Difference Between Condition Monitoring and Protection in Induction Motors
Condition monitoring and protection are closely related concepts; however, their implementation approaches and objectives are fundamentally different. The benefits obtained from condition monitoring are also different from those expected from protection systems.
In many cases, condition monitoring can be used to provide a preliminary level of protection, but its primary function is to identify the development of faults at their earliest possible stage.
It can be argued that when a failure occurs, a machine may already be considered out of service even before fault currents begin to flow. Electrical or mechanical failure modes generally originate from damage to one of the machine components, including mechanical, electrical, magnetic, insulation, or cooling elements, regardless of the type of electrical machine.
If a fault develops gradually over a sufficient period of time and can be detected through measurement, identifying the root cause of the damage provides a means for monitoring the machine before catastrophic failure occurs.
The core objective of condition monitoring is therefore to develop methods—preferably as direct as possible—for measuring parameters that indicate the progression of fault mechanisms. These measurements allow early warnings of an impending failure, making it possible to repair the machine or remove it from service before serious damage occurs.
A limited level of protection can be achieved by increasing the sensitivity of protective relays and providing warning alarms before relay tripping. However, experience has shown that this approach represents a risky form of condition monitoring, often resulting in false alarms and reduced reliability in the monitoring process.
The primary objectives and advantages of condition monitoring of electrical machines are mainly related to early failure prediction. These advantages include:
- Significant reduction in maintenance and repair costs
- Prediction of component failures before they occur
- Optimization of component performance
- Improved accuracy in failure prediction
- Increased system reliability
- Improved safety for both equipment and operators
- Reduction in inspection and maintenance requirements
Considering these advantages, condition monitoring of induction machines has gained significant importance in modern industry.
Selecting an appropriate condition monitoring technique mainly depends on:
- The purpose of maintenance activities
- The components under investigation
- Cost limitations
- The type and size of the machine
For example, visual inspection may be sufficient in some cases to verify operational conditions. In contrast, continuous monitoring systems require more advanced and sophisticated implementation approaches.
Therefore, when selecting a condition monitoring technique, both the monitoring objectives and the associated costs must be carefully considered.
Condition Monitoring Techniques for Induction Motors
Visual Inspection
Visual inspection forms the basic process for evaluating system components and must be performed by personnel with sufficient expertise and technical knowledge.
It is the most commonly used condition monitoring method and may involve simple actions such as observing, rotating, shaking, and even smelling components, which can provide preliminary information about the machine’s condition.
For improved accuracy, visual inspection may be supplemented with handheld instruments, such as:
- Infrared detectors
- Stroboscopes
- Other simple diagnostic tools
However, the major limitation of this method is its inability to detect faults at their early stages, since the inspection is generally limited to externally visible components.
In addition, due to variations in the experience, skills, and expertise of inspectors, different evaluations and conclusions may be reported for the same machine.
Oil Analysis
Lubrication plays a critical role in the operation of rotating machinery, directly affecting the lifetime, efficiency, and performance of the system.
Monitoring lubricating oil helps maintain machine health by predicting equipment condition and enabling preventive actions before catastrophic failures occur.
Lubricating oil analysis can provide valuable information regarding:
- Changes in viscosity
- Chemical contamination
- Presence of wear particles
For example, bearing failures account for approximately 50% of faults in rotating electrical machines. When bearing wear occurs, metallic particles are typically released into the lubricating oil, allowing early detection of potential bearing failures.
Similarly, when internal faults such as partial discharge occur within a machine, contamination of the cooling oil may also be observed.
Vibration Monitoring
Fig.1: Vibration signal sampling with MCM1
All electrical machines generate vibrations, and vibration analysis can provide useful information regarding machine condition.
Even small vibration amplitudes can produce significant noise and measurable signals.
Vibrations in electrical machines may originate from:
- Magnetic forces
- Mechanical forces
- Aerodynamic effects
Faults that can typically be detected through vibration analysis include:
- Bearing faults
- Rotor eccentricity
- Gear faults
- Mechanical imbalance
Vibration monitoring is typically performed using:
- Broadband analysis
- Narrowband analysis
- Signal signature analysis
Despite its effectiveness, the implementation of vibration monitoring may not always be economically feasible, especially for smaller machines. This is primarily due to the high cost of accelerometers and associated wiring.
Another limitation is the requirement for physical access to the machine. For accurate measurements, sensors must be firmly mounted on the machine structure, which requires technical expertise. Additionally, sensors themselves may fail or degrade over time.
Torque Monitoring
Different motor faults produce torque components at specific frequencies within the air-gap torque. However, direct measurement of air-gap torque is not feasible. Differences between estimated torque values may indicate the presence of faults such as broken rotor bars.
At the motor input terminals, instantaneous electrical power includes energy associated with the charging and discharging of winding inductances, meaning that instantaneous power does not directly represent instantaneous torque.
At the mechanical output, the rotor, shaft, and mechanical load form a torsional system with its own natural frequencies. Torque components transmitted through this torsional system are attenuated differently depending on their harmonic order, making direct torque-based fault detection more complex.
Thermal Monitoring
Temperature measurement or thermal monitoring is performed by installing temperature sensors at specific locations, such as:
- Bearings
- Stator windings
This method is useful for detecting both electrical and mechanical faults.
Thermal monitoring effectively reveals changes in system operating conditions, although its effectiveness depends strongly on the location of the installed sensors.
Temperature measurements can be performed using various devices, including:
- Thermal cameras
- Thermometers
- Thermocouples
In general, three primary approaches exist for thermal monitoring:
- Local temperature measurement at specific points using embedded sensors
- Thermal imaging to identify hot spots within the machine
- Measurement of overall temperature distribution within the machine
Electrical Signature Analysis
Many condition monitoring techniques—such as oil analysis, thermal analysis, acoustic analysis, and vibration monitoring—are relatively expensive and require specialized instrumentation. Consequently, these techniques are often applied only to critical machines in industrial facilities.
In contrast, electrical signal–based monitoring methods typically do not require additional instrumentation, since measurement systems are already installed for electrical protection purposes.
Fig.2: Current signal sampling with MCM1
The air-gap magnetic flux inside the machine varies with time. Under ideal conditions, the magnetic flux waveform is symmetrical. However, electrical faults within the machine distort this waveform.
Rotor or stator faults disturb the radial and circumferential flux patterns, which in turn alter the power supplied to the machine.
These changes appear as additional frequency components in the voltage, current, and power signals, and by analyzing these components, the type and severity of the fault can be identified.
