Comparison of methods for assessing reverse osmosis membrane treatment of shrimp process water

Interest in reuse of process water from the food industry has reinforced the importance of controlling and monitoring the effectiveness and reliability of treatment systems regarding removal of organic matter and microorganisms. The ability of adenosine triphosphate bioluminescence, conductivity, turbidometry, absorbance, and multichannel fluorescence spectroscopy for indirectly monitoring the integrity of a reverse osmosis membrane when treating process water recovered from peeling in a shrimp processing line was evaluated. This study demonstrated that reverse osmosis was capable of removing bacteria (ca. 7 log CFU ml(-1)) to the levels required by the regulatory authorities for water recycling within the same food unit operation. Adenosine triphosphate and turbidometry showed a higher sensitivity for detecting compromising conditions at the treatment system (0.1% concentration of feed in permeate) and a better correlation with the aerobic count at lower levels than the other methods investigated. The sensitivity for assessing membrane integrity of conductivity and multichannel fluorescence was 1% of feed in permeate. Impact of feed variations was best leveled out in the permeates for turbidity measurements. Multichannel fluorescence spectroscopy may require laborious calibration procedures and expertise regarding data analysis and interpretation of results, which are not always available in food industries. Absorbance did not respond to changes in membrane integrity and was not well correlated to the aerobic count because of the poor sensitivity of this method for these purposes.