The research presented focuses on the use of Carbon-dioxide (CO₂), Nitrogen (N₂) and Argon (Ar) as purge gases for torrefaction. Torrefaction using CO₂ as a purge gas may further improve the fuel characteristics of the torrefied fuel when compared to N₂ and Ar (which are entirely inert), making it better suited for use as a fuel for co-firing with coal or gasification. Three different biomasses were investigated: Juniper wood chips, Mesquite wood chips, and forage Sorghum. Experiments were conducted using a thermo-gravimetric analyzer (TGA, TA Instruments Model Q-600) to determine the effect of the purge gas over a wide range of torrefaction temperatures (200-300°C). TGA weight traces (thermograms) showed an increased mass loss when using CO2 as a purge gas when compared to N₂. The increased mass loss when CO₂ was used is attributed to a hypothesized reaction between the CO₂ and fixed Carbon contained within the biomass. Torrefaction of biomass, using Ar as the purge gas, produced results similar to torrefaction using N₂. Derivative Thermo-Gravimetric analysis (DTG) was done to determine the temperature ranges over which the three main components of biomass (hemicellulose, cellulose, and lignin) decomposed. The DTG results are in agreement with previously published research. From TGA thermograms and DTG analysis it was determined that torrefaction at higher temperatures (>260°C) likely result in the breakdown of cellulose during torrefaction, an undesired outcome. Proximate, ultimate, and heat value analysis was done on all three biomasses. All three contain a relatively high Oxygen content, which serves to decrease the higher heating value (HHV) of the biomass. The HHV of Juniper, Mesquite, and Sorghum on a dry ash-free (DAF) basis were 20,584 kJ/kg, 20,128 kJ/kg, and 19,389 kJ/kg respectively. The HHV of the three biomasses were relatively constant as expected for agricultural biomass. From TGA analysis (thermograms and DTG), an optimal torrefaction temperature was determined (240°C) based upon the amount of mass lost during torrefaction and estimates of energy retained. Batch torrefaction of all three biomasses at the optimal torrefaction temperature was completed using a laboratory oven. All three biomasses were torrefied using CO₂, N₂, and Ar as a purge gas. Proximate, ultimate, and heat value analysis was done for each of the torrefied fuels and compared. Results of the fuel property analysis showed torrefaction reduced the moisture content and oxygen percentage of the fuel resulting in the torrefied biomass having a larger HHV when compared to raw biomass. Due to inherent mass lost during torrefaction, the amount of energy retained in the torrefied biomass was calculated to determine the percentage of the virgin biomass energy content that remained. Torrefaction using CO2 resulted in the lowest amount of energy retention of all three purge gases tested (78.86% for Juniper); conversely, Nitrogen resulted in the highest amount of energy retention (91.81% for Sorghum.) Torrefaction of the biomass also increased the fixed carbon (FC) content of the fuel. The grindability of the torrefied biomass was investigated via size distribution analysis of the raw and ground biomass. Initial size distribution analysis showed that torrefaction of Mesquite and Juniper resulted in smaller particle sizes; with a greater fraction of the torrefied biomass passing through smaller meshes. Analysis of the ground biomass samples showed that torrefaction improved the grindability of the fuel. The percent of torrefied biomass that passed through an 840 micrometer mesh increased by over 20% for both Mesquite and Juniper when ground. Sorghum exhibited similar increases; however, the amount of increase is less apparent due to the smaller particle size distribution of the raw Sorghum.