MTTR, MTBF, or MTTF? – A Simple Guide To Failure Metrics

Guide to failure metrics

Asset performance metrics like MTTR, MTBF, and MTTF are essential for any organization with equipment-reliant operations. Only by tracking these critical KPIs can an enterprise maximize uptime and keep disruptions to a minimum.

Tracking the reliability of assets is one challenge that engineering and maintenance managers face daily. While failure metrics can be very useful in this context, to use them effectively, you need to know what meaning hides behind their acronyms, how to distinguish between them, how to calculate them, and what does that tell you about your assets.

That’s why we decided to create a simple to follow guide to failure metrics that will help you avoid costly mistakes and successfully monitor equipment performance.

Introduction to Failure Metrics

Even the most efficient maintenance teams experience equipment failures. That’s why it’s critical to plan for them.

But first, what does equipment failure look like?

Failure exists in varying degrees (e.g. partial or total failure) but in the most basic terms, failure simply means that a system, component, or device can no longer produce specific desired results. Even if a piece of manufacturing equipment is still running and producing items, it has failed if it doesn’t deliver the expected quantities.

Managing failure correctly can help you to significantly reduce its negative impact. To help you effectively manage failures, several critical metrics should be monitored. Understanding these metrics will eliminate guesswork and empower maintenance managers with the hard data they need to make informed decisions.

Which failure metrics should be tracked? Across industries and applications, we’ve found that those are MTTR, MTBF, and MTTF. We’ll discuss what each of those acronyms means and how you can use them to improve your operations.

But before that, we need to discuss one thing that is often overlooked – the importance of having reliable data behind your failure metrics.

The Importance of Reliable Data

To make data-backed improvements in equipment failure, it’s crucial for the right data to be collected and for that data to be accurate.

High-level failure statistics require a significant amount of meaningful data. As we’ll show in the failure metrics calculations below, the following inputs must be collected as part of your maintenance history:

  • Labor hours spent on maintenance
  • Number of breakdowns
  • Operational time (can be calculated from total expected operating hours per week – total equipment downtime)
As tedious as recording maintenance figures can be, it’s an essential part of improving operations. This process can be painfully time-consuming when done manually, but it’s made simple with a mobile CMMS like Limble that lets you quickly and easily log reliable data for labor hours and downtime on your phone while you’re performing maintenance tasks. Additionally, Limble runs all the calculations of MTTR and MTBF automatically for you, as seen below.

Calculating failure metrics with Limble CMMS

Collecting inaccurate data can cause a lot of issues. Maintenance technicians might occasionally write down the wrong figure is just one example. A potentially much bigger problem is neglecting to record tasks, which leads to incomplete data.

If data is missing or inaccurate, your failure metrics will be useless in informing decisions on improving operations. Worse still, if you are unaware that the data is unreliable, you might end up making operational decisions that could be counterproductive and harmful.

Now that we got that out of the way, let’s focus on the things you came for.

30+ Most Commonly Used Maintenance Acronyms

Download Our Maintenance Acronym Guide

Did you know that MTTR can mean both “Mean Time To Repair” and “Mean Time To Recovery”?

What is the difference between “CBM” and “FTM”?

Download our list of all common maintenance acronyms.

Limble CMMS


What is Mean Time To Repair (MTTR)?

Mean Time To Repair (MTTR) refers to the amount of time required to repair a system and restore it to full functionality.

The MTTR clock starts ticking when the repairs start and it goes on until operations are restored. This includes repair time, testing period, and return to the normal operating condition.

How do you calculate MTTR?
To calculate MTTR, divide the total maintenance time by the total number of maintenance actions over a given period of time.

How to calculate MTTR

Imagine a pump that fails three times throughout a workday. The time spent repairing each of those breakdowns totals one hour. In that case, MTTR would be 1 hour / 3 = 20 minutes.

A couple of things to note:

  • Typically, every instance of failure will vary in severity. So while some incidents will require days to repair, others could take mere minutes to fix. Hence, MTTR gives an average of what to expect.
  • To obtain reliable results, every repair must be handled by competent and trained personnel that can follow well-defined procedures.
Every efficient maintenance system always needs to look at how to reduce MTTR as much as possible. That can be done in a few different ways.

One approach is through tracking spare parts and inventory levels (thereby saving on downtime while sourcing for parts).

Another way is to implement proactive maintenance strategies like predictive maintenance. Predictive maintenance (PdM) will, among other things, allow you to better monitor the condition of in-service equipment and predict potential failure more accurately by using condition-monitoring sensors mounted directly on those components that are prone to failure.

These sensors can alert them well in advance when to expect failure. At this point, the repair is no longer reactive but predictive, as the manager has enough time to arrange for all the resources needed to execute the job.

Why is MTTR helpful?
Taking too long to repair a system or equipment is not desirable as it can have a highly unpleasant impact on business results. This is especially the case for processes that are particularly sensitive to failure. It often results in production downtime, missed deadlines, loss of revenue, and so on.

Understanding MTTR is an important tool for any organization because it tells you how efficiently you can respond to and repair any issues with your assets. Most organizations seek to decrease MTTR with an in-house maintenance team supported with the necessary resources, tools, spare parts, and CMMS software.

Maintenance managers can use MTTR to inform maintenance decisions such as:

  • when to repair or replace assets
  • quantity of parts and inventory to have on hand
  • whether to lease or buy equipment
Mean Time To Repair vs Mean Time To Recovery
There are several commonly used terms for the acronym “MTTR.” The two most common are “mean time to repair” (discussed above) and “mean time to recovery.”

Mean time to repair vs Mean time to recovery

Mean Time To Recovery is a measure of the time between the point at which the failure is first discovered until the point at which the equipment returns to operation. So, in addition to repair time, testing period, and return to normal operating condition, it captures failure notification time and diagnosis.

Although both terms are often used interchangeably, the need for distinction becomes important in the context of Service Level Agreements (SLAs) and maintenance contracts.

Hence, all parties to such contracts will need to agree on what exactly are they measuring.

What is Mean Time Between Failures (MTBF)?

The second failure metric we’ll cover is Mean Time Between Failures. MTBF measures the predicted time that passes between one previous failure of a mechanical/electrical system to the next failure during normal operation. In simpler terms, MTBF helps you predict how long an asset can run before the next unplanned breakdown happens.

The expectation that failure will occur at some point is an essential part of MTBF.

The term MTBF is used for repairable systems, but it does not take into account units that are shut down for routine scheduled maintenance (re-calibration, servicing, lubrication) or routine preventive parts replacement. Rather, it captures failures that occur due to design conditions that make it necessary to take the unit out of operation before it can be repaired.

So, while MTTR measures availability, MTBF measures availability and reliability. The higher the figure of the MTBF, the longer the system will likely run before failing.

MTBF (Mean Time Between Failures)

How do you calculate MTBF?
Expressed mathematically, the lapses of time from one failure to the next can be calculated using the sum of operational time divided by the number of failures.

How to calculate MTBF

Looking at the example of the pump we mentioned under MTTR, out of the expected runtime of ten hours, it ran for nine hours and failed for one hour spread over three occasions. So, MTBF = 9 hours / 3 = 3 hours.

As you can see from the example above, the repair time is not included in the calculation of MTBF.

Apart from the design conditions mentioned earlier, other common factors tend to influence the MTBF of systems in the field.

A major one of these factors is human interaction. For instance, low MTBF could either indicate poor handling of the asset by its operators or a poorly-executed repair job in the past.

Why is MTBF helpful?
MTBF is an important marker in reliability engineering and has its roots in the aviation industry, where airplane failure can result in fatalities.

For critical assets such as airplanes, safety equipment, and generators, MTBF is an important indicator of expected performance. Therefore, manufacturers use it as a quantifiable reliability metric and as an essential tool during the design and production stages of many products. It is commonly used today in mechanical and electronic systems design, safe plant operations, product procurement, and so on.

Even everyday decisions like buying a particular brand of car or computer are affected by the buyer’s desire for a product with a higher MTBF than what the next brand has to offer.

Although MTBF does not consider planned maintenance, it can still be applied for things like calculating the frequency of inspections for preventive replacements.

If it is known that an asset will likely run for a certain number of hours before the next failure, introducing preventive actions like lubrication or recalibration can help keep that failure to the minimum and extend the uptime of the asset.

What is Mean Time To Failure (MTTF)?

Mean Time To Failure (MTTF) is a very basic measure of reliability used for non-repairable systems. It represents the length of time that an item is expected to last in operation until it fails.

MTTF is what we commonly refer to as the lifetime of any product or a device. Its value is calculated by looking at a large number of the same kind of items over an extended period and seeing what is their mean time to failure.

Mean time to failure (MTTF)

In the manufacturing industry, MTTF is one of the many metrics commonly used to evaluate the reliability of manufactured products. However, there is still a lot of confusion in differentiating between MTTF and MTBF because they are both somewhat similar in definition. The good news is that this is easily resolved by remembering that while MTBF is used only when referring to repairable items, MTTF is used to refer to non-repairable items.

When using MTTF as a failure metric, repair of the asset is not an option.

How do you calculate MTTF?
MTTF is calculated as the total hours of operation, divided by the total number of items being tracked.

How to calculate MTTF

Let’s assume we tested three identical pumps until all of them failed. The first pump system failed after eight hours, the second one failed at ten hours, and the third failed at twelve hours. MTTF in this instance would be (8 + 10 + 12) / 3 = 10 hours.

This would lead us to the conclusion that this particular type and model of the pump will need to be replaced, on average, every 10 hours.

The only surefire way to increase MTTF is to look for higher-quality items made from more durable materials.

Why is MTTF helpful?
MTTF is an important metric used to estimate the lifespan of products that are not repairable. Common examples of these products range from items like fan belts in automobiles to light bulbs in our homes and offices.

In particular, MTTF is important to reliability engineers when they need to estimate how long a component would last as part of a larger piece of equipment. This is especially true where the entire business process is sensitive to the failure of the equipment in question.

In such cases, MTTF becomes the primary indicator of the equipment’s reliability, intending to maximize asset lifetime. Shorter MTTF means more frequent downtime and disruptions.

How do you replace your CMMS

See the Results Red Hawk Enjoys With Limble

Image Description


Final thoughts

One of the top priorities of maintenance managers is to ensure maximum operational availability of their equipment, as well as keeping equipment operations safe and efficient. Understanding the calculations and use of failure metrics will enable maintenance professionals to determine, with greater accuracy, when a critical asset is most likely to fail.

Based on their findings, they can proceed to develop better asset management strategies and improve their overall maintenance processes.

By calculating failure metrics and planning maintenance based on those results, they can also reduce their organization’s dependence on reactive maintenance in favor of planned (predictive) maintenance, which can be just the thing they need to spark their business’s growth.

If you are one of the many managers and technicians that are looking to improve these metrics by implementing a CMMS into their organization, you can start a free trial of Limble CMMS here or contact us for more information.

Image Description
Share:

55 Comments

  • M.IRANI September 22, 2018, 2:14 am

    greetings:
    Regards, please explain my three questions about the MTBF and MTTR Indicators.

    1. According to the formula for calculating the MTBF index, which is equal to the total operating time of the device divided by the number of emergency repairs in a specified interval, when the machine is healthy and not working, this time is considered as the working time and used in the calculation of the index. Or not?
    2. How to calculate the MTBF index for a device that did not have any failures or emergency repairs during a one-month period?
    3. How to calculate the MTTR index for several different devices in a given time period, such as one month?

    • Bryan Christiansen December 14, 2018, 11:34 am

      Hi M.IRANI!

      Those are some specific situation but I’ll try to help as much as I can.

      1. You should not take into account the time when the machine is not operational as that wouldn’t give you a clear picture of how much wear and tear a machine can take before breaking down. You should include any time the machine is operating, healthy or not.

      2. In those scenarios it may be best to get a MTBF index for a group of similar assets. That way you can measuring healthy assets with your problem assets and can have a good measurement of when you will likely experience a breakdown.

      3. You need to look at each device and make a total sum of the time spent to repair and restore it normal operating conditions and then divide it with the total number of repairs that the device had. This data is obviously hard to accurately track manually so we set up Limble to calculate that automatically (as long as your technicians enter the time they’ve spent on the repair).

      I hope this helps!

  • Vivek D January 15, 2019, 12:57 am

    Hi,

    I have a query here. You mentioned that we should not take into account the time when the machine is not operational. So on the same context what is your thought on Starved and Blocked times on the MTBF of an equipment ? We term it as internal MTBF while dealing with the failure analysis at Equipment failure level and not system level.
    So should we exclude starved and blocked times while doing the mtbf analysis because in these 2 states even though my machine is availiable it is not processing anything i.e its in idle state. Appreciate your insights on this !

    Vivek D – +91-9900290780
    vivek.india31@gmail.com

    • Bryan Christiansen January 16, 2019, 5:58 pm

      Hi Vivek!

      Great question.

      The answer is… it depends. In most situations I would personally not include the “Starved and Blocked” times since the machines are not being used and therefore the majority of the wear and tear is not occurring.

      That being said the most important thing is to be consistent and clear with which way you choose to measure. If one month you take include the “Starved and Blocked” times, but the next don’t you will run into inconsistent data. Therefore you should made it very clear for your company’s best practices you are or are not including the “Starved and Blocked” times.

      Another factor to consider is how other teams outside of your own consider what times to include when calculating how long an equipment is running. They may consider the equipment running even when it is in a “Starved and Blocked” time while you don’t. For example, if you tell production to expect a machine to break every x hours of running, yet they think a machine is running even during “Starved and Blocked” times you will cause confusion which will cause problems.

      I hope that helped answer your question… If not let me know!

  • Tushar January 17, 2019, 5:58 pm

    If MTTF of single processor is 10000 hours. What will be MTTF of 1000 processors? Assuming if one fails, they all fail.
    my answer is 10 hours is it correct? Explain plse aslo

  • sunny January 23, 2019, 6:51 am

    How do we determine the index factor between MTBF and MTTR

  • Mariella Tabares January 29, 2019, 6:48 am

    It is possible to calculate the MTBF and MTTR when only occurred 1 failure during the working time period of the machine?

    • Bryan Christiansen February 19, 2019, 11:36 am

      Hi Mariella,

      For MTBF, you need at least 2 failures to calculate “time between failures”. Keep in mind that basing any decision on just one small sample like this might not give you the most accurate information.

      As for MTTR, you can technically calculate it with only one failure. But again, calculating “mean” in “mean time to failure” implies that you are measuring multiple values and calculating an average.

      Another point to bring up is calculating these metrics on a group of similar machines so you can get more actionable results if the number of your data points are low.

      I hope that clears things up a bit.

  • Joko Supriyanto April 22, 2019, 12:43 am

    Hi Bryan,

    MTTR and MTBF is use for calculating every part on the machine (1 part 1 MTTR and 1 MTBF) or 1 machine (any problem part on machine 1 MTTR and 1 MTBF) ?

    • Bryan Christiansen April 25, 2019, 11:35 am

      Hi Joko, It depends on how you want to set it up. Ideally you should have one report that does MTTR and MTBF for the top level equipment and all of its parts and then also another report for each part of that machine. A good CMMS will automatically do these calculations for you so you know exactly what your MTTR and MTBF is for every granular level.

  • Rajiv April 22, 2019, 10:37 am

    we need to see the performance of the 20 equipments individually. Failure rates variying from zero failures to 5 and so on. How I can distinguish the zero failure machine performance, one failure machine performance month wise. Please help.

    • Bryan Christiansen April 25, 2019, 11:33 am

      Hi Rajiv if you are using a good CMMS you can have it automatically do those calculations for you. Other then that it would be hard to give help without looking at the actual data. Fill free to send it over to mail@limblecmms.com and we can take a quick look 🙂

  • jayakumar April 24, 2019, 1:22 pm

    thanku, this is very useful

  • MURUGESH June 2, 2019, 1:02 pm

    Hi,

    Will you suggest me how to calculate MTBF for multiple equipment in single system?
    Example: Petrol station in single city. i want to calculate MTBF of all station having same fuelling pump. How do I calculate the MTBF for whole system or Single station?

  • Luis Fernando Jaramillo June 5, 2019, 6:10 am

    Good Morning
    I was Reading your Internet page, and it’s very interesting but I have a question about how to calculate MTBF with the Formula, in this moment we calculated MTTR and MTBF every month for every machine, my question is:

    If in the last month I didn’t have any fail and the machine was working 150 hours, how I must to use the formula?

    150 / 0 = ∞

    Could you help me to clarify my doubt please?

    Thanks In advance

  • mohamed sayed July 3, 2019, 6:31 am

    I have 4 shift , how i calculate MTBF for every shift although every breakdown have adiffrent frequency for every shift.

    if i calualte MTBF / month for one shift = working hr per month / no of failures is correct or no although for every shift breakdown have diffrent frequeny of failure.

  • Aris August 17, 2019, 5:08 am

    Hi sir,

    Do you have any idea what is repair success rate? is this similar to MTTR?

  • OMAR October 16, 2019, 5:44 am

    Good Morning
    I was Reading your Internet page, and it’s very interesting but I have a question about how to calculate MTBF with the Formula, in this moment we calculated MTTR and MTBF every month for every machine, my question is:

    If in the last month I didn’t have any fail and the machine was working 150 hours, how I must to use the formula?

    150 / 0 = ∞

    Could you help me to clarify my doubt please?

    Thanks In advance

    PLEASE THIS QUES VERY IMPORTANT PLEASE ANSWER IT

    • Senko October 17, 2019, 7:10 am

      Hi Omar,

      If there wasn’t any failure you can’t calculate MTBF. I’m not sure how you use it at your organization but calculating it on a monthly basis doesn’t seem like the best approach, especially if the machine isn’t “breaking” multiple times a month.

      Looking at the overall trend for a longer period of time for every particular machine would probably give you more useful info.

      I hope that helps!

  • Agus October 30, 2019, 12:26 am

    Hello,
    Thank you for the nice introduction, we would like to adopt this failure indicator as our KPI.
    but may I ask as to calculate MTBF the Running time may effect to other non related with Machine. i.e. we are running 24Hrs = 1440 min, and there is Plan D/T for break and change model about 200min. and there is un planned downtime for change parts, set up, change tool and machine problem total about 350 min (in here the D/T is mix and ae we need to segregate?), and machine qty failure is 150 times. can I say that:
    – Running Time = 1440 – 200 – 350 = 890 min or 1440 – 350 = 1090 min ?
    – Machine Problem Qty = 150 times
    so MTBF = 890/150 =6 min 0r 1090/150 = 7.2 min (this based on single machine)
    Thank you.

    • Senko November 6, 2019, 3:02 am

      Hello,

      Since the MTBF is concerned only with operational time, you should use the first calculation (890/150).

      That being said, it might not be a bad idea to track how much of that downtime is planned and how much is unplanned as that will give you a greater insight on what you need to improve or spot potential issues.

  • Alex Gan March 25, 2020, 10:56 pm

    Hello,
    I would like to verify about MTTF and Predictive Maintenance tracking were both similar to calculation for the spare part Usual life cycle?

    Beside there commonly machine maker have the schedule for the replacement of consumable components/ usual life cycle, how we priority the Lean cost saving vs, TPM practical ?

    Pls. advice

    • Senko March 27, 2020, 9:19 am

      I would love to help you Alex but I do not really understand the question. If you could explain it in more details, that would really help.

  • venkat April 15, 2020, 12:34 am

    Hi,
    Can we use MTBF for daily monitoring of machine/equipments ? and if there is no failure in machine, how to calculate MTBF.

    • Senko April 15, 2020, 6:15 am

      Hello,

      MBTF is probably not the metric you would want to use for some kind of daily monitoring. You will be better off making a PM plan based on OEM guidelines and schedule preventive work and inspections accordingly (more info here: https://limblecmms.com/blog/the-ridiculously-simple-preventive-maintenance-plan/)

      If you do not have any failures, then you can’t really measure MTBF 🙂

      I hope this answers your questions.

  • BIKASH MOHANTY April 29, 2020, 1:19 am

    Hii sir,

    some query.
    1 – How to define MTBF a product run time 100days ?
    2- How can improved MTBF a prouct service life ?

    • Senko April 29, 2020, 3:08 am

      Hi Bikash,

      1. To define MTBF for a product, you need to know how much time did the product spent “in use/operating” for those 100 days and how many times did it fail and then you can use the formula given in the article.
      2. MTBF can be used as one of the methods to help you predict when the product might fail so you can schedule preventive maintenance before that happens. The fewer breakdowns an asset experiences, in many cases, the longer the lifespan of that asset.

      Not sure if I got your questions right though, I hope I did!

  • Andrew May 6, 2020, 3:32 am

    Hello
    Thank you so much for this write up. It’s really helpful.
    How can a PM plan be built for a daily monitoring machine?

    • Senko May 7, 2020, 1:07 am

      Thanks Andrew 🙂

      We actually have an in-depth guide on how to create a PM plan. It includes all the steps you need to now, gives examples, and you can even download a helpful checklist.

      I hope that helps!

  • danny townsend May 20, 2020, 11:23 am

    very insightful.
    I am working on calculating the maintenance ratio for two items, discrete numbers for field maintenance and sustainment maintenance. I am focusing the math one a count of one, versus the total number of fielded items. do you think this is the right approach? The MTBF I am using is a projected number, since this is a relatively new piece of equipment.

    • Senko May 21, 2020, 5:43 am

      Thanks for commenting Danny.

      It seems like a pretty specific example. I’m afraid my knowledge of the subject matter doesn’t extend that far so I wouldn’t be comfortable giving you advice that might set you on the wrong path.

      It does sound like you know what you’re doing though 🙂

  • G Bhattacharjee May 28, 2020, 12:06 pm

    Very well explained.

  • manu verma June 13, 2020, 11:40 am

    As we all know that theoretically, MTBF = MTTR (repair) +MTTF (failure) but in your article under MTBF section, there is a figure which shows MTTR and MTBF are two different phases. why?

    • Senko June 15, 2020, 1:47 am

      MTBF does not include the time you spend repairing the asset, you only take “working” hours into account for calculating MTBF.

  • Jason David Carroll June 14, 2020, 11:42 am

    use some simple math:

    10,000/1 = 10,000MTTF
    (10,000(item1)+10,000(item2)+…+10,000(item1,000))/1,000=10,000MTTF

  • ahmed June 16, 2020, 9:48 am

    if there are 10 line of sight microwave links are arranged for a radio receiver. The time availability at the last receiver in the system (due to propagation alone) was found to be 99.82%. Use the most conservative method to find the time availability assigned to each microwave link. Justify how the method is conservative. thanks

    • Senko June 18, 2020, 6:35 am

      That sounds like it was taken from a quiz 🙂

      I’m afraid we can’t help you here, but there’s always a chance that other readers of the article help you out!

  • Saravanan Baskaran June 17, 2020, 11:05 pm

    Hi Sir,
    What is the bench mark for MTTR, MTBF and MTTF for used equipment say 10 yrs old.

    • Senko June 18, 2020, 6:43 am

      That vastly varies from one piece of equipment to the other, from working conditions, from how the equipment is handled, etc.

      I’m afraid your best bet is to look at respective OEM manuals and try to see if you read any relevant info between the lines (or maybe try to google these benchmark for the machines in question).

      Good luck and let us know if you find the answer – to help other readers that might be in the same situation!

  • Tom W. June 22, 2020, 3:42 pm

    Great intro and thanks for sharing. I have a question I’m hoping you can help with, or at least point me in another direction. I have 100 devices that I’m testing and each one has been continuously operating for a different amount of hours. I’ve had 4 fail after operating for 538, 583, 810, and 829 hours. Is the MTTF calculated as (538+583+810+829)/4 or is it all cumulative hours of the 100 devices divided by 100 even though this includes a bunch of devices which haven’t failed yet? My other question is can this data be used to predict what percent of devices will result in failures at a defined time (after 1000 hours for example)? Thanks for your help and insights.

    • Senko June 24, 2020, 3:59 am

      Presuming that we are talking about the same type of device, I would go with your first suggestion “(538+583+810+829)/4”. As more devices fail, you will have a bigger sample and you should be able to use that data to predict how long does a piece lasts on average.

      I’m sure there are some statistical equations you can pump MTTF numbers in to try and predict things like which percentage of devices will last for 1000 hours.

  • Tom W. June 23, 2020, 7:38 am

    Hello – excellent summary, thank you! A couple of clarification questions: 1. I have data for 134 devices that show how many days they have been operational. There are 4 in there that failed at various times. I’m trying to figure out the correct way to calculate the MTTF and not sure if I do the calculation using only the 4 devices that failed, or if I use the entire data set even though there are devices in there that have yet to fail. I’m also wondering if I can use this data to predict some kind of reliability – something like X% of devices are expected to survive to 500 days, Y% to 1000 days, etc. Can you point me in the right direction on this?

  • Tom W. June 24, 2020, 8:36 am

    Thank you! This is helpful and much appreciated!

  • Ashish Batajoo July 4, 2020, 12:25 am

    Thanks for sharing!!! is there any formulae for calculate the service availability depending upon the MTTR MTBF?

  • Saket Fule August 1, 2020, 1:41 am

    It was a mindclear explanation. I was reading through books but honestly I couldn’t understand anything. You’ve reduced my trouble. Thank You so much sir and keep making such nice article.

  • Maouchi Madina September 1, 2020, 5:53 am

    Hi , thank you for the summery. i have a question about the number of failures used to calculate MTBF and MTTR .I am in the ambiguity regarding wich stops are considered as failures. Exp : does stop more than 60s considered as a faillure? even overloading , setting, control….ext ?
    How to set targets of MTBF , MTTR ?
    Thank you in advance.

    • Senko September 2, 2020, 4:10 am

      It is on you to define what you will count as a failure. As a rule of thumb, if it is not something that requires the attention of the maintenance team, I would not count that as a failure.

      The stops you mentioned seem like they would play a bigger role in calculating OEE than these failure metrics.

      Regarding targets, that depends on the type of device/equipment you’re looking at. Sometimes those numbers might come with the OEM manuals so you know what to expect. Otherwise, you can measure the metrics and then set targets for improvements you think are realistic depending on your available resources.

      I hope this helps 🙂

  • Hanne September 23, 2020, 1:25 am

    Hi,
    Thank you for the summary! I still have some questions:
    1. How do you calculate MTBF over more than 1 device? Just sum the different operating hours and divide it by the number of failures? So for example Machine 1 operates for 200hours and fails, machine 2 operates for 300hours and fails, machine 3 operates for 150hours and fails –> (200+300+150)/3
    And thus it depends if the machines are set in serial or parallel?

    2.And what happens if you have multiple machines, but one of them never fails, do you include them into the calculation of MTBF? For example machine 1 runs for 500hours but never fails, Machine 2 runs for 200 hours and fails and machine 3 runs 100 hours and fails. Does MTBF gets calculated as (500+200+100)/2 or as (200+100)/2

    Thanks in advance!

    • Senko September 23, 2020, 2:38 am

      Hi Hanne,
      1. You can do that if you want to find an average for the exact same type of device. I would avoid doing that if devices work in different operating conditions as than I do not know how much value would you get from that average number (just something to keep in mind).
      Series vs parallel discussion depends on what you want to measure exactly, MTBF of the whole system, machine, or machine components. If it is the system, then the serial vs parallel configuration should give diferent results.
      2. I could see an argumeent for using both. Including “500-hours” into the calculation would give you a more accurate average MTBf on such a small sample (with a note that you recalculate it when it eventually fails). If you have a bigger sample, I would not include it in the calculation.

      At the end of the day, as long as you understand the context so you do not missread the final number, you’ll be good to go.

  • Mitchell Stockton September 24, 2020, 8:17 am

    Where can I find “typical” MTTF values for components? It appears that noone publishes “true” values?

    • Senko October 2, 2020, 2:40 am

      If the OEM manufactures don’t include the numbers (or include misleading ones), I’m not sure if you can get “true values” without testing it for yourself. Maybe try to find other managers that used them and see what was their experience like.

  • Hanne September 25, 2020, 2:56 am

    Thank you very much! Is there actually a standard on how to calculate MTBF for mechanical equipment?

  • hicham October 2, 2020, 3:29 am

    Salam Omar,

    from my point of view you need to calculate from your last failure until the moment you report your KPI, e.g
    01/01/2020 first start of the machine and the machine works for 390 H then failure at 31/01/2020 , after repair for 10 hours, works for 350 H in 29/02/2020 then the machines works for 390H until 31/03/2020 then no failure and works 390H until 30/04/2020

    for the MTBF that you will report will be as below :
    Jan/2020 MTBF = 390H
    Feb/2020 MTBF = 370H (390+350)/2
    Mar/2020 MTBF = 565H (390+350+390)/2
    Apr/2020 MTBF = 760H (390+350+390+390)/2

    and so on for the others months

Leave a Comment

Your email address will not be published. Required fields are marked *

*
*

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.