Associated British Ports v Hydro Soil Services NV & Ors [2006] EWHC 1187 (TCC) (23 June 2006)

Neutral Citation Number: [2006] EWHC 1187 (TCC) Case No: HT-04-35 IN THE HIGH COURT OF JUSTICE QUEEN'S BENCH DIVISION TECHNOLOGY AND CONSTRUCTION COURT Royal Courts of Justice Strand, London, WC2A 2LL 23/06/2006 B e f o r e : HIS HONOUR JUDGE RICHARD HAVERY Q. C. ____________________ Between: ASSOCIATED BRITISH PORTS Claimant and HYDRO SOIL SERVICES N. V. and DREDGING INTERNATIONAL (UK) LIMITED and (1) HAECON N. V. (2) GERLING-KONZERN BELGIE N. V. (3) AGF BELGIUM INSURANCE N. V. (4) EUROMAF S. A. Defendants/Part 20 Claimants Part 20 Defendants ____________________ Mr. Martin Bowdery Q. C. and Mr. Patrick Clarke (instructed by Macfarlanes) for the Claimant Mr. Adrian Williamson Q. C. and Mr. Adam Constable (instructed by Davies Arnold Cooper) for the Defendants/Part 20 Claimants Mr. Nigel Jones Q. C. and Mr. Paul Reed (instructed by Pinsent Masons) for the Part 20 Defendants Hearing dates: 17th, 24th, 25th, 26th, 27th October, 2nd, 3rd, 7th, 8th, 9th, 10th, 14th, 17th, 21st, 22nd, 23rd, 24th, 28th, 29th, 30th November, 1st, 5th, 6th, 7th, 8th, 12th, 13th December 2005, 30th, 31st January, 1st, 2nd February 2006. ____________________ HTML VERSION OF JUDGMENT ____________________ Crown Copyright © CONTENTS Introduction The clause 12 claim The condition of the sheet pile wall before the strengthening works began. Proposition number 17 The magnetic sonde Displacement as a guide to bending moment Bending stiffness Curvature Netsurvey 2004 State of 204 Rake of sheet piles Soil pressure The relieving platform Effect of grout pressure on SSP wall Sensitivity to grout pressure Grouting effluent blockages Foreseeability Access to ABP's archive The survey Decision on clause 12 claim Fitness for purpose Meaning of fitness for purpose Fitness for purpose Shearing force Axial load Taylor's paper on shear resistance, aggregate interlock etc. Fitness; factor of safety Factors of safety Characteristic strength The arch Slip circle Bond Loss of composite action Verticality of reinforcement Grouting works Mr. Williamson's submission The coping beam claim Grout disposal claim Terms of contract regarding dumping area Parent company guarantee The Part 20 claim against Haecon Meaning of fitness for purpose His Honour Judge Richard Havery Q. C. : Introduction These proceedings arise out of a contract made in March 2002 between the claimant ("ABP") and the first defendant ("HSS") under which HSS was engaged to strengthen the quay wall of berth 205 at the port of Southampton to allow that berth to be dredged from its existing depth of -12.8mCD (i.e. 12.8 metres below Admiralty Chart Datum) to -16.0mCD plus 0.5m overdredge allowance. The contract was a lump sum contract in the sum of £3,856,465.00. The contract was based on the ICE Conditions of Contract, 6 th edition, as amended by the parties. On 29 th May 2002, ABP consented to the assignment of the benefit of the contract by HSS to the second defendant ("DIUK"). DIUK has been represented before me jointly with HSS. In this judgment I shall use the expression HSS to refer to the first defendant or both defendants indiscriminately. Under the contract, HSS was responsible for the design and construction of the strengthening works. It subcontracted the design of the works to the Part 20 defendant ("Haecon"). Where appropriate, I use the expression "the defendants" to include Haecon. The quay wall that required strengthening was a steel sheet pile wall (sometimes called the SSP wall) at the water's edge topped by a coping beam. It was propped horizontally at the top at intervals by vertical triangular counterforts. The piles were about 27m in length. In cross-section, the piles were approximately U-shaped, alternate piles being oriented in opposite directions, and adjacent piles being interlocked. Behind the wall, at a depth of about 6m, was a relieving platform consisting of a concrete slab. The relieving platform was supported by H-piles (piles with an H-section). Its purpose was to bear the weight of the overlying soil and the cranes and other things on the dock and to transmit that weight to a low level, thereby reducing the lateral forces acting on the wall. The quay wall was constructed on reclaimed land in the early 1970s. The design of the strengthening works incorporated the installation of ground anchors attached to the steel wall and the installation of very high pressure (VHP) vertical grout columns along the length of the berth. The VHP columns were installed by a process known as jet grouting. That process involves pumping grout, a mixture of cement and water, down a pre-cut hole so that it jets out of nozzles in a rotating "monitor" at the bottom of the hole and penetrates and mixes with the surrounding soil. The monitor is gradually drawn upwards while the jetting proceeds. The mixture of soil and grout makes a material commonly called soilcrete, which hardens forming the VHP columns. The columns are commonly called the grout wall or the VHP wall, but the expression refers to a series of discrete columns, whose average separation as designed was approximately 1.7m. Each column was in the form of a pair of cylindrical columns each of design diameter 900mm. The axial separation between the members of each pair was designed to be 700mm. The second was constructed immediately after the first and before the first had hardened, so the two members of the pair were designed to function as a single column. The strengthening works took place between March and November 2002. During the course of the works the sheet piles cracked and bulged outwards by an amount up to one metre. It is common ground that further remedial works are now required to enable berth 205 to be dredged to -16mCD plus 0.5m overdredge allowance. The principal dispute is as to the cause of failure of the strengthening works and which party bears responsibility for payment for the necessary remedial works. ABP claims that the strengthening works were not fit for their purpose. It is said that remedial works costing more than £11 million are required. ABP claims against HSS a declaration that HSS is responsible for defective works and for the cost of remedying them. ABP also claims damages against HSS for breach of contract. HSS counterclaims extra payment for the necessary remedial works on the basis that those works have been rendered necessary by an unforeseeable physical condition, namely that there were plastic hinges [1] in the existing sheet pile wall. In the event that HSS is found liable to ABP by reason of defects in the design of the strengthening works, HSS seeks from Haecon an indemnity, contribution or damages for breach of the design contract. Under its counterclaim, HSS also has two claims against ABP which have been entitled the coping beam claim and the grout disposal claim. I consider those claims below, at paragraphs 299 et seq. and 314 et seq. respectively. The two major disputes are (1) ABP's claim against HSS for the relief mentioned above on the basis that HSS was in breach of contract in that the works were not fit for their purpose, and (2) HSS's claim against ABP for extra payment on the basis that HSS encountered an unforeseeable physical condition. I shall consider the latter first, but before doing so I shall identify the expert witnesses in this case. The following expert witnesses gave evidence before me. Professor Milija Pavlovic Sc. D (Cantab.), Professor of Structural Engineering and Mechanics and Head of Section (Concrete Structures) at Imperial College, Fellow of the Institution of Civil Engineers, Fellow of the Institution of Structural Engineers, Fellow of the Association for Consultancy and Engineering; Mr. Anthony Bracegirdle M. Sc (Imperial College), Member of the Institution of Civil Engineers; Mr. William M. Reid, Fellow of the Institution of Civil Engineers, Fellow of the Institution of Structural Engineers, Fellow of the Royal Academy of Engineering; Dr. David Greenwood Ph. D., Chartered Engineer, Fellow of the Institution of Civil Engineers; and Mr. Anthony O'Brien M. Sc. (Imperial College), Chartered Engineer and Member of the Institution of Civil Engineers. Professor Pavlovic and Mr. Bracegirdle were instructed by ABP; Mr. Reid and Dr. Greenwood were instructed by HSS; and Mr. O'Brien was instructed by Haecon. References in this judgment in the form (day d, page p) are to a daily transcript of the proceedings commissioned by the parties. The clause 12 claim It is common ground that the effect of clause 12 of the contract is that if during the execution of the Works HSS has encountered physical conditions (other than weather conditions or conditions due to weather conditions) or artificial obstructions which conditions or obstructions could not reasonably have been foreseen by an experienced contractor, then HSS is entitled to extra payment for extra work rendered necessary thereby. In this case HSS rely upon a physical condition, namely the physical state of the sheet pile wall. They say that it was overstressed and contained plastic hinges. That meant that it was unable, without bulging, to resist the pressure of the grout applied at the back of the sheet pile wall when the VHP columns were made. HSS contend that it was reasonable to expect that the sheet pile wall would have had a factor of safety of about 2, and that if it had had such a factor of safety it would not have bulged or cracked when the grouting process took place. Mr. Bowdery submitted that the condition of the sheet pile wall was not a physical condition falling within clause 12, since a contractor should not have the benefit of a clause 12 claim if his strengthening works fail because of some pre-existing condition in the works which he has undertaken to strengthen. Clause 12 was not intended to apply to the condition of the very thing that was to be strengthened. In my judgment, the concept of encountering something during the execution of the works tends to import something adjectival that hinders the work, rather than a substantive element of the works themselves. Nevertheless, the words are wide enough to cover the latter, and the clause must be construed against the person (here, ABP) who offers the document for agreement. Accordingly, I reject Mr. Bowdery's argument. The questions remain, what was the condition of the sheet pile wall, and was it unforeseeable by an experienced contractor. I shall first consider the first of those questions. The condition of the sheet pile wall before the strengthening works began In outline, the position is this. It is common ground that before the works were carried out on berth 205, berths 204 and 205 were in the same condition, and that berth 204 is still in the same condition as it was. The present condition of berth 204 has been investigated in order to determine the condition in which berth 205 was before the works were carried out. Surveys of berth 204 above the level of the river bed carried out before the trial showed that most of the sheet piles were not vertical. A small degree of displacement from the chord at mid-depth was to be expected, but generally the displacement from the vertical (where the chord would be expected to lie) was greater than expected. It was said that the maximum bending moments in the sheet piles would be proportional to the maximum displacements from the chord, a displacement of about 300mm indicating a bending moment at relieving platform level sufficient to create a plastic hinge there. Mr. Reid, pointing out that in berth 204 there were many instances of displacements from the vertical of 300mm and more, expressed the opinion that in consequence there must be many instances of plastic hinges in the sheet pile wall. Mr. Bracegirdle considered that the piles had been driven consistently with a rake, so that the toe of each pile would be seaward of the vertical drawn from the top of the pile. Thus the displacement from the chord would be less than the displacement from the vertical, and one could not conclude that there were plastic hinges in the wall. With the object of resolving this issue, a magnetic sonde was used in three locations to measure the position of the sheet piles both above and below the level of the river bed. The defendants submitted that the sonde survey showed conclusively that the piles had been driven vertically, and that the displacements from the vertical were indeed displacements from the chord. However, the position is not so simple as that. In more detail, it is as follows. Proposition Number 17 In the experts' joint statement, the following propositions appeared: Agreement reached that the properties of Berth 204 and 205 prior to strengthening would have been, for practical purposes, the same. No agreement reached on whether the information available indicates that the sheet piles were driven vertically or at a rake. No agreement that Berth 204 is deformed beyond anticipated design values but agreement reached that if the piles can be shown to have been driven vertically, then this would be strong evidence of excessive post-installation deformation. There followed a section headed "Detailed discussion" which consisted of a number of propositions, each followed by the comments of the individual experts. Proposition 15 referred to some data, and went on: From these data it is apparent that the wall of Berth 204 deviates significantly from a vertical plane. The profile of the wall is outside that expected from tolerances normally applied during the installation of sheet piles combined with the displacements to be expected as a result of excavation of the quay. The profile of the wall does not appear to provide operation or serviceability difficulties to the Port. All parties' relevant experts agreed to that proposition. Proposition number 17 was as follows: The location of the toe at the base of the sheet piles has not yet been determined for any location at berth 204. If it can be established that the location of the toe is consistent with the theoretical pitched alignment of the sheet piles, then it can be concluded that the sheet piles at Berth 204 have deformed beyond their elastic limit. Mr. Reid and a representative or representatives of Haecon agreed to that proposition. Mr. O'Brien gave evidence that he understood the significance of it; but the relevant expert acting for Haecon would have been Mr. Foster, who was a general civil and structural engineer. Mr. O'Brien was a general civil engineer specialising in geotechnical engineering. Mr. Foster was not called to give evidence. Mr. Bracegirdle's comment on the proposition was as follows: If this could be shown to be the case consistently throughout Berth 204, there would be strong evidence to support high post-installation deformations. A magnetic sonde was used at Berth 205, although the methods used [rendered?] interpretation of the data provided by URS [2] questionable. Reinterpretation of this survey suggests the toe of the wall lies to seaward of the head of the wall. The word "rendered" is supplied by me. Thus it was an important matter to determine whether the toe of the sheet piles was located vertically below the position of the sheet piles at the level of the relieving platform. In his expert's report, prepared by 9 th September 2005, Mr. Bracegirdle put forward the thesis that the steel sheet piles had systematically been driven with a seaward rake, i.e. so that the toe was seaward of the head. Thus the displacements of the piles from the vertical would be greater than the corresponding displacements from their chords. He illustrated his thesis with diagrams in his report, contained in figure 24. He said that it was evident from the archive material that the reclaimed bank moved seawards during the driving of the sheet piles to an extent of about 2 inches (50mm). That was confirmed by the witness statement of Colin Gilbert [3] . That movement was consistent, said Mr. Bracegirdle, with lateral loading imposed by the reclamation bank. As outlined in figure 24, movement of the reclamation bank and loading of the sheet piles would have caused curvature of the sheet piles. Seaward movement of 50mm at the top of the reclamation bank would be sufficient to induce a curvature of about 1/360 per metre in the sheet piles beneath the relieving platform. In reality, it would not be possible to carry out a definitive analysis of the complex issue of pile wandering. While the analysis presented was simplistic, it was sufficient to illustrate the mechanism of seaward wandering of the piles during driving. Mr. Reid did not agree with the above analysis. He gave a graphic illustration of how the sheet piles could bow outwards below relieving platform level yet have their toes vertically beneath the positions of the piles at relieving platform level. I return to this point in paragraph 82 below. In order to determine whether the toe of the sheet piles was located vertically below the position of the sheet piles at the level of the relieving platform, a magnetic sonde was used in October 2005. The magnetic sonde determined the horizontal distances of a pile from the measuring equipment at different levels at three locations in Berth 204. Those locations have been identified by the references X1, X2 and X3. The magnetic sonde In the magnetic sonde, an electric pulse is passed through a coil. The pulse creates a magnetic field. The creation of that field produces a response from ferromagnetic objects (e.g. steel) within range of the coil. That response creates a signal in the coil which is detected by the electronics controlling the coil. A survey using the magnetic sonde was carried out by GB Geotechnics Limited on berth 204 on 22 nd and 23 rd September 2005. The work was carried out by cantilevering a drilling rig over the side of the quay wall and drilling into the seabed about 18m below the rig. The drilling rig was used to instal a thick-walled cylindrical plastic tube. The plastic tube contained internal grooves which facilitated the orientation of an inclinometer and the sonde. The inclinometer was used to assess the position of any point in the plastic liner in a plane at right angles to the line of the sheet pile wall and in a plane parallel to the sheet pile wall. The sonde was restrained by the grooves so as to maintain the axis of the sonde coil at right angles to the sheet pile wall. Three boreholes designated X1, X2 and X3 were made at different places along the length of the wall. Verticality of the boreholes was assessed by divers using spirit levels and tape measures at various levels from sea bed upwards. Verticality was also assessed by the use of the inclinometers. Inclination was measured by the force necessary to keep a pendulum in the borehole in an axial position. The inclination was a measure of the rate of change of the horizontal position of the instrument with depth. That was integrated to give the horizontal displacement of the instrument from a given datum, in this case the zero datum being the horizontal position of the inclinometer at the level of the bottom of the sheet pile wall. Mr. George Stephen Ballard gave evidence before me. He founded the practice of GB Geotechnics and was Principal and Director of GB Geotechnics Limited. He stated that the magnetic mass sonde had been developed by his company, originally to measure the depth of iron caissons sunk into river beds in the nineteenth century to support railway bridges. It was not then its primary purpose to measure the distance of the caisson from the sonde. He said in his first witness statement, dated 25 th August 2005, that his firm was satisfied that the out of plumbness measured by the system in relation to berth 205 was accurate to better than 5 per cent. of the recorded value. In a further witness statement, dated 29 th September 2005 and relating to the survey of berth 204, he said that both the magnetic sonde and the inclinometer data from borehole X2 showed strong evidence of contamination by wind, tide and wave, in both distance and tilt: the data had therefore been rejected for combination as a tilt corrected profile. He said that data from boreholes X1 and X3 also showed marginal influence from wave and tide action, but the evidence was that any error was considerably less than the amplitude of the real variations recorded. The data could be combined as reasonably valid tilt corrected pile profiles. He wrote: We are satisfied that the tilt corrected profiles are accurate profiles to within ±100mm of the true profile of the steel sheet pile wall….. The data demonstrates a pronounced curvature seawards from the coping to near midheight in the profiles for the wall adjacent to X1 and X3 with a range in excess of 250mm, and predicted to be of the order of 500mm. In both cases the sheet pile walls slope landwards below sea bed level. He also exhibited a letter from him dated 28 th September 2005 in which he wrote: The magnetic sonde has again been demonstrated to be operating well within its expected tolerance limits in the measurement of distance between the sonde and the sheet pile wall, and additional tests have been carried out to identify the effect of potential variation in results in the region of the toe of the pile. The accuracy and repeatability of the measurement of the system remains at ±50mm. Inclinometer measurements have also been taken on this survey to determine the verticality of the tube in which the magnetic mass sonde was placed: this has demonstrated that the tube above sea bed level is not necessarily vertical, and tends to bend in directions defined by currents within the sea. At high current flows and strong wave action the inclinometer tends to act as an accelerometer, and the data may be contaminated by motion. For the purpose of this survey the contamination on boreholes X1 and X3 is minor and can be ignored for the purpose of defining the general trend of the sheet pile profile to an accuracy of about ±100mm. In the case of borehole X2, the error rises to in excess of ±250mm which is unacceptable for the purpose and the data cannot be used for tilt correction with an adequate degree of confidence. Tilt corrected profiles have therefore been supplied for X1 and X3 only which are accurate to ±100mm, and which show a seaward curvature [sic] in excess of 250mm, and probably of the order of 500mm, at approximately midheight. In both cases the sheet pile walls slope landwards below the sea bed level. The curvature of the sheet pile is the rate of change of direction (in a vertical plane perpendicular to the line of the wall) with depth. It may be measured in radians (circular measure of angle) per metre, expressed simply as per metre. The word curvature in the last paragraph of the above quotation clearly means displacement from the datum. At location X3 the toe of the pile was indeed shown to be vertically below the upper position. The pile bulged seaward in between. At location X1, the toe was about 600mm landward of the upper point. The evidence from location X2 is less clear. Mr. Williamson relied on the information encapsulated in an exhibit (X30, page 2) that he put forward, which shows the toe of the pile slightly seaward, say 50mm, from the upper position. Exhibit X30, page 2 shows an average profile from the magnetic sonde results for location X2. I accept that X30, page 2 is the best estimate of that profile available to the court, but the confidence level as to its accuracy is low. Displacement as a guide to bending moment Regarding the shape of the sheet pile at location X1, Mr. Bracegirdle wrote in a report: It is generally accepted that, under normal piling conditions, the deviation of piles from vertical can be expected to be less than 75:1. This corresponds to an offset of about 350mm over the length of the sheet piles; here the offset is nearly double that figure. If the sheet piles were installed vertically, as maintained by HSS, this would have required the top of the sheet pile to have moved seawards by 0.6m. Given that the top of the sheet piles is restrained by the relieving platform and coping beam and that the structure supports the seaward crane rail beam, such a displacement of 0.6m would have resulted in substantial damage to the quay, preventing the operation of the cranes and the port. It is most likely, therefore, that the offset from vertical occurred during driving of the sheet piles. I take the expression "less than 75:1" to mean less than 1 in 75. Subject to that, I find that passage compelling. No one has suggested that the toe of a pile could move landwards after it had been driven. A witness statement from Mr. Colin Gilbert was put before me. He did not give oral evidence, but his written evidence was unchallenged. He was the senior engineering assistant to the resident engineer for the supervision of the "new" (early 1970s) quay wall works to provide four additional berths, namely Berths 202 to 205. In his statement, he said this: The piles were driven in pairs and due to their length (85 and 87 feet) they were restrained during driving by steel frames specifically designed for that contract. The purpose of the frames was to restrain the piles from deviating from a vertical plane during driving….. The driving procedure was to pitch the piles in the No. 1 frame that was some 35 metres high. For each frame location I recall that there were generally six pairs of piles pitched at one time. A diesel D22 pile hammer was used to drive each pair sequentially until they achieved about half penetration, about 40 feet. At this stage the first rig was moved along to the next bay of piles and the second less tall rig was moved into place to continue the driving. I recall that the second rig drove the piles using a D22 hammer down to approximately six feet short of the final set. Two D44 hammers completed the driving using short robust frames….. Apart from a tendency to move a small distance seaward (in the order of 2") I do not recall that there were any difficulties in maintaining the piles in a vertical plane during driving. The piles were regularly checked for alignment and verticality…..and in Berths 204 and 205 the driving went well. Any tendency to lean would have been apparent when the lateral restraint was removed by the change in driving frames and I do not recall that this occurred. Mr. Reid explained that when the piles were driven in pairs the hammer impacted on the clutch between the two. He gave some evidence of which the following is an extract (day 16, page 112): Well, if you ….. drove one of them on its own not in a panel you would be concerned with pile wander perhaps, because it is [a] flexible piece of steel ….. When you drive two together you increase the stiffness by a factor of about 6 or 8. But is still relatively flexible. But when you drive them in panels…..it is totally stiff….. I haven't used Larssen 6 [4] very much, but in the last contract I used them on we hit boulders…..and it was stiff soils there. And the piles we thought would have deflected when they hit the boulders, but they didn't, they stayed vertical and the ends of the piles were damaged on the boulders….. So it is not impossible I guess to drive them other than vertical, and they won't be perfectly vertical. You can't get such a thing as perfect verticality but they will be very close to vertical in my opinion and….. I can't see how they can wander….. in the frame they are kept vertical….. So if in a normal situation there were soils without obstructions even though they are dense I would expect the piles to go in absolutely vertical. There is of course in this case…..the two inch movement. And in the course of cross-examination he said (day 22, page 40): ….. If a contractor were to drive at 1 in 75 with the sheet pile in panels, it would mean he was incompetent. You just don't do that with sheet piles; because you are driving them in panels, you are in total control. All you do if you have a problem is you don't drive them so far ahead of the rest of the panel. If you were to drive them a small distance ahead, it is almost infinitely rigid. British Standard BS EN 12063:1999 was put to Mr. Reid. That standard says at paragraph 8.6.1: The…..verticality of the sheet piles after installation should be in accordance with the recommended values given in Table 2. This table gives values for normal cases. Table 2 states that the tolerance in verticality measured over the top metre shall be less than or equal to 1 per cent, but may amount to 2 per cent. in difficult soils, provided that no strict criteria regarding for example watertightness are specified and de-clutching is not considered to become a problem after excavation. Paragraph 8.7.1 says: When driving in very hard soil layers, the stiffness and stability of the guide frame should receive special attention in order to keep transverse and longitudinal leaning and horizontal displacements of the driven sheet piles within the tolerances given in 8.6.1. Mr. Reid said this (day 22, page 44) in answer to questions from Mr. Bowdery: Q. So would you agree that in difficult soils this tolerance may increase up to 2 per cent.? A. That's what it says there, yes……But this is a general clause that applies to….. all types of sheet piling….. So these are general clauses that a consultant would take and in almost all cases, certainly in something like this, you would be changing it completely because these soils are not difficult soils and these are the heaviest piles that you can possibly get with the heaviest driving frames I have ever seen….. Q. You say they are not difficult soils in this case. I thought it was 12,000 blows per foot? A. That's right down at the bottom. Difficult soils are ones with obstructions in them, boulders and such like, cobbles, things that knock the pile off line. Up at the top levels here, we are talking of gravels at the worst, which are not difficult. Despite that evidence of Mr. Reid and Mr. Gilbert, the measurement of the position of the toe of the pile at location X1 convinces me that the piles cannot all have been driven vertically. Some may have been driven at a rake. Some may have wandered during the course of driving. It is, however, evident that the toes of adjacent piles could not have been widely separated. I thus cannot accept that the horizontal displacement of the piles is a reliable indicator of the stresses or bending moments to which they have been subjected. Given the existence of the profiles, it seemed to me (and in my judgment it is the case) that an obvious way to check whether the piles contained plastic hinges was to determine the curvatures of the piles from the profiles. The bending moment is a function of curvature, depending on the bending stiffness of the piles. Bending stiffness The bending stiffness is the factor, EI, by which the curvature has to be multiplied to arrive at the bending moment. Professor Pavlovic gave evidence that the bending stiffness of the steel sheet piles was 212,000 kNm²/m [5] and that the bending moment capacity of the sheet piles (i.e. the bending moment where the section becomes fully plastic) was 2500 kNm/m. There was a conflict of evidence about that. Mr. Reid presented calculations that what he called the plastic moment of the sheet piles (that is, the bending moment where the outer fibres reach their elastic limit: I shall use the expression maximum elastic moment) was 1776 kNm (by implication, 1776 kNm/m) and that the bending moment capacity was 2103 kNm/m. It was implicit in those calculations that EI was 226,600 kNm²/m. In his finite element analysis, Mr. Bracegirdle used a bending stiffness of 220,000 kNm²/m. Professor Pavlovic said that the value of 212,000 kNm²/m was calculated by both Haecon and ABP, and that the moment capacity of 2500 kN-m/m was calculated by Haecon. Mr. Reid evidently relied on an undated page from a Larssen catalogue. Both witnesses used a figure of 3.55E5 kN/m² as the yield stress (elastic limit) of the relevant steel. (In the expression 3.55E5 the letter E refers to the exponent: it is not to be confused with the element E in the expression EI. The notation 3.55E5 means 3.55 X 10 i.e. 3.55 X 100,000; 6.5E-3 means 6.5 X [10 raised to the power -3], i.e. 6.5 divided by 1000, i.e. 0.0065). The documents relied on by Professor Pavlovic were not identified, but I have researched the files of papers put before me. The figure of 212,000 (2.12E5) for EI appears on page 18 in section 5.2.2 of a document emanating from Haecon which is an as-built design report identified by the code USR2334 637 and marked revision 2 22-08-03. On the same page the figure of 2537 kNm/m appears as the bending moment capacity of the wall. Although I have not had the benefit of oral evidence or submissions on this point it appears that the figure of 2.12E5 kNm²/m is erroneous. The area A per unit length is stated as 4.2488E-2 m²/m, and EA is stated to be 8.92E6 kN/m (i.e. kN/m² times m²/m). From that, E can readily be calculated as 210E6 kN/m². I is given as 1.00952E-3 m . Multiplying that by E does indeed give a figure of EI equal to 2.12E5. But the units must be kNm². The units in which EI is stated are kNm²/m. It is stated on page 9 in section 5.1.1 of the same report that I is 1.00952E-3 m for the combined section, i.e. an inpan and an outpan. It is also there stated (and corroborated by Mr. Reid's page from the Larssen catalogue) that the length of a combined section is 0.84m. Thus I, expressed in m /m, would be the figure stated divided by 0.84, i.e. 1.20E-3. EI, expressed as bending stiffness per metre, is thus 2.52E5 kNm²/m. Those are indeed the units in which EI is wrongly expressed to be 2.12E5 on page 18 of the report. The figure of 2537 kNm/m for the bending moment capacity of the sheet piles is stated to allow for a safety factor. Given the yield stress of 3.55E5 kN/m², also expressed on the same page, 18, of the report, and information given as to the axial force, it is possible readily to calculate that the safety factor is 1.1. Thus the bending moment capacity implied is 2537 kNm/m multiplied by 1.1. That bending moment capacity is arrived at by multiplying the sectional area by the yield stress and by an undefined factor 2a which is evidently 0.185m. (The whole is then divided by the safety factor of 1.1 to arrive at the figure of 2537 kNm/m). To arrive at the bending moment capacity, that is the bending moment which causes (in theory) all the fibres at the relevant section to be subjected to their yield stress, it is necessary to integrate the elementary moments of the forces in the fibres about the neutral axis of the curved sheet pile. The elementary forces are the product of the yield stress (force per unit area) and the area of the element in question. The lever arm is the distance of the elementary area from the neutral axis. Thus the dimensions of the elementary moment (and hence of the integrated moment) are stress times area times distance. Hence the factor of 2a measured in metres. I can find no source which explains how the figure of 0.185m was arrived at. The page from the Larssen brochure gives data from which the factor 2a can be calculated. The figure is the ratio of the plastic section modulus to the sectional area. Both those figures are given on that page of the brochure. The plastic section modulus is stated to be 5.924E-3m³/m; the sectional area is stated to be 4.19E-2 m²/m (I have converted the units). The ratio is 0.141m. If that figure is substituted for 2a, the bending moment capacity comes out at 2127 kNm/m without a safety factor. E is an elasticity modulus which applies in the elastic regime, where stress (force, here longitudinal force, per unit area) is proportional to strain (here, fractional increase in length). Where that regime applies, the ratio is a constant E known as Young's modulus. The value 210E6 kN/m² for E is implicit in the section of the Haecon as-built design report to which I have referred above. The same value is explicit in an earlier design report of Haecon, identified by the reference USR2334 31 and dated 08-11-01, where it appears at page 14. Mr. Reid has taken the value 206E6. For the foregoing reasons, I cannot accept Professor Pavlovic's figures of 2.12E5 kNm²/m for EI or his figure of 2500 kNm/m for the bending moment capacity of the sheet piles. I accept the figures used by Mr. Reid. There are, it is true, minor discrepancies in the parameters A, I and W (see below) between the page of the Larssen brochure that he relied on and the information contained in an undated Larssen Steel Sheet Piling Handbook which was exhibited to the witness statement of a witness who was not called to give evidence before me, and between those sources and the Haecon calculations. In fact, the differences between Mr. Reid and the Haecon reports as to the values of E, I and EI make little difference to the outcome that is relevant for present purposes, that is the curvatures necessary to produce the maximum elastic moment and the moment capacity of the sheet piles. I set out in the table below the figures, implicit or explicit, in the various sources. The parameter W is the so-called section factor or elastic section modulus. The last line of the table gives the curvature required in each case to give the maximum elastic bending moment. Parameter Reid Report 31 Report 637 637 corrected E 206E6 210E6 210E6 210E6 I 1.1E-3 1.12E-3 1.00952E-3 1.2018E-3 EI 2.266E5 2.352E5 2.12E5 2.52E5 W 5.005E-3 5.1E-3 I/y=5.46E-3 I/y=4.59E-3 y=I/W 0.2198 0.2196 Say 0.220 Say 0.220 (fy) 3.55E5 3.55E5 Say 3.55E5 Say 3.55E5 Me=I(fy)/y 1811 1629 1/Re=Me/EI 7.83E-3 7.70E-3 7.68E-3 7.68E-3 Since I have not accepted the value of 2500 kNm/m as the bending moment capacity of the piles, the only acceptable evidence of the ratio between the maximum bending capacity and the maximum elastic moment is that contained in Mr. Reid's extract from the Larssen brochure. That ratio is 5924/5005. It leads to the value of 2103 kNm/m for the maximum bending capacity on Mr. Reid's figures, which I have accepted. A figure of 2100 kNm/m was spoken to by Mr. Bracegirdle. If the moment remained proportional to the curvature beyond the value of 1776 kNm/m, the constant of proportionality being EI, the curvatures which give rise to the capacity bending moment would be respectively 9.27E-3, 9.12E-3, 9.09E-3 and 9.09E-3. The Haecon calculations which I have referred to for the purpose of resolving this conflict of evidence between Professor Pavlovic and Mr. Reid were not put to any witness nor were they the subject of submissions. I disregard them when I come to consider the fitness of the design. The bending moment that mobilizes the elastic limit of the steel is 1776 kNm/m. That causes a curvature of 7.83E-3 per metre. At that bending moment, the outer fibres of the steel are stretched to their elastic limit. They can be stretched further, but offer no increase in resistance. At that point, a plastic hinge starts to form. The hinge is fully plastic at a bending moment of 2103 kNm/m. The moment increases less than proportionally to the curvature as the plastic hinge forms. Thus the curvature of 9.27E-3 given in paragraph 43 above must represent a lower limit. Once a plastic hinge is fully formed over a significant length of the sheet pile, 2 metres or so, the pile, if unrestrained at its sides or adjacent to piles in similar condition, can be pushed out further with little increase in the force acting on it. The same applies to the whole wall, with adjacent piles interlocked, if similar forces are acting along all of it. Curvature I invited the experts through counsel to draw smooth curves representing the best fit of a smooth curve to the profiles. From those curves one can determine the curvature. A smooth curve must be drawn since there will have been random errors in the measurements. The sheet piles could not have been in a shape other than a smooth curve: both Professor Pavlovic and Mr. Reid gave evidence to that effect. A profile of a pile shows a vertical section of the wall, perpendicular to the line of the wall. It expresses the relationship between y and x, where y is the horizontal position of the pile in question, measured at right angles to the line of the wall, at any given depth x. Mr. Reid and Professor Pavlovic responded to that invitation. Both found the smooth curves by computer programs which found the best fit of a profile, subject to the constraint that y was a polynomial function of x of a given degree. The degree of a polynomial is the highest power of x that it contains. Broadly speaking, the higher the degree the more closely can the curve follow the vagaries of the points. But following the points too closely will give a false picture since the shape of the piles is smooth but there are random errors in the positions of the points. Mr. Reid calculated a quartic polynomial as a best fit for the profile of X3. Professor Pavlovic calculated a number of polynomials. He calculated polynomials of all degrees from the third to the ninth for each of X1, X2 and X3. Both gentlemen specified the polynomials in question by reference to their coefficients of the powers of x. Professor Pavlovic also calculated polynomials to fit data presented by Mr. Reid in his report. Those data related to six other piles in berth 204 which had not been examined by the magnetic sonde but had been measured by divers on behalf of HSS. Mr. Reid had drawn profiles of them above the level of the sea bed. Professor Pavlovic used the polynomial curve-fitting procedure in relation to four of those six profiles. He did not specify the relevant polynomials. He concluded that away from the ends of the range of measurements the highest curvature shown by any of the profiles was 6.5E-3. Both Professor Pavlovic and Mr. Reid entered serious caveats against the use of polynomials to determine curvature. Professor Pavlovic said this with reference to the sonde data and the divers' measurements: …..analysing such data in the expectation of obtaining accurate estimates of the level of the shear forces and bending moments acting on the sheet piles is questionable and may lead to misleading conclusions. The main reason for this is that these measurements are sparse (since they were taken every metre along the sheet pile's height) and may contain significant errors. These errors ….. result in unreliable values of the second derivatives, i.e. curvatures and hence bending moments, as the latter are extremely sensitive to even small departures from true local values of displacement. In plotting the above data on a graph…..one can see that the resulting line which represents the profile (deformed shape) of the steel sheet pile is not smooth, as one would expect in an actual sheet pile, but is full of local irregularities. Such local irregularities are the results of errors that do not describe the actual deformed shape of the sheet pile (in fact, it would be physically impossible for the sheet pile to deform in such a discontinuous way). These irregularities affect the shape of the fitted curves used to describe the profile of the deformed shape of the steel sheet pile, especially their second derivatives, thus compromising the accuracy and validity of the analysis of the most relevant parameters, namely the bending moment distribution which is directly proportional to the curvatures, i.e. second derivatives. The use of more complex equations (higher order polynomial expressions) will neither reduce the above errors nor will it improve the accuracy of the curvatures because of the coarseness of the measurements along the height of the steel sheet pile. Mr. Reid delivered a substantially concurring opinion. But the present exercise is not to calculate the correct value of the bending moment at every or any given point, but to investigate the likelihood of there being any plastic hinges in the sheet piles in question. Professor Pavlovic made a second point. He said this: …..towards the end of the pile each polynomial departs from the measured shape because there are no smoothening data beyond this region available for extrapolation. Therefore, such polynomials (and any other curve fitting approach) can only be reliable in the middle region of the pile….. Those objections involve two distinct matters. To take Professor Pavlovic's second point first, there is the error in curvature which best-fit polynomials are prone to exhibit towards the ends of the ranges that they are designed to cover. I accept that that error exists, and that the curvature figures generated for the points towards the ends of the respective ranges are useless. Second, there is the proposition that differentiation of a function, especially double differentiation, is liable to produce wholly unreliable values of curvature, whatever the function, polynomial or other. I shall consider that point further. The essence of the argument is this. Suppose there is a relationship between two physical quantities that can be represented graphically by a smooth curve. Suppose measurements are taken of those quantities in a real experiment or observation. Suppose the measurements are subject to random errors and are plotted as a graph. A "best fit" curve will be a smooth curve showing the trend line, but a curve that passes through all the points will be a squiggly line. It is of the nature of squiggly lines that their curvature rapidly varies from point to point on the line. The values of the curvature will change between high positive values and high negative values, necessarily passing through zero in between. Thus if one measures, or calculates, the curvature at any point it is likely to have a value far removed from the average value in the vicinity of that point, i.e. from the value of the curvature of the trend line in that vicinity. If the squiggly line is represented by a mathematical function, polynomial or other, the curvature can be calculated by differentiation and double differentiation of that function. The answer will be correct, but useless, since it not the curvature of the squiggly line, but that of the trend line, that is desired. The polynomials that have been generated by Professor Pavlovic are (I accept) the best-fit polynomials of the given order in each case that can be found to fit the data, at any rate to as high a degree of accuracy as is necessary for present purposes. They thus represent the smooth curve, not the squiggly line. That the objection to double differentiation is in most cases negligible for present purposes is clear from the actual results. In appendices 1, 2 and 3 to this judgment I set out figures that I have calculated from Professor Pavlovic's polynomials relating to the sonde measurements using boreholes X1, X2 and X3. I have calculated both the y (horizontal position) value for each given value of x (the vertical position) and the second differential of y with respect to x for each given value of x. I have also shown the measured y value for each value of x. Thus it is possible to see how far the smoothed-out curves depart in their y values from the measured values. The second differential of y with respect to x gives a sufficiently accurate value of the curvature of the polynomial function, given that the gradient (the first differential of y with respect to x) is everywhere small. I have not calculated results for every polynomial. But the consistency of these results along each curve is proof enough that the objection to double differentiation does not apply for present purposes. If it did, one would expect rapid oscillations in the values of curvature as the x value gradually changed. However, there remains the point mentioned by Professor Pavlovic that curvatures are extremely sensitive to even small errors in the local values of displacement (i.e. y). I accept that important caveat. The higher polynomials, the seventh to ninth, show increasingly more oscillation than the lower polynomials in the magnitudes and signs of the curvatures. It is the higher degree polynomials that more closely follow the vagaries of the individual readings. The quintic polynomial representing the profile derived from the X2 borehole shows a curvature of 8.36E-3 in the region of x equals -2.25mCD to -5.25mCD. Mr. Ballard rejected the results from borehole X2 since they were highly erratic as a result of stormy weather, which interfered with the measurement of the y-values. Mr. Williamson submitted that an averaging of the various results obtained from borehole X2 gave a good match to the netsurvey results (see paragraph 69 below) for the same sheet pile. Thus, he submitted, they were reliable. There was in my judgment sufficient force in that submission for me to have thought it worth while to calculate, as I have, the curvatures represented by that average. But the profile is far from a smooth curve, and I cannot regard the resulting curvatures as calculated from the polynomial of orders 5 and above as reliable in that case. Further, there is a kink in the y value from borehole X2 at about -15mCD which clearly cannot represent reality (nor has it been suggested that it does) and which is reflected increasingly in the higher polynomials. Apart from the exceptions that I have mentioned above, no curvature greater than 5.42E-3 appears at any point between the relieving platform level (0mCD) and minus 16.5mCD for any of the polynomials up to those of the sixth degree that I have used to calculate the curvature of the sheet piles measured by means of the sonde. Mr. Reid, in a document handed up to me, made a number of criticisms of the method of calculation of curvature by curve-fitting. I consider three of them here. I set them out below together with my comments: 1. Curve fitting is very sensitive to the quality of the data. In particular, if a deformation value is inaccurate at any one location, this changes the bending moment throughout the whole of the member, irrespective of the local curvature at locations remote from that data. I accept that curve fitting is sensitive to the quality of the data. I shall return to that point in relation to borehole X3. The object of drawing a smooth curve is to minimize the effect of random errors in the data. 2. If for a given set of deformation data, a curve is fitted on the basis of the zero x ordinate being at one end of the beam and the bending moment diagram derived by double differentiation, this is different from that derived, from the same data, if the beam is reversed and the zero x ordinate taken at the other end. Clearly this should not be the case. I accept the last sentence. If the first sentence is correct, I can only suppose that the algorithm used by Mr. Reid produces two different "best-fit" curves in the two cases. For a given curve, changing the variable from x to x´=k-x cannot produce the effect contended for. 3. To be credible, a bending moment diagram deduced from deflection data should be compatible with that derived from a calculation where the beam is subjected to the appropriate from of loading. To be given credibility the bending moment for Berth 204 should have a zero value at both capping beam and its base. In addition, the maximum value should occur, either at tie level or at a level of about 8 metres beneath tie level. The first sentence is true provided that the calculation is correct. But the sheet pile must be the best analogue computer of the force field to which it is subjected. Mr. Reid himself made that point. He said (day 22, p.119): …..the combined influence of everything can be seen by the shape of the piles, so you can deduce from the shape and curvature of the piles….. Regarding the second sentence, as I have indicated in paragraph 50 above the curvatures calculated from polynomials for positions towards the end of the range of values of x cannot be relied on. In the same document handed up to me, Mr. Reid expressed the view that curve fitting to a fourth degree polynomial is probably preferable to curve fitting to higher degree polynomials. I have mentioned [6] Professor Pavlovic's view that higher order polynomials will not improve the accuracy of the curvatures. I accept both of those views. The question arises why Professor Pavlovic and Mr. Reid arrive at different values of the curvature represented by the profile of X3. Mr. Reid's quartic polynomial shows a maximum curvature of 7.36E-3/m. Professor Pavlovic's quartic shows a maximum curvature, ignoring the five x-values at each end of the range, of 5.38E-3/m. Professor Pavlovic has put forward reasons for that discrepancy. But in my judgment there is another reason, which has not been mentioned. Professor Pavlovic derived his data from a document identified as X22. That was put forward by Mr. Williamson as a representation of the profile obtained by the sonde. The original profile obtained by the sonde was exhibited to Mr. Ballard's witness statement. It was in the form of a graph of horizontal location y against vertical position x. Thus it showed the shape of the profile, modified by the difference in scale between the horizontal and the vertical axes. The scales were small, and the plot was not on graph paper. Thus it was impossible to read off the precise data from the graph. That was no doubt a perfectly reasonable way to represent the data, given the wide margin of error to which Mr. Ballard spoke. The horizontal scale of X22 was even smaller, and it was even more difficult to read off the horizontal positions accurately. When I read off the data in the two cases, I obtained different results. Mr. Reid's quartic and Professor Pavlovic's quartic do indeed yield different y values for the same x values. I have carried out a least-squares exercise, and I find that Professor Pavlovic's quartic represents the y values as I have read them off X22 better than Mr. Reid's quartic represents the data as I have read them off Mr. Ballard's graph. But in my judgment, if there is a discrepancy between the two sources, Mr. Ballard's is to be preferred to X22, being the original source and easier to read. Mr. Reid's quartic better represents my reading of Mr. Ballard's profile X3 than does Professor Pavlovic's. I have also considered the distance of the bulge from the chord line, where that distance is a maximum. I have taken the chord line from x=0mCD (not the precise position of the relieving platform) to the toe of the pile, as shown on the profiles. The distances, as I have measured them, are 283mm for Mr. Ballard's profile, and 271mm from X22. The corresponding value taken from the curve produced by Mr. Reid's quartic is 279mm from the chord of that curve, that for Professor Pavlovic's quartic is 255½mm from the chord of Professor Pavlovic's quartic. Thus Mr. Reid's quartic appears to be the more accurate. The foregoing discrepancies between the two quartics illustrate the sensitivity of the curvature to the data. The data in this case are not claimed to be accurate. Mr. Reid's quartic intended to match the profile of the wall as measured from borehole X3 showed a maximum curvature of 7.36E-3, correct to 3 significant figures. That was at x=15.1m. Mr Reid calculated a bending moment of 2072 kNm (per metre run) at x=15m, where the curvature calculated to 3 significant figures was the same. The value of EI implicit in the maximum elastic moment capacity of 1776 kNm/m which I have accepted on the evidence of Mr. Reid is 226,600 kNm²/m. The calculated curvature of 7.36E-3 then implies a bending moment of 1668 kNm/m. At the same time as he adduced the figure of 2072 kNm/m, which he did by way of correction to a previous calculation of his, Mr. Reid abjured the use of polynomials to determine curvature, as indicated above. Mr. Reid went on to criticize a response that Professor Pavlovic had given to the same question from me. He said (as I have accepted [7] ) that Professor Pavlovic had quoted too high a value, 2500 kNm/m, for the bending moment capacity of the steel sheet piles. Mr. Reid said this: This is inaccurate, and the correct figures, allowing for 10% loss of steel due to corrosion, are Limit of elastic flexural capacity 1600 kN-m/m Limit of plastic flexural capacity 1900 kN-m/m Clearly the correct figures are a significant reduction from the 2500 kN-m/m used…..to draw conclusions that a large margin of adequacy is available. Mr. Reid's allowance for corrosion misses the point. Mr. Reid's reduction of 10 per cent. for corrosion implies a corresponding reduction in EI. In that case, the bending moment implied by a given curvature is reduced correspondingly. As I have said above [8] , I accept Mr. Reid's figure of 1776 kNm/m as the limit of elastic flexural capacity of the steel. In Mr. Reid's calculation of that figure there is implicit the value of 226,600 kNm²/m for the bending stiffness EI, and he correctly states the radius of curvature as 128m, which is equivalent to the curvature of 7.83E-3 per metre mentioned above [9] . The sheet piles located at X1, X2 and X3 constitute a very small sample of the sheet piles composing berth 204. A survey of the sheet piles of berth 204 was carried out in July 2005. As Mr. Reid explained, the survey was conducted under water using a spirit level equipped with a device to measure offsets from the vertical over one metre lengths. Mr. Reid concluded: This survey confirms that there is significant and variable curvature in the face of the sheet piles at Berth 204. As mentioned in paragraph 47 above, Mr. Reid appended to his report of September 2005 a copy of output from that survey. I have examined all his plots of the output, six in number. They show "deflection" (horizontal position) against level. At location 20, readings were taken from +1mCD to -12.8mCD (sea bed level). There was a bulge with a maximum at -8mCD. The displacement between relieving platform level (0mCD) and the maximum of the bulge was about 95mm. I have made a rough calculation of the average curvature in the region where it appears greatest by measuring the gradient of the plotted curve at two points and calculating the average rate of change of the gradient per metre change in level. The curvature as so calculated is less than 5E-3 per metre. At location 25, where readings go down to -11.72mCD, there is no bulge. The maximum displacement is 12mm and the curvature obviously negligible. At location 160, the graph goes down to -8.37mCD. There is the top part of what may be a bulge at -7 to -8mCD. The maximum displacement is 160mm. The curvature, as calculated by me, is 7E-3. At location 185, where readings go down to the level -9.29mCD, there is a bulge at -5mCD. The maximum displacement is about 65mm. My calculation gives a curvature of 5.5E-3. The plot of location 215 goes down to the level -8.68mCD. No bulge is reached by that level. The maximum displacement shown is 260mm. The curvature is clearly negligible. Finally, location 310 goes down to -11.42mCD. The maximum displacement shown (though apparently increasing with depth at -11.42mCD) is 425mm. The maximum curvature below the relieving platform is negligible. Location 160 is close to location X2. Location 310 is close to location X1. I have mentioned the Netsurvey above. Two such surveys were carried out, one in 2004 and one in 2005. I shall confine my attention to the former, since there is no dispute as to the reliability of its results, but the latter is controversial. The Netsurvey was a multibeam system of measurement of distance from a boat, using a global positioning satellite system and with corrections for movement and angle of the boat. Netsurvey 2004 In the Netsurvey, points were plotted which showed the profiles of the piles above seabed level at 5 metre intervals along the wall in several berths including berths 204 and 205. I have looked at all the profiles covering berths 204 and 205. Berth 204 extends from bollard 216A to a point close to bollard 236; berth 205 extends from that point to bollard 258. Bollard 258 is the zero datum of the survey for measurement along the quay in metres. Bollard 236 is close to the 330m point; bollard 216A is close to the 630m point. There is a striking difference between the profiles for berth 204 and those for berth 205. That is scarcely surprising, since, as is common ground, the piles in berth 205 have been pushed out by the grouting works. The curvature and displacement of the piles is apparent, with the maximum displacement generally at about the level of -9 to -10mCD. The profiles of berth 204, by contrast, are comparatively straight and show less displacement. It is not possible accurately to tell the curvature of the piles from these small-scale profiles. But the curvatures of the vast majority appear to be negligible for present purposes. Those which appeared to me to embody the greatest curvatures are identified in the table below. Chainage (m) Horizontal displacement (mm) Curvature (per metre) near bulge Curvature (/m) near relieving platform 450 at -8½mCD 5E-3 370 at -11mCD 0.02 at -1mCD 200 at -6½mCD 5E-3 150 at -6½mCD 7E-3 at -1½mCD 200 at -11mCD 8E-3 at -1½mCD 300 at -10½mCD 0.015 at -2mCD 350 at -11mCD 8E-3 0.016 at -2mCD 250 at -10mCD 3E-3 500 at -10mCD 3E-3 600 at -11½mCD 5E-3 350 at -10mCD 5E-3 The profiles which appeared to me to embody the greatest curvatures near the maximum of the bulge were at chainages (metrages) 335, 345, 375, 515, 525, 530 and 570. The respective horizontal displacements from the positions at 0mCD are also set out in the table. I made a rough estimate of the maximum average curvatures over a few metres' depth in each of those cases by measuring and calculating the change in direction of the profile in radians per metre depth. The curvatures I obtained were respectively 5E-3, 5E-3, 8E-3, 3E-3, 3E-3, 5E-3 and 5E-3 per metre. Only the third of those represents a bending moment in excess of the maximum elastic moment. The horizontal displacements could not be measured with any precision; the stated curvatures are even less reliable. Nevertheless, it is clear that in the vast majority of these numerous profiles no curvature at the bulge and above the sea bed is shown which approaches that representing the maximum elastic bending moment, let alone a plastic hinge. The profiles which appeared to me to embody the greatest curvatures near the relieving platform were at chainages 340, 350, 355, 360 and 375. The curvatures at chainages 340, 360 and 375 suggest the presence of plastic hinges. But the Netsurvey readings in the vicinity of the relieving platform may be affected, as pointed out by Professor Pavlovic, by the presence of low water corrosion plates attached to the sheet piles by way of repair or reinforcement in that vicinity. Corrosion had to be particularly guarded against in the intertidal zone, which extended down to -0.6mCD. If such plates appear in the Netsurvey profiles, they will have the effect of increasing the apparent curvature of the sheet piles near the relieving platform. Thus I cannot regard the curvatures that I have estimated at that level as reliable. With the single doubtful exception of metrage 615, which possibly shows a small landward offset, all the profiles that show non-zero offsets show seaward offsets. Several piles which appear to be straight have substantial offsets. The following are examples. I give the offset at the lowest point recorded: 390m, offset 380mm at -11mCD; 435m, 500mm at -10mCD; 485m, 250mm at -11mCD; 500m, 350mm at -10½mCD; 595m, 580mm at -12¾mCD. In my study of the Netsurvey of berth 204, which I made without reference to the other evidence, I found seven locations (and only seven) out of the total of about 60 where the visible curvature in the vicinity of the bulge seemed to be significant. One of those, at metrage 375, turned out to be close to X1 (which was at about 380m) and to Mr. Reid's location 310. Another, at 515m, was close to X2 (which was at about 507m) and to location 160. A third, at 570m, was close to location X3 (which was at about 573m). Thus, whether by chance or by design, X1, X2 and X3 were located at or near points of exceptionally high visible curvature. The profile that I have noted in paragraph 69 above at metrage 530 is, I think, the one that appears as having a 650mm displacement, the maximum shown, in figure 4 of Mr. Reid's report. The location mentioned by Mr. Reid as location 185 is situated at about 475m on the Netsurvey metrage. The curvature shown on the Netsurvey in that vicinity is negligible. The profile that Mr. Reid exhibited in his appendix is anomalous in that the maximum is shown at -5mCD. I regard that as implausible. I accept the Netsurvey result. State of 204 Mr. Williamson relied on a finite element analysis produced by Mr. Bracegirdle which showed a bending moment of 600 kNm/m at -10mCD corresponding to a bulge of 100mm at that point. He invited me to conclude that X3, showing a bulge of 300mm, must have been sustaining a bending moment of 1800 kNm/m, a moment in excess of the maximum elastic moment. The sonde profile shows a deviation from the chord of 283mm. If the proportionality held exactly, and the figure of 600 kNm/m were exactly accurate, then the bending moment at the maximum of the bulge on X3 would be 2.83 times 600 kNm/m, i.e. 1698 kNm/m. These figures are not, of course, at all as accurate as their apparent precision. But Mr. Bracegirdle's finite element analysis has the pile almost vertical at 0mCD, and its tangent rotates outward before it starts to rotate inward as one runs one's eye down towards the maximum of the bulge. That may produce a different curvature, and a different calculated bending moment, at the maximum of the bulge, than would be the case for the shape of X3 where (as in the case of X2) the pile is raking seaward even at relieving platform level. It is nevertheless noteworthy that the bending moment of 1698kNm/m is close to the figure of 1668kNm/m arrived at in paragraph 62 above. The above considerations in my judgment tend to confirm rather than otherwise the usefulness of estimating curvature to determine bending moment. Both methods are affected by errors in the original profile measurements. Apart from that, inaccuracies in determining the curvature by using polynomials arise principally, in my judgment, not at all from the process of differentiation but from non-conformance of the polynomials with the smooth curve represented by the true profile. But to extrapolate from the displacements shown in three instances, X1, X2 and X3, (which themselves show that the piles have not in all cases been driven vertically) to the conclusion that the bending moment is in all cases proportional to the displacement of the pile from the vertical passing through the pile at the relieving platform level is in my judgment wholly unjustified. Mr. Bracegirdle said in his report of September 2005, at paragraph 11.11.14, that finite element analyses of the sheet pile wall in berth 204 suggest that the maximum displacements occur at about -10mCD; this, he said, is about the mid-point of the span below the relieving platform and the logical position for maximum displacement. It is common ground that the bending moment at the toe of a sheet pile will be zero. Mr. Reid expressed the view mentioned in his proposition 3 [10] that the maximum curvature will be either at the tie level (i.e., at or near the relieving platform level) or at a depth of about -8mCD. I am thus satisfied that the maximum curvature will generally occur above sea bed level. Thus the points of maximum curvature below the tie level should have been visible in the generality of the Netsurvey results. The nearest candidate to represent X1 that I can find from the Netsurvey is the profile taken at 385m. That shows a maximum offset of 100mm at -7.5mCD and an offset of 50mm at -11.5mCD. If the toe really was at 600mm landward, as measured by the sonde, that would, I think, require a curvature below the sea bed (and thus not visible to the Netsurvey) involving a bending moment in excess of the maximum elastic bending moment. However, the sonde produced no reliable readings for the levels between -4mCD and -14mCD. That fact, coupled with the stated inaccuracy in its measurements, leads me to a state of uncertainty as to the degree of curvature of X1 below the sea bed. If the evidence of curvature stood alone, I should reach the following conclusions. In the case of locations X1, X2 and X3 and their immediate vicinities, I would not be satisfied that there were any bending moments as great as the maximum elastic bending moment, let alone any plastic hinges. Except in the cases of X1, X2 and X3 and their immediate vicinities, I would conclude that the curvature in the sheet piles of berth 204 below the tie level nowhere exceeded about 5E-3/m, thus yielding a factor of safety of about 1.5 in relation to the maximum elastic moment, and of about 1.9 in relation to the formation of plastic hinges. I would conclude that in most places, the factors of safety were substantially greater. Rake of sheet piles When the excavation for the relieving platform was carried out, the pressure of the ground and water on the seaward side of the sheet piles above the level of the relieving platform caused the sheet piles to tilt above that level so that the top of the sheet piles moved landward. Mr. Bracegirdle and Mr. Reid were agreed about that. Mr. Reid expressed the opinion that the sheet piles had been driven vertically, but that when the land on the seaward side of the sheet piles was excavated to -12.8mCD, the piles bulged outwards with a maximum bulge at about -10mCD by reason of the bending moment caused by the tilt above the relieving platform. That movement would reduce the bending moment that had caused it. No calculations were proffered in support of Mr. Reid's theory, but he made a model which graphically illustrated it. It involved rotation of the sheet pile wall outwards "through" the passive soil below -12.8mCD (i.e. moving the passive soil outwards and compressing it) during the course of the excavation to -12.8mCD, with the toe of the pile stationary. Mr. Reid's theory may have some element of truth in it, but I cannot accept it in its entirety, for reasons that follow. Mr. Bracegirdle expounded a theory that because of seaward movement of the heads of the piles while they were being driven, the piles would be driven with a seaward rake, i.e. so that the toe was seaward of the head. It is not controversial that such seaward movement of the heads of the piles occurred. So far as I understand it, Mr. Bracegirdle's theory works as follows. The seaward movement of the top of the pile in the landfill above the Valley gravels at an early stage of driving induced curvature in the pile since the bottom part of the pile was more or less vertical in the Valley gravels. The curvature in the sheet piles was maintained as driving continued, so that the toe of the pile described an arc as it descended, and the upper parts of the pile followed in its track. Thus when the pile was driven it bulged landward. Subsequent excavation of the landfill to construct the relieving platform caused the head of the pile to move landwards, as mentioned above. That more or less relieved the bulge and left the pile more or less straight but raked, with its toe seaward of its head. That appears to me to require lateral movement (rotation) of the pile "through" the Valley gravels at that stage, and to a lesser extent "through" the Selsey sands. When the berth was dredged to -12.8mCD the sheet piles bulged outwards, but retained their general seaward rake. I cannot accept (since I can see no reason for it) that a pile retains its curvature while driven. There is no suggestion that it does not remain elastic: its curvature at any given time depends upon the stresses to which it is subjected at that time. Thus there seems to be no reason why the toe of the pile while being driven should carve out for itself a path with a curvature similar to that of the part of the path which it has already traversed. I prefer Mr. Reid's theory to that of Mr. Bracegirdle. But the actual measurements convince me that neither represents the complete truth. I am satisfied on the evidence of X1 that the piles were not all driven vertically. And the seaward offset of the piles is not always explicable by a bulge. Piles that are straight as far down as the Netsurvey could detect their position show an offset in several cases. Only one shows a landward offset. Thus there was a tendency for the piles to be offset seaward regardless of bulge. Soil pressure In soil mechanics, parameters used in calculations are frequently represented by symbols which consist of a letter, say c, (sometimes primed, say c´) with a following subscript (say u). In this judgment, I shall write such symbols as a pair of letters in brackets, say (cu) or (c ´ u). Mr. Bracegirdle caused to be carried out a finite element analysis designed to show, among other things, the effect of soil pressure on the SSP wall. The validity of that analysis was challenged by reference in particular to a learned article by D. M. Potts entitled Numerical Analysis: a virtual dream or practical reality?: (2003) Géotechnique , No. 6, 535-573. The article pointed out (p.570) that a significant amount of human error is involved, and that implied that the users had insufficient training and/or experience in the use of advanced numerical analysis. It was not put to Mr. Bracegirdle that the persons carrying out the numerical analysis he used had insufficient training or experience. The program he used was called ICFEP: the Imperial College Finite Element Programme. That program had been developed by the same Professor Potts. Mr. Bracegirdle had derived advice and assistance, particularly in relation to small strain analysis, from Professor Richard Jardine of Imperial College. Professor Jardine had developed the non-linear soil model and had applied it to soil mechanics problems for a period in excess of 20 years. Variations in the model and soil parameters had been applied to assess the sensitivity of the model to wall friction and dilation. It was found that the outcome of the model was insensitive to those parameters. More to the point was the question of the validity of some of the other data used in the program. Mr. Bracegirdle's use of a value of 35° for f´ was criticized. Mr. O'Brien said, and I accept, that the model did not allow for variations in the geology along the wall. The analysis modelled only the locations where there was a deepened channel of valley gravel. Elsewhere, the Marsh Farm formation was higher. The model was a 2-dimensional model for a 3-dimensional situation. Mr. O'Brien considered that the assumptions and idealizations made for the ICFEP analysis used by Mr. Bracegirdle meant that the result was of little value in identifying the main issues: there was insufficiently detailed knowledge of the relevant soil parameters to give a reliable outcome. There were other criticisms, e.g. that the grid used was not fine enough. Mr. Reid made a specific criticism of the finite element program by reference to a graph of shear force in the grout wall as a function of depth (run 200) appended to Mr. Bracegirdle's report. Mr. Reid said that it was self-inconsistent. He said that by differentiating the function produced by the program as representing the shear force as a function of depth one would obtain the (horizontal) loading. If that were done, it would appear that the loading below –23m CD was in the same direction as the loading between –7 and –13m CD. He said that that showed an inconsistency. He pointed out (correctly) that Mr. Bracegirdle had said that the force below -23mCD was in a direction from seaward to landward, whereas the loading from –7 to –13m CD must be from landward to seaward. Mr. Bracegirdle in reply stated that above about –14m CD locked-in stresses on the seaward side of the grout column exceeded the stresses on the landward side. It was below that level that the stresses on the landward side exceeded those on the seaward side. Thus the stresses which Mr. Reid said were wrongly shown to be in the same direction were indeed in the same direction, from seaward to landward. Mr. Reid also made the criticism that differentiating the function produced by the program as representing the bending moment as a function of depth would produce a shear force function which differed from that produced by the program. Mr. Bracegirdle pointed out that since the grout wall was subject both to compression and to bending, differentiation of the bending moment plot to derive shear forces was incorrect and would produce erroneous results. I reject both of Mr. Reid's criticisms of the finite element analysis. Mr. Williamson submitted that Mr. Bracegirdle's explanation was untenable for three reasons. His first was this: When dredging takes place from a level of -12.8mCD to a level of -16.5mCD the sheet pile and grout walls will both inevitably deform from landwards to seawards. This will generate strains in front of the wall that will mobilize passive pressures. These will act beneath a level of -16.5mCD in a seaward to landward direction to act on both the sheet piles and the grout columns. While there may be some doubt as to the pressure distribution between the sheet piles and grout columns above a level of -20.7mCD, there can be no such doubt beneath a level of -20.7mCD where only the grout column exists. I shall consider that passage proposition by proposition. (1) The first sentence I accept. (2) "This will generate strains in front of the wall". The strains are the movement, and the movement of the walls will generate strains in the soil. (3) "Strains.....that will mobilize passive pressures". There will already be passive pressures. The strains will tend to counteract the reduction in passive pressures caused by the dredging. (4) "These will act beneath a level of -16.5mCD in a seaward to landward direction". The passive pressures will act on the grout wall in a seaward to landward direction. What is "mobilized", however, an unfortunate word, is a reduction in such pressures. If considered separately as a pressure on its own, it will act on the grout wall from landward to seaward. This is not, of course, a sensible way of looking at the matter. The pressure in the soil between the grout wall and the sheet pile wall will actually operate in an opposite direction on the landward face of the sheet pile wall to that in which it will operate on the seaward face of the grout wall. But I do not doubt that Mr. Williamson was referring to the seaward faces of both walls. Mr. Williamson's second point was this: To suggest that net active pressure (i.e. landward to seaward) exists in this zone is quite ridiculous and contrary to any credible soil-structure model. Mr. Bracegirdle said that the force would be a net force landward to seaward between -14.0mCD and -23.0mCD and a net force from seaward to landward below -23.0mCD. I take it that by "this zone" Mr. Williamson meant the zone between -14.0mCD and -23.0mCD. I cannot accept that there is anything ridiculous or incredible about the suggestion. Mr. Williamson's third point was this: The pressure on the grout column beneath a level of -20.7mCD can only be imposed by the surrounding soil. The magnitude of this pressure both in front of, and behind, the columns will be dependent on whether the column moves landward or seaward. For a large change to occur at one level would require the column to move in two directions at the same level. This cannot occur in the column as modelled in the computer. The point of the first sentence is that the sheet pile wall has no effect below the bottom of the toe. I shall assume that that is correct. The point of the rest of the submission, as I understand it, is that there is a sharp change of direction in the graph at -23mCD. It is clear on the evidence that that arises from the coarseness of the grid used at this point in the finite element analysis. That is a justifiable criticism which I accept. As Mr. Bracegirdle explained, and as Mr. Reid showed in his calculation of the shear force that the grout wall had to resist, there is a net landward to seaward force on the grout wall at the bottom. Mr. Bracegirdle said that that would occur below -23.0mCD. I accept that the results of the finite element analysis will inevitably be inaccurate. I accept that the reliability of the input data must affect, and can seriously affect, the reliability of the outcome. However, no other such analysis has been put before me, and it is the best evidence of many of the matters which are the subject of its outputs. Bending moments to be expected in the sheet pile wall arising from soil pressure have been calculated by Mr. Bracegirdle using the ICFEP finite element analysis and by Mr. O'Brien. Mr. O'Brien's method was, as he stated in his report, Based on a parametric study of the influence of several factors, based on simplified methods using Wallap (a commercially available "beam on elastic-plastic springs" retaining wall analysis program) and hand calculation methods. The ICFEP analysis by which the initial state of the sheet pile wall in berth 205 (i.e. before the strengthening works were carried out) was calculated ("run 200") showed a maximum bending moment of 760 kNm/m, which was at the relieving platform level, and a maximum deflection (distance from chord, assumed vertical) of 96mm. The maximum deflection, and the corresponding local maximum bending moment of 580 kNm/m, both occurred close to a depth of -10mCD. The value of (K0) used in the calculation was 1.5. Although the finite element analysis was criticized, Mr. O'Brien considered that apart from a number of factors that he mentioned, the calculation (or "base case" as he called it) was likely to produce a maximum bending moment in the sheet pile wall of something between 700 and 900 kNm/m. The factors which Mr. O'Brien considered would lead to an increase in bending moment over the base case were these. First, the factor (K0). Second, that the stress relieving platform was not wholly effective. Third, that there had been over-dredging of the berth. As to the factor (K0), it seems to me that Mr. O'Brien has conflated two matters. The first is that the (K0) value chosen for the Marsh Farm deposits of 1.5 was not necessarily realistic: it could be higher, and would certainly vary with depth (his report, paragraph 12.2(ii)). The second matter is that the driving of the H-piles would have increased the value of (K0) (ib., paragraph 12.2(vii)). But in my judgment there is an important difference between the two uses of the parameter (K0). That parameter is a characteristic of the soil, and represents the ratio of the horizontal stress to the vertical stress. In a large mass of soil, the greater the depth, the greater the pressure from the overlying soil, hence the greater the horizontal pressure for a given value of (K0), other things being equal. That pressure applies through the whole mass of the soil at the relevant depth. But the effect of driving the H-piles is different, since it is local to the H-piles. The H-piles bear the weight of the soil above the relieving platform and the surcharge loads on the ground at the top principally by reason of the frictional grip on them of the soil surrounding them. The driving of the H-piles in those circumstances is reflected not only in a vertical but also in a horizontal stress in the soil, especially in the layer of very dense sand, which lies typically between -10 and -11mCD. That horizontal stress is resisted by elastic forces in the soil which produce a horizontal pressure gradient so that the extra horizontal force produced by the H-piles becomes negligible after some distance from them. What Mr. O'Brien said in paragraph 12.2 (vii) of his report is this: It is well known that horizontal stresses in soil adjacent to driven piles will increase beyond their "at rest" values. Hence, the assumption of the sheet pile and bearing piles being "wished-in-place" is highly questionable, and will lead to an incorrect starting point (in terms of initial stresses, bending moments, shear forces etc. in the sheet pile wall) for modelling of subsequent construction phases. In particular, the driving of the H-piles behind the sheet pile wall will cause large increases in insitu horizontal stress in the soil behind the sheet pile wall. This increase in "(K0)" will lead to an increase in the bending moment/shear force in the sheet pile wall, prior to the VHP strengthening works. And in his report at paragraph 12.3(i) Mr. O'Brien expressed the conclusion that the K value (i.e. the value of (K0)) behind the sheet pile wall would be expected to be higher than 2.0 and might be close to 2.5. Mr. O'Brien said (day 26, page 164) that those effects would not have occurred if the H-piles had been driven before the sheet piles were driven. Mr. O'Brien said in the course of his cross-examination by Mr. Bowdery (day 27, page 44) that he could provide some imprecise calculations of the stresses that would develop as the H-pile was driven into the ground. He was asked to do so. The calculations were duly produced. In those calculations, relying on a book by Jardine and Chow (ICP Design Methods for Driven Piles in Sands and Clays, 2005), Mr. O'Brien calculated that the force applied to the sheet piles was about 1.6 times to 5.3 times the capacity of the clutch. I think that those calculations must represent earlier calculations on which Mr. O'Brien based his report, since they were dated 16 th December 2005, after the end of the oral evidence. I think it is fair to say that Mr. O'Brien was not very happy about those calculations. He had said (transcript, day 27, page 44): …..if the court wished I could provide some calculations on that in terms of the impact on the wall, but really I wouldn't describe them as – it is not a precise form of calculation really, because even with the amount of research that has been done so far, there is still an uncertainty about how the radial stresses around the toe of the pile would decay with distance. The following exchange took place shortly afterwards (day 27, page 46): Q. It has been suggested to me by, I think, Professor Jardine, that that analysis [sc., the analysis of Jardine and Chow] cannot be translated directly to berth 205, as a very small amount of deflection will relieve the pressure induced by driving? A. That, well, there are two things there. One thing is the impact during the driving process on the sheet pile wall itself and in particular, the clutches, and the clutches are a potential weak point. In terms of the locked-in stresses, again we get back to the point about the connection detail. If the wall was free to rotate and [sic] Professor Jardine would be exactly right, it would leave and quickly the locked-in stresses would reduce. But I think the real point here, because of the counterforts the wall is not that free to move so it is not that easy for it to deflect, but you can do it at deeper levels when you get remote from the stress-relieving platform, but as you get closer to the stress-relieving platform the counterfort effect becomes more and more important….. The relevant conclusion of the calculation, which I have quoted in part above, was: Applied force is about 1.6 times to 5.3 times larger than capacity of clutch (i.e. factor of safety about 0.2 to 0.6), this occurs between -10 and -11mCD. Thus Mr. O'Brien's point about the counterforts is irrelevant. Mr. O'Brien's calculation shows the force said to be applied to the sheet pile wall by reason of the driving of the H-piles. It does include a calculation of the strength of the clutch, but it does not show the response of the sheet pile wall to the applied force. Until the berth was dredged to -12.8mCD, the sheet pile wall would have been supported by the soil on the passive side of it. It may be that a small movement of the wall would have relieved the net force on the wall. Whilst there may have been some difference between the actual situation and a situation where the H-piles were driven before the sheet piles were driven, no evidence as to the amount of any such effect has been offered. The passages from Mr. O'Brien's report quoted above suggest that the calculation is a calculation of permanent effects of the driving of the H-piles, not of transient effects, even though some of the causes appear to be transient. Part of the force calculated is a force existing below the level of the toe of the H-pile. That force at any given level of the sheet pile wall is transient, occurring only before or at the time when the toe of the H-pile passes the relevant level as it is driven. And in the calculation, the rate of hammering of the H-piles is a relevant factor. That again is clearly a transient cause. (Mr. O'Brien stated in his calculation that at the current state of the art it was not possible to carry out an appropriate dynamic calculation. He applied a factor of 1.2³ [equal to 1.73] to the calculated stresses to allow for the dynamic effect of the hammering. That figure was derived from what appear to have been empirical published figures). I cannot accept Mr. O'Brien's evidence that the effect would not have occurred if the H-piles had been installed before, rather than after, the installation of the sheet piles. It is of course true that the calculated forces would not act on the sheet pile wall if the sheet pile wall were not there. But the stresses calculated to have been in the soil were not dependent upon the presence of the sheet pile wall. If the sheet pile wall were driven subsequently, the forces each side of the wall resulting from those stresses in the soil would be substantially equal and opposite, and hence balanced. They would cause negligible relevant stress to the sheet pile wall. But once the dredging to -12.8mCD took place, the balancing passive force from the soil would be removed. The situation would not then be substantially dependent on the order in which the two sets of piles had been driven. If I am wrong, and the effects that Mr. O'Brien was calculating were transient effects, then the same conclusion applies, since in that case the sheet pile wall would have been supported at the material time by the soil on the passive side. The stresses induced in the sheet pile wall by such transient forces have not been calculated, and there is no material before me to suggest that they might have been significant. There is a further, related, consideration. The H-piles cause a local increase in the horizontal stress in the soil by reason of their displacement of the soil as they are driven. I accept evidence of Mr. Bracegirdle that that cause has an effect on the sheet pile wall which is negligible for present purposes. Whether that consideration is included in the Jardine and Chow method does not appear. But the conclusions expressed in paragraphs 103 and 104 above still apply. Thus I am satisfied that this (K0) effect does not depend to any significant extent on the order of driving the H-piles and the sheet piles. Another point pertaining to the fact that the H-piles were driven after the sheet piles was made by Mr. Reid. He said in his report, paragraph 29(3), that during the driving of the H-piles unwanted horizontal displacement of the sheet piles would occur. The effect would have been exacerbated by the system used to support the piling rig. One end of the piling rig was supported on the ends of the cantilevered sheet piles. Unfavourable forces and vibrations would thereby have been imparted to the sheet pile wall. He explained in oral evidence (day 22, page 102) that the vibrations caused by the hammer and by the horizontal movements of the piling rig would loosen the soil and reduce its capacity to resist the landward movement of the tops of the sheet piles. He said (report, paragraph 29(9)) that the combined influence of all those secondary effects was impossible to assess with accuracy, but it could be deduced that the first plastic hinge that could develop in the sheet pile wall after dredging would occur close to the river bed, rather than at relieving slab level as would otherwise be the case. I need not consider that last deduction. As to the movement of the sheet pile wall caused by the movement of the massive piling frame and cranes, the forces generated would depend on the acceleration of those objects as well as their location. No figure for their acceleration, or indeed any calculation, was offered. I regard this evidence of Mr. Reid as speculative and I attach little weight to it. I am not satisfied that the order of driving the H-piles and the sheet piles had any significant effect on the state of the sheet pile wall before the grouting works were executed. Mr. O'Brien considered that a change in the value of (K0) from 1.5 (as in Mr. Bracegirdle's ICFEP run 200) to 2.5 (Mr. O'Brien said that the K value might be close to 2.5) would increase the maximum bending moment in the sheet piles by 10 to 25 per cent. He did not say how he arrived at the amount of that increase. Mr. Bracegirdle (ICFEP run 206) calculated the maximum bending moment on the basis of a value of (K0) equal to 2.4 at all strata as 850 kNm/m, an increase of 12 per cent. over his figure of 760 kNm/m for (K0)=1.5. Mr. O'Brien's second factor was that the stress relieving platform was only partially effective, because of the relatively stiff competent materials underlying it. He considered (paragraph 12.3(iii) of his report) that the top of the soil below platform level would bear 15 to 30 per cent. of the overburden at platform level. I digress to consider the relieving platform. The Relieving Platform The purpose of the relieving platform was to reduce the seaward stresses on the sheet pile wall imposed by the soil on its landward side. Those seaward stresses arose from the vertical pressure on the soil caused by the weight above it. Some of the weight of the soil and surcharge loads above the platform and resting on it was to be borne by the H-piles, which would transfer that load to lower levels. Mr. Bracegirdle referred (report, section 19.3) to an assumption that the relieving platform would transfer all the soil self weight to the toe of the sheet pile wall. The expression "all the soil self weight" clearly refers to the loads on the platform that I have mentioned above. Mr. Bracegirdle said that that assumption was incorrect and unjustified. However, he did say (report, paragraph 6.11.7) that some small settlement of soil could be expected beneath the relieving platform. In that case, there would be no load on the top of the soil beneath the relieving platform. I thus interpret Mr. Bracegirdle's denial of the assumption to his recognition of the fact that the H-piles would support the load above the relieving platform in part by friction in such a way as to increase the stress in the soil above the level of the toe of the sheet pile wall. Mr. Reid estimated that the percentage of the load transferred directly to the soil underlying the platform was approximately 50 per cent. (report, paragraph 34). He reached that conclusion in part in reliance on his "bulking" theory. His bulking theory is to be distinguished from dilatancy, which is the volume increase of soil caused by shearing forces. The latter is well established; the former appears to be unique to Mr. Reid. As I understand it, the bulking theory asserts that when a load which has compressed soil is released, the soil loosens and expands. Thus when the ground is excavated for the purpose of casting the relieving platform, the soil beneath loosens and expands by heave. When the ground is restored above the platform, the soil immediately beneath the platform bears some of the load. Not all is transferred to the H-piles. In Mr. Reid's own words (report, paragraphs 34 to 38): I have computed the proportion of load that would be carried by the [H-] piles and the proportion that would be carried by the underlying soil. This I have based on the extent of heave that would be anticipated in the soil when the 6 metres excavation took place for the relieving slab. I have adopted this [criterion] since backfilling the slab will, given the over-consolidated soil beneath, be limited to this value [sc., backfilling the slab will cause it to be lowered by an amount limited to the extent of the heave]. I assess that the heave would be in the order of 2mm and this yields a prediction that the percentage of load transferred directly to the soil is approximately 50%. I believe this to be realistic. When the soil in front of the wall is dredged to a level of -12.8mCD, the wall will deflect outwards and a condition of active pressure will develop behind the wall….. While surcharge of the soil underlying the relieving slab is clearly inevitable prior to dredging, the mechanism of load transfer after the wall is allowed to deform is less apparent…..provided the soils behind the wall are not loose, volume increase will take place when the soils deform and loosen to fill the space created by the wall deformation….. The extent of loosening and volume increase of a soil behind a flexible retaining wall is not well understood and difficult to predict with accuracy….. …..it is very likely that the space vacated by the wall deformation must have been accommodated by volume increase and loosening of the soil behind the wall. Mr. Reid explained the matter further in evidence in chief (day 17, pages 36, 37): Q. Mr. Reid, in relation to the various arguments that you have just been discussing, what is the relevance, if any, of the bulking of soils? A. I think Mr. Bracegirdle has said that the soil mechanics textbooks don't mention this phenomenon and it mentions dilation and I agree with that. However, I did back in the late 1970s quite a bit of research on mining subsidence where I was preventing what we call plump holes in the ground under motorway embankments and I stretched membranes over circular holes in the laboratory and I created a plump hole and I measured what happened. What I found was that the soil arched round the hole but it also bulked as it went down so the amount of volume you got by allowing the membrane to sag wasn't mirrored by the amount of settlement you got on the surface. I have spoken to quite a number of eminent soil mechanic people and they say nobody has written a paper on this so therefore it can't happen. I don't subscribe to that, that just because somebody has not written a paper about it, and I didn't either, that does not mean to say it doesn't happen. What, when I looked at the sheet pile, it seems to me that although it is to a wall it will behave like a membrane and if you have dense soil, which you have here, and it is allowed to yield, what will happen is it will loosen, and that is just to let you understand what I'm saying there is that if you were having your driveway concreted and the contractor came along and he dumped a load of gravel on your soil and then he came along with a roller; so if he had 200mm layer of gravel underneath his roller, it would go down to about 175. So it started off loose when you dumped it down there and when you roll it in it becomes dense. What I'm saying is going to happen behind the wall is the reverse of that: the dense soil that is there becomes loose as it yields away and the effect of water and so on in it. Now, that was what I suggested to Babties at the time and they said I was wrong: there would be a gap under the wall [sic: slab?] ….. Although Mr. Reid is reported as having said "wall", I think he must have said, or intended to say, "slab", the point that he was answering being that if the soil did not bulk but filled the bow of the sheet pile wall, the level of soil under the relieving platform would fall. Mr. Reid calculated the percentage of the load on the relieving platform carried by the H-piles in this way. Assuming a certain percentage load carried by the H-pile, he calculated the consequent shortening of the pile. That would involve the lowering of its top. He then computed the heave that would generate the assumed percentage. He used assumed characteristics of the heaved soil to determine the proportion of the load that would bear directly upon it and the proportion that would be supported by the H-pile. Mr. Reid said (day 21, page 128) that in his view the soil would move back down, upon loading of the relieving platform, only by the amount it had heaved when the soil had been excavated. Having assessed the heave, by other means, at 2mm, he then arrived at his figure of 50 per cent. I take it that he used the results of the research he mentioned in order to assess the heave. He said (day 21, page 128) that it was impossible to calculate the heave. However, it appears from the appendix to his report, sheet 17, that the heave that according to his calculation produces the 50 per cent. division is not 2mm but 4.444mm. 2.222mm is the shortening in the length of the H-pile. I am far from saying that bulking cannot happen because nobody has written a paper on it. But Mr. Reid's assessment of 2mm heave is not supported by evidence that has been put before me, and appears to be inconsistent with his own calculation of 50 per cent. Mr. O'Brien did not adopt the bulking theory. He put the matter this way (report, paragraph 12.3(iii)): For the bearing piles to mobilize their resistance they will need to deform. At working load the displacement required to mobilize their shaft resistance is likely to be relatively small, of the order of 5mm. However, the stress relieving platform is underlain by relatively stiff competent materials…[He then considered the characteristics of those materials]…Settlement calculations for the platform (based on the methods of Burland et al., 1977; and Burland and Burbridge, 1983), ignoring the presence of the bearing pile, indicate that at bearing pressures of between 20 and 40kN/m², the platform would settle about 5mm. Hence, there is likely to be a significant load sharing between the bearing piles and the platform itself, i.e. the piles would carry out [sic: about?] 70% to 85% of the overburden above platform level, whilst the soil below platform level would carry the remaining 15% to 30%. This is entirely consistent with the results of rigorous analysis of piled raft behaviour. The unwritten premise is that the bearing pressures on the relieving platform would be more than 20 to 40kN/m². That is clearly so: I accept the figure of 130kN/m² given by Mr. Reid in the appendix to his report at sheet 17. Thus Mr. Bracegirdle considered that the relieving platform imposed no load directly on the soil beneath it, but accepted that there would be an effect on the sheet pile wall by reason of the shaft resistance of the H-piles lower down. Mr. Reid and Mr. O'Brien considered that both effects would exist. I am concerned with the magnitude of the effect, whatever the mechanism of its cause. I revert to Mr. O'Brien's evidence. In table 12.1 in paragraph 12.3(iv) of his report he postulated an increase in the stress borne by the top of the soil below platform level from 0 (the base case) to 40kN/m². The latter figure represents approximately 30 per cent. of the load on the relieving platform (130kN/m²: see paragraph 119 above). He considered that that increase would cause an increase in bending moment in the sheet pile wall of 5 to 15 per cent. He did not say how he arrived at those figures. Mr. Bracegirdle's figure of 760kNm/m was based on the assumption that there was no stress on the top of the soil below the relieving platform. He calculated that if the whole of the load on the relieving platform were borne by the soil beneath it, the maximum bending moment would be increased by 45 per cent. to 1100kNm/m. On the assumption of proportionality, which is the best one can do, that implies that the maximum bending moment if the relieving platform allowed 15 to 30 per cent of the load on it to be transferred to the top of the soil beneath it would be increased from 760kNm/m by 6¾ per cent. to 13½ per cent. Mr. O'Brien considered that the effect of over-dredging from -13.1mCD to -13.7mCD would be to increase the maximum bending moment by 5 to 10 per cent. He did not say how he arrived at those figures. Mr. Bracegirdle, using run 220 for some reason that is not apparent, calculated that over-dredging from -12.8mCD to -14.1mCD would increase the maximum bending moment from 727kNm/m (cf. run 200: 760kNm/m) to 750 kNm/m, an increase of 3 per cent. No calculation has been done of the combined effect of all these changes. The best that can be done to give a rough estimate is to treat them as mutually independent. Applying Mr. Bracegirdle's estimates in that way, the combined effect would be to multiply the figure of 760kNm/m by 1.12 X 1.0675 X 1.03 (=1.23) to 1.12 X 1.135 X 1.03 (=1.31) giving a maximum bending moment of 936 to 995kNm/m. Mr. O'Brien's estimate of the combined effect is 1.2 to 1.5. Thus there is not a wide degree of dispute between Mr. Bracegirdle and Mr. O'Brien on the combined effect of these matters, on the assumption that Mr. O'Brien is right as to the change in the factors mentioned in paragraph 96 above. Taking the maximum figure of Mr. O'Brien's base case, 900kNm/m, and the figure of 2100kNm/m as the bending moment capacity of the sheet pile wall, 43 per cent. of the structural capacity of the sheet pile wall would have been utilized. That accords with his opinion that "one would normally expect the applied bending moments to be a low percentage of the wall capacity, with perhaps less than 50% of the structural capacity being utilized" (his report, conclusion of section 12.3). He concluded that the changes in parameter which he considered appropriate, and which I have considered above, could have increased the utilization factor to about 80% to 90%. Those figures represent increase factors of 1.87 to 2.1 from a base of 900kNm/m, and of 2.4 to 2.7 from a base of 700kNm/m. He concluded that the maximum bending moment moves perilously close to the bending capacity of the sheet pile wall, giving a far lower margin of safety than would normally be expected. But those factors go well beyond the rest of his evidence. If one takes his maximum estimate of the base case, namely 900kNm/m, and his maximum estimate of the uplift factor, namely 1.5, one arrives at a utilization factor of 64 per cent., giving a margin of safety of 55 per cent. I accept Mr. Bracegirdle's evidence as to the individual factors representing the effect of the changes in parameters proposed by Mr. O'Brien. That is to say, I accept that they correctly reproduce the effects of the changes as calculated in accordance with the ICFEP analysis. Mr. O'Brien himself said that ICFEP was a highly respected software package (section 12.2 of his report). It was the assumptions and idealizations that he criticized. In so far as the assumptions did not meet Mr. O'Brien's criteria and gave rise to his suggested increases in bending moment, they have been amended in the manner indicated above. There remain criticisms, including the fact that the analysis is a 2-dimensional analysis of a 3-dimensional problem. But in my judgment, those criticisms are likely to be of lesser force when applied to adjustments than when applied to the original calculations. Thus in my judgment Mr. Bracegirdle's factors of 1.23 to 1.31 are more reliable than Mr. O'Brien's effective factors of 1.87 to 2.7. Mr. O'Brien does not differ from Mr. Bracegirdle in his base case, though he gives a bracket. If one takes Mr. O'Brien's median base case of 800kNm/m and applies a factor of 1.3 to it, one arrives at a figure of 1040kNm/m as the maximum calculated bending moment. That gives a factor of safety of 2 in relation to the formation of plastic hinges. There will be variations along the wall since, as Mr. O'Brien pointed out, there are variations in soil strata along the wall. However, the bending moments calculated from consideration of soil pressure are consistent with the observed curvatures, except in the immediate vicinities of X1, X2 and X3. Effect of grout pressure on SSP wall The defendants contend that the reason why the sheet pile wall came to bulge excessively when the grouting process took place was that there were existing plastic hinges in the sheet pile wall. That proposition was somewhat widened to the proposition that the sheet pile wall was in a state of excessive and unforeseeable bending before the grouting took place. The possible effect of the grouting is thus relevant to determination of the initial state of the sheet pile wall. Before considering that question, I shall give here a fuller description of the grouting process than I have given above. The following is a description given by Dr. Greenwood. In its simplest and original from jet grouting proceeds as follows. A borehole is drilled into soil by any suitable means, of diameter just sufficient to accommodate a drill string and jetting "monitor" (an attachment at the bottom which contains one or more orifices to create radial jets) with an annular space to allow return of circulating waste fluids. This hole is usually 150 to 200mm diameter and it extends to the far limit of intended treatment. The monitor is inserted to the bottom of the hole and the radial jets are turned on whilst the drill string is slowly rotated and withdrawn. The jetting fluid is cement grout. As the jet(s) impinge on the sides of the unlined borehole the soil is eroded and a larger diameter cavity results, extending upwards with the jets as they are simultaneously rotated and lifted. The eroded soil mixes with the grout to form a fluid soil-cement mix in the cavity which, incidentally, helps to support the borehole against collapse. As fresh grout is pumped continuously to maintain the jets a similar quantity of effluent comprising soil-cement is ejected through the annulus alongside the drill string to waste at ground surface. When the desired length of enlarged column is formed the drill string and monitor are withdrawn and the soil-cement remaining in the cavity and bore is allowed to rest and harden in situ. A column of cemented soil results. The equipment used by HSS at Berth 205 had a monitor containing three jets. Two were diametrically opposed and fitted with 3mm diameter nozzles and one, with a 2.45mm nozzle, was directed downwards at an angle of 5° from vertical. In this case the monitor was used as a drill bit during the formation of the initial borehole and the downward jet assisted penetration through the soil in addition to cutting teeth. Moreover, the horizontal jets worked in tandem with the downward jet to pre-cut (or pre-wash) a larger initial bore than would be cut by a simple drill bit. Thus jetting during withdrawal began from a larger base diameter than the conventional process described above. An Ordinary Portland, P 42.5 cement was used in a grout of water/cement ratio 0.82 by dry weight during both penetration and uplift, both phases being used for jetting/soil mixing. This basically simple arrangement and procedure allows larger diameter columns to be cut than by the conventional fluid system, given equivalent energy input and soil resistance to erosion. The target diameters were 0.4m and 0.9m on the respective phases. That is a description of the process as initially operated. It was intended that the columns should be formed in continuous lengths from a depth of -24.75mCD to +0.39mCD (the underside of the relieving platform). However, there frequently occurred cracking of the sheet piles, and in consequence, after about 12.9 per cent. of the columns had been constructed, the process was altered to a two-stage process. The first stage was to construct a column down to -7.0mCD. An admixture was included in the grout to retard its setting. A down-the-hole hammer was used to drill through the hardening upper stage to allow access to construct the second stage, from -24.75mCD to -6.0mCD. The second stage was generally constructed between three and fifteen days after completion of the first stage. Mr. Bracegirdle's ICFEP analysis (Run 200) showed that in the absence of grout loading the sheet pile wall, assumed to have been driven vertically, would have an offset of 96mm from the vertical at a depth of about -10mCD. The maximum bending moment in that vicinity would be 580 kNm/m at approximately the same depth. There would also be a maximum bending moment of about 760kNm/m in the opposite direction at the level of the relieving platform. When the grouting process was working as intended, the high pressure causing the high velocity of the jet of grout was dissipated as the jet was absorbed in the surrounding soil. The pressure that continued in the soil while the grout was liquid was simply the hydrostatic pressure of the grout, plus the pressure difference required to drive the excess grout up the annular tube back to ground level. If the grout return was lost because the annular tube was blocked, then the high pressure of the jet would extend to the liquid grout in the soil for so long (if at all) as the jetting process was continued. The grout return could also be lost because grout was finding its way to contact with the sheet pile wall, forming a lens behind the sheet pile wall and leaving no excess to pass up the annular tube. I derive the foregoing propositions from the evidence of Dr. Greenwood. I have been unable to reconcile two pairs of different statements made by Mr. Bracegirdle in relation to the finite element analysis of the effect of grout pressure on the wall. The differences were not put to him. The first was this. In section 12.8 of his report he said: The outer of the VHP columns is in close proximity to the rear face of the sheet pile wall. On average, the columns project directly on to about 50% of the area of the sheet pile wall. That is, a vertical section through the centre of the seaward row of columns would show that the columns occupy about half of the total area. The unit weight of the fluid grout is estimated to be about 17 kN/m³. The net lateral pressure imposed on the sheet piles by the grout in its liquid state is shown in Figure 72, normalised to account for the area of column exposed to the wall. [Emphasis added]. I understood that passage to mean that the assumption underlying the calculation was that half the area of the sheet pile wall was covered by the liquid grout. (The calculation was a two-dimensional calculation, i.e. it considered a small length of the wall in isolation, or possibly with defined restraints). But in the course of cross-examination by Mr. Williamson (Day 10, page 111) Mr. Bracegirdle said: You have to remember that as each column is formed, it will harden and then the next column will be formed. So you don't get this grout loading over the whole of the wall at the same time, and that's what I am assuming in the finite element analysis. I'm assuming that the grout is spread evenly right across the back of the wall, whereas in reality of course we are only looking at the column of grout…..unless, of course, the grout breaks out to the back of the wall, and then it can spread laterally along the wall, so the grout pressure's only on the back of the wall where the lens is, and no further. The other difference is this. Figure 72, mentioned above, shows the pressure of the liquid grout as being substantially greater at all depths than the active soil pressure in the absence of the grout. Indeed, the passage from Mr. Bracegirdle's report mentioned above continues As may be seen from figure 72, the grout pressures are much greater than the [sc. soil] pressures from limit equilibrium analysis. (I have added the words in square brackets). There was a break during the course of Mr. Bracegirdle's oral evidence to allow the experts to meet to discuss the ICFEP finite element analysis. A transcript of what transpired at the meeting was put before me. From that transcript (page 19) Mr. Bracegirdle is recorded as saying: The excess pressure only occurs between about relieving platform level and, I think, about minus 13, because you have quite high pressures behind the wall anyway. So the grout pressure is only over the very top….. I cannot resolve those apparent discrepancies. However, I conclude, in particular from what Dr. Kovacevic is recorded as having said (transcript of the meeting, pages 17 to 19), the following. At increment 45 the system (including the state of the sheet pile wall) had been modelled on the assumption of no grout pressure. Dr. Kovacevic established the difference between the total horizontal stress which was acting at each level of the wall at increment 45 and the hydrostatic pressure due to liquid grout at the same level. He then applied, using the program, successive increments of 10 per cent. of that difference, assumed to be acting on the wall wherever the grout pressure exceeded the calculated horizontal soil pressure. That showed increasing bending moments in the sheet pile wall and corresponding increasing bulging in that wall. Sensitivity to grout pressure The following table, derived from graphs contained in Mr. Bracegirdle's report, shows the maximum bending moments and displacements of the bulge from the vertical calculated by him using ICFEP on the assumption that the piles were driven vertically. (1) Increment (2) Maximum bending moment at relieving platform level kNm/m (3) Maximum bending moment at bulge kNm/m (4) Maximum displacement mm (5) Ratio (2)/(4) (6) Ratio (3)/(4) 760 96 7.92 6.04 1200 140 8.57 6.21 1620 200 8.10 6.00 2000 255 8.04 6.08 2000 350 5.71 5.71 2000 1040 1.92 1.92 Increment 45 was on the basis of no grout loading. Increment 46 was on the basis that 10 per cent. of the excess of hydrostatic grout pressure over active soil pressure was applied over the whole relevant width of the sheet pile wall. The successive subsequent increments represented successive increments of 10 per cent. of that excess. Increment 50 represented an increment of 50 per cent. It is apparent from the transcript (page 21) of the meeting of the experts that was put before me that the model ceased to converge at a grout pressure of 47 or 48 per cent. of the full excess of the hydrostatic pressure over the active soil pressure. That means that the model showed the wall to keep moving outwards at that stage. It would not reach an equilibrium position. Mr. Bracegirdle said that in the event of hydrofracture the comparison of stresses would be considerably more onerous. He said that loss of slurry return to the surface would cause the immediate outward displacement of the sheet piles and hydrofracture. I accept that if loss of slurry return through blockage of the annular cylinder occurred the grout could apply pressure to the wall far greater than the hydrostatic pressure. Whether it would produce an effect on the wall greater than that calculated by the analysis depends on the extent of the wall subjected to the pressure. The analysis itself assumes a substantial coverage by grout at hydrostatic pressure. But whether or not the process was hydrofracture strictly so called is not of significance. Dr. Greenwood considered that a more likely process was that the hydrostatic pressure of the grout could be transferred either directly or through a few centimetres of sand or gravel to deform the wall. He wrote in his report (paragraph D33): This is a more satisfactory explanation of the observations as the borehole pressures [sc., the pressure of the grout or soil/grout fluid] no doubt caused deformations of the wall and allowed grout to spread behind the wall in whatever direction offered least resistance when the breakout happened. It is this potential for weakening of the sandy soils and consequent spreading of grout behind the wall that was a risk following a deformation, which the designers may not have recognised. (I have added the words in square brackets). Mr. Bracegirdle said, and I accept, that an exact analysis of the effect of liquid grout on the sheet pile wall was not possible. One reason he gave, which I accept, was that the restraining effect of the neighbouring sheet piles was highly uncertain. He said that the calculations nevertheless showed the sensitivity of the sheet pile wall to grout pressure. Dr. Greenwood said this in his report: C110. The location of the seaward columns of each pair [of grout columns] as designed meant that there would be contact of fluid soil-cement with the rear of the sheet piles during the construction of the columns. This would impose a hydrostatic pressure on the wall, varying…..with time for the grout to consolidate and set….. When the grout set and continued to harden this element of load on the wall would eventually be totally relieved. C111. The development of structural rigidity would take about one hour, depending primarily on the consolidation rate of the soil cement mix and on the effectiveness of the retarder included in the mix. The former is likely to have been the determining factor where the soil cement water/solids ratio was…about 0.5 from the laminated clays. At a rate of construction of 5 minutes per metre at least 12m of fluid could be pressing on the wall. In the two phase method of construction this would be limited to the length formed in each phase. C112. If the wall was already deformed before VHP grouting commenced, the distance from column centres to the wall would be increased by the amount of deflection at the relevant elevation and the column may not have had contact with the wall unless it was larger than designed. Since the wall was tied to the relieving slab the relative deviation at that level was zero so some contact was inevitable…… C114……To be precise, a check must also be made on the pressure necessary to maintain the flow through the borehole annulus representing fluid friction losses….. C116. These pressures would be applied to the back of the wall as a vertical strip load where the column was in contact with the wall over a depth range of about 12m. I accept that evidence except in so far as it expresses a limit on the area of the sheet pile wall which could be covered by the liquid grout. It is apparent that the grout could spread widely behind the sheet pile wall. There is uncontroverted evidence that grout was found no less than 16 metres from the edge of the works. Mr. Bracegirdle said that the grout could form a lens behind the wall which would expand as it filled with more grout. Mr. O'Brien said this in the course of his cross-examination by Mr. Bowdery: ….. But I think that, given a lot of the columns were basically in contact with the wall, if the wall did move away, I think the grout would probably just take the line of least resistance just to fill up the void which starts to develop behind the sheet pile wall….. …..given the geometry and location of the jet grout columns relative to the wall, if the wall did start to move away, I think the grout would just fill up the gap. I find the pressure head required to drive the excess grout, or soil/grout fluid, up the annular tube to have been about half a metre when the grouting was being carried out in a single phase, and about one metre when it was being carried out in two phases. Thus in normal operation that extra pressure head would have to be added to the relevant depth to obtain the pressure of the grout while it remained liquid. I obtain the figures of one-half and one metre as follows. The pressure head h is given by h=48.lv∨/g(b-a)² where l is the length (height) of the annular tube; v is the average speed of flow of the fluid up the tube; ∨ is the kinematic viscosity of the fluid; g is the acceleration due to gravity; b is the outer diameter of the annular tube; and a is the inner diameter of the annular tube. The values of those parameters I have taken as follows: l=6.89m (for the single phase and for the first phase of the two-phase working); l=13.5m for the second (lower) phase of the two-phase working; v=0.142m/sec; ∨=1.67E-3m²/sec; g=9.81m/sec²; b=0.213m; a=0.090m. The values of l, v, b and a apply to the upper, narrower, section of the annular tube (of outer diameter 0.213m). Dr. Greenwood accepted the formula for the pressure head and its method of calculation, but he arrived at different values of h. He confined his calculation to the case of l=6.5m. He said that the pressure head in the upper section of the tube was 0.17m equivalent water head. He arrived at the figure of 0.17m by taking the value of l as 6.5m (as he explained) and the value of ∨ as 5.5E-4 m²/sec. (as is apparent). The values of v, b and a are his values, which I accept. Dr. Greenwood explained the method of grouting in his report by reference to a diagram which he identified as figure 3. That shows the upper section of the tube to be of length 13.5m for the second phase of the two-phase working, otherwise 6.89m. That is borne out by the text in paragraphs A16 and A17 of his report. As to the value of the kinematic viscosity ∨, he gave a range of values from 5.5E-4 to 27E-4. My figure, which I obtained by the imprecise method of measuring the gradient of a curve in figure A5-1 of his report, is in the middle of that range. He evidently took the minimum value of the range because the shear rate (velocity gradient) of the fluid flowing in the upper, narrower, section was greater than that of the fluid flowing in the broader section. He took the maximum value of the range for the fluid flowing in the broader section, and arrived at a figure of 0.12m as the equivalent water head loss for that section. Thus his total figure for h was 0.29m water equivalent. He was, however, wrong in thinking that the formula gives a water equivalent value of h. The density of the fluid is implicit in the kinematic viscosity, and the value of h given by the formula is the head of fluid, not of water. Dr. Greenwood said that his figure A5-1 showed that the viscosity was very sensitive to change in shear rate at low values of that rate. I do not accept that his figure A5-1 shows that, though the scale is so small as to make estimation difficult. I find the pressure head required to drive the fluid through the broader section to be negligible. All these conclusions are based on the assumption that the annular cylinders are not obstructed. Dr. Greenwood said that with flowing suspensions of particles tending to pastes we do not know how the velocity difference is distributed across the section as viscous fluid flow transforms into plug flow with reducing liquid phase. The calculations do not account for any difference in resistance between viscous and plug type flows. I accept that evidence. Whilst the figures of ½ metre and 1 metre are in my judgment the best available estimates, they cannot be regarded as reliably accurate. Increments 45 to 50 assume a certain percentage of the excess of the hydrostatic pressure over the geostatic pressure applied over the whole wall. In fact, on the above evidence, the whole of the excess could act, and could well act over a height of about 12 metres and a comparable width. While the effluent was flowing up the annular cylinder, the pressure would be greater than the hydrostatic pressure. Grouting effluent blockages A device known as a Lutz registration device was used by HSS to monitor the installation of the grout columns. According to the unchallenged evidence of Mr. Nico Claessens, who acted as works manager for the installation of the columns, and whose evidence was admitted under the Civil Evidence Act, that device registered all relevant grouting parameters, namely depth, penetration/lifting speed, rotation speed, grout pressure and grout flow. The device showed the data in a digital form and also produced a paper record. It kept a continuous record except on occasions when it malfunctioned; on those occasions, operatives checked the relevant gauges, manometer and flow meter. A visual inspection of the grout return was continuously undertaken. If loss of grout return was experienced, the VHP installation was stopped. Reduced flow rates and pressures were applied, and the rods were moved up and down irregularly, a procedure known as "dancing". The problem was overcome by gradually building up pressure and flow to the pre-determined rates. Mr. van der Eecken, who gave evidence before me, was Product Manager for HSS with responsibility for quay wall projects. He said in his witness statement that when he attended on site, he observed the works including the registration of the data by the Lutz device. He also reviewed the records for the days when he was not on site in order to ensure that relevant grouting parameters and work procedures had been maintained. A site diary of Vanessa Bellardinelli was admitted in evidence under the Civil Evidence Act. Ms. Bellardinelli was a geotechnical engineer, qualified in Italy as a civil engineer. She was site engineer for the quay wall strengthening works. Her site diary entry for 25 th June 2002 stated: VHP grouting activities were finally started. First column to be drilled is the one near bollard 253. Problems encountered with the spoil. At a certain point of the grouting, the spoil doesn't come out any more, but, of contrary, it's suck back in. Her diary entry for 26 th June 2002 includes the following: The second column has been started. There are still "problems" about the slurry. No spoil is coming out. Where is it going? On 28 th June she wrote: Nico [11] shows me the VHP logs. They report an average of 30 tons for column. Nico explains that usually the average expected to be needed for each column is 15 tons. Other 15 tons are expected to be the result spoil. In the first two columns made until now almost no spoil came out and all the 30 tons went straight into the column. Where do they go? The divers have checked several times but nothing had been seen. The reference to the divers is a reference to checking whether grout had emerged on the seaward side of the sheet piles through tie holes. On 30 th June she wrote: 2 columns were made. No spoil at all for the first one. And on 19 th July: Discussion with Koen. He explains that the point -8m was considered the most critical point at which the highest grout pressure occurs. This point is positioned (located) approximately at -2m from the boundary between the gravel layer and the Bracklesham Beds layer (into the BB). The theory is that the pressure builds up into the BB strata and it is caused by the fact that the gravel, which should work as a filter, after a certain point is obstructed by the grout and it forms a cap on the upcoming flux of grout. The reference to Koen is a reference to Mr. van der Eecken. On 13 th July 2002 Mr. van der Eecken sent an e-mail to Marc Voorhuis, Bob Michta and Eric le Moine, with a copy to his manager Lucas Bols, asking them to take certain action and foreshadowing two alternative proposals of action to deal with the problem, the second of which was in the event adopted. The relevant passage for present purposes is this: …..we think that there are only two solutions to doing the VHP which will keep damage to the existing quay (sheet piles and tie rods) to a minimum. These proposals are based on the fact that the problem is caused by the presence of the gravel layer which acts as a filter while the VHP grouting is being done, resulting in blockages and pressure gradually builds up, even with smos return. Mr. van der Eecken explained in oral evidence that smos meant effluent. He was cross-examined about those matters. A statement of Mr. Bracegirdle (report, paragraph 9.3.7) was put to him. The statement was It is evident the blockage of the annulus and loss of spoil return to the surface took place throughout the works. Mr. van der Eecken was asked (day 3, page 127) whether he agreed with that statement. He replied: I'm not too sure about the first blockage of the annulus; loss of spoil return, that was, that happened frequently on the site, but the blockage of the annulus, I don't know where he got that information from, where is that factual information where he based himself upon. Q. Are you saying that you were unaware that there were problems with the blockage of the annulus? A. No. Q. Because there were, weren't there? A. No, I don't see, where does he state – what he is doing, he thinks when you have loss of spoil return, there is blockage of the annulus, but there is no factual evidence of that. I told you, the loss of spoil, of grout return, was most presumably because we were filling the bulge and that was the reason, not the blockage, because also on the Lutz registration there is no indication of any such blockage. Earlier (day 3, page 112) Mr. van der Eecken's e-mail of 13 th July 2002 was put to him. The following exchange took place (day 3, page 113): Q. The blockages caused by the gravel layer, that caused problems with the pumps? A. No, no, no, the blockages you are seeing here has nothing to do with pumps. Q. I thought there was a problem with the gravel layer acting on the boundary between the gravel layer and the [inaudible] face layer, and pressure builds up into the [Bracklesham] beds strata and that— A. That was our understanding of the cause of the problem, because anyhow, there were no – because we checked the wall over its entire length, we didn't find any cracking outside the areas where we had been doing VHP grouting. So there was inevitably a link between VHP grouting and the cracking, which at that time we also had concern about steel quality…. Our understanding was, we are building up a kind of pressure behind the wall which cannot be taken by the wall, so in order to reduce the wall [sic: pressure?] we can do a multiple stage, because we were thinking, okay, the pressure is build up because the gravel layer is blocking the effluent going upwards – even though you have effluent going upwards, you still have a slight increase of pressure and that pressure must be sufficient to damage the wall. That was our understanding at that time and that's why we proposed two solutions in order to mitigate the effect. In spite of Mr. van der Eecken's subsequent denial quoted above (day 3, page 127) I am satisfied that there were, and continued to be, occasions of blockage of the effluent by gravel. I have not been shown, let alone offered an explanation, of the Lutz data claimed to be relevant. Mr. Bracegirdle's statement was correct. Whether blockage was partial or complete, while the jetting process continued the sheet pile wall would have been subjected to higher than hydrostatic pressures of grout wherever (if at all at the material time) liquid grout was in contact with the back of the wall and was liquid continuously between there and the jetting monitor. Partial blockage could raise the excess of pressure over hydrostatic substantially above the extra head of 1 metre of grout that I have mentioned above. Total blockage could raise it to VHP level. To avoid that, the operatives would have had to react quickly to loss of grout return caused by obstructions in the annulus. The higher the pressure of the grout on the back of the sheet pile wall, the less the area of wall that would have to be covered by liquid grout to cause plastic hinges to form in the wall, other things being equal. The amount of wall that would have to be exposed to the grout pressure would also depend on the location of the exposure. It would also depend upon the magnitude of the pre-existing bending moments in the wall. The greater that magnitude, the less the area required, again, other things being equal. Whilst I am not satisfied that there were occasions when the full pressure (i.e. VHP) of the grouting process was applied to the back of the sheet pile wall, I am well satisfied that there were numerous occasions when the hydrostatic pressure was applied over substantial areas. The curvature of the sheet piles before the works were carried out was greater in some parts of the wall than others. Yet the incidence of the damage that occurred was widespread, not confined to limited areas along the length of the wall. No correlation between the locations of the damage and the places of higher curvature has been suggested. I should answer a particular rhetorical question raised by Mr. Williamson. It was this. The finite element analysis was said to support a conclusion that merely 50 per cent. of the grout loading was sufficient to cause an immediate 1 metre displacement of the wall. That conclusion was in no way qualified. If 50 per cent. of the grout loading instantaneously could cause such a massive displacement (1 metre), why did not 100 per cent. of the grout loading, applied repeatedly and over long periods, cause the wall to fail, or, indeed, cause anything like this displacement? My first point in answer is that the figure of 50 per cent. relates to the excess over the geostatic pressure, and is more than 50 per cent. of the hydrostatic pressure of the grout. However, that does not meet the gravamen of the question. The second point is that Mr. Bracegirdle made it clear that the analysis had been carried out in two dimensions (his report, appendix 16, first paragraph). That was a relevant qualification. The question of the extent to which support to the section of the wall in question was afforded by neighbouring sections was not addressed. Third, 100 per cent. of the grout loading was not necessarily ever applied to the wall: the precise area of the wall covered does not appear. Finally, 1 metre displacement and more did occur at many places, as the Netsurvey of 2004 clearly shows. Mr. Williamson submitted that Mr. Bracegirdle sought to evade the question, and he suggested to Mr. Bracegirdle that he was fencing. On the contrary, I have the impression that Mr. Bracegirdle was doing his best to answer the relevant questions and was not fencing. I am satisfied that the normal grouting process was perfectly capable of acting on the sheet pile wall if the latter had an existing bending moment of about 760kNm/m in such a way as to cause it to form plastic hinges. A priori, the grouting process would of course have been more likely to make the wall bulge out if the wall had already been in a state where it had, or was close to having, plastic hinges. The greater the initial bending moment, the less the increment required to reach plasticity. The jet grouting process for strengthening quay walls is a novel process. It was, however, used four times before it was used at Southampton. Haecon have mentioned the following designs which include deepening and strengthening of steel sheet pile quay walls by use of VHP columns and ground anchors: Marshall dock, Antwerp (1999); Quay walls nos. 7 to 9, Klaipeda (1997) Quay walls nos. 5 and 6, Klaipeda (under design); Berths nos. 3 and 4, Plipdeco (Point Lisas), Trinidad (2001). There is no evidence of the details of the works involved in those four contracts, nor of the outcome, which I am prepared to assume was successful in the case of the three completed works. But in my judgment, the success of the previous contracts is only weak evidence that the quay wall of berth 205 was overstressed in bending before the works were carried out. Since berths 204 and 205 must be taken to have been in the same condition before the works were carried out, I conclude, on the whole of the evidence, that berth 204 is and was at the material time in the condition described in paragraph 81 above. That is to say, I reach the same conclusion as is expressed provisionally in that paragraph. The experts' agreement item 17 that the state of berth 205 before the grouting was carried out was the same as that of berth 204 must mean that the two states were comparable in this sense, that the states of bending of the two quays were similar. Thus the reference in paragraph 81 to X1, X2 and X3 must be read, in relation to berth 205, as a reference to three unidentified lengths of about 20m each. I am also satisfied that the state of the sheet pile wall in berth 205 in the locations corresponding to the vicinities of X1, X2 and X3 would, standing alone, be wholly insufficient to account for the widespread cracking and bulging that took place during the grouting process. Foreseeability Amended clause 11 of the contract provides as follows: The Contractor shall be deemed to have inspected and examined the Site and its surroundings and information available in connection therewith and to have satisfied himself before submitting his tender as to: (a) the form and nature of the ground and sub-soil and hydrological conditions; (b) the form and nature of the site; (c) the extent and nature of the Works and materials necessary for the completion of the Works; (d) the means of communication with and access to the site; (e) the accommodation he may require; And in general to have obtained for himself all necessary information as to risks, contingencies and all other circumstances influencing or affecting his tender. The Contractor shall be deemed to have based his tender on his own inspection and examination as aforesaid and on all information whether obtainable by him or made available by the Employer and shall be deemed to have satisfied himself before submitting his tender as to the correctness and sufficiency of the rates and prices submitted by him in the priced Bill of Quantities which shall cover all his obligations under the Contract. The defendants contended that the sheet pile wall was, before the grouting works were carried out, overstressed in bending by reason of three matters: (1) That the H-piles were close behind the sheet pile wall; (2) that the H-piles had been driven after the sheet pile wall had been driven; and (3) that the relieving platform was not 100 per cent. effective. They contended that such overstress was not reasonably foreseeable by an experienced contractor. As to point (1), it is not suggested that the defendants were unaware of the closeness at the time of tendering. The tender documents included a drawing which showed the relevant dimensions. Mr. O'Brien gave evidence (day 26, pages 157 to 162) that the closeness of the H-piles to the sheet pile wall would give rise to extra stresses in the sheet pile wall that were, at the time of tendering and indeed now, incalculable. The combination of closeness and very strong ground was unusual. He accepted that a contractor would know that extra stresses would be imposed. As to point (2), URS Corporation Limited ("URS") was instructed by HSS to provide an opinion on the cause of cracking of the sheet pile wall. In its report, URS stated that the only records describing the construction of the berth were brief descriptions contained in two articles published in the July 1971 and November 1972 editions of the Dock and Harbour Authority. URS appear to have deduced from the information contained in those articles an "envisaged" construction sequence that the H-piles were driven after the sheet pile wall was driven. The archive of ABP contains drawings 6573/P2E and P11A of the contractor Keir which show details of the piling frame for the relieving platform bearing piles (i.e., the H-piles) resting on the sheet piles and the landward rail beam for berth 201. Thus it is clear from that drawing that in the case of berth 201 the sheet piles must have been driven before the H-piles were driven. A similar arrangement was in fact used for the construction of berths 202 to 205. If Haecon and HSS had not already been on enquiry, that archive drawing, if they had seen it, would have put them on enquiry whether berth 205 had been constructed in the same way. If HSS or Haecon had had any reason to suppose, or had supposed, at the time of design or tendering that the order of driving the sheet piles and the H-piles was relevant, they could perfectly well have asked ABP. There is no evidence that they did so. The combined effect of points (1) and (2) appears to have been included in Mr. O'Brien's figure of 10 to 25 per cent. as his estimation of the increase in bending moment in the sheet pile wall caused by his estimated increase in the value of (K0) from 1.5 to 2.5. I do not regard such an increase as beyond the foresight of an experienced contractor. Point (3) is that the relieving platform was unforeseeably not 100 per cent. effective. It is implicit in my finding in paragraph 126 above that I accept that the relieving platform was not 100 per cent. effective, that is, that it did not relieve the soil at levels above the toe of the sheet pile wall of the whole load on the platform. Mr Reid said this about the relieving platform in his report, paragraphs 32 and 36: When the soils beneath the relieving slab are relatively incompressible.....there is an interaction between soil and structure that inevitably results in a portion of the vertical load being transferred directly between the slab and the underlying soil. The remainder of the load that is transmitted by the slab to the piles is not, however, automatically transmitted to a depth that relieves the wall. This is due to the fact that competent soils surround the pile shaft and will inevitably attract load from the pile as it shortens under compressive load. While surcharge of the soil underlying the relieving slab is clearly inevitable prior to dredging, the mechanism of load transfer after the wall is allowed to deform is less apparent...... I conclude from that that there is nothing unforeseeable about a relieving platform not being 100 per cent effective. Nor was there anything unknown to the defendants about the dimensions of the system. The tender documents included a drawing which was put to Mr. O'Brien (day 26, page 168): Q. This provides all the relevant dimensions of the relieving slab and associated elements of the structure. There is nothing missing here, is there? A. Yes, it shows the dimensions. I take that answer to mean that there was nothing missing from the drawing. Of course it did not show soil parameters. I am not satisfied that there was anything unusual or unforeseeable about the relieving platform or about the degree of its effectiveness. In paragraph 13 of Haecon's defence to the part 20 claim, it is pleaded that Haecon undertook a structural analysis of the existing quay wall based on voluminous information which was listed in paragraph 14. Mr. Bowdery put to Mr. O'Brien paragraph 15, which said: Accordingly, sufficient and accurate information was available to determine the properties of the soil acting on the steel sheet pile wall and for the purpose of designing the strengthening works. Mr. O'Brien gave this evidence (day 26, page 171): Q. Can we focus on what Haecon should have done when they designed the strengthening works? Do you agree they had sufficient and accurate information to determine the properties of soil acting on the sheet pile wall – full stop? A. For the properties of the soil in terms of as I have said strength and the thickness of the layers and to assess stability, they had enough information. In terms of doing routine calculations which don't depend on soil stiffness, I think they had enough information, but in terms of calculating the state of stress in a sheet pile wall, which is largely linked to matters such as H-pile driving and the like, I don't think they did have enough information, because that is really a different level of complexity altogether, which is not part and parcel of normal design, so it is a completely different type of calculation, which is not routine design at all. Mr. O'Brien said that it was not possible to calculate the state of stress in the sheet pile wall without high quality ground investigation data which are not available even to-day. His own calculations were, he said, really a sensitivity study. In section 8.6 of his report, Mr O' Brien said that in order to carry out sophisticated analysis of the interaction between the soil and the structure, consistent with the complexity of the structure, it would be necessary to have appropriate input parameters for the analysis, such as (in the case of the soil) the degree of non-linearity between small strains (0.01 %) and large strains (1.0%), the degree of anisotropy of the stiffness and of the strength, and the relationship between strength and mean effective stress. For the Bracklesham Beds, which are not well researched relative to London Clays, it would be particularly important to have such data if geotechnical analysis was to be reliable for design purposes. The relevance and importance of that kind of information is now well known, and Mr. O'Brien cited two papers, one published in 1989 and the other in 1992. I conclude that the relevance and importance of that kind of information was well known at the time of issue of the tender. I am satisfied that Haecon was right in saying that it had sufficient information on the properties of the soil for the purpose of designing the strengthening works. But it went on to plead (ib., paragraphs 20, 21 and 22): Haecon was not informed of the history of the original as built design, amendments to the original design,.....problems with construction and the remedial works to berths 204, 205 and 206..... ABP retained a substantial volume of information regarding all of the above matters but failed to disclose it or to inform Haecon of its existence..... ABP are to be taken to have known the information that was in its possession and the significance of failing to reveal it to Haecon at the pre-tender and tender stage or at any time thereafter. Had Haecon been informed of the as-built design, method of construction and construction of the berths it would have advised ABP and HSS that further opening up works and a thorough investigation of the as built design should be undertaken to ensure that berths 204, 205 and 206 as built accurately reflected the design information in ABP's possession. Haecon avers that had it been supplied with the information in the possession of ABP and a thorough investigation had been undertaken, it would have produced a different design for the strengthening works which would have been significantly more expensive to construct than the design that was constructed. Access to ABP's archive Mr. Gary Brown was the Engineering Manager at the Port of Southampton for ABP. He gave evidence (day 2, page 120) which I accept that at a meeting he had with the pre-tenderers in July 2001 he said that ABP had significant archive information. I am satisfied on the evidence of Mr. Van der Eecken that HSS was not refused any information that it asked for from documents in the archive. I accept his evidence that there was no restraint on what HSS could investigate as part of the site investigation. But none of the investigations that HSS performed, he said, gave any indication of a need to carry out any more detailed investigation. The following exchange took place during the course of his cross-examination by Mr. Bowdery: Q. Do you not think that you should have included a sum or at least foreseen that there may be a risk that the structure, after 30 years, might not be suitable for your works? A. We had latent defect insurance covering the sum. We had an independent checker, a latent defect insurance, so we had a sum, an insurance, the size of the other construction insurances to cover that risk. Q. So the answer to my question is that you not only foresaw that risk but you covered it by insurance? A. That is the usual way we are dealing with this kind of projects…. The designer, Mr. Joep Wijffels, gave evidence before me. He explained that when the tender was submitted the design was a concept design, not a detailed design. He said that the design period (by which he must have meant the period during which the concept design was prepared) stopped at the moment when the contract was signed. During that period, he said, he was not refused access to any information. Haecon obtained the information it needed from HSS and from the tender documents. In his witness statement, Mr. Wijffels said I cannot recall the exact dates but practically every meeting I attended during the execution of the works I asked for documents and information relating to the existing structure. Our requests produced a variety of responses including that the information was no longer available, that ABP no longer knew of its whereabouts, or that the person who had been involved with storing and archiving the documents no longer worked for ABP. In cross-examination (day 5, page 18), Mr. Bowdery asked him Q. What further work was required to actually translate this concept design into a detailed design? A. To enable the contractor to do the orders and to execute the work. So they need to have the details for the construction. Q. Did you require any further information to translate the concept design into a detailed design? A. Yes. Q. What information was that? A. Well, on the steel bars and details for the construction, related to the equipment of the contractor. Q. So the information you needed was from HSS? A. Yes. Q. You did not require any further information from ABP? A. No. I am satisfied that ABP did not withhold any information relevant to the design or construction of the grout wall from any of the defendants. Mr. Reid said in his report at paragraph 46: I would not consider that a tendering contractor is obliged to guess what information may be available within a client's archives. Nor do I consider that it is a contractor's responsibility to carry out a search for the same. To suggest otherwise appears to me to be a totally impractical concept and one that I have not previously encountered. He was cross-examined about that. He said (day 21, pages 121, 122) with reference to a particular archive document: ..... I'm not sure that if the archives had been opened, that there would have been time to go and find that...... With hindsight, there are a number of things in the archive that could have put Haecon on notice. What I'm saying in this paragraph is that in the real world, with the data bundle that there is there, I'm not convinced they would have seen it. The survey Section 101-07A of the specification of the contract contained the following provision: The Contractor is to undertake a condition survey of the quay wall prior to commencement of the Works to satisfy himself of the integrity of the existing structure and its suitability to be strengthened by his proposed method. Mr. Williamson submitted that a survey which would have been sufficient to show the condition of berth 205 at the time of tendering would not ordinarily have been undertaken by an experienced contractor acting reasonably without some indication or evidence that the wall was suffering from any distress. The cost of determining the profiles of the wall as has now been done would have been far more than was envisaged by ABP itself. ABP had allowed a figure of only £5000 for such a survey. HSS's method statement for survey of the quay wall provided for a photographic survey at the surface and for ultrasonic measurements to establish thicknesses of steel, and hence to establish the existence and extent of any corrosion. I accept evidence of Mr. Brown that at a pre-tender meeting of tenderers he told the tenderers that they would need to conduct their own sheet pile wall survey and that there would be a contribution by ABP of £5000 for that survey. That figure was determined by ABP. Mr. Wijffels said in his witness statement (paragraph 34) that he understood the expression "Condition survey" to mean a corrosion survey. That understanding was, he believed, consistent with the amount of money allocated for surveys, the time available and the commercial reality that contractors are most unlikely to incur the cost of surveys prior to acceptance of their tender. In my judgment, it was for the tenderers to decide whether and in what way to survey the wall. If they considered such a survey prohibitively expensive or that the prospect of the work was too risky, they would not tender. Otherwise, they would allow an appropriate amount in their tender. As I have found, most of the wall was not suffering distress; and I am not satisfied that any of it was. Decision on clause 12 claim My conclusion as to the initial state of the quay wall of berth 205 (see paragraphs 158 and 159) means that the sheet pile wall was not in the condition complained of. Nor were the matters upon which the defendants rely as causing the alleged condition unforeseeable by an experienced contractor. Thus this claim fails. Fitness for Purpose Meaning of fitness for purpose The relevant terms of the contract are: Condition 22 : The works shall be designed to be fit for the purpose as described within the Design Parameters and Requirements Employer's Requirements. The Design Parameters and Requirements state in paragraph 301 01, under the heading Employer's Requirements – Generally: The Employer's Requirements are that the Contractor shall develop and complete the design for the Works as set out in this Section of the Specification and thereafter that the Contractor shall construct, complete and rectify defects in the Works all in compliance with the design and the Contract. The Works shall be fit for the intended purpose of strengthening the quay wall to permit deepening of the berth pocket by subsequent dredging together with the strengthening of a short length of the landward and seaward crane beams by provision of additional support and shall be designed and constructed in such a manner as to maintain the integrity and stability of the existing quay wall and crane beam at all times during and on completion of the Works. The Works shall comprise the design and construction of strengthening works to the existing quay wall to permit the subsequent deepening of the berth pocket (by others)….. It is common ground that the strengthening should allow for a design dredge depth of -16mCD and an over-dredge allowance of a further 0.50m. Mr. Jones submitted that the fitness for purpose obligation must be read together with and/or subservient to clause 12 [12] of the contract conditions. Thus the interpretation to be given to "fitness for purpose" in order to give business efficacy to the agreement was that it must take account of the relevant information that was not in the possession or knowledge of Haecon. I treat this submission as also a submission of HSS, where the lack of information in the possession of HSS would be relevant. All that was necessary, submitted Mr. Jones, was for Haecon to raise the assertion that evidence of the physical condition of the quay wall was not foreseen by it or by HSS. Thereafter, ABP must prove that such information was either irrelevant, made available or that it was foreseeable. Mr. Jones submitted that to give any other interpretation to the contractual term would fail to give it business efficacy since it would be analogous to a warranty in respect of latent defects, and in particular the works undertaken in the 1970s, and there would otherwise be an untenable tension between the fitness for purpose obligation and the clause 12 condition. I reject Mr. Jones's submission. The contract term is perfectly clear that the works have to be fit for the purpose. Clause 12 enables the contractor to claim extra payment in certain circumstances but does not relieve him from the obligation, or modify the obligation, as to the fitness for purpose of the works. The burden remains on the contractor to prove his clause 12 claim, not on ABP. The burden does, of course, lie upon ABP to prove that the works are not fit for the purpose. Fitness for purpose Mr. Reid carried out a number of calculations relevant to the question whether the grout wall was fit for its purpose. He wrote in his report that the calculations had been carried out by WMR Consultancy to demonstrate that the wall system as designed by Haecon at Berth 205 was adequate to carry the loading that would have acted on it provided the relieving slab and the supporting piles [i.e. the H-piles] were capable of shielding the wall from overburden pressure. The calculations were based on the assumption that the relieving slab and the support H-piles transmitted the vertical load of the overburden and superimposed load above it to a level beneath where the sheet pile wall would be subject to lateral forces from that vertical loading. The first set of calculations considered the stability of the sheet pile wall under the hydrostatic pressure of liquid grout. Of that set, the first subset related to the case where the whole grout column was executed in one pass. The second represented the upper stage of the two-pass system, and the third, the lower stage of that system. In each such subset of calculations it was assumed that the grout pressure acted over the whole relevant height of the wall. In the first subset, the relevant height of the wall was from 0mCD to -14.8mCD; in the second subset, from 0mCD to -7mCD; and in the third subset, from -6mCD to -14.8mCD. It was assumed that the sheet pile wall was fixed at and below -14.8mCD. It was also assumed that the grout pressure acted on the sheet pile wall over the diameter (0.9m) of one grout column but was resisted by a width of wall equivalent to three such diameters. The maximum bending moments in the sheet pile wall in all cases appeared at relieving platform level, and came to 1800kNm, 1100kNm and 1480kNm respectively. Those figures are equivalent to 667kNm/m, 404kNm/m and 548kNm/m respectively. The calculations also showed shear forces and deflections. The deflections were only a few millimetres. Mr. Reid concluded that on those assumptions the sheet pile wall would have had a significant reserve of strength amounting to about 100 per cent., i.e. a factor of safety of 2. The second and third sets of calculations, which related to serviceability loading of the sheet pile wall alone, I can pass over. The fourth set related to ultimate loading of the sheet pile wall alone, dredged to -16.5mCD. That set contained two cases, corresponding to subclauses (a) and (b) of BS 8002:1994 clause 3.2.5. That clause provided that a retaining wall should be designed to be in equilibrium mobilizing a soil strength of the lesser of two quantities, the one defined in subclause (a), and the other defined in subclause (b). Those calculations showed that although the sheet piles could in both cases carry the applied loading, the embedment in front of the wall was deficient by 270kN in case (a), and by 315kN in case (b). Those calculations related to a width of wall of 1.7m. The conclusion was drawn that each grout column would be required to deliver an ultimate shear resistance of 315kN. That was based on the design spacing of 1.7m. The worst case was a spacing of 2.7m, requiring a shear resistance of 500kN. The conclusion expressed in the report included the following: The only credible location along the grout column [where] collapse could occur due to shear rupture is at the toe of the sheet piles at a level of -20.71. Such a collapse is not possible due to the very high dowel action available from each of the reinforcing tubes [sc., in the grout columns]. A calculation followed which showed that a shear resistance of 525kN was available from that dowel action. A final calculation showed that the H-piles could themselves each provide a shear resistance of 506kN. Mr. Reid explained in the text of his report (section D, paragraph 37): I assess that each [H-] pile is capable of contributing a shear resistance across the failure plane of about 500kN. There are eight piles in each row but I have only considered the influence of the 3 most effective. These are the 3 that are located closest to the wall. The spacing of the rows of piles along the wall is 2.54 metres and therefore I have assessed the potential contribution of lateral resistance from this source to amount to 590kN per metre of wall. It follows that the shear resistance which Mr. Reid calculated as being available from the H-piles was 1003kN for every 1.7m width of the wall. That was more than three times the requirement of 315kN. The conclusions expressed by Mr. Reid at the end of his calculations included these: 3. At ultimate loading conditions, the lateral force required from the grout piles is assessed to be 315kN. This is comfortably within the capacity of the grout columns when dowel action of 520kN per column is mobilized. 4. The shielding effect of the H [-piles] and their ability to resist the formation of a failure plane at the base of the sheet piles has been assessed at 500kN/m² [sic]. That is 1500kN per bay (2.54m). 5. The strengthened wall is adequate to carry the ultimate loading specified in BS8002:1994. There are no credible collapse modes that generate concern for the adequacy of the system as designed and installed. It is implicit in Mr. Reid's calculations and conclusions that, if the sheet pile wall were in the condition which the defendants were contending was the condition to have been expected when the strengthening works were designed, the strengthening works were unnecessary. The H-piles would have done the job. There is no doubt that he regarded the grout wall and the H-piles as acting mutually independently. He said in evidence in chief (day 17, page 52): I need a support there of 315 kilonewtons which I'm going to supply from the grout column or the H-piles, but I need to get another 315 from somewhere. In my judgment, the H-piles ought not to be relied on to afford lateral support. The cranes run on rails and cannot operate if the rails become out of gauge. Mr. Bracegirdle made the point (day 8, page 28): The shielding effect would be where shear is taken by these H-piles in bending….. In a stabilization of landslides this is a very effective method. I have used this on numerous occasions to stabilize landslides, but of course the problem is here that we have a large Goliath crane sitting on top of the quayside relying on this structure and the …..beam…..staying in place. And once we start to see movement, particularly lateral movement, differential lateral movements between the crane well being the seaward side and that of the landward side, we see problem with the cranes and we are talking about millimetres of movements to cause those cranes to become inoperative. So it is quite a serious business relying on these H-piles. They have vertical loading but their principal aim is to support the relieving platform, not to take lateral loading. So if we start to rely on them [sc., to take lateral loading] we are in serious trouble with the serviceability of the structure. Shearing force Mr. Bracegirdle, using the ICFEP finite element analysis, calculated that the grout wall would be required to withstand a shearing force of 200kN/m at -23mCD. (The relevant graph actually showed a figure of 225kN/m). That calculation was criticized. Mr. Reid calculated that each grout column would be required to deliver an ultimate shear resistance of 315kN. That conclusion appeared at page 13 of the appendix to his expert's report. It was based on the design spacing between the soilcrete columns of 1.7m, and amounted to 185kN/m. Professor Pavlovic said (day 16, page 26), and I accept, that that was very close to the figure calculated by Haecon. Mr. Reid stated that the figure of 315kN had to be increased in line with actual spacing for the worst case of an average spacing of 2.7m. That led to a required shear resistance of 500kN per column. Mr. Reid expressed the conclusion that collapse as a result of shear rupture was not possible in view of the very high dowel action available from the reinforcing tubes. A pair of reinforcing tubes had a shear resistance of 525kN, according to his calculation and his oral evidence in chief (day 17, pages 54 to 57). As he explained, for that shear resistance to be mobilized, the soilcrete itself would have had to suffer a brittle failure, with lateral displacement of the parts of the soilcrete column above and below the failure surface. It is clear from page 14 of the same appendix that the amount of that displacement would be about 157mm. Mr. Reid did not say at that point whether or not the grout column had sufficient shear capacity apart from the dowel effect mentioned above. In my judgment, the wall cannot have been fit for its purpose if in the ultimate limit state the wall can be saved from collapse only by the rupture of the soilcrete columns. BS8110-1:1997 provides by section 2.2.2.1 that a design should satisfy the requirement that that no ultimate limit state is reached by rupture of any section under the worst combination of ultimate loads. That in my judgment is also a requirement of fitness for purpose of the grout wall. In response to a pre-trial question for clarification of his report, Mr. Reid wrote I have [a] estimated a value for the shear capacity of the grout column by using the equations in clause 3.4.5.12 of BS 8110: Part 1 : 1997 and making assessment for the vertical load on the pile together with an additional vertical component that is mobilized when the pile is subject to flexure and cracking is initiated. He then set out his calculation. He first calculated the axial load (compression) in the pile at a depth of -20.7mCD (i.e. at the level of the toe of the sheet piles) due to the self-weight of the pile. That came to 198kN. He then referred to an additional load from the soil surrounding the tension side of the column and said that that came to 149kN. Thus the axial load was the sum of the two, 347kN. He then applied clause 3.4.5.12, using equation 6a, and arrived at a shear capacity which, correcting a trivial error, amounts to 549kN. A controversy arose whether equation 6a or equation 6b given in clause 3.4.5.12 should be used. A rubric in clause 3.4.5.12 states: Where it is considered necessary to avoid shear cracking prior to the ultimate limit state, the shear stress should be limited to the value given by equation 6b. Mr. Reid said that equation 6b was intended to be used when it was sought for cosmetic reasons to avoid cracks, and that it was proper to use equation 6a in the present case. I accept that it is proper to use equation 6a in the present case. In the present case, the results are not very different. If one uses equation 6b and adopts Mr. Reid's figures, one arrives at a figure of 507kN as the shear capacity. Using equation 6a, the calculated value of the shear capacity increases linearly with the axial compression. For zero axial compression, the shear strength calculated in the above manner using equation 6a would amount to 396kN. Mr. Reid produced a diagram in which he asserted that downward frictional forces, mobilized by the cracking of a column, would act on the column. I reject that evidence. It was put to Professor Pavlovic in cross-examination (day 6, page 168) that he had not taken account of those forces in his calculations. He said: Well, I have, I have taken it at the beginning. I've shown they were going the other way….. Instead of going downwards, they were going upwards, and I can make an equal case to say, all you need is small amounts of the pile to move down and the frictional forces will be going up, and that will kill any dead weight. I had this in my, in the beginning of my report, but I was advised that Mr. Bracegirdle, that there would be hardly any – to mobilize any degree of friction up or down, as a soil [mechanics] expert, he considered that several millimetres were required, and this was most unlikely to happen. So I began, if you recall my report, by talking about friction in the piles and this is very important. I have done work on that which is in fact published in Géotechnique , the soil mechanical journal, but the advice of Mr. Bracegirdle was that this was not credible. This is even less credible and I'm sure that Mr. Bracegirdle will comment much more competently than I on this. Q. Well, we will have to see about that, Professor, but can I just understand your evidence, then? For the time being, you are saying that you have disregarded the soil friction forces which are shown by Mr. Reid on the advice of Mr. Bracegirdle? A. No. I am saying that my own forces which were going the opposite way, going upwards, I disregarded on the advice of Mr. Bracegirdle. He said, "Theoretically it made sense but as a soil mechanician, I can tell you that this will not happen and you are being unfair to the structure, you are being punitive", so I didn't do it. Further, Mr. Reid's own diagram suggested that those forces would balance out across the width of the column. It was implicit in Mr. Reid's calculation that the whole weight of the grout pile above the level of -20.7mCD was borne by the soil below that level. That is also implicit in Professor Pavlovic's perhaps reluctant acceptance of Mr. Bracegirdle's opinion that upward frictional forces would not be mobilized. It follows from my rejection of the proposition that downward frictional forces would be mobilized that I reject Mr. Reid's addition of 149kN to the axial load represented by the weight of the pile. If the axial load is simply 198kN, then Mr. Reid's calculation of the shear strength of the grout pile would come to 483kN. A more precise calculation using Mr. Reid's dimensional data would, I think, lead to a figure of 506kN. A fallacy underlying Mr. Reid's calculation is that he has ignored the fact that the British Standard relates to concrete. The British Standard in question, BS 8110-1:1997 provides by clause 3.1.7.2 for a minimum strength of concrete. The actual figure does not appear. In Mr. Reid's calculation, table 3.8 in the standard is used which suggests that a characteristic concrete strength of not less than 25N/mm² is envisaged. Mr. Reid accepted (day 18, page 31) that structural concrete is not expected to have a lower value than that. If the characteristic strength of the soilcrete were taken as 12N/mm², the effect would be to reduce a constant term of 0.45 N/mm² in Mr. Reid's calculation by a factor of 0.783 to 0.35N/mm². That would cause a reduction of the values of the shear strengths of the column calculated by Mr. Reid's method by about 86kN. Mr. Reid said in cross-examination (day 18, page 33): ….. I think Professor Pavlovic has criticized my calculations by saying that I have not taken account of [the characteristic strength] being less than 25, and I don't disagree with that criticism….. Mr. Reid explained the following day (day 19, pages 13, 15, 20) that he took the value of 12 for the characteristic strength of the soilcrete and at the same time he removed the material safety factor of 1.25, and that made very little difference: the result was reduced by a factor of 0.98. (That is, indeed, 0.783 X 1.25). In one sense, Mr. Reid conceded too much, since he applied the characteristic strength factor (0.783) to the shear strength of the column, when it should have been applied only to the constant term (independent of axial stress). On the other hand, I reject his justification for removing the material safety factor. He said he was confident he could get 12 as a strength. But one might say the same of 25 for concrete, and the standard does recommend the use of the factor regardless of that. Professor Pavlovic said that he used equation 6b, but he did not think much of the standards and he did his own calculation. What he said in cross-examination was this (day 6, pages 160 to 165): Q. …..the equation you are using, as I understand it, is equation 6b, not equation 6a? A……Yes. Q. You have not, therefore, used the alternative equation 6a? A. No. ….. Q. …..equation 6b, according to the British Standard, is only appropriate where necessary to avoid shear cracking? A. Yes. Q. Is it your opinion, and if so why, that this is such a case? A. Yes, one should definitely avoid shear cracking, so it is an alternative expression, but can I say this? That…..my use of the British code…..was partly cursory because EC2 is a much better document, a more modern document, but also the British code is much more punitive to shear, although it uses lower factors of safety as well, but if you look at the various plots, it is more punitive. Q. By which you mean the results look a lot worse for the design? A. They do, they do. ….. A. Can I say something? This is a problem again with codes of practice. Nobody wants to have shear cracking, shear cracking means failure, so that shows what sort of code this is. Q. You are saying it is not a code you are very impressed by, is that what you are saying? ….. A. I'm not impressed by the shear clauses of this code, I should say. ….. A…..codes, I believe, are very punitive for shear, including EC2, they are punitive for shear, but this is the only information that designers are given, and this is why I felt it necessary to do calculations outside the code. Because, in seven years of my experience before, designing, something which was failing the code in shear was quite safe. All I'm saying, I think the British Standard is more backward than EC2, it's simply chronological, it was drafted before the EC2. That's all. Q….. If you apply question 6a, rather than equation 6b which you have applied, you would get a much higher shear capacity on the numbers that you have adopted? A. Possibly, possibly. Q. You agree with that? A. Yes. ….. JUDGE HAVERY: It is very easy to carry out the calculation? A. Very easy, very easy indeed. MR. WILLIAMSON: Do you accept, as a matter of [general?] proposition, that equation 6a is likely to produce a better result from the point of view of the design than equation 6b? A. Because I usually use it, I would say yes, although it would depend on the parameters. I would not exclude that the opposite may be true in this particular instance, but I would accept either equation, I have to say that. Q. So to put this point to bed, you would need to see a calculation using your numbers in equation 6a, is that right? A. That's right, the geometry of this particular, that's right. In equation 6a, a factor Vh/M appears. Note 2 states that the value of Vh/M should be taken as not greater than 1. Mr. Reid has, without explanation or challenge, taken Vh/M equal to 1. If one takes that value, it is manifest that equation 6b must give a lesser value than equation 6a if, as assumed here, the column is not in tension. Professor Pavlovic set out two graphs in his report showing shear strength per column as a function of axial load. Each graph showed three curves representing different safety factors; one graph related to EC2 and the other related to BS 8110. The latter showed lower shear strengths than the graph for EC2, and was thus, to use his language, more "punitive". He also set out his calculations. They were all apparently based on the formulae in the standards. The other calculations that he said he performed were not put before the court, nor were their results. As to BS 8110, Professor Pavlovic purported to use equation 6b, but I have been unable to see how he arrived at his result. In EC2, the equations corresponding to equations 6a and 6b are labelled 6.2a and 6.2b respectively. It was 6.2a, not 6.2b, that Professor Pavlovic used. He set out his calculation based on that equation. On the basis of the dimensional data that he used, which were not controversial, I accept that his graph of shear strength as a function of axial load, with no safety factors, is a correct representation of equation 6.2a. With no safety factor, the shear capacity of a column at zero axial load comes out at 474kN. At an axial load of 200kN, it comes out at 504kN. That is close to the figure of 506kN mentioned in paragraph 205 above. Axial load Haecon, in their design, used a figure of 560kN as the axial load on a single column. For the purpose of designing columns strong enough to sustain that compression, that was a proper figure to take. But it cannot be used to calculate the shear strength of a column. Professor Pavlovic said this in cross-examination (day 7, page 52): My position is that one should not be taking into account axial loading. Q…..do you accept, Professor, that there are many reputable engineers who would take a different view to you on that issue? A. On the basis of the information I have for the soilcrete piles and the loading, I would say no. Q. Haecon certainly took a different view? A. Yes, and I then found out that most of the load was transient, it was a crane load, so you have to have a crane on 412 piles all the time, otherwise the thing is going to collapse. And in re-examination he said (day 7, page 144): Q. What axial loads is one considering when looking at VHP piles? A. There are, if one ignores the possibility of friction either going upwards or down…..there would be transient loads from the cranes, for example, applied live loads, service loads, at the top of the platform, and there would also be, of course, the self-weight of the pile. Q. And how should one apply those sort of loads in terms of the design calculations of these VHP piles? A. The transient loads should not be applied because…..they would not be there all the time. It is not usual practice to allow for self-weight loads axially, because they are usually small compared to the axial force you expect to boost the design. In the case of the VHP piles, I have made a calculation and I have checked the calculations of others. They do not make a difference to the conclusion on the shear – on the shear strength, even if one were to allow for the self-weight of the piles, they would be increasing in shear strength but it would not be sufficient to even come close to what is required. On self-weight of the columns as axial load, Mr. Reid said this (day 17, pages 66, 68): …..there is an axial load on it and….. I don't mean that it is a huge force. I agree that there isn't going to be load from cranes and so on there permanently, but what we do have is the self-weight of the column there permanently. There was a pressure when the grout was fluid and I see nothing that would have relieved that pressure so at any particular point there is the weight of the grout above you to be taken into account. ….. …..in reality, this grout column looks more like Michelin Man; as the layers of soil become more granular and more clayey, its outer surface fluctuates in and out, so we don't have what you would call a smooth surface. The point made by Mr. Reid was this. While the soilcrete was fluid, there would be a hydrostatic pressure at any given level by reason of the weight of fluid above that level. I accept that. When the soilcrete solidified, that pressure would remain because he could see nothing that would relieve it. I am not satisfied that he is wrong about that. When the soilcrete was fluid, the whole weight of a column of fluid vertically overlying the fluid at a particular level would be borne by the fluid at that level. The weight of any of the overlying fluid in the interstices of the soil at the side surface of the column (where the surface "fluctuates in and out") would be borne, at least in part, by the soil underlying the fluid in the interstice. Whether the whole, or only a part, of the weight of the fluid in the interstice would be borne by the soil depends upon the shape of the interstice. When the fluid solidifies it thereby becomes capable of bearing shear stresses. So the weight of the soilcrete right across any given level could be borne by the soil that has been supporting the fluid in the interstice. That would relieve pressure on the soilcrete below that level. The elastic forces thereby mobilized would involve some movement of the soilcrete relative to the surrounding soil, notwithstanding that, on the evidence of Mr. Reid (day 17, page 67), there is no change in volume of the soilcrete on solidifying. There is no evidence of such movement. Professor Pavlovic, reluctantly I think, accepted that proposition. Consequently, I accept Mr. Reid's evidence that the self-weight of the column will be borne as a compressive stress in the column. I conclude that an axial compression of 198kN in each soilcrete column is to be taken into account for the purpose of calculating the shear strength of the columns. Taylor's paper on shear resistance, aggregate interlock etc. Mr. Williamson submitted that in determining the shearing capacity of the grout columns a number of factors must be taken into account which Professor Pavlovic had ignored. He invited my attention to a paper by Professor H. P. J. Taylor published in 1970. The conclusion expressed in that paper was that there were three factors which contributed to the shear forces in a reinforced concrete beam, unreinforced in shear. Such a beam, or, in this case, column, is, I accept, analogous to the columns of the grout wall in that the latter have tensile reinforcement but are not reinforced in shear. The three factors were the shear forces carried by the compression zone, those carried by aggregate interlock and those carried by dowel action of the main tensile reinforcement. The compression zone is that part of the cross-section that is in compression as a result of bending. Aggregate interlock comes into play only when the beam is cracked. Aggregate interlock refers to the frictional force between lumps of aggregate in the concrete as they rub together when the parts of the beam separate as the crack widens. Whether the phenomenon of aggregate interlock occurs in soilcrete was spoken to by Professor Pavlovic. He had this to say about it (day 7, pages 59, 60): Q….. When you have reached the conclusions that you have on shear capacity in this case…. You have taken no account of aggregate interlock? A. Sorry, where is the aggregate in the soilcrete? I have to ask you that because I am not aware that soil has aggregates. Q. Just answer the question: you have taken no account of that as a factor? A. Of course not. Mr. Reid gave evidence (day 19, pages 66 to 68) that there are two kinds of dowel action: small strain dowel action, such as was referred to in Taylor's paper, and dowel action leading to plastic hinges, as referred to by Mr. Reid in his calculation in the appendix to his report. In my judgment, the difference is simply one of degree. The argument is of little importance unless it be said that the two effects are additive in any given case. I do not accept that they are additive. Fitness; factor of safety The question whether the VHP wall was unfit for its purpose is distinct from the question whether the design of the wall complied with standards. Where, as here, the wall is still standing but not being used as intended, the criterion of fitness for purpose in my judgment is that there should not be an unacceptable risk of failure or unserviceability in the conditions contemplated, viz. the berth having been dredged to -16mCD with an allowance for a further 0.5m. Unfitness is simply non-compliance with that criterion. For the claimant and the part 20 claimant to establish unfitness I have to be satisfied on a balance of probabilities not that the wall will (or would) collapse or be unserviceable, but that the criterion I have mentioned is not satisfied. That inevitably involves applying probabilities to probabilities. The codes can properly be taken into account in deciding whether the criterion is satisfied. Factors of safety It is common ground that there are two kinds of design check. One check is for serviceability, which broadly speaking is a check that the artefact will not suffer undue displacement or deformation in service. The other check is for the ultimate limit state. That is broadly speaking a check that the artefact can sustain foreseeable loads without collapsing. Those propositions are encapsulated in paragraph 2.3.1 of BS 8110-1:1997, which reads: Well-detailed and properly-erected structures designed by the limit state method will have acceptable probabilities that they will not reach a limit state, i.e. will not become unfit for their purpose by collapse, overturning, buckling (ultimate limit states), deformation, cracking, vibration, etc (serviceability limit states)….. In his report at section 5.1.3, Professor Pavlovic said this: …..it is usual practice to impose safety factors (henceforth F) on loading (F L ) and material properties (F M ). The choice of F largely reflects the statistical uncertainty of the parameters in question. For example, in the case of loading, one broadly distinguishes between permanent or 'dead' loads (where the safety factor is, typically, 1.35) and variable or 'live' loads (for which the factor is increased to 1.5). Clearly, one can be more confident about the accuracy in estimating permanent loads (such as, for instance, the self-weight of a floor) than one can in estimating accurately variable loads (e.g. passing traffic on a bridge) – as the word suggests, variability implies a larger degree of uncertainty (and hence potential risk or even danger). Similarly, one can be more certain about the quality (and hence the likelihood of variability) of a material such as steel than, say, concrete; and, in turn, concrete would be a more reliable material ('reliable' in the sense that the design strength value is more likely to be close to the value actually assumed) than, say, brickwork or soilcrete. Thus, one would expect that these various material qualities will attract the appropriate safety factors which will cater for the likelihood of their strength not being achieved on account of the inherent uncertainty of the product. That passage, in so far as it is a statement of principle, is not controversial. But the following quotation from his report was not common ground between the parties: If the soilcrete piles were made of actual concrete (as assumed in the analysis and design as the closest approximation in terms of available codes of practice…..), the appropriate material factor of safety would be F C (subscript C for concrete) = 1.5. However, the quality of the soilcrete material is much less certain than that of concrete. Here, one is not merely talking about the strength (in the sense that the soil/grout mixture (soilcrete) can never match the strength of the grout itself), but the concern arises on account of the very different degrees of variability for the two materials. The very large scatter of results for the soilcrete samples tested …..far exceeds what one would obtain for ordinary concrete and hence the factor of safety for the material of the piles F S (subscript S for soilcrete) needs to be increased substantially. What is a reasonable F S for soilcrete? No definitive standard is available to provide guidelines on this question, but, for comparison purposes, it is helpful here to refer to the material factors of safety used for masonry (brickwork), which can be of the order of 3 (EC6 ….) and even higher (i.e. around 3.5 …..). Such a high resistance safety factor is due to the fact that, unlike precast concrete and steel, there are significant uncertainties on the material quality of masonry products which, in addition, are built or placed to a lower quality control than steel or concrete. Since soilcrete has at least the same uncertainty as masonry, it follows that it should attract a material safety factor F S which ought to be at least 3. I have reached this conclusion with the assistance of Professor Gulvanessian, who is the Chairman of the Eurocode dealing with safety factors, and who actually recommended an F S of 3.5 since masonry, which he judges to be less variable than soilcrete, can attract material factors around 3.5, as stated above. My choice of F S is also supported by the only doctoral thesis dealing with this subject which I know to be available at present …., as the following quotation from it shows: ' Depending on the type of project, the test program and the function of the jet grouted structure, the use of material factors of 2 to 3 is recommended ' (p.122). Similar views are stated in Tony Bracegirdle's report, where a value of F S of 3.0 to 3.5 is considered appropriate for soilcretes fulfilling a structural role. Further support is provided in his report, which discusses the fact that a material factor of 3.3 is apparently being applied in the design of a jet-grout wall at Amsterdam Central Station as part of the currently on-going new metro development. Therefore, the nominal design value of 12 MPa for the soilcrete material of the piles becomes 12/3 = 4 MPa when the minimum appropriate safety factor for the material is introduced. The above choice of material safety factor of 3 for soilcrete, in which the variability of the material (as exhibited by the tested specimens) confirms it to be less reliable than masonry, is, to my mind, further justified by a number of reasons. First, the actual characteristic design strength of the tested specimens was found to be 7.35 MPa …..which is less than the assumed characteristic design strength of 12 MPa (the latter is still adopted throughout most of this report). Secondly, even such measured strength refers to the upper layers of the soil which will produce the stronger soilcrete, while some of the soil strata below will result in considerably weaker soilcrete…... Thirdly, the way in which each circular soilcrete column is constructed appears to result in a larger percentage of grout being concentrated in the central region, making the latter stronger compared with the edges of the column where the grout mixes with the soil and causes these extreme (i.e. edge) fibres of the composite pile to be weakest even though bending is maximum there ….. Fourthly, the previous effect also weakens the overlapping region of the two columns, on whose strength composite action largely relies. Fifthly, the overlapping zone is further weakened by the disturbance of the soilcrete of the first column as the second column is being formed at the overlap. Sixthly, the central region of each column – shown earlier to be the strongest area as it possesses the highest proportion of grout – is later cored out by drilling, thus removing the strongest material from the column. Finally, the damage caused by the drilling so as to insert the reinforcing steel is likely to further undermine the properties of soilcrete. It must be stressed again that my choice of F S = 3 (already favourable to the soilcrete in view of the design and its execution) is also coupled by a very considerable additional concession to the quality of the actual soilcrete produced since I adopt the original assumed design strength of 12 MPa so as to arrive at an allowable stress of 12/3 = 4 MPa. Had I used the measured characteristic strength of soilcrete of 7.35 MPa (which still represents an over-optimistic strength for the VHP pile as a whole – see Appendix 6), the allowable stress would have been even lower irrespective of the F S chosen. For instance, were one to adopt the lowest value of F S found in the literature, which F S = 2 (see the earlier quotation from the doctoral thesis where it is stated that F S ranges from 2 to 3), and which in my opinion would be inappropriate in the present circumstances, the allowable stress would become 7.35/2 = 3.7 MPa, which is less than the 4 MPa I adopt throughout this report. As pointed out in Appendix 4, the use of the much higher assumed strength (12 MPa) instead of the actual strength (7.35 MPa) already implies a lowering of the chosen material safety factor of 3. Characteristic strength Professor Pavlovic's and Dr. Greenwood's opinions differed as to the compressive strength of the soilcrete. The characteristic strength is the strength below which only 5 per cent. of the soilcrete falls. I give a more precise definition below. In situ samples of soilcrete were taken and their compressive strength was measured. The data were set out in two tables that have been put before me, table 3 in Dr. Greenwood's report and a table in appendix 6 to Professor Pavlovic's report. The latter shows results for 20 samples; the former shows results for 21 samples, including the 20 samples shown by Professor Pavlovic. The figures have been processed, and the corresponding figures in the two tables do not agree. I shall consider first Professor Pavlovic's evidence. There is no standard for soilcrete. Professor Pavlovic referred to British and European standards for concrete. The British standards have not been put before me. The European standards are EN 1992-1-1:2003 (EC2) and a draft standard, prEN 13791:2003. Characteristic strength was defined in Professor Pavlovic's appendix as the strength below which no more than 5% of all the strength test results (whether from standard test or in-situ test) are expected to fall. The test should be carried out on the 28 th day after the concrete is cast. Design characteristic compressive strength, (fc), is the characteristic compressive strength measured by standard tests of cylinder specimens made according to the standard method which forms the specimen by placing, into the standard mould, concrete from the same batch as the one used for the actual structure as cast. In-situ characteristic compressive strength, (fc,is) is the characteristic strength of a finished concrete structural element (often obtained by coring a cylinder specimen from the member). The concrete sample may be tested as a standard cylinder or as a standard cube specimen. The cube strength is usually 20 per cent. higher than the cylinder strength. The samples tested were not of standard shapes. In each case, measurements of peak stress were made. The in-situ cube strengths of the samples were calculated from the peak stress by use of a formula provided by BS 1881-120:1983. Concrete (and soilcrete) strengthen with age for a period after casting. The in-situ cube strengths were converted to the 28-day in-situ cylinder strengths by use of the following procedure. First, convert the cube strength to cylinder strength by multiplying the former by a shape factor of 0.8 [sic]. Then, following EC2, paragraph 3.1.2(6), divide the result by the factor (ßcc)(t) where (ßcc)(t) = exp{u[1-v(28/t)]}. The parameter u depends upon the strength of the concrete and has been taken by Professor Pavlovic as 0.25. (I have adopted the symbol u in place of the symbol s used by EC2 to avoid confusion with what follows). Whilst Dr. Greenwood does not agree with the professor's results, the figure of 0.25 has not been challenged. The twenty values of 28-day cylinder strengths varied between 5.5N/mm² and 29.1N/mm². One then works out the sample standard deviation, s, of those values. That is the estimated standard deviation of a much larger population of putative measurements. What is sought is the value of the strength below which only 5 per cent. of that putative population would fall. The design characteristic strength (fc) is evaluated according to the following expression given in EN13791 (which has not been put before me): (fc) = {(fm,is) – 1.48s}/0.85 where (fm,is) is the mean value of the 28-day in-situ cylinder strengths. That value was 15.6N/mm². The sample standard deviation s was 6.86 and the design characteristic strength was 5.88N/mm². Professor Pavlovic observed that the distribution of the values of 28-day in-situ cylinder strengths was far from Gaussian and that in consequence it would be unfair to apply the standard statistical calculations which depended upon such a distribution to obtain the characteristic strength since the result would yield lower strength values. In his opinion, the actual characteristic strength should be found from a cumulative frequency plot. By an unexplained method, he thereby arrived at a design characteristic strength (fc) of 7.35N/mm². Dr. Greenwood adopted two approaches, both different from those of Professor Pavlovic. In his first approach, Dr. Greenwood estimated the water/cement ratio of the grout. I have extracted the following description from his report, paragraphs C125 et seq. The compressive strength of the mixed product (sc., the soilcrete) is determined by the water/cement ratio almost irrespective of the type, sizes and proportion of the other inert mineral content, i.e. the soil, provided that the mineral is stronger than the grout. That is broadly true of all cement based grouts. The relationship between the water/cement ratio and compressive strength has been established empirically. One such correlation was shown graphically in his figure 11. Figure 11 was a family of curves of compressive strength of cement grouts against water/cement ratio for different ages of the grout from 1 day to 1 year. Its source was Farmer, I. W., (1970) "Design of cement grouts", an unpublished internal report of the Cementation Company Limited research department. Dr. Greenwood set out in a table four values of unconfined crushing strength of 17, 14, 10 and 10 N/mm². He had obtained those values from the water/cement ratio and the use of Farmer's curves. He said that the calculated results were at best a guide to expectations based on the chosen parameters. Dr. Greenwood's second method was to use the test results referred to above. He said that the strengths of cemented soils and concretes increased rapidly during the first few days after casting. The strength roughly doubles from 3 to 7 days and almost a third again to about 28 days. Thereafter the rate slows: increases to 90 days or more are relatively small. He said that it was usual to compare strengths at the arbitrarily chosen time of 28 days. A more appropriate comparison for mixes of 0.8 to 1.0 water/cement ratios should be made by allowing a one-third increase of strengths recorded for tests at 8 and 13 days and a one-third reduction of test results for samples over 100 days old. It was by that means that he arrived at his 21 results mentioned above, which differ from those calculated by Professor Pavlovic. The mean of Dr. Greenwood's values of "compressive strength 28 day equivalent" was 15.24N/mm². He stated that the characteristic strength was 12.60N/mm². He did not explain how he arrived at that figure. If one applies to Dr. Greenwood's figures the European code method adopted by Professor Pavlovic the characteristic strength comes out at 5.44N/mm². That figure is to be compared with Professor Pavlovic's figure of 5.88N/mm², not with his figure of 7.35N/mm². I arrive at the figure of 5.44N/mm² as follows. To calculate the sample standard deviation, one sums the squares of each individual value, and multiplies that sum by the number of values (in this case, 21). One subtracts from that product the square of the sum of the individual values. One divides that difference by the product of the number of entries and that number minus one (i.e. by 21 X 20). The sample standard deviation is the square root of the answer. It comes to 7.17. The rest of the calculation is as explained above. I cannot accept Dr. Greenwood's figure of 12.60N/mm² as the characteristic strength. Dr. Greenwood went on to consider twelve crushing tests made on samples cast from effluents. His final conclusion on strengths, taking into consideration all test results, was that the evidence suggested strengths somewhat higher than anticipated from calculation based on design assumptions. Since, like concretes, the soil cement comprises natural, inherently variable materials, variations in its strength from place to place or sample to sample are expected. A publication of A. M. Neville, Properties of Concrete (1995), showed a linear relationship based on numerous test results between mean test cube strengths for concrete and their standard deviation. For strengths in the range 10 to 15Mpa (N/mm²) the standard deviation was about 0.6Mpa. Dr. Greenwood said (quite rightly, as indicated above) that there would be a larger standard deviation for soilcrete than was typical for concretes. He concluded that the characteristic strength of the soil cement would be expected to lie between 8 and 12Mpa. Dr. Greenwood did not explicitly contradict Professor Pavlovic's approach. Professor Pavlovic said this in the course of his cross-examination on Dr. Greenwood's conclusion (day 7, pages 71, 72): My conclusion is that Dr. Greenwood is thinking as a soil mechanician, and he looks at the average mean strength of the samples used, and that is perfectly reasonable in a geotechnical situation, where you know there will be weaknesses, local weaknesses, in the soil. But the collapse of the structure is an overall collapse. In other words, the whole soil mass will move towards collapse, and it makes sense then to average values. So some are stronger, some are weaker: mean strength makes perfect sense. That is simply not permissible in a structural situation. In a structural situation, you are obliged to have a 95 per cent. probability confidence….. In other words, you must have test data that will support your assumed characteristic strength to the effect that no more than 5 per cent. of samples tested can fall below it. All structural codes require that, and even the new geotechnical code requires that now….. In fact, EC2 says very clearly, for example, and this is for concrete which is much more reliable than soilcrete. For those who don't want to do any statistics, it is a very simple formula: characteristic strength equals mean strength minus 8. So let's take the mean strength of what we all agree is around 15, and we get a characteristic strength of 7…..you have to define how you are going to use the soilcrete. And if you are going to use it as a soil mass, then the mean strength is correct. If you are going to use it as a structural beam, then you must use the characteristic strength and looking at…..what Dr. Greenwood has done, and what I have done, we roughly have the same data, we have the same conclusions….. Mr. O'Brien considered the question of compressive strength without any explicit reference to characteristic strength. In his report at section 11.2.2 he referred to a population of 34 relevant core samples that had been obtained. The mean ultimate compressive strength of those samples was 15.6Mpa and the standard deviation was 6.6Mpa. Whether that standard deviation was ordinary standard deviation or sample standard deviation does not appear, since the data for the individual samples are not given. Mr. O'Brien said that there were three approaches to selecting a design strength and the associated material factor (to change from a characteristic to a design strength). First, one could assess a "mean" strength, and use a high partial factor on material strength, say 3.0. Second, one could assess a plausible worst credible strength and use a low partial factor on material strength, say 1.0. Third, one could directly consider the statistical variability of the material strength, and derive the design strength from that. He did not evaluate design strengths by either of the first two of those approaches. But since the mean was 15.6, on the first basis the design strength would be 5.2Mpa. On the second basis, a least strength greater than the lowest value of the sample would scarcely be plausible. That lowest value is not stated, but one might take a figure of 5.5Mpa, being the lowest value in the sample used by Professor Pavlovic (Dr. Greenwood's lowest value was 6.1Mpa). On the third basis, referring to research in Japan (Futaki and Tamura, 2002), Mr. O'Brien used a formula that gave a figure of 7.02Mpa. He said that allowing for a modest strength increase due to age effects would probably lead to a 10 per cent. increase in that design strength. He applied no factor of safety to it. He said nothing else about the ages of the samples in the population. If one applied that third method to Dr. Greenwood's statistics, the answer would be 5.9 or 6.1MPa, depending on which definition of standard deviation is intended in the formula. The corresponding figures for Professor Pavlovic's sample would be 6.2 and 6.5. The strength of each sample in Professor Pavlovic's and in Dr. Greenwood's populations was adjusted to 28-day strength. Mr. O'Brien also stated the following: In Japan, there are 26 case histories of full-scale column load tests which gave a compressive strength Qu equal to 0.8 times the mean [ultimate compressive strength], i.e. for berth 205 this would be equivalent [to] a "field" strength of 0.8 X 15.6 = 12.5MN/m² [Mpa]. I do not find that statement helpful in the present context. Finally, Mr. O'Brien said that there were no available compressive strength tests for the soilcrete in the laminated silts/clays and sands below -12mCD. The strength of the soilcrete was likely to reduce below that level by reason of the increase in silt/clay content. For the laminated silts he believed it to be reasonable to assume that the design soilcrete strength was about 60% to 70% of that in the overlying sands, i.e. between about 4.2 and 4.9Mpa. I prefer the approach by way of characteristic strength, supported as it is by published recognized standards (albeit in relation to concrete), but in truth it makes little difference. I conclude that the characteristic strength of the soilcrete was about 7. If one takes Professor Pavlovic's figure of 7.35, then the shear strength of a column under a compressive stress of 198kN falls from the figure of 504kN (see paragraph 210 above) to 433kN. As to the material factor of safety, Professor Pavlovic has, in my judgment, undoubtedly been generous in arriving at a strength of 4MPa after the application of a safety factor. I reach that conclusion disregarding Professor Pavlovic'c points "secondly" to "finally". I take the characteristic strength as 7.35, so the factor of safety that he has applied is effectively 1.84. Since the strength of the material comes into the calculation of shear strength as a cube root, the effect of the factor of safety is to apply a divisor of the cube root of 1.84, namely 1.225, to that element of the shear strength which is independent of the axial compression, namely 403kN, leading to a figure of 329kN. But EC2 recommends the application of a material factor of safety of 1.5 even in the case of concrete (see section 2.4.2.4 of EC2), which is not to be reduced by taking its cube root. I am entirely satisfied that a factor of safety for soilcrete should be taken which is greater than that recommended for concrete, especially in view of Professor Pavlovic's second point quoted above (that the strength of the soilcrete is likely to be less at the depth which is relevant for present purposes). Application of a factor of safety of 1.5 to a shear strength of 403kN yields a shear strength of 268 kN. To that figure 30kN has to be added to allow for the axial compression, making a figure of 298kN. One of Professor Pavlovic's graphs based on the EC2 formula was drawn on the basis of the following three safety factors: first, a material factor of 1.5 appropriate to concrete; second, a material factor of 1.15 for steel reinforcement; and third, a load factor of 1.35. His graph based on those factors showed a shear strength at zero axial stress of 230kN. I do not know how he arrived at that figure. He may have performed a more sophisticated calculation than is provided by the EC2 design code. Whether one divides 474kN by the product of those three factors, or by the product of the last two and the cube root of the first, one does not arrive at 230kN. In the former case, one arrives at a figure of 204kN. In the latter case, one arrives at a figure of 267kN. At this point I am not concerned with load factors, so I shall for the time being ignore the load factor. The strength of the steel reinforcement does not come into the formula since the reinforcement is tensile reinforcement and the formula is given under the heading "Members not requiring design shear reinforcement" in EC2. Thus I shall ignore the steel material factor. On the totality of the evidence I conclude that for the purpose of determining whether the wall is fit for the purpose so far as the shear strength of the columns is concerned, each column must be taken to have a shear strength of 298kN. Mr. Reid's evidence was that on the basis of the design spacing of 1.7m, each column is required to deliver an ultimate shear resistance of 315kN. Mr. Bracegirdle's ICFEP figure was 200kN/m, i.e. 340kN per grout column. Thus I am satisfied that the design of the wall was unfit for its purpose by reason of shear strength. The position as built is this. Mr. Reid referred to a "worst case of an average spacing of 2.7m" (his report, appendix, sheet 13). In such a locality, the required shear strength of a column is 329kN X 2.7/1.7, i.e. 522kN. The experts had agreed to the following proposition: The local average spacing is appropriate when considering the load on a particular pair of columns. A column pair will respond to the loading associated with half the sum of the spans on either side. The reference to a pair of columns is clearly a reference to a single composite column. Mr. Reid changed his view to the following. Consider five adjacent columns, A,B,C,D, and E. Consider the distance from a point half-way between A and B to a point halfway between D and E. Then the shear force on column C is one third of the shear force aggregated over the distance in question. The original agreement was put to him. He said this (day 20, page 64): That's what was agreed at the expert meeting and I appreciate that. But partly when Mr. Wijffels suggested 15 metres and I had another look at it to see what as a designer I would be willing to move to, I came up with that's [sc. the average of three spaces as described above] as far as I would go. Mr. Reid said (day 20, page 60) that if he were signing the design off he would be looking at the spacings individually again. In my judgment, since there can in practice be variations in individual loadings, the agreed view is the better view. I have mentioned the design average spacing of 1.7m. That is between centres of adjacent columns. In fact, the design was such that the columns (i.e., column pairs) were in triplets spaced at 1.68m, 1.29m and 2.11m. Thus considering any given column, the sum of the half distances between it and its two immediate neighbours was 1.485m, 1.700m or 1.895m, depending upon where in the triplet the given column was situated. Mr. Bracegirdle gave evidence (his report, figure 55) which I accept about actual spacings between adjacent columns as built. Taking the "spacing" of a column as the sum of the half distances between it and its immediate neighbours, Mr. Bracegirdle gave evidence, set out in figure 55 of his report, that 37 columns had spacings of 2m or more. There were eight separate groups of five columns where the average spacing, calculated in accordance with Mr Reid's revised method, amounted to 2.3m or more (up to 2.9m). Thus as built the wall was in this respect unfit for its purpose to a greater degree than it was as originally designed (but see paragraph 265 below). The arch Mr. Reid countered the argument that the grout wall would fail in shear by propounding the opinion that the wall would act as an arch against transverse forces. One has to imagine a vertical arch resisting transverse forces in the same way as a horizontal arch resists weights (and hence supports objects) placed or moving upon it. The grout wall is not, of course, in the shape of an arch, but the proposition was that the forces within the material would react to the horizontal forces in the same way as an arch would react to vertical forces. An arch, in order to develop its strength, needs to be firmly fixed at its two ends by forces pointing inwards. In the case of an ordinary arch, those forces would be horizontal. But in the case of the arch postulated in the grout wall, they must be vertical, downwards at the top of the arch and upwards at the bottom of the arch. One objection to this theory is that the curvature of the grout wall changes direction twice (at least) along its length (an opinion held both by Professor Pavlovic and by Mr. Reid). Mr. Reid said (day 17, page 70; day 18, page 63) that in that case three arches would have to be postulated. The arch in the middle would have its notional concavity facing in the opposite direction to those of the other two arches. It is clear that such an arch would not be sufficiently supported at its ends. Mr. Reid put forward another argument in which there was a single arch, the vertical support being provided by the reaction of the seaward soil to a horizontal point load imposed by the landward soil at the level of the bottom of the sheet piles. Any tendency of the wall to move vertically would be resisted by friction arising from that reaction. That would provide the necessary firm fixing, at any rate at one end of the arch. However, the reaction in question would operate principally at the point where the lateral load was imposed. That is analogous to the load on a horizontal arch being imposed at the end of the arch. The effect of the arch as such is lost. Mr. Reid said this during the course of his cross-examination by Mr. Bowdery (day 23, page 52): JUDGE HAVERY:.....you have put the.....horizontal soil response not opposite to P equal[s] 500 kilonewtons, whereas on your immediately past evidence, it would actually be bang opposite that, would it not? A. Yes, but then it would not be [in the] (sic: an) arch. I find Mr. Reid's arch theory to be unconvincing, and I reject it. Slip Circle In the re-amended defence and counterclaim the following plea appears in paragraph 100(2)(f): ….. HSS note that there is no allegation that the alleged design failures would in fact lead to failure of the VHP grout columns…. In his report, Mr. Reid considered two imaginable modes of failure of the sheet pile/grout wall system. The second such mode was by failure of the soil below the dredge level on the seaward side of the sheet pile wall to resist the pressures imposed on the landward side of the wall. In that scenario, he said, the passive wedge in front of the wall yields and allows the sheet piles to move forward. When that occurs, the grout columns are mobilized to provide additional shear resistance and prevent the wedge of soil behind the sheet piles from translating forward. He considered that the grout columns were indeed sufficient for that purpose, because of the dowel effect [13] produced by the tubular reinforcement in the grout wall. He expressed the opinion that there was no credible mechanism that would generate a shear failure at the base of the sheet piles, provided that the sheet pile wall was not highly stressed prior to the implementation of the strengthening works. Mr. Bracegirdle's ICFEP analysis produced the result that the grout wall would be subjected to a shear force of 200kN/m, which corresponds to a shear force of 340kN for each grout column assumed to be spaced at the average design interval of 1.7m. One might suppose that that was evidence that the system would fail if the grout wall could not sustain that shear force. Indeed, Mr. Bracegirdle gave evidence (day 8, page 27) that the VHP wall did not have enough strength to sustain the loads imposed upon it by the proposed dredge to -16.5mCD. I would interpret Mr. Reid's opinion as merely asserting that the grout wall would sustain the shear loading that he calculated would be imposed upon it. There is no suggestion there that there would be no collapse if the grout wall were subjected to a shear force greater than it could sustain. Mr. Bracegirdle introduced an exhibit illustrating a global mode of landslip with the shear surface passing underneath the sheet pile wall. He gave evidence (day 15, page 4) that he had done so in response to a criticism by Mr. Reid, which I take to be the opinion expressed above, that ABP's finite element analysis did not predict a credible failure mechanism. I find Mr. Bracegirdle's further evidence on this point difficult to follow. He appears to have thought (rightly or wrongly) that what was being said was that even if the grout wall failed there would not be a collapse. He said this (cross-examination, day 15, pages 5 to 8), with reference to a diagram in his report showing movements of the soil: ….we were allowing quite a lot of movement in the sheet pile wall and not relying on the grout wall to provide support. So…..we see quite large movements, large soil movements, and the vectors there clearly delineate potential failure surfaces [sc., in the soil]….. ….. Q…..is that assuming an effective grout wall, or a partially effective grout wall? ….. A….. It is assuming an ineffective grout wall. Q. So ineffective, in the sense that it is not carrying any load at all? A. That's correct, yes. Q. It is as if it is not there. Is that the position? A. Yes. Q. Is that a credible scenario, Mr. Bracegirdle, the grout wall is ignored entirely? A. Yes, it's credible in the sense that the residual capacity of the grout wall after it's [failed] due to shear or bond has very little, no significant residual capacity; once it has gone through its brittle failure, then it has very little capacity. I have looked at residual capacity…..of the steel tubes…..and the restraint they might provide if this failure was to take place, and what I found was they were capable of providing about 180 or 190kN/m as a restraining force. But of course that's very small in relation to the very large forces at work here. My first comment is that 180 or 190kN/m is not very small in relation to 200kN/m, though it is certainly less, which is surely sufficient for the argument. Second, if it were the case that there would be no land slip even if the grout wall were not there, I cannot accept that placing the grout wall there would cause it to be subjected to stresses that it could not sustain. If the soil is itself strong enough to resist the tendency to slip, it would not impose any shear strain on the grout wall, apart perhaps from a small effect caused by introducing the wall itself, since the soil would already be stable. Hence there would be no significant shear stress on the wall. Putting the matter another way, if there were no risk of the soil slipping, there would be no point in introducing the grout wall. The key to the confusion is the distinction between a local landslip and a global landslip. Mr. Bracegirdle was considering the latter as a failure mechanism. Mr. O'Brien produced some diagrams in exhibits X38, X41 and X88 showing the results of some stability calculations that he had carried out in response to Mr. Bracegirdle's evidence as to landslip. The calculations related to the potential for a mechanism for failure of the quay below -20.7mCD, i.e. below the bottom of the sheet pile wall and through the grout wall. The mechanism was a land slip along a failure surface which had a cross-section in the form either of the arc of a circle or of arcs of less regular curves roughly circular. In those exhibits Mr. O'Brien carefully considered the soil characteristics at two different locations along the wall, and displayed diagrams with legends showing his calculated factors of safety, all sufficient to comply with standards. He gave no indication of the forces to which the grout wall would be subjected. Indeed, he did not state whether he followed Mr. Bracegirdle in ignoring the effect of the grout wall. However, it seems that he did. The following is an extract from the conclusions he expressed: Allowing for realistic strength of [the] basal sand layer then a global failure mechanism below -20.7mCD is unlikely since even under sub-artesian ground water pressures the factor of safety is greater than one. Under sub-artesian conditions a factor of safety greater than one would be achieved if the basal sand layer mobilises a friction angle greater than about 31 degrees, once the interbedded nature of the overlying soils is allowed for. If Mr. Bracegirdle's simplified soil profile is assumed, then the basal sand strength needs to be about c´=5 and f´=36° for the overall factor of safety to exceed 1.0 under sub-artesian conditions. Mr. Bracegirdle did not agree with all the soil characteristics postulated by Mr. O'Brien. Mr. O'Brien gave this oral evidence (day 26, pages 77 and 78): Q. Mr. O'Brien, it is clear from what you have already said to his Lordship and the exhibits you have gone through with the factors [of] safety …. on the various bases, both in your local failure exhibit, X88 and your earlier one, X41, that you do not believe that there is going to be this global failure at depth, as propounded by Mr. Bracegirdle? A. That's correct. Q. And as his Lordship has noted, do you consider there to be an adequate and even conservative factor of safety with the dredging to be minus 16? A. That's correct. Q. So can I ask you your opinion, please, in the light of that, about whether, if you were putting your name to it, you would or would not, at the present time, authorize deepening? A. If the original construction, the 1970s construction, you know, effectively wished in place and hadn't been affected by original methods of construction, I would be. But I think, for a number of reasons, I'm concerned about the safety of the sheet pile wall as a result of the original method of construction which I think would have led rise to significantly extra stresses and potentially damage to the sheet pile wall. In what follows, I refer to the friction angle (angle of shearing resistance) f (f´ means effective friction angle) and cohesion c (c´ means effective cohesion). Whilst different kinds of soil have to be treated in different ways in calculations, for the purposes of this part of this judgment, the following can be assumed: (1) The greater the angle f, other things being equal, the greater the strength of the soil; (2) the greater the value of c, other things being equal, the greater the strength of the soil. In particular, the greater the angle f, the less the active pressure coefficient and the greater the passive pressure coefficient. Putting the matter simply, the stiffer or stronger the soil, the less likely it is to push a wall over, and the more effective it will be in supporting the wall against being pushed over. In his exhibit X38, Mr. O'Brien plotted a putative profile of a global failure surface through the soil passing through a level of -21mCD at its lowest point. That was drawn to illustrate his proposition that there would be a factor of safety greater than 1.0 if the stratum below -18mCD had a friction angle f greater than 31°. It showed a dredge level of -16.0mCD. Mr. Jones put to Mr. Bracegirdle that if you dredge down another half metre, that might reduce the safety factor by about 3 per cent. He replied: I couldn't say, but it would not be a huge effect, I wouldn't have thought, but it would nevertheless be an effect.....there would be an effect, but I couldn't say whether it would be 3 per cent. or 5 per cent. or..... Q. That would be Mr. O'Brien's comment on it..... Mr. O'Brien can..... give us his answer in due course..... Nothing further was said in evidence about that. The context of the question was a different profile, which also showed a dredge level of -16.0mCD. That profile represented a factor of safety which Mr. O'Brien had calculated as 1.247. Nevertheless, I am satisfied that a figure of at least 3 per cent. is about right in all similar cases. The factor of safety in the X38 profile was expressed to be 1.010. Thus if dredging to the level -16.5mCD were allowed for, the factor of safety would be not more than about 0.98 as calculated by Mr. O'Brien in that instance. Mr. O'Brien, however, considered that the friction angle was greater than 31° at the relevant level. Mr. O'Brien accepted that a f figure of 28° below -18mCD was possible. The following exchange took place (day 27, page 25): Q.....below minus 18CD, they adopt a phi figure of 28 degrees based on their soil investigations? A. That's right and given their, the work they were doing, they have just made a very cautious estimate of the properties of the soils they are dealing with. I would say these are were [sic: worst?] credible strength parameters they have used so, for example, if you [use] guidance in CIRIA 580, I think this would be described as a worse [sic: worst?] credible analysis, which is something that designers do, so it is not uncommon at all but what I was trying to do is make a realistic appraisal of the strength of the soils and then [identify] cautious estimates of what the peak strength would be so my analysis is not the worst credible analysis. I was trying to make an assessment of what I thought the realistic strength would be and then I have also looked at worse [sic: worst?] credible strengths as well and I think those analyses have been supplied to the court. One such analysis described as worst credible has been put before me. It shows a profile where, at the relevant level, the value of f is 34°. The factor of safety is shown as 1.168. On Mr. O'Brien's own evidence, I cannot accept that a f value of 34° at that level can be the worst credible. In the course of his cross-examination by Mr. Bowdery, Mr. O'Brien gave the following evidence (day 27, page 32): Q. .....you told us you have.....carried out further investigations which had produced what you say is a phi-dash value of between 31 and 42 degrees at depth?..... What is the consequence of those increased phi values? Are you saying the strengthening works were never necessary to allow dredging? A. No, because there is different types of failure mechanism which can develop. So for example, the failure mechanism Mr. Reid investigated is different to what I have been looking at in terms of global stability. So what I was trying to look at in terms of global stability was the sort of failure mechanism that Mr. Bracegirdle was postulating, which is this large global, almost circular slip. So that's a different type of mechanism altogether to even the mechanism that I was looking at using Wallop or the mechanism Mr. Reid has been looking at in his calculations. Q. It has been suggested to me that with your revised phi values, you could suggest that the wall is stable regardless of any design issues with the grout columns? A. I think you would need to look at the effect in terms of soil/structure interaction. The jet grout columns would still be needed. What I'm saying is that, in terms of a global failure mechanism, there is very little risk indeed of a global failure mechanism in terms of a large circular slip. I do not see that as being a credible mechanism. Q. Have you carried out any investigation to see if the grout columns were needed with your revised phi value of 42 degrees? A. In terms of the global mechanism which I have looked at which is in the technical notes, the jet grout columns are not called upon significantly at depth. So in terms of the other analyses I have looked at, when you just look at out-of-balance earth pressures, the demand [at] depth is much, much lower. So you know, you can't confuse the different types of deformation and failure mechanisms, they are very, very different and the demands on the wall are very, very different. If you had global failure at great depth there would be large shear demands, but in terms of my analyses it shows that that type of global mechanism couldn't develop. If that type of mechanism doesn't develop then the demands on the wall at depth are much, much lower. I cannot accept that the risk even of a global failure is negligible, having regard to the factor of safety of less than one even for a friction angle of 31°. Mr. Reid explained in his report (appendix, sheet 7) that he was considering a different mechanism. Although it involved consideration of the global surroundings, it was a study of the balance of forces and identified the extent to which the passive resistance in front of the sheet pile wall was deficient in carrying the ultimate loading applied. A diagram of the mode of failure in question appears as the third illustration (bottom left) in figure 3 at page 22 of the British Standard to which Mr. Reid referred (BS 8002:1994). It would certainly involve local movement of the land, but is to be distinguished, as Mr. O'Brien did, from overall global slip. I do not understand Mr. O'Brien's remarks about the demands on the wall being "much lower" as being intended as a comment on, or as disagreement with, Mr. Reid's evidence. If I am wrong, then I prefer the evidence of Mr. Reid. Mr. Reid's calculation was based on sections 3.2.5, 3.3.3.2, 3.3.3.3 and 3.4.2.3 of BS 8002:1994. Mr. Bracegirdle used the ICFEP finite element analysis, run 210. The respective values of f´ and c´ that they adopted appear in the following table. The letters B and R refer to Mr. Bracegirdle and to Mr. Reid respectively. Depth (mCD) Soil (B) Soil (R) f´ (B) c´ (B) f´ (R) c´ (R) +6.25 to +0.6 Dense Alluvium 30° +0.6 or 0 to -1 Loose alluvium Gravel 30° 35° -1 to -3 Silts and clays Gravel 23° 35° -3 to -5 Sands and gravel Gravel 35° 35° -5 to -9 Sands and gravel Bracklesham sand layer 35° 32° -9 to -18.5 or -19 Bracklesham silts and clays Bracklesham laminated silty clay layer 23° 26° Below -18.5 or -19 Bracklesham medium dense sand Bracklesham laminated sandy clay layer 28° 28° Both experts assumed that the relieving platform was fully effective. The ICFEP analysis extended beyond the end of the relieving platform, and accordingly the soil overlying that level (0mCD) was relevant to it. There is evidently a typing error in Mr. Reid's report (appendix, sheet 2) relating to the depths of the lowest layer but it appears that he intended to refer to a depth below -19mCD. The effect of the differences between the values of those parameters as assessed by Mr. Bracegirdle and by Mr. Reid has not been explored in detail. It will be seen that Mr. Bracegirdle and Mr. Reid were in agreement as to the values of f´=28° and c´=5 at the level of maximum shear force in the grout wall. Those same parameters were used by Haecon in their design to represent the soil below -18mCD. I do not accept that they are "worst credible" values, but they are cautious estimates. Professor Pavlovic gave evidence, and I accept, that an action factor of safety of 1.35 should be applied, that is a factor of safety in relation to the imposed forces. No mention has been made of the application of any action factor of safety in the ICFEP analysis. It seems from a remark of Mr. Bracegirdle (day 13, page 12) with reference to a run of the program which has not been put before me that a factor of safety had to be applied to the result, that no factor of safety was included in the program. Nor was any action factor of safety mentioned by Mr. Reid in relation to his calculation. In order to give a reasonable standard of confidence that the wall will withstand shear stresses imposed on it, and thus be fit for its purpose, each column must in my judgment be capable of withstanding a maximum shear force of 340kN. I am satisfied on a balance of probabilities that as designed it was not capable of withstanding such forces and was thus not fit for its purpose. The position was even worse as built. Bond Bond in this connection means the adhesion between the steel reinforcement and the surrounding soilcrete. Professor Pavlovic gave evidence that the bond was insufficient, thereby rendering the grout columns unfit for their purpose. He calculated that the bond stress, by which he meant the stress tending to break the bond, was 1Mpa (i.e. 1N/mm²), whereas Haecon had determined that the columns should have a designed bond strength of only 0.6Mpa, having calculated that the bond stress would not exceed 0.43Mpa. Thus the design was defective. Mr. Williamson submitted, and I accept, that the relevant comparator to the bond stress was not the design bond strength but the actual bond strength. Dr. Greenwood gave evidence that the ultimate compressive strength of the grout was about 20Mpa. I accept that evidence. He applied a rule of thumb (day 25, page 51) that the bond strength was one-tenth of the ultimate compressive strength to arrive at a figure of 2Mpa as the bond strength of the steel reinforcement to the grout. Professor Pavlovic (day 6, page 138) agreed with the figure of one-tenth. In cross-examination he referred to a grout strength of 11.3Mpa which had been determined in experiments carried out by Professor England and Dr. Tsang of Imperial College. Those experiments had been instigated by Professor Pavlovic and were reported as Appendix 6 to his expert's report. The purpose of the experiments was not to determine the bond, but the report in appendix 6 contained a reference to the grout/steel bond being impaired by channels created by flow of water as the grout "bled". Professor Pavlovic's primary point, however, was that the bond between the grout and the soilcrete had not been determined. He said that he was aware of no data to estimate how good a bond results between grout and soilcrete. Mr. Williamson submitted that Dr. Greenwood in his report had shown that the bond strength between grout and soilcrete was 3Mpa, even higher than the steel/grout bond strength. I reject that submission. What Dr. Greenwood was considering in the relevant section of his report was the shear between the overlapping soil-cement (i.e. soilcrete) columns. The two columns constituting a combined double column were grouted at different times. Dr. Greenwood was concerned with the adhesion of the newly grouted column to the column where the grout had gone some way, or all the way, to solidifying. He concluded that for a compressive strength of 15Mpa there would be a bond strength for a "cold" soil cement (i.e. soilcrete) joint of about 3Mpa. If there were any doubt about what Dr. Greenwood was saying, it would be resolved by his oral evidence (day 25, page 69): ….. Yoshida…..did some laboratory testing from cores taken out of the field across the contact surface between two adjacent columns. He derived these data and it's the best information we have….. Q. Why are you bothered with bond? It is all one material, is it not? A. The two columns are done at different times. Some of them are done simply one after another while the fluid is still not hardened properly, so they would coalesce anyway on contact, but some of them are done with an interval of time between them, so you might have a hard column with a fresh column….. It was said that it was no longer the practice to consider local bond stress, because of anchorage. If a reinforcing bar in concrete is bent round at the end it is thereby anchored into the concrete. It takes a large force to pull it out of the concrete against the hook. The force required to pull it out is the anchorage force. The anchorage length or equivalent anchorage length, though not defined in any document that has been put before me, appears to be the length of a straight reinforcing bar that would require the same force to pull it out. Be that as it may, the standards no longer require the calculation of local bond stress because of the general experience of anchorage. Eurocode 2 (EC2; BS EN 1992-1-1:2004), which relates to concrete structures, provides in section 8.4.2, headed Ultimate bond stress, a requirement of principle that the ultimate bond strength shall be sufficient to prevent bond failure. The design value of the ultimate bond stress, (fbd), for ribbed bars may be taken as: (fbd)=2.25(η1)(η2)(fctd), where (fctd) is the design value of the concrete tensile strength. For present purposes I need concern myself only with (η2). Section 8.4.2 provides that (η2) is related to the bar diameter f in the following way: (η2)=1.0 for φ≤ 32mm (η2)=(132-φ )/100 for φ >32mm. It will be seen that the formula does not work for f greater than 132mm. Professor Pavlovic said this (day 6, page 146): Q. You have not considered, have you, anchorage? A. I have. Q. In this section? A. I'm not allowed to consider it if I use EC2. If you look at EC2, EC2 assumes anchorage stresses become zero as soon as you exceed a value of 132mm; this bar is gargantuan: 168. So what the code says, there is a linear decrease in bond strength. At 132mm, it becomes zero. Clearly, it is not going to be zero, but it is a diplomatic way of saying you can't rely on bond for such large bars. So one can't check anchorage using EC2. However, I have also looked at – I agree anchorage is essential – I have looked at both ends of the pile. Anchorage stresses are bond stresses but at the end of the beam, if I can put it that way, local bond stresses are throughout the beam. They're all important, but I would agree anchorage is more important. Why more important? They are more onerous, and if you satisfy anchorage stresses, usually you satisfy bond stresses. That is why the code makes that simplification….. And a little later (day 6, page 150) the following exchange took place: Q. So both the British Standard and EC2 are telling you a number of things, are they not? First of all, all reinforcing bars are to be anchored? A. Yes. Q. Secondly, anchoring is to be achieved by specifying anchorage length? A. Yes. Q. And thirdly, that length is to be checked by the equations in either of these codes? A. That's right. Q. Which are broadly to the same effect? A. That's right. Q. You have not addressed, have you, in your report any of those issues? A. No, because I have looked at EC2, which is a leading code agreed to be by far the most sophisticated code so far, and there is no point in calculating anchorage lengths if the bond strength is zero, and the reason for that is that all these formulas you quote are for normal bars. Bars of 1, 2 – 1, 2, 3cm maximum, not 16. There is no data….there are no tests, the code never envisaged these huge bars. Nobody knows what the bond strength or the anchorage length of these bars is. You talk to a reinforced concrete man and mention a 16.8cm bar and you will see what response: he will be incredulous. Q….. Can we see where we have got to? You accept, as I understand it, that if we were dealing with a bar [of] diameter 130mm or less …. Then you would have to go into all of these calculations? A. I agree. Q. But your position, so that I understand it, is that because the bar is not 130mm but 168mm….it is outside the contemplation of either the British Standard or these codes? A. Yes. In fact, even 130 would not be correct, because by the time you exceed 100mm, using this formula of EC2, that bond strength is so low, it could not be relied upon, nobody is going to go obviously to zero bond strength. I would say, looking at the formula, 80 to 90mm would be the maximum one would go, in order to ensure that one had reasonable bond strength. So 130 is enough obviously, but even that is absolutely huge for ordinary reinforced concrete… ….. A….. So what the code is saying is, as soon as you exceed 100mm, do not rely on bond strength. The formula given in section 8.4.2 of EC2 applies only to ribbed bars. The reinforcement in the grout columns is smooth. That in itself would tend to reduce the bond (evidence of Professor Pavlovic, day 7, page 44). On the other hand there is this evidence of Mr. Reid (day 17, page 81): There is an end plate at the bottom [of the reinforcement], and every 3.1 metres there is a collar to join one bar to the next which roughens it to that extent, but essentially we are talking about plain bars. Mr. Reid said (day 17, page 82) that because of its large surface area (sc., per unit length) the tube was equivalent to a 52mm bar in bond terms. In strength terms, it was equivalent to a 91mm bar. I think Mr. Reid's reasoning must have been as follows. The cross-sectional area of the steel of the tube reinforcement was 6874mm². The diameter of a solid bar of the same cross-sectional area is 93½mm, say 91mm. Applying a diameter of 91mm in the formula in EC2 section 8.4.2 yields a value of (η 2) equal to 0.41. Increasing that figure in proportion to the curved surface areas per unit length (i.e. in proportion to the diameters) of the tube and of the 93½mm bar, that is, in the ratio of 168.3 to 93½, yields a value of (η2) equal to 0.738, giving f=58.2mm, say 52mm. I may be doing Mr. Reid an injustice by attributing that reasoning to him, but it is fallacious. It is true that the bond strength in terms of force, for any given bond strength in terms of stress, will increase in proportion to surface area. But the strength in terms of tension or compression has nothing to do with the bond strength. I reject Mr. Reid's evidence that in bond terms the reinforcement is equivalent to a 50mm bar. I conclude that Professor Pavlovic was right when he said that nobody knows the bond strength or the anchorage length of the reinforcement tubes. It is unnecessary for me to consider an argument of Mr. Williamson that Professor Pavlovic's calculated figure of 1MPa for the bond stress was too high. I am not satisfied that the bond strength of the grout wall as designed was such as to render the grout wall unfit for its purpose. Loss of composite action The VHP columns were installed in pairs with the design intention that they would act as a composite pair, i.e., combined into a single column. Composite action greatly increases the bending moment capacity of the VHP columns. Professor Pavlovic said that loss of composite action results in a clear bending failure. Loss of composite action can occur if the diameters of the columns are substantially less than the design diameter, or if the axes of the two columns forming the pair are not parallel. Any error in positioning of the initial point of the drilling can also affect the outcome. It appears from Professor Pavlovic's report, though the matter has not been the subject of comment, that loss of composite action actually increases the shear strength of the pair of columns (assuming any given diameter). The question of failure of the wall in bending, as opposed to shear, has not been pursued before me in the evidence and I shall not consider it. As designed, the maximum overlap between the two members of a pair was 200mm, since the design diameter of each column was 900mm and the design separation of their central axes was 700mm. There is no direct evidence of the diameters achieved. The parties' experts were at one in considering that the diameters would be likely to be less than 900mm at depth. Their views were not widely divergent. Mr. Bracegirdle considered it highly unlikely that the diameters would exceed 800mm, and the likely diameter was probably about 600 to 800mm. Dr Greenwood considered that the diameter could fall to 800mm at the toe of the columns, but might be in excess of 900mm at a depth of -10mCD. Mr. O'Brien suggested 750mm to 850mm. It is clear that the bending and shear capacities of the columns will diminish as the diameters of the members diminish, but no calculations have been put before me of the quantitative effect of such diminution below the design diameter of 900mm. I am satisfied that the diameters fell to about 800mm near the toe of the wall. Assuming composite action, that would have increased the margin by which the grout wall had insufficient shear capacity. The question arises whether that is a matter of design, for which Haecon would be responsible to HSS, or a matter of workmanship outside the design, for which HSS would have no remedy against Haecon. Mr. Williamson submitted: The achievability of the design diameter is a function of the input parameters for the jet grouting and the soil types and strength. These are matters over which the designer has ultimate responsibility for the achievability of his design. This is particularly so when the designer is specifically obliged to verify that his design has been satisfactorily constructed, as was Haecon. Haecon so certified in August 2003. On the evidence of Dr. Greenwood, I am satisfied that the diameters of jet-grouted soil-cement columns formed in natural soils depend on the jet momentum and soil properties and structural characteristics which affect its resistance to erosion. The design contract between HSS and Haecon does not specify or make mention of the input parameters for the jet grouting. It does not oblige the designer to verify that his design has been satisfactorily constructed. The design is contained in the execution design report USR 2334/130 revision 2 dated 3 rd June 2002. It states: The grout columns reach down to CD-24.50 with a diameter of 900mm. That report makes no mention of the input parameters for the jet grouting, save the strength of the grout. It makes no mention of the pressures, velocities, jet sizes or configurations. It is true that Haecon gave the certificate mentioned by Mr. Williamson. There is no suggestion that a diameter of 900mm at depth was not achievable. I am satisfied that the failure to achieve diameters of 900mm at depth is a question of workmanship and does not involve a breach of the design contract on the part of Haecon. Mr. Williamson continued: HSS undertook test piles at the specific instruction of Haecon in order to test strength and diameter. Haecon requested such trials before construction of the majority of the columns could commence as part of its responsibilities under its contract with HSS. If the trials were deficient for the purpose of design verification, this is the responsibility of Haecon. In my judgment, Haecon were under no duty to HSS to check the workmanship of HSS. No case has been argued that HSS has a remedy against Haecon on some other ground. Loss of composite action can also arise from divergence of the two members of a pair of columns. Both are intended to be vertical, but it can be that one or both are not vertical. There is no direct evidence as to the verticality of the columns. Mr. Bracegirdle calculated, and I accept, that deviations of only 0.42 per cent. from the vertical in opposite directions can completely remove the interlock between the columns at -18mCD, assuming correct initial positioning of the drills. Mr. Claessens supervised the following elements of the work: the installation of the guidance tubes; the grouting of the VHP columns; drilling works through the VHP columns; and installation of steel reinforcement tubes into the VHP columns. He exhibited to his witness statement a table showing standard procedure under which he was working. That showed that the maximum permissible tolerance in verticality of the columns was 1 per cent. Mr. Claessens wrote: I explained to the drill crew the importance of careful and vertical installed guidance tubes..... I supervised the major part of the installation of the guidance tubes and I am confident that they were accurately and vertically installed. Professor Pavlovic said that the out-of-vertical tolerances were unlikely to be less than 1 per cent. and might be as high as 2 per cent. The European standard on jet grouting EN 12716:2001 at clause 8.4.4 provides: The deviation of drilling from the theoretical axis should be 2% or less for depths of up to 20m. Different tolerances should apply for greater depths..... Mr. Bracegirdle said that the normally accepted tolerance on verticality was "about double" 0.42 per cent. A spirit level had been used to assist in making the columns vertical. Dr. Greenwood considered that a spirit level could have been used which would, in practice, give an accuracy of 1 in 250. He concluded that the deviation was likely generally to be much less than 100mm for "first phase" treatments and less than 150mm for "second phase treatments". The latter would apply, in particular, at a depth of about -20mCD for the work done after the system of making each column of a pair in two passes had been adopted. The figures of 100mm and 150mm allowed for error in the initial placing of the drill. Clause 8.4.3 of EN12716:2001 allows a tolerance of 50mm in the drilling starting point. Dr. Greenwood considered that that was reasonably attainable in normal working circumstances, but not generous. He said that he would expect that positioning the drill rig to the point marked out by surveyors would be within that standard tolerance but not necessarily by much margin. If a spirit level, when used in the circumstances contemplated, can give such accuracy as 1 in 250, it is not clear to me why the standard allows 2 per cent. However that may be, on the totality of the evidence on this point I accept that of Dr. Greenwood set out in the paragraph above. If one column is vertical, the maximum displacement between its axis and that of its pair at -20mCD is likely to be 150mm. If that occurs, and if the diameters of the two columns are 800mm at that depth, then the separation between the columns will be 50mm. If the deviation of the inclined column were in a direction making an angle within ±127.4° of the line joining the intended centres at the top, the axes would be not more than 800mm apart at that depth, and there would not be complete loss of composite action. If the deviation of the inclined column were at a random direction relative to the line joining the intended centres at the top, the probability that there would be complete loss of composite action at a depth of -20mCD would be about 29 per cent. (2 X 52.6/360). What is perhaps more likely in any given case is that each column deviates from its intended position by 75mm at the depth mentioned. That leads to a corresponding probability of about 12 per cent. The calculation is set out in appendix 4 to this judgment. It is a rough calculation, precision not being justified by the imprecision of the data. Other combinations of pairs of deviations from intended axes from 0mm, 0mm to 150mm, 150mm are also possible. The calculation depends on two assumptions: that all angles of deviation from the intended position are equally likely, and that the probabilities of the directions of the deviations of the two members of a pair are mutually independent. In my judgment, those are reasonable assumptions. There is no reason to suppose that any errors in positioning or verticality of the two members of a pair of columns are other than random or are interdependent, nor has the contrary been suggested. There is, in my judgment, in view of the figures given by Dr. Greenwood, a substantial probability that the sum of the deviations (whatever their directions) of the members of a pair of columns will equal or exceed 100mm. I conclude that it is highly likely that several columns have lost their composite action at-20mCD. Mr. Williamson submitted that this statistical approach was inappropriate. There was no direct evidence of any deviations. I would accept that if there were only a small number of columns, the statistical evidence, whatever answer it produced, might well be insufficient to enable the court to make a finding. But here there are 412 columns and I am well satisfied on a balance of probabilities that there must be several instances of loss of composite action. Whether the effect of that in the event of further dredging is likely to be a complete collapse of the wall, or local deformation to an unacceptable or an acceptable extent does not appear. But in my judgment, a wall such as the present where several of the columns have lost their composite action at the relevant level and others have diminished shear resistance and bending moment capacity cannot be regarded as fit for its purpose. Verticality of reinforcement In about 8 per cent. of the columns as made, the reinforcement was outside the grout at the lower levels. That was dealt with by further grouting to cover the reinforcement. It was said that the eccentricity of the reinforcement would weaken the columns. Professor Pavlovic produced calculations to that effect. However, details as to the extent of the regrouting were not in evidence, and in those circumstances it is not possible for me to reach any conclusion as to the effect of the eccentricity of the reinforcement on the strength of the columns. Grouting works I am satisfied that the grouting works were not fit for their purpose in that they caused cracking and bulging or further bulging of the sheet pile wall. The question arises whether that was a result of the design of the works. If the consequence was confined to instances when the annulus was blocked and grout or soilcrete pressure on the back of the sheet pile wall rose substantially above hydrostatic pressure plus the pressure required to drive the effluent up the annulus, then I would not attribute the consequence to design. When blockages occurred and there was continuous fluid between the vicinity of the monitor and a substantial area of the sheet pile wall, it was necessary almost instantaneously to lift the monitor. If there was not continuous fluid, there would be some period of grace. The length of that period is not in evidence. Whilst there may well have been such occasions when the monitor was not lifted in time, I am not satisfied on a balance of probabilities that there were. I am satisfied that the consequence could occur in the absence of such blockages and in such a case the consequence is attributable to the design. However, the evidence is insufficient to enable me to decide which is the more probable: (a) that there were occasions when the monitor was not lifted in time, or (b) that there were no such occasions. Thus I am not satisfied that this aspect of the unfitness of the works for their purpose is attributable to the design of the works. Mr. Williamson's submission Mr. Williamson submitted that ABP's fitness for purpose case was not concerned with the condition of the sheet pile wall, or with the pressure imposed on the sheet pile wall by the liquid grout, but only with Professor Pavlovic's criticisms of the completed grout columns. That submission was made for the first time on the antepenultimate (29 th ) day of the trial. I reject it, for the following reasons. The following is an extract from ABP's amended particulars of claim: PARTICULARS OF BREACHES OF CONTRACT (A) Not Fit for Purpose and/or not Designed and/or not carried out and completed in accordance with the Contractual Requirements ….. 73. Wrongfully and in breach of Special Condition 8(6), …..the Works as designed and/or constructed by HSS are not fit for their intended purpose and/or have not been designed and/or constructed in accordance with the Contract requirements. Without prejudice to the generality of the foregoing, HSS failed to design and/or to construct and/or to complete the Works in the following respects. ….. Construction Phase ..... (7) Failed to take adequate steps to minimise the grout pressures on the inner face of the steel sheet pile wall and deal with fluid-fracture and the spread of grout during the jet grouting process; ….. (13) Failed to act to prevent/limit damage to the steel sheet pile wall during the course of the Works In his opening, under the heading "ABP's case on the defects and deficiencies in the installation of this defective design", Mr. Bowdery referred to the loading on the sheet pile wall caused by pressure of grout. In his opening, Mr. Williamson said that ABP advanced a series of criticisms of the design and workmanship, detailing a positive case that the works were not fit for purpose. He stated one of the broad issues as the impact of the carrying out the works on the sheet pile wall (assuming that it was not already deflected past its elastic limit) and referred to the amended particulars of claim. Thus as regards the statements of case and the submissions of counsel, other than the submission under consideration, the matter was indisputably in issue. But Mr. Williamson relied on the experts' joint statement and Mr. Bracegirdle's evidence. The first item of the experts' joint statement upon which Mr. Williamson relied was part of the preamble, headed Remedial Works. The first paragraph of that item states: Agreed that steel sheet piles that have deformed in flexure and exceeded their elastic limit have not "failed" and that they are capable of sustaining the existing loading. The existing loading must mean the loading in the context of the existing dredge level of -12.8mCD. The second paragraph reads: Agreed that remedial works are necessary but disagree as to why. ABP consider remedial work prior to further dredging is due to the ineffectiveness of the grout columns, Bill Reid, Keith Foster/Tony O'Brien consider that remedial work is required due to the construction of the original wall and the partially ineffective relieving platform. Mr. Williamson's point here is, I think, that ABP do not state that remedial work prior to further dredging is rendered necessary by the distortion of the sheet piles. But it must be implicit in the view expressed by Mr. Reid and Mr. O'Brien (though this item as such was not put to them) that whatever the cause, it was the pre-existing condition of the sheet pile wall that necessitated the remedial works. Thus its condition after the grouting, being, to say the least, no better than it was before, must also on this view have necessitated the remedial works. I do not regard this agreement as removing from controversy in relation to the fitness for purpose issue the question of the cause of the configuration that the sheet pile wall bore after grouting. The remainder of this item is not relevant to the present argument. The second item of the experts' joint statement upon which Mr. Williamson relied was item number 79. That item had not previously been mentioned once during the whole proceedings. It reads: If the sheet pile wall had an adequate factor of safety prior to jet grouting works commencing, and in the event that the grout wall operates as intended, the present condition of the sheet pile wall means that remedial works are unnecessary to enable dredging to -16mCD plus 0.5m allowance for overdredge. Mr. Reid agreed with that proposition. So did "ABP", which must mean Mr. Bracegirdle in this case. So did "Keith Foster/Tony O'Brien". Mr. Foster did not give evidence, but it may well have been Mr. O'Brien who agreed the proposition. It seems that all the relevant experts contemplated the possibility (though perhaps only hypothetically) that the sheet pile wall could have had an adequate factor of safety initially and yet been pushed out of shape by the grout pressure. But given that it had undoubtedly been pushed out of shape, its initial factor of safety appears to be irrelevant to the rest of the proposition. Moreover, it is not entirely clear what is meant by the grout wall operating as intended, since it was intended to operate in concert with the sheet pile wall. Since this cryptic agreement was not put to any of the expert witnesses, I shall not consider it further. The next, and last, item to which Mr. Williamson referred in this context was item number 80. That item reads as follows: The present condition of the grout wall means that remedial works are needed to enable a dredge to -16mCD plus 0.5m allowance for overdredge. [Emphasis added]. The stated comments were these: ABP: Correct. Bill Reid: Agree, but on the basis that the relieving slab cannot be assumed to be effective. Keith Foster/Tony O'Brien: Disagree. Sheet pile wall not adequate anyway due to pre-existing condition of the wall, i.e. partially effective relieving slab, high stresses due to driving of H-piles, overdredging in front of the wall. Mr. Williamson's point here is presumably that ABP's comment does not refer to the condition of the sheet pile wall. But that does not constitute agreement that the condition of the sheet pile wall is irrelevant to the issue of fitness for purpose. Finally in this context, Mr. Williamson referred to evidence of Mr. Bracegirdle. Mr. Bracegirdle said this: …..if the grout wall is effective, there is absolutely no failure at all and there [are] just the shear forces and bending moments that we see projected, and my position at the outset was: providing the grout wall does its job, I don't see a problem with the structure. That was my initial advice to ABP. (Day 14, page 82). Q……by about Christmas last year, you were essentially handing the baton on to Professor Pavlovic? A. Yes, my advice was I think the wall is fine providing the structure can hold up. (Day 14, page 91). Q. And by Christmas of last year you had reached the conclusion, as you told us, I think, yesterday, first of all that HSS's works were not fit for their purpose? A. Yes. I didn't reach that conclusion on my own. The analysis I carried out was suggesting that the wall would be fit for purpose, even though it had two plastic hinges in the sheet piles, and it was entirely the structural aspects of the wall that would control fitness for purpose. So at that stage, in the very early stage, we got opinions on the structure to see whether that would be adequate. All my analyses all the way through have shown that, if the structure itself were adequate, then the strength of the works would be adequate. (Day 15, pages 14, 15). I was also referred to evidence that Mr. Bracegirdle gave a few minutes earlier. None of his evidence in my judgment removes this issue from consideration. The coping beam claim The coping beam claim is a claim by HSS pursuant to clause 12 and clause 52 of the Conditions of Contract that it is entitled to additional sums as a result of having to drill or core through an additional unforeseen thickness of concrete under the coping beam in order to instal the VHP columns. The claim was for an additional payment of £303,424.97 exclusive of VAT. The amount has now been agreed at £69,653.12, subject to proof of liability. The basis of the claim is that the extra thickness of concrete constituted a physical condition or artificial obstruction which was not reasonably foreseeable by an experienced contractor. It is common ground that HSS had to drill through thicknesses of between 250mm and 2400mm of concrete in addition to the 600mm thickness of the coping beam for which HSS made allowance in its price. The effect of clause 12 of the contract is given in paragraph 12 above. Clause 52 relates to valuation of the claim. The only drawing relevant to this claim to which I was referred during the course of the hearing is drawing number 2015/133C, dated 22 nd June 1970. That drawing shows the coping beam with two inches of blinding concrete below it. There is no relevant legend at that point. The two inches is an approximate scale measurement. There is, however, a reference to two-inch blinding below the relieving platform. No other concrete is shown below it. The drawing does show massive concrete in that position at points along the length of the wall where the bollards are founded. It was suggested that absence of concrete in that position except where the bollards were founded was to be inferred from the drawing. Mr. Bowdery relied on the terms of the specification in the contract between ABP and HSS, which, he said, expressly provided that ABP did not warrant the accuracy of drawing 2015/133C. He also submitted that HSS in any event assumed the risk in respect of the thickness of the coping beam and the material around it by virtue of clause 11 [14] . As to the former submission of Mr. Bowdery, the specification provides by clause 101 05: Information given in the contract documents regarding the condition and character of existing structures or any part thereof and the location and nature of pipes, cables and other services are [sic] given without warranty. It is unnecessary for me to decide whether the drawing in question was a contract document, since this claim of HSS is not based on the existence of a warranty. The fact that clause 101 05 exists is not of direct relevance to the question I have to decide: it is simply part of the context. Mr. Bowdery also relied on clause 101 06A, which provided: The procurement of site specific site investigation data is to form part of this contract. The scope and interpretation of the site investigation will be the sole responsibility of the Contractor and no claims for lack or extent of the information in this respect will be entertained. In addition to this further [to] site investigation information is available to the Contractor at the Engineer's office. Regarding the terms of clause 11, as to "information available in connection therewith", such information was in the archive, since the original contractor made a claim for extra payment in respect of the concrete in question. The archive included documents showing the existence and nature of that claim and its rejection. By letter of 19 th May 1971 the then contractor claimed to be unable to obtain the required profile to the backfill material between the culvert and the sheet piling, stated that it had commenced backfilling with concrete to the required profile, and asked that the additional costs be included in its final account. The engineer replied by letter of 25 th May 1971 stating his view that any difficulty encountered in backfilling was the result of the sequence of work adopted by the contractor, and rejected the claim for additional costs. Mr Bracegirdle (day 10, pages 160, 161) considered that HSS should have foreseen that there was a significant risk that the original contractor would have used mass concrete below the coping beam, though not that it should have foreseen that that contractor would probably have used it. In his report, section 15.2, Mr. Bracegirdle said this: 15.2.1 The actual use of mass concrete is likely to vary according to the availability of concrete on site. The construction of berths 202 to 205 was a large project involving the use of large volumes of concrete. It is therefore quite likely that a contractor would make use of mass concrete to ease difficult construction. 15.2.2. The mass concrete used on berth 204 and 205 does not form part of the coping beam. Although the coping beam, the crane rail beam and the service conduits all perform separate structural tasks, they would not be cast separately. The design of the crane rail beam and the service culvert is such that there is a gap between these and the sheet pile wall (refer to figure 96). The reference to figure 96 is a reference to what appears to be a representation of the relevant part of drawing 2015/133C. Mr. Bracegirdle's report continued: An experienced contractor would have known that the quickest and possibly cheapest means of constructing these would be to form the underside of the culvert first, followed by the remaining structure in a single pour. 15.2.3. Casting in a single pour leaves a question over how to support the seaward face of the crane rail beam and to provide a firm base on which to cast the coping beam. Since granular material will not stand vertically, the logical solution is to use mass concrete. 15.2.4. Casting the service conduit and the crane rail beam separately would mean two additional sets of formwork on their seaward side, and breaks in concreting to remove the formwork and place and compact backfill, and place new blinding concrete. The cost of mass concrete is very small when set against the cost of additional formwork, stripping the formwork, backfilling, compacting and placing new blinding concrete. It is not surprising that a contractor, aware of the need to keep both costs and construction programme in check, would use mass concrete to facilitate his works. Mr. Bracegirdle gave this evidence in the course of his cross-examination by Mr. Williamson (day 10, pages 162 to 164): Q. Plainly, Laings, the original contractors, were entitled to take whatever construction technique they thought appropriate? A. Yes. Q. .....it appears that Laing did not anticipate from the drawings that they had that they would need to use mass concrete? A. They used a combination of mass concrete, from what I can see, and the conventional back-fill material; if we look at the claim document, I think you will see quite large variations in the thickness of mass concrete. Q. Again, that is an answer to a different question. Laings did not, it appears, see from the design drawings that they would have to use mass concrete at all? A. No,..... They were apparently using both, concrete back-fill and conventional back-fill, and I'm not sure...why they say they were unable to obtain the required profile. I think they probably could have done if they had spent more money on temporary shuttering and so forth. It is complicated. That's why I say in my report that there is a significant risk of there being concrete in there, because it would just make life a lot easier. Q. .....the factual position appears to be this, does it not? Laings did not anticipate that they would need to use mass concrete at all? A. That's correct. Q. And HSS, my clients, looking at the same drawing, did not anticipate that Laings would have used mass concrete? A. Well, that is correct. I have looked at the same drawings and I came to the conclusion that it would be a strong possibility, because of the difficulty in back-filling between a very awkward area, where you have service culverts and things, that you have a cast concrete wall next to a sheet pile wall, and somehow or other you have got to get shuttering in there, take the shuttering out, put the back-filler in and then do the rest. It is much easier just to form a block of concrete and then cast the beam up against that. It is much easier. On the question of the ease of investigating the thickness of any concrete below the coping beam Mr. Bracegirdle said in his report, paragraph 15.3.1, that it would have been a simple matter for HSS to have used a small diameter hand-held percussive drill. Mr. van der Eecken said this in the course of his cross-examination by Mr. Bowdery: Q..... Can I suggest to you that the problems with the mass concrete were matters that could have been investigated as part of your original site investigation? You could have cored through the coping beam to see what was underneath it? A. That was a possibility, indeed. Q. If you had cored through the coping beam, you would have discovered the mass concrete? A. Depending on which locations, we should have discovered or recorded it. For example, I made an analysis, if we would have caught on the locations where we did the investigation works, so the boreholes 4.5 metre from the front face of the wall, that is borehole 1 and borehole 2 in the Geocone investigations, we would have encountered 80cm. of coping beam which would not give us any indication of excessive increase in concrete under that investigation. Q. That is part of the problem, was it not, that those boreholes were selected before you had decided to actually use the VHP process? A. Yes. Q. And because you had not decided – when you carried out those investigation works – to use the VHP process, you were not having boreholes where perhaps you would have had the borehole if you knew you were carrying out the VHP process? A. Can you repeat the question? Q. If you decide from the outset to use the VHP process, these boreholes would have been in different places? A. No, not necessarily. Q. But would you not have investigated the coping beam with more detail if you were going to bore through it with the VHP process? A. I think there was no indication whatsoever that the drawings which had been made available to us were not reflecting the reality or were unreliable. I think what Mr. van der Eecken must have meant by his second answer quoted above is that boreholes through the coping beam at the longitudinal (along the wall) locations of actual Geocone boreholes 1 and 2 would have shown a thickness of 80cm. I conclude from that evidence a point which may not have been in issue, namely that HSS's site investigation was carried out before, and not after, the decision to bore through the coping beam to create the VHP columns. If so, HSS would have had no occasion at that time to consider the possibility of the existence of mass concrete under the coping beam. Mr. Reid said in his report (section G, paragraphs 3 and 5): I do not agree that an experienced contractor could foresee that there were variable depths of mass concrete present under the coping beam. An experienced contractor would see from the drawings that the designer's intention was that the beam would be conventionally shuttered That this was the designer's intention is now confirmed by documents disclosed by ABP that record that John Laing, the original contractor, made a claim for additional payment for the use of mass concrete. This claim was rejected by the engineer. ..... Had ABP supplied the archive information that indicated the presence of mass concrete then I accept that an experienced contractor would have carried out investigations. I believe, however, the site photographs and correspondence that confirmed the presence of mass concrete were not made available at tender stage. The expression "made available" doubtless comes from clause 11. In isolation, it is ambiguous. It may mean handed to the contractor without any request on the part of the contractor, or it may mean available for inspection on request. The relevant documents were not handed to HSS without request, but I there is no reason to think that ABP would not have allowed the contractor access to the archive if it had been asked. Mr. Brown had told tenderers of its existence. In the context, namely "all information whether obtainable by [the Contractor] or made available by the Employer", I think the words must bear the latter meaning. The information was available and had been made so by the Employer. But HSS was unaware of it. However, the question I have to decide is whether the claim falls within clause 12(1). The concrete in question was undoubtedly a physical condition or artificial obstruction. By common consent, and indeed on the pleadings, although the drawing was not labelled "As built", it was apparent that it was issued for the purpose of constructing the works, and that both HSS and ABP understood it generally to represent the structure of the quay wall. In my judgment, once it had been decided that the VHP columns would require drilling through the coping beam, (and that would have required 412 drillings instead of 24 as originally envisaged), an experienced contractor would carefully consider the implications of drawing 2015/133C. He would ask himself how much drilling would be likely to be required, especially given the size of the job. I am impressed by the evidence of Mr. Bracegirdle. An experienced contractor would be able to deduce, as he did, that the original contractor might well have found it convenient and economic to use mass concrete beneath the coping beam. Notwithstanding the drawing, an experienced contractor in my judgment would at least make enquiries of the employer whether the employer had information on the subject, or would have carried out an investigation using a percussive drill or some other method. This contract is framed in terms onerous to the contractor, but I must give full effect to clause 12. The question is could the mass concrete reasonably have been foreseen by an experienced contractor. I have stated above what in my opinion an experienced contractor would do. But even if I am wrong, an experienced contractor could, I am satisfied, have adopted the reasoning of Mr. Bracegirdle or otherwise have appreciated the risk of the existence of the mass contract. Thus he could reasonably have foreseen it. This claim fails. I am far from saying that an experienced contractor would necessarily do that of which I have found him capable. Under pressure of competition, or in the hope of making a successful claim, he might be willing to take the risk. But in my judgment the drawing would not so set the mind of an experienced contractor at rest as to give him a false sense of security so that he did not appreciate the risk. Grout disposal claim HSS claims an additional payment of £1,028,908.28 plus VAT for the disposal of surplus grout returned to the surface from the jetting process. The volume of grout so returned was 22,566m³. The original tender proposal of HSS was that the surplus grout should be dumped on the river bed, with a view to its being removed when the river bed was dredged. That proposal was not acceptable to ABP. On 28 th November 2001, ABP wrote to HSS in relation to HSS's tender submission of 8 th November 2001 requesting further clarification/confirmation of nine points, including as point e): Disposal of surplus grout mixture by discharge into the river will not be acceptable. Please confirm your alternative method of disposal. HSS replied on 3 rd December 2001, in relation to point e): Other solutions to discharge the surplus grout mixture are amongst others: A dumping area on shore in the vicinity of the work area (maximum 500m from the injection locations) where HSS can discharge the surplus grout. In case such dumping are[a] can not be provided for in the vicinity of the work area but on another location in the port, a barge or trucks can be used to transport the surplus grout from the work area to this dumping location. Please note that HSS offer does not include the extra costs resulting from the change of discharging the surplus grout. A meeting between the parties was held on 3 rd December 2001. Mr. Brown wanted a lump sum price for disposal of grout as an extra to the contract. The minutes, which were drafted by Mr. Brown, note "HSS to provide extra cost for disposal of surplus grout mixture offsite". According to evidence of Mr. van der Eecken, which I accept, Mr. Brown made it clear that the reference to "offsite" was to a landfill site. Mr. Brown gave evidence about that meeting. He said in his witness statement that he told Mr. van der Eecken that ABP required HSS to provide a lump sum price to cover all the costs of the disposal of the surplus grout. That evidence was challenged, Mr. Brown having made no mention of a lump sum price in the minutes of the meeting. Nevertheless, I accept that that is what Mr. Brown said he wanted, having regard to evidence given by Mr. van der Eecken (day 3, page 137) with reference to the costs of grout disposal: A. .....there were too many uncertainties to include it as a lump sum..... I was not willing to include it as a lump sum in the contract, and that's why we agreed to pay it [sic] – the only possibility then – as a provisional sum. Mr. Brown went on in his witness statement: Koen [i.e. Mr. van der Eecken] then calculated the volume of surplus grout that would arise from the VHP works, giving a verbal explanation to indicate what amounts needed to be disposed of. I recall that Koen took the volume of a column and multiplied it by the total number of columns and then by a factor of two thirds. I understood this factor to be a multiplier used in the VHP grouting industry to reflect the proportion of excess grout material to be displaced per column. On this basis Koen calculated a figure of 4,400m³. He then rounded this up to 5000m³ ..... I accept that evidence. Mr. Brown finished the last sentence as follows: indicating that this ensured that it was sufficient to cover the maximum quantity of surplus grout which would be produced and which needed to be disposed of. So far as that passage is concerned, I accept that it represents the impression in Mr. Brown's mind. That is indeed the natural impression to be gained from the earlier part of the passage to which it is the conclusion. A correct calculation would be as follows: Cross-sectional area of grout in pile: 1.181318m²; Length of column: 25.14m (see figure 3 of Dr. Greenwood's evidence); Number of columns: 412; 1.181318 X 25.14 X 412 X 2/3 = 8157m³ (or, for the purposes of a mental calculation, area=1.2, 1.2 X 25 = 30, 2/3 of 30 = 20, 20 X 412 = 8240). In the event, the amount of grout used would have been increased by reason of the change in method, which required the upper parts of the columns to be drilled through in order to grout the lower parts. But on the basis of the 2/3 factor, which I accept that Mr. van der Eecken mentioned, his estimate was a substantial under-estimate at the time that it was made. ABP's and HSS's experts, Mr. Bracegirdle and Dr. Greenwood, agreed, under the heading of grout disposal, that the volume of raw effluent was approximately 10,000m³, and that that figure could have been calculated at the time of tender. For ease of pumping, water was added to the raw effluent, making it into a slurry. That accounted for the fact that the total volume disposed of was more than double the figure of 10,000m³. On 6 th December 2001 Mr. Brown sent his minutes of the meeting of 3 rd December to Mr. van der Eecken referring to an attached document, which must have been HSS's clarification/confirmation submission of 3 rd December. Mr. van der Eecken replied the following day: Extra cost for disposal of surplus grout mixture offsite. A budget of the extra cost for the disposal of the surplus grout mixture offsite is given in annex 1. Please note that this budget outline is based on the disposal of a quantity of maximum 5000m³ of slurry during a period of maximum 12 consecutive weeks. Please be informed I haven't yet received a quotation for the disposal of the surplus grout mixture from PTC but based the local quotation for the transport and disposal of the slurry on a verbal confirmation given by a third party. PTC was a local contractor. Mr. Bowdery submitted that that statement involved in effect a warranty that the volume of 5000m³ would not be exceeded. I reject that submission. Mr. van der Eecken, in his witness statement, said: HSS did not confirm that the maximum surplus grout that would be produced would be 5000m³. The calculation of 5000m³ was based on half of the calculated total injected volume and corresponded with the total volume of the columns. Annex 1 was headed "Disposal of surplus grout mixture offsite" and sub-headed "Budget outline". It gave a total under the heading "Load in on site" of £77,616.00. A breakdown of that figure was given in terms of manpower (plant operator and labourer) and pumping equipment, both in terms of hourly rates for 12 weeks. Annex 1 also gave a total under the heading "Disposal offsite" of £110,000.00 broken down as local quotation, £100,000.00 and 10% GO and risk £10,000.00. Mr. Williamson submitted that having regard to Mr. van der Eecken's letter of 3 rd December it was implicit in annex 1 that the area on site for loading in must be within 500m of the injection locations. I consider that submission below. On 19 th December 2001 the parties met. Mr. Brown's manuscript note of that meeting mentions disposal cost of inert grout of £8 a tonne, density 1.75 tonnes/m³, cost £14/m³; not inert £13 to £20/tonne. It also states "5000m³ calculated amount". The disposal costs mentioned in the note are equivalent to £22.75 to £35/m³. Mr. Brown gave the following evidence, which I accept. At the meeting, Mr. van der Eecken again explained his method of calculating the figure of 4400m³ before rounding the figure up to 5000m³ to allow for contingencies. Mr. Brown felt that Mr. van der Eecken had reassured him as to the maximum volume of grout for disposal, but he recognized that the price given by Mr. van der Eecken had been based on a verbal quotation and he wanted HSS to confirm its price in writing. The two men discussed the question whether HSS could confirm the surplus grout disposal rate at £20/m³. Mr. Brown gave to the court examples from his own experience of £9/m³ at the relevant time and £11 and £12/m³ currently. At the meeting, Mr. van der Eecken explained that the reason the quote was seemingly expensive was that the landfill operators had not been convinced that the surplus grout would be inert. Mr. van der Eecken stated that for inert material disposal costs would be £8 per tonne, and that for material not classed as inert the cost of disposal could range from £13 to £20 per tonne. It was agreed at the meeting that the disposal costs should be based on the surplus grout being considered to be contaminated. As no sample could be provided to the landfill operator until the start of the works there could be no prospect of confirming a lower rate prior to contract. Mr. Brown was confident that the surplus grout would not be classified as contaminated as there was little or no possibility of there being a contaminant in the ground to be grouted. Mr. van der Eecken gave this evidence, which I also accept, about the meeting of 19 th December: I confirmed that the £100,000 figure was based on a rate of £20/m³ for a maximum of 5000m³. I confirmed however that both the rate and quantity of surplus grout [were] uncertain. I agreed with Gary Brown of ABP that it was not possible to estimate with certainty the quantity or rate and therefore the cost of transport and disposal of the surplus grout off site in the sum of £110,000 would be included as a provisional sum. On 21 st December 2001 Mr. Brown sent to HSS his minutes of the meeting of 19 th December, which noted that HSS undertook to confirm that the rate used in the disposal item for the waste grout mixture, provided with their facsimile dated 7 th December 2001, was £20/m³. On the same day, 21 st December, Mr. van der Eecken sent to Mr. Brown a fax message which made reference to HSS's fax message of 7 th December, and stated: Extra cost for disposal of surplus grout mixture off site I hereby confirm we agree to add the extra cost for disposal of surplus grout mixture off site included in above mentioned fax message as a provisional sum to the Bill of Quantities of the tender document which was submitted by HSS on 8 th November 2001. On 19 th March 2002 ABP wrote to HSS accepting HSS's tender for the sum of £3,856,465 "inclusive of a provisional sum of £110,000 for the cost of disposal of surplus grout arisings". The letter listed the documents comprising the contract, including HSS's facsimiles of 7 th and 21 st December 2001 (but not that of 3 rd December). It was signed on behalf of HSS on 26 th March 2002 as having been received. It is evident that the price was also inclusive of the figure of £77,616 as quoted for "Load in on site", since ABP's liability to pay that sum, and the fact that it has been certified and paid, are not in dispute. The dispute over that matter is simply as to the location of the site provided and the question whether ABP is liable to pay more than £77,616. How the figure of £3,865,465 was arrived at from the original tender figure of £3,834,563 is not in evidence. The words "Provisional Sum" are defined as follows in clause 1(1)(l) of the contract: "Provisional Sum" means a sum included and so designated in the Contract as a specific contingency for the execution of work or the supply of goods materials or services which may be used in whole or in part or not at all at the direction and discretion of the Engineer. Mr. Williamson submitted that the sum of £110,000 could not be a provisional sum as defined in the contract, since (1) the sum was not a contingency for works which might not be carried out: the surplus grout was necessarily going to require disposal; and (2) the disposal of grout was not to be at the direction and discretion of the Engineer. Moreover, he submitted, the parties could not have intended the sum to be a maximum since they contemplated that the cost of £20/m³ could well have been exceeded and the volume of 5000m³ could have been exceeded. I reject the latter submission. Having regard to Mr. van der Eecken's calculation, it must at the least have been in the contemplation of reasonable parties that 5000m³ was a likely upper limit to the volume of grout. And the figure of £20/m³ was generous for inert grout, which Mr. Brown considered was the likely description of the grout. It is clear on the evidence of Mr. van der Eecken (day 4, pages 6 and 7) that in the event, the cost paid to the landfill contractor was only £8.67/m³. Mr. van der Eecken said: We included £20/m³ because £6 was for the handling cost to load the material from the area into the lorries or the trucks of disposal, so the 14 was plus [an] add-on for the handling to put it into the trucks and send it to landfill. I reject point (1) of Mr. Williamson's submission. In my judgment, the use in clause 1(1)(l) of the verb "used" is referable only to the sum of money, not to the execution of work or to the supply of goods, materials or services. I accept Mr. Williamson's point (2), however. The expression "provisional sum" in this context cannot mean exactly what it means in clause 1(1)(l) of the contract. But the expression having been used by the parties in the context of the contract, it must in my judgment be taken as having the defined meaning, so far as possible. The reasons put forward for adopting the contrary view are insufficient, in my judgment. It is clear that Mr. van der Eecken understood the meaning of the expression "Provisional Sum" in the contract. He was asked (day 3, pages 137, 139, 140) about his use of the expression in his fax of 21 st December: Mr. Bowdery: ..... Do you remember agreeing to add the extra cost of disposal of surplus grout off-site? A. As a provisional sum, yes. Q. When you wrote that, did you actually look at the contract conditions and see how "provisional sum" had been defined? A. No, I hadn't at that time. My understanding was the provisional sum, for contingencies, and that's why I also used a small P and small S. ..... JUDGE HAVERY: .....you did say a minute ago you did deliberately put "provisional sum" in lower case initials; is that because you [had] this particular clause in mind and wanted to avoid it or what? A. We want to make sure, because we want to make sure that it was not limited and then our in-house counsel told us when you look at the contract and you take a small P, small S, then you have the opportunity that it was not capped, so I take advice from the in-house legal counsel. JUDGE HAVERY: Before you wrote that letter? A. Before – it is obvious you have a sum of £110,000, which I asked advice, okay, also my general manager, he said okay go to the in-house counsel and ask what his understanding of the provisional sum, because no-one wanted, we did not want to put it within the lump sum, so the only other possibility when I spoke to that with Mr. Brown, was a provisional sum but I wanted to be sure at that time that I was, it was not capped because of the too many uncertainties I was dealing with. So that['s] why I took his, that advice was an advice in ten minutes, it was not a written advice or something like that. But he advised me at that time to put it in a small P, small S. JUDGE HAVERY: Would it not have been a good idea to make it clear, in so many words, that you were envisaging that it would not be capped? A. Yes, indeed. It is clear from that evidence that Mr. van der Eecken wanted to make sure that the sum was not capped without telling ABP. I infer that he knew perfectly well that ABP were negotiating on the basis that if they could not have a lump sum, they wanted the sum capped. Given the context that the whole contract was a lump sum contract, not a measure and value contract, the inference in my judgment is that the provisional sum was capped. Mr. van der Eecken's evidence supports, rather than casting doubt on, that inference. Mr. Brown gave this evidence in his witness statement: 136. Despite discussing the costs of the grout disposal in some detail, at no time during the discussions on 3 rd and 19 th December 2001 was the methodology of disposal described in any detail – or even a final method confirmed by HSS. 137. I had anticipated that the surplus grout would be of a consistency suitable for transportation and could be pumped into a truck to be taken off site for disposal..... 138. Nothing in the pre-contract communications with HSS – either in letters or meetings – suggested that any form of interim storage or double handling would be required. I had assumed that, based on their experience, HSS would have taken into account all activities that they considered necessary and that the sum agreed would be adequate for the entire disposal operation. As to the first sentence of paragraph 138, HSS's communication of 3 rd December 2001 did refer to dumping on site as a solution amongst others. Subject to that qualification, I accept the evidence of Mr. Brown given in those paragraphs of his witness statement. The first post-contract meeting between the parties took place on 4 th April 2002. The purpose of the meeting was to resolve as many of the outstanding practical issues as possible so that the strengthening works could begin on site as soon as possible. At that meeting a large number of logistical issues were discussed, including details of where the areas required by HSS during the works would be situated. But there was no discussion of an area for surplus grout handling by HSS. There was no discussion at the meeting that an additional area would be required or that it should be sited within 500m of the grouting operations. HSS did not mention that double grout handling would be required. The only discussion regarding grout disposal was raised by Mr. Brown at the end of the meeting. HSS had not included it as an item on the agenda and the issue was recorded in the minutes under the heading Miscellaneous/Any other business. The minute stated: HSS explained that they were considering a number of options for disposal of grout arisings, and would provide proposals for discussion shortly. The facts stated in this paragraph come from the unchallenged evidence of Mr. Brown (his witness statement, paragraphs 142 to 146), and the minutes in question. A further meeting between the parties was held on 15 th April 2002, when Mr. van der Eecken was absent. HSS's requirements regarding areas on the quay and within the terminal area were again discussed and agreed. But there was still no discussion regarding the disposal of surplus grout or any requirements on the part of HSS for space for the purpose. As recorded in the minutes of the meeting: The method of disposal of grout arisings is still being considered. HSS to provide detailed proposals shortly. On 24 th April 2002 HSS sent a fax to ABP asking for an urgent meeting to discuss a temporary method of grout disposal involving the use of skips. A meeting was held the same day. At that meeting David Holt of Southampton Container Terminals ("SCT") suggested a location for the skips at the back of SCT's offices. Mr. Voorhuis, the project manager, said "It sounds perfect". Ms. Bellardinelli's note of the meeting records that Mr. Voorhuis, Mr. Holt and she visited the potential storage site. It says "The site looks perfect for our plans". The whole of Ms. Bellardinelli's notes appear to be conscientiously and efficiently prepared. I am satisfied that the expression "The site looks perfect for our plans" records something that was said by one of those persons, presumably Mr. Voorhuis. The evidence given by Mr. van der Eecken in his witness statement was that the area was suitable as a temporary storage area, but it was more than 500m away from the VHP locations and it was therefore not possible to pump the grout effluent. I am satisfied from Mr. Brown's plan GWB1 that the area in question was located largely, about four-fifths, within the 500m radius. In examination in chief (day 3, page 87), Mr. van der Eecken gave a different reason why that area was unsuitable for pumping. He said: So you had the quayside, then you had in the back yard you had all the containers stored and in between the containers you had the gangways which were used or occupied for moving the straddler carriers, then up to, let's say where you can see Southampton Container Terminals [sc., an area about 250m from the quay], up to that area, you had a fencing because the container terminal, the terminal itself is completely fenced off for security and safety reasons, and which means that in between the quayside area and the fencing you had the most busiest activity of the container terminal which could not really be interrupted by putting any lines [sc. for pumping] on top of the quay wall. That evidence seems to imply that any area would have been unsuitable for pumping. By the same day, 24 th April 2002, there had already been allocated to HSS an area ("area (2)") averaging about 80m behind the quay wall as a grout batching plant area. HSS had requested of ABP an area of 400m² for the grout batching plant. Mr. Brown gave evidence that the extent of that area was 2500m² which, he said, would have been adequate for a far larger number of skips than were actually used by HSS. And, he said, there was an unused underground conduit from the quay wall to that area which could have been used to service the area. ABP had pleaded that the area was 2500m² and had been available since 22 nd April 2001. HSS pleaded in response that area (2) could not be used and was impractical as a stock pile area. Mr. van der Eecken in his witness statement said simply that area (2) was already occupied by HSS's equipment. He was questioned about that. He said in cross-examination (day 4, page 8): .....the grout batching plant, the commencement of the works, was only 20m by 20m and only sufficient to store our workshop and the grout plant. Then, indeed, that area was expanded but that was due because we had to mobilize more and additional plant and equipment to deal with the night shift and the changed methodology..... In re-examination he said (day 4, page 46): .....it was only at that start of the job it was only 20m by 20m and not suitable to make an area or a temporary storage area for the disposal purposes. And when it was extended, that was much later on, and then we needed that area to be occupied by the additional work plant and equipment which needed to be stored for dealing with night shift and additional drilling equipment needed for the down-hole works through the columns. Also at that time, we already had the solution with the temporary storage facility at berth 109, so there was, it didn't raise any queries any more queries, because the disposal was working now. Minutes of a meeting between the parties held on 15 th April 2002 record, under the heading "Contractor's requirements": .....it was agreed that the grout mixing plant would be within "Bravo" stack, approximately central to both. Approximately 20m X 20m, but HSS to define precise requirements. SCT undertook to have the area cleared by 22 nd April. The hoses/pipes from the grout plant would require to be buried beneath the paving. HSS to instal conduits. Ms. Bellardinelli's notes of the meeting of 24 th April (when Mr. Brown was not present) record Mr. Voorhuis as saying: There are problems with the disposal of the jet grout material. HSS can't find a permanent site to st[ore?] it. People are suspicious of the material potent toxicity. HSS guarantees it is not toxic. It is simple mixture of concrete and soil floating on top of the columns. Arrangements are being made with Jackson to find out a temporary solution. That was the solution described by Mr. Voorhuis as sounding perfect. I am unable to make a finding precisely when the full 2500m² was made available to HSS. But I am satisfied that it could have been made available by 22 nd April if HSS had asked for it. It appears to have been made available when HSS did ask for it (as I infer that they did) for the additional plant and equipment some time later. HSS did not formulate their method statement for disposing of the surplus grout until 10 th May 2002. On 26 th April it transpired that HSS were obtaining quotations for the supply of the skips which would have involved a total cost of some £250,000 for on-site handling, not including the costs of loading trucks for disposal off-site. By 7 th May, it was clear to Mr. Brown that the temporary solution of using skips was far from ideal. No permanent solution appeared to him to be forthcoming. Mr. Brown was active in helping to solve the problem of disposal of surplus grout. On 7 th May 2002 he called a meeting on site, where Mr Voorhuis said that a possible solution was to build a retaining structure where the surplus grout could be stored. Water would be added to the surplus grout to allow it to be pumped to that retaining structure. But additional space would be required and it would have to be within pumping distance of the works. Mr. Voorhuis was estimating that there would be up to 1500m³ of surplus grout per week and that a structure with at least 3000m³ capacity was required. Mr. Brown found a solution to the problem later the same day. He made available an area behind berth 109 where there were a number of large underground tanks where surplus grout could be stored. The area was about 2km by road from berth 205. The tanks were then being used to stockpile glass cullet. HSS cleared the area of glass cullet, the tanks were excavated and the skips were then decommissioned. The volume of the tanks, however, was less than 3000m³. On 29 th July 2002, Mr. Brown made available to HSS an additional storage tank in the same area. On 10 th May 2002 HSS for the first time sent to ABP a method statement for the disposal of surplus grout. HSS obtained two silos from Belgium which arrived on site on 25 th May 2002. They were kept in the jetting area and the surplus grout was pumped into them. The advantages of the silos in Mr. van der Eecken's opinion was that they occupied a relatively small amount of space on the quay wall and they could be moved by crane within the working area. In addition, HSS hired tankers and concrete mixer trucks to transport the surplus grout from the silos to the storage tanks located behind berth 109. Large quantities of water were added to the grout in the silos to ensure that it remained in liquid form prior to its transfer to the tanks. That led to additional costs. I shall adopt Mr. Brown's description of the process of disposal: (1) the surplus grout was transferred by pumping into silos situated immediately adjacent to the quayside—this provided a buffer once the grouting had started on an individual column to enable the grouting to continue to completion; (2) a transport vehicle was then loaded from the silos. It transferred the surplus grout to a storage site (i.e. berth 109 storage tanks); (3) the surplus grout hardened in the storage tanks; (4) the surplus grout was then transferred to another transport vehicle by a 360° tracked excavation; and (5) the surplus grout was taken away to another, unknown, site off the port estate. Mr. Brown said that steps (1) to (4) were not anticipated by ABP and had not been discussed pre-contract. HSS were now storing surplus grout and seeking payment for that additional activity. In addition, the volumes of surplus grout that were being produced were many times greater than those which had been predicted by Mr. van der Eecken at the two meetings prior to Christmas 2001. I accept that there were some steps additional to what had been envisaged. What had been envisaged was pumping or otherwise transporting the grout to a place on site, and then removal off site. On 22 nd May a meeting was held to discuss who would pay for the increased costs of the grout disposal. There is a dispute about the accuracy of ABP's minutes of that meeting. Those minutes, sent to HSS about a month after the meeting, include the following: KV [Mr. van der Eecken] explained that his interpretation of the inclusion of a provisional sum agreed at tender stage was to cover the transfer of arisings from the works to a holding area to allow the grout content to solidify plus the removal of such arisings from the holding area to landfill. It was however underestimated how difficult it would be to obtain a temporary storage site for the holding of the liquid grout/gravel both on or off the port estate. GB [Mr. Brown] explained..... ..... With regard to temporary storage, at no time was an indication made that this would be necessary..... ..... MV [Mr. Voorhuis] confirmed that the tanks provided them with an acceptable solution. He was however still evaluating the true cost of handling the grout arisings. In view of the above GRS [Mr. Steele, the contract engineer] stated that ABP have no doubt that HSS were given opportunity to revise their tender submission to include for all costs associated with the removal of grout arisings. Therefore no additional payment is deemed justified. HSS considered that those minutes did not reflect the discussion of the meeting correctly. Their version, so far as is material, was as follows: KV explained that at tender stage he proposed to dump grout arisings over the quay wall. HSS were informed that this was totally unacceptable on environmental grounds and HSS were required to provide additional costs for disposal of the material to a landfill. This additional cost was made up of two amounts, the first (fixed sum) for removal of the liquid grout/gravel arisings by pumping from the injection locations to another location within the working area and the second (provisional sum) for subsequent transport and disposal from this working area to a landfill. KV explained that because there were so many uncertainties he did not receive firm quotations from subcontractors at tender stage. He therefore based the local quotation for the transport and disposal of the slurry on a verbal confirmation given by a third party. ..... The solution was finally found when ABP offered a temporary holding area through use of ready existing, abandoned ground tanks to the rear of 109 berth. MV confirmed that the tanks provided were an acceptable solution, although he was disappointed that it took so long to come up with such an easy solution. Whilst I do not regard this particular dispute about the minutes as important, I shall quote Ms. Bellardinelli's minutes of the meeting. As resident site engineer for the contract she was charged with the task, among other things, of maintaining records of all site events, including keeping minutes of meetings. Although I was not referred to her minutes of this particular meeting, it is worth quoting from them at length: GB [Mr. Brown]: Problem: Who pays what and when? The tanks solution seems to have cooled down the situation but we still need to discuss what the provisional amount of money agreed in the tender stage includes or not. An extra amount was agreed to cover possible risks, the worst being a contaminated grout. We see this sum as comprehensive of all these possible issues. I acknowledge Marc [Mr. Voorhuis] disagreement but we also know that he wasn't present at the tender discussions. KV [Mr. van der Eecken]: Our interpretation is that the sum should have included the extra labour required for the VHP operations and the expenses for the removal of the grout off-site. We therefore have two separated parts: 1) our pump to pump the slurry into the skips 2) the skip excavated by a third party and then taken away to the final destination. We didn't include any expense of internal removal because at that time we didn't have any idea of what the problems of internal storage and disposal would have been. ..... Our idea was to transport the grout from silos to skips in a liquid form. Then dispose the slurry somewhere where it could have hardened and being finally excavated. Usually, in a normal circumstance, we can dispose of an area where to dig a pit and put the grout in. On this case, this option was impossible. We were very fortunate to have the tanks from ABP. MV [Mr. Voorhuis]: At the moment, the grout goes from the silos into tank wagons. The experience is not good. We are looking into using a "ready to mix" cement trucks that can discharge the grout in the tanks where then the slurry will be treated. The price we came up with which will approximately cover all the expenses (from start to final distribution) is of £25/m³. The price includes disposal and handling of the material. GB: Right from the beginning, the material was always considered as unlikely to be contaminated. We therefore assumed that that extra amount should have covered the other problems, such as transport and disposal. MV: We agree that some of the delays were caused by our own mistakes (gravel) but not this particular one. GB: Now the issue is: should ABP pay for the transport from silos to tanks? I think not. The reason being that we actually never had the knowledge of this all [sic] process therefore we were unable to judge any potential risk. We trusted HSS because experienced in this field, and if this issue was raised at tender stage we could have agreed something different. The truth is not even HSS was aware of the risk! KV: HSS agreed the price on a basis of assuming that the distances on site were not so big and the grout could have been even pumped from a position to another one. GB: Lies! That was never the case. We knew right from the beginning that there would have been serious logistic problems. GF [Mr. Ferguson, of Gardiner & Theobald, acting for ABP]: The truth is that we relied on HSS experience to assess a fair sum of money which could have comprised all the potential problems. The only serious problem was [sic] mentioned was just for the grout being contaminated. HSS misjudged all the situation. MV: My opinion is that the contract is not clear about the disposal of grout on site. G [Mr. Steele]: Maybe we have been a bit naïve to not contemplate all these risks from the start. KV: The agreed £78,000 don't cover in my opinion the internal transport on site. GB: Yes, they do. GF: Why didn't you think of asking for a storage area right from the beginning if you had a clue of what was going to happen? KV: We never expected to not have at least a temporary side [sic: site?] to use. GB: I would like to highlight the fact that if ABP didn't provide you with the tanks area you wouldn't have progressed at all. That area is worth £15,000 for the 3 months of work and it represents a big step toward you to help you out. In my judgment the total sum of £187,616 does cover internal transport on site. As to the question whether ABP provided a suitable area on site within 500m of the quay for the dumping of surplus grout, ABP, in its reply and defence to counterclaim, pleaded that it provided five areas to HSS. The area numbered (4) was identified as "Adjacent SCT office". It is shown on a plan GWB 1 exhibited to Mr. Brown's witness statement. It is there marked in green and labelled "Grout disposal and skip area". Mr. Brown gave evidence, which I accept, that the extent of the area was 3250m². Terms of contract regarding dumping area In my judgment, it was not a term of the contract that ABP would provide a dumping area within 500m of the injection locations. My reasons are these. First, the method of grout disposal proposed by HSS on 3 rd December 2001 gave one option as a dumping area within 500m of the injection locations, and another option where no such location was provided. Moreover, the document indicated that those solutions were "amongst others". Second, the proposal of 3 rd December was not one of the documents listed in the acceptance dated 19 th March 2002 of the tender. Third, grout disposal was included in the contract sum, including the provisional sum. The whole contract was a package. The schedule of prices was a contractual document, being included in the form of tender. The schedule of prices provided in section A: preamble (page 114 of the contract documents) as follows: 4. The prices in the Schedule of Prices shall be for the work finished and complete in every respect and must cover all payments and allowances of every description to any persons employed on the Contract and all incidental and contingent expenses and risks of every description necessary to execute complete and maintain the Works whether such payments allowances expenses and risks are or are not specifically referred to in the Tender Documents and the Contractor shall not claim additional reimbursement in respect of any such payment allowance expense or risk. It is true that the tendered price was amended by agreement between the parties, and the figure that appears in the Schedule of Prices was not amended. In my judgment "the prices in the Schedule of Prices" must be read as meaning the price amended by agreement. The amendment cannot nullify the terms of clause 4 of the preamble. Mr. Williamson's argument was that the quoted sum of £77,616 was agreed on the basis that ABP would provide a suitable storage area within 500m of the grouting works for temporary storage and hardening of the grout before off-site disposal. If and to the extent that that was not provided, the additional expense would "therefore" be borne by ABP. In the circumstances, when considering quantum, it was necessary to do so on the basis that the £77,616 was subject to adjustment by reason of the non-provision of a suitable disposal area within the vicinity of the works. The legal basis of Mr. Williamson's argument was not stated. The claim as pleaded is under clauses 12 and 52 of the contract and "Further and in the alternative, the Engineer has wrongfully refused to certify the aforesaid sums and/or ABP has wrongfully refused to pay such sums claimed and HSS and/or DIUK claim the same by way of damages". Clause 12 is the clause relating to unforeseeable physical conditions or artificial obstructions. It clearly has no place in the present argument. Clause 52 relates to variations. But the disposal of surplus grout, though additional to the tender, was part of the original agreement. Therefore it did not constitute a variation. There remains to be considered the further and alternative argument. It must entail a duty on the part of ABP, in the events which happened, to pay to HSS the costs incurred by HSS in disposing of surplus grout, whatever those costs might be. The basis of the claim must, I think, be one or both of the following. First, that it was a term of the contract, sounding in damages, that ABP would provide a suitable area for grout dumping. Second, that the price of £77,616 included in the contract sum was conditional upon such provision, and that in the absence of such provision the price for grout disposal would be at large. It has not been suggested that there was any collateral contract. The alternative method of disposal of surplus grout given by HSS on 3 rd December 2001 provided for a dumping area within the 500m limit where HSS could "discharge" the surplus grout, or a dumping area outside that limit whither a barge or trucks could be used to transport the surplus grout. At that point it is clear that a 500m limit was not a necessary feature of the method. Pumping was not mentioned, nor in my judgment was it implicit by reason of the use of the word "discharge". The quotation of 7 th December 2001 for £77,616 included an amount of £38,304 for pumping equipment. Whilst no mention is made of barge or trucks, in my judgment the reference to pumping equipment does not imply the 500m limit. If I am wrong about that, then I would accept that the offer was made on the basis submitted by Mr. Williamson, namely that a suitable area within that limit would be provided. I do not accept that it follows from that that if such area were not provided, the sum payable would be at large. That is inconsistent with the terms actually agreed by 19 th March 2001, the sum of £77,616 being part of the contract sum and payable on the terms of the contract. It was not suggested that the purported acceptance on those terms was a mistake. Thus HSS's grout disposal claim fails in point of law. But I find the 2500m² grout batching plant area to have been an area located within 500m of the jetting operation. The balance of the area, amounting to about 2100m², outside the grout batching plant I find to have been suitable for pumping surplus grout there for the purpose of dumping, though not for prolonged storage of surplus grout. HSS could have asked, but did not ask, for that area to be made available for the purpose of dumping surplus grout. That area was made available when asked for, and there is nothing to suggest that it would not have been made available when asked for if it had been asked for earlier. I am not satisfied that it would not have been. Thus if I am wrong in holding that there was no term or condition that ABP make available a suitable area within 500m of the jetting area, I am not satisfied that there was any breach of such term or non-fulfilment of that condition. Parent Company Guarantee ABP seeks specific performance of an obligation on HSS to obtain a parent company guarantee of its obligations under the contract. In the alternative, it claims damages for breach of contract in not obtaining such a guarantee or an entitlement to withhold the sum of £383,456.30 being 10 per cent. of the tender total. Clause 10 of the contract provides, so far as material: Within 28 days after the award of the Contract the Contractor shall if so requested by the Employer provide at his own expense in all respects an ultimate parent company guarantee for the due performance of the Contract in the form annexed to these Conditions of Contract. The form annexed to the contract provides that the expiry date shall be the date of issue by the Engineer of the Defects Correction Certificate as defined in the contract. On 27 th May 2002 HSS sent to ABP a parent company (Dredging International NV) guarantee headed "Corporate Guarantee" expressed to expire on 31 st December 2002. On 29 th July 2002 ABP (by Mr. Steele, the Engineer) wrote to HSS a letter which included the following: With respect to the Parent Company Guarantee (Corporate Guarantee) I note you have limited the extent of your liability to the 31.12.02 when in fact your liability should extend until issuance of the Defects Correction Certificate. I would be grateful if this amendment could be made and the document re issued accordingly. On 20 th December 2002 Mr. Steele wrote again to HSS: I note from our records that the Ultimate Parent Company Guarantee (Corporate Guarantee) .....expires on 31 st December 2002 and as yet the works remain incomplete. I would be grateful if you would re-submit this guarantee to commence from 1 st January 2003 and which is to remain valid until Substantial Completion is achieved. On 8 th January 2003 HSS wrote to ABP: As you hold correctly, the expiry date of the Corporate Guarantee is the 31 st of December, 2002. The guarantee was issued by us as agreed and its terms were fully accepted by you. With respect to this and before proceeding, we'd like to ask you to clarify on which contractual ground your request is based. On 14 th January 2003 ABP replied: I do not concur with your view that the Parent Company Guarantee (Corporate Guarantee) was agreed and accepted by ABP I refer you to my letter dated 29 July 2002.....confirming receipt of the document in which I draw to your attention the matter of the expiry of the guarantee. In that letter I requested that you amend and re-issue the guarantee and it is on this basis that I now request you re-issue the guarantee with sufficient currency to provide a Parent Company Guarantee until completion of all your contractual obligations. Mr. Williamson's arguments were these. First, no particular form of guarantee was annexed to the contract. It was thus for the parties to agree between them an appropriate and acceptable form of guarantee. That argument is based on a mistaken premise. A particular form of guarantee was annexed to the contract. Second, the parent company provided a guarantee expiring on 31st December 2002 which was accepted by ABP, who paid £4,000 for it in July 2002. That satisfied HSS's contractual obligations. If some greater obligation were imposed by clause 10 of the contract, it was expressly varied and/or waived by accepting the guarantee that was in fact provided. Mr. Williamson did not suggest that the £4,000 was paid in settlement of any dispute. It does not appear why ABP paid that sum. Mr. Williamson did not rely on any evidence to support the propositions contained in HSS's letter of 8 th January 2003, namely that the original guarantee was "as agreed" and that its terms were "accepted", which are plainly contradicted by the rest of the correspondence. There is clearly no variation or waiver. HSS is in breach of contract in failing to procure the parent company guarantee beyond the end of December 2002. I am not prepared to order specific performance of an obligation to procure a third party to enter into a guarantee. The parent company has no obligation to enter into the guarantee. The damages claimed for breach of contract are not specified, but if and in so far as they arise out of any incapacity of HSS to satisfy any judgment for damages against it, the claimed relief seems futile. I am, however, prepared to hear counsel further on any relief that should be granted. The part 20 claim against Haecon Meaning of fitness for purpose In the contract between HSS and Haecon ("the design contract") HSS was referred to as the Main Contractor and Haecon was referred to as the Designer. Article 1 of the design contract provided, so far as material: The administrative and technical provisions of the Main Contract.....constitute an integral part of the present contract, taking into account that the Designer takes over all rights and obligations of the Main Contractor in relation to the Principal with regard to the works entrusted to the Designer. The Designer declares to be acquainted with and to accept the terms and conditions of the main contract, insofar as these are related to the works to be carried out by the Designer, which terms and conditions are deemed to be literally repeated and printed. ..... Article 2 provided, so far as material: 2.1. The Designer is instructed by the Main Contractor with the works described herein as part of the Main Contract: to expertly carry out all survey and design activities which according to the specifications of fulfilment of the Main Contract are required for the taking over of the design responsibility, namely: a. the examination of the stability of the quay wall and the seaward and landward crane beams..... ..... 2.5. The assignment entrusted to the designer, comprising the works as described in the present Article 2.1 to 2.3. , constitutes for the account of the designer a duty to achieve a given result..... The contract was subject to Belgian law, but no issue arises in relation to that. It has not been suggested that the obligation of Haecon to HSS under the design contract in relation to fitness for purpose differs from that owed by HSS to ABP under the (main) contract, save that it is confined to design. In the particulars of the part 20 claim against Haecon the claim includes the costs of any remedial work for which HSS is held responsible. HSS allege that in the event that the works are found unfit for their purpose, such unfitness was caused or contributed to by the breach of contract of Haecon under the design contract. There is an alternative allegation of negligence which has not been pursued before me. HSS seeks from Haecon an indemnity in respect of any loss or damage that ABP may be entitled to recover from HSS or, in the alternative, a contribution or damages. The design contract does not provide for an indemnity in the event of its breach. A contribution as such cannot, in my judgment, be ordered. Any remedy must, in my judgment, be by way of damages. The measure of damages will be any loss suffered by HSS which is attributable to breach of the design contract and which would not otherwise have been suffered by HSS. The respects in which I have found the works to have been unfit for their purpose are the following: (1) The insufficiency of the shear strength of the grout wall. The design of the columns was unfit for its purpose. (2) The spacing of the columns was in places greater than the original average design spacing of 1.7m. That increased the insufficiency of the shear strength of the grout wall. (3) There was loss of composite action between the members of some of the column pairs. That was the result of a combination in the case of each such pair of two factors, namely (a) non-parallelism of the columns and errors in setting their positions; and (b) diminution of the diameters of the columns with depth. (4) The grouting works were unfit for their purpose. HSS has failed to satisfy me that that was a matter of design. As to point (2), the original design spacing could not everywhere be adhered to. Revised setting out was designed by Haecon, as appears from the following evidence of Mr. Wijffels given in the course of his cross-examination by Mr. Williamson (day 6, pages 77, 79). Mr. Williamson was putting to Mr. Wijffels questions relating to the setting out of the grout columns: Q. So in relation to the design of the setting-out, as it were, that is a matter that Haecon were taking responsibility for as between themselves and HSS? A. Yes. However they were set, they were not able to instal them at locations that we initially provided. ..... Q. It is right, is it not, that Haecon sent to site a number of personnel physically to set out the columns? A. To assist. Q. To say where they were to go? A. Yes. Thus the as-built spacing was also a matter of design. As to 3(a), the workmanship was within acceptable tolerances. The design should have allowed for such tolerances. Accordingly, (a) was a matter of design. I have found that (b) was not a matter of design for which Haecon is responsible: see paragraph 276 above. Thus I find that Haecon was in breach of the design contract in respect of items (1), (2) and (3)(a). On the question of damages, I make the following finding as to the effect on loss of composite action of 3(b), given 3(a). I have carried out an exercise similar to that mentioned in paragraph 285 above on the hypothesis that the diameters at depth of the columns of a pair were 900mm as designed. I have taken the example of two deviations each of 130mm: the results are set out in Appendix 5. The probability of loss of composite action in that case comes out at approximately 9 per cent. It must be borne in mind that for any possibility of loss of composite action the sum of the individual deviations of the columns must be at least 200mm in this case. I find that the probability of that is rather remote. I am not satisfied that 3(a) would have caused any loss of composite action but for 3(b). I shall hear counsel on the form of the order that I should make, including quantum. Appendix 1 Curvature of sheet pile at location X1 using polynomial representation of sonde profile. Paragraph 52 refers. Notes. 1. The y values are horizontal measurements from the vertical. 2. The figures in the column Read y(m) (1) are read from figure 1 in a note (X97) of Professor Pavlovic. He stated the source as "Mr. Bracegirdle's report". Accuracy ±.006m. 3. The figures in Read y(m) (2) are read from figure 9 in a report of Mr. Bracegirdle dated 7.10.05. Accuracy ±.011m. 4. For x-values between +1 and -14 the sonde was within 200mm of the pile, so the output was saturated. Consequently, the read values of y are unreliable. 5. The polynomials have coefficients A(n) of x to the power n; x to the power 0=1. Their values are as follows: Polynomial A(0) A(1) A(2) A(3) A(4) A(5) A(6) A(7) A(8) A(9) 4th 0.03086 0.0066 7.42715 E-4 -1.24624 E-4 -4.35186 E-6 0 0 5th 0.02925 0.01374 0.00109 -3.15844 E-4 -2.34033 E-5 -4.7216 E-7 0 0 6th 0.01759 0.01768 0.00271 -3.67798 E-4 -5.94258 E-5 -3.01422 E-6 -5.21219 E-8 0 9th 0.01885 0.01505 0.00239 -1.77477 E-4 -2.94541 E-5 -4.58921 E-6 -6.29196 E-7 -4.39401 E-8 -1.41544 E-9 -1.6967 E-11 Appendix 2 Curvature of sheet pile at location X2 using polynomial representation of sonde profile. Paragraph 52 refers. Notes 1. The values of y are horizontal measurements from the vertical 2. The y-values in the second column were read off the profile of X2 in document X30, page 2. The accuracy is no better than ±.015m. 3. The polynomials have coefficients A(n) of x to the power n; x to the power 0=1. Their values appear in the tablebelow. Polynomial coefficients Polynomial A(0) A(1) A(2) A(3) A(4) A(5) A(6) A(7) A(8) A(9) 3rd 0.02005 -0.0389 -0.00223 -3.30918 E-5 0 0 0 4th 0.01745 -0.0393 -0.00203 -4.94899 E-6 8.88107 E-7 0 0 5th 0.0073 -0.0524 -0.00164 5.54704 E-4 5.17728 E-5 1.27092 E-6 0 0 9th -0.01479 -0.05222 0.00457 0.0011 -1.64454 E-4 -3.18206 E-5 -3.45825 E-7 1.7228 E-7 1.03932 E-8 1.81193 E-10 Appendix 3/1 Curvature of sheet pile at location X3 using polynomial representation of sonde profile. Paragraph 52 refers. Notes 1. y is measured horizontally from the vertical. 2. In the second column, the values of y have been read from exhibit X22. The readings cannot be regarded as accurate to within less than ±0.016m. 3. The coefficients of the polynomials are as shown on page 2 of this appendix. The polynomials have coefficients A(n) of x to the power n; x to the power 0=1. Their values are as follows: Polynomial A(0) A(1) A(2) A(3) A(4) A(5) A(6) 4th .03964 -.04662 -.0026 -3.55648 E-5 -1.08208 E-6 0 5th .03129 -.04811 -.00123 2.83723 E-4 2.11814 E-5* 4.93219 E-7 6th .00935 -.04697 -.00139 1.57915 E-4 3.76905 E-6 -4.28843 E-7 -1.69153 E-8 *There appears to be a typing error in Professor Pavlovic's list. Appendix 3/2 Notes 1. y is measured horizontally from the vertical. 2. In the second column, the values of y have been read from exhibit X22. The readings cannot be regarded as accurate to within less than ±0.016m. 3. The polynomials have coefficients A(n) of x to the power n; x to the power 0=1. Their values are given in the table on the next page. Appendix 3/3 Polynomial coefficients. Polynomial A(0) A(1) A(2) A(3) A(4) A(5) A(6) A(7) A(8) A(9) 7th 0.03628 -0.03348 4.77096 E-4 -.00107 -3.40954 E-4 -3.41734 E-5 -1.45534 E-6 -2.26092 E-8 0 8th 0.04283 -0.04641 -0.00646 -4.824E-4 3.01518 E-4 7.66687 E-5 6.9448 E-6 2.77561 E-7 4.12814 E-9 9th 0.05083 -0.04897 -0.01271 -0.00138 7.15576 E-4 2.1559 E-4 2.40921 E-5 1.33171 E-6 3.64409 E-8 3.94035 E-10 Appendix 4 (paragraph 285 refers) This table relates to a horizontal plane at about -20mCD. Consider the points at which the axes of two columns of a VHP column pair cut that plane, and the points at which the intended axes (vertical) of those two columns would have cut that plane. The distance of each actual point from its corresponding intended point is taken as 75mm. The line joining the two points of each pair I call the deviation. The angle θ(1) is the angle between the deviation of column 1 and the line joining the intended point of column 2 to the intended point of column 1, produced. The angle θ(2) is the angle between the deviation of column 2 and the line joining the intended point of column 1 and the intended point of column 2. The range of θ(2) is the range of angles θ(2) such that the distance s between the two actual axes ≥800mm. P[θ(n)] is the probability that the angle θ(n) will lie within the range shown. It is assumed that all directions of displacement are equally probable, so that the probability is proportional to the range. The problem is symmetrical in this sense, that P[θ(1)] = P[360°-θ(1)], and the value of θ(2) corresponding to 360°–θ(1) is 360° minus the value of θ(2) corresponding to θ(1). The last column gives the probability that s≥800mm. The total probability is approximately 12 per cent. It is assumed that the probabilities of the directions of the deviations are mutually independent. Appendix 5 (Paragraph 368 refers) This table relates to a horizontal plane at about -20mCD. Consider the points at which the axes of two columns of a VHP column pair cut that plane, and the points at which the intended axes (vertical) of those two columns would have cut that plane. The distance of each actual point from its corresponding intended point is taken as 130mm. The line joining the two points of each pair I call the deviation. The angle ?(1) is the angle between the deviation of column 1 and the line joining the intended point of column 2 to the intended point of column 1, produced. The angle ?(2) is the angle between the deviation of column 2 and the line joining the intended point of column 1 and the intended point of column 2. The range of ?(2) is the range of angles ?(2) such that the distance s between the two actual axes = 900mm. P[?(n)] is the probability that the angle ?(n) will lie within the range shown. It is assumed that all directions of displacement are equally probable, so that the probability is proportional to the range. The problem is symmetrical in this sense, that P[?(1)] = P[360°-?(1)], and the value of ?(2) corresponding to 360°–?(1) is 360° minus the value of ?(2) corresponding to ?(1). The last column gives the probability that s=900mm. The total probability is approximately 9 per cent. It is assumed that the probabilities of the directions of the deviations are mutually independent. Note 1 For explanation of plastic hinges, see paragraph 45 below. [Back] Note 2 URS Corporation Ltd. See paragraph 159. [Back] Note 3 See paragraph 29 below. [Back] Note 4 Larssen 6 is the type of sheet pile used in this contract. [Back] Note 5 Kilonewton metres squared per metre length of the wall. [Back] Note 6 Paragraph 48 above. [Back] Note 7 Paragraph 42 above. [Back] Note 8 Paragraphs 37 and 42. [Back] Note 9 Paragraph 45. [Back] Note 10 Paragraph 57 above. [Back] Note 11 Mr. Claessens. [Back] Note 12 Paragraph 12 above. [Back] Note 13 Paragraphs 193 and 198 above. [Back] Note 14 Paragraph 160 above. [Back] BAILII: Copyright Policy | Disclaimers | Privacy Policy | Feedback | Donate to BAILII