1

Da Yu Wang, Sheng Yao, David Cabush, DaveRacine

DEER 2007

Ammonia Sensor For SCR NOX Reduction

Agenda

NH3 Sensor Usage in SCR system Overview NH3 Sensing Technologies Functionality of NH3 Sensor NH3 Sensor Design Test Results

– Cross Sensitivity – Test results

Close Loop Control of the SCR System with aNH3 Sensor - System Advantages Summary

2

3

Diesel SCR System

DOC

SCR

Urea tank

Urea Level switch

SCR out T°

Turbine Wide range O2 Delphi NH3 sensor

SCR control sensor SCR control actuator Standard control sensor

Air + urea

Urea Air

Urea pump on/off Compressor on/off

[NH3] out (Delphi) SCR in T°

Dosing Unit

Feedback for closed loop control

Source of Ammonia: Urea through hydrolysis reaction (NH2)2CO + H2O > 2NH3 + CO2

DeNOx Reaction: At favored temperature 4NH3 + 2NO2 + 2NO > 4N2 + 6H2O

•

4

Sensing principle Non-equilibrium electrochemical sensing principle

– Proprietary NH3 sensing electrode materials – Both sensing and reference electrodes exposed to the engine

exhaust – Solid oxide electrolyte used as the sensor body

)(2

)(4

)(3 223 OHONH PLn

e kTPLn

e kTPLn

e kTEMF −−≈

•(Heater circuit not shown)

O2-

O2 + 4e­ 2O2-

Porous Protection

Lean Exhaust

Lean Exhaust

TEMP SENSING CELL

NH3 SENSING CELL

Lean Exhaust

ZIMP

EMF 3O= + 2NH3 N2 + 3H2O + 6e

Theory Semi-log output of EMF versus NH3 concentration

180

60

0 0 20 40 60

NH3 [PPM]

)

EMF

[mV]

kT kT kT 120

EMF≈ Ln(PNH ) − Ln(PO ) − Ln(PH2O3e 4e 2e3 2

– Interference effect due to H2O and O2 concentrations changing due tocombustion can be self-compensating

– The concentration of H2O and O2 vary in opposite directions as functionof A/F ratio minimizing effect

5

6

Sensor structure

Contact pads : signal

+

-

Contact pads : heater

Protective layer (porous and dense alumina)

Sensing electrode

Reference electrode

Exhaust chamber

Structural layer & Temp Sensor (dense alumina & Zirconia)

Heater circuit

Protective layer (dense alumina)

Zirconia layer

+

-

Planar structure – Co-fired zirconia and alumina

layers with NH3 sensing,platinum reference electrode andheater circuit

Key Features – Integrated heater provides fast

time to activity – Temperature sensor included – No air reference – Alumina layers provide electrical

isolation between heater and sensor circuits

– Porous protection provides excellent exhaust poisonresistance

– Small size

Sensing Element Finished sensing elements

– Monolithic thick film multi layer composite substrate

» Alumina / Zirconia composite

» Alumina provides toughness - Zirconia electrolyte – Integral Heater and Temperature Sensor for heater control – Compatible with existing sensor packaging technology

NH3 electrode material applied on substrate surface

A poison protective material is applied over the NH3 electrode

Finished sensing element looks like exhaust oxygen sensing element

7

Sensor package

Based on proven robust production planarsensor packaging

Package has capability beyond dieselexhaust temperatures

Lower shielding can be modified accordingto customer application

8

9

NH3 Sensor System Mechanization Mechanization – Early systems being developed for commercial vehicle applications – Stand-alone electronic interface with CAN link to vehicle – A-sample hardware shown below.

» A-Sample systems provide either Analog or CAN message output

Advanced Development Hardware

10

NH3 Sensor Performance Targets Measurement range: 0 – 100ppm NH3

– Tolerance: ± 5ppm NH3 at 10 ppm NH3. – Acceptable gas content without interference: NO, HC, CO, N2O. – Acceptable gas content with cross sensitivity: O2, H2O.

Temperature range: – Functional: 200 ºC to 450 ºC – Non functional: - 40 ºC to 700 ºC

Durability target: 5.000hours / 250.000km

NOx Exhaust gas content: 0 to 500ppm (sensor performance within spec)

H2O exhaust gas content: 1% to 8% by mass (sensor performance within spec)

Response Time: T60 = 3 s T90 = 5 s

Thermal Shock – Two layers of protection

1. Double layer protective shield 2. System algorithm to disable sensor when liquid water is possible in the exhaust

– Heated ceramic sensors must be protected from contact with liquid water to prevent damage do to thermal shock

NH3 Sensor Interface Electronics

Environment – Ambient Temperature (electronics): - 40 ºC to 105 ºC

Electrical – Sensor system compatible with either a 12V or 24V vehicle electrical

system – Sensor system communicates to vehicle over a CAN bus

Mounting/Installation – Sensor mounts directly to exhaust pipe via a M18x1.5 threaded boss – Sensor must be mounted 10 º above horizontal to prevent pooling of

water in shield

11

12

Water and oxygen interference effect

Water and oxygen have opposite interference effect Self compensation effect is possible in a narrow range as shown in following figure (confirmed by lab gas bench too) Climate air humidity difference is a main concern but can be handled by calibration Model based correction is also possible if A/F ratio and air humidityinformation is available

-15

-10

-5

0

5

10

15

15 25 35 45 55 65 AIR TO FUEL RATIO

D E

MF

[mV]

0

2

4

6

8

10

12

14

16

H2O

, O2

[%]

D EMF [mV]

H2O [%]

O2 [%]

Oxygen

Water

D EMF Δ

NH3 Output in the Presence of NOX

Basic function demonstrated in gas bench testing– NH3 signal free from NOX interference – Some interference below 10ppm set point

-160 160 NOX=0 ppm NO=400 ppm NO/NO2= 200/200 ppm NO2=400 ppmNOX=0 ppm NO=400 ppm NO/NO2=200/200 ppm NO2=400 ppm

NH3=100 ppm

50 ppm

-80 25 ppm 80

10 ppm 5 ppm

0 0

10 ppm Control Point NH3

80 -80 0 20 40 60 80 100

TIME (M)

OU

TPU

T [m

V]

13

14

Sensor performancecross interference

NH3 sensing accords to the theory prediction (semi-logarithms equation)

– Data obtained at lab gas bench, 14% O2, 1.5% H2O Interference from NO, CO, and HC is negligent

#53 BiVO4 (5% Mg, 5% Ta) AFTER 570 H, 800C

0

40

80

120

160

0 20 40 60 80 100 120 NH3 CONCENTRATION (ppm)

OU

TPU

T (m

V)

NH3 ONLY

100 ppm NO

100 ppm NO +1000 ppm CO

100 ppm NO +1000 ppm CO +1000 ppm HC

500 ppm NO

1000 ppm NO

1000 ppm NO

1000 ppm HC

1000 ppm CO

Log. (NH3 ONLY) NH3 cell only

Engine Test Stand PerformanceSteady State Test

NH

3_Se

nsor

[ppm

]

NH3 Sensor Performance--Test 2a (steady-state) NH3 Sensor Performance--Test 2a (steady-state) Temp=200 C; 17-March-2006 Temp=400 C; 17-March-2006

50 50

Sensor C4 Sensor C4 40 Sensor C6 40 Sensor C6

30

20

10 NH

3_Se

nsor

[ppm

]

30

20

10

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

NH3_LDS [ppm] NH3_LDS [ppm] Sensor data

– Steady state performance on HD diesel engine – Test stand fitted with SCR aftertreatment system – Sensors tested at 200 C & 400 C steady-state exhaust gas conditions – Ammonia slip between 0 and 50 ppm – Chart shows sensor output plotted against LDS instrumentation NH3 output signal

Result – Sensor accuracy within specification limits (+/- 5ppm at 10 ppm control point)

15

Engine Test Stand PerformanceESC Cycle

NH3 Sensor output during an ESC cycle on the engine test stand – NH3 sensor output tracks LDS instrumentation over a dynamic test cycle

NH3 Sensor Perfromance Test--Test 2b ESC Engine cycle; 17-March-2006

50

IDEAL 40 Sensor C4

Sensor C6

30

20

10

0 0 10 20 30 40 50

NH3_LDS [ppm]

NH

3_Se

nsor

[ppm

]

No poison effects after 700h test cycle without DPF16

Sample to Sample Variation

100 100

200C 300C

SEN

SOR

_NH

3_[P

PM]

10 S_1_DA_1 Final S_2_DA_2 S_3_DA_1 S_4_DA_1 S_5_DA_1 S_1_DA_1 S_2_DA_2 S_3_DA_1 SE

NSO

R_N

H3_

[PPM

]

10 S_1_DA_1 Final S_2_DA_2 S_3_DA_1 S_4_DA_1 S_5_DA_1 S_1_DA_1 S_2_DA_2 S_3_DA_1

S_4_DA_1 S_5_DA_1 S_4_DA_1 S_5_DA_1 S_1_DA_1 S_2_DA_2 S_1_DA_1 S_2_DA_2 S_3_DA_1 S_4_DA_1 S_3_DA_1 S_4_DA_1 S_5_DA_1 S_1_DA_1 S_5_DA_1 S_1_DA_1 S_2_DA_2 S_3_DA_1 S_2_DA_2 S_3_DA_1 S_4_DA_1 S_5_DA_1 S_4_DA_1 S_5_DA_1 S_1_DA_1 S_2_DA_2 S_1_DA_1 S_2_DA_2 S_3_DA_1 S_4_DA_1 S_3_DA_1 S_4_DA_1 S_5_DA_1 S_1_DA_1 S_5_DA_1 S_1_DA_1 S_2_DA_2 S_3_DA_1 S_2_DA_2 S_3_DA_1 S_4_DA_1 S_5_DA_1 S_4_DA_1 S_5_DA_1

1 1 1 10 100 1 10 100

LDS_NH3_[PPM] LDS_NH3_[PPM]

100 • 30 sensors

400C • Same conversion equation

SEN

SOR

_NH

3_[P

PM] • No water-oxygen adjustment

10 S_1_DA_1 Final S_2_DA_2

S_3_DA_1 S_4_DA_1 • No air humidity adjustment S_5_DA_1 S_1_DA_1

S_2_DA_2 S_3_DA_1

S_4_DA_1 S_5_DA_1

S_1_DA_1 S_2_DA_2

S_3_DA_1 S_4_DA_1

S_5_DA_1 S_1_DA_1 • Red lines mark ±50% at 10 PPM NH3 S_2_DA_2 S_3_DA_1

S_4_DA_1 S_5_DA_1

S_1_DA_1 S_2_DA_2

S_3_DA_1 S_4_DA_1

S_5_DA_1 S_1_DA_1

S_2_DA_2 S_3_DA_1

S_4_DA_1 S_5_DA_1

1 1 10 100

LDS_NH3_[PPM]

17

18

Exhaust Ammonia (NH3) Sensor - Optimized Performance of an SCR Converter

Graphical Representation of Key Factors to SCR Control

NH

3 , N

OX

[ppm

]

Urea Dosing

SCR

Effi

cien

cy

Open Loop Control

Control with NOX Sensor

NOX – Sensor

NH3 – Sensor (NH3 Slip) Tailpipe

NOX Emissions

Euro 4 limits can be achieved by usingopen loop control urea dosing

– SCR efficiency approximately 65%

The realization of maximum NOx conversion (without using a postoxidation catalyst) is only possible withclosed loop controlled Urea dosing:

– A NOx based SCR control does not enable the use of the maximum NOx conversion because of NH3 cross sensitivity of NOx sensor.

– The NH3 based SCR control enables operation at the conversion limit of the catalyst:

– >90% NOx conversion possible for highwadriving conditions

– Minimal catalyst volume

Control with NH3 Sensor @ 10 ppm NH3 Slip

☺☺

19

Conclusion

Sensor demonstration has been done – NH3 sensing accords to concept

Interference with NO, HC, CO is not significant O2 and H2O have opposite interference effect (minimumcompensation through calibration) Air humidity and air to fuel ratio information is required for model based correction of the interference effects of O2 and H2O (only required if highest accuracy is demanded) Response time T60 < 3 second, T90 < 5 sec Sensors are built on existing Delphi exhaust oxygen sensor technology NH3 Sensor provides an opportunity for improved SCR dosing control and system diagnosis

20