Embed Size (px)
Citation preview
DEER CONFERENCE 2008 1
Fuel Injection Strategy for Soot-Filter Regeneration
Fabien A. RioultFlorian Von TrentiniMarius Vaarkamp
DEER CONFERENCE 2008 2
Outline
•
Regeneration of Soot Filters
•
Causes of Soot Filter Failures
•
Modeling of Active Regeneration and Drop-to-Idle events
•
Injection Strategy
•
Conclusion
DEER CONFERENCE 2008 3
Soot Filters in Diesel Exhaust Systems
• Soot thickness increases with time
•
Backpressure in the exhaust system increases = loss of engine power
• Overload can clog the filter
Soot cake
As the soot accumulates in the filter, the backpressure of the exhaust increases. The soot needs to be removed continuously or periodically.
DEER CONFERENCE 2008 4
Passive and Active Regeneration The exhaust temperature is not always hot enough to regenerate
passively the soot filter. Then, active regeneration becomes necessary.
NOCONOC(soot)NO 2CO NO 2 C(soot)
2
22
+→++→+
COO21C(soot)
CO O C(soot)
2
22
→+
→+
Passive Regeneration (optimal at ≈300oC)
Active Regeneration (>550oC)[DieselNet.com]
22 NOO21NO →+
DOC CSF
Fuel injector
DEER CONFERENCE 2008 5
High Temperature is a Risk of Failure for Soot Filters
•
Loss of catalytic activity
•
Sintering of PM particles
•
Alumina polymorphic transformation (1100oC)
•
Ash reaction with cordierite
•
Solid-state reaction with cordierite (1200oC)
•
Filter-ash eutectic
•
Mechanical stress
•
Thermal gradients
•
Fatigue
•
Melting point of Cordierite (1450oC) [Photos from
DieselNet.com]
DEER CONFERENCE 2008 6
Drop-to-Idle: Sudden drop of the exhaust gas flow rate during an active regeneration. The heat from the soot oxidation accumulates in the CSF and leads to a runaway reaction.
• Sudden drop of the exhaust flow rate
Flow
HC Injection
t
D
Variable
Variable
DEER CONFERENCE 2008 7
Speed = x 30
Active Regeneration SimulationAs the temperature increases in the CSF, the soot regeneration rate increases.
Time
DEER CONFERENCE 2008 8
Active Regeneration with Drop-to-Idle Simulation
Speed = x 60
The flow rate is too low to evacuate the heat created by the soot oxidation.
Time
DEER CONFERENCE 2008 9
Filter Model (BASF CatSim) / Assumptions
CO, HC, NO, NH3,…
P, T, F CO2, H2O, N2, NO2, …
Heat and MassTransfer
Thermal losses to ambient
Axial conduction
Thermal conduction to can
Reactions
Soot Accumulation
Soot Regeneration
•
Soot oxidation yields equal amounts of CO and CO2 (CO is converted over the PM function of the catalyst)
•
No axial motion of the soot from the cake•
1D Model (no radial temperature gradients)•
Homogeneous soot distribution•
Uniform deactivation
DEER CONFERENCE 2008 10
Active Regeneration - DOC+ CSF - 600C DOC out - 3.9g/l soot loading
300
400
500
600
700
800
900
200 300 400 500 600 700 800 900 1000 1100 1200Time (s)
Tem
pera
ture
(C)
TC_25mmTC_152mmTC_279mmTC_406mm60 C
Thick line: model
Thin line: experiments
Comparison of Model to Experimental Results. The model underestimates the maximum temperature in the CSF
during an active regeneration (radial thermal gradients are not considered).
23K/hr
DEER CONFERENCE 2008 11
Fuel Injection Strategy: Objectives
•
Propose an injection strategy for active regeneration with the following constraints:
•
The maximum temperature in the CSF during a drop-to-idle event must not be higher than 700oC.
•
Reach 90% regeneration.
•
Decrease the duration of the regeneration.
DEER CONFERENCE 2008 12
Experimental Conditions
•
Initial soot loading: 3.5 gm/l
•
Inlet Gas Temperature: 280oC
•
Constant filter dry gain of 0.25 gm/in3
•
Drop-to-Idle: 10 to 3 K/hr ; 20 to 3 K/hr Uniform DOC, 300/8, 12”x8”
PM : 60 gm/ft3
(4:1)
Uniform CSF, 270/16, 13”x17”
PM : 1.2 gm/ft3
(4:1)
DOC + CSF
DEER CONFERENCE 2008 13
Drop-to-Idle Test (DTI)
•
The flow is decreased from 125 gm/s to 35 gm/s and the HC injection is decreased to 0 gm/s after a variable duration D.
•
Several scenarios (different injection duration D) of drop-to-idle are considered in order to obtain the worst case.
10 to 3K/hr
Flow
HC Injection
tD
Variable
Variable
X
The maximum temperature in the CSF during drop-to-idle test has been calculated for variable injection rates and duration.
DEER CONFERENCE 2008 14
Drop-to-Idle Test (DTI)
10 to 3K/hr
Flow
HC Injection
tD
Variable
Variable
X
Flow
HC Injection
tD
Variable
Variable
X
Increasing the injection rate during the regeneration will allow to avoid high temperatures in case of DTI.
CSF-in: 570oC
CSF-in: 520oC
CSF-in: 560oC
X:
DEER CONFERENCE 2008 15
Comparison between a Constant Injection and a Staged Injection.
Fuel injection rate
Time
X
reference
90 sec
3 step injection
As the soot is consumed, the injection rate can be increased. The following injection represent a 3 additional step injection respectively corresponding to 110, 120 and 130 % of the initial injection. The duration of the steps has been adjusted so that the maximum drop to idle temperature never reaches 700oC.
Fuel injection rate
DEER CONFERENCE 2008 16
0
10
20
30
40
50
60
70
80
90
100
0 200 400 600 800 1000 1200 1400Time (s)
Reg
ener
atio
n (%
)
0
200
400
600
800
1000
1200
1400
1600
1800
2000
Fuel
inje
cted
(gm
)
3 steps - Fuel injected (gm)3 steps - Regeneration (%)
0
10
20
30
40
50
60
70
80
90
100
0 200 400 600 800 1000 1200 1400Time (s)
Reg
ener
atio
n (%
)
0
200
400
600
800
1000
1200
1400
1600
1800
2000
Fuel
inje
cted
(gm
)
3 steps - Fuel injected (gm)Reference - Fuel injected (gm)3 steps - Regeneration (%)Reference - Regeneration (%)
The staged injection shows a significantly shorter regeneration time and lower fuel consumption than the reference injection.
Constant VS Staged Injection: SV (10K/hr)
d0
=260sd1= d2
=100s
10K/hr
10 min
42% time economy
10 min
42% time economy
38% fuel economy
38% fuel economy
90% Regeneration
16 minutes
90% Regeneration
16 minutes
DEER CONFERENCE 2008 17
0
10
20
30
40
50
60
70
80
90
100
0 200 400 600 800 1000 1200 1400Time (s)
Reg
ener
atio
n (%
)
0
500
1000
1500
2000
2500
3000
Fuel
inje
cted
(gm
)
3 steps_Fuel injected (gm)
Reference_Fuel injected (gm)
3 steps_Regeneration (%)
Reference_Regeneration (%)
The staged injection shows a significantly shorter regeneration time and lower fuel consumption than the reference injection.
Constant VS Staged Injection :SV (20K/hr)
d0
=260sd1= d2
=100s
20K/hr
11 min
46% time economy
11 min
46% time economy
43% fuel economy
43% fuel economy
90% Regeneration
14 minutes
90% Regeneration
14 minutes
DEER CONFERENCE 2008 18
Staged injection : DTI testDrop-to-idle events have been simulated for variable injection duration during this stepped injection.
Fuel injection rate
Time
X 1.1X
d0
3 steps
90 sec
d1d2
1.2X 1.3X
D(s), injection time before drop-to-idle140 sec
20 to 3K/hr
10 to 3K/hr
DEER CONFERENCE 2008 19
The staged injection strategy allows to keep the maximum temperature of the uniform CSF (with upstream DOC) below 700oC whenever the drop-to-idle event occurs.
10 to 3K/hr
400
450
500
550
600
650
700
750
800
850
900
0 100 200 300 400 500 600 700D (s)
Max
. Tem
p in
CSF
(C)
3 steps - 10K/hr
3 steps - 20K/hr
20 to 3K/hr
Staged injection : DTI test
DEER CONFERENCE 2008 20
Conclusion / Path Forward
• Staged injection strategy allows:
•
to respect a desired maximum temperature in the CSF when a drop-to-idle event occurs (700oC for this study).
• to decrease the fuel consumption
• to decrease the regeneration duration = decrease risk of DTI
•
A better tuning of the steps duration or the addition of intermediary steps (or even using a continuous function) would allow further improvement of the regeneration.
•
Experimental verification of the injection strategy (Model results validation)
•
Steady-state → Transient
Path Forward
Conclusions
DEER CONFERENCE 2008 21
Acknowledgement
Florian Von Trentini
Dennis Anderson
Marius Vaarkamp
Sanath Kumar
Ken Voss
Alfred Punke
Kevin Hallstrom
Joe Patchett
Matt Larkin
Maurica Fedors
