System Level Aftertreatment Modeling Applied to SCR Thermal Management

Shahidur  Rahman, Cummins Inc.

Diesel exhaust aftertreatment and engine systems have become increasingly complex to meet the progressively stringent emissions standards in the US and Europe.  It has become more and more challenging to design an optimized system to which meets the regulated tailpipe NOx emissions and minimizes the fluid (fuel + DEF, diesel exhaust fluid) economy.  Analysis tools can significantly increase the engineer’s capability to optimize the engine and aftertreatment systems while minimizing the use of experimental testing. To meet current near-zero level NOx emission, significant amount of NOx reduction during cold start is becoming a major challenge. One realistic option is to apply in-engine SCR thermal management (TM) strategy, which is typically a combination of rapid warm-up strategy at cold start that enables earlier urea injection, and a stay SCR warm strategy that prevents faster cool down. In this work, we talk about how in-engine SCR TM was optimized using Analysis Led Design (ALD) tools. The goal of this analysis was to test different engine management schemes to determine the effects on the SCR temperature and thus the NOx conversion efficiency.  This analysis was done for a Cummins ISB 6.7L US engine and thus the EPA’s regulatory FTP heavy-duty transient cycle was taken as a test cycle.  Aftertreatment plant model was created for a Cummins ISB 6.7L 2013 switchback aftertreatment system from component models to represent the actual hardware. The aftertreatment system model was built in AVL Boost software. The baseline model was validated against the test cell data for both thermal performance and NOx reductions.  Map-based engine model developed from the fuel map data was then integrated with the validated aftertreament model in order for the engine calibration to be optimized to achieve better NOx conversion efficiency.  Different engine calibration algorithms were tested on the engine side and the effects were observed for the aftertreatment model. From this analysis we found an optimized solution that gave the opportunity of further reducing NOx emission by about 20% more from the baseline case while minimizing the total fluid consumption.