Hydro Turbine and Governor Modelling: Luz Alexandra Lucero Tenorio

June 2010
Kjetil Uhlen, ELKRAFT

Trond Toftevaag, SINTEF Energy Research
Master of Science in Electric Power Engineering
Submission date:
Supervisor:
Co-supervisor:
Norwegian University of Science and Technology
Department of Electric Power Engineering
Hydro Turbine and Governor Modelling
Electric - Hydraulic Interaction
Luz Alexandra Lucero Tenorio
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
ii
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
models of varying degrees of detail. Firstly, nonlinear models for a simple turbine without surge
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
the most accurate models for any acceptable study of the interaction between hydraulic system
andpowersystem.
The study of dynamic performance and interaction between the hydraulic system and power
system of these extended linearand nonlinear models including the elastic water hammereffect
andvaryingthenonlinearcharacteristicsofthehydraulicturbinemustbestudiedindetail.

iii
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
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
4 SYNCHRONOUSMACHINE.............................................................................................................................26
4.1SYNCHRONOUSGENERATOR..........................................................................................................................................26
4.2SYNCHRONOUSGENERATOREQUATIONS.........................................................................................................................27
4.3SYNCHRONOUSGENERATORMODELS..............................................................................................................................28
4.3.1SynchronousMachineRepresentedbytheClassicalModel...........................................................................29
5 TURBINEGOVERNINGSYSTEMS.....................................................................................................................30
5.1GOVERNINGSYSTEM....................................................................................................................................................30
5.2MECHANICALHYDRAULICGOVERNOR.............................................................................................................................31
5.2.1MathematicalModelling................................................................................................................................31
5.3ELECTROHYDRAULICGOVERNINGSYSTEM.......................................................................................................................34
5.3.1MathematicalModel......................................................................................................................................34
iv
6 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
7 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
9 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
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

vi
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
Figure 66: Nonlinear Model with surge tank assuming elastic water column in penstock and
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
Figure 107: Fault simulation Results: (a) Power angle, (b) Speed, (c) Mechanical Torque, (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
vii
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
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
viii
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
Figure 122: Frequency response of the transfer functions from Gate position to Mechanical
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

ix
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

Preface
In January 2005, an internal project called Turbine and hydropower modelling was started at
SINTEFEnergyResearch.Thatprojectwasconcernedaboutthestudyofdynamicperformanceand
possible interaction between the hydraulic system and power system of a power plant equipped
with Francis turbines. 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
therealsituationsandassuchcouldhavealimitedapplication.
The Masters Thesis work develops improved 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.
Appropriate representations of the hydraulic turbine and water conduit system are developed in
models of varying degrees of detail. Firstly, nonlinear models for a simple turbine without surge
tankconsideringtheinelasticandelastictravellingwaveeffectshavebeendeveloped.Afterthat,
nonlinear models considering the inelastic and elastic travelling wave effects for a 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 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
xi
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,
Kjetil Uhlen, and Lars Lindquist, STRI Sweden, who have assisted me in the dynamic study of the
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,June15
th
2010

xii
List of Symbols
Notation
Uppercasesymbolsnormallydenotephysicalvalues.
Lowercasesymbolsnormallydenoteperunitvalues.
Thesuffix0denotesaninitialsteadystatevalue.
TheprefixAdenotessmalldeviations.
Symbols
a Pressurewavevelocity,[m/s]
A Penstockcrosssectionarea,[m
2
]
A
s
SurgeTankcrosssectionarea,[m
2
]
A
t
constantproportionalityfactor,[]
D Pipeinternaldiameter,[m]
e Pipewallthickness,[m]
E Youngsmodulusofelasticity,[N/m
2
]
f DarcyWeisbachfrictionfactor,[]
g Accelerationduetogravity,[m/s
2
]
H(x,t) Piezometrichead(HGL)orwaterfreesurfacelevel,[m]
H InertiaConstant,[MWs/MVA]
H
f
Headlossesduetofrictioneffectsintheconduit,[m]
i
d
Currentflowinginthedaxisarmaturecoil,[pu]
i
q
Currentflowingintheqaxisarmaturecoil,[pu]
K Bulkmodulusofelasticity,[N/m
2
]
K
D
Dampingtorquecoefficient,[]
K
d
Derivativegain,[s]
K
E
Transientsynchronizingpowercoefficient,[]
K
i
Integralgain,[s
1
]
K
p
Proportionalgain,[pu]
K
s
Servomotorgain,[]
L Lengthoftheconduit,[m]
L
ad
,L
aq
Mutualinductancebetweenthestatorandrotorwindings
L
d
SelfInductanceofthedaxisarmaturewindings
L
q
SelfInductanceoftheqaxisarmaturewindings
M
11
HydraulicturbineUnittorque
M
e
Electromagnetictorque,[Nm]
M
m
Mechanicaltorque,[Nm]
p(x,t) pressure,[Pa]
P
m
Mechanicalpowerontheturbineshaft,[MW]
Q(x,t) Flowdischarge,[m
3
/s]
Q
11
Hydraulicturbineunitflowrate
q
NL
waterflowatzeroelectricalpower,[]
r Piperadius,[m]
R
a
Statorresistance,[pu]
R
e
Reynoldsnumber,[]
R
P
Permanentspeeddroop,[pu]
R
T
Temporaryspeeddroop,[pu]
s Laplaceoperator
xiii
S
N
Ratedapparentpower,[MVA]
t Time,[s]
T
c
Gateclosingtimeconstant,[s]
T
d0
,T
d0
opencircuitdaxistransientandsubtransienttimeconstant,[s]
T
e
Wavetraveltime,[s]
T
g
MainServomotortimeconstant[s]
T
M
MechanicalstartingtimeT
M
=2H,[s]
T
P
Pilotvalveandservomotortimeconstant,[s]
T
q0
Opencircuitqaxissubtransienttimeconstant,[s]
T
R
Resettimeordashpottimeconstant,[s]
T
s
SurgeTankFillingTime,[s]
T
w
Watertimeconstantorwaterstartingtime,[s]
V(x,t) averageVelocity,[m/s]
v Kinematicsviscosityofthefluid,[m
2
/s]
,
d q
v v Voltageacrossthefictitiousdandqaxisarmaturecoils,[pu]
V
g
Voltageatthegeneratorterminals
V
S
Infinitebusbarvoltage
x Distancemeasuredalongthepipeaxis,[m]
x
l
Statorleakagereactance,[pu]
X
d
,X
d
,X
d
directaxissynchronous,transientandsubtransientreactance,[pu]
x
d
,x
d
,x
d
Total directaxis synchronous, transient and subtransient reactance between
(andincluding)ageneratorandaninfinitebusbar,[pu]
X
q
,X
q
,X
q
quadratureaxissynchronous,transientandsubtransientreactance,[pu]
x
s
Internalreactanceofaninfinitebusbar,[pu]
x
T
Seriesreactanceofatransformer,[pu]
y Turbinewicketgateposition,[]
z Elevationofthepipecentreline,[m]

Slopeofthepipeaxis,[]
o Poweranglewithrespecttoaninfinitebusbar
Internalpiperoughness,[m]
,
Dampingratio,[]
q Turbineefficiency
Eigenvalue
Fluidmassdensity,[kg/m
3
]
t
0
Shearforcebetweenthefluidandtheconduitwalls,[N/m
2
]
| Runnerbladeangle
D
,
Q
Totalfluxlinkagesofdamperwindingsindaxisandqaxis
d
,
q
Totaldaxisandqaxisfluxlinkages
f
Totalfluxlinkageofthefieldwinding
e
d
Dampednaturalfrequencyofrotorswings,[rad/s]
e
N
Normalizedvelocityoftheunit,[pu]
e
nat
Undampednaturalfrequencyofrotorswingsforsmalloscillations,[rad/s]
e angularvelocityofthegenerator,[electricalradians]
e
s
Synchronousangularvelocity(equalto 2 f t ),[electricalradians]
Conduitwalltensilestrain
1 Introduction
1.1 Backgroundinformation
InJanuary2005,aninternalprojectcalledTurbineandhydropowermodellingwasstartedat
SINTEF Energy Research. This project was concerned aboutthe study of dynamic performance
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
interaction between hydraulic system and power system. This implies that these models only
reflectpartoftherealsituationsandassuchcouldhavealimitedapplication.
1.2 Objective
ThepurposeofthisMastersThesisworkistodevelopimprovedhydraulicturbinemodelsofa
typicalhydraulicpowergeneratingsystemfortheirrepresentationindynamicstudiesofpower
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
simulationsofapowerplantequippedwithFrancisturbines.
1.3 Scope
TheprecisionoftherepresentationofaHydraulicPowerGeneratingSystemindynamicstudies
isexaminedbymeansofanalysisofthedynamicbehaviourofdifferentmodelsofthehydraulic
machine and water conduit system. The stability study of different models for Synchronous
MachinesandTurbineGoverningSystemarebeyondofthescopeofthisMastersThesiswork.
Appropriate representations of the hydraulic turbine and waterconduit system aredeveloped

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 wave effects for a
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
linearized at an operating point considering both the nonlinear turbine characteristics and the
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.
2 PressureWaterConduitSystem
Dependingon the site,the conduitsystem is composed of a waterdiverting structure; acanal

to carry the water flow to the reservoir; a surge tank; a penstock pipe to convey the water to
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.
Hydroelectric power plants with long conduits may have severe water hammer and governing
stabilityproblems.Theclassicalsolutionistoinsertasurgetankupstreamor/anddownstream
thepowerplanttominimizethehydraulictransienteffects.
This chapter contains the mathematical deduction of the equations describing the transient
state flows in the surge tank and closedconduit systems including travelling wave effects and
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
Hydraulic transient refers to the pressure fluctuations in the water caused during a change in
meanflowconditions.Themaincomponentsofthedisturbancesarepressureandflowchanges
at a rated point that causes propagation of pressure waves throughout the system. The
pressure waves travel with the velocity of sound, which depends on the characteristics of the
conduitsystemandonthewaterelasticity.Whenaclosedconduitisfilledwithmovingwater,
the law governing the changes of pressure and discharge depends upon the conditions under
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
The pressure wave velocity a depends upon the characteristics of the liquid, such as the Bulk
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]
1
K
a
DK
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
e
L
T
a
=

(2.3)
Considering a pipeline in which the flow is initially steady. A sudden rise in the pipe adjacent
gatewillbeproducedwhenaninstantaneousgateclosuretakesplaceatthelowerendofthe
pipe.Thewavetransmissionphenomenonduetothegateclosureisanalyzedthoroughlyinthe
textbookWaterhammerAnalysis.[3]
The theoretical period T

th
for a conduit having constant diameter, constant wall thickness is
givenby
4
th
L
T
a
=

(2.4)

Thesequenceofeventsfollowingthevalveclosureisexplainedbrieflybelow.
Table21:Propagationofpressurewavescausedbyinstantaneousclosureofvalve
Stage Events
0 t L a < s Theflowvelocityatthevalveisreducedtozero,
Thepipeisexpanded,
Thefluiddensityisincreased
2 L a t L a < s Thefluidstartstoflowfromthepipelineintothereservoir.
Anegativewavetravelstowardthevalve.
2 3 L a t L a < s
The velocity is instantaneously changed due to the valve is

completelyclosed.
Thepressureisreducedandthenegativewavepropagatesin
theupstreamdirection.
3 4 L a t L a < s
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 H
f
in a given conduit for a given discharge is usually determined by the
DarcyWeisbachequation.
2
2
f
L V
H f
D g
=

(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
e
VD
R
v
=

(2.6)
The friction factor f in a laminar regime is independent of the wall roughness and inversely
proportional to the Reynolds number. The friction factor for a laminar flow is calculated from
theHagenPoiseuilleequationas
64
e
f
R
=

(2.7)
Forturbulentflow( 4000
e
R > ),thefrictionfactorfisafunctionoftheReynoldsnumberR
e
and
the relative roughness height ( / D c ). ColebrookWhite equation relates the friction factor to
theReynoldsnumberandrelativeroughnessas
10
1 2.51
2log
3.7
e
D
f R f
c
| |
= +
|
|
\ .
(2.8)
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. Any change in the flow condition is assumed to take place instantaneously throughout
thesysteminalumpedsystem.
Mathematically,thetransientsinthedistributedsystemsarerepresentedbypartialdifferential
equations,whereasthetransientsinthelumpedsystemsaredescribedbyordinarydifferential
equations.Thesystemmaybeanalyzedasalumpedsystemwhen L a e 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
tube elasticity, is based in the conservation mass law for a control volume. The radial velocity
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
referred elevations z and slope of the pipe axis. The slope is considered positive for a rising
pipe in the positive direction of x, as shown in Figure 21. The pipe cross sectional area A and
theliquidmassdensityarebothfunctionofpressurep(x,t).[6,89,1516]
A
V
(
)
A
V
A
V
x
x
c
+
A
c
Figure21:Controlvolumeforthederivationoftheequationofcontinuity
Theequationofcontinuityofacontrolvolume
c
ataninstantt,showninFigure21,yields
( ) AV AV AV x
t

c
= + A
c

(2.9)
ExpandingEquation(2.9):
( ) ( )
A V A
A AV A A V AV
t x t t x x x

c c c c c c c
+ = + + + +
c c c c c c c

(2.10)
Rearranging terms, using expressions for the total derivatives and dividing throughout by A ,
Equation(2.9)maybewrittenas:
1 1
0
dA d V
A dt dt x
c
+ + =
c

(2.11)
Foracircularconduithavingradiusr,
2
dA dr
r
dt dt
t =
(2.12)
Intermsoftheconduitwalltensilestraine,Equation(2.12)canbewrittenas
2 2
dA dr d
r A
dt dt dt
t
e
= =

(2.13)
Assumingthattheconduitwallsarealinearlyelastic,then
1
2
2
d D dp
pD
dt e dt
E
e
e
=
(2.14)
ThefirsttermofEquation(2.11)referstochangeinconduitarewithtimebecomes
1 1
2
dA D dp
pD
A dt e dt
E
e
=

(2.15)
The second term of Equation (2.11) refers to the rate of change of fluid density. The bulk
modulusofelasticityofafluidisbydefinition
( ) ( ) / /
dp dp
K
d d
= =

(2.16)
andthesecondtermofEquation(2.11)becomes
1 1 d dp
dt K dt
=

(2.17)
SubstitutingEquation(2.15)and(2.17)intoEquation(2.11)gives
1 1
0
2
dp V
eE p
K dt x
D
| |
|
c
+ + =
|
c
|
\ .

(2.18)
Since 2 p eE D inmostapplications,Equation(2.18)gives
1 1
1 0
dp V
eE
K dt x
DK
| |
|
c
+ + =
|
c
|
\ .

(2.19)
SubstitutingEquation(2.1)andtheexpressionforthetotalderivativeintoEquation(2.19)gives
2
0
p p V
V a
t x x
c c c
+ + =
c c c

(2.20)
2.2.2.2 EquationofMotion
The differential equation of motion for a transient flow is given applying the second Newton's
lawforafluidelementinsideatube.[17]
p
A
p
p
x
A
x
c
|
|
+
A
|
c
\
.
gA x A
2
p x A
p x
x x
c A c | |
+ A
|
c c \ .
0
D x t t A
Figure22:Controlvolumeforthederivationoftheequationofmotion
TheequationofmotionforthecontrolvolumedefinedinFigure22atagiveninstantt,when
thepipeisdeformedduetotransientforces
0
1
sin
2
p p A dV
pA p x A p x x D x gA x A x
x x x dt
t t o
c c c
| | | |
A + + A A A A = A
| |
c c c
\ . \ .

(2.21)
Omitting
2
x A
,
Equation(2.21)yields
0
sin 0
p dV
A D gA A
x dt
t t o
c
+ + + =
c

(2.22)
The force resulting from the shear stress t

0
may be defined in terms of the DarcyWeisbach
expressionforfrictionlossesforaturbulentpipeflow.
0
1
8
f V V t =

(2.23)
SubstitutingEquation(2.23)intoEquation(2.22)anddividingthroughout A ,yields
1
sin 0
2
V V p fV
V g V
t x x D
o
c c c
+ + + + =
c c c

(2.24)

10
2.2.2.3 GeneralRemarks
Inmanyengineeringapplications,theconvectiveaccelerationterms ( ) V p x c c and ( ) V V x c c
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
2
0
p V
a
t x
c c
+ =
c c
(2.25)
1
0
2
V p fV
V
t x D
c c
+ + =
c c

(2.26)
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
neglectingthevariationofandflowareaAduetovariationoftheinsidepressure.Thesmall
variationsofandflowareaAareindirectlytakenintoaccountbyconsideringthewavespeed
a have a finite value. Substituting Equations (2.27) and (2.28) into Equations (2.25) and (2.26)
give
2
0
H z a Q
t t gA x
c c c
| |
+ =
|
c c c
\ .

(2.29)
2
1
0
2
Q H z fQ
g Q
A t x x DA
c c c | |
+ + =
|
c c c
\ .

(2.30)
If the pipe axis is fixed 0 z t c c = and sin z x o c c = . It is possible to eliminate the tube
inclinationeffect, sin 0 z x o c c = = .Therefore,Equations(2.29)and(2.30)become
2
0
Q gA dH
x a dt
c
+ =
c

(2.31)
2
1
0
2
H dQ fQ
Q
x gA dt gA D
c
+ + =
c

(2.32)
11
Equations(2.31)and(2.32)canbenormalizedbysubstitutingfortherelativeterms
0
h H H =

and
0
q Q Q = ,giving
0
2
0
0
H q gA h
x a Q t
c c
+ =
c c

(2.33)
2
0 0
2
0 0
0
2
Q Q h q f
q q
x gAH t gA D H
c c
+ + =
c c

(2.34)
Thewatertimeconstantorwaterstartingtime,T
W
,isassociatedwiththeaccelerationtimefor
water in the penstock between the turbine inlet and the reservoir or the surge tank, if one
exists.Theequationforwaterstartingtimeis:
0
0
W
Q L
T
gA H
=

(2.35)
IntroducingthewatertimeconstantT
W
inEquation(2.33)and(2.34)
2
1
0
W
q L h
x a T t
c c
+ =
c c

(2.36)
0
0
2
W W
T T Q h q
f q q
x L t L AD
c c
+ + =
c c

(2.37)
2.2.3 MathematicalModel
The analysis of transients in hydroelectric systems is subdivided in massoscillation response
andwaterhammerresponse.Ordinarydifferentialequationsdescribethemathematicalmodel
for massoscillation (or rigid water column) between the reservoir and the surge tank. Partial
differentialequationsshowthemathematicalmodelforwaterhammeralongthepenstockand
drafttube.[1520]
The mathematical model of the conduit system taking into account the waterhammer theory
and considering the head losses is based on the linearized Equation (2.36) and (2.37), and on
sinusoidalflowandpressurefluctuations.
UsingtheLaplacetransformtoexpressEquations(2.36)and(2.37)inthefrequencydomain
( )
( ) ( ) ( )
2
1
0 0
W
q s
L
sh s h
x a T
+
c
+ =
c
(2.38)

( )
( ) ( ) ( )
( )
0
0 0
2
W W
h s T T Q
sq s q f q s
x L L AD
+
c
+ + =
c

(2.39)
12
The third term in Equation (2.39) is related to the head losses due to friction. Assuming an
initialsteadystatesituation,
( )
0 0 h
+
= and
( )
0 0 q
+
= ,Equation(2.38)and(2.39)gives
( )
( )
2
1
0
W
q s
L
sh s
x a T
c
+ =
c

(2.40)

( )
( )
0
0
2
W
h s T Q
s f q s
x L AD
c
| |
+ + =
|
c
\ .

(2.41)
DifferentiatingEquation(2.41)withrespecttoxyields
( ) ( )
2
0
2
2
W
h s q s T Q
s f
x L AD x
c c
| |
= +
|
c c
\ .

(2.42)
Substituting ( ) q s x c c byEquation(2.40)intoEquation(2.42)
( )
( )
2
2 0
2 2
1
2
h s Q
s f s h s
x a AD
c
| |
= +
|
c
\ .

(2.43)
Introducingthevariablez
2 2 0
2
fQ
z s s
DA
= +

(2.44)
ThegeneralsolutionforthesecondorderhomogeneousEquation(2.43)is
( )
1 2
z z
x x
a a
h s C e C e
= +

(2.45)
Bydifferentiatingequation(2.45)withrespecttoxyields
( )
1 2
z z
x x
a a
h s
z z
C e C e
x a a
c
=
c

(2.46)
Substituting ( ) h s x c c byequation(2.46)intoequation(2.41)gives
( )
1 2
z z
x x
a a
W
L s
q s C e C e
aT z
| |
=
|
\ .
(2.47)
13
Thewaterpressureandtheflowrateofupstreaminletofthetubeatx=0is
( )
1 2 U
h s C C = + (2.48)

( ) ( )
1 2 U
W
L s
q s C C
aT z
=

(2.49)
Thewaterpressureandtheflowrateofdownstreamoutletofthetubeatx=Lis
( )
1 2
z z
L L
a a
D
h s C e C e
= +

(2.50)

( )
1 2
z z
L L
a a
D
W
L s
q s C e C e
aT z
| |
=
|
\ .
(2.51)
Hence,itfollowsfromEquation(2.50)and(2.51)that
( ) ( )
1
1 1
2 2
z z
L L
W a a
D D
aT z
C e h s e q s
L s

= + (2.52)

( ) ( )
2
1 1
2 2
z z
L L
W a a
D D
aT z
C e h s e q s
L s
=

(2.53)
The substitution of the values of T

e
, C
1
and C
2
from Equations (2.3), (2.52) and (2.53) into
Equations(2.48)and(2.49)yield
( ) ( ) ( ) ( ) ( )
1 1
2 2
e e e e
zT zT zT zT
W
U D D
e
T z
h s e e h s e e q s
T s

= + +
(2.54)
( ) ( ) ( ) ( ) ( )
1 1
2 2
e e e e
zT zT zT zT
e
U D D
W
T s
q s e e h s e e q s
T z

= + +

(2.55)

14
Thetransferfunctionofflowrateandwaterpressureofupstreaminletanddownstreamoutlet
ofthetube,expressedinthematrixnotation,is
( )
( )
( ) ( )
( ) ( )
( )
( )
cosh sinh
1
sinh cosh
e c e
U D
e e U D
c
zT Z zT
h s h s
zT zT q s q s
Z
(
( (
(
=
( (
(

(

(2.56)
Inwhichthecharacteristicimpedanceforthepipeis
W
c
e
T z
Z
T s
=

(2.57)
Theclassicalwavesolutiongivenbyananalysisofthepartialdifferentialequationsintimeand
space defining pressure and flow rate at each point in the conduit taking into account the
elasticwaterhammertheoryandconsideringhydrauliclossasahyperbolictangentfunctionis
( )
( )
1 2 1 2
2 0 0
1 tanh
2 2
W
e
e
h s T fQ fQ
s s T
q s T DAs DA
| |
| | | |
= + + |
| |
|
\ . \ .
\ .
(2.58)
Neglectingthehydraulicfrictionlosses,Equation(2.58)canbesimplifiedas
( )
( )
( ) tanh
W
e
e
h s T
sT
q s T
=

(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
e e
sT sT ~ , Equation (2.58) for small variations around an operating point can be
simplifiedas
( )
( )
W f
h s
T s H
q s
=
(2.60)
Neglectingthehydraulicfrictionlosses,Equation(2.60)yields
( )
( )
W
h s
T s
q s
=

(2.61)

15
2.3 PressureControlSystems
Hydroelectric power plants with long conduits may have severe water hammer and governing
stabilityproblems.Inareactionturbinehavinglongpressuretunnels,thedynamicresponseof
both upstream and downstream water columns must be studied in order to avoid excessive
pressureoroverspeedproblems.Pressuretransientswillonlybeproducedatupstreamsidein
an action turbine; at downstream there will be a canal with a transient free surface flow. 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.
Depending upon the requirement of the hydroelectric plants, the surge tank can be placed at
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]
Theconduitwallsarerigidandthe liquidiscompressible.Thismeans thataflow change

atanypointinthesystemistransmittedinstantaneouslythroughoutthesystem,andthe
liquidmoveslikeasolidslug
The inertia of the liquid in the surge tank is small compared to that of the liquid in the
tunnelandcanthereforebeneglected.
The heads losses in the system during the transient state can be computed by using the
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.
16
Figure23:Simplesurgetank,[6]
ThedynamicequationisgivenapplyingtheNewtonssecondlawofmotion;therateofchange
ofmomentumisequaltotheresultantforce.Thedynamicequationforasimplesurgetankis
givenasfollows
( )
t t
dQ gA
H z fQ Q
dt L
= +

(2.62)
The equation of continuity for the junction of the tunnel and the surge tank, shown in Figure
23,maybewrittenas
t s tur
Q Q Q = +

(2.63)
whereQ
s
istheflowintothesurgetank,Q
tur
istheturbineflow.Since ( )
s s
Q A dz dt = ,Equation
(2.63)becomes
( )
1
t tur
s
dz
Q Q
dt A
=

(2.64)
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:
2
S
t
A L
T
g A
t =

(2.65)
0
S t
L
Z Q
gA A
=

(2.66)
17
2.3.2 SurgeTankMathematicalModel
Inasimplereservoirsurgetanksystemtheunsteadyflowcanbemodelledbyasimpleordinary
differential equation. The inclusion of surge tank effects is warranted in cases where dynamic
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
0
0
s
A H dh
q
Q dt
=

(2.67)
Thesurgetankfillingtime,T
s
,isdefinedas
0
0
s
s
A H
T
Q
=

(2.68)
Thetransferfunctionofflowrateandwaterpressureofupstreamthesurgetankis
( )
( )
1
s
h s
q s sT
=

(2.69)
18
3 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
on the periphery of the runner. As the water after strikingthe bucketsfalls into the tail water
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
19
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
conduit system. Reaction turbines extract power from the kinetic energy of water because of
the difference in pressure between the front and the back of each runner blade as the water
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
Upstream and downstream water columns are sensitive to flow disturbances at reaction type
turbine. Pressure regulators or relief valves areused to temporarily allow part of the water to
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.

20
3.2 GeneralTechnicalAspects
The hydroelectric generation is extremely site dependent because it depends on falling water.
ThebasicelementsoftheHydraulicPowerPlantaredepictedinFigure33.Theturbineproper
is, according to standard practice, taken to begin at the entrance to the turbine casing and to
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 q = (3.1)
The turbine efficiency q 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. The turbine efficiency is defined as the ratio of power supplied by the turbine
(mechanical power transmitted by the turbine shaft) to the absorbed power (hydraulic power
equivalenttothemeasureddischargeunderthenethead).[23]
Thegeneratorefficiencyisusuallyconsideredconstantoverawideoperatingrange.However,
the hydraulic efficiency depends significantly on both the water discharge and the net head.
TheserelationshipsaregenerallyrepresentedinthesocalledTurbineHillChart.
21
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. The plots of the prototype turbine characteristics are based in steadystate model test
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:
( )
11 11
, Q f n y =

(3.2)
( )
11 11
, M f n y = (3.3)
However,therearestillnodetailedandprecisemathematicalmodelsthatcandescribetheunit
flow rate Q
11
and unit torque M
11
. Q
11
and M
11
are usually obtained by interpolation method.
Theexpressionfortheunitrotationspeedn
11
,andunitflowQ
11
maybewrittenasfollows
11
ref
nD
n
H
=

(3.4)

11
2
ref n
Q
Q
D H
=

(3.5)

11 3
ref n
M
M
D H
=

(3.6)
The characteristic curves are extrapolated for small wicketgate openings. Therefore, the flow
should be known when the turbine rotational speed is zero. The windage and friction losses
shouldbeknownatwicketgateopeningsbelowthespeednoloadgate.Thenoloadgateisthe
lowestgateopeningatwhichturbinerotatesatsynchronousspeedwithzerooutput.[6]
A typical Hill Chart for a Francis turbine is shown on Figure 5.2 of Chapter 5, page 131, of the
textbookAppliedHydraulicTransients.[6]
22
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
opening position and runner blade angle. The nonlinear characteristic of a Francis turbine can
bewrittenas,[1,6,8,16,23,25,2831]
( ) , , ,
p
p f h n y | =

(3.7)
( ) , , ,
q
q f h n y | = (3.8)
The effect of the runner blade movement is not considered in the modelling of turbines with
fixed blades, such as Francis turbines. The flow in an impulse turbine depends upon the head
andthenozzleopeningonly.
The turbine dynamics is related to the generator dynamics through mechanical power P
m
produced by the turbine. The relation between mechanical power P

m
and mechanical torque
M
m
,expressedinperunit,isgivenby
m m
s
n
p m
e
=

(3.9)
Normally, a simplifying assumption made is that

s
n e = at synchronous speed. This is not the
sameassayingthatthespeedisconstant,itassumesthatspeedchangesaresmallanddonot
haveasignificanteffect.Thus,equation(3.9)yields
m m
p m =

(3.10)
23
3.3.1 SimplifiedNonlinearTurbinemodel
Thenominalpowerobtainedbyahydraulicturbineisproportionaltothepotentialenergylost
bythefallingfluid.Due totheturbineisnot100%efficient,thenoloadflowq
NL
is subtracted
from the actual flow. A damping effect that is dependent on gate opening is also present, so
thatatanyloadconditiontheturbinepowercanbeexpressedby
( ) ( ) 1
m t NL D N
p Ah q q K y q e = (3.11)
Detailed information is available from turbine manufacturers on the variation of the turbine
efficiency q with variation in the steady operating point of a turbine. The use of this data
involves the use of a stored table and an interpolation procedure. Neglecting the variation of
turbineefficiency,Equation(3.11)yields
( ) ( ) 1
m t NL D N
p Ah q q K y e = (3.12)
TheconstantproportionalityfactorA
t
iscalculatedusingtheturbineMWratingandgenerator
MVAbase.A
t
iscalculatedby
( )
( ) 1
t
NL
turbine power MW
A
h q q generator MVA rating
=

(3.13)
Thepressureheadacrosstheturbineisrelatedtotheperunitflowratebyassumingthatthe
turbinecanberepresentedbythevalvecharacteristic
q y h = (3.14)
3.3.2 Thelinearizedhydroturbinemodel
The most common hydraulic turbine representation for system stability studies consists of a
transfer function obtained by linearizing the turbine characteristic curves around an operating
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
h n y
c c c
A = A + A + A
c c c

(3.15)
q q q
q h n y
h n y
c c c
A = A + A + A
c c c

(3.16)
24
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
oftheturbinerelativetogateopeningpositionyandunitrotationalspeedn
11
.
Theinfluenceoftheturbinecoefficientsonthemodelaccuracyiscritical.Thesevalueshaveto
bemeasuredaccuratelyonthefieldortakenfrommodeltests.
The turbine coefficients of any operating point can be obtained by interpolation before
simulation. For deviations around rated speed and pressure head, the turbine coefficients are
deduced by differentiation of Equations (3.1) and (3.14), [11]. Differentiating Equations (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
Derivatives
IEEE
Model,[11]
IEEE
Model,[37]
q
h
c
c

0
0.5 y
2
y
h
q
n
c
c

0.0 0.0
q
y
c
c

1.0 h
m
h
c
c

0
1.5 y q
3
2
t NL
A y h q
| |
|
\ .
m
n
c
c

0.0
a
D y

m
y
c
c

q
( )
3
2
1
t a N
Ah D e

The partial derivative of torque with respect to gate position, m y c c , 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 c c , is usually considered to be negligible. The deviation of
mechanical torque with rotational speed, m n c c , 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.
25
Table32:TurbineCoefficients,[32]
Partial
Derivatives
GeneratorOperatingConditions
IEEE
model
22.5(MW) 84.3(MW) 112.0(MW)
q
h
c
c

0.06 0.20 0.24 0.50
q
n
c
c
0.13 0.38 0.62 0.00
q
y
c
c
0.80 0.40 0.38 1.00
q
|
c
c

0.00 0.30 0.69 0.00
m
h
c
c

0.40 1.20 1.50 1.50
m
n
c
c

0.39 0.86 0.75 0.00
m
y
c
c

0.88 0.90 0.34 1.00
m
|
c
c

0.00 0.50 0.52 0.00
y
| c
c

0.00 2.30 1.00 0.00
ItisworthytonoticethatthelinearmodelsimplementedintheMastersThesisworkusesthe
standardvaluesaccordingtoIEEEmodeldetailedinTable32.
26
4 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.
The synchronous machine is assumed to have a threephase stator armature winding, a rotor
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) The change in the inductance of the stator windings due to rotor position is sinusoidal
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.

27
4.2 SynchronousGeneratorEquations
The generator windings are magnetically coupled, hence, the flux in each winding depends on
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 0
0 0 0
0 0 0 0
0 0 0 0 0
0 0 0
0 0 0
0 0 0 0
d ad l ad ad d
q aq l aq q
f ad ad f ad f
D ad ad ad D D
Q aq aq Q Q
L L L L i
L L L i
L i
L L L L i
L L L L i
L L L i
+ ( ( (
( ( (
+
( ( (
( ( (
=
( ( (
+
( ( (
( ( (
+
( ( (
+
( ( (

(4.1)
Statorvoltageequations:
0 0 0
0 0
1
0 0
0 0 0
d d q a d
q q d a q
s s
a
v R i
d
v R i
dt
v R i

e

e e
( ( ( ( (
( ( ( ( (
= +
( ( ( ( (
( ( ( ( (

(4.2)
Rotorvoltageequations:
0 0
1
0 0 0
0 0 0
f f f f
D D D
s
Q Q Q
v R i
d
R i
dt
R i
( ( (
(
( ( (
(
= +
( ( (
(
( ( (
(

(4.3)
Thetransformeremfs,the d dt terms,arereferredtothearmatureemfsproportionaltothe
ratechangeoftheflux.Thetransformeremfsareduetochangingcurrentsincoilsonthesame
axisastheoneconsidered.
In addition to the equations describing the electrical dynamic performance, an expression for
theelectromechanicaltorqueisrequired.
s
e d q q d
m i i
e

e
( =

(4.4)

28
RotormechanicalEquations:
Therotormechanicaldynamicaregivenbytheswingequation,whichcanbeexpressedvia,[2]:
( )
2
s
m e D
d
m m K
dt H
e e
e
A
= A (4.5)
s
t
o
e e e
c
= A =
c

(4.6)
Intheaboveequations,allquantitiesareinperunitexcepttimethatisinsecondsandand
s
which are in rad/sec. In some literature, per unit speed

N s
e e e = is used. Therefore,
Equations(4.5)and(4.6)yield
( ) ( )
1
1
2
N
m e D N
d
m m K
dt H
e
e
A
= (4.7)
( ) 1
s N
t
o
e e
c
=
c

(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
e e ~ .
The transformer emfs (
d
d dt and
q
d dt ) are neglected because are small
comparedwiththerotationemfs(
q
e and
d
e + ).
Thedifferentialequationsfortheelectricaldynamics,expressedinorthogonalphasequantities,
usedtomodelanumberofdifferentmodelsare
( )
q
do q q d d d
dE
T E E X X I
dt
= +
(4.9)
( )
d
qo d d q q q
dE
T E E X X I
dt
=

(4.10)
29
( )
q
do f q d d d
dE
T E E X X I
dt
= +
(4.11)
( )
d
qo d q q q
dE
T E X X I
dt
=

(4.12)
4.3.1 SynchronousMachineRepresentedbytheClassicalModel
Thesecondordersynchronousgeneratormodel,referredtoastheclassicalmodel,iswidely
used for simplified analysis of power system dynamics. In the classical model all the voltages,
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
do
T is relatively long and the changes in E
f
and I
d
are small.
Additionallytherotortransientsaliencyisneglected.
'
E o Z
Figure41:Equivalentcircuitoftheclassicalmodelofthegenerator
TheclassicalsynchronousgeneratormodelassumesthatneitherthedaxisarmaturecurrentI
d
nor the internal emf E

f
representing the excitation voltage change very much during the
transientstate.Inthismodel,thegeneratorisrepresentedbyaconstantemfEthroughoutthe
studyperiodbehindthetransientreactanceX
d
andtherotorinertiaequations,Equations(4.5)
and(4.6).[22]
Theequationdescribingthearmaturevoltageyields
( ) ( )
q d d q d d
V E jE jX I jI E jX 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.
30
5 TurbineGoverningSystems
The governing system is a system that regulates the turbinegovernor speed and hence the
frequency and the active power in response to load variation. The turbine governor regulates
theinletofwaterintoaturbine,whichinturnrotatesthegeneratortoproduceelectricity.This
chapter details the modelling of the mechanicalhydraulic and electrohydraulic turbine
governingsystem.
5.1 GoverningSystem
The main function of the turbine governing system is to regulate the turbinegenerator speed
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]
ThepermanentspeeddroopR
P
determinestheamountofchangeinoutputaunitproducesin
response to a change in unit speed. The permanent speed droop can be developed either by
using the wicket gate position or by using the unit output power. For isolated or islanded
operation,thepermanentspeeddroopcanhelptostabilizetheunitspeedcontrolbyproviding
an intermediate feedback that limits the over travel of the turbine control servomotors while
controllingtheunitspeed.Thepermanentspeeddroopdeterminestheamountofparticipation
the unit produces when responding to disturbances in system frequency in operation while
synchronizedtoaninterconnectedpowersystem.
ThetemporarydroopR
T
isusedto limitovershoot oftheturbinecontrolservomotorduringa
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. The response of a hydro turbine to speed changes or to changes in speedchanger
settingisrelativelyslow.
31
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]
The main disadvantages of the Watt centrifugal governor are the present of deadbands and a
relativelylowaccuracy.Thesizeofthedeadbandsalsotendstoincreasewithtimeduetowear
in the moving mechanical elements. The conventional mechanicalhydraulic governing system
hasbeenreplacedbyanelectrohydraulicgovernor.
5.2.1 MathematicalModelling
Figure 52 shows a simplified mechanicalhydraulic governor. The variables used in the
derivationofthetransferfunctionsareperunitderivativesfromtheinitialsteadystatevalues.
[2,22,47]
Figure52:MechanicalHydraulicGoverningSystem.[47]
32
Thetransferfunctionofthedistributingvalveandgateservomotoris
1
K y
b s
=

(5.1)
Thetransferfunctionofthepilotvalveandpilotservomotoris
2
1
p
K b
a T s
=
+

(5.2)
where K
2
is determined by the feedback lever ratio and T
p
by the port areas of the pilot valve
andK
2
.CombiningEquations(5.1)and(5.2)yields
( ) ( )
1 2
1 1
s
P P
K K K y
a s T s s T s
= =
+ +

(5.3)
TheservomotorgainK
s
isdeterminedbythepilotvalvefeedbackleverratioandtheportareas
of the distributing valve and gate servomotor. The governor response time, 1
g s
T K = , is
related to the gate closing time T
c
, where T
c
is the time required for fullgate travel with the
gatesmovingatmaximumvelocity,typically5to10seconds.
Assumingthattheflowofdashpotfluidthroughtheneedlevalveisproportionaltothedashpot
pressure,thedashpottransferfunctionis
1
R
T
R
T s c
R
y T s
=
+

(5.4)
ThetemporarydroopR
T
isdeterminedbytheselectionofpivotpointfortheleverconnectedto
theinputpiston.TheresettimeT
R
isdeterminedbytheneedlevalvesetting.
The pilot valve input signal is produced adding the action of a system of floating levers, the
referencespeed,shaftspeed,permanentdroop,andtemporarydroopsignals.
1
R
ref s P T
R
T s
a R y R
T s
e e =
+

(5.5)
AcombinationofEquations(5.3),(5.4)and(5.5)gives
( )
( ) ( ) ( )
3 2
1
1
1
R
P
ref s g R P T P R g P R g
P P P
T s
R y
T T R R s T T T s T T T s
R R R
e e
+
=
+ + +
+ + +

(5.6)
33
The block diagram of a typical hydro turbine governing system suitable for stability analysis is
showninFigure53.
+
1
1
P
sT +
1
s
1
1
g
sT +
S
K
P
R
1
R
T
R
sT
R
sT +
r
e
+

+
+
Figure53:Modelofgovernorforhydraulicturbines,[2]
As the pilot valve and servomotor time constant T

P
is several times smaller than the time
constantsT
g
andT
R
,itmaybeneglectedtogivethesecondordertransferfunction
( )
( )
( )
1
1
1 1
R
P
ref s g R P T R g
P g R P T
T s
R y
T T R R T T
s s
R T T R R
e e
+
=
| | + + | |
+ +
| |
|
+ +
\ . \ .
(5.7)
Forstableoperationunderislandingconditions,theoptimumchoiceofthetemporarydroopR
T
andresettimeT
R
isgivenby;[2,50]
( ) 2.3 1.0 0.15

W
T W
M
T
R T
T
= (

(5.8)
( ) 5.00 1.00 0.50

R W W
T T T = (

(5.9)
Table51givestypicalvaluesandrangeofparametersforhydraulicgovernors.
Table51:TypicalvaluesandrangeofParameters,[47]
Typicalvalues Range
T
R
5.00 2.5025.0
T
g
0.20 0.200.40
T
W
1.00 0.505.00
T
M
(2H) 8.00 6.0012.0
T
c
5.00 5.0010.0
R
P
0.04 0.030.06
R
T
0.31 0.201.00
34
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
accumulates at a rate proportional to the size of the error input. The integral term also trims
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]
Gate
Speed
Ref.
Speed
+
+
+
R
P
K
i
s
Gate
Servo
Pilot
Servo
K
P
sK
d
+
-
+
-
Figure54:TypicalPIDGovernorController,[2]
The transfer function of the PID governor expressed in terms of proportional, integral and
derivativegains,is
( )
2
2
1
1
d p i
ref s P d p P i
K s K s K
y
R K s K R s K e e
(
+ +
= (
+ + +
(

(5.10)

35
Theoptimumparameterssuggestedin[53]areexpressedasfollows
0.80
M
p
W
T
K
T
=
(5.11)
2
0.24
M
i
W
T
K
T
=
(5.12)
0.27
d M
K T =
(5.13)
Theuseofahighderivativegainwillresultinexcessiveoscillationsandpossiblyinstabilitywhen
the generating unit is connected to an interconnected system. When the derivative gain is set
tozero,thetransferfunctionofPIDcontrollerisequivalenttothatofthemechanicalhydraulic
governor.Thedesignisbasedonlinearcontroltheoryatoneloadconditionandthendetuned
forworstoperatingconditions.Thiscontrollerdesigndoesnotguaranteethecloseloopsystem
toremainstableatalloperatingconditions.ThetransferfunctionofthePIgovernoryields
1
1
p i
ref s P
p i
P
K s K
y
R
K s K
R
e e
+
=
| |
+ +
|
\ .

(5.14)
Addingasmalltimeconstant,Equation(5.14)maybeimprovedas
( )
2
1
1
1
1 1 0.1
p
i
p P ref s P
i P
K
s
K y
K R R
s s
K R
e e
+
=
+ | |
+ +
|
\ .

(5.15)
36
6 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
involve a reaction of the control system. The representation of each component of the power
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
surge tank for Hydropower Systems with long length penstocks are implemented. Finally, the
nonlinearmodelsforaturbinewithlonglength penstocksarelinearizedatan operatingpoint
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.
A hydroelectric power plant may be considered as a system consisting of several subsystems:

the penstock including any surge chamber, the hydraulic machine, the speed governor, the
tailraceand,finally,thegeneratorandelectricalnetwork.ThediagramofFigure61depictsthe
functionalblocksfortheelementsofaHydraulicTurbineGeneratingSystem.
The turbine governing system is modelled by a mechanicalhydraulic turbine governing model

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).
37
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
e
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.
The hydraulic turbine is modelled with varying degrees of detail. The mathematical models of
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
Linear Turbine Model with Surge Tank including Elastic Water Column in Penstock and
turbinecharacteristics

38
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
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
6.2.1 SimplifiedNonlinearTurbineModel
The Simplified Nonlinear Turbine Model is based upon a simple hydraulic turbine with a short
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).
In the modelling of the turbine itself, both its hydraulic characteristics and mechanical output
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
1
Head
-
+
1
Tws
flow
Torque
qnl
At
KD
+
-
-
+
x
2
Figure62:SimplifiedNonlinearTurbineModel
39
6.2.2 NonlinearModelwithoutSurgeTankassumingInelasticWaterColumn
The Nonlinear Model without Surge Tank assuming Inelastic Water Column is based upon the
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 h

f
are proportional to flow squared and depends on the conduit dimensions
andfrictioncoefficientf
p
.Itsufficesinthismodelassumethattheheadlossesareproportional
to flow squared and the head loss coefficient. The transfer function relating the flow rate and
waterpressureofasimplePenstocksystemisdescribedbyEquation(2.60).
In the modelling of the turbine itself, both its hydraulic characteristics and mechanical output
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]
Gate
1
Head
-
+
1
Tws
flow
Torque
qnl
fp
-
At
KD
+
-
-
+
x
2
x
2
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
representedbyadelay
2
e
s T
e
inthehydraulicstructure.
40
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 f
p1
and f
p2
,
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.
41
Thehydraulicmodelofthesurgetankincludesanorificethatdissipatestheenergyofhydraulic
oscillationsandproducesdamping.Headlossesintheorificeofthesurgetankareproportional
tolosscoefficientsf
0
timesflowratetimesabsolutevalueofflowratetomaintainthedirection
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
function are the per unit deviations in gate opening position and turbine torque, respectively.
ThismodelisdepictedinFigure5inthereportHydraulicTurbineandTurbineControlModels
forSystemDynamicStudies.[1]
Figure65:NonlinearModelwithSurgeTankassumingInelasticWaterColumns
6.2.5 Nonlinear Model with Surge Tank assuming Elastic Water Column in Penstock
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
surge tank including elastic water column in penstock and inelastic water column in upstream
tunnel.
Thedynamiceffectsoftheuppertunnelcontributewithlowfrequencyoscillations,intheorder
of0.10Hz.Thedynamicbehaviourofthepenstockcontributeswithhighfrequencyoscillation
duetothepenstockissubjecttoabruptgateorflowchanges.
42
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
proportionaltolosscoefficientsf
0
timesflowratetimesabsolutevalueofflowratetomaintain
thedirectionofheadloss.
The transfer function of flow deviation and water pressure deviation of the upstream water
tunnelisdescribedbyEquation(2.60).Thetransferfunctionofthepenstocktakingintoaccount
the elastic water hammer theory is described in Equation (2.59). The transfer function of flow
rate and water pressure of the surge tank is described by Equation (2.69). The mechanical
power output and hydraulic characteristics of a turbine are described by Equations (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
deviations in gate opening position and turbine torque, respectively. This model is depicted in
Figure 6 in the report Hydraulic Turbine and Turbine Control Models for System Dynamic
Studies.[1]
Gate Head
+
+
Torque
qnl
fp
-
At
-
-
+
x
2
Z
c
Tanh (T
e
s)
x
flow
1
T
W2
s
fp
1
x
1
C
S
s
fp
2
x
1
-
-
-
+
+
-
+
KD
Figure66:NonlinearModelwithsurgetankassumingelasticwatercolumninpenstockandinelastic
watercolumninupstreamtunnel

43
6.3 HydroTurbineLinearModels
The nonlinear characteristics of the hydro power plant models can be approximated by linear
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
Linear Turbine Model with Surge Tank including Elastic Water Column in Penstock and
turbinecharacteristics
These linear models of the hydraulic turbine consider its characteristic curves and coefficients
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
( ) ( )
2 2 2 2 1 W f
T s H q h h + =
(6.1)
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.
44
2 2
1
W f
T s H
q
y
c
c
1 1 W f
T s H
1
S
T s
m
h
c
c
q
h
c
c
m
y
c
c
q
n
c
c
m
n
c
c
Figure67:LinearModelwithSurgeTankassumingInelasticWaterColumns
Equating (2.60), (2.69), (3.15), (3.16) and (6.1) gives the transfer function relating the per unit
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
simple turbine with an upstream tunnel, a surge tank and a penstock. The upstream tunnel is
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.
45
2 2
1
W f
T s H
q
y
c
c
1
S
T s
m
h
c
c
q
h
c
c
m
y
c
c
q
n
c
c
m
n
c
c
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
( ) ( )
2
2 2 1
tanh
W
e
e
T
sT q h h
T
| |
=
|
\ .
(6.2)
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) gives the transfer function relating the per unit
deviationsturbinetorquetogatepositionforaLinearModelwithSurgeTankassumingElastic
WaterColumninPenstock.
46
7 Hydroelectric Power Plant Modelling by

StructureMatrixMethod
This chapter shows an alternative method of Hydroelectric Power Plant modelling for stability
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.
The representation of each component of the Hydro Power System is presented by means of
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
systems require a thorough stability analysis of the speed control of the units in the systems.
TheoriginoftheStructureMatrixMethodmaybetracedtotheStructureAnalysistheoryand
terminologies such as Element, "Node" as well as "Local" and "Global" matrices have been
used.ElementMatrixrepresentationsofeachoftheelementsforthe modellingoftheoverall
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.
47
7.2 Matrix representations of the Basic Elements in Hydro Power

Systems
TheStructureMatrixmethodisbasedonthesamefunctionalequationsthatTransferMethod
and Impedance Method. The essential part of the Structure Matrix Method is to establish the
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.
The dynamic behaviour of the elastic fluid in a closedconduit is described in Chapter 2.2. The
Element Matrix representation for a pipe based on the transfer function detailed in Equation
(2.56)is
( ) ( )
( ) ( )
( )
( )
( )
( )
1 1
1 1
tanh sinh
1 1
sinh tanh
c e c e
i i
i i
c e c e
Z zT Z zT
h s q s
h s q s
Z zT Z zT
+ +
(
(
( (
(
=
( (
(

(

(7.2)
Usingthefollowingexpressions:
( )
1
tanh
c e
T
Z zT
=

(7.3)
( )
1
sinh
c e
S
Z zT
=

(7.4)
Thegeneralpipematrix,Equation(7.2)yields
( )
( )
( )
( )
1 1
i i
i i
h s q s T S
h s q s S T
+ +
( ( (
=
( (
(
(7.5)
7.2.2 Surgetanksorairaccumulators
The Element Matrix expression for surge tank or air accumulator may be written as either a
oneterminal element or a virtual twoterminal element. The Element matrix for a surge tank
writtenasoneterminalelementis
( ) ( )
0
0
s
i i
H A
s h s q s
Q
(
= ( (
(

(7.6)

48
The Element matrix expression for a surge tank written as a virtual twoterminal element,
consideringh
i+1
=q
i+1
=0,is
( )
( )
( )
( )
1 1
0
0
1 0
0
i i
s
i i
h s q s
H A
s h s q s
Q
+ +
(
( (
(
=
( (
(

(

(7.7)
Forafreewatersurfaceofanopensurgetank,themostconvenientrepresentationforasurge
tankisatwoterminalElementbyincorporatingEquations(7.6)and(7.7)together.
( )
( ) ( )
1 1
0
i F
i i
h s T T S
q s h s S T
+ +
( ( (
=
( ( (

(7.8)
where
0
0
s
F
H A
T s
Q
=

(7.9)
7.2.3 LocalLosses
The Element Matrix describing the damping caused by local variations in cross sections of the
conduitorforfixedpartialopenvalvesmaybeestablishedasfollows
( )
( )
( )
( )
1 1
1 1
1 1
re re i i
i i
re re
K K h s q s
h s q s
K K
+ +
(
(
( (
(
=
( (
(

(

(7.10)
Forthrottles,K
re
isgivenby
0
2 2
0
1 1
t
re
s b
QQ
K
gH A A
| |
=
|
\ .
(7.11)
In Equation (7.11), A
s
is the smallest cross section and A
b
is the largest cross section. A loss
constantisincluded.
ForTjoints,K
re
isgivenby
2
0 0
2 2
0 0
3
t
re
s b
kQ QQ
K q
gA H gA H
= +

(7.12)
InEquation(7.12),kisthegeometryconstant.
49
7.2.4 HydroTurbine
Theanalysisofthestabilityisvalidforfrequenciesdomainsnormallyuptoabout6rad/sec,and
resonance frequencies in the turbine runner and guide vane system can be neglected. In this
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
Q H
q
q
c c c
= + +
c c c

(7.13)
Dividingthroughby
0 0 0 0
P gH Q q = ,Equation(7.13)yields
p q h q A =A +A +A (7.14)
The values of Aq, Ah and Aq may be found by studying the characteristic diagram of the
turbine.
TheflowthroughaFrancisturbinedependsuponthenethead,rotationalspeedoftheturbine
andwicketgateopeningposition.Theequationoftheflowyields
( )
1
1
2
y n n
q Q y Q n Q h Q
|
| A = A + A A + A

(7.15)
The values Q
n
, Q
y
and Q
|
may be determined by means of the turbine characteristic diagram.
TheexpressionsforQ
n
,Q
y
andQ
|
are
( )
( )
11
0 11
11 11
0 0
n
n
Q
Q
n Q
c
=
c

(7.16)

( )
0 11
0 11
0
y
Y Q
Q
Y Q
c
=
c

(7.17)

( )
0 11
11 0 0
Q
Q
Q
|
|
|
c
=
c

(7.18)
Theefficiencyisdefinedastheratioofpowersuppliedbytheturbinetotheabsorbedpower.
Therelativeefficiencycanbeexpressedasfollows
( )
1
2
q q n n
E q E E h E n E
|
q | A = A A + A + A

(7.19)
50
ThevaluesE
n
,E
y
andE
|
maybedeterminedbymeansoftheturbinecharacteristicdiagram.The
expressionsforE
n
,E
y
andE
|
are
0
11 0
0
q
Q
E
Q
q
q
c
=
c

(7.20)

( )
11
0
11 0
0
n
n
E
n
q
q
c
=
c

(7.21)

0
0 0
E
|
| q
| q
c
=
c

(7.22)
The runner blade servomotor has a time constant T

L
that describes the runner blades
movement as a function of the guide vane movement. The transfer function describing the
runnerbladesmovementis
1
L
L
K
F
T s
|
=
+

(7.23)
The runner blades normally are stopped mechanically before the guide vanes are fully closed.
The cam transmission constant K
L
must take into account only about 75% of the guide vane
strokeregardedversustherunnerbladestroke.
For turbines with fixed runner blades, such as Francis turbines, the runner blade movement is
zero,|=0.
7.2.4.2 HydroTurbineMatrixRepresentation
Considertheflowpositiveoutoftheturbineonbothsidesandrisingpressuretobepositivein
thesamewayasforpipeelements.ThepressureAhinEquations(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
i
q q A = A

(7.24)

1
0
i i
q q
+
A +A =

(7.25)

1 i i
h h h
+
A = A A

(7.26)
SubstitutingqandhfromEquations(7.19)and(7.26)intoEquation(7.14),yields
( ) ( ) ( )
1
1
1 1 0
2
q q n i i n
E q E E h h E n E p
|
|
+
| |
+ A + + A A + A + A A =
|
\ .

(7.27)
51
Furtherbysubstituting|,qandhbyEquations(7.23)and(7.26)intoEquation(7.15),yields
( ) ( ) ( )
1 1
1
1 1 0
2
q n i i n q i
E E h h E n E F y p E q
| | + +
| |
+ A A + A + A A + A =
|
\ .

(7.28)
Substituting|andqhbyEquations(7.23),(7.24),(7.25)and(7.26)intoEquation(7.27),yields
( ) ( ) ( )
1 1
1 1
1 1
2 2
n i n y n i i
Q h Q n Q Q F y Q h q
| | + +
A + A + A A = A

(7.29)

( ) ( ) ( )
1
1 1
1 1
2 2
n i n y n i i
Q h Q n Q Q F y Q h q
| | +
A + A + A A = A

(7.30)
ThecompleteturbinematrixmaybeestablishedbymeansofEquations(7.28),(7.29)and(7.30)
1
1
0
0
0
i
i
i
i
h
B C B Q q y
J L J K M h
B C B Q q n
p
+
+
(
(
( (
(
( (
( =
( (
(
( (

(
(

(7.31)
The turbine matrix will be established with vectors describing all inputs and variables which
havetobesolved;where
( )
1
1
2
n
B Q =

(7.32)

y
C Q Q F
| |
= +

(7.33)

n
Q Q =

(7.34)

( )
( )
3 1
2 1
n q n
q
E E Q
J
E
+
=
+

(7.35)

1
n
n
q
E
K Q
E
= +
+
(7.36)

1
y
q
E F
L Q Q F
E
| |
| |
= + +
+

(7.37)

1
1
q
M
E
=
+

(7.38)
52
7.2.5 Thesynchronousgeneratorandtheelectricgrid
As a first approach, thesynchronous generatormodel for a generatorinfinite bus systemmay
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
( )
g a g
p p T s F n = +

(7.39)
The transfer function F

g
is the adjusting load coefficient which describes an isolated load
conditionoragridsystem.Taisinertiamasstimeconstantofthegenerator(mechanicalinertia
constant).
Forafrequencyresponseanalysisfor<2rad/sec,thetransferfunction,describedinEquation
(7.39) may be simplified to be
g b
F e = , wheree
b
is the permanent load droop (voltage droop)
or selfregulating in the load as a function of the speed droop of the generator. The worst
scenarioisaresistiveload 0
g
F = .
ThesimplifiedElementMatrixyields:
( ) ( )
0 0 0
1
g
n
p s p s N
( ( (
=
( ( (

(7.40)
where
( )
a g
N T s F = + (7.41)
Forstudiesoftheinteractionofhydraulicsystemandpowersystems,itispossibletoestablish
a complete sub matrix for the electric grid including the torqueangle oscillations of the
synchronousmachine.Thismatrixmaybeincludedintheglobalmatrixofthesystem.
53
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
1
P
n P r ref
GR
F n R p y n
GC
+
+ =
(7.42)
ref r
p p p
e
+ =
(7.43)
where
1
1
p
C
T s
=
+

(7.44)
1
D
T D
T s
G
R T s
+
=
(7.45)
Assuming p
=0 in Equation (7.43) if there is no power adjustment signal from the water level
controller.Equations(7.42)and(7.43)inmatrixformyield
1
0 1 0
P
ref
n P
r
ref
n
GR
n
F R
p CG
p
y
(
+ (
(
(
(
=
(
(
(

(

(7.46)
Equation (7.46) may be simplified assuming P

ref
as a possible port of excitation of little
importanceinstabilityorvibrationanalysis.Hence,Equation(7.46)yields
1
P
n ref
n
GR
F n
y CG
( + (
( =
( (

(7.47)
54
7.2.6.2 PIDGovernor
The following equations describe the turbine governing system when the turbine governor is
assumedaPIDtype.
0 1 0
ref n
r
ref
n
n K E F
p
p
y
(
( (
(
=
( (
(

(

(7.48)
Where
1
p
E R
G
| |
= +
|
\ .
(7.49)
1
p
GR
F
GC
+
=
(7.50)
( )( )
1
1 1
y x
C
T s T s
=
+ +

(7.51)
i
p d
K
G K K s
s
= + +
(7.52)
Assumingp
r
=p
ref
,Equation(7.48)yields
| |
n ref
n
K F n
y
(
( =
(

(7.53)
It should be noted that the constant K

n
is used for choosing between open and closed loop
analysis.ItisassumedK
n
=0fortheopenloopanalysis,andK
n
=1forcloseloopanalysis.

55
7.3 CompositionoftheGlobalStructureMatrix
The Structure Matrix of a complex system could be built from Element and Local Matrixes.
Every element in the system is designed in order with an element number. The points where
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.EachofitsentriesA
ij
isequalto
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
for the hydro turbine set is built by incorporating Equations (7.31), (7.40) and (7.48) together.
Therearesixnodesconnectedtosixindependentvariablesforthecompleteturbineset.
Numbering the variables as h

R
=1, n=2, y=3, p
r
=4, p=5 and h
L
=6. The Structure Matrix may be
determinedas
( )
( )
( )
( )
( )
( )
0 0
0 0 0
0 0
0 0 0 1 0 0
0 0 0 1 0
0 0
R R
ref n
r ref
g
L L
h s q s B Q C B
n n K F E
y J K L M J
p p
p s p s N
h s q s B Q C B
( ( (
( (
(
( (
(
( (
(
=
( (
(
( (
(
( (
(
( (
(
( (

(7.54)

56
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)
( )
( )
( )
( )
1 1 1 1
3 3 1 1
h s q s T S
h s q s S T
( ( (
=
( (
(
(7.55)
TheElement2isthesurgetank,nodenumber2,connectedtonode3
( )
( ) ( )
2 2 2 2
3 3 2 2
0
F
h s T T S
q s h s S T
( ( (
=
( ( (

(7.56)
TheElement3isthetunnelconnectingtothesurgetankandtheturbine,(node3and4)
( )
( )
( )
( )
3 3 3 3
3 3 4 4
T S h s q s
S T h s q s
( ( (
=
( ( (
(7.57)
The Element 4 is the Structure Matrix for the Hydro Turbine Unit. The Group Structure Matrix
forahydroturbinesetisdescribedinChapter7.4.Numberingthenodesasfollowsh
R
=4,n=5,
y=6,p
r
=7,p=8andh
L
=9.
( )
( )
( )
( )
( )
( )
4 4
9 9
0 0
0 0 0
0 0
0 0 0 1 0 0
0 0 0 1 0
0 0
ref n
r ref
g
h s q s B Q C B
n n K F E
y J K L M J
p p
p s p s N
h s q s B Q C B
( ( (
( (
(
( (
(
( (
(
=
( (
(
( (
(
( (
(
( (
(
( (

(7.58)
57
TheElement5isthetailrace,connectingtonode9to10
( )
( )
( )
( )
5 5 9 9
5 5 10 10
T S h s q s
S T h s q s
( ( (
=
( ( (

(7.59)
The Global Structure Matrix expression for the Hydro Power Plant, shown in Figure 71,
accordingtotheindexingis
1 1
2 2 2 2
1 2 1 2 3 3 3
3 3 4
5 5 9
5 5
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0
0 0 0 0 0 0 1 0 0 0
0 0 0 0 0 0 1 0 0
0 0 0 0 0
0 0 0 0 0 0 0 0 0
F
n
r
T S
T T S h
S S T T T S h
S T B Q C B h
K F E n
J K L M J y
p
N p
B Q C B T S h
S T
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(

1
10
0
0
0
0
0
ref
ref
g
q
n
p
p
q
=
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (

(7.60)
Instabilitystudiesforhydropowerplants,thesystemresponsestopowerreferencesettingare
oftenignored.Hence,Equation(7.60)canbereducedas
1 1 1
2 2 2 2
1 2 1 2 3 3 3
3 3 4
5 5 9
5 5
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0
0 0 0 0 0 1 0 0
0 0 0 0
0 0 0 0 0 0 0 0
F
n ref
T S q
T T S h
S S T T T S h
S T B Q C B h
K F n n
J K L M J y
N p p
B Q C B T S h
S T
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (
10
0
g
q
(
(
(
(
(
(
(
(
(
(
(
(
(

(7.61)

58
Foranideallossless,Q
y
=1,Q
n
=0andq=constant.Equation(7.61)issimplifiedto
1 1
2 2 2 2
1 2 1 2 3 3 3
3 3 4
5 5 9
5 5
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0
0 0 0.5 0 0 0.5 0
0 0 0 0 0 0 0
0 0 0 1.5 0 1 1 1.5 0
0 0 0 0 0 1 0 0
0 0 0 0.5 0 1 0 0.5
0 0 0 0 0 0 0 0
F
n
T S
T T S h
S S T T T S h
S T C h
K F n
y
N p
T S h
S T
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (
( (

1
10
0
0
0
0
0
ref
g
q
n
p
q
(
(
(
(
(
(
(
(
(
(
(
(
(

(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
system is determined directly with s replaced by j. The frequency response of the system to
theexcitationcanbeobtainedbysolvingEquation(7.1)repeatedlywithaselectedwithinthe
frequencyrange
1
<>
2
.
Theinputsignalmaybeappliedtoanydeviceandtheoutputsarestudied.Aconstant,usually
1,mustbeassignedtooneofthecomponentsinvectorqinEquation(7.1)accordingtowhere
theexcitationisimposed.
The frequency responses to load disturbances can be calculated assuming p

g
= 1 and n
ref
= 0.
Theresponsen/p
g
isthemostimportantmeasuretostudythestabilityofthesystem.
The openloop speed response to speed setting can be evaluated by the criteria of Nyquist
Bode on openloop n n
ref
responses. The n n
ref
responses can be measured setting n
ref
= 1
andp
g
=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.
59
8 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.
60
8.2 FundamentalsofPowerSystemStability
The behaviour of a dynamic power system is described by a set of n first order nonlinear
differentialequations.Thissetofequationscanbewrittenintheform
( ) , , t = x f x u
(8.1)
Thevectorxisreferredasthestatevector,anditsentriesx
i
asstatevariables.Theinputsofthe
systemarerepresentedbythevectoru.Thederivateofastatevariablexwithrespecttotimet
isdenotedby x .Thestateofasystemistheminimumamountofinformationofthesystemat
anyinstantintime.Thesystemisautonomouswhenthederivativesofthestatevariablesare
notexplicitfunctionsoftime.Equation(8.1)issimplifiedto
( ) , = x f x u
(8.2)
To observe the output variables of the system, the state variables and the input variables can
beexpressedinthefollowingform
( ) , = y g x u

(8.3)
Theoutputsofthesystemarerepresentedbyvectory.Thevectorgisavectorofthenonlinear
functionsrelatingstateandinputvariablestooutputvariables.
Thesingularorequilibriumpointsarethecharacteristicofthedynamicsystembehaviour.The
singularpointmustsatisfythefollowingequation
( )
0
0 = f x

(8.4)
Alinearsystemhasonlyoneequilibriumstate.Theremaybemorethanoneequilibriumpoint
foranonlinearsystem.
The linearized forms of Equations (8.2) and (8.3), evaluated at the equilibrium point about
whichthesmalldisturbanceisanalyzed,are
A = A + A x A x B u (8.5)
A = A + A y C x D u (8.6)
The Laplace transforms of Ax and Ay are the Laplace transforms of the free and zerostate
components of the state and output vectors. These Laplace transforms have a component
dependenton the initial conditions and other component dependent on the inputs. The poles
ofAxandAyaretherootsoftheequation
| |
det 0 s = I A

(8.7)
Equation (8.7) is referred to as the characteristic equation of matrix A. The values of s that
satisfythecharacteristicequationareknownaseigenvaluesofthematrixA.
61
8.3 EigenvalueAnalysis
8.3.1 EigenvaluesandEigenvectors
A scalar parameter is called eigenvalue of a matrix A if there is a nonzero column vector
satisfying
u u A = (8.8)
AnontrivialsolutionofEquation(8.8)isgivenby
| |
det 0 = A I
(8.9)
For every eigenvalues

i
, the vector
i
which satisfies Equation (8.8) is called the right
eigenvectorassociatedwithAtheeigenvalue
i
.Therefore,
i i i
u u A =

(8.10)
Similarly, the left eigenvector associated with the eigenvalue

i
is thevector+
i
which satisfies
Equation(8.11)
i i i
+ + A=

(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
1
i i
+u =

(8.12)
8.3.2 EigenvalueAnalysis
Thetimedependentcharacteristicofamodecorrespondingtoaneigenvalue
i
isgivenbye
it
.
The eigenvalues of Equation (8.7) leads to the following conclusions, which are important for
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
represents a decaying mode. The larger its magnitude, the faster the decay. A positive
realeigenvaluerepresentsaperiodicstability.
E. Each conjugate pair of complex eigenvalues introduces to the response oscillatory
modes.Theoscillatorymodeisstableistherealpartoftheeigenvalueisnegative.

62
Acomplexeigenvalueisrepresentedby
j o e = (8.13)
Therealcomponentoftheeigenvaluesgivesthedampingandtheimaginarycomponentgives
thefrequencyofoscillation.
The damping ratio determines the rate of decay of the amplitude of the oscillation. The
dampingratioisgivenby
2 2
o
,
o e
=
+

(8.14)
Eigenvalue Analysis is performed using special linear power systems analysis programs. 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
shape represents the relative activity of the state variables when a particular mode is excited.
The mode shape represents an inherent feature of a linear dynamic system and does not
dependofthedisturbanceapplied.
Thesensitivityofaneigenvalue
i
totheelementsofthestatematrixisequaltotheproductof
the left eigenvector element and the right eigenvector element. Multiplying the right and left
eigenvectorsasfollows
The participation factor matrix combines the right and left eigenvalues as a measure of the
relationshipbetweenthestatevariablesandthemodes.
ki ki ik
p u + =

(8.15)
Coefficientsp
ki
arereferredtoastheparticipationfactors.Eachparticipationfactorisaproduct
of the kth element of the ith left and right eigenvectors. It quantifies the sensibility of the ith
eigenvalue to the kth diagonal element of the state matrix. Participation factor contains
informationabouttheobservabilityandcontrollability.
Participationfactorscanbeusedtodetermineenhancingsystemstability.
63
8.4 SmallSignalStabilityAnalysisofageneratorinfinitebussystem
Assumethatthegeneratorisconnectedtotheinfinitebusbar,withthegeneratorrepresented
by the synchronous machine classical model detailed in Chapter 4.3.1. The initial disturbed
conditions are
( )
0
0 0 t o o
+
A = = A = and 0 e A = . The incremental swing equation, Equations
(4.7)and(4.8),writteninthematrixformbecomes
'
1
2 2 2
0 0
N
D E
N
m
s
K K
t
m H H H
t
e
e
o o
e
cA (
( (
(
A (
c
( (
= + A (
(
( (
A cA
(

(
c

(8.16)
Thetransientsynchronizingpowercoefficient
' E
K bydefinitionis:
'
cos
e S
E
d
P E V
K
x
o
o
' c
= =
c

(8.17)
where
d d
x X X = + and
T S
X X X = + isthecombinedreactanceofthestepuptransformerand
theequivalentnetwork.
Theeigenvaluesofthisstatematrixcanbedeterminedfromsolving
'
2 '
det 0 2 2
2 2
D E
D E
s
s
K K
K K
H H
H H
e
e
(

(
= + + =
(

(8.18)
Equation(8.18)mayberewrittenasthestandardsecondorderdifferentialequation
2 2
2 0
nat nat
,e e + + =

(8.19)
Thus,theundampednaturalfrequency
nat
e ofrotorswingsforsmalloscillationsisgivenby
'
2
s
nat E
K
H
e
e =

(8.20)
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 1
2 2 2 2
D D
nat
E s
K K
H K H
,
e e
= =

(8.21)

64
Thedampednaturalfrequency
d
e ofrotorswingsis
2
1
d nat
e e , = (8.22)
Thecharacteristicequationhastworoots
1
and
2
givenby
2
1,2
1
nat nat
j ,e e , =

(8.23)
Theresultsoftheroots
1,2
dependsontheactualvaluesofK
E
,K
D
andH.TheinertiaconstantH
isconstantwhilebothK
D
andK
E
dependsonthegeneratorloading.
8.5 DynamicAnalysis
Thedynamicbehaviourofahydraulicpowerplantmaybeanalyzedeitherinthetimedomain
by using Method of the Characteristics or in the frequency domain. The bestknown methods
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.
The nonlinear friction losses and the nonlinear boundary conditions are linearized for solution
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
the surge shaft called massoscillations. The massoscillations can take several minutes
togetthesteadystate.
8.5.1 Naturalfrequencyanalysis
The analysis of transients in hydroelectric systems is subdivided in massoscillation response
andwaterhammerresponse.Ordinarydifferentialequationsdescribethemathematicalmodel
for massoscillation (or rigid water column) between the reservoir and the surge tank. Partial
differentialequationsshowthemathematicalmodelforwaterhammeralongthepenstockand
drafttube.[1520,56,58]
65
Thefirstnaturalfrequencyofthepenstockcorrespondstothefrequencycorrespondingtothe
fundamental period of the system, described in Equation (2.4). The first natural frequency of
thepenstockcanbewrittenas
0
4
a
f
L
=

(8.24)
The period of oscillation of the water surface in a simple surge tank is described in Equation
(2.65). The natural frequency of the massoscillation between the upstream reservoir and the
surgetankisgivenby
0
1 1
2
S
t
f
A L
g A
t
=

(8.25)
Thefrequencyresponseofasystemisthesteadystateresponseofthesystemtoasinusoidal
inputsignal.Thetransferfunctionofasystemcanbydeterminedexperimentallybyfrequency
responsetests.Thedesignofasysteminthefrequencyresponsemethodprovidesthedesigner
with control over the system bandwidth and over the effect of noise and disturbance on the
systemresponse.[43]
Theresponseofalineartimeinvariantsystemtosinusoidalinput ( ) ( ) sin r t A t e = isgivenby
( ) ( ) ( ) sin c t A G j t e e u e = + (

(8.26)
WherethetransferfunctionG(j)isobtainedbysubstitutingjforsintheexpressionforG(s).
ThemostcommongraphicalrepresentationofafrequencyresponsefunctionistheBodePlot.
The frequency response characteristics of the operational parameters and related them to
standard parameters can be examined in SIMPOW. Such characteristics provide useful insight
into the dynamic behaviour of the machine. To demonstrate the dynamic nonlinear effects, a
perturbation signal is applied to any input device in open loop conditions. The input and
outputsmaybeassociatedwiththenetworkequationsaswellasthedynamicdevices.[2]
66
9 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.
Thischapterpresentsthephysicalcharacteristicsandcapabilityofthecomponents ofthetest
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.
Infinite Bus
Transmission
line
Transformer Generator
Hydraulic
Turbine
Figure91:Singlemachineinfinitebuspowersystem
The network reactances of the test system, depicted in Figure 91 are in per unit on 50MVA,
24kV base. The network reactance values are based on the exercise 12.2 of the Chapter 12,
page 732, of the textbook Power System Stability and Control, [2]. Resistances are assumed
negligible. The system parameters and initial operating condition of the electrical network are
asfollows.
67
Table91:Systemparametersandoperatingconditions
Generator Ratedapparentpower 45.00+j15.00[MVA]
Ratedphasetophasevoltage 24.00Z36.01[kV]
Frequency 50.00[Hz]
Numberofpairsofpoles 6.00[]
InertiaTimeConstant 2.70[MWs/MVA]
Transformer Reactance 0.15[pu]
RatedVoltage 24.00/132.00[kV]
Transmissionline Reactance 0.50[pu]
InfiniteBus Ratedvoltage 0.995Z0.00[pu]
9.2 Modelsofthedifferentcomponents
9.2.1 PowerPlantModel
The hydroelectric power plant includes a canal to carry the water flow from the reservoir, a
surge tank, a penstock to convey the water to the powerhouse, a powerhouse for the turbine
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.
68
Table92:HydraulicPowerPlantModel
Tunnel Length 1307.00 [m]
Diameter 3.90 [m]
Frictionconstant,f 0.06 []
SurgeTank Diameter 5.00 [m]
Waterlevelmeasured 300.00 [m]
Penstock Length 435.00 [m]
Diameter 2.00 [m]
FrancisTurbine Ratedmechanicalpower 46.15 [MW]
Ratedspeed 500.00 [rpm]
Rateddischarge 25.00 [m
3
/s]
Ratedhead 200.00 [m]
Referencediameter 3.90 [m]
Hydraulicefficiency 0.96 []
Ratedefficiency 0.94 []
Tailrace/outlet Length 100.00 [m]
Diameter 3.90 [m]
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
The hydraulic turbine is a type Francis. In a reaction turbine having long pressure tunnels, the
dynamicresponseofbothupstreamanddownstreamwatercolumnsmustbestudiedinorder
toavoidexcessivepressureoroverspeedproblems.
Thehydraulicturbineisdesignedbymodelsofvaryingdegreesofdetailforthestabilityanalysis
of The Hydraulic Power Generating System. The hydraulic turbine parameters used in the
modellingandsimulationofthehydroturbinemodelsaregiveninTable93.
Table93:HydraulicTurbineParameters
Parameter Value
TurbineDamping K
D
0.500 []
GatePositionatRatedCondition 0.900 [pu]
GateOpeningatFullLoad y
FL
0.960 [pu]
GateOpeningatNoLoad y
NL
0.172 [pu]
NoloadLoad Q
NL
4.300 [m
3
/s]
ConstantProportionality A
t
1.250 []
WaterTimeConstantinPenstock T
W1
1.764 [s]
WaterTimeConstantinUpstreamTunnel T
W2
1.394 [s]
SurgeTankFillingTime T
s
157.079 [s]
69
The mathematical models of the hydraulic turbine developed, simulated and analyzed for the
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 Nonlinear Turbine Model with Surge Tank assuming Elastic Water Column
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
consists of one field winding, one damper winding in daxis and one damper winding in qaxis
with saturation included. The synchronous machine is modelled without automatic voltage
regulator. This synchronous machine model TYPE 2 is a standard component in the simulation
softwareSIMPOWthatisdescribedindetailintheSIMPOWUserManual.
Thesynchronousmachineparametersusedinthemodellingofthesynchronousgeneratorare
shown in Table 95. These parameters are based on typical values of standard parameters
detailedintheTable4.2oftheChapter4ofthetextbookPowerSystemStabilityandControl,
[2].Thereactancevaluesareinperunitusingthemachineratedvaluesasthebasevaluesand
thetimeconstantsareinseconds.
Table95:SynchronousmachineParameters
Parameter Value
StatorResistance R
a
0.003 [pu]
Directaxissynchronousreactance x
d
1.174 [pu]
Directaxistransientreactance x
d
0.300 [pu]
Directaxissubtransientreactance x
d
0.215 [pu]
Quadratureaxissynchronousreactance x
q
0.770 [pu]
Quadratureaxissubtransientreactance x
q
0.150 [pu]
Statorleakagereactance x
l
0.101 [pu]
Opencircuitdaxistransienttimeconstant T
d0
8.500 [sec]
Opencircuitdaxissubtransienttimeconstant T
d0
0.050 [sec]
Opencircuitqaxissubtransienttimeconstant T
q0
0.250 [sec]

70
9.2.4 GoverningSystemsforHydraulicTurbines
The turbine governing system is modelled by a general speedgoverning model TYPE SG1. The
physicalcharacteristicsandcapabilityoftheturbinegoverningsystemaredetailedinChapter5.
The governor model type SG1 is available in the DSLlibraries in the software SIMPOW that is
documentedintheSIMPOWUser Manual.Therotorspeed
r
issetastheinputspeedsignal.
The input signal for the permanent droop compensation is the pilot or main servomotor
positiony.
The turbine governing system parameters used on the test system are based on the governor
model detailed in the Figure 9.10 of the Chapter 9 of the textbook in Reference [2]. These
valuesareshowningivenastypicalvaluesinTable96.
Table96:TypicalValuesofparametersofturbinegoverningsystem
Parameter Value
Pilotvalvetimeconstant T
P
0.05 [s]
Servogain K
s
5.00 []
Mainservomotortimeconstant T
g
0.20 [s]
Permanentdroop R
P
0.04 []
Temporarydroop R
T
0.40 []
Resettime T
R
5.00 [sec]
Maximumgateopeningrate YPMAX 0.10 [pu/s]
Maximumgateclosingrate YPMIN 0.10 [pu/s]
Maximumgatepositionlimit YMAX 1.00 []
Minimumgatepositionlimit YMIN 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

71
9.4 PowerSystemStabilityAnalysisTheoreticalResults
Thecircuitmodelrepresentingtheinitialoperatingconditionwithalltheparametersexpressed
inperunitisshowninFigure93.Thesynchronousgeneratorismodelledbytheclassicalmodel
andalltheresistancesneglected.
Figure93:Circuitmodelofthetestsystem
Assumingthegeneratorvoltagetobereference,thegeneratorstatorcurrentis
*
0
0.90 0.30
0.9487 17.5760
1.00 36.01
g g
P jQ P jQ j
I
V V
| |
+
= = = = Z
|
|
Z
\ .

(9.1)
Theinternalvoltageforthetransientstateis:
( ) ( )
' '
0
1.00 36.01 0.3 0.95 17.58 1.1229 49.9234
g d
E V jX I j = + = Z + Z = Z

(9.2)
Thetransientsynchronizingpowercoefficient
' E
K is:
'
1.123 0.995
cos cos 49.92 0.7572
0.95
S
E
d
E V
K
x
o
'
= = =
(9.3)
Linearizedsystemequations,fromEquation(8.16),give
0.1852 0.1402 0.1852

100 0 0
N
D N
m
K
t
m
t
e
e
t o o
cA (
(
A ( ( (
c
= + A (
( ( (
A cA
(
(
c

(9.4)
Theeigenvaluesofthestatematrix,Equation(9.4),aregivenby
2
0.1852 0.1402
det 0.1852 14.0212 0
100
D
D
K
K
t
t
(
= + + =
(

(9.5)
72
Theundampednaturalfrequency,fromEquation(8.20),gives
100
0.1402 6.6372 rad/s 1.0563 Hz
2 2.7
nat
t
e = = =
-

(9.6)
Thedampingratio,fromEquation(8.21),is
1
0.0140
2 2 2.7 6.6372
D
D
K
K , = =
- -

(9.7)
ThetheoreticalresultsforthistestsystemfordifferentvaluesofK
D
aredetailedinTable98.
Table98:ResultsoftheSynchronousmachinerepresentedbyClassicalModel
Description Results
Eigenvalues, 0.00j6.6372 0.0463j6.6371 0.09266.6366
Dampingcoefficient,K
D
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
0
1200
0.8305
4 4*435
a
f
L
= = =

(9.8)
The natural frequency of the massoscillation between the upstream reservoir and the surge
tankisgivenby
0
2
1 1 1 1
0.010755
2 2
1307.00 2.50
9.81 3.90
S
t
f
A L
g A
t t
= = =
| |
|
\ .
(9.9)
73
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.
This chapter deals with the study of the system stability characteristics of the different hydro
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.
The dynamic performance of the different Hydro Power System models has been investigated
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].

74
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.
The parameters of the hydro power plant and operating conditions are detailed in Chapter 9.
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
Frequency Dominant
Real,[1/s] [Hz] States
1 23.0168 0.0000
A
D
2 17.5102 0.0000
3,4 0.4073 1.0538 0.06140 1.05383 AandAo
5 6.7770 0.0000
A
Q
6,7 3.7911 0.1221 0.98015 0.12205
8 1.2927 0.0000
9 0.0167 0.0000
10 0.1773 0.0000
A
f
11 1.0000 0.0000
Eigenvalue Damping
Ratio Imag,[Hz]
ItisseeninTable101that
Alleigenvalueshavenegativerealpart.
Thesystemisstable.

1
,
2
,
5
,
8
,
9
,
10
and
11
influencethedynamicsystemwithpuredamping,

3,4
and
6,7
influencethedynamicsimulationwithbothdampingandoscillation.
Theoscillatorymode
3,4
haveadampedfrequencyof1.0538Hz.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,AoandAhaveahighparticipationintheoscillatorymodeassociatedto
the eigenvalues
3,4
. The field flux linkage has a high participation in the nonoscillatory mode
represented by theeigenvalue
10
. The daxis flux linkage has a high participation on the non
oscillatory mode represented by the eigenvalue
1
. The qaxis flux linkage has a high
participationonthenonoscillatorymoderepresentedbytheeigenvalue
5
.
75
Table102:Participationmatrixforamodel1
1

2

3

4

5

6

7

8

9

10
Ao 0.00627 0.00037 0.54104 0.54104 0.02714 0.22094 0.22094 0.06553 0.00391 0.00088
Ae 0.00582 0.00290 0.53081 0.53081 0.01898 0.08866 0.08866 0.00244 0.00043
A
f
0.00575 0.00001 0.00751 0.00751 0.00036 0.00849 0.00849 0.00812 0.00016 0.98838
A
D
1.00647 0.00044 0.01323 0.01323 0.00085 0.00455 0.00455 0.00015 0.00000 0.00888
A
Q 0.00004 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.
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.
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

TIME SECONDS
65.00
65.50
66.00
66.50
67.00
67.50
68.00
68.50
A
N
G
L
E

(
d
e
g
r
e
e
s
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9994
0.9996
0.9998
1.0000
1.0002
1.0004
1.0006
1.0008
S
P
E
E
D

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
45.63
45.64
45.65
45.66
45.67
45.68
45.69
45.70
M
E
C
H
A
N
I
C

T
O
R
Q
U
E

(
M
N
m
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9024
0.9025
0.9026
0.9027
0.9028
0.9029
0.9030
G
A
T
E

O
P
E
N
I
N
G

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.90260
0.90262
0.90264
0.90266
0.90268
0.90270
0.90272
0.90274
F
L
O
W

R
A
T
E

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9994
0.9996
0.9998
1.0000
1.0002
1.0004
P
R
E
S
S
U
R
E

H
E
A
D

(
p
u
)
76
The electrical angle oscillates for a period of 7.80 seconds. The minimum value is 64.9139 at
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. The Mechanic Torque oscillates for a period of 8 seconds, between the values of
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
oscillates between the values of 47.08 MNm and 1.38 MNm. The Electrical Torque at steady
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.
Figure102:FrequencyresponseofthetransferfunctionfromGatepositiontoMechanicalPowerof
model1
10
-3
10
-2
10
-1
10
0
10
1
-150
-100
-50
0
50
M
W
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
M
W
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
77
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.
Figure103:FrequencyresponseofthetransferfunctionfromGatepositiontoMechanicalTorqueof
Model1
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.
Figure104:FrequencyresponseofthetransferfunctionoftheconduitsystemofModel1
10
-3
10
-2
10
-1
10
0
10
1
1
2
3
4
5
6
7
M
T
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-200
-150
-100
-50
0
Frequency [Hz]
M
T
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-50
-40
-30
-20
-10
0
10
H
E
A
D
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-180
-160
-140
-120
-100
-80
Frequency [Hz]
H
E
A
D
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
78
Figure 105 plots the magnitude and phase of the transfer function from gate position to
ElectricalPowerAngle.
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.
Figure106:Frequencyresponseofthetransferfunctionofthemechanicalhydraulicgovernorof
Model1

10
-3
10
-2
10
-1
10
0
10
1
-150
-100
-50
0
50
T
E
T
A
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
T
E
T
A
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-60
-40
-20
0
20
40
G
A
T
E
/
S
P
E
E
D

-

M
a
g
n
i
t
u
d
e
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
G
A
T
E
/
S
P
E
E
D

-

P
h
a
s
e
(
d
e
g
)
79
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.
The parameters of the Hydro Power Plant and operating conditions are detailed in Chapter 9.
This model was implemented in DSL Code Generator which is a tool used for making self
defineddynamicmodelsfordynamicsimulationinSIMPOW.
variablesthathavethehighestparticipationofthetestsystemwithaNonlinearTurbineModel
withoutSurgeTankassumingInelasticWaterColumn.
Table103:EigenvaluesandthestatevariableforModel2
Frequency Dominant
1 23.0319 0.0000
A
D
2 17.5091 0.0000

3,4 0.4203 1.0649
0.06270 1.06485 AandAo
5 6.7511 0.0000
A
Q
6,7 3.7861 0.1176
0.98148 0.11762
8 1.2895 0.0000
A q
9 0.0167 0.0000

10 0.1814 0.0000
A
f
11 1.0000 0.0000

Eigenvalue Damping
Ratio Imag,[Hz]
Thesystemisstable.

1
,
2
,
5
,
8
,
9
,
10
,and
11
influencethedynamicsystemwithpuredamping.

3,4
,and
6,7
Theoscillatorymode
3,4
withatimeconstantof1/0.4203s.Thiscorrespondstoadampingratio,of0.0627.
the eigenvalues
3,4
10
1
5
.
80
Table104:ParticipationfactorsmatrixofaturbineModel2
1

2

3

4

5

6

7

8

9

10
Ao 0.0061 0.0004 0.5412 0.5412 0.0291 0.2246 0.2246 0.0636 0.0036 0.0008
Ae 0.0057 0.0028 0.5316 0.5316 0.0203 0.0893 0.0893 0.0023 0.0004
A
f
0.0057 0.0000 0.0071 0.0071
0.0004 0.0084 0.0084 0.0077 0.0001 0.9886
A
D
1.0062 0.0004 0.0128 0.0128
0.0009 0.0045 0.0045 0.0001 0.0000 0.0087
A
Q
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.
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
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

TIME SECONDS
64.00
64.50
65.00
65.50
66.00
66.50
67.00
A
N
G
L
E

(
d
e
g
r
e
e
s
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9994
0.9996
0.9998
1.0000
1.0002
1.0004
1.0006
1.0008
S
P
E
E
D

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
45.04
45.05
45.06
45.07
45.08
45.09
45.10
M
E
C
H
A
N
I
C

T
O
R
Q
U
E

(
M
N
m
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9025
0.9026
0.9027
0.9028
0.9029
0.9030
G
A
T
E

O
P
E
N
I
N
G

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.89898
0.89900
0.89902
0.89904
0.89906
0.89908
0.89910
F
L
O
W

R
A
T
E

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9914
0.9916
0.9918
0.9920
0.9922
0.9924
P
R
E
S
S
U
R
E

H
E
A
D

(
p
u
)
81
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.
Nonlinear Turbine Model without Surge Tank assuming Inelastic Water Column are plotted in
Figure108.Theperturbationsignalisappliedtothegatepositioninopenloopconditionsand
thesignalmechanicalpowerismonitored.
Model2
10
-3
10
-2
10
-1
10
0
10
1
-200
-150
-100
-50
0
M
W
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
Frequency [Hz]
M
W
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
82
Model2
system of the Model 2. The transfer function of the conduit system relates the Pressure Head
togateposition.
Figure1010:FrequencyresponseofthetransferfunctionoftheconduitsystemofModel2
10
-3
10
-2
10
-1
10
0
10
1
1
2
3
4
5
6
7
M
T
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-200
-150
-100
-50
0
Frequency [Hz]
M
T
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-40
-30
-20
-10
0
10
H
E
A
D
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-180
-160
-140
-120
-100
Frequency [Hz]
H
E
A
D
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
83
Figure1011:FrequencyresponseofthetransferfunctionfromGatepositiontoElectricalAngleof
Model2
The magnitude and phase of the transfer function of the mechanicalhydraulic governor are
plottedinFigure1012.Thetransferfunctionofthegovernorrelatesthegatepositiontospeed
changes.
Figure1012:FrequencyResponseofthetransferfunctionofthemechanicalhydraulicgovernorof
Model2

10
-3
10
-2
10
-1
10
0
10
1
-200
-150
-100
-50
0
T
E
T
A
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
Frequency [Hz]
T
E
T
A
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-60
-40
-20
0
20
40
G
A
T
E
/
S
P
E
E
D

-

M
a
g
n
i
t
u
d
e
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
G
A
T
E
/
S
P
E
E
D

-

P
h
a
s
e
(
d
e
g
)
84
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. The Penstockis
modelledtakingintoaccounttheelasticwaterhammertheoryandthehydraulicfrictionlosses.
The nonlinear characteristics of the hydraulic turbine are neglected in this model. The block
diagramofthisconfigurationisshowninFigure64.
The parameters of the Hydro Power Plant and operating conditions are detailed in Chapter 9.
This model was implemented in DSL Code Generator which is a tool used for making self
defineddynamicmodelsfordynamicsimulationinSIMPOW.
withoutSurgeTankincludingElasticWaterColumnEffects.
Table105:EigenvaluesandthestatevariablefortheTurbineModel3
Frequency Dominant
1 23.0313 0.0000
A
D
2 17.5710 0.0000
A y
3,4 0.5188 1.0587
0.07775 1.05872 AandAo
5 6.8018 0.0000
A
Q
6,7 3.6763 0.2128
0.93977 0.21280 A y
8 1.3377 0.0000

9 0.0167 0.0000
A y
10 0.1814 0.0000
A
f
11 1.0000 0.0000

Eigenvalue Damping
Ratio Imag,[Hz]
Thesystemisstable.

1
,
2
,
5
,
8
,
9
,
10
,and
11

3,4
and
6,7
Theoscillatorymode
3,4
the eigenvalues
3,4
10
1
5
.
85
Table106:ParticipationfactorsmatrixofaTurbineModel3
1

2

3
,
4

5

6
,
7

8

9

10
Ao 0.00614 0.00076 0.55712 0.02676 0.22299 0.10336 0.00337 0.00081
Ae 0.00572 0.00584 0.54235 0.01463 0.08774 0.00410 0.00042
A
f
0.00572 0.00002 0.00742 0.00027
0.00784 0.01188 0.00013 0.98850
A
D
1.00644 0.00087 0.01325 0.00064
0.00436 0.00025 0.00000 0.00873
A
Q
0.00004 0.00013 0.03659 0.97540
0.06619 0.00365 0.00000 0.00192
withoutsurgetankrepresentedbymodel3showsthedynamicbehaviourofthepowerangle,
Figure1013:Faultsimulationresults:(a)angle,(b)speed,(c)mechanicaltorque,(d)gateposition,(e)
flowrateand(f)waterpressureofModel3
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
64.00
64.50
65.00
65.50
66.00
66.50
67.00
A
N
G
L
E

(
d
e
g
r
e
e
s
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9994
0.9996
0.9998
1.0000
1.0002
1.0004
1.0006
1.0008
S
P
E
E
D

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
45.04
45.05
45.06
45.07
45.08
45.09
45.10
M
E
C
H
A
N
I
C

T
O
R
Q
U
E

(
M
N
m
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9024
0.9025
0.9026
0.9027
0.9028
0.9029
0.9030
G
A
T
E

O
P
E
N
I
N
G

(
p
u
)
2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.89895
0.89900
0.89905
0.89910
0.89915
F
L
O
W

R
A
T
E

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9914
0.9916
0.9918
0.9920
0.9922
0.9924
P
R
E
S
S
U
R
E

H
E
A
D

(
p
u
)
86
stateis45.07MNm.
NonlinearTurbineModelwithoutSurgeTankincludingElasticWaterColumnEffectsareplotted
in Figure 108. The perturbation signal is applied to the gate position in open loop conditions
andthesignalmechanicalpowerismonitored.
Model3
10
-3
10
-2
10
-1
10
0
10
1
-200
-150
-100
-50
0
M
W
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
M
W
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
87
Model3
systemoftheModel3.Thetransferfunctionoftheconduitsystemrelatesthewaterpressure
togateposition.
Figure1016:FrequencyResponseofthetransferfunctionoftheconduitsystemofModel3
10
-3
10
-2
10
-1
10
0
10
1
1
2
3
4
5
6
7
M
T
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
M
T
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-40
-30
-20
-10
0
10
H
E
A
D
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-300
-250
-200
-150
-100
Frequency [Hz]
H
E
A
D
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
88
Model3
changes.
Model3

10
-3
10
-2
10
-1
10
0
10
1
-200
-150
-100
-50
0
T
E
T
A
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
T
E
T
A
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-60
-40
-20
0
20
40
G
A
T
E
/
S
P
E
E
D

-

M
a
g
n
i
t
u
d
e
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
G
A
T
E
/
S
P
E
E
D

-

P
h
a
s
e
(
d
e
g
)
89
10.4 MODEL 4: Nonlinear Turbine Model with Surge Tank assuming

InelasticWaterColumns
The hydraulic turbine in the model 4 is represented by a Nonlinear Turbine Model with Surge
TankassumingInelasticWaterColumnsinPenstockandTunnel,describedinChapter6.2.4.The
upstream tunnel and penstock are modelled assuming an incompressible fluid and a rigid
conduit where the travelling pressure wave effects are relatively insignificantly. The nonlinear
characteristics of hydraulic turbine are neglected in this model. The block diagram of this
configuration is shown in Figure 65. The parameters of the Hydro Power Plant and operating
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
Frequency Dominant
1 23.1362 0.0000

A
D
2 17.5026 0.0000

3,4 0.5102 1.1313
0.07160 1.13126 AandAo
5 6.5735 0.0000
A
Q
6,7 3.7469 0.0840
0.99021 0.08405 A y
8 1.1626 0.0000

9 0.0167 0.0000

10,11 0.0392 0.0091
0.56646 0.00909
12 0.2060 0.0000
A
f
13 1.0000 0.0000

Eigenvalue Damping
Ratio Imag,[Hz]
ItisseenTable107that
Thesystemisstable.

1
,
2
,
5
,
8
,
9
,
12
,and
13

3,4
,
6,7
and
10,11
Theoscillatorymode
3,4
the eigenvalues
3,4
12
1
5
.
90
Table108:ParticipationfactorsmatrixofahydraulicturbineModel4
1

2

3
,
4

5

6
,
7

8

9

10
,
11

12
Ao 0.00508 0.00041 0.54296 0.04179 0.27162 0.04706 0.00244 0.00032 0.00048
Ae 0.00465 0.00244 0.53682 0.02836 0.10359 0.00131 0.00035
A
f
0.00556 0.00001 0.00514 0.00049
0.00893 0.00540 0.00006 0.00004 0.98932
A
D
1.00424 0.00030 0.01042 0.00106
0.00451 0.00006 0.00785
A
Q
0.00005 0.00008 0.04992 0.98495
0.14922 0.00199 0.00000 0.00223
with surge tank represented by model 4 show the dynamic behaviour of the power angle,
Figure1019:Faultsimulationresults:(a)angle,(b)Speed,(c)MechanicalTorque,(d)GatePosition,
(e)FlowRateand(f)HeadPressureofModel4
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
57.00
57.50
58.00
58.50
59.00
59.50
A
N
G
L
E

(
d
e
g
r
e
e
s
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9994
0.9996
0.9998
1.0000
1.0002
1.0004
1.0006
1.0008
S
P
E
E
D

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
41.72
41.73
41.74
41.75
41.76
41.77
M
E
C
H
A
N
I
C

T
O
R
Q
U
E

(
M
N
m
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9026
0.9027
0.9028
0.9029
0.9030
0.9031
G
A
T
E

O
P
E
N
I
N
G

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.87808
0.87810
0.87812
0.87814
0.87816
F
L
O
W

R
A
T
E

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9456
0.9458
0.9460
0.9462
0.9464
P
R
E
S
S
U
R
E

H
E
A
D

(
p
u
)
91
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.
Model4
10
-3
10
-2
10
-1
10
0
10
1
-150
-100
-50
0
50
M
W
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
M
W
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
92
Model4
system of model 4. The transfer function of the conduit system relates the Pressure Head to
gateposition.
10
-3
10
-2
10
-1
10
0
10
1
-1
0
1
2
3
4
5
6
M
T
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-200
-150
-100
-50
0
Frequency [Hz]
M
T
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-25
-20
-15
-10
-5
0
5
10
H
E
A
D
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-180
-160
-140
-120
-100
Frequency [Hz]
H
E
A
D
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
93
Model4
changes.
Model4

10
-3
10
-2
10
-1
10
0
10
1
-150
-100
-50
0
50
T
E
T
A
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
T
E
T
A
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-60
-40
-20
0
20
40
G
A
T
E
/
S
P
E
E
D

-

M
a
g
n
i
t
u
d
e
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
G
A
T
E
/
S
P
E
E
D

-

P
h
a
s
e
(
d
e
g
)
94
10.5 MODEL5:NonlinearModelwithSurgeTankassumingElasticWater
ColumninPenstockandInelasticWaterColumninTunnel
The hydraulic turbine in the model 5 is represented by a Nonlinear Turbine Model with Surge
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
Frequency Dominant
1 23.1350 0.0000
A
D
2 17.5566 0.0000
A y
3,4 0.5933 1.1269
0.08350 1.12692 AandAo
5 6.6476 0.0000
A
Q
6,7 3.6479 0.1899
0.95044 0.18991 A y
8 1.3158 0.0000

9,10 0.0392 0.0091
0.56646 0.00909
11 0.0167 0.0000
A y
12 0.2059 0.0000
A
f
13 1.0000 0.0000

Eigenvalue Damping
Ratio Imag,[Hz]
Thesystemisstable.

1
,
2
,
5
,
8
,
11
,
12
,and
13

3,4
,
6,7
and
9,10
Theoscillatorymode
3,4
the eigenvalues
3,4
12
1
5
.
95
Table1010:ParticipationfactorsmatrixforModel5
1

2

3
,
4

5

6
,
7

8

9
,
10

11

12
Ao 0.00510 0.00083 0.55426 0.03864 0.21962 0.09201 0.00019 0.00225 0.00049
Ae 0.00469 0.00500 0.54516 0.02059 0.08119 0.00334 0.00035
A
f
0.00556 0.00001 0.00528 0.00034
0.00662 0.00887 0.00002 0.00006 0.98932
A
D
1.00443 0.00062 0.01068 0.00077
0.00349 0.00017 0.00785
A
Q
0.00005 0.00017 0.05161 0.95721
0.09945 0.00473 0.00000 0.00223
with surge tank represented by model 5 shows the dynamic behaviour of the power angle,

Figure1025:FaultsimulationResults:(a)angle,(b)Speed,(c)mechanicaltorque,(d)gateposition,(e)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
57.00
57.50
58.00
58.50
59.00
59.50
60.00
A
N
G
L
E

(
d
e
g
r
e
e
s
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9994
0.9996
0.9998
1.0000
1.0002
1.0004
1.0006
1.0008
S
P
E
E
D

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
41.73
41.74
41.75
41.76
41.77
M
E
C
H
A
N
I
C

T
O
R
Q
U
E

(
M
N
m
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9027
0.9028
0.9029
0.9030
G
A
T
E

O
P
E
N
I
N
G

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.87805
0.87810
0.87815
0.87820
F
L
O
W

R
A
T
E

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.94570
0.94580
0.94590
0.94600
0.94610
0.94620
0.94630
P
R
E
S
S
U
R
E

H
E
A
D

(
p
u
)
96
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.
Model5
10
-3
10
-2
10
-1
10
0
10
1
-150
-100
-50
0
50
M
W
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
M
W
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
97
Model5
system of model 5. The transfer function of the conduit system relates the Pressure Head to
gateposition.
10
-3
10
-2
10
-1
10
0
10
1
-1
0
1
2
3
4
5
6
M
T
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
M
T
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-40
-30
-20
-10
0
10
H
E
A
D
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-300
-250
-200
-150
-100
Frequency [Hz]
H
E
A
D
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
98
Model5
The magnitude ratio and phase of the transfer function of the mechanicalhydraulic governor
are plotted in Figure 1030. The transfer function of the governor relates the gate position to
speedchanges.
Model5

10
-3
10
-2
10
-1
10
0
10
1
-150
-100
-50
0
50
T
E
T
A
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
T
E
T
A
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-60
-40
-20
0
20
40
G
A
T
E
/
S
P
E
E
D

-

M
a
g
n
i
t
u
d
e
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
G
A
T
E
/
S
P
E
E
D

-

P
h
a
s
e
(
d
e
g
)
99
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
parameters of the Hydro Power Plant and operating conditions are detailed in Chapter 9. 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
Frequency Dominant
1 23.0473 0.0000

A
D
2 17.4271 0.0000
A y
3,4 0.2925 1.0890
0.04270 1.08896 AandAo
5 6.5781 0.0000
A
Q
6 5.2853 0.0000
A y
7 2.5995 0.0000
A y
8 0.0167 0.0000

9,10 0.0037 0.0108
0.05513 0.01076
11 0.1851 0.0000
A
f
12 1.0000 0.0000

Eigenvalue Damping
Ratio Imag,[Hz]
Thesystemisstable.

1
,
2
,
5
,
6
,
7
,
8
,
11
,and
12

3,4
and
9,10
Theoscillatorymode
3,4
the eigenvalues
3,4
11
1
5
.
100
Table1012:ParticipationfactorsmatrixofaModel6
1

2

3
,
4

5

6

7

8

9
,
10

11
Ao 0.00592 0.00018 0.51300 0.03538 0.02546 0.03901 0.00290 0.00038 0.00088
Ae 0.00543 0.00127 0.51623 0.04673 0.02780 0.00641 0.00041
A
f
0.00570 0.00000 0.00634 0.00092
0.00102 0.00203 0.00010 0.00004 0.98949
A
D
1.00554 0.00018 0.01184 0.00200
0.00124 0.00037 0.00000 0.00860
A
Q
0.00004 0.00003 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,
Figure1031:FaultsimulationResults:(a)angle,(b)speed,(c)mechanicaltorque,(d)gateposition,(e)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

TIME SECONDS
63.00
63.50
64.00
64.50
65.00
65.50
66.00
A
N
G
L
E

(
d
e
g
r
e
e
s
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9994
0.9996
0.9998
1.0000
1.0002
1.0004
1.0006
1.0008
S
P
E
E
D

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
44.65
44.66
44.66
44.67
44.67
M
E
C
H
A
N
I
C

T
O
R
Q
U
E

(
M
N
m
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9025
0.9026
0.9027
0.9028
0.9029
0.9030
G
A
T
E

O
P
E
N
I
N
G

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.89940
0.89950
0.89960
0.89970
0.89980
F
L
O
W

R
A
T
E

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
-0.0062990
-0.0062988
-0.0062986
-0.0062984
-0.0062982
-0.0062980
-0.0062978
P
R
E
S
S
U
R
E

H
E
A
D

(
p
u
)
101
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
between 0.9024 and 0.9030. The flow varies between 0.8993 and 0.8999. The values of Gate
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.
Model6

10
-3
10
-2
10
-1
10
0
10
1
-150
-100
-50
0
50
M
W
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
M
W
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
102
Model6
system of a Linear Turbine Model with Surge Tank assuming Inelastic Water Columns. The
transferfunctionofthewatersupplysystemrelatesthePressureHeadtogateposition.
10
-3
10
-2
10
-1
10
0
10
1
-8
-6
-4
-2
0
2
M
T
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
M
T
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-80
-60
-40
-20
0
H
E
A
D
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-300
-250
-200
-150
-100
Frequency [Hz]
H
E
A
D
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
103
Model6
The magnitude ratio and phase of the transfer function of the mechanicalhydraulic governor
are plotted in Figure 1036. The transfer function of the governor relates the gate position to
speedchanges.
Model6

10
-3
10
-2
10
-1
10
0
10
1
-150
-100
-50
0
50
T
E
T
A
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
T
E
T
A
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-60
-40
-20
0
20
40
G
A
T
E
/
S
P
E
E
D

-

M
a
g
n
i
t
u
d
e
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
G
A
T
E
/
S
P
E
E
D

-

P
h
a
s
e
(
d
e
g
)
104
10.7 MODEL 7: Linear Turbine Model with Surge Tank assuming Elastic
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
Frequency Dominant
1 23.0440 0.0000

A
D
2 17.5548 0.0000
A y
3,4 0.4674 1.0695
0.06939 1.06950 AandAo
5 6.7715 0.0000
A
Q
6,7 3.6698 0.1773
0.95688 0.17731 A y
8 1.2029 0.0000

9 0.0167 0.0000
A y
10,11 0.0037 0.0108
0.05504 0.01077
12 0.1846 0.0000
A
f
13 1.0000 0.0000

Eigenvalue Damping
Ratio Imag,[Hz]
Thesystemisstable.

1
,
2
,
5
,
8
,
9
,
12
,and
13

3,4
,
6,7
and
10,11
Theoscillatorymode
3,4
the eigenvalues
3,4
12
1
5
.
105
Table1014:ParticipationfactorsmatrixofaModel7
1

2

3
,
4

5

6
,
7

8

9

10
,
11

12
Ao 0.00599 0.00068 0.55228 0.02888 0.22085 0.07832 0.00280 0.00019 0.00094
Ae 0.00555 0.00502 0.53969 0.01667 0.08406 0.00245 0.00041
A
f
0.00570 0.00002 0.00704 0.00030
0.00789 0.01018 0.00010 0.00002 0.98890
A
D
1.00608 0.00073 0.01279 0.00072
0.00415 0.00014 0.00000 0.00862
A
Q
0.00004 0.00012 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,
Figure1037:FaultsimulationResults:(a)Powerangle,(b)Speed,(c)MechanicalTorque,(d)Gate
Position,(e)FlowRateand(f)HeadPressureofModel7

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
63.00
63.50
64.00
64.50
65.00
65.50
66.00
A
N
G
L
E

(
d
e
g
r
e
e
s
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9994
0.9996
0.9998
1.0000
1.0002
1.0004
1.0006
1.0008
S
P
E
E
D

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
44.71
44.72
44.73
44.74
44.75
M
E
C
H
A
N
I
C

T
O
R
Q
U
E

(
M
N
m
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.9025
0.9026
0.9027
0.9028
0.9029
0.9030
G
A
T
E

O
P
E
N
I
N
G

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
0.89990
0.89995
0.90000
0.90005
0.90010
F
L
O
W

R
A
T
E

(
p
u
)
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
TIME SECONDS
-0.00580
-0.00560
-0.00540
-0.00520
-0.00500
P
R
E
S
S
U
R
E

H
E
A
D

(
p
u
)
106
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
between 0.9025 and 0.9030. The flow varies between 0.8999 and 0.9001. The values of Gate
positionandFlowrateatsteadystateare0.9027,and0.9000,respectively.
LinearModelwithSurgeTankassumingElasticWaterColumninPenstockareplottedinFigure
1038. The perturbation signal is applied to the gate position in open loop conditions and the
signalmechanicalpowerismonitored.
Model7
10
-3
10
-2
10
-1
10
0
10
1
-150
-100
-50
0
50
M
W
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
M
W
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
107
Model7
system of a Linear Model with Surge Tank assuming Elastic Water Column in Penstock. The
transferfunctionoftheconduitsystemrelatesthePressureHeadtogateposition.
10
-3
10
-2
10
-1
10
0
10
1
-10
-5
0
5
10
M
T
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
M
T
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-40
-30
-20
-10
0
10
H
E
A
D
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-300
-250
-200
-150
-100
-50
Frequency [Hz]
H
E
A
D
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
108
Model7
changes.
Model7
10
-3
10
-2
10
-1
10
0
10
1
-150
-100
-50
0
50
T
E
T
A
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
T
E
T
A
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-60
-40
-20
0
20
40
G
A
T
E
/
S
P
E
E
D

-

M
a
g
n
i
t
u
d
e
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
G
A
T
E
/
S
P
E
E
D

-

P
h
a
s
e
(
d
e
g
)
109
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, acoustic calculations in gas pipes on Heidrun, cooling water system for Ormen Lange.
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.
110
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, turbine and PID control, are dragged and dropped from Function Palette >> LVTrans in
the block diagram. These elements must be wired after all the elements are inserted and
named.Eachelementmusthaveauniquename.Themainparametersofthehydrauliccircuit
aresummarizedinTable92.
Figure111:BlockDiagramoftheHydraulicSystemModel
Thedatamustbesetupforeachelementbeforethesimulationbegins.Thedataarestoredin
separate data files. LVTrans sets initial values for each element after running the program for
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.
111
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
the surge shaft called massoscillations. The massoscillations can take several minutes
togetthesteadystate.
Both the waterhammer and massoscillation responses are calculated by shutting down the
turbineintheinterfaceandblockdiagram.Thewaterhammerandmassoscillationsresponses
dependofthelengthoftheconduit.Theoscillationmodecouldbeanalyzedatthegraphofthe
PenstockandTurbineintheblockdiagram.
Figure112showstheoscillationsofthehydraulicpowerplantmodelledinLVTrans86_1.3.1_T.
TheparametersoftheHydroPowerPlantandoperatingconditionsaredetailedinChapter9.
Figure112:SimulationResults:(a)gateopeningposition,(b)PressureHeadoftheHydraulicTest
ModelinLVTrans86_1.3.1_T

100 105 110 115 120 125 130 135 140 145 150
0.92
0.94
0.96
0.98
1
1.02
Time (s)
G
a
t
e
100 105 110 115 120 125 130 135 140 145 150
0.9
1
1.1
1.2
1.3
1.4
Time (s)
P
r
e
s
s
u
r
e

H
e
a
d
112
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.
Figure113:FrequencyResponseofthetransferfunctionfromgatepositiontopressureheadofthe
hydraulicpowerplantsimulatedwithoutdroopinLVTrans8_1.1.2
Figure114showsthemagnitudeandphaseofthefrequencyresponseofthehydraulicpower
plantmodelledwithdroopinLVTrans8_1.1.2.
hydraulicpowerplantsimulatedwithdroopinLVTrans8_1.1.2
10
-3
10
-2
10
-1
10
0
10
1
-80
-60
-40
-20
0
20
H
e
a
d
/
G
a
t
e

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
H
e
a
d
/
G
a
t
e

-

P
h
a
s
e
(
d
e
g
)
10
-3
10
-2
10
-1
10
0
10
1
-80
-60
-40
-20
0
20
H
e
a
d
/
G
a
t
e

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
H
e
a
d
/
G
a
t
e

-

P
h
a
s
e
(
d
e
g
)
113
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.
The hydraulic system model without droop agrees well with the hydraulic system model with
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
oscillation between the reservoir 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. The effects of the droop has very little different on the
phase.
Theoscillationsathigherfrequenciesareduetotheelasticityoftheconduitsystem.
hydraulicpowerplantsimulatedwithoutandwithdroopinLVTrans8_1.1.2
Thefeedbackfromthegateopeningpositiontothecontrollerisincludedwhenthesimulation
isperformedwithdroop.Thedroopisusually4%to8%.

10
-3
10
-2
10
-1
10
0
10
1
-80
-60
-40
-20
0
20
H
e
a
d
/
G
a
t
e

-

M
a
g
n
i
t
u
d
e

(
d
B
)

10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
H
e
a
d
/
G
a
t
e

-

P
h
a
s
e
(
d
e
g
)
without droop
with droop
114
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).
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.
10
-3
10
-2
10
-1
10
0
10
1
-50
-40
-30
-20
-10
0
10
20
H
e
a
d
/
G
a
t
e

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
H
e
a
d
/
G
a
t
e

-

P
h
a
s
e
(
d
e
g
)
115
Figure117:ComparisonoftheFrequencyresponseofthehydraulicpowergeneratingsystem
simulatedwithoutdroopinLVTrans8_1.1.2andcomputedinLVTrans86_1.3.1_T
Comparisonofthemagnitudeandphaseoftherelationshipbetweenthewaterpressurehead
and gate opening position of the hydraulic power generating system simulated 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.
Figure118:ComparisonoftheFrequencyResponseofthehydraulicpowergeneratingsystem
simulatedwithdroopinLVTrans8_1.1.2andcomputedinLVTrans86_1.3.1_T
10
-3
10
-2
10
-1
10
0
10
1
-80
-60
-40
-20
0
20
H
e
a
d
/
G
a
t
e

-

M
a
g
n
i
t
u
d
e
(
d
B
)

10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
H
e
a
d
/
G
a
t
e

-

P
h
a
s
e
(
d
e
g
)
without droop
LVTrans86 1.3.1 T
10
-3
10
-2
10
-1
10
0
10
1
-80
-60
-40
-20
0
20
H
e
a
d
/
G
a
t
e

-

M
a
g
n
i
t
u
d
e
(
d
B
)

10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
H
e
a
d
/
G
a
t
e

-

P
h
a
s
e
(
d
e
g
)
with droop
LVTrans86 1.3.1 T
116
In Figure 117 and Figure 118, the magnitude of the transfer function of the hydraulic power
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
frequency corresponds to the natural frequency of the massoscillation between the reservoir
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.
117
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 in SIMPOW and LVTrans. The models
implemented in SIMPOW are compared and analyzed. After that, these models are compared
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.

118
12.2 SummaryoftheEigenvalueanalysiscomputedinSIMPOW
The dynamic performance of the models implemented in SIMPOW is investigated utilizing the
Eigenvalueanalysis. The Eigenvalue analysis of the models shows that the all themodels have
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
Damping Frequency
Ratio [Hz]
1 SimplifiedNonlinearTurbineModel 0.4073 j 6.6214 0.06140 1.05383
2 Nonlinear Turbine Model without Surge Tank assuming
InelasticWaterColumn
0.4203 j 6.6906 0.06270 1.06485
3 Nonlinear Turbine Model without Surge Tank including
ElasticWaterColumnEffect
0.5188 j 6.6521 0.07775 1.05872
4 Nonlinear Turbine Model with Surge Tank assuming
0.5102 j 7.1079 0.07160 1.13126
5 Nonlinear Turbine Model with Surge Tank assuming
ElasticWaterColumninPenstock
0.5933 j 7.0806 0.08350 1.12692
6 Linear Turbine Model with Surge Tank considering
InelasticWaterColumns,andturbinecharacteristics
0.2925 j 6.8421 0.04270 1.08896
7 Linear Turbine Model with Surge Tank including Elastic
WaterColumninPenstock,andturbinecharacteristics
0.4674 j 6.7198 0.06939 1.06950
Eigenvalue
Models
The error in the frequency of oscillation between the calculated eigenvalues with K
D
=0.50,
Table98,andthecomputedeigenvaluesofthedifferentmodelsareshowninTable122.The
errorcalculatedbetweenthecomputedmodelsandthecalculatedeigenvaluesislessthatthe
10%inalltherepresentations.
Table122:Errorbetweenthecalculatedeigenvaluesandthecomputedeigenvalues
Model Error,[%]
SimplifiedNonlinearTurbineModel 0.1450
NonlinearTurbineModelwithoutSurgeTank
assumingInelasticWaterColumn
0.9068
NonlinearTurbineModelwithoutSurgeTank
includingElasticWaterColumnEffect
0.3193
NonlinearTurbineModelwithSurgeTank
assumingInelasticWaterColumns
7.1986
NonlinearTurbineModelwithSurgeTank
assumingElasticWaterColumninPenstock
6.7817
LinearTurbineModelwithSurgeTankconsidering
InelasticWaterColumns,andturbinecharacteristics
3.1904
LinearTurbineModelwithSurgeTankincluding
ElasticWaterColumninPenstock,andturbinecharacteristics
1.3427

119
12.3 ComparisonoftheresultsofthemodelsinSIMPOW
Frequency response analysis is developed by Frequency Scanning in Linear Analysis in
DYNPOW in order to identify the transfer function and other parameters of the models of the
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. The open loop condition is obtained by assuming H=9999999 for the synchronous
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
simplified nonlinear turbine model (Model 1), the green dashed line represents the nonlinear
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.
Figure121:FaultSimulationResultsofhydraulicturbinerepresentedbymodel1,2and3
0 2 4 6 8 10
62
64
66
68
70
A
N
G
L
E

(
d
e
g
)

0 2 4 6 8 10
0.999
0.9995
1
1.0005
1.001
S
P
E
E
D

(
p
u
)

0 2 4 6 8 10
45
45.2
45.4
45.6
45.8
M
T

(
M
N
m
)

0 2 4 6 8 10
0.902
0.9025
0.903
0.9035
G
A
T
E

(
p
u
)

0 2 4 6 8 10
0.898
0.9
0.902
0.904
0.906
Time (s)
F
L
O
W

(
p
u
)

0 2 4 6 8 10
0.99
0.995
1
1.005
Time (s)
H
E
A
D

(
p
u
)

Model 1
Model 2
Model 3
120
It is seen from Figure 121, the angle and speed of the generator are almost the same for
models2and3,andisslightlydifferentformodel1.Itmeansthattheeffectofdifferenthydro
turbine models on the power system transient stability simulation is comparatively small. The
variablessuchasmechanicaltorque,fluxrate,andpressureheadthatarecloselyrelatedtothe
conduitsystemhavealittledifferencebetweenmodel1,andmodels2and3.Thespeedofthe
generator and gate position are almost the same for the three models. The gate opening
positioninthehydraulicpowersystemmodelchangesslightly.
The dynamic behaviour of Model 2 and Model3 are very similar. It can beconcluded that the
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.
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.
10
-3
10
-2
10
-1
10
0
10
1
-200
-150
-100
-50
0
50
M
W
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)

10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
M
W
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
Model 1
Model 2
Model 3
121
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
function features the first natural frequency of the penstock with a high amplitude at 0.7 Hz.
For higher frequencies, the transfer functions of model 1 and 2 cannot represent the conduit
dynamicswithanyacceptabledegreeofaccuracy.Thisisevidentwhenobservingthefrequency
responsebetween1and10Hz.
Figure124:FrequencyresponseofthetransferfunctionsoftheconduitsystemofModel1,2and3

10
-3
10
-2
10
-1
10
0
10
1
1
2
3
4
5
6
7
M
T
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)

10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
M
T
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)

Model 1
Model 2
Model 3
10
-3
10
-2
10
-1
10
0
10
1
-50
-40
-30
-20
-10
0
10
H
E
A
D
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)
10
-3
10
-2
10
-1
10
0
10
1
-300
-250
-200
-150
-100
-50
Frequency [Hz]
H
E
A
D
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)

Model 1
Model 2
Model 3
122
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, mechanical torque, gate position, flow rate and head pressure of the hydraulic system
modelsimulatedbyModel4,5,6and7canbeexaminedinFigure125.Thestabilityanalysisof
thehydraulicpowergeneratingsystemistestedbysimulatingathreephasefaulttogroundat
t=1s.Thisdisturbanceissimulatedforaperiodof5.0ms.Thesamplingperiodis50s.
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. The gate opening position in the hydraulic power system model changes slightly. The
0 2 4 6 8 10
55
60
65
70
A
N
G
L
E

(
d
e
g
)

0 2 4 6 8 10
0.999
0.9995
1
1.0005
1.001
S
P
E
E
D

(
p
u
)

0 2 4 6 8 10
41
42
43
44
45
M
T

(
M
N
m
)

0 2 4 6 8 10
0.902
0.9025
0.903
0.9035
G
A
T
E

(
p
u
)

0 2 4 6 8 10
0.86
0.88
0.9
0.92
Time (s)
F
L
O
W

(
p
u
)

0 2 4 6 8 10
-0.5
0
0.5
1
Time (s)
H
E
A
D

(
p
u
)

Model 4
Model 5
Model 6
Model 7
123
dynamic behaviour of the linear models, model 6 and 7, are very similar. It can be concluded
thatthesimulationresultsofthenonlinearmodelsareslightlydifferentofthelinearmodels.
Figure126:FrequencyresponseofthetransferfunctionfromgatepositiontoMechanicalPowerof
Model4,5,6and7
Figure 126 compares the gate positionturbine developed power relationship of the hydraulic
turbinemodel4,5,6and7.Themagnitudeofthetransferfunctionsofmodel4,5,6and7are
very similar in the entire bandwidth. Model 4, 5, 6 and 7 shows significant difference in the
phase.
Figure127:FrequencyResponseofthetransferfunctionfromGatepositiontoMechanicalTorqueof
asimplehydraulicturbinerepresentedbyModel4,5,6and7
10
-3
10
-2
10
-1
10
0
10
1
-150
-100
-50
0
50
M
W
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)

10
-3
10
-2
10
-1
10
0
10
1
-400
-350
-300
-250
-200
-150
-100
-50
0
Frequency [Hz]
M
W
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
Model 4
Model 5
Model 6
Model 7
10
-3
10
-2
10
-1
10
0
10
1
-10
-5
0
5
10
M
T
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)

10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
M
T
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
Model 4
Model 5
Model 6
Model 7
124
Figure127comparesthemagnitudeandphaseofthetransferfunctionrelatingthemechanical
torque to the gate position of the turbine represented by model 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.
Figure128:FrequencyresponseofthetransferfunctionoftheconduitsystemofModel4,5,6and7
Thefrequencyresponseofthetransferfunctionofheadpressureandgatepositionofasimple
hydraulic turbine with surge tank represented by model 4, 5, 6 and 7 are compared in Figure
128.Itcaneasilybenoticedthatthefrequencyresponsesoftheelasticmodels,model5and7,
aresimilareventhoughtheirparametricmodelsaresignificantlydifferent.Alargedifferencein
thephaseofmodel4and6areshown.
The transfer functions of model 6 and 7 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 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
and 6 cannot represent the conduit dynamics with any acceptable degree of accuracy. This is
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, which represent the nonlinear characteristic of the turbine, have to be extracted
fromtheHillCharts.

10
-3
10
-2
10
-1
10
0
10
1
-80
-60
-40
-20
0
20
H
E
A
D
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)

10
-3
10
-2
10
-1
10
0
10
1
-300
-250
-200
-150
-100
-50
Frequency [Hz]
H
E
A
D
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)

Model 4
Model 5
Model 6
Model 7
125
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
SIMPOW and model in LVTrans has a slightly difference in the magnitude at low frequencies.
However,significantdifferenceisobservedinthephase.
Figure129:FrequencyResponsesofthetransferfunctionoftheconduitsystemofHydraulicPower
SystemModelsimplementedinSIMPOWandLVTrans(withoutdroop)
10
-3
10
-2
10
-1
10
0
10
1
-100
-80
-60
-40
-20
0
20
H
E
A
D
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)

10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
H
E
A
D
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)
without droop
Model 4
Model 5
Model 6
Model 7
126
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.
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. The effect of the droop has very little influence on the phase. The oscillations at
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
functions of model 4, 5 and 7 in SIMPOW, and models in LVTrans, all of them, feature a high
amplitude at 0.8 Hz that is the natural frequency of the penstock. For higher frequencies, the
transferfunctionsofmodel4and6cannotrepresenttheconduitdynamicswithanyacceptable
degree of accuracy. This is evident when observing the frequency response between 1 and 10
Hz.
It is interesting to note that the magnitude of Model 7, which represents the most complete
modeluntilthisstage,andmodelinLVTransshowagoodfit.Model7featuresalowerpeakin
magnitudeat0.8Hz.Thiscouldbeduetotheinfluenceofthevaluesoftheturbinecoefficients
usedinthismodel.
10
-3
10
-2
10
-1
10
0
10
1
-100
-80
-60
-40
-20
0
20
H
E
A
D
/
G
A
T
E

-

M
a
g
n
i
t
u
d
e

(
d
B
)

10
-3
10
-2
10
-1
10
0
10
1
-400
-300
-200
-100
0
Frequency [Hz]
H
E
A
D
/
G
A
T
E

-

P
h
a
s
e
(
d
e
g
)

with droop
Model 4
Model 5
Model 6
Model 7
127
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
wide variety of hydraulic turbines have the general form with the torque falling off roughly in
proportiontospeedoverwiderangesofspeedandgateopeningposition.Theinfluenceofthe
turbinecoefficientsonthemodelaccuracyiscritical.
The hydraulic power generating system model, which is developed in this Thesis work for the
stability analysis, has been designed with a thirdorder synchronous generator model, a
mechanicalhydraulicgoverningsystemmodel,andahydraulicturbine.Thehydraulicturbineis
modelledwithvaryingdegreesofdetail.Theeffectofthehydraulicpowergeneratingsystemis
studiedbymeansofanalysisofthedynamicbehaviourofthemodelsofthehydraulicmachine
and conduit system about the steadystate operating condition following a small disturbance.
The models of the hydraulic machine and conduit system developed as part of this study and
implementedinSIMPOW,arelistedbelow:
recommendbyIEEE
128
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
determined via analytical approach for a classical simplified power system represented by the
secondordersynchronousmachinemodelconnectedtoaninfinitebusbar.Eigenvaluesarealso
determined in Dynpow of the software SIMPOW. The error in the frequency of oscillation
between the theoretical eigenvalues with K
D
=0.50 and the computed eigenvalues of the
differentmodelsislessthatthe10%inalltherepresentations.
Timedomainanalysis
Thedynamicbehaviourofthehydraulicturbinerepresentedbymodel1,2and3aresimilar.It
can be concluded that the simulation results of model 1 is slightly different. It means that the
effect of different hydro turbine models on the power system transient stability simulation is
comparativelysmall.Thevariablessuchasmechanicaltorque,fluxrateandpressureheadthat
arecloselyrelatedtotheconduitsystemhavealittledifferencebetweenmodel1,andmodel2
and 3. The changes in speed of the generator and gate position are almost the same for the
threemodels.
The changes in speed of the generator is very similar for model 4, 5, 6 and 7; this means that
theeffectofdifferentmodelsofhydroturbinewithsurgetankonthepowersystemtransient
stabilitysimulationiscomparativelysmall.Thespeedofthegeneratorandthegatepositionare
almost the same for the four models. The gate opening position in the test system changes
slightly. The dynamic behaviour of the linear models, model 6 and 7, is very similar. It can be
concludedthatthesimulationresultsofthenonlinearmodelsareslightlydifferentofthelinear
models.
FrequencyResponseAnalysis
Thereisagoodagreementbetweenthefrequencyresponseofthethreetransferfunctionsof
head pressure and gate position of a hydraulic turbine represented by model 1, 2 and 3. The
transfer function features a natural frequency with a high amplitude at 0.8 Hz. This frequency
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
between the reservoir and the surge tank. The transfer functions of model 4, 5 and 7 feature
thenaturalfrequencyofthepenstockwithahigheramplitudeat0.8Hz.Forhigherfrequencies,
the transfer function of model 4 and 6 cannot represent the conduit dynamics with any
acceptabledegreeofaccuracy.
129
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
created specifically for the detailed grid/waterway interaction. The frequency response of the
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
functions of model 4, 5 and 7 in SIMPOW, and models in LVTrans, all of them, feature a high
amplitude at 0.8 Hz that is the natural frequency of the penstock. For higher frequencies, 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, which represent the nonlinear characteristic of the turbine, have to be extracted
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.
130
14 ReferenceBibliography
1. Hydraulic turbine and turbine control models for system dynamic studies. Power
Systems,IEEETransactionson,1992.7(1):p.167179.
2. Kundur, P., N.J. Balu, and M.G. Lauby, Power system stability and control. 1994, New
York;London:McGrawHill.xxiii,1176p.
3. Parmakian,J.,Waterhammeranalysis.1963,[S.l.]:Dover.161p.
4. Wood, D.J., Waterhammer AnalysisEssential and Easy (and Efficient). Journal of
EnvironmentalEngineering,2005.131(8):p.11231131.
5. Bergant,A.,A.R.Simpson,andA.S.Tijsseling,Waterhammerwithcolumnseparation:A
historicalreview.JournalofFluidsandStructures,2006.22(2):p.135171.
6. Chaudhry,M.H.,Appliedhydraulictransients.2nded.ed.1987,NewYork;Wokingham:
VanNostrandReinhold.xvi,521p.
7. BOULOS,etal.,Hydraulictransientguidelinesforprotectingwaterdistributionsystems.
Vol.97.2005,Denver,CO,ETATSUNIS:AmericanWaterWorksAssociation.14.
8. Almeida, A.B.d. and E. Koelle, Fluid transients in pipe networks. 1992, Southampton;
Boston, Mass.; London; New York: Computational Mechanics Publications ; Elsevier
AppliedScience.
9. Zipparro, V.J., H. Hasen, and C.V. Davis, Davis' handbook of applied hydraulics. 4th ed.
ed.1993,NewYork:McGrawHill.
10. (TNSHP), T.N.o.S.h., Guide on How to Develop a Small Hydropower Plant, E.S.H.A..
ESHA,Editor.2004.
11. 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.
12. 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.
13. Romeo, E., C. Royo, and A. Monzn, Improved explicit equations for estimation of the
frictionfactorinroughandsmoothpipes.ChemicalEngineeringJournal,2002.86(3):p.
369374.
14. White, F.M., Fluid mechanics. 6th ed. ed. 2008, New York ; London: McGrawHill. xiii,
864p.
15. 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.
16. Brekke, H., A Stability Study on Hydro Power Plant Governing Including the influence
fromaQuasinonlineardampingofoscillatoryflowandfromtheturbinecharacteristics.
1984,NorgesTekniskeHgskole:Trondheim.
17. Bakken,B.H.,Technicalandeconomicaspectsofoperationofthermalandhydropower
systems.1997,Trondheim:NorwegianUniversityofScienceandTechnology.222.
18. Hongqing, F., et al., Basic Modeling and Simulation Tool for Analysis of Hydraulic
TransientsinHydroelectricPowerPlants.EnergyConversion,IEEETransactionson,2008.
23(3):p.834841.
131
19. Kishor, N., S.P. Singh, and A.S. Raghuvanshi, Dynamic simulations of hydro turbine and
itsstateestimationbasedLQcontrol.EnergyConversionandManagement,2006.47(18
19):p.31193137.
20. Nicolet, C., F. Avellan, and J.J. Simond, Hydroacoustic modelling and numerical
simulationofunsteadyoperationofhydroelectricsystems,inLMH,LME.2007,EPFL.
21. IEEE Guide for the Application of Turbine Governing Systems for Hydroelectric
GeneratingUnits.IEEEStd12072004,2004:p.0_1121.
22. Machowski,J.,J.Bialek,andD.J.Bumby,PowerSystemDynamics:StabilityandControl.
2nded.2008:JohnWiley&Sons,Ltd..
23. Voros, N.G., C.T. Kiranoudis, and Z.B. Maroulis, Shortcut design of small hydroelectric
plants.RenewableEnergy,2000.19(4):p.545563.
24. 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.
25. Haiyan, B., Y. Jiandong, and F. Liang. Study on Nonlinear Dynamical Model and Control
StrategyofTransientProcessinHydropowerStationwithFrancisTurbine.inPowerand
EnergyEngineeringConference,2009.APPEEC2009.AsiaPacific.2009.
26. FraileArdanuy,J.,etal.SpeedOptimisationModuleofaHydraulicFrancisturbinebased
on Artificial Neural Networks. Application to the Dynamic Analysis and Control of an
Adjustable Speed Hydro Plant. in Neural Networks, 2006. IJCNN '06. International Joint
Conferenceon.2006.
27. Prez, J.I., J.R. Wilhelmi, and L. Maroto, Adjustable speed operation of a hydropower
plant associated to an irrigation reservoir. Energy Conversion and Management, 2008.
49(11):p.29732978.
28. De Jaeger, E., et al., Hydro turbine model for system dynamic studies. Power Systems,
IEEETransactionson,1994.9(4):p.17091715.
29. 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.
30. Brekke, H. and X.X. Li. A new approach to the mathematic modelling of hydropower
governingsystems.inControl,1988.CONTROL88.,InternationalConferenceon.1988.
31. Nicolet,C.,etal.,NewToolfortheSimulationofTransientPhenomenainFrancisTurbine
PowerPlants.,inIAHRSymposium.2002,:.
32. Jiang, J., Design of an optimal robust governor for hydraulic turbine generating units.
EnergyConversion,IEEETransactionson,1995.10(1):p.188194.
33. 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.
34. Thorne, D.H. and E.F. Hill, Field Testing and Simulation of Hydraulic Turbine Governor
Performance. Power Apparatus and Systems, IEEE Transactions on, 1974. PAS93(4): p.
11831191.
35. Izena, A., et al. Practical hydraulic turbine model. in Power Engineering Society General
Meeting,2006.IEEE.2006.
36. 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.
37. Strah,B.,O.Kuljaca,andZ.Vukic,Speedandactivepowercontrolofhydroturbineunit.
132
38. 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.
39. IEEE Guide for Synchronous Generator Modeling Practices and Applications in Power
SystemStabilityAnalyses.IEEEStd11102002(RevisionofIEEEStd11101991),2003:p.
0_172.
40. Kanjun, Z., et al. Simulation Analysis of Dynamic Performance for HydroGenerator
UnderLossofExcitationCondition.inUniversitiesPowerEngineeringConference,2006.
UPEC'06.Proceedingsofthe41stInternational.2006.
41. Mouni,E.,S.Tnani,andG.Champenois.Comparativestudyofthreemodellingmethods
of synchronous generator. in IEEE Industrial Electronics, IECON 2006 32nd Annual
Conferenceon.2006.
42. Yongji,C.andY.Jilai.Alongtermdynamicsimulationmethodofpowersystembasedon
variables grouping and seceding grouping strategy. in Electric Utility Deregulation and
RestructuringandPowerTechnologies,2008.DRPT2008.ThirdInternationalConference
on.2008.
43. Saadat, H., Power system analysis. 2nd ed. ed. 2002, Boston ; London: McGrawHill
PrimisCustomPublishing.xix,712p.
44. Kazemi, A., M.R.J. Motlagh, and A.H. Naghshbandy, Application of a new multivariable
feedback linearization method for improvement of power systems transient stability.
InternationalJournalofElectricalPower&EnergySystems,2007.29(4):p.322328.
45. 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.
46. 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.
47. Ramey, D.G. and J.W. Skooglund, Detailed Hydrogovernor Representation for System
StabilityStudies.PowerApparatusandSystems,IEEETransactionson,1970.PAS89(1):
p.106112.
48. Report, I.C., Dynamic Models for Steam and Hydro Turbines in Power System Studies.
PowerApparatusandSystems,IEEETransactionson,1973.PAS92(6):p.19041915.
49. Culberg,J.L.,M.Negnevitsky,andK.M.Muttaqi.Hydroturbinegovernorcontrol:theory,
techniquesandlimitations.inAUPEC'06.2006.Melbourne.
50. Kishor,N.,R.P.Saini,andS.P.Singh,Areviewonhydropowerplantmodelsandcontrol.
RenewableandSustainableEnergyReviews,2007.11(5):p.776796.
51. 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.
52. Sanathanan, C.K., Accurate Low Order Model for Hydraulic TurbinePenstock. Energy
Conversion,IEEETransactionson,1987.EC2(2):p.196200.
53. Hagihara, S., et al., Stability of a Hydraulic Turbine Generating Unit Controlled by P.I.D.
Governor. Power Apparatus and Systems, IEEE Transactions on, 1979. PAS98(6): p.
22942298.
54. Smith, J.R., R. McLean, and J.F. Robbie, Assessment of hydroturbine models for power
systems studies. Generation, Transmission and Distribution, IEE Proceedings C, 1983.
130(1):p.17.
133
55. Kishor,N.,R.P.Saini,andS.P.Singh.Simulationofreducedorderhydroturbinemodelsto
study its hydraulic transient characteristics. in 9th International Multitopic Conference,
IEEEINMIC2005.2005.
56. Nicolet, C., et al., HighOrder Modeling of Hydraulic Power Plant in Islanded Power
Network.PowerSystems,IEEETransactionson,2007.22(4):p.18701880.
57. 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.
58. Nicolet, C., et al. On the hydroelectric stability of an islanded power network. in Power
EngineeringSocietyGeneralMeeting,2006.IEEE.2006.
59. Wozniak, L. and T.L. Filbert, Speed loop cancellation governor for hydrogenerators. I.
Development.EnergyConversion,IEEETransactionson,1988.3(1):p.8590.
60. Pannatier,Y.,etal.,DynamicBehaviorofa2VariableSpeedPumpTurbinePowerPlant,
inICEM'08XVIII18thInternationalConferenceonElectricalMachines.2008,:Vilamoura,
Algarve,Portugal.
61. 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.
62. 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.
63. Hannett, L.N., J.W. Feltes, and B. Fardanesh, Field tests to validate hydro turbine
governor model structure and parameters. Power Systems, IEEE Transactions on, 1994.
9(4):p.17441751.
64. Vournas,C.D.,Secondorderhydraulicturbinemodelsformultimachinestabilitystudies.
65. 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.
66. Persson,J.,UsingLinearAnalysistofindEigenvaluesandEigenvectorsinPowerSystems.
2002:Madrid,Spain.
67. Rogers, G., Demystifying power system oscillations. Computer Applications in Power,
IEEE,1996.9(3):p.3035.
68. Svingen, B., Manual LVTrans for versjon 8_1.2.4. 2007, SINTEF Energy Research:
Trondheim.p.66.

Hydro Turbine and Governor Modelling: Luz Alexandra Lucero Tenorio

Загружено:

Сведения о документе

Исходное описание:

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Hydro Turbine and Governor Modelling: Luz Alexandra Lucero Tenorio

Загружено:

Авторское право:

Доступные форматы

June 2010

Kjetil Uhlen, ELKRAFT

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

Appropriate representations of the hydraulic turbine and waterconduit system aredeveloped

Dependingon the site,the conduitsystem is composed of a waterdiverting structure; acanal

The theoretical period T

The velocity is instantaneously changed due to the valve is

The force resulting from the shear stress t

The substitution of the values of T

Theconduitwallsarerigidandthe liquidiscompressible.Thismeans thataflow change

produced by the turbine. The relation between mechanical power P

Normally, a simplifying assumption made is that

which are in rad/sec. In some literature, per unit speed

nor the internal emf E

As the pilot valve and servomotor time constant T

( ) 2.3 1.0 0.15

( ) 5.00 1.00 0.50

A hydroelectric power plant may be considered as a system consisting of several subsystems:

The turbine governing system is modelled by a mechanicalhydraulic turbine governing model

The head losses h

7 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

The runner blade servomotor has a time constant T

The transfer function F

Equation (7.46) may be simplified assuming P

It should be noted that the constant K

Numbering the variables as h

The frequency responses to load disturbances can be calculated assuming p

For every eigenvalues

Similarly, the left eigenvector associated with the eigenvalue

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

Theresponseofalineartimeinvariantsystemtosinusoidalinput ( ) ( ) sin r t A t e = isgivenby

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

0.1852 0.1402 0.1852

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

10.2 MODEL 2: Nonlinear Turbine Model without Surge Tank assuming

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

10.3 MODEL 3: Nonlinear Turbine Model without Surge Tank including

10.4 MODEL 4: Nonlinear Turbine Model with Surge Tank assuming

10.6 MODEL 6: Linear Turbine Model with Surge Tank assuming

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

LVTrans8_1.1.2 is a version created specifically for the detailed grid/waterway interaction.

12.4 Comparison of the simulation results computed in SIMPOW and

Вам также может понравиться