Publicação
Image analysis in biotechnological processes: applications to wastewater treatment
| Resumo: | In this work four different studies were proposed with a common interest in image analysis meffiodologies and multivariable statistical techniques yet covering distinct objectives. In an activated sludge study, the aggregates and filamentous bacteria contents and morphology were surveyed by image analysis methodologies. The high Sludge Volume Index values denoted the existence of a severe bulking problem of non-zoogleal nature as pointed out by the predominance of normal flocs. Moreover, the high filamentous bacteria per suspended solids ratio values clearly indicated the existence of a filamentous bulking problem, and were able to follow, at some extent, the Sludge Volume Index behaviour. Furthermore, the Partial Least Squares analysis revealed a strong relationship between the Total Sus pended Solids and the Total Aggregates Arca, although it must be emphasized though that for a wastewater treatment plant working in good operating conditions this relationship may not stand frue. In a protozoa and metazoa identification work the main objective resided on the development of an image anaiysis programme to morphologicaily characterize the protozoa and metazoa and treat the coilected data by multivariable statistical techniques. The studied species attained a satisfactory overali recognition levei in terms of global recognition and misciassification performances, whereas for the main protozoa and metazoa groups as weil as for the ciliated protozoa groups, the results were quite good. Such was also the case for the plant conditions assessment as effluent quaiity, aeration, sludge age, and nitrification presence. However, the assessment of critical conditions such as low effluent quality, low aeration and fresh sludge, proved to be poorer. Comparing the two multivariable statistical techniques, the overall results were iower for the Neural Networks than for the Discriminant Analysis with the exception of the critical conditions assessment. Regarding the anaerobic granuiation process, image analysis methodologies were used to follow morphological changes in the granulation process. This survey allowed for the determination of an overall aggregates size and contents increase throughout the experiment as well as the establishment of the granulation time with the formation of granular structures. It was also possible to identify an initial stage involving the predominant growth of the filamentous bacteria followed by a second stage of aggregates growth using the filamentous bacteria as a backbone and a final stage of balanced filamentous bacteria and aggregates contents growth. Moreover, the strong Up-Flow Velocity and Organic Loading Rate increases led to disturbances within the reactor such as the liberation of filamentous bacteria and aggregates size changes. Concerning the granule deterioration triggered by oleic acid study, the results obtained for the outgoing effluent Volatile Sus pended Solids reflected a biomass wash-out phenomenon throughout the experiment. Furffiermore, the aggregates morphological survey allowed determining a decreasing trend in the aggregates size, as well as an aggregate stratification with the larger aggregates in the top section of the reactor. It could also be estabiished that the granule deterioration process triggered by oleic acid led to more freely dispersed structures in terms of filamentous bacteria and lighter aggregates which ultimately rose to the top of the reactor, where the lighter ones suffered a wash-out phenomenon. |
|---|---|
| Autores principais: | Amaral, A. L. |
| Ano: | 2003 |
| País: | Portugal |
| Tipo de documento: | tese de doutoramento |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade do Minho |
| Idioma: | inglês |
| Origem: | RepositóriUM - Universidade do Minho |
| Resumo: | In this work four different studies were proposed with a common interest in image analysis meffiodologies and multivariable statistical techniques yet covering distinct objectives. In an activated sludge study, the aggregates and filamentous bacteria contents and morphology were surveyed by image analysis methodologies. The high Sludge Volume Index values denoted the existence of a severe bulking problem of non-zoogleal nature as pointed out by the predominance of normal flocs. Moreover, the high filamentous bacteria per suspended solids ratio values clearly indicated the existence of a filamentous bulking problem, and were able to follow, at some extent, the Sludge Volume Index behaviour. Furthermore, the Partial Least Squares analysis revealed a strong relationship between the Total Sus pended Solids and the Total Aggregates Arca, although it must be emphasized though that for a wastewater treatment plant working in good operating conditions this relationship may not stand frue. In a protozoa and metazoa identification work the main objective resided on the development of an image anaiysis programme to morphologicaily characterize the protozoa and metazoa and treat the coilected data by multivariable statistical techniques. The studied species attained a satisfactory overali recognition levei in terms of global recognition and misciassification performances, whereas for the main protozoa and metazoa groups as weil as for the ciliated protozoa groups, the results were quite good. Such was also the case for the plant conditions assessment as effluent quaiity, aeration, sludge age, and nitrification presence. However, the assessment of critical conditions such as low effluent quality, low aeration and fresh sludge, proved to be poorer. Comparing the two multivariable statistical techniques, the overall results were iower for the Neural Networks than for the Discriminant Analysis with the exception of the critical conditions assessment. Regarding the anaerobic granuiation process, image analysis methodologies were used to follow morphological changes in the granulation process. This survey allowed for the determination of an overall aggregates size and contents increase throughout the experiment as well as the establishment of the granulation time with the formation of granular structures. It was also possible to identify an initial stage involving the predominant growth of the filamentous bacteria followed by a second stage of aggregates growth using the filamentous bacteria as a backbone and a final stage of balanced filamentous bacteria and aggregates contents growth. Moreover, the strong Up-Flow Velocity and Organic Loading Rate increases led to disturbances within the reactor such as the liberation of filamentous bacteria and aggregates size changes. Concerning the granule deterioration triggered by oleic acid study, the results obtained for the outgoing effluent Volatile Sus pended Solids reflected a biomass wash-out phenomenon throughout the experiment. Furffiermore, the aggregates morphological survey allowed determining a decreasing trend in the aggregates size, as well as an aggregate stratification with the larger aggregates in the top section of the reactor. It could also be estabiished that the granule deterioration process triggered by oleic acid led to more freely dispersed structures in terms of filamentous bacteria and lighter aggregates which ultimately rose to the top of the reactor, where the lighter ones suffered a wash-out phenomenon. |
|---|