Results 1 to 7 of 7

Thread: Model Calibration

  1. #1

    Model Calibration

    I have my water set up completely and have run the required scenarios. I have had my field crew perform random flow tests throughout the system.

    How do I calibrate my model based upon the field conditions?

  2. #2
    Member

    Innovyze Employee



    Join Date
    Mar 2013
    Posts
    72
    InfoWater's extension module of Calibrator can do calibration. Calibrator has three modes - steady state, EPS, and fireflow.

  3. #3
    Member

    Innovyze Employee



    Join Date
    Mar 2013
    Posts
    72
    Left a voice message last week. Are there any further questions about InfoWater Calibrator?

  4. #4
    Here are some tips if you don't want to use the Calibrator extension:

    1. You should have one scenario for each flow test
    2. You will need to set up each scenario with the boundary conditions that existed during the test (tank levels, pumps on/off, reservoir head, PRV settings, etc)
    3. Once all boundary conditions are established and the flow test demand is placed on the test junction, how much is the pressure off?

  5. #5

    Join Date
    May 2013
    Posts
    22
    The default lower bound for pipe groups in the Calibrator is 50. Does anyone use a lower number for the minimum? How low do you go?

    I've got two or three pipe groups that calibrated to the lower bound. Should I lower the minimum roughness for those groups?

  6. #6

    Join Date
    Mar 2013
    Posts
    5
    I probably wouldn't lower the lower bound; I'd rather focus on looking for unknown field conditions like a valve closed or a pipeline that's been abandoned, pipeline diameters that are smaller than on record (you may have already exhausted these options), or offsets related to boundary conditions such as controls or instrumentation. You can also try to divide up the groups that are giving you trouble.

  7. #7
    Forum Moderator

    Innovyze Employee



    Innovyze Employee



    Join Date
    May 2015
    Posts
    405
    JFusilloPE,

    One thing to keep in mind if you use the Calibration tool, is that adjusting roughness factors is not always applicable for all pipes in all water systems.

    Typically only pipes that have unlined cast iron or unlined steel that can corrode internally will have a C factor that will change over time. Unlined cast iron and steel pipes were widely used up until roughly the late 1960's when they were phased out. Systems with a high number of these pipes would be expected to have C-factors that change and these are ideally suited for using the Innovyze Calibrator tool that adjusted c-factors to adjust headloss and better fit model and field data.

    However, for most pipes installed after the late 1960's, they were either cement mortar lined or made of plastic. Neither material will experience heavy interior corrosion like unlined cast iron and would not be expected to have a c-factor that was much different form its initial book published c-factor values. Thus typical plastic pipe such as PVC or HDPE, and other common pipe types such as Ductile Iron, Asbestos Cement (AS), Concrete Cylinder (CCP), or lined Steel pipes installed in the last 20-40 years would not be expected to corrode internally due to the cement mortar lining of these pipes or because they are plastic. As such it would not be justified to adjust pipeline C-factors for these pipelines and one should be careful not to allow more than roughly a 10 point difference from the typical book published average c-factor values for pipes that are non-corroding.

    I have personally seen many pipes removed from service in the field that were 30-40 years old that looked less than 5 years old on the internal surfaces than can attest to what I have noticed doing many calibrations over the years. It has been rare to adjust c-factors except for unlined cast iron and unlined steel pipes. This has also been the experience of many well known modelers I have worked with or talked to in the last 15 years.

    So keep these factors in mind when selecting pipe groupings when you use the calibrator tool.

    Lastly, any c-factor adjustments made should be valid for all flow scenarios. It is not reasonable to expect that the c-factor would be one value in one scenario, and a different one in another. It is the same pipe, and so different tests will allow validation of the c-factors used in different areas of the system. If the c-factor change helps one scenario and hurts in another, the additional headloss that suggested a lower c-factor, was not likely caused by the pipe itself. The problem is something else and needs to be investigated further. This is a common mistake I have seen made before.

    The worst C-factor I have ever personally used in a calibration of a cast iron system was around 55-60 and were for pipes 6 inches and smaller. These were known to be very heavily turburculated. In most models with CI pipe having moderate tuberculation, the typical lower end has been in the 80-90 range. But any lowering of C-factors below 90-100 should be checked to make sure the pipes are known to be either moderately or heavily corroded, and that the headloss is not caused by closed valves in the field.

    These guidelines should be considered when using the calibration tool and when creating pipe groupings that the software will use to make pipeline c-factor adjustments. I always tried to think "is this change justified?" and "does this change make sense?" before making any change in calibration as one can adjust a parameter to better fit data, but if it is not justified the change will cause model results to be off in other scenarios. So any change made should be justified and defendable if questioned.

    Enclosed is a table that has been used by many engineers in the last 10 years and has been found to work very well in over 50+ calibrations. We typically used the average values to start and vary rarely had any observed difference that justified c-factor adjustments for non corroding pipes.


    Suggested C-factor table (click for larger view)
    C-factor table.jpg

    Hope this provides additional assistance.

    Innovyze Support
    Patrick Moore

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •