Förbränningskarakterisering av rapsmjöl och förslag till optimalt nyttjande i olika förbränningsanläggningar

• When rape oil is chemically extracted, rape seed meal, a solid residue remains. Currently, it is used as animal feed. Several plants for the production of rape methyl ester (RME, biodiesel) are in operation or under construction. Combustion properties have been studied for rape seed meal produced as a by product to rape-methyl esther (RME, biodiesel). Composition of the material has been measured, using proximate and ultimate analysis. The lower heating value was 18.2 ± 0,3 MJ/kg d.w. and the ash content was 7-8 percent d.w. The material is rich in nitrogen and sulphur. Concentrations of K, P, Ca and Mg are high in the fuel. Rape seed meal was mixed with bark and pelletised. Bark pellets were also used as a reference fuel. Pellets with 10 and 30 percent rape seed meal were produced. Material with 80 percent rape seed meal and 20 percent planer shavings was also pelletised. Wood had to be added to provide enough friction in the pelletising process, with adapted equipment rape seed meal could probably be easily pelletised). The material was studied using Thermo-Gravimetric Analysis (TGA), and compared with data from tests with wood powder. The pyrolysis of the rape seed meal has a characteristic temperature of 320oC. Devolatilisation starts at 150 oC (at a lower temperature than for wood powder), and proceeds within a rather wide temperature range. The probable cause is the difference in organic content, in particular protein content. The result does not suggest that the material will be difficult to ignite. Experiments in a bench-scale fluidised bed (5 kW) showed that pellets containing only bark, and the mixture rape seed meal/wood had a bed agglomeration temperature well over the normal operational bed temperature. For the fuel mixtures rape seed meal and bark, the agglomeration temperature was slightly over the operational temperature. Particle emissions from fluidised bed combustion and grate combustion were, the latter simulated using a commercial pellet burner, were roughly doubled with fuels containing rape seed meal compared to bark. In the powder burner tests, particle emissions increased with a factor 17 with rape seed meal compared to wood powder. The emitted particles were mainly found in the fine (< 1 µm) mode during grate and powder combustion. During fluidized bed combustion the total particulate matter consisted both of a coarse (>1 µm) and a fine mode fraction. The particles from grate combustion of bark contain mostly K, S, Na and Cl apart from oxygen and carbon. When rape seed meal is present, Cl and Na concentrations decrease considerably and the main contents of the particles are K and S (and O and C). The results from the X-ray Diffraction Spectroscopy (XRD) analyses showed the presence of crystalline K2SO4 och KCl. The fine particles (<1 µm) from powder combustion contain mainly K, P and S. The only identified crystalline phase was K2SO4, suggesting that most phosphorus was in the amorphous phase, i.d. most probably molten. The deposit formation on a cooled probe was studied during the fluidized bed and powder combustion experiments. The fine particles deposited during fluidised bed combustion contained K, Cl and S. When bark was combusted in the fluidised bed, the coarse fraction contained Ca and Si, when rape seed meal in different mixes was combusted this changed to P, K, Ca and Mg. The deposits formed during combustion of rape seed meal in the powder burner were mainly made up of phosphates (Ca-, Mg/K-, Ca/Mg-phosphates) and MgO. Sintered material (slag) from grate combustion of bark contained mainly Si, Ca, K and Al, probably as silicates. Adding rape seed meal tended to increase P, Ca and Mg while Si and Ca content tended to decrease. Through XRD a number o crystalline phases in the sintered material and the rest of the bottom ashes could be identified. NO emissions from the combustions tests increased two to four times with rape seed meal compared to typical wood fuels. For the fluidised bed test, SO2 concentrations were rather high for the rape seed meal pellets (with 20 percent wood), still only about 20 percent of the sulphur in the fuel formed SO2. For the grate combustion and powder burner combustion, 60 percent and 70 percent of the sulphur respectively formed SO2. HCl emissions were low for all tests. The rather high emissions of NOx and SOx mean that the material should be used in large-scale facilities with external SOx and NOx cleaning. In smaller facilities, the material may be used in small amounts mixed with other fuels. The risk of slagging is not very high, and should not rule out grate combustion of pellets with rape seed meal mixed with other fuels. The risk of corrosion of superheater surfaces during combustion is probably low since the smaller-size particles formed at fluidised bed combustion and grate combustion contain K2SO4. However, a large fraction of the particles formed in powder burner combustion probably contains low temperature melting K2PO4, making the risk for deposit formation significant. Rape seed meal for powder burner applications should be used with care. The content of phosphorus in the material may be an advantage when mixes of rape seed meal and other fuels are considered. The high affinity between potassium and phosphorus means that more sulphur in the fuel will be available for sulphatising of any KC. (formed from combustion of many forest and agricultural fuels). Use of rape seed meal as a sulphur containing additive could thus be an option. For grate combustion and fluidised bed combustion, addition of rape seed meal may reduce the risk of slagging and bed agglomeration, respectively. Full scale tests in fluidised beds or grate combustors with problematic biofuels (containing Cl and K) would be useful to test whether ash-reduced operational problems could be reduced through the addition of rape seed meal.

• Effektiva värmevärdet hos rapsmjöl är 18,2 +/- 0,3 MJ/kg TS och askhalten varierar mellan 7-8 %. Bränslet har högt N- och S-innehåll, och bränslet är rikt på K, P, Ca och Mg. Partikelemissionerna från fluidbäddseldning och rosterförbränning med rapsmjölsblandningar var ungefär dubbelt så höga som från barkförbränning. Vid pulverförbränning av rapsmjöl ökade partikelemissionerna med en faktor 17 jämfört med träpulver. De höga NOx och SOx emissionerna från rapsmjölsförbränning innebär att materialet bör utnyttjas i storskaliga anläggningar med extern svavel.- eller NO-rening, eller i relativt låga inblandningsgrader i andra bränslen. Mindre anläggningar med enbart cyklonrening är olämpliga på grund av de höga stofthalterna. Rapsmjöl torde vara ett intressant sameldningsbränsle vid roster- och fluidbäddseldning med klor- och kaliumrika skogs- och åkerbränslen då rapsmjöl eventuellt skulle kunna användas som svaveladditiv för reduktion av klorinducerad korrosion på t ex överhettare. Slaggningstendensen torde minska vid inblandning av rapsmjöl i båda dessa bränslekategorier och bäddagglomereringstendensen torde minska vid inblandning av rapsmjöl i åkerbränslen.

Ämnesord

TEKNIK OCH TEKNOLOGIER  -- Maskinteknik -- Energiteknik (hsv//swe)

ENGINEERING AND TECHNOLOGY  -- Mechanical Engineering -- Energy Engineering (hsv//eng)

Nyckelord

Energy Engineering

Energiteknik

Publikations- och innehållstyp

vet (ämneskategori)

rap (ämneskategori)