by Michael Walsh, GE Energy


Do not let a catastrophic event be your wake-up call to adopt a proactive strategy.

All too often, we have found the tipping point to predictive maintenance is a costly breakdown of motors, pumps and related systems, or worse, a serious catastrophe that not only damages equipment and cripples your operations, but impacts employee safety as well. According to a recent U.S. Department of Energy study, 55 percent of those responsible for industrial plant maintenance admitted to characterizing their program as “reactive” and 31 percent as “preventive” only. It does not have to be that way.

It is established in the industry that predictive, rather than just reactive or preventive maintenance of existing equipment, will likely save money in the long run and can also help prevent the development of serious hazards leading to a safety problem. Whether an organizatin is a pharmaceutical facility using small 25-horsepower motors or an oil and gas plant operator requiring a 60,000-horsepower synchronous machine, the same precepts apply: “Take care of it now or pay later.”

Predictive and Preventive Maintenance—You Need Both

Predictive and preventive maintenance are different, but both are complementary and one should not be conducted exclusive of the other. They each can help protect equipment and people.

Predictive maintenance is a process that is custom-designed for your specific system, built out of regular observation and recordkeeping to understand trends and uncover anomalies. End users can, therefore, leverage this historical data to take future actions to optimize their operational efficiency.

Preventive maintenance is similar to following the maintenance directions in an auto manual—such as when to change the oil, when to check the belts, when to rotate the tires, etc. Most original equipment manufacturers (OEMs) rigorously test their equipment for a battery of conditions that ensure peak performance in many applications. Following included operating and maintenance documentation is always advisable.

Reaching out to the OEMs of major pump platform components may also be a good idea. In many cases, they will have deep engineering expertise, application performance knowledge and global experience that your team could use.

With larger companies, a good knowledge base often exists in-house on most system components. However, in many of the best organizations, this knowledge is refreshed regularly with instruction by acknowledged industry experts with deep domain expertise. This internal expertise can also be supplemented from time to time with consulting experts in advanced diagnostics and troubleshooting technologies. Diagnostics can make all the difference in the world, and used in a healthy predictive maintenance program, will catch problems before they severely impact operations.

A few OEMs may have deep technical knowledge on multiple components of equipment and could be made available as a consultant on the finer points of condition monitoring instrumentation and diagnostic services for monitoring machinery vibration. Casting a wider net for knowledge of the system components will help develop a firm foundation upon which a truly predictive maintenance program can be built.

Economics and Safety

The primary outcomes of predictive and preventive maintenance can have a real impact to an end user's bottom line. These measures, too, can provide savings based on the avoidance of downtime, damage to equipment and employee wellbeing.

Many risk studies use similar numbers to illustrate the inherent advantage of adopting a more proactive maintenance approach. They can also be used as a template to uncover the resulting costs in operations to craft a more realistic model.

Consider that a reactive maintenance strategy would likely contain up to 14 percent risk, which equates to $140,000 of yearly maintenance on every $1 million worth of existing assets. Compare this to a predictive maintenance strategy, which would contain less than half the risk, about 6 percent, which equates to $60,000 of yearly maintenance per $1 million of existing assets. That is a difference of $80,000 per year. End users may find that these resulting savings will easily pay for a predictive program.

Then consider the other savings not mentione—such as unplanned downtime, injured workers and strained customer relationships. The business reasons that justify this path become more evident as the real costs are investigated.

Through experience, making adjustments now—perhaps investing money now—is better than waiting for a disaster to happen and paying ten-fold from having personnel injuries, line stoppages and equipment replacement.

Some recommendations are given below to make end users' systems as fail-safe as possible. These can make maintenance managers, plant operations personnel, financial personnel and the CEO rest better at night.

Based on sound industry practices and experience, a comprehensive proactive maintenance strategy requires a system that captures repetitive failures so that appropriate corrective changes are made. This demands good record keeping of all maintenance programs and a root-cause-analysis of any maintenance performed. These records should be reviewed annually and semi-annually.

Conduct both preventive and predictive maintenance on systems. Follow the manufacturer's minimum maintenance recommendations. Regularly take non-intrusive measurements—such as vibration analysis, infrared (IR) analysis and insulation readings—and then compare these measurements so that equipment failure can be predicted. Making these predictions allows maintenance and production departments to work together to schedule repairs.