Collapsed As-Built / But-For

What it does

Starts from a completed (or largely complete) schedule, sets the chosen excusable delay activities to zero duration, rebuilds the network from its logic and as-built durations, and reports how much earlier the project would have finished without them. The delay is measured against a rebuild of the same schedule with nothing removed, so systematic differences between real dates and logic don't get blamed on the delay.

What it helps you achieve

• Estimate the time attributable to specific events after the fact.
• See which activities would have moved, by how much, and how the finish changes.
• Understand how far the recorded dates depart from the schedule's own logic.

How to run it

Where to find it: Project page → Collapsed As-Built

1. Open the project and choose Collapsed As-Built, then + New CAB analysis.
2. 1. Choose the as-built schedule: a complete or substantially complete update.
3. 2. Choose the excusable delay activities to collapse (only completed activities are offered).
4. Choose Run Analysis.
5. Read the four figures: As-Built Finish (real dates), Rebuilt Finish (nothing collapsed), But-For Finish (collapsed) and the delay attributable.
6. Review the table of activities that shifted, and the Gantt overlay showing the but-for position.

What you get

• As-built, rebuilt and but-for finish dates side by side.
• The delay in calendar days and working hours.
• A per-activity table of how far each moved.
• A disclosed 'logic gap': how far the real finish sits from a logic-only rebuild.
• An Excel export and a saved, reopenable result.

Good to know

• A collapsed activity is set to zero duration and kept in the network, so the schedule's sequencing is preserved. It is not deleted and re-linked, which would require guessing how to reconnect its neighbours.
• A negative or unexpected result usually says something about the schedule, not the delay; the logic-gap note is there to show exactly that.

Frequently asked questions

Why measure against a rebuild rather than the real finish date?

Recorded dates rarely follow a schedule's logic exactly (out-of-sequence work, waiting no relationship captured). Comparing a logic-only recalculation with the real dates would blame that gap on the delay you removed. Measuring against a rebuild with nothing removed cancels it, and the gap is reported separately.

Go deeper

• Collapsed as-built, explained →
• Why a collapsed as-built can show a negative delay →
• But-for delay analysis →

Terms used here

• Collapsed As-Built
• But-For Delay
• As-Built Schedule
• Excusable Delay