Search Results

In this paper, a model-based linear estimator and a non-linear control law for an Fe-zeolite urea-selective catalytic reduction (SCR) catalyst for heavy duty diesel engine applications is presented. The novel aspect of this work is that the relevant species, NO, NO2 and NH3 are estimated and controlled independently. The ability to target NH3 slip is important not only to minimize urea consumption, but also to reduce this unregulated emission. Being able to discriminate between NO and NO2 is important for two reasons. First, recent Fe-zeolite catalyst studies suggest that NOx reduction is highly favored by the NO 2 based reactions. Second, NO2 is more toxic than NO to both the environment and human health. The estimator and control law are based on a 4-state model of the urea-SCR plant. A linearized version of the model is used for state estimation while the full nonlinear model is used for control design.

Model-based control strategies are important for meeting the dual objective of maximizing NOx reduction and minimizing NH3 slip in urea-SCR catalysts. To be implementable on the vehicle, the models should capture the essential behavior of the system, while not being computationally intensive. This paper discusses the adequacy of two different reduced order SCR catalyst models and compares their performance with a higher order model. The higher order model assumes that the catalyst has both diffusion and reaction kinetics, whereas the reduced order models contain only reaction kinetics. After describing each model, its parameter identification and model validation based on experiments on a Navistar I6 7.6L engine are presented. The adequacy of reduced order models is demonstrated by comparing the NO, NO2 and NH3 concentrations predicted by the models to their concentrations from the test data.

The catalyzed particulate filter (CPF), used in conjunction with a diesel oxidation catalyst (DOC) is an important aftertreatment device used to meet Environmental Protection Agency (EPA) heavy-duty diesel emission standards for particulate matter (PM). Numerical modeling of these exhaust after-treatment devices decreases the time and cost of development involved. Modeling CPF active regeneration gives insight into the PM oxidation kinetics, which helps in reducing the regeneration fuel penalty. As seen from experimental data, active regeneration of the CPF results in a significant temperature increase into the CPF (up to 8°C/sec) which affects the oxidation rate of particulate matter (PM). PM oxidation during active regeneration was determined to be a function of filter PM loading, inlet temperature and inlet hydrocarbon concentration.

In this study, a DOC catalyst was experimentally studied in an engine test cell with a2010 Cummins 6.7L ISB diesel and a production aftertreatment system. The test matrix consisted of steady state, active regeneration with in-cylinder fuel dosing and transient conditions. Conversion efficiencies of total hydrocarbon (THC), CO, and NO were quantified under each condition. A previously developed high-fidelity DOC model capable of predicting both steady state and transient active regeneration gaseous emissions was calibrated to the experimental data. The model consists of a single 1D channel where mass and energy balance equations were solved for both surface and bulk gas regions. The steady-state data were used to identify the activation energies and pre-exponential factors for CO, NO and HC oxidation, while the steady-state active regeneration data were used to identify the inhibition factors. The transient data were used to simulate the thermal response of the DOC.

This paper focuses on the development of an Extended Kalman Filter for estimating internal species concentration and storage states of an SCR using NOX and NH₃ sensors. The motivation for this work was twofold. First, knowledge of internal states may be useful for onboard diagnostic strategy development. In particular, significant errors between the outlet NOX or NH₃ sensors, reconstructed from estimated states, and the measured NOX or NH₃ concentrations may aid OBD strategies that attempt to identify particular system failure modes. Second, the EKF described estimates not only stored ammonia but also NO, NO₂ and NH₃ gas concentrations within and exiting the SCR. Exploiting knowledge of the individual species concentrations, instead of lumping them together as NOX, can yield improved closed loop urea controller performance in terms of reduced urea consumption and better NOX conversion.

In this paper, we develop a method for optimizing urea dosing to minimize the downstream readings from a production NOx sensor that has cross-sensitivity to ammonia. This approach favors high NOx conversion and reduced ammonia slip. The motivation for this work is to define a process to identify the maximum selective catalytic reduction SCR performance bounds for a given drive cycle. The approach uses a model structure that has a closed-form optimal solution for the urea injection. Every aftertreatment system has its own, unique model, which must be identified and validated. To demonstrate the approach, a model is identified and validated using experimental SCR input/output NOx sensor data from a 2010 Cummins 6.7L ISB production engine. The optimal control law is then simulated and its performance compared against the simulated performance of the SCR using experimental data for its inlet conditions.

In order to further characterize and optimize the performance of Selective Catalytic Reduction (SCR) aftertreatment systems used on heavy-duty diesel engines, an accurately calibrated high-fidelity multi-step global kinetic SCR model and a reduced order estimator for on-board diagnostic (OBD) and control are desirable. In this study, a Cu-zeolite SCR catalyst from a 2010 Cummins ISB engine was experimentally studied in a flow reactor using carefully designed protocols. A 2-site SCR model describing mass transfer and the SCR chemical reaction mechanisms is described in the paper. The model was calibrated to the reactor test data sets collected under temperatures from 200 to 425 °C and SCR space velocities of 60000, 90000, and 120000 hr-1. The model parameters were calibrated using an optimization code to minimize the error between measured and simulated NO, NO₂, N₂O, and NH₃ gas concentration time histories.

Diesel Oxidation Catalysts (DOC) are used on heavy duty diesel engine applications and experience large internal temperature variations from 150 to 600°C. The DOC oxidizes the CO and HC in the exhaust to CO2 and H2O and oxidizes NO to NO2. The oxidation reactions are functions of its internal temperatures. Hence, accurate estimation of internal temperatures is important both for onboard diagnostic and aftertreatment closed loop control strategies. This paper focuses on the development of a reduced order model and an Extended Kalman Filter (EKF) state estimator for a DOC. The reduced order model simulation results are compared to experimental data. This is important since the reduced order model is used in the EKF estimator to predict the CO, NO, NO2 and HC concentrations in the DOC and at the outlet. The estimator was exercised using transient drive cycle engine data. The closed loop EKF improves the temperature estimate inside the DOC compared to the open loop estimator.