Understanding Feasibility of Medium/Heavy-duty Plug-in ... Hoshing.pdf · Understanding Feasibility of Medium/Heavy-duty Plug-in Hybrid Electric Vehicles from Life-cycle Costs and

Understanding Feasibility of

Medium/Heavy-duty Plug-in Hybrid Electric Vehicles

from

Life-cycle Costs and Emissions Perspectives

Vaidehi Hoshing

Dr. Greg ShaverJune 5th, 2019

Motivation

Introduction

Frameworks and Results

Life-cycle Cost Analysis of PHEV Medium-duty Trucks and Transit Buses

Well-to-wheel Emissions Analysis for PHEV Transit Buses

Conclusions

2

Contents

Motivation

Introduction




Conclusions

3

Contents

Motivation

4Hybridization of vehicles presents a potential to reduce these emissions by reducing fuel consumptionDoes plug-in hybridization also have such a potential?

Light-Duty Vehicles60%

Medium/ Heavy-Duty Trucks 23%

Aircraft9%

Other 4%

Rail 2%Ships and Boats 2%

Greenhouse Gas Emissions in US in 2016

Transportation 29%

Agriculture 9%

Commercial &Residential 12%

Industry 22%

Electricity 29%

US Environmental Protection Agency (2018). Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016

Transit Bus

Two Applications of Interest

5

Medium-duty Truck

Cost is critical to fleet owners!

Transit Bus

Two Applications of Interest

6

MD TruckTransit

Bus

Days used/ year 300 300

Annual Vehicle Miles Traveled

25,000 30,000

Medium-duty Truck

Vehicle Usage Assumptions:

http://www.afdc.energy.gov/data/10309 - US 2015

http://www.afdc.energy.gov/data/10309

Series Architecture

Parallel Architecture

Hybrid Architectures Considered

Engine Generator Battery Motor

Engine

Battery Motor

7

3.33

6

0

2

4

6

2020 2025 2030

Economic Scenario300

150

0

200

400

2020 2025 2030

Economic Scenario

27.415.6

455358

0

200

400

600

0

50

100

2020 2025 2030

Economic Scenario

Fuel

Pri

ce

($/g

al)

Bat

tery

Pri

ce

($/k

Wh

)M

oto

r P

rice

8

Fixed Price ($)

Variable Price ($/kW)

Economic Assumptions

Parameter Value

Electricity Price $0.1/kWh

Electrical AC Charging Efficiency

90%

Battery End-of-Life capacity 70%

Vehicle Life 12 years

Other Constant Economic Assumptions:

At the end of life of the vehicle, the battery is assumed to have a salvage value proportional to the remaining battery life

U.S. Department of Energy. Annual energy outlook 2015, Tech. rep.; 2013V.T. Office. Overview of the DOE advanced power electronics and electric motor R & D program APEEM R & D Program Vehicle, Tech. rep.; 2014B. Nykvist, M. Nilsson, Rapidly falling costs of battery packs for electric vehicles, Nat Clim Change, 5 (4) (2015), pp. 329-332, 10.1038/nclimate2564

https://doi.org/10.1038/nclimate2564

Motivation

Introduction




Conclusions

9

Contents

Framework

10

• Drive cycle: (% time trace missed by > 2mph) < 2%• Gradability (7% at 20 mph)• Payback Period < 2 years• Battery Replacements <= 3 (vehicle life = 12 years)

Constraints for viability

Fuel Consumption

Battery Degradation

Model

Vehicle Parameters

Control Parameters

Sizing Parameters

Drive cycle

Capacity Loss

End of Life

Vehicle Simulation

Model

Economic Post-

Processing

Payback Period

Life-cycle Cost

Economic Scenarios

Application

Drive cycles

P&D Class 6

Refuse Truck

NY Comp.

Man-hattan

Orange County

ChinaNormal

China Agg.

Truck

Series 2030

Parallel

Bus

Series

Parallel

First Scenario of Economic Viability



11

Application

Drive cycles

P&D Class 6

Refuse Truck

NY Comp.

Man-hattan

Orange County

ChinaNormal

China Agg.

Truck

Series 2030 2030 2025

Parallel 2025 2025 2025

Bus

Series 2020 2025 2020 2020

Parallel 2020 2020 2020 2020


Bus application becomes economically viable earlier than Truck.



12

Application

Drive cycles

P&D Class 6

Refuse Truck

NY Comp.

Man-hattan

Orange County

ChinaNormal

China Agg.

Truck

Series 2030 2030 2025

Parallel 2025 2025 2025

Bus

Series 2020 2025 2020 2020

Parallel 2020 2020 2020 2020


Parallel architectures become economically viable earlier.Since they require smaller

motor.



13

Application

Drive cycles

P&D Class 6

Refuse Truck

NY Comp.

Man-hattan

Orange County

ChinaNormal

China Agg.

Truck

Series 2030 2030 2025

Parallel 2025 2025 2025

Bus

Series 2020 2025 2020 2020

Parallel 2020 2020 2020 2020


Urban drive cycles prefer hybridization earlier.



14

Motivation

Introduction




Conclusions

15

Contents

Framework

16

Total WTW Emissions

Vehicle Simulation

ModelInterface

GREET

Electrical Energy Used

Fuel Energy Used

WTW Emissions for Different Energy Sources

GREET (Greenhouse gases, Regulated Emissions, and Energy use in Transportation) is a well-to-wheel emissions database and tool developed by Argonne National Labs

0%

55%

6%

10%

1.5%

7%

5%

7%

7%1%

1%

13%

81%

0.3%

3%

California Indiana

17

Electricity Sources in California vs Indiana in 2016

A majority of California’s electricity comes from Natural gas-fired plants whereas that for Indiana comes from Coal-fired power plants.

http://www.energy.ca.gov/almanac/electricity_data/total_system_power.html https://www.energy.gov/sites/prod/files/2016/09/f33/IN_Energy%20Sector%20Risk%20Profile.pdf

Application

Drive cycles

P&D Class 6

Refuse Truck

NY Comp.

Man-hattan

Orange County

ChinaNormal

China Agg.

Truck

Series 2030 2030 2025

Parallel 2025 2025 2025

Bus

Series 2020 2025 2020 2020

Parallel 2020 2020 2020 2020

Example: Series Transit Bus Solution on Manhattan Cycle

18

MJ of fuel energy used per day

63% FC reduction

19

Daily VMT = 100

Electrical Energy

Diesel Energy

Diesel Energy

0

50

100

150

200

250

300

350

400

450

CO2 GHG100

Emis

sio

ns

(kg)

CO2 GHG100

20

Daily Emissions (kg) from Operation of Series Transit Bus PHEV on Manhattan Cycle

Daily VMT = 100

CaliforniaIndiana

LS Diesel

0

0.05

0.1

0.15

0.2

CO NOx SOx

Emis

sio

ns

(kg)

CO NOX SOX

0

50

100

150

200

250

300

350

400

450

CO2 GHG100

Emis

sio

ns

(kg)

CO2 GHG100

21


Daily VMT = 100

CaliforniaIndiana

LS Diesel

0

0.001

0.002

0.003

0.004

0.005

0.006

0.007

PM10 PM2.5 VOC

Emis

sio

ns

(kg)

PM10 PM2.5 VOC

0

0.05

0.1

0.15

0.2

CO NOx SOx

Emis

sio

ns

(kg)

CO NOX SOX

0

50

100

150

200

250

300

350

400

450

CO2 GHG100

Emis

sio

ns

(kg)

CO2 GHG100

22


Daily VMT = 100

CaliforniaIndiana

LS Diesel

Conclusions

Economic Viability Analysis

• Parallel architectures become viable for hybridization earlier than the series architecture.

• Transit buses become viable earlier than MD truck.

• Urban drive cycles are more favorable for hybridization

WTW Emissions from a Transit Bus PHEV in Indiana and California (not considering battery manufacturing)

• CO2 and GHG emissions reduce by about 60% in both the states

• A reduction in CO, NOx , PM2.5, VOC is shown

• PM10 and SOx emissions are shown to increase

23

Thank You!

24

Back-up

25

Economic Viability Analysis

• Parallel architectures become viable for hybridization earlier than the series architecture.• Transit buses become viable earlier than MD truck.• Urban drive cycles are more favorable for hybridization

26

Application

Drive cycles

P&D Class 6

Refuse Truck

NY Comp.

Man-hattan

Orange County

ChinaNormal

China Agg.

Truck

Series 2030 2030 2025

Parallel 2025 2025 2025

Bus

Series 2020 2025 2020 2020

Parallel 2020 2020 2020 2020

• Drive cycle: (% time trace missed by > 2mph) < 2%

• Gradability (7% at 20 mph)

• Payback Period < 2 years

• Battery Replacements <= 3 (vehicle life = 12 years)


27

• Global earth surface temperature has increased by 1°F since the 1970s.• Carbon dioxide constitutes 81% of the greenhouse gas emissions in the US

Motivation

Carbon Dioxide Information Analysis Center. 2010. http://cdiac.ornl.gov/National Oceanic and Atmospheric Administration. 2010. www.noaa.gov

http://cdiac.ornl.gov/

http://www.noaa.gov/

• 10 parameters; Sparse-sampled DOE size = 1300• Latin Hyper-Cube (LHC) design with “maximin distance” criterion

Parameter Units Min. Value Max. Value

Coefficient of drag - 0.58 0.88

Coefficient of rolling resistance - 0.005 0.007

Vehicle mass kg 12000 18000

M/G peak power kW 150 300

Battery energy capacity kWh 31 198

Engine/generator peak power kW 50 150

Maximum allowed C-Rate for Battery charge/discharge

- 1 4

Power filter parameter T1 - 0 0.5

Slope of Battery SOC regulation power demand (CSHEV mode)

W/SOC 200 20000

% Power demand over which engine is requested (CDHEV mode)

- 0.3 0.9

3 vehicle parameters

4 control strategy parameters

PowertrainSolution

28

Design of Experiments – Series Bus Example

3 sizing parameters

Application

Drive cycles

P&D Class 6

Refuse Truck

NY Comp.

Man-hattan

Orange County

ChinaNormal

China Agg.

Truck

Series(T=1300)

Parallel(T=800)

Bus

Series(T=1000)

Parallel(T=1300)

Scope



29

Application

Drive cycles

P&D Class 6

Refuse Truck

NY Comp.

Man-hattan

Orange County

ChinaNormal

China Agg.

Truck

Series 2030

Parallel

Bus

Series

Parallel




30

Simple Comparison of Drivecycles

31

More start stops and lower speed as compared to Orange County

0.0 1.7 3.3 5.0 6.7 8.3 10 11.7 13.3 15 16.7 18.3 20 21.7 23.3 25 26.7 28.3 30 31.7

Application

Drive cycles

P&D Class 6

Refuse Truck

NY Comp.

Man-hattan

Orange County

ChinaNormal

China Agg.

Truck

Series 2030 2030 2025

Parallel 2025 2025 2025

Bus

Series 2020 2025 2020 2020

Parallel 2015 2020 2020 2015


Bus application becomes economically viable earlier than Truck.



32

Application

Drive cycles

P&D Class 6

Refuse Truck

NY Comp.

Man-hattan

Orange County

ChinaNormal

China Agg.

Truck

Series 2030 2030 2025

Parallel 2025 2025 2025

Bus

Series 2020 2025 2020 2020

Parallel 2015 2020 2020 2015


Parallel architectures become economically viable earlier.Since they require smaller

motor.



33

Application

Drive cycles

P&D Class 6

Refuse Truck

NY Comp.

Man-hattan

Orange County

ChinaNormal

China Agg.

Truck

Series 2030 2030 2025

Parallel 2025 2025 2025

Bus

Series 2020 2025 2020 2020

Parallel 2015 2020 2020 2015


Urban drive cycles prefer hybridization earlier.



34

35

Additional Daily Emissions from Battery ManufacturingEm

issi

on

s (k

g)

Daily VMT = 100Battery Replacements = 3

CD CaliforniaCS CaliforniaCD IndianaCS IndianaLS Diesel

Note: The battery manufacturing emissions data available in GREET for NMC/Graphite was used. This does not consider the emissions due to battery recycling.

Emis

sio

ns

(kg)

Emis

sio

ns

(kg)

36

Emissions According to Source of Electricity (in g/MJ of electricity generated)

Emis

sio

ns

(g/M

J)

Emis

sio

ns

(g/M

J)

Emis

sio

ns

(g/M

J)

Electricity generation from coal produces most CO2 GHG and SOx

emissions.Electricity generation from biomass produces most PM, CO and NOx

emissions.

Natural Gas firedCoal firedNuclearBiomass

GeothermalOther renewables

– NMC+LMO/Graphite chemistry

𝑄loss,% = 𝑎𝑇2 + 𝑏𝑇 + 𝑐 𝑒𝑥𝑝 𝑑𝑇 + 𝑒 𝐼rate × 𝐴ℎthroughput + 𝑓𝑡0.5𝑒𝑥𝑝[−𝐸a/𝑅𝑇]

– Paper presents validation up to 6.5 C, model is used below 4C in our simulations to predict battery life with confidence

– We assume slow overnight charging at C-rates < 2C (No fast charging)

Combined calendar-life + cycle-life prediction model description – Wang et al. (JPS 2014)

Battery Degradation Model

Cyclic Aging Calendar Aging

37

Application

Drive cycles

P&D Class 6

Refuse Truck

NY Comp.

Man-hattan

Orange County

ChinaNormal

China Agg.

Truck

Series(T=1300)

Parallel(T=800)

Bus

Series(T=1000)

Parallel(T=1300)

Scope

38*T = Total simulations

Although electricity by itself produces more WTW emissions than diesel per MJ of usage, FC reduction combined with little electrical energy used for battery charging enables significant WTW emissions reduction.

39

Emis

sio

ns

(kg)

Emis

sio

ns

(kg)


Emis

sio

ns

(kg) CD California

CS CaliforniaCD IndianaCS IndianaLS Diesel

Daily VMT = 100

40

Emis

sio

ns

(kg)

Emis

sio

ns

(kg)


Emis

sio

ns

(kg)

Daily VMT = 100

CaliforniaIndiana

LS Diesel

Documents

Understanding Feasibility of Medium/Heavy-duty Plug-in ... Hoshing.pdf · Understanding Feasibility of Medium/Heavy-duty Plug-in Hybrid Electric Vehicles from Life-cycle Costs and