Critical Chain vs. Critical Path

[pmkb]

Is the management of critical chain buffers any different than managing the float of critical path tasks?

[nileshbiniwale]

According to me Critical chain method would be to identify Bottlenecks on Critical path and monitor and manage this bottleneck resources on critical path.

[pmkb]

Hi nileshbiniwale, welcome to the forum!

Can you elaborate a little bit more on the difference in managing bottlenecks with Critical Chain versus Critical Path?

[BenRam]

Critical chain is the longest set of tasks across the system once both the task effort & the resource balancing is completed. Critical path looks only at the longest set of tasks & thus ignores multiple tasks that need to be completed by the same resource at the same time (and thus can't be done in the time estimated).

Given the critical chain, the theory goes on to say that we need to review the estimates. With any task there is the work that is required for the task & the overhead (preparation, interruptions, student syndrome, previous task not complete, etc). Critical chain says - build your plan & estimates as normal, balance the system for the chain (ie tasks & balanced resources) and then remove from every task a % (often 50%) which accounts for overhead. For the tasks on the critical chain, take that overhead and put it at the very end of the chain to account for the unknowns (estimates off, task harder then expected, small design changes) and manage the people so that the interruptions and other noise is filtered out (when a person is working on a critical chain task they are left alone to complete that task).

The result is a chain through out the system with a buffer that was derived from the actual tasks themselves. The outcome is excellent focus - focus only on the final buffer & whether the chain is decreasing at a rate relative to buffer usage (if it is too high buffer will be used before work is complete). From the buffer it is obvious which tasks are actually taking long & eating into the buffer. So from an extremely complex network you can quickly zero into the task that needs help.

So in answer to the question of the difference between buffers of the methods.
Critical path buffers are created through some guesses, critical chain buffers are integral to the system and are managed to keep the project on track for delivering at the planned for time.

[pmkb]

Hi BenRam, thanks for your response.

Critical chain is the longest set of tasks across the system once both the task effort & the resource balancing is completed. Critical path looks only at the longest set of tasks & thus ignores multiple tasks that need to be completed by the same resource at the same time (and thus can't be done in the time estimated).

Resource constraints can be managed with Critical Path either via soft logic or resource leveling. I really don't see a difference here.

With any task there is the work that is required for the task & the overhead (preparation, interruptions, student syndrome, previous task not complete, etc).

Preparation is the only item from your list that should be built into durations in a critical path schedule (or buffers in a critical chain schedule) IMO. The rest is fat and subject to Parkinson's Law ("Work expands to fill the time available for its completion."). They can be managed by tracking float.

Critical path buffers are created through some guesses, ...

I don't follow you there.

[BenRam]

Hi BenRam, thanks for your response.

Resource constraints can be managed with Critical Path either via soft logic or resource leveling. I really don't see a difference here..

This is similar to resource leveling but it is actually effective across multiple projects. I haven't seen any other system which will allows resource leveling across multiple projects. Also, what is key here is not all resources are leveled, just the constraining resources & by leveling them the rest of the system properly balances itself...

Preparation is the only item from your list that should be built into durations in a critical path schedule (or buffers in a critical chain schedule) IMO. The rest is fat and subject to Parkinson's Law ("Work expands to fill the time available for its completion."). They can be managed by tracking float.

Float is a concept that was thrown out with Critical Chain. The idea was that in a multiproject environment project managers were so busy managing every single task that the critical tasks didn't get the attention they deserved and the less critical tasks got too much attention (critical meaning the tasks that impact the end date). So instead of float per task there is one buffer at the end of the project that effectively reflects the state of the entire project.

Check out this article for indepth details about critical chain.
[url]http://www.tocc.com/Critical_Chain.htm[/url]

The interesting thing about this is having practiced and observed project management for a very long time, CCPM is the first method that I've seen real results where people deliver early & not by adding excessive padding (ie where they reduced program time of a previously run program from 300 to close to 100 days where all other methods of lean and six sigma only helped in the order of a dozen days).

[jblack]

For more information on Critical Chain or Critical Path, try visiting this link and search for "critical"
[url]http://www.pmconnection.com/modules.php?name=Web_Links[/url]

[SemiMike]

I fully support BenRam's comments.
And yes, the pmkb people are also right in theory, but unless the top management team embraces the critical chain concepts that force resource leveling (no dreams of multi-tasking as the solution), they miss a great opportunity.

I have seen MOST projects take too long due to mgmt style "just let the informal network mgmt process take care of the critical resource availability." Ops mgrs do not support "dedicated" PM team members, they want to keep their key people in operatoins. Too many projects started with too much demand on one or two critical resources, resulting in need to add swagge durations and inaction until last minute, due to low pressure to perform on EACH project until they all need same resource at same time, triggering RE PRIORITIZATION as the solution. We can avoid "dedicated" people as long as ops mgmt clearly signs up to supply the key resource when its needed...and that is enforced by top mgmt in support of projects.

Games people play! Passive-aggressive ops mgmt people often stonewall projects in favor of end-of-month shipments, for example.

Software seems to support only the old-school textbook gantt charts, so Concerto is suggested for CCPM efforts. Any way to make OTHER software "do the right things?"

[michaelmz]

I found an excellent article discussing this very same topic. You will find that it is a very interesting read: [url]http://www.tocinternational.com/pdf/Critical%20Path%20vs.%20Critical%20Chain.pdf[/url]

[Scottish Ranger]

I have een looking for case studies on CCPM usage and so far can only see examples in engineering projects...can anyone point mein t direction where this tool has been used in Business Transformation programs....and/or the IT environment...thanks.