Home Insights Force majeure under stress: disaster ‘tails’ too ‘fat’ for infrastructure contracts?

Force majeure under stress: disaster ‘tails’ too ‘fat’ for infrastructure contracts?

All of a sudden, Australia is being assailed by natural disasters of unprecedented scale – from the bushfires of the 2020 summer to the one in 100 year floods of autumn 2021. With recent heatwaves in Europe, snow storms in Texas and forest fires in California and South Africa, Australia is not alone. Forecasts are for these type of events to only increase in frequency and severity as the consequences of climate change unfold.

These rare but destructive events are examples of ‘fat tail’ risks. For fat tail risks, ruin is more likely to come from a single extreme event than from series of bad episodes. Fat tail risks are not more frequent, they are more consequential.

Contracts work well regulating and coordinating parties in the ordinary course of providing and paying for goods and services, but they struggle with extreme circumstances. This is particularly the case for long-term infrastructure contracts where the stakes are high and losses are calculated in the billions.

While relatively rare, the very extreme nature of these events means that it is important that infrastructure contracts address these issues in the interests of all parties in the best manner possible.

Force majeure regimes are a traditional tool but not necessarily one that can get the job done.

In a future world, suffering the consequences of climate change, where significant random events may be expected to be more consequential, and force majeure clauses may be expected to play a more prominent role, it is important to critically review the current approach and, most importantly, explore alternatives to address key shortcomings.

What is force majeure?

In simple terms, a force majeure event is a circumstance beyond the reasonable control of the contracting parties, the occurrence of which is unpredictable and which has a material impact on the parties’ performance of the contract.

We typically think of natural disasters in this category but disasters immediately caused by humans, such as war, also count.

To take a typical example, a force majeure event may be defined as, the occurrence of:

  • an earthquake, natural disaster, bushfire, landslide, seismic activity, tsunami or mudslide and any resulting fire or explosion;

  • hurricane force winds;

  • a one in 100 year flood;

  • war, civil commotion or terrorism; or

  • contamination or radioactivity.

If such a circumstance occurs, a contractual force majeure regime is supposed to respond to the fact that the normal progress of the contract is interrupted and there needs to be a rapid shift to the parties working together to mitigate the consequences and implement solutions which put the project back on track as quickly as possible.

A force majeure regime will be effective if:

  • coverage: it comprehensively addresses circumstances beyond the reasonable control of the parties;

  • utility: it minimises the overall cost, delay and other impacts of the circumstance and promotes co-operative and flexible action by the contracting parties and other stakeholders to resolve the consequences of the circumstance; and

  • flexibility: it provides tools and options which are fit for the purpose of responding to a wide range of unusual, complex and unplanned for circumstances.

Modern force majeure regimes fall short in each of these dimensions.

The coverage problem

Much effort goes into defining what constitutes a force majeure event for the purpose of each contract. This approach to drafting has a number of different motivations some of which conflict.

The lure of lists

Two broad approaches are taken to defining the circumstances to which a force majeure regime applies: the inclusive definition and the exclusive definition. The exclusive definition is more common in large infrastructure contracts. Like the example given above, it sets out a defined list of events which can qualify as events beyond the reasonable control of the parties.

A key issue with an exclusive definition is that it can miss circumstances that the regime would ideally address. Consider the current experience of the COVID-19 pandemic. A number of exclusive force majeure regimes do not list pandemics or epidemics. The result is that it leaves parties trying to jam the pandemic into other listed categories such as ‘biological contamination’ or ‘natural disaster’.

The fact that pandemics and epidemics are not addressed in the force majeure regime is telling. Given SARS, Ebola and MERS outbreaks, all in the last 20 years and therefore the memory of the people who are actually negotiated the current suite of infrastructure contracts, a pandemic or epidemic should have been high on the list of events that may have been anticipated to occur during the lifetime of a long term project.

In the coming decades, as the impacts of the climate crisis become more widely felt, the impacts challenge our imagination. More serve weather events are anticipated: storms, floods, heat waves, droughts. Bush fires have been experienced and are anticipated to become more intense.

But this focus on high impact natural events is still too narrow. Human actions are less discussed.

Consider, for example, the risk of the European Union placing sanctions on Australia because of its inadequate efforts in curbing carbon emission, the integrity of international supply chains if diplomatic relations with China further deteriorate, civil unrest and protests denying access or even increased risk of terrorism (due to attacks on carbon emitters).

An inclusive definition, by contrast, states the general principle that to qualify as force majeure, an event must be beyond the reasonable control of the parties. It may go on to non-exhaustively list some obvious examples. But it does not preclude novel unlisted circumstances from being taken to be force majeure events provided they can be said to be beyond the reasonable control of the parties.

Local versus global

In addition to the use of exclusive category definitions, principles further seek to restrict coverage of force majeure regime by restricting it to events occurring at or in the vicinity of the site of the physical works or infrastructure of the project.

This formulation begs the question of where do force majeure events occur. A cyclone occurs in the atmosphere and the low pressure system with cyclonic strength winds at its core can extend for over a thousand kilometres. What is the physical boundary of the cyclone – is it all the low pressure system or just a part of it? Or is it the effects of the cyclone which need to in some sense touch the site, such as associated rain, wind or flooding?

Similarly, in what sense can an earthquake be said to occur at a project site? The hypocentre of an earthquake will be located many kilometres below the earth’s surface where the fault ruptures. Is the location the epicentre being the point on the earth’s surface immediately above the hypocentre? Or is the earthquake located in an extended zone where vibrations are detected?

Presumably it is not the parties’ intention that the epicentre be located at or in the vicinity of the project site but if it is where vibrations are detected then this probably encompasses the entire globe as seismographs across the world record the earthquake waves.

Restricting the force majeure regime to events in some sense occurring at the project site deliberately or otherwise ignores the fact that any significant project requires on international supply chains and specialist labour and expertise sourced globally.

Force majeure events can legitimately impact these networks well away from the physical site of a project and have as significant an impact on project delivery.

Frustrating in more ways than one

The motivation for this type of drafting is for principals to seek to avoid having to address difficult questions of whether events are beyond the control of the parties and to hopefully deflect spurious contractor claims.[1]

But in attempting this, the category hurdle or the location hurdle or any number of other hurdles mean that force majeure regimes do not fully cover material events which are genuinely beyond the reasonable control of the parties.

If the project has been materially impacted by such a circumstance, this creates the alternative problem that the contractor is left with no other option than to look elsewhere such as common law frustration and take the issue outside the guard rails set by the force majeure regime

In short, a contract is frustrated and automatically terminated if unforeseen circumstances occur, without the fault of the parties after the contract is made, which mean that either party is incapable of performing and, as a result, the obligations under the contract end up being radically different from those contemplated by the parties to the contract when they entered into it.[2]

An express contractual regime will better regulate the response to, and consequences of, an unforeseen event. It provides the opportunity for provision of notices at a relatively early stage, regular updating of information and obligations to take mitigation action.

The common law doctrine of frustration provides no such guidance. It is typically a matter of court or arbitral decision some years after the events in question. In the meantime, it leaves often financially stressed parties working from their own self-interested perspectives to preserve their interests while simultaneously muddling and blustering their way to progressing the project.

Force majeure regimes rarely expressly provide that they are an exclusive code and that the common law doctrine of frustration does not apply to the contract – which it is open to the parties to do. This, combined with the tendency of principals to seek to narrow the circumstances in which the force majeure regime may apply, breathes life into the application of the common law doctrine to fill in the gaps.

Scale is the natural enemy of utility

The effectiveness of a force majeure regime falls off precipitately as the impact of the force majeure event ratchets up.

The force majeure regime is used as an umbrella approach to address events with completely different scale impacts, from small manageable impacts relative to the financial and technical resources of the parties to existential risks. It cannot do both jobs effectively.

While much drafting focuses on the types of events covered, it is the consequences of an event rather than the type of event which is commercially consequential.

There is a level of catastrophe at which society’s institutions break down and contracts becomes irrelevant. The risk of nuclear war at the project site should not detain us too much. Force majeure regimes purport to deal with this type of risk but really their applicability is illusory.

But, at the next level down, there are events which can impact on such a scale that while society as a whole fortunately remains intact, the consequences for the parties involved can wipe out profits for years or threaten a business’s continued existence.

At that point, the only strategy for the party facing the existential threat is to either give up, replace management, exit the market and potentially appoint voluntary administrators, or fight. Even if it is not on the better side of the claim, given the alternative, it is only logical to pursue the claim to the bitter end.

From the perspective of the party with the better claim, there is no acceptable position. It costs so much in money and time to achieve an outcome that the project is significantly delayed.

While it may ultimately be shown to be in the right contractually, its project is not being delivered and typically alternatives such as terminating and procuring an alternative supplier to come into fix a highly distressed project offers no particular comfort.

If the principal is a government, it also needs to deal with media outcry, pressure from laid off workers, cascade failures through the subcontracting chains and impacts on future procurements.

Co-operation is the first casualty

An important part of any force majeure regime is mitigation. This ranges from immediate ‘first responder’ actions, to ensure safety and minimise any further damage, to medium term assessment of impacts and options for recovery and, ultimately, efficient implementation of a long term recovery solution.

At all stages, mitigation efforts will be enhanced by efficient communication and maximum co-operation between the parties with each party able to bring its available resources to bear.

Despite such principles often being provided for in contracts, the commercial incentives to co-operate are typically relatively weak. This means that parties are limited in their ability to deal with the actual circumstances as they find them and implement best for project solutions.

The reason for the failure is that the binary question of liability dominates and clouds all other considerations in responding to the event. It is a winner take all game.

If the event does qualify as a force majeure event, there is typically a sharing of costs in a pre-agreed manner. If it does not qualify, then costs by and large end up with the contractor.

The stakes are high. Therefore each party focuses on ensuring their actions and words do not act as admissions of liability, to be used in evidence in front of a future tribunal. They each assert their rights and seek to comply with their obligations in a manner consistent with the ‘truth’ as they wish it to be seen at odds with the other party’s approach.

The objective should be to remove the need for brinksmanship and not have questions of liability interfere with getting the project back on track.

Tools fit for purpose

By definition, force majeure events are unpredictable and unplanned for. They are uncertain in nature, timing and scope. Logically, the parties need flexible tools and options to respond with the mitigation and recovery strategies to be developed in real time in response to circumstances.

However, many force majeure regimes seek to prescribe options in advance. Rather than providing certainty, the arbitrary bounds place on the parties simply increase the difficulties in managing force majeure events,

Cash flow stress

In particular, force majeure regimes are typically bad at dealing with the passage of time. They imagine there is an almost instant transition from the occurrence of force majeure to implementation of a solution – a series of conditions are satisfied in short order, new rules then apply and the parties are clear on where they stand.

Nothing could be further from the truth especially if the force majeure event is significant. In reality, force majeure can build over time, brings confusion, strains resources, has no clear end point or solution, requires co-ordination among a range of stakeholders often working at cross purposes and drains cash resources.

In the absence of clarity as to funding source, while this is all happening, the natural commercial response is to minimise the cash burn. This can, in turn, lead to behaviours that inhibit implementing solutions.

A well-designed force majeure regime would recognise that a force majeure event can take a long time to resolve and ensure cash flow concerns do not drive the wrong outcomes.

The termination cliff

A force majeure regime should deal with the possibility that the particular force majeure circumstances may be such that the project cannot continue as contemplated. Often the approach is to place time bounds on the duration of a force majeure event so, for example, if the force majeure event continues for a continuous period of six months, then either party may choose to terminate.

A time-bound approach is linked to cash flow questions. What ‘period of delay’ costs can the parties be expected to wear before being able to demobilise and stop the cash losses? For that reason, if there is another source of cash flow – say, delay in start-up insurance - the period may be allowed to run while that insurance pays out.

This approach sets an arbitrary clock ticking. It has no regard to physical state of the project, the prospects for implementation of solutions which will allow the project to continue and in particular for planning of a longer term strategy to complete the project if that makes overall better sense.


A force majeure regime is required for a project to obtain finance if the project has limited or no access to additional funding sources in the case of unforeseen events. This requires that if a project cannot ultimately continue due to a force majeure event, a credit worthy party will commit to pay a termination payment which will provide for repayment of outstanding debt.

Further, aside from the costs of mitigation and implementing a recovery solution, there will be a focus on a source of cash to pay delay financing costs.

In the case of insurable force majeure events this may be addressed by delay in start-up or business interruption insurance. For force majeure risks which are uninsurable, financiers will look to a credit worthy party to fund the delay interest costs in those circumstances.

What is to be done?

If force majeure risk is to be considered seriously, it needs to be looked at through the lens of scale of impact rather than trigger categories. It is important to consider the practical response available to an event.

If it is of modest financial impact, limited duration and the solution is relatively straightforward, then the existing approach to force majeure regimes are sufficient. If a risk actually arises, given the lower order of magnitude of losses, parties will calculate that it is better to mitigate and resolve as quickly and cheaply as possible, get on with the project and litigate later.

For better or worse, that is the standard contracting practice which applies to this and many other contract risks. While it can be improved, there is little appetite for parties to invest in making changes at the time the contact is negotiated.

For bigger scale impacts, however, there is a need to change the commercial calculus from a winner-take-all contest to funnelling the parties’ behaviour back to resolution of the consequences on a best-for-project basis and minimising overall costs and delays regardless of ultimate liability.

The Barley Rule

It is well established principle that a party will have a duty to mitigate its losses in most cases. But that obligation will not assist if the scale of impact is large and one of the parties has assessed that a show down is the optimal strategy. Any genuine steps to mitigate will relief the very pressure that is sought to be generated.

Further, the unavoidable exogenous impacts of the event are magnified by endogenous risks from the parties’ inability to communicate and work together effectively.

What is required is genuine incentive for the parties to work together, to mitigate losses and implement a solution, and sort out liability later, by creating a truce to allow this to happen – or ‘barley’. This should be built into the contract up front rather than having to be negotiated once the crises arises.

The Barely Rule should provide that determination of liability goes last. The rule is triggered by a party plausibly claiming the large impact event has occurred, regardless of whether the other party agrees.

From that point on, no actions or communications are prejudicial to the parties’ ultimate liability claims, they are expressly without prejudice and cannot be referred to in any dispute process.

The parties would be required to immediately and frequently meet to determine a mitigation and resolution plan and oversee its implementation. The costs of mitigation and solution would be recorded and shared in real time.

Ultimate responsibility for these costs remain up for grabs and are to be determined in the final resolution of the liability question. The parties do not know who will ultimately be held to be responsible for them so there is a mutual incentive to keep the other party’s costs reasonable as well as your own.

Certainty is its own reward. The parties therefore do not get to re-litigate what should have been done based on perfect 20-20 hindsight. Whatever liability is determined, the amounts of the costs are to be based on the contemporaneous records.

It is reasonable to expect that once an event is dealt with, and the project is able to continue, much of the heat may come out of the subsequent determination of liability. With a relationship to preserve and a common understanding of what has occurred it may be easier for the parties to then settle liability by negotiation rather than formal dispute.

Stay alive

Big impact events do not get solved overnight. Cash flow over the period while a solution is being implemented needs to be addressed.

Contractors should be asked to share some of the pain but only up to a pre-agreed cap. But once a contractor needs to dip too deeply into its own working capital reserves, less than optimal behaviours will invariably arise to minimise the cash drain. If the principal cannot fund cash requirements above this cap then termination or suspension options need to be available.

Right to suspend termination

The principal needs to be able to suspend the clock ticking on any right to terminate, especially where this time may be an arbitrary fixed period which has no regard to the circumstances of the event.

This would relieve the threat that the contract will disappear if the other party can hold out long enough for the time period to be triggered. In a large scale event, where the required response may be complex and take significant time to implement, this is not such challenging task.

The quid pro quo for suspending the clock will be that there are a range of unavoidable fixed costs which will be incurred to keep the project going, from interests on loans used to finance the project to site specific prolongation costs. These need to be covered by the cash flow arrangements already proposed.

Comprehensive modification rights

The principal needs to be able see the actions necessary to recover from the event implemented. This is required to be able to clearly describe how the project should recover and proceed but again without prejudice to the deferred question of liability.

Most long-term contracts have some type of variation or modification regime. These regimes are typically conceived of, implicitly if not expressly, as providing for scope variations which are reasonably constrained by the original scope of the project to outcomes.[3]

If a significant impact event occurs, there is likely to be a need to rethink fundamental aspects of the project scope, maybe to significantly curtail the project scale or to add on new and unforeseen major elements.

Further, there may be a need to delve into the detailed manner in which the project is being undertaken, including resequencing activities, how they are done and who undertakes them.

If the contractor considers that scope is legitimately outside its expertise and the scope of the project activities, such a regime requires effective means to take work and move it to other entities which are prepared to undertake it efficiently with the ability to integrate that work into the rump of the project.

Step in

Contracts typically contain regimes which on the face of it permit a principal to step in to aspects of a project to seek to remedy an issue and get the project back on track. They are also typically unusable.

A key issue with ‘step in’ regimes is they traditionally require a defined trigger before they can be activated. If the trigger is the subject of the dispute then neither party will concede that it is has occurred and therefore there will be no clear basis to act.

The trigger needs to move from the occurrence of an event such as force majeure to the claim that such an event has occurred. Further, there needs to be explicit recognition that taking action to address impacts is without prejudice to the ultimate determination of liability.

Smaller liability caps for non-manageable risks

Consideration should be given to lower liability caps for major force majeure risks. Recognise that these risks are fat tail risks which occur so infrequently that there is no defensible basis for either party to meaningfully predict their likelihood or impact.

The contractor should have some skin in the game to incentivise an energetic response to fixing the project if such an event occurs but they are not under any parties’ control and should not threaten the contractor with extinction.

If the maintenance of these smaller liability caps is conditional on contractors demonstrating a good faith contribution to resolution of the event’s consequences, this should place important added weight on the internal calculus of their management in favour of choosing to co-operate.

The non-Gaussian distribution of the risk means that contractors cannot be asked to sensibly price such risks. Insurance may be available for a subset of risks but careful consideration reveals how limited this coverage actually is. And insurance may not actually pay out if the risk is systemic, economy wide and undiversified rather than project specific.

It still comes back to human capacity

Finally, and this is something that cannot be covered by the terms of the contract, it is fundamentally important to build and maintain the human capacity to respond to a crisis.

It is not feasible to anticipate, let alone prepare for, every eventuality. The parties need expert teams available with the experience, courage, resources and authority to respond to novel circumstances quickly and effectively when they arise.

The contractual regime needs to empower these teams and give them the flexibility to act. The financial consequences need to be sufficiently muted that an important part of each parties’ interests continue to include the success of the project. These teams will then have the scope to build sufficient trust with their counterparts in the other party and implement solutions rather than accelerate towards meltdowns.

Their responses will not be perfect but will be so much better than a myopic aversion to action for fear of making a mistake.


The contractual technology used to deal with common and smaller scale obligations and risks is ill suited to the pressures of fat tail risks such as large impact force majeure.

As we move into a time where such risks are expected to be more frequent and more consequential, the contract technology for managing the risks needs to be reconsidered.

Without the ability to call ‘barley’ when needed, many of the players are headed for tears with no hope of continuing to play the infrastructure delivery game.

[1] . For simplicity assume there are two parties – one the principal who may be a government or facility owner and the other a contractor such as a builder or facilities management provider, although in practice for large projects there will be many more parties involved by the time government agencies, investors, financiers, joint venturers and subcontractors are taken into account.
[2] Codelfa Construction Pty Limited v SRA of New South Wales (1982) 149 CLR 337
[3] Dorter Sharkey Building and Construction Contracts in Australia, 8.170



Arbitration Construction, Major Projects and Infrastructure Litigation and Dispute Resolution

This publication is introductory in nature. Its content is current at the date of publication. It does not constitute legal advice and should not be relied upon as such. You should always obtain legal advice based on your specific circumstances before taking any action relating to matters covered by this publication. Some information may have been obtained from external sources, and we cannot guarantee the accuracy or currency of any such information.