Modelisation hybride du procede de boues activees dans l'industrie des pates et papiers: Etude et prediction du foisonnement filamenteux

The main objective of this thesis was to develop a systematic approach for predicting, troubleshooting and reducing filamentous bulking problems in the pulp and paper industry. In order to predict the onset of filamentous bulking problems before its momentum becomes unstoppable or before the use of chemicals is imperative, it is important to develop procedures that are based on actionable data that operators can easily follow and incorporate into their decision-making routine.;An activated sludge mechanistic model reflecting the specific characteristics of pulp and paper effluents was developed and was used as a tool for a better understanding of the causes of filamentous bulking. The model called ASMP was calibrated with five pseudo-steady states representing different settleability situations, as characterized by sludge volume indices. This "state by state" approach allowed the following conclusions to be drawn on the bulking phenomenon observed at the mill: (1) The transformations of organic biodegradable soluble nitrogen and phosphorus into readily available forms (ammonia nitrogen and ortho-phosphates) through the ammonification and phosphatification processes were found to be less significant when the activated sludge system was experiencing bulking; (2) The inhibition of the phosphatification and ammonification processes played an important role in stressing the system and in bulking; (3) Oxygen transfer (as the standard oxygen transfer efficiency) seemed impaired as the system was drifting towards a bulking state.;The general objective of these steps was to better understand filamentous bulking, under different angles, and the information that each step brought was centralised into a systematic approach to predicting, troubleshooting and reducing filamentous bulking problems. This constitutes the hybrid approach taking advantage of mechanistic and empirical elements discussed above. This methodology is made of four phases: (1) Phase I---Data acquisition: the information collected in this phase allows the identification of what is already known and of what could possibly provide clues as to the causes and sources of bulking; (2) Phase II---Mill design and operation review: this phase is meant to evaluate whether the activated sludge system is consistently well operated and whether filamentous bulking is the result of periodic events of upsets or if it is related to operational or design deficiencies that may exist. Phase II is further divided into three phases, namely an evaluation of facility housekeeping, an evaluation of operation and an analysis of operating costs; (3) Phase III---Characterization and identification of filamentous bulking problems: the objective of this phase is to characterize and identify the causes and sources of filamentous bulking in the system under study. Phase III is further divided into a characterization of bulking and identification of factors contributing to bulking and a mechanistic identification of factors contributing to bulking, making use of the ASMP model developed in this thesis; (4) Phase IV---Recommendations and monitoring: in this phase, the findings from all previous phases are brought together for interpretation and to establish a set of recommendations to address the problem of bulking at the studied mill. These recommendations can include corrective measures such as process or operational changes or process or equipment modifications. In this final phase, a monitoring program can be initiated by implementing the warning signal system developed with neural networks. (Abstract shortened by UMI.);The project started with a prolonged stay at a pulp and paper mill during which the activated sludge process was monitored with measurements related to adenosine triphosphate (ATP) and the operating costs were evaluated. This analysis identified opportunities to reduce costs related to instability and quantified a bulking event in terms of costs. The daily monitoring of ATP-related parameters established an early warning signal for filamentous bulking and provided information on the health state of the biological system's biomass. It was found that this signal provided an indication of the onset of bulking 1 to 1,5 sludge ages before macroscopic (sludge volume indices) and microscopic (filament count) parameters could, particularly in terms of the capacity of filamentous bacteria to store readily biodegradable substrate under dynamic conditions. Monitoring the fbATP parameter identified the overflow of an upstream whitewater tank as a potential initiator of a bulking cycle.