Abstract

This work uses a fully predictive mass transfer model to simulate the supercritical CO2 extraction of vegetable oils from prepressed oilseeds in the 1-m(3) vessel of an industrial multi-vessel plant operating at 40 degrees C and 30 MPa with the purpose of minimizing the operational cost. The work analyses the effect of particle diameter (0.5, 1, 2, 3, and 4 mm), superficial CO2 velocity (2.76, 5.52, or 11.0 mm/s), and number of extraction vessels (2, 3, or 4) on optimal extraction time and minimal operational cost. Keeping other variables constants, cost diminishes as particle diameter decreases. Although the optimal superficial CO2 velocity increases as particle diameter decreases, in the case of small (<= 1 mm) particles, substrate fluidization may place an upper limit to the superficial velocity. Within the studied region, best superficial CO2 velocities are 11.0 mm/s for particles smaller than 1-2 mm, 2.76 mm/s for particles larger than 3-4 mm, and 5.52 mm/s for particles in between. Keeping other variables constant, the cost of extraction of medium-to-large (>= 2 mm) particles decreases as the number of extraction vessels increases, at the expense of an increase in extraction time. However, because of a sharp transition wave that develops when extracting small (<= 1 mm) particles that separates fully extracted (downstream) from virtually unextracted (upstream) substrate within extraction vessels, two-vessel plants are best for small particles. The lowest operational cost observed in this work was USD 4.08 kg(-1) oil for the extraction of 2-mm particles using 3.30 m(3)/h of CO2 (U=2.76 mm/s) in a four-vessel plant. (C) 2014 Elsevier B.V. All rights reserved.