<img src="" style="display:none;">

Intelligent Lubrication Solutions™

Intelligent Lubrication Solutions™

fluidsmatter_header.png

Measuring Machine Reliability: 5 Key Metrics

Posted by Matt Mohelnitzky on Jan 5, 2017 12:46:07 PM

reliability1

Machine breakdowns are a serious hassle. Not only do they often require costly repairs and replacement parts, but they also prevent you from meeting your production goals. As Plant, Maintenance, or Reliability Managers know, it can be frustrating when a machine breakdown causes you to go over budget and miss your benchmarks or goals—but what are you supposed to do about it? How are you supposed to see the future and predict the health of all your machines?

Here’s where machine reliability metrics come in handy.

Machine reliability metrics help manufacturers measure the probability that their machines will perform without failing for a specified period of time. While this calculation isn’t based on any single calculation or number, there are several key performance indicators that managers can use to measure machine reliability and its impact on operation performance. For reliability data to be useful and actionable, however, you have to consider what the data means in context of operating cost. 

One way that managers can reduce the operating cost of their plant while maximizing their efficiency and preventing machine breakdowns is by using the right lubricant for their machines’ operating conditions. The incorrect choice and use of lubricants is said to account for 43% of all machine failures, and as the “life blood” of machinery, lubrication procedures are a critical factor in maximizing your machines’ reliability.

To help you more accurately forecast your machines’ reliability for your plant’s production and efficiency goals, here are 5 key metrics to track.

1) Mean Time Between Failure (MTBF)

Defining the “mission time” is key to reliability numbers. In other words, how long is the machine expected to perform without failure? MTBF describes the expected time between two failures for a repairable system. In order to maximize your machine reliability, a long MTBF is required. To calculate MTBF, all you have to do is divide the total time your machine is in service (a.k.a. sum of uptime periods) by the number of failures in that timespan. For example: 

15 years of service / 2 failures = 7.5 years MTBF

2) Failure Rate

Similar to MTBF, failure rate refers to the frequency with which a component fails. In fact, failure rate and MTBF are often interchangeable; it really just depends on how you prefer the data to be expressed. To calculate failure rate, simply divide the total number of failures by the total time your machine is in service during that span. For example: 

2 failures / 15 years of service = 1 failure per every 7.5 years

3) Exponential Distribution and Projected Reliability

Exponential distribution is the most basic and commonly used reliability prediction formula for machines with a constant failure rate. Since most industrial machines spend the majority of their lives in the constant failure rate, it is applicable for many companies. However, this calculation is more complex than just failure rate or MTBF. To calculate the exponential distribution (a.k.a. the probability that a piece of equipment will fail in a given amount of time), you must use the base of natural logarithms and failure rate to find the reliability estimate for a given period of time. Here’s what the calculation looks like: 

R(t) = e λt
R(t) = Reliability estimate for a period of time (t)
e = Base of natural logarithms (2.718281828)
λ = Failure rate (1/MTBF)

Now, let’s say you want to figure out the projected reliability for your machine after 5 years of use using the failure rate from above. Here’s what you would get:

R(5) = 2.718281828 (.13333 x 5)
R(5) = .51346 = ~51%

In other words, after 5 years of use, the likelihood that you’ll have operated without a failure in that timespan is 51%. Likewise, you can also assume that 49% of identical equipment operating in an identical environment can be expected to fail in that same timespan (projected failure). Of course, realistically, any machine could fail on Day One while another lasts 20 years; that’s just the nature of probability projections. Just make sure you use the same unit of time when calculating failure rate and the projected reliability (e.g., if failure rate is expressed in years, calculate the projected reliability for years as well).

For manufacturers looking to take this calculation one step further for perhaps a more accurate projection, you can calculate the projected reliability of individual machine components, and multiply them together to find your projected reliability. For example, if your machine has 5 main components, 4 of which are 99% reliable and 1 of which is 75% reliable, your projected reliability calculation would look something like this:

Projected Machine Reliability = .99 x .99 x .99 x .99 x .75 = 72%

4) Machine Availability

Though machine availability and reliability might seem similar, they are not the same thing. In fact, machine availability translates your failure rate to a measure of operating costs and profitability. In other words, failure rate and MTBF measure time, but not money. By calculating the availability, we can calculate the cost of that downtime as a consequence of reliability. To calculate this, divide your downtime by the total time your machine was in service (sum of downtime and uptime). For example:

24 hours of downtime / 240 hours in service = 10% downtime, 90% availability

5) Operating Cost of Downtime

Once you’ve calculated your machine’s availability, you can finally calculate the operating cost of downtime. To do this, first calculate the total number of service hours your machine would have if it were running at 100% availability. For example, let’s say 100% availability means operating 12 hours/day for 260 weekdays each year, or 3,120 hours/year. If your machine only averages 90% availability, however, the total hours of operation decrease to 2,808 hours/year, losing 312 hours/year in production. Once you’ve determined how much production time machine downtime costs you, you can multiply the lost time by your production expectations to arrive at your lost annual production and corresponding revenue. For example, let’s say you say expect to produce 100 tons/hour and can sell each ton for $500:

312 hours of lost production x 100 tons/hour of production x $500/ton = $15,600,000 operating cost of downtime

Now, remember when I said 43% of all machine failures are attributed to incorrect lubrication choices? Let’s imagine that 43% of your lost production hours can be attributed to improper lubrication (312 x .43 = ~134 hours). Plugging those lost production hours into the equation above, it wouldn’t be too much of a stretch to say that improper lubrication choices cost your company as much as $6,700,000 per year!

Thankfully, many lubrication problems can be easily avoided with the proper preventative maintenance procedures and a comprehensive lubrication management plan, but manufacturers have to take the initiative and be proactive with their lubrication strategy. Thorough lubricant analysis helps provide critical early warning information indicative of machine failure, which impacts reliability, availability, and operating costs.

Now that you know some key performance indicators for measuring your machine reliability, tune in for our next post, where we’ll talk about what to do with that data and how to go about making adjustments to improve reliability.

To learn more about how U.S. Lubricants can help your company maximize your machine uptime, contact Tony Springer at TSpringer@uslube.com or by phone at (800) 490-4900 ext. 8823.

CTA-Shearglide_905_case_study

Topics: Machine Reliability