(180b) A discussion on the fundamental aspects of Urea SCR control for transient applications

The last decade has seen the emergence of Urea based Selective Catalytic Reduction (SCR) for NO_x abatement as the after-treatment system of choice in automotive Diesel applications.   This has led to an increased interest amongst researchers, both in industry and academia, in developing an understanding of catalyst behavior and performance constraints for the design of after-treatment systems.  Substantial progress has been made over the years in establishing the basic chemistry that defines the SCR of NO_x by NH₃.  There is a large body of work detailing the kinetics of the NH₃ based SCR reactions relevant to a diverse set of catalysts that may be used for the de-NO_x process. Unfortunately the same degree of unified progress cannot be claimed for the control design of urea SCR systems for fast transient, automotive type applications.  Although the number of publications on control design of urea SCR applications continues to grow, with increasing contributions from academia, the applicability of many of the proposed solutions remains questionable.  This is in part due to limited understanding of the plant dynamics in the context of real time control design as well as On Board Diagnostics (OBD).  We stress the need for transient control design to distinguish the automotive control problem from that relevant to stationary applications such as power-plants that favor Steady State (SS) or pseudo-SS operation. The constraints imposed by the SCR plant dynamics naturally favor such (SS) systems and the control problem may be reduced to an open loop process control design.  In this paper we discuss the challenges associated with the robust control design of automotive urea SCR applications with considerations on the hardware configuration variants that impose additional constraints via catalyst sizing, adequate state sensing, as well as thermal management. We also discuss additional challenges imposed by using internal models that may drift due to input uncertainties and/or catalyst deterioration.  On the OBD front we will discuss the challenges associated with deterministically aging such catalyst and then robustly diagnosing them over designated emissions cycles as well as off cycle scenarios. Last but not least we will also touch upon the sensor set typically used and the associated sensing challenges from sensor cross sensitivity and cost. Overall we aim to provide a solid framework for discussion that adequately represents the usage of such catalytic systems from an OEM’s point of view.