Ove Arup & Partners International Ltd. & Anor v Mirant Asia-Pacific Construction (Hong Kong) Ltd. & Anor

The first claimant, Mirant Asia-Pacific Construction (Hong Kong) Limited, is a company which designs, constructs and commissions coal fired power stations. Mirant was formERLY known as CEPA Slipform Power System Limited (“CEPAS”). The second claimant, Sual Construction Corporation, is a subsidiary of Mirant incorporated in the Phillipines in order to undertake the design, construction and commissioning of a power station at Sual in the Phillipines.

The first defendant, Ove Arup & Partners International Limited (“OAPIL”), are international engineering consultants. The second defendant, Ove Arup and Partners Hong Kong Limited (“OAPHK”), are engineering consultants practising principally in the Far East.

The claimants claim against the defendants damages for breach of contract in relation to the construction of the power station at Sual. The claims arise out of the failure in about April 1997 of two of the foundations of Boiler House No. 1.

This is the defendants’ appeal against a judgment of His Honour Judge Toulmin CMG QC given in the Technology and Construction Court on 11^th June 2003. The judge decided five preliminary issues, one of which was amended by agreement between the parties after the hearing but before the judge gave judgment.

The preliminary issues concerned two agreements. One was for engineering design including the design of the unit 1 boiler foundations: the other was for ground investigation including obligations to inspect foundation excavations for the unit 1 boiler, to inspect formations and to approve ground conditions. There were issues as to the parties who entered into these agreements. The judge decided that the parties to the design agreement were CEPAS and both OAPHK and OAPIL; and that the parties to the inspection agreement were CEPAS and OAPIL. This is no longer contentious but, if the defendants succeed in their appeal on the second issue which the judge decided, it is agreed that the parties to the design agreement were CEPAS and OAPIL. Since the identity of the parties to these agreements is no longer contentious, I shall refer to the claimants as “CEPAS” and to the defendants as “Arup”.

The central issue between the parties is whether both or either of the agreements incorporated the terms of the 1991 FIDIC Client/Consultant Model Services Agreement including a 5 year limitation of liability period and a limit of compensation of £4m.

Those principally involved in the relevant contract negotiations were Mr Elliott on behalf of CEPAS and Mr Roberts and Mr Higson on behalf of Arup. The judge described Mr Elliott as a very successful businessman used to managing large construction projects. The contract with Arup was a relatively small part in the large overall contract to build the power station at Sual. Mr Elliott’s style of management was to look at what he regarded as the important questions and make a decision in principle leaving others to consider the detail. At the time of the hearing, he had left CEPAS. He had obligations under a confidentiality agreement, breach of which might render him liable to very adverse financial consequences. He was called to give evidence by Arup pursuant to a witness summons.

Mr Roberts is the Director responsible for energy products at OAPIL. He is not a lawyer, but he has had years of experience in all aspects of contract and project management, including the drafting of major international contracts. Mr Higson is a civil engineer who has been project director on major projects.

The development of Arup’s proposals for the design work began in 1994. In April 1995 they produced revised proposals. There were discussions in April and May 1995 culminating in a formal letter dated 23^rd May 1995. The judge found that up to this stage FIDIC terms had not been discussed because Arup did not think that these would be acceptable to CEPAS. A version of this letter was signed by Mr Elliott on behalf of CEPAS and Mr Higson on behalf of Arup on 29^th May 1995 and has been referred to as a letter of intent. The judge recorded that Arup no longer pursued a contention that this letter did not constitute a binding agreement between the parties. It clearly did.

The letter of intent defined Arup’s work scope as that detailed in their April 1995 Revision C offer. The fee for concept, preliminary and detailed design of the defined scope of works was to be £3.75 million. It was provided that:

Work on a more detailed draft contract was undertaken by Arup in June and July 1995. There was a meeting on 11^th August 1995 at which Arup handed to CEPAS a Revision A draft of a FIDIC Model Services Agreement. Those present at the meeting included Mr Elliott and Mr Givelin for CEPAS and Mr Roberts and Mr Higson for Arup. Mr Givelin is a partner in the firm of Frank and Vargeson, who are quantity surveyors. Neither Mr Elliott nor Mr Givelin had had any opportunity to see the Revision A draft in advance of the meeting. The judge found that, in general, Mr Elliott was prepared to agree the terms of the contract (which in matters important to him were a reiteration of the letter of intent), but this was subject to a detailed examination of the proposed terms by Mr Givelin. He found that in this very limited sense Mr Elliott signed off the contract on 11^th August 1995 subject to a final agreement. A very brief one page manuscript note of the meeting by Mr Roberts contains the words “FIDIC 91 agreed as basis”. The judge does not appear to have had this note in mind, since he said that there was no record of the meeting. He found that Mr Elliott went no further at the meeting than to say that he expressed no immediate objection to the draft contract. The Revision A proposal included conditions of special application providing a limit of compensation of £4m., a limitation of time for bringing proceedings and a completion date. The judge found that these were matters which Mr Elliott left to his advisors to comment on and, if necessary, to negotiate. He was satisfied that Mr Elliott’s actions were substantially influenced by the fact that, as far as he was concerned, he had concluded the basis of an agreement in May 1995. The important outstanding details for him concerned schedules of rates for payment and the programme of works. It was at one stage Arup’s case that a contract was concluded at this meeting on the Revision A terms. It was however conceded in the course of the trial that there was no such binding agreement.

On 15^th August 1995, Arup wrote enclosing monthly invoices. The letter stated that “the amounts requested reflect the letter of intent between CEPAS and Arup dated 23^rd May and the draft contract issued to you on 11^th August 1995”. This formula was used on each subsequent occasion when Arup submitted invoices up to and including 25^th June 1996.

Mr Givelin reviewed the Revision A draft agreement. He wrote to Mr Elliott on the 17^th August 1995 drawing his attention to salient points. He said that the basic agreement was modelled on the FIDIC standard form “which is fair to both the client and Consultant”. He drew attention to particular matters addressed by the Appendices and Annexes. These included:

There were eleven other matters in Mr Givelin’s list. He advised that, subject to Mr Elliott’s views on these matters, the agreement could be finalised speedily. Mr Givelin’s advice was internal to CEPAS.

On 29^th November 1995, Arup sent to CEPAS a first draft of their Proposal for Ground Investigation. A subsequent version of this, dated March 1996, is a crucial document in this appeal. I shall consider its details later in this judgment. The November 1995 version already contained clause 1, which purported to incorporate the Model Services Agreement “entered into” by CEPAS and Arup for the design of the civil engineering works of the Sual power station. On the judge’s now unchallenged findings, the Model Agreement had not been entered into in November 1995.

On 6^th December 1995, Arup sent to CEPAS Revision B of the draft Model Agreement. They pointed out differences between this version and Revision A which had been discussed on 11^th August 1995. There were changes to the overall project programme. A Schedule of Rates for additional services had been included. A different payment schedule based on the design programme was included with a graph showing payments to be made over time and a list of key progress milestones. Arup said that they had received no comment on the original draft and that they would like to discuss the contract with CEPAS as a matter of priority. The judge said that the changes between Revision A and Revision B were of some importance, although the payment schedule was at all times within the price of £3.75m. agreed in the Letter of Intent.

The Revision B draft FIDIC Model Services Agreement consisted of (a) a form of Agreement, (b) Standard Conditions, (c) Conditions of Particular Application, (d) Appendices and (e) Annexes. The form of Agreement was a two page document providing for formal signature by the parties in the presence of witnesses. It provided for the documents which were to form part of the Agreement, including a Letter of Acceptance, the Standard and Particular Conditions and the Appendices including Appendix C - Remuneration and Payment. The consultant agreed to perform the stipulated services. The client agreed to pay the consultant for the services “such amounts as may become payable under the provisions of the Agreement at the times and in the manner prescribed by the Agreement.”

The Standard Conditions defined the Agreement as meaning “the Conditions … [the Appendices] … Letter of Acceptance and Formal Agreement if completed, or otherwise as specified in Part II.”

(a) Clause 5(i) of the Standard Conditions provided that the consultant should exercise reasonable skill, care and diligence in the performance of his obligations under the Agreement. Clause 16 provided that the consultant should only be liable to pay compensation to the Client if a breach of Clause 5(i) was established against him. Clause 17 provided:

The Conditions of Particular Application (Part II) provided that the duration of liability should be 5 years; that the Limit of Compensation should be £4 million; that the commencement should be on 3^rd June 1995 and completion on 31^st March 1997; that the time for payment should be as stated in Appendix C and that the agreed compensation for overdue payment should be UK base rate plus 3%; that the ruling language should be English and the law to which the Agreement was subject should be English Law; and that Arbitration should be carried out under UNCITRAL Arbitration Rules.

Annex C3 was a Payments Schedule. One page of this was a spreadsheet showing projected monthly payments in sterling between September 1995 and May 1997. Each month’s payment was the sum of individual amounts for each section of the design work with separate amounts for Design Management and Design Management Profit. The details of this spreadsheet had been altered significantly from that submitted with Revision A in August 1995, so that there were larger total payments in the earlier months.

There were meetings in Jakarta between 8^th and 10^th December 1995 attended by Mr Elliott, Mr Roberts and Mr Higson and others. The judge found that the meetings dealt largely with technical matters. It was not until item 19 of the note of the meeting that the contractual question was raised at all. It may not have been discussed until 10^th December after Mr Higson had left the meeting. There is no direct reference to Revision B having been discussed. The judge found that there may have been a general discussion about Revision B, but he was not satisfied on the evidence that the Conditions of Specific Application were discussed in any detail. Mr Bartlett QC on behalf of Arup accepts this, but points out that these conditions were not controversial. The judge said that as far as the main terms of business were concerned, Mr Elliott thought that he had signed off the contract in May 1995. The item, which was more than a detail to him and which was new, was the payment schedule. He did not agree this and it was to be the subject of further discussion. Arup’s evidential case was that Mr Elliott had expressed agreement with all items in Revision B except the payment schedule. Mr Roberts said that he knew Mr Elliott’s methods of working and that, since Mr Elliott did not object to Revision B, that indicated positive agreement. The judge said that this evidence could be tested against subsequent correspondence.

It was at one stage Arup’s case that a contract for the design works was concluded at the Jakarta meetings. The judge held that it was not. There is no challenge to this finding on this appeal.

On 26^th January 1996, Mr Roberts wrote to Mr Elliott saying that, following their discussions in Jakarta, he had revised the payments schedule “as we agreed”. The changes are described. The design activity on roads and site services had been extended to a total of 20 months. There were inserted intermediate milestones in the delivery of construction information. The profit on each of Arup’s activities now related to an early construction milestone. The overall programme had been adjusted to the current programme to which Arup were committed. Mr Roberts asked Mr Elliott to review and agree the payment schedule so that they might insert it into the contract documents between CEPAS and Arup. Should it become necessary to adjust the programme further as a consequence of the effective date of the actual construction contract, Mr Roberts proposed that the payments should be adjusted in the way described in his letter of 15^th November 1995. The letter enclosed a revised payments spreadsheet with a supporting schedule of Key Milestones for each activity with a percentage payment of cost attributed to each.

There was no response to this letter of 26^th January 1996. The judge held that there was no evidence that the payment schedule and the milestone documents were ever agreed.

Arup had a number of internal documents for the purpose of quality control. In one such document dated 29^th January 1996, Mr Roberts wrote that the client’s brief was only currently accepted in principle by CEPAS. There was no written confirmation. “Refer to Letter of Intent. No response yet from CEPAS on contract we submitted.” In a letter dated 31^st January 1996, Mr Fox commented that he believed that, following the contract meeting with Mr Elliott, Mr Higson and Mr Roberts did not agree on what was and was not included in Arup’s offer.

On 1^st February 1996, the main construction contract with GEC Alsthom became effective. There were no further discussions after this date on the Revision B Model Services Agreement.

On 6^th February 1996, Mr Higson sent a fax to Mr Roberts in which he said that Mr Elliott had “signed off verbally” and “instructed to proceed” at a meeting on 15^th September 1995. On the same day, Mr Roberts wrote a long memorandum to Mr Higson. This stated that there had been something of a breakdown in relations with Mr Elliott which called in question the basis of trust between them. Within the last few days, Mr Elliott had separately told Mr Higson and Mr Roberts that he had not approved Arup’s concept drawings and that Arup were “working at risk”. Arup should not be undertaking detailed design work at this stage. The memorandum also referred to the meetings in Jakarta. It stated that Mr Elliott had indicated his willingness to accept Arup’s payments schedule subject to tying payment of their profit element of each activity into the completion of an early construction milestone and with a number of other small changes. The judge noted that there was no mention of either Revision A or Revision B having been orally agreed. Nor was there any suggestion that at that stage Mr Roberts understood that Mr Elliott’s silence could be taken as assent. On the contrary, Mr Elliott had been saying that Arup were working at risk.

On 7^th February 1996, Mr Roberts and Mr Higson produced an internal Project Quality Plan. This stated that CEPAS had issued a letter to Arup expressing its intention to enter into an agreement for the design of certain of the civil engineering works in connection with the Sual Power Station. It also stated under the heading “Agreement with Client”, that, until the Agreement came into force, Arup were providing the Services on the basis of a Letter of Intent signed by the client on 29 May 1995. The Form of Agreement was referred to as the FIDIC Model Services Agreement. Mr Bartlett suggests that this reference shows that CEPAS had agreed to the FIDIC Model Agreement. In my view, this reads too much into the document.

On 9^th February 1996, Mr Fox wrote a memorandum to Mr Higson recording that, following discussions with the client the previous week, Mr Roberts had asked them to review “his draft contract” and to revisit what Arup had undertaken to provide to CEPAS at each stage of the project and to highlight what they had not yet delivered.

On 22^nd February 1996, Arup issued the first Revision of their Proposal for Ground Investigation. This was a revision of the proposals sent to CEPAS on 29^th November 1995. It was further “Revised for Client Agreement” on 8^th March 1996 and presented to Mr Elliott on 15^th March 1996. On that or the following day, Mr Elliott initialled the last page of this document and annotated it “OK”.

The proposal provided, under the heading “Agreement”:

The scope of the ground investigation work was defined. Arup proposed to provide one full time supervising geo-technical engineer. He was to be charged at the monthly rate of £5,150. It was provided generally that the supervision of the geo-technical and geo-physical investigations should be paid for on a reimbursable basis against a schedule of rates. Invoicing was to follow the procedure laid down in item C4 of Appendix C of the Main Agreement. Variations to the scope of the ground investigation services were to be dealt with in accordance with item C3 of Appendix C of the Main Agreement.

The Ground Investigation Proposal was self contained as to payment. Thus, although it needed for its operation to import terms from the Model Services Agreement relating to invoicing and variations, the fact that the payments schedule for the Design Agreement had not been agreed was not material to the operation of the Ground Investigation Proposal.

Arup’s case is that Mr Elliott’s acceptance of the proposal, by signing it “OK” and initialling it, constituted the making of a written agreement. By Clause 1, the terms of the Revision B Model Services Agreement were expressly incorporated. These included the 5 year time limitation and the £4m limitation of liability. Arup’s case further is that, by entering into the Ground Investigation Agreement, Mr Elliott acknowledged that the Main Design Agreement had come into existence on the Revision B terms, alternatively that Mr Elliott’s signature on the Ground Investigation Proposal brought into being the Main Design Agreement on Revision B terms. CEPAS’ case is that, the Main Design Agreement had not been “entered into” up to 15^th March 1996; that Clause 1 of the Ground Investigation Proposal had no content and should be ignored; that accordingly the Ground Investigation Agreement was not upon the Revision B Model Services Agreement terms; and that Mr Elliott’s signature effected an agreement for ground investigation services on the terms on which Arup were proceeding with the Main Design work, which remained those in the Letter of Intent.

Arup continued to submit invoices for the Design Agreement which referred to the Letter of Intent and the draft contract issued on 11^th August 1995.

On 16^th July 1996, Mr Higson and Mr Roberts again signed an internal Arup quality control document stating that action was required to “agree working programme with CEPAS. Obtain agreement to Draft Contract with CEPAS. Obtain agreement on revised schedule of payment with CEPAS.”

On 9^th August 1996, the parties agreed a breakdown of fees and progress assessment for the design work which showed a net amount due to Arup of £693,333 after deducting amounts paid to date. This assessment was not made by reference to any payments schedule which had previously been prepared in relation to the draft Model Services Agreement. It stated Arup’s fee for each section of their work. It then stated a cost for each section. The fee was in each instance the cost plus 10%. Amounts due were calculated for each element as the percentage of the cost which the element was recorded as complete. Neither party made detailed submissions about this document. My own brief comparison between it and the spreadsheet payments schedule which Mr Roberts had sent on 26^th January 1996 indicates that only four or five of the twenty two elements had the same total fee on both documents.

On 14^th November 1996, Mr Fox wrote to Mr Higson and Mr Roberts saying that he considered that there had been no progress in getting the contract agreed and signed. He said that the original payment schedule had now been clearly refuted by the client and the requested revised schedule dismissed. In their place, they had a percentage complete payment release system that appeared to be accepted by the client, although dates or milestones for gross profit payment were not fully defined or agreed.

There were at least three instances after March 1996 when Arup submitted a formal proposal for additional work upon which they were instructed to proceed. These included:

CEPAS instructed Arup to proceed with each of these by documents dated 3^rd August 1996, 9^th December 1996 and 23^rd December 1996 respectively. It is, I think, an open question whether these documents each effected a contract on the terms which Arup had offered, or whether they were no more than instructions to carry out the scope of work upon limited formal terms. It is not in my view necessary to resolve this question. However it might be resolved, Arup relies strongly on their proposals as indicating that a Main Agreement for the design work had by then been entered into.

In April 1997 foundations G2 and G5 failed. It was not then suggested that Arup were responsible for this. In June 1997, CEPAS were taken over. Mr Elliott departed and a new management took over, including Mr Reynolds. On 20^th October 1997, Mr Roberts wrote a letter claiming unpaid fees together with interest at 3% above UK base rate, which was the rate specified in the Revision B Conditions of Particular Application.

On 22^nd January 1998, Mr Reynolds wrote to Mr Roberts saying that there was a dispute between CEPAS and Arup relating to Arup’s outstanding fees. Mr Reynolds proposed an amicable resolution which included a final payment and a proposal to execute a formal agreement for services in relation to the works performed substantially in the form of an agreement between CEPAS and Arup in relation to a different earlier project. Mr Roberts replied on 26^th January 1998. His letter included:

On 16^th February 1998, Mr Higson met Mr Reynolds and another lawyer representing CEPAS. Mr Higson’s memorandum of this meeting to Mr Roberts indicates that his first priority was to secure payment of the outstanding balance of the £3.75m fee. In this he was successful. There was further discussion about the contract issue in the context of negotiating a standard form of agreement for future work. It was evident that Mr Reynolds did not like FIDIC. Mr Higson thought that if they pressed the FIDIC form too hard they might jeopardise “all that lovely future work”. He mentioned that in principle as well as in law Arup probably already had a contract for Sual and both the lawyers remained silent.

On 2^nd May 2001, CEPAS wrote a letter of claim to Arup which is the genesis of the present proceedings. Subject to their contentions on issue 5, which the judge found against them, Arup accept that this letter was a sufficient formal claim for the purpose of Clause 17 of the general conditions of the Model Services Agreement, if that is the agreement the parties entered into.

Arup rely on certain passages in the oral evidence of Mr Elliott, Mr Roberts and Mr Higson. A purist might say that much of this evidence was of questionable admissibility. It consists largely of the unspoken subjective understanding of the witness. Arup rely on passages in the opinions of the House of Lords in Carmichael v National Power plc[1999] 4 All ER 896. I shall refer to this authority later in this judgment.

Mr Elliott was asked about his understanding at the time of the meetings in Jakarta on 8^th to 10^th December 1995. He had already agreed a lump sum price of £3.75m and as far as he was concerned “the deal was done”. He had no objection to the component parts of the Revision B Model Services Agreement. He regarded the arrangements for payment as an administrative matter. “I was quite happy that they had got it all concluded and it would just go in for administration, not that there was any more to agree.” As to the proposal for ground investigation services, he believed that his signature against the word “OK” was more than sufficient to bind the parties. As to Arup’s work on the design of the civil engineering works, his understanding was that Arup were proceeding at a price of £3.75m in accordance with the agreement that had been put on his desk. By this he meant the Revision B Model Services Agreement.

Mr Roberts agreed in relation to the Ground Investigation Proposal that he intended that, if there was a contract or when there became a contract, this variation should be subject to the same terms as the main agreement. In answer to a question from the judge, he agreed that his case was that the ground investigation agreement was consistent with the agreement incorporating the FIDIC conditions which he understood had been made. It was one further piece of evidence to him that Mr Elliott was happy with FIDIC for design and for site supervision. Mr Higson believed that Arup had agreement with CEPAS on all the key terms that he believed were important before this event, but a signature on a document that referred directly in his view to the main agreement suggested that the terms of the main agreement had been accepted. The judge recorded oral evidence of Mr Higson to the effect that an agreement developed between the two parties. Mr Elliott clearly understood what Arup had proposed to him. He would have been very clear if he had disagreed. Mr Higson was absolutely convinced of that. Arup had worked before on a contract with CEPAS that had never been signed. He was familiar with Mr Elliott and signatures. He never had any doubt that Mr Elliott knew he was signing “next to FIDIC”.

Mr Reynolds, who appeared on the scene much later, said that it appeared to have been accepted by both parties, and CEPAS’ contracts team considered, that the scope of the work in Revision B constituted a binding commitment.

As to the nature and quality of the evidence, the judge bore in mind that most of the events took place 7 to 8 years ago. Much had happened to the witnesses in the meantime, which undoubtedly affected the quality of their evidence. There were also occasions when the judge felt that the witnesses were tailoring their evidence to the particular case that was being put forward. He considered that the ability to test the recollection of witnesses years after the events by reference to contemporaneous documents assumed particular importance. The judge referred to oral evidence relevant to the Ground Investigation Proposal which Mr Elliott signed in March 1996. He again said that the accuracy of the evidence of the witnesses called by Arup had to be tested against the documentary evidence of the events which followed. He said that neither Mr Roberts nor Mr Higson suggested that, by signing the Ground Investigation Proposal, Mr Elliott was agreeing to something to which he had not previously agreed.

The issues decided by the judge which are the subject of this appeal are:

As to Issue 2, the judge decided that the agreement for design services between the parties was formed on or about 29^th May 1995 by the signing of the Letter of Intent. Its terms did not include FIDIC terms with a 5 year cap on liability and a £4m cap on the amount of recovery. FIDIC terms including those limitations were not subsequently agreed so as to be incorporated in a varied or substituted agreement.

As to Issue 4(d), the judge decided that the ground investigation agreement was not subject to the FIDIC conditions including conditions of particular application contained in Revision B of the draft contract dated 15^th November 1995. His reasoning was contained in two short paragraphs of his judgment as follows:

As to Issue 5, the judge decided that, if the FIDIC terms had applied, the claimants’ letter dated 2^nd March 2001 was a formal claim within the meaning of Clause 17 of the conditions.

The judge’s decision on Issue 2 dealt with a number of matters which are not the subject of the present appeal. Historically, Arup had variously contended that there was no binding contract on the terms of the Letter of Intent dated 29^th May 1995; that an agreement on FIDIC terms was concluded on 11^th August 1995; and that an agreement on FIDIC terms had been concluded in Jakarta on 8^th to 10^th December 1995. By the conclusion of the hearing before the judge, Arup no longer pursued the first two of these contentions. The judge decided that no agreement on FIDIC terms was concluded in Jakarta in December 1995. On findings of fact which he made, this was clearly correct. In particular, the judge held that the payments schedule, which Mr Elliott regarded as important, was not agreed then or subsequently. Arup do not challenge in this court the judge’s finding that no binding agreement was made in Jakarta in December 1995. Nor, in my view, is the judge’s analysis of the parties’ dealings in January and February 1996 reasonably open to challenge. His conclusion in short was that there was no point at which the parties did or said anything which could constitute the making of an agreement on FIDIC terms; that Arup did not themselves regard Revision B as having been agreed; and that no version of a payments schedule was agreed. He held that thereafter Revision B was not discussed between the parties at all. The judge did not, other than in passing, address the contention, which is the main plank of Arup’s appeal to this court, that, by signing the Ground Investigation Proposal with its explicit reference to the Model Services Agreement entered into between the parties, Mr Elliott thereby acknowledged that those terms applied to the design contract so that a contract on those terms then came into existence. On this submission, the judge only said that the evidence of both Mr Roberts and Mr Higson was that Mr Elliott understood that FIDIC terms already applied, but that neither suggested that by signing the Ground Investigation Proposal, Mr Elliott was agreeing to something to which he had not previously agreed.

It is perhaps understandable that the judge dealt with this matter shortly, since the main burden of his analysis was directed to Arup’s contention that a binding agreement on FIDIC terms had been made in Jakarta.

In the course of his discussion, the judge made a number of factual findings which are not in my view open to challenge in this court. In relation to the meeting on 11^th August 1995 at which Arup proposed Revision A, he said that neither Mr Elliott nor Mr Givelin had seen the draft in advance of the meeting.

The judge found that during the Autumn of 1995 there was no immediate contractual difficulty. Arup were perfectly able to continue to work to a programme which they had until another programme was agreed as part of the main contract. They were also able to agree a mechanism for dealing with change order requests.

Of the meeting in Jakarta in December 1995, the judge found:

The judge held that, in order for an agreement incorporating the FIDIC form to be effective, the pages that were blank were required to be completed and signed by the parties. Mr Bartlett challenges this. He submits that the words “if completed” in the definition of “Agreement”, and the words “if any” in Clause 21 of the Standard Conditions mean that the signing of the Formal Agreement is optional and that the receipt by the Consultant of the Clients Letter of Acceptance is an alternative way of making the Agreement effective. I agree with this submission. He submits that the form of any letter of acceptance is immaterial and that Mr Elliott’s signature besides “OK” on the Ground Investigation Proposal is sufficient as a letter of acceptance. I agree with this to the extent that Mr Elliott’s signature evidenced and constituted his acceptance of the proposal which was before him. It remains to consider what the contractual effect of the agreement so constituted was. More generally, I do not consider than an agreement on these FIDIC terms is incapable in law of being made other than by the formal signing of the Form of Agreement itself; nor, depending on the facts, that an agreement on these FIDIC terms is incapable of being made by means of a sufficiently clear and unequivocal oral acceptance of an appropriate written proposal. Parties would no doubt be well advised to reduce agreements such as these entirely to writing. Mr Bartlett also submits, if necessary, that the parties waived the need for a formal written agreement for the design services. I do not consider that this submission adds anything to the necessary structure of Arup’s case. The same applies, in my view, to arguments based on estoppel. If the agreements for which Arup contend were made by acceptance or conduct, that is sufficient for their case. If either of the agreements were not so made, the necessary ingredients of an agreement by estoppel or waiver of formalities are not established.

The judge considered whether the conduct of the parties subsequent to 26^th January 1996 indicated that they acted in relation to the design agreement on the basis of a common underlying assumption that Revision B had been agreed. He regarded as relevant the fact that Arup sent invoices after December 1995 which made no reference to Revision B and that these were paid by CEPAS. He noted that Arup’s internal documents continued to recognise that there was no signed contract although a draft FIDIC contract had been presented to CEPAS. He noted that Arup did not unequivocally assert that there was a contract on Revision B terms when they were later in dispute with CEPAS’ “new management” about their fees. It was only stated that both parties had generally been working to the spirit of the unsigned contract. The judge considered that working generally to the spirit of the unsigned contract was insufficient to amount to assent to terms which substantially curtailed CEPAS’ rights to bring proceedings and reduce the amount of damages which they could recover. He was not satisfied that the parties acted in a way which related unequivocally to Revision B rather than to an agreement based on the Letter of Intent. Mr Bartlett points out that the judge did not at this stage consider the relevance and effect of the three further proposals for additional work which also referred to a main agreement on FIDIC terms and upon which CEPAS instructed them to proceed.

Orthodox legal analysis habitually identifies certain features necessary for the formation of a binding agreement. There has to be unequivocal acceptance of an offer which stipulates terms essential to the existence of the agreement. The absence of matters of detail may not prevent a binding agreement coming into existence. Acceptance may be by conduct, but the conduct needs to be clearly and unequivocally referable to the agreement contended for. Mere silence will usually not be sufficient, but silence accompanied by a course of conduct may contribute to a conclusion that an agreement came into being. It is an objective analysis. Subjective intention or understanding, unaccompanied by some overt objectively ascertainable expression of that intention or understanding, is not relevant. In order to conclude that a binding agreement came into existence, it is usually regarded as necessary to be able to identify the time at which this occurred. If this is not entirely possible, it is nevertheless necessary to spell out the factual process by which objectively the parties reached a common agreement binding each of them.

In Pagnan v. Feed Products Limited [1987] 2 Lloyds LR 601, Bingham J, as he then was, held on the facts that the parties intended to and did make a mutually binding contract on a particular date. Although certain common terms were not then agreed, neither party intended express agreement on these terms to be a precondition to any concluded contract. Thereafter they were not negotiating terms, agreement of which was to be a precondition of contract, but sorting out details against the background of the concluded contract. The Court of Appeal upheld the judge’s decision.

Certain principles may be taken from Bingham J’s judgment at page 610 to 611 and Lloyd LJ’s judgment at page 619. The court’s task is to review what the parties said and did and from that material to infer whether the parties’ objective intentions as expressed to each other were to enter into a mutually binding contract. The task is to discern and give effect to the objective intentions of the parties. The court is not concerned with what the parties may subjectively have intended. Where correspondence and exchanges between the parties have continued over a period, the court must consider all these exchanges in context and not seize upon one episode in isolation in order to conclude that a contract has been made. Where the parties have not reached agreement on terms which they regard as essential to a binding agreement, there can be no binding agreement until they do agree those terms. But the parties are to be regarded as masters of their contractual fate. It is their intentions which matter and to which the court must strive to give effect. The more important a term, the less likely it is that the parties will have left it for future decision. But there is no legal obstacle preventing the parties agreeing to be bound now while deferring important matters to be agreed later. Even if the parties have reached agreement on all the terms of the proposed contract, nevertheless, they may intend that the contract shall not become binding until some further condition has been fulfilled.

In Atlantic Marine Transport Corporation v Coscol Petroleum Corporation(“The Pina”) [1992] 2 Lloyds LR 103, the issue on appeal to this court was which of two versions of the Shelltime 3 form of charter-party was used by brokers when negotiating a time charter between the parties. It was common ground that a contract was concluded at the latest on 27^th December 1979; that the contract was not varied thereafter except by certain addenda; that the signature of the addenda did not directly vary any term relating to performance; that the charter-party was intended to be signed by both parties; that there was no agreement to dispense with signatures or waiver of that requirement; and that the use during the charter period of working copies of the charter-party had no contractual effect as distinct from affording possible evidence of acquiescence in its terms. It was argued for the owners that the unsigned charter-party constituted the agreement in writing between the parties. The court regarded this as unsustainable. But it was submitted that the fact that the parties entered into five addenda cured the want of signature of “the draft charter-party”. The argument relied on the use in the addenda of the expression:

Mr White QC for CEPAS relies on this decision for the proposition that signature of the Ground Investigation Proposal in March 1996 could not have the effect of bringing into force a contract for the design work which had not otherwise been concluded. Mr Bartlett submits that the facts are not entirely parallel. This is correct, but they are quite close in that (1) in each case there was a previously concluded agreement (the Letter of Intent in the present case), and (2) the agreement contended for (the unsigned charter-party in The Pina and the Revision B Model Services Agreement in the present case) had not been entered into at the time of the addendum said to have the effect of bringing it into force. I regard The Pina as a helpful illustration, but it does not, I think, propound any principle binding this court in the present appeal. The question is whether what the parties did in March 1996 (and on the three subsequent occasions when additional works were instructed) is to be seen objectively as evincing an intention to enter into an agreement for the design work on the Revision B terms which up till then they had not made.

In Carmichael v National Power plc[1999] 4 All ER 897, the issue was whether tour guides who accepted in writing an offer of employment; “on a casual as required basis” were employees under contracts of employment or whether they were not in any contractual relationship when they were not working. The Court of Appeal, reversing decisions of the Industrial Tribunal and the Employment Appeal Tribunal, treated the matter solely as one of construction of the documents, and as one of law rather than of fact. The House of Lords reversed the decision of the Court of Appeal. Lord Irvine LC considered (at page 901e) that it would only be appropriate to determine the issue solely by reference to the documents, if it appeared from their own terms and/or from what the parties said or did then, or subsequently, that they intended them to constitute an exclusive memorial of their relationship. Lord Hoffmann considered (at page 903f) that the rule that construction of documents is a question of law does not apply when the intention of the parties, objectively ascertained, has to be gathered partly from documents but also from oral exchanges and conduct. If that is so, the terms of the contract are a question of fact. The question whether the parties intended a document or documents to be the exclusive record of the terms of their agreement is also a question of fact. Lord Hoffmann said at page 904j that the austere rule which excludes what the parties thought their obligations were would be orthodox doctrine in a case in which the terms of the contract had been reduced to writing. In the case of a contract which is based partly upon oral exchanges and conduct, a party may have a clear understanding of what was agreed without necessarily being able to remember the precise conversation or action which gave rise to that belief. Lord Hoffmann then said at page 905c:

Each party relies on Carmichael, but for different purposes. Arup rely on it for the admissibility and relevance of the understanding of Mr Elliott, Mr Roberts and Mr Higson as expressed in their oral evidence. CEPAS rely on it for the admissibility and relevance of what is said in Arup’s internal documents.

There is, I think, some distinction between evidence as to what parties to an agreement understood the terms of that agreement to be, and evidence of the parties’ understanding of whether they had entered into an agreement. In the present case, the judge considered evidence of Mr Roberts, Mr Higson and Mr Elliott as to what they each understood to be the contractual effect of what occurred and/or had occurred in March 1996 when Mr Elliott signed the Ground Investigation Proposal. The judge considered that this evidence had an overlay of latter day hindsight tailored to the case it was expected to support. That is a judgment which in my view the judge, having seen and heard the witnesses, was entitled to make. There is no basis for this court interfering with it. Equally, although there is here something of a blurred borderline between questions of fact as to what was agreed and questions of law and fact as to what inferences the court should draw from the parties overt dealings with each other as to their objective intentions, the central issue which this evidence sought to address was as much a question of law as a question of fact.

The judge also took account of what was said in Arup’s internal documents. Mr Bartlett submits that he should not have done so. There is, I think, some force in this submission. But the passages from Lord Hoffmann’s opinion in Carmichael do suggest that a party’s understanding of whether a contract had been concluded may be some evidence tending to negative or support that fact. On balance, Arup’s documents do not support their case that a contract on FIDIC terms was entered into for their design services. The evidence may also indicate that Arup did not rely on any representation to the contrary, which might have been a foundation for a subsidiary argument as to estoppel. Mr White submits that the documents were also relevant to the credibility of the oral evidence. I agree with this.

The parties understandably approach Issues 2 and 4 (d) from opposite ends. Mr Bartlett submits that Mr Elliot’s signature to the Ground Investigation Proposal in March 1996 effected a written agreement on FIDIC terms. He points to clause 1 of the proposal as expressly achieving this. He then submits that the same terms must by then be taken to have been agreed for the design services, both because clause 1, taken literally, expressly states that the Main Agreement had been “entered into” on the terms of the Model Services Agreement, and because it would be “absurd” if the Ground Investigation Agreement were on FIDIC terms but the Main Agreement was not. Mr White submits that no agreement on FIDIC terms had been entered into up to 15^th March 1996 (this is now scarcely contentious); that clause 1 of the Ground Investigation Proposal was promoted on an expectation which was never fulfilled; and that a blue pencil should be put through clause 1 because in the circumstances it was meaningless – see Nicolene v. Simmonds[1953] 2 All ER 822. Mr Bartlett says that the court is concerned in the first instance with a question of construction. Mr White says that the first question is one of fact. Each of them, I think, contends for symmetry between the Design Agreement and the Ground Investigation Agreement. Mr Bartlett says that they were both on FIDIC terms: Mr White that they were both on the terms of the Letter of Intent. Mr Bartlett emphasises that the Ground Investigation Agreement required some of the FIDIC terms for its proper operation (those relating to invoicing and variations). He also points out that the Ground Investigation Agreement was self contained as to payment, that is it did not require for its operation the main matter (the payments schedule) which had not been agreed for the design work. Mr White says that the judge held that the payments schedule, which was never agreed, was an essential ingredient of the design agreement. Mr Bartlett tried to say that the payments schedule was inessential because the parties were able to operate without it. Mr White submits that there is nothing unfair in a design contract without limitations as to time and liability, since Arup had tendered for the work without these limitations. They had proceeded on the letter of intent on this basis.

In my judgment, the judge reached the wrong conclusion on issue 4(d). There is no doubt but that Mr Elliott’s initialling of the Ground Investigation Proposal beside the word “OK” signified in writing his agreement to the proposal which was before him. This brought into existence an agreement in writing which the court has to construe in the light of admissible surrounding circumstances. The surrounding circumstances included, not only that a contract for design services on FIDIC terms had not been entered into, but also that the Ground Investigation Proposal, containing clause 1 had originated before Revision B of the proposed design agreement had been sent to CEPAS. The fact that by March 1996 an agreement for design services on FIDIC terms had not been “entered into” does not, in my judgment, begin to justify the use of a blue pencil for clause 1. The terms to which clause 1 referred were clearly identifiable and mutually available to the parties. So far as was relevant to the Ground Investigation Proposal, they were uncontentious. Abjuring, as I do, the use of a blue pencil, the realistic choices of construction are:

In my judgment, (a) is the correct construction. It can be achieved linguistically, if this is necessary, merely by inserting the words “proposed to be” before the words “entered into” in the clause. More importantly, in my view the clause quite clearly shows an objective intention to make an agreement on the FIDIC terms. I am not deterred from this conclusion by the fact that the Main Agreement had not been concluded on these terms when Mr Elliott signed the Ground Investigation Proposal. It would be odd if the two did not march together, but that should not in my view prevent the court from giving effect to the clear objective intention embodied in the words of clause 1.

In my judgment, however, the making of the Ground Investigation Agreement on these terms did not have what I would regard as the wholly incidental effect of bringing into being an agreement for the design works in the terms of the Revision B proposal. Nor did it have the effect of acknowledging the antecedent existence of such an agreement. The parties already had an agreement for this work in the Letter of Intent. Mr Elliott regarded this as sufficient. Arup had proposed and clearly wanted a more detailed agreement. Much of what they tabled was regarded as uncontentious, but the payments schedule, which the Letter of Intent had explicitly anticipated in the words “payment mechanism” and which the judge found Mr Elliott regarded as important, had not been agreed. An agreement was not concluded at the Jakarta meetings, the matter of the payments schedule being left over for future negotiation. This is an indication that the parties regarded agreement of the payments schedule as an essential ingredient of the making of an agreement in substitution for the letter of intent. The payments schedule was never agreed in any form which Revision B anticipated. The late agreement of a payments mechanism was not unequivocably referable to an agreement on FIDIC terms. It was more comfortably referable to the letter of intent.

Although I acknowledge that it may be possible in other circumstances to conclude that an agreement was made without being able to identify the precise moment when this occurred, it is not in my view possible in the present case to find that an agreement was made for the design work on FIDIC terms at some undetermined time by inactivity or exhaustion. The only real candidate for the means whereby a design contract on FIDIC terms was entered into is the signing of the Ground Investigation Proposal in March 1996. There is in my judgment no other event or series of events capable of supporting an agreement by conduct or an estoppel. The subsequent instructions to proceed with extra work on proposals each of which contained a term equivalent to clause 1 in the Ground Investigation Proposal cannot in my judgment do anything which the acceptance of the Ground Investigation Proposal did not itself do. As I have already said, it is debatable whether these three subsequent instructions for additional work should be construed as incorporating the critical terms, but I proceed on the basis that they may have done. Suffice to say, that none of them had the clear written “OK” acceptance which Mr Elliott wrote on the Ground Investigation Proposal. The agreement of a payment mechanism in August 1996 was in significantly different terms from the payment schedules discussed in conjunction with Revision B. In short, there was in my judgment no conduct by CEPAS or between the parties after March 1996 which was unequivocally referable to the existence of an agreement for the design work on the Revision B FIDIC terms.

As to March 1996, Mr Elliott was assenting to the proposal in front of him. The proposal for the design work was not then under consideration. It had not up till then been agreed. There was an important outstanding element which had not been agreed. Looked at objectively, I am unable to conclude that the parties should be seen as having intended by a side wind to bring into existence an agreement which they were not then considering. Mr Bartlett struggled to identify the terms of the design contract that the parties are to be taken as having assented to. He offered in reply that they were assenting to the Revision B terms with the January 1996 payments schedule which would be subject to further revision; or the Revision B terms without the payments schedule.

Although I have concluded that a formal written agreement in terms of the Model Services Agreement was not necessarily required, I do think that as a minimum some direct formal acknowledgment of agreement specifically referable to the design services was necessary. In the circumstances, I do not find this in clause 1 of the Ground Investigation Proposal. I do not think that the judge’s approach to the parties’ oral evidence on this subject should be interfered with. Notwithstanding Carmichael, the admissibility of this evidence was questionable. It had the tendency to divert attention from a proper objective investigation. However that may be, the judge either rejected the evidence or properly held that it did not have objective effect. For these reasons I would uphold the judge’s decision on issue 2.

My conclusions on issues 2 and 4(d) would mean that symmetry between the agreements for the design work and the ground investigation, for which each party contends, was not achieved. This is uncomfortable but not, I think, absurd. The present litigation highlights the critical importance of two terms potentially limiting liability. No doubt at the time Arup would have preferred those limitations to apply throughout. But they cannot then have had the pinnacle significance which they have today. There is nothing to prevent parties making related agreements on different terms, and I think that symmetry may be more attractive forensically than on the ground. The decision of this court in The Pina illustrates circumstances in which making one or more additional contracts did not have the effect of clothing an existing contract with attributes which it did not possess. In the present case, in my judgment, where I consider that the parties did not objectively show a mutual intention to enter into a design agreement on FIDIC terms, I do not think that the argument in favour of such an intention is supported by a putative intention to achieve symmetry.

Arup’s case on issue 5 is that, on the proper construction of clause 17 of the Standard Conditions, a “claim” is not “formally made” other than by means of the formal commencement of court proceedings or arbitration. It is submitted that these are the only formal claims known to English law and procedure, and that only the making of a claim by means sufficient for statutory limitation purposes will suffice. Mr Bartlett accepts that, if this narrow submission is rejected, the letter of 2^nd May 2001 was sufficient in form for the purposes of clause 17. I find this submission entirely unpersuasive. Mr White submits that clause 43 postulates a three stage process of formal claim, attempted agreement and, failing agreement, arbitration. This may be deriving too much structure from clause 43. But in my view there is no warrant for reading clause 17 as referring only to a claim form or notice of arbitration sufficient for limitation purposes. A claim can be “formally made” in correspondence. No doubt it needs to be in writing. No doubt in other cases there could be a debate as to whether a claim advanced in correspondence was sufficiently formal for the purposes of clause 17. The debate does not arise in the present case. In my judgment, the judge reached the correct conclusion on this issue.

For these reasons, I would allow the appeal on issue 4(d), but dismiss it on issues 2 and 5.