Question 5 of 6: Construction Delays — classification and forensic analysis

Question 5: Construction Delays — classification and forensic analysis (20 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

Delay analysis in construction answers two separate questions in sequence: is the contractor entitled to more time, and is it entitled to more money? The classification scheme below exists to keep those two questions apart, because a delay that earns an extension of time does not necessarily earn compensation, and a contractor who conflates them will lose an otherwise good claim. Underlying the whole scheme is a threshold test that is easy to overlook: only a delay that affects the critical path delays the project at all. A delay to an activity with float consumes that float and may be a legitimate disruption claim, but it is not a project delay and generates no extension of time.

Excusable versus non-excusable delays. The distinction turns on responsibility and on what the contract allocates to each party. An excusable delay is one caused by an event outside the contractor's control and not attributable to its own fault or negligence — abnormal weather beyond what a reasonable contractor should have anticipated for the site and season, owner-directed changes, late or defective owner-supplied information or materials, differing site conditions, strikes, and force majeure events. It entitles the contractor to an extension of time, which relieves it of liquidated damages for the period granted. A non-excusable delay is one within the contractor's control or risk: inadequate manpower or equipment, poor planning and sequencing, defective work requiring rework, late procurement, subcontractor default, and failure to prosecute the work diligently. It generates no extension; the contractor bears its own costs and remains liable to the owner for liquidated or actual damages for late completion. In Canada, CCDC 2 general condition 6.5 sets out the mechanism, including the requirement to give notice in writing within a stated period, and the notice requirement is not a formality — late notice is one of the most common reasons entitlement is lost.

Compensable versus non-compensable delays. This second cut applies only within the excusable category, and it decides money rather than time. A compensable delay is an excusable delay caused by the owner or by someone for whom the owner is responsible — the designer, a separate contractor under the owner's control, or the owner's own late decisions, late site access, late approvals or scope changes. It entitles the contractor both to an extension of time and to its delay damages: extended site overhead, extended equipment and supervision, escalation, and in appropriate cases unabsorbed head-office overhead. A non-compensable excusable delay is one caused by neither party — abnormal weather, epidemics, general strikes, acts of a public authority. The contractor gets time but not money, so it is relieved of liquidated damages but carries its own extended costs. This is a deliberate sharing of neutral risk: the owner gives up its damages for late completion and the contractor absorbs its own prolongation costs. It is worth noting that a contract can move the line by express words, and many do — a no-damages-for-delay clause attempts to make all delay non-compensable, and Canadian courts construe such clauses strictly against the drafter and will generally not apply them to delay caused by the owner's own bad faith or fundamental breach.

Concurrent versus non-concurrent delays. A non-concurrent delay is a single delaying event operating alone on the critical path, and its treatment follows directly from the two classifications above. Concurrent delay arises when two or more independent delaying events, attributable to different parties, affect the critical path over the same or overlapping periods, so that each would have delayed completion even in the absence of the other. It is the hardest problem in the field because the classifications now collide: the owner's change and the contractor's under-manning are both delaying the same critical work. The dominant approach in Canadian and English practice, reflected in the Society of Construction Law Delay and Disruption Protocol, is that true concurrency gives the contractor an extension of time but no compensation — time but not money — on the reasoning that the contractor would have been late anyway on its own account and so cannot show that the owner's delay caused its prolongation costs. Two refinements matter in practice. First, genuine concurrency is much rarer than claimed: the events must be truly independent, both critical and effectively simultaneous, and much of what is pleaded as concurrency is in fact sequential delay on the same path, which is apportioned by time rather than shared. Second, the analysis must be done on the critical path as it stood at the time, not as reconstructed at the end, because a delay may be critical in one week and not the next.

A credible method of analysis: time-impact analysis. Of the recognised forensic techniques — as-planned versus as-built, impacted as-planned, collapsed as-built (also called as-built but-for), window analysis and time-impact analysis — the time-impact method is the one most widely accepted by tribunals and the one the Society of Construction Law Protocol recommends for prospective assessment. The procedure is as follows. Establish a validated as-planned critical path method schedule and confirm it is a reasonable and logical plan. Divide the project into windows bounded by the delay events or by contemporaneous schedule updates. At the start of each window, update the schedule with actual progress to that date, so the analysis proceeds from the network as it genuinely stood. Insert a fragnet — a small sub-network representing the delay event and its logic — into that updated schedule, and re-run the critical path calculation. The change in projected completion date is the impact of that single event, measured against the network prevailing at the time it occurred. Repeat event by event and window by window, then assign each increment of delay to the responsible party using the excusable and compensable classifications above.

The method's strength is that it is chronological and contemporaneous: it evaluates each event against the schedule that actually existed when the event occurred, rather than against a plan that had already been overtaken, and it therefore handles shifting critical paths and separates concurrent effects far better than a simple as-planned versus as-built comparison. Its weaknesses are that it is labour-intensive and that it depends entirely on the quality of the baseline schedule and of the contemporaneous updates — which is the practical lesson for the engineer administering the contract. A well-constructed baseline, regular updates preserved with their data dates, accurate daily records of manpower and events, and prompt written notice of delay events are what make any of these methods usable. Without them, the analysis degenerates into competing reconstructions and the claim is decided on impression rather than evidence.