Normalization of wastewater coagulation-flocculation trials and implications in terms of variability in treatment performance and comparison of commercial coagulants.

Normalizations by TSS or P-PO43- initial concentrations are consistent as they are correlated to the Jar-test performances. Jar-tests results are independent of the wastewater quality variations in terms of TSS and P-PO43- and independent of the WWTP origin of the water. A notable variability in the TSS results indicates that the pollutant's initial load has to be taken into account even with normalizations. This variability is lower with normalization by P-PO43-, indicating that this is the best indicator to consider. It is possible to determine that the optimal cation dosage is 60 mol Fe3+ / kg P-PO43- as it guarantees a residual concentration of 0.7-1.0 mgP/L and a good removal of TSS.Then, six commercially available cationic coagulants were compared, demonstrating a comparable effect at a comparable normalized molar dose, whatever the coagulant on both TSS and P-PO43-, as well as on soluble carbon and nitrogen. The differences observed between these types of coagulants in the literature are then probably due to methodological issues. Settling velocity distribution charts were also very similar for the different coagulants. This confirms that the source of cation and the type of cation have no significant effect on physico-chemical settling performances.

International cross-validation of a BOD5 surrogate.

BOD5 dates back to 1912 when the Royal Commission decided to use the mean residence time of water in the rivers of England, 5 days, as a standard to measure the biochemical oxygen demand. Initially designed to protect the quality of river waters from extensive sewage discharge, the use of BOD5 has been quickly extended to waste water treatment plants (WWTPs) to monitor their efficiency on a daily basis. The measurement has been automatized but remains a tedious, time- and resource-consuming analysis. We have cross-validated a surrogate BOD5 method on two sites in France and in the USA with a total of 109 samples. This method uses a fluorescent redox indicator on a 96-well microplate to measure microbial catabolic activity for a large number of samples simultaneously. Three statistical tests were used to compare surrogate and reference methods and showed robust equivalence.