Transport-shifted multi-component non-ideal acid dew-point predictions and generalization/application of the mist-onset 'tangency condition' for ideal vapor mixtures near cool surfaces

摘要

Anticipating condensation conditions at solid surfaces or within thermal boundary layers is important in many industrial applications involving condensible vapor flows in contact with cooler solid surfaces. In the case of hydrocarbon fuel/air combustion products at moderate pressures containing acid precursor species (like NO2, SO3, HCl, etc.) surfaces that operate below the prevailing "acid dew point" (ADP) temperature (which can be significantly higher than the DP expected for H2O alone) become vulnerable to chemical attack. Surfaces cold enough to cause acid "mist" onset (AMO) in the vapor phase create potential environmental problems and become much less efficient for condensate capture. Previous methods for estimating ADP and AMO have several drawbacks, the most prominent of which are: limitation to one acid precursor species, dependence on a local curve lit to earlier thermodynamic ADP calculations, absence of systematic transport effects associated with the local temperature gradient, and "silence" about the resulting liquid add composition. Our present, more fundamental, approach overcomes each of these limitations and yet, will be seen to remain tractable from an engineering viewpoint. While we illustrate our methods for the case of a prototypical ternary (N=3) acid condensate (highly non ideal liquid mixture of H2O + HNO3 + H2SO4) produced from a high temperature ideal gas mixture containing H2O(g) and much smaller concentrations of the aqueous acid precursors: NO2, SO3 (or HNO3(g) and H2SO4(g)), our methods can be formally extended to N = 4,5,... if one has access to VLE data for each of the N(N - 1)/2 participating binary systems (if the resulting N-component liquid mixture does not itself phase-separate). Our present results also properly reduce horn the N-component case to the N-1 case, hence even to the singular case: N=1 (i.e., the DP- and MO-conditions for pure water in the absence of any acid precursor). Because some of the vapor species of interest (e.g., H2O, HNO3, and H2SO4) have molecular weights rather different from the mean molecular weight of the flue gases, Ludwig-Soret transport effects can become non-negligible (in the presence of the temperature gradients expected for high temperature combustion products) both T-W(ADP) and T-W(AMO) will often be seen to be "transport-shifted", although, as discussed in Appendices A, B, in many cases these effects are complicated by the thermal instability of the largest precursor molecules. Further extensions of likely future interest (e.g., effects of homogeneous chemistry within the thermal boundary layer where molecules like HNO3 and H2SO4 may have to be assembled from their lighter precursors: the effects of nucleation kinetic barriers and/or vapor phase non-ideality) are also identified and are the subject of our ongoing research. (C) 2015 Published by Elsevier Ltd.