-------
sq ft) for studies with the aluminum oxide media. Although the fluidized
bed system will save energy compared to activated sludge because no oxygen
is added to the reactor, it is clear that the energy savings could be negated
through excessive pumping requirements.
For any specific set of design parameters, Figure 4 and Equation 4
can be used to estimate the pumping energy requirements for fluidization.
As an example, consider a reactor containing 3.05 m (10 ft) of silica sand
of specific gravity 2.65 at e _ of 0.40. The head loss through the bed is
3.02 m (9.9 ft). If the design called for a 2-hour HRT and no recycle pumping
was contemplated, the particle sizes would need to be exceedingly small
(< 0.2 mm) as shown by the curves in Figure 3. Assuming a wire to water
pumping efficiency of 65 %, the energy expended to overcome the headless
through the bed (excluding the losses in the distribution system) would be
(3.02 m) (9.806 newton/kg) (1000 kg/cu m) = 0.0126 kwh/cu m (47.8 kwh/MG)
(3600 sec/hr) (1000 watt/kw) (0.65 --)
ef f
If the proposed design called for using sand particles of approximately 1 mm
size, the minimum fluidization velocity would increase to 30.5 m/hr (100 ft/
hr) and providing a 2-hour HRT in the 3.05 m (10 ft) bed would require that
the recycle:influent pumping ratio rise to greater than 19:1 to achieve
more than minimum bed expansion. In this case, the pumping requirements at
an overall efficiency of 65 percent would rise to 0.252 kwh/cu m (955 kwh/
MG) of wastewater treated, excluding the additional losses in the distributor
system. The distributor losses will vary with the type of distribution
system and flow rates chosen, and will probably add an additional 0.3 to 1.2 m
(1 to 4 ft) of head loss to the system.
It can be seen that the energy requirements for fluidized beds will be
determined by the HRT required, the size and specific gravity of the media
selected, and the extent to which the bacterial film characteristics alter
the particles behavior. For the thin films observed by Switzenbaum and
Jewell, silica sand particles of around 0.3 to 0.4 mm size should produce
acceptable fluidization characteristics and bacterial concentrations (Table 2)
and result in a headless of 8.5 to 13.7 m (28 to 45 ft) for a 2-hour HRT.
Pumping 3785 cu m/day (1 mgd) with a headless of 15.2 m (50 ft) requires
242 kwh/day at an overall efficiency of 65 percent, so the headloss for such
a system would be reasonable.
In actual practice it is not clear what flow control strategy would be
optimal for plant operation. One approach is to incorporate a flow equali-
zation basin ahead of the reactor to insure that it receives a relatively
constant hydraulic loading and a more uniform organic loading. This approach
will minimize the amount of recycle pumping required. Alternatively, it may
be mor.e desirable to pass the incoming flow directly through the system and
vary the recycle ratio as required for adequate bed expansion. If the ratio
of maximum flow to minimum flow and maximum organic concentration to minimum
concentration synchronously varied by 3:1, the organic mass loading would vary
by 9:1 during the day. The optimal combination of flow equalization, reactor
size and recycle rate can only be calculated when considered in conjunction
with the expansion characteristics of the media selected and the biological
15
-------
kinetic response of the attached growth.
Biological sludge production in anaerobic systems is substantially
less than in aerobic systems. McCarty (21) lists the following growth
constants and endogenous respiration rates:
Endogenous
Growth Respiration Rate
Waste Constant
Fatty Acid
Carbohydrate
Protein
0.054
0.240
0.076
"I
days
0.038
0.033
0*. 014
The combination of a low cell yield coupled with extremely long SRT's
in the reactor will lead to a very low net sludge production. This net
solids production may be low enough to obviate the need for final clarifiers
and still meet secondary effluent standards. The data presented by Jewell
suggest that the net sludge production is low enough so that the excess
solids can be discharged in the effluent. If gas bubble formation and subse-
quent attachment to the particles tends to float media from the reactor,
some stripping and final settling may be required. The settling velocity of
a 0.4 mm particle of 1.1 specific gravity is approximately 21.3 m/hr (70 ft/
hr) so settling these particles can be accomplished in clarifiers/settling
tanks with high overflow rates. Larger or more dense particles will, of
course, settle faster.
The expanded bed system investigated by Jewell was operated on primary
effluent. Hence there would still be primary sludge to be processed and
disposed of. A comparison of sludge quantities and thickening character-
istics between a primary plant and a secondary activated sludge plant
illustrates several potential advantages which may be realized by substituting
an anaerobic reactor for an activated sludge system. Table 3 lists sludge
quantities and volumes for municipal wastewater treatment plants containing
primary clarifiers followed by activated sludge systems which receive incoming
BOD and suspended solids concentrations of 200 mg/1 each for Case No. 1 or
concentrations of 250 mg/1 of each for Case No. 2. In both cases the primary
clarifier is assumed to provide 60 percent suspended solids removal and 35
percent BOD removal. The solids processing scenarios will depend upon the
size of the plant and the options for solids disposal. If the net solids
production in an anaerobic fluidized bed system with a high SRT was 0.075
g VSS/g BOD removed, then substitution of an anaerobic system for the activated
sludge process in the two cases in Table 3 would result in effluent suspended
solids of 12 or 15 mg/1. If the soluble effluent BOD was approximately 10
to 15 mg/1 with an anaerobic system, it would be possible to achieve secondary
effluent quality with no provision for solids capture from the anaerobic
reactor.
Primary sludge can be gravity thickened to around 9 percent solids and
is also easy to dewater with relatively low chemical conditioning dosages
and high solids yields (30). For a small plant, lime conditioning (or
stabilization if required) followed by vacuum filtration and landfilling
may be the least cost sludge disposal option. Where anaerobic digestion
16
-------
TABLE 3. SLUDGE QUANTITIES AND VOLUMES REQUIRING PROCESSING PER MILLION
GALLONS TREATED IN A TYPICAL ACTIVATED SLUDGE PLANT WITH PRIMARY
CLARIFICATION
Influent BOD5 and Suspended Solids, mg/1 of each
Primary Sludge, Ib
Secondary Sludge, Ib
Unthickened Primary Sludge ;
Volume at 4% Solids, gal
'Thickened Primary Sludge
Volume at 9% Solids, gal
Unthickened Secondary Sludge
Volume at 1% Solids, gal
Thickened Secondary Sludge
Volume at 3% Solids, gal
Thickened Combined Sludge
Volume at 5.5%, gal
Primary Sludge Volatile Solids, Ib
Activated Sludge Volatile Solids, Ib
Design Assumptions:
Primary Clarifier Solids Removal 60%
Primary Clarifier Sludge 65% Volatile Solids
Primary Clarifier BOD Removal 35%
Cell Yield 0.75 Ib. VSS/lb BOD5 Removed
Cell Decay 0.07 days ~1
Soluble Effluent 6005 3 mg/1
SRT 5 days
Effluent Suspended Solids 15 mg/1
Effluent Solids are 75% Volatile
Case
No. 1
200
1001
636
Case
No. 2
250
1251
837
3000
1334
7626
2542
3569.
651
4>7
3750
1667
10036
3345
4552
813
628
17
-------
is to be used for sludge stabilization, Table 3 illustrates that the volume
occupied by thickened primary sludge alone is substantially less than achieved
by gravity thickening of primary and activated sludges. The small sludge
volumes require smaller anaerobic digesters. Of course, in any real
design situation the advantages and cost of flotation thickening of secondary
sludge should also-be considered.
The oxygen required for the activated sludge systems in Table 3
operated at a 5-day SRT should be nearly the same as the influent BOD 's to
the aeration basin, i.e., 130 and 163 mg/1 for the two cases shown. If air
is input with 8 psig adiabatic compression, 70 percent efficiency of compress-
or and motor, and an aeration device with a 7 percent oxygen transfer
efficiency, the power requirements are 0.124 and 0.156 kwh/cu m (471 and 590
kwh/MG). A mechanical aerator with an oxygen transfer efficiency of 1.1
kg 0_/kwh (1.8 Ib 0?/Hp-hr) would require 449 kwh or 563 kwh for the higher
oxygen demand. When these values are compared to a fluidized bed reactor
it provides a rough measure of the pumping energy which can be expended in
the anaerobic system and still be competitive with activated sludge on the
basis of energy criteria.
If the activated sludge plants summarized in Table 3 employed anaerobic
digest-ion for the stabilization of the primary and secondary sludges, the
primary sludge volatile solids would comprise about 57 percent of the
total volatile solids loading to the digester. In this hypothetical example
roughly 50 percent of the influent degradable carbon, to the activated sludge
system would be oxidized and the remaining carbon removed would be transformed
into biological solids; ;..some would escape in the effluent. It is this
remaining transformed organic matter that is available for CH, production
in the anaerobic sludge digestion process. In contrast, if an anaerobic
fluidized bed were substituted for an activated sludge system the BOD removal
in a. system? operated with no sludge j was ting or effluent solids capture would
result from CH, formation entirely. ' For the two wastewaters characterized
in Table 3, there would be 130 or 163 mg/1 of BOD,, available for CH, formation
if an anaerobic fluidized bed were used in place of activated sludge. The
amount of CH, formed will be determined by the efficiency of waste utilization
and the net biological sludge production.
An interesting aspect of CH, formation with an anaerobic reactor is
shown by the solubility data in Table 4 (31).
TABLE 4. SOLUBILITY OF CARBON DIOXIDE AND METHANE GASES
Temperature
10
15
20
25
30
Solubility , mg/1
CH4
29.6
26.0
23.2
20.9
19. 0
co2
2318
1970
1688
1449
1257
Solubility as C, mg/1
H4 -C
22.2
19.5
17.4
15.7
14.3
co2-c
632
537
460
395
343
*When the pressure of the gas plus that of the water vapor is 760 mm Hg
18
-------
In contrast to an anaerobic sludge digester where the high sludge feed
concentrations make the amount of CH, exiting in solution negligible in
comparison to the amount which is recovered in the overlying gas phase, the
amount of CH, which leaves the reactor in a dissolved phase from an anaerobic
fluidized bed reactor can represent a substantial part of the CH, formed.
Methane production from the anaerobic decomposition of any organic
compound can be accurately predicted by a number of techniques. Symons (32)
developed the following equation:
C H 0. + (n - a/4 - b/2)H00 * (n/2 - a/8 + b/4)C09 + (n/2 + a/8 - b/4)CH,
nab ^- ^ ^
Equal proportions of methane and CCL result from the decomposition of
carbohydrates and also from acetic acid. Proteins, fats and long chain acids
will yield gas compositions higher in CH. than CO . Typical municipal
wastewaters have total organic carbon concentrations (TOC's) in the primary
effluent of 80 to 180 mg/1. If 85 percent of this TOG were converted to
CO and CH, in an anaerobic system in the ratio of 40:60, the carbon in the
methane produced would range from 41 to 92 mg C/l. A comparison of these
values with the methane solubility data in Table 4 shows that in all cases
the quantity of methane produced which exits as dissolved methane gas must
be considered in any design situation where recovery of the methane from the
gaseous space overlying the reactor will be practiced. These data show
that the amount of CH, which remains dissolved in the liquid phase can be
a significant fraction of the total CH, production. Of course, the partial
pressure of the methane in the gaseous phase will affect the equilibrium
solubility concentration. Whether the dissolved CH, concentration will tend
toward the equilibrium concentration dictated by the overlying partial pres-
sure, remain near the saturation concentrations shown in Table 4, or be
somewhat supersaturated will be influenced by the reactor design, the hydraulic
residence time, and the degree of gas transfer across the gas-liquid
interface. In contrast to an anaerobic sludge digester with an overlying
atmosphere of 25 to 35 percent CO , the CO overlying an anaerobic fluidized
bed reactor will be much less. Assuming influent TOG of 80 to 180 mg/1,
the C09 production would be 27 to 61 mg/1 as C. When these values are
compared to the solubility limits in Table 4 it is clear that the equilibrium
partial pressure of CO will be quite small. The actual values will depend
upon wastewater pH and mass transfer across the gas-liquid interface, but
should be less than 10 percent of the off-gas volume. Also the N_ concen-
tration in the overlying gas volume could be 5 to 15 percent of tne total
gas volume because of evolution of the nitrogen gas initially dissolved
in the wastewater.
Another consideration in estimating methane production is the sulfate
concentration of the wastewater. The sulfate concentration in natural waters
varies widely as shown by the values in Table 5 (33). In anaerobic systems
the sulfate can serve as a terminal electron acceptor in biologically
mediated reactions. This can be represented by the following half reaction
(34)_:
1 SO
2
_L %J\J i » J- */
19 H
16
+ e
_! Hos
16
+ 1 HS
16
19
-------
TABLE 5. SULFATE CONCENTRATIONS AT SELECTED LOCATIONS IN THE UNITED
STATES (33)
2
Location SO^ Range, mg/1 Sampling Period
Connecticut River at 11 - 16 1/66 - 9/66
Thompsonville, CT
Hudson River at 20 - 28 10/65 - 9/66
Poughkeepsi, NY
Neuse River at 5-11 10/65 - 9/66
Goldsboro, NC
Sacramento River at 6-15 4/66-9/66
Freeport, CA
Colorado River near '. 123 - 246 10/65 - 9/66
Grand Canyon, AZ
Ohio River near 100 - 207 10/65 - 12/65
Huntington, WV ,
According to Bryant C35), it is .well known-that methanogenesis in natural
ecosystems does not occur when sulfate is present. Conversion of acetate
to CO- with sulfate reduction to sulfide is thermodynamically more favorable
than acetate conversion to C02 and GH.. With wastewaters containing influent
COD's of 200 to 250 mg/1 and SO ~ concentrations of 200 mg/1 (133 mg/1 as
0_), the majority of the organic material could be oxidized through sulfate
reduction with a corresponding decrease in methane formation. Hydrogen
sulfide gas is extremely soluble in'water (3850 mg/1 at 20°C), whereas
most heavy metals form insoluble sulfides. The partioning of the H S gas
between the liquid and overlying gas phase will depend on the distribution
of sulfur species and the degree to which the equilibrium conditions predicted
by Henry's law are approached.
Process Economics
Bell et al. (36) made a preliminary design for an anaerobic fluidized
bed system treating heat treatment liquor. The reactors were arranged into
four modules with each module consisting of three reactors in series.
Reactor volume was 4106 cu m (145,000 cu ft) with 464 sq m (5000 sq ft)
of surface area. System components ^included individual recycle pumps for
each reactor and controlled gas release. The installed cost, first quarter
1980, was estimated to be $3,436,000. This corresponds to $837. per cu m
($23.70 per cu ft) of reactor volume.
'' 20
-------
Anaerobic fluidized bed treatment was considered for treatment of verti-
cal tube reactor (VTR) effluent in the facility plan for Montrose, Colorado
(37). In the proposed design the VTR was used in place of a primary clari-
fier. The design flow was 10,900 cu m/day (2.88 mgd). Four anaerobic
fluidized bed units, each with dimensions of 7.0 m x 7.0 mx 7.6 m deep with
0;61 m additional freeboard (23 ft x 23 ft x 25 ft + 2 ft freeboard) were
planned to provide a 2-hour HRT at design flow. Reinforced concrete covered
tanks with common wall construction were envisioned. The construction cost
estimate for the anaerobic fluidized bed and recycle pumps was $9.46,0.00
(ENR = 3350). Ignoring the 0.61 m (2 ft) of freeboard, this amounts to
$631. per cu m ($17.88 per cu ft) of reactor volume or $86.66 per cu m
($328,000 per mgd) of design flow capacity.
An anaerobic fluidized bed system is also under consideration for the
town of Hanover, NH (38). JI Associates' initial estimate of the construction
cost is approximately $706. per cu m (.$20. per cu ft), of reactor volume
(March, 1981). This includes pumps and all controls thought to be necessary
for the facility. Thus the estimated construction cost for the 8700 cu m/
day (2.3 mgd) design would be $934,000 or slightly more than $105. per cu m
($400,000 per mgd) of design flow capacity to provide an average 3.6 hour HRT.
Pumping cost per 3785 cu m (1 MG) of flow per 0..305 m (1 ft), head loss
is 24.20 at 5.0 per kwh and 65% overall wire to water efficiency. Hence a
system with a head loss of 3.66 m (12 ft) per pass through the bed and an
overall recycle:influent ratio of 2:1 would represent a power cost of
0.230 per cu m ($8.71 per MG) treated, or an annual cost of 84.00 per cu m/
day ($3180 per mgd) of design capacity.
Summary
The results reported by Jewell (.1,2) and Switzenbaum and Jewell (27)
have demonstrated that better than secondary effluent quality can be obtained
from a laboratory anaerobic expanded bed reactor treating primary effluent
at 20 C. The process was also shown to provide good COD removal with a
glucose feed when the temperature was 10 C and the HRT was four hours or
greater. Since wastewater temperatures in much of the United States fall
to 8 to 12 C during wintertime operation, the response at lower temperatures
is quite important. Previous studies by O'Rourke (.39) with homogenized
primary sludge established that methane fermentation was drastically reduced
at 15 C and that efficient digestion could not be accomplished even at a
60-day retention time. The lipid fraction of the waste was not utilized.
However there was a measurable reduction in the total COD due to the methane
fermentation of formic and acetic acids resulting from cellulose and protein
degradation. Whether or not anaerobic treatment of municipal wastewaters
at low temperature is economically attractive has yet to be demonstrated.
Because of the limited data available, the long time required for such
systems to come to equilibrium, and the scale of the studies reported,
there are a number of questions related to anaerobic fluidized bed technology
which remain to be answered before the design approach can be optimized.
These include:reaction kinetics as a function of temperature and reactor
response under dynamic loading; optimal reactor depth, media density and
21
-------
size; need for equalization basins and an overall flow control strategy
for adequate bed expansion; net solids production; solids levels attainable
in the reactor; biological film properties; effect of biological growth on
media expansion characteristics; solids control strategies in the reactor
if any; need for final clarifiers; influence of wastewater sulfate concen-
tration on the desirability and performance of the process; and long^ term
process stability and reliability at pilot scale. The process is presently
considered to be eligible for funding as innovative technology on a case
by case basis where all relevant factors affecting process performance
have been carefully considered.
ANFLOW
Process Theory
In contrast to fluidized bed systems where extensive surface area is
provided for biological attachment (0.5 mm particles provide about 92.9
sq m (1000 sq ft) of surface area per 0.028 cu m (1 cu ft) of bed), the 2.5
cm (1 in) Raschig ring packing investigated in the ANFLOW system provides
about 5.39 sq m (58 sq ft) of surface area per 0.028 cu m (1 cu ft) (40).
Upflow velocities investigated in the ANFLOW system were normally in the
range of 4.07 - 12.2 m/day (100 - 300 gpd/sq ft) or 0.17 - 0.51 m/hr (0.56 -
1.67 ft/hr). Hence the fluid velocities and the requirements for biological
attachment and growth are substantially different in ANFLOW and fluidized
bed systems. The ANFLOW system relies on sludge settling and solids entrapment
to provide a substantial portion of the BOD removal. Another difference
between the two systems is that ANFLOW treats raw rather than settled
wastewater.
Process Capabilities
Oak Ridge National Laboratory undertook a two-year pilot plant inves-
tigation of an ANFLOW system. The cylindrical reactor had a diameter of
1.52 m (5 ft) and an overall height of 5.58 m (18.3 ft). It contained
3.05 m (10 ft) of 2.5 cm (1 in) unglazed ceramic Raschig ring packing. The
bottom of the cylindrical column was a 45 degree cone with a flanged outlet
at the bottom. A schematic diagram of the process was previously shown
in Figure 2. The column was seeded with a mixture of rumen fluid and anaer-
obically digested sludge. Feed for the ANFLOW unit was taken from the
headworks of the Oak Ridge East Sewage Treatment Plant immediately downstream
of a comminutor. The unsettled wastewater was fed directly to the ANFLOW
column at a constant flow rate which was periodically varied as desired.
An overflow weir and a collection trough in the top of the column were used
to collect reactor effluent. The total off-gas volume was measured by a
wet test meter.
Results from the pilot plant operation have been presented in papers
at three conferences (7,8,9). The data presentations consist primarily of
average monthly values provided in histogram form for selected parameters or
other condensations of process data. None of the papers presents day-to-day
parameter values so that the degree of variation can be quantified.
22
-------
A summary of ANFLQW reactor performance reported by Genung et al. (7)
is shown in Table 6. It can be seen that the ANFLOW reactor does not produce
an effluent of acceptable quality for discharge as secondary effluent.
The ANFLOW reactor produced BOD and TSS removals which averaged 53 percent
and 69 percent respectively, for months 3 through 21. For months 3 through
15, TSS removals averaged 76 percent. Genung et al. (7) drained the
bioreactor (at some point near the end of the study) and washed with waste-
water fed at 30.3 cu m/day (.8000 gpd or 407 gpd/sq ft) for 24 hours to test
the feasibility of periodically removing solids. TSS removal rates were
then reevaluated at 3.78, 18.9 and 26.5 cu m/day (1000, 5000 and 7000 gpd);
in all cases the TSS removals were about 75 percent.
Based on the pilot plant data, an ANFLOW reactor treating municipal
wastewater would be expected to produce effluent qualities as shown in Table
7. Since, the ANFLOW reactor does not produce an effluent of acceptable
secondary quality, further treatment will be required.
TABLE 7. EXPECTED EFFLUENT QUALITIES FROH AN ANFLOW REACTOR
BOD and TSS, mg/1 Effluent BOD, mg/1 Effluent TSS, mg/1
Influent 50 % Removal 60 % Removal 70 % Removal 80 % Removal
200/200
250/250
100
125
80
100
60
75
40
50
Energy Considerations
Since the ANFLOW process is being advocated for energy conservation and
methane production, which is stated to represent a significant and recoverable
energy source (7), it is appropriate to consider the information on gas
generation and recovery potential.
The summary data persented by Koon et al. (8) on average influent
and effluent BOD and TOG values lead to the ratios shown in Table 8.
TABLE 8. AVERAGE TOG:BOD RATIOS MEASURED BY KOON ET AL. (8)
Operating
Period, days
50
72
62
95
36
33
TOG:BOD Ratio
Influent Effluent
0.473
0.725
0.445
0.666
0.766
0.985
0.381
0.680
0.341
0.873
0.689
0.867
The overall time-weighed TOG:BOD ratios for influent and effluent are 0.651
and 0.648, respectively.
23
-------
vO
s
U
0)
60
C
c
0)
O
§ I
bO-u
C m
It"
o &
1-3 M
U
£S
CO
i-l CO
nj co
r
S o
Pd O
M
B-S
4J
a
a)
O bO]t-i
M S1-'
co
CO
H
W
O O O O O OOi-lr-li-lr-l ^- ^- -*'*«* IO IO -*
v£5
vOOr l
c
s
1IrI iI iI i(IliItItI iICMCS
24
-------
Several figures in the paper by Genung et al. (7) provided information
on average monthly parameters for flow, temperature, gas volume and percent
CH,, and influent and effluent BOD. No monthly summary of TOG data was
provided. As shown in Table 9, the amount of off-gas recovered from the top
of the reactor was quite small. In fact, if the factor of 0.651 is used to
estimate the influent TOG (this results in some error for any given month as
shown by the above TOG:BOD variation but does not affect the overall conclu-
sion for the aggregate data), it is clear that only 1 to 2 percent of the
influent TOG was recovered in the gaseous CH, phase when the reactor was
operated near its design loading of 18.9. cu m/'day (.5000 gpd or 255 gpd/sq ft)..
The unweighed average fraction of TOG recovered as methane in the overlying
gas phase for months 10 through 21 was 1.3 percent.
No information was provided on the total amount of methane generated in
the ANFLOW reactor. As shown in Table 10, CH, solubility (when the partial
pressure of CH, and H-O vapor are 760 mm Hg) varied from about 21 to 29 mg/.l
over the range of effluent temperatures which were encountered.
Since the pressure range of operation of a typical wet test gas meter
is 0.76 to 15.2 cm (0.3 to 6 in) of water (41), and a hydraulic head of 5.1 to
10.2 cm (2 to 4 in) of water was adequate to produce flow through the reactor
(7), the pressure overlying the liquid was quite close to atmospheric. The
concentration of dissolved CH, for the case where the CH, in solution is
in equilibrium with the overlying gaseous methane (i.e., net methane flux
is zero) is also shown in Table 10. The concentration of methane exiting
the reactor as dissolved gas should have been between these two limits.
The data in Table 10 were used to develop the estimates of CH, production
shown in Table 11. The MAX and MIN values refer to the upper and lower limits
anticipated for the methane exiting in solution. Again, it is noted that the
values for percent carbon removal or influent and effluent TOG values are only
approximations for any given month because no monthly TOG data were provided
and the values were estimated from the BOD data assuming a constant TOG:BOD
ratio. As shown in Table 10, most of the methane produced exits the reactor
dissolved in the liquid phase. For months 6 through 21 the amount of carbon
removed in the ANFLOW reactor that was converted to methane has averaged
between 25 percent (MIN) and 45 percent (MAX). For the same period the amount
of carbon that entered the reactor and was converted to methane was between
13 percent (MIN) and 23 percent (MAX). The average BOD removal for months
5 through 21 was 53 percent.
The following statements are made by Genung et al. (7). " The methane
produced was approximately 33 percent of that which could theoretically have
been produced as calculated from measurements of the organic carbon removed
from the'wastewater by processes in the bioreactor. This efficiency was
difficult to estimate, however, since carbon was removed by many mechanisms,
some involving solubilization phenomena, for instance, which occurred over
undefined periods." The exact meaning of these statements is not clear,
but the figure of 33 percent may correspond to the 25 and 45 percent minimum
amd maximum concentrations estimated above.
Based upon the above calculations it is clear that any significant
25
-------
o o
CO
"0 0)
0) O
Q) -3-B
> K
O O
o
CU C
oo o
O CM
CM r-H
M
O
JJ
y
n)
g
Q)
.e
§
to
«
o
i
u-i
O
CO
w
T3
0)
n
0)
O
a)
0)
I
4J
0)
S
I
*fl
o
oo
§ %
m
C cfl
O 13
^ bO
cfl
O
C
O >i
H cfl
JJ T3
O ^
3 r-l
-a
c
0)
I-l
M-l
C
M
i-H CM iI CO iI C
vOvOr~-vOvO
4-1
c o
26
-------
M
O
0
cd
Q)
^g '
0
23
0)
43
4J
j-
4-1
H
1*
a
0)
4J
o
OJ
a
«
w
CO
C
0
H
] i
cd
S-i
4-1
a
0)
o
C
o
u
a;
(3
cd
43
4-1
S
-0
CD
I 1
O
co
co
H
P
0
i t
w
hJ
H
C -fc
a) -ft
3 * rH
r-< O -~-~
<4-l O 60
*w H 6
W
M
c
cu -x
3-J,
y3 tl r~H
i-l VJ
«* 3 CX5r~-vDvD\£J
CSICSCNCNiIrHrHT-l^-liI
CN to l/~|
(NCNCN
-------
M
O
4-1
O
cfl
0)
£2
O
i
0)
4J
i-l
C
o
H
O
3
TJ
O
M
PM
(1)
cfl
^
4-1
CU
a
*o
O
0
CU
o
&
cu
Pi
£
o
1-1
Fn
4-1
C
CU
3
t-i
C
r-l
M
pr{
CN
m ON r-l
0 ^
e M
cu e
Pi
oo co m oo o
m co m m r^
4-1
C 0
O £^j
g
m ^o r^ oo o\
o in vo CN
in vo I***- o> c^ CTN oo ^t" r^* r**^ CN ^f
il i 1 i 1 i 1 rl i 1 i 1 r 1
*^" 00 CO CN CN ^O O m CO VO OO I
CNCNCO~d"CNCSCNCNi ICNi li 1
m co co o 'O co vo COCO
covovoosor-£>OO
oo ON o ^^
i li-li-lr-li li 1 i-l i-l r-l i ICNCN
§
H
(1)
a.
13
cu
4J
(X
3
o
o
28
-------
recovery of the methane gas produced In the ANFLQW pilot reactor would have
involved recovery from the liquid phase presumably through vacuum degasifi-
cation. This is in contrast to conventional sanitary engineering design for
anaerobic sludge digestion. With sludge digestion, the incoming waste stream
is thickened to organic carbon concentrations more than two orders of magni-
tude greater than those entering an ANFLOW reactor and as a consequence the
methane exiting in solution represents1 less than one percent of that
generated.
Total methane gas production which represents 13 to 23 percent of the
total TOG which entered a plant as in the ANFLOW pilot reactor, is no more
than produced in a conventional plant with anaerobic digestion of primary
sludge only.
Design and Economic Considerations
Griffith (42) developed cost information for a conceptual ANFLOW reactor
design based on a hydraulic loading of 10.4 m/day (255 gpd/sq ft). The
estimated capital costs for various plant sizes and media costs were
approximately as shown in Table 12.
TABLE 12. CAPITAL COSTS FOR AN ANFLOW REACTOR REPORTED BY GRIFFITH C42)
Design Flow
cu m/day MGD
Capital Cost, Millions of Dollars
Media Cost, $/ cu f.t
5 10 15
1135
3785
18925
0.3
1.0
5.0
0.18
0.59
2.8
0.26
0.85
4.0
0.33
1.1
5.1
Based on a media cost of $353./cum ($10./cu ft), the packing comprised
50 percent of the capital cost for a 3785 cu m (1 MGD) ANFLOW reactor. The
capital costs are based on an ENR index of 2700. The capital cost given by
Griffith can be estimated by:
COST=e[-9731n(FL°W)]
x
where cost is in millions of dollars, flow is in MGD and b = -0.54034,
-0.17735 and .06299 for packing costs of 5, 10 and 15 dollars/cu ft,
respectively.
The capital cost for a 3785 cu m (1 MGD) ANFLOW treatment system based
on packing at $353./cu m ($10./cu ft) and an ENR of 2860 was estimated by
Koon et al. (8) at $2,981,250. Based on the information in Table 4 of Genung
et al. (7), this estimate includes 50 percent extra for related costs plus
a $100,000 base cost for surge and reaeration tanks. Hence, the base cost
of the ANFLOW reactor was estimated at $1,887,500. The conflicting informa-
tion on reactor design criteria makes it impossible to know the basis of
the design with certainty, but in all probability the cost refers to a
reactor receiving an influent flow rate of 5.87 m/day (.0.1 gpm/sq ft or
29
-------
144 gpd/sq ft). In this case the equivalent cost estimate by Griffith (1.77
MGD and ENR of 2860) would be $1,546,000 assuming the 50 percent cost increase
to cover all installed costs.
The ANFLOW system process flow diagram (8) envisioned for a complete
wastewater plant includes the following major components:
1. Bar Screen
2. Grit Chamber
3. Comminutor
4. Equalization Basin
5. Grinder Pumps following the Equalization Basin
6. ANFLOW Column
7. Surge Tank (for sludges and backwash water)
8. Reaeration and Chemical Addition Tank for ANFLOW Column Effluent
9. Upflow Sand Filters following Chemical Addition
10. Chlorine Contact Basin
Hence, any analysis of the design and cost effectiveness of an ANFLOW system
entails more than an economic analysis for just an ANFLOW reactor.
It is desirable to present and discuss, where appropriate, several
observations concerning published information on ANFLOW system design and
economics. Specific points worth noting in the paper by Koon et al. (8)
are as follows:
1. This paper discusses conceptual designs to treat a raw wastewater
with a BOD and TOC of 300 mg/1 and VSS of 275 mg/1 with TSS of
350 mg/1. Plant sizes of 0.05 and 1.0 MGD are discussed.
2. The average hydraulic design loading rate is stated in Table 3 to be
0.1 gpm/sq ft with a peak loading rate of 0.15 gpm/sq ft. In Table
4, the ANFLOW filter is stated to have a detention time of 18 hours
and 14,120 sq ft of surface area for a 1 MGD flow. The total
volume is given as 141,200 cu ft which corresponds to the 10 ft
depth specified on Page 14. However, 0.1 gpm/sq ft and a 10 ft
depth correspond to a detention time of only 12.5 hours. A flow
of 0.1 gpm and 14,120 sq ft of surface area corresponds to an
average flow of 2 MGD for the 1 MGD plant. Furthermore, it is stated
on Page 14 that the design was based on a hydraulic loading rate of
0.15 gpm/sq ft which corresponds to a flow of 3 MGD for the 1 MGD
plant. In addition to the ANFLOW reactor, the design calls for an
equalization basin and aeration system which adds an additional 8
hours of detention time based on the influent flow (i.e., a 1 MGD
flow to a 1 MGD plant). Additional facilities include a grit chamr-
ber, upflow sand filter, and: chlorine contact tank.
3. It was felt that both comminution and subsequent grinding of the
influent to produce particle sizes less than 0.5 mm would be required
to insure that the anaerobic filter did not prematurely clog.
4. The ANFLOW reactor was anticipated to have a soluble effluent BOD of
30
-------
5.
25 mg/1 and effluent TSS of 35 mg/1. The design effluent BOD from
the entire treatment system (ANFLOW reactor, aeration and filtration)
was to be 30 mg/1. A design effluent TSS of 35 mg/1 from the ANFLOW
reactor corresponds to 90 percent TSS removal. If the effluent
TSS are assumed to be 75 percent volatile, this would represent a
solid BOD of 20 mg/1 based on the stated insoluble BOD/VSS ratio of
0.76. Hence, the total effluent design BOD from the ANFLOW reactor
would be roughly 45 mg/1 which corresponds to a BOD removal of 85
percent. Table 6 summarizes the information on,BOD and solids
removal obtained from 21 months operation of the pilot plant reactor;
these data were estimated from the figures in Reference 7. As
shown in Table 6, at no time during their pilot plant studies did
they observe solids removal of 90 percent or BOD removal of 85
percent. Even after the reactor was drained and washed to remove
excess solids (Figure 9, Reference 7), the solids removals after
restarting were only 75 percent. Hence, the entire economic
analysis is based on an ANFLOW reactor which is assumed to perform
significantly better than observed during the pilot plant operation.
Energy available from the gas generation for the 1 MGD plant was
stated to be 225 hp based on an off-gas volume of 35,820 cu ft/day.
How these figures were determined is not given. In the subsequent
section on system costs (Page 19 of Reference 8) it was stated that
a 40 percent conversion efficiency was used to determine the
equivalent cost of recovered power. It should be noted that this
is a higher efficiency than obtained in the average fossil fuel
electric generation plant which is 33 percent (43).
The information on power generation can be used to estimate the
assumed efficiency of TOG conversion to CH, and CO . The heat
of combustion of methane gas at 25°C to H^O, . and CO , . is
21,502 BTU/lb(44). At STP, 35,820 cu ft of^E,. will w4?gh 1595.95
Ib and is equivalent to 224.8 hp at a 40 percent conversion
efficiency. This presumably is the basis for their anticipated
energy recovery and apparently the figure given for the off-gas
refers to the methane only. This conclusion is based on the
power recovery of 225 hp which is specified in Table 5 (in Reference
8).
The wastewater characteristics used for the analysis were 300 mg/1
of TOG or 250.0 Ib of TOC/MG. Of this, 1197 Ib of C is apparently
presumed to show up in the off-gas as CH,. Based on roughly 20 mg/1
of CH^ lost in the effluent (31) an additional 125 Ib of C will exit
in the water as dissolved CH,.
4
The relationship for power generation in hp is stated to be 0.98
Q TOG with Q in mgd and TOG in mg/1. If all incoming C were
converted to CH, only, the factor used to multiply the product of
Q and TOG removed to obtain horsepower is
8.34 JL6 21502 778.1 0.4
12 550 60. 1440
1.566
31
-------
Thus hp - 1.566 Q
TOG
'removed
Since 0.98/1.566 = 0.626, they have apparently assumed a 62.6 %
conversion of TOG to CH, which is not unreasonable.
However, 1322 Ib of CH, as C plus the CO- expressed as C represents
1816 Ib of C converted to CO and CH, per MG treated. If the carbon
removal in the ANFLOW reactor is approximated by BOD removal (85%)
and ignoring the BOD recycled from the filter backwash, then
85.5% of the removed TOG was apparently assumed to be converted
to CO- and CH,. It is clear that this amount of CH, generation
far exceeds anything observed in their pilot plant studies.
6. The solids yield coefficient for the ANFLOW system was given in
Table 3 of Reference 8 as 0.2 g TSS/g BOD removed. If the 255 mg/1
of BOD is removed in the ANFLOW"reactor, this would correspond to
a net solids production of 425 Ib/day for a 1 mgd flow. If an
additional 20 mg/1 of TSS are removed in the filters and assuming
an additional 10 mg/1 production of TSS across the filters due to
BOD removal (since their design assumes BOD removal), the solids
returned to the surge tank in the backwash water would be an addit-
ional 250 Ib for a 1 mgd flow. However, Tables 3 and 4 also
indicate that 10,000 gpd are to be drained from the ANFLOW column
at 2% solids for a stated solids accumulation in the surge tank
of 1650 Ib/day. This presumably means that none of the solids in
the backwash water are assumed to settle in the surge tank but
that they are all recycled to the ANFLOW reactor where they are
all removed.
Of the 1650 Ib/day of sludge to be produced, it appears that the
design calls for the alum and polymer addition to produce 975
Ib of chemical sludge/day for the 1 mgd plant. If one assumes
this sludge ,is all A1PO, (influent P of ^ 30 mg/1) the corresponding
alum dosage is 'about 1600 Ib/day for a 1 mgd flow. Bagged commercial
grade aluminum sulfate is currently selling for $146 - $154/ton (45).
As an alternate calculation, if 255 mg/1 of TOG are removed in
ANFLOW and 14.5% of this is not converted to C02 and CH^, the C
accumulation is 308 lb,/day for 1 mgd. If the organic material is 50%
C and the total solids, accumulation is 70% volatile (this assumption
was not made in the paper being reviewed)., then the total solids
accumulation ignoring backwash, would be 880. lb./day.
7. The effluent from the ANFLOW reactor will receive 3.2 hours of
aeration followed by alum and polymer addition and upflow filtra-"
tion. The sand filter is either 6 ft deep (Table 3 in Reference
8) or 5 ft deep (Page 17 in Reference 8). No data are given on
anticipated alum or polymer dosages.
The design air flow rate of 362 cfm to the 1 mgd aeration system
will provide 9.0.20. Ib of oxygen at 20PC. Providing 8 mg/1 of D.O. for
1 mgd requires 66.7 Ib of O^. At 20°C and a 5% transfer efficiency,
32
-------
the proposed air flow rate would dissolve 451 Ib of 0 per MG.
The stoichiometric combustion of 1596 lb-of CH. to ;CO_ and HO
requires 6384 Ib of 0 . If an additional 167 Ib of CH, is stripped
in the aeration process, the stoichiometric oxygen requirements rise
to 7052 Ib. The volume occupied by 167 Ib of CH4 at 20°C is 4023
cu ft or 0.77% of the air volume supplied to the 1 mgd aeration
basin. The CH, in the contained off-gases from the aeration unit
will not be present in explosive concentrations.
The cost analysis was said to include covering of the equalization
basin for off-gas containment, but no information was given con-
; cerning the treatment/disposal of these gases. However, it is
clear that the off-gas volume is compatible with that required .for
off-gas combustion (based on their assumed CH, generation) so this
is not a problem. In fact, this may have been the basis for-
choosing the 362 cfm air flow rate.
9. According to Table 4 (in Reference 8), 15 hp (.11.2 kw) will be
required to operate the aeration basin in the 1 mgd plant. However,
in Table 10 (in Reference 8) where the power requirements for ANFLOW
are compared to activated sludge, the complete effluent polishing
step (aeration plus filtration) is stated to require only 6.66 kw.
10. The inconsistency in stated power requirements and other information
makes it difficult to estimate backwash frequency for the filters.
The design calls for the removal of 20 mg/1 of TSS in the filters
plus an assumed soluble BOD reduction of at least 5 mg/1 due to
bacterial growth. Table 3 (in Reference 8) lists the maximum filter
headloss and gives a value of 0.05 for specific deposit,defined.as
, Ib SS/sq ft/ft headloss. Filter area is 242 sq ft (for 1 mgd) and
backwash requirements are 350 gal/sq ft. Hence, the removal of 167
Ib of TSS (ignoring solids production in the filter) will require
2.3 backwashes/day and produce 195,000 gal/day of backwash water
for the 1 mgd plant. On the other hand, it was noted in Item 2
above that the design flow for the 1 mgd plant is at ,least 2 mgd.
This indicates that they are estimating a backwash requirement
: which is 100% of the influent flow. Again it is difficult to deter-
mine just what the design actually calls for.
11. It is stated in the section on cost analysis that the comparative
cost information is considered accurate to ± 50%.
12.
A cost credit was taken for the energy to be recovered from CH, gen-
erated in the ANFLOW process at the 1 mgd size (Page 19). However,
no capital costs are assigned to gas collection, cleaning, storage,.
or power generation equipment. Since an overall conversion efficien-
cy of 40% was assumed, both electric generation and waste heat
recovery must have been contemplated. The cost credit is probably
$44,OOQ/year (225hp 365 24 Q.07457 O.Q3£/kwh) although the
value is not specifically stated. The paper states that this cost
was used to offset unit process power costs for the 1 mgd flow rate
case.
33
-------
13. The 1 inch ceramic Raschig rings used in the pilot plant have an
installed cost of $10./cu ft. This results in a capital cost of
$3.85 million for a 1 mgd complete ANFLOW plant and a capital cost
which is 2.03 times that of the activated sludge system against
which it is compared. At 0.05 mgd, the capital costs of the two
systems were essentially the same. When 3 inch plastic ring packing
was used for the 1 mgd ANFLOW cost estimate, the capital cost
estimate decreased to $2.45 million. The 1 inch ceramic rings
used in their pilot studies had an approximate surface area of 58
sq ft/cu ft (40) whereas 3, inch Raschig rings would reduce the sur-
face area to approximately 20 sq ft/cu ft. For a given .film
thickness, the amount of attached biological growth in the system
with 3 inch rings would be 1/3 that in the piloted system. It is
not clear how this change will enhance the performance of the
reactor.
The paper by Genung et al. (7) also gives cost and energy comparisons
between ANFLOW and similar activated sludge configurations to those used in
Reference 8. A few points worth noting in this paper are:
1. The capital costs for the ANFLOW system are all based on using the
3 inch plastic packing. [
2. Capital costs and O&M labor requirements for both ANFLOW and
activated sludge are said to be the same for pumping. Yet Koon et
al. (7) states that grinding to particle sizes of 0.5 mm or less
was felt to be needed for ANFLOW. The flow scheme only calls for
grit removal, comminution and flow equalization prior to the
grinding/pumping operation.
3. The total annual cost for ANFLOW at 1 mgd ($297,000 on Page 35 of
Reference 7 vs. $300,000 in Reference 8) again apparently takes
credit for energy recovery but allocates no capital for this
to occur.
4. There is no breakout of O&M costs for ANFLOW so it is not clear
how much money, if any, has been allocated for alum and polymer.
The sludge production data given in (8) suggest an alum cost alone
of $43,800/yr at 1 mgd. Yet the capital cost for ANFLOW of $2.451
million and labor requirements of 6490 man hr/yr represent an
annual cost of $261,000 at 1 mgd when using the labor rates and
amortization specified in (8). Since the total annual cost is said
to be $297,000, alum and polymer costs were probably not considered.
t
5. Even for the weak wastewater (_BOD = 100 mg/l)_, Table 4 indicates
that power recovery is feasible for ANFLOW and will produce at
least 954 kwh of power. Based on the results of their pilot plant
operation, this is unrealistic.
By this point, it is clear that the economic analysis for a complete
ANFLOW system is based on attaining greater BOD and TSS removals and more CH,
production than ever attained in the pilot plant operation. To make the
34
-------
process competitive with activated sludge requires using an untested packing
which will provide about 1/3 the surface area used in the pilot plant reactor.
Neither pap'er reviewed here provides a flow and materials balance, and it is
not clear upon what the ANFLOW design assumptions are actually based. Items
such as purchasing and maintaining grinder pumps, power recovery equipment,
alum and polymer etc., seem to have been "lost" in the cost accounting
procedures used. The vagueness is exacerbated by the manner in which data
and design assumptions are presented. The limited number of data parameters
makes it impossible to make calculations concerning reactor performance and
the performance of the subsequent processes needed. For example, effluent
BOD's are not subdivided into particulate and dissolved values. Yet there is
the statement (7) that the volatile acids were not efficiently converted
to CH, in the colder months and tended to be discharged with the effluent,
but the acid concentrations are never given. Since acetic acid is the most
prevalent volatile acid intermediate formed in the methane fermentation of
fats, carbohydrates and proteins, and about 70 percent of the methane
produced results from its degradation (46), quantitative data are needed.
Similarly, Figure 3 in Reference 8 indicates that for one steady state period
the COD reduction within the ANFLOW reactor averaged 99%, which is better
than any aerobic system could be expected to perform, but this astounding
observation is not explained or developed. Successfully meeting secondary
effluent standards will require terminal suspended solids and soluble BOD
removal. However, the only results mentioned from the 2-year program are
batch filtration studies with 0.25 to Q.50 mm sand, which is probably too
small for full-scale operation, and 3-weeks operation with a downflow dual
media filter which presented operational difficulties.
35
-------
SECTION 4
COMPARISON WITH EQUIVALENT TECHNOLOGIES
The use of anaerobic systems in place of aerobic systems for wastewater
treatment offers several potential advantages. There are no oxygenation
requirements and biological sludge production is much lower. Energy
requirements may prove to be lower than with conventional systems and a
potentially usable fuel, CH,, is produced.
To obtain a better overall perspective of the cost of anaerobic systems
in relation to conventional activated sludge plants, the cost of primary
treatment plants was compared to conventional activated sludge plants for
flows of 3785 and 37850 cu m/day (1 and 10 mgd). From this comparison,
one could estimate the range of costs for anaerobic reactors that would make
the total system cost competitive with the activated sludge system. Version
1.2 of EXEC/OP (47,48) with the single design evaluation feature was used
to generate the costs of primary and ^activated sludge treatment systems.
This computer program computes cost and energy requirements for a specified
sequence of unit processes and design parameters. A partial listing of the
input design parameters is presented in Table 13. Output parameters for
cost and energy requirements are summarized in Tables 14 and 15.
As previously shown, the sludge handling sequence for an anaerobic
fluidized bed plant would likely be the same as for a primary plant.
Hence a comparison of the processes in Tables 14 and 15 gives an indication
of the costs which can be associated with a fluidized bed reactor and any
needed ancillary equipment (e.g. postaeration) and still result in a treat-
ment sequence competitive with conventional technologies. The two solids
handling options considered in Tables 14 and 15 were either gravity thicken-
ing, anaerobic digestion, elutriation and vacuum filtration; or gravity
thickening, lime stabilization and vacuum filtration. This accounts for the
two different costs given for primary plants in Table 14 and for secondary
plants in Table 15. All energy requirements for heat, fuel or electric
power are expressed in units of equivalent kwh/MG.
The total annual costs for the plants summarized in Tables 14 and 15
are shown in Table 16.
The difference in costs at 1 mgd between the complete activated sludge
and primary plants was 324 and 275 $/MG; the cost of the activated sludge
tank and final settler was $189/MJ. At 10 mgd the difference in costs between
the complete plants was 145 and 138 $,/MG; the costs of the activated sludge
tank and final settlers were 84 and 83 $/MG with the difference reflecting
!. 36
-------
TABLE 13. Summary of ..Selected Input Parameters Used for
: EXEC/OP
Construction Cost Index (3rd quarter 1973 = 1.0)
Wholesale Price Index (1967 = 1.0)
Discount Rate, decimal
Planning Period, years
Direct Hourly Wage, $/hr
Heat Energy Conversion Efficiency, decimal
Process
Raw Wastewater
Pumping
Preliminary Treatment
Primary Sedimentation
Activated Sludge and
Final Settler
Ch1 orination
Gravity Thickening
Anaerobic Digestion
Elutriation
Design Parameter
Influent TSS, mg/1
Influent VSS, mg/1
Influent Suspended BOD, mg/1
Influent Dissolved BOD, mg/1
Pumping Head, ft
Grit Removal
Bar Screens
TSS removal, %
Underflow TSS, %
Effluent BOD, mg/1
Effluent TSS, mg/1
MLVSS, mg/1
Oxygen Transfer Efficiency, %
True Yield Coefficient, Ib/lb BOD
Biqmass Decay Coefficient, I/day
Chlorine Dose, mg/1
Solids Recovery, %
Hydraulic Loading, gpd/sq ft
Solids Loading, Ib/day/sq ft
Underflow TSS, %
Temperature, °C
Retention Time, days
Washwater Ratio
Underflow TSS, %
Solids Recovery Ratio, %
Hydraulic. Loading, gpd/sq ft
Solids Loading, Ib/day/sq ft
1.55
2.4
.07125
20.
7.00
.30
Value
240:
190.
170.
80.
30.
yes
yes
60.
2.
25.
25.
2000.
6.
.7
.12
8.
95^
600.
25. (PRI)
8. (MIX)
9. (PRI)
4.5 (MIX)
35.
20.
3.
9. (PRI)
5. (MIX)
85.
500.
20. (PRI)
10. (MIX)
37
-------
TABLE 13. (Continued)
Process
Lime Stabilization
Vacuum Filtration
Truck Transport and
Land Disposal
Design Parameters
Lime Dose, Ib/ton dry solids
Primary Elutriated Digested Sludge
FeCls Dose, Ib
Lime Dose, Ib.
Mixed Elutriated Digested Sludge
FeCls Dose, Ib.
Lime Dose, Ib
Primary Lime Stabilized Sludge
FeCls Dose, Ib
Lime Dose, Ib
Mixed Lime Stabilized Sludge
FeCls Dose, Ib
Lime Dose, Ib
Dewatering Rate, gph/sq ft
Primaryl Elutriated Digested Sludge
Mixed Elutriated Digested Sludge
Primary Lime Stabilised Sludge
Mixed Lime Stabilized Sludge
FeCl3, $/lb
Lime, $/ton
Hauling Distance, miles
Depreciation Period for Trucks, years
Fuel Cost, $/gal
Land Cost, S/acre
Landfill ing
40.
0.
100.
0.
30.
0.
60.
0.
16.25
10.
17.5
11.25
.064
45.
5.
7.
1.
3000.
yes
38
-------
o
r~H
p&j
1
1-1
ca
o
M-l
o
o
a
O
L4
^.,|
E>"t
&D
J-i
OJ
c
w
"§
>>co
ens
s- -».
co 3:
CM C 3
fri 1 ^X
0
z: -t->
io s:
"Z. CJ o2
- E^.
2p C ^
^^ rtj
0£. CO -r-
0. 4-> CL
o ca
1 CJ
>,o
ens
i QJ ~r-
c: 3
* ' ' ' *vX
O
4->
I 1/1
2:0 s
=3; O o3
1 O
D_ i
> 3
a: c CD
" -fc^ ca
Q- ca -i
4-> a.
co ca
H O
>,0
ens
CU 1C
C 3
LU 2*£
(_
s:
«C -M
i tn
a. o s
O o3
>. O
cc
sj-
co
vo
CO
CM
o
o
LO
cn
CO
CO
CM
o
o
LO
LO
o
s
ro
<0
CO
CM
O
0
10
03
CO
o
CO
It
IO
CM
£
^
CO
00
o
CO
vo
CM
r~~*
1^^
i~
CO
CO
0
to
C\J
r^
i^"
i^
m
LO
CM
f^
in
*
CM
CO
CM
CO
«d-
O
CO
CM
LO
1
1
1
1
1
I
I
1
1
1
1
LO
CO
LO
cn
^>
CO
o
*o
OO
o
co
CO P*!
cn oo
CM CM
CM r
LO CO
CM tO
CM i
CM O
co co
r
« *
CM CM
CM r
CO 0
*r cn
«
CM IO
CM r
CM CM
* *
co !*.
o «
CM CM
CM
CO r
lesj" ^*
»
CM r«~
CM r
cn ^ ro
r~ co i oo
LO l£J
«* to o
O i 1 CM
«3" CM CD
CM CO
^J" O LO
CM CM 1 i
CO r-. o
r--. i co
CO CM
! 1 LO «dr
CO
o
i i r->- CM
ii . ^
Oi -=3-
CM ->J-
CO 00
1 i CM CO
1 1 «
CO
Is** r*^«
ii
CM
i LO
1 1 «5T CM
11
co 10
IO CM
1 1 i LO
1 1
to cn
O
5
r^
*
cn
CO
0
UD
CO
to
0
cn
LO
CM
CM
CM
CO
LO
LO
LO
O
CO
CO
CO
co
O
r-.
CO
O
cn"
cn
cn
CM
CT>
^*
LO
CO
cn
3-
cn
*
LO
CM
CO
cn
O
vo
CO
^*
CO
r
^
*
LO
"^
^_
O
to
to
LO
r"-
CO
O
U
H
cn
c
CL
E
^
i-
co
4_3
ca
3
3
CO
tt:
c:
CO
S
->->
CO
CO
s_
(
^^
i.
ca
c:
E
r
r
0)
s_
Q.
O
r
^J
CO
^J
O)
E
r
co
oo
i-
CO
E
r"
^.
C-
E
4J
o
ai
3
^
tf_
LU
S_
CO
c:
a>
.^
o
r
1
^
1
ca
!_
C
O
f
4->
in
CU
cn
r
0
~o
CO
ca
c:
c:
o
«f.K
4->
ca
r
^_
3
LU
CO
cn
^«
00
c:
o
r
CO
NJ
r-
f
-Q
CO
OO
co
.(..
1
c:
o
r
ca
i.
4^
.,
U-
£
^j
=
i
s- ca
O C
O fO
i.
39
-------
0
I-*
8
S
0
r 1
CB
M
O
U-4
pn
O
***»
O
w
U
m
00
M
(1)
g
W
o
CO
CO
O
o
m
1
C*
i «a-
< ens -^
a. a> 3= i
c: 3
>- LU i«i
s: *-> o
» i IO CO
a: o 21
a. c_j oo co
o
"re CD
3 2:
c: ~-.
< ' oo
i 4J
4J O. CM
o re
CM 1 CO
o
o
«=C to CO
a 0 21
SO o3 CO
8- °
LU ro
OO 3
C CD
C ^r-
v=>'re oo
r -4-J ^3"
re *f^
4J Q. OO
o re CM
si-
ens: >
C3 O> 3T
E: c 3
LU i«i
3
^^ s
_a (/? co
CX> O ff»~
O o3 OO
5- O
CC i
a 3
s= c cs
O EZ 2Z
LU *»=>« CO
OO i^~ 4-* ^~
4J O. OO
o re CM
cn
a.
"-;
o_
s_
CU
H-
>^
S-
re
j~*
r
E
i«*
i«~
OJ
S-
0-
00
CO
LO
CO
CM
CO
oo
CO
CM
CO
<=!
CM
CO
cn
co
cn
CM
^
r
r-.
co
c:
o
] <
c
cu
r
a
cu
00
>}
s_
re
^
r"
t_
O-
1
1
1
1
CM
co
CM
LO
O
00
OO
o
'*
in
Bt
^
LO
CTl
CM
in
CM
^.
OO
CO
in
a
re
c
t »r-
4-J U_
U
CM
CO
CM
in
O
o
*
oo
cn
in
CTl
cn
ID
«5f
CO
LO
LO
r
in
O
l^-
c
o
4-3
re
c
s-
o
P
^*
C-3
4-)
C"
CU
^3
r
H-
M
LU
40
cn
*
CM
CM
r
cn
CM
vo
in
^j-
in
oo
ID
,
OO
1
0
oo
s_
cu
cu
o
r-
<""
1
>}
4-3
i
>
re
i.
cs
tn
<3-
LO
OO
ID
C3
1
t
1
1
ID
in
LO
r
CM
in
ID
in
r
CM
C
O
4-J
cu
cn
r
a
o
o
o
s-
a>
c:
"^
co
CM
CM
O
CM
1
1
1
i
«*
r
5
,
CM
CM
OO
d
O
re
r
s_
3
r
LU
ai
cn
-a
3
r
OO
1
1
1
1
cn
oo
o
ID
*
OO
cn
CO
CM
1
1
1
1
|
C
o
r
4->
re
IM
r
r
r
Q
re
u-3
oo
a>
E
^
l
LO
ID
CM
CM
CM
00
cn
*
I""
a
in
<=)
oo
i
^
CM
OO
O
P-.
cn
CM
oo
CM
oo
CM
OO
CM
C
0
re
^.
4-3
r~~
r*
LU
E
Z3
n
o
*o
>
CO r-
m
in
ro
ro
CM
cn
ID
CO
o
oo
oo
vo
co
o
in
ID
O
CM
CM
in
co
el-
O t/>
Q. O
(/> O.
'i- ro i
-a
u re
CM
UD
CM
cn
cn
ID
co
in
ID
co
in
co
co
=3-
OO
co
cn
CM
o
co
cn
10
co
vo
=c
o
-------
TABLE 16. SUMMARY OF COSTS GENERATED WITH THE EXEC/OP PROGRAM
Plant Size
(mgd)
1
10
Activated Sludge Plant
($/MG)
No. 1 No. 2
871
320 313
Primary Treatment Plant
($/MG)
No. 1 No. 2
547 596
175
the variation in recycle streams from sludge processing. In both'cases it
can be seen that the combined capital and operation cost of an anaerobic
fluidized bed and ancillary equipment can exceed that of an activated
sludge tank and final settler (by 1.45 to 1.73 for the cases considered)
and still produce a treatment system with the same annual cost as the activ-
ated sludge plants modeled. : *
If an anaerobic fluidized bed to provide a 2-hour HRT could be installed
at $1060/cu m ($30/cu ft) for a 3785 cu m/day (1 mgd) flow, the capital cost
would be $334,000 or a total annual cost (at 7 1/8%) of $87/MG for the
capital cost portion of the plant. At 37850 cu m/day (10 mgd) and $706./cu m
($20./cu ft) the annual capital cost would be $58./MG. Pumping costs for an
anaerobic fluidized bed hopefully would by no more than $10/MG (see Section
III). A rough guide to estimated maintenance material costs and labor
requirements can be obtained by considering the reported requirements (49)
for gravity filtration structures. Estimated requirements are shown in
Table 17.
TABLE 17. OPERATION AND MAINTENANCE SUMMARY FOR GRAVITY
FILTRATION STRUCTURES
Total Filter
Area
sq m , sq ft
Building
Energy
kwh/yr
13
65
130
650
1300
2600
140
700
1,400
7,000
. 14,000
28,000
44,120
151,850
279,0.70.
1,190,160
2,165,890
4,123,490
Maintenance
Material
$/yr
80.0
2,510
4,020.
13,200.
21,600.
36,70.0
Labor
hr/yr
9.00
1,50.0.
2,10.0
4,6.0.0.
7,0.00.
18,00.0
Total Cost
$/>yr
11,1201
22,070.
33,390
94,9.00.
156,580
340,400
*Calculated using $.Q3/kwh and $10./hr of labor
When.these cost estimates are compared to the cost differences in Tables
14 and 15, it is clear that anaerobic fluidized bed systems offer potential
cost advantages when compared to conventional treatment systems.
Fluidized bed systems and ANFLOW systems have certain disadvantages or
characteristics which are not typical of aerobic biological systems. These
are:
1. The effluent will contain no dissolved oxygen, Post aeration will
41
-------
be required.
2. Problems associated with hydrogen sulfide production relate to
undesirable odors and the corrosive characteristics of this gas.
Either anaerobic system may find this to be a potential problem.
3. Since the biological sludge production is less than with aerobic
systems, there will be a 'corresponding reduction in N and P removal.
Furthermore, none of the nitrogen will be oxidized.
4. Methane forming bacteria have slow growth rates and are sensitive
to toxic materials. If a plant receives toxic materials there may
be long periods during reestablishment of the methane bacteria
when treatment efficiency suffers.
42
-------
SECTION 5
ASSESSMENT OF NATIONAL IMPACT
Anaerobic systems used for treatment of municipal wastewaters offer
potential for reducing operating energy requirements compared to conventional
activated sludge systems. However, owing to the fact that these systems
are still in the developmental stage, little data are available upon which
to base firm estimates of the energy required to operate an anaerobic
reactor, and the energy which may be recoverable as a result of its opera-
tion. Similarly, since no full scale systems have been designed or construc-
ted, cost assumptions are difficult to justify. For these reasons, no attempt
was made to project the national impact of implementing full scale anaerobic
systems for treating municipal wastewater.
There are a number of studies planned or underway to expand the knowledge
of anaerobic system process performance. The Department of Energy has
recently funded several studies in their program to evaluate submerged media
anaerobic reactor (SMAR) concepts. These include:
1. i A study to identify and evaluate factors affecting SMAR performance
conducted by Dr. Young at Iowa State University.
2. A study to investigate and evaluate the mechanisms of anaerobic
! filter treatment of dilute wastewater by Drs. Rittmann and Pfeffer
at the University of Illinois.
3. A study with Dr. Jeris at Ecolotrol, Inc. to operate an 18.9 cu m
(5000 gpd) pilot plant on municipal sewage to evaluate the anaerobic
.expanded and/or fluidized bed process.
4. Installation of a 189 cu m (50,000 gpd) ANFLOW reactor at Oak Ridge,
Tennessee. This is a joint effort between DOE/'ORNL, the Norton
Company and the City of Oak Ridge.
Projects planned or underway under the sponsorship of EPA include:
1. Installation of pilot scale fluidized bed reactors for anaerobic
wastewater treatment at the EPA Test and Evaluation facility in
Cincinnati, Ohio.
2. An active I/A effort with the City of Hanover, the State of New
Hampshire, and the consulting firms of Hoyle and Tanner and J.I.
Associates to design an expanded/fluidized bed process to treat
43
-------
2 mgd of domestic primary effluent at the City of Hanover plant.
3. A research grant with Dr. Jewell and the City of Hanover to
evaluate the performance of retrofitted activated sludge plants
redesigned to operate anaerobically.
These studies and hopefully others to follow will refine operating and
economic characteristics of alternative approaches to anaerobic wastewater
treatment,
44
-------
SECTION 6
RECOMMENDATIONS
The scale of the proposed ANFLOW project at Oak Ridge, Tennessee should
be more than sufficient to determine the operating characteristics and
performance of an ANFLOW reactor. Presumably the $3.50/cu ft packing will
be installed to test the performance with a more economically competitive
media. However, it is doubtful that this will improve the performance.,
Since the process appears to be possible competitive with conventional systems
only at very small scale, the 189 cu m (50,000 gpd) pilot plant should not
be significantly smaller than any full scale systems which could follow.
Additional supportive experimental studies are judged to be unwarranted
at this time.
Effort should be directed to further evaluation of expanded/fluidized
beds to gather needed design, performance and economic data. Areas where
further information is required are summarized in Section III.
45
-------
REFERENCES
1. Jewell, W. J., et al. "Sewage Treatment with the Anaerobic Attached
Microbial Film Expanded Bed Process", Presented at 52nd Annual Water
Pollution Control Federation Conference, Houston, Texas, October 1979
2. Jewell, W. J. "Development of the Attached Microbial Film Expanded Bed
Process for Aerobic and Anaerobic Waste Treatment", Presented at the
"Biological Fluidized Bed Treatment of Water and Wastewater Conference",
England, April 1980
3. Jeris, J. S. and Owens, R. W. "Pilot Scale High Rate Biological Denitri-
ficaton at Nassau County, N. Y.", Presented at "New York Water Pollution
Control Association Winter Meeting", January 1974
4. "Summary Report - Pilot Plant Test of a Heat Treat Liquor Using a HY-FLO
(TM) Fluidized Bed Treatment System", Ecolotrol Environmental Systems,
Presented at "Workshop Seminar on Anaerobic Filters for Wastewater
Treatment", Howey-in-the-Hills, Florida, January 1980
5. "Summary Report - Pilot Plant Test of a Starch Waste Using a HY-FLO (TM)
Fluidized Bed Treatment System", Environmental Report, Ecolotrol Environ-
mental Systems, New York
6. Jeris, J. S. "Presentation at "Workshop Seminar on Anaerobic Filter for
Wastewater Treatment", Howey-in-the-Hills, Florida, January 1980
7. Genung, R. K., et al. "Energy Conservation and Scale-Up Studies for a
Wastewater Treatment System Based on a Fixed-Film, Anaerobic Bioreactor",
Presented at "Second Symposium on Biotechnology in Energy Production",
Gatlinburg, Tennessee, October 1979
8. Koon, J. H., et al., "The Feasibility of an Anaerobic Upflow Fixed-Film
Process for Treating Small Sewage Flows", Presented at "Energy Opti-
mization of Water and Wastewater Management for Municipal and Industrial
Applications Conference", New Orleans, Louisiana, December 1979
9. Genung, R. K., et al. "Development of a Wastewater Treatment System Based
on a Fixed-Film, Anaerobic Bioreactor", Presented at "Workshop Seminar on
Anaerobic Filters for Wastewater Treatment", Howey-in-the-Hills, Florida
January 1980
10. Jared, D., Oak Ridge National Laboratory, Letter to I. Kugelman of EPA,
April 8, 1980
11. Cillie, G. G., et al. "Anaerobic Digestion - IV. The Application of the
Process in Waste Purification", Water Research, 3^, 623, 1969
12. Mueller, J. A. and Mancini, J. L. "Anaerobic Filter Kinetics and
Application", Proc. 30th Purdue Ind. Waste Conference, 423, 1975
46
-------
13. , Lettinga, 6. et al. "Anaerobic Treatment of Methanobic Wastes",
Water Research, JJ3, 725, 1979
14. Haug, R. T., et al. "Anaerobic Filter Treats Waste Activated Sludge",
Water and Sewage Works, 124, No. 2, 40, 1977
15. Coulter, J. B., et al. "Anaerobic Contact Process for Sewage Disposal",
Sewage and Industrial Wastes, _29, 468, 1957
16. Witherow, J. L., et al. "Anaerobic Contact Process for Treatment of
Suburban Sewage", Proc. American Society of Civil Engineers, Journal
Sanitary Engineering Division, Paper 1849, SA6, November 1958
17. Fall, E. B., Jr. and Kraus, L. S., "The Anaerobic Contact Process in
Practice", Journal Water Pollution Control Federation, ^3, 1038, 1961
18. Pretorius, W. A., "Anaerobic Digestion of Raw Sewage", Water Research,
!5, 681, 1971
19. Young, J. C. and McCarty, P. L., "The Anaerobic Filter for Waste Treatment,"
Technical Report No. 87, Department of Civil Engineering, Stanford
University, March 1968
20. Smith, P.H., "Studies of Methanogenic Bacteria in Sludge," U.S. Environ-
mental Protection Agency, EPA-600/2-80-093, August 1980
21. McCarty, P. L. "Anaerobic Waste Treatment Fundamentals IV. - Process
Design", Public Works, December 1964
22. Miller, D. 6. "Fluidized Beds in Water Treatment - A Short Historical
Introduction", Presented at the Biological Fluidized Bed Treatment of
Water and Wastewater Conference", England, April 1980
23. Cooper, P. F. and Wheeldon, D. H. V., "Fluidized- and Expanded-Bed
Reactors for Wastewater Treatment", Water Pollution Control, 79, 286,
1980
24. Richardson, J. F. "Incipient Fluidization and Particulative Systems",
in Fluidization, edited by Davidson, J. F. and Harrison, D., Academic
Press, 1971
25. Cleasby, J. L. and Baumann, E. R., "Backwash of Granular Filters Used
in Wastewater Filtration", Environmental Protection Technical Series,
EPA-600/2-77-016, April 1977
26. Garalde, J. and Al-Dlbouni, M. "Velocity-Voidage Relationships for
Fluidization and Sedimentation in Solid-Liquid Systems", Ind. Eng.
Chem., Process Des. Dev., 16, 206, 1977
27. Switzenbaum, M. S. and Jewell, W. J. "The Anaerobic Attached Film Ex-
panded Bed Reactor for the Treatment of Dilute Organic Wastes", TID-
29398, National Technical Information Service, Department of Commerce,
Springfield, Virginia, August 1978
47
-------
28. Jeris, J. S., et al. "High Rate Biological Denitrification Using
a Granular Fluidized Bed", Journal Water Pollution Control Federation,
46, 2118, 1974
29. Harremoes, P. "Biofilm Kinetics", in Water Pollution Microbiology
Vol. 2, edited by Mitchell, R., Wiley-Interscience, New York, 19/8
30. Process Design Manual for Sludge Treatment and Disposal, U.S. Environ-
mental Protection Agency Technology Transfer, EPA-625/1-74-006,
October 1974
31. Lange's Handbook of Chemistry, Table 10-1, 12th Edition, McGraw-Hill
Book Company, New York 1979
32. Symons, G. E. and Buswell, A. M., Journal American Chemical Society,
tt, 2028, 1933
33. Quality of Surface Waters of the United States, 1966, Geological
Survey Water Supply Papers 1991, 1992 and 1995, U.S. Department of
Interior
34. McCarty, P.L., "Energetics of Organic Matter Degradation," in Hater
Pollution Microbiology edited by Mitchell, R., Wiley - Interscience,
New York 1972
35. Bryant, M.P., "Growth of Desulfovibrio in Lactate or Ethanol Media Low
in Sulfate in Association with \\2~ Utilizing Methanogenic Bacteria,"
Appl. and Environ. Micro., 33, 1162, 1977
36. Bell, B.A., et al. "Anaerobic Fluidized Bed Treatment of Thermal
Sludge Conditioning Decant Liquor",' The George Washington University,
Washington, D. C., 1980
37. Montrose 201 Facility Plan, Draft Final Report, prepared by Roy F.
Weston, Inc., November 1980
38. Jewell, W. J., "Mini Step 1 Study for Hanover, New Hampshire," J. I.
Associates, Inc., March 1981
39. O'Rourke, J. T., "Kinetics of Anaerobic Waste Treatment at Reduced
Temperatures," PhD Thesis, Stanford University, 1968
40. Perry, R. H., and Chilton, C. H., Chemical Engineers' Handbook, Table
18-6, Fifth Edition, 1973
41. Sargent-Welch Catalog 128, Cincinnati, 1979
42. Griffith, W. L., "Economics of the ANFLOW Process for Municipal Sewage
Treatment", Oak Ridge National Laboratory, ORNL/TM-6574, March 1979
43. Karkheck, J., et al. "Prospects for District Heating in the United
77
48
States", Science, 195, 948] 1977
I
-------
44. Perry, R. H., and Chilton, C. H., Chemical Engineers' Handbook, Table
3-203, Fifth Edition, 1973
45. Chemical Marketing Reporter, Schnell Publishing Company,
April 7, 1980
46. Jeris, J. S. and McCarty, P. L. "The Biochemistry of Methane Fermentation
Using C Tracers", Journal Water Pollution Control Federation, 37,
178, 1965
47. Rossman, L. A. "EXEC/OP Reference Manual. Version 1.2", EPA Municipal
Environmental Research Laboratory, .February 1980
48. Rossman, L. A. "Computer-Aided Synthesis of Wastewater Treatment
and Sludge Disposal Systems", EPA-600/2-79-158, 1979
49. Gumerman, R. C. et al., "Estimating Water Treatment Costs: Volume 2,
U.S. Environmental Protection Agency, EPA-600/2-72-162b, August 1979
49
US.GOVERNMENTPRINTINSOFFICE: 1982-559-09E/3371
-------
-------
-------
-------
-------
CD f;
03 O
C Q. =* ST O. TJ
5 £ < -* £ CD"
T? nt O T 5) Q>
r o t/>
0)
- (D
hd
O\
o
o
N5
I
oo
to
o
o
.p-
?s?
i- o o
o
Ml
ii
i|
01 ^
o 3-
A O>
c
> m
(O 3
CD <
3 =;'
O O
< 3
3
CD
3
TJ
O
CD
O
Q 3
II
0) ~
s*. o'
O3
01
to
05
00
c/>
CO
0
0)
-n
o
00 C
o 3
0 3-
^ O
01
C/l
(A
30
CD
m J>
>!
CO 0
00 <
01
f
-a m
3 <
o o
C±. 3
O ^
3 qj
3
QJ
in i
|0I Ik
|O
r
-n
CD <
CD |
Q> 1
!
i
"%
V
k
M
ml
u
-------