Ammonia Injection Failure Diagnostic and Correction in Engine After-Treatment System by Nox and Nh3 Emissions Observation

To ensure that after-treatment systems (ATS) reduce emissions to the levels for which they were designed, it is essential that the ATS control can rely on the feedback signals from the sensors and actuators that are part of the system. Knowing that the amount of ammonia injected into the catalyst governs the Nitrogen oxides (NOx) reduction, this work addresses the impact of the ammonia injection failure in the Selective Catalytic Reduction (SCR) on the exhaust emissions and describes a model-based fault diagnosis strategy. The proposed approach is based on an artificial neural network (ANN) and a sensor signal analysis (SSA) model of the catalyst, as well as an observer to merge the models and accurately estimate the emissions. The proposed diagnostic strategy is based on the comparison of the observed NOx and ammonia (NH3) emissions of the actual system with those expected in the system without ammonia injection failure. Experimental results show that the proposed strategy can detect failures in ammonia injection above 10%. Once the degradation level is detected, a correction strategy is applied by increasing the ammonia injector opening time according to the estimated degradation to increase the injected ammonia up to levels similar to faultless conditions. When the injection failure was corrected, the proposed strategy was able to mitigate the impact on NOx emissions, reducing them by 23.33% and approaching the NOx levels without injection failure (5.35% increase)