PMP HD_ILCE Informal Meeting at DfT
London 30th March 2009
Preliminary results from VE testing of Iveco Cursor 8 at Ricardo
www.ricardo.com
RD.08/######.#
© Ricardo plc 2009
Equipment – Instruments and Configuration
M E+3022A
DM S500
M E+3010D
SPCS-19
1m, ME + 3010D
CVS
MDLT
Flow rate
~56m3/min
60litres/min
Split ratio
1
1100
Volume
(1800s)
1700m3
PM flow
60litres/min
0.2m
0.2m
90 lpm
0.8m
3m, SPCS-19
5m, DMS500
~4.5m
60 m3/min
Primary dilution tunnel
1800litres
temp
dilution tunnel
Flow direction
60litres/min
Dilution air
LEPA
HEPA
Secondary
dilution
tunnel
act. carbon
Scaling factor
1
~945
SPCS -20
HEPA
Heater (modulated)
MDLT
4m
PM
TX40 47 mm
60 lpm
Identifier
Instrument
Particles
Size range
Measurement(s)
SPCS-19
GPMS-CVS
solid
>23nm
total PN
SPCS-20
GPMS-MDLT
solid
>23nm
total PN
ME_3010D
ALT-CVS
solid
>23nm
total PN
ME_3022A
additional
solid
>7nm
total PN
DMS500
additional
5nm -1000nm
size dist,
total PN
CRT
47±5°C
TX40 47mm
60 lpm
ENGINE
Solid &
volatile
PM
RD.08/######.#
© Ricardo plc 2009
2
Gaseous Emissions
Similar results from raw and dilute measurements except for HCs and to a lesser
extent CO – indicates HC background in CVS
Cursor 8 Gaseous Emissions - Cold WHTC
1 Stdev Error Bars
Cursor 8 Gaseous Emissions - Hot WHTC
1 Stdev Error Bars
70.00
30.00
Raw
Dilute
Raw
60.00
20.00
g/kWh
g/kWh
Dilute
25.00
50.00
40.00
30.00
15.00
10.00
20.00
5.00
10.00
0.00
0.00
CO2/100
CO*10
NOx
HC*100
Cursor 8 Gaseous Emissions - ETC
1 Stdev Error Bars
CO2/100
CO*10
Cursor 8 Gaseous Emissions - WHSC
1 Stdev Error Bars
14.00
Dilute
Raw
Dilute
12.00
10.00
10.00
10.00
8.00
8.00
8.00
g/kWh
12.00
g/kWh
12.00
6.00
4.00
4.00
2.00
2.00
2.00
0.00
0.00
CO2/100
CO*10
NOx
HC*100
Dilute
6.00
4.00
0.00
HC*100
14.00
Raw
6.00
NOx
Cursor 8 Gaseous Emissions - ESC
1 Stdev Error Bars
14.00
Raw
g/kWh

CO2/100
CO*10
NOx
HC*100
CO2/100
CO*10
RD.08/######.#
NOx
HC*100
© Ricardo plc 2009
3
Instrument Validations
All PNCs easily met zero check criterion
3010D PNC ZERO Checks (ME and SPCS Systems)
0.25
0.2
#/cm3 @20°C

0.15
0.1
0.05
0
1
2
3
4
5
6
7
8
9
10
Day
Measured 3010D (ME)
Measured 3010D (SPCS-20)
Measured 3010D (SPCS-19)
Limit
RD.08/######.#
© Ricardo plc 2009
4
Instrument Validations
SPCS systems easily met system zero criterion, but Matter system failed after the first few
days
– Disc wear led to high background particle numbers, so later data discarded
– Matter working on new disc coatings for greater robustness; preliminary results encouraging
#/cm3 @20°C
a
System ZERO Checks (ME and SPCS System s)
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1
2
3
4
5
6
7
8
9
10
Day
Measured 3010D (ME)
Measured 3010D (SPCS-19)
b
Measured 3010D (SPCS-20)
Limit
System ZERO Checks (ME and SPCS System s)
#/cm3 @20°C

20
18
16
14
12
10
8
6
4
2
0
1
2
Measured 3010D (ME)
3
4
5
Day
Measured 3010D (SPCS-19)
6
7
8
Measured 3010D (SPCS-20)
9
10
Limit
RD.08/######.#
© Ricardo plc 2009
5
System Backgrounds

CVS solid particle number backgrounds much higher than partial flow
Background Levels (ME and SPCS System s) CVS and Partial Flow
Solid particle background
#/cm 3 @20°C
50
40
30
20
10
0
1
2
3
4
5
6
7
8
9
10
Day
3010D (ME)

3010D (SPCS-19)
3010D (SPCS-20)
3022A (ME)
DMS shows variable accumulation mode to
be present
– Solid particle background probably
historic soot and low volatility HCs
• Mass background also expected
– Volatile background present as
nucleation mode and also associated
with accumulation mode
RD.08/######.#
© Ricardo plc 2009
6
PM Measurements

General trend that CVS sampled filter masses are higher than partial flow masses
Filter Masses
0.400
CVS
Mass on Filter (mg)
0.350
P° flow
0.300
0.250
0.200
0.150
0.100
0.050
Min
Max
Min
Max
Min
Min
Hot WHTC Hot WHTC
Max
Cold WHTC Cold WHTC
Max
Min
Max
0.000
WHSC
WHSC
ETC
ETC
ESC
ESC
Cycle

The LOD can be determined as 3 x standard deviation of the blank measurement.
– Partial flow mass method LOD:
– Full flow mass method LOD:

3s = 35µg (equates to 1.1mg/kWh ETC)
3s = 108µg (equates to 8.3mg/kWh ETC)
Partial flow system 7 times more sensitive than full flow for mass measurements
RD.08/######.#
© Ricardo plc 2009
7
PN Measurements

PN emissions trends similar between CVS systems, different to partial flow
Brake Specific Mean PN (#/kWh)
Particle Number Emissions - All Cycles - 1 Stdev Error Bars
7.00E+11
6.00E+11
WHTC_cold
WHTC_hot
WHSC
ETC
ESC
5.00E+11
4.00E+11
3.00E+11
2.00E+11
1.00E+11
0.00E+00
PMP ME_3010D [CVS] - BSPN
PMP SPCS 19 [CVS] - BSPN
PMP SPCS 20 [MDLT] - BSPN
Test Cycle Name
PN from CVS and Partial flow - Cold WHTC test 20081218
Both systems indicate
major transient events
– C_WHTC
1000000
1000
SPCS20 [MDLT] - From aligned data file
SPCS19 [CVS] - From aligned data file
PMP 3010D [CVS] - From aligned data file
100000
100
10000
1000
10
100
Conc. sensor p/ccm - PMP 3010D

CVS PN background
much higher than partial
flow
Conc. sensor p/ccm (SPCS-19 and SPCS-20)

1
10
1
0
200
400
600
800
1000
1200
1400
1600
0.1
1800
Time (s)
RD.08/######.#
© Ricardo plc 2009
8
PN Limit of Detection

The LOD can be determined as 3 x standard deviation of the blank
measurement.

Partial flow + SPCS-20
– LOD: 3s = 15 particles/cm3 in the tunnel (after dilution correction)
– equates to ~8.7x108 particles/kWh over the ETC

Full flow + SPCS-19
– LOD:
3s = 3055 particles/cm3 in the CVS (after dilution correction)
– equates to ~1.4x1011 particles/kWh over the ETC

Based upon these results, the limit of detection for the partial flow system is
160 times lower than for the full-flow system
RD.08/######.#
© Ricardo plc 2009
9
Full flow Vs Partial Flow – Correlation between Number Data

Results from cycles with highest emissions levels similar between full and partial flow

Lowest emissions cycles results only discriminated in partial flow system
Comparison Between CVS (SPCS-19) and P.Flow (SPCS-20) Particle Emissions
1.0E+13
#/kWh [SPCS-20]
1.0E+12
y = 0.4893x - 8E+10
R2 = 0.4292
CVS Background
1.0E+11
C-WHTC
H-WHTC
1.0E+10
WHSC
ETC
ESC
Unity
1.0E+09
1.0E+09
SPCS20
1.0E+10
1.0E+11
1.0E+12
#/kWh [SPCS-19]
RD.08/######.#
Linear1.0E+13
(SPCS20)
© Ricardo plc 2009
10
Full flow - Correlation between Different Measurement
Systems

Excellent (2%) agreement between Matter system and SPCS-19, both from CVS

Dilution factors used for comparison
Comparison Between CVS (ME_3010D) and CVS (SPCS-19) Particle Emissions
1.0E+13
y = 1.0195x
R2 = 0.8295
#/kWh [SPCS-19]
1.0E+12
C-WHTC
1.0E+11
H-WHTC
WHSC
ETC
ESC
Unity
1.0E+10
1.0E+10
SPCS19
1.0E+11
1.0E+12
Linear
1.0E+13
(SPCS19)
#/kWh [3010D]
RD.08/######.#
© Ricardo plc 2009
11
Full flow Vs Partial Flow – Correlation between Mass Data

No correlation between mass from different measurement systems

PM from ESC cycle similar from both CVS and partial flow
ESC data similar
between CVS
and MDLT
RD.08/######.#
© Ricardo plc 2009
12
Partial flow Mass Vs Partial Flow Number

No obvious correlation between mass and number, even with very low background partial
flow system: mass method insensitive

Poor correlation on log-log chart
No obvious relationship
between PM and PN
400-fold
range in
particle
number
emissions
<2mg
spread in
mass
RD.08/######.#
© Ricardo plc 2009
13
Which is the ‘best’ system?

Using CoV as the discriminator…
– Background corrected mass should be
avoided
– Number is generally better than mass

Partial flow PM has the lowest average CoV

CVS PN has the lowest particle number CoV

But:
– Partial flow mass method lacks sensitivity
– CVS data has high background

Repeatability not a good metric!

Confirmation of observations in the ILCE_LD
RD.08/######.#
© Ricardo plc 2009
14
Summary

Measurements from Ricardo showed high particle and HC background in the
CVS, but very low background from partial flow system
– CVS >20 years old and used for many projects including DPF regeneration
• DPF calibration work may have contaminated the system leaving particle
precursors that were not removed despite the PMP preconditioning
– Partial flow dilution system new, but also much more practical to clean!

PN from CVS did not correlate well with partial flow PN except for highest
emissions cycles
– CVS limit of detection for PN was ~ 1011/kWh and ~8mg/kWh for mass

PM from partial flow tunnel did not correlate with PN from partial flow or PM from
full flow
– Partial flow limit of detection was <109/kWh and ~1mg/kWh for mass

CoV not seen as a good metric for assessing particle number or mass systems
RD.08/######.#
© Ricardo plc 2009
15
For consideration

If Ricardo CVS is representative of working CVS systems, then a PN limit of
1011/kWh is unfeasible due to high background levels
– 1012/kWh is possible
• Both partial and full flow systems could be used at this level

If partial flow only were permitted, a much lower PN limit could be applied

DPF regeneration has not been studied, and this could lead to higher
backgrounds in all dilution systems
RD.08/######.#
© Ricardo plc 2009
16