The value of advanced maintenance techniques

For a number of years, most industrial sectors have been in harmony with the concept of preventative maintenance. In order to have the best results in preventative maintenance, work is planned based on predetermined time intervals, events or indicators during operation, taking into consideration the age of the equipment and the manufacturer’s recommendations. As a matter of fact, preventative maintenance is the predetermined maintenance. Nevertheless, this time-based approach probably does not represent the actual condition of the equipment and can lead to maintenance work regardless of whether it is necessary or not.

For example, in a packaging factory, a new machine is installed for packing larger items. The needs of the production line impose on alternating between the new machine and the pre-existing one. After the first three months, the new equipment has packed 5,000 items. The manufacturer of the newly introduced machine proposes changing the ball bearing every three months or after every 15,000 packages. So, just three months after installing and minimally using the machine, an expensive change of bearings is necessary.

This is an example of preventive maintenance.

Let’s take another crack at it. Assume the same machine, but in a factory that has implemented a maintenance program that regularly checks the state of the new asset in combination with a series of sensors. After six months and at least 19,000 packaged objects, a new alert pops up on the maintenance personnel task list proposing that the bearings should be changed before the machine packs another 1,000 packages.

This is an example of predictive maintenance. As we can see, predictive maintenance notifies the maintenance personnel in advance and before there is a risk of damage to maintenance work. It provides a time window to schedule a maintenance activity at a convenient timeframe, making the whole process more adapted to the specific use of the equipment. Predictive maintenance has the actual equipment status at its core.

This happens using a combination of sensors, operational characteristics and metrics. These characteristics are categorized and based on prediction algorithms, the manufacturer’s guidelines. Faults that begin to manifest are detected early on and the company is given enough time to plan the maintenance. This allows engineers to fix the damage before it even occurs. 

Vibration sensors that measure engine motion and detect mechanical errors

Vibration analysis is one of the most common predictive maintenance techniques for rotating equipment. Through the use of appropriate instruments, the condition of the equipment is monitored and potential internal errors are determined, measured and quantified for each component. In that way, a critical failure can be avoided while extending the life cycle of mechanical equipment. 

Vibration analysis can cover a wide range of industrial equipment such as motors, gearboxes, agitators, compressors, pumps, fans, blowers, bearings, grinders, hammers or presses. Any rotary equipment can be monitored using vibration analysis:

• Detection and monitoring of bearing wear
• Imbalance and axis misalignment
• Detection and avoidance of resonance
• Pinpoint mechanical damage to joints, bearings etc.
• Identify other malfunctions such as: poor lubrication, soft foot, broken rotor, pump cavitation 

Replacing just a single part reduces the cost of maintenance, the work required for repair and most importantly, the necessary production halt time. Quite often, unplanned shutdown time can cost thousands of euros per hour, while a planned maintenance approach can take into account critical downtime and schedule repairs in non-productive times.

Current analyzers monitor the status of the electrical components of the system

Motor Current Signature Analysis — MCSA — is a technology that is used to dynamically analyze and estimate the operation of an electrically-powered drive. It helps evaluate the state of stator / rotor windings, static and dynamic eccentricity of the engine gap, the condition of the drive system (whether direct, via belts or gearboxes) and the condition of the bearings.

Ultrasonic sensors, which are used for leak detection and inspection of mechanical and electrical components

Valves and valve stems. Valve operation and inspection for leaks or blockages can be carried out with precision while the valve is still in operation. Properly functioning valves are relatively quiet while leaking valves produce a turbulent flow as the fluid moves from the high-pressure side to the low-pressure side. Due to the high sensitivity and wide range of ultrasound frequency selection, all valves, even in noisy environments, can be tested with precision. Similarly, valve stems can be easily tested for any leaks. 

Pressure / vacuum leaks. When any gas (air, oxygen, nitrogen, etc.) goes through a leak hole, it generates a turbulent flow with audibly detectable high frequency parts. By scanning the area, the sound of a possible leak can be heard as a sound (with the help of headphones) or noted on the screen / counter. The closer the instrument is to the leak, the stronger the sound and the better the reading.

Steam Trap Inspection. The largest steam traps manufacturers recommend ultrasonic inspection as one of the most reliable methods of detecting faults. By converting ultrasounds into acoustic frequencies, operators can listen through the headphones and see the exact condition of a steam trap on a screen / meter while it is in operation. Additionally, possible blowing, over dimensioning or obstruction can be easily detected.

Rotary compressor valves. Ultrasonic analysis has become so safe and successful that many engine-analyzer tooling companies now offer instruments with an ultrasonic input port.

Bearing Inspection / Monitoring. Checking a bearing is easy. With just a test point and minimal training, users can learn to test bearings within minutes.

Check for over-lubrication. Add lubricant until the ultrasound meter reaches the desired level. Excessive lubrication is one of the most common causes of bearing failure.

General Mechanical Inspection. Pumps, motors, compressors, gears and gearboxes; all types of equipment can be inspected. Since the instrument operates in a short wave, high-frequency environment, problems such as cavitation in pumps, compressor valve leakage or loss of missing gear teeth can be heard and properly isolated.

Inspection of electrical installations. Electric arc and corona discharge produce ultrasound. These electrical discharges can be quickly detected by scanning the area. The signal sounds as a blowing sound. Inspections can be carried out on switches, distribution bars, transformers, circuit breakers, insulators and other electrical components.

Application extensions

The question is: why is preventative maintenance the standard, when predictive maintenance has such positive results? The main reason is the initial installation cost, which has been high until recently compared to preventative maintenance.

However, predictive maintenance offers significant cost savings by reducing not only the downtime of the production line, but also the cost of other components and spare parts. According to the Roland Berge study for the oil and gas industry, investing in a predictive maintenance has up to 10 times return on investment. These results, combined with the reduction of the cost of sensors and the increase in computational power, make predictive maintenance an affordable and necessary tool for an industrial unit, helping in increasing its capabilities and opening new ways to control and monitor equipment.

With the use of multiple sensors and the development of advanced portable and permanent systems, maintenance can be highly reliable. Additionally, one can easily and quickly measure points of interest, even with multiple different sensors at the same time. Moreover, different types of measurements give more accurate results and help to identify the problems more accurately and assess the asset status.

Progress does not stop there though. The capabilities of cloud-based systems and the network availability offer immediate data storage and integration with production software tools.

Using these methods, the mistakes and omissions of the past in the import of data are eliminated. In addition, one can access the status of the plant from everywhere and make decisions about maintenance and operation. Permanent systems and the Internet of Things (IoT) allow repetitive, accurate and complete measurements in inaccessible or remote locations at any time and without the need for physical access.

The above, when implemented on a software platform that is flexible and safe, can bring results that are not imaginable when compared to the usual maintenance and inspection practices available today.

In summary, the combination of diagnostic methods yields greater credibility to facilities, while the use of the Internet and new technologies offers ease and speed in decision-making.

Predictive maintenance is an attractive option and in the majority of cases it is an investment worth making, and which brings results.

Text: Antonis Kalipetis, Arpedon Private Company