An as-built schedule is a record of what actually happened on a project: when each activity really started and finished, how long it took, and the order in which the work was done. The baseline is the plan and each update is the plan as revised at a data date; the as-built is the record of fact, and it underpins most forensic delay analysis, closeout reporting and lessons learned.
This guide covers what an as-built programme contains, where the as-built data comes from, how to reconstruct one from Primavera P6 updates, how to identify the as-built critical path, and how the as-built is used in as-planned vs as-built and windows analysis.
What is an as-built schedule?
An as-built schedule is a time-scaled representation of the actual sequence and duration of the project's activities. It can be presented as a Gantt chart, a tabular report or a P6 project in which every performed activity carries actual dates. Its key characteristics are:
- Actual dates: every activity that was performed has an actual start and an actual finish.
- Actual durations: the working time (or calendar time, stated which) between actual start and actual finish, including any suspensions or interruptions.
- As-built logic: the sequence in which the work was actually performed, which often differs from the planned logic.
- Traceability: each date can be traced back to a contemporaneous record or a periodic update.
Why does an as-built programme matter?
Delay claims and forensic analysis
Almost every retrospective delay analysis method relies on an accurate record of what happened. Without it, a party cannot show when a delay event actually affected the work, which path was driving completion at the time, or how much time was lost. Weak as-built data is one of the most common reasons an extension of time claim fails on causation.
Closeout and lessons learned
Owners often ask for an as-built programme as part of closeout documentation. For contractors, comparing the as-built with the baseline shows where durations and productivity assumptions were wrong, which improves the next estimate and the next programme.
Contractual records
Most construction contracts require regular progress reporting and the keeping of contemporary records, and some specifications require an as-built schedule at completion. Missing or inconsistent records weaken any later claim, whichever side makes it.
Where does as-built data come from?
The periodic P6 updates are the starting point, but they are only as good as the progress reported in them. Actual dates should be checked against, and where necessary taken from, the primary site records:
- Daily reports and site diaries: activities worked, areas, labour and plant, weather and stoppages.
- Progress records: progress measurement sheets, marked-up drawings and quantity records used to report percent complete.
- Inspection and test records: pour records, inspection requests, test certificates and handover certificates, which pin down when work was finished and accepted.
- Labour and plant returns: timesheets and allocation sheets, which show when crews were on an activity.
- Delivery and procurement records: delivery notes, purchase orders and factory acceptance records for long-lead items.
- Photographs and video: dated progress photographs, drone surveys and webcam records.
- Meeting minutes and correspondence: progress meeting minutes, RFIs, instructions, notices and permit records, which explain why work started, stopped or changed sequence.
Where the update and the site record disagree, the contemporaneous site record usually carries more weight. Record which source each date came from, because the reconstruction will be tested.
How do you build an as-built programme from P6 updates?
Step 1: Gather every schedule update
Collect every XER file or P6 backup from the baseline to completion, together with the data date of each. Gaps in the update series are themselves a finding and should be filled from site records.
Step 2: Extract and reconcile actual dates
For each activity, take the actual start and finish from the update in which the activity started and the update in which it finished. Then compare those actuals across later updates: actual dates should not move once recorded, so any change between updates needs to be explained and resolved against the site records. A side-by-side comparison of consecutive updates, using P6's Claim Digger, a schedule-checking tool or an exported report, shows where actuals were edited.
Step 3: Build the as-built logic
Planned logic often stops reflecting how the work was done. Reconstruct as-built logic by:
- reviewing daily reports, photographs and minutes to see what actually enabled each activity to start;
- using the actual dates to test sequences: if drylining started only after MEP first fix finished in that area, that is the as-built relationship;
- recording interruptions, resequencing and work done out of the planned order rather than forcing it back into the baseline logic.
Scheduling (F9) does not create or correct logic. Once every activity is complete, P6 reports actual dates and does not recalculate them, so as-built relationships have to be set by the analyst from the evidence.
Step 4: Validate the as-built
- No activity has an actual finish before its actual start.
- No actual date falls after the data date of the update that reports it (the DCMA invalid dates check).
- Every completed activity is 100% complete and every achieved milestone has an actual date.
- As-built relationships are consistent with the actual dates, with no successor shown starting before the work that genuinely enabled it.
- Actual durations are measured on the right calendar and include known suspensions.
How do you identify the as-built critical path?
The as-built critical path is the continuous chain of activities that actually determined the completion date. Because completed activities have no float, it cannot be read from a total float column. It is found retrospectively by working back from the actual completion date: identify the activity that finished last and drove completion, then the activity or event that actually prevented it starting or finishing earlier, and so on back to the start, using the as-built logic and the site records at each step.
The critical path usually shifts during a project. A contemporaneous view of the critical path at each update, compared with the as-built critical path, is what allows delay to be allocated period by period rather than in one block at the end.
What are the common pitfalls?
- Using only the last update. Activities added or deleted mid-project, and actuals edited later, are missed. Always use the full update series.
- Ignoring out-of-sequence progress. Where an activity started before its predecessor finished, copying the actual dates into the planned logic produces a sequence that could not have happened. Resolve it by correcting the as-built logic or splitting the activity, as described in out-of-sequence progress in P6.
- Assuming planned logic is as-built logic. The baseline saying A precedes B does not prove it happened that way. Verify against field records.
- Missing activities. If only key milestones were updated, many activities will have no actual dates. Fill the gaps from daily records and say which dates were reconstructed.
- Mixing calendar and working days. An activity that ran from Thursday to the following Monday spans five calendar days but only three working days on a five-day calendar. State which measure is used and apply holidays and weather days consistently.
How is the as-built used in delay analysis?
The as-built is used differently by each method. The choice of delay analysis method depends on the records available, the contract and the question being answered, and guidance such as the SCL Delay and Disruption Protocol and AACE International's recommended practice on forensic schedule analysis discusses each approach.
| Method | How the as-built is used |
|---|---|
| As-planned vs as-built | Compares baseline dates with actual dates to show where and by how much the work slipped. Simple and transparent, but on its own it does not prove which delays drove completion. |
| Windows analysis (contemporaneous period analysis) | Splits the project into windows between updates, and in each window compares the update with the as-built progress to measure delay to completion and identify the critical path that caused it. |
| Collapsed as-built (but-for) | Removes delay events from the as-built, with the as-built logic intact, to estimate when the project would have finished but for those events. |
| Time impact analysis | Inserts a delay fragnet into the contemporaneous update before the event, not into the as-built. The as-built is then used to check that the modelled impact matches what actually happened. |
Whichever method is used, the as-built is the reference point. A carefully reconstructed as-built programme, with traceable dates, evidenced logic and a reasoned as-built critical path, is what allows an analysis to stand up to an opposing expert. Building and testing it is a core part of forensic planning work.
