Residues from biochemical production of transport biofuels in Northern Europe combustion properties and applications

Abstract: Residues from biochemical production of liquid transport biofuels will probably become available for energy use if more gasoline and diesel is substituted. For processes used in northern Europe they amount to 35-65 % of the feedstock energy and despite interest from energy companies, their fuel properties are largely unknown. Combustion-relevant material properties have been characterized and fuel-specific combustion properties determined for powder-, grate- and fluidized bed combustion. Suitable combustion applications have been identified. A techno-economic evaluation of utilization of a selected residue for supplying process heat and electricity to the transport biofuel production, combined with sale of surplus energy has been done. Residues studied are rape-seed meal (RM) from biodiesel production, wheat distillers dried grain with solubles (wheat DDGS) from grain-based ethanol production and hydrolysis residue (HR) from wood-based ethanol production. For RM and wheat DDGS, mixtures with typical forest- and agricultural fuels were also studied. Combustion experiments were performed in a fluidized (quartz) bed (5 kW), an under-fed pellet burner (12 kW), and in a powder burner (150 kW). The calorific value for HR was higher than for wood, for RM and wheat DDGS it was similar to wood. More char was produced from HR, otherwise TGA results showed that thermal kinetics was similar to wood for all fuels. All pulverized residues had better feeding properties than wood powder. While RM and wheat DDGS ash contents were higher than for most common forest and also for some agricultural fuels, HR mostly had very low contents of ash, alkali, Cl, S and N. RM and DDGS had high concentrations of S, N, K and P compared to most other biomass fuels. RM had higher Ca and Mg concentrations than DDGS. The Cl content of wheat DDGS was similar to wheat straw, while RM had a lower Cl content, similar to wood. Combustion of all pulverized residues was stable with CO emissions not higher than for wood powder. While the bed agglomeration tendency of RM was low and comparable to many forest fuels the wheat DDGS bed agglomeration tendency was high and comparable to wheat straw. The K, P and Si contents of wheat DDGS formed layers of K-phosphates/silicates on the quartz grain particles, with low melting temperatures and therefore sticky, resulting in bed agglomeration. For RM, this effect was mitigated by the considerable Ca and Mg concentrations, making the layers formed less sticky, despite the high K and P concentrations. For basically the same reason, the slag formation tendency of RM was moderate and comparable to many forest fuels while wheat DDGS had a slag formation tendency which was even higher than for typical wheat straw. HR had very low bed agglomeration and slagging tendencies. For RM and wheat DDGS, emissions of NO and SO2 were generally high, for HR considerably lower. While HCl emissions for RM were low, they were relatively high for fluidized bed combustion of wheat DDGS. Particle emissions from RM and wheat DDGS were generally high. For powder combustion of RM and wheat DDGS, particle emissions were 15-20 times higher than for wood. The particle emissions from combustion of HR were generally low. For fluidized bed- and grate combustion of RM the finer particles (< 1 μm) contained mainly alkali sulfates. RM addition to bark tended to lower the particle Cl concentrations, potentially lowering the risk of high-temperature corrosion. For fluidized bed combustion of wheat DDGS and wheat DDGS-mixtures the finer particles contained mainly K and S. The Cl concentrations of the fine particles in fluidized bed combustion were reduced when wheat DDGS where added to logging residues and wheat straw in fluidized bed combustion. In grate combustion the Cl- and P-concentrations in the finer particles during combustion of the wheat DDGS-mixtures were considerable higher than during fluidized bed combustion. The fine particles from powder combustion of RM mainly contained P and K, while they mainly contained K, P, Cl, Na and S from wheat DDGS(apart from C and O). A possible use of RM is as a sulfur-containing additive to biomass fuels rich in Cl and K in large-scale fluidized-bed and grate combustors for avoiding ash-related operational problems in fluidized beds and grate combustors originated from high KCl concentrations in the flue gases. Due to its high slagging and bed agglomeration tendencies, the best use of wheat DDGS may be to mix it with other fuels, preferably with high Ca and Mg contents (e.g. woody biomass fuels), so that only a minor fraction of the total ash-forming elements is contributed by the wheat DDGS. Because of their high N- and S contents, RM and wheat DDGS require applications with flue-gas cleaning, economically viable at large-scale. Powder combustion of RM and wheat DDGS should be used with caution, as potassium phosphate particles have low melting temperatures and could therefore increase the risk of deposit formation. Use of HR in small-scale pellet appliances is an interesting option due to low emissions, low ash content and low slagging tendency. While most large-scale combustion uses of HR would be feasible, the low ash and alkali contents and stable powder combustion of HR may be better exploited in a combined-cycle process, as the alkali content can be kept sufficiently low for use in robust gas turbines, simplifying the gas cleaning. In the techno-economic assessment, residue (HR) was assumed to be combusted on site, to supply process steam and electricity to the liquid biofuel production (wood-based ethanol) with surplus residue either sold as solid fuel or used for additional heat and power generation. With a combined cycle to increase electricity production, a location with a large district heating base load is not needed. As electricity replaced is largely generated with fossil fuels, a combined cycle is significantly more effective as a climate mitigation measure than a steam-cycle only, with about 25 percent greater reduction in CO2 emissions per litre of ethanol produced. While it is generally accepted that energy use of the residue is important to the process economy and environmental benefits of ligno-cellulosic ethanol production, it can be concluded from this study that the choice of integrated process design has a significant impact on CO2 emissions.