NERC Wind

Arnold Schwarzenegger
Governor
INTERMITTENCY ANALYSIS PROJECT:
APPENDIX B
IMPACT OF INTERMITTENT
GENERATION ON OPERATION OF
CALIFORNIA POWER GRID

Prepared For:
California Energy Commission
Public Interest Energy Research
Program
Prepared By:
GE Energy Consulting
Xinggang Bai
Kara Clark
Gary A. J ordan
Nicholas W. Miller
Richard J . Piwko

P
I
E
R

F
I
N
A
L

P
R
O
J
E
C
T

R
E
P
O
R
T

J uly 2007
CEC-500-2007-081-APB

Prepared By:
GE Energy Consulting
Richard J . Piwko
1 River Road
Building 2, Room 644
Schenectady, NY 12345
Commission Contract No. 500-02-004
Commission Work Authorization No: MR-017

Prepared For:
Public Interest Energy Research (PIER) Program
California Energy Commission

Dora Yen Nakafuji, Ph.D.
Contract Manager

Gerald Braun
Program Area Lead
PIER Renewables

Ken Koyama
Acting Office Manager
Energy Generation Research Office

Martha Krebs
Deputy Director
ENERGY RESEARCH & DEVELOPMENT DIVISION

B. B. Blevins
Executive Director
DISCLAIMER
This report was prepared as the result of work sponsored by the California Energy Commission. It does not necessarily represent the views of the
Energy Commission, its employees, or the State of California. The Energy Commission, the State of California, its employees, contractors and
subcontractors make no warrant, express or implied, and assume no legal liability for the information in this report; nor does any party represent
that the uses of this information will not infringe upon privately owned rights. This report has not been approved or disapproved by the California
Energy Commission nor has the California Energy Commission passed upon the accuracy or adequacy of the information in this report.

i
Table of Contents
ListofFigures ................................................................................................................................... v
ListofTables.................................................................................................................................... xi
ExecutiveSummary..........................................................................................................................1
StudyOverview...........................................................................................................................1
Conclusions...................................................................................................................................3
Recommendations .......................................................................................................................3
GenerationResourceAdequacy...........................................................................................4
TransmissionInfrastructure..................................................................................................8
RenewableGenerationTechnology,Policy,andPractice.................................................9
Closure...................................................................................... Error!Bookmarknotdefined.
1.0Introduction...............................................................................................................................11
1.1.Challenges.............................................................................................................................11
1.2.Background...........................................................................................................................11
1.3.IntermittentGenerationDefinition ...................................................................................12
1.4.OverviewofProjectObjectives,TasksandParticipants ................................................12
1.5.Participants............................................................................................................................15
2.0 StudyApproach........................................................................................................................17
2.1.StudyScenarios ....................................................................................................................17
2.2.TypesofAnalysis.................................................................................................................20
2.3.Data........................................................................................................................................21
2.4.Terminology .........................................................................................................................22
3.0 StatisticalAnalysis ....................................................................................................................23
3.1.TemporalandSpatialPatterns...........................................................................................23
3.1.1. DailyandSeasonalVariations .............................................................................23
3.1.2. SpatialVariations ...................................................................................................28
3.1.3. YearlyVariationandPenetrationRelativetoSystemLoad.............................29
2006and2010Penetration..............................................................................................33
2020YearlyVariationandPenetration.........................................................................36
3.2.HourlyVariability................................................................................................................38
3.2.1. 2006VariabilityRelativetoLoadLevel ..............................................................40
3.2.2. 2010Variability.......................................................................................................41
RelativetoLoadLevel ....................................................................................................43
RelativetoTimeofDay .................................................................................................45
ii
Sustained3HourChanges ............................................................................................48
ExtremeChanges.............................................................................................................50
3.2.3. 2020VariabilityRelativetoLoadLevelandTimeofDay...............................54
3.3.IntraHourVariability.........................................................................................................57
3.3.1. SelectedPeriods......................................................................................................58
3.3.2. IntraHourVariabilityDefinitions.......................................................................59
3.3.3. 2010XVariabilityatLightLoad...........................................................................62
3.4.VariabilitySummary...........................................................................................................65
3.5.HourlyForecastError .........................................................................................................67
3.5.1. DayAheadForecastError ....................................................................................67
3.5.2. HourAheadForecastError..................................................................................75
3.6.Summary...............................................................................................................................77
4.0 ProductionSimulationAnalysis .............................................................................................79
4.1.1. GeneralDatabaseCreation...................................................................................80
4.1.2. ScenarioDescription..............................................................................................81
4.2.Economics .............................................................................................................................82
4.3.IntermittentRenewableForecasting .................................................................................90
4.4.Operations.............................................................................................................................95
4.4.1. CaliforniaHydroelectricOperation ..................................................................103
4.4.2. CombinedCycleOperation................................................................................106
4.4.3. RampRateandRangeofOperation..................................................................108
4.4.4. Sensitivities ...........................................................................................................114
4.5.Emissions ............................................................................................................................121
4.6.TransmissionPathLoading..............................................................................................122
4.7.Observations.......................................................................................................................126
5.0QuasiSteadyStateAnalysis..................................................................................................129
5.1.OverviewofMethod .........................................................................................................129
5.1.1. StudyPeriods........................................................................................................130
5.1.2. InputData .............................................................................................................131
5.1.3. BoundaryConditions ..........................................................................................132
5.2.SystemPerformanceExamples........................................................................................133
5.2.1. JulyMorningLoadIncrease ...............................................................................133
5.2.2. MayNightLowLoadLevel ...............................................................................141
IncreaseManeuverableGeneration............................................................................144
iii
TemporaryCurtailmentofWindGeneration ...........................................................148
RemoveLargeStepsfromLoadProfile......................................................................150
5.2.3. JuneEveningLoadDecrease..............................................................................153
IncorporateHourlyWindForecast.............................................................................157
TemporaryWindRampRateLimit ............................................................................159
5.3.SummaryofResults...........................................................................................................161
6.0 OperationalImplicationsandMitigationMethods ...........................................................163
6.1.Validation............................................................................................................................163
6.2.StatisticalResultsandOperationalFlexibility...............................................................169
6.2.1. OverallRequirements..........................................................................................170
6.2.2. LightLoadRequirements ...................................................................................172
6.2.3. Extremes................................................................................................................174
6.3.OperationalFlexibility ......................................................................................................175
6.3.1. Forecasting............................................................................................................175
ImplicationsofIgnoringForecasts..............................................................................176
6.3.2. UnitCommitmentandScheduleFlexibility.....................................................177
HydroelectricGenerationShift....................................................................................177
AvailableDispatchRange............................................................................................178
6.3.3. LoadFollowing....................................................................................................179
ImpactofPumpsonLoadFollowing.........................................................................181
ImpliedCostsofLoadFollowing ...............................................................................182
6.3.4. Regulation.............................................................................................................183
ImpactofPumpsonRegulation..................................................................................183
CPS2Discussion ............................................................................................................185
ImpliedCostsofRegulation ........................................................................................186
6.4.MitigationMethods ...........................................................................................................186
6.4.1. UnitCommitmentandScheduleFlexibility.....................................................186
6.4.2. LoadFollowing....................................................................................................187
6.4.3. Regulation.............................................................................................................187
7.0ConclusionsandRecommendations ....................................................................................188
7.1.ObservationsbyTimeFrame ...........................................................................................188
7.1.1. DayAheadandOverallOperation...................................................................188
7.1.2. HourlyScheduleFlexibility................................................................................188
7.1.3. 5MinuteLoadFollowingandEconomicDispatch.........................................189
iv
7.1.4. 1MinuteRegulation............................................................................................190
7.2.Conclusions.........................................................................................................................190
7.3.Recommendations .............................................................................................................190
7.3.1. GenerationResourceAdequacy.........................................................................191
7.3.2. TransmissionInfrastructure ...............................................................................195
7.3.3. RenewableGenerationTechnology,Policy,andPractice ..............................196
7.4.Closure ................................................................................................................................197
8.0 References ................................................................................................................................198
Glossary..........................................................................................................................................199
AppendixA.SummaryofWindProjectsbyScenario.............................................................203
AppendixB.SummaryofSolarProjectsbyScenario ..............................................................209
AppendixC.ApplicationofSolarData.....................................................................................221
Pleasecitethisreportasfollows:
Richard Piwko et. al. 2007. Intermittency Analysis Project: Appendix B: Impact of Intermittent
GenerationonOperationofCaliforniaPowerGrid.CaliforniaEnergyCommission,PIERRenewable
EnergyTechnologiesProgram.CEC5002007081APB.
v
List of Figures
Figure1.FlowchartofIntermittencyAnalysisProject. ...................................................................... 14
Figure2.TimeScalesforGridPlanningandOperations................................................................... 16
Figure3.AverageSystemwideDailyLoad,Wind,Solar,andNetLoadProfilesofJuly2003..... 24
Figure4.AllSystemwideDailyLoad,Wind,andSolarProfilesforJuly2003. .............................. 25
Figure5.AllSystemwideDailySolarProfilesforJuly2003. ............................................................. 25
Figure6.AverageSystemwideDailyLoad,Wind,Solar,andNetLoadProfilesofJanuary2002.
............................................................................................................................................................ 26
Figure7.AllSystemwideDailyLoadandWindProfilesforJanuary2002. ................................... 27
Figure8.AllSystemwideDailySolarProfilesforJanuary2002. ...................................................... 27
Figure9.AllIndividualCaliforniaWindPlantProfilesforJuly21,2003........................................ 28
Figure10.AllIndividualTehachapiRegionWindPlantProfilesforJuly21,2003........................ 29
Figure11.Load,Wind,andSolarDurationCurvesfor2010XScenario. ......................................... 30
Figure12.2010HourlyLoadandNetLoadDurationCurvesfor3Years. ..................................... 31
Figure13.DetailofLoadandNetLoadDurationCurvesfor2010XScenario. .............................. 32
Figure14.2010XWindProductionandPenetrationDurationCurves. ........................................... 33
Figure15.2010XSolarProductionandPenetrationDurationCurves. ............................................ 34
Figure16.2006and2010HourlyAverageWindPenetrationbyDecile. ......................................... 35
Figure17.2006and2010HourlyAverageSolarPenetrationbyDecile........................................... 36
Figure18.2020HourlyLoadandNetLoadDurationCurves. ......................................................... 37
Figure19.2020HourlyAverageWindandSolarPenetrationbyDecile. ........................................ 38
Figure20.HourlyProfilesand1HourDeltasforanExampleJuly2002Day. ............................... 40
Figure21.2010XHourlyLoadandNetLoadDeltaStockChartbyDecile..................................... 44
Figure22.2006and2010StandardDeviationofHourlyLoadandNetLoadDeltas. ................... 45
Figure23.2010XAverageHourlyWindandSolarPenetrationbyHourofDay. .......................... 46
Figure24.2010XHourlyLoadandNetLoadDeltaStockChartbyHourofDay. ........................ 47
Figure25.2010XJulyHourlyLoadandNetLoadDeltaStockChartbyHourofDay. ................ 47
Figure26.2010XJanuaryHourlyLoadandNetLoadDeltasStockChartbyHourofDay. ........ 48
vi
Figure27.2010X3HourLoadandNetLoadDeltasStockChartbyHouroftheDay. ................ 49
Figure28.2010XJanuary3HourLoadandNetLoadDeltasStockChartbyHouroftheDay... 50
Figure29.20103HourPositiveLoadandNetLoadDeltaDurationCurvesfor2004.................. 51
Figure30.20103HourNegativeWindDeltaDurationCurvesfor2004StudyYear.................... 52
Figure31.20101HourNegativeLoadandNetLoadDeltaDurationCurvesfor2004................ 53
Figure32.20101HourPositiveWindDeltaDurationCurvesfor2004StudyYear...................... 54
Figure33.2020HourlyLoadandNetLoadDeltasStockChartbyDecile...................................... 55
Figure34.2020JanuaryHourlyLoadandNetLoadDeltasStockChartbyHourofDay............ 56
Figure35.2010LoadDurationCurveswith3HourPeriodsofInterestIdentified. ...................... 58
Figure36.1MinuteProfilesDuring3HourPeriodofExampleJuly2003Day. ............................ 59
Figure37.ExampleforLoadFollowingandRegulationMetricDefinition.................................... 60
Figure38.LoadFollowingRequirementforJuly2003Example3HourPeriod............................ 61
Figure39.RegulationRequirementforJuly2003Example3HourPeriod..................................... 62
Figure40.2010X5MinuteDeltaDurationCurvesforLightLoad(10thDecile). .......................... 63
Figure41.2010X1MinuteDeltaDurationCurvesforLightLoad(10thDecile). .......................... 64
Figure42.2010XSubHourlyDeltaWindDurationCurvesforLightLoad(10thDecile)............ 65
Figure43.2010Load,Wind,andSolarForecastsandActualsDuringanExampleJulyWeek. .. 68
Figure44.2010Load,Wind,Solar,andNetLoadForecastErrorsDuringExampleJulyWeek. . 69
Figure45.2010XLoad,Wind,Solar,andNetLoadDayAheadForecastErrorDurationCurves.
............................................................................................................................................................ 70
Figure46.2006LoadandNetLoadDayAheadForecastErrorStockChartbyDecile. ............... 74
Figure47.2010XLoadandNetLoadDayAheadForecastErrorStockChartbyDecile.............. 75
Figure48.2010XLoad,Wind,Solar,andNetLoadHourAheadForecastErrorDurationCurves.
............................................................................................................................................................ 76
Figure49.2010XLoadandNetLoadHourAheadForecastErrorStockChartbyDecile............ 77
Figure50.SpotPriceDurationCurvefor2002Shapes(#1). .............................................................. 83
vii
Figure55.SpotPriceDurationCurvefor2003Shapes....................................................................... 86
Figure56.SpotPriceDurationCurvefor2004Shapes....................................................................... 86
Figure57.SpotPriceDurationCurvefor2004Shapes(zoom). ........................................................ 87
Figure58.WECCOperatingCostReductionsDuetoRenewables($M)......................................... 87
Figure59.WECCOperatingCostReductionsDuetoRenewables($/MWh). ................................ 88
Figure60.LoadPaymentReductionsDuetoRenewables($M). ...................................................... 89
Figure61.LoadPaymentReductionsDuetoRenewables($/MWh)................................................ 89
Figure62.NonRenewableGeneratorRevenueReductionsDuetoRenewables($M). ................ 90
Figure63.NonRenewableGeneratorRevenueReductionsDuetoRenewables($/MWh).......... 90
Figure64.ImpactofIntermittentForecastonSpotPrice(2010T). .................................................... 92
Figure65.ImpactofIntermittentForecastonSpotPrice(2010X)..................................................... 92
Figure66.ImpactofIntermittentForecastonSpotPrice(2010Xand2020). ................................... 93
Figure67.TotalOperatingCostImpactofIntermittentForecasting................................................ 94
Figure68.CaliforniaGeneratorRevenueReductionsbyType(2010T). .......................................... 95
Figure69.WECCGeneratorRevenueReductionsbyType(2010T). ............................................... 95
Figure70.CaliforniaEnergyChangeDuetoRenewables(2010T). .................................................. 96
Figure71.WECCEnergyChangeDuetoRenewables(2010T)......................................................... 96
Figure72.AnnualDurationCurvesCaliforniaRenewableGeneration. ...................................... 97
Figure73.AnnualDurationCurvesCaliforniaWindandSolarGeneration. .............................. 98
Figure74.AnnualDurationCurvesCaliforniaNuclear,SteamandGasTurbines. ................... 98
Figure75.AnnualDurationCurvesCaliforniaHydroGeneration. .............................................. 99
Figure76.OneWeekChangeinCaliforniaHydroOperation(2010T). ......................................... 100
Figure77.AnnualChangeinCaliforniaHydroOperation(2010Tand2010X). ........................... 100
Figure78.AnnualChangeinWECCHydroOperation(2010T). .................................................... 101
Figure79.AnnualHistogramofHydroShift(2010T). ..................................................................... 101
Figure80.AnnualDurationCurvesCaliforniaPumpedStorageHydroOperation. ................ 102
Figure81.AnnualDurationCurvesCaliforniaCombinedCycleGeneration. ........................... 102
Figure82.AnnualDurationCurvesCaliforniaImports. ............................................................... 103
viii
Figure83.CaliforniaHistoricalHydroOperationSampleMayWeek. ....................................... 104
Figure84.CaliforniaHistoricalHydroOperationMay. ............................................................... 104
Figure85.CaliforniaHistoricalHydroOperationJune. ............................................................... 105
Figure86.CaliforniaHistoricalHydroOperationJuly. ................................................................ 105
Figure87.CaliforniaHydroHistoricalMonthlyDurationCurves. ............................................... 106
Figure88.2005CEMSDataDeltaEnergyCenter........................................................................... 107
Figure89.2005CEMSDataHaynesGeneratingStation. .............................................................. 107
Figure90.2005CEMSDataValleyGeneratingStation. ................................................................ 108
Figure91.CommitmentWeekofMay10
th
. ..................................................................................... 109
Figure92.DispatchWeekofMay10
th
. ............................................................................................. 109
Figure93.RampRateandRangeCapabilityWeekofMay10
th
. .................................................. 110
Figure94.RampRateDownCapabilityWeekofMay10
th
........................................................... 110
Figure95.RampRateDownCapacityWithoutConventionalHydroWeekofMay10
th
......... 111
Figure96.RampRateDownCapabilityVersusCaliforniaLoad(2010X). .................................... 111
Figure97.RangeDownCapabilityVersusCaliforniaLoad(2010X). ............................................ 112
Figure98.RampRateUpCapabilityVersusCaliforniaLoad(2010X). ......................................... 112
Figure99.RangeUpCapabilityVersusCaliforniaLoad(2010X). .................................................. 113
Figure100.AnnualRampRateDownCapability. ............................................................................ 113
Figure101.AnnualRampRateDownCapability(zoom). .............................................................. 114
Figure102.OperationofHelmsPumpedStorageHydro(2010X).................................................. 114
Figure103.HistoricalHydroDailyMinimum,Maximum,andRange,2006. .............................. 115
Figure104.DailyRangeOfHydroOperation,Summer2004and2006. ....................................... 116
Figure105.ConstrainedVersusBaseHydroforaSampleMayWeek(2010T). ........................... 116
Figure106.AnnualDurationCurveCaliforniaHydroGeneration(2010T)............................... 117
Figure107.RampDownCapacityWithConstrainedHydro,2010T. ............................................ 117
Figure108.RampDownCapacityWithConstrainedHydro,2010T(zoom)................................ 118
Figure 109. Annual Duration Curve California PSH Operation with Constrained Hydro
(2010T). ............................................................................................................................................ 119
Figure110.ComparisonofHelmsOperation,Summer2004and2006. ........................................ 119
ix
Figure111.SampleWeekofPumpingOperation,January2006. ................................................... 120
Figure112.HistoricalPumpingOperation,2006. ............................................................................. 120
Figure113.CaliforniaEmissionReductionsDuetoRenewables(2010T). .................................... 121
Figure114.WECCEmissionReductionsDuetorenewables(2010T). ........................................... 121
Figure115.CaliforniaEmissionReductionsDuetoNewWindandSolarGeneration(2010T). 122
Figure116.WECCEmissionReductionsDuetoNewWindandSolarGeneration(2010T). ..... 122
Figure117.TransmissionFlowDurationCurvesPath15:SouthofLosBanos. ........................ 123
Figure118.TransmissionFlowDurationCurvesPath21:ArizonatoCalifornia ..................... 123
Figure119.TransmissionFlowDurationCurvePath46:WestofColoradoRiver. .................. 124
Figure 120. Transmission Flow Duration Curves Total SCIT (Southern California Import
Transmission). ................................................................................................................................ 124
Figure121.AnnualDurationCurvePath66:COI(2010T). .......................................................... 125
Figure122.Path66:COIFlowsforOneWeekinMay(2010T). ...................................................... 125
Figure123.AnnualDurationCurveforHourlyFlowChangesonPath66:COI(2010T)........... 126
Figure124.TotalCaliforniaLoadDuringtheJulyMorningQSSStudyPeriod. .......................... 134
Figure125.TotalCaliforniaWindGenerationDuringtheJulyMorningQSSStudyPeriod. .... 135
Figure126.TotalCaliforniaSolarGenerationDuringtheJulyMorningQSSStudyPeriod. ..... 135
Figure127.ManeuverabilityVariablesDuringtheJulyMorningQSSStudyPeriod.................. 137
Figure128.QSSPerformanceVariablesDuringtheJulyMorningQSSStudyPeriod. ............... 138
Figure129.EconomicDispatchUnitChangeDuringtheJulyMorningQSSStudyPeriod. ...... 139
Figure 130. Impact of Intermittent Variability on Regulation Duty During July QSS Study
Period............................................................................................................................................... 140
Figure131.ImpactofIntermittentVariabilityonEconomicDispatchandLoadFollowingDuty
DuringJulyQSSStudyPeriod. .................................................................................................... 140
Figure132.TotalCaliforniaLoadDuringtheMayNightQSSStudyPeriod. .............................. 141
Figure133.TotalCaliforniaWindGenerationDuringtheMayNightQSSStudyPeriod.......... 142
Figure134.ManeuverabilityVariablesDuringtheMayNightQSSStudyPeriod. ..................... 143
Figure135.PerformanceVariablesDuringtheMayNightQSSStudyPeriod. ............................ 144
Figure136.ManeuverabilityVariablesforaMayNightwithMoreCombinedCyclePlants. .. 145
Figure137.PerformanceVariablesforaMayNightwithMoreCombinedCyclePlants.......... 146
x
Figure138.ImpactofWindVariabilityonRegulationDutyDuringMayQSSStudyPeriod.... 147
Figure139.ImpactofWindVariabilityonEconomicDispatchandLoadFollowingDutyDuring
MayQSSStudyPeriod. ................................................................................................................. 147
Figure140.TemporaryCurtailmentofWindGeneration................................................................ 149
Figure 141. Impact of Temporary Curtailment of Wind Generation on Economic Dispatch and
LoadFollowingDutyDuringMayQSSStudyPeriod. ........................................................... 149
Figure142.ImpactofTemporaryCurtailmentofWindGenerationonRegulationDutyDuring
MayQSSStudyPeriod. ................................................................................................................. 150
Figure143.ModifiedLoadProfileWithoutLargeSwitchingEvents............................................. 151
Figure144.SelectedCAISO1MinuteLoadDatafromMay2002,2003,and2004. ..................... 151
Figure145.ManeuverabilityVariablesforaMayNightwiththeModifiedLoadProfile. ......... 152
Figure146.PerformanceVariablesforaMayNightwiththeModifiedLoadProfile. ................ 153
Figure147.TotalCaliforniaLoadDuringtheJuneEveningQSSStudyPeriod. .......................... 154
Figure148.TotalCaliforniaWindGenerationDuringtheJuneEveningQSSStudyPeriod. .... 154
Figure149.TotalCaliforniaSolarGenerationDuringtheJuneEveningQSSStudyPeriod. ..... 155
Figure150.ManeuverabilityVariablesDuringtheJuneEveningQSSStudyPeriod. ................. 156
Figure151.PerformanceVariablesDuringtheJuneEveningQSSStudyPeriod......................... 156
Figure 152. Impact of Including Hourly Wind Forecast on Regulation Duty During the June
EveningQSSStudyPeriod. .......................................................................................................... 158
Figure 153. Impact of Including Hourly Wind Forecast on Economic Dispatch and Load
FollowingDuringtheJuneEveningQSSStudyPeriod. .......................................................... 158
Figure154.TemporaryCapwithRampUpRateLimitonWindGeneration. ............................ 160
Figure155.ComparisonofRegulationDuringtheJuneEveningQSSStudyPeriod. ................. 160
Figure 156. Comparison of Economic Dispatch and LoadFollowing During the June Evening
QSSStudyPeriod. .......................................................................................................................... 161
Figure157.ScheduleandLoadFollowingforaSampleDayofCaliforniaOperation................ 164
Figure158.HistoricalInterchangeforaSampleJulyWeek. ........................................................... 165
Figure159.InstantaneousRangeofInterchangeforSampleJulyWeek. ...................................... 165
Figure160.RegulationandInterchangeforaSampleDayofCaliforniaOperation. .................. 166
Figure161.HistoricalACEData. ......................................................................................................... 167
Figure162.PseudoACEfromQSSSimulations. .............................................................................. 168
xi
Figure163.HistoricalLoadandProcuredRegulationDatafor2003. ............................................ 169
Figure164.StandardDeviationsofHourlyDeltas. .......................................................................... 171
Figure165.AnnualStandardDeviationChanges. ............................................................................ 172
Figure166.StandardDeviationsforOneHourDeltasatLightLoad. .......................................... 173
Figure167.LightLoadStandardDeviationChanges....................................................................... 174
Figure 168. Comparison of Historical Forecast and Actual Load and Simulated DayAhead
ForecastandActualWind. ........................................................................................................... 176
Figure 169. 2010X Load and Net Load DayAhead Forecast Error Ignoring Wind and Solar
ForecastStockChartbyDecile. .................................................................................................... 177
Figure170.IntermittentGenerationImpactonHydroOperation.................................................. 178
Figure171.CommittedGenerationRangeandMaximumHourlyNetLoadChange. ............... 179
Figure 172. Committed Generation Ramp Rate Capability and Expected Load Following Duty.
.......................................................................................................................................................... 180
Figure 173. Expanded View of Committed Generation Ramp Rate Capability and Expected
LoadFollowingDuty..................................................................................................................... 180
Figure174.ImpactofWindVariabilitywithPumpsinLoadProfile............................................. 181
Figure175.ImpactofWindVariabilityWithoutPumpsinLoadProfile ...................................... 182
Figure176.MayNight:ImpactofWindVariabilitywithPumpSteps.......................................... 184
Figure177.MayNight:ImpactofWindVariabilityWithoutPumpSteps ................................... 184
Figure178.ExampleConcentratingSolarProjectProfileforaMayDay. ..................................... 222
Figure179.ExampleStirlingSolarProjectProfileforaMayDay. ................................................. 223
Figure180.ExamplePVSolarZIPCodeProfilesforaMayDay.................................................... 224
Figure181.ComparisonofOriginal15MinuteDataandFinalProfilewith1MinuteVariability.
.......................................................................................................................................................... 226
Figure182.IrradiationDataUsedasSourceof1MinuteVariabilityinExample. ...................... 226
xii
List of Tables
Table1.RenewableGenerationMixforFourStudyScenarios. ........................................................ 18
Table2.WindandSolarGenerationinCalifornia. ............................................................................. 19
Table3.LocationsofWindandSolarResourcesforScenario2010T. .............................................. 19
Table4.LocationsofWindandSolarResourcesforScenario2010X. .............................................. 19
Table5.2006HourlyLoadStatistics(MW). ......................................................................................... 41
Table6.2006HourlyNetLoadandIntermittentRenewableStatistics(MWor%). ...................... 41
Table7.2010HourlyLoadStatistics(MW). ......................................................................................... 42
Table8.2010THourlyNetLoadandIntermittentRenewableStatistics(MWor%)..................... 42
Table9.2010XHourlyNetLoadandIntermittentRenewableStatistics(MWor%). ................... 43
Table10.2020HourlyLoadStatistics(MW). ....................................................................................... 57
Table11.2020HourlyNetLoadandIntermittentRenewableStatistics(MWor%). .................... 57
Table12.StatisticsonLoadFollowingRequirementforJuly2003Example3HourPeriod. ...... 61
Table13.StatisticsonRegulationRequirementforJuly2003Example3HourPeriod................. 62
Table14.Summaryof2006and2010FullYearStatisticalAnalysis. ............................................... 66
Table15.Summaryof2006and2010LightLoad(10
th
Decile)StatisticalAnalysis........................ 66
Table16.Summaryof2010and2020HourlyStatisticalAnalysis. ................................................... 67
Table17.2006HourlyLoadForecastStatistics(MW). ....................................................................... 70
Table18.2006HourlyNetLoad,Wind,andSolarForecastStatistics(MW). ................................. 71
Table19.2010HourlyLoadForecastStatistics(MW). ....................................................................... 71
Table20.2010THourlyNetLoad,Wind,andSolarForecastStatistics(MW). ............................... 72
Table21.2010XHourlyNetLoad,Wind,andSolarForecastStatistics(MW)................................ 72
Table22.2010XDayAheadForecastErrorStandardDeviation. ..................................................... 73
Table23.2010XDayAheadForecastErrorEnergy. ........................................................................... 73
Table24.2010XHourAheadForecastErrorStandardDeviation. ................................................... 76
Table25.2010XHourAheadForecastErrorEnergy. ......................................................................... 76
Table26.ProductionSimulationScenarioDescription. ..................................................................... 82
xiii
Table27.AverageVariableOperatingCostReductionsperMWhofRenewableEnergy(2010T).
.......................................................................................................................................................... 126
Table28.AverageLoadPaymentReductionsperMWhofRenewableEnergy(2010T)............. 127
Table 29. Average NonRenewable Generator Revenue Reduction per MWh of Renewable
Energy(2010T)................................................................................................................................ 127
Table30.AnnualLoadPaymentReductionsfromIntermittentGeneration(2010T). ................. 127
Table 31. Average Annual Emission Reductions in WECC per MWh of Renewable Generation
(2010T). ............................................................................................................................................ 127
Table32.CharacteristicsofQSSStudyPeriods. ................................................................................ 130
Table33.WeightedAverageRampRateDatabyUnitType........................................................... 132
Table34.MinimumGenerationOutputLevelbyUnitType. ......................................................... 132
Table35.TotalandLightLoadChangeinFlexibilityRequirements. ............................................ 170
Table36.TotalVariabilityfromStatisticalAnalysis. ........................................................................ 174
Table37.LoadFollowingStatisticsofQSSPumpandWindSensitivityCases.......................... 182
Table38.LoadFollowingStatisticsofLightLoadConditionsforPumpandWindSensitivity182
Table39.RegulationStatisticsofQSSPumpandWindSensitivityCases .................................... 184
Table40.RegulationStatisticsofLightLoadConditionsforPumpandWindSensitivity ........ 184
Table41.WindProjectsIncludedin2006StudyScenario. .............................................................. 203
Table42.IncrementalWindProjectsAddedfor2010TStudyScenario......................................... 204
Table43.IncrementalWindProjectsAddedfor2010XStudyScenario......................................... 205
Table44.IncrementalWindProjectsAddedfor2020StudyScenario. .......................................... 207
Table45.SolarProjectsIncludedin2006StudyScenario. ............................................................... 209
Table46.IncrementalSolarProjectsAddedfor2010TStudyScenario.......................................... 209
Table47.IncrementalSolarProjectsAddedfor2010XStudyScenario. ........................................ 213
Table48.IncrementalSolarProjectsAddedfor2020StudyScenario. ........................................... 213
Table49.StandardDeviationofthe15MinuteVariabilityintheCaliforniaPVData. ............... 225
Table 50. Standard Deviation of the 15Minute Variability in the Golden, CO, Irradiation Data.
.......................................................................................................................................................... 225
xiv

1
Executive Summary
Californiahasoneofthemostdiverseelectricitysupplysystemsinthenationwithalarge
potentialtogenerateelectricityfromrenewablesources,suchaswind,geothermal,biomass,
hydroelectricandsolar.Withprogressiverenewablepolicies,thechallengefacingthestatewill
behowbesttointegrateandmanagerenewableenergyresourceswithtraditionalgeneration
whileensuringareliableelectricitysystem.
Study Overview
TheIntermittencyAnalysisProjectwastailoredtopresentastatewideperspectiveofthe
transmissioninfrastructureandservicesneededtoaccommodatetherenewablepenetration
levelsdefinedinthestatesrenewableenergypolicy.TheIntermittencyAnalysisProjectwas
technicalandwasintendedtoprovidea2020perspectiveonpotentialoperationalneedsand
impactstomeetfuturegrowthanddemand.
ThisreportdocumentsthelaststageofthemultistageIntermittencyAnalysisProject.
Precedingstagesincluded:
Evaluationofpast,present,andfuturewindturbinetechnologies,andtheireffecton
transmissionsystemoperationandperformance,byBEWEngineering,Inc.
Assessmentofworldwideexperiencewithintegratinglargepenetrationsofwind
energy,byKevinPorterofExeterAssociates,Inc.
Developmentoffuturerenewableenergyscenariosandevaluationoftheeffectson
transmissionreliability,byDavisPowerConsultants.
ThefuturerenewablegenerationscenariosdevelopedbyDavisPowerConsultantswerecritical
inputstotheanalysisdocumentedinthisreport.DataprovidedfromDavisPowerConsultants
included:
Detailedlistsofindividualrenewablegeneratingplantsandtheirsite/ratingforeach
scenariostudied,consistentwithCaliforniasRenewablesPortfolioStandardgoals
andlocationsofrenewableresources,and
Powerflowdatasetswithconventionalgeneration,renewablegeneration,and
transmissionsystembuildoutsforeachscenario,consistentwiththeprojected
Californiapeakloadlevel.
TheIntermittencyAnalysisProjectconsideredfourtypesofrenewablegenerationtomeet
Californiasrenewableenergygoals:wind,solar,geothermal,andbiomass.Windandsolar
generationareintermittent,astheirenergysourcesarenotdispatchable:
Thepowerproducedbyawindplantvariesasafunctionofwindspeed.
Thepowerproducedbysolargenerationvariesastheintensityofthesunlight.
Geothermalandbiomassresourcesaredispatchableand,therefore,arenotintermittent
generation.
2
Fourscenarioswereanalyzed,asfollows:
2006BaseCase
o Existing2006transmissionsystemwithexistingmixofgeneration,including
2,100megawatts(MW)ofwindand330MWofsolar.
2010TehachapiCasewith20percentrenewableenergy(designated2010T)
o 7,500MWwindand1,900MWsolarinCalifornia.
o Includes4,200MWofnewandexistingwindgenerationatTehachapi,withnew
500kilovolt(kV)transmissiontosupportit.
2010AcceleratedCasewith33percentrenewableenergy(designated2010X)
o Assumesinteriminfrastructurewithmostofthe2020intermittentrenewable
generation.
2020Casewith33percentrenewableenergy
GeneralElectricEnergyConsultingevaluatedtheeffectofintermittentgeneration(windand
solar)ontheoperationoftheCaliforniapowergrid.Theobjectiveswere:
EvaluateCaliforniagridoperationwithincreasinglevelsofintermittentgeneration,
uptotherenewablepolicylevelsofwindandsolarandusingthefourscenarios
developedforthatpurpose.
Identifyandquantifysystemperformanceandanyoperationalproblems(for
example,loadfollowing,regulation,operationduringlowloadperiods).
Identifyandevaluatepossiblemitigationmethods.
Theevaluationcoveredtimescalesinvolvedingridoperationandincludedthefollowing
specifictypesofanalysisforeachscenario:
Statisticalanalysisofvariabilityduetosystemload,aswellaswindandsolar
generationovertimeframes(hourly,5minute,1minute).
ProductioncostsimulationsoftheCaliforniapowergridandtheWesternElectricity
CoordinatingCouncil,usingtheMultiAreaProductionSimulationprogram,to
evaluatehourbyhourgridoperationfor3yearswithdifferentwindandload
profiles.
Quasisteadystatesimulations,usingPositiveSequenceLoadFlowprogram,to
evaluateminutebyminutetimesequencedpowerflowsfortheentireWestern
ElectricityCoordinatingCouncilgridoverseveralhours,toquantifygrid
performancetrendsandtoinvestigatepotentialmitigationmeasures.
3
Theeffectofwindandsolarforecastingingridoperationsandunitcommitmentwerealso
evaluated.
Conclusions
Twoscenarios(2010Tand2020)representedstepsonanexpectedtrajectorytomeetCalifornias
renewablegenerationgoal.Theartificiallyaccelerated2010Xscenariowasdevelopedto
increasesystemstressandrepresentsthemostchallengingstudycondition.However,the
conclusionsandrecommendationspresentedinthissectionapplytoallscenarios,notjustthe
mostchallenging.Theyareintendedtoenableconsistent,sustainedrenewablegrowththrough
2020.
The2010Xscenarioexaminedatotalof19,800MWofrenewablesinCalifornia,including
12,500MWofwindgeneration,2,600MWofsolar,1,000MWofbiomass,and3,700MWof
geothermal.Thisscenariorepresentsastressedconditiondesignedtotestthesystemwithmore
renewablesthanprojectedfor2010.
ThislevelofrenewablegenerationcanbesuccessfullyintegratedintotheCaliforniagrid
providedappropriateinfrastructure,technology,andpoliciesareinplace.Specifically,this
successfulintegrationwillrequire:
Investmentintransmission,generation,andoperationsinfrastructuretosupportthe
renewableadditions.
Appropriatechangesinoperationspractice,policy,andmarketstructure.
Cooperationamongallparticipants,CaliforniaIndependentSystemOperator
(CaliforniaISO),investorownedutilities,renewablegenerationdevelopersand
owners,nonFederalEnergyRegulatoryCommissionjurisdictionalpowersuppliers,
andregulatorybodies.
Recommendations
ThestudyscenariosrepresentstagesalongatrajectorytomeetCaliforniasrenewable
generationgoal.Thefollowingrecommendationsareasetoftargets,actions,andpolicies
designedtoensuresuccessfulintegrationofsignificantlevelsofintermittentrenewable
generationthrough2020.Theimplementationoftheserecommendationsshouldproceedwith
therenewablegenerationgrowth.Suchevolutionaryimprovementswillallowsecureand
economicintegrationatallstagesalongtherenewablegenerationgrowthtrajectory.
Thechallengeofaccommodatingsubstantialintermittentrenewablegenerationisincremental
tothechallengeofservingexistingandnewload.Longtermplanningmustalwaysconsider
requirementsforgenerationandtransmissionandstrikeanappropriatebalancebetweenthe
two.Further,newconsiderationsspecifictorenewabletechnologiesmustbeincluded.Thus,the
planningprocessmustconsiderthreemajorsystemcomponents:
Generation
Transmission
4
RenewableTechnology
Therecommendationspresentedbelowaregroupedaccordingly.
Generation Resource Adequacy
TheCaliforniaEnergyCommission,CaliforniaPublicUtilitiesCommission,andCaliforniaISO
haveongoingprocessestoprovidethegenerationinfrastructurenecessarytomaintainreliable
operation.Theadditionofbothintermittentandnondispatchablerenewableresourcestothe
Californiagridincreasestherequirementforgenerationresourceflexibility.Itisessentialthat
thisrequirementforflexibilitybeincludedintheoverallassessmentandplanningforresource
adequacy.Itisrecommendedthatspecificattributesofgenerationflexibilitybeinventoried,
maintained,andincreased.Wherepossible,quantitativetargetsaresuggested;othersmaybe
adoptedascircumstancesandunderstandingchanges.Toavoidrepetition,specificpolicyand
technologyrecommendationsaregroupedwiththemostrelevantperformanceissue.However,
manyrecommendationscouldapplytoabroaderrangeofperformancecategories.Further,
noneoftherecommendationsareeitherselfsufficientormutuallyexclusive.Anappropriate
combinationofrecommendationswillbemostsuccessful.
MinimumLoadOperation.TheCaliforniagridshouldtargetacombinationofinstategenerating
resourcesandpowerexchangecapability/agreementswithneighboringsystemsthatallow
operationdowntoaminimumnetload(loadminuswindminussolar)intherangeof
18,000MWto20,000MW.Thesetargetswillmeetthelongterm(2020)needsofthesystemand
allowforoperationwithminimalcurtailmentofintermittentrenewables.
MinimumTurndown.Generatingresourceswithlowerminimumpoweroutputlevels
providegreaterflexibilityandallowsuccessfuloperationatminimumload.New
generatingresourcesshouldbeencouragedand/orrequiredtohavethiscapability;
existinggenerationshouldbeencouragedand/orrequiredtoupgradetheircapability.
Acomparisonoftheloadandnetload(loadwindsolar)forthevariousscenarios
showsthatminimumsarelesswiththeintermittentgenerationonthesystem.The
minimumsystemturndowncapabilitywilldeterminetheamountofrenewable
generationcurtailmentthatisnecessary.Aminimumof20,000MWisexpectedtoresult
incurtailmentduringafewhundredhoursperyearfortheexpectedgrowthtrajectory.
DiurnalStart/Stop.Anotherwaytomeetminimumloadistoincreasetheamountof
generationthatiscapableofreliablediurnalcycling.Thiswillbenefitthesystemby
allowingthecommitmentofunitsthatareeconomicatpeakandshoulderloads,
withoutrequiringtheirnoneconomicoperationatlightload.
LoadParticipation.Activeparticipationbylargeloads,especiallypumps,isanotherway
toassureadequateflexibility.ThepumpscontrolledbyCaliforniaDepartmentofWater
Resourcesarealreadyparticipantsintheenergymarket,butadditionaltypesof
participationandcooperationcouldincreaseoverallsystemflexibility.Forexample,
additionalinvestmentinpumps,controls,orotherloadinfrastructuretotakeadvantage
oflightloadenergypricingcouldbebotheconomicandeffective[8].
5
Californiashouldexploreothermeanstoencourageloadshiftingtowardlightload
conditions.Variousloadshiftingandstoragetechnologies,suchascoldstorage(for
example.forbuildingcoolingorinletaircoolingforgaspeakinggeneration)hold
promiseandmayprovetobeeconomic.Arrangementsthatgivethegridoperator
controloverloadsforacontractualconsiderationorratereductionwillbemore
attractiveaspenetrationofintermittentrenewablesincreases.
PumpedStorageHydro.Useofpumpedstoragehydrofacilitieswasshowntoincreasefor
thescenariosexamined.Theinfrastructureandpolicynecessarytoallowthebestuseof
existingpumpedstoragehydrowithinCaliforniashouldbeenhanced.Additional
pumpedstoragehydrocapabilitycouldalsoenhancesystemschedulingflexibilityand
willlikelyaidotherflexibilityattributesdiscussedbelow.Thisisparticularlytruewhen
conventionalhydroflexibilityislow,duetounusuallyhighrunoffconditions.
HourlyScheduleFlexibility.TheCaliforniagridshouldtargetacombinationofinstategenerating
resourcesthatprovideaminimumlevelofschedulingflexibility.Theanticipatedloadgrowth
to2020willdrivetheoverallsystemflexibilityneedsfromthepresentlevelofabout
4,300megawattsperhour(MW/hr)toabout6,000MW/hr.Theadditionalvariabilityand
uncertaintyassociatedwithintermittentrenewableswillincreasetheamplitudeofsustained
loadramps(bothupanddown),andthefrequencyofgenerationstartsandstops.Forthe
expectedrenewablesgrowthtrajectory(2010T,2020),theoverallhourlyflexibilityrequirement
isexpectedtobeabout130MW/hrgreaterthanthatrequiredforloadalone.Underthe
artificiallyacceleratedrenewableexpansionofthe2010Xscenario,thatincrementalrequirement
isabout400MW/hr.
Duringlightloadconditions,totalrequirementsaresmaller,buttherelativeimpactof
intermittentrenewablesislarger.Theanticipatedloadgrowthto2020willdrivethelightload
systemflexibilityneedsfromthepresentlevelofabout2,000MW/hruptoabout3,000MW/hr.
Fortheexpectedrenewablesgrowthtrajectory(2010T,2020),thehourlylightloadflexibility
requirementisexpectedtobeabout1,000MW/hrgreaterthanthatrequiredforloadalone.
HydroScheduling.Conventionalhydroelectricgenerationplaysakeyroleinlightload
scheduleflexibilityaswellasloadfollowingandregulation.Economicoperationwillbe
enhancedbyhighhydroflexibility.Existingflexibilityshouldbemaintainedatleast,
andinvestmentstoincreasemaneuverabilityshouldbeconsidered.Adocumented
inventoryofcapabilityisimportant.Californiashouldperiodicallyexaminetheamount
andtypeofhydroconstraints,andevaluateinvestmentsorcontractualmechanismsfor
costeffectivereliefofthoseconstraints.
FasterStart/Stop.Uncertaintiesinforecastscreateasomewhatdifferentflexibility
requirement.Evenwithstateoftheartwindforecasting,bothdayaheadandhour
aheadnetloadforecastuncertaintieswillincreaseduetointermittentrenewables.With
anincreasedriskofanactualnetloadsignificantlydifferentfromtheforecastnetload,
shortnoticestart/stopcapabilityduringdailyoperationwillbeanimportantpartofthe
redispatchneededtobalancegenerationandload.TheCaliforniagridshouldtarget
6
sufficientinstategeneratingresourcecapabilitytomeetdayaheadforecasterrorsinthe
rangeof5,000MWofgenerationcapacityandhouraheadforecasterrorsintherange
of2,000MWofgenerationcapacity.Overall,thisrepresentsaboutdoublethepresent
levelofdayaheadloadforecasterrorandabout20percentmorethanthepresenthour
aheadloadforecasterror.
Duringlighterloadperiods,thenetloadforecasterrormaybethreetimestheload
aloneforecasterrorinthedayaheadforecast.Thetargetsrecommendedabovewillalso
besufficientforlightloadconditions.
MultiHourScheduleFlexibility.Flexibilitytargetsshouldalsoaddressperiodsofsustainedload
increasesanddecreases.TherecommendedtargetsarefortheCaliforniagridtohaveenough
resourcestomeetamaximummorningloadincreaseof12,000MWoverthreehoursanda
maximumeveningloaddecreaseof14,000MWoverthreehours.Thisrepresentsanincreaseof
about1,000MWoverthecapabilityneededtomeettheloadalone.
LoadFollowingCapability.TheCaliforniagridshouldtargetacombinationofinstategenerating
resourcesthatprovideaminimumlevelofgenerationrampingcapability,bothupanddown.
Onaverage,thesystemshouldmaintainontheorderof+/130MW/minuteforaminimumof5
minutes.Thisisabouta10MW/minuteincreaseovertherequirementduetoloadalone.
Duringlightloadconditions,approximately70MW/minuteofdownloadfollowingcapability
arerequired.Uploadfollowingrequirementsarelower.Theloadfollowingcapabilityshould
besubjecttoeconomicdispatchfromthesystemoperators.Loadfollowingdutyshouldnotbe
shiftedtounitsprovidingregulation.
Import/ExportScheduling.TheCaliforniagridshouldrecognizethateconomic
incorporationofsubstantialinstaterenewableswillinevitablyinvolvesignificant
displacementofimportedenergy.Regulatoryandcontractualarrangementsforimports
andexportsshouldbestructuredsuchthatthevalueofschedulingflexibilityis
recognized,allowed,andappropriatelycompensated.Inparticular,Californiashould
allowschedulechangestooccurmorefrequentlyandattimesotherthanonthehour.
RegulationCapability.TheCaliforniagridshouldtargetacombinationofinstategenerating
resourcesthatprovideaminimumlevelofregulationcapability.TheCaliforniaISOcurrently
procuresregulationintherangeof300MWto600MW.Theprocuredamountvaries
substantiallyoverallloadlevels.Theimpactofintermittentrenewablesonregulation(20MW)
isconsiderablylessthanthenormalvariabilityintheamountprocured.However,regulation
resourceswillcontinuetobeimportant.Therefore,theCaliforniagridshouldatleastmaintain
thecurrentlevelofregulationcapability.Thislevelofregulationshouldallowthestateto
continuetosatisfytheirregulatoryobligationsforinterchangeandfrequencycontrol,suchas
theNorthAmericanElectricReliabilityCouncilControlPerformanceStandard2compliance
shouldbecontinuallyscrutinizedasintermittentrenewablesareaddedtothegridtorefine
regulationrequirementsandprocurement.
7
RegulationTechnologies.Californiashouldconsiderusingtechnologiesbeyond
conventionalgenerationtoprovideregulation.Theearlierdiscussionaboutload
participationinscheduleflexibilityapplieshereaswell.Functionalrequirementsfor
loadstoprovideregulationaredifferentfromthoseforgeneration.Givenasuitable
regulatoryandmarketstructure,however,itislikelythatothertechnologiesand
participantswillemergetoprovidetherequiredservices.Examplesincludesometypes
ofstoragetechnology,suchasvariablespeedpumpedhydroandthelatestflywheel
energystoragesystems.Policyandmarketstructureshouldencouragediversityof
participantsinprovidingancillaryservices,andtechnicalspecificationsforperformance
shouldbesufficientlyflexibletoallowtheintroductionofnewtechnologies.
NonTechnicalResourceAdequacyConsiderations.Theprecedingrecommendationswereaimedat
securingthetechnicalcapabilitiesnecessaryforsuccessfulintegrationofintermittent
renewables.Thefollowingitemsaddresspolicyandcommercialconsiderations.
MarketDesign.Itmustberecognizedthatwhileoperationalflexibilityisvaluabletothe
grid,itcurrentlyholdslittleattractionforpowersuppliers.Deeperturnback,more
rapidcyclingandloadfollowing,andmorefrequentstartsandstopsallimpose
significantcostsandrevenuereductionsonthesuppliers.Marketandregulatory
structuresmustrecognizethevalueoftheseflexibilityfeatures.Policychangesmay
includeacombinationofexpandedancillaryservicesmarkets,incentives,and
mandates.
ContractualObligations.Muchoftheanalysispresentedinthisreportisbasedonthe
presumptionthatthegridisoperatedrationallythatis,theavailablegeneration
resourcesareusedasefficientlyandeconomicallyaspossible.Theanalysisdidnot
includehistoricalconstraints,suchaslongtermcontractualobligations,thatforcethe
systemtorunlessefficientlythanpossible.Newcontractsunderconsideration,existing
longtermcontractsupforrenewal,orindeedanyexistingcontractsthatcouldbe
renegotiatedshouldbereviewedwithalloftheprecedingresourceadequacy
recommendationsinmind.TheCaliforniagridmustmaintainoperationalflexibility,
andtodoso,itmusthavenotonlythephysicalresourcesnecessary,butalsothe
businessandcontractualarrangementsnecessarytoenabletherationaluseofthose
physicalresources.
Retirements.Generatingplantretirementsthatwerefirmlyscheduledwhenthe
databaseswereassembledwereincorporatedintothisstudy.However,increased
competitionfromnewresources,renewableorotherwise,willtendtopushmarginally
profitablegeneratingresourcesoutofbusiness.Suchspeculative,economicretirements
werenotconsideredinthestudy.Successfulimplementationoftherecommendations
abovewillensurethatresourceswiththenecessaryflexibilityareavailable.Inaddition,
itisrecommendedthatretirementsbeprojected,monitored,andevaluatedduringthe
resourceplanningprocess.
8
Inventory.Duringthisstudy,itwasnotedthatgeneratorcharacteristicsandcapabilities
(forexample,minimumturndown,rampratecapability)werenotalwaysknownwith
sufficientdetailorcertainty.Somedegreeofuncertaintyisinevitable.However,with
theincreasedneedforresourceflexibility,Californiashouldimplementaprogramto
measure,verify,andcataloguetheflexibilitycharacteristicsofthegenerationresources.
AprogramsimilartotheWesternElectricityCoordinatingCouncilgeneratordynamic
testingmightprovesuitable.
Transmission Infrastructure
TheadditionofthousandsofMWofnewgenerationofanyvarietywillrequireexpansionof
thetransmissionsystem.Thisstudyincludedtheadditionofenoughbulktransmission
necessaryforconnectionofthenewrenewablestothegridasdeterminedanddocumentedby
DavisPowerConsultants[6].However,itwasnotadetailedtransmissionstudyandisnota
substituteforone.Policiesmustrecognizethatlocalproblemsmightdevelop,andenablethe
necessarytransmissionadditions.Practiceandpolicythatcorrectproblemsandstrikeabalance
betweeninfrastructureinvestmentandcongestionarenecessary.Toanappreciableextent,this
observationholdsforalltransmissionplanningandallgenerationadditions.Californiacan
economicallybenefitfromchangesinplanningandoperationofthetransmissioninfrastructure
byrecognizingthelocationalandvariablenatureofintermittentrenewables.Thefollowing
recommendationsarespecifictotheseneeds.
ExistingConstraints.Californiahasexistinginfrastructurethatcontributessubstantiallytothe
secureandeconomicoperationofthegridwithhighlevelsofintermittentrenewables.Insome
circumstances,theuseofthatinfrastructureforsystemwidebenefitisconstrainedbylocal
transmissionlimitations.OneexampleofsuchaconstraintistheoccasionalinabilityofHelms
pumpedstoragehydrotoreachfullpumpingpower.Planningandpolicyshouldrecognizeand
enablecorrectionofsuchlocallimitations.
RatingCriteria.Windgenerationisvariable,andthespatialdifferencesbetweenplants
substantiallyeffectsthecoincidentproductionofpowerfromthoseplants.Clearly,awindplant
willreachratedoutputformanyhoursperyear.Thus,normalplanningcriteriarequires
sufficientcapability,suchasthermalrating,onthetransmissioninterconnectiondedicatedto
thatplanttoaccommodateratedpoweroutput.
However,asmorewindplantsvieforaccesstospecifictransmissioncorridors,itwillbe
increasinglyunlikelythatallwindplantswillsimultaneouslyreachtheirmaximumoutput.
Notethatinthreeyearsofdata,allwindplantsinthisstudyneversimultaneouslyreached
maximumoutput.Andthe12,500MWofwindgenerationexceeded10,000MWofproduction
lessthan1percentofthetime.Thus,transmissionplanningtoaccommodatemultiplewind
plantsshouldconsidertheirspatialdifferencesandthestatisticalexpectationofsimultaneous
highpoweroutputlevels.Plantsclosetogetherwillgenerallyrequiretransmissioncapability
equivalenttotheaggregateratingoftheplants.Plantsthatarefartherapartmayrequireless
transmissioncapability.Hence,itisnotnecessarytoguaranteesufficientratingonthebulk
transmissioninfrastructuretoaccommodateallwindprojectsatfulloutput.Existingcriteria
shouldbesufficienttoprovidethisplanningflexibility.
9
Technology.Policyshouldrewardinvestmentintechnologytomaximizeuseoftransmission
infrastructureforrenewables.Suchpoliciesshouldrecognizethatwindgenerationisa
relativelypoorresourceforcapacityandthatcreativeuseoftechnologymayoptimizeuseof
transmission.Regulatoryandcontractualpracticeshouldallowtechnologiessuchasrealtime
lineratings,controlsthatmanageoutputfromintermittentrenewableresources,localshort
termforecasting,andothernonstandardapproachestobalancerenewableenergydelivery
withtransmissioninfrastructurecosts.
Renewable Generation Technology, Policy, and Practice
Withsignificantlevelsofintermittentrenewablegeneration,operationmaybechallengingat
extremelylightloadlevels,underaconstrainedtransmissiongrid,orwithhighwindvolatility.
Undertheseconditions,renewablegenerationmustparticipateinoverallgridcontrol.The
followingrecommendationsarespecifictorenewabletechnologyandareaimedatassuringthat
intermittentrenewablesplayanactiveandpositiveroleinthesecureandeconomicoperationof
thegrid.
Curtailment.Undertherareoccasionsofcoincidentminimumload,highwindgeneration,and
lowconventionalhydroflexibility,itmustbepossibletocurtailintermittentrenewables.The
gridoperatorshouldhavetheabilitytoordersuchareductioninproduction.Regulatoryand
contractualarrangementsforintermittentrenewablesshouldbestructuredsuchthat
curtailmentsarerecognized,allowed,andappropriatelycompensated.Rampratecontrols
couldalsobeconsidered.
AncillaryServices.Intermittentrenewablesmaybeabletoprovideancillaryservicesthatare
bothvaluableandeconomicundersomeoperatingconditions.Forexample,windgeneration
canprovidefrequencyregulation.Suchfunctionalityisarequirementinsomeregions[10].
Regulatoryandcontractualarrangementsforintermittentrenewablesshouldbestructuredsuch
thatprovidingsuchservicesarerecognized,allowed,andappropriatelycompensated.
Forecasting.SuccessfulandeconomicoperationoftheCaliforniagridrequireswindandsolar
forecasting.Thisstudyverifiedsubstantialbenefitsfromtheuseofstateoftheartdayahead
forecastingintheunitcommitmentprocess.Substantialbenefitsareexpectedforimprovements
inbothlongerterm(multiday)andshortterm(hoursandminutesahead)forecasting.
Investmentandpolicymustencouragedevelopmentofhighfidelityintermittentrenewable
forecastingforallintermittentrenewablegenerationinthestate.
Monitoring.Thewindproductionprofilesusedinthisstudyarebasedonhistoricalweather
dataandsophisticatedcomputermodels.Recordeddatafromrealoperatingexperiencewillbe
invaluableinrefiningoperatingpractice,performance,andflexibilityrequirements.Time
synchronizedproductionandmeteorologicaldatafrommanyplantswillprovidevalidationor
correctionofthetrendsandresultspredictedbythisstudy.Theywillshowthebenefitsand
limitationsofspatialdifferences,mesoscalemodeling,andvariouswindplantcontrols.Itis
recommendedthatCaliforniacontinueandexpand,asnecessary,programstomonitor,analyze,
anddisseminateperformanceinformationregardinggridoperationsandplanningfor
intermittentrenewables.
10
ThetargetedlevelsofrenewablegenerationcanbesuccessfullyintegratedintotheCalifornia
gridprovidedappropriateinfrastructure,technology,andpoliciesareinplace.
11
1.0 Introduction
Californiahasoneofthemostdiverseelectricitysupplysystemsinthenationwithalarge
potentialtogenerateelectricityfromrenewablesources,suchaswind,geothermal,biomass,
hydroelectricandsolar.WithprogressiverenewablepoliciesasintheRenewablesPortfolio
Standard(RPS)andstateEnergyActionPlan[1],thechallengefacingthestatewillbehowbest
tointegrateandmanagerenewableenergyresourceswithtraditionalgenerationwhileensuring
areliableelectricitysystem.
1.1. Challenges
Withpolicytargetsof20%renewableenergyby2010and33%by2020,afewofthemain
challengesfacingthestateinclude:
Buildingsufficienttransmissioninfrastructuretosupportandsustainthe
developmentenvisionedfor2020
Balancingtheneedtointegrateincreasinglevelsofrenewableenergywhile
minimizingadverseimpactsonthesurroundingenvironment
Developingtoolswiththeindustrytoproperlyintegratevariablerenewable
resourcesincludingwindandsolar
1.2. Background
TheIntermittencyAnalysisProject(IAP)soughttoaddressthefollowingquestions:
Whataretheimpactsofincreasingrenewableenergyprojectsonsystemreliability
anddispatchability,withaparticularfocusonwindandsolarenergy?
Whatwillthefuturesystemlooklikeandwherewilltheresourcescomefrom?
Howwillthefuturegridneedtorespond?Willitrespondbymarketstructure,
services,ortechnologies?
TheIAPwastailoredtopresentastatewideperspectiveofthetransmissioninfrastructureand
servicesneededtoaccommodatetherenewablepenetrationlevelsdefinedinthestates
renewableenergypolicy.TheIAPwastechnicalinnatureandwasintendedtoprovideayear
2020perspectiveonpotentialoperationalneedsandimpactstomeetfuturegrowthand
demand.Asaresult,certainassumptionsweremadeontechnologyavailability,system
conditionsandconstraints,aswellasmarketconstraints.
Inthisproject,powerflowandproductioncostmodelingwereconductedtoestablishthe
operationalbaselineoftheCaliforniagridasof2006andtodeveloptherenewableresource
mixesforthe2010and2020scenarios.Renewableportfoliomixesaswellasthetransmission
neededtointerconnecttheresourceswereevaluatedinthescenariosbasedonatransmission
benefitcriteria.ThemodelingbuiltandexpandedonpreviousCaliforniaEnergyCommission
(EnergyCommission)fundedtransmissionstudiesthatfocusedonconnectingstatewide
renewableresourcepotentialandtheassociatedtransmissionconsiderations.
12
ThisprojectbuiltuponworkthatwascompletedfortheEnergyCommissionaspartofthe2005
IntegratedEnergyPolicyReport(IEPR)process[2].Moreinformationmaybefoundrelatedto
IEPRontheCommissionwebsite(www.energy.ca.gov).TheIAPeffortleveragedwork
conductedbytheCaliforniaWindEnergyCollaborative(CWEC)[3],theConsortiumfor
ElectricReliabilityTechnologySolutions(CERTS)[4],andtheStrategicValueAnalysis(SVA)
workbyDavisPowerConsultants(DPC).UndertheSVAproject,PublicInterestEnergy
Research(PIER)andDPCassessedtheavailabilityofrenewableresourcesanddefinedan
approachthatminimizestransmissioninfrastructurechangesandmaximizesbenefitsfor
integratingrenewablesontotheCaliforniagridbyavoidingcongestion.Availabilityofinter
stateandintrastaterenewableresourcesandtransmissionrequirementswerealsomodeled
usingtheSVAapproachtoalleviate,oratleastminimize,transmissionconstraints[5].
1.3. Intermittent Generation Definition
TheIAPconsideredfourtypesofrenewablegenerationtomeetCaliforniasrenewableenergy
goals;wind,solar,geothermal,andbiomass.Windandsolargenerationareintermittent,as
theirenergysourcesarenotdispatchable:
Thepowerproducedbyawindplantvariesasafunctionofwindspeed
Thepowerproducedbysolargenerationvariesastheintensityofthesunlight.
Geothermalandbiomassresourcesaredispatchable,andthereforearenotintermittent
generation.
1.4. Overview of Project Objectives, Tasks and Participants
ThisreportdocumentsthelaststageofthemultistageIAP.Precedingstagesincluded:
Evaluationofpast,present,andfuturewindturbinetechnologies,andtheirimpact
ontransmissionsystemoperationandperformance,byBEWEngineering,Inc.
Assessmentofworldwideexperiencewithintegratinglargepenetrationsofwind
energy,byKevinPorterofExeterAssociates,Inc.
DevelopmentoffuturerenewableenergyscenariosconsistentwithCaliforniasRPS,
andevaluationoftheimpactsontransmissionreliability,byDavisPower
Consultants(DPC).
ThefuturerenewablegenerationscenariosdevelopedbyDPCwerecriticalinputstothe
analysisdocumentedinthisreport.Figure1isaflowchartshowingthemajortasksthat
assessedgridimpactsandmitigationmethodsforintermittency,aswellasthetasksthat
producedthescenariosanddatanecessaryforthatanalysis.DPCassessedthepotentialfor
futurerenewableresources(wind,solar,geothermal,biomass)anddevelopedaseriesof
scenarioswithincreasinglevelsofrenewablegeneration,consistentwiththegoalsofthe
CaliforniaRPS.AlloftherenewablegenerationwaslocatedinsideCalifornia.TheDPCteam
alsodevelopedcorrespondingtransmissionexpansionplansforeachrenewablegeneration
scenario.
13
Thus,theresultsoftheDPCtaskincludedthefollowingforeachscenario:
Detailedlistofindividualrenewablegeneratingplantsandtheirsite/rating,
consistentwithCaliforniasRPSgoalsandlocationsofrenewableresources,and
Powerflowdatasetswithconventionalgeneration,renewablegeneration,and
transmissionsystembuildoutsconsistentwiththeprojectedCaliforniapeakload
level.
Thescenariosare:
2006BaseCase
o Existing2006transmissionsystemwithexistingmixofgeneration,including
2,100megawatts(MW)ofwindand330MWofsolar.
2010TehachapiCasewith20%renewableenergy(designated2010T)
o Includes4,200MWofnewandexistingwindgenerationatTehachapi,withnew
500kilovolts(kV)transmissiontosupportit.
2010AcceleratedCasewith33%renewableenergy(designated2010X)
o Assumesinteriminfrastructurewithmostofthe2020intermittentrenewable
generation.
2020Casewith33%renewableenergy
CompleteresultsofthisportionoftheIAParedocumentedinthereportIntermittencyImpacts
ofWindandSolarResourcesonTransmissionReliability,byDavisPowerConsultants[6].
14
Powerflow Data and
Renewable Generation Mix
for 4 Study Scenarios
(from DPC)
Wind Power Profiles
(from AWST)
Solar Power Profiles
(from multiple sources)
Load Profiles and
Historical Operation Data
(from CAISO)
Analysis of Impacts
on
Grid Operations
Statistical Analysis
Quasi-Steady-State
Simulations
Production Cost
Simulations
Conclusions and
Recommendations
Evaluation of Potential
Mitigation Methods
Renewable Resource
Potential and Availability
Assessment
(by DPC)
Transmission
Assessment
Development of Study Scenarios
(by DPC Team)
Analysis of Grid Impacts and Mitigation Methods
(by GE Energy Consulting Team)
Figure 1. Flowchart of Intermittency Analysis Project.

Note:AWST=AWSTruewind,LLC.;CAISO=CaliforniaIndependentSystemOperator;
GEEnergy=GeneralElectricEnergyConsulting.
GeneralElectricEnergyConsulting,(GE)evaluatedtheimpactofintermittentgeneration(wind
andsolar)ontheoperationoftheCaliforniapowergrid.Theobjectiveswere:
EvaluateCaliforniagridoperationwithincreasinglevelsofintermittentgeneration,
uptotherenewablepolicylevelsofwindandsolarandusingthefourscenarios
developedforthatpurpose.
Identifyandquantifysystemperformanceandanyoperationalproblems(e.g.,load
following,regulation,operationduringlowloadperiods).
Identifyandevaluatepossiblemitigationmethods.
Theevaluationcoveredmultipletimescalesinvolvedingridoperation,asillustratedinFigure
2,andincludedthefollowingspecifictypesofanalysisforeachscenario:
Statisticalanalysisofvariabilityduetosystemload,aswellaswindandsolar
generationovermultipletimeframes(hourly,5minute,1minute).
15
ProductioncostsimulationsoftheCaliforniapowergridandtheWesternElectricity
CoordinatingCouncil(WECC),usingGeneralElectricsMultiAreaProduction
Simulation(GEMAPS
TM
)program,toevaluatehourbyhourgridoperationfor3
yearswithdifferentwindandloadprofiles.
Quasisteadystatesimulations,usingPositiveSequenceLoadFlow(PSLF)program,
toevaluateminutebyminutetimesequencedpowerflowsfortheentireWECCgrid
overseveralhours,toquantifygridperformancetrendsandtoinvestigatepotential
mitigationmeasures.
Theimpactofwindandsolarforecastingingridoperationsandunitcommitmentwerealso
evaluated.
Thisreportpresentstheresultsofthatanalysis,aswellasconclusionsandrecommendations
drawnfromtheresults.
1.5. Participants
Inconductingthisproject,GEEnergyConsultingcollaboratedwithseveralotherorganizations
onthefollowingessentialtasks:
AWSTrueWindwindprofiledataforexistingandfuturewindgenerationsites[7]
RumlaproductioncostmodeldatafortheCaliforniapowergridandWECC.
CaliforniaIndependentSystemOperator(CaliforniaISO)loadprofilesand
historicaloperationdatafortheCaliforniapowergrid.
Solardatawasobtainedfrommultiplesources,includingCaliforniaISO,the
UniversityofCaliforniaatDavis,theNationalRenewableEnergyLaboratory
(NREL),StirlingEnergySystems,theCaliforniaPublicUtilitiesCommission(CPUC)
SelfGenerationIncentiveProgram,andAtmosphericResearchScienceCenteratthe
StateUniversityofNewYorkatAlbany.
GEEnergyConsultinggratefullyacknowledgesthevaluablecontributionsofallofthese
organizations.Thisprojectcouldnothavebeenperformedwithoutthem.
16
Un i t Co mmi t men t
an d
Day-Ah ead
Sc h ed u l i n g
L o ad Fo l l ow i n g
(5 Mi n u t e Di s p at c h )
Fr eq u en c y an d
Ti e-L i n e Reg u l at i o n
(AGC)
Day-ah ead an d
Mu l t i -Day
Fo r ec as t i n g
F
a
s
t
e
r

(
s
e
c
o
n
d
s
)

T
i
m
e

F
r
a
m
e

S
l
o
w
e
r

(
Y
e
a
r
s
)
Pl an n i n g an d
Op er at i o n Pr o c es s
Tec h n o l o g y
Is s u es
Ho u r -Ah ead
Fo r ec as t i n g
an d
Pl an t Ac t i ve Po w er
Man eu ver i n g an d
Man ag emen t
Res o u r c e an d
Cap ac i t y Pl an n i n g
(Rel i ab i l i t y)
Uni t Di spatch
0
100
200
300
400
500
600
700
0 2000 4000 6000 8000
Hour
M
W
Real -Ti me an d
Au t o n o mo u s Pr o t ec t i o n
an d Co n t r o l Fu n c t i o n s
(AGC, L VRT, PSS,
Go ver n o r , V-Reg , et c .)
Cap ac i t y Val u at i o n
(UCAP, ICAP)
an d
L o n g -Ter m L o ad
Gr ow t h Fo r ec as t i n g
2001 Aver ag e L o ad vs Aver ag e Wi n d
0
5,000
10,000
15,000
20,000
25,000
30,000
1 6 11 16 21
Hour
N
Y
I
S
O

L
o
a
d

(
M
W
)
0
200
400
600
800
1,000
1,200
1,400
1,600
W
i
n
d

O
u
t
p
u
t

(
M
W
)
J u l y l o ad Au g u s t l o ad Se p t e m b e r l o ad
J u l y w i n d Au g u s t w i n d Se p t e m b e r w i n d
0
500
1000
1500
2000
2500
3000
1 61 121
M i nu t e s
M
W
S eptember Morning A ugus t Morning May Evening October Evening April Af ternoon
1 Year
1 Day
3 Hours
10 Minutes
Figure 2. Time Scales for Grid Planning and Operations.

UCAP=uniformcapacity,ICAP=installedcapacity,AGC=automatedgeneratorcontrol,
LVRT=lowvoltageridethrough,PSS=powersystemstabilizer,VReg=voltageregulation
17
2.0 Study Approach
Theoverallstudyapproachisoutlinedinthissection.Thestudyscenarios,typesofanalysis,
dataprovided,andterminologyarealldescribedbelow.
2.1. Study Scenarios
TherenewablegenerationmixforeachofthefourstudyscenariosissummarizedinTable1.
Thevariousscenarioscanbedescribedintermsofincreasinglevelsofrenewablegeneration
penetration.Butthedefinitionofpenetrationcanbeconfusing.ManyRenewablePortfolio
Standards(RPS)usepenetrationtodescribethepercentofenergytobeprovidedbyallofthe
renewablegeneration,includingwind,solar,geothermal,biomassandsometimeshydroelectric.
TheenergytargetsdiscussedintheIAPfallintothiscategorywith20%penetrationby2010
and33%penetrationby2020.Theenergydefinitionisimportantbecauseitisameasureof
theamountoffossilfuelgenerationthatcanbedisplaced.However,intheanalysisof
intermittentgenerationthetermpenetrationisoftenusedtodescribetheratioofthenameplate
capacityofintermittentgeneration(windandsolar)dividedbythepeakloadofthesystem.Thisis
becausetheimpactonoperationsisoftenafunctionoftheintermittentrenewablepoweroutput
relativetothesystemload.Bothdefinitionsareimportantandbothwillbeusedwithinthis
report.
In2006,theStateofCaliforniahadapeakloadof58,900MW,ofwhich48,900MWwaswithin
theCaliforniaISOoperatingarea.Therewas2,100MWofwindand330MWofsolar
generation,yielding4%intermittentgenerationpenetration(as%ofpeakload)statewideand
5%penetrationwithinCaliforniaISO.
Case2010Trepresentsafuturescenariofortheyear2010,withatotalof7,500MWofwindand
1,900MWofsolargenerationinCalifornia.Inthisscenario,theintermittentgeneration
penetrationis15%statewideand18%withintheCaliforniaISOoperatingarea.Thisscenario
includesover3,000MWofnewwindgenerationintheTehachapiregion,whichisconsistent
withexistingdevelopmentplans.Forthisstudy,theTehachapiregionwasbroadlydefinedto
includeallwindgenerationinregion8(fordetailsseetheAWSTruewindreport,[7]).
Case2020representsafuturescenarioforyear2020with33%renewableenergy,consistentwith
theCaliforniaRPSgoal.Itincludes12,700MWofwindand6,000MWofsolargeneration,
yieldinganintermittentpenetrationof25%inCaliforniaand31%withinCaliforniaISO.
Case2010Xrepresentsanacceleratedscenariowhere33%renewableenergyisintegratedintoa
transmissionsystemsimilartowhatisanticipatedfortheyear2010.Althoughthisscenariois
notarealisticprojectionofrenewableintegrationforyear2010,itprovidesvaluableinsights
relativetotheimpactofintermittentgeneration.The2020scenarioincludesnumeroussystem
expansionassumptionstoaccommodateaprojectedpeakloadof74,300MW,includingnew
transmissionlinesandconventionalgeneratingresources.The2010Xscenario,withapeakload
of62,600MW,doesnotincludethoseextensivegenerationandtransmissionadditions.Assuch,
gridperformanceofthe2010Xscenariocanbedirectlycomparedwithscenarios2010Tand
2006.Theprimarydifferencesbetweenthesescenariosarethelevelsofintermittentgeneration.
18
Similarcomparisonswiththe2020scenarioaremoredifficulttointerpret,sincedifferencesare
notlimitedtointermittentgeneration,butalsoincludesignificantdifferencesinconventional
generation,loadlevel,andtransmissionsysteminfrastructure.
Table 1. Renewable Generation Mix for Four Study Scenarios.
Scenario
2006 2010T 2010X 2020
California Peak Load, MW 58670 64336 64336 80742
California Minimum Load, MW 22804 25006 25006 31383
California Load Factor, % 60% 60% 60% 60%
California ISO Peak Load, MW 48466 53147 53147 66700
California ISO Minimum Load, MW 19066 20908 20908 26239
California ISO Load Factor, % 61% 61% 61% 61%
Total Geothermal Capacity, MW 2,400 4,100 3,700 5,100
Total Biomass Capacity, MW 760 1,200 1,000 2,000
Total Solar Capacity, MW 330 1,900 2,600 6,000
Total Wind Capacity, MW 2,100 7,500 12,500 12,700
Wind Capacity in Tehachapi Region, MW 760 4,200 5,800 5,800
CA Wind+Solar Capacity Penetration, % 4% 15% 23% 23%
California ISO Wind+Solar Capacity Penetration,% 5% 17% 26% 25%
California ISO Wind+Solar Energy, GWH 6201 26,111 43,255 49,933
California Wind+Solar Energy, GWH 6201 27,220 44,365 51,042
CA Wind+Solar Energy Penetration, % 2% 8% 13% 12%
California ISO Wind+Solar Energy Penetration, % 2% 9% 15% 14%
Notes: LoadFactor=(TotalEnergy)/(PeakLoadx8760hours)
CapacityPenetration=(Wind+SolarCapacity)/(PeakLoad)
EnergyPenetration=(Wind+SolarEnergy)/(PeakLoadxLoadFactorx8760hours)
Thewindandsolargenerationresourcesinthisstudyaredistributedamongnumeroussites
acrossCalifornia.Table2summarizesthenumbersofindividualwindandsolarsites
representedineachscenario.Forexample,the2010Tscenarioincludes12concentratingsolar
facilities,136photovoltaicgenerationsites,and98windgeneratingplants,40ofwhicharein
theTehachapiregion.The2020scenarioincludes43concentratingsolarfacilities,228
photovoltaicgenerationsites,and147windgeneratingplants.Detailedlistsofindividualwind
generationandsolargenerationsitesforeachscenarioarepresentedinAppendicesAandB.
19
Table 2. Wind and Solar Generation in California.
Scenario
2006 2010T 2010X 2020
Concentrating Solar (CS)
Number of Sites 7 12 42 43
Total CS, MW 330 1200 2100 3100
Photovoltaic (PV)
Number of Sites 0 * 136 128 228
Total PV, MW 0 * 630 530 2900
Wind Plants
Total Sites in CA 57 98 142 147
Sites in Tehachapi Region 16 40 54 54
Total Wind, MW 2100 7500 12500 12700
*ExistingPVgenerationaggregatedwithload
MostofthehistoricalgridoperationdatausedinthisstudyweresuppliedbytheCalifornia
ISO,andhencecoveredonlytheCaliforniaISOoperatingarea.Operationsdataforother
regions(LosAngelesDepartmentofWaterandPower[LADWP],SacramentoMunicipalUtility
District[SMUD],municipals,etc.)werenotreadilyavailabletothestudyteam.Hence,muchof
thestatisticalanalysisfocusedontheimpactsofintermittentgenerationontheoperationofthe
CaliforniaISOoperatingarea,ratherthantheentirestateofCalifornia.AsshowninTable3and
Table4,thevastmajorityofintermittentgenerationinthestudyscenariosislocatedwithinthe
CaliforniaISOoperatingarea.Forexample,inscenario2010T,89%ofsolarand96%ofwind
generationiswithintheCaliforniaISOoperatingarea.Inscenario2010X,88%ofsolarand93%
ofwindgenerationarewithinCaliforniaISO.GiventhattheCaliforniaISOareahas83%ofthe
totalstateload,CaliforniaISOhasahigherproportionofwindandsolargenerationthanthe
stateasawhole.Therefore,itisreasonableforthisstudytofocusongridperformanceofthe
CaliforniaISOoperatingarea.Thestatisticalanalysislookedattheimpactofallofthewindand
solargenerationforeachscenariocomparedtojusttheCaliforniaISOload.Althoughthiswas
somewhatconservativeinitsapproach,itwasnottoofarfromwhatisprojectedforCalifornia.
Table 3. Locations of Wind and Solar Resources for Scenario 2010T.
Wind Solar Total Wind+Solar
MW % MW % MW %
California ISO 7300 97% 1700 89% 9000 96%
Non-California ISO 200 3% 200 11% 400 4%
Total California 7500 100% 1900 100% 9400 100%
Table 4. Locations of Wind and Solar Resources for Scenario 2010X.

Wind Solar Total Wind+Solar
MW % MW % MW %
California ISO 11600 93% 2300 88% 13900 92%
Non-California ISO 900 7% 300 12% 1200 8%
Total California 12500 100% 2600 100% 15100 100%
20
2.2. Types of Analysis
Theprimaryobjectiveofthisstudywastoidentifyandquantifysystemperformanceandany
operationalproblems,includingloadfollowing,regulation,operationduringlowloadperiods,
etc.Threeprimaryanalyticalmethodswereusedtomeetthisobjective;statisticalanalysis,
productionsimulationanalysis,andquasisteadystateanalysis.
Statisticalanalysiswasusedtoquantifyvariabilityduetosystemload,aswellaswindand
solargenerationovermultipletimeframes(3hour,hourly,5minute,1minute).Thepowergrid
alreadyhassignificantvariabilityduetoperiodicandrandomchangestosystemload.Wind
andsolargenerationaddtothatvariability,andincreasewhatmustbeaccommodatedbyload
followingandregulationwithothergenerationresources.Thestatisticalanalysisquantifiedthe
gridvariabilityduetoloadaloneoverseveraltimescales,aswellasthechangesingrid
variabilityduetowindandsolargenerationforeachscenario.Thestatisticalanalysisalso
examinedthechangesinforecastaccuracyforloadaloneversusloadminuswindandsolar
generation.
ProductionsimulationanalysiswithGEMAPS
TM
wasusedtoevaluatehourbyhourgrid
operationofeachscenariofor3yearswithdifferentwindandloadprofiles.Theresults
quantifiednumerousimpactsongridoperationincluding:
Amountofmaneuverablegenerationonlineduringagivenhour,includingits
availablerampupandrampdowncapabilitytodealwithgridvariabilitydueto
load,windandsolar.
Effectsofload,windandsolarforecastalternatives
Changesindispatchofconventionalgenerationresourcesduetotheadditionofnew
renewablegeneration
Changesinemissionsforoxidesofsulfur(SOX),oxidesofnitrogen(NOX)andcarbon
dioxide(CO2)duetorenewablegeneration
Changesincostsandrevenuesassociatedwithgridoperation,andchangesinnet
costofenergy
Changesintransmissionpathloadings
Quasisteadystate(QSS)simulationwithPSLFwasusedtoquantifygridperformancetrends
andtoinvestigatepotentialmitigationmeasuresintheminutetominutetimeframe.QSS
analysisinvolvesminutebyminutetimesequencedpowerflowsfortheentireWECCgridover
severalhours.Thesetimesimulationsenabledexaminationoftheimpactofintermittent
generationduringchallengingtimeperiods,suchas:
Rapidmorningloadrisewhilewindgenerationisdeclining
Operationduringlowloadperiodswithminimalmaneuverablegenerationonline
Rapideveningloaddecreasewhilewindgenerationisincreasing
21
Theresultsfromthesethreeanalyticalmethodscomplementedeachother,andprovidedabasis
fordevelopingobservations,conclusions,andrecommendationswithrespecttothesuccessful
integrationofwindandsolargenerationintotheCaliforniapowergrid.
2.3. Data
Alargeamountofdatawasrequiredforthisstudy,anditwasobtainedthroughcollaboration
withmanyorganizations.Detailsofthevariousitemsofdataandtheirsourcesareexplained
below.
DPCprovidedpowerflowdata,includingtheintermittentrenewablegenerationmixand
transmissionsystemmodelsforthe2006,2010T,2010X,and2020scenarios.Adetailed
discussionofhowthatdatawasdevelopedisavailableinthereportIntermittencyImpactsof
WindandSolarResourcesonTransmissionReliability,byDavisPowerConsultants[6].
CaliforniaISOprovidedhistoricalloaddatafortheyears2002,2003,2004,includinghourly
loadMW(forecastandactual)and4secondloadMWforabout400days.Whenapplyingthis
loaddatatothestudyscenarios,thedataforallthreeyears(20022004)wasscaleduptothe
projectedpeakloadsforeachofthestudyyears2006,2010,and2020.Inotherwords,each
studyyearhadthreeyearsofhourlyloadprofiles,basedonthehistoricalloadprofiledatafor
years20022004.
CaliforniaISOalsoprovidedhistoricaloperationsdataforhydroelectricgenerationand
DepartmentofWaterResources(DWR)pumploadsforyears2004and2006.
AWSTruewindprovidedhistoricalwinddataforyears2002,2003,and2004,includinghourly
windMW(forecastandactual)and1minutewindMWfor51selectedperiods.Aseparate
windprofilewasprovidedforeachwindfarmincludedintheanalysis.
RumlaInc.compiledproductionsimulationmodelsanddataforCaliforniaandWECCfrom
multiplesources,basedontheirextensiveexperiencestudyingtheCaliforniamarket.
Solardatawasobtainedfrommultiplesources,including:
Hourlyand1minuteMWforSungenandLuzforyears2002,2003,2004(California
ISOandUCDavis)
HourlyStirlingsolarMWforMojaveandImperialfor2002,2003,2004(NRELand
StirlingEnergySystems)
Hourlyand15minutephotovoltaicMWforoneyear,aggregatedbyzipcode
(CPUCSelfGenerationIncentiveProgram)
1minuteand3minutesolarinsolationdataattwosites,forJanuaryandJuly2002
(NRELandAtmosphericResearchScienceCenteratSUNYAlbany)
AppendixCdescribeshowthesolardatawascombinedandprocessedtoproduce
representativesolargenerationprofilesfuturesolargenerationsitesinCalifornia.
22
Theloaddata,winddataandsolardatausedinthisstudyweretimesynchronizeddatasets.
Theload,windandsolardataforeachhour(orminute)werederivedfromrawdata
correspondingtothesamehour(orminute)ofthesamecalendaryear.
2.4. Terminology
Theanalysisofintermittencyinvolvesquantifyingthevariabilityinherentinthepowersystem,
aswellastheabilityofthesystemtoaccommodatethatvariabilitywhilemaintaining
performancewithinacceptableguidelines.Inthisstudy,variabilityisquantifiedbychangesin
operatingpointoverseveraldifferenttimescales.
1HourDelta:Thisreferstothechangefromtheprevioushour,typicallymeasured
inMW.Theabilityoftheoperatingareatoaccommodatehourlychangesinloadis
calledscheduleflexibilityinthisstudy.
5MinuteDelta:Thisreferstothechangefromtheprevious5minuteperiod,
typicallymeasuredinMW.Loadfollowingandeconomicdispatchfunctionsin
Californiaoperateonafiveminutecycle.
1MinuteDelta:Thisreferstothechangefromthepreviousminute,typically
measuredinMW.Regulationfunctionsoperateinthistimeframe.
Throughoutthefollowingsectionsofthisreport,1hourdeltaisusedasameasureofschedule
flexibility,5minutedeltaisusedasameasureofloadfollowing,and1minutedeltaisusedasa
measureofregulation.
Rangeandrampcapabilityaretwotermsusedthroughoutthisreporttodescribegeneration
maneuverability.Bothmeasuretheresponseofthebalanceofportfolio(i.e.,nonrenewable)
generatorstothechangingload,windandsolarconditions.Theyaredefinedasfollows:
Range:Thisreferstotheremainingcapacity(MW)availablebetweenthecurrent
operatingpointandeitherthemaximumorminimum.Uprangeistheremaining
MWcapacitytothemaximum,anddownrangeistheMWcapacityremainingtothe
minimum.Upanddownrangeareameasureofscheduleflexibilityinresponseto
hourlychanges.
Rampratecapability:Thisreferstothespeed(MW/minute)atwhichthesystemcan
usetheremainingupanddownrange.UprampcapabilityistheMW/minavailable
tomoveuptothemaximum,anddownrampcapabilityistheMW/minavailableto
movedowntotheminimum.Upanddownrampratecapabilityareameasureof
loadfollowingcapabilityinresponseto5minutechanges.
Bothtermscouldbeappliedtoindividualgeneratingunits,butaremostoftenusedtodescribe
systemwidegenerationmaneuverability.
23
3.0 Statistical Analysis
Thestatisticalanalysisprovidesabroadviewoftherelativecontributionoftheintermittent
renewablestosystemgenerationandoverallsystemvariability.Alloftheanalysespresentedin
thissectionusethevariableanduncertainbehaviorofsystemloadsasthebenchmarkfor
examiningtheincrementalchangesinvariabilityanduncertaintyduetotheintermittent
renewables.
Afundamentalcharacteristicaddressedbypowersystemoperationandplanningisthediurnal
andseasonalvariationsinsystemload.Itisaxiomaticthatsystemloadshavedailypeaksand
valleysandthatthoseextremesvarywithseasonandbetweenyears.Theinstalledgeneration
mustbecapableofservingtheloadatalltimes.Thatrequiresthatthegenerationhaveboth
sufficientratingandoperationalflexibilitytomeettheload.
Inthefollowingsections,varioustimeframesandoperatingperspectivesareevaluated.
Throughout,theincrementalimpactofintermittentrenewablesispresentedasamodificationto
theload.Thus,anyloadnotsuppliedbytheintermittentrenewablesisservedbytherestofthe
availablegeneration.Thisisreferredtoasthenetload.Evaluationsarebasedonthedata
describedinSection2.3andthesupportingappendices.
3.1. Temporal and Spatial Patterns
Fromasystemwideperspective,theloadandintermittentrenewabledataexhibitsavarietyof
temporalandspatialpatterns.Suchpatternsincludethedaily,seasonalandyearlyshiftsinload
andintermittentrenewablegeneration.Inaddition,thespatialdiversityofindividualwind
projectsaffectsthesystemwidewindgenerationpattern.Thesesystemiccharacteristicsare
discussedinthefollowingsections.
3.1.1. Daily and Seasonal Variations
Inthissection,theoverallimpactofintermittentrenewablegenerationonnetloadisexamined.
Figure3showsanaverageJulydayfor2003withtheloadscaledto2010levels.Thedarkblue
trace(leftyaxisscale)isthetotalCaliforniaISOload.Theshapeisthatofatypicalsummer
diurnalpattern,includingrelativelyhighloadsatmiddayandaneveningloadknuckle.
Thegreentrace(rightyaxisscale)showstheaveragepowerproductionofallCaliforniawind
projectsinthe2010Tcase(7,500MWat98sites),alsoforJuly2003.Thewindpowershowsa
typicalsummerdiurnalpattern,withrelativelylowergenerationmidday,pickingupinthe
afternoon.Theorangetrace(rightyaxisscale)showstheaveragesolarpowerproduction
againforallsolarprojects(PVandconcentrating)inthe2010Tcase(2,200MW).Asexpected,
thesolarproductionpeaksatmidday.
Thelightbluetraceshowsthetotalnetload,i.e.,loadminusthewindandsolargeneration
(abbreviatedasLWS).Thisnetloadmustbeservedbyothergeneratingresources.Notethat
thewindandsolartendtocomplementeachother,withtheresultoflargelymaintainingthe
loadaloneshapeatareducedMWlevel.
24
0
10000
20000
30000
40000
50000
1 5 9 13 17 21
0
1000
2000
3000
4000
5000
Average Load
Average L-W-S
Average Wind
Average Solar
Hour
L
o
a
d

(
M
W
)
W
i
n
d

&

S
o
l
a
r

(
M
W
)

Figure 3. Average System-wide Daily Load, Wind, Solar and Net Load Profiles of July 2003.
Theaveragebehaviormasksthedaytodaydifferencesintotalloadandtotalintermittent
generationproduction.Figure4showsthedailyprofilesofalldaysincludedintheaverages
showninFigure3.Thegeneralshapeoftheloadprofilesissimilar,withtheamplitudevarying
aroundtheaveragepeakbyabout+5,000to10,000MW.Thewindshapesalsoallhavea
diurnalpatternsimilartotheaveragewiththeamplitudevaryingaroundtheaveragepeakby
about+2,000to2,000MW.ThesolarvariabilityisdifficulttodistinguishgiventheMWscaleof
Figure4.Hence,Figure5showsonlythesolartemporalpatternofalldaysofJuly2003.The
generalshapeofthesolarprofilesissimilar,withtheamplitudetheamplitudevaryingaround
theaveragepeakbyabout+200to600MW.Thedaytodaysolarvariationissomewhatless
thanthatofthewind.
25
10000
25000
40000
55000
1 5 9 13 17 21
0
5000
10000
15000
L
o
a
d

(
M
W
)
W
i
n
d

&

S
o
l
a
r

(
M
W
)
Hour
Average Load
Average Wind
Average Solar
10000
25000
40000
55000
1 5 9 13 17 21
0
5000
10000
15000
L
o
a
d

(
M
W
)
W
i
n
d

&

S
o
l
a
r

(
M
W
)
Hour
10000
25000
40000
55000
1 5 9 13 17 21
0
5000
10000
15000
L
o
a
d

(
M
W
)
W
i
n
d

&

S
o
l
a
r

(
M
W
)
Hour
Average Load
Average Wind
Average Solar
Average Load
Average Wind
Average Solar
Figure 4. All Systemwide Daily Load, Wind, and Solar Profiles for July 2003.
S
o
l
a
r

(
M
W
)
Hour
0
500
1000
1500
2000
0 4 8 12 16 20 24
Average Solar Average Solar
Figure 5. All Systemwide Daily Solar Profiles for July 2003.

26
Theaverageload,wind,solarandnetloadprofilesforJanuary2002areshowninFigure6.The
dailypeakloadissignificantlylessthanthatobservedinJuly,andtheeveningpeakloadmay
bethelargestloadoftheday.Notethattheaveragewindcharacteristicisflatterthaninthe
summer,andissomewhatmorecoincidentwiththedailyloadshape.
Alldailyload,windandsolarprofilesforJanuary2002areshowninFigure7andFigure8.
Figure7showsdailyprofilesfortheloadandwind.Thegeneralshapeoftheloadprofilesis
similar,withtheamplitudevaryingaroundtheaveragepeakbyabout+3,000to3,000MW.
ThisrangeislessthanthatobservedinJuly.Thewindshapesalsotendtowardsadiurnal
patternsimilartotheaverage,butwithsubstantiallygreaterdaytodayvariabilitythan
observedinJuly.
Figure8showsthedailysolartemporalpatternofalldaysinJanuary2002.Thegeneralshapeof
thesolarprofilesissimilar,withtheamplitudevaryingaroundtheaveragepeakbyabout+400
to700MW.Asexpected,thesolarproductionislowerduringJanuarythanJuly.
0
10000
20000
30000
40000
50000
1 5 9 13 17 21
0
1000
2000
3000
4000
5000
Average Load
Average L-W-S
Average Wind
Average Solar
L
o
a
d

(
M
W
)
W
i
n
d

&

S
o
l
a
r

(
M
W
)
Hour

Figure 6. Average Systemwide Daily Load, Wind, Solar, and Net Load Profiles of January 2002.
27
Average Load
Average Wind
0
10000
20000
30000
40000
50000
1 5 9 13 17 21
0
2000
4000
6000
8000
10000
L
o
a
d

(
M
W
)
W
i
n
d

(
M
W
)
Hour
Average Load
Average Wind
Average Load
Average Wind
0
10000
20000
30000
40000
50000
1 5 9 13 17 21
0
2000
4000
6000
8000
10000
L
o
a
d

(
M
W
)
W
i
n
d

(
M
W
)
Hour
Average Load
Average Wind
Average Load
Average Wind
0
10000
20000
30000
40000
50000
1 5 9 13 17 21
0
2000
4000
6000
8000
10000
L
o
a
d

(
M
W
)
W
i
n
d

(
M
W
)
Hour
Average Load
Average Wind
Figure 7. All Systemwide Daily Load and Wind Profiles for January 2002.
S
o
l
a
r

(
M
W
)
Hour
0
500
1000
1500
2000
0 4 8 12 16 20 24
Average Solar Average Solar
Figure 8. All Systemwide Daily Solar Profiles for January 2002.

28
3.1.2. Spatial Variations
Thepoweroutputofasinglewindplantcanoccasionallychangesubstantiallyinarelatively
shortperiodoftime.Forsmallsystems,whereasingleplantconstitutesasignificantpercentage
ofthetotalgenerationatagivenpointintime,thiscancreateoperationalproblems.Inlarger
systems,wheretheoutputofmultiplewindplantsmustbecoincidenttocausesubstantial
impactonthebulksystem,thespatialdiversityoftheplantsbecomesimportant.
Figure9showstheproductionofallwindplantsinthe2010Tscenarioforasingleday.Thetwo
heavycurvesrepresentthetotalforallwindplants(blue)andthetotalforallwindplantsinthe
Tehachapiregion(green).Therightyaxisscaleappliesforthesetotals.Theleftyaxisscale
appliestotheindividualwindplants.
Figure10showstheindividualproductionforonlythewindplantsintheTehachapiregion.A
totalof40plantsinwindzone8areincludedintheTehachapiregion.Again,therightyaxis
scaleappliestothetotal,andtheleftyaxisscaleappliestotheindividualwindplants.
Individualplantsmayexhibitsubstantialhourtohourchangesinoutput.However,theplants
arewidelydistributedaroundthestate,whichevensoutthefastvariability,leavinganoverall
diurnalpatternofproduction.ThisobservationalsoholdsfortheTehachapiregion.The
4,200MWofprojectsintheTehachapiregionisdistributedoveranareaofapproximately
500,000acres,asdiscussedintheAWSTruewindreport[7].Thetotaloutputoftheindividual
plantsintheTehachapiregionisnottemporallycoincident,andsubstantialbenefitsfrom
spatialdiversityareachieved.
I
n
d
i
v
i
d
u
a
l

W
i
n
d

P
l
a
n
t

(
M
W
)
Hour
0
50
100
150
200
250
300
1 5 9 13 17 21
0
1000
2000
3000
4000
5000
6000
T
o
t
a
l

W
i
n
d

(
M
W
)
Total Wind
Total Tehachapi Wind
Figure 9. All Individual California Wind Plant Profiles for July 21, 2003.
29

0
50
100
150
1 5 9 13 17 21
0
1000
2000
3000
I
n
d
i
v
i
d
u
a
l

T
e
h
a
c
h
a
p
i

W
i
n
d

P
l
a
n
t

(
M
W
)

Hour
T
o
t
a
l

T
e
h
a
c
h
a
p
i

W
i
n
d

(
M
W
)

Total Tehachapi Wind
Figure 10. All Individual Tehachapi Region Wind Plant Profiles for July 21, 2003.
3.1.3. Yearly Variation and Penetration Relative to System Load
Systemloadpatternsdependonweather,demographic,economicandotherfactors.System
loadsvaryfromyeartoyear,beyondtheoveralltrendofeconomicallydrivenannualload
growth.Productionfromintermittentrenewablegenerationwillalsoexhibitvariationfrom
yeartoyear.
Onewaytolookatannualproductioniswithdurationcurves.Durationcurvesshowallthe
hours(normally8760)ofayear,sortedfrommaximumtominimum.Thisprovidesaviewof
notonlythemaximum(extremeleft)andminimum(extremeright),butalsooftheamplitude
forallhours.Figure11showsthreeload,threewindandthreesolardurationcurvesforthe
2010Xscenario(12,500MWofwind,2,600MWofsolar).Eachcurverepresentsadifferentstudy
year(2002,2003,2004),asdefinedbytheloadshape.
TheloaddurationcurvesarethethreeuppertracesinFigure11.Theleftyaxisscaleappliesto
thesecurves.Therightyaxisscaleappliestothewindandsolarproductioncurves.Thewind
productionrarely(<1%ofhours)exceeds10,000MW,andneverquitedropstozero.As
expected,thesolarproductioniszeroforabouthalfthehoursintheyear.Notethatthe
differencebetweenthethreeloadcurvesisgreaterthanthedifferencesbetweeneitherthewind
curvesorthesolarcurves.Thissuggeststhattheyeartoyearvariabilityinoverallenergy
productionfromtheintermittentrenewablesisrelativelylowerthantheyeartoyearvariability
inoverallloadenergyconsumed.
30
10000
15000
20000
25000
30000
35000
40000
45000
50000
55000
60000
1 501 1001150120012501300135014001450150015501600165017001750180018501
Hour
M
W
0
2000
4000
6000
8000
10000
12000
14000
16000
X-A, CAISO load, actual, MW X-B, CAISO load, actual, MW X-C, CAISO load, actual, MW
X-A, Total Actual Solar X-B, Total Actual Solar X-C, Total Actual Solar
X-A, Total Actual Wind X-B, Total Actual Wind X-C, Total Actual Wind
A = 2002 load shape
B = 2003 load shape
C = 2004 load shape
Figure 11. Load, Wind, and Solar Duration Curves for 2010X Scenario.
Figure11showedtheindividualdurationcurvesforload,windandsolaroverthethreestudy
years.Asnotedabove,thebalanceofthesystemgenerationportfoliomustserveanyloadnot
metbytheintermittentrenewables.Thus,thenetloaddurationcurvesaremoreilluminating.
Figure12showsthreedurationcurves.Theuppercurve(blue)representsthetotalCalifornia
ISOload,themiddlecurve(orange)representsthenetload(LWS)forthe2010Tscenario
(7,500MWofwind,1,900MWofsolar),andthebottomcurve(pink)representsthenetloadfor
the2010Xscenario(12,500MWofwind,2,600MWofsolar).Forthissetofcurves,allthree
yearsofdataareincluded,sothexaxisrangeisabout26,300hours(3years*8760hours/year+
24hoursinaleapday).
Theverticalgridlinesdividethetracesintotenequalpartsstatisticalbinsof1/10
th
ordeciles
ofthetotalsample.Thesetenbinsareusedtoparsethedatainsubsequentsections.Thetop10
percent(peak)loadhoursareincludedintheleftmostbin,B#1.Similarly,thebottom10percent
(light)loadhoursareincludedintherightmostbin,B#10.
Severalobservationscanbedrawnfromthesecurves.First,formostofthedistribution(deciles
2through9)the2010Tnetloadisabout3,000MWlowerthanthe2010Tloadalonecase.
Similarly,the2010Xnetloadisabout5000MWlessthanthe2010loadalonecase.Second,the
intermittentrenewableshaveamodestbeneficialimpactonthepeakload,reducingitby1,342
and2,761MWforthe2010Tand2010Xcases,respectively.Finally,theintermittentrenewables
31
significantlyreducetheminimumload.ThehorizontallineinFigure12highlightsthefactthat
thereareasignificantnumberofhoursforwhichthenetloadislesstheminimumloadalone.
10000
15000
20000
25000
30000
35000
40000
45000
50000
55000
0 2630 5260 7890 10520 13150 15780 18410 21040 23670 26300
Hours
M
W
CAISO load, actual, MW
2010X Total Actual L-W-S
2010T Total Actual L-W-S
B#1 B#2 B#3 B#4 B#5 B#6 B#7 B#8 B#9 B#10
Figure 12. 2010 Hourly Load and Net Load Duration Curves for 3 Years.
Theintermittentrenewableimpactonnetloadatlightloadconditionsisshowinfurtherdetail
inFigure13.Thisfigureshowstheloadandnetloaddurationcurvesforeachofthethreestudy
yearsforthe2010Xscenario,forthehoursfrom4,000to8,760.Forallsixdurationcurves,there
isasharpdownturninthelastonehundredorsohoursofeachyear.Forthenetload(LWS)
curves,thesehourscorrespondtocoincidentperiodsoflowloadandhighwind.Theseextreme
hoursarelikelytorepresentanoperationalchallenge,andareexploredfurtherinsubsequent
sectionsofthisreport.
32
10000
12000
14000
16000
18000
20000
22000
24000
26000
28000
30000
4000 4500 5000 5500 6000 6500 7000 7500 8000 8500
Hour
M
W
X-A, CAISO load, actual, MW X-B, CAISO load, actual, MW X-C, CAISO load, actual, MW
X-A,Total Actual L-W-S X-B, Total Actual L-W-S X-C, Total Actual L-W-S
Figure 13. Detail of Load and Net Load Duration Curves for 2010X Scenario.
Overall,about6%ofhoursinthe2010Tcasehaveaminimumnetloadbelowtheloadalone
minimum,andabout20%ofhoursinthe2010Xcasehaveaminimumnetloadbelowtheload
aloneminimum.Further,theabsoluteminimumisreducedby4734MWand8233MWforthe
2010Tand2010Xcases,respectively.
Fromaplanningperspective,itislikelytobeuneconomictodesignthesystemtohandlethe
absoluteminimumnetload.Rather,atsomepointitwillbeeconomictomodifythenetloadso
astosetafloorontheminimum.ThetwodurationcurvesofFigure12canbeexaminedthis
way.Fortheexpected2010renewablesprofile(scenario2010T),thenetloaddropsbelow
20,000MWfor618hoursoverthethreeyearsabout2.4%ofthetime.Thisrepresentsabout
of1%ofthetotalenergyintermittentrenewableenergy.The2010Tnetloaddropsbelow
18,000MWfor35hoursoverthethreeyears.
Fortheextreme2010Xcase,netloadlessthan20,000MWoccursfor2,849hoursoverthethree
years,accountingfor4.4%oftheenergy.For2010X,netloadlessthan18,000occurs1,105hours
overthreeyears,accountingfor1.2%ofenergy.However,Figure13showsthattheminima
varysignificantlyfromyeartoyear.Thesecasesshowbroadlythataminimumgenerationlevel
ontheorderof18,000to20,000MWwillcoverthevastmajorityofoperationconditionsinthe
2010timeframe.Occasionaldropsbelowtheselevelsmayneedtobehandledbyothermeans,
33
whichcouldincludecurtailment.Resultspresentedlaterforyear2020reinforcethis
observation.
2006 and 2010 Penetration
Renewablepenetrationisterminologyusedwithintheindustrytodescribetheamountof
renewablegenerationrelativetotherestofthesystem.However,thetermhasnostandard
definition.Fromanoperationsperspective,thefractionofthetotalloadbeingsuppliedatany
givenhourisausefulmeasureofpenetration.Theimpactonoperationsisgenerallylowerat
lowerpenetration,althoughotherconsiderationsareimportantaswell.
Figure14showswindproductionandpenetrationdurationcurvesoverthethreestudyyears
forthe2010Xscenario.AswasapparentinFigure11,thetotalwindproductionrarelyexceeds
10,000MW(<1%ofthetime)andneverreachesthenominalmaximumof12,500MW.Forabout
5%ofhours,thewindpenetration(windMW/loadMWforthatspecifichour)exceeds30%,
withonlyabout10hoursexceeding40%inthreeyears.
Figure15showssimilardurationcurvesforthesolargeneration.Thesolarpenetrationexceeds
5%about5%ofhours,andislessthan1%abouthalfofallhours.
0
2000
4000
6000
8000
10000
12000
0 2630 5260 7890 10520 13150 15780 18410 21040 23670 26300
Hour s
M
W
0%
10%
20%
30%
40%
50%
60%
Total Actual Wind (MW)
Total Actual Wind Penetration (%)
P
e
n
e
t
r
a
t
i
o
n
W
i
n
d

M
W
Figure 14. 2010X Wind Production and Penetration Duration Curves.
34
0
500
1000
1500
2000
2500
0 2630 5260 7890 10520 13150 15780 18410 21040 23670 26300
Hour s
M
W
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
10%
Total Actual Solar (MW)
Total Actual Solar Penetration (%)
S
o
l
a
r

M
W
P
e
n
e
t
r
a
t
i
o
n
Figure 15. 2010X Solar Production and Penetration Duration Curves.

Anotherwaytolookatpenetrationistoconsideritinthecontextoftotalsystemloadlevels.
Figure16isthefirstofseveralgraphsinwhichtheresultsaregroupedbythedecilesdescribed
previously.Forthesegraphs,thepenetrationfiguresaretheaverageforeachdecile;i.e.the
averagewindproductionforallhoursinthatloaddeciledividedbytheaverageloadofthat
decile.Threelevelsofwindgenerationareplotted,whichcorrespondtothe2006,2010Tand
2010Xscenarios.
Notethatwindpenetrationtendstobehigheratlighterload.ThisisconsistentwithFigure3
andFigure6.Asexpected,windpenetrationlevelsincreaseaswindresourcesareaddedtothe
system.The2006penetrationofabout4%atlightloadincreasestoabout16%forthe2010T
scenario,andtoabout28%forthe2010Xscenario
35
0%
5%
10%
15%
20%
25%
30%
1 2 3 4 5 6 7 8 9 10
Bin
2006 Wind Penetration
2010T Wind Penetration
2010X Wind Penetration
Decile
%

Figure 16. 2006 and 2010 Hourly Average Wind Penetration by Decile.
Averagesolarpenetrationbydecileforthe2006,2010Tand2010XscenariosisshowninFigure
17.Again,averagepenetrationinadecilewascalculatedasaveragesolarproductionforall
hoursinthatloaddeciledividedbytheaverageloadofthatdecile.
Notethatsolarpenetrationispositivelycorrelatedwithsystemload,producingthemostpower
duringpeakloadperiods.Asexpected,penetrationlevelsincreaseassolarresourcesareadded
tothesystem.The2006penetrationofabout0.6%atheavyloadincreasestoabout2.1%forthe
2010Tscenario,andtoabout3.2%forthe2010Xillustrativescenario.
36
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
3.5%
1 2 3 4 5 6 7 8 9 10
Bin
2006 Solar Penetration
2010T Solar Penetration
2010X Solar Penetration
Decile
%

Figure 17. 2006 and 2010 Hourly Average Solar Penetration by Decile.
2020 Yearly Variation and Penetration
Substantialloadgrowthisexpectedbetween2010and2020.Incontrast,theincreasein
intermittentrenewablesbetweenthe2010Xand2020scenariosisrelativelysmall.Figure18
showstheloadandnetloaddurationcurvesforthethreestudyyears(2002,2003,2004)forthe
2020scenario.ComparedtoFigure12andFigure13,therelativeimpactoftheintermittent
renewableisless.Theminimumnetloadlevelofabout20,000MWissignificantlyhigherthan
thatinthe2010scenarios,andisonlyslightlylessthantheminimumloadalonefor2010.This
minimumnetloadreinforcestheobservationsmadeearlierforthe2010cases:aminimumnet
loadcapabilityof20,000MWwillsuittherequirementsofthe2020system.Intheeventthat
renewablesgrowthoutpacesloadgrowthintheyearsleadingto2020,adeeperminimumor
temporarymitigationmeasuresmayberequired.
37
10000
20000
30000
40000
50000
60000
70000
1 501 1001 1501 2001 2501 3001 3501 4001 4501 5001 5501 6001 6501 7001 7501 8001 8501
Hour
M
W
A, CAISO load, actual, MW B, CAISO load, actual, MW C, CAISO load, actual, MW
A,Total Actual L-W-S B, Total Actual L-W-S C, Total Actual L-W-S
Figure 18. 2020 Hourly Load and Net Load Duration Curves.
Figure19showsthe2020averagewindandsolarpenetrationbydecile.Again,average
penetrationinadecilewascalculatedasaveragewindorsolarproductionforallhoursinthat
loaddeciledividedbytheaverageloadofthatdecile.Windpenetrationdropsfromthe2010X
case,butsolarpenetrationincreasesduetothesubstantialadditionsofsolarfromthe2010X
scenariotothe2020scenario.
A=2002LoadShape
B=2003LoadShape
C=2004LoadShape
38
0%
5%
10%
15%
20%
25%
1 2 3 4 5 6 7 8 9 10
Bin
M
W
Wind Penetration
Solar Penetration
Decile
A
v
e
r
a
g
e

P
e
n
e
t
r
a
t
i
o
n

Figure 19. 2020 Hourly Average Wind and Solar Penetration by Decile.
3.2. Hourly Variability
Theproductionofpowerfromtheintermittentresourcesvariescontinuously.Thisvariability
contributestothevariabilitythatexistswithloads.Thepowersystemmusthavetheabilityto
notonlysatisfythenetloaddemandatanypointintime,itmustalsohavetheflexibilityto
successfullymovefromconditiontocondition.
Thechangesinpowerthatcharacterizeloadandintermittentrenewablesoccurinamultitude
oftimeframes.Inordertoquantifyvariability,thesechangesareanalyzedstatistically.
Changesinloadandnetload(loadminuswindminussolar)relatetothreetypesofoperating
requirements:scheduleflexibility,loadfollowingandregulation.Thesegroupings,whilenot
entirelyindependent,allowformeaningfulquantificationofsystembehavior.Eachofthethree
groupingarerelatedtoaspecifictimeinterval,overwhichthechange(delta)ismeasured.
Specifically:
1hrDelta=Scheduleflexibility
5minDelta=LoadFollowingcapability/EconomicDispatch
1minDelta=Regulation
Distributionofdeltasacrossthedatasetsexaminedinthisstudytendtobenormal
distributions.Therefore,standarddeviation()ofthedistributionsbecomesausefulmeasureof
variability.Highervaluescorrespondtohighervariability.3timesstandarddeviation()isa
proxyformaneuverability/flexibilityrequirements;thevastmajority(99.7%)ofeventsfall
39
within+/3(inanormalpopulation).Increasein3isonemeasureofrequirementfor
additionalmaneuverability/flexibilityduetoincreasedvariability
Figure20showsvariousprofilesforanillustrativesummerday.Thesolidcurvesshowthat
dayshourlyload,totalloadminuswind,andnetload(loadminuswindminussolar)profiles.
Thesetracesusetheleftyaxisscale.LiketheaveragedayofFigure3,thecombinationofwind
andsolarproductiontendtoreducetheoverallnetload.Thecurveswithsquaresymbols
representthe1hourdeltasforeachofthethreeprofiles.A1hourdeltaisdefinedasthechange
fromonehourtothenext.Therightyaxisscaleappliestothe1hourdeltas.
Notetherapidmorningloadrisefromthe7
th
to10
th
hours.Thisparticularsummermorninghas
relativelyhighrateofmorningloadrise,whichreachesamaximumofabout3,600MW/hrin
the7
th
hour(bluelinewithsquaresymbol).Duringthissametimeperiod,thetotalwindpower
isdeclining,whichincreasestherateofnetloadrise.Themaximumrateofrise,consideringjust
loadandwind,increasesbyabout600MW/hrto4,200MW/hr(greenlinewithsquaresymbol).
Alsoduringthistimeperiod,solargenerationisincreasing.Thistendstooffsetthedecreasein
wind.Thenetload(orangeline,LWS)exhibitsamaximumrateofloadrise(orangelinewith
squaresymbol)ofabout4,000MW/hr.Thisisanincreaseofabout400MW/hrorabout10%
overtheloadalone.Intheafternoon,thedecreaseinsystemloadandincreaseinwind
generationreversesthesituation.Thisresultsinafasterdeclineintheprofiles(solidlines)and
morenegative1hourdeltas(lineswithsquaresymbols).Forthisparticularday,thefastestrate
ofloaddecline(about4,200MW/hr)occursatthe22
nd
hour.Theaggregateimpactofthe
intermittentrenewablesinthistimeperiodistoreducethatrateofdeclinetoabout
3,500MW/hr.Ingeneral,fasterratesofloadriseanddeclinearemorechallengingforthe
system.Thus,theimpactofintermittentrenewablesonthehourtohourchangesinloadis
examinedfurther.
40
20000
25000
30000
35000
40000
45000
50000
1 5 9 13 17 21
-5000
-2500
0
2500
5000
Load
Load-Wind
Load-Wind-Solar
1Hr Load Deltas
1Hr L-W Deltas
1Hr L-W-S Deltas
H
o
u
r
l
y

d
e
l
t
a
s

(
M
W
/
h
r
)

L
o
a
d

(
M
W
)

Hour

Figure 20. Hourly Profiles and 1-Hour Deltas for an Example July 2002 Day.
3.2.1. 2006 Variability Relative to Load Level
Examinationofthehourtohourvariabilityofthesystemwithandwithoutwindandsolar
providesthemostinsight.Table5showsthe2006hourlyloadstatisticalresults,groupedinto
tendecilesonthebasisofloadalone.The10percentofpeakloadhoursareincludedindecile1,
andthe10percentoflightloadhoursareincludedindecile10.Thetableshowsthemaximum,
minimum,andaverageloadaswellasthestandarddeviation(),maximum,minimum,and
average1hourloaddeltas.
Table6showsthe2006hourlynetload,windandsolarstatisticalresults,groupedintoten
decilesonthebasisofLWS.Thefirstrowsinthetableshowthemaximum,minimum,and
averagenetload(LWS).ThenextrowshowstheaverageloadaloneineachLWSdecile.This
averageisdifferentfromtheaverageloadshowninTable5becausedecilessortedbyLWSare
differentfromdecilessortedbyloadalone.Theremainingrowsshowthestandarddeviation
(),maximum,minimum,andaverage1hournetloaddeltasaswellastheaveragewind
productionandpenetration,andtheaveragesolarproductionandpenetration.
All2006hourlyvariabilityfigurescitedintheremainderofthereportarefromthesetables.
41

Table 5. 2006 Hourly Load Statistics (MW).
Load Decile
Load Trait 1 2 3 4 5 6 7 8 9 10
All
Year
Maximum 48113 36776 33503 31859 30655 29430 27776 26093 24615 23074 48113
Minimum 36779 33505 31860 30656 29430 27777 26093 24616 23074 19443 19443
Average 40163 34995 32598 31240 30063 28657 26911 25353 23840 22058 29587
Delta 1254 1555 1471 1363 1555 1816 1663 1436 1075 669 1436
Maximum Delta 4529 4854 4776 4375 6123 6071 3824 3245 2862 1706 6123
Minimum Delta -4334 -4446 -4382 -4372 -5122 -4017 -3535 -3200 -2868 -2567 -5122
Average Delta 225 317 209 130 117 -19 -178 -262 -309 -229 0.2

Table 6. 2006 Hourly Net Load and Intermittent Renewable Statistics (MW or %).
Net Load Decile
Net Load (L-W-S)
Trait
1 2 3 4 5 6 7 8 9 10
All
Year
Maximum 47736 35985 32800 31197 29994 28722 27068 25376 23849 22407 47736
Minimum 35988 32801 31197 29994 28722 27069 25377 23851 22407 18567 18567
Average 39398 34243 31938 30591 29386 27927 26188 24620 23136 21374 28880
Load Alone Average 40142 34971 32563 31228 30075 28681 26895 25349 23840 22133 29587
Delta 1294 1547 1456 1399 1587 1829 1700 1446 1049 699 1451
Maximum Delta 4924 4729 4857 4581 6091 5981 3947 3241 2914 2448 6091
Minimum Delta -4295 -4533 -4450 -4592 -5155 -4174 -3956 -3452 -2814 -2613 -5155
Average Delta 242 320 183 149 129 -53 -146 -265 -307 -250 0.1
Wind Average 492 571 521 549 611 693 663 705 699 746 625
Wind Penetration 1.2% 1.6% 1.6% 1.8% 2.0% 2.4% 2.5% 2.8% 2.9% 3.4% 2.1%
Solar Average 253 156 105 88 77 61 44 25 6 13 83
Solar Penetration 0.6% 0.4% 0.3% 0.3% 0.3% 0.2% 0.2% 0.1% 0.0% 0.1% 0.3%
3.2.2. 2010 Variability
Table7showsthe2010Thourlyloadstatisticalresults,groupedintotendecilesonthebasisof
loadalone.Table8showsthe2010Thourlynetload,windandsolarstatisticalresults,grouped
intotendecilesonthebasisofLWS.Table9showsthe2010Xhourlynetload,windandsolar
statisticalresults,groupedintotendecilesonthebasisofLWS.Theformatofthesetablesis
thesameasdescribedaboveforTable5andTable6.
All2010hourlyvariabilityfigurescitedintheremainderofthereportarefromthesetables.
42

Load Decile
Load Trait 1 2 3 4 5 6 7 8 9 10
All
Year
Maximum
52761 40329 36738 34936 33616 32272 30459 28613 26993 25302 52761
Minimum
40331 36741 34937 33617 32272 30460 28613 26994 25302 21321 21321
Average
44042 38375 35747 34258 32966 31425 29510 27801 26143 24189 32445
Delta
1375 1706 1613 1495 1705 1992 1824 1575 1179 733 1575
Maximum Delta
4967 5323 5237 4797 6714 6657 4193 3559 3139 1871 6714
Minimum Delta
-4753 -4875 -4806 -4794 -5617 -4405 -3877 -3509 -3145 -2815 -5617
Average Delta
247.0 347.9 229.5 142.4 128.6 -22.1 -195 -287 -339 -251 0.2
Table 8. 2010T Hourly Net Load and Intermittent Renewable Statistics (MW or %).
Net Load Decile
Net Load (L-W-S)
Trait
1 2 3 4 5 6 7 8 9 10
All
Year
Maximum 51418 37348 33909 31991 30432 28854 27211 25537 23953 22222 51418
Minimum 37349 33909 31991 30432 28856 27211 25538 23953 22224 16587 16587
Average 41210 35434 32889 31210 29655 28033 26378 24741 23102 20759 29341
Delta 1541 1786 1670 1694 1818 1870 1747 1540 1211 933 1623
Maximum Delta 6234 5883 5725 6108 5144 6312 4852 4283 3464 2939 6312
Minimum Delta -4752 -5236 -4858 -5283 -5713 -4507 -4303 -4100 -3728 -3427 -5713
Average Delta 298 335 243 148 68 -53.3 -151 -303 -286 -298 0.2
Wind Average 1707 2018 2060 2388 2811 3045 2800 2806 3120 4039 2680
Solar Average 927 638 560 549 489 424 319 182 111 50 425
43

Table 9. 2010X Hourly Net Load and Intermittent Renewable Statistics (MW or %).
Net Load Decile
Net Load (L-W-S)
Trait
1 2 3 4 5 6 7 8 9 10
All
Year
Maximum 50000 35686 32437 30415 28648 26841 25134 23518 21830 19795 50000
Minimum 35687 32438 30418 28648 26842 25135 23520 21831 19798 13088 13088
Average 39511 33841 31392 29516 27751 25961 24319 22682 20860 18020 27385
Delta 1602 1792 1741 1859 1955 1911 1767 1591 1364 1081 1704
Maximum Delta 6672 6168 6607 6132 7219 6220 5541 4974 4384 3250 7219
Minimum Delta -4671 -5690 -5224 -5543 -5986 -5046 -4769 -4649 -4422 -4141 -5986
Average Delta 327.8 328.0 215.6 175.2 89.0 -72.4 -138.9 -296.9 -290.2 -335.8 0.1
Wind Average 2649 3055 3177 3875 4476 4756 4824 5039 5729 7230 4481
Wind Penetration 6.1% 8.1% 9.0% 11.4% 13.6% 15.2% 16.3% 18.0% 21.4% 28.5% 14%
Solar Average 1389 910 761 698 600 521 397 243 163 100 578
Relative to Load Level

Figure21showsanopenhighlowclosestockchartfortheextremeintermittentrenewables
2010Xcasehourlyloaddeltas(red)andhourlynetloaddeltas(yellow)bydecile.Thetopofa
barrepresentstheaveragevalueofhourlydeltaplusthestandarddeviation.Thebottomofa
barrepresentstheaveragevalueofhourlydeltaminusthestandarddeviation.Thetopofa
verticallinerepresentsthemaximumhourlydeltas,andthebottomofaverticallinerepresents
theminimumhourlydelta.
Thechartshowsthatwindandsolarhavethemostimpactonhourlyvariation,including
extremes,underlightloadconditions.Thestandarddeviationoftheloadalonevariabilityfor
the10thdecileis733MW,increasingby348MWto1081MWwithwindandsolar.Thismeans
thatwithinthatlightloaddecile,thereisa99.7%expectation(3)thathourtohourchanges
willbelessthan2199MWwithoutwindandsolar.The3expectationwithwindandsolaris
forthehourlychangeslessthan3240MW.Undertheselightloadconditions,thelargestload
aloneincreaseis1871MWandthelargestnetloadincreaseis3250MW.Thus,themaximum
loadaloneincreaseislessthanthe3expectation,andthemaximumnetloadincreaseis
slightlymorethanthe3expectation.Thelargestloadalonedecreaseis2815MWandthe
largestnetloaddecreaseis4141MW.Boththeloadaloneandnetloaddecreasesaremorethan
the3expectationunderlightload.
44
Load
L-W-S
Load
L-W-S
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10
M
W
Decile

Figure 21. 2010X Hourly Load and Net Load Delta Stock Chart by Decile.
Figure22showsthestandarddeviationofloadandnetloadhourlydeltasforthe2006,2010T
and2010Xscenarios.Lefttoright,thefivebarsineachdecilerepresentthestandarddeviations
of2006loaddelta,2006netloaddelta,2010loaddelta,2010Tnetloaddeltaand2010Xnetload
delta.Thechartshowsthatwindandsolarinboth2006and2010scenariohavethemostimpact
onhourlyvariation,includingextremes,underlightloadconditions.
45
0
500
1000
1500
2000
2500
1 2 3 4 5 6 7 8 9 10
2006 Sigma Delta L
2006 Sigma Delta L-W-S
2010 Sigma Delta L
2010T Sigma Delta L-W-S
2010X Sigma Delta L-W-S
Decile
M
W

Figure 22. 2006 and 2010 Standard Deviation of Hourly Load and Net Load Deltas.
Relative to Time-of-Day
Figure23showstheaveragehourlywindandsolarpenetrationforeveryhourofthedayforthe
extreme2010Xscenario.Thewindandsolarpenetrationisdefinedasaverageproduction
dividedbyaverageloadatthespecifiedhourforanentireyear.Windpenetrationdipsdownto
about8percentinthemorning,andrisesbackuptoabout21percentatmidnight.Solar
penetrationreachesitspeakof4percentatmidday.
46
0%
5%
10%
15%
20%
25%
1 3 5 7 9 11 13 15 17 19 21 23
Hour
P
e
n
e
t
r
a
t
i
o
n
Solar Penetration
Wind Penetration
Figure 23. 2010X Average Hourly Wind and Solar Penetration by Hour of Day.
Periodsofpeakdemandandrapidriseinloadaregivenspecialattentionbysystemoperators.
Thesummermorningloadrise,especiallyduringperiodsofsustainedhotweather,presents
oneofthemoreseveretestsforthesystem.
TheopenhighlowclosestockchartsofFigure24throughFigure26weredevelopedto
identifytheimpactofwindandsolaroverthecourseofasummerorwinterday,respectively.
Thesefiguresshowthe2010Xhourlyloaddeltas(red)andhourlynetloaddeltas(yellow)by
timeofday.Asdescribedpreviously,thetopofabarrepresentstheaveragevalueofhourly
deltaplusthestandarddeviation.Thebottomofabarrepresentstheaveragevalueofhourly
deltaminusthestandarddeviation.Thetopofaverticallinerepresentsthemaximumhourly
deltas,andthebottomofaverticallinerepresentstheminimumhourlydelta.
Figure24showshourlydeltasforthe2010Xloadandnetloadscenariosbasedonthethree
yearsof2002,2003and2004.Themaximumoverallhourlyloaddeltaisabout6,500MW/hr,at
7am.Themaximumoverallhourlynetloaddeltaisabout7,000MW/hr,alsoat7am.The
largestincreaseinmaximumhourlydelta,betweentheloadandnetloadcases,isabout
2,000MW/hrat5pm.Theminimumoverallhourlyloaddeltaisabout5,500MW/hr,at10pm.
Theminimumoverallhourlynetloaddeltasisabout6,000MW/hr,alsoat10pm.Thelargest
increaseinminimumhourlydelta,betweentheloadandnetloadcases,isabout1,500MW/hrat
6pm.
47
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour
M
W
Load L-W-S Load L-W-S
Figure 24. 2010X Hourly Load and Net Load Delta Stock Chart by Hour of Day.
Figure25isbasedonallJulydaysinthethreestudyyearsof2002,2003,and2004forthe2010X
scenario.DuringJulymornings,thenetloadrises500MW/hrto1000MW/hrfasterthanload
alone.Theafternoonandeveningswinginnetloadisalsomorepronouncedat6pmthanwith
loadalone.
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour
M
W

Figure 25. 2010X July Hourly Load and Net Load Delta Stock Chart by Hour of Day.
48
Figure26isbasedonallJanuarydaysinthethreestudyyearsof2002,2003,and2004forthe
2010Xscenario.ItshowsthattheaverageHolidayLightnetloadriseat6pmisabout
440MW/hror12%higherthantheaverageloadaloneriseof3600MW/hr.Themaximumnet
loadriseisabout1350MW/hror25%higherthanthemaximumloadaloneriseof5300MW/hr.
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10111213141516171819 20 21 22 2324
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10111213141516171819 20 21 22 2324
Load
L - W - S Load
L - W - S
Hour
M
W

Figure 26. 2010X January Hourly Load and Net Load Deltas Stock Chart by Hour of Day.
Sustained 3-Hour Changes
Figure20showedanillustrativesummerdayinJuly2002.Asnotedpreviously,thisparticular
summermorninghasarelativelyhighrateofmorningloadrise,whichreachesamaximumof
about3,600MW/hrinthe7
th
hour(bluelinewithsquaresymbol).Thenextfewhourshave
aboutthesamelevelofhourlyloadrise.Thepeak3hourperiodofloadriseisduringthe6
th
,7
th
and8
th
hours.The3hourloaddeltaduringthisperiodisabout9,300MW/3hrs.The3hournet
loaddeltaduringthisperiodisabout9,900MW/3hrs.Sustainedloadincreasesanddecreases
representachallengingoperatingcondition.Thus,theimpactoftheintermittentrenewableson
3hourloaddeltaswasevaluatedfurther.
Figure27shows3hourdeltasforthe2010Xload(red)andnetload(yellow)scenariosbasedon
allstudyyears(2002,2003,2004)relativetotimeofday.Themaximum3hournetloadriseat
6pmincreasedbyabout2,000MW/3hrsfromthemaximum3hourloadriseofabout
8,000MW/3hrs.At8am,themaximumnetloaddeltaalsoincreasedbyabout2,000MW/3hrs
fromthemaximum3hourloaddeltaof10,000MW/3hrs.However,theoverallmaximumnet
loaddeltaofabout12,000MW/3hrsisonly1,000MW/3hrsgreaterthantheoverallmaximum
loaddeltaofabout11,000MW/3hrs.
Similarly,theminimumnetloaddeltaat7pmincreasedbyabout3,000MW/3hrsfromthe
minimumloaddeltaabout6,000MW/3hrs.At11pm,theminimumnetloaddeltaincreasedby
49
about1,000MW/3hrsfromtheminimumloaddeltaofabout13,000MW/3hrs.Hence,the
overallminimumloaddeltahasincreasedfromabout13,000MW/3hrstoanoverallminimum
netloaddeltaofabout14,000MW/3hrs.
-14000
-12000
-10000
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
10000
12000
14000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
-14000
-12000
-10000
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
10000
12000
14000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Load

L-W-S
Load
Hour
M
W

Figure 27. 2010X 3-Hour Load and Net Load Deltas Stock Chart by Hour of the Day.
Figure28shows3hourdeltasforthe2010Xload(red)andnetload(yellow)scenariosin
January,relativetotimeofday.Asexpected,the3hourdeltasduringtheHolidayLight
eveningloadrisearelargerthanthehourlydeltas.Themaximum3hournetloadriseat6pm
increasedbyabout2,000MW/3hrsfromthemaximum3hourloadriseofabout8,000MW/3hrs.
However,theoverallmaximumisstillobservedinthemorning.At7am,themaximumnetload
deltaincreasedbyabout1,000MW/3hrsfromthemaximum3hourloaddeltaof
10,000MW/3hrs.Hence,theoverallmaximumnetloaddeltaisabout11,000MW/3hrs.
Theminimumnetload3hourdeltasincreaseinseveraltimeperiodsbyabout2,000MW/3hrs.
Theoverallminimumnetload3hourdeltaincreasedbyabout2,000MW/3hrsfromthe
minimumloaddeltaofabout8,000MW/3hrs.Hence,theoverallminimumnetloaddeltais
about10,000MW/3hrs.
50
-14000
-12000
-10000
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
10000
12000
14000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
-14000
-12000
-10000
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
10000
12000
14000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Load L-W-S Load L-W-S
Hour
M
W

Figure 28. 2010X January 3-Hour Load and Net Load Deltas Stock Chart by Hour of the Day.
Extreme Changes
Thesystemmustbepreparedforextremesofloadandrenewableconditions.Oneusefulmeans
ofcharacterizingtheseverityandfrequencyofeventsiswithdeltadurationcurves.Selectednet
loadincreaseanddecreaseexamplesarepresentedinthissection.The2004studyyearload,
windandsolarshapesarepresentedasrepresentative.Allyearswereanalyzed.
First,extremesof3hournetloadriseareexamined,sincethesemaypresentoneofthemore
severesystemchallenges.Figure29shows3hourpositivedeltadurationcurvesforstudyyear
2004andthe2010loadalone,2010Tnetload,and2010Xnetloadscenarios.Theyaxis
representsthenumberofhourswithapositivechange,ofloadornetload,thatexceedsthe
levelshowninthexaxis.Forexample,boththe2010load(greenline)and2010Tnetload(pink
line)scenarioshaveabout300hourswith3hourpositivedeltasinexcessof7,000MW.The
2010Xnetload(blueline)scenariohasabout300hourswith3hourpositivedeltasinexcessof
7,500MW.
Themaximumloadalone3hourpositivedeltaisabout9,500MW.The2010Tnetloadscenario
hasabout2hourswith3hourpositivedeltasabovethatlevel,andthe2010Xnetloadscenario
hasabout10hours.Therefore,netload3hourpositivedeltashigherthantheprojectedload
onlymaximumarerelativelyinfrequent.
51
0
50
100
150
200
250
300
350
400
450
500
6000 7000 8000 9000 10000 11000 12000
MW
T
i
m
e
s
X-C 3 Hr Delta Total Actual L-W-S
T-C 3 Hr Delta Total Actual L-W-S
T,X-C 3 Hr Delta load, actual, MW
C
o
u
n
t

o
f

H
o
u
r
s

p
e
r

Y
e
a
r
MW Increase

Figure 29. 2010 3-Hour Positive Load and Net Load Delta Duration Curves for 2004.
Figure30shows3hournegativewinddeltadurationcurvesforthe2010Tand2010Xscenarios
(2004studyyear).Theseareofinterestbecausedecreaseofwindpowerthatoccurs
simultaneouslywithloadincreaseaggravatestheoperationschallenge.Again,theyaxis
representsthenumberofhourswithanegativechangeinwindthatexceedstheMWlevelon
thexaxis.Themaximumnegative3hourwinddeltasareabout2,600MWforthe2010T
scenarioandabout4,300MWforthe2010Xscenario.ComparingFigure29andFigure30clearly
showsthatcoincidentextremeloadriseandwinddeclinearenottobeexpected.
52
0
500
1000
1500
2000
2500
3000
3500
4000
4500
-4500 -4000 -3500 -3000 -2500 -2000 -1500 -1000 -500 0
MW
T
i
m
e
s
X-C 3 Hr Delta Total Actual Wind
T-C 3 Hr Delta Total Actual Wind
C
o
u
n
t

o
f

H
o
u
r
s

p
e
r

Y
e
a
r
MW Decrease

Figure 30. 2010 3-Hour Negative Wind Delta Duration Curves for 2004 Study Year.
Rapidloaddeclinecouldalsobeproblematic.Forloaddecline,shortertimewindowsareof
concern.Figure31shows1hournegativedeltadurationcurvesforstudyyear2004andthe
2010loadalone,2010Tnetload,and2010Xnetloadscenarios.theyaxisrepresentsthenumber
ofhourswithanegativechange,ineitherloadornetload,thatexceedstheMWlevelonthex
axis.Theminimum1hournegativedeltaisabout5,400MWforallthreescenarios.
53
0
100
200
300
400
500
-6000 -5500 -5000 -4500 -4000 -3500 -3000 -2500 -2000
Delta (MW)
T
i
m
e
s
X-C 1 Hr Delta Total Actual L-W-S
T-C 1 Hr Delta Total Actual L-W-S
T,X-C 1 Hr Delta load, actual, MW
C
o
u
n
t

o
f

H
o
u
r
s

p
e
r

Y
e
a
r
MW Decrease
Figure 31. 2010 1-Hour Negative Load and Net Load Delta Duration Curves for 2004.
Figure32shows1hourpositivewinddeltadurationcurvesforthe2010Tand2010Xscenarios
(2004studyyear).Asbefore,theyaxisrepresentsthenumberofhourswithanegativechange
inwindthatexceedstheMWlevelonthexaxis.Themaximumpositive1hourwinddeltasare
about2,100MWforthe2010Tscenarioand3,500MWforthe2010Xscenario.Aswiththenet
loadrise,itisclearthatextremesofwindarenotexpectedtocoincidewithextremesofload
decline.
54
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0 500 1000 1500 2000 2500
Delta (MW)
T
i
m
e
s
X-C 1 Hr Delta Total Actual Wind
T-C 1 Hr Delta Total Actual Wind
C
o
u
n
t

o
f

H
o
u
r
s

p
e
r

Y
e
a
r
MW Increase
Figure 32. 2010 1-Hour Positive Wind Delta Duration Curves for 2004 Study Year.
3.2.3. 2020 Variability Relative to Load Level and Time-of-Day
Figure33showsanotheropenhighlowclosestockchartforthe2020loadandnetload
scenario.Theredbarsrepresenttheloadscenarioandtheyellowbarsrepresentthenetload
(LWS)scenario.Thetopofabarrepresentstheaveragevalueofhourlydeltaplusthe
standarddeviation.Thebottomofabarrepresentstheaveragevalueofhourlydeltaminusthe
standarddeviation.Thetopofaverticallinerepresentsthemaximumhourlydeltas,andthe
bottomofaverticallinerepresentstheminimumhourlydelta.
Thechartshowsthatwindandsolarhavethemostimpactonhourlyvariation,including
extremes,underlightloadconditions.Thestandarddeviationoftheloadalonevariabilityfor
the10thdecileis920MW.Itincreasesby324MWto1244MWwiththeadditionofthe
intermittentrenewables.Therefore,underlightload,thereisa99.7%expectation(3)thathour
tohourchangeswillbelessthan2760MWwiththeloadalone,and3732MWwiththe
additionalwindandsolar.Themaximumhourlydeltais2349MWforloadalone,and
3923MWfornetload.Theminimumhourlydeltais3533MWforloadalone,and5100MW
fornetload.Themaximumandminimumhourlydeltas,forbothloadandnetload,exceedthe
3expectation(99.7%)underlightloadconditions.Broadly,thesystemrequireshourly
scheduleflexibilityatlightloadofabout4000MWby2020.
55
Decile
M
W
Load
L-W-S
Load
L-W-S
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
10000
1 2 3 4 5 6 7 8 9 10
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
10000
1 2 3 4 5 6 7 8 9 10
Figure 33. 2020 Hourly Load and Net Load Deltas Stock Chart by Decile.
Figure34showshourlydeltasforthe2020loadandnetloadscenariosbasedonallJanuary
daysinthethreestudyyearsof2002,2003,and2004.Thisfigureshowsthe2020hourlyload
deltas(red)andhourlynetloaddeltas(yellow)bytimeofday.NotethattheHolidayLight
averageloadriseat6pmincreasesbyabout500MW/hr,or11%basedonanaverageloadrise
of4,500MW/hr,withtheadditionoftheintermittentrenewables.Themaximumevening
hourlydeltaincreasesbyabout1,500MW/hr,or22%basedonmaximumloadrise
6,700MW/hr,withwindandsolar.
56
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
10000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
10000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Load
L-W-S
Load
L-W-S
Hour
M
W

Figure 34. 2020 January Hourly Load and Net Load Deltas Stock Chart by Hour of Day.
Table10showsthe2020hourlyloadstatisticalresults,groupedintotendecilesonthebasisof
loadalone.The10percentofpeakloadhoursareincludedindecile1,andthe10percentof
lightloadhoursareincludedindecile10.Thetableshowsthemaximum,minimum,and
averageloadaswellasthestandarddeviation(),maximum,andminimum1hourloaddeltas.
Table11showsthe2020hourlynetload,windandsolarstatisticalresults,groupedintoten
decilesonthebasisofLWS.Thefirstrowsinthetableshowthemaximum,minimum,and
averagenetload(LWS).ThenextrowshowstheaverageloadaloneineachLWSdecile.This
averageisdifferentfromtheaverageloadshowninTable10becausedecilessortedbyLWS
aredifferentfromdecilessortedbyloadalone.Theremainingrowsshowthestandard
deviation(),maximumandminimum1hournetloaddeltasaswellastheaveragewind
productionandpenetration,andtheaveragesolarproductionandpenetration.
All2020hourlyvariabilityfigurescitedinthisreportarefromthesetables.
57

Load Decile
Load Trait 1 2 3 4 5 6 7 8 9 10
All
Year
Maximum 66215 50612 46107 43844 42188 40502 38226 35910 33876 31754 66215
Minimum 50616 46110 43846 42189 40502 38227 35910 33877 31755 26758 26758
Average 55273 48160 44862 42993 41373 39438 37035 34891 32809 30357 40719
Delta 1725 2140 2024 1876 2140 2500 2289 1976 1480 920 1977
Maximum Delta 6233 6680 6573 6021 8427 8355 5262 4466 3939 2349 6233
Minimum Delta -5965 -6118 -6031 -6016 -7049 -5528 -4866 -4404 -3947 -3533 -7049
Average Delta 310 437 288 179 161 -28 -245 -360 -426 -315 0.1
Table 11. 2020 Hourly Net Load and Intermittent Renewable Statistics (MW or %).
Net Load Decile
Net Load (L-W-S)
Trait
1 2 3 4 5 6 7 8 9 10
All
Year
Maximum 61290 44529 40692 38346 36333 34262 32233 30388 28462 26183 61290
Minimum 44529 40693 38347 36333 34262 32234 30388 28463 26185 18400 18400
Average 48928 42399 39449 37339 35309 33220 31288 29434 27383 24230 34898
Delta 1904 2203 2099 2198 2327 2276 2037 1860 1549 1244 2019
Maximum Delta 8216 7779 8261 7797 8747 6203 6483 5701 5025 3923 8216
Minimum Delta -6869 -6393 -6515 -6424 -7351 -6080 -5670 -5359 -5030 -5100 -7351
Average Delta 411 442 259 180 92 -110 -194 -352 -333 -394 0.1
Wind Average 2891 3294 3365 3972 4530 4884 4893 4919 5539 7060 4535
Solar Average 2873 1850 1651 1525 1410 1186 1021 612 405 329 1286
3.3. Intra-Hour Variability

Adetailedanalysisofselected3hourperiodsexhibitinginterestingbehaviorwasperformed.
Thedefinitionofinterestingbehaviorincluded:
Large1hourand3hourchangesinload
Highlevelsofwindandsolaroutput
Highpenetrationofwindandsolar
Lightload
Large1hourand3hourchangesinwindandsolar
Sustainedhighlevelsofwindoutputwithlowvariability
58
Specific3hourperiodswereselectedfromthetop20ofeachcategoryforvarietyinstudyyear,
season,timeofday,typeofbehavior(e.g.,largechangeorhighpenetration),directionof
change(e.g.,positiveornegative).
3.3.1. Selected Periods
Theselected3hourperiodsareidentifiedonthethree2010loaddurationcurvesshownin
Figure35.Theselectedperiodsaredistributedacrosstheloaddeciles.Thus,highload,light
load,andintermediateload(oftenperiodswithlargechanges)areallrepresented.Thethree3
hourperiodsselectedforthequasisteadystate(QSS)analysisarehighlighted,anddiscussedin
detailinSection5.1.1.
10000
15000
20000
25000
30000
35000
40000
45000
50000
55000
0 2630 5260 7890 10520 13150 15780 18410 21040 23670 26300
Hours
M
W
CAISO load, actual, MW
2010X Total Actual L-W-S
2010T Total Actual L-W-S
A1
A2
A3
A1
A4
A5
A
A
A8
A1
A9
A11
A12
A1
3
A14
B1
B4
B2
B3
G4
B5
G3
B6
B7
H3
J 1
C1
G10
G11
C2
I2
C3
G8
I3
J 2
C4
D1
D2
F1
F2
G1
G2
G5
G
G7
G9
H
H2
I1
B#1 B#2 B#3 B#4 B#5 B#6 B#7 B#8 B#9 B#10
July 19, 2004
6-9AM QSS
June 24, 2004
4-7PM QSS
May 15. 2003
1-4 AM QSS
Figure 35. 2010 Load Duration Curves with 3-Hour Periods of Interest Identified.
Foreachselected3hourperiod,1minuteprofilesweredevelopedforeachwindandsolar
resourceand1minutedataforCaliforniaISOloadwasprovided.Forillustration,Figure36
showsaspecific3hourperiod,6amto9am,inJuly2003.Inthemorning,loadincreasesatarate
ofabout10,000MW/3hrs,whilethewinddecreasesatarateofabout600MW/3hrs.
59
25000
30000
35000
40000
45000
6:00 7:00 8:00 9:00
0
500
1000
1500
2000
Load
Load-Wind-Solar
Wind
Solar
Time
L
o
a
d

(
M
W
)
W
i
n
d

&

S
o
l
a
r

(
M
W
)

Figure 36. 1-Minute Profiles During 3-Hour Period of Example July 2003 Day.
3.3.2. Intra-Hour Variability Definitions
Afterthe1minutewindandsolarprofilesandloaddataweresynchronized,asubhourly
statisticalanalysiswasperformedontheloadalone,windalone,andnetload.Theanalysiswas
similartothatperformedonthe1hourdataanddescribedinprevioussections.
Figure37usesanexample1minuteprofile(redlinewithblackdots)todefinetheload
followingmetricusedinthisanalysis.Thefirststepindevelopingthismetricwastocalculate
the15minuterollingaverage(blueline)ofthe1minutedata.Then,a5minutedeltawas
calculatedbasedonthe15minuterollingaverage.Forexample,therollingaverageattime
equalto5minutesisabout42MW.Fiveminuteslater,therollingaverageisabout52MW.Thus
the5minutedelta,orloadfollowingmetric,isabout10MW.
Thesamefigurecanbeusedtodescribetheregulationmetricusedinthisstudy.Again,thefirst
stepindevelopingthismetricwastocalculatethe15minuterollingaverageofthe1minute
data.Then,the1minutedeltabetweenthedataandthe15minuterollingaveragewas
calculated.Forexample,therollingaverageattimeequalto14minutesisabout56MW.The
dataatthistimeisabout40MW.Thusthe1minutedelta,orregulationmetric,isabout16MW.
60
10
20
30
40
50
60
70
80
0 5 10 15
Time in Minutes
M
W

Figure 37. Example for Load-Following and Regulation Metric Definition.
Asanexample,Figure38showsthe5minutedeltasonthe15minuterollingaverageforthe
load,netload,wind,andsolardataonaJulymorning.Theleftyaxisscaleappliestotheload
andnetload,therightyaxisscaleappliestothewindandsolar.Theloadfollowing
requirement(i.e.,5minutedeltas)rangesfrom120MWto400MWper5minutes.Thewind
deltaisgenerallynegative,andthesolardeltaisgenerallypositive.
The5minutestatisticalresultsassociatedwiththisstudyperiodareshowninTable12.The
standarddeviationofnetload5minutedeltais2MWlessthantheloadalone5minutedelta
standarddeviationof54MW.
5-minute
delta
1-minute
delta
61
-100
0
100
200
300
400
500
6:00 7:00 8:00 9:00
-50
0
50
100
150
200
250
Load
Load-Wind-Solar
Wind
Solar
L
o
a
d

(
D
e
l
t
a

M
W

p
e
r

5

M
i
n
u
t
e
s
)
Time
W
i
n
d

&

S
o
l
a
r

(
D
e
l
t
a

M
W

p
e
r

5

M
i
n
u
t
e
s
)

Figure 38. LoadFollowing Requirement for July 2003 Example 3-Hour Period.
Table 12. Statistics on Load Following

Requirement for July 2003 Example 3-Hour Period.
Mean (MW) (MW)
Load 265 54
Wind -16 18
Solar 7 13
L-W-S 275 52
Forthesamestudyperiod,Figure39showsthe1minutedeltasfromthe15minuterolling
averagefortheload,netload,windandsolardata.Theleftyaxisscaleappliestotheloadand
netload,therightyaxisscaleappliestothewindandsolar.The1minutewinddeltaiswithin
20MW/perminute,andthe1minutesolardeltarangesbetween60MW/perminute.
The1minutestatisticalresultsassociatedwiththisstudyperiodareshowninTable13.The
standarddeviationofnetload1minutedeltaisabout5MWlargethanthe1minuteloadalone
deltastandarddeviationof56MW.
62
-600
-300
0
300
6:00 7:00 8:00 9:00
-100
0
100
200
Load
Load-Wind-Solar
Wind
Solar
L
o
a
d

(
D
e
l
t
a

M
W

p
e
r

1

M
i
n
u
t
e
)
Time
W
i
n
d

&

S
o
l
a
r

(
D
e
l
t
a

M
W

p
e
r

1

M
i
n
u
t
e
)

Figure 39. Regulation Requirement for July 2003 Example 3-Hour Period.
Table 13. Statistics on Regulation Requirement

for July 2003 Example 3-Hour Period.
Mean (MW) (MW)
Load 6.5 56
Wind -0.3 9
Solar 0.0 22
L-W-S 6.8 61
3.3.3. 2010X Variability at Light Load
Thesectionfocusesontheloadfollowingrequirements(i.e.,5minutedeltas)forthe2010X
scenariounderlightloadconditions.Figure40showsthe5minutedeltadurationcurvesfor
loadalone,netloadandwindunderthoseconditions.Theyaxisshowsthepercentoflightload
hourswitha5minutedeltalargerthantheMWlevelshownonthexaxis.Bothpositiveand
negative5minutedeltasareshownonthexaxis.Loaddecreasesaremoreimportantthanload
increasesatlightloadintheloadfollowing/economicdispatch(5minute)timeframe.Theload
alone(lightblueline)andnetload(pinkline)durationcurvesaresimilar.Thelargest5minute
loadaloneandnetloaddecreasesarebothabout600MW.
63
0
10
20
30
40
50
60
70
80
90
100
-600 -500 -400 -300 -200 -100 0 100 200 300 400
Delta (MW)
T
i
m
e
s

(
%
)
5min L-W-S delta (+)
5 min Load delta (+)
5min L-W-S delta (-)
5 min Load delta (-)
5min Wind delta (-)
5min Wind delta (+)
P
e
r
c
e
n
t

o
f

L
i
g
h
t

L
o
a
d

H
o
u
r
s
MW Change
Figure 40. 2010X 5-Minute Delta Duration Curves for Light Load (10th Decile).
Figure41showsthe1minutedeltadurationcurvesforloadalone,netloadandwindunder
2010Xlightloadconditions.Again,theyaxisshowsthepercentoflightloadhourswitha1
minutedeltalargerthantheMWlevelshownonthexaxis.Bothpositiveandnegative1minute
deltasareshownonthexaxis.Loaddecreasesremainmoreimportantthanloadincreasesin
theregulation(minutetominute)timeframe.Theloadalone(lightblueline)andnetload(pink
line)durationcurvesaresimilar.Thelargest1minutedecreasesinbothloadandnetloadare
about350MW.
64
0
10
20
30
40
50
60
70
80
90
100
-350 -300 -250 -200 -150 -100 -50 0 50 100 150 200
Delta (MW)
T
i
m
e
s

(
%
)
1min L-W-S delta (+)
1min Load delta (+)
1min L-W-S delta (-)
1min Load delta (-)
1min Wind delta (-)
1min Wind delta (+)
Figure 41. 2010X 1-Minute Delta Duration Curves for Light Load (10th Decile).
Figure42shows1minuteand5minutewinddeltadurationcurvesfor2010Xlightload
conditions.Again,theyaxisshowsthepercentoflightloadhourswithadeltalargerthanthe
MWlevelshownonthexaxis.Bothpositiveandnegative1minuteand5minutedeltasare
shownonthexaxis.Thelargest5minutewinddecreaseisabout200MW,approximately
fourtimesofthelargest1minutewinddecrease.
65
0
10
20
30
40
50
60
70
80
90
100
-200 -150 -100 -50 0 50 100
Delta (MW)
T
i
m
e
s

(
%
)
5min Wind delta (+)
1min Wind delta (+)
5min Wind delta (-)
1min Wind delta (-)
Figure 42. 2010X Sub-Hourly Delta Wind Duration Curves for Light Load (10th Decile).
3.4. Variability Summary
Asummaryofthehourlyandsubhourlyvariabilityforthe2006,2010Tand2010Xscenariosis
showninTable14.Theloadandnetloadvariabilityforeachscenario,aswellasthedifference
betweenthem,isdisplayedineachsetofthreerows.Thefirstthreecolumnsshowthestandard
deviation()ofthe1hour,5minuteand1minutedeltas.Thefinalthreecolumnsshowthe
maximumandminimumofthe1hour,5minuteand1minutedeltas.
Themaximumandminimum1hournetloaddeltasforthe2010Tscenarioare402MWand
96MWless,respectively,thanwithloadalone.Thus,the2010Twindandsolarscenarioreduces
theextreme1hourscheduleflexibilityrequirementsintheCaliforniaISOsystem.
66

Table 14. Summary of 2006 and 2010 FullYear Statistical Analysis.
Standard Deviation, Maximum, Minimum
1-Hour s
(MW)
5-Min s
(MW)
(1)

1-Min s
(MW)
(2)

1-Hour s
(MW)
5-Min s
(MW)
(1)

1-Min s
(MW)
(2)

2006 Load 1436 189.3 44.8 6123, -5122 526, -480 803, -305
2006 L-W-S 1451 189.9 44.9 6091, -5155 550, -481 803, -306
Change 15 0.3 0.1 -32, -33 24, -1 0, -1
2010 Load 1575 207.6 49.1 6714, -5617 577, -527 881, -334
2010T L-W-S 1623 214.5 50.7 6312, -5713 699, -522 887, -323
Change 48 6.9 1.6 -402, -96 122, 5 6, 11
2010 Load 1575 207.6 49.1 6714, -5617 577, -527 881, -334
2010X L-W-S 1704 221.8 52.4 7219, -5986 722, -530 884, -335
Change 129 14.2 3.3 505, -37 145, -3 3, -1
Notes: (1)5minutechangeinMWon15minuterollingaverage
(2)1minutedifferenceinactualMWfrom15minuterollingaverage
Asummaryofthelightloadhourlyandsubhourlyvariabilityforthe2006,2010Tand2010X
scenariosisshowninTable15.Theloadandnetloadvariabilityforeachscenario,aswellasthe
differencebetweenthem,isdisplayedineachsetofthreerows.Thefirstthreecolumnsshow
thestandarddeviation()ofthe1hour,5minuteand1minutedeltas.Thefinalthreecolumns
showthemaximumandminimumofthe1hour,5minuteand1minutedeltas.
Table 15. Summary of 2006 and 2010 Light Load (10
th
Decile) Statistical Analysis.
Standard Deviation, Maximum, Minimum
1-Hour s
(MW)
5-Min s
(MW)
(1)

1-Min s
(MW)
(2)

1-Hour s
(MW)
5-Min s
(MW)
(1)

1-Min s
(MW)
(2)

2006 Load 669 86.5 40.8 1707, -2567 154, -257 200, -194
2006 L-W-S 699 89.2 40.9 2448, -2613 174, -257 198, -193
Change 30 2.7 0.1 741, -46 20, 0 -2, 1
2010 Load 734 94.9 44.8 1871, -2815 169, -282 219, -213
2010T L-W-S 933 109.1 45.9 2939, -3427 231, -259 213, -228
Change 199 14.2 1.1 1068, -612 62, 23 -6, -15
2010 Load 734 94.9 44.8 1871, -2815 169, -282 219, -213
2010X L-W-S 1081 114.7 47.7 3250, -4141 254, -250 203, -224
Change 347 19.8 2.9 1379, -1326 85, 32 -16, -11
Asummaryofthehourlyvariabilityforthe2010T,2010Xand2020scenariosisshowninTable
16.Forthe2010Xscenario,the1hournetloaddeltastandarddeviationis129MWlargerthan
the20101hourloadalonestandarddeviation.Thisisasignificantlylargerdifferencethanis
observedforthe2020scenario.Underthoseconditions,thestandarddeviationincreasesby
42MWbetweenloadaloneandnetload.Therefore,therelativeimpactofintermittent
renewablegenerationislessinthe2020scenariothaninthe2010Xscenario.
67
Table 16. Summary of 2010 and 2020 Hourly Statistical Analysis.

Standard Deviation,
1-Hour s
(MW)
Maximum, Minimum
1-Hour s
(MW)
2010 Load 1575 6714, -5617
2010T L-W-S 1623 6312, -5713
Change 48 -402, -96
2010 Load 1575 6714, -5617
2010X L-W-S 1704 7219, -5986
Change 129 505, -37
2020 Load 1977 8427,-7049
2020 L-W-S 2019 8747,-7351
Change 42 321,-302
Thereisasubstantialincreaseinthehourlyscheduleflexibilityrequirementsbetween2010and
2020.TheoperationalimplicationsofthestatisticspresentedinTable14throughTable16will
beexaminedinSection6.2.
3.5. Hourly Forecast Error
Adistinctionofsomeimportanceisthatbetweenvariabilityanduncertainty.Thephysical
natureofbothsystemloadsandintermittentrenewablesissuchthattheyvary.Muchattention
hasbeengiveninthisreporttoquantifyintherelativeamountofchangeorvariationinload
andrenewablepowerthatcanoccurindifferenttimeframes.Theabilityofthepowersystem
torespondtochangesisafunctionofthecapability(physicalandotherwise)ofavailable
resourcestochangeoutput.Oneimportantaspectofthisabilityisknowledge,apriori,ofthe
outputrequirement.Variabilityalonedoesnotnecessarilyimplyunpredictability.Forexample,
tidalpower,whichtypicallyvariesfromzerotomaximumpowertwiceadayishighlyvariable,
butnearlyperfectlypredictable.Theintermittentrenewablesunderconsiderationinthisstudy,
solarandwind,areratherlessperfectlypredictable.Ingeneral,themoreunpredictablethe
variation,themoreagilethegeneratingresourcesmustbe.
Thevariabilityofintermittentresourcesisreflectedindeltastatistics.Theuncertaintyof
intermittentresourcesisreflectedinforecasterrorstatistics,whichareexaminedinthissection.
3.5.1. Day-Ahead Forecast Error
Figure43showsdayaheadhourlyactualandforecastdataforanexample2010summerweek.
Theactualload(darkblueline)andforecastload(lightblueline)curvesarebasedonhistorical
datascaledtothe2010loadlevel.Theactualwind(redline)andforecastwind(orangeline)
weredevelopedasdescribedintheAWSreport[7].Theactualsolar(darkgreenline)was
developedasdescribedinSection2.3.Theforecastsolar(lightgreenline)wasdefinedasthe
monthlyaverageoftheactualsolar.Theleftyaxisscaleappliestotheloadcurves,andtheright
yaxisscaleappliestothewindandsolarcurves.
68
Theloadactuallargelyfollowstheforecast,withsomesignificantdifferencesobservedduring
thedailypeakloadperiod.Bothunderforecasts(forecastlessthanactual)andoverforecasts
(forecastmorethanactual)areobserved.Thesolaractualfollowstheforecastasexpected,given
theforecastwasderivedfromtheactual.Thewindactualhasthesameoverallshapeasthe
forecast,buttheforecastisconsistentlylessthantheactualoutput.
0
10000
20000
30000
40000
50000
1 25 49 73 97 121 145
-5000
0
5000
10000
15000
20000
Actual Load Forecast Load Actual Wind Forecast Wind Actual Solar Forecast Solar
W
i
n
d

&

S
o
l
a
r

(
M
W
)
L
o
a
d

(
M
W
)
Hour of Week
Figure 43. 2010 Load, Wind, and Solar Forecasts and Actuals During an Example July Week.
Figure44showsthedayaheadhourlyforecasterrorcurvesforthesameexampleweek.Load
forecasterror(darkblueline)wascalculatedasforecastminusactual.Thewind(orangeline)
andsolar(greenline)forecasterrorswerecalculatedasactualminusforecast.Thenetload
forecasterrorisloadforecasterrorminuswindforecasterrorminussolarforecasterror.The
loadforecasterrorrangesbetweenabout+/3,000MW.Thewindforecasterrorrangesfrom
about0MWto3000MW.Thisparticularforecastincludesabiastendstowards
underproduction.Inlessflexiblesystems,suchabiascanbeameansofhedging
underproductionrisk.Aswillbeshownlater,underforecastinghaseconomicpenalties,so
subsequentresultspresentedherearefornominallyunbiasedwindforecasts.Thesolarforecast
erroriswithinabout+/200MW.
69
Hour of Week
-5000
-4000
-3000
-2000
-1000
0
1000
2000
3000
4000
5000
1 25 49 73 97 121 145
Load Forecast Error (F-A)
Wind Forecast Error (A-F)
Solar Forecast Error (A-F)
L-W-S Forecast Error
Load Forecast > Actual Load
Actual Load > Load Forecast
F
o
r
e
c
a
s
t

E
r
r
o
r

(
M
W
)
Figure 44. 2010 Load, Wind, Solar, and Net Load Forecast Errors During Example July Week.
Justasvariabilityofintermittentresourcesismeaningfulonlyinthecontextofexistingload
variability,sotoo,theuncertaintyofintermittentresourcesisonlymeaningfulinthecontextof
existingloadforecasterror.
Figure45showsdayaheadforecasterrordurationcurvesforthe2010Xscenario.Thecurves
includeloadalone(darkblueline),netload(orangeline),wind(greenline)andsolar(lightblue
line)dayaheadforecasterrors.Theuncertaintyofwindforecastworsensthenetloadforecast.
70
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
0 2630 5260 7890 10520 13150 15780 18410 21040 23670 26300
Hours
M
W
Load Forecast-Actual
L-W-S Forecast-Actual
Wind Actual-Forecast
Solar Actual-Forecast
Act ual Load > Load Forecast
Figure 45. 2010X Load, Wind, Solar, and Net Load DayAhead Forecast Error Duration Curves.
Table17showsthe2006hourlyloadforecaststatisticalresults,groupedintotendecilesonthe
basisofloadalone.The10percentofpeakloadhoursareincludedindecile1,andthe10
percentoflightloadhoursareincludedindecile10.Thetableshowstheaverage,standard
deviation,maximum,andminimumofthedayaheadhourlyloadforecasterrors.Table18
showsthe2006hourlynetload,windandsolarforecaststatisticalresults,groupedintoten
decilesonthebasisofLWS.Thefirstrowsinthetableshowtheaverage,standarddeviation,
maximum,andminimumofthedayaheadhourlynetload(LWS)forecasterrors.Thenextset
ofrowsshowstheaverage,standarddeviation,maximum,andminimumofthedayahead
hourlywindforecasterrors.Thefinalsetofrowsshowstheaverage,standarddeviation,
maximum,andminimumofthedayaheadhourlysolarforecasterrors.
Table 17. 2006 Hourly Load Forecast Statistics (MW).
Load Decile
Load Trait 1 2 3 4 5 6 7 8 9 10
All
Years
Average Error 157 313 141 178 147 93 36 -2 -24 59 110
Error 1318 1023 746 704 640 672 623 558 589 520 781
Maximum Error 5825 6533 4761 4813 3791 3585 4173 2531 3841 2208 6533
Minimum Error -6281 -3896 -3063 -3400 -3423 -3589 -2759 -2778 -1940 -1675 -6281
71
Table 18. 2006 Hourly Net Load, Wind, and Solar Forecast Statistics (MW).
Net Load Decile
Net Load Trait 1 2 3 4 5 6 7 8 9 10
All
Year
Average Error 231 448 247 301 276 261 202 185 142 269 256
Error 1339 1060 803 749 714 739 684 661 636 571 829
Maximum Error 6402 6555 4463 4519 3959 3397 4687 4974 3979 2321 6555
Minimum Error -6443 -3954 -3036 -3205 -3343 -3556 -2668 -1773 -1743 -1701 -6443
Average Wind Error -79 -128 -111 -126 -134 -170 -159 -177 -171 -203 -146
Wind Error 196 227 230 240 246 254 260 261 273 285 251
Maximum Wind Error 537 630 646 707 618 698 720 701 612 713 720
Minimum Wind Error -1104 -1134 -992 -1083 -1059 -1079 -1073 -1079 -1136 -1137 -1137
Average Solar Error -14 -4 -1 3 5 5 4 1 0 -3 -0.4
Solar Error 60 60 53 55 54 51 49 39 25 37 50
Maximum Solar Error 306 286 205 237 278 225 210 223 175 131 306
Minimum Solar Error -175 -246 -241 -659 -717 -724 -656 -671 -343 -548 -724
Table19showsthe2010hourlyloadforecaststatisticalresults,groupedintotendecilesonthe
basisofloadalone.Thetableshowstheaverage,standarddeviation,maximum,andminimum
ofthedayaheadhourlyloadforecasterrors.Table20showsthe2010Thourlynetload,wind
andsolarforecaststatisticalresults,groupedintotendecilesonthebasisofnetload(LWS).
Thefirstrowsinthetableshowtheaverage,standarddeviation,maximum,andminimumof
thedayaheadhourlynetloadforecasterrors.Thenextsetofrowsshowstheaverage,standard
deviation,maximum,andminimumofthedayaheadhourlywindforecasterrors.Thefinalset
ofrowsshowstheaverage,standarddeviation,maximum,andminimumofthedayahead
hourlysolarforecasterrors.
Table21showsthe2010Xhourlynetload,windandsolarforecaststatisticalresults,grouped
intotendecilesonthebasisofnetload(LWS).TheformatisthesameasthatusedinTable20.
Table 19. 2010 Hourly Load Forecast Statistics (MW).
Load Decile
Load Trait 1 2 3 4 5 6 7 8 9 10
All
Years
Average Error 172 344 154 196 162 102 39 -2 -26 65 121
Error 1445 1122 818 772 701 737 683 612 646 570 857
Maximum Error 6387 7164 5221 5278 4157 3932 4576 2776 4212 2422 7164
Minimum Error -6888 -4272 -3359 -3729 -3753 -3935 -3026 -3046 -2128 -1836 -6888
72
Table 20. 2010T Hourly Net Load, Wind, and Solar Forecast Statistics (MW).
Net Load Decile
Net Load Trait 1 2 3 4 5 6 7 8 9 10
All
Year
Average Error 246 611 544 585 758 846 721 639 799 1327 708
Error 1535 1356 1280 1246 1276 1366 1303 1249 1251 1174 1333
Maximum Error 7127 6978 7003 7059 6260 6996 7209 6954 7596 5106 7596
Minimum Error -7672 -3690 -3513 -4144 -4528 -3921 -4028 -3741 -3688 -4065 -7672
Average Wind Error -150 -281 -334 -420 -619 -758 -662 -614 -767 -1265 -587
Wind Error 646 765 806 895 993 1049 1031 1003 1035 1081 987
Average Solar Error -7 -10 -4 -7 18 12 2 1 1 -5 0
Solar Error 170 148 176 173 186 172 153 110 91 62 150
Table 21. 2010X Hourly Net Load, Wind, and Solar Forecast Statistics (MW).
Net Load Decile
Net Load Trait 1 2 3 4 5 6 7 8 9 10
All
Year
Average Error -383.9 14.7 5.9 103.3 227.1 297.8 278.8 213.2 419.5 1221 240
Error 1676 1450 1453 1500 1566 1681 1610 1635 1592 1542 1620
Maximum Error 6521 6311 5625 6141 7854 7148 7317 7081 6545 6920 7854
Minimum Error -9455 -6160 -5706 -6618 -5447 -6695 -6339 -5021 -6055 -5653 -9455
Average Wind Error 489 283 193 83 -112 -219 -211 -149 -384 -1166 -119
Wind Error 945 1033 1080 1197 1295 1414 1393 1427 1411 1469 1350
Average Solar Error -26 -15 -3 -11 30 19 5 1 8 -8 0
Solar Error 268 248 257 255 266 246 208 173 153 127 226
Table22summarizesthestandarddeviationofthedayaheadload,netloadandwindplus
solarforecasterrorsunderthe2010Xscenario.Thestandarddeviationofnetloadforecasterror
increasesby763MW,or90%,fromtheloadaloneforecasterrorstandarddeviationof857MW.
Table23showsthepositiveandnegativeenergyassociatedwiththedayaheadload,netload
andwindplussolarforecasterrorsunderthe2010Xscenario.Theloadalonepositiveforecast
errorenergyisabout9,600GWhr.Theadditionofwindandsolarforecasterrorsresultsina
73
netloadpositiveforecasterrorenergyofabout19,500GWhr.Thenegativenetloadforecast
errorenergyisalsoabouttwicethenegativeloadaloneforecasterrorenergy.
Table 22. 2010X Day-Ahead Forecast Error Standard Deviation.
Standard Deviation (MW)
Load (Forecast-Actual) 857
Wind+Solar (Actual-Forecast) 1566
L-W-S (Forecast-Actual) 1620
Table 23. 2010X Day-Ahead Forecast Error Energy.

Positive Energy (GW-hr) Negative Energy (GW-hr)
Load (Forecast-Actual) 9612 -6442
Wind (Actual-Forecast) 14723 -11583
Solar (Actual-Forecast) 1814 -1814
L-W-S (Forecast-Actual) 19453 -13142
Figure46showsanotheropenhighlowclosestockchartwith2006loadandnetloadday
aheadforecasterrorscenario.Theredbarsrepresenttheloadscenarioandtheyellowbars
representthenetload(LWS)scenario.Thetopofabarrepresentstheaveragevalueofhourly
forecasterrorplusthestandarddeviation.Thebottomofabarrepresentstheaveragevalueof
hourlyforecasterrorminusthestandarddeviation.Thetopofaverticallinerepresentsthe
maximumhourlyforecasterror,andthebottomofaverticallinerepresentstheminimum
hourlyforecasterror.
Thechartshowsthatwindandsolarhavesomeimpactonforecasterrors,includingextremes,
underallconditions.Thestandarddeviationoftheloadalonevariabilityforthe10thdecileis
520MW.Thisincreasesby50MWto570MWwithwindandsolar.Hence,underlightload
conditions,thereisa99.7%expectation(3)thathourtohourchangeswillbelessthan
1560MWwithloadalone,and1710MWwiththeadditionalwindandsolar.
74
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10
F
o
r
e
c
a
s
t

E
r
r
o
r

(
M
W
)
Load
L-W-S
Load
L-W-S
Decile
Figure 46. 2006 Load and Net Load Day-Ahead Forecast Error Stock Chart by Decile.
Figure47showsasimilaropenhighlowclosestockchartforthe2010Xloadandnetload
scenarios.
Thechartshowsthatwindandsolarhaveasignificantimpactonforecasterrors,theincluding
extremes.Theoveralluncertaintystandarddeviationrangesfrom857MW(2010loadalone)to
1,333MW(2010T)to1620MW(2010X).Thelatterisapproximatelydoublethestandard
deviationforloadalone.Theenergyerror,asreportedinTable23,confirmsthisapproximately
twotooneratio.Butatlighterloadlevels,andnotjustthelightest,theuncertaintyis
significantlyhigher.Atthelightestload,10
th
decile,theerrorincreasesfrom570MW(2010load
alone)toaboutdoublefor2010T(1174MW)andtoalmosttriplefor2010X(1542MW).Thelatter
isanincreaseof970MWovertheloadalonestandarddeviation.Hence,thereisa99.7%
expectation(3)thathourtohourchangeswillbelessthan1710MWforloadalone,
increasingto3,522MWfortheexpected2010Tscenario,andto4,620MWfortheextreme
scenario.Underlightloadconditions,thelargestpositiveloadalonedayaheadhourlyforecast
errorisabout2,420MW.Thelargestpositivenetloaddayaheadhourlyforecasterrorisabout
6,920MW.Thelargestnegativeloadalonedayaheadhourlyforecasterrorisabout1840MW.
Thelargestnegativenetloaddayaheadhourlyforecasterrorisabout5,650MW.The
maximumandminimumlightloadforecasterrors,forboththeloadaloneandnetload
scenarios,slightlyexceedthe3level.Theworstloadaloneoutliersareabout+/7,000MW,and
theworstnetloadoutliersareabout+/7,700MW.Underlighterloadconditions(lessthan
75
median)thedifferenceissubstantiallylarger.Theloadaloneerrorroughlydoublesfromthe
+/2,0004,000MWrangetothe+/5,0007,000MWrange.Overall,thesystemmustbeprepared
fordayaheaduncertaintyofabout+/5,000MW.
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10
F
o
r
e
c
a
s
t

E
r
r
o
r

(
M
W
)
Decile
Load
L-W-S
Load
L-W-S
Figure 47. 2010X Load and Net Load Day-Ahead Forecast Error Stock Chart by Decile.
3.5.2. Hour-Ahead Forecast Error
Figure48showshouraheadforecasterrordurationcurvesforthesame2010Xscenario.The
curvesincludetheloadalone(darkblueline),netload(orangeline),wind(greenline)andsolar
(lightblueline)houraheadforecasterrors.Asexpected,theuncertaintyofthehourahead
forecastislessthandayaheadforecast.However,theuncertaintyofwindforecaststillworsens
thenetloadforecast.
76
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
0 2630 5260 7890 10520 13150 15780 18410 21040 23670 26300
Hours
M
W
Load Forecast-Actual
L-W-S Forecast-Actual
Wind Actual-Forecast
Solar Actual-Forecast
Act ual Load > Load Forecast
Figure 48. 2010X Load, Wind, Solar, and Net Load Hour-Ahead Forecast Error Duration Curves.
Table24summarizesthestandarddeviationofhouraheadloadalone,netloadandwindplus
solarforecasterrors.Thestandarddeviationofthenetloadforecasterroris156MW,or26%,
largerthantheloadaloneforecasterrorstandarddeviationof606MW.
Table25showsthepositiveandnegativeenergyassociatedwiththehouraheadforecasterrors.
Thepositiveenergyofthenetloadhouraheadforecasterroris2240GWhr,or34%,largerthan
thepositiveloadaloneforecasterrorenergyofabout6500GWhr.Similarly,thenegative
energyofthenetloadhouraheadforecasterroris1700GWhr,or36%,morethanthenegative
loadaloneforecasterrorenergyof4800GWhr.Thustheimpactoftheintermittentrenewables
ismuchlessinthehouraheadforecastthaninthedayaheadforecast.
Table 24. 2010X Hour-Ahead Forecast Error Standard Deviation.
Standard Deviation (MW)
Load (Forecast-Actual) 606
Wind+Solar (Actual-Forecast) 706
L-W-S (Forecast-Actual) 762
Table 25. 2010X Hour-Ahead Forecast Error Energy.

Positive Energy (GW-hr) Negative Energy (GW-hr)
Load (Forecast-Actual) 6508 -4810
Wind (Actual-Forecast) 4797 -4796
Solar (Actual-Forecast) 1703 -1703
L-W-S (Forecast-Actual) 8252 -6553
Figure49showsanotheropenhighlowclosestockchartoftheloadandnetloadhourahead
forecasterrorsunderthe2010Xscenario.Thechartshowsthatwindandsolarhavesome
impactonforecasterrors,includingextremes,underallconditions.Underlightloadconditions,
77
thestandarddeviationoftheloadalonehouraheadforecasterroris460MW.Thisincreasesby
140MWto600MWwiththeadditionalwindandsolar.Hence,underlightloadconditions,
thereisa99.7%expectation(3)thathourtohourchangeswillbelessthan1,380MWforload
alone,and1,800MWfornetload.Stillunderlightloadconditions,thelargestpositivehour
aheadloaderroris3,160MW.Thecorrespondingnetloaderrorisabout2,200MW.Thelargest
negativehouraheadloaderrorunderlightloadconditionsis1,600MW.Thecorresponding
netloaderrorisabout2400MW.Boththeloadandnetloadhouraheadforecasterrorextremes
aresomewhatgreaterthan3.
Overall,thecontributionofwindtohouraheaduncertaintyisrelativelymuchlessthanday
ahead.Thehouraheaderrorwithwindisonlyslightlyworseacrosstheboard,consistentwith
theincreaseinsigmaofabout20%(fromabout600toabout720MW).Overall,thesystems
shouldplanfor2,000MWhouraheaderror.
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
1 2 3 4 5 6 7 8 9 10

Figure 49. 2010X Load and Net Load Hour-Ahead Forecast Error Stock Chart by Decile.
3.6. Summary
Thestatisticalresultsfortheexpectedgrowthofrenewables(2010T,2020)aresummarizedas
follows:
Intermittentrenewablegenerationchangesdailyandseasonally.Totalstatewide
productionofwindtendstobeanticoincidentwithload.Solartendstoberelatively
coincidentwithload.
Averagewindpenetrationwillrangefromabout4%atpeakloaduptoabout20%at
lightload.Averagesolarpenetrationatpeakloadwillgrowtoabout5%.
78
Spatialvariationinproductionbetweenwindplantsissubstantial,andevenwind
plantsingeneralgeographicproximity,i.e.Tehachapiarea,demonstratesignificant
diversity.
Dailyperiodsofrapidloadriseandloaddeclinewillexperienceincreasedratesof
change.Temporallycoincidentextremechangesinloadandrenewablegeneration
arenotexpected.Themostextremesustainednetchanges,bothupanddown,are
expectedtobeabout1,000MWgreaterover3hoursthanthatduetoloadalone.
Onaverage,variabilityduetointermittentrenewablesisabout3%to7%largerthan
thatduetoloadalone.Therelativeimpactonhourlyscheduleflexibilityandload
followingatlightloadwillbegreater.Incrementalregulationrequirementsare
relativelyunaffectedbyloadlevel.
Thedayaheadforecasterrorisabout5,000MW,abouthalfofwhichisdueto
intermittentrenewables.Thehouraheaderrordropstoabout2,000MW,towhich
intermittentrenewablescontributeabout20%.
TheoperationalimplicationsofthestatisticspresentedherewillbeexaminedinSection6.2.
79
4.0 Production Simulation Analysis
AneconomicsimulationoftheCaliforniasystemwasperformedtodeterminetheoperational
impactofintermittentgeneration.BecauseCaliforniaissuchasignificantportionoftheWECC
systemitwasnecessarytomodeltheinterconnectedoperationoftheentireWECCgrid.The
WECCwasmodeledbroadlyassixregions:Arizona,California,Canada,Northwest,Rocky
MountainandMexico.Theseregionswerefurtherdividedinto73separateloadareas,each
withtheirownchronologicalloadshapefortheyear.Loadswithinanareawerethenassigned
toindividualbussesintheloadflow.Over2600generatingunitsweremodeledwithinthe
WECCsystem.Thermalunitsweremodeledwithminimumoperatingpointsandmultiple
incrementalcostsegmentsaswellasrampingandstartupconsiderationsthatmirrortheir
physicalcapabilities.Theunitcommitmentdecisionwasbasedoninformationtypically
availableinthedayaheadmarket,i.e.forecastedloadsandforecastedwindandsolar
generation.Thoseunitsthatwereturnedonwerethendispatchedinaleastcostmanner
similartothehouraheadmarket.Fromthisdispatchthehourlymarginalprices,orspotprices
astheyareoftenreferredto,weredeterminedforeachbuswithaveragescalculatedonanarea,
regionalandsystembasis.Boththecommitmentanddispatchdecisionsrecognizedthe
individualbuslocationsofthegeneratorsandloadandthetransmissionconstraintspresentin
thenetwork.AlltransmissionconstraintsintheWECCPathRatingCataloguewereconsidered.
Conventionalhydrogenerationwasrepresentedwithamonthlyminimumandmaximum
operatinglimitaswellasamonthlyenergyavailability.Withintheseconstraintsthehydro
generationinCaliforniawasscheduledonalocalareabasisatthetimesofhighestload.Hydro
intherestofWECCwasscheduledprimarilyonaregionalbasiswithafewofthelargestsites
adjustingtheirschedulesbasedonloadsthroughoutthesystem.Varioussensitivitieswere
examinedconcerningtheflexibilityofthehydrogeneration.PumpedStorageHydro,PSH,was
alsoconsidered.Themodelrecognizesthecostofenergyavailableforpumping,thevalueof
theenergywhenitisusedforgeneration,thelimitsimposedbythestoragereservoirandthe
overallcycleefficiencytodeterminetheeconomicoperationofthePSHunits.
Thesystemwasexaminedforincreasinglevelsofrenewablegeneration,bothintermittent(solar
andwind)andconstant(biomassandgeothermal).Theconstantrenewablesweredispatchable
butwereassumedtobidinatarelativelylowcostsothattheiroutputdidnotvary.The
intermittentrenewablegenerationwasassumedtobeapricetakerandsowasbidinatzero
cost.Themodelwascapableofsheddingthisenergyiftransmissionconstraintsarose.Thecost
savingsdeterminedwiththeadditionoftherenewablesisthenagrossvalueoftheenergyand
needstobeoffsetbytheenergypaymentsfortherenewableenergyaswellasthecapitalcosts
ofanysystemimprovementsrequiredinordertodeterminetheoverallsystemeconomics.
TheprimarypurposeofthesimulationanalysiswasNOTtodeterminetheeconomicvalueof
therenewablesbutrathertodeterminetheoperationalimpactonthebalanceofthesystem.
Doesloadsheddingoccur?Whattypeofgenerationisdisplaced,coalorgas?Whathappensto
transmissionloading?Whatistheimpactonemissions?Whatisthespotpriceimpactof
introducingalargeamountofpricetakerstothesystem?Whatlevelofmaneuverabilityis
80
desiredfromthehydrogeneration?Whatistherampingrateandrangeofthebalanceofthe
system?Thesearethetypesofresultsaddressedintheremainderofthissection.
TheoperationalanalysiswasperformedusingtheGEMultiAreaProductionSimulation
program(MAPS).ThisprogramhasbeenusedforyearsthroughoutNorthAmericatoassist
planners,developersandregulatorsintheanalysisofelectricpowersystems.MAPSisahighly
detailedmodelthatcalculateshourbyhourproductioncostswhilerecognizingtheconstraints
onthedispatchofgenerationimposedbythetransmissionsystem.Whentheprogramwas
initiallydevelopedoverthirtyyearsago,itsprimaryusewasasagenerationandtransmission
planningtooltoevaluatetheimpactsoftransmissionsystemconstraintsonthesystem
productioncost.Inthecurrentderegulatedutilityenvironment,themodelhasbeenusefulin
studyingissuessuchasmarketpowerandthevaluationofgeneratingassetsoperatingina
competitiveenvironment.TheuniquemodelingcapabilitiesofMAPSincludeadetailed
electricalmodeloftheentiretransmissionnetwork,alongwithgenerationshiftfactors
determinedfromasolvedacloadflow,tocalculatetherealpowerflowsforeachgeneration
dispatch.Thisenablestheusertocapturetheeconomicpenaltiesofredispatchingthe
generationtosatisfytransmissionlineflowlimitsandsecurityconstraints.Thechronological
natureofthehourlyloadsismodeledforallhoursintheyear.Intheelectricalrepresentation,
theloadsaremodeledbyindividualbus.Inadditiontothetraditionalproductioncosting
results,MAPScanprovideinformationonthehourlyspotpricesatindividualbusesandflows
onselectedtransmissionlinesforallhoursintheyear.
4.1.1. General Database Creation
TheMAPSmasterinputfile(MIF)wasconstructedspecificallyfortheIntermittencyAnalysis
Project(IAP).TheMIFconsistedofGEsWesternElectricityCoordinationCouncil(WECC)base
inputs,Californiaspecificdataandrenewablegenerationinformation.ThebaseMIFinputs
coveredtheWesternInterconnection,whichisprimarilycomprisedofthestateswestofthe
RockyMountainsandportionsofCanadaandMexico.DataforareasoutsideofCalifornia
remainedasoriginallyconceivedfortheWECCfortheIAPMAPSruns.Thestepstakento
createthefinalinputsare:
IncorporatingCaliforniadatainputsdevelopedbyRumla,
UpdatingCaliforniagenerationunitsandfuelprices,and
Integratingrenewablegenerationdevelopmentscenarios.
CaliforniaSpecificData
MAPSinputsspecifictoCaliforniawereextractedfromadatabasedevelopedovertimeby
Rumla.Thedataincorporated20032005CECandCaliforniaISOgenerationinformation.
Updatedelementsincludedinstallationandretirementdates,CaliforniaISOreliabilitymust
run(RMR)designations,andgeothermalunitratings.Forthermalunits,fueltypeassignments
andmonthlypriceswereobtainedfromtheCECandimplemented.AtestMAPSrunverified
thattheintegrationproducedreasonableresults.
81
IntermittencyAnalysisProject(IAP)
UpdatesperformedforIAPunitsweremadeconsistentwiththeMAPSCaliforniadatabase
reflecting20052006information.AsintheinitialCaliforniaupdates,changesinCaliforniaISO
RMRdesignations,andCEConlineunitstatusandfuelpriceswereimplemented.
Hydroelectricgenerationlevelsforeachunitwereadjustedtomirrorhistoricalmonthly
CaliforniahydroelectricgenerationfromutilityownedunitspublishedbytheCEConits
website.
RenewableGeneration
Renewablegenerationadditionsconsistedofbiomass,geothermal,windandsolarunits.IAP
unitswereappointedidentificationlabelsandassignedtobusnodes.Totheextentpossible,
IAPidentificationlabelswerematchedtoexistingMAPSunitnames.Anyremaining
unmatchedIAPrenewablesweretheninsertedintothedatabase.Windandsolargeneration
weremodeledashourlyresources.Forthesetwocategoriesofrenewables,existingCalifornia
windandsolarunitswereremovedfromthedatabaseandtheIAPdatainserted.Thetotal
levelsofrenewablegenerationvariedbyscenarioasdescribedpreviously.
FuelPriceAssumptions
Generationwasassumedtobecommittedanddispatchedonaminimumcostbasis.Historical
heatratedatawasusedandthefollowingfuelpriceswereassumed:
NaturalGas~$5.70/MBTU
DistillateOil~$6.50/MBTU
Coal~$1.5/MBTU
4.1.2. Scenario Description
Theoverallanalysisexaminedthesystemoperationforthe2006,2010and2020timeframes.
Becauseitcanbedifficulttocompareresultswithamultiplicityofdifferencesintheinput,
studycaseswerecreatedwithminimalchangesbetweenthem.Thevariousscenariosare
describedinTable26.Allscenariosusethesamefuelprices.Theyalsoholdthegenerationmix,
loadsandtransmissionconstantoutsideofCalifornia.Thefirstthreescenarios(basedon2010T)
holdthetransmissionconstantwithinCaliforniaandonlyvarythelevelofrenewable
generation.The2010Xscenariofurtherexpandsthelevelofrenewablesandalsoadds
additionaltransmissionwithinCaliforniatofacilitatetheadditionalgeneration.Thefinal
scenario,2020,usedthesamegenerationandtransmissionmixasthe2010Xcase,butexpanded
theloadstoreachthe2020projectedlevels.
82
Table 26. Production Simulation Scenario Description.

Scenario Title Scenario Description
2010T system No new renewables Geothermal, Biomass, Wind and Solar generation held at
2006 levels
2010T system No new intermittents Geothermal and Biomass increased to 2010 projections but
Wind and Solar held to 2006 levels
2010T system All renewables Geothermal, Biomass, Wind and Solar generation increased
to 2010T projections
2010X system Wind and Solar further expanded to reach 33% Renewables
2020 system Similar to the 2010X system with the loads increased to
2020 values

Theadvantageofthesescenariosisthattheyallowtheevaluationofdifferentlevelsandtypes
ofrenewablegenerationwhilethefuelcostandbalanceofthesystemremainconstant.Eachof
thesescenarioswascreatedusingeitherthe2002,2003or2004historicalhourlyloadandwind
profiles.Thesameistruefortheconcentratingsolarplants.The2004photovoltaic(PV)shapes
wereappliedforallcases.
4.2. Economics
ThecurveinFigure50showsadurationplotoftheaverageCaliforniahourlyspotpriceforthe
firstscenario(2010Tsystemnonewrenewables)basedonthe2002loadandwindshapes.
Thisplotwascreatedfromthechronologicalhourlyvaluesthatwerethensortedtomoreeasily
seetheimpactofchanges.Thisisapictureofthesystemcostsbeforetheadditionofany
renewablegenerationbeyondthe2006level.Thiswillserveasabasisforcomparisontothe
otherscenarios.Thecostswereconservativeinthatcostbasedbidswereassumedforall
generation.Higherbidswouldonlyincreasethevalueoftherenewablegeneration.
83
0.00
10.00
20.00
30.00
40.00
50.00
60.00
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
S
p
o
t

P
r
i
c
e

(
$
/
M
W
h
)
2010T No new Renewabl es
Figure 50. Spot Price Duration Curve for 2002 Shapes (#1).
InFigure51thenewgeothermalandbiomassunitsareintroducedintothesystem.These
generatorsaredispatchable,butareassumedtohavefirmcontractsandsoarebidintothe
commitmentanddispatchwithlowvalues.Thesecausethespotpricestodecrease,butonly
slightly.Theimpactonspotpricesisslightbecausethisrenewableenergyisintroducedintothe
systeminaconstant,predictablemanner,whichallowstheremainderofthesystemto
smoothlyreadjustinthecommitmentanddispatch.
0.00
10.00
20.00
30.00
40.00
50.00
60.00
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
S
p
o
t

P
r
i
c
e

(
$
/
M
W
h
)
2010T No new Intermi ttents
84
Figure52nowintroducessignificantamountsofnewintermittentrenewablegeneration.Two
separate,chronologicalshapeswereusedtodescribeeachwindplantoverthecourseofthe
year.Thefirstoneappliedistheactualwindgenerationbasedonthemeteorologicalconditions
thatarepresentduringthehour.Thisiswhatisinjectedintothesystemintheactualdispatch.
Thesecondshapeusedistheforecastofthishourlyprofilebasedontheexpected
meteorologicalconditionsfromroughlytwodaysaheadoftime.Boththeintermittencyandthe
errorinforecastcontributetothevariabilityintroducedwiththenewcurve.Whenwindis
overforecasted(i.e.forecastishigherthanwhatactuallyoccurs)thentheexistingcommitted
generationmustrunmorethanwasplannedandpossiblypeakinggenerationneedstobecalled
upontosupplementtheothergeneration.Bothofthesefactorswillincreasethespotprices.In
otherhoursthewindisunderforecastedresultinginexcessgenerationbeingonlinethatwill
causeadepressionofthespotprices.Lowpricescanalsoresultevenwhenthereisnoforecast
error.Increasesinwindgenerationduringlowloadnighttimehourscansignificantlylowerthe
spotpricesatatimewhenthebalanceofthesystemgenerationisunable(orunwilling)toback
down.Alloftheseimpactscanbeseeninthe2010TSystemcurvewhichincludesallofthe
projectedrenewablesforthebasecase.
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
S
p
o
t

P
r
i
c
e

(
$
/
M
W
h
)
2010T No new Intermi ttents
2010T System
Theseissuesareamplifiedwhenthesystemisexpandedto33%renewables.Figure53addsthe
resultsofthe2010Xscenario.Althoughtherearenoexpectationsthattherenewableswillreach
thislevelbythe2010timeframethisanalysisisintendedtoshowtheimpactonacomparable
scaleofahighpenetrationofintermittentrenewablegeneration.
85
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
S
p
o
t

P
r
i
c
e

(
$
/
M
W
h
)
2010T No new Renewables
2010T No new Intermittents
2010T System
2010X System
ThefinalcurveinFigure54showsthatastheloadgrowthoccursthespotpricesreturntoator
abovetheinitialvaluesandthedropoffatthetailofthecurveissignificantlyreduced.
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
80.00
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
S
p
o
t

P
r
i
c
e

(
$
/
M
W
h
)
2010T System
2010X System
2020 System
Havingdemonstratedthebuildofthevariousscenariosforthe2002shapes,Figure55and
Figure56showsimilarresultsforthe2003and2004profiles.Figure57expandsthecurvefor
the2004shapestohighlightthelowspotpriceimpacts.Thesharpdropinspotpricesindicate
periodswhenthesystemistryingtoshedallbutthemostcriticalresources.Itisnotnecessary
forthemodeltospillanyhydrooractuallydumpanyenergy,butclearlyitisstartingtoreach
86
deeperintothestackoflowcostgeneration.Theadditionofthegeothermalandbiomass
generationexpandsthenumberoflowpricedhoursfromroughly15hourstoabout75.The
20%renewablepenetrationinthe2010Tscenarioincreasesthistojustover300hours.The
2010Xcasewith33%penetrationisnowuptoabout750hoursthatthewindenergywillbeof
minimaleconomicvalue.The2020curveisbackdowntolessthan100hoursinthetail.
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
80.00
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
S
p
o
t

P
r
i
c
e

(
$
/
M
W
h
)
2010T System
2010X System
2020 System
Figure 55. Spot Price Duration Curve for 2003 Shapes.

0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
80.00
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
S
p
o
t

P
r
i
c
e

(
$
/
M
W
h
)
2010T System
2010X System
2020 System
Figure 56. Spot Price Duration Curve for 2004 Shapes.

87
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
45.00
50.00
7000 7500 8000 8500
Hours
S
p
o
t

P
r
i
c
e

(
$
/
M
W
h
)
2010T System
2010X System
2020 System
Figure 57. Spot Price Duration Curve for 2004 Shapes (zoom).
ThereductionsonoverallWECCsystemvariableoperatingcostsareshowninFigure58.Itis
importanttonotethatthisjustrepresentsthereductioninthecostofoperatingthebalanceof
thesystem.Thisdoesnotincludethecostthatneedstobepaidfortherenewableenergyorfor
thecapitalexpendituresrequiredfortherenewablegenerationandforthetransmission
enhancementsrequiredtosupportthem.
0
100
200
300
400
500
600
700
2002 2003 2004
W
E
C
C

O
p
e
r
a
t
i
n
g

C
o
s
t

I
m
p
a
c
t

(
$
M
)
Bi omass & Geothermal
Wi nd & Sol ar
Figure 58. WECC Operating Cost Reductions Due to Renewables ($M).

ThebarsinFigure59showthesameresultsexpressedin$/MWhofrenewableenergy.The
biomassandgeothermalgenerationdemonstrateabouta$10/MWhhighervaluethanthewind
88
andsolarenergy.ThisisNOTanintegrationcostbutsimplyarecognitionthatintermittent
generationisnotasvaluabletothesystemasenergyfromaconstantsource.
0
5
10
15
20
25
30
35
40
45
2002 2003 2004
W
E
C
C

O
p
e
r
a
t
i
n
g

C
o
s
t

I
m
p
a
c
t

p
e
r
M
W
H

o
f

R
e
n
e
w
a
b
l
e

(
$
/
M
W
h
)
Wi nd & Sol ar
Figure 59. WECC Operating Cost Reductions Due to Renewables ($/MWh).

Inamarginalcostmarkettheloadpaymentsareequaltothehourlyloadtimesthehourlyspot
pricefortheload.Thenexttwocharts,Figure60andFigure61,showtheimpactsontheload
paymentsforbothCaliforniaandWECCexpressedinbothmillionsofdollarsandin$/MWhof
renewableenergy.Aswithotherfactorsbeingexamined,theimpactofadditionalrenewable
generationinsideCaliforniaextendswellbeyondthestatesborder.
89
-
100
200
300
400
500
600
700
800
Cal i forni a
2002
Cal i forni a
2003
Cal i forni a
2004
WECC 2002 WECC 2003 WECC 2004
L
o
a
d

P
a
y
m
e
n
t

R
e
d
u
c
t
i
o
n
s

(
$
M
)
Wi nd & Sol ar
Figure 60. Load Payment Reductions Due to Renewables ($M).

0
5
10
15
20
25
30
35
40
Cal i forni a
2002
Cal i forni a
2003
Cal i forni a
2004
L
o
a
d

P
a
y
m
e
n
t

R
e
d
u
c
t
i
o
n

p
e
r

M
W
h

o
f

R
e
n
e
w
a
b
l
e

E
n
e
r
g
y

(
$
/
M
W
h
)
Wi nd & Sol ar
Figure 61. Load Payment Reductions Due to Renewables ($/MWh).

Figure62andFigure63showasimilarimpactonthegenerationrevenueforallofthenon
renewablegeneratorsinCaliforniaandWECC.Therevenuereductionhastwocomponents.
First,theenergyfromtheothergeneratorshasbeenreducedbecauseitwasdisplacedbythe
additionoftherenewablegeneration.Second,theeconomicvalueofthegenerationthatisstill
beingproducedhasbeenreducedduetotheintroductionoflargeamountsofpricetakers
intothegenerationbidstack.Thisadverseeffectonothergenerationcouldpossiblycausesome
existingmarginallyprofitablegenerationtooptforearlyretirementortocausesomeproposed
90
additionstobepostponed.Itisimportantforsufficientincentives,suchasthecapacitymarket,
tobeinplacetoensurethatreliabilitydoesnotsuffer.
0
200
400
600
800
1000
1200
1400
1600
Cal i forni a
2002
Cal i forni a
2003
Cal i forni a
2004
G
e
n
e
r
a
t
o
r

R
e
v
e
n
u
e

R
e
d
u
c
t
i
o
n
s

(
$
M
)
Wi nd & Sol ar
Figure 62. Non-Renewable Generator Revenue Reductions Due to Renewables ($M).

0
10
20
30
40
50
60
70
80
Cal i forni a
2002
Cal i forni a
2003
Cal i forni a
2004
G
e
n
e
r
a
t
o
r

R
e
v
e
n
u
e

R
e
d
u
c
t
i
o
n
s

p
e
r

M
W
h

o
f

R
e
n
e
w
a
b
l
e

G
e
n
e
r
a
t
i
o
n

(
$
/
M
W
h
)
Wi nd & Sol ar
Figure 63. Non-Renewable Generator Revenue Reductions Due to Renewables

($/MWh).
4.3. Intermittent Renewable Forecasting
Theimpactofforecastingforintermittentrenewablegenerationisofsufficientimportanceto
meritaseparatesection.Historically,forecastingofintermittentenergywasoftenignored.This
wasperfectlyadequateforsystemswithinstalledwindgenerationlessthan0.1%ofpeakload.
91
Thevariabilityofthewindgenerationwaslostinthevariabilityoftheloadandcouldeasilybe
absorbedinthedispatchwhenandifitshowedup.Astheenergycontributionfrom
intermittentunitsexpandsto5%,10%,20%andeveninexcessof30%thisisnolongerpossible.
Thebasicassumptionusedinthisanalysiswasthathydrooperationandthermalunit
commitmentwouldbeadjusted(totheextentpossible)basedontheprojectedwindandsolar
generation,whichwasdeterminedfromtheforecastedmeteorologicalconditionsfromroughly
twodaysinadvanceoftheactualoccurrence.Stateoftheartforecastingwasassumedand
hourlyprofilesweredevelopedasdescribedelsewhereinthisreport.Sensitivityanalyseswere
performedtoexaminetheimpactofeitherperfectforecasting,(forecastequalsactual
generationthatoccurs)orofignoringtheforecastandsimplyallowingtheintermittentenergy
tojustshowupinthehouraheadmarket.
Figure64showstheimpactoftheforecastontheCaliforniaaveragesystemspotpricesforthe
2010Tscenario.Thegreencurve,basedonestimatedforecasts,istheonethathasbeenshownin
Figure50.Theredcurveshowstheeffectofaperfectforecastfortheintermittentenergy.The
higherspotpricesarereducedbecausethesystemisnolongeroverforecastingandrequiring
peakingunitstomakeuptheshortages.Similarly,thelowerspotpricesareincreasedduetothe
eliminationofunderforecasting.Significantly,thenumberofhourswithasharpdropinspot
pricehavebeenreduced.Thebluecurve,whereintermittentsareignoredinthecommitment,
showstheotherextreme.Spotpricesareseverelydepressedthroughouttheyearduetothe
constantovercommitmentofresourcesandthenumberofhourswithasharpdropinprices
hasmorethandoubled.Whilesomemaythinkspotpricereductionsaregood,right?this
severedropcouldhaveadverseeffectsonthestabilityoftheoverallmarket.Combiningthe
hourlyspotpricewiththeloadwouldshowthattheloadpaymentsandthecorresponding
generatorrevenueswoulddecreasebybillionsofdollars.Thissignificantdropingenerator
revenuescouldcausegenerationtoleavethesystem,causingareductioninoverallsystem
reliability.
92
0
10
20
30
40
50
60
70
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
S
p
o
t

P
r
i
c
e

(
$
/
M
W
h
)
2010T Perfect forecast for commitment
2010T Estimated forecast for commitment
2010T No forecast for commitment
Figure 64. Impact of Intermittent Forecast on Spot Price (2010T).

ThecurvesinFigure65showthattheimpactisevenmorepronouncedinthe2010Xscenario
with33%renewablegeneration,aswouldbeexpected.Figure66showsthatastheloadsare
increasedtothe2020levelsthespotpricesincreasetohigherlevels.
0
10
20
30
40
50
60
70
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
S
p
o
t

P
r
i
c
e

(
$
/
M
W
h
)
2010X Perfect forecast for commitment
2010X Estimated forecast for commitment
2010X No forecast for commitment
Figure 65. Impact of Intermittent Forecast on Spot Price (2010X).

93
0
10
20
30
40
50
60
70
80
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
S
p
o
t

P
r
i
c
e

(
$
/
M
W
h
)
2010X Perfect forecast for commitment
2010X Estimated forecast for commitment
2010X No forecast for commitment
2020 Estimated forecast for commitment
Figure 66. Impact of Intermittent Forecast on Spot Price (2010X and 2020).
ThechangesinthesystemvariablecostsofoperationaredemonstratedinFigure67.Thischart
showsthatthereissignificantvalueinimplementingstateoftheartestimatingtechniques
acrossthesystem.Andthatthereisstilladditionalvaluethatcanbegainedbyimprovingthe
existingmethodologies.Whendividedbytheintermittentenergytheestimatedforecasthasa
benefitof$4.37/MWhandtheperfectforecastcouldaddanadditional$.95/MWh.Asdiscussed
before,ignoringthewindinthedayaheadcommitmentcanresultinseriousovercommitment
ofthermalgeneration.Eveniftheforecasterroris20%,a5,000MWforecastwouldresultin
actualgenerationintherangeof4,000MWto6,000MW.Itisfarbettertobeoverorunder
committedby1,000MWthantobealwaysovercommittedbytheentire4,000MWto6,000
MW.ThecurrentdirectionoftheCaliforniaISOtoforecastofallintermittentgenerationonthe
Californiagridiscritical.Supportingandenhancingthesystemcapabilitiesinthisareaarein
thebestinterestoftheoverallsystem.
94
-
50
100
150
200
250
2010T A 2010T B 2010T C 2010X A 2010X B 2010X C
T
o
t
a
l

V
a
r
i
a
b
l
e

C
o
s
t

c
h
a
n
g
e
s

(
$
M
)
no forecast - perfect forecast
no forecast - esti mated forecast
esti mated forecast - perfect forecast
Figure 67. Total Operating Cost Impact of Intermittent Forecasting.

Figure68andFigure69showthegeneratorrevenuereductionsbygeneratortypeforboth
CaliforniaandWECCrespectively.Thefirstcolumnshowstheimpactofnonintermittent
renewablegeneration(biomassandgeothermal)andthesecondcolumnshowstheimpactof
theintermittentgeneration(windandsolar).Thisassumedanestimatedforecastforthe
intermittentgeneration.Thebulkoftheimpactisonthecombinedcycleunits.Latersections
willshowthatthesearetheunitsthatarelargelydisplacedbytherenewablegeneration.The
revenuereductionsonothergenerationtypeswereduelargelyorcompletelytothereduction
inspotprices.Thethirdcolumnshowstheincrementalrevenuereductionsduetoignoringthe
intermittentgenerationinthecommitment.Althoughtheimpactoncombinedcycleunitsis
increasedslightlyitislargelythehydro,nuclearandsteamgenerationthatisimpactedmost.
Althoughlittleornoenergyisdisplaced,thespotpriceimpactsontheirrevenueissignificant
duetotheamountofenergyproduced.Thisunderscoresthefactthatifintermittentgeneration
isbeingaddedtothesystemitisintheinterestofallofthegenerationtousethebest
intermittentforecastpossible.
95
0
100
200
300
400
500
600
700
800
900
1000
Bi omass &
Geothermal
Wi nd & Sol ar Esti mated forecast -
No forecast
G
e
n
e
r
a
t
o
r

R
e
v
e
n
u
e

R
e
d
u
c
t
i
o
n
s

(
$
M
)
Steam
PSH
Nucl ear
Hydro
Gas Turbi ne
Combi ned Cycl e
Figure 68. California Generator Revenue Reductions by Type (2010T).

0
1000
2000
3000
4000
5000
6000
Bi omass &
Geothermal
Wi nd & Sol ar Esti mated forecast -
No forecast
G
e
n
e
r
a
t
o
r

R
e
v
e
n
u
e

R
e
d
u
c
t
i
o
n
s

(
$
M
)
Steam
PSH
Nucl ear
Hydro
Gas Turbi ne
Combi ned Cycl e
Figure 69. WECC Generator Revenue Reductions by Type (2010T).

4.4. Operations
Thissectionlooksattheoveralloperationalimpactoftheadditionofrenewablegenerationand
lessattheeconomicside.Figure70showsthatvirtuallyalloftheenergydisplacedinside
Californiawillcomefromcombinedcycleunitsandthatroughly60%ofalloftherenewable
energywillbeabsorbedbyreductionsintheimportsfromoutsidethestate.
96
-30000
-20000
-10000
0
10000
20000
30000
40000
CC STM GT PSH Imports Renewabl es
E
n
e
r
g
y

C
h
a
n
g
e

(
G
W
h
)
2002 shapes
2003 shapes
2004 shapes
Figure 70. California Energy Change Due to Renewables (2010T).

LookingatthesituationfromtheWECCperspectiveinFigure71showsthatthedisplaced
Californiaimportsarealmostallfromcombinedcycleunitsoutsidethestate.
-40000
-30000
-20000
-10000
0
10000
20000
30000
40000
CC STM GT PSH Renewabl es
E
n
e
r
g
y

C
h
a
n
g
e

(
G
W
h
)
2002 shapes
2003 shapes
2004 shapes
Figure 71. WECC Energy Change Due to Renewables (2010T).

Thesecurvesarenottoimplythatalargeportionoftheregulationandloadfollowingburden
ofanaggressiveCaliforniarenewableenergyexpansionwillbefoistedupontheirneighbors.
Thesearescheduledreductionsintheroughly10,000MWofimportsthatroutinelycrossthe
Californiaborders.Mostofthereductionswouldbeexpectedtotakeplaceinthedayahead
marketwithslightadjustmentstoimportsinthehouraheadmarketduetochangesinthe
intermittentenergyforecastsinmuchthesamewaythatchangesaremadenowdueto
fluctuationsintheloadforecast.Intrahourvariationsintheintermittentgenerationwouldbe
accommodatedbyinstategeneration.Duetotherelativelyhighlevelofimportscurrently
comingintoCalifornia,itisonlynaturaltoexpectthattheadditionofanynewgeneration
sourcewithinCaliforniawouldresultinareductionintheseimports.
97
ThecurvesinFigure72showchangesintheCaliforniarenewablegenerationforthevarious
scenarios.Thelowestcurverepresentstheexistinglevelofbothintermittentandnon
intermittentgenerationwithinthestate.Thenextcurveuprepresentstheadditionofnew
geothermalandbiomassgeneration.Ascanbeseen,thefirsttwocurvesareroughlyequidistant
fortheentireyearsincethenewnonintermittentgenerationisfairlyconstantfortheyear.The
thirdcurveshowstheadditionofthenewwindandsolargenerationinthe2010Tscenario.The
finalcurverepresentsthe2010Xscenariopenetrations.Figure73showsjustthewindandsolar
valuesforthesamescenarios.Notethatevenintheextremescenario(2010X)thetotal
intermittentgenerationislessthan4,000MWforoverhalfoftheyear.
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
0 2000 4000 6000 8000
Hours
G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T no new Renewabl es
2010T no new Wi nd & Sol ar
2010T base case
2010X
Figure 72. Annual Duration Curves California Renewable Generation.

98
0
2000
4000
6000
8000
10000
12000
0 2000 4000 6000 8000
Hours
G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T base case
2010X
Figure 73. Annual Duration Curves California Wind and Solar Generation.
ThesetsofcurvesinFigure74showtheannualdurationcurvesforthegenerationfrom
Californianuclear,steamandgasturbineunits.Ascanbeseen,thereisnotmuchchangein
theiroveralloutput.
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
0 2000 4000 6000 8000
Hours
G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T base case
2010X
0
1000
2000
3000
4000
5000
6000
7000
8000
0 2000 4000 6000 8000
Hours
G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T base case
2010X
0
200
400
600
800
1000
0 2000 4000 6000 8000
Hours
G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T base case
2010X
Not e scale
Steam
Gas Turbine
Nuclear
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
0 2000 4000 6000 8000
Hours
G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T base case
2010X
0
1000
2000
3000
4000
5000
6000
7000
8000
0 2000 4000 6000 8000
Hours
G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T base case
2010X
0
200
400
600
800
1000
0 2000 4000 6000 8000
Hours
G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T base case
2010X
Not e scale
Steam
Gas Turbine
Nuclear
Figure 74. Annual Duration Curves California Nuclear, Steam, and Gas Turbines.
ThecurvesinFigure75showtheannualgenerationfromtheCaliforniahydrogeneration.This
wasinitiallyquitesurprisingbecauseitseemedtoindicatethatnotmuchchangedinthehydro
operationwhentherenewablegenerationwasadded.Althoughthegeothermalandbiomass
99
energyisproducedinafairlyconstantmannerwhichwouldnotbeexpectedtoimpactthe
hydro,theintermittentnatureofthewindandsolarwasexpectedtocausechanges.
0
2000
4000
6000
8000
10000
12000
14000
0 2000 4000 6000 8000
Hours
G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T base case
2010X
Figure 75. Annual Duration Curves California Hydro Generation.

ThecurvesinFigure76showtheoperationoftheCaliforniahydrogenerationforoneweek
fromthesimulationswithandwithoutthenewwindandsolargeneration.Itcanbeseenhere
thatthehourlyhydrogenerationdidshift,althoughthetotalgenerationfortheweekremained
approximatelyconstant.Figure77plotsthisdeltaoperationfortheentireyear.Thesmooth
yellowcurvesuperimposedovertheotheristhechronologicaldeltassortedfromhightolow
fortheyear.Fromthiscurveitcanbeseenthattheshiftinthehydrogenerationisgenerallyless
than+/1,500MWoverthecourseoftheyear.Theshiftinoperationonlyexceed+/1,000MW
forroughly300hoursoftheyearandwaslessthan+/500MWalmost85%ofthetime.This
wasfortheadditionofalmost7,000MWofnewintermittentgeneration.Eventheexpansionto
the2010Xscenarioonlybroadenedtheshouldersofthecurveslightly.Laterinthissectionwe
willexaminetheabilityofthehydrotoacceptthismuchchange.
100
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
0 50 100 150
Hours
H
y
d
r
o

G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T base case
2010T - No new Wi nd & Sol ar
Figure 76. One Week Change in California Hydro Operation (2010T).

-2500
-2000
-1500
-1000
-500
0
500
1000
1500
2000
0 2000 4000 6000 8000
Hours
C
h
a
n
g
e

i
n

H
y
d
r
o

O
p
e
r
a
t
i
o
n

(
M
W
)
Hydro shi ft (2010T)
Del ta durati on curve 2010T
Del ta durati on curve 2010X
Figure 77. Annual Change in California Hydro Operation (2010T and 2010X).
TherewassomeconcernthattheanalysismightrelytooheavilyonhydrooutsideofCalifornia.
Withtheexceptionofafewofthelargerunits(ChiefJoseph,HooverandGlenCanyon)the
hydrooutsideofCaliforniawasscheduledagainsttheloadcurvesforthelocalregion,notthe
entiresystem.HydrowithinCaliforniawasscheduledagainstthelocalarealoads(i.e.PG&E,
SCE,SMUD,etc.).AlatersectionexaminestheflowsontheCaliforniaOregonInterfaceand
theoverallimpactontheNorthwesthydro.Figure78showsthechangeinhydrogenerationfor
allofWECC.ThiscurvehasasimilarshapetothehydrochangesinCaliforniawiththechanges
101
nowfallingmostlywithin+/2000MW.Figure79isanannualhistogramofthehourlychanges
bothinsideCaliforniaandforallofWECC.Thesimilarityofthecurvesdemonstratesthatmost
oftheshiftishappeningwithintheCaliforniaborders.
-4000
-3000
-2000
-1000
0
1000
2000
3000
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
C
h
a
n
g
e

i
n

W
E
C
C

H
y
d
r
o

O
p
e
r
a
t
i
o
n

Figure 78. Annual Change in WECC Hydro Operation (2010T).
0
500
1000
1500
2000
2500
3000
3500
4000
4500
<

-
4
0
0
0
<

-
3
5
0
0
<

-
3
0
0
0
<

-
2
5
0
0
<

-
2
0
0
0
<

-
1
5
0
0
<

-
1
0
0
0
<

-
5
0
0
<

0
>

0
>

5
0
0
>

1
0
0
0
>

1
5
0
0
>

2
0
0
0
>

2
5
0
0
#

H
o
u
r
s
WECC
Cal ifornia
Figure 79. Annual Histogram of Hydro Shift (2010T).

Figure80showstheannualdurationcurveoftheoperationoftheCaliforniapumpedstorage
hydro,PSH.Asexpected,thePSHoperationincreasedasthelevelofrenewablesincreased.
Althoughthemaximumoperatinglevelswerereachedforafewhundredhoursperyearitdoes
102
notappearfromthesecurvesthatadditionalPSHresourcesarerequiredforthesystem.Aswill
beshownlater,theutilizationofthePSHincreasesifthemaneuverabilityoftheconventional
hydroissignificantlyconstrained.
-5000
-4000
-3000
-2000
-1000
0
1000
2000
3000
4000
5000
0 2000 4000 6000 8000
Hours
G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T base case
2010X
Figure 80. Annual Duration Curves - California Pumped Storage Hydro Operation.
So,whatgenerationisbeingdisplaced?AsdiscussedbeforeandshowninFigure81,the
Californiacombinedcyclegenerationdropssignificantlyastherenewablegenerationisadded.
Thesimulationassumedthatcombinedcycleunitscouldonlyrampdownto50%oftheir
capacity.Anyfurtherreductionwouldrequirethemtobedecommittedforaperiodoftime.As
willbeshown,thisiswellwithinthedesigncharacteristicsandhistoricaloperatingcapabilityof
mostexistingandfuturecombinedcycleunits.
0
2000
4000
6000
8000
10000
12000
14000
16000
0 2000 4000 6000 8000
Hours
G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T base case
2010X
Figure 81. Annual Duration Curves - California Combined Cycle Generation.

103
ThecurvesinFigure82showthefinalareaofimpact.Asmentionedearlier,notsurprisinglythe
importsintoCaliforniadecreasedsubstantiallywhensignificantamountsofnewgeneration
wereaddedtotheCaliforniagrid.BasedoneconomicsfortheentireWECCthesimulationonly
calculatedimportsintoCalifornia.HistoricallyCaliforniahasalsoexportedenergytothe
northwestduringlowloadnighttimeperiods.Theseexportswouldhelptoreduceany
difficultiesthatmightbeencounteredinthelowloadperiodsandwouldimprovetheabilityof
thesystemtointegrateadditionalrenewablegeneration.
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
0 2000 4000 6000 8000
Hours
G
e
n
e
r
a
t
i
o
n

(
M
W
)
2010T base case
2010X
Figure 82. Annual Duration Curves - California Imports.

4.4.1. California Hydroelectric Operation
Figure77showedthatthesystemmightneedtoshiftthehydroroughly+/1,500MWtofully
accommodatethevariabilityfromtheadditional7,000MWofwindandsolargenerationinthe
2010Tscenario.ThequestionisHowmuchvariabilityisthereinthehydro?Figure83shows
thehistoricaloperationoftheCaliforniahydroforsampleweeksinMayof2004and2006.This
datawasprovidedbytheCaliforniaISO.Theyear2004wasatypicalhydroyearandthe
curveinFigure83showsthehydrovaryingover4,000MWoverthecourseoftheweek.The
year2006,however,wasahighhydroyearandthevariationsinthehydroweremuchless
pronounced.Buteveninthishighhydroweektherewasalmost2,000MWofdifference
betweenthepeakandvalleyoperationofthehydro.
104
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
May, fi rst week (10 mi nute val ues)
C
a
l
i
f
o
r
n
i
a

H
y
d
r
o

(
M
W
)
2004
2006
Figure 83. California Historical Hydro Operation - Sample May Week.

ThecurvesinFigure84showthehistoricaloperationfortheentiremonth.Thisshowsthatthe
hydromightbeexpectedtovaryitsoperationbyasmuchas2,500MWevenoverthecourseof
averywetmonth.Figure85andFigure86showthechronologicalCaliforniaISOhydro
operationforthemonthsofJuneandJulyof2004and2006.Figure87presentsthedatafromthe
lastthreecurvesinadurationformat.Theseshowthateventhehighhydromonthsmaybe
expectedtohaveupto4,000MWofvariationwithinthemonthandthattypicallythevariation
mightbeasmuchas5,000MWorevenmore.Thiswouldseemtoindicatethatcertainlyin
typicalyears,buteveninhighhydroyearsthereshouldbesufficientflexibilityinthehydroto
accommodatesignificantamountsofintermittentrenewableadditions.
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
May (10 mi nute val ues)
C
a
l
i
f
o
r
n
i
a

H
y
d
r
o

(
M
W
)
2004
2006
Figure 84. California Historical Hydro Operation May.

105
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
June (10 minute values)
C
a
l
i
f
o
r
n
i
a

H
y
d
r
o

(
M
W
)
2004
2006
Figure 85. California Historical Hydro Operation June.

0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Jul y (10 mi nute val ues)
C
a
l
i
f
o
r
n
i
a

H
y
d
r
o

(
M
W
)
2004
2006
Figure 86. California Historical Hydro Operation July.

106
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
10 mi nute val ues
C
a
l
i
f
o
r
n
i
a

H
y
d
r
o

(
M
W
)
May 2004 June 2004 Jul y 2004
May 2006 June 2006 Jul y 2006
Figure 87. California Hydro Historical Monthly Duration Curves.

Ithasbeenpointedoutthatthehydrohasmanyconstraintsplaceduponitandisnotunderthe
directcontroloftheCaliforniaISO.Whilethisistrue,itisalsotruethatthehydrowillvaryits
operations,withinthelimitsplacedonit,inordertochasethehighermarginalcostvaluesin
thehigherloads.ItwouldseemthatiftheCaliforniaISOweretopublishtheforecastedwind
andsolaroperationalongwiththeforecastedloadsthatthehydrooperatorswouldstriveto
operatetothenetloadminusintermittentschedulesastheycurrentlyoperatetothe
forecastedloads.ThevariabilityofthehydroisexaminedinmoredetailinSection4.4.4
4.4.2. Combined Cycle Operation
Basedonourunderstandingofthephysicallimitationsoftheunitsanddiscussionswithother
operatorsthecombinedcycleunitsweremodeledwithaminimumoperatingpointequalto
50%oftheirratedcapacity.Ifgenerationhadtobereducedbelowthe50%levelthentheunit
wouldbedecommittedforaminimumof8hours.Themodelfactoredintheminimumdown
timeforeachoftheunitsaswellastheirstartupcoststodecideiftheunitsshouldbeleftonat
theirminimumoperatingpointorturnedoff.Inordertovalidateourassumptionswe
examinedthehourlyoutputfromvariouslargecombinedcycleunitsoperatinginCalifornia.
ThisoutputwasdeterminedfromtheContinuousEmissionsMonitoringSystem,CEMS,
databasefor2005.TheresultsforthreeoftheseunitsareshowninFigure88,Figure89and
Figure90.Thefirstunitcyclesfromabove800MWtoroughly400MWonadailybasisand
occasionallycyclesdowntoabout260MWorroughly30%.Thesecondunitcyclesdownto60%
routinelyalthoughtherearemultipleinstanceswhereitcyclestoabout30%.Thethirdunit
operatesinasimilarmanner.Althoughthiswasnotanexhaustivesearch,itindicatesthat50%
isareasonableassumptionandmightevenbeconservative.
107
Newercombinedcycleunitscancycleto30%outputandstillmaintainemissionlevels.In
addition,manyexistingunitscouldberetrofittedtoimprovetheircyclingcapability.Higher
penetrationsofintermittentgenerationrequireasystemthatcanrespondtotheirfluctuations.It
wouldbebeneficialforthemarkettoencourageincreasedcyclingcapabilityinthegeneration
mix.
0
100
200
300
400
500
600
700
800
900
1
/
1
/
2
0
0
5
2
/
1
/
2
0
0
5
3
/
1
/
2
0
0
5
4
/
1
/
2
0
0
5
5
/
1
/
2
0
0
5
6
/
1
/
2
0
0
5
7
/
1
/
2
0
0
5
8
/
1
/
2
0
0
5
9
/
1
/
2
0
0
5
1
0
/
1
/
2
0
0
5
1
1
/
1
/
2
0
0
5
1
2
/
1
/
2
0
0
5
D
e
l
t
a

E
n
e
r
g
y

C
e
n
t
e
r

(
C
P
N
)

N
e
t

G
e
n
e
r
a
t
i
o
n

O
u
t
p
u
t

(
M
W
)
Figure 88. 2005 CEMS Data Delta Energy Center.

0
100
200
300
400
500
600
1
/
1
/
2
0
0
5
2
/
1
/
2
0
0
5
3
/
1
/
2
0
0
5
4
/
1
/
2
0
0
5
5
/
1
/
2
0
0
5
6
/
1
/
2
0
0
5
7
/
1
/
2
0
0
5
8
/
1
/
2
0
0
5
9
/
1
/
2
0
0
5
1
0
/
1
/
2
0
0
5
1
1
/
1
/
2
0
0
5
1
2
/
1
/
2
0
0
5
H
a
y
n
e
s

G
e
n
e
r
a
t
i
n
g

S
t
a
t
i
o
n

N
e
t

U
n
i
t

O
u
t
p
u
t

(
M
W
)
Figure 89. 2005 CEMS Data Haynes Generating Station.
108
0
100
200
300
400
500
600
1
/
1
/
2
0
0
5
2
/
1
/
2
0
0
5
3
/
1
/
2
0
0
5
4
/
1
/
2
0
0
5
5
/
1
/
2
0
0
5
6
/
1
/
2
0
0
5
7
/
1
/
2
0
0
5
8
/
1
/
2
0
0
5
9
/
1
/
2
0
0
5
1
0
/
1
/
2
0
0
5
1
1
/
1
/
2
0
0
5
1
2
/
1
/
2
0
0
5
V
a
l
l
e
y

G
e
n
e
r
a
t
i
n
g

S
t
a
t
i
o
n

N
e
t

U
n
i
t

O
u
t
p
u
t

(
M
W
)
Figure 90. 2005 CEMS Data Valley Generating Station.

4.4.3. Ramp Rate and Range of Operation
Oneofthekeyoperationalfactorsthatisofinterestishowmuchrampingcapabilityisavailable
eachhourbasedonthecommitmentanddispatchofthegenerationandthecorrespondingunit
rampingcapabilities.Figure91showsasampleunitcommitmentschedulebygenerationtype
foraweek.Figure92showsthecorrespondingunitdispatch.Basedontherampratesofthe
individualunitstheamountoframpupanddowncapability,inMW/minute,wasthen
calculatedaswellastheavailablerangeupanddown,inMW.Asanexample,aunits
contributiontothesystemrangeupcapabilityequalsitcapacityminusitsdispatchforthehour.
Theunitsrampupcapabilityisequaltoitscapacitytimesitsramprateaslongasthisisless
thanthecalculatedrange.TheresultingtotalsareshownforthesampleweekinFigure93.Since
therampdowncapacitycanbecritical,particularlyduringlowloadperiods,thisisshownin
Figure94withabreakdownbyunittype.Althoughthehydrogenerationisobviously
providingthemajorityoftherampingcapability,Figure95showsthatevenwithoutthehydro
thesystemwouldgenerallyhaveatleast200MW/minuteavailable.Basedonthestatistical
analysispresentedintheprevioussection,threestandarddeviationsoftheoneminutedeltais
expectedtobeabout150MW.
109
0
10000
20000
30000
40000
50000
60000
1 25 49 73 97 121 145
Hour
G
e
n
e
r
a
t
i
o
n

(
M
W
)
IMPORTS
HYDRO
PSH
WIND
SOLAR PV
SOLAR CON.
GAS TURBINE
COMB. CYCLE
STEAM
BIOMASS
GEOTHERMAL
NUCLEAR
Figure 91. Commitment Week of May 10

th
.
0
5000
10000
15000
20000
25000
30000
35000
40000
1 25 49 73 97 121 145
Hour
G
e
n
e
r
a
t
i
o
n

(
M
W
)
IMPORTS
HYDRO
PSH
WIND
SOLAR PV
SOLAR CON.
GAS TURBINE
COMB. CYCLE
STEAM
BIOMASS
GEOTHERMAL
NUCLEAR
Figure 92. Dispatch Week of May 10

th
.
110
-15000
-10000
-5000
0
5000
10000
15000
20000
1 25 49 73 97 121 145
Hours
M
W
,

M
W
/
m
i
n
u
t
e
Range up (MW)
Ramp up (MW/mi n.)
Ramp Down (MW/mi n.)
Range down (MW)
Figure 93. Ramp Rate and Range Capability - Week of May 10

th
.
0
500
1000
1500
2000
2500
3000
3500
1 25 49 73 97 121 145
Hours
R
a
m
p

R
a
t
e

D
o
w
n

C
a
p
a
c
i
t
y

(
M
W
/
m
i
n
)
PSH
GAS TURBINE
STEAM
COMB. CYCLE
HYDRO
Figure 94. Ramp Rate Down Capability - Week of May 10

th
.
111
0
100
200
300
400
500
600
700
800
900
1000
1 25 49 73 97 121 145
Hours
R
a
m
p

R
a
t
e

D
o
w
n

C
a
p
a
c
i
t
y

(
M
W
/
m
i
n
)
PSH
GAS TURBINE
STEAM
COMB. CYCLE
Figure 95. Ramp Rate Down Capacity Without Conventional Hydro Week of May 10
th
.
ThescatterplotsinFigure96andFigure97werecreatedtoshowtheramprateandrangedown
capabilityversustheCaliforniaload.Asexpected,thesevaluestypicallyexceedwhatis
generallyneededinallbutahandfulofthelowloadhours.Astheloadsincreasetheramprate
andrangedownarenolongercritical.TheMay15
th
valuesthatareflaggedareoneofthetime
framesexaminedinthequasisteadystateanalysisinSection5.0
0
500
1000
1500
2000
2500
3000
3500
4000
4500
20000 25000 30000 35000 40000 45000 50000 55000 60000
California Load (MW)
R
a
m
p

R
a
t
e

D
o
w
n

C
a
p
a
b
i
l
i
t
y

(
M
W
/
m
i
n
)
Ramp rate down capabi l i ty
May 15th val ues
Figure 96. Ramp Rate Down Capability Versus California Load (2010X).
112
0
5000
10000
15000
20000
25000
20000 25000 30000 35000 40000 45000 50000 55000 60000
R
a
n
g
e

D
o
w
n

C
a
p
a
b
i
l
i
t
y

(
M
W
)
Range down capabi l i ty
May 15th val ues
Figure 97. Range Down Capability Versus California Load (2010X).

Inasimilarmanner,Figure98andFigure99showtheramprateandrangeupcapability.These
curveswouldindicatethatthereisalwayssufficientramprateandrangeupcapabilityonthe
system.
0
500
1000
1500
2000
2500
3000
3500
4000
4500
20000 25000 30000 35000 40000 45000 50000 55000 60000
R
a
m
p

R
a
t
e

U
p

C
a
p
a
b
i
l
i
t
y

(
M
W
/
m
i
n
)

Figure 98. Ramp Rate Up Capability Versus California Load (2010X).
113
0
5000
10000
15000
20000
20000 25000 30000 35000 40000 45000 50000 55000 60000
R
a
n
g
e

U
p

C
a
p
a
b
i
l
i
t
y

(
M
W
)

Figure 99. Range Up Capability Versus California Load (2010X).
ThecurvesinFigure100showtheannualrampdowncapacityindurationcurvesforthe
variousscenariosexamined.Althoughthereissomevariationitdoesntseemtochange
significantlyforthedifferentscenarios.
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
R
a
m
p

D
o
w
n

C
a
p
a
c
i
t
y

(
M
W
/
m
i
n
.
)
2010T - Base system
2010T - No new renewabl es
2010X system
Figure 100. Annual Ramp Rate Down Capability.

ThecurvesareexpandedinFigure101tohighlightthehourswithlowerrampingcapacity.
Again,itappearsthatonlyahandfulofhoursmayfallwithinthe150MWvalueneededto
coverthreestandarddeviations.
114
0
200
400
600
800
1000
1200
1400
7000 7200 7400 7600 7800 8000 8200 8400 8600
Hours
R
a
m
p

D
o
w
n

C
a
p
a
c
i
t
y

(
M
W
/
m
i
n
.
)
2010T - Base system
2010T - No new renewabl es
2010X system
Figure 101. Annual Ramp Rate Down Capability (zoom).

4.4.4. Sensitivities
ItwasindicatedthattheHelmspumpedstoragehydro(PSH)plantisoftenconstrainedinthe
pumpingmodesuchthatonlytwoofthethreepumpsarecapableofoperating.Asensitivity
casewasrunwiththepumpingcapabilityreducedfortheentireyear.Figure102showsthat
undertheseconditionstheoperationchangesslightly,butislimitinginthepumpingmodefor
onlyafewadditionalhours.ThevariablecostsofoperationinCaliforniaincreasedabout$2
millionperyearwiththereducedPSHcapability.
-1500
-1000
-500
0
500
1000
1500
0 2000 4000 6000 8000
Hours
M
W
Three pumps
Two pumps
Figure 102. Operation of Helms Pumped Storage Hydro (2010X).

Figure103showsthedailyminimum,maximumandrangefortheconventionalhydroforallof
2006.Figure104showstheaveragedailyhydrogenerationandrangefortheMarchthrough
115
Augustperiodforboth2004and2006.Sincehydrowasanimportantcontributortotheoverall
systemrampingcapabilitytwosensitivitycaseswererunwiththehydroseverelyconstrained.
Thehydrooutputwasconstrainedtobebetween50%and75%oftheunitrating,energylimits
permitting.Figure105showstheimpactonthehydrooperationforasampleweekinMayfor
the2010Tscenario.ThefirstsensitivityconstrainedthehydroforJanuarythroughMay,
inclusive,andthesecondsensitivityconstrainedthehydrofortheentireyear.Figure106shows
theimpactontheCaliforniahydrooverthecourseoftheentireyearforthebasecaseand
sensitivities.Whenthehydrowasconstrainedfortheentireyearthetotalrangeofoperation
decreasedfromover10,000MWtoroughly4,000MW.Whenconstrainedforthefirstfive
monthstheoverallannualrangedidntchange,buttheimpactwasstillsignificant.Thefullyear
ofconstraintsincreasedtheannualoperatingcostinCaliforniabyalmost$70million.The
Californiaoperatingcostincreasedbylessthan$30millionperyearwhenthehydrowasonly
constrainedforthefirstfivemonths.
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
0 50 100 150 200 250 300 350
Day (2006)
H
y
d
r
o

m
a
x
,

m
i
n

a
n
d

r
a
n
g
e

(
M
W
)
Maximum
Minimum
Range
Figure 103. Historical Hydro Daily Minimum, Maximum, and Range, 2006.
116
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
0 30 60 90 120 150 180
March to August (dai l y val ues)
A
v
e
r
a
g
e

a
n
d

R
a
n
g
e

o
f

H
y
d
r
o

G
e
n
e
r
a
t
i
o
n

(
M
W
)
Average 2004
Range 2004
Average 2006
Range 2006
Figure 104. Daily Range Of Hydro Operation, Summer 2004 and 2006.
0
2000
4000
6000
8000
10000
12000
14000
0 50 100 150
Hours
C
a
l
i
f
o
r
n
i
a

H
y
d
r
o

D
i
s
p
a
t
c
h

(
M
W
)
Base case
Constrai ned hydro
Figure 105. Constrained Versus Base Hydro for a Sample May Week (2010T).
117
0
2000
4000
6000
8000
10000
12000
14000
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
C
a
l
i
f
o
r
n
i
a

H
y
d
r
o

D
i
s
p
a
t
c
h

(
M
W
)
base case
5 month constrai ned Hydro
Ful l year constrai ned Hydro
Figure 106. Annual Duration Curve California Hydro Generation (2010T).

Figure107andFigure108showthenewimpactsontheCaliforniarampdowncapability.The
extremecaseseemedtohavesufficientrampingcapacityforalmostallofthehours,although
therampdowncapabilitydiddropbelow200MW/minforabout500hoursintheyear.
Althoughthefivemonthconstrainedassumptionsarestillquiteconservativetheresultsshow
muchlessimpact,particularlyforthelowloadhours.
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
R
a
m
p

D
o
w
n

C
a
p
a
c
i
t
y

(
M
W
/
m
i
n
)
Base
Figure 107. Ramp Down Capacity With Constrained Hydro, 2010T.

118
0
200
400
600
800
1000
1200
1400
6000 6500 7000 7500 8000 8500
Hours
R
a
m
p

D
o
w
n

C
a
p
a
c
i
t
y

(
M
W
/
m
i
n
)
Base
Figure 108. Ramp Down Capacity With Constrained Hydro, 2010T (zoom).
Whentheflexibilitywasremovedfromtheconventionalhydrothenthesystemdependedmore
onthePumpedStorageHydrogeneration.Figure109showsthechangeinoperationofthe
CaliforniaPSH.ThenumberofhourswhentheexistingPSHiscapacityconstrainedmorethan
tripledwhenthehydrowasconstrainedallyear.ThePSHalsoincreasedsomewhatwhenthe
hydrowasconstrainedforthefirstfivemonths.ThecurvesinFigure110showthehistorical
changeinoperationfortheHelmsPSHplantoverthesixmonthperiodfromMarchthrough
August.AswassimulatedinFigure109,thetotalPSHoperationincreasesasthevariabilityon
theconventionalhydroisreduced.
119
-5000
-4000
-3000
-2000
-1000
0
1000
2000
3000
4000
5000
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
C
a
l
i
f
o
r
n
i
a

P
S
H

O
p
e
r
t
i
o
n

(
M
W
)base case
Figure 109. Annual Duration Curve California PSH Operation with Constrained Hydro
(2010T).
-1000
-800
-600
-400
-200
0
200
400
600
800
1000
1 4231 8461 12691 16921 21151 25381
March to August (10 mi nute val ues)
H
e
l
m
s

P
S
H

d
u
r
a
t
i
o
n

c
u
r
v
e

(
M
W
)
Total Hel ms, 2004
Total Hel ms, 2006
Figure 110. Comparison of Helms Operation, Summer 2004 and 2006.

Thepumps(bothPSHandconventionalpumpingloads)arealsoapotentiallyimportantsource
ofsystemmaneuverability.Figure111showsatypicalweekofpumpingoperationbothwith
andwithoutthepumpedstoragefacilityatHelms.Figure112showsthedailyrangeof
operationforallof2006.CaliforniacurrentlyhasovertwothousandMWofpumpsthatcycle
duringthedaytotakeadvantageofthelowoffpeakcostofenergy.Ifincreasedlevelsofwind
120
generationcausethelowcostperiodstoshiftthenthepumpswilllikelyfollow.Infact,itmay
beeconomicallyattractivetoaddadditionalpumpstoincreasethemaneuverablerange
available.
-3,000
-2,500
-2,000
-1,500
-1,000
-500
0
0 20 40 60 80 100 120 140 160
Hours (1st week of January, 2006)
P
u
m
p

L
o
a
d

(
M
W
)
Total Pump Load
Pump w/o Hel ms Pump
Figure 111. Sample Week of Pumping Operation, January 2006.

-2,500
-2,000
-1,500
-1,000
-500
0
0 50 100 150 200 250 300 350
Days (2006)
P
u
m
p

L
o
a
d

(
M
W
)
Range of Total Pump Load (-1436)
Range of Pump w/o Hel ms Pump (-903)
Figure 112. Historical Pumping Operation, 2006.

121
4.5. Emissions
Oneofthemajorbenefitsofrenewablegenerationisthereductionofemissionscausedbyfossil
fuelfiredgeneration.Figure113showsthattheNOXemissionsinCaliforniawoulddecreaseby
almost500tonsperyearandthatSOXemissionswoulddecreasebyover250tonsperyearfor
the2010Tscenariowithallnewrenewablegenerationadded.Butsincemuchofthegeneration
shifthappensoutsideofCaliforniaitisimportanttolookattheWECCimpact.Figure114
showsthatwithinWECCtheNOXemissionswoulddecreaseby2500tonsandtheSOX
emissionswoulddecreaseby500tons.Figure115andFigure116showthecorresponding
reductionsforthenewwindandsolargenerationonly.
0
100
200
300
400
500
600
NOX reduction SOX reduction
E
m
i
s
s
i
o
n

R
e
d
u
c
t
i
o
n
s

(
t
o
n
s
)
2002 shapes
2003 shapes
2004 shapes
Figure 113. California Emission Reductions Due to Renewables (2010T).

0
500
1000
1500
2000
2500
3000
E
m
i
s
s
i
o
n

R
e
d
u
c
t
i
o
n
s

(
t
o
n
s
)
2002 shapes
2003 shapes
2004 shapes
Figure 114. WECC Emission Reductions Due to renewables (2010T).

122
0
50
100
150
200
250
300
E
m
i
s
s
i
o
n

R
e
d
u
c
t
i
o
n
s

(
t
o
n
s
)2002 shapes
2003 shapes
2004 shapes
Figure 115. California Emission Reductions Due to New Wind and Solar Generation
(2010T).
0
200
400
600
800
1000
1200
1400
E
m
i
s
s
i
o
n

R
e
d
u
c
t
i
o
n
s

(
t
o
n
s
)
2002 shapes
2003 shapes
2004 shapes
Figure 116. WECC Emission Reductions Due to New Wind and Solar Generation (2010T).
4.6. Transmission Path Loading
Thefinaloperationalimpactthatwasexaminedwastheimpactonthetransmissionloading.
Althoughsomenewtransmissionwillberequiredtoconnectthenewgenerationintothegrid,
thefocuswasonthemajorinterfacesintoandwithinthestateofCalifornia.Figure117shows
theflowsonPath15increasingintheSouthtoNorthdirection,butdecreasingintheopposite
direction.Figure118showstheflowsconsistentlyreducingfromArizonatoCaliforniaasmore
resourcesareaddedontotheCaliforniagrid.Inparticular,thehourswithcongestionare
123
significantlyreduced.SimilarimpactsareshowninFigure119andFigure120fortheWestof
ColoradoRiverandSouthernCaliforniaImportinterfaces.Sowhilesomenewtransmissionis
required,congestiononmanyoftheexistingtransmissioncorridorsarerelieved.
-4000
-3000
-2000
-1000
0
1000
2000
3000
4000
5000
6000
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
F
l
o
w

(
M
W
)
2010T No new renewabl es
2010T No new i ntermi ttents
2010T base case
2010X system
Sout h t o Nort h 5400 MW
Nort h to Sout h 3265 MW
MIDWAY 500-LOSBANOS500
GATES 500-LOSBANOS500
GATES 230-PANOCHE 230
GATES 230-HENTAP1 230
S
o
u
t
h

t
o

N
o
r
t
h
N
o
r
t
h

t
o

S
o
u
t
h
-4000
-3000
-2000
-1000
0
1000
2000
3000
4000
5000
6000
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
F
l
o
w

(
M
W
)
2010T base case
2010X system
S
o
u
t
h

t
o

N
o
r
t
h
N
o
r
t
h

t
o

S
o
u
t
h
S
o
u
t
h

t
o

N
o
r
t
h
N
o
r
t
h

t
o

S
o
u
t
h
Figure 117. Transmission Flow Duration Curves Path 15: South of Los Banos.
0
1000
2000
3000
4000
5000
6000
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
F
l
o
w

(
M
W
)
2010T base case
2010X system
Rati ng : 5700 MW
NAVAJO 500-CRYSTAL 500
MOENKOPI 500-ELDORDO 500
LIBERTY 345-PEACOCK 345
PALOVRDE500-DEVERS 500
HASSYAMP500-N.GILA 500
RNDVLYTP 230-PEACOCK 230
EAGLEYE 230-PARKER 230
HARCUVAR230-PARKER 230
0
1000
2000
3000
4000
5000
6000
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
F
l
o
w

(
M
W
)
2010T base case
2010X system
Rati ng : 5700 MW
NAVAJO 500-CRYSTAL 500
MOENKOPI 500-ELDORDO 500
LIBERTY 345-PEACOCK 345
PALOVRDE500-DEVERS 500
HASSYAMP500-N.GILA 500
RNDVLYTP 230-PEACOCK 230
EAGLEYE 230-PARKER 230
HARCUVAR230-PARKER 230
Figure 118. Transmission Flow Duration Curves Path 21: Arizona to California
124
0
2000
4000
6000
8000
10000
12000
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
F
l
o
w

(
M
W
)
2010T base case
2010X system
Figure 119. Transmission Flow Duration Curve Path 46: West of Colorado River.
0
2000
4000
6000
8000
10000
12000
14000
16000
0 1000 2000 3000 4000 5000 6000 7000 8000
Hours
F
l
o
w

(
M
W
)
2010T base case
2010X system
Figure 120. Transmission Flow Duration Curves Total SCIT (Southern California
Import Transmission).
Figure121showstheflowdurationcurvesfortheCaliforniaOregonInterfaceforthebasecase
andwithouttheadditionofthenewintermittentgeneration.Althoughthetotalshiftwassmall
therewasconcernthatthevariabilityoftheintermittentgenerationmightintroducesignificant
swingsinthepowerflowsonthisinterface.
125
-1000
0
1000
2000
3000
4000
5000
6000
0 2000 4000 6000 8000
Hours
P
a
t
h

6
6
:

C
O
I

F
l
o
w
s

(
M
W
)
2010T base case
Figure 121. Annual Duration Curve Path 66: COI (2010T).

Figure122showsthechronologicalflowsontheinterfaceforthefirstweekinMay.Although
theimportsintoCaliforniaarereducedwhentheintermittentgenerationisaddedtheredoesnt
seemtobeasignificantchangeintheoverallpattern.
0
1000
2000
3000
4000
5000
6000
0 24 48 72 96 120 144 168
Hours
C
O
I

F
l
o
w

(
M
W
)
2010T base case
Figure 122. Path 66: COI Flows for One Week in May (2010T).
ThehourlychangeinflowswascalculatedforthetwocurvesinFigure122fortheentireyear
andthevaluesweresortedfromhightolowanddisplayedinFigure123.Fromthesecurvesit
canbeseenthatthereisnoappreciabledifferenceinthevariabilityinthelineflowsduetothe
introductionoftheintermittentgeneration.
126
-4000
-3000
-2000
-1000
0
1000
2000
3000
4000
0 2000 4000 6000 8000
Hours
H
o
u
r
l
y

F
l
o
w

C
h
a
n
g
e
s

(
M
W
)
2010T base case

Figure 123. Annual Duration Curve for Hourly Flow Changes on Path 66: COI (2010T).
4.7. Observations
TheintroductionofanynewgenerationintoCaliforniawillimpactthecostofenergyas
reflectedinthespotpricesnotonlywithinCaliforniabutthroughoutWECC.Thisis
particularlytrueofrenewablegeneratorsthattendtoenterthemarketaspricetakers(even
thoughthepricepaidfortheenergyisnotzero).Theimpactsweregenerallyrathermoderate,
althoughtheycouldbemoresevereinsomehours,withthesignificantdropinspotprices
indicatingthatthesystemisrunninglowongenerationthatcanbemaneuvereddownward.
The2010Tscenariohadabout250hoursintheyearwiththismoresevereimpact.The33%
penetrationofrenewablesinthe2010Xscenarioincreasedthistoabout500hours.
Table27showstheaveragereductionsinoperatingcostsforallofWECCperMWhofthe
renewablegeneration.Thedecreasedvalueofthewindandsolargenerationisduetotheir
intermittentnature.
Table 27. Average Variable Operating Cost Reductions
per MWh of Renewable Energy (2010T).
WECC
($/MWh)
Biomass & Geothermal 39.48
Wind & Solar 29.90
Table28showstheaverageloadpaymentreductionsinbothCaliforniaandallofWECCper
MWhoftherenewablegeneration.
127
Table 28. Average Load Payment Reductions per MWh
of Renewable Energy (2010T).
California
($/MWh)
WECC
($/MWh)
Biomass & Geothermal 8.97 19.66
Wind & Solar 11.41 28.86
Thenonrenewablegeneratorshavetheirrevenuereducedfromtwosides.Thesystemspot
pricesingeneralhavedecreasedsothattheyreceivelessrevenueforeachMWhofenergy
producedandinadditiontheamountofenergythattheyareproducinghasbeenreduceddue
todisplacementbytherenewablegeneration.Table29showstheimpactinCaliforniaandallof
WECCforeachtypeofrenewablegeneration.
Table 29. Average Non-Renewable Generator Revenue Reduction
per MWh of Renewable Energy (2010T).
California
($/MWh)
WECC
($/MWh)
Biomass & Geothermal 23.98 69.17
Wind & Solar 17.78 47.42
Itisofcriticalimportancethattheintermittentrenewablegenerationbecentrallyforecastedand
publishedinthedayaheadmarketalongwiththeloadforecastinorderthatthemarketcan
properlyrespondtotheintermittentgeneration.Theimpactofignoringtheintermittentsinthe
commitmentprocesswasanorderofmagnitudegreaterthantheimpactofthedisplacement
whentheywereproperlyaccountedfor,assummarizedinTable30.
Table 30. Annual Load Payment Reductions from Intermittent Generation
(2010T).
California
($ Millions)
WECC
($ Millions)
Wind & Solar with an estimated
forecast
240 607
Ignoring Wind & Solar forecast in
the commitment process
2231 6330
TheadditionofrenewablegenerationreducedthetotalemissionsthroughoutWECC.Table31
showstheaverageemissionreductionsinWECCperMWhofrenewablegeneration.
Table 31. Average Annual Emission Reductions in WECC per MWh
of Renewable Generation (2010T).
NOx
(lbs/MWh)
SOx
(lbs/MWh)
CO2
(lbs/MWh)
Biomass & Geothermal 130 10 830
Wind & Solar 117 46 810
IngeneralitwasshownthattheenergydisplacementbothinsideCaliforniaandthroughout
WECCwaspredominantlyonthecombinedcyclegeneration.Roughly40%ofthedisplaced
128
energyfellonunitswithinCaliforniaand60%displacedimportscomingfromtherestof
WECC.ThehydrogenerationwithinCaliforniawasadjustedbasedonthedayaheadforecast
fortheintermittentgeneration.Inthe2010Tscenario,over7500MWofWindgenerationonly
introducedshiftsofroughly+/1500MWinthehydrooperation.Somehydroschedulesoutside
ofCaliforniawereaffectedbutthebulkofthemodificationsoccurredinstate.Historical
operationindicatedthattheseshiftsshouldbewellwithinthecapabilityofthesystemmostof
thetime.Pumpedstoragehydrooperationincreasedasadditionalrenewablegenerationwas
addedbuttheredidntappeartobeaneedforadditionalPSHcapacity.Theexceptiontothis
mightoccuriftheconventionalhydroisseverelyconstrained.Inthatcase,thePSHactivity
increasedsignificantly.
Overall,withproperapplicationofexistingtechnology,usingtheexistingandplanned
infrastructureandwithappropriatepolicychangestoencourageflexibilityinoperation,there
wasnotanysignificantoperationalproblemsidentifiedatthehourlyleveltotheintroductionof
renewableenergypenetrationsofupto33%.
129
5.0 Quasi-Steady State Analysis
Thestatisticalandproductionsimulationanalysesprovidedabroadviewoftheimpactof
significantintermittent(windandsolar)generationonsystemperformance.Theycovereda
rangeoftimescales,fromminutetominutetohourlyaswellasdaily,seasonalandannual.In
contrast,thequasisteadystate(QSS)analysisconsistsofdetailedtimesimulationsofspecific3
hourperiods.Itisdesignedtoillustratekeyaspectsofsystemperformanceandpotential
mitigationmeasureswithinthebroadercontextofthestatisticalandproductionsimulation
analyses.Assuch,theQSSanalysisistightlylinkedtobothoftheseanalyses.Forexample,the
QSSstudyscenarioswereselectedonthebasisofthestatisticalanalysis,andtheQSSboundary
conditionsweresetbytheproductionsimulationanalysis.
Thedata,methods,assumptions,studyscenariosandresultsfortheQSSanalysisaredescribed
inthefollowingsubsections.AllQSSanalysiswasperformedusingGEsPSLF(Positive
SequenceLoadFlow)softwarepackage.
5.1. Overview of Method
TheprimaryobjectivesoftheQSSanalysisweretoevaluatetheimpactofsignificant
intermittentgenerationonloadfollowing(5minutetimescale)andregulation(1minutetime
scale)requirementswithinCalifornia.
Thiswasaccomplishedbyperformingtimesimulationsconsistingofaseriesofpowerflow
solutionstosimulateCaliforniasystemperformanceonaminutebyminutebasisoverselected
3hourintervals.Eachpowerflowintheseriesrepresentedsystemconditionsataparticular
minuteofthesimulation.AllCalifornialoadsvariedfromminutetominute,allCaliforniawind
projectpoweroutputsvariedfromminutetominute,andallCaliforniasolarprojectoutputs
variedfromminutetominute.AnypowernecessarytobalancetotalCaliforniagenerationand
loadinthe1minutetimeframewasprovidedbyaproxyAGCunit.Thepoweroutputofthis
proxyunitapproximatedtheamountofregulationrequiredofallunitsonAGCbetween5
minuteredispatchesofthesystem.
At5minuteintervals,asimplifiedeconomicdispatchwasperformedtomeetthefollowing
objectives:
UpdatethehouraheadscheduleforeachdesignatedCalifornialoadfollowingunit
basedonaperfectloadandsolargenerationforecast,andapersistencebasedwind
generationforecast(i.e.,thenexthourwillbethesameasthecurrenthour)
RedistributethepowerfromtheproxyAGCunitontothedesignatedloadfollowing
units
AsubsetofallCaliforniaunitsparticipatedinthisloadfollowing.Rampratelimits
(MW/minute)andabsolutepowerlimits(maximumandminimumMW)wererespectedon
eachunit.Theidentificationoftheseunits,andhowtheyshareloadfollowingduty,is
discussedinSection5.1.3.
130
Theresultsofeach3hourQSSsimulationincludedavarietyofperformancemetrics,aswellas
totalCaliforniaload,totalCaliforniawindgeneration,totalCaliforniasolargeneration,total
poweroutputofallunitscontrolledbytheeconomicdispatch,andselectedinterfaceflows.
5.1.1. Study Periods
TheQSSstudyperiodswereselectedbaseduponthehourlystatisticalanalysis.Theprimary
objectivewastoidentifychallenging,butcredible,systemconditions.Therefore,thehourly
statisticalanalysiswasusedtoidentifystudyperiodswithanyofthefollowingcharacteristics:
Large1hourand3hourchangesinsystemload
Large1hourand3hourchangesinwindandsolargeneration
Highlevelsofwindandsolargeneration
Highlevelsofwindandsolarpenetration
Lowloadlevels
ThreeprimarystudyperiodswerechosenaJulymorningwithbothaloadincreaseandanet
decreaseinwindandsolargeneration,aMaynightwithbothalowloadlevelandhighwind
penetration,andaJuneeveningwithbothaloaddecreaseandasignificantincreaseinwind
generation.AdditionalcharacteristicsofthesestudyperiodsareshowninTable32.
Table 32. Characteristics of QSS Study Periods.
July Morning
Load Increase
May Night
Light Load
June Evening
Load Decrease
Renewable Scenario 2010T
(1)
2010X
(2)
2010X
(2)

Load Year 2004 2003 2004
Time Period
(3)
5:00 8:00 am 1:00 4:00 am 4:00 7:00 pm
Initial Total Load 34,300 MW 25,100 MW 41,900 MW
Initial Total Wind 3,400 MW 10,200 MW 2,200 MW
Initial Total Solar 400 MW 0 MW 1,350 MW
Initial Wind and Solar Penetration
(4)
11% 41% 8%
Total Load Change 8,200 MW -700 MW -2,700 MW
Total Wind Change -2,100 MW -600 MW 4,400 MW
Total Solar Change 450 MW 0 MW -900 MW
Total Net Change (L-W-S) 9,850 MW -100 MW -6,200 MW
Final Total Load 42,500 MW 24,400 MW 39,100 MW
Final Total Wind 1,300 MW 9,600 MW 6,600 MW
Final Total Solar 850 MW 0 MW 450 MW
Final Wind and Solar Penetration
(4)
5% 39% 18%
Notes: (1)2010Tscenarioincludesabout7,500MWofwindcapacityand1,900MWofsolarcapacity.
(2)2010Xscenarioincludesabout12,500MWofwindcapacityand2,600MWofsolarcapacity.
(3)AlltimesareinPacificStandardTime.
(4)Instantaneouspenetrationwasdefinedastotalwindandsolargeneration(MW)dividedby
totalload(MW).
131
Systemperformanceduringeachofthesestudyperiodswillbeillustratedinsubsequent
sections.Theimpactofvariousmitigationmeasureswillalsobeaddressed.
5.1.2. Input Data
SeveraltypesofdatawereusedintheQSSanalysis:powerflowdatabases,individualwind
projectoutputprofiles,individualsolarprojectoutputprofiles,atotalCalifornialoadprofile,
andindividualunitramprate(MW/minute)capabilities.
DPCprovidedthepowerflowsrepresentingtherenewablegenerationscenariosfor2010Tand
2010X.TheinitialQSSsimulationoftheJulymorningstudyperiodusedthe2010Trenewable
scenario,withabout7,500MWofwindgenerationcapacityand1,900MWofsolargeneration
capacity.AllotherQSSsimulationsusedthe2010Xrenewablescenario,withabout12,500MW
ofwindcapacityand2,600MWofsolarcapacity.
AWSTruewindprovided1minuteoutputprofilesforeachQSSstudyperiod.Theseprofiles
coveredallexistingandnewwindfarmsidentifiedinthe2010Trenewablescenario,aswellasa
varietyofadditionalprofilesforfuture(e.g.,2020)projects.Ingeneral,asingleprofilewas
appliedtoeachwindfarmrepresentedinthe2010Tpowerflow.However,multipleprofiles
wereappliedtoanylarge(e.g.,500to1,000MW)windprojectsrepresentedassingleequivalent
unitsinthepowerflow.Forexample,asingleequivalent500MWwindfarmprojectatthe
Tehachapi500kVbuswasrepresentedintheQSSanalysisasfiveindividualfarmsrated
100MW,100MW,102MW,103MW,and112MW.
The2010Xrenewablescenarioacceleratedmany2020windprojectsintothe2010timeframe.
AWSTruewindprofileswereassignedtothese2020projectsonthebasisofwindregion.Thus,
a2020windprojectinwindregion9wasassignedawindprofilefromthatregion.Theprofiles
werescaled,asneeded,tobettermatchtheratingoflargewindfarms.Forexample,a500MW
windprojectinthe2010XrenewablescenariowasrepresentedintheQSSanalysisbyfour
individualfarmsrated110MW,110MW,125MWand155MWandassignedAWSTruewind
profilesscaledfromtheoriginal100MW,100MW,114MWand119MWratings.
Limitedhighresolutionsolardatawasavailable.Therefore,thenecessary1minutesolar
profileswerecreatedfromavarietyofdatasources.ProfilesforthePVsiteswerecreatedby
superimposing1minuteGolden,COirradiationdataon15minutePVdatafor13California
zipcodes.AdiscussionoftheapplicabilityofoutofstatesolardatatoCaliforniaPVsitesis
providedinAppendixC.ProfilesforthelargeStirlingsolarfacilitieswerecreatedby
superimposing3minuteDesertRock,NVirradiationdataonhourlyStirlingplantdata.Profiles
forallotherconcentratingsolarplantswerecreatedbyscalingthe1minuteprofilesofthe
existingSungenandLuzprojects.
CaliforniaISOprovidedtotalloaddataforthestudyperiodswith4secondresolution.This
datawassampledat1minuteintervalstocreatethenecessaryprofilesfortheQSSanalysis.In
thatanalysis,anychangeintotalloadfromoneminutetothenextwasspreadacrossall
Californialoads,proportionaltothesizeofanindividualload.
132
CaliforniaISOalsoprovidedconfidentialrampratedata(MW/minute)forindividual
generatingunits.Aweightedaveragerampratewascalculatedforeachtypeofunit(e.g.,
Steam,Hydroelectric)includedinthedata.Thisgenericramprate,showninTable33,wasthen
usedintheQSSanalysis.
Table 33. Weighted Average Ramp Rate Data by Unit Type.
%/Minute
Combined Cycle 3.8
Combustion Turbine 13.5
Hydroelectric 22.3
Steam 3.1
Finally,CaliforniaISOprovideddatashowingtheamountofregulationprocuredonanhourly
basis.TheminimumprocuredduringalloftheQSSstudyperiodswasabout320MWofup
regulationand320MWofdownregulation.Thiswasaconservativeassumption,asshownby
theplotsofprocuredregulationdatainSection6.0.TheresultsoftheQSSanalysisassociated
withregulationwerecomparedagainstthesethresholds.
Manyoftheminimumgeneratingunitoutputsasdefinedbythepowerflowswerezero.This
doesnotreflecttherealitythatgeneratingunitshaveminimumloadlevelsthatmustbe
respected.TheminimumpoweroutputassumptionsforCaliforniagenerationusedinthis
study,byunittype,areshowninTable34.Theminimumoutputofcombustionturbineswas
leftatzerotoemulatetheirfaststart/stopcharacteristics.
Table 34. Minimum Generation Output Level by Unit Type.
% of Rating
Combined Cycle 50
Combustion Turbine 0
Hydroelectric 20
Steam 25
5.1.3. Boundary Conditions

Theresultsoftheproductionsimulationanalysiswereusedtodefinetheboundaryconditions
(i.e.,initialandfinalsystemstates)foreachQSSstudyperiod.Thepowerflowsprovidedby
DPCweremodifiedtorepresentthedesiredinitialandfinalconditionsasspecifiedbythe
productionsimulationanalysis.Theseconditionsincludedthedispatchofallgeneratingunits
inWECC,individualarealoadlevels,HVDCtieflows,andinterareaACtieflows.
AcomparisonoftheCaliforniagenerationdispatchbetweentheinitialandfinalpowerflows
identifiedthoseunitswhoseoutputchangedoverthestudyperiod.Asubsetoftheseunits(e.g.,
thosewitha3hourchangeindispatchof>25MW)wereusedintheQSSeconomicdispatch
model.Therefore,fewerunitswereassignedtoloadfollowingintheQSSsimulationsthanin
theproductionsimulationanalysis.Asaresult,theQSSsimulationswereconservative
comparedtotheexpectedcapabilityoftheCaliforniasystem.
133
Oncetheeconomicdispatchunitswereidentified,eachwasassignedaparticipationfactorfor
theQSSanalysis.Theparticipationfactorallottedafractionoftheeconomicdispatch
requirements(MW)toeachidentifiedunit.Theallottedfractionwasproportionaltothechange
ingenerationdispatchobservedontheunitintheproductionsimulationresults,comparedto
thetotalchangeingenerationdispatchoverthe3hourstudyperiod.Asanequation,the
participationfactorcanbedefinedasfollows:
PF=MWi/MWtotal
where:MWi=MWchangeoni
th
unitover3hourperiod
MWtotal=totalMWchangeonalleconomicdispatchunitsover3hourperiod
ThecomparisonofinitialandfinalpowerflowsalsoidentifiedchangesinHVDCtieflowsand
thelevelofCaliforniaimports.IntheQSSanalysis,HVDCtieflowswererampedfromthe
initialconditiontothefinalcondition.However,the3hourCaliforniaimportincreaseswere
ignoredtoensurethatallnecessaryinstateloadfollowingandregulationwasperformedby
Californiaunits.ThegreaterWECCsystemwas,therefore,largelyunchangedduringaQSS
simulation.
5.2. System Performance Examples
TheresultsoftheQSSanalysisarediscussedinthissection.TheanalysisoftheJulymorning
studyperiodispresentedfirst.ThatdiscussionwillfocusonillustratingtheQSSanalysis
procedureanddefiningtheresultingperformancemetrics.TheanalysesoftheMaynightand
Juneeveningstudyperiodswillfollow.Thosesectionswillfocusonillustratingkeyaspectsof
systemperformanceandpotentialmitigationmeasures.
5.2.1. July Morning Load Increase
TheJulymorningstudyperiodwasevaluatedunderthe2010Trenewablescenario,which
includesabout7,500MWofwindgenerationcapacityat98sitesand1,900MWofsolar
generationcapacityat12concentratingsolarplantsand136PVsites.Theloadwasderived
fromthe2004data.
AsshowninTable32,thisstudyperiodincludedalargeincreaseinload(8,200MW)combined
withanetdecreaseinwindandsolargeneration(1,650MW).Thisresultsinanetloadincrease
ofabout9,850MW.Ascontext,notethatthestatisticalanalysisofthreeyearsofhourlydata
showedthelargest3hourloadincreasewas10,900MW,andabout1%ofthe3hourload
increasesweregreaterthan8,200MW.Themaximum3hourdropinwindgenerationover
thosethreeyearsofdatawasabout2,700MW,anda3hourwinddecreasegreaterthan
2,100MWoccurred17times.Similarly,themaximum3hournetload(i.e.,loadminuswind
minussolar)increasewas11,300MW.A3hournetloadincreasegreaterthan9,850MW
occurred20times,anda3hourloadonlyincreasegreaterthan9,850MWoccurred6times.
ThetotalloadprofilefortheJulymorningstudyperiodisshowninFigure124.Thesingle
largest1minutechangeinloadisapproximately300MWandoccursjustpast7:30am.This
mayrepresentapumpedstoragehydropumpstopping.Thetotalwindgenerationprofilefor
134
thisstudyperiodisshowninFigure125,andthetotalsolargenerationprofileisshownin
Figure126.ThesethreefiguresprovideanoverviewoftheinputtotheQSSsimulation.Forload
alone,theloadfollowingrequirementwasabout2,700MW/hour(i.e.,8,200MW/3hours).With
thenetdecreaseinwindandsolargeneration,theloadfollowingrequirementincreasedto
about3,300MW/hour(i.e.,9,850MW/3hours).
34000
36000
38000
40000
42000
44000
5:00 AM 6:00 AM 7:00 AM 8:00 AM
Figure 124. Total California Load During the July Morning QSS Study Period.
M
W
135
1000
2000
3000
4000
5:00 AM 6:00 AM 7:00 AM 8:00 AM
Figure 125. Total California Wind Generation During the July Morning QSS Study Period.
0
200
400
600
800
1000
5:00 AM 6:00 AM 7:00 AM 8:00 AM

Figure 126. Total California Solar Generation During the July Morning QSS Study Period.
M
W
M
W
136
TheoutputvariablesoftheQSSanalysisweredividedintotwocategories.Onegroupwas
associatedwithmaneuverability,andmeasurestheresponseofthebalanceofportfolio(i.e.,
nonrenewable)generatorstothechangingload,windandsolarconditions.Theothergroup
wasassociatedwithperformance,andmeasuresnetload(e.g.,loadminuswindminussolar)
variability,aswellastheresultingregulationandloadfollowingrequirements.AllCalifornia
load,windandsolargenerationarerolledintothesemetrics.However,onlythesubsetofall
otherCaliforniageneratingunitsthatparticipatesintheQSSeconomicdispatchisincludedin
themetrics.TheseunitswillbecalledeconomicdispatchorEDunitsthroughouttheremainder
ofthissection.
Specifically,theQSSmaneuverabilityvariablesaredefinedasfollows:
PUPtot=RemainingrangeofEDunitstoincreaseoutput=(PimaxPi)
PDNtot=RemainingrangeofEDunitstodecreaseoutput=(PiminPi)
Pmntot=MinimumpowerofEDunits=Pimin
Pmxtot=MaximumpowerofEDunits=Pimax
SumEDMW=OutputofEDunits=Pi
RUPtot=RemainingratecapabilityofEDunitstorampup=(Ratei,ifPi<Pimax)
RDNtot=RemainingratecapabilityofEDunitstorampdown=(Ratei,ifPi>
Pimin)
where
Pi=Realpoweroutputofthei
th
EDunit(MW)
Pimax=Maximumrealpoweroutputofthei
th
EDunit(MW)
Pimin=Minimumrealpoweroutputofthei
th
EDunit(MW)
Ratei=Weightedaverageramprateofthei
th
EDunit(MW/minute)
Thefirstfivemaneuverabilityvariables(PUPtot,PDNtot,Pmntot,Pmxtot,SumEDMW)arein
unitsofMW.Thelasttwovariables(RUPtot,RDNtot)areinunitsofMW/minute.
ThemaneuverabilityresultsfortheJulystudyperiodareshowninFigure127.Thegreenline
representsPmxtot,whichisaconstant15,000MWduringtheQSSsimulation.Similarly,the
lavenderlinerepresentsPmntot,whichiszerothroughout.TheblacklineshowsSumEDMW,
orthetotaloutputoftheEDunits.Asthenetloadincreasesoverthis3hourperiod,theoutput
oftheEDunitsalsoincreases.Theremaininguprangeavailableonthoseunits(PUPtot)is
representedbytheredline.Notethatitdecreasesastheoutputoftheunitsincreases.
Conversely,theremainingdownrange(PDNtot),representedbytheturquoiseline,increasesas
thetotaloutputincreases.Thefinaltwotracesrepresenttheup(RUPtot,yellowline)anddown
(RDNtot,pinkline)rampratecapabilityoftheEDunits.Aswiththeuprange,theupramprate
capabilityisreducedastheoutputincreases.Aswiththedownrange,thedownramprate
capabilityisincreasedastheoutputincreases.
137
Thesignconventionissuchthatuprangeorrampcapabilityisshownasapositivenumber,
anddownrangeorrampcapabilityisshownasanegativenumber.
-15000
-10000
-5000
0
5000
10000
15000
5:00 AM 6:00 AM 7:00 AM 8:00 AM
PUPtot
PDNtot
RUPtot
RDNtot
Pmntot
Pmxtot
Sum ED MW
MW
MW
MW/Min
MW
MW
MW/Min
Figure 127. Maneuverability Variables During the July Morning QSS Study Period.
TheQSSperformancevariablesaredefinedasfollows:
DeltaLWS=Changeintotalloadminuswindgenerationminussolargeneration
DeltaEDMW=TotalchangeinEDunitoutputrequiredtofollowload
Preg=Totalregulatingpowernecessarytobalanceloadandgeneration
Thefirsttwoperformancevariables(DeltaLWS,DeltaEDMW)areinunitsofMW/minute.
Theothervariable(Preg)isinunitsofMW.
TheperformanceresultsfortheJulystudyperiodareshowninFigure128.Thebluetraceshows
theminutetominutechangeinnetload(DeltaLWS).Thegreentraceshowstheamountof
powerrequiredona1minutebasistobalancegenerationandload(Preg).Theyellowlines
(RegUp,RegDown)showtheminimumregulationprocuredduringthisstudyperiod.Note
thatthelargestchangeinnetload,aswellasthelargestneedforregulation,coincideswiththe
300MWdecreaseinloadobservedjustafter7:30aminFigure124.Asecondlargeregulation
requirementoccursjustafter7:00am.ThisisassociatedwithapersistentincreaseinDelta
LWS,drivenbyashorttermincreaseintheslopeoftheloadprofile.
138
ThepinklineshowsthetotalchangeinEDunitoutput(DeltaEDMW)requiredtofollowload
andaccommodatetherequiredminuteminuteregulation.Sincetheeconomicdispatchis
performedevery5minutes,theDeltaEDMWsignalisconstantforeachsuccessive5minute
period.Initially,thetotalincreaseinEDunitoutputislessthan10MW/minute.Itisalmost
alwayspositiveduringthismorningloadrise,andpeaksatabout135MW/minutejustafter
7am.
-400
-300
-200
-100
0
100
200
300
400
5:00 AM 6:00 AM 7:00 AM 8:00 AM
Delta L-W-S
Delta ED MW
Preg
Reg Up
Reg Down
MW
MW
MW/Min
MW
MW
MW/Min
Figure 128. QSS Performance Variables During the July Morning QSS Study Period.
Aspreviouslynoted,theproductionsimulationanalysisdefinedtheboundaryconditionsfor
theQSSanalysis.Assuch,theEDunitswereidentifiedonaneconomicbasis.Thetotalchange
inEDunitoutput,bytype,isshowninFigure129fortheJulystudyperiod.Bydefinition,each
ofthesetracesstartsatzero.ThebluelinerepresentsthetotalchangeinallhydroelectricED
unitoutput,andthepinklinerepresentsthetotalchangeinallcombinedcycleEDunitoutput.
Thiscaseshowsthatmostoftheloadfollowingisperformedbycombinedcycleplants,
followedbyhydroelectricunits.
139
0
2000
4000
6000
8000
5:00 AM 6:00 AM 7:00 AM 8:00 AM
Delta GT
Delta Hydro
Delta CC
Delta Steam
Figure 129. Economic Dispatch Unit Change During the July Morning QSS Study Period.
Anadditionalsensitivitycasewasperformedtoevaluatetheimpactofwindandsolar
variabilityonsystemperformance.ForthisQSSsimulation,allwindandsolarplantswereheld
constantattheirinitialvaluesduringthe3hourstudyperiod.Constantwindandsolar
generationisequivalenttoanynondispatchablegeneration(e.g.,biomass,nuclear).
CrossplotsofthesensitivitycaseresultswiththeprimaryresultsareshowninFigure130and
Figure131.Inbothfigures,thebluelinerepresentsthesensitivitycasewithconstant
intermittentrenewableoutput,andtheyellowlinerepresentstheprimarycasewithvariable
windandsolaroutput.
Figure130showstheimpactofintermittentrenewablevariabilityonthedutyimposedonthe
proxyAGCunit.Theseresultsindicatethattheregulationdutyisdueprimarilytoload
variation,butisoffsetupbythenetdecreaseinwindandsolarpowerover3hours.
Figure131showstheimpactofintermittentrenewablevariabilityontheeconomicdispatchand
loadfollowingrequirements.Theseresultsshowthattheloadfollowingdutyisincreasedfrom
anaverageofabout50MW/minutetoabout60MW/minuteduetothenetdecreaseinwindand
solarpower.
Ingeneral,theJulymorningQSSanalysisshowsthataneconomicallyrationalunitcommitment
anddispatchprovidesadequateloadfollowingcapability.
M
W
140
-600
-400
-200
0
200
400
600
5:00 AM 6:00 AM 7:00 AM 8:00 AM
Preg Constant W&S
Preg Variable W&S
Reg Up
Reg Down
Figure 130. Impact of Intermittent Variability on Regulation Duty During July QSS Study Period.
-50
0
50
100
150
5:00 AM 6:00 AM 7:00 AM 8:00 AM
Delta ED MW Constant W&S
Delta ED MW Variable W&S
Figure 131. Impact of Intermittent Variability on Economic Dispatch and Load Following Duty
During July QSS Study Period.
M
W
M
W
/
M
i
n
u
t
e
141
5.2.2. May Night Low Load Level
TheMaynightstudyperiodwasevaluatedunderthe2010Xrenewablescenario,whichincludes
about12,500MWofwindgenerationcapacityat142sitesand2,600MWofsolargeneration
capacityat42concentratingsolarplantsand128PVsites.Theloadwasderivedfromthe2003
data.AsdescribedinSection5.1.3,theunitsavailableintheQSSanalysisareasubsetofboth
thetotalnumberofunitsinCaliforniaaswellastheunitsavailableintheproductionanalysis.
Thus,theanalysiswasconservative.
AsshowninTable32,thisstudyperiodrepresentedarelativelylightloadcondition
(25,100MW)withahighlevelofinstantaneouswindpenetration(10,200MWor41%).Neither
theloadnorthewindchangedsignificantlyoverthe3hourstudyperiod.Ascontext,notethat
thestatisticalanalysisofthreeyearsofhourlydatashowedthepeakwindgenerationoutput
was11,500MW,andwindgenerationwasgreaterthan10,000MWforlessthan1%ofthe
hours.Theaveragehourlyintermittentpenetrationatlightloadwasabout29%,andthepeak
hourlyintermittentpenetrationwasabout39%.
ThetotalloadprofilefortheMaynightstudyperiodisshowninFigure132.Severalloadsteps
greaterthan+/200MWareobserved.Thesearelikelyduetoswitchingpumpedstoragehydro
facilitypumps,orotherlargeloads.Alatersensitivitycasewillinvestigatesystemresponse
withouttheseloadsteps,sincetheymaybepredictableand/orscheduled.Thetotalwind
generationprofileforthisstudyperiodisshowninFigure133.Thereisnosolargeneration
profileatnight.ThesetwofiguresprovideanoverviewoftheinputtotheQSSsimulation.
23000
23250
23500
23750
24000
24250
24500
24750
25000
25250
25500
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Figure 132. Total California Load During the May Night QSS Study Period.
M
W
142
9500
9600
9700
9800
9900
10000
10100
10200
10300
10400
10500
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Figure 133. Total California Wind Generation During the May Night QSS Study Period.
TheQSSmaneuverabilityvariables(definedinSection5.2.1)fortheMaystudyperiodare
showninFigure134.ThegreenlinerepresentsPmxtot,whichisaconstant5,500MWduring
theQSSsimulation.Similarly,thelavenderlinerepresentsPmntot,whichisconstantatabout
1,100MW.Theuprampratecapability(RUPtot)oftheEDunitsisrepresentedbytheyellow
line,andthedownrampratecapability(RDNtot)isrepresentedbythepinkline.Similarly,the
remaininguprangeavailableontheEDunits(PUPtot)isrepresentedbytheredline,andthe
remainingdownrange(PDNtot)isrepresentedbytheturquoiseline.Theblacklineshowsthe
totalactualoutputoftheEDunits(SumEDMW).Asthenetloaddecreasesoverthis3hour
period,theoutputoftheEDunitsalsodecreases.Atabout1:30am,theEDunitshaverunoutof
downrampratecapability.Atabout1:45am,theEDunitshaverunoutofdownrange.Thus,
alldownmaneuverabilityhasbeenexhausted.Afterabout3:15am,thenetloadhasincreased
suchthatsomedownmaneuverabilityisrecovered.
TheQSSperformanceresults(alsodefinedinSection5.2.1)fortheMaystudyperiodareshown
inFigure135.Thebluetraceshowstheminutetominutechangeinnetload(DeltaLWS).The
greentraceshowstheamountofpowerrequiredona1minutebasistobalancegenerationand
load(Preg).Theyellowlines(RegUp,RegDown)showtheminimumregulationprocured
duringthisstudyperiod.
Twoobservationscanbemade.First,thelargestepsinloadhavemoresignificantimpactonthe
regulationduty(Preg)thanthevariabilityofloadandwind.Thisisobservedinthelarge
changesinPregatabout1:30am,2:30amand3:15am.
M
W
143
Second,insufficientdowncapability(bothrangeandramprate)shiftsloadfollowingdutyto
regulation,whichmaythenbecomeexhausted.ThisinturnmayresultinaviolationofCPS2
criteria.Theshiftisobservedbetween1:30amand2:30amwiththepersistentlyhighlevelof
negativeoutputfromtheAGCproxyunitwhiletheloadfollowingcapability(DeltaEDMW)is
zero.
-2000
-1000
0
1000
2000
3000
4000
5000
6000
1:00 AM 2:00 AM 3:00 AM 4:00 AM
PUPtot
PDNtot
RUPtot
RDNtot
Pmntot
Pmxtot
Sum ED MW
MW
MW
MW/Min
MW
MW
MW/Min
Figure 134. Maneuverability Variables During the May Night QSS Study Period.
144
-600
-400
-200
0
200
400
600
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Delta L-W-S
Delta ED MW
Preg
Reg Up
Reg Down
MW
MW
MW/Min
MW
MW
MW/Min
Figure 135. Performance Variables During the May Night QSS Study Period.
Increase Maneuverable Generation
Asensitivitycasewasperformedtoevaluatetheimpactofincreasingthemaneuvering
capabilityofthebalanceofportfoliogenerationbychangingthegenerationcommitment.This
wasaccomplishedbyreplacing2,200MWofbaseloadgenerationwith2,200MWofcombined
cyclegeneration.
TheQSSmaneuverabilityvariablesforthissensitivitycaseareshowninFigure136.Thegreen
linerepresentsPmxtot,whichisaconstant8,400MWduringtheQSSsimulation.Similarly,the
lavenderlinerepresentsPmntot,whichisconstantatabout2,600MW.Theupramprate
capability(RUPtot)oftheEDunitsisrepresentedbytheyellowline,andthedownramprate
capability(RDNtot)isrepresentedbythepinkline.Similarly,theremaininguprangeavailable
ontheEDunits(PUPtot)isrepresentedbytheredline,andtheremainingdownrange
(PDNtot)isrepresentedbytheturquoiseline.Theblacklineshowsthetotalactualoutputofthe
EDunits(SumEDMW).Asthenetloaddecreasesoverthis3hourperiod,theoutputoftheED
unitsalsodecreases.Unliketheoriginalcase(Figure134),however,neitherdownrangenor
downrampratecapabilityisexhausted.Therefore,theloadfollowingrequirementsarenot
shiftedtotheAGCproxyunit.
ThisisconfirmedbytheQSSperformanceresultsshowninFigure137.Thebluetraceshowsthe
minutetominutechangeinnetload(DeltaLWS).Thegreentraceshowstheamountofpower
requiredona1minutebasistobalancegenerationandload(Preg).Theyellowlines(RegUp,
RegDown)showtheminimumregulationprocuredduringthisstudyperiod.Unlikethe
145
originalcase(Figure135),theproxyAGCunitoutput(Preg)islargelywithintheprocured
regulationrange.Theexcursionsbeyondthosethresholdsareduetolargestepsinloadatabout
1:30am,2:30am,and3:15am.
Thissensitivitycaseshowsthatreplacingnonmaneuverablegenerationwithmaneuverable
generationinthebalanceofportfolio(i.e.,nonrenewablegeneration)effectivelymitigatesloss
ofbothdownrangeanddownrampratecapabilityunderlightloadconditions.
-4000
-2000
0
2000
4000
6000
8000
10000
1:00 AM 2:00 AM 3:00 AM 4:00 AM
PUPtot
PDNtot
RUPtot
RDNtot
Pmntot
Pmxtot
Sum ED MW
MW
MW
MW/Min
MW
MW
MW/Min
Figure 136. Maneuverability Variables for a May Night with More CombinedCycle Plants.
146
-600
-400
-200
0
200
400
600
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Delta L-W-S
Delta ED MW
Preg
Reg Up
Reg Down
MW
MW
MW/Min
MW
MW
MW/Min
Figure 137. Performance Variables for a May Night with More CombinedCycle Plants.
Anothersensitivitycasewasperformedtoevaluatetheimpactofwindvariabilityonsystem
performance.ThisQSSsimulationagainusedthemodifiedgenerationcommitmentwithmore
maneuverablecombinedcycleplantsavailable.Inaddition,allwindplantswereheldconstant
attheirinitialvaluesduringthe3hourstudyperiod.Aspreviouslynoted,constantwind
generationisequivalenttoanynondispatchablegeneration.
CrossplotsofthissensitivitycaseresultswiththepreviousresultsareshowninFigure138and
Figure139.Inbothfigures,thebluelinerepresentsthesensitivitycasewithconstantwind
output,andtheyellowlinerepresentsthemodifiedcommitmentcasewithvariablewind
output.
Figure138showstheimpactofwindvariabilityonthedutyimposedontheproxyAGCunit.
Theseresultsconfirmthattheregulationdutyisdueprimarilytoloadvariation,butis
increasedbythevariabilityassociatedwith10,000MWofwindgeneration.
Figure139showstheimpactofwindvariabilityontheeconomicdispatchandloadfollowing
requirements.TheseresultsshowthatthevariabilityofwindincreasestheEDunitdutyboth
theMW/minuteandthefrequencyofsignchangesincrease.
147
-600
-400
-200
0
200
400
600
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Preg Modified Commitment, Constant Wind
Preg Modified Commitment, Variable Wind
Reg Up
Reg Down
Figure 138. Impact of Wind Variability on Regulation Duty During May QSS Study Period.
-150
-100
-50
0
50
100
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Delta ED MW Modified Commitment, Constant Wind
Delta ED MW Modified Commitment, Variable Wind
Figure 139. Impact of Wind Variability on Economic Dispatch and Load Following Duty During May
QSS Study Period.
M
W
M
W
/
M
i
n
u
t
e
148
Temporary Curtailment of Wind Generation
Theprevioussectionexploredonemitigationmethodmodifyingthegenerationcommitment
toensuresufficientmaneuverability.Anothermitigationmethod,thetemporarycurtailmentof
windgeneration,isexaminedinthissection.TheoriginalMaywindprofile(blueline)andthe
curtailedwindprofile(yellowline)areshowninFigure140.Forthisillustration,the
curtailmentconsistedofarelativelyfastreductioninwindgenerationofabout500MWover5
minutes.Thisreductionwasappliedatabout1:15ambeforealldownmaneuverabilitywas
exhaustedintheoriginalMaysimulation(Figure134).Totalwindproductionwascurtailedfor
thenexttwohours.Thecurtailmentwasremovedaftersomedownmaneuverabilitywas
recoveredintheoriginalMaysimulationatabout3:30am.Windfarmoutputwasallowedto
increaseoverabout10minutesbackuptothatavailableduetotheprevailingwind.This
temporarycurtailmentwasimplementedatallwindfarms,witheachreducingitsoutputbya
shareofthe500MWproportionaltoitsrating.Thewindenergylostduringthecurtailmentwas
about1,140MWhoutofabout30,000MWh,or3.8%,forthis3hourperiod.Asnotedabove,the
statisticalanalysisofthreeyearsofhourlydatashowedtotalwindgenerationwasgreaterthan
10,000MWforlessthan1%ofthehours.Hence,thisisarareoccurrence.
Crossplotsoftheresultsfromthecurtailedwindsimulationandtheoriginalwindsimulation
areshowninFigure142andFigure141.Inbothfigures,thebluelinerepresentsthecasewith
theoriginalwindprofile,andtheyellowlinerepresentsthecasewiththecurtailedwind.
Figure141showstheimpactofwindvariabilityontheeconomicdispatchandloadfollowing
requirements.Theresultsfromtheoriginalwindprofilesimulationshowthatalldown
maneuverabilityislostbetweenabout1:45amand2:45amwhentheEDunitoutputiszero.
Windcurtailmentensuresthattheremainingnonrenewablegenerationmaintainssufficient
maneuverabilitytofollowload,asshownbythenonzeroEDunitoutputinthatcase.
Figure142showstheimpactofwindvariabilityonthedutyimposedontheproxyAGCunit.
Undertheoriginalwindprofile,thelackofloadfollowingcapabilityhasshiftedthatdutytothe
proxyAGCunit.Thecasewithwindcurtailmenthassufficientloadfollowingcapability,sothe
regulationrequirementsarenowlargelybetweentheprocuredregulationthresholds.As
previouslynoted,theexcursionsareduetolargechangesintheloadprofile.
TheresultsoftheseQSSsimulationsshowthatthecurtailmentofwindeffectivelymitigatesloss
ofmaneuverabilityunderextremelightload,highwindconditions.Curtailmentwouldalso
allowthewindplantstoprovideregulation.
149
8000
9000
10000
11000
12000
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Original Wind
Curtailed Wind
Figure 140. Temporary Curtailment of Wind Generation.

-150
-100
-50
0
50
100
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Delta ED MW Original Wind
Delta ED MW Curtailed Wind
Figure 141. Impact of Temporary Curtailment of Wind Generation on Economic Dispatch and
LoadFollowing Duty During May QSS Study Period.
M
W
M
W
/
M
i
n
u
t
e
150
-600
-400
-200
0
200
400
600
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Preg Original Wind
Preg Curtailed Wind
Reg Up
Reg Down
Figure 142. Impact of Temporary Curtailment of Wind Generation on Regulation Duty During
May QSS Study Period.
Remove Large Steps from Load Profile
TheMaysimulationresultshaveallshownthatlargestepsinloadhaveamoresignificant
impactonregulationthanwindvariability.Asnotedpreviously,theselargeloadstepsare
probablyduetoswitchingpumpedstoragehydrofacilitypumps.Whileschedulable,such
pumpswitchingeventsstillhaveasignificantimpactonsystemperformance.
Nonetheless,asensitivitycasewasdevelopedtoevaluateloadandwindvariabilitywithoutthe
disruptionoflargeloadswitchingevents.Thiswasaccomplishedbyremovingallloadsteps
withanabsolutevaluegreaterthan200MWfromtheoriginalloadprofile,asshowninFigure
143.Inthisfigure,theoriginalloadprofileisrepresentedbythebluelineandthemodifiedload
profileisrepresentedbytheyellowline.
Ascontext,selectedMaynightloadprofilesfromtheCaliforniaISOprovideddataareshown
inFigure144.TheloadprofileforthisMayQSSstudyperiodisrepresentedbytheyellowline.
Whilethisprofilesharesthegeneralshapeofalloftheotherplottedprofiles,itistheonlyone
withsignificantstepdowninloadbetweenmidnightand3am.Allotherprofilesshowpositive
loadstepsinthistimeframe,whichisconsistentwiththepracticeofswitchinginpumped
storagehydrofacilitiesatlightload.Therefore,theMayQSSstudyperiodisoneofthemost
severetestsoftheneedtomaintaindownmaneuverabilityunderlightloadconditions.
M
W
151
24000
24500
25000
25500
26000
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Original Load
Modified Load
Figure 143. Modified Load Profile Without Large Switching Events.

18000
19000
20000
21000
22000
23000
24000
25000
12:00 AM 1:00 AM 2:00 AM 3:00 AM 4:00 AM 5:00 AM 6:00 AM
Figure 144. Selected California ISO 1-Minute Load Data from May 2002, 2003, and 2004.
M
W
M
W
MayQSSProfile
152
TheQSSmaneuverabilityvariables(definedinSection5.2.1)withthemodifiedloadprofileare
showninFigure145.ThegreenlinerepresentsPmxtot,whichisaconstant5,500MWduring
theQSSsimulation.Similarly,thelavenderlinerepresentsPmntot,whichisconstantatabout
1,100MW.Theuprampratecapability(RUPtot)oftheEDunitsisrepresentedbytheyellow
line,andthedownrampratecapability(RDNtot)isrepresentedbythepinkline.Similarly,the
remaininguprangeavailableontheEDunits(PUPtot)isrepresentedbytheredline,andthe
remainingdownrange(PDNtot)isrepresentedbytheturquoiseline.Theblacklineshowsthe
totalactualoutputoftheEDunits(SumEDMW).Withtheoriginalloadprofile,thedown
maneuverabilityoftheEDunitswasexhausted,asshowninFigure134.Withthemodifiedload
profile,thedownmaneuverabilityoftheEDunitswasnotexhausted.
TheQSSperformanceresults(alsodefinedinSection5.2.1)withthemodifiedloadprofileare
showninFigure146.Thebluetraceshowstheminutetominutechangeinnetload(Delta
LWS).Thegreentraceshowstheamountofpowerrequiredona1minutebasistobalance
generationandload(Preg).Theyellowlines(RegUp,RegDown)showtheminimum
regulationprocuredduringthisstudyperiod.Theseresultsindicatethataccommodatingthe
variabilityofboththeloadandsignificantwindgenerationiswithinthecapabilityofthe
system.
-2000
-1000
0
1000
2000
3000
4000
5000
6000
1:00 AM 2:00 AM 3:00 AM 4:00 AM
PUPtot PDNtot
RUPtot RDNtot
Pmntot Pmxtot
Sum ED MW
MW
MW
MW/Min
MW
MW
MW/Min
Figure 145. Maneuverability Variables for a May Night with the Modified Load Profile.
153
-400
-300
-200
-100
0
100
200
300
400
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Delta L-W-S
Delta ED MW
Preg
Reg Up
Reg Down
MW
MW
MW/Min
MW
MW
MW/Min
Figure 146. Performance Variables for a May Night with the Modified Load Profile.
5.2.3. June Evening Load Decrease
TheJuneeveningstudyperiodwasevaluatedunderthe2010Xrenewablescenario,which
includesabout12,500MWofwindgenerationcapacityat142sitesand2,600MWofsolar
generationcapacityat42concentratingsolarplantsand128PVsites.Theloadwasderived
fromthe2004data.
AsshowninTable32,thisstudyperiodincludedadecreaseinload(2,700MW)combinedwith
asignificantincreaseinwindgeneration(4,400MW)andadecreaseinsolargeneration(
900MW).Thisresultsinanetloaddecreaseofabout6,200MW.Ascontext,notethatthethree
yearsofhourlydatashowedthiscasehasthesinglelargest1hourincreaseinwindgeneration
ofabout3,500MW.Thedataalsoshowedthatthemaximum3hourincreaseinwind
generationwasabout4,900MW,anda3hourwindincreasegreaterthan4,400MWoccurred13
times.
ThetotalloadprofilefortheJuneeveningstudyperiodisshowninFigure147.Thisevening
loaddropisnotmonotonic,andincludesasteepdeclineatabout4:15pm.Thetotalwind
generationprofileforthisstudyperiodisshowninFigure148.Thereisarapidandsustained
windincreaseduringthefirsthour,whichislargelymonotonic.Thetotalsolargeneration
profileisshowninFigure149.Thedeclineinsolaroutputtendstocoincidewiththeload
decrease.ThesethreefiguresprovideanoverviewoftheinputtotheQSSsimulation.
154
38000
39000
40000
41000
42000
43000
44000
4:00 PM 5:00 PM 6:00 PM 7:00 PM
Figure 147. Total California Load During the June Evening QSS Study Period.
2000
2500
3000
3500
4000
4500
5000
5500
6000
6500
7000
4:00 PM 5:00 PM 6:00 PM 7:00 PM
Figure 148. Total California Wind Generation During the June Evening QSS Study Period.
M
W
M
W
155
0
200
400
600
800
1000
1200
1400
4:00 PM 5:00 PM 6:00 PM 7:00 PM
Figure 149. Total California Solar Generation During the June Evening QSS Study Period.
TheQSSmaneuverabilityvariables(definedinSection5.2.1)fortheJunestudyperiodare
showninFigure150.ThegreenlinerepresentsPmxtot,whichisaconstant10,800MWuntil
about6:30pmwhenalargeamountofpumpedstoragehydro,whichhadbeengenerating
power,wasswitchedoff.ThelavenderlinerepresentsPmntot,whichisconstantatabout
2,800MWthroughoutthesimulation.Theuprampratecapability(RUPtot)oftheEDunitsis
representedbytheyellowline,andthedownrampratecapability(RDNtot)isrepresentedby
thepinkline.Similarly,theremaininguprangeavailableontheEDunits(PUPtot)is
representedbytheredline,andtheremainingdownrange(PDNtot)isrepresentedbythe
turquoiseline.TheblacklineshowsthetotalactualoutputoftheEDunits(SumEDMW).The
downmaneuverabilityoftheEDunitsisalmostcompletelyexhaustedbytheendofthe
simulation.
TheQSSperformanceresults(alsodefinedinSection5.2.1)fortheJunestudyperiodareshown
inFigure151.Thebluetraceshowstheminutetominutechangeinnetload(DeltaLWS).The
greentraceshowstheamountofpowerrequiredona1minutebasistobalancegenerationand
load(Preg).Theyellowlines(RegUp,RegDown)showtheminimumregulationprocured
duringthisstudyperiod.Theseresultsagainindicatethattheloadstepatabout4:15pmis
largelyresponsiblefortheexcursionbeyondtheprocuredregulationthresholds.However,the
sustainedwindincreaseoverthefirsthourislargelyresponsiblefortheoffsetindicatedbythe
heavyyellowline.DuringthisperiodtheEDunitsarenotprovidingsufficientloadfollowing,
sothatrequirementisshiftedtotheAGCproxyunit(Preg).
M
W
156
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
10000
12000
4:00 PM 5:00 PM 6:00 PM 7:00 PM
PUPtot
PDNtot
RUPtot
RDNtot
Pmntot
Pmxtot
Sum ED MW
MW
MW
MW/Min
Figure 150. Maneuverability Variables During the June Evening QSS Study Period.
-600
-400
-200
0
200
400
600
4:00 PM 5:00 PM 6:00 PM 7:00 PM
Delta L-W-S
Delta ED MW
Preg
Reg Up
Reg Down
MW
MW
MW/Min
MW
MW
MW/Min
Figure 151. Performance Variables During the June Evening QSS Study Period.
157
Incorporate Hourly Wind Forecast
AsnotedinthedescriptionoftheQSSmethod(Section5.1),thesimplifiedeconomicdispatch
modelincludesanhouraheadschedulefunction.Thisfunctionnominallyincludesaperfect
forecastforbothloadandsolargeneration,andapersistencebasedforecastforwind
generation.Apersistenceforecastassumesthatthewindgenerationinthenexthourwillbethe
sameasitisinthecurrenthour.
TheJuneQSSresultsdescribedaboveshowedthatloadfollowingrequirementswereshiftedto
regulationduringperiodswithasustainedincreaseinwindpower.Modifyingtheeconomic
dispatchmodeltoincludeashorttermwindforecastisonepotentialmitigationmethodto
preventthisshift.
CrossplotsoftheQSSsimulationresultswithandwithoutanhouraheadwindforecastare
showninFigure152andFigure153.
Figure152showstheimpactofincorporatinganhouraheadwindforecastonthedutyimposed
ontheproxyAGCunit.Thebluelinerepresentstheoriginaleconomicdispatchmodelwitha
persistencebasedwindforecast,andtheyellowlinerepresentstheimprovedeconomic
dispatchmodel.Thegreenlinesrepresenttheminimumupanddownregulationprocured
duringthisstudyperiod.Originally,theproxyAGCunitoutputwasoffsetbyanaverageof
about180MW,asindicatedbytheheavyblueline.Withanhouraheadforecastincluded,the
offsetisreducedtoabout60MW,asindicatedbytheheavyyellowline.Thisshowsthatthe
EDunitsarebetterabletofollowloadwiththeimprovedwindforecastincorporatedintothe
dispatchfunction.
Figure153showstheimpactofanimprovedwindforecastontheeconomicdispatchandload
followingrequirements.Thebluelinerepresentstheoriginaleconomicdispatchmodelwitha
persistencebasedwindforecast,andtheyellowlinerepresentstheimprovedeconomic
dispatchmodel.Withtheoriginaleconomicdispatchmodel,theEDunitswereunabletokeep
upwiththerapidincreaseinwindgeneration.Withtheincorporationofanimprovedwind
forecast,theEDunitsdoprovidesufficientloadfollowingduringthekeytimeframethefirst
15minutesofthestudyperiod.
Thus,incorporatingashorttermforecastintotheeconomicdispatchmodelimprovessystem
performance.Thisstudyscenariousedaperfectforecastforillustration.Astateoftheartshort
termforecastwouldstillprovideimprovedperformanceoverall.However,therewouldbe
timeswhentheshorttermforecastwaswrong.Duringthesetimes,incorporationoftheforecast
intotheeconomicdispatchcouldhaveanadverseimpactonsystemperformance.
158
-600
-400
-200
0
200
400
600
4:00 PM 5:00 PM 6:00 PM 7:00 PM
Preg Original
Preg Wind Forecast
Reg Up
Reg Down
Figure 152. Impact of Including Hourly Wind Forecast on Regulation Duty During the June
Evening QSS Study Period.
-200
-150
-100
-50
0
50
4:00 PM 5:00 PM 6:00 PM 7:00 PM
Delta ED MW Original
Delta ED MW Wind Forecast
Figure 153. Impact of Including Hourly Wind Forecast on Economic Dispatch and Load Following
During the June Evening QSS Study Period.
M
W
M
W
/
M
i
n
u
t
e
159
Temporary Wind Ramp Rate Limit
Theprevioussectionexploredonemitigationmethodforrespondingtorapidincreasesinwind
generationincorporatinganhouraheadwindforecastintotheeconomicdispatchmodel.
Anothermitigationmethod,atemporaryrampratelimitonincreasingwindgeneration,is
examinedinthissection.TheoriginalJunewindprofile(blueline)andtherampratelimited
windprofile(yellowline)areshowninFigure154.Forthisillustration,acaponwind
generationisappliedatapproximately4:15pm.ThiswastriggeredbytheAGCwhenthe
regulationrequirementapproachedtheprocuredregulationthreshold.Totalwindproduction
wascappedforthenext15minutes.Thecapwasthenremovedandwindfarmoutputwas
allowedtoincreaseatnomorethan1%/minuteforthedurationofthesimulation.Hence,the
totalwindoutputwiththecapandrampratelimitneverreturnstotheoriginallevel.This
temporarycapandrampratelimitwasimplementedatallwindfarms.Thewindenergylost
duringthecurtailmentwasabout550MWhoutofabout16,300MWh,or3.5%,forthis3hour
period.
CrossplotsofallthreesetsofQSSresultsfortheJunestudyperiodareshowninFigure155and
Figure156.Inbothfigures,thedarkbluelinerepresentstheoriginalcasewithnowind
generationcapandrampratelimitandwithapersistencebasedwindforecast.Theyellowline
representsthefirstsensitivitycasewithnowindgenerationcapandrampratelimitbutwithan
houraheadwindforecastincorporatedintheeconomicdispatch.Thelightbluelinerepresents
thesecondsensitivitycasewiththewindgenerationcapandrampratelimitandwitha
persistencebasedwindforecast.
Figure155showstheoutputoftheproxyAGCunitforthethreescenarios.Notethattheshort
termdepletionofregulationcapabilityisgreatlyreducedbyeitherincorporatingawind
forecastintotheeconomicdispatchmodelorlimitingthewindgenerationduringperiodsof
significantincrease.
Figure156showstheEDunitsignalforthethreescenarios.Again,bothmitigationmethods
allowtheEDunitstoprovidesufficientloadfollowingduringthekeytimeframethefirst15
minutesofthestudyperiod.
160
0
1000
2000
3000
4000
5000
6000
7000
8000
4:00 PM 5:00 PM 6:00 PM 7:00 PM
Original Wind
1% Ramp Rate Up Limit
Figure 154. Temporary Cap with RampUp Rate Limit on Wind Generation.
-600
-400
-200
0
200
400
600
4:00 PM 5:00 PM 6:00 PM 7:00 PM
Preg Original
Preg Wind Forecast
Preg Wind Ramp Rate
Reg Up
Reg Down
Figure 155. Comparison of Regulation During the June Evening QSS Study Period.
M
W
M
W
161
-200
-150
-100
-50
0
50
4:00 PM 5:00 PM 6:00 PM 7:00 PM
Delta ED MW Original
Delta ED MW Wind Forecast
Delta ED MW Wind Ramp Rate
Figure 156. Comparison of Economic Dispatch and LoadFollowing During the June Evening QSS
Study Period.
5.3. Summary of Results
AsummaryoftheQSSanalysisineachofthestudyperiods(e.g.,Julymorningloadrise,May
nightlightload,Juneeveningloaddecrease)isprovidedinthissection.
TheJulymorningloadrisesimulationsillustrated:
Loadfollowingincreasesduetoanetdecreaseinrenewablegeneration
Mostoftheloadfollowingisperformedbycombinedcycleplants,followedby
hydroelectricunits
Theregulationdutyisdueprimarilytoloadvariation,butisoffsetupbythe
decreaseinwindpower
Ingeneral,aneconomicallyrationalunitcommitmentanddispatchprovides
adequateloadfollowingcapability
TheMaynightlightloadsimulationsillustrated:
Largestepsinloadhavemoresignificantimpactonregulationthanwindvariability
Insufficientdowncapability(bothrangeandramprate)shiftsloadfollowingdutyto
regulation,whichmaythenbecomeexhausted
VariabilityofwindincreasestheEDunitduty,bothintermsofMW/minuteandthe
frequencyofsignchanges
M
W
/
M
i
n
u
t
e
162
Theimportanceofusingavailablemaneuverability,andavoidingnontechnical
constraints
Changeincommitmentorcurtailmentofwindeffectivelymitigateslossof
maneuverability
TheJuneeveningloaddecreasesimulationsillustrated:
Regulationpicksupsustainedchangesinwindpower
Thisincreasesthelikelihoodthatregulationwillbeexhausted,andlessableto
respondtorapidchangesinload
Incorporatingashorttermwindforecastintotheschedulecanimproveload
followingandreducetheshifttoregulationduty
Shorttermcurtailmentofwindrise(withrampratelimitation)canalsoachievethat
goal.Somewindenergyproductionislosttoachievethisbenefit.
163
6.0 Operational Implications and Mitigation Methods
TheprecedingsectionspresentedtheimpactofintermittentrenewablesontheCalifornia
systemfromdifferentperspectives:statisticalexpectationsofvariationanduncertainty(Section
3.0),overalldaytodayoperationandeconomicbehavior(Section4.0),andselectedillustrations
offaster,intrahourbehavior(Section5.0).Ineachsection,resultsandobservationsparticularto
thatanalyticalviewpointwerepresented.
Inthissection,windvariabilityanduncertaintyarefurtherexaminedwithinthecontextofthe
relationshipbetweentheseanalyticalperspectives.Thisprovidesadditionalinsightintothe
overalloperationalimplicationsofintermittentrenewablesandpotentialmeansofmitigating
adverseimpacts.
Throughoutthisreport,thetimeframesforvariabilityandtherequirementsforoperational
flexibilityhavebeenconsistentlydefinedasfollows:
1hourdeltareferstothechangefromtheprevioushour.Theabilityoftheoperating
areatoaccommodatehourlychangesiscalledscheduleflexibility.
5minutedeltareferstothechangefromtheprevious5minuteperiod.Load
followingandeconomicdispatchfunctionsoperateinthistimeframe.
1minutedeltareferstothechangefromthepreviousminute.Regulationfunctions
operateinthistimeframe.
Inpractice,theboundariesbetweenthesetimeframesarenotcrisplydefined,andtheresources
andpracticesthatimpactonewilloftenimpactanother.
Thefollowingsectionswillvalidatethestudyapproachagainsthistoricaldata,examinethe
systemflexibilityrequirementsfromanoperationalperspective,anddescribeselected
mitigationstrategies.
6.1. Validation
Inthissection,thestudyapproachandresultsarecompared,andvalidated,againsthistorical
dataandperformance.
Atypicaldayofoperation,asrecordedbytheCaliforniaISO,isshowninFigure157.Thisis4
secondresolutiondata.Therefore,itshowsallofthetimeframesexaminedinthisproject.The
totalCaliforniaISOloadisrepresentedbythepinkline,thetotalgenerationunderCalifornia
ISOcontrolisrepresentedbythelightblueline,andthetotalscheduledinterchangeis
representedbythebrownline.Theleftyaxisscaleappliestotheloadandgeneration,andthe
rightyaxisscaleappliestotheinterchange.Theinterchangescheduleislargelyblock
scheduledonthehour,withroughly10minutetransitions.Asexpected,theinstategeneration
closelytracksthesystemload.
Thisvalidatesabasicanalyticalpremiseofthisproject.Specifically,WECCwideunit
commitmentanddispatchisproperlycapturedwiththeonehourresolutionusedinthe
productionsimulationanalysis.Thisfigurealsoshowsthatessentiallyalloftheloadfollowing
164
isperformedbyinstategeneration.Themaneuveringcapabilityevaluation(Section4.0),was
basedonthisassumption.Equallyimportant,thiswasabasicboundaryconditionfortheQSS
analysis(Section5.0):onlyasubsetofavailableCaliforniagenerationsunitswereusedto
examineloadfollowingbehavior.
Figure158showsaweekofactualhistoricalinterchange.Notethattheinterchangefollowsone
hourmodifications,andthereisconsiderabledaytodayvariation.Theinstantaneousrange,
fromminimumtomaximum,ofthesesevendaysisshowninFigure159.Themaximum
variationininterchangewasabout3,000MW,andconsistentlyvariesfrom500MWto
2,000MWduringlightloadperiods.ThisvalidatestheeconomicvariabilityinCalifornias
powerexchangewithneighboringsystemsasshownintheproductionsimulationanalysis.
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
50000
0:00:00 4:00:00 8:00:00 12:00:00 16:00:00 20:00:00 0:00:00
L
o
a
d

a
n
d

G
e
n
e
r
a
t
i
o
n

M
W

-8000
-7000
-6000
-5000
-4000
-3000
-2000
-1000
0
I
n
t
e
r
c
h
a
n
g
e

S
c
h
e
d
u
l
e

M
W
Load - MW
Gen Total
Total Interchange Schedule
Figure 157. Schedule and Load Following for a Sample Day of California Operation.
165
-10000
-8000
-6000
-4000
-2000
0
0:00:00 4:00:00 8:00:00 12:00:00 16:00:00 20:00:00 0:00:00
M
W

7/19/04
7/20/04
7/21/04
7/22/04
7/23/04
7/24/04
7/25/04
Figure 158. Historical Interchange for a Sample July Week.

0
1000
2000
3000
0:00:00 4:00:00 8:00:00 12:00:00 16:00:00 20:00:00 0:00:00
M
W

Figure 159. Instantaneous Range of Interchange for Sample July Week.
166
Acloserinspectionoftheinterchangeshowsexpectedperformanceintheregulationtime
frame.Figure160showsthescheduled(brownline)andactual(orangeline)interchangefora
sampleday.ItalsoshowstheISOcontrolareacontrolerror(ACE,darkblueline).Thelefty
axisscaleappliestotheinterchangesignals,therightyaxisscaleappliestotheACEsignal.
Thetotalinterchangescheduleoccursinonehourblockswithtransitionsonthehour.
Deviationsfromschedule,i.e.,thedifferencebetweenscheduledandactualinterchange,
mirrorstheACEsignal.ACEgenerallyvariesintherangeof200MW,whichisconsistentwith
theCaliforniaISOsprocurementofregulationservices.NotethatthelargestACEexcursions
correspondtothehourlyschedulechanges.Thus,theschedulechangesmaybecausing
avoidableACEviolations.The5minuteeconomicdispatchworkstocorrecttheseexcursions,
butthedatasuggeststhatmorefrequentinterchangeschedulechanges(e.g.,onthequarteror
halfhour)couldalsoreducetheACEexcursions.Suchschedulechangeswouldalsoincrease
theflexibilityavailabletoaddressintermittentrenewablevariability.Thisissueisaddressed
furtherinthediscussionofCPS2(Section6.3.4).
NotealsothatthecontinuouschangesinloadshowninFigure157arenotobservedineither
theACEorthescheduleofFigure161.Thisconfirmsthatthemajorityofloadfollowingis
performedbyinstategeneration,aswasassumedforthisstudy.
-10000
-9000
-8000
-7000
-6000
-5000
-4000
-3000
-2000
-1000
0
0:00:00 4:00:00 8:00:00 12:00:00 16:00:00 20:00:00 0:00:00
I
n
t
e
r
c
h
a
n
g
e

M
W

-2000
-1500
-1000
-500
0
500
1000
A
C
E

M
W
Interchange Actual
Total Interchange Schedule
CAISO Control ACE
Figure 160. Regulation and Interchange for a Sample Day of California Operation.
167
Themethods,assumptions,andresultsoftheQSSanalysisarevalidatedbyacomparisonof
Figure161andFigure162.Figure161showstheactualCaliforniaISOACEforthree3hour
periods:May15,2003from1amto4am(blueline),June24,2004from4pmto7pm(yellow
line),andJuly19,2004from5amto8am(greenline).Figure162showstheoutput(Preg)ofthe
proxyAGCunitfromtheQSSanalysisofcomparabletimeperiodsasprojectedfor2010.Again,
thebluelinerepresentstheMaystudyperiod,theyellowlinerepresentstheJunestudyperiod,
andthegreenlinerepresentstheJulystudyperiod.
NotethattheyaxisscaleoftheQSSresultsislargerthanthatofthehistoricaldata.TheQSS
resultsrepresentthe2010Xscenariowithsignificantlymoreload,aswellaswindandsolar
generation,incomparisontothe2003and2004historicaldata.
Aseconddifference,oneofsignconvention,isalsoobserved.ThehistoricalACEisactual
interchangeminusscheduledinterchange,ignoringthefrequencyterm.Thus,apositive
historicalACEindicatesthatgenerationinCaliforniaISOexceedsload,resultinginanincrease
inpowerexport.IntheQSSresults,Pregisthepowerneededtobalancegenerationandload.
Thus,apositivePregindicatesthatgenerationwaslessthanload,resultinginoutputfromthe
proxyAGCunit.
Thekeyobservationisthattheoverallcharacteroftheplotsissimilar.Thereisalargestepin
historicalACEatabout20minutesintotheJune24,2004data.Thereisacomparable,albeit
larger,stepinPregfromtheJuneeveningQSSanalysis.Therearesimilarlargestepsinthe
historicalACEdatafortheMaytimeperiodatabout30minutes,90minutes,and135minutes
thatarealsoreflectedintheQSSresults.
-400
-200
0
200
400
0 60 120 180
May 15 2003
J une 24 2004
J uly 19 2004
Figure 161. Historical ACE Data.

M
W
168
-600
-400
-200
0
200
400
600
0 60 120 180
May Night Preg
J une Evening Preg
J uly Morning Preg
Figure 162. Pseudo-ACE from QSS Simulations.

Figure163showsadurationcurve(blueline)ofhistorical2003loaddata.Itissortedfrom
highestloadtolowestload,andtheMWscaleisontheleftyaxis.Thetwoothertracesinthis
figureshowthehistoricalamountofupregulation(yellowline)anddownregulation(green
line)procuredateachloadlevel.Theregulationdataisnotsorted,butistimesynchronizedto
theloaddata.TheregulationMWscaleisontherightyaxis.
Theprocuredupregulationrangesfromabout300MWto800MW.Ingeneral,thereismoreup
regulationprocuredathigherloadlevelsthanatlowerloadlevels.Thisisconsistentwiththe
needtomaintainloadservingcapability,particularlyduringpeakloadtimes.
Theprocureddownregulationisnotascorrelatedwithloadlevelastheupregulation.Itranges
fromabout300MWto550MW.
Thisdatasupportstheuseofageneric+320MWofupregulationand320MWofdown
regulationintheQSSanalysis.Thisisaconservativeassumptionfortworeasons.First,the
amountofregulationprocuredisoftengreaterthan320MW.Second,theQSSanalysisis
focusedontheartificiallyaccelerated2010Xscenariowhichincludessignificantlymoreload,as
wellaswindandsolargeneration,thanisreflectedinthe2003data.
M
W
169
0
10000
20000
30000
40000
50000
1 731 1461 2191 2921 3651 4381 5111 5841 6571 7301 8031
-600
0
600
1200
1800
2003 Load
2003 Reg Up
2003 Reg Down
Figure 163. Historical Load and Procured Regulation Data for 2003.
6.2. Statistical Results and Operational Flexibility
Inthissection,selectedstatisticalresultsfromSection3.0arerevisitedinthecontextofthe
productionsimulationandQSSresults.Ingeneral,thestatisticsprovideinsightintooverall
systemvariabilityaswellastheimpactofintermittentrenewables.However,thestatistic
analysisdidnotaddresssystemoperationsandperformance.Withtheresultspresentedin
Sections4.0and5.0,itispossibletoextractoperationalrequirementsfromthestatistical
analysis.Specifically,theproductionsimulationanalysisidentifiedthemixofresources
availableatanygiventimetomeetthemaneuverability/flexibilityrequirements.TheQSStime
simulationsillustratedtherelationshipbetweentheboundaryconditionsfromtheproduction
simulationanalysisandtheminutetominutebehaviorofthesystem.
Thestatisticalanalysisexaminedtheexpectedvariabilityofloadandintermittentrenewablesin
differenttimeframes.Throughout,theanalysisassumedthatthisvariabilityexhibitedanormal
distributionmeasurablewithastandarddeviation().Examinationofthedistributions
supportedthisapproximation.Inanormalpopulation,threetimesthestandarddeviation
coversthevastmajorityofevents.Thatis,99.7%ofeventsfallwithin+/3.Thus,+/3isa
proxyfortheflexibilityrequirements.Similarly,anincreasein3isonemeasureofthe
additionalflexibilityrequiredduetotheincreasedvariabilityfromintermittentresources.
M
W
M
W
170
The3measureisusedinTable35,wherethetotalandlightloadstatisticalresultsare
revisited.Theloadandnetloadvariabilityforeachscenario,aswellasthedifferencebetween
them,isdisplayedineachsetofrows.Thefirstthreecolumnsshowthetotalstandarddeviation
()ofthe1hour,5minuteand1minutedeltas.Thefinalthreecolumnsshowthelightload
standarddeviation()ofthe1hour,5minuteand1minutedeltas.Thevariabilityduetoload
aloneisreportedaswellastheincreaseinvariabilityduetointermittentrenewables.Ineach
comparison,theincrementalrequirementisbasedonthechangein3.
Table 35. Total and Light Load Change in Flexibility Requirements.
Total Light Load (10
th
Decile)
1-Hour
(MW)
5-Min
(MW)
(1)

1-Min
(MW)
(2)

1-Hour
(MW)
5-Min
(MW)
(1)

1-Min
(MW)
(2)

2006 Load 1,436 189.3 44.8 669 86.5 40.8
2006 L-W-S Change 15 (+1%) 0.3 (+0.2%) 0.1 (+0.2%) 30 (+4%) 2.7 (+3%) 0.1 (+0.2%)
Increased Need (3) 45 0.9 0.3 90 8 0.3
2010 Load 1,575 207.6 49.1 734 94.9 44.8
2010T L-W-S Change 48 (+3%) 6.9 (+3%) 1.6 (+3%) 199 (+27%) 14.2 (+15%) 1.1 (+3%)
Increased Need (3) 144 21 5 597 42.6 3.3
2010 Load 1,575 207.6 49.1 734 94.9 44.8
2010X L-W-S Change 129 (+8%) 14.2 (+7%) 3.3 (+7%) 347 (+47%) 19.8 (+21%) 2.9 (+7%)
Increased Need (3) 387 42.6 9.9 1041 59.4 8.7
2006 Load 1,436 189.3 44.8 669 86.5 40.8
2010 Load Change 139 (+10%) 18.3 (+10%) 4.3 (+10%) 65 (+10%) 8.4 (+10%) 4.0 (+10%)
Increased Need (3) 417 54.9 12.8 195 25.2 12
Overall,theincrementalvariabilityduetotheexpectedgrowthofintermittentrenewables(i.e.
the2010Tscenario)increasesabout~3%acrossalltimeframes.Theartificiallyacceleratedcase
(i.e.,the2010Xscenario)resultsinanoverallincreaseinvariabilityofabout7%to8%acrossall
timeframes.Duringthelightestloadperiods,therequirementsarelowerbuttherelative
impactoftheintermittentrenewablesisgreater.
6.2.1. Overall Requirements
ThefirstdatacolumninTable35showsaprogressionofonehourloadandnetloadvariability
forthe2006and2010scenarios.Thiscolumnshowsquantitativelytheincreasingrequirements
foroverallflexibilityinhourlyscheduling.Theexpectedhourtohourchangeincreaseswith
loadgrowthandwiththeadditionofintermittentrenewables.
Inthistimeframe,thebasiclevelofflexibilityrequiredtoserveloadnowisabout4,300MW/hr.
Thisisthreetimesthestandarddeviationofloadalonevariability,or3x1,436.The2006
intermittentrenewablesincreasethatrequirementslightlyto4,353MW/hr.Thesystemload
growthto2010drivestheloadalonevariabilityrequirementupto4,725MW/hr,whichisan
increaseofabout400MW/hrduetoload.Theexpectedgrowthofintermittentrenewables,
representedbythe2010Tscenario,furtherincreasesthehourlyflexibilityrequirementto
4,869MW/hr,anincreaseof144MW/hrorabout35%.
171
By2020,theincreaseinloadhasdriventheloadaloneflexibilityrequirement(Table10in
Section3.0)uptoabout6,000MW/hr.Withintermittentrenewables,afurtherincreaseof
126MW/hrappears.
Intheartificiallyaccelerated2010Xcase,thehighlevelofintermittentrenewablesoutpacesload
growth.Thus,thehourlyflexibilityrequirementincreasedanother387MW/hrto5112MW/hr.
ThetimerelationshipofthechangeinvariabilityisshowninFigure164.Thelowertrace(dark
blueline)showstheexpectedincreaseinhourlyvariabilityduetoloadgrowthfrom2006to
2020.Theuppertrace(pinkline)showstheincreaseduetobothloadandtheexpected
intermittentrenewablegrowth(2010T,2020).Forthisexpectedtrajectory,loaddominatesthe
increasingrequirement.Theimpactoftheintermittentrenewablesin2010Tisabout1/3ofthe
loadimpact,andtheintermittentrenewableimpactin2020isabout10%oftheloadimpact.
Theincreaseinthehourlyrequirement(i.e.,threetimesthedifferencebetweenthetwocurves)
isabout130MWoveralltheyears.Incontrast,theincreaseduetoloadaloneisabout1,700MW
(i.e.,threetimesthedifferenceinloadalonehourlydeltafrom2006to2020).Theimpactof
acceleratedadditionofintermittentrenewablesappearsasthesinglepoint(orangesquare)for
2010X.Thisconfirmsthatafastergrowthofrenewableswilladvancethetimingofincreased
flexibilityrequirements.
1400
1600
1800
2000
2200
2006 2008 2010 2012 2014 2016 2018 2020
Year
1

H
o
u
r

D
e
l
t
a

S
i
g
m
a

(
M
W
/
h
r
)
Load Only
Load - Wind -Solar
2010X
Figure 164. Standard Deviations of Hourly Deltas.

ThesecondandthirddatacolumnsofTable35showsimilartrendsforsubhourlyvariability.
The20065minuterequirementis568MWper5minutes(i.e.,3x189.3),withessentiallyno
overallimpactfromwindandsolar.Loadgrowthto2010increasesthisrequirementto623MW
per5minutes,anincreaseof54.9MWper5minutesorabout10%.Theincremental
172
requirementduetointermittentrenewablesaddsupto42.6MWper5minutes(about8%)fora
totalof665MWper5minutes.Theresultsaresimilarforregulationrequirements.
Figure165showstheannualchangeinthestandarddeviationsinthethreetimeframes.The
changesduetowindandsolargenerationforthe2006,2010Tand2010Xscenariosaswellasthe
changeduetoloadalone(2010)areshown.Themagnitudeoftheimpactofintermittent
renewablesislessthanthatoftheloadchangesacrossalltimeframes.
0
20
40
60
80
100
120
140
160
1 hr 5 min 1 min
One Standard Deviation delta
M
W
2006 wi nd & sol ar
2010T wi nd & sol ar
2010X wi nd & sol ar
l oad '06 to '10
Figure 165. Annual Standard Deviation Changes.

Thehourlyand5minuterequirementscanalsobecompared.Overall,theloadfollowing
requirement(i.e.,threetimesthestandarddeviation)willgrowtoabout600MWto700MWper
5minutes.Notethatthescheduleflexibilityrequirementisabout5,000MWperhour.
Assuminglinearity,thisisabout400MWper5minutes.Bydefinition,multipleperiodsof3
changewouldnotbeexpected.Therefore,therequirementfor600MWto700MWper5
minutes,orabout130MW/minute,isstatisticallyconsistentwiththehourlyrequirement.
6.2.2. Light Load Requirements
Atlightload,therequirementforscheduleflexibilityislower,buttherelativeimpactof
intermittentrenewablesishigher.ThethreerightdatacolumnsinTable35showthechangesin
flexibilityrequirementforthelightloadconditionsin10%oftheyear(i.e.,10
th
decile).In2006,
thelightloadhourlyschedulerequirementduetoloadaloneisabout2,000MW/hr.In2010,the
lightloadrequirementisabout2,200MW/hr,growingtoabout3,000MW/hrin2020.With
intermittentrenewables,therequirementgrowstoabout4,000MWin2020.Notethatthislight
loadrequirementismuchlessthantheoverallrequirementofabout6000MW/hr.
Figure166showsthegrowthinhourlyvariabilityatlightload.Thelowertrace(darkblueline)
showstheexpectedincreaseinhourlyvariabilityduetoloadgrowthfrom2006to2020.The
uppertrace(pinkline)showstheincreaseduetobothloadandtheexpectedintermittent
173
renewablegrowth(2010T,2020).Theartificiallyaccelerated2010Xscenarioisrepresentedbythe
orangesquare.
Thehourlyand5minutechangeinrequirementsduetoloadareabouthalfoftheoveralllevels.
Intermittentrenewablesareresponsiblefortheotherhalfoftheincrease.Thus,therelative
impactofintermittentrenewables,primarilywindatlightload,isgreaterduringlightload
periods.
Light Load Hourly Variability
600
800
1000
1200
1400
2006 2008 2010 2012 2014 2016 2018 2020
Year
1

H
o
u
r

D
e
l
t
a

S
i
g
m
a

(
M
W
/
h
r
)
Load Only
Load - Wind -Solar
2010X
Figure 166. Standard Deviations for OneHour Deltas at Light Load.

Similarobservationsapplytotheloadfollowingrequirements,althoughtheimpactofwindis
somewhatless.Specifically,theloadalonerequirement(i.e.,3)is285MWper5minutesunder
lightloadconditions,and623MWunderallloadconditions.Withwind,thelightload
requirementisasmuchas344MWper5minutes,orabout70MW/min.Again,thisisstillless
thantheoverallrequirement,butabout20%higherthantherequirementforloadalone.
Figure167showstheannualchangeinthestandarddeviationsunderlightloadconditions.The
hourlychangesaregreater,butthelightloadrequirementisstilllessthantheoverall
requirement.However,providingthisloadfollowingcapabilityinthedowndirectionatlight
loadmayprovechallengingforsomeoperatingconditions.
Unlikethelongertimeframes,changesinthestandarddeviationofregulation(i.e.,1minute)is
notaffectedbyloadlevel.
174
0
50
100
150
200
250
300
350
400
1 hr 5 min 1 min
One Standard Deviation delta
M
W
2006 wind & solar
2010T wind & solar
2010X wind & solar
load '06 to '10
Figure 167. Light Load Standard Deviation Changes

6.2.3. Extremes
Table36summarizesthehourlyloadandnetloadvariabilityforthe2010T,2010X,and2020
scenarios.Theloadandnetloadvariabilityforeachscenario,aswellasthedifferencebetween
them,isdisplayedineachsetofthreerows.The1hourstandarddeviationisshown,aswellas
thelargestsingleloadincreaseanddecrease.
Ingeneral,thetrendsinthemaximaandminimaaresimilartothoseobservedinthemore
statisticallymeaningfulstandarddeviation.Hence,astudyapproachbasedonindividualworst
caseshassomeintuitiveappeal.However,relianceonasingledatapointisproblematic.For
example,themaximum2010Tnetload1hourdeltaisabout400MWlessthanthe1hourdelta
ofthe2010loadalone.Astudyapproachbasedonextremeswouldconcludethatthenetload
requirementshouldbelessthantheloadalonerequirement.Thus,astatisticaloutlierhas
distortedtheresults,andtherefore,theconclusions.
Table 36. Total Variability from Statistical Analysis.
1-Hour Maximum 1-Hour Minimum 1-Hour
2010 Load 1575 MW 6714 MW -5617 MW
2010T L-W-S 1623 MW 6312 MW -5713 MW
Change 48 MW -402 MW -96 MW
2010 Load 1575 MW 6714 MW -5617 MW
2010X L-W-S 1704 MW 7219 MW -5986 MW
Change 129 MW 505 MW -369 MW
2020 Load 1977 MW 8427 MW -7049 MW
2020 L-W-S 2019 MW 8747 MW -7351 MW
Change 42 MW 321 MW -302 MW
175
6.3. Operational Flexibility
Inthissection,thesystemwidemaneuverabilityrequirementsareexaminedintheforecasting,
dayaheadunitcommitmentanddispatch,hourlyschedule,5minuteloadfollowingand1
minuteregulationtimeframes.
6.3.1. Forecasting
ThedistinctionbetweenvariabilityanduncertaintywasdiscussedinSection3.5.Bynature,
bothsystemloadsandintermittentrenewablesvary.Therefore,muchofthisstudyfocusedon
theimpactofintermittentrenewablevariabilityrelativetoloadvariability.Thepowersystems
responsetothisvariabilitydependsuponthecapabilityoftheavailableresources.One
importantaspectofthisresponseisknowledge,apriori,oftheoutputrequirement.Ingeneral,a
moreunpredictablevariationrequiresmoreagilegeneratingresources.
Loadandwindforecasting(Section5.0)haveasignificantimpactonsystemoperatingcosts.
Figure168showsamixofhistoricalloadforecastsfromtheCaliforniaISO,andwindforecasts
developedforthisstudy.Thefouruppercurvesrepresentactualload,houraheadloadforecast,
dayaheadloadforecastandtwodayaheadloadforecast.Thelowercurvesrepresenttheactual
windanddayaheadwindforecast.Thisfigurevalidatesboththeanalyticalapproachandthe
recommendationsforflexibilityinmultipletimeframes.
Notethattheaccuracyoftheloadforecastimprovesasthetimehorizondecreases.Further,the
amplitudeoftheforecasterrorissomewhatcorrelatedtotheloadmagnitude.Thedetailed
forecasterrorstatistics(Section3.0)confirmthis.Thewindforecasterror,ontheotherhand,is
notcorrelatedtosystemload,butratherhassomecorrelationtothewindpower.Froman
operationalperspective,thisdifferenceisimportant.Overall,thedayaheadforecastincluding
windandsolarintroducesabouttwicetheuncertaintyastheloadforecastalone.However,
significanterrorsinwindforecastatlowloadperiodshavealargerimpactrelativetothe
balanceofgenerationavailable.Theuncertaintyduetointermittentrenewablescanbethree
timesgreaterthantheuncertaintyduetoloadaloneatmoderatetolightloadlevels.
Theintroductionofintermittentrenewablestendstomaketheamplitudeofoperations
uncertaintylesscorrelatedtoloadlevel.Thestatisticalanalysisshowsthatarangeofabout
+/5,000MWboundsthedayaheaduncertaintywithintermittentrenewables.However,asthe
operationalhourapproaches,uncertaintyinboththeloadandrenewablegenerationforecasts
dropsconsiderably.Theforecaststatistics(Section3.0)showthatintermittentrenewables
increasethehouraheaduncertaintyabout20%overtheloadaloneuncertainty.Arangeof
+/2,000MWboundsthehouraheaduncertaintywithintermittentrenewables.
176
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
50000
0:00 4:00 8:00 12:00 16:00 20:00
2004 System load, actual
2004 System load, hour ahead forecast
2004 System load, day ahead forecast
2004 System load, 2 day ahead forecast
Actual Wind
Forecast Wind
Figure 168. Comparison of Historical Forecast and Actual Load and Simulated Day-Ahead
Forecast and Actual Wind.
Implications of Ignoring Forecasts
Systemoperationsalreadyfaceloadforecastuncertainty.Animperfectloadforecast,aswellas
animperfectintermittentrenewableforecast,playsacriticalroleinsecureandeconomic
operationofsystems.Systemswithsmallamountsofintermittentrenewablescanlargelyignore
themindayaheadoperations.Asthepenetrationofintermittentrenewablesincrease,however,
suchapracticebecomesuntenable.
Figure169showsanopenhighlowclosestockchartofthe2010Xdayaheadhourlyloadand
netloadforecasterrorswhenbothwindandsolarforecastsareignored.Inalloperatingtime
frames,butespeciallyunderlightloadconditions,ignoringstateoftheartforecastshasahuge
impactonthenetloadforecasterrors.Allstatisticalquantities(average,standarddeviation,
extremes)aresignificantlyincreased.Asnotedabove,thestandarddeviationoftheloadalone
forecasterrorforthe10thdecileis570MW.Thenetloadforecasterrorreachesintothe
6,000MWto8,000MWrange,withevengreateroutliers.Theeconomicanalysis(Section4.2)
showedthattheeconomicpenalty,intermsofoperationalinefficiency,completelyswampsany
benefitsduetotheadditionoftherenewables.Thepenaltyismeasuredinthebillionsofdollars.
M
W
177
1 2 3 4 5 6 7 8 9 10
-8000
-6000
-4000
-2000
0
2000
4000
6000
8000
10000
12000
14000
16000
1 2 3 4 5 6 7 8 9 10
F
o
r
e
c
a
s
t

E
r
r
o
r

(
M
W
)
Decile
Load Error
Total Error
Load Error
Total Error Load Forecast > Actual Load
Figure 169. 2010X Load and Net Load Day-Ahead Forecast Error Ignoring Wind and Solar Forecast
Stock Chart by Decile.
6.3.2. Unit Commitment and Schedule Flexibility
Theimpactofintermittentrenewablegenerationonhydrooperationandtheavailabledispatch
rangearediscussedinthefollowingsections.
Hydroelectric Generation Shift
Resultsoftheproductionsimulationanalysis(Section4.0)supportedbytheQSS
analysis(Section5.0)showthathydrooperationplaysanimportantroleforoperationwith
significantintermittentrenewablegeneration.Theabilitytomodifythehydrodispatchwith
smalleconomicpenaltymakesitanaturalcountertovariationfromintermittentrenewables.
Theproductionsimulationresultsshowedthathydrooperationwithinthestateproducesthe
sameamountofenergy,butistemporallyshiftedwiththeadditionofwindandsolar
generation.
ThenumberofhoursofhydrodisplacementforasingleyearisshowninFigure170.Thered
barshowsthenumberofhourswithadisplacementgreaterthan500MW,theyellowbarshows
thehourswithadisplacementgreaterthan1,000MW,andthelightbluebarshowsthehours
withadisplacementgreaterthan1,500MW.Themajorityofthehoursinayearhavea
displacementlessthan500MWandarenotshowninthisfigure.Notethatthechangeinhydro
operationisgreaterthan1,000MWabout8%ofthetimeinthe2010Xscenarioandonly3.5%of
thetimeinthe2010Tscenario.Theconventionalhydroshouldhavethecapabilitytoprovide
thismaneuverability,particularlywhenaugmentedbytheavailablepumpingloads.
178
0
500
1000
1500
2000
2500
2010T 2010X
#

o
f

h
o
u
r
s

w
i
t
h

a
b
s
o
l
u
t
e

v
a
l
u
e

o
f

h
y
d
r
o

s
h
i
f
t

>

s
p
e
c
i
f
i
e
d

M
W
> 500
> 1000
> 1500
Figure 170. Intermittent Generation Impact on Hydro Operation.

Available Dispatch Range
Theproductionsimulationanalysis(Section4.4.3)showedthedistributionofavailabledispatch
range,bothupanddown,asafunctionofsystemload.Figure171isbasedonthesameyearof
data,butispresentedasafunctionoftimeofday.Thefirstgroupoftracesshowsthe
maximumandminimum(solidbluelines)andtheaverageavailable(dottedblueline)up
dispatchrange.Thesecondgroupoftracesshowsthemaximumandminimum(solidorange
lines)andtheaverageavailabledowndispatchrange(dottedorangeline).Thethirdgroupof
tracesshowsthehourlychangeinnetload(loadminuswindminussolar)forthe2010Xcase.
Thesolidgreenlinesarethemaximumandminimumandthedottedgreenlineistheaverage.
Theplotshowstheexpectedtendencyforlowerdownrangeatlightloadandloweruprangeat
highload.Inaddition,therearefewpointsatwhichtheextremerequirement(maximumor
minimumofthegreenlines)impingeonthecorrespondingleastflexibledayoftheyearfora
givenhour.Periodsofextrememorningandeveningloadrisepresentsomerisk.However,itis
importanttonotethatnosimultaneousextremeseveroccurredinthedata.Itisunlikelyforthe
uprangerequirementtoexceedcapability.Similarly,extremehourswiththeleastrangedown
capabilityoccurduringlateeveningandearlymorninghours.Aswiththerangeup,no
simultaneousextremeseveroccurredinthedata.Again,therequirementisunlikelytoexceed
capability.
179
-25000
-20000
-15000
-10000
-5000
0
5000
10000
15000
20000
25000
0 6 12 18 24
Maximum of Range Up Average of Range Up Minimum of Range Up
Maximum L-W-S Range Need Average L-W-S Range Need Minimum L-W-S Range Need
Maximum of Range Down Average of Range Down Minimum of Range Down
Figure 171. Committed Generation Range and Maximum Hourly NetLoad Change.
6.3.3. Load-Following
TheproductionsimulationresultsusedtocreateFigure171alsoproducedmeasuresof
availablerampingcapability.Figure172presentstherampingcapabilityandrequirementsina
similarfashion.Thisfigureshowsthemaximum,minimum,andaverageoftheup(bluelines)
anddown(orangelines)rampingcapabilityinMW/min.Therequirementtracesarederived
andpresentedslightlydifferently.Thesolidgreencurveistheaveragerampingrequirementfor
eachhour,basedonthenetloadvariabilityofthathour.Therequirementrange(dottedgreen
lines)isbasedontheoverallloadfollowing(5minutedelta)statisticalvariation:3x5minute
=120MW/min(i.e.600MW/5minutes).
Notethattypicallytherampcapabilitygreatlyexceedsexpectedrequirement.Inaddition,the
extremehourswiththeleastrampdowncapabilityoccurduringearlymorninghours.Figure
173showsthesamedataasFigure172,butwiththeverticalscalereducedtoshowthelight
loaddownrampingcapabilityconstriction.Duringearlymorninghourstheavailableramp
downcapabilityonthelimitingdayjustmeetstherequirements.Again,nosimultaneous
extremeseveroccurredinthedata,andtherequirementisunlikelytoexceedcapability.
M
W
180
-5000
-4000
-3000
-2000
-1000
0
1000
2000
3000
4000
5000
0 6 12 18 24
Maximum of Ramp Rate Up Average of Ramp Rate Up Minimum of Ramp Rate Up
Average-3*Sigma Average L-W-S Ramp Rate Need Average+3*Sigma
Maximum of Ramp Rate Down Average of Ramp Rate Down Minimum of Ramp Rate Down
Figure 172. Committed Generation Ramp Rate Capability and Expected LoadFollowing Duty.
-1000
-800
-600
-400
-200
0
200
400
600
800
1000
0 6 12 18 24
Minimum of Ramp Rate Up
Average-3*Sigma
Average L-W-S Ramp Rate Need
Average+3*Sigma
Minimum of Ramp Rate Down
Figure 173. Expanded View of Committed Generation Ramp Rate Capability and Expected Load
Following Duty.
M
W
/
m
i
n
u
t
e
M
W
/
m
i
n
u
t
e
181
Impact of Pumps on Load-Following
Thestatistical,productionsimulationandQSSanalysesallshowedthatpumps,betheyDWR
pumpsorpumpedstoragehydropumps,canhaveasignificantimpactduringlightload
periods.Therelativeimpactofpumpswitchingandwindvariabilityonloadfollowingis
illustratedinFigure174andFigure175.
Figure174showstheeconomicdispatchfromapairofQSSsimulationsforaMaynight.The
bluetraceshowstheeconomicdispatchraise/lowersignal(MW/min)foracasewithconstant
windoutput,andtheorangetraceshowsthesamesignalforacasewithvariablewind.The
variabilityinthebluetraceisdueonlytotheload,whichincludespumpswitchingevents.The
variabilityofwind(orangetrace)increasesthefrequencywithwhichtheeconomicdispatch
signalchangessign.Figure175showsthesamepairofcases,exceptthatthepumpstepshave
beenremovedfromtheloadprofile.Thebehaviorofthetwocasesshowslessoverallvariability
duetoalessvariableloadprofile.
StatisticalmeasureswereextractedfromthissetoffourcasesandareshowninTable37.The
impactofthepumpsislargerthanthatofthewind,andtheimpactofthewindisrelativelyless
inthemorevariablecase.Afurtherstatisticalexaminationofalllightloaddataissummarized
inTable38.Undernominallightloadconditions,windvariabilityincreasesloadfollowingby
about20MW/min,whichisaboutthesameasthecontributionduetothepumpsteps.
-150
-100
-50
0
50
100
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Delta ED MW Modified Commitment, Constant Wind
Delta ED MW Modified Commitment, Variable Wind
Figure 174. Impact of Wind Variability with Pumps in Load Profile.

M
W
/
m
i
n
u
t
e
182
-100
-50
0
50
100
100 AM 200 AM 300 AM 400 AM
Delta ED MW Constant Wind, no PSH Steps
Delta ED MW Variable Wind, no PSH Steps
Figure 175. Impact of Wind Variability Without Pumps in Load Profile
Table 37. LoadFollowing Statistics of QSS Pump and Wind Sensitivity Cases
Without Pump Steps With Pump Steps
Constant
Wind
Variable
Wind
Change Constant
Wind
Variable
Wind
Change
Delta ED 14 MW 18 MW 29% 30 MW 32 MW 7%
Zero Crossings 11 21 91% 19 21 11%
Table 38. LoadFollowing Statistics of Light Load Conditions for Pump and Wind Sensitivity
Without Pump Steps With Pump Steps
Load L-W-S Change Load L-W-S Change
5-Minute in 10
th
Decile 75 MW 110 MW 47% 95 115 21%
Implied Costs of Load-Following
Theproductionsimulationsshowthateconomicoperationofthesystemresultsinaunit
commitmentanddispatchwithadequateloadfollowingcapability.Consequently,the
economicimpactofprovidingloadfollowingisbuiltintotheeconomicdispatch.
Itis,nevertheless,aninterestingexercisetopostulateaseparateloadfollowingfunctionthat
isolatestheimpactoftheintermittentrenewablesfromnormalloadfollowingandeconomic
dispatch.Thiscanbeapproximatedbyassumingthatallincrementalloadfollowing
requirementswillbeimposedontheregulationmarket.
Thestatisticalanalysisshowedthattheyearroundincrementalloadfollowingrequirementis
43MWperfiveminutes(i.e.3=3x14.2MW/5minutes).Ifthisincrementalloadfollowingis
assignedtoregulation,thenadditionalregulationcapabilitymustbeprocuredtocover5
minutesofincrementalloadfollowing.Thus,43MWofupregulationanddownregulation
mustbeprocured.Athistoricalaverageregulationprices(fromCaliforniaISO),thisextra
procurementcosts$18.5Mperyearor48/MWh,i.e.43MWx($28/MWperhourup+$21/MW
perhourdown)x8760hr/year=$18.5M.
M
W
/
m
i
n
u
t
e
183
Asnoted,lightloadoperatingconditionspotentiallypresentthemostchallengingoperating
condition.Iftheincrementalregulationrequirementforonlythelightload(i.e.,10thdecile)is
considered,theimpliedcostsareless.Duringlightload,theincrementalregulationrequirement
is60MW/5minutes.Procurementoftheextraregulationforonlythislightestloadperiod
wouldcost$2.5Mperyear.Thecalculationis(3x19.8)x($28+$21)x8760X0.1=$2.5M,which
worksoutto40/MWhofwindpowerproducedduringthe10
th
decile.
Anothermechanismtoprovideoperationalflexibilityisselectivecurtailment.Althoughthe
statisticalandproductionsimulationresultssuggestthatcurtailmentisunlikelytoeverbe
necessaryforaneconomicallyoperatedsystem,curtailmentwouldtendtoresultinwind
energylossduringperiodsoflowspotprice.Thecostimplicationsofsuchacurtailmentcanbe
estimated.Forexample,a5%curtailmentduringallminimumloadperiodswouldresultin
~300,000MWhoflostwindproduction.Sincetheaveragespotpriceis~$23/MWh,thisresults
inanannualcostofabout$7M,or18/MWhofwindproduction.
6.3.4. Regulation
Theimpactofpumpsonregulation,theimpactofintermittentrenewablesonCPS2,andthe
impliedcostsofregulationarediscussedinthefollowingsections.
Impact of Pumps on Regulation
Theprevioussectioncomparedtheimpactofpumpsandintermittentrenewablesonload
followinginlightloadperiods.Therelativeimpactofpumpswitchingandwindvariabilityon
regulationisillustratedinFigure176andFigure177.
Figure176showstheregulationdutyfromapairofQSSsimulationsforaMaynight.Theblue
traceshowstheregulation(MW)foracasewithconstantwindoutput,andtheorangetrace
showsthesamesignalforacasewithvariablewind.Theimpactofthewindvariabilityis
visible,butrelativelysmallcomparedtothepumpswitchingeventswhicharethethreelarge
excursionsatabout1:30,2:30and3:15.Figure177showsthesamepairofcases,exceptthatthe
pumpstepshavebeenremovedfromtheloadprofile.Theimpactofthewindismorevisible.
StatisticsfortheseQSScasesaresummarizedinTable39.Afurtherstatisticalexaminationofall
thelightloaddataissummarizedinTable40.Undernominallightloadconditions,wind
variabilityincreasesoverallregulationbyabout10%,aboutthesameasthevariabilitydueto
switchingpumps.
184
-600
-400
-200
0
200
400
600
1:00 AM 2:00 AM 3:00 AM 4:00 AM
Preg Modified Commitment, Constant Wind
Preg Modified Commitment, Variable Wind
Reg Up
Reg Down
Figure 176. May Night: Impact of Wind Variability with Pump Steps
-400
-300
-200
-100
0
100
200
300
400
Preg Constant Wind, no PSH Steps
Preg Variable Wind, no PSH Steps
Reg Up
Reg Down
Figure 177. May Night: Impact of Wind Variability Without Pump Steps
Table 39. Regulation Statistics of QSS Pump and Wind Sensitivity Cases
Without PSH Steps With PSH Steps
Constant
Wind
Variable
Wind
Change Constant
Wind
Variable
Wind
Change
Preg 60 68 13% 111 114 3%
Table 40. Regulation Statistics of Light Load Conditions for Pump and Wind Sensitivity
Without PSH Steps With PSH Steps
Load L-W-S Change Load L-W-S Change
1-Minute in 10
th
Decile 33.7 40.9 22% 38.6 42.3 10%

M
W
M
W
185
CPS2 Discussion
TheNERCControlPerformanceStandard2(CPS2)istheUSindustrystandardmetricfor
determiningcontrolareaadherencetoscheduledinterchangeandfrequency.Inbroadterms,
CPS2allocatesresponsibilityforregulatingfrequencyandinterchangeaccordingtothesizeofa
controlarea[9].CPS2isahighlynonlinearfunction,whichallowsdeviationwithinaband
aroundtheschedule.Largercontrolareasareallowedgreaterdeviationfromschedule,withthe
understandingthattheybearagreaterburdenandresponsibilitytoparticipateinthesecure
andstableoperationofthesystem.
CPS2,givenin%,isnormallycompiledonamonthlybasis.ACPS2performanceof100%
meansthattheareacontrolerror(ACE)neverwentoutsidetheallowedband.NERCcriteria
requireCPS2performanceofatleast90%.InmostsystemsCPS2ishigherthan90%,i.e.,
performanceisbetter.
Fundamentally,CPS2adherencerequiresadequateregulationcapabilitytoaccommodatefast
netloadvariabilityandmeettheinterchangeschedule.Forinstance,lowloadvariability,
sufficientavailableregulationandnoschedulechangeswillensurecompliance.Bycontrast,
highloadvariabilitywithinsufficientavailableregulationorsignificantschedulechangeswill
violatecriteria.
Fourassumptionsarerequiredtoperformastatisticalestimationoftheimpactofincreased
variabilityonCPSperformance.Theyare:
AllCPS2violationsareduetoloadornetloadvariations
Loadandnetloadvariabilitydistributionsarenormal
Interchangescheduleremainsfixed
Regulationresourcesandstrategyareunchanged
Withtheseassumptions,CPS2performancecorrespondstoasymmetricconfidenceintervalin
thestandardnormaldistributionofloadonlyvariability.Forexample,90%CPS2performance
correspondstoa90%symmetricconfidenceinterval.ByapplyingChebyshevsTheoremitis
thenpossibletocalculatethenarrowingoftheconfidenceintervalassociatedwiththemore
variablenetloaddistribution.ThisprovidesanestimateofthenewCPS2performance.
Throughoutthisstudy,the1minutedeltawasusedasameasureofthefastvariability.
Therefore,thestatisticaldistributionsofthe1minutedeltasarereasonableproxiesfor
anticipatedchangeinCPS2.
Fortheprojectedrenewablesgrowth(2010T),theincreasein1minutedelta dueto
intermittentrenewablesis1.6MW/min(from49.1MW/minto50.7MW/min).Thechangein
expectedCPS2performanceforthisscenarioisasfollows:
90%CPS2wouldbeexpectedtodeclineto88.9%
95%CPS2wouldbeexpectedtodeclineto94.2%.
186
Fortheartificiallyaccelerated2010Xscenario,theincreasein1minutedeltadueto
intermittentrenewablesis3.3MW/min(from49.1MW/minto52.4MW/min).Thechangein
expectedCPS2performanceforthisscenarioisasfollows:
90%CPS2wouldbeexpectedtodeclineto87.7%
95%CPS2wouldbeexpectedtodeclineto93.3%.
Therefore,CPS2performancewouldbeexpectedtodeclineapproximately1%to2%duetothe
increaseinfastvariabilitywithoutadditionalregulation.
IftheexistingCPS2isatleast92.2%,thennoadditionalregulationisrequiredtomeetthe90%
criteriaunderthe2010Xscenario.IftheexistingCPS2isatleast91.1%,noadditionalregulation
isrequiredtomeetthe90%criteriaforthe2010Tscenario.
SomesystemsareknowntoholdtheirCPS2performancelevelsfarabovethatrequiredby
NERCoperatingstandards.However,thereisasignificantoperationalcostassociatedwith
maintainingahigherthanrequiredlevelofperformance.Theneedtoincursuchoperatingcosts
shouldbeexamined.Further,CPS2violationsdrivenbyhourlyschedulechangescouldbe
reducedbymodifyingschedulingpractice.
Implied Costs of Regulation
Thestatisticalanalysisshowsanincreaseof20MWinregulationrequirement.Asnotedin
Section6.1,thisissmallcomparedtotherangeofregulationregularlyprocured,whichis
roughly300MWto600MW.Theaveragecostofregulation,perCaliforniaISOdata,is
$28/MWupand$21/MWdown.Thus,thecosttoprocureoneMWyrofupregulationisabout
$245,000,andoneMWyrofdownregulationisabout$184,000.Toprocureanadditional
20MWineachdirectionwouldcostatotalof$8.6M/year,or22/MWhofintermittent
renewableenergy.
6.4. Mitigation Methods
Inthissection,selectedmitigationoptionsareexaminedfurther.Thesemitigationoptionsare
primarilyfocusedontheadverseimplicationsofvariabilityinagiventimeframe.
6.4.1. Unit Commitment and Schedule Flexibility
TheQSSMayexample(Section5.2.2)showedthatchangingthecommitmentbysubstituting
maneuverableunitsforfixeddispatchunitswouldincreaseavailablerange.Windcurtailment
wasalsoshowntoincreaseavailablerange.However,therearecosttradeoffsbetweenthe
modifiedcommitmentandwindcurtailmentmitigationmethods.Decommittingabaseload
unitmaymeanthatitwillbeunavailableduringthenexthighloadperiodandbeyond.Inthat
case,shorttermwindcurtailmentwillprobablybethelowertotalcostoption.Conversely,de
committingbaseloadunitsmaybemorecosteffectiveifthecombinedloadandrenewable
forecastindicatesanextendedperiodofsignificantwindenergycurtailment.
Providingdeeperrunbackcapabilityalsomitigatesthelightloadmaneuverabilityproblem,and
eliminatesthecurtailment/decommitmentdecision.Generatorsrealizefurtherbenefitsby
avoidingstart/stopcosts.AsnotedinSection5.2.2,combinedcyclepowerplantsmaypresent
187
anopportunityfordeeperrunback.Similarly,energystoragereducestheneedforother
mitigationmethods.Somestoragetechnologiesmayalsoprovidebenefitsinothertimeframes.
Theuseofpumpedstorageincreasedwhenconventionalhydroflexibilitydecreased.Itis
possiblethatsomevariationwithinthedaycanbeaccommodatedwithgasstorageinpipelines.
Also,shorttermmodificationofinterchangeschedulewouldprovidesimilarbenefits.
Theneedforthesemitigationmethodsdropsasloadincreases(Section6.2).
6.4.2. Load-Following
TheQSSJuneexample(Section5.2.3)showedthatimposingshorttermwindcurtailmentwitha
ratelimitonrecoveryrelievestemporarydepletionoframpdowncapability.TheQSSMay
nightexampleshowedthatcurtailmentcanincreaseavailablerampcapability.Similarly,
energystoragecanincreaseavailablerampcapability.Variablespeedpumpedstoragecan
providerampcapabilityduringpumping.Addingloads,e.g.increasedparticipationof
controlledpumploads,hassimilarbenefits.
6.4.3. Regulation
Thestatisticalanalysisshowedanincreaseinregulationrequirementandtheproduction
simulationanalysisshoweda1MWhincreaseingasturbinegenerationper20MWhofwind
andsolarenergy.Sincegasturbine(GT)usageislikelytoincreasewithsignificantlevelsof
intermittentgeneration,theymaybeabletoprovideregulationandloadfollowingservicesas
well.
Moderngasturbineshaveaminimumramprateofabout10%MW/minutefromacoldstart.
Theyarealsoextremelyflexible,withanoperatingrangeof20%to100%ofnameplate.By
contrast,atypicalexistinggasturbinehasanoperatingrangeofabout50100%ofnameplate.
Table35showsa1minute3increaseof10MW/minute(i.e.,regulation),anda5minute3
increaseof43MW/5minutes(i.e.,loadfollowing)fortheaccelerated2010Xscenario.
Therefore,100MWofnewGTwouldcoverthesystemwideincreaseinregulationforthe
2010Xscenario,i.e.,10MW/minregulationneed/10%MW/minuteramprate=100MW.And,
54MWofnewGTwouldcoverthesystemwideincreaseinloadfollowingfor2010X,i.e.,
43MWloadfollowingneed/80%MWoperatingrange=54MW.
Thus,about200MWofnewGTshouldmeetalladditionalregulationandloadfollowing
requirementsforallstudyscenarios,providedthatthehourlyoperationflexibilityrequirements
aremetbyothermeans.
188
7.0 Conclusions and Recommendations
Fourscenarioswithincreasinglevelsofintermittentrenewablegenerationwereevaluatedin
thisstudy:2006(2,100MWwind,330MWofsolar),2010T(7,500MWwind,1,900MWsolar),
2010X(12,500MWwind,2,600MWsolar),and2020(12,700MWwind,6,000MWsolar).The
2010Tand2020studyscenariosrepresenttwostepsonanexpectedtrajectorytomeet
Californiasrenewablegenerationgoal(Section2.1).Theartificiallyaccelerated2010Xscenario
wasdevelopedtoincreasesystemstressandrepresentsthemostchallengingstudycondition.
However,theobservations,conclusionsandrecommendationspresentedinthissectionapply
toallscenarios,notjustthemostchallenging.Theyareintendedtoenableconsistent,sustained
renewablegrowththrough2020.
7.1. Observations by Time Frame
Theanalyticalworkpresentedthroughoutthisreportproducedextensivequantitativeresults.
Inthissection,alargelyqualitativesynopsisofthekeyfindingsispresentedtoprovidecontext
fortheconclusionsandrecommendations.
7.1.1. Day-Ahead and Overall Operation
Intermittentrenewablegenerationwilldisplaceothermoreexpensivegeneration,i.e.,
generationwithahighermarginalvariablecost.Sincenaturalgascombinedcyclepowerplants
arefrequentlyonthemargininWECC,mostofthedisplacedgenerationiscombinedcycle.
RoughlyhalfofthedisplacedgenerationisinsideCaliforniaandtheotherhalfoutside(Section
4.4).Thedisplacementofoutofstategenerationbylowermarginalcostinstaterenewablesis
aneconomicbenefitoftheaddedrenewables.Thisdisplacementisnotanexportofwind
variabilitytoneighboringsystems.
Conventionalhydroelectricfacilityoperationwithinthestateistemporallyshiftedduetowind
andsolargeneration.However,thechangeinoperationislessthan1,000MWmorethan90%of
thetime,onahydrosystemwithover9,000MWofcapacity(Sections4.4,4.4.1).The
conventionalhydrofacilitiesshouldbeabletoprovidethismaneuverability,particularlywhen
augmentedbytheavailablepumpingloads.
Dayaheadoperationswillbelesscertainasintermittentresourcesincrease.Totalloadand
intermittentrenewableforecasterrorswillberoughlytwicethatoftheloadforecasterroralone.
Thismayincreasetheoperationofpeakinggenerationwhenothergenerationisunder
committedduetooverforecastingofintermittentrenewables.Theincreaseduseofpeaking
combustionturbinesoffsetssomeoftheeconomicvalueoftheintermittentrenewables.This
uncertaintywasincludedintheanalysis.Systemperformancebasedoncurrentlyavailableload
dataandwindforecastingtechnologywassatisfactory(Sections3.5.1,4.7).Substantialeconomic
benefitwillberealizedwiththeuseofdayaheadwindandsolarforecastinginunit
commitmentandscheduling(Sections4.3,4.7).
7.1.2. Hourly Schedule Flexibility
Therequirementsforhourlyscheduleflexibilityincreaseovertimeduetobothsystemload
growthandadditionalintermittentrenewables.Threetimesthestandarddeviationofonehour
189
change,whichwastheprimarymetricofrequiredscheduleflexibility,showstherelative
impactofthesetwocomponents.Theincreaseinhourlyvariabilityduetoloadissubstantially
greaterthantheimpactofwindandsolarfortheexpectedintermittentrenewablescenarios
(2010T,2020).Theloadgrowththrough2020continuallyincreasesthescheduleflexibility
requirement.Theincrementalrequirementduetoloadgrowthfrom2006to2020isabout1,700
MW/hr.Theintermittentrenewablesuniformlyincreasethescheduleflexibilityrequirement
abovetheloadalonerequirementbyabout130MW/hrforthesescenarios(Sections3.2,6.2.1).
Acomparisonbetweenthe2006andtheartificiallystressed2010Xscenariosshowstheincrease
inscheduleflexibilityrequirementduetoloadaloneisaboutequaltothatduetothe
acceleratedadditionofintermittentrenewables(Section6.2.1).
Theadditionofzeromarginalcostintermittentrenewableswilldisplaceothergenerationwith
highermarginaloperatingcosts.Thismeansthatthecommitmentanddispatchoftheother
generationresourcesnecessarytoprovideoperationalflexibilitywillchange.Ingeneral,the
abilitytodispatchdownatlightloadandtodispatchupatheavyloadrepresentthelimiting
conditions(Section4.4.3).
Range(i.e.,theremainingcapacity(MW)availablebetweenthecurrentoperatingpointand
eitherthemaximumorminimum)measurestheabilityofthebalanceofportfolio(i.e.,non
renewable)generatorstorespondtochangingload,windandsolarconditionsinthehourly
timeframe.Theanalysisoftheavailablerangetodispatchupatpeakloadshowedno
limitations.Theavailablerangetodispatchdownatlightloadwasalsoadequate.However,
limitationsmaybeencounteredwithcoincidentminimumload,highwindgenerationandlow
conventionalhydroflexibility.Conventionalandpumpedstoragehydrogenerationaswellas
pumploadsplayanimportantroleinprovidingthenecessaryscheduleflexibility(Section
6.3.2).
Duringdailyrealtimeoperation,theincrementalhouraheaduncertaintyduetointermittent
renewablesismuchlessthandayaheadvalues.Thecombinedhouraheadforecasterroris
about20%greaterthanthatforloadalone(Section3.5.2).
Theanalysisfoundthatarational,i.e.,leastcost,dispatchandcommitmentofavailable
resourcesresultsinsatisfactoryoperationinthistimeframe(Section4.7).
7.1.3. 5-Minute Load Following and Economic Dispatch
Anexaminationofthechangeinthestandarddeviationsinthe5minutetimeframeshowsthe
relativeimpactofloadgrowthandadditionalintermittentrenewablesontheloadfollowing
requirements.Loadgrowthwillincreasetheloadfollowingrequirementabout10%by2010and
about35%by2020.Theintermittentrenewableswillfurtherincreasethatrequirementby3%to
7%.Althoughtherelativeincreaseisgreateratlightload,thelightloadrequirementitselfisless
thantheoverallrequirement(Sections3.3,6.2).
Theloadfollowingcapabilityatanygivenoperatingpointisdictatedbyunitcommitmentand
dispatch.Rampratecapability(i.e.,thespeed(MW/minute)atwhichthesystemcanusethe
remainingupanddownrange)measurestheabilityofthebalanceofportfolio(i.e.,non
190
renewable)generatorstorespondtochangingload,windandsolarconditionsinthe5minute
timeframe.Theavailablerampingcapabilityofonlineunits,bothupanddown,wasfoundto
belargelyadequate.Underlightloadconditions,variousmitigationstrategies(e.g.,selective
windcurtailmentandthermalunitrecommitment)wereeffectiveinrelievingloadfollowing
limitationsshouldtheyoccur(Sections5.2.2,6.3.2).Throughouttheanalysisandresultsofthis
study,allintrahourvariabilityimpactsofintermittentrenewablesarehandledbyinstate
resources.Variabilityimpactsarenotexportedtoneighboringsystems.
7.1.4. 1-Minute Regulation
The1minutevariabilityalsoincreaseswithloadgrowth,andisfurtherincreasedbythe
additionofintermittentrenewables.Onapercentagebasis,theincreaseinregulation
requirementduetoloadgrowthandduetointermittentrenewablesaresimilartotheload
followingincreases.Unlikethehourlyandloadfollowingtimeframes,however,regulation
requirementsarerelativelyuncorrelatedtosystemloadlevel.Theincreaseinregulation
requirementduetotheintermittentrenewablesisabout3%to7%.
Insufficientloadfollowingcapabilityincreasestheneedforregulationcapability.Rapid
variationinloadaswellasintermittentrenewableproductionwillincreasetheareacontrol
error(ACE),whichalsodrivesagreateruseofregulation.AnyincreaseinACEmaydegrade
CPS2performance.Ifnochangesaremadetothepresentregulationprocurement,theimpact
onCPS2isabout2%(Sections3.3,6.3.4,5.2.3).
7.2. Conclusions
The2010Xscenarioexaminedatotalof19,800MWofrenewablesinCalifornia,including
12,500MWofwindgeneration,2,600MWofsolar,1,000MWofbiomassand3,700MWof
geothermal.Thisscenariorepresentsastressedconditiondesignedtotestthesystemwithmore
renewablesthanprojectedfor2010.
ThislevelofrenewablegenerationcanbesuccessfullyintegratedintotheCaliforniagrid
providedappropriateinfrastructure,technology,andpoliciesareinplace.Specifically,this
successfulintegrationwillrequire:
Investmentintransmission,generationandoperationsinfrastructuretosupportthe
renewableadditions,
Appropriatechangesinoperationspractice,policyandmarketstructure,
Cooperationamongallparticipants,e.g.,CaliforniaISO,investorownedutilities,
renewablegenerationdevelopersandowners,nonFERCjurisdictionalpower
suppliers,andregulatorybodies.
7.3. Recommendations
ThestudyscenariosrepresentstagesalongatrajectorytomeetCaliforniasrenewable
generationgoal.Thefollowingrecommendationsareasetoftargets,actionsandpolicies
designedtoensuresuccessfulintegrationofsignificantlevelsofintermittentrenewable
generationthrough2020.Theimplementationoftheserecommendationsshouldproceed
191
concurrentlywiththerenewablegenerationgrowth.Suchevolutionaryimprovementswill
allowsecureandeconomicintegrationatallstagesalongtherenewablegenerationgrowth
trajectory.
Thechallengeofaccommodatingsubstantialintermittentrenewablegenerationisincremental
tothechallengeofservingexistingandnewload.Longtermplanningmustalwaysconsider
requirementsforgenerationandtransmission,andstrikeanappropriatebalancebetweenthe
two.Further,newconsiderationsspecifictorenewabletechnologiesmustbeincluded.Thus,the
planningprocessmustconsiderthreemajorsystemcomponents:
Generation
Transmission
RenewableTechnology
Therecommendationspresentedbelowaregroupedaccordingly.
7.3.1. Generation Resource Adequacy
TheCEC,CPUCandCaliforniaISOhaveongoingprocessestoprovidethegeneration
infrastructurenecessarytomaintainreliableoperation.Theadditionofbothintermittentand
nondispatchablerenewableresourcestotheCaliforniagridincreasestherequirementfor
generationresourceflexibility.Itisessentialthatthisrequirementforflexibilitybeincludedin
theoverallassessmentandplanningforresourceadequacy.Itisrecommendedthatspecific
attributesofgenerationflexibilitybeinventoried,maintainedandincreased.Wherepossible,
quantitativetargetsaresuggested;othersmaybeadoptedascircumstancesandunderstanding
changes.Toavoidrepetition,specificpolicyandtechnologyrecommendationsaregrouped
withthemostrelevantperformanceissue.However,manyrecommendationscouldapplytoa
broaderrangeofperformancecategories.Further,noneoftherecommendationsareeitherself
sufficientormutuallyexclusive.Anappropriatecombinationofmeanswillbemostsuccessful.
MinimumLoadOperation.TheCaliforniagridshouldtargetacombinationofinstategenerating
resourcesandpowerexchangecapability/agreementswithneighboringsystemsthatallow
operationdowntoaminimumnetload(loadminuswindminussolar)intherangeof
18,000MWto20,000MW.Thesetargetswillmeetthelongterm(2020)needsofthesystem,and
allowforoperationwithminimalcurtailmentofintermittentrenewables(Section3.1.3).
MinimumTurndown.Generatingresourceswithlowerminimumpoweroutputlevels
providegreaterflexibility,andallowsuccessfuloperationatminimumload.New
generatingresourcesshouldbeencouragedand/orrequiredtohavethiscapability;
existinggenerationshouldbeencouragedand/orrequiredtoupgradetheircapability.
Acomparisonoftheloadandnetload(loadwindsolar)forthevariousscenarios
showsthatminimumsarelesswiththeintermittentgenerationonthesystem.The
minimumsystemturndowncapabilitywilldeterminetheamountofrenewable
generationcurtailmentthatisnecessary.Aminimumof20,000MWisexpectedtoresult
incurtailmentduringafewhundredhoursperyearfortheexpectedgrowthtrajectory
(Sections4.4.2,3.1.3).
192
DiurnalStart/Stop.Anotherwaytomeetminimumloadistoincreasetheamountof
generationthatiscapableofreliablediurnalcycling.Thiswillbenefitthesystemby
allowingthecommitmentofunitsthatareeconomicatpeakandshoulderloads,
withoutrequiringtheirnoneconomicoperationatlightload(Section4.2).
LoadParticipation.Activeparticipationbylargeloads,especiallypumps,isanotherway
toassureadequateflexibility.ThepumpscontrolledbyCaliforniaDepartmentofWater
Resourcesarealreadyparticipantsintheenergymarket,butadditionaltypesof
participationandcooperationcouldincreaseoverallsystemflexibility.Forexample,
additionalinvestmentinpumps,controlsorotherloadinfrastructuretotakeadvantage
oflightloadenergypricingcouldbebotheconomicandeffective[8],(Sections5.2.2,
6.3.4).
Californiashouldexploreothermeanstoencourageloadshiftingtowardslightload
conditions.Variousloadshiftingandstoragetechnologies,suchascoldstorage(e.g.for
buildingcoolingorinletaircoolingforgaspeakinggeneration)holdpromise,andmay
provetobeeconomic.Arrangementsthatgivethegridoperatorcontroloverloadsfora
contractualconsiderationorratereductionwillbemoreattractiveaspenetrationof
intermittentrenewablesincreases.
PumpedStorageHydro.Useofpumpedstoragehydro(PSH)facilitieswasshownto
increaseforthescenariosexamined.Theinfrastructureandpolicynecessarytoallow
optimaluseofexistingPSHwithinCaliforniashouldbeenhanced.AdditionalPSH
capabilitycouldalsoenhancesystemschedulingflexibility,andwilllikelyaidother
flexibilityattributesdiscussedbelow.Thisisparticularlytruewhenconventionalhydro
flexibilityislow,duetounusuallyhighrunoffconditions(Section4.4.4).
HourlyScheduleFlexibility.TheCaliforniagridshouldtargetacombinationofinstategenerating
resourcesthatprovideaminimumlevelofschedulingflexibility.Theanticipatedloadgrowth
to2020willdrivetheoverallsystemflexibilityneedsfromthepresentlevelofabout
4,300MW/hrtoabout6,000MW/hr.Theadditionalvariabilityanduncertaintyassociatedwith
intermittentrenewableswillincreasetheamplitudeofsustainedloadramps(bothupand
down),andthefrequencyofgenerationstartsandstops.Fortheexpectedrenewablesgrowth
trajectory(2010T,2020),theoverallhourlyflexibilityrequirementisexpectedtobeabout
130MW/hrgreaterthanthatrequiredforloadalone.Undertheartificiallyaccelerated
renewableexpansionofthe2010Xscenario,thatincrementalrequirementisabout400MW/hr
(Section6.2.1).
Duringlightloadconditions,totalrequirementsaresmallerbuttherelativeimpactof
intermittentrenewablesislarger.Theanticipatedloadgrowthto2020willdrivethelightload
systemflexibilityneedsfromthepresentlevelofabout2,000MW/hruptoabout3,000MW/hr.
Fortheexpectedrenewablesgrowthtrajectory(2010T,2020),thehourlylightloadflexibility
requirementisexpectedtobeabout1,000MW/hrgreaterthanthatrequiredforloadalone
(Section6.2.2).
193
HydroScheduling.Conventionalhydroelectricgenerationplaysakeyroleinlightload
scheduleflexibilityaswellasloadfollowingandregulation.Economicoperationwillbe
enhancedbyhighhydroflexibility.Existingflexibilityshouldbemaintainedatleast,
andinvestmentstoincreasemaneuverabilityshouldbeconsidered.Adocumented
inventoryofcapabilityisimportant.Californiashouldperiodicallyexaminetheamount
andtypeofhydroconstraints,andevaluateinvestmentsorcontractualmechanismsfor
costeffectivereliefofthoseconstraints(Sections4.4.1,4.4.4).
FasterStart/Stop.Uncertaintiesinforecastscreateasomewhatdifferentflexibility
requirement.Evenwithstateoftheartwindforecasting,bothdayaheadandhour
aheadnetloadforecastuncertaintieswillincreaseduetointermittentrenewables.With
anincreasedriskofanactualnetloadsignificantlydifferentfromtheforecastnetload,
shortnoticestart/stopcapabilityduringdailyoperationwillbeanimportantpartofthe
redispatchneededtobalancegenerationandload.TheCaliforniagridshouldtarget
sufficientinstategeneratingresourcecapabilitytomeetdayaheadforecasterrorsinthe
rangeof5,000MW,andhouraheadforecasterrorsintherangeof2,000MW.Overall,
thisrepresentsaboutdoublethepresentlevelofdayaheadloadforecasterror,and
about20percentmorethanthepresenthouraheadloadforecasterror(Sections3.5,
6.3.1).
Duringlighterloadperiods,thenetloadforecasterrormaybethreetimestheload
aloneforecasterrorinthedayaheadforecast.Thetargetsrecommendedabovewillalso
besufficientforlightloadconditions.
MultiHourScheduleFlexibility.Flexibilitytargetsshouldalsoaddressperiodsofsustainedload
increasesanddecreases.TherecommendedtargetsarefortheCaliforniagridtohaveresources
adequatetomeetamaximummorningloadincreaseof12,000MWoverthreehours,anda
maximumeveningloaddecreaseof14,000MWoverthreehours.Thisrepresentsanincreaseof
about1,000MWoverthecapabilityneededtomeettheloadalone(Sections3.2.2,6.3.1).
LoadFollowingCapability.TheCaliforniagridshouldtargetacombinationofinstategenerating
resourcesthatprovideaminimumlevelofgenerationrampingcapability,bothupanddown.
Onaverage,thesystemshouldmaintainontheorderof+/130MW/minforaminimumof5
minutes.Thisisabouta10MW/minuteincreaseovertherequirementduetoloadalone
(Sections3.4,6.2.1,6.3.3).
Duringlightloadconditions,approximately70MW/minofdownloadfollowingcapabilityare
required.Uploadfollowingrequirementsarelower.Theloadfollowingcapabilityshouldbe
subjecttoeconomicdispatchfromthesystemoperators.Loadfollowingdutyshouldnotbe
shiftedtounitsprovidingregulation.
Import/ExportScheduling.TheCaliforniagridshouldrecognizethateconomic
incorporationofsubstantialinstaterenewableswillinevitablyinvolvesignificant
displacementofimportedenergy.Regulatoryandcontractualarrangementsforimports
andexportsshouldbestructuredsuchthatthevalueofschedulingflexibilityis
recognized,allowedandappropriatelycompensated.Inparticular,Californiashould
194
allowschedulechangestooccurmorefrequentlyandattimesotherthanonthehour
(Section6.1).
RegulationCapability.TheCaliforniagridshouldtargetacombinationofinstategenerating
resourcesthatprovideaminimumlevelofregulationcapability.TheISOcurrentlyprocures
regulationintherangeof300MWto600MW.Theprocuredamountvariessubstantiallyover
allloadlevels.Theimpactofintermittentrenewablesonregulation(20MW)isconsiderablyless
thanthenormalvariabilityintheamountprocured.However,regulationresourceswill
continuetobeimportant.Therefore,theCaliforniagridshouldatleastmaintainthecurrent
levelofregulationcapability.Thislevelofregulationshouldallowthestatetocontinueto
satisfytheirregulatoryobligationsforinterchangeandfrequencycontrol,i.e.NERCCPS2
performance.CPS2performanceshouldbecontinuallyscrutinizedasintermittentrenewables
areaddedtothegridtorefineregulationrequirementsandprocurement(Sections6.1,3.4,
6.3.4).
RegulationTechnologies.Californiashouldconsidertheuseoftechnologiesbeyond
conventionalgenerationtoprovideregulation.Theearlierdiscussionaboutload
participationinscheduleflexibilityapplieshereaswell.Functionalrequirementsfor
loadstoprovideregulationaredifferentfromthoseforgeneration.Givenasuitable
regulatoryandmarketstructure,however,itislikelythatothertechnologiesand
participantswillemergetoprovidetherequiredservices.Examplesincludesometypes
ofstoragetechnology,suchasvariablespeedpumpedhydroandthelatestflywheel
energystoragesystems.Policyandmarketstructureshouldencouragediversityof
participantsinprovidingancillaryservices,andtechnicalspecificationsforperformance
shouldbesufficientlyflexibletoallowtheintroductionofnewtechnologies.
NonTechnicalResourceAdequacyConsiderations.Theprecedingrecommendationswereaimedat
securingthetechnicalcapabilitiesnecessaryforsuccessfulintegrationofintermittent
renewables.Thefollowingitemsaddresspolicyandcommercialconsiderations.
MarketDesign.Itmustberecognizedthatwhileoperationalflexibilityisofconsiderable
valuetothegrid,itcurrentlyholdslittleattractionforpowersuppliers.Deeper
turnback,morerapidcyclingandloadfollowing,andmorefrequentstartsandstopsall
imposesignificantcostsandrevenuereductionsonthesuppliers.Marketand
regulatorystructuresmustrecognizethevalueoftheseflexibilityfeatures.Policy
changesmayincludeacombinationofexpandedancillaryservicesmarkets,incentives,
andmandates.
ContractualObligations.Muchoftheanalysispresentedinthisreportisbasedonthe
presumptionthatthegridisoperatedinarationalfashionthatis,theavailable
generationresourcesareusedasefficientlyandeconomicallyaspossible.Theanalysis
didnotincludehistoricalconstraints(i.e.longtermcontractualobligations)thatforce
thesystemtorunlessefficientlythanpossible.Newcontractsunderconsideration,
existinglongtermcontractsupforrenewal,orindeedanyexistingcontractsthatcould
berenegotiatedshouldbereviewedwithalloftheprecedingresourceadequacy
195
recommendationsinmind.TheCaliforniagridmustmaintainoperationalflexibility,
andtodoso,itmusthavenotonlythephysicalresourcesnecessary,butalsothe
businessandcontractualarrangementsnecessarytoenabletherationaluseofthose
physicalresources.
Retirements.Generatingplantretirementsthatwerefirmlyscheduledwhenthe
databaseswereassembledwereincorporatedintothisstudy.However,increased
competitionfromnewresources,renewableorotherwise,willtendtopushmarginally
profitablegeneratingresourcesoutofbusiness.Suchspeculative,economicretirements
werenotconsideredinthestudy.Successfulimplementationoftherecommendations
abovewillensurethatresourceswiththenecessaryflexibilityareavailable.Inaddition,
itisrecommendedthatretirementsbeprojected,monitored,andevaluatedduringthe
resourceplanningprocess.
Inventory.Duringthisstudy,itwasnotedthatgeneratorcharacteristicsandcapabilities
(e.g.,minimumturndown,rampratecapability)werenotalwaysknownwithsufficient
detailorcertainty.Somedegreeofuncertaintyisinevitable.However,withthe
increasedneedforresourceflexibility,Californiashouldimplementaprogramto
measure,verify,andcataloguetheflexibilitycharacteristicsofthegenerationresources.
AprogramsimilartotheWECCgeneratordynamictestingmightprovesuitable.
7.3.2. Transmission Infrastructure
TheadditionofthousandsofMWofnewgenerationofanyvarietywillrequireexpansionof
thetransmissionsystem.Thisstudyincludedtheadditionofsufficientbulktransmission
necessaryforconnectionofthenewrenewablestothegridasdeterminedanddocumentedby
DPC[6].Howeveritwasnotadetailedtransmissionstudyandisnotasubstituteforone.
Policymakersmustrecognizethatlocalproblemsmightdevelop,andenablethenecessary
transmissionadditions.Practiceandpolicythatcorrectproblemsandstrikeabalancebetween
infrastructureinvestmentandcongestionarenecessary.Toanappreciableextent,this
observationholdsforalltransmissionplanningandallgenerationadditions.Californiacan
economicallybenefitfromchangesinplanningandoperationofthetransmissioninfrastructure
byrecognizingthelocationalandvariablenatureofintermittentrenewables.Thefollowingare
recommendationsthatarespecifictotheseneeds(Section4.6).
ExistingConstraints.Californiahasexistinginfrastructurethatcontributessubstantiallytothe
secureandeconomicoperationofthegridwithhighlevelsofintermittentrenewables.Insome
circumstances,theuseofthatinfrastructureforsystemwidebenefitisconstrainedbylocal
transmissionlimitations.OneexampleofsuchaconstraintistheoccasionalinabilityofHelms
pumpedstoragehydrotoreachfullpumpingpower(Section4.4.4).Planningandpolicyshould
recognizeandenablecorrectionofsuchlocallimitations.
RatingCriteria.Windgenerationisvariableandthespatialdiversitybetweenmultipleplants
substantiallyimpactsthecoincidentproductionofpowerfromthoseplants.Clearly,an
individualwindplantwillreachratedoutputformanyhoursperyear.Thus,normalplanning
196
criteriarequiressufficientcapability(i.e.,thermalrating)onthetransmissioninterconnection
dedicatedtothatplanttoaccommodateratedpoweroutput.
However,asmorewindplantsvieforaccesstospecifictransmissioncorridors,itwillbe
progressivelylesslikelythatallwindplantswillsimultaneouslyreachtheirmaximumoutput.
Notethatinthreeyearsofdata,allwindplantsinthisstudyneversimultaneouslyreached
maximumoutput.And,the12,500MWofwindgenerationexceeded10,000MWofproduction
lessthan1%ofthetime.Thus,transmissionplanningtoaccommodatemultiplewindplants
shouldconsidertheirspatialdiversityandthestatisticalexpectationofsimultaneoushigh
poweroutputlevels.Plantsincloseproximitywillgenerallyrequiretransmissioncapability
equivalenttotheaggregateratingoftheplants.Plantsthatarefartherapartmayrequireless
transmissioncapability.Hence,itisnotnecessarytoguaranteesufficientratingonthebulk
transmissioninfrastructuretoaccommodateallwindprojectsatfulloutput.Existingcriteria
shouldbesufficienttoprovidethisplanningflexibility.
Technology.Policyshouldrewardinvestmentintechnologytomaximizeuseoftransmission
infrastructureforrenewables.Suchpoliciesshouldrecognizethatwindgenerationisa
relativelypoorresourceforcapacityandthatcreativeuseoftechnologymayoptimizeuseof
transmission.Regulatoryandcontractualpracticeshouldallowtechnologiessuchasrealtime
lineratings,controlsthatmanageoutputfrommultipleintermittentrenewableresources,local
shorttermforecasting,andothernonstandardapproachestobalancerenewableenergy
deliverywithtransmissioninfrastructurecosts.
7.3.3. Renewable Generation Technology, Policy, and Practice
Withsignificantlevelsofintermittentrenewablegeneration,operationmaybechallengingat
extremelylightloadlevels,underaconstrainedtransmissiongrid,orwithhighwindvolatility.
Undertheseconditions,renewablegenerationmustparticipateinoverallgridcontrol.The
followingrecommendationsarespecifictorenewabletechnology,andareaimedatassuring
thatintermittentrenewablesplayanactiveandpositiveroleinthesecureandeconomic
operationofthegrid.
Curtailment.Undertherareoccasionsofcoincidentminimumload,highwindgenerationand
lowconventionalhydroflexibility,itmustbepossibletocurtailintermittentrenewables.The
gridoperatorshouldhavetheabilitytoordersuchareductioninproduction.Regulatoryand
contractualarrangementsforintermittentrenewablesshouldbestructuredsuchthat
curtailmentsarerecognized,allowedandappropriatelycompensated.Rampratecontrolscould
alsobeconsidered(Section5.2.2).
AncillaryServices.Intermittentrenewablesmaybeabletoprovideancillaryservicesthatare
bothvaluableandeconomicundersomeoperatingconditions.Forexample,windgeneration
canprovidefrequencyregulation.Suchfunctionalityisarequirementinsomeregions[10].
Regulatoryandcontractualarrangementsforintermittentrenewablesshouldbestructuredsuch
thatprovidingsuchservicesarerecognized,allowedandappropriatelycompensated.
Forecasting.SuccessfulandeconomicoperationoftheCaliforniagridrequireswindandsolar
forecasting.Thisstudyverifiedsubstantialbenefitsfromtheuseofstateoftheartdayahead
197
forecastingintheunitcommitmentprocess.Substantialbenefitsareexpectedforimprovements
inbothlongerterm(multiday)andshortterm(hoursandminutesahead)forecasting.
Investmentandpolicymustencouragedevelopmentofhighfidelityintermittentrenewable
forecastingforallintermittentrenewablegenerationinthestate(Sections4.3,6.3.1).
Monitoring.Thewindproductionprofilesusedinthisstudyarebasedonhistoricalweather
dataandsophisticatedcomputermodels.Recordeddatafromrealoperatingexperiencewillbe
invaluableinrefiningoperatingpractice,performanceandflexibilityrequirements.Time
synchronizedproductionandmeteorologicaldatafrommanyplantswillprovidevalidationor
correctionofthetrendsandresultspredictedbythisstudy.Theywillshowthebenefitsand
limitationsofspatialdiversity,mesoscalemodeling,andvariouswindplantcontrols.Itis
recommendedthatCaliforniacontinueandexpand,asnecessary,programstomonitor,analyze
anddisseminateperformanceinformationrelevanttogridoperationsandplanningfor
intermittentrenewables.
7.4. Closure
ThetargetedlevelsofrenewablegenerationcanbesuccessfullyintegratedintotheCalifornia
gridprovidedappropriateinfrastructure,technology,andpoliciesareinplace.
198
8.0 References
[1] CaliforniaEnergyCommission:2005EnergyActionPlanII,2005,
http://www.energy.ca.gov/energy_action_plan/index.html.
[2] CaliforniaEnergyCommission:2005IntegratedEnergyPolicyReport,2005,
http://www.energy.ca.gov/2005_energypolicy/documents.
[3] CaliforniaWindEnergyCollaborative(CWEC):CaliforniaRPSRenewableGeneration
IntegrationCostAnalysisMultiyearReport,CaliforniaEnergyCommissionConsultant
ReportCEC5002006064,June2006,http://energy.ca.gov/2006publications/CEC5002006
064/CEC5002006064.PDF.
[4] ConsortiumforElectricityReliabilityTechnologySolutions(CERTS)andElectricPower
Group,LLC:AssessmentofReliabilityandOperationalIssuesforIntegrationofRenewable
Generation,CaliforniaEnergyCommissionConsultantReportCEC7002005009D,2005,
http://energy.ca.gov/2005publications/CEC7002005009/CEC7002005009D.PDF
[5] DavisPowerConsultants:StrategicValueAnalysisforIntegratingRenewableTechnologiesin
MeetingTargetRenewablePenetration,CaliforniaEnergyCommissionConsultantReport
CEC5002005106,IEPR2005Proceedings,July1,2005,
http://www.energy.ca.gov/2005_energypolicy/documents/2005_index.html#070105
[6] DavisPowerConsultants,IntermittencyImpactsofWindandSolarResourcesonTransmission
Reliability,CaliforniaEnergyCommissionConsultantReport
[7] AWSTruewind,CharacterizingNewWindResourcesInCalifornia,CaliforniaEnergy
CommissionConsultantReport
[8] Kirby,B.J.,JohnD.Kueck,SpinningReservefromPumpLoad:AReporttotheCalifornia
DepartmentofWaterResources,ORNL/TM2003/99,OakRidgeNationalLaboratory,
April2003.
[9] NERCPerformanceStandardsReferenceDocument,version2,November21,2002.
[10] ESBNationalGrid,GridCode,version1.2,May2005,www.eirgrid.com
199
Glossary
ACE Areacontrolerror
AGC Automatedgeneratorcontrol
AMWCO aggregatemegawattcontingencyoverload
ATC Availabletransfercapability
CABPS CaliforniaBulkPowerStorage
CaliforniaISO CaliforniaIndependentSystemOperator
CERTS ConsortiumforElectricReliabilityTechnologySolutions
COI CaliforniaOregonIntertie
CO2 Carbondioxide
CPS2 ControlPerformanceStandard2
CPUC CaliforniaPublicUtilitiesCommission
CS Concentratingsolar
CSP concentratedsolarpower
CWEC CaliforniaWindEnergyCollaborative
DPC DavisPowerConsultants
DWR DepartmentofWaterandPower
EAO ElectricityAnalysisOffice
EnergyCommission CaliforniaEnergyCommission
FERC FederalEnergyRegulatoryCommission
GE GeneralElectricEnergyConsulting
GEMAPS GeneralElectricsMultiAreaProductionSimulation
GW gigawatt
IAP IntermittencyAnalysisProject
ICAP installedcapacity
IEPR IntegratedEnergyPolicyReport
IID ImperialIrrigationDistrict
200
kV kilovolt
LADWP LosAngelesDepartmentofWaterandPower
LCOE Levelizedcostofenergy
LVRT Lowvoltageridethrough
m/s meterspersecond
MVA megavoltamperes
MVAR megavoltamperesreactive
MW megawatt
NERC NorthAmericanElectricReliabilityCouncil
NOx oxidesofnitrogen
NREL NationalRenewableEnergyLaboratory
OASIS OpenAccessSameTimeInformationSystem
PG&E PacificGasandElectricCompany
PIER PublicInterestEnergyResearch
PSH Pumpedstoragehydro
PSLF PositiveSequenceLoadFlow
PSS powersystemstabilizer
PV photovoltaic
QSS quasisteadystate
RA rollingaverage
RAS Remedialactionschemes
RD&D research,developmentanddemonstration
RPS RenewablesPortfolioStandard
RTBR RenewableTransmissionBenefitRatio
SCE SouthernCaliforniaEdisonCompany
SDG&E SanDiegoGasandElectricCompany
SGIP SelfGenerationIncentiveProgram
SMUD SacramentoMunicipalUtilityDistrict
201
SOx oxidesofsulfur
SVA StrategicValueAnalysis
UCAP uniformcapacity
VAR voltageamperereactive
VReg voltageregulation
WECC WesternElectricityCoordinatingCouncil
w/m
2
wattspersquaremeter
WTRL WeightedTransmissionLoadingRelieffactor

202
203
Appendix A. Summary of Wind Projects by Scenario
Thewindprojectsincludedineachstudyscenarioareshowninthefollowingtables.Eachtable
identifiesaprojectbyitspowerflowbusnumber,busname,andvoltagelevelinthefirstthree
columns.Thefourthcolumnshowstheratedpoweroutputforanindividualproject,andthe
finalcolumnshowsthesitenumberfromtheAWSwindprofiledata.
Table41liststhewindprojectsincludedinthe2006studyscenario.Table42liststhe
incrementalprojectsaddedtothe2006scenariotocreatethe2010Tscenario.Table43liststhe
incrementalprojectsaddedtothe2010Tscenariotocreatethe2010Xscenario.Table44liststhe
incrementalprojectsaddedtothe2010Xscenariotocreatethe2020scenario.
Table 41. Wind Projects Included in 2006 Study Scenario.
Power Flow Data
Bus # Bus Name kV Total MW Rating AWS Site #
24009 APPGEN1G 13.8 55 14
24010 APPGEN2G 13.8 55 7
24136 SEAWEST 230 263 1
24152 VESTAL 66 8 3
24152 VESTAL 66 50 8
24422 PALMDALE 66 1 6
24436 GOLDTOWN 66 13 2
24457 ARBWIND 66 22 9
24458 ENCANWND 66 113 4
24459 FLOWIND 66 41 34
24460 DUTCHWND 66 14 5
24465 MORWIND 66 56 32
24826 INDIGO 115 21 23
24914 MTNVIEW1 13.8 63 29
24915 MTNVIEW2 13.8 63 30
25632 TERAWND 115 23 22
25633 CAPWIND 115 20 18
25634 BUCKWND 115 21 20
25635 ALTWIND 115 50 17
25636 RENWIND 115 13 31
25637 TRANWND 115 60 27
25639 SEAWIND 115 27 21
25640 PANAERO 115 30 28
25645 VENWIND 115 45 16
25646 SANWIND 115 28 19
28501 MIDWIND 12 18 13
28502 SOUTHWND 12 35 10
28503 NORTHWND 12 19 12
28504 ZONDWND1 12 26 15
28506 BREEZE1 12 13 11
32168 USWINDPW 9.11 50 25
204
Table 41 (continued). Wind Projects Included in 2006 Study Scenario.
Power Flow Data
Bus # Bus Name kV Total MW Rating AWS Site #
32172 HIGHWNDS 34.5 160 24
33170 WINDMSTR 9.11 38 37
33171 TRSVQ+NW 9.11 28 38
33175 ALTAMONT 9.11 13 39
33175 ALTAMONT 9.11 16 40
33834 KALINA 9.11 7 41
33836 USWP_#4 9.11 24 42
33836 USWP_#4 9.11 41 43
33836 USWP_#4 9.11 17 44
33838 USWP_#3 9.11 77 45
33840 FLOWD3-6 9.11 19 26
33840 FLOWD3-6 9.11 19 46
33840 FLOWD3-6 9.11 19 47
33840 FLOWD3-6 9.11 19 48
33842 PATTERSN 9.11 2 50
33842 PATTERSN 9.11 30 51
33842 PATTERSN 9.11 70 52
34342 INT.TURB 9.11 13 33
35310 LFC FIN+ 9.11 22 35
35312 SEAWESTF 9.11 13 36
35314 WALKER+ 9.11 100 53
35316 ZOND SYS 9.11 20 54
35318 FLOWDPTR 9.11 9 55
35320 USW FRIC 12 10 56
35320 USW FRIC 12 5 57
Table 42. Incremental Wind Projects Added for 2010T Study Scenario.
Power Flow Data
Bus # Bus Name kV Added MW Rating AWS Site #
24056 ETIWANDA 230 168 59
24520 TEHACHPI 500 500 61
24520 TEHACHPI 500 500 62
24520 TEHACHPI 500 105 82
24520 TEHACHPI 500 100 83
24520 TEHACHPI 500 108 84
24520 TEHACHPI 500 100 85
24520 TEHACHPI 500 150 86
24520 TEHACHPI 500 110 87
24520 TEHACHPI 500 105 88
24520 TEHACHPI 500 101 89
24520 TEHACHPI 500 100 90
24520 TEHACHPI 500 100 91
205
Table 42 (continued). Incremental Wind Projects added for 2010T Study Scenario.
Power Flow Data
24520 TEHACHPI 500 102 92
24520 TEHACHPI 500 103 93
24520 TEHACHPI 500 112 94
24520 TEHACHPI 500 103 95
24520 TEHACHPI 500 142 96
24520 TEHACHPI 500 103 97
24520 TEHACHPI 500 100 98
24520 TEHACHPI 500 183 99
24815 GARNET 115 21 76
24828 WINTEC9 13.8 61 77
25632 TERAWND 115 158 63
25633 CAPWIND 115 158 64
25634 BUCKWND 115 158 65
25635 ALTWIND 115 154 66
25636 RENWIND 115 158 67
25637 TRANWND 115 158 68
25639 SEAWIND 115 158 69
25645 VENWIND 115 154 70
25646 SANWIND 115 156 71
25902 NEWSD138 138 90 58
26135 WTG 0.57 120 75
26160 LA-Wind 230 120 74
28020 WINTEC6 115 38 78
28060 SEAWEST 115 76 79
28061 WHITEWTR 33 66 80
28280 CABAZON 33 43 81
30529 HIWD TAP 230 165 72
38610 DELTAPMP 230 80 73
Table 43. Incremental Wind Projects Added for 2010X Study Scenario.
Power Flow Data
21915 IMPERHWD 500 110 106
21915 IMPERHWD 500 110 136
21915 IMPERHWD 500 111 138
21915 IMPERHWD 500 131 158
21915 IMPERHWD 500 138 189
22465 SDGEHWD 500 110 165
22465 SDGEHWD 500 110 176
22465 SDGEHWD 500 125 199
22465 SDGEHWD 500 155 220
24056 ETIWANDA 230 100 105
206
Table 43 (continued). Incremental Wind Projects Added for 2010X Study Scenario.
Power Flow Data
24056 ETIWANDA 230 104 110
24056 ETIWANDA 230 100 112
24056 ETIWANDA 230 102 124
24056 ETIWANDA 230 71 128
24520 TEHACHPI 500 104 113
24520 TEHACHPI 500 145 116
24520 TEHACHPI 500 106 119
24520 TEHACHPI 500 113 121
24520 TEHACHPI 500 158 125
24520 TEHACHPI 500 100 126
24520 TEHACHPI 500 124 127
24520 TEHACHPI 500 132 130
24520 TEHACHPI 500 105 131
24520 TEHACHPI 500 125 134
24520 TEHACHPI 500 101 146
24520 TEHACHPI 500 100 147
24520 TEHACHPI 500 92 148
26160 LA-Wind 230 80 153
30105 COTWD_E 230 134 108
30529 HIWD TAP 230 103 100
30529 HIWD TAP 230 102 102
30529 HIWD TAP 230 114 103
30529 HIWD TAP 230 117 104
30529 HIWD TAP 230 107 107
30529 HIWD TAP 230 107 109
30529 HIWD TAP 230 114 111
30529 HIWD TAP 230 152 114
31665 WESTWOOD 230 104 122
31665 WESTWOOD 230 100 133
31665 WESTWOOD 230 100 137
31665 WESTWOOD 230 100 157
31665 WESTWOOD 230 114 166
31665 WESTWOOD 230 131 232
34342 INT.TURB 9.11 40 172
38610 DELTAPMP 230 40 73
207
Table 44. Incremental Wind Projects Added for 2020 Study Scenario.
Power Flow Data
24098 MOORPARK 66 50 161
25903 INYOWIND 115 100 162
33170 WINDMSTR 9.11 28 37
38610 DELTAPMP 230 80 73
208
209
Appendix B. Summary of Solar Projects by Scenario
Thesolarprojectsincludedineachstudyscenarioareshowninthefollowingtables.Eachtable
identifiesaprojectbyitspowerflowbusnumber,busname,andvoltagelevelinthefirstthree
columns.Thefourthcolumnshowstheratedpoweroutputforanindividualprojectandthe
fifthcolumnshowstheprojecttype(i.e.,concentratedorphotovoltaic).Thefinalcolumnshows
thetypeofprofileusedinthesubhourlystatisticalandQSSanalyses.
Table45liststheconcentratedsolarprojectsincludedinthe2006studyscenario.All2006PV
siteswereincorporatedintotheload,andnotrepresentedindividually.Table46liststhe
incrementalprojectsaddedtothe2006scenariotocreatethe2010Tscenario.Table47liststhe
incrementalprojectsaddedtothe2010Tscenariotocreatethe2010Xscenario.Table48liststhe
incrementalprojectsaddedtothe2010Xscenariotocreatethe2020scenario.
Table 45. Solar Projects Included in 2006 Study Scenario.
Power Flow Data
Bus # Bus Name kV Total MW Rating Project Type Profile Type
24737 LUZ8 G 13.8 80 Concentrated Sungen/Luz
24738 LUZ9 G 13.8 80 Concentrated Sungen/Luz
24754 SUNGEN3G 13.8 34 Concentrated Sungen/Luz
Table 46. Incremental Solar Projects Added for 2010T Study Scenario.
Power Flow Data
Bus # Bus Name kV Added MW Rating Project Type Profile Type
21032 EMESA1 92 11 Concentrated Sungen/Luz
21038 HIGHLINE 230 11 Concentrated Sungen/Luz
21043 LEATHERS 92 11 Concentrated Sungen/Luz
21045 MIDWAY X 230 11 Concentrated Sungen/Luz
22068 BOLDRCRK 69 7 Concentrated Sungen/Luz
22360 IMPRLVLY 500 300 Concentrated Stirling
24097 MOHAVE 500 500 Concentrated Stirling
24751 SEGS 1G 14 20 Concentrated Sungen/Luz
24752 SEGS 2G 14 32.6 Concentrated Sungen/Luz
24902 VSTA 66 50 Concentrated Sungen/Luz
31678 GRYS FLT 60 23 Concentrated Sungen/Luz
21004 AVE58 92 3 PV Zip Code 920
22008 ASH 69 2 PV Zip Code 920
22048 BATIQTOS 138 2 PV Zip Code 920
22056 BERNARDO 69 3 PV Zip Code 921
22108 CANNON 138 2 PV Zip Code 920
210
Table 46 (continued). Incremental Solar Projects Added for 2010T Study Scenario.
Power Flow Data
22112 CAPSTRNO 138 4 PV Zip Code 920
22124 CHCARITA 138 2 PV Zip Code 920
22160 DEL MAR 69 2 PV Zip Code 920
22208 EL CAJ ON 69 2 PV Zip Code 920
22216 ELLIOTT 69 2 PV Zip Code 921
22252 ENCNITAS 69 2 PV Zip Code 920
22256 ESCNDIDO 69 2 PV Zip Code 920
22272 ESCO 69 2 PV Zip Code 920
22288 FELICITA 69 2 PV Zip Code 920
22316 GENESEE 69 3 PV Zip Code 921
22336 GRANITE 69 2 PV Zip Code 920
22364 J AMACHA 69 2 PV Zip Code 920
22372 KEARNY 69 2 PV Zip Code 921
22396 LAGNA NL 138 9 PV Zip Code 920
22408 LOSCOCHS 69 2 PV Zip Code 921
22432 MARGARTA 138 15 PV Zip Code 920
22440 MELROSE 69 3 PV Zip Code 920
22444 MESA RIM 69 2 PV Zip Code 921
22448 MESAHGTS 69 2 PV Zip Code 921
22480 MIRAMAR 69 2 PV Zip Code 921
22496 MISSION 69 3 PV Zip Code 921
22516 MONTGMRY 69 2 PV Zip Code 921
22532 MURRAY 69 2 PV Zip Code 920
22592 OLD TOWN 69 2 PV Zip Code 921
22620 PACFCBCH 69 2 PV Zip Code 921
22632 PALOMAR 138 2 PV Zip Code 920
22656 PICO 138 4 PV Zip Code 920
22660 POINTLMA 69 2 PV Zip Code 921
22676 R.CARMEL 69 2 PV Zip Code 920
22704 SAMPSON 13 2 PV Zip Code 921
22708 SANLUSRY 69 2 PV Zip Code 920
22724 SANMRCOS 69 2 PV Zip Code 920
22760 SHADOWR 138 2 PV Zip Code 920
22800 STREAMVW 69 2 PV Zip Code 921
22852 TELECYN 138 2 PV Zip Code 921
22856 TOREYPNS 69 3 PV Zip Code 921
22868 URBAN 69 2 PV Zip Code 921
24024 CHINO 66 12 PV Zip Code 91A
24055 ETIWANDA 66 8 PV Zip Code 91A
24111 PADUA 66 12 PV Zip Code 91A
24133 SANTIAGO 66 11 PV Zip Code 91A
24135 SAUGUS 66 18 PV Zip Code 91A
211
Power Flow Data
24157 WALNUT 66 10 PV Zip Code 91A
24160 VALLEYSC 115 5 PV Zip Code 91A
24205 EAGLROCK 66 3 PV Zip Code 91A
24207 J OHANNA 66 6 PV Zip Code 91A
24211 OLINDA 66 7 PV Zip Code 91A
24216 VILLA PK 66 10 PV Zip Code 91A
24418 LANCSTR 66 6 PV Zip Code 91A
24422 PALMDALE 66 3 PV Zip Code 91A
24424 QUARTZHL 66 5 PV Zip Code 91B
24426 SHUTTLE 66 3 PV Zip Code 91B
24602 VICTOR 115 6 PV Zip Code 91B
24603 APPLEVAL 115 5 PV Zip Code 91B
24605 HESPERIA 115 5 PV Zip Code 91B
24608 SAVAGE 115 5 PV Zip Code 91B
24816 SANTA RO 115 3 PV Zip Code 91B
24817 EISENHOW 115 3 PV Zip Code 91B
24821 TAMARISK 115 2 PV Zip Code 91B
24822 INDIAN W 115 1 PV Zip Code 91B
24902 VSTA 66 10 PV Zip Code 91B
25002 GOODRICH 33 10 PV Zip Code 91B
25202 LEWIS 66 8 PV Zip Code 91B
26013 GLENDAL 230 3 PV Zip Code 900
26061 RINALDI 230 4 PV Zip Code 900
26063 RIVER 230 6 PV Zip Code 900
26068 STJ OHN 230 3 PV Zip Code 900
26076 FAIRFAX 138 7 PV Zip Code 900
26078 TOLUCA 230 17 PV Zip Code 900
26081 ATWATER 230 5 PV Zip Code 905
26085 HOLYWDLD 138 7 PV Zip Code 905
26086 NRTHRDGE 230 9 PV Zip Code 905
26088 OLYMPCLD 138 7 PV Zip Code 905
26093 TARZANA 230 12 PV Zip Code 905
26102 VALLEY 138 4 PV Zip Code 905
30535 TIDEWATR 230 1 PV Zip Code 94A
30545 ROSSMOOR 230 1 PV Zip Code 945
30555 SANRAMON 230 3 PV Zip Code 945
30561 TASSAJ AR 230 1 PV Zip Code 945
30585 LS PSTAS 230 5 PV Zip Code 945
30720 SARATOGA 230 3 PV Zip Code 94A
30730 HICKS 230 3 PV Zip Code 94A
30841 FIGRDN 1 230 2 PV Zip Code 95A
212
Power Flow Data
30850 ASHLAN 230 4 PV Zip Code 95A
32971 MEDW LNE 115 1 PV Zip Code 945
32973 LAKEWD-C 115 1 PV Zip Code 945
32974 LAKEWD-M 115 1 PV Zip Code 945
33714 HAMMER 60 4 PV Zip Code 95A
33801 STAGG_5 21 4 PV Zip Code 95A
34372 MALAGA 115 2 PV Zip Code 95A
34404 WST FRSO 115 2 PV Zip Code 95A
34408 BARTON 115 3 PV Zip Code 95A
34410 MANCHSTR 115 2 PV Zip Code 95A
34414 WOODWARD 115 3 PV Zip Code 95A
34416 BULLARD 115 3 PV Zip Code 95A
34706 WESTPARK 115 4 PV Zip Code 95A
34718 KERN OIL 115 4 PV Zip Code 95A
34736 MAGUNDEN 115 4 PV Zip Code 95A
35353 MT VIEW 115 2 PV Zip Code 94A
35354 STELLING 115 2 PV Zip Code 94A
35363 LAWRENCE 115 2 PV Zip Code 94A
35368 BRITTN 115 2 PV Zip Code 94A
35610 MONTAGUE 115 3 PV Zip Code 95A
35612 TRIMBLE 115 3 PV Zip Code 95A
35620 EL PATIO 115 3 PV Zip Code 95A
35622 SWIFT 115 3 PV Zip Code 945
35624 MILPITAS 115 3 PV Zip Code 945
35626 MCKEE 115 2 PV Zip Code 95A
35638 EDENVALE 115 3 PV Zip Code 95A
35646 MRGN HIL 115 2 PV Zip Code 94A
36850 KIFER 115 3 PV Zip Code 94A
36852 SCOTT 115 3 PV Zip Code 94A
37101 CARMICAL 69 7 PV Zip Code 956
37103 ELVERTA1 69 3 PV Zip Code 956
37114 ORANGVL1 69 7 PV Zip Code 956
37649 LLNLAB 115 4 PV Zip Code 945
38028 PLO ALTO 115 4 PV Zip Code 94A
38280 SYLVAN 69 13 PV Zip Code 95A
213
Table 47. Incremental Solar Projects Added for 2010X Study Scenario.
Power Flow Data
22068 BOLDRCRK 69 13 Concentrated Sungen/Luz
31678 GRYS FLT 60 -23 Concentrated Sungen/Luz
Table 48. Incremental Solar Projects Added for 2020 Study Scenario.
Power Flow Data
21043 LEATHERS 92 69 Concentrated Sungen/Luz
22360 IMPRLVLY 500 200 Concentrated Stirling
24097 MOHAVE 500 350 Concentrated Stirling
24809 YUCCA 115 -14 Concentrated Sungen/Luz
24810 HI DESER 115 18 Concentrated Sungen/Luz
24817 EISENHOW 115 -1 Concentrated Sungen/Luz
24819 CONCHO 115 -1 Concentrated Sungen/Luz
24824 CARODEAN 115 18 Concentrated Sungen/Luz
24902 VSTA 66 -50 Concentrated Sungen/Luz
24903 VSTA 66 18 Concentrated Sungen/Luz
24909 PEPPER 115 23 Concentrated Sungen/Luz
24911 HOMART 115 18 Concentrated Sungen/Luz
24912 SHANDIN 115 18 Concentrated Sungen/Luz
25602 DVLCYN34 115 18 Concentrated Sungen/Luz
25632 TERAWND 115 -1 Concentrated Sungen/Luz
25633 CAPWIND 115 -1 Concentrated Sungen/Luz
25635 ALTWIND 115 -30 Concentrated Sungen/Luz
25636 RENWIND 115 20 Concentrated Sungen/Luz
25639 SEAWIND 115 -30 Concentrated Sungen/Luz
25645 VENWIND 115 -1 Concentrated Sungen/Luz
25646 SANWIND 115 -1 Concentrated Sungen/Luz
25650 MHV SPHN 115 -17 Concentrated Sungen/Luz
31678 GRYS FLT 60 23 Concentrated Sungen/Luz
31665 WESTWOOD 230 200 Concentrated Sungen/Luz
21002 AVE42 92 19 PV Zip Code 920
21005 COACHELA 92 20 PV Zip Code 920
22008 ASH 69 8 PV Zip Code 920
214
Table 48 (continued). Incremental Solar Projects Added for 2020 Study Scenario.
Power Flow Data
22024 B 69 10 PV Zip Code 921
22048 BATIQTOS 138 8 PV Zip Code 920
22056 BERNARDO 69 7 PV Zip Code 921
22108 CANNON 138 8 PV Zip Code 920
22124 CHCARITA 138 5 PV Zip Code 920
22132 CHOLLAS 69 3 PV Zip Code 920
22160 DEL MAR 69 1 PV Zip Code 920
22208 EL CAJ ON 69 8 PV Zip Code 920
22216 ELLIOTT 69 1 PV Zip Code 921
22252 ENCNITAS 69 5 PV Zip Code 920
22256 ESCNDIDO 69 13 PV Zip Code 920
22276 F 69 7 PV Zip Code 921
22316 GENESEE 69 7 PV Zip Code 921
22336 GRANITE 69 8 PV Zip Code 920
22364 J AMACHA 69 8 PV Zip Code 920
22372 KEARNY 69 8 PV Zip Code 921
22396 LAGNA NL 138 11 PV Zip Code 920
22408 LOSCOCHS 69 5 PV Zip Code 921
22432 MARGARTA 138 37 PV Zip Code 920
22440 MELROSE 69 14 PV Zip Code 920
22444 MESA RIM 69 8 PV Zip Code 921
22448 MESAHGTS 69 3 PV Zip Code 921
22480 MIRAMAR 69 8 PV Zip Code 921
22496 MISSION 69 7 PV Zip Code 921
22516 MONTGMRY 69 1 PV Zip Code 921
22532 MURRAY 69 8 PV Zip Code 920
22576 NOISLMTR 69 7 PV Zip Code 921
22580 NORTHCTY 138 3 PV Zip Code 920
22592 OLD TOWN 69 8 PV Zip Code 921
22620 PACFCBCH 69 1 PV Zip Code 921
22636 PARADISE 69 3 PV Zip Code 921
22656 PICO 138 6 PV Zip Code 920
22660 POINTLMA 69 1 PV Zip Code 921
22664 POMERADO 69 3 PV Zip Code 920
22668 POWAY 69 3 PV Zip Code 920
22676 R.CARMEL 69 1 PV Zip Code 920
22696 ROSE CYN 69 7 PV Zip Code 921
22704 SAMPSON 12.5 8 PV Zip Code 921
22708 SANLUSRY 69 13 PV Zip Code 920
22724 SANMRCOS 69 8 PV Zip Code 920
22734 SANTEE 138 7 PV Zip Code 920
22760 SHADOWR 138 8 PV Zip Code 920
215
Power Flow Data
22800 STREAMVW 69 5 PV Zip Code 921
22820 SWEETWTR 69 3 PV Zip Code 921
22852 TELECYN 138 5 PV Zip Code 921
22856 TOREYPNS 69 7 PV Zip Code 921
22868 URBAN 69 5 PV Zip Code 921
24007 ALMITOSW 66 10 PV Zip Code 91A
24024 CHINO 66 59 PV Zip Code 91A
24028 DELAMO 66 10 PV Zip Code 91A
24032 AMERON 66 13 PV Zip Code 91A
24039 EL NIDO 66 27 PV Zip Code 91A
24055 ETIWANDA 66 32 PV Zip Code 91A
24083 LITEHIPE 66 15 PV Zip Code 91A
24111 PADUA 66 38 PV Zip Code 91A
24133 SANTIAGO 66 30 PV Zip Code 91A
24135 SAUGUS 66 63 PV Zip Code 91A
24157 WALNUT 66 40 PV Zip Code 91A
24160 VALLEYSC 115 25 PV Zip Code 91A
24201 BARRE 66 21 PV Zip Code 91A
24205 EAGLROCK 66 22 PV Zip Code 91A
24207 J OHANNA 66 14 PV Zip Code 91A
24211 OLINDA 66 18 PV Zip Code 91A
24212 RECTOR 66 28 PV Zip Code 91A
24213 RIOHONDO 66 29 PV Zip Code 91A
24216 VILLA PK 66 11 PV Zip Code 91A
24407 ANAVERDE 66 10 PV Zip Code 91A
24418 LANCSTR 66 24 PV Zip Code 91A
24421 OASIS SC 66 10 PV Zip Code 91A
24422 PALMDALE 66 12 PV Zip Code 91A
24424 QUARTZHL 66 13 PV Zip Code 91B
24426 SHUTTLE 66 12 PV Zip Code 91B
24602 VICTOR 115 7 PV Zip Code 91B
24603 APPLEVAL 115 5 PV Zip Code 91B
24604 AQUEDUCT 115 10 PV Zip Code 91B
24605 HESPERIA 115 5 PV Zip Code 91B
24606 PHELAN 115 10 PV Zip Code 91B
24607 ROADWAY 115 10 PV Zip Code 91B
24608 SAVAGE 115 8 PV Zip Code 91B
24610 BLKMTN 115 10 PV Zip Code 91B
24622 PERMANTE 115 10 PV Zip Code 91B
24623 GOLDHILS 115 10 PV Zip Code 91B
24815 GARNET 115 10 PV Zip Code 91B
24818 FARREL 115 10 PV Zip Code 91B
216
Power Flow Data
24903 VSTA 115 -6 PV Zip Code 91B
25002 GOODRICH 33 11 PV Zip Code 91B
25202 LEWIS 66 12 PV Zip Code 91B
25655 VIEJ O66 66 31 PV Zip Code 91B
26061 RINALDI 230 21 PV Zip Code 900
26063 RIVER 230 19 PV Zip Code 900
26068 STJ OHN 230 22 PV Zip Code 900
26076 FAIRFAX 138 28 PV Zip Code 900
26078 TOLUCA 230 28 PV Zip Code 900
26081 ATWATER 230 20 PV Zip Code 905
26085 HOLYWDLD 138 28 PV Zip Code 905
26086 NRTHRDGE 230 36 PV Zip Code 905
26088 OLYMPCLD 138 28 PV Zip Code 905
26093 TARZANA 230 33 PV Zip Code 905
26102 VALLEY 138 9 PV Zip Code 905
30430 FULTON 230 8 PV Zip Code 954
30472 PEABODY 230 14 PV Zip Code 94A
30505 WEBER 230 10 PV Zip Code 95A
30535 TIDEWATR 230 4 PV Zip Code 94A
30545 ROSSMOOR 230 4 PV Zip Code 945
30554 CASTROVL 230 15 PV Zip Code 945
30555 SANRAMON 230 7 PV Zip Code 945
30561 TASSAJ AR 230 4 PV Zip Code 945
30565 BRENTWOD 230 25 PV Zip Code 945
30585 LS PSTAS 230 21 PV Zip Code 945
30711 S.L.A.C. 230 5 PV Zip Code 94A
30720 SARATOGA 230 6 PV Zip Code 94A
30730 HICKS 230 6 PV Zip Code 94A
30850 ASHLAN 230 6 PV Zip Code 95A
30941 STCKDLEB 230 5 PV Zip Code 95A
30950 BKRSFLDA 230 5 PV Zip Code 95A
30951 BKRSFLDB 230 14 PV Zip Code 95A
31239 MONROE2 115 8 PV Zip Code 954
31240 SNTA RSA 115 12 PV Zip Code 954
31246 BELLVUE 115 5 PV Zip Code 954
31467 J ESSUP 115 5 PV Zip Code 95A
31496 NORD 1 115 5 PV Zip Code 95A
31498 SYCAMORE 115 14 PV Zip Code 95A
31500 BUTTE 115 5 PV Zip Code 95A
217
Power Flow Data
31502 CHICO B 115 5 PV Zip Code 95A
32010 J AMESON 115 5 PV Zip Code 95A
32258 DMND SPR 115 5 PV Zip Code 95A
32263 CLRKSVLE 115 20 PV Zip Code 95A
32265 SHPRING 115 5 PV Zip Code 95A
32971 MEDW LNE 115 4 PV Zip Code 945
32973 LAKEWD-C 115 4 PV Zip Code 945
32974 LAKEWD-M 115 4 PV Zip Code 945
32978 LMEC 115 9 PV Zip Code 945
33311 BAY MDWS 115 8 PV Zip Code 94A
33312 BELMONT 115 5 PV Zip Code 94A
33370 REDWOOD 60 5 PV Zip Code 94A
33372 BLLE HVN 60 5 PV Zip Code 94A
33548 TRACY 115 18 PV Zip Code 95A
33555 STKTON A 115 10 PV Zip Code 95A
33704 STAGG 60 7 PV Zip Code 95A
33714 HAMMER 60 6 PV Zip Code 95A
34362 CLOVIS-1 115 10 PV Zip Code 95A
34364 CLOVIS-2 115 10 PV Zip Code 95A
34372 MALAGA 115 8 PV Zip Code 95A
34404 WST FRSO 115 8 PV Zip Code 95A
34408 BARTON 115 7 PV Zip Code 95A
34410 MANCHSTR 115 8 PV Zip Code 95A
34414 WOODWARD 115 7 PV Zip Code 95A
34416 BULLARD 115 7 PV Zip Code 95A
34706 WESTPARK 115 1 PV Zip Code 95A
34718 KERN OIL 115 1 PV Zip Code 95A
34736 MAGUNDEN 115 1 PV Zip Code 95A
34752 KERN PWR 115 5 PV Zip Code 95A
34754 TEVIS 115 5 PV Zip Code 95A
34762 RENFRO 115 20 PV Zip Code 95A
34911 FRUITVLE 70 13 PV Zip Code 95A
35062 DISCOVRY 13.8 10 PV Zip Code 95A
35106 MT EDEN 115 29 PV Zip Code 945
35110 FREMNT 115 15 PV Zip Code 945
35111 J ARVIS 115 15 PV Zip Code 945
35120 NEWARK D 115 9 PV Zip Code 945
35350 AMES BS1 115 6 PV Zip Code 94A
35351 AMES BS2 115 5 PV Zip Code 94A
35352 WHISMAN 115 6 PV Zip Code 94A
218
Power Flow Data
35353 MT VIEW 115 4 PV Zip Code 94A
35354 STELLING 115 7 PV Zip Code 94A
35355 WOLFE 115 6 PV Zip Code 94A
35363 LAWRENCE 115 6 PV Zip Code 94A
35368 BRITTN 115 4 PV Zip Code 94A
35606 AGNEW 115 6 PV Zip Code 95A
35610 MONTAGUE 115 5 PV Zip Code 95A
35612 TRIMBLE 115 6 PV Zip Code 95A
35620 EL PATIO 115 6 PV Zip Code 95A
35622 SWIFT 115 3 PV Zip Code 945
35624 MILPITAS 115 6 PV Zip Code 945
35626 MCKEE 115 6 PV Zip Code 95A
35636 EVRGRN 1 115 6 PV Zip Code 95A
35638 EDENVALE 115 6 PV Zip Code 95A
35646 MRGN HIL 115 6 PV Zip Code 94A
35907 PAUL SWT 115 31 PV Zip Code 94B
35918 SALINAS2 115 10 PV Zip Code 94B
36420 STONE 115 6 PV Zip Code 95A
37101 CARMICAL 69 2 PV Zip Code 956
37102 ELKGROV1 69 10 PV Zip Code 956
37107 HEDGE 3 69 25 PV Zip Code 956
37108 HURLEY 1 69 15 PV Zip Code 956
37109 HURLEY 2 69 19 PV Zip Code 956
37111 LAKE 1 69 8 PV Zip Code 956
37116 POCKET 1 69 10 PV Zip Code 956
37117 POCKET 2 69 19 PV Zip Code 956
37122 LAKE 2 69 8 PV Zip Code 956
37123 NATOMAS 69 15 PV Zip Code 956
37583 TRACYPP1 13.8 5 PV Zip Code 945
37584 TRACYPP2 13.8 6 PV Zip Code 945
37649 LLNLAB 115 11 PV Zip Code 945
37912 CANBY2 12.5 5 PV Zip Code 959
37913 CANBY3 12.5 16 PV Zip Code 959
37949 SULP1 12.5 5 PV Zip Code 959
38028 PLO ALTO 115 5 PV Zip Code 94A
38228 SNTA CRZ 115 10 PV Zip Code 95A
38262 BRGGSMRE 69 10 PV Zip Code 95A
38264 ENSLEN 69 10 PV Zip Code 95A
219
Power Flow Data
38268 12TH ST 69 10 PV Zip Code 95A
38280 SYLVAN 69 -3 PV Zip Code 95A
38314 STODDARD 69 5 PV Zip Code 95A
38460 CERES 69 6 PV Zip Code 95A
220
221
Appendix C. Application of Solar Data
Theassumptionsandtechniquesemployedtodevelopthenecessaryhighresolution(i.e.,1
minute)profilesforeachsolarprojectaredescribedinthisappendix.Theseprofileswereused
inboththesubhourlystatisticalanalysisandtheQSSanalysis.Additionaldetailsonthose
analyses,includingtheselectionoftimeperiodsforevaluation,arecontainedinthebodyofthe
report.
AvailableSolarData
Limitedhighresolutionsolardatawasprovidedforthisstudy.Asnotedinthemainreport,the
availabledatawas:
1minutedataforthreeyears(20022004)ofnetSungenandLuzplantoutput(MW)
15minutedataforoneyear(2004)of13CaliforniazipcodebasedPVprofiles(pu)
1minutedataforJanuary2002andJuly2002ofGoldenCOirradiation(W/m
2
)
3minutedataforJanuary2002andJuly2002ofDesertRockNVirradiation(W/m
2
)
Inaddition,thefollowinghourlysolardatawasavailable:
Threeyears(20022004)ofnetSungenandLuzplantoutput(MW)
Threeyears(20022004)ofStirlingprojectoutput(MW)fortwosites
Oneyear(2004)ofCaliforniazipcodebasedPVprofiles(pu)
Unlessotherwisenoted,theprofilesforagivenstudytimeperioduseddatafromthattime
period.
Developmentof1MinuteProfiles
ThreetypesofprofilesweredevelopedforthisstudySungen/Luzbasedconcentratedsolar
profiles,Stirlingenginebasedprojectprofiles,andzipcodebasedPVprofiles.
Asimplescalingprocedurewasusedtocreateconcentratingsolarprojectprofilesfromthenet
SungenandLuzdata.Thisprocedurewasusedtocreate1minuteprofilesfortheexistingunits
atthosetwosites,aswellasforanynewconcentratingsolarprojects.Projectsidentifiedas
Stirlingenginebasedwereexcluded.Theprocedureconsistedofmultiplyingthedatainthe
desiredtimeperiodbyafactorequaltotheprojectratingdividedbytheexistingnetSungen
andLuzrating.AnexampleprofileforthecombinedSungenandLuzfacilitiesisshownin
Figure178.
222
0
100
200
300
400
500
5 10 15 20
Sungen & Luz
Figure 178. Example Concentrating Solar Project Profile for a May Day.
TwonewStirlingprojectswereincludedinthefuturestudyscenarios.Sincenohighresolution
datawasavailable,thenecessaryprofilesweredevelopedbysuperimposingthe3minute
DesertRockNVdirectirradiationdataonthehourlydata.TheJanuaryirradiationdatawas
usedforstudyperiodsfromNovemberthroughApril,andtheJulyirradiationdatawasused
forstudyperiodsfromMaythroughOctober.
Thefirststepinthisprocedurewastointerpolatebetweenthehourlydatapointstocreatea
rampratherthanastepfunction.
Next,arolling1houraverageofthe3minutedatawascalculated.Then,thedifferencebetween
the3minutedataandtherollingaveragewasdeterminedandsuperimposedonthehourly
datatocreatea3minuteprofile.Thissumwaslimitedtoensurethattheprojectoutputdidnot
exceedprojectrating.
Finally,the3minuteprofilewasinterpolatedtocreatea1minuteprofile.
Arandomnumbergeneratorwasusedtopickdifferentdaysofirradiationdatafordifferent
studyintervalsandsites.Exampleprofilesforboththe300MWand500MWStirlingprojects
areshowninFigure179.
M
W
HourofDay
223
0
100
200
300
400
500
5 10 15 20
Stirling 500MW
Stirling 300MW
Figure 179. Example Stirling Solar Project Profile for a May Day.
Morethan200PVprojectswereincludedinsomeofthefuturestudyscenarios.Thehighest
resolutionCaliforniaPVdatawasbasedonzipcodesandin15minuteincrements.Therefore,
eachPVsitewasassignedanappropriatezipcode.AllPVsitesinagivenzipcodeusedthe
sameprofile.
SincenohigherresolutionCaliforniadatawasavailable,thenecessary1minuteprofileswere
developedbysuperimposingthe1minuteGoldenCOdirectirradiationdataonthe15minute
Californiazipcodebaseddata.TheJanuaryirradiationdatawasusedforstudyperiodsfrom
NovemberthroughApril,andtheJulyirradiationdatawasusedforstudyperiodsfromMay
throughOctober.
Thefirststepinthisprocedurewastointerpolatebetweenthe15minutedatapointstocreatea
rampratherthanastepfunction.
Next,arolling15minuteaverageofthe1minutedatawascalculated.Then,thedifference
betweenthe1minutedataandtherollingaveragewasdeterminedandsuperimposedonthe
15minutedatatocreatea1minuteprofile.Thissumwaslimitedtoensurethattheproject
outputdidnotexceedprojectrating.
M
W
HourofDay
224
Arandomnumbergeneratorwasusedtopickdifferentdaysofirradiationdatafordifferent
studyintervalsandzipcodes.Exampleprofilesforthe10zipcodesusedinthe2010scenarios
areshowninFigure180.
0
50
100
150
5 10 15 20
Zip Code 920
Zip Code 91a
Zip Code 95a
Zip Code 91b
Zip Code 905
Zip Code 900
Zip Code 921
Zip Code 956
Zip Code 94a
Zip Code 945
Figure 180. Example PV Solar ZIP Code Profiles for a May Day.
BoththeStirlingandPVsolarprofileshavemorenotchesthantheSungen/Luzprofile,whichis
consistentwiththeirrelativelyshortthermaltimeconstantsandsensitivitytovariablecloud
cover.
ApplicabilityofOutofStateSolarDatatoCaliforniaPVSites
AbriefstatisticalevaluationwasperformedtotesttheapplicabilityoftheGoldenCOdatato
thevariousCaliforniaPVsites.First,thedifferencebetweeneachintervalinthe15minutezip
codebasedCaliforniadatawascalculatedforthemonthsofJanuaryandJuly.Thestandard
deviationofthis15minutevariabilitywascalculatedandisshowninTable49.
Then,the15minuteaverageofthe1minuteGoldenCOdatawascalculated,followedbya
calculationofthedifferencebetweeneach15minutevalue.Thestandarddeviationofthis15
minutevariabilitywasalsocalculatedandisshowninTable50.
M
W
HourofDay
225
Table 49. Standard Deviation of the 15-Minute Variability in
the California PV Data.
CA Zip Code Based Data (pu) January July
91A 0.024 0.033
91B 0.026 0.033
900 0.025 0.036
905 0.029 0.047
920 0.034 0.033
921 0.031 0.032
945 0.021 0.026
94A 0.023 0.033
954 0.030 0.034
94B 0.017 0.029
956 0.028 0.035
959 0.049 0.047
95A 0.029 0.032
Table 50. Standard Deviation of the 15-Minute Variability in

the Golden, CO, Irradiation Data.
Golden, CO (pv) January July
15-min Average 0.042 0.066
Thestandarddeviationofthe15minutevariabilityofthe1minuteGoldendataisgenerally
higherthanthatoftheCaliforniadata.Theexceptioniszipcode959(coveringYuba,Sutterand
othercounties)inJanuary,whichhasthehigheststandarddeviationofanyoftheCaliforniazip
codes.Therefore,usingtheGoldendatatoprovide1minutevariabilityisbothconservativeand
credible.
Thefollowingexampleillustratestheprocedureforsuperimposingthe1minuteGoldenCO
dataonthe15minuteCaliforniazipcodedatatocreate1minuteprofilesforallCaliforniaPV
sites.
Step1istointerpolateat1minuteintervalsbetweenthe15minutedatapointstoachievea
smoothratherthanstairstepfunction.
Step2istosuperimposethe1minutevariabilityinglobalinsolationfromtheGoldendataonto
the15minuteinterpolatedprofile.TheJanuarydatawasusedforNovemberthroughApril,
andtheJulydatawasusedforMaythroughOctober.Variabilityforindividualdayswas
appliedinarandomfashion.
Figure181showsthe15minutedataforaJulydaywith1minuteinterpolation(blueline),and
thefinal1minuteprofileincludingthevariabilityfromtheGoldendata(redline).Forreference
the1minuteGoldeninsolationdatausedinthisexampleisshowninFigure182.Notethe
similarityinthe1minutevariability.Alsonotethatthelongertermvariability(e.g.15minute
toanhour)observedaroundinterval720intheinsolationprofileisnotsuperimposedonthe
226
15minuteCaliforniazipcodedata.Thisisimportantbecausethe15minutevariabilityis
alreadypresentintheCaliforniadata,andonlythe1minvariabilityneedstobesuperimposed.
0.00
0.20
0.40
0.60
0.80
1.00
0 240 480 720 960
1-mi n Interval s
P
o
w
e
r

O
u
t
p
u
t

(
p
u
)
15-minute w/ Interpolation
15-minute CA + 1-minute CO
Figure 181. Comparison of Original 15-Mminute Data and Final Profile with 1-Minute
Variability.
0
200
400
600
800
1000
1200
1400
0 240 480 720 960
1-mi n Interval s
I
n
s
o
l
a
t
i
o
n

(
W
/
m
^
2
)

Figure 182. Irradiation Data Used as Source of 1-Minute Variability in Example.
Giventhepaucityofavailablesolardata,thisevaluationconfirmsthereasonablenessofusing
theGoldenCOirradiationdatatoprovide1minutevariabilityfortheCaliforniaPVsites.

NERC Wind

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

NERC Wind

Încărcat de

Drepturi de autor:

Formate disponibile

Arnold Schwarzenegger

Figure 1. Flowchart of Intermittency Analysis Project.

Figure 2. Time Scales for Grid Planning and Operations.

Table 4. Locations of Wind and Solar Resources for Scenario 2010X.

Figure 5. All Systemwide Daily Solar Profiles for July 2003.

Figure 8. All Systemwide Daily Solar Profiles for January 2002.

Figure 14. 2010X Wind Production and Penetration Duration Curves.

Figure 15. 2010X Solar Production and Penetration Duration Curves.

Relative to Load Level

3.3. Intra-Hour Variability

Table 12. Statistics on Load Following

Table 13. Statistics on Regulation Requirement

Table 16. Summary of 2010 and 2020 Hourly Statistical Analysis.

Table 23. 2010X Day-Ahead Forecast Error Energy.

Table 25. 2010X Hour-Ahead Forecast Error Energy.

Table 26. Production Simulation Scenario Description.

Figure 55. Spot Price Duration Curve for 2003 Shapes.

Figure 56. Spot Price Duration Curve for 2004 Shapes.

Figure 58. WECC Operating Cost Reductions Due to Renewables ($M).

Figure 59. WECC Operating Cost Reductions Due to Renewables ($/MWh).

Figure 60. Load Payment Reductions Due to Renewables ($M).

Figure 61. Load Payment Reductions Due to Renewables ($/MWh).

Figure 62. Non-Renewable Generator Revenue Reductions Due to Renewables ($M).

Figure 63. Non-Renewable Generator Revenue Reductions Due to Renewables

Figure 64. Impact of Intermittent Forecast on Spot Price (2010T).

Figure 65. Impact of Intermittent Forecast on Spot Price (2010X).

Figure 67. Total Operating Cost Impact of Intermittent Forecasting.

Figure 68. California Generator Revenue Reductions by Type (2010T).

Figure 69. WECC Generator Revenue Reductions by Type (2010T).

Figure 70. California Energy Change Due to Renewables (2010T).

Figure 71. WECC Energy Change Due to Renewables (2010T).

Figure 72. Annual Duration Curves California Renewable Generation.

Figure 75. Annual Duration Curves California Hydro Generation.

Figure 76. One Week Change in California Hydro Operation (2010T).

Figure 79. Annual Histogram of Hydro Shift (2010T).

Figure 81. Annual Duration Curves - California Combined Cycle Generation.

Figure 82. Annual Duration Curves - California Imports.

Figure 83. California Historical Hydro Operation - Sample May Week.

Figure 84. California Historical Hydro Operation May.

Figure 85. California Historical Hydro Operation June.

Figure 86. California Historical Hydro Operation July.

Figure 87. California Hydro Historical Monthly Duration Curves.

Figure 88. 2005 CEMS Data Delta Energy Center.

Figure 89. 2005 CEMS Data Haynes Generating Station.

Figure 90. 2005 CEMS Data Valley Generating Station.

Figure 91. Commitment Week of May 10

Figure 92. Dispatch Week of May 10

Figure 93. Ramp Rate and Range Capability - Week of May 10

Figure 94. Ramp Rate Down Capability - Week of May 10

Figure 97. Range Down Capability Versus California Load (2010X).

Figure 100. Annual Ramp Rate Down Capability.

Figure 101. Annual Ramp Rate Down Capability (zoom).

Figure 102. Operation of Helms Pumped Storage Hydro (2010X).

Figure 106. Annual Duration Curve California Hydro Generation (2010T).

Figure 107. Ramp Down Capacity With Constrained Hydro, 2010T.

Figure 110. Comparison of Helms Operation, Summer 2004 and 2006.

Figure 111. Sample Week of Pumping Operation, January 2006.

Figure 112. Historical Pumping Operation, 2006.

Figure 113. California Emission Reductions Due to Renewables (2010T).

Figure 114. WECC Emission Reductions Due to renewables (2010T).

Figure 121. Annual Duration Curve Path 66: COI (2010T).

5.1.3. Boundary Conditions

Figure 140. Temporary Curtailment of Wind Generation.