IEA AMF Annex 51: Methane Emission Control

The mechanisms behind formation of unburned methane from natural gas engines were studied in this project. Furthermore, measures to address reduction of methane emissions were enlightened qualitatively and quantitatively in order to estimate the influence of different technical solutions on the effect of the emissions.

These solutions includes developments during the main combustion in the combustion chamber, and catalytic after treatment of the exhaust. The project has resulted in a detailed understanding of methane formation mechanisms. It is concluded, that there are a number of items of special importance. These include: misfire/bulk quenching, wall quenching, crevice volumes, post oxidation and valve timing/overlap.

It is concluded that particularly low-pressure dual fuel engines are associated with high values of methane emissions. These methane emissions most likely off-sets the advantage of applying natural gas instead of, particularly, diesel fueled engines with respect to greenhouse gas emissions. Since the unburned methane emissions origins from areas near the combustion chamber walls the sensible way to go now is towards direct injection of natural gas/bio-methane in order to reduce emissions.

The project included 7 different case studies. In case study 1 the knowledge about the formation mechanisms was implemented in a phenomenological mathematical model, TECMU (Thermodynamic Engine Cycle Modeling of Unburned Hydrocarbons), which was used to analyze the influence of the different mechanisms.

The model was used to simulate emissions from a 2014 model medium-speed 4 stroke dual fuel engine. The model results were in good agreement with experimental results, and demonstrated that all of the above mentioned mechanisms were important in relation to the resulting emissions of unburned methane.

The model showed, among other things, that a straight forward way to reduce the emission could be achieved by changing the valve timing. This would reduce unburned emissions by 12-62 %, depending on the engine load, a result that was in good agreement with practical measurements on a new engine generation, where the valve timing was changed accordingly.

Further evaluation of the mechanism understanding, was carried out in case study 2 by applying a special exhaust measurement technique in a single cylinder research engine. In these experiments a socalled FFID (Fast Flame Ionization Detection) was applied to distinguish between the contribution from the different mechanisms to the unburned hydrocarbon emissions.

The study focussed on investigation of the influence of different operating parameters on the unburned methane emissions from an engine operating at moderately lean air fuel mixtures. The influence of compression ratio, intake pressure and excess air ratio was studied. It was concluded that both the crevice mechanism and the quench layer mechanism was important for in the investigated operating ranges.

Post oxidation seems also to play an important role in contradiction to the medium speed engine investigated in Case Study 1, where post oxidation was negligible. This most likely is due to the fact that the two engines operate at different excess air ratios - the medium speed engine operating at much leaner conditions than the engine applied in Case Study 2.

In case study 3, the influence of mixing hydrogen to the fuel was investigated. Methane was applied as fuel with and without hydrogen admixture in a Euro-4 vehicle with stoichiometric operated, naturally aspirated, manifold injected 4 cylinder engine with an engine capacity of 2.0 l. The vehicle was equipped with an external fuel supply, an access to the electronic engine control (ECU), a modified catalytic converter as well as an internal cylinder pressure measurement in the combustion chamber of the engine.

The vehicle was driven with two different driving cycles (NEDC and WLTC) on a chassis dynamometer. Emissions were measured before and after the catalytic converter. The blending of hydrogen into the CNG is not readily feasible in terms of materials, as both the gas network and the gas-carrying components in the vehicle must be designed for this.

However, operation in conventional CNG vehicles would lead to significantly lower pollutant emissions than those already present in CNG operation. NOx emissions could be practically eliminated in the entire operating range, and T.HC emissions, which mainly consist of methane, which is primarily a strong greenhouse gas, could also be reduced by an average of one third.

However, the volumetric energy density, which decreases by almost 30% with an H2 content of 25 mole %, would be operationally disadvantageous. It could be compensated by increasing the pressure to 350 bar. For future combustion processes with diluted mixture formation (lean or EGR operation), H2 blending could be an interesting option due to the difficult ignition of (diluted) methane gases.

In case study 4, a series of Rh/zeolite catalysts design for oxidation of exhaust CH4 were tested. 1 wt.% Rh/zeolite catalyst had higher activity compared with the commercial catalyst under same operation condition. The activity of the Rh/zeolite catalyst can be significantly enhanced by elevating the operation temperature to 475 ℃ and limiting the SO2 concentration to a low level.

It seems promising to be used in the real engine exhausted gas condition where the SO2 concentration is 1-2 ppm. Regeneration by removing SO2 from the reaction gas can partly restore the catalyst activity, but a more efficient regeneration method is still being sought. Case study 5 investigated the Pd based catalyst performance.

Pd is believed to be the best converting precious metal for methane catalysts. In this study, some key factors were found, which led to enhance the activity and durability of current Pd-based CNG catalyst. Two critical contributions were from optimal support material and optimal characteristics of Pd.

Pd dispersion were achieved by selecting support with optimal surface property. Pd-Pt alloying and the use of electronic modifiers such as OSC and promoters were effective to make CNG catalyst more durable. SO2 and water are known to inhibit oxidation of methane in a catalytic converter. In case study 6, a regeneration method by hydrogen was studied.

With a catalyst aged to a conversion efficiency of 37%, it was possible to maintain this level, and even increase the efficiency after regeneration and ageing again applying regeneration gases containing 2,5% hydrogen. In case study 7, a number of vehicles were tested for tailpipe methane emissions as well as other methane emissions.

The amount of tested vehicles was 60. Twelve of the tested vehicles (15%) did not have any leakage. Thirty buses (50 %) did have leakage at fuel filling fittings corresponding to 2.9 ug/day on average. There were no leakage from the gas tanks and roof fittings and only 3 % of the vehicles did have a diffuse, not quantified, increased level in the engine room.

The major source of methane was inside the tailpipe corresponding to 0.88 mg /day and bus. The result indicates the major contribution of methane originates from slip during driving.