Research ArticleNumerical Simulation and Computational Flow CharacterizationAnalyses of Centrifugal Pump Operating as Turbine

Du Jianguo1 Guanghui Chang1 Daniel Adu 12 Ransford Darko3

Muhammad A S Khan1 and Eric O Antwi4

1School of Management Science amp Eng Jiangsu University Zhenjiang 212013 China2Faculty of Engineering Accra Technical University Barnes Road Accra Ghana3Department of Agricultural Engineering University of Cape Coast PMB 233 Cape Coast Ghana4Department of Energy and Environmental Engineering University of Energy and Natural Resources Sunyani Sunyani Ghana

Correspondence should be addressed to Daniel Adu adudaniel39yahoocom

Received 29 April 2021 Revised 14 July 2021 Accepted 13 August 2021 Published 21 August 2021

Academic Editor Xiaoqing Bai

Copyright copy 2021 Du Jianguo et al (is is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Using a pump in reverse mode as a hydraulic turbine remains an alternative for hydropower generation in meeting energy needsespecially for the provision of electricity to remote and rural settlements(e primary challenge with small hydroelectric systems isattributed to the high price of smaller size hydraulic turbines A specific commercial pump model with a flow rate of 125m3hhead 32m pressure side diameter of 50mm impeller out and inlet diameters of 160mm and 6mm respectively was chosen forthis research(is research aimed to investigate a pumprsquos flow characteristics as a turbine to help select a suitable pump to be usedas a turbine for micro- or small hydropower construction Numerical methodologies have been adopted to contribute to thethoughtful knowledge of pressure and velocity distribution in the pump turbine performance In this study the unsteady flowrelations amongst the rotating impeller and stationary volute of the centrifugal pump made up four blades and four splittersIntermittent simulation results of pressure and velocity flow characteristics were studied considering diverse impeller suctionangles (e study was conducted by considering a wide range of rotational speeds starting from 750 rpm to 3250 rpm From theresults it was found that PAT operation was improved when operated at low speeds compared to high-speed operation (usspeeds between 1500 rpm and 2000 rpm were suitable for PAT performance (is research helps to realize the unsteady flowphysiognomies which provide information for future research on PAT (is study makes useful facts available which could behelpful for the pump turbine development Future studies should focus on cost analysis and emission generation inenergy generation

1 Introduction

Centrifugal pumps are commonly used for domestic andindustrial applications such as water supply energy gen-eration flood control irrigation for agricultural purposesand transportation of liquid-solid mixtures Designs ofhydropower plants are based on its ability to function at all-out efficiencymatching with acceptable speed Hydroturbineperformance rests on its speed therefore it is manufacturedto function at the greatest attainable efficiency in a specificrapidity (e role head is to provide speed since the speedrests on its head (e ability of the head variation can affect

the performance of the turbine (erefore it is essential forthe adjustment of rotational speed to result in a possible all-out efficiency [1 2] PAT A can in general take care ofhigher flow rates which sequentially could implicate higheroutput energy ranges [3] Nevertheless since pump pro-ducers have not made available pump performance curves inturbine mode it is usually not easy to select an appropriatepump turbine to run suitably at specific operating condi-tions which has become a significant challenge [4] Severalinvestigational and theoretical research studies were carriedout on the performance of pumps running in reverse [5](eeasiest way is the pump model-based prediction since the

HindawiComplexityVolume 2021 Article ID 9695452 9 pageshttpsdoiorg10115520219695452

user only needs to have access to the necessary informationof the pump performance studies for example QP HP andηP as it makes finding correspondent in turbine modeoperation easier using simple scheming As a result muchattention was given to pump mode performance parametersby theoretical researchers such as [6ndash8] Many researcherslike [9ndash11] have established some PAT relationships withprecise pump speed Various researchers have continuallyconducted several research pieces to provide a PAT generalperformance prediction method [12ndash14] Comparative re-search was conducted by Hossain et al [15] not forgetting[8 16] methods to come out with a knowledge edge that isbetter than the others(emethod of [16ndash18] was discoveredto offer a deeper understanding of PAT performancecharacteristics they consist of the ldquoarea ratiordquo which was atheoretical method developed in [19ndash21] A study wasconducted to investigate material consumption and carbonemissions in selected towns and provinces in China (eoutcome was that total infrastructure building delivery in theselected towns was not leveled (us infrastructure buildingsupply remained steady along with economic growth [22]Cao et al carried out an experiment and numerical study onsnowdrift on a typical large-span retractable roof utilizingthe Euler-Euler technique It was revealed that not oneunchanging snow delivery on the roofs was different whichmust be well thoughtout [23] Other researchers alsoresearched the various turbines for power generation axialflow pumps and energy conservation [24ndash26] A stan-dardized pump can operate as a hydraulic turbine deprivedof modifications to the casing or the impeller geometryConsideration should only be given to the selection of ap-propriate flow rate (Q) head (H) and speed (n) for a site Itis conceivable to attain an equally high PAT efficiency levelas that of pump mode operation in most cases At bestturbine mode operation the impact of PAT on pressuremovement is controlled (e dissimilarity of shared shockloss controls a major part of the energy from the fluid whichis then transported open-air through the shaft by PAT (epower take out could be utilized again either in amechanicalor electrical form (e dissimilarity between the turbine andthe pump rests on their rotational bearing which is indi-cated by an arrow as shown in Figure 1 Additionally theturbine head usually has a higher head and flow at BEP (bestefficiency point) than the pumpmode at equal speed Table 1shows the comparison of methods to obtain BEP for PATproposed by previous researchers

2 Centrifugal Pump Model

A specific commercial pump model which has four mainblades and four splitter head at 16m flow rate of 125m3hand running at a speed of 2900 rpm was carefully chosen forthe paper (e pump is mainly composed of long straightpipes as shown in [27] (e entire computational flowdomains were generated using three-dimensional (3D) ProE 50 software (e long straight pipes at the impeller inletand volute outlet were adequately extended to make thesimulation more accurate Figure 2 displays the completemodel of the selected commercial pump

Summary of the scheme parameters and hydraulicevaluation description of the selected centrifugal pumpmodel under the study are presented in Table 2

3 Numerical Simulations

31 Computational Domains Meshing is the process ofdiscretizing a region under consideration to a set of thecontrol volume Mesh for diverse PAT scheme parts such asvolute covering wearing the ring and front plus backchambers is generated in ANSYS ICEM 145 (e CFDcalculation method was used An unstructured grid with atetrahedral mesh was selected to represent and create all theparts of PAT fluid domains Since mesh of this type hasoutstanding compliance particularly for flows which hascomplicated border situations Mesh independence studywas performed by Zhang et al [28] and the results show thatmesh size of 165 was selected for creating the computationaldomain bearing in mind the structure of the appliedcomputing for which Figure 3 provides the pictures of thewhole mesh areas

32 Transient Simulation Setup In carrying out a numericalstudy ANSYS CFX code 16 was chosen for creating PATgeometry (e K-omega (k-ω) turbulence model was se-lected for the study [8]

zkzt

+ Ujzkzxj

Pk minus βlowastkω +z

zxjv + σkvT( 1113857

zkzxj

(1)

Specific dissipation rate

zωzt

+ Ujzωzxj

prop s2 minus βω2+

z

zxjv + σωvT( 1113857

zωzxj

+ 2 1 minus F1( 1113857σω21ω

zkzx

zωzxi

(2)

(e values k and ω follow closely after the transportequations for turbulent kinetic energy as well as the dissi-pation range of turbulence [6]

Pump TurbineQP QT

PTnT

nT = knostnP=konst

nT =0

Md= 0

HTHP

QD

QP QT

QP QT

HD

PP

nPHS HS

HP=HDndashHS HT=HDndashHS

Figure 1 Structure of dissimilarity amongst pump and turbinemodes

2 Complexity

All boundary conditions were set (e steady calcula-tions were first performed to make an initial flow fieldavailable for the unsteady simulation In each time step1times 12minus5 was set (e simulation was conducted for 12complete impeller rotations to achieve a steady simulationoutcome Flow field pressure and absolute velocity datawere further studied in the study (ere was one interfacebetween the two domains when simulating a PAT model(is interface was found between the stationary volute andthe rotating impeller domain (e interface between do-mains is created using the ldquointerfacerdquo icon on the toolbar(e entire interface was designed and the interface type wasset to a fluid-fluid interface (e general grid interface (GGI)connection option was selected as the interface model for thetwo separate individual rotary and stator fields (e sta-tionary blade was used for frame change

4 Results and Discussion

41 Hydraulic PAT Performs Characteristics Curves (espeed of a turbine is determined by the amount of load usedAt constant flow rate and head the turbine speed decreaseswhen the turbine uses a higher load more than the designload with the decrease of the burden placed on a turbinewhereas the head and flow rate remain constant theturbine speed increases [29] (e same principle applieswhen hydraulic pumps are used as a turbine at a specifiedsite with a continuous flow rate and contribute signifi-cantly to determining the turbinersquos performance charac-teristics As the PAT load is further reduced the PATrsquosspeed reaches a maximum value at which the PAT candeliver no torque (is speed is called runaway speed (ePATrsquos runaway speed was obtained to be approximately2250 rpm as can be seen in Figure 4 while Figure 5 showsthe relationship between speed and power From the re-sults as the speed of PAT increases the power output alsorises until it reaches 1750 rpm (512 kw) after which thepower began to drop progressively Increase in PATrsquosefficiency with the increase in impellerrsquos rotational speeduntil it reaches 1500 rpm should be also mentioned asshown in Figure 6 It means that when PAT operatesbeyond 1500 rpm the efficiency of the PAT will decreasePAT efficiency behaves equally as the power output of thePAT (erefore it can be concluded that the best-oper-ating conditions for selected PAT could be achieved whenit operates between 1000 and 1750 pm

Table 1 Comparison methods to obtain BEP for PAT

Ns (m m3s) ηPmax () D (m) h q h q h q h q h q146 655 0250 205 156 163 128 178 145 220 209 214 148147 (Williams 1992) 46 0125 287 163 217 147 254 186 284 24 271 172207 (micro-hydropower development 2003) 60 0160 224 173 184 142 24 166 222 204 229 174230 76 0250 195 159 141 119 149 129 178 176 194 160348 (Joshi S 2005) 83 mdash 171 155 128 114 134 120 149 136 170 148364 (Singh 2005) 744 0175 172 154 181 134 143 127 173 178 171 150377 865 0250 173 148 124 111 127 116 134 116 165 144452 (Singh 2005) 80 0200 140 138 156 125 131 12 151 149 151 133556 87 0250 134 115 123 111 126 116 132 113 138 118

Back ChamberImpeller

Outlet

VoluteFront Chamber

Inlet

Figure 2 Complete 3D model of the selected centrifugal pump

Table 2 Scheme parameters of the selected pump model

Pressure side diameter D1 50mm Flow rate qd 125m3hImpeller out diameter D2 160mm Head hd 32mImpeller out width b2 6mm Efficiency η 56Long blades 4 4Splitter 4 4Splitter inlet diameter dimm 104mmPower P 3 kW Speed n 2900 rpm

Wearing ring

Back chamber

Volute

Front chamber

Delivery pipe

Wearing ring

ImpellerTong region

Figure 3 (e whole created mesh areas of the selected centrifugalpump model

Complexity 3

42 Performance of PAT Predicted Based on PumpPerformance PATsimulation results acquired from ANSYSCFX was used to evaluate the performance parameters of thePAT It is essential to recognise these parameters for PATperformance under different flow rate conditions (eyconsist of the head power and nondimensionless coeffi-cients such as head flow and power coefficient Pumpmodedata can predict PAT performance by using the pump BEP(is section will determine the PAT performance using theBEP of the pump and compare anticipated results with theCFD simulation result (e performance parameters of PATcould be predicted by considering pumpmode performance(e public affinity laws are used to establish relationships

between the performance parameters of similar machinesoperating under similar kinematic conditions It is essentialto know the following dimensionless parameters which arevital in determining the turbinersquos performance character-istics(ey consist of a head flow and power coefficient(esimilarities between these parameters are that the variableswith the particular coefficient are made dimensionlessthrough normalising with a diameter (D) and rotation speed(N) Except for the flow coefficient previous and latterparameters call for one additional variable to normalise (edimensionless coefficients result realised from the PATnumerical simulation has also been plotted as displayed inFigure 5 showing the head coefficient flow coefficient andpower coefficient (ese results show the differences whichexist among these coefficients (us Figure 6 indicates thedifference in the head coefficient at different flow coefficientlevels and different rotational speeds It can be witnessedfrom the results that the increase in the head coefficientincreases the flow coefficient but this occurs at a differentrotational speed thus increase in rotational speed decreasesthe head coefficient as shown in Figure 7 Figure 8 shows thedifference in the flow coefficient via the power coefficientOne can observe from the figure that the power coefficientperforms comparably with the PAT flow coefficient underdifferent flow rate conditions It is characterised by reducingthe value as the flow coefficient drops Figure 9 shows thehead coefficient vs flow coefficient curves of a selected PATat a different flow rates

(e nondimensional parameters at various rotationalspeeds have been studied (let us say ϕ 08) as tabulated inTable 3 (e study discovered that an increase in speed in-creases power coefficient (π) and efficiency η Neverthelesstheir differences in ψ were not much except at higher speeds(e results are as shown in the curves below with the BEPpoints at different rotational speeds also indicated (e poweroutput and efficiency were found better in the speed range of1000ndash1200 rpm As PAT operates in the opposite direction ofpumpmode operation the energy converted from hydraulic tomechanical might not be effectual at higher speeds causing itsbetter performance at the lower speed level

43 Transient Static Pressure Supply under Diverse Flow Ratein Lieu of Centrifugal Pump Impeller Pressure distributionunder different flow conditions and rotational speed in pumpas turbine fluid domain are displayed in Figure 10 From theresults pressure progressively intensified in a streamwise waywith the pressure side recording the maximum Flow partingcould be observed at the impeller blade leading edge (e flowat the inlet is not at a tangent with a blade with uniformed flowobserved near the blade which causes flow parting to occur onthe blade surface Minimal pressure was found at the impellerinlet along the suction area representing location cavitationwhich usually occurs in pumps

It is observed from Figure 11 that static pressure wasasymmetrically disseminated (e minimum pressure hap-pens at the blade suction area Pressure increases as the flowrate reaches the lowest values while it decreases with in-creasing flow rates as shown in Figure 11(b) Comparing the

130

230

330

430

530

08 1 12 14

P(kW

)

Q (QD)

1000RPM1750RPM1500RPM

Figure 5 Power vs rotational speed for selected PAT

750 1000 1250 1500 1750 2000 2250 2500N=(RPM)

η=(

)

72

70

68

66

64

62

60

Nrunaway

NBE

P

Figure 6 Power vs efficiency for selected PAT at Qt 12m3s

750 1000 1250 1500 1750 2000 2250 2500

Plot Are

N=RPM

530

480

430

380

330

280

230

P(kw

)

Nru

naw

ay

NBE

P

Figure 4 Power vs rotational speed for selected PAT atQt 12m3s

4 Complexity

two selected rotational speeds under three flow conditionsfromQ 008m3s toQ 12m3s a rise in the flow rate alsobrought about an increase in drop closer to the exit duct ofthe volute Proper recirculation of flow could be observed atthe impeller suction area due to lower flow rates though theflow on the pressure side is smooth On the contrary as theflow rate increases flow parting along the pressure side

occurs bringing about continuous flow movement in thepressure side (is movement in the suction side of theimpeller blade drops as the flow rate rises

44 Transient Velocity Distribution in the PAT Model for aSelected Centrifugal Pump Hydrodynamics of flow in PATwas numerically calculated using CFD to find the supply ofactual parameters defining hydrodynamics flow in thepump which interconnects the PAT characteristicsrsquo geo-metric model Figure 12 shows the velocity in the PATcolored by relative velocity distribution Considering thevolute shape one could observe the velocity increase with arise flow rate Flow thenmoves smoothly in the volute regionat a minimal flow rate Moreover it slows down after strikingthe trailing edge of the impeller after which relative fluidvelocity begins to once again increase to some extent (eCFD model shows a trend of velocity differences withchanges in rotational speed It was also observed that aswater enters the suction side relative velocity intensifiednext to the curving of the blade in the streamwise directionas the passage bit by bit gets smaller In summary a portionof the simulation was performed by the PAT impeller as afeedback turbine as a result of reaction from the pressuredrop with a change in kinetic energy also playing a role

45 Transient Static Pressure Distribution for CentrifugalPump Impeller In analysing flow features of the centrifugalpump in pump mode pressure and velocity contour plotsand vectors are employed (e streamlines of the 3D modelsare plotted to examine the flow design in a centrifugal pumpFigure 13 displays the static pressure transfer of the des-ignated PAT model under a diverse flow rate atn 1750 rpm(e results show that lower static pressure wasobserved at the blade leading edge As kinetic energy flowedand transported to fluid the static pressure progressivelyincreases (e figure shows that maximum pressure wasmonitored on the blade side whereas minimal pressure wasalso detected on the blade suction side at the same radialposition As the fluid leads it is a way to the volute casing(e static pressure further increased as shown in Figure 13which could be attributed to the conversion of the fluidrsquoskinetic energy to pressure energy because of the increasingcasing cross-segment region Figure 14 also displays thevelocity delivery of the designated PATmodel under diverseflow rate n 1750 rpm

Table 3 Parameters at ϕ 08 at different rotational speeds for theoriginal impeller

N (rpm) ψ π ϕ η ()1250 88 016 771 0002021500 71 007 645 0002071750 66 008 553 0001842000 64 004 484 000106

65697377818589

1250 1500 1750 2000Rotational speed=(RPM)

Hea

d co

effic

ient

(ψ)

Figure 7 Rotational speed vs head coefficient characteristics curve of selected PAT at Qt 08m3s

00150018002100240027

0030033

00002 00004 00006 00008 0001 00012 00014 00016

Hea

d co

effic

ient

(ψ)

Power coefficient (π)

Figure 8 Power coefficient vs flow coefficient curve of selectedPAT at Qt 08m3s and n 1500 rpm

5

7

9

11

13

06 08 1 12

Hea

d C

oeffi

cien

t (ψ)

Q (m3h)

150017502000

Figure 9 Head coefficient vs flow coefficient curves of a selectedPAT at a different flow rates

Complexity 5

Pressure1877e+005

1780e+005

1683e+005

1585e+005

1488e+005

1391e+005

1294e+005

1197e+005

1100e+005

1003e+005

9060e+004[Pa]

Q=08 m2s Q=10 m2s Q=12 m2s

(a)

Pressure

[Pa]

1251e+005

1128e+005

1005e+005

8827e+004

7599e+004

6372e+004

5144e+004

3917e+004

2689e+004

1462e+004

2345e+003 Q=08 m2s Q=10 m2s Q=12 m2s

(b)

Figure 11 (a) 3D pressure dissemination for the PATmodel under the dissimilar flow rate n 1750 rpm (b) 3D pressure dissemination forthe PAT model under different flow rates n 1500 rpm

Total Pressure Total Pressure1294e+004-2749e+002-1349e+004-2670e+004-3991e+004-5312e+004-6633e+004-7954e+004-9275e+004-1060e+005-1192e+005

1870e+0044273e+003-1015e+004-2457e+004-3900e+004-5342e+004-6785e+004-8227e+004-9669e+004-1111e+005-1255e+005

[Pa] [Pa]Q=08m3s at n=1750rpm Q=10m3s n=1750rpm

Figure 10 Total pressure distribution on the PAT impeller at a different flow rate

6 Complexity

1734e+001

1561e+001

1387e+001

1214e+001

1041e+001

8671e+000

6937e+000

5203e+000

3469e+000

1734e+000

0000e+000[m s^-1]

Velocity in Stn Frame

Q=01m3s Q=10m3s Q=12m3s

Figure 12 Absolute velocity delivery of the designated PAT model under diverse flow rate n 1750 rpm

2400e+005

2240e+005

2080e+005

1920e+005

1760e+005

1600e+005

1440e+005

1280e+005

1120e+005

9600e+004

8000e+004[Pa]

Pressure

Q=08m3s Q=10m3s Q=12m3s Q=14m3s

Figure 13 Static pressure delivery of the designated PAT model under diverse flow rate n 1750 rpm

1101e+0019910e+0008809e+0007708e+0006607e+0005506e+0004404e+0003303e+0002202e+0001101e+0000000e+000

[m s^-1]

Velocity

Q=08m3s Q=10m3s Q=12m3s Q=14m3s

Figure 14 Velocity distribution for the selected centrifugal pump model at different flow rates N 1750 rpm

Complexity 7

5 Conclusion

In this study the comparison of numerical studies resultsobtained under different flow and speed conditions wasinvestigated From the study the following conclusions weremade

(A) (e speed of a turbine is determined by the amountof load used At constant flow rate and head theturbine speed decreases when the turbine uses ahigher load more than the design load with thedecrease of the burden placed on a turbine whereasthe head and flow rate remain constant the turbinespeed increases (e same principle applies whenhydraulic pumps are used as a turbine at a specifiedsite with a continuous flow rate and contributesignificantly to determining the turbinersquos perfor-mance characteristics As the PAT load is furtherreduced the PATrsquos speed reaches a maximum valuethat the PAT can deliver no torque (is speed iscalled runaway speed (e runaway speed of thePAT was obtained to be approximately 2250 rpm

(B) Increase in PATrsquos efficiency with the increased ro-tational speed of the impeller till it reaches 1500 rpmshould also be mentioned It means that when PAToperates beyond 1500 rpm the efficiency of the PATwill decrease In comparison with Figures 6 and 7PAT efficiency behaves equally as the power outputof the PAT (erefore it can be concluded that thebest-operating conditions for selected PAT could beachieved when it operates between 1000 rpm and1750 rpm

(C) Pressure distribution under different flow condi-tions and rotational speed in the pump as turbinefluid domain From the results pressure progres-sively intensified in a streamwise way with thepressure side recording the maximum Flow partingcould be observed at the impeller blade leading edge(e flow at the inlet is not at a tangent with the bladewith ununiformed flow observed near the bladewhich causes flow parting to occur on the bladesurface Minimal pressure was found at the impellerinlet along the suction area representing locationcavitation typically occurs in pumps

In summary PAT simulation results acquired fromANSYS CFX have been used to evaluate the PATrsquos per-formance parameters It is essential to recognise these pa-rameters for PAT performance under different flow rateconditions (ey consist of the head power and non-dimensionless coefficients such as head flow and powercoefficient Pump mode data can predict PAT performanceby using the pump BEP

Data Availability

(e data are openly available in a public repository thatissues datasets with DOIs (e authors confirm that the datasupporting the findings of this study are available within thearticle and could be obtained upon request (e data

generated at a central large-scale facility are available uponrequest Raw data were generated at Jiangsu University Lab(e derived data supporting the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

(e authors declare that they have no conflicts of interest

Acknowledgments

(is study was supported by the National Science Foun-dation of China under Grant nos 71974081 71704066 and71971100 and special funds of the National Social ScienceFund of China (Grant no 18VSJ038)

References

[1] J Gonzalez and C Santolaria ldquoUnsteady flow structure andglobal variables in a centrifugal pumprdquo Journal of FluidsEngineering vol 128 no 5 pp 937ndash946 2006

[2] B K Gandhi S N Singh and V Seshadri ldquoPerformancecharacteristics of centrifugal slurry pumpsrdquo Journal of FluidsEngineering vol 123 no 2 pp 271ndash280 2001

[3] H Nautiyal A Varun and A Kumar ldquoReverse runningpumps analytical experimental and computational study areviewrdquo Renewable and Sustainable Energy Reviews vol 14no 7 pp 2059ndash2067 2010

[4] S V Jain and R N Patel ldquoInvestigations on pump running inturbine mode a review of the state-of-the-artrdquo Renewable andSustainable Energy Reviews vol 30 pp 841ndash868 2014

[5] H Nautiyal V Varun A Kumar and S Y S Yadav ldquoEx-perimental investigation of centrifugal pump working asturbine for small hydropower systemsrdquo Energy Science andTechnology vol 1 pp 79ndash86 2011

[6] A J Stepanoff Centrifugal and Axial Flow Pumps Design andApplications John Wiley amp Sons New York NY USA 1957

[7] S Childs ldquoConvert pumps to turbines and recovers HPrdquoHydrocarbon Processing and Petroleum Refiner vol 41pp 173-174 1962

[8] R K Sharma Small Hydroelectric ProjectsmdashUse of CentrifugalPumps as Turbines Kirloskar Electric Co Bangalore India1984

[9] S Gopalakrishnan ldquoPower recovery turbines for the processindustryrdquo in Proceedings of the ltird International PumpSymposium pp 3ndash11 Texas AampM University Houston TXUSA September 1986

[10] H Diederich ldquoVerwendung von kreiselpumpen als turbinenrdquoKSB Techn Ber vol 12 pp 30ndash36 1967

[11] K Grover Conversion of Pumps to Turbines GSA Intercorporate Katonah NY USA 1982

[12] A Morabito and P Hendrick ldquoPump as turbine applied tomicro energy storage and smart water grids a case studyrdquoApplied Energy vol 241 pp 567ndash579 2019

[13] M Rossi A Nigro and M Renzi ldquoExperimental and nu-merical assessment of a methodology for performance pre-diction of pumps-as-turbines (PaTs) operating in off-designconditionsrdquo Applied Energy vol 248 pp 555ndash566 2019

[14] B Luo and X C L Wang ldquoNumerical simulation of flow-induced vibration of double-suction centrifugal pump asturbinerdquo Journal Drainage Irrigation Mechanical Engineeringvol 37 pp 313ndash318 2019

8 Complexity

[15] I M Hossain S Ferdous and T Jamal ldquoPump-as-turbine(PAT) for small-scale power generation a comparativeanalysisrdquo in Proceedings of the 2014 3rd International Con-ference on the IEEE Developments in Renewable EnergyTechnology (ICDRET) pp 1ndash5 Dhaka Bangladesh May 2014

[16] J-M Chapallaz P Eichenberger and G Fischer ldquoManual onpumps used as turbinesrdquo Vieweg vol 11 1992

[17] S Derakhshan and A Nourbakhsh ldquo(eoretical numericaland experimental investigation of centrifugal pumps in re-verse operationrdquo Experimental ltermal and Fluid Sciencevol 32 no 8 pp 1620ndash1627 2008

[18] P Vasanthakumar J Krishnaraj S Karthikayan T Vinothand S K Arun Sankar ldquoInvestigation on reverse character-istics of centrifugal pump in turbine mode a comparativestudy by an experimentation and simulationrdquo MaterialsToday Proceedings vol 4 no 2 pp 693ndash700 2017

[19] M Gantar ldquoPropeller pumps running as turbinesrdquo in Pro-ceedings of the Conference on Hydraulic Machinery pp 237ndash248 Ljubljana Slovenia August 1988

[20] H H Anderson ldquoBorehole pumpsrdquo Centrifugal Pumps andAllied Machinery vol 4 pp 177ndash185 1993

[21] D Adu J Zhang M Jieyun S N Asomani and M OsmanKoranteng ldquoNumerical investigation of transient vortices andturbulent flow behaviour in centrifugal pump operating inreverse mode as turbinerdquo Materials Science for EnergyTechnologies vol 2 no 2 pp 356ndash364 2019

[22] H Wang Y Wang C Fan et al ldquoMaterial consumption andcarbon emissions associated with the infrastructure con-struction of 34 cities in northeast Chinardquo Complexityvol 2020 Article ID 4364912 20 pages 2020

[23] Z Cao M Liu and P Wu ldquoExperiment investigation andnumerical simulation of Snowdrift on a typical large-spanretractable roofrdquo Complexity vol 2019 Article ID 598480414 pages 2019

[24] K Ren H Li S Li and H Dong ldquoVoltage stability analysis offront-end speed controlled wind turbine integrated into re-gional power grid based on bifurcation theoryrdquo Complexityvol 2020 Article ID 8816334 11 pages 2020

[25] D Florez-Orrego I B Henriques T-V Nguyen et al ldquo(econtributions of prof Jan Szargut to the exergy and envi-ronmental assessment of complex energy systemsrdquo Energyvol 161 pp 482ndash492 2018

[26] Y Zhang Y Xu Y Zheng et al ldquoMultiobjective optimizationdesign and experimental investigation on the axial-flow pumpwith orthogonal test approachrdquo Complexity vol 2019 ArticleID 1467565 14 pages 2019

[27] X Xu X Yuan M Cui Y Yuan and L Qi ldquoJoint optimi-zation of energy conservation and migration cost for complexsystems in edge computingrdquo Complexity vol 2019 Article ID6180135 14 pages 2019

[28] Y L Zhang S Q Yuan Z Jinfeng et al ldquoNumerical in-vestigation of the effects of splitter blades on the cavitationperformance of a centrifugal pumprdquo IOP Conference SeriesEarth and Environmental Science vol 22 Article ID 0520032014

[29] Z Jinfeng A Daniel S Lv and K Majid ldquoSmall hydropowerby using pump as turbine for power generation in SouthernAfricardquo Current Science vol 118 no 3 2019

Complexity 9

All boundary conditions were set (e steady calcula-tions were first performed to make an initial flow fieldavailable for the unsteady simulation In each time step1times 12minus5 was set (e simulation was conducted for 12complete impeller rotations to achieve a steady simulationoutcome Flow field pressure and absolute velocity datawere further studied in the study (ere was one interfacebetween the two domains when simulating a PAT model(is interface was found between the stationary volute andthe rotating impeller domain (e interface between do-mains is created using the ldquointerfacerdquo icon on the toolbar(e entire interface was designed and the interface type wasset to a fluid-fluid interface (e general grid interface (GGI)connection option was selected as the interface model for thetwo separate individual rotary and stator fields (e sta-tionary blade was used for frame change

4 Results and Discussion

41 Hydraulic PAT Performs Characteristics Curves (espeed of a turbine is determined by the amount of load usedAt constant flow rate and head the turbine speed decreaseswhen the turbine uses a higher load more than the designload with the decrease of the burden placed on a turbinewhereas the head and flow rate remain constant theturbine speed increases [29] (e same principle applieswhen hydraulic pumps are used as a turbine at a specifiedsite with a continuous flow rate and contribute signifi-cantly to determining the turbinersquos performance charac-teristics As the PAT load is further reduced the PATrsquosspeed reaches a maximum value at which the PAT candeliver no torque (is speed is called runaway speed (ePATrsquos runaway speed was obtained to be approximately2250 rpm as can be seen in Figure 4 while Figure 5 showsthe relationship between speed and power From the re-sults as the speed of PAT increases the power output alsorises until it reaches 1750 rpm (512 kw) after which thepower began to drop progressively Increase in PATrsquosefficiency with the increase in impellerrsquos rotational speeduntil it reaches 1500 rpm should be also mentioned asshown in Figure 6 It means that when PAT operatesbeyond 1500 rpm the efficiency of the PAT will decreasePAT efficiency behaves equally as the power output of thePAT (erefore it can be concluded that the best-oper-ating conditions for selected PAT could be achieved whenit operates between 1000 and 1750 pm

Table 1 Comparison methods to obtain BEP for PAT

Ns (m m3s) ηPmax () D (m) h q h q h q h q h q146 655 0250 205 156 163 128 178 145 220 209 214 148147 (Williams 1992) 46 0125 287 163 217 147 254 186 284 24 271 172207 (micro-hydropower development 2003) 60 0160 224 173 184 142 24 166 222 204 229 174230 76 0250 195 159 141 119 149 129 178 176 194 160348 (Joshi S 2005) 83 mdash 171 155 128 114 134 120 149 136 170 148364 (Singh 2005) 744 0175 172 154 181 134 143 127 173 178 171 150377 865 0250 173 148 124 111 127 116 134 116 165 144452 (Singh 2005) 80 0200 140 138 156 125 131 12 151 149 151 133556 87 0250 134 115 123 111 126 116 132 113 138 118

Back ChamberImpeller

Outlet

VoluteFront Chamber

Inlet

Figure 2 Complete 3D model of the selected centrifugal pump

Table 2 Scheme parameters of the selected pump model

Pressure side diameter D1 50mm Flow rate qd 125m3hImpeller out diameter D2 160mm Head hd 32mImpeller out width b2 6mm Efficiency η 56Long blades 4 4Splitter 4 4Splitter inlet diameter dimm 104mmPower P 3 kW Speed n 2900 rpm

Wearing ring

Back chamber

Volute

Front chamber

Delivery pipe

Wearing ring

ImpellerTong region

Figure 3 (e whole created mesh areas of the selected centrifugalpump model

Complexity 3

42 Performance of PAT Predicted Based on PumpPerformance PATsimulation results acquired from ANSYSCFX was used to evaluate the performance parameters of thePAT It is essential to recognise these parameters for PATperformance under different flow rate conditions (eyconsist of the head power and nondimensionless coeffi-cients such as head flow and power coefficient Pumpmodedata can predict PAT performance by using the pump BEP(is section will determine the PAT performance using theBEP of the pump and compare anticipated results with theCFD simulation result (e performance parameters of PATcould be predicted by considering pumpmode performance(e public affinity laws are used to establish relationships

between the performance parameters of similar machinesoperating under similar kinematic conditions It is essentialto know the following dimensionless parameters which arevital in determining the turbinersquos performance character-istics(ey consist of a head flow and power coefficient(esimilarities between these parameters are that the variableswith the particular coefficient are made dimensionlessthrough normalising with a diameter (D) and rotation speed(N) Except for the flow coefficient previous and latterparameters call for one additional variable to normalise (edimensionless coefficients result realised from the PATnumerical simulation has also been plotted as displayed inFigure 5 showing the head coefficient flow coefficient andpower coefficient (ese results show the differences whichexist among these coefficients (us Figure 6 indicates thedifference in the head coefficient at different flow coefficientlevels and different rotational speeds It can be witnessedfrom the results that the increase in the head coefficientincreases the flow coefficient but this occurs at a differentrotational speed thus increase in rotational speed decreasesthe head coefficient as shown in Figure 7 Figure 8 shows thedifference in the flow coefficient via the power coefficientOne can observe from the figure that the power coefficientperforms comparably with the PAT flow coefficient underdifferent flow rate conditions It is characterised by reducingthe value as the flow coefficient drops Figure 9 shows thehead coefficient vs flow coefficient curves of a selected PATat a different flow rates

(e nondimensional parameters at various rotationalspeeds have been studied (let us say ϕ 08) as tabulated inTable 3 (e study discovered that an increase in speed in-creases power coefficient (π) and efficiency η Neverthelesstheir differences in ψ were not much except at higher speeds(e results are as shown in the curves below with the BEPpoints at different rotational speeds also indicated (e poweroutput and efficiency were found better in the speed range of1000ndash1200 rpm As PAT operates in the opposite direction ofpumpmode operation the energy converted from hydraulic tomechanical might not be effectual at higher speeds causing itsbetter performance at the lower speed level

43 Transient Static Pressure Supply under Diverse Flow Ratein Lieu of Centrifugal Pump Impeller Pressure distributionunder different flow conditions and rotational speed in pumpas turbine fluid domain are displayed in Figure 10 From theresults pressure progressively intensified in a streamwise waywith the pressure side recording the maximum Flow partingcould be observed at the impeller blade leading edge (e flowat the inlet is not at a tangent with a blade with uniformed flowobserved near the blade which causes flow parting to occur onthe blade surface Minimal pressure was found at the impellerinlet along the suction area representing location cavitationwhich usually occurs in pumps

It is observed from Figure 11 that static pressure wasasymmetrically disseminated (e minimum pressure hap-pens at the blade suction area Pressure increases as the flowrate reaches the lowest values while it decreases with in-creasing flow rates as shown in Figure 11(b) Comparing the

130

230

330

430

530

08 1 12 14

P(kW

)

Q (QD)

1000RPM1750RPM1500RPM

Figure 5 Power vs rotational speed for selected PAT

750 1000 1250 1500 1750 2000 2250 2500N=(RPM)

η=(

)

72

70

68

66

64

62

60

Nrunaway

NBE

P

Figure 6 Power vs efficiency for selected PAT at Qt 12m3s

750 1000 1250 1500 1750 2000 2250 2500

Plot Are

N=RPM

530

480

430

380

330

280

230

P(kw

)

Nru

naw

ay

NBE

P

Figure 4 Power vs rotational speed for selected PAT atQt 12m3s

4 Complexity

two selected rotational speeds under three flow conditionsfromQ 008m3s toQ 12m3s a rise in the flow rate alsobrought about an increase in drop closer to the exit duct ofthe volute Proper recirculation of flow could be observed atthe impeller suction area due to lower flow rates though theflow on the pressure side is smooth On the contrary as theflow rate increases flow parting along the pressure side

occurs bringing about continuous flow movement in thepressure side (is movement in the suction side of theimpeller blade drops as the flow rate rises

44 Transient Velocity Distribution in the PAT Model for aSelected Centrifugal Pump Hydrodynamics of flow in PATwas numerically calculated using CFD to find the supply ofactual parameters defining hydrodynamics flow in thepump which interconnects the PAT characteristicsrsquo geo-metric model Figure 12 shows the velocity in the PATcolored by relative velocity distribution Considering thevolute shape one could observe the velocity increase with arise flow rate Flow thenmoves smoothly in the volute regionat a minimal flow rate Moreover it slows down after strikingthe trailing edge of the impeller after which relative fluidvelocity begins to once again increase to some extent (eCFD model shows a trend of velocity differences withchanges in rotational speed It was also observed that aswater enters the suction side relative velocity intensifiednext to the curving of the blade in the streamwise directionas the passage bit by bit gets smaller In summary a portionof the simulation was performed by the PAT impeller as afeedback turbine as a result of reaction from the pressuredrop with a change in kinetic energy also playing a role

45 Transient Static Pressure Distribution for CentrifugalPump Impeller In analysing flow features of the centrifugalpump in pump mode pressure and velocity contour plotsand vectors are employed (e streamlines of the 3D modelsare plotted to examine the flow design in a centrifugal pumpFigure 13 displays the static pressure transfer of the des-ignated PAT model under a diverse flow rate atn 1750 rpm(e results show that lower static pressure wasobserved at the blade leading edge As kinetic energy flowedand transported to fluid the static pressure progressivelyincreases (e figure shows that maximum pressure wasmonitored on the blade side whereas minimal pressure wasalso detected on the blade suction side at the same radialposition As the fluid leads it is a way to the volute casing(e static pressure further increased as shown in Figure 13which could be attributed to the conversion of the fluidrsquoskinetic energy to pressure energy because of the increasingcasing cross-segment region Figure 14 also displays thevelocity delivery of the designated PATmodel under diverseflow rate n 1750 rpm

Table 3 Parameters at ϕ 08 at different rotational speeds for theoriginal impeller

N (rpm) ψ π ϕ η ()1250 88 016 771 0002021500 71 007 645 0002071750 66 008 553 0001842000 64 004 484 000106

65697377818589

1250 1500 1750 2000Rotational speed=(RPM)

Hea

d co

effic

ient

(ψ)

Figure 7 Rotational speed vs head coefficient characteristics curve of selected PAT at Qt 08m3s

00150018002100240027

0030033

00002 00004 00006 00008 0001 00012 00014 00016

Hea

d co

effic

ient

(ψ)

Power coefficient (π)

Figure 8 Power coefficient vs flow coefficient curve of selectedPAT at Qt 08m3s and n 1500 rpm

5

7

9

11

13

06 08 1 12

Hea

d C

oeffi

cien

t (ψ)

Q (m3h)

150017502000

Figure 9 Head coefficient vs flow coefficient curves of a selectedPAT at a different flow rates

Complexity 5

Pressure1877e+005

1780e+005

1683e+005

1585e+005

1488e+005

1391e+005

1294e+005

1197e+005

1100e+005

1003e+005

9060e+004[Pa]

Q=08 m2s Q=10 m2s Q=12 m2s

(a)

Pressure

[Pa]

1251e+005

1128e+005

1005e+005

8827e+004

7599e+004

6372e+004

5144e+004

3917e+004

2689e+004

1462e+004

2345e+003 Q=08 m2s Q=10 m2s Q=12 m2s

(b)

Figure 11 (a) 3D pressure dissemination for the PATmodel under the dissimilar flow rate n 1750 rpm (b) 3D pressure dissemination forthe PAT model under different flow rates n 1500 rpm

Total Pressure Total Pressure1294e+004-2749e+002-1349e+004-2670e+004-3991e+004-5312e+004-6633e+004-7954e+004-9275e+004-1060e+005-1192e+005

1870e+0044273e+003-1015e+004-2457e+004-3900e+004-5342e+004-6785e+004-8227e+004-9669e+004-1111e+005-1255e+005

[Pa] [Pa]Q=08m3s at n=1750rpm Q=10m3s n=1750rpm

Figure 10 Total pressure distribution on the PAT impeller at a different flow rate

6 Complexity

1734e+001

1561e+001

1387e+001

1214e+001

1041e+001

8671e+000

6937e+000

5203e+000

3469e+000

1734e+000

0000e+000[m s^-1]

Velocity in Stn Frame

Q=01m3s Q=10m3s Q=12m3s

Figure 12 Absolute velocity delivery of the designated PAT model under diverse flow rate n 1750 rpm

2400e+005

2240e+005

2080e+005

1920e+005

1760e+005

1600e+005

1440e+005

1280e+005

1120e+005

9600e+004

8000e+004[Pa]

Pressure

Q=08m3s Q=10m3s Q=12m3s Q=14m3s

Figure 13 Static pressure delivery of the designated PAT model under diverse flow rate n 1750 rpm

1101e+0019910e+0008809e+0007708e+0006607e+0005506e+0004404e+0003303e+0002202e+0001101e+0000000e+000

[m s^-1]

Velocity

Q=08m3s Q=10m3s Q=12m3s Q=14m3s

Figure 14 Velocity distribution for the selected centrifugal pump model at different flow rates N 1750 rpm

Complexity 7

5 Conclusion

In this study the comparison of numerical studies resultsobtained under different flow and speed conditions wasinvestigated From the study the following conclusions weremade

(A) (e speed of a turbine is determined by the amountof load used At constant flow rate and head theturbine speed decreases when the turbine uses ahigher load more than the design load with thedecrease of the burden placed on a turbine whereasthe head and flow rate remain constant the turbinespeed increases (e same principle applies whenhydraulic pumps are used as a turbine at a specifiedsite with a continuous flow rate and contributesignificantly to determining the turbinersquos perfor-mance characteristics As the PAT load is furtherreduced the PATrsquos speed reaches a maximum valuethat the PAT can deliver no torque (is speed iscalled runaway speed (e runaway speed of thePAT was obtained to be approximately 2250 rpm

(B) Increase in PATrsquos efficiency with the increased ro-tational speed of the impeller till it reaches 1500 rpmshould also be mentioned It means that when PAToperates beyond 1500 rpm the efficiency of the PATwill decrease In comparison with Figures 6 and 7PAT efficiency behaves equally as the power outputof the PAT (erefore it can be concluded that thebest-operating conditions for selected PAT could beachieved when it operates between 1000 rpm and1750 rpm

(C) Pressure distribution under different flow condi-tions and rotational speed in the pump as turbinefluid domain From the results pressure progres-sively intensified in a streamwise way with thepressure side recording the maximum Flow partingcould be observed at the impeller blade leading edge(e flow at the inlet is not at a tangent with the bladewith ununiformed flow observed near the bladewhich causes flow parting to occur on the bladesurface Minimal pressure was found at the impellerinlet along the suction area representing locationcavitation typically occurs in pumps

In summary PAT simulation results acquired fromANSYS CFX have been used to evaluate the PATrsquos per-formance parameters It is essential to recognise these pa-rameters for PAT performance under different flow rateconditions (ey consist of the head power and non-dimensionless coefficients such as head flow and powercoefficient Pump mode data can predict PAT performanceby using the pump BEP

Data Availability

(e data are openly available in a public repository thatissues datasets with DOIs (e authors confirm that the datasupporting the findings of this study are available within thearticle and could be obtained upon request (e data

generated at a central large-scale facility are available uponrequest Raw data were generated at Jiangsu University Lab(e derived data supporting the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

(e authors declare that they have no conflicts of interest

Acknowledgments

(is study was supported by the National Science Foun-dation of China under Grant nos 71974081 71704066 and71971100 and special funds of the National Social ScienceFund of China (Grant no 18VSJ038)

References

[1] J Gonzalez and C Santolaria ldquoUnsteady flow structure andglobal variables in a centrifugal pumprdquo Journal of FluidsEngineering vol 128 no 5 pp 937ndash946 2006

[2] B K Gandhi S N Singh and V Seshadri ldquoPerformancecharacteristics of centrifugal slurry pumpsrdquo Journal of FluidsEngineering vol 123 no 2 pp 271ndash280 2001

[3] H Nautiyal A Varun and A Kumar ldquoReverse runningpumps analytical experimental and computational study areviewrdquo Renewable and Sustainable Energy Reviews vol 14no 7 pp 2059ndash2067 2010

[4] S V Jain and R N Patel ldquoInvestigations on pump running inturbine mode a review of the state-of-the-artrdquo Renewable andSustainable Energy Reviews vol 30 pp 841ndash868 2014

[5] H Nautiyal V Varun A Kumar and S Y S Yadav ldquoEx-perimental investigation of centrifugal pump working asturbine for small hydropower systemsrdquo Energy Science andTechnology vol 1 pp 79ndash86 2011

[6] A J Stepanoff Centrifugal and Axial Flow Pumps Design andApplications John Wiley amp Sons New York NY USA 1957

[7] S Childs ldquoConvert pumps to turbines and recovers HPrdquoHydrocarbon Processing and Petroleum Refiner vol 41pp 173-174 1962

[8] R K Sharma Small Hydroelectric ProjectsmdashUse of CentrifugalPumps as Turbines Kirloskar Electric Co Bangalore India1984

[9] S Gopalakrishnan ldquoPower recovery turbines for the processindustryrdquo in Proceedings of the ltird International PumpSymposium pp 3ndash11 Texas AampM University Houston TXUSA September 1986

[10] H Diederich ldquoVerwendung von kreiselpumpen als turbinenrdquoKSB Techn Ber vol 12 pp 30ndash36 1967

[11] K Grover Conversion of Pumps to Turbines GSA Intercorporate Katonah NY USA 1982

[12] A Morabito and P Hendrick ldquoPump as turbine applied tomicro energy storage and smart water grids a case studyrdquoApplied Energy vol 241 pp 567ndash579 2019

[13] M Rossi A Nigro and M Renzi ldquoExperimental and nu-merical assessment of a methodology for performance pre-diction of pumps-as-turbines (PaTs) operating in off-designconditionsrdquo Applied Energy vol 248 pp 555ndash566 2019

[14] B Luo and X C L Wang ldquoNumerical simulation of flow-induced vibration of double-suction centrifugal pump asturbinerdquo Journal Drainage Irrigation Mechanical Engineeringvol 37 pp 313ndash318 2019

8 Complexity

[15] I M Hossain S Ferdous and T Jamal ldquoPump-as-turbine(PAT) for small-scale power generation a comparativeanalysisrdquo in Proceedings of the 2014 3rd International Con-ference on the IEEE Developments in Renewable EnergyTechnology (ICDRET) pp 1ndash5 Dhaka Bangladesh May 2014

[16] J-M Chapallaz P Eichenberger and G Fischer ldquoManual onpumps used as turbinesrdquo Vieweg vol 11 1992

[17] S Derakhshan and A Nourbakhsh ldquo(eoretical numericaland experimental investigation of centrifugal pumps in re-verse operationrdquo Experimental ltermal and Fluid Sciencevol 32 no 8 pp 1620ndash1627 2008

[18] P Vasanthakumar J Krishnaraj S Karthikayan T Vinothand S K Arun Sankar ldquoInvestigation on reverse character-istics of centrifugal pump in turbine mode a comparativestudy by an experimentation and simulationrdquo MaterialsToday Proceedings vol 4 no 2 pp 693ndash700 2017

[19] M Gantar ldquoPropeller pumps running as turbinesrdquo in Pro-ceedings of the Conference on Hydraulic Machinery pp 237ndash248 Ljubljana Slovenia August 1988

[20] H H Anderson ldquoBorehole pumpsrdquo Centrifugal Pumps andAllied Machinery vol 4 pp 177ndash185 1993

[21] D Adu J Zhang M Jieyun S N Asomani and M OsmanKoranteng ldquoNumerical investigation of transient vortices andturbulent flow behaviour in centrifugal pump operating inreverse mode as turbinerdquo Materials Science for EnergyTechnologies vol 2 no 2 pp 356ndash364 2019

[22] H Wang Y Wang C Fan et al ldquoMaterial consumption andcarbon emissions associated with the infrastructure con-struction of 34 cities in northeast Chinardquo Complexityvol 2020 Article ID 4364912 20 pages 2020

[23] Z Cao M Liu and P Wu ldquoExperiment investigation andnumerical simulation of Snowdrift on a typical large-spanretractable roofrdquo Complexity vol 2019 Article ID 598480414 pages 2019

[24] K Ren H Li S Li and H Dong ldquoVoltage stability analysis offront-end speed controlled wind turbine integrated into re-gional power grid based on bifurcation theoryrdquo Complexityvol 2020 Article ID 8816334 11 pages 2020

[25] D Florez-Orrego I B Henriques T-V Nguyen et al ldquo(econtributions of prof Jan Szargut to the exergy and envi-ronmental assessment of complex energy systemsrdquo Energyvol 161 pp 482ndash492 2018

[26] Y Zhang Y Xu Y Zheng et al ldquoMultiobjective optimizationdesign and experimental investigation on the axial-flow pumpwith orthogonal test approachrdquo Complexity vol 2019 ArticleID 1467565 14 pages 2019

[27] X Xu X Yuan M Cui Y Yuan and L Qi ldquoJoint optimi-zation of energy conservation and migration cost for complexsystems in edge computingrdquo Complexity vol 2019 Article ID6180135 14 pages 2019

[28] Y L Zhang S Q Yuan Z Jinfeng et al ldquoNumerical in-vestigation of the effects of splitter blades on the cavitationperformance of a centrifugal pumprdquo IOP Conference SeriesEarth and Environmental Science vol 22 Article ID 0520032014

[29] Z Jinfeng A Daniel S Lv and K Majid ldquoSmall hydropowerby using pump as turbine for power generation in SouthernAfricardquo Current Science vol 118 no 3 2019

Complexity 9

42 Performance of PAT Predicted Based on PumpPerformance PATsimulation results acquired from ANSYSCFX was used to evaluate the performance parameters of thePAT It is essential to recognise these parameters for PATperformance under different flow rate conditions (eyconsist of the head power and nondimensionless coeffi-cients such as head flow and power coefficient Pumpmodedata can predict PAT performance by using the pump BEP(is section will determine the PAT performance using theBEP of the pump and compare anticipated results with theCFD simulation result (e performance parameters of PATcould be predicted by considering pumpmode performance(e public affinity laws are used to establish relationships

between the performance parameters of similar machinesoperating under similar kinematic conditions It is essentialto know the following dimensionless parameters which arevital in determining the turbinersquos performance character-istics(ey consist of a head flow and power coefficient(esimilarities between these parameters are that the variableswith the particular coefficient are made dimensionlessthrough normalising with a diameter (D) and rotation speed(N) Except for the flow coefficient previous and latterparameters call for one additional variable to normalise (edimensionless coefficients result realised from the PATnumerical simulation has also been plotted as displayed inFigure 5 showing the head coefficient flow coefficient andpower coefficient (ese results show the differences whichexist among these coefficients (us Figure 6 indicates thedifference in the head coefficient at different flow coefficientlevels and different rotational speeds It can be witnessedfrom the results that the increase in the head coefficientincreases the flow coefficient but this occurs at a differentrotational speed thus increase in rotational speed decreasesthe head coefficient as shown in Figure 7 Figure 8 shows thedifference in the flow coefficient via the power coefficientOne can observe from the figure that the power coefficientperforms comparably with the PAT flow coefficient underdifferent flow rate conditions It is characterised by reducingthe value as the flow coefficient drops Figure 9 shows thehead coefficient vs flow coefficient curves of a selected PATat a different flow rates

(e nondimensional parameters at various rotationalspeeds have been studied (let us say ϕ 08) as tabulated inTable 3 (e study discovered that an increase in speed in-creases power coefficient (π) and efficiency η Neverthelesstheir differences in ψ were not much except at higher speeds(e results are as shown in the curves below with the BEPpoints at different rotational speeds also indicated (e poweroutput and efficiency were found better in the speed range of1000ndash1200 rpm As PAT operates in the opposite direction ofpumpmode operation the energy converted from hydraulic tomechanical might not be effectual at higher speeds causing itsbetter performance at the lower speed level

43 Transient Static Pressure Supply under Diverse Flow Ratein Lieu of Centrifugal Pump Impeller Pressure distributionunder different flow conditions and rotational speed in pumpas turbine fluid domain are displayed in Figure 10 From theresults pressure progressively intensified in a streamwise waywith the pressure side recording the maximum Flow partingcould be observed at the impeller blade leading edge (e flowat the inlet is not at a tangent with a blade with uniformed flowobserved near the blade which causes flow parting to occur onthe blade surface Minimal pressure was found at the impellerinlet along the suction area representing location cavitationwhich usually occurs in pumps

It is observed from Figure 11 that static pressure wasasymmetrically disseminated (e minimum pressure hap-pens at the blade suction area Pressure increases as the flowrate reaches the lowest values while it decreases with in-creasing flow rates as shown in Figure 11(b) Comparing the

130

230

330

430

530

08 1 12 14

P(kW

)

Q (QD)

1000RPM1750RPM1500RPM

Figure 5 Power vs rotational speed for selected PAT

750 1000 1250 1500 1750 2000 2250 2500N=(RPM)

η=(

)

72

70

68

66

64

62

60

Nrunaway

NBE

P

Figure 6 Power vs efficiency for selected PAT at Qt 12m3s

750 1000 1250 1500 1750 2000 2250 2500

Plot Are

N=RPM

530

480

430

380

330

280

230

P(kw

)

Nru

naw

ay

NBE

P

Figure 4 Power vs rotational speed for selected PAT atQt 12m3s

4 Complexity

two selected rotational speeds under three flow conditionsfromQ 008m3s toQ 12m3s a rise in the flow rate alsobrought about an increase in drop closer to the exit duct ofthe volute Proper recirculation of flow could be observed atthe impeller suction area due to lower flow rates though theflow on the pressure side is smooth On the contrary as theflow rate increases flow parting along the pressure side

occurs bringing about continuous flow movement in thepressure side (is movement in the suction side of theimpeller blade drops as the flow rate rises

44 Transient Velocity Distribution in the PAT Model for aSelected Centrifugal Pump Hydrodynamics of flow in PATwas numerically calculated using CFD to find the supply ofactual parameters defining hydrodynamics flow in thepump which interconnects the PAT characteristicsrsquo geo-metric model Figure 12 shows the velocity in the PATcolored by relative velocity distribution Considering thevolute shape one could observe the velocity increase with arise flow rate Flow thenmoves smoothly in the volute regionat a minimal flow rate Moreover it slows down after strikingthe trailing edge of the impeller after which relative fluidvelocity begins to once again increase to some extent (eCFD model shows a trend of velocity differences withchanges in rotational speed It was also observed that aswater enters the suction side relative velocity intensifiednext to the curving of the blade in the streamwise directionas the passage bit by bit gets smaller In summary a portionof the simulation was performed by the PAT impeller as afeedback turbine as a result of reaction from the pressuredrop with a change in kinetic energy also playing a role

45 Transient Static Pressure Distribution for CentrifugalPump Impeller In analysing flow features of the centrifugalpump in pump mode pressure and velocity contour plotsand vectors are employed (e streamlines of the 3D modelsare plotted to examine the flow design in a centrifugal pumpFigure 13 displays the static pressure transfer of the des-ignated PAT model under a diverse flow rate atn 1750 rpm(e results show that lower static pressure wasobserved at the blade leading edge As kinetic energy flowedand transported to fluid the static pressure progressivelyincreases (e figure shows that maximum pressure wasmonitored on the blade side whereas minimal pressure wasalso detected on the blade suction side at the same radialposition As the fluid leads it is a way to the volute casing(e static pressure further increased as shown in Figure 13which could be attributed to the conversion of the fluidrsquoskinetic energy to pressure energy because of the increasingcasing cross-segment region Figure 14 also displays thevelocity delivery of the designated PATmodel under diverseflow rate n 1750 rpm

Table 3 Parameters at ϕ 08 at different rotational speeds for theoriginal impeller

N (rpm) ψ π ϕ η ()1250 88 016 771 0002021500 71 007 645 0002071750 66 008 553 0001842000 64 004 484 000106

65697377818589

1250 1500 1750 2000Rotational speed=(RPM)

Hea

d co

effic

ient

(ψ)

Figure 7 Rotational speed vs head coefficient characteristics curve of selected PAT at Qt 08m3s

00150018002100240027

0030033

00002 00004 00006 00008 0001 00012 00014 00016

Hea

d co

effic

ient

(ψ)

Power coefficient (π)

Figure 8 Power coefficient vs flow coefficient curve of selectedPAT at Qt 08m3s and n 1500 rpm

5

7

9

11

13

06 08 1 12

Hea

d C

oeffi

cien

t (ψ)

Q (m3h)

150017502000

Figure 9 Head coefficient vs flow coefficient curves of a selectedPAT at a different flow rates

Complexity 5

Pressure1877e+005

1780e+005

1683e+005

1585e+005

1488e+005

1391e+005

1294e+005

1197e+005

1100e+005

1003e+005

9060e+004[Pa]

Q=08 m2s Q=10 m2s Q=12 m2s

(a)

Pressure

[Pa]

1251e+005

1128e+005

1005e+005

8827e+004

7599e+004

6372e+004

5144e+004

3917e+004

2689e+004

1462e+004

2345e+003 Q=08 m2s Q=10 m2s Q=12 m2s

(b)

Figure 11 (a) 3D pressure dissemination for the PATmodel under the dissimilar flow rate n 1750 rpm (b) 3D pressure dissemination forthe PAT model under different flow rates n 1500 rpm

Total Pressure Total Pressure1294e+004-2749e+002-1349e+004-2670e+004-3991e+004-5312e+004-6633e+004-7954e+004-9275e+004-1060e+005-1192e+005

1870e+0044273e+003-1015e+004-2457e+004-3900e+004-5342e+004-6785e+004-8227e+004-9669e+004-1111e+005-1255e+005

[Pa] [Pa]Q=08m3s at n=1750rpm Q=10m3s n=1750rpm

Figure 10 Total pressure distribution on the PAT impeller at a different flow rate

6 Complexity

1734e+001

1561e+001

1387e+001

1214e+001

1041e+001

8671e+000

6937e+000

5203e+000

3469e+000

1734e+000

0000e+000[m s^-1]

Velocity in Stn Frame

Q=01m3s Q=10m3s Q=12m3s

Figure 12 Absolute velocity delivery of the designated PAT model under diverse flow rate n 1750 rpm

2400e+005

2240e+005

2080e+005

1920e+005

1760e+005

1600e+005

1440e+005

1280e+005

1120e+005

9600e+004

8000e+004[Pa]

Pressure

Q=08m3s Q=10m3s Q=12m3s Q=14m3s

Figure 13 Static pressure delivery of the designated PAT model under diverse flow rate n 1750 rpm

1101e+0019910e+0008809e+0007708e+0006607e+0005506e+0004404e+0003303e+0002202e+0001101e+0000000e+000

[m s^-1]

Velocity

Q=08m3s Q=10m3s Q=12m3s Q=14m3s

Figure 14 Velocity distribution for the selected centrifugal pump model at different flow rates N 1750 rpm

Complexity 7

5 Conclusion

In this study the comparison of numerical studies resultsobtained under different flow and speed conditions wasinvestigated From the study the following conclusions weremade

(A) (e speed of a turbine is determined by the amountof load used At constant flow rate and head theturbine speed decreases when the turbine uses ahigher load more than the design load with thedecrease of the burden placed on a turbine whereasthe head and flow rate remain constant the turbinespeed increases (e same principle applies whenhydraulic pumps are used as a turbine at a specifiedsite with a continuous flow rate and contributesignificantly to determining the turbinersquos perfor-mance characteristics As the PAT load is furtherreduced the PATrsquos speed reaches a maximum valuethat the PAT can deliver no torque (is speed iscalled runaway speed (e runaway speed of thePAT was obtained to be approximately 2250 rpm

(B) Increase in PATrsquos efficiency with the increased ro-tational speed of the impeller till it reaches 1500 rpmshould also be mentioned It means that when PAToperates beyond 1500 rpm the efficiency of the PATwill decrease In comparison with Figures 6 and 7PAT efficiency behaves equally as the power outputof the PAT (erefore it can be concluded that thebest-operating conditions for selected PAT could beachieved when it operates between 1000 rpm and1750 rpm

(C) Pressure distribution under different flow condi-tions and rotational speed in the pump as turbinefluid domain From the results pressure progres-sively intensified in a streamwise way with thepressure side recording the maximum Flow partingcould be observed at the impeller blade leading edge(e flow at the inlet is not at a tangent with the bladewith ununiformed flow observed near the bladewhich causes flow parting to occur on the bladesurface Minimal pressure was found at the impellerinlet along the suction area representing locationcavitation typically occurs in pumps

In summary PAT simulation results acquired fromANSYS CFX have been used to evaluate the PATrsquos per-formance parameters It is essential to recognise these pa-rameters for PAT performance under different flow rateconditions (ey consist of the head power and non-dimensionless coefficients such as head flow and powercoefficient Pump mode data can predict PAT performanceby using the pump BEP

Data Availability

(e data are openly available in a public repository thatissues datasets with DOIs (e authors confirm that the datasupporting the findings of this study are available within thearticle and could be obtained upon request (e data

generated at a central large-scale facility are available uponrequest Raw data were generated at Jiangsu University Lab(e derived data supporting the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

(e authors declare that they have no conflicts of interest

Acknowledgments

(is study was supported by the National Science Foun-dation of China under Grant nos 71974081 71704066 and71971100 and special funds of the National Social ScienceFund of China (Grant no 18VSJ038)

References

[1] J Gonzalez and C Santolaria ldquoUnsteady flow structure andglobal variables in a centrifugal pumprdquo Journal of FluidsEngineering vol 128 no 5 pp 937ndash946 2006

[2] B K Gandhi S N Singh and V Seshadri ldquoPerformancecharacteristics of centrifugal slurry pumpsrdquo Journal of FluidsEngineering vol 123 no 2 pp 271ndash280 2001

[3] H Nautiyal A Varun and A Kumar ldquoReverse runningpumps analytical experimental and computational study areviewrdquo Renewable and Sustainable Energy Reviews vol 14no 7 pp 2059ndash2067 2010

[4] S V Jain and R N Patel ldquoInvestigations on pump running inturbine mode a review of the state-of-the-artrdquo Renewable andSustainable Energy Reviews vol 30 pp 841ndash868 2014

[5] H Nautiyal V Varun A Kumar and S Y S Yadav ldquoEx-perimental investigation of centrifugal pump working asturbine for small hydropower systemsrdquo Energy Science andTechnology vol 1 pp 79ndash86 2011

[6] A J Stepanoff Centrifugal and Axial Flow Pumps Design andApplications John Wiley amp Sons New York NY USA 1957

[7] S Childs ldquoConvert pumps to turbines and recovers HPrdquoHydrocarbon Processing and Petroleum Refiner vol 41pp 173-174 1962

[8] R K Sharma Small Hydroelectric ProjectsmdashUse of CentrifugalPumps as Turbines Kirloskar Electric Co Bangalore India1984

[9] S Gopalakrishnan ldquoPower recovery turbines for the processindustryrdquo in Proceedings of the ltird International PumpSymposium pp 3ndash11 Texas AampM University Houston TXUSA September 1986

[10] H Diederich ldquoVerwendung von kreiselpumpen als turbinenrdquoKSB Techn Ber vol 12 pp 30ndash36 1967

[11] K Grover Conversion of Pumps to Turbines GSA Intercorporate Katonah NY USA 1982

[12] A Morabito and P Hendrick ldquoPump as turbine applied tomicro energy storage and smart water grids a case studyrdquoApplied Energy vol 241 pp 567ndash579 2019

[13] M Rossi A Nigro and M Renzi ldquoExperimental and nu-merical assessment of a methodology for performance pre-diction of pumps-as-turbines (PaTs) operating in off-designconditionsrdquo Applied Energy vol 248 pp 555ndash566 2019

[14] B Luo and X C L Wang ldquoNumerical simulation of flow-induced vibration of double-suction centrifugal pump asturbinerdquo Journal Drainage Irrigation Mechanical Engineeringvol 37 pp 313ndash318 2019

8 Complexity

[15] I M Hossain S Ferdous and T Jamal ldquoPump-as-turbine(PAT) for small-scale power generation a comparativeanalysisrdquo in Proceedings of the 2014 3rd International Con-ference on the IEEE Developments in Renewable EnergyTechnology (ICDRET) pp 1ndash5 Dhaka Bangladesh May 2014

[16] J-M Chapallaz P Eichenberger and G Fischer ldquoManual onpumps used as turbinesrdquo Vieweg vol 11 1992

[17] S Derakhshan and A Nourbakhsh ldquo(eoretical numericaland experimental investigation of centrifugal pumps in re-verse operationrdquo Experimental ltermal and Fluid Sciencevol 32 no 8 pp 1620ndash1627 2008

[18] P Vasanthakumar J Krishnaraj S Karthikayan T Vinothand S K Arun Sankar ldquoInvestigation on reverse character-istics of centrifugal pump in turbine mode a comparativestudy by an experimentation and simulationrdquo MaterialsToday Proceedings vol 4 no 2 pp 693ndash700 2017

[19] M Gantar ldquoPropeller pumps running as turbinesrdquo in Pro-ceedings of the Conference on Hydraulic Machinery pp 237ndash248 Ljubljana Slovenia August 1988

[20] H H Anderson ldquoBorehole pumpsrdquo Centrifugal Pumps andAllied Machinery vol 4 pp 177ndash185 1993

[21] D Adu J Zhang M Jieyun S N Asomani and M OsmanKoranteng ldquoNumerical investigation of transient vortices andturbulent flow behaviour in centrifugal pump operating inreverse mode as turbinerdquo Materials Science for EnergyTechnologies vol 2 no 2 pp 356ndash364 2019

[22] H Wang Y Wang C Fan et al ldquoMaterial consumption andcarbon emissions associated with the infrastructure con-struction of 34 cities in northeast Chinardquo Complexityvol 2020 Article ID 4364912 20 pages 2020

[23] Z Cao M Liu and P Wu ldquoExperiment investigation andnumerical simulation of Snowdrift on a typical large-spanretractable roofrdquo Complexity vol 2019 Article ID 598480414 pages 2019

[24] K Ren H Li S Li and H Dong ldquoVoltage stability analysis offront-end speed controlled wind turbine integrated into re-gional power grid based on bifurcation theoryrdquo Complexityvol 2020 Article ID 8816334 11 pages 2020

[25] D Florez-Orrego I B Henriques T-V Nguyen et al ldquo(econtributions of prof Jan Szargut to the exergy and envi-ronmental assessment of complex energy systemsrdquo Energyvol 161 pp 482ndash492 2018

[26] Y Zhang Y Xu Y Zheng et al ldquoMultiobjective optimizationdesign and experimental investigation on the axial-flow pumpwith orthogonal test approachrdquo Complexity vol 2019 ArticleID 1467565 14 pages 2019

[27] X Xu X Yuan M Cui Y Yuan and L Qi ldquoJoint optimi-zation of energy conservation and migration cost for complexsystems in edge computingrdquo Complexity vol 2019 Article ID6180135 14 pages 2019

[28] Y L Zhang S Q Yuan Z Jinfeng et al ldquoNumerical in-vestigation of the effects of splitter blades on the cavitationperformance of a centrifugal pumprdquo IOP Conference SeriesEarth and Environmental Science vol 22 Article ID 0520032014

[29] Z Jinfeng A Daniel S Lv and K Majid ldquoSmall hydropowerby using pump as turbine for power generation in SouthernAfricardquo Current Science vol 118 no 3 2019

Complexity 9

two selected rotational speeds under three flow conditionsfromQ 008m3s toQ 12m3s a rise in the flow rate alsobrought about an increase in drop closer to the exit duct ofthe volute Proper recirculation of flow could be observed atthe impeller suction area due to lower flow rates though theflow on the pressure side is smooth On the contrary as theflow rate increases flow parting along the pressure side

occurs bringing about continuous flow movement in thepressure side (is movement in the suction side of theimpeller blade drops as the flow rate rises

44 Transient Velocity Distribution in the PAT Model for aSelected Centrifugal Pump Hydrodynamics of flow in PATwas numerically calculated using CFD to find the supply ofactual parameters defining hydrodynamics flow in thepump which interconnects the PAT characteristicsrsquo geo-metric model Figure 12 shows the velocity in the PATcolored by relative velocity distribution Considering thevolute shape one could observe the velocity increase with arise flow rate Flow thenmoves smoothly in the volute regionat a minimal flow rate Moreover it slows down after strikingthe trailing edge of the impeller after which relative fluidvelocity begins to once again increase to some extent (eCFD model shows a trend of velocity differences withchanges in rotational speed It was also observed that aswater enters the suction side relative velocity intensifiednext to the curving of the blade in the streamwise directionas the passage bit by bit gets smaller In summary a portionof the simulation was performed by the PAT impeller as afeedback turbine as a result of reaction from the pressuredrop with a change in kinetic energy also playing a role

45 Transient Static Pressure Distribution for CentrifugalPump Impeller In analysing flow features of the centrifugalpump in pump mode pressure and velocity contour plotsand vectors are employed (e streamlines of the 3D modelsare plotted to examine the flow design in a centrifugal pumpFigure 13 displays the static pressure transfer of the des-ignated PAT model under a diverse flow rate atn 1750 rpm(e results show that lower static pressure wasobserved at the blade leading edge As kinetic energy flowedand transported to fluid the static pressure progressivelyincreases (e figure shows that maximum pressure wasmonitored on the blade side whereas minimal pressure wasalso detected on the blade suction side at the same radialposition As the fluid leads it is a way to the volute casing(e static pressure further increased as shown in Figure 13which could be attributed to the conversion of the fluidrsquoskinetic energy to pressure energy because of the increasingcasing cross-segment region Figure 14 also displays thevelocity delivery of the designated PATmodel under diverseflow rate n 1750 rpm

Table 3 Parameters at ϕ 08 at different rotational speeds for theoriginal impeller

N (rpm) ψ π ϕ η ()1250 88 016 771 0002021500 71 007 645 0002071750 66 008 553 0001842000 64 004 484 000106

65697377818589

1250 1500 1750 2000Rotational speed=(RPM)

Hea

d co

effic

ient

(ψ)

Figure 7 Rotational speed vs head coefficient characteristics curve of selected PAT at Qt 08m3s

00150018002100240027

0030033

00002 00004 00006 00008 0001 00012 00014 00016

Hea

d co

effic

ient

(ψ)

Power coefficient (π)

Figure 8 Power coefficient vs flow coefficient curve of selectedPAT at Qt 08m3s and n 1500 rpm

5

7

9

11

13

06 08 1 12

Hea

d C

oeffi

cien

t (ψ)

Q (m3h)

150017502000

Figure 9 Head coefficient vs flow coefficient curves of a selectedPAT at a different flow rates

Complexity 5

Pressure1877e+005

1780e+005

1683e+005

1585e+005

1488e+005

1391e+005

1294e+005

1197e+005

1100e+005

1003e+005

9060e+004[Pa]

Q=08 m2s Q=10 m2s Q=12 m2s

(a)

Pressure

[Pa]

1251e+005

1128e+005

1005e+005

8827e+004

7599e+004

6372e+004

5144e+004

3917e+004

2689e+004

1462e+004

2345e+003 Q=08 m2s Q=10 m2s Q=12 m2s

(b)

Figure 11 (a) 3D pressure dissemination for the PATmodel under the dissimilar flow rate n 1750 rpm (b) 3D pressure dissemination forthe PAT model under different flow rates n 1500 rpm

Total Pressure Total Pressure1294e+004-2749e+002-1349e+004-2670e+004-3991e+004-5312e+004-6633e+004-7954e+004-9275e+004-1060e+005-1192e+005

1870e+0044273e+003-1015e+004-2457e+004-3900e+004-5342e+004-6785e+004-8227e+004-9669e+004-1111e+005-1255e+005

[Pa] [Pa]Q=08m3s at n=1750rpm Q=10m3s n=1750rpm

Figure 10 Total pressure distribution on the PAT impeller at a different flow rate

6 Complexity

1734e+001

1561e+001

1387e+001

1214e+001

1041e+001

8671e+000

6937e+000

5203e+000

3469e+000

1734e+000

0000e+000[m s^-1]

Velocity in Stn Frame

Q=01m3s Q=10m3s Q=12m3s

Figure 12 Absolute velocity delivery of the designated PAT model under diverse flow rate n 1750 rpm

2400e+005

2240e+005

2080e+005

1920e+005

1760e+005

1600e+005

1440e+005

1280e+005

1120e+005

9600e+004

8000e+004[Pa]

Pressure

Q=08m3s Q=10m3s Q=12m3s Q=14m3s

Figure 13 Static pressure delivery of the designated PAT model under diverse flow rate n 1750 rpm

1101e+0019910e+0008809e+0007708e+0006607e+0005506e+0004404e+0003303e+0002202e+0001101e+0000000e+000

[m s^-1]

Velocity

Q=08m3s Q=10m3s Q=12m3s Q=14m3s

Figure 14 Velocity distribution for the selected centrifugal pump model at different flow rates N 1750 rpm

Complexity 7

5 Conclusion

In this study the comparison of numerical studies resultsobtained under different flow and speed conditions wasinvestigated From the study the following conclusions weremade

(A) (e speed of a turbine is determined by the amountof load used At constant flow rate and head theturbine speed decreases when the turbine uses ahigher load more than the design load with thedecrease of the burden placed on a turbine whereasthe head and flow rate remain constant the turbinespeed increases (e same principle applies whenhydraulic pumps are used as a turbine at a specifiedsite with a continuous flow rate and contributesignificantly to determining the turbinersquos perfor-mance characteristics As the PAT load is furtherreduced the PATrsquos speed reaches a maximum valuethat the PAT can deliver no torque (is speed iscalled runaway speed (e runaway speed of thePAT was obtained to be approximately 2250 rpm

(B) Increase in PATrsquos efficiency with the increased ro-tational speed of the impeller till it reaches 1500 rpmshould also be mentioned It means that when PAToperates beyond 1500 rpm the efficiency of the PATwill decrease In comparison with Figures 6 and 7PAT efficiency behaves equally as the power outputof the PAT (erefore it can be concluded that thebest-operating conditions for selected PAT could beachieved when it operates between 1000 rpm and1750 rpm

(C) Pressure distribution under different flow condi-tions and rotational speed in the pump as turbinefluid domain From the results pressure progres-sively intensified in a streamwise way with thepressure side recording the maximum Flow partingcould be observed at the impeller blade leading edge(e flow at the inlet is not at a tangent with the bladewith ununiformed flow observed near the bladewhich causes flow parting to occur on the bladesurface Minimal pressure was found at the impellerinlet along the suction area representing locationcavitation typically occurs in pumps

In summary PAT simulation results acquired fromANSYS CFX have been used to evaluate the PATrsquos per-formance parameters It is essential to recognise these pa-rameters for PAT performance under different flow rateconditions (ey consist of the head power and non-dimensionless coefficients such as head flow and powercoefficient Pump mode data can predict PAT performanceby using the pump BEP

Data Availability

(e data are openly available in a public repository thatissues datasets with DOIs (e authors confirm that the datasupporting the findings of this study are available within thearticle and could be obtained upon request (e data

generated at a central large-scale facility are available uponrequest Raw data were generated at Jiangsu University Lab(e derived data supporting the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

(e authors declare that they have no conflicts of interest

Acknowledgments

(is study was supported by the National Science Foun-dation of China under Grant nos 71974081 71704066 and71971100 and special funds of the National Social ScienceFund of China (Grant no 18VSJ038)

References

[1] J Gonzalez and C Santolaria ldquoUnsteady flow structure andglobal variables in a centrifugal pumprdquo Journal of FluidsEngineering vol 128 no 5 pp 937ndash946 2006

[2] B K Gandhi S N Singh and V Seshadri ldquoPerformancecharacteristics of centrifugal slurry pumpsrdquo Journal of FluidsEngineering vol 123 no 2 pp 271ndash280 2001

[3] H Nautiyal A Varun and A Kumar ldquoReverse runningpumps analytical experimental and computational study areviewrdquo Renewable and Sustainable Energy Reviews vol 14no 7 pp 2059ndash2067 2010

[4] S V Jain and R N Patel ldquoInvestigations on pump running inturbine mode a review of the state-of-the-artrdquo Renewable andSustainable Energy Reviews vol 30 pp 841ndash868 2014

[5] H Nautiyal V Varun A Kumar and S Y S Yadav ldquoEx-perimental investigation of centrifugal pump working asturbine for small hydropower systemsrdquo Energy Science andTechnology vol 1 pp 79ndash86 2011

[6] A J Stepanoff Centrifugal and Axial Flow Pumps Design andApplications John Wiley amp Sons New York NY USA 1957

[7] S Childs ldquoConvert pumps to turbines and recovers HPrdquoHydrocarbon Processing and Petroleum Refiner vol 41pp 173-174 1962

[8] R K Sharma Small Hydroelectric ProjectsmdashUse of CentrifugalPumps as Turbines Kirloskar Electric Co Bangalore India1984

[9] S Gopalakrishnan ldquoPower recovery turbines for the processindustryrdquo in Proceedings of the ltird International PumpSymposium pp 3ndash11 Texas AampM University Houston TXUSA September 1986

[10] H Diederich ldquoVerwendung von kreiselpumpen als turbinenrdquoKSB Techn Ber vol 12 pp 30ndash36 1967

[11] K Grover Conversion of Pumps to Turbines GSA Intercorporate Katonah NY USA 1982

[12] A Morabito and P Hendrick ldquoPump as turbine applied tomicro energy storage and smart water grids a case studyrdquoApplied Energy vol 241 pp 567ndash579 2019

[13] M Rossi A Nigro and M Renzi ldquoExperimental and nu-merical assessment of a methodology for performance pre-diction of pumps-as-turbines (PaTs) operating in off-designconditionsrdquo Applied Energy vol 248 pp 555ndash566 2019

[14] B Luo and X C L Wang ldquoNumerical simulation of flow-induced vibration of double-suction centrifugal pump asturbinerdquo Journal Drainage Irrigation Mechanical Engineeringvol 37 pp 313ndash318 2019

8 Complexity

[15] I M Hossain S Ferdous and T Jamal ldquoPump-as-turbine(PAT) for small-scale power generation a comparativeanalysisrdquo in Proceedings of the 2014 3rd International Con-ference on the IEEE Developments in Renewable EnergyTechnology (ICDRET) pp 1ndash5 Dhaka Bangladesh May 2014

[16] J-M Chapallaz P Eichenberger and G Fischer ldquoManual onpumps used as turbinesrdquo Vieweg vol 11 1992

[17] S Derakhshan and A Nourbakhsh ldquo(eoretical numericaland experimental investigation of centrifugal pumps in re-verse operationrdquo Experimental ltermal and Fluid Sciencevol 32 no 8 pp 1620ndash1627 2008

[18] P Vasanthakumar J Krishnaraj S Karthikayan T Vinothand S K Arun Sankar ldquoInvestigation on reverse character-istics of centrifugal pump in turbine mode a comparativestudy by an experimentation and simulationrdquo MaterialsToday Proceedings vol 4 no 2 pp 693ndash700 2017

[19] M Gantar ldquoPropeller pumps running as turbinesrdquo in Pro-ceedings of the Conference on Hydraulic Machinery pp 237ndash248 Ljubljana Slovenia August 1988

[20] H H Anderson ldquoBorehole pumpsrdquo Centrifugal Pumps andAllied Machinery vol 4 pp 177ndash185 1993

[21] D Adu J Zhang M Jieyun S N Asomani and M OsmanKoranteng ldquoNumerical investigation of transient vortices andturbulent flow behaviour in centrifugal pump operating inreverse mode as turbinerdquo Materials Science for EnergyTechnologies vol 2 no 2 pp 356ndash364 2019

[22] H Wang Y Wang C Fan et al ldquoMaterial consumption andcarbon emissions associated with the infrastructure con-struction of 34 cities in northeast Chinardquo Complexityvol 2020 Article ID 4364912 20 pages 2020

[23] Z Cao M Liu and P Wu ldquoExperiment investigation andnumerical simulation of Snowdrift on a typical large-spanretractable roofrdquo Complexity vol 2019 Article ID 598480414 pages 2019

[24] K Ren H Li S Li and H Dong ldquoVoltage stability analysis offront-end speed controlled wind turbine integrated into re-gional power grid based on bifurcation theoryrdquo Complexityvol 2020 Article ID 8816334 11 pages 2020

[25] D Florez-Orrego I B Henriques T-V Nguyen et al ldquo(econtributions of prof Jan Szargut to the exergy and envi-ronmental assessment of complex energy systemsrdquo Energyvol 161 pp 482ndash492 2018

[26] Y Zhang Y Xu Y Zheng et al ldquoMultiobjective optimizationdesign and experimental investigation on the axial-flow pumpwith orthogonal test approachrdquo Complexity vol 2019 ArticleID 1467565 14 pages 2019

[27] X Xu X Yuan M Cui Y Yuan and L Qi ldquoJoint optimi-zation of energy conservation and migration cost for complexsystems in edge computingrdquo Complexity vol 2019 Article ID6180135 14 pages 2019

[28] Y L Zhang S Q Yuan Z Jinfeng et al ldquoNumerical in-vestigation of the effects of splitter blades on the cavitationperformance of a centrifugal pumprdquo IOP Conference SeriesEarth and Environmental Science vol 22 Article ID 0520032014

[29] Z Jinfeng A Daniel S Lv and K Majid ldquoSmall hydropowerby using pump as turbine for power generation in SouthernAfricardquo Current Science vol 118 no 3 2019

Complexity 9

Pressure1877e+005

1780e+005

1683e+005

1585e+005

1488e+005

1391e+005

1294e+005

1197e+005

1100e+005

1003e+005

9060e+004[Pa]

Q=08 m2s Q=10 m2s Q=12 m2s

(a)

Pressure

[Pa]

1251e+005

1128e+005

1005e+005

8827e+004

7599e+004

6372e+004

5144e+004

3917e+004

2689e+004

1462e+004

2345e+003 Q=08 m2s Q=10 m2s Q=12 m2s

(b)

Figure 11 (a) 3D pressure dissemination for the PATmodel under the dissimilar flow rate n 1750 rpm (b) 3D pressure dissemination forthe PAT model under different flow rates n 1500 rpm

Total Pressure Total Pressure1294e+004-2749e+002-1349e+004-2670e+004-3991e+004-5312e+004-6633e+004-7954e+004-9275e+004-1060e+005-1192e+005

1870e+0044273e+003-1015e+004-2457e+004-3900e+004-5342e+004-6785e+004-8227e+004-9669e+004-1111e+005-1255e+005

[Pa] [Pa]Q=08m3s at n=1750rpm Q=10m3s n=1750rpm

Figure 10 Total pressure distribution on the PAT impeller at a different flow rate

6 Complexity

1734e+001

1561e+001

1387e+001

1214e+001

1041e+001

8671e+000

6937e+000

5203e+000

3469e+000

1734e+000

0000e+000[m s^-1]

Velocity in Stn Frame

Q=01m3s Q=10m3s Q=12m3s

Figure 12 Absolute velocity delivery of the designated PAT model under diverse flow rate n 1750 rpm

2400e+005

2240e+005

2080e+005

1920e+005

1760e+005

1600e+005

1440e+005

1280e+005

1120e+005

9600e+004

8000e+004[Pa]

Pressure

Q=08m3s Q=10m3s Q=12m3s Q=14m3s

Figure 13 Static pressure delivery of the designated PAT model under diverse flow rate n 1750 rpm

1101e+0019910e+0008809e+0007708e+0006607e+0005506e+0004404e+0003303e+0002202e+0001101e+0000000e+000

[m s^-1]

Velocity

Q=08m3s Q=10m3s Q=12m3s Q=14m3s

Figure 14 Velocity distribution for the selected centrifugal pump model at different flow rates N 1750 rpm

Complexity 7

5 Conclusion

In this study the comparison of numerical studies resultsobtained under different flow and speed conditions wasinvestigated From the study the following conclusions weremade

(A) (e speed of a turbine is determined by the amountof load used At constant flow rate and head theturbine speed decreases when the turbine uses ahigher load more than the design load with thedecrease of the burden placed on a turbine whereasthe head and flow rate remain constant the turbinespeed increases (e same principle applies whenhydraulic pumps are used as a turbine at a specifiedsite with a continuous flow rate and contributesignificantly to determining the turbinersquos perfor-mance characteristics As the PAT load is furtherreduced the PATrsquos speed reaches a maximum valuethat the PAT can deliver no torque (is speed iscalled runaway speed (e runaway speed of thePAT was obtained to be approximately 2250 rpm

(B) Increase in PATrsquos efficiency with the increased ro-tational speed of the impeller till it reaches 1500 rpmshould also be mentioned It means that when PAToperates beyond 1500 rpm the efficiency of the PATwill decrease In comparison with Figures 6 and 7PAT efficiency behaves equally as the power outputof the PAT (erefore it can be concluded that thebest-operating conditions for selected PAT could beachieved when it operates between 1000 rpm and1750 rpm

(C) Pressure distribution under different flow condi-tions and rotational speed in the pump as turbinefluid domain From the results pressure progres-sively intensified in a streamwise way with thepressure side recording the maximum Flow partingcould be observed at the impeller blade leading edge(e flow at the inlet is not at a tangent with the bladewith ununiformed flow observed near the bladewhich causes flow parting to occur on the bladesurface Minimal pressure was found at the impellerinlet along the suction area representing locationcavitation typically occurs in pumps

In summary PAT simulation results acquired fromANSYS CFX have been used to evaluate the PATrsquos per-formance parameters It is essential to recognise these pa-rameters for PAT performance under different flow rateconditions (ey consist of the head power and non-dimensionless coefficients such as head flow and powercoefficient Pump mode data can predict PAT performanceby using the pump BEP

Data Availability

(e data are openly available in a public repository thatissues datasets with DOIs (e authors confirm that the datasupporting the findings of this study are available within thearticle and could be obtained upon request (e data

generated at a central large-scale facility are available uponrequest Raw data were generated at Jiangsu University Lab(e derived data supporting the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

(e authors declare that they have no conflicts of interest

Acknowledgments

(is study was supported by the National Science Foun-dation of China under Grant nos 71974081 71704066 and71971100 and special funds of the National Social ScienceFund of China (Grant no 18VSJ038)

References

[1] J Gonzalez and C Santolaria ldquoUnsteady flow structure andglobal variables in a centrifugal pumprdquo Journal of FluidsEngineering vol 128 no 5 pp 937ndash946 2006

[2] B K Gandhi S N Singh and V Seshadri ldquoPerformancecharacteristics of centrifugal slurry pumpsrdquo Journal of FluidsEngineering vol 123 no 2 pp 271ndash280 2001

[3] H Nautiyal A Varun and A Kumar ldquoReverse runningpumps analytical experimental and computational study areviewrdquo Renewable and Sustainable Energy Reviews vol 14no 7 pp 2059ndash2067 2010

[4] S V Jain and R N Patel ldquoInvestigations on pump running inturbine mode a review of the state-of-the-artrdquo Renewable andSustainable Energy Reviews vol 30 pp 841ndash868 2014

[5] H Nautiyal V Varun A Kumar and S Y S Yadav ldquoEx-perimental investigation of centrifugal pump working asturbine for small hydropower systemsrdquo Energy Science andTechnology vol 1 pp 79ndash86 2011

[6] A J Stepanoff Centrifugal and Axial Flow Pumps Design andApplications John Wiley amp Sons New York NY USA 1957

[7] S Childs ldquoConvert pumps to turbines and recovers HPrdquoHydrocarbon Processing and Petroleum Refiner vol 41pp 173-174 1962

[8] R K Sharma Small Hydroelectric ProjectsmdashUse of CentrifugalPumps as Turbines Kirloskar Electric Co Bangalore India1984

[9] S Gopalakrishnan ldquoPower recovery turbines for the processindustryrdquo in Proceedings of the ltird International PumpSymposium pp 3ndash11 Texas AampM University Houston TXUSA September 1986

[10] H Diederich ldquoVerwendung von kreiselpumpen als turbinenrdquoKSB Techn Ber vol 12 pp 30ndash36 1967

[11] K Grover Conversion of Pumps to Turbines GSA Intercorporate Katonah NY USA 1982

[12] A Morabito and P Hendrick ldquoPump as turbine applied tomicro energy storage and smart water grids a case studyrdquoApplied Energy vol 241 pp 567ndash579 2019

[13] M Rossi A Nigro and M Renzi ldquoExperimental and nu-merical assessment of a methodology for performance pre-diction of pumps-as-turbines (PaTs) operating in off-designconditionsrdquo Applied Energy vol 248 pp 555ndash566 2019

[14] B Luo and X C L Wang ldquoNumerical simulation of flow-induced vibration of double-suction centrifugal pump asturbinerdquo Journal Drainage Irrigation Mechanical Engineeringvol 37 pp 313ndash318 2019

8 Complexity

[15] I M Hossain S Ferdous and T Jamal ldquoPump-as-turbine(PAT) for small-scale power generation a comparativeanalysisrdquo in Proceedings of the 2014 3rd International Con-ference on the IEEE Developments in Renewable EnergyTechnology (ICDRET) pp 1ndash5 Dhaka Bangladesh May 2014

[16] J-M Chapallaz P Eichenberger and G Fischer ldquoManual onpumps used as turbinesrdquo Vieweg vol 11 1992

[17] S Derakhshan and A Nourbakhsh ldquo(eoretical numericaland experimental investigation of centrifugal pumps in re-verse operationrdquo Experimental ltermal and Fluid Sciencevol 32 no 8 pp 1620ndash1627 2008

[18] P Vasanthakumar J Krishnaraj S Karthikayan T Vinothand S K Arun Sankar ldquoInvestigation on reverse character-istics of centrifugal pump in turbine mode a comparativestudy by an experimentation and simulationrdquo MaterialsToday Proceedings vol 4 no 2 pp 693ndash700 2017

[19] M Gantar ldquoPropeller pumps running as turbinesrdquo in Pro-ceedings of the Conference on Hydraulic Machinery pp 237ndash248 Ljubljana Slovenia August 1988

[20] H H Anderson ldquoBorehole pumpsrdquo Centrifugal Pumps andAllied Machinery vol 4 pp 177ndash185 1993

[21] D Adu J Zhang M Jieyun S N Asomani and M OsmanKoranteng ldquoNumerical investigation of transient vortices andturbulent flow behaviour in centrifugal pump operating inreverse mode as turbinerdquo Materials Science for EnergyTechnologies vol 2 no 2 pp 356ndash364 2019

[22] H Wang Y Wang C Fan et al ldquoMaterial consumption andcarbon emissions associated with the infrastructure con-struction of 34 cities in northeast Chinardquo Complexityvol 2020 Article ID 4364912 20 pages 2020

[23] Z Cao M Liu and P Wu ldquoExperiment investigation andnumerical simulation of Snowdrift on a typical large-spanretractable roofrdquo Complexity vol 2019 Article ID 598480414 pages 2019

[24] K Ren H Li S Li and H Dong ldquoVoltage stability analysis offront-end speed controlled wind turbine integrated into re-gional power grid based on bifurcation theoryrdquo Complexityvol 2020 Article ID 8816334 11 pages 2020

[25] D Florez-Orrego I B Henriques T-V Nguyen et al ldquo(econtributions of prof Jan Szargut to the exergy and envi-ronmental assessment of complex energy systemsrdquo Energyvol 161 pp 482ndash492 2018

[26] Y Zhang Y Xu Y Zheng et al ldquoMultiobjective optimizationdesign and experimental investigation on the axial-flow pumpwith orthogonal test approachrdquo Complexity vol 2019 ArticleID 1467565 14 pages 2019

[27] X Xu X Yuan M Cui Y Yuan and L Qi ldquoJoint optimi-zation of energy conservation and migration cost for complexsystems in edge computingrdquo Complexity vol 2019 Article ID6180135 14 pages 2019

[28] Y L Zhang S Q Yuan Z Jinfeng et al ldquoNumerical in-vestigation of the effects of splitter blades on the cavitationperformance of a centrifugal pumprdquo IOP Conference SeriesEarth and Environmental Science vol 22 Article ID 0520032014

[29] Z Jinfeng A Daniel S Lv and K Majid ldquoSmall hydropowerby using pump as turbine for power generation in SouthernAfricardquo Current Science vol 118 no 3 2019

Complexity 9

1734e+001

1561e+001

1387e+001

1214e+001

1041e+001

8671e+000

6937e+000

5203e+000

3469e+000

1734e+000

0000e+000[m s^-1]

Velocity in Stn Frame

Q=01m3s Q=10m3s Q=12m3s

Figure 12 Absolute velocity delivery of the designated PAT model under diverse flow rate n 1750 rpm

2400e+005

2240e+005

2080e+005

1920e+005

1760e+005

1600e+005

1440e+005

1280e+005

1120e+005

9600e+004

8000e+004[Pa]

Pressure

Q=08m3s Q=10m3s Q=12m3s Q=14m3s

Figure 13 Static pressure delivery of the designated PAT model under diverse flow rate n 1750 rpm

1101e+0019910e+0008809e+0007708e+0006607e+0005506e+0004404e+0003303e+0002202e+0001101e+0000000e+000

[m s^-1]

Velocity

Q=08m3s Q=10m3s Q=12m3s Q=14m3s

Figure 14 Velocity distribution for the selected centrifugal pump model at different flow rates N 1750 rpm

Complexity 7

5 Conclusion

In this study the comparison of numerical studies resultsobtained under different flow and speed conditions wasinvestigated From the study the following conclusions weremade

(A) (e speed of a turbine is determined by the amountof load used At constant flow rate and head theturbine speed decreases when the turbine uses ahigher load more than the design load with thedecrease of the burden placed on a turbine whereasthe head and flow rate remain constant the turbinespeed increases (e same principle applies whenhydraulic pumps are used as a turbine at a specifiedsite with a continuous flow rate and contributesignificantly to determining the turbinersquos perfor-mance characteristics As the PAT load is furtherreduced the PATrsquos speed reaches a maximum valuethat the PAT can deliver no torque (is speed iscalled runaway speed (e runaway speed of thePAT was obtained to be approximately 2250 rpm

(B) Increase in PATrsquos efficiency with the increased ro-tational speed of the impeller till it reaches 1500 rpmshould also be mentioned It means that when PAToperates beyond 1500 rpm the efficiency of the PATwill decrease In comparison with Figures 6 and 7PAT efficiency behaves equally as the power outputof the PAT (erefore it can be concluded that thebest-operating conditions for selected PAT could beachieved when it operates between 1000 rpm and1750 rpm

(C) Pressure distribution under different flow condi-tions and rotational speed in the pump as turbinefluid domain From the results pressure progres-sively intensified in a streamwise way with thepressure side recording the maximum Flow partingcould be observed at the impeller blade leading edge(e flow at the inlet is not at a tangent with the bladewith ununiformed flow observed near the bladewhich causes flow parting to occur on the bladesurface Minimal pressure was found at the impellerinlet along the suction area representing locationcavitation typically occurs in pumps

In summary PAT simulation results acquired fromANSYS CFX have been used to evaluate the PATrsquos per-formance parameters It is essential to recognise these pa-rameters for PAT performance under different flow rateconditions (ey consist of the head power and non-dimensionless coefficients such as head flow and powercoefficient Pump mode data can predict PAT performanceby using the pump BEP

Data Availability

(e data are openly available in a public repository thatissues datasets with DOIs (e authors confirm that the datasupporting the findings of this study are available within thearticle and could be obtained upon request (e data

generated at a central large-scale facility are available uponrequest Raw data were generated at Jiangsu University Lab(e derived data supporting the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

(e authors declare that they have no conflicts of interest

Acknowledgments

(is study was supported by the National Science Foun-dation of China under Grant nos 71974081 71704066 and71971100 and special funds of the National Social ScienceFund of China (Grant no 18VSJ038)

References

[1] J Gonzalez and C Santolaria ldquoUnsteady flow structure andglobal variables in a centrifugal pumprdquo Journal of FluidsEngineering vol 128 no 5 pp 937ndash946 2006

[2] B K Gandhi S N Singh and V Seshadri ldquoPerformancecharacteristics of centrifugal slurry pumpsrdquo Journal of FluidsEngineering vol 123 no 2 pp 271ndash280 2001

[3] H Nautiyal A Varun and A Kumar ldquoReverse runningpumps analytical experimental and computational study areviewrdquo Renewable and Sustainable Energy Reviews vol 14no 7 pp 2059ndash2067 2010

[4] S V Jain and R N Patel ldquoInvestigations on pump running inturbine mode a review of the state-of-the-artrdquo Renewable andSustainable Energy Reviews vol 30 pp 841ndash868 2014

[5] H Nautiyal V Varun A Kumar and S Y S Yadav ldquoEx-perimental investigation of centrifugal pump working asturbine for small hydropower systemsrdquo Energy Science andTechnology vol 1 pp 79ndash86 2011

[6] A J Stepanoff Centrifugal and Axial Flow Pumps Design andApplications John Wiley amp Sons New York NY USA 1957

[7] S Childs ldquoConvert pumps to turbines and recovers HPrdquoHydrocarbon Processing and Petroleum Refiner vol 41pp 173-174 1962

[8] R K Sharma Small Hydroelectric ProjectsmdashUse of CentrifugalPumps as Turbines Kirloskar Electric Co Bangalore India1984

[9] S Gopalakrishnan ldquoPower recovery turbines for the processindustryrdquo in Proceedings of the ltird International PumpSymposium pp 3ndash11 Texas AampM University Houston TXUSA September 1986

[10] H Diederich ldquoVerwendung von kreiselpumpen als turbinenrdquoKSB Techn Ber vol 12 pp 30ndash36 1967

[11] K Grover Conversion of Pumps to Turbines GSA Intercorporate Katonah NY USA 1982

[12] A Morabito and P Hendrick ldquoPump as turbine applied tomicro energy storage and smart water grids a case studyrdquoApplied Energy vol 241 pp 567ndash579 2019

[13] M Rossi A Nigro and M Renzi ldquoExperimental and nu-merical assessment of a methodology for performance pre-diction of pumps-as-turbines (PaTs) operating in off-designconditionsrdquo Applied Energy vol 248 pp 555ndash566 2019

[14] B Luo and X C L Wang ldquoNumerical simulation of flow-induced vibration of double-suction centrifugal pump asturbinerdquo Journal Drainage Irrigation Mechanical Engineeringvol 37 pp 313ndash318 2019

8 Complexity

[15] I M Hossain S Ferdous and T Jamal ldquoPump-as-turbine(PAT) for small-scale power generation a comparativeanalysisrdquo in Proceedings of the 2014 3rd International Con-ference on the IEEE Developments in Renewable EnergyTechnology (ICDRET) pp 1ndash5 Dhaka Bangladesh May 2014

[16] J-M Chapallaz P Eichenberger and G Fischer ldquoManual onpumps used as turbinesrdquo Vieweg vol 11 1992

[17] S Derakhshan and A Nourbakhsh ldquo(eoretical numericaland experimental investigation of centrifugal pumps in re-verse operationrdquo Experimental ltermal and Fluid Sciencevol 32 no 8 pp 1620ndash1627 2008

[18] P Vasanthakumar J Krishnaraj S Karthikayan T Vinothand S K Arun Sankar ldquoInvestigation on reverse character-istics of centrifugal pump in turbine mode a comparativestudy by an experimentation and simulationrdquo MaterialsToday Proceedings vol 4 no 2 pp 693ndash700 2017

[19] M Gantar ldquoPropeller pumps running as turbinesrdquo in Pro-ceedings of the Conference on Hydraulic Machinery pp 237ndash248 Ljubljana Slovenia August 1988

[20] H H Anderson ldquoBorehole pumpsrdquo Centrifugal Pumps andAllied Machinery vol 4 pp 177ndash185 1993

[21] D Adu J Zhang M Jieyun S N Asomani and M OsmanKoranteng ldquoNumerical investigation of transient vortices andturbulent flow behaviour in centrifugal pump operating inreverse mode as turbinerdquo Materials Science for EnergyTechnologies vol 2 no 2 pp 356ndash364 2019

[22] H Wang Y Wang C Fan et al ldquoMaterial consumption andcarbon emissions associated with the infrastructure con-struction of 34 cities in northeast Chinardquo Complexityvol 2020 Article ID 4364912 20 pages 2020

[23] Z Cao M Liu and P Wu ldquoExperiment investigation andnumerical simulation of Snowdrift on a typical large-spanretractable roofrdquo Complexity vol 2019 Article ID 598480414 pages 2019

[24] K Ren H Li S Li and H Dong ldquoVoltage stability analysis offront-end speed controlled wind turbine integrated into re-gional power grid based on bifurcation theoryrdquo Complexityvol 2020 Article ID 8816334 11 pages 2020

[25] D Florez-Orrego I B Henriques T-V Nguyen et al ldquo(econtributions of prof Jan Szargut to the exergy and envi-ronmental assessment of complex energy systemsrdquo Energyvol 161 pp 482ndash492 2018

[26] Y Zhang Y Xu Y Zheng et al ldquoMultiobjective optimizationdesign and experimental investigation on the axial-flow pumpwith orthogonal test approachrdquo Complexity vol 2019 ArticleID 1467565 14 pages 2019

[27] X Xu X Yuan M Cui Y Yuan and L Qi ldquoJoint optimi-zation of energy conservation and migration cost for complexsystems in edge computingrdquo Complexity vol 2019 Article ID6180135 14 pages 2019

[28] Y L Zhang S Q Yuan Z Jinfeng et al ldquoNumerical in-vestigation of the effects of splitter blades on the cavitationperformance of a centrifugal pumprdquo IOP Conference SeriesEarth and Environmental Science vol 22 Article ID 0520032014

[29] Z Jinfeng A Daniel S Lv and K Majid ldquoSmall hydropowerby using pump as turbine for power generation in SouthernAfricardquo Current Science vol 118 no 3 2019

Complexity 9

5 Conclusion

In this study the comparison of numerical studies resultsobtained under different flow and speed conditions wasinvestigated From the study the following conclusions weremade

(A) (e speed of a turbine is determined by the amountof load used At constant flow rate and head theturbine speed decreases when the turbine uses ahigher load more than the design load with thedecrease of the burden placed on a turbine whereasthe head and flow rate remain constant the turbinespeed increases (e same principle applies whenhydraulic pumps are used as a turbine at a specifiedsite with a continuous flow rate and contributesignificantly to determining the turbinersquos perfor-mance characteristics As the PAT load is furtherreduced the PATrsquos speed reaches a maximum valuethat the PAT can deliver no torque (is speed iscalled runaway speed (e runaway speed of thePAT was obtained to be approximately 2250 rpm

(B) Increase in PATrsquos efficiency with the increased ro-tational speed of the impeller till it reaches 1500 rpmshould also be mentioned It means that when PAToperates beyond 1500 rpm the efficiency of the PATwill decrease In comparison with Figures 6 and 7PAT efficiency behaves equally as the power outputof the PAT (erefore it can be concluded that thebest-operating conditions for selected PAT could beachieved when it operates between 1000 rpm and1750 rpm

(C) Pressure distribution under different flow condi-tions and rotational speed in the pump as turbinefluid domain From the results pressure progres-sively intensified in a streamwise way with thepressure side recording the maximum Flow partingcould be observed at the impeller blade leading edge(e flow at the inlet is not at a tangent with the bladewith ununiformed flow observed near the bladewhich causes flow parting to occur on the bladesurface Minimal pressure was found at the impellerinlet along the suction area representing locationcavitation typically occurs in pumps

In summary PAT simulation results acquired fromANSYS CFX have been used to evaluate the PATrsquos per-formance parameters It is essential to recognise these pa-rameters for PAT performance under different flow rateconditions (ey consist of the head power and non-dimensionless coefficients such as head flow and powercoefficient Pump mode data can predict PAT performanceby using the pump BEP

Data Availability

(e data are openly available in a public repository thatissues datasets with DOIs (e authors confirm that the datasupporting the findings of this study are available within thearticle and could be obtained upon request (e data

generated at a central large-scale facility are available uponrequest Raw data were generated at Jiangsu University Lab(e derived data supporting the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

(e authors declare that they have no conflicts of interest

Acknowledgments

(is study was supported by the National Science Foun-dation of China under Grant nos 71974081 71704066 and71971100 and special funds of the National Social ScienceFund of China (Grant no 18VSJ038)

References

[1] J Gonzalez and C Santolaria ldquoUnsteady flow structure andglobal variables in a centrifugal pumprdquo Journal of FluidsEngineering vol 128 no 5 pp 937ndash946 2006

[2] B K Gandhi S N Singh and V Seshadri ldquoPerformancecharacteristics of centrifugal slurry pumpsrdquo Journal of FluidsEngineering vol 123 no 2 pp 271ndash280 2001

[3] H Nautiyal A Varun and A Kumar ldquoReverse runningpumps analytical experimental and computational study areviewrdquo Renewable and Sustainable Energy Reviews vol 14no 7 pp 2059ndash2067 2010

[4] S V Jain and R N Patel ldquoInvestigations on pump running inturbine mode a review of the state-of-the-artrdquo Renewable andSustainable Energy Reviews vol 30 pp 841ndash868 2014

[5] H Nautiyal V Varun A Kumar and S Y S Yadav ldquoEx-perimental investigation of centrifugal pump working asturbine for small hydropower systemsrdquo Energy Science andTechnology vol 1 pp 79ndash86 2011

[6] A J Stepanoff Centrifugal and Axial Flow Pumps Design andApplications John Wiley amp Sons New York NY USA 1957

[7] S Childs ldquoConvert pumps to turbines and recovers HPrdquoHydrocarbon Processing and Petroleum Refiner vol 41pp 173-174 1962

[8] R K Sharma Small Hydroelectric ProjectsmdashUse of CentrifugalPumps as Turbines Kirloskar Electric Co Bangalore India1984

[9] S Gopalakrishnan ldquoPower recovery turbines for the processindustryrdquo in Proceedings of the ltird International PumpSymposium pp 3ndash11 Texas AampM University Houston TXUSA September 1986

[10] H Diederich ldquoVerwendung von kreiselpumpen als turbinenrdquoKSB Techn Ber vol 12 pp 30ndash36 1967

[11] K Grover Conversion of Pumps to Turbines GSA Intercorporate Katonah NY USA 1982

[12] A Morabito and P Hendrick ldquoPump as turbine applied tomicro energy storage and smart water grids a case studyrdquoApplied Energy vol 241 pp 567ndash579 2019

[13] M Rossi A Nigro and M Renzi ldquoExperimental and nu-merical assessment of a methodology for performance pre-diction of pumps-as-turbines (PaTs) operating in off-designconditionsrdquo Applied Energy vol 248 pp 555ndash566 2019

[14] B Luo and X C L Wang ldquoNumerical simulation of flow-induced vibration of double-suction centrifugal pump asturbinerdquo Journal Drainage Irrigation Mechanical Engineeringvol 37 pp 313ndash318 2019

8 Complexity

[15] I M Hossain S Ferdous and T Jamal ldquoPump-as-turbine(PAT) for small-scale power generation a comparativeanalysisrdquo in Proceedings of the 2014 3rd International Con-ference on the IEEE Developments in Renewable EnergyTechnology (ICDRET) pp 1ndash5 Dhaka Bangladesh May 2014

[16] J-M Chapallaz P Eichenberger and G Fischer ldquoManual onpumps used as turbinesrdquo Vieweg vol 11 1992

[17] S Derakhshan and A Nourbakhsh ldquo(eoretical numericaland experimental investigation of centrifugal pumps in re-verse operationrdquo Experimental ltermal and Fluid Sciencevol 32 no 8 pp 1620ndash1627 2008

[18] P Vasanthakumar J Krishnaraj S Karthikayan T Vinothand S K Arun Sankar ldquoInvestigation on reverse character-istics of centrifugal pump in turbine mode a comparativestudy by an experimentation and simulationrdquo MaterialsToday Proceedings vol 4 no 2 pp 693ndash700 2017

[19] M Gantar ldquoPropeller pumps running as turbinesrdquo in Pro-ceedings of the Conference on Hydraulic Machinery pp 237ndash248 Ljubljana Slovenia August 1988

[20] H H Anderson ldquoBorehole pumpsrdquo Centrifugal Pumps andAllied Machinery vol 4 pp 177ndash185 1993

[21] D Adu J Zhang M Jieyun S N Asomani and M OsmanKoranteng ldquoNumerical investigation of transient vortices andturbulent flow behaviour in centrifugal pump operating inreverse mode as turbinerdquo Materials Science for EnergyTechnologies vol 2 no 2 pp 356ndash364 2019

[22] H Wang Y Wang C Fan et al ldquoMaterial consumption andcarbon emissions associated with the infrastructure con-struction of 34 cities in northeast Chinardquo Complexityvol 2020 Article ID 4364912 20 pages 2020

[23] Z Cao M Liu and P Wu ldquoExperiment investigation andnumerical simulation of Snowdrift on a typical large-spanretractable roofrdquo Complexity vol 2019 Article ID 598480414 pages 2019

[24] K Ren H Li S Li and H Dong ldquoVoltage stability analysis offront-end speed controlled wind turbine integrated into re-gional power grid based on bifurcation theoryrdquo Complexityvol 2020 Article ID 8816334 11 pages 2020

[25] D Florez-Orrego I B Henriques T-V Nguyen et al ldquo(econtributions of prof Jan Szargut to the exergy and envi-ronmental assessment of complex energy systemsrdquo Energyvol 161 pp 482ndash492 2018

[26] Y Zhang Y Xu Y Zheng et al ldquoMultiobjective optimizationdesign and experimental investigation on the axial-flow pumpwith orthogonal test approachrdquo Complexity vol 2019 ArticleID 1467565 14 pages 2019

[27] X Xu X Yuan M Cui Y Yuan and L Qi ldquoJoint optimi-zation of energy conservation and migration cost for complexsystems in edge computingrdquo Complexity vol 2019 Article ID6180135 14 pages 2019

[28] Y L Zhang S Q Yuan Z Jinfeng et al ldquoNumerical in-vestigation of the effects of splitter blades on the cavitationperformance of a centrifugal pumprdquo IOP Conference SeriesEarth and Environmental Science vol 22 Article ID 0520032014

[29] Z Jinfeng A Daniel S Lv and K Majid ldquoSmall hydropowerby using pump as turbine for power generation in SouthernAfricardquo Current Science vol 118 no 3 2019

Complexity 9

[15] I M Hossain S Ferdous and T Jamal ldquoPump-as-turbine(PAT) for small-scale power generation a comparativeanalysisrdquo in Proceedings of the 2014 3rd International Con-ference on the IEEE Developments in Renewable EnergyTechnology (ICDRET) pp 1ndash5 Dhaka Bangladesh May 2014

[16] J-M Chapallaz P Eichenberger and G Fischer ldquoManual onpumps used as turbinesrdquo Vieweg vol 11 1992

[17] S Derakhshan and A Nourbakhsh ldquo(eoretical numericaland experimental investigation of centrifugal pumps in re-verse operationrdquo Experimental ltermal and Fluid Sciencevol 32 no 8 pp 1620ndash1627 2008

[18] P Vasanthakumar J Krishnaraj S Karthikayan T Vinothand S K Arun Sankar ldquoInvestigation on reverse character-istics of centrifugal pump in turbine mode a comparativestudy by an experimentation and simulationrdquo MaterialsToday Proceedings vol 4 no 2 pp 693ndash700 2017

[19] M Gantar ldquoPropeller pumps running as turbinesrdquo in Pro-ceedings of the Conference on Hydraulic Machinery pp 237ndash248 Ljubljana Slovenia August 1988

[20] H H Anderson ldquoBorehole pumpsrdquo Centrifugal Pumps andAllied Machinery vol 4 pp 177ndash185 1993

[21] D Adu J Zhang M Jieyun S N Asomani and M OsmanKoranteng ldquoNumerical investigation of transient vortices andturbulent flow behaviour in centrifugal pump operating inreverse mode as turbinerdquo Materials Science for EnergyTechnologies vol 2 no 2 pp 356ndash364 2019

[22] H Wang Y Wang C Fan et al ldquoMaterial consumption andcarbon emissions associated with the infrastructure con-struction of 34 cities in northeast Chinardquo Complexityvol 2020 Article ID 4364912 20 pages 2020

[23] Z Cao M Liu and P Wu ldquoExperiment investigation andnumerical simulation of Snowdrift on a typical large-spanretractable roofrdquo Complexity vol 2019 Article ID 598480414 pages 2019

[24] K Ren H Li S Li and H Dong ldquoVoltage stability analysis offront-end speed controlled wind turbine integrated into re-gional power grid based on bifurcation theoryrdquo Complexityvol 2020 Article ID 8816334 11 pages 2020

[25] D Florez-Orrego I B Henriques T-V Nguyen et al ldquo(econtributions of prof Jan Szargut to the exergy and envi-ronmental assessment of complex energy systemsrdquo Energyvol 161 pp 482ndash492 2018

[26] Y Zhang Y Xu Y Zheng et al ldquoMultiobjective optimizationdesign and experimental investigation on the axial-flow pumpwith orthogonal test approachrdquo Complexity vol 2019 ArticleID 1467565 14 pages 2019

[27] X Xu X Yuan M Cui Y Yuan and L Qi ldquoJoint optimi-zation of energy conservation and migration cost for complexsystems in edge computingrdquo Complexity vol 2019 Article ID6180135 14 pages 2019

[28] Y L Zhang S Q Yuan Z Jinfeng et al ldquoNumerical in-vestigation of the effects of splitter blades on the cavitationperformance of a centrifugal pumprdquo IOP Conference SeriesEarth and Environmental Science vol 22 Article ID 0520032014

[29] Z Jinfeng A Daniel S Lv and K Majid ldquoSmall hydropowerby using pump as turbine for power generation in SouthernAfricardquo Current Science vol 118 no 3 2019

Complexity 9