Reducer Fittings Decrease Pipe Size to Avoid Failure (Second of Two Parts)

Part One of this series (Pumps & Systems, August 2014), discussed the new American National Standards Institute/Hydraulic Institute (ANSI/HI 9.8) design standard requirements. It also discussed how to calculate the length of a reducer (see Equation 1) and the angle of a reducer (see Figure 1). L_r = 4 (D_L – D_s) Equation 1 Where: L_r = Length of the reducer D_L = Larger pipe diameter D_s = Smaller pipe diameter Part Two covers the assessment of the standard using computational fluid design (CFD). A recent study observed the velocity distributions resulting from different concentric and eccentric reducer angles (a constant diameter reduction with a variation in reducer length). During the study, the velocity distributions for the recommended allowable eccentric angle (20 degrees) without the requirement of additional pipe length fell outside of the time averaged velocity distribution requirements presented by ANSI/HI 9.8 (see Table 1).

CFD Assessment of the Design Standards

The CFD assessments were performed with upstream flow velocities [at DL (larger pipe diameter)] of 1 meter per second (m/s), 1.5 m/s, 2 m/s and 2.4 m/s. The range represented the typical flow velocities experienced in suction pipe work. To assess the criteria specified by ANSI/HI 9.8, the velocity distributions at the downstream end of the reducer along the y-axis were recorded and plotted together with 1.1 times the average velocity and 0.9 times the average velocity for a visual presentation of the results. The position of this axis for velocity measurement is illustrated in Figure 2 (for a 10-degree concentric reducer). Time averaged velocities were not used because a steady state CFD analysis was performed. The reducers studied were modeled on a nominal diameter (DN) 200 (210.1-millimeter inner diameter) to DN 150 (156.7-millimeter inner diameter), representing a single pipe reduction. The concentric reducer angles ranged from 2 to 20 degrees. The eccentric reducer angles ranged from 2.5 to 30 degrees. Twelve reducer geometries were modeled. Lengths of 3 DL were added upstream and downstream of the reducer to assess the extent of the velocity distribution. Typical results from the study are provided in Figures 3 and 4 (page 100), which show the CFD velocity scalar scenes for a concentric reducer with an angle of 10 degrees and an eccentric reducer with an angle of 20 degrees at a flow velocity of 2.4 m/s. Table 2 provides the upstream, downstream, and minimum and maximum velocities for the ANSI/HI 9.8 acceptance criteria. The velocity distribution results along the y-axis for the 10-degree concentric reducer, 5-degree eccentric reducer and 10-degree eccentric reducer are provided in Figures 5, 6 and 7, respectively (Figures 6 and 7, page 102). The results show the velocity distribution for the 10-degree concentric reducer is symmetrical with respect to the x-axis and falls within the ANSI/HI 9.8 criteria for all four velocities modeled. ANSI/HI 9.6.6’s requirement of zero straight lengths of pipe at the pump suction for a one-pipe-size-reduction 10-degree concentric reducer is confirmed in this study and corresponds to the ANSI/American Water Works Association (AWWA) C208 standard concentric reducer (see Table 1). For the 5-degree and 10-degree eccentric reducers, the flow is not symmetrical with respect to the y-axis. This asymmetry is caused by an acceleration of flow along the sloped (bottom) side of the reducer. The velocities for the 5-degree eccentric reducer (excluding the near-wall velocities, which tend to zero) fall within the ANSI/HI 9.8 10 percent velocity distribution criteria for all four of the velocities modeled. However, the 10-degree eccentric reducer’s velocities extend past or are on the limit of the ANSI/HI 9.8 10 percent velocity distribution criteria. The extent of the asymmetrical acceleration through the eccentric reducer increases as the reducer angle increases. If the 10 degrees do not pass the ANSI/HI 9.8 10 percent velocity distribution criteria, eccentric reducers with an angle greater than 10 degrees will also not pass the criteria. The ANSI/HI 9.6.6 requirement of zero straight lengths of pipe at the pump suction for a one-pipe size reduction, 20-degree eccentric reducer is therefore insufficient and should be questioned (see Table 1). The superior performance of the concentric reducer explains why ANSI/HI 9.6.6 prescribes a concentric reducer for vertical inlet (suction) pipes or horizontal installations where there is no potential for air vapor accumulation. The requirement for a minimum length of straight pipe upstream of the pump inlet that is based on the number of pipe size reductions as presented by ANSI/HI 9.6.6 is also not confirmed by the results. The velocity distribution resulting from a reducer is directly related to the reducer angle, not the number of pipe reductions. A five-pipe size reduction with a concentric angle of 10 degrees or an eccentric angle of 5 degrees will result in velocity distributions that fall within the ANSI/HI 9.8 10 percent velocity distribution criteria. A concentric reducer with an angle that is large enough to allow air to be hydraulically transported past it or an additional straight length of pipe to be placed after the eccentric reducer should be investigated as alternative options for the ANSI/HI 9.6.6 eccentric reducer prescription. The capacity of a reducer to hydraulically transport air through a concentric reducer can be determined similarly to the assessment of the hydraulic transportation of air through a pipeline with the same angle as that of the reducer. Van Vuuren, van Dijk and Steenkamp (2004) provides details on the assessment of the hydraulic transportation of air. References 1. ANSI/AWWA C208-07. 2008. Dimensions for fabricated steel water pipe fittings. American Water Works Association, Denver. 2. ANSI/HI 9.8-1998. 2000. American National Standard for pump intake design. Hydraulic Institute, New Jersey. 3. ANSI/HI 9.6.6-2009. 2009. American National Standard for rotodynamic pumps for pump piping. Hydraulic Institute, New Jersey. 4. Van Vuuren, S.J., Van Dijk, M. and Steenkamp, J.N. 2004. Guidelines for effective de-aeration of large diameter water pipelines. WRC Report No. 1177/2/04. Water Research Commission, Pretoria.