Search

The stability with respect to filamentous bulking of two activated sludge fully-aerobic systems, one with a completely mixed tank and one with a channel reactor, fed either by a synthetic wastewater or by a primary-settled municipal wastewater, of variable composition and flow rate, has been investigated. The morphological characteristics of the biomass in terms of floc size and roughness and of filamentous bacteria abundance have been monitored by image analysis. Severe bulking was only observed in the well-mixed tank fed at a constant flow rate by synthetic substrate of constant concentration, when the channel reactor fed in a similar manner was fully stable. Variations of biomass characteristics as well as of settling properties were observed on both systems fed with the real wastewater, but these events were related to the characteristics of the wastewater, as similar changes were observed on the full-scale plant fed with the same substrate. In any case, automated image analysis was an efficient way to monitor in detail the fate of the activated sludge at pilot and full scale.The stability with respect to filamentous bulking of two activated sludge fully-aerobic systems, one with a completely mixed tank and one with a channel reactor, fed either by a synthetic wastewater or by a primary-settled municipal wastewater, of variable composition and flow rate, has been investigated. The morphological characteristics of the biomass in terms of floc size and roughness and of filamentous bacteria abundance have been monitored by image analysis. Severe bulking was only observed in the well-mixed tank fed at a constant flow rate by synthetic substrate of constant concentration, when the channel reactor fed in a similar manner was fully stable. Variations of biomass characteristics as well as of settling properties were observed on both systems fed with the real wastewater, but these events were related to the characteristics of the wastewater, as similar changes were observed on the full-scale plant fed with the same substrate. In any case, automated image analysis was an efficient way to monitor in detail the fate of the activated sludge at pilot and full scale.

The reliability of a gas membrane sensor, used to monitor volatiles in fermentation, is studied in this work. The influence of some important components of the broth on the sensor response is investigated: mineral salts, yeast extract, antifoaming agent and solid particles. All experiments were performed with ethanol as volatile. The presence of the various components affects slightly the output signal of the sensor, but the effect is negligible. Correlations are proposed to predict the effect of mineral salts.

A model of broth pH has been established based upon the mineral requirements of the microorganism, the substrate consumption and the metabolites production. It has been tested on acetic acid, citric acid and yeast fermentations. It has then been used to predict the mineral composition of the feed for fedbatch growth of Saccharomyces cerevisiae.

On-line estimation of lactic acid produced by Lactobacillus casei in a pilot scale reactor coupled with a microfiltration module is performed by measuring the conductance on the permeate effluent. Non-interference by other ions (inorganic and organic) present in the medium has been verified. This method is found more reliable that the estimation of lactate concentration from the amount of alkaline solution supplied for the pH control.