Torrefaction of biomass for power production

In order to increase the share of biomass for sustainable energy production, it will be an advantage to utilize fuels as straw, wood and waste on large suspension fired boilers. On a European scale, currently large straw resources are available that are not fully utilized for energy production. Straw can be co-fired with coal in suspension fired power plants with a maximum straw share of 10 to 20 wt%. However, 100% straw firing induced several problems that can impede both boiler availability and power efficiency. Straw is highly fibrous and tenacious in nature, therefore a relatively high amounts of energy is needed to pulverize the straw to a size where a good burn out can be obtained. Also the large alkali and chlorine content in straw often induce severe chlorine rich deposit formation on super heaters. The chlorine rich deposits are corrosive and to prevent high superheater corrosion rates, relatively low superheater temperatures have to be applied, which in turn lower the power efficiency.
The idea for this Ph.D. project is to develop a biomass pretreatment method that could provide the heating value of the fuel for the boiler, but in a way such that the fuel is easily pulverized and the superheating can be done without an exposure of alkali rich flue gas on superheaters. A potential pretreatment process is to use a ball mill with an integrated torrefaction process. The char produced is very fragile and can be easily pulverized down to a size where a high burn out is obtained. The present Ph.D. thesis focus on the following subjects: 1) the development of experimental procedures for a novel laboratory scale reactor (simultaneous torrefaction and grinding) and a study on the torrefaction of straw and wood; 2) study the influence of biomass chemical properties such as ash content, ash composition and carbohydrate composition on torrefaction characteristics by using a broader range of biomasses; and 3) quantification of chlorine and sulfur release during torrefaction.
A novel laboratory scale experimental setup which combines torrefaction and a ball mill has been constructed for studies of the influence of feedstock type, temperature and residence time on the product yields and particle size reductions. The laboratory set up was used to investigate the torrefaction properties of Danish wheat straw and spruce chips. A standard experimental procedure was developed based on initial experiments which evaluated the influence of reactor mass loading, gas flow and grinding ball size and material. The particle size reduction capability of the torrefaction process has been evaluated by using the relative change in d50 of the product size distribution, and this method was compared with the Hardgrove Grindability Index (HGI), showing reasonably similar results.
Significant differences in torrefaction behavior have been observed for straw and spruce chips torrefied at 270 – 330 °C. Torrefaction of straw for 90 minutes yielded a higher mass loss (27 – 60 wt %) and a larger relative size reduction (59 – 95%) compared to spruce (mass loss of 10 – 56 wt% and a size reduction of 20 – 60%). The two types of biomass investigated differ with respect to hemicellulose type, lignocellulosic composition, particle morphology and ash composition where straw has higher alkali content. Experiments with separate particle heating and grinding showed a swift grinding of the torrefied biomass which implies that the rate limiting step in the laboratory reactor is the heat transfer, and not the grinding process.
Different torrefaction characteristics are observed from straw and wood chips, therefore an improved understanding and ability to predict the torrefaction characteristic of different biomass types are desired. In this study, the influence of biomass chemical properties (carbohydrate composition and alkali content) on the torrefaction behavior with respect to mass loss and grindability is investigated. Six raw biomass samples (Danish wheat straw, miscanthus, spruce, beech, pine, and spruce bark) with different chemical and physical properties were pyrolyzed by Simultaneous Thermal Analysis (STA) and torrefied in the simultaneous torrefaction and grinding reactor. The effect of biomass alkali content on torrefaction characteristics were furthermore investigated by washing or impregnating (KCl and K2CO3) of selected biomass. The solid yields at the investigated torrefaction temperatures (270 and 300 °C) are strongly influenced by the biomass potassium content as well as to some extent the lignocelluloses composition. High biomass potassium content leads to a relatively low solid yield; however in a single case (spruce bark), a high lignin content leads to a relatively high solid yield even in the presence of relatively high potassium content. In summary both potassium content and lignocelluloses composition affect the solid yield obtained by torrefaction. A significant decrease in d50 value of the torrefied products was observed when the alkali content is increased from 0 to 0.2 wt% db, while no additional effect is seen for higher potassium contents.
The release of chlorine (Cl) and sulfur (S) to gas phase during biomass torrefaction has been investigated via experiments in laboratory-scale reactors by using six biomasses which cover a wide range of ash content and ash-forming elements in the temperature range of 150 – 500 °C. The relative release of chlorine and sulfur was calculated based on mass balance and analysis of the biomass before and after torrefaction. In few cases, measurement of methyl chloride (CH3Cl) in the gas released from straw torrefaction has been conducted. Initial release of chlorine was observed at 250 °C and about sixty percent of chlorine was released from straw at 350 °C. The analysis of methyl chloride from the released gas showed that most of chlorine was released as CH3Cl. By using a large amount of straw (40 g compared to 5 – 20 g), less Cl is released, probably due to more reactive sites available for secondary reactions. The secondary reactions can be reaction with relatively stable basic functionalities on the char surface or reaction with potassium to generate KCl. Almost complete release of chlorine was observed for woody biomass at 350 °C. This result showed an agreement with the previous studies reported that the biomass with a lower chlorine content released a higher fraction of chlorine during the pyrolysis process. Significant sulfur release (about 60%) was observed from the six biomasses investigated at 350 °C. It is seen that the initial sulfur content in biomass did not influence the fraction of sulfur release during torrefaction.