A two-stage NOx reduction system has been proposed to address the challenges associated with exhaust treatment from high efficiency lean burn natural gas engines. This concept is based on oxidation of NO to NO 2 followed by the subsequent reduction of NO 2 to N 2 with a hydrocarbon reducing agent. Titania supported cobalt catalysts prepared by both incipient wetness and modified sol-gel techniques are effective for the oxidation of NO to NO 2 , being capable of reaching equilibrium conversions at or below 300°C. Screening of several formulations of NO 2 reduction catalysts are presented, with silver supported on alumina being the most promising. Kinetic experiments are monitored by gas chromatography and chemiluminescent NO x analysis. Characterization of the catalysts is performed by BET surface area measurements, thermogravimetric analysis, and in-situ X-ray diffraction.

Recently, industry and government have joined together to develop high efficiency, low emissions, natural gas fueled, industrial, reciprocating engines for power generation. The California Energy Commission targets fuel-to-electricity efficiency at over 50% and NOx less than 0.01 gm/BHP-hr by the year 2010 [1]. The Department of Energy’s ARES program has targeted 50% efficiency and 0.1 gm/BHP-hr NOx by 2010 [2]. The engine manufacturers have determined that these goals cannot be met with current ignition system technology. They have jointly developed a specification for the next generation ignition system [3], which will support meeting the engine cost, efficiency and emissions goals. The Electrically Controlled Combustion Optimization System (ECCOS) is a new technology (patent pending) which is designed to meet or exceed this specification. This ignition system generates a high voltage, low current, radio frequency electrostatic field inside the combustion chamber to efficiently ionize the air and fuel mixture and initiate multiple flame fronts. The system is able to reliably introduce much higher ionization energy to the combustion chamber than conventional ignition systems because the ionization is done with a high voltage electric field, not high temperature. Conventional ignition systems generate up to 30,000 deg F of temperature in the spark plug gap. This temperature is created in the plug gap by a high current, low voltage plasma arc. The reliance of the conventional ignition system on temperature to initiate combustion limits the maximum energy that can be delivered because the high temperatures erode the electrodes. Since the ECCOS does not generate these high temperatures, electrode erosion should not be a problem. This paper presents a comparison of combustion characteristics between a conventional ignition system and the ECCOS igniting various mixture rations of propane and air in a constant volume combustion test chamber. Pressure rise rates as well as combustion photographs of the ignition and flame propagation processes are presented. In addition, experimental data obtained from the natural gas, single-cylinder engine operating with a conventional ignition system and the ECCOS are presented. Combustion rates, ignition delay, fuel consumption and emissions are presented at various air-fuel ratios.

Novel bifunctional catalysts combining two active phases, typically Cu-ZSM-5 and a modifier, were prepared and tested for the selective catalytic reduction of nitrogen oxides using propylene in order to overcome the hindering effects of water typically seen for single-phase catalysts such as Cu-ZSM-5. The catalysts were made by typical preparation techniques, but parameters could be varied to influence the catalyst. The physical characterization of the materials showed that the modification phase was added strictly to the external surface of the zeolite without hindering any internal surface area. Chemical characterization by temperature-programmed reactions, DRIFTS and x-ray absorption spectroscopy indicated strong interaction between the two phases, primarily producing materials that exhibited lower reduction temperatures. Two improvements in NOx reduction activity (1000 ppm NO, 1000 ppm C 3 H 6 , 2% O 2 , 30,000 hr −1 GHSV) were seen for these catalysts compared with Cu-ZSM-5: a lower temperature of maximum NOx conversion activity (as low at 250°C), and an enhancement of activity when water was present in the system. The use of a second phase provides a way to furthure tune the properties of the catalyst in order to achieve mechanistic conditions necessary to maximize NOx remediation.

Predictive feedback control is applied to achieve reduction in cyclic dispersion in an analytical, lean, spark-ignition model and a two-cylinder, four-stroke, natural gas Kohler Command 25 engine operating at lean conditions. Recent observations of the combustion dynamics are used to define a desired target point for control and to predict future combustion events which may stray from the target point. Fueling perturbations are applied to steer the system back toward the desired behavior. Overall control perturbations are constrained to maintain a constant average fuel-to-air ratio. We present two methods for obtaining the prediction of future combustion events. In the first method, the recent history of cycle heat release is used to construct an adaptive, low-order map which relates the current-cycle heat release to the next-cycle heat release. The second method uses symbolic analysis to determine the relative frequency of successive-cycle combustion events and predict the most probable successor to the current cycle. Results are presented which show a moderate reduction in cycle-to-cycle variation near the lean limit in both the model and the engine. Similarities in behavior have been shown to exist-ignition engines suggesting that a similar prediction strategy could be successfully applied to control cyclic dispersion in large-scale reciprocating engines.

An experimental apparatus and an analytical model have been developed to investigate and determine the lubrication condition and frictional losses at the interface between a piston ring and cylinder liner. The experimental apparatus features twin fiber optic displacement sensors to accurately measure the lubricant film thickness and a tri-axial piezoelectric force transducer to simultaneously measure frictional force. An analytical mixed lubrication model featuring Elrod cavitation and a stochastic/deterministic approach for asperity contact was used to investigate the effects of boundary and mixed lubrication conditions at the ends-of-stroke. A comparison between experimental and analytical results indicated that they are in good agreement. The results illustrate the transition through all of the different lubrication regimes (i.e. boundary, mixed and hydrodynamic lubrication) the piston ring and liner experience during a stroke. The twin displacement sensor arrangement is capable of producing accurate, repeatable measurements of lubricant film thickness that are in agreement with the analytical predictions. The analytical model developed for this study can capture the different lubrication regimes that the piston ring and liner experience.

J-type spark plugs composed of Ni-base alloy electrodes with a pure Ir tip in the center electrode and a Pt-W alloy tip in the ground electrode were examined as-manufactured and after use in natural gas reciprocating engines by spectroscopic and metallurgical techniques. The spectroscopic examination indicated Ni emission from the Ni alloy electrodes in new plugs, but a strong Ca signal in engine used plugs. This was confirmed by metallurgical examination, which showed the presence of Ca containing glassy oxide phase(s) (with the electrode alloy components) in the used spark plug electrodes. Intergranular cracking was observed on the Ir and Pt-W alloy electrode insert tips. The interface between the Pt-W insert and the Ni alloy ground electrode also became a site for extensive cracking and oxidation during service. These oxidation/corrosion and metallurgical issues may represent a significant component of the wear mechanism of these plugs in natural gas engines.

Premixed, lean burn combustion research has focused for years on extending the lean flammability limit while maintaining both stables ignition and turbulent flame propagation. Operating with a leaner air-fuel mixture results in a lower temperature conversion of reactants to products (i.e. reduced NOx) while maintaining thermal efficiency. The lean limit, at some level, is dependent on both the fuel transport and chemical properties. This work sets out to numerically explore the effect of reformed fuels on both fundamental flame stability and the performance/emissions tradeoffs of the engine. The numerical simulations were conducted for a range of reformed fuel blends (10–40%) as well as mixture equivalence ratios (0.35–0.6). The laminar flame speed results clearly define the regime of stable flame propagations for equivalence ratio/reformed fuel blend combinations. Subsequently, a validated and predictive quasi-dimensional engine simulation is used to simulate the performance/emissions characteristics of the complete engine system operating on the reformed fuel blends (10–50%) for a range of ignition timings, and air-fuel ratios. The performance trends define not only the misfire and detonation limits associated with the air-fuel blends but also the thermal efficiency/NOx tradeoffs.

Evermore demanding market and legislative pressures require stationary lean-burn natural gas engines to operate at higher efficiencies and reduced levels of emissions. Higher in-cylinder pressures and leaner air/fuel ratios are required in order to meet these demands. Contemporary ignition systems, more specifically spark plug performance and durability, suffer as a result of the increase in spark energy required to maintain suitable engine operation under these conditions. This paper presents a discussion of the need for an improved ignition source for advanced stationary natural gas engines and introduces laser spark ignition as a potential solution to that need. Recent laser spark ignition engine testing with natural gas fuel including NOx mapping is discussed. A prototype laser system in constructed and tested and the results are discussed and solutions provided for improving the laser system output pulse energy and pulse characteristics.

Friction from the power cylinder represents a significant contribution to the total mechanical losses in internal combustion engines. A reduction in piston ring friction would therefore result in higher efficiency, lower fuel consumption, and reduced emissions. In this study, models incorporating piston ring dynamics and piston secondary motion with elastic skirt deformation were applied to a Waukesha natural gas power generation engine to identify the main contributors to friction within the piston and ring pack system. Based on model predictions, specific areas for friction reduction were targeted and low-friction design strategies were devised. The most significant contributors to friction were identified as the top ring, the oil control ring, and the piston skirt. Model predictions indicated that the top ring friction could be reduced by implementing a skewed barrel profile design or an upward piston groove tilt design, and oil control ring friction could be reduced by decreasing ring tension. Piston design parameters such as skirt profile, piston-to-liner clearance, and piston surface characteristics were found to have significant potential for the reduction of piston skirt friction. Designs were also developed to mitigate any adverse effects that were predicted to occur as a result of implementation of the low-friction design strategies. Specifically, an increase in wear was predicted to occur with the upward piston groove tilt design, which was eliminated by the introduction of a positive static twist on the top ring. The increase in oil consumption resulting form the reduction in the oil control ring tension was mitigated by the introduction of a negative static twist on the second ring. Overall, the low-friction design strategies were predicted to have potential to reduce piston ring friction by 35% and piston friction by up to 50%. This would translate to an improvement in brake thermal efficiency of up to 2%, which would result in a significant improvement in fuel economy and a substantial reduction in emissions over the life of the engine.

This paper investigates a purely thermal control system for a 6-cylinder HCCI engine. Thermal energy from exhaust gas and from compression is used to condition the charge for the desired engine output. HCCI engine operation is analyzed with a detailed chemical kinetics based engine cycle simulation code. This cycle simulation code is linked to an optimizer that determines the operating conditions that result in maximum brake thermal efficiency, while meeting the restrictions of low NO x , and peak cylinder pressure. The results show the values of the operating conditions that yield optimum efficiency as a function of brake power for a constant engine speed (1800 rpm). It has been determined that a thermally controlled HCCI engine can successfully operate at high efficiency and low emissions over a wide range of conditions from idle to full load. The results show that a 42% brake thermal efficiency can be reached while the NOx emissions are kept under 2 parts per million. The analytical results shown here are expected to guide the ongoing experimental effort of converting a heavy-duty stationary engine to HCCI mode. The experimental work has the goal of meeting the very aggressive efficiency and emissions targets set by the California Energy Commission (CEC) Advanced Reciprocating Internal Combustion Engine (ARICE) Program.

Ignition of extremely lean mixtures is a very challenging problem, especially for the low speed, high load conditions of large-bore natural gas engines. This paper presents initial results from testing a high energy ignition system, the railplug, which can assure ignition of very lean mixtures by means of its high energy deposition and high velocity jet of the plasma. Comparisons of natural gas engine tests using both a spark plug and a railplug are presented and discussed in this paper. The preliminary engine test show that the lean stability limit (LSL) can be extended from an equivalence ratio, φ, of ∼0.63 using a spark plug down to 0.56 using a railplug. The tests show that the railplug is very promising ignition system for lean burn natural gas engines and potentially for other engines that operate with very dilute mixtures. The ignition characteristics of different railplug geometric and circuit designs are also discussed.

A non-fueled prechamber ignition system was developed to provide for controlled and stable combustion to best support the goals of both Waukesha Engine and the ARES program. This paper will provide details and results of efforts undertaken in optimization of the following design aspects: tangential angle of prechamber orifice channels in relation to head-induced cylinder swirl, prechamber orifice diameter, prechamber volume (spark plug recess in precombustion chamber), and recession of the entire precombustion chamber into the cylinder head. A number of important conclusions will be verified and presented based on the development work done on a laboratory engine.

We present theoretical models to simulate spark discharge and ignition. Two models are presented. The first model considers simplified fluid mechanics with chemistry effects. The second model utilizes more sophisticated flow with no chemistry. The simplified model incorporates physical models of breakdown and chemical kinetics of combustion with species and energy equations solved using a control volume method. A three-step mechanism is used to simulate chemical kinetics for fuel combustion and nitrogen chemistry. Dissociation and ionization in the plasma are included by assuming local thermal equilibrium. The second model simulates a spark discharge in air with more complete physics of flow using a high order numerical method, which shows the evolution of the shock wave and a torus-like plasma kernel.

This paper presents enhancements to a previously developed mixed-lubrication ring-pack model that has been used extensively in the automotive industry in predicting piston-ring/liner oil film thickness, friction and oil-transport processes along the liner. The previous model considers three lubrication regimes, shear thinning of the lubricant, and the unsteady wetting conditions of the rings at the leading and trailing edges. The model incorporates the effects of surface roughness by using Patir and Cheng’s average flow model and the Greenwood and Tripp statistical asperity contact model, assuming a Gaussian distribution of surface roughness. However, as a result of the methods used to machine a cylinder liner and the wear-in process observed in engines, the cylinder liner finish is highly non-Gaussian. The purpose of this current study is to understand the effects of additional surface parameters other than Gaussian root-mean-square surface roughness on piston ring-pack friction in the context of a natural gas reciprocating engine ring/liner interface. In general, the surface roughness of a cylinder liner is negatively skewed. Applying similar methodology published in the literature, a wide variety of non-Gaussian probability density functions were generated in terms of the skewness of the cylinder liner surface. These probability density functions were implemented into the Greenwood and Tripp asperity contact model, and subsequently into the traditional MIT ring-pack friction model. The effects of surface skewness on flow were approximated using Gaussian flow factors and a simple truncation method. The enhanced model was studied in conjunction with results from an existing ring-pack dynamic model that provided the dynamic twists of the rings relative to the liner and inter-ring pressures. In this manner, a detailed analysis of the effects of engineered cylinder liner finish on reducing friction losses was performed.

During a recent project at Southwest Research Institute ® (SwRI ® ), a medium-duty turbo-charged V-8 diesel engine was converted to natural gas operation for low-emission power generation applications. Initial cycle simulation studies indicated significant levels of cylinder-to-cylinder variation with the stock induction and exhaust system, which can limit the potential performance gains of a lean-burn spark-ignited engine due to spark knock and lean misfire. Cycle simulation techniques were used to investigate the sources of the cylinder-to-cylinder variation and determine potential solutions. The results indicated that significant improvements could be made with new intake and exhaust manifolds and a different turbocharger configuration.

This paper presents two piston ring and cylinder liner lubrication models and compares the friction predictions against the experimental results from a corresponding bench test. The first model aims to solve the average Reynolds equation with corrective flow factors, which describe the influence of surface irregularities on the lubricant flow under mixed lubrication condition. The second model takes account of the lubricant film rupture and cavitation. Meanwhile, a stochastic rough contact sub-model quantifies the relation between contact pressure and mean surface separation in both cases. Numerical results on the top compression ring simulation show that both models capture hydrodynamic, mixed, and boundary lubrication regimes, which depend on the real surface topographies of the piston ring and the cylinder liner. Whenever hydrodynamic action is insufficient to maintain the equilibrium position of the ring, the restoring force will be augmented by multi-asperity contacts lubricated by a thin boundary film. Total friction will originate mainly from shearing of viscous lubricant and shearing of asperity conjunctions. The purpose of this modeling effort is to compare both lubrication models to data from an experimental test-rig. This test rig eliminates many of the factors that can make analysis of predictions for real engine operating conditions difficult.

The emissions profile from a natural gas (NG) stationary power, stoichiometric, engine have been measured. Of particular interest are the unregulated emissions (especially formaldehyde) from this engine class. A relatively small 250 kW Caterpillar G3406 engine without a catalyst was used as the test platform with the emissions being characterized as a function of applied load and external conditions. In addition, multiple formaldehyde measurement techniques have been used and are compared. This paper presents the formaldehyde results, comparison of formaldehyde with various other parameters, a comparison of the different measurement techniques, and measurement of other toxic emissions from this engine.

A single cylinder spark ignited gasoline engine was modified to operate with natural gas. In such an engine, laser ignition was successfully demonstrated while transmitting the high-power laser pulses via solid core optical fibers. Subsequently, ignition studies were performed while using laser ignition (LI) and conventional spark ignition (SI). However, due to limitations imposed by the engine hardware the adverse conditions for ignition could not be simulated, i.e., of lean operation and high-pressures. As a result, the scope of the study was limited to comparing LI and SI ignition characteristics at various ignition timings. It was observed that both LI and SI resulted in reliable combustion over all ignition timings. Furthermore, LI resulted in higher rates of pressure rise and higher peak cylinder pressures. However, the higher NO x emissions resulting from such conditions might not be representative as the final performance of an engine as it is determined by optimizing ignition timing and other operating parameters.

NO x control in a lean burn natural gas engine is typically achieved with appropriate management of air/fuel ratio and ignition timing. A novel approach for further reduction involves the capture of NO x by first adsorbing the NO x from the exhaust stream, followed by the periodic desorption of NO x from an aftertreatment medium. Then, by passing the desorbed NO x gas into the intake air stream and back through the engine, a percentage of the NO x will be converted to harmless gases during the combustion process. The objective of this paper is to report the NO x conversion phenomenon during a lean combustion process. The results of this testing will be used to develop an optimal system for the conversion of NO x with a NO x adsorber. A 1993 Cummins L10-G spark ignited natural gas engine was used to conduct the experiments. Commercially available nitric oxide (NO, 98.8% purity) was injected into the engine intake to mimic the NO x stream from the desorption process to obtain NO conversion rates at various steady-state engine operating points. The NO injection system was capable of injecting NO at varying flow rates and time intervals. NO was injected into the intake manifold for ten and twenty second periods, and the conversion rates were calculated. When the injected NO amount increased from 0.22 g/s to 1.2 g/s and engine loads varied from 200ft-lb to 400ft-lb at 800 RPM, the NO conversion rates increased from 5% to 47%. It was observed that the air/fuel ratio, injected NO quantity and the engine load greatly effected the NO conversion rates. It was also noted that engine speed had a negligible affect when the intake NO concentration was held constant.

The use of Corona (Transient Plasma) discharge as an ignition source alternative to spark ignition (SI) for a Natural Gas fired engine presents an opportunity for increased efficiency and reduced NOx emissions. The multiple ignition sites created by a Corona discharge produce a faster and more complete burn than conventional SI. Bench tests show a faster combustion rise time (∼3×) and higher peak combustion pressures with corona discharge compared to SI. There are challenges encountered in retrofitting a corona discharge ignition system onto an IC engine. Electrode design and placement are critical in preventing arc formation and engine knock.