Abstract
This study was conducted with the aim of optimising co-combustion of Hwange coal (HC) and Pinus sawdust (PS) using computational fluid dynamics (CFD) simulation. The study was motivated by the need to sustainably harness energy from PS, a waste product in need of environmentally friendly disposal techniques. Secondary to this motivation, was the need to study combustion parameters and reduce emissions from originally coal fired boilers by use of CFD simulations. Combustion parameters such as ignition, burnout, intensity, stability, and emission factors have shown varied results with respect to coal and biomass co-combustion. This limits the ability to predict the operation of co-combustion systems hence the need to investigate appropriate combustion parameters with respect to Hwange coal and Pinus sawdust.
To address the problem at hand, various objectives were postulated. At first, the prepared fuel blends where chemically and physically characterised by use of proximate, elementary, and thermogravimetric analysis techniques (TGA). These fuel samples were collected from the Southern African region, bituminous coal from Hwange and Pinus sawdust from the Eastern Highlands of Zimbabwe. Blended samples were prepared with a PS substitution by mass ranging from 0 to 30%, thus producing five samples, namely:100HC, 90HC10PS, 80HC20PS, 70HC30PS, and 100PS. Fuel combustion was then carried out using appropriate experimental techniques inside a drop tube furnace (DTF) equipped with important measuring instruments connected to a data logger. The co-combustion model was then developed using ANSYS Fluent 2022 and C++ to cater for the customisations which required user defined functions. A discrete phase submodel was employed to capture solid fuel particle transport whilst an eddy dissipation concept submodel was developed to capture chemical reaction mechanisms.
TGA results for a 20°C/min heating rate showed a decrease in ignition temperature from 457.70 to 265.76°C as PS blending increased from 0 to 30%. The burnout temperature also decreased from 820.05 to 699.97°C which corresponded to a burnout index improvement from 39.44 x 10-5 to 566.90 x 10-5. Combustion stability index decreased by 16.9% as blending increased to 20% and further dropped 50.5% when blending increased to 30%. The distributed
iv
activated energy model (DAEM) was applied to determine the activation energy of the devolatilization, volatile combustion and char combustion stages separately. The volatile combustion activation energy increased from 92.98 to 106.05 kJ/mol as blending increased from 0 to 30%. Similarly, char combustion activation energy increased from 52.90 to 103.85 kJ/mol as blending increased from 0 to 30%.
Co-combustion modelling of the DTF under a wall temperature loading of 1200°C demonstrated an overall decrease in particle residence time from 3.81s to 1.97s as blending increased since the furnace was downfired. Blending did not offset any zones of recirculation within the DTF as the average bulk fluid velocity remained within the laminar flow regime. The reaction zones became more intense as blending increased as demonstrated by the increase in devolatilization rate from 2.42 x 10-12 to 2.62 x 10-12 kg/s for the 1000°C furnace wall temperature loading. As such, blending increased the tendency of devolatilization occurring near the injection point highlighting the propensity of having unstable flames for such cases. Char combustion showed an increase in burnout rate from 5.28 x 10-10 to 6.21 x 10-10 kg/s as blending increased for the 1000°C furnace wall temperature loading. Nitrogen oxides ppm emissions showed an average decrease of 22% at a blending ratio of 30% PS. Evaluation of co-combustion experimental and modelling findings suggested a limit of 20% with regards to PS blending if no modification of existing coal fired boilers is envisaged.
Keywords: Co-combustion, CFD modelling, coal, biomass,FLUENT