What are we doing to a piece of equipment? Or put another way, can we really determine what services we are actually performing on a given piece of equipment? In a large, capital-intensive organization, this is not as straightforward as it seems. Determining what we do to equipment largely depends on how well our basic maintenance systems (e.g., preventive maintenance program, planning, scheduling, work execution, etc.) are designed and institutionalized. It also depends on how well we use the data associated with these systems. Data weaknesses typically come in two forms: the data was never generated in the first place or the data may be somewhere in the computerized maintenance management system (CMMS) or predictive maintenance (PdM) databases, but cannot be readily accessed. The value of the second question stems from the fact that it is very difficult to improve current maintenance practices if one does not know what the current practices actually entail.
WHEN DATA IS NOT GENERATED IN THE FIRST PLACE – LACK OF WORK ORDER DISCIPLINE
The first case arises when work is performed outside of work orders or against blanket work orders. This is a common shortcut that is seen as facilitating the execution of work in the present, but does so at the expense of the future. The loss of information that could be useful in preventing future failures is often not a consideration because the people performing the paperwork have never seen the benefits of such analysis.
An extension on the first case is where work orders are closed without proper coding or feedback. Here, the mechanism is slightly different, but the cause -- expediency in the present -- is the same as with working outside the system. The result is also the same, the loss of work history.
Both are work order discipline issues and can be addressed through management’s setting of expectations, properly designed and displayed metrics that highlight work order discipline behavior, the visible use of work order history data and routine work process audits.
WHEN DATA CANNOT BE READILY ACCESSED – LACK OF DATA SYSTEMS STRUCTURE
Is there a consistent and well-known place in your CMMS where you can determine all the services actually performed to a piece of equipment? This does not mean those that are supposed to be performed, but those that are actually performed. In the case of specialized databases, such as those for fluid and vibration analysis data, this single point can be expanded to a short list of information sources. The point here is that equipment service information is readily accessible without special knowledge of any nooks and crannies that might have been built into the system for convenience or special purposes. This doesn’t mean just work order history, it also means the data collected on routes, such as those for lubrication, condition monitoring, contractor services, etc.
Here are a couple of examples to better illustrate this point. In the first example, data proved to be difficult to extract because the equipment register within the CMMS had been set up to facilitate the segregation and execution of work, as opposed to housing of equipment history. “Dummy” pieces of equipment had been set up to group work performed by certain individuals or work by type of task. While this may have made some aspects of work execution easier, it made any analysis of work history very laborious, if not impossible.
Another example involves a situation where CMMS data showed that vibration routes had been consistently completed on time. However, the data in the vibration analysis database showed there were pieces of equipment on these routes that had never had a reading taken. Further study revealed that some of the routes were too long to be completed in the time allotted, so the tail ends of these routes were never read. The organization believed this equipment was being monitored, but in fact, it never had been. As a result, the organization was bearing more risk than it thought it was. The ability to readily extract work history data depends largely on system setup, consistent execution to a preventive maintenance strategy, and work process discipline and feedback.
Because of all the various CMMS systems in use, this topic is highly specific to individual organizations, however, there are some general guidelines that can be offered. The foremost guideline is to set up your CMMS with the mindset of future data extraction. As straightforward as this sounds, there are many exceptions to this. The people involved in system setup and configuration often don’t have the background or the visionary agenda to keep eventual data extraction in the forefront. It is often the case that CMMS implementation supersedes the implementation of a reliability function. This results in reliability having to work with a system that was not set up for the reliability function. In extreme cases, it may be necessary to rewrite or reconfigure certain portions of the CMMS to facilitate data extraction.
CONSISTENT EXECUTION TO A PREVENTIVE MAINTENANCE STRATEGY
Organizations with the ability to make modifications to preventive maintenance (PM) content and frequency tend, over time, to end up with either low value PM, more PM than they can ever accomplish, or both. PMs that are set up in the absence of an overall asset management strategy tend to be driven by the emotional responses stemming from equipment failures. Emotional stimulus results in many “good ideas” as to what could or should be done. Unfortunately, these ideas may not be cost effective and build into the system more work than the available resources can accommodate. By applying a strategy that uses business goals and available data to balance risks with available resources, a PM program can be constructed that manages risk and actually can be executed. This is where tools, such as reliability centered maintenance (RCM), come into play. Just remember that the utility derived from RCM depends on how well the two questions can be answered for the equipment in question.
Another aspect of PM program design is load balancing. This is where the workload is matched to the available resources on a period-by-period basis. Proper PM content and frequency are of little value if the work bunches up into untenable piles. Ideally, PMs would be set up on a throughput or usage basis. However, most maintenance organizations have some constraint, such as available weekly man-hours, available truck shop bays, availability of overhead cranes, or some other specialized equipment. These constraints often make it advisable to set up PMs on a calendar basis so the work can be performed with the available resources.
The aspects of adhering to an asset management strategy and applying load balancing dictate that the ability to alter PM content and frequency be confined to a limited population equipped with the proper training and skills. This is what really defines the reliability function within the organization.
WORK PROCESS DISCIPLINE AND FEEDBACK
As indicated, work order discipline is necessary for the capture of equipment data. Let’s expand on this concept to the entire work management process. Having well-designed PM content in place does not add much value unless it is consistently applied. In the case of PM, it is better to be consistently wrong than to be inconsistently right. This is because consistency allows performance verification, inconsistency does not. To give a simple example, if you believe changing the oil in your car’s engine every 3,000 miles has a positive effect on engine life, you can only validate this if you indeed change your oil every 3,000 miles. It’s this type of thinking that gives rise to metrics, such as PM compliance. There are different forms of this metric, but typically it is a ratio of how many PM work orders were due and actually completed within a period over the total that were due within the period. There are often penalties included for things, such as uncompleted PMs from prior periods, the lack of manpower, or resource charges to closed PMs. Given this, PM compliance is really a work process discipline metric. If an organization does not have good work process discipline, it is impossible to have legitimately good PM compliance. Without good work process discipline, it is very hard to answer the second question.
Another aspect of work process discipline is the utilization of feedback. In this context, feedback relates to the work order work steps. If the work process does not have enough discipline to utilize feedback from work execution, then the execution will drift from the documented work steps. People don’t like to perform work if they feel it is poorly thought out, inefficient, or ineffective. Over time, they will do what they believe is right and not bother with what is written on the work order. The result is a loss of control over what is being done to the equipment. One of the primary drivers of this behavior is generic job plans that are so broadly applicable as to not be really applicable at all. While it is not possible to utilize all feedback, if for no other reason than some will counter the asset management strategy, it is possible to acknowledge all feedback with honest dialogue. This makes people feel included and also spreads knowledge as to why the asset management strategy is the way it is.
THE PATH FORWARD
A lot of information has been presented here so far, but what’s of use to you? Well, that depends on where your organization is in its development. Any bells that went off during the reading of this article might be clues as to where to go next. If just a few bells went off, then maybe you have a higher performing organization and are ready for some of the higher end tools, such as RCM. If however, so many bells went off as to create a confusing racquet, then consider the following.
While addressing both questions is necessary for good asset management, from a value-added perspective, it may be beneficial to focus on one or the other. If your business is currently process constrained, meaning you can sell all you can make, then focus on the first question in the area that constrains the process. The first question is more throughput centric, while the second question is more cost centric. With process constrained, the cost of lost production typically dwarfs the cost of maintenance. However, if your business is market constrained, meaning you have excess capacity, focus on the second question. Just don’t lose sight of the fact that both questions are important and neither question is 100 percent throughput or cost centric.
If your maintenance processes are either immature, weak in the basics, or you just don’t know where to start, condition monitoring and delay analysis offer fertile ground. Both deal with how the equipment is failing and both add value in the near term. The addition of value is critical because it builds political capital that can be traded for time or resources.
Condition monitoring is a big topic that should be explored in earnest. However, here are some considerations.
Start with vibration analysis. It’s okay to start with a contractor, but if at all possible, build in-house capabilities and work towards bringing the program in-house. This helps build understanding and thereby faith in the technology. When an issue is found, get it planned, scheduled and executed. Get the work order properly closed with the work history captured. This will help build the work management processes. Then estimate the cost savings or avoidance, taking into account both maintenance and production losses. Keep these figures conservative, but realistic, and then publicize heavily. This is where the political capital is created. Keep moving, but don’t move faster than you can build competence because wrong or missed calls on equipment issues cause setbacks.
Next, take on fluid analysis. This includes lubricants, fuels and coolants. Find a qualified and competent lab that not only can handle the volume you are likely to generate, but also offers coaching and guidance on program development. Get some members of your team adequately trained in how to read test results and in understanding the underlying meaning of the values and the limits of the various tests. As with vibration analysis, broadcast the findings and value. Use every avenue and occasion to build and spread understanding of these technologies and their value.
The remaining major condition monitoring technologies include thermography, passive and active ultrasound, and motor current analysis. Where you go next depends on the nature of your equipment and a value evaluation.
Delay analysis is as big a topic as condition monitoring. While covered in Part 1 of this series, it’s important to re-emphasize that there is always a story in the data and it’s your job to find it. Don’t let imperfect data hold you back, but rather strive to understand and work with the data’s limitations. Blend data analysis with interviews. There is no substitute for face-to-face discussion with the people who operate and maintain the equipment. This communication serves two purposes: it validates or invalidates the results and it shows people that the data is being used. Used data breeds better data. Do not allow yourself to completely discount the value of the data. Doing nothing because the data is perceived as being inaccurate does not advance anything. As with condition monitoring, act on the findings and then broadcast the value.
Achieving good equipment reliability is a journey without end. However, no matter where your organization is along this path, the two questions should always be in play. If your efforts cannot be linked directly back to the two questions, then maybe you should question what you are doing. After all, it’s all about the equipment.
Peter Chalich, CRL, is a Professional Engineer and Owner of Equipment in Motion, LLC and Senior Associate of Caravel Solutions, LLC. He has over 25 years of experience in industrial process reliability, including 7 years as Director of Equipment Reliability at Cliffs Natural Resources. Mr. Chalich also served as a Maintenance and Reliability Systems Consultant for 12 years with The Sinclair Group, Global Performance, Day & Zimmerman and Fluor.
Keep reading...Show less
“R.A.I.” the Reliability.aiTMChatbot
You can ask "R.A.I." anything about maintenance, reliability, and asset management.