When updating equipment protection systems in a plant, it often pays to think long term. Today’s protection systems can do more than avoid catastrophic failure — they are the foundation of an enterprise-wide predictive maintenance program
Chemical process industries (CPI) plants often operate expensive, complex equipment across a number of critical applications, and maintenance teams need to make sure that equipment continues to run to prevent critical processes from shutting down. If that dynamic is not stressful enough, many of those same pieces of equipment can pose dangers. If a compressor operating at 10,000 rpm liberates a blade, the results can be disastrous to personnel, property and reputation.
Fortunately, solutions to prevent catastrophic failure have existed for decades, and plants have been equipping their most critical and dangerous equipment with basic machine-protection devices for a long time. But today, that dynamic is changing. As decades-old equipment is heading into its last rounds of life, and as competitive organizations expand and build out new operations, teams are making changes and updating systems. As they do so, they are considering a new model of maintenance and reliability for critical equipment.
A key element of the coming maintenance and reliability evolution is founded in a boundless automation mindset, as teams are no longer looking to simply protect their equipment and trip it before catastrophic failure occurs. Today’s teams know that data give them the power to improve performance, safety and sustainability, so they are looking to collect and move those data seamlessly from intelligent sensors in the field, through the edge, and into analytic and diagnostic systems in the cloud.
Instead of replacing legacy protection systems with the same basic relays that have been used for years, forward-thinking teams are instead using modernization opportunities to build a long-term roadmap to more reliable operation. That roadmap is built on a layered approach, starting with smarter protection equipment that can later be expanded with additional modern machinery- health technologies, ultimately creating a holistic, real-time maintenance program across the enterprise.
Traditional protection programs
Maintenance and reliability programs are built on protection systems for critical rotating machinery, but traditional protection systems often have significant limitations and fall short of modern needs. Most basic protection systems cannot easily tie back into a plant’s distributed control system without complex engineering work. Twenty years ago, when plants had a deep bench of experienced personnel, engineering and maintaining complex, custom connections in the face of system updates was a reasonable task.
Today, however, teams are typically leaner and less experienced. But even for teams with the expertise to configure such a connection, when experts move on, the team likely must start from scratch as they try to manage and maintain their custom solutions.
Traditional protection systems also strain lean teams with their need for technicians to travel to the control room or field cabinet to view data. In a single plant, it may be feasible for a technician to regularly travel to the equipment to collect critical data, but this procedure still has users wasting time on low-value tasks. And in a fleet of many plants, traveling to a site could mean someone crossing the continent or the globe if a local site does not have the expertise necessary to analyze data and determine root causes. These extra steps can dramatically increase the time and cost of repair.
If there is no standardization for the protection systems an organization installs across its fleet, technicians can quickly find themselves needing to be experts in dozens of different systems. Plants must also maintain an inventory of hundreds of parts from many manufacturers, further increasing cost and complexity.
Perhaps the most significant problem, however, is that most traditional protection systems are not designed to also support the predictive data necessary for providing advanced warning of developing problems. They may alert maintenance teams that vibration is high, and that protection system will likely — if properly maintained — shut off the machine before catastrophic failure occurs. By the time the system triggers, however, equipment has likely undergone significant damage, and operations have likely suffered from poor performance for some time.
Protection system flaws
As an example, for one global industrial-process manufacturer, a disparate array of legacy protection equipment across its fleet had become a serious problem. The team had hundreds of protection systems across dozens of sites, all from a wide variety of manufacturers. Over time, many of those systems became poorly maintained. Bad wiring, changed relays and loose parts on protection systems occurred at many of the sites.
Trying to maintain a depot of hundreds of parts from different manufacturers had become untenable, and each new technician added to the team had to learn a variety of different systems and techniques to be effective in the field. Reliable maintenance was simply unsustainable, and many pieces of critical equipment were running to failure.
As part of its modernization, the manufacturer chose to standardize across its fleet with a new protection system built for predictive maintenance. The system they installed provided the required powerful protection options, with the flexibility to expand configuration to provide prediction data so the team could better see what was happening on their critical devices.
More importantly for the success of their program, however, was standardization on a single system type across the fleet (Figure 1). The team now has consistency of maintenance, and the company now needs only a single supply center, with a smaller inventory of parts. This new supply system not only reduces the cost of inventory, it also helps the maintenance team more quickly learn which parts they need for any given task.
FIGURE 1. A unified machinery-health interface makes it easier for any technician to function at their best, regardless of location, helping small teams accomplish more over a larger area
In addition, the maintenance team only needs to know a single type of protection system, which reduces training requirements and mistakes in the field. Everything looks the same, so it is easier for anyone on staff to perform maintenance work, whether they are working in their home plant or traveling to another site elsewhere in the fleet. Today, it is rare that the team must return to the supply center because they brought the wrong parts or tools to address an issue.
Machinery health visibility
The maintenance and reliability team working for the aforementioned process manufacturer faced two significant problems that helped them determine the next steps in their roadmap. First, critical equipment could only be taken offline at rare times for short periods because the organization needed to continuously manufacture its product. Second, because the team was small, any outages needed to be highly organized, so the maintenance team would not go in “blind.”
If technicians were going to work on a piece of equipment, they needed to know exactly what work was required. They also needed to ensure they had parts and the right tools on hand so there would be no delays in maintenance.
Fortunately, because the organization standardized on a protection system with prediction capabilities, they were able to expand their investment with additional hardware that could be configured to acquire prediction information. Vibration data coming from the protection system, for instance, offered additional value that could help a maintenance team better predict what is happening to the equipment. For example, specific spectrum and waveform data can help differentiate among different common conditions. Those foundational data can be sent to a historian, where analysts can use it to differentiate among rubbing, imbalance, poor lubrication or other conditions.
It is at this stage where the foundation built on powerful, modern protection solutions truly begins to show its value. Once teams have predictive data available for their critical equipment, they can move those data from the historian into fit-for-purpose machinery-health software. Using machinery health software, the team can collect and store data from multiple machines in a single location, where they can use comprehensive analysis tools to assess the health of every piece of equipment quickly and easily (Figure 2).
FIGURE 2. When machinery health software collects data from multiple machines and delivers it to a centralized location, like the control room, maintenance and reliability personnel can more easily monitor the health of critical equipment
Machinery health software starts maintenance and reliability teams down the path of continuous condition monitoring. The most effective solutions provide a comprehensive health score for each critical asset in an intuitive format. Technicians can instantly see which machines need attention, and they can drill down further in the software to see the reason for the alert, alongside suggestions for intervention.
More importantly, teams can identify failing equipment well before that failure begins to cause irreparable damage, and long before it trips the protection system. Armed with this rich data, teams can more easily and effectively schedule maintenance outages, limiting disruption to production.
Bringing the plant together
When basic condition monitoring and predictive maintenance is built on a healthy foundation of standardized, powerful predictive systems, it becomes much easier to reach the next step in a maintenance and reliability roadmap, bringing the whole plant together. A powerful machinery-health solution can connect to the plant’s other predictive maintenance technologies, such as device and instrument management software, valve management solutions and more to help build the context necessary for better analysis.
Instead of simply looking at discrete data for a single system, teams can now take the vibration data, for example, from their protection and prediction systems and see it in context of operational data. With the many sensors available in the field, teams can loop in temperatures, pressures, flows and more — as well as information about changes in the process — to see what is really happening in the plant.
For example, consider a team at a chemical facility that recently went into a short-window outage trying to determine the cause of high thrust measurements on a centrifugal compressor. The team used much of its outage time examining equipment, and eventually discovered that a control valve was no longer functioning as intended. That valve failure created a compressor surge condition that was seen in the high axial-thrust measurements and high radial vibration caused by a mechanical rub.
By the time the team discovered the cause, the compressor was already damaged. As a result, during the outage the team not only had to tear down and rebuild the valve, but they also had to perform significant compressor repairs.
The protection system functioned as intended by keeping the compressor from catastrophic failure, but this was not enough to prevent equipment damage. If the team had been using a machinery health solution tied into both the vibration readings from the compressor and the signatures from the valve, they would have had the contextual data necessary to identify exactly what was happening long before they started shutting equipment down, allowing them to save time and money (Figure 3).
FIGURE 3. As fans in rotating machinery operate, the blades undergo stress caused by vibration and resonance. Machinery health solutions can help reliability teams identify excessive vibration before it causes damage
Today’s industrial equipment is relatively stable, and it is unlikely that a part like a fluid film bearing will fail on its own. As a result, it has become far more important to see a holistic picture of how every asset in the plant contributes to the health of critical equipment. Moreover, such visibility also reduces or eliminates recurring problems. If a bad valve or bad lubrication-oil system is causing bearings to fail, those bearings will keep failing until the root cause is remedied. Without a holistic view of plant health, such problems are very hard to find and fix.
Better analytics
As organizations improve their analysis on the plant level, they will ultimately want to move their data up to the enterprise level, where cross-functional teams can use it to innovate and improve performance. When a plant’s maintenance and reliability solution is built on a connected, predictive foundation — from protection to machinery health to balance-of-plant monitoring — high-level teams can more easily make critical business decisions.
For example, if a team can not only discern that a problem exists, but also identify the steps to remediation and the amount of time the system can run without creating additional costly damage to critical equipment, the organization can better decide whether to shut equipment down immediately for maintenance or wait until the next scheduled outage.
Even more significantly, however, those same data can tie into enterprise-level software to help connect operations and maintenance, and empower them with advanced analytics, automated workflows, decision support and more. Armed with such tools, teams can more easily and effectively make collaborative decisions, such as whether to extend intervals between maintenance to drive increased production to meet an emerging need.
Think lifecycle
As the time comes to update protection systems around the plant and across the enterprise, it often pays to think long term. While like-for-like system swaps will help ensure critical protections are in place, they can also cause additional headaches and are often not designed with a long-term holistic maintenance program in mind.
To meet the dynamic needs of today’s marketplace, organizations are better served by implementing a foundation of advanced protection systems, designed to be expanded into predictive maintenance, and, ultimately, an enterprise-wide holistic reliability and maintenance system. Such a solution delivers improved maintenance today, while preparing the team to continue improvement over the lifecycle of their equipment. ■
Edited by Dorothy Lozowski
Acknowledgement
All figures courtesy of Emerson
Author
Erik Lindhjem serves as vice president and general manager of Emerson’s Reliability Solutions business (pss.marcom.emerson.com). In this role since June 2021, he has been focused on driving digital transformation through plant asset management of automation assets and machinery that enables clients to reach top-quartile performance. Lindhjem joined Emerson as vice president of Reliability Solutions and Consulting, Asia Pacific in August of 2018 and was based in Singapore. He holds a B.S. degree in mechanical engineering from the University of Virginia and an M.S. degree in business administration from Wake Forest University.