The days of frequency vibration analysis started well before a day in 1986, when journalist Dan Rather’s altercation with two assailants was immortalized by them yelling, “Kenneth, what is the frequency?” Yet, that was the exact year I started my career as a vibration test engineer for a large public utility company.
In the world of predictive maintenance and condition-based monitoring, vibration analysis is the grandfather to all other technologies. In the 1970s and 1980s, vibration analysis equipment was very expensive and could only be afforded by the military and regulated power utility companies.
My fortunate hire as a field vibration test engineer was a perfect match for me, with my mechanical engineering degree and my background in auto restoration and customization. I grew fast and couldn’t get enough information about vibration analysis.
At the utility company, my vibration test group serviced the entire company and its assets, which included fossil steam production, nuclear steam production, and electric and gas distribution. Power utility companies are not nine-to-five jobs. They are 24-7, 365 days a year. My group felt that with a lot of overtime.
One company asset that demanded 24-7 service was the Salem Nuclear Power Plant. Our group created a satellite office on the site and rotated every two weeks out of six. This involved hotel stays for everyone. To build our on-site presence, we took on every job that came around that was even just a little bit related to vibration analysis. One job was to remove all the seismic acceptance sensors and their cabling. This was hundreds of sensors and miles of cables weaved throughout the entire enclosure of the containment. Completed in three weeks with four men, we gained a reputation of doing anything. Two of the men in the group were honored with having the highest completion rate in one year out of all employees in the plant over many years.
In the 1980s, one nuclear steam turbine produced $ million worth of electricity to the grid in one day. The turbine generator vibration was monitored by a third-party asset condition monitoring company, with proximity and seismic probes at each bearing. There was no continuous data collection and storage. If there was an issue, our vibration test group would cable in with oscilloscopes, digital vector filters, spectrum analyzers and reel-to-reel tape recorders. Hundreds of pounds of equipment was set up outside containment.
The site experienced a high vibration event on a monitoring system’s vibration dials for Salem’s No 2 TG high pressure rotor. The high vibration was also accompanied by an unusual high-pitched noise coming from the rotor location. Our group got set up and was asked to manually monitor the turbine vibration continuously, 24-7. We scheduled one man every twelve hours, with me having the overnight shift. All we had to link us to the control room operators was a two-way radio. Site engineers were our liaisons. They knew us by name. Yet, we were essentially just a continuous monitoring system that could talk.
Manual continuous monitoring setup consisted of eight channel reel-to-reel tape recorders and dozens of reel-to reel-tapes. Tapes were recording and all analyzers were on and watching 24-7. The recording tape speed was set at a balance between high enough vibration frequency capture and tape run time.
After about a week of my midnight shift, there still wasn’t any high-pressure rotor, high vibration, or noise. Then around 2 am a night later, I was watching my trusty spectrum analyzer and the high-pressure rotor noise made itself known. It was definitely a wake-up call. The spectral data indicated a 143Hz component in the rotor’s vibration data. That was it!! We had it, yet what did we have? Turbine rotor vibration almost always is a onetime running speed, with maybe much lower amplitude harmonics of that running speed. Turbines run with journal bearings. No bearing tones. This 143Hz is considered asynchronous to the 30HZ running speed of the turbine. All this analysis happened within seconds.
And in seconds, the two-way radio goes off. “What’s the frequency, Kenneth?” I obviously gave them the frequency of 143Hz and that was it. Nothing else. Later, as morning rose, I was pulled into a meeting with the turbine engineers and site vice president. I just had to confirm the frequency. This was truly vibration analysis in a silo. No RCA or defect elimination. They had the answers.
In the end, it was a loose fit steam diaphragm seal in resonance.
Later in 1994, I heard REM’s new song, “What’s the Frequency, Kenneth.” It brings back the entire event in my head as if it was yesterday.
