Page 1
Optimization of Centrifugal Pump Used as Impulse Turbine
by
Pattiya Thongkruer
A thesis submitted in partial fulfillment of the requirements for the
degree of Master of Engineering in
Industrial and Manufacturing Engineering
Examination Committee: Assoc Prof. Erik L. J. Bohez (Chairperson)
Dr. Mongkol Ekpanyapong
Prof. Stanislav S. Makhanov (External Export)
Nationality: Thai
Previous Degree: Bachelor of Engineering in Electrical Engineering
Thammasat University
Thailand
Scholarship Donor: Thailand (HM King)
Asian Institute of Technology
School of Engineering and Technology
Thailand
May 2015
Page 2
ii
ACKNOWLEDGEMENTS
It has been a strikingly experience studying at Asian Institute of Technology where they
supply me not only Industrial and Manufacturing Engineering knowledge, but also spending
life experience with various nationalities and daring group works.
My special gratitude will be granted to Assoc. Prof. Erik L. J. Bohez, my thesis advisor for
giving me suggestions and supporting and providing me an opportunity to conduct my
thesis work on. I am very respectful having such an open-mind and professional advisor for
my thesis as well. Besides, I would like to thank all committee members, Dr. Mongkol
Ekpanyapong and Prof. Stanislav S. Makhanov for precious recommendations.
Last but not least, I would like to express my appreciation to my family members who have
been supporting me all my life. They work hard in order to support my entire life to success
in every step. They always abet me when I am discouraged and try to solve a problem when
I am faced with on matter how hard it is.
Page 3
iii
ABSTRACT
Pump as Turbine (PAT) is one of the ways to reduce the equipment cost in micro
hydropower projects and is one attractive alternative to conventional hydraulic turbine.
However, the drawback of PAT is lacking of flow controlling device, which causes
rapidly dropping performance when PAT operate at off-design point. Thus, the new
concept by using the flow regulating device namely, spear valve, nozzle and deflector
will be proposed in this study. The objective in this research is to optimize the pump
working as turbine in Impulse mode by using the flow controlling device. To optimize
Impulse PAT, the geometries of a Centrifugal pump have to be known. However, the
pump vendors are reluctant to supply or give typical geometries of their pump. In order
to achieve this study, there are two main processes as follow: pump selection and
Impulse PAT optimization. The pump’s geometries derived from pump calculation
based on their given geometries and characteristic curves. These calculated geometries
will be used to be the input parameter to optimize the Impulse PAT. In the optimization
procedure, there are two options to inject the water in different direction. In other
words, when the water jet velocity is lower than rotational speed, the first option will be
applied in spite of second option. The result from optimizing Impulse PAT shows that
the second option namely, when velocity of water jet is higher than rotational speed is
solely adapted in pump operating as Impulse turbine. Moreover, the result shows that
the maximum efficiency of Impulse PAT is occurred when the rotational speed of
Impulse PAT performs nearly half of water jet velocity.
Keywords: Pump as Turbine, Conventional Hydraulic turbine, Characteristic
curves, Impulse PAT, Optimization, Flow regulating device
Page 4
iv
TABLE OF CONTENTS
CHAPTER TITLE PAGE
TITLE PAGE i
ACKNOWLEDGEMENTS ii
ABSTRACT iii
TABLE OF CONTENTS iv
LIST OF FIGURES vi
LIST OF TABLES viii
LIST OF ABBREVATIONS ix
1 INTRODUCTION 1
1.1 Background 1
1.2 Statement of the problem 2
1.3 Objectives of the study 2
1.4 Scope and limitation 2
2 LITERATURE REVIEW 3
2.1 Pump as turbine (PAT) 3
2.1.1 Optimization in PAT 5
2.1.2 Selection of Pump using as turbine 6
2.2 Turbine 7
2.2.1 Impulse turbine 7
2.2.2 Jet wheel 9
2.3 Centrifugal pump 10
2.3.1 One dimensional calculation with velocity triangles 10
2.3.2 Euler’s pump equation 12
2.4 Pump losses 13
2.4.1 Shock losses (Incidence losses) 15
2.4.2 Friction losses 16
2.4.3 Recirculation losses 17
2.4.4 Leakage losses 17
2.5 Pump as Impulse turbine 18
2.6 Conclusion 18
3 METHODOLOGY 21
3.1 Research background 21
3.2 The overview of research concept 21
4 PUMP SELECTION PART 26
4.1 The overview of selection part 26
4.1.1 Data acquisition 27
4.1.2 Model the total losses 27
4.1.3 Estimate the losses in selected pump 28
4.1.4 Algorithm to identity geometry of selected pump 30
4.1.5 Find the number of water jet 31
4.2 Case study 32
Page 5
v
CHAPTER TITLE PAGE
5 OPTIMIZATION OF IMPULSE PAT IN CASE STUDY 38
5.1 The overview of optimization of Impulse PAT 38
5.2 The algorithm to optimize Impulse PAT 40
5.3 The results of Impulse PAT optimization 43
5.3.1 The results of first option without friction losses 43
5.3.2 The results of second option without friction losses 43
5.3.3 The results of first option with friction losses 46
5.3.4 The results of second option with friction losses 46
5.4 Conclusion 46
6 HYDRAULIC TURBINE DESIGN 50
6.1 Spear valve and nozzle design 50
6.2 The 2D and 3D model of regulating flow rate device 52
6.3 Performance of spear valve and nozzle 56
6.4 Performance of deflector 57
6.5 Position the water jet with a selected pump 60
7 CONCLUSION AND RECOMMENDATION 65
7.1 Conclusion 65
7.2 Recommendation for case study 66
REFERENCES 67
APPENDIXES 69
Page 6
vi
LIST OF FIGURES
FIGURE TITLE PAGE
Figure 2.1 Direction of flow in sectional view of PAT 3
Figure 2.2 Pump impeller inlet and outlet velocity vector diagram 4
Figure 2.3 Turbine impeller inlet and outlet velocity vector diagram 4
Figure 2.4 Block diagram of the consolidated model for pumps as turbines
with optimization routine [6]
6
Figure 2.5 Flow chart for the PAT selection model [6] 6
Figure 2.6 Uncertainly bands for and 7
Figure 2.7 Impulse turbine 8
Figure 2.8 (a) Velocity diagram top view of bucket (b) Typical velocity
diagram [12]
9
Figure 2.9 Velocity vector diagrams of the Inward-flow jet wheel 9
Figure 2.10 The fluid path through the centrifugal pump in side and
frontal view [9]
10
Figure 2.11 (a) Velocity triangles diagram of centrifugal pump impeller
(b) Velocity triangles at outlet and inlet impeller [11]
11
Figure 2.12 Control volume for an impeller [11] 12
Figure 2.13 Hydraulic loss curve [17] 14
Figure 2.14 Reduction of theoretical Euler head due to losses [11] 15
Figure 2.15 Shock loss at inlet of impeller [11] 15
Figure 2.16 Shock loss as function of the flow curve [11] 16
Figure 2.17 Friction losses as function of the flow curve [11] 16
Figure 2.18 Recirculation losses in impeller [11] 17
Figure 3.1 Flow chart of research concept 22
Figure 3.2 Pump selection on Grundfos website [14] 24
Figure 3.3 The example of typical data and characteristic curve of selected
pump from Grundfos website [14]
25
Figure 4.1 A flow chart of pump selection part 26
Figure 4.2 Flow chart of calculating pump’s geometries of inlet of impeller 28
Figure 4.3 Flow chart of calculating total losses in selected pump 30
Figure 4.4 Flow chart of calculating remaining pump’s geometries 31
Figure 4.5 Given data of selected pump from Grundfos website [14] 33
Figure 4.6 Given data of a new selected pump from Grundfos [14] 36 36
Figure 5.1 Velocity diagram of a pump with backward blade 38
Figure 5.2 Impulse PAT with reverse operation (first option) 39
Figure 5.3 Impulse PAT with backward mode (second option)39 39
Figure 5.4 Flow chart of optimization process 42
Figure 5.5 Graph between some geometry at inlet and rpm of
Impulse PAT without friction losses
45
Figure 5.6 Graph between some geometry at inlet and rpm of
Impulse PAT with friction losses
48
Figure 5.7 Graph between friction loss in Impulse PAT and rpm 49
Figure 6.1 The means of controlling the speed of Pelton turbine with
a spear valve [28]
50
Figure 6.2 Description of spear valve and nozzle [27] 51
Figure 6.3 Dimensions for the spear valve and nozzle [27] 51
Figure 6.4 Deflector operation [25] 52
Page 7
vii
FIGURE TITLE PAGE
Figure 6.5 Dimensional drawing for the spear valve and nozzle 53
Figure 6.6 Dimensional drawing for the deflector 53
Figure 6.7 3D model of spear valve 54
Figure 6.8 3D model of nozzle 54
Figure 6.9 3D model of deflector 55
Figure 6.10 Assembly of spear valve, nozzle and deflector 55
Figure 6.11 The position of spear valve [28] 56
Figure 6.12 Graph between the flow rate and stoke of needle 56
Figure 6.13 The angle of deflector 57
Figure 6.14 The angle of deflector in different position 58
Figure 6.15 Fully deflected position 58
Figure 6.16 Half deflected position 59
Figure 6.17 Completely open deflector 59
Figure 6.18 Graph between the flow rate from jet and angle of deflector 60
Figure 6.19 The diameter of scroll 61
Figure 6.20 The position of 2 water jets with a selected pump 61
Figure 6.21 The position of 2 water jets with a selected pump in side view 62
Figure 6.22 The position of 2 water jets with a selected pump in top view 62
Figure 6.23 The position of 2 water jet when deflector completely opens 63
Figure 6.24 The position of 2 water jet when deflector completely closes 63
Page 8
viii
LIST OF TABLES
TABLE TITLE PAGE
Table 2.1 Summary of the governing equations for simulation 5
Table 2.2 The different types of losses in pump and their affect
on characteristic curve
14
Table 2.3 List of researches related to PAT 19
Table 4.1 Proportion of each loss in shut-off point 28
Table 4.2 Input parameter of selected pump from Grundfos for using to
calculate the remaining geometries
32
Table 4.3 Results from pump selection procedure 34
Table 4.4 Results from calculating diameter and the number of water jet 35
Table 4.5 Input parameter of a new selected pump from Grundfos for
using to calculate the remaining unknown geometries
35
Table 4.6 Results from a new pump selection procedure 37
Table 4.7 Results from calculating new diameter and the number of water
jet
37
Table 5.1 Impulse PAT performance evaluation in second option without
friction losses
44
Table 5.2 Impulse PAT performance evaluation in second option with
friction losses
47
Table 6.1 The results of spear design 52
Page 9
ix
LIST OF ABBREVIATIONS
Nomenclature Description Unit
A Area m2
b The width of impeller m
D Diameter m
g Gravitational acceleration m2/s
H Head m
k Friction coefficient
N Rotational speed rpm
P Power kW
Q Flow rate m3/s
u Peripheral speed m/s
v Absolute velocity m/s
w Relative velocity m/s
Greek symbols Description Unit
α Absolute velocity angle degree
β Relative velocity angle degree
η Efficiency %
ρ density Kg/m3
Specific weight N/m3
ω Omega radius
Discharge number
Turbine specific diameter m
Subscripts Description
0 At exit of pump
1 Inlet of impeller In pump mode
2 Outlet of impeller In pump mode
3 Inlet of impeller In turbine mode
4 Outlet of impeller In turbine mode
f Friction losses
j Jet
l Leakage losses
r Radial
s Shock losses
t Tangent
re Recirculation losses
BEP Best efficiency point
Page 10
1
CHAPTER 1
INTRODUCTION
1.1Background
Energy plays a key role in almost all many countries especially developing countries.
However, high price, fast depletion rate and environmental implication of fossil fuels create
many problems in electricity generation; worldwide, about 67% of electricity is generated
from fossil fuels [1]. The appropriate solution for these problems is generating electricity
though renewable sources [2].
“After the oil crisis in recent years, the small hydroelectric power generating stations became
attractive because the cost of energy produced by these stations is higher than large capacity
hydroelectric power plants. In order to reduce the cost of produced electrical energy, simple
turbine is used.” However, one of the efficient alternatives for generating and recovering
power through small and micro hydropower schemes is reverse running centrifugal pump (or
normally calls Pump-as-Turbine). A Centrifugal pump is a common hydraulic machine and it
is easily available at all places. “In addition, the maintenance and installation costs of
centrifugal pumps are less. Pump used as turbines (PATs) are available in wide variety of 1.7
kW and 160 kW ranges [16] and the cost of PAT is less than that of conventional turbine
around 50% [15].” The objective of using a Centrifugal pump is to transportation of liquid,
industrial process, heating systems and so forth. “Besides, handling water, pump, however,
can be used to generate electricity when operate in reverse direction. ” Using reverse running
centrifugal pump for generating electricity is not new. The researches related to this field had
been started around 80 years ago [3, 4].
“Although pump working as turbine (PAT) can replace many goal-built turbines in the small
and micro hydro turbine system, the main limitation of centrifugal pump running as turbine
mode is lower efficiency at part load operating condition as compared to simple turbines. ” To
improve the utilization of PAT, there are many researchers that have attempted variety of
modification to improve performance of the PAT by carrying out experimental and numerical
investigation. For example, some researchers demonstrated relations of best efficiency points
between pump and turbine based on experimental data and theoretical analyses. “The results
from these relations had almost ±20% deviation from experimental data [5]. ”
In order to defeat the challenges of these problems in PAT, few researchers tried to propose
the optimization of PAT. For instance, Singh and Nestmann [6] presented an optimization
routine that improve the accuracy and reliability of prediction and selection models on 9
PATs covering a specific speed range of 20 to 80 rpm.
Normally, pump vendors do not provide characteristic curve when pumps are worked in
turbine mode. “This problem causes difficulty to select an appropriate pump to use as turbine
for a specific operating condition ”. Thus, many researchers have been studied predicting
performance of PAT based on its pump mode performance of pump geometric parameters.
However, the PAT behavior is complex and difficult to find a relation to cover all pumps
behavior in reverse mode [22].
Page 11
2
1.2 Statement of the problem
Currently, pump working as turbine are normally a type of reaction turbine. Besides, this type
of PAT is studied in many researches in order to predicting performances of PAT by
fulfilling experimental and numerical exploration. Although PAT can replace many goal-built
turbines in the small and micro hydro turbine system, the main drawback substitute of
common PAT is its inability to regulate flow because simple centrifugal pumps do not have
guide vanes in the impeller casing as the case with purpose-built turbines”. Absence of guide
vanes in common centrifugal pumps make governing of PAT by controlling flow not
possible. “Thereby, one of improvement that can be made on PAT is development of a flow
regulating system to be retrofitted into the simple centrifugal pump ”. Hence, the purpose
method in this thesis is using the spear valve with nozzle from impulse turbine couple with
impeller from centrifugal pump running as turbine. There are no literatures that study in this
method until now. Moreover, before predicting performance of PAT, the geometries of the
selected pump using as turbine are known. There are almost no literatures present how to get
the geometries of selected pump. They only show the geometries of the selected pump that
they have or follow the geometries of selected pump from other researches. Thus, this thesis
will be focus on the procedure for selecting a Centrifugal pump to work as Impulse turbine as
well.
1.3 Objectives of the study
The specific objectives are:
1) optimization and selection of a centrifugal pump working as an impulse turbine for
given flow rate and head
2) mechanical design Hydraulic power system for PAT
1.4 Scope and limitation
The specific scopes and limitations are:
1) develop 1D geometry calculation of pump by using losses and Head-Capacity curve
2) develop 1D geometry optimization model of selected pump working as Impulse
turbine
3) consider only pump from Grundfos company
4) consider only the inward-flow jet wheel
5) not include force and torque computation and computational fluid dynamic (CFD)
Page 12
3
CHAPTER 2
LITERATURE REVIEW
In this chapter, a Centrifugal pump working as turbine (PAT), the theoretical foundation of
existing Impulse turbine and Centrifugal pump and the losses in the centrifugal pump and
PAT in impulse mode will be described. Also, this chapter will express how they operate and
how to estimate the losses in pump. Moreover, the PAT in impulse mode will be introduced
as well.
2.1 Pump as turbine (PAT)
Pump-as-Turbine is a pump running in turbine mode by changing direction of flow as shown
in figure 2.1 and also the direction of rotation of the impeller as shown in figure 2.3. When a
pump operate in normal mode, the water will flow from the eye of pump namely; a centre of
a pump and go along the blades then flow out at the outlet of impeller. Figure 2.2 shows the
velocity vector diagram of pump in normal mode. From figure2.2, the number 1 and 2 are
represented to the inlet and outlet of impeller of a pump. An increase in fluid pressure is rose
by transferring mechanical energy from the shaft of motor to the fluid through the rotating
impeller. Whereas, in PAT mode, the pump will rotate in reverse direction namely, the water
into high pressure will flow into the pump at the outlet of impeller and flows through impeller
blades to extract energy from the fluid and transfer most of that energy to form of mechanical
energy output and then flows out at the eye of pump at low pressure [10] as shown in figure
2.1. Figure 2.3 shows the velocity vector diagram of PAT. The number 3 and 4 are
substituted inlet and outlet of direction that water flow into PAT.
Figure 2.1: Direction of flow in sectional view of PAT
Page 13
4
Figure 2.2: Pump impeller inlet and outlet velocity vector diagram
Figure 2.3: Turbine impeller inlet and outlet velocity vector diagram
Page 14
5
The absolute flow angle position 2 (α2) can be approximated as same for both modes of
operation in respect to the shape of the volute. This angle can be estimated as the volute
angle. In turbine mode, figure 2.3 shows the blade-outlet flow angle (β3) can be assumed
closed to the blade-inlet angel (β1) in pump mode when whirl does not happen at exit [7].
Yang et al [26] presented a theoretical method for prediction of PAT performance. The
mentioned method is developed based on theoretical analysis and empirical correlation on the
basis of its performance of pump. The authors used some data reported such as head ratio and
discharge ratio from other researches. Moreover, the graphs head and discharge of PATs are
presented versus specific speed and maximum efficiency of pump for finding out head ratio
and discharge ratio equations of empirical correlation.
2.1.1 Optimization in PAT
Singh [18] presented various possibilities of modifying the pump geometry to improve the
performance of a given pump in turbine mode. The study showed that impeller rounding was
the most beneficial modification.
“Singh and Nestmann [6] presented an experimentally validated optimization routine for the
turbine-mode operation of centrifugal pumps.” “The optimization routine improves upon
previous uncertainties without altering the philosophy of the methodology that is used in the
basic PAT, as shown in table 2.1.” The basic PAT model comprises of the prediction model
and selection model, whose functions are briefly illustrated in Figure 2.4. The prediction
model performs the complete synthesis of the turbine characteristics of a given pump of
known shape and size by evaluating the BEP as well as the no-load points using the
governing equations summarized in table 2.2, whose origin comes from experimental tests on
corresponding 13 PATs. In the similar way, the governing equations for the basic prediction
model based on 9 PAT data are also developed.
Table 2.1: Summary of the governing equations for simulation
Patel et al [8] researched the installation of the fixed guide vanes inside the volute casing of
pump for improve efficiency of PAT. Moreover, they designed additional space for having
enough space for guide vane mechanism providing. The guide vane positions were optimized
by carrying out numerical simulation of casing by putting the guide vane at different angles.
The results of optimum angle is that the loss in kinetic energy in the casing was found to be
minimum at fixed guide vanes angle have an angle of 75 for the PAT.
Page 15
6
Figure 2.4: Block diagram of the consolidated model for pumps as turbines with
optimization routine [6]
Figure 2.5: Flow chart for the PAT selection model [6]
2.1.2 Selection of Pump using as turbine
Singh and Nestmann [6] also designed selecting the most appropriate pump working as
turbine models. The fixed input parameters namely the head and flow are used along with the
turbine speed, which is the control parameter, to determine the turbine mode specific speed
(Nqt) and Mean Cordier PAT Line (). The pump shape (N) can define by using the
experimental relationship between the turbine mode specific speed (Nqt) and the pump mode
specific speed (Nqp), given by equations in this study. The pump size (range of specific
diameter () and diameter (D)) can find by using the Cordier PAT line and uncertainly bands
for and as shown in figure 2.5. And then, these final output parameters of selection model
namely, pump shape and pump size are used to select the pump from manufacturer’s
catalogue.
Page 16
7
Figure 2.6: Uncertainly bands for and
2.2 Turbine
“Turbines have been used to change utilizable mechanical energy from rivers and wind to be
mechanical work via a rotating shaft for many years. ” Hydro turbines can be classified two
types namely, impulse and reaction. In this thesis, impulse turbine is focused on. Some detail
of impulse turbine as below.
2.2.1 Impulse turbine
“Impulse turbine composes of a set of nozzles and a row of blades. ” In an impulse turbine, the
available mechanical energy from the fluid sent through a nozzle is converted into kinetic
energy. And then the fluid flowing from the high-speed jet will impact bucket that the kinetic
energy is transferred to be turbine shaft as shown in figure 2.7[9]. The bucket converts
momentum-flow direction that gives an impulsive to bucket.
The conservation mass is applied to analyze the force F and power out P of this turbine
that are defined as
1 cosjF Q V u (2.1)
1 cosjP Fu uQ V u (2.2)
Where u r is the bucket linear velocity and r is the wheel radius.
Page 17
8
Figure 2.7: Impulse turbine
As compared to wheel radius, the size of bucket and the diameter of jet are small.
Consequently, 1r and 2r is approximated as r shown as figure 2.7. According to figure 2.7,
the velocity of reflected flow is turned through angle and is equal jV r . If 180 , the
power is maximum theoretically. Nevertheless, if these situations occur, the reflective fluid
runs into the back side of its neighbor coming along behind it. These cause decreasing torque
and power. Therefore, should equal around 160 165to in practice. The efficiency factor
because of 180 is
,
,
1 cos
1 cos180
shaft actual
shaft ideal
W
W
(2.3)
From equation 2.15 the theoretical power of an impulse turbine is maximum when 0dP du
that mean speed u equal as
1
2ju r V (2.4)
For a perfect nozzle, the entire available head would be converted to jet velocity 2jV gH .
Actually, since there are 2 to 8 percent nozzle losses, a velocity coefficient vC is used
2j vV C gH (2.5)
Page 18
9
Figure 2.8: (a) Velocity diagram top view of bucket (b) Typical velocity diagram [12]
2.2.2 Jet wheel
Normally, jet wheel can be divided 3 types depending on the general path of the water
through runner as [13];
a. Axial-flow jet wheel
b. Inward-flow jet wheel
c. Outward-flow jet wheel
In this thesis, only the inward-flow jet wheel, however, will be focus. In this type of jet
wheel, the water will enter from outside of wheel to inside of wheel as shown below figure.
Figure 2.9: Velocity vector diagrams of the Inward-flow jet wheel
From figure2.9, the subscript 1 and 2 refer to outlet and inlet, respectively. If all losses are
ignored, the head can be calculated as 2 2
1 2
2 2
Q v Q vH
g g
(2.6)
Page 19
10
From triangle law and above velocity diagram in figure2.9, the Euler law equation from A to
B is 2 2 2 2
1 2 1 2
2 2 2
w w u u
g g g
(2.7)
If friction losses between inlet and outlet of turbine is considered (2
2
2f
wk
g), the equation2.9
will be changed as 2 2 2 2 2
1 2 1 2 2
2 2 2 2f
w w u u wk
g g g g
(2.8)
Where fk is friction coefficient
2.3 Centrifugal pump
Centrifugal pumps are turbomachines that is used in order to transporting liquids by means of
increasing the fluid pressure level from the inlet to outlet when the pump is operating. The
fluid flows into suction side of pump at low pressure and absorbs energy by the impeller and
pump transfer most of that energy to the fluid, usually via a rotating shaft. The increase in
fluid energy is usually an increase in the pressure of the fluid. The fluid path through the
centrifugal pump is shown in Figure 2.1.
Figure 2.10: The fluid path through the centrifugal pump in side and frontal view [9]
2.3.1 One dimensional calculation with velocity triangles
Velocity triangle is used for predicting performance of a centrifugal pump in connection with
changes of e.g. speed, impeller diameter and blade angle [10].
The velocity diagram is shown in Figure 2.2. The fluid is assumed to run into the impeller at
1r r . The circumferential speed 1 1u r matches the tip speed of the impeller. The relative
velocity 1w is the fluid velocity compared to the rotating impeller. Hence, its absolute
Page 20
11
entrance velocity is the vector sum of 1w and 1u , shown as 1C (also called 1V ) . Similarly, the
flow exits at 2r r , speed 2 2u r , relative velocity 2w and sum of both 2C (also called 2V ).
Figure 2.11:(a) Velocity triangles diagram of centrifugal pump impeller (b) Velocity
triangles at outlet and inlet impeller [11]
Page 21
12
The angles 1 and 2 is defined as the angle of departure of the absolute velocity vector ( C )
from circumferential speed ( u ). In addition, the angles 1
and2
is defined as the angle
between the relative velocity ( w ) and the circumferential speed ( u ).Velocity triangles can be
illustrated in two different ways and both ways are shown in figure 2.11a and b. As seen from
the figure the same vectors are repeated. Figure 2.11a shows the vectors compared to the
blade, whereas figure 2.11b shows the vectors forming a triangle. By drawing the velocity
triangles at inlet and outlet, the performance curves of the pump can be calculated by means
of Euler’s pump equation which will be described in section 2.2.2.
2.3.2 Euler’s pump equation
“Euler’s pump equation is the most important equation in connection with pump design. ” The
equation can be derived in many different ways. This described method includes a control
volume which limits the impeller, the moment of angular momentum equation which
describes flow forces and velocity triangles at inlet and outlet [11]. To evaluate the rotating
shaft torque, an angular momentum to a turbomachine relation for control volume is an
applied. Figure 2.3 show a control volume between 1 and 2.
Figure 2.12: Control volume for an impeller [11]
Euler’s turbine equation is
2 2 1 1t tT Q r C rC (2.9)
Where 2tC and 1tC are the absolute circumferential velocity components of the flow.
Page 22
13
The brake hoursepower can be found as
2 2 1 1bhp t tP T Q u C u C (2.10)
Where is the shaft angular velocity. From previous equation, net head can be delivered as
2 2 1 1
1t tH u C u C
g (2.11)
From figure 2.2, the geometry equation is obtained as 2 2 2 2 cosw u C uC and
costC C Substituting this in equation (2.3)
2 2 2 2 2 2
2 1 2 1 2 1
1
2H C C u u w w
g (2.12)
Where tC is velocity paralleled to the blade surface that is defined as
cott rC u C (2.13)
According to conservation of mass, the volume flow rate at area 1 and 2 is same. Thus the
normal components of velocity at 1r and 2r is defined as
1
1 12r
QC
rb (2.14)
2
2 22r
QC
r b (2.15)
Where 1b and 2b are the blade widths at inlet and exit.
The power delivered to the flow is
P gQH (2.16)
Normally, bhp is less than P due to losses. The efficiency of pump is defined as
shaft
P gQH
bhp T
(2.17)
The design flow rate is normally estimated by assuming that the fluid enter the impeller is
mostly axial 1 90 so the circumferential component of the absolute inflow velocity is
1 0tC
cott rC C (2.18)
Hence, the bhp can be rewrite as
2 2 2 1 1 1cot cotr rbhp Q u C u C (2.19)
If the parameters 1 2 1 2, , ,r r and are known, equation (2.11) and (2.19) can be used to
calculate bhp and head versus discharge
2.4 Pump losses
“The types of losses can be divided mainly two primary types of losses namely; Hydraulic
losses, and Mechanical losses which can be divided as subgroup as shown in table 2.1[11]. ”
This table shows the types of losses which affect the delivered Head, capacity and power
consumption. All losses of head that occur between the suction and discharge pressure effect
from Hydraulic losses. Besides, there are losses of capacity in pump known as Leakage losses
which take place through gaps between the rotating and fixed parts of pump. The other losses
such as Mechanical losses, disk friction losses and other losses do not affect the head.
Page 23
14
The following section will express the detail and equation of H-Q curve. In general, the
hydraulic losses can be subdivided 2 types; the friction losses and the shock losses.
Stepanoff[17] presented the both losses curve as shown in Figure 2.4
Table 2.2: The different types of losses in pump and their affect on characteristic curve
Figure 2.13: Hydraulic loss curve [17]
“From figure 2.13, the sum of the friction and shock losses is the total of hydraulic losses and
determines the point of the maximum efficiency. ” When the flow rate is zero, there are only
the shock losses occurring in the pump. However, there are other losses in pumps as shown in
below figure [11]. In this section, the different types of losses and introduce some simple
models for calculating the magnitude of the losses will be described.
Page 24
15
Figure 2.14: Reduction of theoretical Euler head due to losses [11]
2.4.1 Shock losses (Incidence losses)
Shock losses occur when there is a difference between the flow angle and blade angle at the
impeller as shown in below figure.
Figure 2.15: Shock loss at inlet of impeller [11]
Stepanoff [17] presented a general formula of shock losses as a parabola with zero at the
best efficiency point.
2( )s s BEPH K Q Q (2.20)
Where sK is the coefficient of shock losses
Page 25
16
Figure 2.16: Shock losses as function of the flow curve [11]
2.4.2 Friction losses
“Friction losses occur in all components which the fluid flows through.” Stepanoff [17]
presented a general formula of friction losses as quadratic equation.
2
f fH K Q (2.21)
Where fK is the coefficient of friction losses
Figure 2.17: Friction losses as function of the flow curve [11]
Moreover, LeConte [13] has estimated losses in the pump. “First is an elbow loss at the
suction side, which can be estimated as 2
1 1 2k v g .” A loss at the channels of the impeller
from inlet to outlet (impeller friction losses) equal to 2
2 2 2k w g . “The loss where the high
velocity discharge 2v impacts into the lower velocity ov of the water in the scroll (sudden
expansion losses), which can be estimated as2 2 2
2 2 0 2 2( cos ) / 2 sin / 2v v g v g .” The
last is a loss in the scroll itself due to friction, which equal to 2
0 0 2k v g . Therefore, the
total friction losses can be calculated as
Page 26
17
2 2 2 2 2 2
1 1 2 2 0 0 2 2 0 2 2( cos ) sin
2 2 2 2 2f
k v k w k v v v vH
g g g g g
(2.22)
2.4.3 Recirculation losses
“Recirculation losses occur when the flow is below the design flow and can occur in inlet,
impeller, return channel or volute casing.” Figure 2.20 shows an example of recirculation in
the impeller. The formula of head of recirculation is [23];
3 2 2.5
1 (1 )re re
BEP
QH K D
Q (2.23)
Where reK is the coefficient of recirculation losses
The recirculation losses of head are high when the flow rate decrease and this loss will
disappear at the best efficiency point.
Figure 2.18: Recirculation losses in impeller [11]
2.4.4 Leakage losses
“Leakage losses occur because of smaller circulation through gaps between the rotating and
fixed parts of the pump.” Leakage loss results in a loss in efficiency because the flow in the
impeller is increased compared to the flow through the entire pump. The model of leakage
can be calculated by subtracting the head across the gap from the head generated by impeller
as [11]: 2 2
22
, ,
( )
8
gap
stat gap stat impeller fl
D DH H
g
(2.24)
Where
fl is rotational velocity of the fluid in cavity between impeller and casing of pump
gapD is inner diameter of the gap
Page 27
18
,stat impellerH is impeller static head rise
Ravi et al [24] present the way to analysis the losses of centrifugal pump that have origin
formula and data of centrifugal pump from research written in 2008 by Khin Cho Thin [23].
They use the MATLAB program to generate the loss curve for each type of losses. When
they received the all losses, the actual curve of head can be found by subtracting all losses
from theoretical head.
2.5 Pump as Impulse turbine
In case study of special study, Karan [19] studied about the feasibility study of using pump
working as turbine in impulse mode by empirical and numerical method. Also, design the
experimental setup for testing PAT in impulse mode. Due to lacking information involved
with parameters of centrifugal pump from vendors, the author has to consider parameters of
centrifugal pump for using as Impulse PAT by using the centrifugal sizing method from S.
Korpela [20]. After the most optimal parameters of pump were investigated, these will be
used to compute the performance of Impulse PAT.
The evaluation procedure and flow vector diagram of PAT in Impulse mode is applied. In this
study, the pump rotational speed at 1,500 and 3,000 rpm were been selected and turbine
rotational speed at 1,000, 1,500 and 3,000 rpm were evaluated. The result is the maximum of
theoretical efficiency from the calculation is 72.3% when Impulse PAT operates at 1,500 rpm
without including losses when water jet impact impeller, loss from hydraulic friction and
mechanical losses. In addition, the author presents selecting pump from pump manufacture
by using GRUNDFOS WebCAP program and compare the impeller radius between
calculation method and real pump obtained from pump vendor. The deviations from both are
4.89% and 3.55% for pump operation at 1,500 and 3,000 rpm, respectively.
2.6 Conclusion
From literature review, it can be found that the concept of using nozzle with PAT is new and
never explored by other researcher. In the theoretical study, Euler’s equation is used in order
to compute the performance of Impulse PAT. The H-Q curve of pump may be able to predict
by assuming in manner the head-capacity curve of idealized pump. Thus, being possible,
when the shape H-Q curve is known, the losses can be found by using developed H-Q curve
equation. Then, the losses can be used to calculation the geometry of centrifugal pump. The
table 2.3 summarizes the literatures that involve in PAT.
Page 28
19
Table 2.3: List of researches related to PAT
Researcher Pump type/RPM Pump data Type of method Detail
Sigh and
Nestmann
(2010)
Centrifugal pump
with specific
speeds 18.2, 19.7
and 44.7 rpm
-Head 32.5 m
-Discharge 29.2 lite/sec
-PAT efficiency 74%
Theoretical,
Numerical and
Experimental
- Use only the head, flow rate and rpm of
pump to find pump size based on the
experimental uncertainty of the model PATs
at BEP (the graph between error bands for
specific diameter, discharge number and
specific speed)
- Optimize the accuracy and reliability of
prediction and selection models
Patel et al
(2013)
Centifugal pump
-250 mm diameter of
impeller
Experimental - Do not explain how to select pump ,
considering the diameter pump that they have
- Determine the optimum position of fixed
guide vanes by varying the guide vane angles
between 45 to 80 degree
- 8 numbers of fixed guide vanes provided
angle at 75 degree
Derakhshan
and
Nourbakhsh
(2008)
4 Centifugal pumps
with specific speed
14.6, 23, 37.7 and
55.6 m3/sec
- Power 30 kW
- Head 25 m
- 0.15 m3/sec
- Efficiency for each pump;
65.5, 76, 86.5 and 87 %
Theoretical,
Numerical and
Experimental
- Do not explain how to select pumps
- Follow data pump from other researches
- Present 2 experimental equations to estimate
the complete characteristic curves of PAT
based on BEP
- Present a procedure to sect the selecting an
appropriate PAT in small hydro-site
Yang et al
(2012)
Centifugal pump
- 102 mm inlet impeller
diameter, 235 mm outlet
impeller diameter 39 and
28.22 degree for blade
angles at inlet and outlet,
outlet width impeller; 15.30
mm
Theoretical,
Numerical and
Experimental
- Measure geometric parameters from a real
pump in lab
- Develop a theoretical method by using
theoretical analysis and empirical correlation
- Predict the performance of PAT by using
CFD
Page 29
20
Table 2.3: List of researches related to PAT (continue)
Researcher Pump type/RPM Pump data Type of method Detail
Khin Cho Thin
et al(2008)
Centrifugal pump
2900 rpm
- Head 10 m
- Discharge 0.00293 m3/sec
- 95, 45, 10,8 and 16 for
outlet diameter of impeller,
eye diameter, hub diameter,
outlet width and hub length
Theoretical and
Numerical
-Do not explain how to get the geometries of
pump
-Design the pump by using the head, flow rate
and pump speed
-Some geometry of pump is known and used
to calculate rest of geometry
Karan (2014) Centifugal pump
with rotational
speed 1500 and
3000 rpm
- Head 100 m
- Discharge 0.2 m3/sec
-Maximum optput power
195.2 kW
Theoretical and
Numerical
- Design the pump by using the head, flow rate
and pump rotational speed
-Compare the pump’s outlet diameter of
impeller from calculation and real pump from
catalogue.
- Evaluate the performance Impulse PAT
Page 30
21
CHAPTER 3
METHODOLOGY
This chapter will describe the overview of the optimization of Centrifugal Pump used as
Impulse Turbine. The main concepts of methodology in thesis can be divided two part
namely; selection part and optimization part.
3.1 Research background
Pump as Turbine (PAT) selection part is a process of selecting a pump for a given site, with
head and flow rate. A great number of theoretical and experimental researches have been
finished for predicting performance of PAT as explained in chapter 2. However, these
experimental researches have detailed geometric and design data of pump in their lab. There
are no researches that present how to calculate the geometry of the selected pump to operate
as turbine. In addition, some researches follow the detail of pump geometry from other
researches.
Design centrifugal pump is the process of calculating geometry of a pump based on given
head and flow fate data from site. Several studies describe the method how to design a
centrifugal pump and how to calculate the pump design parameters. In design procedure,
they, however, need some input parameter apart from the given head and flow rate such as
speed of pump number of blade and so forth.
In addition, there are researches that present the way how to analysis losses in a centrifugal
pump and also explain the means to identify the losses of centrifugal pumps in order to
predetermine the head-capacity curve of their pumps. However, these studies still need some
input parameter apart from the given head and flow rate to calculate the losses.
Abovementioned, PAT selection part is not an easy task because if pump vendors are
reluctant to supply some needed information of pump aside from the given head and flow
rate, these methods cannot be apply to calculation. Even through some vendors give
characteristic curve, there are no researches study the way how to use the characteristic curve
to identity the geometry of pump. On the whole researches, they study in opposite way; in
other word, they use the geometry to determine pump characteristics.
Moreover, several researches present the way how to predict performance of centrifugal
pump working as turbine or in reverse mode. But, there are on literatures consider when the
centrifugal pump working as Impulse turbine.
3.2 The overview of research concept
This study is to select and optimize pump operating as Impulse turbine (PAT). Before
optimizing PAT, however, selecting pump should be considered because operating
centrifugal pump as impulse turbine has to be considered geometry of pump.
The data of site namely, head and discharge have origin from Ban Khun Pae Hydropower
Plant by Provincal Electricity Authority (PEA) is consist of 0.1166 – 0.2179 m3/s discharge
and 30 – 100 m head [13]. From these data, we select the design value for this study as 0.1
m3/s discharge and 100 m. Figure 3.1 shows the flow chart of concept for this study. First, the
centrifugal pump will be selected at the design point from catalogue after the design point of
Page 31
22
head and discharge data from site are received. As aforementioned in previous section, the
geometries of a centrifugal pump are needed in order to use to calculate optimization. In this
case study, some geometry of pump can be received from some Grundfos company [14] that
is a famous company and they have several pumps. Moreover, they also give typical
information of pump, dimensional drawing and characteristic curve for their pumps as shown
in figure3.3.The detail of technical term of a pump will be explained in appendix A1. These
pumps can be searched on their website as shown in figure 3.2.
Figure 3.1: Flow chart of research concept
Selection part
Optimization part
END
No
Design flow rate control
(spear valve and nozzle)
Best efficiency of
Impulse PAT ?
Yes
Select Pump from
manufacturer’s catalogues and
H-Q curve
Model the losses based on
H-Q curve
Algorithm to identify
geometry of pump
Optimize PAT in impulse mode
by varying the rotating speed
Page 32
23
However, only the typical geometries that shown in figure 3.2 cannot to use to be the input
parameter to optimization. Thereby, this study will present the method to find the remaining
unknown geometries by using the characteristic curve. As mentioned in previous section, the
losses can be found by calculating from the geometries. However, this study will do in
opposite way. The losses in a centrifugal pump can be estimated if the head-capacity (H-Q)
and efficiency curves are known. The detail how to estimate the losses will be described in
next chapter. After the losses are found, these losses will be used to calculate the geometry by
using algorithm that will be explained in next chapter.
After selection part, the next process is to optimize the pump operating as Impulse turbine.
The selected pump will be operated as Impulse turbine by first fixed the velocity of inject
water from nozzle. The speed of the inject water is fixed because of the stable design point of
head from site. The selected centrifugal pump will be optimized by varying the speed of
pump until find the best efficiency of PAT in Impulse mode. Moreover, this process will
include the losses from nozzle and the selected pump such as friction loss from nozzle and
selected pump and so on. The detail of optimization will be presented in chapter 4.
After optimization part, the next process is hydraulic turbine design that composes of spear
valve, nozzle and deflector. The details to calculate dimensional drawing of spear valve,
nozzle and deflector will be presented in chapter 6. After then, the spear valve, nozzle and
deflector will be built in 3D model in order to set up with a selected pump.
The algorithm of calculation the geometries of selected pump will be explained in chapter 4
and algorithm and the results of optimization pump working as Impulse turbine will be shown
in the chapter 5.
The program for calculation the geometries of the centrifugal pump , all concepts of optimize
pump using as Impulse turbine and calculate the efficiency of PAT in Impulse mode is
MATLAB. Solidworks program is performed to build the 2D and 3D model of hydraulic
turbine design.
Page 33
24
Figure 3.2: Pump selection on Grundfos website [14]
Page 34
25
Figure 3.3: The example of typical data and characteristic curve of selected pump from
Grundfos website [14]
Page 35
26
CHAPTER 4
PUMP SELECTION PART
This chapter will present the detail of method how to select a pump to perform as Impulse
turbine. As mentioned in chapter 3, Grundfos company; one famous pump company provides
the typical information and characteristic curves of a pump such as the actual impeller
diameter, rpm and so on. However, they do not provide the geometries such as the width of
outlet of impeller and the blade angle that is essential data to be the input parameter to
optimization. Thus, the objective in this chapter is to find the geometries of the pump that
will be used to optimization by using the given typical data and characteristic curves from
Grundfos company.
4.1 The overview of selection part
Figure 4.1 shows the flow chart of the sub-process in select part. This process consists of four
activities; select a pump, model the losses, identity geometries of pump and find the number
of water jet. The input of this process is the design point of head and discharge from site.
Figure 4.1: A flow chart of pump selection part
Input: Head and flow rate
Calculate the geometries and velocity at inlet of impeller
Get the total losses
Model the
losses
Select pump from catalogue based on head and flow rate
point
Get typical geometries and characteristic curve
Select pump
Calculate the unknown of geometries of selected pump Identity geometry
Is the number of jet less than
4?
Optimization part
No
Yes
Find the number of water jet
Split proportion for each type of total losses at the shut-off
point (Q=0)
Consider the new
operating design
point of head and
flow rate of pump as
same power
Page 36
27
4.1.1 Data acquisition
According to the data site from Ban Khun Hydropower Plant by Provincal Electricity
Authority (PEA), the design point of the head and flow rate are 100 m and 0.1 m3/s. The
detail of a pump will be illustrated in case study that will be presented after.
4.1.2 Model the total losses
After select a pump from catalogue and receive typical data of pump such as the head (H),
flow rate (Q), rpm (N), the efficiency (η), actual outlet of impeller diameter (D2), diameter of
outlet of scroll of pump (D0), the diameter of eye of pump (Deye) and the diameter of shaft
(Dshaft), next procedure is calculating pump’s geometries at the inlet of impeller and
geometries at the inlet of impeller that can be able to find. The algorithm to calculate these
geometries will show figure4.3. However, these data are not enough to use to calculation of
optimization. Thus, characteristic curve will be focus in order to find the rest of geometry of
selected pump.
After receiving pump data, geometries of pump at the inlet of impeller can be calculated as
the steps below.
1. Find the peripheral outlet speed of pump 2u as
22
60
D Nu
(4.1)
2. Calculate the absolute tangential velocity component 2tv from Euler’s equation
2
2
t
gHv
u (4.2)
3. From figure 2.2 the relative velocity at outlet of impeller in tangent component can be
calculated as
2 2 2t tw u v (4.3)
4. The area at exit of pump can be calculated 2
00
4
DA
(4.4)
5. The velocity at exit of pump
0
0
Qv
A (4.5)
6. Calculate the inlet area by first assuming as
2 2
14
e sA D D
(4.6)
7. The peripheral speed at inlet of impeller
160
eyeD rpmu
(4.7)
8. Calculate radius at inlet of impeller
Page 37
28
1*60
12* *
ur
pi rpm (4.8)
9. Due to angle of departure of the absolute velocity at inlet from radial velocity 1
1 90 , the absolute velocity equal circumferential speed ( 1 1rv v ).
1 1
1
r
Qv v
A (4.9)
10. From figure 2.2,the relative velocity at inlet of impeller 1w can be calculated as
2 2
1 1 1w v u (4.10)
11. The inlet impeller angle 1 is
11
1
arctanv
u
(4.11)
12. The width of inlet impeller 1b is
1
1 12
Qb
rv (4.12)
Figure 4.2: Flow chart of calculating pump’s geometries of inlet of impeller
4.1.3 Estimate the losses in selected pump
According to the type of losses shown section 2.4, the types of losses at the zero point of flow
rate (shut-off point) are recirculation loss, shock loss and leakage loss. Besides, when the
flow rate increase the recirculation loss will decrease and this loss is zero at the BEP. In
contrast, the friction loss will increase as quadratic from zero from shunt off point. Thus,
from figure 2.4, this study will divide the proportion of each type of losses as shown in table
4.1. Then, the each loss can be found by using the formula as shown in section2.4.
Table 4.1: Proportion of each loss in shut-off point
Description Abbreviation Value (%) Unit
The portion of Recirculation loss 70 %
The portion of Shock loss 22.5 %
The portion of Leakage loss 7.5 %
2 2
1 eye shaftA D D 2
00
4
DA
22
60
D rpmu
0
0
Qv
A 2
2
t
gHv
u
2 2 2t tw u v
1
1
Qv
A
2 2
1 1 1w u v 160
eyeD rpmu
Page 38
29
Then, the all types of losses in pump can be found as the step as;
1. Find the ideal head by using Euler’s equation
2
2 2 2
2 2
coteuler
u uH Q
g D b
(4.13)
If the flow rate is zero, the ideal head is 2
2, 0euler Q
uH
g (4.14)
2. Calculate the total head losses
1
2loss
HH
(4.15)
At shunt off point the total head losses is
@ 0 @ 0loss Q euler QH H H (4.16)
3. Recirculation loss, Leakage loss and shock loss at the shunt off point can be
calculated as
@ 0 @ 0
@ 0
@ 0 @ 0 @ 0
re Q loss Q
l loss Q
s Q loss Q re Q l
H H
H H
H H H H
(4.17)
Where and are the proportions of recirculation and leakage losses in shut-off
point
4. As mentioned formula of shock loss in section 2.4.1, coefficient of Shock loss can be
found at the shunt off point (Q=0)
@ 0
2
s Q
s
BEP
Hk
Q
(4.18)
5. Then the shock loss in different flow rate can be calculated as
2
s s BEPH k Q Q (4.19)
6. Also coefficient of recirculation loss can be found from recirculation formula in
section 2.4.3 at the shunt off point (Q=0) as
@ 0
3 2
1
re Q
re
Hk
D
(4.20)
7. The recirculation loss can be found in 2.5
3 2
1 1re re
BEP
QH k D
Q
(4.20)
8. The friction loss can be found by subtracting the recirculation loss, shock loss and
leakage loss from the total loss(4.15)
f loss re l sH H H H H (4.21)
Page 39
30
Figure 4.3: Flow chart of calculating total losses in selected pump
4.1.4 Algorithm to identity geometry of selected pump
After the each type of losses are found, the rest of geometry of selected pump can be
calculated as the step below.
1. According to Leconte’s friction losses formula [13], the friction losses can be found
as
2 2 2 2 2 2
1 1 2 2 0 0 2 2 0 2 2( cos ) sin
2 2 2 2 2f
k v k w k v v v vH
g g g g g
(4.22)
In this study, the coefficients namely, 1 2 0,k k and k that are represented by fk ,
however, will be the same value and that will be expressed later. Besides, this study
develops the equation of impeller friction losses in order get more accurate by using
the average of the relative velocity at the inlet and outlet as 2 2
1 2
2 2
fk w w
g
. Therefore,
the total friction losses can be calculated as
2 22 2 2 2 21 01 2 2 2 0 2 2( ) ( cos ) sin
2 2 2 2 2 2
f f f
f
k v k k vw w v v vH
g g g g g
(4.23)
Expand above equation 2 22 2 2 21 01 2 2 0 0 2 0 2 2( ) 2 2 cos
2 2 2 2 2 2 2 2
f f f
f
k v k k vw w v v v u v wH
g g g g g g g
(4.24)
But 2 2 2
2 2 2 2 2 22 cosv u w w u
Page 40
31
Then, the above equation can be reduced to
2 2 2 2 2 2 2
1 1 2 0 2 2 2 2 2 0 0 2 0 2 2( ) 2 cos 2 2 cos2
2
f
f f
fric
kk v w w k v u w u w v v u v w
Hg
(4.25)
From equation 4.25, there are 2 unknown parameter namely, coefficient of friction
loss and the relative velocity at outlet of impeller. However, the friction losses in the
different flow rate are known, so the coefficient of friction loss and the relative
velocity at outlet of impeller can be solved. After formatting equation 4.25 in order to
solve the relative velocity at outlet, the equation 4.25 will be change as quadratic
equation. In this study, MATLAB will be used in order to solve the relative velocity
at outlet. After the relative velocity at outlet of impeller is known, the remaining
unknown geometries of selected pump can be calculated as the algorithm shown in
figure 4.4.
Figure 4.4: Flow chart of calculating remaining pump’s geometries
4.1.5 Find the number of water jet
After all geometries of a selected pump are known, next procedure is finding the number of
water jet by comparison between the width at outlet of impeller and the diameter of the water
jet. The objective of this procedure is to avoid some of waste energy will occur when the
width at outlet of impeller is smaller than the diameter of water jet. Firstly, the diameter of
the water jet can be calculated if the head and the flow rate at design point from site are
known. The velocity of water jet can calculated as
2jv gH (4.26)
Then, the diameter of the water jet can be found as
4j
j
QD
v (4.27)
Page 41
32
After then, the width at outlet of impeller will be compared with diameter of water jet. If the
width at outlet of impeller is smaller than the diameter of water jet, the number of water jet
will be more than one. The number of water jet can be calculated as
2
2
( )jD
numberof jetb
(4.28)
However, if the number of water jet increases a lot, for example, 6 or more than that, the loss
and the cost to set up will also increase. To avoid this occurrence, a new pump from
catalogue will be selected again and this process will be iterated until the number of water jet
is appropriated; in other word this study defines the number of water jet that should be less
than 4. For the iteration, the new pump at the other design point but the same power will be
selected. For example, the new pump that can operate at the half of head and double of flow
rate design point or double of head and half of flow rate will be considered. Moreover, the
case study will show the iteration process. After obtained the suit number of water jet, next
process is optimization selected pump as Impulse turbine.
4.2 Case study
After receiving the head and flow rate at the design point, a pump from catalogue will be
selected. The given data, dimensional drawing and characteristic curve of selected pump are
shown in figure 4.5.
From the head- flow rate curve, the selected pump can operate at the design points (H=100m
and the flow rate=0.1 m3/s). However, the data, n is 103% means that the rpm perform in
103% with the normal 2980 rpm. That means this selected pump has to perform at 3069.4
rpm to operate at the design points. The lists in the table 4.1 are all the input parameters that
are got from Grundfos catalogue and used to calculation the remaining unknown geometries
of its pump. An actual head at shut-off point (Q=0) can be estimated on the given H-Q curve
from Grundfos as shown figure4.1. The head at the zero of flow rate is 108.61m.
Table 4.2: Input parameter of selected pump from Grundfos for using to calculate the
remaining geometries
Parameter Abbreviation value unit
Flow rate @BEP Q 0.1 3m s
Head@ BEP H 100 m
Efficiency @BEP .806
Rotating speed of pump rpm 2980 rpm
Actual impeller diameter 2D 0.27 m
Pump flange inlet diameter eyeD 0.125 m
Pump flange outlet diameter 0D 0.1 m
Shaft diameter shaftD 0.032 m
Gravity g 9.81 2m s
Actual Head at Q=0 , 0act QH 108.61 m
Then, these input parameters in table 4.2 will be used to calculation by use the algorithm as
mentioned from section 4.1.2 to 4.1.5. The table 4.3 shows the results of geometry of selected
pump from catalogue after calculation.
Page 42
33
Figure 4.5: Given data of selected pump from Grundfos website [14]
From the table 4.3, the geometry at the inlet of impeller and at the exit pump and some
velocity at the outlet of impeller can be solved directly by using given data from catalogue of
Grunfos company. In contrast, the geometry at the outlet of impeller will be found by using
characteristic curve to estimate the losses in pump. After friction loss is found, the relative
velocity at outlet of impeller and friction coefficient will be solved by using the different
points on head-discharge curve. Because there are two unknown parameters namely, the
relative velocity at outlet of impeller and friction coefficient, they should have two equations.
Hence, the friction loss in different point of head-discharge point will be considered. After
getting the relative velocity at outlet of impeller and friction coefficient, the rest of geometry
of the pump at outlet of impeller can be calculated.
Page 43
34
The diameter of the water jet can be calculated from velocity of water jet as shown as section
4.1.5. The results from calculation will be shown in table 4.4. The velocity of water jet will
be calculated from the head from site (100 m). Then, the width of the blade at outlet will be
compared with the diameter of the jet. As the result, the width of the blade at outlet is smaller
than the diameter of the jet. The number of water jet should be more than one. After
calculating, the number of jet is 8.1722 that mean 9 jets. This number is very high and leads
to more loss in the jet and also raises the cost to set up the jet. Consequently, the new pump at
the other design point will be select namely, decrease the head 50 m and increase the flow
rate 0.2 m3/s. The reason why the head 50 m and increase the flow rate 0.2 m
3/s will be
choose because the number of jet will be decreased and the power of the pump will be same.
But this pump still is used for the head and discharge as the same site.
Table 4.3: Results from pump selection procedure
Description Abbreviation Unit Value
The peripheral speed at outlet of impeller 2u m/s 30.2729
The absolute velocity in tangential component at
outlet of impeller 2tv m/s 21.4320
The relative velocity in tangential component at outlet
of impeller 2tw m/s 12.8764
The area at exit of pump 0A m
2 0.0491
The velocity at exit of pump 0v m/s 4.0744
The inlet area 1A m
2 0.0452
The peripheral speed at inlet 1u m/s 21.3691
The radius at inlet of impeller 1r m 0.0605
The absolute velocity at inlet of impeller 1v m/s 3.0701
The relative velocity at inlet of impeller 1w m/s 21.5885
The inlet impeller angle 1 degree 11.6888
The width of inlet impeller 1b m 0.0480
The relative velocity at outlet of impeller 2w m/s 11.2890
Friction coefficient k 0.3422
The outlet impeller angle 2 degree 22.7485
The absolute velocity at outlet of impeller 2v m/s 33.2694
The absolute velocity in radial component at outlet of
impeller 2rv m/s 4.3653
The angle of absolute velocity at outlet 2 degree 7.5395
The outlet area 2A m
2 0.0159
The width of outlet impeller 2b m 0.0188
Page 44
35
Table 4.4: Results from calculating diameter and the number of water jet
Table 4.5: Input parameter of a new selected pump from Grundfos for using to
calculate the remaining unknown geometries
Then, a new alternative pump is also selected from catalogue of Grundfos. The given data
and characteristic curve of a new selected pump are shown in figure 4.6.
From the head- flow rate curve in figure 4.7, a new selected pump can operate at the design
points (H=50 m and the flow rate=0.2 m3/s) when the pump operate at 1360.4 rpm because
76% of the full rpm, 1790, is 1360.4. The lists in the table 4.5 are all the input parameters that
are got from Grundfos catalogue and used to calculation the remaining unknown geometries
of its pump. The head at the zero of flow rate is 48.7 m.
The results of geometry of a new selected pump from the same calculation with first selected
pump are shown in table4.6. Also, the diameter of the water jet can be calculated from
velocity of water jet as the same frist selected pump.The results from calculation will be
shown in table4.7. The velocity of water jet will be calculated from the head from site (100
m). Then, the width of the blade at outlet will be compared with the diameter of the jet. As
the result, the width of the blade at outlet is smaller than the diameter of the jet. The number
of water jet should be more than one. After calculating, the number of jet is 1.3590 that
means is 2 jets. Thus, this new selected pump will be used to optimization. Next process is
optimization and the detail of this part will be presented in next chapter.
Description Abbreviation Unit Value
The velocity of water jet jv m/s 44.2945
The diameter of water jet j
D m 0.0536
The number of water jet jnum
9
Parameter Abbreviation value unit
Design point of flow rate from site Q 0.1 3m s
Design point of head from site H 100 m
Flow rate @BEP BEPQ 0.2264 3m s
Head@ BEP BEPH 46.9 m
Efficiency @BEP 0.868
Rotating speed of pump N 1360.4 rpm
Actual impeller diameter 2D 0.425 m
Pump flange inlet diameter eyeD 0.3 m
Pump flange outlet diameter 0D 0.25 m
Shaft diameter sD 0.06 m
Actual Head at Q=0 , 0act QH 48.7 m
Page 45
36
Figure 4.6: Given data of a new selected pump from Grundfos [14]
Page 46
37
Table 4.6: Results from a new pump selection procedure
Description Abbreviation Unit Value
The peripheral speed at outlet of impeller 2u m/s 43.3926
The absolute velocity in tangential component at
outlet of impeller 2tv m/s 28.0490
The relative velocity in tangential component at outlet
of impeller 2tw m/s 15.3436
The area at exit of pump 0A m
2 0.0079
The velocity at exit of pump 0v m/s 12.7324
The inlet area 1A m
2 0.0068
The peripheral speed at inlet 1u m/s 20.0892
The radius at inlet of impeller 1r m 0.0625
The absolute velocity at inlet of impeller 1v m/s 14.7212
The relative velocity at inlet of impeller 1w m/s 24.9056
The inlet impeller angle 1 degree 36.2337
The width of inlet impeller 1b m 0.0173
The relative velocity at outlet of impeller 2w m/s 9.1256
Friction coefficient k 0.331
The outlet impeller angle 2 degree 14.3507
The absolute velocity at outlet of impeller 2v m/s 21.5510
The absolute velocity in radial component at outlet of
impeller 2rv m/s 2.2618
The angle of absolute velocity at outlet 2 degree 6.0245
The outlet area 2A m
2 0.0614
The width of outlet impeller 2b m 0.046
Table 4.7: Results from calculating new diameter and the number of water jet
Description Abbreviation Unit Value
The velocity of water jet jv m/s 44.2945
The diameter of jet j
D m 0.0536
The number of water jet jnum
2
Page 47
38
CHAPTER 5
OPTIMIZATION OF IMPULSE PAT IN CASE STUDY
After the geometries of selected pump are found, next procedure is optimization of PAT in
Impulse mode. The goal of this part is to find the performance of pump operated as Impulse
turbine namely, the rpm of the selected pump that can give the maximum efficiency. This
process will be divided in two main cases; not including the losses that happen in the pump
and the control flow rate device such as the nozzle and spear valve and including the losses.
5.1 The overview of optimization of Impulse PAT
According to the chapter 4, the velocity of the water jet will be calculated from the design
point of head from site. This velocity is constant because the value of head is constant as
well. Then the peripheral speed at the inlet of impeller of Impulse PAT will be varied in order
to optimize the efficiency of PAT in Impulse mode. In this case study, the Impulse PAT will
be operated from 0 to 2500 rpm. Then a selected pump will be performed evaluation of
Impulse PAT. This process can be divided in 2 options. First option is reverse rotating
direction in order to working as turbine. Normally, pump run as backward mode as shown
figure5.1 so Impulse PAT operated as first option will run in forward mode as shown
figure5.2. First option can be considered when the velocity of the water jet smaller than the
peripheral of Impulse turbine at the inlet. Secondly, a pump is not operated reverse running
direction namely; the pump still run in backward mode but the direction to inject the water
changes. The figures5.3 shows the diagram of vector at inlet and outlet impeller of second
option. Second option can be considered when the velocity of the water jet bigger than the
peripheral of Impulse turbine at the inlet.
Figure 5.1: Velocity diagram of a pump with backward blade
Page 48
39
Figure 5.2: Impulse PAT with reverse operation (first option)
Figure 5.3: Impulse PAT with backward mode(second option)
Page 49
40
5.2 The algorithm to optimize Impulse PAT
The subscribe 1 and 2 will represent inlet and outlet of impeller when the selected pump work
in normal mode (pump mode). The subscribe 3 and 4 will represent inlet and outlet of
impeller when the selected pump work as Impulse turbine. From here, the known parameters
that consist of 2 1 2 1 2 1, , , , , , ,b b d d rpm Qand H will be used to calculation.
We assume that 2 3 1 4 2 3 1 4 2 3 1 4 3, , , , , jb b b b d d d d and v v
1. Normally, the velocity of water jet can be calculated as
2jv gH (5.1)
2. Rotating speed at inlet impeller is
33
60
d rpmu
(5.2)
3. The relative velocity ( 3w ) at inlet impeller in first option and second option can be
calculated solving quadratic equation from cosine law as equation3.30 and 3.31.
2 2 2
3 3 3 3 3 32 cosv w u w u (5.3)
2 2 2
3 3 3 3 3 32 cos(180 )v w u w u (5.4)
4. Normal velocity at inlet will be calculated as
3 3 3sin( )rv w (5.5)
5. The angle between velocity and circumferential speed at inlet impeller 3 can be
found from cosine law
1 33
3
sin rv
v
(5.6)
6. Tangent velocity at inlet of impeller will be calculated as
3 3 3cos( )tv v (5.7)
7. Rotating speed at outlet impeller is
44
60
d rpmu
(5.8)
8. The relative velocity at outlet impeller can be calculated by using the formula [13] as
2 2 2 2
4 3 3 4( )
2 2 2
w w u u
g g g
(5.9)
9. The absolute velocity at outlet impeller can be calculated in first option and second
option can be calculated solving quadratic equation from cosine law as
2 2 2
4 4 4 4 4 42 cos( )v w u w u (5.10)
2 2 2
4 4 4 4 4 42 cos(180 )v w u w u (5.11)
Page 50
41
10. The angle between velocity and circumferential speed at outlet impeller 4 can be
found from cosine law
1 44
4
sin rv
v
(5.12)
11. Tangent velocity at inlet will be calculated as
4 4 4cos( )tv v (5.13)
12. Efficiency can be calculated as
2 2
3 4
2
3
100%v v
v
(5.14)
After optimizing Impulse PAT included losses, next process is design the control flow rate
devices such as spear valve and nozzle. The details of regulating devices will be presented in
chapter 6. Then spear valve and nozzle will be included to optimize PAT in Impulse mode by
include the friction loss of spear valve and nozzle.
Next step, the all type of losses in Impulse PAT and control flow rate devices will be
included in optimization. The algorithm to optimize Impulse PAT with the losses is same as
that without losses. But velocity of water jet and efficiency will be changed as
2j vv C gH (5.15)
2 2 2
3 4 4
2
3
100%fv v k w
v
(5.16)
Where the efficiency of the nozzle ( vc ) is 0.95 because of friction losses in nozzle. However,
the efficiency of the nozzle is 1 in first case study owing to no including the all type of losses
when pump operating as Impulse turbine. The coefficient of friction losses of Impulse PAT
( fk ) is similar to that of pump in normal mode.
Page 51
42
Figure 5.4: Flow chart of optimization process
2 2 2 2
4 3 3 4
2 2 2
w w u u
g g g
1Option 2Option
2 2 2
3 3 3 3 3 3
3 3 3
( 2 cos(180 )) ( 2 cos(180 )) 4( )
2
u u u vw when v u
2 2 2
3 3 3 3 3 3
3 3 3
( 2 cos ) ( 2 cos ) 4( )
2
u u u vw when v u
3 2vv C gH3
360
D rpmu
4
460
D rpmu
3 3 3sinrv w 1 33
3
sin rv
v
2 2
4 4 4 4 4 4 3 32 cosv u w u w when v u 2 2
4 4 4 4 4 4 3 32 cos(180 )v u w u w when v u
3 3 3costv v 4 4 4sinrv w
2
4
2
wFriction loss k
g
4 4 4costv v 2 2 2
3 4 4
2
3
v v kw
v
1Option 2Option
Page 52
43
5.3 The results of Impulse PAT optimization
This case study is divided two main cases; not including the losses and including the losses.
The section 5.3.1 and 5.3.2 will show the results of optimization by not including the losses
in spite of section 5.3.3 and 5.3.4.
5.3.1 The results of first option without friction losses
For the first option, the rpm of Impulse PAT will be varied from 2000 to 2500. The absolute
velocity, relative velocity and the peripheral velocity at the inlet of Impulse PAT (subscript 3)
can be calculated as the steps in previous chapter. However, the all types of velocity at the
outlet of Impulse PAT cannot find because the value of the relative velocity is complex
number that mean it cannot give realistic results. Thereby, this option is not satisfied. The
velocity vector diagram for first option will be shown in appendix B1.
5.3.2 The results of second option without friction losses
In the second option, the rpm of Impulse PAT will be varied from 500 to 1790. The geometry
of Impulse PAT can be calculated by following the steps in previous section. The absolute
velocity, relative velocity and the peripheral velocity at the inlet of Impulse PAT (subscript 3)
can be found. The mentioned velocity at the outlet of Impulse PAT can be found in low rpm
from 0 to 1790 rpm. After optimize the Impulse PAT in different rpm, the Impulse PAT can
perform when range of rpm is from 0 t0 1174 as shown in figure 5.5. If Impulse PAT is
operated at the higher rpm than 1174, the results from calculating is not realistic because they
are complex number. Figure 5.5 shows the graph of peripheral velocity at inlet, water jet, the
angle of absolute velocity at inlet of Impulse PAT and efficiency when Impulse PAT operates
in different point of rpm. From figure 5.5, the trend of efficiency increases when rpm is rose.
When the Impulse PAT operates at rpm 1174, the maximum efficiency is 81.972%. The
peripheral velocity is increased as linear from 0 to 26.125 m/s due to increasing rpm. The
velocity of the jet is thoroughly constant (44.2945 m/s) all of rpm. The angle of absolute
velocity at inlet (alfha3) is decreased from 14.351 to 5.945 degree. The efficiency is rose like
a quadratic from 0 to 81.972%. From figure 5.5 at the maximum efficiency point and rpm
1174, the peripheral velocity is nearly half of the velocity of jet. However, this calculation
does not include the losses from control flow rate device and pump. Thus, the next section
will be present the results of optimization- two options- including the friction losses from the
spear valve and nozzle and also Impulse PAT. The velocity vector diagram for second option
will be shown in appendix B2.
Page 53
44
Table 5.1: Impulse PAT performance evaluation in second option without friction losses
Description Abbreviation Unit Value
Pump Rotational Speed Npump RPM 1360.4
Turbine Rotational Speed Nturbine RPM 500 750 1000 1174
Turbine Blade Tip Tangential Speed at outlet U3 m/s 11.1265 16.6897 22.2529 26.125
Turbine Tip Absolute Velocity (Jet velocity) V3 m/s 44.2945
Turbine Reverse Flow Angle of Absolute Velocity at
outlet
α3 degree 10.781 8.992 7.198 5.9447
Turbine Tip Relative Velocity at outlet W3 m/s 33.429 27.932 22.391 18.5088
Turbine Absolute Velocity in Radial Direction at outlet V3r m/s 8.286 6.923 5.55 4.5875
Turbine Absolute Velocity in Tangential Direction at
outlet
V3t m/s 43.5126 43.7501 43.9454 44.0563
Turbine Tip Absolute Velocity at inlet V4 m/s 40.2095 36.9208 31.4468 18.8071
Turbine Absolute Velocity in Radial Direction at inlet V4r m/s 6.5817 5.1271 3.2219 0.0757
Turbine Absolute Velocity in Tangential Direction at
outlet
V4t m/s 39.6672 36.5631 31.2813 18.807
Turbine Reverse Flow Angle of Absolute Velocity at inlet α4 degree 9.4209 7.9823 5.8806 0.2306
Turbine Tip Relative Velocity at inlet W4 m/s 32.4869 25.3069 15.9031 0.3736
Turbine Blade Tip Tangential Speed at inlet U4 m/s 7.854 11.781 15.708 18.4411
Efficiency η % 17.5939 30.5227 49.5973 81.9721
Page 54
45
Figure 5.5: Graph between some geometry at inlet and rpm of Impulse PAT without friction losses
Page 55
46
5.3.3 The results of first option with friction losses
The condition in process is same as calculating without the friction losses namely; the rpm of
Impulse PAT will be varied from 2000 to 2500. The velocity of jet is dropped from 44.2945
to 42.9656 m/s because friction losses of spear valve and nozzle. Also the results of first
option with friction losses are similar to results of first option without friction losses. The
absolute velocity, relative velocity and the peripheral velocity at the inlet of Impulse PAT
(subscript 3) can be calculated as the steps in previous section. However, the all types of
velocity at the outlet of Impulse PAT cannot find because the value of the relative velocity is
complex number as the same as first option that means it cannot give realistic results.
Thereby, this option is not satisfied.
5.3.4 The results of second option with friction losses
The rpm of Impulse PAT will be varied from 0 to 1790. The geometries of Impulse PAT can
be calculated by following the steps in previous section. The absolute velocity, relative
velocity and the peripheral velocity at the inlet of Impulse PAT (subscript 3) can be found.
As the result, the mentioned velocity at the outlet of Impulse PAT can be found in low rpm
from 487.8 to 1138.87 rpm as shown in table 5.2. Figure 5.6 illustrates the trend of peripheral
velocity at inlet, water jet, the angle of absolute velocity at inlet of Impulse PAT and
efficiency when Impulse PAT operates in different point of rpm. From figure 5.6, the trend of
efficiency increases when rpm is rose. The range of rpm that pump can operate as Impulse
turbine is from 487.8 to 1138.87. When the Impulse PAT operates at rpm 1138.87 and the
maximum efficiency is 82.493%. In contrast, the mentioned velocity at the outlet of Impulse
PAT cannot be found in high rpm of Impulse PAT because of complex number that means
rpm that is out of the range of rpm from 487.8 to 1138.87 is not realistic. The peripheral
velocity is rose as linear from 10.9362 to 25.343 m/s due to increasing rpm. The velocity of
the jet is thoroughly constant (42.9565 m/s) all of rpm. The angle of absolute velocity at inlet
(alfha3) is decreased from 10.761 to 5.944 degree. The efficiency is rose like a quadratic
from 0 to 82.493%. From figure 5.4 at the maximum efficiency point and rpm 1138.87, the
peripheral velocity is nearly half of the velocity of jet. Figure 5.7 illustrate the friction losses
in Impulse PAT. The friction losses are drop from 16.5723 to nearly 0 m while the rpm of
Impulse PAT is increased due to the dropping relative velocity at the outlet of Impulse.
5.4 Conclusion
Only the second option namely; when the velocity of jet is bigger than the inlet peripheral
velocity can be applied for optimize pump as Impulse turbine. In a case without friction
losses, the Impulse PAT can operate from 0 to 1174 rpm. The maximum of efficiency is
81.972% at the 1174 rpm. In a case with friction losses, the Impulse PAT can operate from
487.8 to 1138.87 rpm. The maximum of efficiency is 82.493% at the 1138.87 rpm. For two
cases, the peripheral velocity is almost half of the velocity of the jet at the maximum
efficiency point. That is similar to the theoretical performance of the Pelton turbine, at the
maximum efficiency, the peripheral velocity is half of jet velocity. However, at these
theoretical maximum efficiency points, the outlet relative velocity is zero that means the flow
rate cannot go through Impulse PAT. However, if Impulse PAT operates at the lower rpm,
the flow rate can go through in Impulse PAT. For without losses, the maximum rpm is 1005
rpm and the maximum efficiency point is 49.838%. For without friction losses, the maximum
rpm is 960 rpm and the maximum efficiency point is 43.784%.
Page 56
47
Table 5.2: Impulse PAT performance evaluation in second option with friction losses
Description Abbreviation Unit Value
Pump Rotational Speed Npump RPM 1360.4
Turbine Rotational Speed Nturbine RPM 500 750 1000 1138.87
Turbine Blade Tip Tangential Speed at outlet U3 m/s 11.1265 16.6897 22.2529 25.3432
Turbine Tip Absolute Velocity (Jet velocity) V3 m/s 42.9656
Turbine Reverse Flow Angle of Absolute Velocity at outlet α3 degree 10.671 8.826 6.975 5.9441
Turbine Tip Relative Velocity at outlet W3 m/s 32.098 26.597 21.052 17.9516
Turbine Absolute Velocity in Radial Direction at outlet V3r m/s 7.956 6.592 5.218 4.4494
Turbine Absolute Velocity in Tangential Direction at outlet V3t m/s 42.223 42.457 42.648 42.7346
Turbine Tip Absolute Velocity at inlet V4 m/s 38.8389 35.4425 29.5083 17.9782
Turbine Absolute Velocity in Radial Direction at inlet V4r m/s 6.3038 4.8269 2.827 0.0128
Turbine Absolute Velocity in Tangential Direction at outlet V4t m/s 38.3239 35.1123 29.3726 17.9782
Turbine Reverse Flow Angle of Absolute Velocity at inlet α4 degree 9.3408 7.8275 5.4976 0.0407
Turbine Tip Relative Velocity at inlet W4 m/s 31.1151 23.8254 13.954 0.0898
Turbine Blade Tip Tangential Speed at inlet U4 m/s 7.854 11.781 15.708 17.8893
Friction loss at impeller
m 16.5723 9.5766 3.2849 0.0001
Efficiency η % 0.928 21.775 49.341 82.4913
Page 57
48
Figure 5.6: Graph between some geometry at inlet and rpm of Impulse PAT with friction losses
Page 58
49
Figure 5.7: Graph between friction loss in Impulse PAT and rpm
Page 59
50
CHAPTER 6
HYDRAULIC TURBINE DESIGN
This chapter will present the details to design the contracture of distributor consisting of
spear valve, nozzle and deflector.
6.1 Spear valve and nozzle design
A spear valve (or needle) is a regulator flow rate device. In figure 6.1, the flow rate is
adjusted by closing and opening the nozzle mouth by moved axially the spear valve
within the nozzle. The position of spear valve is controlled via a servomechanism. In
addition, the appropriate position of deflector as shown in figure 6.4 can adjust the flow
rate as well.
Figure 6.1 : The means of controlling the speed of Pelton turbine with a spear
valve [28]
The dimensions of spear valve and nozzle in this study will follow the design from
Provincial Electricity Authority report [27] as shown in figure 6.2 and 6.3.
From the figure 6.3, the nozzle diameter can calculated as
n n jD C D (6.1)
The largest spear diameter is
sl s jD C D (6.2)
The spear rod diameter can be found as
rod rod jD C D (6.3)
The spear tip length
s sl jL C D (6.4)
The pipe diameter (Nozzle)
pipe pipe jD C D (6.5)
The Nozzle holder radius
1.2n pipeR D (6.6)
In this study, the spear tip angle is 55 degree and the nozzle holder angle is 80 degree
[27]. According to previous chapter, the diameter of one jet is 53.8 mm. In this case
study, the number of jet, however, is 2 so the area of one jet will be divided into two jets
with the same area. The diameter of each jet ( jD ) is 37.9 mm
The coefficient of all parameters is corresponding to the efficiency of 94 to 96 %. The
dimensions of spear valve and nozzle from calculation are shown in table 6.1.
Page 60
51
Figure 6.2: Description of spear valve and nozzle [27]
Figure 6.3: Dimensions for the spear valve and nozzle [27]
Where
jD is jet diameter and 37.9jD mm
nC is nozzle diameter coefficient and 1.25nC
sC is largest spear diameter coefficient and 1.65sC
Page 61
52
rodC is spear rod diameter coefficient and 0.6rodC
slC is spear tip length coefficient and 3.5slC
pipeC is pipe coefficient and 3.2pipeC
Table 6.1: The results of spear design
Description Abbreviation Unit Value
The nozzle diameter nD mm 47.4
The largest spear diameter slD mm 62.6
The spear rod diameter rodD mm 22.7
The spear tip length sL mm 132.7
The pipe diameter (Nozzle) pipeD mm 121.3
The Nozzle holder radius nR mm 145.6
Normally, the spear valve move slowly to open and close the hold of nozzle. If the spear
valve closes quickly, the water hammer will occur. To avoid this problem, the deflector
will be used. The deflector acts to protect overspeeding and allows time for the slower
acting spear valve to move to a new position. Figure 6.4 shows the deflector operation.
Figure 6.4: Deflector operation [25]
There are no researches presenting the way how to calculate the deflector dimension.
Hence, the estimated dimension of deflector in this study is followed from Nechleba
[25]. The dimensional drawing of spear valve, nozzle and deflector will be illustrated in
next section.
6.2 The 2D and 3D model of regulating flow rate device
After calculating 1D dimension of spear valve and nozzle, the 1D result from
calculating of spear valve and nozzle will be built in 2D and 3D model. Solidworks
program is used to sketch and build 2D and 3D models. Figure 6.5 and 6.6 show the 2D
dimension of spear valve, nozzle and deflector. Figure 6.7 to 6.9 show the 3D model of
spear valve, nozzle and deflector, respectively. Moreover, the figure 6.10 will show the
assembly of the hydraulic turbine device to control the flow.
Page 62
53
Figure 6.5: Dimensional drawing for the spear valve and nozzle
Figure 6.6: Dimensional drawing for the deflector
Page 63
54
Figure 6.7: 3D model of spear valve
Figure 6.8: 3D model of nozzle
Page 64
55
Figure 6.9: 3D model of deflector
Figure 6.10: Assembly of spear valve, nozzle and deflector
Page 65
56
6.3 Performance of spear valve and nozzle
As mentioned in previous section, the flow rate through the nozzle solely depends on
the position of the spear valve. Figure 6.11 shows the position of the spear valve
(needle). In figure 6.11, S represents the stoke needle. If the spear valve is nearly to
close the mouth of nozzle (S≈0), the diameter of the jet will decrease that mean the flow
rate will increase. In contrast, if the spear valve move far from the mouth of nozzle, the
diameter of the jet will increase that mean the flow rate will decrease.
Figure 6.11: The position of spear valve [28]
Figure 6.12: Graph between the flow rate and stoke of needle
0
0.02
0.04
0.06
0.08
0.1
0.12
0.00
1.11
2.29
3.47
4.65
5.83
7.01
8.19
9.37
10.56
11
.74
12.92
14
.10
15.28
16
.46
17.64
18
.82
20.00
21
.18
22.36
23
.54
24.72
25
.90
27.08
28
.26
29.44
30
.62
31.80
32
.98
34.16
35
.34
36.42
Flo
w r
ate
m^3
/se
c
stroke of needle in mm
Flow rate vs stroke of needle
Page 66
57
The typical functional dependency of the discharge and stoke of needle (spear valve) is
depicted in figure 6.12. The curve of the flow rate and stoke of needle is quadratic.
Moreover, the maximum stoke needle is 36.415 mm that means the flow rate is
maximum around 0.105 m3/sec. The flow rate can be calculated from the constant jet
velocity and the area at exist of jet as shown in equation 6.7.
j
j
vQ
A (6.7)
The area between the needle and nozzle at exist of jet is Truncated cone and the way to
calculate this area will be presented in appendix C1.
6.4 Performance of deflector
From figure 6.4, the deflector has to always be closed to the water jet. For example,
when it is required to reduce the output water immediately from 100% to 90%, the
deflector cuts off the essential portion of the jet and controls the flow rate. Figure 6.13
illustrates angle’s position of deflector. The angle of deflector is represented by αd.
Figure 6.14 shows the different position of the angle’s deflector. For instance, the angle
of deflector is 22.2 degree, the deflector absolutely close the diameter of the jet as
shown in figure 6.15. Figure 6.16 and 6.17 show the performance when deflector has
half position and completely open as angle is 43 and 57.54 degree, respectively.
Figure 6.13: The angle of deflector
Page 67
58
Figure 6.14: The angle of deflector in different position
Figure 6.15: Fully deflected position
Page 68
59
Figure 6.16: Half deflected position
Figure 6.17: Completely open deflector
Page 69
60
Figure 6.18: Graph between the flow rate from jet and angle of deflector
Figure 6.18 shows the relative between the flow rate exiting from water jet and the
angle of deflector in different positions. The flow rate is increased from 22.20 to 57.54
degree. The flow rate is zero when angle of deflector performs from 5 to 22.5 degree.
Then, the flow rate is reduced to be 50% when the angle is 43 degree. In addition, the
deflector will completely open when angle is 57.54 degree that means the flow rate is
maximum. The flow rate can be calculated as equation 6.7. The Area of the exiting
water can be calculated as shown in appendix C2.
6.5 Position the water jet with a selected pump
After designing distributors such as spear valve, nozzle and deflector, the scroll will be
built in 3D model. Figure 6.19, the inside diameter is 426 mm that is bigger than
impeller of pump 1 mm. The diameter of outlet and inlet of pump are 125 mm and 46
mm, respectively. Then, these distributors and scroll will be position by first set up into
the scroll of a pump as shown in figure 6.20 to 6.22. Figure 6.21 and 6.22 show the side
view and top view of the position of setting up water jet in scroll, respectively. Each
water jet is position in opposite way namely, 180 degree. The position of water jet
should be near the tip of impeller as much as impossible so the water jet will be inside
scroll and the scroll will be drilled. However, it should be far from the outlet of impeller
properly because the deflector is always moved up and down to deflect the water exiting
from jet. Figure 6.23 and 6.24 show that the distance from a center of scroll – at red
point– to a center of joint of deflector– at yellow point is 319.09 mm when the deflector
completely open and close. The distance from tip of impeller to the mouth of nozzle is
105.10 mm.
Page 70
61
Figure 6.19: The diameter of scroll
Figure 6.20: The position of 2 water jets with a selected pump
Page 71
62
Figure 6.21: The position of 2 water jets with a selected pump in side view
Figure 6.22: The position of 2 water jets with a selected pump in top view
Page 72
63
Figure 6.23: The position of 2 water jet when deflector completely opens
Figure 6.24: The position of 2 water jet when deflector completely closes
Page 73
64
According to figure 6.20, there are problems to locate these distributors into the scroll
of a pump. In other words, due to volute scroll, the deflector a number 2 is bigger than
the diameter of scroll that means this deflector cannot entirely move to open and close.
Moreover, when the water is injected from nozzle, the water will be deflected by
deflector. After deflecting of water, water is departed from the pump; this water will
flow along with the scroll and flow out at exit of scroll. In order to solve first problem,
some solution will be considered, for example, the design of deflector should be new
modify. In this study will not include the way how to solving these problems.
Page 74
65
CHAPTER 7
CONCLUSION AND RECOMMENDATION
7.1 Conclusion
Although Pump working as Turbines are used to replace the conventional turbine in
Micro hydropower projects, they still have a main drawback; they cannot control the
flow. Thus, Pump working as Turbine in Impulse mode is a new concept to reduce this
drawback by incorporating turbine distributor such as spear valve and nozzle from
Impulse turbine with a Centrifugal pump.
Before optimization of Impulse PAT, however, the geometries of pump have to be
known such as the width of outlet of impeller, the blade angel and so on. Although,
pump vendors provide typical information of pump, they do not provide the required
geometries of pump that will be used in optimization Impulse PAT. Thus, this study
firstly develop a pump selection algorithm to calculate these geometries based on given
typical data and characteristic curve of pump from Grundfos’ pump company. From
these given data such as outlet diameter of impeller, revolution of pump, inlet diameter
of impeller, H-Q curve and efficiency curve, the required geometries to design and
optimize the PAT in turbine mode can be estimated. Once a feasible pump has been
identified, the peripheral speed of Impulse PAT will be optimized in order to find the
best efficiency point of PAT in impulse mode.
The design point of head and flow rate that have origin from Ban Khun Pae
Hydropower Plant are 100 m and 0.1 m3/s. The given typical data dimensional drawing
and characteristic curves of selected pump from Gundfos’ catalogue are shown in table
4.2, table 4.5 and figure 4.5, 4.6. After use the given data to calculate the required
geometries of a selected pump, the number of jet will be considered. In case study, if the
number of jet is more than 4, a new selected pump from pump’s catalogue will be
considered in condition; different operating point of head and flow rate but the same
power. In this case study, two pumps in different operating design point of head and
flow rate are considered because the number of water jet for first pump is more than 4.
Thus, a new selected pump with operating point of a half of head and double flow rate
will be consider. The results from finding the number of jet are shown in table 4.4 and
4.7. However, the accuracy of calculating geometries of selected pump still has to be
validated because the validation method required the acquisition of geometries of
geometry of a real pump.
The optimization of Impulse PAT shows that its peripheral speed can be operated lower
than water jet’s velocity. Moreover, when Impulse PAT operates at the peripheral speed
that is nearly a half of velocity of water jet, its efficiency will be close to maximum.
That is similar to the theoretical performance of the Pelton turbine, at the maximum
efficiency, the rotational velocity is half of jet velocity. In this case study, the losses
from spear valve with nozzle and selected pump will be considered in order to
calculating the efficiency more accuracy. Hence, two cases studies namely; first case is
optimization of Impulse PAT without including the losses from spear valve with nozzle
and selected pump. The second case is optimization of Impulse PAT with including the
losses from spear valve with nozzle and selected pump. For the first case study, Impulse
PAT can operate from 0 to 1174 rpm and the maximum efficiency without losses from
calculation is 81.972% at the 1174 rpm. However, this point of maximum efficiency
(theoretical point) can occur when the outlet relative velocity is zero that means there
are no flow rate can go through in Impulse PAT. Moreover, the practical point of rpm
that the flow can go through the Impulse PAT is 1005 rpm. And the efficiency at 1005
Page 75
66
rpm is 49.833%. For the second case study with including the friction losses, Impulse
PAT can operate from 487.8 to 1138.87 rpm and the maximum efficiency including
losses is 82.493% at the 1138.87 rpm. However, this point of maximum efficiency can
occur when the outlet relative velocity is zero that means there are no flow rate can go
through in Impulse PAT. Moreover, the practical point of rpm that the flow can go
through the Impulse PAT is 960 rpm. And the efficiency at 960 rpm is 43.784%.
7.2 Recommendation for case study
The recommendations for this thesis are as follow:
1. To make more reliable, the validation between the calculation’s results and the
real pump should be considered by experimental or Computational Fluid
Dynamics (CFD).
2. From a graph result in figure 5.6, the efficiency of Impulse PAT is maximum
when operate at 1138.87 rpm. However, this point of rpm is not easy to operate.
Thus, there are three recommendations to make it easy.
Firstly, an accessory device such as gearbox with 50 Hz will be used in
order to adjust the speed of a pump. Nevertheless, setting a gearbox lead
to more cost and the efficiency of pump after apply gearbox will drop a
little as more or less 2% by estimate.
Second alterative is operating Impulse PAT at 1000 rpm because it is
easier to operate rpm as 1000 rpm. But the efficiency of Impulse PAT is
decreased from 82.493% to 49.341%.
Third suggested alternative is change the best efficiency point (BEP) at
1000 rpm or 1500 rpm because it is easier to operate Impulse PAT at
these rpm. Hence, reducing the rpm that have a BEP can be done by
modified by selecting a new pump. In other word, the size of Impulse
PAT will be changed. For example, according to figure5.6, if the BEP of
Impulse PAT is on 1000 rpm, the diameter of new selected pump should
be bigger because the speed of Impulse PAT reverses variation with the
diameter.
3. According to the result from optimization Impulse PAT, the rotational speed is
nearly a half of velocity of water jet at the maximum efficiency point. Hence,
developing the selection pump for using as Impulse turbine is select a pump that
have the rotational speed that is closely a half of the velocity of water jet.
4. Standardize spear valve design and fixture in standard sizes which would be
fabricated in advance.
Page 76
67
REFERENCES
[1] Akella, K., Saini, P., Sharma, P. (2009). Social, Economical and Environmental
Impacts of Renewable Energy Systems. Renewable Energy, 34(2), 390-396.
[2] Nautiyal, H., Varun, A. and Yadav, S. (2011). Experimental Investigation of
Centrifugal Pump Working as Turbine for Small Hydropower Systems. Energy
Science and Technology.
[3] Tamm, A., Braten, A., Stoffel, B. and Ludwig, G. (2000). Analysis of a Standard
Pump in Reverse Operation Using CFD. 20th IAHR-Symposium, Charlotte North
Carolina USA.
[4] N. G. (1992). Inexpensive Small Hydropower Stations for Small Areas of
Developing Countries. Conference on Advance in Planning-Design and
Management of Irrigation Systems as Related to Sustainable Land Use,
Louvain, Belgium.
[5] William, A. (1992). Pumps as Turbines Users Guide. International Technology
Publications, London.
[6] Singh, P. and Nestmann, F. (2010). An Optimization Routine on a Prediction
and Selection Model for the Turbine Operation of Centrifugal Pumps.
Experimental Thermal and Fluid Science, 34,152-164.
[7] Derakhshan S. (2008). Theoretical, Numerical and Experimental Investigation of
Centrifugal Pumps in Reverse Operation. Experimental Thermal and Fluid
Science, 32, 1620-1627.
[8] Patel, V. A., Jain, S. V., Motwani, K. H., and Patel, R. N. (2013). Numerical
optimization of guide vanes and reducer in pump running in turbine mode.
Procedia Engeneering, 51,797-802.
[9] Cengel, A. and John M. (2010). Fluid Mechanics: Fundamentals and
Applications. 2nd McGraw-Hill.
[10] Friedrich, J. (2008). Centrifugal Pumps. Springer 2008.Heidelberg
[11] Jacobsen, C. B. (1998). The Centrifugal Pump, GRUNDFOS Research and
Technology
[12] Frank, M.W. Fluid Mechanics. McGraw-Hill, New York.
[13] LeConte, J. N. (1928). Hydraulic. McGraw-Hill, New York.
[14] Information of a pump and characteristic curve of Grundfos, retrieved from
http://net.grundfos.com/Appl/WebCAPS/PdfStreamer.pdf?productno=9662686
[15] Motwan, K. H., Jain, S. V. and Patel, R. N. (2013). Cost Analysis of Pump as
Turbine for Pico Hydropower Plants. Procedia Engeneering, 51, 721-726.
[16] Laghari, J.A. (2013). A Comprehensive Overview of New Design in the
Hydraulic, Electrical Equipments and Controllers of Mini Hydro Power Plants
Making It Cost Effective Technology. Renewable and Sustainable Energy
Reviews, 20, 279-293.
[17] Stepanoff, A. J. (1957). Centrifugal and Axial Flow Pumps, Design and
Applications. John Wiley and Sons, New York.
[18] Singh, P. (2005). Optimization of the Internal Hydraulic and of System Design
in Pumps as Turbines with Field Implementation and Evaluation. Ph.D. Thesis,
University of Karlsruhe, Germany.
[19] Karan, S. (2014). The Feasibility Study of Using PUMP-as-TURBINE (PaT) in
the Impulse Turbine Mode. Speacial study report, Asian Institute of Technology,
Thailand.
[20] Korpela, S. (2011). Principle of Turbomachinery. John Wiley and Sons.
Page 77
68
[21] Logan, J. E. (1981). Turbomachinery; Basic Theory and Applications. Marcel-
Dekker, New York
[22] Nautiyal, H. and Kumar, V. (2010). Reverse Running Pumps Analytical,
Experimental and Computational Study: A review. Renew Sustain Energy 2010.
[23] Khin,T., Mya, K. and Khin, A. (2008). Design and Performance Analysis of
Centrifugal Pump. World Academy of Science, Engineering and Technology.
[24] Ravi, S., Singh, A. and Singh, M. (2014). Analysis About Losses of Centrifugal
Pump by MATLAB. International Journal of Computational Engineering
Research.
[25] Nechleba, M. (1957). Hydraulic Turbines: Their Design and Equipment. ARTIA,
Prague.
[26] Yang, S. (2012). Theoretical, Numerical and Experimental Prediction of Pump
as Turbine Performance. Renewable Energy, 507-513.
[27] PEA. (2013). The Design and Development of Efficient Turbine for Micro
Hydropower Plant of Provincial Electricity Authority. School of Engineering
and Technology Asian Institute of Technology 2013, Thailand.
[28] Karadzic, U., Bergant, A. and Vukoslavcevic, P. (2009). Anovel Pelton Turbine
Model for Water Hammer Analysis. Journal of Mechanical Engineering , 369-
380.
[29] Information of a circular segment formula and the detail of formula, retrieved
from http://en.wikipedia.org/wiki/Circular_segment
Page 78
69
APPENDIXES
Appendixes A
A 1. General data of pump from Grundfos company [11]
Appendixes B
B 1. Pump using as Impulse turbine in reverse mode in first option at rotational speed
from 2000 to 2500 rpm
B 2. Pump using as Impulse turbine in backward mode in second option at rotational
speed from 500 to 1790 rpm
Appendixes C
C 1. Area between the needle and nozzle at exist of jet
C 2. Area of the circular segment at the exit of jet when deflecting deflector
Appendixes D
D 1. MATLAB program code for selected pump with the head is 100 m and
discharge 0.1 m3/s
D 2. MATLAB program code for selected pump with the head is 50 m and discharge
0.2 m3/s
D 3. MATLAB program code for optimization without including the losses
D 4. MATLAB program code for optimization with including the losses
Page 79
70
Appendixes A
A 1. General data of pump from Grundfos company [11]
Figure A 1.1: General data of Grundfos’s pump
Page 80
71
Appendixes B
B 1. Pump using as Impulse turbine in reverse mode in first option at rotational speed
from 2000 to 2500 rpm
Figure B 1.1 : Pump using as Impulse turbine in reverse mode in first option at
rotational speed 2000 rpm
Figure B 1.2: Pump using as Impulse turbine in reverse mode in first option at
rotational speed 2250 rpm
Page 81
72
Figure B 1.3: Pump using as Impulse turbine in reverse mode in first option at
rotational speed 2500 rpm
B 2. Pump using as Impulse turbine in backward mode in second option at rotational
speed from 500 to 1790 rpm
Figure B 2.1: Pump using as Impulse turbine in backward mode in second option at
rotational speed 500 rpm
Page 82
73
Figure B 2.2: Pump using as Impulse turbine in backward mode in second option at
rotational speed 750 rpm
Figure B 2.3: Pump using as Impulse turbine in backward mode in second option at
rotational speed 1000 rpm
Page 83
74
Figure B 2.4: Pump using as Impulse turbine in backward mode in second option at
rotational speed 1250 rpm
Figure B 2.5: Pump using as Impulse turbine in backward mode in second option at
rotational speed 1360.4 rpm
Page 84
75
Figure B 2.6: Pump using as Impulse turbine in backward mode in second option at
rotational speed 1790 rpm
Page 85
76
Appendixes C
C 1. Area between the needle and nozzle at exist of jet
Area between the needle and nozzle at exist of jet is Truncated cone and can be
calculated as 2 2( ) ( )jA a b b a h
Where ,a band h represent to the position that is shown in figure 6.13.
Figure C 1.1: Typical example of exit area of jet
C 2. Area of the circular segment at the exit of jet when deflecting deflector
When the deflector is deflected in different position, the area of the exit of jet will
change. The area of the exit of jet by deflecting deflector is a circular segment and can
be calculated as [29]
2
( sin )2
RA
Where θ is the central angle in radians that is shown in the figure C 2.1. This angle can
be found as [29] 2 cosd
acrcR
Where
R is the radius of the jet.
d is the height of the triangular portion.
h is the height of the segment
c the chord length
s the arc length
Page 86
77
Figure C.2.1: A circular segment (in green) [29]
Page 87
78
Appendixes D
D 1. MATLAB program code for selected pump with the head is 100 m and
discharge 0.1 m3/sec
clear all % Pump geometry estimation from Grundfos % we select a grundfos pump from the catalog online % Pump data are given by Grundfos
g=9.81; % m/s^2 rho=998.2; % Water density QBEP=0.1; % 360 m^3/min or 0.1 m^3/s HBEP=100; % Head BEP in m H0=108.6111; % Static head of pump when Q=0 taken
from pump H/Q curve from Grundfos rpm= 2980*1.03; % Revolutions/min effBEP=0.806; % Pump efficiency Q=[0;50;100;150;200;250;300;360;400;449]; % The flow rate in
different point from given H-Q curve in m^3/hour Q=Q./3600; % The flow rate in
different point from given H-Q curve in m^3/sec H=[108.6;111;113;113;112;110;106;100;93.9;82.5]; % The head
in different point from given H-Q curve in
m eff=[0;0.259;0.451;0.59;0.689;0.754;0.792;0.806;0.796;0.76]; % The
efficiency in different point from given efficiency-
Q curve di= 0.270; % Outlet impeller diameter in m d2=di; r2=d2/2; % Impeller radius at outlet in m deye=0.125; % The diameter of the eye from pump
drawing given by Grundfos reye=deye/2; % The radius of the eye ueye=deye*pi*rpm/60; % The peripheral speed of pump's eye dshaft=0.032; % The diameter of shaft in m rshaft=dshaft/2; % The radius of shaft in m u1=ueye % The peripheral speed at the inlet
of impeller A1=pi*(reye-rshaft)^2 % The area at inlet of impeller dout=0.1; % The diameter at the exit of the
scroll from drawing A0=(pi*dout^2)/4; % The area at the exit of the scroll v0=Q./A0; % The velocity at the exit of scroll
% We can now compute the approximate geometries of the impeller u2=pi*d2*rpm/60 % The peripheral speed at the outlet
of impeller omega=2*pi*rpm/60; % Omega H0_euler=(u2^2)/g % The Euler's head at the shut-off
point in m H0_loss=(H0_euler-H0)/2 % The Total losses in pump at shut-
off point in m v1=Q./A1 % Absolute velocity at inlet of
impller d1=u1*60/(pi*rpm) % Inlet impeller diameter in m b1=A1/(pi*d1) % Width of impeller at inlet in m w1=(u1^2+v1.^2).^0.5 % Relative velocity at the inlet of
impeller beta1=acosd(u1./w1) % Blade angle at inlet of impeller in
degree
Page 88
79
beta1_radial=acos(u1./w1); % Blade angle at inlet of impeller in
radius op_eff=(1-eff); Hloss=(op_eff.*H./eff)/2 % The Total losses in pump
in different point in m Hloss(1,1)=H0_loss HlossBEP=((1-effBEP)*HBEP/effBEP)/2 % The Total losses in pump
at BEP in m v2t=g.*H./(eff.*u2) % The tangential velocity
at outlet of impeller v2t(1,1)=u2 w2t=u2-v2t % Tangential relative
velocity at outlet of impeller con=u2^2-2*u2.*w2t+v0.^2-2*u2.*v0+2.*v0.*w2t; % The constant
velue of term to solve the equation power four to find the outlet
relative velocity M=v1.^2+v0.^2+0.25.*(w1.^2); % we divide the proportion for shock loss, recirculation loss and
leakage % loss at Q=0 p=0; for i=1:1:9 m=(0.7-p); Hr0(i,1)=(0.7-p)*H0_loss; % Recirculation loss
at shut-off point % Hr0=(kr*(omega^3)*(d1^2)) kr(i,1)=Hr0(i,1)/((omega^3)*(d1^2)); % Coefficient of
Recirculation loss
for ii=1:1:2 Hl0(i,ii)=0.075*ii*H0_loss; % Leakage losses at
the shut-off point Hs0(i,ii)=H0_loss-(Hl0(i,ii)+Hr0(i,1)); % Shock loss at the
shut-off point ks(i,ii)=Hs0(i,ii)/(QBEP)^2; % Coefficient of
Shock loss for j=1:1:length(Q) % Num 1 represent for leakage loss 7.5 percent % Num 2 represent for leakage loss 15 percent % hs(j,ii)=ks(i,ii)*(Q(j,1)-QBEP)^2; % Start to calculate the losses if ii==1 hs1(j,i)=ks(i,ii)*(Q(j,1)-QBEP)^2; %Shock loss for
leakage 7.5% at shut-off point Hl01(j,i)=Hl0(i,ii); %Leakage loss is
7.5% of total losses at shut-off point and other point % The total losses if j==1 Hloss(1,i)=H0_loss; % The total losses
at the shut-off point end if j>1 Hloss(j,i)=Hloss(j,1); % The total losses
in other point end end if ii==2 hs2(j,i)=ks(i,ii)*(Q(j,1)-QBEP)^2; %Shock loss for
leakage 15% at shut-off point Hl02(j,i)=Hl0(i,ii); %Leakage loss is
15% of total losses at shut-off point and other point
Page 89
80
end q(j,1)=QBEP/Q(j,1); %The ratio between
Qat BEP and Q in other point hr(j,i)=kr(i,1)*((omega^3)*(d1^2))*(1-(1/q(j,1)))^2.5;
% Recirculation loss t(j,i)=isreal(hr(j,i)); if t(j,i)~=1 hr(j,i)=0; end end
end p=p+0.05; end kr=kr hf1=Hloss-hs1-Hl01-hr % Friction loss for
leakage 7.5% at shut-off point hf2=Hloss-hs2-Hl02-hr % Friction loss for
leakage 15% at shut-off point % The constant to solve the equation power four to find the outlet
relative velocity for i=1:1:length(hf1) c1(i,:)=2*g.*hf1(i,:)-con(i,1); c2(i,:)=2*g.*hf2(i,:)-con(i,1); A(i,1)=(q(i,1)^2)/4; B(i,1)=(q(i,1)^2)*w1(i,1)/2; ll(i,1)=Q(i,1); if Q(i,1)==QBEP CBEP1=c1(i,1); CBEP2=c2(i,1); w1m=w1(i,1); Mm=M(i,1); end end n=0; % The method to calculate relative velocity at outlet of impeller for i=1:1:length(Q)
for mm=1:1:length(c1(i,:)) n=n+1; C1(i,mm)=(q(i,1)^2)*M(i,1)-1-CBEP1/4+c1(i,mm)*(q(i,1)^2)/4; C2(i,mm)=(q(i,1)^2)*M(i,1)-1-CBEP2/4+c2(i,mm)*(q(i,1)^2)/4; D1(i,mm)=0.5*w1m*q(i,1)*C1(i,mm)-CBEP1*w1(i,1)/2; D2(i,mm)=0.5*w1m*q(i,1)*C2(i,mm)-CBEP2*w1(i,1)/2; E1(i,mm)=c1(i,mm)*Mm-CBEP1*M(i,1); E2(i,mm)=c2(i,mm)*Mm-CBEP2*M(i,1); p1(n,:)=horzcat(A(i,1),B(i,1),C1(i,mm),D1(i,mm),E1(i,mm)); p2(n,:)=horzcat(A(i,1),B(i,1),C2(i,mm),D2(i,mm),E2(i,mm)); end end R1=zeros(length(p1),4); R2=zeros(length(p1),4); for ii=10:1:length(p1)
R1(ii,:)=roots(p1(ii,:)); R2(ii,:)=roots(p2(ii,:));
for j=1:1:length(R1(ii,:)) t1(ii,j)=isreal(R1(ii,j));
Page 90
81
t2(ii,j)=isreal(R2(ii,j)); if t1(ii,j)~=1 y1(ii,j)=R1(ii,j); end if t2(ii,j)~=1 y2(ii,j)=R2(ii,j); end end end root1=R1; w21=R1-y1; % Outlet relative velocity for
leakage 7.5% at shut-off point root2=R2; w22=R2-y2; % Outlet relative velocity for
leakage 15% at shut-off point for ll=1:1:length(c1) for kk=1:1:length(w21) for jj=1:1:length(w21(kk,:)) if w21(kk,jj)~=0 % Find the coeficient of
friction loss for leakage 7.5% at shut-off
point k1(kk,jj)=(c1(ll,jj)-
(w21(kk,jj)^2))/(M(ll,1)+0.5*w1(ll,1)*w21(kk,jj)+0.25*(w21(kk,jj)^2)); else k1(kk,jj)= 0; end end end end k1=k1; for ll=1:1:length(c1) for kk=1:1:length(w22) for jj=1:1:length(w22(kk,:)) if w22(kk,jj)~=0 % Find the coeficient of
friction loss for leakage 15% at shut-off
point k2(kk,jj)=(c2(ll,jj)-
(w22(kk,jj)^2))/(M(ll,1)+0.5*w1(ll,1)*w22(kk,jj)+0.25*(w22(kk,jj)^2)); else k2(kk,jj)= 0; end end end end k2=k2; oo=1; for mm=1:1:length(w21) for hh=1:1:length(w22(kk,:)) beta21_radial(mm,hh)=acos(w2t(oo,1)/w21(mm,hh)); %blade angle
in radius at outlet for leakage 7.5% at shut-off point beta21(mm,hh)=acosd(w2t(oo,1)/w21(mm,hh)); %blade angle
in degree at outlet for leakage 7.5% at shut-off point beta22_radial(mm,hh)=acos(w2t(oo,1)/w22(mm,hh)); %blade angle
in radius at outlet for leakage 15% at shut-off point beta22(mm,hh)=acosd(w2t(oo,1)/w22(mm,hh)); %blade angle
in degree outlet for leakage 15% at shut-off point v21(mm,hh)=sqrt(u2^2+R1(mm,hh)^2-2*u2*w2t(oo,1)); %absolute
velocity at outlet for leakage 7.5% at shut-off point v22(mm,hh)=sqrt(u2^2+R2(mm,hh)^2-2*u2*w2t(oo,1)); %absolute
velocity at outlet for leakage 15% at shut-off point
Page 91
82
vr21(mm,hh)=w21(mm,hh)*sin(beta21_radial(mm,hh)); %radial
velocity at outlet for leakage 7.5% at shut-off point vr22(mm,hh)=w22(mm,hh)*sin(beta22_radial(mm,hh)); %radial
velocity at outlet for leakage 15% at shut-off point A21(mm,hh)=Q(oo,1)/vr21(mm,hh); %Area at
outlet for leakage 7.5% at shut-off point A22(mm,hh)=Q(oo,1)/vr22(mm,hh); %Area at
outlet for leakage 15% at shut-off point b21(mm,hh)=A21(mm,hh)/(2*pi*r2); %Width of
impeller at outlet for leakage 7.5% at shut-off point b22(mm,hh)=A22(mm,hh)/(2*pi*r2); %Width of
impeller at outlet for leakage 15% at shut-off point i1(mm,hh)=isreal(v21(mm,hh)); %Cheack the
ral number and the imagine number i2(mm,hh)=isreal(v22(mm,hh)); if i1(mm,hh)~=1 l1(mm,hh)=v21(mm,hh); else l1(mm,hh)=0; end if i2(mm,hh)~=1 l2(mm,hh)=v22(mm,hh); else l2(mm,hh)=0; end V21(mm,hh)=v21(mm,hh)-l1(mm,hh); % Absolute velocity
at outlet for leakage 7.5% at shut-off point that is not
complex number V22(mm,hh)=v22(mm,hh)-l2(mm,hh); % Absolute velocity
at outlet for leakage 15% at shut-off point that is not complex
number end pp=mod(mm,9); if pp==0 oo=oo+1; end if pp~=0 oo=oo+0; end uu=oo; end oo=1; for mm=1:1:length(V21) for hh=1:1:length(V22(kk,:)) alfha21(mm,hh)=acosd(v2t(oo,1)/V21(mm,hh)); %Angle of absolute
velocity in degree at outlet for leakage 7.5% at shut-off point alfha21_radial(mm,hh)= alfha21(mm,hh)*pi/180; %Angle of absolute
velocity in radial at outlet for leakage 7.5% at shut-off point alfha22(mm,hh)=acosd(v2t(oo,1)/V22(mm,hh)); %Angle of absolute
velocity in degree at outlet for leakage 15% at shut-off point alfha22_radial(mm,hh)=alfha22(mm,hh)*pi/180; %Angle of absolute
velocity in radial at outlet for leakage 15% at shut-off point end pp=mod(mm,9); if pp==0 oo=oo+1; end if pp~=0 oo=oo+0; end uu=oo;
Page 92
83
end vjet=sqrt(2*g*H(8,1)) %Velocity of the water jet djet=sqrt(Q(8,1)*4/(vjet*pi)) % Diameter of the water jet num1=(djet/b21(46,4))^2 % Appropriate number of the
water jet for leakage 7.5% at shut-off point num2=(djet/b22(46,4))^2 % Appropriate number of
the water jet for leakage 15% at shut-off point
D 2. MATLAB program code for selected pump operating with the head is 50 m and
discharge 0.2 m3/sec
The code is a same with D1. However, The Data of the head, flow rate and efficiency
on the characteristic curves. clear all % Pump geometry estimation from Grundfos % we select a grundfos pump from the catalog online % Pump data are given by Grundfos
g=9.81;
rho=998.2; QBEP=825/3600;
HBEP=47.7; H0=55.2;
rpm= 2980*0.8;
effBEP=0.852;
Q=[0;100;200;300;400;500;600;720;825]; Q=Q./3600; H=[55.2;58.8;61.3;60.5;60;56.7;54.9;50;47.7]; eff=[0;32.1;53.1;66.4;74.6;80;82.7;84.1;85.2]; eff=eff./100 di= 0.265; d2=di; r2=d2/2; deye=0.2;
reye=deye/2; ueye=deye*pi*rpm/60; dshaft=0.042; rshaft=dshaft/2; ushaft=dshaft*pi*rpm/60; dout=0.15; A0=(pi*dout^2)/4;%m from pump drawing u1mean=(ueye+ushaft)/2; u1=ueye A1=pi*(reye^2-rshaft^2) v0=Q./A0; %we can now compute the approximate geometry of the impeller u2=pi*d2*rpm/60 omega=2*pi*rpm/60; H0_euler=(u2^2)/g H0_loss=(H0_euler-H0)/2 v1=Q./A1 d1=u1*60/(pi*rpm) b1=A1/(pi*d1) w1=(u1^2+v1.^2).^0.5 beta1=acosd(u1./w1) beta1_radial=acos(u1./w1); op_eff=(1-eff); Hloss=(op_eff.*H./eff)/2 Hloss(1,1)=H0_loss
Page 93
84
HlossBEP=((1-effBEP)*HBEP/effBEP)/2 v2t=g.*H./(eff.*u2) v2t(1,1)=u2 w2t=u2-v2t con=u2^2-2*u2.*w2t+v0.^2-2*u2.*v0+2.*v0.*w2t; M=v1.^2+v0.^2+0.25.*(w1.^2); % we divide the proportion for shock loss, recirculation loss and
leakage % loss at Q=0 p=0; for i=1:1:9 m=(0.7-p); Hr0(i,1)=(0.7-p)*H0_loss; % Hr0=(kr*(omega^3)*(d1^2)) kr(i,1)=Hr0(i,1)/((omega^3)*(d1^2));
for ii=1:1:2 Hl0(i,ii)=0.075*ii*H0_loss; Hs0(i,ii)=H0_loss-(Hl0(i,ii)+Hr0(i,1)); ks(i,ii)=Hs0(i,ii)/(QBEP)^2; for j=1:1:length(Q) % num 1 represent for leakage loss 5 percent % num 2 represent for leakage loss 10 percent % hs(j,ii)=ks(i,ii)*(Q(j,1)-QBEP)^2; if ii==1 hs1(j,i)=ks(i,ii)*(Q(j,1)-QBEP)^2; Hl01(j,i)=Hl0(i,ii); if j==1 Hloss(1,i)=H0_loss; end if j>1 Hloss(j,i)=Hloss(j,1); end end if ii==2 hs2(j,i)=ks(i,ii)*(Q(j,1)-QBEP)^2; Hl02(j,i)=Hl0(i,ii);
end q(j,1)=QBEP/Q(j,1); hr(j,i)=kr(i,1)*((omega^3)*(d1^2))*(1-(1/q(j,1)))^2.5; t(j,i)=isreal(hr(j,i)); if t(j,i)~=1 hr(j,i)=0; end end
end p=p+0.05; end kr=kr hf1=Hloss-hs1-Hl01-hr hf2=Hloss-hs2-Hl02-hr
for i=1:1:length(hf1) c1(i,:)=2*g.*hf1(i,:)-con(i,1); c2(i,:)=2*g.*hf2(i,:)-con(i,1); A(i,1)=(q(i,1)^2)/4; B(i,1)=(q(i,1)^2)*w1(i,1)/2; ll(i,1)=Q(i,1); if Q(i,1)==QBEP
Page 94
85
CBEP1=c1(i,1); CBEP2=c2(i,1); w1m=w1(i,1); Mm=M(i,1); end end n=0; for i=1:1:length(Q) for mm=1:1:length(c1(i,:)) n=n+1; C1(i,mm)=(q(i,1)^2)*M(i,1)-1-CBEP1/4+c1(i,mm)*(q(i,1)^2)/4; C2(i,mm)=(q(i,1)^2)*M(i,1)-1-CBEP2/4+c2(i,mm)*(q(i,1)^2)/4; D1(i,mm)=0.5*w1m*q(i,1)*C1(i,mm)-CBEP1*w1(i,1)/2; D2(i,mm)=0.5*w1m*q(i,1)*C2(i,mm)-CBEP2*w1(i,1)/2; E1(i,mm)=c1(i,mm)*Mm-CBEP1*M(i,1); E2(i,mm)=c2(i,mm)*Mm-CBEP2*M(i,1); p1(n,:)=horzcat(A(i,1),B(i,1),C1(i,mm),D1(i,mm),E1(i,mm)); p2(n,:)=horzcat(A(i,1),B(i,1),C2(i,mm),D2(i,mm),E2(i,mm)); end end R1=zeros(length(p1),4); R2=zeros(length(p1),4); for ii=10:1:length(p1)
R1(ii,:)=roots(p1(ii,:)); R2(ii,:)=roots(p2(ii,:));
for j=1:1:length(R1(ii,:)) t1(ii,j)=isreal(R1(ii,j)); t2(ii,j)=isreal(R2(ii,j)); if t1(ii,j)~=1 y1(ii,j)=R1(ii,j); else y1(ii,j)=0; end if t2(ii,j)~=1 y2(ii,j)=R2(ii,j); else y2(ii,j)=0; end end end root1=R1; w21=R1-y1; root2=R2; w22=R2-y2; for ll=1:1:length(c1) for kk=1:1:length(w21) for jj=1:1:length(w21(kk,:)) if w21(kk,jj)~=0 k1(kk,jj)=(c1(ll,jj)-
(w21(kk,jj)^2))/(M(ll,1)+0.5*w1(ll,1)*w21(kk,jj)+0.25*(w21(kk,jj)^2)); else k1(kk,jj)= 0; end end end end k1=k1; for ll=1:1:length(c1) for kk=1:1:length(w22)
Page 95
86
for jj=1:1:length(w22(kk,:)) if w22(kk,jj)~=0 k2(kk,jj)=(c2(ll,jj)-
(w22(kk,jj)^2))/(M(ll,1)+0.5*w1(ll,1)*w22(kk,jj)+0.25*(w22(kk,jj)^2)); else k2(kk,jj)= 0; end end end end k2=k2; oo=1; for mm=1:1:length(w21) for hh=1:1:length(w22(kk,:)) beta21_radial(mm,hh)=acos(w2t(oo,1)/w21(mm,hh)); beta21(mm,hh)=acosd(w2t(oo,1)/w21(mm,hh)); beta22_radial(mm,hh)=acos(w2t(oo,1)/w22(mm,hh)); beta22(mm,hh)=acosd(w2t(oo,1)/w22(mm,hh)); v21(mm,hh)=sqrt(u2^2+R1(mm,hh)^2-2*u2*w2t(oo,1)); v22(mm,hh)=sqrt(u2^2+R2(mm,hh)^2-2*u2*w2t(oo,1)); vr21(mm,hh)=w21(mm,hh)*sin(beta21_radial(mm,hh)); vr22(mm,hh)=w22(mm,hh)*sin(beta22_radial(mm,hh)); A21(mm,hh)=Q(oo,1)/vr21(mm,hh); A22(mm,hh)=Q(oo,1)/vr22(mm,hh); b21(mm,hh)=A21(mm,hh)/(2*pi*r2); b22(mm,hh)=A22(mm,hh)/(2*pi*r2); i1(mm,hh)=isreal(v21(mm,hh)); i2(mm,hh)=isreal(v22(mm,hh)); if i1(mm,hh)~=1 l1(mm,hh)=v21(mm,hh); else l1(mm,hh)=0; end if i2(mm,hh)~=1 l2(mm,hh)=v22(mm,hh); else l2(mm,hh)=0; end V21(mm,hh)=v21(mm,hh)-l1(mm,hh); V22(mm,hh)=v22(mm,hh)-l2(mm,hh); end pp=mod(mm,9); if pp==0 oo=oo+1; end if pp~=0 oo=oo+0; end uu=oo; end oo=1; for mm=1:1:length(V21) for hh=1:1:length(V22(kk,:)) alfha21(mm,hh)=acosd(v2t(oo,1)/V21(mm,hh)); alfha21_radial(mm,hh)= alfha21(mm,hh)*pi/180; alfha22(mm,hh)=acosd(v2t(oo,1)/V22(mm,hh)); alfha22_radial(mm,hh)=alfha22(mm,hh)*pi/180; end pp=mod(mm,9); if pp==0 oo=oo+1;
Page 96
87
end if pp~=0 oo=oo+0; end uu=oo; end vjet=sqrt(2*g*100) djet=sqrt(0.1*4/(vjet*pi)) num1=(djet/b21(28,4))^2
D 3. MATLAB program code for optimization without including the losses
clear all; % Turbine performance evaluation procedure % Reverse running pump % We inject the water opposite with outlet velocity of centrifugal
pump % we represent subscribe number 3 as the outlet of centrifugal pump
working as turbine % and use number 4 as the outlet of centrifugal pump working as
Impulse turbine also % jet velocity, we can calculate from Head g=9.81; % Gravity rho=998.2; % Water density Q=0.1; % The design point of flow rate H=100; % The design point of head b4=0.048; % Width of outlet of Impeller of
Impulse PAT b3=0.0460; % Width of inlet of Impeller of
Impulse PAT d4=0.3000; % Outlet diameter of Impulse PAT d3=0.4250; % Inlet diameter of Impulse PAT A3=pi*d3.*b3 % Inlet area of Impulse PAT A4=pi*d4*b4 % Outlet area of Impulse PAT beta4=8.1758 % Outlet blade angle of Impulse
PAT beta3=14.3507 % Inlet blade angle of Impulse
PAT beta_radial4=beta4*pi/180; beta_radial3=beta3.*pi./180; xx=0; % Iterate the RPM of Impulse turbine from 0 to 1174 for pp=1:1:(1174/50)+2 rpmt(pp,1)=xx*10 xx=xx+5; if rpmt(pp,1)>1174 rpmt(pp,1)=1174 end end rpm=1790*.76; % RPM of pump in normal mode v3=sqrt(2*g*H) % Velocity of water jet % From here,the known parameters that consist of beta1, beta2, b1, b2,
r1, % r2 will be used to calculation for kk=1:1:length(b3) for ll=1:1:length(rpmt) u3(kk,ll)=pi*d3*rpmt(ll,1)/60; % Peripheral speed at inlet
of Impulse PAT u4(kk,ll)=pi*d4*rpmt(ll,1)/60; % Peripheral speed at inlet
of Impulse PAT if v3>u3(kk,ll) % Condition for the option 2
Page 97
88
w3(kk,ll)=(-(-2*u3(kk,ll)*cos(pi-
beta_radial3(kk,1)))+sqrt((-2*u3(kk,ll)*cos(pi-beta_radial3(kk,1)))^2-
4*(u3(kk,ll)^2-v3^2)))/2; w3_1(kk,ll)=(-(-2*u3(kk,ll)*cos(pi-beta_radial3(kk,1)))-
sqrt((-2*u3(kk,ll)*cos(pi-beta_radial3(kk,1)))^2-4*(u3(kk,ll)^2-
v3^2)))/2; else % Condition for the option 1 w3(kk,ll)=(-(-2*u3(kk,ll)*cos(beta_radial3(kk,1)))-sqrt((-
2*u3(kk,ll)*cos(beta_radial3(kk,1)))^2-4*(u3(kk,ll)^2-v3^2)))/2; w3_1(kk,ll)=(-(-2*u3(kk,ll)*cos(beta_radial3(kk,1)))+sqrt((-
2*u3(kk,ll)*cos(beta_radial3(kk,1)))^2-4*(u3(kk,ll)^2-v3^2)))/2; end v3r(kk,ll)=w3(kk,ll)*sin(beta_radial3(kk,1)); % The radial
velocity at
inlet of Impulse PAT alfha3(kk,ll)=asind(v3r(kk,ll)/v3); % Absolutely
velocity angle at Inlet of Impulse
PAT in degree alfha3_radial(kk,ll)=asin(v3r(kk,ll)/v3); v3t(kk,ll)=v3*cos(alfha3_radial(kk,ll)); % Tangential
velocity at inlet of
Impulse PAT w4(kk,ll)=sqrt(w3(kk,ll)^2-(u3(kk,ll)^2-u4(kk,ll)^2)); % Relative
velocity at outlet of
Impulse PAT
if v3>u3(kk,ll) % The absolute velocity at outlet of Impulse PAT in
condition option 2 v4(kk,ll)=sqrt(w4(kk,ll)^2+u4(kk,ll)^2-2*u4(kk,ll)*w4(kk,ll)*cos(pi-
beta_radial4)); else % The absolute velocity at outlet of Impulse PAT in condition
option 1 v4(kk,ll)=sqrt(w4(kk,ll)^2+u4(kk,ll)^2-
2*u4(kk,ll)*w4(kk,ll)*cos(beta_radial4)); end v4r(kk,ll)=w4(kk,ll)*sin(beta_radial4) % The radial velocity at
outlet of Impulse PAT t1(kk,ll)=isreal(v4(kk,ll)); % Cut off the complex number of the outlet absolute velocity if t1(kk,ll)~=1 y1(kk,ll)=v4(kk,ll); else y1(kk,ll)=0; end V4(kk,ll)=v4(kk,ll)-y1(kk,ll) % Absolute velocity angle at outlet of Impulse PAT if V4(kk,ll)>0 alfha4(kk,ll)=asind(v4r(kk,ll)/V4(kk,ll)); else alfha4(kk,ll)=0 end alfha4_radial(kk,ll)=alfha4(kk,ll)*pi/180; % Tangentail velocity at outlet of Impulse PAT
v4t(kk,ll)=v4(kk,ll)*cos(alfha4_radial(kk,ll)); % Find efficiency of Impulse PAT if V4(kk,ll)==0 Pw(kk,ll)=0; Turbine_eff(kk,ll)=0 else Turbine_eff(kk,ll)=((v3^2-v4(kk,ll)^2)/v3^2)*100 end end
Page 98
89
end
D 4. MATLAB program code for optimization with including the losses
clear all; % Turbine performance evaluation procedure % Reverse running pump % We inject the water opposite with outlet velocity of centrifugal
pump % we represent subscribe number 3 as the outlet of centrifugal pump
working as turbine % and use number 4 as the outlet of centrifugal pump working as
Impulse turbine also % jet velocity, we can calculate from Head g=9.81; % Gravity rho=998.2; % Water density Q=0.1; % The design point of flow rate H=100; % The design point of head b4=0.048; % Width of outlet of Impeller of
Impulse PAT b3=0.0460; % Width of inlet of Impeller of
Impulse PAT d4=0.3000; % Outlet diameter of Impulse PAT d3=0.4250; % Inlet diameter of Impulse PAT A3=pi*d3.*b3 % Inlet area of Impulse PAT A4=pi*d4*b4 % Outlet area of Impulse PAT beta4=8.1758 % Outlet blade angle of Impulse
PAT beta3=14.3507 % Inlet blade angle of Impulse
PAT beta_radial4=beta4*pi/180; beta_radial3=beta3.*pi./180; xx=0; % Iterate the RPM of Impulse turbine from 0 to 1174 for pp=1:1:(1174/50)+2 rpmt(pp,1)=xx*10 xx=xx+5; if rpmt(pp,1)>1174 rpmt(pp,1)=1174 end end rpm=1790*.76; % RPM of pump in normal mode k=0.3310; % Coefficient of nozzle Cv=0.97; % Nozzle velocity coef Cn=1.25; % Nozzle dimater coef CS = 1.65; % Largest spear dia. coef. (1.5-
1.8) CSrod = 0.6; % Spear rod dia. coef. (0.58-0.7) CSL = 3.5; % Spear tip length coef. (3.25-
3.66) CFS = 1.16; % Stroke coef. CDN = 3.2; % Pipe dia. coef. (2.5-3.5) n_jet=2; %number of jet % Spear valve design v3=sqrt(2*g*H) % Velocity of water jet Vj = Cv*sqrt(2*g*H); % Jet velocity U1 = 0.46*Vj; % Circumferential speed dj = sqrt((4*Q)/(n_jet*pi*v3)); % Jet size (water) dn = Cn*dj; % Nozzle Diameter
Page 99
90
ds_large = CS*dj; % Largest spear diameter ds_rod = CSrod*dj; % Spear rod diameter SL = CSL*dj; % Spear tip length Stroke = CFS*dj; % Full stroke of spear valve DN = CDN*dj; % Pipe diameter (Nozzle) RN = 1.2*DN; % Nozzle holder radius alpha = 55*pi/180; % Spear tip angle (42-60
degree) beta = 80*pi/180; % Nozzle holder angle (60-90
degree)
% From here,the known parameters that consist of beta1, beta2, b1, b2,
r1, % r2 will be used to calculation for kk=1:1:length(b3) for ll=1:1:length(rpmt) u3(kk,ll)=pi*d3*rpmt(ll,1)/60; % Peripheral speed at inlet
of Impulse PAT u4(kk,ll)=pi*d4*rpmt(ll,1)/60; % Peripheral speed at inlet
of Impulse PAT if v3>u3(kk,ll) % Condition for the option 2 w3(kk,ll)=(-(-2*u3(kk,ll)*cos(pi-
beta_radial3(kk,1)))+sqrt((-2*u3(kk,ll)*cos(pi-beta_radial3(kk,1)))^2-
4*(u3(kk,ll)^2-v3^2)))/2; w3_1(kk,ll)=(-(-2*u3(kk,ll)*cos(pi-beta_radial3(kk,1)))-
sqrt((-2*u3(kk,ll)*cos(pi-beta_radial3(kk,1)))^2-4*(u3(kk,ll)^2-
v3^2)))/2; else % Condition for the option 1 w3(kk,ll)=(-(-2*u3(kk,ll)*cos(beta_radial3(kk,1)))-sqrt((-
2*u3(kk,ll)*cos(beta_radial3(kk,1)))^2-4*(u3(kk,ll)^2-v3^2)))/2; w3_1(kk,ll)=(-(-2*u3(kk,ll)*cos(beta_radial3(kk,1)))+sqrt((-
2*u3(kk,ll)*cos(beta_radial3(kk,1)))^2-4*(u3(kk,ll)^2-v3^2)))/2; end v3r(kk,ll)=w3(kk,ll)*sin(beta_radial3(kk,1)); % The radial
velocity at
inlet of Impulse PAT alfha3(kk,ll)=asind(v3r(kk,ll)/v3); % Absolutely
velocity angle at Inlet of Impulse
PAT in degree alfha3_radial(kk,ll)=asin(v3r(kk,ll)/v3); v3t(kk,ll)=v3*cos(alfha3_radial(kk,ll)); % Tangential
velocity at inlet of
Impulse PAT w4(kk,ll)=sqrt(w3(kk,ll)^2-(u3(kk,ll)^2-u4(kk,ll)^2)); % Relative
velocity at outlet of
Impulse PAT
if v3>u3(kk,ll) % The absolute velocity at outlet of Impulse PAT in
condition option 2 v4(kk,ll)=sqrt(w4(kk,ll)^2+u4(kk,ll)^2-2*u4(kk,ll)*w4(kk,ll)*cos(pi-
beta_radial4)); else % The absolute velocity at outlet of Impulse PAT in condition
option 1 v4(kk,ll)=sqrt(w4(kk,ll)^2+u4(kk,ll)^2-
2*u4(kk,ll)*w4(kk,ll)*cos(beta_radial4)); end v4r(kk,ll)=w4(kk,ll)*sin(beta_radial4) % The radial velocity at
outlet of Impulse PAT t1(kk,ll)=isreal(v4(kk,ll)); % Cut off the complex number of the outlet absolute velocity if t1(kk,ll)~=1 y1(kk,ll)=v4(kk,ll);
Page 100
91
else y1(kk,ll)=0; end V4(kk,ll)=v4(kk,ll)-y1(kk,ll) % Absolute velocity angle at outlet of Impulse PAT if V4(kk,ll)>0 alfha4(kk,ll)=asind(v4r(kk,ll)/V4(kk,ll)); else alfha4(kk,ll)=0 end alfha4_radial(kk,ll)=alfha4(kk,ll)*pi/180; % Tangentail velocity at outlet of Impulse PAT
v4t(kk,ll)=v4(kk,ll)*cos(alfha4_radial(kk,ll)); % Find efficiency of Impulse PAT if V4(kk,ll)==0 Pw(kk,ll)=0; Turbine_eff(kk,ll)=0 else Turbine_eff(kk,ll)=((v3^2-v4(kk,ll)^2)/v3^2)*100 end end end