I went to the IBM Maximo conference today. IBM is going to take Maximo to new areas such as work tickets for service repairs, IT support "help desk", and other places that don't have anything to do with reliability based maintenance. With the new focus, will they ever build the tools in Maximo to perform task such as RCFA, MTBF and maintenance cost vs replacement, or any other tools that help us control our maintenance. There needs to be more attention to put on the proactive side, rather than the reactive.
Read what they say about the new reliability tools at http://www.datapaste.com/ reliability/CBM/MeanTimeBetweenFailures

Dear all,

I very much appreciated to hear opinions of all the professionals here, such as the authority Paul Barringer, on the issue "MTTF versus MTBF", which is undoubtedly prone to confusions as stated by the previous posts.

To prove that these two terms are misleading, please compare what Paul Barringer in http://www.barringer1.com/TWMS.htm#MTBF_MTTF accepts to be an appropriate definition: "May 2003 What is the difference between MTTF and MTBF? Dr. Bob's reply: Mean Time To Failure (MTTF) is the average life of the part. Mean Time Between Failures (MTBF) is the average time between failures for all the parts in the fleet. To estimate MTBF divide the total operating time on all parts by the number of failures." and the term MTTF which he wrote in cell P22 of the spreadsheet which I attach onceagain.

Please, don't take me wrong. I am not criticizing anybody; I am just calling your attention to the fact that there are too many definitions for the same terms as stated by Joe Petersen and it becomes consequently real hard to avoid controversies.

And, meanwhile we diverted our attention from the original question initiated by Terrence O´Hanlon...

Regards,

Rui

This message has been edited. Last edited by: <Rui Assis>,

Air_compressor.JPG (168 Kb, 6 downloads)

FYI; if I strive for TPM utilizing RCM incorporating RCA, I must have BS for an ME so I can perform as an RE thus lend credibililty to PdM so I'll know MTTF (average life) from MTBF (mean time between failures or average life). All I need is one more acronym for another ICON to become fully literate on how to get the job done or I could dazzle 'em with brillance rather than baffle 'em with another acronym. Is this getting a full prespective? Or do I need a WO to CYA in the CMMS system? Is outsourcing an option?

So from a practical point of view - you could have high availability with low reliability if you get really good and fast at fixing (or maintaining) the failure.

Terry O

Yes Terry and you still can have the opposite: low availability with high reliability if you are not so good at fixing the failure. It sounds like a paradox but I know of a few cases concerning high-tech machinery. Sometimes failures are so few that maintenance people tend to forget how to deal with them and have to look through the documentation thus wasting time.

Regards,

Rui

This message has been edited. Last edited by: <Rui Assis>,

I think when we have high availability with low reliability, most likely the life cycle cost is high due to a lot of repairs to improve reliability and keep the plant running.

When we have low availability with high reliability, the plant may be not running due to no damands for its product or the product cost is too low (not economical) for the plant to run?

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

quote:

Originally posted by Terrence O'Hanlon:
So from a practical point of view - you could have high availability with low reliability if you get really good and fast at fixing (or maintaining) the failure.

Terry O

Terry, I do not buy on maintenance getting good at and fast at fixing failures. This only means that failures tend to repeat itself. If one person becomes good at fixing failure then he encounter the failure too frequently.

Maintenance and reliability is not synonymous to repair but maintenance has much more to do with thinking on what to do to make our assets more reliable.

Most maintenance becomes good at repairing failures because they make it a habit to do so without trying to analyze once and for all why the failure occur on its own.

As I have posted out that Availability is not a measure of reliability but only a measure on when the equipment is available and even if the equipment is available and utilized still it is not a quarantee that the equipment is optimized and reliable as speed can be lowered down.

Terry, Rui,
Of course you are both stating a mathematical truth.

If we are aimimg for world class performance, the math is less relevant than what is between people's ears. A culture that accepts poor reliability (read - frequent failures) is highly unlikely to be great at fixing things well, even if done quickly. These jobs are quite often of poor quality, causing infant mortality problems. Some 30% of breakdowns are attributed to maintenance work itself; poor workmanship (both Ops and Maint) during restarts are a significant issue here. So, while the math may be faultless, poor reliability is a sure start of the downward spiral.
I would argue that maintainability is important, but our first focus must always be on reliability. Once we get that resolved, we can move to improving maintainability. Reliability is a quality issue, maintainability is a productivity issue.

Vee, / Terry

Of course Vee is correct in his reading of the reliability - availability relationship. But there are occasions where Terry's statement rings true.

For example; repeat nuisance alarms within a processing plant will often give the effect of high availability and low reliability due to the short time required to reset them.

Agreed there are underlying issues, sometimes related to design in my recent experience, but not immediately obvious from the availability figure alone.

Another case where reliability suffers and availability does not is in older process plants where there are continual nuisance trips for all sorts of spurious reasons.

Often the only way to fix these things is to spend an amount of money that nobody can afford, so the situation is allowed to continue.

In the mid 1980's I worked at a mine where there were continual errors related to failing brake lights. These were the filament type globes in those days and normally it was spread over the fleet rather than on one particular truck.

The result, high availability of the fleet or of specific trucks, low reliability over the fleet due to the short time taken to replace the globe. (These sorts of situations led to the replacement of filament globes with LED style indicators in mobile mining equipment fleets globally)

In each of these instances you could say that the technicians involved had become "very good at repairing the problems".

The truth is probably that they became very good at returning equipment to service.

I have a section on the relationship between availability and reliability in The Maintenance Scorecard which goes into some of the examples such as these.

Is it true that generally, most would want high availability and high reliability?

Just thought I would add to the discussion in regards to MTBF and MTTF. MTBF (Mean-Time-Between-Failure) is only applicable with the underlying distribution has a constant failure rate (e.g. exponential). In this case, MTBF and MTTF are equivalent. However, if the failure rate is not constant then these terms are not equal. Generally, MTBF is more applicable when dealing with repairable systems.

David

Reliability and availability are two entirely separate notions.

A reliable machine may not be available at the time (power to that section of the plant may be cut off for a different reason.); likewise, an available machine may not be reliable.

Therefore, a good maintenance engineer's job is to make his machines both reliable and available.

I am hope I am not way off. If so blame english being my second language!

Regards
Ibrahim

Reliability vs. maintainability

Production and maintenance folks are often at odds over equipment performance. The debate frequently involves what to measure and how to measure it. Because of this, I found the need to define two commonly used performance measurement terms: reliability and maintainability.
Confusion stems from many thinking reliability relates to everything about a piece of equipment, including maintainability. Although the terms are closely related, they involve very different approaches to maintenance performance and monitoring. I find the following definitions most useful.
Reliability
Reliability is the probability that a component, equipment or line will complete its assigned work within a defined time period. Think of it as the reflection of the design and operation of a piece of equipment throughout its effective life.
But in reality, the definition is, will the equipment or process work when you want it? It's that simple. Reliability deals with the frequency of failure, so reducing how often something fails improves its reliability and decreases downtime.
When analysis of equipment problems indicates the need to improve reliability (decrease the number of production interruptions) on equipment, consider any of the following actions:
"¢ review operating procedures
"¢ improve installation procedures
"¢ increase predictive (PdM) and preventive (PM) maintenance service
"¢ specify stronger components
"¢ improve operating environment
"¢ modify the equipment
"¢ review maintenance and production scheduling
Maintainability
Maintainability measures the impact failure has on production.
It reflects the speed, difficulty, and cost of repairs and PM/PdM on equipment. Its primary concern is lost production. So, improving asset maintainability reduces downtime. Reduced downtime increases production potential. To improve maintainability (reduce the time of production interruption), look at these:
"¢ installation of standby units
"¢ improve PdM testing
"¢ order or increase the number of spare parts
"¢ order emergency spares
"¢ increase maintenance or operator inspections
"¢ review maintenance or operator training
"¢ provide standby labor
"¢ develop contingency plans
"¢ review maintenance/production scheduling and coordination
While many of these are excellent solutions, having extra parts, people and redundant equipment is an expensive way to improve maintainability. It often makes more sense to initially purchase more costly equipment "” designed for reliability and maintainability "” that, in the long run, proves less expensive.
Everybody plays a part
Once you understand the performance measures, who can impact them is the next issue.
Many believe the responsibility and greatest role in equipment reliability and maintainability lies with maintenance. That's not necessarily true. Production, engineering, purchasing and sales all impact equipment performance.
Maintenance is just one group that influences reliability. Sales and marketing determines the product offering and mix.
Sales determines lead times and lot sizes due to customer expectations. The expectations and offerings set the type and length of product runs. The runs, in turn, determine patterns of equipment wear and tear. These affect breakdown frequency.
Also impacting reliability are:
"” production, through operating procedures, methods and scheduling equipment for maintenance;
"” purchasing, by considering life and quality when ordering replacement parts and process materials;
"” engineering, by determining original equipment specifications and design.
One of the most important ways to improve both reliability and maintainability is planning and scheduling all work or runs. If you read the definitions of both terms, it's the only common element on the lists. Scheduling involves developing and following a plan.
A plan ensures the right production status, procedures, parts, resources and people are available. A complex plan needs all the different groups to come together and form a team.
That team works to achieve peak performance for your plant, equipment and systems. Remember, everyone's actions impact your reliability and maintainability.

Welcome to the forum Tkuhn.

I couldn´d agree more with your words. Scientific knowledge, experience and willingness to share are features higly appreciated by us all.

Thanks.

Rui

Sheesh! what a convoluted mess!

Availability - Is ready and able to run means 100% available. If it isn't ready or able, 0% available. No in between numbers allowed.

Reliability - Percentage of time it is available.

You math guys must drive yourselves crazy, besides each other. Interesting discussion!

Math proficiency prevents one from writing nonsense on serious matters.

Regards,

Opinions were asked for on what the two terms, availability and reliability, meant to us. I gave mine. I think most everyone else left the topic a long time ago, but it was interesting. I will move on and leave you experts to argue your equations.
Cheers!