Rockwater Ltd v Coflexip SA & Anor

This action was commenced by Rockwater Limited (“Rockwater”) which sought the revocation of European Patent (UK) 0 478 742 entitled “Device and Process for Unrolling Flexible Tubular Conduits Essentially Vertically”. The defendant is Coflexip SA (“Coflexip”), the registered proprietor of the patent. Coflexip and Technip Offshore UK Limited, which is said to be the exclusive licensee under the patent, counterclaimed for infringement. That infringement involves Rockwater’s ship, the Toisa Perseus, which was deployed in the execution of two pipelaying contracts in the summer and autumn of 2001. One contract was carried out in the Leadon field in water depths of about 120m, and the other in the East Foinaven field in water depths of about 300-500m. The jurisdiction under s.132(4) of the Patents Act 1977 is not in dispute.

The patent has been litigated before in this country. Coflexip sued Stolt Comex Seaway MS Limited and two associated companies (together referred to as “Stolt”) in respect of the operation of Stolt’s pipelaying ship, the Seaway Falcon. Stolt attacked the validity of the patent relying in large part on allegations of obviousness based upon the prior use and disclosure of another pipelaying vessel called Apache. That action came before me. In a judgment dated 22 January, 1999, I found for the Coflexip both in relation to infringement and validity. That decision was upheld by the Court of Appeal, save that my finding of non-infringement in relation to claim 3 of the patent, was reversed.

Whatever the outcome of the proceedings against Stolt, this is a different action which must be determined on the basis of the evidence relied on here. It should be noted that the alleged infringement relates to a vessel which is by no means to the same design as Stolt’s Seaway Falcon. Furthermore, although allegations of anticipation and obviousness are raised here, the prior art relied on is quite different to that relied on in the previous action. There the primary prior art was the Apache. Here Apache is not pleaded. Instead, the primary prior art relied on consists of US Patent No. 4,721,411 and UK Patent Application No. 2,178,129, referred to during the trial as Recalde US and Recalde GB respectively

The technical background to the patent is set out below. Although part of this is derived from the opening passages of my judgment in the Stolt action, it has had to be revised and expanded in certain areas to take into account the different evidence and different issues which arise in this action.

The patent is concerned with a process and equipment for laying pipes and other conduits in deep water. It has particular application in off shore oil and gas fields. A variety of pipes and cables, some of very great length, may have to be used, for example, to connect the well to a site on dry land. The patent in suit is concerned particularly with the laying of flexible pipes and cables from a pipelaying ship and overcoming the problems associated with passing the pipe or cable from the ship into the water. It is convenient to start by considering the properties and behaviour of submarine pipes.

There are different types of submarine pipes. Although they have many features in common and the same laws of physics applies to them all, in the art they are divided into two broad classes, namely rigid pipes and flexible pipes. The former are normally made of steel. Sometimes they are coated in concrete or plastics materials. They are capable of being laid in very deep water. The latter are normally made up of a number of layers of composites and reinforcing materials such as steel braids. Because their walls are made up of a number of interacting layers, those walls tend to be very thick. Most members of the general public would regard them as stiff or rigid. However, in the technical world of submarine pipes, they are considered to be flexible. As a practical matter those in the art have little difficulty in distinguishing one type of pipe from the other. There are difficulties associated with laying both, particularly in deep water. The two broad classes of pipes cannot be distinguished on the basis of size alone. Each can be made in a variety of sizes and their size ranges overlap. However, as Mr Nash, Coflexip’s expert, explains, because of their thick walls, flexible pipes are much heavier than the same sized rigid pipes.

The differences between typical rigid and flexible pipes can be explained as follows. A copper pipe used in a domestic water system can be regarded as rigid. If one end is clamped and the pipe is not too long, it will maintain an almost horizontal position without support at its free end. Its reluctance to bend can be referred to as its bending stiffness. If the end of the pipe is pulled down a bit it will flex. It is a bit “springy”. When the force is removed, it will return to its original position. However, if more force is applied, a point will arrive at which parts of the pipe’s surface will begin to stretch. It can be made to take up a permanently bent configuration. A plumber will do this by applying force to a copper pipe in a pipe bender. It is said to have been subject to plastic deformation. These characteristics were explained by Mr. Nash in his first report as follows:

“15.

Rigid pipe has a finite bending stiffness. That is to say, if a sufficient load is applied to a length of rigid pipe it will deflect (i.e. bend). A rigid pipe can be bent elastically up to its elastic limit. This means that provided the pipe is not bent beyond its elastic limit (or yield stress), it will return to its original shape after the bending force has been removed. If a rigid pipe is bent beyond its elastic limit, this will result in the pipe being plastically deformed. This means that if the bending force is removed, the pipe will not return to its original profile, ie a permanent curvature will be induced into the pipe.

16.

A plastically deformed pipe can, within limits, be returned to its original profile. This is achieved by applying a sufficient bending force to the pipe in the opposite direction to that which originally caused the plastic deformation. In other words, if a straight section of pipe has been plastically bent one has to overstress the pipe (i.e. apply a stress greater than the yield stress) in the opposite direction in order to straighten it. This process, if carried out correctly, will not affect the pressure containing properties of the pipe, nor its resistance to hydrostatic forces. However, if a rigid pipe is plastically bent beyond a certain minimum radius of curvature, called its ultimate bending radius, the pipe will suffer permanent localised buckling or crimping. This will irreparably damage the integrity of the pipe.”

These characteristics are exhibited by rigid steel submarine pipes. As long as they are not subjected to too much force they will be springy. If more force is applied they can be plastically deformed so as to take on a permanently bent shape. If they are bent beyond their ultimate bending radius, they buckle. It may take quite a lot of force to plastically deform a rigid pipe. If such a pipe is plastically deformed so that it adopts a curved shape, it will need to be plastically deformed again to return it to its straight configuration. Depending on the size and composition of the pipe, this also may involve the use of a lot of force.

The behaviour of a typical flexible pipe is somewhat different. The bending stiffness of flexible pipe is far less than that of rigid pipe. The concepts involved can be explained sufficiently accurately for the purpose of this action by reference to the behaviour of a garden hose. If it is held horizontally at one end it will tend to droop. It has a much smaller ultimate bending radius than a rigid pipe, so it will have to be bent much more acutely before it buckles. On the other hand it is less readily plastically deformed. If bending force is removed from a flexible pipe or hose which has not buckled it will tend to return to its original shape. It is less prone than rigid pipes to acquiring a permanent bent shape. However if a garden hose is fed over a hose reel and a weight is put on the free end, the latter will tend to pull the hose down onto the reel. The greater the weight, the more pronounced this effect will be. Because the walls of the hose are comparatively flexible this has the effect of squashing the hose onto the reel. The cavity in the centre of the hose will become oval and, at some point, it will close completely. When this happens the pipe is said to be crushed. Again these characteristics were explained by Mr. Nash in his first report:

“18.

The bending stiffness of flexible pipe is several orders of magnitude less than that of rigid pipe. This is due to the lower bending stiffness of the materials which make up the flexible pipe and also because each layer is, to a degree, able to move relative to its neighbouring layer (in non-bonded flexible pipe). This means that flexible pipe can be bent to a much smaller radius of curvature than rigid pipe without exceeding its elastic limit. If one were to bend a flexible pipe beyond its elastic limit (also referred to as damaging bend radius or minimum bend radius), the pipe would be irreparably damaged so that it could not be straightened back to its original shape and its pressure containing properties would probably be impaired. However, flexible pipe can be bent to a very much smaller radius of curvature than rigid pipe without reaching its damaging bend radius. There is no plastic deformation with flexible pipe.”

Rockwater’s expert, Professor Witz, explains that the distinction between rigid and flexible pipes is not sharply defined. He says that when bending is applied to rigid and flexible pipe, there is much in common ground in their resistance and behaviour. Both flexible and rigid pipe exhibit finite bending stiffness. He disagrees with Mr Nash’s statement that there is no plastic deformation with flexible pipe. If a flexible pipe is subject to severe bending (i.e. it is subject to a “sharp” as opposed to a “gentle” bend), the innermost steel carcass may seize and plastically deform. But he accepts that the limiting bending criteria for rigid pipe are reached at a bending radius higher than the corresponding bend radius for an equivalent flexible pipe and that flexible pipe will have different parameters to rigid pipe. He points out that buckling and crushing can occur in both rigid and flexible pipes. However, because the walls of a rigid pipe are, by definition, comparatively rigid, they withstand crushing better. Thus rigid pipes are less sensitive to crushing and more sensitive to plastic deformation and buckling than their flexible equivalents. As was not in dispute, before laying a new pipe, load calculations should always be carried out to ensure that the pipe can be laid successfully with the apparatus available.

When rigid submarine pipes were first laid in water from ships, they were fed off the back of the ship in a more or less horizontal direction. The ship is loaded with numerous lengths of straight pipe which are welded one by one on board to the end of the pipe being fed into the water. As more and more pipe is fed off the back of the ship, the weight of the unsupported pipe grows. This pulls down on the end of the pipe causing it to curve. This curvature of the pipe at the end near the laying vessel is called the overbend. To avoid permanent deformation here, the pipe is supported by a long curved guide, called a stinger. The stinger maintains a set minimum radius of curvature so that there is no plastic deformation of the pipe. With very rigid pipes (e.g. pipes of large diameter) being laid in very deep water, the stinger might have to be very long indeed. For example, a stinger of about 130 metres might be needed to lay rigid pipe in water 1000 metres deep. In such a case, the tension created by the 1000 metres of suspended pipe could be in the region of 60 metric tonnes. For that reason, where rigid pipe is to be laid in deep water, the stinger could extend well beyond the stern of the pipe-laying vessel. This could compromise the vessel’s stability, particularly in rough weather.

Furthermore in all cases steps have to be taken to ensure that the pipe does not contact the end of the stinger (or is carefully controlled on contact). The reason for this is that if the vessel suddenly backed up or moves forward too slowly relative to the pipe pay-out speed, there is a risk that the pipe will be bent more sharply (i.e. be subjected to a smaller bend radius) at the end of the stinger, thereby risking damage to the pipe. It may buckle at that point. Similarly, in adverse weather, the vessel may pitch and roll. If the pipe is in contact with the very end of the stinger and the stern of the ship rises as the prow falls, there will be a tendency to bend the pipe at an acute angle (i.e. a small bend radius) at the end of the stinger where it is no longer supported.

The greater the depth of the water and the bigger the pipe, the greater the weight of the suspended pipe. Furthermore the vessel will be pulling in the direction of lay. The weight of suspended pipe and the tension imparted by relative movement of the vessel away from the pipe lying on the seabed will tend to pull the pipe towards the sea bed and off the ship. In some cases, the weight of the suspended pipe can be reduced by adding floatation devices to it as it is paid out. Indeed, pipe is normally laid in a sealed condition in which it is full of air since this will give it some buoyancy. Notwithstanding the adoption of these procedures, there will continue to be a substantial tension in the pipe tending to pull it off the vessel. This has to be prevented by means of some device on the vessel which pulls in the opposite direction. This can be done by a mechanical “hand” which pulls on the end of the pipe as it leaves the ship and stops it going overboard. The mechanical hand is called a tensioner. It must have the capacity to match the tension trying to pull the pipe off the vessel. Using more technical terminology, the tensioner “reacts” the tension in the pipe. In other words it acts as a brake on the pipe. Whichever terminology is used, the concept is the same; the suspended pipe under the influence of gravity pulls towards the seabed. To prevent the pipe moving in that direction in an uncontrolled manner, an equal and opposite pull has to be exerted, for example by the tensioner. The pipe can be likened to the rope being pulled by two teams in a tug-of-war. A great deal of energy may be expended by each team, i.e. each is applying a lot of tension in opposite directions, even though the rope does not move at all or only moves slowly. It can be said that the tension created by one team is being reacted by the other and vice versa. If either team lets go, the other will fall backwards. Similarly, if the pipe under the vessel is severed near the surface, the seaboard side will fall to the seabed and the vessel-side pipe will be brought back sharply on board by the effect of the tensioners.

Thus the tensioner must have the capacity to exert a tension on the pipe which matches (i.e. reacts) the tension generated by the suspended pipe as it is being laid. In fact, it may need to have a higher capacity than that. As mentioned above, when the pipe is being laid it will normally be full of air. This gives it buoyancy. However the tensioner may be used to recover the pipe from the seabed in which event the latter may be full of water. The buoyancy will have disappeared. The pipe may now have an effective weight considerably greater than it had when being laid. The tensioner may be made up of a number of smaller tensioners in sequence capable of exerting a total pull which exceeds that of the suspended weight of pipe. Pipelaying in this fashion can be illustrated as follows:

In this Figure, the pipe-laying ship is being driven towards the right. If it were not, the end of the pipe where it leaves the stinger would be pulled hard down by the weight of the pipe below it. It would therefore be pulled down and might hit the end of the stinger. This could cause high localised forces which tend to buckle the pipe. This horizontal thrust of the ship results in the pipe lifting off the stinger before it reaches the end of it. Because the horizontal thrust lifts the pipe off the stinger (in other words reduces its tendency to fall down vertically), this can be used to reduce the overall length of the stringer to some extent. However increased horizontal thrust means greater use of the vessel’s engines and this costs money. The location at which the pipe lifts off the surface of the stinger to enter the water is called the pipe suspension or lift off point. It will be appreciated that the location of this point will vary from time to time throughout the laying operation. The lighter the pipe being suspended and the greater the forward thrust of the vessel’s engines, the “flatter” the take off trajectory of the pipe. This means that the pipe suspension point moves up the stinger towards the vessel. On the other hand heavier pipe (for example pipe being laid in deeper water) and lower engine thrust will result in a steeper take off trajectory of the pipe. In such circumstances the pipe suspension point moves down the stinger towards its seaboard end. Changes of the pipe suspension point will also be caused by pitching of the vessel due to adverse weather conditions. Professor Witz explained during his oral testimony that the stinger will be sized so the lift off point is broadly in the middle and the installer has some degree of latitude in adjusting the horizontal tension and, therefore, the effective lift off point to keep the pipe within its lay configuration.

Figure 1 above also illustrates how the rigid pipe flexes. It describes an “S” shape in the water. As a result, this type of rigid pipelaying is called “S-lay”. The deeper the water, the nearer the vertical the central section of the pipe will be as illustrated below.

Because in deep water the pipe will tend to take up the shape shown in Figure 2, an alternative type of rigid pipelaying can be used in such situations. The angle at which the pipe is fed off the end of the ship is matched, as nearly as possible, to the natural angle which the pipe will take up in the water (its so-called catenary shape). The result is that the pipe is fed off the end of the ship at an angle to the surface of the sea. This is illustrated in Figure 3 below. Because the pipe takes up a generally “J” configuration, this form of pipelaying is called “J-Lay”.

Needless to say, the deeper the water, the more near the vertical the top part of the pipe will be. There is a minor dispute between the parties as to how extensively J-Lay was practised before the priority date of the patent, but there is no dispute that it was well recognised as a possible way of laying rigid pipe and had been used.

In both S-Lay and J-Lay one of the objectives is to prevent the pipe being bent at too acute an angle (i.e. with a small bend radius). If this is not avoided, there is a risk of the pipe buckling. A similar problem exists at the seabed. It will be seen in all the Figures above that the pipe curves up from its horizontal position on the seabed. This is called the sag bend. Once again, the pipelaying must be conducted in a way which ensures that the bend radius at this point is not too small for the pipe being laid. This is a point which will be considered more fully in a moment.

Up to this point, pipelaying of rigid pipe has been described which involves welding together straight lengths of rigid pipe. Rigid pipe can be bent and spooled onto a large reel. This involves plastic deformation of the pipe. If the pipe from such a reel is laid from a pipe-laying vessel, it will need to be plastically deformed back into its substantially straight configuration. The reel is supplied with drive motors. These may be used to spool pipe onto the reel. They can also be used to apply axial pull to the pipe so as to counter or react the axial tension created by the pipe suspended under the vessel.

With flexible pipe the standard form of laying used in modest depths of water is to pass it more or less horizontally from a reel (sometimes called a winch) over a curved chute (sometimes called an “overboarding gutter”) or wheel (called an “overboarding wheel”, “laying wheel” or “sheave”) and down into the water. The catenary of such pipe from the point where it leaves the gutter or wheel to the sea bed is J-shaped although, because of the shape of the path taken by the pipe from the reel to the seabed, this is sometimes referred to as S-Lay. Once again, to prevent the weight of pipe under the ship from stripping the rest of the pipe from the reel, some form of tensioning device is used to react the tension created by the suspended pipe. This can take the form of tensioners or the wheel on which the pipe is loaded can be powered. The “Recommended Practice for Flexible Pipe” API 17B published by the American Petroleum Institute in June, 1988 illustrates both of these methods and describes them as the most common. These illustrations are set out as Figures 4 and 5 below.

In 1990, powered reels were not suitable to lay flexible pipe in deep water where a tension capacity of more than typically 30 tonnes was required. Mr Coutarel, Coflexip’s Product Research and Development Manager, said that at that time the tension capacity provided by the reel was around 10-30 metric tonnes. This was to be compared with linear tensioners. Mr Nash explained that in the Norske Shell Draugen field in about 1992, tensioners of 60 tonne capacity were used and Mr Coutarel explained that Coflexip used a tensioner of 125 tonne in 1991. Furthermore tensioners can be used in series. Accordingly, linear tensioners were used to provide the tension when greater than 30 tonne tension capacity was required. An arrangement using tensioners is illustrated in Figure 5 below.

The use of the laying wheel illustrated in this Figure and the chute in Figure 4 is to prevent the pipe from being crushed (i.e. flattened) by the effect of the weight of the line below the ship. The increased crushing forces produced by laying heavier pipe in deeper water can be offset by increasing the radius of the chute or wheel over which the flexible pipe is overboarded into the water. This was described by Mr Nash in his first report in a passage which he said was common general knowledge:

“Crushing of the pipe at the overboarding point

36.

As the water depth increased, the tensile load in the pipe also increased due to the greater weight of suspended pipe. The increased tensile load became a particular concern where the pipe passed across the overboarding gutter (or wheel). The gutter provided the reaction to a combination of the tensile load in the pipe between the gutter and the installation reel or tensioner, and the free hanging pipe suspended from the gutter. This reaction was spread approximately uniformly over the length of the contact between the pipe and the gutter. If the radius of the gutter and, therefore, the pipe contact length, were to remain the same, the crushing force per unit length of contact increased as the water depth and pipe suspended weight increased. To avoid or minimise the increase in crushing load as the water depth increased, one had to increase the radius of the gutter, and hence the contact length. In this way the increased crushing load was compensated by an increase in the support length.”

In the case of flexible pipes, just as with rigid pipes, there will be a sag bend at the junction between the pipe lying horizontally on the seabed and the near vertical portion leading down from the pipe-laying vessel. Again, if this sag bend is too acute, there is a risk that the pipe will be deformed at the bend. To avoid this, the vessel is driven in the pipe-laying direction so as to impart tension to the line in a horizontal direction. It will be appreciated that the same effect is achieved with rigid pipe. This is, in substance, a mirror of what happens on the stinger – as the vessel is driven forward the pipe takes a more shallow trajectory. Whether considering rigid or flexible pipe, additional tension applied in the direction of lay reduces the angle of the sag bend and thereby reduces the risk of damage to the pipe at that location. This effect is illustrated in the following figure.

The drawing on the left illustrates a case where there is no forward motion of the pipe-laying vessel. The one on the right is an exaggerated depiction of the effect of driving the vessel to the right. In the left hand drawing, the only tension in the pipe is the vertical one trying to pull the pipe towards the sea. The vessel has to counter or react this tension. In this case it is the reel to the right of the vessel which reacts that tension. In the right hand drawing, the tension in the pipe can be considered to be made up of two components. One is the vertical component generated by the effect of gravity on the suspended pipe. It is pulling the pipe towards the seabed. The second is a horizontal component which is trying to pull the pipe to the left. Both of these components have to be reacted on the pipe-laying vessel. Mr Nash explained the use of horizontal forces to avoid problems at the sagbend as follows:

“In order to prevent the pipe buckling at the sagbend a horizontal tension was applied to the pipe by tensioners situated on the deck of the vessel. In this way a minimum radius of curvature was maintained at the sagbend. This avoided the risk of the weight of the pipe tending to straighten itself vertically and creating overstress in the sagbend (caused by a small radius of curvature).” (First Report paragraph 85)

Although in this passage he was discussing the sagbend problem in relation to rigid pipes, the same principles apply to flexible conduits as well.

Flexible pipes are normally made in the factory in very great lengths. Mr Coutarel gave evidence that a flexible pipe with an 8 inch internal diameter may be made in individual lengths up to 10 km and a 12 inch pipe may be made in individual lengths of up to 5 km. These lengths have to be joined together. Each length of pipe has an end fitting at each end. To extend a flexible pipe, the end fitting on one length must be connected to the end fitting on another to make what is known as an intermediate connection. Mr Coutarel explained that end fittings vary between 0.5 and 4 metres in length depending on the size of the pipe and its application. So an intermediate fitting may be from 1 to 8 metres in length. End fittings are made of steel and usually have a diameter twice the diameter of the flexible pipe to which they are fitted. They are rigid. Other types of rigid accessories which are installed on flexible pipe during production are bending stiffeners which are used on pipes called dynamic risers. Bending stiffeners are generally up to 7 metres long and are made of steel and polymer material. The diameter of a bending stiffener is normally 3-4 times the size of the diameter of the flexible pipe. There are other rigid accessories such as buoyancy modules and anodes.

These rigid accessories have to be overboarded. This means that they have to be fed over the gutter or wheel and fed into the water. If such an accessory is fed round a wheel or gutter, the tension in the pipe will act on the pipe/accessory junction to give very high buckling forces over a very short distance. The problem is illustrated in Figure 7 below.

This propensity for local buckling has been a serious problem where the laying of flexible pipe is concerned. A significant number of possible solutions have been proposed and quite a few of them have been put into practice. Coflexip gave evidence relating to, and produced drawings of, some of these. Mr. Coutarel explained that one way of overboarding the rigid accessories without damaging the pipe is to use a crane or the so-called A-frame which is sometimes located at the stern of the vessel. As the rigid accessory is unrolled from the reel, the laying operation is halted before the rigid accessory reaches the overboarding gutter. The crane hook is attached to the rigid accessory. In this way the tension load from the weight of the suspended pipe is transferred from the reel to the crane which then operates to lift the rigid accessory up and over the overboarding gutter. The crane then lowers the pipe and the rigid accessory attached to it below the level of the overboarding gutter. If the rigid accessory is an intermediate connection between two lengths of pipe, the tension is then taken up once again by the reel and the crane is disconnected from the rigid accessory. The laying process then continues. In this case, during the operation of the crane, the powered reel no longer pulls the pipe. The load is taken by the crane.

Another arrangement used and published before the priority date of the patent in suit operates as follows.

This shows the flexible pipe being fed from the tensioners at the right towards the laying wheel. The drawing illustrates a case in which some 300 tonnes of pipe is suspended off the end of the vessel. The rigid accessory is approaching the upper surface of the laying wheel. At this point a lift off device which is located between the tensioner and the laying wheel and is normally not in contact with the pipe, is lifted up so as to support the underside of the pipe and push it in an upward direction. This has the effect of lifting the right hand end of the accessory. This is illustrated in Figure 8(b) below.

The lift off device continues to rotate until it is in a position to allow the accessory to travel vertically past the laying wheel as shown in figure 8(c) below.

Finally the lift off device is pulled back to its original position.

It will be seen that although the pipe and the accessory are kept under tension at all times, the arrangement avoids there ever being an occasion on which there is any bending at the accessory/pipe junction.

Further methods for overboarding involve moving the accessory round the laying wheel or overboarding gutter under very little tension. One method of doing this, which was the subject of a Coflexip patent, is illustrated below.

A collar is fitted around one of the end fittings and cables (9) are attached to it in order to connect it to the winch used when a pipe has to be abandoned, e.g. because of bad weather, and recovered (hence the name “A&R” winch). The tension load is transferred from the tensioners to the A&R winch. The tensioners are then opened and the A&R winch cable is unreeled to allow the intermediate connector to pass through the tensioners. Unreeling is continued until the collar reaches a tilting frame (5). When the collar reaches the tilting frame it automatically engages with the frame. By continuing to unreel the A&R winch cable, the tilting frame is able to rotate and thereby lift the intermediate connector off the wheel. The tension applied inboard of the tilting frame, e.g. by the tensioners, is turned off or reduced. This means that there is bending but no significant tension on the vessel side and tension but no bending on the sea side of the intermediate connector. It will be appreciated that this is similar in principle to the use of the A-frame discussed in paragraph 28 above. This is illustrated in the following Figure.

Once the tilting frame has completed its rotation, the collar automatically disengages from the frame. The tension is transferred back to the tensioners and the A&R winch cable is disconnected. The tilting frame is brought back on board and normal laying is resumed.

From what has been said above, it can be seen that among the problems facing those wishing to lay flexible pipe were the following. First there is a need to avoid bending the pipe at the interface between the pipe and the rigid accessory at the same time as a high axial tension is being applied at that point. If a high axial tension is applied to the pipe at the interface at the same time as it is subjected to bending stress, there is a risk that the pipe will become damaged. Second there could be a problem which arises out of the radial contact loads induced by the axial tensile load (pipelay tension) when the flexible pipe is bent around the overboarding gutter or wheel. These contact loads are generally acceptable in shallow or medium water depth, but they tend to crush the pipe in deep water. The same problem exists if high axial tension is applied to the pipe while it is spooled on its reel.

The patent was analysed and the claims were construed in the Stolt action. Neither Mr Miller QC for Coflexip nor Mr Thorley QC for Rockwater has invited me to depart from any of the relevant holdings made there. However the issues of validity and infringement in this action are different and matters which were of no or little importance in the Stolt action are central to this dispute. For that reason, and to ensure that this judgment is self-contained, it is necessary to look again at the patent and its claims.

The opening paragraph of the specification states that the invention relates to a method and a device for effecting the laying of “flexible conduits, in particular of tubular flexible conduits”. Later it says that the invention is to cover “the laying of cables, for example traction cables and above all electric cables. Electric cables are understood to mean any cable comprising electrical conductors, as well as power cables and cables carrying information”. In the Stolt action, the Court of Appeal held that the latter passage:

“establishes that the words “flexible conduit” should be interpreted as including more than what was known as flexible pipe, but not so widely as to cover what was known in the industry as rigid pipe.” (paragraph 18)

Only one point need be added to that. There is nothing in the specification or the claims which imposes a lower or upper limit on the diameter of flexible conduits covered by the patent nor is there any limit on the type of construction adopted or, for example, the surface characteristics of the conduit. The expression is wide but excludes rigid pipes.

The specification goes on to describe some of the prior art and to identify some of the problems which are to be addressed by the invention. It refers to the methods of laying of the type illustrated in Figures 4 and 5 above and comments that:

“For the purpose of being able to lay flexible conduits with relatively large diameters in great depths of water, one has to use tensioning means and deflecting elements whose dimensions and cost pose problems and which create an excessive space requirement on the bridge of the laying vessel” (page 1 lines 33 to 36)

As explained above, a concern when laying flexible pipe in deep water is the possibility of crushing. This can be addressed by employing overboarding wheels or chutes of larger diameter. This is addressed in the specification. It comments on the effect of laying such large diameter flexible pipes in deep water on the dimensions of the pipe reel or winch on which the pipe is stored onboard:

“Such a winch, as well as the deflecting element, assume dimensions and space requirements that are excessive when the diameter of the conduit and the depth of water increase. As the diameter of the tubular flexible conduit and the depth increase, the size of the wheel becomes increasingly larger. Such a wheel may have a diameter of the order of 10 metres for a depth substantially equal to 500 metres.” (page 1 lines 43 to 49).

It then goes on to state that:

“With the devices of the known type, it is not possible to exceed this order of depth. Wheels with a larger diameter are difficult to make.” (page 1 lines 51 to 54)

The specification then states that there is an additional problem with overboarding accessories. It acknowledges that there are known ways of overcoming this problem, but suggests that the equipment for doing it becomes excessively bulky when trying to overboard large pipes in deep water. Once again it says (page 2 lines 16 to 20) that it is not possible to exceed a depth of water of the order of 500 metres when the internal diameter of the tubular flexible conduit to be laid reaches approximately 30 cm (this would be a pipe with an outside diameter of about 18 inches). The specification then sets out the purpose of the invention as follows:

“The object of the present invention is the laying of flexible conduits at depths that are substantially greater than those which are feasible by using the known means, as, for example, a depth of the order of 1000 to 2000 metres. The device in accordance with the present invention must be capable of withstanding considerable tractive forces which may reach and even exceed 250 tons in the case of a conduit with a diameter substantially equal to 30 cm for a depth of 1000 metres.” (page 2 lines 37 to 44)

Thus one of the objectives of the patent is to enable large flexible pipe to be laid in water deeper than about 500 metres. However, as will be seen when considering the claims, there is nothing to limit them to a laying process or equipment at any particular depth. For that reason, the fact that the two pleaded acts of pipelaying by the Toisa Perseus took place in water significantly shallower than 500 metres is irrelevant to the issue of infringement. Furthermore, there is no lower limit on the tractive forces which apparatus falling within the claims must be capable of handling.

A major part of Coflexip’s invention is the idea of positioning the tensioning means vertically so that the flexible pipe leads directly from the tensioning means down into the water. The way this works is illustrated as follows in the patent:

This depicts a laying vessel (1), which has a basket (2) in which the flexible pipe or conduit (3) is stored. The pipe is led from the basket up over a wheel or, as illustrated, a chute (4) which is located at the top of a derrick (5) towards the middle of the ship. The chute acts as a guide. The conduit hangs down from it inside the derrick in a substantially vertical direction. Inside the derrick the conduit passes down between tensioning devices (6) of the caterpillar variety (11). These grip the conduit and support it. Because the tensioning devices are disposed vertically, they make it possible to lower the conduit in a substantially vertical direction through a hole in the middle of the ship, generally referred to as a moonpool (8). Just before the moonpool there is a working table (7) which can be used if work has to be performed on the pipe before it enters the water, for example if it is necessary to add an additional accessory to it.

As shown in this drawing, before it enters the mouth of the first tensioner at the top of the derrick, the flexible conduit is subject to very little tension. This is depicted by the sag (3) between the basket (2) and the chute (4). Because at this point the conduit is not under tension it sags under its own weight. The drawing depicts two conduits in this location. The specification explains that this is to demonstrate the sag which will exist in a small conduit (the top one depicted in broken lines in the Figure) and that which will exist where a larger one is being laid (the bottom one in solid lines). The specification explains the advantage of using vertical tensioners:

“In these conditions, the tension ahead of the tensioning means being exerted on the portion of the flexible conduit up the line and coming from the storage means is very low and preferably virtually negligible. It has been found that in these conditions, it becomes possible to subject a portion of the conduit situated in the portion ahead, between the storage means and the tensioning means, to relatively extensive bending, this being explained by the fact that in this case, there is no combination of the bending with an axial tractive load” (page 2 line 52 to page 3 line 7)

The patent also describes how the parts of the conduit carrying accessories or end fittings are passed through the tensioners. Because accessories and end fittings have a diameter significantly larger than that of the conduit, the tensioners are opened up to allow them to pass through. While this is happening the open tensioners can apply no tension to the conduit. Consequently the accessory is attached to a winch which takes the strain while tensioners are open. Once the accessory has passed through, the tensioners are closed and the accessory can be disconnected from the winch. How this is done in relation to an end fitting is illustrated in the patent as follows:

Drawing (a) shows the conduit (3) being supplied over the chute (4) to the top of the derrick (5). It passes down between, and is gripped by, the caterpillar tensioners (6, 11), passes through the open working table (7) and down through the moonpool (8). Above the derrick is a winch (19), which is also just visible at the top of the derrick in Figure 11 above, from which hangs a cable (20). What happens when an end fitting has to be lowered through the derrick and attached to the end fitting at the beginning of the next length of conduit is illustrated in drawings (b) to (h). The cable is attached to the fitting and the winch takes up the strain. The caterpillar tensioners are then opened and the winch allows the end fitting to be lowered to the level of the working table which then shuts to prevent the end fitting from moving about laterally. The end fitting is clamped in this position and the cable (20) is released and wound back up by the winch. The end fitting at the beginning of the next length of conduit is then passed down through the derrick. The two end fittings (21a and 21b) are joined together. The caterpillar tensioners are closed and can resume taking the strain. The working table is opened and the vessel resumes laying. The same procedure, but without the need to join two end fittings, is used if the accessory is pre-fitted to the conduit.

The specification contains no details of the design and operation of the caterpillar tensioners. This is because such devices were well known at the priority date of the patent. They have opposing tracks which press on the outside of the conduit. Operation of the motors which drive the tracks causes the conduit to be moved forward or backwards. Normally they have reversible motors. Thus there are mechanisms which can drive the opposing tracks towards each other so as to grasp the conduit to stop it slipping through the tensioner and or drive them apart so as to release the conduit. There are also motors designed to apply axial tension to the gripped conduit. What is happening can be understood by reference to someone lifting a rope at the end of which is a heavy weight. The muscles in the hand enable the hand to grasp the rope and the muscles in the arm apply the force to do the lifting.

As the above description indicates, the tracks can be moved towards or away from each other. This is also necessary to allow the tensioner to operate with different diameter conduits. As Mr Nash agreed, opening a tensioner was a conventional method of allowing accessories to pass (Transcript Day 2 page 234). The concept of opening and closing the tensioner to allow accessories to pass adds nothing inventive to the patented system. I do not understand Mr Miller to suggest otherwise.

There is one other passage which is of significance to the issues I have to decide. At the end of the specification is the following passage:

“One of the original features of the invention lies in the absence of any means for guiding the tubular flexible conduit after it has left the main tensioning means. However, it should be noted that the main tensioning means 6 can themselves ensure guidance for the tubular flexible conduit and a deviation relative to the vertical. For example, the multi-caterpillar tensioning means generally permit a deviation of more or less 10 to 15^o relative to the vertical.” (page 10)

The second half of this passage contemplates that the tensioner is designed in such a way that it can operate to “guide” the conduit when the top of its catenary is up to 15^o from the vertical.

There are nine claims. Claims 1 and 2 are process claims. Claims 3 to 9 are apparatus claims. Although many of these are said to be in issue, Claims 1 and 3 are the most important and it is convenient to consider them first. Claim 1 is as follows:

“A process for laying from a floating support (1), a flexible conduit (3) comprising a rigid accessory (21, 21a, 21b) mounted on the said flexible conduit (3) and having an outer dimension larger than the outer diameter of the latter, wherein one unrolls a flexible conduit (3) gripped at its outer surface by linear winch-type tensioning means (6) with a substantially vertical axis, and wherein

(a)

to cause the section of the flexible conduit (3) whereon the accessory is mounted to pass through the said tensioning means (6), the said rigid accessory (21) is connected to auxiliary tensioning means (19, 20), so as to take up the pull exerted by the flexible conduit (3) by the said auxiliary tensioning means (19, 20), while the flexible conduit (3) is lowered through the free space between the said laterally moved-apart tensioning means (6),

(b)

the outer surface of the flexible conduit (3) is gripped ahead of the rigid accessory by the said tensioning means (6) after they have been brought together, so that the pull exerted by the flexible conduit (3) is again taken over by the said tensioning means (6);

(c)

after the auxiliary tensioning means (19) have been released, the unrolling of the flexible conduit (3) is resumed by the said tensioning means (6), the latter comprising the last means for guiding the conduit at the level of the floating support.”

Although this covers a process designed to achieve the benefits described by the specification, it should be noticed that there is no limitation to operating in deep water or using small overboarding equipment nor is there any limit to the size, weight or construction of the flexible conduit. The claim covers a method of laying flexible conduit, whether or not anything else, for example a rigid pipe, is being laid at the same time. Furthermore, the claim does not refer to the design of the equipment upstream of the tensioner. As can be seen in Figure 11 above, the drawings in the patent illustrate the use of a chute (4) over which the conduit and accessories pass before taking up a vertical orientation. This is described in the specification. However there is no requirement in the claim that that there should be any such device or, if there is a device, that it should be a chute. Thus the claim would cover the process carried out on a vessel fitted with an overboarding wheel or sheave (see Figure 5 above) feeding the conduit and accessories into the top of the vertical tensioner. Furthermore, there is no requirement that there be a moonpool. The pipe could be paid out over the stern of the vessel. Thus the purely structural features of the claim would be met by a device of the following general arrangement:

I have already discussed the meaning of the expression “flexible conduit” (see paragraph 39 above). In the previous action, the meaning of the words “the pull exerted by the flexible conduit is again taken over by the said tensioning means” was considered. The Court of Appeal held as follows:

“21.

The pull referred to in the claim is that exerted by the flexible conduit seaside of the main tensioners. Thus the words “the pull” refer to all the tension caused by the part of the pipe seaside of the main tensioners. The sentence bridging pages 2 and 3 of the specification, relied on by Stolt, does not suggest a contrary meaning. All that is said is that, if the seaside pull is taken by the main tensioners, the tension in the pipe after the main tensioners “is very low and preferably virtually negligible”. There will inevitably be some tension due for example to the weight of the pipe and the friction, but that does not detract from the teaching of the specification that the tension from the pipe seaside of the tensioners is to be taken by the main tensioners or, when an accessory is passed, by the auxiliary tensioner.

22.

Of course, the words “the pull” have to be construed in context. As Lord Diplock pointed out in Catnic Components Limited v Hill & Smith Limited [1982] RPC 183 a word such as “vertical” had to be construed in context; in that case in the context of manufacture of lintels for the building trade. Thus the requirement that a metal beam should be vertical did not mean that it should be exactly 90^o to the horizontal. A similar approach should be adopted in the present case so that the requirements as to “the pull” should be construed in the context of pipelaying from a ship.”

It is apparent that some tension can exist on the vessel side of the tensioner. This is said to be “very low”. In the Stolt action it was not necessary to consider just how low this tension had to be. To apply the purposive construction suggested in Catnic, it is necessary to have in mind the reason why the inventor requires all or most of the pull to be taken by the tensioner. As pointed out above, when laying flexible pipe in deep water there is an increased problem with high tension on the pipe causing crushing on the overboarding chute or wheel and also with that tension making it more difficult to overboard accessories. A major purpose of the invention is to avoid these problems. Taking all or a large part of the pull on the vertical tensioner has the effect of reducing the tension upstream, thereby solving these problems. Therefore these benefits of the invention are secured if the tension upstream of the tensioner is kept low enough that overboarding of accessories and crushing are no longer a problem. Subject to ensuring that those requirements are met, a small amount of tension upstream of the tensioner is permissible. For example, spooled pipe tends to unravel itself from the reel unless it is maintained under some tension. A process which would otherwise fall within the claim does not fall outside it by reason of the fact that sufficient tension to avoid unspooling is maintained between the tensioner and the reel of conduit as long as that tension is well short of the level to cause crushing and overboarding problems.

The only other part of Claim 1 which need to be considered are the words “last means for guiding” in the phrase “the said tensioning means (6) comprising the last means for guiding the flexible conduit (3) on board the floating support”. Similar words are to be found in claim 3, the first product claim. It is convenient to consider them in relation to that claim.

Claim 3 is in the following terms:

“A device for operating the process according to any one of the claims 1 and 2, comprising;

- linear winch-type tensioning means (6) with a substantially vertical axis, capable of ensuring the normal lowering of the flexible conduit (3) by gripping the outer surface of the flexible conduit (3) and capable of being laterally moved apart, the said tensioning means (6) comprising the last means for guiding the flexible conduit (3) on board the floating support (1),

- auxiliary tensioning means (19, 20) comprising at least one elongate movable traction element (20) capable of being connected to the rigid accessory (21, 21a, 21b) mounted on the flexible conduit (3).”

This product claim consists of two parts. First, the device must be “for” operating the process of Claims 1 and 2. Second it sets out a number of structural integers which must be present in the device.

In the Stolt action I had held that the words “a device for operating” should be construed as meaning “when used for”. As the Court of Appeal pointed out, that construction would have made the apparatus claims otiose. Such claims could only be infringed when there was also infringement of the process claims so the former would have become largely redundant. The Court of Appeal held that the words should be construed in the normal way so that the claim covers devices “suitable for operating” the patented process (see Court of Appeal judgment paragraphs 23 to 27). For example if the pipelaying vessel depicted in the drawing in the patent (see Figure 11 above) had a large motor attached to the reel (2) and it was always used to provide most of the tension needed to keep the pipe suspended in the sea, it would still be within the scope of the patent because the apparatus would still be suitable for or capable of operating the process covered by claims 1 and 2. The construction adopted by the Court of Appeal not only avoids the redundancy which would have existed had my interpretation been correct, but it also creates certainty. Instead of the apparatus passing in and out of infringement depending on how it is used, it always infringes as long as it has the capacity to operate in accordance with the process claims and meets the structural requirements of the claim. The vessel depicted in Figure 11 falls within the device claim even if it has no pipe on board and is in dry dock.

This is of considerable importance in this case because Coflexip accepts that Rockwater has not infringed the process claims. Its case of infringement is that the Toisa Perseus meets all the hardware requirements of the device claims and is capable of being operated in a manner which would fall within the process claims even though Rockwater has not operated it in that manner and, as far as I am aware, has never threatened to do so. Since this is so, the alleged infringement must consist of the commercial presence of the Toisa Perseus in the Leadon and East Foinaven fields, not the pipelaying operation conducted there.

Specific functional capability is a feature of a number of other integers in this claim. Thus, the linear winch-type means (i.e. the tensioner) has to be “capable of” ensuring normal lowering of the conduit and “capable of” being moved laterally apart. Similarly the auxiliary tensioning means (i.e. the A&R winch) has to be “capable of” being connected to a rigid accessory mounted on the flexible conduit. Neither the tensioner nor the A&R winch need be performing these functions for the requirements of the claim to be met. It should be noticed that this claim does not require the presence of a flexible conduit or any rigid accessory attached to it. Once again, this means that a pipelaying vessel can fall within the claim even if it is not carrying out a pipelaying operation. I do not understand there to be any dispute between the parties on this issue.

The only dispute as to the meaning of the claim centers on the requirement that the hardware in the apparatus must include a tensioning means “comprising the last means for guiding the flexible conduit on board the floating support”.

Mr Miller advances two constructions for the words “means for guiding”. First he argues that, in the context of the patent, these words should be taken to be a reference to the well known wheels, gutters or chutes of the prior art. As he puts it in his opening skeleton argument, “that is the sort of device that must be absent after the tensioning means”. The alternative construction proceeds as follows. He accepts that the word “for” in “means for guiding” should be interpreted in the same way as the Court of Appeal interpreted it in the context of “device for operating”. In other words it means “suitable for”. He argues that the words in the claim should then be construed as meaning “capable of comprising the last means for guiding the flexible conduit” (Footnote: ¹).

I do not accept the first of these arguments. There is nothing in the specification to suggest that the word “means” is to be limited to wheels, gutters or chutes of the prior art and devices of the same sort. “Means” is one of the most frequently used words in patent specifications. It is used to encompass anything capable of carrying out a specified function. This is consistent with how it is used in Coflexip’s patent. The only passage in the specification which deals with this feature of the invention is the one set out at paragraph 53 above. There the inventor refers to “any means”. In my view, the word has its normal meaning in claim 3. It covers any structure for guiding the flexible pipe.

I also do not accept Mr Miller’s second argument. The claim does not say “capable of comprising the last means for guiding the flexible conduit” as he suggests. Having accepted that the word “for” means “capable of”, he has then re-written the claim as if it read “for comprising the last means for guiding” by notionally inserting “for” in front of “comprising”. It will be recalled the relevant section of the claim provides that the tensioner is to be “capable of” ensuring normal lowering of the flexible conduit and “capable of” being laterally moved apart but it must also “comprise” the last means for guiding the flexible conduit. In my view this imposes a mandatory requirement that the tensioner be the last means for guiding. It is not enough that it is “capable of being” the last means as Mr Miller suggests.

To determine the meaning of the expression “the last means for guiding”, it is necessary to consider separately the meaning of “means”, “for” and “guiding”. The first of these has been considered already. Mr Miller and Mr Thorley agree that the second has the same meaning as in the words “for operating the process” in claim 3 namely “suitable for”. The point of difficulty is as to the meaning of the word “guiding”.

In the Stolt action, the meaning of “means for guiding” was considered. The relevant part of my judgment reads as follows:

“By the end of the trial, only one issue on infringement remained. Claim 1 ends with a requirement that the vertically disposed tensioning devices are “the last means for guiding the conduit at the level of the floating support.” Similar wording is to be found at the end of Claim 3. This wording reflects the fact that since the patented system delivers the flexible conduit at the same angle as the catenary below the vessel, no further guides, such as stingers or overboarding wheels or chutes, are required. As the drawing in the patent (Figure 14 above) illustrates and the text describes, below the lowest tensioner there is the working table (7) and the moonpool (8). Neither the table nor the sides of the moonpool are described or referred to in the specification as guides. In normal and stable conditions, they do not even touch the conduit.

The layout of the Seaway Falcon, the Stolt Comex vessel which is alleged to infringe, is essentially the same as that illustrated in Figure 14. It includes bumpers or deflectors located below the tensioners to prevent pipe which is being laid from hitting the moonpool or moonpool doors. It was accepted by Stolt Comex that in normal use these devices do not touch the pipe. They are used to prevent the pipe from being damaged, for example by smashing against the edge of the moonpool doors during stormy weather. Mr. Willis, who had never seen the Seaway Falcon in operation, said that the pipe would touch the deflectors “every other second”. I do not accept that evidence. When the whole of his cross examination on this subject is reviewed and read in the light not only of Mr. Seeley’s evidence but also Mr. Coutarel’s evidence, it is apparent that the vessel is designed to be operated without the pipe contacting the deflectors at all. The same point comes out of the written instructions given to the operators of the Seaway Falcon. They are told to set up the vessel to ensure that there is no “clashing with the bottom edge of the moonpool”. Those instructions go on to explain that if the pipe does come into contact with the bottom edge:

“This will not cause any damage to the [pipe] if performed in a fully controlled manner as the base of the vessel moon pool has a 7.5 metre radius”

This means that because the edge of the moonpool is rounded (because of the presence of the deflectors) pulling the pipe past it will not cause damage if done carefully.

I do not accept that the deflectors in the Seaway Falcon can reasonably be called guides. They are no more guides than the bumpers at the front of a car or a crash helmet worn by a motorcyclist. The defendants infringe.”

This passage was approved in the Court of Appeal. As Mr Thorley points out, whilst this conclusion was sufficient to dispose of the question of infringement in that action, it does not help to determine what the scope of the expression “means for guiding” is. In particular it does not resolve the ambit of the word “guiding”.

I think the simplest starting point is to consider an example discussed during the trial. If a man and a woman walk down a straight road while holding hands, neither is guiding the other. If the man gently pushes the woman in a different direction, he is guiding her. He is applying a lateral force to ensure that she takes another direction. As Mr Miller argues, if a pipe is hanging vertically in water and it passes through a hole in a plate, it is not, in that condition, being guided by the edge of the plate. Furthermore, if the pipe just touches the edge of the plate but there is no lateral force applied by the one to the other, again, in that condition, the pipe is not being guided. On the other hand, if the edge of the plate applies force to the side of the pipe so as to alter its direction, in that condition it is guiding.

This issue was the subject of considerable evidence from Professor Witz. In his first report he said that he would interpret “means for guiding” to refer to any device:

“which directs or diverts the angle [of the conduit] under conditions of average contact. By “average contact” I mean that the “flexible conduit” is in contact with it for the majority of the time. This element directs the angle of the “flexible conduit” into the angle of the catenary in the water column.”

Therefore the “last means for guiding the flexible conduit” as required by claim 3 would be the last device which directs or diverts the conduit under conditions of average contact.

Professor Witz also referred to the fact that the Coflexip patent refers to the existence of a working table which may be put in place temporarily during pipelaying so that technicians can carry out work on the pipe. In relation to this he said:

“The arrangement in the Patent may include a working table which may be opened or closed when the pipe is being laid. There may be occasions when the pipe comes into contact with the closed working table as a result of vessel motions but this does not mean the working table acts as a means for guiding, as it is not designed for average contact. This is clear from the Patent which refers to the linear tensioner as the last means for guiding the pipe. If the working table was designed for average contact, it would in my view be the last means for guiding the pipe.” (Second Report paragraph 14)

In the Stolt judgment it was held that the deflectors in the Seaway Falcon could not reasonably be called guides. It was said that they were no more guides than the bumpers at the front of a car or a crash helmet worn by a motorcyclist. Using the terminology of Professor Witz, the deflectors, bumpers and helmet are not designed for average contact. They are designed for an occasional contact outside the normal operating envelope of the vessel, car or motorcyclist, as the case may be.

Professor Witz’s analysis can be explained a bit further. The purpose of the guide means is to divert the pipe into its natural catenary below the pipelaying vessel. It applies a lateral force to the pipe. If one considers paragraphs 24 and 25 above again, a guide will react the horizontal tension in the pipe. In Claim 3 it is the winch-type tensioning means which must be the last means for guiding the conduit. That is to say, it is the tensioner which has to be the last piece of apparatus capable of reacting the horizontal tension in the conduit. As Mr Thorley points out, this is consistent with the passage on page 10 of the patent referred to in paragraph 53 above; it is the tensioner which ensures guidance of the conduit and a deviation relative to the vertical. This means that the tensioner has to be constructed in a way which allows it to react the horizontal tension in the conduit.

This imposes some constructional constraints on the design of the tensioner. Not only must it be constructed in such a way that it is capable of withstanding the very considerable horizontal forces which it may encounter during normal pipelaying but also its outlet, that is to say the lowermost part of the tensioner (referred to as the “fore”), must be shaped so as to avoid damage to the conduit entering the sea. It will be recalled that one of the problems which exists when an accessory on a flexible conduit is overboarded under tension, is the risk of damage at the rigid/flexible junction (see Figure 7 above) Similar sorts of forces can act on the flexible conduit as it leaves the tensioner. This is illustrated by a drawing produced by Professor Witz in his First Report:

Construing the patent, save in relation to expressions having special technical meaning, is for the court. However, with one qualification, I think Professor Witz’s analysis of these words in the claim is correct. He refers to “average contact” as meaning contact during the majority of pipelaying. It seems to me that a “means for guiding” is a part which is capable of applying that lateral force during a significant part of the operation of the pipelaying vessel within its normal operational envelope. For example a guide which would be in almost constant contact with the conduit during pipelaying in shallow water may not be in contact with it for much of the time when pipelaying in particularly deep water where the top of the catenary is practically vertical. Thus for the majority of the time in deep water the guide does not react the horizontal tension. It would nevertheless still be a guide means or means for guiding.

As mentioned above, Coflexip only alleges infringement of the apparatus claims in the patent. For the purpose of this action, Rockwater produced a Process and Product Description together with a series of drawings. These are confidential. However the relevant design features of the Toisa Perseus are set out in the drawings of Rockwater’s own patent application, UK 2,379,259. The general arrangement of the vessel is shown below:

Flexible conduit can be led from the reel (5), over a chute (6) and down through two vertical tensioners (8). The conduit passes out of the vessel through a moonpool (9). The design of the exit equipment is more clearly shown in the following drawing.

This shows the conduit (3) passing out of the bottom of the lower tensioner (8b). It is then illustrated passing through a clamp (11) which can be ignored for present purposes. The moonpool (9) has cover doors (13). On this is mounted a flared device (12) which is referred to as a pipe guide. This is designed to deflect the conduit through a small angle into the top of the conduit catenary. If the vessel is operating in very inclement weather, the conduit may be forced against the bottom outlet of the moonpool. To lessen the risk of damage to the conduit in such conditions, the outlet of the moonpool is curved (15). Before the conduit reaches the seabed, and assuming calm seas, it may pass through the centre of the flared device (12). However once the pipe reaches the seabed and a catenary is formed, the vessel must be driven in the pipelaying direction to avoid damage at the sagbend. When that takes place the conduit can engage the surface of the flared device (12). This is illustrated in the following drawing which shows the circular pipe (3) being pressed against the two walls of the flared device (12) on the left hand side:

The point to notice about this arrangement is that the lower tensioner (8b) is not designed to react the horizontal tension in the conduit. That tension is reacted on the surface of the flared device (12) and it is that device which then transfers the horizontal tension to the rest of the vessel. Mr Thorley says that the flared device (12) is a means for guiding the conduit. For this reason there can be no infringement of any of the apparatus claims in Coflexip’s patent because in the Toisa Perseus the tensioners (8) do not comprise “the last means for guiding the flexible conduit”. It should be made clear that no one suggests that the curved surfaces on the moonpool (15) are means for guiding. They are like the bumpers on the Seaway Falcon in the Stolt action. To apply the definition set out above, they are not capable of applying lateral force to the pipe during a significant part of the operation of the pipelaying vessel during its normal operational envelope.

Mr Miller advances two types of argument to support the claim to infringement. First he says that the flared device (12) is not a “means for guiding” within claim 3. Alternatively he says that if it is a means for guiding, then, although this would mean that the tensioner is not the last such means, this feature of the claim can be ignored.

Mr Miller says that the flared device (12) is not a “means” for guiding the pipe because it is not a wheel, gutter or chute or other device of the sort intended by the patent. For reasons set out above, I reject that argument. Alternatively he says that in the context of the laying operation in the Leadon and East Foinaven fields, it is not “for guiding” the conduit after the tensioner. He says that sometimes, for example when the conduit has not reached the seabed, the latter is hanging vertically and, in calm weather, will not be touching the device. At such times it is not guiding it and is not a means for guiding.

I reject that argument too. The rear end of a stinger is a guide, as Mr Miller accepts, even though for much of its normal operation it is not in contact with the pipe. If the pipe is being laid in shallow water, the rear end of the stinger may never touch it. Nevertheless it is a means for guiding the pipe because it is capable of guiding during a significant part of the operation of the pipelaying vessel within its normal operational envelope. A means for guiding is not a means which is guiding. This is the same error as I made in the Stolt action when I held that “a device for operating the process” in claim 3 meant “a device which is operating the process”. Furthermore, as Mr Thorley points out, Mr Miller’s argument is self defeating. If one considers those instances when the pipe is hanging vertically, not only is the flared device (12) not applying any lateral force to the conduit, but neither are the vertically orientated tensioners because they also are not diverting the conduit from its vertical path. So, on this argument, they are not means for guiding either and the claim, which requires them to be the last means for guiding, is not met. In my view Mr Nash’s evidence supports Rockwater’s case:

“185.

During normal lay there will be an angle of lay which is dependent upon the submerged weight of the pipe, the depth of water and the bottom tension applied by the pull of the vessel. A lay angle of 1º-2º would be typical for the sizes of pipe and water depth of Foinaven. Vessels such as the Toisa Perseus can lay in the forward, sideways or reverse directions. The frequency of contact, and forces operating, between the pipe guide and the pipe will depend upon the direction of lay because the pipe guide is offset from the centre line of the tensioner. In the case of the Foinaven contract when laying in the forward direction, once a catenary is established, it is likely there would have been some contact between the pipe guide and the pipe but the extent to which the pipe was bent around the guide cannot be known for certain (because the angle of lay is not known for certain) although in normal circumstances it could be very small. If laying is taking place with the vessel moving in reverse then, under median conditions, the pipe would not contact the pipe guide up to an angle of lay of 1.2º (see table 6 below).” (First Report paragraph 185)

This passage was the subject of cross-examination:

“Q. … Let us get to the stage where we make contact with the seabed, which is paragraph 185 of your statement. As I understand it, you accept that in this state, with a lay angle of one to two degrees, there is likely to be some contact with the guide means, and that is [paragraph 185]. I just read it wrong. "... it is likely there would have been some contact between the pipe guide and the pipe but the extent to which the pipe was bent around the guide cannot be known for certain". A. Yes, that is right.

Q. At that stage the horizontal tension due to the catenary will be taken on that pipe guide. A. Yes. Making the assumption that the vessel was going forward and you have an angle, yes, that is correct” (Transcript Day 4 page 525)

On that evidence, in Foinaven, the flared chute (12) was not only capable of acting as a guide but, although not strictly relevant, it was guiding.

There is something of an inconsistency in Mr Miller’s final argument on infringement. Both in opening and in reply he made it clear that he was not asserting that this feature was an inessential integer in the claim. On the other hand he argues that, if the flared device (12) is a means for guiding the conduit, the Toisa Perseus still infringes. He says that the reference to the tensioner being “the last means for guiding” has to be construed in accordance with the Protocol on Article 69. He then relies on Aldous LJ’s judgment in Wheatley v Drillsafe [2001] RPC 133 as indicating that the structured approach adopted in Improver v Remington Consumer Products [1990] FSR 181 assists in applying the Protocol and that a strict literal application of the claims is outlawed.

Following a structured approach can be helpful. However it is not a straight-jacket. The court is trying to determine not what the inventor intended to be covered by his claims but the meaning of what he said he intended to cover. In doing that it can pay less regard to any wording used in the claim which appears to import a limitation which the skilled addressee would believe could not reasonably have been intended. The claims therefore have to be read in the context of the specification taking into account not only the words used but the explanation given by the patentee for how the invention is believed to work and what he states its benefits to be. In some cases where a party argues that he does not infringe because his apparatus or process omits a feature in the claim, the issue can be put simply as follows; would the notional skilled addressee who reads the whole patent understand that the particular feature in the claim has to be present. If that question is asked here, the answer is in the affirmative. Claim 3 only incorporates a few integers. This makes it less likely that any of them would be read as redundant. But, whether that is so or not, the passage on page 10 of the specification picks out for special mention the alleged originality of the feature that there are no means for guiding the conduit after the tensioner. The skilled addressee would not ignore this integer.

The same conclusion is arrived at if one follows the Improver structured approach. First, does the incorporation of the flared device (12) in the Toisa Perseus have a material effect on the way the invention works? The answer to this requires an understanding of what the invention is said to be. The passage on page 10 of the specification points out the absence of a further downstream guide means that the tensioner “ensures guidance of the tubular flexible conduit”. In other words the invention includes the alleged benefit of having a tensioner which not only applies vertical tension to the line but is capable of reacting horizontal tension for the purpose of guiding the conduit. The benefit of all the forces being able to be reacted by the tensioner is missing in the Toisa Perseus. The difference between the design of that vessel and the design in Coflexip’s patent was explained by Professor Witz:

“One of the consequences of using a linear tensioner as the last means of guiding (as in the Patent) is that the catenary top angle is directed or diverted over the fore of the tensioner from the vertical (see Fig [14] above ). In effect, the pipe is bent under tension over an arc at the fore of the linear tensioner. The associated bend radius is relatively small. It is generally undesirable to bend pipe at such relatively small radii under high tension, particularly for larger angles as this increases the risk of pipe collapse and wear and tear on the linear tensioner. An advantage of the pipe guide is that it has a radius which is larger than the radius of the fore of typical linear tensioners. This results in a reduced risk of collapse of the pipe, particularly at larger angles and alleviates the linear tensioner’s components from the mechanical demands placed by the horizontal force.” (Second Report paragraph 24)

The third Improver question is; would the reader skilled in the art have understood from the language of the claim that strict compliance with the primary meaning was an essential requirement of the invention? As noted already, Mr Miller accepts that this feature of the claim is not inessential. In those circumstances it is difficult to see how the third Improver question can be answered in Coflexip’s favour. In any event, the passage on page 10 of the specification would indicate to the skilled reader that the absence of any guide means downstream of the tension was an essential requirement of the invention.

For these reasons, the allegation of infringement of claim 3 fails. Since all the other apparatus claims are dependent on this claim, all allegations of infringement fail also.

As mentioned above, Rockwater alleges that the claims in the patent are invalid on the grounds of anticipation and obviousness. At the commencement of the trial, three pieces of prior art were relied on, namely Recalde GB, Recalde US and UK Patent No. 1,059,932 (“IFP”). The latter relates to the opening and closing of the jaws of caterpillar tensioners. As a result of the oral evidence given by Mr Nash in relation to how such tensioners were normally used, Mr Thorley decided that he no longer needed to rely on this piece of prior art.

Although Recalde US and Recalde GB are equivalents and contain much common information, there are some differences which are said to be material to the issues in this case. For that reason it is necessary to consider them separately. The parties concentrated most of their attention on Recalde US and I will do likewise. Before considering this material, it is worth mentioning a preliminary matter. There is a dispute between Coflexip and Rockwater as to what each of the prior art patents discloses. In support of their respective constructions they served evidence, Coflexip relying on Mr Nash’s interpretation and Rockwater relying on Professor Witz’s. However, in relation to those passages in the specifications where the parties are in disagreement, there are no disputed terms of art. The disputes therefore relate to the construction of passages in the documents which are non technical and use ordinary English words and grammar. I understand Mr Miller and Mr Thorley to agree that that is so. In the circumstances it appeared to me that much of the evidence was inadmissible. Mr Miller and Mr Thorley agree that the construction of the prior art is a matter for the court alone. They agree that I am not constrained by anything the witnesses say, but they also agree that the evidence allows each side a convenient avenue by which to put its case on construction before the court. Each side also criticizes the other’s expert. Mr Thorley says that Mr Nash has been steeped in the litigation concerning the patent in suit for such a long time that, no doubt subconsciously, he has interpreted the prior art in a way which is most favourable to Coflexip. It is suggested that he has done this to the extent of ignoring or misreading the clear wording of the documents. Mr Miller argues that Professor Witz read the Recalde specifications knowing the points in dispute between the parties. As a result, again subconsciously, he has resolved alleged obscurities in the text in a manner which suits Rockwater. Having seen both these experts in the witness box, there is no reason to doubt the sincerity of the views they have expressed.

Normally, it would be appropriate to reject this type of evidence out of hand. But I agree with Counsel that in this case it has served a useful purpose. Professor Witz said that it took him about two days to read Recalde US. I am not surprised. Although it is not suggested that there are any complex concepts discussed in these documents, each of them is crammed full of information. This makes for a difficult and, frankly, turgid read. Much of what is disclosed has no or only peripheral relevance to the issues in this trial. Those passages which are or may be relevant are scattered throughout the text. Digesting the disclosures is a considerable task. As Counsel say, each expert has put into writing his side’s interpretation of the document. By doing so, the task of understanding not only the documents but also the arguments between the parties has been eased. Nevertheless, ultimately construction of the prior art is not a matter for the experts. The expert evidence directed to these prior art documents is also admissible in so far as it relates to the issue of obviousness.

In the following analysis of the Recalde patents, I refer to some of their drawings. They are annexed to this judgment. To aid comprehension, in some cases I have coloured the particularly significant components and have removed the numbers and lead lines which refer to components which are of no relevance to the issues. This has not altered in any way the structure of the Recalde apparatus. It has simply made it possible to understand the drawings without the distraction of references to irrelevant material. The letter “p” at the end of a number indicating a particular part indicates that that part is on the port side of the vessel, e.g. 330p. The letter “s” similarly indicates starboard side, e.g. 330s. The letter “a” indicates aft and “f” indicates fore.

This patent describes apparatus for, and methods of laying, a number of submarine pipes and other conduits simultaneously. It refers to a variety of prior proposals for laying rigid pipes in deep water including rigid pipe which has been spooled onto a reel. It refers to prior art which discloses the use of a combined straightener-level winder to spool the rigid pipe onto the reel and for straightening it as it is unspooled. In passages dealing with the prior art it also contains a section which concentrates on the design of the Apache in which spooled rigid pipe was fed down a ramp at an acute angle to the sea (i.e. using a J-Lay configuration). It goes on to note that the Apache was not equipped to layout multiple lines, apparently because it had a single main reel (for rigid pipe) and insufficient deck space for convenient placement of auxiliary reels. It comments that there was an early suggestion during the Apache’s construction phase to mount portable reels on its desk to permit smaller lines to be bundled with the pipeline (that is to say put together with and connected to the rigid pipe) at the stern of the vessel. It notes:

“These smaller lines were not required to be passed through the pipe handling equipment with the main reel pipeline according to the suggestion …” (Column 3 lines 58 to 60)

In a section entitled “Summary of the Invention”, Recalde US points to the increasing requirement in the offshore petroleum industry for laying multiple operational lines in water more than 3,000 ft deep. The expression “operational lines” is used throughout the specification as a generic term to include rigid and non-rigid pipes and other types of conduits. For example it includes not only rigid pipes but also electrical cables and tension support cables (Column 11 lines 50 to 57). It is not in dispute that this encompasses the type of flexible conduit referred to in Coflexip’s patent. Recalde US goes on to say:

“To be commercially viable a pipelaying vessel must also be capable of laying either single or multiple operational lines in shallow waters of less than 200 ft. up to 3,000 ft. depth.” (Column 4 lines 2 to 5)

The summary of the main objective of the patent reads as follows:

“A principal feature of the present multi-reel pipelaying vessel is that an operational lines laying device is mounted adjacent to the stern of the vessel. A plurality of operational lines are unspooled from the reels mounted on the vessel and are laid out into simultaneous contact with the laying device which includes an operational lines supporting means adapted for providing moving contact with the operational lines. The preferred laying device of the present invention changes the direction of movement of the plurality of operational lines from horizontal to vertical and can be used for laying operational lines in shallow waters of under 200 ft down too much greater depths of 7,500 ft and beyond. The supporting means is adapted for gathering the operational lines into an initial juxtaposed configuration which is parallel with the direction of forward vessel motion. All of the operational lines are moved at the same linear velocity due to the contact thereof with the supporting means of the operational lines laying device.

The preferred operational lines laying device also includes straightening and tensioning devices which are adapted to straighten and provide tension for the operational lines while maintaining the same in a juxtaposed array which is aligned with the direction of forward vessel motion. The straightening means is adapted for imparting a reverse bending force to the rigid walled pipeline(s) which are among the operational lines being laid out.” (Col 4 lines 16 to 44)

The general layout of the equipment can be best understood by reference to Annex 1 to this judgment which contains drawings derived from Figures 1 and 4 of the patent. The top drawing is a schematic view of the multi-reel vessel and the bottom drawing is a larger scale version of the stern half of that vessel. Towards the centre of the vessel there is a main reel (20 – red) which is adapted to provide storage for a series of wraps of rigid walled pipeline. The drawings show a single pipeline (22) being unspooled from that onto the pipe takeoff assembly (shown generally at 24 – orange and dark blue). This pipe takeoff assembly includes a pipe take-off drum (26 – dark blue) which is located over the stern of the vessel. The pipe take-off assembly also includes a pipe take-off structure (30 – orange). The pipe take-off structure incorporates a straightening device (32) and a tensioning device (34). The designs of the latter devices are not clearly depicted in these drawings. They will be discussed below by reference to other Annexes. Next to the main reel (20) there is a first auxiliary reel (36 – green). Next to that is a second auxiliary reel (40 – yellow). Operational lines (44 and 46) are taken off these reels and over the pipe take-off drum (26) side by side with the pipe (22) from the main reel (20). As Recalde US explains:

“All of these operational lines are gathered together into an initial horizontal juxtaposed configuration at the top of the take-off drum 26 and are maintained in continuous contact with the periphery thereof as the direction of movement of these operational lines is changed from horizontal to vertical at the stern end of the pipe take-off drum 26.

The contact between the operational lines 22, 44 and 46 with the periphery of the pipe take-off drum 26 results in the linear velocity of layout of all of these lines being equal. Upon changing of the direction of movement from horizontal to any angular pipe take-off position the operational lines are then passed through the straightening means 32 and the tensioning device 34 so that the array of these lines passes downwardly over the stern 28 at nearly a 90^o angle as shown in FIG 1 for deep water layout.” (Col. 10 lines 13 to 30)

It should be noted that in this passage and throughout the specification the group of operational lines being laid simultaneously is referred to as an “array”. The pipe take-off drum rotates round an axle (116). The pipe take-off structure (30) is constructed integrally with a support frame structure (114). This is mounted on the same axle (116). This is connected to motors which enable the pipe take-off structure (30) to rotate so as to accommodate different angles at which the array (35) enters the water. The vessel also includes winches. One of these is shown beside the main reel (20) on the upper drawing in Annex 1. The arrangement consists of a storage reel (50). On it are stored two cables (52, 54). These pass under the main reel (20) by way of a twin groove sheave (56) and rollers (58, 60) to a twin drum traction winch (62).

The internal structure of one version of the pipe take-off structure (30) is shown in Annex 2. This is based on Figure 8 of Recalde US. This illustrates the passage of an array of pipes and conduits (35 - light blue) into the top of the pipe take-off assembly past the straightening device (32 - yellow), between the co-operating caterpillars of the tensioning device (34 - green and pink) and through an opening, called the operational line array opening, (368) down towards the sea.

Considering first the straightening device (32), this consists of a first track assembly (102) made up of a main carriage (308) at either end of which is a sprocket gear wheel (318, 320) which carries an idler roller assembly (328). The idler rollers and guide rollers (310p) force a flexible chain into engagement with the surface of the operational line array (35). Behind the main carriage (308) is an internal guide frame (280p) in the middle of which is a hydraulic positioning ram (306p - mauve) with a hydraulic ram piston (316). The function of this ram is to force the face of the straightening device towards or away from the operational line array (35) as designated by the arrows S1. The straightening device (32) is also articulated which allows it to pivot as indicated by the arrows S2.

The operational line array (35) then passes into the tensioning device (34). On one side is the second track assembly (104 - green). This consists of a carriage at either end of which are motors (348, 346) which engage and drive a flexible chain rack (349). Behind this track assembly is an internal guide frame (287p) in the middle of which is a hydraulic ram (330p) which acts through a piston to force the track assembly towards or away from the operational line array (35). On the opposite side of the operational line array (35) is the third track assembly (106 – pink). This has a structure which is the same as the second track assembly (104) save that its hydraulic ram (338p) is somewhat shorter. On the seaboard side of the tensioning device (34), the operational line array (35) passes through the opening (368). In the region of this opening, there is a retractable clamp (367 - red). This can be operated to lock the operational line array (35) in position, for example if work has to be done on the array. During normal pipelaying it is retracted. In this region there is also a pipe aligner clamp set (370 - brown) which can be used to manipulate the array when it is locked by the clamp (367). There are also pivotally adjustable floor panels (358, 356) on which workers can stand for the purpose of manipulating the array. Finally there is an A&R (i.e. abandonment and recovery) winch (362 – orange).

In Annex 3, the top drawing is of a cross section of the straightener (32) depicted in Annex 2. Its function is to plastically deform the bent pipe which is being carried to it on the outer rim of the pipe take-off drum (26 – blue) which can be seen in Annexes 1 and 2. To do this, the pipe has to be clamped between a pipe-receiving surface (brown) on the pipe take-off drum and support pads (green) carried on flexible chain racks (322). In the illustrated embodiment there are openings for 6 pipes or conduits between the drum and the support pads. They are grouped into three pairs. The flexible chain rack carrying each pair of pads can be moved towards and away from the take-off drum (26) by means of manually-operated jactuator adjusters (329 – red). The composite component carrying the flexible chain racks, jactuator adjustors and pipe-supporting pads can be moved forwards or backwards by actuation of hydraulic positioning rams (306p and 306s – mauve). The latter forward and backward movement is facilitated by guide rollers (310p and 310s) which run in lateral tracks. The positioning rams (306p and 306s) are mounted on ears (314p and 314s) which allow the mechanism to pivot. It will be noticed that the straightener contains no drive means. The rigid pipe which is to be bent has to be drawn round the take-off drum and through the straightener. Depending on the dimensions and construction of the pipe, this takes considerable force.

The tensioner device is shown in cross-section in the bottom drawing. It consists of two co-operating track assemblies (104 and 106) which can be forced towards or away from each other by the hydraulic rams (330p, 330s, 388p and 338s – mauve). This movement is facilitated by guide rollers (336p, 336s, 342p, 342s) carried on the outer surfaces of the track assemblies which co-operate with tracks. The second track assembly (104) includes hydraulic motors on each side (348p, 348s). With the assistance of an idler roller assembly (352), these motors drive flexible chain tracks (349). On either side of the track assembly (104) there are adjustment cylinders (353p and 353s) which permit variation in the applied force for the idler roller assemblies. The second and third track assemblies (104, 106) carry opposing operational lines supporting pads pairs (354a – red and 354f – green). The way in which the tensioner is intended to work, the subject of considerable dispute during the trial, will be considered below. It should be noticed that in this embodiment, there is a single supporting pad on each track assembly (104, 106). That is to say, the 6 pipes or conduits are caught between two opposing pads, each pad having 6 indentations in its surface to accommodate the pipes or conduits.

Recalde US also discloses an alternative design for the straightener and tensioners of Annex 2. These are described by reference to three drawings, Figures 17 to 19 in the patent. Annex 4 contains drawings based on those Figures. The description of these drawings in the specification commences with the following:

“A preferred configuration of a straightener/tensioner track assembly which can be used for assembly 102 and, with some modification to provide for adequate hydraulic power, also for assemblies 104 and 106 is shown in FIGS. 17 – 19.” (Col 19 lines 29 to 33)

Thus, subject only to the need for adequate hydraulics, the design depicted can be used either for the straightener or the tensioners. The top drawing is a side elevation of the preferred device, the middle drawing is a transverse cross-section and the bottom drawing is a vertical section taken along the lines 19 – 19 of the top drawing. The device carries three endless tracks (322 – light green) on which are mounted support pads (624, 626, 628, 630, 632 and 634 – yellow). Each such pad is designed to interface with up to two pipes or conduits. This can be seen particularly clearly in the bottom drawing. Each endless track is mounted on a pair of sprocket gears (540/544, 539/543 and 538/542 – lilac). These sprocket gears are attached to axles (534, 536 – dark blue) which are carried in axle housings (546, 556, 558 and 560 – turquoise) which can slide inside C-shaped brackets (554 and 556 – dark green). The axles (534, 536) can be pushed away from each other by means of hydraulic rams (550, 552, 562, 564 – light mauve) which act on the axle housings (546, 556). Sprocket chain tension gears (324, 326 – grey) bear against the back of the tracks so as to maintain the tracks in tension. Each of the three tracks is mounted over a row of nine idler rollers (602, 604, 606 – dark pink). Each idler roller is carried on a mounting frame (612 – beige) which is connected to one end of an adjusting screw (614 – red). The other end of each adjusting screw passes through a jactuator (616 – pink) which can be operated manually to move the mounting frame with associated idler roller forward or backwards. As depicted in the bottom drawing, the jactuators are adjusted so as to accommodate the different outside diameters of the pipes or conduits in the operational lines array. Finally, the whole device can be moved forwards or backwards (as depicted in Annex 2) by the operation of hydraulic rams which are connected to it (317p, 317s – mauve).

In the embodiment of Recalde US depicted in Annexes 1, 2, 3 and 4, the tensioner and straightener are shown as discrete pieces of equipment. However the patent also discloses another embodiment in which one device carries out both the tensioning and straightening functions. This is said to be the preferred form of apparatus with respect to operating efficiency and capital cost minimization. This is illustrated in Annexes 5 and 6 which are derived from Figures 36, 37 and 38 of the patent. Annex 5 has two drawings on it. The top one is a side elevation. The bottom one is a cross-section taken on the lines 37 – 37 of the top drawing. Annex 6 contains a more detailed view of the interacting caterpillar mechanisms of the straightener/tensioner. It will be appreciated that many of the features of this device are similar to those illustrated in Annex 4.

Considering the top drawing in Annex 5 first, this shows the operational line array (light blue) entering from top right and exiting bottom left. It enters the straightener/tensioner multiple track assemblies (1020 and 1022 – respectively green and pink). These are equipped with multiline tracks (1042 and 1044). The straightener/tensioner multiple track assemblies can be pushed towards or away from each other by operation of the hydraulic rams (1034, 1036 – mauve). The straightener/tensioner multiple track assemblies contain a number of hydraulic motors (1046s, 1048s, 1050s and 1052s [there will also be four motors on the port side]) to drive the multiline tracks. The locations marked A, B and C identify zones where force is applied to straighten any rigid pipe entering the straightener/tensioner device which needs straightening.

The cross-section drawing in the bottom half of Annex 5 shows the same parts identified by the same numbers. However this drawing also shows the pipe/conduit retaining pads (respectively yellow and green) on the left and right chain track sets (1042 and 1044). The individual pipes and conduits are located between opposing pads. The drawing also depicts the adjusting screws and jactuators (red) which were identified (614, 616) in the top drawing in Annex 3 and Annex 4 as part of the tensioner device.

A more detailed view of the combined straightener/tensioner is to be seen in Annex 6 which is a view from the side. This depicts the operational lines array (35 – light blue) entering between the opposing chain track sets (1042, 1044 – respectively yellow and green). Once again, the opposing hydraulic rams (1034s, 1036s – mauve) are shown. The hydraulic motors (1046s, 1048s, 1050s, 1052s – grey) are shown. The individual adjusting screws (red) which can be adjusted to press down on the chain track sets (1042, 1044 – respectively yellow and green) are inside the straightener/tensioners. However there are sight openings (1104, 1106) in the walls of the apparatus through which it is possible to observe the rotation of the jactuator screws, which are used to adjust the setting of each adjusting screw. The only other part of this adjustment mechanism which can be seen from the side is the bottom part of the idler rollers at the point where they face each other on either side of the pipeline array (35). On the left-hand device in the drawing, the outline of the jactuator/adjusting screw mechanisms inside are represented in ghost outline (pink). The drawing also illustrates a series of jactuator adjustment openings (1100, 1102 – brown). Recalde US explains that:

“These jactuators adjustments permit the turning of internally mounted screws in order to position the multiple roller guides independently.” (Col. 30 lines 1 to 4)

Annex 6 shows that there are nine roller guides (red) on either side of the operational line array (35). The bottom drawing also shows that there are three endless chain track sets on each caterpillar straightening/tensioning device, that is to say, the tracks for the top two small conduits, the track for the middle two large conduits and the track for the bottom two small conduits. Recalde US explains:

“As shown in FIG 38 (Footnote: ²) nine roller guide sets are provided for each of three endless sprocket chains which are employed to support the pipe pads for three pairs of lines which constitute the pipeline array 35 as shown in FIGS 37 (Footnote: ³) and 38.” (Col. 30 lines 4 to 8).

It will be appreciated that if one of the auxiliary reels (36, 40) in Annex 1 is loaded with a flexible conduit, as Recalde US teaches, that conduit passes over the take-off drum 26, which is a form of sheave, and down to the water through the tensioner (34). This arrangement is the same as that depicted schematically in Figure 13 above. In other words the major items of hardware required by claims 1 and 3 of the patent in suit are present. Major issues between the parties concern the meaning of Recalde US’s description of the operation of the apparatus and how the apparatus can be operated. Coflexip’s case is that virtually all the axial tension is taken on the reels (20, 36 and 40). The teaching of the specification is that flexible conduit passing through the straightening/tensioning section with the rest of the operational lines array (35) would only be in touching contact (if in contact at all) with the tensioner, so that no part of the axial load in the conduit would be taken by the tensioner and that the apparatus is of a design which would not allow the tensioners to take up the axial load of the flexible conduit. Rockwater argues that the tensioner (34) takes all or most of the axial tension in the flexible conduit and that, even if that is not described, the apparatus is capable of being operated in that way.

As mentioned above, the description in Recalde US is detailed. The passages which throw light on the issues to be decided in this action are distributed throughout the specification. It is necessary to gather them together to understand how pipelaying is achieved and, in particular, how the tensioner at the rear of the vessel operates. It must also be borne in mind that the specification describes more than one method of operation.

Coflexip argues that this patent is primarily concerned with laying rigid spooled pipe. For that reason it needs to have a straightener. The tensioner co-operates with the straightener, or is incorporated within it, to assist in the removal of the curve from the rigid pipe. The result is that although flexible conduits take the same route as the rigid pipe(s) in the operational lines array (35) over the take-off drum (26) and down through the straightener (32) and tensioner (34) into the water, there is no need for them to be straightened, because they are not plastically deformed in the first place. Therefore their passage through the straightener is essentially passive. They are in contact with the straightener but have no force applied to them by it. Since this is so, there is also no need for the tensioners to apply tension to them for the purpose of reacting the tension created by the line hanging below the vessel. The flexible conduits may, but need not, contact the pads on the tensioner, but that is all. Mr Nash summarised the point:

“There is no technical reason why Recalde US would wish to have the tensioners grasp the operational lines which are not rigid pipes.” (First report paragraph 153).

It is true that a major objective of Recalde US is to provide for the straightening and laying of rigid pipe and it is for that reason that the straightening facility is present. In all the described embodiments of the invention, the operational line array must include at least one rigid pipe (see, for example, Col 6 lines 2 to 6, and lines 58 to 62). The existence and use of the straightener and tensioner are central to the patent. Furthermore all of the reels described and illustrated are powered. The reason for this is explained:

“The storage reels are fitted with hydraulic motors for imparting motive power to the reel flanges or rims in order to provide for spooling up of the lines. The hydraulic motors are also fitted with hydraulic braking systems for controlling tension of the lines during unspooling and to control the rate of line(s) layout.” (Col. 6 lines 17 to 23)

Thus, these motors can be used to pull pipe onto each of the reels while they are being loaded (“spooling up”). Furthermore there are provisions to apply a brake to each of the reels during pipelaying (“unspooling”). This means that the braking system can be used to oppose or react the tension created by the pipe suspended under the vessel. Furthermore there is an indication in the patent of how powerful the motors or brakes on the reels are expected to be:

“In deep waters beyond 3,000 ft the weight of the pipeline(s) is sufficient to elastically straightening the rigid walled pipe. In this embodiment it is possible to use the hydraulic braking systems on the operational lines reel motors to provide tensioning of the lines, thus permitting pipe layout in the absence of separate straightening and/or tensioning devices.” (Col 5 lines 3 to 9)

In this embodiment, the totality of the tension generated by the suspended pipe in the deepest water can be reacted by the braking system on the reels. The specification therefore teaches the design of equipment in which all or substantially all of the tension is capable of being taken by the reels upstream of the take-off drum. When operated in that mode, the equipment is being used in the manner indicated in Figures 4 and 5 above. Although this embodiment illustrates the intended versatility of the Recalde design, it is an embodiment which eschews use of the straightener and tensioner and is therefore somewhat peripheral to the main teaching of the patent. On the other hand it should be noticed that this passage states that in deep water the rigid pipe can be straightened without using the straightening and/or tensioning devices. It could be said that this means that in some embodiments, the straightening device is still used but the tensioning device is not and that, in such a case, it is the tension generated by the weight of suspended pipe which renders the tensioner redundant. In other words, the tension which had been applied by the tensioner is now applied by the weight of suspended pipe. If this is what the passage means, the tensioner (prior to its redundancy) was pulling in a seawards direction, not towards the vessel and, for that reason, again all the tension (save for that necessary to force the rigid pipe through the straightener) is taken on the reels. However, even if so, this does not affect the general teaching of Recalde US which is considered below. The parties agree that caterpillar tensioners are normally reversible. This means that the tensioners shown in this patent can either pull towards or away from the seabed. The last above cited passage may indicate that sometimes the tensioner pulls towards the sea. In any event, it is likely that when laying commences, the operational lines array will have to be fed through the straightener and tensioner. Since, at the very beginning of this operation, there will be no weight of suspended pipe to help pull the array through and to overcome the resistance generated by the straightener, it is likely that the tensioner will be used to pull in a seaward direction.

Notwithstanding the above, the important questions are (i) what does Recalde US teach about the way in which its apparatus operates during normal pipelaying and, (ii) of particular significance to Claim 3 of Coflexip’s patent, how can the apparatus be operated?

In my view Recalde US teaches that all the lines in the array, including the flexible conduits, are grasped by the tensioner and have axial tension in an upward direction applied to them by it. Further it teaches that most, if not all, of the tension generated by the suspended pipe can be reacted or taken by the tensioner. If that is so, little axial tension exists in the operational lines array passing over the take-off drum. Set out below are the passages in the specification which appear to support that conclusion. It should be borne in mind that the expressions “array” and “operational lines array” are used throughout the specification to designate all the pipes and conduits which are being laid simultaneously.

A convenient starting point is the summary of the invention which commences on column 3 of Recalde US. It is worth repeating one passage which has already been set out earlier in this judgment:

“The preferred operational lines laying device also includes straightening and tensioning devices which are adapted to straighten and provide tension for the operational lines while maintaining the same in a juxtaposed array which is aligned with the direction of forward vessel motion. The straightening means is adapted for imparting a reverse bending force to the rigid walled pipeline(s) which are among the operational lines being laid out.” (Col 4 lines 36 to 44)

There is no dispute that the straightener only operates on the rigid pipes since they are the only ones which need to be bent back to the straight configuration. The above passage points out that the straightener works on the rigid pipes. However it also says that the tensioning device is adapted to provide tension to the operational lines array. The natural meaning of this is that the tensioner applies tension to each of the lines in the array, including any flexible ones. Coflexip’s argument, as expressed by Mr Nash, is that the flexible lines or conduits are neither tensioned nor straightened. If that were so, this paragraph would have been directed to rigid pipes alone rather than all the lines in the array.

The next significant passage is as follows:

“The tensioning device 34 is formed by a second track assembly 104 and a third track assembly 106 which act on opposite sides of the operational line array 35, respectively, in order to provide tension for supporting the pipeline weight which is suspended from the pipe take-off assembly 24.” (11-1, 11-8)

There is no dispute that this passage indicates that the tensioner is exerting a force in the upwards direction away from the seabed. Mr Thorley argues that the reference to providing “tension for supporting the pipeline weight” indicates that substantially all the tension is reacted by the tensioner with the consequence that upstream of the tensioner there is little or no tension. I do not accept that argument. Taken by itself, the passage could just as well mean that the tensioner provides tension to assist in the support of the weight of the suspended pipes. On the other hand Mr Miller argues that this passage only refers to supporting “the pipeline” and this must be a reference to the rigid pipeline. I do not accept that argument either. Both on its own and in the context of other passages to which reference will be made below, this suggests that all the lines in the array have tension supplied to them. That means that they are all grasped by the tensioner and pulled in an upward direction.

Coflexip relies upon the following passage which is dealing with the drawings reproduced in Annex 1 to this judgment:

“Operating examples are that the line 22 can be a 6” o.d. Rigid walled pipeline; the line 44 can be of 4” o.d. Rigid walled pipeline; and the line 46 can be either a single or dual set of electrical lines. All of these lines are passed over the pipe take-off drum 24 and are also passed through the straightening device 32 and tensioning device 34 even though the electrical line 46 does not require straightening.” (11-57, 11-64)

It is said that the reference to the lines passing “through” the straightening and tensioning devices suggests that they are not in load-bearing contact, like a thread passing through the eye of a needle. I do not accept that. The paragraph refers to all the lines passing through the devices. That includes the rigid pipes which everyone agrees are being pulled by the tensioners. To read the reference to the lines passing “through” the tensioning device as requiring them to be in non load-bearing contact would be counter to the express teaching in relation to the rigid pipes. There is nothing in the passage to suggest that in this respect rigid pipes are being treated any differently to flexible conduits.

Later, in relation to the drawings reproduced in Annex 2, the specification states:

“Another advantage of the straightener device 32 and tensioning device 34 in the locations shown in Figures 3 and 8 with respect to drum 26 is that only a few mechanical devices are required for providing both pipe straightening and tensioning. This configuration permits the contacting of the operational lines array 35 by the straightening device 32 prior to engagement of the array by the tensioning device 34. This permits the proper functioning of the tensioner device 34 which must be operated in order to have the same force exerted on both sides of the operating lines array 35.” (Col 22 lines 1 to 11)

This teaches that “the array”, meaning all the components within the array including the flexible conduits, is “engaged” by the tensioner. The latter applies force to both sides of the “array”, again meaning all the components within it, including the flexible conduits. This means that the tensioner applies force to all the conduits within the array including the flexible conduits. The suggestion made on behalf of Coflexip that the “engagement” referred to here indicates no more than that the flexible conduits can touch the tensioner but not be pulled by it, is not the natural meaning of the passage.

The specification describes how the auxiliary reels 36 and 40 are fitted with hydraulic motors “which are used for spooling of operation lines on to the reels” (Col 24 lines 30 to 34). As made clear throughout the specification, the operational lines referred to include flexible conduits. The specification continues:

“During the unspooling operation the hydraulic systems providing power to the two hydraulic motors 806 and 808 can be operated in order to provide breaking force for the real 40 in order to provide additional tension for the operational lines which are being paid out over the drum 26 for layout.” (Col 24 lines 41 to 47)

This is an important passage. This teaches that the hydraulic motors “can” be operated to provide “additional” tension during unspooling, i.e. pipelaying. This must be tension to react or counter the downward pull exerted by the suspended pipe. The passage indicates that the tension generated by the hydraulic motors is additional to tension generated elsewhere. On the other hand the operator may choose not to operate the hydraulic motors in which case all the tension will be reacted elsewhere. The natural meaning of this is that all or most of the tension generated by the suspended array can be reacted in the tensioner and in the straightener (as far as the latter is concerned, some force is necessary to pull the rigid pipe round the straightening surface). Furthermore, since the flexible conduits do not need to be straightened, they pass through the straightener without substantial load being applied to them by that device. In respect of such conduits, this passage is teaching that the tensioner can react substantially all of the tension. Upstream of the tensioner there will be substantially no tension on the flexible conduits. For the avoidance of doubt, there is no reason to read the words “operational lines” in this passage atypically as referring to the rigid lines only.

The same teaching that the tensioner applies tension to all the lines in the arrays rather than to the rigid line(s) alone is consistent with the following passage:

“For these reasons, the tensioner track assemblies 104 and 106 as shown in Figure 8 are utilised solely for providing tension to the operational lines array. These are not usable for straightening since they do not permit curvature adjustment of the type required for use in laying devices described herein.” (Col 27 line 68 to 28 line 5)

If Coflexip and Mr Nash were correct, this passage would have said that the track assemblies are utilised for providing tension to the rigid pipes. It does not do so.

In my view, Coflexip’s construction also does not read onto the description of the combined straightener/tensioner (see Annexes 5 and 6).

“Figures 37 and 38 illustrate in greater detail the straightening/tensioning device 1024. The two assemblies 1020 and 1022 which comprises device 1024 are of identical construction except that each of the assemblies has the track sets mounted thereon in a configuration to grasp the various lines in the pipeline array in order to exert tension there along.” (Col 29 lines 4 to 10)

Although, as noted above, Mr Nash expressed the view that there was no technical need to grasp the flexible conduits in the array, this passage unambiguously states that the “various” lines in the array are grasped. The natural meaning of this is that the tensioner grasps, for the purpose of applying tension, each of the lines in the array, including the flexible conduits. This is consistent with another passage describing the combined straightener/tensioner:

“Upon activation of the hydraulic ram pairs 1034 and 1036 the two assemblies 1020 and 1022 can be closed on either side of the pipeline array so that the individual lines are caught between the opposing line support pads which are mounted on the endless sprocket chain track sets 1042 and 1044 as shown in figure 37.” (Col 29 lines 47 to 53)

The use of words “the individual lines are caught” reads naturally onto Rockwater’s construction that the tensioner (or in this case straightener/tensioner) grasps and applies tension to each line in the array, including the flexible conduits. On the other hand it appears to be inconsistent with Coflexip’s construction that each of the flexible lines in the array passes through the straightener/tensioner while, at most, merely touching the internal surfaces of it.

Furthermore there is another passage in the description of the straightener/tensioner which suggests that it is providing all or most of the tension which reacts the tension of the suspended array beneath the vessel:

“The tensioning function is provided by motive force import through the hydraulic motors 1046s, p, 1048s, p, 1050s, p, and 1052s, p, which are connected to the main axles as described above. The input power from these motors permits the tensioning along the pipeline array in an upward direction as shown in Figure 38 in order to maintain desired operating tension on the pipeline array which passes downwardly through the pipe array clamp 367 and then into the water.” (Col 30 lines 32 to 40)

There is one other passage describing the operation of the straightener/tensioner which supports Rockwater’s construction:

“The multiple line track sets 1042 and 1044 of each of the assemblies 1020 and 1022 are interconnected to one another through the sprocket gears and main axles 534 and 536. This interconnection provides for moving the lines in the operational array at a common velocity in the same manner as provided by the interconnected grooves of the pipe take-off drum 26 in Figures 1-7. The pipeline support pads on the track sets provide the supporting means for the array.” (30-41, 30-49)

This is teaching that drives for the caterpillar tracks in the straightener/tensioner ensure that all the lines in the array move at a common velocity. However the tracks could only do this if, through the support pads, they are gripping the lines in the array so as to ensure that they advance together at the same speed. Again, this is inconsistent with Coflexip’s interpretation that the flexible lines in the array do no more than touch the pads in the straightener/tensioner and are not pulled by it.

All the above passages in Recalde US support Rockwater’s construction of the document. There are no passages which are inconsistent with that construction. On the other hand each of the passages above taken separately and together are inconsistent with the construction advanced by Coflexip.

Mr Miller runs two other arguments in support of his clients’ construction of Recalde US. First, he says that the size of the tensioner in the specification indicates that the patentee did not intend all the tension of the operational lines (whether rigid or flexible) to be taken by the tensioner. Second, he says that it does not teach, or purport to teach, a tensioner with the ability to apply different tensions to rigid pipes on the one hand and flexible pipes on the other.

The first argument may be summarised as follows. Mr Miller says that the tensioner depicted in the drawings to Recalde US is, to use his word, “puny”. He relies in particular on some evidence given by Mr Nash. He had been asked to look at a model of Recalde US which had been made by Rockwater for the purpose of the litigation. Under cross-examination he said:

“A.

Yes. I have already covered it briefly this morning but if I can just say it again. That model appears to have been built in general in the way Recalde envisaged where no load was carried on the tensioner. The tensioner is a very modest tensioner. … [goes through various calculations] .... At the stern of the pipe take-off drum down to its bottom, it is about nine metres. There are three elements with this, but if I can just stick with the two. I believe what it is showing is a tensioner which is designed to pull out rigid pipe to straighten it. It is not designed to take any vertical load. If it was going to be representative of a system which was designed for the tensioner to take the load, then it is completely wrong because the height of the tensioners would be such as to extend from the bottom of the pipe take-off drum to a substantial height above the top of the pipe take-off drum. The scale is just out of order. It is just not relevant. If you want me to go back and point you to the numbers in my table as to the weight of the pipe and then I will take you to the SAS manual where it gives you the length of the tensioners, it is very easy to show that even looking at 1,000 metre water depth and not the 2,200 metre water depth as a tensioner basis it is just .... I am afraid the dimensions are just not right, assuming the tensioner takes the load.” (Transcript Day 4 pages 508-9)

Mr Miller relies particularly on the two italicised sentences.

It is difficult to reconcile Mr Nash’s suggestion that the tensioner is not designed to take any load with the express teaching of the specification, referred to above. Even if one modifies Mr Nash’s evidence to an assertion that the tensioners are not designed to take all or the majority of the load, not only is this inconsistent with the teaching of the specification but it appears to have been reached by an illegitimate route. The Rockwater model was made by scaling dimensions from the drawings in Recalde US. Mr Nash then measured the length of the tensioner and came to the conclusion that it was too short to support the whole weight of the array of pipes in very deep water. There are a number of problems with this. First, although the drawings in Recalde US are helpful in understanding the teaching of the patent, they are not blueprints, as everyone agreed, and there is no basis for treating them as scale drawings. Second, there is no suggestion that they set out the only designs of tensioner possible. Indeed there was extensive evidence that the “pulling capacity” of tensioners could be greatly enhanced by careful selection of the design of the support pads on the caterpillar tracks. Third, even if one comes to the conclusion that there is insufficient space to fit a tensioner with capacity to take all or substantially all of the load generated by the suspended array below the vessel in very deep water, that does not address the versatility of the apparatus. As mentioned above, Recalde US states that the apparatus can be used in water less than 200 feet deep. There was no suggestion from Mr Nash that even modest sized tensioners would be unable to support all the weight of the lines in shallow water. I reject this criticism of the prior art.

Mr Nash’s second point as to the difficulty which would be experienced in trying to apply different tensions to rigid pipes on the one hand and flexible pipes on the other was the basis for a submission by Mr Miller to the following effect. The skilled man fairly interpreting Recalde US would not be inclined to construe it as advancing an unworkable system (in which the tensions of different operational lines in the array are taken on the tensioner) when a perfectly reasonable alternative interpretation of Recalde US exists (in which the tension is taken on the reel) (Footnote: ⁴). In the light of the conclusions set out above, I do not accept that the Coflexip construction of this prior art is a reasonable one as suggested. But in addition to this, I do not accept the premise underlying this submission. No doubt it would be difficult to set up the apparatus illustrated in the figures of the patent to unspool simultaneously 6 or 8 pipes and flexible conduits of very differing outside diameters, surface characteristics and weights. I am not persuaded that it would be impossible to do so. However this misses the point. Recalde US teaches the design and operation of a piece of apparatus which is intended to be very versatile both as to the nature of the lines being laid and the depth of water in which the vessel operates. In particular it teaches that the apparatus can be used to lay from 2 to about 8 operational lines (Recalde US: Col 11 lines 65-6).

There is no credible evidence that it would not be possible to lay simultaneously two lines, one rigid and the other flexible, with the tensioners described in Recalde US. As explained in relation to Annexes 5 and 6 above, the specification teaches the use of multiple tracks within the caterpillar tensioner, each one of which carries support pads. Each of those is adjustable by the jactuators. The purpose of the jactuators is stated in the specification by reference to Figures 17 to 19 of the specification (see Annex 4):

“Each of the idlers in the idler sets 602, 604 and 606 is similarly provided with an adjustment screw and a jactuator for adjusting the position of the idlers in order to contact the operational lines array with the pipeline support pads 624, 626, 628, 630, 632 and 634 as shown in FIG. 19. This individual adjustability feature for each idler roller in the roller sets 602, 604 and 606 then permits various curvatures to be established for each of the operational lines in the array.

If desired the pipeline support pads 624-634 can be joined into a single transverse pad extending across the width of the operational lines array 35 when different curvatures between the operational lines are not needed.” (Col 20 line 60 to Col 21 line 5).

It must be recalled that this passage is describing the Annex 4 type of device. As noted above, although the particular description is directed to the device when used as the straightener, it is also to be used as the tensioner (subject to the addition of sufficient hydraulic power). The function of the adjustment screws and jactuators is to force the idler rollers, tracks and conduit support pads forwards. This is designed to force the rigid pipes to be plastically deformed back into a straight configuration. When used as the opposing faces of the caterpillar tensioning device, it is apparent that the adjustment screws and jactuators will be operated to ensure that the pipes and conduits are sufficiently grasped between the support pads to have tension applied to them by the tensioner. There does not appear to be any point in operating them in any other way. Furthermore the specification teaches that the apparatus can be adjusted to cater for the laying needs of each individual pipe or conduit. Thus it teaches that there can be a single support pad on a single track to interact with a single line and that each pad is replaceable so as to best fit the pipe or line it is supporting:

“The tracks of 322 can be arranged with operational lines support pads extending across two or three sprocket chains so that the operational lines array is contacted at a given position by a single support pad. Alternatively, separate support pads can be mounted on each of two or three sprocket chains in order to contact single operational lines or pairs of lines as shown in figure 10.” (Col 19 lines 16 to 22)

and

“In this manner the pipe support pads 624-634 can be exchanged and/or replaced in order to accommodate various operational line arrays having different diameter operational lines therein.” (Col. 21 lines 48 to 51)

There is no reason to believe that this apparatus cannot be operated successfully, for example to lay two lines, one being flexible, in shallow water with all or substantially all of the tension being taken on the tensioner. In particular, I reject Mr Miller’s argument that the apparatus is so clearly incapable of applying tension to the flexible conduit that a man in the art would try to construe it as excluding that possibility.

Thus far I have considered how the inventor described the way in which Recalde US should be operated. Assume, contrary to the findings above, that Mr Miller is right and that it would be impossible with the apparatus described to lay rigid and flexible conduits simultaneously in a way in which any or most of the axial tension in the latter was taken or reacted by the vertical tensioner. On this assumption, since the patent always requires a rigid pipe to be present, there is no teaching of laying a flexible conduit with the tension being taken by the vertical tensioner. However, particularly for the purposes of considering the issues of obviousness and the validity of claim 3 of the Coflexip patent, it is necessary to ask whether or not the Recalde apparatus is “suitable for” handling a flexible conduit in that way. In my view there can be little doubt that it is. For example, if the operator of a vessel built to the Recalde design only wanted to lay a single flexible conduit in shallow water, he could use it to do so. In that eventuality the support pads and tensioners could be set up to grasp that conduit to apply axial tension to the line. In such a case there would be none of the alleged difficulty caused by co-laying flexible and rigid conduits. Therefore the apparatus is capable of being used in that way. It should be recalled that in the passage from Recalde US set out at paragraph 99 above, the authors state that to be commercially viable a pipelaying vessel must be capable of laying a single operational line is shallow water. It is difficult to believe that Recalde US was not intended to be commercially viable. This leads me to Recalde GB.

Recalde US and Recalde GB are discrete prior publications. They have to be assessed on their own teachings. In fact most of the teaching is identical. Many of the same drawings are used. There are only two points of difference which are of significance to the issues in this action. First, the description of the way in which the array passes through the straightening and tensioning sections is worded somewhat differently in the two patents. The relevant passages in Recalde US are set out at paragraphs 123 and 127 above. Recalde GB is clearer:

“Each of these lines can be passed over the pipe take-off drum 26 and then passed through the straightening device 32 and the tensioning device 34, even though the plastic, electrical, and support lines may not require straightening and hence are passed through without the straightening device being in operative contact in order to use the layout drum 26” (Page 7 lines 29 to 37)

What this passage appears to be saying is that all lines pass through the straightening and tensioning stations but the flexible ones miss out operative contact with the straightener because they do not need straightening. The inference is that the flexible conduits are in operative contact with the tensioner. That means that the tensioner applies tension to them. Coflexip’s riposte to this was left to Mr Nash:

“126.

… Recalde GB teaches that, although such a line is to be passed through the straightening device 32 and tensioning device 34, the straightening device is not in operative contact (see page 7, lines 29 to 37). As Recalde GB makes clear, the reason for passing the line through the straightening and tensioning device in this manner is so that the line may use the layout drum 26.

127.

Recalde GB only states that the straightening device is not in operative contact with the plastic line, electrical cable or tension support cable. However in the case of the dual-function straightener/tensioner device, this will inevitably mean that the tensioner is also not in operative contact. Thus, so far as concerns the dual function straightener/tensioner device, the system is intended to work without use of the tensioner. But that being so, I do not believe that the system is intended to work any differently when the two-part straightener and tensioner is utilised. And, of course, the tensioner is in fact a part of the straightener. In my opinion, when Recalde GB says (at page 7, lines 29 to 37) that the straightening device is not in operative contact, it is clear that the intention is that the line will not be in operative contact with the straightening device (32) and the tensioning device (34).” (Nash First Report)

In my view this interpretation is neither accurate nor convincing. What Mr Nash has done is to incorrectly apply the teaching in the passage in Recalde GB to the case of the dual-function straightener/tensioner so as to change its sense and then has applied that modified sense back to the case of the separate tensioner and straightener. The error occurs in the first three sentences of his paragraph 127. He appears to have read the passages in Recalde GB as if it prohibited contact between the flexible conduit and the straightener and, for that reason, it must prohibit contact with the dual-function device. That “inevitably” means that the tensioner is not in operative contact in that case. Following from that it is “intended” that the system should work without the tensioner. That is not what the passage says. It merely notes that, because flexible conduits do not need straightening, they do not need to be in operative contact with the straightener.

The meaning of the above passage in Recalde GB is clear. The operative contact between the straightener and the flexible conduits is not necessary, but there is operative contact between the tensioner and each of the pipes/conduits. This patent teaches that the tensioner reacts the axial tension in the flexible conduits.

The second point is as follows. Whereas it is said that Recalde US does not expressly teach the laying of flexible conduit alone, a matter which is only of importance if Coflexip is correct in its assertion that co-laying of rigid and flexible conduits would render it impossible for the tensioner to apply tension to the latter, Mr Thorley says that Recalde GB expressly teaches laying flexible conduit by itself. If so, it must also teach that the tensioners react all or substantially all of the axial tension in that conduit.

Mr Miller does not suggest that there is any passage in Recalde GB which expressly states, as Recalde US does, that the apparatus is for laying at least two lines simultaneously one of which must be a rigid pipe. There is no reason to read any such requirement into the specification. Furthermore Mr Thorley points to the following passage:

“The pipelaying system described herein is adapted for laying out single operational lines including rigid walled pipelines onto the bottom of bodies of water varying in depth from about 200 feet to over 3,000 feet. …

The pipelaying system comprises one or more storage reels 20, 36 and 40 from which operational lines including a rigid walled pipeline can be sequentially unspooled and laid out on the bottom of the body of water by passing each of the operational lines, in turn, through the pipeline laying device.” (page 18 lines 89 to 114)

I accept Mr Thorley’s submission that this teaches that the apparatus is to be used to lay a single line which can be flexible. In such circumstances the alleged difficulties caused by trying to lay rigid and non-rigid conduits at the same time do not arise. The apparatus is therefore intended, inter alia, to lay flexible conduits alone. In doing so the tensioner reacts some or most of the axial tension in the conduit. Furthermore, for reasons explained in relation to Recalde US, the equipment described in Recalde GB is capable of being used for laying flexible conduit, whether alone or with other conduits, in a pipelaying operation in which the axial tension in the conduit is taken by the tensioner.

There is only one other issue which needs to be addressed in relation to the Recalde US and Recalde GB prior art, namely the overboarding of accessories. As discussed at the beginning of this judgment, flexible conduits normally have accessories or fittings attached to them which have to be passed through or over the pipelaying equipment into the sea. They will include the end fittings which begin and terminate such conduits. These, like other fittings, will have a larger outside diameter, and be more rigid, than the conduit. Neither of the Recalde patents refers expressly to this part of the pipelaying operation. Mr Miller argues that it would be impossible to pass accessories through the tensioner device at the same time as it is laying rigid pipe. To pass accessories, it would be necessary to open the jaws of the tensioner. Opening and closing of tensioners was well known and both pieces of prior art describe how the caterpillar tracks can be moved towards and away from each other. Recalde US expressly says that the tensioner can be opened in this way (see Col 30 lines 12 to 15). However Mr Miller says that opening the tensioner would cause insurmountable problems if a rigid and flexible pipe were being laid together. As the jaws of the tensioner are opened to allow the accessory on the flexible conduit to pass, the tension on the rigid pipe would be released. This would make it extremely difficult or impossible to lay the latter pipe. For that reason the Recalde apparatus will not work.

I do not accept this submission for two reasons. First, what is primarily in issue is the capabilities of the Recalde apparatus. As mentioned already, each of the pipe reels is powered. It is not disputed that these reels would be supplied with sufficient power from their motors to support the dependent pipes without the assistance of the tensioners. The patent says as much in the passage in which it states that the total weight of the suspended pipe can be taken on the reels with the tensioner and straightener disengaged. Indeed it is part of Coflexip’s case that in all cases all or most of the tension is taken on the pipe reels and the take-off drum. It must follow that if the tensioner is opened to allow an accessory on a flexible conduit to pass, all the tension on the rigid pipe could be taken on or reacted by the reel associated with that pipe. In other words, the Recalde apparatus is capable of working in that way. Furthermore, for reasons set out above, Recalde US would be capable of being used for passing flexible conduit by itself and Recalde GB expressly discloses laying pipes and conduits singly. Where only a flexible conduit is being laid this alleged problem does not exist.

This leads onto the only other point which needs to be made in relation to the overboarding operation. The Recalde patents disclose the existence of two A&R winches. One (362) is located on the take-off structure. It can be seen in Annex 2. The other is the twin drum winch (62) with its associated cable storage reel (50) on the top drawing of Annex 1. It is towards the prow of the vessel. The latter A&R winch mounts two cables (52, 54). As their names indicate, these winches are for assisting in the abandonment and recovery of pipes and conduits. For example, during bad weather it may be necessary to stop pipelaying and to disengage the conduit from the pipelaying equipment. This involves letting the end of the pipe down into the water and subsequently recovering it when the weather has calmed. To do this, the cable on an A&R winch was connected to the conduit, it was used to allow the pipe to pass to the seabed in a controlled manner. Later the cable with the conduit attached would be brought back on board. A&R winches were also used to facilitate the overboarding of accessories. This was explained by Mr Coutarel in explaining how a conventional A-frame could be used:

“46.

If the rigid accessory was a single end fitting (that is to say, the end of the length of pipe had been reached) as the pipe left the reel, an abandonment cable was attached to the end fitting. The cable was unwound until just before the end fitting reached the gutter. The rigid accessory was connected to the crane which then took up the tension from the reel. The crane was operated in the same way as described above. Once the end fitting had been lowered beneath the level of the gutter by the crane, the tension was taken up again by the cable attached to the reel and the crane disconnected. The end of the pipe was then lowered to the sea bed using the reel cable in the normal way (the powered reel being used as an abandonment winch).

47.

A similar procedure (but in reverse) can be used to recover a pipeline abandoned on the seabed.

48.

Sometimes when overboarding an end fitting, rather than attaching it to a cable on the powered reel, the end fitting was instead attached to a deck winch to carry out the abandonment (or recovery) operation. As the end of the pipe came off the reel, the tension was transferred from the reel to the deck winch. The end fitting was overboarded using the crane in the same way as I have described above. The deck winch would typically have a tension capacity of 10-30 tonnes depending on what was required which would, of course, depend on the size of pipe and depth of water in which it was to be laid.” (Coutarel Witness Statement)

The cable on A&R winch (362) or one of the cables on double A&R winch (62, 50) could be used to pass an end fitting or an accessory through the opened tensioner. If the front A&R winch (62, 50) were to be used, the cable would be taken over the top of the take-off drum (24) and down through the open mouth of the tensioner (34). Professor Witz gave evidence to that effect in his first report (paragraph 100 et seq) and he returned to it in his third report (see, in particular, page 15). He said he would have preferred to use the double winch. None of this was challenged during cross-examination. In fact the only comment made by Mr Miller in relation to the use of the A&R winches in the Recalde specifications for passing accessories through the tensioner was the following passage in his closing skeleton argument:

“(a)

Auxiliary tensioning means capable of being connected to the rigid accessory. The winch relied upon is an A&R winch for recovering pipe to the pipe clamp and is plainly not inevitably sheaved for lowering accessories through the tensioner.” (page 22)

This is only directed to the stern A&R winch (362). In my view there is no challenge to the assertion that the forward winch could readily be used in the way that Professor Witz explained.

I can now consider the issues of validity. Each side attacked the competence of the other’s expert witness and each said that the other’s expert’s views had been clouded, in Mr Nash’s case, by too close a connection with this and prior litigation and, in Professor Witz’s case, by knowing what the issues were in the action before reading the prior art. In my view the notional skilled team employed to design a pipe-laying vessel and considering the disclosure of the two Recalde patents would have included the expertise of the kind possessed by both Professor Witz and Mr Nash. Both were qualified to give evidence in this case. Furthermore I reject any suggestion, if it was made, that either of these witnesses modified his evidence to suit the needs of the parties. In my view they were both honest and doing their best to help. At times I felt that Mr Nash’s reading of some passages in Recalde US would not have been the reading of someone who was less connected with this litigation. But the fact that Mr Nash seemed to read the prior art through the eyes of the party for whom he was giving evidence was, I am sure, not deliberate and certainly not cynical but a reflection of his close association with the issues over a long time. I found Professor Witz an admirable witness.

Allegations of anticipation were levelled at claims 1 and all of claims 3 to 9. By the time of his speech, Mr Thorley had decided to downplay, if not expressly abandon, that attack as far as claim 1 is concerned.

To succeed on anticipation, Rockwater would need to prove that the pleaded prior art contains clear and unmistakable directions to do something falling within the scope of the claims. In the terminology of the EPO, it must be possible to derive the features of the invention unambiguously from the teaching of the prior document. In the case of Claim 1, this is a non-starter. Much of this process claim is concerned with the passing of accessories through the pipelaying equipment. None of that is even referred to, let alone described, in the pleaded prior art.

As far as claim 3 is concerned, the position is not so simple. The claim requires certain pieces of hardware “for operating the process according to any one of claims 1 and 2”. For reasons set out above, notwithstanding the density of the text, Recalde US and Recalde GB both disclose apparatus which meets all the hardware requirements of the claim. Thus they disclose linear winch-type tensioning means with a substantially vertical axis which is capable of ensuring the normal lowering of a flexible conduit by gripping the outer surface of the conduit. That is the tensioner or straightener/tensioner illustrated in the Annexes to this judgment. That tensioning means is capable of being laterally moved apart. It is not disputed that it is the last means for guiding the conduit. Furthermore there is a disclosure of A&R winches with cable attached. The latter come within the requirement of the claim for an auxiliary tensioning means comprising at least one elongate movable traction element capable of being connected to a rigid accessory mounted on the conduit. However what has to be disclosed is a piece of hardware “for operating the process” of claim 1. Thus it must disclose a piece of hardware capable of operating that part of the process which involves connecting the accessory to an “auxiliary tensioning means” and then lowering the accessory through the open tensioner. The auxiliary tensioning means relied on by Rockwater is the A&R winch and its associated cable.

If one has regard to the A&R winch (362) in the take-off structure, there is no disclosure of any means by which its cable can be passed upwards so as to enter the top end of the vertical tensioner. To allow that to happen, an additional chute or sheave would be needed above the tensioner. No doubt such a component could be installed with no difficulty and it may well be that it would be obvious to do so, but there is no disclosure of such an arrangement. Therefore, in this respect the hardware in Recalde is not suitable for operating the claim 1 process. However this is not the only A&R winch. The other winch (62) is located between the main reel (20) and the take-off drum. In this case there is suitable equipment to allow the winch cable to be introduced into the top of the tensioner. That equipment consists of the take-off drum (26). For this reason, claim 3 is anticipated.

Coflexip argues for independent validity of all the subsidiary claims 4 to 9 if claim 3 is invalid for anticipation only. It is not necessary to set out all of these claims here. They are dealt with briefly in paragraphs 83 to 88 of Mr Thorley’s closing skeleton argument. Mr Miller addressed none of these claims. In my view there is nothing in any of them. If claim 3 is anticipated, so are they.

This leaves the issue of obviousness. This has to be considered in relation to Claim 1. In view of the substantial arguments put before me and the possibility that a different view on anticipation may be taken by an appellate court, it is also appropriate to consider the obviousness attack in relation to the apparatus claims. Rockwater only relies on Recalde US and Recalde GB. It does not pursue a case of obviousness over the common general knowledge alone. Claim 1 will be considered first.

In the Stolt action, I described the inventive concept in that claim as follows:

“52.

The inventive concept embodied in the patent is the provision of linear winch-type tensioners with substantially vertical axes so arranged that, in use, they take all or substantially all of the pull exerted by the dependent conduit. As explained above the method of passing accessories through the tensioners by opening and closing them to allow the accessories through was known. In fact it is difficult to see how accessories could be made to pass through linear tensioners without opening them and no alternative mechanism was advanced as a possibility. There is no peculiar or unexpected interaction between the method of passing the accessories and the use of the vertical tensioners. The former can be ignored for the purpose of assessing obviousness since it adds nothing to the inventive concept in the patent”.

Mr Thorley says that the analysis is correct. Mr Miller does not disagree. Mr Thorley invites me to follow the structured approach suggested by Oliver LJ in Windsurfing International Inc v Tabur Marine (Great Britain) Ltd [1985] RPC 59 at 73. According to that, it is necessary to identify the differences between inventive concept and what is taught by the prior art. Mr Thorley identifies only two possible candidates; (i) all or substantially all of the tension being taken on the tensioner and (ii) the particular sequence adopted for passing accessories. The first of these is not a difference at all if the analysis of the Recalde patents as set out above is correct. However, for the purpose of considering this attack on validity, I will assume that neither of these prior art documents expressly discloses this feature.

The last stage is to determine whether, viewed without any knowledge of the alleged invention, either or both of these differences constitute steps which would have been obvious to the skilled man or whether they require any degree of invention. In answering that question it is necessary to keep in mind the width of claim 1. The claim will be invalid if it is proved to be obvious to operate anywhere within its scope. In particular, the claim is not restricted to operations in deep water, nor is Recalde. It is convenient to consider how a man in the art would operate the latter in shallow water. Further it should be assumed that he has decided to lay a single flexible conduit. This is a reasonable starting point because Recalde GB expressly tells him that he can do so and, in the case of Recalde US, it is obvious that it can be used in this way. It is apparent that a pipelaying vessel which can be used to lay different numbers of pipes and conduits simultaneously and is versatile enough to accommodate different sizes of pipes, could be used to lay one alone.

Professor Witz told me that it would be standard practice for the control circuits to the pipe reel and the tensioner to be independent of each other. It would be possible to turn one on and the other off, to drive them in different directions or to make them take different amounts of the tension in the line. At the very least, Recalde teaches taking some of the tension on the vertical tensioner. The question is whether it would be obvious to operate it so as to take most of the tension. In my view, it would. In shallow water where the weight of suspended conduit was well within the pulling capacity of even a small tensioner, the skilled addressee would know that he could take all the tension on the tensioner, or the reel or a mixture of the two. Which he would use would be a matter of operational convenience. It would be obvious to take all or most of the tension on the tensioner.

In my view the same answer is obtained if one considers the operation of Recalde in deep water. Both Recalde patents teach the design of a pipelaying vessel which can lay up to 8 pipes and conduits simultaneously. Assume that such a vessel is now being operated so as to lay a single small flexible conduit in deep water. As Mr Thorley points out, the evidence was that any skilled operator would know that he would have to consider all the loads operating on a conduit when deciding how best to lay it. This would be particularly important when the conduit is to be laid in circumstances where it will be subject to large loads, such as when it is being laid in deep water. He would know, because it is a matter of common general knowledge, that the high axial loads on a conduit while it passed over a curved surface would increase the risk of crushing. For that reason and as a matter of course he would need to consider the loads which are generated in the conduit as it passes over the take-off drum and on the reel and he would be aware that reducing the axial tension in the pipe on those locations would reduce the risk of damage to the conduit. He could not help but know that operation of the tensioner to support the weight of the suspended conduit (as expressly taught by the Recalde patents) would inevitably reduce the crushing loads on the conduit. In those circumstances, it is almost inevitable that he would think of operating the tensioner so as to reduce the upstream tension in the conduit as much as possible. For these reasons, taking most of the tension on the vertical tensioner is obvious.

Indeed the practical application of the Recalde design makes this all the more likely. It will be recalled that by 1990 reels for pipe had a pulling capacity of 10 to 30 tonnes. Mr Coutarel gave evidence that systems comprising a powered reel and overboarding gutter continue to be used today to lay flexible pipe. Although the motorization units of the reels have improved slightly, even now their maximum capacity is only in the region of 30-35 metric tonnes (see Coutarel Witness Statement paragraph 54). However for laying flexible pipe in deep water much greater capabilities than that are required. This was explained by Mr Nash:

“…the in-water laying weight of flexible pipe with diameters of 3" and 12" suitable for water depths of 1000m would be approximately 29kg per metre and 173kg per metre respectively. This would mean that the theoretical maximum pipe tension when laying an 3" flexible pipe in 1000m would be approximately 29 tonne and laying a 12" flexible pipe in the same water depths would be approximately 173 tonne. However, in actual practice a dynamic amplification factor would need to be applied to the maximum pipe tensions which would result in required lay tensioner capacities of approximately 40 tonne and 230 tonne for the 3" and 12" flexible pipe respectively.” (First Report paragraph 69)

Thus, someone laying flexible line in deep water would know and be used to the fact that most of the tension would normally be taken by the tensioners. That would be the experience of someone considering the operation of the Recalde vessel. The fact that the tensioner is orientated vertically would not deter him from considering taking most of the tension on it.

As far as passing an accessory is concerned, it appears that there is really little option but to open the tensioner to allow it to pass. If this is done, the pulling power of the tensioner is lost. The tension must be taken by some other equipment. The use of an A&R winch to perform one of its normal functions, namely to react the tension on the suspended conduit while an accessory is overboarded, is a standard and obvious thing to do. I can not see how this difference between what is described in Recalde and what is claimed in claim 1 can be inventive. Inevitably this will involve arranging for the cable on one of the A&R winches to be presented to the top of the tensioner. As explained in relation to the arguments on anticipation, this would involve no additional hardware if the forward A&R winch (62) were to be used. Its cable would simply be taken over the take-off drum. In the case of the stern A&R winch (362), this might involve use of additional pulleys or a chute, but there is no suggestion that any of this would be other than a trivial and common or garden engineering modification. It follows that claim 1 fails for obviousness. Coflexip did not assert independent validity for claim 2.

For the reasons set out above, the Recalde vessel contains all the hardware required of the apparatus claims. If it is obvious to operate Recalde US or Recalde GB in accordance with the process claims, then the apparatus claims must also be obvious. Although Coflexip asserted that some of the subsidiary claims were independently valid, no arguments were advanced to support that position. All the apparatus claims fail for obviousness.

In the result, Rockwater succeed both on the issue of infringement and invalidity.