Warning Signs: How to take advantage of data from condition-monitoring tools
A plant engineer’s ability to diagnose, detect and monitor equipment condition issues is advancing all the time, thanks to ongoing developments with vibration, thermography (infrared), oil analysis and ultrasound tools, just to name a few.
So once you have all the fancy new tools, do you know how best to take advantage of them?
We’re here to help. Along with the sophistication of the tools available, ways to synthesize and integrate data so that maintenance teams can make immediate use of it and also monitor trend issues over a period of time are also progressing. PEM asked leading technology providers to share the latest in their condition monitoring tech developments, how best to integrate them, and where the future is headed.
Over the last few years, infrared cameras have improved significantly in terms of resolution and now come with more options as well, says Paul Frisk, manager of the Infrared Training Center in Burlington, Ont. (the training arm of infrared camera-maker FLIR Canada Ltd.). “Infrared cameras now have the ability to incorporate wireless data from digital clamp meters and other instruments and make that all available at one glance,” he explains. “Some cameras now available immediately generate a single-page report. This summary can be transferred for printing and archiving by download to an office computer or through wifi to a plant’s CMMS system.”
Frisk says the primary value of an infrared camera is in its ability to initially determine whether a device is working properly or not while it’s running. “With some other diagnostic tools, you have to shut down the device, which obviously impacts production,” he notes. However, as with many types of detection and monitoring technology, there are misconceptions about what infrared cameras can provide.
“From watching movies and TV, people think infrared cameras can allow you to see through walls, water, etc., but they only measure released infrared energy,” he explains. “A properly trained thermographer can determine temperatures from infrared readings using conversion factors, knowing the material and so on, but infrared cameras cannot overcome the physics of all materials under all conditions.” He also stresses that infrared images can easily be misinterpreted, and proper training is absolutely necessary.
In addition to using handheld infrared cameras and connecting them with your plant’s CMMS, standalone infrared cameras can send data to the process PLC (programmable logic controller). “Based on the camera’s readings, things like process speed, fans or heat can automatically be adjusted if the material needs to be kept at a certain temperature,” Frisk notes.
With regard to the future of infrared condition monitoring technology, he foresees more improvement in resolution and smaller camera size, along with a continued drop in cost.
Ultrasound instruments have changed a great deal over the past decade, according to Alan Bandes, vice president of marketing at UE Systems. Analog detectors, which required manual entry of test results for basic trouble-shooting, have been replaced by software-driven digital systems capable of analyzing trends and reporting on a wide range of operating conditions. Newer models offer things like sound analysis, cameras, non-contact infrared thermometers, and even touch screen controls. “There are a lot of professionals that haven’t looked at ultrasound technology closely and view the instruments as basically leak detectors,” Bandes says. “Others feel, incorrectly, that ultrasound is too subjective, which is often due to experience only with older analog units.”
Bandes says it’s very easy to integrate ultrasound technology into plant processes. “Due to the sophistication of on-board software and external supportive software, users can create routes, establish baseline information and upload and download route data,” he explains. With baselines set, the software can notify personnel with low-level alarms (for example, lubrication starvation) or high alarms (failure) through headphones or other means.
Some instruments provide inspectors with the option of opening up a spectral analysis screen to analyze bearing faults, gear mesh issues and electric emissions while in the field. Recorded sound samples can be played in real-time and viewed with an image of the spectral screen. “This feature is very useful for electrical emissions as well as mechanical operations,” he notes.
Software associated with ultrasound instruments can provide specialized reporting for things like steam traps, valves and bearings. “Regarding leak surveys, downloaded test results can be converted into reports that provide important information for cost analysis and greenhouse gas emissions,” Bandes says. Regarding the future of machine monitoring by ultrasound, he believes “we are only limited by the software we can develop.”
More vendors now supply in-plant oil analysis sensors and the means to communicate with those sensors. “It’s no longer necessary to rely solely on a lab for analyzing oil samples to determine fluid condition,” says Darren German, Bosch Rexroth national service manager. “In the plant, we can now get real-time results on of oil cleanliness (particle count), water content and temperature when sensors are coupled with a data acquisition device.” These devices can record and track trend parameters in real time for any given time period, but German cautions maintenance teams that monitoring equipment should be considered as a compliment to a bottle sampling program; reports from an oil analysis lab still provide the most oil condition information. The role of monitoring equipment is to provide additional protection between bottle sampling periods, he says. “If, for example, a heat exchanger ruptures and releases water into the oil the day after a bottle sample was taken,” he notes, “this will likely go unnoticed until production stops if there is no oil analysis sensors in place.”
The many oil-monitoring systems on the market range in complexity and price. “Some of the data acquisition systems also provide the ability to add a threshold or alarm which will signal the moment the results vary from a ‘baseline normal,’ ” he says. “We suggest that before investing, you should understand what it is that you want to accomplish — what parameters are important to monitor.” He recommends that maintenance groups consult with their engineering groups prior to purchasing a system, as the ability for a machine to communicate with a sensor often already exists within the machine HMI.
German predicts that down the road, the capacity to measure reliable viscosity and TAN (total acid number) will be developed, along with a sensor that can measure the amount of air in hydraulic fluid. “ ‘Smart’ sensors and wireless sensors are often mentioned as coming down the pipe as well,” he says.
Advances over the last few years in sensor, recording, and analysis technology have put vibration analysis within the reach of even small companies, says John Bernet, product and application specialist at Fluke Corp. “Easier measurement procedures (triaxial sensors), combined with vibration diagnosis programs (expert systems) now enable maintenance teams with minimal training and experience to use vibration to evaluate machine health and determine required maintenance,” he notes.
Bernet says vibration can identify problems before other symptoms, such as heat, sound, electrical consumption and lubricant impurities, are detected. “Measuring the vibration of motors, pumps, and other common machines can reveal valuable information about machine health or impending failures,” he notes. “However, instead of focusing on the patterns of the hundreds of faults that vibration analysis can reveal, we should focus on the four most common mechanical faults: imbalance, misalignment, wear, and looseness.” He adds that studies have found that many vibration analysis programs don’t collect all the data needed to make an accurate diagnosis — to diagnose machine condition correctly, vibration data is needed from all three axes of a rotating shaft.
The key to automating vibration analysis, he notes, is to compare new data with data from a similar machine known to be functioning properly. Automated diagnostic programs perform a sophisticated analysis, comparing hundreds of data points with the fault patterns of similar machines to give easy-to-understand results.
Bernet foresees that the benefits of vibration analysis will be expanded to the entire plant in future. “A plant’s reliability team can use high-end analysis programs on the few complex machines, while the maintenance team can use simple diagnostic tools on the basic machines,” he says. p
Treena Hein is a freelance writer based in Pembroke, Ont.