Particulate and gaseous emissions from manually and automatically fired small scale combustion systems

Abstract

In an extensive wood combustion source test series 2 automatically and 2 manually fired appliances, and 8 fuel types were investigated with respect to their particulate (PM10) and gaseous emissions. Softwood briquettes, beech, oak and spruce logs, wood pellets as well as further biogenic fuels: wood chips, miscanthus (elephant grass) pellets and triticale (“energy crop”) pellets were tested. Gaseous emissions were measured continuously while PM10 was sampled with a dilution system and averaged over standard test cycles. Manually fired stoves exhibited highly variable emissions resulting in an uncertainty of 30% for most measured compounds, determined in a series of replicate experiments. Average PM10 emissions from manually fired appliances were around 130 mg m −3 (standard conditions for temperature and pressure (STP), 13%O 2 , dry gas), equivalent to 90 mg MJ −1 . Wood pellets and chips combustion under full load operation with automatically fired appliances emit almost one order of magnitude less PM10, respectively: 12–21 mg m −3 (STP, 13%O 2 , dry gas), or 8–14 mg MJ −1 . Around 30% of total particle mass from manually fired systems account for elemental carbon and 30–40% for organic carbon, resulting in carbonaceous fraction content of around 90%. On average around 5% of PM10 emitted by manually fired stoves consisted of levoglucosan while this anhydrous sugar was below detection limit in full- and part load operation of automatically fired systems. Generally, emissions from automated systems were relatively constant for the same fuel type predominantly consisting of inorganic constituents. Emissions are mainly influenced by the mode of operation, start-up, full load or part load for a given fuel type. Surprisingly high emissions were observed for triticale pellets: 184 mg m −3 (125 mg MJ −1 ,) PM10 and 466 mg m −3 (395 mg MJ −1 ) NO x, (under full load operation, STP, 13%O 2 , dry gas), originating from high chlorine and nitrogen contents of the fuel. ► Emissions were tested for combustion of briquettes, logs and pellets in modern stoves. ► Automatically fired systems emit an order of magnitude less PM10, mainly inorganics. ► Levoglucosan accounts for 5% of PM10 emitted by manually fired stoves. ► Levoglucosan was mostly below detection limit in automatically fired systems. ► Levoglucosan/mannosan ratios were of 2–4 for softwood and >10 for hardwood.