none
Work Variance Incorrect in Project 2010 RRS feed

  • Question

  • I have 2 projects with the same start date in the future.  No work has been done on them.  Each has the same following values, but the resulting work variance at the project summary (0) is different.  The formula as I understand it : Work Variance = Baseline Work (Actual Work - Remaining Work).  Why the difference?

    Work Variance (project 1) = zero; Work Variance (project 2) = 485

    Baseline Work = (NULL)

    Actual Work = zero

    Remaining work = 485

    We are running Project Server 2010, SP2, Patched to April 2014.

    Friday, November 7, 2014 5:53 PM

All replies

  • Two questions:

    With those values, having a WV of 485 is correct.

    1) Is the Calculation set to Off in the project that shows 0 hours? Force caluclate by hitting F9 and see if it changes anything.

    2) Are you really looking at the project summary task? some times PMs create a task as the summary task for the project which is not task 0.


    Cheers,

    Prasanna Adavi, Project MVP

    Blog:   Podcast:    Twitter:    LinkedIn:   

    Friday, November 7, 2014 6:08 PM
    Moderator
  • In addition to Prasanna's advice, I'd mention that the exact formula for the Work Variance is [Work]-[Baseline Work].

    If the issue persists, I'll be glad to have a look to you file if you can send them at guillaumeDOTrouyreATgcosiDOTcom (replace cap letters).


    Hope this helps,


    Guillaume Rouyre, MBA, MVP, P-Seller |

    Friday, November 7, 2014 7:30 PM
    Moderator
  • Prasanna.

    Both the calculation and the project summary task are set properly. 

    I discovered that when the PM attempted to build a Resource Plan, using the "resource plan" for the "calculate from" option, it would not allow a publish.  The PM had to first open the newly created (template) project, publish, and check it in.  In these cases, the Work Variance has the 485 hours.  In those that didn't require this action, had 0 Work Variance.  Therefore, why is it necessary to "publish" some new projects before a Resource Plan is built and in most others it is not required?

    Thanks.

    Friday, November 7, 2014 8:28 PM
  • Indeed I noticed the same behavior with PS2010, I guess this is by design that the project plan has to be published once before being able to publish the resource plan. Then if you have work in the resource plan and the capacity calculation is set from the resource plan, it is normal to have a work variance equal to the resource plan work evenif you don't have any work in the project plan.

    Hope this helps,


    Guillaume Rouyre, MBA, MVP, P-Seller |

    Friday, November 7, 2014 8:33 PM
    Moderator
  • Guillaume,

    Interesting you state the formula that way.  I searched "work variance" in the forum before and found this formula from Ben Howard: work variance = baseline work - (actual work + remaining work).  Your formula gives the accurate value.

    See my reply to Prasanna for my reasoning as to why the values are different on my report.

    Thank you.

    Friday, November 7, 2014 8:41 PM
  • Guillaume,

    Most new projects didn't seem to require the publish-first requirement, but a few did.  Why the inconsistency?

    Thanks,

    Friday, November 7, 2014 8:54 PM
  • I saw your reply indeed and I responded accordingly. I must admit that I experienced also this inconstency about the resource plan publication in PS2010 but have no clue for it.

    About the work variance, if you refer to the office reference here, you'll find the formula I gave. But assuming that work = actual + remaining, both formula are mathematically equivalent. If actual is null, then work = remaining.


    Hope this helps,


    Guillaume Rouyre, MBA, MVP, P-Seller |

    Friday, November 7, 2014 9:08 PM
    Moderator
  • I still cannot understand the value of Work Variance.  See my sample below, which came straight from the reporting DB.  Am I grabbing the correct variance column?  If not, where is the correct one?

    ProjectWork ProjectBaseline0Work ProjectActualWork ProjectRemainingWork ProjectWorkVariance
    1714.349983 916.8 832.5 881.849983 587.25

    Friday, November 7, 2014 9:11 PM
  • If the resource plan has been used and populated with some work, then you also need to grab the resource plan work. Note that a thrid type of work called "combined work" calculated the work based on the capacity parameter of the resource plan (resource plan, project plan or project plan until). If you use the resource plan with the capacity parameter then you should look for the combined work.

    Hope this helps,


    Guillaume Rouyre, MBA, MVP, P-Seller |

    Friday, November 7, 2014 9:18 PM
    Moderator
  • The Resource Plan Work was not populated in this case.  When I open the project and view "Work Variance" it is stated as 797, not 587 as in the above listing.  Does Project Pro grab the value from some place other than the UserView for Project in the Reporting DB?  If so, where?

    Thanks

    Friday, November 7, 2014 9:22 PM
  • Dumb question: has the project been correctly published?

    Hope this helps,


    Guillaume Rouyre, MBA, MVP, P-Seller |

    Friday, November 7, 2014 10:47 PM
    Moderator
  • Guillaume,

    I am convinced that the work variance stored in the Reporting Database, Project UserView, is incorrect.  I have checked several projects that have been recently published in Project Pro and seen tasks that did not have a work variance although it should have been there.  The Project Summary was correct.  I checked the "schedule" in PWA and it was incorrect at the project summary level (same project).  Given this fact, I am going to do my own calculation of work variance.  I think this is a bug.  Do you agree?

    Monday, November 10, 2014 6:55 PM
  • If I were you, I'd try to query the draft and publish DB to check for potential synchronisation discrepencies between DB.

    You might have Reporting DB sync issues and need to refresh the Reporting DB.


    Hope this helps,


    Guillaume Rouyre, MBA, MVP, P-Seller |

    Monday, November 10, 2014 7:14 PM
    Moderator
  • Guillaume,  I have decided to calculate the work variance within the SQL query to avoid this problem. 

    Thanks.

    Wednesday, November 12, 2014 3:51 PM