Select Committee on Defence Fourth Report

WARSHIP BUILDING STRATEGIES

8. In April 2001, the MoD commissioned consultants RAND Europe[11] to investigate potential procurement strategies for the MoD's warship programme over the next 10-15 years.[12] The study's impetus came from difficulties encountered in negotiating the contract for the Type-45 destroyer, and many of its conclusions were directed towards mapping out that programme's future direction.

Type-45 procurement

9. The previous Committee produced a report on the Type-45 destroyer programme (and its associated PAAMS anti-air missile) in 1999.[13] It highlighted the delays, cost overruns and collaboration difficulties of its predecessor programme—the Common New Generation Frigate—which led the UK to withdraw from it in April 1999 and launch its own national Type-45 destroyer programme. The MoD had appointed Marconi Electronic Systems (subsequently BAE Systems Electronics) as the Type-45 prime contractor in November 1999, without competition because the firm had previously been involved in the aborted collaborative programme.[14] The MoD's plan, announced in July 2000, was for two principal sub-contractors to build the first three of what could be "up to" 12 vessels in the programme—BAE Systems Marine to build two, and Vosper Thornycroft one. The intention was to allow both BAE Systems Marine and Vosper Thornycroft to be able to compete for the assembly of further Type-45s, with ship modules, or 'blocks', to be able to be supplied by these and other shipyards.

10. In late 2000, however, one of the subcontracting shipyard firms, BAE Systems Marine, submitted an unsolicited proposal to build all 12 Type-45 destroyers.[15] The MoD "could see that the Type-45 procurement strategy was not delivering results,"[16] and commissioned the RAND study to help inform its consideration of the new proposals. Sir Robert Walmsley told us that at December 2000, when the MoD placed the prime contract to build the first three destroyers, the ship's design immaturity was such that the prime contractor could not place sufficiently specific sub-contracts with either Vosper Thornycroft or BAE Systems Marine. The MoD had been looking for a risk-sharing agreement between the prime contractor and each of these two shipyards that would require the shipyards to take a risk that work would be added subsequently to the specification. As the Chief of Defence Procurement put it, "we were unable to persuade the prime contract office and the shipbuilders to share that risk."[17]

11. While the MoD subsequently rejected the unsolicited BAE Systems Marine proposals, the RAND study allowed the Department to accept a new joint BAE Systems Marine/Vosper Thornycroft proposal for BAE Systems Marine to assemble six batch-one vessels from 'blocks' built by both firms. The revised contract, for six vessels, was placed with BAE Systems Electronics (the prime contractor) in February 2002.[18] The RAND study also helped the MoD evaluate construction options for the Future Carrier (paragraph 25).

12. RAND took as its starting point (or 'reference case') the option of one shipyard producing all of the Type-45s, taking advantage of the cost efficiencies possible from 'learning' over a long production run, and then evaluated the possible additional costs or savings should the MoD instead follow one of four alternative approaches—

(a)  dividing (whole) ships between two shipyards, through competitions for each batch of three vessels;
(b)  keeping the then already planned division of work between BAE Systems and Vosper Thornycroft yards for the first three vessels, and then dividing subsequent ships between them through competition for each subsequent batch of three vessels (this essentially mirrored the procurement approach then already being developed by the MoD);
(c)  dividing the (whole) ships in the entire class between two shipyards, according to the MoD's direction and without competition beyond the initial decision point; and
(d)  dividing not just ship assembly but also the construction of major 'blocks' between two shipyards, by MoD direction and without competition beyond the initial decision point.

13. These four options for dividing the work would add to the production costs that could be achieved by a single shipyard, taking advantage of production efficiency savings that are possible when it is decided at the outset to place all of the work with one shipyard. According to RAND's calculations, the extra costs of dividing the work were 10-13% in the case of a whole-ship division between two yards (options a, b and c above), and about 5% for a division of ship 'blocks' between two yards (option d above). These extra costs would reflect the way that Type-45 construction overheads at the shipyards would be spread over fewer ships in each location when the work was divided, and Astute submarine construction overheads spread over fewer ships at the BAE Systems Marine yard at Barrow. The study then examined the likelihood that savings from competition at intervals (as in options a and b) could more than offset these extra costs. There is little reliable evidence on the savings possible from warship competition, and RAND's analysis gave a wide range of figures. Data from 31 ship and missile programmes suggested savings from competition averaging only 7%,[19] with some programmes producing competition savings of up to 40% on the one hand, and others with extra costs of 30% on the other.

14. RAND concluded from their analysis that "historical competitive programme data suggest that there is approximately a 50:50 chance that competition, if it can be sustained over the programme as a whole, will result in equal or lower costs than a sole source. Therefore, there is no dominant answer to whether competition or sole source would be likely to lead to lower costs".[20] The Chief of Defence Procurement put that in context against past MoD procurement policy—

If you compare that with the robustness of [RAND's] conclusions about the Joint Strike Fighter, where they were absolutely against the introduction of competition, I found this slightly reassuring because we have persuaded ministers that a competitive strategy was good, partly to not put all your eggs in one basket and partly we thought it would probably lead us to lower costs. Here were RAND validating that initial strategy, not saying it was obviously right, but certainly not saying it was wrong.[21]

15. Informed by the RAND analysis, the MoD decided to allocate the Type-45's six 'blocks' between the two shipyards, with the same blocks for each ship to be allocated to specific shipyards. Vosper Thornycroft will build virtually all the ship forward of the bridge (two blocks), as well as both masts, the funnel and the upper works. The BAE Systems Marine Barrow shipyard will build the engine rooms (two blocks) and, after ship-one, will assemble the ships. And BAE Systems Marine's Clyde yard will make the stern and the operations room (each a separate block), as well as assembling ship-one.[22]

16. The unsolicited proposal from BAE Systems Marine was rejected because of a number of factors. First, BAE Systems had attached conditions about other equipment programmes in order to generate the proposed cost savings, and this would have entailed, we were told, "signing up to quite a major shift in our policy, which was to distribute warship building for competition reasons."[23] We noted, nevertheless, that with the 'block' allocation strategy dividing all of the work at the outset, it provides, according to RAND, "no opportunity for continuing competition throughout the production programme, no basis for expecting any reduction from [initially agreed] costs, and minimal incentive for innovation. Therefore, this [block allocation] strategy must be evaluated on criteria other than that of [a comparison of the costs of other options]".[24] The consultants concluded that—

The option of having specified shipyards build the entire production run of some blocks should lead to costs that are slightly higher than the sole-source option but lower than the estimated cost of competitive dual sources for the entire ship, assuming little or no cost reduction through competitive pressures. The cost of a directed buy of blocks from two shipyards falls within the range deemed affordable by the MoD, retains the advantages of sustaining two shipyards, and is not critically dependent on subsequent, and uncertain, reductions of cost through competitive pressures.[25]

17. Second, the MoD told us that under BAE Systems' unsolicited bid a large component of the Type-45 programme would have been undertaken at Barrow, resulting in congestion in that shipyard between modules of Type-45 destroyers and Astute submarines in the Devonshire Dock Hall: "Either Astute or Type-45 would not have to have much of a hiccup for the whole place to come to a grinding gridlock."[26] We noted, however, that the RAND study also foresaw such risks with the 'block allocation' strategy that was finally settled upon,[27] and they cautioned that under the MoD's new plan "scheduling of the construction and delivery of the blocks must be closely managed; a block that arrives late at the assembly yard [at Barrow] may cause significant delays in not only the Type-45 programme but also in the Astute programme".[28]

18. In adopting the RAND strategy of seeking 'learning curve' efficiencies from allocating ship modules over a long production run, we questioned the MoD about its decision to offer a construction contract for only six vessels rather than for the entire twelve of the class. CDP told us that committing to 12 would have reduced the flexibility of the defence programme. He assured us, however, that the conditions for pricing the seventh and subsequent ships were linked to the productivity secured in the first batch ordered, while retaining flexibility over its equipment fit—

I feel very comfortable with six. I feel very comfortable with the productivity that we will get on ship-seven onwards. I feel very comfortable that the economies of scale that we have on the equipment production by going for six are very sensibly close to those we would get for 12, but it is not as good, I accept that. But we now have the flexibility. We will decide when we order the second six ... what the modification state ... is for the equipment, and whether we want [for example] a different sonar or not. If we had gone for all 12, all that would be locked up now. The last ship comes off in 2014. That is a long time ahead. I do not like predicting the future with that much certainty. However,... we do have a budget for all 12 ships. There is a very sound operational case for all 12 ships.[29]

19. It appears to us that that 'sound case' for all 12 of the anti-air warfare destroyers will be the stronger if our carriers are not to have a dedicated air-defence aircraft (we discuss later in this report the early withdrawal of the air-defence Sea Harrier and the non-air-defence role of its Joint Strike Fighter replacement). In its response to our predecessor's report on the Type-45 programme, the MoD stated that the balance of Type-45s and Future Surface Combatants—set by the SDR at 12 and 20 ships respectively—would be subject to "continuing critical operational analysis".[30] It is our hope that the 'sound case' to which CDP refers now settles the result of that analysis, and that the MoD has accepted that it needs at least 12 destroyers.

Future competition

20. Another factor in rejecting the BAE Systems unsolicited proposal was the value of retaining the potential for warship competition between UK shipyards in the future. The Department considers that the new arrangements for the Type-45 give it a strategy which looked "a good way of keeping two companies in the warship building business."[31] The previous plan of a head-to-head competition between Vosper Thornycroft and BAE Systems Marine for the fourth and subsequent ships had the potential of the loser going out of warship building,[32] as the loser would be faced with building one of the three vessels in each batch at the price offered by the winning bidder.[33] Instead, CDP told us, "if you allow competition to kill off one of two effective companies, then you have no competition for ever. The Type-45 strategy retains an element of competition, a needle if you like, between two yards, each of whom are trying to perform better than the other, without putting them into a head-to-head, winner-takes-all, loser-disappears situation. I think it is quite constructive."[34] The RAND consultants acknowledged that the solution now settled upon by the MoD "helps ensure that both shipbuilders will remain viable and able to compete on future MoD programmes",[35] but were cautious about the nature and extent of that future competitive capability—

For future programmes such as the [Future Surface Combatant] and [the Future Carrier], Vosper Thornycroft's continued presence increases the chance of Vosper Thornycroft being in the market to compete for these programmes. Sole-source and, to some degree, the directed buy of blocks may not allow Vosper Thornycroft to maintain its warship construction capabilities, leading to future programmes facing a monopoly in warship construction ... Keeping Vosper Thornycroft active in building warships will be positive for future MoD programmes. A caveat must be stated for the option of a directed buy of blocks, however: Once that paradigm is chosen, it may be difficult to choose another paradigm for future programmes.[36]

21. The RAND analysis also underlines a wider concern, about the availability of competition for future warship programmes within the UK. Their report noted the risks, had Vosper Thornycroft withdrawn from warship building, of BAE Systems Marine securing a monopoly position[37] (since it is government policy that only UK shipyards should be allowed to build warships). They noted widespread concern within industry about the prime contractor—BAE Systems— for the Type-45 and Astute submarines, and possibly for the Future Carrier, also controlling three shipyards involved in those same programmes.[38] In the short term at least, demand for warship building work would remain low, and as the industry consolidated there would be less impetus for innovation.[39] The RAND study identified a clear market trend over the next 15 years. Demand for blue-collar workers is set to decline slightly between now and 2006, after which demand will increase rapidly with work on the Type-45s and Future Carriers to a peak of nearly 7,000 'direct' workers in 2010—almost double the level in 2005. Consequently, RAND identified a risk for the MoD's future warship building programme in being able to maintain and expand the labour force in the next several years while managing a small dip in demand.[40] In his evidence to us, Sir Robert Walmsley too highlighted the need to retain shipyards with their special warship building skills.[41] He distinguished, however, the need for the MoD to help maintain shipbuilding capacity needed for warship construction — where the yards building the Type-45 destroyers would be sufficient — and for its non-warfighting ships where many other yards were available and provided competition without MoD intervention.[42]

22. Maintaining sufficient warship construction capability in the UK remains essential not just for political reasons. There is no international reciprocity in the shipbuilding industry, and the relatively low design content of warships in terms of their total cost reduces the impetus (compared with aircraft programmes, for example) for international collaboration to share the non-recurring costs.[43] To maintain UK shipbuilding capacity until the expected rise in demand, Sir Robert considered that industry too had a part to play by improving its export performance—

For as long as I can remember, we have been dealing with over-capacity in shipyards. The problem has not been shortage of labour; the problem has been shortage of orders to sustain the shipyards.[44]

We should be doing better in terms of securing warship export business. The Type-45 gives a platform and both companies can look for a certain future from the Ministry of Defence over a long period. Of course there will be some uncertainty about it but the Type-45 should be enough to help launch some export programmes.[45]


If we do not use as a nation the opportunity presented by our warship building programme over the next ten years, in order to present designs that are attractive to other countries, we will only have ourselves to blame for what happens to the warship building industry in the next decade after that.[46]

23. Shrinking demand in a market with surplus capacity appears to be the backdrop for many areas of the defence industry, and perhaps it can be said to be a more intractable problem for the warship construction industry (it also affects the ship repair industry as we shall see later in this report). It is clear to us that in the Type-45 programme the MoD has sought to weave a strategy through often conflicting pressures, not just for minimising costs and maximising competition (when it can be effective), but also balancing what might be best in the long-term as well as the near-term. And, with the complexities of warship procurement, the MoD appears at least to have made reasonable choices, backed by expert advice. There are clearly dangers still in the approach adopted. There is uncertainty about the prospects for genuine competition for future programmes, like the Future Surface Combatant, from shipyards which may only by then have experience of fabricating 'blocks' rather than assembling ships. The MoD will have to determine how to support warship building capacity for the lean short-term to ensure that it is available when needed in nearly a decade's time. And even the smooth passage of ship work at Barrow's Devonshire Dock Hall will have to be monitored carefully.

24. But at least the problems of this industry are starting to be addressed. It is clear to us that the sort of work commissioned from RAND provides a valuable and timely analysis of the state of the warship construction market in the UK. From their examination of the procurement options available to the MoD, it is evident that a clear strategy is now needed for warship procurement. The RAND analysis would be a useful starting point.

The construction of the Future Carrier

25. Like the Type-45 programme, the Future Carrier's development programme has been revised following representations from the bidders (BAE Systems and Thales). In this case, the firms had sought a longer 'assessment' phase to reduce risk in the designs. In response, the MoD introduced a new phase-3 to the assessment studies. The plan now is to down-select to a preferred bidder at the end of a shorter phase-2 in November 2002, with a Demonstration and Manufacture contract to be awarded in early 2004.[47] In the current phase-2 work, the two contractors are working up designs able to take either a 'short take-off and vertical landing' (STOVL) or a conventional 'Carrier Variant' aircraft,[48] pending the selection of the particular variant of Joint Strike Fighter later this year (paragraph 85).

26. There have been recent reports[49] suggesting a possible smaller than previously planned order for the STOVL version from the aircraft's US customers, which could increase the unit cost of a STOVL aircraft for the UK if it chose to adopt that variant (paragraph 86). In terms of the external size and shape of the Future Carrier, however, we were told that there will be little difference between STOVL and Carrier Variant versions of the ship.[50]

27. Nevertheless, there will still be challenges for building a 50,000 tonne carrier in the UK for the first time in half a century.[51] Shipyard capacity is such now that the Future Carrier cannot be built at one location. The Type-45 programme, however, is providing a model for its construction. Shipyards will have to cooperate to build modules in different locations and then assemble them elsewhere. RAND highlighted how the 'hook-up' of the blocks will be a greater challenge with the Future Carrier than with the Type-45,[52] and in that light CDP saw the experience with the Type-45 and the Alternative Landing Ship Logistic (ALSL) as particularly valuable. The ALSL is being constructed on the Clyde to a Swan Hunter design,[53] and that requires co-operation between two yards "who do not instinctively cooperate—this is practising at it."[54] Sir Robert saw in these developments the beginning of a "network" of shipyards that could co-operate to build ships.[55]

28. As we have discussed above, the construction of the Future Carrier and Future Surface Combatant (the successor to the frigate force) will require extra shipyard capacity beyond about 2009. This was mainly a question, CDP told us, of securing sufficient labour capacity in the shipyards at that time: "It will not be a question of building some huge crane in some lonely dry-dock somewhere. There are very few issues relating to facilities, other than modernising."[56] To assemble a modular Future Carrier, CDP identified three likely candidates with an infrastructure large enough for its final assembly — Harland & Wolff in Belfast, Inch Green on the Clyde and Cammell Laird in Birkenhead.[57] We understand that Swan Hunter are also developing a large dry-dock which will be able to accommodate the vessel, and that Rosyth are working on similar plans.

---

11   RAND's consultancy report, The Royal Navy's New-Generation Type-45 Destroyer: Acquisition Options and Implications, has been placed in the Library of the House (HC Deb, 2 July 2002, c233w). Back

12   Ev 103 Back

13   Eighth Report, Session 1998-99, op cit Back

14   Ev 99, para 8 Back

15   Ev 99, para 11 Back

16   Q 2 Back

17   QQ 2, 17 Back

18   Ev 99, para 13 Back

19   Mean of 7%, but a range median of 11% Back

20   RAND Report, op cit, page 52; Q7 Back

21   Q 8 Back

22   RAND Report, op cit, p 64; Q 8 Back

23   Q 17 Back

24   RAND Report, op cit, p 56 Back

25   RAND Report, op cit, p 58 Back

26   Q 17 Back

27   RAND Report, op cit, p 62 Back

28   RAND Report, op cit, p 66 Back

29   QQ 19-21 Back

30   Fourth Special Report, 1999-2000, HC 223, para 16 Back

31   Q 9 Back

32   Q 8 Back

33   RAND Report, op cit, p 47 Back

34   Q 16 Back

35   RAND Report, op cit, pp 64-65 Back

36   RAND Report, op cit, pp 60-61 Back

37   RAND Report, op cit, p 85 Back

38   RAND Report, op cit, p 73 Back

39   RAND Report, op cit, p 75 Back

40   RAND Report, op cit, p 70 Back

41   Q 10 Back

42   QQ 14-15 Back

43   Q 11 Back

44   Q 25 Back

45   Q 15 Back

46   Q 26 Back

47   Ev 92, para 14  Back

48   Ev 106, para 4 Back

49   see Aviation Week, 1 April 2002 Back

50   QQ 37-39 Back

51   The last such carrier, the previous HMS Ark Royal, was completed in 1955. Back

52   Q 32 Back

53   Strictly speaking, a design put together by Swan Hunter and Royal Schelde, its Dutch strategic partner, according to MoD evidence submitted to the Scottish Affairs Committee's inquiry into Employment in Shipbuilding on the Clyde (Fifth Report, Session 2001-02, HC 865). Back

54   Q 28 Back

55   ibid Back

56   Q 24 Back

57   QQ40-43 Back

---

Prepared 10 July 2002