All Member Open Forum

 View Only
Back to discussions
Expand all | Collapse all

Measuring MTTR and MTBF at plant level

  • 1.  Measuring MTTR and MTBF at plant level

    Posted 10-30-2020 07:38 AM

    Dear M&R professionals,

    I am aware MTTR and MTBF should be measured at equipment or component level.At plant level we should be measuring OEE.
    Can MTTR (maintenance metric) and MTBF ( reliability metric) be measured at plant level?

    I would appreciate your response.



    ------------------------------
    Abid Baqir
    ------------------------------


  • 2.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-02-2020 06:31 AM
    It is possible but could be compared to measuring fuel consumption for a fleet of trucks. It will hide the data for bad actors and inefficiencies where the improvement focus should be.

    ------------------------------
    Marius Basson
    President
    Aladon
    Wilmington NC
    ------------------------------

    Original Message


  • 3.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-03-2020 06:26 AM
    Hi Abid,

    As some others have shared already, it's possible but could be completely meaningless unless you are focusing on metrics for your most critical machine, i.e. your bottleneck (which should also be the pacemaker at your plant). This requires doing first a criticality analysis of the equipment.

    Making a comparison to the production world, it's like measuring OEE* for the plant. It's only meaningful if you are measuring it at your pacemaker/bottleneck.

    Best
    Alex Sierra

    *OEE=Overall Equipment Efficiency
    --
    Alejandro Sierra





    Original Message


  • 4.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-04-2020 10:55 AM
    Good morning Abid -

    What Alejandro said makes sense.  I would like to expand on your question what should be measured at the plant level.  I have worked on two LNG facilities developing bad actors list using data automation.  I want to see the comparison of these two facilities - what causes the most heartburn, which then provides guidance for reliability engineers to solve the right problems.  The criteria used were production loss, maintenance cost, and repeat failures. I worked with the IT teams using a business analytics application and connected into production loss and CMMS (maintenance records) databases.  The final product, an ad-hoc reporting tool, identifies a list of top 20 bad actors for both LNG facilities. The bad actors list is measured at a plant level and provides meaningful information to identify the right problem where a reliability improvement plan should be developed.  Another benefit of the plant level comparison is lessons learned.  If one equipment fails often while not in another plant, we may be able to look into the differences and transfer knowledge. 


    Suzy Jiang

    Reliability Engineering and Maintenance Professional

    LNG Operations and Upstream Technology

     713.962.0546 cell 

    Suzy.Jiang11@gmail.com

    www.linkedIn.com/in/suzy-jiang




    Original Message


  • 5.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-02-2020 08:12 AM
    I would say that is not a good idea. As Marius put it, it will hide your bad actors. I would add that the metrics should drive behavior. Looking at these 2 at the plant level may put the focus on the wrong things.


    Scott Fortner
    Honda of America
    ​

    ------------------------------
    Scott Fortner
    Equipment Service Coordinator
    Honda of America Manufacturing
    Marysville OH
    ------------------------------

    Original Message


  • 6.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-02-2020 05:15 PM
    Agree as others have stated.  If you want to get a gauge of MTTR as a global maintenance metric for a whole plant, perhaps a wrench time study or productivity analysis is more appropriate?

    MTBF should be grouped by equipment class, failure type, etc for weibull analysis.  Our plant tracks rotating equipment MTBR, thats about a general as we get.

    ------------------------------
    Ronald Chaumont
    INEOS Olefins & Polymers USA
    Friendswood TX
    ------------------------------

    Original Message


  • 7.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-02-2020 05:17 PM
    Abid, yes, you can measure MTBF and MTTR on the site using site data, in fact that is how the 2 metrics are used.  Mean Time Between Failure is exactly that, the measurement of time it takes to repair a piece of equipment.  As example, a production line is running and all of a sudden a piece of the equipment fails and hence shuts the equipment line down.  Maintenace immediately responds and it takes them 2 hours to respond, another 1/2 hour to gather the necessary tools, mechanics, parts, etc.  The mechanics spend 3 hours replacing the failed equipment, another hour to test the fixed equipment, and another hour to test for produced product.  Finally, after 7.5 hours the production line is up and running producing product in accordance with operation's specs.  Each time the failure occurs the work done along with the duration is recorded and MTBF is measured.  The desire is to measure MTBF and over time take action to decrease it so that downtime is minimized.  MTBF measures how often the same failure occurs.  If the equipment failure over time (might be failure every few months or every few days).  Reliability, and there might be dozens of reliability failures, should use the duration between the same equipment failures (MTBF) and over time the MTBF should increase while MTTR should decline.  OEE is one way of gathering the info needed to make these improvements.
    I hope this helps some, if not let me know.
    Roger Hucker
    Principal Consultant
    Improved Production Capability



    Original Message


  • 8.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-03-2020 09:02 AM

    Roger,
    Just a correction (I am sure it just slipped by) but the following is not true:

    Mean Time Between Failure is exactly that, the measurement of time it takes to repair a piece of equipment (not true).  MTTR is used to calculate the repair time. MTBF is used to calculate the equipment reliability and may include the MTTR. 

    Also, each time the failure occurs the work done along with the duration is recorded and MTBF is measured.  This is not entirely true. The desire is to measure MTBF and over time take action to decrease it so that downtime is minimized. The downtime is reduced through improving the MTTR. This is improving the effectiveness of the crew and does not have the same bearing as improving MTBF. 

    The intent is to reduce the number of failures, increase reliability therefore extend the MTBF. 

    MTBF measures how often the same failure occurs.  This is true If the equipment failure over time (might be failure every few months or every few days).  Reliability, and there might be dozens of reliability failures, should use the duration between the same equipment failures (MTBF) and over time the MTBF should increase (improve reliability) while MTTR (reduce downtime) should decline.  OEE is one way of gathering the info needed to make these improvements.

    I hope this helps some, if not let me know.


    ------------------------------
    Marius Basson
    President
    Aladon
    Wilmington NC
    ------------------------------

    Original Message


  • 9.  RE: Measuring MTTR and MTBF at plant level

    Posted 12-07-2020 09:34 AM
    Hi Roger,

    I totally agree with what  Mr. Marius Basson. OEE still stands for overall equipment efficiency, not overall plant efficiency !!?? so if you will measure the efficiency of equipment it will be under the MTBF & MTTR.

    what I'm saying is as much the equipment reliable as mush the efficiency is high as much as the equipment not reliable as much as the efficiency is low. losing production is not due to pant failure. it is due to equipment failure.

    that's how I see it plant wise 
    ------------------------
    Tariq AlRashdi
    Mechanical supervisor 
    PETROKEMYA SABIC Company
    T.I.

    ------------------------------
    Tariq Alrashdi
    T.I.
    Tariq Alrashdi
    Al Jubail
    ------------------------------

    Original Message


  • 10.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-03-2020 09:56 AM
    It's quite possible but the resolution and the quality of the final image you will have regarding your facility condition is like observing moon with a telescope or a kids binocular. MTTR and MTBF will not provide you a detailed image of what's going on at your facility.

    ------------------------------
    Yaser Sahebi PhD
    Air Liquide Large Industries
    Houston TX
    ------------------------------

    Original Message


  • 11.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-03-2020 07:07 PM
    It is a good question. what you say about the OEE is also valid although the SMRP advises that it be used at the equipment level.
    Dear greeting.

    ------------------------------
    Ronald Carrasco ME
    Ingeniero Mecánico Electricista
    CCCC
    Marcona
    ------------------------------

    Original Message


  • 12.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-03-2020 09:28 PM
    Yes.  While OEE provides production information and identifies the big downtime causes, MTTR / MTBF can and will measure and report Mean Time To Repair and Mean Time Between Failure for a single piece of equipment.  Over time these metrics will lead you to the causes of lost production.
    Roger Hucker



    Original Message


  • 13.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-04-2020 09:37 AM
    Abid,

    I'll take a slightly different thought process to the previous replies. I would say there is very little value for measuring MTBF and MTTR for an entire plant. I would say that if you look only at the equipment in the systems that directly impact production you could use production system MTBF in a manner similar to how you can use overall vibration level averages to approximate the health of rotating equipment. If you practice precision alignment and balancing you would see the overall vibration level average values trend lower. Same philosophy can be used with MTBF. But as others have said, it's not useful for discrete machines; it's really about how you intend to use the information.

    With regard to MTTR, I would be very careful about using that measure. If you want a reliable organization you want your tradesmen to do precision work. Trying to drive MTTR down may send the message you want fast repairs, not quality repairs. It's much, much better to take an extra couple of hours to do a job correctly, with precision, and avoid having to do another repair a short time later because it was not done right the last time.

    ------------------------------
    Tom Moriarty
    President, Alidade MER, Inc.
    Satellite Beach FL
    http://www.alidade-mer.com
    (321) 773-3356
    tjmpe@alidade-mer.com
    ------------------------------

    Original Message


  • 14.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-04-2020 10:12 AM
    There are only two rules:

    Do the right maintenance (improve MTBF) and do maintenance right (improve MTTR)

    ------------------------------
    Marius Basson
    President
    Aladon
    Wilmington NC
    ------------------------------

    Original Message


  • 15.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-04-2020 02:45 PM
    Several good responses.  My 2 cents.  Yes I guess it could but I don't think there is a lot of value at plant level for either.  I mostly would only use to evaluate isolated equipment performance or common equipment comparisons for MTBF.  That can highlight a blue herring in the statistical footprint that may have some issues.

    For MTTR, I would again only use in a comparison to other like or similar equipment to help identify some issues or opportunities.  MTTR is not a good metric by itself or used at a plant or area level to evaluate maintenance performance or quality.  Most crews already focus on getting the job done quickly to a fault sometimes, they don't need a metric for that.  Something done faster can mean shortcuts taken which is not the best maintenance result.  Doing a job right may increase MTTR but can also increase MTBF if the root cause of failure is infant mortality related to installation or assembly errors. 

    A long MTTR can also indicate there are other issues in the maintenance process such as stores material supply, craft training or maint procedures.  Picking an ideal MTTR may be a slippery slope on each side.

    ------------------------------
    Randy Riddell, CMRP, PSAP, CLS
    Reliability Manager
    Essity
    Cherokee AL
    ------------------------------

    Original Message


  • 16.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-05-2020 03:57 PM
    Healthy discussion here. I agree that there are many metrics that are traditionally applied at the equipment level (MTBF, MTTR, OEE, etc.) but they in fact can be applied to asset systems and eventually the plant level. It's really just a matter of perspective.  A plant is a collection of assets (equipment) much like a piece of equipment is a collection of components (and yes you can even apply MTBF at the component level if you wish). Applying these at the plant level can provide an indication of the effectiveness of your maintenance and reliability program as a whole (OEE as number and trend, MBTF and MTTR as trends only). All are correct in that you will not be able to know where your problems lie however. But regardless of where the problems may lie in a complex system of assets known as the plant, where let's be frank very few people have the ability to measure any of these everywhere they would like - would you not be proud to say you increased plant OEE and MTBF over time (where F is now represents a plant shutdown) and reduced MTTR (where R is a restored plant)? I know I would.

    ------------------------------
    Jean-Pierre Pascoli PEng, CAMA, CMRP
    Director, Asset Management & Reliability
    Cameco Corporation
    Port Hope ON
    ------------------------------

    Original Message


  • 17.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-05-2020 05:30 PM
    Hello Abid,

     

    There are some great insights from everyone contributing to this discussion. I will share my experience in working for an organization that made the decision to track MTTR and MTBF at the plant level. I have been in the Maintenance team for just over 7 years now in various roles. The maintenance program was in its infancy at the time I joined the team. Leadership wanted a way to measure the success of the newly established maintenance program and MTTR and MTBF were chosen as plant level metrics along with targeted improvements from the baseline of the data available on the assets from the previous year. If I knew then what I know now, I would caution against that just as others have in this discussion.  Here is why. During that first year, the Maintenance team would meet with the Factory Manager each month to review the status of the maintenance program which always involved reviewing the plant level metrics. Well the metrics were always Green (Green = Good, Red = Bad) and trending favorably each month. The Factory Manager would always stand up from the table and ask how the metrics can be green when all he gets is complaints from Operations about the amount of downtime and the lines not able to hit their planned jobs per hour each day.  The Team would leave those meetings feeling like kids that got scolded for something they did not do. Well, we didn't know what we didn't know. And in hindsight, the Factory Manager's complaints were valid.

    So, when you calculate MTTR and MTBF using all your assets and 80 to 85% of those assets experience very little to no failures, then the 15 to 20% of the assets that are causing you grief will go unnoticed. Simply because the impact is watered down by the assets performing well.  Measuring this way may not give you a true assessment or indicator as to the performance of the plant. That is exactly what we were experiencing in the early years.

    Fast forward to present day, we still maintain what are considered plant level metrics. However, the MTBF and MTTR are focused on the constraint work center for each value stream and the critical equipment as defined by an asset criticality analysis. For us, the unscheduled downtime of these work centers and assets have a significant impact on velocity and throughput for the plant and it is reflected in our MTTR and MTBF metrics. The metrics then guide the prioritization efforts of RCCA and RCM activity.

    That is my input. I hope this discussion helps you find that which you are seeking.



    ------------------------------
    Zachary Witkowski
    Facilities Engineering Supervisor, CMRP
    Caterpillar
    San Antonio TX
    ------------------------------

    Original Message


  • 18.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-06-2020 01:56 PM
    Edited by Scott Hamilton 11-06-2020 01:57 PM
    Actually we break up our production lines into areas based what process is being done and we measure MTTR and MTBF in each individual area.  This allows Management to look at certain areas that are not performing well.  Since different personnel are responsible for different areas it helps to determine if training of specific associates is needed or if PM's of a specific area may be ineffective.  To have one number at a department level we combine the data from all the areas for both MTTR and MTBF and calculate Equipment Availability.  We use this number to track trends but have the specific data by area to use for analysis.

    ------------------------------
    Scott Hamilton CRL;CMM
    Staff Engineer
    Honda Mfg. of Alabama
    Lincoln, AL
    ------------------------------

    Original Message


  • 19.  RE: Measuring MTTR and MTBF at plant level

    Posted 11-21-2020 12:40 AM
    Edited by Abid Baqir 11-21-2020 12:40 AM
    Dear All,

    I just wanted to thank you all for taking out your valuable time to reply to my thread. I really appreciate.


    ------------------------------
    Abid Baqir
    ------------------------------

    Original Message


  • 20.  RE: Measuring MTTR and MTBF at plant level

    Posted 12-04-2020 11:40 PM
    Dear Abid, thats a question that I made to myself several times. My answer was like, hey, you can only measure MTTR and MTBF at a component level, not to a whole plant. Then I thought in a plant, as an equipment. At the end, aren't both systems? So I started to run some numbers and realize that they really make sense. I calculate the Availability, MTTR and MTBF of each of the plants we have and compare them in a daily basis. The major issue that I had was how to consider that failures that don't stop the whole plant. I solved that applying a single factor (or wheight if you will) to each equipment that fail. For example if I have one part of the process where 4 equipments of the same type are used, working in parallel, and doing the same task I apply 25% to each one. If one of them fails I apply the 0.25 factor to obtain the total time impact. If that failure lasted 4 hours, I consider only one in the calculation of the availability, MTTR and MTBF. I did a BI report that allows me to track in realtime that KPI's along all the plants of the company. So in my experience: yes, you can measure MTTR and MTBF at a plant level. 


    ------------------------------
    Gustavo Soto
    Callao
    ------------------------------

    Original Message


  • 21.  RE: Measuring MTTR and MTBF at plant level

    Posted 12-09-2020 12:46 PM
    The central theme from everyone's super insightful responses is that the level of your functional location hierarchy where you evaluate your metric depends on:

    - Understanding what you are trying to achieve with this measurement, and
    - How you are measuring it

    The successful scenarios above all have well defined and clearly understood use case and objectives, in addition to an understanding of how the metric (and input data) supports that.

    I wanted to add to this thread: ISO 14224, a data standards for oil and gas, has a great discussion and table relating general maintenance and reliability data and parameters to different levels of the functional location hierarchy.  For example, it points out that "For data used in availability analyses, the reliability at the equipment-unit level can be the only data required, while an RCM analysis and root-cause analysis can require data on failure mechanism at the component/maintainable item, or parts, level."

    There is also a really nice way of presenting the stuff discussed in this thread more generally (page 32 in the 2016 edition), relating different types of things you may want to measure (such as impact of failure on safety) against the functional location hierarchy.  It's a nice, organized way to more generally think about this.

    ------------------------------
    Sarah Lukens
    GE Digital
    Roanoke VA
    ------------------------------

    Original Message


  • 22.  RE: Measuring MTTR and MTBF at plant level

    Posted 12-08-2020 10:00 AM
    Abid,
    At the plant level you are measuring mean time between maintenance (MTBM) and mean time to maintain (MTTM). At this level you are into operational availability as compared to inherent and achievable. The types and measures are defined in the document titled, Availability is the Defining Framework, at https://analytics4strategy.com/ddassetmgt 

    Actually measurement of availability is complex because level of performance is the core issue.  The is seen in the definition of availability as the probability that a plant, subsystem or item will be in a state to perform a required function at specified standards of performance under given conditions when called for; assuming cost-effective support with respect to working conditions, processes and resources. Until the plant is returned to the level we are actually measuring MTBM and MTTM of a different "system."

    I believe that we should establish what takes place within the maintenance and reliability operations to influence MTBM and MTTM and thene focus on moving them within the system.

    ------------------------------
    Richard Lamb
    Analytics4strategy.com
    rchrd.lamb@gmail.com Houston TX
    http://analytics4strategy.com
    ------------------------------

    Original Message


  • 23.  RE: Measuring MTTR and MTBF at plant level

    Posted 12-16-2020 05:42 AM
    I've used MTBF by equipment groups such as centrifugal pumps or discipline levels at individual production areas and overall site levels for benchmarking. The intent is a measure for checking the effectiveness of bad actor elimination programs and the same data will generate the lists of bad actors. At equipment group level, you can benchmark with industry. However, the value in the measurement is not the absolute numbers but the trends.

    ------------------------------
    Tee Yeow Hum
    Reliability Manager
    Shell
    Singapore
    ------------------------------

    Original Message


Related Content