CLEAN TECHNOLOGIES FOR WATER RECOVERY IN ALCOHOLIC
GAVACH Marjorie, FARGUES Claire*, SAGNE Camille, DECLOUX Martine, LAMELOISE
UMR GENIAL 1145, AgroParisTech, 1 avenue des Olympiades 91300 MASSY
Industries including a fermentation step use large volumes of water to dilute their musts. In
beet alcohol distilleries, water is recovered as condensates during stillage concentration.
Recycling this effluent into the fermentation step would enable reducing both water
consumption and wastewater rejection; but it was shown that some compounds they contain
cause the inhibition of the alcohol production (Morin Couallier, Payot et al. 2006). Reverse
osmosis treatment is an interesting process within this framework. A pre-selection step led to
choose the ESPA2 (Hydranautics) as the most appropriate membrane for this application
(Sagne, Fargues et al. 2008).
A study was undertaken on the treatment of an industrial condensate at both pilot- and semi-
industrial scales. The Trans-membrane Pressure (TMP = 5 to 30 bar) and Volume Reduction
Factor (VRF = 1 to 15) influences on the process performances were first studied on a 2540
spiral-wound RO pilot (membrane surface area S = 2.6 m2). The contents of the main
inhibitory molecules (acetic, propionic, butyric and valeric acids, 2-phenylethanol) were
measured by HPLC in the feed, retentate and permeate solutions. The respective inhibitions of
alcoholic production by these solutions were also determined (via batch fermentation tests).
The rejection was calculated according to Rs ( CfeedCperm eate) .
For solutes with MW > 80 g mol-1, the process was very efficient whatever the pressure (Rs >
95 % - FRV1). For all the solutes, the rejection increased from TMP = 5 to 15 bar and no
significant improvement was noticed beyond. Acetic acid was the less retained solute (37% <
Rs < 67%), due to its high polarity (logP = -0.17) and its small size (MW = 60 g/mol). For
TMP over 10 bar, the flux permeate exceeded the recommended limit of 30 Lh-1m2. A TMP
of 10 bar was then chosen to study the VRF influence, which showed for most solutes a fall of
the rejection between FRV 2 and 4. Concerning the acetic acid, its concentration in the
permeate exceeded its feed concentration for VRF 4. VRF = 2 was then considered as a
limit for this application. These results were confirmed on a semi-industrial RO pilot (3x4040
spiral-wound modules - S = 23.7 m2), implemented in a French distillery.
In order to improve the carboxylic acids rejection, the pH effect was studied (for TMP 10 bar;
VRF 2). As expected, a treatment at pH = 6 gave better rejections for the carboxylic acids
than at natural pH of the condensates (pH = 3.5), due to electrostatic repulsions between the
carboxylate anions and the negatively charged membrane. Continuous fermentation tests are
being performed to confirm the interest of neutralizing the effluent before treatment.
Morin Couallier, E., T. Payot, et al. (2006). "Recycling of distillery effluents in alcoholic
fermentation: role in inhibition of 10 organic molecules." Applied Biochemistry and
Biotechnology 133(3): 217-238.
Sagne, C., C. Fargues, et al. (2008). "Screening of reverse osmosis membranes for the
treatment and reuse of distillery condensates into alcoholic fermentation."
Desalination 219: 335-347.