My first introduction to equipment reliability came in 1988 at Eastman Kodak Company in Rochester New York. Jerry Haggerty, one of the founding members of SMRP (Society of Maintenance and Reliability Professionals) had begun to assemble a steering committee of Kodak Maintenance professionals who would begin working together sharing information on equipment reliability. As a member of this team I remember well the emphasis Jerry put on understanding the P-F curve and the P-F interval. Jerry knew that if we could get our managers to understand the P-F curve we could begin to make the transition into predictive technologies and reduce the amount of Reactive Maintenance being performed at our plants.
Nearly twenty years later, I can now say with confidence that Jerry was somewhat correct. Understanding the P-F curve, the P-F interval, and the scheduling of predictive technologies is fundamental in building a sound PdM program. The understanding of the P-F curve as most of us know it will help a maintenance or reliability manager to sell the need for predictive technologies such as vibration analysis, lubrication analysis, ultra-sound and IR inspections. This will also, if properly implemented, reduce the amount of Reactive Maintenance being performed at your plant. What the original P-F curve will not do is maximize the benefit of PdM program.
Below is the P-F curve most people are familiar with. The x-axis of the curve represents Time (T) or Operating Age, and the y-axis represents resistance to failure. Starting at the top left part of the curve and moving right we encounter point P, known as Potential Failure. This is the point in time that, when using some form of Predictive Technologies, one can first detect resistance to failure. As we continue to move right along this curve, resistance to failure continues to fall until we encounter point F, known as Functional Failure. This is the point in time when the components resistance to failure has deteriorated to a point where it can no longer perform its intended function. The time elapsed between point P and point F is known as the P-F interval. The value of knowing the P-F interval of a component for a specific failure mode is that we now can set the interval of the condition based (PdM) inspection. In setting the interval we should now, with a high level of confidence, be able to detect the failure of this component, plan a replacement or restoration task and repair the component before the failure occurs. In doing so, we have now replaced what once was a reactive task with a PdM task.
The introduction of the P-F Curve and on-condition PdM tasks provided a much needed innovative change into a world where Preventive Maintenance was viewed as the only option to avoid emergency/demand maintenance. Excitement around the P-F Curve quickly evolved the world of maintenance into a new age of “proactive maintenance” for companies who could afford the new and costly predictive technologies associated with Predictive Maintenance. Companies who invested in technologies such as Vibration Analysis, Lubrication Analysis, and Thermographic Analysis paid large sums for the equipment and training to develop in-house Predictive Maintenance groups and not long after making these investments began sharing stories of success and the savings that could be generated from detecting failures, and the costly secondary damage associated with emergency maintenance.
Today, it would be no stretch at all to make the connection between the P-F Curve, Predictive Maintenance and the birth of SMRP, ReliabilityWeb.com, IMC, and Marts. As word on Predictive Maintenance spread around the world, PdM course offerings and PdM service providers exploded making the new technologies both more attractive and affordable. By 1995, if you were benchmarking your company’s maintenance organization, you could not be world class if you were not involved in Predictive Maintenance.
Something is Missing
Several years after working with Jerry Haggerty I left Eastman Kodak to form Reliability Solutions, Inc. a consulting firm specialized around training people in reliability tools and measures. As part of our company services we also offer on-site consulting and mentoring and it was at one of these on-site visits that I learned the P-F curve was incomplete.
Our client in the past two years had invested a substantial amount of money to develop a Predictive Maintenance Program and while they were quite proud of that program, they also revealed that they were at the same time disappointed that it was not delivering the savings at a rate they had hoped for. Our clients PdM service provider collected data from hundreds of pieces of rotating equipment around their plant they prepared reports for our clients that showed nearly all of their rotating equipment was in the process of failure (someplace between points P and F). As trained, our client would then open a maintenance work order to schedule replacement of the asset prior to failure. Using precision alignment techniques, the rotating equipment would be replaced and to their extreme disappointment months later they would again be informed that the same asset was again failing. The P-F curve detailed below represents our clients experience with Predictive Technologies and the P-F Interval. Notice the saw tooth effect that designates each time the asset is replaced or repaired.
While the Saw Tooth P-F Curve still effectively eliminates running costly rotating equipment to failure, it can lull maintenance managers into the illusion that PdM is all maintenance has to offer regarding these types of failures. While one could celebrate that this company successfully detected and responded to three potential failures over a short period of time and avoided the costly secondary damage associated with each failure, I would want them to question why each failure occurred. The most important things we need to understand about the P-F Curve and the Saw Tooth P-F Curve is this:
Detecting potential failure is simply not enough today to consider your PdM program a success. For each detected potential failure, we must also determine the specific cause of failure. We need to know: what has caused this potential failure and most important, can this cause be eliminated?
If we ask and answer the above questions our maintenance organization, PdM service providers and our company are ready to gain the full benefit of the Modified P-F Curve utilizing not only Predictive Maintenance but Pro-Active maintenance techniques and Reliability Tools.
Failure Modes that Result in the Saw Tooth P-F Curve
- Soft Foot
- Pipe Stress
- Lack of Lubrication
- Improper Lubrication
- Lubrication Breakdown
- Undersized Foundations
- Improper Belt Tension – Too Tight/Too Loose
- Over Torque of Electrical Connections
- Dirt/Dust/Moisture Contamination of Electrical Connections
- Improper Sized Wiring, Overloads or Heaters
- Improper Torque of Piping Connections that Result in Leaks
- Improper Gasket Materials
- Improper Design or Application
While each of these failure modes could be detected using some form of Predictive Technologies, and then corrected prior to total failure, if the failure mode is not properly identified the failure will occur again.
The above should clearly highlight the need take your Predictive Maintenance program a step further by asking the above listed questions each time a component has been determined to have reached point P. In doing so we can now pinpoint the specific cause of each failure and use RCM decision logic and sound Proactive Maintenance techniques to eliminate these causes and the saw tooth effect.
Completing the P-F Curve
As we performed an RCM Blitz™ analysis of several assets at our clients facilities it became clear why some were not having the success they had expected from their PdM program. In working with their PdM service provider to set up their PdM program they had simply generated a list of assets for each specific technology, the list generated for critical assets, set up PdM routs and intervals for each asset based on the providers recommendations. In most cases, Vibration Analysis and Airborne Ultrasonic tasks were performed on a monthly basis, Theromgraphic inspections were set up on a quarterly basis, and Motor Current Evaluation was performed every six months. Not one single PdM inspection detailed the failure modes the tasks were looking to detect. While we all understood the P-F curve and the P-F interval we failed to understand or determine why the assets were failing over and over again.
While the technologies our client had invested in were successfully detecting failures, our client had never asked the service provider why some assets continued to fail over and over again. This is where the addition to the P-F curve comes in. In the representation below note the difference in this P-F curve. Starting at the far left at point I (Installation) and moving right we have a very long flat line going between point I and point P (Potential Failure) this is what we call the I-P interval. The I-P interval represents the time it takes move from the point of installation to the point where Potential Failure is first detected. The objective of all world-class maintenance and reliability organizations should be to work to maximize the I-P Interval. This can only be achieved through a thorough understanding of your assets, proactive maintenance techniques and reliability tools. In viewing the P-F curve in this manner it became clear to our client that a large percentage of the failure modes they were detecting through the use of predictive technologies could in fact be identified and eliminated using RCM and proactive maintenance techniques. As an example, one of the failures our client was seeing over and over again was on a blower that was mounted to an undersized foundation. Each time they replaced the blower they used precision alignment to ensure the blower and motor sheaves were properly aligned. Without proper foundational support, continued stopping and starting of the blower over time resulted in misalignment and degradation of the blower and motor bearings. In performing the RCM Blitz analysis of this asset we listed all of the probable failure modes for the blower and determined that the blower base and foundation would need redesign to eliminate the failure mode. The result – a blower that had failed three times in eighteen months has not failed in over four years.
In completing the P-F curve we have identified several proactive maintenance techniques and reliability tools that can be used to extend the I-P Interval. These tools and techniques are listed left of the modified P-F curve.
The Value in Understanding the Modified P-F Curve
While many companies and maintenance organizations around the world have seen the value in understanding the original P-F curve, I want you all to understand the additional value provided by our Modified P-F Curve. To do this we start at the far right end of the P-F Curve at the point of complete failure (where the P-F curve contacts the x-axis). Moving from here back to the left and up to point F (Functional Failure), this interval between Functional Failure and Failure is the interval where Reactive Maintenance takes place. It’s the area if time where this piece of rotating equipment starts smoking, shaking, stinking, and squealing. As a result we quickly send someone out to shut the asset down so that it can be replaced. Performing maintenance in this area is costly and minimizes maintenance effectiveness to less than ten percent.
Moving back to the left and up from point F we encounter point P, this is the well known P-F interval, the time frame where Predictive Maintenance (PdM) is employed. The value of performing maintenance here is we can detect failures are in the process of occurring, then plan and schedule repair or replacement to minimize equipment damage and reduce operations down time. Performing maintenance in P-F interval provides a cost benefit that increases maintenance effectiveness to as high as fifty percent.
Finally, we now move left on the P-F curve from point P (Potential Failure) back to point I (Installation). The I-P interval is the time frame from installation (I) to potential failure (P), this interval should take years to elapse provided the correct Proactive reliability tools are employed and precision maintenance techniques and tools are used at installation. Performing these Proactive Maintenance techniques will provide a cost benefit that increases maintenance effectiveness to one-hundred percent!
To reach this level of effectiveness one will need to understand how Proactive Maintenance Techniques and Reliability Tools can increase the I-P interval of your assets.
Understanding Proactive Maintenance Techniques and Reliability Tools
While it would take a full text book be able to completely explain the value of each technique and tool, I will list each out here and provide a summary of how each can extend your I-P Interval.
Reliability Centered Maintenance – RCM is a Reliability Tool that uses a structured team approach to analyze a process or piece of equipment. In performing a RCM analysis your team will assess all likely failure modes for the asset and develop a maintenance strategy to mitigate the consequences for each failure mode. The value in performing RCM is the proactive assessment of these failure modes and the resulting tasks developed to eliminate reoccurring failures.
FMEA – Failure Modes and Effects Analysis – Similar to RCM, FMEA is a Reliability Tool used in the design phase to identify likely failure modes. In performing FMEA your design team will discuss these failure modes and attempt to design out failure modes that result from poor design and installation decisions.
The Five Rights of Reliability – Design it right, purchase it right, build it right, operate it right and maintain it right. An overall reliability program focused on educating employees at all levels and organizations on the importance of reliability. The five rights of reliability develops a reliability plan across engineering, purchasing, construction, operations and maintenance that clearly describes how each business unit can improve reliability.
Select Supplier Agreements – Often a part of your reliability plan, select supplier agreements should be made consulting engineering, operations, maintenance and purchasing. These agreements should be developed using your company’s reliability data while working with suppliers to provide the most robust and reliable assets. Inferior parts or components are a common cause for reoccurring failures.
Requirements Documents – If this is not part of your companies capital design and engineering program it needs to be. Requirements documents are binding agreements written to ensure the highest level of reliability in design and installation. As an example many companies now have requirements documents written for the acceptable level of vibration on start up of new rotating equipment. The document will clearly state what that acceptable measure will be and the resulting action taken if the requirement is not achieved. Again, the intent of these documents is to eliminate failure modes inherent to poor design or installation practices.
Design standards – Company design standards should always be used as a tool to improve equipment reliability. Used in combination with select suppliers, and requirements documents, design standards will help your company ensure all new installations are safe and reliable. Some examples of design standards that will eliminate reoccurring failure modes; standard mass requirements for pump foundations, standards requirements for piping supports, standards for starter panel installations.
Precision Alignment and Balancing – Precision maintenance tools known for increasing the life of rotating equipment. While these tools have been available for several years few of us have taken advantage of their use. Precision alignment and balancing will both dramatically reduce vibration that results in reoccurring failures of bearings, seals and couplings.
Installation Standards – Used for both new installation and maintenance, these standards are put in place to ensure proper craft skills are used when working on equipment/assets. Some examples of installation standards would be the identification of the proper type and grade of flange hardware and gasket material. Developing installation standards eliminates reoccurring failures such as leaks caused by using incorrect gasket material.
Torque Specifications – While almost everyone working in maintenance knows what a torque wrench is and what torque specifications are, they are seldom used. Leaking connections, loose rotating equipment, and sight glass failure are often the result of improper torque. While using a torque wrench and following the specifications may take more time the resulting reliability will increase your I-P Interval.
Precision Tools – If you want to ensure proper maintenance and installation practices your people will need precision tools to do the work. As I work with companies I can quickly assess the level of understanding concerning reliability with a quick look in the tool boxes of their maintenance people. Hammers, channel-locks, pry bars and screw drivers alone will begin to ensure reliability. Precision work requires precision tools and if your people don’t have these tools don’t expect your results to improve.
Some Things to Work On
For those of us who have been directly involved in making the cultural changes in moving from Reactive Maintenance to Predictive Maintenance, we understand these changes take training and time. These changes do not take place overnight; they come in the form of small victories as people learn these techniques do work. As we now look to move back on the P-F curve to include and employ proactive maintenance techniques, a first step in the transition should be to require your PdM Technologists to list the Failure Mode associated with failures resulting equipment entering the P-F curve. In doing so we can identify and eliminate the failures that result in failure reoccurrence or the saw tooth P-F curve.
The most effective way to enter into and employ Proactive Maintenance techniques is to become involved in RCM. Reliability Centered Maintenance is the most effective way to develop a complete maintenance strategy that includes both Proactive and Predictive Maintenance tasks, and it should be applied to all critical process equipment. RCM is the only way to quickly identify and eliminate reoccurring failures through redesign and effective preventive maintenance tasks. Correctly applied and implemented RCM will always provide an effective return on the money invested for training, analyses, and implementation.
Some simple things to think about along these same lines is the number of Failure Modes that can be detected using Predictive Technologies that could simply have been eliminated altogether had we taken the time to identify these failure modes using Reliability Centered Maintenance. While it is extremely important for all maintenance organizations to celebrate the detection of potential failures through Predictive Technologies, it is far more important for us to eliminate failures whenever possible. Failure identification and elimination offers the greatest savings and presents your maintenance group as truly a world-class maintenance and reliability organization.
In closing, I would like to thank my colleague Mr. Terry Harris for his help in providing information on maintenance effectiveness regarding the different zones associated with the P-F curve. Like me, Terry brings years of experience in the trenches of maintenance to our various courses and articles. I would also like to thank our clients who through the past seven years have worked with us to make Reliability Solutions, Inc. and RCM Blitz™ a success.