B & D Clays & Chemicals Limited v Castle Water Limited [2026] EWCA Civ 171

[2026] EWCA Civ 171Case No CA-2024-001829
IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE COUNTY COURT AT BRIGHTON
HIS HONOUR JUDGE SIMPKISS
Case No. H8QZ0D16
Venue Royal Courts of Justice, Strand, London, WC2A 2LLDate 25 February 2026
Before
LORD JUSTICE PETER JACKSONLORD JUSTICE LEWISLORD JUSTICE SNOWDEN
Between
B & D CLAYS & CHEMICALS LIMITEDAppellant/CASTLE WATER LIMITEDRespondent/Claimant
David Parratt KC, Christopher Edwards and Zhen Ye (instructed by Excello Law) for Appellant.Neil Kitchener KC and Jarret Huang (instructed by Addleshaw Goodard) for RespondentHearing Hearing date: 3 December 2025
Approved JudgmentThis judgment was handed down remotely at 10.30am on 25 February 2026 by circulation to the parties or their representatives by e-mail and by release to the National Archives..............................
[1]This appeal concerns the proper interpretation of the Water (Meters) Regulations 1988 (“the Regulations”) governing charges for the supply of water to premises. The core issue is whether water meter readings are definitive evidence of water volume supplied unless the meter is proven to register incorrectly through prescribed statutory tests, or whether other evidence can show that a meter is unreliable.[2]In summary, the respondent, Castle Water Limited (“Castle”) is a water undertaker which supplied water to the appellant, B & D Clays & Chemicals Limited (“Clays”). Castle issued proceedings to recover charges of £28,014.37 for water supplied. Clays contended that one of the meters from which a reading had been taken was defective (even though it had not been found to be registering water supply incorrectly when tested). Clays sought to rely upon the evidence of a jointly instructed expert witness to show that, although the water meter had not been shown to be registering incorrectly, it was defective and so could not be relied upon as accurately recording the amount of water supplied.[3]Deputy District Judge Ginesi (“the deputy district judge”) held that, on a proper interpretation of regulation 8 of the Regulations, even where a water meter was tested and was found to be functioning within the prescribed limits, other evidence could be adduced to show that the water meter was defective. Further, she found on the basis of the joint expert’s comments, and the difference in the daily average water consumption before and after certain dates, that it was more likely than not that the increase in volume of water supplied shown in those readings was due to malfunctions or defects in the water meter.[4]On appeal, HHJ Simpkiss (“the judge”) held that, properly interpreted, the Regulations provided a statutory regime for the regulation of water meters and the use of water meter readings as evidence of the supply of water. Those readings could be relied upon to establish the amount of water supplied unless the meter was shown to be registering incorrectly. It was not, therefore, open to the court to consider other evidence to suggest that the readings were incorrect. Further, the judge found that the deputy district judge had erred in relying on the evidence of the expert in reaching a conclusion that the water meter was defective. The judge allowed the appeal.[5]Clays appeal from the decision of the judge on three grounds, namely:(1) the judge incorrectly interpreted regulation 8 of the Regulations;(2) the judge wrongly overturned the findings of the deputy district judge; and(3) the judge wrongly substituted an incorrect finding rather than remitting the case for further evidence.

THE STATUTORY FRAMEWORK

[6]The Public Utility Transfers and Water Charges Act 1988 (“the 1988 Act”) makes provision, amongst other things, for fixing charges by water undertakers. At the material time, section 2(1) of the 1988 Act provided that a statutory water company “shall have power” amongst other things “to fix such charges for the services performed” as the company thought fit and to recover those charges. Section 2(2) provided that a statutory water company could fix its charges by means of a scheme made under section 31 of the Water Act 1972 or by agreement.[7]Section 5 provided that schedule 1 to the 1988 Act had effect. That Schedule provided powers to enter premises and carry out works consisting in the installation and connection of any meter for use in determining the amount of any charges.[8]Section 5(2) and (3) provided, so far as material, that: “(2) The Secretary of State may by regulations make such provision, supplementing the provisions of this Act, as appears to him to be appropriate with respect to the installation of meters, with respect to the connection, disconnection, use, maintenance, authentication and testing of meters and with respect to any related matters. (3) Without prejudice to the generality of subsection (2) above, regulations under that subsection may – ….. (c) provide for a reading from a meter to be proved in such manner as may be specified in the regulations and for a reading of from a meter to be such evidence as may be so specified of the volume of water supplied to, or of effluent discharged from, any premises; (d) fix the method of determining the amount of the charges to be paid where it appears that a meter has given, or may have given, an incorrect reading; …..” (c) provide for a reading from a meter to be proved in such manner as may be specified in the regulations and for a reading of from a meter to be such evidence as may be so specified of the volume of water supplied to, or of effluent discharged from, any premises; (d) fix the method of determining the amount of the charges to be paid where it appears that a meter has given, or may have given, an incorrect reading;[9]The Regulations were made in the exercise of powers conferred by section 5(2) to (4) of the 1988 Act. Part II of the Regulations deal with the installation and positioning of water meters. Part III is headed “Testing and Reading of Meters”. Regulation 6 deals with testing of meters and provides that:
“(1) The undertakers or any person duly authorised by them may at any time, and shall if so requested in writing by the customer, test the accuracy of any meter which has been installed. (2) A meter shall be tested in accordance with Regulation 6(2) and (4) of the 1988 Regulations in order to determine whether it falls within the prescribed limits of error (whether or not it is a meter to which those Regulations apply). (3) The undertakers or any person duly authorised by them may remove a meter from premises where it is necessary to do so for the purpose of testing. (4) The equipment used for testing shall comply with clause 4.6 of BS 5750: Part 3: 1987. (5) The undertakers shall provide the customer with a copy of the results of the test.”
[10]The reference to the 1988 Regulations is a reference to the Measuring Equipment (Cold-water Meters) Regulations 1988. Those regulations set out the method of testing meters (essentially by passing a volume of water through them over a period of time which is sufficient to rotate a pointer on a verification scale). A Schedule to those regulations sets out the prescribed limits of error in the recording of the flowrate range by the meter.[11]Regulation 8, also in part III of the Regulations, is at the heart of the appeal. It provides:
“Method of proof and effect of reading meters 8 (1) Where undertakers fix charges payable by the customer in relation to any premises by reference to volume, a reading from the meter installed in relation to those premises proved in accordance with this regulation shall be evidence of the volume of water supplied to those premises, unless the meter is proved to register incorrectly. (2) A reading from a meter may be proved by producing the certificate of a person duly authorised by the undertakers to read the meter and certify the reading. (3) A meter shall be regarded as registering incorrectly if on being tested it is found to exceed the prescribed limits of error. (4) Except where it is proved to have begun to register incorrectly on some later date, a meter which is found to exceed the prescribed limits of error shall be deemed to have done so since the last occasion but one on which the undertakers or any person duly authorised by them last read the meter for the purpose of ascertaining the volume of water supplied to the premises in relation to which it is installed.”
[12]Regulation 9 deals with adjustment to charges and provides so far as material to this appeal that:
“(1) Subject to paragraphs (2) and (3), an account shall be taken of the sums payable by the undertakers to the customer or by the customer for any period during which the meter has or is deemed to have registered incorrectly; and the balance shall be paid or allowed by the undertakers or paid by the customer, as the case may be. …..”
THE FACTUAL BACKGROUND AND THE JUDGMENTS BELOW The premises and the charges levied

THE FACTUAL BACKGROUND AND THE JUDGMENTS BELOW

[13]Clays is the owner of premises at 10 Wandle Way, Mitcham in Surrey. The premises contain a number of different units which it leases to 21 different commercial or industrial tenants. Water was supplied to the premises by Thames Water until 1 April 2017. Thereafter, Thames Water transferred its non-household (i.e. its commercial, rather than domestic) water undertaking to Castle, who continued to supply water to the premises.[14]All of the units at the premises are serviced by one of two water meters which measures the volume of water supplied. Clays was invoiced for charges in the sum of £28,014.37 for water supplied between 1 April 2020 and 31 March 2021. Those charges were based on the volume of water recorded as being supplied in readings from the two meters. This appeal concerns readings taken from one of those meters, namely meter 04A198064 (“the meter”). Clays disputes the accuracy of the readings from that meter. Clays did not pay the charges.

The claim

[15]Castle issued proceedings for the recovery of the sums said to be owed. In its amended defence, Clays averred that the meter was defective and that the readings it produced were inaccurate. It averred that the first two digits in the meter had slipped or jumped to a higher number and therefore incorrectly recorded the volume of water supplied. The amended defence referred to the fact that the figures calculated on the base of the two readings were considerably higher than that those recorded by previous readings.[16]The parties instructed a joint expert to consider a number of issues. Those included whether the two water meters at the premises were currently and were previously likely to have been functioning as expected, whether an occurrence of either or both water meters malfunctioning or slipping was possible, and whether the meter readings could be relied upon as accurate for the period 2017 to 2021 and continuing.[17]The expert’s report recorded that the two water meters had been removed and tested. Each was operating with the prescribed limits set by the 1988 Regulations. The expert however indicated that there were other situations that could affect a water meter. He said at paragraph 4.3.2 that a malfunction due to general wear and tear could not be ruled out. At paragraph 5.2.3 of his report, he said that digital dials, or the meter itself, could become enveloped with scale deposited over time and that this could interfere with a meter, causing it to advance by two digits instead of one (e.g. to advance by 11, rather than 1). He said that he had tested hundreds of water meters, and had been able to infer that that had happened on several occasions and he had himself observed it happen on one occasion.[18]He noted the increase in water consumption in the disputed period as compared with other periods. He said he had considered three possibilities as to how the volume of water supplied could have increased. These were: an increase in the number of clients at the premises, a failure within the water pipe work which had remained undiscovered for some time, or a single tap or taps left running for a long time. The expert appeared to dismiss the three possibilities he had identified. The first he said was inconsistent with meter readings taken by one of Clays’ directors, Mr Bigos, and one of their employees, Mr Smith. He considered that these showed a stable usage. The second he dismissed, based on what he was told by Mr Smith who said that he had been on site during each working day (including the period which coincided with the Covid pandemic). Mr Smith told the expert that he had seen no such failure. The expert considered that the third possibility would have required one tap to be left running for a considerable period of time or multiple taps being left running for a lesser period. He dismissed this because of what he was told by Mr Smith. The expert’s conclusions are at section six of his report and are in the following terms:
“6.0.1 I confirm that insofar as all the facts stated in my report are within my own knowledge, I have made clear which they are and I believe them to be true and that the opinions I have expressed represent my true and complete professional opinion. 6.0.2 Both Water Meters were tested at the WRc (Water Research Centre) as previously mentioned. Both were within the tolerance band required and therefore passed the accuracy limits. 6.0.3 However it is to be understood that just because the meters passed the accuracy test does not mean they can’t be faulty in other ways. Meter 04A198064 is the Meter that exhibits problematic readings. Meter 99A803647 does not exhibit any issues so far as I can tell from the information provided to me. 6.0.4 Due to the paucity of readings by Castle Water I am limited in my understanding of the history in terms of the average usage of this device. We are unable to discover whether there could have been a huge discrepancy within just one month or a series of smaller continuous discrepancy’s every week, month, or other period. Whatever the type of failure the time scale over which this must have occurred is lengthy. It is equally possible there is a fault with the meter indicating device itself. It is to be noted that Ofwat state “If you have a water meter, it should be read at least once a year and read by your water company at least once every two years. Some water companies might read your meter more frequently. 6.0.5 Should there be a fault with the reading given by Meter 04A198064 there is a possible explanation for this of which I am aware. Whilst previously testing this type of meter at WRc Test Laboratory I witnessed two digits of the seven digits on the meter roll over at the same time, indicating a mechanical device failure on the dials themselves. I have no proof that this has occurred on this Meter at some period in the past. It is doubtful if further long term testing could repeat such an indicator error. 6.0.6 RAH Consultancy are not convinced by the alternative explanations of increased water usage: catastrophic pipe failure or water devices left running as the time period this would have to have taken place would appear to me to be excessive. I do not believe that on the balance of probabilities and given the witness statements from Mr Smith that leaks would have been allowed to remain on-going for such long periods. Professionally they would be a dereliction of his duty of care. 6.0.7 Further to the above, if such a large-scale leak had occurred there is likely there would have been physical evidence on the site, insurance claims submitted and reports of damage to the property and/or the tenant’s items such as stock, tools etc.”
[19]The expert attached the test report for each meter. The condition of the meter in issue (number 04A198064) was recorded as “Good”.[20]The expert was asked a number of questions in writing by both parties. He was asked to confirm that he had not found evidence of general wear and tear in either water meter which had caused a malfunction of the device. He answered that he could not confirm whether such evidence would or would not exist. That would require dismantling of the meters which was something he had not done (as they were the property of Thames Water and dismantling them would require its consent). The expert gave a similar answer to the question as to whether there was any evidence of the meters being enveloped with scale. He was asked by Clays whether the meter readings could be relied upon as an accurate record of water usage at the property and replied:
“The short answer is no. We cannot prove for certain at this stage what has happened however it is fair to say something is not right, and in our opinion, there is something untoward with meter 04198064”
.

The judgment of the deputy district judge

[21]On the question of the proper interpretation of regulation 8 of the Regulations, the deputy district

judge concluded that:

“16. I prefer the interpretation of Regulation 8 put forward by the defendant, namely: that if a water meter fails the statutory test, it is defective, However, if the meter does not fail the statutory test, other evidence can be relied on to prove the meter is defective. This is the ordinary common sense interpretation of the wording used in Regulation 8.”
[22]Consequently, she concluded that:
“19. The effect of my findings in relation to issue 1, is that the fact that the water meters in this case both passed the statutory test is not conclusive evidence in itself, as the claimant has argued, that the water meters were not defective or registering incorrectly. The court may go on and consider if there is other evidence which indicates on the balance of probabilities that the water meters were defective.”
[23]The deputy district judge therefore addressed the issue, as it was put in paragraph 20 of her judgment, of “whether the evidence in this case proves on a balance of probabilities that the water meters were defective and incorrectly recording the water usage”. She noted that she had heard evidence from among others, from Mr Bigos. She noted that a witness statement from Mr Smith had been served on behalf of Clays but he had not attended and his evidence had not been tested. She therefore gave limited weight to his evidence (paragraph 21). She also noted that the court had been referred to expert evidence of the joint expert.[24]The deputy district judge noted that the expert’s evidence was that the meters were operating within the prescribed limits and the accuracy of the meters was not in dispute. She referred to the answer given by the expert set out at paragraph 20 above that the meters could not be relied upon. She noted his evidence that the expert could not rule out the possibility of a malfunction caused by wear and tear or that it was equally possible that there was fault in the meter such as a build-up of scale. She noted that the expert could not say that was the case without examining the meter.[25]The deputy district judge said that the starting point was that there was an increase or spike in the meter readings between July 2018 and July 2020 when the average daily consumption increased and that increase was very large and unusual. She considered that the key to deciding if the water meters were defective was deciding what the likely explanation was for the spike or increase in the recorded consumption. She noted the three most likely explanations put forward by the expert. She said that she agreed with his conclusions and found that these were unlikely to have been the cause of the increase. Her essential conclusions are set out at paragraphs 76 to 78 of her judgment in the following terms. “76. I find that taking into account the joint expert’s comments, and the vast difference in the daily average recorded water consumption before and after the spike (from 3.5 cubic metres to 14.2 cubic metres), on the balance of probabilities, it was more likely than not that the reason for the spike and the increase in readings was down to the malfunction or defects in the water meters. 77. I find as a matter of fact that it is more likely than not that the water meters were defective for a number of possible reasons, whether that be scale, age, wear and tear and/or a fault in the meter indicating device. As the joint expert has said, he cannot prove “for certain” what has happened at this stage, but in his view “something is not right” and there is “something untoward with the meters”. The exact cause of the defect(s) could not be identified without further more invasive testing. However, the burden of proof is not to provide conclusive evidence of the defects, or of which particular defect existed. 78 I am satisfied that the defendant has proven on the balance of probabilities that the increased recorded usage of the meters was due to defects in the meters and that a number of likely defects have been identified.”[26]Consequently, the deputy district judge concluded at paragraph 80 that she was “not satisfied that the meter readings can be relied upon as a correct measurement of the amount of water used, and the basis to substantiate” the claim for charges made. At paragraph 81 she added that, since the claim had been put forward on the basis of the invoices and the claimant had failed to prove that it was entitled to those payments, the claim for water charges would be dismissed. The claimant could choose to recalculate the invoices but at that stage the task had not been undertaken. No other basis for calculation had been put forward so it was not possible for the court to consider what the revised invoices should be.

The judgment of the judge

[27]Castle appealed and the judge allowed its appeal. At paragraph 6, he identified the two issues that arose on the appeal: first, the interpretation of the relevant regulations, and secondly the evidence dealing with Clays’ contention that the meter was defective.[28]On the question of interpretation, the judge held that the Regulations provided a comprehensive regime for charging based on readings from the relevant water meter unless that meter was recording incorrectly in that it did not meet the test identified in Regulation 6. His reasoning and conclusion are set out at paragraphs 37 to 42 in the following terms:
“37. When one looks at regulations 6, 8 and 9 as a whole it is quite clear to me that what Parliament intended was to provide a statutory regime for the regulation of water meters and the use of water meter readings in evidence to recover water charges. Anything short of Mr O’Sullivan’s interpretation would lead to a chaotic situation, as when I come to the second ground, will become apparent. 38. The way in which I read regulation 8 is that the reading on the meter is evidence of the volume of water supplied to those premises and 8.2 provides for the certification of the reading as being evidence that that is the reading which can be relied upon in court. The proviso in 8.1 is that the reading will not be evidence if the meter is proved to register incorrectly. 39. 8.3 is a direct reference back to the testing regime in regulation 6. Regulation 6 tests the accuracy of the meter, not whether it is defective, and in order to do that the tests are prescribed and Mr Hurst has explained all that in his report. That is why 8.3 defines registering incorrectly as a meter which, on being tested, is found to exceed the prescribed limits of error. 40. This is because the only circumstance in which the evidence will not be admissible in court is if it is registering incorrectly. If the meter passes the test then there is no need to prescribe what should happen because that has already been done, the evidence of the volume of water supplied. 41. There was no other evidence of the volume of water supplied produced in this case, at least not in a form which could be used to establish the charges nor was it anticipated there would be because the only evidence is the meter reading. Since the test is designed to establish whether or not the prescribed limits of error have been exceeded, this is a reference direct to regulation 6 and the test prescribed under that regulation. 42. Therefore, it is not open to the Court to consider other evidence that the readings were incorrect, let alone evidence that the meters were defective, which is not the wording used in the regulation. Any interpretation other than the above would lead to a great deal of uncertainty in metered water bills where a party produced evidence that perhaps the premises had been unoccupied.”
[29]In relation to the second issue, the judge noted that the deputy district judge had referred to the expert’s comment that something was untoward with the meter but noted that she did not explain what was untoward with it. He noted that the expert had produced a number of reasons why a meter might not produce a correct reading. In that regard, the judge noted that there was no evidence as to the use of the premises at the material time. Mr Bigos (who gave evidence on behalf of Clays) was not at the premises at the material time. Mr Smith did not give evidence. Secondly, so far as the question of possible leaks, the deputy district judge needed to examine the evidence on that matter not simply accept the expert’s comments. Thirdly, so far as concerned a fault in the meter, the expert had not examined the meters, but sent them for tests. He would have had to dismantle the water meters to identify any possible defect.[30]The judge concluded that he would have allowed the appeal on ground 2, namely that the deputy district judge wrongly decided that Clays had satisfied the burden of proof, if he had not already allowed it on ground 1 (the issue of statutory interpretation). His essential reasoning is at paragraphs 53 and 54 where he said this:
“53. The basis of the Judge’s finding is that there is a spike which, if she accepts the Defendant’s witness evidence, is unexplained and completely out of the normal and that there are various possible reasons why the meter might have given a wrong reading. 54. None of that, in my judgment, is sufficient for a Judge to make a finding that the reading was wrong, particularly given the tests that were carried out following the incident, which showed the statutory prescribed tests which are designed to establish whether or not a meter is reliable.”

GROUND 1 – THE PROPER INTERPRETATION OF THE REGULATIONS

[31]Mr Parratt KC, with Mr Edwards and Dr Ye, who appeared for Clays, submitted in essence that regulation 8(1) of the Regulations provided a means of proving the amount of water usage, namely a reading from a meter (unless that meter was shown to be recording incorrectly). But, he submitted, regulation 8(1) left open the possibility of other evidence being adduced to demonstrate that a meter was in some way defective or that a reading could not be relied upon as an accurate record of the volume of water suppled. That interpretation, he submitted, was reinforced by the fact that the testing of the meter contemplated by regulation 6 would demonstrate whether the meter was registering correctly at the time of testing not whether it was registering incorrectly previously. Other evidence could be admitted to show that. Further, he submitted that regulation 8(4) contemplated that other evidence might be relied upon.[32]Mr Kitchener KC, with Mr Huang, for Castle submitted that the Regulations do provide a comprehensive statutory code governing the circumstances in which meter readings will, and will not, be probative of the volume of water supplied to a customer. That permitted readings to be relied upon unless the meter was shown to be registering incorrectly, i.e. when tested, it was found to have exceeded the prescribed limits of its operation as recognised by regulation 8(3).

Discussion and conclusion

[33]The issue concerns the proper interpretation of regulation 8(1) of the Regulations. That involves considering the words of the statutory provisions, read in context and having regard to the purpose underlying the statute, and bearing in mind any legitimate aids to statutory interpretation. Regulation 8(1) needs to be read in the context of the Regulations as a whole, and the enabling provisions, as that provides the relevant context for ascertaining, objectively, what the words in the legislation mean. See generally, R (O) v Secretary of State for the Home Department [2023] AC 255 at paragraphs 29–31.[34]In the present case, the context is the method of charging consumers for the provision of water. As appears from the material before us, the supply of water was historically charged for by reference to the annual value, sometimes referred to as the rateable value, of the property supplied. The charges were not related to the volume of water supplied to the individual property. As late as the mid-1980s, over 99% of domestic households were charged in this way for water supplied. There was a greater use of meters to measure the volume of water supplied in respect of non-domestic properties. There were concerns at the fairness of charging for water by reference to the value of the property receiving a water supply rather than the volume of water supplied. There were also arguments as to whether charging by reference to volume supplied would provide a greater incentive to reduce the unnecessary use of water.[35]It was in that context that the 1988 Act was enacted. The Act provided a power to enter into premises and install water meters. Section 5 provided power to make regulations with respect to the “connection, disconnection, use, maintenance and authentication and testing of meters”. That power included the making of regulations which, amongst other things, provided(1) for readings from a meter to be proved in such manner as was specified,(2) for a reading from a meter to be “such evidence as may be so specified of the volume of water supplied to… any premises” and(3) for fixing “the method of determining the amount of the charges to be paid where it appears that a meter has given, or may have given an incorrect reading” (see sections 5(2) and (3)(c) and (d) of the 1988 Act).[36]Turning then to the Regulations themselves, Part II deals with the positioning of meters. Regulation 3 provides that meters “shall be so installed as to ensure that they are reasonably accessible for reading, inspection, testing and maintenance”.[37]Part III of the Regulations deals with the testing and reading of meters. Regulation (1) provides for a water company to be able to “test the accuracy of a meter” and must do so if a customer makes a written request. Regulation 6(2) provides for it to be tested to determine if it falls within the prescribed limits of error.[38]Regulation 8(1) then deals, as the heading to that regulation makes clear, with the “Method of proof and effect of meter reading”. Regulation 8(1) applies when water companies fix charges in relation to premises by volume. The regulation essentially then provides that(1) a reading from a meter shall be evidence of the volume of water supplied to the premises(2) if the reading is proved in accordance with regulation 8, and(3) unless the meter is proved to register incorrectly. The second and third of those elements are governed by Regulations 8(2) and (3).[39]Regulation 8(2) provides how a meter reading is to be proved. It may be provided by producing the certificate of a person duly authorised to read the meter and certify the reading.[40]Regulation 8(3) deals with when a meter is proved to register incorrectly. That will occur “if on being tested it is found to exceed the prescribed limits of error”. That is a reference back to regulation 6, which provides for the testing of meters and provides a means of identifying the prescribed limits within which a meter must be operating, if it is not to be regarded as registering incorrectly.[41]In my judgment, read naturally, and in context, Regulation 8 is providing a comprehensive code by which the volume of water supplied to any premises is to be measured and hence, how charges are to be determined. If a properly certified reading is taken from a meter that is found on testing to be registering correctly, the reading shows the volume of water supplied to the premises and is the basis upon which charges may be determined and recovered. Neither regulation 8, nor any other regulation, contemplates that different evidence may be utilised to determine if the meter was functioning correctly or to measure the volume of water supplied in some other way.[42]That interpretation is also reinforced by regulation 8(4). If a meter is tested, and is not registering correctly (i.e. it is operating outside the prescribed limits) provision has to be made for deciding the consequences of that finding. The testing will only show that the meter was not registering correctly on the date when the tests were carried out. Regulation 8(4), therefore, provides that a meter that is found to exceed the prescribed limits of error “shall be deemed to have done so since the last occasion but one” when the meter was tested. That is subject to an exception, namely where the meter is proved to have been registering incorrectly on a later date. The reference to “registering incorrectly” is again a reference to testing in accordance with regulation 6. That exception permits the water company to show that the meter was tested, and was operating within the prescribed limits, at some date later than the start of the two charging periods before the test showing that the meter was not working.[43]In other words, the customer will get the benefit of an assumption that the meter was not working correctly for the last two periods before the testing, (unless the water undertaker can show the meter was tested, and was functioning within the prescribed limits during that time). That reinforces the view that the measurement of the volume of water supplied, and the determination of charges, is to be done on the basis of the reading from the meter unless the meter is not registering correctly, in which case the regulations provide a means of determining the period when the reading cannot be relied upon.[44]This interpretation of Regulation 8, which contemplates that volume will be measured and charges determined by reference to a reading from a meter (unless the meter is tested and shown to be operating outside the prescribed limits of error), accords with the purpose underlying the Regulations. The aim was to move to a system where charges were linked to the volume of water supplied to premises, rather than an annual sum based on the value of the premises, irrespective of the amount of water consumed. The purpose underlying the enabling power in section 5 of the 1988 Act, and the Regulations, was to provide a system by which meters were installed, and the volume of water supplied would be measured by readings from those meters (unless they were not functioning within the prescribed limits of error).[45]There is nothing unjust or arbitrary in that system. It provides a method of measuring the volume of water by reference to meters operating within prescribed limits of error. Customers are given the right to of have the meter tested if they consider that the readings do not reflect their water usage. If the meter is not functioning within prescribed limits when it is tested, the statutory regime strikes a balance and assumes that that is the case for the last two charging periods (unless, during that time, the meter had been tested and had been shown to be functioning within the prescribed limits).[46]The question was raised during the hearing as to the possibility of an intermittent fault occurring, that is, a fault which corrected itself before the meter was tested. The question was whether that might lead to a reading that did not accurately reflect the actual water use even though the meter, when actually tested, would be found to be operating within the prescribed limits (because the fault has corrected itself before testing). There is nothing to indicate that the Regulations when made contemplated the use of evidence other than readings from meters because of a possibility of an intermittent fault. Instead, the statutory regime was constructed on the basis that meters meeting prescribed standards would be installed, that their accuracy would be tested, and that readings from such meters would be used to measure the volume of water supplied. It might be said that the regulations could have been drafted to state expressly that only evidence from a meter functioning within prescribed limits was to be used in measuring volume. It is always possible for primary or secondary legislation to be drafted differently, or for matters to be expressed more clearly. But the task for the court is ascertain objectively the meaning of the language under consideration (see the observation of Lord Nicholls in R v Secretary of State for the Environment, Transport and the Regions, ex p Spath Holme Ltd [2001] 2 AC 249 at page 396.)[47]I have had the benefit of reading the judgment of Peter Jackson LJ. I entirely agree with that judgment.[48]For those reasons, the judge was correct to conclude that it was not open to the court to consider other evidence about whether a meter might be defective in circumstances where, as here, readings were taken from meters which were not shown to be registering incorrectly, that is, from meters which, when tested, were shown to be functioning within the prescribed limits.

GROUNDS 2 AND 3 – THE RELIANCE ON EXPERT EVIDENCE

[49]In the light of the finding on ground 1, the deputy district judge should not have gone on to consider the question of whether there was other evidence which showed that one of the meters was defective and that the readings from that meter could not be relied upon as a basis for measuring the volume of water supplied. Ground 2 of the appeal does not, therefore, arise. For completeness, I note briefly the principal arguments of the parties, and my conclusion, on ground 2 if it had arisen.[50]In brief, Mr Parratt submitted that the judge was wrong to overturn the findings of the deputy district judge. She had found as a fact that there was a large and usual increase in water consumption between July 2018 and July 2020. In those circumstances, she was entitled to rely on the expert’s opinion that there were three likely explanations for this: an increase in users at the property, an undiscovered failure in the water pipe, or a single or multiple taps running, and to accept the expert’s opinion that none of those explanations were likely.[51]Mr Kitchener KC accepted that the principles summarised by Lewison LJ in Volpi v Volpi [2022] EWCA Civ 464; [2022] 4 WLR 48 at paragraphs 2 to 5 set out when an appellate court may overturn a finding of fact by a lower court. He submitted that the trial judge has an obligation to analyse expert evidence, relying on Woolley v Essex County Council [2006] EWCA Civ 752 at paragraph 38. He identified a number of matters where the deputy district judge was at error. These included the fact that there was no evidence for concluding that there was a defect in the meter and, further, that in dismissing the three likely possibilities of any increase in consumption, the expert had relied on what he had been told by a particular witness, Mr Smith. That was factual evidence which was a matter for the court to assess, not the expert, and in the event Mr Smith did not give evidence.[52]The principles governing the circumstances in which an appellate court may reverse a finding of fact by a trial judge are well-established. In essence, I consider that the judge, when hearing the appeal, was entitled to reach the conclusion that he did, essentially for two reasons. First, the evidence did not establish that the readings from the relevant meter were flawed. The meter had been tested and it was functioning within the prescribed limits. Its condition was described as good. The expert said that he could not establish if there were wear and tear or a build-up of scale without dismantling the meter and he had not done that. In those circumstances, the position essentially was that the expert thought that something untoward had happened and, therefore, looked for some explanation for that. In truth, there was no evidential basis for concluding that the meter was defective so that the readings from the meter should not be relied upon.[53]Secondly, in relation to the three possible explanations for any increase in water consumption identified by the expert, he dismissed those largely on the basis of things he was told by others, in particular by Mr Smith, However, it was for the trial judge to make findings of fact to determine whether, on a balance of probabilities, certain things had or had not happened (such as the whether the number of commercial tenants remaining unchanged, whether there had been on failure in the water pipes, or whether taps had or had not been left running). That did not fall within the scope of the expert’s role. Mr Smith had made a witness statement but did not attend and his evidence was not tested. The deputy district judge gave it little weight. In those circumstances, the judge was correct to conclude that, if he had not allowed the appeal on ground 1 (the issue of interpretation), he would have found that the evidence was insufficient for the deputy district judge to make a finding that the reading from the meter was wrong, as explained in paragraph 54 of his judgment.[54]I would not, therefore, have allowed this appeal on ground 2. Ground 3 does not, in the circumstances, arise.

CONCLUSION

[55]I would dismiss this appeal. On a proper interpretation, Regulation 8 provides a code by which the volume of water supplied to any premises may be measured and hence the charges for the supply of water determined. Charges may be recovered in respect of the volume of water recorded in a reading taken from a meter unless that meter is shown to be registering incorrectly (i.e. that, on testing, the meter is shown to be operating outside the prescribed limits of error). Regulation 8 does not enable different evidence to be adduced to determine if a meter was defective or to measure the volume of water supplied in some other way.

LORD JUSTICE SNOWDEN:

[56]I am grateful to Lord Justice Lewis for setting out the background, and I shall adopt the same abbreviations in this judgment as he does. I agree with him that the appeal should be dismissed. However, I have reached that conclusion by a different route.[57]In short, for reasons that I will explain, I do not agree that the Regulations mean that if a water meter is shown by a subsequent test to be operating within the prescribed limits of error, a customer is precluded from adducing any other evidence seeking to show that the volume of water actually supplied to them in an earlier period was different from the amount recorded on the meter. The Regulations do not say that, and there is nothing in the statutory scheme of which they form part to indicate that it was the legislative intent. I would therefore allow the appeal under Ground 1.[58]However, reflecting the obvious point that a customer will face an uphill struggle to show that a meter which has passed a subsequent accuracy test was nonetheless registering inaccurately in an earlier period, I would dismiss the appeal in the instant case on Ground 2. In that regard, I agree with Lord Justice Lewis that the evidence adduced by Clays in this case was insufficient to support the deputy district judge’s conclusion that there must have been something wrong with the meter in this case.

Interpretation of legislation

[59]When interpreting legislation, the Court’s purpose is to determine the true meaning of the words used by the legislature. In that regard, the primary source is the text of the legislation itself, read in context and having regard to its purpose: see e.g. R v Secretary of State ex parte Spath Holme Ltd [2001] 2 AC 349 at 396, and R (O) v Secretary of State for the Home Department [2023] AC 255 at paragraphs [29]–[31].[60]Generally, the courts will assume that if the legislature had meant to provide for something, it would have said so expressly: see e.g. Okedina v Chikale [2019] EWCA Civ 1393 at [45]. However, the meaning of a provision is not limited to the express language used, but will also include anything that can properly be implied from the express language. So, for example, a meaning may be implied because it is compellingly clear from the express words and context: see e.g. Pwr v DPP [2022] UKSC 2 at [34].[61]In interpreting legislation, the court should also have regard to the likely consequences of the rival constructions, and will generally seek to avoid a construction that produces a result that is absurd, unworkable or anomalous: see e.g. Project Blue v HMRC [2018] UKSC 30 at [31].

The legislative scheme

[62]The relevant provisions of the 1988 Act have been set out by Lord Justice Lewis. As he indicates, the purpose of the 1988 Act was to provide for a change in the method by which customers were charged for water, from a scheme based upon the rateable value of the relevant premises, to a scheme based upon the volume of water actually supplied. That new scheme is based upon the use of water meters to measure the volume of water supplied. To that end, section 5(2) of the 1988 Act contains broad enabling powers for the Secretary of State to make detailed regulations with respect to the installation, connection, disconnection, use, maintenance, authentication and testing of water meters and any related matters.[63]One such related matter concerns how a water undertaker might use a meter reading to prove in court proceedings the volume of water supplied to a customer. Section 5(3)(c) specifically provides that regulations promulgated under section 5(2) may, “provide for a reading from a meter to be proved in such manner as may be specified in the regulations and for a reading from a meter to be such evidence as may be so specified of the volume of water supplied to … any premises” (my emphasis) (my emphasis)[64]On its face, section 5(3)(c) contains an explicit invitation to the Secretary of State not only to specify how a meter reading should be proved in court, but also to specify the extent to which a meter reading so proved should be evidence of the volume of water supplied.[65]Against that background, Regulation 8 provides, “Method of proof and effect of reading meters 8(1) Where undertakers fix charges payable by the customer in relation to any premises by reference to volume, a reading from the meter installed in relation to those premises proved in accordance with this regulation shall be evidence of the volume of water supplied to those premises, unless the meter is proved to register incorrectly.(2) A reading from a meter may be proved by producing the certificate of a person duly authorised by the undertakers to read the meter and certify the reading.(3) A meter shall be regarded as registering incorrectly if on being tested it is found to exceed the prescribed limits of error.(4) Except where it is proved to have begun to register incorrectly on some later date, a meter which is found to exceed the prescribed limits of error shall be deemed to have done so since the last occasion but one on which the undertakers or any person duly authorised by them last read the meter for the purpose of ascertaining the volume of water supplied to the premises in relation to which it is installed.”

Analysis

[66]The first point to note is that whilst Regulation 8(1) provides that a meter reading proved by a certificate under Regulation 8(2) “shall be evidence of the volume of water supplied to those premises, unless the meter is proved to register incorrectly”, it does not say that such reading shall be the only admissible evidence of the volume of water supplied. That could easily have been done if, for example, Regulation 8(1) had provided that a meter reading “shall be conclusive evidence of the volume of water supplied, unless the meter is proved to register incorrectly”. Conclusive evidence clauses are a long-established legislative technique, and the omission of such wording from Regulation 8(1) suggests that a meter reading was not meant to be the only evidence by which the volume of water supplied to a customer could be proved.[67]Nor does Regulation 8(1) prescribe exhaustively the manner in which a customer can prove that a meter is registering incorrectly. It does not, for example, say that a meter reading “shall be evidence of the volume of water supplied, unless the meter is proved to register incorrectly on a test conducted in accordance with Regulation 6”. Regulation 8(3) simply provides, in non-exhaustive terms, that a meter “shall be regarded as registering incorrectly if on being tested it is found to exceed the prescribed limits of error”. That wording does not expressly preclude the possibility that the meter might be proved to be registering incorrectly by some other evidence.[68]These points are of particular importance in a situation in which a meter has suffered from an intermittent (or transient) fault. The expert evidence before the deputy district judge was that water meters can suffer from mechanical faults, whether through wear and tear, or from the build-up of scale within the device. This may cause the adjoining dials of the meter to advance by two digits instead of one – i.e. to advance by 11 rather than 1, 22 rather than 2 and so on. Faults of this type could be intermittent – e.g. scale between the dials could cause an inaccurate reading as the dials move, but then be subsequently dislodged, so that the meter would read correctly again. Whilst such intermittent faults might be rare, Mr. Kitchener KC did not dispute that they could occur.[69]In this regard, consider the position of a customer who, for no apparent extraneous reason (e.g. a leak or tap left running), suddenly receives a quarterly bill for ten times the normal amount. Assume that the meter was subsequently tested and found to be measuring within limits, but was then opened and more closely inspected, and found to be heavily scaled. In such a case it would be entirely plausible that the sudden spike in measured water usage was due to an intermittent fault with the meter of the type that I have described.[70]However, even if the consensus of expert opinion in such a case was that an intermittent or transient fault with the meter was the probable cause of the aberrant reading, the consequence of Castle’s argument, as Mr. Kitchener KC candidly accepted, was that the customer would not be permitted to adduce such evidence and hence would be prevented from challenging the vastly increased bill. Indeed, the logic of this argument would be that no matter how high the meter reading might be - even if so high that it could not sensibly be thought that the volume of water measured could have been supplied in the relevant period - the same result would necessarily follow and the customer would be liable to pay the bill, no matter how exorbitant and ruinous that might be.[71]Such an interpretation of the Regulations would be absurd and manifestly unjust. I do not consider that such an extreme intention should be attributed to the legislature unless it was necessary to make the regime under the 1988 Act of charging by the volume of water supplied, rather than by rateable value, workable. But in my judgment it is not.[72]In the vast majority of cases, the simple regime of proof of a meter reading, and subsequent testing of the meter, will be all that is necessary to determine the amount that a customer has been properly charged. Intermittent faults will be rare. Permitting other evidence to be adduced in a very small number of cases to show that the volume of water supplied was not that shown by a meter that passed a subsequent test will not undermine the proper operation of the Regulations.[73]Nor will it “open the floodgates” to unmeritorious challenges by customers as Mr. Kitchener KC suggested. There was no evidence to suggest that significant numbers of customers would readily, or without good reason, embark upon the costly exercise of obtaining expert evidence to challenge the reading from a meter that has passed a test. And if a challenge based upon such evidence were to fail, then the water company can be awarded its additional costs in the proceedings in the usual way.[74]I would add that this conclusion would in no way be inconsistent with the provisions of Regulation 9. Regulation 9(1) provides for an account to be taken of the sums payable by the water undertaker or by the customer “for any period during which the meter has or is deemed to have registered incorrectly”. Regulation 9(2) also provides for there to be a statutory cap on the balance payable by a domestic customer following such account where a meter is “proved to have registered less than the volume of water supplied” to the house. Whilst the wording of Regulations 9(1) and (2) would certainly apply where a meter had exceeded the prescribed limits of error on a subsequent test, they are both couched in neutral language, and could equally apply if the meter had passed a subsequent test, but had been proven by other evidence to have been “registering incorrectly” during the period in question.

The evidence in the instant case

[75]For reasons that I have explained, I consider that Clays was entitled to rely on the evidence from the joint expert in seeking to show that it had not been provided with the volume of water measured by the meter(s) in this case. However, I agree with Lord Justice Lewis that on proper analysis, such evidence did not support the deputy district judge’s conclusion that Clays had established that the suspect meter had been reading inaccurately.[76]In short, the reason for that is that the expert did not examine the meter internally to ascertain whether it showed any signs of wear and tear or the build-up of scale. His conclusion that something untoward had happened was therefore pure speculation based upon his acceptance of what he had been told by Mr Smith. That was not a matter for the expert, but for the judge to assess. I therefore agree with the judge and with Lord Justice Lewis’s analysis in paragraphs 50 and 51 above.

Disposal

[77]I would dismiss the appeal. LORD JUSTICE PETER JACKSON[78]In agreement with Lewis LJ, I would dismiss the appeal on grounds 1 and 2 for the reasons he gives, and for the following concurring reasons in relation to ground 1.

Ground 1: The Regulations

[79]The Water (Meters) Regulations 1988 have stood, virtually unamended, for the better part of 40 years. They embody Parliament’s decision to bring about a fundamental change in the basis of charging for water. When charges were based on rateable values, there was a weak and notional correlation between charging and consumption. The clear purpose of the 1988 legislation was to correlate charging with consumption as closely as is practically possible. The only way to achieve that was by regulating the use of water meters, because water meters are the only way in which water consumption can be measured.[80]The statutory framework, described in outline by Lewis LJ at paragraphs 6-12 above, provides for the installation, connection, disconnection, use, maintenance, authentication and testing of meters in accordance with closely prescribed standards and procedures. As a means of achieving the statutory purpose, the scheme is coherent and clear, and it has proved durable. That is critically important for an enterprise on this scale, involving millions of domestic and non-domestic consumers.[81]The legislation foresaw that water meters cannot be 100% accurate in all conditions. It acknowledged and catered for this in two ways. First, it set prescribed limits of error. Meters are tested against five parameters, four of which set a tolerance of ±2.5% and the fifth a tolerance of ±6%. (The meter in the present case was found on testing to be at least 99% accurate on all five measures.) Second, the legislation provided for the consequences where a meter has been found to be registering incorrectly on testing.[82]The legislation does not identify or provide for a third anomaly: the meter with a transient fault that corrects itself before testing (‘transient’ is preferable to ‘intermittent’, which supposes repetition). Is that silence mere reticence, allowing for a bespoke charging regime for customers (or suppliers) who wish to challenge the readings of meters that have passed the statutory test? In my view it is not, for these reasons.[83]The legislation has to my mind all the characteristics of a comprehensive scheme. There is no indication that any part of the charging regime falls outside its statutory footprint. It is not stated in terms that meter readings are to be the “only” evidence of the volume of water supplied because that does not need saying: it is inherent in the scheme that meters are the only way to measure water supply. The only exception is where a meter has been tested and found wanting, and in that case Reg. 9 (entitled “Adjustment of charges”) applies. Here, and only because the meter cannot be relied upon, an account has to be taken, with a statutory cap on the charges that can be levied on a domestic consumer. There is no similar provision for adjustment of charges in Reg. 9 or elsewhere because of a transient, self-correcting meter. That is a far more speaking silence than the absence of a superfluous conclusive evidence clause.[84]It is not surprising that Parliament has not provided for challenges of the present kind. Transient, self-correcting faults, if they exist, will surely be very rare. I say “if they exist” because there was no evidence in the present case of a self-correcting fault ever having been definitively established in any circumstances whatever: the most that the expert could speculate was that it was “quite possible” that items obstructing the mechanism had been washed out (SB118). Further, it is not clear how a transient, self-correcting fault would be reliably identified, even if the meter was opened after passing its test. Finally, if a fault was somehow identified, the sub-optimal process of taking an account would have to be embarked upon, and without any statutory cap. In the present case, the Deputy District Judge disallowed the invoices and dismissed the claim because no other basis for calculating the charges had been put forward. She did not explain what that other basis might be.[85]Parliament is most unlikely to have contemplated the existence of such a state of affairs after a meter has been tested and found to be working within tolerances. A process of the kind described above, where evidence about water consumption is at large, is not only unsatisfactory but also likely to be highly disproportionate. The present case offers a dire example. By the end of the trial before the Deputy District Judge, the consumer’s legal costs were £124,000 and the supplier’s were £69,000. After two appeals, the parties’ total costs are in the region of £0.6 million. Much of the appeal expenditure is no doubt attributable to the point of statutory construction, but even so the sums involved are wildly disproportionate to the invoices. It is therefore not correct to say that there is no evidence to suggest that customers would readily, or without good reason, embark upon the costly exercise of obtaining expert evidence to challenge the reading from a meter that has passed a test. That is exactly what has happened in this case. I therefore do not share the view that customers are likely to be deterred by the cost of challenging bills, or that water companies can be adequately compensated in costs for unsuccessful challenges. A system that allowed parties (customers or suppliers) to deploy lay and expert evidence in an attempt to override tested meter readings would inevitably lead to attempts to do so, and many consumers will be unable to compensate suppliers for the costs incurred in such disputes. It would, as the judge rightly said, be chaos.[86]I do not consider that hypothesising extreme and improbable possibilities of injustice is a useful guide to the interpretation of a universal statutory scheme where clarity and efficient use of resources are so important. It is telling that even in the present case, where an expert cast doubt on the meter reading, the challenge has failed.[87]My conclusion is that, on a true construction of the legislation, evidence other than that obtained by statutory testing, is not admissible to show that a water meter is unreliable or that the volume of water supplied should be measured in some other way. Any other construction would be unworkable and anomalous.

Ground 2: the present case

[88]I agree that the appeal would also have failed on this ground, for the reasons given by Lewis LJ at paragraphs 52 and 53.

Outcome

[89]The appeal is dismissed.