Collapsed As-Built Dashboard Larchmere Library and Community Hub (Demo Data) · AACE RP 29R-03 §3.8 (Modeled / Subtractive / Single Simulation)

Removing the seven delay events from the as-built schedule brings completion 20 working days earlier, from 4 September 2026 to 7 August 2026. Removed in turn, the Owner's events account for 12 working days and the Contractor's for 8.

The windows analysis of the same schedules measures 86 working days of movement against the 20 working days found here, a gap of 76.7%; section 3 explains why.

The finish dates here are the opening of the working day after the last day worked: the as-built schedule’s last day of work is 3 September 2026.

The as-built update was filed at data date 5 August 2025 but records actual progress up to 14 October 2025, so it is read at 15 October 2025, as section 4 sets out.

Each event was taken out of a copy of the as-built schedule, which was then recalculated: removed alone, the finish moves by the event’s own effect; removed in turn, each event adds to the ones before it.

The windows analysis measures 86 working days of movement between the schedules; removing the events here gives 20 working days, 66 working days apart.

The two methods measure different things. The windows analysis measures the whole movement of the finish between the schedules, whatever caused it. Collapsed as-built takes out only the events it is given, each as a shortened duration or a deleted activity, and holds completed work at its actual dates, so movement that comes from changed logic or progress lost is not an event it can remove.

The results differ by more than 15%, so this result does not corroborate the windows figure: what it measures is the effect of the events it removes, not the whole movement of the finish.

The 5% and 15% bands are CPP practice thresholds, not figures from AACE RP 29R-03; exhibit A-04 sets out the comparison.

The figures come from the as-built schedule recalculated with each event taken out; completed work stays where it happened.

Method: AACE 29R-03 MIP 3.8 (modeled, subtractive, single simulation). For each event the as-built schedule is copied, the event removed (its added duration shortened, or the activity deleted with its links) and the schedule recalculated; the change in the finish is that event’s effect. Removed together, the events give the cumulative figure. The default removal is shorten.

Completed activities are held at their actual dates, so a delay that sits on completed work cannot be removed here: such an event is marked locked and reports 0 working days, which is not a finding that it had no effect.

Anchor check: the schedule’s stored finish and the recalculated finish agree, so every date here starts from the schedule as filed.

How the Schedule Was Read

Basis of Measurement

Diagnostics

The run raised four diagnostics; exhibit A-05 lists every one with what it means for the result.

The exhibits behind these findings follow in full, each under its drawing code:

• A-01Methodology
• A-02Per-Event Subtractive Pass
• A-03Cumulative Roll-Up
• A-04Dual-Method Gap
• A-05Diagnostics and Warnings

For each delay event below, the as-built network was cloned, the event's impact removed using the indicated method, and the CPM forward pass re-run. The resulting "but-for" finish and computed impact are shown.

Methods disagree by more than 15%. The forensic conclusion is NOT defensibly grounded under practitioner-consensus convergence cut-offs. Significant further investigation is required before submitting the claim.

Thresholds: AGREE ≤ 5% · WARN 5-15% · INVESTIGATE > 15%. Bands are CPP heuristic thresholds. They are NOT codified numeric cut-offs in AACE RP 29R-03 (which permits more than one method in §5 but describes no cross-check and publishes no numeric tolerance percentages).

4 diagnostics across 4 detector codes. The triage strip below states, per detector, what its findings mean for the reliability of the but-for result; the full registers follow, grouped by detector, every entry listed.