England and Wales High Court (Administrative Court) Decisions

IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
ADMINISTRATIVE COURT

		Royal Courts of Justice Strand, London, WC2A 2LL
		21/03/2019

Mrs Justice May DBE:

Introduction

1. The Gwynt-y-Môr wind farm is located off the shore of North Wales. It comprises 160 wind turbines and produces a very large amount of renewable electricity. It is owned and operated by the Claimant. The electricity it generates is transmitted onshore via a number of subsea electricity cables and other transmission assets that are owned and operated by the First Interested Party ("the GyM OFTO^[1]").
2. In these proceedings, the Claimant seeks a review of the decision by the Defendant ("Ofgem") dated 8 September 2017, that the physical failure of one of the GYM OFTO's subsea electricity cables should be treated as an "Income Adjusting Event" ("IAE"). The meaning of that term is explained below.
3. The effect of Ofgem's decision was that the GyM OFTO's pre-determined 20-year revenue stream could be adjusted to take account of the event, by allowing relevant costs associated with the cable failure to be "passed through". It is common ground that the effect of this pass-through, in the context of the regulatory charging regime, is to re-route the majority of any IAE award in respect of costs associated with repairing the damaged cable to the Claimant.
4. The Claimant's PAP letter was dated 7 November 2017, to which Ofgem responded on 23 November 2017. Proceedings were issued on 7 December 2017, and permission was given by Mostyn J on 20 July 2018.

Factual and regulatory background

Offshore transmission regulatory regime

5. Generators of electricity like the Claimant are required to be separate from the owners of transmission equipment. In the case of the transmission infrastructure that connects an offshore wind farm to the National Grid on the mainland (National Grid Electricity Transmission Plc ("NGET")), the Authority is required to operate a competitive tender process for the right to own and operate those transmission assets^[2]. The successful tenderer becomes an Offshore Transmission Owner ("OFTO") and is granted an offshore transmission licence by the Authority. The OFTO owns, operates and maintains the transmission assets for a period of 20 years, in exchange for a regulated revenue payment.
6. Bidding to become an OFTO is on the basis of the amount of the required revenue stream for the 20-year period, based on each tenderer's required return on investment, the cost of financing its investment in acquiring the assets and the ongoing cost of financing, operating and managing the assets. For the 20-year period after purchasing the assets, the OFTO receives the revenue stream set out in its licence, which reflects the amount of its bid subject to various adjustments ("the Tender Revenue Stream").

Costs adjustment to the Revenue Stream

7. Amended Standard Condition E12-J3 (Restriction of Transmission Revenue: Allowed Pass-Through Items) of an OFTO's licence ("the IAE Condition") sets out cost adjustment terms for eight different "Allowed pass-through items". If the relevant conditions are satisfied, each adjustment term allows costs incurred by the OFTO to be "passed through", thereby adjusting the revenue stream received by the OFTO for the relevant year. Paragraph 1 of the IAE Condition provides:

"The purpose of this condition is to provide for revenue adjustments to reflect certain costs that can be passed through to consumers as part of allowed transmission owner revenue."

8. The criteria for whether an event is an IAE are set out in paragraph 15 of the IAE Condition, as follows:

"An income adjusting event in relevant year t may arise from any of the following:

(a) An event or circumstance constituting force majeure under the STC;

(b) An event or circumstance resulting from an amendment to the STC not allowed for when allowed transmission owner revenues of the licensee were determined for the relevant year t; and

(c) An event or circumstance other than listed above which, in the opinion of the Authority, is an income-adjusting event and is approved by it as such in accordance with paragraph 21 of this licence condition,

Where the event or circumstance has, for relevant year t, increased or decreased cost and/or expenses by more than £1,000,000 ("the STC threshold amount")."

(STC stands for System Operator Transmission Owner Code).

It is the application of (c) above which has been the focus of these proceedings. The parties referred to this provision as "Limb (c)".

9. Paragraph 21 of the IAE Condition provides for Ofgem to consult with "the Licensee and such other persons as it considers desirable".
10. Paragraph 22 of the IAE Condition provides for a time limit of 3 months from the latest of (i) the service by an OFTO of an event said to be an IAE, or (ii) the receipt by Ofgem of any further information which it has requested, for Ofgem to make its determination as to whether or not the claimed event is an IAE. In default of a determination within the prescribed period, the full costs claimed are deemed to have been passed through.

The Charging regime - TNUoS

11. Payment of the Tender Revenue Stream to the OFTO is made by NGET. NGET recovers this sum through Transmission Network Use of System Charges ("TNUoS") which all network users, including offshore generators such as the Claimant, are required to pay to NGET.
12. The terms of NGET's electricity transmission licence require NGET to establish and maintain a TNUoS charging methodology in accordance with certain stipulated objectives. The calculations of the amounts paid and received by relevant parties are set out in the TNUoS charging methodology contained in section 14 of the Connection and Use of System Code ("CUSC").
13. The intricacies of TNUoS calculations under the CUSC were set out in detail in the witness statement of Min Zhu, Deputy Director of Electricity Transmission at Ofgem. Happily, it was not necessary to examine the CUSC charging regime or details of the various charging algorithms at the hearing as the effect of the regime on the Claimant was not in dispute: through the system of TNUoS charges some 75% of the costs of repair of the relevant undersea cable (approximately £13.2 million) will, as a result of the IAE pass-through, be charged to, and paid by, the Claimant.

GyM cable failures and insurance cover

14. The Gwynt-y-Môr wind farm and its four subsea cables were originally constructed by a joint venture of developer companies. The cables were known as SSEC1-4, SSEC being an acronym for sub-sea electricity cable.
15. The transmission assets at GyM were transferred to the GyM OFTO on 11 February 2015, pursuant to the terms of a Sale and Purchase Agreement of the same date ("the SPA"). The SPA contained various indemnities and warranties as between the Claimant as vendor and the GyM OFTO as purchaser.
16. At the time of purchasing the assets the GyM OFTO had in place an Operating All Risks ("OAR") insurance policy dated 6 February 2015 (commencing on 17 February 2015) ("LEG2 cover"). Clause 14 of this policy, entitled "Exclusions", included at sub-paragraph (g) the "LEG [London Engineering Group] 2/96" exclusion. The effect of the various LEG exclusion clauses is conveniently summarised in a presentation by the GyM OFTO's brokers, Willis Towers Watson ("Willis") in February 2017:

"The cover provided for internal faults is addressed primarily by way of the defects exclusions which, depending on the level of cover afforded, writes back an element of cover.

The cover written back from the exclusion is referred to as a LEG clause. The London Engineering Group (LEG) is an informal association of engineering underwriters which has drafted various clauses to address the extent of defects cover which may be provided. These fall into three categories:

LEG1 – "full exclusion"

LEG2 – "consequences design cover". The policy would exclude the cost which would have been incurred if replacement or rectification of the insured property had been put in place immediately prior to the damage (e.g. would exclude the cost of replacing/rectifying the faulty parts but would cover the consequential losses).

LEG3 – "full defects design cover". The policy would exclude the cost incurred to improve the original material workmanship, design plan or specification (e.g. would cover also the replacement of the faulty part but not the costs of rectifying/improving the original design/workmanship).

…

As might be expected, the broader the cover selected the greater the premium rate applied."

As appears from the above, a LEG2 policy provides a lesser degree of cover than a LEG3 policy.

17. On 2 March 2015 one of the four cables owned by the GyM OFTO suffered a physical failure ("the SSEC1 failure"). The GyM OFTO notified a claim to its insurers and in August 2015 insurers made an advance payment to the GyM OFTO in respect of the SSEC1 failure. (The GyM OFTO was later obliged to repay this sum when insurers denied liability under the terms of the LEG2 cover).
18. Engineers Edif ERA undertook a detailed laboratory examination of the faulty section of the SSEC1 cable on 2 September 2015. The technical report concluded that the SSEC1 failure stemmed from "puncture of the cable during manufacture after laying up of the cores and prior to armouring".
19. On 24 September 2015 insurers informed the GyM OFTO that as the detailed examination had indicated a likely manufacturing fault, the LEG2 exclusion applied and they had no liability for the costs of repair. They sought repayment of the advance payment which they had earlier made.
20. On 25 September 2015 the GyM OFTO sustained a second cable failure, this time to the SSEC2 cable ("the SSEC2 failure"). The subsequent technical investigation reported that the most likely cause was "damage that punctured the FOC sheath during manufacture of the [fibre optic cable] or assembly of the Export cable"
21. On 13 and 16 June 2016 the GyM OFTO served notices under paragraph 14 of the IAE Condition inviting Ofgem to determine whether the SSEC1 and SSEC2 failures respectively constituted an IAE. The costs associated with each failure were identified as £10.2 million in respect of SSEC1 and £14.2 million in respect of SSEC2.
22. The person at Ofgem charged with responsibility for making these IAE decisions was Akshay Kaul, then Director of Networks ("AK").
23. On 23 May 2017 Ofgem decided that the failure of the SSEC1 cable did not qualify as an IAE and thus that the GyM OFTO would itself have to bear the costs of that failure (the "SSEC1 Decision").
24. Having requested further information from the GyM OFTO concerning the second cable failure, the 3-month period for a determination under paragraph 22 of the licence (referred to above) expired on 8 September 2017.
25. On 8 September 2017, Ofgem came to the opposite conclusion in relation to the failure of SSEC2, namely that it did qualify as an IAE (the "SSEC2 Decision"). Ofgem was aware of the effect that its decision would have on the Claimant; an internal presentation which AK made on 31 August 2017 entitled "OFTO Income Adjusting Events" included the following:

"The costs of an IAE are mostly passed back to the wind farm generator through the local transmission charge. …the generator that built the asset will be considerably worse off compared to competing generators."

The SSEC1 Decision

26. By its decision dated 23 May 2017 the Authority concluded that the SSEC1 failure was not an IAE under either sub-paragraph (a) or (c) of the IAE Condition, being the two limbs on which the GyM OFTO had sought to rely. In particular, as to Limb (c), Ofgem stated that it would apply the approach for assessing claims that it had adopted in its earlier decision on an IAE claim from Blue Transmission London Array Limited dated 27 October 2016 ("the BTLAL decision").
27. The BTLAL decision set out four factors that Ofgem would take into account when exercising its discretion under Limb (c) (numbering added):

"1. Whether the Licensee knew of the event or circumstance before it arose or ought to have known of it;

2. Whether the risk of damage of that type was reasonably foreseeable (even if the particular way in which the damage has occurred may not have been);

3. Whether there are nevertheless exceptional factors in the relevant case that mean that the event or circumstance, or its consequences, could not have been reasonably foreseeable; and

4. The ability of the Licensee to manage the risk or impact by putting in place and pursuing risk management arrangements such as insurance, commercial recourse against third parties and/or operating practices."

("the BTLAL factors")

28. In paragraphs 42 and 44 of its decision on SSEC1, Ofgem set out its policy behind the determination of IAEs as follows:

"42. In assessing whether an event or circumstance is an IAE under Limb (c), we have considered, consistent with the BTLAL Decision, the balance of risk and whether the Licensee is the most appropriate party to manage the risk of the event. To determine this, we have considered the extent to which the Licensee was, or should have been, in a position to foresee the event or circumstance and the level of control it had to mitigate the impact of such an event.

…

44. As noted in the BTLAL Decision, we consider that such an approach [applying the BTLAL Factors] is consistent with the overarching design of the Offshore regime and with the Authority's statutory duties, in particular its principal objective to protect the interests of existing and future consumers in relation to electricity conveyed by transmission systems. For example, we do not consider it to be in the interests of consumers to pass through those costs arising from a type of damage that was (or should have been) foreseeable to a bidder/OFTO, solely because the precise damage of the type that occurred was not foreseeable; we therefore consider it appropriate to adopt a narrower, rather than a broader construction of Limb (c) in this regard. Such an approach also seeks to ensure that bidders are properly incentivised to conduct due diligence in respect of the assets, to put in place appropriate commercial arrangements prior to asset transfer and to pursue any relevant third parties who may be liable (such as developers, manufacturers, installers and insurers). The OFTO regime facilitates commercial transactions for large-scale infrastructure investment. We consider that the OFTO is responsible for managing its investment including adopting what it considers are suitable risk management measures."

29. Having considered the BTLAL factors, Ofgem found that the SSEC1 failure was not an IAE under sub-paragraph (c) for the following reasons (references in square brackets are to paragraphs of the SSEC1 Decision):

(1) the risk of the event was foreseeable and the GyM OFTO was the most appropriate party to manage the risk of the event [45].

(2) on the facts the GyM OFTO could not reasonably have known about the specific fault that arose [47]. However the invitation to tender document for the transmission licence had made clear that the risk of a 'major default event' had to be addressed by OFTOs including those relating to internal cable failures; and the GyM OFTO had also identified the risk of an internal cable failure in its bid for the transmission licence, having set out its approach to repair costs arising due to a major failure event including any internal able failures. Such failures were therefore a type of risk that was foreseeable [50].

(3) There were no exceptional factors that meant that the event or its consequences could not have been reasonably foreseen. [51]

(4) Ofgem expects licensees to pursue third parties for remedies in respect of their negligent or below standard work, and to put in place other commercial arrangements and risk management practices. [53]

(5) Ofgem also expects licensees to put in place appropriate insurance arrangements to manage risks and to satisfy themselves that the insurance cover was suitable for their needs. In this case, the GyM OFTO took out an OAR insurance policy subject to a LEG2 exclusion, which was a commercial decision that it made at its own risk, even though Ofgem understood that LEG3 cover was available in the market at the time that the GyM OFTO bid for the Licence. Ofgem stated that, in general, it did not consider it appropriate for consumers to bear a risk that arose due to the interpretation of commercial arrangements between the Licensee and third parties [54].

30. Ofgem arrived at the following conclusion regarding assumption of risk [56]:

"In summary there is nothing in the information provided that suggests to the Authority that the risk of the Event was of a type so unforeseeable or exceptional that a prudent licensee would not have contemplated that risk in assessing the project prior to submitting its tender or fixing the revenue entitlement. We note that the Licensee may be able to recover only part of the costs incurred through commercial arrangements. However, whether or not the Licensee can recover any or all of the costs it incurred through commercial arrangements, the Authority considers that the Licensee is the most appropriate party to manage the risk of the Event."

31. At [59] and [64] Ofgem went on to explain and clarify risk apportionment under the generator-build model:

"59. As stated in the BTLAL Decision, an overarching premise of the generator build model was that the developer bears the risks associated with construction of the transmission assets, such as increased costs from construction overruns and the failure to complete the assets on time during the construction phase. In contrast the OFTO is responsible for owning and operating the transmission assets from the point of asset transfer, and for owning and operating the transmission assets from the point of asset transfer, and for the associated risks arising from ownership of the assets. The Offshore regime was not designed to insulate OFTOs from all risks that could somehow be traced back to the construction of the assets such as a latent defect.

…

64. We have considered the potential wider implications for the OFTO regime as a result of this determination where relevant. In our view, this particular determination should not materially impact future OFTO tenders. We consider that the regime will continue to bring benefits to consumers by providing long term visibility on transmission costs through a competitive procurement process. Further, to address the issues covered in this determination, we believe it is incumbent on developers to work actively with OFTOs to ensure future issues are prevented through good practice in project development. We are keeping a watching brief on developments occurring in the industry regarding offshore transmission assets."

The SSEC2 Decision

32. Ofgem subsequently decided in relation to the SSEC2 failure that it did constitute an IAE, applying its discretion under Limb (c). It reiterated its previous approach, namely to apply the BTLAL factors. Its conclusions in relation to the first two BTLAL factors were identical to the conclusions that it reached in the earlier SSEC1 Decision: the GyM OFTO had identified the risk of an internal cable failure in its bid and such failures were of a type that was foreseeable (paragraph 52 of the SSEC2 Decision; hereafter references in square brackets are to paragraphs of the SSEC2 Decision).
33. As to the third of the BTLAL factors, Ofgem concluded that there were no exceptional matters rendering the event or its consequences not reasonably foreseeable, commenting as follows:

"On the present facts the Cable failure suffered by the Licensee arose from a foreseeable type of event, namely a latent defect in the transmission assets. The independent technical report identified that the Cable Failure "presented many similarities" to the SSEC1 Failure. It is possible that the root cause of the two failures was the same. However, the fact of the SSEC1 Failure occurring did not increase the chances of the Cable Failures occurring; they were independent failures…." [55]

34. Ofgem went on to find that the SSEC2 cable failure was an IAE, on the basis of the fourth BTLAL factor, namely the ability of the OFTO to manage the risk or impact by putting in place and pursuing risk management arrangements such as insurance, commercial recourse against third parties and/or operating practices.
35. At [57] and [58], Ofgem made the following statements about the apportionment of risk under the OFTO regime, in substantially the same terms as paragraphs 52 to 53 of the SSEC1 Decision:

"57. Similar to any other transaction involving a purchase of assets, a licensee should enter into such transactions with the awareness that it is assuming any risks arising from damage or defects that it has not been able to discover through its due diligence. The Offshore regime was not designed to insulate licensees from all such risks [footnote 15: The framework for the Offshore regime also reflects this through the STC which deems the OFTO, for the purpose of the STC, to have been the party that developed the transmission assets from the point of asset transfer…]. Even if a licensee believes, having conducted a reasonable level of due diligence, that the construction of the assets had been undertaken properly and to the level of reasonable skill and care expected, we do not consider it appropriate for the licensee to be able to pass on the risks arising from defective work in the construction of the assets to consumers.

58. We expect licensees to pursue third parties for remedies in respect of their negligent or below standard work and to put in place other commercial arrangements and risk management practices to ensure they can bear the consequences of risks in the event there may not be any such recourse. We also expect licensees to put in place appropriate insurance arrangements to manage risks and satisfy themselves that the insurance cover is suitable for their needs."

36. Ofgem noted in [59] that the GyM OFTO's insurer had refused to cover the costs of the failure of SSEC2 under the LEG2 cover; it repeated paragraph 54 of the SSEC1 Decision as follows:

"We understand that the Licensee took out an insurance policy that it believed would cover the costs of repair arising as a result of such a cable failure (under its 'operating all risk' policy with a LEG2 exclusion). This was a commercial decision it made at its own risk, and our understanding is that 'operating all risks' insurance cover at LEG3 level was available in the market at the time the Licensee bid for the Licence."

37. Critically, however, Ofgem went on to consider a further argument which had been raised by the GyM OFTO relating to the continued availability of (ex hypothesi, hypothetical) LEG3 cover by the time of the SSEC2 failure, at [60]:

"60. However, the Licensee has argued that its ability "to manage the risk of subsequent cable failure including the [SSEC2 Cable Failure] was impaired by the [SSEC1 failure] in March 2015, which limited the levels of coverage that insurers would have been and subsequently were prepared to offer". On consideration of the evidence, we agree with the Licensee that it would not in any event have been able effectively to manage the risk of the [SSEC2] Cable Failure by putting in place insurance cover. This is because we consider that it is more likely than not that an exclusion would have been imposed on the insurance cover of the Licensee's subsea cables following the SSEC1 Failure, had the insurer considered itself liable to pay out for such failures."

38. At [61] to [64] Ofgem set out three examples of situations in which other OFTOs had "had exclusions placed on their insurance directly following the occurrence of [fibre optic cable] issues". The decision then proceeded to recite the chronology of the two SSEC failures at GyM, before observing, at [65]:

"We therefore consider that the insurer would most likely have imposed an exclusion on any LEG3 policy between technical examination of the SSEC1 failure [on 2 September 2015] and the Cable Failure [on 25 September 2015]."

39. The consequence of this for the IAE decision relating to the SSEC2 failure is set out at [67] and [69]:

"67. Therefore, even if the Licensee had purchased LEG3 cover prior to the SSEC1 Failure, we consider it more likely than not that it would not have had insurance cover for the [SSEC2] Cable Failure. The Cable Failure was therefore likely to have been effectively uninsurable.

…

69. We consider that the likely uninsurability of the Cable Failure renders it an event in respect of which the Licensee was not, and could not have been, able effectively to manage the risk, with the consequence that the Cable Failure should be identified to be an IAE under Limb (c)."

The Hypothetical Question

40. As is apparent from the above, Ofgem's decision as to whether or not the SSEC2 failure qualified as an IAE turned on the answer to a single question concerning hypothetical insurance cover as follows:

"Is it more likely than not that hypothetical insurers providing LEG3 cover to the GyM OFTO from February 2015 would have managed to obtain an exclusion and/or to downgrade cover mid-term in the 23 days from 2 September 2015 to 25 September 2015?" ("the Hypothetical Question")

41. In making its decision on SSEC2, Ofgem answered the Hypothetical Question in the affirmative.

Grounds for seeking review

42. The Claimant's case is that the SSEC2 Decision was unlawful on the following Grounds:

i) Ground 1: Ofgem failed to apply its own stated policy on how to determine whether an event constituted an IAE. In particular, Ofgem did not take into account the fact that the SSEC2 Decision would result in the Claimant having to bear the very significant costs mentioned above.

ii) Ground 2a: Ofgem adopted an erroneous approach on the law regarding the GyM OFTO's insurance position. It expressly but wrongly assumed that if the GyM OFTO had purchased insurance cover for the cable failure which was commercially available in the market, the insurer would nonetheless have been entitled to decline cover (mid-term) for the SSEC2 cable failure, in light of an earlier failure of SSEC1.

iii) Ground 2b: when answering the hypothetical question that Ofgem posed for itself in making the SSEC2 Decision, Ofgem reached an irrational conclusion; and/or failed to take into account relevant considerations; and/or failed to make sufficient enquiries.

43. Before turning to deal with each of these grounds, I should like to record my gratitude to all counsel for their oral and written submissions, all of which were of the very highest order.

Ground 1

Ofgem's failure to apply its own policy

44. Sam Grodzinski QC, for the Claimant, contended that application of the BTLAL factors necessarily involves a comparative exercise as between the OFTO on one hand and the Generator (in this case the Claimant) on the other. He pointed to Ofgem's statement of its approach to Limb (c) at paragraph 44 of the SSEC2 Decision:

"In assessing whether an event or circumstance is an IAE under Limb (c), we have considered, consistent with the BTLAL Decision, the balance of risk and whether the Licensee is the most appropriate party to manage the risk of the event… "

45. Mr Grodzinski argued that it made no sense to consider "balance of risk" without taking into account the other side of the scales; likewise, deciding "the most appropriate party" could only be done by reference to other parties in addition to the OFTO.
46. In response, Monica Carss-Frisk QC, for Ofgem, pointed out that the challenge under Ground 1 had gone through several iterations: at first, the Claimant contended that Ofgem's policy regarding the exercise of its Limb (c) discretion should have engaged a comparative approach. Later, its case changed to one where it was asserted that the policy did in fact require a comparative approach.
47. Ms Carss-Frisk submitted that decisions setting out policy are not to be construed legalistically as one might a statute: she referred me to the collection of citations from cases gathered by Ms Sara Cockerill QC (as she then was) at paragraph 48 of her decision in R (London School of Science and Technology) v. Secretary of State for the Home Department [2017] EWHC 423 (Admin). Her case was that the first sentence of paragraph 44 of the SSEC2 Decision cannot be read alone but must be read together with what follows. The first sentence is merely introductory, given particularity by the second sentence starting "In determining this…". She argued that these words, and what follows in paragraphs 44 and 45, made it clear that Ofgem's policy was to focus on the Licensee and matters relating to the Licensee when considering the fourth BTLAL factor.
48. Ms Carss-Frisk argued further that the decision under Limb (c), as under Limbs (a) and (b), involved Ofgem in an essentially binary consideration of whether an event was or was not an IAE. Construing Limb (c) "in the context of the two preceding limbs", as the BTLAL reasoning required, involved a decision as to whether the event was reasonably foreseeable for the Licensee and if it was one in respect of which the Licensee could mitigate the risk. In any event, she submitted, the onward impact of an IAE decision on other parties within the regulatory regime was regulated by the provisions of the CUSC, over which Ofgem had no control.
49. Mr Grodzinski responded that Limb (c) engaged a wider ambit of enquiry for the exercise of Ofgem's discretion, which was the purpose of its inclusion at paragraph 15 of the IAE Condition. Ofgem was aware that under the charging regime of the CUSC, the result of allowing the IAE would be that the Claimant would bear the costs of the cable failure; in those circumstances Ofgem should have considered a number of other things including the Claimant's capitalisation and its ability to protect itself by insurance, together with the contractual allocation of risk under the SPA and the terms of the deeming provision in the STC referred to at paragraph 35 above.
50. I accept Ms Carss-Frisk's analysis of Ofgem's approach to Limb (c). Ofgem's policy as set out in the BTLAL decision appears to me to be clear that in addressing whether the Licensee is the most appropriate person to bear the costs associated with an event, Ofgem will focus on the position of the Licensee, in this case the GyM OFTO. In my view Ms Carss-Frisk is right that the first sentence of the relevant paragraph of the SSEC2 Decision (paragraph 48) is an introduction, which is thereafter particularised. This view is supported by the fact that in no previous IAE decision regarding OFTOs (there had been three prior to the SSEC2 Decision) had Ofgem adopted the comparative exercise now advocated by the Claimant. Moreover in this case none of the Claimant's submissions to Ofgem had included any detailed information on its circumstances, for instance by providing information on its capitalisation or insurance cover, as might have been expected if the Claimant had sought from Ofgem the comparative approach which it now says it ought to have adopted.

Failure to take into account risk allocation under the SPA

51. I had some difficulty in understanding how the Claimant put its case on this point, in terms of identifying a relevant public law error. The reasons for the Claimant's dissatisfaction were easy to grasp, but I could see no obvious connection between the grievance and a public law error on the part of Ofgem in making the SSEC2 Decision.
52. The grievance arises as follows: the SPA contained an indemnity. I was told that the GyM OFTO is currently pursuing the Claimant in proceedings issued in the Commercial Court under that indemnity for the cost of repair and/or replacement of the defective cables. Ofgem has already determined that the final amount to be passed through on the IAE must take account of any recoveries from third parties made by the GyM OFTO, including against the Claimant in those proceedings.
53. If, however, the GyM OFTO fails in recovering all or part of the cost of repair/replacement in its Commercial Court action, then the effect of the SSEC2 Decision will be to make the Claimant pay anyway, rendering its (on this assumption, by then successful) defence in the Commercial Court action nugatory. If the Claimant succeeds in its defence in the Commercial Court action, it will be a pyrrhic victory, which Mr Grodzinski submits cannot be right; hence the grievance.
54. Ms Carss-Frisk argued that the SPA could not be allowed to trump the regulatory regime. As it has indicated, Ofgem will take claims under contracts with third parties into account in applying the BTLAL factors so as to prevent an OFTO obtaining double recovery but it would not be appropriate, she submitted, to permit the terms of private contracts to dictate the direction of regulatory decision making.
55. Michael Fordham QC, for the GyM OFTO, further pointed out that the BTLAL factors were set in the context of the known contractual arrangements between generator/developer and OFTO. He emphasised that there has been no challenge to those factors in any of the many decisions which Ofgem has made concerning IAEs, nor is there any challenge to the BTLAL factors brought by the Claimant in this case.
56. Mr Grodzinski responded that it is not a matter of the SPA dictating a result in the application of the BTLAL factors, but rather a requirement that in applying those factors Ofgem should have had regard to the ability of an OFTO to put in place commercial arrangements which would have protected it from foreseeable loss. It was irrational, Mr Grodzinski argued, for Ofgem to have asked what insurance the GyM OFTO could have put in place whilst not also asking itself what commercial arrangements it had arrived at, or could have arrived at, with relevant third parties.
57. I pressed Mr Grodzinski on how this would have worked in practice here: how could Ofgem have taken into account the efficacy (or otherwise) of any other terms of the SPA, which are currently the subject of Commercial Court proceedings? He pointed out that Ofgem is in any event awaiting the result of those proceedings in order to determine the amount to be passed through as an IAE; he suggested that Ofgem could either have arrived at its own view of the effect of relevant terms in the SPA, or waited until the outcome of proceedings in the Commercial Court before making its decision on the IAE.
58. These were interesting arguments, but they represented a considerable development of Mr Grodzinski's case set out in his Grounds and his skeleton. As Ms Carss-Frisk pointed out, they appeared for the first time in Mr Grodzinski's reply on the third day of the (three-day) hearing. She invited me not to entertain the new points and I thought that she was right to do so.
59. In any event I agreed with Ms Carss-Frisk that there is a limit to what Ofgem's policy required it to take into account when considering Limb (c). The IAE decision itself could not have waited until the outcome of the Commercial Court action, still less the result of any appeal, as the terms of the IAE Condition required Ofgem to make a decision within three months of notification/provision of further information (see paragraph 10 above). The SPA was concluded against the backdrop of the (known) regulatory regime. As to the availability of some other, or additional, form of contractual risk protection, this would have required Ofgem to speculate upon what other terms the parties might have agreed at the time of concluding the SPA; by contrast the availability in the market of LEG3 insurance cover was not something that called for any speculation.
60. Once it is accepted, as I do, that Ofgem was entitled to apply its policy by reference to the position of the Licensee then it follows that the contractual arrangements under the SPA are relevant only to the extent that any recovery by the GyM OFTO in the current Commercial Court proceedings will reduce the amount allowed by Ofgem as an IAE. I conclude that Ofgem made no public law error by omitting any wider consideration of contractual allocation of risk under the SPA.

Account taken of macro-economic considerations affecting OFTOs

61. Mr Grodzinski's final point under the first Ground criticised Ofgem's decision on the basis that AK had in fact taken into account wider macro-economic factors concerning the financial position of OFTOs generally, not simply the hypothetical insurance position of the GyM OFTO, in arriving at the SSEC2 Decision. Disclosure of internal documentation showed, he said, that Ofgem's thinking process in making the IAE decision on SSEC2 had engaged matters which were wholly outside the narrow compass of a hypothetical enquiry into the continued availability of LEG3 insurance cover. Although the published reasons turned on the answer to the Hypothetical Question, the contemporaneous material suggested that Ofgem had considered much wider questions about the economics of cable failures and the ability of OFTOs generally to bear the costs.
62. Mr Grodzinski referred me to the chronology of the SSEC2 decision-making as evidenced by contemporaneous documentation. As indicated above, the GyM OFTO applied for the costs associated with both cable failures to be declared IAEs in a submission to Ofgem on 30 June 2016. In December 2016 Ofgem sent out a draft decision ruling out both the SSEC1 and SSEC2 cable failures as IAEs. The GyM OFTO then wrote several letters seeking to persuade Ofgem why it and OFTOs in general should not be expected to bear such costs, including, in a letter dated 17 January 2017, a submission to the effect that the most appropriate person to bear the costs of such failures would be the generator (ie the Claimant):

"We appreciate that an IAE would give the OFTO relief for expenditure and have wider consequences in terms of costs to consumers (including the generator). However the consumer base is realistically the only constituency able to withstand the risk of multiple unexpected cable failures. A generator would have other access to the consumer base through power prices, that is not available to an OFTO with its tightly controlled revenue stream set on the basis of this being intended to be a low risk asset with no construction risk.."

63. At this stage the matter of continuing insurance had not been raised. The genesis of the insurance argument appears to have been at a meeting between Ofgem and GyM OFTO shareholders on 9 May 2017 (the meeting also discussed an IAE claim by the same shareholders for a cable failure offshore at Thanet, the GyM and Thanet OFTOs both being owned by the same shareholders). The note of that meeting records that Ofgem remained of the view that none of the cable failures at GyM or Thanet was an IAE but goes on:

"[Kate Kendall] said that the Authority was sympathetic to OFTOs where assets have become uninsurable. We were proposing to issue further guidance on IAEs by the summer and one possible way forward might be to make licence amendments where an OFTO had been able to demonstrate that their assets were no longer insurable…

[Shareholders] argued that the second failure at GYM would not have been covered by insurance. [Steve Taylor] said that the Willis report stated that the insurance position only tightened in November 2015, after the second failure had occurred. This implied that the second failure could have been covered if the OFTO had taken out LEG3 at the outset. [Shareholders] said that the position was arguable – the OFTO would like an opportunity to set out the timeline on Insurance more clearly and discuss this point further with the Authority. He also asked what would happen if a cable failed before the guidance was issued or licences amended."

64. Shortly after this meeting, Ofgem published its decision on the SSEC1 cable failure, finding that it was not an IAE (see above).
65. On 11 May 2017 AK sent an internal email to Jonathan Brearley (Senior Partner, Networks) setting out his thoughts on the SSEC2 IAE issue:

"J

This decision is delegated to me, so no particular action needed from you, but note the following:

…

(b) for the second failure at GyM (SSEC2) we have decided to seek further information on the insurance position. We were minded to reject their claim on the basis that they would have had cover for these costs had they put in place the correct insurance policy…The bidder has made a late claim (this week) that they have evidence to show that even if they had put in place the correct insurance cover at the beginning, the insurers would have withdrawn cover after the first failure event, and no other insurer in the market would have been willing to cover them, leaving the OFTO uninsurable – and therefore, unviable. We consider this unlikely, but I have deemed it safer to ask them for the evidence to avoid any risk of legal challenge later. We have done so, and this gives us a bit more time to decide this third claim.

More generally, I briefly described to you the un-insurability issue for OFTOs, which threatens to make these assets unfinanceable unless the IAEs respond as insurer of last resort in the event the insurance market withdraws from covering an important risk category. We may for instance have a situation at GyM where the OFTO now knows it needs LEG3 cover, is willing to pay for it but cannot find an insurer who would write the policy. If a third failure occurs, it may well be out of cash, and will face a default situation unless there is fresh injection of equity. If equity withdraws and the lenders step in and try to replace the obligor, any new investor coming in will want cover for the costs of repair and installation (the costs of the cable itself will generally be recoverable from the original manufacturers under their warranties). Ofgem would face the same situation if it tries to retender or bring in an OFTO of Last Resort.

So the most efficient solution in this case – provided a claimant can provide clear evidence that the insurance market as a whole has withdrawn from covering the risk of a latent defect failure event – is for IAEs to respond, and for us to pass most of these costs back to the generator through the recharging regime via National Grid.

Our plan is to publish some guidance for the OFTO market on this un-insurability point later in the summer, which will help clarify the issue for lenders to TR4 and TR5 bidders."

66. On 9 June 2017 the GyM OFTO made its final submission to Ofgem including the following, at section G, entitled "Impact on Funding":

"35. From our dialogue with the Authority we note that the Authority recognises that the issue of the IAE mechanism to support property damage claims where insurance is no longer available is of fundamental importance in meeting the Authority's responsibility to ensure that the OFTO regime can be funded by project finance, at high equity to debt gearing.

36. In the case of the GyM OFTO the senior debt and equity funders all invested in the expectation that in such circumstances, the Income Adjustment mechanism would allow recovery of reasonably incurred costs.

37. If this principle is not retained then the impact on the wider market will be (i) increased funding costs as a consequence of the increased risk from narrowing the availability of IAE claims, but more importantly (ii) that bank and bond financing markets will withdraw support from the OFTO regime…."

67. On 13 July 2017, Kate Kendall ("KK"), Head of Interconnectors and OFTOs at Ofgem, sent a long internal email to a number of Ofgem personnel, including AK, reporting the results of their analysis to date. Her email included the following:

"…The key issue raised in the GyM letter of 9 June 2017 is GyM's hypothetical argument that if GyM had taken out LEG3 insurance cover at licence grant, its insurance would have been downgraded following its first cable failure…Having now investigated this issue, we conclude that it is possible that GyM's insurer would have downgraded its insurance, and our current view is that we should be providing OFTO's [sic] protection where something becomes uninsurable, but given it is a hypothetical it is not possible to know. Set out below are the considerations we have identified in determining whether or not to accept GyM's argument and as a consequence grant the IAE claim…

…

Factors that support accepting the un-insurability argument

1. The experience of the Thanet OFTO is that an insurance exclusion was imposed 7 days after the cause of the outage was identified…

..

8. Lenders are concerned with the potential contagion effect of multiple cable failures. An OFTO can absorb one failure through its reserves. For a second event it has no reserves, no-one will lend to it at this point and equity holders have no incentive to inject capital and remain in the market when they have no certainty of any future return on that asset. The potential contagion impact that follows an unfunded repair includes: stranded developer calls on its business interruption insurance (up to 18 months); excessive pay out of BI insurance could drive insurers out of the market; no lending without insurance and therefore the market would ultimately collapse.

9. Payment of a claim on an 'un-insurable' asset is likely to be consistent with our proposed uninsurability guidance, which we are currently developing and propose to release alongside the decision letter in early September 2017.

10. Moody's has downgraded its assessment of the regime from AA to BAA, see note on how regime is assessed attached….

11. The high probability of a repeat failure or a failure where the fibres in the [fibre optic cable] have begun to deteriorate is resulting in insurers placing exclusions on polices…If we don't act as insurer of last resort then OFTOs will fall over leading to OFTO of last resort where it is likely that we will have to recompense the new provider for taking on the assets and so are likely to pay either way.

12. OFTOs are not responsible in any way for the failure of these assets and they are doing everything possible to ensure that they maintain the assets."

68. Mr Grodzinski relied on the above email primarily for its recitation at paras 8-11 of the wider macro-economic considerations which KK and her team appeared to consider relevant to the IAE decision on the SSEC2 cable failure. He drew my attention also to an important error at paragraph 12 of the email: under the terms of the deeming provision in the STC referred to and relied upon by Ofgem in the SSEC1 and the SSEC2 Decisions (set out at paragraph 35 above) an OFTO is taken to have assumed responsibility for all the transmission assets together with latent defects from the time of asset transfer.
69. In a further email dated 17 July 2017 entitled "RE: GYM IAE update" KK reiterated her concern with these issues:

"…it is true that we are seeing new lenders to the market becoming pretty aggressive in terms of pricing but they are blind to these issues surrounding IAEs as they have no knowledge of what is happening on the ground re uninsurability. As I attached to my previous email, the market is taking a different view of OFTO assets and Moodys have downgraded the regulatory regime following the first failure. My fear is this will be exacerbated where we decide not to protect OFTOs from uninsurability and we will see further downgrades.

One of the original constructs of the OFTO regime was to create an asset class which looked and felt like PPP – in SOPC5 there is coverage for debt providers for uninsurability. If we go forward with a regime where we leave the risk of uninsurabiliity with the OFTO it is highly likely that either they will be required to put in significant reserves to cover cable failures (which will result in significant cost to all future OFTO's going forward which will be of a contingent nature) OR we won't be able to finance OFTO's going forward. The risk of uninsurability could potentially push the OFTO asset class to junk status; bond holders and debt just won't entertain the risk.

I realise that the decision has already been taken about the issue of latent defect risk and we are not trying to unpick that. This is a genuinely new argument and is about whether we want to ensure a stable regulatory regime that insulates debt from risks that are completely outside its control. I think we have some work to do about how a mechanism might work, and whether it protects Equity. I would have thought as a minimum we might want to say that any event is equity risk up to a level whereby all its contingencies and equity surplus funds have been used but as soon as it impacts/eats into debt then it is for the consumer to bear that risk, as they are benefitting from the exceptionally low costs of finance attracted to the OFTO regime as a result of clearly packaging up risks that the OFTO can't and shouldn't deal with.…"

70. On 31 August 2017 AK made an internal presentation at Ofgem setting out and discussing the ability of OFTOs in general, and the GyM OFTO in particular, to bear the costs of cable failures. His presentation included the following observations:

OFTOs are very thinly capitalised vehicles…

If OFTO bidders were asked to price the risk of un-insurable latent defects, we know exactly how they would behave because this has been attempted numerous times in PPP transactions…

In contrast, a much better value for money solution is for consumers to provide insurance of last resort through the IAE mechanism in the licence…

The costs of an IAE are mostly passed back to the wind farm generator through the local transmission charge…the generator that built the asset will be considerably worse off compared to competing generators.

…

The current claim (Gwynt Y Mor SSEC2) is the second cable fault on the same project, and the OFTO has adduced evidence from a number of other OFTOs to show that insurers withdrew insurance cover for latent defects within a very short time after the reasons for the first failure were known. It argues that after the first failure, the assets were essentially uninsurable for latent defects and no efficient OFTO would have been able to protect itself. On the balance of the evidence, we agree the assets were probably uninsurable after the first failure."

We recommend the following course of action

Say "Yes" to the current claim (GyM SSEC2) on the grounds that the evidence indicates that insurance cover would not have been available to an efficient OFTO…"

71. Mr Grodzinski submitted, looking at these internal documents, that the SSEC2 Decision took into account an array of economic factors arising out of cable failures impacting the financial position of OFTOs generally, and was not based solely upon the answer to an hypothetical question about insurability.
72. Overnight following the first day of argument, Ofgem served a second witness statement from AK, amongst other things seeking to address these points. As there was no objection taken to the late admission of this evidence, I allowed it in. Ms Carss-Frisk made the point, which I accepted, that the matters Mr Grodzinski had raised in opening represented an extension of his case on Ground 1 which had not been foreshadowed in his amended pleadings or his skeleton argument and that her client had therefore not had a chance to deal with it in evidence.
73. At paragraphs 5 and 6 of his second statement AK states as follows:

"5. Ofgem's analysis of the BTLAL Factors, including its answer to the relevant Hypothetical Question, had concluded by 31 August 2017. That meant that, by this date, it was clear that the Authority would be allowing the IAE under the SSEC2 Decision. It was also clear that the insurance analysis was central to this outcome under the policy as formulated in the BTLAL Factors.

6. That conclusion triggered a wider review within Ofgem, in which Ofgem considered whether, for the purpose of future cable failures, it was content that it had adopted the appropriate policy or whether a different policy should be imposed. That review involved Ofgem considering broader economic arguments about the appropriate IAE policy in respect of uninsurable latent defects. Dermot Nolan (CEO of Ofgem), in particular, was concerned that the SSEC2 Decision should not be published until he had reassurances that the wider aspects of the policy (to be applied prospectively) would be consulted on and altered if appropriate.

…

8. It is therefore wrong for the Claimant to suggest that I reached the decision under limb (c) of the SSEC2 Decision by reference to anything other than the BTLAL Factors, or that Ofgem in fact had a policy to take any other factors into account or to carry out any comparative exercise as contended for by the Claimant. It is true that I received representations, both from Ofgem employees and from industry participants, regarding the broader potential impacts of the SSEC2 Decision on the OFTO regime…But considerations separate from the BTLAL factors formed no part of the basis on which I reached the SSEC2 Decision."

74. In dealing with this new evidence, Mr Grodzinski referred me to an email from AK to KK on 31 August 2017 asking whether the second failure was on SSEC1 again, or on a different cable; he suggested that a basic enquiry of this nature belied a case that AK's analysis had ended on that day, and that his decision had by then been reached. Mr Grodzinski referred also to the email from AK to Dermot Nolan (Chief Executive of Ofgem) on 6 September 2017 setting out "our proposed recommendation on the current IAE claim based on the legal advice we have received our proposed next steps and our view of the risks involved". This followed an email exchange with KK on 5 September 2017 in which AK had responded to her enquiry about how "the meeting with Dermot" had gone in these terms:

"Badly. He's adamant that the OFTOs understood that they were taking the risk of latent defect, and of un-insurability; and that they should have been prepared for the fact that the licence will never respond under any circs to cover latent defects.

Can you do something for me? Dermot wanted a quantitative demonstration of my assertion that it is better value for money for consumers to pay out for un-insurability protection on a few claims, than for bidders to price in a capped reserve of (say) £20m into the TRS [Tender Revenue Scheme], with the consumer covering any costs above this level?"

75. Mr Grodzinski submitted that the matters canvassed in these emails were manifestly policy considerations, circulated to a number of people including in particular AK, to be fed into the decision-making process on SSEC2.
76. Further, Mr Grodzinski invited me to conclude that AK's evidence contained in his second statement, to the effect that the SSEC2 Decision had been made by 31 August 2017, was inconsistent with AK's "recommendation" made in his presentation on 31 August 2017 and his "proposed recommendation" sent to Dermot Nolan on 6 September 2017. He contended that in the light of all this it was fanciful to suppose that policy matters had been kept "hermetically sealed" (as he put it) from an answer to the Hypothetical Question.
77. I was directed to the review of authorities on the provision of late reasons undertaken by Burnton J in the case of Nash v. Chelsea College of Art and Design [2001] EWHC Admin 538 from which he drew the following propositions (at [34]-[35]):

"(i) Where there is a statutory duty to give reasons as part of the notification of the decision, so that (as Laws J put it in Northamptonshire County Council ex p D) "the adequacy of the reasons is itself made a condition of the legality of the decision", only in exceptional circumstances if at all will the Court accept subsequent evidence of the reasons.

(ii) In other cases the Court will be cautious about accepting late reasons. The relevant considerations include the following, which to a significant degree overlap:

(a) Whether the new reasons are consistent with the original reasons.

(b) Whether it is clear that the new reasons are indeed the original reasons…

(c) Whether there is a real risk that the later reasons have been composed subsequently in order to support the tribunal's decision, or are a retrospective justification of the original decision. This consideration is really an aspect of (b).

(d) The delay before the later reasons were put forward.

(e) The circumstances in which the later reasons were put forward. In particular, reasons put forward after the commencement of proceedings must be treated especially carefully….

35. To these I add two further considerations. The first is based on general principles of administrative law. The degree of scrutiny and caution to be applied by the Court to subsequent reasons should depend on the subject matter of the administrative decision in question. Where important human rights are concerned, as in asylum cases, anxious scrutiny is required; where the subject matter is less important, the Court may be less demanding, and readier to accept subsequent issues."

78. Applying these observations, Mr Grodzinski invited me to scrutinise the evidence in AK's second statement carefully and to treat it with caution. Ms Carss-Frisk argued that, in circumstances where her clients had had no notice prior to the hearing of the point that Mr Grodzinski had sought to develop, AK's evidence should not be regarded as a late addition of the kind referred to in Nash. She submitted that it would be wrong to go behind his account of how the SSEC2 Decision had been reached, maintaining that the new case made at this late stage by the Claimant was tantamount to an allegation of bad faith on the part of Ofgem.
79. In my judgement, as the case concerning wider macro-economic factors taken into account was developed very late in proceedings, there was no call to treat AK's second witness statement, immediately responsive as it was to the new argument, with particular caution on this point.
80. Considering the chronology set out above there is, on one view, an apparent inconsistency between AK's evidence given in his second statement and what he said earlier, at paragraph 74 of AK1:

"74. It was a complex decision, with competing factors on either side. Individuals within Ofgem shared their views on the relative balance of the competing factors during the decision-making process and as we were gathering relevant evidence. See internal emails between relevant members of Ofgem."

The internal emails referred to by AK in this paragraph (footnoted) included the KK emails dated 13 and 17 July 2017, set out at paragraphs 67 and 69 above. None of the matters addressed by KK in her email of 17 July concerned insurance cover or any matter touching on continuing insurability. In these circumstances, Mr Grodzinski submitted, the court could not safely conclude that AK's most recent evidence was a genuine attempt to recall the approach he had taken, and not an attempt, long after the event, to re-construct his decision-making process.

81. In my view, however, the same material can also be read consistently with AK's evidence that matters brought to Ofgem's notice in the course of considering the IAE position at GyM prompted a wider policy review. It is entirely possible for decisions to be made on a narrow basis, for that purpose setting to one side considerations which the decision-maker is nevertheless aware will generate and inform a wider debate about how similar decisions should be taken in future. I am satisfied, on AK's later evidence, that that is what happened here.
82. For these reasons I reject the Claimant's case that the SSEC2 Decision should be set aside on Ground 1.

Ground 2

83. Ground 2 seeks to challenge Ofgem's answer to the Hypothetical Question. There was originally a criticism, characterised as Ground 2a, directed at an alleged failure properly to appreciate the legal position regarding an insurer's ability to "impose" an exclusion of cover mid-term. As I understand it the Claimant now accepts that Ofgem was aware of the ordinary legal position: namely that, apart from a situation in which extraordinary factors can be said to alter the nature of the risk, and absent specific contractual terms permitting a change, insurers are not entitled to alter the terms of cover during the agreed policy period (Kausar v Eagle Star Insurance Co Ltd [2000] 1 Lloyds Rep IR 154). Indeed contemporary material shows that AK himself was aware of this, as he referred to it in an email to KK and others on 21 July 2017:

"Another commercial point you should consider is the contractual rights of the insured party. Normally, the insurer has full ability to impose exclusions or raise premia when the contract comes up for renewal. But in my experience it is untypical for them to be able to do so mid-way through an existing policy without the customer having any right of redress."

84. At the hearing, Mr Grodzinski addressed the residuum of Ground 2(a) in the context of one aspect of the challenge to Ofgem's approach to the Hypothetical Question. I deal with this further below.

The correct approach to Ofgem's evaluation of the Hypothetical Question

85. Ms Carss-Frisk submitted that, as the expert regulator, Ofgem's decision on the Hypothetical Question should be accorded particular respect and that the court should be very slow to impugn the approach Ofgem took, or the answer at which it arrived. She relied in this respect on the following observations of Mr Justice Lightman in R v. Director General of Telecommunications ex p. Cellcom Ltd [1999] ECC 314, at [26] and [27] (footnoted references omitted):

"[26] … Where the Act has conferred the decision-making function on the Director, it is for him, and him alone, to consider the economic arguments, weigh the compelling considerations and arrive at a judgment. The applicants have no right of appeal: in these judicial review proceedings so long as he directs himself correctly in law, his decision can only be challenged on Wednesbury grounds. The court must be astute to avoid the danger of substituting its views for the decision-maker and of contradicting (as in this case) a conscientious decision-maker acting in good faith with knowledge of all the facts.¹⁰ As Lord Brightman said in R. v. Hillingdon London Borough Council, ex parte, ex parte Pulhofer.

"Where the existence or non-existence of a fact is left to the judgment and discretion of a public body and that involves a broad spectrum ranging from the obvious to the debatable to the just conceivable, it is the duty of the court to leave the decision of that fact to the public body to whom Parliament has entrusted the decision-making power save in a case where it is obvious that the public body, consciously or unconsciously, is acting perversely."

If (as I have stated) the court should be very slow to impugn decisions of fact made by an expert and experienced decision-maker, it must surely be even slower to impugn his educated prophesies and predictions for the future. Guidance as to the appropriate approach to the written reasons of the decision-maker for his decision (in that case the Secretary of State for Education) was given by Lord Wilberforce in Secretary of State for Education v. Tameside Borough Council:

"These documents are to be read fairly and in bonam partem. If reasons are given in general terms, the court should not exclude reasons which fairly fall within them: allowance must be fairly made for difficulties in expression. The Secretary of State must be given credit for having the background to this situation well in mind, and must be taken to be properly and professionally informed as to educational practices used in the area, and as to the resources available to the local education authority. His opinion, based, as it must be, upon that of a strong and expert department, is not to be lightly overridden."

The guidance is particularly apposite in this case where (as Ms Chambers states in her first affirmation) the Director (assisted by his expert staff) made his decisions of 2 April 1998 after a prolonged consultation period in the light of "his experience of monitoring and regulating the mobile telephony market over a long period.

[27]  The court may interfere with a decision if satisfied that the Director has made a relevant mistake of fact or law. But a mistake is not established by showing that on the material before the Director the court would reach a different conclusion. The resolution of disputed questions of fact is for the decision-maker, and the court can only interfere if his decision is perverse, e.g. if his reasoning is logically unsound, as it was found to be in R. v. Director General of Electricity Supply, ex parte Scottish Power Plc. The court may interfere if the Director has taken into account an irrelevant consideration or has failed to take into account a relevant consideration. But so long as the Director takes a relevant consideration into account, the weight to be given to that consideration and indeed whether any weight at all should be given to that consideration is a matter for the Director alone, so long as his decision is not perverse."

86. Mr Grodzinski drew a distinction between the type of decision being made by the Director General in Cellcom and the nature of the task which Ofgem set itself here, in answering the Hypothetical Question. Ofgem was not expert in addressing what were essentially insurance market questions about exclusion/downgrading of insurance mid-term, Mr Grodzinski submitted.
87. I accept that there is a distinction to be made here. On Ofgem's case, it refined its application of the fourth BTLAL factor to asking a single question, namely the Hypothetical Question. Had its decision on SSEC2 been cast more widely, had this been a challenge, for instance, to Ofgem's evaluation of the macro-economic considerations seen referred to in its internal emails discussed above, then Ms Carss-Frisk's point would have been unassailable. But Ofgem's case is that it asked and answered a single question on the continued availability, unrestricted, of LEG3 insurance cover. As to this, Ofgem was not an expert in the insurance market and specifically whether and how cover might be excluded/downgraded mid-term, as its own staff recognised: an email from Stephen Taylor ("ST") Manager, OFTO Licensing at Ofgem to KK and others dated 30 August 2017 noted, "[c]learly we do not have any expertise in this area ourselves."
88. Accordingly, I accept Mr Grodzinski's submission that I am not required to adopt an overly hands-off approach. Having said this, I have reminded myself that my task is not to substitute my own decision for that of Ofgem's on the Hypothetical Question, but to ask whether there were sufficiently serious public law errors such that the final decision cannot stand.
89. I also acknowledge the validity of Mr Fordham's reminder that there is no challenge to the process adopted by Ofgem, in particular to the identification by Ofgem of the Hypothetical Question as the proper question to be asked.

Specific public law errors alleged by the Claimant

90. Ofgem's reasons for arriving at an affirmative answer to the Hypothetical Question appear from paragraphs 61 to 68 of the SSEC2 Decision. These are clarified and expanded upon by AK in paragraphs 73-94 of AK1 and may be summarised as follows:

(i) Ofgem was aware of three other OFTOs which "had exclusions placed on their insurance directly following the occurrence of [fibre-optic cable] issues" (paragraph 61 of the SSEC2 Decision). It is accepted that the OFTOs referred to are Thanet, Greater Gabbard and A OFTO^[3]. AK refers to these OFTOs in his evidence as "three separate examples of an exclusion being imposed mid-term" (paragraph 86 of AK1).

(ii) Of these three examples Thanet was treated as the closest comparable to the GyM OFTO because of:

(a) the similarity in nature and circumstances of the cable failures (paragraph 76 of AK1). By comparison with Thanet the GyM cable was likely to have been considered especially vulnerable (paragraph 78 of AK1),

(b) the timing of the Thanet insurer's involvement and of its response requiring an exclusion (paragraph 79-80 of AK1),

(c) the fact that Thanet had LEG3 cover (paragraph 81 of AK1),

(d) the fact that the Thanet and GyM OFTOs had a common insurer, Codan (paragraph 82 of AK1),

(e) also the same loss adjuster (paragraph 83 of AK1), and

(f) Thanet had had a previous cable issue in 2011 and Ofgem was aware of a "similar issue" at GyM OFTO (paragraph 84 of AK1).

(iii) Although the GyM OFTO's LEG2 insurers did not in fact seek to change the terms until renewal, this could be explained by the fact that they had rejected cover for the SSEC1 cable failure. There were considerable difficulties in arranging cover at renewal, following the SSEC1 and SSEC2 cable failures (paragraph 88 of AK1).

(iv) Information given by Codan at a presentation given in Berlin in 2017 (paragraph 89 of AK1). Advice provided by JLT "did not materially alter or assist Ofgem's understanding of the relevant position" (paragraph 72 of AK1).

91. The Claimant's case is that Ofgem materially erred in relation to seven of the nine factors above. It is said that, whether taken individually or together, these errors were sufficiently serious as to render the SSEC2 Decision unlawful and/or unreasonable.

Timing – a 23-day window

92. It was not in dispute that timing was a key factor in addressing the Hypothetical Question: any exclusion/downgrade of insurance cover would have had to have taken effect between 2 September 2015 (when Edif ERA performed the detailed investigation of the cable fault on SSEC1) and 25 September 2015 (when the SSEC2 cable failed). AK's evidence was that

"The evidence before us led us to consider that, within the 23-day period following technical examination of the SSEC1 Failure, a hypothetical LEG3 insurer would have been sufficiently concerned about future liability for cable failures that it would have acted to limit its liability." (paragraph 91 of AK1)

In addressing the Hypothetical Question, therefore, Ofgem's task was to consider a 23-day window in September 2015.

93. I now turn to the individual criticisms of Ofgem's answer to the Hypothetical Question advanced by Mr Grodzinski at the hearing.

(1) Factual error – the 7-day issue

94. Paras 61 and 62 of the SSEC2 Decision recorded as follows:

"61. We are aware of three OFTOs that have had exclusions placed on their insurance directly following the occurrence of the FOC issues…

62. The Thanet OFTO had an exclusion placed on its insurance policy following a cable failure which occurred on 23 February 2015…The chronology of the Thanet OFTO events is as follows: the cable failure occurred on 23 February 2015; technical examination of the damaged cable occurred on 21 September 2015; and Thanet was informed by its insurance broker on 28 September 2015 – one week after the technical examination – that a LEG1 defects exclusion was imposed on the cable with immediate effect.." (emphasis added)

95. The Claimant says that these passages demonstrate a significant misunderstanding on the part of Ofgem: it appears from its reasoning that Ofgem understood the Thanet OFTO's insurer to have unilaterally imposed a defects exclusion within a week of the technical examination. However on closer examination the chronology and the process leading to the exclusion of cover at Thanet was very different: on 28 September 2015 Thanet's brokers (Willis) notified Thanet that the insurer, Codan, wanted an exclusion. On 30 September 2015 Willis expressed the belief that if agreement could not be reached Codan would "force our hand and issue 30 days' notice of cancellation". There followed various negotiations covering matters such as whether cover could continue upon payment of extra premium (an offer which Codan refused) and it was not until 13 October 2015, 22 days after the technical examination, that an exclusion was finalised.
96. Mr Grodzinski emphasised two aspects of this mistaken understanding bearing importantly, he said, on Ofgem's decision: in the first place there was no unilateral "imposition" of an exclusion by insurers; secondly, the process of negotiation and agreement took 22 days, not 7. These factors were highly material, he submitted, going both to the weight properly to be attached to Thanet as an example and the likelihood of hypothetical insurers obtaining an exclusion within 23 days in the case of the GyM OFTO.
97. Ms Carss-Frisk accepted that Ofgem had made an error as between 7 and 22 days, but she contended that it was immaterial. In his second statement AK dealt with it in this way, at paragraph 12:

"…if it were to be said that the subsequent disclosure demonstrates that there was a mistake in paragraph 62 of the SSEC2 Decision referring to the exclusion being imposed on 28 September 2015, I confirm that the knowledge that the exclusion was finally imposed on 13 October 2015 would not have altered the outcome of the SSEC2 Decision. The question I was asking in relation to this aspect of the analysis was whether the Thanet OFTO's insurer acted quickly to impose an exclusion, in particular within a period of 23 days. An exclusion being imposed on 13 October 2015 likewise provides an affirmative answer to that question…"

98. Ms Carss-Frisk argued that in circumstances where Codan had achieved an exclusion within 22 days at Thanet the error was immaterial as the 22-day period was still within the 23-day window identified above.
99. AK's evidence in his second statement is to be contrasted, however, with his evidence at paragraph 80 of his first statement:

"80. The Thanet OFTO insurer notified the OFTO about its intention to impose the exclusion seven days after the technical examination of the cable. The relevant question we had to answer was whether the hypothetical GyM OFTO insurer would have reacted between 2 September 2015 (when the SSEC1 technical examination took place) and 25 September 2015 (when the SSEC2 Failure occurred). This 23-day period was therefore a longer period than the mere seven days to react taken by the Thanet OFTO insurer…"

100. AK's use of the phrase "mere seven days" suggests to me that the extent of the difference between 7 and 23 days was an important factor, offering as it did a working margin of some 16 days. On this point, the observations of Burnton J in Nash, referred to above, are particularly apposite. The position advanced by AK in his second witness statement is in my view inconsistent with his approach to the 7 days appearing from the passage in AK1 referred to above.
101. I accept Mr Grodzinski's submission that Ofgem was bound to have approached a 22-day period, giving a margin of just 1 day, with a great deal more circumspection. In the context of a 23-day window an exclusion which (a) has had to be negotiated, (b) over 22 days is a very different matter to one that has been unilaterally imposed by insurers within 7 days. Given AK's evidence in his first statement referred to above, I conclude that the distinction was a critical one, particularly in circumstances where, as the internal documents show, the SSEC2 Decision was finely-balanced.

(2) Further points on timing

102. At paragraph 65 of the SSEC2 Decision Ofgem relied on the fact that the GyM OFTO's actual (LEG 2) insurers notified their decision to decline cover for the costs of repairs to SSEC1 on 24 September 2015, the day before the SSEC2 failure on 25 September 2015.
103. Mr Grodzinski submitted that the timing of a decision to decline a claim on the basis that the policy wording does not cover it is an entirely different matter to the process of negotiating and agreeing an exclusion half-way through the term of a policy, so as to exclude losses which the policy wording would otherwise have been apt to cover. He pointed out that Ofgem appeared to have taken no account of the importance of this distinction, in particular as it affected timing, in arriving at an answer to the Hypothetical Question.
104. Mr Grodzinski also raised other factors impacting timing which Ofgem appeared to have disregarded. In the first place whilst the Thanet OFTO had a single insurer, Codan, the GyM OFTO's LEG 2 cover was written by a pool of insurers, with Codan subscribing to just 4.3%. He pointed out that the process of negotiation and agreement with a multiplicity of insurers, even assuming that every insurer in the pool would have sought an exclusion, must have taken longer than with a single insurer like Codan.
105. Ms Carss-Frisk responded that Ofgem had taken account of the fact that the GyM OFTO's existing LEG2 cover was provided by a pool of insurers, pointing to AK's evidence at para 82 of his first statement:

"…Of course, for Thanet OFTO, Codan was the sole insurer, whereas for GyM OFTO it was one of a pool of insurers. Further, we could not know (and did not consider it necessary to speculate) on the identity of the insurance pool in the hypothetical where GyM OFTO had LEG3 insurance – but it could well have been a different composition from that in fact [sic] obtained."

106. As to the effect of a pool on timing, Ms Carss-Frisk submitted that there was no evidence to suggest that a pool of insurers would have taken any longer to propose, insist and agree the wording of an exclusion than a single insurer like Codan had.
107. In my view Ofgem was in error in failing to distinguish between the process of insurers taking a decision on the scope of existing policy wording and that of embarking on negotiations to introduce a change to policy wording, mid-term; also in failing to address the likely effect on timing (possibly also on the outcome itself) of negotiating with multiple insurers. The fact that negotiating and reaching agreement with many different insurers will take more time than with one alone seems to me to be self-evident; but in any event Ofgem had before it a submission from Willis from which increased delay was to be inferred (see paragraph 126 below).

(3) Different circumstances of cable failures at Thanet and GyM

108. At paragraphs 75 and 76 of his first statement AK explains that Ofgem regarded Thanet OFTO as the closest factual scenario to the hypothetical insurance cover for the GyM OFTO which Ofgem was considering for the purposes of answering the Hypothetical Question.
109. The Claimant criticises this reasoning, contending that the circumstances of the Thanet cable failures were materially different to those at GyM, such that the actions of the single insurer at Thanet, Codan, could not be regarded as indicative of what hypothetical insurers at GyM would have done. Mr Grodzinski referred me to the email dated 28 September 2015 from Willis to Lynn Gladwell of Balfour Beatty (shareholders of both the Thanet and GyM OFTOs) concerning the Thanet insurance, referring to a "wider array of known issues that could potentially cause damage to the cable in the future"; also to a further email dated 1 October 2015 from Willis setting out Codan's view that "what we see is significant issues apparently affecting the whole asset." Mr Grodzinski argued that Codan's concerns at Thanet arose from a separate examination of the remaining 174m length of recovered cable. The engineer's report regarding this length of recovered cable noted 56 surface anomalies together with sheath defects every 3.1m along the full length. An email from Codan to Willis dated 25 September 2015 stated:

"This information brings about a material change in circumstances from Codan's perspective. We now have a wider array of known issues that could potentially cause damage to the cable in future".

110. Mr Grodzinski submitted that what prompted Codan to seek an exclusion at Thanet, as evidenced by the contemporary emails, was not the single-point damage caused by an electrochemical reaction between the fibre optic cable and the core, but the array of additional problems and defects found along the length of cable recovered when that had been separately examined. It was these endemic issues, specific to the Thanet cable, Mr Grodzinski said, which had caused Codan to act to exclude liability, and not simply the electrochemical reaction causing a defect in one location. The SSEC1 cable at GyM had not had any of the additional problems seen at Thanet.
111. Mr Grodzinski referred me to the conclusions of engineers Edif ERA in their report on the SSEC1 cable at GyM from which it is clear, he said, that the fault at GyM constituted a single puncture of the cable during manufacture. It was irrational, he suggested, in these circumstances to take Codan's reaction to the situation at Thanet as any sort of reliable guide to what hypothetical insurers would have done in response to the SSEC1 cable failure at GyM.
112. In response Ms Carss-Frisk pointed out that Ofgem had been aware of the different circumstances and had taken them into account, as explained by AK in his first statement at para 76. The critical point, she said, was that the cable failures at both Thanet and GyM were examples of a new form of reaction between FOC and power core sheaths causing a power core failure, thought to be occasioned by a manufacturing fault. Mr Grodzinski responded that the impact upon Codan of the endemic problems identified in the Thanet cable, quite separately from the fibre optic cable/core sheath interaction fault, was simply not referred to by AK in his statement, where the only point of difference identified (in parentheses) was that "the cable type is different between the two OFTOs". This entirely missed the point, Mr Grodzinski argued, which was that the damage to the cable at Thanet was of a completely different order.
113. Ms Carss-Frisk relied on a further factor, identified by AK at para 78 of his first statement:

"…the GyM OFTO is likely to have been considered to have been particularly vulnerable, because the GyM OFTO cables contained two fibre optic cables (as opposed to the single fibre optic cable in the other OFTOs that had experienced cable failures). As the Willis Report highlighted in relation to the SSEC1 Failure…"the [fibre optic cable] which failed was the 'spare' FOC which was not monitored and hence there would have been no advance warning of a developing fault through loss of fibres". We considered that the lack of visibility of the condition of one of the two fibre optic cables in the remaining two GyM OFTO cables would have posed legitimate concerns to insurers…"

The lack of visibility for the dormant fibre optic cable, Ms Carss-Frisk argued, rendered the GyM OFTO cables riskier, with the potential for damage occasioned by the identified electrochemical reaction developing unseen.

114. As to this last point, I agree with Mr Grodzinski that it was illogical to use a matter specific to the GyM cables to support the use of Thanet as a comparable. In any event, however, the point itself is ill-founded as AK's evidence in the above passage involves a mis-reading of the Willis report to which he refers. The Willis Report noted that the interaction with the spare fibre optic cable occurred in the SSEC2 cable, suggesting that the "visibility issue" only became apparent as a problem once the SSEC2 failure had occurred. That being the case, the presence of an additional fibre optic cable in the SSEC cables at GyM was irrelevant, since the Hypothetical Question was concerned only with matters which would have been known to insurers within the 23-day window in September 2015, prior to the SSEC2 cable failure.
115. Mr Fordham developed another point arising from the presence of two fibre optic cables in the GyM cables, suggesting that the inclusion of two such cables, as opposed to one, created twice the opportunity for the type of electrochemical interaction between fibre optic cable sheath and power core, and that this potential for "double trouble" at GyM OFTO was a further reason for hypothetical insurers to have been sufficiently concerned to seek a speedy exclusion. Ingenious as Mr Fordham's "double trouble" point was, it was clearly not a matter of concern to any party or to Ofgem at the time: it was not mentioned in any of the submissions made to Ofgem, or in the SSEC2 Decision, or anywhere in AK's evidence.
116. I conclude that the absence from the SSEC2 Decision of any analysis of the very much more serious and extensive issues discovered upon examination of the Thanet cable is a material omission from Ofgem's reasoning.

(4) GyM SSEC3 cable "hotspot"

117. At paragraphs 84-85 of AK1 there is a reference to a previous issue with the GyM cables, namely a "hotspot" identified in the SSEC3 cable in 2011 prior to the GyM OFTO taking responsibility for the equipment. At paragraph 85 AK states "[w]e considered that significant defects in two of the four cables was likely to be perceived as high risk that damage may also have occurred to the other cables."
118. Mr Grodzinski submitted that this was an inappropriate matter for Ofgem to have taken into account, as the hotspot had not resulted in any cable failure, nor was there any reason to link the hotspot identified in 2011 with the SSEC1 cable failure in 2015. The GyM OFTO's own evidence provided to Ofgem in December 2015 was that the hotspot had been a one-off installation incident "specific to the field joints" and that thermal imaging "provided confidence that there were no similar defects elsewhere". The hotspot at GyM was an entirely dissimilar issue to that which had occurred at Thanet in 2011, Mr Grodzinski pointed out.
119. By contrast, the 2011 cable issue at Thanet had been attributed by investigating engineers engaged at the time to an interaction between the fibre optic cable and core, caused by a fault during the manufacturing process. Mr Grodzinski submitted that when there was a further similar incident at Thanet in 2015 (even setting aside the wider array of difficulties found with that cable referred to above), it was unsurprising that Codan had been concerned: given the similar nature of the incident at Thanet in 2011, the further fault in 2015 would rightly have been seen as a second incident.
120. Mr Grodzinski contended that there was no parallel to be drawn with the situation at GyM; he directed me to this passage in an internal email dated 11 July 2017 suggesting that ST shared the same view:

"That is one way to read the Thanet example. However it seems to me that it could also be read quite differently. We were told by Codan last week that they made a major insurance payment for repair costs following the first failure event at Thanet, prior to asset transfer, and told by Steve Griffin on 25 June that insurers had now paid out for the second Thanet claim….So the position at Thanet seems to be that under LEG3 policies (first held by Codan with the generator, and then by Codan with the OFTO) the insurers paid for both the first and second failure events, but moved very quickly to make an exclusion to the policy after the technical examination for the second failure. So the Thanet example could equally be read as implying that a LEG3 insurer at GYM might only have made an exclusion to the policy after the second failure had occurred."

121. Ms Carss-Frisk responded that Ofgem was rationally entitled to consider that the GyM OFTO had experienced a cable failure against the background of prior cable issues, in the most general sense. Ofgem had never suggested that there were no differences between the previous occasions of cable defects at Thanet and at GyM, she pointed out. Mr Fordham supported this, drawing my attention to the fact that at renewal, one of the GyM OFTO's insurers, AXA, had approached cover on the basis that there had been problems with 3 of the 4 cables (ie including the cable affected by the hotspot).
122. On this point I agree with Ms Carss-Frisk and Mr Fordham that there was no error in Ofgem's reasoning concerning the treatment of previous cable issues as between the Thanet and GyM OFTOs. Ofgem was entitled, in my view, to regard "previous cable issues" in general as a matter that would have raised (hypothetical) insurers' concerns at GyM in the same way as had occurred at Thanet.

(5) Codan - a common insurer at Thanet and GyM

123. One of the factors which Ofgem relied on in regarding Thanet as its favoured comparable was the fact that the Thanet and GyM OFTOs had a common insurer, Codan.
124. The Claimant says that Ofgem erred in failing to consider that whilst Codan was the sole insurer on the LEG3 cover obtained by the Thanet OFTO, it was only one of a pool of insurers underwriting the GyM OFTO's LEG2 cover. Codan had a 4.3% share of the GyM OFTO's LEG2 cover.
125. Mr Grodzinski submitted that there were two important aspects arising from this difference: first, other members of a (hypothetical) pool providing LEG3 cover for the GyM OFTO may not have reacted in the same way as Codan did at Thanet. In one of the other OFTO examples available to Ofgem only one member of a pool of insurers had sought any change prior to renewal (at Greater Gabbard, see further below). Second, even if each of the other members of the (hypothetical) pool had wanted to secure an exclusion, the process of negotiation and agreement would invariably have taken longer than for a single insurer writing 100% of the risk.
126. The importance of such considerations was identified in a report dated March 2016 prepared by the GyM OFTO's brokers, Willis, and submitted to Ofgem for the purposes of the SSEC2 Decision. The report included the following passage:

"It is common practice for policies in this sector to be "scheduled", meaning that a number of insurers provide a specified percentage of the cover totalling 100%...for some time the largest lead line offered was 15% which would have resulted in an increased number of insurers being required to achieve a 100% placement. Policy amendments and often claims agreements may need to be sought from each participating market and there is no guarantee of a uniform approach in any matter. "

127. At paragraph 82 of his first statement AK refers to the fact that the GyM OFTO's insurance was written by a pool, going on to say that "we could not know (and did not consider it necessary to speculate) on the identity of the insurance pool in the hypothetical where GyM OFTO had LEG3 insurance..".
128. I accept that this evidence indicates a degree of caution arising from the circumstance of the GyM OFTO's LEG2 cover being written by a pool. I also accept that it was not irrational for Ofgem to have had regard to the position of Codan as an insurer on both policies.
129. Nevertheless I agree with Mr Grodzinski that there is an inconsistency on the one hand in relying on the commonality of Codan as an insurer on both policies, whilst at the same time rejecting, as a matter of unnecessary speculation, any need to identify or consider the impact on timing that a pool might have had. This is particularly so where, as the Willis report highlighted, it was common practice for risks in this sector to be written by a pool. Moreover it is evident from the passage in the Willis report set out above that the process of negotiation with multiple insurers was likely to be more cumbersome and time-consuming.
130. In my view Ofgem erred in dismissing such considerations on the basis that it was unnecessary to speculate on the make-up of any pool. It was the fact of a pool with its impact on timing, not the identity of pool members, which mattered.

(6) Mid-term exclusion mechanism

131. This is the criticism incorporating the remains of Ground 2a. The Claimant's case is that Ofgem erred in failing to consider the precise mechanism by which insurers might have achieved a restriction/exclusion to the hypothetical LEG3 cover. The SSEC2 Decision simply recorded "three OFTOs that have had exclusions placed on their insurance directly following the occurrence of FOC issues". There was no further consideration apparent on the face of the SSEC2 Decision given to the means by which such exclusions were achieved, nor how hypothetical LEG3 insurers at GyM might have achieved an exclusion, had they sought one following the SSEC2 cable failure. In his evidence AK said this (at paragraph 92 of AK1):

"..We did not decide on one particular hypothetical legal or factual mechanism, and did not consider it necessary to do so. However we were aware of the ability to impose a contractual exclusion (as per [A OFTO]), and of the ability to threaten removal of cover if an exclusion was not adopted (as per Gabbard OFTO) – these were lawful, factual examples before us of exclusions being imposed. We were also aware, from the Thanet OFTO and Gabbard OFTO examples in particular, of the strong bargaining position that insurers were put in because of the real difficulties that would face an OFTO that had its insurance cancelled."

There is no further analysis in AK's evidence of the mechanisms by which the exclusions were achieved in each of the three examples upon which Ofgem relied.

132. Mr Grodzinski emphasised that it was not his case that Ofgem should have identified each and every clause in the hypothetical LEG3 policy, but that it should have turned its mind to the mechanism by which an exclusion might have been achieved, and the implications for timing, when answering the Hypothetical Question. For Ofgem to rely without more on examples of situations where insurers had sought exclusions without even analysing what mechanism(s) had been used to achieve them was irrational, he argued.
133. Ofgem's response was that it was not irrational to rely on factual examples without engaging in a detailed analysis – taking a "high level" approach, as Ms Carss-Frisk termed it.
134. I have noted, at paragraph 83 above, the email from AK dated 21 July 2017 raising concerns about the ability of an insurer to amend cover mid-term. AK received the following answer to these concerns, in an email from Diane Mailer of Ofgem the same day:

"Thanks Akshay, a good question and one that we can ask JLT when we meet with them…however advice from Thanet, Gabbard and Ormonde is that they have all add (sic) exclusions imposed on their insurance mid-term directly following a cable failure or significant breakdown of fibres in the fibre optic cable."

135. The three examples identified by Ms Mailer were the ones relied on by Ofgem in arriving at an answer to the Hypothetical Question. Mr Grodzinski drew my attention to the circumstances in which policy changes were made in each case.

(i) Thanet

136. I have noted above the chronology and the process by which Codan obtained an exclusion to the LEG3 cover at Thanet. I take note of the fact that although AK refers in his witness statement to "the strong bargaining position" of insurers at Thanet, the SSEC2 Decision itself contains no reference to any negotiation under threat of cancellation. Instead, paragraph 61 of the SSEC2 Decision simply records that "one week after the technical examination..a LEG1 defects exclusion was imposed on the cable with immediate effect". I have already given my reasons for concluding that there was a material error in this reasoning.

(ii) Greater Gabbard

137. A chronology of insurance developments at Greater Gabbard was provided to Ofgem under cover of a letter from the Greater Gabbard OFTO dated 18 May 2016. As appears from this letter, on 12 February 2016 Greater Gabbard OFTO made "full disclosure of the Fibre Optic Cable losses to its insurers". Engineers Edif ERA clearly regarded the failure of optical fibres as a precursor to a power core failure of the kind seen at GyM.
138. One of the pool of insurers, Ark, with 12% of the risk, sought an exclusion, which was negotiated and ultimately agreed for its 12% on 22 April 2016. An email to Greater Gabbard OFTO dated 15 April 2016 from brokers Willis suggested that Ark had taken the view that there had been non-disclosure to it of fibre optic issues the previous November.
139. Insurers of the remaining 88% did not seek to negotiate or agree any change to the Greater Gabbard cover until renewal, some months later.
140. This chronology of events at Greater Gabbard is quite inconsistent with the picture as recorded at paragraph 61 of the SSEC2 Decision. At paragraph 45 of AK1, AK relies on a letter from Greater Gabbard to Ofgem dated 5 April 2017 referring to an "immediate exclusion of cover by one of the panel of insurers…", but as the above chronology demonstrates, the exclusion was very far from "immediate", when considered in the context of a 23-day window.

(iii) A OFTO

141. The policy covering A OFTO incorporated a clause which permitted insurers to amend cover in the event of claims exceeding a certain proportion of the value of the premium. It seems that, following an incident on 3-4 February 2017, an exclusion was applied with effect from the annual renewal date on 1 April 2017.
142. Ms Carss-Frisk submitted that as this was a long-term policy, albeit with annual renewal dates, the amendment was effectively a mid-term exclusion; Mr Grodzinski responded that an alteration (i) pursuant to a contractual entitlement and (ii) adopted from renewal, 2 months after notification of the incident, could not in any meaningful sense be described or viewed as a mid-term exclusion.
143. In my view the circumstances in which the amendment came to be made to A OFTO's cover are not properly to be characterised as "imposed…directly following.." as Ms Mailer described it in her email to AK, or as "placed.. directly following..", being the description used by Ofgem at paragraph 61 of the SSEC2 Decision. I agree with Mr Grodzinski that a term introduced at an annual renewal date, albeit during the currency of a long-term agreement, is not the same as a mid-term amendment; but in any event, even if it is so regarded, the timing was well in excess of 23 days.
144. In my judgment these distinctions were of critical importance when evaluating whether or not hypothetical insurers of the GyM OFTO would have managed to negotiate and achieve an exclusion within 23 days.

B OFTO

145. At the hearing Mr Grodzinski referred to a fourth example, not cited in the SSEC2 Decision but referred to in the Grounds of Defence and in AK1. In the case of the B OFTO, Codan was the sole insurer on a LEG3 policy. There was a cable failure in 2016 which was found to be due to the same electrochemical reaction between fibre optic sheath and core seen at Thanet and at GyM the year before. Nevertheless Codan waited until renewal over a year later in 2017 to amend the policy. The GyM OFTO's explanation for this given to Ofgem in 2017 was that Codan regarded the failure of the cable as a one-off, also that the cables at B OFTO had the benefit of a full five-year warranty.
146. Mr Grodzinski relied on B OFTO as a yet further example of a cable failure where the insurance had not been changed mid-term, still less immediately following discovery of the fault.
147. On Ms Carss-Frisk's high-level approach there would be force in this point, but upon analysis of the particular circumstances under which Codan was prepared to wait until renewal then in my view it was not unreasonable or irrational to disregard B OFTO when considering the Hypothetical Question.

(7) The JLT Advice

148. The final criticism levelled by the Claimant under Ground 2 concerns Ofgem's treatment of a report from independent brokers JLT. Ofgem had commissioned JLT to provide advice to assist it in arriving at an answer to the Hypothetical Question. The report is dated August 2017. Mr Grodzinski drew my attention to two passages in particular:

(at p.4-5)

"Within normal market practices it is not necessarily the case that an OFTO would not be able to gain LEG 3 cover even after a loss. In some cases, dependent on the loss the LEG 3 cover may be amended to be subject to a sub-limit. Moreover, further evidence may be required to prove that mitigation has been put in place to prevent a reoccurring event…

Finally, with regards to LEG 3 cover being available following a failure event. It is our opinion that this is possible, whether at full limit or sub-limited…"

(at p.8)

"Similar to the above, it may be the case that the incumbent insurers who have suffered losses and paid a claim on a cable loss will be unwilling to continue providing LEG 3 cover post loss at a comparative pricing level to expiry. This would likely only be restricted at renewal where insurers may look to increase premium to maintain LEG 3 cover and/or to invoke sub limits. However, this is not to say that following a cable failure that the OFTO would not be able to obtain full LEG 3 insurance cover." (emphasis added)

149. There is no mention of the JLT report in the SSEC2 Decision itself. AK deals with Ofgem's treatment of the JLT advice at paragraph 72 of AK1:

"..we found that the advice provided by JLT did not materially alter or assist Ofgem's understanding of the relevant position. In Ofgem's meetings with JLT, and to a certain extent in its written advice, JLT emphasised the fact-specific nature of our inquiries."

In the same paragraph, AK quotes from contemporaneous views expressed by ST, in an email to KK dated 30 August 2017:

"JLT refused to be drawn on the question of whether the hypothetical LEG3 insurer at GYM might have withdrawn LEG3 cover before the second failure occurred…"

150. Mr Grodzinski submitted that Ofgem's decision, essentially to put the JLT report to one side, was simply irrational, given the clear advice it contained. He pointed out that, contrary to the view expressed by ST in the above email, JLT had not refused to be drawn on the Hypothetical Question: when asked it in correspondence JLT had specifically referred Ofgem to the passages at pp.5 and 8 of its report set out above.
151. Why, Mr Grodzinski asked rhetorically, would Ofgem dismiss the JLT report as of no assistance whilst relying, as AK says he did (paragraph 89 of AK1), on information from Codan given at a presentation in Berlin in June 2017? Codan gave no specific information at that presentation about the possibility of a mid-term exclusion being imposed, let alone about the likelihood (or otherwise) of insurers seeking to take such a step in September 2015. Moreover the Codan presentation was dealing with concerns which insurers had about cable issues in 2017, some two years after the time in 2015 with which the Hypothetical Question was concerned.
152. Ms Carss-Frisk, responding, pointed out that the Codan presentation in June 2017 spoke of a trend in cable failures occurring over the previous 2 years, ie from June 2015. As to JLT, she emphasised that although they had been asked in correspondence and at meetings to engage with the specific facts relating to the cable failures at GyM, JLT had consistently failed to do so.
153. In my view it was irrational to ignore the JLT report. In the first place, JLT were brokers with very wide experience of the OFTO market and the insurance issues which Ofgem was considering. Ofgem, as the email from ST referred to above pointed out "[did] not have any expertise [them]selves", accordingly it made sense to seek assistance from a broker with experience in the particular market. Second, JLT were independent, having no connection with, or interest in, insurances at GyM OFTO, or at Thanet or Greater Gabbard.

Discussion and conclusions

154. The Claimant's case on Ground 2 is that, for each of the above reasons, taken alone or in any combination, the SSEC2 Decision cannot stand. Mr Grodzinski's case is that the answer to the Hypothetical Question was itself irrational and/or the decision-making process was so flawed as to undermine the result.
155. Although in her submissions Ms Carss-Frisk carefully and helpfully tackled each of the specific criticisms made by the Claimant and discussed above, her overriding argument was that this was an illegitimate approach. In drawing out and attacking individual strands of the determination, she said, the Claimant was in effect inviting the court to re-make the decision. She pointed to AK's evidence as to the extent of consultation and enquiry Ofgem had had, both internally and externally, with a variety of different people over many months. The Hypothetical Question was given extensive consideration and was subject to lengthy internal debate and review. Ofgem had initially been of the view that neither cable failure qualified as an IAE but in the process of consultation and review it had changed its mind in relation to the later SSEC2 failure. Standing back and considering that process, she asked, can it really be right to conclude that no reasonable decision-maker would have arrived at the answer to the Hypothetical Question which Ofgem did here?
156. Ms Carss-Frisk and Mr Fordham both emphasised the value which Ofgem placed on actual examples of insurers' reactions to cable failures at other sites. However, I agree with Mr Grodzinski that, in what was an evaluative exercise, Ofgem acted unreasonably in taking a "high-level" approach to the examples at Thanet, Greater Gabbard and A OFTO. Arriving at an answer to what, on the balance of probabilities, hypothetical insurers providing LEG3 cover at GyM would (a) have sought to do, never mind (b) achieved by way of mid-term exclusion in a 23-day window in September 2015 in my view demanded a more in-depth analysis of the examples used for comparison than Ofgem engaged in. It was Ofgem's duty to grapple with the material provided to it, examining the detail of the circumstances of insurers' reactions at each of the example sites, on account of the critical importance to its decision of the timing.
157. It can be demonstrated in this way: in inviting me to stand back and look at the matter in the round, Ms Carss-Frisk submitted that of the four actual examples which Ofgem had in front of it when making the SSEC2 Decision, insurers had achieved a mid-term exclusion in three of them. However, as demonstrated above, when each of the three examples is considered in more detail the picture that emerges is a great deal more nuanced, and very different.
158. I did not conclude (as Mr Grodzinski invited me to do) that it was irrational for Ofgem to have used Thanet as a comparator at all, for all the reasons which Ms Carss-Frisk advanced; however, in my view Ofgem materially erred in failing to consider and reflect in its decision all the differences between the position at Thanet and that at GyM OFTO, with the effect on timing that those differences would have entailed.
159. It was necessary to engage with the detail of Thanet and the other examples because the window of opportunity for obtaining any exclusion in the hypothetical situation was very tight – just 23 days; moreover Ofgem was having to consider not just whether and if so what restriction hypothetical LEG3 insurers might have sought to achieve in 23 days, but what it was more probable than not they would have achieved in that time; and in doing so to guard against hindsight, given that the Hypothetical Question required the analysis to focus on the market in 2015.
160. I was invited for these purposes to compare the task which Ofgem set itself in answering the Hypothetical Question with a situation where an issue of continuing cover would fall for consideration in the context of an action against brokers for a negligent failure to advise. A causation defence run by such brokers to the effect that any LEG3 cover obtained after non-negligent advice would have been restricted/excluded in the 23 days before the second cable failure would have been subjected to very detailed scrutiny. Examples of what insurers had done in response to cable failures at other places would perforce have been picked over very carefully.
161. In my view this comparison affords a useful insight. Whilst Ofgem is not, of course, required to adopt the same approach to evidence as a court, the Hypothetical Question which it set itself, incorporating a balance of probability evaluation, in my judgment required a much more considered and analytical approach to the examples before it than Ofgem adopted.
162. Further, the "high level" approach to the examples, as described by Ms Carss-Frisk, was in my view inconsistent with rejecting the JLT advice on the basis that JLT had not engaged with the detail. Even if it is accepted that JLT did not provide detailed answers to specific questions about the position at GyM, the general advice which JLT provided was clear: LEG3 cover would have been available in 2015 and any exclusions would not have been imposed until renewal. That was relevant, authoritative and independent advice which Ofgem should have taken into account when making a decision on the Hypothetical Question. It was irrational to leave it out of account.

Conclusion

163. I have considered Ms Carss-Frisk's submission that the SSEC2 Decision would have been the same, notwithstanding any error of fact or analysis. She relied in this respect on Caroopen v Secretary of State for the Home Department [2017] 1 WLR 2339. I am quite unable to accept however, that the answer to the Hypothetical Question would inevitably have been the same, upon a proper analysis of the information which Ofgem had.
164. It is unnecessary to determine whether, and if so which of, the individual complaints made under Ground 2 and discussed above would afford a stand-alone public law ground for vitiating Ofgem's decision. Taken altogether, I am quite satisfied that the affirmative answer to the Hypothetical Question, and thus the SSEC2 Decision, must be set aside.
165. In those circumstances the proper order is to quash the SSEC2 Decision and to remit it back to Ofgem for re-consideration.

Note 1   OFTO is an acronym for Offshore Transmission Owner.     [Back]

Note 2   The tender process relevant to this case was provided for by the Electricity (Competitive Tenders for Offshore Transmission Licences) Regulations 2010, SI 2010/1903, reg 30(1).    [Back]

Note 3   The identity of this and another OFTO (both non-parties) are anonymised in this judgment at Ofgem’s request.     [Back]