I am looking for advice from people who have success with establishment of a set of specifications for Engineering to use in building predictive maintenance into new capital equipment purchases.

Done with one of my clients within GM, and a few others. Happy to share what I am allowed to (and, no, not a solicitation for consulting work). Feel free to email me at howard@motordoc.net and I can answer a few questions and put you in direct contact with a few people who have implemented the above.

There are also military standards for specifying the same thing that may assist you. We use these standards in our US Coast Guard and US Navy work. Happy to point you in that direction, as well.

Sincerely,
Howard

What capital projects we are talking about here? My experience in downstream Oil & gas industry is as follows:
1) Online condition monitoring for major equipment like compressor-turbines in DCS/Control room which monitors vibration levels, lub process parameters, inlet & outlet process etc. And the data trending is done by another software via data acquisition. (I heard online monitoring results can be connected to CMMS in one vendor presentation so automatically a work notifiation will be raised if any faults occur.)

2) Offline condition monitoring using handheld devices such as:

a) vibration monitoring for pumps, motors, fans which are not included in online monitoring

b) lub oil sampling for lab analysis for all eqpt mentioned above

c) thermograpghy for hot eqpt like reformers, reactors, MCC, refractory linings, hotspots etc. We also monitor our highest temp reactor (>1200C) using a camera & the thermogaphic results plotted in Control room. Temp indicating can also be used to indicate hotspots

d) ultrasonic/noise monitoring for wall thinning, leakages, steam traps, bearing damages etc as per Reliabilityweb page

e) Non-instrusive inspection methods such as:
- replica test for detecting microstructural changes in high-temp coils, superheaters, steam drums, boilers to estimate remainant life after undergoing creep
- IRIS for heat exchanger tube inspection to detect tube wall thinning
- TOFD for weld joint inspection for pressure vessels operating under cyclic stress & thus subjected to fatique
- crawler UT for reformer tube weld joint inspection
- intelligent pigging for crude oil or gas pipelines

f) other condition monitoring methods avaialble in Maubray's RCMII book

3) Online process condition monitoring - I believe the above should be complimented further with good process condition monitoring with appropriate instruments installed. Example monitoring temps of heat exchangers or even boilers can indicate the amount of fouling inside & outside their tubes.

4) Heat & material balance eg for steam also can indicates process and equipment conditions.

5) Recently I read about instrument loop performance monitoring in PTQ via a computer in Control room - this is good for a big plants with many trips & control loops

6) Power consumption monitoring to indicate motor efficiency etc

7) I believe visual inspection with other senses is the oldest man's tool for predictive maintenance even though requires experience.

Like to add some more:

8) Signature test for control valves - to determine which control valves to overhaul rather a blanket strategy to overhaul all.

9) Trevi-test for PSVs - to determine which PSVs need to be overhauled rather than a blanket strategy overhaul all regardless of service eg water service is more than steam. Simpler option is to pre-pop before overhaul, if pop at set pressure (+_5%), no need to overhual & reinstall.

10) Weibull analysis -never done this but I heard from experts it's useful

11) Motor circuit analysis - may be MotorDoc can explain this further.

12) Instrument trip & control loop testing - to check which loops have hidden failures

P/s Perhaps you don't mind share a copy of your spec once completed to see how many of the above are included. TQ

This message has been edited. Last edited by: Josh,

Paul,
Typically one would do a Criticality Analysis to determine the criticality of the equipment based on many factors such as health, safety, enviromental, reliability, critical spares, production and any others that may be relavent to your situation. Then a RCM analysis to determine if Run to Failure, PM, PdM or Redesign is appropriate. I then follow up with a simplified FMEA to determine a Risk Factor as well as how that risk is affected by PdM technologies that may be applied. The Risk Factor takes into account Severity, Probability of Occurance, and Detectability of the Failure mode. This is done at the lowest maintainable component level. This process give you an idea of what to include in your spec. Of course it all has to be looked at in the form of cost justification.

I'll close with this.

Is something like a $25 limit switch a critical component?

Paul,

A good reference might be "Operations & Maintenance Best Practices - A Guide to Achieving Operational Efficiency" publsihed by the Department of Energy. It compares various maintenance strategies and technologies, and includes maintenance guidelines for different machinery types. The price is right.....you can download it from the web at:

http://www.eere.energy.gov/femp/pdfs/omguide_complete.pdf

Jon
Spintelligent Labs

Jon:

Excellent reference. Reviewed it this week and will be using it as a reference in a project that I am involved in.

In the meantime, there are a number of MIL Handbooks and Standards for FMECA. I am also sure you will receive recommendations for civilian processes, as well. The MIL-STD for FMECA can be downloaded from my site, along with the original Nowlan & Heap book, "Reliability-Centered Maintenance."

No, they are for free, just too large to attach here:

http://www.motordiagnostics.com/presentations.htm

If you have any trouble, or questions, please feel free to email me: howard@motordoc.net

Howard