The use of fluidized-bed bioreactors in waste and drinking water treatment has several advantages, the most significant of which is the specific removal rate, which is an order of magnitude higher than that of equivalent activated sludge processes. In this paper, the usual concept of nitrification-denitrification in separated units is replaced by a new concept in which the two processes are performed together in a single high-performance fluidized-bed. Based on the former nitrifying reactor, new equipment was designed. This reactor contained a fluidized bed with separated aerobic and anoxic sections for nitrification and denitri fication respectively. This was accomplished by the use of different-diameter sand as carrier material and appropriate reactor shape, recirculation, feed and aeration conditions. The reactor (20 L fluidized-bed volume) was operated for 3 months. It was fed with synthetic waste water (50 L/h) containing 25-40 mg NH4 +-N/L. Propionic acid and ethanol in a 1:4 ratio were used as the carbon source (2.3 g C/L) for deni trification, fed in at different points of the reactor. Ammonium removal reached 50%, while denitrification was 75%. The total nitrogen removal rate was 0.8-1.2 kg N/m 3.d. A new simple hydrostatic pressure method was used to monitor biofilm thickness in the fluidized bed. During the experiments the oxidation-reduc tion potential (ORP) was tested as a tool to monitor reactor performance; its use for the control of the process was found to be limited.

The use of fluidized-bed bioreactors in waste and drinking water treatment has several advantages, the most significant of which is the specific removal rate, which is an order of magnitude higher than that of equivalent activated sludge processes. In this paper, the usual concept of nitrification-denitrification in separated units is replaced by a new concept in which the two processes are performed together in a single high-performance fluidized-bed. Based on the former nitrifying reactor, new equipment was designed. This reactor contained a fluidized bed with separated aerobic and anoxic sections for nitrification and denitrification respectively. This was accomplished by the use of different-diameter sand as carrier material and appropriate reactor shape, recirculation, feed and aeration conditions. The reactor (20 L fluidized-bed volume) was operated for 3 months. It was fed with synthetic waste water (50 L/h) containing 25-40 mg NH4+-N/L. Propionic acid and ethanol in a 1:4 ratio were used as the carbon source (2.3 g C/L) for denitrification, fed in at different points of the reactor. Ammonium removal reached 50%, while denitrification was 75%. The total nitrogen removal rate was 0.8-1.2 kg N/m(3).d. A new simple hydrostatic pressure method was used to monitor biofilm thickness in the fluidized bed. During the experiments the oxidation-reduction potential (ORP) was tested as a tool to monitor reactor performance; its use for the control of the process was found to be limited.

A fluidized bed denitrifying reactor was run to examine the vertical segregation of sand particles on the basis of different biofilm coverage, so far neglected when modelling fluidized beds. The segregation was found to be significant and it can be directly correlated with the vertical hydrostatic pressure profile in the bed. A procedure was developed for the rapid determination of biofilm thickness from hydrostatic pressure data using a recently published method based on the use of the novel criteria “expansion coefficient” and “specific occupied volume”. A key feature of the procedure is the “particle content”, which can be calculated from particle characteristics and is correlated in this study with the hydrostatic pressure gradient. The method was verified by directly measuring biofilm thickness as a function of the vertical position in the bed. This way biofilm thickness can be calculated from a readily measurable hydrostatic pressure profile with an error of 0.04–0.06 mm. This error is believed to be due to N2 gas entrapment in the denitrifying biofilm and to the original inaccuracy of the determination of particle size and volume. The method is rather insensitive to the exact biofilm density when the usual high-density carrier material is used.

Biofilm processes are widely used; the most effective of these fluidized bed reactors, the design and modelling of which is still difficult. A fluidized sand bed denitrifying reactor was run to clarify contradictions in the applicability of the Richardson-Zaki model for the description of fluidization hydrodynamics and to determine fully the true biofilm parameters experimentally. Biofilm density was found to be 1.055 ±0.018 g/cm3. A fundamental parameter of the Richardson-Zaki model, the terminal settling Reynolds number, was found to be not characteristic of the particles. Consequently a new approach was developed for the description of biofilm-coated particle fluidization. The model is based on two new parameters: the ‘expansion coefficient’ and the ‘specific occupied particle volume at zero flow’, which are readily determinable and characteristic parameters of the fluidized particles, being independent of reactor size and shape, liquid velocity or of the quantity of carrier particles. The model is suitable for modelling bed porosity or biomass concentration as a function or the biofilm thickness and upflow liquid velocity. We found that there can be an optimal biofilm thickness above which not only can the diffusion limitation increase, but the overall biomass concentration decreases at a given liquid velocity.

AbstractChemInform is a weekly Abstracting Service, delivering concise information at a glance that was extracted from about 100 leading journals. To access a ChemInform Abstract of an article which was published elsewhere, please select a “Full Text” option. The original article is trackable via the “References” option.

The control of the granulation process seems to be a kind of microbial selection: the enrichment of Methanothrix sp. against the Methanosarcina sp. Because of a lower KS value of Methanothrix, keeping a low (below 200 mg.dm−3) acetate concentration has been advised to be beneficial for granulation. This method results a 70-100 days start-up time. This approach assumes Monod-type kinetics, although a substrate inhibition model may describe better the biomethanation of acetate. We found that the best fit was yielded by a Haldane-type equation modified by an inhibition response coefficient (n). The major difference between the kinetics of raw and granular sludges has been manifested in this dimensionless parameter. The n was 4 times higher for raw sludge (3.6-4.1) than for granular sludge (0.95-1.13) which means that the granular sludge (formed mainly by Methanothrix sp.) is less sensitive to substrate inhibition. Continuous UASB experiments gave a similar result: the n-value continuously decreased (from 2.3 to 0.2) following granule formation. On the basis of the above findings we developed a new strategy for granulation control which ensured fast (35-40 day) granulation on carbohydrate-containing wastewaters.

Several types of high organic matter pollutants containing (COD-range: 3-50 kg.m-3) industrial waste waters were treated in laboratory scale (1.2-23 dm3) sludge blanket (UASB) and UASB-fixed bed hybrid (UBF) reactors. In most cases higher than 80% of COD-removal efficiency has been attained. The CO2 content of the biogas developed was mainly influenced by the neutralization (base to acid) ratio related to feed pH.Cell immobilization by granule formation was considered as a change in microbial population: enrichment and aggregate formation of Methanotrix-like filamentous microorganisms. Based on physiological and physical indexes of microbial selection and with regard to the different sensitivities of microorganisms to substrate inhibition, a new start-up method was developed for rapid (40-45 days) granulation of raw digested sludge.

If the wastewater to be treated contains various forms of nitrogen, three biological treatment steps are required for nitrogen removal: (a) in a bio-oxidation step, organic nitrogen is anerobically broken down to ammonia nitrogen; (b) in a subsequent nitrification step, ammonia nitrogen in the wastewater is aerobically converted to nitrate nitrogen; and (c) in a final denitrification step, nitrate nitrogen is anaerobically or anoxically converted to nitrogen gas. This chapter discusses bio-oxidation, nitrification and denitrification process steps, their principles, and design considerations in detail.

In steady state, attained by continuous aeration after oxygen saturation of water in a bubble column, vertical composition distribution of liquid and gas phases has been determined. It has been assumed that, as a result of absorption at the bottom of the column, desorption in the upper section and vertical dispersion of dissolved oxygen flux, a closed oxygen circulation is created. Determination of the axial dispersion coefficient from hydrodynamic and oxygen transfer data verifies the mathematical model proposed. The results allow conclusions to be drawn about supersaturation and desorption and other phenomena expected in biological systems.

A continuous fixed-bed denitrification system operating with high capacity and 100% efficiency has been developed for laboratory scale experiments. The data required for stable operation were obtained by determination of the kinetic constants of the system.