by Lev Nelik, Pumping Machinery LLC; Hiram Tanner, DC WASA, Washington, DC; & Phil Higgins, DC WASA, Washington, DC

Vibration monitoring can be an early warning of problems developing inside rotating machinery. As companies migrate from a passive run-to-failure approach to a proactive/preventive maintenance method, questions are asked about how to start, what approach to choose and how sophisticated a vibration monitoring plan needs to be.

In past issues of Pumps & Systems, I have touched on different methods that can be used to effectively strike a balance between sophistication and pragmatism. Not every plant needs a full-blown transient electric current analysis as described in “Transient Startup Problems for Pumps with Synchronous Motors,” September 2007. In this article, high-speed, real-time, start-up fluctuations of the motor amps were analyzed, and motor tripping problems were solved after several months of complex testing.

A better way to begin is to identify critical equipment. Every piece of equipment does not need to be part of a vibration monitoring program. Once the critical equipment has been selected, begin recording the trend of vibrations with a simple hand-held vibration analyzer, which shows the overall root mean square amplitude (RMS) values of vibration. See “Trending Revelations in Vibrations Analysis,” June 2009. This article described the manner and logistics of vibration program philosophies that plants tend to follow. The machinery survey usually precedes the formal launch of the program to determine the most critical machinery  to monitor. See “Pump Reliability and Energy Surveys,” December 2009. 

Figure 1. Vibration data for failing pumps

By the end of 2009, I received feedback and many questions regarding the proper balance between simple, overall vibration velocity value trending and complex fast Fourier transform (FFT) spectral analysis. This feedback was summarized in “Think Simple Before Rushing to Fancy Analysis,” August 2010; “Vibration Analysis Simplified,” November 2010; and “Facts and Fiction About Vibration,” March 2011.

“Vibration Analysis to Improve Pump Reliability and Save Energy,” September 2011, described how an initial simplified vibration program gradually expanded to include information regarding the pumps’ inefficiencies and energy consumption. Simple trending vibration methods and on-line, real-time continuous monitoring present a next step for plants that are interested in elevating their program to the next level. By using this combination, plants will be better able to predict internal problems with the machine before the problem becomes a failure. The remainder of this column focuses on a case study in which a wastewater facility in Washington, D.C., successfully used this method. 

Figure 2. Vibration data from the continuous monitoring system

Wastewater Pump Vibration Trending

In 2007, a municipality in the Washington, D.C., area began tracking the reliability of its wastewater treatment facility pumps using vibration analysis. Starting with a simple analyzer, initial efforts involved the in-house rotating machinery technician examining any pump with high vibrations. 

Because the municipality had several plants and a large number of pumps, servicing the units at the early signs of trouble was difficult. By the time a problematic unit was examined, the internal damage (bearings, leaky seals, etc.) had progressed substantially, resulting in the need for expensive repairs and downtime.

The monitoring system

In 2009, the municipality shifted to a more regimented systematic approach, trying to catch the problem closer to its inception. This approach helped reduce the number and cost of catastrophic failures significantly. However, while the majority of pumps under the trending program were monitored monthly, occasional, undetected failures still occurred. They happened despite the trending chart showing no obvious signs of early distress, as recorded by the overall vibration data, with overall vibration being significantly below the warning (0.3 inch per second) and alarm (0.5 inch per second) values. See Figure 1 for the data.

Figure 3. Full-spectrum data analysis

To understand the issue better, the plant technician and an outside consultant looked into the issue more closely. After interviewing several operators of all three shifts, they learned that the pumps tend to exhibit high vibrations at occasional and unpredicted times.

On closer review, some problems appeared to be pump- or motor-related, and others were system-related. A first discovery was an incorrect setting of the motor’s minimum speed. It  was not set low enough to accommodate low-flow periods (usually at night). This caused the motor to shut down and then restart a few minutes later as the wet-well level built up again. Readjusting the minimum speed setting solved the problem of frequent starts/stops, significantly improving the impact on the motor life. However, the high vibration spikes persisted.

Figure 4. Bump test results