Delay analysis methods in Australia

That sentence has to carry more weight than it should, because nothing else in Australian law resolves the tension it papers over. White Constructions Pty Ltd v PBS Holdings Pty Ltd [2019] NSWSC 1166 rejected both parties' delay experts and preferred a lay witness who had no delay-analysis expertise at all, and at [191] held that the SCL Delay and Disruption Protocol gives a party's chosen methodology no standing either way. Six years later, Santos Ltd v Fluor Australia Pty Ltd [2025] QSC 184 held that the experts' relative adherence to the Protocol was "dispositive" of which one to prefer, and Freeburn J called it "an accepted professional standard". Both statements are current law. Neither has overruled the other, and a delay expert who arrives in an Australian tribunal armed only with one of them is under-prepared for cross-examination on the other.

It is worth being precise about what changed and what did not, because the practitioner literature tends to compress this into a single headline. "Australia has softened on the Protocol". That overstates the shift. White Constructions was never a holding that the Protocol is wrong, unhelpful, or to be avoided. It was a holding that the Protocol confers no legal status on a method, so citing it does not, by itself, get an expert's analysis over the line. Santos v Fluor does not disturb that legal proposition at all. What it adds is a factual observation about how two competing experts were assessed against each other, and once a court is choosing between two experts, some basis for preferring one has to exist, and adherence to a well-known industry document turned out to be a defensible one. The Protocol went from having no legal standing to having no legal standing and some evidential weight. Those are not the same as "softened into a rule".

The trajectory: four decisions, four positions

Three reconciliations are available, and none of them is a full answer. First, the decisions may operate at different stages: Alstom and Santos v Fluor go to method recognition and expert credibility, while White Constructions goes to causation. Whether the methodology, whatever it was, actually proved the facts. Second, Santos v Fluor came through referees, whose reasoning the judge adopted. It is a step removed from a first-instance finding made directly by a judge on the evidence. Third, White Constructions rejected both experts, so its remarks on the Protocol were not strictly load-bearing to the result. The case turned on the lay witness, not on which expert's Protocol-conformity won.

None of that resolves the tension. It explains why it exists, and why a competent Australian delay expert briefs both positions rather than picking one. Fluor's appeal in Santos was heard in July 2026, and as at this review no appellate judgment had been reported in the sources available to this corpus. Check the current status before relying on Santos v Fluor as settled. A Queensland Court of Appeal decision bearing the citation [2021] QCA 204 sometimes surfaces in commentary on this dispute. It is an earlier, interlocutory decision in the same underlying litigation and it does not carry the Protocol-credibility finding described here. Do not cite it for that proposition.

The practical instruction that follows from the trajectory is to plead the methodology twice over: once as a factual case, built from records and tied to the critical path, so that it survives on White Constructions' terms even if the Protocol argument is never reached, and once as a credibility case, built from disciplined, documented adherence to a recognised framework, so that it survives on Santos v Fluor's terms if the tribunal does reach for a comparative assessment of the experts. A report that only does the second is vulnerable to exactly the failure that sank both experts in White Constructions. A report that only does the first leaves an easy credibility point on the table.

The threshold that sits under every method question

Before any methodology question arises, Australian law asks whether there was "actual delay" at all, and "actual delay" means critical-path delay. Delay to an activity that carries float, or that does not push out the completion date, does not found an entitlement. However elegant the analysis applied to it. Every method comparison below assumes this threshold has already been cleared. A report that skips it, however methodologically sophisticated, has not actually engaged with Australian law.

Concurrency is defined narrowly at the same first-instance level. Australian Development Corporation Pty Ltd v White Constructions (ACT) Pty Ltd (1996) 12 BCL 317, at 345, holds that delays are not concurrent where the contractor-risk event starts first, causes actual delay, and ends after the qualifying event has finished. A causation-first sequencing test that disposes of many pleaded concurrency cases before any rule about the consequence of concurrency is even reached. It is functionally the same move as the English Adyard line, arrived at independently. The full treatment is at concurrent delay in Australia.

The Protocol has no standing either way. Methodology must be driven by the factual evidence.

That is White Constructions at [191], and it is still the operative statement of principle even after Santos v Fluor softened its practical bite. The two decisions can both be true: the Protocol confers no legal status on a method, and a court can still find that closer adherence to it happens to track which expert did the more careful, more credible work.

Method-by-method reception

No Australian court since White Constructions has accepted a named delay methodology by name. A negative finding worth stating plainly, because it means no technique carries a presumption of admissibility and none is presumptively barred.

Two consequences follow directly. If the contract's EOT clause specifies or implies a method. As V601 found the clause before it did. That displaces the analyst's preference entirely. Read the clause before choosing a technique. And where the contract is silent, an Australian court's only hard exclusion on the record is the Alstom rejection of an unorthodox, non-standard method. Orthodoxy is rewarded not because a protocol requires it, but because an unfamiliar technique is harder to defend as fact-led.

Notice what the table does not contain: a single Australian decision that compares two accepted methods against each other and prefers one on the merits. Every entry above is either an acceptance-in-principle, a rejection of an outlier, or a contractual mandate. None is a judicial ranking of, say, windows analysis against collapsed as-built on the same facts. That gap matters for how a practitioner should read the table. It is not a hierarchy. It is a record of what has and has not been tested. A method sitting in the "accepted as a starting point" row has not been endorsed as sufficient on its own. It has simply never been struck out. The absence of judicial ranking is itself consistent with the fact-led posture running through the whole trajectory: Australian courts have shown no interest in adjudicating a contest between methodologies in the abstract, only in whether the method actually used proved the facts in the case before them.

Float ownership: the question Australian law has not needed to answer

There is no Australian authority and no Australian standard-form provision directly on float ownership. That is not an oversight so much as a structural consequence of the critical-path threshold above: if a delay does not hit the critical path, there is no claim, so a court never has to decide whose float was consumed to get there. The question is displaced, not resolved.

It has a latent trigger, however. AS 4000-1997 cl 34.4 apportionment. Apportioning a single period of delay between concurrent causes. Requires some view about whose float was used before the causes overlapped. That is the argument most likely to force an Australian court to decide float ownership for the first time, and it has not yet happened. See concurrent delay in Australia for the apportionment mechanics.

Two things follow from having no authority rather than a settled rule. First, neither side gets to assert a default position. A contractor cannot simply argue that float belongs to the project and an owner cannot simply argue that float belongs to whoever the contract happens to favour, because no Australian decision has said either. The question, when it finally arrives, will arrive attached to a live apportionment dispute under a specific clause, and it will be decided on that clause's wording rather than as a general proposition. Second, an expert report that assumes an answer to float ownership without flagging the assumption is exposed in exactly the way White Constructions punished. An unstated methodological choice standing in for a factual finding.

Why the field is this thin

Layer. The sweep counted reported decisions available to this corpus. It is not a claim of legal-database completeness, and Australian case law generally is the least-verified block in this corpus.

The reason is structural, not accidental: Australian standard forms predominantly route disputes to arbitration, so the reported-decision docket is thin by design, and it is thin in a specific way. Roughly a third of the reported cases in the sweep were security-of-payment adjudication review applications, and exactly one concerned delay analysis. The same pattern shows up, for the same underlying reason, in other Commonwealth arbitration-heavy jurisdictions.

The structural point should govern how much weight White Constructions carries. Its dominance in practitioner discussion is a function of scarcity: it is a single first-instance judgment that happens to be the only recent one on point. Not a function of settled authority. Treat it as the best available guidance, not as a rule that has survived appellate testing, because it has not been tested at all.

The same caution applies with equal force to Santos v Fluor, and for a slightly different reason: it did not reach its finding through a trial judge directly weighing competing expert evidence, but through referees whose reasoning a judge then adopted. A referee's determination is a recognised and respected feature of Australian complex-dispute practice, but it is a step removed from a first-instance trial finding, and it carries a correspondingly narrower footprint if the appeal now on foot addresses the Protocol point directly. Until that appeal is resolved, the honest position is that Australian delay-analysis doctrine rests on four first-instance determinations. One of them itself a step removed from a full trial. In four different states, decided between 2012 and 2025, with no appellate court having considered a method question at all. A practitioner citing "Australian law on delay methodology" should be precise about which of those four decisions the proposition actually comes from, because none of them binds a court outside its own state, and none of them binds any court at all in a formal sense beyond the persuasive weight a first-instance decision commands.

Where an Australian delay analysis actually fails

Taken together, the five failure points describe a discipline more than a doctrine: build the analysis from records that capture prevention, tie every delay to the critical path, read the contract before selecting a method, and treat the Protocol as a description of good practice a court might reward rather than as a source of legal entitlement. That is what four first-instance judgments, thin as they are, actually agree on.

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