Somewhere in North America today, a pump that is responsible for delivering residential drinking water through a municipal pipeline will stop working. Whatever the cause, the result will be largely the same: water supplies will be diverted, lives will be disrupted, and emergency repair funds will be spent.
The good news is that events like these may not be typical for much longer.
Thanks to a precipitous drop in monitoring technology costs and powerful new predictive analytics, it is now far easier—and less expensive—than ever to avoid pipeline, pumping and other system failures. And municipal water systems are the tip of the iceberg of beneficiaries. New stand-alone predictive maintenance solutions like these can be integrated into new and existing applications across a wide range of industries.
What a Difference a Sensor Makes
Monitoring of pumping systems and pipelines is not new. Water pressure has long been monitored by the nation’s water utilities. Regulations require midstream oil and gas pipeline companies to invest in leak detection systems.
But traditional electronic monitoring processes like these are not particularly effective at signaling operators when system failure might be imminent. That takes data on other warning signs, such as equipment vibration and temperature.
To obtain that kind of data, operators traditionally have been required to contract third-party specialists who survey conditions along a specific route. The data they collect can later be compared to previous reports to identify signs of deterioration. But because the data is collected only intermittently, it can be difficult, if not impossible, to pinpoint when or how a specific problem is triggered.
Today’s advanced remote monitoring systems can deliver all that, and more. By adding sensors to motors, pumps and other equipment, it is now possible to collect continuous, real-time data that suggests a pump is approaching an unexpected shutdown or experiencing conditions that might negatively impact process performance. Depending on the application, that data can include everything from current and voltage information, to flow and pressure monitoring.
This is just the kind of information a maintenance service company or in-house staff can use to prevent unexpected and costly equipment failure, improve performance and lengthen equipment service life.
Predictive Analytics Add Value
This advanced form of monitoring gains even more value when predictive analytics are added to the equation. Using algorithms, the data collected through monitoring can be modeled to identify patterns that help users not only intervene on imminent breakdowns, but predict future behaviors and events.
Thanks to decades of route-based monitoring, some organizations have already amassed considerable data that can be more fully exploited. Using predictive analytics, experts can use this data to pinpoint the historical causes
of past problems and identify the best next steps for impending ones.
Consider how such a system might be used on an aging large motor driving any kind of equipment. These technologies make it possible to set an alarm that alerts a user if, for example motor fan vibration exceeds an acceptable or preset value.
Today’s systems can do even more. A monitoring system can also be designed to alert the user if the reading starts to change and approach that value. With the knowledge that operating conditions have changed, maintenance crews can take corrective measures to address situations that are trending.
The algorithms that support these alerts have another benefit. The predictive analytics born from crunching enormous amounts of data is not subjective. Vibration experts can use this information to make unbiased judgements on the cause of particular problems, minimizing unnecessary or improperly focused repairs.
From Preventive to Predictive Maintenance
Thanks to real-time asset condition monitoring, the ultimate benefit of these solutions is their ability to move organizations from a preventive maintenance model to a predictive approach.
Instead of following a predetermined schedule for maintenance and repairs, organizations that use these solutions can focus instead on the actual issues that are currently or might soon be affecting system performance. A sensor that picks up on changes in vibration, for example, might lead a maintenance team to maintain a motor bearing in lieu of performing general maintenance on a rigid schedule, producing substantial cost savings in the process.
All this information can be accessed from a browser-enabled computer or mobile device at a fraction of the cost of just a few years ago. Sensors and microchips that once cost more than $500 are now available for $25 or less. The cost to transmit and store data in the cloud has also been slashed by as much as 95 percent opening the door to new ways of doing business.
Even so, an organization is wise to develop a monitoring and analytics strategy before jumping in head-first. Rather than monitor everything, it may make better economic sense to deploy these technologies strategically on mission-critical systems with the greatest value or on those that pose the greatest risk.