Castle Water Ltd v Thames Water Utilities Ltd

This judgment arises out of disputes about disclosure that were heard remotely at the second CMC in these proceedings on 19 and 20 May 2020.

On 1 April 2017 a degree of liberalisation of the non-household water market in England occurred with the opening of the Non-Household Retail Market. With an eye to that liberalisation, the Claimant purchased the Defendant’s non-household water and sewerage retail business by two agreements executed on 18 July 2016, which are known as “the Transfer Agreement” and “the Operating Agreement”. The consideration paid by the Claimant for the Defendant’s business was £100 million. The Agreements provided for data about the workings and customer base of the business to be migrated from the Defendant to the Claimant. The Agreements are suitably long and complicated and do not need to be set out extensively here.

Disputes have arisen which have led to proceedings involving claim and counterclaim, the details of which do not matter for present purposes save that the Court was told that both claim and counterclaim are in the region of £40 million. The parties have engaged representation of the highest order and are investing considerable resources in the prosecution of the litigation.

The parties have exchanged long pleadings. As a result of other orders made at the second CMC there will be a split trial with Stage 1 covering all issues going to the Defendant’s alleged breach, including issues of principle about recoverability of loss. If there is a Stage 2, it will deal with factual causation and proof of loss, and issues specific to the Claimant that go to reduce recoverable loss (i.e. contributory negligence and failure to mitigate).

The parties have produced a List of Issues for the Stage 1 Trial. Separately, and as required by PD51U, they have collaborated to produce a List of Issues for Disclosure. The List of Issues for Disclosure is largely agreed, as are the questions to be asked where Model C is adopted. The discussions between the parties have, however, left some issues that have required decision by the Court. Hence this judgment.

In general terms, the disputes that have not been resolved by agreement have arisen because the Claimant has taken a more expansive view of the obligations that arise under PD51U than has the Defendant. It is therefore necessary to review the applicable principles before turning to the detail of the disputes.

The principles emerge with reasonable clarity from the terms of PD51U itself. They have been helpfully clarified by two judgments of the Chancellor (UTB LLC v Sheffield United Limited [2019] EWHC 914 (Ch) at [75]-[79] and McParland & Partners Limited v Whitehead [2020] EWHC 298 (Ch) at [3]-[4], [44]-[54], [58]) and one (on slightly different points from those arising here) by Falk J (Brearley v Higgs & Sons [2020] EWHC 376 (Ch)). It is therefore not necessary to set out the terms of the practice direction or reprint extensively what has been said by others already. I bear in mind the passages cited above and merely highlight some of the points that are

The one area where there is as yet an absence of authoritative clarification and which may not be common ground is that of “adverse” and “known adverse” documents. A party is under an obligation once proceedings are commenced against it to disclose known adverse documents, regardless of any order for disclosure made, unless they are privileged: paragraph 3.1(2). This is described in the Practice Direction as a continuing obligation. The obligation arises under Models A and B, and is expressly mentioned as applying to known adverse documents “arising from a search directed by the Court” under Models C, D and E.

Paragraphs 2.7, 2.8 and 2.9 of the Practice Direction lay down that “a document is “adverse” if it or any information it contains contradicts or materially damages the disclosing party’s contention or version of events on an issue in dispute, or supports the contention or version of events of an opposing party on an issue in dispute”. “Known adverse documents” are documents that a party is “actually aware (without undertaking any further search for documents than it has already undertaken or caused to be undertaken) both (a) are or were previously within its control and (b) are adverse.” For this purpose “a company or organisation is “aware” if any person with accountability or responsibility within the company or organisation for the events or the circumstances which are the subject of the case, or for the conduct of the proceedings, is aware.”

The question then arises what the obligation of a party may be to discover whether it has any “known adverse” documents that must be disclosed. Paragraph 2.9 states that “for this purpose it is also necessary to take reasonable steps to check the position with any person who has had such accountability or responsibility but who has since left the company or organisation.” This provision, taken in conjunction with the fact that there needs to be a degree of assurance that adverse documents will not simply be ignored or buried, leads to the conclusion that a party is obliged to take reasonable steps to check whether it has any known adverse documents. The Practice Direction gives no guidance on what has to be done to amount to “reasonable steps to check” and the specific steps to be taken will be fact and context sensitive. However, it may be asserted with some confidence that, in a case of any complexity at all or an organisation of any size, reasonable steps to check whether a company or organisation has “known adverse documents” will require more than a generalised question that fails to identify the issues to which the question and any adverse documents may relate. Similarly, it will not be sufficient simply to ask questions of the leaders or controlling mind of an organisation, unless the issue in question is irrelevant to others.

There is a clear distinction between carrying out checks and carrying out searches. A known adverse document is one of which a party is aware without undertaking any further search for documents: see paragraph 2.8. However, the requirement to disclose known adverse documents would be emasculated if there was no obligation at all to look for adverse documents of which the party is aware. Paragraph 3.4 states that “where there is a known adverse document but it has not been located, the duty to disclose the document is met by that fact being disclosed, subject to any further order that the court may make.” To take an example cited by counsel during argument, it would be absurd if a party were able to say “I know I have an adverse document, but I don’t know whether it is in the left-hand drawer or the right. I have therefore not located it.”

Adopting the Practice Direction’s touchstone of what is “reasonable and proportionate” I would hold that a party must undertake reasonable and proportionate checks to see if it has or has had known adverse documents and that, if it has or has had known adverse documents, it must undertake reasonable and proportionate steps to locate them. Any other conclusion seems to me to be a rogue’s charter, which is not the intended purpose or function of the Practice Direction. If, however, the provisions about known adverse documents are operated in the manner I have suggested, the other party and the court should have some assurance that the most significant adverse documents are likely to be disclosed as a matter of routine without an order for Extended Disclosure.

Finally, the question arises: what is meant by the reference to the obligation to disclose known adverse documents being a “continuing” obligation? A clue is to be found in the references to the obligation to disclose known adverse documents arising from the search directed by the Court under Models C, D and E. Provided that the party makes appropriate checks when proceedings have been commenced against it and the case is formulated by reference to the pleadings, that should cause the party to discharge its initial obligation to disclose known adverse documents. The Practice Direction does not say or imply that, if nothing in the litigation context changes, a party is then obliged to revisit the question and renew its checks on a continuing basis (whether periodically or otherwise); and to interpret the “continuing” duty as implying such an obligation would be unduly onerous. However, circumstances might change (for example a party’s case might shift materially) which could raise the need for the other party to review whether it has documents that were not previously known adverse documents but now are; and the Practice Direction makes clear that if a party becomes aware of adverse documents after its initial checks and disclosure, it is under an obligation to disclose them, including where that awareness arises in the course of searches directed by the Court. This approach, which reflects the traditional approach to the continuing obligation of a party to give disclosure of relevant documents of which it is or becomes aware, is both reasonable and proportionate.

The issues that are the subject of dispute are set out in the table below. Where there is a dispute about the wording of the Issue, the disputed passage appears in italics.

The parties agree that this Issue for Disclosure goes to the factual matrix relevant to the formation of their contracts. The Issue for Trial to which this Issue for Disclosure is linked emerges from paragraph 2.1.9 of the Amended Particulars of Claim, which alleges that it was an implied term of the Transfer Agreement that the information transferred by the Defendant to the Claimant pursuant to that agreement would be “reasonably complete and reasonably accurate and/or that the Defendant would take reasonable care to ensure that it was complete and accurate.” The basis for this is said by Mr Kitchener QC in oral submissions to be the commercial necessity test for the implication of implied terms. The Claimant goes on to plead that whether the information transferred from the Defendant to the Claimant was reasonably complete and reasonably accurate is to be assessed by reference to industry standards and practices at the Transfer Date. It also pleads that the industry standard for reasonable completeness and reasonable accuracy was that the data in question would have no (alternatively no material or more than de minimis) errors. It is therefore apparent that the Claimant maintains that there were industry standards and practices about the completeness and accuracy of data which form part of the factual matrix for the parties’ contracts.

The Defendant, at paragraphs 11-28 of the Amended Defence, sets out its case on the factual, legal and regulatory matrix. With one exception, it pleads matters of public record, the one exception being a statement made by the Claimant’s CEO and published on the Claimant’s website about the likelihood of errors occurring in data. It does not plead any published or formal standards (as commonly understood) about data accuracy, and it later denies that any standards existed; but it alleges at paragraph 29(5) of the Amended Defence that the relevant benchmark adopted across the industry for data accuracy was “only that the quantity and quality should be good enough to allow Market Opening”. There is, accordingly, a wide gap between the positions for which the Claimant and the Defendant respectively contend.

The Claimant proposes Model C; the Defendant proposes Model B.

The Defendant emphasises that “factual matrix” has a specific meaning in this context, namely the background knowledge which would reasonably have been available to the parties in the situation in which they found themselves at the time of the contract. By reference to the present formulation of Issue 1 it submits that the only documents that could be relevant are either those that were shared between the parties or that were reflected in public information, such as published standards.

Since it is common ground that Issue 1 goes to the factual matrix, the use of the word “shared” is potentially misleading. The factual matrix is not dependent upon information being shared between contracting parties but is to be identified by reference to an objective assessment of what was reasonably available to the parties, whether or not they individually availed themselves of the knowledge or shared the information between them. That said, documents evidencing the actual knowledge of one party could, at least in theory, be of evidential value when deciding what information was reasonably available to both.

To focus the issue more closely on the requirements for inclusion in the factual matrix, Issue 1 should be reformulated as:

“What background knowledge would reasonably have been available to the parties in the situation in which they found themselves at the time of the contract about the expectations that existed within the industry at the time of entering the Transfer Agreement and the Operating Agreement as to the completeness and accuracy of the types of customer data that were to be transferred to CWL?”

The question then arises, should Extended Disclosure be given and, if so, has the Claimant shown that Model C and its requests are appropriate? The Defendant submits that nothing beyond Model B is required because (a) if there were relevant expectations within the water industry they would be reflected in published information such as public standards, which each party will be able to access and disclose the standards upon which they rely, and (b) if documents were shared between the parties, they have them already. The second objection falls because of the adjustment to the wording of Issue 1, removing the need for “sharing”. Turning to the first objection, the Claimant submits that the Defendant, as a major player in the water industry over the years, is likely to have in its possession documents that evidence the existence of generally accepted standards or practices other than published standards such as those to which the Defendant refers. The Claimant submits that, although not a nationwide monopoly before market liberalisation, the Defendant was in a monopolistic position over a substantial geographic and demographic area that makes it inherently likely that it would have engaged in internal discussions as well as in discussions with public bodies such as Ofwat or external industry consultants, among others, about what standards of data accuracy were generally accepted practice within the water industry. It points to the importance of this Issue for Disclosure for major parts of its case.

Amplifying these submissions orally, Mr Kitchener submitted that the question to be answered is whether the Defendant was likely to have in its possession documents that may be adverse to its case. The example he chose was of a document that would be adverse to evidence to be given by a witness on behalf of the Defendant. His assertion in submissions was that the question of data accuracy was a “huge” issue in the industry at the time and that therefore internal debates and communications with Ofwat are both likely to exist and to be contemporaneous documents that are required in order to determine fairly whether there were in fact relevant practices and standards.

Although these submissions are not supported by a witness statement made for the purposes of this application, the Claimant’s pleadings are pleaded by Counsel who are well aware of the need for sufficient evidence to justify pleading contentious matters of fact, and they are supported by Statements of Truth. Furthermore, it is apparent from the competing Statements of Case that the scope of the factual matrix, with particular reference to whether there were accepted standards of (or practices relating to) data accuracy is both highly contentious and a major issue in the litigation. If, as I am told by leading counsel, it was a major issue for the industry as well, it will have generated documentation. That being so, it seems highly likely that the Defendant would have documents relating to such practices, as it would have required to be knowledgeable about them in the normal course of its business. Furthermore, I do not accept the Defendant’s submission that the documents it (or the Claimant) is likely to have in its possession are limited to published standards, or that factual matrix documents are necessarily limited to documents that are common to the parties. Particularly with a major player such as the Defendant, it is inherently likely that it would have been involved in the discussion about and development of such standards or practices, even if the fact that it was not a national monopoly means that it does not have a monopoly in determining or promulgating standards. Its involvement in such debates and developments is likely to have been influential and its documents are likely to be informative about whether there were such generally accepted industry standards and practices and, if so, what they were.

For these reasons I am satisfied that disclosure in relation to Issue 1 is important for the fair resolution of the Issue for Trial that underpins it; and I am satisfied that Extended Disclosure is necessary, reasonable and proportionate. It is easy to envisage that there may be a vast number of documents that broadly relate to the role of standards or practices applying to data accuracy of which only a relatively small number will be truly relevant to this Issue 1. I therefore agree with the Claimant’s submission that Model C is appropriate.

The devil, however, is in the detail – as will often be the case for Model C because of the requirement that parties request “particular documents or narrow classes of documents”. Here, the Claimant’s requests are as follows:

The preliminary paragraph is far too wide to be appropriate for a Model C request, for two main reasons. First, it is not sufficiently focused on the Issue for Disclosure either as originally drafted or as I have decided it should be. Second, it is so broad as to approximate to a Model D search exercise.

Turning to the sub-categories:

In my judgment, at this stage of the proceedings and on the basis of present information, the Claimant has shown that it is reasonable and proportionate for the Defendant to provide Model C disclosure of documents that it is likely to have and which evidence the existence of industry standards, practices or benchmarks that could be relevant to the factual matrix as follows:

The questions that arise under Issue 2 are similar to those for Issue 1 but are focused on the existence of industry standards and practices as at the Transfer Date about what levels of accuracy and completeness of data should be achieved. By paragraph 42(3K)(a) of the Amended Defence, the Defendant denies that there were any relevant industry standards or practices at the Transfer Date. The parties agree that Model C is appropriate, but the Defendant submits that the Claimant’s proposed requests are too broad.

The Claimant requests:

The Defendant’s suggestion is:

The Defendant submits that paragraph 1 of the Claimant’s request is too broad because it captures all documents (e.g. emails) “referring to” published standards

It appears that the first of the Claimant’s requests is intended to go to (published) standards while the second is intended to go to industry practices. By aggregating the requests, clarity is lost; and the Claimant’s request is too broad because of the breadth of types of documents and the broad meaning of “referring to”. On the other hand, it is likely that the Defendant will have important documents over and above any published standards, for reasons explained under Issue 1.

In my judgment, at this stage of the proceedings and on the basis of present information it would be reasonable and proportionate for the Defendant to search for and disclose:

The Defendant wishes to delete the disputed words because there is no pleaded issue about what it considered putting in place or intended to put in place. This is an overly granular approach. The Claimant brings its claims in both contract and tort, alleging that the Defendant was negligent (and/or acted in breach of a contractual duty to exercise reasonable care) in failing to ensure that Customer Data was complete and accurate, as much more particularly set out in the Amended Particulars of Claim. As an Issue for Disclosure it is potentially important to know what the Defendant may have considered putting in place or may have intended to put in place but did not. This is because the fact that the Defendant considered doing something or intended to do something that it subsequently did not do, and the reasons why they did not do it, may inform the question whether their not doing it was negligent or not.

The disputed words should be included.

SPID account information relating to Relevant Customers;

Allowance and abatement information relating to Relevant Customers;

Business Assessed rates and charges information relating to Relevant Customers;

Legacy Debt balances owed by Relevant Customers; and

Tariffs, balances, meters and associations between meters and DPIDs in respect of Trade Effluent accounts relating to Relevant Customers (the “Relevant Categories”)?

SPIDS are unique reference numbers (Supply Point IDentifier numbers) that identify locations where water and/or sewerage services are provided. Paragraphs 1-5 adopt a degree of shorthand to indicate the heads of claim being pursued by the Claimant: they do not need further translation or explanation here.

The parties have agreed the terms of the issue and agree that Model C should apply to the Claimant and Model D to the Defendant. The dispute is about the words “For the avoidance of doubt… appendices to the APOC.” The footnote records that the Claimant agrees this limitation but asserts that the Defendant must disclose any known adverse documents relating to accounts or SPIDs which the Claimant has not expressly identified. The Claimant insists on the inclusion of this proviso because it alleges that it is still discovering inaccurate data some three years after the Transfer Date.

The Defendant resists this proviso for three reasons:

I reject these submissions:

For these reasons, I endorse the Claimant’s proviso as set out in the footnote.

The dispute between the parties is the same as for issue 3, as is the outcome, for the same reasons.

The Operating Agreement required the Defendant to provide to the Claimant information about “Legacy Debt”, which were debts owed by customers for services provided by the Defendant in the period prior to the Transfer Date and which were to be collected by the Claimant and paid to the Defendant after the Transfer Date. To that end, the Defendant was meant to provide the Claimant with an Opening Statement and a draft Closing Statement that would enable the calculation of the

By paragraph 67(b) its Reply, the Claimant responds as follows:

On this state of the issues as disclosed by the Pleadings, the difference in approach between the formulation of Issue 17 proposed by the Claimant and the more limited approach proposed by the Defendant becomes clear. The Claimant bases its formulation on the general joinder of issue about whether or not the Defendant supplied it with reliable information. The Defendant latches onto the words in the numbered sub-paragraphs from the Amended Reply as set out above.

In my judgment the Defendant’s approach is (a) too granular and (b) mischaracterises the nature of the issue between the parties. On a plain reading of the Amended Reply, the numbered sub-paragraphs are stated to be non-exhaustive examples of categories of errors and the issue between the parties remains founded on the Claimant’s assertion that the migration data provided to it in respect of Legacy Debt was unreliable. The Defendant’s knowledge that its data was unreliable is highly material to that issue, which can only be fairly resolved with access to relevant documentation.

I therefore endorse the Claimant’s formulation.

Under the Operating Agreement, the Defendant was responsible for providing meter reading services. The Defendant’s obligation was to act in conformity with “Good Industry Practice”, which was defined in the Operating Agreement as:

“the exercise of that degree of skill, care, prudence, efficiency, diligence, foresight and timeliness as would reasonably and ordinarily be expected to be exercised by a skilled and experienced company engaged in the same type of undertaking under the same or similar circumstances and shall as a minimum be to a standard at least as high as the standard achieved by Thames in respect of the Relevant Customers”.

It is inevitable even in the best regulated circumstances that some meters will not get read: this is called meter skipping. The Claimant claims that, if the Defendant (or its sub-contractors) had acted in conformity with Good Industry Practice, the meter skip rate would have been about 12-13% but that it was in fact between 25% and 35%.

The parties agree the wording of Issue 36 and that Model C disclosure is appropriate. I agree.

The Claimant’s proposed Model C question is:

The Defendant’s counter-proposal is:

This dispute raises similar questions to Issues 1 and 2 above. The Claimant’s formulation is too wide, being akin to the scope of Model D disclosure. At the same time, the Defendant’s restriction to published standards and data is too narrow, particularly in light of the Defendant’s plea that it “is unaware of any published materials setting out the rate of skipped meter readings which would be consistent with Good Industry Practice.”

For reasons similar or identical to those set out in relation to Issues 1 and 2 above, requests of essentially the same scope as I have determined for Issue 2 would be reasonable and proportionate for Issue 36. I do not have the necessary information to know whether any or all of Oxera, PA Consulting, Open Water, Water UK, Ofwat and/or MOSL are the appropriate entities for Issue 36. I therefore direct that the parties liaise about the entities to be included in the request for this Issue. Subject to

I thank both Leading Counsel and their supporting teams for written and oral submissions of consistently high quality.