Coriolis Process Temperature Effect

Dear All

I need some help in estimating the temperature effect on the overall uncertainty of the Emerson Coriolis meter. In the datasheet, under the title of "Temperature Measurement", it says the following:

Temperature is a measured variable that is available as an output. The temperature is also used internal to the sensor to compensate for temperature influences on Young’s Modulus of Elasticity.

Then it defines the accuracy associated with this temperature sensor under the title of "Accuracy and repeatability on temperature".

It then tells us the "Process Temperature Effect" on mass flow accuracy under the title "Process Temperature Effect". It is also stated that if the meter is zeroed at operating condition, then this effect can be zero. 

Now imagine that the process temperature is varying over a certain range, it means that the temperature effect has to be considered. Also, the temperature sensor used to calculate the difference between operating and calibration temperature will be the integrated temperature sensor, therefore, its uncertainty should also be considered as mentioned under the heading of "Accuracy and repeatability on temperature". 

But surprisingly, in AGA 11, we do not see the temperature effect as a valid uncertainty contributor in Coriolis mass flow. Why do you think its like that?

  • Hi Shaiq.

    Perhaps I can help explain.

    The "Process Temperature Effect" specification that you see in Micro Motion Coriolis meter Product Data Sheets is the anticipated maximum random variation of the meter zero in response to temperature changes from the last temperature that the meter zero was checked and/or calibrated at. This is why you see the statement about how zero checking and/or calibrating the meter while at the operating temperature will minimize this source of random uncertainty because it verifies the meter zero at the process temperature. Note that the temperature measurement of the onboard RTD has no relation to the meter zero; only the mass flow rate span is corrected for changes in Young's Modulus of Elasticity due to changes in temperature of the flow tube walls. Mass flow span uncertainty associated with the uncertainty of the RTD temperature measurement as that effects the uncertainty of the Young's Modulus related correction is not stated separately, as it is already incorporated into the overall mass flow uncertainty statement.
  • In reply to Marc Buttler:

    Hi Marc

    Thanks for your response.

    So when we say, zeroing the meter at process temperature, that could be one instantaneous temperature at which the meter is zeroed. In the actual process, we may have a temperature range over which the process temperature is varying. In that case, for every degree change in temperature, the meter mass flow accuracy will be impacted by the factor stated under the title of "process temperature effect". please correct if I'm wrong:

    Let us suppose that deltaT = Operating temperature - calibration temperature

    Now as per the datasheet, this internal RTD (temperature sensor) has an uncertainty associated with it which is clearly mentioned in the micromotion datasheet. So the uncertainty with the internal RTD will contribute towards the uncertainty of the deltaT and thus impact the overall mass flow accuracy. Please confirm if Im correct.

    Now can you help me where is it mentioned in the datasheet that the stated mass flow accuracy includes the uncertainty of the RTD temperature measurement?

    Regards

    Shaiq
  • In reply to Shaiq Bashir:

    To be clear, the "Process Temperature Effect" stated is an expanding range of potential random uncertainty. That is to say, for every degree of temperature change from the latest temperature where the zero has last been verified as correct, the meter mass flow zero value may or may not be impacted by the amount up to the factor stated. It is possible, and would not even be unusual, for the temperature to change considerably and yet result in no change to the zero value. It is also worth noting that the process temperature effect factor is stated in terms of absolute flow rate because it is the potential for ZERO variation, which means that the impact as a percentage of the flow rate becomes smaller and smaller as the actual flow rate increases.
  • In reply to Marc Buttler:

    Hi Marc

    Thanks for this tremendous guidance. Please allow me to understand it in steps.

    So when we calculate uncertainty, we talk about the worst credible scenario. So, would you agree that Process Temperature Effect can cause an impact on the mass flow accuracy if the Operating temperature is away from the last calibrated temperature? This impact can be approximated by the factor given in the datasheet.?

    Regards

    Shaiq
  • In reply to Shaiq Bashir:

    Correct. Although it is also possible to verify the zero at any new temperature without necessarily having to actually calibrate or adjust it, so long as the zero was not changed by the change in temperature.
  • In reply to Marc Buttler:

    Dear Marc

    Thanks a lot. Now the last question in this regard.

    The internal temperature sensor of the Coriolis has a measurement uncertainty associated with it which is given by the specification under the title of "Accuracy and repeatability on temperature". This is the same internal temperature sensor responsible for reading process temperature as well?

    Please confirm if this understanding is correct.

    Regards
  • In reply to Marc Buttler:

    Dear Marc

    Thanks a lot. Now the last question in this regard.

    The internal temperature sensor of the Coriolis has a measurement uncertainty associated with it which is given by the specification under the title of "Accuracy and repeatability on temperature". This is the same internal temperature sensor responsible for reading process temperature as well?

    Please confirm if this understanding is correct.

    Regards
  • In reply to Shaiq Bashir:

    That is correct. The "accuracy and repeatability on temperature" specification is meant to convey the expected accuracy of the internal RTD for representing the approximate process fluid temperature. It is not as accurate as a typical thermowell because it is designed primarily to provide a measurement of the flow meter vibrating tube wall temperature. Because the tube wall temperature is typically very close to the process fluid temperature, this also provides a coarse measurement of the process fluid temperature as a secondary output. However, if more accurate process fluid temperature is required, a thermowell would typically provide considerably better accuracy because it is immersed in the flow stream.
  • In reply to Shaiq Bashir:

    That is correct. The "accuracy and repeatability on temperature" specification is meant to convey the expected accuracy of the internal RTD for representing the approximate process fluid temperature. It is not as accurate as a typical thermowell because it is designed primarily to provide a measurement of the flow meter vibrating tube wall temperature. Because the tube wall temperature is typically very close to the process fluid temperature, this also provides a coarse measurement of the process fluid temperature as a secondary output. However, if more accurate process fluid temperature is required, a thermowell would typically provide considerably better accuracy because it is immersed in the flow stream.
  • In reply to Marc Buttler:

    Dear Marc

    Yes thanks a lot for such a great help and explanation. Now I understood it. I will calculate this overall uncertainty and would like to share with you the uncertainty budget once its ready for your review for a typical Emerson Coriolis Meter. Hope to get more technical insights from you on it.

    Regards