The analysis set out in this paper represents the author’s views alone. The author acknowledges that they may diverge from the perspectives of those who subscribe to the prevailing English orthodoxy.

Introduction

The law of concurrent delay in England remains conceptually unsettled. Despite repeated judicial engagement, the case law continues to proceed on the basis of a definition whose foundations are fragile and whose practical consequences are uncertain. Courts persist in applying a formulation that emphasises simultaneity and equal causative potency, even though neither requirement is demanded by general principles of causation. Recent decisions have not clarified the position; if anything, they have exposed the extent to which the orthodox approach is shaped less by principle than by concerns about administrability, contractual machinery, and the prevention principle.

This paper argues that concurrency should be understood as a same‑damage phenomenon: independent causes, arising at any point in the project, each sufficient to produce the same period of delay. This approach aligns construction law with established causation doctrine and avoids the artificiality of insisting on simultaneity or equal potency. It does so whilst recognising the concerns grounding the anticipated counter‑arguments. Finally, the paper revisits Peak Construction, demonstrating that it remains a legitimate – and under‑appreciated – authority for apportionment.

1. The Orthodox Definition

The dominant English definition of concurrency – based on Marrin^[1]John Marrin, Concurrent Delay, (2002) 18(6) Const. L.J. 436. – was treated as a useful working definition in Adyard Abu Dhabi v SD Marine Services.^[2][2011] EWHC 848 (Comm). Its influence is evident in in North Midland Building Ltd v Cyden Homes Ltd^[3][2018] EWCA Civ 1744, CA. where the Court of Appeal accepted concurrency as a recognised concept, albeit in the context of a bespoke clause. It requires:

• two effective causes,
• of approximately equal causative potency,
• operating at the same time,
• both delaying completion.

Whilst this formulation has become entrenched, its doctrinal foundations are precarious. Neither simultaneity nor equal potency is a requirement of general causation doctrine. Their elevation in construction law represents a bespoke rule that sits uneasily with the courts’ own treatment of multi‑cause loss.

The courts have never explained why construction delay should be treated differently from other areas of private law. The result is a definition that is both conceptually narrow and practically unhelpful.

2. The Emphasis on Timing

The insistence that concurrent causes must operate “at the same time”^[4]A notable example is De Beers v Atos Origin IT Services UK Ltd [2010] EWHC 3276 (TCC), where the court stated: “177. The general rule in construction and engineering cases is that where there is … Continue reading is not easily reconcilable with the general law of causation. English courts have long recognised that a cause may be legally effective even if another sufficient cause arises – earlier or later.

There is no requirement of simultaneity. The law asks whether each cause explains the outcome, not whether they overlapped temporally. The insistence on simultaneity in construction delay is therefore to create a bespoke causation rule without clear doctrinal justification.

Nor, absent express contractual provision to the contrary, does the mere sequence in which independent causes of delay to progress arise warrant treating one as superseding or displacing the other, particularly where the Contractor’s obligation is to achieve a single contractual completion date (or defined sectional dates) such that no breach occurs unless and until that date is missed.

3. The Dubious Elevation of the Equal Potency Requirement

The requirement that concurrent causes be of “approximately equal potency” is equally problematic.

Modern causation theory – applicable across all areas of private law – provides a coherent and principled framework for analysing situations involving multiple sufficient causes. Wright’s NESS analysis^[5]RW Wright, Causation in Tort Law, California Law Review vol. 73 (1985), pp. 1735–1828. and anticipated in Mackie’s INUS conditions,^[6]JL Mackie, Causes and Conditions, American Philosophical Quarterly 2 (1965), pp. 245-64. provides a coherent framework. Under that approach, a factor is a cause of an outcome if it constitutes a necessary element of a sufficient set for that outcome. Where two independent sequences of events – each sufficient to delay project completion – operate on parallel critical paths, each is properly regarded as a cause of the resulting period of overrun. The relative magnitude of the two causes is immaterial; each is a complete cause in law, as long as it is an effective cause.

Consider, for example, where sub-contractors to two package contractors on a construction‑management project each fail to perform their respective works on parallel critical paths. Each omission forms part of a sequence of activities on its own path that is independently sufficient to delay completion, even if the other path contains its own sequence of activities that would likewise have delayed completion. In such circumstances, each omission is a legally effective cause of the same period of project overrun, irrespective of differences in magnitude or potency.

Similarly, a minor contractor delay and a major employer delay may each independently be sufficient to delay completion. The fact that one is “stronger” or “weaker” does not affect their status as causes, even if the strength of each cause could be objectively verified (a challenge also for those who have argued for a “dominant cause” concept).

The orthodox definition therefore excludes precisely the sorts of situations in which concurrency most naturally arises. It imposes conditions – simultaneity and equal potency – that find no parallel elsewhere in private law and which distort the inquiry into factual causation.

4. A More Coherent Definition?: Same Damage, Not Same Time

The more principled approach is to treat concurrency as a damage‑based concept. Two causes are concurrent where they are independent, where each would (in the counterfactual sense) have delayed completion, and where their effects converge upon the same head of delay. Temporal overlap is not required; nor is equivalence of potency. What matters is that each cause is independently sufficient to produce the same damage.

This approach aligns with Glanville Williams’ seminal definition of concurrent wrongdoers^[7]Following Glanville Williams, Joint Torts and Contributory Negligence, Stevens, 1951 (p. xxxix): “…concurrent wrongdoers… (are)… two or more persons… liable for the same damage, whether by … Continue reading and reflects the reality of project delivery. It also accords with the reasoning in Saga Cruises BDF Ltd. and Another v. Fincantieri S.p.A.,^[8][2016] EWHC 1875 (Comm). in which the court focused on whether two causes contributed to the same head of loss, not whether they were simultaneous or equally potent. See also Thomas Barnes & Sons Plc v. Blackburn with Darwen Borough Council.^[9][2022] EWHC 2598 (TCC).

It also reflects the reality of project delivery, where delays often arise from independent sequences of events that do not neatly overlap.

Concurrency, properly understood, is about same damage – not same time, and not identical or coextensive periods of delay.

5. The Doctrinal Status of Apportionment

The orthodox view is that apportionment has no place in English delay analysis. Whilst this position is repeated frequently, its doctrinal basis appears thin. Once examined, the hostility to apportionment proves to be precarious, both on principle and authority.

(a) General causation doctrine permits apportionment

English law, which already recognises multiple sufficient causes in other contexts apportions liability where multiple causes contribute to an indivisible loss. Tort law does so as a matter of course. Contract law does so in loss‑of‑chance cases and in claims involving multiple causal factors. There is no satisfactory rule that contractual causation must be all‑or‑nothing.

(b) Contractual risk allocation does not prohibit apportionment

Standard forms allocate responsibility for certain events, but they do not dictate the method of causal analysis where multiple events are involved. Thus, none of the standard forms of construction contract, e.g., JCT, NEC or FIDIC requires simultaneity or equal potency, nor do they prohibit apportionment.

(c) The prevention principle is not a bar

The prevention principle precludes an employer from enforcing liquidated damages for periods of delay it has caused. It does not require that all delay be attributed to the employer merely because the employer contributed to some of it. The principle is binary as to enforceability of LADs, not as to causal analysis.

(d) Malmaison does not prohibit apportionment

Henry Boot v Malmaison is often invoked as though it bars apportionment. The case, however, simply found that where a relevant employer event is a cause of delay, the contractor is entitled to an extension of time. It says nothing about how to identify the period of delay attributable to each cause, nor does it preclude apportionment where delay is indivisible. Moreover, the court adopted a party‑stipulated definition of concurrency that it never interrogated.^[10](1999) 70 Con. LR 32, in which the court said, at [13]: “… it is agreed that if there are two concurrent causes of delay, one of which is a Relevant Event, and the other is not, then the … Continue reading

(e) Peak Construction: the overlooked authority

The most significant authority is the one most often ignored. In Peak Construction (Liverpool) Ltd v McKinney Foundations Ltd,^[11](1970) 1 BLR 111. Salmon LJ (as then was) expressly contemplated that where delay is indivisible (as it would at that time have been considered to be given the mainstream absence of managerial and organisational techniques such as critical path analysis, which it predated) and caused by both employer and contractor, and where it is unrealistic to separate “employer weeks” from “contractor weeks”, the court may apportion the delay on a common‑sense basis.^[12]Salmon LJ said, at pp. 119 – 120:“It would, of course, have been impossible for the defendants to argue that their breach caused no delay… It seems to me equally impossible to find that … Continue reading

This is a clear judicial acknowledgment that apportionment is permissible. It has unjustly received limited attention in subsequent discussion. The reason Peak has not been treated as an apportionment authority is not because the courts have rejected such a reading, but because no party has ever advanced it. The apportionment passage has been overlooked, not overruled. If apportionment was permissible when delay analysis was less sophisticated, it is difficult to justify its exclusion now.

(f) Apportionment Legislation

Apportionment may in any event be available under the Civil Liability Contribution Act 1978, which provides a statutory mechanism for allocating responsibility between parties whose conduct has contributed to the same damage.

6. Why the Orthodoxy Persists: Institutional Caution?

The persistence of the orthodox definition may be best explained by institutional considerations:

• Administrability: tribunals prefer simple filters to complex retrospective programming.
• Contractual machinery: EOT clauses require a single certified period of delay.
• Prevention principle anxiety: courts are wary of expanding the circumstances in which employer delay defeats LADs.
• Evidential caution: reluctance to engage in detailed critical path analysis.

These concerns are understandable, but they do not justify distorting causation doctrine or ignoring Peak. Indeed, the irony is that the current approach – insisting on simultaneity and equal potency – creates greater uncertainty by forcing tribunals to engage in artificial inquiries into “true concurrency” that rarely arise in practice.

A causation‑based approach, with apportionment where appropriate, is both principled and workable.

Other jurisdictions – including those within the common law family have demonstrated that neither a broader conception of concurrent delay nor the use of apportionment presents insurmountable difficulties. A clear illustration is found in Schindler Elevator Corporation v. Walsh Construction Company of Canada,^[13]2021 ONSC 283. where the Ontario Superior Court rejected the narrow English insistence on simultaneity, stating at [346]:

“I do not accept… that concurrent delay requires two co‑critical and co‑controlling activities that are parallel in time and identical in duration.”

The same judgment recognised, at [336-337], that apportionment of responsibility for delay is both permissible and workable; it need not be a precise calculation, provided it is grounded in evidence rather than conjecture. Precision is not required; rationality is.

Concerns that apportionment invites impressionistic assessment are therefore overstated. Structured methodologies exist. One such example, was offered in Mastrandrea, “Concurrent Delay: An Alternative Proposal for Attributing Responsibility”^[14][2014] 30 Const. L.J. 173. proposing a causation‑based methodology of undertaking such apportionment. Others could surely be devised.

Scotland, being an obvious example given its geographical proximity, is another jurisdiction taking a more flexible approach.^[15]See, there the formative case of City Inn Ltd. v. Shepherd Construction Ltd. [2008] BLR 269. Civil law systems appear routinely to apportion responsibility where multiple causes contribute to the same harm. The principle of proportional liability seems embedded in civil law doctrine, not requiring simultaneity or equal potency, focusing instead on causal contribution and fairness.

Conclusion

English law’s definition of concurrent delay is conceptually fragile, doctrinally anomalous, and practically unhelpful. Its insistence on simultaneity and equal potency finds no footing in orthodox principles of causation and sits uneasily alongside the courts’ own treatment of multi‑cause loss. A more coherent formulation would focus on independent causes – arising at any point in the life of a project – that each produce the same period of delay. Framed in terms of the same damage, such an approach would align construction law with established causation doctrine rather than standing apart from it.

Likewise, the entrenched hostility to apportionment is not demanded by principle. It is contradicted by general causation doctrine, unsupported by any compelling reading of standard contractual terms, and directly undermined by Peak Construction. English law is not merely capable of apportionment; it already recognises it in principle, even if it has been reluctant to acknowledge the implications.

The path to coherence lies not in refining the orthodox definition, but in adopting instead a causation‑based approach – and acknowledging the legitimacy of apportionment.

---

About the author

Franco Mastrandrea is a Chartered Quantity Surveyor and Chartered Arbitrator with over 40 years of experience in the construction industry. He has acted as expert on more than 50 international project management, delay and quantum-related disputes.

From London to Australia and Canada to Antarctica, Franco has extensive and diverse dispute resolution expertise. With an established record in drafting, interpreting and applying commercial terms in contracts, he successfully combines both knowledge and experience under traditional cross-examination and hot-tubbing.

Franco’s expert commissions cover a wide range of industries including oil and gas in Australia, Canada, Kazakhstan, North Africa and South East Asia; transport infrastructure in the Caribbean and the UK; power generation, from CCGT to windfarms, in the UK; and numerous building and infrastructure projects around the world.