Key points

  • AACE Recommended Practice 52R-06 describes time impact analysis: insert a fragnet for the delay event into the accepted update whose data date is just before the event, reschedule, and measure the change in completion.
  • Any correction made to the unimpacted update must be disclosed, because it can move the result as much as the fragnet does.
  • Record the P6 scheduling options (Retained Logic or Progress Override, critical by total float or longest path), because they change the calculated dates.

In brief: A time impact analysis (TIA) measures the effect of a delay event by adding a fragnet to the schedule update current before the event and recalculating the schedule. AACE Recommended Practice 52R-06 sets out how to do this. In Primavera P6, most of the work is in choosing the right update, building a fragnet that the other side can follow, and keeping the scheduling options the same between runs.

What AACE 52R-06 covers

AACE International Recommended Practice 52R-06, Time Impact Analysis As Applied in Construction, describes a modelled, additive method. It is mainly a prospective tool, used to estimate the effect of a delay event at or near the time it happens so that the parties can agree an extension of time before the work is done.

The RP can be used retrospectively, but it warns that the results then need to be checked against what actually happened. Once the as-built facts are known, a tribunal may prefer a method that uses them, such as a windows or as-planned versus as-built analysis. AACE RP 29R-03 on forensic schedule analysis covers those methods.

On US projects, many specifications require a TIA for any request for a time extension, and some name 52R-06 directly. Read the scheduling specification before you start. It often sets the format, the submission period and the software settings to be used.

Step 1: select the unimpacted schedule

Use the accepted schedule update with a data date as close as possible to, and before, the start of the delay event. That update is the unimpacted schedule. Do not use the baseline unless no update exists, because the baseline does not reflect the progress and logic changes made since the start of the job.

In P6, copy the project and give the copy a clear name and ID, for example the update number followed by “TIA-03 unimpacted”. Never work on the live update. Export the copy to XER so that the starting point can be reproduced.

Step 2: check the update before you use it

A TIA is only as reliable as the schedule it is built on. Before inserting anything, check the unimpacted schedule for:

  • activities without predecessors or successors, other than the start and finish milestones;
  • hard constraints that stop dates moving, such as Mandatory Start or Finish On;
  • out-of-sequence progress and actual dates after the data date;
  • remaining durations that do not match the reported progress;
  • calendars that do not reflect the working pattern on site.

If you correct anything, keep a log of each change and the reason for it. Corrections made to the unimpacted schedule must be disclosed, because they can change the result as much as the fragnet does.

Step 3: record the scheduling options

P6 scheduling options affect the calculated dates. Record them for the unimpacted run and use exactly the same options for the impacted run.

P6 settingWhy it matters
Retained Logic, Progress Override or Actual DatesChanges how out-of-sequence activities are scheduled
Define critical activities as Total Float or Longest PathChanges which activities show as critical
Compute Total Float as Start, Finish or SmallestChanges the float values reported
Make open-ended activities criticalCan hide or create critical paths
Calendar for scheduling relationship lagChanges lag durations across non-working days
Ignore relationships to and from other projectsMatters where external links exist

Run the schedule (F9) with Log to file switched on, and keep the schedule log for both runs.

Step 4: build the fragnet

The fragnet is a small network of new activities that represents the delay event. It might show the time to issue a revised design, approve a submittal, or carry out added work. Build it from the records, not from estimates made after the event.

  • Use a separate activity code or ID prefix so the fragnet activities are easy to find, for example “TIA03-“.
  • Base each duration on a record: the date of an RFI and its response, the contractual review period, or the planned output rate for the added work.
  • Tie the fragnet into the existing network with logic, at the points where the event actually affects the work.
  • Do not use constraints to force the fragnet into place.
  • Do not change existing activity durations or logic, unless the change is part of the event and is explained.

Write a short description for each fragnet activity and each new relationship. The reviewer will check every one.

Step 5: schedule the impacted model and compare

Copy the unimpacted project, insert the fragnet and schedule it with the same data date and options. To compare the two, assign the unimpacted project as the project baseline of the impacted copy (Project, Assign Baselines). Columns such as BL Project Finish and Variance – BL Project Finish Date then show the movement of each activity.

The difference between the completion dates, or the contract milestone dates, is the modelled delay. Also report:

  • the critical path before and after the fragnet is added;
  • the total float on the affected path in both runs;
  • any float used up without a change to completion.

If the event only consumes float, the analysis shows no delay to completion. Whether the contractor is entitled to time in that case depends on the contract, in particular on who owns float.

Several events

Where there is more than one event, analyse them in the order in which they occurred. Each TIA should be built on the update current at the time of that event. If two events overlap in the same update period, you can model them together and then one at a time to show the contribution of each. Explain the approach in the report.

Contractor delays in the same period need to be considered. The update will already include them if progress was recorded correctly. A TIA that ignores them is likely to be criticised.

Step 6: the report

The submission should let the reviewer repeat the analysis. It normally includes:

  1. a short narrative of the delay event and the contract clauses relied on;
  2. the identity of the unimpacted update and any corrections made to it;
  3. a list of fragnet activities, durations and logic, with the source of each duration;
  4. the scheduling options and the schedule logs;
  5. a comparison of completion dates and critical paths;
  6. the unimpacted and impacted XER files.

Keep the language plain. A short, clear TIA that the owner’s scheduler can open and rerun is more likely to be accepted than a long one that cannot be checked.

This article is general guidance, not legal advice. The contract, the scheduling specification and the facts of each project decide how a TIA is assessed.

Related: Delay analysis methods compared · CPM scheduling services · REA vs claim under the FAR

Need support with this topic?

Solvix prepares and reviews time impact analyses in Primavera P6 for contractors in the USA and internationally. Book a free 30-minute call to talk through your project with us.

Ahmed Elsayed
Ahmed Elsayed
Managing Director

Managing Director of Solvix International. Ahmed leads the firm's delay analysis, construction claims and project controls work, and holds an MSc with a specialism in delay analysis. He has prepared delay analyses, extension of time claims and Primavera P6 programmes for contractors, employers and legal teams on projects across the Middle East, Africa, Europe, North America and Australia.

Leave a Reply

Your email address will not be published. Required fields are marked *