposed a twofold condition, namely (i) that a judgment was expected as a matter of fact and (ii) that the expected judgment was one which was capable of recognition and, where applicable, enforcement. Expectation as a matter of fact imported a test that there must be a real prospect of such judgment being given. He relied on this as being the natural meaning of the language, and further submitted that it would be a nonsense if a stay could be granted where a conflicting judgment was posited in theory but was not a real prospect in practice. In response Mr Toledano argued that the paragraph only imposed a single question, namely whether the action which was pending was of such a character as to give rise to a judgment capable of recognition, or where applicable, enforcement. In the alternative, if it required a real prospect of a judgment in the 155bn CPA being delivered, there is such a prospect in this case, based on the evidence of Professor Didier; the claimants’ contrary evidence would not prevent a real prospect threshold being met.

Mr Choo Choy’s second argument was that a judgment in the 155bn CPA proceedings would only be capable of recognition if decided in favour of the claimants, not if decided in favour of the Brazilian Companies; and that such an asymmetric outcome meant that the judgment could not fulfil article 34.1(b). Mr Toledano submitted that all that was required was a judgment “capable” of recognition, or where applicable enforcement, and that if one possible outcome was a recognisable and enforceable judgment, that was sufficient.

In our view Mr Toledano’s submission on the meaning of article 34.1(b) is to be preferred and the Judge’s conclusion was correct, although not for the reasons given by Fancourt J in Privatbank which the Judge adopted.

Article 34.1 is, by its opening words, engaged if, and only if, an action is pending before the third court state. That will ordinarily involve an expectation of a judgment in due course, absent a settlement. Article 34.1(b) is not concerned to introduce a factual question as to the degree of likelihood of expectation, but to characterise the nature of the pending action by reference to its capability of resulting in a recognisable or, where applicable, enforceable judgment. A number of factors point towards that conclusion.

First, recital (23) is informative of the purpose of article 34.1(b) whereas recital (24) is clearly directed to the purpose of article 34.1(c). Recital (23) says nothing about any factual inquiry into the degree of likelihood of a judgment being given; rather it focuses on the nature and quality of a judgment in the pending action by reference to its recognisability and enforceability. It is in recital (24) where there is to be found a reference to the discretionary factors to be taken into account in determining the proper administration of justice requirement in article 34.1(c), one of which is “whether or not the third state can be expected to give a judgment within a reasonable time”. It is here, in the context of article 34.1(c), not article 34.1(b) that factual inquiries into expectation are raised. It would do no violence to the language or scheme of recital (24) to treat it as implicitly including the words “or at all” at the end.

Secondly, if article 34.1(b) were intended to introduce not one, but two threshold conditions, one might expect to find them separately enumerated in separate subparagraphs, which is the drafting technique adopted for the article as a whole.

Thirdly, articles 29 and 30 have no equivalent provision for the recognisability of the foreign court judgment. That is because they are concerned with actions pending in another member state and so with member state judgments, which by articles 36.1 and 39 are made recognisable and enforceable in other member states. That explains the additional requirement in article 34.1(b), because the recognisability and enforceability of judgments in third states will be a matter for the differing domestic rules of individual member states. However, if there were intended to be an additional threshold requirement of a sufficient degree of likelihood of the foreign judgment occurring, one would expect such a requirement to apply equally to the lis pendens provisions applicable between member states provided for in article 30. Its absence there suggests that it is not intended to form a threshold requirement in article 34 cases either.

Fourthly, recital (23) suggests that what is intended is a flexible mechanism, which militates against putting a restrictive interpretation on what is a threshold condition.

Fifthly, there is some support for this approach in the travaux préparatoires to Brussels Recast. Article 34 emerged late in the drafting process, the new provisions being initially limited to what became article 33 aimed at third state actions between the same parties involving the same cause of action. The draft text of what became article 33 in the original proposal of the European Commission of 14 December 2010 imposed a requirement that “it may be expected that the court in the third state will, within a reasonable time, render a judgment that will be capable of recognition, and, where applicable, enforcement in that Member State”. The explanatory memorandum made clear that, as its language suggests, these were two separate requirements, that the judgment would be expected within a reasonable time and that it would be capable of recognition and enforcement in the member state. At that stage the recitals included what became recital (23) in its final form, but there was no equivalent to recital (24). A draft of what became article 24 was first proposed, by the Committee of Legal Affairs of the European Parliament on 25 September 2012. It contained the same wording as the final form of article 34.1(b) and added a recital in the terms of the final form of recital (24), including there a reference to whether a judgment was to be expected in a reasonable time, and removing any such reference from what became article 33.1(a), and not including it in article 34.1(b). This drafting was adopted in the Committee’s final proposal of 15 October 2012 and in the Regulation itself. This suggests, to put it no higher, that factual questions as to whether and when a judgment was expected were not intended to be part of the article 33.1(b) inquiry, and the identical 34.1(b) inquiry, but were assigned to that under article 34.1(c) to which recital (24) referred.

Against this, the points taken by Mr Choo Choy appear to us to have little weight. The language of the paragraph does not point firmly towards the interpretation he seeks to put on it, and in any event, it is not to be construed in the same was an English statute. A purposive approach to interpretation is often adopted by the CJEU, and its predecessor the European Court of Justice, and the recitals are a surer guide to the purpose in such a teleological approach. Nor is there anything in the point that it cannot have been intended to permit a stay where there was no prospect at all of any judgment in the pending action: this is a factor which can be taken into account in the article 34.1(c) inquiry, and would no doubt in most cases be fatal to a stay. It need not be part of the 34.1(b) threshold requirement.

We would in any event have accepted Mr Toledano’s alternative submission, that if what was required was a real prospect of a 155bn CPA judgment being given, that threshold was met by the expert evidence adduced by the defendants, notwithstanding the contrary evidence which gave rise to substantial uncertainties on that question.

On this issue, too, we consider that the defendants are correct, as the Judge found. All that is required is that the judgment in the pending action is “capable” of recognition. It is common ground that a judgment in the 155bn CPA against the Brazilian Companies would fulfil the common law criteria for recognition. If that is one possible and not fanciful outcome, a judgment in the 155bn CPA is capable of recognition, notwithstanding that it would not be if decided in favour of the Brazilian Companies. Mr Choo Choy argued that this involved the court having to make an assessment of the outcome of the pending proceedings on their merits. We do not see that it does so. The exercise at this stage is a conceptual one, looking at the type of judgment to which the third state pending action may give rise, and evaluating whether it attracts recognition, or where applicable enforceability. The exercise is not premised on a factual inquiry as to what will happen; on the contrary, it assumes that the court cannot embark at this stage upon an assessment of the outcome on its merits, but focuses on what outcomes the pending action is “capable” of giving rise to.

The claimants’ interpretation would have the unacceptable consequence of removing the availability of relief in appropriate cases. Suppose there were 20 permutations of outcome, 19 of which would result in a judgment capable of recognition and enforcement, but one of which would not. If Mr Choo Choy’s argument were correct, article 34.1(b) would bar relief at the threshold stage. That is, to our minds, antithetical to the scheme and purpose of article 34 itself, and the flexibility identified in recital (23), that a stay should be available in such cases if necessary for the proper administration of justice. It would prevent the condition being met where, for example, the foreign defendant advanced a challenge to the personal jurisdiction of the foreign court, which was at that stage unresolved, irrespective of the possibility of such challenge being unsuccessful.

As we have explained, it was accepted by the defendants that, so far as the 58 were concerned, the 155bn CPA would not make any determination which would be binding on them. The slightly more nuanced position in the evidence means that it is tolerably clear that the 13 large businesses cannot benefit from a generic sentence in the 155bn CPA. So far as they are concerned it will have no effect on their claims, whichever way it is decided: the expected judgment is not capable of being recognised as making any determination of the merits of their claims, whatever its outcome. Each of the claimants must be considered individually for these purposes, and it is irrelevant whether their claims are brought in a single action or separate actions which are consolidated or to be heard together.

When we asked Mr Toledano in the course of the hearing whether it was his submission that the 58 (or 13) lost their right to bring a claim which could not be stayed under article 34 by it being coupled with claims by claimants which are susceptible to an article 34 stay, his response was that this formed no part of his submission. Rather, he argued, the risk of irreconcilable judgments test was sufficiently wide, applying the Sarrio criteria, that it could still be said of the 58 or 13 that a 155bn CPA would give rise to a risk of irreconcilable judgments, even though they would not be parties to it; article 34 is applicable where the parties are not the same; and where article 34(1)(a) is fulfilled, the judgment need only be capable of recognition as between the parties to it, not the parties in the member state. We would accept this argument. The effect of the assumption we are asked to make that article 34(1)(a) is fulfilled means that the 58, or the 13, are not in a different position from the other claimants merely by reason of the fact that they will not be bound by the 155bn CPA judgment.

There was debate in the written and oral argument before us as to whether a stay was available or could be justified on a “consolidation” basis, or on a “wait and see” basis. Mr Toledano advanced both bases as alternatives.

In the 5^th cumulative supplement to the 15^th edition of Dicey Morris and Collins The Conflict of Laws, the editors say this at pp. 171-172:

“In the fourth place, the rules applicable as between Member States contemplate that the court will not merely stay its proceedings. Provided that the court first seised confirms its own jurisdiction, other courts must decline jurisdiction in a lis pendens case (Art.29(3)) and may do so in the context of related actions (Art.30(2)). By contrast, under the new procedures applicable to relations with the courts of non-Member States, the court of the Member State will only ever stay its proceedings until the proceedings in the non-Member State have resulted in a judgment capable of recognition or enforcement in the Member State. At that point, the Member State court must dismiss its action in [a] case involving lis pendens (Art.33(3)) and may do so in a related action case (Art.34(3)). But until that time the Member State court may continue the proceedings at any time if the non-Member State proceedings are themselves stayed or discontinued; or are unlikely to be concluded within a reasonable time; or continuation is “required for the proper administration of justice”: Art.33(2). In the case of related actions, all of these grounds apply and continuation may also be ordered if there is no longer a risk of irreconcilable judgments: Art.34(2)(a). The stay in all such cases is therefore temporary until the foreign proceedings result in a judgment enforceable in the Member State. The rule thus operates as a tool of case management, rather than allocating jurisdiction as it does as between the courts of Member States.”

The distinction between the article being a tool of case management and one of allocation of jurisdiction is not as binary as this passage suggests. The focus of the article is the risk of irreconcilable judgments. If the claim being advanced in the member state can, in the alternative, be brought in the third state and heard at the same time by the same tribunal as the pending action, whether or not formally consolidated, the risk of irreconcilable judgments will be eliminated if that takes place. The availability of such “consolidation” can therefore properly be a basis for a stay under the article, although it will not necessarily mandate a stay in all cases. If granted on that basis, the stay would envisage an allocation of jurisdiction, rather than just temporary case management, on the footing that the entire claim advanced in the member state can and should be decided in the third state. On the other hand, the width of the concept of related actions, and uncertainties over the outcome of the third state action, may result in the stay being envisaged as a temporary one to wait and see whether and to what extent the outcome affects the action in the member state.

Where the availability of consolidation is uncertain, one aspect of the “wait and see” approach may be to wait and see whether it proves to be available. In a case such as the present, however, in which it is clear that the claimants would not seek to pursue the proceedings in the third state, we think the Judge was right to ignore that possibility. It is also this uncertainty as to whether consolidation is available in this case which renders irrelevant the statement in Privatbank at para. [210] that the unavailability of consolidation will usually be a compelling reason to refuse a stay, absent some strong countervailing factor.

The real question in this case, therefore, is whether the Judge was right that a “wait and see” stay is necessary for the proper administration of justice. Does waiting for the outcome of the 155bn CPA give rise to advantages which sufficiently outweigh any disadvantages such that a stay is necessary?

This in turn gives rise to an issue as to the factors which may properly be taken into account. A narrow view, which might at first sight appear justified from the terms of the article itself, would be that it is restricted to an assessment of the advantages from the judgment in the pending third state action (here the 155bn CPA judgment) and the advantages for the claims in the member state action which is to be stayed. If so, the focus would be confined to the extent of overlap. The broader view, for which the defendants contend, is that it involves any advantage to the parties or the court from a temporary stay. This would include as a relevant consideration, for example, the redress which the claimants might receive from the other processes in Brazil whilst awaiting a 155bn CPA judgment, and those arising from typical forum non conveniens considerations.

We consider that the wider approach is the correct one. Recital (24) provides that, when taking into account the proper administration of justice, the court should assess all the circumstances of the case. This is very broad wording. The wide introductory wording and non-exhaustive character of the listed examples mean that connections of the case to the member state are relevant to the inquiry (see Ness Global Services Ltd v Perform Content Services Ltd [2020] EWHC 3394 (Comm),  [2021] 1 WLR 4146, at paras. [66]-[67]). However, the relevant factors are not so confined, as the examples which follow make clear: they include aspects of the third state proceedings. The first example given is “connections between the facts of the case and the parties and the third state concerned” which has nothing to do with the advantage from the third state judgment for the member state proceedings, but is rather the sort of connecting factor commonly taken into account in addressing arguments of forum non conveniens. As this court said in Ness at para. [67], the Judge was correct at para. [206] of the Judgment to reject the claimants’ argument that forum non conveniens factors were irrelevant.

In Starlight Shipping Co v Allianz Marine & Aviation Versicherungs AG (The Alexandros T) [2013] UKSC 70; [2014] 1 All ER 590, Lord Clarke addressed the discretion arising under the predecessor to article 30 at para. [92] by citing with approval as relevant the three factors identified at paras. 74-79 of the Opinion of Advocate General Lenz in Owens Bank Ltd v Bracco (No 2) Case C-129/92 [1994] QB 509. These were (1) the extent of relatedness and the risk of mutually irreconcilable decisions; (2) the stage reached in each set of proceedings and (3) the proximity of the courts to the subject matter of the case. That was in the context of the predecessor to article 30, applying between member states, which does not include the “necessary for the proper administration of justice” wording. Nevertheless the lis pendens context is the same, and we would accept that these are relevant factors, amongst “all the circumstances of the case”, for the purposes of the proper administration of justice requirement in article 34.1(c).

The fact that the relevant factors are unlimited in scope does not, of course, mean that they are to be accorded equal weight. We consider the first factor identified by Advocate General Lenz to be of particular importance where a stay is to be granted on a “wait and see” basis. If a stay is to be granted to await the outcome of a pending action in the third state only if, and because, it gives rise to the risk of irreconcilable judgments, the extent to which it may do so, and the extent to which a judgment in the pending action may assist or impact upon the resolution of the member state action, must be central to the question of whether a stay is necessary. We therefore regard the extent of overlap between the 155bn CPA and the claim against these defendants, and the extent to which any judgment will impact on the action here, to be factors of particular significance. In Easygroup Ltd v Easy Rent a Car Ltd  [2019] EWCA Civ 477, [2019] 1 WLR 4630, at para. [67] David Richards LJ described the degree of overlap as “a factor of great importance” for the purposes of article 30.

In this respect the Judge appears to us to have fallen into an error which permeated his reasoning at various stages of the analysis under article 34.1(c). The stay granted by him, as sought in its final form, was only until conclusion of the 155bn CPA. The 155bn CPA proceedings will conclude, if not settled, and, if resumed at all, with a generic sentence. Any liquidation proceedings by which the claimants could obtain monetary compensation thereafter would be separate proceedings. Such liquidation proceedings could not engage article 34 because they were not pending when the English court became seised of the current action. It follows that, for the purposes of the article 34 application, the nature and extent of overlap which falls to be considered when addressing whether and to what extent there is a risk of irreconcilable judgments, and in considering whether that risk weighs in favour of a stay being necessary for the proper administration of justice, is limited by reference to that which might be decided in the 155bn CPA. The overlap between issues which will arise in this action and issues which would arise in Brazilian liquidation proceedings following a generic sentence in the claimants’ favour, if such eventuated, form no part of the relevant comparison. Equally the risk of overlap between these proceedings and other proceedings in Brazil outside the 155bn CPA form no part of the inquiry.

The Judge’s conclusion at para. [181] was that the list of areas in which potentially irreconcilable judgments are liable to arise was “almost endless” and included, by way of example only, his four identified areas. However, category (iv) (the appropriate quantum of damages in any individual case) is not an issue which will be addressed in the 155bn CPA, but only potentially in any subsequent liquidation proceedings. Moreover, the Judge makes no mention of the evidence of the claimants’ experts that generic issues of causation and quantum would not be addressed in the 155bn CPA at all which, if accepted, would mean that there would also be no risk of irreconcilable judgments in respect of his remaining exemplar categories (i) (health consequences attributable to pollution), (ii) (permissible heads of damage in Brazilian law) and (iii) (geographical effect of the pollution). This error must have informed his conclusion at para. [220] that, save for the identity of the defendants, the degree of relatedness between the Brazilian and English proceedings is “very close indeed”. That cannot be justified if the Brazilian proceedings being considered are, as they should be, confined to the 155bn CPA, as we analyse in more detail below.

A similar error is apparent in his further reasoning at para. [228] relying on all the factors he had identified for the purposes of the abuse argument as making the action unmanageable, by reason of parallel proceedings in Brazil, such that in his view the combined pursuit of them would lead to pandemonium which would be the very antithesis of the proper administration of justice. We have already rejected his reasoning that the current proceedings would be procedurally unmanageable, which is an error which infects his reasoning on article 34.1(c) which here adopts it; but the further error lies in the fact that his reasoning is based not on pursuit of the 155bn CPA, but rather on pursuit of individual claims by claimants in Brazil either in ordinary civil claims, or in liquidation proceedings following a generic sentence. The pandemonium which he foresaw did not arise from the parallel pursuit of the 155bn CPA.

Moreover, the other factors for or against a stay must, in our view, be judged by reference to the article 34 stay being one which is only to last, at the longest, until the conclusion of the 155bn CPA. Article 34.2 would permit a lifting of the stay before then, but there is no outcome of the 155bn CPA which could engage art 34.3 so as to permit dismissal of the claim. The stay ordered by the Judge will, by its terms, come to an end upon conclusion of the 155bn CPA, and the claimants will thereupon be free to pursue their claims here against these defendants. Whilst in theory the court would have power at that stage to impose a case management stay to await the outcome of liquidation proceedings, it is difficult to conceive that it could properly do so in the light of Owusu, because it would involve granting a stay of a claim against a defendant domiciled in a member state on what would essentially be forum non conveniens grounds.

Mr Toledano argued that we could assume that at that stage, in the event of a favourable judgment in the 155bn CPA proceedings, the claimants would not wish to pursue this action and would choose not to do so: they would already have the benefit of a finding of liability against the Brazilian Companies with findings on causation and quantum, and in kind relief from which even the 58 would benefit; or in the event of a final settlement which avoided an 155bn CPA judgment, the claimants would have the benefit of a settlement endorsed by the MPF looking after the interests of all claimants, which it is to be assumed would provide better redress than they could hope for in this action. He did not suggest that that was a certainty, but rather that it was likely, or at least something of which there was a real prospect.

Given the uncertainties as to what may happen in the GTAC negotiations or the 155bn CPA proceedings, it is impossible to rule out the possibility that some claimants will reconsider their position as a result of a settlement or court decision, but we cannot regard it as likely, on the current state of the evidence, that a significant number of them will do so. It must be borne in mind that the claimants have already taken a view that they are better off pursuing their claims here, notwithstanding that Samarco has not, with limited exceptions, sought to deny liability in Brazil, and notwithstanding that the principle underlying GTAC and Renova’s existing programmes is said to be full redress; and that for the reasons we have identified in relation to the abuse arguments, there is a proper objective basis for the claimants’ subjective belief that they will achieve more, more quickly and more cheaply in this action than in Brazil. It is a matter of uncertainty whether a 155bn CPA judgment, if one is delivered at all, will deal with issues of causation or quantum in a way which assists the claimants in this action. We would accept that there is a prospect, which is not fanciful, that a settlement would provide for further entitlements or modes of achieving redress for at least some of the claimants in this action beyond what is currently available; but that too is subject to uncertainty as to whether the manner in which that is or is not implemented provides adequate redress in practice; the evidence of Renova’s activities to date suggests that the claimants would have legitimate grounds for scepticism on that score. We have already expressed our view that the fact that any settlement must satisfy the MPF is no guarantee that it would advance the interests of all claimants.

We do not think that Mr Toledano’s argument can be put higher than that there is a prospect of some of the claimants abandoning the current action after a 155bn CPA judgment or earlier settlement, but it is not likely that a significant proportion would do so.

Accordingly, in determining whether a stay is necessary for the proper administration of justice, it must be kept in mind that what is under consideration is in substance a temporary, albeit potentially lengthy, interruption to the progress of this action, which will resume after conclusion of the 155bn CPA.

Mr Choo Choy also placed emphasis on the word “necessary” and submitted that this itself imposed a high threshold on defendants seeking a stay. We do not agree. The exercise of addressing the proper administration of justice will involve the balancing of a number of potentially competing factors. If the result of that exercise is that a stay is in the interests of the proper administration of justice, a stay will be necessary for that purpose; if not, it will not. We doubt that the word “necessary” adds anything to the balancing exercise involved, save perhaps to leave refusal of a stay as the default position where the court is left in doubt on the state of the evidence.

Mr Choo Choy submitted that whether a stay was necessary admitted only of a right or wrong answer, and the function of this court was to reach its own conclusion on whether a stay was necessary, without deference to any reasoning of the Judge. Mr Toledano submitted that the determination was akin to the exercise of a discretion, and that the constraints on interfering with an exercise of discretion were applicable. In our view the correct approach is that the exercise required by article 34.1(c) is not discretionary but evaluative, involving the balancing of factors, and an appellate court will be reluctant to interfere with the balancing exercise undertaken by the judge if there has been no error of approach in principle: Aldi Stores at para. [16]; see para. [143] above. The dispute is of no significance in this case because in our view the Judge made a number of significant errors in his approach. We have already identified some, including his reliance on his reasoning on the abuse application. The Judge himself expressly recognised at para. [234] that if his findings underpinning his conclusion on the abuse issue were successfully challenged on appeal, his conclusion on art 34.1(c) would fall for fresh consideration (he expressed this by reference to the exercise of his discretion but we think that he intended to refer to his article 34.1(c) decision, which he regarded as dispositive of the exercise of discretion).

The Judge fell into further error at para. [188] in holding that a homologation of a further TAC in the 155bn CPA would be a relevant judgment such that the whole of the range of issues in the 155bn CPA which it compromised would fall within the range of potentially irreconcilable judgments. As the Judge observed, this was not an argument advanced by the defendants and the defendants did not seek to uphold it on the appeal. They were right not to do so. Any homologation would not be a judgment capable of recognition for current purposes because it would not involve the determination of the merits of any legal issues in the 155bn CPA or any legal rights of the parties; it would merely approve a consensual settlement agreement.

Accordingly it is for this court to consider afresh whether a stay is necessary for the proper administration of justice.

Of first and central relevance to this question is the claim against BHP Australia, which for present purposes can be taken to be identical to the claim against BHP England. We have concluded, for reasons set out below, that the forum non conveniens application (and case management stay application) by BHP Australia must fail, for reasons which are not dependent on the article 34 application being rejected. The consequence is that the claimants can progress that claim now in this action. Again there is no reason to think that they will not do so. That being so, it is again difficult to see how there could be said to be any advantage in the claim against BHP England being put on hold until the conclusion of the 155bn CPA. Again, doing so would not contribute to any attenuation of any risk of irreconcilable judgments, nor could it involve the advantages (if any) which might arise from avoiding parallel proceedings in this action and the 155bn CPA. In short, a stay pending conclusion of the 155bn CPA would for this reason be pointless.

The decision and reasoning of the Supreme Court in Vedanta does nothing to undermine this conclusion. In that case the claimants, some 1800 members of poor rural Zambian farming communities, brought proceedings in negligence alleging personal injury, damage to property and other heads of loss, arising from toxic discharges from a Zambian copper mine. The copper mine was owned and operated by the second defendant, a Zambian company (“KCM”), which was an indirect subsidiary of the first defendant, Vedanta Resources plc (“Vedanta”), a UK company. Jurisdiction against Vedanta was founded on its domicile here pursuant to article 4 of Brussels Recast. Permission to serve KCM out of the jurisdiction was granted under the necessary and proper party gateway. Vedanta challenged jurisdiction on the basis that (1) the evidence disclosed no triable issue against it as a matter of law and/or (2) that its suit here was an abuse of EU law because it was being used as an anchor defendant solely to enable the claim against the real target, KCM, to be brought in England. Both submissions were rejected at first instance and at every level of appeal. The issue of potential relevance in the current context was the effect of the claim brought against Vedanta as of right, under article 4, on the forum non conveniens arguments involved in KCM’s application to set aside the order for service out of the jurisdiction.

Lord Briggs, giving the judgment of the court, said at paras. [39]-[40]:

“39.

Following Owusu v Jackson the English court has one hand tied behind its back. No more can it stay the proceedings against the anchor defendant on forum conveniens grounds. This is the precise ratio of Owusu v Jackson, and the Court of Justice was fully aware of the difficulties which that conclusion would be likely to cause in the traditional exercise of the English court’s forum conveniens jurisprudence in such cases. The result is, in a case (such as the present) where the English court is persuaded that, whatever happens to the claim against the foreign defendant, the claimants will in fact continue in England against the anchor defendant, the risk of irreconcilable judgments becomes a formidable, often insuperable, obstacle to the identification of any jurisdiction other than England as the forum conveniens. Thus not only is one of the court’s hands tied behind its back, but the other is, in many cases, effectively paralysed. In the context of group litigation about environmental harm, the defendants say that it has the almost inevitable effect that, providing a minimum level of triable issue can be identified against an English incorporated parent, then litigation about environmental harm all around the world can be carried on in England, wherever the immediate cause of the damage arises from the operations of one of that group’s overseas subsidiaries.

40.

Two consequences flow from that analysis. The first is that, leaving aside those cases where the claimant has no genuine intention to seek a remedy against the anchor defendant, the fact that article 4 fetters and paralyses the English forum conveniens jurisprudence in this way in a necessary or proper party case cannot itself be said to be an abuse of EU law, in a context where those difficulties were expressly recognised by the Court of Justice when providing that forum conveniens arguments could not be used by way of derogation from what is now article 4. The second is that to allow those very real concerns to serve as the basis for an assertion of abuse of EU law would be to erect a forum conveniens argument as the basis for a derogation from article 4, which is the very thing that the Court of Justice held in Owusu v Jackson to be impermissible. In my view, if there is a remedy for this undoubted problem, it lies in an appropriate adjustment of the English forum conveniens jurisprudence, not so as to permit the English court to stay the proceedings against the anchor defendant, if genuinely pursued for a real remedy, but rather to temper the rigour of the need to avoid irreconcilable judgments which has, thus far, served to disable the English court from concluding that any jurisdiction other than its own is the forum conveniens or proper place for the litigation of the claim against the foreign defendant. As will appear, I consider that there is a solution to this difficulty along those lines, where the anchor defendant is prepared to submit to the jurisdiction of the domicile of the foreign defendant in a case where, as here, the foreign jurisdiction would plainly be the proper place, leaving aside the risk of irreconcilable judgments.”

In addressing the position of the foreign defendant, KCM, he said:

“70 In cases where the court has found that, in practice, the claimants will in any event continue against the anchor defendant in England, the avoidance of irreconcilable judgments has frequently been found to be decisive in favour of England as the proper place, even in cases where all the other connecting factors appeared to favour a foreign jurisdiction: see e g OJSC VTB Bank v Parline Ltd [2013] EWHC 3538 (Comm) at [16], per Leggatt J.

71 That is a fair description of the judge’s reasoning in the present case. Having found that, looking at the matter as between the claimants and KCM, all the connecting factors pointed towards Zambia, the judge concluded that, factoring in the closely related claim against Vedanta, which he found as a matter of fact that the claimants were likely to pursue in England in any event, the risk of irreconcilable judgments arising from separate proceedings in different jurisdictions against each defendant was decisive in identifying England as the proper place: see paras 160-168. He said that: “The alternative—two trials on opposite sides of the world on precisely the same facts and events—is unthinkable.”

…

75 I have however been much more troubled by the absence of any particular focus by the judge upon the fact that, in this case, the anchor defendant, Vedanta, had by the time of the hearing offered to submit to the jurisdiction of the Zambian courts, so that the whole case could be tried there. This did not, of course, prevent the claimants from continuing against Vedanta in England, nor could it give rise to any basis for displacing article 4 as conferring a right to do so upon the claimants. But it does lead to this consequence, namely that the reason why the parallel pursuit of a claim in England against Vedanta and in Zambia against KCM would give rise to a risk of irreconcilable judgments is because the claimants have chosen to exercise that right to continue against Vedanta in England, rather than because Zambia is not an available forum for the pursuit of the claim against both defendants. In this case it is the claimants rather than the defendants who claim that the risk of irreconcilable judgments would be prejudicial to them. Why (it may be asked) should that risk be a decisive factor in the identification of the proper place, when it is a factor which the claimants, having a choice, have brought upon themselves?

…

79 After anxious consideration, I have come to the conclusion that Leggatt J’s analysis of this point [in OJSC VTB Bank v Parline], followed by the judge, is wrong. At the heart of it lies the proposition that, because a claimant has a right to sue the anchor defendant in England, there is “no reason why the claimant should be expected or required to relinquish that right in order to avoid duplication of proceedings”. In my judgment, there is good reason why the claimants in the present case should have to make that choice, always assuming that substantial justice is available in Zambia (which is a necessary but hypothetical predicate for the whole of the analysis of this issue).”

…

83 The recognition that claimants seeking to avail themselves of their article 4 rights to sue an anchor defendant are nonetheless exposed to a choice whether to do so at the risk of irreconcilable judgments, even in cases where article 8 is not available, but another proper, convenient or natural forum is available for the pursuit of the case against all the defendants is, to my mind, the answer to the conundrum posed in para 40 above. It does not in any way bring into play forum conveniens considerations as a reason for denying the claimants access to the jurisdiction of England as a member state, against the anchor defendant. It simply exposes the claimants to the same choice, whether or not to avoid the risk of irreconcilable judgments, as is presented by the combination of article 4 and article 8 in an intra-EU context.

84 That analysis does not mean, when the court comes to apply its national rules of private international law to the question whether to permit service out of the jurisdiction upon KCM, that the risk of irreconcilable judgments is thereby altogether removed as a relevant factor. But it does in my view mean that it ceases to be a trump card, and that the basis upon which the judge, following Leggatt J in the OJSC VTB Bank case [2013] EWHC 3538 (Comm), regarded it as decisive, involved an error of principle….

…

87 In conclusion, it is sensible to stand back and look at the matter in the round. This case seeks compensation for a large number of extremely poor Zambian residents for negligence or breach of Zambian statutory duty in connection with the escape within Zambia of noxious substances arising in connection with the operation of a Zambian mine. If substantial justice was available to the parties in Zambia as it is in England, it would offend the common sense of all reasonable observers to think that the proper place for this litigation to be conducted was England, if the risk of irreconcilable judgments arose purely from the claimants’ choice to proceed against one of the defendants in England rather than, as is available to them, against both of them in Zambia. For those reasons I would have concluded that the claimants had failed to demonstrate that England is the proper place for the trial of their claims against these defendants, having regard to the interests of the parties and the ends of justice.”

We address below the impact of this decision on BHP Australia’s forum non conveniens application by reference to the claim against BHP England not being stayed under article 34, where the application is analogous to that made by KCM, albeit in the context of service as of right rather than permission to serve out of the jurisdiction. The question here is the converse: what is the effect of the claimants’ entitlement to pursue BHP Australia here, which is a suitable forum, on a discretionary stay of the claim against BHP England, a defendant against whom jurisdiction is established pursuant to article 4 of Brussels Recast, as was jurisdiction against Vedanta in that case. The Vedanta decision is therefore not analogous. Nor in our view can its reasoning in respect of KCM’s forum non conveniens application be applied to assist BHP England in relation to article 34.1(c). What was critical to the reasoning was (1) that it was only by asserting a risk of irreconcilable judgments arising from pursuit of Vedanta here that the claimants could seek to argue at the first stage of the forum non conveniens test that Zambia was not the natural forum for its claim against KCM; and (2) because pursuing Vedanta here was a matter of choice for the claimant as to whether to create that risk, it was not a trump card. In the current case, the relevance of the claim proceeding against BHP Australia here is not that it creates a risk of irreconcilable judgments, but rather that a stay would serve no useful purpose; and, moreover, insofar as a risk of irreconcilable judgments is relevant at all, it is BHP England, not the claimants, who raise it and seek to rely upon it. In our view there is nothing in the Vedanta decision or reasoning which means that the claimants’ choice to pursue BHP Australia here cannot be relied on as demonstrating that a stay of the claim against BHP England under article 34 is not necessary for the proper administration of justice.

This is sufficient in itself to require that the appeal on the article 34 application be allowed.

However, in case we are wrong, we will address the position on the assumption that any continued pursuit of a claim against BHP Australia here is to be ignored.

Our conclusion on that hypothesis is also that a stay of the claim against BHP England is not necessary for the proper administration of justice.

It is important to start by identifying the disadvantages of a stay. The most optimistic estimate of when the CPA will be resolved at first instance is mid-2024 to mid-2026, and it may well be significantly later, quite possibly by many years; and if, as must be a real possibility, there are further appeals through two levels up to the STJ, that will probably add a further 5 years before the 155bn CPA is finally resolved. There is, in other words, a real possibility that final resolution of the 155bn CPA, if it resumes at all, is well over a decade away. In the course of argument one possibility which was floated was that consideration could be given to lifting the stay following a first instance decision in the 155bn CPA without waiting for the appeal process in Brazil to conclude. That would be a very unattractive course, involving proceedings resuming here for four years or more whilst remaining subject to the risk of a further stay thereafter. Where the court is concerned with disputes about jurisdiction it is desirable that they should be resolved speedily and conclusively, so that the parties can get on with incurring the time and cost of pursuing or defending the claim in the known forum without fear of delay or wasted expense. Any delay or uncertainty in allocation of jurisdiction is itself undesirable.

The claims in this action are already three years old and the only reason they have not substantively progressed is the delay caused by the applications which have been made by the defendants. For there to be a further delay of years, and quite possibly over a decade, before they could resume would cause very substantial prejudice to the claimants in obtaining relief, and would be inimical to the efficient administration of justice as a result of all the well-known problems which delay brings to the process. Moreover, as we have observed, by the time they were ready to resume, events in Brazil will likely have moved on in terms of forms of redress, and it is entirely conceivable that a renewed abuse application would be mounted based on additional material to that now before the court. Mr Gibson would not rule out such a possibility when asked in the course of argument.

Against these serious disadvantages inherent in a lengthy but temporary interruption of the claims being pursued here, the potential benefit of awaiting any 155bn CPA judgment is beset with uncertainties which in combination render the prospect of any such benefit relatively remote and its value relatively small.

First, there is the considerable uncertainty whether the 155bn CPA will resume at all. It is common ground that it will not do so if there is a final settlement. If the claimants’ expert evidence is correct, it will not do so even if there is no settlement or the MPF can only achieve a settlement which falls short of full redress.

Secondly, there is the uncertainty as to what issues it will address if it does resume. It may not address all the claims raised in the original complaint, but only those which are specifically referred for decision. It may not address the liability of the Brazilian Companies generally, or of BHP Brazil in particular, nor any other issue of potential relevance to the claims in the current action.

Thirdly, if it does address the liability of BHP Brazil, there is uncertainty of outcome: it can only create any res judicata, even in Brazil, if it determines that issue against the company. If it decides that there is no liability, that would have no effect in subsequent litigation in Brazil, and consequently could have none in this action.

Fourthly, the degree of overlap with the claim against BHP England is relatively limited, if it arises at all, so far as can be ascertained at present and so far as can be established by the defendants, upon whom the burden lies on this issue. It is fair to say that the issues have not been crystallised in great detail in either set of proceedings. In the 155bn CPA there is simply the MPF complaint and BHP Brazil’s defence, and the scope of the issues may well be widened or narrowed further if the 155bn CPA proceeds to a judgment. In the current action, there is only the MPOC to go on in order to identify the issues, without there yet being any responsive pleading. That leaves it unclear, for example, whether BHP England will challenge the pleaded content of Brazilian law (as opposed to its application) and if so in what way or to what extent.

The current state of the evidence suggests to us the following.

Taking all this into account, we consider that the defendants, upon whom the burden lies, have failed to establish any overlap to an extent which would amount to a weighty factor in favour of a stay.

Fifthly, the status of such relevant determinations as may be made in a 155bn CPA judgment means that they will be of limited assistance in the claim here. None of the parties is the same and the overlap is limited. Nothing it decides will be res judicata or give rise to issue estoppel in this action. Mr Choo Choy submitted that in accordance with the rule in Hollington v Hewthorn [1943] KB 587 any factual findings in the 155bn CPA would be inadmissible in proceedings in the present action. That is not true of findings of Brazilian law, and there are a number of first instance cases suggesting that the rigour of Hollington v Hewthorn may be to some extent diluted. We do not need to consider them further because, even assuming such dilution and consequent admissibility, the utility to the English Court of anything said in a 155bn CPA Judgment is likely to be relatively small at best.

Mr Toledano relied on four other points which we regard as of modest weight in favour of a stay.

Weighing all these factors, we conclude that it would not be necessary for the proper administration of justice to stay any of the claims, even if one were to ignore the position of the 13 large businesses and the claims against BHP Australia.

For all these reasons we allow the appeal against the order made pursuant to article 34.

This application is (and can only be) made by BHP Australia, and references in this part of the judgment to the defendant or defendants should be construed accordingly.

At para. [236] the Judge identified the two stage test identified in Spiliada as being:

“(a)

At stage one, the court's task is to analyse whether the foreign forum is an available forum that is clearly or distinctly more appropriate for any trial of the dispute (i.e. the 'natural' forum). This typically requires analysis of the competing connecting factors as between England and the foreign forum. The burden of persuasion is on the applicant.

(b)

If, following the first stage, it is established that the foreign forum is the natural forum, then the court will grant a stay subject to the second stage. At this second stage, the claimant may seek to establish that "there are circumstances by reason of which justice requires that a stay should nevertheless not be granted. In this enquiry, the court will consider all the circumstances of the case, including circumstances which go beyond those taken into account when considering connecting factors with other jurisdictions. One such factor can be the fact, if established objectively by cogent evidence, that the plaintiff will not obtain justice in the foreign jurisdiction". The claimant bears the burden on stage two.”

He then identified four factors as cumulatively significant in pointing to Brazil as the natural forum:

He then cited extensively from Vedanta, including the passages we have set out earlier in this judgment. He observed at para. [241] that both defendants had offered to submit themselves to the jurisdiction of Brazil, which he said “attenuated” the force of any suggestion of a risk of irreconcilable judgments (if BHP England’s applications were unsuccessful). He concluded that “notwithstanding the conspicuously close corporate relationship between the two defendants I am satisfied that the remaining arguments concerning the appropriate forum are, when taken as a whole, so strong as to lead to no other conclusion than that the first stage of Spiliada is made out.”

In his last paragraph on stage one (para. [242]), he observed that if BHP England were unsuccessful in its applications, the strong likelihood is that the claimants would not be interested in pursuing BHP Australia in Brazil if its forum non conveniens application succeeded.

Turning to stage two, he noted that the burden now shifted to the claimants and stated that it is not sufficient for the claimants to show that proceedings in Brazil would merely be less advantageous than proceedings in England, citing the following passage from the speech of Lord Goff in Connelly v RTZ Corporation Plc (No 2) [1998] AC 854 (“Connelly”) at pp. 872G-873A:

“… [I]f a clearly more appropriate forum overseas has been identified, generally speaking the plaintiff will have to take that forum as he finds it, even if it is in certain respects less advantageous to him than the English forum. He may, for example, have to accept lower damages, or do without the more generous English system of discovery. The same must apply to the system of court procedure, including the rules of evidence, applicable in the foreign forum. This may display many features which distinguish it from ours, and which English lawyers might think render it less advantageous to the plaintiff…. But that is not of itself enough to refuse a stay. Only if the plaintiff can establish that substantial justice cannot be done in the appropriate forum, will the court refuse to grant a stay…”

At para. [246] he emphasised that the court must have regard to the strong desirability of achieving comity, citing Lord Collins’ dictum in AK Investment CJSC v Kyrgyz Mobil Tel Ltd [2011] UKPC 7, [2012] 1 WLR 1804, at para. [79] that “Comity requires the court to be extremely cautious before deciding that there is a risk that justice will not be done in the foreign country by the foreign court, and that is why cogent evidence is required”.

He then expressed his conclusion in para. [247] that “the claimants’ evidence falls far short of establishing, upon sufficiently cogent evidence, that substantial justice cannot be done in Brazil”, before going on to explain that conclusion in the subsequent paragraphs, which contain the following main points.

The debate in the evidence about the relative ease or difficulty of suing BHP Australia in Brazil and via what alternative procedural routes “[came] close to generating a somewhat sterile debate … because no one is under any illusion that the claimants would ever be likely to contemplate bringing the proceedings in Brazil in any event” (para. [248]).

Because stage two required the court to consider “all the circumstances of the case” it was relevant to consider not simply a claim against the applicant defendant in the foreign jurisdiction, but also the option to achieve comparable redress in respect of the same loss and damage against another defendant in the foreign jurisdiction. The claimants had no interest in compensation being paid by BHP Australia rather than any other company within the corporate structure. They had not sought to “join” the two defendants within the Brazilian jurisdiction “because for perfectly understandable reasons they did not want to.” Therefore he “would place little weight upon the alleged additional challenges which would face the claimants were they to choose hypothetically to proceed against the English defendants in Brazil” (his emphasis, para.[249]).

Even if he were wrong on this last point it would not lead to a different conclusion. “Much time has been devoted to speculation about what may or may not be procedurally achievable, under the umbrella of the 155bn CPA or otherwise, but no efforts have been made by the claimants to take active steps to test the waters in this regard whether by approaching the prosecutor or otherwise. The evidence relied upon thus lacks the cogency to be expected under stage two of Spiliada” (para. [250]).

Delay was not a relevant factor in this case because he was entirely unpersuaded that proceedings in England would be concluded more promptly than would proceedings in Brazil, for reasons identified in numbered sub-paragraphs of para. [255].

The evidence did not establish that any impecuniosity on the part of the claimants would be a major factor in stifling legitimate claims; legal aid was available, and this was to be contrasted with the position in England in which claimants would be required to pay 30% of their winnings to their solicitors (para.[258]).

The claimants argued that the Judge’s error at stage one was in failing to analyse and apply Vedanta correctly. In particular, the Judge failed to take account of the “heavily qualified” nature of the defendants’ undertaking to submit to the jurisdiction of the Brazilian Courts, which was confined to the claims being advanced in a new CPA; and to take account of the fact that in this case both defendants had been sued as of right and were two halves of a single economic entity operating through a single management structure, with identical causes of action against each of them, as distinct from the position in Vedanta of an English parent as an “anchor” defendant. In the particular circumstances of this case the need to avoid two simultaneous and identical trials should plainly have been treated as a decisive factor. The “conspicuously close corporate relationship” which the Judge correctly recognised as existing made the separate trial of the claim against BHP Australia nonsensical.

At stage two, it was submitted, the Judge had made two errors of law. The first was that the Judge had held at paras. [236(b)], [244] and [247] that there was a burden on the claimants to show that they will not obtain justice in the foreign jurisdiction, whereas the burden is only to show a real risk that justice will not be done, as explained by Lord Collins in paras [89]-[95] of Kyrgyz Mobil. The second was in taking into account, at para.[249], the availability of redress against persons other than BHP Australia in Brazil.

The claimants further criticise the Judge for failing to engage with their evidence of the difficulties in suing the defendants in Brazil via a CPA or other proceedings, the limitations on the availability of legal aid, the likely delays which would be involved, and the obstacles to obtaining substantial justice which were said to be evident from the 19 witness statements.

It was submitted on behalf of BHP Australia that the fact that the claimants had made clear they were not interested in suing either of the defendants in Brazil if the applications failed, and would obviously have no interest in doing so against BHP Australia if the action proceeded here against BHP England, meant that their complaints were artificial. The Judge had correctly applied Vedanta at stage one in taking account of the risk of irreconcilable judgments, but holding that it was outweighed by the strength of the other connecting factors to Brazil. The defendants’ undertaking to submit to Brazilian jurisdiction was superfluous because it was common ground between the experts that the Brazilian courts had jurisdiction; an undertaking to submit to a new CPA was sufficient; and in any event further undertakings had been given in the course of the appeal to submit to all forms of proceedings in Brazil. The position of BHP England and BHP Australia is not materially distinguishable from that of Vedanta and KCM Mobile respectively in Vedanta.

In relation to stage two, the Judge obviously had in mind the test of “real risk” because it was articulated in the passage at para. [79] of Kyrgyz Mobil which he quoted. Even if he did not, he found that there was not even a real risk by reference to the evidence on delay and legal aid, or alternatively this court should make those findings on the evidence. Mr Gibson accepted that the relevant comparison was between redress available against BHP Australia in the two jurisdictions, with the result that the Judge was wrong to consider the availability of redress against other parties in Brazil; but he emphasised that the Judge made clear at para. [250] that he would reach the same conclusion without taking that into account. The Judge did address the claimants’ evidence about difficulties in bringing a new CPA but correctly held that there was a fundamental gap in it because the claimants had not “tested the water”; and it was also undermined because a large number of CPAs had been commenced since the dam collapse including by some of the claimants. The Judge’s conclusions on delay and legal aid were sound, and the Judge took into account all the relevant evidence.

The basic principles which apply where a defendant seeks a stay on forum non conveniens grounds of an action in which it has been served here as of right, were authoritatively identified in Spiliada and Kyrgyz Mobil. The defendant must discharge the evidential burden of satisfying the court that there is another available forum of competent jurisdiction which is clearly and distinctly more appropriate as the forum in which the case may be tried more suitably for the interests of all the parties and the ends of justice: Spiliada at pp. 476C, 476E, 477E. This is stage one. If the defendant satisfies the burden, the court will nevertheless refuse a stay if the claimant satisfies it, by cogent evidence, that there are circumstances by reason of which justice requires such refusal, including in particular if it is established by cogent evidence that there is a real risk that the claimant will not obtain justice in the foreign forum: Spiliada at p. 478D-E, Kyrgyz Mobil at paras. [91]-[95]. This is stage two.

There is a tension in the authorities as to how the two stages are to be applied where a foreign forum of competent jurisdiction has been identified by the defendant, but a question arises as to whether it is available to the claimant in practice. In particular, does stage one require the defendant to satisfy the court that the identified forum is available in practice? Or is that a question which only arises at stage two, with the burden resting on the claimant? The question is of potential significance in this case where the claimants rely on evidence that a new CPA, which constitutes the proceeding in the foreign court of competent jurisdiction upon which Mr Gibson relied as his primary case, is unavailable to them.

In Mohammed v Bank of Kuwait [1996] 1 WLR 1483, the claimant Iraqi national sued his employer, a Kuwaiti bank, which was amenable to service here through its London office. On the bank’s stay application, the first instance judge held that Kuwait was a clearly more appropriate available forum. One of the claimant’s arguments was that the Kuwait court was not a forum available to him in practice because he could not go to Kuwait (Iraqi nationals being banned following the liberation of Kuwait after the first Gulf War), and because he would have difficulty in instructing lawyers. The bank argued that these were matters for consideration at stage two, and that all that was required for the availability criterion at stage one was identifying a foreign forum in which the claimant could bring proceedings as of right. In rejecting this argument Evans LJ, with whom Saville LJ and Morritt LJ agreed, said at p.1490F:

“As to the meaning of ‘available’ as well as ‘appropriate’ forum … I would hold that the judge was correct to define the test as ‘available in practice to this plaintiff to have his dispute resolved’”.

He concluded that the stay application failed at stage one because the defendant had not established that Kuwait was an available jurisdiction in the light of the evidence that the claimant could not visit Kuwait, and he would face insuperable difficulties in instructing Kuwaiti lawyers to act on his behalf: see pp. 1495D-F, 1496F.

In Connelly, the House of Lords accepted that the defendants had discharged the burden at stage one of showing that the foreign forum, Namibia, was the appropriate forum, and addressed at stage two the significance of the fact that the claimant would be unable to fund his claim in Namibia but would in England: see per Lord Goff at p. 872B-D.

In Askin v Absa Bank Ltd [1999] I.L.Pr 471 (“Askin”) this court was concerned with a stay application by a South African bank which was amenable to suit here through its London office. South Africa was clearly the more suitable appropriate forum, but the claimant argued, amongst other things, that it was not available to him because were he to return to South Africa he was likely to be arrested and imprisoned on remand on criminal charges which had been brought against him, which would make it practically impossible for him to conduct a civil claim. Tuckey LJ, with whom Peter Gibson LJ and Robert Walker LJ agreed, referred at para. [27] to the decision in Mohammed and went on to say at para. [28] that whilst the decision could be explained on its special facts, what Evans LJ had said about “another available forum” had been criticised by learned commentators. The essence of the criticism was that it elided the two stages of the test, which were simple to apply; that Lord Goff could not have intended “available” to mean “available in practice” since his statement of the principle was derived from Lord Kinnear’s statement in the Scottish case of Sim v Robinow (1892) 19 R 665 at p. 668, which made no mention of “availability”; that at the first stage of the test the court was concerned only to identify and evaluate the connecting factors with the competing courts of competent jurisdiction, with questions relating to availability in practice arising only at the second stage; and that this was the approach Lord Goff had taken in Connelly where he considered the question of whether the Namibian courts were available to the claimant because he could not afford lawyers or experts there at the second stage. Tuckey LJ concluded at para. [29] that there was substance in these criticisms of Mohammed.

He went on at para. [29] to observe, however, that the point is only of any practical importance in a case which turns on the onus of proof. He stated that Askin was not such a case, and that “in a case which does [depend on onus of proof] it may be necessary for this court to consider whether its decision in Mohammed can stand with Spiliada and Connelly.”

In Cherney v Deripaska [2008] EWHC 1530 (Comm) Christopher Clarke J was concerned with an application for permission to serve out of the jurisdiction. He determined that Russia was the natural forum for the litigation (para. [254]), but that permission would be granted because the claimant had established a well-founded fear that if it took place in Russia (1) he would be at greater risk of assassination; (2) he would face criminal prosecution for what on the evidence would arguably be trumped up charges (paras. [257]); and (3) that there was a significant risk that any trial in Russia would be affected by improper influence of state actors such that justice would not be done (para. [260]). The relevance of the decision for present purposes is that (1) and (2) were treated as factors to be considered under the stage two inquiry, rather than going to availability at stage one. The judge explained this at para. [252], characterising them as “question[s] whether a foreign court is one in which justice may not be done” and treating Connelly in this regard as authoritative despite Mohammed being “an earlier Court of Appeal decision to the contrary.” Christopher Clarke J’s decision was affirmed on appeal without any express consideration of the point here being considered: [2009] EWCA Civ 849; [2010] 2 All E.R. (Comm) 456.

In these cases the courts were concerned with aspects affecting availability which were specific to the claimants’ personal circumstances, rather than being an aspect of the foreign jurisdiction itself. If it were necessary to choose between Mohammed and the later authorities, we would be inclined to treat the latter as authoritative and in line with Spiliada itself: these considerations are aspects of whether the claimant can get justice in the foreign forum of competent jurisdiction and are properly to be taken into account at stage two, not stage one, of the test.

However, not all of the factors upon which the claimants in this case rely as to the unavailability of the relevant forum in Brazil necessarily fall into a similar category. Questions as to the availability of a new CPA might be characterised as aspects of the CPA jurisdiction itself; and to go to the very question of whether there is an available foreign court of competent jurisdiction as a stage one question.

Nevertheless we are content to consider all the factors upon which the claimants rely in this case in the context of a stage two analysis. This is for three reasons. First, the case was argued before us on the basis that these were stage two considerations, and presumably also before the Judge, who addressed them as part of stage two. Secondly, as Tuckey J observed in Askin, it only matters in a case in which the onus of proof is critical, and this is not such a case. It makes no difference to our decision. Thirdly, we would agree with the editors of Dicey & Morris at 12-032 when discussing what is meant by “available” that:

“… the line which divides the two limbs of Spiliada from each other is neither completely impermeable nor drawn in such a way that there are no factors which do not appear on both sides of it: from time to time a court will locate under one limb of Spiliada material which, arguably at least, might more comfortably belong to the other. But when it is recalled that the overall test is one which asks what the interests of justice require, and when it is to be remembered that the analysis in Spiliada is designed to manage, rather than constrain, that test, it will rarely be a matter of legitimate complaint that this has happened.”

It is important to start by identifying with precision the alternative foreign forum relied on and the nature of the proceedings which render it available. It is not permissible merely to refer to the Brazilian courts in general, when there are different forms of proceedings in different courts, each of which raise their own questions of appropriateness and suitability. The Judge did not identify the form of proceedings he was considering in applying Spiliada.

The defendants’ primary case was that the alternative forum was a federal court in which all the claims of all the defendants could heard in a single action through the commencement of a new CPA, followed by multiple liquidation proceedings brought by individual claimants following a generic sentence. Mr Gibson maintained a fallback position that the alternatives were (a) federal courts hearing multiple CPAs brought by and/or on behalf of different claimants; or (b) federal or state courts in multiple ordinary or special civil claims brought by individual claimants.

We do not think that reliance on either of these two alternatives is realistic; and, to be fair to the defendants, they were not developed to any extent in the written or oral argument. Nor were they really supported by the defendants’ experts, who both implicitly recognised the disadvantages and limitations in seeking to advance the claims through individual lawsuits in a case of this kind where liability is in issue. They would all involve a multiplicity of liability proceedings resulting in proliferation of time effort and expense, an acute risk of inconsistent judgments, with few, if any of the advantages of case management and synergy of cost sharing and expertise available in a group action here. The small claims courts would not have expert evidence and would be overwhelmed. The deficiencies in such multiplicity of proceedings are inherent in the optional initiatives set up pursuant to TTAC/GTAC and are well articulated in the Baixo Guandu Judgment. In our view the defendants do not establish that these are clearly and distinctly more appropriate forums in which the case can more suitably be tried than in a single action in England, even taking into account the four Brazilian connecting factors identified by the Judge. This may be the reason why until shortly before the appeal the undertaking by the defendants to submit to Brazilian jurisdiction was confined to a new CPA.

We therefore proceed on the basis that the defendant’s only realistic case is its primary case that the federal court in an action commenced by a new CPA is the relevant alternative forum to determine issues of liability (and on its case generic issues of causation and quantum) followed by liquidation proceedings. That case does not depend on whether the new CPA could be consolidated with the 155bn CPA, although if it could, that would be a further factor to be considered.

We find it convenient first to address and apply the Spiliada tests to the claim against BHP Australia by ignoring the existence of a claim against BHP England. We then consider the effect, if any, of the fact that we have rejected the article 34 application (on grounds which are not dependent on the outcome of the forum non conveniens application), with the result that the claim will proceed against BHP England here in any event.

The only argument advanced by the claimants in respect of stage one of the test related to the application of Vedanta. That is of no relevance if the action against BHP England is being ignored.

As we have said, there are a number of points on the “availability” of a new CPA which were dealt with under stage two and which we will also address in that context. There is, however, one aspect of a new CPA which potentially has a direct bearing on the appropriateness of the forum in which it would be decided, and which in our view can only properly be addressed at stage one. It is the asymmetric nature of a new CPA: it will determine the rights of the parties if, but only if, it is favourable to the claimants; if it is unfavourable to them, it will not be determinative of anything and will not be the forum in which any aspect of the dispute between the parties will be resolved; rather the putative alternative “forum” would become, if the claimants wished to pursue their claims against BHP Australia in Brazil, the multitude of different fora in which ordinary civil proceedings fell to be pursued, which we have already identified as failing the stage one test.

We think it is also relevant in this context to consider the timescale involved. The defendants’ case is that a new CPA against BHP Australia could and should be consolidated with the 155bn CPA so as to be heard at the same time. There is real uncertainty as to when that will be resolved, with a real prospect that it may take more than a decade. If there were no consolidation it would be longer. Whilst delay is normally a stage two factor, it can in our view be relevant to stage one when the relevant foreign proceedings may not conclude in any binding resolution of any of the parties’ rights and obligations.

These are matters which were not raised or relied upon by the claimants, and upon which we were not addressed. We therefore reach no firm conclusions on them, and they do not affect the outcome of the appeal. However, our provisional view is that, even assuming the availability of a new CPA, its asymmetric nature and timescale would be fatal to the defendant’s application at stage one of Spiliada. The four cumulative connecting factors identified by the Judge would be powerful factors in favour of a stay in a case involving a single alternative forum in which the disputes would be resolved. However, given the “clearly and distinctly” burden on the defendant, it could be said that they are not sufficient to outweigh the disadvantages of a process which may last for over a decade or more before producing a result which may be one which does not advance the resolution of the dispute at all, and requires the dispute thereafter to be resolved in multiple fora which are clearly less appropriate.

We have set out at paras. [113]- [122] above our conclusion that the claimants have established by cogent evidence that there is a real risk that a new CPA against the defendants is not an available route for any of the claimants in Brazil. This means that the only realistic alternative forum identified by the defendants, as the one in which the case can more suitably be tried, is one in which there is a real risk that the claimants cannot obtain substantial justice. The consequence is that justice requires that a stay should be refused.

The Judge dismissed the significance of this evidence on the grounds that (1) what was relevant was redress against other parties, not just BHP Australia (para. [249]); (2) the claimants would not be likely to bring proceedings against BHP Australia in Brazil anyway (para. [248]); and (3) that the claimants’ evidence lacked cogency because they had failed to take active steps “to test the waters in this regard whether by approaching the prosecutor or otherwise” (para. [250]).

We consider that the Judge’s approach suffered from a number of errors.

First, we accept the claimants’ submission that the Judge applied the wrong test at stage two by asking whether the claimants would not obtain substantial justice, rather than whether there was a real risk of them not doing so. This was how he formulated the legal test at paras. [236(b)] and [244], citing the speech of Lord Goff in Spiliada without the modification expounded in Kyrgyz Mobil; and this was the way he expressed his conclusion at para. [247] in terms of the claimants’ failure to establish upon sufficiently cogent evidence “that substantial justice cannot be done in Brazil”. The fact that the previous paragraph (para. [246]) which he quoted from para. [79] of Kyrgyz Mobil included a reference to a “risk that justice will not be done” does not detract from that conclusion. Para. [79] was quoted by him on an entirely different point, namely to support the proposition that the court must have regard to the strong desirability of achieving comity. Had he taken from it the proposition that the relevant test was one of risk of injustice, he would no doubt have referred to the passages in Kyrgyz Mobil at paras [91ff] which addressed that question when articulating the principles he derived from Spiliada at para. [236] or elsewhere; and he would not in the very next paragraph have expressed his conclusion in the terms he did.

Secondly, the Judge erred in taking into account the redress potentially available against other parties in Brazil. That was not a relevant consideration, as Mr Gibson accepted; but as he correctly submitted it would not be fatal to the Judge’s reasoning because he went on to say that he would reach the same conclusion without taking it into account.

Thirdly, the view expressed at para. [248] that the claimants would not be likely to commence proceedings in Brazil against BHP Australia may well be justified, but is not here relevant in circumstances where there was cogent evidence from the claimants’ experts, which is that upon which they could reasonably be expected to act, namely that there were insuperable obstacles to such a course in respect of a new CPA. As a matter of principle, if a claimant establishes by cogent evidence that there is a real risk that he cannot obtain substantial justice in the identified foreign forum because, as he is advised, it is unavailable to him, it should not be an answer that he has not sought to embark upon that course, and is unlikely to, but rather has sued the defendant where the latter is amenable to service here, where substantial justice can be done.

It is possible that the Judge had in mind the position if the article 34 application of BHP England failed, which he addressed at para. [242]. We agree with his conclusion there that if the forum non conveniens application of BHP Australia were to fail, and the action would be proceeding against BHP England alone in England, the strong likelihood is that the claimants would not sue BHP Australia in Brazil. But if this was what he had in mind at para. [248] as a reason for dismissing the evidence of the obstacles to a new CPA in Brazil, it would involve putting the cart before the horse: if there is a real risk that a suitable forum is not available to the claimants against BHP Australia in Brazil, the forum non conveniens application should fail irrespective of the fate of the claim against BHP England.

Fourthly, the criticism that the claimants had not “tested the water” is for similar reasons an irrelevant consideration; and is in any event, in our view, an unfair one. Given the claimants’ expert evidence, it was entirely understandable that they had not “tested the water” by seeking to persuade the MPF to commence a new CPA, or by bringing a new CPA or CPAs themselves. The Judge was wrong to treat the failure to test the water as having any bearing on the cogency of their expert evidence on this issue.

It is, of course, well established that the fundamental question of the appropriate forum is a matter within the discretion of the judge; and that an appellate tribunal can only interfere with the exercise of the discretion in accordance with well-settled principles: see amongst many other cases Kyrgyz Mobil at para. [139]. For the above reasons, those well-settled principles are engaged in this case.

Accordingly we conclude that the forum non conveniens application must fail at stage two, if it does not fail at stage one, irrespective of the outcome of the article 34 application.

We will, nevertheless, go on to consider the impact, if any, of our conclusion that the article 34 application fails with the consequence that the action will be proceeding against BHP England here in any event. We can do so relatively briefly in the light of the fact that it is not determinative of the application.

The only additional points which fall to be considered in this context arise at stage one and depend upon the proper application of Vedanta. The stage two issues have already been addressed and are unaffected by the existence of the continuing action here against BHP England.

We have already cited the important passages from the judgment of Lord Briggs in Vedanta. It is authority for the proposition that where a defendant is amenable to suit in a foreign forum which is the appropriate forum for the claim against it, it is not a trump card, although it remains a relevant consideration, that the claimant asserts a risk of irreconcilable judgments arising from his pursuit of another defendant here as of right. In Vedanta itself, such risk was held not to be sufficient, on the facts, to prevent the foreign forum being the appropriate forum.

The Judge cited the relevant passages from Vedanta, recognised that the risk of irreconcilable judgments was a relevant factor but not a trump card, and concluded at para. [241] that the factors in favour of Brazil as the appropriate forum outweighed the risk of irreconcilable judgments where the claimants had a choice to sue BHP England in Brazil. On the assumption that Brazil was an available forum for such a claim, that was a conclusion he was entitled to reach in the exercise of his discretion and we should not interfere with it unless persuaded by the claimants that there was some error of principle in his interpretation or application of Vedanta.

None of the points advanced by the claimants reveal any such error of principle.

The qualified nature of the defendant’s undertaking to submit to the jurisdiction of the Brazilian Courts, being limited to a new CPA, was irrelevant. The experts agreed that all Brazilian courts had jurisdiction over claims against BHP England; in any event the undertaking offered is no longer so qualified.

The fact that BHP Australia is amenable to suit here through its London office, so that it is a stay case rather than one of service out of the jurisdiction, as it was in Vedanta, is an immaterial distinction. The same principles apply in either case, subject only to a difference in the burden of establishing the clearly more appropriate forum. The part which reliance on a risk of irreconcilable judgments plays in that analysis is no different in either case.

Nor is it a relevant distinction that the claim against BHP Australia is materially identical to that against BHP England. The extent of overlap between the claims in the two fora will inform the extent of the risk of irreconcilability. But the risk of irreconcilability was present in Vedanta to a very substantial degree, and there is no material distinction from the extent of the risk in this case.

The argument that it would be “nonsensical” to have a separate trial of the claims against BHP Australia, is no more than repackaging of the point which was rejected in Vedanta. This is not a trump card, and it was taken into account by the Judge.

Moreover we consider that the risk of irreconcilable judgments was an entirely irrelevant consideration, for the reasons given by the Judge at para. [242], namely that if the forum non conveniens application of BHP Australia were to fail, and the action would be proceeding against BHP England alone in England, the strong likelihood is that the claimants would not sue BHP Australia in Brazil. It does not appear from his Judgment that he regarded this as an independent reason why a risk of irreconcilable judgments was irrelevant at stage one, but in our view it is conclusive. The factual premise, that the claimants would not sue BHP Australia in Brazil, is not in this context dependent on any difficulties or obstacles in so proceeding, but rather is simply the logical result of the two claims being materially identical: there would be no advantage in the claimants additionally seeking relief against BHP Australia, in Brazil, if identical relief were to the same extent available in proceedings against BHP England which are to be pursued in any event. In those circumstances there would be no risk of irreconcilable judgments in the absence of a stay, because there would be no foreign proceedings, as the direct result of pursuit of the claim against an adequate alternative defendant in England.

For these reasons the forum non conveniens application must be dismissed and the appeal in respect of it allowed.

The court has a discretion to order a stay to await the outcome of foreign proceedings as part of its case management powers pursuant to s. 49(3) Senior Court’s Act 1981 and CPR 3.1(2)(f). It will only exercise it in rare and compelling circumstances (Reichhold Norway ASA v Goldman Sachs International [2000] 1 WLR 173, 186C). It cannot do so in a manner inconsistent with Brussels Recast (see e.g. Mazur at para. [69]).

For the reasons we have given when addressing the arguments on the abuse and art 34 applications, a stay is not in the interests of justice in this case, which dictate that the claimants should be permitted now to proceed with the claims in the action. Moreover a case management stay of the claim against BHP England would be inconsistent with article 4 and article 34 of Brussels Recast.

Accordingly the case management stay application must be dismissed (as the Judge recognised would be the result if his conclusions on abuse were not upheld). The appeal is in that respect, as others, allowed.

We would accordingly allow the appeal and dismiss each of the applications.

Each of the Individuals has suffered and/or is suffering ongoing loss and damage for which he or she is entitled to be compensated under one or more of the following heads:

Each of the Businesses has suffered and/or is suffering ongoing loss and damage for which it is entitled to be compensated under one or more of the following heads:

Each of the Faith-Based Institutions has suffered and/or is suffering ongoing loss and damage for which it is entitled to be compensated under one or more of the following heads:

Each of the Municipalities has suffered and/or is suffering ongoing loss and damage for which it is entitled to be compensated, either on its own behalf or in respect of its citizens or the community, under one or more of the following heads:

Each of the Utility Entities has suffered and/or is suffering ongoing loss and damage for which it is entitled to be compensated under one or more of the following heads: