One of our customers recently encountered the perplexing situation whereby total float was greater than zero on some of their critical activities in Primavera P6. Here's how http://ow.ly/S1Pr30cfV3l
Primavera P6 - Float on Critical Activities?
Forum Sponsor
Top Posters
sairedz25
0 posts
Ahmed Awad
2 posts
Syed Shoeb
0 posts
Vimukthi
0 posts
bal aji
2 posts
Lee Mallek
23 posts
Viet Tran
9 posts
Ola Gbotoso
0 posts
Jaturapit Multongka
1 posts
James Williams
74 posts
If the software cannot consistently yield reliable resource leveled float values no matter what the threshold is then the output is not reliable.
the critical activities can be defined at any total float value threshold that you set (the default is 0).
You can have a critical path that has positive total float.
Under resource leveling P6 Longest Path calculations are flawed, even if you could make P6 to disclose reliable resource leveled float values.
This is so because Longest Path Theory was devised before CPM of the 70's. The workarounds cannot deal with more modern and advanced models (and a few old ones) that in addition to date constraints, multiple calendars and renewable resources also include leveling for consumable resources, spatial resources, financial resources and a long list of model enhancements not considered 50 years ago.
Even for the simplest of consumable resource leveled schedule Longest Path will not be disclosed. Elementary if there are not enough bricks activity will be delayed until replenished, no rocket science. Still resource leveling consumable resources can be a challenge unless the computer algorithm does it for you.
Have you ever wondered why the controversial AACE International Delay Analysis RP does barely mention resource constraining and gives no guidance at all with regard to this issue? Maybe they are so much into Longest Path that they can't see the forest for the trees.
Resource allocation can affect the results of a delay analysis, performing a schedule analysis without considering resource allocations may increase the owner’s or contractor’s risk of assuming delay responsibility which is not his or her fault.
Hi Mike,
Your curing calendar is a typical example of the necessity for cross-calendar links along logical paths. There are many more: shiftwork, weather, seasonal shutdowns, contractual restrictions, 5d office vs. 6d field vs. 7d contract milestones, monthly board meetings, etc. Yes, Longest Path (and Multiple Float Path w/ Free Float option) overcomes this problem in P6.
Such options don't exist out-of-the-box in MSP. What is the Asta approach?
Hi Tom
When you use a task to represent a curing period - or a lead lag for that matter - it has to be set as a 24/7 calendar.
When the end of that curing task lands in a non work calendar zone all the up stream criticality disapears.
The longest path restores the critical path.
Best regards
Mike Testro
This article demonstrates why Total Float is not a reliable indicator of the Critical Path in real-world project schedules, and why Longest Path - i.e. the driving path to project completion - is generally preferred in P6. It provides concrete examples for understanding the impact of activity calendars on P6's computation of activity dates and float - namely when a relationship crosses a time interval that is working time for one activity but not for the other. This is all good.
The subsequent recommendation not to link activities on different calendars is a ridiculous concession to those who haven't understood anything else in the article.
".... While it’s quite legitimate to use multiple calendars in any schedule, it can present problems if we attempt to mix those calendars within the logic of the schedule. That is to say activities on one calendar should not be linked to activities on a different calendar if at all possible. Doing so can generate some very perplexing results in not only the dates, but the critical path."
Even in a well-constructed schedule with multple calendars, cross-calendar links are unavoidable, and the existence of any of them can make Total Float unreliable as an indicator of driving logic. Reducing such links from many to few isn't guaranteed to improve anything. (Late constraints and resource leveling, though not addressed in the article, similarly degrade the value of Total Float.)
Unfortunately, recommendations like the one above have a tendency to be repeated, and before you know it some 170-point schedule health checker will start flagging cross-calendar links as "poor scheduling practice." That would be one more step downward in the quality of project schedules for their intended purpose.