How does it compare and where does it fit in your plant’s Reliability Program?
It is well known that rotating machinery exhibit specific characteristics when faults cause the geometric centerline of the shaft to periodically move. For more than 70 years, Machinery Vibration Analysis (MVA) has been used to identify and determine the severity of these faults and is an integral part of many successful plant reliability programs.
Recent experience and research has proven that many of these same faults can be identified using Electrical Signature Analysis (ESA). ESA also evaluates and identifies faults with power coming into the plant, the power supplied to the motor, as well as electrical and mechanical faults within the motor system. Moreover, ESA is emerging as a very important technology within some Electrical Reliability programs.
Some plants are using it as the main detection tool to identify both electrical and mechanical problems on machines, which are driven by electric motors. This article will examine both of these dynamic Predictive Maintenance Program (PdM) technologies and identify the strengths and weaknesses of each technology and try to determine where these two technologies best fit into a Reliability Program.
Motor Circuit Analysis is a powerful tool that is simple and intrinsically safe (offline test). The testing range and potential payback is almost immediate.
Electrical Signature Analysis
ESA is a predictive maintenance (PdM) technology that uses the motor’s supply voltage and operating current to identify existing and developing faults in the entire motor system. These measurements act as transducers and any disruptions in the motor system cause the motor supply current to vary (or modulate). By analysing these modulations, it is possible to identify the source of these motor system disruptions.
Machinery Analysis
Historically, vibration analysis has been the basis for rotating machinery analysis to assess the condition of rotating equipment and has been used very effectively for over 70 years. Modern electronics and micro-processors have matured this process, from simple vibration amplitude measurements using a coil, magnet and a meter to measure overall vibration amplitudes to quickly assessing the mechanical condition of rotating machinery. It soon became apparent that machines with high levels of vibration generally were in poor mechanical condition and led to the development of various vibration severity charts, all of which are based solely on users’ experience.
Maintenance Philosophies
Companies with a large population of capital equipment either provide a service or produce a product with this highly capital-intensive equipment. In order to protect this equipment and keep it in operating order, it is necessary to perform maintenance. Pressure, over time, continues for companies to produce higher quality products at a lower cost, while trying to gain higher profits. Service providers are also subjected to providing more reliable services at a much lower cost. This requires the maintenance department to not only properly maintain this equipment, but to do it at a lower cost.
The overall increased costs, not including wasted product, would be 93.6% would be due to lost production, 3.1% due to increased power consumption, 1.2% due to decreased motor life and 2.1% due to increased demand costs (Figure 1).
These pressures have led to the evolution of maintenance practices or philosophies. Early maintenance practices were known as “run till failure” (RTF), but industry pressure has evolved these practices to precision (or proactive) maintenance.
Run Till Failure
This approach requires little involvement other than turning the machine on and off and supplying a product. In this manner the machines continue to operate without interruption. However, when failure does occur they are usually very severe and result in failure to the original component as well as damage to other components of that machine, such as connected machines and the foundation. This additional harm often results in damage to components, which usually do not fail, and are seldom found in the in-plant’s spares.
Repairing or replacing these components require manufacturing them in-house or purchasing them from the original manufacturer at premium cost and long lead times, thus resulting in lengthy shutdowns. Therefore, RTF results in the most expensive method of maintaining plant equipment. This is without considering the lost production costs.
Preventive Maintenance
This maintenance philosophy is based on the assumption that mechanical equipment will wear and fail over time. Machine designers and manufacturers, research and study their machines to determine the recommended maintenance requirements and inspection intervals for their machinery. The recommended maintenance and inspections are then performed at these predetermined time intervals.
However, a reliability study written by Nolan and Heap, determined that machines do not fail on time. They fail either too early or too late. Machines that fail too early have the same problems and costs associated with the “run till failure” maintenance, whereas machines that fail too late result in many hours of unnecessary maintenance and premature replacement of components.
Predictive Maintenance
An additional reduction in maintenance costs was achieved through the use of condition monitoring. In the early 1960’s companies recognized that when rotating equipment began to fail, its operating conditions would change. By routinely monitoring these operating conditions, an advanced warning of these changes provides sufficient time to remove the machine from operation, before catastrophic failure occurs.
Machinery Vibration Analysis (MVA)
The process of machinery vibration analysis identifies the frequencies that are present in the machinery vibration and then correlates them to the frequencies of the forces that are created by mechanical and electrical faults. To determine the frequencies that are present in the measured signal, the analyser performs a Fast Fourier Transform (FFT) on the signal. This mathematical process converts the collected complex time based signal from the time domain to the frequency domain. The FFT identifies the amplitudes and frequencies that are combined together to make up this complex signal.
Mechanical Faults
There are numerous charts, tables and papers that describe the frequencies that each of these mechanical faults generate, when these faults are present. Several of these faults generate the same fault frequencies. Faults such as unbalance, misalignment, bent shaft, cracked shaft and an eccentric rotor all are created by faults on the rotor and will generate forces that are related to the shaft rotational speed. In many cases, it is necessary to perform additional measurements or use additional technologies to further define these similar problems.
Other problems such as rolling element bearing defects have frequencies that are dependent on the stage of the defect as well as the geometry of the bearing. One of the problems with rolling element bearing defects is that defects in the early stages generate very low amplitude signals and are difficult to identify in the early stages of a developing fault.
Electrical Faults
Electric motors operate by the interaction of magnetic fields on the rotor and the stator. If the magnetic field on either the stator or the rotor become unbalanced or distorted, it will create unbalanced electrical forces inside the motor. These forces will cause the rotor to move inside the motor as the rotating magnetic field passes the distorted or unbalanced fields.
Stator Electrical Faults
The shape of the core is determined by magnetic field. Both the stator core and the rotor are normally designed to be perfectly round.
Other Faults
Includes unequal air gaps, loose winding/stator iron, rotor electrical faults, eccentric rotor, broken rotor bars, thermally sensitive rotor among the most common.
Electrical Signature Analysis (ESA)
Electrical Signature Analysis measures all three phases of current and voltage at the motor controller, while the machine is operating. By measuring all three phases of voltage and current, a complete analysis of the power being supplied to the motor is performed each time the ESA data is taken. Additionally, an FFT is performed on the voltage and current waveform.
Testing and research has shown that many mechanical and electrical faults in the motor system will cause the motor current to modulate at the frequency of the fault.
Power Analysis
Power analysis will not only identify problems relating to the motor, but will also identify any incoming power issues such as excessive harmonic content, voltage unbalance, voltage mismatch, current unbalance, the power factor of the motor system and the motor system efficiency. Also, since ESA simultaneously measures all three phase of voltage and current, it can very accurately determine the load on the motor. This allows the ESA software to accurately determine the actual rotor speed; typically the running speed is measured within 1 RPM.
FFT Analysis (fast Fourier transform)
The FFT of the current identifies faults in the motor system similar to MVA or other signature analysis techniques. However, performing the FFT on both the motor voltage and current waveforms provides additional diagnostic capabilities, when compared to MVA and Motor Current Signature Analysis (MCSA). Both MVA and MCSA measure the response of the motor system only.
If there are large spectral peaks in either the current spectrum or the vibration spectrum, which is a result of a carrier frequency in the incoming power, this is undetectable with either of these techniques. However, by performing a FFT on both the voltage and current, any spectral peaks that are present are coming from the incoming power. However, if there are no spectral peaks in the voltage spectrum that are present in the current spectrum then the fault is coming from either the motor or the driven machine.
Stator Faults
Stator faults in ESA are categorised as either electrical or mechanical in nature.
Stator Mechanical Faults
Faults categorised as stator mechanical faults are created when either the stator core becomes loose in the motor frame, or if the windings are loose in the stator slots. Either of these faults will cause magnetic fields created by the discontinuities in the stator iron the windings are placed to modulate. These frequencies are known as stator slot passing frequencies, which are determined by multiplying the number of stator slots by running speed.
Stator Electrical
If the insulation between stator windings and ground breaks down, a winding fault or ground fault will occur. These faults result in localized heating and further insulation degradation until the winding eventually burns and completely destroys the winding and in severe cases warps or burns the inner laminar insulation.
When these faults occur, the winding’s weaknesses causes the stator slot passing frequencies to modulate at line frequency, as the magnetic field rotates around the stator. These will further modulate as the shaft turns, which will create running speed sidebands around the line frequency sidebands.
ESA can identify faults that are classified as stator electrical, but to confirm fault type, performing Motor Circuit Analysis with the motor de-energised is recommended.
Again these modulations in the stator slot passing frequency are so slight and the forces created are very small and are usually undetectable with MVA.
Rotor Faults
Common rotor faults detected using ESA are Static Eccentricity (unequal air gaps), Dynamic Eccentricity (Eccentric Rotor), and broken rotor bars.
Static Eccentricity
When the rotor core is concentric and centered in the magnetic field, the current flowing through the rotor bars will be equal and polar opposites on opposing sides of the rotor. But, if the rotor is not centered in the magnetic field then the strength of the magnetic field in the rotor bars that are closest to the stator will be stronger than the opposite side.
Dynamic Eccentricity
If the rotor is centered, but the rotor core is eccentric this will create a narrow air gap that rotates around, the inside of the air gap with the rotor. The narrow air gap creates two times line frequency (2xLF) sidebands around rotor bar passing frequency, but because the narrow clearance is rotating around in the air gap, at rotor speed, it will cause the 2xLF sidebands to modulate at rotor speed. This creates running speed sidebands around the 2xLF sidebands.
Broken Rotor Bars
When the dead spot on the rotor passes under a magnetic field there will be no inductance between the stator’s magnetic field and the rotor. This will cause the motor current to modulate at PPF, this creates PPF frequency sidebands around line frequency in the current spectrum.
What is the future of ESA?
Preliminary testing has indicated that ESA is one of the most powerful tools available for screening motor driven machines. In almost all cases the faults appear much earlier in the ESA data than in MVA since the force of the fault does not have to be sufficient to move the entire machinery structure, as does mechanical vibration. Also, ESA is capable of determining the condition of the power supplied to the motor system as well as determine the motor efficiency, and most importantly the exact running speed of the motor at the time the data was taken. This measurement is critical when using ESA and MVA, since most faults in the motor system are speed dependent and an accurate determination of the running speed is crucial to accurate spectrum analysis.
Frequency Response
Since ESA uses changes in motor current as identifiers of faults, even very low and very high frequency faults can be detected. MVA has limits based upon the measurement type (relative or absolute) and the frequency response of the sensor.
Deep Well Vertical Pumps
Experience with vertical pumps has shown that faults in the pump are not transmitted to the motor. To determine what is going on in the pump it is necessary to place transducers on the pump itself. Pump faults are not detected on the motor until the pump is usually completely destroyed. Preliminary testing has shown that small amounts of cavitation and even vane passing frequencies in the pump can be easily detected using ESA. MVA spectrum taken at the same time showed no evidence of either fault.
Variable Frequency Drives
When using ESA to test motors driven by VFDs, not only can motor system faults be detected, but aged capacitors and other electrical problems in the drive are very readily apparent.
Summary
Successful implementation of PdM programs requires a thorough understanding of the PdM process and the efficient utilisation of highly trained PdM personnel together with special and often expensive equipment. Reliability engineers agree that developing faults need to be identified as early as possible and ESA fulfills this requirement. As a detection tool, ESA usually identifies most mechanical faults in the motor system before mechanical methods like machinery vibration analysis (MVA). Additionally, ESA accurately identifies electrical problems in the motor system that MVA or other PdM technologies cannot identify. In the analysis phase, ESA more accurately determines the system’s rotational speed and more precisely identifies the mechanical and electrical faults that lead to reduced plant availability and uptime.
A range of articles and white papers on this subject can be found on the All-Test Pro website here >>