COSMO-based solvent selection and Aspen Plus process simulation for tar absorptive removal

Document Type

Journal Article

Publication Title

Applied Energy


Elsevier Ltd


School of Engineering




Funding information available at: https://doi.org/10.1016/j.apenergy.2019.113314


Zhang, X., Pan, J., Wang, L., Qian, T., Zhu, Y., Sun, H., ... Liu, C. (2019). COSMO-based solvent selection and Aspen Plus process simulation for tar absorptive removal. Applied Energy, 251, Article 113314. Available here


Tar accumulation is a major problem that has critically hindered further development of biomass gasification. Current studies on oil washing technologies have made progress, but experimental screening of core absorbents and process performance lack theoretical insight. Therefore, based on the design of Quench Coupled with ABsorption Technology (QCABT), the benzene, toluene, phenol, and naphthalene mixture representing real tar components was used as a tar simulant. The method of screening absorbents by Conductor-like Screening Model for Real Solvents (COSMO-RS) combined with thermodynamic relationships was established, and the tar removal process was investigated by Aspen Plus simulation. The results revealed that the developed semi-quantitative absorbent screening method is more efficient than that using only infinite dilution activity coefficients. A mixture of screened soybean oil, and N-formylmorpholine (NFM) is the optimal tar absorbent. The maximum relative error between Aspen Plus simulation and the experiment is 9.5‰ indicating good reliability. The addition of 30 v/v% of NFM could effectively promote the absorption of benzene and toluene. Higher liquid–gas ratio of absorption, lower absorption temperature, and better regeneration could ensure the satisfactory gas purification degree. The loss tendency and rate of oleic and linoleic acids (soybean oil simulant) are basically the same, but the loss rate of NFM is 4–6 orders of magnitude higher—this has the biggest contribution to solvent loss. The 90 °C condenser at the outlet of the desorber can control the solvent loss rate effectively. Under optimal conditions, the content of tar, energy loss, as well as loss ratio can be reduced to below 30 mg/Nm3, 1600 kJ/kg, and 0.5‰, respectively. As a result, this fascinating method provides quantitative guidance for tar absorption and removal.