Search Results

The engine selected for this work was a Caterpillar 3176 engine. Engine exhaust emissions, performance, and heat release rates were measured as functions of engine configuration, engine speed and load. Two engine configurations were used, a standard 1994 design and a 1994 configuration with EGR designed to achieve a NOx emissions level of 2.5 gm/hp-hr. Measurements were performed at 7 different steady-state, speed-load conditions on thirteen different test fuels. The fuel matrix was statistically designed to independently examine the effects of the targeted fuel properties. Cetane number was varied from 40 to 55, using both natural cetane number and cetane percent improver additives. Aromatic content ranged from 10 to 30 percent in two different forms, one in which the aromatics were predominantly mono-aromatic species and the other, where a significant fraction of the aromatics were either di- or tri-aromatics.

A zero-dimensional cycle simulation model coupled with a chemical equilibrium model and a two-zone combustion model has been extended to predict nitric oxide formation and emissions from spark-ignited, lean-burn natural gas engines. It is demonstrated that using the extended Zeldovich mechanism alone, the NOx emissions from an 8.1-liter, 6-cylinder, natural gas engine were significantly under predicted. However, by combining the predicted NOx formation from both the extended Zeldovich thermal NO and the Fenimore prompt NO mechanisms, the NOx emissions were predicted with fair accuracy over a range of engine powers and lean-burn equivalence ratios. The effect of injection timing on NOx emissions was under predicted. Humidity effects on NOx formation were slightly under predicted in another engine, a 6.8-liter, 6-cylinder, natural gas engine. Engine power was well predicted in both engines, which is a prerequisite to accurate NOx predictions.

Results from work focusing on the development of an ultra low emissions, high efficiency, natural gas-fueled heavy- duty engine are discussed in this paper. The engine under development was based on a John Deere 8.1L engine; this engine was significantly modified from its production configuration during the course of an engine optimization program funded by the National Renewable Energy Laboratory. Previous steady-state testing indicated that the modified engine would provide simultaneous reductions in nonmethane hydrocarbon emissions and fuel consumption while maintaining equivalent or lower NOx levels. Federal Test Procedure transient tests confirmed these expectations. Very low NOx emissions, averaging 1.0 g/bhp-hr over hot-start cycles, were attained; at these conditions, reductions in engine-out nonmethane hydro-carbons emissions (NMHC) were approximately 30 percent, and fuel consumption over the cycle was also reduced relative to the baseline.

An experimental program to develop a practical natural gas-fueled locomotive engine was conducted. Six natural gas-fueled combustion systems for an EMD 710-type locomotive engine were developed and tested. The six systems were evaluated in terms of NOx and CO emissions, thermal efficiency, knock tolerance, and other practical considerations. Each combustion system was tested at Notch 5, 100-percent load, Notch 8, 80-percent load, and Notch 8, 100-percent load conditions. In general, all of the technologies produced significantly lower NOx emissions than the baseline diesel engine. Based on the results of the tests and other analyses, a late cycle, high-injection pressure (LaCHIP) combustion system, using a diesel pilot-ignited, late cycle injection of natural gas with a Diesel-type combustion process, was determined to provide the most practical combustion system for a natural gas-fueled, EMD 710-powered locomotive.

The effect of humidity on the lean misfire limit and emissions from a lean burn, natural gas engine is described in this paper, along with a description of a practical humidity compensation method for incorporation into an electronic control system. Experiments to determine the effects of humidity on the lean limit and emissions are described. Humidity increases were shown to decrease the rate of combustion, reduce NOx emissions, and increase the levels of unburned hydrocarbon (HC) and carbon monoxide (CO) emissions. Data and calculations are also presented which demonstrate that increases in humidity will cause enleanment in a typical closed loop control system utilizing a universal exhaust gas oxygen (UEGO) sensor. A prototype system for humidity sensing and subsequent compensation based on these findings was implemented, and the system was found, through additional testing, to compensate for humidity very effectively.

An emissions reduction strategy was developed and demonstrated to significantly reduce cold-start hydrocarbon (HC) and CO emissions from a spark ignition (SI), gasoline-fueled engine. This strategy involved cold-starting the engine with an ultra-fuel rich calibration, while metering near-stoichiometric fractions of air directly into the exhaust ports. Using this approach, exhaust constituents spontaneously ignited at the exhaust ports and burned into the exhaust manifold and exhaust pipe leading to the catalytic converter. The resulting exotherm accelerated catalyst heating and significantly decreased light-off time following a cold-start on the FTP-75 with a Ford Escort equipped with a 1.9L engine. Mass emissions measurements acquired during the first 70 seconds of the FTP-75 revealed total-HC and CO reductions of 68 and 50 percent, respectively, when compared to baseline measurements.

A menu-driven, PC-based model, ALAMO_ENGINE, has been developed to predict the nitrogen oxides (NOx) reductions in direct-injected, diesel engines due to exhaust gas recirculation (EGR), emulsified fuels, manifold or in-cylinder water injection, fuel injection timing changes, humidity effects, and intake air temperature changes. The approach was to use a diesel engine cycle simulation with detailed gas composition calculations for the intake and exhaust gases (including EGR, water concentration, fuel-type effects, etc.), coupled with a code to calculate stoichiometric, adiabatic flame temperatures and expressions that correlate measured NOx emissions with the flame temperature. Execution times are less than 10 seconds on a 486-66 MHz PC.

Development of a heavy-duty natural gas engine to improve its lean operating characteristics is detailed in this paper. Testing to determine the lean misfire limit is described, as well as investigations into the cause of lean misfire in this engine. Details of engine modifications to improve the lean misfire limit are also included. The development process resulted in a significant improvement in the lean performance of the engine (i.e., an extended lean misfire limit, better combustion stability, and lower hydrocarbon emissions).

Heating neat (100 percent) methanol fuel (M100) is shown to improve dramatically the atomization of the fuel from a production, automotive, port fuel injector of pintle design. This improvement is particularly noticeable and important when compared with atomization at low fuel temperatures, corresponding to those conditions where cold-start is a significant problem with neat methanol-fueled (M100) vehicles. The improved atomization is demonstrated with photographs and laser-diffraction measurements of the drop-size distributions. Fuel temperatures were varied from -34°C (-29°F to 117°C (243°F), while the boiling point of methanol is 64.7°C (148.5°F). Air temperatures were ambient at about 24°C (75°F). For temperatures above the boiling point, some flash boiling and vaporization were presumably occurring, and these may have contributed to the atomization, but the trends for drop size did not shown any discontinuity near the boiling point.