Nuisance — Flooding — Liability of sewerage undertaker — Flooding from culvert causing damage to property — Whether sewerage undertaker having statutory duty to maintain — Culvert forming part of channel originally consisting of natural stream — Whether character of culvert altered to that of sewer — Relevance of changes to other parts of channel
A property owned by the first respondent housing association suffered damage owing to flooding from an adjoining watercourse channel. The respondent brought proceedings in nuisance against the appellant, as the relevant statutory sewerage undertaker, and the second respondent local authority on the ground that the relevant part of the channel was a sewer, the maintenance of which was the responsibility of one or other of them. Both contended that responsibility for the relevant section of channel lay with the first respondent as the riparian owner. A preliminary issue was tried as to the status of the channel where it adjoined the first respondent’s land.
The channel consisted of a formerly natural stream, which had been lined with concrete for its entire length and covered for much of that length, although it formed an open, uncovered culvert where it ran past the respondent’s property. The Environment Agency had given permission for a foul sewer overflow into controlled waters at times of flash floods, such that foul sewage flowed through the culvert at times of high rainfall. Over the years, parts of the channel had been subjected to major works for the purpose of providing drainage of surface water from roads and a railway line. The county court judge held that the character of the natural stream had changed to such an extent that the culvert was now to be characterised as a surface-water sewer. He took into account that: (i) the channel as a whole was piped or lined in concrete; (ii) the culvert carried foul water on occasion; and (iii) the flow through the channel included a great deal more surface water than previously, including surface water brought to it from surface-water sewers. The appellant appealed. It submitted that the nature of the flow alone was relevant, not the physical characteristics of the channel along which it flowed, and that the change in the former was not sufficiently substantial as to turn the culvert into a sewer.
Held (Toulson LJ dissenting): The appeal was allowed.
Where the question for decision is as to the status of a particular stretch of a stream or flow, it must be addressed by reference to that specific stretch and not to the channel as a whole. It does not follow that, because one part of a channel is a sewer, another part must also be. The status of a stretch of the channel downstream from the culvert was unlikely to be relevant to the status of the culvert itself, and the judge had erred in taking into account changes to the channel downstream from the culvert. With regard to the other factors considered by the judge: (i) the fact that a watercourse has been lined or piped or covered does not affect the status of the flow, unless the works have been carried out pursuant to a relevant statutory authority, for which there was no evidence in the instant case; (ii) the effect of the discharge of foul sewage into a channel will depend upon the statutory provisions under which it is done, which had not been before the court in the instant case, such that it was not possible to base any conclusion on that factor; and (iii) the fact that a channel receives some, or even a great deal of, surface water cannot of itself transform what was once a watercourse into a sewer. It will be relevant in that context whether the channel still carries a flow of natural groundwater as well as the surface water. Since the culvert had originally carried groundwater and remained partly open, the natural assumption was that it still did so, and the burden of proof lay upon the first respondent to show otherwise. The status of the culvert had not changed to that of a sewer, however much its appearance might have changed in part and whatever changes there had been to the character and status of other parts of the channel, both upstream and downstream.
The following cases are referred to in this report.
Attorney-General v Lewes Corporation [1911] 2 Ch 495; [1911] 55 Sol Jo 703
British Railways Board v Tonbridge and Malling District Council (1981) 79 LGR 565, CA
George Legge & Son Ltd v Wenlock Corporation [1938] AC 204
Sefton Metropolitan Borough Council v United Utilities Water Ltd [2001] EWCA Civ 1284
Shepherd v Croft [1911] 1 Ch 521
This was an appeal by the appellant, Southern Water Services Ltd, from a decision of HH Judge Iain Hughes QC, sitting in Southampton County Court, determining a preliminary issue on a claim by the first respondent, Raglan Housing Association, in nuisance against the appellant and the second respondents, Southampton City Council.
Clifford Darton (instructed by the legal department of Southern Water Services Ltd) appeared for the appellant; Graham Chapman (instructed by Jacobs & Reeves, of Poole) represented the first respondent; Jeremy Burns (instructed by the legal department of Southampton City Council) appeared for the second respondents.
Giving the first judgment, Lloyd LJ said:
Introduction
[1] The Bitterne Stream was once a natural stream running westwards from Humm Hole to the River Itchen. What now exists, from the same starting point, is a channel through which water flows, passing through Bitterne, effectively a suburb of Southampton, and eventually flowing out into the Itchen. For much of its length, it is covered over, and the entire bed of the stream is lined with concrete. Part of one of the open stretches runs along the southern boundary of the first respondent’s property at 60-64 Cobbett Road, the present buildings on which were constructed in 1998. This area is prone to flooding. Several occasions of flooding occurred in 1999. In these proceedings, the first respondent seeks to establish that one or other of the second respondents and |page:156| the appellant is responsible for the maintenance of the channel and for making good any damage caused by a failure properly to maintain it.
[2] The first respondent contends that the damage to its property caused by the flooding gives rise to a claim in nuisance against the party responsible for the section of the channel that is adjacent to its property. Given that both the second respondents and the appellant deny such responsibility, the first respondent sought declarations as to both the ownership of the relevant section of the channel and the responsibility for its maintenance. District Judge Ainsworth directed that a preliminary issue be determined as to whether “the culvert” is a sewer and whether the appellant, as sewerage undertaker, has statutory responsibility to maintain it. In this context, the culvert means the section of the channel adjacent to the first respondent’s property. I will describe it more precisely later in this judgment.
[3] HH Judge Iain Hughes QC heard the preliminary issue, with evidence from four witnesses heard over three days, and made a site visit on a fourth. He handed down judgment on 14 March 2006 and, on 25 May 2006, he heard submissions as to the order to be made. His order, made on that date, answered the preliminary issue by saying that the culvert was a surface-water sewer, but not a public sewer. He refused permission to appeal to both the second respondents and the appellant. Only the appellant pursued the question of permission to appeal, which was refused on paper by Jonathan Parker LJ, but granted on oral renewal by the Chancellor.
[4] The issue before us is whether the judge was right to hold that the relevant part of what was once a natural watercourse (the Bitterne Stream) has become a sewer. It appears to be accepted that other parts of it have become a surface-water sewer, but that is not accepted as being true of the relevant part.
[5] Debate on this subject involves the use of terminology that may be open to confusion. A flow of water, or the channel along which it flows, may be a watercourse or a sewer, but according to the correct use of legal language it cannot be both. Therefore, to use the word watercourse risks begging the question, even though it might be a natural word to use as a description of a particular feature according to the ordinary use of language. There may also be confusion between the physical structure along which the water flows, on the one hand, and the flow of water itself, on the other: the container and the contents. That might not matter but for the argument, for the appellant, that the only thing that is relevant is the character of the flow, and that the physical characteristics of what it flows through or along is irrelevant.
[6] The judge used two phrases: “the channel” referred to the entire course of what used to be the Bitterne Stream from Humm Hole to the Itchen, and “the culvert” meant that part of the channel that is immediately adjacent to the first respondent’s property. Part of this stretch is open and part, at the western end, is covered. I will use the same phrases, and will seek to make it clear, if it is necessary to make the distinction, whether I am referring to the flow of liquid along the relevant stretch or to the physical structure.
Facts
[7] The judge found that the channel is what used to be the Bitterne Stream, originally a natural watercourse. I will summarise what the judge said about its present course:
(i) At the eastern end, at Humm Hole, it is still partly open. According to the public sewer map (PSM), which the judge saw and which was the only map shown to us, it then passes through a covered section going westwards, part of which is along Beech Avenue. Part of this stretch was piped (by the borough of Southampton) in order to assist in the draining of an area that was a boggy marsh and that is now occupied by Beechwood Junior School.
(ii) From there, it emerges into an open section. At this point, the channel is joined by another, both of them issuing into the single channel with which this appeal is concerned. The open section continues as far as Midanbury Lane, passes under that road, and then is again open for a stretch that includes the culvert, just to the south of the first respondent’s property, before becoming covered again just before it goes under Cobbett Road, which marks the western end of the first respondent’s property. The judge noted that the covered stretch adjacent to the first respondent’s property was shown as being covered over on the 1949 Ordnance Survey map, and he inferred that this had been done to allow for the construction of what became 62 and 64 Cobbett Road, properties that were eventually demolished and replaced by the first respondent’s property: see [52] of the judgment.
(iii) From that point on, the channel remains covered for some distance, passing under what is now a gyratory road system involving Cobbett Road and Buller Road. The construction of the gyratory system, in the early 1960s, seems to have involved changes to the course of the channel, so that it now has several right-angle bends. The judge held that the major works to the channel in the area of the gyratory system had been undertaken by the borough, in their capacity as the statutory sewerage undertaker, in order to deal with surface water: see [67].
(iv) West of the gyratory system, the channel goes under Bitterne Road West and under the railway lines just south of Bitterne station. Part of the work involved in this area was the subject of an agreement dated 27 December 1951 between the borough of Southampton and the British Transport Commission, which refers to “a 45 inches internal diameter Surface Water Sewer” to be constructed under the commission’s land in a given position. The borough, the predecessors in title of the city council, undertook to maintain and repair the sewer.
(v) Further west from the railway, the channel passes in a covered section as far as just north of Bitterne Manor School. There, it becomes open again and remains open, apart from passing under Quayside Road, until it reaches the Itchen.
[8] The stretch of the channel that lies east of the point at which it emerges into an open section near Beech Avenue is shown on the PSM as being a public sewer. The stretch to the west of that point used to be so shown but was removed from the PSM by the appellant in 2001.
[9] The judge recounted the changes in the course of the channel that appear from successive Ordnance Survey maps. In 1878, the entire channel was shown as a natural stream, the only interference with which was the construction, by then, of a single-track railway. In the first part of the 20th century, the channel was piped in part to the west of the culvert and also to the east, near Beech Avenue. In the 1960s or thereabouts, the gyratory system was put in, and more piping was done. By the 1949 edition, the culvert is shown sinking in order to allow for the construction of 62 and 64 Cobbett Road, later demolished. At that stage, the entire culvert was shown as being within the curtilage of 64 Cobbett Road, whereas now it is in the property to the south, from which the judge inferred that the line of the culvert had been moved a little to the south.
[10] The judge said that the 1949 edition was the first that showed the culvert, and, in [116], that there was no evidence as to who had built the culvert (meaning, here, the container — the concrete lining along which the water flows). He was satisfied that parts of the channel to the east and to the west had been constructed by the city council or their municipal predecessors, which were at the time responsible both for foul and for surface-water sewers, but he could make no finding as to who had constructed this part. He accepted that there was no record that the city council had done so: see [25].
[11] As for the flow along the channel, he found that, in 1878, it was still a natural stream. At that stage, therefore, it carried both natural groundwater and surface water. According to the evidence, there is still a continuous clear water flow in dry weather. The judge said that he had no evidence as to the source of this water flow, although he referred to one witness statement that the constant base flow of clear water in dry weather showed that groundwater was entering the conduit at many points and that it was acting as land drainage for the catchment area. However, the judge said that this water flow had not been tested to see whether it was natural groundwater, leaks from water supply pipes, discharges from domestic or commercial premises, run-off drainage or a combination of these. The culvert is in a natural valley, which will naturally receive surface drainage from both north and south. |page:157|
[12] In addition, there are some points in the channel upstream of the culvert at which there is a CSO — combined sewage overflow [or consented storm overflow1]. These are locations at which the regulator (the Environment Agency) has given permission for a foul sewer to overflow into “controlled waters” at times of flash flood: see section 88 of the Water Resources Act 1991. As the judge said, as a result, at times of high rainfall, foul sewage will flow through the channel, including the culvert, together with the surface water. These CSOs are known to exist because they are shown on the PSM, but their terms are not known because the relevant documents were not in evidence. Mr Clifford Darton submitted that they show that the channel downstream is a watercourse because a surface-water sewer is not within the definition of “controlled waters”. I agree with the judge that, in the absence of sight of the consents, this is not a point that can be relied upon: see [83].
—————————————————————————
1 In para 29, the judge gives “combined sewage overflow” as the phrase for which CSO stands. In the appellant’s skeleton argument, para 10 says the same, but para 4(v), footnote 1 gives “Consented Storm Overflows”: which is correct?
—————————————————————————
Law
[13] Rights and obligations in respect of sewers and, in particular, public sewers, have been regulated by statute for a very long time. The Public Health Act 1936 (the 1936 Act) is one of the significant stages in the evolution of statutory regulation in this area. More recently, the legislation relating to the privatisation of the water industry (the Water Act 1989 (the 1989 Act) and subsequent Acts, including the Water Industry Act 1991 (the 1991 Act) and the Water Resources Act 1991) has affected the position. However, we were referred to few statutory provisions. Instead, as before the judge, counsel’s submissions focused on a number of decided cases.
[14] In George Legge & Son Ltd v Wenlock Corporation [1938] AC 204, the House of Lords held that the discharge of sewage into a natural watercourse since 1876 could not convert the stream into a sewer because the discharge of sewage was unlawful under the Rivers Pollution Prevention Act 1876 (the 1876 Act). Lord MacMillan pointed out, at p213, that a channel may be a sewer although it may carry no sewage and its contents consist solely of innocuous surface drainage. Since any natural stream or watercourse that is still open will almost always contain some surface water, at any rate in or after wet weather, the fact that a flow of water does contain such surface water cannot show that it is a sewer, and cannot transform what was once a watercourse into a sewer.
[15] In Shepherd v Croft [1911] 1 Ch 521, at pp526-527, Parker J said that “the mere fact that a natural watercourse is culverted or piped by the several owners of the lands which are intersected by it does not make it a drain or sewer so as to vest it in the local authority” under the Public Health Act 1875. Shortly afterwards, in Attorney-General v Lewes Corporation [1911] 2 Ch 495, Swinfen Eady J had to consider a converse case, where crude sewage was discharged by the local authority into an intermittent stream, which was partly tidal. He said, at p508:
The question then arises, is the culvert a sewer? The plaintiffs contend it is. The defendants dispute it. The mere pollution of a natural stream or watercourse by turning sewage into it does not convert it into a sewer. On the other hand, if the watercourse has become substantially a sewer, the fact that at certain periods of the year clean water flows into it will not in my opinion prevent it from being a sewer. The question is one of fact and degree in each case. See Falconar v South Shields Corporation (1895) 11 TLR 223. In that case Lindley LJ pointed out that the stream had changed its character completely and had become a sewer in the ordinary sense of the word, ie, a channel for the reception and carrying away of sewage. It was a dirty, filthy sewer.
[16] The principal modern case on the point is the decision of the Court of Appeal in British Railways Board v Tonbridge and Malling District Council (1981) 79 LGR 565 (BRB). Oliver LJ gave the judgment of the court. The question was whether a culvert under a railway carried a sewer or a watercourse. It appeared that the construction of the railway had interrupted three ancient natural channels or watercourses that drained a large catchment area, and the culvert was constructed in order to carry away the water from the ancient watercourses despite the obstacle created by the railway embankment. If it was a sewer, the board could seek to have it vested in the local authority under section 17 of the 1936 Act, but this did not apply if it was not a sewer but a watercourse. At p572, Oliver LJ said:
Granted that in certain circumstances that which started life as a watercourse can become a sewer, that is not easily established where all that has happened is that water, whether surface water or foul water, has been made to flow through an outfall into an existing natural stream. One has to ask whether the circumstances are such that the stream has substantially lost its original character and taken on the character of a sewer and that does not occur simply because the stream is made to carry a quantity of sewage.
[17] Later, at p573, he went on:
What is clear is that something very much more than the mere discharge of sewage into a stream (and, a fortiori, the mere discharge of pure surface water) is required before its status is changed to that of a sewer (see, for instance, Glasgow, Yoker & Clydebank Railway Company v MacIndoe (1896) 24 R (Ct of Sess) 160). Thus, for instance, if circumstances are such that what was originally an agricultural stream comes to carry sewage in such substantial quantities that its character is completely changed (as occurred in Falconar v Corporation of South Shields (1895) 11 TLR 223) it may no doubt become a sewer within the ordinary meaning of the word.
[18] He then cited a passage from the speech of Lord Maugham in George Legge, and went on to say, at pp574-575:
In the instant case there has been nothing approaching the situation envisaged by Lord Maugham in the passage I have read. All that has happened is that outfalls have been constructed channelling the surface drainage of the built-up area into the existing streams so as to increase to some extent the flow of surface water which they carry away; and it is, Mr Nugee submits, quite impossible to say that these streams have become, as a result of such increased flow of surface water, “sewers” within the ordinary meaning of the word.
We find Mr Nugee’s argument persuasive. On the facts as found or agreed at the trial it is, in our judgment, clear that there has been no alteration in the essential character of the three watercourses and the culvert since 1840 and the mere fact that the surface drainage of the built-up area has been collected and diverted into them through a number of outfalls so as to produce a significant increase in the volume of water carried off, cannot possibly constitute them, either individually or collectively, sewers or a sewer within the ordinary meaning of that term.
In that case, no question arose as to works done by a statutory authority that might have been done in the capacity of a sewerage undertaker, and might therefore have had a bearing on the status of the resulting works. All the work had been done by the railway company under its own statutory powers.
[19] We were shown a more recent decision of the Court of Appeal where such a question did arise: Sefton Metropolitan Borough Council v United Utilities Water Ltd [2001] EWCA Civ 1284. That case concerned part of Maghull Brook, which passed under the densely populated centre of Maghull, in Merseyside, in an enclosed culvert constructed in around 1958. The question was whether this part had become a sewer before 1 April 1974 because of the culverting work. The question arose indirectly. The British Transport Commission was the successor to the Leeds and Liverpool Canal Company. The Leeds and Liverpool canal passed over this part of the Maghull Brook by way of an arch or tunnel constructed under statutory authority that the commission was bound, under the same statute, to maintain. West Lancashire Rural District Council (the RDC) were, before 1974, the relevant local authority for the area and were also the sewerage authority for the area under Part II of the 1936 Act. In 1961, the commission and the RDC entered into a deed under which each of them accepted responsibility for one-half of the cost of repairing, maintaining and renewing the arch and channel under the canal. In 1974, Sefton Metropolitan Borough Council became the relevant local authority, but North West Water Authority became the sewerage undertaker. The question was which of these two authorities (or, rather, its successor |page:158| in the case of the sewerage undertaker, namely United Utilities) was responsible for the liabilities of the RDC under the 1961 deed. This turned on the question in which capacity the RDC had acted when they entered into the deed. In turn, this depended upon the capacity in which the RDC had carried out the work of culverting in 1958. Had it been undertaken under Part II of the 1936 Act, which applies to local authorities’ functions and duties in respect of sewerage, or was it under Part XI of the 1936 Act, which set out a number of miscellaneous powers of local authorities? The trial judge had held that the relevant stretch of the Maghull Brook had always been a watercourse, and had not changed its character as such. Sefton appealed and argued, among other grounds, that the judge was wrong not to hold that the character of this part of the Maghull Brook had changed to a sewer before 1 April 1974, but this argument was not pursued on appeal. The appeal focused instead on the evidence as to the basis upon which the culverting work had been undertaken in 1958. Robert Walker LJ said that the references in the minutes were not all one way, but that overall he was satisfied that the work had been undertaken under Part XI of the 1936 Act, under general local authority powers, and not under Part II, under powers conferred on a local authority in respect of sewerage. Given the different focus of the arguments in that case, on the fact that the culverting work had been carried out by a particular authority that had two capacities and on finding out in which capacity it had acted, it is not surprising that attention was given not to the flow along the relevant channel but to the construction work undertaken and to the basis upon which it was undertaken. Robert Walker LJ referred to BRB, which had been cited to the judge although not to the Court of Appeal, because of the abandonment of the ground of appeal to which it was relevant. It does not seem to me that Sefton adds to or qualifies what was said in BRB, save to demonstrate that construction work carried out in respect of a channel may make a difference, depending upon who did it and in what capacity.
[20] It seems to have been common ground that the Maghull Brook was still a natural watercourse, both upstream of the culvert, in a small undeveloped area east of Maghull, and also downstream, towards its junction with the River Alt. This fact seems not to have been regarded as being incompatible in itself with an intermediate part of the brook having become a sewer by virtue of works carried out under a particular statutory authority, but since the decision was that the work had been done by the RDC under general powers, not under sewerage powers, the point did not arise for decision as to whether it is legally possible to have a flow that is a watercourse for part of its length and a sewer for another part, with all or part of the watercourse downstream from the sewer.
Judge’s decision
[21] Judge Hughes decided that the culvert had become a sewer, but not a public sewer. He posed the question as to “whether the character of the natural stream has changed to such an extent that the channel has become a sewer in the ordinary, non statutory sense of the word”: see [41], based upon BRB. Later, in [46], he formulated the test as “whether there has been so substantial a change in the character of the channel as to transform it into a sewer”, as a question of fact and degree in every case in the light of the guidance given by BRB, and said that it was not an easy test to satisfy.
[22] The judge then reviewed the facts, and came to consider their effect, starting with the evidence as to a continued flow of groundwater. He rejected this on the basis that Southern Water Services Ltd (Southern Water) had not proved that the continued flow of clear water in dry weather (which he accepted there was) was attributable to ground water: see [88] to [90].
[23] The factors that he did regard as relevant were these. First, the complete change in the structure of the channel; second, the change in the route of the channel; third, the change in the nature of the flow; fourth, the fact that the channel is used for the drainage of buildings and yards appurtenant to buildings; fifth, the fact that many tributary conduits are themselves sewers properly so called; sixth, the description given to the channel in documents from 1951 onwards; seventh, the capacity of the channel; and, eighth, the occasional presence of foul sewage in the channel.
[24] He went on to consider the second question, namely whether the sewer was a public sewer. Although there is no appeal against his decision that it was not a public sewer, it is pertinent to note what he said on this point. Under section 219(1) of the 1991 Act, it would be a public sewer only if it were a sewer vested in a sewerage undertaker in its capacity as such, whether vested in that undertaker pursuant to a scheme under provisions of the 1989 Act or the 1991 Act (which provided for the transfer of functions, assets and liabilities in respect of the privatisation of the water industry) or under section 179 of the 1991 Act (which it was not) “or otherwise”. The judge said that, to be so vested, it would have to have been a public sewer at the date of the transfer of the property rights and liabilities of the Southern Water Authority, the former sewerage undertaker, to the appellant as part of the privatisation process under the 1989 Act. In order to qualify in that way, he said that it had to come within statutory requirements of section 20 of the Public Health Act 19392. This offered three possibilities, namely the sewer had been:
(i) constructed prior to 1 October 1937;
(ii) built as a sewer by the municipal predecessors of the city council after 1 October 1937, or acquired by them;
(iii) the subject of a specific declaration of vesting on the part of Southern Water or its municipal predecessor under section 17 of the 1936 Act.
—————————————————————————
2 This is what the judge said at para 113, but should the reference be to the 1936 Act?
—————————————————————————
[25] He rejected the first of these on the basis that the channel was not constructed in its present form before 1937 and, in particular, the culvert was first shown to exist as such in the 1949 Ordnance Survey. He rejected the second because there was no evidence that the municipal predecessors had done anything to the culvert (unlike to other parts of the channel). There was no evidence of any declaration of vesting, so the third possibility was also irrelevant. Accordingly, given that he had held that it was a sewer, it followed that it was a private rather than a public sewer.
Discussion
[26] Mr Darton, for Southern Water, the sewerage undertaker, submitted that most of the factors relied upon by the judge for his finding that it was a sewer rather than a watercourse are irrelevant in law, and that the only relevant matter is the nature of the flow, basing this submission on BRB, and that the change in the nature of the flow along the culvert was nothing like substantial enough to satisfy the test set out by Oliver LJ in that case.
[27] Mr Graham Chapman, for Raglan Housing Association Ltd (Raglan), supported the judge’s approach and conclusion, in particular submitting that more factors than just the nature of the flow were relevant to be considered in addressing the question of fact and degree as to whether the nature of what had been a watercourse had changed so that it was a sewer. Mr Jeremy Burns, for Southampton City Council, the relevant local authority, submitted that the only issue before the court was as to the status of the culvert, rather than the rest of the channel, and made common cause with Mr Chapman in supporting the judge’s approach on the evidence.
[28] The judge’s order relates only to the status of the culvert, not to the entire channel, even though the judge made findings as to what had happened to other parts of the channel. Mr Burns was therefore correct in pointing out that the appeal is therefore concerned only with the culvert as such. It does not follow that because one part of the channel is a sewer another part of it is also a sewer. It seems to me that, logically, the status of a stretch downstream from the culvert is unlikely to be relevant to the status of the culvert itself. It might seem odd to find that a section of the channel that is upstream of the culvert was a sewer, but that the culvert itself was a watercourse. It may be less odd if one remembers that a flow may be a sewer even though it carries no foul sewage but only surface water. I note that, at several points in his |page:159| judgment, the judge did pose the question in relation to the status of the channel as a whole: see [41], [46] and [109]. Expressing his conclusion in [109] (quoted in full below in [34]), he said:
I am satisfied on this evidence that there has been so substantial a change in the character of the channel, and therefore the culvert, as to transform it into a sewer.
[29] Where the question for decision is as to the status of a particular stretch of a stream or flow, such as the culvert in the present case, I would not regard evidence as to what has happened to other parts of the stream or flow as being irrelevant, but it does seem to me that the question needs to be addressed by reference to the specific stretch at issue, rather than the stream or flow as a whole. Otherwise, there is a risk that matters relating to other parts of the stream will be regarded as having more significance than they should.
[30] The judge found that two parts of the channel downstream from the culvert were the subject of works done by the relevant sewerage undertaker as such: see [55] to [58] and [64] to [67]. That applies to the length under and near the railway and that under the gyratory system, immediately to the west of the culvert.
[31] So far as the culvert itself is concerned, part of it is piped and covered, and the rest of it is lined with concrete. The judge had no evidence as to who had done this work over the length of the culvert, nor when (except that the piped section had been created by 1949) or upon what basis: see [116]. Of course, by itself, the fact that a watercourse is lined, or is piped and covered over, does not make a difference to its status: see Shepherd, quoted in [15] above.
[32] The judge referred to a different point in [94], namely the fact that, in addition to the pipe carrying the flow along the line of the original watercourse, a second pipe discharges into the channel by a junction created upstream of the culvert, which, as the judge said “has nothing to do with the original stream but is intended to assist with the removal of surface water sewage and, on occasions, foul sewage”. It seems to me that this point goes with other aspects of the change in the flow through or along the channel. He said, in [98], that “the channel now receives flow from a number of other sewers, principally surface water sewers but also on occasion foul sewers”. In [99], he continued: “The evidence suggests that the primary function of the channel is now surface water sewage. Most of the surface water sewers in the area discharge directly or indirectly into it.” He mentioned the fact that the significant part of the channel that runs under Beech Avenue is still shown on the PSM (maintained by Southern Water) as a public surface-water sewer, and said that since that part of the channel is no different in character from the part that lies downstream, it is difficult to see why one part should be treated differently from the other in terms of legal classification.
[33] Mr Darton submitted, on the basis of BRB, that the fact that a watercourse carries away surface water cannot make it a sewer: see [17]. The judge also relied upon the fact that the channel (including the culvert) carries foul sewage on occasion, pursuant to the CSOs already mentioned, at times of flash flood. In [107] and [108], he said:
107. The mere presence of foul sewage cannot alter the character of a watercourse. However the channel passes through a densely populated area of Southampton and for three lengths (excluding Humm Hole) flows in an open culvert. The risk of flooding is at its greatest at the very times when the presence of foul sewage is most likely.
108. On the evidence of past history flooding will take place at the property where the channel sinks. On the other side of the culvert at this point is another development of residential flats. I consider the occasional but deliberate discharge of foul sewage into a channel of this nature, with these characteristics, to be significant and quite different in nature and effect to a discharge of foul sewage into a river or watercourse in a rural area. I consider it proper to take this into account when considering this issue because of the impact of a flood including foul sewage on a population that did not exist when the channel was Bitterne stream.
[34] However, his conclusion, in [109], was that the channel, including the culvert, had become a sewer, but a surface-water sewer. He said:
109. The effect of these various features is cumulative, although some are plainly more important than others. Considering the features I have identified and the channel as a whole it no longer can sensibly be called a stream or a watercourse without undue strain to the ordinary meaning of those terms. I am satisfied on this evidence that there has been so substantial a change in the character of the channel, and therefore the culvert, as to transform it into a sewer. More precisely, it has become a surface water sewer.
[35] Among the factors to which he referred was the change in the route of the channel. The significant changes in this respect were all downstream from the culvert. It does not seem to me that these can affect the proper view of the status of the culvert itself. He also referred to the fact that the channel was described, from time to time, as a sewer in formal documents. Again, these references were to stretches downstream from the culvert. For these parts of the channel, the description may have been significant because they showed the basis upon which the work had been done. In my judgment, they do not affect the position upstream. He also referred to the fact that the capacity of the piped section of the channel is much greater than would be necessary to cope with only natural flow. That applies to sections both downstream and upstream from the culvert. It seems to me that it goes with the fact that the channel does now carry away a lot of surface water.
[36] Thus, taking the various points together, there are three main separate factors, or groups of factors, that the judge regarded as relevant: (i) the channel as a whole, and the culvert in particular, is now piped or at least (where it is open) lined in concrete; (ii) the flow includes a great deal more surface water than it used to, including surface water brought to the channel by way of conduits that are or include surface-water sewers, and it might be said, therefore, to form part of the surface-water sewer network; and (iii) the channel, and in particular the culvert, carries foul water on occasion, at times of flash flood, pursuant to consents given under the relevant legislation.
[37] Mr Darton’s submissions on those factors, based upon BRB, were clear and simple: piping or lining a flow of water does not affect its status; adding more surface water to a watercourse makes no difference; and the presence of foul sewage is not sufficient to change a watercourse into a sewer unless the quantity of sewage is very substantial, which clearly it is not in the present case.
[38] Given that, as George Legge showed, the discharge of foul sewage into a natural watercourse could not convert a watercourse into a sewer because the passing of the 1876 Act made such discharge illegal, it seems to me that nowadays the effect of the discharge of foul sewage into anything other than a foul sewer is likely to depend upon the statutory provisions under which it is carried out. If it is not done with consent, the position would be the same as under the 1876 Act. If it is done with consent (as in the case of the CSOs referred to above in the present case), it will depend upon the terms of the consent and of the provision under which the consent is given. The judge was not shown the consent, nor, I think, was he taken to the statutory provisions under which the consent may have been given. In those circumstances, it seems to me impossible to base any conclusion on the discharge of the foul sewage into the channel. It is fair to say that, from the terms of his conclusion in [109], it does not seem that the judge did rely upon this aspect to any substantial extent.
[39] By itself, it must be right that to put a flow of water into a concrete lining, or even into a pipe, is irrelevant to the status of the flow. So, the real point is whether the fact that the culvert now receives surface water to a much larger extent than before, brought to it by way of surface-water sewers, so that it can be said to form part of the local network of surface water drainage for the area, makes it a surface-water sewer itself. It lies downstream from a number of surface-water sewers and further downstream from it parts of the channel appear, from the judge’s findings, to have been the subject of works carried out by the sewerage authority as such, so that these parts are therefore presumably also sewers. (I should record that Mr Darton did not accept that the mere fact that a sewerage undertaker carries out works of construction to the channel of an existing flow, as opposed to constructing a new channel for a new flow, makes it a sewer if it is not one already.) |page:160|
[40] Where work has been done to the structure of a channel through or along which a watercourse flows, the statutory authority for the work could be such that it changes the character of the flow from that of a watercourse to that of a sewer. Whether it does will depend upon the facts of the given case and the terms of the statute. No such point arises in the present case with regard to the culvert because it is not known who undertook such work as has been carried out to the culvert itself, as distinct from other parts of the channel, and, therefore, it is also not known whether there was any relevant statutory authority for it. There is no evidence that connects the work carried out with the city council, so, on the evidence, it had not been done by a relevant authority. If it had been, it might have qualified as a public sewer.
[41] What is left, therefore, is the fact that the culvert, as part of the channel, now carries a great deal more surface water than it did originally, and the question as to whether it still carries any groundwater.
[42] The channel carries a continuous base flow of clear water in dry weather, but the judge had no evidence as to the source of this flow. He commented that Southern Water had not tested the water flow to determine whether it was natural groundwater. He regarded it as possible that it was attributable to one or more of leaks from water supply pipes, discharges from domestic or commercial premises or run-off drainage, rather than natural groundwater. He held that this point was not decisive because if a channel is a sewer, the fact that it does receive some natural groundwater, even on a continuous basis, does not alter its status as a sewer. I agree with that proposition, but it is also true that if it is a watercourse, the fact that it receives some, or even a great deal of, surface water does not change it into a sewer. Moreover, the judge’s proposition does beg the question: if it is in issue whether the flow is a sewer or a watercourse, the fact (if it be so) that it continues to carry a flow of natural groundwater, as it did historically, is relevant to answering the question. If there were no natural groundwater, it could be much easier to establish that it was now a sewer because it would be carrying only surface water, rather than carrying such water as an unavoidable addition to natural groundwater. The judge approached this question on the basis that it was for Southern Water to show that there was still a continuous flow of natural groundwater. I respectfully disagree. Since it is clear that the channel was originally a watercourse, the burden of proof of relevant facts, where in dispute, would be on the party seeking to show that the situation has changed, that is to say, on Raglan rather than on Southern Water.
[43] The judge may have been influenced in this by the view he took of the appellant’s decision, without having gone through the procedures that it had laid down for itself to follow, to remove the part of the channel west of Beech Avenue from the PSM. As the judge said, in [71], this was not determinative because although a sewerage undertaker is under a statutory duty to maintain a PSM (see section 199 of the 1991 Act), the contents of the PSM do not define what is or is not a public sewer. It is not of the same significance as a definitive map of rights of way under the Wildlife and Countryside Act 1981, for example. Nevertheless, it is striking that the appellant did remove this part of the channel from the PSM shortly after the first letter from the first respondent’s solicitor. The judge may have felt that, before taking such a step, the appellant ought to have tested the clear-water flow to assess its source or sources.
[44] The judge also paid regard to the evidence of Mr Rawson Burns, a retired chartered engineer who had worked for Southern Water and previously for Havant Borough Council and who gave evidence (among other things) as to the practice of making corrections to a PSM. The judge referred to the criteria that Mr Burns mentioned in this respect (see [85]), and considered the appellant’s decision to remove part of the channel from the PSM in the light of those criteria: see [86]. It was in that context that he made the comments referred to above as to the absence of tests on the part of the appellant as to the source of the clear water. Those comments seem to me entirely legitimate in the context of considering the appellant’s decision to remove part of the channel from the PSM. They do not necessarily reflect the correct legal approach to the issue that the judge had to decide.
[45] However, the judge went on to say, in [91], that he was not prepared to determine the case solely upon the basis of the existence of a significant clear flow during periods of dry weather. I agree that it would not have been appropriate to decide the case solely on such a basis. The main question is as to the significance of the largely increased flow of surface water emanating from the different sources mentioned by the judge, including a number of upstream tributaries, some of which appear to be surface-water sewers.
[46] In respect of this, Mr Darton relied heavily upon Oliver LJ’s words in BRB, quoted above in [17]:
What is clear is that something very much more than the mere discharge of sewage into a stream (and, a fortiori, the mere discharge of pure surface water) is required before its status is changed to that of a sewer.
[47] The judge’s decision is based upon a number of matters, not all of which I would regard as being of equal relevance. The increase in the amount of surface water in the stream is one of the most important aspects of his reasoning. However, it can fairly be argued that he did not proceed upon the basis of a mere discharge of surface pure water into the stream. The discharge is, in part, the result of substantial works carried out by the sewerage undertaker upstream of the culvert, for the drainage of the surface water in what seems likely to be a wider catchment area than that of the original stream. That produces a curious result, on the judge’s finding. The stretches of the channel upstream (as well as downstream), where the then sewerage undertaker undertook works under statutory authority, may have become sewers (they were not directly in issue, and I do not say they necessarily did, although, clearly, the judge thought they did). If so, it seems likely, at lowest, that they would have become public sewers that would be the responsibility of Southern Water as sewerage undertaker. By contrast, the judge’s conclusion that the culvert had changed from a watercourse to a sewer is largely based upon work done elsewhere in the channel by the sewerage undertaker, but because there is no evidence that the then undertaker did any work to the culvert as such, that part of the channel, if it became a sewer, remained a private sewer and is not the responsibility of the sewerage undertaker. Whose responsibility it is remains to be decided: Raglan argues that the city council are responsible, but the council deny this. I find it surprising that work carried out by the sewerage undertaker elsewhere in the channel should have had this side-effect as to the status of the culvert to which the sewerage undertaker did nothing in terms of construction, and for which it (rightly, according to the judge) accepts no responsibility in terms of maintenance.
Conclusion
[48] I have found the issue on this appeal difficult to decide. I see the force of the judge’s comments concerning the change in the nature of what used to be the natural and open Bitterne Stream, and the oddity, on Southern Water’s contention, of having a fairly short stretch of watercourse in the channel with extensive stretches of what appear to be sewers (and probably public sewers at that, even if only or mainly for surface water, not foul sewage) both upstream and downstream of the part in question. On the other hand, I also find it curious, as I have said, that work upstream and downstream under statutory authority can have not only whatever effect it has upon the parts of the channel to which the work was done but also an incidental effect upon the status of the culvert and, what is more, a different effect, because it makes it a sewer but not a public sewer.
[49] I would pay tribute to the judge’s judgment, which is a model of clarity and deals very fully with the issue before him, in what were clearly far from ideal circumstances, with regard to the identification of the preliminary issues, and the manner in which they came before him for decision, including the serious underestimate of the time required. I have, however, come to the conclusion that the judge’s decision is wrong, and that the culvert is still a watercourse, not having become a sewer. My reasons for disagreeing with the judge are these.
[50] First, I consider that there are some errors in his reasoning. He ought not to have regarded it as being for the appellant to prove that the continuous base flow of clear water in dry weather still included |page:161| natural groundwater. On that point, the burden of proof was the other way. He ought not to have taken into account (to the extent that he did, which may not have been great) the discharge of foul sewage under the CSOs. I do not regard the fact that the course of the stream now lies over or through concrete culverts or pipes as a matter that, by itself, can have any relevance to the decision. The circumstances in which the construction works were carried out could be relevant if they were carried out under statutory authority, but none were relevant to the culvert. The change in the route of the channel is also irrelevant of itself, especially since almost all the significant changes are downstream from the culvert, and matters downstream seem to me to be inherently of little or no relevance to the status of the culvert. I would not attach any importance to the descriptions used in relation to parts of the channel, especially since none applied to the culvert itself (as the judge said, in respect of the second issue, in [116]). It seems to me also that the judge approached the question wrongly by considering the status of the channel as a whole, rather than that of the culvert, albeit in the context of the rest of the channel so far as relevant.
[51] What certainly is relevant is the nature of the flow through the culvert. The stream started as a watercourse, carrying natural groundwater, as well as, inevitably, some surface water. It still starts in the same place and it is still partly open there. It seems to me that, in those circumstances, the natural assumption should be that it still carries natural groundwater. If Raglan or the city council wished to contend otherwise, it seems to me that the burden of proof would have been on them. There having been no evidence on the point, it seems to me that the matter should proceed on the basis that the channel still carries some natural groundwater. It has not changed in that respect.
[52] Clearly, it also carries surface water and, no doubt, a lot more than it used to. It does occasionally carry some foul sewage, but I do not regard that as relevant because of the absence of evidence as to the terms and basis of the CSOs. So one is left with the large increase of surface water, the fact that there have been substantial works upstream that presumably add to the flow of surface water and that some of what there is upstream, by way of tributaries, so to speak, does seem to have the status of sewers, and, apparently, public sewers.
[53] If the question is posed specifically in relation to the culvert, it seems to me that the only relevant change that has occurred is the addition of a great deal more surface water to the flow along that part of the channel. Nothing has happened to change the status of the culvert in terms of works carried out under statutory authority. The sewerage undertaker has done nothing to that stretch of the channel that can affect its status. Therefore, if there has been a change, it must be by virtue of the change in the flow because the fact that the flow is now channelled and partly piped is not sufficient. It seems to me that, viewed in that light and by reference to the culvert itself, there is no sufficient change to constitute the flow a sewer rather than a watercourse. I come back to the words of Oliver LJ in BRB, quoted in [17] above. There is no relevant discharge of foul sewage, and although there is an increased discharge of surface water, something very much more than that is needed. In my judgment, on the true analysis of the facts concerning the culvert, there is nothing more than that increased discharge that can be relied upon. Accordingly, the status of the stream as it flows through the culvert as a watercourse has not changed, however much its appearance may have changed in that part, and whatever changes there may have been to its character and even its status elsewhere in the channel, both upstream and downstream.
[54] For that reason, I would allow this appeal.
[55] The judge was understandably critical of the way in which public bodies had incurred substantial cost in pursuing expensive and time-consuming litigation, rather than coming to “a sensible negotiated settlement of what is in reality a relatively minor local issue”. It may be, however, that, particularly at a time when flooding is at risk of becoming more prevalent, the issue raised in this appeal has a wider potential significance for sewerage undertakers and others. Whether this decision itself is of wider relevance remains to be seen. Given the extensive flooding experienced in many parts of England in June and July 2007, the importance of issues of this kind is all too plain. I hope that, if such a question arises in future, the court will not have to decide it (as the judge had to in this case) without knowing the statutory basis upon which any relevant works of construction were done or even that of the known consents to the occasional discharge of foul sewage.
Giving the second judgment, Toulson LJ said:
[56] HH Judge Hughes QC expressed the view at the beginning of his judgment that the preliminary issue that he was being asked to try was inappropriate. I agree, but I am not sure that the judge appreciated the full extent of the problems that it created. I say that in no spirit of criticism because I, too, did not appreciate the full extent of the problems until I came to preparing this judgment with the benefit of reading in draft the judgment of the Chancellor. I will come back to this subject but will begin (with the benefit of Lloyd LJ’s full summary of the background and the facts) by considering the arguments advanced on behalf of the appellant water authority by Mr Darton.
[57] The primary submission made by Mr Darton can be summarised as follows: in considering whether a watercourse has become a sewer: (a) the court is concerned only with what flows through the channel, and not with any questions about how, why or in what circumstances any changes had been made to the construction of the channel; and (b) a watercourse can become a sewer only if the contents of the flow consist so substantially of lawfully deposited foul sewage that it has become essentially a foul-water sewer.
[58] I reject that submission. It goes further than anything said in the judgment of Oliver LJ in British Railways Board v Tonbridge and Malling District Council (1981) 79 LGR 565 (BRB) or in the cases cited by him, and it is inconsistent with Sefton Metropolitan Borough Council v United Utilities Water Ltd [2001] EWCA Civ 1284.
[59] In both BRB and Sefton, the courts had to consider whether a natural watercourse had become a sewer within the meaning of the 1936 Act. Section 343 of that Act provided:
“Drain” means a drain used for the drainage of one building or of any buildings or yards appurtenant to buildings within the same curtilage,
and
“sewer” does not include a drain as defined in this section but, save as aforesaid, includes all sewers and drains used for the drainage of buildings and yards appurtenant to buildings.
[60] Materially identical language appears in earlier and later statutes dealing with the same subject matter.
[61] During the argument, the language used by Oliver LJ in BRB was subjected to analysis of the kind that would be more appropriate if it appeared in the statute. Arising from Oliver LJ’s acceptance of the submission advanced in that case by Mr Nugee QC that the critical question was “whether the circumstances are such that the stream has substantially lost its original character and has taken on the character of a sewer”, we heard argument whether the word “stream” in that sentence should be construed so as to include the bed or be confined to the flow.
[62] The important point is, as summarised by Robert Walker LJ in Sefton, in [16], that the authorities:
indicate that a natural watercourse can become a sewer within the meaning of the 1936 Act (which I will refer to as a statutory sewer) and that whether this has occurred is a question of fact and of degree.
[63] In considering what might or might not bring about a change from a watercourse to a sewer, Oliver LJ said in BRB, at p573:
What is clear is that something very much more than mere discharge of sewage into a stream (and, a fortiori, the mere discharge of pure surface water) is required before its status is changed to that of a sewer (see, for instance, Glasgow, Yoker and Clydebank Railway Company v MacIndoe (1896) 24 R (Ct of Sess) 160). Thus, for instance, if circumstances are such that what was originally an agricultural stream comes to carry sewerage in such substantial quantities that its character is completely changed (as occurred in Falconar v Corporation of South Shields (1895) 11 TLR 223) it may no doubt become a sewer within the ordinary meaning of the word. |page:162|
[64] That passage provides authoritative but not exclusive guidance. It does not, in my judgment, provide a platform for the argument that the only circumstance in which a watercourse can become a sewer is where it has come to carry substantial quantities of foul sewage.
[65] If that were so, it would mean that Robert Walker LJ was tilting at a windmill for most of his judgment in Sefton. In that case, part of a watercourse had been enclosed in a mile-long stretch of culvert, constructed between 1958 and 1959, under what had become the densely populated centre of Maghull. It received surface-water drainage (and some unintended contamination from foul water) from many houses and other buildings in the area, as well as flows from other sources (including undeveloped land and highway drains). In other words, it primarily carried surface water from an urban area and any element of foul water was accidental. On Mr Darton’s argument, the culverted stretch could not be a public sewer. That is not how the Court of Appeal approached the matter. It identified the issue as being whether the culverting of the stream had had, in the relevant circumstances, the legal effect of causing the culverted watercourse to become a sewer. The culverting had been done by West Lancashire Rural District Council acting under statutory powers under the 1936 Act. West Lancashire were not only a local authority but a sewerage authority. They therefore had the powers of a sewerage authority under Part II of that Act and miscellaneous powers as a local authority under Part VI. The Leeds and Liverpool canal crossed over the culverted section. After the culverting works had been done, West Lancashire contracted with the canal authority to share the cost of maintaining the crossing. Under a later local government reorganisation, Sefton Metropolitan Borough Council became West Lancashire’s successors as local authority, and West Lancashire’s rights and liabilities as sewerage authority became vested in the North West Water Authority. The question arose as to whether the new local authority or the new sewerage authority were responsible for West Lancashire’s contractual obligations in relation to the crossing of the culvert with the canal. For this purpose, it was necessary to consider whether the culverted section had become a public sewer.
[66] Robert Walker LJ in his judgment, with which Henry and Rix LJJ agreed, noted, in [21], that, in BRB, the court did not enquire into the statutory power under which part of the natural watercourse had been piped, but that in the instant case both parties had recognised that the essential issue was whether, in carrying out the culverting work, West Lancashire had been acting as a sewerage authority under their Part XI powers. The judge had held the latter, and his decision was upheld. Accordingly, the work had not been done by West Lancashire as a sewerage authority and, in those circumstances, the works did not convert the watercourse into a public sewer.
[67] Robert Walker LJ observed, in [41], that both parties were agreed that West Lancashire could not have been exercising powers under both Part II and Part XI simultaneously. He added:
It is not obvious to me why a pipe substituted for a watercourse under s262(1) might not in some circumstances be a statutory surface-water sewer from the time of its construction, if at that time the watercourse was already receiving flows from other surface-water drains and sewers, and might not be a public sewer if constructed by a sewerage authority. But that line of argument was not adopted by experienced counsel on either side.
[68] The observation was obiter, but it is instructive to note in what respect. Robert Walker LJ was there saying that he could see no reason why if a sewerage authority (which were also a local authority) installed a pipe for surface water in place of a watercourse already receiving surface water, this might not be a public sewer, albeit that the statutory power used for the construction of the pipe was under section 262(1), that is, within Part XI. The point “not adopted by experienced counsel” was that a sewerage authority could use Part XI powers in that way, as distinct from proceeding under Part II. If, however, the court had held that West Lancashire had been using Part II powers to construct the culvert (and nobody suggested that they could not have done so), it is clear that the court would have held that the culverted section would, in those circumstances, have become a public sewer, although it was predominantly for surface water.
[69] In the present case, there was no evidence before the judge as to who had built the section of culvert adjoining the first respondent’s premises. However, upstream of that section, the local authority had undertaken substantial work, piping what had originally been the watercourse in order to take surface water from urban development and also to permit combined sewage overflows. A combined sewage overflow, or CSO, is a connection between a foul- and surface-water sewer in order to relieve pressure on foul sewers during periods of high rainfall. The judge found as a fact that the city council’s predecessors had constructed the channel under Beech Avenue to provide a surface-water sewer. He was entitled so to find (although I note, as a matter of detail, from the public sewer map that the channel beneath Beech Avenue also appears to have a number of CSOs).
[70] Mr Darton submitted that this was wrong. He submitted that even where a sewerage authority in purported exercise of statutory powers laid pipework in order to take surface water along the line of a previous watercourse, however small the previous flow may have been and however large the subsequent flow might be, the resulting work could not be a sewer in law, and that in so far as Robert Walker LJ’s approach in Sefton suggested otherwise, that approach was wrong. I do not accept that.
[71] Mr Darton further argued that if he were wrong in that submission, the matter is of no relevance when considering the status of the culvert adjacent to the claimant’s property. This brings me to what I regard as the really difficult part of the case, and it is a novel point. Taking it as an established fact that the channel under Beech Avenue was constructed as a sewer under statutory authority, what is the effect upon the culvert with which the judge was concerned? The judge recorded that the two pipes beneath Beech Avenue had diameters of 450mm and 800mm. Their contents flowed into the culvert adjacent to Raglan’s land and, on a common-sense view, must have constituted the vast majority of what entered the culvert.
[72] As the Chancellor points out, the judge was not asked to determine the status of the culvert within a statutory context. The Chancellor has referred, in [79], to the definition of “sewer” in section 219 of the 1991 Act (which is substantially the same as in the 1936 Act), but, as he observes, the judge did not consider that section to be of assistance. It is not satisfactory to decide a difficult point of law on an ill-drafted preliminary issue. I also suspect that the real issues between Raglan and the second respondent and the appellant involve wider considerations. Raglan is concerned about the risk of flooding. If a sewerage authority constructs a sewer in such a way that it is liable to cause flooding of land in the proximity, I can see that it may be liable at common law for the damage caused, whether the water issues directly from the authority’s sewer or floods the bed of a natural small watercourse and so floods the neighbouring land. However, none of this falls within the preliminary issue.
[73] I have real doubts about the appropriateness of the exercise that the judge was asked to carry out. However, I am also conscious that the parties have already incurred considerable expense, much of which will fall on the public. So, with those reservations, I proceed to give my view on the question of whether the judge was entitled to find that the relevant culvert is a sewer (using the word as interpreted in previous authorities under statutes containing a similar definition to that in the 1991 Act).
[74] If the channel immediately upstream of the section adjacent to the claimant’s property has been turned into a public sewer to accommodate principally surface water in much increased quantity, but also some foul water when the occasion requires, the effect of those changes upon the section by the first respondent’s property is, in my view, properly to be taken into account in considering whether that section has changed its character. Whether there has been such a change is ultimately a question of fact and degree. Mr Darton submitted that it would be ludicrous that a change in the character of one part of a channel should necessarily produce a change several miles away. I agree. A common-sense view has to be taken on the particular facts. |page:163| The fact that the water authority’s own public sewer map showed the relevant section as a public sewer until the present dispute arose, and the water authority issued a new version showing it not as a public sewer, was material that the judge was entitled to take into account as evidence of how the authority itself regarded the relevant section at the time the map was originally prepared.
[75] In his careful and thorough judgment, after reviewing the authorities, the judge summarised the basis upon which he had to proceed as follows:
I proceed on the basis that the correct test is whether there has been so substantial a change in the character of the channel as to transform it into a sewer. This is a question of fact and degree in each case, taking into account the above guidelines. The authorities make it plain that the test is a high one.
[76] He then proceeded to set out the facts in full and careful detail. In my judgment, he was entitled to conclude that, at the relevant time, the culvert with which he was concerned had become but a section in a network of a public sewerage system, predominantly but not exclusively for surface water, and that its character had therefore changed to the extent that it had become a sewer. I would therefore dismiss the appeal.
Giving the third judgment, Sir Andrew Morritt CVO, said:
[77] The issue before Judge Hughes QC was “whether the culvert is a sewer and whether Southern Water have the statutory responsibility to maintain the same”. The culvert was described by the judge as “adjacent to [60-64 Cobbett Road, Bitterne, Southampton] and running along the southern boundary but not within Raglan’s ownership”. By his order made on 25 May 2006, the judge declared the culvert to be “a surface water sewer but… not a public sewer within the statutory meaning of the same”. Lloyd LJ has described both the culvert and the channel of which it forms part and I gratefully adopt his descriptions.
[78] The judge described the circumstances in which the issue arose in [8] and [9] of his judgment. Thus, Raglan argued that the culvert was a sewer, whether public or not, and not a watercourse in its endeavour to impose liability for its flooding on one or other of Southampton City Council or Southern Water Services Ltd. The city council contended that the culvert was a public sewer so as, they hoped, to shift any liability onto Southern Water. Southern Water argued that the culvert was either a watercourse or a non-public sewer so as to divert any liability back to Raglan as riparian owner or to the city council. In those circumstances, the judge considered, in [10], that he should approach the preliminary issue in two parts: “First, is the culvert a sewer at all? Secondly, if the culvert is a sewer is it a public sewer?”
[79] The statutory meaning to which the judge referred in his order is that contained in the Water Industry Act 1991. Section 219(1) contains a number of definitions of potential relevance “except in so far as the context otherwise requires”. Thus:
“public sewer” means a sewer for the time being vested in a sewerage undertaker in its capacity as such…
“sewer” includes (without prejudice to subsection (2) below) all sewers and drains (not being drains within the meaning given by this subsection) which are used for the drainage of buildings and yards appurtenant to buildings;
“surface water” includes water from roofs;
“watercourse” includes all rivers, streams, ditches, drains, cuts, culverts, dykes, sluices, sewers and passages through which water flows… .
Although he referred, in [37] of his judgment, to the fact that both Raglan and the city council relied upon section 219 to support their respective arguments, he concluded that those provisions did not determine the matter: see [37] and [43].
[80] Instead of applying any of those statutory definitions, the judge adopted as the proper test, which he understood by the end of the argument to be common ground, that explained by Oliver LJ in British Railways Board v Tonbridge and Malling District Council (1981) 79 LGR 565 (BRB), at p572, quoted by Lloyd LJ in [16] of his judgment. The judge summarised that test, in [46] of his judgment, as:
whether there has been so substantial a change in the character of the channel as to transform it into a sewer. This is a question of fact and degree in each case, taking into account the above guidance. The authorities make it plain that the test is a high one.
[81] In my view, that proposition is open to criticism in two respects. First, it looks to the channel as a whole not just the culvert to which the preliminary issue was rightly confined. Elsewhere in his judgment, for example in [10], the judge altered the word “channel” to “culvert”. However, it does not appear that the use of the word “channel” in [46] is a mistake because there are later passages in his judgment in which the judge refers expressly to parts of the “channel” when considering the status of the “culvert”. As Lloyd LJ has pointed out in [28] of his judgment, logically the status of a stretch of the channel downstream of the culvert can be of little, if any, relevance once it is accepted that it is possible for different sections of the channel to have a different status. The second error lies in the omission of any reference to the contents of the channel. Whether or not the composition of the contents is the only consideration, it is one of the most relevant, as the judgment of Oliver LJ in BRB shows.
[82] The judge then traced the history of the channel (in [47] to [91]) and summarised the material facts (in [93] to [108]) under the headings: (a) “The structure of the channel has completely changed”; (b) “The route of the channel has changed to a significant extent”; (c) “The nature of the flow has changed”; (d) “The channel is used for the drainage of ‘buildings and yards appurtenant to buildings’”; (e) “Many of the tributary conduits are themselves sewers, properly so-called”; (f) “The description given to the channel after the war”; (g) “The capacity of the channel”; and (h) “the occasional presence of foul sewage in the channel”. The judge’s conclusion, expressed in [109] of his judgment, is as follows:
The effect of these various features is cumulative, although some are plainly more important than others. Considering the features I have identified and the channel as a whole it no longer can sensibly be called a stream or watercourse without undue strain to the ordinary meaning of those terms. I am satisfied on this evidence that there has been so substantial a change in the character of the channel, and therefore the culvert, as to transform it into a sewer. More precisely it has become a surface water sewer.
[83] In my view, the focus for the judge’s consideration and his conclusion, namely the channel as a whole, was wrong in both the respects to which I have referred in [81] above. The issue is the status of the culvert. Much of the judge’s consideration was devoted to the construction of the channel downstream of the culvert and independently of its flow or contents. As such, it was of marginal, if any, relevance.
[84] The judge then considered the second issue that he had analysed and concluded that the culvert was not a public sewer within the meaning of section 219 of the 1991 Act. That depended upon the identity of the undertaker in which it was vested. There is no appeal from that conclusion. Thus, this court is left in the unsatisfactory position of being expected to determine the first of the issues to which the judge referred, namely whether “the culvert is a sewer at all”, by reference to “the natural and ordinary meaning of the word ‘sewer’” divorced from a particular statutory context in the 1936 Act, the 1991 Act or any other relevant statutory provision.
[85] That, as I understand it, is why all parties have relied upon the judgment of Oliver LJ in BRB. As he made plain, at p571, he was considering, at least in the passages that followed, the natural and ordinary meaning of the word “sewer”, notwithstanding that the case itself concerned the meaning of the word in the context of the 1936 Act. Those passages have been quoted by Lloyd LJ in [16] to [18], and I need not repeat them. They clearly show that the nature of the contents or flow of the culvert is the most significant factor in the determination of the question as to whether what was once a natural watercourse has become a sewer. However, I agree with Toulson LJ ([64]) that the decision of the Court of Appeal in that case does not establish that alterations to the bed and banks of the culvert or of an upstream section of the channel are irrelevant to the question. It is easy to envisage cases where the additional flow of sewage is inconclusive but works to the relevant bed and banks resolve the issue. |page:164|
[86] On the other hand, I agree with Lloyd LJ that the decision of this court in Sefton Metropolitan Borough Council v United Utilities Water Ltd [2001] EWCA Civ 1284 does not carry the matter much further. As he has explained in [19] to [20], the reference in that case to the decision of the Court of Appeal in BRB was only in connection with a ground of appeal that had been abandoned. However, the decision in Sefton does confirm that construction work on the culvert and upstream bed and banks may be relevant, at least if who did it and in what capacity is relevant to the issue.
[87] I agree with Lloyd LJ that, for the reasons he gives ([38], [40] and [50]), various considerations the judge took account of have to be excluded from consideration or substantially discounted. Accordingly, as he said in [41], one is left with:
the fact that the culvert, as part of the channel, now carries a great deal more surface water than it did originally, and the question whether it still carries any ground water.
[88] If the question is whether the culvert is a sewer within the meaning of the definition contained in section 219 of the 1991 Act I have quoted in [79] above, it may be that the presence of surface water in the volume now to be found in the culvert would be sufficient to alter its essential nature from watercourse to sewer. However, I do not understand that to be the question because there is no appeal from the judge’s conclusion on the second part of the preliminary issue. If, by contrast, the question is whether on the test propounded by Oliver LJ in BRB for a change from natural watercourse to sewer is satisfied, the increased flow of surface water through the culvert cannot be enough. Even if one adds the occasional presence of raw sewage pursuant to the CSOs to the natural groundwater and surface water, the test still cannot be satisfied. Nor, in my view, can it matter if some part of the channel upstream of the culvert is classified as a sewer for some specific statutory purpose unless the statutory test is the same as that propounded by Oliver LJ.
[89] For these reasons, I agree with the conclusion of Lloyd LJ and respectfully differ from that of Toulson LJ. I would allow this appeal.
[90] The fact that, by a majority, we are differing from the conclusion of the judge should not obscure the fact that he was presented with an impossible task and dealt with it as well as any judge could. The preliminary issue should never have been ordered to be tried in the form in which it was put forward or, perhaps, in any form. There was no common standard by which the issue as to “whether the culvert is a sewer” was to be judged. Each party was seeking to establish, avoid or shift liability for the alleged nuisance suffered by Raglan. They sought to establish, avoid or shift liability in a variety of ways to which the issue of whether the culvert was a sewer might be relevant but by reference to different meanings of the word. Second, the case had not been properly prepared by some or all the parties. Third, the judge was, for whatever reason, given inadequate time to pre-read or hear the case or to finalise his judgment. Fourth, the trial of the preliminary issue has merely increased costs and delay in the resolution of this dispute and was never likely to do anything else. These points are dealt with by the judge in [11] to [24] in tones of studied moderation. His moderation should not minimise the ample justification for his criticisms.
[91] We will hear further argument on the form of order that we should make to give effect to the conclusion of the majority. We shall wish to be addressed on whether we should merely allow the appeal and set aside the order of the court below or whether we should go further and make some declaration of our own and, if so, in what terms. We shall also wish to consider what directions for the future conduct of this action should be given so as to ensure that it is progressed to a conclusion without further delay and at the least expense.
Appeal allowed.
