Field Evaluation of Different Check Valve Designs and Surge Pressure Measurement in a Municipal Water Pumping Station
Overview: During a municipal water pump station upgrade, hydraulic engineers got approval to perform field tests to assess the waterhammer (pressure surge) transients. Surge pressure measurements were taken after an induced pump power loss and subsequent check valve slam. They also developed a computer model to see how well available software could simulate the produced transients. This data was gathered to provide guidance to the municipality on future project work.
Category: Blogs, Case Studies December 12, 2024
Intended Audience: Waterhammer/surge transient engineers of all experience levels.
Objective: The primary project involved the replacing and upgrading of equipment in an aging, municipal water pumping station. As part of the project scope, additional goals were added to gather data to guide future project efforts:
- Compare the slamming effect of two different check valves: a conventional swing check valve vs. a nozzle check valve.
- Reinforce the importance of proper check valve selection.
- Discuss the importance of energy dissipation during a transient event on complex systems.
- Explore agreement between computed and measured pressures during a transient event.
- Improve understanding of the robustness and reliability of computational modeling tools to perform transient analyses.
Description of System: The Tibidabo pumping station in Barcelona, Spain, is part of a municipal water network over 125 years old and serving over three million users. The pump station received water through a 2 mile (3.1 km) long stainless steel pipe which was 12-inch (300 mm) in diameter. The pump station then supplied a main pipeline which was 12-inch (300 mm), mostly ductile iron pipe about 0.8 miles (1.3 km) in length. The study was focused on this ductile iron pipeline. The pipeline transferred water to a tank 450 ft (137 m) higher in elevation. The system supplies water to the highest neighborhood in Barcelona, known as Vallvidrera. The ductile iron pipeline was rated to a maximum pressure of 275 psig (20 bar-g).
There were two pumps past their design life and the operating equipment was outdated. The existing flow capacity was 950 gpm (215 m3/hr). The system functioned with one pump in operation and one on standby. Each pump had traditional swing check valves at each pump discharge.
The system had two surge vessels: a 1,150 gal (4.3 m3) vessel on the pump suction line and a 1,650 gal (6.2 m3) vessel on the pump discharge line.
Description of Intervention: The pump station and the two pumps needed upgrades and replacement. Further, it was desired to increase the flow capacity by 33% – to 1,270 gpm (288 m3/hr). The new pumps were rated at TDH of 510 ft (155 m) with constant speed at 2900 RPM. The rated pump power was 300 hp (220 kW).
As described previously, part of the project involved gathering data to guide future projects and developing a surge computer model. To get the desired data, a temporary high frequency pressure transducer (HFPT) was installed after the check valves. Further, one swing check valve was replaced with a new swing check valve. The other check valve was replaced with a new nozzle check valve. See Figure 1.
Figure 1: Pictures of the axial flow INDAR pump sets (at left) and the surge vessels (pump suction/discharge, at right) installed at Tibidabo Pump Station. The high frequency pressure transducer (HFPT) was installed downstream of the check valves in the common discharge line.
After the refurbishment was completed, tests were performed in January 2017. The tests were run with one pump on and the other off as per normal procedure. Each pump was intentionally tripped and data gathered.
The surge model was created with the commercial AFT Impulse™ software. Figure 2 shows a simplified version of the model.
Figure 2: Surge model diagram with Pipe 1 stainless steel and roughly 2 miles (3.1 km) long, and from Object 2 to 16 predominantly ductile iron and roughly 0.8 miles (1.3 km) long. Not to scale.
Summary of Results: Field test results are shown in Figs. 3 and 4 and are cross plotted vs. the surge model predictions model. The first conclusion from Fig. 3 is that comparing swing and nozzle check valves, the nozzle valves reduce the surge pressure spike after valve slam by over 80%. It is well known that nozzle check valves reduce surge pressures, and this field test confirmed that.
The second conclusion is that, as shown in Figures 3 and 4, the surge model shows good agreement with field data.
Figure 3: Modelled pressure vs. recorded pressure oscillation for swing check valve pump set (at top) and nozzle check valve pump set (at bottom) during the minute after the trip. Note that the top and bottom scales are different.
Figure 4. Modelled water level vs. recorded level in discharge side surge vessel. Swing check valve pump set (at top) and nozzle check valve pump set (at bottom) during the minute after the trip.
Finally, surge engineers often use dynamic characteristic performance data for check valves in the form of Figure 5. Figure 5 shows where the Tibidabo Pump Station check valve behavior falls within existing experimental data. Here one can see that the swing check valve data differs significantly from the curve fit of previous test data (the DN 200/Ballun data). It is difficult to assess the nozzle check valve data due to the lack of experimental data for DN 200. After completing the field test, the existing swing check valve in Figure 1 was replaced with a nozzle check valve in order to reduce surge pressures for operations with either pump.
Figure 5: Dynamic performance of different check valves, from Thorley and Ballun experimental tests, along with estimated performance of Tibidabo installed valves. This chart covers data for 4, 8 and 12-inch (100, 200 and 300 mm) line sizes. For Thorley and Ballun references see the Case Study Source section.
Conclusion: The refurbishment of the Tibidabo Pump Station in Barcelona Spain provided a unique opportunity to measure and assess the surge pressures after check valve slam of two common types of check valves – swing vs. nozzle. It was shown that nozzle valves reduced the surge pressure spike by over 80%. Further, it was shown that surge modeling software predictions showed good agreement with the measured data. For the water utility in Barcelona, this study provided a good reference for replacing existing swing check valves in their network. It also improved confidence that the pumping stations could be modeled with commercially available software with good accuracy.
Written by:
This study was taken from the published paper “Surge transients due to check valve closure in a municipal water pumping station” which was presented at the 13th International Conference on Pressure Surges 2018, in Bordeaux, France.
The authors of the paper were David Lozano Solé and Roger Bosch Segarra with Aquatec Proyectos para el Sector del Agua SAU (SUEZ Group), Barcelona, Spain, and Trey Walters with Applied Flow Technology, Colorado Springs, Colorado, USA.
SUBSCRIBE TODAY
Get the latest pump industry news, insights, and analysis delivered to your inbox.