Failure date OREDA-Interpretation

[RajONG]

Hi, Can anyone give a guidance on the interpretation of No. of failures and failure rate in OREDA.
Is the failure rate indicated per year or 10^6hrs/8760hrs = 114yrs.And is it for a single unit or the total population given on the header of the table?
I have a 20 yr old project without any record on No. of Failures. Need to find the Remaining Useful life using Minitab(Monte Carlo method). On referring in OREDA(all modes failure), there are so many doubts arising on the header. Like population(no. of units surveyed), agg. time in service, calendar time, failure rate per 10^6 hours.
For eg.: I have a Regenerator Reflux Pump(Centrifugal) in Gas Processing. I need the no. of failures per month as per OREDA.
Thanks.

 

[IRstuff]

Doesn't the top of the column say "Failure rate per 10^6 hours"? The failure rate is normalized to a single unit, otherwise, it would be nonsensical.

The version I have shows the number of units and installations. Oreda explains that the statistics are not from a full lifetime, but of selected windows of time, since that avoids infant mortality and wearout phases of life. If you are trying to find remaining life, then using the constant failure rate data indiscriminately will possibly result in an overly optimistic answer.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[georgeverghese]

Though I dont have answers to any of your queries, this post brings back memories from over 20years ago - I used to be familiar with working these Monte Carlo simulations for plant availability for various modes of operation with MTBF data from OREDA for a brief period of time - useful info for concept design when you need to set up sparing philosophy for high CAPEX machinery and processing trains. All forgotten now. Vaguely recall Shell has its own data on MTBF values to supplement the OREDA database.

 

[hacksaw]

Most plants are designed with a 20 life span, based on one startup perday.
Beyond 20 years statisics doen matter, you need to start testing metal thickness, and te rest!