External power contracting, offering a single point of accountability, has reigned supreme for a decade as the principal power station contract. However, power generation projects can benefit from alternative models providing an equivalent degree of security.

Stuart Connor, Alstom, Switzerland

When the International Federation of Consulting Engineers (FIDIC) published its new Silver Book, Conditions of Contract for EPC Turnkey Projects, in 1999, it was not the first time the term “turnkey” had been used. But you could say it marked the ascendancy of the turnkey contract, with its single point of accountability model as the contract of choice for owners of power plants.

In theory, the model allows the plant owner to transfer more of the risks of a power plant project to contractors than possibly any other model in common use, short of situations where the contractor makes a financial investment. However, as business sophistication ratchets up a notch, a full range of alternative contracting models have now become available.

There are many ways to analyze the spectrum of available contracting models. FIDIC for example has, since 1999, differentiated its contract books according to the level of responsibility of the contractor. The Red Book (1957) is essentially a construct-only contract. However, subsequent books – the third edition of the Yellow Book (1987) and the first edition of the Orange Book (1995) – introduced design responsibility for the contractor. Perhaps the final step was the full turnkey package laid out in the Silver Book (1999).

However, in my view, an equally important analysis should be performed on the remuneration mechanism – indeed this has been the area of greatest development in recent years. Is the contract paid on a lump sum, cost reimbursable or incentivized target-cost basis?

That said, even with contract models with the same level of design responsibility and payment mechanism, risk allocation can vary significantly. What are the consequences for the contractor of late delivery or a failure to achieve performance guarantees? To get a more complete picture, we must conduct a more in-depth examination of risk allocation for particular models and projects.

In addition to the standard EPC model, two other contract models are available to modern power generation projects: alliance contracts and engineering, procurement and construction management (EPCm) contracts.

There is no commonly accepted definition of the EPCm contract, and the risk allocation under contracts termed EPCm can vary dramatically. I place an EPCm contract somewhere on the spectrum between a full turnkey model and a professional services model.

In this article I want, first, to consider how these models sit alongside each other, then to look at more specific aspects of the newer models. Finally, I will consider the type of projects that can benefit from an alternative contract model.

Model types

I have assumed that the scope of each model is broadly as follows:

EPC contract: the contractor is responsible for, and directly contracts for, all elements of the project, but most particularly, the engineering, equipment procurement, transport, civil works, erection and commissioning.


The Sino Iron 450 MW combined-cycle plant project, currently under construction in Western Australia, was contracted using the EPCm model Source: AE&E Group GmbH
Click here to enlarge image

Alliance contract: the owner, contractor and other expert parties form an “alliance” and the alliance is responsible for the full EPC scope.

EPCm (professional services): the most common “EPCm” contract is where the contractor only provides professional services, namely engineering, and management of procurement and construction. The owner lets separate contracts with vendors and civil and erection contractors.

EPCm (full turnkey): at the other end of the spectrum, the contractor itself procures equipment and takes a fuller construction management role, so that owner involvement is limited to strategic-level input in construction management and signing construction subcontracts.

There are a couple of ways of analyzing risk allocation. Traditionally, commentators speak of “responsibility” in terms of legal liability. If the costs overrun, who pays? If delivery is late, are there liquidated damages?

An alternative is to consider responsibility in more practical terms. Are the contractor’s risk and reward entitlements dependent on its performance in cost, time and performance outcomes of the project?

The risk allocation for the various contract models is shown in Table 1 (on page 26). The models are ordered in decreasing order of risk allocated to the contractor from left to right.


Table 1. Allocation of risks between owner and main contractor under alternative contract models.
Click here to enlarge image

Some people think alliance contracts are soft: no liquidated damages, no liability for breach of contract and payment on a cost reimbursable basis. On the other hand, the alliance gives an EPC wrap, bears a share of cost overruns, takes responsibility for lateness by the subcontractors and must achieve 100 per cent of performance guarantees. Also it is the only model that minimizes the opportunity for the contractor to claim for compensation events and force majeure.

Likewise, although EPCm (turnkey) contractors are not legally responsible for the performance of the civil and erection subcontractors, they do often share the risk of cost overruns in the construction scope and can share responsibility for the civil and erection subcontractor’s lateness through a painshare/gainshare regime.

EPCm model

Under all forms of the EPCm model the contractor only provides construction management services for the civil and erection component. The subcontractors are engaged directly by the owner. And that is pretty much where the similarities finish.

For many process plants, professional engineering and construction management firms are available that deliver EPCm services. These firms have developed a model towards the professional services end of the spectrum that has been applied to numerous power generation projects. They rarely offer liquidated damages for late takeover of the plant. They charge on a cost reimbursable basis for their own services, and bear only a moderate risk of the overall project out-turn cost. Perhaps the key legal liability is for negligence in carrying out their professional duties.

But do not confuse legal liability with practical responsibility. This model does involve the contractor deeply in the project, and it is fair to assume that those EPCm contractors that achieve good project results are ones that will succeed in the long run.

One significant factor differentiating the power generation industry from others is the existence of contractors who own technology, have an in-house engineering capability and sell major power plant components. This allows them to expand the scope of the EPCm role – the EPCm turnkey model.

This model also has significant possible variations:

Payment: the engineering, procurement and commissioning of the plant can be done on a lump sum basis, or a mix of lump sum and cost reimbursable. Onsite contractors are paid directly by the owner, but there can be a risk sharing arrangement where the contractor bears a share of cost overruns and underruns against a target cost.

Time: it is possible to incorporate liquidated damages into the model, either for late delivery by the contractor of engineering, equipment or commissioning, or for late handover of the plant. Alternatively, for an owner looking to place practical responsibility for timeliness on the contractor, they can omit liquidated damages and include a painshare/gainshare regime with, perhaps, lower stakes but a target date that covers the whole contract scope.

Performance: it is possible for the owner and supplier of the power generation technology to give performance guarantees, as it does under the EPC model. If defects in the construction affect plant performance, since the owner has recourse to the subcontractors, this is arguably an exception to the performance guarantees.

Alliance contracting

Alliance contracting is a model of co-operative contracting that has taken root strongly in the Australian market.

The alliance contracting concept evolved as a response to two perceived failings in the traditional model of construction contracting.

The first was the excessively adversarial nature of the construction market, driven by the traditional lump sum construction contract with liquidated damages.

The second was the barrier to outstanding project performance that was imposed by the commercial structure of traditional contracting. There was a perception that, both commercially and technically, construction contracts were failing the industry.

At the heart of “alliancing” are two concepts: the alignment of the commercial interests of the participants and the removal of barriers to outstanding project performance. The model is directed at the commercial managers and lawyers who put contracts together and administer projects under them as well as the project managers responsible for their execution.

With the aid of an excellent project manager, the model can also speak to the team that actually implements the project. Alliancing achieves this in three key ways:

Performance-based remuneration: the participants are reimbursed their actual costs plus a margin for corporate overheads and profit. The participants then share, on a 50/50 basis, cost overruns or underruns against the cost target, and pay or receive painshare or gainshare, depending on the time and performance outcomes of the alliance. Performance can be managed by reference to key performance indicators. Painshare is capped at 100 per cent of the contractor’s margin.

No blame: neither party is liable to the other for breach of contract or negligence, except in the case of insolvency or wilful default.

Unanimous decision making: the alliance is led by an alliance leadership team with representatives of all participants. Its decisions must be unanimous.

These three elements are illustrated in the alliance equation – performance based remuneration + no blame + unanimous decision making = collective responsibility – the sum of which is collective responsibility. This means that the outcome for the owner and the rewards for the participants depend on the performance of the alliance as a whole. The performance of individual participants is irrelevant.

Because blame is irrelevant, the owner or the contractor cannot make claims, except for wilful default. Without claims, the alliance has much greater freedom to implement innovations in terms of plant optimization and engineering, and procurement and construction processes.

You can think of an alliance as a quasi-company or joint venture. The parties join together to form a body managed and operated in a similar way to an entirely separate organization.

Alliance contracting has been successfully implemented on hundreds of civil and engineering projects in Australia, and it has begun to be introduced in Europe. Its use on power projects is a work in progress, with only a handful of alliance power generation projects in Australia and New Zealand.

Statistical outcomes of alliance contracts include:

  • the almost complete elimination of claims; court disputes on less than 1 per cent of projects
  • out-turn costs within a comparatively small range of the target cost
  • good timeliness, including early completion for projects that looked like running significantly late
  • excellent results in safety and environment, with some alliances winning awards
  • strong outcomes in other performance areas, with most alliances achieving a performance score of between 60 and 80 per cent.

Value for money

Unquestionably, the EPC contract has its strengths. Chiefly, in theory at least, the out-turn time, cost and performance of the plant are fixed for the owner. Unquestionably, for the short- to medium-term, EPC contracting will remain the norm.

However, owners and contractors should be aware of the capacity of alternatives to provide value for money.

A key benefit of the EPCm model is the reduction of price contingencies. As the owner takes back some of the project risks, the contractor can often reduce contingencies. This is particularly so, where the project gains an increasing level of certainty about key construction risks such as ground conditions and bills of quantities as it progresses. The model also allows some flexibility, as key contract packages are let separately to the EPCm contract, allowing movement to account for design evolution.

The alliance model has a host of benefits. It is very useful where the project requires unusual flexibility or the specification is not well defined. This is because of the much reduced scope for contractor claims due to normal project evolution. Indeed, where claim minimization is a chief concern, alliances have performed very well.

The minimal opportunities for claims, performance remuneration and “no blame” concepts create an environment that is apt to foster innovation. Accordingly, an alliance contract works best where there is the chance and the desire to incentivize new ways of working (to achieve good time and cost outcomes) as well as the optimization of plant performance.