Pipe ID Difference with respect to AGA 9 Requirements (Ultrasonic Flow meters)


We are facing one technical issue regarding compliance of AGA 9.
We are trying to get a 24" USM flow calibrated from an external calibration lab. Our existing pipe in the plant is 24" sch60. But that lab does not have the same upstream and downstream pipe spools. They have 24" sch40. So, there will be a 2.5% pipe id difference between calibration and actual field installation. As per AGA-9 Clause#7.2.3, such differences should be within 1%. 
Is there any way we can work around this problem? I understand that there will be impact on the calibration uncertainty due to this difference. What do you say?

3 Replies

  • AGA 9 section 7.2.3 (3rd edition from July 2017) is in regards to Field Verification and states:

    7.2 Field Verification
    The manufacturer shall provide a field verification test procedure to the operator that will allow the USM
    to be functionally tested to ensure that the meter is operating properly. These procedures may include a
    combination of a zero-flow verification test, speed-of-sound measurement analysis, individual path
    measurement analysis, internal inspection, dimensional verification and other mechanical or electrical tests.
    1 . The field verification of a USM consists of comparing current meter diagnostic data against initial
    diagnostic baseline values or to a prior-known good value to identify possible changes in the USM
    performance. It is recommended that field verifications be conducted following commissioning. The
    frequency of verifications should be guided by the meter data history, volume, operating conditions,
    and/or operator policy.
    2. Evaluation of any changes to these diagnostic indicators and their potential cause may guide the
    operator in determining any impact on the meter performance and the need for any repair, flow
    performance test (in-situ or laboratory), adjustment to maintenance interval or design improvement.
    3. USMs provide serial data that can be collected through end-user polling systems. End-user custom
    algorithms and commercial data-analysis packages can be utilized to provide real-time continuous
    evaluation of the USM’ s performance that can help predict maintenance timing.
    Note: See Appendix E (Informative) for a field verification checklist sample.

    Based on your remark, I assume that you are in fact an end user of this USM and not the manufacturer that has to supply calibration data to a customer.

    I think it all depends on who or which entity requires compliance with AGA 9 and what you are measuring.
    So maybe a couple of questions to clarify:
    1. What media flow are you measuring with the USM? (Is it really just natural gas?)
    2. Is this a custody flow measurement? (If yes, does it supply one or multiple customers?)
    3. Is this a measurement due to regulatory or legal requirements?

    Also, how old is the existing pipe in your plant and what is the internal condition of that pipe?
    From my experience with USM, the accuracy is highly depended on how well you can replicate the installation and mounting conditions (even if you can assure that the pipe is "exactly" the same and you follow all AGA 9 requirements).

    Is it possible for you to supply that testing lab with a section of pipe (24" sch60) that is long enough to perform the calibration? If it is really such an important measurement, then you might have a section of that exact same pipe laying around, since it is part of the USM measurement installation.

    In the end, your company will have to decide if AGA 9 has to be followed exactly for this measurement or not. And if it is a must, then there are really only two options for you:
    1. Find a new calibration lab, that can perform the calibration.
    2. Work with the existing lab and see if you can supplythe pipe.

    There might be a third option:
    Is there any possible way to perform a calibration in the field? (I assume not, but maybe there is a different calibration company that can bring their calibration equipment to your site?)

  • The actual pipe step change is 1.17% and in a larger meter such as yours the effect will be none minimal if anything at all. When we put this in the report it was a catch all for all meter sizes so from 4" to whatever, all the work that we have done shows that a step change of up to 3% has no effect on the uncertainty so I would be too concerned. One thing to remember is that AGA 9 is a report and not a specification so its a guideline where we expect common sense will prevail!!
  • In reply to martin schlebach:

    Thanks a lot Martin. Im surprised that why 3% is not mentioned in the AGA report, why it has been made so stiff. Also, will 16" meter will also be OK upto 3% of step change in ID?