Lag Calendars

[TMBoyle]

I am using MSP 2003 (Std) on a large engineering and construction project
where some tasks are scheduled on a 7-day work week and some are scheduled on
a 5-day work week. I am becoming more confused regarding MSP's schedule
computations when lagging relationships and such non-standard task calendars
are involved. I have read in this forum and elsewhere that MSP uses the
Project Calendar to account for lag time, but the results in my working
schedule are not consistent with this.
To check, I prepared a little three-task A-B-C network. For simplicity,
all three tasks are fixed duration tasks of 20d with no resources. A is the
predessessor of both B (SS+10d) and C (FF+10d). I then systematically
switched the task calendars for A, B, and C between "None" (i.e. the Standard
Project Calendar, 5-day workweek), and a newly-created 7-day workweek. In
all cases, the start dates for the successor activities accounted for the
specified lag based on the SUCCESSOR TASK CALENDAR, not the Project Calendar.
Can someone else please repeat this 10-minute exercise and confirm the same
result?

In the current project, I have a number of construction activities with a
7-day task calendar, while the overall project calendar is a 5-day calendar.
I ignore weekends in specifying the lag times between these activities (i.e.
7d lag is Monday-to-Monday, not Monday-to-Wednesday), and they have all been
scheduled correctly. This would not have occurred if the Project Calendar
was the basis for all lag calculations.

So, is the schedule impact of lag actually computed using the Successor Task
Calendar as observed, not the Project Calendar as commonly stated?

Thanks for any and all contributing opinions.

tmb

 

[John]

I am using MSP 2003 (Std) on a large engineering and construction project
where some tasks are scheduled on a 7-day work week and some are scheduled on
a 5-day work week. I am becoming more confused regarding MSP's schedule
computations when lagging relationships and such non-standard task calendars
are involved. I have read in this forum and elsewhere that MSP uses the
Project Calendar to account for lag time, but the results in my working
schedule are not consistent with this.
To check, I prepared a little three-task A-B-C network. For simplicity,
all three tasks are fixed duration tasks of 20d with no resources. A is the
predessessor of both B (SS+10d) and C (FF+10d). I then systematically
switched the task calendars for A, B, and C between "None" (i.e. the Standard
Project Calendar, 5-day workweek), and a newly-created 7-day workweek. In
all cases, the start dates for the successor activities accounted for the
specified lag based on the SUCCESSOR TASK CALENDAR, not the Project Calendar.
Can someone else please repeat this 10-minute exercise and confirm the same
result?

In the current project, I have a number of construction activities with a
7-day task calendar, while the overall project calendar is a 5-day calendar.
I ignore weekends in specifying the lag times between these activities (i.e.
7d lag is Monday-to-Monday, not Monday-to-Wednesday), and they have all been
scheduled correctly. This would not have occurred if the Project Calendar
was the basis for all lag calculations.

So, is the schedule impact of lag actually computed using the Successor Task
Calendar as observed, not the Project Calendar as commonly stated?

Thanks for any and all contributing opinions.

tmb

TMBoyle,
I'm not sure the stuff you read about lag time being associated with the
Project calendar is applicable in the context of your scenario. I tried
your example and indeed the lag time in the predecessors follows the
calendar of the successor task - and that makes sense. Why, because the
predecessor really "belongs" to the successor. In other words, the
predecessor is a modifier for the successor start date, (or finish date
as the case may be), and since the start date is determined by the task
calendar, so should the predecessor.

Without knowing the details of how your construction tasks dovetail with
your other tasks, it is difficult to make any suggestions as to how to
best set up links. However, you might like to know that you can also use
elapsed days in a link (e.g. 1SS+10ed). That may or may not help in your
case.

John
Project MVP

 

[John]

Thanks very much for the reply John. It seems had I (and others)
misinterpreted a statement by another MVP from a May 2006 post. So in-fact
MSP computes the working-time lag based on the successor Task Calendar, not
necessarily the default (Project standard) calendar. I also appreciate the
logic for using the successor task rather than the predessesor task.

Thanks again.

TMBoyle,
You're welcome. Just to check the context, I did a quick search on the
newsgroup questions for May 2006 but I didn't find anything relating to
working time lag, although I didn't read every post. The only comment I
can make is that for a project file with a single calendar, (the
"normal" plan scenario), the lag is based on the Project calendar,
because it is the only calendar. However, that is not your case, so
there are exceptions to every "rule".

John
Project MVP