Changes to the Baseline must be accommodated so that the baseline accurately represents the current plan. Timely incorporation of changes is necessary so that managers are not tracking to obsolete plans and generating "false" variances.

Baseline Change Management Procedure

To keep baseline variance can be a monumental task in complex schedules. Is it worth it?

Extract from a Planning Planet discussion, August 2017:

http://www.planningplanet.com/forums/planning-scheduling-programming-discussion/613378/baseline-change-management-procedure-p6

• Department of Defense Earned Value Management Implementation Guide
• 2.5.2.4.3 Re-planning of the Remaining Baseline. - Maintaining a realistic PMB may occasionally require the re-planning of some, or all, of the remaining baseline.
• 2.5.2.5.1 During contract execution, the contractor may conclude that the budget and schedule, for performing the remaining work, is decidedly insufficient and no longer represents a realistic plan. At this point the contractor should prepare and submit a request to implement an Over Target Baselines (OTB) and/or Over Target Schedules (OTS).
• 2.2.3.7 Exclusions for Firm Fixed Price (FFP) Contract Type. The application of EVM on FFP contracts and agreements is discouraged, regardless of dollar value. - Same as DOD I find EVM unnecessary on Fixed Price contracts.
• Contrary to DOD I am not hooked on EVM and even find it flawed. - Earned Value Management as a tool for Project Control

However, an S-curve is a result of schedule logic. On the level of activities, projects tend to have many paths, some of them critical, other not. A tiny (if compared to total project duration) schedule variance of an activity can become a cause for substantial rearrangements in the project network, it can change critical path or order of activities, enforce a total reorganization of works. This cannot be captured by either Earned Value or Earned Schedule calculations - potential problems might be masked by compensating positive and negative schedule deviations.

• We simply issue a recovery schedule from last updated schedule that meets contractual requirements and set it as new schedule.
• This could be set as new baseline but erases prior variance details.
• EVM is flawed because it does not distinguish available float, it does not distinguish between critical and non-critical activities. Tracking variance without taking into account activity float makes no sense.
  • Earned Value Analysis does not distinguish between the works done on critical activities and activities with sufficient floats. A project could be late but EVA will not notice this problem if Earned Value exceeds Planned Value.
  • Earned Value Analysis motivates project managers to do expensive tasks first delaying cheaper activities that could have higher priorities.
  • Earned Value Analysis suggests forecasting future performance basing on past experience that may be wrong if resources that planned to be used in the future are not the same as in the past.
• DOD the father of the creature is against the use of EVM in fixed price contracts. For the DOD our few jobs have been fixed price. In such cases if required to compare against some baseline we use the recovery schedule until it needs to be revised again.
• The schedule contract milestones give us all we need to know if under/over target. It is simple; no esoteric procedures, no EVM jargon only a few at the field understand.
• We pay attention to critical activities and resources where EVM is of little or no help.

OTB-OTS-Guide-121205.pdf

3.5.6.2 Adjusting Variances: A key consideration in implementing an OTB is to determine what to do with the variances against the pre-OTB baseline. There are essentially five basic options. This is a far more detailed effort than these simple descriptions imply, as these adjustments have to be made at the detail level (control account or work package). (See Appendix B for examples.)

1)  Eliminate all Variances: …

2)  Eliminate the Schedule Variance (SV) Only: … After evaluating the cumulative information in the CPR/IPMR, the two PMs may agree that the cost variance represents meaningful performance measurement information that the CAMs should continue to focus on and that only the SV should be eliminated.

3)  Eliminate the Cost Variance (CV) Only: … If, after evaluating the cumulative performance measurement information, the two PMs agree that the schedule variance contains valid performance measurement information, the OTB can be implemented by eliminating only the CV

4)  Eliminate Selected Variances: … the two PMs may choose to implement an OTB for only that portion of the contract. In this case, all other variances and performance measurement elements would remain intact. The OTB reporting provisions would only apply to the items selected for OTB. The resulting TAB will be greater than CBB and will vary by which elements are reset.

5)  Retain All Variances: … the contractor and customer have agreed to retain all variances.

• As per this guide it seems like if there is no agreement on any of the options that keep some or all variance then Eliminate all Variance (always my choice) is the way to go.
• No need for esoteric procedures that complicate and delay the updating of baseline schedules.
• Your current contract conditions are your contract targets; these essentially are the contract milestones and budget amounts you keep updated on your schedule updates, everything else belongs to your means and methods.
• If you use P6 take a look at ORACLE P6 online documentation regarding baseline maintenance.
• P6 Professional Release 8.3 Documentation Library - Update a baseline
• To keep all variance requires a un-statused baseline.
• To keep a un-statused baseline when there are many/complex schedule revisions can become a monumental task.
• If easy to keep all variance then why so many people struggle with it?
• To keep a un-statused baseline when there are many/complex schedule revisions can become a monumental task.

As said before, a tiny (if compared to total project duration) schedule variance of an activity can become a cause for substantial rearrangements in the project network, it can change critical path or order of activities, enforce a total reorganization of works.

Schedule variance can be a change in plans, a change in scope, a change order, an act of god etc.  A substantial rearrangement in the project baseline might be needed.

Even a single activity might require substantial and complex rearrangement. Imagine Activity A (footing excavations) is schedule for first six weeks but it rained on week 1, then on week 2 an unforeseen condition stop it until evaluated, at week 12 a completely redesigned foundation was issued, from spread footing to pile foundation.

Changing the baseline for this single activity takes time if to be un-statused to keep all variance. It will have to be scheduled as intermittent, it will require additional activities with different resources when changed from a spread footing to a pile foundation activity(es). It  will require re-arrangement of budget amounts.

Un-statused re-baseline that keeps all variance is not as easy as a single click of the mouse.

Such a waste of time is not worth it, all is needed is an updated schedule that meets current contract conditions.

As a construction manager I have seen hundreds of times when substantial rearrangement of activities was necessary. 

Eliminate all variance is the option that makes sense on complex schedules.

Construction schedules are usually complex, even the schedule for a local hospital might require thousands of activities.  To apply the same poison to all schedules is bad medicine.

I suspect DOD is aware of this and therefore leaves it open the option to eliminate all variance.  For some inexplicable reason some schedulers insist on avoiding the option that makes sense on complex schedules, eliminate all variance.