On 5 November 2015, the Fundão Dam in southeast Brazil collapsed, killing 19 people, destroying villages, and causing widespread destruction and damage. One estimated figure for the remediation costs and compensation has been put at £25 billion. The dam was owned and operated by Samarco Mineração S.A. (“Samarco”), a joint venture company owned by the appellant (“Vale”) and BHP Brasil Ltda (“BHP Brazil”). The second respondent in these proceedings is the ultimate parent company of BHP Brasil. I shall refer to the respondents collectively as “BHP”.

Arising out of this disaster, there are different sets of civil proceedings in Brazil, both individual claims and class actions, which have been brought against Samarco, Vale and BHP Brasil. There have been very few, perhaps a dozen, claims in Brazil against BHP. On 2 and 5 November 2018, around 200,000 claimants issued proceedings against BHP in this jurisdiction and on 3 May 2019, a further claim form was issued against both BHP defendants. On 24 February 2023, a new claim form was issued against BHP, increasing the total number of claimants to approximately 732,000.The claims are all advanced under Brazilian law. BHP originally obtained an order that the claims should be struck out an abuse of process and this jurisdiction was not the appropriate forum for those claims, but that order was overturned by the Court of Appeal. Subsequently BHP have sought to join Vale into these proceedings as a Part 20 defendant. Vale challenged the jurisdiction of the English Court on the basis that there was no serious issue to be tried, and that this jurisdiction was not the appropriate forum for the Part 20 claim.

By a judgment dated 7 August 2023 ([2023] EWHC 2630 (TCC)) O’Farrell J (“the judge”) refused those applications. In September, Vale sought permission to appeal (“PTA”) against her decision, although they had not first sought permission from the judge. I directed that they do that first. The judge refused PTA on 10 October 2023 and also refused Vale’s application for a stay of the Part 20 claim. The application for PTA then came back to me in late October, albeit in a very different form to the September application. I called in the application for PTA to ensure the necessary efficiency and speed, given the judge’s other directions and the trial date fixed for the Autumn of 2024. The hearing of the PTA application took place on 15 November 2023.

The focus of the PTA application is the appropriate forum for the Part 20 claim. The leading authorities are Spiliada Maritime Corp v Cansulex [1987] AC 460 (“Spiliada”), AK Investment CJSC v Kyrgyz Mobil Tel Limited [2011] UKPC 7, [2012] 1 WLR 1804 (“Altimo”), and Vedanta Resources Plc v Lungowe UKSC 20, [2020] AC 1045 (“Lungowe”). All are referred to and analysed in the judge’s judgment.

The application for PTA does not suggest, save for one possible exception made orally, that the judge failed to identify the correct principles from these cases. The complaint is that she did not apply them correctly to the facts: Ground 2A, which is at the heart of the PTA application, is headed “Failure correctly to apply the Spiliada principles”. That is a criticism of the judge’s evaluation exercise, and so is not a promising starting point for any application for PTA. The general position has been recently restated by Lewison LJ in Volpi v Volpi [2022] EWCA Civ 464; [2022] 4 WLR 48 as follows:

These propositions have been recently restated in an appeal concerned with the application of the Spiliada principles. In Samsung Electronics Co Ltd & Ors v LG Display Co Ltd [2022] EWCA Civ 423, Males LJ said:

Whatever the precise formulation of these propositions, they necessarily amount to a high bar. The civil justice system requires first instance judges regularly to undertake the sort of multi-faceted evaluation which the judge undertook here, and this court will only grant permission to appeal if there is a real prospect of demonstrating that the judge was “plainly wrong, in the sense of being outside the generous ambit where reasonable decision makers may disagree”: see Global Torch Ltd v Apex Global Management Ltd (No.2) [2014] UKSC 64, [2014] 1 WLR 4495 at 4500. For the reasons set out below, I am firmly of the view that Vale has no prospect of persuading the full court that they have cleared – or even got close to clearing – that high bar.

Ground 1 was originally very detailed, but the vast majority of the points originally taken in the Notice of Appeal have been deleted. The only surviving point is the complaint that BHP continue to claim a “contribution”, which is not a concept known to Brazilian law, and that, because BHP have no cause of action to recover payment until they themselves have made payment to the claimants, the only relief BHP are entitled to claim in the Part 20 proceedings is a declaration.

In my judgment, there is nothing in this remnant of Ground 1. I note that, at the hearing on 10 October 2023, various amendments were proposed by BHP to reflect the agreed position that BHP must make payment to the claimants before seeking any reimbursement from Vale. As the judge said at that hearing, the continuing references to “contribution” stem from the potential difficulties caused by transposing concepts of Brazilian law into the English law framework. But whether BHP are right to say that they are entitled both to a declaration and to an order that contribution/ reimbursement should be ordered in respect of any sums paid to the claimants, the nature of BHP’s case is clear, and the precise nature of the relief claimed is immaterial to the application for PTA.

The important point is that, on any view, it cannot now be said that there is no serious issue to be tried. Precisely how the relief is framed is irrelevant to that question; even if Vale is right, and the only claim open to BHP is a claim for a declaration, that can have no material effect on the outcome of the dispute as to the appropriate forum. For those reasons, Ground 1 of the proposed appeal goes nowhere. In all the circumstances, and in the light of what was said and decided at the hearing on 10 October, I consider that Ground 1 has no real prospect of success.

I consider this to be the heart of the application for PTA. This ground of appeal is expressly based on the “failure correctly to apply the Spiliada principles” (emphasis added). Accordingly, the Volpi/Samsung approach is directly applicable. Moreover, when considering this ground, it is important to consider the judgment as a whole, and not simply to isolate particular paragraphs or even individual sentences. So, although the relevant analysis in the judgment is at [82]-[100], it is important to consider those paragraphs in the context of the lengthy preceding paragraphs setting out the background and some of the law. It is also important to read those particular paragraphs together and to avoid the sort of minute textual analysis which Lewison LJ deprecated in Volpi.

Although it did not feature in his skeleton argument, during his oral submissions, Mr Salzedo KC took issue with the third of the four applicable principles identified by the judge at [84]. Although he also referred to the incorrect reference to [74] of Lungowe which was cited in relation to the fourth principle, it was clear that the judge intended to refer to [70] not [74], and Mr Salzedo did not take issue with the principle itself.

The third principle which the judge identified at [84(iii)] was:

In my view, the judge’s summary of the principle to be derived from [68] of Lungowe was correct. Lord Briggs expressly referred to “the interests of all the parties” as referring to the case as a whole, and therefore including all parties, including, in that case, what was referred to as the anchor defendant. Later on, as he put it, “the court is looking for a single jurisdiction in which the claims against all the defendants may most suitably be tried”. That was the principle summarised by the judge. I reject as unarguable the suggestion that the summary at [84(iii)] was somehow an invention of the judge’s own.

We are therefore back to the judge’s evaluation exercise: how should the Spiliada principles be applied to the facts of this case? Vale’s application for PTA raises four alleged errors under Ground 2A, and I address each in turn.

The argument is that the judge failed to acknowledge or apply what is said to be the important statement of principle in Altimo at [73], that “…it must never become the practice to bring foreign defendants here as a matter of course, on the ground that the only alternative requires more than one suit in more than one different jurisdiction” (Footnote: ¹).

However, as the judge correctly said when she refused permission to appeal, she never at any time suggested that this was the practice, let alone the practice that she adopted in this case. Instead, the judge was very careful to analyse the particular balancing factors in this complex case, one against the other. The words before the passage in Altimo on which Vale relies expressly state that “caution must always be exercised in bringing foreign defendants within our jurisdiction”. The judge exercised the necessary caution in her analysis.

Furthermore, I consider that Error 1 not only tilts at a windmill that is not there, but adopts an unrealistic approach to the particular issues in this case. As Lord Briggs noted at [70] of Lungowe, “the avoidance of irreconcilable judgments has frequently been found to be decisive in favour of England as the proper place, even in case where all the other connecting factors appear to favour a foreign jurisdiction”. It is never a question of practice or routine; but it some cases, a consideration of actual and/or hypothetical proceedings in another jurisdiction is required as a matter of practicality. This was one such case.

Finally, there is a strong similarity between Vale’s argument on this point, and one of the unsuccessful arguments in Samsung, addressed at [29] of Males LJ’s judgment. He said that, just as here, the judge made no decision about where contribution claims should usually be heard, and focussed instead on the particular contribution claim in those proceedings. As to that, Males LJ said “he [the judge] recognised that if the underlying claim is proceeding here…that will be a powerful and sometimes overwhelming factor in favour of hearing the contribution claim here. I would endorse that view”. Even allowing for the fact that Vale contends that the Part 20 claim here cannot, in Brazilian law, be a contribution claim as such, it is in substance such a claim, and the same “powerful and sometimes overwhelming factor” is therefore on any view a relevant factor in the balancing exercise in the present case.

For those reasons, I consider that the arguments based on Error 1 have no real prospect of success.

The complaint is that the judge failed to find that the position in Altimo was directly analogous to the present case. In my judgment, there are three complete answers to that.

First, the judge was not obliged to go through every authority, and identify whether or not each was comparable, in whole or in part, to the present case. Secondly, it is not said (other than Error 1, which I have rejected) that the judge failed to apply any of the principles in Altimo. Thirdly, Altimo was, on any view, not directly analogous with the present case. In Altimo, the parties were agreed that Kyrgyzstan was the natural forum for the pursuit of their claims: the issue was whether the counterclaiming defendants could get a fair (or indeed any) trial in Kyrgyzstan. That issue does not arise here. Furthermore, in Altimo, it was found that if there was no trial in the Isle of Man, there would be no trial anywhere. That issue does not arise here either. Altimo was therefore a completely different case, with a different result to the one urged by Mr Salzedo. To put the point another way, Lord Collins said that Altimo was an unusual and complex case, so analogies with it will be rare. This case is also unusual and complex, albeit for very different reasons.

For these reasons, I consider that the arguments based on Error 2 have no real prospect of success.

The first of Vale’s principal complaints is that the judge took an erroneous starting point in her analysis, because she put the fact that there were claims against BHP in this jurisdiction over and above all other relevant factors. That is said to have skewed the conclusion of the Spiliada balancing test. The particular paragraph of the judgment that came under fire in this context was [87], where the judge said:

The fundamental difficulty with this argument is that it concentrates on [87], and makes no reference to either the preceding or subsequent parts of the judgment. I have already indicated [84], where the judge set out the applicable principles, and would in this context emphasise the agreed principle at [84(ii)], that the court must identify “the forum in which the case could most suitably be tried for the interests of all parties and for the ends of justice”. Moreover, the judge was very aware of all the proceedings involving Vale in Brazil. Those are identified at [9]-[15], and again at [92]-[98]. Against that, at [88] and [89], the judge addressed the claims against BHP in England and the significant “overlap” between those claims and the Part 20 claims which BHP now seek to bring against Vale. The judge therefore plainly had in mind both the Brazilian and the UK sets of proceedings. In undertaking the Spiliada balancing test, the judge was entitled to ascribe particular weight to some factors and less weight to others. That was an essential part of her evaluative decision.

It is unhelpful to focus simply on [87] of the judgment, and to separate out one particular sentence within that paragraph so as to argue that in some way the judge prioritised the relevance of the English proceedings above everything else. All the judge was saying at [87] was that BHP’s claims against Vale had to be seen for what they were: not fresh, standalone claims (where the natural forum would be Brazil) but Part 20 claims which were part of substantial proceedings brought by 732,000 claimants against BHP. That was a reflection of the same point made in Samsung (paragraph 19 above).

In my view, this summary provides a complete answer to alleged Error 3. The judge identified all of the relevant elements as part of the Spiliada balancing exercise and, as she was entitled to do, gave more weight to some elements and less to others. That cannot be an arguable ground of appeal.

To the extent that it matters, I should add that the arguments about Error 3 also ignore the detailed exercise that the judge actually undertook. She acknowledged the various proceedings in Brazil, but she made three very important points about them at [97] and [98]. First, she noted that there was no single set of ongoing proceedings in Brazil that would determine liability in respect of the dam collapse for a significant proportion of the 732,000 claimants. The proceedings in Brazil were scattered across different claimants and different jurisdictions and the evidence suggested that they involved far fewer claimants than in this jurisdiction. They could themselves lead to a series of different and conflicting results, regardless of the claim in this jurisdiction, but in contrast to the claimants’ claim here, none of them could be regarded as a main or majority or lead action.

Secondly, following on from that at [98], the judge made the point that none of the Brazilian proceedings were identified as proceedings to which BHP could be made a party, in an attempt to avoid further multiplicity. In those circumstances, as the judge said, BHP would be forced to issue fresh proceedings against Vale in Brazil. It was plainly a relevant factor in the balancing exercise that, if BHP’s claims against Vale were not to be heard as part of the claim in this jurisdiction, entirely fresh proceedings would have to be started in Brazil. That led to the judge’s third point as to the risk of inconsistent decisions (which she indicated would be increased if BHP’s claims against Vale were not dealt with in this jurisdiction), and which I address in greater detail under Error 4 below.

Accordingly, for these reasons, I consider that the arguments based on Error 3 have no real prospect of success.

The second of Vale’s principal complaints is that the judge failed to give sufficient weight to the risk to Vale of inconsistent judgments if the additional claim proceeded in England, because of the existing Brazilian claims. I disagree.

The fact that there were many different claims in Brazil, in many different places, featuring many different claimants, meant that there was always some risk to Vale of irreconcilable judgments within Brazil itself. Prima facie, that risk was not greatly exacerbated by the claimants’ claims in this jurisdiction; on the contrary, the limited overlap between the underlying claims in Brazil and the claims in this jurisdiction had formed part of the reasoning of the Court of Appeal when they rejected BHP’s own jurisdiction application, a point the judge made at [95]. But the risk to Vale of irreconcilable decisions would be increased if BHP’s Part 20 claims against Vale did not proceed in this jurisdiction, because there would then have to be new and separate proceedings between BHP and Vale in Brazil, in addition to the existing claims.

As the judge saw this aspect of the exercise, the comparison was this: on Vale’s case, BHP would have to defend proceedings in this jurisdiction and bring a separate claim against Vale in Brazil, with the real risk to BHP of inconsistent findings. But fresh proceedings in Brazil between BHP and Vale would also increase the risk to Vale of inconsistent results within Brazil (because there were many individual claims already, and this would add another, different sort of claim). By contrast, on BHP’s case, if their Part 20 claims against Vale were determined as part of this 732,000 claimant action, there would be one consistent result binding the vast majority of the claimants, BHP and Vale. In those circumstances, the judge was entitled to conclude that Brazil would not be “the forum in which the case could most suitably be tried for the interests of all parties and for the ends of justice”, and that this jurisdiction would be such a forum.

I consider therefore that the judge considered the risk to Vale of irreconcilable decisions as part of the balancing exercise, and reached a view which, at the very least, she was entitled to reach. The arguments based on Error 4 have no real prospect of success.

Although, on behalf of Vale, Mr Salzedo had points - many of them very minor - about some parts of paragraphs [87] and [100], it is unnecessary to deal with them one by one. The judge was not obliged to, and could not in any event, address every element of the relevant evidence. She identified some of the points that she considered important. The reference at [96] to “only” about 5,000 claims still live in Brazil, which Mr Salzedo criticised, was not a point of criticism at all: that number was in marked contrast to the 732,000 claims live in these proceedings. To my mind, none of these points led to any arguable way in which the judge’s analysis might be questioned, let alone set aside.

It is important to stand back and consider the judge’s evaluation in the round. She saw that this was a particularly complex case, and that the Court of Appeal’s refusal of BHP’s earlier application meant that there was always some risk of irreconcilable decisions between this jurisdiction and Brazil. She had to apply the Spiliada test in those unusual circumstances. The facts that the proceedings in this jurisdiction, uniquely, could determine the liability of BHP and Vale in respect of the majority of claimants, and that the alternative was for BHP to commence yet further proceedings in Brazil in order to establish Vale’s liability to them led the judge to conclude that the best course was for the claims against Vale to proceed as additional claims in these proceedings. She applied the right principles to the particular facts of this case. She reached a conclusion on that issue that she was entitled to reach.

The argument is that the judge was wrong to conclude that Vale could be ready to participate in the threshold liability issues which are due for trial in the Autumn of 2024. In my view, given both the history and the discretionary nature of the judge’s case management orders in respect of this aspect of the case, this argument has no prospect of success.

There was a three-day CMC on 29 and 30 March 2023 and 20 April 2023 at which both BHP and Vale were represented. BHP were seeking an adjournment of the trial date, then fixed for April 2024, to 2025. Vale’s position, as recorded by the judge at [45] of her May judgment, was that it was entitled to refuse to engage on the substance of the claim prior to the resolution of its jurisdictional challenge (the hearing of which had by then been listed for July 2023). Vale made it plain that any determination of that issue would be too late for Vale to participate in the trial in April 2024. I was told during argument that they supported BHP’s application for an adjournment until Spring 2025, although I note that their skeleton argument for the CMC said that they were neutral.

As is often the way, the judge reached a compromise and adjourned the trial, but to the Autumn of 2024, not the Spring of 2025 (Footnote: ²). That was always a real possibility, not only because such halfway houses are commonplace in such case management disputes, but also because it was expressly raised with the parties. Indeed, I understand that it was in that context that Vale made clear their potential representation difficulties if the trial was adjourned to Autumn 2024.

In her May case management judgment, the judge made plain at [82] that one factor in favour of the adjournment was that it would “give the parties a more relaxed, achievable timetable and would allow Vale and others to participate if necessary”. So there can be no doubt that one of the factors that led to the adjournment to Autumn 2024 was to accommodate BHP’s claims against Vale.

Although Mr Salzedo referred to the previous paragraph of the same judgment, where the judge said that “it would not be appropriate for the court to manage this large, heavy litigation around Vale or any possible appeals”, and suggested that that was inconsistent with the later conclusion, I disagree. The judge was there simply making the point that she could not manage this entire litigation around Vale and its appeals. That was sensible and correct. But that did not mean that she was not prepared to adjourn the trial so as to ensure (doubtless amongst other things) that Vale could participate if necessary; on the contrary, that is precisely what she did.

In her principal judgment of August at [99] the judge reiterated that point, saying that “the threshold liability trial has been fixed for hearing in October 2024 allowing sufficient time for Vale to prepare its case and participate in the common issues”. That was her answer to Mr Salzedo’s submission that the court “could not assume” that Vale could be ready for such a trial. Although he made some further submissions purporting to flesh this out, they were made without substantial evidence and only in reply. Given the history, the judge was entitled to reach the conclusion she did at [99].

It may be that Vale woke up to this point rather late, because it chose to put in further evidence as to its alleged timetabling difficulties at the consequentials hearing in October. On the face of it, that was much too late: that ship had sailed at the time of the April judgment, and certainly by the time of the August judgment. None of that later material can be relevant to the judge’s conclusion in her May judgement, reiterated at [99] of her August judgment, that Vale could be ready to participate in the trial in Autumn 2024.

The directions that the judge has so far made in relation to the threshold trial (and Vale’s participation in it) are in any event relatively limited. Vale has to serve their acknowledgment of service and their defence to the claim by 1 December 2023. Vale must provide a draft disclosure review document by 15 December 2023. Vale and BHP must seek to agree additions to the current list of issues for the first stage trial by 12 January 2024. All other directions are to be considered at the CMC in January. By then, the judge will have heard Vale’s separate application to stay the Part 20 claims for arbitration.

In my view, it follows from this history that Ground 2B has no prospect of success. The judge carefully considered when BHP and Vale could be ready for a first stage trial and, as long ago as May 2023, identified a date in the Autumn of next year. That decision was not the subject of any appeal. It is now much too late for Vale to complain about that order and the directions made thereafter. In any event, those were case management decisions which would not be revisited by this court; they were matters for the judge, and no-one else.

By late January 2024, Vale will have complied with the various directions identified above. The outcome of their application to stay for arbitration will probably also be known. The CMC in late January will therefore be the perfect opportunity for considering the precise shape of the threshold liability trial in the Autumn of 2024 and the detailed directions leading to it. Vale could even apply for an adjournment of the first stage trial at that stage, although I imagine that might be met with opposition from the claimants. It would be wrong in principle, and wholly unjustified, for this court to interfere in this ongoing case management process.

For those reasons, I regard Ground 2B as wholly unarguable.

Ground 3 raises two separate points about the judge’s orders as to the acknowledgments of service, and Vale’s continued participation in the trial preparations, notwithstanding the outstanding application for PTA in respect of jurisdiction.

These two points are predicated on the basis that Vale has an arguable case in respect of jurisdiction. For the reasons that I have already set out, it do not have such a case. In those circumstances, it is unnecessary to say anything more about Ground 3; it falls away following the refusal for permission to appeal on Grounds 2A and 2B.

However, I should make this observation. Vale’s stance, as demonstrated by Ground 3 (Footnote: ³), was that, unless and until their jurisdiction argument had been finally refused with no further right of appeal, it could rely on the very existence of that application to avoid substantial participation in the ongoing proceedings. I do not accept that. It would mean that a defendant could always drive this sort of complex multi-party litigation into the ditch by taking a threshold point and sticking to it, avoiding any engagement with the case management process until all appeal rights had been exhausted. That is not how case management in the Business and Property Courts is designed to work. Of course, parties in the position of Vale are entitled to take whatever threshold points they consider appropriate. But they also have to plan for the possibility that their threshold position will fail, and that they may ultimately be found to be the proper subject of the court’s directions. It is therefore incumbent upon them to provide some level of cooperation with the other parties and the court pending the final resolution of the relevant application. Otherwise a party in the position of Vale could achieve what they wanted – in this instance, not being an active participant in the first stage trial in 2024 – by reference to a jurisdiction argument which is unfounded.

For these reasons, I have concluded that this application for permission to appeal has no real prospect of success. I therefore refuse PTA. I am grateful to both parties for attending on 15 November: it was the only way in which the application could be resolved efficiently and within the tight timetable required by the judge’s many other orders.

I agree.