Critical Chain Project Management
In the 1990s, Goldratt applied the Theory of Constraints to project management – and Critical Chain Project Management (CCPM) was born (Goldratt, Eliyahu M.: Critical Chain, The North River Press, 1997). In the "production system" of a project, CCPM ensures the creation of steady production while considering the resources. The production system is the value creation through the project – represented by the project plan. There, the sequence of activities, their dependencies and the use of resources are defined.
Critical Chain Project Management should not be confused with the critical path (CP). In the latter, the "critical" processes are searched for, i.e., the steps whose delay would endanger the project timing. See the figure below:
The red activities are on the critical path - a delay there will endanger the end date of the entire project; the blue activity, on the other hand, has more than 7 weeks of time buffer.
CCPM not only addresses the predicted duration of tasks, but recognizes that uncertainties and risks affect the availability of people and equipment. Goldratt therefore proposes several countermeasures to achieve a steady flow of work through the project system.
Multi-tasking, i.e., jumping between different activities, is inefficient. After all, time is lost by mentally switching to the next topic - experts claim this number is around 40%. So, one should not start too many projects at the same time in the first place.
Remember the drum-buffer rope above? Only as much new material (projects) should be released to the system as the bottleneck (humans) can handle. In the figure below, this becomes clear through a "Tetris" screenshot of a resource histogram. Paula seems to be tied up in many – probably too many – projects at once.
CCPM, just like the TOC, looks for the bottleneck. Bottleneck resources are the people or equipment that are overloaded, which projects must constantly wait on. However, unlike the CP method, resources determine the sequence of operations whose delay endangers the end of the project. This is also possible across chains.
All activities in which bottleneck resources are planned must therefore be "buffered" or, accounted for, during project planning. How?
a) Allow enough time around the use of these resources to anticipate that they will not be available in time. Goldratt refers to these buffers as "feeding buffers"
b) Seek replacement personnel or additional personnel, even if they are not equally efficient. In the following figure, it is clear that "Peter" determines the critical chain here.
Goldratt assumes that buffers are factored into every activity. This is bad for two reasons, he says: if there was too much buffer, the result is an unnecessary wait for the "official end" of the activity - in fact, it could have been started earlier. If there was too little buffer, the project schedule gets messed up.
Goldratt's solution: instead of estimating buffers into each task, all buffers are added in their entirety to the end of the project. As soon as a task is completed, the next one is immediately started. If a task takes longer, the total buffer ("project buffer") is used up.
Intermediate milestones for measuring progress are consequently not used. In contrast to the "earned value analysis", progress is not measured in terms of completed work packages, but along the time spent. And it is precisely this that is compared to the time buffer consumed. So it's not primarily about the effort spent, but about the progress over time. Excuse me? That will lead to effort overruns! Maybe - yes, these are not measured any more than effort underruns - it is about the delivery of the result. Have you heard that before?
Sounds kind of agile, doesn't it?
CCPM encourages project teams to be open and to communicate both within and across teams. Bottlenecks should be looked for and solutions worked out together, without waiting for management intervention.