Probing Diverse Biomass Combustion Modes through Experiments with a Controlled Atmosphere Cone Calorimeter

Vilhelm B. Malmborg, Ioannis Sadiktsis, Axel Eriksson, Johannes Rex, Dan Madsen, Patrick van Hees, ANDREW GRIESHOP, Joakim Pagels, Lund University, Sweden

     Abstract Number: 323
     Working Group: Carbonaceous Aerosols

Abstract
The composition of biomass combustion emissions depends on both fuel properties and combustion conditions, but few studies have systematically controlled combustion type to explore its influence on emission yields and properties. A controlled atmosphere cone calorimeter is a device that varies multiple combustion parameters during heating and ignition of a fuel sample. Forty experiments were conducted with flat birch wood samples with varied heat flux (HF) and gas flow rate and composition (pure N2 or air) to generate pyrolysis, and under-ventilated (oxygen-starved) or well-ventilated combustion conditions. During tests, both online aerosol measurements and integrated samples (filters, TEM grids) were collected. Highly distinct particle yields and characteristics were demonstrated across combustion conditions and between test phases. Tests had similar characteristics during the initial pre-ignition ‘pyrolysis’ phase, but differed greatly during combustion, where the under-ventilated tests show much higher emissions of OA, eBC and especially PAHs (from SP-AMS analysis). Phase-specific OA yields and SP-AMS spectra differ greatly between phases, for example between N2 pyrolysis and ‘pyrolysis phase’ (pre-ignition) and combustion-phase emissions. Nontarget analysis via reversed phase LC-high-resolution MS was conducted on methanol extracts from representative test filters for 5 conditions (pyrolysis in N2 at 3 different heat flux levels and under-ventilated and well-ventilated combustion). The molecular composition from under-ventilated combustion diverges from those from other conditions, which appear more similar. However, the pyrolysis sample at the lowest HF (15 kW m-2) shows an enrichment in hydrophobic compounds (indicated by higher retention times) as well as lower contributions from low mass-to-charge features. We hypothesize that these compounds could be molecular signatures of larger breakdown products of lignin that further break down into smaller molecules at higher combustion/pyrolysis temperatures (higher HF) and may be linked with viscosity/morphology examined in future analyses.