Technology, culture changes drive improved maintenance
An audit completed recently at the utilities section of a large U.S. paper mill revealed that overtime was accounting for 38% of maintenance activities. Work orders were backlogged and costly reactive (corrective) maintenance was rampant. Employees were spending so much time ‘fighting fires’ just to keep the boilers operating that they didn’t have time for most of the maintenance tasks scheduled to prevent the very equipment failures they were battling.
Unfortunately, this is the rule in a great many industrial plants %%MDASSML%% not the exception. The sad irony is that maintenance supervisors in most plants have no time to implement the very technologies that are supposed to remove the overtime burden from their employees. Personnel in these plants are in a continuous state of anxiety trying to solve problems as they arise, with precious little time left for other maintenance.
Figure 1 depicts three types of maintenance organizations. The size of each disc reflects the relative cost of maintenance in the different types of organizations. The largest disc represents those organizations that achieve only “adequate” maintenance. The majority of their time is devoted to corrective maintenance (CM), which typically requires large amounts of overtime that simply devour maintenance budgets. The ‘better’ organizations do more preventive maintenance (PM) and even a little predictive maintenance (PdM), while the ‘best’ employ a variety of methods. The disc on the right symbolizes the lowest cost producers. They save money by limiting the amount of time spent on expensive corrective maintenance.
So, how do those groups on the left of the chart in Figure 1 begin to eliminate the chaos of reacting to emergencies and move toward becoming a low-cost producer?
Computerized Maintenance Management Systems are obviously not the whole answer. Process industry companies have spent millions of dollars on CMMS only to be disappointed. The workforce rarely has time to respond to all the PMs and work orders generated by the CMMS databases. Sometimes, they simply sign off without doing a task just to clear the paperwork jam %%MDASSML%% a dangerous precedent.
Technology alone is not the answer. Too often, companies purchase a diagnostic technology with the notion that it would solve their problems and save money. Unfortunately, diagnostics rarely are integrated into daily maintenance programs, and the expected benefits never materialize. Worse yet, much of this technology is simply gathering dust on a shelf. People are so busy fighting fires that they have no time for technology.
Most maintenance personnel are forced to react to what’s going on around them. They do preventive maintenance when they have time, because it makes them feel as though they are preventing something bad from happening %%MDASSML%% but the failures keep occurring anyway.
The asset optimization process
The asset optimization process offers a way out of the woods, beginning with an identification of business objectives and key drivers. This is followed by determining what is actually happening within the maintenance organization and designing new work practices in the form of failure defense plans. When such a plan is implemented and the necessary changes occur, greater equipment availability, increased plant throughput and lower maintenance costs are sure to follow. Finally, a periodic review measures performance against the benchmark to be sure the changes remain in effect and the improvements continue.
Most maintenance employees work hard and respond quickly when something breaks. Given enough time, they’re able to fix most problems, but they don’t do much proactively to prevent those problems in the first place. Few maintenance departments are willing or able to change the fix-it-when-it breaks culture by changing the way they operate.
It often takes knowledgeable outsiders, using a systematic, well-tested and rigorous plant assessment and benchmarking methodology, to uncover the weaknesses in such an organization. For example, a hard look at the amount of overtime the workers are asked to put in can be very revealing. While 10% is a world-class number, anything greater than 20% should raise a red flag. If overtime averages more than 30%, there are simply too many emergencies.
An in-depth assessment involves evaluating an organization on some 20 factors, including work planning and execution, management systems, leadership, goal making, accountability, diagnostics technologies in use, methods of improvement, internal communications, training and others (Fig. 2). It might take a knowledgeable team two or three weeks to complete such an assessment, but the result is a comprehensive benchmark against which improvements can be measured and the basis for development of a failure defense plan.
The failure defense plan serves as a roadmap to getting the plant back on track. It provides the justification and basis for implementing an appropriate blend of reactive, preventive and predictive maintenance. And it identifies the technical basis for each maintenance task. The failure defense plan incorporates asset prioritization, work process optimization on the top 20% to 50% of critical assets and the application and integration of diagnostic technologies into the work processes.
Since no plant has time to do all the PM that equipment manufacturers call for (mostly to protect themselves), it’s essential that all maintenance be prioritized so the most important work gets done at the right time. This includes addressing repair procedures that are appropriate but may be too frequent. The worst thing you can do is tear down a piece of good equipment just because some periodic schedule calls for that kind of service. Too-frequent teardown can introduce the possibility of failure that did not exist previously.
In establishing priorities, every asset in the plant is evaluated for its importance to safety, the environment, production, cost and product quality, and each asset receives a criticality ranking from 1,000 to zero, with 1,000 as the highest priority. Interestingly, we find that those assets with a criticality ranking in the top 20% to 30% cause 70% to 80% of the problems, reflecting their importance to productivity. Still, critically important machinery receives no more attention in many plants than other pieces of equipment, but when that key part of the process breaks down, problems ensue.
Obviously, the heaviest maintenance activity should be focused on equipment with the highest criticality ranking. Generally, the lower 70% to 80% on the criticality scale can be left for preventive maintenance, with many assets at the bottom of the list allowed to run-to-fail.
More bang for the buck
Reliability-centered maintenance implies that all equipment with a high criticality rating is subject to work processes developed to address any conceivable failure mechanism that could occur. In the real world of process manufacturing, the possibility of many of these failure modes ever occurring is minimal, and developing work procedures to address each one would be extremely time consuming and costly.
To get the biggest bang for the buck, streamlined reliability-centered maintenance is implemented. Maintenance requirements are evaluated on the basis of probability that a particular failure mode will actually occur. This allows users to concentrate on tasks applicable to restoring or maintaining the inherent reliability of a piece, and each task is justified.
Finally, existing technology is evaluated to determine what may be applicable, which ones might be added and what kind of maintenance %%MDASSML%% predictive, preventive or reactive %%MDASSML%% should be employed. For example, a turbine with a high criticality ranking will need continuous online vibration monitoring, periodic oil analysis and ongoing thermodynamic analysis to ensure its continuing performance. Key process control valves may need digital valve controllers with predictive diagnostics, whereas less critical valves can be monitored with advanced diagnostics controllers, and some valves need not be monitored at all.
Predictive programs take advantage of available technologies to provide information on the condition of equipment in the field. Then, when signals are raised that the performance of a certain piece of equipment is slipping, maintenance managers and schedulers determine how long that equipment can be allowed to run before taking it out of service for repair. Their decisions depend on knowledge of how important the equipment is to the continued operation of the plant and its current condition. Predictive maintenance is based on informed decisions regarding when to take action to prevent unexpected downtime.
A plan like the one just described can be fairly costly, but it is virtually worthless unless put into action. That’s why the asset optimization process requires a transition from planning to implementation %%MDASSML%% the period during which the cost is amortized while performance and plant availability rise.
Most maintenance departments simply won’t be able to change the way they function overnight; there will still be emergencies to settle. They will need guidance in the adaptation of the asset optimization process, but given a chance, it will work, and they will slowly make the transition, reducing their dependence on reactive maintenance while increasing predictive activity. As that happens, overtime hours will drop and the accountants will notice a return on the investment in change.
Remember, asset optimization depends on a combination of technology, expertise and the right work processes. Users can buy technology and have people trained with the necessary expertise, but unless a culture change is made in the way things are done, chances for real improvement are remote.
Technology is here to stay. Those who embrace it and make it work will enjoy increased plant availability and productivity. Those who don’t have time for technology will be mired in the past, wondering why they can’t get ahead of their competitors even though their employees are working so hard and putting in so much overtime.
|Mark Nord is a 1990 Mechanical Engineering graduate of the University of North Dakota. He is currently a Plant Web Services Engineer with Emerson Process Management’s Asset Optimization Division. He has worked for Emerson for 10 years, and for the past six years has been responsible for helping customers implement reliability-based maintenance programs with a focus on Control Valves and Instrumentation. Nord is located at the Fisher Valve Division facility in Marshalltown, IA. His e-mail address is Mark.Nord@EmersonProcess.com .|
The Bottom Line…
A variety of maintenance strategies need to be employed on the plant floor to achieve optimal success.
Finding time to implement technological solutions is difficult when a manager faces urgent maintenance issues.
Emerging maintenance technologies are tools to assist the maintenance staff in identifying problems that need immediate attention.
Changes in culture and technology won’t happen overnight, but like any maintenance program, they must be scheduled and monitored to ensure success.