THE EFFECT OF CLEAN ELECTRICITY ON CO2 EMISSIONS FROM

PLUG-IN ELECTRIC VEHICLES

Samaneh Babaee*and Joseph DeCarolisNorth Carolina State University

Department of Civil, Construction, & Environmental Engineering

33rd USAEE/IAEE North American ConferenceOctober 2015 - Pittsburgh

* This work was performed while I was a doctoral student and postdoctoral fellow at NC State University. I am currently ORISE fellow at U.S. EPA, Research Triangle Park, NC.

2

Motivation

1. U.S. oil security and availability, anthropogenic climate change, and urban air quality

2. Policy or technological change in the electric sector electricity prices plug-in electric vehicle (PEV) deployment 3. Effect of clean electric sector scenarios that could reduce emissions over time and the efficacy of using PEVs to further reduce CO2 emissions 4. Several plausible options for targeted electric sector policy (e.g., the recent U.S. EPA proposed emission guidelines for CO2

emissions from fossil fuel-fired power plants)

Electric

Plug-in Hybrid

Hybrid Electric

Plug-in Hybrid

Electric

Electric

3

Objective

Under assumptions favorable to PEV deployment (high oil prices and low battery cost), quantify the effect of clean electricity scenarios on:

Technology deployment in the electric and light duty vehicle sectors

Electricity price Electric sector CO2 intensity and system-wide CO2 emissions

Electric

4

Overview on TIMES and NUSTD

Utilize a technology-rich energy system model comprised of: Model generator: The Integrated MARKAL-EFOM System (TIMES)

Energy-economy optimization framework to identify the least-cost way to satisfy end-use demands over the model time horizon

Inputs: National U.S. TIMES Dataset (NUSTD) Represents the U.S. energy system at the national scale and over the next 40 years Data include:

Model outputs Optimal installed capacity and utilization by technology, equilibrium energy prices, and emissions

Fuel prices Technology cost and performance estimates End-use demands

5

Sector Representation in NUSTD

NUSTD is publicly available at http://www.energy-modeling.org/

Model time horizon = 2010 to 2050 Model time periods = 5 years 3 seasons and 4 diurnal time segments A single region Social discount rate = 5%

6

Scenario Description

5 scenarios examined with and without PEV availability:

1. Base Case Scenario2. Renewable Portfolio Standard (RPS) Scenario3. EPA CO2 Rules Scenario4. Clean Energy Standard (CES) Scenario5. Low Wind and Solar Cost Scenario

7

Base Scenario Focused on High PEV Deployment

Highest PEV deployment scenario without new policy:High oil prices, Reference natural gas prices, Low battery cost, No RPS, No CO2

.

.

.

.

Scenario Nat Gas Price Oil Price RPS CO2 Policy Battery Cost EDV (% LDV

market)

1 High High Yes Yes Low 0.42

2 High High Yes Yes Ref 0.29

3 High High Yes Yes High 0.27

4 High High Yes No Low 0.42

5 High High Yes No Ref 0.27

6 High High Yes No High 0.24

7 High High No Yes Low 0.42

8 High High No Yes Ref 0.29

9 High High No Yes High 0.27

10 High High No No Low 0.42

11 High High No No Ref 0.28

12 High High No No High 0.24

13 High Ref Yes Yes Low 0.42

14 High Ref Yes Yes Ref 0.25

15 High Ref Yes Yes High 0.22

45 Ref High No Yes High 0.27

46 Ref High No No Low 0.42

47 Ref High No No Ref 0.28

48 Ref High No No High 0.24

49 Ref Ref Yes Yes Low 0.42

Base scenario:Highest PEV deployment without new policy (34% of LDVs).

.

.

.

8

Low Carbon Electricity Scenarios

Renewable Portfolio Standard (RPS) Scenario High oil prices, Reference natural gas prices, Low battery cost, RPS, No CO2

Based on the Title I of the American Clean Energy and Security Act of 2009 (H.R. 2454)

The minimum requirement for renewable energy generation is 9.5% in 2015, which gradually increases to 20% by 2020 and is held constant from 2020 to 2050.

Wind, solar PV and concentrating thermal, biomass gasification, and

incineration of municipal solid waste

9

Low Carbon Electricity Scenarios

The EPA CO2 Rules Scenario(High oil prices, Reference natural gas prices, Low battery cost, No RPS, No CO2) + EPA CO2 Rules in the electric sector

Based on the U.S. EPA proposed new source performance standard for CO2 emissions from new fossil fuel-fired electric generating units: 1100 lbs/MWh for new fuel-fired boilers, IGCC, and small gas-fired combustion turbines 1000 lbs/MWh for large gas-fired combustion turbines Emissions rate limits are applied to applicable new capacity in model year 2015 and remain in place through 2050. Based on the U.S. EPA proposed emission guidelines for existing fossil fuel-fired power plants (Option 1) which requires: A reduction in electric sector CO2 emissions below 2005 levels of

26% in 2020, 29% in 2025, and 30% in 2030 The 30% upper bound constraint on total CO2 emissions is

extended from 2030 to 2050.

10

Low Carbon Electricity Scenarios

Clean Energy Standard (CES) Scenario Based on Clean Energy Standard Act of 2012 (S. 2146) A minimum requirement for electricity purchase from clean power plants including solar, wind, geothermal, municipal solid waste, biomass, new nuclear, coal-based IGCC-CCS, and NGCC-CCS

Year Percent Clean Energy

CES 2012 This study

2015 24.0 NA

2020 39.0 24.0

2025 54.0 39.0

2030 69.0 54.0

2035 84.0 69.0

2040 NA 84.0

2045 NA 84.0

2050 NA 84.0

(High oil prices, Reference natural gas prices, Low battery cost, No RPS, No CO2) + Clean Energy Standard in the electric sector

11

Low Carbon Electricity Scenarios

Low Wind and Solar Cost Scenario(High oil prices, Reference natural gas prices, Low battery cost, No RPS, No CO2) + Low Wind and Solar Cost

Based on Azevedo et al. (2013), historical mean learning rate for wind is 16%, for solar PV and thermal is 22%.

Assuming four-fold increase in wind and solar, the capacity capital cost reduction is 30% for wind and 40% for solar PV and thermal by 2050 relative to 2015.

The capacity constraint requiring one unit backup capacity of natural gas turbine for every capacity unit of wind, solar thermal, and solar PV was removed.

12

Effect of Electricity Scenario on LDV Market Share

No PEVs With PEVs

Base and 4 Low Carbon Electricity Scenarios

0

1000

2000

3000

4000

5000

6000

7000

2010 2015 2020 2025 2030 2035 2040 2045 2050

LDV

Dis

tanc

e Tr

avel

ed (1

09km

)

Year

GaolineE85XDieselDiesel Hybrid

0

1000

2000

3000

4000

5000

6000

7000

2010 2015 2020 2025 2030 2035 2040 2045 2050

LDV

Dis

tanc

e Tr

avel

ed (1

09km

)

Year

GasolineE85XDieselDiesel HybridPHEV60BEV160

13

Average Annual Electricity Price

With PEVs

0

20

40

60

80

100

120

140

2010 2015 2020 2025 2030 2035 2040 2045 2050

Elec

tric

ity P

rice

($/M

Wh)

Year

Base

RPS

CES

EPA Rules

Low Wind/Solar Cost

14

Electricity Production

0

1000

2000

3000

4000

5000

6000

Elec

ltric

ity G

ener

ation

(TW

h/yr

)

Scenario with PEVs

NGCC-CCS

New solar

Existing solar

New geothermal

Existing hydro

New wind

Existing wind

New nuclear

Existing nuclear

New natural gas

Existing natural gas

New coal

Existing coal

2010 2030 2050

15

Electric Sector CO2 Intensity

0

100

200

300

400

500

600

700

2010 2015 2020 2025 2030 2035 2040 2045 2050

CO2

Inte

nsit

y in

Ele

ctri

c Sec

tor

(kg

CO2/

MW

h)

Year

Base: No PEVS BaseRPS: No PEVS RPSCES: No PEVS CESEPA Rules: No PEVS EPA RulesLow Wind/Solar Cost: No PEVS Low Wind/Solar Cost

HICC, MRO

RFCSERCTRE,FRCC

ASCC

WECC

NPCC

SPP

16

National U.S. CO2 Emissions

0

1000

2000

3000

4000

5000

6000

7000

2010 2015 2020 2025 2030 2035 2040 2045 2050

Tota

l sys

tem

-wid

e CO

2em

issi

ons (

mill

ion

tonn

es)

Base: no PEVS BaseRPS: no PEVS RPSCES: no PEVS CESEPA Rules: no PEVS EPA RulesLow Wind/Solar Cost: no PEVS Low Wind/Solar Cost

-4%; +0.5%

-9%; -6%

-20%; -3%

-36%; -5%

Incremental PEV Change

0%; +0.6%

Scenario Change

17

No difference in LDV market share across the various electricity scenarios. The direct effect of electric sector policies reducing national CO2 emissions is much larger than the effect produced by PEV deployment. The EPA Rules and Clean Energy Standard scenarios produce the largest incremental reduction in CO2 emissions across low carbon scenarios due to

PEV deployment. Wind and solar compete favorably against other low carbon options. In the absence of policy constraints, coal remains a viable generation option, particularly as natural gas prices increase over time. Reducing electric sector CO2 emissions will increase the efficacy of using PEVs to further reduce emissions without producing a significant effect on PEV cost-effectiveness. Policymakers must be attentive to electric sector developments when considering policy related to PEV deployment.

Insights

18

Questions and Comments …?

19

20

Overview on TIMES

Utilize a technology rich energy system model comprised of: Model generator: The Integrated MARKAL-EFOM System (TIMES)

Energy-economy optimization framework to identify the least-cost way to satisfy end-use demands over the model time horizon

Natural Gas Electricity

Combined-Cycle Gas Turbine

Capital CostVariable & Fixed O&M CostThermal EfficiencyEmissions Coefficients

.

.

.

Size of the box represents installed capacity

21

The TIMES Model Generator

The Integrated MARKAL (MARket ALlocation)-EFOM (Energy Flow Optimization Model) System (TIMES): Developed under the auspices of the International Energy Agency by the Energy Technology Systems Analysis Program (ETSAP1):

A widely used bottom-up energy system model Represents energy system as a network of technologies linked

together via flows of energy commodities Performs linear optimization to minimize the discounted cost of

energy over the model time horizon, represented by present value, subject to user-imposed constraints such as emissions limits. Decision variables are technology capacity and commodity flows over time(1) http://www.iea-etsap.org/web/index.asp

22

Reference Energy System

Gargiulo et al. (2011)

23

Scenario Description

Scenario development is focused on five factors likely to affect the cost-effectiveness of EDVs relative to other vehicle technologies:

Natural gas price Crude oil price EDV battery cost A federal cap on CO2 emissions A federal renewable portfolio standard (RPS)

Scenarios

TIMES+NUSTD(Bottom-Up

Energy System Model)

Fuel Prices Battery Cost CO2 Policy, RPS

Tech Deployment

Emissions

108

Factor Assumption 1 Assumption 2 Assumption 3

Natural gas price Reference Low High

Crude oil price Reference Low High

Federal CO2 cap No Yes

Federal RPS No Yes

Battery development Reference Low High

24

Baseline Assumptions

In the electric sector:

Constraints on SO2 and NOx emissions based on Annual Energy Outlook (AEO) results, Mercury and Air Toxics Standards (MATS), and the Cross-State Air Pollution Rule (CSAPR)

The overall minimum share of renewable energy for all states (including state-level RPSs): 2% in 2010 and 13% by 2025.

25

Baseline Assumptions

In the transportation sector:

New CAFE standard and the corresponding GHG emissions rate limit for LDVs: (49.6 mpg and 163 grams CO2 per mile in model year 2025) Renewable fuel requirements based on the Energy Independence and Security Act of 2007 and AEO2012

The effect of existing fuel subsidies and tax credits for new vehicles is included in the baseline cost assumptions (AEO).

NUSTD: Fuel Supply Sector

A set of exogenous and period-specific fuel prices for oil, coal, natural gas (NG), biomass, and nuclear

Coal supply: maintain distinction between coal types and sulfur level Oil, natural gas, nuclear, and biomass supply: include exogenously specified prices of refined oil, natural gas, nuclear, and biomass feedstocks based on the projected AEO1 fuel price

Emissions factors associated with the fuel production process (EPA2)

1 Annual Energy Outlook 2012 with projections to 2035, U.S. Energy Information Administration

2 U.S. Environmental Protection Agency, National MARKAL Database Documentation

26

27

NUSTD: End-Use Demand Sectors

Commercial, Industrial, and Residential sectors are comprised of: Single aggregate energy demand (AEO) Fuel share constraints (AEO) Emission factors associated with in-sector fossil fuel combustion (EPA, AEO)

0

2000

4000

6000

8000

10000

12000

14000

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2010 2015 2020 2025 2030 2035 2040 2045 2050

Tota

l Ene

rgy

Dem

and

(PJ)

Fuel

Sha

re (%

)

Electricity

Natural Gas

Distillate

LPG

Commercial sector

28

NUSTD: Electric Sector

32 electric generation technologies (coal, nuclear, natural gas, oil, hydro, wind, solar) 71 emission retrofit technologies to capture CO2, NOX, and SO2

emissions (EPA) Cost (capital, variable and fixed O&M), availability factor, efficiency, start year, lifetime, peak fraction, existing capacity (EPA and AEO) The price of electricity is determined endogenously.

29

NUSTD: Transportation Sector

Transportation sector: light duty vehicles, heavy duty vehicles, and off-highway technologies Light Duty Transportation Sector: 85 Light Duty Vehicles (LDVs)

7 vehicle size classes (Mini-compact, Compact, Full, Minivan, Pick-up, Small SUV, Large SUV)

6 fuels types (E10, E85X, Diesel, Electric, CNG, H2) 13 vehicle types Vehicles cost and performance data is driven from EPA, but is updated based on AEO. The total demand for vehicle miles is drawn from AEO. PEVs: plug-in hybrid (PHEV20 and PHEV60) and electric (BEV160) PEVs performance data is based on GREET model. Hurdle rates of 10% (hybrid, plug-in hybrid, electric, diesel, CNG, and

hydrogen fuel cell vehicles)

30

Technology Deployment in Two Extreme Scenarios

Lowest EDV deployment (Low oil, base NG & bat, no RPS & CO2) Highest EDV deployment (High oil, Low NG & bat, RPS & CO2)

0

1000

2000

3000

4000

5000

6000

7000

2010 2015 2020 2025 2030 2035 2040 2045 2050

LDV

Dis

tanc

e Tr

avel

ed (1

09km

)

Year

GSL E85X Diesel

0

1000

2000

3000

4000

5000

6000

7000

2010 2015 2020 2025 2030 2035 2040 2045 2050

Elec

tric

Gen

erati

on (T

Wh)

Year

Existing coal Existing natural gas New natural gas Existing nuclearNew nuclear Existing windNew wind Existing hydroNew geothermal New solar

0

1000

2000

3000

4000

5000

6000

7000

2010 2015 2020 2025 2030 2035 2040 2045 2050

LDV

Dis

tanc

e Tr

avel

ed (1

09km

)

Year

GSL E85XDiesel Diesel HybridPHEV60 BEV160

0

1000

2000

3000

4000

5000

6000

7000

2010 2015 2020 2025 2030 2035 2040 2045 2050

Elec

tric

Gen

erati

on (T

Wh)

Year

Existing coal New coalExisting natural gas New natural gasExisting nuclear New nuclearExisting wind New windExisting hydro New geothermalNew solar

Lowest EDV deployment (Low oil, High NG & bat, no RPS & CO2) Highest EDV deployment (High oil, Low NG & bat, RPS & CO2)

31

Effect of Scenario Drivers on EDV Deployment

32

Effect of EDV Deployment & Scenario Drivers on Emissions

Larger bubbles: higher oil price

0

1000

2000

3000

4000

5000

6000

7000

0% 10% 20% 30% 40% 50%

Estim

ated

205

0 CO

2 Em

issi

ons i

n (M

tonn

es)

2050 Light Duty Market Share of Electric Drive Vehicles

Battery Cost: LowBattery Cost: ReferenceBattery Cost: High

CO2

Scenarios with CO2 policy

33

Effect of EDV Deployment & Scenario Drivers on Emissions

• Larger bubbles indicate higher oil price• Scenarios in the dashed boxes are with CO2 policy

0

1000

2000

3000

4000

5000

6000

7000

0% 10% 20% 30% 40% 50%

Estim

ated

205

0 SO

2Em

issi

ons (

kton

nes)

Battery Cost: LowBattery Cost: ReferenceBattery Cost: High

SO2

34

Effect of EDV Deployment & Scenario Drivers on Emissions

• Larger bubbles indicate higher oil price• Scenarios in the dashed boxes are with CO2 policy

0100020003000400050006000700080009000

100001100012000

0% 10% 20% 30% 40% 50%

Estim

ated

205

0 N

Ox

Emis

sion

s (kt

onne

s)

Battery Cost: LowBattery Cost: ReferenceBattery Cost: High

NOx

35

The Sector-Specific Differences in 2050 Emissions

-400

-200

0

200

400

Supply Electricity LDV HDV OH End-Use Net

Emis

sion

s Ch

ange

(E

DV H

igh−

ED

V Lo

w)

Sector

SO2 (ktonnes) NOX (ktonnes) CO2 (Mtonnes)

-400

0

400

800

1200

Supply Electricity LDV HDV OH End-Use Net

Emis

sion

s Ch

ange

(E

DV H

igh−

ED

V Lo

w)

Sector

36

Research Contributions

Previous modeling work on PEVs My modeling work on PEVs

Employs sector-specific models that ignore the interaction of PEVs with the rest of the energy system.

Utilizes an energy system model, which accounts for interactions across all energy sectors.

Considers a single point in time. Incorporates a model time horizon extending from 2010 to 2050 with stage-wise decisions every 5 years.

Utilizes a complex energy system model to study a limited number of scenarios that are not specifically focused on the effect of PEV deployment.

Examines several factors likely to affect PEV deployment to create a large number of scenarios, which are used to quantify emissions impacts.

37

High oil prices and low EDV battery costs are the strongest drivers of EDV deployment. Across the 108 scenarios studied, there is no clear and consistent trend towards lower system-wide emissions of CO2, SO2, and NOX in the U.S. as EDV deployment increases. The incremental CO2 emissions benefit associated with PEV deployment largely depends on the evolving electricity generation mix and to a lesser extent changes across the broader energy system. Time-of-day PEV charging does not produce a significant impact on electricity prices, PEV deployment, or total system-wide CO2 emissions in the U.S. through 2050. The net effect of PEVs over time on national emissions will depend on a variety of factors beyond vehicle deployment numbers (e.g., new energy and environmental policies, prevailing fuel prices, and technology innovation). Incentivizing the purchase of PEVs will not automatically lead to emissions reductions.

Thesis Summary

38

Future Work

Develop a U.S. regional energy system database

Quantify the potential local air quality benefits associated with PEV deployment

Examine the effects of large scale EDV deployment on crude oil consumption, imports, and overall energy security

39

PHEV Representation in NUSTD

PHEV20 and PHEV60: E10-Electric and E85x-Electric:

Compact/full Minivan/Small SUV Large SUV/Pickup truck

Fuel economy improves from 2010 to 2020 based on GREET2012 Constant efficiency and fuel ratio from 2020 to 2050 Fractional distance traveled annually in CD mode for PHEV20=28%

and PHEV60=47% (Michalek et al., 2011)

Ratio (gasoline/electricity) =

Total gasoline consumption in CD CS modes

Total electricity consumption in CD CS modes

40

PHEV Representation in NUSTD

CD Efficiency Electricity CS (109 mile/PJ) Gasoline Compact/full E10-Electric (2010) PHEV20 0.28 0.326 6.35

0.72 93.65• All electric range: 20 km• Blended control strategy

PHEV60

0.47 0.352 17.5 0.53 82.5• All electric range: 60 km• All electric control strategy

CD/CS (Michalek, 2011), Fuel economy (GREET2012)

41

Hurdle rate

Compare the year-by-year cumulative discounted cost (estimated fuel cost and the annualized capital cost) for the alternative vehicle with the conventional gasoline vehicle The vehicle-specific hurdle rate is set such that the year-by-year cumulative discounted cost of the alternative vehicle exceeds that of the gasoline vehicle for 3-5 years past the 15-year lifetime of the vehicle. 2 categories of the alternative vehicles based on the current availability of required refueling infrastructure The highest vehicle-specific hurdle rate in each group is assigned to all the vehicles within the group Hybrid, plug-in hybrid, Electric, CNG, hydrogen fuel cell, and diesel vehicles = 10%

42

Time Slice Representation in NUSTD

Model time horizon = 2010 to 2050 (with 5-year time periods) Each year breaks down to 3 seasons: summer, winter, and intermediate (Fall + Spring) and 4 diurnal time segments: morning, afternoon/evening, peak, and night Use sub-annual timeslices to describe the changing electricity load within a year, which may affect the required electricity generation capacity and other commodity flows 2010 2015 2020 2025 2030 2035 2040 2045 2050

Summer Winter Intermediate

Morning Afternoon Night Peak

SM SA SN SP WM WA WN WP IM IA IN IP

Years

Seasons

Times of Day

Time Slices

43

Assumptions to the Annual Energy Outlook 2010 , page 91*Electricity Market ModuleTable 8.2. Cost and Performance Characteristics of New Central Station Electricity Generating Technologies

Samaneh:I updated these costs based on AEO2012 data

Technology

Online

YearSize (MW)

Leadtimes (Year

s)

Base Overnight Costs in

2010 ($2009/k

Project

Contingency Factor

Technologic

al Optimi

sm

Total Overnight

Cost in 2010 (2009

Variable O&M

($2010 mills/kWh

)Fixed O&M ($2010/kW)

Heatrate in 2011

(Btu/kWhr)

Heatrate nth-of-a-

kind (Btu/kWr)

efficiency, %

efficiency, %

Scrubbed Coal New 2013 600 4 2,658 1.07 1 2,844 4.25 29.67 8,800 8,740 39% 39%Integrated Coal-Gasification Combined-Cycle 2013 550 4 3,010 1.07 1 3,220 6.87 48.9 8,700 7,450 39% 46%IGCC with Carbon Sequestration 2016 380 4 4,852 1.07 1.03 5,348 8.04 69.3 10,700 8,307 32% 41% 3412/10700 3412/8307Conv Gas/Oil Comb Cycle 2012 250 3 931 1.05 1 977 3.43 14.39 7,050 6,800 48% 50% 0.3188785 0.410738Adv Gas/Oil Comb Cycle (CC) 2012 400 3 929 1.08 1 1003 3.11 14.62 6,430 6,333 53% 54%ADV CC with Carbon Sequestration 2016 400 3 1,834 1.08 1.04 2,060 6.45 30.25 7,525 7,493 45% 46%Conv Combustion Turbine 2011 160 2 927 1.05 1 974 14.7 6.98 10,745 10,450 32% 33% 3.14917937 3.06272Adv Combustion Turbine 2011 230 2 634 1.05 1 666 9.87 6.7 9,750 8,550 35% 40% 2.85756155 2.505862Fuel Cells 2012 10 3 5,918 1.05 1.1 6,836 0 350 9,500 6,960 36% 49%Advanced Nuclear 2016 1350 6 4,619 1.1 1.05 5,335 2.04 88.75 10,460 10,460 33% 33%Distributed Generation -Base 2012 2 3 1366 1.05 1 1,434 7.46 16.78 9,050 8,900 38% 38%Distributed Generation -Peak 2011 1 2 1,640 1.05 1 1,722 7.46 16.78 10,056 9,880 34% 35%Biomass 2013 80 4 3,519 1.07 1.05 3,859 5 100.55 13,500 13,500 25% 25%MSW - Landfill Gas 2010 30 3 7,694 1.07 1 8,233 8.33 378.76 13,648 13,648 25% 25%Geothermal 2010 50 4 2,393 1.05 1 2,513 9.64 108.62 9,760 9,760 35% 35%Conventional Hydropower 2013 500 4 2,134 1.1 1 2,347 2.55 14.27 9,760 9,760 35% 35%Wind 2009 50 3 2,278 1.07 1 2,437 0 28.07 9,760 9,760 35% 35%Wind Offshare 2013 100 4 4,345 1.1 1.02 5,974 0 53.33 9,760 9,760 35% 35%Solar Thermal 2012 100 3 4,384 1.07 1 4,691 0 64 9,760 9,760 35% 35%Photovoltaic 2011 5 2 4,528 1.05 1 4,755 0 16.7 9,760 9,760 35% 35%

http://www.eia.gov/forecasts/aeo/assumptions/pdf/electricity.pdf