When navigating the Las Vegas Strip, one can’t help but be impressed by the array of water features, waterfalls and fountains. At the heart of each are the pumps that drive them. Proper pump application and reliability are paramount to maintaining these attractions.
One such water feature, an indoor casino waterfall using eight pumps regulated to produce the desired fall effects, is the subject of this article. This indoor water feature provides the backdrop for an exclusive luxury bar and lounge. However, running all eight pumps resulted in excessive failure rates, compromising the feature and diminishing the site’s appeal.
The system uses horizontal overhung pumps, each driven by a variable frequency electric motor (see Image 1). The pumps are constructed of cast iron and, with the exception of impeller diameters, identical.
The mean time between repair (MTBR) had become unacceptable at one-month intervals. Seal failures were typical casualties, but significant impeller and casing cavitation damage was also becoming apparent. To identify the problems, a comprehensive hydraulic and mechanical assessment was commissioned to determine the status of the pumps and provide recommendations encompassing materials, maintenance and operations.
Pump & System Assessment
Instrumentation was installed, and each pump was tested under typical operating conditions. Casino patrons visit the feature more frequently during the evening, which provided an opportunity earlier in the day to assess the system in configurations that included pump pairings and speed regulation. The following operating parameters were measured:
- discharge flow
- suction and discharge pressures and temperatures
- pump speeds
- vibration signatures
- net positive suction head available (NPSHa)
The measured data was compared to the manufacturer’s curves. Also, a system resistance curve was generated from site data. The as-found performance for all eight is averaged in Figure 1.
The total developed head (TDH) of all the pumps was approximately 5 percent below expected at the measured flows. The desired water feature esthetic required all eight pumps to operate at 22 percent of the best efficiency point (BEP) capacity and significantly below the manufacturer’s recommended minimum continuous stable flow (MCSF) of 2,050 gpm. The NPSHa was determined at the measured flow rate. However, the net positive suction head required (NPSHr) curve did not extend to the measured field. This was concerning since NPSHr increases at low flow rates for some designs.
Concurrent with the hydraulic data acquisition, vibration signatures were recorded at the locations indicated in Image 2. The unfiltered amplitudes are presented in Figure 2.
The amplitudes were higher than expected for pumps of this type. A review of the associated spectra provided further insight. The spectrum and time waveform presented in Figure 3 is typical of the majority of the signatures collected from all of the pumps. Note the spectrum is replete with random frequencies commonly referred to as spectral floor noise, which is indicative of hydraulic anomalies such as recirculation or cavitation. Further evidence of these anomalies is presented in the associated time waveform where random impacting is evident, a typical energy signature associated with cavitation. Additionally, during the low flow testing, audible noise associated with extreme cavitation was present.
System Resistance Curve
Centrifugal pumps always operate at the intersection of the pump head capacity and the system resistance curves. The pump head capacity curve is supplied by the manufacturer while the system curve is a plot of the head requirements as a function of flow, calculated by the system designer.
System resistance curves include three components: