Select Committee on Trade and Industry Fourth Report



53. We have examined the steel rail sector because of the unusual conjunction of an unexpected increase in domestic demand for the product following the Hatfield rail crash and Railtrack's award in May 2000 of a new five year contract which meant that henceforth most of the rail to be laid will be produced abroad, either in Corus' French plant or in Austria or Italy. Rail may not be typical of the various segments of the steel market. There are however some features which do seem to be replicated in other steel markets —

  • UK domestic demand is insufficient to sustain a UK industrial base, with the result that the UK plant is dependent for its survival on exports, which are significantly affected by the sterling exchange rate as well as more complex demand and supply issues;

  • there is overcapacity in Europe and steadily falling prices, although we understand that rail still commands a premium price over steel beam, more so in the European than the US market;

  • penetration of non-UK companies in the UK market has caused loss of market share in the UK for the UK company;


54. Corus operate a rail mill located at Workington on the coast of Cumbria.[124] The plant has been involved in the manufacture of steel rails since 1877. It now employs around 325 people, working on five shifts. It manufactures heavy rail, steel sleepers, light rail and track accessories. The steel is delivered from the Corus works at Teesside. The steel is subjected to various industrial processes before finishing in lengths of up to 40 metres of finished rail. Rails are generally produced in a standard 36 metre length. Some are delivered to the customer in that length. Others are welded into longer "strings" of rail, either at a separate part of the Workington plant, commissioned in the early 1990s, or at the Corus welding plant at Castleton in Lancashire, acquired in 1993.[125]

55. For many years steel rails from Workington were in practice the only source of rails for the UK market. Workington rails and rail products were also exported to other parts of Europe and to the USA. In recent years exports have accounted for over half of the revenue from Workington rail products sales. There has been a generally buoyant world market for rail, including to other parts of the EU, Central and Eastern Europe, North Africa and the Middle East. Rail produced to a sufficiently high quality standard commands a premium price over steel beams, although European prices have fallen steadily over recent years.

Sales to Railtrack

56. Railtrack are the principal UK purchasers of the heavy rail output from Workington.[126] Every few years they enter into a fixed price contract for the supply of rail for track renewal. Railtrack also awards long term contracts for infrastructure maintenance; contractors can source the rail for this purpose from any approved source. Railtrack told us that in future this policy may change towards rail being issued to all contractors by Railtrack. British Steel sales of rail to BR and then Railtrack fell steadily from around 150,000 tonnes a year in the 1960s and 1970s to under 50,000 tonnes in 1993 and 1994. Since then they have begun to rise slowly. The 1995 Railtrack long term contract led to around 90% of rail being supplied by British Steel, the rest coming from Voest Alpine of Austria.

New contract

57. In September 1999 Railtrack invited tenders for the supply of rail for the next five years in 216 metre welded strings with as few factory welds as possible. It also issued an invitation to tender for the design and build of a rail welding depot in the south of England and for the provision of a welding service. The contracts were placed in May 2000. Railtrack are to buy around 95,000 tonnes a year in 2000-01 and 2001-02. Around half is to be purchased from Corus. Most of the rest is to be divided between Voest Alpine of Austria and Lucchini of Italy. The small quantity remaining can be ordered from wherever Railtrack can "optimise volume discounts".[127]

58. The significance of the May 2000 contract for Corus was twofold.

  • its market share of the UK rails market fell from around 90% to around 60/70%, and it lost its monopoly of welding services for rails, with the establishment of a new welding centre in the south of England. It also reduced its prices to be able to compete with imports benefiting from the weak euro;

  • from now on much of the supply of rails to Railtrack would have to be from its French plant. In October 1999 Corus purchased the Sogerail plant in Hayange, north east France, from Usinor for £83 million. This plant was one of those converted in the 1990s to make longer rail. The long rails produced there are to be shipped to Teessport. Rail supplied by Voest Alpine and Lucchini will from this year also be in longer as-rolled lengths.[128]

Following the contract award, the number of shifts at Workington was reduced from 10 to 5, attributed at that time in a Corus Press release to an "export downturn".

Post-Hatfield requirement

59. Following the Hatfield rail crash and the subsequent rail replacement programme, there is now an additional requirement for around 25,000 tonnes of rail, predominantly in short lengths. Railtrack told us that, "due to expediency", the majority of this emergency requirement was likely to be obtained from Corus.[129] The additional volume already allowed for in Railtrack's May 2000 plans is also likely to come from Corus.

Long rail

60. In the 1990s British Steel supplied BR and Railtrack with welded strings of 180 metres, comprising eight 32-metre lengths. These had four "factory" welds carried out at Workington or Castleton. The strings were then welded together on site when laid as track. In the course of the 1990s many of the main West European railway operators moved cautiously but steadily towards seeking longer continuous lengths of welded rail, so as to reduce the number of factory and on-site welds needed. Longer as-rolled lengths are more expensive. On-site welds had, however, been identified as a principal cause of rail breaks.[130] We understand that there is no evidence that factory butt welds used to make up the strings are a significant cause of failure. The weaknesses in on-site welds has however led to a general perception that all welds were a potential cause of weakness.

61. The operators therefore sought both longer stretches of original rolled rail and longer strings of welded rail for laying and subsequent on-site welding. Providing longer strings by increasing the number of rolled steel rail lengths to be welded in specialised welding plant did not present major problems for producers. The constraint arose from the need to adapt railway rolling stock to carry longer strings.[131] Longer rolled lengths, however, required the producers to commit to major investment in conversion of plant.

British Steel response

62. Most of the major West European manufacturers — including Thyssen, Lucchini, Voest Alpine, and Sogerail — made the requisite investment in the course of the decade to enable them to produce longer lengths of rolled rail. British Steel did not make this investment. Instead, they responded to the demand for longer strings by investing in modern welding facilities at Workington and Castleton, and so "ensured a stable and profitable market for the Workington product throughout the 1990s".[132] Railtrack's decision announced in 1999 to seek longer rolled lengths in the new contract lay behind Corus's decision to buy Sogerail, with its long rail capability, thus giving it an in-house long rail producer and also taking a potential competitor out of the market.

63. In its evidence to us, Corus gave several reasons for its decision not to install the capacity at Workington to make long (72 metre) rail[133] —

  • the market for long rail is almost exclusively a continental European one, with " no demand for long rolled rail from the rest of world customer base" . There is already excess capacity in Europe and little chance of export given the size of the product and the extent to which the home markets of long rail users are tied up by their domestic producers. There is little or no premium for long rolled rail over short despite the additional investment required.

  • Most rail demand worldwide is for short lengths; Corus told us that around 40% of demand in Europe was for short rail.[134] Corus hope to supply the rail for the Channel Tunnel Rail link in welded strings of 288 metres made up of eight 36 metre lengths.

  • Installing a 72 metre scheme at Workington "would not increase its market share" and would not meet the threat from 108 metre rolled rail.

  • It was not until 1998 that Railtrack made it clear that they would prefer to use long rail.

  • The costs of the necessary investment was too high given layout constraints, and the low level of investment in the UK rail infrastructure.

Railtrack and long rail

64. Corus told us that Railtrack stated their preference for longer rolled lengths in September 1998.[135] Railtrack suggested in their first memorandum that it "would have preferred longer input lengths from Corus" in the second half of the 1990s.[136] In response to our further queries, Railtrack provided details of internal documentation which suggested that British Rail procurement policy had leaned towards long as-rolled rail from the early 1990s, and that British Steel had been well aware of this preference and seeking ways of meeting it. Railtrack also told us that Corus have offered discounted rail rates if Railtrack continued to use short rail lengths, and noted that Railtrack's other suppliers were offering lengths beyond the 72 metres from the Hayange mill which Corus can offer.[137]

65. Other rail producers in Europe made the decision to invest in long rail production facilities through the 1990s. Corus have referred to the review carried out in 1993-94 of the practicality of adapting the Workington plant to production of longer as-rolled lengths.[138] The option of purchasing a low cost central or eastern European producer also seems to have been considered some years ago. It must have been common knowledge in the 1990s that Railtrack in common with many — although not all — European operators were thinking of moving towards specifying long rolled rail. British Steel and Corus must have been aware of the need either to persuade Railtrack of the virtues of ever better welded rail or to ensure that they could supply rail in the lengths the UK customer wanted. Whether a good thing or not, the decision by Railtrack to seek longer rolled lengths of rail was scarcely a bolt from the blue.

Workington and long rolled rail

66. It is questionable whether British Steel/Corus could conveniently have adapted Workington to produce longer rails had they wished to. Corus suggested that they were constrained from doing so by the costs involved. Sir Brian Moffat even suggested that it was impossible because of the nature of the site.[139] The company instituted an appraisal in 2000 of the capital costs of the investment necessary to produce 72 metre long rolled rail at Workington. They provided us in confidence with the detailed results, which suggest a cost of around £48 million. This mirrors the results of the similar exercise conducted in the mid-1990s.

67. The estimated costs of adapting Workington includes the cost of heat-treating rails, which makes them harder but possibly also more brittle. Railtrack have decided not to use any mill heat treatment rail pending investigations following the Hatfield incident, a practice which Corus describe as "inconsistent with the practice adopted by railways worldwide". Corus told us that in any event the outline plans for a 72 metre scheme at Workington did not include a heat treatment capability.[140]

68. It is of course a matter for judgement whether British Rail in the mid-1990s or Corus now should have made an investment in Workington to equip the plant to produce the longer lengths of rail which its principal customer wanted. It may be that the judgement was made that Railtrack alone would not provide a sufficient customer base for long rail and that exports of long rail to the continental Europe seemed improbable. We are however surprised that —

  • Corus could find £83 million to acquire a French plant, but not £50 million to invest in a UK plant;

  • that it has proved practical to import French-produced long rail through Teessport but would not have proved possible to export long rail from a UK source through a similar route;

  • that Corus should accept that Railtrack should be switching to Italian and Austrian steel, without making a vigorous effort to protect its domestic market in a sector where the logistical difficulties of importing should give the domestic incumbent supplier a natural advantage.

Bearing in mind the likely changes noted in the mid-1990s in the probable demand for long rail, we are surprised at the decision by British Steel not to invest in conversion of the Workington plant to long rolled rail in the 1990s, when it was making substantial profits. The purchase by Corus of the Hayange long rail plant in France in 1999, in response to Railtrack's express requirement for longer lengths of rolled rail, was the consequence of that earlier misjudgement, resulting in the company spending at least £83 million in acquiring a French plant rather than investing £50 million in the Workington facility.

Workington prospects

69. There is continuing demand from Railtrack and from other customers for shorter lengths of rail. In the short term, as a result of the post-Hatfield renewal programme, UK demand for short rail is buoyant. Rail production at Workington has increased substantially as a result.[141] Indeed, Corus told us that the modifications proposed at the Castleton welding plant to enable it to weld 72 metre lengths— modifications which Railtrack told us had been a condition of the new contract— are being delayed so as to enable Corus to meet the demands of the emergency re-railing programme. 72 metre rail from Hayange has even been cut into shorter lengths for welding at Castleton.[142] Sir Brian Moffat referred to " a total reversal of a market through an accident".[143]

70. In the medium to longer term, however, prospects look less bright. Even future UK short rail sales would be liable to competition from other suppliers, all the more as lower levels of production with fixed overheads at Workington will make the plant less competitive. Some other European railways still use shorter rail; but Corus may seek to supply that market from their plant at Hayange in France, which enjoys some logistical advantage for deliveries in some parts of continental Europe. We have been told that Hayange is indeed producing a great deal of short rail for the continental European market. There is also a strong possibility that lower cost suppliers will soon be able to produce rail steel of sufficient quality to take that relatively small proportion of national markets in western Europe which are not in practice the preserve of the domestic producer. Hayange also has "an established position " in the US market and has historically supplied " significant volumes" to North America.[144] There may be a role for Workington as a niche producer of high quality medium-cost producer of short rails and sleepers predominantly for export to continental Europe, including eastern Europe, to North Africa and the Middle East, including Turkey, and to the USA. The opportunities for Workington are slight, unless Corus were to make it plain that its short rail deliveries were to be from there rather than Hayange.

71. It is of course for the owners of Corus to decide if there is a future for Workington. We find it hard to believe that there is the will to make the investment now in long rail facilities, since the company has the Hayange plant. We urge Corus to consider making Workington the company centre of excellence in short rail and sleeper production for the domestic and international market. We also urge Railtrack to give proper weight to the advantages of having a viable domestic rail producer close at hand when considering its future policy on ordering.

123  Ev, p 78, A1 Back

124  For details, see Ev, pp 74-77 Back

125  ibid, para 6 Back

126  For details, see pp 79-80 Back

127  ibid Back

128  Ev, p 78, A2 Back

129  Ev, p 76, (c) Back

130  Eg Ev, p 79, (d) (ii) Back

131  ibid Back

132  ibid, p 76, para 15 Back

133  ibid, paras 7ff Back

134  ibid, p 78, A3 and p 76, para 13 Back

135  Eg Ev, p 76, para 14 Back

136  Ev, p 79, (b) Back

137  Ev, pp 80-1 Back

138  Ev, p 76, para 17 Back

139  Q 190 Back

140  Ev, p 78, A4 Back

141  Ev, p 77, para 22 and p 78, A3 Back

142  ibid, A2 Back

143  Q 186 Back

144  Ev, p 78, A5 Back

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2001
Prepared 14 March 2001