THESIS NTNU MST HYDROTURBINE AND GOVERNOR MODELLING 2010 PDF

Hydro Turbine and Governor Modelling
Electric - Hydraulic Interaction
Luz Alexandra Lucero Tenorio
Master of Science in Electric Power Engineering

Submission date: June 2010
Supervisor:
Kjetil Uhlen, ELKRAFT
Co-supervisor:
Trond Toftevaag, SINTEF Energy Research
Norwegian University of Science and Technology

Department of Electric Power Engineering
Problem Description
Traditionally, mathematical models for hydraulic power plants, normally found in relevant
literature and power system analysis tools are often simplified models. Approaches based on
approximate linear models assuming an ideal lossless turbine and ignoring the elasticity of the
conduit system, are not suitable for the accurate study of the interaction between hydraulic system
and power system. This implies that these models only reflect part of the real situations and as
such could have a limited application.
Assignment given: 25. January 2010

Supervisor: Kjetil Uhlen, ELKRAFT
HydroTurbineandGovernorModelling
Abstract
ThisMastersThesisworkdealswiththedevelopmentofimprovedhydroturbinemodelsforthe
evaluationofahydraulicpowergeneratingsystemperformanceinresponsetosmalldisturbances
in power system analysis tool. These improved models must be able to reflect the possible
interaction between the hydraulic system and power system in the computer simulations of a
powerplantequippedwithFrancisturbines.
The accuracy of a Hydraulic Power Generating System is studied by means of analysis of the
dynamicbehaviourofdifferentmodelsofthehydraulicmachineandconduitsystem.Thestability
study of different models for Synchronous Machines and Turbine Governing System are beyond
thescopeofthiswork.
Appropriaterepresentationsofthehydraulicturbineandconduitsystemaredevelopedinvarious
modelsofvaryingdegreesofdetail.Firstly,nonlinearmodelsforasimpleturbinewithoutsurge
tankconsideringtheinelasticandelastictravellingwaveeffectshavebeendeveloped.Afterthat,
nonlinear models considering the inelastic and elastic travelling wave effects for a turbine with
surge tank for Hydropower Systems with long length penstocks are implemented. Finally, the
nonlinear models for a turbine with long length penstocks are linearized at an operating point
consideringboththenonlinearturbinecharacteristicsandthetravellingwaveeffects.
The Masters Thesis work is divided into three parts. The first part, comprising Chapters 2 to 8,
reviews the physical characteristics and mathematical models of the components of a hydraulic
power generating system. The influence of each component of the power system by means of
appropriate mathematical models is essential for the understanding of system stability. The
second part, comprising Chapters 9 to 11, deals with the dynamic study of the system stability
characteristics of the different hydraulic power generating system models implemented in
SIMPOWandLVTrans.Finally,thethirdpart,Chapter12andChapter13,presentsthediscussion
ofthesimulationresultsofthehydroelectricpowersystemmodels,anddrawsgeneralconclusions
onthisworkandsuggestspossibilitiesfortheapproachfurtherwork,respectively.
Itwasconcludedthatapproachesbasedonnonlinearandlinearmodelsincludingtheelasticityof
the conduit system and the nonlinear turbine characteristics extracted from the Hill Charts, are
themostaccuratemodelsforanyacceptablestudyoftheinteractionbetweenhydraulicsystem
andpowersystem.
The study of dynamic performance and interaction between the hydraulic system and power
systemoftheseextendedlinearandnonlinearmodelsincludingtheelasticwaterhammereffect
andvaryingthenonlinearcharacteristicsofthehydraulicturbinemustbestudiedindetail.
ii
Table of Contents
Page
ABSTRACT...............................................................................................................................................................II
TABLEOFCONTENTS..............................................................................................................................................III
LISTOFFIGURES.....................................................................................................................................................VI
LISTOFTABLES.......................................................................................................................................................IX
PREFACE..................................................................................................................................................................X
LISTOFSYMBOLS...................................................................................................................................................XII
1
INTRODUCTION...............................................................................................................................................1
1.1BACKGROUNDINFORMATION...........................................................................................................................................1
1.2OBJECTIVE....................................................................................................................................................................1
1.3SCOPE.........................................................................................................................................................................1
1.4OUTLINEOFTHEPROJECT................................................................................................................................................2
PRESSUREWATERCONDUITSYSTEM...............................................................................................................3
2.1HYDRAULICTRANSIENTSFUNDAMENTALS..........................................................................................................................3
2.1.1Pressurewavevelocityinconduits....................................................................................................................4
2.1.2Wavepropagationandreflectionsintheconduit.............................................................................................4
2.1.3HeadlossesduetoFriction...............................................................................................................................5
2.2CLOSEDCONDUITSYSTEMS.............................................................................................................................................6
2.2.1BasicAssumptions.............................................................................................................................................6
2.2.2BasicDifferentialEquationsforTransientFlow................................................................................................6
2.2.2.1EquationofContinuity.................................................................................................................................................7
2.2.2.2EquationofMotion......................................................................................................................................................9
2.2.2.3GeneralRemarks........................................................................................................................................................10
2.2.3MathematicalModel......................................................................................................................................11
2.3PRESSURECONTROLSYSTEMS........................................................................................................................................15
2.3.1GoverningEquations.......................................................................................................................................15
2.3.2SurgeTankMathematicalModel....................................................................................................................17
3
HYDRAULICTURBINES...................................................................................................................................18
3.1HYDRAULICTURBINESOVERVIEW...................................................................................................................................18
3.1.1Impulseturbines..............................................................................................................................................18
3.1.2ReactionTurbines............................................................................................................................................19
3.2GENERALTECHNICALASPECTS.......................................................................................................................................20
3.2.1NeatHead,PowerandEfficiency....................................................................................................................20
3.2.2TurbineHillCharts...........................................................................................................................................21
3.3HYDRAULICTURBINEMODELLING...................................................................................................................................22
3.3.1SimplifiedNonlinearTurbinemodel................................................................................................................23
3.3.2Thelinearizedhydroturbinemodel................................................................................................................23
SYNCHRONOUSMACHINE.............................................................................................................................26
4.1SYNCHRONOUSGENERATOR..........................................................................................................................................26
4.2SYNCHRONOUSGENERATOREQUATIONS.........................................................................................................................27
4.3SYNCHRONOUSGENERATORMODELS..............................................................................................................................28
4.3.1SynchronousMachineRepresentedbytheClassicalModel...........................................................................29
TURBINEGOVERNINGSYSTEMS.....................................................................................................................30
5.1GOVERNINGSYSTEM....................................................................................................................................................30
5.2MECHANICALHYDRAULICGOVERNOR.............................................................................................................................31
5.2.1MathematicalModelling................................................................................................................................31
5.3ELECTROHYDRAULICGOVERNINGSYSTEM.......................................................................................................................34
5.3.1MathematicalModel......................................................................................................................................34
iii
HYDRAULICPOWERPLANTMODELS..............................................................................................................36
6.1HYDROPOWERPLANTMODELS......................................................................................................................................36
6.2NONLINEARTURBINEMODELS.......................................................................................................................................38
6.2.1SimplifiedNonlinearTurbineModel................................................................................................................38
6.2.2NonlinearModelwithoutSurgeTankassumingInelasticWaterColumn.......................................................39
6.2.3NonlinearModelwithoutsurgetankincludingElasticWaterColumnEffect.................................................39
6.2.4NonlinearModelwithSurgeTankassumingInelasticWaterColumns..........................................................40
6.2.5NonlinearModelwithSurgeTankassumingElasticWaterColumninPenstockandInelasticWaterColumn
inUpstreamTunnel..................................................................................................................................................41
6.3HYDROTURBINELINEARMODELS...................................................................................................................................43
6.3.1LinearTurbineModelwithSurgeTankassumingInelasticWaterColumns...................................................43
6.3.2LinearTurbineModelwithSurgeTankassumingElasticWaterColumninPenstock.....................................44
HYDROELECTRICPOWERPLANTMODELLINGBYSTRUCTUREMATRIXMETHOD.............................................46
7.1DEFINITIONOFTHEMETHOD.........................................................................................................................................46
7.2MATRIXREPRESENTATIONSOFTHEBASICELEMENTSINHYDROPOWERSYSTEMS....................................................................47
7.2.1PipesandTunnels............................................................................................................................................47
7.2.2Surgetanksorairaccumulators.....................................................................................................................47
7.2.3LocalLosses.....................................................................................................................................................48
7.2.4HydroTurbine.................................................................................................................................................49
7.2.4.1HydroTurbineCharacteristics....................................................................................................................................49
7.2.4.2HydroTurbineMatrixRepresentation.......................................................................................................................50
7.2.5Thesynchronousgeneratorandtheelectricgrid............................................................................................52
7.2.6TurbineSpeedGovernor..................................................................................................................................53
7.2.6.1TraditionalGovernor..................................................................................................................................................53
7.2.6.2PIDGovernor..............................................................................................................................................................54
7.3COMPOSITIONOFTHEGLOBALSTRUCTUREMATRIX...........................................................................................................55
7.4STRUCTUREMATRIXOFTHEHYDROTURBINEUNIT............................................................................................................55
7.5HYDROPOWERPLANTSTRUCTUREMATRIX......................................................................................................................56
7.6DYNAMICANALYSIS.....................................................................................................................................................58
7.6.1FrequencyResponseAnalysis..........................................................................................................................58
7.6.2FreeVibrationAnalysis....................................................................................................................................58
8
POWERSYSTEMSTABILITYANALYSIS.............................................................................................................59
8.1POWERSYSTEMSTABILITY.............................................................................................................................................59
8.1.1SmallSignalStabilityAnalysis.........................................................................................................................59
8.2FUNDAMENTALSOFPOWERSYSTEMSTABILITY.................................................................................................................60
8.3EIGENVALUEANALYSIS..................................................................................................................................................61
8.3.1EigenvaluesandEigenvectors.........................................................................................................................61
8.3.2EigenvalueAnalysis.........................................................................................................................................61
8.3.3ModalandSensitivityAnalysis........................................................................................................................62
8.4SMALLSIGNALSTABILITYANALYSISOFAGENERATORINFINITEBUSSYSTEM............................................................................63
8.5DYNAMICANALYSIS.....................................................................................................................................................64
8.5.1Naturalfrequencyanalysis.............................................................................................................................64
8.5.2FrequencyResponseAnalysis..........................................................................................................................65
TESTSYSTEMMODEL.....................................................................................................................................66
9.1OVERVIEWOFTHETESTSYSTEM.....................................................................................................................................66
9.2MODELSOFTHEDIFFERENTCOMPONENTS........................................................................................................................67
9.2.1PowerPlantModel..........................................................................................................................................67
9.2.2HydraulicTurbine............................................................................................................................................68
9.2.3SynchronousGenerator...................................................................................................................................69
9.2.4GoverningSystemsforHydraulicTurbines.....................................................................................................70
9.3POWERFLOWANALYSIS...............................................................................................................................................70
9.4POWERSYSTEMSTABILITYANALYSISTHEORETICALRESULTS.............................................................................................71
9.4.1EigenvalueAnalysis.........................................................................................................................................71
9.4.2Naturalfrequencyanalysis.............................................................................................................................72
10 DYNAMICSIMULATIONSINSIMPOW.............................................................................................................73
iv
10.1MODEL1:SIMPLIFIEDNONLINEARTURBINEMODEL......................................................................................................74
10.1.1EigenvalueAnalysis.......................................................................................................................................74
10.1.2DynamicSimulationAnalysis........................................................................................................................75
10.1.3FrequencyResponseAnalysis........................................................................................................................76
10.2MODEL2:NONLINEARTURBINEMODELWITHOUTSURGETANKASSUMINGINELASTICWATERCOLUMN.................................79
10.2.2DynamicResponseAnalysis..........................................................................................................................80
10.3MODEL3:NONLINEARTURBINEMODELWITHOUTSURGETANKINCLUDINGELASTICWATERCOLUMNEFFECTS........................84
10.4MODEL4:NONLINEARTURBINEMODELWITHSURGETANKASSUMINGINELASTICWATERCOLUMNS.....................................89
10.5 MODEL 5: NONLINEAR MODEL WITH SURGE TANK ASSUMING ELASTIC WATER COLUMN IN PENSTOCK AND INELASTIC WATER
COLUMNINTUNNEL..........................................................................................................................................................94
10.6MODEL6:LINEARTURBINEMODELWITHSURGETANKASSUMINGINELASTICWATERCOLUMNS............................................99
10.6.2DynamicSimulationAnalysis......................................................................................................................100
10.6.3FrequencyResponseAnalysis......................................................................................................................101
10.7MODEL7:LINEARTURBINEMODELWITHSURGETANKASSUMINGELASTICWATERCOLUMNINPENSTOCK...........................104
10.7.1EigenvalueAnalysis.....................................................................................................................................104
10.7.2DynamicSimulationAnalysis......................................................................................................................105
10.7.3FrequencyResponseAnalysis......................................................................................................................106
11 POWERSYSTEMMODELLEDINLVTRANS.....................................................................................................109
11.1DESCRIPTIONOFLVTRANS........................................................................................................................................109
11.2HYDRAULICSYSTEMMODELLING................................................................................................................................110
11.3DYNAMICSIMULATION.............................................................................................................................................111
11.3.1FrequencyResponseAnalysiscomputedinLVTrans8_1.1.2.......................................................................112
11.3.2FrequencyResponseAnalysiscomputedinLVTrans86_1.3.1_T.................................................................114
11.3.3ComparisonoftheFrequencyResponseAnalysisoftheHydraulicPowerPlantcomputedinLVTrans8_1.1.2
andLVTrans86_1.3.1_T.........................................................................................................................................114
12 DISCUSSION................................................................................................................................................117
12.1SUMMARYOFSIMULATIONRESULTSOFTHEMODELSCOMPUTEDINSIMPOW...................................................................117
12.2SUMMARYOFTHEEIGENVALUEANALYSISCOMPUTEDINSIMPOW..................................................................................118
12.3COMPARISONOFTHERESULTSOFTHEMODELSINSIMPOW...........................................................................................119
12.3.1ComparisonofNonlinearTurbineModelswithoutsurgetank...................................................................119
12.3.2ComparisonofTurbineModelswithSurgeTank........................................................................................122
12.4COMPARISONOFTHESIMULATIONRESULTSCOMPUTEDINSIMPOWANDLVTRANS...........................................................125
13 CONCLUSIONS.............................................................................................................................................127
13.1CONCLUSIONS.........................................................................................................................................................127
13.2FURTHERWORK......................................................................................................................................................129
14 REFERENCEBIBLIOGRAPHY..........................................................................................................................130
List of Figures
Figure21:Controlvolumeforthederivationoftheequationofcontinuity.....................................7
Figure22:Controlvolumeforthederivationoftheequationofmotion..........................................9
Figure23:Simplesurgetank,[6]......................................................................................................16
Figure31:ImpulseTurbine...............................................................................................................18
Figure32:FrancisTurbine................................................................................................................19
Figure33:SchematicofHydraulicPowerPlantwithareactionturbine.........................................20
Figure34:FunctionalBlockDiagramofthehydroturbine,[1,28]..................................................22
Figure41:Equivalentcircuitoftheclassicalmodelofthegenerator..............................................29
Figure51:SchematicdiagramoftheGoverningSystem,[48].........................................................31
Figure52:MechanicalHydraulicGoverningSystem.[47]...............................................................31
Figure53:Modelofgovernorforhydraulicturbines,[2].................................................................33
Figure54:TypicalPIDGovernorController,[2]...............................................................................34
Figure61:FunctionalBlockDiagramoftheHydraulicTurbineGeneratingSystem........................37
Figure62:SimplifiedNonlinearTurbineModel...............................................................................38
Figure63:NonlinearModelwithoutSurgeTankassumingInelasticWaterColumn......................39
Figure64:NonlinearModelwithoutsurgetankincludingElasticWaterColumnEffect................40
Figure65:NonlinearModelwithSurgeTankassumingInelasticWaterColumns..........................41
Figure66:NonlinearModelwithsurgetankassumingelasticwatercolumninpenstockand
inelasticwatercolumninupstreamtunnel.....................................................................42
Figure67:LinearModelwithSurgeTankassumingInelasticWaterColumns................................44
Figure68:LinearModelwithSurgeTankassumingElasticWaterColumninPenstock.................45
Figure71:LayoutofHydroPowerPlantforStructureMatrixMethodModelling..........................56
Figure91:Singlemachineinfinitebuspowersystem......................................................................66
Figure92:AgeneralLayoutofHydroPowerPlant,[19]..................................................................67
Figure93:Circuitmodelofthetestsystem.....................................................................................71
Figure 101: Fault simulation results: (a) angle, (b) speed, (c) mechanical torque, (d) gate
position,(e)flowrateand(f)headpressureofModel1................................................75
Figure 102: Frequency response of the transfer function from Gate position to Mechanical
Powerofmodel1.............................................................................................................76
TorqueofModel1...........................................................................................................77
Figure104:FrequencyresponseofthetransferfunctionoftheconduitsystemofModel1.........77
Figure105:Frequencyresponseofthetransferfunctionfromgatepositiontoelectricalangle
ofModel1........................................................................................................................78
Figure 106: Frequency response of the transfer function of the mechanicalhydraulic
governorofModel1........................................................................................................78
Figure107:FaultsimulationResults:(a)Powerangle,(b)Speed,(c)MechanicalTorque,(d)
GatePosition,(e)FlowRateand(f)HeadPressureofModel2......................................80
PowerofModel2.............................................................................................................81
TorqueofModel2...........................................................................................................82
Figure1010:FrequencyresponseofthetransferfunctionoftheconduitsystemofModel2.......82
Figure 1011: Frequency response of the transfer function from Gate position to Electrical
AngleofModel2..............................................................................................................83
vi
Figure 1012: Frequency Response of the transfer function of the mechanicalhydraulic

Figure 1013: Fault simulation results: (a) angle, (b) speed, (c) mechanical torque, (d) gate
position,(e)flowrateand(f)waterpressureofModel3...............................................85
Figure1014:FrequencyresponseofthetransferfunctionfromGatepositiontoMechanical
PowerofModel3.............................................................................................................86
TorqueofModel3...........................................................................................................87
Figure1016:FrequencyResponseofthetransferfunctionoftheconduitsystemofModel3......87
AngleofModel3..............................................................................................................88
Figure 1019: Fault simulation results: (a) angle, (b) Speed, (c) Mechanical Torque, (d) Gate
Position,(e)FlowRateand(f)HeadPressureofModel4...............................................90
PowerofModel4.............................................................................................................91
TorqueofModel4...........................................................................................................92
AngleofModel4..............................................................................................................93
Figure 1025: Fault simulation Results: (a) angle, (b) Speed, (c) mechanical torque, (d) gate
position,(e)flowrateand(f)waterpressureofModel5...............................................95
PowerofModel5.............................................................................................................96
TorqueofModel5...........................................................................................................97
AngleofModel5..............................................................................................................98
Figure 1031: Fault simulation Results: (a) angle, (b) speed, (c) mechanical torque, (d) gate
position,(e)flowrateand(f)waterpressureofModel6.............................................100
PowerofModel6...........................................................................................................101
TorqueofModel6.........................................................................................................102
Figure1034:FrequencyResponseofthetransferfunctionoftheconduitsystemofModel6....102
AngleofModel6............................................................................................................103
governorofModel6......................................................................................................103
Figure1037:FaultsimulationResults:(a)Powerangle,(b)Speed,(c)MechanicalTorque,(d)
GatePosition,(e)FlowRateand(f)HeadPressureofModel7....................................105
vii
PowerofModel7...........................................................................................................106
TorqueofModel7.........................................................................................................107
Figure1040:FrequencyResponseofthetransferfunctionoftheconduitsystemofModel7....107
AngleofModel7............................................................................................................108
governorofModel7......................................................................................................108
Figure111:BlockDiagramoftheHydraulicSystemModel...........................................................110
Figure112:SimulationResults:(a)gateopeningposition,(b)PressureHeadoftheHydraulic
TestModelinLVTrans86_1.3.1_T.................................................................................111
Figure113:FrequencyResponseofthetransferfunctionfromgatepositiontopressurehead
ofthehydraulicpowerplantsimulatedwithoutdroopinLVTrans8_1.1.2..................112
ofthehydraulicpowerplantsimulatedwithdroopinLVTrans8_1.1.2........................112
ofthehydraulicpowerplantsimulatedwithoutandwithdroopinLVTrans8_1.1.2...113
ofthehydraulicpowerplantsimulatedinLVTrans86_1.3.1_T.....................................114
Figure117:ComparisonoftheFrequencyresponseofthehydraulicpowergeneratingsystem
simulatedwithoutdroopinLVTrans8_1.1.2andcomputedinLVTrans86_1.3.1_T.....115
Figure 118: Comparison of the Frequency Response of the hydraulic power generating
system simulated with droop in LVTrans8_1.1.2 and computed in
LVTrans86_1.3.1_T.........................................................................................................115
Figure121:FaultSimulationResultsofhydraulicturbinerepresentedbymodel1,2and3........119
Figure122:FrequencyresponseofthetransferfunctionsfromGatepositiontoMechanical
PowerofasimplehydraulicturbinerepresentedbyModel1,2and3........................120
TorqueofahydraulicturbinerepresentedbyModel1,2and3..................................121
Figure124:FrequencyresponseofthetransferfunctionsoftheconduitsystemofModel1,2
and3..............................................................................................................................121
Figure125:FaultSimulationResultsofhydraulicturbinerepresentedbyModel4,5,6and7....122
Figure 126: Frequency response of the transfer function from gate position to Mechanical
PowerofModel4,5,6and7.........................................................................................123
Figure 127: Frequency Response of the transfer function from Gate position to Mechanical
TorqueofasimplehydraulicturbinerepresentedbyModel4,5,6and7...................123
Figure128:FrequencyresponseofthetransferfunctionoftheconduitsystemofModel4,5,
6and7............................................................................................................................124
Figure129:FrequencyResponsesofthetransferfunctionoftheconduitsystemofHydraulic
PowerSystemModelsimplementedinSIMPOWandLVTrans(withoutdroop)..........125
Figure1210:FrequencyResponsesofthetransferfunctionoftheconduitsystemofHydraulic
PowerSystemModelsimplementedinSIMPOWandLVTrans(withdroop)...............126
viii
List of Tables
Table21:Propagationofpressurewavescausedbyinstantaneousclosureofvalve.......................5
Table31:Turbinecoefficients,[11,37]............................................................................................24
Table32:TurbineCoefficients,[32].................................................................................................25
Table51:TypicalvaluesandrangeofParameters,[47]..................................................................33
Table91:Systemparametersandoperatingconditions..................................................................67
Table92:HydraulicPowerPlantModel...........................................................................................68
Table93:HydraulicTurbineParameters..........................................................................................68
Table94:HydraulicTurbineModels.................................................................................................69
Table95:SynchronousmachineParameters...................................................................................69
Table96:TypicalValuesofparametersofturbinegoverningsystem.............................................70
Table97:PowerFlowresults............................................................................................................70
Table98:ResultsoftheSynchronousmachinerepresentedbyClassicalModel............................72
Table101:Eigenvaluesandthestatevariableforamodel1..........................................................74
Table102:Participationmatrixforamodel1..................................................................................75
Table103:EigenvaluesandthestatevariableforModel2.............................................................79
Table104:ParticipationfactorsmatrixofaturbineModel2..........................................................80
Table105:EigenvaluesandthestatevariablefortheTurbineModel3.........................................84
Table106:ParticipationfactorsmatrixofaTurbineModel3.........................................................85
Table107:EigenvaluesandthestatevariablefortheturbineModel4..........................................89
Table108:ParticipationfactorsmatrixofahydraulicturbineModel4..........................................90
Table109:EigenvaluesandstatevariablesforModel5..................................................................94
Table1010:ParticipationfactorsmatrixforModel5......................................................................95
Table1011:EigenvaluesandthestatevariableforModel6...........................................................99
Table1012:ParticipationfactorsmatrixofaModel6...................................................................100
Table1013:EigenvaluesandthestatevariablefortheModel7...................................................104
Table1014:ParticipationfactorsmatrixofaModel7...................................................................105
Table121:Summaryoftheelectromechanicaloscillatorymodeofthemodelsofthemodels...118
Table122:Errorbetweenthecalculatedeigenvaluesandthecomputedeigenvalues................118
ix
Preface
In January 2005, an internal project called Turbine and hydropower modelling was started at
SINTEFEnergyResearch.Thatprojectwasconcernedaboutthestudyofdynamicperformanceand
possibleinteractionbetweenthehydraulicsystemandpowersystemofapowerplantequipped
with Francis turbines. Traditionally, mathematical models for hydraulic power plants, normally
found in relevant literature and power system analysis tools are often simplified models.
Approachesbasedonapproximatelinearmodelsassuminganideallosslessturbineandignoring
the elasticity of the conduit system, are not suitable for the accurate study of the interaction
betweenhydraulicsystemandpowersystem.Thisimpliesthatthesemodelsonlyreflectpartof
therealsituationsandassuchcouldhavealimitedapplication.
The Masters Thesis work developsimproved hydro turbine models of a typical hydraulic power
generating system for their representation in dynamic studies of power systems in response to
any small disturbance. These improved models must be able to reflect the possible interaction
between the hydraulic system and power system in the computer simulations of a power plant
equippedwithFrancisturbines.
The study of dynamic performance and interaction of hydraulic system and power system of a
powerplantequippedwithFrancisturbinesintheMastersThesisworkisbasedonthetextbook
Power System Stability and Control by P. Kundur, and the paper Hydraulic Turbine and Turbine
ControlModelsforSystemDynamicStudiesbyIEEEWorkingGroup.[12]
Theprecisionoftherepresentationofahydraulicpowergeneratingsystemindynamicstudiesis
examined by means of analysis of the dynamic behaviour of different models of the hydraulic
machine and conduit system. The stability study of different models for Synchronous Machines
andTurbineGoverningSystemarebeyondofthescopeofthisMastersThesiswork.
Appropriaterepresentationsofthehydraulicturbineandwaterconduitsystemaredevelopedin
modelsofvaryingdegreesofdetail.Firstly,nonlinearmodelsforasimpleturbinewithoutsurge
tankconsideringtheinelasticandelastictravellingwaveeffectshavebeendeveloped.Afterthat,
nonlinear models considering the inelastic and elastic travelling wave effects for a turbine with
surgetankforhydraulicpowersystemswithlonglengthpenstocksareimplemented.Finally,the
nonlinear models for a turbine with long length penstocks are linearized at an operating point
consideringboththenonlinearturbinecharacteristicsandthetravellingwaveeffects.
The stability analysis of these implemented models contains powerflow calculation, linear
analysis and timedomain simulation in the simulation software SIMPOW and the dynamic
simulationtoolLVTrans.
The Masters Thesis work is divided into three parts. The first part, comprising Chapters 2 to 8,
reviews the physical characteristics and mathematical models of the components of a hydraulic
power generating system. The influence of each component of the power system by means of
appropriate mathematical models is essential for the understanding of system stability. The
second part, comprising Chapters 9 to 11, deals with the dynamic study of the system stability
characteristics of the different hydraulic power generating system models implemented in
SIMPOWandLVTrans.Finally,thethirdpart,Chapter12andChapter13,presentsthediscussion
x
ofthesimulationresultsofthehydroelectricpowersystemmodels,anddrawsgeneralconclusions
onthisworkandsuggestspossibilitiesfortheapproachfurtherwork,respectively.
Chapter2to5presentthephysicaldescriptionandthemathematicaldeductionoftheequations
describing appropriate models of the conduit systems and pressure control systems considering
the waterhammer theory and friction head losses, hydraulic turbines, synchronous generators
and turbine governing systems, respectively, for their representation in power system dynamic
studies. Chapter 6 deals with the development of improved mathematical models of each
component of a typical Hydraulic Power Generating System equipped with Francis turbines for
theirrepresentationinpowersystemdynamicstudies.Chapter7showsanalternativemethodof
HydroelectricPowerPlantmodellingforstabilitystudies.Chapter8describesfundamentalaspects
and analytical techniques in the study of smallsignal stability of dynamic, and identifies factors
influencingthem.
Chapter 9 presents the physical characteristics and capability of the components of the test
system. The general configuration of the test system consists of a single synchronous machine
connectedtoalargepowersystemthroughatransmissionline.Chapter10dealswiththestudyof
the system stability characteristics of the different Hydraulic Turbine models within a typical
Hydraulic Power Generating System implemented in the software SIMPOW. Chapter 11 studies
the dynamic system characteristics of the Hydraulic System modelled in the dynamic simulation
toolLVTrans.
The development of this work has the main contribution of Trond Toftevaag, Bjrnar Svingen,
KjetilUhlen,andLarsLindquist,STRISweden,whohaveassistedmeinthedynamicstudyofthe
differentmodelsandtheirimplementationinthesimulationsoftwareSIMPOWandthedynamic
I would like to thank Trond Toftevaag, Bjrnar Svingen and Kjetil Uhlen for their continuous
supportintheproject.AspecialthankstoTrondToftevaag,whowasalwaystheretolistenandto
give advice. He showed me different ways to approach a research problem and the need to be
persistent to accomplish any goal, and to ask me good questions to help me think through my
problems(whetherphilosophical,analyticalorcomputational).
AlexandraLuceroT.
Trondheim,June15th2010
xi
List of Symbols
Notation
Uppercasesymbolsnormallydenotephysicalvalues.
Lowercasesymbolsnormallydenoteperunitvalues.
Thesuffix0denotesaninitialsteadystatevalue.
Theprefixdenotessmalldeviations.
Symbols
a
Pressurewavevelocity,[m/s]
A
Penstockcrosssectionarea,[m2]
As
SurgeTankcrosssectionarea,[m2]
At
constantproportionalityfactor,[]
D
Pipeinternaldiameter,[m]
e
Pipewallthickness,[m]
E
Youngsmodulusofelasticity,[N/m2]
f
DarcyWeisbachfrictionfactor,[]
g
Accelerationduetogravity,[m/s2]
H(x,t)
Piezometrichead(HGL)orwaterfreesurfacelevel,[m]
H
InertiaConstant,[MWs/MVA]
Headlossesduetofrictioneffectsintheconduit,[m]
Hf
id
Currentflowinginthedaxisarmaturecoil,[pu]
iq
Currentflowingintheqaxisarmaturecoil,[pu]
K
Bulkmodulusofelasticity,[N/m2]
KD
Dampingtorquecoefficient,[]
Kd
Derivativegain,[s]
Transientsynchronizingpowercoefficient,[]
KE
Integralgain,[s1]
Ki
Kp
Proportionalgain,[pu]
Ks
Servomotorgain,[]
L
Lengthoftheconduit,[m]
Lad,Laq
Mutualinductancebetweenthestatorandrotorwindings
SelfInductanceofthedaxisarmaturewindings
Ld
Lq
SelfInductanceoftheqaxisarmaturewindings
M11
HydraulicturbineUnittorque
Me
Electromagnetictorque,[Nm]
Mm
Mechanicaltorque,[Nm]
p(x,t)
pressure,[Pa]
Pm
Mechanicalpowerontheturbineshaft,[MW]
Q(x,t)
Flowdischarge,[m3/s]
Q11
Hydraulicturbineunitflowrate
qNL
waterflowatzeroelectricalpower,[]
r
Piperadius,[m]
Statorresistance,[pu]
Ra
Re
Reynoldsnumber,[]
RP
Permanentspeeddroop,[pu]
R T
Temporaryspeeddroop,[pu]
s
Laplaceoperator
xii
SN
t
Tc
Td0,Td0
Te
Tg
TM
TP
Tq0
TR
Ts
Tw
V(x,t)
v
vd , vq
Ratedapparentpower,[MVA]
Time,[s]
Gateclosingtimeconstant,[s]
opencircuitdaxistransientandsubtransienttimeconstant,[s]
Wavetraveltime,[s]
MainServomotortimeconstant[s]
MechanicalstartingtimeTM=2H,[s]
Pilotvalveandservomotortimeconstant,[s]
Opencircuitqaxissubtransienttimeconstant,[s]
Resettimeordashpottimeconstant,[s]
SurgeTankFillingTime,[s]
Watertimeconstantorwaterstartingtime,[s]
averageVelocity,[m/s]
Kinematicsviscosityofthefluid,[m2/s]
Voltageacrossthefictitiousdandqaxisarmaturecoils,[pu]
Vg
VS
x
xl
Xd,Xd,Xd
xd,xd,xd
Voltageatthegeneratorterminals
Infinitebusbarvoltage
Distancemeasuredalongthepipeaxis,[m]
Statorleakagereactance,[pu]
directaxissynchronous,transientandsubtransientreactance,[pu]
Totaldirectaxissynchronous,transientandsubtransientreactancebetween
(andincluding)ageneratorandaninfinitebusbar,[pu]
quadratureaxissynchronous,transientandsubtransientreactance,[pu]
Internalreactanceofaninfinitebusbar,[pu]
Seriesreactanceofatransformer,[pu]
Turbinewicketgateposition,[]
Elevationofthepipecentreline,[m]
Slopeofthepipeaxis,[]
Poweranglewithrespecttoaninfinitebusbar
Internalpiperoughness,[m]
Dampingratio,[]
Turbineefficiency
Eigenvalue
Fluidmassdensity,[kg/m3]
Shearforcebetweenthefluidandtheconduitwalls,[N/m2]
Runnerbladeangle
Totalfluxlinkagesofdamperwindingsindaxisandqaxis
Totaldaxisandqaxisfluxlinkages
Totalfluxlinkageofthefieldwinding
Dampednaturalfrequencyofrotorswings,[rad/s]
Normalizedvelocityoftheunit,[pu]
Undampednaturalfrequencyofrotorswingsforsmalloscillations,[rad/s]
angularvelocityofthegenerator,[electricalradians]
Synchronousangularvelocity(equalto 2 f ),[electricalradians]
Conduitwalltensilestrain
Xq,Xq,Xq
x s
x T
y
z
D,Q
d,q
f
d
N
nat
xiii
Introduction
1.1 Backgroundinformation
InJanuary2005,aninternalprojectcalledTurbineandhydropowermodellingwasstartedat
SINTEFEnergyResearch.Thisprojectwasconcernedaboutthestudyofdynamicperformance
and possible interaction between the hydraulic system and power system of a power plant
equippedwithFrancisturbines.Traditionally,mathematicalmodelsforhydraulicpowerplants,
normally found in relevant literature and power system analysis tools are often simplified
models. Approaches based on approximate linear models assuming an ideal lossless turbine
andignoringtheelasticityoftheconduitsystem,arenotsuitablefortheaccuratestudyofthe
interactionbetweenhydraulicsystemandpowersystem.Thisimpliesthatthesemodelsonly
reflectpartoftherealsituationsandassuchcouldhavealimitedapplication.
1.2 Objective
ThepurposeofthisMastersThesisworkistodevelopimprovedhydraulicturbinemodelsofa
typicalhydraulicpowergeneratingsystemfortheirrepresentationindynamicstudiesofpower
systemsinresponsetoanysmalldisturbance.Theseimprovedmodelsmustbeabletoreflect
the possible interaction between the hydraulic system and power system in the computer
simulationsofapowerplantequippedwithFrancisturbines.
1.3 Scope
TheprecisionoftherepresentationofaHydraulicPowerGeneratingSystemindynamicstudies
isexaminedbymeansofanalysisofthedynamicbehaviourofdifferentmodelsofthehydraulic
machine and water conduit system. The stability study of different models for Synchronous
MachinesandTurbineGoverningSystemarebeyondofthescopeofthisMastersThesiswork.
Appropriaterepresentationsofthehydraulicturbineandwaterconduitsystemaredeveloped
in models of varying degrees of detail. Firstly, nonlinear models for a simple turbine without
surge tank considering the inelastic and elastic travelling wave effects have been developed.
After that,nonlinear models considering the inelastic and elastic travelling waveeffects fora
turbine with surge tank for hydraulic power systems with long length penstocks are
implemented. Finally, the nonlinear models for a turbine with long length penstocks are
linearizedatanoperatingpointconsideringboththenonlinearturbinecharacteristicsandthe
travellingwaveeffects.
The stability analysis of these implemented models contains powerflow calculation, linear
analysis and timedomain simulation in the simulation software SIMPOW and the dynamic
1.4 Outlineoftheproject
Chapter2containsthemathematicaldeductionoftheequationsofcontinuityandmomentum
describing the transient state flows in closedconduit systems and pressure control systems
includingtravellingwaveeffectsandfrictionlosses.
Chapter 3, Chapter 4 and Chapter 5 presents the physical description and the mathematical
deduction of the equations describing appropriate models of hydraulic turbines, synchronous
generators and turbine governing systems, respectively, for their representation in power
system dynamic studies. Standard models are normally found in considerable number of
relevant literature related to hydraulic power generating system modelling. These chapters
representthebasisforthedynamicmodellingofthedifferentconfigurationsofthetestsystem.
Chapter6dealswiththedevelopmentofimprovedmathematicalmodelsofeachcomponentof
a typical hydraulic power generating system equipped with Francis turbines for their
representationinpowersystemdynamicstudies.
Chapter 7 shows an alternative method of Hydroelectric Power Plant modelling for stability
studies.ThehydraulicturbineandhydroelectricpowerplantmaybemodelledbytheStructure
MatrixMethod.Thismodelcoversawiderangeofparameterssuchasinfluenceoftheturbine
characteristics, frictional damping of oscillatory flow in elastic conduits, influence of the
generatorloadandtheanalysisoftheturbinegoverningsystem.
Chapter8describesfundamentalaspectsandanalyticaltechniquesinthestudyofsmallsignal
stability of dynamic, and identifies factors influencing them. The stability of the generator
infinitebusbarsystemfollowingasmalldisturbanceisdiscussed.
Chapter 9 presents the physical characteristics and capability of the components of the test
system such as water upstream tunnel, surge tank, penstock, hydraulic turbine, speed
governor, generator, and electrical network. The general configuration of the test system
consists of a single synchronous machine connected to a large power system through a
transmissionline.
Chapter10dealswiththestudyofthesystemstabilitycharacteristicsofthedifferentHydraulic
Turbine models within a typical Hydraulic Power Generating System implemented in the
softwareSIMPOW.Theobjectiveofthedynamicmodellingofthetestsystemistoanalyzethe
stabilitycharacteristicsoftheHydroPowerSystemaboutthesteadystateoperatingcondition
followingathreephasefaulttoground.
Chapter 11 studies the dynamic system characteristics of the Hydraulic System, described in
Chapter 9, modelled in the dynamic simulation tool LVTrans. The dynamic behaviour of the
HydraulicPowerSystemisanalyzedintheprogramLVTransversionLVTrans8_1.1.2andversion
LVTrans86_1.3.1_T.
Finally,Chapter12presentsthediscussionofthesimulationresultsofthehydroelectricpower
system models implemented in SIMPOW and LVtrans, and Chapter 13 draws general
conclusionsontheMastersThesisworkandsuggestspossibilitiesfortheapproachandscope
forfurtherwork.
PressureWaterConduitSystem
Dependingonthesite,theconduitsystemiscomposedofawaterdivertingstructure;acanal
tocarrythewaterflowtothereservoir;asurgetank;apenstockpipetoconveythewaterto
thepowerhouse;and,atailracethroughwhichthewaterisreleasedbacktotheriver.Hydraulic
transients in closedconduits consist of pressure disturbances when the system undergoes a
change from one operational steadystate condition to another. The disturbances in a
hydroelectricplantareinitiatedbyanadjustmentinthesettingofacontrolvalveorthechange
in operation of the hydro turbines. The damping of pressure transients is achieved when the
systemenergylossoccursintheformofconduitfrictionorminorlosses.
Hydroelectricpowerplantswithlongconduitsmayhaveseverewaterhammerandgoverning
stabilityproblems.Theclassicalsolutionistoinsertasurgetankupstreamor/anddownstream
thepowerplanttominimizethehydraulictransienteffects.
This chapter contains the mathematical deduction of the equations describing the transient
stateflowsinthesurgetankandclosedconduitsystemsincludingtravellingwaveeffectsand
friction losses. These equations are usually referred to as equations of continuity and
momentum. Hydraulic transient analysis is essential to good design and operation of conduit
systems.
2.1 HydraulicTransientsFundamentals
Hydraulictransientreferstothepressurefluctuationsinthewatercausedduringachangein
meanflowconditions.Themaincomponentsofthedisturbancesarepressureandflowchanges
at a rated point that causes propagation of pressure waves throughout the system. The
pressurewavestravelwiththevelocityofsound,whichdependsonthecharacteristicsofthe
conduitsystemandonthewaterelasticity.Whenaclosedconduitisfilledwithmovingwater,
thelawgoverningthechangesofpressureanddischargedependsupontheconditionsunder
whichtheflowoccurs.[37]
Any disturbance in the water caused during a change in mean flow conditions initiates a
sequence of transient pressure. Typical events of the causes of transients in engineering
systemsinclude:
Pumpstartuporshutdown;
Valveopening,closingorchattering;
Changesinboundarypressures;
Hydraulicturbinestartup,acceptingorrejectingload;
Vibrationsofthevanesofarunneroranimpeller;
Suddenchangesinthecanalinfloworoutflowbyopeningorclosingthecontrolgate;
Damfailureorcollapse;
2.1.1 Pressurewavevelocityinconduits
Thepressurewavevelocityadependsuponthecharacteristicsoftheliquid,suchastheBulk
modulus and density, the characteristics of the pipe material, including the conduit size, wall
thickness and wall material; the external constraints include the type of supports and the
freedomofconduitmovementinthelongitudinaldirection.Thewavevelocityinathinwalled
elasticconduitwithmultiplejointsisgivenbytheclassicalKortewegsequation.[1,6,89]
K
DK
1
eE
(2.1)
Expressionsforafordifferentsupportconditionsandforotherconduitparametersaregivenin
Section22ofthetextbookDavisHandbookofAppliedHydraulics.[9]
Thewavevelocityforaperfectlyrigidpipe,whereEisinfinite,simplifiesto
K
a
(2.2)

Typicalvaluesforwavevelocityareintherangeof1000to1200m/s.
2.1.2 Wavepropagationandreflectionsintheconduit
The wave transmission phenomenon due to the gate closure is characterized by the wave
propagationandreflectionalongapipelineuntiltheyaredampedoutbyfriction.Accordingto
thewaterhammertheorydevelopedbyAllieviintheearly1900s,pressurewavesmayarisein
longpenstockspropagatingwithwavevelocitya.[3,6,1012]
Thetimetakenforthepressurewavetotravelthelengthofthepenstocktotheopensurfaceis
givenby
L
Te
(2.3)
a
Consideringapipelineinwhichtheflowisinitiallysteady.Asuddenriseinthepipeadjacent
gatewillbeproducedwhenaninstantaneousgateclosuretakesplaceatthelowerendofthe
pipe.Thewavetransmissionphenomenonduetothegateclosureisanalyzedthoroughlyinthe
textbookWaterhammerAnalysis.[3]
The theoretical period Tth for a conduit having constant diameter, constant wall thickness is
givenby
4L
Tth
(2.4)
a
Thesequenceofeventsfollowingthevalveclosureisexplainedbrieflybelow.
Table21:Propagationofpressurewavescausedbyinstantaneousclosureofvalve
Stage
0t L a
Events
Theflowvelocityatthevalveisreducedtozero,
Thepipeisexpanded,
Thefluiddensityisincreased
L a t 2 L a Thefluidstartstoflowfromthepipelineintothereservoir.
Anegativewavetravelstowardthevalve.
2 L a t 3 L a The velocity is instantaneously changed due to the valve is
completelyclosed.
Thepressureisreducedandthenegativewavepropagatesin
theupstreamdirection.
3 L a t 4 L a The fluid flows towards the valve and the pressure head is
restored.
2.1.3 HeadlossesduetoFriction
Theheadlossisameasureofthereductioninthetotalheadofthefluidasitmovesthrougha
pipeline.Headlossesarepresentduetothefrictionofthefluidagainstthepipewalls.Thehead
losses are usually the result of boundary losses and form losses. Boundary losses are those
arising from the effects associated with the crosssectional shape that affect the ratio of the
flowareatothewettedperimeter,andfromshearforcesbetweenthefluidandtheboundary
materials. Form losses arise from recirculating eddies produced by the geometry of the
containingvesselsuchasbendsandeitherexpandingorcontractingtransitions.[910,1314]
Headlossesalongthepipewallarecalledfrictionlossesorheadlossesduetofriction.Thehead
loss due to friction Hf in a given conduit for a given discharge is usually determined by the
DarcyWeisbachequation.
L V2
Hf f
D 2g
(2.5)
Thedimensionlessfrictionfactorfisafunctionofvelocity,roughness,viscosityofthefluidand
conduit diameter. The evaluation of the friction factor under these widely varying conditions
has been made possible by the contribution of Reynolds. The Reynolds criterion relates the
inertial forces per unit of volume to the viscous forces per unit of volume. The Reynolds
numberforfullflowingcircularpipescanbeexpressedas
Re
VD
v
(2.6)
The friction factor f in a laminar regime is independent of the wall roughness and inversely
proportionaltotheReynoldsnumber.Thefrictionfactorforalaminarflowiscalculatedfrom
theHagenPoiseuilleequationas
64
Re
(2.7)
Forturbulentflow( Re 4000 ),thefrictionfactorfisafunctionoftheReynoldsnumberReand

the relative roughness height ( / D ). ColebrookWhite equation relates the friction factor to
theReynoldsnumberandrelativeroughnessas

1
2.51
2log10
(2.8)
3.7 D R f
f
e
2.2 ClosedConduitSystems
2.2.1 BasicAssumptions
Hydraulic transients arise as a consequence of rapid variations in the flow conditions in
pressureconduits.Duetotheelasticityofthematerialandthecompressibilityofthefluid,such
variationsinthepressureand/orflowvelocitypropagateintheconduitsatveryhighspeeds.
In order to obtain the basic physical laws of the hydraulic transients in a closed conduit, it is
assumedthat:
The pipe is uniform and flow is onedimensional; for quasiincompressible fluids, the
velocityandpressuredistributionsareuniformineachcrosssectionoftheconduit.
Pipe deformations are proportional to the stresses (Hookes law) and the liquid
compressibilityeffectscanbecharacterizedbyaconstantbulkmodulus.
Novaporizationoftheliquidoccursduringthehydraulictransient.
The formulas used for the calculation of head losses in steady state flow remains valid
duringtransientconditions.
Nodistributedlateralflowsareconsidered.
Theanalysisoftransientsinclosedconduitissubdividedintwotypes:distributedsystemsand
lumped systems. In the former case, the transient phenomenon occurs in form of travelling
waves.Anychangeintheflowconditionisassumedtotakeplaceinstantaneouslythroughout
thesysteminalumpedsystem.
Mathematically,thetransientsinthedistributedsystemsarerepresentedbypartialdifferential
equations,whereasthetransientsinthelumpedsystemsaredescribedbyordinarydifferential
equations.Thesystemmaybeanalyzedasalumpedsystemwhen L a ismuchlessthan1.
2.2.2 BasicDifferentialEquationsforTransientFlow
Equations for the conservation of mass and momentum describe the transient state flows in
closed conduits. The condition of dynamic equilibrium requires that Newtons second law of
motionbesatisfied.Theconditionofcontinuityfortheelementrequiresthatallavailablespace
insidetheboundariesoftheconduitbeoccupiedbywateratalltimes.[3]
The partial differential equations of continuity and motion are expressed in terms of two
dependentvariables,thepressurep(x,t)andflowvelocityV(x,t)andtwoindependentvariables
time(t)anddisplacement(x).
2.2.2.1 EquationofContinuity
Thecontinuitydifferentialequation,whichtakesintoaccountthewatercompressibilityandthe
tubeelasticity,isbasedintheconservationmasslawforacontrolvolume.Theradialvelocity
due to radial expansion and contraction is not included in the analysis. The distance x, flow
discharge Q and velocity V will be considered positive from the left (upstream) to right
(downstream). Heads H are measured relatively to a horizontal datum to which will be also
referredelevationszandslopeofthepipeaxis.Theslopeisconsideredpositiveforarising
pipeinthepositivedirectionofx,asshowninFigure21.ThepipecrosssectionalareaAand
theliquidmassdensityarebothfunctionofpressurep(x,t).[6,89,1516]
V
A
AV
x
) x
V
A
Figure21:Controlvolumeforthederivationoftheequationofcontinuity
Theequationofcontinuityofacontrolvolume c ataninstantt,showninFigure21,yields
AV AV AV x
(2.9)
t
ExpandingEquation(2.9):
A
V
A
(2.10)
A AV A A V AV
t
x
t
t
x
x
x
Rearrangingterms,usingexpressionsforthetotalderivativesanddividingthroughoutby A ,
Equation(2.9)maybewrittenas:
1 dA 1 d V
0
(2.11)
A dt dt x
Foracircularconduithavingradiusr,
dA
dr
2 r
dt
dt
(2.12)
Intermsoftheconduitwalltensilestrain ,Equation(2.12)canbewrittenas
dA
dr
d
2 r 2A
dt
dt
dt
(2.13)
Assumingthattheconduitwallsarealinearlyelastic,then
d
1
D dp
dt E pD 2e dt
2e
(2.14)
ThefirsttermofEquation(2.11)referstochangeinconduitarewithtimebecomes
1 dA
1 D dp
A dt E pD e dt
2e
(2.15)
The second term of Equation (2.11) refers to the rate of change of fluid density. The bulk
modulusofelasticityofafluidisbydefinition
dp
dp
d / d /
(2.16)
andthesecondtermofEquation(2.11)becomes
1 d 1 dp
(2.17)
dt K dt
SubstitutingEquation(2.15)and(2.17)intoEquation(2.11)gives
1
1 dp V
0
(2.18)
eE p
dt
x
D 2
Since p 2 eE D inmostapplications,Equation(2.18)gives
1
1 dp V
0
(2.19)
1 eE
K
dt x
DK
SubstitutingEquation(2.1)andtheexpressionforthetotalderivativeintoEquation(2.19)gives
p
p
V
V
a2
0
(2.20)
t
x
x
8
2.2.2.2 EquationofMotion
ThedifferentialequationofmotionforatransientflowisgivenapplyingthesecondNewton's
lawforafluidelementinsideatube.[17]
0 Dx
p
p x A
pA
p x A
p
x
x 2 x
gAx
Figure22:Controlvolumeforthederivationoftheequationofmotion
TheequationofmotionforthecontrolvolumedefinedinFigure22atagiveninstantt,when
thepipeisdeformedduetotransientforces
p
1 p A
dV
pA p x A p
x x 0 Dx gAx sin Ax
(2.21)
x
dt
2 x x
Omitting x ,Equation(2.21)yields
dV
p
A 0 D gA sin A
0
(2.22)
dt
x
The force resulting from the shear stress 0 may be defined in terms of the DarcyWeisbach
expressionforfrictionlossesforaturbulentpipeflow.
1
0 f V V
(2.23)
8
SubstitutingEquation(2.23)intoEquation(2.22)anddividingthroughout A ,yields
V
V 1 p
fV
V
g sin
V 0
(2.24)
t
x x
2D
2
2.2.2.3 GeneralRemarks
Inmanyengineeringapplications,theconvectiveaccelerationterms V p x and V V x
are usually considered negligible in comparison with the remaining terms. The slope term is
also possible neglect. The system parameters and D remain constant with respect to time.
Therefore,eliminatingthesetermsfromthegeneralEquations(2.20)and(2.24)become
p
V
a2
0
(2.25)
t
x
V 1 p fV
V 0
(2.26)
t x 2D
In the analysis of hydraulic transients, the pressures in the pipelines are usually expressed in
terms of the piezometric head, H, above a specified datum. Additionally the flow velocity is
replacedbythedischarge,Q.Thepressureandtheflowvelocitymaybewrittenas
p g H z
(2.27)
Q VA
(2.28)
Assuming the fluid is slightly compressible and the conduit walls are slightly deformable and
neglectingthevariationof andflowareaAduetovariationoftheinsidepressure.Thesmall
variationsofandflowareaAareindirectlytakenintoaccountbyconsideringthewavespeed
ahaveafinitevalue.SubstitutingEquations(2.27)and(2.28)intoEquations(2.25)and(2.26)
give
2
H z a Q
t t gA x
fQ
1 Q
H z
g

Q 0
2
A t
x x 2 DA
(2.29)
(2.30)
If the pipe axis is fixed z t 0 and z x sin . It is possible to eliminate the tube
inclinationeffect, z x sin 0 .Therefore,Equations(2.29)and(2.30)become
Q gA dH
0
(2.31)
x a 2 dt
H 1 dQ
fQ
Q 0
(2.32)
x gA dt 2 gA2 D
10
Equations(2.31)and(2.32)canbenormalizedbysubstitutingfortherelativeterms h H H 0
and q Q Q0 ,giving
q gA H 0 h
0
(2.33)
x a 2 Q0 t
Q q
Q02
h
f
0
q q 0
(2.34)
x gAH 0 t 2 gA2 D H 0
Thewatertimeconstantorwaterstartingtime,TW,isassociatedwiththeaccelerationtimefor
water in the penstock between the turbine inlet and the reservoir or the surge tank, if one
exists.Theequationforwaterstartingtimeis:
TW
L Q0
gA H 0
(2.35)
IntroducingthewatertimeconstantTWinEquation(2.33)and(2.34)
q 1 L h
0
(2.36)
x a 2 TW t
T Q
h TW q
f W 0 q q 0
(2.37)
x L t
L 2 AD
2.2.3 MathematicalModel
The analysis of transients in hydroelectric systems is subdivided in massoscillation response
andwaterhammerresponse.Ordinarydifferentialequationsdescribethemathematicalmodel
formassoscillation(orrigidwatercolumn)betweenthereservoirand thesurgetank.Partial
differentialequationsshowthemathematicalmodelforwaterhammeralongthepenstockand
drafttube.[1520]
Themathematicalmodeloftheconduitsystemtakingintoaccountthewaterhammertheory
andconsideringtheheadlossesisbasedonthelinearizedEquation(2.36)and(2.37),andon
sinusoidalflowandpressurefluctuations.
UsingtheLaplacetransformtoexpressEquations(2.36)and(2.37)inthefrequencydomain
q s 1 L
2
sh s h 0 0
(2.38)
x
a TW
h s TW
T Q
(2.39)
sq s q 0 f W 0 q s 0
L
L 2 AD
x
11
The third term in Equation (2.39) is related to the head losses due to friction. Assuming an
initialsteadystatesituation, h 0 0 and q 0 0 ,Equation(2.38)and(2.39)gives
q s
x
1 L
sh s 0
a 2 TW
h s TW
L
x
(2.40)
Q0
s f
q s 0
2 AD
DifferentiatingEquation(2.41)withrespecttoxyields
2 h s TW
Q q s
s f 0
2
x
L
2 AD x
(2.41)
(2.42)
Substituting q s x byEquation(2.40)intoEquation(2.42)
2h s 1 2
Q0
s h s
2 s f
2
2 AD
a
x
Introducingthevariablez
fQ0
z2 s2
s
2 DA
ThegeneralsolutionforthesecondorderhomogeneousEquation(2.43)is
h s C1e
z
x
a
C2 e
z
x
a
Bydifferentiatingequation(2.45)withrespecttoxyields
h s
z az x
z az x
C1 e C2 e
a
a
x
Substituting h s x byequation(2.46)intoequation(2.41)gives
(2.43)
(2.44)
(2.45)
(2.46)
z
x
L s az x
a
q s
C1e C2e
aTW z
(2.47)
12
Thewaterpressureandtheflowrateofupstreaminletofthetubeatx=0is
hU s C1 C2
qU s
(2.48)
L s
C1 C2
aTW z
(2.49)
Thewaterpressureandtheflowrateofdownstreamoutletofthetubeatx=Lis
hD s C1e
z
L
a
C2 e
z
L
a
qD s
L s
C1e
aTW z
(2.50)
z
L
a
C2e
z
L
a
(2.51)
Hence,itfollowsfromEquation(2.50)and(2.51)that
1 az L
1 aTW z az L
(2.52)
C1 e hD s
e qD s
2
2 L s
z
z
1 L
1 aTW z a L
(2.53)
C2 e a hD s
e qD s
2
2 L s
The substitution of the values of Te, C1 and C2 from Equations (2.3), (2.52) and (2.53) into
Equations(2.48)and(2.49)yield
hU s
1 zTe zTe
1T z
e e hD s W e zTe ezTe qD s
2
2 Te s
(2.54)
1 Te s zTe zTe
1
e e hD s e zTe ezTe qD s
2 TW z
2
(2.55)
qU s
13
Thetransferfunctionofflowrateandwaterpressureofupstreaminletanddownstreamoutlet
ofthetube,expressedinthematrixnotation,is
cosh zTe
Zc sinh zTe
hU s
hD s
cosh zTe qD s
qU s Z sinh zTe
(2.56)
Inwhichthecharacteristicimpedanceforthepipeis
Zc
TW z
Te s
(2.57)
Theclassicalwavesolutiongivenbyananalysisofthepartialdifferentialequationsintimeand
space defining pressure and flow rate at each point in the conduit taking into account the
elasticwaterhammertheoryandconsideringhydrauliclossasahyperbolictangentfunctionis
12
12
2
h s
TW
fQ0
fQ0
tanh
1
s
s
T
(2.58)
e

2DA
q s
Te 2DAs
Neglectingthehydraulicfrictionlosses,Equation(2.58)canbesimplifiedas
h s
T
W tanh sTe
q s
Te
(2.59)
Thepenstockismodelledassuminganincompressiblefluidandarigidconduitforhydroelectric
power plants with short or medium penstock where the travelling pressure wave effects are
relatively insignificant. With the assumption of an inelastic water column effect,
tanh sTe sTe , Equation (2.58) for small variations around an operating point can be
simplifiedas
hs
TW s H f
q s
(2.60)
Neglectingthehydraulicfrictionlosses,Equation(2.60)yields
h s
TW s
q s
(2.61)
14
2.3 PressureControlSystems
Hydroelectricpowerplantswithlongconduitsmayhaveseverewaterhammerandgoverning
stabilityproblems.Inareactionturbinehavinglongpressuretunnels,thedynamicresponseof
both upstream and downstream water columns must be studied in order to avoid excessive
pressureoroverspeedproblems.Pressuretransientswillonlybeproducedatupstreamsidein
anactionturbine;atdownstreamtherewillbeacanalwithatransientfreesurfaceflow.The
classical solution is to insert a surge tank upstream or/and downstream the power plant to
minimizethehydraulictransienteffects.However,thesurgetankmaybeverycostlystructure
orcausesomeenvironmentalproblems.[6,8]
Asurgetankisanopenstandpipeconnectedtotheconduitsofthehydroelectricpowerplant.
The main functions of a surge tank are reduce the amplitude of pressure fluctuations by
reflecting the incoming pressure waves; improve the regulating characteristics of a hydraulic
turbine;and,storeorprovidewater.
Dependingupontherequirementofthehydroelectricplants,thesurgetankcanbeplacedat
upstreamand/ordownstreamsideofthepowerhouse.
Dependinguponitsconfiguration,asurgetankmaybeclassifiedas:
SimpleSurgeTankisatankoranopenshaftdirectlyconnectedtothepipeline,andthere
isverylittleheadlossbetweenthetankandthepipeline.
AnOrificeSurgeTankthatisasimplesurgetankwiththeentrancerestrictedbymeansof
anorifice.
ADifferentialSurgeTankisanorificetankhavingariser.
In a Oneway Tank; the liquid flows from the tank into the pipeline only when the
pressureinthepipelinedropbelowtheliquidlevelinthesurgetank.
AClosedSurgeTankhasthetopofthetankclosedandthereiscompressedairbetween
thewatersurfaceandthetopofthetank.
2.3.1 GoverningEquations
Tosimplifythederivationofthedynamicandcontinuityequationsthatdescribetheoscillations
ofthewaterlevelinthetankthefollowingassumptionsaremade;[6]
Theconduitwallsarerigidandtheliquidiscompressible.Thismeansthataflowchange
atanypointinthesystemistransmittedinstantaneouslythroughoutthesystem,andthe
liquidmoveslikeasolidslug
The inertia of the liquid in the surge tank is small compared to that of the liquid in the
tunnelandcanthereforebeneglected.
Theheadslossesinthesystemduringthetransientstatecanbecomputedbyusingthe
steadystateformulasforthecorrespondingflowvelocities.
Figure 23 depicted a typical simple surge tank system. In a simple tank, the tank is directly
connected to the pipeline. A flow variation results in the oscillation of the liquid level in the
surgetank.
15
Figure23:Simplesurgetank,[6]
ThedynamicequationisgivenapplyingtheNewtonssecondlawofmotion;therateofchange
ofmomentumisequaltotheresultantforce.Thedynamicequationforasimplesurgetankis
givenasfollows
dQt gAt
(2.62)
H z fQ Q
dt
L
Theequationofcontinuityforthejunctionofthetunnelandthesurgetank,showninFigure
23,maybewrittenas
Qt Qs Qtur
(2.63)
whereQsistheflowintothesurgetank,Qturistheturbineflow.Since Qs As dz dt ,Equation
(2.63)becomes
dz 1
Qt Qtur
(2.64)
dt As
Equations (2.62) and (2.64) are for a simple surge tank located at the upstream side of a
turbine.Theseequationsarevalidforatailracesurgetank.
The period and the amplitude of the oscillations of the water surface in the surge tank for a
frictionlesssystemaregivenby:
L AS
T 2
(2.65)
g At
Z Q0
L
gAS At
(2.66)
16
2.3.2 SurgeTankMathematicalModel
Inasimplereservoirsurgetanksystemtheunsteadyflowcanbemodelledbyasimpleordinary
differentialequation.Theinclusionofsurgetankeffectsiswarrantedincaseswheredynamic
performance is being simulated over many seconds to minutes. The surge tank equation is
derived from the continuity of flow at the two junctions, Equation (2.64), and where the
hydrauliclossesatorificesofthesurgetankareneglected.NormalizingEquation(2.64)yields
A H dh
q s 0
(2.67)
Q0 dt
Thesurgetankfillingtime,Ts,isdefinedas
AH
Ts s 0
(2.68)
Q0
Thetransferfunctionofflowrateandwaterpressureofupstreamthesurgetankis
h s 1
q s sTs
(2.69)
17
HydraulicTurbines
This chapter contains the physical description of the hydraulic turbines and the deduction of
the equations describing the appropriate mathematical models of the hydraulic turbine for
theirrepresentationinpowersystemdynamicstudies.Theturbinedynamicsarecharacterized
by the variations in flow and output mechanical torque with respect to turbine speed, gate
opening,runnerblademovementandthedifferenceinpressurebetweentheturbineinletand
outlet. Standard models are normally found in considerable number of relevant literature
related hydropower plants modelling. Approaches based on approximate linear models
assuming an ideal lossless turbine and ignoring the elasticity of the conduit system, are not
suitable for the accurate study of the interaction between the hydraulic system and power
system.
3.1 HydraulicTurbinesOverview
Hydraulicturbinesderivethepotentialenergyofthefluidintokineticenergyandaconversion
of kinetic energy, or both kinetic and potential energy, into useful work. Hydraulic turbines
derive power from the force exerted by water as it falls from an upper to a lower reservoir.
Hydraulicturbinesaredividedaccordingtotheirhydraulicactionintotwomainclasses:impulse
turbinesandreactionturbines.[1,6,8,10,2122]
3.1.1 Impulseturbines
An impulse turbine has a runner with numerous spoonshaped buckets attached to its
periphery, which are driven by one or more jets of water issuing from fixed or adjustable
nozzles.Thekineticenergyisintheformofahighspeedjetthatstrikesthebuckets,mounted
ontheperipheryoftherunner.Asthewaterafterstrikingthebucketsfallsintothetailwater
with little remaining energy, the casing can be light and serves the purpose of preventing
splashing. Many impulse turbines use deflectors to allow rapid reduction of the power
deliveredtotheturbinerunnerincaseswheretheflowcannotbechangedquickly.
Figure31:ImpulseTurbine
18
ImpulseoractiontypeturbinesarerepresentedbythePeltonwaterwheel,Turgoturbinesand
Crossflowturbines.Impulseturbinesareusedinhighhead,300metersormore,hydroelectric
powerplants.
Pressure transients in impulse turbine will only be induced at upstream side; at downstream
therewillbeacanalwithatransientfreesurfaceflow.Eachnozzlewillthenactuatelikeaflow
controlvalve.
3.1.2 ReactionTurbines
In reaction turbines, the entire flow from the headwater to tailwater takes place in a closed
conduitsystem.Reactionturbinesextractpowerfromthekineticenergyofwaterbecauseof
thedifferenceinpressurebetweenthefrontandthebackofeachrunnerbladeasthewater
flows through the runner. The water pressure can apply a force on the face of the runner
blades.Theturbinecasing,withtherunnerfullyimmersedinwater,mustbestrongenoughto
withstandtheoperatingpressure.
Reaction turbines are represented by radialflow Francis turbines; axialflow Kaplan turbines
withadjustablerunnerblades;axialflowpropellerturbineswithfixedrunnerblades;diagonal
flowDeriazturbineswithadjustablerunnerblades.
Figure32:FrancisTurbine
Upstreamanddownstreamwatercolumnsaresensitivetoflowdisturbancesatreactiontype
turbine.Pressureregulatorsorreliefvalvesareusedtotemporarilyallowpartofthewaterto
bypass the runner by flowing from the turbine spiral case directly to the tailrace during the
quick closure of the wicket gates. The pressure regulators or relief valves prevent excessive
water hammer, which is a severe pressure rise that results from rapid deceleration of the
penstockwatercolumn.
19
3.2 GeneralTechnicalAspects
Thehydroelectricgenerationisextremelysitedependentbecauseitdependsonfallingwater.
ThebasicelementsoftheHydraulicPowerPlantaredepictedinFigure33.Theturbineproper
is,accordingtostandardpractice,takentobeginattheentrancetotheturbinecasingandto
endatasectionofthetailracejustbeyondthephysicalendofthedrafttube.[9]
Figure33:SchematicofHydraulicPowerPlantwithareactionturbine
3.2.1 NeatHead,PowerandEfficiency
The Neat Head is the vertical height between the entrance level into the penstock and the
dischargelevelfromtheturbinedrafttube.
Themechanicalpowerdevelopedbytheturbineisproportionaltotheproductoftheflowrate,
theheadandtheefficiency.Thepoweriscontrolledbyregulatingtheflowintotheturbineby
wicket gates on the reaction turbines and by a needle on the impulse turbine. The nominal
powerisgivenbythefollowingequation
P gQH
(3.1)
The turbine efficiency represents the actual utilization of the available potential energy of
the system. The turbine efficiency depends on the working fluid flow rate and the turbine
characteristics.Theturbineefficiencyisdefinedastheratioofpowersuppliedbytheturbine
(mechanicalpowertransmittedbytheturbineshaft)totheabsorbedpower(hydraulicpower
equivalenttothemeasureddischargeunderthenethead).[23]
Thegeneratorefficiencyisusuallyconsideredconstantoverawideoperatingrange.However,
the hydraulic efficiency depends significantly on both the water discharge and the net head.
TheserelationshipsaregenerallyrepresentedinthesocalledTurbineHillChart.
20
3.2.2 TurbineHillCharts
ThemathematicalmodellingofthedatafortheflowandpoweroutputofaFrancisturbine,and
itstransientbehaviourisbasedonthecharacteristiccurveofthehydraulicmachinecalledHill
Charts. The Turbine Characteristics are curves representing the relationship between the net
head and discharge. Although such curves are seldom used in specifications due to of
difficulties of measurement, the speedtorque characteristics of a wide variety of hydraulic
turbineshavethegeneralformwiththetorquefallingoffroughlyinproportiontospeedover
widerangesofspeedandgateopeningposition.[6,18,2427]
Staticcharacteristicrelationshipsofhydraulicturbinescanbestudiedthroughthesocalledhill
charts.Theplotsoftheprototypeturbinecharacteristicsarebasedinsteadystatemodeltest
results.Theseturbinecharacteristicsareassumedvalidduringthetransientstate.Theturbine
efficiency for any operating point given by runner speed, net head and gate position can be
extractedfromthehillcharts.
Anoperatingpointofahydraulicturbineischaracterizedbythespecificenergy,thedischarge,
the rotational speed, the torque and the gate opening position. Therefore, the graphical
representationofaturbinecharacteristicrequirestheeliminationofoneofthesequantitiesby
theuseofthehydraulicmachinessimilitudelaws.Fortransientanalysis,itismoreconvenient
to use dimensional factors where the specific energy is eliminated, [20]. The mathematical
modelofthecharacteristicofhydraulicturbinemaybewrittenas:
Q11 f n11 , y
(3.2)
M11 f n11 , y
(3.3)
However,therearestillnodetailedandprecisemathematicalmodelsthatcandescribetheunit
flowrateQ11andunittorqueM11.Q11andM11areusuallyobtainedbyinterpolationmethod.
Theexpressionfortheunitrotationspeedn11,andunitflowQ11maybewrittenasfollows
n11
nDref
Q11
(3.4)
Q
2
ref
(3.5)
Hn
M
M 11 3
Dref H n
(3.6)
Thecharacteristiccurvesareextrapolatedforsmallwicketgateopenings.Therefore,theflow
should be known when the turbine rotational speed is zero. The windage and friction losses
shouldbeknownatwicketgateopeningsbelowthespeednoloadgate.Thenoloadgateisthe
lowestgateopeningatwhichturbinerotatesatsynchronousspeedwithzerooutput.[6]
AtypicalHillChartforaFrancisturbineisshownonFigure5.2ofChapter5,page131,ofthe
textbookAppliedHydraulicTransients.[6]
21
3.3 HydraulicTurbineModelling
The hydraulic turbine dynamics have a considerable influence on the dynamic stability of the
power system. The block diagram of Figure 34 shows the basic elements of a hydro turbine
withinthepowersystemenvironment.
Figure34:FunctionalBlockDiagramofthehydroturbine,[1,28]
Thepowerdevelopedbytheturbineisafunctionofthewaterflow,therunnerbladeangleand
thenethead.Theflowthroughtheturbineisafunctionofthenethead,rotationalspeed,gate
openingpositionandrunnerbladeangle.ThenonlinearcharacteristicofaFrancisturbinecan
bewrittenas,[1,6,8,16,23,25,2831]
p f p h, n, y,
(3.7)
q f q h, n, y,
(3.8)
Theeffectoftherunnerblademovementisnotconsideredinthemodellingofturbineswith
fixedblades,suchasFrancisturbines.Theflowinanimpulseturbinedependsuponthehead
andthenozzleopeningonly.
The turbine dynamics is related to the generator dynamics through mechanical power Pm
produced by the turbine. The relation between mechanical power Pm and mechanical torque
Mm,expressedinperunit,isgivenby
n
pm
m
(3.9)
m
s
Normally,asimplifyingassumptionmadeisthat n s atsynchronousspeed.Thisisnotthe
sameassayingthatthespeedisconstant,itassumesthatspeedchangesaresmallanddonot
haveasignificanteffect.Thus,equation(3.9)yields
pm mm
(3.10)
22
3.3.1 SimplifiedNonlinearTurbinemodel
Thenominalpowerobtainedbyahydraulicturbineisproportionaltothepotentialenergylost
bythefallingfluid.Duetotheturbineisnot100%efficient,thenoloadflowqNLissubtracted
fromtheactualflow.Adampingeffectthatisdependentongateopeningisalso present,so
thatatanyloadconditiontheturbinepowercanbeexpressedby
pm At h q qNL K D N 1 y
(3.11)
Detailed information is available from turbine manufacturers on the variation of the turbine
efficiency with variation in the steady operating point of a turbine. The use of this data
involvestheuseofastoredtableandaninterpolationprocedure.Neglectingthevariationof
turbineefficiency,Equation(3.11)yields
pm At h q qNL K D N 1 y
(3.12)
TheconstantproportionalityfactorAtiscalculatedusingtheturbineMWratingandgenerator
MVAbase.Atiscalculatedby
turbine power MW
1
At
(3.13)
h q q generator MVA rating
NL
Thepressureheadacrosstheturbineisrelatedtotheperunitflowratebyassumingthatthe
turbinecanberepresentedbythevalvecharacteristic
(3.14)
q y h
3.3.2 Thelinearizedhydroturbinemodel
The most common hydraulic turbine representation for system stability studies consists of a
transferfunctionobtainedbylinearizingtheturbinecharacteristiccurvesaroundanoperating
point. The dynamics of the analyzed system with respect to the operating point can be
observed by changing operating point. For the controller synthesis using linear synthesis
method, Equations (3.7) and (3.8) can be represented by the Taylor series approximation for
smallvariationsinthevicinityofanoperatingpointas,[2526,3237]
m
m
m
m
h
n
y
(3.15)
h
n
y
q
q
q
h
n
y
h
n
y
(3.16)
23
The partial derivatives of flow and torque with respect to head, rotational speed and gate
position are called Turbine Coefficients. The turbine coefficients represent the nonlinear
characteristicsofahydraulicturbine.Thepartialderivativesvarywiththeoperatingconditions
oftheturbinerelativetogateopeningpositionyandunitrotationalspeedn11.
Theinfluenceoftheturbinecoefficientsonthemodelaccuracyiscritical.Thesevalueshaveto
bemeasuredaccuratelyonthefieldortakenfrommodeltests.
The turbine coefficients of any operating point can be obtained by interpolation before
simulation.Fordeviationsaroundratedspeedandpressurehead,theturbinecoefficientsare
deducedbydifferentiationofEquations(3.1)and(3.14),[11].DifferentiatingEquations(3.12)
and (3.14), the turbine coefficients are deduced for models neglecting the variation of the
turbineefficiency,[37].TheseparametersaredetailedinTable31.
Table31:Turbinecoefficients,[11,37]
Partial
IEEE
Derivatives Model,[11]
q
0.5 y0
h
q
0.0
n
q
1.0
y
m
1.5 y0
h
m
n
m
y
IEEE
Model,[37]
y
2 h
0.0
h
At y h qNL
2
Da y
0.0
At h
Da N 1
The partial derivative of torque with respect to gate position, m y , is called Turbine Gain.
Turbine Gain is a critical parameter for an accurate approximation of hydro power plants
dynamics, and has to be measured precisely in the field. The partial derivative of flow with
respect to rotational speed, q n , is usually considered to be negligible. The deviation of
mechanical torque with rotational speed, m n , is known as Turbine Self Regulation. The
valueoftheturbineselfregulationisnegativewithanabsolutevalueusuallynearunity.
Standard values for an ideal lossless turbine at full load and three linearized values of the
turbinecoefficientsareshowninofTable32.Moreaccuratevaluesofturbineparameterscan
be deduced from the Characteristic Curves of the turbine socalled Hill Charts, detailed in
Chapter3.2.2.
24
Table32:TurbineCoefficients,[32]
GeneratorOperatingConditions
Partial
IEEE
Derivatives 22.5(MW) 84.3(MW) 112.0(MW) model
q
0.06
0.20
0.24
0.50
h
q
0.13
0.38
0.62
0.00
n
q
0.80
0.40
0.38
1.00
y
q
0.00
0.30
0.69
0.00

m
0.40
1.20
1.50
1.50
h
m
0.39
0.86
0.75
0.00
n
m
0.88
0.90
0.34
1.00
y
m
0.00
0.50
0.52
0.00
0.00
2.30
1.00
0.00
y
ItisworthytonoticethatthelinearmodelsimplementedintheMastersThesisworkusesthe
standardvaluesaccordingtoIEEEmodeldetailedinTable32.
25
SynchronousMachine
This chapter deals with the mathematical modelling of the synchronous generator.
Synchronous machines play an important role in power system stability. The physical
characteristicsofthesynchronousgeneratorsandtheirperformanceaffectthesystemstability.
Thecompletemathematicalmodellingofthesynchronousmachineisfairlycomplexsystemfor
stabilityanalysisofpowersystems.Theclassicalmodelisusedforsimplifiedanalysisofpower
systemdynamics.
4.1 SynchronousGenerator
Synchronous generators are the main source of electric energy in power systems. Hence,
understandingofthetheoryandperformanceofthesynchronousmachinesisfundamentalto
the study of power system stability. Synchronous generators can be classified as either high
speed generators, called turbo generators, driven by steam or gas, or lowspeed generators
drivenbyhydraulicturbines.
Thesynchronousmachineisassumedtohaveathreephasestatorarmaturewinding,arotor
field winding and two rotor damper winding one in the daxis and one in the qaxis. The
armaturewindingcarriestheloadcurrentandsuppliesthepowertothesystem.Thearmature
winding usually operates at a voltage considerable higher than the field voltage. The rotor
excitation winding is supplied with a direct current to produce a rotating magnetic flux. The
rotordamperwindinghelpsdampmechanicaloscillationsoftherotor.Thenotationfollowsthe
normalIEEEconvention.
The following assumptions are made in order to develop the mathematical model of a
synchronousmachine.[2,22,3840]
a) Thethreephasestatorwindingissymmetricaldistributed.
b) Thecapacitanceofallthewindingscanbeneglected.
c) Eachofthedistributedwindingsmayberepresentedbyaconcentratedwinding.
d) Thechangeintheinductanceofthestatorwindingsduetorotorpositionissinusoidal
anddoesnotcontainhigherharmonics.
e) Hysteresis loss is negligible but the influence of eddy currents can be included in the
modelofthedamperwindings.
f) Inthetransientandsubtransientstatestherotorspeedisnearsynchronousspeed.
g) The magnetic circuits are linear (not saturated) and the inductance values do not
dependonthecurrent.
26
4.2 SynchronousGeneratorEquations
Thegeneratorwindingsaremagneticallycoupled,hence,thefluxineachwindingdependson
the currents in all the windings. The electrical dynamic performance of the machine is
describedbythefluxcurrentrelationsandthevoltageequations.Thetransformationofallthe
generator windings into the rotor reference frame is referred as the 0dq transformation or
Parkstransformation.[4143]
Theelectricaldynamicperformanceofthemachineintermsofthedq0coordinatesystemmay
bedescribedbythefollowingsetofequations.
Fluxlinkageequations
0
0
Lad
Lad
0 id
d Lad Ll
0
Laq Ll 0
0
0
Laq iq
q
0 0
0
L0
0
0
0 i0
(4.1)

0
0 Lad L f
Lad
0 if
f Lad
D Lad
0
0
Lad
Lad LD
0 iD

Laq
0
0
0
Laq LQ iQ
Q 0
Statorvoltageequations:
vd
d
q Ra 0 0 id
1
d

(4.2)
q dt q d 0 Ra 0 iq
s
s
v0
0
0 0 0 Ra i0
Rotorvoltageequations:
f R f 0
0 i f
v f

0 1 d 0 R
0 iD
(4.3)
D
dt D
s
Q 0 0 RQ iQ
0

Thetransformeremfs,the d dt terms,arereferredtothearmatureemfsproportionaltothe
ratechangeoftheflux.Thetransformeremfsareduetochangingcurrentsincoilsonthesame
axisastheoneconsidered.
Inadditiontotheequationsdescribingtheelectricaldynamicperformance,anexpressionfor
theelectromechanicaltorqueisrequired.
me
s
i qid
dq
(4.4)
27
RotormechanicalEquations:
Therotormechanicaldynamicaregivenbytheswingequation,whichcanbeexpressedvia,[2]:
d s
(4.5)
mm me K D
2H
dt
s
(4.6)
Intheaboveequations,allquantitiesareinperunitexcepttimethatisinsecondsandands
which are in rad/sec. In some literature, per unit speed N s is used. Therefore,
Equations(4.5)and(4.6)yield
d N
1
(4.7)
mm me KD N 1
2H
dt
s N 1
(4.8)
4.3 SynchronousGeneratorModels
The mathematical model of a synchronous machine is made up of Parks equations for the
electrical dynamics, Equations (4.1), (4.2) and (4.3), and, two equations for mechanical
dynamics,Equations(4.5)and(4.6).[4445]
The larger number of differential equations required in the model, the greater the model
complexityandthegreaterthetimerequiredtosolvethedifferentialequations.Thecomplete
ninth order nonlinear model is a complex system for stability analysis of power systems. It is
highlydesirable,torephraseandsimplifythefullsetofgeneratormodelinordertogetamore
acceptableformandeasiertointerfacetothepowersystemnetworkequations,[22].Someof
theessentialsimplificationsare:
Therearenozerosequencecurrentsinbalancedoperations.
Changesinthegeneratorspeedaresmall, s .
The transformer emfs ( d d dt and d q dt ) are neglected because are small
comparedwiththerotationemfs( q and d ).
Thedifferentialequationsfortheelectricaldynamics,expressedinorthogonalphasequantities,
usedtomodelanumberofdifferentmodelsare
do
dEq
dt
Eq Eq X d X d Id
(4.9)
Tqo
dEd
Ed Ed X q X q I q
dt
(4.10)
28
do
dEq
dt
E f Eq X d X d Id
(4.11)
dEd
T
Ed X q X q I q
dt
qo
(4.12)
4.3.1 SynchronousMachineRepresentedbytheClassicalModel
Thesecondordersynchronousgeneratormodel,referredtoastheclassicalmodel,iswidely
usedforsimplifiedanalysisofpowersystemdynamics.Intheclassicalmodelallthevoltages,
emfs and the currents are phasors in the network reference frame rather than their
componentsresolvedalongthedandqaxis.
AssumethatthegeneratorisconnectedtotheinfinitebusasshowninFigure41andthatall
the resistances and shunt impedances associated with the transformer and network are
neglected. The time constant Tdo is relatively long and the changes in Ef and Id are small.
Additionallytherotortransientsaliencyisneglected.
E '
Figure41:Equivalentcircuitoftheclassicalmodelofthegenerator
TheclassicalsynchronousgeneratormodelassumesthatneitherthedaxisarmaturecurrentId
nor the internal emf Ef representing the excitation voltage change very much during the
transientstate.Inthismodel,thegeneratorisrepresentedbyaconstantemfEthroughoutthe
studyperiodbehindthetransientreactanceXdandtherotorinertiaequations,Equations(4.5)
and(4.6).[22]
Theequationdescribingthearmaturevoltageyields
V Eq jEd jX d I q jI d E jX d I
(4.13)
The simplest secondorder model is very approximate and only suitable for representing
remote machines in the analysis of very large interconnected power systems or generators
locatedalongwayfromthepointofdisturbance.Itcanalsobeusefulforevaluatinggenerator
behaviourduringthefirstrotorswing.
29
TurbineGoverningSystems
The governing system is a system that regulates the turbinegovernor speed and hence the
frequencyandtheactivepowerinresponsetoloadvariation.Theturbinegovernorregulates
theinletofwaterintoaturbine,whichinturnrotatesthegeneratortoproduceelectricity.This
chapter details the modelling of the mechanicalhydraulic and electrohydraulic turbine
governingsystem.
5.1 GoverningSystem
Themainfunctionoftheturbinegoverningsystemistoregulatetheturbinegeneratorspeed
andhencethefrequencyandtheactivepowerinresponsetoloadvariation.Thespeedcontrol
mechanism includes equipment such as relays, servomotors, pressure or power amplifying
devices, levers and linkages between the speed governor and governorcontrolled gates. The
speedgovernornormallyactuatesthegovernorcontrolledgatesthatregulatethewaterinput
totheturbinethroughthespeedcontrolmechanism.[12,6,2122,4649]
Hydroturbinegoverningsystemsarestronglyinfluencedbytheeffectsofwaterinertiaand,as
aresult,twoservomotorsareusedtoprovidetherequiredforcetomovethecontrolgate.The
first pilot servomotor, low power, operates the distributor or relay valve of the second main
gateservomotor,highpower.Thepilotservomotorhasapilotvalvethatiscontrolledeitherby
amechanicalgovernororbyanelectronicregulator.Theoutputofthespeedsensingdevicesis
thedeviationfromthereferencespeed.[22]
ThepermanentspeeddroopRPdeterminestheamountofchangeinoutputaunitproducesin
responsetoachangeinunitspeed.Thepermanentspeeddroopcanbedevelopedeitherby
using the wicket gate position or by using the unit output power. For isolated or islanded
operation,thepermanentspeeddroopcanhelptostabilizetheunitspeedcontrolbyproviding
anintermediatefeedbackthatlimitstheovertraveloftheturbinecontrolservomotorswhile
controllingtheunitspeed.Thepermanentspeeddroopdeterminestheamountofparticipation
the unit produces when responding to disturbances in system frequency in operation while
synchronizedtoaninterconnectedpowersystem.
ThetemporarydroopRTisusedtolimitovershootoftheturbinecontrolservomotorduringa
transientcondition.Thetemporarydroopmaybedevelopedeitherconnectingadashpotfrom
thewicketgatepositiontothegovernorerrorsummingpoint,oraddingafilteredderivativeof
wicketgatepositiontothegovernorerrorsummingpoint.
Hydro turbine governors are designed to have relatively large transient droop compensation
with long resetting times. This ensures stable frequency regulation under isolated operating
conditions.Theresponseofahydroturbinetospeedchangesortochangesinspeedchanger
settingisrelativelyslow.
30
5.2 MechanicalhydraulicGovernor
TheMechanicalHydraulicGoverningSystemusestheWattcentrifugalmechanismasthespeed
governor. Speed sensing, permanent droop feedback and computing functions are achieved
through mechanical components and functions involving higher power are achieved through
hydraulic components. The dashpot feedback is required in order to achieve stable
performance.Theschematicdiagramofthemechanicalhydraulicgoverningsystemisshownin
Figure51.
Figure51:SchematicdiagramoftheGoverningSystem,[48]
ThemaindisadvantagesoftheWattcentrifugalgovernorarethepresentofdeadbandsanda
relativelylowaccuracy.Thesizeofthedeadbandsalsotendstoincreasewithtimeduetowear
inthemovingmechanicalelements.Theconventionalmechanicalhydraulicgoverningsystem
hasbeenreplacedbyanelectrohydraulicgovernor.
5.2.1 MathematicalModelling
Figure 52 shows a simplified mechanicalhydraulic governor. The variables used in the
derivationofthetransferfunctionsareperunitderivativesfromtheinitialsteadystatevalues.
[2,22,47]
Figure52:MechanicalHydraulicGoverningSystem.[47]
31
Thetransferfunctionofthedistributingvalveandgateservomotoris
y K1
(5.1)
b
s
Thetransferfunctionofthepilotvalveandpilotservomotoris
b
K2
(5.2)
a 1 Tp s
whereK2isdeterminedbythefeedbackleverratioandTpbytheportareasofthepilotvalve
andK2.CombiningEquations(5.1)and(5.2)yields
Ks
K1 K 2
y
(5.3)
a s 1 T s s 1 T s
P
TheservomotorgainKsisdeterminedbythepilotvalvefeedbackleverratioandtheportareas
of the distributing valve and gate servomotor. The governor response time, Tg 1 K s , is
related to the gate closing time Tc, where Tc is the time required for fullgate travel with the
gatesmovingatmaximumvelocity,typically5to10seconds.
Assumingthattheflowofdashpotfluidthroughtheneedlevalveisproportionaltothedashpot
pressure,thedashpottransferfunctionis
c
TR s
RT
(5.4)
y
1 TR s
ThetemporarydroopRTisdeterminedbytheselectionofpivotpointfortheleverconnectedto
theinputpiston.TheresettimeTRisdeterminedbytheneedlevalvesetting.
The pilot valve input signal is produced adding the action of a system of floating levers, the
referencespeed,shaftspeed,permanentdroop,andtemporarydroopsignals.
TR s
a ref s RP y RT
(5.5)
1 T s
R
AcombinationofEquations(5.3),(5.4)and(5.5)gives
1
1 TR s
y
RP
ref s TPTRTg s 3 TP TR Tg s 2 Tg TR RP RT s
1
RP
RP
RP
(5.6)
32
Theblockdiagramofatypicalhydroturbinegoverningsystemsuitableforstabilityanalysisis
showninFigure53.
1
1 sTP
KS
1
s
1
1 sTg
RP
RT
sTR
1 sTR
Figure53:Modelofgovernorforhydraulicturbines,[2]
As the pilot valve and servomotor time constant TP is several times smaller than the time
constantsTgandTR,itmaybeneglectedtogivethesecondordertransferfunction
1
1 TR s
RP
y
(5.7)
ref s Tg TR RP RT
TRTg
s 1
s
1
RP
Tg TR RP RT
Forstableoperationunderislandingconditions,theoptimumchoiceofthetemporarydroopRT
andresettimeTRisgivenby;[2,50]
T
RT 2.3 TW 1.0 0.15 W
(5.8)
TM
TR 5.00 TW 1.00 0.50 TW

(5.9)
Table51givestypicalvaluesandrangeofparametersforhydraulicgovernors.
Table51:TypicalvaluesandrangeofParameters,[47]
TR
Tg
TW
TM(2H)
Tc
RP
R T
Typicalvalues
5.00
0.20
1.00
8.00
5.00
0.04
0.31
Range
2.5025.0
0.200.40
0.505.00
6.0012.0
5.0010.0
0.030.06
0.201.00
33
5.3 ElectroHydraulicGoverningSystem
The ElectroHydraulic Governing System operation is very similar to that of mechanical
hydraulic governors. Speed sensing, droop compensation, and computing functions are
performed electrically. The output signal drives an electricalmechanical transducer, which
operates a pilot valve and pilotvalve servomotor. The turbine rotor speed is measured
electronicallywithhighaccuracy.[2,18,4749]
5.3.1 MathematicalModel
The ElectroHydraulic governor uses the threeterm controllers with proportionalintegral
derivative action, PIDcontroller, to perform the lowpower functions. The proportional term
produces a control action proportional to the size of the error input, and an immediate
responsetoanerrorlevelinput.Theproportionaltermresponsehasasignificantinfluenceon
the stability of the governed system. The integral term produces a control action that
accumulatesatarateproportionaltothesizeoftheerrorinput.Theintegraltermalsotrims
out the error input to the governor controller to determine the steadystate accuracy of the
governedsystem.Thederivativetermproducesacontrolactionthatisproportionaltotherate
of change of the error input. The derivative term helps to extend the stability limits of the
governedsystembyallowinghigherproportionalandintegralgainswhilemaintainingastable
controlsystem.[1,21,35,40,47,5152]
The block diagram of a PID controller is shown in Figure 54. The permanentdroop feedback
may be obtained from generator terminal power rather than from gate position to eliminate
thenonlinearityofthegatepositionversuspowercurve.[47]
KP
Speed
Ref.
+
Ki
s
+
-
Speed
Pilot
Servo
Gate
Servo
Gate
sKd
RP
Figure54:TypicalPIDGovernorController,[2]
The transfer function of the PID governor expressed in terms of proportional, integral and
derivativegains,is
K d s 2 K p s Ki
1
y
(5.10)
ref s RP K d s 2 K p 1 RP s K i
34
Theoptimumparameterssuggestedin[53]areexpressedasfollows
T
K p 0.80 M
(5.11)
TW
T
K i 0.24 M2
(5.12)
TW
K d 0.27TM
(5.13)
Theuseofahighderivativegainwillresultinexcessiveoscillationsandpossiblyinstabilitywhen
thegeneratingunitisconnectedtoaninterconnectedsystem.Whenthederivativegainisset
tozero,thetransferfunctionofPIDcontrollerisequivalenttothatofthemechanicalhydraulic
governor.Thedesignisbasedonlinearcontroltheoryatoneloadconditionandthendetuned
forworstoperatingconditions.Thiscontrollerdesigndoesnotguaranteethecloseloopsystem
toremainstableatalloperatingconditions.ThetransferfunctionofthePIgovernoryields
K p s Ki
y
1
ref s RP
(5.14)
1
Kp
s Ki
RP
Addingasmalltimeconstant,Equation(5.14)maybeimprovedas
K
1 p s
Ki
1
y
(5.15)
ref s RP K p RP 1
2
s 1 0.1s
1
K i RP
35
HydraulicPowerPlantModels
Hydroelectric power generating system is a highorder, uncertain and nonlinear system. The
power output of a hydro power plant varies constantly according to the network needs that
involveareactionofthecontrolsystem.Therepresentationofeachcomponentofthepower
system by means of appropriate mathematical models is essential for the understanding of
systemstability.
Thischapterdealswiththedevelopmentofadvancedmathematicalmodelsofeachcomponent
of a typical Hydraulic Power Generating System equipped with Francis turbines for their
representationinpowersystemdynamicstudies.
The stability analysis of different models for synchronous machines and turbine governing
systemarebeyondofthescopeoftheMastersThesiswork.Theenhancementofthehydraulic
powergeneratingsystemisstudiedbymeansofanalysisofthedynamicbehaviourofdifferent
modelsofthehydraulicmachineandwaterconduitsystem.
Appropriate models of the hydraulic turbine and water conduit system are implemented in
various degrees of detail. Firstly, nonlinear models for a simple turbine without surge tank
considering the inelastic and elastic travelling wave effects have been developed. After that,
nonlinearmodelsconsideringtheinelasticandelastictravellingwaveeffectsforaturbinewith
surgetankforHydropowerSystemswithlong lengthpenstocksareimplemented.Finally,the
nonlinearmodelsforaturbinewithlonglengthpenstocksarelinearizedatanoperatingpoint
aredevelopedconsideringthenonlinearturbinecharacteristicsandthetravellingwaveeffects.
6.1 HydropowerPlantModels
A hydroelectric development includes in some form a waterdiverting structure, conduit to
carry the water to the turbines and governors, generators, control and switching apparatus,
housing for the equipment, transformers, and electrical network. Depending upon the
requirement of the hydroelectric plants, a surge tank can be placed at upstream and/or
downstreamsideofthepowerhouse.
Ahydroelectricpowerplantmaybeconsideredasasystemconsistingofseveralsubsystems:
the penstock including any surge chamber, the hydraulic machine, the speed governor, the
tailraceand,finally,thegeneratorandelectricalnetwork.ThediagramofFigure61depictsthe
functionalblocksfortheelementsofaHydraulicTurbineGeneratingSystem.
Theturbinegoverningsystemismodelledbyamechanicalhydraulicturbinegoverningmodel
and it is detailed in Chapter 5. The subsystem of the governor control considers the turbine
speeddeadzone,valvesaturation,andlimitation.Theblockdiagramofatypicalhydroturbine
governingsystemusedforthestabilityanalysisisdepictedinFigure53.Thetransferfunction
ofamechanicalhydraulicgovernorisgivenbyEquation(5.6).
36
For stability and control studies, the synchronous generator is modelled by a detailed model
with damping, which consists of one field winding, one damper winding in daxis and one
damperwindinginqaxiswithsaturationincluded.Themodellingofthesynchronousmachine
isdetailedinChapter4.
0
s
Figure61:FunctionalBlockDiagramoftheHydraulicTurbineGeneratingSystem
The most common classification of the hydraulic power plant models consists in linear and
nonlinearmodels.Thisclassificationisbasedonthecomplexityoftheequationsinvolvedinthe
models. The accurate mathematical models of the hydraulic components of a hydro power
plant include the dynamic representation of penstock, surge tank, upstream and/or
downstreamtunnelandthehydrauliclossesofalltheseelements.[1,6,11,18,24,2829,31
32,3637,42,46,48,50,5464]
Thehydraulicpowergeneratingsystemforthestabilityanalysishasbeendesignedwithathird
order synchronous generator model, a mechanicalhydraulic governing system model, and a
hydraulicturbine.
Thehydraulicturbineismodelledwithvaryingdegreesofdetail.Themathematicalmodelsof
the hydraulic turbine developed for the stability analysis of a Hydraulic Power Generating
Systemare
SimplifiedNonlinearTurbineModel
NonlinearTurbineModelwithoutSurgeTankassumingInelasticWaterColumn
NonlinearTurbineModelwithoutSurgeTankincludingElasticWaterColumnEffect
Nonlinear Turbine Model with Surge Tank assuming Inelastic Water Columns in
PenstockandTunnel
Nonlinear Turbine Model with Surge Tank assuming Elastic Water Column in Penstock
andInelasticWaterColumninTunnel
LinearTurbineModelwithSurgeTankconsideringInelasticWaterColumnsinPenstock
andTunneland,turbinecharacteristics
LinearTurbineModelwithSurgeTankincludingElasticWaterColumninPenstockand
turbinecharacteristics
37
6.2 NonlinearTurbineModels
Nonlinear turbine models are required when speed and power changes are large during an
islanding, load rejection and system restoration conditions. Hydrodynamics and mechanic
electric dynamics are involved in nonlinear systems. It is important in a system with long
penstock.
Thenonlinearmodelsofthehydraulicturbinedevelopedandstudiedare
NonlinearTurbineModelwithoutSurgeTankassumingInelasticWaterColumn
NonlinearTurbineModelwithoutSurgeTankincludingElasticWaterColumnEffect
Nonlinear Turbine Model with Surge Tank assuming Inelastic Water Columns in
PenstockandTunnel
Nonlinear Turbine Model with Surge Tank assuming Elastic Water Column in Penstock
andInelasticWaterColumninTunnel
6.2.1 SimplifiedNonlinearTurbineModel
TheSimplifiedNonlinearTurbineModelisbaseduponasimplehydraulicturbinewithashort
length penstock, unrestricted head and tailrace, and without surge tank. The penstock is
modelled assuming an inelastic conduit and incompressible fluid. Neglecting friction losses in
thepenstock,thetransferfunctionofflowrateandwaterpressureofasimplePenstocksystem
isdescribedbyEquation(2.61).
Inthemodellingoftheturbineitself,bothitshydrauliccharacteristicsandmechanicaloutput
power must be modelled. The nonlinear characteristics of hydraulic turbine are neglected in
thismodel.ThehydraulicturbinemodelisdescribedbyEquations(3.12)and(3.14).
The block diagram representing the dynamic characteristics of a simplified nonlinear turbine
modelisshowninFigure62.Equating(2.61),(3.12)and(3.14)givesthetransferfunctionfora
simplifiednonlinearturbinemodel.Theinputandoutputsignalofthetransferfunctionarethe
perunitdeviationsingateopeningpositionandturbineoutputtorque,respectively.
Gate
x2
Head
-
1
Tws
+
1
+
flow
At
-
Torque
+
-
qnl
KD
Figure62:SimplifiedNonlinearTurbineModel
38
6.2.2 NonlinearModelwithoutSurgeTankassumingInelasticWaterColumn
TheNonlinearModelwithoutSurgeTankassumingInelasticWaterColumnisbaseduponthe
Simplified Nonlinear Turbine Model, described in Chapter 6.2.1. The penstock is modelled
assuming an inelastic conduit and incompressible fluid where the travelling pressure wave
effectsarerelativelyinsignificant.
The head losses hf are proportional to flow squared and depends on the conduit dimensions
andfrictioncoefficientfp.Itsufficesinthismodelassumethattheheadlossesareproportional
toflowsquaredandtheheadlosscoefficient.Thetransferfunctionrelatingtheflowrateand
waterpressureofasimplePenstocksystemisdescribedbyEquation(2.60).
Inthemodellingoftheturbineitself,bothitshydrauliccharacteristicsandmechanicaloutput
power must be modelled. The hydraulic turbine model is described by Equations (3.12) and
(3.14).Figure63showstheblockdiagramofaNonlinearModelwithoutSurgeTankassuming
InelasticWaterColumn.
Equating (2.60), (3.12) and (3.14) gives the transfer function for a nonlinear model without
surge tank assuming inelastic water column. The input and output signal of the transfer
function are the per unit deviations in gate opening position and turbine output torque,
respectively. This model is depicted in Figure 2 in the report Hydraulic Turbine and Turbine
ControlModelsforSystemDynamicStudies.[1]
x2
Gate
x2
fp
Head
-
1
Tws
+
flow
At
-
Torque
+
-
qnl
KD
Figure63:NonlinearModelwithoutSurgeTankassumingInelasticWaterColumn
6.2.3 NonlinearModelwithoutsurgetankincludingElasticWaterColumnEffect
The modelling of inelastic water column effect is adequate only in short or medium length
penstocks. The travelling wave effects and the travel time of the pressure are considered in
hydropower plant models with long penstocks. Penstock is modelled taking into account the
elasticwaterhammertheoryandneglectingthehydraulicfrictionlosses.
Theclassicalwavesolutiontakingintoaccounttheelasticwaterhammertheoryisdescribedin
Equation (2.59). The water hammer occurrence in penstock in mathematical terminology is
s 2Te
representedbyadelay e
inthehydraulicstructure.
39
The mechanical power output and hydraulic characteristics of a turbine are described by
Equations (3.12) and (3.14), respectively. The block diagram of the Nonlinear Model without
SurgeTankincludingElasticWaterColumnEffectisshowninFigure64.
Equating (2.59), (3.12) and (3.14), the transfer function for a Nonlinear Model without Surge
TankincludingElasticWaterColumnEffectrelatingtheturbineoutputtorquetothedeviations
ingatepositionisgotten.ThismodelisdepictedinFigure4inthereportHydraulicTurbineand
TurbineControlModelsforSystemDynamicStudies.[1]
Figure64:NonlinearModelwithoutsurgetankincludingElasticWaterColumnEffect
6.2.4 NonlinearModelwithSurgeTankassumingInelasticWaterColumns
Hydro power plants with long conduits use surge tank to provide some hydraulic isolation of
the turbine from the head deviations generated by transients in the conduit. The Nonlinear
ModelwithSurgeTankassumingInelasticWaterColumnsisbaseduponasimpleturbinewith
anupstreamwatertunnel,asurgetankandalonglengthpenstock.
TheblockdiagramoftheNonlinearModelwithSurgeTankassumingInelasticWaterColumnsis
shown in Figure 65. The upstream water tunnel and penstock are modelled assuming an
incompressible fluid and a rigid conduit where the travelling pressure wave effects are
relatively insignificantly. It suffices in this model assume that the upstream water tunnel and
penstock head losses are proportional to flow squared through loss coefficients fp1 and fp2,
respectively. The transfer function relating the flow rate and water pressure of a simple
PenstocksystemisdescribedbyEquation(2.60).
The mechanical power output and hydraulic characteristics of a turbine are described by
Equations(3.12)and(3.14),respectively.Thenonlinearcharacteristicsofhydraulicturbineare
consideredinthismodel.
40
Thehydraulicmodelofthesurgetankincludesanorificethatdissipatestheenergyofhydraulic
oscillationsandproducesdamping.Headlossesintheorificeofthesurgetankareproportional
tolosscoefficientsf0timesflowratetimesabsolutevalueofflowratetomaintainthedirection
of head loss. The transfer function of flow rate and water pressure of the surge tank is
describedbyEquation(2.69).
Equating(2.60),(2.69),(3.12)and(3.14)givesthetransferfunctionforaNonlinearModelwith
Surge Tank assuming Inelastic Water Columns. The input and output signal of the transfer
functionaretheperunitdeviationsingateopeningpositionandturbinetorque,respectively.
ThismodelisdepictedinFigure5inthereportHydraulicTurbineandTurbineControlModels
forSystemDynamicStudies.[1]
Figure65:NonlinearModelwithSurgeTankassumingInelasticWaterColumns
6.2.5 NonlinearModelwithSurgeTankassumingElasticWaterColumninPenstock
andInelasticWaterColumninUpstreamTunnel
TheNonlinearModelwithSurgeTankassumingElasticWaterColumninPenstockandInelastic
WaterColumninUpstreamTunnelisbaseduponaNonlinearModelwithSurgeTankassuming
Inelastic Water Columns in cases where the travelling wave effects in the penstock are
essential. The block diagram of Figure 65 is modified to a block diagram that represents the
dynamiccharacteristicofasimpleturbinewithanupstreaminelastictunnel,asurgetankanda
long length elastic penstock. Figure 66 shows the block diagram of a nonlinear model with
surgetankincludingelasticwatercolumninpenstockandinelasticwatercolumninupstream
tunnel.
Thedynamiceffectsoftheuppertunnelcontributewithlowfrequencyoscillations,intheorder
of0.10Hz.Thedynamicbehaviourofthepenstockcontributeswithhighfrequencyoscillation
duetothepenstockissubjecttoabruptgateorflowchanges.
41
The hydraulic model of the surge tank includes an orifice, which dissipates the energy of
hydraulic oscillations and produces damping. Head losses in the orifice of the surge tank are
proportionaltolosscoefficientsf0timesflowratetimesabsolutevalueofflowratetomaintain
thedirectionofheadloss.
The transfer function of flow deviation and water pressure deviation of the upstream water
tunnelisdescribedbyEquation(2.60).Thetransferfunctionofthepenstocktakingintoaccount
theelasticwaterhammertheoryisdescribedinEquation(2.59).Thetransferfunctionofflow
rate and water pressure of the surge tank is described by Equation (2.69). The mechanical
poweroutputandhydrauliccharacteristicsofaturbinearedescribedbyEquations(3.12)and
(3.14),respectively.Thenonlinearcharacteristicsofhydraulicturbinearenotconsideredinthis
model.
Equating (2.59), (2.60), (2.69), (3.12) and (3.14) gives the transfer function for a Nonlinear
ModelwithSurgeTankassumingElasticWaterColumninPenstockandInelasticWaterColumn
in Upstream Tunnel. The input and output signal of the transfer function are the per unit
deviationsingateopeningpositionandturbinetorque,respectively.Thismodelisdepictedin
Figure 6 in the report Hydraulic Turbine and Turbine Control Models for System Dynamic
Studies.[1]
qnl
-
x2
fp
-
Gate
flow
Zc
Tanh (Tes)
Head
+
+
1
CSs
Torque
At
+
-
KD
x
fp1
+ 1
TW2s
-
x
fp2
Figure66:NonlinearModelwithsurgetankassumingelasticwatercolumninpenstockandinelastic
watercolumninupstreamtunnel
42
6.3 HydroTurbineLinearModels
Thenonlinearcharacteristicsofthehydropowerplantmodelscanbeapproximatedbylinear
models.Linearequationsorlowersteptransferfunctionareusedforstudiesofcontrolsystem
stability limited to small perturbations around a stable operating point. The dynamics of the
analyzed system with respect to the operating point can be observed by changing operating
point. The nonlinear turbine characteristics and the travelling wave effects are considered in
thesemodels.[15,18,33]
The linear models of the hydraulic turbine, including the nonlinear turbine characteristics
extractedfromtheHillCharts,developedandstudiedare
LinearTurbineModelwithSurgeTankconsideringInelasticWaterColumnsinPenstock
andTunneland,turbinecharacteristics
LinearTurbineModelwithSurgeTankincludingElasticWaterColumninPenstockand
turbinecharacteristics
Theselinearmodelsofthehydraulicturbineconsideritscharacteristiccurvesandcoefficients
thatmustbeextractedandcalculatedfromtheHillChartsofthetestmodel.
6.3.1 LinearTurbineModelwithSurgeTankassumingInelasticWaterColumns
The Linear Model with Surge Tank assuming Inelastic Water Columns is based upon a simple
turbine with an upstream tunnel, a surge tank and a penstock. The upstream tunnel and
penstock are modelled assuming an inelastic conduit and incompressible fluid where the
travelling pressure wave effects are relatively insignificant. The nonlinear characteristics of
hydraulic turbine and the inelastic water hammer effect are considered in this model. This
model is based on the paper Basic Modelling and Simulation Tool for Analysis of Hydraulic
TransientsinHydroelectricPowerPlants.[18]
The transfer function of flow deviation and water pressure deviation of the outlet of the
upstream water tunnel is described by Equation (2.60). The transfer function relating
incrementalwaterpressureandflowofpenstockcanbewrittenas
(6.1)
TW 2 s H f 2 q2 h2 h1
wherethesuffix1isrelatedtotheupstreamtunnelandthesuffix2correspondstoPenstock.
ThetransferfunctionofflowrateandwaterpressureofthesurgetankisdescribedbyEquation
(2.69)wherethehydrauliclossesintheorificeareneglected.
ThemathematicalmodeloftheFrancisturbineisrepresentedbyEquations(3.15)and(3.16)for
smallvariationsinthevicinityofanoperatingpoint.Thesixpartialderivatives,whichrepresent
the nonlinear characteristic of the turbine, are taken from model tests. Prototype turbine
characteristics are plotted in the Hill Charts. The turbine efficiency for any operating point
givenbyrunnerspeed,pressureheadandgatepositioncanbeextractedfromtheHillCharts.
43
m
y
q
h
1
TS s
1
TW 2 s H f 2
q
y
TW 1s H f 1
q
n
m
h
m
n
Figure67:LinearModelwithSurgeTankassumingInelasticWaterColumns
Equating(2.60),(2.69),(3.15),(3.16)and(6.1)givesthetransferfunctionrelatingtheperunit
deviations turbine torque to gate position for a Linear Model with Surge Tank assuming
InelasticWaterColumns.Figure67showstheblockdiagramofaLinearModelwithSurgeTank
assumingInelasticWaterColumns.
6.3.2 Linear Turbine Model with Surge Tank assuming Elastic Water Column in
Penstock
TheLinearModelwithSurgeTankassumingElasticWaterColumninPenstockisbasedupona
simpleturbinewithanupstreamtunnel,asurgetankandapenstock.Theupstreamtunnelis
modelledassuminganinelasticconduitandincompressiblefluidwherethetravellingpressure
wave effects are relatively insignificant. Penstock is modelled taking into account the elastic
waterhammertheoryandneglectingthehydraulicfrictionlosses.Thenonlinearcharacteristics
of hydraulic turbine are considered in this model. This model is based on the paper Basic
Modelling and Simulation Tool for Analysis of Hydraulic Transients in Hydroelectric Power
Plants.[18]
The block diagram of Figure 67 is modified to a block diagram that represents the dynamic
characteristicofasimpleturbinewithaninelasticupstreamtunnel,asurgetankandanelastic
penstock. Figure 68 shows the block diagram of a Linear Model with Surge Tank assuming
ElasticWaterColumninPenstock.
44
m
y
q
h
1
TS s
1
TW 2 s H f 2
m
h
q
y
q
n
m
n
Figure68:LinearModelwithSurgeTankassumingElasticWaterColumninPenstock
The transfer function of flow deviation and water pressure deviation of the outlet of the
upstream water tunnel is described by Equation (2.60). The transfer function relating
incrementalwaterpressureandflowofpenstockcanbewrittenas
TW 2
tanh sTe q2 h2 h1
(6.2)
Te
Wherethesuffix1isrelatedtotheupstreamtunnelandthesuffix2correspondstoPenstock.
ThetransferfunctionofflowrateandwaterpressureofthesurgetankisdescribedbyEquation
(2.69)wherethehydrauliclossesintheorificeareneglected.
ThemathematicalmodeloftheFrancisturbineisrepresentedbyEquations(3.15)and(3.16)for
smallvariationsinthevicinityofanoperatingpoint.Thesixpartialderivativeswhichrepresent
the nonlinear characteristic of the turbine are taken from model tests. Prototype turbine
characteristics are plotted in the Hill Charts. The turbine efficiency for any operating point
givenbyrunnerspeed,pressureheadandgatepositioncanbeextractedfromtheHillCharts.
Equating(2.60),(2.69),(3.15),(3.16)and(6.2)givesthetransferfunctionrelatingtheperunit
deviationsturbinetorquetogatepositionforaLinearModelwithSurgeTankassumingElastic
WaterColumninPenstock.
45
Hydroelectric Power Plant Modelling by

StructureMatrixMethod
ThischaptershowsanalternativemethodofHydroelectricPowerPlantmodellingforstability
studies. The Hydroelectric power plant is modelled by the Structure Matrix Method. Hermod
Brekke introduced this mathematical method. This improved model covers a wide range of
parameters such as influence of the turbine characteristics, frictional damping of oscillatory
flow in elastic conduits, influence of the generator load and the analysis of the turbine
governingsystem.
The Hydroelectric Power Plant mathematical model may consist of many different elements.
TherepresentationofeachcomponentoftheHydroPowerSystemispresentedbymeansof
appropriate block diagrams. The Structure Matrix Method uses the Laplace transformed
equations for the system stability study. The turbine characteristics are linearized at an
operating point. The stability study of a dynamic system is based in the frequency response
analysisandfreevibrationanalysis.Thismethodisonlypresentedforillustrationpurposes.
7.1 DefinitionoftheMethod
The Structure Matrix Method was introduced by Hermod Brekke for the governing stability
analysis and the study of frictional damping of oscillatory flow in complex conduits. The
combination of isolated operation, long transmission lines, and long and complex tunnel
systemsrequireathoroughstabilityanalysisofthespeedcontroloftheunitsinthesystems.
TheoriginoftheStructureMatrixMethodmaybetracedtotheStructureAnalysistheoryand
terminologies such as Element, "Node" as well as "Local" and "Global" matrices have been
used.ElementMatrixrepresentationsofeachoftheelementsforthemodellingoftheoverall
systemareusedasbrickstoconstructtheGlobalStructureMatrix.[6,16,30]
AnElementMatrixrepresentationhasthefollowinggeneralform
A s b = c
(7.1)
ThedimensionoftheElementMatrixA(s)ismxn.Thevectorb,dimensionnx1,isoftenreferred
toasthepressurevector.Thevectorcofdimensionmx1isreferredtoasflowvector.The
ElementMatrixisthelowestleveloftheStructureMatrix.
A matrix representation of the group of elements interconnected may be established by

incorporatingalltheElementMatricestogether.TheGlobalStructureMatrix,whichisnormally
regular, can be solved if the number of known components in the vector b is equal to the
numberofunknowncomponentsinthevectorc.
46
7.2 Matrix representations of the Basic Elements in Hydro Power

Systems
TheStructureMatrixmethodisbasedonthesamefunctionalequationsthatTransferMethod
andImpedanceMethod.TheessentialpartoftheStructureMatrixMethodistoestablishthe
Element Structure Matrices of each component. The Hydro Power Plant mathematical model
may consist of many different elements such as tunnels, surge tanks, throttles, junctions,
pumps, turbine sets and so on. The representation of each component of the Hydro Power
Systemispresentedbymeansofappropriateblockdiagrams.
7.2.1 PipesandTunnels
Thepipeortunnelisoneofthemostimportantelementsofthehydropowerplantmodelling.
ThedynamicbehaviouroftheelasticfluidinaclosedconduitisdescribedinChapter2.2.The
Element Matrix representation for a pipe based on the transfer function detailedin Equation
(2.56)is
1
1
Z tanh zT Z sinh zT h s q s
e
c
e
c
i i
(7.2)
hi 1 s qi 1 s
1
1
sinh
Z
zT
Zc tanh zTe
e
c
Usingthefollowingexpressions:
1
Zc tanh zTe
(7.3)
1
Zc sinh zTe
(7.4)
Thegeneralpipematrix,Equation(7.2)yields
T
S
S hi s qi s
T hi 1 s qi 1 s
(7.5)
7.2.2 Surgetanksorairaccumulators
The Element Matrix expression for surge tank or air accumulator may be written as either a
oneterminalelementoravirtualtwoterminalelement.TheElementmatrixforasurgetank
writtenasoneterminalelementis
H 0 As
s hi s qi s
(7.6)
Q0
47
The Element matrix expression for a surge tank written as a virtual twoterminal element,
consideringhi+1=qi+1=0,is
0
1
hi 1 s qi 1 s
H
A
(7.7)
0
s
0
s hi s qi s
Q0
Forafreewatersurfaceofanopensurgetank,themostconvenientrepresentationforasurge
tankisatwoterminalElementbyincorporatingEquations(7.6)and(7.7)together.
T TF
S
S hi s 0
T hi 1 s qi 1 s
(7.8)
where
TF
H 0 As
s
Q0
(7.9)
7.2.3 LocalLosses
TheElementMatrixdescribingthedampingcausedbylocalvariationsincrosssectionsofthe
conduitorforfixedpartialopenvalvesmaybeestablishedasfollows
1
1
K
K re hi s qi s
re
(7.10)
1 hi 1 s qi 1 s
1
K
Kre
re
Forthrottles,Kreisgivenby
Kre
Qt Q0 1
1
2 2
gH 0 As Ab
(7.11)
In Equation (7.11), As is the smallest cross section and Ab is the largest cross section. A loss
constantisincluded.
ForTjoints,Kreisgivenby
kQ02
QQ
K re
q t2 0
(7.12)
2
gAb H 0
3gAs H 0
InEquation(7.12),kisthegeometryconstant.
48
7.2.4 HydroTurbine
Theanalysisofthestabilityisvalidforfrequenciesdomainsnormallyuptoabout6rad/sec,and
resonancefrequenciesintheturbinerunnerandguidevanesystemcanbeneglected.Inthis
frequency range, the influence from the turbine may be evaluated from the turbine
characteristic diagram, which is assumed valid for small oscillations with low frequencies
around a steady state value. A linearization around the steady state may also be carried out
withinacceptableaccuracy.
7.2.4.1 HydroTurbineCharacteristics
ThemechanicalpowerdevelopedbytheturbineisgivenbydifferentiatingEquation(3.1)
P
P
P
dP
dQ
dH d
(7.13)
Q
H

Dividingthroughby P0 gH00Q0 ,Equation(7.13)yields
p q h
(7.14)
The values of q, h and may be found by studying the characteristic diagram of the
turbine.
TheflowthroughaFrancisturbinedependsuponthenethead,rotationalspeedoftheturbine
andwicketgateopeningposition.Theequationoftheflowyields
q Qy y Qn n
1
1 Qn h Q
2
(7.15)
ThevaluesQn,QyandQmaybedeterminedbymeansoftheturbinecharacteristicdiagram.
TheexpressionsforQn,QyandQare
Qn
Q11 n11 0
n11 0 Q11 0
Qy
Y0
Q11
Y 0 Q11 0
(7.16)
(7.17)
0
Q11
0 Q11 0
(7.18)
Theefficiencyisdefinedastheratioofpowersuppliedbytheturbinetotheabsorbedpower.
Therelativeefficiencycanbeexpressedasfollows
Eq q
1
Eq En h Enn E
2
(7.19)
49
ThevaluesEn,EyandEmaybedeterminedbymeansoftheturbinecharacteristicdiagram.The
expressionsforEn,EyandEare
Eq
En
Q0
Q11 0 0
(7.20)
n11 0
n11 0 0
(7.21)
0
0 0
(7.22)
The runner blade servomotor has a time constant TL that describes the runner blades
movement as a function of the guide vane movement. The transfer function describing the
runnerbladesmovementis
KL
1 TL s
(7.23)
Therunnerbladesnormallyarestoppedmechanicallybeforetheguidevanesarefullyclosed.
The cam transmission constant KL must take into account only about 75% of the guide vane
strokeregardedversustherunnerbladestroke.
Forturbineswithfixedrunnerblades,suchasFrancisturbines,therunnerblademovementis
zero,=0.
7.2.4.2 HydroTurbineMatrixRepresentation
Considertheflowpositiveoutoftheturbineonbothsidesandrisingpressuretobepositivein
thesamewayasforpipeelements.Thepressure hinEquations(7.15)and(7.19)isregarded
to pressure difference between the right hand side and the left hand side. The equations of
continuityforflowthroughtheturbine,headupstreamanddownstreamoftheturbineisgiven
by
q qi
(7.24)
qi1 qi 0
(7.25)
h hi1 hi
(7.26)
SubstitutingandhfromEquations(7.19)and(7.26)intoEquation(7.14),yields
1 E q 1 12 E
q
En hi 1 hi En n E p 0
(7.27)
50
Furtherbysubstituting,qandhbyEquations(7.23)and(7.26)intoEquation(7.15),yields
1 Eq En hi 1 hi En n E F y p 1 Eq qi 1 0
2
(7.28)
SubstitutingandqhbyEquations(7.23),(7.24),(7.25)and(7.26)intoEquation(7.27),yields
1
1
(7.29)
1 Qn hi1 Qn n Qy Q F y 1 Qn hi qi1
2
2
1
1
(7.30)
1 Qn hi1 Qn n Qy Q F y 1 Qn hi qi
2
2
ThecompleteturbinematrixmaybeestablishedbymeansofEquations(7.28),(7.29)and(7.30)
hi 1

B C B Q 0 y qi 1
J
L J K M hi 0
(7.31)

B C B Q 0 n qi
p
The turbine matrix will be established with vectors describing all inputs and variables which
havetobesolved;where
1
B 1 Qn
(7.32)
C Qy Q F
(7.33)
Q Qn
(7.34)
3 En 1 Eq Qn
2 1 Eq
K Qn
En
1 Eq
(7.36)
L Qy Q F
(7.35)
E F
(7.37)
1 Eq
1
1 Eq
(7.38)
51
7.2.5 Thesynchronousgeneratorandtheelectricgrid
Asafirstapproach,thesynchronousgeneratormodelforageneratorinfinitebussystemmay
besimplified.Thesimplifiedgeneratormodelonlyincludestheanalysisofthehydraulicsystem
with the influence of the frictional damping and the turbine characteristics. The synchronous
generatormodelforageneratorinfinitebussystemmaybesimplifiedbyassuming:
The electric load is pure resistive and there is no rotating mass (except the generator
itself)inthesystem,
The transients of the electric system are much faster than those of hydraulicmechanic
system,andthereforemaybeignored.
Thesimplifiedequationofthegeneratortakingaccountofloadcharacteristicmaybesimplified
bythefollowingsimplefunction
p pg Ta s Fg n
(7.39)
The transfer function Fg is the adjusting load coefficient which describes an isolated load
conditionoragridsystem.Taisinertiamasstimeconstantofthegenerator(mechanicalinertia
constant).
Forafrequencyresponseanalysisfor<2rad/sec,thetransferfunction,describedinEquation
(7.39)maybesimplifiedtobe Fg eb ,whereebisthepermanentloaddroop(voltagedroop)
or selfregulating in the load as a function of the speed droop of the generator. The worst
scenarioisaresistiveload Fg 0 .
ThesimplifiedElementMatrixyields:
0 0 n 0
N 1 p s p s
where
N Ta s Fg
(7.40)
(7.41)
Forstudiesoftheinteractionofhydraulicsystemandpowersystems,itispossibletoestablish
a complete sub matrix for the electric grid including the torqueangle oscillations of the
synchronousmachine.Thismatrixmaybeincludedintheglobalmatrixofthesystem.
52
7.2.6 TurbineSpeedGovernor
Despite of the large variety of turbine governors, there is no great difference on the general
mathematicalrepresentationforthem.Themostcommonusedgovernorsmaybedividedinto
twotypes:
Mechanicaldashpotgovernor(PItype)
ElectricHydraulicPIDgovernor
7.2.6.1 TraditionalGovernor
Thetraditionalgovernorisdescribedwiththefollowingequations
GRP 1
Fn n RP pr
y nref
GC
pref p pr
(7.42)
(7.43)
where
1
1 Tp s
(7.44)
1 TD s
RT TD s
(7.45)
Assumingp=0inEquation(7.43)ifthereisnopoweradjustmentsignalfromthewaterlevel
controller.Equations(7.42)and(7.43)inmatrixformyield
GRP 1 n
F
R
nref
n
P
CG pr
(7.46)
pref
0
1
0
Equation (7.46) may be simplified assuming Pref as a possible port of excitation of little
importanceinstabilityorvibrationanalysis.Hence,Equation(7.46)yields
GRP 1 n
F
nref
(7.47)
n
CG y
53
7.2.6.2 PIDGovernor
The following equations describe the turbine governing system when the turbine governor is
assumedaPIDtype.
n
K n E F nref
(7.48)
0 1 0 pr p
ref
y
Where
E Rp
G
GR p 1
F
GC
1
C
1 Ty s 1 Tx s
(7.49)
(7.50)
(7.51)
G Kp
Ki
Kd s
s
(7.52)
Assumingpr=pref,Equation(7.48)yields
n
K n F nref
(7.53)
It should be noted that the constant Kn is used for choosing between open and closed loop
analysis.ItisassumedKn=0fortheopenloopanalysis,andKn=1forcloseloopanalysis.
54
7.3 CompositionoftheGlobalStructureMatrix
The Structure Matrix of a complex system could be built from Element and Local Matrixes.
Everyelementinthesystemisdesignedinorderwithanelementnumber.Thepointswhere
the elements terminals join are referred as nodes. The Element Matrixes are connected
together in series or in parallel according to some general rules. The Element Matrixes are
incorporatedintotheGlobalSystemStructureMatrixaccordingtothesystemindexing.
ThegeneralrulesforthecompositionoftheSystemStructureMatrixaredetailedbelow.
1. The dimension of a Global Structure Matrix A is equal to the number of nodes of the
completesystem.
2. Anmdimensionelementhasconnectiontomnodes.
3. Twoelementsmaysharemcommonnodesiftheysharemcommonvariables.
4. TheGlobalStructureMatrixAhasadimensionofnxn.EachofitsentriesAijisequalto
thesumofElementStructureMatrixes.
5. A group of elements, incorporated together, may be treated as an element, and the
LocalStructureMatrixcanbeusedastheElementMatrix.
In stability analysis for hydro power plants, the following boundary conditions have been
establishedinordertobuilttheCompleteStructureMatrix
Foropenwatersurfaces,thevectorhisequaltozero(h=0),
Forclosedends,thevectorqisequalto0(q=0)
7.4 StructureMatrixoftheHydroTurbineUnit
The Group Structure Matrix for a hydro turbine set includes the hydraulic turbine, generator
andturbinespeedgovernor.TheElementMatrixesoftheturbine,generatorandPIDgovernor
have been described in Equations (7.31), (7.40) and (7.48), respectively. The Structure Matrix
forthehydroturbinesetisbuiltbyincorporatingEquations(7.31),(7.40)and(7.48)together.
Therearesixnodesconnectedtosixindependentvariablesforthecompleteturbineset.
Numbering the variables as hR=1, n=2, y=3, pr=4, p=5 and hL=6. The Structure Matrix may be
determinedas
B Q C 0 0 B hR s qR s
0 K
F E 0
0 n nref
n
J
K
L 0 M J y 0
(7.54)
p
p
0
0
0
1
0
0
r
ref
0
N
0 0 1
0 p s pg s

h
s
q
s
B
Q
C
B
0
0
L L
55
7.5 HydroPowerPlantStructureMatrix
The hydro power plant chosen for this study is shown in Figure 71. This hydro power plant
consistsof5elements.Theelementscanbewritteninmatrixesaccordingtotheindexing.
Figure71:LayoutofHydroPowerPlantforStructureMatrixMethodModelling
TheElement1isthetunnelconnectingtothereservoirandthesurgetank(node1and3)
T1 S1 h1 s q1 s
S T h s q s
1
1 3
3
(7.55)
TheElement2isthesurgetank,nodenumber2,connectedtonode3
TF 2 T2 S2 h2 s 0
S
T2 h3 s q3 s
(7.56)
TheElement3isthetunnelconnectingtothesurgetankandtheturbine,(node3and4)
T3 S3 h3 s q3 s
S T h s q s
3 4
3
4
(7.57)
TheElement4istheStructureMatrixfortheHydroTurbineUnit.TheGroupStructureMatrix
forahydroturbinesetisdescribedinChapter7.4.NumberingthenodesasfollowshR=4,n=5,
y=6,pr=7,p=8andhL=9.
B Q C 0 0 B h4 s q4 s
0 K
0 n nref
F E 0
n
J
K
L 0 M J y 0
(7.58)
p
0
0
0
1
0
0
r pref
0
0 0 1
0 p s pg s
N

B Q C 0 0 B h9 s q9 s
56
TheElement5isthetailrace,connectingtonode9to10
T5 S5 h9 s q9 s
5 T5 h10 s q10 s
(7.59)
The Global Structure Matrix expression for the Hydro Power Plant, shown in Figure 71,
accordingtotheindexingis
0
S1
0
0
0
0 0
0
0 0 q1
T1
0 T T
S2
0
0
0
0 0
0
0 h2 0
2
F2

S2
S3
0
0
0 0
0
0 h3 0
T1 T2 T3
S1

0
0
S3
B
T3 B Q C 0 0
0
h4 0
0
0
0
0
0
0 n nref
Kn
F E 0

(7.60)
0
0
0 M
0 y 0
J
K
L
J
0
0
0
0
0
0
0
1 0
0
0 pr pref

0
0
0
0
0 1
0
0 p pg
N
0
0
0
0
0 0 B T5 S5 h9 0
B
Q
C

0
0
0
0
0
0 0
T5 0 q10
S5
0
Instabilitystudiesforhydropowerplants,thesystemresponsestopowerreferencesettingare
oftenignored.Hence,Equation(7.60)canbereducedas
0
S1
0
0
0
0
0
0 0 q1
T1
0 T T
0
0
0
0
0
0 h2 0
S2
F2
2

T1 T2 T3
S2
S3
0
0
0
0
0 h3 0
S1

T3 B Q C 0
0
S3
B
0
0
h4 0
0
Kn
F
0
0
0
0
0
0 n nref
(7.61)

J
J
K
L M
0
0
0 y 0
0
0
N
0
0
0
0
1
0
0 p pg

B
Q
C
0
0
0 B T5 S5 h9 0
0
0
S5
T5 0 q10
0
0
0
0
0
0
57
Foranideallossless,Qy=1,Qn=0and=constant.Equation(7.61)issimplifiedto
0
S1
0
0
0
0
0
0 0 q1
T1
0 T T
S2
0
0
0
0
0
0 h2 0
F2
2

T1 T2 T3
S2
S3
0
0
0
0
0 h3 0
S1

T3 0.5 0 C 0
S3
0
0.5
0
0
h4 0
0
Kn
F
0
0
0
0
0
0 n nref
0
0
0
1.5
0
1
1
1.5
0
y 0
0
N
0
0
0
0
1
0
0 p pg

0
0
0.5
0
1
0 0.5 T5 S5 h9 0
0
0
S5
T5 0 q10
0
0
0
0
0
0
(7.62)
7.6 DynamicAnalysis
Thestabilityanalysisofadynamicsystemisbasedinthefrequencyresponseanalysisandfree
vibrationanalysis.
7.6.1 FrequencyResponseAnalysis
Thefrequencyresponseanalysisisbasedonthestudyofthesteadyoscillatoryflowresponses
to a large harmonic excitation over a range of frequencies. The frequency response of the
systemisdetermineddirectlywithsreplacedbyj.Thefrequencyresponseofthesystemto
theexcitationcanbeobtainedbysolvingEquation(7.1)repeatedlywithaselectedwithinthe
frequencyrange1<>2.
Theinputsignalmaybeappliedtoanydeviceandtheoutputsarestudied.Aconstant,usually
1,mustbeassignedtooneofthecomponentsinvectorqinEquation(7.1)accordingtowhere
theexcitationisimposed.
Thefrequencyresponsestoloaddisturbancescanbecalculatedassumingpg=1andnref=0.
Theresponsen/pgisthemostimportantmeasuretostudythestabilityofthesystem.
The openloop speed response to speed setting can be evaluated by the criteria of Nyquist
Bodeonopenloopnnrefresponses.Thennrefresponsescanbemeasuredsettingnref=1
andpg=01.
7.6.2 FreeVibrationAnalysis
Thefreevibrationanalysisisastudyofthedynamiccharacteristicsofthesystemassumingno
external excitation imposed on the system. The dynamic characteristic of a linear system
depend on its eigenvalues and the mode shapes. The eigenvalues determine the natural
frequenciesanddampingofthesystem,andthemodesdeterminetheshapeofthevibrations.
58
PowerSystemStabilityAnalysis
Stability problems in a power system may be manifested in different ways depending on the
systemconfigurationandoperatingmode.Analysisofelementarypowersystemconfigurations
bymeansofidealizedmodelsillustratessomeofthefundamentalstabilitypropertiesofpower
systems.Smallsignalstabilityistheabilityofapowersystemtoreachasteadystateoperating
point, which is identical, or close to, the initial condition after being subjected to small
disturbances.Smallsignalanalysisuseslineartechniquestoprovidevaluableinformationabout
thedynamiccharacteristicsofthepowersystem.
Thischapterdescribesfundamentalaspectsandanalyticaltechniquesinthestudyofstabilityof
dynamicsystems,andidentifiesfactorsinfluencingthem.Thestabilityofthegeneratorinfinite
busbarsystemfollowingasmalldisturbanceisdiscussed.
8.1 PowerSystemStability
Power System Stability is defined as the ability of a power system to remain in a state of
operatingequilibriumundernormalconditionsandtoregainanacceptablestateofequilibrium
after being subjected to a physical disturbance. An electrical power system is composed of
many individual elements connected together to form a large, complex and dynamic system
thatisabletogenerate,transmitsanddistributeenergyoverageographicalarea.Thetwomain
causes of power system dynamic are the behaviour of the system to both a changing power
demandandtovarioustypesofdisturbances.[2,22,45,6566]
The power system stability can be divided into rotor angle stability, frequency stability and
voltagestability.Rotorangleandvoltagestabilitycanbestudyintosmallandlargedisturbance
stability.
8.1.1 SmallSignalStabilityAnalysis
A system is steady state stable for a particular operating point if, following any small
disturbance, it reaches a steadystate operating point which is identical, or close to, the
predisturbance condition. This is known as smallsignal stability. Smallsignal analysis uses
linear techniques to provide valuable information about the dynamic characteristics of the
power system. Instability may be a result of steady increase in rotor angle due to lack of
sufficient synchronous torque, or rotor oscillations of increasing amplitude due to lack of
sufficientdampingtorque.[22,67]
The dynamics of the generator, and its stability, are affected by automatic control of the
generatorandtheturbine.Thestabilityofalinearsystemisindependentoftheinput,andthe
stateofastablesystemwithzeroinputwillalwaysreturntotheoriginofthestatespace.The
stabilityofanonlinearsystemdependsonthetypeandmagnitudeofinputandtheinitialstate.
59
8.2 FundamentalsofPowerSystemStability
The behaviour of a dynamic power system is described by a set of n first order nonlinear
differentialequations.Thissetofequationscanbewrittenintheform
x f x, u,t
(8.1)
Thevectorxisreferredasthestatevector,anditsentriesxiasstatevariables.Theinputsofthe
systemarerepresentedbythevectoru.Thederivateofastatevariablexwithrespecttotimet
isdenotedby x .Thestateofasystemistheminimumamountofinformationofthesystemat
anyinstantintime.Thesystemisautonomouswhenthederivativesofthestatevariablesare
notexplicitfunctionsoftime.Equation(8.1)issimplifiedto
x f x, u
(8.2)
Toobservetheoutputvariablesofthesystem,thestatevariablesandtheinputvariablescan
beexpressedinthefollowingform
y g x, u
(8.3)
Theoutputsofthesystemarerepresentedbyvectory.Thevectorgisavectorofthenonlinear
functionsrelatingstateandinputvariablestooutputvariables.
Thesingularorequilibriumpointsarethecharacteristicofthedynamicsystembehaviour.The
singularpointmustsatisfythefollowingequation
f x0 0
(8.4)
Alinearsystemhasonlyoneequilibriumstate.Theremaybemorethanoneequilibriumpoint
foranonlinearsystem.
The linearized forms of Equations (8.2) and (8.3), evaluated at the equilibrium point about
whichthesmalldisturbanceisanalyzed,are
(8.5)
x Ax Bu
y Cx Du
(8.6)
The Laplace transforms of x and y are the Laplace transforms of the free and zerostate
components of the state and output vectors. These Laplace transforms have a component
dependentontheinitialconditionsandothercomponentdependentontheinputs.Thepoles
ofxandyaretherootsoftheequation
det sI A 0
(8.7)
Equation (8.7) is referred to as the characteristic equation of matrix A. The values of s that
satisfythecharacteristicequationareknownaseigenvaluesofthematrixA.
60
8.3 EigenvalueAnalysis
8.3.1 EigenvaluesandEigenvectors
AscalarparameteriscalledeigenvalueofamatrixAifthereisanonzerocolumnvector
satisfying
(8.8)
A =
AnontrivialsolutionofEquation(8.8)isgivenby
det A I 0
(8.9)
For every eigenvalues i, the vector i which satisfies Equation (8.8) is called the right
eigenvectorassociatedwithAtheeigenvaluei.Therefore,
Ai = i i
(8.10)
Similarly,thelefteigenvectorassociatedwiththeeigenvalueiisthevector iwhichsatisfies
Equation(8.11)
i A = i i
(8.11)
The left and right eigenvectors corresponding to different eigenvalues are orthogonal. The
eigenvectors can be normalized. The product of the normalized left and right eigenvectors
correspondingtothesameeigenvalueis
i i = 1
(8.12)
8.3.2 EigenvalueAnalysis
Thetimedependentcharacteristicofamodecorrespondingtoaneigenvalueiisgivenbyeit.
TheeigenvaluesofEquation(8.7)leadstothe followingconclusions,whichareimportantfor
theanalysisofthepowersystemdynamics:
A. The original system is asymptotically stable when the eigenvalues have negative real
parts.
B. Theoriginalsystemisunstablewhenatleastoneoftheeigenvalueshasapositivereal
part.
C. Thestabilityoftheoriginalsystemisnotpossibletodeterminatewhentheeigenvalues
haverealpartsequaltozero.
D. A real eigenvalue corresponds to a nonoscillatory mode. A negative real eigenvalue
representsadecayingmode.Thelargeritsmagnitude,thefasterthedecay.Apositive
realeigenvaluerepresentsaperiodicstability.
E. Each conjugate pair of complex eigenvalues introduces to the response oscillatory
modes.Theoscillatorymodeisstableistherealpartoftheeigenvalueisnegative.
61
Acomplexeigenvalueisrepresentedby
j
(8.13)
Therealcomponentoftheeigenvaluesgivesthedampingandtheimaginarycomponentgives
thefrequencyofoscillation.
The damping ratio determines the rate of decay of the amplitude of the oscillation. The
dampingratioisgivenby
2 2
(8.14)
EigenvalueAnalysisisperformedusingspeciallinearpowersystemsanalysisprograms.These
determinatetheeigenvaluesandeigenvectorsassociatedwiththesystemmodel.Localmodes
ofoscillationhavenaturalfrequenciesofabout1to2Hz.
In practice, damping of rotor swings is considered to be satisfactory if the damping ratio

0.05 .
8.3.3 ModalandSensitivityAnalysis
Thelefteigenvectorscarryinformationaboutthecontrollabilityoftheindividualmodalvariable
byindividualstatevariables.Thelefteigenvectorweightsthecontributionoftheactivityofthe
statevariablestomodei.
The right eigenvectors carry information about the observability of the individual modal
variable by individual state variables. The right eigenvector gives the mode shape. The mode
shaperepresentstherelativeactivityofthestatevariableswhenaparticularmodeisexcited.
The mode shape represents an inherent feature of a linear dynamic system and does not
dependofthedisturbanceapplied.
Thesensitivityofaneigenvalueitotheelementsofthestatematrixisequaltotheproductof
thelefteigenvectorelementandtherighteigenvectorelement.Multiplyingtherightandleft
eigenvectorsasfollows
The participation factor matrix combines the right and left eigenvalues as a measure of the
relationshipbetweenthestatevariablesandthemodes.
pki = ki ik
(8.15)
Coefficientspkiarereferredtoastheparticipationfactors.Eachparticipationfactorisaproduct
ofthekthelementoftheithleftandrighteigenvectors.Itquantifiesthesensibilityoftheith
eigenvalue to the kth diagonal element of the state matrix. Participation factor contains
informationabouttheobservabilityandcontrollability.
Participationfactorscanbeusedtodetermineenhancingsystemstability.
62
8.4 SmallSignalStabilityAnalysisofageneratorinfinitebussystem
Assumethatthegeneratorisconnectedtotheinfinitebusbar,withthegeneratorrepresented
by the synchronous machine classical model detailed in Chapter 4.3.1. The initial disturbed
conditions are t 0 0 0 and 0 . The incremental swing equation, Equations
(4.7)and(4.8),writteninthematrixformbecomes
N
t K D K E ' N 1
2H
2H
(8.16)
2 H mm
0
0
t s
Thetransientsynchronizingpowercoefficient KE' bydefinitionis:
P EVS
KE ' e
cos
(8.17)
xd
where xd X d X and X X T X S isthecombinedreactanceofthestepuptransformerand

theequivalentnetwork.
Theeigenvaluesofthisstatematrixcanbedeterminedfromsolving
K
K
D E '
K
K
2
det
2H
2 H D s E ' 0
(8.18)
2H
2H

s
Equation(8.18)mayberewrittenasthestandardsecondorderdifferentialequation
2
2 2nat nat
0
(8.19)
Thus,theundampednaturalfrequency nat ofrotorswingsforsmalloscillationsisgivenby
nat K E '
(8.20)
2H
The damping ratio determines the amount of damping present in the system response
expressing how quickly the amplitude of rotor swings decreases during subsequent periods.
Thedampingratiois
1 KD
1
KD
2 2Hnat 2 K E ' 2Hs
(8.21)
63
Thedampednaturalfrequency d ofrotorswingsis
d nat 1 2
(8.22)
Thecharacteristicequationhastworoots1and2givenby
1,2 nat jnat 1 2
(8.23)
Theresultsoftheroots1,2dependsontheactualvaluesofKE,KDandH.TheinertiaconstantH
isconstantwhilebothKDandKEdependsonthegeneratorloading.
8.5 DynamicAnalysis
Thedynamicbehaviourofahydraulicpowerplantmaybeanalyzedeitherinthetimedomain
byusingMethodoftheCharacteristicsorinthefrequencydomain.Thebestknownmethods
forfrequencydomainanalysisareImpedanceMethodandTransferMatrixMethod.
The partial differential equations describing the unsteady flow are converted into ordinary
differential equations in the Time Domain Analysis. The nonlinear friction losses and the
nonlinearboundaryconditionsmaybeincludedintheMethodoftheCharacteristics.
The equations describing the unsteady flow in the time domain are converted into the
FrequencyTimeDomainbyassumingasinusoidalvariationofthepressureheadandtheflow.
Thenonlinearfrictionlossesandthenonlinearboundaryconditionsarelinearizedforsolution
inthesemethods.Thesystemresponseisdeterminedbysuperpositionofindividualresponses.
As the frequency response for the analysis is small, these methods are suitable for general
studies.
TheanalysisoftransientsintheHydraulicPowerSystemModelissubdividedinwaterhammer
responseandmassoscillationresponse.
The elastic pressure frequency describes the oscillation along the penstock. The
characteristic frequency does not vary with the load and depends only upon the
dimensionsofthepenstock.Thenaturalfrequencyistypicallyintheorderof1.00Hz.
The natural frequency resulting from the pendulum action between the reservoir and
thesurgeshaftcalledmassoscillations.Themassoscillationscantakeseveralminutes
togetthesteadystate.
8.5.1 Naturalfrequencyanalysis
The analysis of transients in hydroelectric systems is subdivided in massoscillation response
andwaterhammerresponse.Ordinarydifferentialequationsdescribethemathematicalmodel
formassoscillation(orrigidwatercolumn)betweenthereservoirand thesurgetank.Partial
differentialequationsshowthemathematicalmodelforwaterhammeralongthepenstockand
drafttube.[1520,56,58]
64
Thefirstnaturalfrequencyofthepenstockcorrespondstothefrequencycorrespondingtothe
fundamental period of the system, described in Equation (2.4). The first natural frequency of
thepenstockcanbewrittenas
a
f0
(8.24)
4L
The period of oscillation of the water surface in a simple surge tank is described in Equation
(2.65).Thenaturalfrequencyofthemassoscillationbetweentheupstreamreservoirandthe
surgetankisgivenby
1
1
f0
2 L AS
(8.25)
g At
Thefrequencyresponseofasystemisthesteadystateresponseofthesystemtoasinusoidal
inputsignal.Thetransferfunctionofasystemcanbydeterminedexperimentallybyfrequency
responsetests.Thedesignofasysteminthefrequencyresponsemethodprovidesthedesigner
with control over the system bandwidth and over the effect of noise and disturbance on the
systemresponse.[43]
Theresponseofalineartimeinvariantsystemtosinusoidalinput r t A sin t isgivenby
c t A G j sin t
(8.26)
WherethetransferfunctionG(j)isobtainedbysubstitutingjforsintheexpressionforG(s).
ThemostcommongraphicalrepresentationofafrequencyresponsefunctionistheBodePlot.
The frequency response characteristics of the operational parameters and related them to
standardparameterscanbeexaminedinSIMPOW.Suchcharacteristicsprovideusefulinsight
intothedynamicbehaviourofthemachine.Todemonstratethedynamicnonlineareffects,a
perturbation signal is applied to any input device in open loop conditions. The input and
outputsmaybeassociatedwiththenetworkequationsaswellasthedynamicdevices.[2]
65
TestSystemModel
A hydroelectric development includes in some form a waterdiverting structure, conduit to

carry water to the turbines and governors, generators, control and switching apparatus,
housingfortheequipment,transformers,andtransmissionlinestoelectricalnetwork.Insome
cases,asurgetankisprovidedtominimizethehydraulictransienteffects.
The general configuration of the test Electric Power System consists of a single synchronous
machineconnectedtoalargesystemthroughatransmissionline.Thesynchronousmachineis
drivenbyahydraulicturbinetypeFrancis.Forthepurposeofstabilityanalysis,thesystemwas
reducedtoasinglemachineinfinitebusbarconfiguration.
Thischapterpresentsthephysicalcharacteristicsandcapabilityofthecomponentsofthetest
system such as water upstream tunnel, surge tank, penstock, a hydraulic turbine, speed
governor, generator, and electrical network. The dynamic behaviour of these components is
described by a set of nonlinear equations and is modelled in the power simulation software
SIMPOW.
9.1 OverviewofthetestSystem
Thetestsystemmodelusedforstabilitystudiesiscomposedofasynchronousmachineinfinite
bus power system shown in Figure 91. The hydraulic turbine is connected to a synchronous
generator and the generator is connected to an infinite bus through a transmission line. The
generated power is fed into the transmission system via a stepup transformer. The
transmissionsystemnominalvoltageis132kV.
Hydraulic
Turbine
Generator
Transformer
Transmission
line
Infinite Bus
Figure91:Singlemachineinfinitebuspowersystem
Thenetworkreactancesofthetestsystem,depictedinFigure91areinperuniton50MVA,
24kV base. The network reactance values are based on the exercise 12.2 of the Chapter 12,
page732,ofthetextbookPowerSystemStabilityandControl,[2].Resistancesareassumed
negligible.Thesystemparametersandinitialoperatingconditionoftheelectricalnetworkare
asfollows.
66
Table91:Systemparametersandoperatingconditions
Generator
Transformer
Transmissionline
InfiniteBus
Ratedapparentpower
Ratedphasetophasevoltage
Frequency
Numberofpairsofpoles
InertiaTimeConstant
Reactance
RatedVoltage
Reactance
Ratedvoltage
45.00+j15.00[MVA]
24.0036.01[kV]
50.00[Hz]
6.00[]
2.70[MWs/MVA]
0.15[pu]
24.00/132.00[kV]
0.50[pu]
0.9950.00[pu]
9.2 Modelsofthedifferentcomponents
9.2.1 PowerPlantModel
The hydroelectric power plant includes a canal to carry the water flow from the reservoir, a
surgetank,apenstocktoconveythewatertothepowerhouse,apowerhousefortheturbine
and generator, and a tailrace. The synchronous generator is mechanically coupled to the
turbine,andtheelectricaloutputpowerofthegeneratoriscarriedbythetransmissionlinesto
theelectricgrid.Agoverningsystemisprovidedtocorrectanychangesinthesystemfrequency
byopeningorclosingthewicketgatesoftheturbine.[6,10]
Figure92:AgeneralLayoutofHydroPowerPlant,[19]
ThelayoutofhydroelectricpowerplantusedforthestabilitystudiesisshowninFigure92.The
water conduit system has a static head between 220.0 m and 200.0 m. The power plant is
composedofanupstreamreservoir,a1307meterslonggallery,asurgetank,a435meterslong
penstock,aturbineand100meterslongdischargetailrace.ThepowerplanthasaFrancisunit.
ThemainparametersofthehydrauliccircuitaresummarizedinTable92.
67
Table92:HydraulicPowerPlantModel
Tunnel
SurgeTank
Penstock
FrancisTurbine
Tailrace/outlet
Length
1307.00 [m]
Diameter
3.90 [m]
Frictionconstant,f
0.06 []
Diameter
5.00 [m]
Waterlevelmeasured
300.00 [m]
Frictionconstant,f
0.05 []
Length
435.00 [m]
Diameter
2.00 [m]
Frictionconstant,f
0.01 []
Ratedmechanicalpower
46.15 [MW]
Ratedspeed
500.00 [rpm]
Rateddischarge
25.00 [m3/s]
Ratedhead
200.00 [m]
Referencediameter
3.90 [m]
Hydraulicefficiency
0.96 []
Ratedefficiency
0.94 []
Length
100.00 [m]
Diameter
3.90 [m]
Frictionconstant,f
0.06 []
The value considered for the pressure wave velocity a is 1200 m/s. The dimensionless pipe
friction factor is a function of the Reynolds number and the relative roughness height,
describedinChapter2.1.3.ThefrictionfactorisselectedfromtheMoodydiagram.
9.2.2 HydraulicTurbine
ThehydraulicturbineisatypeFrancis.Inareactionturbinehavinglongpressuretunnels,the
dynamicresponseofbothupstreamanddownstreamwatercolumnsmustbestudiedinorder
toavoidexcessivepressureoroverspeedproblems.
Thehydraulicturbineisdesignedbymodelsofvaryingdegreesofdetailforthestabilityanalysis
of The Hydraulic Power Generating System. The hydraulic turbine parameters used in the
modellingandsimulationofthehydroturbinemodelsaregiveninTable93.
Table93:HydraulicTurbineParameters
Parameter
TurbineDamping
GatePositionatRatedCondition
GateOpeningatFullLoad
GateOpeningatNoLoad
NoloadLoad
ConstantProportionality
WaterTimeConstantinPenstock
WaterTimeConstantinUpstreamTunnel
SurgeTankFillingTime
Value
KD
0.500 []
0.900 [pu]
yFL
0.960 [pu]
yNL
0.172 [pu]
QNL
4.300 [m3/s]
At
1.250 []
TW1
1.764 [s]
TW2
1.394 [s]
Ts 157.079 [s]
68
Themathematicalmodelsofthehydraulicturbinedeveloped,simulatedandanalyzedforthe
stabilityanalysisofahydraulicpowergeneratingsystemarelistedinTable94.Itisworthyto
noticethatthelinearmodelsusesthetypicalvaluesrecommendbyIEEE,detailedinTable32.
Table94:HydraulicTurbineModels
Model Description
1
SimplifiedNonlinearTurbineModelassumingInelasticWaterColumn
2
Nonlinear Turbine Model without Surge Tank assuming Inelastic Water
Column
3
Nonlinear Turbine Model without Surge Tank including Elastic Water
ColumnEffect
4
Nonlinear Turbine Model with Surge Tank assuming Inelastic Water
ColumnsinPenstockandTunnel
5
NonlinearTurbineModelwithSurgeTankassumingElasticWaterColumn
inPenstockandInelasticWaterColumninTunnel
6
LinearTurbineModelwithSurgeTankconsideringInelasticWaterColumns
inPenstockandTunnel,andturbinecoefficientsrecommendbyIEEE
7
Linear Turbine Model with Surge Tank including Elastic Water Column in
Penstock,andturbinecoefficientsrecommendbyIEEE
9.2.3 SynchronousGenerator
The synchronous generator is modelled by a detailed model with damping, TYPE 2, which
consistsofonefieldwinding,onedamperwindingindaxisandonedamperwindinginqaxis
with saturation included. The synchronous machine is modelled without automatic voltage
regulator.ThissynchronousmachinemodelTYPE2isastandardcomponentinthesimulation
softwareSIMPOWthatisdescribedindetailintheSIMPOWUserManual.
Thesynchronousmachineparametersusedinthemodellingofthesynchronousgeneratorare
shown in Table 95. These parameters are based on typical values of standard parameters
detailedintheTable4.2oftheChapter4ofthetextbookPowerSystemStabilityandControl,
[2].Thereactancevaluesareinperunitusingthemachineratedvaluesasthebasevaluesand
thetimeconstantsareinseconds.
Table95:SynchronousmachineParameters
Parameter
StatorResistance
Directaxissynchronousreactance
Directaxistransientreactance
Directaxissubtransientreactance
Quadratureaxissynchronousreactance
Quadratureaxissubtransientreactance
Statorleakagereactance
Opencircuitdaxistransienttimeconstant
Opencircuitdaxissubtransienttimeconstant
Opencircuitqaxissubtransienttimeconstant
Ra
xd
xd
xd
xq
xq
xl
Td0
Td0
Tq0
Value
0.003 [pu]
1.174 [pu]
0.300 [pu]
0.215 [pu]
0.770 [pu]
0.150 [pu]
0.101 [pu]
8.500 [sec]
0.050 [sec]
0.250 [sec]
69
9.2.4 GoverningSystemsforHydraulicTurbines
TheturbinegoverningsystemismodelledbyageneralspeedgoverningmodelTYPESG1.The
physicalcharacteristicsandcapabilityoftheturbinegoverningsystemaredetailedinChapter5.
ThegovernormodeltypeSG1isavailableintheDSLlibrariesinthesoftwareSIMPOWthatis
documentedintheSIMPOWUserManual.Therotorspeedrissetastheinputspeedsignal.
The input signal for the permanent droop compensation is the pilot or main servomotor
positiony.
Theturbinegoverningsystemparametersusedonthetestsystemarebasedonthegovernor
model detailed in the Figure 9.10 of the Chapter 9 of the textbook in Reference [2]. These
valuesareshowningivenastypicalvaluesinTable96.
Table96:TypicalValuesofparametersofturbinegoverningsystem
Parameter
Pilotvalvetimeconstant
Servogain
Mainservomotortimeconstant
Permanentdroop
Temporarydroop
Resettime
Maximumgateopeningrate
Maximumgateclosingrate
Maximumgatepositionlimit
Minimumgatepositionlimit
TP
Ks
Tg
RP
R T
TR
YPMAX
YPMIN
YMAX
YMIN
Value
0.05 [s]
5.00 []
0.20 [s]
0.04 []
0.40 []
5.00 [sec]
0.10 [pu/s]
0.10 [pu/s]
1.00 []
0.00 []
9.3 PowerFlowAnalysis
Thepowerflowanalysisinvolvesthestudyoftheflowofactiveandreactivepower,losses,and
thevoltageprofilefornormalsymmetricalsteadyoperatingconditioninthegivensystem.The
PowerFlowmoduleOPTPOW,partofSIMPOW,calculatesinitialpowerflowsinthenetwork.
The test system, shown in Figure 91, contains three buses. The synchronous generator is
connected in the bus 1. The generation at bus 1 is modelled as UPnode. The transformer is
connected between bus 1 and bus 2. Bus 3 is the swing bus and represents an infinite bus.
Table97showsthepowerflowresultsfromtheOPTPOWcalculation.
Table97:PowerFlowresults
Name
Up.u.
UkV
FI(u)Deg. P1MW
Q1Mvar P2MW
Q2Mvar
BUS1
1.0000 24.0000 36.0109
TR2BUS1BUS20
45.0000 15.0107 45.0000
8.2597
PRODSYNKRONGEN
45.0000 15.0107
0.0000
0.0000
BUS2
0.9645 127.3090 27.9646
LINEBUS2BUS30
45.0000
8.2597 45.0000 14.2433
TR2BUS1BUS20
45.0000 15.0107 45.0000
8.2597
BUS3
0.9950 131.3400
0.0000
LINEBUS2BUS30
45.0000
8.2597 45.0000 14.2433
PRODSWING
45.0000 14.2433
0.0000
0.0000
70
9.4 PowerSystemStabilityAnalysisTheoreticalResults
Thecircuitmodelrepresentingtheinitialoperatingconditionwithalltheparametersexpressed
inperunitisshowninFigure93.Thesynchronousgeneratorismodelledbytheclassicalmodel
andalltheresistancesneglected.
Figure93:Circuitmodelofthetestsystem
Assumingthegeneratorvoltagetobereference,thegeneratorstatorcurrentis
P jQ P jQ
0.90 j 0.30
I0
0.948717.5760
Vg
V
1.00
36.01
(9.1)
Theinternalvoltageforthetransientstateis:
E' Vg jX d' I0 1.0036.01 j0.3 0.9517.58 1.122949.9234
Thetransientsynchronizingpowercoefficient KE' is:
E VS
1.123 0.995
KE '
cos
cos 49.92 0.7572
xd
0.95
Linearizedsystemequations,fromEquation(8.16),give
N
t 0.1852 K D 0.1402 N 0.1852
mm
0 0
100
t
Theeigenvaluesofthestatematrix,Equation(9.4),aregivenby
0.1852 K D 0.1402
det
2 0.1852 K D 14.0212 0
100

(9.2)
(9.3)
(9.4)
(9.5)
71
Theundampednaturalfrequency,fromEquation(8.20),gives
nat 0.1402
100
6.6372 rad/s 1.0563 Hz
2 2.7
(9.6)
Thedampingratio,fromEquation(8.21),is
KD
1
0.0140K D
(9.7)
2 2 2.7 6.6372
ThetheoreticalresultsforthistestsystemfordifferentvaluesofKDaredetailedinTable98.
Table98:ResultsoftheSynchronousmachinerepresentedbyClassicalModel
Description
Results
Eigenvalues,
0.00j6.6372 0.0463j6.6371 0.09266.6366
Dampingcoefficient,KD
0.00
0.50
1.00
DampingRatio,
0.00
0.0070
0.0140
Dampedfrequency,d
1.0563[Hz]
1.0563[Hz]
1.0562[Hz]
Undampednaturalfrequency,nat 1.0563[Hz]
1.0563[Hz]
1.0563[Hz]
9.4.2 Naturalfrequencyanalysis
Thefirstnaturalfrequencyofthepenstockcorrespondstothefrequencycorrespondingtothe
fundamental period of the system, described in Equation (2.4). The first natural frequency of
thepenstockis
a
1200
f0
0.8305
(9.8)
4 L 4*435
The natural frequency of the massoscillation between the upstream reservoir and the surge
tankisgivenby
1
1
1
1
f0
0.010755
2 L AS 2 1307.00 2.50 2
(9.9)
g At
9.81 3.90
72
10
DynamicSimulationsinSIMPOW
Once the Hydro Power Plant models are established, the different models are implemented,
simulated and studied in the software SIMPOW. SIMPOW is a software system for digital
SimulationandanalysisofelectricalPowersystems.Digitalmodels,staticanddynamic,ofmost
elementsinapowersystemareavailablefortheconventionalcalculationofpowerflow,fault
currents and time simulation of shortterm dynamics. SIMPOW has been designed with a
numericaltechniqueandeventshandling,whicharerobustandgiveaccurateresults.
Thischapterdealswiththestudyofthesystemstabilitycharacteristicsofthedifferenthydro
turbine models within a typical hydraulic power generating system implemented in the
softwareSIMPOW.Theobjectiveofthedynamicmodellingofthetestsystemistoanalyzethe
stabilitycharacteristicsoftheHydroPowerSystemaboutthesteadystateoperatingcondition
followingathreephasefaulttogroundatt=1s.Thisdisturbanceissimulatedforaperiodof5.0
ms.Thesamplingperiodis50seconds.
The hydraulic power generating system for the stability analysis has been designed with a
detailedsynchronousgeneratormodelwithdamping,ageneralspeedgoverningmodel,anda
turbine. The hydraulic turbine is modelled with varying degrees of detail. The mathematical
modelsofthehydraulicturbinearedetailedinTable94,andlistedbelow
Model1 SimplifiedNonlinearTurbineModel
Model2 NonlinearTurbineModelwithoutSurgeTankassumingInelasticWaterColumn
Model3 NonlinearTurbineModelwithoutSurgeTankincludingElasticWaterColumnEffect
Model4 NonlinearTurbineModelwithSurgeTankassumingInelasticWaterColumns
Model5 Nonlinear Turbine Model with Surge Tank assuming Elastic Water Column in
PenstockandInelasticWaterColumninUpstreamTunnel
Model6 Linear Turbine Model with Surge Tank considering Inelastic Water Columns in
Penstock and Tunnel, and turbine characteristics based in the turbine coefficients
recommendbyIEEE
Model7 Linear Turbine Model with Surge Tank including Elastic Water Column in Penstock,
andturbinecharacteristicsbasedintheturbinecoefficientsrecommendbyIEEE
Frequencyresponseanalysisaswellastransientanalysisisperformedtoevaluatetheeffectsof
the detailed modelling of the turbine and conduit system to the stability analysis and the
dynamicperformanceofthehydraulicpowergeneratingsystem.
ThedynamicperformanceofthedifferentHydroPowerSystemmodelshasbeeninvestigated
utilizingtheeigenvalueanalysistoanalyzethebehaviourofasingleunitconnectedtoapower
system. The computed eigenvalues are analyzed in Dynpow of the software SIMPOW. The
values of the real part of the eigenvalues are measured in [1/s] and the imaginary part is
measuredin[Hz].
73
10.1 MODEL1:SimplifiedNonlinearTurbineModel
The hydraulic turbine in the model 1 is design by a Simplified Nonlinear Turbine Model,
described in Chapter 6.2.1. The penstock is modelled assuming an inelastic conduit,
incompressible fluid and neglecting friction losses. The nonlinear characteristics of hydraulic
turbine are neglected in this model. The block diagram of this hydro power plant model is
showninFigure62.
TheparametersofthehydropowerplantandoperatingconditionsaredetailedinChapter9.
The simplified nonlinear turbine model, type HYTUR, is available in the SIMPOW DSLlibrary
thatisdocumentedintheSIMPOWUsersManual.
Table 101 shows the realvalued eigenvalues, nonrealvalued eigenvalues and the state
variables that have the highest participation of the test system with a simplified nonlinear
turbinemodel.
Table101:Eigenvaluesandthestatevariableforamodel1
1
2
3,4
5
6,7
8
9
10
11
Eigenvalue
Damping
Ratio
Real,[1/s] Imag,[Hz]
23.0168
0.0000
17.5102
0.0000
0.4073
1.0538 0.06140
6.7770
0.0000
3.7911
0.1221 0.98015
1.2927
0.0000
0.0167
0.0000
0.1773
0.0000
1.0000
0.0000
Frequency
[Hz]
1.05383
0.12205
Dominant
States
D
and
Q
ItisseeninTable101that
Alleigenvalueshavenegativerealpart.
Thesystemisstable.
1,2,5,8,9,10and11influencethedynamicsystemwithpuredamping,
3,4and6,7influencethedynamicsimulationwithbothdampingandoscillation.
Theoscillatorymode 3,4haveadampedfrequencyof1.0538Hz.Theoscillationsdecay
with a time constant of 1/0.4073 seconds. This corresponds to a damping ratio of
0.0614.
The magnitudes of the participation factors, computed in SIMPOW, are shown in Table 102.
The angles of the participation factors do not provide any useful information. From the
participationmatrix, and haveahighparticipationintheoscillatorymodeassociatedto
theeigenvalues 3,4.Thefieldfluxlinkagehasahighparticipationinthenonoscillatorymode
representedbytheeigenvalue 10.Thedaxisfluxlinkagehasahighparticipationonthenon
oscillatory mode represented by the eigenvalue 1. The qaxis flux linkage has a high
participationonthenonoscillatorymoderepresentedbytheeigenvalue5.
74
Table102:Participationmatrixforamodel1
1
0.00627
0.00582
2
0.00037
0.00290
3
0.54104
0.53081
4
0.54104
0.53081
5
0.02714
0.01898
6
0.22094
0.08866
7
0.22094
0.08866
8
0.06553
0.00244
9
0.00391
10
0.00088
0.00043
0.00575
0.00001
0.00751
0.00751
0.00036
0.00849
0.00849
0.00812
0.00016
0.98838
D 1.00647
0.00044
0.01323
0.01323
0.00085
0.00455
0.00455
0.00015
0.00000
0.00888
Q 0.00004
68.50
1.0008
68.00
1.0006
67.50
1.0004
SPEED (pu)
ANGLE (degrees)
0.00006 0.03267 0.03267 0.99570 0.07283 0.07283 0.00203 0.00000 0.00185
10.1.2 DynamicSimulationAnalysis
Figure101showstheresultofsimulatingathreephasefaulttogroundonbusBUS2att=1s.
Thisdisturbanceissimulatedforaperiodof5.0ms.Thesimulationresultsofasimpleturbine
withoutsurgetankrepresentedbymodel1showsthedynamicbehaviouroftheangle,speed,
mechanicaltorque,gateposition,flowrateandheadpressure.
67.00
66.50
1.0002
1.0000
0.9998
66.00
0.9996
65.50
0.9994
65.00
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0
2.0
4.0
TIME SECONDS
8.0
10.0
12.0
14.0
10.0
12.0
14.0
10.0
12.0
14.0
TIME SECONDS
0.9030
45.70
45.69
0.9029
GATE OPENING (pu)
MECHANIC TORQUE (MNm)
6.0
45.68
45.67
45.66
45.65
45.64
0.9028
0.9027
0.9026
0.9025
45.63
0.9024
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0
2.0
4.0
TIME SECONDS
6.0
8.0
TIME SECONDS
0.90274
1.0004
PRESSURE HEAD (pu)
FLOW RATE (pu)
0.90272
0.90270
0.90268
0.90266
0.90264
0.90262
1.0000
0.9998
0.9996
0.9994
0.90260
0.0
1.0002
2.0
4.0
6.0
8.0
TIME SECONDS
10.0
12.0
14.0
0.0
2.0
4.0
6.0
8.0
TIME SECONDS
Figure101:Faultsimulationresults:(a)angle,(b)speed,(c)mechanicaltorque,(d)gateposition,(e)
flowrateand(f)headpressureofModel1
The initial value of the Mechanical Power is 45.5214 MW. After the fault is simulated, the
mechanical power drops to 0.00 MW. The mechanical power oscillates for a period of 6
seconds,approximately.
75
Theelectricalangleoscillatesforaperiodof7.80seconds.Theminimumvalueis64.9139at
1.74 seconds and the maximum value is 68.51 at 1.22 seconds. The value of the electrical
angleatsteadystateis66.61.
Therotationalspeedoscillatesbetweenthevaluesof0.9993puand1.0008pu.Therotational
speedrecoversstabilityafter8seconds,demonstratingthatthedynamicresponseofthehydro
powerplantisefficient.
The initial value of Mechanical Torque and Electric Torque are 45.67 MNm and 45.66 MNm,
respectively.TheMechanicTorqueoscillatesforaperiodof8seconds,betweenthevaluesof
45.62 MNm and 45.70 MNm. The Mechanical Torque at steady state is 45.6678 MNm. The
period of oscillation of Electrical Torque is 5.50 seconds, approximately. Electrical Torque
oscillatesbetweenthevaluesof47.08MNmand1.38MNm.TheElectricalTorqueatsteady
stateis45.67MNm.
The variables such as gate opening, flow rate and water pressure head are measured in per
unit. The gate position oscillates between 0.9024 and 0.9030. The flow rate varies between
0.9025and0.9027.Thepressureheadoscillatesbetween0.9993and1.0005.Thevaluesofgate
position, flow rate and water pressure at steadystate are 0.9027, 0.9027 and 1.0000,
respectively.
ThemagnitudeandphaseofthetransferfunctionfromgatepositiontoMechanicalPowerofa
SimplifiedNonlinearModelareplottedinFigure102.Theperturbationsignalisappliedtothe
gatepositioninopenloopconditionsandthesignalmechanicalpowerismonitored.
MW/GATE - Magnitude (dB)
50
-50
-100
-150 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
MW/GATE - Phase(deg)
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Figure102:FrequencyresponseofthetransferfunctionfromGatepositiontoMechanicalPowerof
model1
76
MT/GATE - Phase(deg)
MT/GATE - Magnitude (dB)
Therelationshipbetweenthemechanicaltorqueandthevariationofgatepositionisstudiedin
ordertoanalyzethedynamicbehaviourofthesystem.Theperturbationsignalisappliedtothe
gate position in open loop conditions and the signal mechanical torque is monitored. The
magnitude and phase of the frequency response of the transfer function of the hydraulic
turbineareplottedinFigure103.
7
6
5
4
3
2
1 -3
10
0
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-50
-100
-150
-200 -3
10
Frequency [Hz]
Figure103:FrequencyresponseofthetransferfunctionfromGatepositiontoMechanicalTorqueof
Model1
HEAD/GATE - Magnitude (dB)
Figure 104 plots the computed frequency response of the transfer function of the conduit
system of a Simplified Nonlinear Turbine Model. The transfer function of the conduit system
relatesthewaterpressuretogateposition.
10
0
-10
-20
-30
-40
-50 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
HEAD/GATE - Phase(deg)
-80
-100
-120
-140
-160
-180 -3
10
Frequency [Hz]
Figure104:FrequencyresponseofthetransferfunctionoftheconduitsystemofModel1
77
TETA/GATE - Phase(deg)
TETA/GATE - Magnitude (dB)
Figure 105 plots the magnitude and phase of the transfer function from gate position to
ElectricalPowerAngle.
50
-50
-100
-150 -3
10
0
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Figure105:Frequencyresponseofthetransferfunctionfromgatepositiontoelectricalangleof
Model1
The magnitude and phase of the transfer function of the mechanicalhydraulic governor of a
Simplified Nonlinear Model are plotted in Figure 106. The transfer function of the governor
relatesthegatepositiontospeedchanges.
GATE/SPEED - Phase(deg)
GATE/SPEED - Magnitude
40
20
0
-20
-40
-60 -3
10
0
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Figure106:Frequencyresponseofthetransferfunctionofthemechanicalhydraulicgovernorof
Model1
78
10.2 MODEL 2: Nonlinear Turbine Model without Surge Tank assuming

InelasticWaterColumn
The hydraulic turbine in the model 2 is represented by a Nonlinear Turbine Model without
Surge Tank assuming Inelastic Water Column, described in Chapter 6.2.2. The Penstock is
modelledassuminganinelasticconduitandincompressiblefluidwherethetravellingpressure
wave effects are relatively insignificant. The nonlinear characteristics of the hydraulic turbine
areneglectedinthismodel.TheblockdiagramofthisconfigurationisshowninFigure63.
TheparametersoftheHydroPowerPlantandoperatingconditionsaredetailedinChapter9.
This model was implemented in DSL Code Generator which is a tool used for making self
defineddynamicmodelsfordynamicsimulationinSIMPOW.
variablesthathavethehighestparticipationofthetestsystemwithaNonlinearTurbineModel
withoutSurgeTankassumingInelasticWaterColumn.
Table103:EigenvaluesandthestatevariableforModel2
Eigenvalue
Damping
Ratio
1 23.0319
0.0000
2 17.5091
0.0000
Frequency
[Hz]
Dominant
States
D
0.4203
1.0649
0.06270
1.06485
and
6.7511
0.0000
3.7861
0.1176
0.98148
0.11762
1.2895
0.0000
0.0167
0.0000
10
0.1814
0.0000
11
1.0000
0.0000
3,4
5
6,7
q
f
Thesystemisstable.
1,2,5,8,9,10,and11influencethedynamicsystemwithpuredamping.
3,4,and6,7influencethedynamicsimulationwithbothdampingandoscillation.
withatimeconstantof1/0.4203s.Thiscorrespondstoadampingratioof0.0627.
79
Table104:ParticipationfactorsmatrixofaturbineModel2
1
10
0.0061
0.0057
0.0004
0.0028
0.5412
0.5316
0.5412
0.5316
0.0057
0.0000
0.0071
0.0071
0.2246
0.0893
0.0084
0.2246
0.0893
0.0084
0.0636
0.0023
0.0077
0.0036
0.0291
0.0203
0.0004
0.0001
0.0008
0.0004
0.9886
1.0062
0.0004
0.0128
0.0128
0.0009
0.0045
0.0045
0.0001
0.0000
0.0087
0.0000
0.0001
0.0350
0.0350
0.9947
0.0801
0.0801
0.0021
0.0000
0.0019
10.2.2 DynamicResponseAnalysis
withoutsurgetankrepresentedbymodel2showsthedynamicbehaviourofthepowerangle,
rotationalspeed,mechanicaltorque,gateposition,flowrateandheadpressure.
1.0008
67.00
1.0006
1.0004
66.00
SPEED (pu)
ANGLE (degrees)
66.50
65.50
65.00
1.0002
1.0000
0.9998
64.50
0.9996
64.00
0.9994
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0
2.0
4.0
TIME SECONDS
6.0
8.0
10.0
12.0
14.0
10.0
12.0
14.0
10.0
12.0
14.0
TIME SECONDS
0.9030
0.9029
45.09
GATE OPENING (pu)
45.10
45.08
45.07
45.06
45.05
0.9028
0.9027
0.9026
0.9025
45.04
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0
2.0
4.0
TIME SECONDS
0.9924
PRESSURE HEAD (pu)
0.89908
FLOW RATE (pu)
8.0
TIME SECONDS
0.89910
0.89906
0.89904
0.89902
0.89900
0.9922
0.9920
0.9918
0.9916
0.9914
0.89898
0.0
6.0
2.0
4.0
6.0
8.0
TIME SECONDS
10.0
12.0
14.0
0.0
2.0
4.0
6.0
8.0
TIME SECONDS
Figure107:FaultsimulationResults:(a)Powerangle,(b)Speed,(c)MechanicalTorque,(d)Gate
Position,(e)FlowRateand(f)HeadPressureofModel2
Mechanical Power drops to 0.00 MW. The Mechanical Power oscillates for a period of 5.2
80
1.69secondsandthemaximumvalueis67.2254at1.22seconds.Thevalueoftheelectrical
stateis45.07MNm.
The variables Gate Opening, Flow rate and Pressure Head are measured in per unit. Gate
opening oscillates between 0.9024 and 0.9030. The flow varies between 0.8989 and 0.8991.
Thepressureheadoscillatesbetween0.9919and0.9925.ThevaluesofGateposition,Flowrate
andPressureHeadatsteadystateare0.9027,0.8990and0.9919,respectively.
NonlinearTurbineModelwithoutSurgeTankassumingInelasticWaterColumnareplottedin
Figure108.Theperturbationsignalisappliedtothegatepositioninopenloopconditionsand
thesignalmechanicalpowerismonitored.
-50
-100
-150
-200 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Model2
81
7
6
5
4
3
2
1 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-50
-100
-150
-200 -3
10
Frequency [Hz]
Model2
systemoftheModel2.ThetransferfunctionoftheconduitsystemrelatesthePressureHead
togateposition.
10
0
-10
-20
-30
-40 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-120
-140
-160
-180 -3
10
Frequency [Hz]
Figure1010:FrequencyresponseofthetransferfunctionoftheconduitsystemofModel2
82
-50
-100
-150
-200 -3
10
-150
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Figure1011:FrequencyresponseofthetransferfunctionfromGatepositiontoElectricalAngleof
Model2
The magnitude and phase of the transfer function of the mechanicalhydraulic governor are
plottedinFigure1012.Thetransferfunctionofthegovernorrelatesthegatepositiontospeed
changes.
40
20
0
-20
-40
-60 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Figure1012:FrequencyResponseofthetransferfunctionofthemechanicalhydraulicgovernorof
Model2
83
10.3 MODEL 3: Nonlinear Turbine Model without Surge Tank including

ElasticWaterColumnEffects
The hydraulic turbine in the model 3 is represented by a Nonlinear Turbine Model without
SurgeTankincludingElasticWaterColumnEffects,describedinChapter6.2.3.ThePenstockis
modelledtakingintoaccounttheelasticwaterhammertheoryandthehydraulicfrictionlosses.
The nonlinear characteristics of the hydraulic turbine are neglected in this model. The block
diagramofthisconfigurationisshowninFigure64.
This model was implemented in DSL Code Generator which is a tool used for making self
defineddynamicmodelsfordynamicsimulationinSIMPOW.
withoutSurgeTankincludingElasticWaterColumnEffects.
Table105:EigenvaluesandthestatevariablefortheTurbineModel3
Eigenvalue
Damping
Ratio
1 23.0313
0.0000
2
3,4
5
6,7
Frequency
[Hz]
Dominant
States
0.0000
0.5188
1.0587
0.07775
1.05872
and
6.8018
0.0000
17.5710
3.6763
0.2128
0.93977
0.21280
1.3377
0.0000
0.0167
0.0000
10
0.1814
0.0000
11
1.0000
0.0000
Thesystemisstable.
84
Table106:ParticipationfactorsmatrixofaTurbineModel3
1
0.00614
0.00572
2
0.00076
0.00584
3,4
0.55712
0.54235
5
0.02676
0.01463
0.00027
6 ,7
0.22299
0.08774
0.00784
8
0.10336
0.00410
0.01188
0.00572
0.00002
0.00742
D 1.00644
0.00087
Q 0.00004
0.00013
9
0.00337
0.00013
10
0.00081
0.00042
0.98850
0.01325
0.00064
0.00436
0.00025
0.00000
0.00873
0.03659
0.97540
0.06619
0.00365
0.00000
0.00192
withoutsurgetankrepresentedbymodel3showsthedynamicbehaviourofthepowerangle,
1.0008
67.00
1.0006
1.0004
66.00
SPEED (pu)
ANGLE (degrees)
66.50
65.50
65.00
1.0002
1.0000
0.9998
64.50
0.9996
64.00
0.0
0.9994
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
10.0
12.0
14.0
10.0
12.0
14.0
TIME SECONDS
45.10
0.9030
45.09
0.9029
GATE OPENING (pu)
TIME SECONDS
45.08
45.07
45.06
0.9028
0.9027
0.9026
45.05
0.9025
45.04
0.0
2.0
4.0
6.0
8.0
10.0
12.0
0.9024
0.0
14.0
2.0
4.0
TIME SECONDS
6.0
8.0
TIME SECONDS
0.9924
PRESSURE HEAD (pu)
FLOW RATE (pu)
0.89915
0.89910
0.89905
0.89900
0.9922
0.9920
0.9918
0.9916
0.9914
0.89895
2.0
4.0
6.0
8.0
TIME SECONDS
10.0
12.0
14.0
0.0
2.0
4.0
6.0
8.0
TIME SECONDS
Figure1013:Faultsimulationresults:(a)angle,(b)speed,(c)mechanicaltorque,(d)gateposition,(e)
flowrateand(f)waterpressureofModel3
85
stateis45.07MNm.
NonlinearTurbineModelwithoutSurgeTankincludingElasticWaterColumnEffectsareplotted
inFigure108.Theperturbationsignalisappliedtothegatepositioninopenloopconditions
andthesignalmechanicalpowerismonitored.
-50
-100
-150
-200 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Model3
86
7
6
5
4
3
2
1 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Model3
systemoftheModel3.Thetransferfunctionoftheconduitsystemrelatesthewaterpressure
togateposition.
10
0
-10
-20
-30
-40 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-150
-200
-250
-300 -3
10
Frequency [Hz]
Figure1016:FrequencyResponseofthetransferfunctionoftheconduitsystemofModel3
87
-50
-100
-150
-200 -3
10
0
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Model3
changes.
40
20
0
-20
-40
-60 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Model3
88
10.4 MODEL 4: Nonlinear Turbine Model with Surge Tank assuming

InelasticWaterColumns
Thehydraulicturbineinthemodel4isrepresentedbyaNonlinearTurbineModelwithSurge
TankassumingInelasticWaterColumnsinPenstockandTunnel,describedinChapter6.2.4.The
upstream tunnel and penstock are modelled assuming an incompressible fluid and a rigid
conduitwherethetravellingpressurewaveeffectsarerelativelyinsignificantly.Thenonlinear
characteristics of hydraulic turbine are neglected in this model. The block diagram of this
configurationisshowninFigure65.TheparametersoftheHydroPowerPlantandoperating
conditions are detailed in Chapter 9. This model was implemented in DSL Code Generator
which is a tool used for making selfdefined dynamic models for dynamic simulation in
SIMPOW.
withSurgeTankAssumingInelasticWaterColumnsinPenstockandTunnel.
Table107:EigenvaluesandthestatevariablefortheturbineModel4
Eigenvalue
Damping
Ratio
1 23.1362
0.0000
Dominant
States
0.0000
0.5102
1.1313
0.07160
1.13126
and
6.5735
0.0000
3.7469
0.0840
0.99021
0.08405
1.1626
0.0000
0.0167
0.0000
0.0392
0.0091
0.56646
0.00909
3,4
5
6,7
10,11
17.5026
Frequency
[Hz]
12
0.2060
0.0000
13
1.0000
0.0000
ItisseenTable107that
Thesystemisstable.
3,4,6,7and10,11influencethedynamicsimulationwithbothdampingandoscillation.
89
Table108:ParticipationfactorsmatrixofahydraulicturbineModel4
1
0.00508
0.00465
2
0.00041
0.00244
3 ,4
0.54296
0.53682
5
0.04179
0.02836
0.00049
6,7
0.27162
0.10359
0.00893
8
0.04706
0.00131
0.00540
0.00556
0.00001
0.00514
D 1.00424
0.00030
0.01042
0.00106
0.00451
0.00006
Q 0.00005
0.00008
0.04992
0.98495
0.14922
0.00199
9
0.00244
10,11
0.00032
0.00006
0.00004
12
0.00048
0.00035
0.98932
0.00785
0.00000
0.00223
with surge tank represented by model 4 show the dynamic behaviour of the power angle,
59.50
1.0006
59.00
1.0004
SPEED (pu)
ANGLE (degrees)
1.0008
58.50
58.00
1.0002
1.0000
0.9998
57.50
0.9996
57.00
0.9994
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0
2.0
4.0
TIME SECONDS
6.0
8.0
10.0
12.0
14.0
10.0
12.0
14.0
10.0
12.0
14.0
TIME SECONDS
0.9030
GATE OPENING (pu)
0.9031
41.77
41.76
41.75
41.74
41.73
0.9029
0.9028
0.9027
41.72
0.0
2.0
4.0
6.0
8.0
10.0
12.0
0.9026
0.0
14.0
2.0
4.0
TIME SECONDS
0.9464
PRESSURE HEAD (pu)
FLOW RATE (pu)
8.0
TIME SECONDS
0.87816
0.87814
0.87812
0.87810
0.9462
0.9460
0.9458
0.9456
0.87808
0.0
6.0
2.0
4.0
6.0
8.0
TIME SECONDS
10.0
12.0
14.0
0.0
2.0
4.0
6.0
8.0
TIME SECONDS
Figure1019:Faultsimulationresults:(a)angle,(b)Speed,(c)MechanicalTorque,(d)GatePosition,
(e)FlowRateand(f)HeadPressureofModel4
90
41.7182MNmand41.7776MNm.TheMechanicalTorqueatsteadystateis41.7492MNm.The
oscillates between the values of 43.1007 MNm and 1.43839 MNm. The Electrical Torque at
steadystateis41.7492MNm.
opening oscillates between 0.9026 and 0.9030. The flow varies between 0.878059 and
0.878164. The pressure head oscillates between 0.945495 and 0.946452. The values of Gate
position, Flow rate and Pressure Head at steadystate are 0.9028, 0.8781 and 0.946032,
respectively.
Nonlinear Turbine Model with Surge Tank assuming Inelastic Water Columns in Penstock and
Tunnel are plotted in Figure 1020. The perturbation signal is applied to the gate position in
openloopconditionsandthesignalmechanicalpowerismonitored.
50
-50
-100
-150 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Model4
91
6
5
4
3
2
1
0
-1 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-50
-100
-150
-200 -3
10
Frequency [Hz]
Model4
system of model 4. The transfer function of the conduit system relates the Pressure Head to
gateposition.
10
5
0
-5
-10
-15
-20
-25 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-120
-140
-160
-180 -3
10
Frequency [Hz]
92
50
-50
-100
-150 -3
10
0
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Model4
changes.
40
20
0
-20
-40
-60 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Model4
93
10.5 MODEL5:NonlinearModelwithSurgeTankassumingElasticWater
ColumninPenstockandInelasticWaterColumninTunnel
Thehydraulicturbineinthemodel5isrepresentedbyaNonlinearTurbineModelwithSurge
Tank assuming Elastic Water Column in Penstock and Inelastic Water Column in Tunnel,
describedinChapter6.2.5.Penstockismodelledtakingintoaccounttheelasticwaterhammer
theory.Thenonlinearcharacteristicsofhydraulicturbinearenotconsideredinthismodel.The
blockdiagramofthisconfigurationisshowninFigure66.ThismodelwasimplementedinDSL
Code Generator which is a tool used for making selfdefined dynamic models for dynamic
simulationinSIMPOW.TheparametersoftheHydroPowerPlantandoperatingconditionsare
detailedinChapter9.
variables that have the highest participation of the test system with a Nonlinear Model with
SurgeTankassumingElasticWaterColumninPenstockandInelasticWaterColumninTunnel.
Table109:EigenvaluesandstatevariablesforModel5
Eigenvalue
Damping
Ratio
1 23.1350
0.0000
Frequency
[Hz]
Dominant
States
0.0000
0.5933
1.1269
0.08350
1.12692
and
6.6476
0.0000
3.6479
0.1899
0.95044
0.18991
1.3158
0.0000
0.0392
0.0091
0.56646
0.00909
11
0.0167
0.0000
12
0.2059
0.0000
13
1.0000
0.0000
2
3,4
5
6,7
8
9,10
17.5566
Thesystemisstable.
Themagnitudesoftheparticipationfactors,computedinSIMPOW,areshowninTable1010.
94
Table1010:ParticipationfactorsmatrixforModel5
1
0.00510
0.00469
2
0.00083
0.00500
3 ,4
0.55426
0.54516
5
0.03864
0.02059
0.00556
0.00001
0.00528
1.00443
0.00062
Q 0.00005
0.00017
f
D
0.00034
6 ,7
0.21962
0.08119
0.00662
8
0.09201
0.00334
0.00887
0.01068
0.00077
0.00349
0.00017
0.05161
0.95721
0.09945
0.00473
9 ,10
0.00019
11
0.00225
0.00002
0.00006
12
0.00049
0.00035
0.98932
0.00785
0.00000
0.00223
60.00
1.0008
59.50
1.0006
1.0004
59.00
SPEED (pu)
ANGLE (degrees)
with surge tank represented by model 5 shows the dynamic behaviour of the power angle,
58.50
58.00
1.0002
1.0000
0.9998
57.50
0.9996
57.00
0.9994
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0
2.0
4.0
TIME SECONDS
6.0
8.0
10.0
12.0
14.0
10.0
12.0
14.0
TIME SECONDS
0.9030
GATE OPENING (pu)
41.77
41.76
41.75
41.74
0.9029
0.9028
0.9027
41.73
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0
6.0
8.0
0.94630
PRESSURE HEAD (pu)
FLOW RATE (pu)
4.0
TIME SECONDS
0.87820
0.87815
0.87810
0.87805
0.0
2.0
TIME SECONDS
0.94620
0.94610
0.94600
0.94590
0.94580
0.94570
2.0
4.0
6.0
8.0
TIME SECONDS
10.0
12.0
14.0
0.0
2.0
4.0
6.0
8.0
TIME SECONDS
10.0
12.0
14.0
Figure1025:FaultsimulationResults:(a)angle,(b)Speed,(c)mechanicaltorque,(d)gateposition,(e)
95
TheinitialvalueofMechanicalTorqueandElectricTorqueare41.7515MNmand41.75MNm,
oscillates between the values of 43.1092 MNm and 1.43937 MNm. The Electrical Torque at
steadystateis41.7493MNm.
Nonlinear Model with Surge Tank assuming Elastic Water Column in Penstock and Inelastic
Water Column in Upstream Tunnel are plotted in Figure 1026. The perturbation signal is
applied to the gate position in open loop conditions and the signal mechanical power is
monitored.
50
-50
-100
-150 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Model5
96
6
5
4
3
2
1
0
-1 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Model5
system of model 5. The transfer function of the conduit system relates the Pressure Head to
gateposition.
10
0
-10
-20
-30
-40 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-150
-200
-250
-300 -3
10
Frequency [Hz]
97
50
-50
-100
-150 -3
10
0
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Model5
Themagnituderatioandphaseofthetransferfunctionofthemechanicalhydraulicgovernor
areplottedinFigure1030.Thetransferfunctionofthegovernorrelatesthegatepositionto
speedchanges.
40
20
0
-20
-40
-60 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Model5
98
10.6 MODEL 6: Linear Turbine Model with Surge Tank assuming

Thehydraulicturbineinthemodel6isrepresentedbyaLinearTurbineModelwithSurgeTank
assuming Inelastic Water Columns in Penstock and Tunnel, described in Chapter 6.3.1. The
nonlinear characteristics of hydraulic turbine and the inelastic water hammer effect are
considered in this model. The block diagram of this configuration is shown in Figure 67. The
parametersoftheHydroPowerPlantandoperatingconditionsaredetailedinChapter9.This
model was implemented in DSL Code Generator which is a tool used for making selfdefined
dynamicmodelsfordynamicsimulationinSIMPOW.
variables that have the highest participation of the test system with a Linear Turbine Model
withSurgeTankassumingInelasticWaterColumnsinPenstockandTunnel.
Table1011:EigenvaluesandthestatevariableforModel6
Eigenvalue
Damping
Ratio
1 23.0473
0.0000
Frequency
[Hz]
Dominant
States
0.0000
0.2925
1.0890
0.04270
1.08896
and
6.5781
0.0000
5.2853
0.0000
2.5995
0.0000
0.0167
0.0000
0.0037
0.0108
0.05513
0.01076
11
0.1851
0.0000
12
1.0000
0.0000
2
3,4
9,10
17.4271
Thesystemisstable.
1,2,5,6,7,8,11,and12influencethedynamicsystemwithpuredamping.
99
Table1012:ParticipationfactorsmatrixofaModel6
1
0.00592
0.00543
2
0.00018
0.00127
3 ,4
0.51300
0.51623
5
0.03538
0.04673
7
0.03901
0.00641
0.00203
8
0.00290
9 ,10
0.00038
0.00092
6
0.02546
0.02780
0.00102
0.00570
0.00000
0.00634
D 1.00554
0.00018
Q 0.00004
0.00003
11
0.00088
0.00041
0.98949
0.00010
0.00004
0.01184
0.00200
0.00124
0.00037
0.00000
0.00860
0.03431
1.13006
0.06890
0.00521
0.00000
0.00198
with surge tank represented by model 5 shows the dynamic behaviour of the power angle,
1.0008
66.00
1.0006
1.0004
65.00
SPEED (pu)
ANGLE (degrees)
65.50
64.50
64.00
63.50
1.0002
1.0000
0.9998
0.9996
63.00
0.9994
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0
2.0
4.0
TIME SECONDS
6.0
8.0
10.0
12.0
14.0
10.0
12.0
14.0
TIME SECONDS
GATE OPENING (pu)
0.9030
44.67
44.67
44.66
44.66
0.9029
0.9028
0.9027
0.9026
44.65
0.9025
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0
2.0
4.0
TIME SECONDS
6.0
8.0
TIME SECONDS
-0.0062978
-0.0062980
PRESSURE HEAD (pu)
FLOW RATE (pu)
0.89980
0.89970
0.89960
0.89950
-0.0062982
-0.0062984
-0.0062986
-0.0062988
0.89940
-0.0062990
0.0
2.0
4.0
6.0
8.0
TIME SECONDS
10.0
12.0
14.0
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
TIME SECONDS
Figure1031:FaultsimulationResults:(a)angle,(b)speed,(c)mechanicaltorque,(d)gateposition,(e)
100
The electrical angle oscillates for a period of 6.80 seconds. The minimum value is 62.696 at
speed recovers stability after 7.8 seconds, demonstrating that the dynamic response of the
hydropowerplantisefficient.
Electrical Torque oscillates between the values of 46.0743 MNm and 1.3905 MNm. The
ElectricalTorqueatsteadystateis44.6633MNm.
The variables Gate Opening and Flow rate are measured in per unit. Gate opening oscillates
between0.9024and 0.9030.Theflowvariesbetween0.8993and0.8999.ThevaluesofGate
positionandFlowrateatsteadystateare0.8996,and1.0794,respectively.
Linear Turbine Model with Surge Tank assuming Inelastic Water Columns in Penstock and
Upstream tunnel are plotted in Figure 1032. The perturbation signal is applied to the gate
positioninopenloopconditionsandthesignalmechanicalpowerismonitored.
50
-50
-100
-150 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Model6
101
2
0
-2
-4
-6
-8 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Model6
system of a Linear Turbine Model with Surge Tank assuming Inelastic Water Columns. The
transferfunctionofthewatersupplysystemrelatesthePressureHeadtogateposition.
-20
-40
-60
-80 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-150
-200
-250
-300 -3
10
Frequency [Hz]
102
50
-50
-100
-150 -3
10
0
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Model6
Themagnituderatioandphaseofthetransferfunctionofthemechanicalhydraulicgovernor
areplottedinFigure1036.Thetransferfunctionofthegovernorrelatesthegatepositionto
speedchanges.
40
20
0
-20
-40
-60 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Model6
103
10.7 MODEL7:LinearTurbineModelwithSurgeTankassumingElastic
WaterColumninPenstock
Thehydraulicturbineinthemodel7ismodelledbyaLinearModelwithSurgeTankassuming
ElasticWaterColumninPenstock,describedinChapter6.3.2.ThePenstockismodelledtaking
intoaccounttheelasticwaterhammertheoryandneglectingthehydraulicfrictionlosses.The
nonlinearcharacteristicsofhydraulicturbineareconsideredinthismodel.Theblockdiagramof
this configuration is shown in Figure 68. The parameters of the Hydro Power Plant and
operating conditions are detailed in Chapter 9. This model was implemented in DSL Code
Generatorwhichisatoolusedformakingselfdefineddynamicmodelsfordynamicsimulation
inSIMPOW.
variablesthathavethehighestparticipationofthetestsystemwithaLinearModelwithSurge
TankassumingElasticWaterColumninPenstock.
Table1013:EigenvaluesandthestatevariablefortheModel7
Eigenvalue
Damping
Imag,[Hz]
Ratio
Real,[1/s]
1 23.0440
0.0000
2 17.5548
0.0000
3,4
5
6,7
Frequency
[Hz]
Dominant
States
0.4674
1.0695
0.06939
1.06950
and
6.7715
0.0000
3.6698
0.1773
0.95688
0.17731
1.2029
0.0000
0.0167
0.0000
0.0037
0.0108
0.05504
0.01077
10,11
12
0.1846
0.0000
13
1.0000
0.0000
y
f
Thesystemisstable.
104
Table1014:ParticipationfactorsmatrixofaModel7
1
0.00599
0.00555
2
0.00068
0.00502
3 ,4
0.55228
0.53969
5
0.02888
0.01667
0.00570
0.00002
0.00704
1.00608
0.00073
Q 0.00004
0.00012
f
D
8
0.07832
0.00245
0.01018
9
0.00280
10 ,11
0.00019
0.00030
6 ,7
0.22085
0.08406
0.00789
12
0.00094
0.00041
0.98890
0.00010
0.00002
0.01279
0.00072
0.00415
0.00014
0.00000
0.00862
0.03773
0.97929
0.07153
0.00246
0.00000
0.00197
with surge tank represented by model 7 show the dynamic behaviour of the power angle,
1.0008
66.00
1.0006
1.0004
SPEED (pu)
ANGLE (degrees)
65.50
65.00
64.50
1.0002
1.0000
64.00
0.9998
63.50
0.9996
0.9994
63.00
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0
2.0
4.0
TIME SECONDS
6.0
8.0
10.0
12.0
14.0
10.0
12.0
14.0
10.0
12.0
14.0
TIME SECONDS
GATE OPENING (pu)
0.9030
44.75
44.74
44.73
44.72
0.9029
0.9028
0.9027
0.9026
0.9025
44.71
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0.0
2.0
4.0
TIME SECONDS
8.0
TIME SECONDS
0.90010
-0.00500
PRESSURE HEAD (pu)
FLOW RATE (pu)
6.0
0.90005
0.90000
0.89995
-0.00520
-0.00540
-0.00560
-0.00580
0.89990
0.0
2.0
4.0
6.0
8.0
10.0
TIME SECONDS
12.0
14.0
0.0
2.0
4.0
6.0
8.0
TIME SECONDS
Figure1037:FaultsimulationResults:(a)Powerangle,(b)Speed,(c)MechanicalTorque,(d)Gate
Position,(e)FlowRateand(f)HeadPressureofModel7
105
speed recovers stability after 7.0 seconds, demonstrating that the dynamic response of the
hydropowerplantisefficient.
ElectricalTorqueoscillatesbetweenthevaluesof46.146MNmand1.389MNm.TheElectrical
Torqueatsteadystateis44.7304MNm.
The variables Gate Opening and Flow rate are measured in per unit. Gate opening oscillates
between0.9025and 0.9030.Theflowvariesbetween0.8999and0.9001.ThevaluesofGate
positionandFlowrateatsteadystateare0.9027,and0.9000,respectively.
LinearModelwithSurgeTankassumingElasticWaterColumninPenstockareplottedinFigure
1038.Theperturbationsignalisappliedtothegatepositioninopenloopconditionsandthe
signalmechanicalpowerismonitored.
50
-50
-100
-150 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Model7
106
10
-5
-10 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Model7
system of a Linear Model with Surge Tank assuming Elastic Water Column in Penstock. The
transferfunctionoftheconduitsystemrelatesthePressureHeadtogateposition.
10
0
-10
-20
-30
-40 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-50
-100
-150
-200
-250
-300 -3
10
Frequency [Hz]
107
50
-50
-100
-150 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Model7
changes.
40
20
0
-20
-40
-60 -3
10
0
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Model7
108
11
PowerSystemModelledinLVTrans
LVTrans is a general hydraulic dynamic simulation tool used to calculate all kinds of systems
consistingmainlyoffluidfilledpipesandopenchannels.ThegeneralconfigurationoftheHydro
Power Plant modelled in LVTrans contains the essential components of a hydraulic power
system such as reservoir, an upstream tunnel to carry the water from the reservoir, a surge
tank, a penstock, a hydraulic turbine, speed governor and a tailrace through the water is
releasedbacktotheriver.
Thischapterstudiesthesystemdynamiccharacteristicsofthetesthydraulicsystem,detailedin
Chapter 9, modelled in the dynamic simulation tool LVTrans. The dynamic behaviour of the
hydraulicpowersystemisanalyzedintheprogramLVTransversionLVTrans8_1.1.2andversion
LVTrans86_1.3.1_T. Afterwards, the frequency response of the transfer function relating the
Pressure Head to Gate position simulated in the two different versions are compared and
studied.
11.1 DescriptionofLVTrans
LVTransisageneralhydraulicdynamicsimulationtoolbuiltinLabVIEW.LVTransmaybeused
tocalculateallkindsofsystemsconsistingmainlyoffluidfilledpipesandopenchannels.Itcan
also be used to calculate the acoustics of gasfilled tubes. LVTrans uses the Method of
Characteristic to solve the differential equations describing the hydraulic conduits. A detailed
onedimensionEulerequationsystemisusedtocalculatethefulldynamicbehaviourofthetest
system. LVTrans has been used in several projects in the North Sea, including the fire water
system,acousticcalculationsingaspipesonHeidrun,coolingwatersystemforOrmenLange.
HydropowerisaspecialfieldforLVTrans,asmostitemsarecustommadeforthispurpose.[68]
LVTrans has had many enhancements. The dynamic behaviour of the test Hydraulic Power
SystemissimulatedandanalyzedintwoversionsofLVTrans.
LVTrans version LVTrans8_1.1.2 is a version created specifically for the study of the dynamic
performanceandinteractionbetweenthehydraulicsystemandpowersystemofapowerplant
equipped with Francis units. This version was specially designed for detailed grid/waterway
interactioninJanuary2005.
LVTrans version LVTrans86_1.3.1_T is the newest version of the program. LVTrans version
86_1.3.1_Tincludesmanyupgradestothefrontpanel,blockdiagramandrelatedfunctionality
tothestudyofthedynamicperformanceofthehydraulicsystem.ThenewversionofLVTrans
doesnothavesomeofthetoolsdevelopedintheversionLVTrans8_1.1.2.
109
11.2 HydraulicSystemModelling
ThegeneralconfigurationoftheHydroPowerPlantmodelledinLVTranscontainstheessential
componentsofahydraulicpowersystemsuchasareservoir,anupstreamtunnel,asurgetank,
a penstock, a hydraulic turbine, speed governor and a tailrace through the water is released
backtotheriver.ThepowerplanthasaFrancisunit.Thesimplifiedsketchofatypicalhydraulic
power system is depicted in Figure 92. The accuracy of the dynamic analysis of the Hydro
PowerPlantmodelledinLVTransisatleast90%duetothedampingoftheseelementscanbe
adjusted.
Figure 111 shows the block diagram of the test system modelled in LVTrans. The elements
needed in the block diagram to model the sketch, shown in Figure 92, such as pipes, surge
shaft,turbineandPIDcontrol,aredraggedanddroppedfromFunctionPalette>>LVTransin
the block diagram. These elements must be wired after all the elements are inserted and
named.Eachelementmusthaveauniquename.Themainparametersofthehydrauliccircuit
aresummarizedinTable92.
Figure111:BlockDiagramoftheHydraulicSystemModel
Thedatamustbesetupforeachelementbeforethesimulationbegins.Thedataarestoredin
separatedatafiles.LVTranssetsinitialvaluesforeachelementafterrunningtheprogramfor
firsttime.Thesedatafilesareediteddirectlyineachelementwhiletheprogramisrunbypress
theindicatordatainthewindowofeachelementorinaplaintexteditor.
The input data used for each of the components of the test system model are described in
detailinChapter9.ThegivenparametersaresetaccordingtotheHydraulicGradeLine(HGL),
whichshowstheheightcorrespondingtoelevationandpressurehead.TheHGListheheightto
whichtheliquidwouldriseinapiezometertubeattachedtotheflow.[14]
The value considered for the pressure wave velocity a is 1200 m/s. The dimensionless pipe
friction factor is a function of the Reynolds number and the relative roughness height,
describedinChapter2.1.3.ThefrictionfactorisselectedfromtheMoodydiagram.
110
11.3 DynamicSimulation
Thefrequencyresponseanalysisaswellastransientresponseanalysistoevaluatethedynamic
performanceoftheHydraulicPowerSystemModel,depictedinFigure111,isimplementedin
the dynamic simulation tool LVTrans version LVTrans8_1.1.2 and version LVTrans86_1.3.1_T.
TheHydraulicPowerSystemModelissimulatedwithandwithoutdroopinLVTrans8_1.1.2.The
frequency response of the transfer function relating the water pressure to gate opening
positionimplementedinthedifferentversionsarecomparedandstudied.
Thefeedbackfromthegateopeningpositiontothecontrollerisincludedwhenthesimulation
isperformedwithdroop.Thedroopisusually4%to8%.
TheanalysisoftransientsintheHydraulicPowerSystemModelissubdividedinwaterhammer
responseandmassoscillationresponse.
The elastic pressure frequency describes the oscillation along the penstock. The
characteristic frequency does not vary with the load and depends only upon the
dimensionsofthepenstock.Thenaturalfrequencyistypicallyintheorderof1.00Hz.
The natural frequency resulting from the pendulum action between the reservoir and
thesurgeshaftcalledmassoscillations.Themassoscillationscantakeseveralminutes
togetthesteadystate.
Both the waterhammer and massoscillation responses are calculated by shutting down the
turbineintheinterfaceandblockdiagram.Thewaterhammerandmassoscillationsresponses
dependofthelengthoftheconduit.Theoscillationmodecouldbeanalyzedatthegraphofthe
PenstockandTurbineintheblockdiagram.
Figure112showstheoscillationsofthehydraulicpowerplantmodelledinLVTrans86_1.3.1_T.
1.02
Gate
1
0.98
0.96
0.94
0.92
100
105
110
115
120
125
130
135
140
145
150
130
135
140
145
150
Time (s)
Pressure Head
1.4
1.3
1.2
1.1
1
0.9
100
105
110
115
120
125
Time (s)
Figure112:SimulationResults:(a)gateopeningposition,(b)PressureHeadoftheHydraulicTest
ModelinLVTrans86_1.3.1_T
111
11.3.1 FrequencyResponseAnalysiscomputedinLVTrans8_1.1.2
The frequency response of the transfer function from gate position to pressure head of the
hydraulicpowerplantmodelledwithoutdroopinLVTrans8_1.1.2isshowninFigure113.
Head/Gate - Magnitude (dB)
20
0
-20
-40
-60
-80 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
Head/Gate - Phase(deg)
-100
-200
-300
-400 -3
10
Frequency [Hz]
Figure113:FrequencyResponseofthetransferfunctionfromgatepositiontopressureheadofthe
hydraulicpowerplantsimulatedwithoutdroopinLVTrans8_1.1.2
Figure114showsthemagnitudeandphaseofthefrequencyresponseofthehydraulicpower
plantmodelledwithdroopinLVTrans8_1.1.2.
20
0
-20
-40
-60
-80 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
hydraulicpowerplantsimulatedwithdroopinLVTrans8_1.1.2
112
Comparison between the magnitude and phase of the frequency response of the transfer
functions from gate position to pressure head of the hydraulic system model simulated with
andwithoutdroopcanbeexaminedinFigure115.
InFigure115,thedynamicperformanceofthebluesolidlinerepresentsthehydraulicsystem
model simulated without droop, and the green dashed line represents the hydraulic system
modelsimulatedwithdroop.
Thehydraulicsystemmodelwithoutdroopagreeswellwiththehydraulicsystemmodelwith
droop representation model. The magnitudes of both models reach a peak at a resonance
frequency around 0.01 Hz. This frequency corresponds to the natural frequency of the mass
oscillationbetweenthereservoirandthesurgetank.Themagnitudesofbothmodelsreacha
higher peak at a resonance frequency around 0.8 Hz. This frequency corresponds to the first
natural frequency of the penstock. The effects of the droop has very little different on the
phase.
Theoscillationsathigherfrequenciesareduetotheelasticityoftheconduitsystem.
20
0
without droop
with droop
-20
-40
-60
-80 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
hydraulicpowerplantsimulatedwithoutandwithdroopinLVTrans8_1.1.2
Thefeedbackfromthegateopeningpositiontothecontrollerisincludedwhenthesimulation
isperformedwithdroop.Thedroopisusually4%to8%.
113
11.3.2 FrequencyResponseAnalysiscomputedinLVTrans86_1.3.1_T
Themagnitudeandphaseofthefrequencyresponseofthetransferfunctionfromgateposition
toPressureHeadoftheHydraulicPowerSystemisplottedinFigure116.
Themagnitudeofthetransferfunctionreachesapeakvalueat0.01Hzthatcorrespondstothe
natural frequency of the massoscillation between the reservoir and the surge tank. The
theoretical value is given by Equation (9.9). The magnitude reaches a second peak value at
approximate 1 Hz. This value corresponds to the resonance frequency of the penstock which
theoreticalvalueisgivenbyEquation(9.8).
20
10
0
-10
-20
-30
-40
-50 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
hydraulicpowerplantsimulatedinLVTrans86_1.3.1_T
11.3.3 ComparisonoftheFrequencyResponseAnalysisoftheHydraulicPowerPlant
computedinLVTrans8_1.1.2andLVTrans86_1.3.1_T
Thefrequencyresponseofthehydraulicpowersystemmodelhasbeensimulatedandanalyzed
in the dynamic tool LVTrans version LVTrans8_1.1.2 and LVTrans86_1.3.1_T. The hydraulic
powersystemmodelincludestheessentialcomponentssuchasthereservoir,upstreamtunnel,
asurgetank,apenstock,aFrancisturbine,aPIDgovernorandtailrace.
In Figure 117 and Figure 118, the blue solid line represents the frequency response of the
hydraulic power generating system simulated in LVTrans8_1.1.2 and the green dashed line is
thehydraulicpowerplantsimulatedinLVTrans86_1.3.1_T.
The magnitude and phase of the relationship between the gate opening position and water
pressure head of the hydraulic power generating system simulated without droop in
LVTrans8_1.1.2 and computed in LVTrans86_1.3.1_T can be compared in Figure 117. The
magnitude of the hydraulic power generating system simulated in LVTrans8_1.1.2 is slightly
differenttothatcomputedinLVTrans86_1.3.1_T.
114
Head/Gate - Magnitude(dB)
20
0
-20
-40
-60
-80 -3
10
without droop
LVTrans86 1.3.1 T
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Figure117:ComparisonoftheFrequencyresponseofthehydraulicpowergeneratingsystem
simulatedwithoutdroopinLVTrans8_1.1.2andcomputedinLVTrans86_1.3.1_T
Comparisonofthemagnitudeandphaseoftherelationshipbetweenthewaterpressurehead
and gate opening position of the hydraulic power generating systemsimulated with droop in
LVTrans8_1.1.2 and computed in LVTrans86_1.3.1_T can be analyzed in Figure 118. The
magnitude of the hydraulic power generating system simulated in LVTrans8_1.1.2 is slightly
differenttothatcomputedinLVTrans86_1.3.1_T.
Head/Gate - Magnitude(dB)
20
0
-20
-40
-60
-80 -3
10
with droop
LVTrans86 1.3.1 T
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Figure118:ComparisonoftheFrequencyResponseofthehydraulicpowergeneratingsystem
simulatedwithdroopinLVTrans8_1.1.2andcomputedinLVTrans86_1.3.1_T
115
InFigure117andFigure118,themagnitudeofthetransferfunctionofthehydraulicpower
generating system simulated in LVTrans8_1.1.2 is a little different to that computed in
LVTrans86_1.3.1_T. The frequency response of the hydraulic power generating system
simulatedinLVTrans86_1.3.1_Thasmoredamping.
LVTrans8_1.1.2 is a version created specifically for the detailed grid/waterway interaction.

LVTrans8_1.1.2hasmoredetailsthanthenewestversionLVTrans86_1.3.1_T.Themodellingof
thegenerator,gridandPIDgovernorismoredetailedinLVTransversionLVTrans8_1.1.2.
The magnitude of the hydraulic power generating system simulated LVTrans86_1.3.1_T has a
lower value at the early frequency range but slowly increases. The magnitude of the model
simulated in LVTrans86_1.3.1_T reaches higher values than the magnitude of the model
simulatedinLVTrans8_1.1.2.
The magnitudes of both models reach a peak at a resonance frequency around 0.01 Hz. This
frequencycorrespondstothenaturalfrequencyofthemassoscillationbetweenthereservoir
and the surge tank. The magnitudes of both models reach a higher peak at a resonance
frequency around 0.8 Hz. This frequency corresponds to the first natural frequency of the
penstock.Theoscillationsathigherfrequenciesareduetotheelasticityoftheconduitsystem.
116
12
Discussion
The simulation system of a typical hydroelectric power plant has been implemented in the
software SIMPOW and the simulaton tool LVTrans. The hydroelectric power system has a
reservoir, an upstream tunnel, a long length penstock with an upstream surge tank and is
equippedwithaFrancisturbine.Inordertotesttheaccuracyandeffectivenessofthehydraulic
powersystem,somesimulationshavebeenconducted.
Thischapterpresentsthesummaryoftheresultsofthefrequencyresponseanalysisaswellas
transient response analysis of the models implemented inSIMPOW and LVTrans. The models
implementedinSIMPOWarecomparedandanalyzed.Afterthat,thesemodelsarecompared
withthemodelsdevelopedinLVTrans8_1.1.2.
12.1 SummaryofSimulationResultsofthemodelscomputedinSIMPOW
The hydraulic power generating system, which is developed in this work for the stability
analysis,hasbeenimplementedinSIMPOWwithadetailedsynchronousgeneratormodelwith
damping, a general speedgoverning model, and a hydraulic turbine. The hydraulic turbine is
modelledwithvaryingdegreesofdetail.Themathematicalmodelsofthehydraulicturbineare
detailedinTable94,andlistedbelow
recommendbyIEEE
In order to test the accuracy and effectiveness of the hydro power generating system,
frequencyresponseanalysisaswellastransientresponseanalysisisperformedtoevaluatethe
effectsofthedetailedmodellingoftheturbineandconduitsystemtothestabilityanalysisand
thedynamicperformance.
The frequency response analysis has been performed to the relationships of the mechanical
power, mechanical torque, and pressure head to gate position. The angular frequency of the
inputsinusoidalwasvariedbetween0.001and10Hz.
117
12.2 SummaryoftheEigenvalueanalysiscomputedinSIMPOW
ThedynamicperformanceofthemodelsimplementedinSIMPOWisinvestigatedutilizingthe
Eigenvalueanalysis.TheEigenvalueanalysisofthemodelsshowsthattheallthemodelshave
one oscillatory mode associated with the rotor angle and speed. This mode is called
electromechanical oscillatory mode and corresponds to the synchronous machine natural
frequencythatdependsonrotatinginertias.Table121showstheelectromechanicaloscillatory
modeofthemodelsoftheHydroPowerSystemimplementedinChapter10.
Table121:Summaryoftheelectromechanicaloscillatorymodeofthemodelsofthemodels
Models
1 SimplifiedNonlinearTurbineModel
2 Nonlinear Turbine Model without Surge Tank assuming
InelasticWaterColumn
3 Nonlinear Turbine Model without Surge Tank including
ElasticWaterColumnEffect
4 Nonlinear Turbine Model with Surge Tank assuming
5 Nonlinear Turbine Model with Surge Tank assuming
ElasticWaterColumninPenstock
6 Linear Turbine Model with Surge Tank considering
InelasticWaterColumns,andturbinecharacteristics
7 Linear Turbine Model with Surge Tank including Elastic
WaterColumninPenstock,andturbinecharacteristics
Eigenvalue
Damping Frequency
Ratio
[Hz]
0.4073 j 6.6214 0.06140 1.05383
0.4203 j 6.6906 0.06270
1.06485
0.5188 j 6.6521 0.07775
1.05872
0.5102 j 7.1079 0.07160
1.13126
0.5933 j 7.0806 0.08350
1.12692
0.2925 j 6.8421 0.04270
1.08896
0.4674 j 6.7198 0.06939
1.06950
The error in the frequency of oscillation between the calculated eigenvalues with KD=0.50,
Table98,andthecomputedeigenvaluesofthedifferentmodelsareshowninTable122.The
errorcalculatedbetweenthecomputedmodelsandthecalculatedeigenvaluesislessthatthe
10%inalltherepresentations.
Table122:Errorbetweenthecalculatedeigenvaluesandthecomputedeigenvalues
Model
Error,[%]
0.1450
NonlinearTurbineModelwithoutSurgeTank
0.9068
assumingInelasticWaterColumn
NonlinearTurbineModelwithoutSurgeTank
0.3193
includingElasticWaterColumnEffect
NonlinearTurbineModelwithSurgeTank
7.1986
assumingInelasticWaterColumns
NonlinearTurbineModelwithSurgeTank
6.7817
assumingElasticWaterColumninPenstock
LinearTurbineModelwithSurgeTankconsidering
3.1904
InelasticWaterColumns,andturbinecharacteristics
LinearTurbineModelwithSurgeTankincluding
1.3427
ElasticWaterColumninPenstock,andturbinecharacteristics
118
12.3 ComparisonoftheresultsofthemodelsinSIMPOW
Frequency response analysis is developed by Frequency Scanning in Linear Analysis in
DYNPOWinordertoidentifythetransferfunctionandotherparametersofthemodelsofthe
hydraulic turbine and conduit system, detailed in Table 94. To demonstrate the dynamic
nonlinear effects, a perturbation signal is applied to the gate opening position in open loop
conditions.TheopenloopconditionisobtainedbyassumingH=9999999forthesynchronous
machinemodel.Thatmeansthattheeffectsofthespeedintheblockdiagramcanbeomitted.
12.3.1 ComparisonofNonlinearTurbineModelswithoutsurgetank
Herein, frequency response analysis as well as transient response analysis is performed to
evaluate the effects to the stability analysis of a simple hydraulic turbine without surge tank
modelledby
thesimplifiednonlinearmodel(Model1),
thenonlinearmodelassuminginelasticwatercolumn(Model2)and
thenonlinearmodelincludingelasticwatercolumneffects(Model3)
In Figure 121 to Figure 124, the dynamic performance of the blue solid line represents the
simplifiednonlinearturbinemodel(Model1),thegreendashedlinerepresentsthenonlinear
modelassuminginelasticwatercolumn(Model2),andthereddashdottedlinerepresentsthe
nonlinearmodelincludingelasticwatercolumneffects(Model3).
Comparison between the dynamic behaviour of variables such as power angle, rotational
speed, mechanical torque, gate position, flow rate and head pressure of a hydraulic power
system model simulated by Model 1, 2 and 3 can be examined in Figure 121. The stability
analysis of the hydraulic power system model is tested by simulating a threephase fault to
groundatt=1s.Thisdisturbanceissimulatedforaperiodof5.0ms.
1.001
SPEED (pu)
ANGLE (deg)
70
68
66
64
62
0
0.9995
0.999
0
45.6
45.4
45.2
10
10
10
0.9025
0.902
0
10
1.005
0.904
HEAD (pu)
FLOW (pu)
0.903
0.906
0.902
0.9
0.898
0
0.9035
GATE (pu)
MT (MNm)
10
45.8
45
0
Model 1
Model 2
Model 3
1.0005
Time (s)
10
1
0.995
0.99
0
Time (s)
Figure121:FaultSimulationResultsofhydraulicturbinerepresentedbymodel1,2and3
119
It is seen from Figure 121, the angle and speed of the generator are almost the same for
models2and3,andisslightlydifferentformodel1.Itmeansthattheeffectofdifferenthydro
turbinemodelsonthepowersystemtransientstabilitysimulationiscomparativelysmall.The
variablessuchasmechanicaltorque,fluxrate,andpressureheadthatarecloselyrelatedtothe
conduitsystemhavealittledifferencebetweenmodel1,andmodels2and3.Thespeedofthe
generator and gate position are almost the same for the three models. The gate opening
positioninthehydraulicpowersystemmodelchangesslightly.
ThedynamicbehaviourofModel2andModel3areverysimilar.Itcanbeconcludedthatthe
simulationresultsofModel1isslightlydifferent.Thedifferencebetweenmodel1andmodel2
and3issmall.Thedifferencebetweenmodels1,andmodel2and3islittledifferentbecauseof
thesimplifiednonlinearmodelneglectsthehydrauliclosses.
Thestaticbehaviourisestablishedbytherelationshipbetweenthesteadystatevaluesofgate
position and turbine developed power. Figure 122 compares the gate positionturbine
developed power relationship of the hydraulic turbine model 1, 2 and 3. There is a good
agreementbetweenthemagnitudeofthetransferfunctionsofmodel2and3.Themagnitude
ofthetransferfunctionsofmodel1and,model2and3havealittledifferenceduetothehead
losses. It is observed a significant difference in the phase of model 1, 2 and 3 at very low
frequenciesupto0.002Hz.Thephaseisslightlydifferentfortherestofthebandwidth.
50
Model 1
Model 2
Model 3
0
-50
-100
-150
-200 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Figure122:FrequencyresponseofthetransferfunctionsfromGatepositiontoMechanicalPowerof
asimplehydraulicturbinerepresentedbyModel1,2and3
Figure123comparesthemagnitudeandphaseofthetransferfunctionrelatingthemechanical
torquetovariationsofthegatepositionofthehydraulicpowersystemrepresentedbymodel1,
2and3.Itisobservedthatmodel1,2and3haveaslightdifferencebetweeneachotheratlow
frequencies up to 0.7 Hz. However, significant difference in model 3 is observed at high
frequencies.Model3showsoscillationsduetotheelasticityoftheconduitsystem.Thenatural
frequenciescorrespondingtohighermodesoftheconduitsystemoscillationsarevisible.
120
7
6
Model 1
Model 2
Model 3
5
4
3
2
1 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Figure123:FrequencyresponseofthetransferfunctionfromGatepositiontoMechanicalTorqueofa
hydraulicturbinerepresentedbyModel1,2and3
Thefrequencyresponsesoftransferfunctionofheadpressureandgatepositionofahydraulic
turbine represented by model 1, 2 and 3 are compared in Figure 124. The magnitude and
phase of the three transfer functions are similar for frequencies up to 1.0 Hz. The transfer
functionfeaturesthefirstnaturalfrequencyofthepenstockwithahighamplitudeat0.7Hz.
Forhigherfrequencies,thetransferfunctionsofmodel1and2cannotrepresenttheconduit
dynamicswithanyacceptabledegreeofaccuracy.Thisisevidentwhenobservingthefrequency
responsebetween1and10Hz.
10
0
-10
-20
-30
-40
-50 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-50
-100
-150
-200
-250
-300 -3
10
Model 1
Model 2
Model 3
Frequency [Hz]
Figure124:FrequencyresponseofthetransferfunctionsoftheconduitsystemofModel1,2and3
121
12.3.2 ComparisonofTurbineModelswithSurgeTank
Thefollowingstudiesinvolvefrequencyresponseanalysisaswellastransientresponseanalysis
to evaluate the effects to the stability analysis of a simple hydraulic turbine with surge tank
modelledby
thenonlinearmodelassumingInelasticwatercolumnsinpenstockandtunnel(Model4)
thenonlinearmodelassumingelasticwatercolumninpenstock(Model5)
thelinearmodelassuminginelasticwatercolumnsinpenstockandtunnel(Model6)
thelinearmodelincludingelasticwatercolumneffectsinthepenstock(Model7)
In Figure 125 to Figure 128, the dynamic performance of the blue solid line represents the
model 4, the green dashed line represents the model 5, the red dotted line represents the
model6,andtheturquoisedashdottedlinerepresentsthemodel7.
Comparison between the dynamic behaviour of variables such as power angle, rotational
speed,mechanicaltorque,gateposition,flowrateandheadpressureofthehydraulicsystem
modelsimulatedbyModel4,5,6and7canbeexaminedinFigure125.Thestabilityanalysisof
thehydraulicpowergeneratingsystemistestedbysimulatingathreephasefaulttogroundat
t=1s.Thisdisturbanceissimulatedforaperiodof5.0ms.Thesamplingperiodis50s.
1.001
SPEED (pu)
ANGLE (deg)
70
65
60
55
0
0.9995
44
Model 4
Model 5
Model 6
Model 7
43
42
10
10
10
10
0.9025
0.902
0
1
HEAD (pu)
FLOW (pu)
0.903
0.92
0.9
0.88
0.86
0
0.9035
GATE (pu)
MT (MNm)
0.999
0
10
45
41
0
1.0005
Time (s)
10
0.5
0
-0.5
0
Time (s)
Figure125:FaultSimulationResultsofhydraulicturbinerepresentedbyModel4,5,6and7
ItisseenfromFigure125,theangleofthegeneratorarealmostthesameformodel4and5,
and is slightly different for the model 6 and 7. The speed of the generator is very similar for
model4,5,6and7;thismeansthattheeffectofdifferenthydroturbinemodelsonthepower
system transient stability simulation is comparatively small. The variables such as mechanical
torque,fluxrate,andpressureheadthatarecloselyrelatedtotheconduitsystemhavealittle
differencebetweennonlinearmodelsandlinearmodels.
Thechangesofspeedofthegeneratorandthegatepositionarealmostthesameforthefour
models.Thegateopeningpositioninthehydraulicpowersystemmodelchangesslightly.The
122
dynamicbehaviourofthelinearmodels,model6and7,areverysimilar.Itcanbeconcluded
thatthesimulationresultsofthenonlinearmodelsareslightlydifferentofthelinearmodels.
50
Model 4
Model 5
Model 6
Model 7
-50
-100
-150 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
0
-50
-100
-150
-200
-250
-300
-350
-400 -3
10
Frequency [Hz]
Figure126:FrequencyresponseofthetransferfunctionfromgatepositiontoMechanicalPowerof
Model4,5,6and7
Figure126comparesthegatepositionturbinedevelopedpowerrelationshipofthehydraulic
turbinemodel4,5,6and7.Themagnitudeofthetransferfunctionsofmodel4,5,6and7are
very similar in the entire bandwidth. Model 4, 5, 6 and 7 shows significant difference in the
phase.
10
Model 4
Model 5
Model 6
Model 7
-5
-10 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Figure127:FrequencyResponseofthetransferfunctionfromGatepositiontoMechanicalTorqueof
asimplehydraulicturbinerepresentedbyModel4,5,6and7
123
Figure127comparesthemagnitudeandphaseofthetransferfunctionrelatingthemechanical
torque to the gate position of the turbine represented bymodel 4, 5, 6 and 7. It is observed
that model 4, 5 and 7 have a slight difference at low frequencies up to 0.8 Hz. However,
significantdifferenceisobservedathighfrequencies.Thenaturalfrequenciescorrespondingto
highermodesoftheconduitsystemofmodel5and7areshown.
20
0
-20
-40
-60
-80 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-50
-100
-150
-200
-250
-300 -3
10
Model 4
Model 5
Model 6
Model 7
Frequency [Hz]
Figure128:FrequencyresponseofthetransferfunctionoftheconduitsystemofModel4,5,6and7
Thefrequencyresponseofthetransferfunctionofheadpressureandgatepositionofasimple
hydraulicturbinewithsurgetankrepresentedbymodel4,5,6and7arecomparedinFigure
128.Itcaneasilybenoticedthatthefrequencyresponsesoftheelasticmodels,model5and7,
aresimilareventhoughtheirparametricmodelsaresignificantlydifferent.Alargedifferencein
thephaseofmodel4and6areshown.
Thetransferfunctionsofmodel6and7featureanaturalfrequencywithahighamplitudeat
0.01Hz.Thisfrequencycorrespondstothenaturalfrequencyofthemassoscillationbetween
the reservoir and the surge tank. The calculated natural frequency of the massoscillation is
given by Equation (9.9). The transfer functions of model 4, 5 and 7 feature the natural
frequency of the penstock with a higher amplitude at 0.8 Hz. The natural frequency of the
penstockisdescribedinEquation(9.8).Forhigherfrequencies,thetransferfunctionofmodel4
and6cannotrepresenttheconduitdynamicswithanyacceptabledegreeofaccuracy.Thisis
evidentwhenobservingthefrequencyresponsebetween1and10Hz.
The simulation results of Model 4 and 5 do not show the natural frequency of the mass
oscillation between the reservoir and the surge tank. The implementation of Model 4 and 5
mustbecheckedinordertoobtainabetterfitwiththeresultsinLVTrans.
Model6and7shouldbestudiedindetail,varyingthevaluesoftheturbinecoefficients.These
coefficients,whichrepresentthenonlinearcharacteristicoftheturbine,havetobeextracted
fromtheHillCharts.
124
12.4 Comparison of the simulation results computed in SIMPOW and

LVTrans
Thefollowingstudiesinvolvefrequencyresponseanalysistoevaluatetheeffectstothestability
analysis of hydraulic power generating system implemented in SIMPOW and LVTrans. The
hydroturbineimplementedinSIMPOWismodelledby
thenonlinearmodelassumingInelasticwatercolumnsinpenstockandtunnel(Model4)
thenonlinearmodelassumingelasticwatercolumninpenstock(Model5)
thelinearmodelassuminginelasticwatercolumnsinpenstockandtunnel(Model6)
thelinearmodelincludingelasticwatercolumneffectsinthepenstock(Model7)
Thefrequencyresponseofthehydraulicpowersystemmodelhasbeensimulatedandanalyzed
with droop and without droop in the dynamic tool LVTrans version LVTrans8_1.1.2. The
feedback from the gate opening position to the controller is included when the simulation is
performedwithdroop.Thedroopisusually4%to8%.
InFigure129andFigure1210,thedynamicperformanceofthebluesolidlinerepresentsthe
hydropower generating system without droop and with droop computed in LVTrans,
respectively.Thegreendashedlinerepresentsthemodel4,thereddashdottedlinerepresents
the model 5, the turquoise dotted line represents the model 6, and the purple dashed line
representsthemodel7.
Figure129comparesthemagnitudeandphaseofthetransferfunctionofwaterpressureand
gateopeningofahydraulicpowersystemrepresentedbymodel4,5,6and7,inSIMPOW,and
thehydraulicpowersystemmodelwithoutdroopinLVTrans.Itisobservedthatbothmodelsin
SIMPOWandmodelin LVTranshasaslightlydifferenceinthemagnitudeatlowfrequencies.
However,significantdifferenceisobservedinthephase.
20
0
-20
-40
-60
-80
-100 -3
10
without droop
Model 4
Model 5
Model 6
Model 7
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
Frequency [Hz]
Figure129:FrequencyResponsesofthetransferfunctionoftheconduitsystemofHydraulicPower
SystemModelsimplementedinSIMPOWandLVTrans(withoutdroop)
125
Figure1210comparesthemagnitudeandphaseofthetransferfunctionofwaterpressureand
gateopeningofahydraulicpowersystemrepresentedbymodel4,5,6and7,inSIMPOW,and
model with droop implemented in LVTrans. It is observed that both models in SIMPOW and
model in LVTrans has a slightly difference in the magnitude at low frequencies. However,
significantdifferenceisobservedinthephase.Theoscillationsathighfrequencyofthemodel5
and7,andthemodelcomputedinLVTransareduetotheelasticityoftheconduitsystem.
20
0
-20
-40
-60
-80
-100 -3
10
10
-2
10
-2
10
-1
10
10
10
-1
10
10
-100
-200
-300
-400 -3
10
with droop
Model 4
Model 5
Model 6
Model 7
Frequency [Hz]
Figure1210:FrequencyResponsesofthetransferfunctionoftheconduitsystemofHydraulicPower
SystemModelsimplementedinSIMPOWandLVTrans(withdroop)
FromthemodelsimplementedinLVTrans,itcanbeconcludedthatthehydraulicsystemmodel
withoutdroopagreeswellwiththehydraulicsystemmodelwithdroop.Themagnitudeofboth
transfer functions shows two peaks at a resonance frequency of 0.01 Hz and 0.9 Hz. These
frequencies correspond to the natural frequencies of the massoscillation and penstock,
respectively.Theeffectofthedroophasverylittleinfluenceonthephase.Theoscillationsat
higherfrequenciesareduetotheelasticityoftheconduitsystem.
The transfer functions of model 6, 7 in SIMPOW, and models in LVTrans feature a natural
frequency with a high amplitude at 0.01 Hz. This frequency corresponds to the natural
frequency of the massoscillation between the reservoir and the surge tank. The transfer
functionsofmodel4,5and7inSIMPOW,andmodelsinLVTrans,allofthem,featureahigh
amplitudeat0.8Hzthatisthenaturalfrequencyofthepenstock.Forhigherfrequencies,the
transferfunctionsofmodel4and6cannotrepresenttheconduitdynamicswithanyacceptable
degreeofaccuracy.Thisisevidentwhenobservingthefrequencyresponsebetween1and10
Hz.
It is interesting to note that the magnitude of Model 7, which represents the most complete
modeluntilthisstage,andmodelinLVTransshowagoodfit.Model7featuresalowerpeakin
magnitudeat0.8Hz.Thiscouldbeduetotheinfluenceofthevaluesoftheturbinecoefficients
usedinthismodel.
126
13
Conclusions
13.1 Conclusions
This Masters Thesis work deals with the development of improved hydro turbine models for
the evaluation of a hydraulic power generating system performance in response to small
disturbancesinpowersystemanalysistool.Theseimprovedmodelsmustbeabletoreflectthe
possible interaction between the hydraulic system and power system in the computer
simulations of a power plant equipped with Francis turbines. Traditionally, mathematical
models for hydraulic power generating systems, normally found in relevant literature and
powersystemanalysistools,aresimplifiedmodels.Thisimpliesthatthesemodelsonlyreflect
partoftherealsituationsandassuchcouldhavealimitedapplication.
The dynamics of hydraulic turbines have a considerable influence on the dynamic stability of
the power system. The mathematical modelling of the flow and power output of a Francis
turbine, and its transient behaviour is based on the characteristic curve of the hydraulic
machinecalledHillCharts.Theturbineefficiencyforanyoperatingpointgivenbyrunnerspeed,
net head and gate position can be extracted from the Hill Charts. These values have to be
measuredpreciselyviafieldtestsortakenfrommodeltests.Althoughsuchcurvesareseldom
usedinspecificationsduetodifficultiesofmeasurement,thespeedtorquecharacteristicsofa
widevarietyofhydraulicturbineshavethegeneralformwiththetorquefallingoffroughlyin
proportiontospeedoverwiderangesofspeedandgateopeningposition.Theinfluenceofthe
turbinecoefficientsonthemodelaccuracyiscritical.
Thehydraulicpowergeneratingsystemmodel,whichisdevelopedinthisThesisworkforthe
stability analysis, has been designed with a thirdorder synchronous generator model, a
mechanicalhydraulicgoverningsystemmodel,andahydraulicturbine.Thehydraulicturbineis
modelledwithvaryingdegreesofdetail.Theeffectofthehydraulicpowergeneratingsystemis
studiedbymeansofanalysisofthedynamicbehaviourofthemodelsofthehydraulicmachine
andconduitsystemaboutthesteadystateoperatingconditionfollowingasmalldisturbance.
Themodelsofthehydraulicmachineandconduitsystemdevelopedaspartofthisstudyand
implementedinSIMPOW,arelistedbelow:
recommendbyIEEE
127
Frequencyresponseanalysisaswellastransientanalysisisperformedtoevaluatetheeffectsof
the detailed modelling of the turbine and conduit system to the stability analysis and the
dynamic performance of the hydraulic power generating system. The hydraulic power
generatingsystemhasalsobeenimplementedinthehydraulicsystemanalysistoolLVTrans.
Eigenvalueanalysis
ThedynamicperformanceofthedifferentHydroPowerSystemmodelsconnectedtoapower
system has been investigated utilizing Eigenvalue analysis. The theoretical eigenvalues are
determinedviaanalyticalapproachforaclassicalsimplifiedpowersystemrepresentedbythe
secondordersynchronousmachinemodelconnectedtoaninfinitebusbar.Eigenvaluesarealso
determined in Dynpow of the software SIMPOW. The error in the frequency of oscillation
between the theoretical eigenvalues with KD=0.50 and the computed eigenvalues of the
differentmodelsislessthatthe10%inalltherepresentations.
Timedomainanalysis
Thedynamicbehaviourofthehydraulicturbinerepresentedbymodel1,2and3aresimilar.It
canbeconcludedthatthesimulationresultsofmodel1isslightlydifferent.Itmeansthatthe
effectofdifferenthydroturbinemodelsonthepowersystemtransientstabilitysimulationis
comparativelysmall.Thevariablessuchasmechanicaltorque,fluxrateandpressureheadthat
arecloselyrelatedtotheconduitsystemhavealittledifferencebetweenmodel1,andmodel2
and 3. The changes in speed of the generator and gate position are almost the same for the
threemodels.
Thechangesinspeedofthegeneratorisverysimilarformodel4,5,6and7;thismeansthat
theeffectofdifferentmodelsofhydroturbinewithsurgetankonthepowersystemtransient
stabilitysimulationiscomparativelysmall.Thespeedofthegeneratorandthegatepositionare
almost the same for the four models. The gate opening position in the test system changes
slightly.Thedynamicbehaviourofthelinearmodels,model6and7,isverysimilar.Itcanbe
concludedthatthesimulationresultsofthenonlinearmodelsareslightlydifferentofthelinear
models.
FrequencyResponseAnalysis
Thereisagoodagreementbetweenthefrequencyresponseofthethreetransferfunctionsof
headpressureandgatepositionof ahydraulicturbinerepresentedbymodel1,2and3.The
transferfunctionfeaturesanaturalfrequencywithahighamplitudeat0.8Hz.Thisfrequency
correspondstothefirstnaturalfrequencyofthepenstock.Forhigherfrequencies,thetransfer
functionsofmodel1and2cannotrepresenttheconduitdynamicswithanyacceptabledegree
ofaccuracy.
Thefrequencyresponseofthetransferfunctionofheadpressureandgatepositionofasimple
hydraulicturbinewithsurgetankrepresentedbymodels4,5,6and7showdifferencesatlow
frequencies.Transferfunctionsofmodel4and5showsanexcellentfitatlowfrequenciesupto
1Hz.Thetransferfunctionsofmodel6and7featureanaturalfrequencywithahighamplitude
at 0.01 Hz. This frequency corresponds to the natural frequency of the massoscillation
betweenthereservoirandthesurgetank.Thetransferfunctionsofmodel4,5and7feature
thenaturalfrequencyofthepenstockwithahigheramplitudeat0.8Hz.Forhigherfrequencies,
the transfer function of model 4 and 6 cannot represent the conduit dynamics with any
acceptabledegreeofaccuracy.
128
The dynamic behaviour of the Hydraulic Power System has been analyzed in the program
LVTrans version LVTrans8_1.1.2 and version LVTrans86_1.3.1_T. LVTrans8_1.1.2 is a version
createdspecificallyforthedetailedgrid/waterwayinteraction.Thefrequencyresponseofthe
hydraulic power generating system simulated in LVTrans8_1.1.2 is slightly different to that
computed in LVTrans86_1.3.1_T. The frequency response of the hydraulic power generating
system simulated in LVTrans86_1.3.1_T has more damping. Based on the simulation results
fromthemodelsimplementedinLVTrans,itcanbeconcludedthatthehydraulicsystemmodel
withoutdroopagreeswellwiththehydraulicsystemmodelwithdroop.Themagnitudeofboth
models show two peaks at a resonance frequency of 0.01 Hz and 0.9 Hz. These frequencies
correspond to the natural frequencies of the massoscillation and penstock, respectively. The
effectofthedroophasverylittleinfluenceonthephase.Theoscillationsathigherfrequencies
areduetotheelasticityoftheconduitsystem.
The transfer functions of model 6, 7 in SIMPOW, and models in LVTrans feature a natural
frequency with a high amplitude at 0.01 Hz. This frequency corresponds to the natural
frequency of the massoscillation between the reservoir and the surge tank. The transfer
functionsofmodel4,5and7inSIMPOW,andmodelsinLVTrans,allofthem,featureahigh
amplitudeat0.8Hzthatisthenaturalfrequencyofthepenstock.Forhigherfrequencies,the
transferfunctionofmodel4and6cannotrepresenttheconduitdynamicswithanyacceptable
degreeofaccuracy.
The simulation results of Model 4 and 5 do not show the natural frequency of the mass
oscillation between the reservoir and the surge tank. The implementation of Model 4 and 5
mustbecheckedinordertoobtainabetterfitwiththeresultsinLVTrans.
Model6and7shouldbestudiedindetail,varyingthevaluesoftheturbinecoefficients.These
coefficients,whichrepresentthenonlinearcharacteristicoftheturbine,havetobeextracted
fromtheHillCharts.
It is interesting to note that the magnitude of Model 7, which represents the most complete
modeluntilthisstage,andmodelinLVTransshowagoodfit.Model7featuresalowerpeakin
magnitudeat0.8Hz.Thiscouldbeduetotheinfluenceofthevaluesoftheturbinecoefficients
usedinthismodel.
13.2 FurtherWork
TheimplementationofModel4and5mustbecheckedinordertoobtainabetterfitwiththe
results in LVTrans. The effects of the surge tank must be better represented in the model to
develop in SIMPOW. Additionally, these models could be improved, including the nonlinear
characteristicsoftheturbineextractedfromtheHillCharts.
Model6and7shouldbestudiedindetail,varyingthenonlinearcharacteristicsofthehydraulic
turbine, in order to draw more accurate conclusions about the dynamic behaviour of the
models. The influence of the turbine coefficients on the model accuracy is critical and these
valuesshouldbeextractedfromtheHillChartsandlinearizedaroundanoperatingpoint.
Additionally, the hydroelectric power plant system may be modelled by the Structure Matrix
Method,andbecomparedwiththemodelsimplementedinSIMPOWandLVTrans.
129
14
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
ReferenceBibliography
Hydraulic turbine and turbine control models for system dynamic studies. Power
Systems,IEEETransactionson,1992.7(1):p.167179.
Kundur, P., N.J. Balu, and M.G. Lauby, Power system stability and control. 1994, New
York;London:McGrawHill.xxiii,1176p.
Parmakian,J.,Waterhammeranalysis.1963,[S.l.]:Dover.161p.
Wood, D.J., Waterhammer AnalysisEssential and Easy (and Efficient). Journal of
EnvironmentalEngineering,2005.131(8):p.11231131.
Bergant,A.,A.R.Simpson,andA.S.Tijsseling,Waterhammerwithcolumnseparation:A
historicalreview.JournalofFluidsandStructures,2006.22(2):p.135171.
Chaudhry,M.H.,Appliedhydraulictransients.2nded.ed.1987,NewYork;Wokingham:
VanNostrandReinhold.xvi,521p.
BOULOS,etal.,Hydraulictransientguidelinesforprotectingwaterdistributionsystems.
Vol.97.2005,Denver,CO,ETATSUNIS:AmericanWaterWorksAssociation.14.
Almeida, A.B.d. and E. Koelle, Fluid transients in pipe networks. 1992, Southampton;
Boston, Mass.; London; New York: Computational Mechanics Publications ; Elsevier
AppliedScience.
Zipparro,V.J.,H.Hasen,andC.V.Davis,Davis'handbookofappliedhydraulics.4thed.
ed.1993,NewYork:McGrawHill.
(TNSHP), T.N.o.S.h., Guide on How to Develop a Small Hydropower Plant, E.S.H.A..
ESHA,Editor.2004.
Konidaris, D.N. and J.A. Tegopoulos, Investigation of oscillatory problems of hydraulic
generating units equipped with Francis turbines. Energy Conversion, IEEE Transactions
on,1997.12(4):p.419425.
Oldenburger, R. and J. Donelson, Dynamic Response of a Hydroelectric Plant. Power
Apparatus and Systems, Part III. Transactions of the American Institute of Electrical
Engineers,1962.81(3):p.403418.
Romeo, E., C. Royo, and A. Monzn, Improved explicit equations for estimation of the
frictionfactorinroughandsmoothpipes.ChemicalEngineeringJournal,2002.86(3):p.
369374.
White, F.M., Fluid mechanics. 6th ed. ed. 2008, New York ; London: McGrawHill. xiii,
864p.
Souza, O.H., Jr., N. Barbieri, and A.H.M. Santos, Study of hydraulic transients in
hydropower plants through simulation of nonlinear model of penstock and hydraulic
turbinemodel.PowerSystems,IEEETransactionson,1999.14(4):p.12691272.
Brekke, H., A Stability Study on Hydro Power Plant Governing Including the influence
fromaQuasinonlineardampingofoscillatoryflowandfromtheturbinecharacteristics.
1984,NorgesTekniskeHgskole:Trondheim.
Bakken,B.H.,Technicalandeconomicaspectsofoperationofthermalandhydropower
systems.1997,Trondheim:NorwegianUniversityofScienceandTechnology.222.
Hongqing, F., et al., Basic Modeling and Simulation Tool for Analysis of Hydraulic
TransientsinHydroelectricPowerPlants.EnergyConversion,IEEETransactionson,2008.
23(3):p.834841.
130
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
Kishor,N.,S.P.Singh,andA.S.Raghuvanshi,Dynamicsimulationsofhydroturbineand
itsstateestimationbasedLQcontrol.EnergyConversionandManagement,2006.47(18
19):p.31193137.
Nicolet, C., F. Avellan, and J.J. Simond, Hydroacoustic modelling and numerical
simulationofunsteadyoperationofhydroelectricsystems,inLMH,LME.2007,EPFL.
IEEE Guide for the Application of Turbine Governing Systems for Hydroelectric
GeneratingUnits.IEEEStd12072004,2004:p.0_1121.
Machowski,J.,J.Bialek,andD.J.Bumby,PowerSystemDynamics:StabilityandControl.
2nded.2008:JohnWiley&Sons,Ltd..
Voros, N.G., C.T. Kiranoudis, and Z.B. Maroulis, Shortcut design of small hydroelectric
plants.RenewableEnergy,2000.19(4):p.545563.
Undrill, J.M. and J.L. Woodward, Nonlinear Hydro Governing Model and Improved
Calculation for Determining Temporary Droop. Power Apparatus and Systems, IEEE
Transactionson,1967.PAS86(4):p.443453.
Haiyan,B.,Y.Jiandong,andF.Liang.StudyonNonlinearDynamicalModelandControl
StrategyofTransientProcessinHydropowerStationwithFrancisTurbine.inPowerand
EnergyEngineeringConference,2009.APPEEC2009.AsiaPacific.2009.
FraileArdanuy,J.,etal.SpeedOptimisationModuleofaHydraulicFrancisturbinebased
on Artificial Neural Networks. Application to the Dynamic Analysis and Control of an
AdjustableSpeedHydroPlant.inNeuralNetworks,2006.IJCNN'06.InternationalJoint
Conferenceon.2006.
Prez, J.I., J.R. Wilhelmi, and L. Maroto, Adjustable speed operation of a hydropower
plant associated to an irrigationreservoir. Energy Conversion and Management, 2008.
49(11):p.29732978.
De Jaeger, E., et al., Hydro turbine model for system dynamic studies. Power Systems,
IEEETransactionson,1994.9(4):p.17091715.
Zhihuai, X., et al. Identifying of Hydraulic Turbine Generating Unit Model Based on
Neural Network. in Intelligent Systems Design and Applications, 2006. ISDA '06. Sixth
InternationalConferenceon.2006.
Brekke, H. and X.X. Li. A new approach to the mathematic modelling of hydropower
governingsystems.inControl,1988.CONTROL88.,InternationalConferenceon.1988.
Nicolet,C.,etal.,NewToolfortheSimulationofTransientPhenomenainFrancisTurbine
PowerPlants.,inIAHRSymposium.2002,:.
Jiang, J., Design of an optimal robust governor for hydraulic turbine generating units.
EnergyConversion,IEEETransactionson,1995.10(1):p.188194.
Hongqing, F. and S. Zuyi. Modeling and Simulation of Hydraulic Transients for
Hydropower Plants. in Transmission and Distribution Conference and Exhibition: Asia
andPacific,2005IEEE/PES.2005.
Thorne, D.H. and E.F. Hill, Field Testing and Simulation of Hydraulic Turbine Governor
Performance.PowerApparatusandSystems,IEEETransactionson,1974.PAS93(4):p.
11831191.
Izena,A.,etal.Practicalhydraulicturbinemodel.inPowerEngineeringSocietyGeneral
Meeting,2006.IEEE.2006.
Jadid, S. and A. Salami. Accurate model of hydroelectric power plant for load pickup
during power system restoration. in TENCON 2004. 2004 IEEE Region 10 Conference.
2004.
Strah,B.,O.Kuljaca,andZ.Vukic,Speedandactivepowercontrolofhydroturbineunit.
131
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
Zhijun, L. and C. Lijuan. The building and analyzing of the fifthorder model of the
synchronous generator in standalone infinite system. in Electrical Machines and
Systems,2008.ICEMS2008.InternationalConferenceon.2008.
IEEE Guide for Synchronous Generator Modeling Practices and Applications in Power
SystemStabilityAnalyses.IEEEStd11102002(RevisionofIEEEStd11101991),2003:p.
0_172.
Kanjun, Z., et al. Simulation Analysis of Dynamic Performance for HydroGenerator
UnderLossofExcitationCondition.inUniversitiesPowerEngineeringConference,2006.
UPEC'06.Proceedingsofthe41stInternational.2006.
Mouni,E.,S.Tnani,andG.Champenois.Comparativestudyofthreemodellingmethods
of synchronous generator. in IEEE Industrial Electronics, IECON 2006 32nd Annual
Conferenceon.2006.
Yongji,C.andY.Jilai.Alongtermdynamicsimulationmethodofpowersystembasedon
variables grouping and seceding grouping strategy. in Electric Utility Deregulation and
RestructuringandPowerTechnologies,2008.DRPT2008.ThirdInternationalConference
on.2008.
Saadat, H., Power system analysis. 2nd ed. ed. 2002, Boston ; London: McGrawHill
PrimisCustomPublishing.xix,712p.
Kazemi,A.,M.R.J.Motlagh,andA.H.Naghshbandy,Applicationofanewmultivariable
feedback linearization method for improvement of power systems transient stability.
InternationalJournalofElectricalPower&EnergySystems,2007.29(4):p.322328.
Xianshan, L., et al. The Effect of Hydro Turbines and Governors on Power System Low
Frequency Oscillations. in Power System Technology, 2006. PowerCon 2006.
InternationalConferenceon.2006.
Murty,M.S.R.andM.V.Hariharan,AnalysisandImprovementoftheStabilityofaHydro
Turbine Generating Unit With Long Penstock. Power Apparatus and Systems, IEEE
Transactionson,1984.PAS103(2):p.360367.
Ramey, D.G. and J.W. Skooglund, Detailed Hydrogovernor Representation for System
StabilityStudies.PowerApparatusandSystems,IEEETransactionson,1970.PAS89(1):
p.106112.
Report, I.C., Dynamic Models for Steam and Hydro Turbines in Power System Studies.
PowerApparatusandSystems,IEEETransactionson,1973.PAS92(6):p.19041915.
Culberg,J.L.,M.Negnevitsky,andK.M.Muttaqi.Hydroturbinegovernorcontrol:theory,
techniquesandlimitations.inAUPEC'06.2006.Melbourne.
Kishor,N.,R.P.Saini,andS.P.Singh,Areviewonhydropowerplantmodelsandcontrol.
RenewableandSustainableEnergyReviews,2007.11(5):p.776796.
Yin Chin, C., K.M. Muttaqi, and M. Negnevitsky. Modelling of hydraulic turbine for
dynamic studies and performance analysis. in Power Engineering Conference, 2007.
AUPEC2007.AustralasianUniversities.2007.
Sanathanan, C.K., Accurate Low Order Model for Hydraulic TurbinePenstock. Energy
Conversion,IEEETransactionson,1987.EC2(2):p.196200.
Hagihara,S.,etal.,StabilityofaHydraulicTurbineGeneratingUnitControlledbyP.I.D.
Governor. Power Apparatus and Systems, IEEE Transactions on, 1979. PAS98(6): p.
22942298.
Smith,J.R.,R.McLean,andJ.F.Robbie,Assessmentofhydroturbinemodelsforpower
systems studies. Generation, Transmission and Distribution, IEE Proceedings C, 1983.
130(1):p.17.
132
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
Kishor,N.,R.P.Saini,andS.P.Singh.Simulationofreducedorderhydroturbinemodelsto
study its hydraulic transient characteristics. in9th International Multitopic Conference,
IEEEINMIC2005.2005.
Nicolet, C., et al., HighOrder Modeling of Hydraulic Power Plant in Islanded Power
Network.PowerSystems,IEEETransactionson,2007.22(4):p.18701880.
Vournas, C.D. and N. Daskalakis. Governor tuning and regulating capacity of
hydroelectric units. in WESCANEX 93. 'Communications, Computers and Power in the
ModernEnvironment.'ConferenceProceedings.,IEEE.1993.
Nicolet,C.,etal.Onthehydroelectricstabilityofanislandedpowernetwork.inPower
EngineeringSocietyGeneralMeeting,2006.IEEE.2006.
Wozniak, L. and T.L. Filbert, Speed loop cancellation governor for hydrogenerators. I.
Development.EnergyConversion,IEEETransactionson,1988.3(1):p.8590.
Pannatier,Y.,etal.,DynamicBehaviorofa2VariableSpeedPumpTurbinePowerPlant,
inICEM'08XVIII18thInternationalConferenceonElectricalMachines.2008,:Vilamoura,
Algarve,Portugal.
Aguero, J.L., et al. Hydraulic transients in hydropower plant impact on power system
dynamic stability. in Power and Energy Society General Meeting Conversion and
DeliveryofElectricalEnergyinthe21stCentury,2008IEEE.2008.
Huimin, G. and W. Chao. Effect of Detailed Hydro Turbine Models on Power System
Analysis. in Power Systems Conference and Exposition, 2006. PSCE '06. 2006 IEEE PES.
2006.
Hannett, L.N., J.W. Feltes, and B. Fardanesh, Field tests to validate hydro turbine
governormodelstructureandparameters.PowerSystems,IEEETransactionson,1994.
9(4):p.17441751.
Vournas,C.D.,Secondorderhydraulicturbinemodelsformultimachinestabilitystudies.
Slootweg, J.G., et al., A Study of the Eigenvalue Analysis Capabilities of Power System
Dynamics Simulation Software, in 14th Power Systems Computation Conference,
PSCC'02.2002:Sevilla,Spain.p.8.
Persson,J.,UsingLinearAnalysistofindEigenvaluesandEigenvectorsinPowerSystems.
2002:Madrid,Spain.
Rogers, G., Demystifying power system oscillations. Computer Applications in Power,
IEEE,1996.9(3):p.3035.
Svingen, B., Manual LVTrans for versjon 8_1.2.4. 2007, SINTEF Energy Research:
Trondheim.p.66.
133

THESIS NTNU MST HYDROTURBINE AND GOVERNOR MODELLING 2010 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

THESIS NTNU MST HYDROTURBINE AND GOVERNOR MODELLING 2010 PDF

Uploaded by

Copyright:

Available Formats

Hydro Turbine and Governor Modelling

Electric - Hydraulic Interaction

Luz Alexandra Lucero Tenorio

Master of Science in Electric Power Engineering

Norwegian University of Science and Technology

Assignment given: 25. January 2010

Figure 1012: Frequency Response of the transfer function of the mechanicalhydraulic

Forturbulentflow( Re 4000 ),thefrictionfactorfisafunctionoftheReynoldsnumberReand

TR 5.00 TW 1.00 0.50 TW

Hydroelectric Power Plant Modelling by

A matrix representation of the group of elements interconnected may be established by

7.2 Matrix representations of the Basic Elements in Hydro Power

Dividingthroughby P0 gH00Q0 ,Equation(7.13)yields

In practice, damping of rotor swings is considered to be satisfactory if the damping ratio

Thetransientsynchronizingpowercoefficient KE' bydefinitionis:

where xd X d X and X X T X S isthecombinedreactanceofthestepuptransformerand

Thus,theundampednaturalfrequency nat ofrotorswingsforsmalloscillationsisgivenby

2 2Hnat 2 K E ' 2Hs

1,2 nat jnat 1 2

Theresponseofalineartimeinvariantsystemtosinusoidalinput r t A sin t isgivenby

A hydroelectric development includes in some form a waterdiverting structure, conduit to

E' Vg jX d' I0 1.0036.01 j0.3 0.9517.58 1.122949.9234

Thetransientsynchronizingpowercoefficient KE' is:

0.00006 0.03267 0.03267 0.99570 0.07283 0.07283 0.00203 0.00000 0.00185

GATE OPENING (pu)

MECHANIC TORQUE (MNm)

PRESSURE HEAD (pu)

FLOW RATE (pu)

MW/GATE - Magnitude (dB)

MT/GATE - Magnitude (dB)

HEAD/GATE - Magnitude (dB)

TETA/GATE - Magnitude (dB)

10.2 MODEL 2: Nonlinear Turbine Model without Surge Tank assuming

GATE OPENING (pu)

MECHANIC TORQUE (MNm)

PRESSURE HEAD (pu)

FLOW RATE (pu)

MW/GATE - Magnitude (dB)

MT/GATE - Magnitude (dB)

HEAD/GATE - Magnitude (dB)

TETA/GATE - Magnitude (dB)

10.3 MODEL 3: Nonlinear Turbine Model without Surge Tank including

GATE OPENING (pu)

MECHANIC TORQUE (MNm)

PRESSURE HEAD (pu)

FLOW RATE (pu)

MW/GATE - Magnitude (dB)

MT/GATE - Magnitude (dB)

HEAD/GATE - Magnitude (dB)

TETA/GATE - Magnitude (dB)

10.4 MODEL 4: Nonlinear Turbine Model with Surge Tank assuming

GATE OPENING (pu)

MECHANIC TORQUE (MNm)

PRESSURE HEAD (pu)

FLOW RATE (pu)

MW/GATE - Magnitude (dB)

MT/GATE - Magnitude (dB)

HEAD/GATE - Magnitude (dB)

TETA/GATE - Magnitude (dB)

GATE OPENING (pu)

MECHANIC TORQUE (MNm)

PRESSURE HEAD (pu)

FLOW RATE (pu)