LMT Surgical P/L v Allianz Australia Insurance Ltd [2013] QSC 181

[1] JACKSON J: When a back-flow not only prevents rain and surface water from draining through a storm water pipe but also overflows from the pipe, and there is damage caused by inundation of a nearby building, is that damage that was occasioned by or happened through flood? The answer to the question in the context of this insurance dispute depends on how the policy of insurance defined "flood".

[2] The defendant insurer contends that its liability to indemnify the plaintiff is excluded under the policy because the damage was "damage occasioned by or happened through flood" within the meaning of cl 3 of the perils exclusions ("the flood exclusion"). The plaintiff insured claims declarations as to the defendant's liability to indemnify for loss occasioned by water damage.

[3] On 11 January 2011, the plaintiff's business premises were situated at unit 11/97 Castlemaine Street, Milton. From about 6 pm on that day they were inundated by a combination of water from local run-off and from a back-flow of water from the Brisbane River through two local storm water drainage pipes (collectively "the pipes"). The water level rose to and above the level or levels of the relevant obverts for the pipes near the plaintiff's premises.

[4] Section 1 of the applicable industrial special risk policy, headed "Material loss or damage", provided:

In the event of any physical loss, destruction or damage (hereinafter in Section 1 referred as 'damage' with 'damaged' having a corresponding meaning) not otherwise excluded happening at the Situation to the Property Insured described in Section 1 the Insurer(s) will, subject to the provisions of this Policy including the limitation on the Insurers liability, indemnify the Insured in accordance with the applicable Basis of Settlement."

"In the event of any building or any other property or any part thereof used by the Insured at the Premises for the purpose of the Business being physically lost, destroyed or damaged by any cause or event not hereinafter excluded (loss, destruction or damage so caused being hereinafter termed 'Damage') and the Business carried on by the Insured being in consequence thereof interrupted or interfered with, the Insurer(s) will, subject to the provisions of this Policy including the limitation of the Insurer(s) liability, pay to the Insured the amount of loss resulting from such interruption or interference in accordance with the applicable Basis of Settlement."

"The insurer shall not be liable under sections 1 and/or 2 in respect of: ...

1. Physical loss, destruction or damage occasioned by or happening through:

(a) flood, which shall mean the inundation of normally dry land by water overflowing from the normal confines of any natural watercourse or lake (whether or not altered or modified), reservoir, canal or dam".

[7] Within the meaning of those provisions, a number of relevant matters are not in dispute. First, there was physical damage. Secondly, the damage was occasioned by or happened through inundation of land. Thirdly, the relevant land is normally dry land. Fourthly, the inundation was by water.

[8] The dispute resolved into the following questions:

1. were the pipes an "altered" or "modified" natural watercourse and, if so, was the inundation "by water overflowing from the normal confines" of that natural watercourse?

2. alternatively, were the pipes a "canal" and, if so, was the inundation by water overflowing from the normal confines of that canal?

3. alternatively, was the river a relevant natural watercourse and, if so, was the inundation "by water overflowing from the normal confines" of that natural watercourse?

[9] The premises are situated on the corner of Castlemaine Street and Black Street, Milton, an inner Brisbane suburb. Joint expert evidence of hydraulic engineers described the immediately surrounding area as a depression. As previously stated, it was flooded by the January 2011 flood event which affected the river. The area in general is susceptible to flooding during significant flood events from the river, as are other areas of Milton and the adjoining suburb of Rosalie.

[10] Adjacent to the premises, on its present alignment, Castlemaine Street runs approximately in a southerly direction to the point where it intersects with Milton Road. On the southern side of Milton Road is an embankment on which the Brisbane-Ipswich railway line is constructed. Milton Road and the parallel railway line run approximately in an east-west direction. On the southern side of the railway embankment is land which is relatively flat. In earlier times it was part of a swamp. The river lies to the south and runs downstream in approximately an east by north-easterly direction at that point. The northern bank of the river is thus over 200 metres to the south of the area of the premises.

[11] Prior to 1886, there was a small watercourse near the site of the premises that at that point ran downstream in a southerly direction. By 1895, works described as "Milton Drain", comprising a masonry lined open drainage channel, had been constructed on or to the east of and parallel to the watercourse, which had otherwise been filled in running towards Milton Road. Thus, Milton Drain drained towards the railway embankment. By 1930, it appears that Milton Drain was piped by an 8 foot pipe from Caxton St to Milton Rd where it ran under and through the railway embankment.

[12] From the southern side of the railway embankment, the watercourse had originally meandered towards the river, running approximately in a south-westerly direction for a short distance before it turned east and continued to meander until it reached the river bank at the approximate alignment of what has become Hale Street leading onto the Go Between Bridge.

[13] By 1895 another significant change had been made to that part of the watercourse on the southern side of the railway embankment. The open watercourse was replaced by an 8 foot pipe to a point near the river bank. However, in this instance, the piping did not follow the path of the prior meandering watercourse. Instead, it proceeded from the point of crossing of the railway embankment in a straight and southerly direction, continuing in the approximate line which the watercourse and Milton Drain had taken approaching the embankment from the north.

[14] As presently constructed, the two pipes from Castlemaine Street through to the river on that approximate alignment have diameters of approximately three metres and 2.4 metres respectively, depending on the particular section. The pipes receive storm water drained from a catchment of over 140 hectares. The catchment includes a low lying area to the north and which is adjacent to and surrounds the Brisbane City Council swimming pool, formerly described as the Ithaca Baths. The pipe or pipes run from there under Caxton Street, draining the land to the north and from east and west.

[15] The experts' opinions were that water inundated the premises between approximately 7:55 pm to 8:40 pm on 11 January 2011. The approximate percentage of river water (meaning water backed up via the pipes) at the time of initial inundation was between 10% and 14%, the remainder being stormwater or local runoff. Thereafter, as the river level quickly rose the percentage of river water in the mix rapidly increased.

[16] The plaintiff did not submit that the inundation was not by water from the pipes, for the purpose of the application of the flood exclusion. It submitted that "river water represented the dominant source of water that caused inundation of the subject premises above floor level."

[17] The genesis and object of the transaction is insurance of the plaintiff's property and business against the risk of damage and loss. The policy is a commercial contract. The applicable principles for the task of interpretation are:

"A policy of insurance, even one required by statute, is a commercial contract and should be given a businesslike interpretation. Interpreting a commercial document requires attention to the language used by the parties, the commercial circumstances which the document addresses, and the objects which it is intended to secure." (footnotes omitted)[1]

[18] The scope of the cover provided for under section 1 is not confined to particular risks of physical damage. The structure of the policy is such that the limits of the scope of cover under section 1 emerge from the application and operation of a number of exclusions, including the flood exclusion. This is a common structure for industrial risks and other forms of all risk property insurance.

[19] The risk of damage to property caused by water is one risk over which the parties to such a contract of insurance may bargain as they choose. Damage caused by flood is one common sub-set of the risk of damage to property caused by water. But there are other potential mechanisms or means of damage caused by water.

[20] The cover or exclusion of damage caused by flood is an everyday subject matter of contracting for insurance cover. The scope of the exclusion and the cover depends on the language deployed in the particular policy on that subject matter. Different insurers deploy different language for similar policies. An insurer and insured may bargain over the scope of the cover and the extent of the risk by adding to the cover provided for under the insurer's usual form of policy by indorsement. The permutations and combinations possible are many.

[21] The only present significance of these somewhat obvious points is that the particular allocation of risk presented by the meaning of the specific language deployed in the policy under consideration is not determined by the meaning of other policies which deploy other language or by broad statements as to purpose or object. The statement that a policy is intended to exclude damage caused by flood does not define the inquiry where the policy itself defines what is meant by flood. The surest approach is a close consideration of the contractual text in its context.

[22] Were the pipes a natural watercourse as modified or altered?

[23] The process of piping the watercourse which flowed past the area of the premises and continued to the northern bank of the river changed that natural watercourse. Reprising, from the descriptions set out above, the changes included:

(b) piping from north of the area of the premises to the south up to the point of the railway embankment;

(d) piping from the southern side of the railway embankment to the northern river bank;

(e) changing the alignment of the prior watercourse from the southern side of the railway embankment to the northern river bank;

(g) repairing or replacing the pipes at various stages.

[24] There is no visual remnant of the original watercourse. The connection between the watercourse and the pipes which replaced it is largely functional. The pipes serve the drainage function previously served by the watercourse. The catchment may be altered but will be similar to that created by the original natural topography of the area. The evidence did not explore that comparison.

[25] The point at which changes to a watercourse are such that the outcome is not a modified or altered watercourse but something else is not capable of precise definition. Just because a storm water drainage pipe fulfils the drainage function that a natural watercourse previously filled does not constitute it as a "natural watercourse (whether modified or altered)".

[26] Outside the premises in the present case the pipes replaced the function which had originally been performed in that vicinity by the natural watercourse. The process of change had started with the construction of the Milton Drain. The evidence about that is not precise. But by the 1930s the open drain had been piped and covered. Other drainage works necessary to drain into the piping must also have been carried out. In the 1990s a second parallel pipe was placed alongside the first pipe to improve the drainage. The construction of the Milton Drain or piping the watercourse in the vicinity of the premises alone might have taken it outside the meaning of "natural watercourse (whether modified or altered)".

[27] As well, the installation of the original pipe from the southern side of the railway embankment to the northern bank of the river did not follow the path of or arrive at the same destination as the prior natural watercourse. The insurer did not contend that consideration of the changes to the natural watercourse should be confined to the immediate vicinity of the premises.

[28] In my view, the overall changes broke the link to the prior natural watercourse. It is not an ordinary use of language to describe the pipes so located as the natural watercourse, or as that natural watercourse, modified or altered. The pipes as they are constituted are a functional replacement for the prior natural watercourse through the area, which was filled in.

[29] Were the pipes a canal? The insurer relied upon both the meaning of the latin root of canal, that is "canalis", and upon dictionary meanings of "canal" that include a pipe or tube for conveying fluids.

[30] Construing "canal" in the flood exclusion to include a pipe would significantly affect the operation of the flood exclusion. For example, if a water supply pipe were a canal, damage through inundation by water overflowing from the normal confines of a water supply pipe which bursts or is leaking would be excluded. The insurer sought to confine the classes of pipes to which the meaning of canal would extend, but there was no warrant for that distinction which I could discern from either the text or context of the flood exclusion.

[31] In the flood exclusion, "canal" is used in association with other man-made works which hold water on land, being a reservoir or dam. Although a tank which makes up part of a closed hydraulic system may constitute a reservoir, and there are other meanings of that word in the context of medicine and the natural sciences, a primary meaning of reservoir and the relevant meaning in the context of the flood exclusion is a natural or man-made lake or pool used for collecting or storing water, often for public or industrial use. Similarly, a dam is commonly understood to be a barrier constructed to hold back water to create a body of water confined by the barrier. In association with those words, a reference to "canal" would also connote a man-made water holding structure, and through which water or traffic passes. It would not ordinarily be the word used to refer to an underground storm water drainage pipe.

[32] In my view, in the context of the flood exclusion "canal" does not include the pipes.

[33] It is not in dispute that the river is a natural watercourse, whether modified or altered. Water from the river backed up into the pipes via the outlets on the northern bank of the river. As the river level rose, river water filled the pipes. The point was reached where water from the river discharged from the pipes in the area of the premises, because the obverts of the pipes were below the level of the river, even though the river had not overtopped its northern bank in the area. The water from the pipes then overflowed onto the land and inundated the premises.

[34] Was that overflowing "from the normal confines of [the river]"? It might be said that the water overflowed from the pipes but not from the normal confines of the river, in the sense that the relevant overflowing is required to be from the banks as the normal confines. But the water has left the normal confines of the river. And it has overflowed onto and inundated the premises.

[35] A requirement that the damage caused is by water overflowing might not refer to inundation by overflowing the banks as opposed to water which has risen through a pipe. The requirement of overflowing, as a matter of ordinary meaning, excludes water from the atmosphere, and connotes a surface flow over relevant land. The requirement that the water overflowing be from a natural watercourse or other identified sources is intended to exclude surface flows from other unidentified sources, including natural run-off not sourced from a natural watercourse. But a surface overflow from a river via a pipe is no less an overflow from the river than a surface overflow by overtopping the banks of the river.

Overflowing not from the normal confines of the river

[36] The most troubling part of the flood exclusion is the requirement that the water overflowing be "from" the "normal confines" of the relevant source, at least where that source is a natural watercourse. What, for example, would constitute water overflowing from the river but not from the normal confines of the river?

[37] The river flooded, in the sense that the water levels rose to flood levels and inundated land in the vicinity including the premises. However, the flooding of the area of the premises did not occur because of the overtopping of the river bank resulting in overland flow to that area.

[38] Instead, the flooding occurred because water backed up through the pipes, due to the water level in the river rising to a level above the obverts of the pipes at some drainage point or points in the vicinity of the area of the premises.

[39] Was the inundation by water which backed up in the pipes an inundation by water overflowing from the normal confines of the river? If the pipes are treated as being within the normal confines of the river, it would follow that water backing up through the pipe and overflowing onto the premises was water overflowing from the normal confines of the river.

[40] It is, of course, readily foreseeable that a river in flood will enter and back up the pipes of a storm water drainage system which usually empty into the river. In a natural state, the banks of a river are its normal confines. A flood plain onto which it overflows would not be part of the normal confines. There may be man-made works which could affect the normal confines, such as levee banks intended to operate during flood events.

[41] However, a stormwater drainage pipe system would not be commonly thought of as part of the normal confines of the river. The result is not different, in my view, because the pipes in question function in place of a prior natural watercourse, which has been filled and the pipes covered. Even if the prior natural watercourse still existed and flooding were to occur because water from the river backed up into the watercourse, that watercourse would not be the normal confines of the river. It would have its own normal confines as a watercourse.

[42] Thus, in my view, the normal confines of the river did not include the pipes.

[43] What is intended by requiring that the inundation be by water overflowing from the normal confines of a natural watercourse? One possible example is where water inundates by overflowing but was never within a river because the flood water within the river prevents the inundating water from flowing into the normal confines. In a sense, the water may be said to be overflow from the river, being part of the swollen stream, but it could not be said to have been from within the normal confines. If this is the purpose of the requirement that the damage is by water overflowing from the normal confines, it would exempt from the operation of the flood exclusion inundation caused by water which was never within the normal confines of the river, but might not exempt inundation by water overflowing from the normal confines of the river via the pipes.

[44] An alternative and narrower view is that the requirement, that the inundation be by water overflowing from the normal confines, restricts the operation of the flood exclusion to inundation by water which has overflowed the banks where those banks are the normal confines. On this approach the phrase "normal confines" may be seen as wider than "banks" because of the possibility of artificial works comprising part of the normal confines such as levees or because "normal confines" also qualifies the other water sources identified in the flood exclusion, namely a reservoir, dam or canal.

[45] The ordinary meaning of the text supports the narrower view. That view gives effect to the syntax of the provision that the inundation is by "water overflowing from the natural confines". The prepositional phrase "from the natural confines" modifies "overflowing", not "water". The ordinary meaning of the words is directed to the place from where the overflowing occurred, not the place from where the water was sourced.

[46] Is there a contextual reason to adopt the wider meaning? None that I can find. Is the answer affected by other principles of contractual construction which may apply to an insurance contract as a commercial contract? Because the ordinary meaning is the narrower view, in my view it is unnecessary to consider whether the principle that an exemption clause should be strictly construed or the principle that an ambiguous clause should be construed against an insurer as the proferens has any role to play in the contractual construction of this commercial insurance policy.

[47] I accept, as stated above, that a commercial contract should be construed in a businesslike way. But I do not consider that one view or the other in this case is more businesslike. In the end, it is a matter of construction of the text in context applied to the facts. The task is also not assisted by referring to cases decided upon other clauses in materially different language in similar policies, except to the extent that such cases show the approach that should be taken to the precise constructional question. For that reason, I have not essayed the application of the cases on other clauses or facts to which I was helpfully referred in argument.[2]

[48] In the result, in my view the insurer's liability to indemnify the plaintiff for the damage occasioned by inundation of the premises in the present case was not negated by the flood exclusion. As discussed above, the inundation of the premises in the present case through back-flow from the pipes running from the river bank, was not by water overflowing from the natural confines of the Brisbane River.

[49] It follows that the claim must succeed and the applicant is entitled to an appropriate form of declaration.

[1] McCann v Switzerland Insurance Australia Ltd [2000] HCA 65 at [22]; [2000] HCA 65; (2000) 203 CLR 579 at 589; Wilkie v Gordian Runoff Ltd [2005] HCA 17 at [15]; [2005] HCA 17; (2005) 221 CLR 522 at 528-529; CGU Insurance Limited v Porthouse [2008] HCA 30 at [43]; [2008] HCA 30; (2008) 235 CLR 103 at 116.

[2] Provincial Insurance Australia Pty Ltd v Consolidated Wood Products Pty Ltd (1991) 25 NSWLR 541; K Sika Plastics Ltd v Cornhill Insurance Co Ltd [1982] 2 NZLR 50; Hams v CGU Insurance Ltd [2002] NSWSC 273; Peterson v Union des Assurances de Paris IARD (1997) 9 ANZ Ins Cas 61-366; Elilade Pty Ltd v Nonpareil Pty Ltd [2002] FCA 909; (2002) 124 FCR 1.