BanditKeith
Mechanical
- Aug 9, 2012
- 11
Hi Everybody,
I have recently become a reliability expert in training at my company (old reliable, they're gonna call me). Although I have a fairly decent grasp on the majority of the basics, there's a few minor things regarding the basics that I could use some clarification with:
1. I know that one of the key parameters in reliability is the mean time before failure:
MTBF = Timespan/Number of failures
I feel really dumb asking this, but how far exactly do you go to classify a "failure" for reliability purpose when calculating MTBF. I know that a failure is when the equipment is unable to perform its intended function, which can range from a component needing to be replaced to resume operation (bearing failure, seal failure, etc.), up to the equipment not performing to its intended design conditions/output (not enough pressure, not enough flow, etc.). I guess electric breaker trips and start-up issues are considered failures as well since they render the equipment unusable unexpectedly, but are you interested in these as well from a reliability view (assuming they're part of the system your analyzing). Even if something like a valve needs to be replaced or a suction screen for a pump needs to be cleaned, are they classified as failures if they are a part of your system and if the equipment needs to be isolated/pulled out of service to fix or replace? In a nutshell, are you just interested in the physical failures as the ones previously mentioned, or is any unplanned shutdown/downtime of an equipment (including breaker trips, inability to start-up, valve replacements, screen cleaning, etc.) considered when calculating MTBF? I would like some clarification simply because, obviously, the MTBF heavily depends on it.
2. I know quite well that there is a difference between availability and reliability, but I know lots of people can get the two confused. I know reliability at it's most basic form is defined as:
R(t) = exp(-t/MTBF)
where t is the mission time, and MTBF is the mean time before failure from before, and F is the failure rate. However, I've seen a few websites, a colleague, and I believe somewhere in this forum that reliability is defined by:
R = Uptime/(Uptime + Unplanned downtime due to failure)
However, this second definition seems more like availability instead of reliability. Is the second definition closer to availability, or can it describe reliability if you just look at the downtime due to failure? My gut says this is availability and not reliability.
I appreciate any help/clarification you all can provide me with. Thanks
I have recently become a reliability expert in training at my company (old reliable, they're gonna call me). Although I have a fairly decent grasp on the majority of the basics, there's a few minor things regarding the basics that I could use some clarification with:
1. I know that one of the key parameters in reliability is the mean time before failure:
MTBF = Timespan/Number of failures
I feel really dumb asking this, but how far exactly do you go to classify a "failure" for reliability purpose when calculating MTBF. I know that a failure is when the equipment is unable to perform its intended function, which can range from a component needing to be replaced to resume operation (bearing failure, seal failure, etc.), up to the equipment not performing to its intended design conditions/output (not enough pressure, not enough flow, etc.). I guess electric breaker trips and start-up issues are considered failures as well since they render the equipment unusable unexpectedly, but are you interested in these as well from a reliability view (assuming they're part of the system your analyzing). Even if something like a valve needs to be replaced or a suction screen for a pump needs to be cleaned, are they classified as failures if they are a part of your system and if the equipment needs to be isolated/pulled out of service to fix or replace? In a nutshell, are you just interested in the physical failures as the ones previously mentioned, or is any unplanned shutdown/downtime of an equipment (including breaker trips, inability to start-up, valve replacements, screen cleaning, etc.) considered when calculating MTBF? I would like some clarification simply because, obviously, the MTBF heavily depends on it.
2. I know quite well that there is a difference between availability and reliability, but I know lots of people can get the two confused. I know reliability at it's most basic form is defined as:
R(t) = exp(-t/MTBF)
where t is the mission time, and MTBF is the mean time before failure from before, and F is the failure rate. However, I've seen a few websites, a colleague, and I believe somewhere in this forum that reliability is defined by:
R = Uptime/(Uptime + Unplanned downtime due to failure)
However, this second definition seems more like availability instead of reliability. Is the second definition closer to availability, or can it describe reliability if you just look at the downtime due to failure? My gut says this is availability and not reliability.
I appreciate any help/clarification you all can provide me with. Thanks
