Jouni Pyykönen

Modelling alkali chloride superheater deposition and its implications

Abstract Alkali chlorides are often involved in superheater tube fouling and corrosion problems, especially during biofuel combustion. We have carried out a modelling study on fume particle and condensable alkali chloride vapour behaviour in superheater tube boundary layers. Two models of boundary layer aerosol behaviour are presented: one is a general computational fluid dynamics (CFD)-based model and the other one is a one-dimensional boundary layer theory-based model for the stagnation point. Simulations of NaCl behaviour in recovery boiler panel heat exchangers indicated that both (1) boundary layer condensation and subsequent particle thermophoretic deposition and (2) direct vapour deposition are significant deposition mechanisms for NaCl. In the first panel tube boundary layer, NaCl is clearly supersaturated, while further in the panel section boundary layer, it seems to be closer to the thermodynamic equilibrium. As the surface temperature of the ash deposit layer increases during a soot-blowing cycle from 480 to 670 °C, the overall deposition rate of NaCl remains fairly constant (increases only by 52%) due to boundary layer aerosol dynamics. Homogeneous nucleation within boundary layers may take place at certain conditions. Additionally, we have made estimates of the subsequent behaviour of deposited alkali chlorides. The rates of deposit sulphation and generation of HCl(g) do not seem to be sensitive to alkali chloride deposition rates provided that SO 2 levels are sufficiently low.

Modelling Ash Deposition during the Combustion of Low Grade Fuels

Deposit formation on external surfaces, such as heat exchangers, is a serious problem in many combustion processes. Deposits form from residue ash particles and from volatilized species that deposit either as tiny condensed particles or as vapour. With modelling it is possible to make predictions about the rate and the chemical composition of deposition resulting from a specified fuel fired in different process conditions. This paper discusses the modelling of different deposit formation mechanisms that are used in the ABC (Aerosol Behaviour in Combustion) code (Jokiniemi et al., 1994). When modelling deposition it is essential to know the ash particle size distribution and chemical composition and well as the composition of the flue gases. The emphasis in the ABC code and in this paper is in the deposition of species that have volatilized during combustion. The chemical composition of these species is often such that in deposits they make surfaces sticky for residue ash particles or their presence may lead to the sintering of the deposited layer or their presence can cause high temperature corrosion. The ABC model couples the description of deposition mechanisms with the description of ash particle formation from volatilized species and the respective gas phase chemistry. The amount of residue ash particles is given in input for the ABC code. Detailed descriptions of the deposition models applicable to the process conditions are given in this paper. As example cases we present model results of deposition in a kraft recovery boiler and in a medium speed diesel engine.

A Novel Porous Tube Reactor for Nanoparticle Synthesis with Simultaneous Gas-Phase Reaction and Dilution

A novel porous tube reactor that combines simultaneous reactions and continuous dilution in a single-stage gas-phase process was designed for nanoparticle synthesis. The design is based on the atmospheric pressure chemical vapor synthesis (APCVS) method. In comparison to the conventional hot wall chemical vapor synthesis reactor, the APCVS method offers an effective process for the synthesis of ultrafine metal particles with controlled oxidation. In this study, magnetic iron and maghemite were synthesized using iron pentacarbonyl as a precursor. Morphology, size, and magnetic properties of the synthesized nanoparticles were determined. The X-ray diffraction results show that the porous tube reactor produced nearly pure iron or maghemite nanoparticles with crystallite sizes of 24 and 29 nm, respectively. According to the scanning mobility particle sizer data, the geometric number mean diameter was 110 nm for iron and 150 nm for the maghemite agglomerates. The saturation magnetization value of iron was 150 emu/g and that of maghemite was 12 emu/g, measured with superconducting quantum interference device (SQUID) magnetometry. A computational fluid dynamics (CFD) simulation was used to model the temperature and flow fields and the decomposition of the precursor as well as the mixing of the precursor vapor and the reaction gas in the reactor. An in-house CFD model was used to predict the extent of nucleation, coagulation, sintering, and agglomeration of the iron nanoparticles. CFD simulations predicted a primary particle size of 36 nm and an agglomerate size of 134 nm for the iron nanoparticles, which agreed well with the experimental data. Copyright 2015 American Association for Aerosol Research