ARTICLES RISK ALLOCATION IN PROCESS ENGINEERING PROJECTS Andrew Stephenson* Many of the issues which arise in process engineering contracts are similar to those which arise in other engineering contracts. However, because of the risk associated with process engineering projects the usual participants are unlikely to be willing to take full risk. Accordingly, there is usually significant residual risk left for the project vehicle and/or debt. Identifying that risk is important for determining the commercial viability of the project, after considering both the risks and rewards and developing strategies for managing that risk. The principal areas of risk associated with such projects relate to failures of the process engineering, defects in design of other aspects of the engineering (civil, structural, mechanical, electrical or instrumentation), defects in construction, delay in reaching mechanical completion, delay in commissioning and ramp up, and inadequacy of insurance. This paper deals with the balancing of these risks between the four principal players in such projects, namely, the process supplier, the EPC/EPCM Contractor, the Trade or Subcontractors and the insurers. INTRODUCTION The purpose of this paper is to consider the special challenges which arise in respect of process engineering contracts. There are four principal participants in such projects. First, the process supplier, who is usually an engineering company with proprietary technology such as Kellogg in the oil industry and Sherritt/Dynatec in the hydrometallurgical industry. Many of these organisations are not full service. Their primary business is to provide the process engineering design or technology which principally consists of the chemistry necessary to achieve the process which produces the product from the process plant. Secondly, the EPC/EPCM Contractors who are primarily responsible for the other engineering design and construction of the project. However, the risks accepted by these two types of contractors are very different. The risk accepted is a function of the different contracts. Both forms of contract will be discussed in this paper. In each case the EPC or EPCM contractor will be responsible for the co-ordination of the detailed design and construction of the vessels, pipework and other equipment to facilitate process to a timetable and a budget. Thirdly, Trade or Subcontractors engaged by the EPC/EPCM contractor. While sometimes these contractors will have limited design responsibility they are primarily responsible for construction. Fourthly, insurers who assist the parties by taking some of the risk. The nature and extent of the risk which insurers are prepared to take is in a state of flux. This paper will also deal with recent experiences in the market relating to such insurances. * Partner, Clayton Utz, Melbourne. 323 Recent Developments (2003) 22 ARELT ENGINEER, PROCURE AND CONSTRUCT The EPC form of contract is popular with smaller mining companies, particularly those seeking to finance on a limited recourse basis. The principal advantages of this form of contract are that the contractor will take the principal risks relating to the cost of construction, the time for completion and the quality of the finished product. The extent to which these risks are assumed will depend upon the market from time to time, the nature of the project and the number of competitors in the market place available to tender for the work. It is unusual for EPC contractors in the process engineering area to assume full responsibility for cost, quality and time. It is common that caps or limitations on liability are negotiated. Notwithstanding this, there are number of examples of contracts having been let in the last 10 years where significant risk has been assumed by the EPC contractor. How risk is allocated under an EPC contract Risks and their allocation are considered under the following headings: (i) (ii) (iii) (iv) (v) (vi) Process engineering Civil, structural, mechanical and electrical engineering Construction Time for mechanical completion and liquidated damages Commissioning and ramp up Works insurance. Process engineering It is common for the process engineering to be provided by someone other than the EPC contractor. Therefore, it is necessary for the owner to enter into some licence arrangement with the process provider. In relation to mineral processing no two ore bodies are likely to be the same. Therefore, a process which works in one part of the world will not necessarily work (unamended) with an ore body in another part of the world. It is therefore usual for the owner of the process engineering technology to be engaged to do laboratory and other testing to modify successful processes used elsewhere to accommodate unique aspects of a particular ore body. The bargaining position of process engineering providers is usually high. They own the intellectual property which the owner wishes to access. Usually the principal has taken care to ensure that the process it has selected is the best available in the market. Therefore the owner of that technology can insist on its terms or there will be no deal. This produces significant problems in a risk sense, where the process provider insists upon extensive limitations on liability. For example, in a recent project the contract limited the process provider's liability to: (a) (b) reperformance of its workscope; and a further cap which limited the value of such rework to $3m. Accordingly, there was no allowance for consequential loss in the event that the plant was built and the process did not work. At the time of contract it was recognised that there was significant risk in the event that the process did not work and further engineering and construction was required (let alone consequential losses, such as loss of revenue). Accordingly an attempt was made to bolster the warranty given by the process provider by the owner taking out, on the process provider's behalf, $50m professional indemnity insurance. That insurance was a conventional "claims made" policy. Such insurance only responds when: (a) (b) a claim is made; and the process provider is liable. Given that there was a limit of liability in the contract between the process provider and the owner of some $3m, the insurance policy could only respond in respect of a claim up to $3m. However, in this case, the excess under the policy was $3m. Accordingly, the premium had been paid whilst no insurance was actually provided. In this circumstance the owner's sole rights associated with a process failure were limited to having the process provider redo the design and test work which was the subject of the licence and ancillary service agreements. A failure of the process engineering may be capable of being fixed relatively easily by introducing more equipment, for example, to deal with impurities that are adversely affecting the product. However, more fundamental problems with the process can result in total project failure as a consequence of the plant being unable to process the particular ore to produce the required product. Accordingly, care needs to be taken to ensure that the risk is adequately sheeted to an appropriate party. This will usually require careful drafting of the relevant agreements with the process provider together with detailed support from insurers. Civil, structural, mechanical and electrical engineering design Pursuant to an EPC contract, the contractor is responsible for both the engineering (ie design) other than process design and the construction. Accordingly, the risk of the proper performance of the design work is with the EPC contractor. 323 Recent Developments (2003) 22 ARELT There are 2 common issues which give rise to difficulty in this area being: (a) (b) early works contractors; insurance issues. Early works contractors It is common in the initial phases of the project for the owner to engage engineers on an EPCM basis (discussed below) for the purpose of drawing design parameters which define the workscope. These design parameters become a contract document in the subsequent EPC contract. At some point it becomes critical to the overall program for the project that long lead time items are ordered. If this time arises before the EPC contract is let, the owner will enter into the necessary contract as principal. If these contracts remained with the owner, then the risk allocation in relation to the project would be split between the EPC contractor and the early works contractors. From a legal and administrative point of view this is undesirable. Therefore, it is common to provide mechanisms whereby these earlier works contracts can be novated to the EPC contractor so that it can take responsibility for the early works contractors. This requires appropriate provisions in both the early works contracts and the EPC Contract to effect the necessary novation. However, difficulties arise where the early works contracts are less favourable to the owner than that ultimately negotiated with the EPC contractor. In these circumstances, the EPC contractor (properly advised) will be unwilling to accept responsibility for the early works contractors in circumstances where it cannot pass the liability it is assuming back to those contractors because of unfavourable terms originally negotiated by the owner under the early works contract arrangement. It is therefore important to ensure that all contracts, including early works contracts, are consistent or that the risk allocation in the early works contracts is superior to that contained or to be contained in the EPC contract. Insurance As with the process provider, it is common for the EPC contractor to take out professional indemnity insurance to cover itself in the event of a claim by the owner. Often, the EPC contractor will seek to limit its liability to the extent of the insurance available to it. The example given above in respect of insurance obtained under agreements for the provision of process engineering services is applicable here. Obviously, if the limit of liability under the EPC contract is less than the insurance limit, the insurance will be limited to the value of the limit of liability in the EPC Contract (if the insurance contract is written on a claims made basis, as the vast majority are). The relationship between liability under the EPC contract and the insurer's liability gives rise to significant levels of anxiety in contractors. There is a tendency for contractors to try and negotiate provisions which state that they will only be liable: (a) if the insurance responds; and (b) only to the extent of such response. Implicit in such clauses is that the owner takes the risk of the solvency of the insurer. However, more importantly, much of the drafting which tries to achieve this outcome produces a Catch-22 along the following lines: (c) (d) the claims made insurance policy only responds where the EPC contractor is liable; the EPC contractor is not liable unless the insurance responds. Such arrangements produce a mismatch between the EPC contract and the insurance contract which provide the insurer with a respectable argument that it can never be liable. This is because its liability is dependant upon the contractor being liable but the contractor is not liable until the insurer is liable. Accordingly, great care needs to be taken to ensure that liability will arise in the contractor which will satisfy the requirement for the insurer to be liable under the relevant claims made policy. The amount of cover available to a contractor is (obviously) also dependant upon the wording of the relevant insurance policy. Not all professional indemnity policies are the same. A common exclusion under professional indemnity policies relates to liability assumed by a professional not being in excess of that which would have been imposed by the common law. Professionals (such as designers) are usually only liable at common law for a failure to exercise reasonable care.1 However, many EPC contracts stipulate a contractual liability to perform the work such that the design is "fit for its intended purpose". Such wording places a significantly higher level of liability upon the contractor than that imposed by the common law. Accordingly, if a claim is made pursuant to such a provision and the insurer has a standard exclusion in respect of liability in excess of that usually assumed, the insurer may decline liability or argue that the policy does not respond. To avoid this outcome, it is good practice to obtain written confirmation from the insurer that it has seen the relevant EPC contract and that its policy extends to cover the liability stipulated. Finally, on this issue it will be important for the owner to understand the terms of the contractor's insurance policy, particularly the exclusions of liability. Accordingly the owner should require the EPC contractor to provide copies of any insurance that it takes out from time to time (in accordance with the contract) and preferably the wording should be procured from the contractor prior to entering into the EPC contract. 1 Voli v Inglewood Shire Council (1963) 110 CLR 74; Brickhill v Cooke [1984] 3 NSWLR 396; Pullen v Gutteridge Haskins & Davey Pty Ltd [1993] 1 VR 27. 323 Recent Developments (2003) 22 ARELT Construction Pursuant to the usual EPC terms, the contractor will assume responsibility for completing the works: (a) (b) (c) to the standard specified (if any); using materials that are of merchantable quality and are otherwise fit for their intended purpose; in a proper and workmanlike manner. In the process engineering area, EPC contractors will (by reference to general caps) seek to limit their liability in relation to this aspect of the work as well. Generally, insurance is not available for defective workmanship or materials (see discussion below in relation to works insurance). However, the contractor does have a method of sharing the risk associated with defective materials and workmanship. Pursuant to most EPC contracts, the EPC contractor will enter into a myriad of: (a) (b) supply agreement; and subcontracts. At this level of the contractual chain, there are many more competitors able to do components of the work. For example, the EPC contractor will usually enter into a subcontract with a civil engineering contractor to provide civil and structural engineering services. Limits of liability (particularly in relation to materials and workmanship) in this market are rare. Accordingly, such contractors are often willing to provide uncapped warranties in relation to construction. However, as the subcontracts become more exotic and the number of available suppliers for the equipment in a question reduces, then there is usually a corresponding increase in the difficulty of securing complete warranties in relation to those contracts. Accordingly, from a commercial point of view, it seems that an owner should try to either: (a) (b) secure a warranty from the EPC contractor that its liability in relation to the quality of construction be limited only to the extent that it is unable to pass that liability on to third party contractors; and at least in respect of the significant subcontractors, secure collateral contracts to take advantage of the benefit of the warranties which are provided. Time for mechanical completion and liquidated damages Pursuant to the usual EPC contract terms the contractor will accept responsibility to complete the works by a particular time, subject to an extension of time provision. Obviously, the extent of the risk assumed by the contractor will be a function of the extension of time provision. It is usual to provide extensions of time for limited force majeure events (ie events beyond the control of either the owner or the contractor) and for acts of prevention by the owner or those for whom the owner is responsible. In relation to the second class of delay, delay arising as a consequence of acts or omissions of the owner, it is essential that the owner provide a proper mechanism for granting extensions of time to the contractor. Failure to do so is likely to render the time provisions in the contract unenforceable as a consequence of the prevention principle.2 Pursuant to that principle if circumstances arise whereby: (a) (b) the owner or those for whom the owner is responsible (principally its agents and servants) cause a delay to the contractor; and Either: (i) (ii) there is no mechanism in the contract for granting an extension of time for such delay; or there is such a mechanism but it does not work in the circumstances; then, at common law, the time provisions and associated liquidated damages clauses will be unenforceable. It is also common in such contracts to provide that the contractor is liable to pay the owner liquidated damages in the event that the project is delayed. Such liquidated damages must be a genuine pre-estimate of the loss.3 A common mistake is to estimate the total revenue likely to be lost as a consequence of a delay in mechanical completion. If the process plant is processing a mineral reserve, a delay in mechanical completion merely causes a delay in the revenue. The revenue is not lost. It is therefore inappropriate to calculate the liquidated damages by reference to the anticipated lost income during the period of delay. Assume that the mine has reserves of 30 years. The project is delayed by a year. No income is earned in the first year in which income was intended to be earned. That income has effectively been postponed to year 31. Therefore, the appropriate calculation of the anticipated loss is: (total lost revenue during period of delay) minus (the net present value of that income in year 31) Where contracts exclude liability for consequential loss, except to the extent of liquidated damages, it is prudent to provide that if the liquidated damages clause fails for any reason, then the 2 3 Peak Construction Limited v McKinney Foundations Pty Ltd (1970) 1 BLR 111; SMK Cabinets v Hill Modern Electrics Pty Ltd [1984] VR 391; Turner Corp Ltd (in liq) v Co-ordinated Industries Pty Ltd (1995) 11 BCL 202; MacMahon Constructions Pty Ltd v Crestwood Estates [1971] WAR 162; Turner Corporation Pty Ltd (Receiver and Manager Appointed) v Austotel Pty Ltd (1994) 13 BCL 378. Dunlop Pneumatic Tyre Co Ltd v New Garage & Motor Co Ltd [1915] AC 79; (1994) 34 BCL 378; Esanda Finance Corp v Plessnig (1989) 84 ALR 99; Wollondilly Shire Council v Picton Power Lines Pty Ltd (1994) 33 NSWLR 551; AMEV UDC Finance Ltd v Austin (1986) 162 CLR 170. 323 Recent Developments (2003) 22 ARELT owner will be entitled to general damages for consequential loss to the limit stipulated by the liquidated damages clause. Otherwise, the owner can find itself in the invidious position of not being able to enforce the liquidated damages provision and not being able to recover general damages as a consequence of an independent consequential loss limitation. It is extremely common in EPC forms of contract for there to be a significant limit on the amount of liquidated damages available, perhaps 10% of the contract sum or less. This is particularly important for the providers of debt who are anticipating that they will be paid interest and capital from the income flow of the project. Extended liquidated damages insurance In the last 5 or 6 years some projects have had the benefit of insurance which is designed to pick up the risk after the contractor's liability has been exhausted. Such insurance has been available pursuant to a line slip through Adam Brothers of New York. The extent to which such is insurance currently available, post 11 September 2001, is uncertain. However, these policies mirror the construction contract provisions in relation to time and put the insurer on risk after the date for mechanical completion (as extended pursuant to the EPC contract) and contain an excess which is usually equal to the extent of the liquidated damages which the contractor has agreed to pay for delay. These policies are relatively unusual and bespoke. The insurers will (increasingly) engage in an exercise of due diligence. They will: (a) (b) (c) (d) be particularly interested to understand the quality of the contractor and its previous performance. Therefore, insurance is more likely to be obtained if the contractor is an international organisation of repute; be particularly interested to understand the extent to which the contractor has accepted liability for the losses due to the delay (therefore if the liability is modest, the insurer will be more uncomfortable, because they will want to ensure that somebody in control of the risk has a legitimate interest in ensuring that the risk does not occur); want to do a due diligence (at the insured's expense) in relation to the design. Any bankability studies by independent engineers are of assistance, but the insurers will usually engage engineers to do a further review; be keen to establish that there is no novel or experimental engineering associated with the project. They will seek disclosure in proposal forms relating to this issue. They will also seek to make it a warranty (entitling recision of the policy if breached) to the effect that the project does not contain any such novel or experimental technology. If the insurance is governed by English law (not unusual given its bespoke nature), the insured may (notwithstanding a contrary legislative intent) not have the benefit of the Australian Insurance Contracts Act 1984 ("the Act") and will run a high risk of the policy being avoided if there is a non-disclosure or breach in respect of these issues, even if the novel and experimental technology does not adversely affect the project or cause the delay4. A large claim has been made by (and paid to) Anaconda Operations Pty Ltd (the joint venture vehicle for the Murrin Murrin Nickel Mine) against such a policy in 2000. The settlement sum (after international arbitration in London) was A$113m. Accordingly it can be expected that any other process engineering projects in Australia will attract significant premiums, which may adversely affect the commerce of taking out such insurance. 4 Insurers are likely in the future to insist that the insurance policy be subject to English law. If they are successful then the insured will not be able to avail itself of the protection provided by the Act. The draftsmen of the Act anticipated such a policy and provided for it in section 8, which reads: "8. Application of Act (1) (2) Subject to section 9, the application of this Act extends to contracts of insurance and proposed contracts of insurance the proper law of which is or would be the law of a State or the law of a Territory in which this Act applies or to which this Act extends. For the purposes of subsection (1), where the proper law of a contract or proposed contract would, but for an express provision to the contrary included or to be included in the contract or in some other contract, be the law of a State of or a Territory in which this Act applies or to which this Act extends, then, notwithstanding that provision, the proper law of the contract is the law of that State or Territory." However, if proceedings are instituted in England (for example) first, then a judge or arbitrator will apply English conflicts rules. The Act (including section 8) will be irrelevant unless and until the Court or arbitrator determines that Australian law applies. This will not occur where there is an express provision in the policy that English law applies because such clauses are determinative pursuant to English law. Under the Contracts (Applicable Law) Act 1990, the provisions of the EEC Convention on the Law Applicable to Contractual Obligations are incorporated into the law of the United Kingdom. Article 3, paragraph 1 of this Convention provides that, in relation to contractual obligations in any situation [other than those expressly excluded] involving a choice between the laws of different countries, a contract will be governed by the domestic system of law chosen by the parties. Accordingly, for the Australian Act to become relevant the entity wishing to rely upon section 8 of the Act probably needs to: - - commence proceedings in Australia; and seek an anti-suit injunction preventing the commencement of proceedings outside Australia, to avoid the outcome in Akai Pty Ltd v People's Insurance Co Ltd (1996) 188 CLR 418 where proceedings were simultaneously conducted in Australia and England (see also Akai Pty Ltd v People's Insurance Co [1988] 1 Lloyd's Rep 90). Even where an Australian Court has the sole conduct of the proceedings the application of the Australian conflicts rules may result in a law other than Australian applying. The relevant Australian conflicts rule requires a consideration (absent a proper law clause in the contract) of the systems of law with which the transactions has its closest and most real connection (Bonython v Commonwealth (1950) 81 CLR 486; [1951] AC 201 at 219. See also Oceanic Sun Line Special Shipping Co Inc v Fay (1988) 165 CLR 197 at 217; Mendelson-Zeller Co Inc v T & C Providores Pty Ltd [1981] 1 NSWLR 366; KA & C Smith Pty Ltd v Ward [1988] 45 NSWLR 702 at 720-1 per Austin J. 323 Recent Developments (2003) 22 ARELT Commissioning and ramp up Each process plant is unique. Accordingly, it is always anticipated that it will take some time from mechanical completion to commission the plant and ramp it up to full production. At this stage of the project, the owner invariably wants to be in control of the plant. However, particularly with inexperienced owners, there is significant risk that it will not have the technical or other expertise to effectively ramp up the plant to full production or educate its operators as to how to operate the plant. For these reasons, where: (a) (b) there are inexperienced owners; or the project is financed on a limited recourse basis the owners will sometimes enter into an agreement whereby the EPC contractor or perhaps the process provider, provides commissioning and ramp up assistance. Such contracts often stipulate a ramp up curve and liquidated damages for failure to achieve production levels anticipated at the time of contract. Usually, the liquidated damages stipulated are a cap on the contractor's liability. Again, the prevention principle (discussed above) is very important. If the production levels set are not achieved as a consequence of an act or omission of the owner then, pursuant to the prevention principle, it is highly arguable that the liquidated damages payable pursuant to such a ramp up agreement will be unenforceable. When coupled with an exclusion of all consequential loss liability the owner will arguably be left with no remedy. This outcome is a distinct possibility in circumstances where the owner has taken over responsibility for the operation of the plant (as is the usual case). Accordingly it is recommended that there be a mechanism, similar to an extension of time provision under a conventional construction contract, whereby the thresholds stipulated in the contract can either be adjusted or the period in which performance is to be achieved can be changed to take account of acts or omissions of the owner. Again, it is also important to ensure that the liquidated damages clause is not a penalty and is a genuine pre-estimate of the loss. Works insurance Works insurance or Construction All Risk (CAR) insurance is a fundamental component of risk mitigation in process engineering projects for both the owner and the EPC contractor. However it should be noted that this type of insurance has significant limitations. In addition there is a marked difference between the risk that is covered by policies that cover projects occurring offshore and those occurring on-shore. The primary purpose of CAR insurance is to insure the physical works or facilities being constructed against loss or damage arising from an "occurrence". In other words this type of insurance is an event based insurance where the insurance will trigger if an "occurrence" happens which causes physical loss or damage to the works. "Occurrence" will be defined by the policy wording but in general it will include events such as fire, natural disaster as well as the negligence of or mistake made by one of the parties. Both the owner and the EPC contractor are named as insured and it is common for either the owner or the EPC contractor to take out the insurance. With such policies care should be taken, where a party to the construction contract give indemnities to the other party which are also given by an insurer, to: (a) (b) avoid creating coordinate liabilities between the indemnifying party and the insurer which would give the insurer the right to seek contribution from the indemnifying party; ensure that the policy of insurance taken out in the joint names of the parties has a waiver of subrogation clause, to avoid the insurer subrogating to the non-indemnifying party's right and seeking to recover from the indemnifying party5 Finally, exclusions of liability can become problematic. The extent and nature of the exclusions will have a significant impact on the risk allocation that will be agreed as between contractor and owner. In this regard there is a major difference between the likely approach for an on-shore project and an off-shore project given that the nature of exclusions in off-shore CAR policies has undergone significant recent change. Onshore projects In the on-shore scenario defective design or workmanship is usually excluded. This type of exclusion generally contains a carve out for consequential physical loss or damage caused by a defective part or defective workmanship. For example, if the foundations to a vessel in a process plant are defective and as a result the foundations fail causing consequent damage to the vessel and associated pipe, the cost of: (a) (b) rectifying the foundations would not be covered; rectifying the vessel and associated pipework would be. As a consequence it is common for the EPC contractor to indemnify the owner for any physical loss or damage caused to the work. Accordingly the EPC contractor will usually accept risk for: (a) (b) (c) defective work not insured; physical loss or damage caused to the work as a consequence of defective work -insured; and any economic losses which flow (i.e. liquidated damages if the completion date is delayed as a consequence of performing reinstatement) ­ not insured. 5 See Caledonia North Sea Ltd v London Bridge Engineering [2000] Lloyds Rep IR249; Maintenance Australia Pty Ltd v Hamersley Iron Pty Ltd (2000) 23 WAR 291 and the very useful discussion by Mark Williams "House of Lords Final Chapter in Elf saga" (2002) 17(5) ILB 33. 323 Recent Developments (2003) 22 ARELT The only carve out would be the "excepted risks" which include things such as war and nuclear risk (almost invariably excluded from CAR cover). Accordingly: (a) (b) (c) the CAR insurer would ultimately accept risk for the physical loss or damage (ie the cost of reinstatement) but not the defective work or the economic losses; the EPC contractor accepts (usually and subject to exclusions and limits of liability) the risk of defective work; and The owner accepts the risk of the limited "excepted risks" and any risk in excess of the exclusions or limitations in respect of which the EPC contractor and the owner have agreed will rest with the owner. Off-shore projects The situation for off-shore process engineering projects has been similar in the past although in the last few years there has been a significant change in risks which the insurance market is prepared to accept, which in turn has had a profound impact on risk allocation as between owner and EPC contractor. Clearly events like the sinking of Petrobras P36 platform (one of the World's largest floating platforms at the time) have influenced underwriters. There is a very limited pool of insurance for off-shore construction risk operating out of London (which is one of the only markets that will insure this type of risk). The Wellington syndicate is the lead underwriter and has effectively set the market precedent by preparing a "standard" policy wording for off-shore construction risk. The latest wording (called Welcar 2020) has formalised some significant changes which have been driven by the underwriter's concerns that they were shouldering too much responsibility for owner's and EPC contractor's mistakes. Welcar 2020 has the same exclusion for defective parts as discussed in the context of on-shore risk. However it differs significantly in the sense that: (a) (b) (c) there are differing tiers of insured namely "principal assured" (generally the owner) and "other assured" (generally the EPC contractor and possibly sub-contractors); there is an exclusion for losses arising out of any failure to comply with the "principal assured's" QA/QC requirements (which must be written into the EPC contract). Arguably this type of exclusion has been introduced because (in the words of the CEO of Petrobras prior to P36 sinking), "onerous quality requirements and outdated concepts of inspection and client control" had been stripped away in order to deliver the project more quickly and cheaply; the CAR insurer has a right of subrogation as against the "other assured" where loss arises as a consequence of the failure to comply with QA/QC requirements, although the "principal assured" remains insured. The long-term effect of these changes is hard to judge. However, the following is immediately apparent: (a) (b) (c) the market is unsure about what might constitute a failure to comply with QA/QC requirements. There is a general lack of confidence in the insurance as a consequence; this lack of confidence has manifested itself in EPC contractors no longer being willing to accept risk for loss of or damage to the works at all or for only the level of the deductible under the policy.; many EPC contractors are demanding that the owner indemnify them for loss or damage caused above a particular limit (again, commonly the level of the deductible). The indemnity is to avoid, from the contractor's point of view, the operation of the insurer's right of subrogation against the "other assured". This leaves an owner in the uncomfortable situation of taking the risk for loss of or damage to the work caused by the contractor's failure to comply with QA/QC requirements even when the loss or damage is caused by the contractor. Understandably this has created a level of uncertainty as to how these risks should be balanced. ENGINEERING PROCUREMENT CONSTRUCTION MANAGEMENT Pursuant to this form of contract the contractor is responsible for: (a) (b) (c) design; procurement of necessary materials and equipment; management of the construction. The important difference between the EPCM and EPC form of contract is that the contractor is not a principal in relation to: (a) (b) procurement of plant and materials; construction. Rather, the EPCM contractor acts as the owner's agent and creates contractual relationships with suppliers and trade contractors. Each trade contract is a head contract directly between the owner and the trade contractor. Accordingly, the principal liabilities of the EPCM contractor relate to: (a) (b) negligence in the performance of the design work; and negligence in managing the procurement and construction work. 323 Recent Developments (2003) 22 ARELT Advantages/disadvantages of EPCM and EPC contracts Obviously the advantage of an EPC contract is that the owner defines the work usually by reference to a performance specification. The contractor takes responsibility on a single point basis (except perhaps in relation to process engineering) in respect of the following: (a) (b) (c) cost of completion (subject to limited adjustments); the time for completion (subject to extensions of time); the quality of the design and work (subject to any exclusions). However, the major disadvantage of the EPC contract is that the detailed design is the contractor's prerogative. Accordingly, great care needs to be taken in defining the design parameters so that the owner obtains a project of the required standard. This usually requires more than simply stipulating performance criteria in relation to the output of the plant. Particularly, in a lump sum environment, the commerce of the transaction will result in the contractor wanting to deliver the project at the lowest possible cost. One of the ways in which the contractor can lower costs is, where it has a design discretion, choosing the cheapest possible design outcome consistent with the design criteria. It is therefore important to ensure the design criteria deal with all aspects which will be important to an owner including: (a) (b) (c) (d) performance; whole of life costs; operability (including maintenance); mechanical availability (insofar as this is not covered by a performance criteria expressed in terms of output). Even where these things are stipulated, the usual provisions in an EPC contract require the contractor to provide copies of the designs to the owner. The EPC contract, if properly drawn, will have 2 provisions which are relevant to this exchange being: (a) (b) a provision entitling the owner to reject the design if it does not comply with the design criteria; and a variations clause. Many disputes arise in this form of contract where the owner has rejected designs on the basis that they do not comply with the design specification (as the owner understands it). Often the contractor will argue that its design did comply and that the owner's rejection of its design is based upon engineering preference (often driven by important issues such as whole of life costs, operability and maintenance). In these circumstances the contractor will argue that the directions rejecting the design (which the contractor contends complies with the specification) are instructions to vary. This will leave the owner exposed to unintended claims for variations. It is for this reason that experienced project managers of EPC contracts avoid commenting on the design produced by the EPC contractor during the course of construction. They, the owners and debt ultimately rely on the warranties by the contractor that it will produce a plant which meets the design criteria. Such reliance, where the contractor fails to meet the design criteria, can cause significant delay either in mechanical completion or commissioning and ramp up phases. This is obviously problematic, but more so where there are significant limitations on liability. For example, if there is an overall limit of liability of 20% of the contract sum and there is a significant failure of design and construction which requires rectification costs of 50% of the original contract sum, then the owner will be left with 60% of the risk associated with the failure of the contractor to achieve the contractual objectives. In addition the consequential losses will be limited by any cap on liquidated damages. While these figures seem to be extraordinary, and are probably unlikely to regularly occur, litigation is currently on foot (the subject of an arbitration) where the owner contends that the defects in the plant exceed 50% of the value of the original contract price. Accordingly, if the cost of an EPC contract includes significant limitations on liability, the owner should consider whether EPCM is a more appropriate method of project delivery because it gives the owner superior control to avoid things going wrong. Because the EPCM contractor is the agent of the owner and does not take responsibility for the final cost of construction or the time for completion (except to the extent of its management role), the owner can interfere extensively during the course of design and construction to ensure that it achieves what it requires. This would appear to be an appropriate course where the owner has significant technical resources who will be able to control the design process. How risk is allocated under an EPCM contract It should be understood that this does not mean that the owner is left assuming all of the risk associated with: (a) (b) (c) (d) cost; time; process design; other design. Under conventional EPCM contracts those risks are dealt with as discussed below. Cost Primarily the EPCM contractor will be responsible for developing budgets and managing the cost in accordance with those budgets. The EPCM contractor will not be liable to the owner simply because the cost of construction exceeds the budget. However, it does have an obligation to 323 Recent Developments (2003) 22 ARELT exercise reasonable care from a financial point of view. This is a relatively low duty and cost blowouts will not generally be the responsibility of the EPCM contractor. However, as design proceeds, packages of work will be developed and let to trade contractors (usually) on a lump sum basis with a conventional risk allocation allowing increases in cost due to specified events only. By the time all of those contracts are let the owner will have to recast its budget to a sum equal to the aggregate of those contracts plus an allowance for contingencies. At this point, there is reasonable certainty as to the price outcome, but the risk profile is still less attractive than under a conventional EPC contract. This is because the owner will generally take the co-ordination risk as between the trade contractors. Therefore, if: (a) (b) (c) trade contractor A delays trade contractor B (because for example trade contractor A must complete its work before trade contractor B can start its); or trade contractor A damages trade contractor B's work; or trade contractor A causes the owner to be in breach for any other reason of trade contractor B's contract, then the owner may (depending upon the drafting of the trade contract) be liable to trade contractor B and may (again depending upon the wording of the trade contracts) have a cause of action against trade contractor A. In a clear-cut case where the facts are well understood and/or do not depend upon disputed engineering opinion, the owner should be able to effectively transfer this risk to trade contractor A (preferably by way of setoff or having access to any security provided pursuant to the trade contract). However, in less clear circumstances a situation often develops where each trade contractor holds the other responsible for some loss or damage. This heightens the potential for litigation arising out of the project which, apart from its distraction, can be very expensive. Time Again, the EPCM contractor is responsible for managing the co-ordination of the various trade contractors in an attempt to ensure that the program is met. The EPCM contractor will not be liable for breaches of contract by the trade contractor, unless induced as a consequence of some negligence by the trade contractor. Each trade contractor takes responsibility to complete its package of work by the date stipulated within the program. However, it will be difficult for the owner to transfer the risk of co-ordination or delay induced to a particular trade contractor by another trade contractor. In those circumstances the owner will, probably, be subjected to claims by trade contractors for non-performance of other trade contractors and will have to seek contribution from those other trade contractors. A further disadvantage of the EPCM method of project delivery is that the amount of liquidated damages available is likely to be less. Each trade contractor is likely to seek a limit of liability based on its contract price. As each contract price under the trade contract will be significantly less than that which would have prevailed under an EPC arrangement, the total liquidated damages available are likely to be significantly less, because, delay is likely to be caused by a small number of the contractors. Design Design liability in these circumstances is similar to that under an EPC arrangement (discussed above). However, under an EPCM arrangement an owner, with appropriate in-house resources, will be able to supervise and check the design so as to catch any defects at any early stage. The tendency under EPC contracts, for the reasons discussed above, is to refrain from comment, to avoid claims for variations. Construction Through the trade contracts the owner should be able to allocate the risk associated with the proper construction of the works (ie use of material of merchantable quality, fit for its intended purpose and proper workmanship) to the various trade contractors. Further, at this level of the market, many of the trade contractors are prepared to accept significant risk of the quality of their own work (as opposed to the design). CONCLUSION 1. The EPC form of contract provides an attractive risk allocation in relation to most aspects of the construction of the project including: Costs; (a) (b) (c) Time for mechanical completion; Quality of the design; Quality of the construction. 2. 3. However, such arrangements will not usually relate to the process design. The supplier of the process technology has a good bargaining position. It will therefore be difficult to negotiate a contract in which the process provider assumes extensive liability. The process supplier will be more likely to accept liability if there is insurance available to cover that liability. Great care is required to ensure that the insurance will be available if the process design fails. Most EPC contracts currently being negotiated have significant caps on liability (some as low as 20%). As a result the owner is effectively taking the risk beyond the 20% cap. Accordingly, it is necessary to consider whether the advantages associated with the EPC form of contract are out weighed as a consequence of the: (a) limitation on the owner's capacity to interfere with the EPC contractor's work for good commercial reasons; and 323 Recent Developments (2003) 22 ARELT (b) the cap on liability. 4. 5. The answer to the previous question will depend on a number of matters, including commercial issues, such as the likelihood of the 20% cap being exceeded. If the EPC is unattractive for these reasons then risk can be managed by use of an EPCM contract by: (a) (b) Ensuring that the owner has sufficient in-house or other resources available to it to monitor and check the performance of the EPCM contractor during the design phase, to ensure that the owner is getting exactly what it wants in terms of performance, operability, maintenance and whole of life costs; Passing as much of the risk in relation to the cost of construction, time for completion and quality of the construction work to the trade contractors by effective trade contracts. 6. The project can be enhanced significantly by appropriate: (a) (b) (c) (d) Professional indemnity insurance (in respect of the design risk); Effective works insurance, although particularly in respect of off-shore projects, the scope of cover is reducing as a consequence of the general tightening of the insurance market and recent failures; Extended liquidated damages insurance of the type written by Adams Brothers of New York (although availability in the current market is questionable); Risk allocation by way of exotic insurance such as Bermuda bonds which provide either standby credit or standby equity upon the happening of specified events.