D. Thompson

NOx formation in combustion

Nitric oxide formation in the flame and post-flame gases of a lean methane-air flame (7.3–9.2% methane) has been studied in a 2.54 cm i.d. tubular combustor with a multiport flameholder. The combustor was naturally oscillatory (frequency 500–1250 Hz; amplitude 0.07 kg cm−2). Nitric oxide formation has been found to be substantially unchanged by the introduction of combustion driven oscillations when relatively low amplitude temperature oscillations were associated with the oscillatory combustion. It is shown that superequilibrium [O] concentrations may be calculated from the ratio [H2]/[H2]eq in the fuel lean system. The use of the modified chain rate equation d [ NO ] d t = 4.31 × 10 13 exp ⁡ ( − 134.7 R T ) [ O 2 ] 1 / 2 [ N 2 ] [ H 2 ] [ H 2 ] e q mole liter − 1 sec ⁡ − 1 , (E = 134.7 kcal/mole) for the prediction of nitric oxide formation in the post flame gases of fuel lean hydrocarbon flames above 1800°K gave satisfactory agreement with experimental results.

Studies into the formation of dioxins in the sintering process used in the iron and steel industry. 1. Characterisation of isomer profiles in particulate and gaseous emissions

Dioxin analyses using a DB5MS chromatography column are presented for samples of dust from an electrostatic precipitator and emissions from the main stack of a UK sinter plant. The analyses have been repeated using the SP2331 column, for which the elution order has been determined for the whole range of tetra to octachlorodibenzo-p-dioxins and dibenzofurans, and which provides a larger degree of separation than the DB5MS column. Polychlorinated dibenzofurans are present in significantly larger amounts in both the dust and stack emissions, and are the main contributors to the I-TEQ. Analysis using the SP2331 column for dioxins show considerable similarity to those of Bacher et al. for soot from a wood-burning fire, but differ markedly from those of Addink et al. for simulation of waste incineration. The similarities to Bachers results are less for the dibenzofurans. Comparison of absolute concentrations with reported vapour pressures shows the retention of significant dioxin in the dust to involve mechanisms other than pure condensation.

Enthalpies of formation and entropies of chlorinated dibenzo-p-dioxins and dibenzofurans; selected data for computer-based studies

Abstract Values for calculating the enthalpies of formation and standard entropies of chlorinated dibenzo- p -dioxins and dibenzofurans are obtained. The enthalpy of formation for dibenzo- p -dioxin derived from values for related compounds is −55 kJ mol −1 . The value of Δ f H for dibenzofuran obtained by Chirico et al., 55.2 kJ mol −1 , is recommended. Basic reductions in the enthalpy of formation of 29.5 kJ mol −1 for each chlorine substitution, with positive corrections of 7.5 kJ mol −1 for each ortho- and 2.5 kJ mol −1 for each para-substitution, are derived from the enthaplies of formation of chlorinated benzenes. It is noted that a recent evaluation of the enthalpies of formation and sublimation of the mono- and dichlorophenols indicates that the enthalpies of formation of these and possibly other chlorinated aromatic compounds vary more regularly with degree of chlorination than previously thought, and the model of chlorination based on phenols due to Shaub is thus, in general, unsatisfactory. A previous model of dibenzo- p -dioxin and dibenzofuran chlorination based on benzoquinone chlorination due to the author may underestimate the reduction in enthalpy of formation for the higher degrees of chlorination. The values 388 J mol −1 K −1 (Shaub) and 374.4 J mol −1 K −1 (Chirico et al.) for the standard entropies of dibenzo- p -dioxin and dibenzofuran are recommended, with each chlorination increasing the entropy by 22.6 J mol −1 K −1 .

An evaluation of the heat of formation of chlorinated dioxins and its application to isomer abundance predictions

Abstract Heats of formation for dibenzo-p-dioxin and all chlorinated dioxins are derived and combined with standard entropies and heat capacities to obtain a computational database for prediction for these compounds. Databases based on a published preditive method for dioxins, and on general group additivity data have also been constructed. The predicted relative abundances of the isomers for all degrees of chlorination from the three databases are compared. The abundances of the tetra- to heptachlorodioxin isomers are compared with those predicted by Unsworth using MOPAC, and with experimentally determined concentrations from combustion systems.

Thermodynamic considerations in dibenzodioxin and dibenzofuran formation: Concentrations of chlorinated dioxins and furans in model fuel-rich combustion gases

Abstract A database for chlorinated dioxins (DBDs), with heats of formation derived by analogy with chlorination of quinone and benzene, has already been shown to predict relative isomer distributions nearer to those observed, and those predicted using MOPAC, than an earlier database due to Shaub. The model has been extended to obtain a similar database for chlorinated dibenzofurans (DBFs), and comparisons with observed isomer concentration distributions are illustrated. Calculations are carried out using the NPL MTDATA suite of programs. Isomer group values for the properties of the original compounds and each degree of chlorination have been derived, and the use of these values leads to a reduction in the data obtained in computations to more manageable proportions. The reduced database obtained is combined with other databases to carry out computer predictions for simulated chlorine-containing combustion gases. Data for inorganics substances and the simplest organic substances (methane. ethane and related species) are obtained from the SGTE database. More extensive data for aliphatic and aromatic hydrocarbons from various standard sources have been incorporated in an additional computational database. In order to examine the relative concentrations of DBFs and DBDs a database for chlorinated hydrocarbons has been established. Predictions for simulated fuel rich combustion products for the C, H, O, Cl system are presented.

Process-related patterns in dioxin emissions: a simplified assessment procedure applied to coke combustion in sinter plant

Analyses for dioxins present in the windlegs of sinter plant using coke breeze as fuel which were carried out originally to monitor the 17 targeted isomers have been re-examined in order to establish the variation in isomer profiles with location of the sampling point relative to the beginning of the sinter strand. The analysis has been carried out using peak height as a measure of isomer abundance to allow assessment of a large number of peaks reasonably rapidly. It is found that the isomer profiles of the tetra- to heptachlorodibenzofurans, which dominate sinter plant emissions in the exhaust gases from the majority of the bed are similar. However, analysis shows that whilst some isomers contribute a similar percentage of the isomer group at the beginning of the strand, there are more, which vary significantly from the mean. Ways in which this localised difference in isomer distribution could arise are discussed.

PREDICTION OF THE PHASES PRESENT IN FLY ASH, THEIR COMPOSITION AND THE INFLUENCE OF THESE FACTORS ON ITS UTILITY AND DISPOSAL.

Two computational packages, F*A*C*T and MTDATA have been used in predictions of the mineralogy of ash and the distribution of elements between ash and emissions. Kinetic constraints are considered. Experience in monitoring fly ash in two field sites where the fly ash is 17 and 40 years old is described together with the results of batch and column leaching tests on fresh ash that demonstrate rapid release of elements which are mainly surface associated. Porewater extracted from borehole samples demonstrate significant leaching of elements thought to be more associated with the glassy phase. The use of fly ash in blended cements is reviewed noting the enhanced properties attainable compared with conventional mixes. These results are discussed using the computer program CEM-CHEM to predict phases formed in mixtures of Portland cement and ash. Combustion, emissions, mineralogy, fly ash, leaching, blended cements.

Formation of NO in a methane-air flame

Factors affecting NO formation in fuel-lean methane combustion, in the temperature range 1800°–2000°K, have been theoretically and experimentally investigated. Application of rate data indicates that post-flame nitric oxide formation in this range, in cases where equilibrium oxygen-atom concentrations may be assumed, will be dominated by the Zeldovich mechanism (rate-determining step N 2 + O ⇌ N O + N ), with a much smaller contribution by the bimolecular mechanism (rds, N 2 O + O ⇌ 2 N O ). Theoretical assessment of the effects of nonequilibrium [O], under conditions of balance of the hydrogen-oxidation reactions, shows that the nonlinear rate relationship of the latter mechanism with [O]/[O] eq (=[H 2 ]/[H 2 ] eq ) is not sufficiently marked for it to become significant by comparison with the (linearly dependent) chain mechanism. Experimental investigations were carried out in a 1-in.-diam tubular oscillating combustor (500–1250 Hz) with a multiport flameholder, burning methane-air mixtures (7.3%–9.2% CH 4 ). Mixing in the post-flame region was low. Gas samples were withdrawn through quartz microprobes and analyzed for nitrogen oxides by a modification of the Saltzman technique, and by mass spectroscopy for other species. Temperatures were measured by bare-wire thermocouple. NO concentrations up to 25 ppm were observed in the flame region, rising by up to 18 ppm in the first 2 in. (approx. 2 msec residence time) in the post-flame gases. No significant changes in NO formation, due to oscillations, were observed. The mean post-flame rise in NO concentration was an order-of-magnitude greater than predicted by the Zeldovich mechanism with equilibration of [O] and [O 2 ] assumed, but results were in satisfactory agreement with predictions from the modified rate equation d [ N O ] / d t = 1 . 5 0 3 × 1 0 1 7 T - 1 / 2 [ O 2 ] 1 / 2 [ N 2 ] r exp ( - 1 3 4 . 7 / R T ) ppm se c - 1 , , where [O 2 ] and [N 2 ] are expressed as percentages, and r =[O]/[O] eq , derived as [H 2 ]/[H 2 ] eq (always more than 2.4 in the present work).

Comparison of observed and predicted equilibrium PAH concentrations in coke oven emissions: I. Relative abundances of carbon-hydrogen PAH.

The relationship between the equilibrium concentrations of the carbon-hydrogen species in polycyclic aromatic hydrocarbons (PAH) found in coke oven emissions by Kirton et al and those which they measured has been investigated. A thermodynamic database for the species involved for use with the MTDATA computational package has been constructed and used to predict concentrations for local equilibrium between the group of PAH species at varied temperature. The temperature at which the concentration of each species observed would be in equilibrium with representative methane and hydrogen concentrations in coke oven gases has also been calculated. In each case the most satisfactory agreement between equilibrium and observed concentrations occurs above 1100K. In the group calculation, the higher PAH are over-predicted and this is consistent with the known kinetic restriction on soot formation. Some substituted benzenes and the biphenyl series are seriously under-predicted within the PAH group, and in the calculations for the model coke oven gas, do not equilibrate with methane and hydrogen at temperatures representative of coke ovens: this is consistent with their formation by reaction during quenching. The predicted equilibrium concentrations of the majority of species are within an order of magnitude of group equilibrium concentration at 1150K.

Removal of vapour phase PCDD/Fs in electric arc furnace steelmaking emissions by sorption using plastics

Plastics are potentially suitable for the removal of vapour phase PCDD/Fs in emissions from the electric arc furnace (EAF) steelmaking process. Three different commercial plastics, i.e. polypropylene BE170MO (Borealis A/S, Denmark), polypropylene in the form of 5 mm spheres (The Precision Plastic Ball Co. Ltd., UK) and polyethylene LD605BA (ExxonMobil Chemical, Belgium), have been studied using a novel experimental apparatus for the removal of vapour phase PCDD/Fs. Polypropylene BE170MO was identified to be the most suitable product amongst the three plastics in terms of PCDD/F sorption and potential industrial application. The optimum temperature for PCDD/F sorption on polypropylene BE170MO was below 90 degrees C for a removal efficiency of >99% at an average vapour phase PCDD/F concentration of 3.5 ng I-TEQ/Nm(3). At 130 degrees C, 53% of the PCDD/Fs trapped on polypropylene BE170MO were desorbed.