“Gathering machine data through the use of an automated system can help facility managers plan and foresee production disturbances far more accurately. This allows them to predict failures before they occur and plan maintenance schedules and parts replacements, accordingly.”
SMB Bearings MD Chris Johnson encapsulates the vital role that effective condition monitoring can play in ensuring the downtime associated with equipment failure can be minimised. And indeed the part that automation and digitalisation can play in making this still more effective.
It’s a point reinforced by Sulzer’s data analytics and smart technologies manager Matthew Anderson, who emphasises the need for his industrial customers to gain better insights into asset behaviour in order to proactively plan maintenance across their production facilities.
As always this is achieved by measuring parameters such as bearing unit temperature and vibration to assess asset condition.
Traditional manual methods though, have their limitations, namely:
- Manual measurements are taken infrequently, problems are not identified early
- Not all equipment runs 24/7; single measurements may give the wrong picture
- Safety aspects: personnel must access installation areas that might include a safety risk
In effect, summarises Sulzer, the best the manual approach can do is to gather snapshots. And if the results prove effective, the chances are that is simply down to chance and lucky timing. Not the safest foundation for making key decisions that may ultimately impact the business balance sheet.
Automated and digitalised products now offer a widening range of refinements to and applications for condition monitoring.
Atlas Copco is one of many companies addressing the challenges in the pumps and compressors market. Its Smartclamp solution helps overcome the lack of built-in controllers, explains business line manager of Atlas Copco compressor technique service sales and marketing, Paul Clark.
“With Smartclamp, no extra energy source is needed for operation, as it is simply attached to the incoming power cord to harvest energy from the machine while it is running.”
Designed for easy installation, the product garners insights on the running hours and service status of equipment rated up to 30 kW.
Gathering machine data through the use of an automated system can help facility managers plan and foresee production disturbances far more accurately
Chris Johnson, MD, SMB Bearings
The in-built permanent SIM chip enables remote data monitoring via 3G smartphone connection to the equipment installation, enabling the Smartclamp to upload running hours from the compressor/dryer/pump at specified intervals. Data extracted from condition monitoring can subsequently be employed to identify when a need for the next service is approaching.
Vibration measurement remains a key component for monitoring. Contrinex’s analogue inductive sensors make use of this especially in the measuring of components in assembly jigs and the measurement of steel threads inside rubber belts, judging the wear in seals as two parts slowly move towards each other – used in one case for a hydro-electric as well as the more commonplace assembly jigs.
However, condition monitoring specialists Sensonics point out that, while the advantages of applying vibration condition monitoring to rotating equipment is well established, the technique can only provide its fullest benefit if one can be certain of applying the correct measurement techniques.
For example, there may be a need to make a change to the measurement to optimise early fault detection. Whilst industrial vibration sensors offer a broadband performance suitable for most machines, the signal processing is what makes the difference, says the company.
But the problem, as it acknowledges, is that too often the higher cost of more refined vibration monitoring systems providing such flexibility cannot be seen as a justifiable expense in the case of smaller machinery.
Hence the development of Sensonics’ DN26 G3 protection monitor as a standalone monitor in addition to the firm’s range of rack-mounted products. The dual channel din rail mountable unit provides monitoring of bearing vibration, shaft vibration, or shaft position with fully programmable signal conditioning and with a range of measurement algorithms and sensor options.
The unit is designed for optimum flexibility; as a universal module (single hardware platform) it is field upgradable and fully programmable. Special protection options include slow rotation vibration monitoring down to 0.2 Hz and narrow band filter measurements for specific vibration signatures.
Cost effectiveness and enhanced protection is a powerful selling point, not least when engineering experts must seek to persuade colleagues with less technical knowledge but a greater influence upon budgetary accountability.
Schaeffler’s recent condition monitoring intervention for a utilities client enabled significant cost savings and also avoided more wholesale investment.
With many of the company’s network pumping stations unprepared for an instant response to floods, it was deemed essential to prevent unexpected failures of the drives or the pumps themselves. Schaeffler’s answer was to install condition monitoring devices on pumps and motors to provide advanced warnings of impending bearing failures.
It recommended installing six CONCEPT2 automatic systems for lubrication of bearings in the motors and pumps. The system provided two lubrication points independently of each other and withstood pressures up to 50 bar.
Schaeffler also recommended use of nine SmartCheck condition monitoring devices to oversee the trio of pump/motor units. Two SmartChecks were installed per unit to monitor the upper and lower motor bearings, plus a third situated on the pump bearings.
Deviations or changes in vibration behaviour were detected by SmartCheck and reported via the customer’s control system, thus allowing incipient bearing damage to be detected early.
On average, plants might monitor key assets such as pumps and heat exchangers every one to two months. Lengthy gaps increase the opportunity for unexpected failures to occur, with direct impact on efficiency and downtime.
Emerson’s AMS Asset Monitor overcomes the problem of supervising remote and difficult-to-access machinery, as well as adding the opportunity to access predictive analytics.
Plants are always looking for ways to improve profitability by increasing productivity. A percentage point or two can equal millions of dollars per year or more
John Turner, product manager for online prediction, Emerson
This edge analytics device digitalises essential asset data and analytics for better operations performance and improved decisionmaking. It provides actionable insights into essential assets that were previously monitored only with infrequent assessments.
In contrast to devices transporting data to a historian or the Cloud, the AMS Asset Monitor provides the analytics at the edge, making its calculations at the device.
Data collection is continuous, using embedded logic to identify and diagnose issues. Individual problems such as reliability imbalance, misalignment, bearing faults, lubrication issues or fouling are consolidated into an overall asset health score.
“The AMS Asset Monitor enables personnel across the plant to see the current health of essential assets along with suggested actions to improve asset health. This allows them to make informed decisions to maintain reliability, increase uptime and maximise productivity,” says John Turner, Emerson product manager for online prediction.
While the future of monitoring – and by extension predictive maintenance – lies with technology that is digital, connected and analytical, Sulzer’s Dr. Marc Heggemann and Seth Tate caution that this needs to be aligned with clear goals and objectives.
“Advanced analytics solutions that aim to improve business performance can only be deemed successful if they deliver a measurable benefit to safety, the environment or business finances,” says the duo.
“Digital solutions are a value enabler, giving operations the opportunity to take the best action based on the provided intelligence.”
Pumping systems, for example, can vary greatly by design and cost, they emphasise. While mathematical models learn from historical data to identify similar patterns, in some cases, pump failures may be infrequent and most operators do not have the history of operation and maintenance readily available in a digitised format.
Sulzer uses unsupervised machine learning techniques, where the models are trained with the recent operational history of the pumps together with physical pump modelling.
The company cites the success of its complementary approach to asset performance on behalf of a solar power installation, with its Blue Box system operating alongside the in-house data science team on analysis and anomaly detection.
Emerson’s Turner concludes: “Plants are always looking for ways to improve profitability by increasing productivity. A percentage point or two can equal millions of dollars per year or more.”