SESSION 2

POLLUTANTS

ABSTRACT Reduction of nitric oxides in the exhaust of Diesel engines by urea SCR is a well known technique for stationary operating engines. Conversion rates of up to 99% are possible without slip of ammonia. However, the application of urea-SCR for vehicles like trucks or passenger cars fitted with Diesel-engines requires additional information about the dynamic behavior of different system components under various conditions of exhaust gas temperature space velocity and NO_x- concentration level.

This paper presents results of investigations sponsored by a research organization of the German industry, FVV (Forschungsvereinigung Verbrennugskraftmaschinen e.V., Frankfurt) The results indicate that NO_x conversion is mainly dependant on exhaust-gas temperature, feed a (it means the mole number of ammonia formed from urea addition related to the total mole number of nitric oxides) and finally the space velocity. Conversion rates of up to 90% have been obtained at stationary engine-conditions and, more than 70% at dynamic engine operation starting with cold engine.

Keywords: DeNO_x, selective catalytic reduction, SCR, urea, ammonia

ABSTRACT

The present study addresses validation of the explicit expressions for instantaneous NO formation rates to be used in a NO_x post-processor for CFD modeling. The combustion of lean, stoichiometric and rich hydrogen-air mixtures in well-stirred reactors has been modeled in the temperature range 1500 - 2200 K employing (1) detailed H/N/O reaction scheme, (2) the same H/N/O scheme but without the NNH pathway, and (3) H/O reaction scheme together with algebraic expressions for the NO production rate as a function of local gas composition and temperature. Thermal-NO, nitrous-oxide, and NNH mechanisms of nitrous oxide formation have been taken into account. It has been demonstrated that the calculations of the NO concentration employing full chemistry and explicit expressions for the instantaneous NO formation rates are in good qualitative and acceptable quantitative agreement for lean mixtures in the whole temperature range under consideration. Only for 2200 K at long residence times (> 20 ms) the difference between the full and the reduced mechanisms becomes important. For the rich mixtures the reduced mechanism overpredicts the NO levels for all temperatures.

The evaluation of the NNH path showed that this pathway is important at low residence times for high temperatures, except for the lean mixtures. At low (1500 K) and moderately high (1900 K) temperatures the NNH-path is important for all residence times, for the stoichiometric low temperature case this route of the NO formation is most pronounced.

ABSTRACT

Soot formation during near isothermal pyrolysis of some aromatic hydrocarbon mixtures diluted in helium for nominal temperature of 1473 K was investigated. An inhibition of soot particle formation takes place during pyrolysis of the model hydrocarbon mixtures. As a result, the mixture when pyrolyzed yields coarse soot compared with that produced during pyrolysis of the pure components.

During pyrolysis of hydrocarbon mixtures, which contain acetylene there, was demonstrated that only acetylene generates soot particle nuclei, whereas the other components are consumed by heterogeneous growth of the particles. This observation could be considered as an experimental evidence contradicting the agreed-upon mechanism of soot formation through a PAH generation stage under conditions of the flat flame combustion.

Key words: PAH, acetylene, pyrolysis, soot, soot particle nucleation, surface area, sooting tendency.

ABSTRACT

This paper describes the activities carried out in the domain of combustion modelling at ENEL Research Generation Branch. This activity concerns the development and the use of CRFD codes applied to the design of industrial furnaces. The objective is to predict the emission of pollutants, such as nitrogen oxides, in combustion flue gases. This prediction may be correctly accomplished using a detailed kinetic mechanism, which cannot be directly implemented on CRFD codes due to actual computational limits (both in terms of memory and CPU time consumption). A new approach has been developed: analysing the 3D CRFD flow fields, an "equivalent" chemical reactors network model is extracted with corresponding residence time distributions and overall reactor properties, and the detailed kinetic calculation is performed on this simpler scheme.

The approach has been successfully applied to different scale of furnaces such as pilot plants and industrial boilers, low- NOx burners and glass furnaces.

In the papers a description of the CRFD codes is given, and the methodology to extract a chemical ideal reactor network from CFRD fields is presented. Finally, an application of the procedure on the Monfalcone #3 steam generator is discussed.

ABSTRACT

This paper presents an application of the laminar flamelet approach to the calculation of a turbulent methane/air diffusion flame and formation of soot. The mean flow field is modeled by the Favre averaged Navier-Stokes equations and related transport equations, together with the k-e model. The soot number density and soot volume fraction are modeled by two transport equations with empirical expressions for the source terms. The source terms are functions of the flame temperature, density and mole fraction of C₂H₂. Different approaches for modeling the effect of fluctuations of the flame temperature, density and mole fraction of C₂H₂ on soot formation are investigated. The calculated mean mixture fraction and temperature are in fairly good agreement with the available experimental data. A sensitivity study of the results to the empirical model constant is performed.

ABSTRACT

A stochastic model of combustion and NOx formation in premixed turbulent methane flames is described. The model is based on the combination of Computational Fluid Dynamics and Monte Carlo methods for the solution of the joint scalar Probability Density Function. A GRI-derived reduced-chemistry model allows for the consideration of finite chemical rates.

The model is applied to the investigation of a lean, premixed, bluff body- stabilized flame, for which experimental data exist. In these conditions, the prediction of NO formation is very challenging, both because of the very low values (typically a few parts per million), and because all NO formation-routes are relevant. The results obtained show good accuracy for velocities, main species and temperature, and are very encouraging in respect of minor species, including NO.

ABSTRACT

The spectroscopic analysis in the uv-visible and infrared region of the electromagnetic spectrum was carried out on the high molecular weight species condensed and sampled in the soot inception region of a rich premixed ethylene flame.

The PAH content of these species, preliminary analysed by gas chromatography/mass spectrometry, was found to take into account for just a fraction of the total condensed species.

High pressure liquid chromatography allowed the separation of the condensed species in two large fractions: the aromatic fraction including substituted and unsubstituted PAH and the tar fraction including heavier components.

The absorptivities of the aromatic and tar fractions in the 215-400nm wavelength region was quantitatively evaluated and an increase of the absorptivity of both was found in correspondence of soot inception followed by a decrease. The higher visible absorptivity of the tar fraction with respect to the aromatic fraction suggest a higher molecular weight of aromatic species contained in the tar. The mixed aliphatic/aromatic character of the aromatic and tar species was shown by means of FT-IR analysis in the 2700-3400 cm^-1 where the aliphatic and aromatic C-H stretch can be detected.

The determination of hydrogen linked to aliphatic and aromatic carbon was performed by quantitative FT-IR analysis of the total CS. The hydrogen content of the condensed species changed along the flame reflecting the transformations that the condensed species undergo as soot is formed.

ABSTRACT

The detailed structure of counter-flow, ethylene diffusion flame has been studied under atmospheric conditions. The temperature and mole fraction profiles of major, minor, aromatic and polycyclic aromatic hydrocarbon (PAH) species have been determined under the strain rate of 57 s^-1. The sampling from the flame was achieved by using a heated quartz microprobe connected to a silica coated stainless steel sampling line which transferred the sample to an online Gas Chromatograph/Mass Spectrometer (GGMS). The temperature profile of the flame was obtained by using a silica coated Pt/Pt+13%Rh thermocouple. Visually, the flame showed 3 different color zones: Blue, yellow and orange. Blue zone was located in the oxidizer side of the flame. Luminous yellow and orange zones were located below the blue zone. The mole fraction profiles of major products peaked in the blue zone; where the maximum flame temperature of 1529 °C was also reached. This flame showed the same major characteristics of the lower strain rate ethylene flame studied before. However, the flame temperature was about 100 °C cooler in this flame as a result of larger strain rate. The most abundant hydrocarbon pyrolysis product was C₂H₂ with 3.5 % peak level located in the yellow zone. All the other pyrolysis products were also maximized within this zone. Benzene was measured as the most abundant aromatic species with 220 ppm maximum level, located in the yellow zone. Naphthalene was the most abundant PAH formed with 50 ppm peak level, followed by pyrene at 5 ppm. All the trace species were lower in this flame than the lower strain ethylene flame. Increasing strain rate results in decreasing residence times and cooler flame temperature thereby leading to the observation of lower peak levels of aromatics and PAH when compared to the lower strain rate ethylene flame.

ABSTRACT

The paper focuses on the very important problem of the humanity concerning the reduction of the released CO₂ amount from stationary fossil fuelled combustion facilities. The state of the art indicates that by improving the general efficiency of the classic power plants, especially that of the combustion facility, it might be possible to reduce the emission of the carbon dioxide, most responsible for the greenhouse effect.

Pre drying of the moisture Lignite is an alternative to this general purpose. First a theoretical background is developed. Thus, by eliminating the water vapours before the fuel is entering the combustion chamber, more stable conditions for the ignition and combustion of the fuel are realised, that is for the benefit of the combustion efficiency. By eliminating the vapours also the final cold end temperature of the exhaust flue gases might be lowered, thus improving the boiler efficiency, also reducing the acid contamination of the neighbourhood. If a suitable recovery of the heat enthalpy of the vapours is organised, and the gained energy is used, being thus a supplementary energy supply alternative, the general power plant efficiency is recovering. Thus the CO₂ emission is reduced, for the same used thermal power output of the plant.

The second part of the paper presents possible directions in order to pre-dry the Lignite. An inventory of drying systems is given, and, in extension, a proposal for a Romanian pre-dryer at lab scale is presented, as it will be errected for both demonstration and research purposes.

ABSTRACT

The effect of NO on the activity of a Cu/V/K/Cl based catalytic filter in the combustion of soot has been investigated. Temperature Programmed Oxidation (TPO) of soot, collected on the filter at the exhaust of a gas-oil burner, was performed in a flow microreactar either in the absence and in the presence of NO in the gas feed. Similar tests were also carried out with uncatalytic filter. For the catalytic filter the presence of NO resulted in a substantial increase of the rate of carbon combustion, while for the uncatalytic filter the effect of NO was negligible, but for what concerns the selectivity to CO₂. The influence of NO and O₂ concentration in the gas fed to the reactor on the overall activity of the catalyst has been also investigated.

Nitrogen oxides Temperature Programmed Desorption (TPD) showed that the simultaneous presence of NO and O₂ in the inlet gas stream gives rise to the formation of surface species whose desorption generates both NO and NO₂.

Altogether the results suggest that in the presence of NO the enhancement of soot catalytic oxidation is mainly due to carbon oxidation by NO₂ in turn formed by NO oxidation on the same catalyst. Therefore, the Cu/V/K/Cl catalyst is capable of activating the oxidation of both soot and NO.

Keywords: Soot, catalytic combustion, NO, NO₂

ABSTRACT

The quantitative characterization of sooting flames in terms of soot load and total soot surface area continues to be a major challenge in both fundamental investigations and practical applications. More particularly, a major interest can be found to discriminate the two classes of particulate, namely soot and cenosphere, presents in power generation plants fuelled with heavy fuel oils.

The Laser-Induced Incandescence (LII) technique is widely employed for the determination of soot volume fraction but LII may also be used to derive the particle diameter, profiting from the size-dependent cooling behavior of the Laser-heated particles. The temporally resolved LII emissions has been successfully used in a heavy fuel oil spray flame to distinguish the two classes of carbonaceous particles produced during the heavy fuel oil combustion. Moreover, it has been possible to follow along the flame axis the formation and burn-out of these two classes of particulate and their relative amount in terms of volume fraction.

ABSTRACT

In this work the attention was focused on the adsorption and desorption of elemental mercuric on HGR activated carbon produced by Calgon-Carbon Corp.. The study was performed in an apparatus at laboratory scale in which Hg⁰ vapors in a nitrogen gas stream, at a given temperature and mercury concentration, flowed through a fixed bed of adsorbent material. The experiments showed that the adsorption phenomena are faster than the desorption phenomena. SEM micrographs of the fresh carbon and of the carbon after adsorption or desorption had occurred have shown that mercury is adsorbed on the surface on particular sites where high sulfur concentration exists.

ABSTRACT

The catalytic reduction of nitric oxide with methane in lean conditions on cobalt-exchanged synthetic ferrierite has been investigated. Several samples were prepared by changing the cationic form of the parent zeolite, the time, and the temperature of ion-exchange.

Physico-chemical characterization evidenced the presence of different Co- containing species inside ferrierite, depending on the preparation conditions and the cobalt loading.

Catalytic test of NO reduction with CH₄ were carried out achieving SO% maximum NO conversion at 100% methane conversion. The presence of water in the feed gas resulted in a decrease of NO conversion. Better catalytic performance was exhibited by catalysts containing aggregated Co species with respect to isolated ones.

The contemporary occurrence of other reactions such as NO catalysed methane combustion together with NO reduction has been evidenced.

Preliminary kinetic investigation suggests that water competes with NO for adsorption on catalytic sites. Water effect is reversible for Co richest catalysts.