One of the reasons that international parties choose arbitration in England is finality. The Arbitration Act 1996 provides very limited grounds for appeal and these are often restricted further by the parties' agreement or by their choice of arbitral rules. Once the parties have chosen arbitration, the English courts are generally reluctant to interfere; whether during the arbitration or after an award has been made.

For the winner, the benefits are obvious. But for both parties, there is often some commercial sense in having disputes resolved comparatively quickly, rather than have the uncertainty and expense of consecutive appeals to higher courts (which in some jurisdictions can entail full retrials) hampering the parties' abilities to 'move on'. That is not to say, however, that the parties are stuck with whatever the arbitrator decides, no matter how bizarre or irrational. Arbitrators can make mistakes and sometimes there will be real grounds to appeal. In those cases the courts will intervene.

The 1996 Act permits an award to be challenged (ie, appealed) on the basis of:

  • the arbitrators' lack of substantive jurisdiction (s.67);
  • a serious irregularity affecting the arbitrators, the proceedings or the award (s.68); or
  • a point of law (s.69).

The third of these grounds is often excluded, either by agreement or by the institutional rules governing the arbitration (see,e.g., Art. 26.9 LCIA Rules, and Art. 34.6 ICC Rules). In contrast, the first two grounds are mandatory (s.4(1) and Schedule 1 of the 1996 Act). Even so, it is probably fair to say that appeals on the basis of the arbitrators' lack of jurisdiction are less common once a substantive award has been made because this right of appeal may be lost if jurisdiction is not challenged promptly. Often this means at an earlier stage of the arbitration. As such, it is the second ground – a serious irregularity – which is most often invoked in appeals against substantive awards. This might also be explained on the basis that s.68 appears on its face to be quite broad and it is usually possible to couch almost any kind of complaint in terms which, at least superficially, fall within its scope.

Needless to say, however, challenging an award under s.68 and doing so successfully are two quite different things. The section defines a list of nine circumstances which can constitute a "serious irregularity" (s.68(2)) and each has generated its own body of case law.

A very recent case before the English High Court have provided some helpful clarification about what will and what will not be sufficient in an appeal under one of these sub-sections, namely: "failure by the tribunal to deal with all the issues that were put to it," (s.68(2)(d)).

Petrochemical Industries Company (K.S.C.) v The Dow Chemical Company1 was an appeal, from the award in a US$ multi-billion arbitration before a panel of arbitrators, comprising the former head of chambers at one of the leading English arbitration sets, a former Judge of the Iran-US Claims Tribunal in the Hague and a former English Law Lord.

The Arbitration and the Award

The arbitration concerned the liability of the Claimant, Petrochemical Industries Company (K.S.C.) ("PIC") for losses suffered by the Defendant, The Dow Chemical Company ("Dow"), as a result of the withdrawal of the former from a US$ 7.5 billion corporate transaction. The tribunal awarded US$ 2.05 billion in damages for consequential losses incurred by Dow in re-financing another, separate corporate acquisition. PIC contended that Dow was not properly entitled to recover such losses because they were too remote from the alleged breach of contract; ie, being neither losses arising naturally in the usual course of things, nor in the reasonable contemplation of the parties as a result of special circumstances known to both of them (respectively, the first and second limbs of the test in the well-known case of Hadley v Baxendale).2 PIC had further argued that even if such losses did fall within the scope of either of the limbs of Hadley v Baxendale, they still ought not to be recoverable because there was specific evidence that PIC had not assumed responsibility for this type of loss. It was this latter argument which gave rise to the appeal. PIC claimed that although the tribunal had addressed the 'Hadley v Baxendale question', it had neglected to take into account PIC's arguments and evidence regarding 'assumption of responsibility'. As such, PIC contended that there was a serious irregularity affecting the tribunal, because it had "fail[ed]… to deal with all the issues that were put to it" and, as a result, PIC had suffered substantial injustice (s.68(2)(d) of the 1996 Act). Accordingly PIC sought an order that part of the award be remitted back to the tribunal for reconsideration.

The High Court's Judgment

The English High Court judge disagreed. In reaching his conclusion, he provided some helpful commentary on the court's approach to an appeal under s. 69(2)(d). He framed the test in three parts, with a fourth question which would have to be answered if the first three parts were satisfied, as follows:

"i) Whether the assumption of responsibility question was an 'issue'

ii) If so, whether it was 'put to' the Tribunal.

iii) Whether the Tribunal failed to 'deal with' it.

If the answer to all these specific issues is 'yes', a further issue arises: whether the failure has caused or will cause PIC substantial injustice."

An 'issue'

First of all, the judge noted that the authorities draw a distinction between 'issues' and 'arguments', 'points' or 'lines of reasoning'. The former satisfy the test, while the latter do not and this distinction reflects a concern in the authorities, "to maintain the “high threshold” that has been said to be required for establishing a serious irregularity" He went on to say that trying to identify a universal definition of an 'issue' was an "impossible task" and rejected three possible "yardsticks" proposed by Dow, instead, preferring to adopt a more holistic approach. He accepted that the 'assumption of responsibility' question was an 'issue' within the meaning of the sub-section, noting that it was "not simply a way of presenting the question of foreseeability, and not simply an argument in support of a contention that losses were not within the First Limb or the Second Limb of Hadley v Baxendale." He also took into account that:

  • The whole of Dow’s claim could have depended on how the 'assumption of responsibility' question was resolved;
  • Treating the 'assumption of responsibility' question as an 'issue' best accorded with what he considered to be, "the ordinary and natural meaning of the word"; and
  • Assuming the question was put to the tribunal, "fairness demanded that [it]… be 'dealt with' and not ignored or overlooked"

Being 'put'

Based on the content of PIC's written memorials, filed in the arbitration – the second of which contained a section titled "PIC did not assume responsibility for Dow's losses," the judge concluded that the 'issue' had indeed been 'put' to the tribunal. He rejected the argument that PIC had withdrawn the issue simply because, in oral argument, the parties' counsel had at times referred to it as part of the wider question of 'foreseeability' and whether the losses fell within either limb of Hadley v Baxendale.

'Dealing with' it

Finally, the judge confirmed that in order to determine whether an issue has been 'dealt with', the court must look to the award. In so doing:

  • The court should not necessarily look for the tribunal to have, "set out each step by which they reach their conclusion or deal with each point made by a party…"3
  • A tribunal has not failed to deal with an issue which it decides without giving reasons and nor is it, "required to deal with each issue seriatim: it can sometimes deal with a number of issues in a composite disposal of them."
  • The approach of the court is to read the award in a "reasonable and commercial way expecting… that there will be no substantial fault that can be found with it." This may involve taking account of the parties’ submissions during the arbitration and awards "are not to be interpreted in a vacuum."4

Within this framework, he accepted PIC's submission that certain evidence relevant to the 'assumption of responsibility' question was "nowhere referred to in the award…" Nevertheless, he did not agree that the tribunal had failed to deal with that issue. Rather, he considered that it was dealt with "admittedly succinctly, in the first sentence of paragraph 146 of the award," which simply stated.

The Judge summarised PIC's case as really amounting to two complaints – either that the tribunal had conflated the 'foreseeability' question and the 'assumption of responsibility' question, or that it had not explained why it had not given much weight to PIC's evidence on this point. But, on the principles stated earlier in his judgment, neither could it properly be considered to, "…amount to a complaint of a 'serious irregularity' within section 68."


Having concluded that paragraph 146 of the award, was sufficient to show that the tribunal had 'dealt with' the 'assumption of responsibility question', the Judge did not need to decide whether PIC had been caused substantial injustice. Nevertheless, he observed that, "if the Tribunal… had entirely ignored the assumption of responsibility question, PIC would have been caused substantial injustice" By doing so, he seems to have made clear that the appeal only failed this one element of the s.68(2)(d) test.

This raises an intriguing question about how close PIC in fact came to succeeding. At the very least, it is surprising that an issue upon which, "almost the whole of Dow’s claim could have depended," should only receive only a single line's attention in the award – and even then not an explicit mention.

It is perhaps understandable, therefore, that the judge sought to emphasise that he could and would have reached the same conclusion on a different basis. He said that if, "the first sentence of paragraph 146 of the award is directed to the [assumption of responsibility] question but… fell short of 'dealing with'…, then I would conclude that no substantial injustice was or will be caused: the first sentence would still have 'cast the die'." However, this reasoning might be criticised:

  • First, it seems to conflate two distinct matters without any explanation – namely: (i) whether, qualitatively, the award indicates that the 'issue' was addressed at all; and (ii) what quantitative standard of attention is necessary for an issue to be said to have been properly 'dealt with'; and
  • Second, the judge seems to have thought that the tribunal had already made up its mind, so that even if the 'assumption of responsibility question' was only 'addressed' but not 'dealt with', PIC would not have suffered any substantial injustice. This is directly at odds with a sentence appearing just a few lines earlier in the judgment ("…PIC would have been caused substantial injustice…"). Further, the argument is somewhat circular, since it requires an assumption that the tribunal would have reached the same conclusion in any event, in order to justify the conclusion that the 'issue' which was not 'dealt with' would not have made any difference.

We might speculate that the judge could have been influenced in his decision by his opinion on the substantive question (i.e., whether 'assumption of responsibility' had a role to play in the recoverability of consequential losses this case). Similarly, where the question is one which appears to have been so fundamental to the outcome of the case, it is difficult to imagine that a tribunal of this calibre could have overlooked it completely.

In view of the sums at stake, it will be interesting to see whether PIC now takes the case to the Court of Appeal.