Radial Variations in Axial Velocity Affect Supercritical CO2 Extraction of Lipids from Pre-pressed Oilseeds

• Published in: Food engineering reviews Vol. 13; no. 1; pp. 185 - 203
• Main Authors: del Valle, José M, Núñez, Gonzalo A., Díaz, J. Facundo, Gelmi, Claudio A.
• Format: Journal Article
• Language: English
• Published: New York Springer US 01-03-2021; Springer Nature B.V
• Subjects: Biot number; Carbon dioxide; Chemistry; Chemistry and Materials Science; Chemistry/Food Science; Computational fluid dynamics; Dimensionless numbers; Fluid flow; Food Science; Independent variables; Lipids; Mass transfer; Packed beds; Particle size; Plant extracts; Plug flow; Porosity; Prediction models; Seeds; Substrates; Supercritical gas extraction; Variation; Velocity; Vessels; Nonlinear solute partition; Radial changes in CO; Linear driving force model; Pre-pressed oilseed; Packed bed extraction vessel; Radial variations in bed porosity; velocity; extraction; Mathematical simulation; Supercritical CO
• ISSN: 1866-7910; 1866-7929
• DOI: 10.1007/s12393-020-09232-1

Packed beds of spherical particles in a cylindrical vessel have a high porosity region next to the vessel wall that allows preferential fluid flow. Consequently, there are radial variations in porosity (
ε
) and superficial fluid velocity (
U
) that depend on the vessel-to-particle diameter ratio (
D
/
d
p
) and the flow regime of the fluid. This work ascertained if these radial variations affected SuperCritical (SC) CO
2
extraction curves of oil from pre-pressed seeds at 40 °C and 28 MPa, as compared with the commonly adopted plug flow condition. It focused specifically on comparing extraction curves as a function of the controlling mass transfer mechanism (characterized by the dimensionless Biot number,
Bi
) and
D
/
d
p
ratio. A predictive model was adopted to describe the SC-CO
2
extraction of oil from sheared seeds comparing plug flow with radial variations in superficial CO
2
velocity,
U
(
r
), from literature correlations. Selected independent variables included the initial oil content of the substrate (132.7 ≤ 
C
o
 ≤ 397.2 g/kg),
d
p
(1 or 2 mm),
U
(1–4 mm/s), and vessel volume (0.038–495 L).
C
o
markedly affected the effective diffusivity of the oil (0.780 ≤ 
D
e
 ≤ 6.24 × 10
−10
 m
2
/s), whereas
d
p
and
U
moderately affected the film mass transfer coefficient (2.44 ≤ 
k
f
 ≤ 7.40 × 10
−5
 m/s). Radial variations in superficial CO
2
velocity decreased extraction rates, with differences between extraction curves when considering plug flow or adopting
U
(
r
) diminishing as
Bi
increased for
D
/
d
p
 = 20, or as
D
/
d
p
increased for
Bi
 = 18.
Bi
increased by increasing
U
and
k
f
, or decreasing
C
o
and
D
e
, whereas
D
/
d
p
increased by increasing vessel volume. The radial variations in porosity in a packed bed and associated changes in superficial CO
2
velocity may have a more pronounced negative impact in laboratory or pilot plant extraction vessels (small
D
) than industrial vessels (large
D
), mainly when extracting small particles and applying large superficial CO
2
velocities. A proxy for the SC-CO
2
extraction of oil from pre-pressed seeds in an industrial extraction vessel (495-L capacity,
D
/
d
p
 = 270) would be plug flow using the porosity, and superficial CO
2
velocity predicted for the axis of the extraction vessel (
ε
o
and
U
o
, respectively). Literature correlations predict a value of
ε
o
slightly less than
ε
, and value of
U
o
slightly less than
U
. The remainder of the CO
2
bypassing the vessel along a high porosity region near the vessel wall, containing a small fraction of the loaded substrate.