Project AssureThere is no single "correct" method of delay analysis, there is the method that fits the evidence, the contract and the moment. Choosing badly is one of the most common ways a genuine entitlement is lost, because a tribunal will discount an analysis whose method the records cannot actually support.
What is the difference between prospective and retrospective delay analysis?
The first fork is timing. Prospective methods assess the likely impact of an event at the time it arises, using the programme as it stood then. Retrospective methods look back, after the effect is known, and reconstruct what actually drove the critical path.
Time Impact Analysis (prospective)
Time Impact Analysis (TIA) inserts a fragnet modelling a delay event into the accepted programme current at the time, and reads off the predicted effect on completion. It is the method NEC4 effectively requires for compensation-event assessment, and it is strongest when the contemporaneous programme was well maintained and the event is discrete and datable. Its weakness: it is a forecast, and a tribunal may prefer what actually happened.
Windows & as-planned vs as-built (retrospective)
A windows (time-slice) analysis divides the project into periods and identifies the critical path and its drivers in each, the strongest method for separating concurrent causes, but only if the contemporaneous updates captured both. As-planned vs as-built compares the intended and actual sequences; it is intuitive but weaker on causation where logic is thin.
What does the SCL Protocol say about choosing a delay analysis method?
The Society of Construction Law Protocol favours contemporaneous, programme-based analysis. And repeatedly makes the same point: the method must be one the project records can actually sustain.
That is the real decision criterion. A method is only as good as the contemporaneous record beneath it. Where records are strong and the question is concurrency, windows analysis usually wins; where the contract is NEC4 and the event is being assessed as it arises, TIA is the fit.
How do we choose a delay analysis method on live mandates?
On live mandates we let the records and the contract choose the method, not the other way around. And we keep the underlying schedule assurance-grade with Schedule Intelligence so the programme can carry whichever analysis the dispute demands. See this in practice in our Hafeet Rail forensic-claims engagement.
| Method | Viewpoint | How it works | Fit and limits |
|---|---|---|---|
| Time Impact Analysis (TIA) | Prospective | Inserts a fragnet modelling a delay event into the accepted programme current at the time, and reads off the predicted effect on completion | The method NEC4 effectively requires for compensation-event assessment. Strongest when the contemporaneous programme was well maintained and the event is discrete and datable. Its weakness: it is a forecast, and a tribunal may prefer what actually happened |
| Windows (time-slice) | Retrospective | Divides the project into periods and identifies the critical path and its drivers in each | The strongest method for separating concurrent causes, but only if the contemporaneous updates captured both |
| As-planned vs as-built | Retrospective | Compares the intended and actual sequences | Intuitive, but weaker on causation where logic is thin |