Search Results

The use of catalyzed diesel particulate filter (CDPF) systems in light duty diesel (LDD) vehicles is becoming increasingly common. The primary functions of the system are to remove carbon monoxide (CO) and hydrocarbons (HC) from the vehicle exhaust stream, while simultaneously reducing the level of particulate matter (PM) emissions to ambient background levels. These systems can comprise either a separate diesel oxidation catalyst (DOC) and a downstream CDPF, or a single unit CDPF with the DOC functions incorporated within the CDPF. The single CDPF unit provides higher regeneration efficiency as it is located nearer to the engine and also cost benefits, as only a single unit is required compared to the alternative separate DOC and CDPF arrangement. A model describing the performance of the single unit CDPF for emissions control has been developed, with particular emphasis on achieving predictions of the CO and HC emissions over transient vehicle drive cycles.

The use of vanadia selective catalytic reduction (V-SCR) catalysts for NOX reduction from diesel engine exhaust is well known. These catalysts are also active for hydrocarbon (HC) and particulate matter (PM) oxidation. This dual functionality (oxidation and reduction) of V-SCR catalysts can help certain applications achieve the legislative limits with an improved margin. In this work, NOX reduction, HC and CO oxidation over V-SCR were studied independently and simultaneously in microreactor tests. The effect of various parameters (HC speciation, concentration, ANR, and NO₂/NOX ratio) was investigated and the data was used to develop a kinetic model. Oxidation of CO, C₃H₆, and n-C₁₀H₂₂ is first order in CO/HC, while C₇H₈ oxidation is less than first order in C₇H₈. All these reactions were zero order in O₂. Oxidation activity decreased in order: C₇H₈ ≻ n-C₁₀H₂₂ ≻ C₃H₆ ≻ CO. HC oxidation was inhibited by NH₃.

Particulate Matter (PM) emissions are of increasing importance, as diesel emissions legislation continues to tighten around the world. Diesel PM can be controlled using Diesel Particulate Filters (DPFs), which can effectively reduce the level of carbon (soot) emissions to ambient background levels. The Johnson Matthey Continuously Regenerating Trap (CRT®) [1], which will be referred to as the Continuously Regenerating DPF (CR-DPF) for the remainder of this paper, has been widely applied in Heavy Duty Diesel (HDD) applications, and has been proved to have outstanding field durability [2]. To widen the potential application of this system, addition of a platinum based catalyst to the DPF has been shown to lead to a higher PM removal rate under passive regeneration conditions, using the NOx contained in the exhaust gases.

The aftertreatment challenge in the non-road market is making the same system work and fit not just in one machine, but in hundreds of different machines, some of which can be used for many different purposes. This huge diversity of applications and the relatively small unit numbers for each application, coupled with the rapid introduction of new standards and the very high performance needed from the engines and machines, requires a sophisticated approach to product development. Furthermore, as emissions requirements become ever more stringent, designing a system to meet the legislation subject to packaging and cost constraints becomes progressively more difficult. This is further exacerbated by increasing system complexity, where more than one technology may be required to control all the legislated pollutants and/or an active control strategy is involved. Also a very high degree of component integration is required.

A 1-D numerical model describing the ammonia selective catalytic reduction (SCR) de-NOx process has been developed based on data measured on a laboratory microreactor for a vanadia-titania washcoated catalyst system. Kinetics for various NH3-NOx reactions were investigated, as well as those for ammonia, CO and hydrocarbon oxidation. The model has been successfully validated against engine bench measurements, over light-off and ESC tests, under a wide range of conditions, e.g. flow rate, temperature, NO2/NO ratio, and ammonia injection rate. A very good agreement between the experimental data and the model has been achieved. The model has now been used to predict the effect of NO2/NO ratio on NOx conversion, and the effect of different ammonia injection rates on the efficiency of the SCR process.

A one-dimensional numerical model for a Cu-zeolite SCR catalyst has been developed. The model is based on kinetics developed from laboratory microreactor data for the various NH₃-NOX reactions, as well as for NH₃ oxidation. The kinetic scheme used is discussed and evidence for it presented. The model is capable of predicting the conversion of NO and NO₂, NH₃ slip and the formation of N₂O, as well as effects associated with NH₃ storage and desorption. To obtain a good prediction of catalyst temperature during cold start tests, it was found necessary to include storage and desorption of H₂O in the model; storage of H₂O is associated with a sizable exotherm and the subsequent desorption of this water produces a correspondingly large endotherm.

The influence of SCR (selective catalytic reduction) activity on soot regeneration was investigated using engine test measurements with and without urea dosing on a vanadia-SCRF®1, also known as a vanadia SCR coated diesel particulate filter (V.SCR-DPF). The extent and rate of passive soot regeneration is significantly reduced in the presence of SCR activity especially at high temperatures (>250 °C). The reduction in soot regeneration is because some of the NO2, which would otherwise react with the soot, is consumed by SCR reactions and consequently the rate of soot regeneration is lower when urea is dosed. The converse effects of soot oxidation on SCR activity were studied separately by analysing steady-state light-off engine measurements with different initial soot loadings on the V.SCR-DPF. The measurements show an increase in NOX conversion with increasing soot loading.

Asymmetric particulate filters (PF), where the inlet channel is wider than the outlet channel, are commonly used because of their greater capacity for ash. Somewhat surprisingly, very few models for asymmetric PFs have been published and none of these gives a correct/detailed description of the geometry. For example, octahedral channels may be treated as if they were square or the tapering walls between the inlet and outlet channels treated as if they were rectangular in cross section. Alternatively, the equations may be presented in generic form in terms of channel cross-sectional areas and perimeters, but without giving any indication of how to calculate these. This paper aims to address these deficiencies with a model that correctly describes the geometry of square and octo-square asymmetric PFs. Expressions for the solid fraction of the PF (which affects thermal mass) and channel cross section and perimeter (both when clean and soot/ash loaded) are presented.

Diesel emissions legislation continues to tighten around the world, and Particulate Matter (PM) emissions are currently the focus of much attention. Diesel PM can be controlled using Diesel Particulate Filters (DPFs), which can effectively reduce the level of carbon (soot) emissions to ambient background levels. In the Heavy Duty Diesel (HDD) area, the Continuously Regenerating Trap (CRT®) [1] has been widely applied in the retrofit market. This system will henceforth be referred to as the Continuously Regenerating DPF (CR-DPF). There are currently over 100,000 of these systems in use in retrofit applications worldwide. This system comprises a specially formulated Diesel Oxidation Catalyst (DOC) upstream of a DPF; the NO2 generated by the DOC is used to combust the carbon collected in the DPF at low temperatures. A model describing the performance of the CR-DPF has been developed.

Asymmetric particulate filters (PF), where the inlet channel is wider than the outlet channel, are commonly used because of their greater ash capacity. Surprisingly, very few models for asymmetric PFs have been published. This paper considers how to model the soot cake in octo-square asymmetric PFs. Some previous studies have neglected the octahedral shape of the inlet channel and instead assumed that the inlet channels were square. As the correct approach for modelling the soot cake is not obvious, three options are considered. The calculation of soot-loaded channel perimeter and hydraulic diameter (which are important for heat and mass transfer), soot thickness and backpressure as a function of soot loading are given for each geometry. In option 1, the shape of the soot-loaded channel is assumed to be geometrically similar to the soot-free channel.

In recent years magnetic resonance imaging (MRI) has been shown to be an attractive method for fluid flow visualization. In this work, we show how MRI velocimetry techniques can be used to non-invasively investigate and visualize the hydrodynamics of exhaust gas in a diesel particulate filter (DPF), both when clean and after loading with diesel engine exhaust particulate matter. The measurements have been used to directly measure the gas flow in the inlet and outlet channels of the DPF, both axial profiles along the length and profiles across the channel diameter. Further, from this information we show that it is possible to indirectly ascertain the superficial wall-flow gas velocity and the soot loading profiles along the filter channel length.

Selective Catalytic Reduction (SCR) technology has been widely studied for removal of NOX from the exhaust of diesel engines. To design and optimize diesel engine aftertreatment systems including an SCR catalyst component, a reliable SCR model is a very useful tool, to aid in system integration and control algorithm testing. In this paper, the development of a one-dimensional numerical model for a Fe-Zeolite-based SCR catalyst (hydrothermally aged for 100 hours at 650°C in 10% H₂O in air) is presented, followed by its validation and application. The resulting model is capable of predicting NOX reduction efficiency under various operating conditions as a function of gas hourly space velocity (SV), temperature, NO₂/NOX ratio and NH₃ to NOX (ANR) ratios; NH₃ slip and N₂O formation are also correctly predicted by the model. Extensive validation of the model has been carried out against engine test data for both steady state light-off and the heavy-duty FTP transient cycle (HD-FTP).

Porous wall permeability is one of the most critical factors for the estimation of backpressure, a key performance indicator in automotive particulate filters. Current experimental and analytical filter models could be calibrated to predict the permeability of a specific filter. However, they fail to provide a reliable estimation for the dependence of the permeability on key parameters such as wall porosity and pore size. This study presents a novel methodology for experimentally determining the permeability of filter walls. The results from four substrates with different porosities and pore sizes are compared with several popular permeability estimation methods (experimental and analytical), and their validity for this application is assessed. It is shown that none of the assessed methods predict all permeability trends for all substrates, for cold or hot flow, indicating that other wall properties besides porosity and pore size are important.

The development of a one-dimensional model for the prediction of backpressure across a gasoline or diesel particulate filter (PF) is presented. The model makes two innovations: Firstly, the term for momentum convection in the gas momentum balance equations includes the loss (or gain) of axial momentum in the direction perpendicular to the channels; neglecting this results in the momentum convection term being too large. Secondly, equations for the pressure change due to the abrupt contraction at the PF entrance and for abrupt expansion at the exit are derived which take into account the fact that the velocity profile across the channels is not flat; often workers have used equations appropriate for high Reynolds numbers which assume flat velocity profiles. The model has been calibrated/tested against cold flow data for more than one length of PF. The use of more than one length allows along-filter pressure losses to be separated from entrance and exit effects.

Microreactor, engine bench tests and modelling studies have been carried out to understand the influence of cold start (low temperatures) on the performance of NH3/urea-SCR automotive exhaust aftertreatment systems. Water storage experiments using Simultaneous Thermal Analysis (STA) coupled with numerical modelling demonstrated that the exo/endo-therms associated with water adsorption and desorption at temperatures below 150°C strongly influence the catalyst temperature. Appreciable amounts of NO and NO2 could be stored on the catalyst during reactor or engine testing in the absence of any NH3 (blank tests). Modelling studies at different inlet NO2/NOx ratios demonstrated some of the influences of these surface adsorbed species on the performance of the SCR system during cold start.

Future heavy-duty diesel (HDD) engines are designed to have higher engine out NOx, for improved fuel economy, while reduction of the emission control technology footprint is also desired. Consequently, higher NOx reduction across compact emission control systems is required. Selective catalytic reduction (SCR) catalyst coating combined with a wall flow particulate filter (SCRF®1) is a technology that enables abatement of NOx emissions in addition to oxidation of soot from diesel engine exhausts. Vanadia based-SCR is well known for NOx reduction and is active for hydrocarbon (HC) and particulate matter (PM) oxidation. This dual functionality (oxidation and reduction reactions) of the V.SCR catalysts plus the filtration achieved by the filter substrate can help certain diesel engine applications achieve the legislative limits with a reduced packaging volume.