How condition-based monitoring reduces downtime

In a recent blog, Schneider Electric considered the latest trend in condition-based monitoring technology for rotating equipment. In particular, the urgent need for industrial businesses to find a better solution than traditional motor monitoring solutions.

Reducing costly downtime is an industry-wide challenge. Average downtime costs for industry are estimated to be between £20k and £35k an hour. But for some industries like automotive manufacturing, this can be as high as £35 a minute. These costs relate to revenue losses, recovery costs and, in some cases, penalties and fines based on service level agreements or regulations.

Electric motor failures are a common cause of unplanned downtime. Rotating machinery driven by low or medium voltage AC induction motors represent most industrial applications, from oil and gas, to mining, marine, airports, and logistical centres. Beyond motors, rotating equipment can include pumps, compressors, conveyors, blowers or fans, rolls or mills, etc.

Condition-based monitoring

With 20 to 25 percent of electric motors being critical to operations, and a typical annual failure rate of up to 7 percent, motors are having a large impact on downtime and losses. Besides, any needed repairs are often done during operating hours, causing further downtime. So, finding ways to improve efficiency can make a big difference to a company’s bottom line.

For these reasons, organisations must move to a more predictive maintenance strategy, to help avoid unscheduled repairs and downtime. Furthermore, it improves efficiency, reduces costs, and extends the lifespans for rotating equipment. Doing so requires the support of condition-based monitoring and predictive analytics.

Predictive versus traditional motor maintenance

A variety of maintenance strategies are commonplace in the industrial environment. If a facility team uses a ‘run-to-failure’ method, they perform minimal or no maintenance on a motor until it fails completely. The organisation then must accept unplanned downtime as part of its regular operations.

When using a preventative strategy, maintenance takes place at set intervals, either by calendar or running hours. The aim here is to achieve a level of availability based on mean-time-between-failure statistics. Yet this may mean maintenance is either too late, following a costly failure, or too early, incurring unnecessary operational expenditures. Neither of these strategies consider the actual condition of the motor.

A better approach is predictive maintenance, with work performed only when the motor needs it, i.e., when performance falls-off or a failure is predicted. But this approach requires continuous monitoring of the motor condition. If possible, the technology should detect risks at an early stage. Examples of risk conditions include bearing degradation, rotor or coupling eccentricities, mechanical unbalance, stator winding looseness, pump cavitation, harmonics disturbance, or axis misalignment. It is important to detect risks early to reduce potential damage, minimise wasted energy and importantly, minimise unplanned downtime.

Problems with traditional motor monitoring technology

A variety of methods are available to monitor motor conditions, and each has strengths, but most have significant weaknesses. For example:

– Noise or vibrations from the surrounding environment can fool vibration sensors

– Oil and vibration analysis are not able to spot electrical problems

– Acoustic sensors are sensitive to background noise and interference from other objects

– Thermal cameras may be sensitive to ambient temperature and the thermo-optical properties of the objects under observation.

For all the above technologies, sensors must be on or near the asset under observation. This means they are not usable for motors in inaccessible or remote locations. Additionally, most sensors need power from a  wired source or by a battery requiring periodic replacement. Finally, these types of sensors can suffer damage if used in harsh conditions.

A new way

An established technology that is again coming to the fore is motor current signature analysis (MCSA). When combined with AI-based technology, it compares the electrical signals feeding motors with conditions against a library of data fingerprints. Its advantages include simpler installation in a wider range of applications as it does not need installing at the motor’s location.