All Member Open Forum

 View Only
  • 1.  Sensors in CBM: Are You Seeing Real Value?

    Posted 16 days ago

    Good morning everyone,

    I wanted to revisit a topic from a slightly different angle-how many of you are using sensors (vibration, oil analysis, infrared, etc.) as part of your condition-based maintenance strategy?

    More importantly, are you seeing measurable value from that data-fewer failures, improved planning, or even cost impact?

    If not, what's the biggest challenge: budget, bandwidth, or turning data into actionable insight?

    Curious how others are making this work in practice.



    ------------------------------
    Linda Perry
    Senior Business Development Executive
    The Viswa Group
    American Canyon CA
    ------------------------------


  • 2.  RE: Sensors in CBM: Are You Seeing Real Value?

    Posted 14 days ago

    Some large rotating assets, such as turbines or centrifuges, have real-time hard-wired RMS vibration sensors, so they can be automatically shut down if there is a problem that causes vibration to reach an unsafe level. We have a few critical pumps that run infrequently, and that can't be started manually if there doesn't happen to be enough liquid in the supply tank, so I plan to put wireless full-spectrum vibration sensors on them. I'm also considering oil analysis sensors for air compressors, where it's difficult to pull a sample. Other than that, for most of my assets I have yet to see the value in installing permanent sensors. To cover every asset with good vibration sensors would cost 4.5 to 6 million dollars. (I could do it for much less than that if I bought RMS-only sensors, but those don't provide enough information to be useful.) 

    For less than a few million dollars I have hired some technicians and trained them in ultrasound-based lubrication. They visit every rotating asset on some frequency (which depends on each asset's criticality) and they listen to it. The more critical assets also get checked with route-based vibration analysis. This way, everything gets a PdM check and everything gets lubricated properly. Also, by paying attention and using all of their senses, the technicians identify and correct all kinds of problems that fixed sensors will never pick up, such as leaks, unusual noises, damaged or missing guards, tripping hazards, and so on.

    One could argue that, by not being online, we aren't seeing faults the moment they appear. Most faults take a long time to develop, so (for most assets) it's normally not necessary to see them immediately. I've been pleased with the reduction in unplanned downtime that we have experienced with this approach.



    ------------------------------
    Dale Nicholson, PE, CMRP, CRL
    Reliability Engineering Mgr
    Evonik Corp
    ------------------------------



  • 3.  RE: Sensors in CBM: Are You Seeing Real Value?

    Posted 14 days ago

    Linda, 

    We do use various condition-based monitoring tools here at Boardwalk including oil analysis, vibration analysis, and remote monitoring of assets using a large variety of sensors on the equipment. 

    We do see real value in monitoring the equipment, it's helped us avoid failures and keep the equipment running in good condition and helps with our shutdown planning.

    The primary thing you need to look for when starting a program is you need to identify specific failure modes that you want to detect. 

    Bad compressor valves, 

    Failed prechamber check valves, 

    Bad spark plugs, 

    failed transformers, 

    ruined turbochargers,

    fouled heat exchangers,

    soft foot,

    valve lash,

    fouled lube oil filters,

    unstable combustion, 

    loose wastegate, 

    cracked head gaskets, 

    failed deactivators. 

    Anomaly detection is only valuable if you can act on it.   

    Hope this helps!

    Best Regards, 



    ------------------------------
    Zachary Whear
    Engineer II
    Boardwalk Pipelines
    Owensboro KY
    ------------------------------



  • 4.  RE: Sensors in CBM: Are You Seeing Real Value?

    Posted 13 days ago

    I"m aligned with the prior two responses.  There is not a one size fits all but in general,  I would say having online digital monitoring for all predictive technologies is cost prohibitive.   We use online vibration sensor and Steam Trap sensors in critical processes or on equipment that can't safely be accessed while running.  We have routes set up with a couple of technicians to use a hand held vibration probe and feed that data back into our server for diagnostics and trending.   We are currently setting up a couple of production lines with FDI technology to monitor valves and instrumentation as this cost is minimal IE cost of a server and license which will enable alerts using NAMUR protocol.   CBM is about using technology where it makes sense to give early warnings for equipment failures.  We run about 2000 instruments and about 1500 valves in one of our facilities so by capturing basic PDM technologies such as Vibration,  Ultrasonic, Infrared,  oil analysis and motor testing only on large motors coupled with digital signals for monitoring VFD's, Insturents and Valves we have covered 85% of our failure modes.   We do see results using CBM and with onslaught of technology to enable prescriptive analytics we will even enhance troubleshooting with this information in the future.   To capture the savings we identify the failure that was prevented and estimate the cost that failure would have incurred had it happened.  This is an area that most companies struggle with as trying to put a dollar value to something that never happened is difficult but you will need some methodology while imperfect give directionally the cost savings to the company so you can business justify some of this expense.



    ------------------------------
    Mark Pospisil
    Program Manager Maintenance Excellence AN Division
    Abbott Laboratories
    Sunbury OH
    ------------------------------



  • 5.  RE: Sensors in CBM: Are You Seeing Real Value?

    Posted 12 days ago

    Hi Linda,

    I would also recommend a means to monitor sensor reliability and availability, we use sensors as a data feed for performance modelling, which are only as good as the sensor uptime. I agree with the other comments, we should not rely on sensors alone, equipment knowledge and familiarity play a big part in our CBM utilising 3 data sources: performance, quantitative and operational. 



    ------------------------------
    David Cargill
    Lead Maintenance & Reliability Engineer
    NOV Rig Technologies
    Forfar
    ------------------------------



  • 6.  RE: Sensors in CBM: Are You Seeing Real Value?

    Posted 11 days ago

    I think our experience with sensors falls in line with the majority of responses here. 

    We use wireless sensors with cloud based data storage BUT it's used in conjunction with an already robust PdM program with monthly route-based vibration, thermography, oil analysis and ultrasonic greasing.

    We see value in having sensors as we can get real time temperature and vibration excursions which we follow up with an off route PdM inspection (2140 and C3 FLIR)   

    Sensors weren't installed randomly either:

    If our historical failure data shows a P-F happening in between our monthly route based inspection a sensor was installed-  additionally where machines ran on a process based operation, such as a pump intermittently running to maintain a critical tank level, or if the physical data collection was at high safety risk to the PdM technician.... among a few other criteria.

    Short answer is Yes, our SAP shows more PdM identified failures

    Biggest challenge: Cost, and Time for initial set up ie setting alarm parameters and since wireless does not integrate with the established route based database, we need to monitor another program in addition to the current slew.

    Hope this helps!



    ------------------------------
    Ravi Ramdeo
    Supervisor , Predictive Maintenance
    K+S Potash Canada
    Bethune SK
    ------------------------------