England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE QUEEN'S BENCH DIVISION
(Mr Justice Collins)

		Royal Courts of Justice Strand, London, WC2A 2LL Friday 7 December 2001

LORD JUSTICE SIMON BROWN:

1. On 3 October 2001, nearly five years after British Nuclear Fuels plc (BNFL) first applied for an authorisation, the respondent Secretaries of State decided that the practice of manufacturing mixed oxide fuel (MOX) – which BNFL proposed to undertake at their Sellafield Mox Plant (SMP) – is justified within the meaning of Article 6.1 of Directive 96/29/Euratom. Article 6.1 provides:

"Member States shall ensure that all new classes or types of practice resulting in exposure to ionizing radiation are justified in advance of being first adopted or first approved by their economic, social or other benefits in relation to the health detriment they may cause."

2. The appellants challenge that decision, contending that in evaluating the economic benefits, which the respondents concluded were sufficient to justify the "very minor radiological detriments" which would result from the practice, they failed to take into account the capital costs inherent in the practice.
3. BNFL (who appear in the proceedings as an interested party) intend if possible to take "irreversible implementation steps" on 20 December 2001. Hence the need to decide the challenge urgently. Collins J heard it on 8 and 9 November and dismissed it on 15 November. We heard the appeal on 27 and 28 November, and deliver our judgments at the earliest opportunity as understandably we were invited to do. If in the result these are less comprehensive than otherwise they might have been, we shall hope to be forgiven.
4. With that briefest of introductions let me set out the essential background to this challenge, drawing heavily for the purpose upon Collins J's helpful judgment below.
5. Most fuel for nuclear reactors is made from enriched uranium oxide. During the process some of the uranium is converted into plutonium. In addition, waste products are generated. The waste products in time make the fuel less efficient and so it is sent for reprocessing in the course of which the waste products are removed and the plutonium is reclaimed. It then has to be stored or recycled. The reprocessing is carried on at Sellafield in the Thermal Oxide Reprocessing Plant (THORP) and the plutonium, which belongs to whichever customer has sent the fuel for reprocessing, has either to be stored or returned to the customer.
6. It has been known for some time that nuclear reactors can operate efficiently using a fuel called MOX, which is a mixture of plutonium oxide and uranium oxide. The manufacture of MOX enables the reclaimed plutonium to be recycled. This has the advantage of reducing the amount of stored plutonium and saving the use of fresh uranium so that the environmental hazards of mining new uranium can be reduced. In addition, it avoids the need to transport the plutonium back to the customers or for reprocessing in a third country. MOX fuel in the form of what are known as ceramic pellets is said to be less attractive to terrorists and safer than plutonium (which is transported in the form of plutonium oxide powder).
7. Between 1994 and 1996 BNFL constructed a mixed oxide fuel plant at Sellafield to enable it to manufacture MOX fuel. The plant is known as SMP. In November 1996, it applied to the Environment Agency (EA) for variations to the gaseous and liquid discharge authorisations. EA asked BNFL to provide information specifically relating to SMP and began a process of consultation to decide whether the manufacture of MOX was justified. On 3 November 1998 EA published its proposed decision which was that the operation of SMP to produce MOX fuel was justified, but it referred the matter to the then responsible Secretaries of State. It stated in paragraph 3.3:

"In reaching its proposed decision, the [EA] has not taken any view on the wider policy issues of plutonium management strategy. The [EA] is concerned about these wider policy issues and considers that major developments at Sellafield are national and international matters and that, given the significant political and economic issues, relevant government departments should be involved in considering the [EA's] proposed decision".

8. Paragraph 1.7 of the EA's letter should also be noted:

"The [EA] received the application from BNFL in November 1996, when construction of the MOX plant was virtually completed and after the capital cost (£300 million) had been incurred. It is unsatisfactory that the [EA] has no powers under the Radioactive Substances Act 1993 to require an application to be submitted for a new plant prior to its construction. The time at which an application is received is crucial to the [EA's] involvement in the regulation of new plant. The [EA] is dissatisfied that it was unable to consider the full economic case for the MOX plant. It is seeking a change in the legislation to prevent a similar situation occurring in future."

The EA's concern was that the construction costs had to be disregarded in accordance with standard economic practice in assessing the economic case for SMP because by the time the application was made those costs had been incurred and so were what are known as "sunk costs". It is this disregard which is said to be unlawful so as to vitiate the decision under attack.

9. On 11 June 1999 the Secretaries of State then responsible indicated a provisional conclusion in favour of justification of the full operation of SMP but concluded that they needed further consultations on the economic case for it. At the same time, they decided that uranium commissioning could go ahead but without prejudice to the ultimate decision on the justification for SMP. If that decision was unfavourable, BNFL would have to bear the costs of decommissioning. In the event a further £170 million was expended between December 1997 and July 2001, some of it on the uranium commissioning permitted in 1999.
10. Following the June 1999 permission, further consultations took place, the claimants each being at all material times consulted. In the Spring of 2001 the Secretaries of State decided to instruct independent analysts to evaluate BNFL's updated economic case and A.D. Little (ADL) were accordingly instructed in April 2001. ADL reported in July and on 27 July 2001 a suitably redacted version of their report was published. Friends of the Earth responded in August and on 3 October 2001 the decision now impugned was made by the Secretaries of State. The judicial review challenge was lodged on 5 October 2001.
11. The duty to justify the manufacture of MOX fuel arises under what is now the relevant Council Directive which is 96/29/EURATOM of 13 May 1996. That Directive came into force on 13 May 2000 and governs the decision made by the Secretaries of State. When EA made its report in November 1998, however, the governing Directive was in a somewhat different form and the relevant considerations were not identical. In particular the previous Directive required that a site specific practice be justified whereas the present Directive requires justification of a generic type or class of practice.
12. The Radioactive Substances Act 1993 contains no requirement for justification, but, on the principle set out by the ECJ in Marleasing SA v La Comercial Internacional de Alimentacion SA [1990] ECR 1-4135, a Member State is obliged if possible to apply its legislation in conformity with a relevant Directive. This may mean reading words into legislation or applying it through conforming administrative action. Here it involved the Secretaries of State reaching a view on justification under Article 6 when deciding how to exercise their powers under sec.23 of the 1993 Act to give directions to the EA with regard to authorising MOX production.
13. The 1996 Directive, like its predecessors, is based on the recommendations of the International Commission on Radiological Protection (ICRP) – see particularly recital 6 of the Directive referring to ICRP Recommendation No. 60 (of 1990). The system of protection recommended by the ICRP has three general principles. The first is the justification of a practice, that most relevant to the present challenge. The other two are the optimisation of protection (keeping exposure as low as reasonably achievable) and the specification of individual dose and risk limits. The following paragraphs relate to the first principle:

"112(a) No practice involving exposures to radiation should be adopted unless it produces sufficient benefit to the exposed individuals or to society to offset the radiation detriment it causes.

115. … The Commission recommends that, when practices involving exposure, or potential exposure, to radiation are being considered, the radiation detriment should be explicitly included in the process of choice. The detriment to be considered is not confined to that associated with the radiation – it includes other detriments and the costs of the practice. Often, the radiation detriment will be a small part of the total. The justification of a practice thus goes far beyond the scope of radiological protection. It is for these reasons that the Commission … requires only that the net benefit be positive. To search for the best of all the available options is usually a task beyond the responsibility of radiological protection agencies."

The Directive:

14. Recital 9 of Directive 96/29 reads:

"Whereas the Member States in order to ensure compliance with the basic standards, are required to submit certain practices involving a hazard from ionizing radiation to a system of reporting and prior authorisation or to prohibit certain practices"

15. Article 1 defines practice to mean:

"a human activity that can increase the exposure of individuals to radiation from an artificial source, or from a natural radiation source where natural radionuclides are processed for their radioactive, fissile or fertile properties, except in the case of an emergency exposure."

16. Article 2.1 applies the Directive to inter alia "all practices which involve a risk from ionizing radiation emanating from an artificial source … , namely … the … processing … of radioactive substances". Article 4 deals with authorisation of practices including the operation and decommissioning of any facility of the nuclear fuel cycle.
17. Title IV, covering Articles 6 to 14, is headed: "Justification, optimisation and dose limitation for practices".
18. Article 6 comes under a sub-heading: "General principles". I have already set out Article 6.1. Article 6.2 allows a review of classes or types of practice "whenever new and important evidence about their efficacy or consequences is acquired". Article 6.3 requires in addition that Member States ensure that all exposures are kept as low as possible (optimisation) and that the sum of the doses from all relevant practices shall not exceed the dose limits for those exposed to them.
19. Article 14 reads:

"Exposure of the population as a whole

Each Member State shall take reasonable steps to ensure that the contribution to the exposure of the population as a whole from practices is kept as low as reasonably achievable, economic and social factors being taken into account.

The total of all such contributions shall be regularly assessed."

20. Finally, Article 54:

"This Directive establishes the basic safety standards for the protection of the health of workers and the general public against the dangers arising from ionizing radiation with the aim of their uniform implementation by Member States."

21. It is the reference in Article 6.1 to "classes and types of practice" and the requirement that they be justified in advance of being "first adopted or first approved" which indicate that the Article is concerned with a generic approach to a class or type of practice (i.e., reading in the Article 1 definition, a generic approach to a class or type of human activity) and not with a site specific justification.
22. Before coming to the decision letter there are two further documents to which I should briefly refer. First, the government's White Paper "Review of Radioactive Waste Policy" of July 1995 (at a time, therefore, when the ruling Directive was concerned with site specific justification) which stated:

"60. In the preliminary conclusions of the Review, it was proposed that applicants might be encouraged to apply for an authorisation under RSA 93 at an early stage in a project so that justification could be considered fully, before major capital investment had taken place. If an authorisation was given, the developer could proceed with construction of the plant confident that its subsequent operation would be allowed, providing that extant safety and environmental standards were met. At the same time, justification would have been considered without the need to take account of any substantial sunk costs. …

62. … For major projects, it is expected that developers will make early applications for disposal authorisations. This would be at about the same time that they seek full planning permission for the project, although it would be determined separately. The regulators would then be able to decide on authorisations before major commitments of money and effort had been made. If the regulators are content, the authorisations could be granted containing conditions which if met at specified stages should lead to approval to start operations in due course when the plant is built and commissioned. For some projects – e.g. a power station of the same design as one already built – the design may be well developed at the outset and, if the site has no unusual features, the authorisation may need no further amendment. In other cases, the design may evolve as the project progresses and decisions will be needed about whether to revise or vary the authorisation. … Legislation is not necessary in order to introduce a system of early authorisations under RSA 93. The procedure will be available under existing legislation for any applicants who choose this route. They would not be required to do so, but the greater certainty it could provide will give applicants an incentive to apply early in the process in appropriate cases. … "

23. Secondly, a document entitled "Communication from the Commission concerning the implementation of Directive 96/29", expressing views which are not binding on Member States but are intended to assist them in applying the Directive. With regard to Article 6 it states:

"This Article sets out the basic principles of radiation protection and thus requires Member States to base their procedures on these principles, namely justification, optimisation and dose limitation.

Determination of the justification of any new classes or types of practice is the duty of the Member State. It should take place before the introduction of the class or type of practice and as early as possible to reduce the influence of the already incurred costs in balancing economic and social factors against health detriment. … The newly introduced proviso of Article 6(2) reflects that there might be the need to review the justification of existing classes or types of practices."

The decision letter

24. The decision letter extends to twenty-five pages. Paragraph 22 notes that Article 6 requires "a generic rather than a site specific assessment of justification of ionizing practices" and paragraph 24 states that "the Secretaries of State have therefore considered whether the class or type of practice comprising the manufacture of MOX fuel is justified" (that justification being stated by a footnote to include "the necessary steps to be taken associated with the manufacture of MOX fuel, including commissioning, operation and decommissioning of plants for that purpose.") I should set out paragraphs 25 and 26:

"25. Although the provisions of the 1996 Directive require a generic assessment to be carried out, it should be borne in mind that BNFL is the only manufacturer of MOX fuel in the UK and that this is likely to continue to be the case for the foreseeable future. In addition, BNFL intends to produce MOX fuel only at the SMP. The company has indicated that the MDF [MOX Demonstration Facility] at its site in Sellafield will not be used again as a commercial production facility but instead will carry out development work in support of the SMP.

26. Therefore, in applying the revised justification test to the present case, it is relevant to have regard to the detailed information available to the Secretaries of State about the specific instance of a MOX fuel fabrication plant in respect of which they have access to extensive data, namely the SMP. This information has been used to inform and test the decision made by the Secretaries of State in respect of whether the manufacture of MOX fuel is justified. The Secretaries of State have therefore reviewed the particular benefits and disbenefits which are likely to result from the full operation of the SMP."

25. Having then dealt with the history of the application and the consultation process the decision letter considers "safety and security issues" and concludes that the detriments would be very small and, indeed, that there would be safety and security advantages in re-using separated plutonium since it avoids the risk inherent in its transportation and processing elsewhere. The decision then turns to "economic issues" and this section I must set out at some length although I can omit certain paragraphs:

"71. In applying the generic test for justification explained above, the Secretaries of State consider that an important factor to be taken into account is the economic benefits to be derived from the particular class or type of practice in question: here, the manufacture of MOX fuel.

72. It is clear that there is a significant demand from a range of countries for the manufacture of MOX fuel. MOX manufacturing plants already operate in France and Belgium, and the construction of another in Japan is planned. It appears likely that a MOX manufacturing capacity will be an important adjunct for nuclear fuel processors, such as BNFL, who seek to maintain their market position.

73. The Secretaries of State consider that, as a result, the manufacture of MOX fuel carries economic benefits which are capable of justifying it as a class or type of practice. Overall, the Secretaries of State consider that, given the ability to carry on that type of practice with very minor radiological detriments (see above), the economic benefits are sufficient to justify it.

74. This general conclusion of the Secretaries of State has in part been informed by, and has been tested against, their consideration of the specific prospects of the SMP. They have consulted on BNFL's business case for the SMP and are satisfied that the operation of the SMP will provide significant economic benefits. This supports their conclusion that the class or type of practice comprising the manufacture of MOX fuel is justified on application of the generic test (see above) by the economic benefits it makes it possible to achieve.

75. The main conclusion reached by ADL in its report to the Secretaries of State is that the operation of the SMP would be in the national economic interest, with a net present value ('NPV') of £216M, when compared to the cost of the SMP not going ahead. The consultants calculated the NPV of operating the SMP at £159M, on a conservative basis, as compared to a loss of £58M if the plant did not operate.

76. ADL's assessment took account of the possibility of various adverse events, such as delays to the delivery schedule, loss of some anticipated orders and more extreme events, including a major interruption to the SMP's operations or an early shut down. For reasons of prudence, ADL did not consider unexpected but possible events which could benefit the LSMP, such as an interruption to the competitors' operations.

77. ADL also left out of account in its calculation the benefits which would flow from operation of the SMP from BNFL's other business, including nuclear fuel reprocessing. The Secretaries of State consider that these additional economic benefits, although difficult to quantify with any precision, are also likely to be substantial.

84. For the purposes of the current justification assessment, the Secretaries of State have tested their conclusion as to the justification for the manufacture of MOX as a class or type of practice by reviewing the BNFL business case specific to the SMP, both including and excluding an element of Japanese Magnox-derived plutonium as a source of material to be processed in the SMP. On both bases, the NPV of the SMP is substantially positive and well in excess of £100M. Consistently with their practice in the course of the various consultation exercises, the Secretaries of State do not go into greater detail about the breakdown of the figures in this decision, so as not to prejudice BNFL's commercial opportunities for the SMP.

85. The Secretaries of State accept ADL's advice that BNFL's overseas customers wish to buy BNFL's full target volume of MOX fuel arising from the THORP baseload contracts. The Secretaries of State also accept that, even taking into account some countries' changing policies on reprocessing spent fuel, customers are not attracted by alternative ways of dealing with their spent fuel which has already been reprocessed or committed for reprocessing. Therefore, even allowing for delays, price reductions and other risks and factors considered by ADL, the Secretaries of State accept the consultants' conclusion that operation of the SMP will yield a substantial net economic benefit as compared to its non-operation.

86. In assessing the economic issues and the NPV of operating the SMP, the Secretaries of State consider that it is appropriate and consistent with the justification test in the 1996 Directive to focus on the present circumstances both on the costs of operating the SMP in the future and on the present assessments of available commercial opportunities. The Secretaries of State do not consider it appropriate to deduct BNFL's sunk costs from the consultants' calculation of the NPV of the SMP – which costs were themselves incurred at a time when a different assessment of commercial opportunities may have been made.

87. Further, the Secretaries of State consider that the class or type of practice consisting of the manufacture of MOX fuel will support potentially significant levels of employment. This is illustrated by the assessment in the ADL report that the operation of the SMP is likely to support up to 480 jobs in total in West Cumbria."

The judgment below

26. It is clear from Collins J's judgment that the argument below proceeded on a very different basis from that before us. Paragraph 7 noted that the present Directive requires justification of a generic type or class of practice rather than a site specific practice and continued:

"The claimants say that this makes no difference in the circumstances of this case since the practice has to be carried out at Sellafield, there being no other site or potential operators, and so the question of economic justification inevitably and rightly focused on SMP."

27. The appellants' essential argument was summarised in paragraph 14 as follows:

"It is common ground that the manufacture of MOX fuel is a new type of practice and so covered by Article 6.1. In deciding what are the economic benefits of the new type of practice, it is necessary to include the costs of enabling it to come into being. It cannot be carried out in the abstract. It cannot be right that the timing of the application for authorisation will determine that cost, it being apparently accepted that any cost to be incurred after the application is made is to be included. This would mean that the applicant could improve his chances of establishing economic benefit by waiting until enough has been expended, which will be 'sunk', so that little remains to affect the likely economic benefits resulting from the type of practice. It is not, I should make clear, suggested that BNFL have acted in any way wrongly or in bad faith in this case: this argument is used to emphasise the potentially arbitrary element in the exercise if timing is to determine the amount of cost. Furthermore, the Directive presupposes a uniform implementation in Member States (Article 54) and that is only to be achieved if there is a uniform approach adopted to assessing economic benefit. It is not in the circumstances of radiological protection a case of choosing next steps on what may be described as a corporate finance basis but of acting as an environmental regulator applying an objective test of justification by reference to all its cost."

28. The Secretaries of State's main argument was that Article 6 affords them the widest of discretions as to what benefits should be taken into account and as to how any benefit or detriment should be evaluated; accordingly, it was open to them to take account of the capital costs of the MOX project or not as they saw fit.
29. BNFL's primary argument, apparently adopted at the judge's own suggestion, was that "capital costs should always be left out of account".
30. In paragraph 17 Collins J rejected the Secretaries of State's argument:

"Either sunk costs are to be ignored in accordance with standard economic practice and the Treasury Guidelines or they are to be included or excluded because the Directive on its true construction requires such inclusion or exclusion. The Secretaries of State cannot in my view choose whether or not to include them. They can, however, once the net positive value (NPV) is determined, decide that other factors which have not been taken into account, such as are referred to in [paragraph 77 of the decision letter], can tip the balance in favour of SMP. In addition, they must put social and other benefits in respect of which they have a wide discretion into the balance. All this means, as Lord Lester accepted, that even if he were right and sunk costs should have been included, it would not automatically follow that SMP would not be justified. Apart from anything else, the figure of £470M (which would, it seems, have to be set against the £216M NPV) cannot be taken at face value. Included in the cost of construction will be benefits in the form of employment and other benefits to the local and perhaps national economy. A more detailed exercise will be needed to assess the true disbenefit."

31. The judge finally concluded, in paragraphs 19 and 20, that the Directive on its true construction requires the capital costs of any new practice to be ignored; alternatively, in paragraph 21, that if that approach be wrong, the Secretaries of State were right, or at least entitled, to adopt the classic economic approach to sunk costs. These three paragraphs I should set out in full:

"19. I was initially impressed and even beguiled by Lord Lester's argument. It seems to me that the assessment of economic benefit should not be influenced by the time at which the application or the assessment is made. That would be likely to lead to uncertainty and inconsistency. Naturally, different States may reach different conclusions on the types of practice in question. Indeed, we know that the manufacture of MOX fuel is carried on in Belgium and France. It was carried on in Germany but no longer is. But they should all so far as possible be approaching the question of justification in the same way. However, the answer in my judgment is not that put forward by the claimants but rather that the cost included in setting up a particular plant to enable the type of practice to be put into effect is not to be set against economic benefits. While I have not had to reach a final conclusion whether the generic approach is right, I believe that it probably is and in any event the adjective 'first' must be given its proper weight. Justification is clearly not site specific. Once the practice of manufacturing MOX fuel is accepted to be justified, the exercise will not have to be repeated (absent any new and important evidence which may engage Article 6.2) if manufacture of MOX fuel at any other site is proposed. It may be said that that is unrealistic since manufacture by BNFL at Sellafield is the only feasible option now and in the foreseeable future. That is not the point since I am here concerned with the correct meaning of and approach to Article 6.1 and that cannot depend on the factual position in an individual Member State.

20. The capital costs involved in enabling the activity to go ahead will vary from site to site. Existing facilities may well affect substantially the amount to be spent. Furthermore, there will be considerable difficulty in assessing any net disbenefit in any given case. What is needed is a justification of the type of practice. It may be too expensive and not cost effective or there may be no satisfactory market for the end product. I accept that it will be necessary to identify which costs can truly be said to be included in the practice itself. Clearly capital costs involved in constructing the plant will not, nor should any costs peculiar to the site, such as those incurred because of a data falsification incident, save insofar as that incident affected the wholly future economic benefits of the type of practice in the United Kingdom generally. It follows that on any view here there was a positive NPV and the decision would have been that justification was established.

21. I have, however, to consider the position on the basis that that is wrong. In the alternative, both Mr Sales and Mr Pannick submit that it cannot be wrong in law for the Secretaries of State to adopt a proper, indeed a classic economic approach to sunk costs. There is nothing in Article 6.1 which in terms supports Lord Lester's construction and nothing is said in Article 6.1 itself or in any domestic legislation about when an application should be made, other than that it must obviously be in advance of the adoption or approval of the type of practice in question. It is the absence of any such requirement that has led the EA to be concerned and may underline the Commission's comments. Since bad faith is not suggested, the time at which the application was made cannot be held against BNFL and so the usual principle ought to be applied. Accordingly, no error of law has been established in the approach taken by the Secretaries of State. That submission is in my view correct."

The argument before us

32. The strength of the appellants' argument seems to me to lie in two main points: first, that the decision whether or not to approve a new practice ought not to depend upon when, in relation to the expenditure of capital costs, the decision as to justification is taken; secondly that, on a generic approach to justification, it is in any event impermissible to ignore the sunk costs of a particular project since the decision will apply generally and permit the practice to be carried on at other plants whose capital costs would not be sunk.
33. Whereas below Lord Lester's argument appears to have focused principally, if not exclusively, upon the first of those points, before us he emphasised the second. Both points, of course, are met by Collins J's main ground of decision: if capital costs are always to be ignored, then it matters not when in relation to their expenditure the decision is taken, nor does it affect the position that other plants may be constructed in future. That ground of decision, however, is not easily supportable and, indeed, before this court neither the respondent Secretaries of State nor BNFL advanced it as their principal line of argument. The highest that Mr Sales put it for the respondents was that they were entitled to leave capital costs out of account – not least because they are entitled to regard such costs as much as a benefit (to other areas of the economy) as a detriment. For BNFL, Mr Pannick QC's main argument assumed that unexpended capital costs would ordinarily fall to be brought into account as part of the detriment but, in a two-part submission, he contended first that sunk costs should always be ignored, and secondly, since in reality SMP is the only plant contemplated for this manufacturing process, that the generic nature of the decision is immaterial. Essentially, therefore, Mr Pannick seeks to support Collins J's alternative basis of decision in paragraph 21 (although, of course, paragraph 21 said nothing as to the generic nature of the decision, doubtless because so little was made of this below).
34. For my part I cannot accept Collins J's first ground of decision. It seems to me that Lord Lester is right (subject to the argument on sunk costs) in submitting that the capital costs inherent in a new type of practice (mainly, no doubt, the costs of constructing the necessary plant) are indeed a cost of the practice and relevant, therefore, when evaluating the overall economic benefit (or detriment) likely to result from adopting the practice. Assume, for example, that approval were sought to operate a number of nuclear power stations and that no such stations had previously been constructed. To my mind it could not reasonably be suggested that the capital costs of the project should simply be ignored. Of course, even if taking these costs into account were to result in a projected net loss, it would not follow that the regulatory authorities could not properly regard the practice as justified under Article 6: they might well regard the expenditure of part at least of these capital costs as economically beneficial to the wider economy and they might in any event perceive a variety of other benefits to accrue from the availability of nuclear power. What, however, if the capital costs have already been expended? That is the critical question.

Sunk costs

35. It is convenient to consider this question initially without the added complication of the generic point. Let me at once set out the classic economic approach to sunk costs, an approach graphically explained in a text book on the principles of corporate finance:

"Forget Sunk Costs. Sunk costs are like spilled milk: They are past and irreversible outflows. Because sunk costs are bygones, they cannot be affected by the decision to accept or reject the project, and so they should be ignored.

This fact is often forgotten. For example, in 1971, Lockheed sought a federal guarantee for a bank loan to continue development of the TriStar airplane. Lockheed and its supporters argued it would be foolish to abandon a project on which nearly $1 billion had already been spent. Some of Lockheed's critics countered that it would be equally foolish to continue with a project that offered no prospect of a satisfactory return on that $1 billion. Both groups were guilty of the sunk-cost fallacy; the $1 billion was irrecoverable and, therefore, irrelevant."

36. Lord Lester's argument on sunk costs is essentially two-fold. First he submits that the classic economic approach is simply not the appropriate way to resolve the issue of justification under Article 6.1. The classic economic approach, he points out, is apt for determining which future course of action is commercially most efficient. It is the approach to be taken by investors when deciding what to do in the light of already committed funds. Article 6.1 is concerned not with choosing a further step but rather with justifying a practice. The Secretaries of State should approach the question as regulators, not as accountants. Secondly, Lord Lester submits that the sunk costs approach creates a perverse disincentive against early application for approval under Article 6. He prays in aid in this regard paragraph 1.7 of the EA's letter of 3 November 1998 (see paragraph 8 above), paragraph 115 of ICRP Recommendation No. 60 (see paragraph 13 above), paragraph 60 of the White Paper of July 1995 (see paragraph 22 above), and the Commission's comments on Article 6 in their Communication document (see paragraph 23 above).
37. Attractively though the argument was presented, Mr Pannick's response to it is straightforward and to my mind entirely convincing. Article 6 requires of the Member State no more than that before a particular practice is first adopted or approved it shall be justified. The question for the Secretaries of State here was whether approving the practice would produce an economic benefit outweighing the health detriment. That was a question necessarily falling to be resolved as at the date of the decision. There is nothing in Article 6 which requires the Secretaries of State in reaching their decision to disapply standard economic principles including that of ignoring sunk costs. Nor would it make any sense for them to do so. Take this very case. Assume that the Secretaries of State at the time of their decision were quite satisfied that the economic future for MOX manufacture at SMP was very bright and that it was clearly in the national interest to proceed with it. Is it to be suggested that they should, or even could, nevertheless have felt obliged to withhold their approval on the basis that, had the sunk costs not already been expended, a different view might have been taken of the project's overall net profitability and, accordingly, its prospective justifiability. That surely would be the height of irrationality and, indeed, an adverse decision on that basis would to my mind be open to legal challenge by BNFL. The economic benefit of the practice (i.e. the human activity of manufacturing MOX) is the same whether the matter is being considered by an accountant or a regulator; it is merely that the latter must take account of many additional considerations in deciding whether to approve the project.
38. This further difficulty arises, moreover, on the approach invited by Lord Lester. How precisely should the Secretaries of State take into account the sunk costs? Are they somehow to attempt to recreate the position existing before those costs were expended? As was stated in the final sentence of paragraph 86 of the decision letter, those "costs were themselves incurred at a time when a different assessment of commercial opportunities may have been made". Plainly BNFL would not have embarked upon this huge expense unless at the time it had appeared to make sound economic sense for them to do so. And difficult as it is to assess (from a national, not merely BNFL's standpoint) the economic benefits and disbenefits of a project viewed prospectively in real time, it would be immeasurably more so to attempt the exercise in retrospect.
39. Of course it makes sense for approval to be sought and the issue of justification to be decided as soon as reasonably possible. In the first place it is obviously desirable that before prospective operators expend large sums of money they should know whether their projects will be approved. But in addition the policy which plainly underlies Article 6 will best be served by early decision-making so that unprofitable schemes, unless justifiable on other grounds, can be avoided rather than (as theoretically could happen) permitted because of sunk costs and late decisions.
40. It is, I apprehend, the essential purpose of Article 6 to deter radiation-producing practices unless sufficient net benefits will result. So much, indeed, appears from paragraph 112(a) of ICRP Recommendation No. 60 (see paragraph 13 above). But that is not to say that when taking decisions in individual cases Member States are required to assess economic benefit otherwise than by reference to classic economic principles.
41. I could understand an argument to the effect that the Secretaries of State, in order to promote early decision-making under Article 6, could themselves adopt a policy of taking some limited account of sunk costs. Even that, however, might well be challengeable by the operators: they, after all, will in effect be being penalised for delay over which they might have little control and despite there being nothing in Article 6 requiring decisions to be taken any earlier than before the practice is first adopted or approved. Be that as it may, it is not in any event Lord Lester's argument. Rather his contention is that capital costs, sunk or not, must always be brought into account in full measure.
42. Turning then to the documents on which Lord Lester seeks to rely, I confine myself to the following brief comments.

i) Paragraph 1.7 of the EA's November 1998 letter, so far from supporting the appellants' argument, seems to me logically inconsistent with it. The reason why the EA "was unable to consider the full economic case for the MOX plant" is because the capital costs incurred in constructing it were "sunk". The EA was "dissatisfied" no doubt for the policy reasons I have endeavoured to explain.

ii) The reference in paragraph 115 of the ICRP Recommendation to "the detriment to be considered [is to include] … the costs of the practice" is neutral: it says nothing as to whether to ignore or take account of sunk costs.

iii) Paragraph 60 of the July 1995 White Paper may be rather more helpful to the appellants' argument. The suggestion that, were authorisation to be sought at an early stage, "justification would have been considered without the need to take account of any substantial sunk costs", seems to imply that even if it were sought late, sunk costs would nevertheless be taken into account. Even then, however, it is perhaps unclear in whose favour. The White paper cannot in any event dictate the correct approach to Article 6. In truth, both paragraphs 60 and 62 were making the policy point that it would be desirable for applicants to seek early approval for their projects, for "the greater certainty it could provide".

iv) The reference in the Commission's Communication to the Article 6 determination taking place "as early as possible to reduce the influence of the already incurred costs" was discussed at some length before us as it had been in the court below. Whilst, however, its meaning is far from clear, it is difficult to see how it can support the appellants' argument. Lord Lester's case is that the timing of the decision is irrelevant. Yet the one thing clearly implicit in this sentence is that the incidence of sunk costs will influence the balance and thus the decision. That must be so either because such costs are to be ignored, as Mr Pannick submits, or because the Commission supposes that they will be brought into account in favour of approval i.e. they are guilty on the first limb of the sunk cost fallacy.

The generic approach

43. I turn now to consider whether the fact that Article 6 requires a generic assessment of justification affects the approach to sunk costs in the present case. Lord Lester submits that it does since the approval attaches to the MOX manufacturing practice generally and inevitably this would involve capital costs in future were it to be undertaken elsewhere than at SMP. Lord Lester does not, of course, suggest that the full £470 million already expended on SMP should be brought into account £ for £ against ADL's assessed NPV of £216 million for SMP's future operation. Rather he recognises the need not merely for the sort of exercise referred to at the end of paragraph 17 of the judgment below (quoted in paragraph 30 above) but also for some adjustment to reflect the fact that future plants might not cost the same as SMP.
44. Mr Pannick's contrary argument is, as stated, that MOX manufacture at SMP is in reality the only game in town and that to refuse it on the basis that approval would apply to the practice generally would be absurd.
45. Once again I have found Mr Pannick's argument entirely convincing. That SMP alone was under consideration for approval is surely plain. Paragraph 25 of the decision letter in truth puts it somewhat conservatively, suggesting merely that BNFL's position as the only UK manufacturer of MOX "is likely to continue to be the case for the foreseeable future". As was stated in the appellants' original detailed grounds of challenge (footnote 2 to paragraph 5):

"The decision document refers to justification as a 'generic test', but it is accepted in this case that what has to be justified is SMP."

46. So much, indeed, appears to have been expressly accepted at first instance (see paragraph 26 above).
47. In these circumstances it cannot in my judgment be said that the Secretaries of State were bound to take into account costs which had already been incurred in constructing SMP, which plainly cannot be recovered, and which equally plainly are not going to be incurred anywhere else. Secretaries of State are entitled to decide these cases in the real world. To bring into account sunk costs on the fictional basis that equivalent costs would be incurred were the approval to be invoked to operate the practice elsewhere in future would be absurd. It would be to sacrifice reason on the altar of blind theory. I cannot accept that Article 6 on its true construction requires such an economically nonsensical approach.
48. Little remains to be said. There was much debate before us as to how satisfactorily these questions were dealt with in the decision letter. Ultimately, however, as Mr Sales rightly emphasised, this challenge is not to the reasoning of the Secretaries of State but rather to their entitlement to ignore sunk costs. The appellants' case stands or falls upon the issue of principle it raises as to the proper construction of Article 6. Lord Lester's argument is that as a matter of law the Secretaries of State were bound to take into account (in the sense of deducting from the net economic benefit of the project) the capital costs of constructing SMP even though these costs were sunk. In my judgment, for the reasons already given, that argument fails.
49. With regard to the decision letter, therefore, it is sufficient to note that it identifies "a significant demand … for the manufacture of MOX fuel" (para. 72); that it regards it as "likely that a MOX manufacturing capacity will be an important adjunct for nuclear fuel processors, such as BNFL, who seek to maintain their market position" (para. 72); that the radiological detriments of the practice would be "very minor" (para. 73); that accordingly the economic benefits are sufficient to justify the manufacture of MOX fuel (para. 73); that this general conclusion is supported by the specific economic prospects of SMP (paras. 74 and 75); that there are likely to be substantial additional economic benefits for BNFL's other business, including nuclear fuel re-processing, flowing from the operation of SMP (para. 77); and that the future operation of SMP is likely to support up to 480 jobs in West Cumbria (para. 87). That the economic benefits of operating SMP are being evaluated prospectively and ignoring, therefore, the sunk costs of construction is plain from several of the quoted paragraphs of the decision letter and, of course, is expressly stated in paragraph 86. That approach was in my judgment plainly open to the Secretaries of State.
50. I would accordingly dismiss this appeal.

LORD JUSTICE WALLER:

51. I agree.

LORD JUSTICE DYSON:

52. I also agree.
Order: Appeal dismissed no order for costs. (Order does not form part of the approved judgment)