Probability of Failure (PoF)

Table of Contents

For any business, the ability to accurately assess and predict expected failure rates for equipment in their plant is vital for proper maintenance planning and performance optimization. Understanding what Probability of Failure (PoF) is, how it’s calculated, and its implications on preventive maintenance schedules can help organizations stay ahead of failure events before they occur. In this post, we will explore PoF from all angles so that you have the insight required to make the best decisions about maintaining your machinery.

What is Probability of Failure?

Organizations that rely on equipment use the metric, probability of failure, to schedule equipment maintenance and replacement. Probability of failure makes up for the fact that vital components may not be easily visible for visual inspections, meaning that you have little insight to anticipate asset failure. Additionally, it might be challenging to forecast when a breakdown will occur, even when the crucial component is obvious.

Types of Failure

PoF behaviors typically fall into one of three categories:

  • Failures at Random – The probability of random failures is constant. Examples include sophisticated electrical assets and electronic components.
  • Aging Asset Failures – Failures due to deteriorating assets as they age or are used more frequently cause a rise in PoF.
  • Infant Mortality – While it’s unlikely, there is a probability that brand-new parts will fail. Maybe the probability is only one out of 1,000, but it is important to include this possibility.

Collecting Probability of Failure Data

To calculate probability of failure, you need data. This includes:

  • Data from both inside and outside the company
  • Information from manufacturers and other sources
  • Knowledge from your team’s experience and other experts
  • Conditioning monitoring data
  • Predictive models
  • Weibull failure distribution analysis data

Asset managers can more accurately predict the useful life of assets with the use of a thorough asset management history, such as that found in a CMMS (Computerized Maintenance Management System). Your organization’s CMMS will greatly assist in collecting and analyzing the data necessary for the probability of failure calculation. It will be a resource of information based on your actual usage, maintenance, and application of the asset.

While it is impossible to precisely predict when an asset will fail, a CMMS, however, can assist in identifying the assets that are nearing the end of their useful lives. With a combination of the insight from your CMMS and condition monitoring efforts, you will have the best estimate of the effective life currently available.

How to Calculate Probability of Failure

To calculate probability of failure, use the following formula:

PoF = 1 / MTBF = R / T

MTBF is the mean time between failures, R is the total number of failures, and T is the total amount of time.

If you’d like to calculate the probability of failure per hour, you can divide the total number of failures by the total number of hours. For example, a machine with 5 failures and a run time of 3,000 hours will have a PoF of 0.0017 or 0.16% probability of failure per hour. The failure rate per hour in this instance is so negligible that it is almost irrelevant.

The Steps of the PoF Process

There are three basic steps in the PoF procedure:

  1. Keep a record of the asset’s implementation date.
  2. Assess and record as accurately as possible each asset’s effective life.
  3. Regularly monitor condition! If you have limited resources, prioritize condition monitoring for the most important assets first.

Nevertheless, there are numerous things to keep in mind along the way. Calculating probability of failure entails:

  • Determining the system’s most vital component(s), such as the compressor in an air conditioning system.
  • Figuring out how old the important component is. The serial number easily reveals the age of the complete equipment, but you must also track when you replace important components.
  • Estimating the expected life based on one’s own experience, historical data, or statistical data.
  • Assigning a general probability of failure. For example, low, medium, or high.
  • If necessary, modify the probability of failure in light of your facility’s conditions. The operating environment of equipment can greatly impact its life cycle.

A Vital Metric for Asset Management

Calculating and understanding probability of failure is vital for asset management. That’s because asset performance or condition is typically correlated with the probability of failure. Tracking this metric will increase your awareness of different modes of failure, your asset’s conditions, and the reliability and accuracy of your asset data. Only quality data will provide an accurate estimate of PoF.

4.7 Star Rating