Notes on „Soot“ Measurement of Diesel Engines · Soot measurement of Diesel engines 2004.-08-17 | Page 6 “Filter Smoke Number (FSN)” Measurement Advantages: – Simple method,

Notes on „Soot“Measurement of Diesel Engines

Wolfgang SchindlerWolfgang Singer

AVL List GmbH, Graz, Austria

Soot measurement of Diesel engines 2004.-08-17 | Page 2

Why bother about Soot Measurement ?

The measurement of “Soot” is a widely accepted method in Engine R&D since several decades. It provides information about the quality of the combustion processPMP has decided that agglomeration mode particles should be counted - Such particles typically consist of “soot” in Diesel exhaustThere is some evidence that soot is more relevant with regard to health concerns than the homogeneous, volatile “Nanoparticles”

HC Fuelcond.

Soot “ C “

Metal & Others

WaterHC Oilcond.

Gravimetric measurement oftotal particulate mass

after partial-flow or full-flow dilution


Properties of Soot from Diesel Engines

Soot is a combustion product. (“Carbon black” is no “soot”). Soot consists of “Carbonaceus Particles”Soot has a combustion temperature in air of >450 °C,it is not volatile or liquid.(compare the measuring proposal of PMP)The size is mainly in the “agglomeration mode” range, the structure irregular (fractal dimension 1.8 to 2.5) Soot is “really black”


Methods for Diesel Soot Measurement

Analysis of filtered Diesel Particulates

– chem. extraction– thermogravimetry– Coulometry

PMP method:Thermodesorption and particle counting

Via its optical properties:– The mass extinction coefficient

of Soot is nearly constant for D < 100 nm

– The scattering coefficient goes with D/λ - 4

Mass extinction with Lamda= 560 nm

1.E-02

1.E-01

1.E+00

1.E+01

10 100 1000Diameter (nm)

Mas

s ex

t. co

eff R

(m2/

g)

ScatterExtinktion

typical mean size of Diesel particulate


How well can various systems measure Soot Mass? C

arbo

nC

arbo

n „C

“ „

C“

„HC“„HC“

SOSO44--

Ash

F ilte

r Gra

vim

e try

F SN

Op a

city

PhotoacousticSensor

Opaci-meter

Smoke Meter

Criteria:accuracy - selectivity -measuring artefactsconcentration:

– resolution, detection limit– dynamic range

repeatability, reproducibilitytime resolution

Target: to assess the physical limits of various methods and the practical problems encountered when measuring Diesel Soot.


“Filter Smoke Number (FSN)” Measurement

Advantages:– Simple method, proven reliability,

widely used in R&D– Selectivity: for soot content >15%, FSN is

clearly related to Soot concentration– Detection limit: with high sampling volume

20 µg/m3 or 0,002 FSN achievable⇒ dynamic range > 1: 10000 (concentration)

– Repeatability: 5% achievable (see below)

Challenges:– Due to the simplicity of the method it is

often used uncritically without observing the basic principles of particle measurement.

– A realistic theory of the relation between FSN and Soot concentration does not exist

Drawback:time resolution >10 sec by principle

currentenginesfutüre

FSN

0 1 2 3 4 5


FSN measurement: Challenges

Influences to FSN Measurement:– Sample Conditioning. FSN is measured from the raw exhaust ⇒

concentrations are high, particle deposition is of considerable concern.– Hangup after measurement of large FSN: purging is required to avoid

re-entrainment during “white level determination” ( <2 µg allowed !)– Filter paper: if two different filter papers, at opposite ends of the ISO

10054 specification band are used, results may vary by >20%.(AVL makes a special paper check to guarantee ± 5%).

– Soot content: in aerosols with a very small soot core (<15% by mass) “the absorption coefficient is increased”

FSN reproducibility ESC, Euro 5 HD engine

0.00

0.02

0.04

0.06

0.08

0.10

0.12

1 2 3 4 5 6 7 8 9 10 11 12 13 mode

FSN

ESC 251

ESC 250

ESC 234

ESC 235

If measurements are carried out carefully, even the emissions of a Euro 5 HD engine can be measured reproducibly.


Opacity Measurement

Advantages:– Good time resolution, ≤ 0.1 sec– Theoretically well defined ( for fixed R)– Detection limit: k = 0.002 m-1

corresponding to ≈ 300 µg/m3

(sufficient for non-trap engines)Challenges:

– To achieve a stable “Zero” value I0– The mass extinction coefficient R of

soot is slightly size dependent (typical uncertainty ± 20%)

– The influence of NO2 on the signal is substantial,the influence of HC and SO4 only at high concentrations

Drawback:– At the current stage the technology is

at its limits


Challenges for good Opacity Sensitivity

I/ I0= e -k . L = ( 1 - N/100 )N [%] Opacityk [m-1] Absorption

Good resolution in k is achieved if:• I0 is stable• The measuring length L is large

To some extent the two requirements contradict each other:

hot

hot

cold

If the temperature of gas and chamberwalls are not stable, the thermal lens(„mirage effect“) changes I0Temperature changes of 1.5 °C resultin an I0 change of 0.1%

U(Det) Änderung mit Gas-ein T

y = 0.0547x + 0.5119

-0.5

2.0

4.5

-20 30 80T(Zelle) - T(Gas-ein)

% U

(Det

)


“Accuracy” - is Opacity related to Soot?

Practical experience:A constant (± 20%) relation betweenOpacity and Soot is given if:• NO2 concentration is < 100ppm• The particulates consist to less than

50% of HC and SO4

• The exhaust temperature in the measuring chamber is ≥ 100°C

• The particle size is in the “typical”Diesel exhaust accumulation mode range

In other words: isk (m-1) = R • Conz (mg/m3)

with konstant R?

FSN_Opazitätskorrelation

y = 0.0043x + 0.0403

0.00.10.20.30.40.50.60.70.80.9

0 50 100 150 200mg/m3 aus FSN

k-W

ert

BMW Werte

Prüfstand 13

Prüfstand 4

MAN Werte

Linear (BMW Werte)

Quantities influencing R:• Particulate size.• Particulate composition: (HC and

SO4)

The Particulate size dependence of R decreases with increasing λ-> IR extinction (drawback: R decrease with increasing wavelength)

Comparison of Partial and Full-flow Opacimeters to Smoke values FSN < 0.1

-0,1

0,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0 0,02 0,04 0,06 0,08 0,1 0,12FSN

N (%

)

Partial Flow 430 mmFull-flow, 6 inch


How about light Scattering?

I0

Is IS = Fkt( ϑ, ϕ, d/λ, n, ff)with n = m+ik,

ff = form faktor

IS ∝ d-6

In Principle: The measurement of Is at different ϑ, ϕ and at different λ canyield a wealth of information about size, size distribution, number, shape.....

In Practice: • Single particle scattering: with reasonable effort the lowersize limit is ≈ 100 nm due to low scattering intensity.

• Multi particle scattering: large particles dominate the Signal.• Multi-wavelength and -angle scattering: the addition of

measuring uncertainties introduces ambiguities into thedata evaluation.

⇒ A robust and accepted light scattering instrument has not made its way to the market despite substantial and repeated effort in the last 20 years


Photoacoustics - AVL Micro Soot Sensor

Periodic heating by a modulated laser beam and subsequent cooling generates an acoustic signal that is detected by a microphone

Advantages:– Sensor is “directly” sensitive to soot

concentration– high sensitivity

(detection limit < 10 µg/m³)

Modulatedlaser beam

Soot particle

Modulatedheating

Modulatedexpansion

Acoustic waveDetector

microphone

Challanges:– signal evaluation at low

concentrations– repeatability / reproducibility /

calibration– currently for diluted exhaust only


Accuracy: Selectivity, Linearity and Comparability

Selectivity:No cross-sensitivity from typical or untypical exhaustgases, if the proper near IR Laser wavelength is used (H2O, HC, SO4, NO2....).

Due to the physical principle thesignal is linear with sootconcentration

Linearity:

Comparability:According to experience in various laboratories the signal iscomparable to the sootdetermined with gravimetric sootmass determination

Comparability of Photoacoustics with

Gravimetry

y = 0.9773x - 6E-05R2 = 0.9914

0.00

0.01

0.02

0.03

0.00 0.01 0.02 0.03PM soot mass / g km-1

AVL4

83 /

g km

-1

Courtesey V. Scheer, FFAResearch & Advanced Engineering


Reproducibility, Repeatability, Calibration

Reproducibility:The near IR frequency of the laser used guarantees, that D/λis small for typical Diesel particulates. Therefore the mass extiction coefficient R is nearly constant

The experiments show that twophotoacoustic systems of different build and different calibrationmethods show equal results

Repeatability:

Calibration:Calibration must be performed in comparison to gravimetric methods. However, the calibration factor in the„PASS“ has not been changed in thelast 2 years, and yields the sameresults as the AVL Micro Soot Sensor

Courtesey V. Scheer, FFA

Comparability of two Photoacoustic instruments

(IWC Munich and AVL)

y = 0.9912x + 0.0004R2 = 0.9963

0.00

0.01

0.02

0.03

0.00 0.01 0.02 0.03PASS / g km-1

AVL

483

/ g k

m-1

Research & Advanced Engineering


Resolution (concentration, time)

ETC 3 (without PM_KAT) and 5,6 (with PM_KAT)

0

500

1000

1500

2000

1200 1400 1600 1800time(s)

µg/

m3 ETC 3

ETC 5ETC 6

Resolution:Challenges: Good resolution (below10 µg/m3) can be obtained by:– Signal evaluation with high

frequency selectivity– Low noise in the gas path– Correct signal evaluation,

subtracting the baseline signal fromsooted windows phase-corrected

Photoacoustic Signal Evaluation

01

234

567

89

0.0 0.2 0.4 0.6 0.8mg/m3

Sign

al (m

V)

naivphase sensitive

filtered air, but dirty windows

Dynamic range:Photoacoustics allows a dynamicrange of 1: 10000 (important to detect peaks on low background

Time resolution:1 sec τ90 is easily achievable, bettertime resolution currently sacrificesconcentration detectivity


Laser Induced Incandescence (LII)

ExhaustExhaustpipepipe

ExhaustExhaustadapter ringadapter ring

Laser

PMTPMTDetectorDetector

Gas flowGas flow

High sensitivity for ultra-low emissiongood time resolution for transient testingIn-situ raw exhaust gas measurement

However:– substantial costs, due to expensive

laser– serious laser hazards


Summary

Several Methods exist to measure Soot on the basis of its strong absorption coefficient.If these methods are used with attention and some basic knowledge about particulate measurement they can give

– reliable soot data– even for very low soot

concentration in the exhaust The method proposed by PMP for measuring particles seems feasible -the interlaboratory exercise has not yet started - but practically it measures only the soot fraction of the Diesel exhaustA second thought on mass-proportional soot measurement methods as a viable alternative seems appropriate.

Mass extinction with Lamda= 560 nm

1.E-02

1.E-01

1.E+00

1.E+01

10 100 1000Diameter (nm)

Mas

s ex

t. co

eff R

(m2/

g)

ScatterExtinktion

typical mean size of Diesel particulate

Documents

Notes on „Soot“ Measurement of Diesel Engines · Soot measurement of Diesel engines 2004.-08-17 | Page 6 “Filter Smoke Number (FSN)” Measurement Advantages: – Simple method,