Great Western Trains Company Ltd v Network Rail Infrastructure Ltd

This case concerns what is essentially a short point of construction, embedded in the complex arrangements governing the operation of the rail industry. By an Interim Award dated 24^th April, 2009, Mr. Rhodri Davies QC (“the award” and “the arbitrator”, as appropriate) determined a preliminary issue of construction concerning Part G of the “Network Code” (of which more in due course), adversely to the Appellant (“FGW”) and in favour of the Respondent (“Network Rail”). From that decision, FGW appeals to this Court.

Much of the background is common ground and can be summarised relatively shortly.

The Industry structure: Following privatisation, the rail industry operates on the basis of a division between the infrastructure (“the Network”) and the operation of passenger trains.

Network Rail (formerly Railtrack plc) is and has at all material times been the owner and operator of the Network, under licence from the Office of Rail Regulation (the “ORR”). The Network is owned, managed, repaired and maintained by Network Rail. Network Rail grants access to the Network for passenger services according to Track Access Agreements with train operating companies (“TOCs”).

For their part, TOCs enter into contracts:

With the relevant franchising authority (“the Authority”) to obtain the right to operate specified services on the basis of a Franchise Agreement. TOCs submit bids to the Authority for the right to a Franchise Agreement.

With Network Rail, to govern the terms of access to the Network, on the basis of a Track Access Agreement regulated by the ORR.

FGW was, between 18^th December, 1994 and 31^st March, 2006, a TOC operating train services (“the Services”) between London and the West of England. FGW provided the Services pursuant to a Franchise Agreement dated 19^th December, 1995 (as amended from time to time, “the Franchise Agreement”), between it and the Authority. FGW gained access to the Network pursuant to Track Access Agreements dating from the 18^th December, 1994 (“the Track Access Agreement”) from time to time in force with Network Rail.

The Track Access Agreement: The Track Access Agreement is a regulated contract, subject to approval by the industry regulator, presently the ORR. Pursuant to the Track Access Agreement, Network Rail granted FGW access to the Network and, in return, FGW agreed to pay track access charges to Network Rail. The level of track access charges is set by the ORR. Such track access charges comprise Network Rail’s main funding stream, and are required to finance the Network.

As noted in the award, Network Rail’s funding requirement is subject to regular regulatory review, by way of a “charges review” conducted by the ORR. Under a charges review, charges may be increased or reduced, depending on the regulatory assessment of Network Rail’s funding requirements. As the arbitrator put it:

For present purposes, two charges reviews are relevant. Track access charges payable by FGW to Network Rail were altered, following reviews, first, with effect from 1^st April, 2001 and, secondly, with effect from 1^st April, 2004. The periods prior to the implementation of charges reviews are known as Control Periods (“CPs”); CP1 ran from 1^st April, 1996 to 31^st March, 2001; CP2 ran from 1^st April, 2001 to 31^st March, 2004 and CP3 ran from 1^st April, 2004. As will be seen in due course, the transition from the CP1 to the CP2/3 charges regimes formed an important part of the argument before me, as it did before the arbitrator.

Pausing here, so far as concerns the CP3 Notice itself, though it was debated before the arbitrator, the parties are agreed that its effect need not be considered as part of this appeal. This agreement was recorded in a helpful document (“the CP3 document”) under the names of both leading counsel. The reasons for this agreement are contained in the CP3 document, should it ever be necessary to refer to them; it is not, however, necessary to set out those reasons here.

By way of “logical corollaries” (as the arbitrator expressed it) of FGW’s agreement to pay charges to Network Rail for access to the Network, the Track Access Agreement includes provision for Network Rail to compensate FGW in the event of poor performance of the Network. It is also fair to say that the Track Access Agreement provides for FGW to make payments to Network Rail in the event of “increased” performance. The Track Access Agreement additionally provides a mechanism to allow Network Rail to make changes to the Network, with provision to compensate FGW for loss caused by the making of such changes. These provisions are summarised in the paragraphs which follow.

Schedule 4 of the Track Access Agreement deals with restricted access to the Network caused by planned engineering work carried out by Network Rail. In that event, Schedule 4 provides, in effect and by way of formulae, for the payment of liquidated damages to FGW.

Schedule 8 of the Track Access Agreement contains a performance regime, under which additional payments may be made to Network Rail for performance above a benchmark and either party may be required to make rebates or payments for performance below the benchmark. By contrast with Schedule 4, Schedule 8 deals with unplanned performance failures (i.e., delays or cancellations of scheduled services) on the part of Network Rail. Insofar as either Network Rail’s or FGW’s performance is below the benchmark, formulae in Schedule 8 provide (in effect) for the payment of liquidated damages. In the hearing before me, FGW referred to the liquidated damages provisions contained in Schedules 4 and 8 as providing “off the peg” compensation.

Network Change is a broad concept, contained in Part G of the “Network Code”, itself incorporated into the Track Access Agreement by cl. 5.1 thereof. As indicated in an Explanatory Note to Part G of the Network Code, “Network Change” is widely defined, to include any changes which are likely to have a material effect on the operation of the Network or of trains operated on the Network. Network Changes can be either physical (e.g., changes to the condition or layout of the track) or operational (e.g., the introduction of a speed restriction on a section of the track) but operational changes are only Network Changes if they last, or are likely to last, for more than six months.

In very general terms, a Network Change is likely to give rise to compensation under Schedules 4/8 – but the distinctive feature of a Network Change is that it extends beyond the scope of those Schedules to matters outside the normal range of performance. Network Changes are not confined to those having a negative impact; a Network Change may also be beneficial in nature (e.g., enlargement of capacity on a stretch of track). Part G of the Network Code contemplates that Network Changes should be preceded by a notification procedure; pursuant to this procedure, Network Rail will propose the change and the TOC will respond – a matter to which some further (if brief) reference must be made below. It was not, however, in dispute before me that a deterioration of the Network, which had not been the subject of this notification procedure, could constitute a Network Change and trigger the right to compensation, dealt with next.

Conditions G2.2 and G2.3 of the Network Code address the question of compensation to be paid by Network Rail to the TOC for the consequences of the implementation of the Network Change. Unlike Schedules 4 and 8, Part G does not provide for liquidated damages in accordance with formulae; at the hearing before me, FGW referred to Part G as providing for “bespoke” compensation. As will be seen in a moment, Part G compensation focuses on the net rather than the gross position of the TOC as a consequence of the implementation of the proposed change. Conditions G2.2 and 2.3 provide as follows:

The Franchise Agreement: This is the contract between FGW and the Authority, pursuant to which FGW was awarded the relevant franchise by the Authority. The Franchise Agreement deals with payments to be made by FGW to the Authority, service details and performance levels. It also assured FGW of the necessary Track Access Agreement with Network Rail, by way of the regulatory structure under which the ORR could require Network Rail to enter into a Track Access Agreement in prescribed terms.

FGW (as with any aspiring TOC) bids for a franchise on the basis of the charges then payable under the Track Access Agreement; self evidently, that is the only basis on which a TOC can plan. However, as already noted, later charges reviews can alter the amounts payable by the TOC to Network Rail. So, here, the period of the Franchise Agreement spanned CPs 1, 2 and 3 and the charges reviews which resulted in the alteration of the charges payable by FGW to Network Rail with effect from the 1^st April, 2001 and the 1^st April, 2004. Such later charges reviews can affect the profits of the TOC and the economics of the franchise. But the charges reviews, though designed to ensure that Network Rail receives sufficient income to run the Network to the appropriate standard, are neither intended to penalise the TOC (if charges rise above the level prevailing when the franchise was entered into) nor to confer a windfall on the TOC (if charges are reduced below that level). As between the Authority and the TOC, cl. 18.1 of the Franchise Agreement is intended to “insulate” (the word used in the award and before me) FGW from such effects by the making of corresponding adjustments to the franchise payments passing between FGW and the Authority. Accordingly, although the effect of the CP2/3 charges reviews was to increase the level of payments payable under Schedules 4 and 8 of the Track Access Agreement between FGW and Network Rail, FGW’s position as a CP1 TOC was to be maintained through the insulation furnished by cl. 18.1 of the Franchise Agreement; so far as FGW is concerned, it was always entitled to be in the position of a CP1 TOC.

Insofar as material, cl. 18.1 of the Franchise Agreement provided as follows:

The overall dispute between FGW and Network Rail was succinctly summarised by the arbitrator as follows:

The parties have, however, recognised that if FGW has a valid claim against Network Rail under Part G of the Network Code, there is in any event a stark dispute as to how it should be calculated. It is that dispute which has given rise to the Preliminary Issue, formulated in the following (agreed) terms:

In the event, sub-issue (1) proved not to be in dispute. It became and is common ground that payments made or liable to be made under Schedules 4 and 8 of the Track Access Agreement fall to be taken into account when determining compensation under Part G. That common ground was reflected in the arbitrator’s formal answer to sub-issue (1), set out in para. 126 of the award and need not be repeated here.

However, sub-issue (2) was and remains very much in dispute. As the arbitrator put it:

In the event, the arbitrator concluded that cl. 18.1 payments should be excluded when calculating compensation under Part G. Accordingly, the arbitrator’s formal answer to sub-issue (2) of the Preliminary Issue was as follows (at para. 126 of the award):

In essence, it is this conclusion which has given rise to the point of law forming the subject-matter of the appeal: namely, FGW’s contention that in construing Part G of the Network Code, the arbitrator erred in law in leaving cl. 18.1 payments, or the loss of such payments, out of account.

The most convenient introduction to the arbitrator’s reasoning is by way of a brief outline of the debate before him, including, equally importantly, the “illustrative” figures used by the parties. Although these figures may well bear some relationship with reality, given that the arbitration and this appeal have been concerned with a Preliminary Issue, the figures were correctly treated by the arbitrator as illustrative – they assume that there has been a Network Change, a matter which remains in dispute.

Turning at once to those figures, they begin with the supposition that there was no Network Change:

Under the CP1 charging regime, Network Rail would have paid £11 million (“m”) to FGW under Schedules 4/8 (of the Track Access Agreement);

Under the CP2/3 charging regime, FGW would have paid £27m to Network Rail under Schedules 4/8;

Therefore, the Authority would have paid £38m to FGW under cl. 18.1 (of the Franchise Agreement) by way of adjustment to franchise payments to compensate FGW for the difference between its position under CP1 and CP2, i.e., + £11m (under i) above) compared with -£27m (under ii) above).

The next supposition proceeds on the basis that there was Network Change:

Under the CP1 charging regime, Network Rail would have paid £68m to FGW under Schedules 4/8;

Under the CP2/3 charging regime, Network Rail would have paid £86m to FGW under Schedules 4/8;

Therefore, ignoring at this stage any claim by FGW under Part G (of the Track Access Agreement), FGW would have paid £18m to the Authority under cl. 18.1, so as not to benefit from receiving £86m (under ii) above) rather than £68m (under i) above) from Network Rail;

The total costs, losses and expenses of a trading nature incurred by FGW in consequence of the Network Change are £160m (before any contribution from Network Rail under Schedules 4/8 or Part G and excluding any payments under cl. 18.1).

In a nutshell, the nature of the argument before the arbitrator involved, as he put it, not so much a dispute over the application of Part G or cl. 18.1 when viewed in isolation but:

FGW’s approach was (and is) to treat the difference in cl. 18.1 payments between the scenarios (i.e., with and without Network Change, or as they have been referred to, “actual” and “hypothetical”) as a “cost, loss or expense” recoverable from Network Rail under Part G. FGW’s order of calculation was and is as follows:

First, take into account Schedule 4/8 payments between Network Rail and FGW;

Secondly, take into account cl.18.1 payments between FGW and the Authority (without taking into account Part G compensation payable by Network Rail to FGW);

Thirdly, calculate the Part G compensation payable by Network Rail (taking account of both incremental Schedule 4/8 payments and incremental cl. 18.1 payments).

On FGW’s primary case, the Part G compensation thus due amounted to £103m. On FGW’s alternative case (the difference, as explained in more detail below, focussing on the treatment of the £18m payable by FGW to the Authority on the assumption that there was Network Change), the Part G compensation amounted to £85m.

By contrast, Network Rail’s approach was and is to treat the cl.18.1 adjustments as irrelevant to the calculation of Part G compensation – effectively, res inter alios acta - and recoverable from the Authority rather than from Network Rail. Network Rail’s order of calculation was and is as follows:

First, take into account the Schedule 4/8 payments between Network Rail and FGW;

Secondly, calculate the Part G compensation payable by Network Rail to FGW (taking account of incremental Schedule 4/8 payments as a result of Network Change);

Thirdly, consider the cl.18.1 position as between FGW and the Authority (taking account of the incremental Schedule 4/8 payments and the Part G compensation already calculated).

On Network Rail’s approach, the Part G compensation due to FGW amounted to £47m. The difference between its case and FGW’s primary case (£103m - £47m = £56m), turns on the treatment of the £56m (£38m + £18m, to which separate considerations apply) cl. 18.1 adjustments. The difference between the Network Rail case and FGW’s alternative case (£85m - £47m = £38m), turns on the treatment of the £38m cl. 18.1 adjustment.

In the award, the arbitrator helpfully set out in tabular form the illustrative figures and calculations involved in (1) FGW’s primary case; (2) Network Rail’s case; and (3) FGW’s alternative case. These tables are reproduced in the Annexe to this judgment (“the Annexe”). I record that at the hearing before me, the parties provided further tables and a diagram; these were likewise helpful but it is unnecessary to reproduce them here; the arbitrator’s tables will suffice.

A number of noteworthy features of the dispute can be discerned and are common ground:

The Track Access Agreement and the Franchise Agreement, though involving different parties, are to be read together so as to provide a coherent contractual structure for the running of the railways.

The particular complexity of this dispute arises because of the concurrent operation of three factors: (1) A charges review resulting in the CP2/3 regime; (2) a CP1 TOC, so engaging cl.18.1 once faced with the CP2/3 regime; (3) a Network Change.

There is no dispute that FGW is in principle entitled to be made whole, in respect of its loss; there ought to be no “compensatory black hole” (subject only to any dealings between FGW and the Authority); the question is instead whether Network Rail or the Authority is liable to pay to FGW the sums in dispute.

The Authority is not a party to these proceedings; no judgment can be made against it and it is not bound by any judgment in these proceedings. However, on both parties’ cases, I should only find against FGW and in favour of Network Rail, if and to the extent persuaded (on the material and the parties before the Court) that FGW should be entitled to succeed against the Authority.

I turn without further delay to the reasoning of the arbitrator. Rightly, he treated both Part G of the Track Access Agreement and cl. 18.1 of the Franchise Agreement as forming part of a “single contractual structure”. On that footing, he approached them on the basis that they “may be expected to interact in a rational manner which makes sense in the overall context”. As to the overall context, he remarked on the “conjunction of three factors” – the charges review, that FGW was a CP1 TOC and a Network Change. All three of those factors needed to be present to generate the problem.

The arbitrator alluded to the fact that, on FGW’s case, Network Rail’s liability under Part G differed depending on whether it was dealing with a CP1 or CP2/3 franchisee, whose position was otherwise identical. He also noted that the compensation provisions in Part G applied both to “virtuous” Network Changes and those which were the result of Network Rail’s dereliction of duty. Pulling these threads together the arbitrator said this:

These considerations led the arbitrator to conclude that Part G and cl. 18.1 interacted in a more rational manner on Network Rail’s case than on FGW’s case. On FGW’s case, the arbitrator viewed the interaction between Network Changes and new and old franchisees as “haphazard and irrational”. Further, on FGW’s case, Network Rail’s costs of implementing a “benign” Network Change would be affected depending on whether it was dealing with a CP1 or CP2/3 franchisee. Accordingly, if not itself decisive, the rationality of the overall interaction between cl. 18.1 and Part G, favoured the case of Network Rail.

Turning to the detailed operation of the provisions, the arbitrator took the view that FGW’s case assumed what it needed to prove: namely, that the effects of cl. 18.1 ought to play a part in the calculation of Part G compensation. FGW further submitted that Network Rail could not put forward calculations on the basis that it had delivered a good Network performance when it had not done so; in particular, Network Rail could not bring into account the £38m that the Authority would have paid to FGW under cl. 18.1 if there had been no Network Change because (on the assumed facts) the deterioration of the Network giving rise to the Network Change meant that the £38m never became payable. Network Rail’s response was that it referred to the hypothetical position (i.e., no Network Change) in order to make the necessary calculation under Part G (a comparison between FGW’s position following a Network Change with its position without a Network Change). The arbitrator preferred Network Rail’s argument, observing that Network Rail was seeking to apply Part G and:

In an important passage, the arbitrator set out the strongest of the points made by Network Rail, as he saw them, on the operation of Part G and cl. 18.1:

For all these reasons, the arbitrator concluded that he preferred Network Rail’s case on the detailed operation of Part G and cl. 18.1.

The arbitrator observed that he had not extracted a great deal of assistance from the wording of Part G and cl. 18.1; the parties’ cases here tended to be circular. In this regard, the only point which struck the arbitrator as having some force was the following:

FGW’s alternative case took out of account the £18m already referred to. The arbitrator rejected this case as well. Although it served to “disarm” certain of his concerns as to FGW’s primary case, it remained open to the objection that:

In large measure, the argument on the appeal appeared to traverse the same ground as was covered before the arbitrator. I can therefore summarise the rival cases on the appeal considerably more briefly than would otherwise have been the case.

For FGW, Mr. Wolfson QC advanced three principal grounds of appeal:

The arbitrator paid insufficient regard to the agreed fact that Part G was a compensatory provision. The arbitrator had focussed on the position of Network Rail, the paying party, rather than on the position of FGW, the party receiving compensation.

The arbitrator had wrongly considered it to be “haphazard” or “irrational” for different amounts of compensation to be paid to TOCs affected by a Network Change based on the particular impact of the Network Change on those TOCs. Part G was not a “one size fits all” provision; it provided bespoke compensation for the actual financial impact experienced by the TOC in question from a particular Network Change. The fact that a TOC happened to be an old rather than a new franchisee was just as relevant as any other characteristic relating to the financial effects of the Network Change on the TOC.

The award was internally inconsistent. Part G was intended to provide full compensation to FGW for the effects of a Network Change. Consequently, if a sum could not be recovered from another source (including the Authority under cl. 18.1), it must be recoverable from Network Rail under Part G. The arbitrator, however, held that the (illustrative) sum of £38m relating to payments which would have been made to FGW by the Authority if no Network Change had occurred was not recoverable from Network Rail under Part G; as it was also not recoverable from the Authority – because cl. 18.1 does not deal with hypothetical facts – there was a “compensatory black hole”. The arbitrator erred in concluding that the sum in question was not recoverable from Network Rail.

Developing his submissions, Mr. Wolfson took issue with the suggestion that the real question on the appeal was to be characterised as “temporal” – i.e., in which order the calculation was to be carried out. Mr. Wolfson said this:

The £38m represented the sum for which the Authority would have been liable to FGW on the “hypothetical” scenario – i.e., no Network Change. It was the funding to which FGW was entitled to insulate it from paying increased CP2/3 regime sums to Network Rail following the charges reviews, so restoring its position as a CP1 TOC. But, by reason of the Network Change, FGW had received a poorer Network than that to which it was entitled; there was no principled basis for FGW to claim from the Authority the £38m to which it would have been entitled in different circumstances; in other words, cl.18.1 and the Authority dealt and dealt only with the “real world”, not the hypothetical scenario. The Authority, as the “custodian of the public purse” was not obliged to fund FGW for Schedule 4/8 payments which FGW might have made if circumstances had been different but which were not now payable; there was no “contractual mechanism” for FGW to recover the £38m from the Authority. The consequences of a Network Change were for Network Rail, not the Authority; it would be odd if the dislocation of service flowing from a Network Change was to be funded by the Authority. This was not a “windfall” for the Authority; it merely meant that the circumstances in which the £38m payment would otherwise have been made had not arisen. As to cl. 18.1, Mr. Wolfson summarised his argument as follows:

Mr. Wolfson criticised Network Rail and the arbitrator for producing a halfway house between the CP1 and CP2/3 regimes. That was inconsistent. The matter could be considered either in terms of the CP1 regime alone or by way of following through with the calculation and addressing both the CP1 and CP2/3 regimes. The answer, in terms of FGW’s net position, had to be the same. That it was on the basis of FGW’s approach demonstrated the correctness of that approach.

As to FGW’s alternative case, this turned on the £18m (illustrative and assumed) payment made by FGW to the Authority under cl. 18.1 of the Franchise Agreement, reflecting the difference between the £68m payable by Network Rail to FGW under the CP1 charging regime and the £86m payable by Network Rail to FGW under the CP2 charging regime, in both instances on the basis that there had been a Network Change. On FGW’s primary case, it is assumed that this £18m payment is irrecoverable from the Authority. However, FGW’s alternative case acknowledges the argument that the £18m payment is recoverable from the Authority. The reasoning is as follows. The £18m payment was made before a Network Change had been declared; had the Network Change been declared sooner, this payment would not have needed to be made. Accordingly, the £18m payment is excluded under FGW’s alternative case, because:

There would have been no need to make it, in that the higher Schedule 4/8 payments together with a smaller Part G payment would have meant that there was no scope for a clause 18.1 payment; FGW had neither made a gain nor sustained a loss as a result of the application of the CP2/3 charges regime; and/or

The payment is recoverable from the Authority because it was not properly payable under cl. 18.1 in the circumstances of Network Change and should therefore be offset against Part G compensation, in accordance with Part G2.3.

The effect of Network Rail’s case was to treat cl. 18.1 of the Franchise Agreement as sui generis, in excluding cl. 18.1 payments alone when calculating FGW’s financial position under the hypothetical scenario. By contrast, Network Rail accepted that payments made by FGW to the Authority, pursuant to cl. 17 of the Franchise Agreement and Schedule 7 thereof, were payments which constituted an element of FGW’s loss for the purposes of calculating Part G compensation. Accordingly, even on Network Rail’s case, there were instances where in the course of calculating Part G compensation, it was necessary to perform a calculation as between FGW and the Authority under the Franchise Agreement – thus involving switching between contracts. In Mr. Wolfson’s submission, the same considerations applied to cl. 18.1 of the Franchise Agreement. Furthermore, the fact that Schedule 7 payments were individual and not uniform as between TOCs illustrated how the same Network Change could result in different financial consequences for different TOCs.

The reason why Schedule 4/8 payments were assessed without reference to cl. 18.1 was because they involved liquidated damages in accordance with contractual formulae. By contrast, Part G involved bespoke compensation, calculated on “normal principles”. The loss of the £38m payment was a part of the Part G compensation to which FGW was entitled. It was:

For Network Rail, Mr. Choo-Choy QC acknowledged that there was no doubt that the combined application of Part G and cl. 18.1 “in whatever order” will produce the same result. FGW would be “transported” from the actual state it was in, having (as it were) suffered both a charges review and a Network Change, to “CP1 land”. Accordingly, from FGW’s perspective, it did not matter in what order Part G and cl. 18.1 were applied. But the order in which those provisions were applied did matter to Network Rail and the Authority, because the order of their application was relevant to the proportions of what each had to pay to transport FGW from its actual state back to CP1 land without Network Change. In summary:

Whatever FGW’s disavowal of the relevance of order of application, FGW had itself assumed that cl. 18.1 was to be applied after Schedules 4/8 but before Part G. That was no more than an assumption; merely pointing to the compensatory nature of Part G did not assist as to the order of application intended by the parties in a coherent contractual structure.

There was no fundamental difference between the compensation regimes in Schedules 4/8 on the one hand and Part G on the other. Indeed, when a Network Change occurred, it usually, if not invariably, gave rise to a claim for compensation under Schedules 4/8. Recoveries under those Schedules were, as was common ground, to be credited against the Part G claim, so serving to reduce it. Network Rail’s case involved the successive calculation of all the elements of compensation recoverable under Schedules 4/8 and Part G, thus the calculation of compensation due under the Track Access Agreement as a whole – before the calculation under cl. 18.1 of the Franchise Agreement was undertaken. Network Rail’s case therefore avoided the need to jump between agreements, a feature of FGW’s case.

Cl. 18.1 was essentially collateral, extraneous or res inter alios acta, so far as Part G compensation was concerned. As to the FGW submission, that the “lost” cl. 18.1 payment of £38m was an item of loss when applying Part G, Mr. Choo-Choy countered: (1) there was no such loss; the Authority remained liable for that sum; (2) it was only a “loss” if the assumption was made that cl. 18.1 was to be applied before Part G; if Part G was applied first, then the “lost” cl. 18.1 payment would not feature as a loss under Part G. Furthermore, once the consequences of the Network Change had been calculated (under Schedules 4/8 and Part G) what remained were consequences of the charges review which were and remained the liability of the Authority. If FGW was “exposed” vis-à-vis the Authority that was not because of a flaw in Network Rail’s case but because of FGW’s dealings with the Authority – to which Network Rail (and, for that matter, the Court) was not privy.

As to Part G, its function was not to restore FGW from the CP2 to the CP1 position. Part G’s function was only to compensate FGW for the adverse effects of the Network Change, assuming that there had been a charges review (i.e., so comparing FGW’s position under the CP2/3 regime with and without a Network Change). The upshot was a net loss of £47m, comprised of the difference between: (1) FGW’s trading costs, losses and expenses incurred in consequence of the Network Change, amounting to £160m; less (2) the “improvement” in FGW’s Schedule 4/8 position flowing from the Network Change of £113m (FGW was no longer required to pay £27m to Network Rail and instead received £86m from Network Rail).

The Authority’s position was and ought to be unchanged. The Authority was neither concerned with nor responsible for the consequences of Network Change; its position remained unchanged – a liability to pay £38m to FGW (the difference between FGW’s entitlement to receive £11m from Network Rail under CP1 absent Network Change and FGW’s obligation to pay £27m to Network Rail under the CP2/3 regime, again absent Network Change).

As to the FGW objection that in the “real world” (i.e., with Network Change), the Authority was not liable to make the £38m payment, Mr. Choo-Choy’s response was that the objection assumed an intended “correlation” between what was payable under cl.18.1 and the level of performance. But there was no such correlation. Cl. 18.1 was not a performance regime. The calculation of cl. 18.1 payments was based on a comparison of the effect of a charges review on payments due between Network Rail and FGW – not on the quality of performance delivered on the Network. Network Rail did not quibble with FGW’s proposition that cl. 18.1 was concerned with the actual scenario (i.e., with Network Change); however:

Focussing on the terms and nature of cl. 18.1, Mr. Choo-Choy said this in his skeleton argument:

Mr. Choo-Choy was dismissive of FGW’s alternative case. If FGW was right that cl.18.1 was to be applied before Part G, the £18m payment was irrecoverable. But if Network Rail was right that cl. 18.1 only arose for consideration after the Part G calculation had been undertaken, then not only would FGW not be liable to pay the £18m but it would be entitled to receive the £38m. There was, accordingly, no room for this halfway house. The choice lay between FGW’s primary case and Network Rail’s case.

As to the principal criticisms of the arbitrator’s reasoning advanced by FGW, Mr. Choo-Choy submitted that, as demonstrated by the award, the arbitrator was well aware of the compensatory nature of Part G. The arbitrator was entitled to take into account the fact that, on FGW’s approach, the amount of compensation payable by Network Rail would vary depending on the charges regime applicable to the TOC – even though the extent of the costs, losses and expenses of a trading nature incurred by that train operator as a result of a Network Change would not be affected by whether it was an “old” or a “new” franchisee. FGW’s reliance on what Mr. Choo-Choy termed the “egg-shell skull” principle – i.e., Network Rail taking its TOCs as it found them – begged the question of the order of application of Part G and cl. 18.1. That, as the arbitrator held, was irrational where the difference turned not on operating or trading losses incurred as a result of a Network Change but instead on the question of whether the TOC in question was an old or new franchisee. As to cl. 17 and Schedule 7 of the Franchise Agreement, this was simply another head of trading loss and was therefore rightly taken into account in calculating compensation under Part G. Finally, there was no internal inconsistency in the award; there was no compensatory black hole; on the arbitrator’s reasoning and Network Rail’s case, FGW was entitled to recover the £38m sum from the Authority.

As will be apparent, I was most grateful to both Mr. Wolfson and Mr. Choo-Choy for the sustained quality of their written and oral submissions.

(1) Introduction: As was not in dispute, the question for me was whether the arbitrator was right or wrong in his conclusion on sub-issue (2) of the Preliminary Issue. This is not a case where there is a range of permissible outcomes; in that sense, there is no discretionary element involved.

As was further not in dispute, the question is not whether FGW should be made whole for the consequences flowing from the (assumed) Network Change and the charges reviews but by whom – i.e., whether the entirety of FGW’s loss should be borne by Network Rail or whether the £38m and £18m in dispute are properly (in principle) the liability of the Authority.

During the course of the hearing, I expressed misgiving at the prospect of expressing a view on the Authority’s liability in the absence of the Authority. I continue to regard that situation as unsatisfactory but both parties were united in saying that I should do so – and, indeed, given the course taken by the proceedings to date, I have in effect no option. I cannot refuse to entertain the appeal and nothing can be done at this stage to join the Authority into the proceedings. So far as I am aware, no steps were taken to join the Authority into the arbitration or to commence other proceedings to which the Authority was a party or to reach an agreement binding the Authority to the outcome. Nonetheless, the premise of both FGW’s and Network Rail’s cases before me was that if the disputed sums or any part thereof is not recoverable from Network Rail, then, in principle (i.e., subject only to any dealings between FGW and the Authority) they would be recoverable from the Authority. In the event, therefore, while this Judgment cannot be a Judgment against the Authority (should I take the view that liability rests with the Authority) and does not bind the Authority, as between FGW and Network Rail my conclusions impact on the presumed liability of the Authority. Troubling though this is, I console myself with the reflection that it must be unlikely that the Authority would have developed arguments not covered in Mr. Wolfson’s and Mr. Choo-Choy’s submissions.

Before proceeding further, I should allude to another matter which gave me pause for thought. Both before the arbitrator and on the appeal, the parties have relied on illustrative calculations in support of their rival contentions. This is permissible, provided that the illustrative calculations are kept within their proper confines. As it seems to me, they can be used to illustrate the consequences of rival constructions of the Track Access and Franchise Agreements. It is necessary, however, to keep in mind that subsequent conduct of the parties cannot be used as an aid to construction. At all events, I am content that the parties have not infringed that principle.

Reverting to the appeal, for my part, the nature of the inquiry was correctly introduced by Mr. Wolfson: the real question goes not to the order in which the calculations are performed but whether any amounts excluded – especially the £38m - from the Part G calculation are (in principle) recoverable from the Authority. As I ventured to suggest in argument, given an agreed set of figures, it did not seem to me that the order in which the calculations were performed could determine the outcome – albeit that rival orders were obviously useful for the making of forensic points. But the nature of the inquiry does not end even with Mr. Wolfson’s characterisation. Underlying his formulation, there is the need to go on and determine for what the sums in dispute, especially the £38m, truly compensate FGW for, or insulate FGW from – for the Network Change (in which case they should be Network Rail’s liability) or from the charges review/s (in which case they should be the liability of the Authority). I put the matter this way because it was incontrovertible and not (realistically) in dispute, that Part G is intended to compensate FGW for the consequences of a Network Change, whereas cl. 18.1 is intended to insulate FGW from the financial consequences of a charges review. To my mind, the principal difficulty in this case lies in determining on which side of this line the £38m and £18m fall.

In fairness to Mr. Choo-Choy, I should indicate that insofar as the order of performing the calculation matters or assists, I do broadly prefer the Network Rail approach – of proceeding by way of Schedules 4/8, then Part G and then cl. 18.1. Although Schedules 4/8 provide for compensation by way of formulae whereas Part G leaves compensation to be determined (as Mr. Wolfson put it) in accordance with “normal principles”, they are closely related provisions; in performing the calculation, it seems natural to move from one to the next, before turning to consider the Franchise Agreement. Further and in general (and not overlooking other occasions when cl.17 and Schedule 7 of the Franchise Agreement will be relevant), that approach reduces the need to jump between the Track Access and Franchise Agreements. As already indicated, however, I do not think that the order of performing the calculations is or can be determinative of the true nature of the sums in dispute and hence the question of where liability in that regard should rest. By way of illustration, the adoption of Mr. Choo-Choy’s proposed order of calculation leaves open the question of where Part G compensation ends – i.e., what is to be included in or excluded from compensation under that head.

(2) The contractual wording: As and to the extent that the point is one of construction, I start with the contractual wording and seek to highlight a number of features of Part G and cl. 18.1.

As to Part G:

Part G is focussed on Network Change. It provides compensation for “costs, direct losses and expenses (including loss of revenue)” (to which, for brevity, I shall some times refer as “loss”) consequential upon a Network Change. As G2.3 makes plain, this will be a net figure. To attract compensation under Part G, the loss must flow from the Network Change.

Whether implicitly or expressly, the necessary comparison is between the TOC’s position without and with Network Change – in other words, between the so-called hypothetical and actual scenarios.

To my mind, as a matter of language, cl. 18.1 payments are capable of coming within the “costs, direct losses and expenses” for which Part G gives compensation. Indeed, it is to be noted that the meaning of “costs, direct losses and expenses (including loss of revenue)” resulting from a Network Change is left at large. Part G could have been but was not defined so as to exclude cl. 18.1 payments from its scope.

Additionally, it is not at all apparent that the wording of Part G points to the exclusion from its scope of the loss of a cl.18.1 payment that would have been payable under the hypothetical scenario but is not payable under the actual scenario.

Turning to cl. 18.1:

The focus is on the “direct result” of charges reviews (sub-cll. (e) and (f)).

As to its nature, cl.18.1 is an “insulating” provision, a matter underlined by the wording: “the Franchise Operator suffers no net financial loss and makes no net financial gain” as a “direct result” of the increase or reduction in charges contained in a charges review.

The relevant comparison is set out in sub-cl. (i) and is the “difference between the relevant charges that would have been payable in the absence of the Charge Variation and those that are payable following the implementation of the Charge Variation” – i.e., for present purposes, the difference between CP1 and CP2/3 charges.

Plainly, as this comparison is between two actual charges regimes, cl. 18.1 only operates in the “real” or “actual” world. If, however, Part G compensation is to be considered as an element of this comparison, then cl. 18.1’s focus on the actual world does not preclude the need to compare the hypothetical and actual scenarios, for the purpose, if only for the purpose, of undertaking the relevant Part G calculation (i.e., a comparison between the hypothetical and actual scenarios).

As it seems to me, cl. 18.1 does not provide a “performance regime” - it compares sums payable under two charges regimes rather than hinging on Network performance. However, if and insofar as Part G compensation forms an element of the cl. 18.1 comparison, then to such extent Network performance is not irrelevant, given its centrality in determining Part G compensation.

With these reflections in mind, I agree with the arbitrator thus far. First, that the contractual wording is not itself decisive; both parties’ contentions are possible constructions, given the contractual language. Secondly, there is a good deal in the parties’ cases which tends towards the circular. Thirdly, as to the apparent oddity if cl. 18.1, an insulating provision, was itself capable of giving rise to a cost, direct loss or expense recoverable by a TOC under G2.2 – a matter lending support to the Network Rail case.

That said, there seem to me to be additional matters arising from the contractual language which likewise require emphasis. First, that the loss under Part G is at large and not, as a matter of language, the subject of any restrictive definition. Secondly, that, on the face of it, there is nothing in the language to exclude the loss of a cl. 18.1 payment from a Part G claim, subject always to such loss being a consequence of a Network Change. Thirdly, that cl. 18.1 only operates in the real world. All these matters may be thought to lend support to the FGW case.

(3) The contractual scheme: As is common ground, the Track Access Agreement and the Franchise Agreement, though involving different parties, are to be read together so as to provide a coherent contractual structure for the running of the railways. For present purposes, a most important corollary is that Part G and cl. 18.1 should both be given effect – but, each within its own proper sphere: Part G to provide compensation for loss flowing from Network Change and cl. 18.1 to provide insulation against the results of the charges reviews. Necessarily, the parties’ cases must be tested with such considerations in mind.

I begin with the award. The arbitrator’s conclusions, both as to the overall interaction between the Track Access Agreement and the Franchise Agreement and their detailed operation, appear to have been central to his decision. In both instances, he preferred the Network Rail case. With respect and without losing sight of the attractions in the Network Rail case, I am unable to agree. As it seems to me, there are anomalies in the cases of both parties and much that is circular in the Network Rail case or which turns on the order of calculation. My reasons follow.

First and admittedly supporting Network Rail’s case, it is at least at first blush anomalous that, on the FGW case, all other factors being equal, the compensation payable by Network Rail under Part G will (or will likely) differ according to whether the TOC is a “new” or “old” franchisee. The anomaly is all the more striking, given that it applies even in the case of a benign Network Change. I do find that troubling, because the consequences of charges reviews are intended to be for the Authority rather than Network Rail. I am unable to accept Mr. Wolfson’s answer in response (comprising one of his grounds of appeal): namely, that approaching the matter this way involved overlooking the compensatory nature of Part G. It is instead doing no more than considering the ramifications of rival constructions of Part G, as a compensatory provision, with regard to the contractual structure as a whole. It is of course true, as Mr. Wolfson submitted, that as compensation under Part G constitutes or is analogous to “unliquidated damages”, the amount payable will not necessarily be the same as between different TOCs; that difference is acceptable provided it is confined to the consequences of a Network Change and does not stray into the territory of charges reviews – as explored further, below.

In any event, matters and anomalies do not end there. The Network Rail case involves – at least on the illustrative figures in this dispute - less Part G compensation due following a Network Change under the CP2/3 regime than under the CP/1 regime. On the face of it, this is a striking saving and not at all easily attributable or explicable as a direct result of the charges reviews.

It follows that both parties’ cases are capable of generating curiosities or apparent curiosities. Perhaps that is unavoidable, given the complexity of the Agreements and the combination of factors present in a case of this nature.

Secondly, so far as it is a matter of rational contractual structure – rather than simply the language of the contracts – there is apparent force in Mr. Choo-Choy’s submission that it is irrational or circular to treat cl. 18.1 both as an insulating provision, presumably pre-supposing that a Part G calculation has already been undertaken and as a component of the Part G calculation. That said, it is at least implicitly assumed in this submission that the loss of a cl.18.1 payment is of an “insulating” nature, rather than simply another “cost, direct loss or expense” for which compensation is payable under Part G. Accordingly, this argument too is not devoid of an element of circularity.

Thirdly, as to the “strongest” Network Rail points, emphasised by Mr. Choo-Choy and singled out by the arbitrator (in para. 77 of the award):

Here too, so far as concerns items (1) and (2), there is, with respect, circularity.

The suggestion of a “windfall” to the Authority (or the Authority being £56m or £38m better off) depends on whether the remainder of the Network Rail argument is well-founded. If, for instance, the £38m exposure, properly analysed, arose as a consequence of Network Change rather than a result of the charges reviews, then all that has happened is that the benefit for Network Rail of increased payments under CP2/3 has been eliminated - on the illustrative figures in this case - by the scale of the dislocation caused by the Network Change.

So too, the arbitrator’s conclusion that the burden of protecting FGW from the effects of the charges review is (on FGW’s case) transferred from the Authority to Network Rail, depends on the sums in question truly resulting from the charges review/s rather than from Network Change.

Items (3) and (4) turn on the order of calculation – a matter with which I have already dealt.

Without more ado I come to the true nature of the £38m in dispute and the question of giving proper effect to Part G and cl. 18.1. I confess that it is this topic, going to the true nature of the principal sum in dispute, which has given me the most pause for thought. But, having anxiously considered and reconsidered the arbitrator’s and the parties’ (illustrative) tables and diagrams, I have reached a clear conclusion.

It is here, as it seems to me, that the FGW case enjoys a decisive advantage.

Although the change from the CP1 to the CP2/3 regimes, in the absence of Network Change, would have resulted in increased charges payable to Network Rail, the (assumed) Network Change entailed dislocation to the Network and hence the provision of “worse” services.

Against this background, it is right that the £38m first entered into the parties’ calculations as an insulating payment, resulting from the charges review/s - but that was in the hypothetical scenario, absent Network Change. On that hypothesis, FGW would have been liable to make a payment to Network Rail (given the increase in the charges payable) and the Authority would, in turn, have been liable to make a compensating (or insulating) payment to FGW.

The question which next arises goes to the position under the actual scenario – i.e., with Network Change. As it seems to me, the premise of the Authority’s liability to make the £38m payment under cl.18.1 has gone. Given the dislocation in service, FGW is no longer liable to make the increased payments to Network Rail. To the contrary, in the actual scenario with Network Change, Network Rail is liable to make increased payments to FGW, whether under CP1 or CP2/3. In this situation, involving the actual scenario with Network Change, it is, accordingly, not easy to see why the Authority should make an “insulating” payment to FGW: there is nothing from which to “insulate” FGW.

But it follows, on this reasoning, that the loss of the £38m payment is attributable to the Network Change and FGW’s £38m exposure is not attributable to the charges review/s – notwithstanding the fact that the £38m payment first arose (in the hypothetical scenario) as a consequence of the charges review. The Network Rail argument that the £38m is and remains a liability of the Authority appears to me to look at the comparison between the CP2/3 and CP1 regimes in isolation and without paying due regard to the obligations of the Authority in the actual scenario. As Mr. Wolfson put it, the public purse was not there to fund TOCs in respect of payments which they did not in fact make. Suffice to say, I agree.

As already discussed, there is no or no realistic dispute that the role of Part G is to compensate a TOC for Network Change and that the role of cl.18.1 payments is to insulate a TOC from the results of a charges review. On the illustrative facts of this case, in my judgment, the loss of the £38m flows from the Network Change not from the charges review/s. The Authority does not remain liable to pay the £38m for which it would have been liable by reason of the charges review, regardless of the Network Change. To the contrary, the Authority is no longer liable to make a payment of £38m because of the dislocation in service flowing from the Network Change. That is not because cl. 18.1 provides a “performance regime”; it is instead because the dislocation in service flowing from Network Change has resulted in moneys flowing from Network Rail to FGW rather than from FGW to Network Rail, so that the basis for the Authority compensating (or “insulating”) FGW, has gone. It follows that the loss of the £38m properly falls within Part G.

Put another way, as the parties before me are agreed that the £38m must be payable (in principle) by either Network Rail or the Authority, then, if there is no proper basis for imposing liability for this sum on the Authority, it must be the liability of Network Rail.

I am reinforced in this conclusion by the consideration that Mr. Wolfson’s tables point to the same Part G compensation, whether the matter is addressed under the CP1 or CP2/3 regimes. To me, that suggests that the £38m gap is properly attributable to Network Change rather than charges reviews.

A claim for Part G compensation is a claim for, or in effect for, unliquidated damages. The sums claimed necessarily turn on the circumstances and facts of the case. With respect to the arbitrator’s view to the contrary, it is neither here nor there that the sums payable by Network Rail under Part G might differ between different TOCs, provided only that, as here, the “loss” in question is consequential upon a Network Change.

For completeness, the order of calculation matters not. On Mr. Choo-Choy’s and Network Rail’s order (see the Annexe), the £38m is most simply added to the £47m Part G figure under the CP2/3 regime, so producing a nil liability under cl. 18.1. But there are no doubt a variety of ways of undertaking the calculation, all of which ought to lead to the same result.

(4) Conclusion as to the £38m: Pulling the threads together, this conclusion as to the £38m appears to me to be soundly based as a matter of contract, context and principle. It accords with the contractual wording; it involves reading the Track Access and Franchise Agreements together as a coherent whole; it does so in a manner which has the practical benefit of allocating liability for the £38m exposure to the party responsible for it - Network Rail – and, in so doing, permits effect to be given to both Part G and cl.18.1, while confining each to its proper sphere on the illustrative facts. Though the arguments were on the whole relatively finely balanced, I am, in particular, persuaded that the £38m exposure is a consequence of the Network Change rather than the charges review/s and is irrecoverable from the Authority.

With respect to the arbitrator and his demonstrably careful consideration of the dispute, it follows that the appeal must be allowed as to the £38m. In my judgment, with regard to sub-issue (2) of the Preliminary Issue, account is to be taken of the £38m payment/s “….made or liable to be made – or which, but for the alleged Network Changes, would have been made or liable to be made…” between FGW and the Authority pursuant to cl. 18.1, when determining the amount of compensation which FGW may be entitled to claim from Network Rail under Part G.

(5) The £18m: I turn next to consider the question of the (illustrative) £18m in dispute. I do so, having concluded that, for the reasons given, FGW is entitled to recover from Network Rail for the “lost” £38m, which in the actual scenario FGW would not have been entitled to recover from the Authority. I am not, however, persuaded by Mr. Wolfson that the same result should follow with regard to the £18m; nor am I dissuaded from reaching this conclusion by Mr. Choo-Choy’s forensic contention that there is no room for FGW’s alternative case. In my view, FGW’s alternative case is soundly based and is to be preferred to both FGW’s primary case and Network Rail’s case.

Put summarily, the £18m (assumed) payment by FGW to the Authority was not payable to and/or is recoverable from, the Authority, essentially for the reasons given by Mr. Wolfson - set out above, illustrated in the tables upon which FGW relied and which do not benefit from elaboration. I should underline the simple and straightforward basis of this decision; it does not entail, contrary to Mr. Choo-Choy’s submissions, any broader conclusions as to the overall order of calculation or either party’s case as to the £38m. In the result, Network Rail is not liable for the £18m (assumed) payment by FGW to the Authority. FGW must recover this £18m payment, if it is to recover it at all, from the Authority. It follows that the FGW appeal fails in respect of the £18m. Any risk/s of failure to recover from the Authority – either because of dealings between FGW and the Authority or because of a failure to join the Authority to these proceedings or to bind the Authority to their outcome - must be borne by FGW.

(6) Overall conclusion: Accordingly, the FGW appeal succeeds as to the £38m in question and fails as to the £18m. It follows that, on the illustrative figures, FGW is entitled to Part G compensation from Network Rail in the total sum of £85m (i.e., £47m which Network Rail admits plus the £38m).

I shall be grateful for the assistance of counsel in drawing up an appropriate order, including the formal answer to sub-issue (2) of the Preliminary Issue and as to all matters relating to costs.

(7) Postscript: Two points of procedure arise under this heading – both are intended constructively:

First, as to permission to appeal. I record that, with respect, I was somewhat surprised to discover that the application for permission to appeal was considered and granted by a Judge sitting in the Bristol Mercantile Court, before the case was transferred to the Commercial Court. Rightly or wrongly, I had assumed that all applications for permission to appeal from an arbitration award under s.69 of the Arbitration Act 1996 (“the Act”) were to come, at least in the first instance, to the Commercial Court, where they would be considered or transferred (out), depending on the view taken by a Judge of the Commercial Court. It appears, however, that the wording of Part 62 CPR is by no means clear in this regard – cf. PD62.2.3 and PD62.14.1. – a matter which to my mind requires further attention and clarification. In saying this, I should underline that I intend no criticism of the decision in the present case to grant permission. My interest lies instead with the appropriate practice in this area.

Secondly, as to the volume and nature of the materials before the Court. In the vast majority of cases, an appeal from an arbitration award, under s.69 of the Act, requires little, if any, documentation beyond the award itself and the parties’ skeleton arguments (together with any necessary authorities). In the present case there was justification for the introduction of certain additional background materials. But I was wholly unable to share the parties’ enthusiasm for documents such as the pleadings in the arbitration, together with documents relating to the application for permission to appeal. Save exceptionally, such documents have no place in any appeal under s.69 of the Act. I mention this matter by way of guidance, rather than by way of criticism of those involved in the present case (where, in the event, no time at the hearing of the appeal was lost as a result).