Co-firing of cassava rhizome and eucalyptus bark in a fluidized-bed combustor using a reburning method to reduce NOx emissions

Cassava rhizome and eucalyptus bark are promising bioenergy resources in Thailand. Based on the availability and calorific values, the domestic annual energy potentials of these biomass residues account for about 45 PJ and 50 PJ, respectively. However, individual burning of these biomass fuels in a fluidized-bed combustion system are associated with some operational problems, such as (i) noticeable NO emission during combustion of cassava rhizome with its elevated fuel-N and (ii) flame instability when firing high-moisture bark. Furthermore, elevated potassium contents in fuel-ash indicate a propensity of both fuels for bed agglomeration when fired individually in a fluidized bed of silica/quartz sand. In this thesis work, pelletized cassava rhizome (a base fuel) and shredded eucalyptus bark (a reburn fuel) were co-fired in a 200 kW fluidized-bed combustor, using a reburning technique to reduce NO emission from the combustor. At the first stage of the study, the selected fuels were (co-)combusted at the constant heat input to the reactor, using conventional bed material (silica sand), for variable operating conditions. In this test series, the energy fraction of the secondary fuel (EF2) varied within 0–0.25, with excess air (EA) of 20–80% (at fixed EF2) and secondary-to-total air ratio (SA/TA) of 0.10–0.25 (at fixed EA). During a test run at fixed operating parameters, the temperature and gas concentrations (O2, CO, CxHy as CH4, and NO) were measured along the axial and radial directions inside the combustor, as well as at stack, to investigate combustion and emission characteristics of the reactor. A cost-based optimization technique was used to quantify the optimal operating parameters: EF2 = 0.15, EA ≈ 60%, and SA/TA ≈ 0.25, to minimize the “external” (emission) costs of the combustor. Under these operating conditions, a 50% NO emission reduction is achievable, compared to burning the base fuel alone. However, when using silica sand as the bed material, a fairly small proportion of agglomerates were found in the bed after 8 h co-firing tests with reburning. The agglomerates were mainly formed with binding materials, such as K2O–CaO–SiO2 and K2O–SiO2 eutectics with low melting points, generated in “coating-induced” and “melt-induced” agglomeration mechanisms.At the second stage of the thesis work, two alternative bed materials (alumina sand and alumina/silica sand mixture in equal proportions) were therefore employed to inhibit bed agglomeration during long-term (30 h) combustor operation for co-firing the selected fuels with a reburning technique under the optimal EF2, EA, and SA/TA. To verify a steady-state operation, the bed temperature and the pressure drop across bed were measured, along with the CO, CxHy as CH4, and NO emissions, during the entire test period. At this stage, a special attention was given to the behavior of the bed materials in the long-term co-combustion tests. Physiochemical characteristics of the used/reused bed materials and those of particulate matter (PM) sampled from the combustor were analyzed at different operating times using SEM–EDS and XRF techniques, as well as a particle size analyzer. No evidences of bed agglomeration were found in the proposed combustor during these 30 h co-firing tests with reburning. However, the two alternative bed materials showed the time-related changes in their physiochemical characteristics, pointing at a gradual decrease of the bed capability to resist bed agglomeration, particularly when using a mixture of alumina/silica sand.