CFD Analysis of Flow in Pump Sump to Check Suitability

for Better Performance of Pump

Kadam Pratap M. & D.S. Chavan.

Department of Mechanical Engineering, R.I.T, Sakharale, Sangli, Maharashtra, India
E-mail: [email protected]

prevented and their disappearance must be verified by

Abstract – During the pump operation, abnormal flow
phenomena such as cavitation, flow separation, pressure model test in the construction of pump station.
loss, vibration and noise occur often by flow unsteadiness To reduce these vortices and for the advanced pump
and instability. Especially, free and subsurface vortices sump design with high performance, it is very important
containing air occurred in sump pumps seriously damage
to know the detailed flow information in sump system.
to pump system. According to certain in the JSME
Standard, the appearance of such vortices is not For this purpose, to date many researchers have made
permissible for ordinary pump sumps, experiments with experimental and numerical studies on the flow in
scale models usually have been performed to assess pump-sump
performance of sump or to improve the design
configuration of a sump.
Now days it is common practice in India , that
while engineering big lift Irrigation , Drinking water
This study attempts to model the flow characteristic in a scheme or Boiler feed pump unit ,Sump model study is
pump sump of physical model by using Computational conducted to confirm suitability of pump sump , also
Fluid Dynamics (CFD) code FLUENT. The experimental parallel CFD analysis is done to check suitability of
procedures include the data collection using a flow meter
sump. Initially while engineering pumping scheme
and swirl meter (Rotometer /Vortimeter). Two types of
measurements were conducted which are flow, and swirl general guidelines of appropriate standards HIS or
angle. A visual test that involves the dye tracing technique BHRA are used to decide the sump dimensions and then
was also carried out to characterize the flow. sump model study is conducted in presence of
manufacturer and customer representatives.
The CFD analysis is done at critical cases, Grid generation
is done in ICEM-CFD and numerical analysis are carried The flow conditions at entry to a pump depend upon
out in FLUENT , and flow is analyzed with the help of flow conditions in approach channel, sump geometry,
velocity stream lines and vector plot and velocity contour location of pump intake with respect to the walls,
at the entrance of pump chamber, in CFD-POST software. velocity changes and obstructions such as piers, screens
And concluded on experimental and cfd results
etc., and rotational tendencies in flow produced
Hence this work is aimed at determining the feasibility of upstream of the pump bays. Analytical determination of
commercial CFD software as a design optimization tool for the flow conditions in a sump is not an easy task due to
pump sumps. the complex nature of the flow. Moreover the analytical
Keywords : Sump, numerical simulation (CFD), streamline solution may not completely predict the actual
pattern conditions in the sump due to the assumptions made for
simplifying the analysis. Thus at present model studies
I. INTRODUCTION are the only tool for developing a satisfactory design of
a pump sump, yet numerical simulation is a very good
The main aim of sump is to provide water with
facility for reducing the time and cost involved in the
uniform velocity during the pump operation, abnormal
design process.
flow phenomena such as cavitation, flow separation,
pressure loss, vibration and noise occur often by flow II. METHODOLOGY
unsteadiness and instability. Especially, free and
The methodology includes experiments completed
subsurface vortices containing air occurred in sump
in a physical model along a series of numerical
pumps seriously damage to pump system. According to
simulations using a package CFD code. First, physical
the HI standard of Hydraulic Institute or JSME criteria
experiments were set up and completed using a simple
for a pump sump design, these vortices should be

ISSN (Print): 2321-5747, Volume-1, Issue-2, 2013

59
 International Journal on Mechanical Engineering and Robotics (IJMER)

pump intake model. A series of test were completed to

ensure the apparatus could provide experimental
conditions that represented flow approaching a typical
pump intake. Experiments were completed using this
simple pump intake model under various flow rates to
define a set of flow conditions with varying degrees of
vortex development.
Next, the series of simulations was completed using
a commercial CFD software package that is ANSYS 13-
FLUENT-V6, to characterize the flow in the vicinity of
a pump intake. The numerical model configuration was
developed to simulate the actual conditions in the
physical experimental setup. The approach for the
numerical modelling involved the initial use of the
parameters used in the physical experimental model, and Fig.1 Detail arrangement of pump-sump model
direct comparison between experimental results and
model predictions. It also included subsequent Principle of similarity
sensitivity analyses to address the sensitivity of the
model predictions to the inherent assumptions and As an experimental criterion of vortex occurrence
associated parameters. estimation, HI standard [8] is adopted. The adopted
principle of similarity between model and prototype is
III. EXPERIMENTAL METHODS as followings.
The Shirapur Lift Irrigation scheme in Solapur Same Froude-number velocity (observing free-
District in Maharashtra is proposed to irrigate 10000 surface flow): Froude No. (Fr) is kept the same for the
Hectares of Command area. The lift irrigation scheme model and prototype. The Froude-number is the ratio of
comprises of two Pumping stages with associated rising gravitational and inertial forces. The condition of free
mains. surface is fundamentally governed by the Froude No.
Overall performance of any lift irrigation scheme Using Froude criterion for essentially free surface flow
not only depends upon technical suitability of machinery Fm = Fp
but also on proper intake structure and design of a sump.
A faulty design of pump sump or intake is one of the Also,
major causes for unsatisfactory performance of pumping Vm / Vp = (Lm / Lp)1/2
plant. Adverse conditions of flow approaching the
intake sections may lead to occurrence of noise, If Lm / Lp = 1 / 10, then
vibrations and cavitation in the pumps and also cause
Vm / Vp = (1 / 10)1/2
reduction in pump efficiency. Though some standards
are available which furnish guidelines and Flow relation is given by, Qm / Qp = (1 / 10) 2.5
recommendations for ideal design of pump sump and
Hence, Vm = 0.3162 Vp
intake, in many cases these standards cannot be adopted
fully due to site constraints or any other governing Qm = 0.003162 Qp
factors. In such cases, it is essential to assess the
Hence for Shirapur Stage II., If Qp = 3276 cubic m/ hr.
performance of such sump which may not conform to
at duty point for prototype then Qm = 2.88 Lit/sec for
standard design conditions & intake in advance by
model pump.
conducting investigations on a geometrically similar
model.
IV. CFD STUDIES
In order to check the suitability of proposed
prototype sump at site, a geometrically similar model From a purely numerical perspective the geometric
sump with scale ratio 1:10 was constructed in Mild steel complexity of the problem is such that it demands the
sheets and angle frames and acrylic materials and run full power of modern computational fluid dynamics
with scale model pumps to study the flow patterns in the (CFD) to solve the equations of motion and turbulence
sump. The sump is to be suitable for 4 no’s .of models in domains that involve multiple surfaces.
horizontal centrifugal pumps. Fig 1 shows the detail Additional difficulties are associated with modeling free
drawing of the model setup. surface and vortex phenomenon, the physics of which is
not yet fully understood. Inspite of the practical

ISSN (Print): 2321-5747, Volume-1, Issue-2, 2013

60
 International Journal on Mechanical Engineering and Robotics (IJMER)

importance of the problem, the literature on numerical The computational investigations were performed
modeling of pump intake flows is rather limited. using ANSYS ICEM CFD 13.0 and ANSYS FLUENT-
V6 software’s. ANSYS ICEM CFD 13.0 was used for
Review of the available literature reveals that site
mesh generation while the analysis was done using
specific computational studies have been undertaken for
ANSYS FLUENT-V6. The inputs and outputs of both
certain projects [4]. The only generalized study has been
the software’s are in easily accessible formats enabling
conducted at IIHR [1],[2] and that too pertains to
full integration with any CFD software. For the CFD
development of CFD code for simple rectangular sump
model in the present study, volumetric meshing with
having one and two pumps. The developed software
unstructured tetra meshing option was adopted for grid
cannot be used as general software for CFD analysis of
generation in the pump sump geometry which is shown
any intake structure. In the present work an attempt has
in figure 3 In general the mesh generated for different
been made to simulate and predict the flow conditions
variants had about 8 to 12 lakh elements with the
such as vortices and swirl for multiple pump intakes in a
number of nodes varying from 1 to 2 lakhs.
single sump, with an aim to determine the viability of,
commercially available computational fluid dynamic
software – ANSYS FLUENT as an important design
optimization tool for intake sumps.

Geometry Of Computational Model

Computational study was conducted for the same
pumping system that is Shirapur Lift Irrigation scheme
of having four pumps, and space for one pump is
provided on one side for future scope. The layout
includes a leading channel, approach channel, forebay,
pump sump and intake. A model of the prototype at a
scale of 1:10 was used for hydraulic analysis. The
geometry of the simulation model starts with the inlet to
the sump followed by a short approach section and a
vertically sloping section which ends in an expanding
forebay. After the forebay is the rectangular portion of
the sump consisting of four identical pump bays. Fig.3 Tetrahedral grid of fluid domain.
Towards the end of each of the bay is placed the suction
pipe of a pump at required clearances from the The domain was specified as a Non-Buoyant,
boundaries. The length of the suction pipes is extended Stationary Fluid domain with working fluid as water at
above the sump boundary to some distance. 25°C and reference pressure as 1atm. The turbulence
model was selected as K-Є model. As boundary
conditions for the steady state calculation, zero pressure-
inlet condition at the inlet and averaged velocity of
0.7478 m/s at the outlet of pipe at 1F run condition and
1.4967 m/s at 2F run condition are applied considering
the equal velocity flow condition. Water level is kept to
Different levels that are at HFL and MDDL. As the wall
conditions, free-slip condition is applied to the upper
side of pump sump model passage. No-slip condition is
applied to side and bottom walls and at pipe outside and
inside.
The ANSYS FLUENT-V6 Solver module of
ANSYS-13 was used to obtain the solution of the CFD
problem. The solver control parameters were specified
in the form of solution scheme and convergence criteria.
Table 1 shows the fluent solution setting.
Fig.2 3-d model of sump
Figure 2 SHOWS 3-D model of sump which is
created in Pro-e wildfire-4. This model is transferred
into ICEM-CFD using .stp or .igs format.

ISSN (Print): 2321-5747, Volume-1, Issue-2, 2013

61
 International Journal on Mechanical Engineering and Robotics (IJMER)

TABLE 1 Details of FLUENT Solution settings V. ANALYSIS OF SIMULATION RESULTS

The results of the computational simulation can be
Sr.
Settings Choice analyzed using number of variables. In this study it has
No. been restricted to the comparison of results based on the
1 Simulation 3D pattern of streamlines of flow and the velocity profiles.
The streamline pattern (Figure 4) shows that a very
2 Solver Pressure Based large rotating mass of fluid is created in the forebay
Viscous standard k- portion but it not affects the smooth entry of water in
3 Model each pump due to the length of the forebay is more.
e, standard wall Fn
Also there flow is maximum on the side where space for
4 Material Water future expansion is kept blank, similarly in experimental
we get the maximum flow on blank pump chamber side
Pressure Velocity which is analyzed by using wool thread which is shown
5 Simple
Coupling scheme in figure 5. And there is no any surface or submerged
vortices in the pump chamber.
Green-Gauss Node
6 Gradient
Based
Discretization
7 STANDARD
Pressure
Discretization
8 First Order upwind
Momentum

Pressure – 0.3,
Density - 1 ,
Body Forces - 1,
Under-Relaxation
10 Momentum - 0.4
factors
and
Turbulent kinetic
Energy – 0.8
Fig. 4 Velocity stream lines starting from the inlet

11 Compute from Inlet

Upwind scheme was specified for the solution

while for convergence the residual target for RMS
values was specified as 10-6. With the above specified
convergence criteria it took about 30hrs for the solution
of one simulation variant.
The results of the CFD analysis were analyzed
using CFD-post, the post processor module of ANSYS -
13. Initially, numerical simulation of flow through the
sump model (scale 1:10) was performed and the results
was analyzed. Results of numerical simulation of flow
in the original model showed flow disturbances only in
the forebay but length of forebay is more so it can’t
affect the performance of sump. In experiment vortex
Fig.5 Photographic view showing flow pattern
was observed when water level reduces 150mm below
Minimum water level (MDDL) so this case was also Fig 6 shows the velocity vector plot at XY plane at
analyzed in the CFD and we get the location of vortex middle position which shows when water enters in
by using CFD. forebay there is hydraulic jump occurs in the forebay

ISSN (Print): 2321-5747, Volume-1, Issue-2, 2013

62
 International Journal on Mechanical Engineering and Robotics (IJMER)

due to back flow of water. Similarly in experimental Fig.8 shows velocity stream lines near and inside
study hydraulic jump occurs at near about same position the suction pipe which shows there is smooth entry in
in approach channel. But it does not affect smooth flow belmouth without any swirl.
at suction because the length of fore bay is more and due
Similarly Fig.9 show smooth stream line entry of
to this swirl vanishes up to pump chamber. Fig. 7 shows
water in bell mouth and pipe using wool thread.
photographic view of hydraulic jump at entrance of the
forebay in experimental study.

Fig. 6Velocity vector showing hydraulic jump in Forebay

Fig.9 Stream line flow at belmouth without swirl

Fig 10 shows velocity vector at top surface which

shows direction of water particles.

Fig. 7Photographic view showing hydraulic jump

Fig.10 Velocity vector plot

A comparison of the velocity vector plots inside the

pump column (Figure 11) also show that the velocities
have increased at the inlet as well as in the belmouth
section in the final variant of the model, however the
swirl has reduced.

Fig.8 Stream line flow at belmouth without swirl

ISSN (Print): 2321-5747, Volume-1, Issue-2, 2013

63
 International Journal on Mechanical Engineering and Robotics (IJMER)

Fig.13 Concentrated pressure drop at top surface at

Fig.11 Velocity vector near belmouth. water level below M.D.D.L
Fig.12 shows velocity contour at entrance of pump
chamber where maximum velocity of water at entrance
of pump chamber is 1.64e-2m/s which is within limit as
per standard which is 0.5 m/s.

Fig.14 Top view showing surface vertices at water level

150mm below MDDL

Fig.12 Velocity contour at entrance of pump chamber.

Fig 13 shows pressure distribution at top surface

when water level reduces 150 mm below MDDL, a
reduced pressure is concentrated near pipe 4 region
which means water particles concentrated at that point
where pressure is less or vortex is created at low
pressure region due to water particles flows towards less
pressure area with increased velocity. This can be also
seen with the help of velocity stream lines which are
shown in Fig.14 and Fig.15 which shows helical path of
streamlines that is vortex formation. Similarly in
experimental study when water level reduces below
MDDL, vortex occurs near the pipe 4 which is observed Fig.15 Velocity streamlines showing surface vortex at
by die injection. This is shown by Fig.16 water level 150 mm below MDDL

ISSN (Print): 2321-5747, Volume-1, Issue-2, 2013

64
 International Journal on Mechanical Engineering and Robotics (IJMER)

[3] Kshirsagar J .T. and Joshi S.G. 2003,

“Investigation of air entrainment-A Numerical
Approach,” Fluid Engineering Division’s
Summer Meeting 2003-45415, Proceedings of the
Forth ASME-JSME Joint Fluid Engineering
Conference, Honolulu, Hawaii.

[4] Joshi, S.G. and Shukla, S.N. (2000),

“Experimental and Computational Investigation
of Flow through a Sump”, Pumps & Systems
Asia 2000. Nakato Tatasuaki., Weinberger Marc.
and Logden Fred. (1994), “A Hydraulic model
study of Korea Electric Power Corporation
Ulchin Nuclear Units 3 and 4 circulating-water
Fig. 19 Photographic view showing position of vortex and essential-service-water Intake structure”,
when water level reduces below MDDL IIHR Technical Report No 370.
[5] W.T. Kang, K.H. Y and B.R. Shin “An
VI. CONCLUSIONS Investigation of Surface Vortices Behavior in
A uniform pump approaching flow existed in the Pump Sump” The 11th Asian International
pump chambers. Even water level goes below MDDL Conference on Fluid Machinery and the 3rd Fluid
up to 110mm there is no any surface or submerged Power Technology Exhibition
vortices are occurred this was indicated from the results [6] Jong- Woong Choi, Young-Do Choi , Chang-Goo
of pre-rotation tests. The commercial CFD package Kim and Young-Ho Lee “Flow uniformity in a
ANSYS FLUENT V6 was used to predict the three multi-intake pump sump model” Journal of
dimensional flow and vortices in a pump sump model. Mechanical Science and Technology 24 (7)
The CFD model predicts the flow pattern in detail and (2010) 1389~1400
the location, and nature of the vortices. However,
considerable post-processing of the basic data is needed [7] Shin, C.S., E.M. Greitzer, W.K. Cheng, C.S. Tan,
to fully comprehend the details of the flow. Thus CFD and C.L. Shippee, 1986, “Circulation
model can be used to study the effect of various measurements and vortical structure in an inlet-
parameters which reduces time as well as cost and hence vortex field”, Journal of Fluid Mechanics, Vol
can become an important tool for optimization of pump 162, Pp 463-487.
sump geometry. [8] Hydraulic Institute Standard, 1998, American
National Standard for Pump Intakes Design,
VII. REFERENCES ANSI/HI 9.8.
[1] Constantinescu G.S., and Patel, V.C. (1998), [9] JSME, 1984, Standard Method for Model Testing
“Numerical model for simulation of pump-intake the Performance of a Pump Sump, JSME S 004.
flow and vortices”, ASCE Journal of Hydraulic
Engineering, Vol.124, No.2, 123-124. [10] ANSYS, 2013, Fluent theory guide, Version 13
[2] Constantinescu1, G.S. and Patel, V.C. (2000),
“Role of Turbulence Model In Prediction of 
Pump-Bay Vortices”, ASCE Journal Of
Hydraulic Engineering, Vol.126, No.5, 387-390.

ISSN (Print): 2321-5747, Volume-1, Issue-2, 2013

65