Relationships Between Instantaneous and Measured Emissions in Heavy Duty Applications 2001-01-3536

Selective Catalytic Reduction (SCR), using urea injection, is being examined as a method for substantial reduction of oxides of nitrogen (NO_x) for diesel engines, but the urea injection rates must be controlled to match the NO_x production which may need to be predicted during open loop control. Unfortunately NO_x is usually measured in the laboratory using a full-scale dilution tunnel and chemiluminescent analyzer, which cause delay and diffusion (in time) of the true manifold NO_x concentration. Similarly, delay and diffusion of measurements of all emissions cause the task of creating instantaneous emissions models for vehicle simulations more difficult. Data were obtained to relate injections of carbon dioxide (CO₂) into a tunnel with analyzer measurements. The analyzer response was found to match a gamma distribution of the input pulse, so that the analyzer output could be modeled from the tunnel CO₂ input. The relationship between measured carbon dioxide and diffused power was established for a heavy-duty engine and was used to predict instantaneous emissions of carbon dioxide for various engine test cycles. The gamma function was then used to diffuse the implied carbon dioxide concentrations to yield a modeled analyzer output. The transient model and actual analyzer levels responded well. Similarly, the emissions measurements for oxides of nitrogen were modeled and agreed well. This research represents verification of the “forward transform” between actual “engine out” and measured emissions. Future research will examine the “reverse transform” and will consider issues of uniqueness of the solution.