Electric Signature Analysis

	ABOUT THE AUTHORS Nissen Burstein is a Sr. Principal Engineer with the Engineering Unit of Framatome ANP. He has written many articles on electric signature analysis, published technical papers, and presented numerous case history studies before various industrial, commercial, and governmental technical forums, including the ASME, NRC, Vibration Institute, IEEE, JOAP, and several P/PM symposia. Mr. Burstein has earned Bachelor of Science and Master of Science degrees in Engineering from The Pennsylvania State University, a Master of Administration in Industrial Management from Lynchburg College, and a Bachelor of Arts degree. Donald Ferree has a B.Sc. from Ohio State University and an M.Sc. from Tennessee. Mr. Ferree has worked for Framatome ANP and its predecessor Babcock & Wilcox for over 27 years. He was the engineering manager for the development of the EMPATH system and has been instrumental in the continued development of this sophisticated motor diagnostics tool.

Electric Signature Analysis (ESA) is a technique for analyzing motors, generators, alternators, transformers, and other electric equipment. The technology can test operating electrical equipment and identify several mechanical and electrical problems; for example, ESA traces can analyze the driven load, the power supply, and perform inrush testing on motors. As a preventative-maintenance tool, you use ESA for either single or periodic testing to track equipment performance.

ESA is remote, non-intrusive, and invisible to the monitored equipment. Data acquisition takes less than two minutes per-motor and LAN-based, continuous monitoring of motors is readily accomplished.

Current and voltage data are acquired directly from the Motor Control Center (MCC), while the equipment is in operation. The collected data can help determine phase imbalance, motor load, power factor, power harmonics, and the impact of the driven equipment on the motor (Figure 1). ESA also assesses rotor bar as well as stator health and rotor-stator eccentricity (air gap) characteristics. In addition, you can observe degraded bearings from the traces (Figure 2). ESA is particularly helpful in accessing mechanical conditions when it is not possible or convenient to make vibration measurements (Figure 3).

Figure 1:  Three phase current sine waves for a 15 HP motor with a severe turn-to-turn short driving a pump. The unbalance in current is about 38%, much higher than acceptable to continue running this motor. When the motor was taken from service and opened up, a turn-to-turn short encompassing nearly half of one phase was observed.

Figure 2:  The current spectrum of one phase of a 300 HP motor driving a compressor. The rotor has broken rotor bars as evidenced by the amplitude of the pole passing sidebands around the line frequency peak at 59.97 Hz.

Figure 3:  The peaks are evident in the current spectrum of a 200 HP motor that has a broken bearing. The peaks highlighted by colored cursors are a result of the modulation of the current draw of the motor by the broken bearing.

For DC motors and motors with separate power supplies, such as variable-frequency drives, ESA monitors the power supply and can point out its problems (Figure 4).

Figure 4:  Armature current spectrum of a DC motor showing drive characteristics. The large peak at 360 Hz indicates the DC drive is full-wave-rectified. The large peaks at 120 and 240 Hz indicate problems in the control circuitry of the DC drive.

A one-time test rarely provides this data. However, trend-tracking provides an indication of how quickly a motor condition is changing. For example, if an indication of rotor degradation appears, it may not be clear from one test how rapidly the rotor circuit is degrading. Testing over several weeks or months will confirm if the rotor is stable and not changing (Figure 5).

The number of starts and stops that a motor experiences is very important regarding rotor change. A motor with a high on-off duty-cycle is much more likely to show rapidly increasing rotor degradation than a motor that runs constantly. These types of operating conditions can be factored into the trend-tracking data to provide a much clearer indication of motor health.

Figure 5:  The current spectrum of a 1750 HP motor that has "soft foot". The "soft foot" shows up as static eccentricity, or air gap variation. The peaks highlighted by the colored cursors represent the rotor bar passing peaks indicative of static eccentricity. This condition can be determined with a single test. Degrading static eccentricity will be seen as the peaks grow in amplitude.

Figure 6:  The demodulated current spectrum of a motor driven by a variable frequency drive. The motor drives a belt that drives a fan. The peaks in the spectrum are the belt passing peak at 5.64 Hz, the second harmonic of belt passing at 11.24 Hz, motor running speed at 22.05 Hz, and fan blade passing at 41.44 Hz. The VFD was running at about 45 Hz and the motor has four poles.

Figure 7:  The demodulated spectrum of a DC motor driving a gearbox. The peaks at 8.62 Hz and multiples are from one of the shafts in the gearbox. Sidebands are evident around the peak at 8.62 Hz; these come from gear meshing modulation on the shaft. The numerous peaks at the lower end of the spectrum come from the gearbox shafts and from the hunting tooth frequencies in the gear meshing.

Figure 8:  The current spectrum of VFD showing line frequency and chopping frequency peaks. Note first the peak at 52 Hz, the drive output line frequency. Then note the large number of peaks near 2975 Hz, which arise from the chopping frequency of the VFD. In this case, the motor powered by this VFD was being destroyed because of the high frequency ripple riding on the VFD line frequency.

Figure 9:  The inrush current of a 9000 HP motor and pump that requires almost 11 seconds to start. Most motors achieve standard operating speed within a few seconds or less, but because this motor/pump represents a very large mass, a considerable amount of time is required to bring it up to speed.