Dr. Nelik (aka "Dr. Pump") is president of Pumping Machinery, LLC, an Atlanta-based firm specializing in pump consulting, training, equipment troubleshooting and pump repairs. Dr. Nelik has 30 years of experience in pumps and pumping equipment. He may be reached at pump-magazine.com. For more information, visit www.pumpingmachinery.com/pump_school/pump_school.htm.
Regarding the vibration spectrum on page 12 of the February 2016 issue of Pumps & Systems ("Using Pump Efficiency Monitoring to Make Faster Decisions," Figure 1, see image below and read the whole article here), your question of what the spectrum is showing needs more information to enable an accurate answer:
- Where is the sensor located?
- What kind of pump?
- What is the orientation of the sensor? X, Y or Z axis?
- Is the data averaged, peak hold or instantaneous?
- Is the pump speed constant or changing?
- Is the flow steady or changing?
- How many vanes are on the impeller?
- Are the units root mean square, peak or other?
- What is the power required by the pump?
- Is this a Category I or II pump per International Standards Organization (ISO) 10816-7?
Making some assumptions about the data, an unbalance problem appears to be present. It is not misalignment or cavitation. However, if you look at the spectrum, there is a slight peak just above one time per revolution at maybe 70 hertz (Hz). This could be resonance. If so, the running speed is less than 20 percent away from this natural frequency, and the response could be amplified by this resonance. While unbalance is indicated, it may be a resonance problem, and balancing is not, in my opinion, a permanent fix.
Assuming this is a small pump, judging by the operating speed, the amplitude is probably acceptable per ISO 10816-7. The overall level of vibration is probably less than 0.098 in/s-RMS, which is the Zone A/B boundary for new machines. But if the small peak at 70 Hz is a resonance, then all bets are off. If I were the end user, I would not be happy with a resonance less than 20 percent above running speed. As the machine wears, the resonance will drop in frequency, potentially to the point where the running speed will sit right on the natural frequency. With the low amount of damping present in the spectrum, this would create amplitudes well in excess of the ISO 10816-7 limit.
—Reader from California
Let me start with your ending notes first: very observant! The small peak just past the 1X is a potential problem. With wear and lowered stiffness of the system over time, the 70 Hz peak may creep right onto the 60 Hz resonance. Frequent or even continual monitoring would be important. The values are still small, and the unit is small, judging from data shown on the performance curve above the spectral plot. Unbalance (1X) is small, but the proximity to the possible critical frequency may lead to issues—though it is hard to predict how soon. While immediate pump pull and fix is not necessary, monitoring is.
Regarding your other points, the location of the sensor is relevant, but they are most often on bearing housings. The type of pump would not make much difference, although if it is vertical, then any unbalance would be a serious issue. Impeller wear way below the soleplate is not often detectable by the analyzers at the top of the structure (motor area), and vertical turbine vibrations are not easily detectable by the vibration instruments at the top.
Your point on sensor location has another important aspect. If the signal comes from the axial direction, then the angular misalignment may be an issue, which is often shown not as a 2X but a 1X component. In such cases, having another sensor direction is important. If it is at the offset direction (parallel of vertical, for example) and shows 2X (twice running speed), then misalignment is likely (presence of 1X and 2X).
The pump speed is constant in this case, but the cloud of higher head data is an indication of the second pump joining in at some time, with two of them now generating higher head, but at a flow only slightly higher—a typical situation with two pumps running in parallel.
Vibration data are at RMS, as shown at the vertical axis. The pump is small, and vibrations are well within the limits, as you noted. Interestingly, the resonance concern was expressed in this particular case, which is why continual monitoring is planned.
Thank you for your insightful comments and observations. If you want to be involved in our committee work, let me know, and we will get you involved. For more information, visit www.pumpingmachinery.com.