Regarding the issue of false values in the shaft vibration monitoring of axial flow compressors
Thread Content
The factory’s four-in-one nitric acid production unit has experienced several unplanned shutdowns due to interlocks; upon examining the historical data from various monitoring points, it was found that none of them reached the levels that would trigger an interlock shutdown. After the manufacturer checked the historical trend waveforms using an oscilloscope, it was found that the shaft vibration of the axial flow compressor reached 472 micrometers, while the interlock value is 80. Dear experts, could it be that the period during which this abnormal value occurred was too short, and as a result this interlock value was not detected in the historical curve, which led to the shutdown via interlock?
It’s difficult to obtain instantaneous values for historical trends; there is a sampling period involved, which is usually at least 1 second, though in many cases it’s more than 1 second
If your card supports the SOE function, it is capable of recording events on a millisecond level; however, if it is a regular card, the system’s own ability to collect historical data cannot achieve such precision at the millisecond level. If it’s an instantaneous fluctuation, it cannot be recorded. Personally, I think the control systems of compressor units should all have a SOE function.
The SOE function is an optional feature – not available on all models – and this is also a selling point!
Is the controller system of the poster a DCS? Are there other systems for the unit interlocks? Typically, unit protection systems come with a separate system, such as Bentley 3500 or ITCC+3500, which features robust event recording capabilities that allow for the identification of interlock points.
General historical themes are only available in DCS; they are not present in the SIS system. The DCS data refresh is usually 500 milliseconds, so it is normal if there is no update. The SOE record refresh for SIS is 100 milliseconds or less.
The instrument circuit needs to be checked, including the junction box, preamplifier, etc., to identify the root cause
Remove the vibration interlock; displacement is sufficient.
Interlocking is a safety measure; interlocking should indeed occur
The phenomenon you mentioned is not either a false value
The minimum scanning cycle here is 1 second.