Emerson Exchange 365

• Not Answered

Is there any possibility that coriolis meter has a biased mass flow rate?

Hello. 

I recently replaced a flowmeter and transmitter in a facility for measuring the mass flow rate and the density of liquid oxygen.

I plotted a figure which shows the relationship between the measured mass flow rate and the calculated one from the flow coefficient of an object at the end of the facility. (below graph)

As you can see, at first, the flow meter showed coincident data(red circles, o) with the old one(black circles, o).

But, recently the flowmeter seems to show somewhat biased values(purple circles, o) from the previous data.

I wonder if there is any possible reason of getting biased mass flow rates from the coriolis meter and then why.

I also want to know how to check the integrity of the sensor.

Thank you.

Here are the flow meter and transmitter specs.

Model: CMF100HB33NRAIEZZX_36020 CMF100H MICRO MOTION ELITE CORIOLIS METER, 1 INCH (DN25), NICKEL ALLOY C22 B33 Process Connections: 1-inch CL900/1500 ASME B16.5 F304/F304L Lap joint flange N06022 stub N Case Options: Standard pressure containment R Electronics Interface: 9-wire polyurethane-painted aluminum J-box A Conduit Connections: No gland I Approvals: IECEx Zone 1 E Language: English installation manual Z Calibration Options: 0.10% mass flow and 0.0005 g/cc density calibration Z Measurement Application Software: No measurement application software X Factory Options: ETO Product 36020 ETO Number: CRYO TEMP CYCLING MODIFICATIONS WITH J-BOX TO -240C(AMBIENT AND FLUID) ON A CMF100H Model 2700R12BBIEZZZPK 2700 MICRO MOTION 2700 FIELD AND INTEGRAL-MOUNT MULTIVARIABLE FLOW AND DENSITY TRANSMITTER R Mounting: 4-wire Remote Mount Transmitter (Polyurethane-Painted Aluminum Housing) includes10 ft. [3m] 4 wire Shielded PVC Cable 1 Power: 18 to 100 VDC and 85 to 265 VAC; Self Switching 2 Display: Backlit Dual Line Display for CSA, UL, and IIB + H2 ATEX IECEx and NEPSI Ratings B Output: One mA; Two Configurable IO Channels - Default Configuration of 2 mA, 1 FO B Conduit Connections: 1/2-inch NPT - No gland I Approvals: IECEx- Equipment Category 2 ( Zone 1 - Flameproof Terminal Compartment ) E Language: English Installation Manual and English Configuration Manual Z Software Options 1: Flow and Density Variables (Standard) Z Software Options 2: No Software Options 2 Z Factory Options: Standard Product PK Add on Option: 2-inch Pipe Mount U-Bolt Kit for Electronics

• Coriolis meters
• biased
• inaccurate

Dokeun Hwang

• 7 Dec 2020 2:44 AM
• Cancel

4 Replies

• Jared Drennen

  

  • 7 Dec 2020 4:07 PM

  Hello Dokeun, there are many reasons why two measurements may not match - but for two measurements that previously matched, the first question to ask yourself is - what changed? Was a valve or regulator replaced, a check valve installed, or a spool piece added? If the line is vacuum jacketed, is the vacuum still intact? Is the fluid still in single-phase liquid at both measurement points? Were the meter's flow calibration factors or outputs rescaled recently? (check the calibration paperwork to validate). Also, how is the reference flow coefficient calculated, and is the methodology valid at all flow rates? (vs. a gravimetric weigh scale test for instance). I would start with this type of investigation. Regarding self-test capability of the meter, the two part numbers you provided are not directly compatible with one another, so one of them might be incorrect (or were retro-fitted using other existing equipment - namely a remote core processor). Feel free to send your serial numbers to flow.support@emerson.com to explore if your meter may be equipped with Smart Meter Verification capabilities. I hope this helps.

  • Cancel
• Marc Buttler

  

  • 7 Dec 2020 4:20 PM

  This is unexpected behavior in a application like yours.  Typically, the only thing that could cause a change in a Coriolis meter mass flow bias error as suggested by this data would be some physical damage to the flow tubes inside the meter.  Is it possible the meter may have experienced any extreme internal pressure that would be higher than the rated pressure?  We have diagnostics called Smart Meter Verification (SMV) that can detect damage to the tubes, but I am not sure if your meter has this feature.  

  Please contact me by email at marc.buttler@emerson.com so that we can engage our engineers directly to help diagnose the problem for you.

  • Cancel
• Marc Buttler

  

  • 7 Dec 2020 4:24 PM

  I see that Jared Drennen has also responded to you. You can contact either one of us for assistance, we work together.

  • Cancel
• Jeffrey Mach

  

  • 8 Dec 2020 3:32 AM

  I think the recommendations so far are the most important. A couple other things to check further down the list:
  
  Do you take the flow rate digitally or with the 4-20 mA analog signal? If you are using the analog signal, it wouldn't hurt to check that the analog output and your recording system's analog input are accurate (or at least agree with each other). If you connect to it digitally or via HART, you can verify that the flow that it is measuring matches the value being recorded. (In other words, if your data has changed, by all means check the instrument, but also check the data system, too.)
  
  One scary possibility that comes to mind is that when the unit was last calibrated, it was not zeroed first or worse, zeroed when there was flow in the system, and then adjusted only at the higher end, which would explain the convergence as the flow rate increases, and divergence as it decreases, but that seems to be a disturbingly large zero effect. Another calibration error would be if the LRV and URV of the unit were somehow altered. It is not impossible that the calibrator could have reset the LRV and URV for their calibration and then forget to set it back. Fortunately, this should be easy to check if you can communicate with the transmitter.
  
  On the more unusual end, but you did mention using liquid oxygen: the user manual also notes there can be issues caused by 2-phase flow (you would have to ask Emerson about that). Has the temperature of your process, or at least the flowmeter changed? Was some insulation removed or the measurement made before the instrument could reach normal operating temperature? It doesn't take much to get LOx to boil. They recommend you check out: www.emerson.com/.../white-paper-entrained-gas-handling-in-coriolis-flowmeters-micro-motion-en-64446.pdf
  
  Do you compensate for line pressure, and if so, has your process changed with respect to that recently? Even before the pressure effects are high enough to cause a permanent calibration shift, it can be a significant shift if you are at the low end of the flow rate range the instrument is designed for. As the line pressure increases, the vibrating tubes stiffen, which (slightly) alters the output. What is your output relative to the maximum flow for the meter and can you calculate the line pressure correction (provided in the instrument specification)? This seems unlikely due to the disturbingly large effect, though. Typical values would be on the order of –0.0002% of rate per psi change.

  • Cancel