KSCE Journal of Civil Engineering (0000) 00(0):1-9 Copyright ⓒ2016 Korean Society of Civil Engineers DOI 10.1007/s12205-016-0714-z

Water Engineering

pISSN 1226-7988, eISSN 1976-3808 www.springer.com/12205

TECHNICAL NOTE

Numerical Analysis of the Effects of Anti-Vortex Device Height on Hydraulic Performance of Pump Sump Hyung-Jun Kim*, Sung Won Park**, and Dong Sop Rhee*** Received March 19, 2016/Accepted June 12, 2016/Published Online August 30, 2016

··································································································································································································································

Abstract Pump stations are very important flood control facilities in that they are designed to mitigate inundation in urban lowland areas. In urban areas, however, it is often difficult to acquire the land needed to increase the capacity of a pump station. Thus, the discharge capacity should be maximized despite these restricted conditions. The installation of a new pump with a larger capacity would be one way of increasing the capacity of a pump station. However, the addition of facilities could affect the flow characteristics, such as increasing the flow velocity and vorticity, thus causing severe problems with the operation of the pump. To overcome those problems, an Anti-vortex Device (AVD) can be applied. In Korea, the design of a pump station demands that engineers refer to advanced design criteria derived in other countries, owing to the lack of appropriate domestic design standards. In addition, the Korean design criteria for an AVD are based on experience, as there is insufficient available data. In this study, a numerical simulation was applied to simulate the swirl flow in a sump with various AVD heights to derive information that can be applied to studies aimed at improving the design and efficiency of pump station. Keywords: sump, pump station, vortex, anti-vortex device, swirl ··································································································································································································································

1. Introduction Climate change has led to an increase in the number of severe rainstorms, while ongoing urbanization means that much of the ground in a city is impermeable, such that flood damage is becoming a serious problem. As a result, there has been an increase in interest in research to improve flood safety of lowlands through the application of advanced drainage facilities. To improve flood safety in lowlands that are vulnerable to flooding, particularly in urban areas, it is necessary to design efficient pump stations for draining the inner basins. With the increase in the risk of flooding due to rainfall, pump stations are required to reduce the damage to urban areas. As such, they must be of an appropriate design and offer an adequate level of performance. The rainwater that falls in urban areas reaches the pump stations through the sewers, after which it enters the sump and is then pumped away and discharged to the foreland. Given this process, rainwater discharged through the pump is affected by the shape of the sump; thus, the design of the sump is an important factor affecting the design of a pump station. There are several factors affecting the design of the sump, such as the pump output and piping specifications, but owing to the difficulty

of acquiring land in urban areas, the sump dimensions will be limited by the location of the pump station. The performance of a pump station is generally affected by the size and shape of the sump. If parameters such as sump shape, water level, or output from the pump are inappropriate, a vortex will be generated in the vicinity of the pump suction side. The vortex generated near the pump suction entrains air from the free surface and this leads to a drop in the local pressure. Consequently, cavitation can occur in the pump, causing vibration and noise. Owing to these effects of a vortex, the efficiency of pump stations decrease relative to the designed performance (Kim and Kim, 2002). Therefore, this can act as a major factor for increasing the inundation risk in the lowlands of urban areas in the event of a flood. Many studies have been conducted in Korea and overseas with the goal of maintaining the efficiency of the pump by preventing the formation of a vortex in the sump. A review of the literature indicates that the sumps of pump stations have been studied extensively by Rho et al. (2002), who undertook a numerical analysis based on Computational Fluid Dynamics (CFD). Choi (2003) built an experimental model of a suction cistern and used Particle Image Velocimetry (PIV) to observe the flow characteristics in the sump. Park and Roh

*Researcher, Hydro Science and Engineering Research Institute, Korea Institute of Construction Technology, Goyang 10223, Korea (E-mail: [email protected]) **Member, Researcher, Hydro Science and Engineering Research Institute, Korea Institute of Construction Technology, Goyang 10223, Korea (E-mail: [email protected]) ***Member, Researcher, Hydro Science and Engineering Research Institute, Korea Institute of Construction Technology, Goyang 10223, Korea (Corresponding Author, E-mail:[email protected]) −1−

Hyung-Jun Kim, Sung Won Park, and Dong Sop Rhee

(2007) reproduced the flow of a Japanese-standard simple sump model as defined by the Turbomachinery Society of Japan (TSJ) by using a CFD model and observed the generation of a freesurface vortex and an underwater vortex. Kim et al. (2008) used the ANSYS-CFX model to analyze the effects of the gap between the underwater pump inlet and the sump on the suction performance. Choi et al. (2009) simulated the flow homogeneity of pump stations with multiple sumps by conducting a numerical analysis and further analyzed the effects of an Anti-Vortex Device (AVD) on the flow field. Choi et al. (2012) reproduced a benchmark experiment conducted by the TSJ by using a numerical model and analyzed the effects of the AVD on the pump efficiency by considering an additional AVD. Rajendran et al. (1998) installed a vertical suction pipe in a rectangular channel, measured the flow characteristics through a PIV experiment, and conducted a numerical analysis. Nagahara et al. (2003) conducted experimental and numerical analyses on the flow around the suction pipe due to the asymmetric flow in the direction of the water course in the suction cistern. The TSJ (2005) amended the standard for pump suction cisterns, and reviewed the applicability of numerical models to the same analysis target. Okamura et al. (2007) additionally conducted a hydraulic experiment similar to the TSJ benchmark test, and verified the accuracy of the numerical models. Shabayek (2010) verified the improvements in the flow characteristics by improving the shape of the suction side by conducting a hydraulic model experiment. As such, various studies were conducted to reproduce the flow characteristics in a pump station sump and to investigate how to prevent the formation of vortexes. However, the design guidelines relating to the AVD of a pump station are still insufficient. In Korea, there are river design criteria (KWRA, 2009) and sewage system criteria (ME, 2011). These, however, only suggest limited criteria for the size and location of the suction side as well as measures to avert the formation of vortexes. The AVDs currently in use in Korea were designed based on overseas design guidelines for sump facilities, but without the aid of any quantitative standards. Thus, there is a need for quantitative data that could be applied to the design of pump stations. Accordingly, the present study reproduced the flow formed in the sump with different sizes of AVD and went on to analyze the features.

2. Hydraulic Experiment and Numerical Simulation Reproducing a Vortex in a Sump 2.1 Benchmark Test The present study compared the results of a numerical simulation with those of hydraulic experiments that were conducted previously to verify the accuracy of the numerical model. Okamura et al. (2007) installed the channel and suction part of a design that was similar to that used in the hydraulic experiment conducted by TSJ (2005), and performed a hydraulic experiment generating a stationary surface vortex, a non-stationary surface vortex, and a stationary subsurface vortex. The width of the channel was 300

Fig. 1. Experimental Channel

mm, the diameter D of the suction pipe was 145 mm, and the inlet was installed 100 mm from the bottom of the channel, and 110 mm from the back. The center of the suction pipe was angled 10 mm to the right of the centerline of the channel, while the water level was 230 mm and the output was 1.0 m3/min (Fig. 1). 2.2 Benchmark Test Numerical Simulation To reproduce the flow in the suction cistern, this study used the ANSYS-CFX three-dimensional model. CFX is a three-dimensional flow analysis program developed by ANSYS, which enables the use of both structural and non-structural tetrahedral grids, thereby allowing efficient grid formation. It is also capable of accurate flow analysis for the contact surface between the solid and liquid, that is, the shear stress on the wall or bottom of the channel. Regarding the governing equation, it conforms to the continuity and momentum equations, that is, Eqs. (1), (2), and (3). ∂ρ ------ + ∇ ⋅ ( ρU ) = 0 ∂t

(1)

( ∂ρU ) --------------- + ∇ ⋅ ( ρU × U ) = – ∇p + ∇ ⋅ τ + SM ∂t

(2)

T 2 τ = μ ∇U + ( ∇U ) – --- δ∇ ⋅ U 3

(3)

Here, ρ is the density of water (kg/m3), U is the flow vector, p is the pressure, τ is a stress tensor related to the strain rate, and SM is the external momentum generation term. The calculation grid for the flow simulation in the sump was an adapted 3-D hexahedral-type grid, representing the target area as a grid of approximately one million nodes (Fig. 2). To reproduce the flow, the boundary conditions for the pressure and outflow at the inlet and outlet, respectively, were applied, and the Shear Stress Transport (SST) model for the turbulence model was applied to calculate the steady state. The SST model is an improved turbulence closure model that combines the advantages of the k-ε and k-ω models to achieve a wide range of applications. To this end, the SST model adopts a blending function F1, which is equal to 1 near the solid surface and equal to 0 for the flow domain where the wall friction is not dominant. It solves the k-ω equations in the near wall region and the k-ε equations for the

−2−

KSCE Journal of Civil Engineering

Numerical Analysis of the Effects of Anti-Vortex Device Height on Hydraulic Performance of Pump Sump

Fig. 2. Numerical Geometry and Mesh for Benchmark Test: (a) Analysis Geometry, (b) Computational Grid

rest of the flow. Through this approach, the attractive near-wall performance of the k-ω model can be used without the potential errors resulting from the free stream sensitivity of the model. In addition, the SST model also features a modification of the definition of the eddy viscosity, which can be interpreted as a variable cμ, where cμ in the k-ε model is constant (CarregalFerreira et al., 2002). This modification is required to accurately capture the onset of separation under pressure gradients. In a recent NASA Technical Memorandum, the SST model was rated the most accurate model in its class (Bardina et al., 1997). The SST model is similar to the standard k-ω model for which the equations are given as follows. ∂ ∂k ∂ ∂ ---- ( ρk ) + ------ ( ρkui ) = ------ ⎛ Γk ------⎞ + Gk – Yk ∂t ∂xj ⎝ ∂xj⎠ ∂xi

(4)

∂ ∂ ∂ ∂ω ---- ( ρω ) + ------ ( ρωui ) = ------ ⎛ Γω -------⎞ + Gω – Yω + Dω ∂xj ⎝ ∂xj⎠ ∂t ∂xi

(5)

In these equations, ρ is the density, k is the turbulent kinetic energy, and ω is the specific dissipation rate. Gk represents the generation of turbulent kinetic energy due to mean velocity gradients. Gω represents the generation of ω. Γk and Γω represent the effective diffusivity of k and ω, respectively. Yk and Yω represent the dissipation of k and ω owing to turbulence. Dω represents the cross diffusion term. Comparisons were made between the 3-D numerical results of the present study and those of a previous study that used STARCD, a commercial model for the analysis of the flow and vorticity components, as well as with the experimental results obtained by Okamura et al. (2007) using LS-PIV measurement

Fig. 3. Comparison of Numerical Results with Those of Previous Study: (a) x-directional Velocity Results, (b) y-directional Velocity Results, (c) z-directional Velocity Results, (d) Vorticity Results Vol. 00, No. 0 / 000 0000

−3−

Hyung-Jun Kim, Sung Won Park, and Dong Sop Rhee

at the y-axis observation line crossing the center of the inlet in the x-y plane. This was located at z = 85 mm from the channel bottom as indicated in Fig. 3. The x-axis flow velocity results are in better agreement with the observed flow velocity than previously simulated results, while the results for the y and z flow velocities exhibited marked improvements over the results obtained in previous studies. Regarding the vorticity, the results were considerably larger than those of previous numerical simulations and observed values for the right side of the channel, but the results were mostly similar to those of previous numerical simulations and observed values at the center or on the left. From a comparison of the numerical simulations, we can conclude that the numerical model implemented in this study accurately reproduces the flow characteristics at the suction side.

3. Flow Characteristics Changes Caused by AVD 3.1 AVD Apparatus Many pump stations encounter flow problems in intakes. One of the principal flow problems is a vortex in the pump sump. If the vortex problems are not resolved, the effectiveness of the intake is reduced and the operating costs for pump operation may increase. According to TSJ standard S002, the vortices in the pump sump can be classified as an air-entraining vortex and a submerged vortex. The air-entraining vortex, which is referred to as a free surface vortex starts at the free water surface and is sucked into the pump bell. When the vortex is fully developed, a full air core to the pump bell is established. The submerged vortex, which is referred to as a sub-surface vortex starts at the bottom surface or at the side walls of the sump and enters the pump bell. This is a form of vortex cavitation caused by low pressure at the vortex core (Okamura and Kamemoto, 2005). An AVD is intended to prevent vortex formation by improving the flow characteristics in the sump. This study addressed the floor cone (Fig. 4(a)) and fillet-splitter combined (Fig. 4(b)) AVDs, which are commonly adopted for pump stations in Korea. These were incorporated into the channel of the hydraulic experiment, and we observed the changes in the flow characteristics depending on the size after reproducing the changes through numerical

Table 1. Location of Lines for Results Comparison Height of surface from bottom of channel (mm) Line direction

Line 1

Line 2

Line 3

Line 4

25

50

75

100

y -axis crossing center of inlet

simulation. Floor cone AVDs are commonly installed in pump stations in Korea. However, the related design criteria do not provide quantitative guidelines for the installation height or the size, and research into the changes in the flow characteristics based on the installation size is also lacking. Thus, this study performed a numerical simulation by applying the conditions of d = 0.1, 0.2, 0.3, 0.4, and 0.5D of the installation height for AVDs, assuming that the diameter of the suction pipe does not change. A combined AVD is intended to improve the flow characteristics by grading the contact point of the channel, the wall of the suction cistern, and the shape of the bottom beneath the inlet. This type is rarely used in pump stations in Korea. This study undertook a numerical simulation, applying the conditions of d = 0.1, 0.2, 0.3, and 0.4D for the AVD height. To compare the performance for different heights of AVD, we set lines at different heights, as shown in Table 1, and compared the results of the flow velocity and vorticity in each case. 3.2 Analysis of Impact of AVD in Sump In the present study, we set out to analyze how the flow characteristics in the suction cistern changed depending on the AVD height, using a numerical simulation. Fig. 5 shows the results of the analysis and compares the vortex formation area, as simulated for each condition, together with the values of the swirl strength. Fig. 5(a) indicates the results for the case in which an AVD was not installed, showing that a vortex forms at the back of the inlet. Fig. 5(b)-(j) shows the vortex formation areas after performing simulations with different floor cone and filletsplitter AVD installation heights. These results show that the area in which the vortex forms, between the wall and the inlet, decreased as a result of the AVD installation. The vortex area was most significantly reduced with the floor cone AVD with an

Fig. 4. Anti-vortex Devices for Numerical Case: (a) Floor-cone Type AVD Geometry, (b) Fillet-splitter AVD −4−

KSCE Journal of Civil Engineering

Numerical Analysis of the Effects of Anti-Vortex Device Height on Hydraulic Performance of Pump Sump

Fig. 5. Comparison of Calculated Vortex Region: (a) Without AVD, (b) Floor Cone (d = 0.1D), (c) Floor Cone (d = 0.2D), (d) Floor Cone (d = 0.3D), (e) Floor Cone (d = 0.4D), (f) Floor Cone (d = 0.5D), (g) Fillet-splitter (d = 0.1D), (h) Fillet-splitter (d = 0.2D), (i) Fillet-splitter (d = 0.3D), (j) Fillet-splitter (d = 0.4D) Vol. 00, No. 0 / 000 0000

−5−

Hyung-Jun Kim, Sung Won Park, and Dong Sop Rhee

Fig. 6. Comparison of Calculated Vorticity with Floor Cone AVD: (a) z = 25 mm, (b) z = 50 mm, (c) z = 75 mm, (d) z = 100 mm

installation height of d = 0.1D, but as the installation height increased, the vortex area increased again. For the fillet-splitter AVD, the vortex area clearly decreased as the installation height increased. These results were obtained from an analysis of the qualitative trend through schematized results; thus, it was necessary to analyze the effects of the AVD based on a quantitative comparison. Comparisons of each case with the floor cone AVD installation are plotted in Fig. 6. We set the y-axis line crossing the center of the inlet on the x-y plane to z = 25, 50, 75, and 100 mm from the channel bottom and then compared the vorticity. Fig. 6(a) shows the results obtained with Line 1 from the channel bottom. The vorticity decreased after the installation when the AVD height was 0.1D, but increased along with the AVD height; particularly, at 0.2D, the vorticity rapidly increased at the center of the inlet. Fig. 6(b) shows the results for Line 2, for which the vorticity decreased when the AVD height was 0.1D, but increased in other cases, exhibiting the highest value at 0.4D. Fig. 6(c) shows the result for Line 3, which exhibits improved flow characteristics when the AVD height was 0.1D, as it did in the two aforementioned results. However, as the AVD height is increased at the center of the inlet, the vorticity also exhibited a rapid increase. Fig. 6(d) shows the results observed for Line 4 for an inlet installation height of z = 100 mm, for which the value of the vorticity is extremely high at both ends of the inlet. This result was due to the rapid change in the flow direction as there was a streamline from the outside to the inside of the inlet. The results obtained for the inlet center showed that the vorticity

gradually increased with the AVD height. Figure 7 shows the results of the analysis of the changes in the flow characteristics with the AVD installation by comparing the maximum values of the flow velocity and vorticity for each line. Fig. 7(a) indicates the maximum value of the x-directional flow velocity as calculated from each line based on the AVD height. As shown in the figure, the flow velocity decreased when the AVD height was 0.1D, but increased along with AVD height, relative to when the AVD was not installed. Fig. 7(b) shows the results of comparing the y-directional flow velocity. As can be seen from the x-directional result, the flow velocity decreased when the AVD height was 0.1D. Fig. 7(c) shows the result for the z-directional flow velocity, for which the trend is generally similar. In the vorticity comparison shown in Fig. 7(d), it can be seen that when the AVD height is 0.1D, the vorticity decreased. Based on the above results, we can conclude that if the installation height of a floor cone AVD exceeds the optimal value, it will reduce the flow area and thus increase the flow velocity and vorticity, which may adversely affect the flow conditions. Among the installation heights addressed in this study, 0.1D of the inlet diameter proved to be the optimal value for stabilizing the flow when operating the pump while reducing the flow velocity and vorticity. Next, the performance of an AVD combining a fillet-splitter was analyzed. Fig. 8(a) shows the change in the vorticity based on the AVD height for Lines 1-4. Fig. 8(a) shows the result obtained for Line 1, which shows that the vorticity on the righthand side, where the inlet is inclined, is high prior to the

−6−

KSCE Journal of Civil Engineering

Numerical Analysis of the Effects of Anti-Vortex Device Height on Hydraulic Performance of Pump Sump

Fig. 7. Comparison of Numerical Results for Floor Cone AVD: (a) x-directional Velocity, (b) y-directional Velocity, (c) z-directional Velocity, (d) Vorticity

Fig. 8. Comparison of Calculated Vortex with Fillet-splitter AVD: (a) z = 25 mm, (b) z = 50 mm, (c) z = 75 mm, (d) z = 100 mm

installation of the AVD, while that on the left-hand side is low. As a result of installing the AVD, the vorticity increased when the height was 0.1D, but as the installation height was gradually Vol. 00, No. 0 / 000 0000

increased, the vorticity decreased. Fig. 8(b) compares the results of the numerical simulation for Line 2. When the AVD height was 0.1D, the vorticity increased after the installation of the

−7−

Hyung-Jun Kim, Sung Won Park, and Dong Sop Rhee

Fig. 9. Comparison of Numerical Results with Fillet-splitter AVD: (a) x-directional Velocity, (b) y-directional Velocity, (c) z-directional Velocity, (d) Vorticity

AVD, with a large change in the vorticity being apparent throughout flow area. Fig. 8(c) shows the vorticity results for Line 3. When the AVD height was 0.1D, the vorticity increased significantly, adversely affecting the flow conditions relative to before the installation of the AVD. Fig. 8(d) shows the results for Line 4 of the inlet installation height, for which the values of the vorticity were large on both sides of the inlet, which was equivalent to the aforementioned results and for which the greatest vorticity was obtained when the AVD height was 0.1D. Figure 9 shows the maximum values for the flow velocity and vorticity in each direction for Lines 1-4, comparing the changes based on the fillet-splitter AVD height. Fig. 9(a) shows the results of comparing the x-directional flow velocity changes, indicating that the presence of the AVD generally reduced the flow velocity. Fig. 9(b) shows the results obtained for the ydirectional flow velocity, with the highest velocity being obtained when the AVD height was 0.1D, but otherwise exhibiting similar results for other conditions. Fig. 9(c) shows the z-directional flow velocity, in which case the AVD does not notably affect the flow velocity, thereby producing similar results overall. Fig. 9(d) shows the results for the vorticity where the reduction effects were excellent for Lines 1-3 which were some distance from the inlet, but for Line 4, the vorticity was similar owing to the effects at both ends of the inlet. The results of comparing and analyzing the flow velocity and vorticity for Lines 1-4 showed that, if the height of the fillet-splitter AVD was not optimal, it adversely affected the flow characteristics in the sump station, thereby hindering the pump operation.

4. Conclusions In this study, we conducted a numerical simulation of the flow inside the sump of a pump station with the ultimate goal of improving flood safety in areas vulnerable to inundation through inner basin drainage. We analyzed the changes in flow characteristics based on the height of the AVD installation. Prior to analyzing the AVD effect, a numerical model is applied to a benchmark case to compare experimental observations and the numerical simulation results. The results showed accurate and reasonable results; thus, it can be concluded that the numerical simulation is appropriate for regenerating the flow characteristics in a pump sump. Then, the numerical model adopted to simulate the changes in the flow characteristic according to the heights of the installed AVD. We installed a floor cone and fillet-splitter AVD with different heights and derived the following conclusions by quantitatively and qualitatively analyzing the results. 1. The installation of a floor cone AVD generally increases the flow velocity by reducing the flow area near the bell-mouth, but the optimal size to reduce flow velocity. The optimal size of a floor cone AVD was determined to be d = 0.1D. 2. As the flow velocity is increased, the vorticity near the bellmouth is also increased. Only the optimal size case showed improved flow characteristic results. This implies that to achieve the best operation conditions with a floor cone AVD, the height of the AVD has to be d = 0.1D. 3. The installation of a fillet-splitter AVD generally decreases

−8−

KSCE Journal of Civil Engineering

Numerical Analysis of the Effects of Anti-Vortex Device Height on Hydraulic Performance of Pump Sump

the flow velocity, though the AVD reduced the flow area near the bell-mouth. However, at the AVD height of d = 0.1D, the flow velocity increased. 4. As the flow velocity decreased, the vorticity near the bellmouth also decreased. Only the case of a fillet-splitter AVD height of d = 0.1D showed worse flow characteristic results. This implies that to achieve improved operation conditions with the fillet-splitter AVD, the height of the AVD has to be more than d = 0.2D. 5. It is necessary to install an AVD considering the optimal height, which differs according to the AVD shape. To install the floor cone AVD in the pump sump, the design height should be d = 0.1D. For the fillet-splitter AVD, the minimum height of installation is d = 0.2D. As described above, an AVD with suitable standards can help pump operation by improving flow characteristics in pump sump. According to the results of this study, different AVD types yielded different flow characteristic changes in the pump sump. To achieve appropriate flow conditions during pump operation, the optimized height of the AVD must be considered in design process. The floor cone AVD can improve the flow conditions when the installation height is 0.1 times the suction pipe diameter D, while the fillet-splitter AVD requires a minimum height of 0.2D for improving the flow conditions. From this study, it is clear that the appropriate installation of an AVD can improve the flow characteristics, such as flow velocity and vorticity in the pump sump, and secure stable conditions for pump operation. Furthermore, the present results can contribute to the suitable design of pump sumps.

Acknowledgements This study was funded by the Water Management Research Project by the Ministry of Land, Infrastructure, and Transport (13AWMP-B066744-01).

References Bardina, J. E., Huang, P. G., and Coakley, T. J. (1997). “Turbulence modeling validation, testing, and development.” NASA Technical Memorandum 110446. Carregal-Ferreira, J., Holzwarth, A., Menter, F., Esch, T., and Luu, A. (2002) “Advanced CFD analysis of aerodynamics using CFX.” AEA

Vol. 00, No. 0 / 000 0000

Technology GmbH, Otterfing, pp. 1-14. Choi, J. W. (2003). A study on the flow characteristics around intakes within a sump in a pump station by PIV (Master Dissertation), Korea Maritime University (in Korean). Choi, J. W., Choi, Y. D., Lim, W. S., and Lee, Y. H. (2009). “Numerical analysis on the flow uniformity in a pump sump model with multi pump intake.” Journal of Fluid Machinery, KSFM, Vol. 12, No. 4, pp. 14-22 (in Korean). Choi, J. W., Park, N. S., Kim, S, S., Park, S. S., and Lee, Y. H. (2012). “Study on performance analysis of pump within sump model with AVD installation by CFD.” Journal of Korean Society of Water and Wastewater, Vol. 26, No. 3, pp. 463-469. Kim, J. H., Choi, Y. S., and Lee, K. Y. (2008). “A numerical study on the suction performance of a submerged cargo pump.” Journal of Fluid Machinery, KSFM, Vol. 11 No. 6, pp. 18-23 (in Korean). Kim, S. W. and Kim, M. H. (2002) “Study on optimal determination of pump specification (I).” Proceedings of the KSFM Conference, pp. 69-74. Korea Water Resources Association (2009). Design criteria of the hydraulic structures and river management in rivers, KWRA (in Korean). Ministry of Environment (2005). Drainage Design Criteria, ME (in Korean). Nagahara, T., Sato, T., and Okamura, T. (2003). “Measurement of the flow around the submerged vortex cavitation in a pump intake by means of PIV.” Proceedings of the 5th International Symposium on Cavitation, Cav03-OS-6-011. Okamura, T. and Kamemoto, K. (2005). “CFD simulation of flow in model pump sumps for detection of vortices.” Proc. of 8th Asian International Fluid Machinery Conference, Yichang, China. Okamura, T., Kamemoto, K., and Matsui, J. (2007). “CFD Prediction and model experiment on suction vortices in pump sump.” Proceedings of the 9th Asian Conference on Fluid Machinery, pp. 110. Park, S. E. and Roh, H. W. (2007). “CFD Prediction on vortex in sump intake at pump station.” Journal of Fluid Machinery, KSFM, Vol. 10 No. 4, pp. 39-46 (in Korean). Rajendra, n V. P., Constantinescu, G. S., and Patel, V. C. (1998). “Experiments on flow in a model water-pump intake sump to validate a numerical model.” ASME, FEDSM98-5098. Roh, H. W., Kim, J. S., and Suh, S. H. (2002). “Flow analysis around within sump in a pump station using by the CFD.” Proceedings of the KSFM Conference, pp. 89-94. Shabayek, S. A. (2010). “Improving approach flow hydraulics at pump intakes.” International Journal of Civil & Environmental Engineering, Vol. 10, No. 6, pp. 23-31. Turbomachinery Society of Japan (2005). Standard method for model testing the performance of a pump sump, TSJ S002.

−9−