Chapter 01
The anatomy of a delay claim
How a delay claim is assembled, what evidence supports each part, how the contract form shapes the substantiation requirement, and where the most common failure modes lie.
01
The four parts of a delay claim
A delay claim has four parts that have to hold together: the event, the activity affected, the path the activity sits on, and the milestone that moves as a result. Each part has its own evidence shape. The event is dated and described, with a contemporaneous note tied to the work front. The activity is identified by package, area, and trade, not by a generic description that could refer to multiple operations. The path is shown in the programme extract as it stood at the moment of the event, which is rarely the programme that exists when the formal claim is drafted six weeks later. The milestone is the contractual reference that determines the value of the time at large, whether sectional completion, key date, or practical completion. A claim that names the event but skips the path or the milestone tends to collapse on review because the assessor cannot connect the disturbance to the money. The discipline is to think of the claim file as a chain rather than a narrative. Each link is a discrete artefact with its own provenance. The narrative reads the chain back; it does not replace it. When the chain is short and continuous, the claim survives external review. When the chain has breaks, the breaks become the focus of the response, regardless of the merit of the underlying claim. Practitioners who have lost a claim usually report that the loss happened on a missing link rather than on the substantive argument.
02
How the contract form shapes the substantiation
The contract form determines how the four parts have to be assembled and on what timeline. Under NEC, the early-warning regime and compensation-event mechanism push the substantiation forward in time. The notice is short, the quotation arrives within weeks, and the project manager assesses against records that are expected to be live. A late submission under NEC is not just commercially weaker; under clause 61.3 it can extinguish the claim entirely. Under JCT, the relevant-event mechanism allows a longer assembly window, but the contractor still has to demonstrate the cause-effect link with reasonable diligence and provide such information as the architect or contract administrator may reasonably require. Under FIDIC, the time-bar at clause 20 is a hard procedural cliff that has been the subject of repeated reported decisions. Under IChemE for process plants, the substantiation requirement leans heavily on as-built records that are often weaker than they should be. The practical implication is that the claim file should match the contract clock. A NEC project that runs claims on a JCT cadence loses time. A JCT project that ignores the relevant-event mechanism loses standing. The team that knows the form and runs the file to the form is the team that does not lose claims on procedure.
03
Programme impact analysis and the evidence it consumes
Programme impact analysis is the technique that connects the event to the milestone. Time impact analysis, windows analysis, and as-planned versus as-built each have their adherents and their critics. What unites them is appetite for time-aligned evidence. The analyst needs the programme as it stood at the event date, the progress as it actually occurred, and the disturbance evidence that links the event to the activity on the critical path. Without all three, the analysis is contestable. Site capture closes the loop on the third. A 360 walk taken in the relevant week, a fixed-camera view of the work front, and a dated entry in the diary together form the disturbance evidence that the analyst can drop into the report. The analyst who has to invent the disturbance evidence from emails and weekly reports is producing a thinner report than the one who can cite a tagged clip and a programme extract. Prolongation cost calculations sit downstream of the impact analysis. They draw on time-related preliminaries, supervision, accommodation, and equipment standing time. The cost figures are rarely the point of contention; the point of contention is whether the time was caused. The discipline of the impact analysis is therefore the discipline of the prolongation case.
04
Common failure modes
The most common failure is retrospective assembly. The team opens the file when the formal notice goes out, by which point the captured base has rotated past the event and the programme has been updated several times. The second failure is a programme that was not preserved at the moment the event occurred, so causation has to be argued from a programme that has already absorbed the event. The third is correspondence not aligned to the same timeline, so the narrative has to be rebuilt from email rather than from a coherent record. The fourth is the overreach claim, where the contractor pushes a soft argument alongside a strong one and the assessor rejects the overall package on the weakest link. The remedy in each case is the same. Open the file at event moment. Preserve the programme at event moment. Keep the correspondence aligned to the same timeline. Do not bundle a weak argument with a strong one. The claim that survives is the claim that does not require apology.
Practice
01. Open the most recent delay claim you submitted or defended. Identify the contract clause that governs the notice period and write down whether the notice landed inside or outside that window.
Look for: A strong response names the specific clause (for example NEC clause 61.3 or FIDIC clause 20.1), states the notice deadline calculated from the event date, and confirms whether the notice was timely with the supporting register entry.
02. For a current programme, list which time-aligned evidence sources would have shortened the substantiation cycle on your last delay claim by more than a week.
Look for: A strong response names at least three sources — for example a preserved programme extract at event date, a 360 walk of the affected front, and a dated diary entry — and links each to the part of the claim it would have shortened.
Checkpoint
For a delay event last week on your project, can you name the four parts and produce the evidence you have for each within an hour?
Recommended reading