Search Results

Emissions of nitrogen oxides (NOx) from heavy-duty diesel engines are subject to more stringent environmental legislation. Selective catalytic reduction (SCR) over metal ion-exchanged zeolites is in this connection an efficient method to reduce NOx. Understanding durability of the SCR catalyst is crucial for correct design of the aftertreatment system. In the present paper, thermal and chemical ageing of Fe-BEA as NH3-SCR catalyst is studied. Experimental results of hydrothermal ageing, and chemical ageing due to phosphorous and potassium exposure are presented. The catalyst is characterized by flow reactor experiments, nitrogen physisorption, DRIFTS, XRD, and XPS. Based on the experimental results, a multisite kinetic model is developed to describe the activity of the fresh Fe-BEA catalyst.

The emissions from a direct injection diesel engine measured according to the ECE R49 13-mode cycle and as a function of exhaust gas recirculation are compared for diesel fuel without water addition, and for water-in-diesel as emulsion and microemulsion. The effect of water addition on the soot emissions was remarkably strong for both the emulsion and microemulsion fuels. The average weighted soot emission values for the 13-mode cycle were 0.0024 and 0.0023 g/kWh for the two most interesting emulsion and microemulsion fuels tested, respectively; 5-fold lower than the US 2007 emission limit.

Sulphur poisoning and regeneration of NOx trap catalysts have been studied in synthetic exhausts and in an engine bench. Sulphur gradually poisoned the NOx storage sites in the axial direction of the NOx trap. During sulphur regenerations, hydrogen was found to be more efficient than carbon monoxide in removing the sulphur from the trap. The sulphur regeneration became more efficient the richer the environment (λ<1) and the higher the temperature (at least 600°C). H2S was found to be the main product during the sulphur regeneration. However, it was possible to reduce the H2S formation and instead produce more SO2 by running with lambda close to one or by pulsing lambda. Even if a relatively large amount of sulphur was removed from the NOx trap, these methods gave a much less efficient regeneration per sulphur atom removed than when running relatively rich constantly. Finally, a model that could explain this observation was proposed.