Extension of time in Australia

The harder-hitting half of that thesis is the one competing pages tend to soften. Australia gives the strictest answers in the common-law world on the questions that actually decide money. Strictest on global claims. A claim that partly fails, fails entirely, with no English tolerance and no Scottish safety net. No formula jurisprudence at all. Not a preference for actual cost, but a genuine absence of authority either way. No constructive-acceleration doctrine, and a penalties test that catches clauses England's would not. A claim that would survive in London or Toronto can fail in Sydney on quantum alone.

The counterweight, and the note this page ends on, is that Australia also gives claimants two advantages they routinely fail to plead: Hungerfords v Walker's compound finance costs as substantive damages, and Cessnock City Council v 123 259 932 Pty Ltd's facilitation of proof, which shifts the evidential onus onto the party whose breach created the gap.

What an extension of time actually does

An extension of time moves the date against which liquidated damages are measured. It is defensive, not compensatory: winning an EOT stops the principal's LD clock. It does not, by itself, pay for a single dollar of delay cost. Time and money are separate gateways in every jurisdiction this corpus covers, and Australia is no exception: an EOT protects the contractor from liquidated damages for the extended period, while a separate claim. Grounded in breach, a compensable event, or a specific contractual right. Is what recovers prolongation cost. Conflating the two is the most common drafting and pleading error in this area, and it recurs in Australian practice exactly as it does in Canada and the United States.

What is different in Australia is how that date usually moves. In England the contractor argues prevention to destroy the completion date outright. In Australia, the more common and more reliable route is that the Superintendent was obliged to exercise a unilateral power to extend. Which produces a new date, not no date at all. That distinction organises the rest of this article.

One common law, eight statutes

Australia has a single body of common law that applies uniformly across every state and territory on the doctrines that decide entitlement: the prevention principle and the Superintendent's unilateral power, the penalties doctrine, the causation principles for concurrent delay, the rule against apportioning a failed global claim, and delay-analysis methodology. None of that changes when a project crosses a state border.

That table is the whole orientation. A reader who has confirmed which project they are working on can immediately tell whether they need anything below the doctrine layer at all. Most of the time, for the questions that decide entitlement, they do not.

One statistic should sit behind every doctrinal statement on this page: 84% of Australian standard forms are amended. 100% on projects above A$500 million. The EOT clause specifically is amended in 76% of contracts, and delay or liquidated-damages provisions in 71% (Melbourne Law School study, 295 respondents). The famous AS 2124/AS 4000 concurrency divergence that every Australian delay article leads with describes clauses that are simply absent from roughly three-quarters of real contracts. Read the executed contract. The standard form tells you what was probably negotiated away.

The Superintendent's unilateral power: the thing that is not in the English textbooks

This is the structural centrepiece of Australian delay law, and it is the single biggest reason an English practitioner's instincts mislead in Australia. Most Australian standard forms give the Superintendent (or equivalent certifier) a unilateral power to grant an extension of time, independent of any contractor claim. That power must be exercised honestly and fairly.

Peninsula Balmain Pty Ltd v Abigroup Contractors Pty Ltd [2002] NSWCA 211 is the leading statement: on AS 2124, the Superintendent's unilateral power must be exercised honestly and fairly, and the Superintendent was required to grant an extension of time even though the contractor's own claim was out of time. 620 Collins Street Pty Ltd v Abigroup Contractors Pty Ltd (No 2) [2006] VSC 491 confirms the same duty for an independent certifier more generally.

Probuild Constructions (Aust) Pty Ltd v DDI Group Pty Ltd [2017] NSWCA 151 sharpens the consequence. On an amended AS 4303-1995 subcontract, the subcontractor completed 144 days late without timely EOT claims, and the head contractor sought to set off liquidated damages. The Court held that to claim those damages, the head contractor was obliged to exercise its unilateral power to extend time for delays it had itself caused. The Court indicated an implied duty of good faith in exercising the discretion. Confined to the prevention rationale, so it obliges extension only for delays the principal actually caused, not an at-large fairness obligation running more broadly.

Land the consequence plainly: this duty largely displaces the prevention principle in Australia, and it produces a new date rather than no date. Where an English contractor would argue prevention to destroy the completion date, an Australian contractor argues that the Superintendent was obliged to extend. A narrower, more predictable, and in practice much easier remedy to win.

V601 Developments Pty Ltd v Probuild Constructions (Aust) Pty Ltd [2021] VSC 849 is the modern high-water mark for enforcing that duty. The Superintendent was found to have colluded with the Principal. The liquidated-damages certificates were set aside and the LD claim failed wholly. Acceleration costs were recovered as breach damages. Though the court expressly declined to treat the EOT refusal itself as a "direction" to accelerate, a point developed further on the acceleration spoke. V601 is the case to cite whenever the Superintendent has not acted independently: it converts the Peninsula Balmain duty from a principle into an enforceable remedy.

Three practical consequences follow. First, the contractor's first argument is the discretion, not the doctrine. Ask what the Superintendent knew, when, and whether the power was considered at all. Second, certifier independence becomes the battleground: a Superintendent acting as the principal's agent rather than an independent certifier is the target, not the underlying delay facts. Third, time at large is correspondingly rarer in Australia than English case law would suggest, because the unilateral power keeps a completion date alive even where an English contractor would be arguing that no date survives at all.

One structural caution: principals have found a device that neutralises the whole architecture. Some amended forms. Reportedly including Queensland government TMR contracts. Provide that an absolute discretion need not be exercised for the contractor's benefit, directly attacking the premise that the power must be exercised honestly and fairly. Where that clause appears, the Peninsula Balmain/Probuild v DDI Group line has correspondingly less to work with. It is the single most consequential amendment a principal can make to an Australian EOT regime, and it is worth checking for first, before any of the doctrine below is pleaded.

Time at large, and two things everybody gets wrong

Two corrections do most of the work in this section, and together they explain why time-at-large arguments are rarer and weaker in Australia than practitioners trained elsewhere expect.

First: only one Australian decision has ever found time at large. Gaymark Investments Pty Ltd v Walter Construction Group Ltd (1999) 16 BCL 449 held that a contractor's failure to lodge EOT applications within a contractual time bar, for delays the Principal caused, put time at large and defeated the liquidated damages. It is a first-instance Northern Territory decision, widely criticised, rejected in England by Jackson J in Multiplex Constructions (UK) Ltd v Honeywell Control Systems Ltd (No 2) [2007] EWHC 447 (TCC), and distinguished rather than disapproved by every later Australian court that has considered it. It has never been expressly overruled. It has been declined repeatedly since: Growthbuilt Pty Ltd v Modern Touch Marble & Granite Pty Ltd [2021] NSWSC 290 at [80], CMA Assets Pty Ltd v John Holland Pty Ltd (No 2) [2015] WASC 217, and Bensons Property Group Pty Ltd v Key Infrastructure Australia Pty Ltd [2021] VSCA 69, which put it plainly: "the touchstone must be the terms of the contract".

Gaymark retains argumentative value in exactly one configuration: a contract with a time bar and no unilateral EOT power. Because most Australian standard forms do have that power, the configuration is rare. Which is precisely why Gaymark has never needed to be overruled outright rather than merely distinguished.

Second, and this correction alone is worth the section: SMK Cabinets v Hili Modern Electrics Pty Ltd [1984] VR 391 is widely mis-described in Australian delay commentary as a time-at-large case. It is not. SMK Cabinets disallowed liquidated damages prospectively from the act of prevention: it did not set time at large. Do not cite it for that proposition, and treat any source that does with suspicion.

The honest Australian answer to "is time at large available?" is: almost never, and the contract. Specifically, whether a unilateral EOT power exists and was properly exercised. Is where the argument is actually won or lost. That is a narrower doctrine than the English case law would suggest, and a different one from the position in the United States, which has no time-at-large doctrine at all, or in British Columbia, which has one candidate decision of its own. See the dedicated prevention principle and time at large spoke for the full doctrinal treatment, including the under-litigated interaction with AS 2124's concurrency bar.

Concurrent delay is decided by your contract, not by a court

Australia is the only jurisdiction in this corpus with no judicial default on concurrent delay. The question is answered by the contract, not by a common-law rule filling a contractual gap. AS 4000-1997 clause 34.4 apportions concurrent delay. AS 2124-1992 clause 35.5 denies an EOT altogether to the extent delays are concurrent. Two standard forms, the same continent, opposite answers on identical facts.

Cite that divergence carefully. It belongs to AS 4000-1997 specifically. Whether apportionment survives into AS 4000:2025 is unverified, and should not be asserted. And immediately undercut the divergence with the amendment statistics above: 84% of contracts are amended, EOT clauses in 76%, so the textbook contrast between the two forms describes clauses that are absent from most of the contracts actually in use. Read the executed contract before relying on either default.

The full treatment. The narrow first-instance definition of concurrency Australian courts have actually applied, the interaction between AS 2124's concurrency bar and the Probuild v DDI Group duty to extend for principal-caused delay (an under-litigated, genuinely open tension, not one this corpus resolves), and the Queensland device that can defeat the whole architecture. Is in the dedicated concurrent delay spoke.

Notice, time bars, and the two states where a time bar can be void

Notice regimes in Australia sit under the ordinary common-law condition-precedent analysis, amended forms, and. Increasingly. A statutory overlay unique to two states. Western Australia's section 16 (Security of Payment Act 2021, commenced 1 August 2022) and Victoria's section 13A (Security of Payment Act 2002, commenced 15 April 2026, retrospective to all contracts already in force in the state) allow a notice-based time bar. Including an EOT time bar affecting a payment or security entitlement. To be declared unfair and of no effect, where compliance "is not reasonably possible" or "would be unreasonably onerous". The test is decided by an adjudicator, court, arbitrator or expert determiner, so it reaches arbitration and litigation as well as adjudication.

The fact that makes this genuinely publishable is the negative: no case law has applied either provision, anywhere, ever, as at this review date. Both statutes are live and untested. Queensland has a clean negative of its own. No equivalent unfair-time-bar power exists there at all, a fact worth stating affirmatively rather than leaving as a silent gap.

These two statutes substantially reduce the practical need for a prevention or time-at-large argument in Western Australia and Victoria specifically, because a contractor can attack a time bar directly through the statutory test rather than arguing that the principal's conduct destroyed the completion date altogether. In effect, they do by statute, and with a structured test. Much of what Gaymark tried and mostly failed to do at common law. The full jurisdiction-by-jurisdiction notice mechanics, including the corrected position on NSW's security-of-payment notice-deeming provisions, are in the notice and time bars spoke.

What you can actually recover: and the formulae that do not exist here

This is the section every international delay article gets wrong about Australia, because every one of them assumes the standard prolongation-cost formulae are universal. They are not. No Australian decision adopts or rejects Hudson, Emden, or Eichleay. That is a genuine absence of authority, not a judicial preference for actual cost dressed up as silence. Say so plainly, and correct any assumption that Australia has quietly picked a side.

The practical consequence is that off-site overhead recovery in Australia runs on actual cost, proved, or nothing. There is no formula available as a fallback where records are thin, which raises the evidentiary stakes on every Australian prolongation claim relative to a comparable US or English one, where a recognised formula can at least provide a starting point.

The counterweight belongs here too, and it should carry real weight in how an Australian prolongation claim is pleaded: Hungerfords v Walker (1989) 171 CLR 125 establishes that finance costs are recoverable as substantive damages, at compound rates. Not as a discretionary interest award bolted on at the end. On a multi-year prolongation claim this is frequently the largest single line item, and it is routinely left out of Australian pleadings entirely. The full treatment of recoverable heads, the measured-mile authorities, and the pleading checklist are in the prolongation and delay costs spoke.

Global claims: the strictest position in the common-law world

Mainteck Services Pty Ltd v Stein Heurtey SA [2014] NSWCA 184 rejected apportionment as an answer to a failed global claim in the plainest possible terms. "that is not the law". There is no English Walter Lilly tolerance and no Scottish apportionment safety net available in Australia: a global claim that fails, fails entirely. Impracticability of apportionment must be proved, not merely asserted, and asserting it is the single most common way an Australian global claim collapses.

This is worth stating as a comparative fact, because it is easy to understate: of the common-law jurisdictions surveyed across this corpus, Australia gives the strictest answer on this question. A pleading strategy that would survive in London on Walter Lilly's permissive-but-not-forgiving approach, or that would find some shelter in Scotland's apportionment practice, has no equivalent shelter in Australia.

Two citation traps sit in this territory and must never be repeated. The correct citation for the case connecting global claims to the doctrine of apportionment by attribution is John Holland Pty Ltd v Kvaerner R J Brown Pty Ltd (1996) 82 BLR 81. A materially different citation from a fabricated version that circulates elsewhere and should never be used. And a separate fabricated citation, attached to a real 2019 decision discussed in the delay-analysis-methods section below, circulates elsewhere in connection with global-claims commentary. Do not use it here or anywhere in this cluster. Only the correct 2019 New South Wales Supreme Court citation given in that section is accurate.

The counterweight closes this section, as it closes the whole page: Cessnock City Council v 123 259 932 Pty Ltd [2024] HCA 17, unanimous, decided 12 May 2024, holds that where a defendant's own breach caused or increased the evidential uncertainty a claimant faces, a rebuttable presumption of recoupment arises and the onus shifts to the defendant. "a fair wind, not a free ride". It is the modern Australian analogue to the US reasonable-approximation line, and the strongest available answer to a defendant's "your records are inadequate" objection. It has apparently never yet been run on delay facts specifically, which makes it this cluster's clearest forward-looking note. The full doctrinal detail is in the global claims and total cost spoke.

Liquidated damages, and a penalties test that is nobody else's

Andrews v Australia and New Zealand Banking Group Ltd [2012] HCA 30 is the reason an English drafting playbook does not transplant cleanly into Australia: a stipulation can be a penalty even where there is no breach of contract. Australia did not follow England's Cavendish reform, which requires a breach as the gateway into penalties analysis at all. The Australian ceiling instead asks whether the sum is "out of all proportion to the legitimate interests" the clause protects (Ringrow. Paciocco v Australia and New Zealand Banking Group Ltd [2016] HCA 28).

State the honest limit plainly, because it matters for how much weight this doctrine should carry in a live dispute: no Australian construction decision has yet been located actually striking a delay liquidated-damages clause on the Andrews no-breach gateway. This is a real and structurally significant doctrinal risk. Australian liquidated-damages clauses face a challenge English clauses do not, but it remains, on the current state of the corpus, an undemonstrated one in the specific context of delay LDs. The comparative point is nonetheless a genuine gap in the market: Australia did not follow Cavendish, and neither did Canada, for entirely different reasons of its own. A two-country divergence from England that most comparative delay commentary does not address at all. The full drafting and litigation detail is in the liquidated damages and penalties spoke.

Acceleration without a doctrine

Constructive acceleration is not a recognised doctrine in Australia. And V601 Developments v Probuild (cited above on the Superintendent's duty) shows precisely why that gap matters less than it sounds. The Superintendent was found to have colluded with the Principal. The liquidated-damages certificates were set aside and the LD claim failed wholly, and acceleration costs were recovered as breach damages even though the court expressly declined to treat the EOT refusal as a "direction" to accelerate. The claimant reached a result at least as favourable as a successful US constructive-acceleration claim, without needing the doctrine that underpins the US claim at all.

That is worth reframing directly for a reader trained on the US position: where a US claimant must plead and prove each of the Fraser five elements of constructive acceleration, an Australian claimant in the same factual position may do better by pleading straightforward breach. The Superintendent's failure to exercise its unilateral power honestly and fairly, and letting acceleration costs follow as damages for that breach. The full comparative treatment, including the interaction with AS 4000:2025's acceleration provisions and the standard forms that do and do not address acceleration expressly, is in the acceleration spoke.

Proving it: methodology, records, and the expert who was not an expert

Delay-analysis method in Australia is decided by the facts and the record, not by a protocol. White Constructions Pty Ltd v PBS Holdings Pty Ltd [2019] NSWSC 1166, and this is the correct citation. A version of this case name circulates elsewhere attached to a fabricated 2002 Court of Appeal citation that must never be repeated. Held that the SCL Delay and Disruption Protocol is not a rule of law. The court went further than merely declining to prefer one expert's methodology: it rejected both parties' delay experts and preferred a lay witness with no delay-analysis expertise at all. The lesson the corpus draws from that outcome is unglamorous but decisive: the site diary in that case recorded what was done, not what was prevented, and a delay analysis built on records that cannot answer the causation question loses to a witness who can answer it plainly, however unsophisticated the method.

That position has since softened, not reversed. In Santos v Fluor [2025] QSC 184, the referees held that the experts' relative adherence to the SCL Protocol was "dispositive", and the judge went so far as to call the Protocol "an accepted professional standard". The two decisions sit in genuine, unresolved tension rather than one simply overruling the other: White Constructions establishes that no protocol commands legal deference, while Santos v Fluor shows a tribunal treating adherence to one as practically decisive on the facts before it. Both propositions should be pleaded, not collapsed into a single rule. The full methodology-by-methodology survey, the records that actually carried Australian decisions, and the expert-evidence admissibility line are in the delay analysis methods and records and expert evidence spokes.

Clocks: adjudication, limitation and the long-stop nobody sees coming

Since 15 April 2026, delay costs are adjudicable in every Australian jurisdiction. Victoria was the last holdout, repealing its excluded-amounts regime the same day section 13A commenced. That is a genuinely new, dateable development, and it changes the calculus for where an Australian delay dispute is actually likely to be resolved: adjudication is now available everywhere, not merely in the states that adopted security of payment earliest.

It matters more than it might otherwise because of what happens after an adjudication determination issues: an adjudicator's delay findings are effectively unreviewable. Probuild Constructions (Aust) Pty Ltd v Shade Systems Pty Ltd [2018] HCA 4 confirms that judicial review is unavailable for non-jurisdictional error, and delay evaluation is quintessentially non-jurisdictional. The practical instruction that follows is blunt: get the adjudication right the first time, because there is very little prospect of correcting it afterwards.

Limitation and the building-action long-stops sit underneath adjudication reach as the final set of clocks a delay claim must clear. Limitation is six years in most of Australia. Except the Northern Territory, which is three. Building-action long-stops of roughly ten years exist in Victoria, South Australia, the Northern Territory, Tasmania and the ACT. New South Wales has its own variant, and there is no equivalent long-stop at all in Queensland or Western Australia. In Victoria specifically, the long-stop replaces the ordinary six-year limitation period rather than sitting alongside it as an additional bar. A distinction worth stating precisely, because treating the two as cumulative overstates a Victorian claimant's actual exposure. The full state-by-state table, and the unresolved question of whether a delay claim itself counts as a "building action" for long-stop purposes, are in the limitation and long-stops spoke, alongside the security-of-payment mechanics in the security of payment spoke.

That leaves the two advantages this page opened with, and they belong here as the closing, forward-looking note rather than as an afterthought. Hungerfords v Walker (cited above). Compound finance costs as substantive damages, frequently the largest single line on a multi-year prolongation claim, and routinely left unpleaded. And Cessnock City Council (cited above). Facilitation of proof, where a defendant's own breach shifts the onus onto it once it has caused the evidential gap a claimant would otherwise have to fill. Neither appears to have been run on delay facts yet. Given how strict the rest of Australian quantum law is. No formula jurisprudence, no apportionment safety net on global claims, no constructive-acceleration doctrine. These two, still-unused advantages are where an Australian claimant's next real gain is most likely to come from.

The three-way comparison

Same question, three common-law answers. Canada apportions concurrent delay and has no statutory no-damage-for-delay regime anywhere in the country. The United States has no construction adjudication forum at all and applies statutes of repose that can extinguish a claim regardless of when it accrued. Australia has adjudication everywhere as of 15 April 2026, and gives the strictest quantum rules of the three. No formula jurisprudence, no apportionment safety net on a failed global claim, and a penalties doctrine that reaches non-breach stipulations.

Layer. Comparative and inferential across three jurisdictions surveyed in this corpus. The Australian row rests on the least-verified case law in the corpus. Verify every citation before pleading any comparative point drawn from this table.

That comparison is this cluster's best asset for a reader deciding which jurisdiction's rules actually apply to their dispute, and it is also the clearest illustration of the thesis this page opened with: one common law, eight statutes, and the strictest quantum rules in the common-law world. With two significant, still-unpleaded advantages sitting underneath all of it.

Sobre este material El contenido de la biblioteca es información general sobre la práctica de las reclamaciones de construcción, no asesoramiento jurídico. El derecho a reclamar, los plazos y el procedimiento se rigen por su propio contrato y por el derecho del lugar donde se ejecuta la obra. Revisado el 1 de septiembre de 2026 · Informar de una corrección