United Kingdom Upper Tribunal (Lands Chamber)

Neutral Citation Number: [2025] UKUT 138 (LC) 

Case No: LC-2023-404

IN THE UPPER TRIBUNAL (LANDS CHAMBER)

IN THE MATTER OF A NOTICE OF REFERENCE

2 May 2025                                     

 

TRIBUNALS, COURTS AND ENFORCEMENT ACT 2007

 

COMPENSATION - COMPULSORY PURCHASE - loss of profits - site of production of railway sleepers - past and future profits and/or losses in the real world and the "no scheme world" - interim decision

 

BETWEEN:

    CEMEX UK OPERATIONS LIMITED

Claimant

-and-

 

   SECRETARY OF STATE FOR TRANSPORT

Compensating Authority

 

   Land at Washwood Heath,

Aston Church Road,

Birmingham,

West Midlands,

B8 1QF

 

 

   Upper Tribunal Judge Elizabeth Cooke and Mr Peter D McCrea OBE FRICS FCIArb

27, 29, 30 January, 3, 10-14 February, 3-4 March 2025   

 

Royal Courts of Justice

             

Michael Humphries KC, Rebecca Clutten and Daisy Noble, instructed by Bryan Cave Leighton Paisner LLP, for the claimant

Richard Glover KC, Cain Ormondroyd and Brendan Brett, instructed by DLA Piper LLP, for the compensating authority

 

© CROWN COPYRIGHT 2025

The following cases were referred to in this decision:

 

 

Bishopsgate Space Management Ltd v London Underground Ltd [2004] 2 EGLR 175

 

Introduction

1.            This is a claim for compensation by Cemex UK Operations Limited following the compulsory purchase of its factory and land at Washwood Heath in Birmingham in 2020 pursuant to the High Speed Rail (London to West Midlands) Act 2017. The land was required as an operational hub for the new HS2 rail link; it is perhaps ironic that the claimant's business was, among other things, the production of concrete railway sleepers, albeit of a kind not used in the construction of HS2.

2.            The claimant was represented by Mr Michael Humphries KC, Ms Rebecca Clutten and Ms Daisy Noble, and the Secretary of State for Transport as the compensating authority by Mr Richard Glover KC, Mr Cain Ormondroyd and Mr Brendan Brett. We offer our grateful thanks to them all; their succinct and focussed oral and written submissions were of great assistance to us in this complex and lengthy claim.

3.            The claimant held land at Washwood Heath ("WWH") on a number of long leases, and conducted three businesses: a factory for the manufacture of railway sleepers, an asphalt plant, and an aggregate supply business. It was not possible to find a site to which all three businesses could move and so the claimant had to move to three different sites. It is entitled to compensation not only for the value of the land taken from it but also for relocation costs, lost profits for all three businesses, professional fees and other miscellaneous costs. Many of the elements of the compensation have now been agreed. The main outstanding issue is the loss of profits of the sleeper business. The parties' positions can be summarised as follows.

4.            The claimant says that it had a very successful business at WWH, selling a high volume of sleepers, and had a favoured position in the market because of its efficiency, flexibility, and expertise, operating from an ideal location close to the West Coast Main Line ("WCML"). It has lost its premises, was unable to produce sleepers at all for two years, and has now resumed production, using a less profitable method, at an inferior location in Rochester, Kent. The claimant seeks compensation for loss of profits of just over £59 million.

5.            The compensating authority says that the years from 2002 to 2016 saw a very high demand for sleepers, for a number of reasons including the renewal of the WCML, but after that the demand for sleepers dropped considerably and is now at an all-time low.  The claimant's only competitor is operating efficiently and is now the majority supplier. The claimant's move to a smaller and more efficient factory has protected it from the down-turn in the market, and as a result it has enjoyed a financial gain of nearly £4.5 million.

6.            The most important evidence in this reference is therefore the expert evidence called by both parties relating to the market in sleepers, Mr Paul Jarvis for the claimant and Mr Peter Heubeck for the authority. We have to assess what has happened and, as best we can, what will happen to that market, and compare that to what would have happened, to date and in the future, in the imaginary "no-scheme world" where HS2 was cancelled. We have to make factual findings on the basis of the evidence of the witnesses of fact and of the experts on sleepers, but those findings cannot dispose of the case; it will be for the forensic accountants called by each party to compute the claimant's loss (or gain) on the basis of our factual findings.

7.            The hearing was listed for three weeks from 27 January 2025, but had to be adjourned for unavoidable reasons after we had heard evidence from all the witnesses except the forensic accountants. It has therefore been agreed that we will issue this interim decision on the basis of the evidence we have heard; a hearing has been listed at which the forensic accountants will give their evidence about the claimant's loss (or gain) on the basis of the Tribunal's findings. This has the advantage of making it possible for the accountants to carry out their analysis on the facts that the Tribunal has found, rather than having to work on the basis of multiple possible scenarios. The expert reports already made by the forensic accountants will inevitably be superseded by our findings and we have given directions for them to exchange and file further reports. The issues on which we make determination in this decision are matters that fall not within the accountants' expertise but within that of the sleepers experts.

The legal background

8.            The legal basis of the claimant's entitlement to compensation is not in dispute. It is well established that compensation for compulsory purchase should reflect not just the market value of the land taken but also the value of the land to the owner; an owner is therefore entitled to compensation for being disturbed from the enjoyment of the land. That entitlement is reflected in what is known as "rule 6", in section 5 of the Land Compensation Act 1961, although it is set out there rather obliquely. But at any rate it is not in dispute.

9.            As we explained above, we have to compare the claimant's financial position in the real world, to date and in the future, with the position it would have been in in the "no scheme world", as it is often called - a shorthand for the world where the scheme was cancelled. In both worlds we have to look into the future, real and imagined, and we have to decide how far to look; there will come a time in both worlds after which we cannot now predict the claimant's financial position, because other factors, known and unknown, will have a more pronounced effect upon the claimant's financial position than does the presence or absence of the scheme.

The factual background

10.         The account that follows is not in dispute except where we indicate to the contrary. It is based upon the parties' agreed chronology, upon facts agreed by the expert witnesses, and on the evidence of two witnesses of fact: Mr Stuart Neil, who was from 2000 the claimant's Operations Manager at WWH and later its Business Manager for Rail until he retired in January 2022, and Mr Damien Allen who has worked for the claimant since 2005 and has been its Business Manager for Rail since April 2021, sharing the role with Mr Neil until January 2022.

The supply and demand for sleepers in Great Britain

11.         Railway track is formed of two rails attached to sleepers which hold them the correct distance apart. Beneath the sleepers is ballast, made of small stones. There are some 20,000 miles of railway in the UK, and about 26 million sleepers. Sleepers used to be made of timber, but most of those have now been replaced; a few are made of steel; the vast majority are made of concrete. New materials for making sleepers, including recycled plastic, are being developed with a view to minimising the carbon footprint of the network. The life-span of concrete sleepers is not infinite and so Network Rail ("NR"), which since its creation in 2002 has owned the railway track in Great Britain and is required to maintain it, requires a constant supply of sleepers, mostly for the cyclical replacement of all the sleepers in the network and a small proportion for new track. The size of that requirement has varied over time and is now at an all-time low, and the reasons for that are not agreed; nor do the experts agree how many sleepers NR will require in the future. There are other customers for sleepers, notably Transport for London Limited, but NR's purchases constitute more than 90% of the market in Great Britain. Conventional sleepers have not been used in the construction of HS2, and therefore that project has had no effect on the market in sleepers.

12.         The vast majority - about 95% - of sleepers used in Great Britain are of the G44 and the slightly longer EG47 type. The remaining 5% is made up of a number of specialist sleepers for use where sizing is non-standard (for example in tunnels and viaducts). In recent years NR has required some sleepers to be fitted with Under Sleeper Pads ("USPs"); these are rubber-based pads fixed to the underside of the sleeper during the manufacturing process which cushion the weight of the train and therefore reduce damage to the ballast. Mr Allen said that while NR would like 50% of sleepers to have USPs, they are more expensive and only about 16% of sleepers are made with them.

13.         Track is attached to the sleepers by means of a clip, and when the clip works loose the sleeper has to be replaced. The "Pandrol" clip was introduced in 1963, and according to Mr Heubeck it "solved the problem of the long-term stability of the rail secured to the sleeper"; by 1969 all sleepers were fitted with Pandrol clips.

14.         Nevertheless, how long a sleeper lasts is variable; the sleepers experts disagreed about their usual lifespan. Mr Jarvis told us that it is generally reckoned to be 30 years, while Mr Heubeck for the authority told us that they could last 60 years or more. We say more about that disagreement later, but both views demonstrate that sleepers are needed in substantial quantities for maintenance. Mr Jarvis's evidence was that NR reckoned on an average of 750,000 a year, and he recalled one year when it purchased 930,000. Neither of those figures is agreed, and NR's requirement has been much lower in recent years, reaching an (agreed) all-time low of 307,000 or so in the last twelve months.

The Washwood Heath factory

15.         The sleeper factory at WWH was purpose-built in 1989 and commenced operation in 1990. It had a maximum annual production capacity of 600,000 sleepers. According to Cemex it provided up to 60% of NR's annual requirement until 2016; the authority does not challenge that but points out that it is impossible to check on the information available. Because of the location of WWH, and the way the railway network is laid out (with more north-south lines than east-west since the Beeching cuts), the claimant has mostly supplied the west and south-west of England, including the WCML, and Wales. The eastern side of the country was served in the 1990s by a factory owned by Tarmac Building Products at Tallington on the East Coast Main Line, which ceased production in 2014. Since 2014 the other supplier in this duopoly has been a factory at Doncaster, built in 2012 and operated by Trackwork-Moll Limited ("TWM"), a joint venture between Trackwork Limited and Leonhard Moll Betonwerke AG. It has a maximum annual production capacity of approximately 400,000 sleepers and was working at full capacity by the end of 2016.

16.         Cemex held WWH on five leases from GB Railfreight Ltd, four relating to areas which the parties have called areas A-D, and the fifth being of an access roadway - area R. The leases of areas A, C, D and R were granted in the 1980s and 1990s and were within the protection of the Landlord and Tenant Act 1954.   Those leases were due to expire on 24 March 2025, with Cemex having the statutory right to renew.  The lease of Area B was granted in 2006, with a term commencing in 2000 and expiring in 2025.  It was contracted out of sections 24-28 of the Landlord and Tenant Act 1954, meaning Cemex had no statutory right to renew.  Area B was located to the southwest of the remaining land and was used primarily for the storage of sleepers.  

17.         There are two principal methods of making sleepers. WWH used, and the TWM factory uses, the long line method whereby sleepers are cast in moulds laid out in a long line, with concrete being brought to the moulds and poured in around long steel cables under tension. The alternative is the carousel method, in which sleeper moulds are moved around the factory, with each sleeper being cast in moulds which radiate from a central point, each having separate steel cables under tension. The long line method needs more space, and more capital investment; the carousel method can be accommodated in a smaller space and is cheaper to build, but it has more moving parts, equipment needs replacing more frequently, and each tension requires periodic checking. As Mr Jarvis put it, "within the industry there are differing schools of thought as to which is the optimum method."

Local Distribution Centres and the rail network

18.         The location of the Doncaster factory is important. NR manages the materials used in the maintenance of the railway by using a network of 11 Local Distribution Centres ("LDCs"), shown on the plan below, which was produced by Mr Heubeck:

19.         Some materials, in particular ballast, are stockpiled at the LDCs for use when needed; sleepers are not, and therefore sleeper factories have to have storage capacity in order to cope with times of high demand. In the early 2000s NR adopted a policy of seeking to have sleepers produced at LDCs, so as to eliminate the cost of moving them from the factory to the nearest LDC. The TWM factory is at the Doncaster LDC. The WWH factory by contrast is not at an LDC, although it is not far from the LDC at Bescot, to which sleepers have to be taken for onward movement to a work site. The round trip from WWH to Bescot and back is 38 miles. There are 11 LDCs, and the following plan produced by Mr Heubeck indicates their geographical distribution:

20.         Because WWH was not at an LDC, NR incurred an additional "trip cost" when it bought sleepers from WWH, being the cost of transport to the nearest LDC at Bescot (estimated by NR to be around £750,000 per annum; see paragraph 27 below). It was Mr Jarvis's evidence that this was not a significant concern for NR and certainly did not give rise to any preference for buying from TWM, because more important than cost was deliverability. WWH's proximity to the WCML made it an important and reliable source of supply for the western side of the country. Mr Heubeck disagreed with that evidence, and we shall have to say more about that disagreement later.

Contracts and tenders

The 2011 and 2012 contracts: MGVs and price banding

21.         The contract with TWM was made in 2011, so that the Doncaster factory was under development during the last three years of operations at Tallington; the contract will expire in 2026. In 2012 NR entered into a five-year contract with the claimant for production at WWH; the contract included an option for NR to extend its term by up to two individual years to 31 March 2019. The TWM and WWH contracts both featured two important terms:

a.       A minimum guaranteed volume ("MGV"); NR agreed to buy a minimum of 250,000 sleepers each year from the claimant, and 200,000 a year from TWM once production at Doncaster was under way. As Mr Jarvis explained, and we accept, this term was mutually beneficial; it gave the producer a guaranteed income, and it meant that the factory overheads could be spread over the term of the contract rather than being front-loaded on to prices.

b.      A price-banding system. In the WWH contract, if NR bought fewer than 350,000 in a year a surcharge was applied to all sleepers ordered: 12% if between 300,000 and 349,000 were bought, 24% if the order was between 250,000 and 299,000. Fewer than 250,000 was not possible because of the MGV. On the other hand, price reductions were applied to sleepers bought in excess of 350,000, again in bands of 50,000. TWM's contract contained similar provisions. From NR's point of view that had an effect on what was ordered from each supplier; as Mr Jarvis put it in cross-examination:

"we would play tunes on the contract price banding with both Cemex and Trackwork Moll, to try and get the best possible rate for Network Rail."

 

22.         Both contracts also made provision for annual indexation of prices, calculated by a formula dependent on the price of materials. Thus the prices initially charged under the 2012 WWH contract increased each year.

23.         After 2012, with contracts in place for supply from Doncaster until 2026 and from WWH until 2017 and potentially 2019, NR was able to turn its attention to the possibility of getting a sleeper factory established at an LDC on the west side of the country. It would have done so in any event, because the policy of co-location of production at LDCs was important to NR, but from 2014 onwards it became obvious that another problem was coming down the track: the growing likelihood that the WWH site would be compulsorily purchased as a result of the construction of HS2. A Bill making provision for the first phase of its construction had its first reading in Parliament in November 2013. The claimant was aware that its site at WWH was identified as being subject to compulsory acquisition, and petitioned against the Bill; its petitions were withdrawn shortly before the end of 2016 and the Bill became law on 23 February 2017. To jump ahead, the property at WWH vested in the compensating authority on 15 May 2020.

The P1 and P2 procurement processes

24.         The procurement process known to the parties as P1 began in June 2015 when NR issued an Invitation To Tender ("ITT") for a 15-year sleeper supply contract from one of eight sites, some of them LDCs including Bescot, but the Invitation also stated that alternative sites could be considered. The plant was to provide up to 600,000 sleepers per year, and the factory was to be fully operational by October 2017. The claimant submitted a number of bids, including one for supply from WWH. In April 2016 the P1 process was discontinued because, NR said, its requirements had changed.

25.         The procurement process known as P2 began in September 2016, and was again for the grant of a 15-year contract requiring the construction of a factory capable of producing 600,000 sleepers per annum; this time bids had to be for a factory at Bescot. NR owned the LDC at Bescot and it wanted to have an ownership interest in the sleeper factory. However, there was of course no existing sleeper factory at Bescot, nor was there planning permission to build one.

26.         The claimant submitted two bids. The first was a bid to construct and supply sleepers from a factory at Bescot. A second separate bid, acknowledged to be non-compliant, was to supply sleepers from WWH. It included an offer to transfer Cemex's leasehold interests in the land there to NR. Both the claimant's bids were unsuccessful. The claimant's Bescot bid was ranked in third place; in June 2017 the contract was awarded to RailOne GmbH. NR then sought planning permission for a new sleeper manufacturing facility at Bescot; the planning statement made by NR in support of the planning application has been a source of evidence about NR's view of the volume of sleepers it was going require from Bescot, to which we return later.

27.         In July 2017 the claimant commenced a High Court action challenging the decision to abandon P1 and replace it with P2. The challenge was subsequently discontinued, but not before witness statements had been filed. One of those statements, made by Jonathan Curtis, NR's Head of Commercial Materials, Route Services Contracts and Procurement Team, has also been a source of information about NR's thinking and expectations at the time. For example, it is the source of the figure we have been given for the annual "trip cost" of taking sleepers from WWH to the LDC at Bescot.

28.         The failure to win the P2 contract was of course not inevitably the end of the road for WWH; the Bescot plant needed planning permission, and then needed to be built, so a replacement supplier was not yet in place and was not guaranteed. And individuals at WWH - among them Mr Neil - cherished the hope that the compulsory purchase might yet be fended off.

29.         Furthermore, NR was taking steps to protect its supply of sleepers in light of the impending compulsory acquisition of WWH by building up a stockpile of sleepers, first at Carlisle, the most north-westerly LDC, and later also at Tyne Yard just south of Newcastle. According to Mr Jarvis, the first sleepers to be stockpiled, at the end of 2014, were the last 50,000 or so produced at Tallington before its closure. More were purchased from both TWM and the claimant in the years 2017 to 2019. The size of the stockpile is not agreed, and we discuss that later, but on any reckoning was somewhere between 700,000 and 900,000. About 440,000 sleepers have been withdrawn from the stockpile, which is perhaps surprisingly low given that (as we shall see) the claimant was unable to produce any sleepers between 2020 and 2022 so that one of NR's only two suppliers was out of action for two years; but demand was low during those years, not least because of the effects of the pandemic. So, a substantial stockpile remains.  We should mention at this point that it is agreed that the stockpile is a real world factor, and as we explain later would not have occurred in the no scheme world.

30.         It was therefore not in NR's interests for WWH to close at the end of the five-year term of the 2012 contract on 31 March 2017, because that would leave only TWM, with its smaller capacity which, on the basis of the previous few years, would certainly not have been adequate to meet NR's requirement. So the WWH contract was extended for two one-year periods, and two further short-term extensions were put in place. Sleepers continued to be acquired from WH until its closure in 2020 both for NR's requirement for track and for the stockpile. The claimant was able to negotiate a significant increase in its prices in those four short contracts, and as a result was able to continue to operate profitably until shortly before WWH closed.

The P3 procurement exercise and contract

31.         By early summer 2019, therefore, WWH is still producing sleepers, NR is still trying to get planning permission at Bescot, and the site at WWH is going to vest in the Secretary of State within the next 12 months. Even if planning permission was going to be granted for Bescot - and it was no doubt obvious to everyone that that was not a given, in light of local opposition - NR's supply of sleepers was looking vulnerable because a new factory at Bescot was going to take two to four years to be built and get going.

32.         So in September 2019 NR issued an ITT for a contract for the supply of sleepers, which the parties have referred to as P3. It was for a four-year term, potentially extendable by up to four single years (so a maximum term of eight years); the notice stated that the successful supplier would receive an initial order for 200,000 sleepers and thereafter 50,000 per annum "although this may increase to a range of 250,000 to 400,000 a year should unforeseen difficulty in supply arise." No site was specified.

33.         The claimant submitted two bids in November 2019. One was for supply from its site at Rochester. This was a small site, already producing some specialist sleepers. The bid was fully worked-up including all the details one would expect including the pricing of the two standard types of sleeper and the wide range of specialist sleepers, as well as price banding and so on. The other was a single page letter, signed by Mr Neil and sent without approval from the Cemex board. It was for supply from WWH; it offered significantly lower prices than the claimant had been charging in 2016 (before the increased prices under the contractual extensions) and lower prices than those contained in its P2 bid, and it listed only three types of sleeper. Mr Neil explained that the letter was "a sprat to catch a mackerel"; he wanted to open up a dialogue and raise the possibility that NR might decide that it still needed WWH and might therefore apply pressure for it to be reprieved from compulsory acquisition. We might describe the letter as a last minute plea for mercy written on behalf of a business that had nothing to lose. As Mr Neil put it, nothing more was heard of that bid.

34.         Both bids were submitted in November 2019. In December 2019, planning permission for Bescot was refused. NR did not appeal that refusal. Mr Jarvis spoke eloquently of the sense of let-down and frustration experienced within NR; Bescot had been a project for several years, a lot of effort had gone into it, and all for nothing. The RailOne contract came to an end.

35.         There remained for NR the immediate problem of supply, with the disappearance of the WWH factory now imminent. The claimant's Rochester bid in response to the P3 invitation was successful, and the claimant is now manufacturing sleepers there under a four-year contract which we outline below.

36.         The Rochester site is agreed by all concerned to be less than ideal. It is not at an LDC, and it is on the wrong side of London for supply to the west side of the country. Access to the main rail network is via the congested Hoo Junction. Vehicular access is very difficult (through a tunnel).  It is too small; while its maximum capacity is a notional 400,000 sleepers per annum, it has insufficient storage capacity to operate comfortably at a production rate of over 200,000. Due to its size, it is a carousel plant, with machinery supplied by a Dutch company and specially adapted for the layout of the site. Installation of the plant commenced in June 2021 and commissioning was completed in April 2022. Low level production for testing purposes began in December 2021. The plant became fully operational in April 2022. It has yet to generate a profit.  

37.         It will be appreciated from what we have said that while NR has a dual source of sleeper supply at the moment, the future is uncertain. The TWM contract will expire in 2026 and therefore there will have to be a procurement exercise; TWM may or may not secure a new contract; any new manufacturer will need time to get going - hence, it is thought, the retention of a significant part of the stockpile. The Rochester contract can continue until 2030, but the factory has limited capacity which may become a problem when NR's requirement recovers from its current low level; in any case the location of Rochester is so poor that one suspects almost any site to the west of London, whether or not at an LDC, would be preferable. So the future of the claimant's business is hard to predict.

The issues in the appeal

38.         As explained above, the Tribunal has to decide the profit or losses the claimant would have made absent the compulsory purchase in the imaginary "no scheme world", and the profit or losses made already and to be made in the future in the real world. We can then determine whether the claimant's sleepers business has suffered a loss of profits as a result of the compulsory purchase, and if so the amount of compensation payable, or, as the Secretary of State says, has made a gain which must be set against other sums payable to it.

39.         We have heard evidence so far from the claimant's witnesses of fact, and from expert witnesses called by both parties in relation to the manufacture and supply of sleepers. That evidence enables us to make findings about the market in sleepers, the claimant's share of that market, and the prices it has charged, will charge and - in the no scheme world - would have charged, and also about the likely duration of the claimant's business, in both the real and no scheme worlds. But we cannot reach a final determination because we have not yet heard evidence about the claimant's costs, which will generate sums for profit and loss, of which the future elements have to be discounted to give an overall figure at the valuation date. That missing evidence will be supplied by the forensic accountants from whom we have yet to hear, and their evidence and calculations must follow from the findings we make about the evidence we have heard. It is for the sleepers experts and not for the accountants to give evidence about price, about the volume of sleepers required in the market, and about the claimant's market share.

40.         Important features of those two worlds are known. In the real world, the claimant is supplying sleepers from Rochester under a contract made in 2020, for a supply of sleepers for four years with a possible four one-year extensions, and with no MGV. The price of sleepers in the past and under the present contract is known; the volumes required by NR 'to track' (i.e. not sent to stockpile) to date, from 2017/18 onwards, are the subject of some dispute, and the size of the future market in sleepers and the claimant's share of that market are hotly contested. We also have to make findings about the duration of the claimant's business, at Rochester and elsewhere, in the real world.

41.         In the no scheme world it is agreed that NR's requirement for sleepers is the same as in the real world, past and future. It is agreed that in the no scheme world the 2012 contract would have been extended as it was in the real world (paragraph 30 above), that there would have been a P3 ITT in 2019, and that the claimant would have been awarded the P3 contract to supply sleepers from WWH. We have to determine the likely duration of the claimant's business at WWH, and make findings about pricing and about market share to date and in the future in the no scheme world.

42.         We address the issues between the parties under the following heads.

43.         First, the volume of sleepers required by NR in the real world and the no scheme world to date and in the future; it is agreed that the two worlds are identical in this respect, although in the real world there is some distortion because of the effect of the purchases made by NR for its stockpile, which as we have said is agreed would not have been made in the no scheme world. By "to date" we mean the period from 2017to the end of March 2025 because the parties agree that the requirement before that period makes no difference to the claimant's claim since sales figures and prices before then are known and are the same in both worlds. We address the volume issue under two separate heads, one relating to the position to date and one to the future. As to the requirement to date, we have to consider whether figures supplied by NR are accurate and whether Mr Jarvis's recollection should be preferred for the years where he disagrees with NR's figures. The authority's position is that NR's figures are accurate, and that those figures enable us to see what the claimant supplied to the stockpile and therefore would not, according to the authority, have sold in the no scheme world until later when they were required. As to the future, the parties are a long way apart. The claimant asserts that while the current demand for sleepers is at an all-time low it will pick up again and reach a high level in the next ten to 15 years; the authority says that demand will rise, but by far less.

44.         Second, we make findings about the likely duration of the claimant's business, in the real world and the no scheme world.

45.         Once those two issues are determined we do not need to make any further findings about the real world; the number of sleepers sold by the claimant to date and the prices it charged, are known, future pricing under the current contract is known, and it is agreed that the claimant's share of the market for as long as it is operating from the Rochester site will be 30%. Those findings and agreements are the basis on which the accountants will give their evidence about costs and about the calculation of profits and losses past and future.

46.         Then we have to turn to the no scheme world. We look first at the period from 2017 to early summer 2020, between the expiry of the 2012 contract and the start of the P3 contract at WWH. We have heard only very limited evidence about what would have happened in the no scheme world during this time of just over two years. Price, sales volumes and market share (the latter informed by our findings about NR's requirement to date) are closely linked during this period, and we address them together.

47.         Then we turn to the major disagreement, about which extensive evidence has been heard, about the likely terms of the WWH P3 contract; we look separately at prices, at whether there would have been an MGV, and at the market share that would be achieved in the no scheme world.

48.         In our summary of the issues above, and our substantive discussion below, we have discussed the market as if NR were the only customer. Since NR accounted for at least 90%, perhaps 95%, of the claimant's business that was the approach adopted by the parties and we are content to adopt that approach.

49.         On all the issues the evidence given by the parties' experts on sleepers has been of central importance. The claimant called Mr Paul Jarvis, who worked in the railway industry for 38 years from 1984 until his retirement in 2022. He was a Chartered Member of the Institute of Logistics and Transport and obtained an MBA from Sheffield Business School in 1997. From 2009 to 2022 he was NR's Business Manager with responsibility for the supply of sleepers for the railway network. He explained that he would devise a strategy for supply and advise the Contracts and Procurements department on what was needed, then that department would carry out a procurement process; once the contract was in place he and his team would take over contract management. The endeavour to establish a factory at the LDC at Bescot was his strategy.

50.         So for 13 years Mr Jarvis knew how many sleepers NR needed, and he ensured that it had them. That was his responsibility in the real world, and it would have been his responsibility in the no scheme world. However, he did not manage all aspects of the process; he determined demand, but the Contracts and Procurement team arranged ITTs and agreed the terms of NR's contracts with its suppliers.

51.         Mr Heubeck joined British Rail in 1979; from 1998 to 2004 he led the implementation of new distribution arrangements, and from 2004 to 2009 led the national track materials procurement team, managing the supply of rail, sleepers and ballast. In 2009 he joined Trackwork Limited and formed the joint venture with Leonhard Moll of Munich, which became TWM; from 2011 to 2015 he was general manager of TWM and from 2015 to 2023 he worked as a freelance consultant assisting clients with tenders for materials supply contracts, including Tarmac Major Projects for the supply of slab track to HS2.

52.         Both experts were impressive in terms of their knowledge and experience in the railway industry, and we are confident that both gave the Tribunal their honest opinions. Both on occasions were over-optimistic in favour of the party who instructed them. It is not possible to say that on the whole we preferred the evidence of one or the other and we have had to consider their views issue by issue, along with the rest of the evidence.

Issue 1(1): the volume of sleepers required by NR to date in the real world and the no scheme world

The information provided by NR

53.         As we said above, it is agreed that the number of sleepers required by NR is unaffected by the scheme and so is the same in the real world and the no scheme world. The timing of its requirement does differ, because in the no scheme world there is no stockpile, but the total number of sleepers required is the same. We use the term "required" (etc) to mean what NR needed for work on the track, and "demand" to mean what it bought from its suppliers each year. Thus in the years when NR was stockpiling sleepers demand exceeded requirement. In the no scheme world there is no stockpile and sleepers would have been bought when required; thus we need to ascertain NR's requirement to date. We have to do so because figures supplied by NR about its requirement to date and about the stockpile are not agreed to be accurate. The volume of NR's requirement will then be needed at a later stage in our decision when we turn to the number of sleepers the claimant would have sold, to date, in the no scheme world; in order to determine the market share it would have achieved we have to make a finding about the size of that market.

54.         NR has provided information in response to questions asked of it by the parties in the course of proceedings. The claimant has concerns about its accuracy and indeed it is clear that some figures cannot be right. That causes difficulties because there is no evidence from NR. We do not know how or by whom the figures have been produced, and the person who produced them cannot be asked for an explanation or clarification and cannot be cross-examined. An added problem is that it is difficult for NR to quantify the stockpile. A physical stock count is awkward because the sleepers are stacked close together, so that it is not possible to walk between the stacks, in order to prevent vandalism. Mr Jarvis explained that a drone is used to survey the stacks from above, but there is obvious scope for error. Discrepancies in the available data about the stockpile are said by the claimant to cast doubt on the figures as a whole, and as we shall see there is some truth in that.

55.         The following table sets out what NR has said about its requirement, for sleepers for the track, from 2009/10 onwards (with financial years ending on 31 March in each year). The first column contains the figures for 2009/10 to 2015/16 taken from a document produced by NR in 2016 for tenderers in the P2 procurement exercise. In the second column are the figures given by NR for the sleepers it purchased from both TWM and the claimant for work on the track, in response to questions addressed to it in this litigation. The shaded cells are those where NR itself stated that the figures were "believed to be unreliable". The figure for 2024/25 was provided in December 2024 and is estimated so far as the final three months are concerned. As we said above, we are not asked to determine the requirement before the year 2017/18, but we have included the earlier figures because it will be important to see the sort of quantities NR required before that date. In the third column are the numbers of sleepers that NR has told the parties, in response to their questions, that it has withdrawn from the stockpile. In the final column are the totals of the second and third column, adding NR's purchase for track to its withdrawals from the stockpile to give NR's total requirement, according to NR's own figures, for the years from 2017/18 to date. Throughout the table the figures that are challenged by the claimant are emboldened.

Table 1: NR's data for the sleepers purchased, and withdrawn from stockpile, for work on the track 2009/10 to 2020/21

Financial year	NR's 2016 figures for its requirement	NR's 2024 figures for purchases for track from TWM and the claimant	Withdrawals from stockpile	Total requirement
09/10	523,672
10/11	573,738
11/12	702,896
12/13	716,314
13/14	821,419
14/15	869,625	739,346
15/16	633,854	730,879
16/17		622,930 or 624,241
17/18		327,796		327,796
18/19		369,390	4,000	373,390
19/20		292,853	71,597	364,450
20/21		325,690	83,432	409,122
21/22		392,051	104,452	496,503
22/23		337,048	133,151	470,199
23/24		347,139	44,575	391,714
24/25		307,309	0	307,309

           

 

56.         The figures given in the second column for NR's purchases for track in 2016/17 onwards are further broken down by NR into purchases from the claimant and TWM. According to NR, the claimant provided the following numbers of sleepers for track; again, the figures from 2017/18 onwards are not accepted by the claimant.

         Table 1a: sleepers sold by the claimant for work sites

16/17	361,819
17/18	153,941
18/19	28,774
19/20	66*

         *sic, not 66,000 

57.         The claimant has supplied figures for its total sales to NR to date (for track and stockpile together), and they are unchallenged. The figure given in table 1a for 2016/17 is about 10,000 lower than the claimant's own figure for its sales to NR in that financial year, which bears out NR's own observation that the figure is unreliable. NR has also supplied a break-down of what the two suppliers sold by sleeper type. An earlier version of the figures which appeared to show that the claimant did not sell any sleepers with USPs (see paragraph 12 above) was obviously incorrect and NR corrected it.

58.         If the figures in table 1a were accepted by the claimant then we would know how many sleepers the claimant sent to the stockpile, but those figures are not agreed; and the claimant is unable to supply its own figure for the sleepers it sent to the stockpile. Mr Heubeck thought the claimant should be able to do so on the basis that sleepers going to the stockpile were identifiable because they went out by road not rail, and did not have clips fitted; but Mr Jarvis and Mr Neil's evidence was that although most went by road, not all did, and that some sleepers went by road to track. We accept the evidence given for the claimant that it does not know how many sleepers went from WWH to the stockpile.

59.         NR has supplied its own figures for the number of sleepers sent by the claimant and by TWM to the stockpile. The figures are detailed, being broken down by sleeper type; in total, NR said that the claimant sent 304,161 and TWM sent 321,273, giving a total of 625,434. However, NR's chart setting out those figures is annotated as follows:

"Please note a physical stock count was undertaken in 2020 showing a further ~100,000 sleepers. However, these stock counts are known to be inaccurate due to the nature of the activity."

60.         So we can take it that the stockpile at some point in 2020 contained, according to a physical count (and we have noted the limitations in such counts, see paragraph 54 above) about 725,000 sleepers. In December 2024 NR stated that a physical count of the stockpile in July 2022 indicated that it contained 526,278 sleepers - and of course we know that by then a substantial number had been withdrawn (the third column of table 1 above).

61.         In setting out what the parties say about NR's requirement to date it is easiest to start with the authority's position, because it advanced a positive case on the basis of NR's figures.

The authority's case about NR's requirement to date

62.         For the authority, Mr Heubeck expressed confidence in NR's data, and it was argued that because NR expressed doubt about some of its figures (the shaded cells in tables 1 and 1a above) the rest could be taken as correct. Mr Heubeck took the physical stockpile count from July 2022 (526,278) and observed that by 31 March 2022 NR had used 263,481 sleepers (column 3 of table 1 above). 133,151 sleepers are said to have been withdrawn in 2022/23, and so he suggested that one quarter of 133,151 might have been used between April and July 2022. Accordingly, Mr Heubeck said, if the July 2022 count was right, then the stockpile in total was (526,278 + 263,481 + 33,288) = 823,047.  

63.         Mr Heubeck then took the claimant's own sales data, which have not been challenged, and subtracted from them the number of sleepers said by NR to have been purchased from the claimant for use at work sites (table 1a, paragraph 55 above). If those figures are correct then the claimant sent 525,386 sleepers to the stockpile:

Table 2: Mr Heubeck's calculation of sleepers sent by the claimant to the stockpile

	Claimant's sales data	Subtract claimant's sales for work sites	Totals
17/18	303,437	153,941	149,496
18/19	222,881	28,774	194,107
19/20	181,849	66	181,783
			525,386

 

64.         We have not yet heard evidence from the forensic accountants, but it is useful to be aware of the use the authority's forensic accountant made of that figure. Mr Adam Smith, for the authority, expressed the view that sleepers sold for the stockpile in the real world would have been sold for track between 2020 and 2024. He therefore subtracted 500,000 (rounding Mr Heubeck's 525,386) from the claimant's sales data in 2017/18 to 2019/20 (making the assumption that what the claimant would have sold during those years would otherwise have been unchanged) and then spread the 500,000 evenly across the following four years. Thus on the authority's (as yet untested) case 500,000 sleepers sold at an inflated price in the real world would have been sold for far less, and later, in the no scheme world. The claimant resists that conclusion. This decision is not about the accountants' evidence, but we mention Mr Smith's calculation in order to explain why the parties attached a great deal of importance to the stockpile figures.

The claimant's position about NR's requirement to date

65.         The claimant's evidence about NR's demand to date comes from Mr Jarvis, whose task it was until 2022 to supply that demand. In his first witness statement he said that he experienced a high of approximately 950,000 sleepers in 2014 and a low of approximately 500,000 as a result of the pandemic in 2020 and 2021. He added that a working average assumption for demand was always in the region of 750,000 per annum, and pointed out that Mr Jonathan Curtis said the same in his witness statement (see paragraph 27 above). The difficulty Mr Jarvis has is that he left NR in 2022 and no longer has access to NR's data; so he was able to recollect in terms of round numbers but has no access to precise figures.

66.         NR's figures for its requirement from 2020/21 are accepted by the claimant, on the basis that volumes were heavily impacted by the pandemic and by the financial constraints that followed (when NR was required to make cuts after the heavy subsidies provided during lockdown and beyond).

67.         As to the figures in dispute, as we said above we are not asked to make a decision about 2014/15 but it may be helpful to explain that Mr Jarvis thought the 869,625 given by NR in 2016 was too low (let alone the even lower figure NR supplied later - see the second column of table 1 above). In cross-examination he recalled giving a presentation to colleagues at the end of that year and saying that he had sourced 930,000 sleepers for NR that year. That is the sort of thing one remembers, and we think he was probably right, but we bear in mind that NR's figures may be different because of the way purchases have been allocated across the financial year-end.

68.         Mr Jarvis also disagreed with the 4,000 given for the first withdrawal from the stockpile in 2018/19 (the first figure in the fourth column of table 1 above); he thought it was too low. He pointed out that the wagons on which sleepers were transported carried multiples of 56 so that 4,000 did not look right.

69.         Mr Jarvis was adamant that the figures given by NR for its requirement in 2017/18, 2018/19 and 2019/2020 were too low. His recollection was that the requirement was nearer to 500,000 in those years.

70.         In light of Mr Jarvis' evidence, the claimant's position is that NR's figures are in general too unreliable for any weight to be placed upon them. In support of that position the claimant made three points.

71.         First, other documents indicate that the volumes NR was expecting, going forwards, were much higher than they should have been expecting on the basis of the figures now given, in particular Jonathan Curtis' witness statement (indicating an average requirement of 750,000 (see paragraph 27 above) and the planning statement for Bescot indicating an annual requirement of "up to 1,000,000 sleepers each year".

72.         Second, certain errors are obvious; in paragraph 57 we noted the problem with NR's data about the supply of USPs, which was demonstrably wrong, and the discrepancy between NR's figure and the claimant's sales data.

73.         Third, Mr Heubeck's calculation of the claimant's contribution to the stockpile (table 2 above), based on what NR says the claimant sent to track, is about 225,000 higher than what NR said the claimant sent to the stockpile (paragraph 59 above). Both those figures cannot be right, and thus NR's own figures are inconsistent with each other. Mr Heubeck's calculation of the claimant's contribution to the stockpile is also inconsistent with his own  calculation of the size of the stockpile: using round numbers for a moment, if the claimant sent 525,000, TWM sent 320,000, and Tallington contributed 50,000 (see paragraph 29 above) then the stockpile comprised about 895,000 sleepers which is considerably larger than Mr Heubeck's 820,000, again based on figures provided by NR (and than Mr Jarvis's recollection that the stockpile was 750,000 or more, but not 800,000).

74.         On that basis, it was argued for the claimant, no reliance should be placed upon NR's figures insofar as they differ from Mr Jarvis's recollection. Moreover, the size of the stockpile, and of the claimant's contribution to it, cannot be known, and in particular Mr Heubeck's figure for the claimant's contribution to the stockpile should not be used as Mr Smith does to calculate the number of sleepers the claimant would have sold in the no scheme world to date.

Discussion and conclusions on NR's requirement to date

75.         It is clear that the Tribunal does not have a complete set of accurate data about NR's demand or requirement from 2017/18 onwards to date (which, since this decision is issued in April 2025, we take as the end of the financial year 2024/25). However, we do accept as accurate figures that are unchallenged, in particular the claimant's sales data, NR's withdrawals from the stockpile from 2019/20 onwards, and NR's requirement from 2020/21 onwards.

76.         We also accept that 4,000 sleepers were withdrawn from the stockpile in 2018/19, in the absence of anything specific from Mr Jarvis.

77.         So the real challenge is to the years 2017/18 to 2019/20. Mr Jarvis's recollection was that NR's requirement was close to 500,000 in each of those years. NR says it purchased 327,796, 369,390 and 292,853 sleepers respectively; when added to the withdrawals from the stockpile the total requirement was:

2017/18: 327,796

2018/29: 373,390

2019/20: 364,450

 

78.         We know that there is an internal inconsistency within NR's own data for those years; as the claimant demonstrated, the figures given for the claimant's sales to track and for its contributions to the stockpile cannot both be right. The figures given for the claimant's sales to track (table 1a) are a component of NR's figures for its total requirement (table 1), and in light of Mr Jarvis's recollection we agree with him that the figures given by NR for its requirement for these three years must be too low.

79.         We said above that we do not know how NR's figures were produced; but there is an exception to that. We have been told the outcome of a physical count of the stockpile in 2020, and of another in July 2022. Mr Heubeck observed that the count in July 2022, when combined with what was withdrawn by that date, indicates a stockpile of around 820,000 sleepers (we round the number because the figure for withdrawals between April and July is an estimate). We know that at some point in 2020 the stockpile stood at around 725,000 (paragraph 59 above), and we have accepted that by 31 March 2020 about 75,000 sleepers had been withdrawn. We do not know when in 2020 the count was made, but if, as in 2022, the count took place in the summer, then more than 75,000 had probably been withdrawn by then, again consistent with a stockpile of something over 800,000 sleepers. We are well aware that the count might have been made in December, in which case more would have been withdrawn and the total would be higher. But we think that it is significant that if we take the sort of quantity likely to have been withdrawn by June or July we get a figure consistent with the 2022 count. That may be a coincidence, but we take the view that it is more likely to mean that the physical stock counts were not so unreliable after all. We find, on the balance of probabilities, that the stockpile comprised, very roughly, in the region of 820,000 sleepers.

80.         That indicates, as the claimant suggested, that Mr Heubeck's figure for what the claimant sent to the stockpile was too high.

81.         It also means that NR's figures for the contributions made by TWM and the claimant (320,000 and 300,000 respectively, see paragraph 74 above) are too low. If we add to those two figures the contribution of 50,000 from Tallington, we have about 670,000, which is about 150,000 too low. There is no challenge to the figure from Tallington and we conclude that TWM and the claimant together sent 770,000 sleepers to the stockpile, so that the figures given by NR for the two companies' contributions are, taken together, wrong (as NR itself indicated by adding a note to the effect that the physical count of the stockpile showed "~100,000" more) by about 150,000. The error may be in the TWM figure, the claimant's figure, or both. We have no information about TWM's sales.

82.         Accordingly we have to conclude that the claimant sent no less than 300,000 and no more than 450,000 to the stockpile. Any less than 300,000, and the stockpile is too small. Any more than 450,000, and it is too big (as a little experimentation with table 2 above demonstrates). If the claimant's contribution to the stockpile was 450,000 (so that all the error is in NR's figure of 300,000 for the claimant's contribution) then NR's figures for what the claimant sent to track are 75,000 too low (again, this can be seen by playing with table 2). If the claimant's contribution was 300,000 then NR's figures for what the claimant sent to track are 225,000 too low.

83.         We have no way of determining how big is the error in what NR says the claimant sent to track. But we do not have to decide how many sleepers the claimant sent to the stockpile; we have to decide NR's overall requirement, which will determine what the claimant would have sold in the no scheme world once we have determined its market share in that world (issue 3 below). The figures given by NR for what the claimant sold for work on the track in these three years can be seen to have been between 75,000 and 225,000 too low. We have no basis however for saying that what NR said TWM sold to the track (again, components of the figures in column 2 of table 1, which we have not set out here) was incorrect.

84.         NR's own figures give its total requirement in these three years as 1,065,636 (see the final column of table 1 above, and paragraph 77). Mr Jarvis's recollection that the figures were nearer 500,000 in each of those three years indicates that NR's totals are over 400,000 short. We bear in mind that Mr Jarvis does not claim to recall precise figures, but we can rely on his memory at least so far as to say the error must be at the upper end of the range of 75,000 to 225,000. We find therefore (a) that NR's figures for what it required for track in total in the three years 2017/18 to 2019/20 were too low and (b) that they were too low by about 225,000. It is fair to say that we have far more confidence in (a) than (b), but we nevertheless make that second finding on the balance of probabilities.

85.         We are conscious that we have laid a lot of weight on the size of the stockpile, on physical counting of the stockpile, and on the figures for withdrawals from the stockpile. As we say, we are doing the best we can on the basis of the limited and conflicting information we have. We make the following findings of fact:

a.       That a total of approximately 820,000 sleepers were stockpiled (on the basis of two different stock counts and the unchallenged figures for withdrawals).

b.      That the claimant sent between 375,000 and 450,000 sleepers to the stockpile, and that therefore the figures given by NR for the claimant's contribution to the stockpile are too low.

c.       That NR's figures for its requirement for 2017/18 and 2018/19 and 2019/20 were too low by around 225,000 in total.

86.         The accountants need figures to work on for the years from 2017/18 to date. If we share the 225,000 shortfall arbitrarily between those three years, purely for the sake of the accountants' calculations, then NR's total requirement to date (giving rounded figures for the years where we have added an approximate figure) was:

2017/18:   400,000

2018/19:   450,000

2019/20:   440,000

2020/21:   409,122

2021/22:   496,503

2022/23:   470,199

2023/24:   391,714

2024/25:   307,309

87.         We do not comment in this interim decision on the accountants' methodology, but the forensic accounting experts should use these figures for NR's requirement to date in the no scheme world rather than constructing their own version of that requirement by re-distributing, in the no scheme world, the sleepers sold for the stockpile in the real world.

Issue 1(2): NR's future requirement for sleepers in the real world and the NSW

88.         Now we turn to the future. We have to decide how many sleepers NR will require in the future so that we can determine how many sleepers the claimant will sell in the real world, and would have sold in the no scheme world.

The background to future demand

89.         The context for this part of our decision begins with the figures we set out in the first two columns of table 1 (at paragraph 55 above) above, from which it can be seen that NR's requirement over the last 16 years rose to a peak in 2014/15, dropped, fluctuated, and is now at an all-time low. The sleepers experts disagree about the reasons why it is so low at the moment; they agree that it will pick up, but not by how much.

90.         The context also requires an explanation of two things; first, of the use by NR of track renewal trains and, second, of how NR is funded.

91.         As to track renewal trains: Mr Jarvis explained that the conventional approach to track renewals involves the old track being cut and removed in 60 ft lengths, loaded on to flat wagons and taken away. New sleepers, delivered to an LDC, are then shunted and brought to the worksite, lifted off using manually operated worksite plant, and placed into the track bed. Rails are threaded on to the sleepers and clipped into place, again using manually operated plant.  Such work is generally done at weekends with the track closed for 28 or even 52 hours, with replacement bus services provided in the meantime.

92.         A track renewal train is a train that renews track in a fully automated process. It is loaded with sleepers and rails; at the worksite it removes the rails and the old sleepers, places in new sleepers, and replaces the rails. It consists of automated machinery and equipment that can be loaded with new sleepers and is then hauled to a site of work on the railway network.  It then replaces the rails and sleepers, taking the old sleepers away to be unloaded. A track renewal train can renew half a mile of sleepers on a midweek night, involving closure of the track only between 22:30 and 05:00, with almost no disruption to services. Although fast and efficient, track renewal trains put a lot of pressure on sleeper manufacturers because large quantities of sleepers are needed, together with extra space for loading and unloading. They also have an impact upon the site where sleepers are manufactured because extra-long sidings are needed to accommodate the wagons used to pick up sleepers to be loaded onto the TRTs. The first track renewal train used on the network, Mr Heubeck explained, was bought for the WCML Upgrade project which was completed in 2009, and supplied with sleepers from WWH; NR bought a second one in 2011. After 2015 a lower volume of work was needed; only one track renewal train was in use after 2018/19 and the other operated only until 2024; they are no longer in use and there is no prospect of their being used in the future.

93.         According to Mr Heubeck, track renewal trains increased NR's requirement for sleepers because, whereas traditional renewal methods would replace only sleepers that were truly life-expired, it was inefficient to use the track renewal trains like that; they needed to replace long stretches of track in one pass. The use of trains therefore increased the requirement for new sleepers; conversely, according to Mr Heubeck, the cessation of their use is a reason for the current low requirement. Mr Jarvis expressed the view that the difference in the amount of sleepers used was minimal.

94.         As to NR's funding: Mr Jarvis explained that NR's track maintenance and renewal is funded in five-year Control Periods, or CPs. The relevant CPs are:

CP6 1 April 2019 to 31 March 2024

CP7 1 April 2024 to 31 March 2029

CP8 1 April 2029 to 31 March 2034

CP9 1 April 2034 to 31 March 2039

95.         At the date of this decision the first year of CP7 has just come to an end.

96.         NR is informed of the budgetary settlement for each Control Period and then decides, in a Strategic Business Plan, how to spend the money - whether on signalling schemes, structure renewals (such as bridges and tunnels), earthworks, or track renewals including sleepers. Mr Heubeck explained that the business plan is then submitted to the Office of the Rail Regulator, which examines its efficiency and effectiveness and issues a determination about NR's funding and planned activities. This is a protracted process; once it is completed the resulting requirement for sleepers is then submitted to NR's centralised Supply Chain Operations, which is responsible for supply chain management and decides how the supply is to be split between the two sleeper suppliers.

97.         Importantly, therefore, NR cannot control how much money it gets but it can make its own decisions about how the money is allocated, subject to the scrutiny of the ORR. We have been assisted in our consideration of future requirement by the terms of the Strategic Business Plan for CP7.

98.         With those aspects of the context in mind, we turn to what the parties said about future requirement. It is agreed that the NR's current requirement (just over 307,000 in 2024/25) is an all-time low. In their Statement of Agreed Facts and Issues ("SAFI") the sleepers experts said:

"4.51 NR's demand for concrete sleepers in the current 2024/2025 year is at a

historically low level of demand which is due to a combination of factors:

 

4.51.1 High volumes of concrete sleepers being replaced between 2002 and 2018. 

4.51.2 Modern sleepers, which are part of an improved overall track system, lasting longer than their predecessors.

4.51.3 NR ceasing to use Track Renewal Trains as a means of delivering high volumes of "strategic" track renewals. 

4.51.4 Improved track condition data being routinely provided to local track engineers, enabling more precise specification of smaller scale "tactical" track renewals.   

4.51.5 Within the track engineering discipline, rerailing, re-ballasting and certain switch and crossing renewals all being accorded a higher priority than the replacement of concrete sleepers. 

4.51.6 A reduced spend on track renewals overall, due to an urgent need to increase the spend on other assets, at a time when the overall funding settlement is relatively constrained."

The claimant's case about future requirement

99.         For Mr Jarvis, those factors are temporary and do not indicate that requirement will remain low. Expenditure on track renewal is both cyclical and unavoidable. It is cyclical in that it rises and falls across several decades. In the 1980s and 1990s in the run-up to privatisation money was tight; there was the crash at Hatfield in October 2000 when four people died; and then expenditure rose in the next two decades with NR "playing catch-up", with "all sorts of money being spent on track renewals" - including the WCML project between Stoke and Manchester, which was completed in 2009. Expenditure reached a high in 2014/15, and it is now at a low point in the cycle. His view of the points identified in the sleepers' experts' SAFI, set out above, was that they were temporary and did not represent a systemic change. The track renewal trains did use some extra sleepers but not many; the additional work done in the first two decades of the millennium will, before long, need re-doing because of the finite lifespan of sleepers; improved detection systems do not mean that sleepers do not have to be replaced; and as to funding constraints, in his experience where sleepers require renewal money is found, because NR puts safety first.

100.     In Mr Jarvis's view the low demand in CP6 is explained almost entirely by the pandemic and its after-effects; government funding sustained the railways during the pandemic, but efficiencies were demanded thereafter and expenditure has had to be cut back. But it must revive, because it is driven by the lifespan of a sleeper, which he reckoned to be about 30 years and because sleepers are safety-critical components, the renewal of sleepers is unavoidable. NR will have to catch up from its current low levels.

101.     In his first witness statement, written in June 2024 before NR's requirement figures were available, Mr Jarvis predicted that once the stockpile has been exhausted demand will rise and revert to its usual average of 750,000 per annum. In his second witness statement he referred to the planning statement supporting NR's application for planning permission for the factory at Bescot, where it was said that NR needed "up to 1,000,000 sleepers each year" and to Mr Curtis' witness statement where it was said that the average annual demand was 750,000. When that statement was written NR had produced figures for its requirement up to 2019/20 but not beyond, and Mr Jarvis predicted a minimum requirement of 500,000 for each of the latter years of CP6. He noted that the Strategic Business Plan for CP7 indicated a 20% drop in expenditure on track renewal in CP7 and therefore predicted an average of 400,000 per annum. But he insisted that because sleepers are a safety critical product demand would then have to rise "substantially" into CP8 and beyond to offset the current underspend; he predicted a "bow wave" effect - by which he meant a spike in demand - in the latter part of CP 7 and/or into CP8 and CP9. He therefore expected that the average sleeper requirement for "CP6, CP7, CP8 and beyond" would be "in the order of 600,000 to 650,000" per annum. He pointed out that this was a conservative forecast and assumed a sleeper lifespan of 42 years.

102.     Essentially Mr Jarvis' view rested on two convictions: that sleepers have a lifespan of about 30 years, and that when they need replacing money will be found to do so.

103.     In support of his views about lifespan Mr Jarvis referred to a presentation given by Mr Darren Sharp, a Principal Engineer for Network rail, in December 2023 entitled "Decarbonising and Sustaining Track". The material available comprises PowerPoint slides rather than the text of the presentation; and the subject matter was, obviously, strategies for reducing carbon emissions by means for example of recycling ballast. One of the slides set out a "Life Cycle Assessment" for sleepers, and included a table one of whose columns set out "Average Life Expectancy (Track System)" for different types of sleepers. Figures ranged from 30 years for concrete sleepers without USPs, to 40 years for recycled plastic composite sleepers; but it was impossible to be clear whether the figures given were for when the sleeper wears out, or for when the track as a whole wears out. Sleepers are more durable than either ballast or rails, and so if the figures refer to the life expectancy of the track then they tell us nothing about the lifespan of the sleeper. We were therefore not assisted by the presentation.

104.     Mr Jarvis was asked for his comments on Mr Heubeck's view that sleepers can last up to 60 years, and said he thought that could not be right because the industry is still "working to" a lifespan of 30 years. But it is difficult to know what that means; Mr Jarvis agreed that replacement of sleepers is driven by their condition, not by their age.

The authority's case about future requirement

105.     For the Secretary of State, Mr Heubeck said that Mr Jarvis's estimate of an average demand of 750,000 sleepers per annum was "only broadly correct" during the years of high demand and that the "true average" was lower. Turning to the future, for him the factors set out in the SAFI (paragraph 98 above) mark a long-term change such that the requirement for sleepers will not return to the levels seen in the 2010s. The track renewal trains replaced significantly more sleepers than actually needed to be replaced. The high level of work done between 2002 to 2018 put the system in very good shape and sleeper lifespan is now much longer than 30 years. Improved detection systems, assessing the condition of sleepers remotely, mean that targeted renewal can be done much more accurately than it was in the days when assessment was done by visual inspection. Resources, on the other hand, are tight and look set to remain so.

106.     In support of his position about the lifespan of a sleeper, Mr Heubeck relied upon three documents. One was a table produced in 2008 by Richard Spoors Associates Limited, and the second a study by that company in 2009/10 which examined sleepers at track renewal sites in order to assess whether track was being renewed prematurely. He claimed that it showed that sleepers with pre-Pandrol clips could achieve a lifespan of 40 years. Ms Clutten in cross-examination put it to him that most of the sleepers in the table would have had Pandrol clips, on the basis of their age, and he agreed. The third document was a study by Purbrick and Cope in 1981 which predicted that sleepers with Pandrol clips could achieve a life of 50 years.

107.     For Mr Heubeck, a major factor that will drive down rates of renewal of sleepers is climate change. Increased frequency of extreme weather leads to floods, to track and embankment being washed away, and to prioritisation of safety-critical work other than the renewal of sleepers.

108.     For his predictions of future requirement Mr Heubeck took the current demand as the baseline, but out of caution took it to be 350,000 rather than the actual 307,000. He predicted that the requirement may be less in CP7 (as did Mr Jarvis, and as does NR's Strategic Business Plan for CP7), but that there would be modest growth in CP8 and stronger growth in CP9, thus recognising that some recovery is going to be needed. In an endeavour to put figures to those words, he considered a number of scenarios: no growth, low growth (10% to 20% in CP8), medium growth (30 - 40% in CP8) and high growth (50% or more in CP8), and concluded that, on the basis of the funding likely to be available, low growth would be the most likely and the other scenarios all unlikely. He quantified low growth as 350,000 per annum in CP7, 70,000 more per annum (up to a maximum of 420,000) in CP8, and another 70,000 in CP9 up to a maximum of 490,000 (thus adding 20% of 350,000 in CP8 and again in CP9).

 

Discussion and conclusion about future requirement

109.     Neither expert has attempted to forecast NR's requirement for sleepers beyond CP9 and nor do we. We say more about the extent to which the future can be predicted under Issue 2 below; for now, our findings are confined to the period up to and including 2039 - in other words, the end of CP9 in 15 years' time. The experts' views about NR's requirement over those periods are some distance apart, as described above.

110.     We begin by saying that in the absence of a proper study of sleeper lifespan, working from an appropriate sample, it is not possible to determine the normal lifespan of a sleeper. None of the studies referred to by the sleepers experts set out to be such a study. We are also unable to determine matters such as the number of additional sleepers a track renewal train would replace, compared to conventional methods, since all we have is the assertions of the two experts. As we indicated above, the Darren Sharp report relied on by Mr Jarvis was impossible to interpret. As to the papers relied on by Mr Heubeck, we were not assisted by the Richard Spoors report (paragraph 106 above) because it related to what it described as a "relatively small sample" of sites (43) where renewal work was being carried out. It is not possible to make any deductions about the lifespan of sleepers from a small sample of sleepers already identified as requiring renewal. The Purbrick and Cope paper recorded expectations for the lifespan of sleepers with Pandrol clips, but did not and could not (at that date) test the reality.

111.     We do however accept Mr Heubeck's view that climate change has given rise to new track maintenance challenges for NR, and that budget constraints are limiting the renewal of sleepers at present and are likely to continue to do so; both propositions chime with commonsense and one's general knowledge of the physical and economic climate.

112.     It may be helpful if we repeat our findings about NR's requirement to date:

2017/18: 400,000

2018/19: 450,000

2019/20: 440,000

2020/21: 409,122

2021/22: 496,503

2022/23:  470,199

2023/24:  391,714

2024/25:  307,309

113.     The average requirement across CP6 (2019-2024) was about 440,000 sleepers per annum; and we think that it is useful to look at averages across CPs so as to avoid the distortion arising from individual years of particularly high or low demand.

114.     Both experts took a view about future trends, described those trends and then put a figure on them. For Mr Jarvis, demand was going to revert to its usual levels, and while he regarded the long-term average as 750,000 he adopted a cautious approach and put it at between 600,000 and 650,000; at the hearing Mr Ormondroyd took this to mean 625,000 and Mr Jarvis stood by that figure.

115.     His approach was, as we said, based on a conviction that sleeper replacement is driven by sleeper lifespan, and that when a sleeper is worn out money will always be found to replace it. He was not prepared to accept that climate change might require that higher priority be given to other safety-critical work, nor that it might prove simply impossible to replace all sleepers so that difficult decisions might have to be made instead to close lines or impose speed restrictions.

116.     When we turn to the numbers Mr Jarvis attached to his predictions we find that they are too far adrift from reality. He forecast a minimum demand of 500,000 per annum in CP6, which can be seen to be too high. His average of 400,000 across CP7 looks high on the basis of the first year of that period (307,000). Requirement could catch up, but to generate an average of 625,000 across "CP6, CP7, CP8 and beyond", taking an average of 440,000 for CP6 and even accepting Mr Jarvis's average of 400,000 for CP7, an average of 1,035,000 per annum would be needed in CP8. To generate that average over CP6, CP7, CP8 and CP9 an average of 830,000 per annum would be needed across CP8 and CP9, which Mr Jarvis himself accepted in cross-examination was not realistic.

117.     So we cannot accept Mr Jarvis's predictions about future volume.

118.     We have no difficulty in agreeing with Mr Heubeck's description of future demand as reduced or static in CP7, followed by modest growth in CP8 and stronger growth in CP9. We say that because, as we said above, we agree with what he says about climate change and about restricted resources, and - importantly - because his description is consistent with the Strategic Business Plan for CP7, which refers to

"... acceptance that we will not return to CP6 levels of asset performance in the near future, due to the cost and deliverability challenges but this can be mitigated to increase maintenance activity, as well as improve management through technological advancements...

Maintaining CP7 exit levels of asset performance in the long run will be challenging from a funding and deliverability perspective and likely prohibitively disruptive to our customers, but we also recognise that it will be important to take action to redress asset condition in CP8 and beyond. We therefore consider that it would take two control periods to achieve this, CP8 and CP9."

119.     The difficulty we have with Mr Heubeck's evidence is that his translation of that description into numbers is as pessimistic as Mr Jarvis's figures were optimistic. Despite the average requirement for CP6 being 440,000, he starts from 350,000 which is lower than every year for which we have a figure except 2024/25. His maximum of 420,000 in CP8 means that even in 2034 the requirement is below the average for CP6. And Mr Heubeck's maximum of 490,000 in CP9 means that it will be 10 to 15 years from now before requirement returns to the level seen only three years ago in 2021/22. These predictions are implausible.

120.     So we cannot accept Mr Heubeck's predictions either. Nevertheless we accept his general characterisation of the coming years. NR's Strategic Business Plan predicts a reduction of 20% in track maintenance in CP7, followed by growth across CP8 and CP9, and for that to be meaningful in terms of catching up on renewals left over from CP6 and CP7 it must represent a higher average demand than that seen in CP6 and CP7. But it is unlikely to see a full return to the levels seen between 2011 and 2017 (see table 1 above, paragraph 55).

121.     If we take that 20% and apply it to the average requirement in CP6, that generates an average requirement of just over 350,000 per annum in CP7, which is not out of kilter with the 307,000 or so required in 2024/25. We find that the best estimate that can be made of NR's future requirement, expressed as averages across Control Periods so as to represent modest growth but real growth over CP8 and stronger growth in CP9, is:

CP7 350,000

CP8 450,000

CP9 600,000

 

122.     We take the view that it is not possible for us to make findings about levels of requirement in individual years because of the inevitable ups and downs in demand; as we said above, an average is the best we can do. But the accountants will need to use figures for individual years for discounting purposes. In the table below we have set out figures for their use which produce the averages which we have found are the best estimate for CP7, CP8 and CP9 (beginning with a rounded figure for 2024/25 for convenience):

	24/25	310,000
	25/26	330,000
CP7	26/27	350,000
	27/28	370,000
	28/29	390,000
	29/30	410,000
	30/31	430,000
CP8	31/32	450,000
	32/33	470,000
	33/34	490,000
	34/35	520,000
	35/36	560,000
CP9	36/37	600,000
	37/38	640,000
	38/39	680,000

 

Issue 2: the duration of the claimant's business in the real world and the no scheme world

NR's quest for a new sleeper factory at an LDC

123.      It is common ground that NR has not abandoned its ambition to secure a site for a sleeper factory at or close to an LDC - which it achieved at Doncaster but failed to do so at Bescot.  It remains NR's policy in the real world to have sleepers manufactured at an LDC - not just because of the very unsatisfactory nature of the site at Rochester (being too small, and in the wrong place) but because in any event it wishes to avoid the trip costs, and the potential lack of flexibility, associated with sites that are not at LDCs. For those reasons NR would pursue the same policy in the no scheme world despite the continued availability, in that world, of the site at WWH. If NR is eventually successful in identifying a new site at an LDC, the experts agree that it would then take approximately four years before sleeper production can begin. 

124.      The claimant's case is that, despite NR's policy, this is not going to happen. All NR's efforts to identify a new LDC site came to nothing, because none of the LDCs is suitable for the installation of a sleeper factory, for reasons of space, or for environmental or planning reasons. There is no evidence to show that an LDC site is going to be found during CP7 - CP9 (April 2024 to March 2039) or indeed ever.  Despite NR's efforts, Bescot was the only suitable location identified, and that failed.

125.      The Secretary of State's position is that a suitable site could be found, in the real world, by 2028.   Mr Heubeck was asked to comment on the likelihood of a site being found at Carlisle Kingmoor, Crewe, Toton or Westbury, and he maintained that all those sites would be suitable. The basis on which he said so was unconvincing; for example, he dismissed environmental concerns about a site at Carlisle on the basis of internet searches, but he accepted in cross-examination that he had no expertise in planning or ecology and was not qualified to comment on environmental matters. Indeed he appeared to have very little understanding of the environmental issues at Carlisle. He also disagreed with Mr Jarvis' emphasis on the importance of finding a new site on the west of the country, maintaining that the west can easily be supplied from the east and arguing that new rail links could be provided where necessary.

126.      Overall, Mr Heubeck's solutions to the problem of locating a sleeper factory at an LDC were simplistic; if it were as easy as he said, NR would have been able to find a suitable LDC, either as an additional possibility back in 2016 in the P2 procurement exercise or since then. It has not done so. In the planning statement supporting the application for Bescot NR explained why the other sites were unsuitable; and while we can see that that planning statement might have downplayed the other sites we accept that the reasons given were genuine rather than merely pessimistic - as is borne out by the fact that NR today is no nearer to finding a new site at an LDC.  

127.      We therefore reject Mr Heubeck's view that an LDC site will be found by 2028 in the real world.

128.      That is not to say that a site will never be found, given the obvious advantages of being at or near an LDC and given the fact that the need for one will become pressing both in the real world and in the no scheme world. In the real world, requirement is going to rise and put pressure on Rochester in a few years' time, and NR will want to secure its supply before that happens. In the no scheme world, the site at WWH is getting old (we say more about that below). We accept Mr Jarvis' view that it will need to be in a position where it can serve the west side of the country more readily than can be managed from Doncaster. What has changed in the last few years, and even since the refusal of planning permission for Bescot, is the reduction in NR's requirement for sleepers. We have found that that reduction was not just the result of the pandemic; we have accepted Mr Heubeck's view that systemic changes mean that although the requirement will rise, it is not going to reach, in the foreseeable future, the levels seen ten years ago.

129.      As Mr Ormondroyd submitted, a change in size requirement might well yield more positive results in the future. A new plant is not going to need the capacity looked for in the P2 exercise. Nor will it need the long sidings required in the past for loading track renewal trains (see paragraph 92 above). A new plant might be a carousel plant, which needs less space than does the long line method. We find it more likely than not that NR will eventually locate a suitable site at or near an LDC.  

The No Scheme World

130.      As we have said, it is common ground that in the no scheme world there would have been no stockpile, which in the real world was made in contemplation of the disappearance of WWH. In the no scheme world, by 2017 after the outcome of the P2 procurement exercise, there was a threat to supply arising from the absence of planning permission for the new Bescot factory, but there was no need for a stockpile because WWH could continue to supply sleepers until Bescot came on stream (if indeed it did). The sleepers experts in their SAFI said:

"4.47 ... in a No Scheme World, stockpiling would not have been required and would not have happened, because NR could have taken actions to maintain supply from the Washwood Heath factory.  

 

4.48 Such a continuation of supply would have been achieved by the award of two

contracts. Firstly, a short-term contract lasting only a few months from the end of the extended 2012 Contract until a new medium-term contract was in place. Secondly, just as in reality, NR would have prepared and issued an Invitation to Tender, ("ITT") in 2019, which would have sought offers for supply for a medium-term contract. This second contract would have commenced no earlier than April 2020 and most likely from June 2020 for a likely term of up to 8 years in duration via either a 4 year contract with the option to extend for an additional 4 years on a 1+1+1+1 basis or via a 5 year contract with the option to extend for an additional 3 years on a 1+1+1 basis. The ITT would have described NR's supply requirements, in the same way as the P3 contract described in the Scheme World.

 

4.49 Given the availability and capabilities of the WWH factory and its workforce, it is a reasonable conclusion that Cemex would have bid for and been awarded both the short term contract and the subsequent medium-term contract. As a result, there should have been no gap in supply."

 

131.      So it is agreed that in the no scheme world the 2012 contract would have been extended as it was in reality, that there would have been a P3 ITT in 2019, and that the claimant would have bid successfully in that exercise. They agreed that it was unlikely that anyone else would have bid for that contract; after the failure of the planning application for Bescot potential competitor bidders would have had little confidence in NR.

132.      Accordingly the claimant would have continued to produce sleepers at WWH. We have to consider in due course (issue 4 below) what would have been the terms of that contract and what would have been the claimant's market share; first, we have to consider how long the claimant's business at WWH would have continued.

133.      It agreed that WWH had a 600,000 sleeper production capacity, with a proven track record.  It was better located than Rochester, close to the WCML.  According to Mr Jarvis NR considered it as the 'optimum' location for supplying sleepers on the west side of the country. The claimant's case is that there is no reason to foresee an end to the business at WWH in the no scheme world; it had the capacity, the personnel, the track record, the location (good if not ideal) and there is no reason to foresee its replacement by a different manufacturer or any move to a different location despite NR's policy of co-location at LDCs. In effect WWH was, so far as location was concerned, good enough and its advantages outweighed any disadvantages.

134.      The weakness in that case is that WWH would not have lasted for ever.  It was built, as we said, in 1989. Mr Neil accepted that by 2032 the factory would have been 'really quite old', and that work would be required to renew or replace it.  Mr Jarvis accepted that renewal work could not be put off forever if WWH were to continue to supply sleepers and pass the necessary safety requirements.  Renewal would be costly, that cost would be reflected in the price of sleepers, and that cost would in turn have an effect upon the claimant's competitiveness in future procurement exercises following the end of the current contract.

135.      The authority's case, and the view expressed by Mr Heubeck, was that operations would cease at WWH by 2036 because the plant there would by then have become too old and another site would by then have been found.

136.      Counsel for Cemex argued that Mr Heubeck conceded, in cross-examination, that if the Tribunal were to find that no alternative location at an LDC  was available to replace WWH, and if we found that there was no reason to suppose that NR would wish to relocate production to a site that was not at an LDC, then there would be no basis for assuming that supply of sleepers from WWH would end. What Mr Heubeck said, when that view was put to him, was "I believe that's correct, if I follow your logic"; but he did not concede either that no alternative LDC location was available or that NR would not wish to relocate to a non-LDC location.

137.      In any event, we have found (paragraph 129 above) that eventually NR would have found an alternative LDC location, and therefore Mr Heubeck's concession, if such it was, does not take effect. We find that the claimant's case, that WWH would continue to operate in perpetuity, is unrealistic. We accept the alternative view put forward by the authority that a new site at an LDC would have been found by 2036.

138.      Beyond that date the state of the claimant's business cannot be predicted. There would be procurement exercises and the claimant would probably bid. There is no evidence about what its chances of success would be and no reason to suppose that the claimant would have any better or worse chance of success in bidding for future contracts than any of its competitors, which would of course be drawn from a worldwide field and have access - as would the claimant - to constantly developing technology. There would also be new materials in play; for example, Mr Heubeck drew to our attention research programmes developing sleepers made out of recycled plastic, aimed at reducing carbon emissions. We cannot say anything about how successful such endeavours will be. The future is unknown because the technical landscape is changing so fast.

139.      In closing Mr Ormondroyd described the decision we have to make here as being whether the accountants should model cash flows into perpetuity or to 2036. In our judgment it is impossible to make any findings about the claimant's business in the no scheme world beyond 2036 and therefore the accountants should model to that date. 

The real world

140.      It is uncontroversial that Rochester is an inferior location to WWH, with a smaller site, challenging access conditions, and in a relatively remote part of the country in relation to the WCML.  It has a de facto annual capacity of 200,000 sleepers.  It is also common ground that NR would be alive to the risk of only having one supplier. It will plan ahead to avoid failure at times of peak demand, and therefore will need to find an alternative site in the future.  

141.      The experts agreed that the Rochester contract would probably be extended by NR to its maximum extent, to 2030. They also agreed that Rochester could potentially continue to supply sleepers to 2036, although that was dependent on the level of NR's demand. Certainly it could carry on beyond that date at the levels of requirement predicted by Mr Heubeck; we have found that NR's requirement will be rather larger than he predicted, but will not have reached 600,000 per annum by 2036.

142.      The claimant argued that even in the absence of an LDC location, an alternative will 'simply have to be found' at a non-LDC location, which while challenging should be somewhat easier than identifying a location at an LDC.   Mr Jarvis thought that after 2030 NR, who he thought are pre-occupied with the re-tender of the TWM contract (due to expire in 2026), would grant one further contract of two years, again with four one-year extensions sufficient to allow NR to secure a replacement site better located in relation to the WCML, taking operations at Rochester to 2036.  His oral evidence was slightly vague, but we understand from closing submissions that the claimant's case is that relocation would have to occur by 2032.    

143.      In his written evidence, Mr Heubeck thought that it would be necessary to relocate from Rochester by early- to mid-CP9, around 2036.  He accepted that there would be benefits of relocating earlier to avoid unreliability of supply, on his case by 2034.

144.      We accept that a new site will have to be found because the useful life of the Rochester plant is limited by its capacity. However, standing as we do in early 2025 when there is no sign of NR yet taking steps to find a new site, and when it is about to be preoccupied by a new procurement exercise at Doncaster, we find it difficult to suppose that a new site could replace Rochester by 2032. Even in the real world, prediction into the 2030s is very difficult. Both experts at different stages in the proceedings favoured 2036 as the date by which sleeper production will move away from Rochester, and that is our finding.  

145.      There will therefore be a procurement exercise and the claimant will no doubt wish to tender. The authority's case is that it will have a competitive advantage in that exercise because it can relocate the carousel plant from Rochester, and so is likely to win that contract. Mr Heubeck thought it eminently possible; and he pointed out that it was believed that the Bescot P2 bid had been won by RailOne on the basis that it was going to use a carousel plant relocated from Iowa.

146.      Carousel plant is sold on the basis that it can be relocated because it does not require the same degree of foundations required to keep long cables under tension as needed by the long line system.  The claimant said that despite this, the Rochester plant would not be moved. It has been specially adapted to fit the awkward Rochester site. The claimant called Mr Lawrence, who gave expert evidence that while it was physically possible to relocate elements of the carousel plant, it would not be economic to do so as the costs of dismantling, transporting, reassembling and then commissioning would easily exceed the installed value of the plant. 

147.      Mr Lawrence's expertise is in insolvency; he assesses for the benefit of third parties whether it is economic to sell plant. But that is not relevant here. We accept that the plant at Rochester could not economically be re-sold, but the question is a different one; could its owner move it for use elsewhere, and would that be cheaper for the owner than buying new? Mr Lawrence accepted that on that basis up to 50% of the plant at Rochester could be relocated and re-used by the claimant.

148.      Nevertheless, we have little hesitation in rejecting the notion that in reality this possibility would give the claimant a competitive advantage.  Mr Heubeck's evidence to that effect carried no weight; he is not an expert in plant and machinery, and his information about the RailOne relocation was, as he put it, little more than 'informed gossip' in the industry. Relocation would be expensive; and the plant at Rochester is specially adapted to the awkward site and might be difficult to fit elsewhere. There is no basis on which we can make a finding that the claimant would be more likely than any other competitor to win the competition for a contract at the replacement site - particularly in light of the uncertainties we identified at paragraph 138 above.

149.      Accordingly in the real world by 2036 the claimant is in a very similar position to that which we described in the no scheme world; its future cannot be predicted, and it has no more chance of winning a contract and continuing to operate than any of its competitors.

Conclusions about the real world

150.      We have made findings about the volume of NR's past and future requirement in the real world. We have determined that we cannot make any finding that the claimant is likely to continue in the business of producing concrete sleepers in Great Britain beyond 2036 and for the purposes of this decision we therefore need not go beyond that date. It is agreed that the claimant's market share to that date will be 30%.

151.      The rest of this decision relates to the no-scheme world.

Issue 3: the terms of the extension contracts from April 2017 to April 2020

152.     We explained above that in the no scheme world as in the real world the 2012 WWH contract would have been extended as it was in reality (paragraphs 30 and 41), and further short-term contractual arrangements made, to bridge the gap between 31 March 2017 when the fixed term of the 2012 contract expired and the beginning of the P3 contract in early summer 2020.

153.     We have to make decisions that will enable the accountants to express their opinions about what the claimant's profit or loss would have been in that period in the no scheme world. We have made findings about the volume of sleepers required by NR for work on the track during this period; it is agreed that that requirement would have been the same in the no scheme world. Ms Clutten in closing suggested that in the no scheme world perhaps more work might have been done on the track during this period, without the distraction of the stockpile and of the imminent closure of WWH; we reject that suggestion, first because it is speculative and second because it goes against the agreed basis on which we have looked at requirement.

154.     So the volume of NR's requirement is now a given. How many of those sleepers would the claimant have supplied, and at what price?

155.     Mr Humphries KC in closing argued that the contract extensions in the real world remained subject to the MGV in the 2012 contract, and would equally have been subject to it in the no scheme world; aside from that the claimant's position was that we should leave it to the accountants to fill in the detail about price and market share during this period. Ms Fowler has expressed her own views as to the price the claimant would have charged, and the market share it would have achieved.

156.     For the authority it was argued that price during this period would have been the same as in the real world; Mr Smith, the forensic accountant, has expressed his own views about the volume the claimant would have sold.

157.     In our view these are not issues for the accountants. The accountants are not experts in the supply of sleepers; price and market share are for the sleepers experts. Our approach is therefore to use what we have learnt from the evidence we have heard so far in order to make findings about price and market share in this short interlude between the end of March 2017 and early summer 2020.

Market share and MGV

158.     We look at market share in two stages. We address first the claimant's argument that the MGV in the 2012 contract (of 250,000 per annum) applied to the extensions to that contract and to the short-term contracts that followed it, and that it would have applied likewise in the no scheme world. This is a question primarily of construction of the 2012 contract and the extension and short-term contracts.

159.     The MGV in the 2012 contract was expressed simply as follows: "Network Rail shall purchase a minimum of 250,000 sleepers per annum". That term is set out along with prices, and the price banding terms we explained above at paragraph 21.

160.     As to the agreements made for the period beyond 31 March 20197we have copies only of two documents. The first is the second one-year extension, dated 23 May 2018; under the heading "Description of change" it said:

"1. Extension to the Cemex contract for a further 12 months to end 31st March 2019.

2. Reduction in scope of supply during this period and price increase requested by Cemex...."

161.     The document went on to set out in more detail the changes that it effected, and referred to NR's need to continue stockpiling and its requirement of "further volume from Cemex than originally anticipated." It is not clear what that means. On the following page is a table of prices and estimated volumes which appears to say that the volume required was estimated to be 128,900 sleepers, well below the MGV; in light of that, the reference to "further volume" earlier in the document does not appear to be to an increase in production, but rather to a continuation of production where none had been anticipated. In fact, we know from Cemex's sales data (Table 2 above at paragraph 63) that the claimant sold 222,881 sleepers to NR in 2018/19, well above 128,000 but still below the MGV.

162.     The other copy agreement we have is a two-page document headed "Contract Variation Proforma", dated 18 January 2019. It states:

"There are 3 parts to this variation:

1.      Extension to the Cemex contract to manufacture sleepers for 10 weeks from 1/4/19 to 14/6/19.

2.      Sleeper pricing, profile and volumes from 1/4/19 to 14/6/19.

3.      Pricing for sleepers produced and/or dispatched from 01/02/19 to 31/03/19."

163.     The document goes on to set out prices and payment terms, and then states that a total of 68,160 sleepers will be produced during the 10 weeks following 1 April 2019 (equivalent to about 340,000 over a year).

164.     What we can see from these two documents is that volume and pricing was agreed afresh on each occasion. On ordinary principles of contract law we see no reason why the MGV in the 2012 contract should have applied automatically to the extensions or the short standalone contracts. The MGV is not expressed in the 2012 contract to be something the extensions have to include; and we can see from the two documents we have that the parties were free to make a new deal about price and volume. These are short-term agreements, entered into if and only if NR so required; it would make no sense for its choice to be restricted by an MGV agreed five years earlier. Just as the extension contract was an opportunity to renegotiate the price (as Mr Jarvis agreed, and as we can see did happen) so it was an opportunity to renegotiate volume (and we can see that sales fell below that volume in 2018/19 without any suggestion of any penalty for NR). We find that the extension contracts, and the short-term contracts that followed the extensions, were not subject to the MGV.

165.     That means that in the no scheme world, as in the real world, the claimant's share of the reduced market in sleepers during this period was for NR to determine, one extension or contract at a time, and it would have depended partly on the location of the work that needed to be done to the track but also, crucially, on price.

Price in the short-term contracts

166.     The 2012 contract included provisions for the prices to be indexed. Those provisions did not carry over into the extensions in the real world; instead, prices under the extensions and the following short-term contracts were significantly higher than they would have been under the 2012 contract formula. The claimant had to maintain profitability in a very uncertain period and the higher prices enabled it to do so; we imagine that for NR a short period of high prices was acceptable if it meant that the supply was secure - particularly in view of the need to build up a stockpile.

167.     In the no scheme world that latter motivation was absent, so NR was in a stronger position. Moreover, NR was buying only its requirement for track because it was not stockpiling, so overall sales would have been considerably lower (by 820,000 sleepers or so, see paragraph 79 above) than it was in the real world. In those circumstances the claimant had to bear in mind the competition from TWM, which in the no scheme world could in fact have provided most of NR's requirement - an outcome the claimant would have been keen to avoid. Accordingly we find that the prices in the short-term contracts in the no scheme world would not have been the inflated prices charged in the real world. Nor would they have been as low as they were in the claimant's P2 bids for supply from WWH, because the claimant was not in competition with a range of international bidders as it was in the P2 procurement exercise.

168.     Pricing was complicated because there were so many different types of sleeper and we have insufficient information to construct a set of prices higher than the P2 bid price and lower than the prices in the short-term contracts in the real world. But in between those two sets of prices lie the prices the claimant was charging under the 2012 contract at the end of its fixed term in March 2017. They could be further updated by means of the formula set out in the contract, which depended upon the price of raw materials as well as inflation. We find that that is what would have been done; the 2016 prices would have been further updated in each contract until the P3 contract began. We expect the accountants to be able to agree the updating in order to produce the necessary range of prices for their calculations.

Market share  

169.     In light of that decision about price, we return to the question of market share. We remind ourselves (from paragraph 112 above) that NR's requirement from 1 April 2017 to 31 March 2020 was:

     2017/18: 400,000

     2018/19: 450,000

     2019/20: 440,000

 

170.     How many of those sleepers would the claimant have sold? One of the questions addressed to NR by the parties in September 2024 was "In the period 2012 to 2020, how did NR decide from which supplier to purchase sleepers?". The answer was:

"The decision as to which supplier to purchase sleepers from under their respective contracts was largely driven by volume constraints of supply from the respective suppliers.  TWM could supply a maximum of up to around 380,000 to 400,000 G44 or EG47 sleepers per year.  The other specialised sleepers plus the balance of G44 and EG47 sleeper demand was produced by CMX.  The average requirement for NR each year was approximately 700,000 to 900,000, declining later in the period, and with the balance of what could not be produced by TWM met by the stockpile."

 

171.     For the authority Mr Ormondroyd put it to Mr Jarvis that that meant that NR took as many sleepers as possible from TWM and then turned to the claimant for the balance; in other words, NR bought from TWM in preference to the claimant. Mr Jarvis disagreed; that was not how orders were allocated between the two suppliers during the time when he was responsible for ordering. Instead, orders for sleepers were placed so as to get the best price for NR bearing in mind the price banding in the two contracts (and we saw how he put that in paragraph 21 above). This is a further instance of information provided by NR which is not evidence, and where the maker of the statement cannot be asked for an explanation. We accept Mr Jarvis' evidence that orders were allocated between the two suppliers so as to get the best price, and that there was no preference for TWM.

172.      One of the factors in that decision in the no scheme world (where NR is buying only its requirement and not stockpiling) is the MGV in the TWM contract. Of the requirement listed at paragraph 169 above 200,000 sleepers in each year had to be bought from TWM. Mr Jarvis expressed the view that WWH would have commanded a 60% market share throughout its operation, but that cannot be right during these years when TWM's MGV gave it nearly 50% of the requirement.

173.     The prices that we have found would have been payable under the extensions and the short-term contracts would have been neither as uncompetitive as the prices in the real world extensions and short-term contracts, nor so low as in the P2 bid. Mr Heubeck's evidence was that TWM's operations were generally more efficient and its pricing generally lower than the claimant's but we have no specific information about TWM's prices and we do not know how far its prices rose when levels dropped to near the MGV. Doing the best that we can we find that in the no scheme world, from 1 April 2017 to the time when the P3 contract began at WWH, the claimant would have had approximately a 50% share of the market.

174.     The accountants now have the volume of NR's requirement in the no scheme world in the period from 1 April 2017 to the beginning of the P3 contract, the range of prices the claimant would have charged, and the market share it would have had.

Issue 4: the terms of the P3 contract in the no scheme world

175.     We now turn to the P3 contract which it is agreed would have been made in the no scheme world for continued production at WWH. We look separately at price, at whether there would have been an MGV, and at the market share the claimant would have achieved under this contract. In light of our findings on duration (Issue 2, above) we work on the basis that production would have continued at WWH beyond the fixed term and extensions of the P3 contract, but we have been offered no evidence about what the terms of future agreements might be and we expect the accountants to make their projections up to 2036 on the basis of our findings about the P3 contract.

Price in the P3 contract in the no scheme world

176.     As we noted above, pricing is complicated because there are so many different types of sleeper; supply contracts listed up to about a dozen different sleepers and prices. In addition there were banding provisions which varied the price depending upon the quantity ordered. In considering the possibilities for pricing in the P3 contract in the no scheme world we have to look at pricing in the real world and decide which of a number of different sets of prices would most likely have been replicated in the no scheme world; it is not practicable to adopt a more fine-grained approach by looking at specific types of sleepers, and no witness has constructed a set of prices specifically for the no scheme world.

177.     No-one suggests that the prices charged by the claimant in the real world between 2017 and 2020 would have been replicated in the P3 contract in the no scheme world; they were inflated by the unusual circumstances caused by the scheme. Ms Fowler, the claimant's forensic accountancy expert, in her witness statements has expressed the view that the pricing would have been calculated on the basis of what was charged in the last year of the fixed-term of the 2012 contract, further adjusted for inflation. We have not heard her evidence but, as we said above, pricing is for the sleepers experts, and neither of them espoused this view. The claimant's case at the outset of the hearing appeared to be as set out in Mr Jarvis' witness statement, namely that the P3 prices would have been the prices offered in the claimant's non-compliant P2 bid for production at WWH (see paragraph 26 above), adjusted for inflation. They were lower than the claimant was charging under the 2012 contract at the time, and lower than those in the claimant's compliant P2 bid for supply from Bescot (which were not low enough to win the contract); we do not know how the prices in the non-compliant bid compared with those offered by the successful bidder, RailOne.

178.     In the real world the claimant did submit a bid for a P3 contract at WWH, as we saw above (paragraph 33) in a short letter setting out prices for three types of sleeper only, at what was clearly intended to be an attractively low level. It said:

"This assumes that HS2 does not serve a CPO order on Cemex to vacate [its] existing site at Washwood Heath and Cemex can continue to manufacture and supply sleepers to Network Rail from it."

179.     The compensating authority characterised this as a no scheme world bid because it is made on the assumption that the scheme is cancelled; Mr Heubeck expressed "a great deal of confidence" that this would have been the pricing agreed in the P3 contract at WWH in the no scheme world.

180.     At the hearing the claimant's position appeared to change. Mr Neil, as a witness of fact, did not express any opinion in his witness statements about what would have been the pricing in the P3 contract in the no scheme world, although he provided some information about the P2 WWH bid prices. In re-examination Mr Humphries KC asked Mr Neil whether he would have offered those P2 prices for the P3 contract in the no scheme world, given that in that world WWH was not under threat of compulsory acquisition, and he said he would not. Mr Jarvis a week later, giving evidence in chief, said that in light of Mr Neil's evidence the P2 bid pricing was "perhaps not the right place to start in terms of pricing". In cross examination on that point he said "I've kind of only changed my view based on what Stuart said last week."

181.     That left the claimant's position on pricing rather uncertain; Mr Jarvis had withdrawn his view, while Mr Neil's view was given in re-examination and could not therefore be challenged. Moreover it was a negative view, rejecting the P2 pricing but leaving open the question what the price would actually be. The only positive evidence at that point was Mr Heubeck's opinion that the prices in the P3 contract in the no scheme world would replicate those offered in the one-page letter bid for the P3 contract at WWH in the real world.

182.     We have confidence in Mr Neil's evidence of fact, but he did not give evidence as an expert witness. In his second witness statement he said that the P2 price offered by the claimant in its non-compliant bid for supply from WWH was influenced by the scheme, and that it offered discounted rates in the hope that it would "commence a dialogue" with NR in the hope of fending off the compulsory purchase. He did not say in so many words that he would not have offered that price in the P3 process in the no scheme world; but when asked a direct question about that in re-examination he said he would not.

183.     That question did not relate to a matter that had been the subject of cross-examination. If the strict rules of evidence applied in the Tribunal Mr Neil's answer would be inadmissible but those rules do not apply (rule 16(2) of the Tribunal Procedure (Upper Tribunal) (Lands Chamber) Rules 2010). Nevertheless we have decided to disregard it because it would be unfair to rely on it. Had he been an expert witness he would have given his views on the P3 contract as evidence in chief and could have been cross-examined about it; to place any weight on what he said about the P2 prices would be unfair to the compensating authority.

184.     We also disregard Mr Jarvis' change of mind. It was prompted, on his own account, not by Mr Neil's witness statement but by what Mr Neil said at the hearing; if it is unfair for us to have regard to one it is equally unfair to have regard to the other.

185.     Accordingly we cannot disregard what Mr Jarvis said in his witness statement about the P2 pricing. In any event, pricing at that level makes sense. It was a bid designed to be attractive, albeit not as dramatically discounted as the P3 letter. It was a set of prices offered for production at WWH and therefore at a level at which the claimant must have expected to make a profit. It was lower than the prices then being charged under the 2012 contract, and much lower than the inflated prices charged during 2017-2020. And while the claimant in 2019, bidding for the P3 contract in the no scheme world, was "the only ticket in town" as Ms Clutten put it (because WWH was the obvious and the only way to bridge the gap that NR could see coming up) the claimant still had to be competitive. Its competitor was TWM, not for the contract but for orders, since orders of sleepers were dictated (as we have seen) by pricing and price banding.

186.     The P3 letter, by contrast, can hardly be described as a bid. It was a last-ditch attempt to fend off closure, made by a committed employee of the company doing his best but without board approval; its pricing was so low that it smacked of desperation; it omitted all the detail that would be required in a bid. It could certainly not be characterised as a bid made on the assumption of a no-scheme world; it assumed that WWH could remain in operation but there is no assumption that HS2 itself was cancelled.

187.     Accordingly we reject Mr Heubeck's view that the prices in the P3 letter would have been replicated in the P3 contract in the no scheme world.

188.     The only other candidate for pricing in the P3 contract in the no scheme world is the rates charged under the 2012 contract, as updated under that contract until 2016/17 and then further updated for inflation. Neither of the sleepers experts supported it and in our view those prices would have been too high to give the claimant the substantial market share that it no doubt hoped for.

189.     Bearing all the above in mind we find that the prices in the P3 contract in the no scheme world would have been the prices offered in the non-compliant P2 bid for a contract at WWH, adjusted for inflation.

Would there have been an MGV in the P3 contract in the no scheme

190.      It was the claimant's case that there would have been an MGV in the P3 contract in the no scheme world, of 250,000 sleepers per annum, as there had been in the 2012 contract. Mr Jarvis emphasised the mutually beneficial nature of such a guarantee, enabling the manufacturer to spread the cost of their overheads and thus to keep the price down. He was adamant about this, and in response to cross-examination asked rhetorically; "why wouldn't I have given an MGV?" He pointed out that in the no scheme world the failure of the Bescot planning application would have been a major blow to NR, as it was in the real world; he spoke of the despondent "mood in the camp" and said that in the no scheme world he would have said to the Contracts and Procurement team "just get me a Cemex contract at Washwood Heath". The easiest way to do that, he argued, would have been to roll over the terms of the existing contract including the MGV. He suggested that arrangements might have been made for a "single tender action" to facilitate that.

 

191.     Mr Heubeck on the other hand thought that there would have been no MGV. Low demand from 2019 going forwards would have meant that either no MGV was offered or one at a low level of 100,000 per annum.

192.     We accept that if the P3 ITT had been issued with a view to securing the second supplier in the established duopoly (as were the ITTs for P1 and P2) NR would have offered an MGV. And we accept that Mr Jarvis would have regarded an MGV as a routine provision in such a contract. It is significant that there was an MGV in both the 2012 contract at WWH and TWM's 2011 contract, and that an MGV of 200,000 per annum was offered in the P1 and P2 ITTs. It was a wholly normal term.  Of all the contracts we have been told about, only the current Rochester contract does not have one, but we were told and we accept that there were two reasons for this. One was that the contract was intended as a stop-gap, to provide for a dwindling supply while Bescot was built and started production - the ITT having been issued in September 2019 when the Bescot planning application was still live. The other was that NR provided £6.2 million for capital costs at Rochester (funded by HS2), so that the claimant was really not in a position to ask for an MGV as well.

193.     But the difficulty for the claimant on this point is that the sleepers experts have said, in carefully agreed terms in their SAFI, that in the no scheme world:

"just as in reality, NR would have prepared and issued an Invitation to Tender, ("ITT") in 2019 ... The ITT would have described NR's supply requirements, in the same way as the P3 contract described in the Scheme World." (emphasis added)

 

194.     In light of that wording we take that agreement to mean that the ITT would have been issued at the same time and on the same terms as in the real world, before the refusal of the Bescot planning application and on the basis that the contract was going to be for an interim supply to tide NR over until Bescot came into production. Accordingly no MGV was offered in the ITT.

195.     Despite that agreement, at the hearing there was some argument as to whether the ITT for the P3 contract would have been issued later in the no scheme world. The claimant argued that it would have been delayed, because with WWH still operating there would not have been the urgency there was in the real world and NR would have been able to wait until it knew the outcome of the Bescot application. Once it knew that the P3 contract would secure its second supplier rather than providing a stop-gap it would have issued the ITT and offered an MGV. We think that most unlikely; in the no scheme world in September 2019 there would have been much the same urgency because the outcome of the Bescot planning application was in doubt, and even if permission was granted Bescot had still to be developed and start production Importantly, that argument runs counter to the sleepers experts' agreed evidence on this point.

196.     So that argument is unconvincing. Equally unconvincing is Mr Jarvis' suggestion at the hearing that there would have been a "single tender action" (which he admitted would have needed to go through a rigorous approvals process) in order to simply roll over the Cemex contract. In the SAFI the sleepers experts agreed that there would have been a procurement process just as in the real world, and there is no room in the context of that agreement for a wholly different process. And such a process would not have been necessary; by the time the Bescot planning permission was refused the claimant had already made its bid in response to the P3 invitation, with a price and terms that did not involve rolling over the existing contract. The Contracts and Procurements team would have been able to proceed in response to that tender rather than initiating a new and difficult single tender process.

197.     We have not heard evidence about the rules of the procurement process which would enable us to say whether it would have been open to NR to offer an MGV, contrary to the terms of the ITT, in the early months of 2020 in the no scheme world when it was clear both that Cemex' bid was successful and that the Bescot project was dead in the water. But even if that would have been possible we fail to see why it would have done so. As Mr Jarvis explained, NR's motive for offering an MGV was to keep prices down. But pricing was already fixed in the bid submitted in November 2019. Moreover, in the early months of 2020 the pandemic was gathering momentum; whilst the final P3 arrangements were being made it would have been obvious to everyone that the coming months or more were going to be very difficult indeed for the railways as for all other industries; it is difficult to see a reason for NR to offer an MGV to Cemex in those circumstances.

198.     Accordingly we find that the P3 contract in the no scheme world would not have contained an MGV.

Market share during the P3 contract in the no scheme world

199.     As to the market share that the claimant would achieve in the P3 contract in the no scheme world, the parties' positions mirror each other. Mr Jarvis predicted that the claimant would secure 60% of the market during the P3 contract and beyond, and TWM 40%; he said that on the basis of WWH's good record and of its position with access to the WCML, and because it produced far more specialist sleepers than did TWM; overall Mr Jarvis clearly regarded the claimant as the better supplier. Mr Heubeck on the other hand took the view that the claimant would secure only 40% of the market with TWM as the cheaper, more efficient and better-positioned supplier selling 60% of the sleepers required. He regarded the cost of trip trains taking sleepers from WWH to Bescot as an important reason for that outcome.

 

200.     We are unable to accept either expert's prediction, for three reasons.

201.     The first is that it is not possible to find that either of the suppliers in the duopoly was in general in any sense better, cheaper or more efficient. We take seriously Mr Jarvis' enthusiastic commendation of the claimant as NR's trusted supplier, always flexible and able to provide any kind of sleeper however specialist and its go-to adviser for any kind of sleeper problem. But inevitably by 2030 or so in the no scheme world things have moved on, personnel will have changed, and the claimant's plant at WWH is getting old. Cemex's market share approached 60% in the real world at a time of high demand, but it has not achieved that level of market share since TWM has been at full production. Overall requirement is now, and will be in all the years now being considered, far lower than it was when Cemex had that share. Even the trip cost, which was so important a factor in Mr Heubeck's assessment, is unknown when we look years ahead. Furthermore, by 2026 a new contract at Doncaster would be in place, probably but not inevitably with an MGV, probably with prices designed (whether by TWM or another manufacturer) to be competitive against the claimant's prices. So the years ahead are hard to predict and for that reason alone neither expert's view is wholly persuasive.

202.     The second reason why neither can be entirely correct relates to the period from April 2020 to date and arises from what we have found about levels of requirement and about the absence of an MGV in the P3 contract at WWH. In order to explain that, it may be helpful if we repeat our findings for NR's sleeper requirement in the relevant period and beyond, setting out with each finding the figures for 40%, 50% and 60% of the total figure. For the future, beyond 31 March 2025, our findings take the form of averages across each five year period rather than of an annual requirement:

 

 

 

 

 

 

 

Table 3: requirements and proportions

 

		40%	50%	60%
2020/21	409,122	163,648	204,561	245,473
2021/22	496,503	198,601	248,251	297,901
2022/23	470,199	162,879	203,599	244,319
2023/24:	391,714	156,685	195,857	235,028
2024/25:	307,309	122.923	153,654	184,385
CP7 (2024-29	350,000 (average)	140,000	175,000	210,000
CP8 (2029-34)	450,000 (average)	180,000	225,000	270,000
CP9 (2034-39)	600,000 (average)	240,000	300,000	360,000

 

203.     In those first five years to 31 March 2025, in which the requirement is known rather than being a finding by the Tribunal, we can see that in this period there is no possibility of the claimant securing 60% of the market because in each year TWM's MGV guarantees it over 40% of the requirement. In 2023/24 the MGV accounts for more than 50%, and in 2024/25 for more than 60% so that Cemex could not even have achieved 40%.

204.     In the remaining years until the end of 2036 (in accordance with our findings on duration we go no further) we have made findings only about average requirements across Control Periods. During those remaining years, if there is still an MGV in the Doncaster contract and if it still stands at 200,000, then the market share the claimant can achieve does not reach 50% until (probably and approximately) 2030. However, whilst we think it likely that there will be an MGV in the Doncaster contract from 2026 onwards we think that in view of current levels of requirement it would be set at less than 200,000 per annum in the no scheme world. It would be more likely to be 150,000 or 100,000; on that basis either supplier might achieve 60% of the market at any stage from now on; but, as we said above, we have insufficient information about either supplier to say that either would be likely to do so.

205.     The third reason why neither expert appears to us to be correct arises from the fact that we are looking at a duopoly in a situation where NR's annual requirement is, and will be for some time, much lower than it has been in the past. NR has only two suppliers. The experts agreed that NR would not want to create a situation in which it had only one supplier, and we find that it would not be in NR's interests to favour either supplier in the future in the no scheme world where requirements are so low that both suppliers may be operating near the margins of profitability. We remind ourselves of the price banding system; purchasing just the MGV from one supplier would be more expensive per sleeper than purchasing rather more than the MGV - but how much more expensive that would be, and how many more would need to be purchased to achieve a lower price band, is unknown, as is the MGV at Doncaster after 2025. Even if the price banding dictated that in one particular year one supplier got the majority of the orders and the other could not pay its rent, we do not think that NR, taking the long view about the need both to ensure supply and to preserve a market (however limited) would let that happen. It would be in NR's interests to hold a balance between the two suppliers even if that meant that the cheapest option was not always taken.

206.     Doing the best we can with the information available we find that the claimant cannot show that it would have achieved more than an equal market share in the P3 contract in the no scheme world, and that the compensating authority cannot show that it would have achieved less than 50% overall (even though as we observed above that would be the outcome in two identifiable years). We find that the claimant's market share in the duopoly on the balance of probabilities would be 50%, subject to one quite tricky point.

207.     That tricky point is the authority's argument that once the claimant's lease of Area B comes to an end in 2026 it will have insufficient storage capacity to cope with a demand of more than 196,000 sleepers per annum. To that argument we now turn.

The Area B problem

208.     It will be recalled that Cemex's lease of area B was contracted out of the security of tenure provisions of the Landlord and Tenant Act 1954 and accordingly at the end of the lease in March 2025, Cemex did not have a statutory right of renewal (paragraph 16 above). Area B was used primarily for the storage of sleepers, but Cemex had also erected some buildings on it, including a testing laboratory.  It was common ground that the capacity of the site at WWH to produce sleepers (and therefore the profits that could be generated) would, in part, be affected by the number of sleepers that could be stored at the site. It will be recalled that this was because sleepers could not be stored at LDCs and NR therefore required its suppliers to have storage capacity to cope with times of high demand (paragraph 19 above). It was a term of the claimant's 2012 contract that it keep a minimum 20,000 sleepers in stock.

209.     The Secretary of State submitted that since the lease of Area B was contracted out of the 1954 Act, the law dictates that compensation must be assessed on the twin assumptions that (1) the tenancy would have determined on the earliest date of termination that could have resulted under a landlord's notice given on the date of entry (24 March 2025) and (2) that no account can be taken of any possibility of renewal that would in fact have existed in the absence of the acquisition: Bishopsgate Space Management Ltd v London Underground Ltd [2004] 2 EGLR 175 (a decision of the Tribunal).  Any assessment of the profitability of the business absent the acquisition must therefore take account of the fact that, after 24 March 2025, it would have had to operate from a site not including Area B.      

210.     In its statement of case the claimant argued that owing to the contractual relationships between Network Rail, GB Railfreight Ltd (which was Cemex's direct landlord of Area B), and Cemex, it should be assumed for the purpose of assessing compensation that the sub-lease would have been renewed.  Subsequently, in their skeleton argument, counsel for Cemex submitted that there was no reason to assume that Cemex would be required to give up possession when the lease expired, and that we should assume that it would be able to continue occupation of Area B beyond 2025 'whether pursuant to a renewed lease or by licence' (our emphasis). In their closing argument, that narrowed, with Cemex's final position being that it would have remained in occupation of Area B on licence, and that could be assumed to be the case without any inconsistency with the decision in Bishopsgate.

211.     The Secretary of State's position also developed as the hearing proceeded, originally pleading that owing to the assumed loss of Area B, Cemex's ability to store sleepers, and thus its overall production capacity, would have been reduced such that its market share of Network Rail's annual requirement (which Mr Heubeck put at an average of 490,000 per annum in CP9) would fall from 40% to 35%.  In his evidence in chief, Mr Heubeck explained that he had changed his mind and now accepted that Cemex would be able to accommodate sufficient sleepers to maintain a 40% share, on the Secretary of State's projected demand figures. In cross-examination he went further and conceded that it might be able to accommodate 60% of 490,000 (294,000) sleepers, but there might be difficulties at peak demand

212.     Cemex maintains that if it lost Area B it could have relocated the stored sleepers to the remaining part of the yard. Mr Neil gave evidence about the storage capacity at WWH; it will be recalled (paragraph 10 above) that he was from 2000 the claimant's Operations Manager at WWH and later its Business Manager for Rail.  He explained that in around 2002, Cemex increased its sleeper storage capacity by increasing stack height by 66% (from 12/14 layers per stack to 20/22 layers - in each case for G44/EG47 sleepers), and laying a hard surface on area B and, later, in area C.  This increased the storage capacity of the site to around 135,000 sleepers - a notional figure based on Cemex being asked to store three months' production, assuming no sleepers left the site.  In fact, he said, they had never been asked to store more than 120,000 at any time.  In oral evidence, he suggested that Cemex could probably have stored more than that figure. 

213.     Mr Ormondroyd pointed out various mathematical discrepancies, and tested the evidence on access, width between stacks of sleepers, and so on, but Mr Neil was unshakeable in his belief that even were area B to be lost, he could still have accommodated around 94,000 sleepers by moving some buildings and operations if necessary; it would be tight, and would have to be managed, but there was sufficient space to store the stock. 

214.     While Mr Heubeck was familiar with WWH, having visited it half a dozen times between 2004 and 2009, he accepted that he did not have Mr Neil's detailed knowledge of the site.  Mr Heubeck's calculation of available storage was based upon looking at a 2012 aerial photograph with the aid of a magnifying glass in good quality light; he calculated a storage capacity of areas A and C (i.e. without area B) of between 55 and 60,000 sleepers, assuming that the stacks on the photograph were at their full height. We do not accept that Mr Heubeck's evidence of the storage capacity at WWH, based on the examination of a photograph from some years ago with a magnifying glass, can have any claim to accuracy.

215.     Mr Neil's job was to store sleepers on the site, he knew it far better than Mr Heubeck did, and it would have been his responsibility to manage the available space had area B been lost. We have no difficulty in accepting his evidence that without Area B the site at WWH could have stored up to 94,000 sleepers (rejecting Mr Heubeck's 55,000 to 60,000). Mr Heubeck's evidence was that with 55 to 60,000 sleepers able to be stored the claimant could have coped with a demand of 294,000 per annum albeit with some difficulty at peak times. We can safely find that with a 94,000 storage capacity the claimant would have no difficulty supplying considerably more than that. Mr Heubeck also agreed that the Trackwork Moll site might provide a useful indicator of storage capacity; the Doncaster site had a production capacity of 400,000 sleepers a year, and an effective stockyard capacity of 70,000 sleepers.   Using the same ratios, a yard with a capacity to store 94,000 sleepers should be consistent with a production capacity of 540,000 sleepers, well above any quantity that might be required of it on our findings as to NR's annual requirement.  We therefore do not need to say any more about Bishopsgate.

216.     Therefore our finding about market share, at paragraph 206 above, is unaffected by the Area B problem.

Conclusion

217.     We now pass the matter back to the parties so that their forensic accountants can work on the findings we have made. We have directed that they exchange reports and also that they file an agreed statement setting out where they differ and why.

218.     As we said at the beginning, this judgment has told a story that lacks its final chapter. We hope that, equipped with the story as we have told it and with the help of the forensic accountants, the parties can write the final chapter together without further assistance from us; but in case that is not possible a hearing has been listed in September 2025 for the accountants to give their evidence on the basis of our findings in this interim decision.

 

Upper Tribunal Judge Elizabeth Cooke                        Mr Peter D McCrea OBE FRICS FCIArb

 

                                                               2 May 2025

 

Right of appeal 

 

Any party has a right of appeal to the Court of Appeal on any point of law arising from this decision.  The right of appeal may be exercised only with permission. An application for permission to appeal to the Court of Appeal must be sent or delivered to the Tribunal so that it is received within 1 month after the date on which the Tribunal's subsequent decision, disposing of all the issues in the reference, is sent to the parties (unless an application for costs is made within 14 days of that subsequent decision being sent to the parties, in which case an application for permission to appeal must be made within 1 month of the date on which the Tribunal's decision on costs is sent to the parties).  An application for permission to appeal must identify the decision of the Tribunal to which it relates, identify the alleged error or errors of law in the decision, and state the result the party making the application is seeking.  If the Tribunal refuses permission to appeal a further application may then be made to the Court of Appeal for permission.