A reader who came upon one of the posts in the abnormal situation prevention category asked about ways early fault detection could reduce unplanned downtime in a refinery fluid catalytic cracking (FCC) unit.
Predictive diagnostics can definitely help improve the on-stream availability of your FCC unit. In the absence of historical data specific to your current availability of the FCC unit, generally refiners who have implemented predictive maintenance technologies have reduced their outages by half.
I’m assuming that gross generalizations are not sufficient for your purposes so I’ll share some of the diagnostics and incidents that can be avoided in FCC process units using diagnostics to enable predictive maintenance. The FCC process is unique within a refinery in that it is extremely dynamic-the unit can be in trouble one minute and safely recover the next minute if proper actions are taken, or the process may need to be shut down for extended periods for repair if proper actions are not taken.
There are diagnostics that continually update the status of the transmitter, detecting internal faults and alerting the operating or maintenance staff of problems with the transmitter itself. However, there are also diagnostics and secondary measurements that can inform operations of impending faults in the process in time to take action to avoid process upsets. A common example of a secondary measurement is the internal temperature of the transmitter. This is useful to detect the failure of heat tracing, warning operations of potential freezing problems. Although not an issue in warm climates, it is an issue with viscous oil streams within the FCC unit (feed, main column bottoms), and it can also be useful to detect transmitter overheating.
Within the reactor section of the FCC unit, key process measurements such as the catalyst level and slide valve differential pressure (DP) are measured with differential pressure instruments that require a special purged process connection to keep the lines clear from plugging with catalyst. Rosemount pressure transmitters with
Statistical Process Monitoringplugged impulse line diagnostics can detect the onset of plugging, before the line is so plugged that the measurement is compromised. This diagnostic is very useful for avoiding incidents caused by faulty measurements. This same technology can be used to detect the onset of flooding in fractionation columns and furnace flame instability detection.
Reactor temperature is one of the key operating variables. The advanced diagnostics within the Rosemount temperature transmitter can warn of impending failure of the thermocouple element used to measure the process temperature. Depending upon the actual installation, the transmitter can be configured to automatically switch to a spare thermocouple before failure.
The diagnostics within Fisher digital valve controllers can warn of potential problems with any control valve, before control is lost. Issues with low air supply, torn actuator diaphragm, packing friction (too tight or too loose) can be detected.
Instrument diagnostics can also improve the availability and reliability of Safety Instrumented Systems with DeltaV SIS that can use HART parameters transmitted by the field devices.
The instrumentation specified by the process designer and shown on the Piping and Instrument Diagrams (P&IDs) is typically the minimum amount of instrumentation required to operate the unit. However, operating the process with greater efficiency and reliability often requires additional measurements going beyond the P&ID. WirelessHART-based measurements offer ways to cost effectively add new measurements to monitor pump health, heat exchanger fouling, air-cooled exchanger vibration, fired heater efficiency, etc. These additional measurements also have diagnostics to support the overall reliability improvement of the process.
Hopefully I have provided you with a good idea of a number of ways you can improve reliability of the FCC process unit that you can compare with your own experiences.
Gary, thanks for the great examples!
Update: An astute reader in Australia noted that there is a separate plugged impulse line diagnostic. I turned to Bill Zhou, whom you may recall from earlier pressure measurement posts for clarification. Bill responded:
Statistical Process Monitoring is a flexible diagnostic that can detect decrease and also increase in variability, allowing for numerous possible applications in addition to plugged line detection. This diagnostic is available in HART and Foundation fieldbus.
There is also a separate diagnostic called “Plugged Line Diagnostic” which is only used to detect plugged lines and is only available in Foundation fieldbus.
I’ve updated the post to reflect this correction. Thanks for the sharp eyes, mate!
Update 2: I added the bolded phrase “in addition to plugged line detection” in the update above to clarify that SPM can help to detect plugged lines in HART devices.