Just as duration uncertainty may have a significant effect on schedule risk, cost uncertainty may have a material effect on cost risk. And the cost uncertainty affects the cost probability distribution more as the length of the schedule increases.
Deltek Acumen uses the Monte Carlo analysis to compute a probability distribution of the expected completion date of a respective project. The schedule analyst may consider duration uncertainty, schedule risk events, and schedule mitigation efforts.
Likewise when it comes to cost risk the analyst may consider cost uncertainty, cost risk events, and cost mitigation efforts. But in cost risk the analyst must take into account duration uncertainty and the probability distribution of schedule length in addition to the cost uncertainty.
This is important because labor and equipment costs increase as the length of the schedule extends. In Deltek Acumen the scheduler may generate a cost estimate with the schedule probability distribution overlaid onto this cost estimate.
This article demonstrates the effect of duration uncertainty on the schedule and the schedule cost risk.
We have in Figure 1 our demonstration Microsoft Project schedule.
This is a simple cost loaded piping repair and improvement project. In Figure 2 we have our demonstration project imported into Deltek Acumen.
Figure 3 displays the results of the Acumen schedule quality diagnostic.
Our schedule quality diagnostic is flagging two activities for insufficient detail. These activities have short duration, but in relation to the brief schedule duration they may be long. At this point we would recommend shortening these two activities to provide more detail in the schedule on these two efforts.
In our demonstration, however, we simply exclude these two activities from the schedule quality analysis. We also exclude the two activities that are missing logic, as these activities represent the start and finish of the schedule. Figure 4 displays the resulting schedule quality analysis.Figure 4
We achieved an overall schedule quality analysis score of 95%, which is good; we are ready to perform a risk analysis of our schedule. We save the workbook, Figure 5.
In Figure 6 we choose the S3 // Risk tab and create cost estimate | generate from schedule | generate without overlay.
Without the schedule overlay the schedule probability distribution is not considered in the cost probability distribution analysis. In Figure 7 we set the cost uncertainty to very aggressive which corresponds to min, most-likely, and max uncertainty factors of 100%, 100%, and 150%.
We would therefore expect the Monte Carlo cost analysis to provide an on budget or worse schedule cost. The simulation scenario is set to uncertainty only (no risk events), Figure 7. Figure 8 displays the histogram and statistical results.
As expected we barely missed the budget or worse. The best case came in at $30.3K and worst case $39.6K. The deterministic cost estimate is $29.2K. In Figure 9 we choose the S3 // Risk tab and create cost estimate | generate from schedule | generate with overlay.
Note, the duration uncertainty is set to zero. We repeat our cost uncertainty setting of very aggressive and again run the Monte Carlo analysis for uncertainty only, Figure 10.
We would expect the Monte Carlo cost analysis, again, to provide an on budget or worse schedule cost. The results are exactly the same as before without the schedule overlay, Figure 11.
The schedule overlay therefore had no effect on our cost probability distribution. But this is because the duration uncertainty is currently none, which means Acumen overlays the deterministic schedule onto the cost output. Now let’s see what happens when we set the duration uncertainty to very aggressive, Figure 12.
In Figure 12 we return to the pipe repair schedule and adjust the duration uncertainty to very aggressive. Let’s run the Monte Carlo analysis one more time, and we would expect the worst case scenario to have an increased cost. Figure 13 displays the histogram and statistical results.
This increased the worst case cost is due to the schedule overlay, and very aggressive duration uncertainty. Again, labor and equipment costs increase when the schedule extends. Our schedule with very aggressive duration uncertainty would barley meet our schedule or worse.
This very aggressive duration uncertainty extends the length of the schedule, which in turn increases cost. This is because the labor and equipment cost continue through the lengthened schedule.
Both Deltek Acumen schedule risk and cost risk consider uncertainty (duration or cost), risk events (schedule or cost), and mitigation efforts (schedule or cost). What makes the cost probability distribution analysis particularly unique and insightful is that it may include the impact of the schedule probability distribution. In this way the cost probability distribution may account for the impact of schedule duration uncertainty.
As explained, labor and equipment costs increase as the schedule extends. And we would expect a schedule and aggressive duration uncertainty to increase cost in the probability cost distribution.