Comparative study by numerical simulation of two methods for automatic flow control in centrifugal pumps

International Journal of Power Electronics and Drive Systems (IJPEDS)
Vol. 13, No. 3, September 2022, pp. 1365~1379
ISSN: 2088-8694, DOI: 10.11591/ijpeds.v13.i3.pp1365-1379  1365
Journal homepage: http://ijpeds.iaescore.com
Comparative study by numerical simulation of two methods for
automatic flow control in centrifugal pumps
Rogger José Andrade-Cedeno1,2
, Jesús Alberto Pérez-Rodríguez3
, Carlos David Amaya-Jaramillo4
,
Ciaddy Gina Rodríguez-Borges5
, Endrickson Ramón Vera-Cedeno2
, Luis Santiago Quiroz-Fernández6
,
Yolanda Eugenia Llosas-Albuerne3
1
Investigador asociado al Posgrado en Instrumentación, Facultad de Ciencias, Universidad Central de Venezuela (UCV), Caracas,
Venezuela
2
Unidad Académica de Formación Técnica y Tecnológica, Universidad Laica Eloy Alfaro de Manabí (ULEAM), Tosagua, Ecuador
3
Departamento de Electricidad, Facultad de Ciencias Matemáticas, Físicas y Químicas, Universidad Técnica de Manabí (UTM),
Portoviejo, Ecuador
4
Departamento de Electricidad, Facultad de Ciencias de la Ingeniería. Universidad Técnica Estatal de Quevedo (UTEQ), Quevedo,
Ecuador
5
Departamento de Ingeniería Industrial, Facultad de Ciencias Matemáticas, Físicas y Químicas, Universidad Técnica de Manabí (UTM).
Portoviejo, Ecuador
6
Departamento de Ingeniería Civil, Facultad de Ciencias Matemáticas, Físicas y Químicas, Universidad Técnica de Manabí (UTM),
Portoviejo, Ecuador
Article Info ABSTRACT
Article history:
Received May 12, 2022
Revised May 29, 2022
Accepted June 14, 2022
Flow rate and pressure are variables of great interest in the process control
industry, especially in pumping systems. With the help of modeling and
simulation, it is possible to understand the operation of these systems prior
to their construction, in addition to allowing their behavior to be analyzed in
various operational scenarios and testing different control strategies. In this
work, the hydraulic, mechanical, electrical and electronic models of the
different components of the system are studied. Two study cases for
automatic flow control are assembled, using Simscape library from
MATLAB/Simulink R2019b. This study cases are: i) with a control valve
and ii) with a variable speed drive. The simulations determined that both
cases keep the flow constant in the face of pressure disturbances, with a
good dynamic response. Case 2 consumes less power than case 1, between
24 and 64% less, especially at low flow rates. It also reduce unwanted
mechanical and electrical problems due to sudden starts and stops. Case 2
produced harmonic pollution on both the grid and motor sides, which
implies a potential risk for the motor and the electrical grid. Case 2 was
experimentally validated, obtaining errors of less than 10%.
Keywords:
Energy efficiency
Flow control
Modeling and simulation
Pumping systems
Variable speed drive
This is an open access article under the CC BY-SA license.
Corresponding Author:
Rogger José Andrade-Cedeno
Investigador asociado al Posgrado en Instrumentación, Facultad de Ciencias, Universidad Central de
Venezuela (UCV)
Caracas 1040, Distrito Capital, Venezuela
Email: rogger.andrade@gmail.com
1. INTRODUCTION
Pressure and flow control systems are widely used in the pumping, transportation, distribution and
handling of fluids. They are found in crude oil transportation systems [1], in aqueducts for pumping and
water distribution [2], and in pressurized irrigation systems [3], among others. It is common in these systems
to find centrifugal pumps, induction motors, variable frequency drives, control valves, electronic

 ISSN: 2088-8694
Int J Pow Elec & Dri Syst, Vol. 13, No. 3, September 2022: 1365-1379
1366
instrumentation and controllers, since the combination of several of them allows the implementation of an
automatic pressure or flow control system according to the needs of each application, and a laboratory
experimental application can be seen in [4]. In the case of pumping systems, there are two methods to
regulate the capacity: i) by throttling with a control valve and ii) by varying the speed of the electric pump
with a variable speed drive, the second method being the most efficient [5]. However, this does not mean that
the control valve has lagged behind; on the contrary, it is still considered the most used final control element
in the process control industry [6]. Gevorkov et al. [7] consider that throttling control is still widely used in
industry, and simulated an electric actuator coupled to a control valve for a pumping application, the results
of which determined that the pressure could be regulated by precise position control achieved by the electric
actuator. In the same way, Vijayalakshmi et al. [8], they simulated and implemented a closed loop flow
control using a control valve with proportional integral derivative (PID) algorithm. On the other hand,
Andrade-Cedeno and Perez-Rodriguez [9] simulated a pulse-width modulation voltage-source inverter (VSI-
PWM) type variable speed drive, governed by a scalar control strategy with a constant V/f ratio combined
with the space vector modulation (SVM) technique. The analysis determined that this technology performs
well for applications with induction motors coupled to centrifugal machines such as pumps, compressors and
fans. Several investigations on automatic control in pumping systems have been developed. Shankar, et al.
[10] simulated two centrifugal pumps in parallel at variable speed and studied the pressure and flow response
for various speeds. Furthermore, they determined that a 50% reduction in speed produces an 80% reduction
in power. Gevorkov et al. [11] simulated a constant pressure system on a variable speed centrifugal pump,
adding a pressure sensor for feedback and an automatic proportional integral (PI) controller. The simulated
model was experimentally validated, finding differences in steady state from 7 to 14%. Gevorkov et al. [12]
carried out a mixed simulation and experimentation study of a flow control system with a variable speed
centrifugal pump, using a flow sensor for feedback and a PI controller. The inaccuracy of the proposed model
varied between 0.3 and 4%. In the research by Wu et al. [13], they compared the use of a variable speed drive
(speed regulation) versus the use of a control valve (discharge throttling), and it was determined that the
speed control method requires 37% less power than the discharge throttling method.
This research describes the models of the typical components of a water pumping and distribution
system, made up of: the piping system, centrifugal pump, induction motor, variable speed drive, control
valve, and control system. Using Simscape library from MATLAB/Simulink R2019b, two case studies are
simulated to achieve automatic flow control: i) with a control valve and ii) with variable speed drive. The
dynamic response of the system, the starting current, the harmonic distortion and the power saving are
analyzed.
2. METHOD
2.1. Capacity regulation in centrifugal pumps
Figure 1 shows the characteristic curves of a centrifugal pump with the operating points for different
capacity control methods. In nominal conditions, the pump must operate at the best efficiency point (BEP),
which is very suitable in systems with fixed demand that are well designed and dimensioned. If the demand
changes, the system curve also changes and it operates at a different point of the BEP.
Figure 1. Methods for capacity regulation in centrifugal pumps
Operating a centrifugal pump too far from the BEP, either to the right or to the left of the curve, can
put the equipment at risk from adverse effects such as cavitation, vibration, impeller damage, suction or
discharge recirculation, and reduced life of seals and bearings [14]. To prevent these problems, the

Int J Pow Elec & Dri Syst ISSN: 2088-8694 
Comparative study by numerical simulation of two methods for automatic … (Rogger José Andrade-Cedeno)
1367
centrifugal pump must be operated in the preferred operating region (POR), which is between 70 and 120%
of BEP [15]. When regulating the output of the pump with a control valve, the valve changes the dynamic
head of the system, causing the system curve to intersect with the pump curve at the new operating point.
While on the one hand the throttling reduces the flow rate, on the other hand the pump is forced to produce a
higher dynamic head (higher pressure). When using variable speed control the system curve remains
unchanged, while the pump curve moves down according to the speed change, and the operating point moves
to a new intersection between the two curves. The benefit of controlling the speed of the pump is that the
efficiency drops much less compared to the use of a control valve [16].
2.2. Study cases
Figure 2 shows the two study cases to be simulated. Case 1 is shown in Figure 2(a), where a control
valve with an electric actuator is used on the pump discharge for automatic flow control. Case 2 can be seen
in Figure 2(b), and in this case a variable frequency drive coupled to the electric motor is used. The models
are obtained from the Simscape library of Simulink, in version R2019b of MATLAB. These tools are
specialized in the modeling and simulation of physical, mechanical, hydraulic, electrical, electromechanical,
mechatronic, and automatic control systems.
(a)
(b)
Figure 2. Study cases for (a) flow control system based on control valve with electric actuator and (b) flow
control system based on variable speed drive on the pump
2.3. Piping system
For the pipe system model, the pressure loss due to friction is calculated using the Darcy equation.
The friction factor in the turbulent regime is determined with the Haaland approximation. The friction factor
during the transition from laminar to turbulent regimes is determined by linear interpolation between the
extreme points of the two regimes [17].

 ISSN: 2088-8694
1368
2.4. Pumping system
Figure 3 shows the model of the pumping system, made up of the centrifugal pump as the main
component. For centrifugal pump characterization, the performance curves provided by the manufacturer are
used to form two one-dimensional search tables: i) differential pressure vs flow rate and ii) brake power vs
flow rate. Both characteristics are specified at the same reference angular velocity, and at the same fluid
density. To obtain the differential pressure at another angular velocity, the laws of affinity are used [18].
Figure 3. Pumping system model
2.5. Control valve with electric actuator
Figure 4 shows the integrated model of the control valve with electric actuator. The control valve is
modeled using the flow coefficient provided by the manufacturer [19]. The driving source of the actuator is a
DC motor [20]. The mechanical output of the DC motor enters the ideal, non-planetary, fixed gear ratio
gearbox [21]. The gear ratio is determined as the ratio of the angular velocity of the input shaft to that of the
output shaft. The gearbox output connects to the rack and pinion block, which converts rotational to
translational motion [22]. The DC motor drive has a control component and a power component. The control
component consists of a pulse-width modulation (PWM) system, which receives the signal from the main
control system to adjust the trigger signals to the power component. The power component consists of power
semiconductors (usually BJTs, MOSFETs, or IGBTs) in an H-bridge connection scheme.
Figure 4. Model of control valve with electric actuator

1369
The PWM system modulates the switching of the power semiconductors, thus controlling the
polarity and amplitude of the voltage fed to the DC motor, and therefore its speed and rotational direction
[23]. The main control system for the linear electric actuator consists of a three-loop cascade control. The
innermost loop is a PI controller of the motor current, whose feedback signal is achieved by a current sensor,
and directly modifies the PWM signal going into the H-bridge. The middle loop consists of a PI controller of
the motor speed, which is fed back from a speed sensor. The outermost loop is for linear position control of
the system, and receives feedback from a linear position sensor. For fast response and safe operation, each
control system loop must be precisely designed [24], [25].
2.6. Induction motor
The induction motor is based on the asynchronous machine model. The stator and rotor windings are
Y-connected with internal neutral. A sixth-order state space model for the double-cage machine and a fourth-
order state space model for the single-cage machine represent the electrical component. The mechanical
component is modeled as a second order system. All electrical variables and parameters are referred to the
stator; furthermore, the stator and rotor quantities are in an arbitrary two-axis frame of reference [26]. The
two-axis reference frame, known as the dq reference frame, is the result of applying the Clarke and Park
transforms [27]. The induction motor is used in two schemes: with direct connection to the grid (frequency
and nominal speed, Figure 5(a)), and as a part of a variable speed drive (frequency and variable speed,
Figure 5(b)).
(a)
(b)
Figure 5. Three-phase induction motor models with their drives and meters for (a) direct connection to the
grid and (b) variable speed drive system

 ISSN: 2088-8694
1370
2.7. Variable speed drive
The variable speed drive model is based on the block drive AC2 “Space vector PWM VSI induction
motor drive”, which contains an asynchronous machine controlled by a variable frequency drive [28]. The
asynchronous machine is configured as a three-phase induction motor, and its variable frequency drive
contains the following components: the three-phase rectifier formed by diodes; the DC link formed by a
capacitor; the three-phase inverter formed by IGBTs; and the control system formed by a speed controller
and a space vector modulator (SVM) [29], as shown in Figure 6(a).
The control system has a closed-loop speed control, made up of the classic V/f scalar control plus a
slip compensator based on a PI algorithm [30], as seen in Figure 6(b). The speed control loop receives the
speed reference (N*
) and subtracts it from the current speed (N) measured with the speed sensor, to obtain the
motor slip (speed error). Then, the PI algorithm computes and compensates for the slip in the current speed
(N). The PI compensator parameters, Kp and Ki, were tuned using the Ziegler-Nichols method.
(a)
(b)
Figure 6. AC2 drive model block diagrams for (a) general block diagram and (b) block diagram of
control system
The frequency reference (Freq*
), and the voltage reference (Volts*
), are obtained from the constant
V/f ratio block, and are used as inputs to the space vector modulator (SVM) stage, as seen in Figure 6(b). The
SVM technique starts with a three-phase signal generator, whose signals are then converted to the αβ
reference frame using the Clark transform block. This is necessary to run the SVM algorithm, which
produces the controlling signals of IGBTs. SVM has better performance than other PWM techniques because
of better utilization of the DC bus and least harmonic distortion production [31].
2.8. Automatic control system
Figure 7 shows the block diagram of the automatic flow control system, when a control valve with
an electric actuator is used (Figure 7(a)), and when a variable speed drive is used (Figure 7(b)). In both cases,
the main control loop is executed with the PI flow controller, which commands the controls embedded in the
electric actuator or in the variable speed drive (described above). Feedback is achieved with a flow
transmitter (flowmeter).

1371
(a)
(b)
Figure 7. Block diagram of the automatic flow control system for (a) based on control valve and electric
actuator and (b) based on variable speed drive
3. RESULTS AND DISCUSSION
For simulation, a 110 gpm @ 32 m centrifugal pump was used, with a 5 hp, 3 ph, 220 V, 60 Hz,
3450 rpm induction motor, in addition to a globe type control valve, equal percentage, 2 inches and Cv of
33.2.
3.1. Dynamic response of the control system
Figure 8 shows the dynamic response of several variables of interest, for the two cases studied. Both
cases met the objective of maintaining constant flow at a set point of 70 gpm, in the face of pressure
disturbances produced by the disturbance valve. The control system specifications, such as the response time,
overshoot and position error, will depend on the requirements and characteristics of each process to be
controlled.
In case 1 (Figure 8(a)), with the pump at nominal speed and the control valve in discharge, the
mechanical and hydraulic variables of the pump remained more or less constant during the steady state, with
transients produced by pressure changes. The pressure drop across the control valve resulted from the
difference between the pump discharge pressure and the system pressure, and increased when the control
valve had a lower position. This pressure drop represents energy losses for the system [13]. Also observed
was the dynamic response of the three variables from the cascade control of the actuator: current (internal
loop), speed (medium loop), and position (external loop). The response time was lower the more internal the
control loop was, which is required for the adequate performance of the cascade control [24].
In case 2 (Figure 8(b)), there was no restrictive element in the pump discharge, and the speed was
automatically regulated to meet the flow requirement in the presence of pressure disturbances. The speed, the
total dynamic head and the torque increased or decreased as the pressure disturbance did. This action allows
power saving compared to case 1. In addition, the slip remained at zero thanks to the slip compensator part of
the control system of the variable speed drive. This helps improve the operating performance of the induction
motor [32].

 ISSN: 2088-8694
1372
(a)
(b)
Figure 8. Dynamic response of several variables of interest for (a) case 1: fixed speed pump with control
valve with electric actuator and (b) case 2: variable speed pump with the use of variable speed drive
3.2. Operational efficiency, consumption and power savings
Table 1 shows a register of the various hydraulic and mechanical variables of interest. With the flow
and pressure data the hydraulic power was calculated, with the torque and the angular speed the mechanical
power in the shaft was obtained, and with these results the efficiency of the pump was calculated. In all
scenarios, it was possible to meet the set point, but case 2 did so with greater variability. Reducing the set
point reduced the efficiency of the pump, more significantly in case 1. With respect to mechanical power
consumption, case 2 consumed less power than case 1, and this difference increased when reducing the set
point. With this, a saving in mechanical power of case 2 was obtained with respect to case 1, from 24 to 64%.
In this analysis, the effect of the variable speed drive was not considered, given that under nominal
conditions electric motors are more efficient than pumps, and variable speed drives more efficient than
motors. However, the efficiency of the motor can change with variations in load torque and speed [33]-[35].
Burt et al. [35] analyzed that the efficiency of the electrical system with variable speed drive can be reduced
by up to 8% compared to the system with direct connection to the grid. However, the study concluded that
this reduction is not significant in relation to the energy savings achieved with a variable speed drive, since
the drive adjusts the speed of the process to the changing demand conditions.

1373
Table 1. Operational data recording
Set
point
Flow rate Pressure
Angular
speed
Torque
Pump
efficiency
Shaft
power Power
saving
Case
1
Case
2
Case
1
Case
2
Case
1
Case
2
Case
1
Case
2
Case
1
Case
2
Case
1
Case
2
gpm gpm gpm psi psi rpm rpm Nm Nm % % kW kW %
100 100 100.00 35 35.01 3440 3080 8.67 7.33 73.81 74.55 3.12 2.36 24.34
100 100 100.00 30 30.01 3440 2924 8.67 6.78 73.81 74.41 3.12 2.08 33.55
100 100 100.10 25 25.00 3440 2758 8.67 6.21 73.81 74.03 3.12 1.79 42.58
70 70 70.00 35 34.99 3459 2875 7.54 5.61 65.90 70.67 2.73 1.69 38.22
70 70 70.01 30 30.00 3459 2707 7.54 5.10 65.90 72.00 2.73 1.45 47.08
70 70 69.99 25 25.00 3459 2528 7.54 4.58 65.90 73.24 2.73 1.21 55.62
40 40 40.07 35 35.03 3478 2742 6.40 4.25 46.66 55.03 2.33 1.22 47.66
40 40 39.93 30 29.96 3478 2562 6.40 3.78 46.66 57.19 2.33 1.01 56.43
40 40 40.11 25 25.04 3478 2376 6.40 3.34 46.66 59.87 2.33 0.83 64.35
3.3. Starting current and harmonic distortion
Figure 9 shows the waveforms of the currents and voltages of the induction motor during the
starting process and in the stable state. Figure 9(a) shows the direct on-line starter (DOL) through direct
connection to the grid (case 1), where the transient lasted approximately 250 ms (15 cycles). It is seen that
the starting current was about 7.4 times the steady state current. On the other hand, Figure 9(b) shows the
starting process through an acceleration ramp of 3600 rpm/s, provided by the variable speed drive (case 2). In
this second case, the starting current did not exceed 2.8 times the steady-state current. When a pump operates
at variable speed (case 2), the waveforms of the voltages and currents, on both the drive and the motor side,
are as shown in Figure 9(c). The total harmonic distortion (THD) was obtained using the fast fourier
transform (FFT), a tool of the specialized “powergui” block of Simulink. On the line side, a THDi of 82.69%
and THDv of 4.31% were obtained, and on the motor side, a THDi of 20.55% and THDv of 48.42% were
obtained. The variable speed drive (case 2) affects the quality of the energy on the motor and grid side.
3.4. Experimental results and model validation
Figure 10 shows case 2 experimental setup. The module is made up of a Gould 3656 centrifugal
pump with a Baldor JMM3212T induction motor, electrical panels, control panel, and instrumentation. The
control panel contains an ABB ACS310 variable frequency drive (VFD), and an XT sensor TUF-2000M
ultrasonic flowmeter. The flow transducers, pressure transducer, and pressure gauge were installed in the
discharge pipe. The flow transmitter and pressure transducer were connected with the analog inputs of the
VFD, using 4-20 mA signals. The control algorithms (PI flow controller, PI slip compensator, V/f scalar
control, and space vector modulation), were configured on the VFD. For data collection, a Dell Inspiron PC
was used, with an Intel Core i7 processor and Windows 7 operating system, interconnected with the VFD and
the Hantek 6022BL digital oscilloscope. The specifications of the centrifugal pump, at nominal conditions,
are: capacity=110 GPM, total dynamic head=32 m, speed=3500 rpm, efficiency=75%. The selected three-
phase induction motor has the following specifications at rated conditions: output power=5 hp (3.73 kW),
voltage=208-230/460 V, current=13-12.2/6.1 A, speed=3450 rpm, frequency=60 Hz, power factor=89%,
efficiency=85.5%. The ultrasonic flowmeter has the following specifications: accuracy=±1%, temperature=-
30~160 °C, flowrate=0~±7 m/s, damping=10 s, output signal=4-20 mA, power supply=8 a 36 Vdc.
Figure 11(a) shows the pressure, flow, speed, and torque trend graphs during the experimental test,
data obtained from the VFD using ABB's DriveWindow Light software. Since the software only graphs in
percentages, all four variables were converted to percentages based on the range of the measurement. For the
test, a set point of 50 GPM (46.67%) was set, and the manual valve in the pump discharge was manipulated
to generate up and down pressure disturbances. As can be seen in Figure 11(a), the control system fulfills its
objective of keeping the flow rate constant at the set point, in the face of pressure disturbances. In the
transient state, the overshoot was less than 10% and the settling time was less than 20s, mainly influenced by
the flowmeter time constant (10s) and PI flow controller parameters.
Figure 11(b) shows the pressure, flow, speed, and torque trend graphs resulting from the simulation,
observing the similarity with the experimental results. Simulink models parameterization was done using
equipment plate data and manufacturers' technical specifications. The parameterization of the centrifugal
pump was carried out with the performance curves obtained from the catalog. In the case of the induction
motor, the parameterization was done by entering the nameplate data into the optimizer provided by the
asynchronous machine block in Simulink. For the simulated flowmeter, a transfer function of a first-order
system with a time constant of 10 s was considered, to be consistent with the damping of the real-world
flowmeter. Table 2 shows the relative error of the simulation concerning the experimentation for the steady-
state condition, observing that, in none of the cases, the error exceeds 6%.

 ISSN: 2088-8694
1374
(a)
(b)
(c)
Figure 9. Currents and voltages at startup and steady state for (a) direct start, (b) soft start by means of the
variable speed drive, at 3600 rpm/s, and (c) steady state signals when operating with variable speed drive
Figure 10. Experimental setup of flow control system with variable speed drive (case 2)

1375
(a)
(b)
Figure 11. Model validation for (a) experimental results and (b) simulation results
Table 2. Hydraulic and mechanical variables validation
Simulation Experimentation Relative error
Pressure Flow Speed Torque Freq. Pressure Flow Speed Torque Freq. Pressure Flow Speed Torque Freq.
% % % % Hz % % % % Hz % % % % %
48.07 46.67 92.8 67 58.05 48.5 46.9 96.2 63.7 59.4 0.89 0.49 3.53 5.18 2.27
22.25 46.67 44.24 20.33 27.26 21.9 46.9 41.8 21.1 26.5 1.6 0.49 5.84 3.65 2.87
33.35 46.67 69.27 41.22 42.91 32.8 46.8 70.42 41 44.3 1.68 0.28 1.63 0.54 3.14
Figure 12 shows the electrical variables obtained with the oscilloscope during the experimental tests.
What was predicted in the simulation concerning the starting current is corroborated, being more abrupt in
the direct starting (Figure 12(a)), and softened when the acceleration ramp of 1s is configured in the VFD
(Figure 12(b)). Therefore, the use of a variable speed drive limited the starting current by 62%, which is
advantageous to prevent some hydraulic, mechanical, and electrical problems, such as water hammer and
pressure surges [36], vibrations in the pipes [37], voltage drops in the motor and the electrical grid [38], and
mechanical stress that reduces the useful life of the motor [39].
The line-side current has high harmonic distortion (Figure 12(c)), and a non-symmetrical waveform
is observed, unlike what was obtained in the simulation. This asymmetry is due to the voltage unbalance in
the power supply to the VFD [40]. The motor-side current is closer to the sine wave (Figure 12(d)), with the
characteristic ripple produced by semiconductor switching. The line-side voltage has the lowest harmonic
distortion among all the analyzed waves (Figure 12(e)), corroborating what was obtained in the simulation.
The voltage transients on the motor-side, produced during the switching of the power semiconductors, are not
observed in the simulation but are appreciated in the experimentation (Figure 12(f)). This leads to the
phenomenon of the reflected wave, which is amplified if the distance between the VFD and the induction
motor increases, generating damaging overvoltages for the conductors and the induction motor, and
damaging overcurrents for the VFD [41], [42]. The motor-side voltage also has a high harmonic distortion as

 ISSN: 2088-8694
1376
a consecuence of the PWM control signal. Harmonics affect the power factor and efficiency of the induction
motor and can also affect the behavior, reliability, and useful life of some of its mechanical components [43].
Harmonic pollution also affects the electrical grid and equipment connected to it, causing overload and
overheating in transformers, as well as unwanted tripping of electrical protections, affecting lighting systems
and causing interference in electronic equipment and communication systems [44]. Harmonics can be treated
by implementing filters [45]. The validation of the electrical variables is shown in Table 3, where it is
observed that the relative error of the simulation results does not exceed 10% concerning the real
experimental data.
(a) (b)
(c) (d)
(e) (f)
Figure 12. Oscilloscope waveforms for (a) direct start current, (b) soft start current (acceleration ramp set
in 1 s), (c) line side current at 60 Hz, (d) motor side current at 60 Hz, (e) line side voltage at 60 Hz, and
(f) motor side voltage at 60 Hz
Table 3. Electrical variables validation
Simulation Experimentation Relative error
Line
Current
Motor
Current
Line
Voltage
Motor
Voltage
Freq.
Line
Current
Motor
Current
Line
Voltage
Motor
Voltage
Freq.
Line
Current
Motor
Current
Line
Voltage
Motor
Voltage
Freq.
A A V V Hz A A V V Hz % % % % %
12.3 10.9 220 244 60 11.2 9.9 222 246 59.8 9.82 9.8 0.9 0.73 0.33
Note: RMS values at steady state condition

1377
4. CONCLUSION
In this research, the models of some hydraulic, electromechanical, and mechatronic components
belonging to the Simscape libraries in Simulink were studied. These models were integrated to form automatic
water flow control systems, based on two case studies: with a control valve (case 1) and with a variable speed
drive (case 2). In both cases, the goal of maintaining a constant flow in the event of pressure disturbances
generated in the system was met. These case studies do not correspond to a specific process or industry, where
there may be specific requirements for the performance of the control system. In those scenarios, a suitable
control must be selected and the control system must be carefully designed and configured. Case 2 has some
advantages over case 1: at low flow rates, it is more energy-efficient and has less impact on current and torque
transients due to the acceleration and deceleration ramps configured in the variable speed drive. A disadvantage
of case 2 is the production of harmonics on both the motor and grid sides, which can be reduced with the
implementation of harmonic filters. Case 2 model was validated through experimentation, obtaining errors of
less than 6% for the hydraulic and mechanical variables, and less than 10% for the electrical ones.
REFERENCES
[1] E. B. Priyanka, K. Krishnamurthy, and C. Maheswari, “Remote monitoring and control of pressure and flow in oil pipelines
transport system using PLC based controller,” in 2016 Online International Conference on Green Engineering and Technologies
(IC-GET), Nov. 2016, pp. 1-6, doi: 10.1109/GET.2016.7916754.
[2] R. Andrade-Cedeno, “Energy management of a pumping station through the use of statistical process control. case study: “la
esperanza” aqueduct - pacific refinery (in Spanish: Gestión energética de una estación de bombeo mediante el uso del control
estadístico de procesos. estudio de caso: acueducto "la esperanza" - refinería del pacífico),” Revista Politécnica, vol. 40, no. 2,
pp. 7-18, 2018.
[3] M. D. Azari, A. P. Rizi, and A. Ashrafzadeh, “Hydraulic design and operation of variable-speed pumps as the water–energy
saving strategies in pressurized irrigation systems,” Clean Technologies and Environmental Policy, vol. 23, no. 5, pp. 1493-1508,
2021, doi: 10.1007/s10098-021-02043-w.
[4] R. J. Andrade-Cedeno, “Teaching module for controlling water level and flow, by means of SCADA system, PLC and PID
algorithm (in Spanish: Módulo didáctico para controlar nivel y caudal de agua, mediante sistema SCADA, PLC y algoritmo
PID),” RIEMAT, vol. 4, no. 2, pp. 50-62, 2019, doi: 10.33936/riemat.v4i2.2196.
[5] V. K. A. Shankar, S. Umashankar, S. Paramasivam, and N. Hanigovszki, “A comprehensive review on energy efficiency
enhancement initiatives in centrifugal pumping system,” Applied Energy, vol. 181, pp. 495-513, 2016, doi:
10.1016/j.apenergy.2016.08.070.
[6] M. Paun, Control valve handbook. 5 ed. Marshalltown, USA: Fisher Controls International LLC, 2019.
[7] L. Gevorkov, A. Rassõlkin, A. Kallaste, and T. Vaimann, “Simulink based model of electric drive for throttle valve in pumping
application,” in 2018 19th International Scientific Conference on Electric Power Engineering (EPE), 2018, pp. 1-4, doi:
10.1109/EPE.2018.8395996.
[8] S. Vijayalakshmi, C. Anuradha, R. C. Ilambirai, and V. Ganesh, “Real-time monitoring and control of flow rate in transportation
pipelines using matlab-based interactive GUI and PID controller,” International Journal of Power Electronics and Drive System
(IJPEDS), vol. 11, no. 4, pp. 1767-1774, 2020, doi: 10.11591/ijpeds.v11.i4.pp1767-1774.
[9] R. J. Andrade-Cedeno and J. A. Perez-Rodriguez, “Analysis of V/f control with SVM in a variable speed drive (in Spanish:
Análisis del control V/f con SVM en un accionamiento de velocidad variable),” Dominios de la Ciencias, vol. 7, no. 6, pp. 38-62,
2021.
[10] V. K. A. Shankar, S. Umashankar, S. Paramasivam, and H. Norbert, “Real time simulation of variable speed parallel pumping
system,” Energy Procedia, vol. 142, pp. 2102-2108, 2017, doi: 10.1016/j.egypro.2017.12.612.
[11] L. Gevorkov, V. Vodovozov, and Z. Raud, “Simulation study of the pressure control system for a centrifugal pump,” in 2016 57th
International Scientific Conference on Power and Electrical Engineering of Riga Technical University (RTUCON), 2016, pp. 1-5,
doi: 10.1109/RTUCON.2016.7763086.
[12] L. Gevorkov, A. Rassõlkin, A. Kallaste, and T. Vaimann, “Simulink based model for flow control of a centrifugal pumping
system,” in 2018 25th International Workshop on Electric Drives: Optimization in Control of Electric Drives (IWED), 2018, pp.
1-4, doi: 10.1109/IWED.2018.8321399.
[13] Y. Wu, X. Guo, G. Li, and C. Lu, “Analysis of shaft power of centrifugal pump under variable speed condition,” in 2017 12th
IEEE Conference on Industrial Electronics and Applications (ICIEA), 2017, pp. 248-252, doi: 10.1109/ICIEA.2017.8282851.
[14] I. J. Karassik, J. P. Messina, P. Cooper, and C. C. Heald, Pump Handbook. 3ed. New York, USA: McGraw-Hill, 2001.
[15] B. Kelechava, “Rotodynamic (centrifugal and vertical) pumps - guideline for allowable operating region,” American National
Standard Institute (ANSI), 2017. [Online]. Available: https://blog.ansi.org/2017/07/ansihi-963-2017-rotodynamic-pumps-aor-bep/
[16] J. Viholainen, “Energy-efficient control strategies for variable speed driven parallel pumping systems based on pump operation
point monitoring with frequency converters,” M.S. thesis, Institute of Energy Technology, Lappeenranta University of
Technology, Lappeenranta, Finlandia, 2014.
[17] MathWorks, “Hydraulic resistive tube,” 2022. [Online]. Available:
https://la.mathworks.com/help/physmod/simscape/ref/hydraulicresistivetube.html [Accessed Jan. 12, 2022]
[18] MathWorks, “Centrifugal pump with choice of parameterization options,”. [Online]. Available: https://la.mathworks.com
/help/physmod/hydro/ref/centrifugalpump.html [Accessed Jan. 12, 2022].
[19] MathWorks, “Flow coefficient parameterized valve (TL),” 2022. [Online]. Available:
https://la.mathworks.com/help/physmod/hydro/ref/
flowcoefficientparameterizedvalvetl.html [Accessed Jan. 12, 2022].
[20] MathWorks, “DC motor model with electrical and torque characteristics and fault modeling,” 2022. [Online]. Available:
https://la.math
works.com/help/physmod/sps/ref/dcmotor.html [Accessed Jan. 12, 2022].
[21] MathWorks, “Gear box in mechanical systems,” 2022. [Online]. Available:
https://la.mathworks.com/help/physmod/simscape/ref/gearbox.html [Accessed Jan. 12, 2022].

 ISSN: 2088-8694
1378
[22] MathWorks, “Rack and pinion gear coupling translational and rotational motion, with adjustable pinion radius and friction
losses,”. [Online]. Available: https://la.mathworks.com/help/physmod/sdl/ref/rackpinion.html [Accessed Jan. 12, 2022].
[23] MathWorks, “PWM-controlled DC motor,”. [Online]. Available: https://la.mathworks.com/help/physmod/sps/ug/pwm-controlled-
dc-motor.html [Accessed Jan. 12, 2022].
[24] L. Jiang, L. Liu, X. Peng, and Z. Xu, “Design and analysis of a fully variable valve actuation system,” Energies, vol. 13, no. 23,
2020, doi: 10.3390/en13236391.
[25] G. Tosun, O. C. Kivanc, E. Oguz, Y. Mutlu, and O. Ustun, “Design of a position controlled electric actuator used in fluid control
valves,” in 2016 IEEE International Power Electronics and Motion Control Conference (PEMC), 2016, pp. 551-556, doi:
10.1109/EPEPEMC.2016.7752055.
[26] MathWorks, “Model dynamics of three-phase asynchronous machine, also known as induction machine, in SI or pu units,”.
[Online]. Available: https://la.mathworks.com/help/physmod/sps/powersys/ref/asynchronousmachine.html [Accessed Jan. 12,
2022].
[27] MathWorks, “Implement Clarke and Park transforms for motor control (in Spanish: Implemente transformadas de Clarke y Park
para el control de motores),”. [Online]. Available: https://la.mathworks.com/solutions/power-electronics-control/clarke-and-park-
transforms.html [Accessed Jan. 12, 2022].
[28] MathWorks, “Simulate an AC motor drive,”. [Online]. Available: https://la.mathworks.com/help/physmod/sps/powersys/
ug/simulating-an-ac-motor-drive.html [Accessed Jan. 12, 2022].
[29] MathWorks, “Implement universal power converter with selectable topologies and power electronic devices,”. [Online].
Avaolable: https://la.mathworks.com/help/physmod/sps/powersys/ref/universalbridge.html [Accessed Jan. 12, 2022].
[30] MathWorks, “Implement space vector PWM VSI induction motor drive,”. [Online]. Available: https://la.mathworks.com
/help/physmod/sps/powersys/ref/spacevectorpwmvsiinductionmotordrive.html [Accessed Jan. 12, 2022].
[31] S. Ojha and A. K. Pandey, “Close loop V/F control of voltage source inverter using sinusoidal PWM, third harmonic injection
PWM and space vector PWM method for induction motor,” International Journal of Power Electronics and Drive System
(IJPEDS), vol. 7, no.1, pp. 217-224, 2016.
[32] H. Sarhan and R. Issa, “Selection of slip compensation technique in three-phase induction motors drive system,” International
Journal of Emerging Technology and Advanced Engineering, vol. 9, no. 12, pp. 9-15, 2019.
[33] Y. Lozanov, S. Tzvetkova, and A. Petleshkov, “Study of the effectiveness of a variable frequency drive of an induction motor,” in
2019 11th Electrical Engineering Faculty Conference (BulEF), 2019, pp. 1-6, doi: 10.1109/BulEF48056.2019.9030775.
[34] E. Agamloh, A. Cavagnino, and S. Vaschetto, “Induction machine efficiency at variable frequencies,” in 2019 IEEE Energy
Conversion Congress and Exposition (ECCE), 2019, pp. 1655-1662, doi: 10.1109/ECCE.2019.8912969.
[35] C. Burt, X. Piao, F. Gaudi, B. Busch, and N. F. Taufik, “Electric motor efficiency under variable frequencies and loads,” Journal
of Irrigation and Drainage Engineering-ASCE, vol. 134, pp. 129-136, 2008, doi: 10.1061/(ASCE)0733-9437(2008)134:2(129).
[36] S. Greenwood, “Soft starter benefits in pump control,” World Pumps, vol. 2015, no. 2, pp. 24-27, 2015, doi: 10.1016/S0262-
1762(15)70025-8.
[37] D. Lale, M. Oršulić, and I. Palunko, “Modelling and soft-start control of measurement and transport line,” 2020 International
Conference on Smart Systems and Technologies (SST), 2020, pp. 59-64, doi: 10.1109/SST49455.2020.9264077.
[38] N. Matanov, “Study of the impact of induction motors starting on the supply voltage,” 2019 16th Conference on Electrical
Machines, Drives and Power Systems (ELMA), 2019, pp. 1-5, doi: 10.1109/ELMA.2019.8771585.
[39] M. Habyarimana and D. G. Dorrell, “Methods to reduce the starting current of an induction motor,” 2017 IEEE International
Conference on Power, Control, Signals and Instrumentation Engineering (ICPCSI), 2017, pp. 34-38, doi:
10.1109/ICPCSI.2017.8392319.
[40] E. B. Agamloh, S. Peele, and J. Grappe, “Operation of variable frequency drive motor systems with source voltage unbalance,”
2017 Annual Pulp, Paper And Forest Industries Technical Conference (PPFIC), 2017, pp. 1-9, doi: 10.1109/PPIC.2017.8003869.
[41] B. Narayanasamy, A. S. Sathyanarayanan, A. Deshpande, and F. Luo, “Impact of cable and motor loads on wide bandgap device
switching and reflected wave phenomenon in motor drives,” 2017 IEEE Applied Power Electronics Conference and Exposition
(APEC), 2017, pp. 931-937, doi: 10.1109/APEC.2017.7930808.
[42] B. Narayanasamy, A. S. Sathyanarayanan, F. Luo, and C. Chen, “Reflected wave phenomenon in SiC motor drives:
consequences, boundaries, and mitigation,” IEEE Transactions on Power Electronics, vol. 35, no. 10, pp. 10629-10642, 2020,
doi: 10.1109/TPEL.2020.2975217.
[43] H. G. Beleiu, V. Maier, S. G. Pavel, I. Birou, C. S. Pică, and P. C. Dărab, “Harmonics consequences on drive systems with
induction motor,” Applied Sciences, vol. 10, no. 4, 2020, doi: 10.3390/app10041528.
[44] J. P. Srividhya, D. Sivakumar, and T. Shanmathi, “A review on causes, effects, and detection techniques of harmonics in the
power system,” 2016 International Conference on Computation of Power, Energy Information and Commuincation (ICCPEIC),
2016, pp. 680-686, doi: 10.1109/ICCPEIC.2016.7557309.
[45] A. Kalair, N. Abas, A. R. Kalair, Z. Saleem, and N. Khan, “Review of harmonic analysis, modeling and mitigation techniques,”
Renewable and Sustainable Energy Reviews, vol. 78, pp. 1152-1187, 2017, doi: 10.1016/j.rser.2017.04.121.
BIOGRAPHIES OF AUTHORS
Rogger José Andrade-Cedeno he was born in Chone, Ecuador, in 1985. He
received the Bachelor’s Degree in Electronics and Control Engineering from the National
Polytechnic School in 2010. He received the Master’s Degree in Quality Management Systems
from the Private Technical University of Loja in 2017, and the Master’s Degree in Electrical
Engineering (mention: Power Systems) from the Technical University of Manabí in 2022. He is
currently an assosiate professor at the Technical and Technological Training Academic Unit,
ULEAM. Mr. Andrade has collaborated on different projects and operations related to the oil
industry, pumping stations, pipelines and aqueducts, and water distribution systems. His
research interests are in energy and power systems, electric machines and drives, automation
and control systems. He can be contacted at email: rogger.andrade@uleam.edu.ec.

1379
Jesús Alberto Pérez-Rodríguez he was born in Cagua, Venezuela in 1963. He
graduated in Electrical Engineering at the University of Carabobo, Valencia, Venezuela in
1988, and M.Sc. and Ph.D. in Sciences (mention: Instrumentation) at the Central University of
Venezuela, Caracas, Venezuela, 2002 respectively. His works were developed in the field of
robotics and automation, in addition to works on the teaching of Engineering. He has
participated in numerous projects aimed at energy efficiency. He can be contacted at email:
jesus.perez@utm.edu.ec.
Carlos David Amaya-Jaramillo full-time professor in the Department of
Electricity of the Quevedo State Technical University (UTEQ), Quevedo, Ecuador. He
graduated in Electronic and Control Engineering in 2010 at the National Polytechnic School,
Quito, Ecuador. He defended his doctorate in the Polytechnic University of Madrid in 2018. He
can be contacted at email: camaya@uteq.edu.ec.
Ciaddy Gina Rodríguez-Borges she was born in Puerto Ordaz, Venezuela in
1970. She graduated in Industrial Engineering, Master in Business Management (mention:
Industrial Management), from the University of Zulia, Venezuela 1999, and Doctor in
Technical Sciences, from the Technological University of Havana, Cuba 2012. She has been a
professor at the University of Zulia and is currently a Principal Professor at the Technical
University of Manabí in the Department of Industrial Engineering. She has been a worker and
researcher for the electricity company in Venezuela. Her research work has addressed issues
associated with energy efficiency, the use of hybrid energy systems and in areas related to the
teaching of engineering. She can be contacted at email: ciaddy.rodriguez@utm.edu.ec.
Endrickson Ramón Vera-Cedeno He was born in Chone, Ecuador in 1969. He
graduated in Electrical Engineering in 2013 at the Eloy Alfaro Secular University of Manabí
(ULEAM), and is currently an associate professor in the Academic Unit of Technical and
Technological Training of the same institution. He is a specialist in the maintenance of
electrical systems. He can be contacted at email: endrickson15@gmail.com.
Luis Santiago Quiroz-Fernándes Civil Engineer, Ph.D. in Technical Sciences,
current Rector-elect of the Manabí Technical University, Former Director of the Postgraduate
Institute of UTM, specialist in the area of Hydraulic Engineering and outstanding manager of
public institutions and as a professor of the Civil Engineering career of the UTM. He can be
contacted at email: santiago.quiroz@utm.edu.ec.
Yolanda Eugenia Llosas-Albuerne full-time principal professor in the
Department of Electricity of the Technical University of Manabí (UTM), Portoviejo Ecuador.
She graduated from the Department of Electrical Engineering in Automatic Control in 1979 at
the Universidad de Oriente, Cuba, and defended her doctorate in Technical Sciences in 1992.
She can be contacted at email: yolanda.llosas@utm.edu.ec.

Comparative study by numerical simulation of two methods for automatic flow control in centrifugal pumps

More Related Content

Similar to Comparative study by numerical simulation of two methods for automatic flow control in centrifugal pumps (20)

More from International Journal of Power Electronics and Drive Systems (IJPEDS) (20)

Recently uploaded (20)

Comparative study by numerical simulation of two methods for automatic flow control in centrifugal pumps