Files
Abstract
A methodology is developed and tested for sizing convergence, resource, and completion buffers for networks using Critical Chain Project Management (CCPM). The testing was done using computer simulation. Activities along the critical chain were started upon completion of the preceding activity, often called the roadrunner method. The normal distribution was used to model all activity durations in the models. Activities along non-critical paths were started upon completion of both the previous activity and the scheduled start date. This second method of starting activities was found to modify the normally distributed activity times on the non- critical path to where they behaved like truncated normal activity times. As a result of the behavior of the non-critical paths, the path durations did not behave as expected in the literature.