Methodology
Forensic Delay Analysis
Reconstructing, from the programme and the record, how specific events actually affected completion — so entitlement rests on evidence another analyst could reproduce, not on the story a claim would prefer to tell.
This sits beneath Forecast Integrity in the wider framework, and alongside the capability set out under delay analysis and Extension of Time.
A claimed extension-of-time position that has to be tested against the record.
Windowed, longest-path forensic analysis — validity, criticality and causation, event by event.
The baseline, the contemporaneous updates, and the instruction and correspondence record.
The Employer-risk critical delay that is actually demonstrable.
A defensible, evidence-based entitlement rather than a negotiated number.
Why it becomes difficult
The simple version collapses under its own complexity.
By the time a delay is disputed, “the job is late, so we’re owed time” has usually stopped being true in any simple sense. A credible analysis has to hold several hard things at once.
Concurrent events
Several delays running at once, each arguably driving completion over the same window.
A migrating critical path
The path that controls completion moves as the work progresses — what was critical in month four may not be in month nine.
Baseline integrity
If the accepted programme is weak, everything measured against it inherits the weakness.
Mitigation & acceleration
Recovery effort masks delay; the analysis has to see through it to what actually happened.
Contractor vs Employer risk
The same slip can be excusable, compensable, or neither — depending on whose risk it was.
Contemporaneous records
A claim is only as strong as the record made while events unfolded; reconstruction reads differently.
Causation, not correlation
Two things going wrong together is not proof that one caused the other.
Effect on the milestone
Only delay that reaches a contractual completion milestone moves entitlement.
The analytical workflow
From contract to entitlement.
Every step is measured against the contemporaneous record. Skip one and the result stops being reproducible — which is the only thing that makes a determination defensible.
Contract
↓What the contract requires — notice provisions, risk allocation, the completion milestones that matter.
Baseline
↓The accepted, logic-linked programme the analysis is measured against.
Updates
↓Contemporaneous progress — establishing where the driving path actually ran when each event bit.
Delay events
↓Each event isolated, evidenced, and tested for a valid contractual trigger.
Critical path
↓Which path drove completion in each window — and how it migrated across the programme.
Causation
↓Cause and effect demonstrated rather than asserted; concurrency identified honestly.
Time impact
↓The days each event genuinely added to completion — net of float and mitigation.
Entitlement
The Employer-risk critical delay that supports an extension of time — kept separate from the money question.
Contract → Baseline → Updates → Delay events → Critical path → Causation → Time impact → Entitlement
Choosing the method
No single correct method — a correct method for the evidence.
AACE 29R-03 recognises a family of techniques — impacted as-planned, time impact analysis, as-planned versus as-built, windows, and collapsed as-built — each with a proper application and a characteristic weakness. Selection is driven by the records, the programme quality, the timing and the purpose; the real test is whether an independent analyst, given the same records, would reach a comparable result. Where the choice of method is doing the work the evidence should be doing, it is advocacy, not assessment. This is set out in full in how a delay is measured, with the honest treatment of concurrency and risk ownership alongside it.
In practice
How the analysis was undertaken.
Seventeen delay events across a long period, with the critical path migrating repeatedly. A windowed retrospective analysis was the only way to separate the days that genuinely delayed completion from those that did not. Read the case study →
A 291-day claim across twenty-one events and thirteen milestones, assessed to sixty-four approved days — a figure that sat between the Engineer’s sixty-seven and an independent specialist’s sixty. The distance from the claim was traceability, event by event. Read the case study →
Educational commentary drawn from completed determinations; not legal advice.
Get in touch
Discuss forensic delay analysis, critical-path cause and effect, and Extension of Time.