Identifying Second & Third Longest Path's in Primavera P6

1. We increase probabilities of success by targeting a tight schedule with a finish somewhat before contract delivery date. In the absence of Schedule Risk Analysis our hands down approach is to use a schedule that is 20% shorter.
2. We avoid/reduce claims for concurrent delays when we do not show delays.
3. Pacing is usually claimed by the other party as a contractor’s delay.

Being said so our strategy considers mitigating concurrent delays as follows.

• We use two versions of the schedule in our jobs, the baseline meeting all contractual conditions and a tight schedule.
• We use the baseline schedule as our reference for contract/claim administration that shows longer activity durations than our tight schedule.
• On site use the tight schedule to manage our activities.
• We avoid pacing as suggested in the following reference, please click the hyperlink.
• Pacing Delays  – “Contractors seeking to maximize recovery due to compensable delay should not attempt to pace an owner delay”.
• In this way we avoid/minimize the possibility of apparent concurrent delays that might jeopardize valid claims. 

This is a very good point you have raised for discussion.

Thank you

What about resource calendars? 

• In a portfolio of jobs with hundreds of jobs the resource calendars count can be in the hundreds.  
• In our portfolios limited resources each with their particular calendars are shared and must be leveled.
• Frequently teams of resources work on their particular shift calendar.
• Some resources such as equipment can work on several shifts.
• Hundreds of calendars must be considered, no big deal.
• Any serious software should be able to handle all of them without this invalidating float calculations.

TIA procedure does not mention longest path but float, search the AACE International Recommended Practice No. 52R-06 for float and you will get some hits, search for longest path and you will get none. 

52R-06: Time Impact Analysis - As Applied in Construction

• STEP 7:Determine the actual dates of the delay. Using the original schedule update, determine when the successor activity to the delay impact actually became a project critical activity. On schedules without negative float, the activity will be predicted to become project critical on the computed late start date. The first date of delay due to this impact will be the next day after the activity late start date. For original update schedules that do show negative float, the start or delay date will simply be the first day of the delay event. Every day after this start of delay will be labeled a delay day (counting work days or calendar days as appropriate) until the number of delay days is exhausted.

On the other hand concurrency analysis in Windows TIA mentions longest path, critical and near critical Activities.

(CDR-2849) Time Impact Analysis in Windows - Concurrency Analysis

But as documented on the web P6 float and longest path calculations are unreliable when using multiple calendars, more so when performing resource leveling.

Does this means when using multiple calendars and/or resource leveling P6-TIA is unreliable?

• I guess so.

The link is on slide 16 and it is working, you must open the presentation and when in slide 16 click on it.  For your convenience I am pasting the link as follows.

http://www.spiderproject.com/

Be reminded longest path gets broken as soon as the schedule model must deal with multiple calendars and all sort of constraints, it soon becomes useless.  

I would not recommend switching scheduling and resource management software in the middle of such a big job.  P6 longest path and float values might be unreliable but you can still perform what-if analysis.  I do not expect P6 early dates to be flawed. Then within available time learn and explore Spider Project. 

Take a look at the following image I got from the web, P6 free float calculations seems to be equally unreliable.

A1020 is critical but shows free float, no matter what it is weird, very weird ...

I use Spider Project, the link is on the presentation, just click the hyperlink.

The longest path is broken when activities are no longer driven by relationships, it has come of age, it is time to let it rest.

Project schedule constraints include resource constraints, finance and supply constraints, calendar constraints and imposed dates. The critical path in the projects with imposed dates and different activity and resource calendars can consist of only one final activity. Managing project schedules, one should pay more attention to activity floats. The activities with shorter floats are more critical than the activities with longer floats. However the float should be calculated with all schedule constraints as well as the network logic taken into account.

Activity float is a quantitative measure of criticality.  That some software cannot deal with it is another thing.  It is better to switch to software capable of yielding reliable activity float values than to add patches to the software hoping the bandage will cure the cancer.

Sean,

1. As an SA user, you’re going to understand SA better than I, but what you describe doesn’t match what Ron is selling.  His website implies that every single activity in the P6 schedule is assigned (in UDFs) a “Longest Path Value” (LPV) and an LPV sequence number.  LPV=0 designates the Longest Path, and successively higher numbers indicate relative distance from the Longest Path.  Grouping and sorting by these UDFs may provide a better quantitative comparison of the relative float paths compared to MFP results.
2. Your desire to repeatedly display the same activity as a member of multiple distinct, intersecting, full-length paths is understandable, to simplify presentations for your audience.  In the long run, however, such paths in complex project will tend to be largely duplicative, and a special effort will be needed to highlight their differences and their intersections.  I’m not sure that’s worth the trouble.  In any case, its going to be difficult to flag activities as part of multiple paths without using a 100% external tool or report.
3. Have a look at “Deep Schedule Analysis” from hcp-consulting.com.  This claims to give you exactly what you are looking for, in report form.  (Rafael makes a good point re. the resource leveling.  If you use SA’s Logic League module to handle those aspects ahead of time, then HCP may be able to give you the resource-constrained Critical and near-Critical paths.)

Good luck, tom

P6 users should not assume that P6 will always produce a reliable critical path even if the poor scheduling practices are avoided.

https://www.linkedin.com/pulse/primaveratitis-critical-path-identification-disorder-wail-menesi-phd/?trackingId=olTYas1sw237ZApRAb%2BaTw%3D%3D

If the software cannot reliably identify 1st longest path for simple models do not expect 2nd, 3rd and so on will be reliable.  As you add complexity such as multiple calendars and date constraints to name a few we can expect longest path calculations to be less reliable even if you use Schedule Analyzer Logic League.

Sean,

The Longest Path Value module in Schedule Analyzer Software (http://scheduleanalyzer.com/sa_brochure.htm) might be useful.  I understand it works directly in the P6 database.  Combining LPV with the Logic League module of the same software might give you the resource-leveled longest and near-longest paths.  I say "might" because SA promises to do this stuff, but I'm not an SA user nor do I know any users who have successfully exploited these two modules in a modern P6 schedule.

Besides the linked reference above, Longest Path Value is described in AACE Recommended Practice 49R-06, Identifying the Critical Path.  Longest Path Value and Multiple Float Path analysis are the only two non-float-dependent analysis methods included in another relevant AACE Recommended Practice: 92R-17, Analyzing Near Critical Paths.

You haven't spelled out your problems with MFP analysis in P6.  Are you using the Free float option, as recommended?  Unfortunately, MFP can choke on concurrent branches to the Longest Path when multiple calendars are in play; it also chokes on ALAP-constrained predecessors that are not driving.  When these exist, then you need some manual checking.  In the end, you can always manually-define the relevant paths by click-tracing through the schedule network using the GoTo button - paying close attention to relationship free float and relationship successor free float.

Good luck, tom