rainer vogt, ulf kirchner and matti maricq, · 1.4e+10 2.1e+10 pn / # s-1 and speed / km h-1 soot...
TRANSCRIPT
Rainer Vogt, Ulf Kirchner and Matti Maricq, Ford Forschungszentrum Aachen GmbH, Süsterfeldstr. 200, D-52072 Aachen, Germany Ford Motor Company, Dearborn, USA Particle number emissions are measured with two instruments according to the upcoming European emission regulations for light-duty diesel passenger vehicles and compared to data from other methods, including the current regulatory total particulate matter (PM) mass, photo-acoustic soot sensor (PASS) and engine exhaust particle sizer (EEPS). At the very low emission levels of diesel particulate filter (DPF) equipped vehicles, the solid particle number data correlate well with soot mass and with particle number measured by EEPS, if only those particles belonging to the accumulation mode are considered in the latter case. PN differences of >100% between tests of the same vehicle are observed. Comparison of the two PN instruments and the photoacoustic soot sensor show that these are systematic differences which originate primarily with the vehicle and not from instrument uncertainties. After accounting for this, a repeatability of <8% and a reproducibility of <27% are estimated for the particle number method. A large body of new vehicle data is presented from tests carried out immediately after production. The data show that nearly all vehicles would be below the anticipated Euro-5/Euro-6 particle number limit. The observed variability is significant and needs further investigation.
Research & Advanced Engineering
1
Investigation of EURO-5/6 Level
Particle Number Emissions of
European Diesel Light Duty
Vehicles
Rainer Vogt, Ulf Kirchner and Matti Maricq
Ford Motor Company
Aachen/Dearborn
14th ETH Conference, Zürich, August 1-4th, 2010
Research & Advanced Engineering
2
Introduction
Solid particle number (PN) counting required in upcoming European
emission regulation for light-duty diesel: evaporation of volatiles and
dilution of sample from CVS tunnel.
PSP: Particle
Sample
Probe
PTT: Particle
Transfer
Tube
PCF: Particle
Classifier
PND: Particle
Number
Diluter
Research & Advanced Engineering
3
Experimental Setup
• Diesel passenger car 1.6 L w DPF
• New European drive cycles (NEDC) after DPF regeneration /
conditioning.
• In addition to regulated components (CO, HC, NOx, total PM
mass): Two fully calibrated PN instruments in parallel plus
micro soot sensor (MSS). All sampling from full flow CVS.
• Use of calibrated PN instruments: average Particle
Concentration Reduction Factors (PCRF) for 30, 50 and 100
nm. Counter desensitized for small particles, 50% detection
efficiency for 23 nm particles according to upcoming European legislation.
Research & Advanced Engineering
4
PN compared to Soot Mass
All three parallel methods (2x PN and soot mass) show consistent trend.
0.00
0.02
0.04
0.06
0.08
0.10
A1 A2 A3 A4 A5 B1 B2 B3 B4
Test Number
PA
SS
So
ot
Mass [
mg
/km
]
0.0E+00
3.6E+10
7.2E+10
1.1E+11
1.4E+11
1.8E+11
Part
icle
Nu
mb
er
[km
-1]
Soot mg/km
FFA-PMP #/km
AVL489 #/km
first tests after
DPF regeneration
and conditioning
Research & Advanced Engineering
5
Regression Analysis of PN vs. Soot
Very good (R2>0.98) correlation of PN with soot; 1 mg soot corresponds to ~2 x 1012 particles.
y = 1.81E+12x
R2 = 9.82E-01
0.0E+00
2.0E+10
4.0E+10
6.0E+10
8.0E+10
1.0E+11
1.2E+11
1.4E+11
1.6E+11
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Soot Mass [mg/km]
Pa
rtic
le N
um
be
r [k
m-1
]
PN [ km^-1]
Linear (PN [ km^-1])
Research & Advanced Engineering
6
Emission Time Traces (1)
Strong decrease of soot and PN emissions after ~200 s, more
than engine out decrease => increasing filter efficiency.
High efficiency level, initial ~99.7% increase to >99.9%
0.0
0.1
0.2
0.3
0 200 400 600 800 1000 1200
time / s
mg
so
ot
/ m
3
0.0E+00
7.0E+09
1.4E+10
2.1E+10
PN
/ #
s-1
an
d s
peed
/ k
m h
-1
soot [mg/m^3]
PN [#/s]
NEDC speed [km/h]
0 km/h
70 km/h
140 km/h
Research & Advanced Engineering
7Emission Time Traces (2)
CO
2 [
%]
0
5
10
15
20
Sp
eed
[km
/h]
-120
-60
0
60
120
To
tal N
um
ber
[Part
./cm
³]
0
2
4
6
8
EEPS
Part
icle
Siz
e D
p [
nm
]
1
10
100
1000
dN
/dlo
g(D
p)
[#/c
m³]
0
Pa
rtic
le S
ize
Dp
[n
m]
CO
2
8.0x106
6.0x106
4.0x106
2.0x106
0.0
0
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
Time [s]
• Euro-4 DPF vehicle
• tailpipe sampling with FPS and EEPS
M. Bergmann et al., ETH Conference 2008; Atmospheric Environment 43, 1908-1916 (2009)
=> some PN measurable during cold start
=> very low PN during remainder
almost no nucleation mode particles
Research & Advanced Engineering
8
Solid vs. total PN
Total PN measured by TSI Engine Exhaust Particle Sizer (EEPS) ishigher than solid PN >23 nm as expected.
Best match for EEPS >40 nm: different measurement principle, no specific calibration.
0.0E+00
2.0E+11
4.0E+11
6.0E+11
8.0E+11
1.0E+12
1.2E+12
1.4E+12
1.6E+12
1.8E+12
T1 T2 T3 T4
Test Number
Part
icle
Nu
mb
er
[#/k
m]
0
2E+11
4E+11
6E+11
8E+11
1E+12
1.2E+12
1.4E+12
1.6E+12
1.8E+12FFA-PMP #/kmAVL489 #/kmEEPS #/kmEEPS >40 nm #/kmEEPS >23 nm #/km
Research & Advanced Engineering
9Particle mass vs. PN
• Non-DPF vehicles: very good correlation of PM mass with PN.
• DPF vehicles: correlation overwhelmed by artifacts of filter methods (~0.5 mg/km). Good correlation for soot, same slope as non-DPF => similar particles (size, density).
y = 1.81E+12x
R2 = 9.82E-01
y = 2.17E+12x
R2 = 9.72E-01
1.0E+09
1.0E+10
1.0E+11
1.0E+12
1.0E+13
1.0E+14
0.00 0.01 0.10 1.00 10.00 100.00
PM or Soot Mass [mg/km]
Part
icle
Nu
mb
er
[km
-1]
PN vs. SootPN vs. PM massDISI PN vs. PMLinear (PN vs. Soot)Linear (PN vs. PM mass)
soot & PM of the same 9 tests
1 mg =
~2 x 1012
particles
Non-DPF
vehicles
(13 tests)
DPF vehicle (9
tests, each with
PM mass and
soot measured
by a PASS)
Research & Advanced Engineering
10
Testing Variability
• Testing of new, conditioned vehicles at the production plant.
• Large variability of different vehicles of same type and model.
Even for the same vehicle >10% difference at 2nd test.
• Effect of vehicle or method? Error analysis required.
0.0E+00
1.0E+11
2.0E+11
3.0E+11
4.0E+11
5.0E+11
6.0E+11
7.0E+11
8.0E+11
1 1 2 3 4 5
Test Number
PN
[#/k
m]
0.0E+00
1.0E+11
2.0E+11
3.0E+11
4.0E+11
5.0E+11
6.0E+11
7.0E+11
8.0E+11
10 20 30 40 50
same vehicle type and modelvehicle tested twice
Research & Advanced Engineering
11
Error Estimation I: Repeatability
Error propagation of inputs in formula to calculate PN results:
Main contributions to repeatability identified:
• PNC flow. Daily measurements over several months: ±2%.
• Repetition of PCRF calibration, i.e. stability of PN dilution: ± 5%.
CVS volume(∆V/V) 1%
distance (∆d/d) 0.2%
PNC counting accuracy (∆k/k) 0%
PNC flow (∆C/Cs) 2%
PCRF uncertainty, daily (∆fr/fr) 5%
Volatile particles 0%
CVS tunnel background 0.02% - 0.2%
total repeatability ± 8%
Repeatability (one CVS / PN instrument)
d
fCkV rsN
610××××
=
Research & Advanced Engineering
12
Error Estimation II: Reproducibility
Main contributions to reproducibility:
• PNC flow. Daily measurements of 2 instruments: ±5%.
• PNC counting accuracy. Difference between 2 instruments,
which are within legal limits: ±10%.
• PCRF uncertainty. Comparison of 2 instruments with real diesel exhaust size distribution
- instrument “at the limit”, 70%, 80% and 100% penetration
- “low losses” instrument, 90%, 95% and 100% penetration for
particle sizes 30, 50 and 100 nm, respectively: ±10%.
CVS volume(∆V/V) 1%
distance (∆d/d) 0.2%
PNC counting accuracy (∆k/k) 10%
PNC flow (∆C/Cs) 5%
PCRF uncertainty (∆fr/fr) 10%
Volatile particles 1%
CVS tunnel background 0.02% - 0.2%
total reproducibility ± 27%
Reproducibility (several CVS facilities / PN instruments)
Research & Advanced Engineering
13
Summary and Conclusions I
• PN differences of >100% between tests of the same vehicle.
• Comparison of two PN instruments and correlation to a
photoacoustic soot sensor at extremely low emission levels
(<0.08 mg/km): differences are mostly variability of the vehicle.
• Estimations for PN method repeatability: ~8% and reproducibility
~27%. Good numbers for aerosol measurements.
• Following regeneration the DPF efficiency increases from 99.7%
to 99.97% relative to the same vehicle without a DPF.
Formation of a soot cake is a likely explanation, but needs
further investigation.
• Solid PN emissions of newly manufactured DPF Diesel vehicles
showed test to test variability with a CoV up to 110%. This
needs to be taken into account when emission standards are discussed.
Research & Advanced Engineering
14
Summary and Conclusions II
• Good correlation observed between PN emissions and soot mass, which corresponds to 2 x 1012 particles per mg.
• Correlation of solid PN to soot mass extends from DPF
equipped to non-DPF vehicles. This is consistent with soot as dominant component of modern, oxidation catalyst equipped,
light duty diesel vehicles and a uniformly high filtration efficiency
of DPFs versus particle size.
• The upcoming European legal limit of 6 x 1011 particles/km
corresponds to 0.3 mg/km soot, which is over an order of
magnitude more stringent than the PM mass limit of 4.5 mg/km.
• Comparison of solid PN emissions to the accumulation mode
particles measured with an EEPS shows that nucleation
particles are effectively removed by the regulation compliant
solid particle count instruments.
• Please also see Kirchner et al. SAE 2010-01-0789