Analysis of co-firing biomass with South African coal in pulverised coal boilers

摘要

In this study, the effect on the extent of emissions reduction with co-combustion ofudsmall proportions (0-20%) of biomass with coal in a pulverised fuel combustor, isudinvestigated. Such emissions include CO2, SO2 and NOx. South African coal hasudhigh ash content and low volatile content that affects its ignitability. The effect ofudbiomass co-firing on ignitability and slagging is also investigated. While higherudpercentages of biomass could reduce emissions more and improve ignitability,udbiomass tends to increase the undesirable fouling and slagging propensities, henceudlimiting the amount of biomass that can be co-fired.udTwo types of biomass, namely grass and sawdust (calorific value 16-18 MJ/kg), andudone coal (calorific value 21 MJ/kg) were used in this study. Combustion tests wereudcarried out using the Eskom 1 MW Pilot Scale combustion Test Facility (PSCTF).udThe coal chosen was representative of an average coal burned at the Eskom’s coalfiredudpower stations. For each of the types of biomass, three blends of biomass andudcoal were used, resulting in seven different feed fuels including coal alone. Theudratios of biomass to coal, on an energy basis, in the three blends were 10%:90%,ud15%:85% and 20%:80%. Seven tests were also carried out with the same fuelsudusing a drop tube furnace (DTF) to determine the reaction kinetics of the baselineudcoal and its three different blends with each type of biomass. The reactionudparameters obtained from these tests were used as input data in numericaludsimulations of the tests. Simulation using CFD software was used to predictudcombustion characteristics of each fuel in the PSCTF, which in turn can beudextrapolated to predict the performance in a full scale commercial boiler. Theudsimulation results were validated by the experimental data from the PSCTF;udcomparison of the combustion and emissions characteristics with experimental dataudfrom the PSCTF showed that the simulation procedure was capable of predictingudthese characteristics with generally good accuracy. The results coming out of this work are positive. Co-firing with grass at the PSCTFudwas found to reduce the emissions by between 13% and 50% for NOX, between 12%udand 23% for CO2 and between 21% and 29% for SO2 as the proportion of biomassudincreased from 10% to 20%. The maximum emissions reduction for sawdustudoccurred at 20% co-firing ratio; these are, 29% for NOx, 17% for CO2 and 15% forudSO2.udFor grass based co-firing, the combustion efficiency of the coal used was improvedudby 0.55% and 0.62% for 15% and 20% co-firing ratios respectively; whereas that ofudsawdust was between 0.78% and 0.38% as co-firing ratio was increased from 10%udto 20%. Combustion efficiency for 10% grass dropped by 0.65%. The DTF resultsudindicate that both grass and sawdust were able to improve the ignitability of the coaludused in a temperature range of 1000°C to 1200°C. The combustion efficiencyuddetermined from DTF results for this range indicate an improvement of betweenud0.1% and 1.4% for grass and between 0.8% and 2.45% for sawdust. The QEMSCANudanalysis of slag deposits when co-firing biomass indicated that they should generallyudbe weak enough to be handled by soot blowing equipment.udIn conclusion grass (herbaceous) biomass resulted in greater emissions reductionsudthan sawdust (ligneous). It was also found that slagging would be less of a problemudwhen using grass, because the slag deposits are more friable than those of sawdust.udAt higher co-firing ratios grass based co-firing was found to improve coal ignitabilityudbetter than that can be achieved with sawdust. The optimum co-firing ratio with grassudwould appear to be about 15% on an energy basis.udThis project was carried out to obtain fundamental technical information on some ofudthe salient effects of using biomass co-firing with pulverised power station coal.udBefore any implementation can be carried out, the next step would be to conduct auddetailed technical and economic feasibility study into possible large-scaleudapplications, using full systems engineering principles.