Introduced in 1997, the critical chain schedule (CCS) is a relatively new tool to the world of project managers.18A CCS is a network diagram that strives for accomplishment of drastically faster and more reliable schedules (see Figure 6.7). It uses several unique
A:16d Activities
B:16d
D:8d
G:16d
H:16d C:8d
I:8d F:16d
J:8d
D:8d E:16d
Critical Chain (CC) Activity Non-CC Activity
CC Feeding Buffer (CCFB) Project Buffer (PB) Key:
A B C D E F CCFB1
G H I J CCFB2
K PB
Days (d) 16 32 48 64 80 96 112 128
Figure 6.7: Example Critical Chain Schedule
THE CRITICAL CHAIN SCHEDULE 169
approaches. First, the CCS focuses on the critical chain, the longest path of dependent events that prevents the project from completing in a shorter time. Unlike the critical path, the critical chain never changes. Second, its activity durations are estimates with 50 percent probability. For this reason, they are significantly shorter than those used in other scheduling tools, which are often with 95 percent probability. Third, in contrast to the critical path, the critical chain is defined by the resource dependencies.
Fourth, buffers are built to protect the critical chain during the course of project implementation.
Developing a Critical Chain Schedule
Because of the inherent challenge of faster schedules that it seeks to make possible, a CCS’s quality is even more dependent on the depth and degree of definition of inputs than other schedule development tools.
While the scope, responsibilities, and schedule management system will provide information about the what, and the who elements, as well as how to schedule project activities faster, the real emphasis is on the CCS’s requirement for dedicated team resources, meaning that team members work full-time on one project only. Because of this, the logic goes that members of the dedicated project teams are more productive than members who are shared by multiple project teams. A reason for this is that the switching time cost created by one’s work in multiple projects is eliminated as discussed in the previous section. Although this is generally true, there are some exceptions. A study found that when a team member who is focused on a single project is assigned a second one, productivity often increases a bit because the team member no longer has to wait for the activities of other members working on the initial project (see Figure 6.8).
Rather, the team member can shift his or her attention to the second project.19When a third, fourth, and fifth project is added, however, the productivity plummets rapidly, and the team member becomes a bottleneck of all projects he or she is involved in. This is why the CCS approach insists on using dedicated teams.
10 20 30 40 50 60 70 80 90 100
1 2 3 4 5
Efficiency
Number of projects working on Figure 6.8: Productivity of Multiproject Team Members
Determine the Level of Detail and Identify Activities
Thenumberof activities in a CCS is closely related with thesizeof activities. Therefore, choosing to have approximately 100 or 300 or 500 activities will help determine how large individual activities will be. To illustrate this point, consider one company’s golden rule: Large projects, including between 5,000 and 10,000 person-hours, will have around 500 project activities that range in duration from two to four calendar weeks. Not only does this clearly tell everyone that neither 180-activity schedule nor 15-week activities are tolerated, but it also spells out the company’s belief about the right level of detail.
Given the complexity and size of the project, such level of detail provides sufficient information to manage the project without making it unnecessarily burdensome and time demanding.
Once the decision has been made about the level of detail, these actions should be taken:
■ Brainstorm and identify activities that are necessary to complete in order to finish the project. As with other scheduling tools, resort to the WBS for the activity identifi- cation. In this process, disregard how large the activities are; rather, ensure that all necessary activities are on the list.
■ Go back to the chosen level of detail. If the list of activities is below the intended num- ber of activities, continue breaking down larger activities; if it is over the target number, combine similar activities to reach the desired goal.
Sequence Activities
Sequencing means arranging the activities in a logical workflow and identifying the dependencies between the various activities. Deep knowledge of workflow is a prereq- uisite here. The principle of sequencing is to know that a preceding activity produces outputs that become inputs to an activity that follows. If the obtained diagram fails to observe the principle, it is likely that we are missing the logic of the project work, result- ing in rework and delays in project execution.
Assign Resources and Estimate Activity Durations
Since people and material resources are needed to complete project activities, they dic- tate activity durations. Therefore, a natural starting point for estimating the durations is: “What resources do I need to successfully complete this activity?” The answer should provide the names of resources and work time for each to complete the activity—for example, 100 hours of work from a programmer. The key point here is that CCS uses a unique technique of activity duration estimating that does not allow for contingencies (see “When Estimating Durations, No Contingency Safety Allowed”). Considering that the critical chain approach requires dedicated teams, and knowing the company’s work calendar—5 days a week, 10 hours a day—those 100 hours turn into 10 workdays or 14 calendar days. Naturally, the estimation of each activity should undergo this process.
THE CRITICAL CHAIN SCHEDULE 171
Identify the Critical Chain
The critical chain is the longest path in the network diagram, considering activity and resource dependencies. Stated a different way, it is the sequence of dependent events that keeps the project from completing in a shorter time.
Add the Resource Buffers
Critical chain schedules always consider the resource constraints and include the resource dependencies that define the overall longest path. Practically, this is handled by adding resource buffers to protect the critical chain from unavailability of resources.
Resource buffers are added to the critical chain only, do not take any time in the critical chain, and are termedresource flags. For example, any time a new resource will be used in a critical chain activity, we will add a resource buffer. This signals to the project manager and resource provider when to make the resource available to work on a critical chain activity. Since timely resource availability is critical to the rapid execution that CCS advocates, some companies use incentives to reward behavior of early delivery of activity outputs and standby time of resources.20
Create a Project Buffer
Unlike other schedule development tools, the CCS uses a novel concept of the project buffer. Its purpose is to protect the project completion date by aggregating risk contin- gency time in the form of the project buffer at the end of the critical chain (for manage- ment of the buffer, see the Buffer Chart section in Chapter 12). There are several methods to determine the buffer duration. One of them is to divide the duration of the critical chain by two (called the “50 percent buffer sizing rule”). The buffer is used to absorb uncertainty or disruptions that may occur on the critical chain and has no work assigned to it (see Figure 6.7).
Create Feeding Buffers
Protecting the critical chain with the project buffer is not enough. There is a signifi- cant risk that activities that are not on the critical chain but feed into it may slip to the point of pushing out the critical chain. To protect the critical chain from the risk, we can aggregate contingency time at all points where noncritical activities feed into the critical chain (see Figure 6.7). These contingency times are termedcritical chain feeding buffers.
During the project implementation, these buffers will be used to absorb uncertainty or disruptions that may occur in noncritical chain activities. To determine these buffers, use one-half of the sum of the activity durations in the chain of activities preceding the buffer. No work is assigned to the buffers.
When Estimating Durations, No Contingency Safety Allowed
Most project managers tend to include contingency time into each activity esti- mate without specifying it. The reason is simple: add the safety time. The CCS strives to eliminate the safety. Here is what it means. Figure 6.9 shows a typical distribution of activity time performance. The solid line (the left ordinate) tells us the incremental probability of a given activity duration time on thex-axis.
The dotted line indicates the cumulative probability (the right ordinate) that the activity will be finished in a time less than or equal to the activity duration time on thex-axis.21
Incremental Probability
0.000 0.004 0.008 0.012 0.016
0 0.020
0.2 0.1 0.0 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.00
Contingency
Activity Duration (Days)
Cumulative Probability
Inc Prob Cum Prob
100 80
60 40
20
Figure 6.9: A Typical Distribution of Activity Time Performance
When project managers include contingency time within an activity, they really go for 95 percent probable estimate (cumulative probability). As the cumulative curve shows in the figure, it is duration equal to or less than 50 days. Without the contingency time, the duration is less or equal to 20 days.
That is a 50 percent probable estimate. The difference between the 95 percent probable estimate and the 50 percent probable estimate is contingency time, 30 days in this example. To avoid excessive activity duration, and speed up the schedule, the CCS eliminates the contingency time, using only 50 percent probable estimates.
Using the Critical Chain Schedule
The most appropriate application of the CCS is for a dedicated project team seeking a significant reduction of the project cycle time in a company with an outstanding performance culture. The only job of this team is their project. Equipped with all