Hi All,
I am putting together a project S-curve analysis presentation for project managers and I am looking for a reliable source that can provide a comprehensive scurve analysis practice, specifically Early and Late data S-curve with actual progress.
an analysis the indicate if the project is in a good standing or not, what is the likelihood of delay if actual curve falls very close to late curve, etc...
I highly appreciate it if anyone provides me a document, link, etc to this sort of project analysis for project managers.
Thanks in advance
Too much thought is put into the "S" Curve. I use only two standard curves for all EPC Projects. One a 10-80-10 curve noting that 10% Progress would be completed in the first 25% of total schedule duration - 80% in the next two quarters and 10% at the end and the second a 15 - 75 - 10 curve indicating 15% Progress in the first 25% of the total duration. This is based on the simple fact that if you could place maximum workforce on site on day one and earn progress at optimum monthly values it would be 25 - 50 -25 so thus highly unlikely. Once you understand EPC projects are all about opening work space early in the process with little manpower and cost value - it makes common sense that no more than 10 to 15% progress would be achieved in the first 8 months of a 2 year project. Same occurs at the end with workspace dimminishing . I typically allow 3% either side to indicate a valid schedule to account for reasonable deviations. If its a vertical project like a highrise, has a lot of early demolition and hazardous earthwork - its a 10-80-10. If its a greenfield lowrise with little to no evident complications at the start. No matter what the project or what you are plotting (Cost, Manpower or earned value) if you significantly stray outside these boundarys yo hav a problem with your resource loading that would indicate numerical bust in your schedule resources or if you're trying to validate a contractor's schedule - I would look for FEL of cost.
Ok Rodel has the same approach as me use the early and late dates for the curve and then use and average.
You can also do as Rafael shows
here is a link to a free schedule validator that based on the status of the project perdicts the likehood of completing on time.
the basis is the DCMA 14 point schedule analysis with a few other factors in the mix.
https://schedulevalidator.com/
Please be reminded that:
It’s usually done by average of early and late. Or you can run a range analysis using P80. You can also use the critical Path or longest path for median.
Thanks, I appreciate it.
my other question would be how should we define the median curve ? should it be 50-50 split of early and late unit or 60% or early and 40% of late unit?
Thanks Rafael Davila . Appreaciate it
There is much misconception about the use and interpretation of the S-curves. I find the AACE International Recommended Practice No. 55R-09 is an incomplete document regarding S-Curve Analysis. Among other issues, the AACE RP does not mentions how under negative float P6 S-Curves are drawn and how are to be interpreted when late bars are projected earlier than early bars.
55r 09 Analysing s Curves - [Download PDF]
A complete set of S-Curves will not identify the critical path but will tell you if the job is on time or not, is limited on what it disclose. For decades top management have been looking at these curves for a fast birds-eye view on actual status and projections.
When updating the schedule and the period of concentrated work is marked as 100% completed on an uncompleted activity with minimal remainig work the schedule and the S curves might be distorted unless the issue is properly addressed. If there is minimal work remaining say 5% and the activity was marked 100% complete as to then the remaining 5% must still be scheduled while avoiding Out-Of-Sequence logic.
The Great Divorce: Cost Loaded Schedule Updating