modeling the details of u.s. renewable energy … · 2014. 9. 30. · national renewable energy...

Modeling the Details of U.S. Renewable Energy ElectricityRenewable Energy Electricity

2010 GTSP Technical Workshop

W lt Sh tWalter Short

May 27, 2010May 27, 2010

National Renewable Energy Laboratory Innovation for Our Energy Future

Contents

Complexities of modeling Renewable gElectric Technologies

Hierarchy of modelsHierarchy of models

U.S. Electric Sector Capacity Expansion -R EDS d lReEDS model

Hourly modeling - GridView model

From Global to Local – GCAM/ReEDS project

National Renewable Energy Laboratory Innovation for Our Energy Future2

project

Complexities of Modeling Renewable Electric Technologies - Attributes

• Resource variability• TemporallyTemporally

• Diurnally, seasonally, annually, susceptibility to climate• Uncertainty – forecast errors

• Spatially• Spatially• Locally • Regionally /diversity• Non portability of the resource wind solar hydro• Non-portability of the resource – wind, solar, hydro

• Transmission requirements

• High Capital, low operating costs• Nascent industries

• Modularity of the technology – costs, transportability• Environmental impacts visual large land areas etc• Environmental impacts – visual, large land areas, etc.


Complexities of Modeling Renewable Electric Technologies - Impacts

• Resource variability• Temporal – the smaller the time step, the better

• Capacity value, operating reserves, curtailments, transmission req’ts,Capacity value, operating reserves, curtailments, transmission req ts, storage requirements

• Dynamic

• Spatial – the smaller the regions, the betterp g ,• Inter-region transmission • Diversity difficult to capture, even with small regions

• High capital costs; low operating costs• High capital costs; low operating costs• Economic and financial parameter values have huge impacts• Nascent industry growth rates impact prices

• Modularity• Modularity of many renewables not recognized by most models• World markets for technologies – solar windWorld markets for technologies solar, wind

• Environmental impacts – visual, sound, local codesNational Renewable Energy Laboratory Innovation for Our Energy Future4

Actual Models

• Limitations• Run time and memory requirements• Data inputs• Analyst time

• Existing model limitations• Existing model limitations• Very aggregated time slices (GCAM 4, NEMS 11, ReEDS

17)Li it d b f i (GCAM 1 f U S NEMS 13• Limited number of regions (GCAM 1 for U.S., NEMS 13, ReEDS 358)

• Limited transfers between regions – no optimal power flow• Aggregated economic, financial, tax structure• Continuous plant sizes• World technology-supply markets largely ignoredgy pp y g y g• Market structure largely ignored


Hierarchy of Models

• No one model can “do it all”• Different model purposes call for different areas of

focusGCAM i t t d t d li• GCAM – integrated assessment modeling

• NEMS – national forecast and policy analysis• ReEDS – potential of central renewable electric technologies• SolarDS – potential of photovoltaics on buildings• GridView – 8760 hours/year production cost with optimal

power flow model (ABB product)p ( p )

• A hierarchy of models allows one to come closer to “doing it all”


NREL’s Electricity Market Model Hierarchy

does it work

ReEDS GridViewhourly?

ft PV

SolarDS

rooftop PVpenetration


SolarDS

PNNL/NREL Modeling Construct

GCAMGCAM

SGridView

does it work hourly?

ReEDS



SolarDS

C it i & di t h f th US l t i it t t t 2050 i l di

ReEDS Overview• Capacity expansion & dispatch for the US electricity sector out to 2050 including

transmission & all generator types – AP coal, IGCC coal w/wout CCS, Gas CTs, Gas CCs w/wout CCS, oil/gas steam, nuclear, hydro, PHS, major renewables

Mi i i t t l t t i h 2 i t t i d• Minimize total system cost in each 2-year investment periodo All constraints (e.g. balance load, reserves, etc.) must be satisfied o Investment decision based on 20 year present value cost of each technology

M lti i l• Multi-regional o 3 interconnections – separately synchronizedo 13 NERC subregions – fuel price differenceso 48 states – incentives, state RPSs, etc.

134 t l (PCA ) b l ti & l do 134 power control areas (PCAs) – balance generation & loado 356 wind/csp regions – resource availability/quality, grid connectiono Enables transmission capacity expansiono Enables treatment of the variability of wind/solar

• Temporal Resolutiono 17 timeslices in each year

4 typical days representing the 4 seasonsh t i l d t d b 4 ti li ( i ft i i ht) each typical day separated by 4 timeslices (morning, afternoon, evening, night)

1 superpeak summer afternoon timesliceo Statistical treatment of variability of wind & non-dispatchable solar


Region DefinitionsRegion Definitions

Time-slice DefinitionsTime-slice Definitions

ReEDS Processing

Loads and load growthLoads and load growth Transmission RequirementsTransmission Requirements

Load RequirementsLoad Requirements

Initial Gen CapacitiesInitial Gen Capacities

Fuel price forecastsFuel price forecasts

Installed CapacityInstalled Capacity

New Generating CapacityNew Generating Capacity

New Transmission CapacityNew Transmission Capacity

2-yearReEDS2-yearReEDS

Technology Cost/ Performance ForecastsTechnology Cost/ Performance Forecasts

Resource dataResource data

Fuel PricesFuel Prices

Technology Cost/ Performance DataTechnology Cost/ Performance Data

Wind VariabilityWind Variability

DispatchDispatch

CapacityCapacityReEDS Optimization

ReEDS Optimization

Transmission DataTransmission Data

State/Federal Rules/IncentivesState/Federal Rules/Incentives

System RequirementsSystem Requirements

Wind Variability ParametersWind Variability Parameters

System RequirementsSystem Requirements

Economic AssumptionsEconomic Assumptions

Electricity PriceElectricity Price

National Renewable Energy Laboratory Innovation for Our Energy Future

Fuel Demand ForecastsFuel Demand Forecasts

Electricity Price ForecastsElectricity Price Forecasts Electricity PriceElectricity Price

Fuel UsageFuel Usage

Wind and PV Capacity ValueConv - Load WindConv Load Wind

Wind

NameplateCapacity

Available Capacity

Load

Conventionals

0 1000 2000 3000 4000 5000 6000 7000

Capacity

0 1000 2000 3000 4000 5000 6000 7000

Conv - LoadConv - Load

Wind + Conv - Load

-1000 -500 0 500 1000 1500 2000 2500 3000

LOLP

-1000 -500 0 500 1000 1500 2000 2500 3000

Wind Capacity Value/MW =DL/Wind Nameplate Capacity

W+C-(L+DL) W+C-L

DL

DL/Wind Nameplate Capacity

-1000 -500 0 500 1000 1500 2000 2500 3000 3500

LOLP

Value of Diversity

Wind and PV CurtailmentsConventionals and Load WindConventionals and Load Wind

Load

Conventionals

Wind

NameplateCapacityAvailable

0 1000 2000 3000 4000 5000 6000 70000 1000 2000 3000 4000 5000 6000 7000

CapacityAvailable Capacity

Load

Conventionals

Must Run

0 1000 2000 3000 4000 5000 6000 7000

0 1000 2000 3000 4000 5000 6000 7000

Must Run - Load

Must Run - Load

Wind + Must Run - Load

-4000 -3500 -3000 -2500 -2000 -1500 -1000 -500 0-5000 -4000 -3000 -2000 -1000 0 1000 2000 3000

Surplus

Wind and PV Induced Operating Reserve

O ti d i bOperating reserves driven by– contingency and regulation reserves – Load/wind following reservesLoad/wind following reserves

Load/wind following reserves – Based on errors in load and wind forecasts

Operating reserves comprised of – Spinning reserves– Quick-start capability– Interruptible load

20% Wind Energy by 2030(Results from ReEDS/WinDS model)

6000

5000

6000

Wh)

3000

4000

y G

ener

ated

(TW

Direct Electric System Cost

1000

2000

Elec

tric

ity

WindHydroNatural GasNuclear

20% Wind Scenario Costs 2% More

0

2000

2002

2004

2006

2008

2010

2012

2014

2016

2018

2020

2022

2024

2026

2028

2030

Coal


20% Wind Energy by 2030

Detailed Studies Following the 20% Wind Energy by 2030 StudyWestern

EASTERN WIND INTEGRATIONAND TRANSMISSION STUDY


ReEDS Now Includes All Major RETsWINDWIND

CSP

GEO


BIO


GCAMGCAM

SGridView


ReEDS



SolarDS

GridViewCommercial production-cost (8760 hour simulation with unit commitment), optimal DC power flow model by ABB

Used by ISOs, utilities, and others for wide variety of purposes

11 000 85 000 transmission11,000 generators

85,000 transmission lines

GridView Representation of ReEDS Results

Buses/substations

R EDS iReEDS region

transmission lineslines

GridView - 2050 U.S. Electricity Demand

700000

800000

500000

600000

400000

500000

MW

200000

300000

0

100000

0

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

RE Generation from the High RE Scenario

700000

800000

500000

600000

400000

500000

MW

200000

300000

0

100000

0


Thermal generation

700000

800000

500000

600000

400000

500000

MW

200000

300000

0

100000

0


Actual renewable generation

700000

800000

500000

600000

400000

500000

MW

200000

300000

0

100000

0


Curtailment (14% of renewable gen)

700000

800000

500000

600000

400000

500000

MW

200000

300000

0

100000

0


Curtailment by PCA (TWh)

< 5

Curtailment (TWh)

50 ‐ 75

25 ‐ 50

5 ‐ 25

< 5

> 125

100 ‐ 125

75 ‐ 100


GCAMGCAM

SGridView


ReEDS



SolarDS

GCAM/ReEDS Project Purpose

• Compare model results and identify reasons for p ydifferences

• Learn from each model and adopt improvements

Develop a joint capability that allows analysis from• Develop a joint capability that allows analysis from global to local


Brief Comparison of ReEDs and GCAM US Electric Sectors

GCAM ReEDSGCAM ReEDSRegional Coverage

Global model with 14 regions. U.S. is one region.

U.S. only; 358 subregions built from county data

Sectoral Coverage

Full agricultural and energy sector coverage Electricity sector only

Solution M h i

Dynamic recursive, i ilib i

Linear Optimization sequentially h h i i h li i d f i hMechanism economic equilibrium through time with limited foresight

Technology Choice

Logit choice approach forces diversification

Optimization with regional variations in resource, prices

Capacity Expansion

New capacity is coupled with new generation

Explicit generation capacity expansion

Reserves Not modeled Reserve margin and operating Reserves Not modeled. g p greserve requirements

DispatchLogit choice for investment, continued operation of

it l t k i tOptimal dispatch in 17 timeslices

30

capital stock vintages

Transmission Average transmission costs, no capacity constraints

Explicit transmission capacity expansion and transmission use

First Comparison Runs – Carbon Cap Case

GCAM ReEDS

6000

7000

8000

6000

7000

8000

4000

5000

TWh/yr

4000

5000

TWh/yr

1000

2000

3000

1000

2000

3000

Coal w/o CCS Coal‐CCS Oil w/o CCS Oil‐CCS Gas w/o CCS

0

2005 2020 2035 2050

0

2006 2020 2034 2048

31

Gas‐CCS Nuclear Hydro Geothermal Biopower

Wind CSP Central PV Distributed PV

Harmonization Steps

Advanced Supply (allow nuclear and CCS construction)Advanced Supply (allow nuclear and CCS construction)1. Unharmonized (only load and carbon cap are synchronized)2. Harmonized model inputs (tech and fuel costs, emissions rates) 3 Harmonized capital cost financing (use fixed charge rate)3. Harmonized capital cost financing (use fixed charge rate)4. Harmonized annual generation for biopower and distributed PV5. Harmonized annual generation for nuclear power6 Harmonized annual generation from all conventionals6. Harmonized annual generation from all conventionals7. Completely harmonized (renewables and conventional

Harmonized Runs - Carbon Cap Case

GCAM ReEDS8000

Tech costs, finance and taxes, emission rates, biopower, nuclear, PV

6000

7000

8000

6000

7000

8000

4000

5000

TWh/yr

4000

5000

TWh/yr

1000

2000

3000

1000

2000

3000

Coal w/o CCS Coal‐CCS Oil w/o CCS Oil‐CCS Gas w/o CCS

0

2006 2020 2034 2048

0

2005 2020 2035 2050

33

Gas‐CCS Nuclear Hydro Geothermal Biopower

Wind CSP Central PV Distributed PV

Continued collaboration will lead to improvements of both models and a joint capability

GCAM benefits from updated supply curves from ReEDS for renewable technologies.g

GCAM also benefits from ReEDS temporal detail with addition of base, intermediate, and peak competition in the electric sector.

GCAM will benefit from ReEDS modeling of the grid integration ofGCAM will benefit from ReEDS modeling of the grid integration of intermittent resources and reliability.

ReEDS benefits from GCAM experience with technologies that are not currently fully functional in ReEDScurrently fully functional in ReEDS.

– Rooftop PV.– Biogasification Combined Cycle with and without CCS.

ReEDS will benefit from information about other sectors from GCAM.ReEDS will benefit from information about other sectors from GCAM.– Biofuels influence on biopower.– Influence of heating demand on natural gas prices.

34

Conclusions

• Simulating renewable electric technologies requires extra-ordinary detail

• A hierarchy of models provides insights and checks on higher level models


modeling the details of u.s. renewable energy … · 2014. 9. 30. · national renewable energy...

Documents