Coriolis Flowmeter Slow Density Response (Troubleshooting / Possible Cause)?

1.  Response of Coriolis technology

A research paper from 2004 shows the relative response rates of DP, Magnetic, Vortex, and Coriolis.

https://www.researchgate.net/publication/236986366_The_dynamic_response_of_Coriolis_mass_flow_meters_Theory_and_applications#pf10

Check the graph on the far left, "Nominal Step Response"

You can see that the Coriolis was the slowest responding of the four technologies, and by extrapolation you can see that the response to a square change at the input was still only 2 seconds, not 30 minutes.  So I agree that if your density is constant, 30 minutes is an unreasonable response for density measurement.

2.  Real density change?

Sometimes the approach is to "kill the messenger" because the message is not what you want to hear.  In this case, is it possible that your "process fluid" has a lower density at the beginning of the flow session, and the density actually increases over 30 minutes to approach what is considered 'normal'?  In other words, the meter is showing you that there is,  in fact, a density change?  (Factors that typically affect density are temperature and/or material blend/composition.)


3.  Entrained Air

Only specific brands/models claim to measure 2 phase flow: liquid with entrained gas.  Siemens' 2100 muffler/trombone Coriolis meter is not one of them.  Could there be some low level of entrained gas (high levels will fault many Coriolis meters) that changes over time?
(sarcasm mode on) But, I know already that you're going to tell me your fluid has no entrained gas.   That's OK, nobody's fluid has entrained gas.  It's just mentioned in all the documentation literature for the other guy's entrained gas. (sarcasm mode off)

4.  Process fluid solidifying

Is the process fluid solidifying on the inside of the Coriolis flow tube(s), affecting tubes' resonant frequency, which would affect the density measurement which is frequency dependent?   And then over time, the solidification is being eroded by the flow, thus producing a measurement that approaches the correct density value over time as the solidification is removed?

5.  Flowmeter Zero

Is the meter zero being shifted over time due changes in the physical installation due to temperature changes during the flow?

Emerson's MicroMotion, arguably the leader in Corolis technology, has a paper on Coriolis for mid-stream Oil & Gas.  But the principles apply for other industries.

https://asgmt.com/wp-content/uploads/2016/02/026_.pdf

Here's the paper's comments on Meter Zero:

"Meter performance, specifically meter zero, can be affected (by the physical piping's) axial, bending, and torsion stresses. When these stresses exist, pressure, weight, and thermal expansion effects can amplify them."

"Solids, sediment, plugging, coatings, or trapped liquids can affect the meter performance, especially when present during zeroing of the meter."

It's my opinion that too many Coriolis installations ignore the fact that Coriolis is not self-zeroing like Magnetic flow meters are.  Coriolis needs, in fact I would opine, demands, periodic zero checks and compensation.

"The piping configuration of a Coriolis installation should consist of block valves up and downstream of the Coriolis meter with bleed valves to facilitate purging of the piping, zeroing of the meter, and maintenance procedures."

6.  Environmental vibration

David Simpson makes people mad because he knows marketing BS when he hears it.  Here's David's comment on Coriolis from his book, Practical Onshore Gas Field Engineering

Proponents and manufactures of Coriolis meters claim that these meters “directly measure density” and “directly measure mass flow.”  Both claims are utter nonsense.  The device has a bent tube located within some very sensitive sensors.  The frequency of vibration of the tube in response to excitation is said to be a direct measure of density. It is actually a direct measure of vibration frequency.  A slug of gas will change the vibration frequency, but latency in the measurement of the frequency will cause the indication to lag the change and for a time that can be as much as 2 minutes, the density used in the calculation will have no relationship with actual density.  There are other normal conditions (e.g., intermittent road traffic and vibration frequencies imposed on the pipe by compression, etc.) where a change unrelated to changing density can change the vibration frequency."
https://www.sciencedirect.com/topics/engineering/coriolis-meter

7.  Time to warm up to operating temperature

All electronics need to achieve an operating temperature in order to produce accurate measurement values.

Flow meters need to be powered up for 20-30 minutes before measurements need to be made so the electronics can get to their operating temperatures.  Measurements cold will not be the same for a given input as measurements made when meter is at operating temperature.