Article

Fuel Sulfur Effect on Membrane Coated Diesel Particulate Filter

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

The diesel particulate filter (DPF), in conjunction with fuel reformulation technologies such as ultralow sulfur fuel or Fischer-Tropsch diesel, represents a promising solution for reducing particulate emissions. In this work, membrane-coated and conventional uncoated SiC diesel particulate filters were tested on a 4-cylinder Volkswagen TDI diesel engine under four different engine load conditions at constant engine speed. Di-tert-butyl-sulfide was added to the base fuel to increase the sulfur content from 39 ppm to approximately 300 PPM. Gaseous and particulate mass emission measurements, as well as, PM morphology and composition have been used to address how engine-out and post-DPF emissions and post-oxidation catalyst emissions change with increasing fuel sulfur content. The influence of fuel sulfur on emissions was compared for the membrane coated and uncoated SiC filter using the same diesel oxidation catalyst (DOC) located downstream of the DPF. With the base fuel (39 ppm), the membrane coated filter shows lower PM emissions and higher PM filtration efficiency compared to the uncoated filter during the initial filling stage with a slight penalty in fuel consumption. Fuel sulfur had no effect on PM reduction efficiencies, with only a small increase in the soot+sulfate (i.e., insoluble) fraction and no change in the soluble organic fraction (SOF) of the PM mass. Particulate morphology, as seen by transmission electron microscopy (TEM), shows that the uncoated filter passes agglomerated particulate structures similar to the engine-out particles until a filter cake of PM is present. However, the membrane coated filter only allows small particles to pass at all times because the membrane coating provides a higher filtration efficiency.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... found that increasing porosity in DPFs decreases the pressure drop across the filter [33]. Boehman and colleagues found that applying coatings of granular silicon carbide to particulate filters can increase the filtration efficiency by replicating a soot cake [34]. While the coated filters had an increased backpressure, they allowed for the use of courser substrates with thinner walls, which can decrease backpressure [34,35,36]. ...
... Boehman and colleagues found that applying coatings of granular silicon carbide to particulate filters can increase the filtration efficiency by replicating a soot cake [34]. While the coated filters had an increased backpressure, they allowed for the use of courser substrates with thinner walls, which can decrease backpressure [34,35,36]. Therefore, it is reasonable to think that the combination of a coating to increase the filtration efficiency and using thin, highly porous walled GPFs has potential method for co-optimizing filtration efficiency and backpressure, which is the motivation for this work. ...
... To examine the hypothesis that a pseudo-cake would improve filtration efficiency without a large backpressure effect, two powdered materials were used to load the GPFs: silicon carbide (SiC) and calcium sulfate (CaSO 4 ). The materials were used because of the similarity of CaSO 4 to vehicular ash and the previous filtration efficiency improvement to DPFs that both materials exhibited [34]. Furthermore, silicon carbide can be bonded to the filter walls and kept in place by heating, which would make manufacturing simpler. ...
... found that increasing porosity in DPFs decreases the pressure drop across the filter [33]. Boehman and colleagues found that applying coatings of granular silicon carbide to particulate filters can increase the filtration efficiency by replicating a soot cake [34]. While the coated filters had an increased backpressure, they allowed for the use of courser substrates with thinner walls, which can decrease backpressure [34,35,36]. ...
... Boehman and colleagues found that applying coatings of granular silicon carbide to particulate filters can increase the filtration efficiency by replicating a soot cake [34]. While the coated filters had an increased backpressure, they allowed for the use of courser substrates with thinner walls, which can decrease backpressure [34,35,36]. Therefore, it is reasonable to think that the combination of a coating to increase the filtration efficiency and using thin, highly porous walled GPFs has potential method for co-optimizing filtration efficiency and backpressure, which is the motivation for this work. ...
... To examine the hypothesis that a pseudo-cake would improve filtration efficiency without a large backpressure effect, two powdered materials were used to load the GPFs: silicon carbide (SiC) and calcium sulfate (CaSO 4 ). The materials were used because of the similarity of CaSO 4 to vehicular ash and the previous filtration efficiency improvement to DPFs that both materials exhibited [34]. Furthermore, silicon carbide can be bonded to the filter walls and kept in place by heating, which would make manufacturing simpler. ...
... Reducing the pressure drop across the filter by altering the substrate properties, such as porosity and pore diameter, can often have a negative impact on the filtration efficiency [16]. Application of a membrane to the filter surface has been shown to be effective at improving filtration efficiency during the filling stage of a DPF [17], but this may cause increases in pressure drop that are not tolerable for GPF applications. An alternate approach is to operate the filter at lower filtration velocities which can also positively impact capture efficiency [18]. ...
... By comparison, the FWHM of mass-selected spherical particles is typically smaller than 5% [17]. Figure 4 (a) also shows the large sensitivity of mass of the fractal agglomerates to small shifts in the peak of dva distribution [18]. ...
Thesis
Full-text available
The harmful nature of combustion-generated nano-particle emissions from modern spark-ignited direct-injection (SIDI) engines has motivated particle-number based emissions regulations in different parts of the world. Gasoline particulate filters (GPFs) provide an effective means of reducing particulate matter (PM) emissions from the SIDI exhaust. Filter pressure drop and filtration efficiency (FE) are the primary metrics used to govern GPF design. In the current study, a series of filter wall-scale experiments were performed to understand the impacts of filter and particle properties and filtration conditions on the evolution of capture efficiency starting from a clean filter until transition to soot-cake filtration where the FE exceeds 99%. PM emissions from different steady-state SIDI operating conditions were characterized using a wide range of instruments. Although rich operation resulted in the largest fraction of agglomerated particles in the exhaust, heavy-load operation resulted in a wider range of particle shapes with shape factors in the transition and free-molecular regimes varying between 1.6-1.85 and 1.8-3, respectively. The relationship between particle mass and mobility diameter was independent of engine condition and the mass-mobility exponent was observed to lie within a narrow range of ∼2.3±0.15. An integrated particle size distribution (IPSD) method was used to estimate the mass concentrations in the SIDI exhaust. Average IPSD mass concentrations were 77 −32 +47 % of the gravimetric measurements performed using Teflon filters. Filter samples with a wide range of properties were characterized using multiple techniques to understand the filter microstructure. Mercury intrusion porosimetry (MIP) results showed that the samples had porosities ranging between 42-67% and median pore diameters (MPD) between 11-27 µm. The MIP results were also used to evaluate the width of the pore size distribution. ii Permeability measurements indicated a smaller pressure drop across filters with narrower pore size distribution when all other parameters were similar. Capillary flow porometry (CFP) measurements showed that the pore throat sizes fell within a narrow range of 1-30 µm. Trends from the CFP measurement generally agreed with the MIP results for the different samples. Porosity distribution profiles obtained using X-Ray CT technique showed lower porosities at the filter surface compared to the rest of the wall. The exhaust filtration analysis system (EFA) developed at the University of Wisconsin-Madison was used to perform wall-scale filtration experiments at a nominal temperature of 125°C for two different filtration velocities of 2.75 and 5.5 cm/s. The IPSD trapped mass in the filter was observed to be ~70% lower than corresponding gravimetric measurements. Particle penetration showed weak dependence on inlet PSD or the range of particle shapes explored in this study. A 10% difference in surface porosity was found to have little impact on the evolution of particle penetration with trapped mass. A significant impact of pore size distribution was observed on particle capture within the filter walls. Higher filtration velocity appeared to have a larger impact on penetration of 50-nm particles (50) for filters with a narrower pore size distribution. A simplified cylindrical pore flow model was used to explain the trends observed from the filtration experiments. The penetration results from the model showed larger deviation from experiments for filters with higher porosity. The EFA was modified to allow high-temperature experiments for future studies on the filter regeneration process. Increasing the filtration temperature from 125 °C to 550 °C resulted in a more than 10-fold decrease in 50. Preliminary filtration experiments at 600°C demonstrated the ability of the upgraded EFA system to perform in-situ filter regeneration studies. iii Acknowledgement
... The porous walls of the filter allow gas to flow through, capturing diesel particles, depending upon the porosity of material used in DPF (Twigg, 2007), while DPF may have limited influence in controlling the nonsolid fraction of diesel particulate emissions -the volatile organic fraction and sulfate particles. These volatile organic fraction emissions can be minimized or removed by typical oxidation catalysts, while sulfate particulates can be minimized or avoided by use of ultra-low sulfur fuels (Song et al., 2002). ...
Chapter
Full-text available
... Hence, these systems may be advanced by addition of functional components targeting the SOFtypically oxidation catalysts, while sulfate particulates may be minimized or avoided by use of ultra-low sulfur fuels. [45] DPFs accumulate a large amount of soot due to a low density of diesel PMs (~0.1 g/cm 3 ) and old generation heavy-duty engines can generate more than a few liters of soot per day. Removal of these particulates, also called filter regeneration, is necessary as filter clogging results in high exhaust gas pressure drop in the filter that negatively affects the engine operation. ...
Article
Full-text available
In this review, we have systematically discussed diesel particulate composition and its formation, understanding of which is essential to design the effective catalyst compositions. The most commonly used after treatment strategies such as diesel oxidation catalysts, diesel particulate filters, and partial flow filters are described followed by chronological and category-wise discussions on various groups of reported soot oxidation catalysts. A detailed review is also presented on mechanistic and kinetics aspects of non-catalytic direct particulate matter (PM) or soot oxidation in air/O2 and NO2. Recent progress in catalyst development with a focus on the low-cost catalyst for diesel PM oxidation has been given more emphasis considering their renewed importance.
... Reducing the pressure drop across the filter by altering the substrate properties can often have a negative impact on the FE. 16 Application of a membrane to the filter surface has been shown to be effective at improving FE during the filling stage of a DPF, 17 but this may cause increases in pressure drop that are not tolerable for GPF applications. An alternate approach is to operate the filter at lower filtration velocities which can also positively impact capture efficiency. ...
Article
Size-resolved particle mass and number concentrations were obtained from different operating conditions using a spark-ignition direct-injection engine and a heavy-duty diesel engine. Particle mass versus mobility diameter results obtained for the engines showed weak dependence on the operating condition. The particle mass–mobility data enabled the use of an integrated particle size distribution method to estimate the particulate matter mass concentration in the exhaust stream. Average mass concentrations determined with the integrated particle size distribution method were of the gravimetric measurements performed using Teflon filters. Despite the relatively low elemental carbon fraction (∼0.4 to 0.7), the integrated particle size distribution mass for stoichiometric spark-ignition direct-injection exhaust was 83% ± 38 % of the gravimetric measurement. Exhaust from the spark-ignition direct-injection engine was also used to perform wall-scale filtration experiments on identical cordierite filter samples with properties representative of diesel particulate filters. The filters were sequentially loaded with particulate matter from four spark-ignition direct-injection engine operating conditions, in order of increasing particulate matter mass concentration. Simultaneous particle size distribution measurements upstream and downstream of the filter sample were used to evaluate filter performance evolution and the instantaneous trapped mass within the filter for two different filter face velocities. The filtration experiments focused on the filter wall loading stage where the estimated trapped mass was < 0.3 g/m². The evolution of filtration performance at a fixed filtration velocity was found to only be sensitive to the trapped mass, despite using particulate matter from different operating conditions. Higher filtration velocity resulted in a more rapid shift of the most penetrating particle size toward smaller mobility diameters.
Conference Paper
During the first decade of Diesel particle filter development and deployment in cars, trucks, buses and underground sites, DPF regeneration methods were engineered that were compatible with the then prevalent high sulfur content in the fuel > 2000 ppm. The mainly used methods were burners, electrical heaters, replaceable filters and non-precious metal fuel additives. Low sulfur Diesel fuel became only available from 1996 in Sweden, 1998 in Switzerland, and after 2000 everywhere in Europe. Thus, the deployment of precious metal catalytic converters was feasible both as original equipment and retrofitting of in-use engines. The so-called CRT particle filters using PGM-catalysis for providing NO 2 for low temperature regeneration became very successful wherever ULSD was available. However, in many applications, e.g. off-road and in the construction industry, Diesel engines continued to run on high sulfur fuel and in many emerging countries, even on-road Diesel fuel still contains between 1000 and 2000 ppm sulfur. These countries suffer very much from air pollution through increasing Diesel particle emissions and the high impact of black carbon particles on human health as well as on the global warming is worrying. Hence, the necessity for modern particle filters that are compatible with high sulfur content of the fuel. In the context of Chinese megalopolis, this paper reports investigation of a fuel which is typical for China (containing > 1000 ppm sulfur) and compares results with European standard Diesel fuel. The test objects were two modern SiC particle filters, which were regenerated using different iron-based FBC. The combustion attributes of the soot were investigated by TGA and their EC/OC composition was examined. The results indicate that at the given test conditions the fuel sulfur does not significantly change the filters' physical and chemical properties. Neither the filter particle loading process nor the filter regeneration is noticeably different for the high sulfur test fuel compared to the ultra-low sulfur European fuel. Therefore VERT-verified iron-based FBC-type DPF can be used in countries where ULSD is not yet available.
Article
Measurement of particulate mass concentration in the exhaust of automotive engines by means of the filter sampling method prescribed by the USA's Environmental Protection Agency (EPA) is tedious, costly, time consuming and labour intensive. This paper presents a new approach, which is capable of measuring real-time particulate mass emissions in the exhaust of a diesel engine by using a tapered-element oscillating microbalance (TEOM) and a fast-response flame ionization detector (FID). The measuring principle rests on the fact that, by operating the TEOM at 50 °C, water interference is minimized, albeit at the expense of reducing the hydrocarbon content condensed on the particles. This is corrected by a method that consists of a measurement of hydrocarbon gas concentration, used with an empirical correction factor to recover what the filter sampling method would produce. The time resolution for the measurement using this new method is limited by the sampling window of the TEOM which is 0.42 s, though the response of the FID offers a time resolution of milliseconds. Results show that the mass concentration of particulate matter measured by TEOM and FID is well correlated to that based on the conventional EPA filter sampling method. Keywords: particulate mass concentration, tapered-element oscillating microbalance, flame ionization detector, diesel exhaust particulate measurement, correlation to EPA (USA) standard
Article
An account is given of a comparative examination of the formation and behaviour of sulphur carbons resulting from the reaction of representative polymer carbons with sulphur or hydrogen sulphide. Non-stoichiometric retention of sulphur was accompanied by reasonably systematic changes in ESR shown by the carbons; but this behaviour was distinctly less systematic in the reaction with hydrogen sulphide than with sulphur vapour. The thermal stability and chemical behaviour of the sulphurised carbons, including desulphurisation by hydrogen, was also examined.It was concluded that the sulphurised carbons probably comprised chiefly carbon-sulphur complexes, in the formation of which the nature and availability of the surface and the presence of unpaired spin centres and of hydrogen played an important part. Indications of the close association between loss of unpaired spin centres and sulphur retention also suggested that coupling of the reacting sulphur with spin centres present initially or formed by dehydrogenation, was a feature of the carbon-sulphur vapour reaction. It is suggested that the bonding chiefly concerned is akin to that in thioethers or disulphides, and may effectively form peripheral heterocyclic structures on the exposed layer lattices in the carbon surface, or sulphur bridges between neighbouring layers. This suggestion conforms with the general observation that the most disordered carbons are most susceptible to sulphurisation.
Article
The effects of sulfur dioxide on soot formation and on concentrations of polycyclic aromatic hydrocarbon (PAH) species have been investigated in a premixed ethylene-oxygen-argon flame. Sulfur dioxide reduces PAH concentrations and reduces the soot volume fraction without affecting the number density, coagulation rate, or the decay constant for the surface reactivity of the soot particles. these effects parallel those previously observed for oxygen addition to rich flames, and are consistent with consumption of SOâ in the sooting region. The oxygen and sulfur atoms become available for the formation of compounds such as COâ and CSâ, thereby reducing the amount of carbon available for soot growth.
Article
Laser-light scattering and absorption measurements of soot volume fractions and particle number densities in premixed ethylene-air flat flames at atmospheric pressure show that additions of up to 3%, by volume of the unburned gas, of N2, H2O, H2, NH3, CH3NH2, NO, H2S, SO2, SO3, HCl, CH4, C2H4 or O2 do not influence the coagulation of soot particles although the soot volume fraction may be significantly affected. Pro-soot additives are CH4, C2H4 and CH3NH2 while additives which reduce the soot volume fraction are O2, NO, NH3, H2S, SO2 and SO3. Whatever their effects, these additives do not alter the specific surface growth rate for soot particle mass addition.The various additives retain their pro- or anti-soot tendencies when considered in terms of their effects on the critical C/O-ratio for the appearance of soot. Addition of H2 promote soot under these conditions.Special attention is paid to the effect of SO3—no evidence of any pre-soot effects at or beyond the critical C/O-ratio is found in aliphatic or aromatic flames.The mode of action of the additives is not generally understood, although the formal equivalence of the addition of one molecule of NH3 with removal of one fuel-C atom is consistent with the expected rapid formation of inert HCN in the flame and with the relatively weak effects of CH3NH2 addition.