This one almost slipped by me. Back in October, Stuart wrote . . . The organisation may select one of a number of methods to actually implement resource levelling (the nitty-gritty work), such as Critical Chain scheduling . . .
I'd like to add a bit of clarification on resource levelling and Critical Chain as it applies to multi-project environments.
Within individual projects, it is absolutely necessary to account for resource dependencies if you are going to make rational promises, hence the need for resource leveling in the Critical Chain approach. As a matter of fact, that levelling occurs prior to the determination of what is "critical."
However, when we move to the system that is a multi-project environment, it becomes an onerous (and usless) effort to try to level resources across projects. Onerous, because everytime you add a new project into the mix, there is the possibility of having to revisit previous schedules and commitments. Useless, because after you spend all that time and effort levelling everything and everyone, the moment we move to execution, and reality deviates from the schedule, all the nicely lined up resources will be unlevelled.
The application of Critical Chain to multi-project systems involves identifying one or two heavily used, commonly used resources, and launching the projects based on levelling only them. If they are commonly used across projects and heavily used relative to other resources, then the other resources will have ample time (over the long haul) to get what they need to do done.
If, as a result, there contention across projects for the use of a resource a a particular point of time, that's what buffers are for. Project and feeding buffers will absorb the possible delay of attention. And Buffer Management will provide guidance on which of those contending tasks is the best use of the resource's time.
So yes, there's a bit of [prudent] "nitty-gritty" work in leveling resources within projects, but not across projects when using Critical Chain-based multi-project management.