2 Biofilm

Figure 18.1 Schematic Diagram of an FBBR.

bioparticles develop without the presence of a carrier particle. Most FBBRs are two-phase systems, containing only water and bioparticles, and if oxygen is required it is dissolved in the recirculation flow prior to its return to the reactor. However, recent advances in system design have allowed the incorporation of a gas phase, thereby allowing oxygen transfer to occur directly in the bioreactor.'* Although the popularity of three-phase systems is increasing, they are considerably more complicated to model than two-phase systems, and thus, our discussion is limited to the latter. It should be noted that the designation of two-phase and three-phase is made with regard to reactants. Some so-called two-phase systems, such as denitrifying and methanogenic systems, actually have a gas phase in them because of the gas produced by the biological reactions. Nevertheless, they can generally be considered to be two phase for modeling purposes as long as the gas flow rate is small relative to the liquid flow rate.'2 This restriction may not be met for systems that are very heavily loaded, however, and care should be exercised in applying two-phase models to such FBBRs.

Because the bioparticles are retained in the reactor, the effluent from an FBBR often contains a sufficiently low suspended solids concentration to allow its discharge without clarification. Maintenance of the appropriate velocity to achieve the desired degree of suspension usually requires recirculation of bioreactor effluent. As biomass grows, the bioparticles become larger, causing the bed to expand in height. To prevent uncontrolled bed expansion, leading to loss of the bioparticles in the effluent, they are usually removed in a systematic manner to maintain a desired bed height. If the bioparticles contain a carrier particle, the excess biomass is removed in a separator, allowing the carrier particles to be returned. In this way a constant quantity of biomass can be kept in the system while maintaining an effluent low in suspended solids.

In general, FBBRs can be divided into two categories, depending on the nature of the bioparticles.' Tower bioreactors are those in which the bioparticles are composed entirely of biomass, without a carrier particle at the center, whereas supported-film bioreactors are those in which the biomass grows as a film on a carrier particle like sand, anthracite, or activated carbon. The sorptive properties of activated carbon provide distinct advantages in some cases, but complicate the analysis of FBBRs using it as the carrier particles. Consequently, its use is not discussed in this chapter. Rather, the reader should consult other sources for more information.'1 Distinction between the two types of FBBRs is necessary because the presence or absence of a carrier particle has a strong influence on the way the bioparticles behave as they grow larger, as we will see later. The upflow anaerobic sludge blanket (UASB) bioreactor is an important example of a tower bioreactor. In it the bioparticles grow as small spherical granules containing the complex microbial community associated with methanogenic systems. Most other FBBRs are supported-film bioreactors, and in fact, the two terms are usually used synonymously, as we will do here. Our discussion here is limited to supported-film bioreactors.

The main advantage of FBBRs over other attached growth bioreactors is that the small size of the carrier particles provides a very large specific surface area for biomass growth. Whereas the media in packed towers and rotating disk bioreactors have specific surface areas on the order of 100 rnVm' (see Table 19.2 and Section 20.1.1), the specific surface area provided by typical carrier particles in an FBBR is on the order of 1,000 to 3,000. " This allows the maintenance of very high biomass concentrations, ranging from 15,000 mg/L in acrobic FBBRs to 40,000 mg/L in anoxic ones. " s This, in turn, allows very short hydraulic residence times (HRTs) to be used, often on the order of minutes. While packed towers could theoretically contain media of similar size, the downward flow of liquid would cause excessive pressure drop and be prone to clogging as biomass growth occurred. It would also be subject to plugging through entrapment of suspended solids that might enter in the influent. Such solids can pass through FBBRs, however, because of the open structure of the fluidized bed. More information about the characteristics of these unique bioreactors can be found elsewhere.' '

18.1.2 Nature of the Biofilm

In Chapter 15, we discussed the concept of a steady-state biofilm. In an FBBR, development of a steady-state biofilm would require the excess biomass to be continually removed from the bioparticle surface and carried away in the effluent. It would also require the bed height to be sufficiently large to accommodate the quantity of biomass associated with the steady-state biofilm. Neither of these requirements is particularly desirable in practice. As noted above, one advantage of the FBBR is that it can be operated in a way that eliminates the need for a final clarifier. This would not be possible if biomass were constantly being sheared from the bioparticles in the FBBR. Furthermore, the thickness associated with a steady-state biofilm is likely to exceed the active thickness of the biofilm,1 i.e., the depth to which reactants penetrate. This means that the bed height and volume associated with a steady-state biofilm would be greater than that required to achieve the desired effluent quality with a fully active biofilm. Consequently, one characteristic of most FBBRs is the continual wastage of biomass from the top of the bed to maintain a constant bed height less than that associated with a steady-state biofilm. Another complicating factor is that wastage occurs from the top of the bed because that is where the bioparticles with the thickest biofilm reside. After the biomass on the removed bio-particles has been reduced by subjecting them to surface shear forces, the carrier particles are returned to the bed where they again serve as a support for biomass growth. Initially, the carrier particles fall to the bottom of the bed, but they migrate upward as biofilm builds up on them. (The reason for this behavior is explained later.) Consequently, the biofilm thickness on any individual bioparticle is continually changing, which differs from the assumptions associated with steady-state biofilms. Nevertheless, many models of FBBRs assume the existence of a steady-state biofilm to reduce computational complexity. Although such models are very useful for understanding the major factors influencing the behavior of FBBRs, it should be recognized that they differ significantly from the characteristics of most operating FBBRs.

One interesting attribute of FBBRs is that the dry density of the biofilm on a bioparticle depends on the thickness of that biofilm."2:1 :s Biofilm dry density is defined as the attached dry biomass per unit wet biofilm volume and is the same as the biomass concentration in a biofilm, XHIM, used in Chapter 15. However, the term dry density, and its associated symbol, pw, is used here to make clear the distinction between the amount of biomass per unit volume of biofilm and the amount per unit volume of bioreactor, X(!. The exact relationship between density and thickness varies from study to study, but Figure 18.2 s shows one that has been used in FBBR modeling. An important characteristic is that dry densities are very high (on the order of

Figure 18.2 Effect of biofilm thickness, L,, on the dry density, p,,„ of a denitrifying biofilm. (From W. K. Shieh and J. D. Keenan, Fluidized bed biofilm reactor for wastewater treatment. Advances in Biochemical Engineering!Biotechnology 33:131-169, 1986. Copyright © Springer-Verlag New York, Inc.; reprinted with permission.)

Figure 18.2 Effect of biofilm thickness, L,, on the dry density, p,,„ of a denitrifying biofilm. (From W. K. Shieh and J. D. Keenan, Fluidized bed biofilm reactor for wastewater treatment. Advances in Biochemical Engineering!Biotechnology 33:131-169, 1986. Copyright © Springer-Verlag New York, Inc.; reprinted with permission.)

70 g/L) for thin biofilms (on the order of 200 |xm or less), but decrease markedly as the biofilm grows thicker. Although the reason for this behavior is poorly understood, it is probably related to the activity of the biofilm. Thin biofilms tend to be fully penetrated by both electron donor and electron acceptor, whereas the interior of thick biofilms is devoid of one or both of these substances. The resulting environment leads to microbial reactions that can decrease the quantity of viable microbial cells, thereby decreasing the density. Another possibility is that the hydrody-namic conditions that lead to thin biofilms bring about a morphological change in the biofilm that make it denser." Regardless of the mechanism, however, it is clear from this behavior that thicker biofilms do not necessarily lead to more active biomass. This is one reason that FBBRs are commonly operated to give thin biofilms.

The hydrodynamic conditions required to give a thin biofilm are fairly complex and somewhat counterintuitive. Thinner biofilms develop in systems that have higher first-order detachment rate coefficients, b,,. While at first glance it might appear that b|, will increase whenever the upward velocity of the fluid (superficial velocity) is increased, this is not the case. Generally, the superficial velocities used in FBBRs result in low Reynolds numbers (<10), which means that surface shear is likely to be small.:s Lower superficial velocities, however, result in a smaller degree of bed expansion, which means that there is a higher probability of collisions among particles. The attrition caused by those collisions has a larger effect on the detachment coefficient than the fluid velocity past the biofilm surface.1'Consequently, higher values of b,, have been observed at lower superficial velocities." Empirical models are available that account for the various factors influencing the detachment rate coefficient."

Was this article helpful?

0 0

Post a comment