A time impact analysis (TIA) is a method for measuring how much a specific change or delay moves a project’s completion date. You model the event as a small group of activities, insert it into the schedule that was current when the event arose, recalculate, and compare the completion date before and after. The difference is the time impact.
It is the analysis most CPM scheduling specifications ask for when a contractor requests more time, and the one owners most often have to evaluate. Done promptly and simply, it settles time alongside the change order. Done late, it becomes the opening exhibit in a dispute.
Key points
- A TIA compares one schedule with and without a modeled event. The difference in completion is the impact.
- Start from the last accepted update with a data date before the event began, and check that update before adding anything.
- A delay that never reaches the critical path uses float and usually earns no time extension.
- Prepare it when the event happens. An analysis written a year later invites argument.
When is a time impact analysis required?
Most specifications that require a critical path method (CPM) schedule also require a TIA — sometimes called a fragnet analysis — to support any request for a time extension. The recognized industry description of the method is AACE International’s Recommended Practice 52R-06.
The contract usually sets two things you cannot ignore: the form the analysis must take, and the deadline for giving notice. Notice periods are short. The 2017 edition of the widely used AIA A201 general conditions, for example, requires a claim to be initiated within 21 days of the event or of first recognizing it, and notice of concealed or unknown site conditions within 14 days of first observing them. Your contract may be stricter, so read the time-extension and notice clauses before anything else.
Typical triggers include owner-directed changes, late or revised design information, differing site conditions, delayed owner-furnished items, and unusually severe weather where the contract treats it as excusable.
How to prepare a TIA, step by step
- Define the event. What changed, when it started, and which documents record it: the RFI, bulletin, directive, daily reports and correspondence.
- Choose the right schedule. Use the most recent accepted update with a data date before the event began. Check that it is properly statused and that its logic is sound. An analysis is only as credible as the schedule it starts from.
- Build the fragnet. Model the event as activities with durations: for example redesign, resubmittal, review, procurement and the added work. Each duration should trace to a document.
- Tie it into the network. Give the fragnet a predecessor where the event begins, and successors at the work that cannot proceed until it ends.
- Recalculate and compare. Run the schedule with the same data date and settings. Compare the completion date and the critical path with the unimpacted version.
- Check concurrency and mitigation. Was something else delaying the same path in the same period? Could resequencing reasonably reduce the impact?
- Write it up. A short narrative, a graphic of the fragnet, the before-and-after comparison and the supporting records.
A worked example
A five-story building is under construction. The latest accepted update forecasts substantial completion on March 12, with interior close-in and finishes on the critical path and no float.
The owner then issues a bulletin relocating the level 2 electrical room. The work it affects — level 2 wall close-in — was due to start 10 working days after the bulletin arrived.
The fragnet has four activities, 20 working days in total:
- Revise shop drawings — 3 days
- Engineer’s review — 5 days
- Procure revised conduit and supports — 8 days
- Install revised rough-in — 4 days
It starts when the bulletin is received and is tied to level 2 close-in. Close-in can no longer start in 10 working days; it has to wait 20. Because close-in is critical, the 10-day push carries straight through to the end: completion moves from March 12 to March 26. The time impact is 10 working days, or 14 calendar days.
Change one fact and the answer changes. Had close-in been carrying 15 days of float, the same event would have used 10 of them and moved nothing. The event would still be worth documenting, because the float it consumed is no longer there for the next problem.
Time impact analysis or windows analysis?
The two are often confused. A TIA is a modeled method: it adds an event to a schedule and forecasts the effect. It works best prospectively, for one event at a time, while the project is under way.
A windows analysis is an observational method. It steps through the project one update period at a time and measures how the critical path and the completion date actually moved in each. It suits a look back over a job where many events overlapped and regular updates exist.
Both families are described in AACE International’s Recommended Practice 29R-03 on forensic schedule analysis. As a rule of thumb: use a TIA to price time into a change as it happens, and a windows analysis to explain a late finish afterward.
How to run a time impact analysis in Primavera P6
- Make a copy of the accepted update. Keep the original untouched as the unimpacted schedule.
- Before changing anything in the copy, save a baseline of it (Project, Maintain Baselines) and assign it, so the unimpacted dates stay on screen.
- Add the fragnet activities, coded so they can be filtered and shown on their own.
- Add the logic ties into and out of the fragnet. Do not use date constraints to hold it in place.
- Schedule the copy with the same data date and the same calculation settings as the original.
- Show the baseline bars beside the current bars in one layout, and compare completion dates and critical paths.
- Export the fragnet layout and the before-and-after comparison for the narrative, and keep both native files.
The same steps work in Microsoft Project: save a copy, add and link the fragnet tasks, and compare against a saved baseline.
Common mistakes that get a TIA rejected
- Starting from the wrong schedule. Using the original baseline, or an update long after the event, instead of the update current at the time.
- Unsupported durations. Fragnet activities with durations no document backs up.
- Constraints instead of logic. Pinning the fragnet in place with dates hides what is really driving it.
- Ignoring concurrent delay. If the contractor’s own work was late on the same path, the reviewer will find it.
- Bundling events. Several changes inserted at once make it impossible to see what each one did. Insert them in the order they happened.
- Leaving it too late. Missing the notice period can cost the entitlement, however good the analysis.
This guide is general information about construction scheduling, not legal or contract advice. Your contract governs.