Final Report: Reduced NOx/Hydrocarbon Emissions via Oxygen Enriched Lean Burn Engines

EPA Contract Number: 68D98154
Title: Reduced NOx/Hydrocarbon Emissions via Oxygen Enriched Lean Burn Engines
Investigators: Nemser, Stuart
Small Business: Compact Membrane Systems Inc.
EPA Contact: Richards, April
Phase: II
Project Period: September 1, 1999 through September 1, 2000
Project Amount: $225,000
RFA: Small Business Innovation Research (SBIR) - Phase II (1999) Recipients Lists
Research Category: Air Quality and Air Toxics , SBIR - Air Pollution , Small Business Innovation Research (SBIR)


The Phase II project included 11 tasks, with the major focus placed on actual lean burn engine testing and membrane development. The major task was to supply 24 percent oxygen-enriched air (OEA) from perfluoro-2-2-dimethyl-1-2 dioxole-tetrafluoroethylene (PDD-TFE) high-flux membranes to a current design direct-injected (DI) lean burn spark ignition (SI) engine. Another task was to investigate the effect of adding OEA to extend the lean burn limit (the limit greater than 14.7:1 air-to-fuel [AFR] ratio, where combustion becomes unstable) beyond that when normal air is supplied. If the lean burn limit was successfully extended, Compact Membrane Systems, Inc. (CMS) would determine if NOx, hydrocarbon (HC), and CO emissions were reduced beyond the levels currently produced by lean burn engines operating with normal air. CMS also attempted to determine whether OEA would help to restore the power loss associated when operating with lean AFRs while simultaneously reducing fuel consumption.

Developing PDD-TFE membranes to withstand the harsh environments when installed under the hood of a typical gasoline powered automobile was also investigated during the Phase II project. The membranes were exposed to gasoline vapors and pressure cycling to determine the effect on membrane life and durability. Also, housings to protect the membrane were to be developed that would minimize both weight and cost.

Summary/Accomplishments (Outputs/Outcomes):

Gasoline Direct-Injected (GDI) Engine Test

The University of Wisconsin (UW) was contracted to perform engine testing on a current design lean burn DI engine. CMS and UW were successful in supplying the necessary OEA concentration of 24 percent and flow rate of 10 ft3/min at standard conditions to a Ford single-cylinder DI engine built to the Ricardo Hydra Platform. The lean burn limit could not be determined for the DI engine operating on normal air due to control problems experienced by UW. DI lean burn engines operate in an "unthrottled" mode to reduce pumping losses associated with partial throttle operation. The fuel is injected into the engine cylinder for combustion late (1201 crank angle before piston top dead center [BTDC]) in the cycle to try to achieve a stratified lean charge around the centrally located spark plug. The Ford engine used could not achieve stratified lean burn combustion when operating in an unthrottled mode. The engine did not incorporate "wall-guided" or "swirl-guided" air-fuel mixing used by current lean burn GDI engines. Many attempts were made to establish lean AFRs when operating on normal air, but higher than normal coefficient of variation of indicated mean effective pressure (COV of IMEP) values were experienced (less than 10%). Changes were made to the engine by replacing the cylinder head and modifying the fuel injection strategy to an early timing mode. The changes improved the COV of IMEP values, but reduced the AFR to a maximum of 12.5 from a previous value of 19, with unacceptable COV of IMEP results. To achieve an AFR greater than 14.7, the DI engine was switched to a partial-throttle early injection mode (270 BTDC, volumetric efficiency of 33%) to create a homogenous air fuel charge for better combustion. The homogenous air fuel charge mode operation provided acceptable COV values and allowed a lean AFR of 18.5 to be achieved on normal air and an AFR of 20 when operating on 24 percent OEA, thus showing that adding OEA will help to extend the lean burn limit. The torque also was increased slightly over the normal air case for the partial throttle mode as the AFR was extended beyond 14.7. However, it could not be shown that adding OEA during a stratified unthrottled DI mode would extend the lean burn limit, most likely due to control problems with the experimental engine.


Emissions of NOx, HC, CO, and CO2 were measured for the lean partial-throttle mode only. Partial throttle operation was the only mode of operation for which the addition of OEA provided a greater AFR over the normal air case. There was no decrease in NOx emissions with OEA over the normal air case at the lean limit. HC emission was reduced with the initial addition of OEA, but increased over the normal air case as the AFR became lean due to cylinder misfiring. CO emission increased slightly with OEA due to the extra oxygen available for formation, but could be controlled by the use of a catalyst. Emissions of CO2 were virtually unchanged.

Membrane Technology

CMS developed hollow fiber membrane modules in parallel while the engine testing was conducted at UW. CMS, along with its business partner, successfully developed a membrane substrate that can withstand the harsh environment experienced under the hood of a typical gasoline-powered automobile. A thick wall fiber was developed to withstand the pressure cycling from the turbocharger on heavy-duty on-highway diesel engines. The stronger fiber was produced in response to use the hollow fiber membrane developed by CMS to produce nitrogen-enriched air (NEA) exclusively for NOx reduction by one of the big four diesel engine manufacturers. The fiber was designed to withstand the extremely harsh environment experienced by heavy-duty on-highway diesel engines and to last a maximum if 10,000 operating hours.

CMS also exposed the membrane to gasoline vapors of 20-30 percent for volatile organic compound removal in underground storage tanks. The CMS membrane is used to remove the air from a gasoline vapor stream that is exposed directly to the membrane. There has not been any detectable decline in membrane performance due to gasoline vapor exposure. Because the CMS membrane in this Phase II project was designed for gasoline engines, the parallel development conducted by CMS and its business partner produced a membrane substrate to withstand the harsh environment experienced by gasoline engines.


At this time, it appears that the addition of 24 percent OEA does not help to extend the lean burn limit for DISI lean burn engines. Lean burn engines are still a relatively new engine technology, and these engines are not yet produced in the United States. The combustion process is very complex and requires careful fuel injection control. It is not surprising that UW investigators had problems controlling the Ford DI engine and could not develop a stratified lean burn mode. To date, Mitsubishi Motors is the only commercial manufacturer of GDI engines. CMS has developed a complete family of PDD-TFE hollow fiber membrane modules suitable for under-the-hood use for both gasoline and diesel engines that were designed to last for up to 10,000 hours of operation. CMS can provide OEA streams up to 25 percent and NEA streams up to 81 percent for emissions control on both gasoline and diesel engines. The hollow fiber membrane modules are commercially available thorough CMS in sizes of 2- and 6-inch diameter, ranging in membrane area from 1 ft2 to 105 ft2. CMS has licensed this and broader engine use to Praxair. The target is commercial introduction for NOx reduction in diesel engines in 2006/2007.

Supplemental Keywords:

small business, SBIR, air emissions, engineering, chemistry, EPA, lean burn engine, membrane, NOx, direct injection, gasoline, air-to-fuel ratio, spark ignition, combustion, hydrocarbon, oxygen-enriched air, RFA, Scientific Discipline, Toxics, Air, Sustainable Industry/Business, air toxics, cleaner production/pollution prevention, Environmental Chemistry, Sustainable Environment, Chemistry, HAPS, Technology for Sustainable Environment, New/Innovative technologies, tropospheric ozone, Engineering, Engineering, Chemistry, & Physics, Nitrogen Oxides, Nox, hydrocarbon, oxidation, oxidation of volitale organic hydrocarbons, nitrogren oxides (NOx), membrane technology , membranes, ambient air, carbon monoxide (CO), emissions, membrane modules, carbon monoxide, hydrocarbons, lean burn engine, carbon monoxide , nitrogen oxides (Nox), pollution prevention, oxygen enriched air, membrane technology, lean burn engines

SBIR Phase I:

Reduced NOx/Hydrocarbon Emissions via Oxygen Enriched Lean Burn Engines  | 1998 Progress Report