System Simulations of Hybrid Electric Vehicles with Focus ... · System Simulations of Hybrid Electric Vehicles with Focus on Emissions Author: Zhiming Gao Subject: Comparative simulations

Managed by UT-Battellefor the Department of Energy

System Simulations of Hybrid Electric Vehicles with Focus on Emissions

Zhiming Gao†Veerathu K. Chakravarthy†

Josh Pihl†C. Stuart Daw†

Maruthi Devarakonda‡Jong Lee‡

† Oak Ridge National Lab‡ Pacific Northwest national Lab

2010 DOE-DEER Conference September 27, 2010

Sponsor : Lee SlezakVehicle Technologies Program

U.S. Department of Energy

2

US fuel efficiency policy in 2009•Average fuel economy standard of 35.5 mpg in 2016•Nearly 10 miles per gallon better than the current average

Hybrid electrical vehicles (HEVs) technology•HEV with stoichiometric engines or lean-burn engines•Challenges for emissions control in HEVs

– Intermittent engine operation – A longer cold start at the beginning – Multiple cold starts during driving cycles– For diesel HEV, minimize fuel penalty without hurting emissions

DOE Vehicle Systems Analysis Technical Team (VSATT)• Powertrain System Analysis Toolkit (PSAT) developed at ANL•ORNL is tasked with studying after-treatment options

Background

3

VSATT Modeling Team

ORNL team : Stuart Daw, Kalyan Chakravarthy, Zhiming Gao

Testing data support: CLEERS, OEMs, National Labs (ORNL/PNNL)

Our mission is to evaluate the technologies and performance characteristics of advanced automotive powertrain components and subsystems in an integrated vehicle systems context. • Transient engine model and engine maps• Diesel Oxidation Catalyst (DOC)• Diesel Particulate Filter (DPF)• Lean NOx Traps (LNT)• Selective Catalytic Reduction (SCR)• Three-Way Catalyst (TWC)• HEV & Plug-in HEV

4

HEV simulations are integrated with transient engine and aftertreament models

Vehicle simulation framework● PSAT- ANL developed forward-

looking simulation package● Transitioning to AUTONOMIE,

which will replace PSAT

ORNL transient engine model● Estimate transient engine exhaust

properties and fuel economy based on corrections to steady-state maps

Aftertreatment models ● Low-order, physically consistent● TWCs, DOCs, LNTs, DPFs, SCRs● Development ongoing ORNL/PNNL

ORNL Aftertreatment

Models

ORNL Transient Engine Model

Fuel & Engine Exhaust Maps

PSATExperiment

PSATExperiment

Fuel Consumption & Tailpipe Emissions

5

ORNL transient engine model demonstrates a good prediction for engine-out emissions, exhaust temperature, and fuel economy associated with cold and warm starting conditions

* Int. J. Engine Res., 11(3), 2010, 137-151

Results :

Integrated mileage and engine-out emissions*

Example Simulation: Mercedes 1.7L diesel engine A UDDS cycle with cold start Civic vehicle configuration

Mileage (mpg)

CO (g/mi)

HC (g/mi)

NOx (g/mi)

PM (g/mi)

Exp 40.3 2.28 0.54 0.74 0.14Simu 40.4 2.29 0.54 0.89 0.12

050

100150200250300350400450

0 200 400 600Time (s)

Exha

ust T

empe

ratu

re (o

C)

ExperimentPSAT

0102030405060708090

0 200 400 600Time (s)

NO

x Em

issi

ons

(mg/

s) ExperimentPSAT

6

0

100

200

300

400

500

600

700

0 100 200 300 400Time (min)

NO

, NO

2, N

H3

(ppm

)

NO NH3

NO Inlet

Temperature (oC)

NO2

200ppmNH3

300ppmNH3

400ppmNH3

500ppmNH3

600ppmNH3

700ppmNH3

800ppmNH3

SCR

0.00.20.40.60.81.0

0 20 40 60 80 100Time (s)

CO

(g/s

)

PredictionExperimentTWC Inlet

TWC

Simulated Aftertreatment System

Stoichiometric engines• 1D TWC

Lean-burn engines• 1D LNT (SAE 2010-01-0082)• 1D SCR (Cu-ZMS-5)• DOC/LNT/CDPF • DOC/SCR/CDPF

The models were validated with public domain experimental data 0

100200300400500600700800

0 500 1000 1500 2000Time (s)

NO

x Em

issi

ons

(ppm

)

-1200-1000-800-600-400-200020040060080010001200

LNT-out SimulationLNT-out ExperimentLNT-inlet condition

LNT

0.0

5.0

10.0

15.0

20.0

0 2 4 6Time (hr)

Pres

sure

Dro

p (K

Pa)

PredictionExperiment

100% load

75% load

50% load

25% load

DPF

7

Results and Analysis

8

Std. gasoline vs. diesel baseline HEV comparison indicates large diesel fuel economy benefit

Diesel vs. Gasoline HEV:+13% energy density for diesel+6% engine efficiency for diesel

= +19% fuel economy for diesel (mpg)

Simulation parameters: 1450 kg HEV One UDDS cycle at hot start 1.3 kWhr battery charge (65%) 1.5-L gasoline & diesel engines No emissions control device

Diesel HEV without any aftertreatment: • 84.2 mpg diesel• 36% cycle average engine efficiency

Gasoline HEV without any aftertreatment:• 70.7 mpg gasoline (71.2mpg @ SAE 2007-01-0281)• 34% cycle average engine efficiency

+13%

+6%

Engine CO HC NOx PMGasoline (g/mi) 3.74 0.65 1.76 0.00

Diesel (g/mi) 0.44 0.11 1.14 0.62

Engine-out Emissions

0%

10%

20%

30%

40%

50%

0 10 20 30 40Engine Power (kW)

Engi

ne E

ffici

ency

GasolineDiesel

9

PHEV baseline comparison also indicates large potential diesel efficiency benefit similar to HEVSimulation condition:PHEV (1450kg HEV+120kg battery)Cold start, 5 kWh charge (100%)5 consecutive UDDS cycles 1.5-L gasoline and diesel enginesNo emissions control device

Results:Overall 19.9% better mpg for diesel (6% higher energy efficiency)

0

20

40

60

80

0 1372 2744 4116 5488 6860Time (s)

Vehi

cle

Spee

d (m

ph)

0.00

0.10

0.20

0.30

0.40

Fuel

Con

sum

ptio

n (g

al)

Vehicle Speed Diesel PHEV SI PHEV

0

50

100

150

200

250

1 2 3 4 5The Five Consecutive UDDS Cycles

Fuel

Eco

nom

y (m

pg)

0

50

100

150

200

250

Exp_SI_PHEVSimu_SI_PHEVSimu_diesel_PHEV

0

1

2

3

4

5

6

0 1 2 3 4 5The Five Consecutive UDDS Cycles

Bat

tery

Ene

rgy

(kW

hr)

0

1

2

3

4

5

6Exp_SI_PHEVSimu_SI_PHEVSimu_diesel_PHEV

* SAE 2007-01-0283

* *

10

However, lean NOx control has a big impact on expected diesel HEV efficiency advantage

Results:78.8 mpg diesel vs. 67.3 mpg gasoline 0.12 g/mile NOx vs. 0.12g/mile NOxLNT fuel penalty for diesel: 2.8%With LNT diesel efficiency advantage

just over 3%

Simulation parameters:1450kg HEV 1.3 kWhr battery (65% charge)Cold start UDDS drive cycle1.5-L gasoline w/ 2.2-L TWC 1.5-L diesel engines w/ 2.4-L LNTEngine fueling modulation for LNT

regeneration (e.g. 60s lean vs. 3 rich) 030

60

90

120

150

0 500 1000 1500Time (s)

NO

x em

issi

ons

(mg/

s) Gasoline engine-outTWC-out

0

20

40

60

0 500 1000 1500Time (s)

NO

x em

issi

ons

(mg/

s) Diesel engine-outLNT-out

LNT fuel efficiency penalty has spurred interest in SCR NOx control

11

Urea SCR saves diesel efficiency advantage, but causes extra NH3 slip and a slightly less NOx reduction

Results:

Simulation parameters:1450kg HEVCold start UDDS drive cycle1.3 kWhr battery (65% charge)1.5-L diesel engines w/ 2.4-L SCRUrea inj. for SCR (1:1 NH3 to NO)No NH3 slip control for SCR

TWC LNT SCRTailpipe NOx (g/mi) 0.12 0.12 0.20

Fuel Economy (mpg) 67.3 78.8 80.9Fuel penalty (%) 0.00 2.80 0.00

NH3 slip (g/mi) 0.00 0.00 0.04

0600

12001800240030003600420048005400600066007200

0 200 400 600 800 1000 1200 1400Time (s)

Engi

ne S

peed

(rpm

)

Veh

icle

Spe

ed (k

m/h

)

200

406080

100120

Engine off Engine on

0

20

40

60

0 500 1000 1500Time (s)

NO

x em

issi

ons

(mg/

s) Diesel engine-outSCR-out

12

We are using this SCR model to compare PHEVs with different NOx control technologies

Results:Tailpipe NOx: SCR=0.16g/mi; LNT=0.15g/miSCR generated 0.068g/mile NH3 emissions Fuel econ.: SCR=136.4mpg; LNT=133.8mpg1.9% penalty in fuel efficiency for LNT vs.

SCR (less LNT penalty than previous HEV case due to less NOx removal)

Simulation parameters: PHEV powered by 1.5-L diesel engine 5 kWh, 24 Ah battery charge (full charge) 5 UDDS cycles beginning with cold start 2.4-L LNT, 2.4-LUrea SCR LNT regeneration: 60s lean vs. 3 richUrea inj. for SCR (1:1 NH3 to NO)No NH3 slip control for SCR

01020304050607080

0 1372 2744 4116 5488 6860Time (s)

NO

x Em

issi

ons

(mg/

s) Engine-outSCR Outlet

High NOx spike

01020304050607080

0 1372 2744 4116 5488 6860Time (s)

NO

x Em

issi

ons

(mg/

s) Engine-outLNT Outlet

13

Key issue is the impact of the integration of aftertreatment device models on diesel HEV fuel efficiency and EmissionsSimulation parameters: 1450kg HEV w/ 80 UDDS drive cycles 1.5-L gasoline w/ TWC 1.5-L diesel engines w/

1): DOC/LNT/CDPF; 2): DOC/SCR/CDPF(DPF regen.: 600s if pressure drop >7.5kPa)

Results: Better fuel economy for DOC/SCR/CDPF One DPF regeneration event (for diesel HEV) CO, HC, and PM meet Tier 2 Bin 5, except NOx

0.00

0.02

0.04

0.06

0.08

0.10

TWC DOCLNTDPF

DOCSCRDPF

HC

Em

issi

ons

(g/m

ile)

Tier 2-Bin 5

0.00

1.00

2.00

3.00

4.00

5.00

TWC DOCLNTDPF

DOCSCRDPF

CO

Em

issi

ons

(g/m

ile)

Tier 2-Bin 5

0.00

0.10

0.20

0.30

0.40

TWC DOCLNTDPF

DOCSCRDPF

NO

x Em

issi

ons

(g/m

ile)

Tier 2-Bin 5

Challenge

0.00

0.01

0.02

TWC DOCLNTDPF

DOCSCRDPF

Soot

Em

issi

ons

(g/m

ile)

Tier 2-Bin 5

90%

95%

100%

105%

Gasoline Diesel withDOC/LNT/DPF

Diesel withDOC/SCR/DPF

Diesel withoutaftertreatment

Nor

mal

ized

Ene

rgy

Usa

ge

3.4%6.0%

-0.12%

14

We recently studied the effect of thermal insulation on HEV fuel efficiency and emissionsSimulation parameters: 1450kg HEV with 80 UDDS drive cycles Insulation material: 3.0mm mineral fiberAftertreatment device: Shell, Can,

Insulation layer, and Catalyst

Results: Improve fuel economyEnhance CDPF self regenerationReduce NOx and NH3 emissions

0.00

0.10

0.20

0.30

0.40

TWC DOCLNTDPF

DOCSCRDPF

NO

x Em

issi

ons

(g/m

ile)

w/o insulationw/ insulation

Tier 2-Bin 5

0.00

0.02

0.04

0.06

0.08

0.10

W/Othermal

insulation

W/thermal

insulation

NH

3 Slip

(g/m

ile)

01020304050607080

0 40000 80000 120000Time (s)

Soo

t Lay

er T

hick

ness

( µm

)

DOC/SCR/DPFw/o InsulationDOC/SCR/DPF w/Insulation

DPF regen

90%

95%

100%

105%

Gasoline Diesel withDOC/LNT/DPF

Diesel withDOC/SCR/DPF

Diesel withoutaftertreatment

Nor

mal

ized

Ene

rgy

Usa

ge w/o insulationw/ insulation

3.46.0

-0.12%

2.2

2.0

15

Summary Diesel HEV/PHEV achieve 19% higher fuel economy (mpg) than

gasoline HEV/PHEV without emission control device• 13% higher energy density for diesel• 6% better engine efficiency for diesel

NOx/PM emission control reduce diesel HEV/PHEV fuel efficiency advantages • LNT add about 2%-4% fuel penalty in HEVs • SCR saves diesel efficiency advantage, but causes extra NH3 slip • DPF add about 2%-3% fuel penalty in HEVs

The integrated system of DOC/SCR/CDPF • Save 3.4% diesel efficiency advantage• Meet Tier 2 Bin 5 regulation for CO, HC, and PM, except NOx• Good insulation boosts diesel efficiency advantage and emission reduction

NOx emissions control is still challenging for gasoline and diesel HEV/PHEV

16

Acknowledge

The authors thank Lee Slezak and U.S. Department

of Energy for support to this research

17

Thanks

18

ORNL 1D TWC model has been validated against independent OEM data

Model validation conditions: • Vehicle chassis data for gasoline engine• Supplied by OEM collaborator• UDDS cycle, cold start

Integrated emissions:• CO (g/mi): 0.833(exp) vs. 0.836(simu) • NOx (g/mi): 0.156(exp) vs. 0.157(simu) • HC (g/mi): 0.139(exp) vs. 0.148(simu)

0.00.20.40.60.81.0

0 20 40 60 80 100Time (s)

CO

(g/s

)


0.0

3.0

6.0

9.0

0 500 1000 1500Time (s)

Cum

ulat

ive

CO

(g)


0.00

0.03

0.06

0.09

0 20 40 60 80 100Time (s)

HC

(g/s

)


0.0

1.0

2.0

3.0

0 500 1000 1500Time (s)

Cum

ulat

ive

HC

(g)


0.00

0.04

0.08

0.12

0 20 40 60 80 100Time (s)

NO

x (g

/s)


0.0

1.0

2.0

3.0

0 500 1000 1500Time (s)

Cum

ulat

ive

NO

x (g

)


0

250

500

750

1000

0 500 1000 1500Time (s)

Tem

pera

ture

(K)


19

ORNL basic 1D DOC model has been validated against open literature experimental data

ORNL DOC model

• Three global reactions: (1) CO oxidation, (2) NO oxidation, (3) HC oxidation

Model validation conditions:

• Experimental data from the open literature (P. Triana, Dissertation, MTU, 2005)

Example results

• 5%-100% engine load at the engine speed of 1400rpm-2200rpm

0

100

200

300

0 200 400 600Temperature (oC)

CO

(ppm

)

ExpModel

0

200

400

600


NO

,NO

2 (pp

m) Exp_NO

Mode_NOExp_NO2Model_NO2

0%

40%

80%

120%


CO

Con

vers

ion

ExpModel

0

100

200

300


HC

(ppm

)

ExpModel

0%

30%

60%

90%


NO

Con

vers

ion Exp

Model

0%

40%

80%

120%


HC

Con

vers

ion

ExpModel

20

We have utilized literature data to construct an initial 1-D transient Simulink module for SCR-NOx control

SCR model assumptions:•Currently CuZSM5 catalyst•NH3 adsorption/desorption•NO SCR reaction•NO2 SCR reaction•Fast SCR reaction (NO + NO2)•NO and NH3 oxidation

SCR vs. LNT:•SCR uses urea for NOx reduction

•SCR does not require PGM catalyst

•No modulation of engine is required

•LNT causes fuel penalty for NOx reduction

•LNT requires PGM catalyst

SAE 2009-01-0897

Steady State Response

ORNL Model

Points from experiments by Olsson et.al, Applied Catalysis B: Environmental, 81(2008), 203-217. Lines from ORNL simulation.

0

100

200

300

400

500

600

700

0 100 200 300 400Time (min)

NO

, NO

2, N

H3 (

ppm

)NO NH3

NO Inlet

Temperature (oC)

NO2

200ppmNH3

300ppmNH3

400ppmNH3

500ppmNH3

600ppmNH3

700ppmNH3

800ppmNH3

21

ORNL DPF model for simulating soot emissions filtration compares well with open literature

Model validation conditions:

• Experimental data from the open literature (SAE 2003-01-0841)

Validation Results:• 25%-100% engine load at the

engine speed of 1800rpm

CDPF model assumptions:•Two-layer model•Thermal and catalytic oxidation

reactions

0.0

5.0

10.0

15.0

20.0

0 2 4 6Time (hr)

Pres

sure

Dro

p (K

Pa)


100% load

75% load

50% load

25% load

Engine load 25% 50% 75% 100%Exp (g) 18 35 103 55

Model (g) 18 34 108 53

Cumulative soot

(Cited from SAE 1999-01-0469)

22

0100200300400500600700800

0 500 1000 1500 2000Time (s)

NO

x Em

issi

ons

(ppm

)

-1200-1000-800-600-400-200020040060080010001200

LNT-out SimulationLNT-out ExperimentLNT-inlet condition

0100200300400500600700800

0 30 60 90 120 150 180Time (s)

NO

x Em

issi

ons

(ppm

) SimulationExperiment

LNT Inlet

LNT Outlet

ORNL LNT model for simulating lean NOx emissions compares well with observations

Simualtion Parameters• Engine: 1.7-L Mercedes diesel engine

• Steady-state engine operation (A): 57s Lean burn and 3s rich combustion

• FTP driving engine operation (B): a combined driving cycle of UDDS and US06

A

B

Steady-state operation

FTP driving operation

Successfully demonstrates reasonably agreement with observations for both steady-state and FTP driving engine operation (SAE 2010-0882).

LNT-out NOx(g/mi)

NOx Reduction(%)

Simu 0.05 94.1Exp 0.04 95.5

23

Thermal insulation improve on aftertreatment device thermal operation conditionsSimulation parameters:PHEV powered by 1.5-L diesel engine

5 kWh, 24 Ah battery charge (full charge)

5 UDDS cycles beginning with cold start

2.4-LUrea SCR (Cu-ZSM-5 catalyst)

Insulation material: 3.0mm mineral fiberAftertreatment device: Sheel, Can,

Insulation layer, and Catalyst

Results:Avoid multiple cold start in PHEVReduce NOx and NH3 emissions

0

50

100

150

200

250

300

350

0 1000 2000 3000 4000 5000 6000 7000Time (s)

Cat

alys

t Tem

pera

ture

(oC

)

0 mm insulation2 mm insulation5 mm inulsationAdiabatic

A long cold startat beginning

Multiple cold starts

0

10

20

30

40

50

60

0 1000 2000 3000 4000 5000 6000 7000Time (s)

SCR

-out

NO

x (m

g/s)

0 mm insulation2 mm insulation5 mm inulsationAdiabatic

24

ObjectivesDevelop engine maps and aftertreatment models for simulating the

performance of conventional, advanced hybrid and plug-in hybrid vehicles operating with gasoline, diesel and alternative fuels as well as advanced engine combustion modes

In the presentation

We focus on reporting simulated comparison of gasoline and diesel HEVs with explicit consideration of

the impacts of emissions control

25

Future work

Smart control strategies for utilization of ammonia in SCRs and reductant in LNTs

Potential impact of cold start or low temperature emissions traps technology

Optimization of integrated DOC/LNT/SCR/CDPF systems

Effects of advantage engine combustion (e.g. GDI, HCCI, PCCI) on improving fuel economy and emission reduction

Effects of waste energy recovery on improving fuel economy and emission reduction

Coordination• Close to coordination with OEM, national laboratories, and

universities to maintain relevance to the latest engine/emissions technologies for HEV/PHEV and industry needs

Documents

System Simulations of Hybrid Electric Vehicles with Focus ... · System Simulations of Hybrid Electric Vehicles with Focus on Emissions Author: Zhiming Gao Subject: Comparative simulations