NOx-Particulate Filter (NPF): Evaluation of an After-Treatment Concept to Meet Future Diesel Emission Standards
Marco Ranalli, Stefan Schmidt, Lee Watts
ArvinMeritor LVS, Air and Emission Technologies
Abstract
Simultaneous particulate and NOx reduction represents the next step to the reduction of diesel emissions. One of the most promising concepts to achieve this target involves the combination of two technologies already in use in the after-treatment technology - Diesel Particulate Filter and NOx Storage Catalyst - in the same component. The major issue to be solved is the design of a complex thermal strategy, for the regeneration of NOx emissions, particulate matter and possibly sulfates. For this set-up to function properly the engine must periodically generate a rich spike to induce the NOx de-sorption process. The system must also increase the exhaust gas temperature to induce the soot oxidation process. Complicating matters further, the regeneration process of the filter must also be controlled to avoid substrate or washcoat damage.
- To fully understand the process an extensive investigation of an NPF (NOx Particulate Filter) was conducted using a 6-cylinder diesel engine on a test bed. To support de-NOx and induce the soot regeneration conditions an in-house developed diesel fuel vaporizer was used, which was placed upstream of the DOC-NPF assembly. To enable the system to be tested, a complete regeneration strategy for soot and NOx after-treatment was then developed and tested.
The main objective of this work was to verify if these three technologies could be reasonably combined in order to obtain a reliable diesel after-treatment concept. The investigations showed interesting benefits in combining NSC and DPF technology, like an increased NOx storage capacity in presence of soot. Important effects of the regeneration strategy design over the thermal behavior of the NPF are also outlined.
Introduction
One of the aims of this exercise was to investigate if a combined NOx and Soot regeneration was possible (Figure 1) as the time intervals between regenerations are very different (minutes for NOx and hours for soot). It was recognized what was being attempted was extremely challenging. Therefore the process is considered to be more of a joining of two processes than merging of two processes. The investigation of possible interactions/benefits of a combined system and strategy were also a main focus of this work.
Figure 1:
Regeneration strategy
It is important to mention that, in order to achieve rich gas conditions, an in-house developed diesel fuel vaporizer was used. Main advantages of this device are the lower fuel amount necessary for the soot oxidation events and the reduced risk of oil dilution.
Measurement set-up
The test set-up consisted of a 3 litre Euro 3 compliant diesel engine that had modifications to the EGR, VGT, injection strategy and the intake throttling, therefore enabling the engine to produce rich spikes. The pre-catalyst was placed 100 mm from the turbo outlet after which followed the vaporizer, then the DOC (300mm from vaporizer) and NPF as illustrated in Figure 2. The NPF was a 5.66x6 300/12 SiC substrate (high porosity) coated with a NOx adsorbing and reducing catalyst. The total NOx storage capacity was around 0.8 g NOx (or 0.32 g/liter) so originally, the system was composed of two NPF in parallel, in order to provide sufficient storage capacity to the 3l engine. However, only one of the two substrates was used at a time in a single-line exhaust system (see Fig. 2), because decreasing the trap volume in comparison to the engine displacement enabled phenomena like conversion efficiency and soot deposition to be more easily recognized.
The two main measurements performed with this set-up were:
1) A series of only NOx regeneration events (rich spike)
2) A series of soot and NOx regeneration events (see strategy)
A diesel fuel with low sulfur content (10 ppm) was used for all investigations.
Figure 2: system set-up
Regeneration strategy
The combined regeneration of the NPF has a two-phase structure: soot oxidation first and then NOx desorption (de-NOx). The strategy was first tested without NPF, with emissions and temperature values measured downstream of the DOC. Figure 3 shows the increase of vaporized hydrocarbons (VAP FREQ) with the consequent reduction of lambda (LAMBDA) and temperature increase after the DOC (TdownCat). When the temperature achieves 580°C-600°C (soot oxidation temperature) the soot regeneration takes place. After a short period of time the engine activates its measures (increased EGR, injection parameters, intake throttling) to induce a low AFR period (rich spike). As a result, lambda decreases causing a rich mixture resulting in a higher production of CO, which is the optimal condition for NOx regeneration.
One of the main benefits of this strategy, is the low level of oxygen present after the soot has ignited. Therefore any kind of exothermic reaction either on the DOC (hydrocarbon oxidation) or in the trap (soot oxidation) is first slowed down and then eventually stopped, resulting with a reduction of the temperature inside the substrates. This strategy has also some benefits due to the switch from lean-high temperature to rich-low temperatures. The transition produced a phase with high temperatures and low AFR, which works as a desulphurisation phase.
Figure 3: combined regeneration strategy
no X rEGENERATION (RICH SPIKE FOR DESORPTION AND REDUCTION) CYCLES
Figure 4: de-NOx cycles Figure 4 shows a series of NOx regenerations, which are clearly represented by the peaks in CO and the significant reduction in Lambda and percentage of oxygen present. However, the most interesting factor is the relatively long time interval between NOx regenerations before an efficiency of approximately 50% is reached. During the investigation it was identified that the apparent NOx storage during the lean phases was approximately 1.8 g/l during the lean period, at which it was not saturated. This is at least twice as much as originally expected. The regeneration intervals seemed also to increase as the soot accumulates on the filter.
As a result of the rich spike settings, the soot emissions increased by a factor 4 (from 1 to 6 FSN) during the de-NOx events.
Figure 5: back-pressure and temperature inside the NPF
Figure 5 provides a clear representation of the increase in the back-pressure [mbarx10] after the NOx regeneration events. The peaks in the back-pressure during the rich phases are a consequence of the increased mass flow due to the engine settings. The plot also illustrates the temperature increase within the trap at different locations, which also corresponds with the rich spikes that are highlighted by the changes in Lambda. The temperature increase after the de-NOx events is a result of the oxidation of residual hydrocarbons when the engine turns lean again. As a consequence of the soot deposition during the rich phases, the back-pressure curve shows a constant increase. This would probably soon lead to unacceptable backpressure values, if a soot regeneration process did not occur.
Figure 6: cumulative NOx emissions
Figure 6 shows the cumulative NOx and NO emissions upstream and downstream of the NPF. The reduction in NOx is approximately 80% and the reduction in NO is approximately 67%. A possible explanation for the difference in reduction is thought to be a result of the NO2 reacting with the soot, as a result producing NO. In addition NO2 is preferentially absorbed by the NPF coating.
sOOT & noX rEGENERATION (t INCREASE ABOVE 580°C + RICH SPIKE) CYCLES
As outlined in the previous chapter, the system back pressure would increase without periodical soot regeneration events. A soot regeneration phase was added before the periodical de-NOx phases.
Figure 7 Shows a series of combined NOx and Soot regeneration at the same operation point. This effect is highlighted by the different Lambda trace, which now comprises of two phases, the initial drop caused by the vaporizer and the second phase being caused by the engine parameters.
Figure 7: NOx emissions during soot and NOx regeneration cycles
Here, like in the NOx – only - regeneration cycle, NOx emissions are low and the intervals to overcome 50% reduction efficiency relatively long.
In this case a temperature increase over 600°C produces the necessary conditions for the soot regeneration. As an expected side effect, the NOx storage capacity of the NPF decreases considerably during this phase, resulting in a NOx release clearly visible during the Soot regeneration phases.
It is understood that during the high temperature lean excursions (soot regeneration), a certain amount of NOx slip occurs. However, it is believed that this can be reduced by improving the regeneration strategy. Investigations with different regeneration strategy parameters aimed to reduce the NOx slip are ongoing.
Figure 8: Backpressure during soot and NOx regeneration cycles
In Figure 8 the effect of the periodical soot regeneration is clearly visible from the back-pressure curve. The initial small increase in backpressure is a result of the soot emission produced during the de-NOx phase. However this soot deposit is soon removed when the next soot regeneration takes place. After that, the soot loading does not increase, as a result of the cyclic soot removal through regeneration.
The temperature inside the NPF rises very rapidly during the de-soot phase. The high PGM loading on the NPF promotes the oxidation of the available hydrocarbons during the soot regeneration phase. In addition, when soot is regenerating the temperature increase is even faster. It is by reducing the oxygen content during the de-NOx phase that the increase in the temperature can be first controlled and then inverted. All the maximum measured temperatures are between 600’C and 650’C, all of which demonstrate low maximum temperature inside the NPF and well-matched exothermic reactions.
Figure 9: cumulative NOx emissions
Figure 9 illustrates the cumulative NOx and NO emissions, similar to that of Figure 6. The results on Figure 9 although being worse than those with the low temperature NOx regeneration. The conversion efficiency is still relatively high. As previously mentioned, these results are from the preliminary investigation to prove the principles of operation. Therefore with further development and optimization better results should be achievable.
Conclusion
These investigations showed potentials and problems of a promising strategy for the control of diesel emissions. Major effort seems to be necessary to reduce the NOx slip and to keep regeneration temperature under acceptable levels, but the approach seems to be correct. With the correct tuning of the strategy and design, a complete PM filtration and a NOx reduction system to be feasible. An appropriate strategy design can keep control of the temperature issues, caused by the soot regeneration. The fuel penalty, as of yet, remains to be investigated in detail. In regards to the additional investigations being conducted, it has been decided to use the current EURO3 engine with DPF system as a point of reference.
Acknowledgments
Thanks are due to Andrew Chiffey and Philip Blakeman who provided the NPF systems and to Hans-Juergen Bruene for the precious support in the engine strategy design.
References
K. Ohno, K. Shimato, N. Taoka, H. Santae, T. Ninomiya, T. Komori and O.Salvat: Characterization of SiC DPF for Passenger Car. SAE Paper 2000-01-0185
Fekete et al.: Evaluation of NoOx Adsorber Catalyst System to Reduce Emissions of Lean Running Gasoline Engines. SAE Paper 970746
A.Shoji, S.Kamoshita, T.Watanabe, T.Tanaka; Development towards Serial Production of a Direct Injection Diesel Engine for Light-Duty Truck with Simultaneous Reduction System of NoOx and PM. Internationales Wiener Motorensymposium 2003
P.Blakeman, P.Andersen, H.Chen, D.Fonsson, P.Phillips and M.Twigg: Performance of NOx Adsorber Emissions Control Systems for Diesel Engines SAE paper 92003-01-0045
FVV-Abschlussbericht: Diesel-NOx-Speicherkatalysator. Heft 719.- 2001 (in German)
U.Pfahl, W. Hirtler and T.Cartus: Passenger Car Investigations of NOx-Adsorber and DPF Combination to Fulfill Future Diesel Emission Limits. SAE Paper 2003-01-0043
S.Michon, C.Gerentet, L.Pierron, T.Colliou, J.B.Dementhon and B,Martin; Application of a NOx trap for Emission Reduction of a Heavy-Duty Engine. Internationales Wiener Motorensymposium 2001
Contact
Marco Ranalli, COC Emissions – New Technologies
marco.ranalli@arvinmeritor.com
ZEUNA STAERKER GmbH & Co.KG, Biberbachstrasse 9 D-86154 Augsburg , Germany
Definitions, Acronyms, Abbreviations
de-NOx NOx desorption
AFR Air-Fuel Ratio
BP Back-pressure
DOC Diesel Oxidation Catalyst
DPF Diesel Particulate Filter
NPF N Ox-Particulate Filter
NSC NOx Storage Catalyst
|
