Vaccari D.A., Strom P.F., Alleman J.E. Environmental Biology for Engineers and Scientists

Подождите немного. Документ загружается.

towers, where the attached bioﬁlm is extremely thin and tightly bound, these reactors may

not even require a fol low-up clariﬁer.

The typical hardware components used with a conventional system (Figure 16.8)

include a reactor ﬁlled with media, an underdrain platform at the base of the reactor on

which the medium rests, a rotating inﬂuent distribution arm (or, less commonly, a collec-

tion of ﬁxed spray distributors), and the necessary valves, pipes, and pumps required to

route, and possibly recirculate, this ﬂow through the reactor. In addition, there may be a

downstream settling tank to clarify the outgoing waste stream prior to discharge, in which

solids pulled or released (i.e., sloughed) from the bioﬁlm surface can be settled.

As for the design strategies used by environmental engineers to conﬁgure and size

these processes, hydraulic and substrate loadings typically represent the key parameters,

as depicted in Figure 16.9. Table 16.2 provides an overview of the associated design

Influent

Effluent

Recycle

Waste Solids

Rotating Distributing Arm

Packed or Random Media

Clarifier

Attached Growth Reactor

Underdrain Support Platform

Clarifier

Figure 16.8 Attached-growth process schematic.

LOADING CRITERIA

- Hydraulic Loading

- Organic Loading (per volume)

Area

cross-

Area

media

= V*Area

specific-media-surface-area

R = Radius

Volume

reactor

= Area

cross-section

Area

cross-section

π*R

H = Height

BOD

eff

BOD

Recycle

BOD

Volume

eff

Area

cross

section

BOD

Volume

reactor

Effluent

eff

Influent

Figure 16.9 Attached-growth process design considerations.

588

BIOLOGICAL APPLICATIONS FOR ENVIRONMENTAL CONTROL

values typically used for attached-growth systems, as well as information regarding

expected reactor depths and recirculation ratios. Hydraulic loading is a primary design

factor based on the ﬂow of inﬂuent wastewater (which may or may not include recycle

ﬂow) applied per cross-sectional surface area of the medium (with typical units of,

/daym

, or when simpliﬁed, m/day). One of the more subtle aspects of these systems

is that bioﬁlm surfaces will naturally retard the movement of liquid across a media’s sur-

face in proportion to the amount of dynamic drag imparted by the physical roughn ess of

its liquid–bioﬁlm interface. As a result, for any given system, an applied total hydraulic

loading (THL), inclusive of inﬂuent plus recycle ﬂow, coul d well result in considerably

different operating values for the amount of time the ﬂow actually spends in transit

through the reactor, as would be quantiﬁed using a hydraulic residence time (HRT).

Without an attached bioﬁlm, wastewater ﬂow over smooth, clean media surfaces would

experience little hydraulic resistance, and in turn would have a low hydraulic residence

time. Conversely, mature bioﬁlms have an inherent degree of surface roughness associated

with microbes extending physically into the bulk ﬂuid layer, which would then slow down

the ﬂow and increase the HRT. By analogy, a pail of water splashed across a ﬂoor will

travel much faster across smo oth and recently waxed tile (i.e., a clean medium) rather

than a ﬁbrous carpet (i.e., a highly nonhomogeneous microbial surface). As a result,

real-world observations of attached-growth systems have shown a considerable difference

in hydraulic residence times at equivalent THL loadings, by roughly a factor of 10 (e.g.,

1 minute for 3 m of clean media vs. 10 minutes when passed over mature bioﬁlm)

(CH2M-Hill, 1984).

Attached-growth trickling ﬁlter systems used with wastewater treatment implicitly

beneﬁt from this phenomena, since it gives the bioﬁlm more time to effectively treat an

overﬂowing waste stream. Granted, compared to the sort of hydraulic residence times

used with suspended-growth wastewater treatment (e.g., measured in hours) or lagoons

and wetlands (e.g., measured in days, weeks, and possibly even months) , bioﬁlm systems

have far lower contact times (e.g., 10 to 20 minutes) in which the desired treatment

can be completed. On the other hand, this sort of bioﬁlm growth and surface roughness

would be considered problematic in many other instances of attached-growth formation,

including that of cooling towers, ship hulls, or your own teeth. In the particular case of

TABLE 16.2 Typical Attached-Growth System Design Criteria

Fixed-Film

Design Factor

Hydraulic

Loading Range

/daym

)

Substrate

Range

(kg/m

day)

Media

Depth

(m)

Recirculation

Ratio Q

=Q,

Relative to

Inﬂuent Flow

Low-rate

bioﬁlm system

1.2–3.6 0.08–0.4

(as CBOD)

2–2.5 0

High-rate

(‘‘roughing’’)

bioﬁlm system

48–150 1.6–8

(as CBOD)

5–12 0–1

Nitriﬁcation

bioﬁlm system

1.2–4.8 0.0015–0.002

(as N)

3–12 0–1

Source: Adapted from Metcalf and Eddy (2003).

WASTEWATER TREATMENT

589

cooling tower operations, for instance, bioﬁlm growth on heat transfer surfaces would

reduce the overﬂow velocity while increasing the depth and temperature of the overlying

water stream, such that the tower would then have a lower thermal gradient and slower

rate of heat release.

As for the second, substrate loading, design parameter, this variable considers the mass

of substrate applied per unit time and per unit volume of the media (typically, kg sub-

strate/daym

). The magnitude of these terms, of course, varies in relation to the type

of substrate and the degradation rates expected. For example, even the low end of the

organic loading rates typically used (e.g., 0.08 kg CBOD

/daym

) is far higher than

that of nitriﬁcation rates (e.g., usually less than 0.002 kg ammonia-nitrogen/daym

)

(Characklis and Marshall, 1990).

Example 16.1: Preliminary Trickling Filter Design A newly proposed suburban commu-

nity near Chicago, Illinois (i.e., Deer Creek, with 400 homes in a standard three-bedroom

conﬁguration) will require a dedicated wastewater treatment facility given the remote,

unsewered location. During a preliminary public utilities hearing, the community’s con-

sulting engineer is asked to present a preliminary design est imate for sizing a standard

attached-growth trickling ﬁlter, recommended as one possible option for the biological

secondary stage in this proposed system. With an expected per-home family size of

four and a projected average per-person wastewater discharge rate of 378.5 L/day (i.e.,

100 gal/day-person), the daily wastewater ﬂow for the entire community would be

605,700 L/day [(150 homes) (4 people/home) (378.5 L/person day)] or 605.7 m

/day.

In turn, and given that the carbonaceous organic waste [i.e., typically quantiﬁed as a car-

bonaceous biochemical oxygen demand (CBOD)] in this sort of municipal wastewater is

typically around 200 mg CBOD/L, the daily organic wastewater loading generated by this

community would be 121.2 kg CBOD/day [i.e., (605.7 m

/day) (0.2 g CBOD/L)], of

which approximately 40% (or 48.5 kg BOD/day) would be removed by primary treat-

ment, leaving a ﬁnal CBOD load on this trickling ﬁlter system of 72.7 kg CBOD/day.

Preliminary Design Details

Hydraulic loading,

system sizing, and

retention times

Design hydraulic loading rate: 8 m

day (see Table 16.2)

Design media speciﬁc surface area: 98.4 m

day (see

Table 16.2)

Design media depth: 3 m (see Table 16.2)

Cross-sectional tower area: (605.7 m

/day)/(8 m

day) ¼

75.7 m

Circular tower diameter : [(4) (75.7 m

)/p]

0.5

¼ 9.82 m

diameter

Projected tower volume: (3 m media depth) (75.7 m

) ¼

227.13 m

Projected total media surface area: (98.4 m

)(227.13m

) ¼

22,350 m

Projected average liquid trickling ﬁlm depth: 0.2 mm

Projected total tower liquid volume: (0.2 mm) (22,350 m

) ¼

4.47 m

Projected hydraulic retention time: (4.47 m

)/(605.7 m

/day) ¼

10.6 min

590 BIOLOGICAL APPLICATIONS FOR ENVIRONMENTAL CO NTROL

Biomass values Estimated active biomass depth: 0.15 mm (or 150 mm)

Estimated total biomass depth: 1 mm

Estimated biomass density at 5% solids ¼ 50 kg/m

Estimated active biomass mass:

ð0:15 mmÞð22;350 m

Þð50 kg=m

Þ¼167:7kg

Estimated total biomass mass:

ð1mmÞð 22;350 m

Þð50=kg=m

Þ¼1118 kg

Organic loading Projected volumetric organic loading:

ð72:7 kg CBOD=dayÞ=ð227:13 m

Þ¼0:32 kg=m

 dayÞ

Related note Although this determination was simplistically developed for

a single tower, multiple towers are typically designed by

engineers to provide beneﬁcial system redundancy. In simi-

lar fashion, while this design estimate was based on an aver-

age daily ﬂow, engineering designs are typically based on

higher likely ﬂows, such as the facility’s daily peak ﬂow,

in order to physically accommodate these types of hydraulic

transient events.

The third and fourth design factors listed in Table 16.2, which are somewhat secodary

in importance relative to the preceding loading terms, refer to the typical depths and levels

of recirculation provided with these ﬁxed-ﬁlm systems. Older, rock-media designs were

seldom built with media depths beyond 2 to 3 m, but many newer, plastic-media units are

considerably taller (with depths up to 10þ m) and are designed to handle much higher

loads. As for the notion of incorporating a recycle stream, this capability is considered

necessary with inﬂuent ﬂows that tend to have extreme lows (e.g., smaller towns, day-

time-only industrial operations), such that the bioﬁlm can be kept wetted during periods

with inﬂuent lulls rather than having it dry out.

Extending beyond these heuristic design criteria, there are also a number of empirical

models used to qualify the reac tion-rate kinetics and expected reactor performance levels

with attached-growth reactors. These models, typically named in honor of the responsi ble

author or group, include the following variations: Eckenfelder, Rose, NRC, Howland,

Schulze, Germain, Velz, and modiﬁed-Velz. Each of these models is considered to have

its own merits, but the modiﬁed-Velz model, given in equation (16.1), is one of the more

widely used strategies. As is the case with many of these competing models, the modiﬁed-

Velz model follows a ﬁrst-order format:

out

expðkD

T20

THL

ð16:1Þ

where

out

¼ outlet substrate concentration ðmg=LÞ

¼ inlet substrate concentration ðmg=LÞ

k ¼ empirical reaction-rate constant

WASTEWATER TREATMENT 591

D ¼ media depth ðmÞ



T20

¼ temperature correction factor ðrelative to 20



CÞ

THL ¼ hydraulic loading ðbased on influent plus recycle flowÞ

n ¼ emp irical loading constant

Figure 16.10 accordingly presents a representative substrate removal proﬁle derived

with this model, as might be derived for a particular range of media heights relative to

a given set of constant values for k, n, and , In turn, this type of model does provide

an informat ive characterization of bioﬁlm performance, at least on a qualitat ive basis.

However, on closer review, it would also appear that these types of models bear an inher-

ent degree of fuzziness on several important issues.

The parameters of the modiﬁed Velz equation are sometimes reported with high pre-

cision (e.g., n ¼ 0:4332; CH2M-Hill, 2000). However, it should be recognized that these

types of documented constants reﬂect cumulative, averaged values derived from real-

world bioﬁlms whose actual site- and point-speciﬁc metabolic activities undoubtedly

vary to a signiﬁcant degree, depending on time, location, season, and so on. Indeed,

these constants could well vary from reactor top vs. reactor bottom locations, and even

from one substrata depth within a bioﬁlm to another.

Yet another level of ‘‘fuzziness’’ exists with the total hydraulic loading rate (THL). The

THL depends on inﬂuent ﬂow and cross-sectional media area, but it does not actually

account for the manner in which the inﬂuent is applied. For example, two parallel systems

of equal design, construction, and hydra ulic loading could have identical THLs, yet have

sizably different levels of performance. In this case, the issue of ‘‘how’’ the inﬂuent is

applied represents an important factor, as opposed to simply ‘‘how much,’’ relative to

both the rate of movement of the rotating inﬂuent distributor and the corresponding pul-

sing of ﬂow across the media. With each successive radial pass, this rotating arm distri-

butes a liquid stream whose downward passage across the media appears as a moving

wave whose cyclic repetition follows that of the arm’s travel speed. This pulsing and

Attached-Growth BiotowerDepth (m)

12020 40 60 80 100

Substrate (CBOD

Attached-Growth Biotower Depth (m)

12020 40 60 80 100

Substrate (CBOD

) Concentration (mg/L)

Figure 16.10 Representative pattern of exponential substrate removal relative to attached-growth

biotower depth.

592

BIOLOGICAL APPLICATIONS FOR ENVIRONMENTAL CONTROL

wavelike movement of liquid into and across the depth of the media would have an amplitude

(i.e., height of the wave) that increased as the rotating arm was slowed. In turn, higher-

amplitude waves would impose higher levels of physical shear and drag against the face

of the bioﬁlm, to an extent that would probably yield thinner attached-growth depths.

Undoubtedly, though, one of our largest levels of uncertainty with understanding

attached-growth systems is our appreciation of the bioﬁlm complexity in terms of biolo-

gical makeup, physical conformation, and related metabolic behavior. Attached-growth

system models akin to the modiﬁed-Velz equation implicitly suggest that bioﬁlms are

built and behave as homogeneous surfaces, with uniform levels of depth, kinetic activity,

and structural composition, but this is hardly the case. Instead, bioﬁlms comprise a highly

nonhomogeneous, three-dimensional matrix with a decidedly complex physical structure.

In short, attached-growth bioﬁlms are conﬁgured as a densely arrayed microcosm of life,

composed largely of bacteria but with several higher life-forms, including protozoans,

rotifers, and nematodes, and possibly even fungus and insect forms as well.

This complexity within bioﬁlm structures starts at an early stage during an initial per-

iod of startup and adhesion, where a preliminary layering of bacterial cells (i.e., a mono-

layer) binds to a clean med ia surface within a relatively short period of time (i.e., minutes

to hours). This adhesive phenomenon stems from the presence of a sticky coating of exo-

cellular polymeric materials secreted by bacterial cells growing under conditions of lim-

ited substrate availability, thereby allowing them to bind not only to media surface s but

also to other cells. In time, this monolayer coverage subsequently extends in thickness,

progressively melding new and old cells as well as various enmeshed inorganic precipi-

tates and particulate solids, to a point where the monolayer eventually reac hes depths ran-

ging from a few hundred to several thousand micrometers.

An apt analogy for mature bioﬁlm growth is that of a forest canopy, with overarching

branched limbs (newly grown cells) stretched across open glades (internal cavities gouged

out by ongoing microbial sloughing and ﬂuid shear) and underlying brush (deeper depos-

its of dead or inert cells). However, even then, this canopy will be apt to collect an added

assortment of intertwined solids and cells, including a variety of higher life-forms living

within the bioﬁlm as well as entrapped wastewater solids. Given the chem ical complexity

of this ﬁlm, yet another group of enmeshed, precipitated solids could well be wrapped into

this matrix, including such materials as sulﬁde-, hydroxide-, phosphate-, or carbonate-

based precipitates of iron, manganese, calcium, magnesium, and aluminum.

Of course, commensurate with this increased microbial depth, substrate and product

transport through the ﬁlm also becomes considerably more constrained, to a point

where most bioﬁlms eventually take on a vertically stratiﬁed layering. For those micro-

organisms living on the outer, aerobic edge of this bioﬁlm (depicted schematically in

Figure 16.11), their proximity to energy-rich substrates, nutrients, and oxygen received

from the overlying bulk solution will allow them to reach the highest rates of aerobic

metabolic activity. Figures 16.6 and 16.7 depict the type of bacterial growth found at

this topmost, aero bic layer, including the considerable presence of ﬁlamentous forms,

which give the bioﬁlm its inherent outer roughness.

In particular, Figures 16.12 and 16.13 depict progressive enlargements of an extensive

overlying growth of a ﬁlamentous Beggiatoa-type growth. Organisms such as this are

often found living at this top-level bioﬁlm region, catabolically using sulﬁde diffusing

to the surface from the underlying anoxic–anaerobic (i. e., sulfate-reducing) zone. Strictly

aerobic higher life-forms (e.g., most protozoans, rotifers, worms) are also commonly seen

in this outer region, grazing steadily through this relatively active region.

WASTEWATER TREATMENT 593

Moving downward into the bioﬁlm, however, metabolic uptake will progressively

deplete the available substrates and nutrients , imposing metabolic limitations that progres-

sively retard the activity of these lower cells. Although no exact threshold can be pre-

dicted for the point at which the oxygen will be exhausted, interstitial measurements

taken with microscale probes suggest that this point will occur roughly in the neighbor-

hood of 150 mm.

Anaerobic sub -layer

(~1000-5000 microns)

Wastewater

stream

across

biofilm

face

Anoxic-anaerobic

sub-layer

(~typically

1000-5000 µm

depth)

Aerobic top layer

(

~100

150

Anaerobic sub -layer

(~1000-5000 microns)

Anaerobic sub -layer

(~1000-5000 microns)

Wastewater

stream

across

biofilm

face

Attachment media

Anoxic-anaerobic

sub-layer

(~typically

1000-5000 µm

depth)

Aerobic top layer

(~100-150 µm depth)

Figure 16.11 Bioﬁlm layering (aerobic top vs. anaerobic bottom) schematic.

Figure 16.12 Highly ﬁlamentous bioﬁlm morphology at top surface (800).

594

BIOLOGICAL APPLICATIONS FOR ENVIRONMENTAL CONTROL

Moving deeper into the bioﬁlm, therefore, the aero bic metabolic viability of the cells

will decline as a direct consequence o f diminished substrate and oxygen availability.

Oxygen, in particular, declines steadily in availability as depth increases, to the point

where it is depleted and anaerobic conditions begin. One such zone of transition from

an overlying aero bic top layer to an underlying anoxic and/or anaerobic zone is depicted

in Figures 16.14 and 16.15.

With oxygen reaching limiting conditions within the subsurface bioﬁlm strata, anoxic

respiration will accordingly deplete whatever alternative electron acceptors (i.e., nitrates,

nitrites, etc.) might be present, eventually leading to sulfate reduction and a metabolic

release of sulﬁdes. Tucked just below the overlying Beggiatoa ﬁlaments shown in

Figures 16.14 and 16.15 is a distinctly different type of bacteria, Desulfovibrio, which

Figure 16.13 Bioﬁlm surface close-up with dense Beggiatoa-type ﬁlaments (2000).

Figure 16.14 Bioﬁlm transition zone between aerobic top and anaerobic bottom (2000).

WASTEWATER TREATMENT 595

is responsible for this sulfate reduction. These lower cells are more tightly packed and

have a characteristic comma-shaped curvature. The last photograph in this series,

shown in Figure 16.16, provides a close-up look at this group of cells.

The sulﬁdes produced at this lower level will often scavenge and precipitate a variety

of metals (probably dominated by iron sulﬁde, FeS, but undoubtedly including many other

metal species), leading to a sort of cementation behavior that further inhibits chemical

transport at these lower depths. Whereas optimal activity would be achieved with limited

bioﬁlm depths, therefore, the aging process experienced with bioﬁlms leads to progressive

metabolic constraints at the lower depths. Burie d at a depth that limits their access to key

Figure 16.15 Bioﬁlm transition zone (aerobic to anaerobic) close-up ( 4000).

Figure 16.16 Bioﬁlm anaerobic bottom zone close-up with Desulfovibrio-type bacteria (4000).

596

BIOLOGICAL APPLICATIONS FOR ENVIRONMENTAL CONTROL

substrates and nutrients, as well as hampered by transport problems brought about by

entrapped and precipitated solids, the underlying cells must inevitably shift to anaerobic

(or at least anoxic) life-styles.

However, when viewed in terms of microbial diversity and potential treatment efﬁcacy,

the circumstance of having adjacently layered aerobic and anaerobic strata within a bio-

ﬁlm does appear to provide a potentially beneﬁcial arrangemen t. Indeed, these strata

could well complement one another, such as a reductive transformation of what might

otherwise comprise recalcitrant organics (e.g., chlorophenolics, nitrophenolics, aniline

dyes) within the anaerobic sublayer, followed by a concluding oxidation within the over-

lying aerobic layer.

At least in theory, it would also appear that it might hypothetically be possible to

develop a beneﬁcially layered sequence of nitrifying and denitrifying bioﬁlms, such

that nitrates produced in the oxidative overlying strata would then be directly inside

reduced the lower anoxic–anaerobic depth. However, this conceptual scheme would be

difﬁcult to attain since the slow-growing nitriﬁers are prone to overgrowth and submer-

gence by faster multiplying heterotrophs, at which point competition for oxygen and other

essential nutrients severely constrains their viability. As a result, nitrifying bioﬁlm sys-

tems designed for oxidation of wastewater ammonia-nitrogen are usually conﬁgured as

stand-alone processes, supplied with a pretreated wastewater whose biodegradable

organic levels available for heterotrophic growth have been largely removed, and pro-

vided with specially designed media offering a far higher speciﬁc surface area more con-

ducive to these slow-growing lithotrophs.

Once these attached-growth systems have been placed into use, though, and their mul-

tilayered bioﬁlm conﬁgurations have been established, operating personnel have surpris-

ingly few monitoring and control options with which they might try to improve or

optimize performance. Merely measuring the attached mass or depth of bioﬁlm within

a reactor is quite difﬁcult, although in rather rare instances, provisions have been made

to track bioﬁlm mass using removable sample coupons removed from inside the reactor.

Other than shifting reactor loading patterns by taking bioﬁlm towers off- or online, the

most signiﬁcant control factor is probably that of altering the rotation speed for the dis-

tributors. In recent years, there has been a distinct shift to slow down the rotational speed

of these distributing arms, dropping the number of rotations per minute (rpm) from values

of 1 to 2 by a factor of 10 or more (i.e., to 0.1 or even lower rpm rates). Slowing down

these distributing arms while maintaining the same inﬂuent waste ﬂow means that the

hydraulic loadings applied at the point of release are considerably higher. This increase

subsequently raises the level of hydraulic shear as it moves downward across the face of

the bioﬁlm, and the added shear helps to keep the bioﬁlm thinner and more active than

would be the case with a much slower-moving water ﬁlm.

Notwithstanding these operational constraints with control and monitoring, though,

several sign iﬁcant developments have occurred with attached-growth media materials

and conﬁgurations over the past several decades. In the realm of media options, rock

media largely gave way to lighter plastic media with far higher speciﬁc surface areas mid-

way through the twentieth century, and today’s vendor options include a wide range of

modular and random-packed versions. Table 16.3 provides speciﬁc technical details

regarding several of these newer media forms com pared against the older rock option.

These newer plastic media provide several times more surface area per unit volume,

which in turn yields considerable improvements in bioﬁlm growth and loading capacities.

In addition, plastic media are distinct ly lighter and have far higher void fractions (i.e., the

WASTEWATER TREATMENT 597