Process and Device Diagnosis Modern Flow Meters Also Monitor The Immediate Process Environment
Meters know more about the process than commonly supposed. Flow meters, for instance, can detect critical process states such as partially filled pipelines, non-homogeneous media and deposits. The question is only how users can get hold of this information.
There are, essentially, three different types of real-time diagnosis in continuous operation: device diagnosis, verification of the accuracy of the device, and process diagnosis. Device diagnosis has now become state-of-the-art. “Verification of the device — i.e., whether it is still within specification — is a challenge that we have already solved with our magnetic inductive flow meters,” explains Günter Pinkowski, Director Marketing at Krohne Messtechnik.
It remains difficult, though, to track down the cause of an error message. Jens Niemann Gøbel, Director Product Management at Siemens Flow Instruments, gives an example: “With Coriolis sensors, for instance, air or gas can generate an amplitude or frequency pattern that is similar to a pulsating, intermittent flow or vibrations in the system.” New and more sophisticated algorithms for filtering and detection can help to clearly distinguish between the various causes of faults and provide more differentiated diagnoses.
Diagnostic Functionalities for Monitoring The Process
The diagnostic functionalities for monitoring the process are implemented and assessed differently by the various manufacturers. Many devices on the market come with an empty-pipe or partial-filling detection facility, for instance, but not the ability to detect gas bubbles and monitor conductivity or temperature. ABB, for example, has developed a flow measuring system that allows gas bubbles to be detected from an air fraction of as little as 1.5 to two percent. “Other standard devices do not offer such diagnostic functions until the air fraction in the medium is about ten percent,” according to Cornelia Giebenhain-Wagner, Product Application Specialist for flow measurement technology at ABB. That means the user can respond much sooner to critical states such as a pump running dry.
The density value also affords a very good insight into the process. “It means we can very quickly identify gas pockets and incorrect mix ratios resulting from process faults. The ability of our MID to detect increased process noise allows good insight into the process,” explains Volker Kramer, Product Manager Flow for Emerson Process Management in Germany.
Almost every piece of equipment that is now launched on the market is based on Namur Recommendation NE 107. It deals with the introduction of the Namur status signals such as “Failure”, “Function check”, “Maintenance required” and “Out of specification”. With the AXR two-wire magnetic inductive flow meter brought out by Yokogawa in 2010, for instance, the warning messages are in future to be classified as “F”, “C”, “S” and “M” so that the device is adapted even better to the requirements of NE107.
For Endress+Hauser, NE 107 also contains an extensive wish-list from the user’s point of view. Right at the top is the trend towards “preventative maintenance”. The subject of “application reliability and system availability” continues to gain in importance. For Peter Dietrich, Head of Flow Marketing at Endress+Hauser, the challenge is not just about detecting changing states, though, but first and foremost the unambiguous interpretation of the information. Or, as Namur puts it, better no diagnosis than a wrong diagnosis.