Anaerobic bacterial aggregates Variety and variation

Chemical conversions that are mediated by micro-organisms exhibit, necessarily, an autocatalytic nature. In Chapter I, this point is elaborated, and it is shown that in continuous-flow systems, the maximum specific growth rate of micro—organisms sets an undesirably low limit to the rate at which a bioreactor may be operated. When bacteria can be grown in the form of clumps (called aggregates in this thesis) , it becomes possible to operate a continuous-flow reactor at far higher loadings, as such aggregates are less subjected to wash-out from the reactor, and may be retained with some efficiency. This hold-up or retention of biomass finds its physical basis in the settling characteristics of aggregates. Unfortunately, however, mathematical descriptions for such systems are rather simplistic and unsatisfactory, as knowledge on physical characteristics of aggregates is lacking. The research in this thesis therefore aimed at a description and understanding of the development of bacterial aggregates, with particular emphasis on physical characteristics. As a model process, anaerobic acidification by a mixed bacterial population was studied in a mineral medium with glucose as the growth-limiting carbon and energy source.
Chapter Il gives a description and justification of the anaerobic gas-lift reactor (AGLR) used in this research. A start-up routine employing a dilution-rate shift-up ensured the rapid formation of aggregates from freely suspended cells, provided carrier material (sand) was available.
Chapter Ill presents more detailed data on the complicated role of carrier material during the first stages of aggregate formation.
Chapter IV illustrates the development of aggregates with data from scanning and transmission electronmicroscopy, and from light microscopy. The final type of steady-state aggregate was found to have high cellular densities at its periphery of 1012 cells per milliliter, which occupied 25 % of the local volume. In the centre of large aggregates, however, a pronounced void with a low biomass concentration was observed. The results were summarized in a working hypothesis on aggregate dynamics, involving cellular lysis due to substrate insufficiency.
In Chapter V, data are presented on prolonged AGLR operation. Although aggregate formation in itself was accomplished rapidly, the stabilization of reactor operation was delayed to some extent. Rather large aggregates were found initially, but were seen to disappear from the AGLR once substrate became depleted. This loss in retention could be attributed to the loss in specific weight, and the concomitant decrease in settling velocity. After this shift in biomass hold —up, the reactor was operated for a considerable time span under stable conditions. The events during this rearrangement of aggregate characteristics could be explained in terms of the previously developed working hypothesis.
In Chapter VI, the previous findings on the development of aggregates are summarized in a mathematical form. A model is presented describing overall reactor characteristics in terms of the dynamic behaviour of individual aggregates. As a result of substrate consumption, aggregates are growing, and thus increase in size. On the other hand, substrate depletion in the centre of large aggregates causes an internal deterioration by cellular lysis. As a result, large aggregates desintegrate into smaller ones. Overall biomass retention was formulated in terms of the settling velocity of individual aggregates. On the basis of these highly mechanistic formulations, retention was found to depend on the dilution rate, but not on the influent substrate concentration. Reactor biomass concentrations exhibited a complex pattern, depending on the choise of the fragmentation pattern. The importance of the result resides in the possiblity to predict reactor biomass concentration and retention without taking recourse to commonly used, but unattractive, assumptions on the absence of bacterial growth and the uniformity of aggregates.