Most force majeure clauses fail when they’re needed most. Not because the event doesn’t qualify, but because the clause was poorly drafted, the notice was late, or nobody built an evidentiary file. COVID-19, the CrowdStrike outage, the QatarEnergy declarations, the Red Sea crisis: each one exposed the same gap. The clause that seemed fine at signing became useless under real pressure. This article explains what every force majeure clause must contain, how courts actually interpret them, and what separates the provisions that hold from the ones that don’t.
Key questions this article answers
- What are the essential components every force majeure clause must include to be enforceable?
- What is the difference between force majeure and the common-law doctrines of frustration and impossibility?
- How should force majeure clauses be tailored for construction, energy, technology, maritime, and manufacturing contracts?
- What drafting techniques defeat restrictive judicial canons like eiusdem generis?
- How do you invoke force majeure correctly and what procedural failures will defeat an otherwise valid claim?
6. What do COVID-19, the CrowdStrike outage, the QatarEnergy declarations, and the Suez Canal blockage teach us about modern force majeure drafting?
Who this article is for
Written for lawyers and in-house counsel drafting commercial contracts; procurement and contract managers negotiating cross-border deals; risk officers mapping exposure to geopolitical and technology disruption; and executives who want to understand what the force majeure clause in their agreements actually does. Covers English, US, and South African law, plus civil law frameworks and key instruments: ICC, UNIDROIT, FIDIC, BIMCO, UCC, CISG.
The doctrinal foundation
The tension at the heart of every contract
Contracts rest on a simple principle: pacta sunt servanda — agreements must be honoured. Force majeure is the exception. It says: when something truly extraordinary happens, and it prevents you from performing through no fault of your own, the law won’t hold you to your promise. The clause translates that principle into a specific, negotiated allocation of risk.
Note what it isn’t. Force majeure is not an escape route from an unprofitable deal. It’s not a shield for poor planning. Courts read it narrowly because that’s how the parties wrote it, and because the alternative would make contracts unenforceable every time circumstances shifted.
The common law and civil law divide
In common law jurisdictions—England, the US, Australia, South Africa—force majeure doesn’t exist in the general law. There’s no inherent right to be excused from a contract because something went wrong. Without an express clause, you’re left with the much narrower doctrines of frustration (England), impossibility (US), or supervening impossibility (South Africa). These are blunt instruments: they terminate the contract automatically, and they apply in far fewer situations than most people expect.
However, there’s a trap here most drafters miss. A badly written force majeure clause doesn’t just fail to help you; it actively makes things worse. Courts apply the clause first. Frustration and impossibility only come into play if the clause doesn’t cover the event. So a clause that triggers but then fails on procedural grounds—say, because notice was late—can leave you worse off than if you’d written no clause at all.
Alternatively, civil law systems are different. France, Germany, and China all codify force majeure by statute. A contractual clause can refine the statutory framework, but even without one, courts have a default regime. This gives civil law drafters more flexibility, but also less control, since mandatory statutory rules may constrain what the parties can agree.
Governing law is not a formality. It determines how much work the clause has to do and how courts will interpret it. Choose it deliberately.
A force majeure clause is a contractual allocation of risk for supervening events, not a general escape hatch from an unprofitable bargain. Courts read them narrowly, and with good reason.
From absolute impossibility to commercial impracticability
Traditional common law demanded physical impossibility before excusing performance. The US moved on from this with UCC Section 2-615, which allows relief when performance becomes “impracticable”, so excessively burdensome that it defeats the contract’s basic assumptions.
This matters for one specific drafting reason: “prevents”, “hinders”, and “impedes” are not synonyms. They impose different thresholds.
- “Prevents” means impossible.
- “Hinders” means seriously obstructed.
- “Impedes” is softer still.
The word you choose determines who wins a borderline claim and courts won’t rewrite your choice in your favour.
Core components: The must-haves
A professional-grade force majeure clause has eight structural components. Each one must be drafted explicitly, because courts won’t invent anything you leave out.
C1 | The definition of triggering events
The triggering event definition is where the clause either does its job or doesn’t. Three approaches exist.
- Exhaustive enumeration — listing every covered event — is precise but full of gaps. No drafter is omniscient. Worse, courts apply expressio unius: if it’s not listed, it was intentionally excluded. A clause covering “fire, flood, earthquake, and war” gives you nothing when a government shuts your industry down.
- A general catch-all alone — “any event beyond reasonable control” — creates uncertainty. Strict jurisdictions like New York may find it too vague.
- The hybrid approach — a non-exhaustive illustrative list plus a broad qualifier — is now the gold standard. The list signals what was contemplated. The qualifier handles what couldn’t be foreseen. You need both.
Post-2020, contracts must explicitly address: pandemics and public health emergencies; government-mandated closures; cyberattacks and infrastructure failures; sanctions and trade embargoes; geopolitical conflict; and climate events benchmarked against historical averages. Each of these has been litigated. Each has failed against boilerplate that didn’t name it.
Eiusdem generis and how to defeat it
Eiusdem generis: where a general term follows a specific list, courts read it to cover only events similar to those listed. A clause listing “fire, flood, and storm” followed by “or any other event” may not cover a government export ban, because bans aren’t natural disasters.
To defeat this canon, include one of these:
| Expansive formulation | “…whether or not similar to the events listed above…” |
| Unlimited formulation | “…including, without limitation, the following events…” |
| Canon exclusion | “The doctrine of eiusdem generis shall not be applied to the interpretation of this Clause.” |
C2 | The causation standard
Defining the event isn’t enough. The clause must also define how much that event must affect performance before relief kicks in. This is the element most often missing from boilerplate.
It’s not enough that a force majeure event occurred. You have to prove it actually caused the non-performance. Courts consistently reject claims where performance was technically possible but commercially undesirable. In multi-cause situations—where the FM event is one of several contributing factors—English courts apply the “but for” test: you must show you would have performed but for the event. Pre-existing financial problems, poor planning, or supply issues on your side will break the causal chain.
The choice of verb determines everything:
| Prevents | Highest threshold. Performance must be physically or legally impossible. |
| Hinders / Impedes | Moderate threshold. Performance significantly obstructed but not rendered impossible. |
| Delays | Lowest threshold. Covers timing disruption; ultimate performance obligation remains. |
| Makes impracticable | UCC standard. Performance so excessively burdensome as to defeat basic assumptions. |
C3 | The foreseeability requirement and how to override it
Courts often read an unforeseeability requirement into force majeure clauses even when the drafting doesn’t say so. The logic: if you could see it coming and signed anyway, you took the risk. A seasonal hurricane in Florida, a South African power outage during the Eskom era: these may fail unforeseeability scrutiny under a generic clause.
There’s a paradox worth knowing. Some courts have held that naming a specific risk in the clause proves it was foreseeable which then undermines the clause’s general unforeseeability requirement. Specificity can work against you.
The clean fix: include “whether or not foreseeable at the date of this agreement”. This removes the question from judicial interpretation and allocates the risk consciously, not accidentally.
C4 | Notice requirements
Notice is the procedural gateway. Unlike frustration, which operates automatically, force majeure requires you to act. In many contracts— specially energy and maritime—late notice is an absolute bar: you get nothing, regardless of whether the event qualifies. In others, relief runs only from the date notice is received. Either way, late notice is one of the most common reasons valid claims fail.
A proper notice provision must specify:
- Timeframe — 24–72 hours in energy and maritime; up to 14 days under FIDIC in construction.
- Form — written, via a specified method. Casual emails and calls get contested.
- Content — the event, its expected duration, the affected obligations, and what you’re doing to mitigate.
- Consequences — does late notice bar all relief, or just reduce it?
- Ongoing updates — a requirement to report regularly throughout the event, not just at the start.
That last one—continuous updates—is missing from most boilerplate. A single initial notice creates no obligation to keep the other side informed. Regular updates let both parties plan, mitigate, and build the evidentiary record that will matter later.
C5 | The duty to mitigate
Force majeure protects the genuinely blameless. It doesn’t excuse inaction. Every well-drafted clause requires the affected party to take commercially reasonable steps to overcome the event (e.g., alternative sourcing, rerouting, business continuity plans) and to resume as soon as practicable once it ends.
The standard is “commercially reasonable”, not “all possible”. Mitigation should be proportionate to the scale of the obligation.
One important limit, established by the UK Supreme Court in MUR Shipping v. RTI: you’re not required to accept non-contractual performance to mitigate. The court held a shipowner wasn’t obliged to accept euro instead of the contractually specified US dollars, even though the currency switch would have worked around the sanctions triggering the force majeure event. Mitigation is mandatory. Rewriting your own contract to mitigate is not.
C6 | Consequences, duration, and termination
Most force majeure clauses spend all their energy defining what qualifies and almost nothing on what happens next. That’s backwards. The consequences section is where both parties will be reading the clause under maximum stress.
Specify explicitly:
- Suspension or extinguishment — suspension is almost always the right answer in commercial contracts.
- What happens to payment — obligations to pay for goods or services already received usually survive a force majeure event and should be carved out.
- Bilateral vs unilateral suspension — does the non-affected party’s performance also pause?
- The long-stop date — the point after which either or both parties can terminate. Typically 30–180 days. Without it, contracts can be stranded indefinitely.
- Termination consequences — return of prepayments, handover of work-in-progress, settlement of costs.
- Unjust enrichment restitution — if partial deliveries were made before the event was declared, the clause should require compensation for value received. Without this, the performing party is penalised for partial compliance.
The long-stop termination right is not optional. Without it, a temporary disruption can harden into a permanent impasse.
C7 | The subcontractor and supplier double test
What happens when the disruption starts two tiers up the supply chain? The contracting party may be willing and able to perform, but physically can’t because a supplier was hit by a qualifying event. Most clauses don’t address this.
The ICC model establishes what practitioners call the “double test”. To invoke force majeure based on a subcontractor’s failure, you must prove two things: (1) that your own non-performance satisfies the force majeure conditions, and (2) that your subcontractor’s failure independently satisfies those same conditions. Pointing to your supplier’s default as the cause of yours requires proving the supplier was also genuinely force-majeured. Simply establishing that they let you down is not enough.
Address this explicitly in supply agreements: whether subcontractor failures are covered at all; what evidence is required to establish the subcontractor’s qualifying impediment; and whether alternative sourcing must be attempted before invoking the clause.
C8 | The duty to resume performance
Force majeure suspends obligations; it doesn’t end them. Once the impediment ceases, performance must resume. But this is only enforceable if the clause says so. Without an explicit resumption provision, a party that delays resuming after the event ends may argue it remains excused because no clear contractual trigger fired.
The clause should state: performance resumes as soon as reasonably practicable after the event ceases; the affected party must notify the counterparty when the impediment ends; and failure to resume at that point is a new and independent breach, not a continuation of the original excused non-performance.
Advanced risk-allocation tools
The eight components above are the legally necessary minimum. These additional provisions are what turn a standard clause into a genuinely resilient one.
Step-in rights
In critical infrastructure, essential services, and PPPs, the non-affected party may need to take over performance during a force majeure period—directly or through a third party—without triggering a formal breach. Step-in rights should specify the trigger threshold, the scope of authority, the duration, and the handback process.
Hardship and price escalation provisions
Force majeure addresses impossibility. Hardship addresses situations that fall short, i.e., where extreme economic disruption changes the contract’s fundamental assumptions without preventing performance entirely. They need separate clauses. The ICC publishes a model Hardship Clause alongside its Force Majeure Clause precisely for this reason.
In practice: a 15–20% escalation trigger on key input costs, tied explicitly to a geopolitical or regulatory event, gives parties a defined renegotiation mechanism. Without it, courts will reject economic hardship claims under the force majeure clause: consistently and regardless of how dramatic the cost increase.
Payment carve-outs
Force majeure does not automatically excuse payment. Most clauses should include an express carve-out: the obligation to pay for goods or services already received survives the event. Under South African law, the inability to pay money is “subjective impossibility”: money is fungible, so cash flow problems don’t discharge monetary obligations. If you want payment suspended, say so.
The mandatory documentation requirement
The party invoking force majeure carries the entire burden of proof. Build the documentation requirement into the clause rather than leaving it to litigation. A clause-mandated “Force Majeure File” should include: time-stamped photographs and video; official government orders and regulatory notices; meteorological data for weather claims; Chamber of Commerce certificates; and operational logs documenting what was attempted and why performance was prevented.
One specific note on Chinese FM certificates: the CCPIT issues these routinely for widespread events. COVID-19 was the prominent example. They are evidentiary, not self-executing. In common-law jurisdictions they carry weight but not legal authority. For contracts with China-facing exposure, the clause should address their evidentiary status explicitly: either recognising them as prima facie evidence of a qualifying event, or stating that they are neither sufficient nor necessary. Silence has generated substantial litigation.
Partial performance and pro-rata allocation
Most force majeure clauses address total non-performance. They rarely address partial performance and that’s a real gap. A supplier who could deliver 60% of contracted volumes but delivers nothing risks losing the force majeure defence for the portion it could have performed. Courts have rejected claims where partial performance was possible but not tendered.
The UCC imposes a fair allocation duty when a supplier genuinely can’t fill all orders. But the methodology for calculating pro-rata shares must be specified contractually. Include: obligations excused only to the extent prevented; proportional allocation among affected buyers; the basis for calculating each buyer’s share; and the notification procedure.
Governing law and jurisdictional tailoring
New York courts apply strict construction: catch-all provisions are limited to events similar to those listed. California is more liberal. English courts require the event to fall squarely within the clause’s language. Civil law courts may apply statutory criteria that override the contract. Every structural choice in the clause must be calibrated to the judicial construction rules of the governing jurisdiction.
Industry-specific calibration
A force majeure clause that ignores the operational realities of the specific industry it governs is, at best, a starting point. Here’s what changes sector by sector.
| Construction: Critical Path, Weather Thresholds, and FIDIC |
| The most common construction force majeure claims are weather-related. But courts and arbitral panels require more than proof that it rained: the delay must have hit the critical path—tasks that directly affect the completion date. Weather disrupting non-critical activities, or absorbed by scheduling float, won’t qualify. Define weather thresholds by reference to historical data: rainfall or temperatures exceeding the 10-year or 50-year average for the project location. This gives both parties a measurable, objective standard and removes the arguments about what counts as “unusual”. Under the FIDIC 2017 Rainbow Suite, “force majeure” was replaced with “exceptional events”, and a four-limb test applies: beyond control, not reasonably foreseeable, not reasonably avoidable, not substantially attributable to the other party. Cost recovery is available only in specified cases: war or rebellion in the country of performance. Many contractors assume FIDIC gives them more than it does. |
| Energy: Geopolitical Chokepoints, Take-or-Pay, and Regulatory Intervention |
| Long-term LNG contracts—spanning 15 to 25 years—operate in an environment of concentrated infrastructure risk and geopolitical volatility. The 2019 attack on Saudi Aramco’s Abqaiq facility removed 5.7 million barrels per day from global supply. The 2026 QatarEnergy declarations following strikes on the Ras Laffan complex showed how single-point-of-failure concentration translates directly into contract invocations. These aren’t tail risks; they’re recurring events. In take-or-pay agreements, force majeure must explicitly address whether the payment obligation is suspended. Without that, a buyer who can’t receive gas because a terminal was destroyed still owes minimum volume payments—a catastrophic outcome. The Strait of Hormuz, which carries 20–21% of global LNG supply, should be named. Generic “war” or “hostilities” language may not cover a regional conflict short of declared war. |
| Technology: Acts of Code, CrowdStrike, and the Data Security Carve-Out |
| The CrowdStrike incident in 2024 crashed more than 8 million systems and caused an estimated $10 billion in losses. Most boilerplate force majeure clauses listed natural disasters and government actions. Almost none listed “faulty software updates” or “third-party cloud provider failures”. The result was widespread litigation about whether a Blue Screen of Death qualified as a force majeure event. It shouldn’t have been a close question. Technology contracts must distinguish two very different scenarios. Service interruptions caused by a cyberattack or infrastructure failure may qualify as force majeure. Data breaches arising from negligent security practices or non-compliance with GDPR almost never do, regardless of whether an attack was the proximate cause. That distinction must be explicitly drafted. Leaving it to judicial interpretation is a gamble you’ll lose. |
| Maritime: BIMCO 2022, Piracy, and the Own-Employee Strike Carve-Out |
| BIMCO’s 2022 clause is the most sophisticated standardised force majeure framework currently in use. Two provisions matter most. On piracy: BIMCO recommends deleting piracy from the force majeure list if the charterparty already has a dedicated war risk clause, otherwise you end up with two clauses offering inconsistent remedies for the same event. On strikes: only general or regional strikes qualify. A strike limited to the invoking party’s own employees or their subcontractors does not. This carve-out prevents parties from using force majeure to cover labour disputes they arguably had the power to avoid. |
| Manufacturing: The 135% Rule and the Allocation Duty |
| US courts applying UCC Section 2-615 have repeatedly rejected impracticability claims based on price increases, including increases of 135% and more. Sophisticated manufacturers are expected to price commercial risk into their contracts. The threshold for statutory relief is much higher: “extreme scarcity” making it physically impossible to source a unique component, not simply making it expensive. Economic hardship requires a hardship clause, not force majeure. When a manufacturer does successfully invoke force majeure due to genuine supply scarcity, the UCC imposes a duty to allocate remaining supply among all customers fairly. That allocation methodology and the procedure for notifying affected buyers must be specified contractually. |
| Hospitality and events: Frustration of purpose and the post-pandemic problem |
| Hospitality contracts expose a mismatch between doctrine and commercial reality. A venue may be physically operational. A hotel may be open. But government gathering restrictions make the event illegal. This is not force majeure—which requires the event to prevent performance—but frustration of purpose, which applies when a supervening event makes one party’s performance valueless to the other. They require different language. Post-2021, unforeseeability of pandemic risk is increasingly unavailable as a defence. Courts are finding that because parties now know pandemics happen, a frustration of purpose defence fails unless the contract explicitly names pandemics as a qualifying event. Hospitality contracts drafted without that language carry significantly elevated enforcement risk if another major health emergency occurs. |
Invoking the clause: Procedural discipline
Good drafting isn’t enough. Procedural failures, such as late notice, inadequate documentation, failure to mitigate, are the most common reasons valid force majeure claims fail. When the event hits, do this:
| Step 1 | Read the clause immediately. Locate every deadline—notice periods, mitigation obligations, update schedules—and calendar them. |
| Step 2 | Assess whether the event qualifies. Does it fall within the listed events? Does it satisfy the causation threshold? Is it genuinely beyond reasonable control? |
| Step 3 | Give formal written notice within the required window. Include the event, date of onset, expected duration, obligations affected, and initial mitigation steps. State the notice is given “without prejudice” to all other rights. |
| Step 4 | Start building the Force Majeure File: government orders, weather data, photographs, operational logs, Chamber of Commerce certificates. |
| Step 5 | Implement all commercially reasonable mitigation steps. Document each one. |
| Step 6 | Send regular written status updates as required by the clause. |
| Step 7 | Check cross-document implications: credit agreements, insurance policies, related contracts. |
| Step 8 | Watch the long-stop date. Be ready to exercise or respond to termination rights before they default. |
| Step 9 | Notify when the event ends. Resume as soon as reasonably practicable. Delayed resumption is a new breach. |
The waiver and estoppel trap
You can accidentally waive the right to invoke force majeure before you’ve formally claimed it. This happens more often than it should.
If you keep accepting late deliveries, keep representing that you’ll perform, or negotiate a commercial resolution without reserving your position, you may have waived the clause entirely. If you act inconsistently with a force majeure claim (partially performing while asserting you’re excused from full performance), the counterparty has an estoppel argument.
The fix is simple but must be habitual: every force majeure notice should state explicitly that it’s given “without prejudice” to all other rights under the contract. If you’ve already made concessions, get legal advice before invoking the clause. Silence, delay, and inconsistent conduct are the three routes through which otherwise valid claims are defeated on procedural grounds.
Legal notice and disclaimer
This article is published by ITLawCo for informational and educational purposes only. It does not constitute legal advice and must not be relied upon as a substitute for advice from a qualified legal practitioner in the relevant jurisdiction. The law of force majeure varies significantly across jurisdictions and is subject to ongoing judicial development. Force majeure clauses must be tailored to the specific facts, industry, governing law, and commercial context of each transaction.
No attorney-client relationship is created by reading or relying on this article. Case references and statutory citations are provided for illustrative purposes and should be verified before reliance. Legal counsel should be consulted when drafting, reviewing, negotiating, or invoking force majeure provisions in any commercial agreement.
