A statistical comparison of protein and carbohydrate characterisation methodology applied on sewage sludge samples.

Biochemical characterization of organic matter is becoming of key importance in wastewater treatment. The main objectives are to predict organic matter properties, such as granulation or flocculation, and hence treatment performance. Although standardized methods do exist for some organic molecules, such as volatile fatty acids or lipids, there are no standard methods to measure proteins and carbohydrates content, both biochemical families being the main components of sewage sludge. Consequently, the aim of the present work is to investigate the efficiency of several colorimetric methods to determine proteins and carbohydrates content as well as their compatibility with the sludge matrices. The different methods have been evaluated based on statistical criteria such as sensitivity, linearity, accuracy, rightness, and specificity using standard molecules such as Bovine Serum Albumin (BSA), glucose, cellulose and a certified reference product. The Lowry and the Dubois methods have been shown to be the best compromise for the considered criteria after having been tested on sewage sludge samples obtained from different locations in a wastewater treatment plant. In average, the measured volatile fatty acids, lipids, proteins and carbohydrates contents represented 80 ± 7% (% volatile solids) of the organic matter. Proteins and carbohydrates represented in average 69 ± 3%. This study underlined that the choice of a relevant methodology is of great importance for organic matter measurement.