When we develop the Baseline CPM Schedule for a construction project, we need to develop the WBS to establish the project deliverables.
To what level of detail is this required?
The general idea is to decompose the project deliverable down to the Work Package level. For construction, this could be taking the WBS down to an area of structural foundation for a large project. It is really taking the deliverable down to the point that it can be easily identified and measured.
We then develop the “work” activities that need to happen to produce this deliverable.
Basically, we develop the list of activities, sequence them, assign the resource(s) and determine the Activity Duration, and assign the activity’s Budgeted Cost. There are also the Activity Code, Calendar, and Role assignments…
Many specifications have limits on the number of Work or Calendar Day duration a “work” activity may be assigned. A good rule of thumb is “not to exceed the reporting period”. I like to cap my “work” activities at 20 Work Days or less. An activity should only have one responsible party performing the work. If an activity only has one party performing the work and the duration of the work is going to exceed 20 Work Days, perhaps it is best to break the work area down to allow smaller durations. It is difficult to accurately measure the performance of activities with large areas or durations.
The use of activities with large Activity Durations also leads to the use of Start-to-Start and/or Finish-to-Finish relationships to model the work which is now running concurrently with other large duration activities. This should be limited. The overuse of Start-to-Start and/or Finish-to Finish relationships is a basic problem for many detailed project schedules. Makes a pretty Gantt Chart, but does not allow adequate control and measurement of scheduled work.
Most deliverables should have activities with durations much less than 20 Work Days. They should not all be 1 Work Day unless that is actually what it takes to model the work of several parties. Usually, the “work” could be broken down to 5 days to excavate the foundation, 2 days to install reinforcing steel and any vapor barrier or blockouts, and 1 day to place concrete and an activity with a period of time for concrete curing. Breaking the work down to this level allows the work to be easily measured and managed. Too often I see a “Place Foundation Concrete” activity with a duration of 10 Work Days. No cure activity. No idea if cure time is included or not…. No idea when the excavation should complete or when the reinforcement placement starts…. Most likely, the formwork and reinforcement work will run concurrently at some point and the concrete placement will be driven by the finish of the reinforcement work and inspection. If there is a delay in delivery of reinforcement steel, we should be able to accurately model that…..
The idea is to develop a schedule which is manageable and has enough detail to control the work. Pretty simple.
If there is more than one contractor performing the work, you probably need to break down the work further. If a series of activities are all running concurrently, perhaps it would be best to break the work down into smaller areas. The goal is to use Start-to-Start and/or Finish-to-Finish relationships as little as possible. Using Finish-to-Start relationships is almost always the best way to model the work. This usually requires breaking the work down into smaller pieces…
I’m sure many of you have comments or additional insight into this subject. Please share!
I’d love to hear what you think!
Please visit https://conschmanservices.com to learn more about Construction and Schedule Management Services, LLC
Please visit my LinkedIn account to learn more about me.
Please visit my “The Blue Book” ProView.
Paul Epperson CCM, PMP, PSP, PMI-SP