Fundamental Processes Controlling Ash Accumulation in ...€¦ · Alexander Sappok V. Wong, C. Kamp, S. Munnis, I. Dimou, C. Chiou, Y. Wang, I Govani, M. Bahr, E. Cross, J. Kroll

1

Fundamental Processes Controlling Ash Accumulation in Diesel Particulate Filters and Impacts on DPF Performance

May 2, 2012

Alexander SappokV. Wong, C. Kamp, S. Munnis, I. Dimou, C. Chiou, Y. Wang, I Govani, M. Bahr, E. Cross, J. Kroll

Massachusetts Institute of TechnologySloan Automotive LaboratoryCambridge, MA

2012 DOE Crosscut Workshop

on Lean Emissions Reduction Simulation

MIT Diesel Engine, Lubricant, & Aftertreatment Conso rtium

2

DPF

Lube/ Fuel

Lubricant

Additives

Ash Function Of:• Incoming ash/soot particles

• Lubricant composition

• Exhaust conditions/Regeneration

• DPF design parameters

• DPF operating history

Engine-Out Emissions

Engine-Out Ash Emissions

Ash Accumulation Reduces DPF Life and Engine Effici ency

Efficiency & DPF Life

DPF Ash Accumulation

DPF ∆P Catalyst

Source: K. Aravelli

More Ash than Soot in DPF!

Image: Adapted from ORNL

3

Ash Accumulation and Deposit Formation Differs from PM!

~ 100 nm

τcr~ 1 µm

4

Program Approach & Consortium Activities

Existing Knowledge Base

Targeted Experiments

Address Specific Knowledge Gaps

Models and Theory

Build on Existing Framework

Ash Properties Analysis

Lab and Field

“Holistic approach considering lubricant chemistry, engine operation, and aftertreatment design through combination of fo cused experiments and theoretical models.”

Enhance fundamental understanding of key parameters controlling ash properties and impact on aftertreatment perform ance.

5

Experimental Facilities

∆P, T Sensors

DPF Core

FTIR

DPF Bench Reactors

Cummins ISB 300� Variable geometry turbocharger

� Cooled EGR

� Common rail fuel injection� Fully electronically controlled

� Gaseous and PM emissions measurement systems

Accelerated Ash Loading

6

Analytical Test Facilities

Center for Materials Science (CMSE)

• Materials Analysis•FTIR, Raman, XPS, Optical Microscopes

•Thermal Analysis• TGA, ICP-OES

•Electron Microscopy•SEM – EDX, TEM, FIB

•X-Ray Diffraction

Extensive and increasing use as part of DPF post-mortem analysis and ash characterization

Ash Exposed to Elevated Temperatures

XY

7

Key Parameters Controlling Ash Deposits and DPF Imp acts

Engine-Out Ash Emissions and Transport• Form of ash in exhaust/feed gas entering DPF; ash trapping efficiency

Ash Deposit Accumulation and Build-Up in the DPF

• Agglomerate formation and ash mobility/distribution in DPF

Ash Impact on DPF Pressure Drop Response

• Ash composition and properties relevant to DPF performance

Sensitivity of DPF Design Parameters to Ash Accumul ation

• Substrate materials, pore size & distribution, porosity

Role of Engine Control Strategies and Exhaust Condi tions

• Temperature, flow, and feed gas conditions affecting ash deposits

Regeneration Processes

• Real-time optical studies of ash formation and mobility

Ash – Catalyst Interactions

• Chemical and physical interactions of ash and catalyst/washcoat

8

Lubricant-Derived Ash Precursors Bound to PM

�Lubricant-derived ash transported to DPF bound to c arbonaceous PM

� Size of ash precursors of same order or smaller than PM agglomerates

� No lubricant-derived ash particles found separate from PM

� Cu peaks due to background from copper TEM grid

0

10

2030

40

50

60

7080

90

100

0.00.40.81.21.62.02.42.83.23.64.04.44.85.25.66.06.46.87.27.68.08.48.89.29.610.010.4

Energy [keV]

Cou

nts

A1A2

100 nmA1

A2Zn

Zn

S

P Cu

Cu

C

100 nm100 nm

SAE 2007-01-0318

9

Ca Fe Mg P S ZnTrapping Efficiency

99.87% 92.43% 98.01% 85.09% 64.89% 99.67%

Nearly All Metallic Ash Components Trapped in DPF

• Elemental emission rates determined from ICP analys is

• Post-DPF PM sampling 20 hours for sample size of 2- 3 mg

1682 rpm, 25% load (Steady-State)

Measured elemental trapping efficiency in DPF

10

Coating Evaporating

Core Evaporating

Particle Mass Spectrometer for Ash Measurements

Time-Resolved Mass-Spec

ASME ICEF2011-60100

11

Real-Time Measurement of Exhaust Ash Emissions (I)

40x103

30

20

10

0

Tota

l SM

PS

Mas

s (u

g/m

3)

5:30 PM10/11/10

6:00 PM 6:30 PM 7:00 PM 7:30 PM 8:00 PM

Date and Time

50

40

30

20

10

0

Hz

(ions

/s)

2

3

4

567

100

2

8

6

4

2

0

Oil Injection R

ate (ml/m

in)

inject_rate SMPS Mass Boron

diam

eter

(nm

)

ASME ICEF2011-60100

12

Real-Time Measurement of Exhaust Ash Emissions (II)

40x103

30

20

10

0

Tota

l SM

PS

Mas

s (u

g/m

3)

5:30 PM10/11/10

6:00 PM 6:30 PM 7:00 PM 7:30 PM 8:00 PM

Date and Time

14x103

12

10

8

6

4

2

0

Hz

(ions

/s)

2

3

4

567

100

2

sign

al, H

z

8

6

4

2

0

Oil Injection R

ate (ml/m

in)

inject_rate SMPS Mass Zn Boron S K Na P Ca

diam

eter

(nm

)

ASME ICEF2011-60100

13


Engine-Out Ash Emissions and Transport

• Form of ash in exhaust/feed gas entering DPF; ash trapping efficiency

Ash Deposit Accumulation and Build-Up in the DPF• Agglomerate formation and ash mobility/distribution in DPF











14

Ash First Accumulates Along DPF Channel Walls

0.0

0.3

0.5

0.8

1.0

1.3

1.5

0 25 50 75 100 125 150

Cha

nnel

Wid

th [m

m]

57 mm 133 mm

12.5 g/l Ash

(70K mile equiv.)

Center Center - 18 mm Center -36 mm

Center Center - 18 mm Center -36 mm

0.0

0.3

0.5

0.8

1.0

1.3

1.5

0 25 50 75 100 125 150

Axial Distance [mm]

57 mm 133 mm

42 g/l Ash

(240K mile equiv.) Plug

� Ash accumulation reduces DPF volume and affects pre ssure drop

ASME ICEF2009-14080

15

Benefit from

ash layer

formation

Initial Ash Deposition and Layer Formation

Ash layer build-up evident by changing ∆P profiles.

Pre

ssu

re D

rop

[K

pa

]10.7 g/l Ash

3.0 g/l Ash

0 g/l Ash 6.9 g/l Ash

Soot Load [g/l]

DPF Ash PM

Ash layer not fully established until 10 g/L or ~ 50,000 miles

Soot Load [g/l]

Pre

ssu

re D

rop

[K

pa

]

10.7 g/l Ash

3.0 g/l Ash

0 g/l Ash

6.9 g/l Ash

16

Application of Tracer Produces Stratified Ash Layer s

Ca Ash

Zn Ash

Mg Ash

DPF Substrate0

100

200

300

400

500

600

700

0 57 111Axial Distance [mm]

Ca Zn Mg

Ash

Lay

er T

hick

ness

[µm

]

0

200

400

600

800

1000

1200

1400

0 25 50 75 100 125 150

Axial Distance [mm]

Cha

nnel

Wid

th [µ

m]

Periphery

Centerline

Ash Thickness and Distribution

Ash Plug

ASME ICEF2011-60072

17

Voids in Ash Plug – Opportunities to Improve Packing

Si

O

Al

Start of plug and ash layer

Large voids throughout ash layer

Voids may explain low ash packing density…

Ca

Zn

Mg

ASME ICEF2011-60072

18






Ash Impact on DPF Pressure Drop Response• Ash composition and properties relevant to DPF performance









19

Additive Chemistry Impact on Ash Properties

Lubricant matrix all formulated to 1% sulfated ash, except base oil.

Composition of ash directly related to lubricant ad ditive chemistry.

Ca Mg Zn P S B Mo

ppm ppm ppm ppm ppm ppm ppm

Base <1 <1 <1 8 60 1 <1

Base + Ca 2,928 5 <1 2 609 3 <1

Base + Mg* <1 2,070 <1 <1 460 <1

Base + ZDDP <1 <1 2612 2,530 6,901 1 <1

Base, Ca+ZDDP* 2480 <1 1280 1,180 2,750 <1

Base, Mg+ZDDP* <1 1730 1280 1,180 2,840 <1

Commercial CJ-4 1,388 355 1,226 985 3,200* 586 77

Lubricant

Density Melting Point Description

g/cm 3 °C

CaSO4 Calcium Sulfate 2.96 1,460 Sinters/Decomposes ~1,250 °C

CaZn2(PO4)2 Calcium Zinc Phosphate 3.65

Zn2(P2O7) Zinc Pyrophosphate 3.75 Sintering begins ~ 800 °C

Zn3(PO4)2 Zinc Phosphate 4.00 900 Sintering begins ~ 800 °C

Zn2Mg(PO4)2 Zinc Magnesium Phosphate 3.60

MgO Magnesium Oxide 3.58 2,832

MgSO4 Magnesium Sulfate 2.66 1,124 Decompostion 900-1,100 °C

Major Ash Components

20

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 5 10 15 20 25 30 35 40 45

Total Ash Load [g/L]

Pre

ssur

e D

rop

[kP

a] .

Ca Ash Shows 2X Increase in ∆P Over Zn & Mg Ash

� Lubricant additive chemistry affects ash properties and pressure drop

� Ca-based ash shows much larger effect on pressure drop than Zn ash

Flow Bench @ 25 °C, Space Velocity: 20,000 hr -1

Equivalent Miles*

Equivalent Hours* 2,840 4,260 5,670

110 K 160 K 220 K

CJ-4Ca

Zn

* Assumes: 15 g/hr avg. oil consumption, avg. speed of 40 mph, and full size DPF of 12 L volume

Mg

ρPack = 0.25 g/cm 3

ε = 91%


ε = 91%


ε = 95%ρPack = 0.19 g/cm 3

ε = 95%

Ash: Zn

Ash: Ca

ASME ICES2012-81237

21

0

3

6

9

0 1 2 3 4 5 6 7 8Cummumlative PM Load [g/l]

0.0

0.3

0.6

0.9

1.2

1.5

0 25 50 75 100 125 1500.0

0.3

0.6

0.9

1.2

1.5

0 25 50 75 100 125 150

Ash Chemistry Impacts Ash Properties and DPF ∆P

Cha

nnel

Wid

th [m

m]

Axial Distance [mm] Axial Distance [mm]

Pre

ssur

e D

rop

[kP

a]

CJ-4 33 g/LCa 29 g/L

Ca + ZDDP 25 g/L

Mg 28 g/L

No Ash

Mg + ZDDP 23 g/L

ZDDP 28 g/L

Cha

nnel

Wid

th [m

m]

Cordierite 200/12

Ash Distribution Profiles

Ca - Ash

Differences in DPF ∆P due to ash

properties ( ρ, ε, Dp)

ASME ICES2012-81237

22

Detailed Understanding of Ash Properties Required

wvK

P wP

SootAshWall ⋅⋅

=∆ µ

//

( )PDf ,ε=K

lTheoretica

Packing

ρρ

ε −= 1

Ash Propertieso Provide critical information to explain

fundamental differences in ∆P

o Complex mixture of metal oxides, sulfates, phosphates

o Characterization of particle morphology, physical, chemical properties challenge

New Techniques and Diagnostics

o C. Kamp Presentation (12/2010)

23

CJ-4 Ash Composition and Porosity

Zn2Mg(PO4)2

CaSO4

CaSO4

Zn2Mg(PO4)2

Zn2Mg(PO4)2

AshWall

DensityTheoretical

DensityAsh

Porosity

[g/l] [g/cm3] [g/cm3] [%]

42.0 0.30 3.4 91.1

12.5 0.18 3.4 94.6

lTheoretica

Packing

ρρ

ε −= 1

SAE 2010-01-1213

24

Ash – PM Layer Interface Clearly Defined

Unlike PM depth filtration in DPF surface pores, ve ry little soot penetrates into ash layer.

Focused Ion Beam (FIB) Milling Coupled with SEM

SAE 2012-01-0836

25

Individual Ash Particles Also Porous

Particle most likely formed from sintering/agglomer ation of ash precursors.

Focused Ion Beam (FIB) Milling Coupled with SEM

SAE 2012-01-0836

26

Ash Particles and Agglomerates Porous Shells!

• Ash agglomerates consist of porous particles

• Consistent with low packing density and high porosity measurements

Images: C. Kamp

SAE 2012-01-0836

27

Ash Accumulation Also Influences Soot Properties

0

2

4

6

8

10

0 1 2 3 4 5 6 7 8 9 10 11

Cummumlative PM Load [g/l]

Pre

ssur

e D

rop

[kP

a]

No Ash 42 g/l Ash 42 g/L Ash (Adjusted)

Ash deposits displace soot in DPF – higher local soo t loads

Adjust specific soot load to account for actual available trap volume

RPS iCleaniAsh PM

P

PM

P

,,

∂∆∂÷

∂∆∂=

SAE 2010-01-0811

28

Variation in PM Layer Properties with DPF Flow Well -Known

Kso

otx1

0-14

[m2 ]

Por

osity

D

dUPe imaryw Pr⋅

=

Diffusion

InertiaPe =

Source: SAE 2002-01-1015

Source: SAE 2002-01-1015

Image: Menelaos, T., Ph.D. Thesis, Yale, 1991.

Image: Menelaos, T., Ph.D. Thesis, Yale, 1991.

SAE 2002-01-1015SAE 2002-01-1015SAE 2002-01-1015

29

-20%

-15%

-10%

-5%

0%

5%

10%

15%

20%

0 10 20 30 40 50

Ash Load [g/L]

Rel

ativ

e C

hnag

e [%

] 50K GHSV

75K GHSV Soot deposited at flows > 75K 1/hr insensitive to ash loading

Soot Properties

• Estimated from empirical Pe number correlations for constant flow rate

• Low space velocity conditions most strongly affected by ash loading

-100%

-50%

0%

50%

100%

150%

200%

0 10 20 30 40 50

Ash Load [g/L]

Rel

ativ

e C

hnag

e [%

]

Variation of Soot Properties Due to Ash Deposits

Packing Density Permeability

25K GHSV

75K GHSV

50K GHSV

SAE 2010-01-0811

30








Sensitivity of DPF Design Parameters to Ash • Substrate materials, pore size & distribution, porosity







31

Matrix PorosityMean Pore

Size

1st TrialLowHigh

Low

2nd Trial HighLowHigh

3rd Trial Moderate Low

Multi-Cartridge Filter Holder12 DPF

Segments

Sensitivity of DPF Design Parameters to Ash

Ash Loading

Additional DPF Parameters

• Filter/substrate materials

• DPF coatings and catalysts

• Filter geometry and cell configuration

32

Average Ash Deposition after 60 hrs of operation: 20 g/L

Inlet Temperature (before the cone)

610 °C (±25)

7 g/L ash

0 g/L ash

14 g/L ash

20 g/L ash

(Space Velocity range: 6,500 – 55,000 [1/hr])

DPFs Experience Even Ash Loading and Temperatures

• Ash evenly distributed in core samples

• Little ∆P variation between duplicate samples

Pressure Drop VariationAsh [g/L]

9.1 %

4.3 %

0.0 %

4.8 %

9.1 %

2.2 %1

2

3

4

5

6

33

SiC (50%,15*µm) with 0 g/L ash

Cd (50%,15*µm)

with 0 g/L ash


SiC (40%,15*µm)

with 0 g/L ash


Similar ∆P Response to PM Accumulation for All DPFs

DPF Pressure Drop Response 0 g/L Ash

* Indicates pore and porosity size class

34

(50%)

(60%)

(40%) 23 %

15 %3 %

Sensitivity of DPF Porosity to Ash Accumulation Varie s

Low Ash Load (5 g/L)

• Sensitivity of ∆P to ash accumulation increases with decreasing DPF porosity at low filter ash levels

• At high ash loads, ash dominates ∆P, which is insensitive to initial DPF porosity of filter, over range tested

(50%)

(60%)

(40%)

Pre

ssu

re D

rop

[k

Pa

]

Moderate Ash Load (20 g/L)

High porosity DPFs reduce ash ∆P sensitivity

0 g/L

5 g/L

∆P Insensitive to DPF Porosity at Elevated Ash Levels.

35










Engine Control Strategies and Exhaust Conditions• Temperature, flow, and feed gas conditions affecting ash deposits





36

Exhaust Conditions Are Continually Changing

SAE 2009-01-1262Corning Deer 2006

DPF Temperature Distribution

(300-420K Miles, HD Diesel) Typical Highway Drive Cycle

DPF Inlet Temp

Air Flow

Speed

• Potential for short excursions above 700 °C over DPF operating history

• Exhaust flow rates also vary considerably, even over highway drive cycle

37

-30

-25

-20

-15

-10

-5

0

5

0 200 400 600 800 1000 1200 1400

Temperature [C]

Cha

nge

in L

engt

h [%

]

Exhaust Temperature Significantly Affects Ash Volum e

Field CJ-4: *705 C

Lab CJ-4: *940 C

�Large decrease in ash volume for temperatures over 700 °C� Reduction in ash weight over temperature ranges less than 10%

� Typical ash porosities 85% - 95% means large potential to reduce volume

Lost contact with probes

Lab Ca: *1,260 CLab Zn: *800 C

Change in Length = dL/L o

*Sintering Onset

Competing Effects on ∆P Based on Ash Distribution

SAE 2012-01-1093

38

-2.2%

1.7%

-5.3%

2.4%

-2.4%

-6.0%

-4.0%

-2.0%

0.0%

2.0%

4.0%

6.0%Ca (28 g/L, Per) Zn (28 g/, Per)CJ-4 (13 g/L, Per) CJ-4 (33 g/L, Con)CJ-4 (42 g/L, Per)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

600 700 800 900 1000 1100

Elevated Temperatures Exert Large Effect on Ash Pac king

Control (No Ash)

13 g/L CJ-4 Ash

42 g/L CJ-4 Ash

28 g/LCa Ash28 g/L ZDDP Ash

Nor

mal

ized

∆P

Cha

nge

in ∆

P

High Flow Exposure to 200,000 hr -1 High Temperature Exposure

0 min 5 min

CJ-

4 La

b A

sh

Ash Volume Reduction Fast

DPF core heated to 880 °C in 5 min, then quenched.

Lab Ca Field Lab CJ4 Field

650 °C

800 °C

1,000 °C

SAE 2012-01-1093

39

Large Reduction in Ash Volume at Elevated Temperatu res

650 °C 700 °C 800 °C 900 °C 1,000 °C 1,100 °C

ZDDP

Calcium

CJ-4 (P)

CJ-4 (C)

28 g

/L28

g/L

42 g

/L33

g/L

SAE 2012-01-1093

40

High Temperatures Cause Ash Layer Cracking/Shrinkin g

42 g/L CJ-4 Ash (Periodic) 33 g/L CJ-4 Ash (Continuous)

650 °C

700 °C

800 °C

900 °C

1,000 °C

1,100 °C

Despite large volume reduction, ash weigh change < 7%

SAE 2012-01-1093

41

Similar Behavior in Lab/Field Ash May Be Due to Che mistry

650 °C 700 °C 800 °C 900 °C 1,000 °C 1,100 °C

Calcium

CJ-4 (P)

Lab

Lab

Fie

ld A

Fie

ld D

SAE 2012-01-1093

42

Chemical and Physical Changes in Ash at High Temps.

Ca

Zn

Tem

pera

ture

Equilibrium Crystal Structures formed at High Temperatures

)min( SSdA∫γ

Ash Composition Changes Irreversibly

SAE 2012-01-1093

43

Applications to Understand Field DPF History

Wetting on SubstrateAsh Necking & Agglomeration

Prior High Temperature

Exposure

“Normal” Field Ash

Field Ash Exposed to Thermal Event

SAE 2012-01-1093

44

Conceptual Description of Temperature Effects on As h

Possible Effect of Temperature on Ash Deposits

Competing Processes Require Detailed Understanding

• Elevated temperatures result in significant ash volume reduction

• Location of ash deposits (channel vs. wall) plays a large role in impact on ∆P

• Deterioration of ash cake layer could result in increased ∆ P with soot

Ash Deposits in Typical DPF

Ash Deposits After High Temp.

DPF Pores Exposed!

45












Regeneration Processes• Real-time optical studies of ash formation and mobility



46

Optical Access System for DPF Regeneration Studies

� Understand Influence of Regeneration on Ash Propertie s� Active/Passive strategies may impact ash agglomerat ion and mobility

� Role of soot interactions with ash during regenerat ion important

DPF Substrate

Regeneration Parameters• Thickness of PM Layer

• Role of NO2 from DOC vs. CDPF

• Temperature and flow conditions

• Catalysts interactions

Air

2nd Generation

47

Video: PM Oxidation on Clean DPF Surface (Heavy PM )

48

Video: PM Oxidation with Ash (DPF Cross Section)

49

Quartz Glass

Gasket

27mm by 1.44mm Slit in Matting and One Channel of the DPF

Current Optical Setup for Flow Reactor Testing

Channel View

50

Video: PM Oxidation on Clean DPF Surface (Heavy PM )

51

Ash Deposition on Top of PM Layer: Regeneration

Coat surface of soot cake with thin layer of ash

• Allows for visualization of ash/PM mobility during regeneration

• Enables visualization of ash agglomerate formation

Substrate

PM (Genset)

Ash

• Thin ash layer not fully-established

• Ash applied via accelerated loading system

• Thick, fully-established PM layer

• PM loading accomplished using genset

52

Ash Deposition on Top of PM Layer: Regeneration

Regeneration with thin ash layer covering PM surfac e

1 2

3 4

53

Video: Ash Deposited on Top of PM Layer

Substrate

54

Ash Agglomeration Process During Soot Oxidation

~ 100 nm~ 1 µm

Ash residence time in DPF is long ~ 100,000 + Miles

Internal void shows walls composed of ~nm scale particles

55














Ash – Catalyst Interactions• Chemical and physical interactions of ash and catalyst/washcoat

56

Summary and Conclusions

I. Ash Build-Up: Ash loading of ~ 10 g/L or around 50,000 miles required to form fully-established ash layer.

II. Ash Morphology: Two porosity scales identified in ash layer and ash primary particles, which are themselves hollow.

III. Lube Chemistry: Ash properties and DPF pressure drop strong function of additive composition.

IV. Exhaust Conditions: Transient changes in temperature induce much larger variations in ash packing than high flow rates.

V. DPF Parameters: DPF pressure drop relatively insensitive to original substrate porosity following ash layer build-up.

VI. Regeneration Effects: Preliminary optical studies highlight importance of regeneration parameters but requires further study.

Detailed understanding of all system parameters imp ortant to reduce impact of ash on DPF degradation and fuel ef ficiency.

57

Acknowledgements

� Research supported by: MIT Consortium to Optimize L ubricant and Diesel Engines for Robust Emission Aftertreatme nt Systems

� We thank the following organizations for their supp ort:

- Caterpillar - Chevron/Oronite - Cummins

- Detroit Diesel - Infineum - Komatsu

- NGK - Oak Ridge National Lab - Süd-Chemie

- Valvoline - US Department of Energy

- Ciba - Ford - Lutek

� MIT Center for Materials Science and Engineering

Documents

Fundamental Processes Controlling Ash Accumulation in ...€¦ · Alexander Sappok V. Wong, C. Kamp, S. Munnis, I. Dimou, C. Chiou, Y. Wang, I Govani, M. Bahr, E. Cross, J. Kroll