Abstract: Abstract: This study aims to experimentally investigate the effect of steam in the heating medium on the biomass torrefaction and subsequent pyrolysis. A systematic investigation was made on the effect of energy that consumed by the torrefaction pretreatment on the economics of biomass utilization. The temperature of the waste heat flue gas was close to the biomass torrefaction temperature for the potential as a heat source. As such, the biomass utilization process was more economically feasible and environmentally sustainable using the waste heat flue gas as the carrier gas and heat source for the torrefaction process. However, the waste heat flue gas also contained steam, CO2 and O2 in addition to nitrogen. The complex composition of the heating medium posed a great challenge on the utilization for the biomass torrefaction. It was necessary to clarify the effects of the flue gas atmosphere on the torrefaction and the subsequent pyrolysis. In the experiments, the elm woody biomass was torrefied under the different atmospheres (nitrogen and nitrogen-steam) and various temperatures (200, 230, 260, and 290 ℃). Then, the torrefied biomass was pyrolysed on the thermogravimetric mass spectrometry and pyrolysis chromatography-mass spectrometry. The results showed that the torrefaction was reduced the oxygen content of biomass, but the presence of steam during torrefaction was inhibited the reduction of oxygen content and the generation of volatiles. The biomass sample that torrefied at 290?C (TNS290) presented the lowest relative content of volatiles (52.31%). The calorific value of the biomass increased with the torrefaction temperature, compared with the raw biomass sample. However, this increase in the calorific value of the torrefied samples was hindered by the presence of steam in the heating medium. The Dehydration, decarbonylation, and demethylation of elm wood occurred during torrefaction. The pyrolysis weight loss rate of each torrefied sample was lower than that of raw sample, where the peak weight loss rate of TN290 sample was the minimum of 0.3%/℃. At the same time, the presence of steam was greatly contributed to the maximum weight loss peak of nitrogen-steam torrefaction samples backward. Specifically, n=1 was used to calculate the kinetics of biomass pyrolysis. The pyrolysis of elm was divided into two stages, including a low temperature stage (the main reaction zone of pyrolysis) and a high temperature stage. The activation energies in the two-stage pyrolysis of torrefied biomass samples significantly increased, compared with the raw sample. In the main reaction temperature zone of pyrolysis, the steam at the torrefaction temperature of 200 ℃ and 290 ℃ increased the activation energy that required for the main pyrolysis reaction of the elm sample, but the steam was reduced the activation energy at the torrefaction temperature of 230 ℃ and 260 ℃. There was the very small difference between the activation energy of nitrogen-steam and nitrogen samples at the same torrefaction temperature, indicating the insignificant effect of steam on the activation energy of elm pyrolysis in the high-temperature stage of pyrolysis. The torrefaction increased the CO yield in the non-condensable gases from the pyrolysis, but decreased the H2 yield. However, the presence of steam during the torrefaction process was reduced the release of H2, CH4, H2O, CO, and CO2 during the pyrolysis of torrefied elm. Therefore, the presence of steam during torrefaction was also more favorable for the removal of acetic acid from the pyrolysis oil.