Dear All:

I try to calculate equipment reliability from failure data that capture from CMMS using by Excel.

I would like to know how to calculate ETA parameter from EXCEL. The attachment file that I calculated, I take Beta and Eta parameter from Weibull software but I want to take by Excel. Have somebody guide to me.

380-EA2C.xls (26 Kb, 42 downloads)

OK! I got it.

Sorry! I'm being confuse!

I concern about reliability calculate for current situation of equipment that illustrate how many percentage of reliability.

Suppose, I have failure time interval(from desired period 5 yrs past)

Unit of measure (hour)
Fault TTF
#1 8376.00
#2 2952.00
#3 3312.00
#4 2616.00
#5 1152.00
#6 888.00
#7 1896.00
#8 3552.00
#9 4824.00
#10 3384.00
#11 5256.00
#12 336.00

I would like to know what is Reliability of this equipment at the moment.

Thank for help

Again!

Alpha = Eta ???

Hi Aromatics,

When you calculate the plot values, call
ln [-ln R(t)] "PlotCDF" and the ln(Ti) values (LogTime)

Beta is the slope of this graph, ie SLOPE(PlotCDF,LogTime)
Now you can calculate the Eta value as:

Eta = EXP(Beta/INTERCEPT(PlotCDF,LogTime))

In the Excel WEIBULL(x,Alpha,Beta,Cumulative) function, Alpha = Beta and Beta = Eta

Hope this helps,

Badders

Thank you, Relman!

That make me clear. I would like to make one thing clear....Failure time or time to failure that use to calculate?

Either.

However, I see from your table above, that you have the headings "Fault" and "TTF" - that stands for Time To Failure, so call it that.

I made a mistake in the formula I posted: it should be

Eta = EXP(-INTERCEPT(PlotCDF,LogTime)/Beta)

Profound apologies if this confused you more!

Hi all,

The attached EXCEL template allows you to automatically calculate the three parameters of a Weibull probability distribution that fits the best to your empirical data (TTF). The template allows you also to censor any data values if it applies.

Regards

Rui Assis

Fit_Weibull_Bernard.xls (256 Kb, 42 downloads)

Hi Aromatics Thailans
I noted you have been asking about Blocksim on www.weibull.com. I am still the UK partner but am now accepted advisers position on a SAP project in Angola for the nationa oil company. if I can help with BS please e mail me on dave@ramsoftuk.com

Thank you Rui!

I've to download more your Excel file that a good learning for new engineer, thanks you again!

Hi David,

Now, our company would to implement project we called "Maintenance Reliability Programs", we have more strategy in place but the system reliability prediction not are in place. You know that Oil&Gas Plant very complexity. So, I need some software to illustrate our system reliability.....I've one tools that I know, BlockSim that easy to use and another AVsim it difficult to learn. Yesterday I've got online demonstrated from Reliasoft Asia about BlockSim, Weibull and a little bit RCM++.

Thank you for your help, I'll contract you.

Dear Rui,

I would like to know what failure distribution is appropriate for equipment? Bartlett's test is the way to solve?

Hello Panuphan B.

Avsim is not a difficult package to learn provide you have the correct training. I would suggest attending a 2 day training course in which you would be able to build availability models of your plant and get results. More advanced user training is also available.

It like driving a car, once you know the basics its easy, but without good instruction it could end in disaster.

If you would like more information on Avsim training see http://www.reliability.com.au/index.asp?pgid=126

Cheers Gary

Reliability Week 2008 is being held in Brisbane, Australia from September 1-5

“BUILDING AND MAINTAINING RELIABLE PLANT”
Click here to find out more information

Dear GLT

As per I said AVsim it difficult to learn because I think when I build complex system in our plant, BlockSim it very easy to do that.

For instance: I have one block that have two direction it mean two function, BlockSim can do that but AVsim could not.

Hi Panuphan,

When several groups of failure originated in different components (failure modes) are combined, it is easy to check by a best of fit test (maximum likelihood, qui-square test or the like) that resulting TTF´s conform to an exponential probability distribution, even if some of the groups fit a Weibull (or any other) probability distribution. So the answer to your question is: an exponential distribution is appropriate to model the failure behaviour of any piece of equipment. The parameter MTTF can be found by simulation using either BlockSim (if you can afford the money) or EXCEL (if you have the skill).

With regard to the Bartlett's test, this specific test is used to test if k samples have equal variances (if homogeneity of variances exist). When you are seeking the parameters of the Weibull that has the best adherence to your data, please note that the data refers to a unique failure mode (a unique sample composed by n TTF´s). Therefore, the Bartlett's test doesn’t apply at all.

Best regards,

Rui Assis

Hello Panuphan B

quote:

For instance: I have one block that have two direction it mean two function, BlockSim can do that but AVsim could not

We have to remember that the a Reliability Block Diagram (RBD) is not a Plant Flow Diagram (PFD). I think that is a common mistake that new users to RBD’s actually make. They take the PFD’s and start to map them only to struggle when a piece of equipment has more than one function or flow path.

An RBD only has one input and one output. It represents the interdependencies between equipment items and this represents the impact on overall availability and in the case of AVsim it also represents plant capacity.

With basic training this concept is easy to understand and follow.

Cheers Gary

Thank again Rui!

I'm clear about how to fit the plot using by mathematics calculation.


Dear GLT

For my instance: I've found the feature to do that in BlockSim (ReliaSoft) they called Mirror.

I found different between AVsim and BlockSim are block and only one input and output but BlockSim can many input and output. How do you think about different?

Hello Panuphan B,

The Avsim module can use the same piece of equipment in several areas of the RBD. In avsim it is not called 'mirror' it is simply a copy and paste functionality.

If you would like an online webex demonstration to share with you the key features and simplicity of Avsim please let me know.

Cheers - Gary

Sure! very thank you GLT

I'd to say something....I knew AVSim before BlockSim, I request DemoCD from Isograph in 2003, 5 year ago. But I knew BlockSim a recent 2 years.

So, please tell me how to webex online demostration from you. I've webex online demon. from ReliaSoft in last week.

Hello Panuphan B,

That would be no problem.
If you want to contact me at my email address below. I will then contact you directly and arrange the avsim webex demo at a mutually convenient time.

My email address: gtyne@globalreliability.com

Cheers Gary