- Real-world case studies with EMI
- Wireless monitoring solutions
A recent webinar presented by Cutsforth Inc focused on electromagnetic interference monitoring (EMI), a valuable diagnostic tool for detecting impending problems with generators, motors, isophase bus, bearings, and other plant equipment. Primary presenter and discussion leader was Kent Smith, well respected in the electric power industry for his deep knowledge of EMI, honed by years of service as the lead generator expert for one of the world’s largest utilities and as the chairman of the Generator Users Group, one of the planets in the Power Users universe.
Plant personnel not able to participate in the webinar when it aired can access a recording here via the expansive Cutsforth webinar library. The editors believe you will benefit professionally from Smith’s case studies which illustrate findings by way of data scans. Smith, who was supported by Cutsforth’s Steve Tanner, VP business development, shared several case histories, including these:
Generator monitoring with EMI. Water was affecting the calibration of hydrogen analyzers. A cooler leak was found and the unit repaired. Generator reached end of life without a winding replacement.
Motor monitoring with EMI. Plant’s six pump motors had started multiple times without cooldown between starts; the possibility of damage to the induction-motor rotor bars was a concern. EMI and motor-current signature analyses were performed. One motor registered higher EMI values than the others. It was found to have salt-encrusted winding and some cooling passages plugged with salt. Cleaning was the fix.
Excitation power rectifier. Data revealed significant arcing and discharge in the lower frequency band and suggested a loose connection in the excitation system. Connections checked when the unit was in a “not-in-demand” (NID) state were found loose and tightened. EMI data returned to normal after exiting the NID state.
Wet stator bars and loose wedges. Generator was going into a rotor-out outage. Retaining-ring and stator-wedge replacements were scheduled, plus a hydraulic integrity test (HIT). Replacements were made with no issues.
However, when performing the HIT skid test, plant personnel couldn’t pull the required vacuum on the unit. Capacitance mapping and helium leak testing was performed. The findings: four significant clip leaks, minor plumbing leaks, wet bar found on the “B” phase. Corrective action: Leaks were repaired, bar dried out, and the unit HiPot-tested to an operational level. The generator was returned to service with a rewind planned for the following year.
Bearing electrolysis. EMI trending located a loose ground lead.
Isophase-bus flex link. One unit had been monitored for years because of its high EMI readings. The brushing box was suspected because of the frequency content, time-domain waveform, and sniffer readings. Overheated flex links and moisture intrusion found by transformers was repaired, but improved EMI improved only slightly. During the next outage flex links under the generator were removed and inspected. The results of that investigation were not available at the time of the webinar.
Water pumping stations. Electromagnetic signature data were collected and analyzed for two pumping stations, each having nine synchronous-motor-driven pumps. The motors were equipped with rotating pilot exciters and rotating main-exciter-to-feed-motor main fields.
The worst-case motor at one pumping station was found with pilot-exciter brush rigging and commutator arcing. It was experiencing alignment/rotor wobble and had loose connections at the bus connector and/or insulator.
The worst-case motor at the second pumping state had similar issue characteristics, plus loose windings in the slot causing slot discharge.
Bearing electrolysis. A generator was removed from service with high vibration on the No. 1 turbine bearing attributed to electrolysis, which caused pitting and melting of the babbitt material. An enhanced shaft grounding system, with a sensing point for voltage, was installed. Plus, a ground current monitor was installed for the turbine. Instrumentation was connected to the main server to access the EMSA data.
Engineers believed there were the following three possible sources of the high voltage:
- Static voltage build-up because of a brush rider in the turbine blading.
- A magnetic driving force from turbine-shaft magnetism.
- Static exciter thyristor firing voltage transition.
Turbine bearing data are presented both graphically and in tabular form.
The unit was removed from service several times because of high vibration, with bearing electrolysis believed to be the cause. The bearing was replaced and clearances validated. Vibration analysis suggested electrolysis was still occurring. The fix was installation of a high-frequency blocking filter on the exciter field circuit. That eliminated the high vibrations and electrolysis.
Wireless monitoring solutions
Chuck Requet, principal applications engineer, and Steve McAlonan, director of business development, began their presentation by explaining the value proposition of the company’s InsightCM™ architecture, which can accommodate multiple measurement technologies in a single platform.
Vibration monitoring was a focal point of the presentation, which can be accessed here. InsightCM was said to support industry-standard viewers for vibration—including trend, waveform, spectrum, waterfall, orbit, polar, bode, shaft centerline, full spectrum, envelope (amplitude demodulation), order (even angle), time synchronous average, and autocorrection.
Wireless is particularly advantageous for monitoring the many common assets—such as pumps, fans, compressors, etc—that would benefit from more attention. Also, when assets are not deemed critical enough to warrant 24/7 screening, or may be located in remote, difficult, or hazardous locations. InsightCM supports two wireless families, NI and Erbessd, which, in turn, support Bluetooth 5.
An example illustrating the value of wireless was for a large user with 35,000 sensors. This project was said to have three-year breakeven cost for hardwired vibration of $80-million. Wireless reduced the install cost by 70% and the planned major design effort was shifted to “minor modification.” The breakeven went from three years to 18 months.