measuring delay with s curve

Hi Rao

It is not easy to measure delay using only an S Curve - particularly if it is cost generated - because there is no direct link between expenditure and progress.

The S Curve is an indication that delay has occurred but there is no demonstration of cause and effect.

Stick to the proven method of cause and effect on the critical path.

Best regards

Mike Testro

I have seen many S Curves without the Projections shown, just the Baseline and Actuals, in such case you have less than zero.

In the above curve you can see actual at Data Date are near the average of early and late baselines curves but it does not tells you why the cost nor delay overruns, with the projection side you will see how much cost overrun and time delay is expected but not why. It is a birds-eye view and as such shall be interpreted. 

Hello Rao,

It's only if the costs being captured are are exactly the same as you had previously budgeted for where you can then measure progress.

However, this is very unlikely and you need to refer to your baselined programme to get this information.

Point for the future, your budget is firm and fixed at a given point from which to measure, the costs being incurred are variable but you should know the reasons why for reporting purposes.

Best wishes,

Ian

The flaws on using traditional EV is well known and documented, it has been suggested to update its use with the concept of an extension known as Earned Schedule.

From:

http://www.pmforum.org/library/second-edition/2011/PDFs/jan/SE-LIPKE.pdf

"The analysis indicates that the aberrant behavior of the EVM schedule indicators, SV and SPI, is overcome by employing the Earned Schedule (ES) computation methods. The application of Earned Schedule provides a set of schedule indicators, which behave correctly over the entire period of project performance."

I believe no method is perfect because there is no way to know the future, you can only forecast based on a set of assumptions, even probabilistic methods are just predictions whose predictions shall be updated continuously. Any method that is based on a fixed baseline is wrong.

If you look at my previous posting you will see two independent Banana Curves, one based on a Baseline perhaps an already obsolete baseline or an updated baseline and another set based on actual predictions using the updated schedule. In order to know where you are in comparison to original plan you must show all three banana curves, at minimum. I would say you should compare preferably in two separate diagrams the progress of your job. One diagram displaying the original baseline banana curves versus actual baseline and another comparing latest updated schedule banana curve against current/revised contractual baseline. Everyone knows the data displayed by S Curves is somewhat limited, just a birds-eye view and as such shall be interpreted, as something subject to change, as something dynamic.

Good management practice requires for the documentation and analysis of changes, it starts after your updates, it do not ends there.