United States
Environmental Protection
Agency
Water Engineering
Research Laboratory
Cincinnati OH 45268
Research and Development
EPA/600/S2-85/040 Sept. 1985
Project Summary
The Loves Creek Anaerobic,
Upflow (ANFLOW) Pilot Plant:
Performance Summary
Michael T. Harris, Terrence L. Donaldson, Richard K. Genung, Angel L
Rivera, and Charles W. Hancher
The performance of an anerobic, up-
flow (ANFLOW) fixed-film bioreactor
was studied on a near-commercial
scale in Knoxville, Tennessee, with a
190-m3/day facility from August 1981 to
October 1983.
During treatment of low-strength
municipal wastewater before primary
sedimentation, the effluent met the
EPA secondary treatment discharge
limits of 30 mg/L for TSS and BOD the
vast majority of the time, with only an
occasional increase to the 30- to 40-
mg/L range for perhaps 1 to 2 days.
Loading rates were -0.25 kg/m3-day of
TSS and BOD each, and the hydraulic
retention time was 9 to 10 hours. This
performance was maintained in ambi-
ent cold weather tests (~12°C water
temperature), though the rate of solids
accumulation in the bioreactor was
higher in cold weather because of de-
creased biological activity, which nor-
mally converts solids to off-gas.
The primary mechanism for remov-
ing TSS and BOD appears to be bio-
physical filtration. Approximately 20%
of the influent carbon was converted to
methane, and 30% remained in the
bioreactor as sludge. The balance of the
carbon was converted to CO2 or re-
mained in the liquid effluent as TSS and
BOD/COD. More than half of the
methane was dissolved in the effluent
water; this methane could potentially
be recovered along with the methane in
the off-gas.
Dry solids accumulated in the biore-
actor at the rate of -150 kg/3800 m3
(3800 m3 = 1 million gal) of wastewater
treated, which represents a 75% to 80%
reduction in solids production com-
pared with primary sedimentation and
activated sludge or trickling filtration.
The rate was slightly higher in cold
weather and somewhat lower in warm
weather. A simple material balance
model was shown to predict the sludge
accumulation using measurable influ-
ent and effluent parameters.
Gas production increased at higher
loading rates with higher strength
wastewater. Solids accumulation also
increased, probably as a partial result of
the solids content of the feed, which
was formulated by mixing primary clar-
ifier underflow with the raw waste-
water. Performance of the ANFLOW
process under these conditions was
ambiguous, however, because the
bioreactor was heavily loaded with ac-
cumulated solids when this test was
run.
After ~2 years of operation, the pilot
plant was decommissioned and the site
was returned to the City of Knoxville, as
required by contractual agreements.
This Project Summary was devel-
oped by ERA'S Water Engineering Re-
search Laboratory, Cincinnati, OH, to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
Recent development of the upflow
anaerobic filter, the upflow anaerobic
sludge blanket, and the attached film
expanded bed, along with a better un-
derstanding of the microbiology of
anaerobic processes have suggested
the feasibility of anaerobic bioactivity
for treating low-strength wastewater at
temperatures <20°C. The fixed-film pro-
-------
cesses are especially attractive because
the bacteria are retained on the packing
material and are not washed off in the
effluent. Thus mean cell residence
times on the order of 100 days can be
obtained. Also, the fixed-film systems
have been postulated to be less suscep-
tible to toxic materials because films ex-
posed to these materials may slough
off, leaving lower levels of film viable
for continued wastewater treatment.
The development of anaerobic filter
technology for treating municipal
wastewater has been pursued by Oak
Ridge National Laboratory (ORNL) since
1976. The ANFLOW (ANaerobic up-
FLOW) process has been operated suc-
cessfully at the laboratory on a
19-m3/day and a 190-m3/day (5,000- to
50,000-gal/day) scale. Performance over
this range has been consistent and pre-
dictable; no scale-up problems have
been encountered. A commercial AN-
FLOW unit of the latter size could handle
the municipal wastewater produced by
250 to 500 persons. Larger systems
could be constructed using this basic
model.
Facility Description
The ANFLOW pilot plant was located
at the Loves Creek Wastewater Treat-
ment Plant in Knoxville, Tennessee. The
pilot plant was designed to process
-190 m3day (50,000 gal/day) of munici-
pal wastewater after pretreatment (bar
screening and grit settling) but before
primary clarification.
The ANFLOW bioreactor was basi-
cally a cone-bottomed, mild-steel tank
constructed to American Petroleum In-
stitute (API) 620 Code Specifications.
The tank was 4.9 m (16 ft) in diameter
and 5.4 m (18 ft) high with a 3-m-(10-ft-)
high packed section, the internal sur-
faces were coated with Amercoat* coal-
tar epoxy paint. The entire system was
mounted on truck scales to provide a
continuous readout of its weight.
The 3-m section above the cone was
packed with 3-in. polypropylene Pall
rings manufactured by Glitsch, Inc.,
which were floated into place randomly.
This material is relatively inexpensive,
gives a high void fraction in the bioreac-
tor, has an acceptably high packing sur-
face area, and has an open structure to
facilitate intermittent backwashing of
solids from the bioreactor. Sampling
ports were located at several axial and
'Mention of trade names or commercial products
does not constitute endorsement or recommenda-
tion for use.
radial positions in the bioreactor. The
gas headspace above the normal liquid
volume was ~7.5 m3. The off-gas was
monitored and safely vented through
the roof. The incoming wastewater
could be preheated to ~25°C in a verti-
cal, double-pipe heat exchanger. Liquid
effluent was recycled to the inlet of the
Loves Creek Plant.
The ANFLOW bioreactor was inocu-
lated with sewage sludge in August
1981. Operation during the start-up pe-
riod (through September 1982) is de-
scribed elsewhere. (A. L. Rivera, T. L.
Donaldson, R. K. Genung, M. T. Harris,
and C. W. Hancher, The Loves Creek
Anaerobic, Upflow (ANFLOW) Pilot
Plant: Design and Start-Up, ORNL/TM-
8828, Union Carbide Corp. Nuclear Div.,
Oak Ridge Nat I. Lab., April 1984). This
report describes the performance of the
system under ambient temperatures
during the 1982-83 winter and under in-
creased loading rates during the sum-
mer of 1983.
Performance Summary
During treatment of low-strength mu-
nicipal wastewater before primary sedi-
mentation (when TSS and BOD were
typically 100 mg/L each), the effluent
met the EPA secondary treatment dis-
charge limits of 30 mg/L for TSS and
BOD the vast majority of the time, with
only an occasional increase to the 30- to
40-mg/L range for a brief period of time
(perhaps 1 to 2 days). Loading rates
were -0.25 kg/m3«day of TSS and BOD
each, and the hydraulic retention time
was 9 to 10 hours.
The bioreactor was operated at ambi-
ent temperatures from October 1, 1982,
to March 16, 1983. Then the feed pre-
heater was turned on to simulate the
spring and summer warm-up period.
After that followed high-strength waste-
water study in which primary clarifier
underflow was added to the wastewater
feed.
The performance throughout this pe-
riod is summarized in Table 1 for the
various operating conditions during the
period from October 1982 to September
1983. In general, the good removal of
TSS and BOD achieved during the
start-up period was maintained in ambi-
ent cold weather tests (~12°C water tem-
perature), although the rate of solids ac-
cumulation in the bioreactor was higher
in cold weather because of decreased
biological activity, which normally con-
verts solids to off-gas. These solids
were partially digested later during the
warm weather operation.
The primary removal mechanism for
TSS and BOD appeared to be biophysi-
cal filtration. Approximately 20% of the
influent carbon was converted to
methane, and 30% remained in the
bioreactor as sludge. The balance of the
carbon was converted to C02 or re-
mained in the liquid effluent as TSS and
BOD/COD. More than half of the meth-
ane was dissolved in the effluent water;
this methane could potentially be recov-
ered along with the methane in the off-
gas.
Gas production increased at higher
loading rates during the high-strength
wastewater study. Solids accumulation
also increased, probably as a partial re-
sult of the high solids content of the
wastewater formulated by mixing pri-
mary clarifier underflow with the raw
wastewater. In terms of effluent water
quality, performance of the ANFLOW
process under these conditions is am-
biguous because the bioreactor was
heavily loaded with accumulated solids
when this test was run.
Analyses for total phosphates, ortho-
phosphates, TKN, NH3-N, sulfates, sul-
fides, and volatile acids were done on
influent and effluent samples periodi-
cally. The average total concentrations
of phosphates and orthophosphates
were 2.5 and 2.0 mg/L (three samples),
respectively, in the influent, and 3.0 and
2.0 mg/L (four samples), respectively, in
the effluent. Respective averages for in-
fluent and effluent total nitrogen were
7.8 and 9.0 mg of N/L (three samples
each), and for NH3-N, they were 7.0 and
7.8 mg/L (four samples each). These re-
sults suggest that <15% of the TKN was
organic nitrogen.
Daily influent and effluent sulfate con-
centrations averaged 37 and 26 mg/L,
respectively (four samples). Sulfide
concentrations were 0 and 6 mg/L for
the influent and effluent streams. These
data correspond with a sulfate reduc-
tion and sulfide production rate of -1
kg/day (or 0.013 kg/m3-day as sulfide.
The concentration of sulfide in the liquid
effluent was well below toxic levels
(e.g., >100 mg/L) that have been known
to inhibit methanogenesis.
Liquid samples were obtained period-
ically at 0,4, and 8 ft from the bottom of
the packed section. Removal of TSS and
total COD/BOD was greater in the lower
portion of the packed section, whereas
removal of soluble COD/BOD was ap-
proximately uniform throughout the
packed section.
During the cold weather period, the m
-------
Table 1. Operation and Performance Summary for the 190-m3/day ANFLOW Pilot Plant*
Raw Wastewater
Plus Primary
Clarifier
- Underflow,
Cold Temperature Temperature
Weather Control to >20°C Control to >20°C
Raw Wastewater
Operating or Performance
Parameter
Temperature of feed, °C 12-18 20-25 22
Period of operation, days 137 213 61
Hydraulic loading rate, 10.0 7.3 7.3
m3/m2-day
TSS loading rate, kg/m3-day 0.15 0.12 0.73
BOD loading rate, kg/m3-day 0.13 0.23* (0.14)* 0.40
COD loading rate, kg/m3-day 0.35 0.30 1.2
TSS removal efficiency, % 80 80 78
BOD removal efficiency, % 63 69f(50>* 70
COD removal efficiency, % 50 57'(45)* 71
Methane yield coefficient, 50 160 110
L CH4/kg COD removed
Volumetric CH4 production rate, 9 24 94
L CH4/m3 reactor-day
Concentration of CH4 in 3-9 63-76 60-80
off-gas, %
Sludge production rate, 170 200*1120)* 575
kg dry solids/3800 m3 of
wastewater treated
"All average values were derived from monthly averages.
1Values observed during the start-up period.
* Values observed during the increasing-temperature period.
influent and effluent concentrations of
volatile acids were <1 mg/L. These data,
along with the fact that very little off-gas
was produced during this period, indi-
cated a substantial decrease in anaero-
bic bioactivity in the ANFLOW column
during the cold weather period. How-
ever, effluent remained good because
removal of TSS and BOD continued by
biophysical filtration.
rate of VSS rate of VSS
+ production — destruction
(kg/d) (kg/d)
rate of inert rate of inert
+ solids in - solids out
(kg/d) (kg/d).
Sludge Accumulation
The amount of sludge in the bioreac-
tor was estimated independently using
a material balance model and the mea-
sured change in weight of the ANFLOW
column. The mass balance equation fol-
lows:
Rate of TSS
accumulation
(kg/d)
rate of VSS rate of VSS
in (kg/d) - out (kg/d)
The quantities on the right side of the
equation can be expressed in terms of
measurable or otherwise known param-
eters.
The accumulation of dry solids in the
bioreactor is plotted in Figure 1. Accord-
ing to the model, -850 kg of dry solids
(or 170 kg solids/3800 m3 wastewater
treated) accumulated in the column dur-
ing the cold weather period (days 483
through 588). The data in Figure 1 also
show the solids accumulation esti-
mated by the measured change in
bioreactor weight. To convert bioreac-
tor weight to the mass of solids in the
reactor, it was assumed that the total
volume occupied by the liquid and
solids was constant, and that the densi-
ties of the water, volatile solids, and dry
inert solid had constant values of 1.0,
1.0, and 2.5 g/mL, respectively. The data
in Figure 1 show excellent agreement
between the model and the experimen-
tal data over the entire operating period
of -750 days.
Discharge of solids was carried out
several times late in the project. Simple
draining of the bioreactor removes
some solids, but not as much as is prob-
ably desired. Techniques need to be de-
veloped to improve the solids discharge
procedures. Solids can be dislodged
from the packing very easily by a gentle
stream of water from a hose. Approxi-
mately 5500 kg of dry solids was re-
moved from the bioreactor when the pi-
lot plant was decommissioned, which is
in reasonable agreement with the
—6800 kg indicated in Figure 1.
Residence Time Distribution
Tests
Residence time distribution tests
were conducted before inoculation and
after —5000 kg of dry solids had accu-
mulated in the bioreactor. An inert fluo-
rescein dye was used. The flow patterns
before inoculation were adequately de-
scribed by three unequal-volume tanks
in series with volume ratios of 1:3:6.
The latter flow patterns could be de-
scribed by two equal-volume tanks in
series. However, a substantial reduction
occurred in the mean residence time;
approximately one-sixth of the total re-
actor volume was used, while the other
five-sixths were occupied by solids
and/or stagnant dead-volume zones.
Metal Test Coupons
Type 316 stainless steel metal test
coupons were evenly coated with seven
types of paint and placed in the top of
the ANFLOW column at the gas-liquid
interface for exposure to both the liquid
and gas phases. The Amercoat 395/395
paint sample showed no signs of peel-
ing and no indication of reactivity with
the corrosive atmosphere after -800
days. The Sikagard 62-gray also
showed high resistance to the corrosive
atmosphere. Other paints showed sig-
nificant deterioration. Overall, the gas
phase was more reactive to the paints
than the liquid phase.
Energy Conservation
ANFLOW is an energy-conserving
technology. Earlier studies have indi-
-------
10.000
8,000
6,000
4,000
2,000
T
T
• Bioreactor Weight
— Material Balance Model
I
I
I
I
\
I
Figure 1 .
0 200 400 600
Time After Inoculation (days)
A ccumulation of solids in the A NFL O W bioreactor.
800
100O
cated that at the 3800-m3/day (1-mgd)
scale, a treatment plant using the AN-
FLOW process would need only 60% of
the energy required for a plant using
activated sludge to treat weak waste-
waters, or it would need only 30% of the
energy required to treat strong waste-
waters. When only the biotreatment
step is considered, ANFLOW is pro-
jected to need less than 10% of the en-
ergy required by activated sludge.
Methane recovery is not included in
these comparisons, but recovery of
methane and conversion to electricity at
40% efficiency are projected to provide
more than sufficient energy to operate
the ANFLOW-based treatment plant.
The full report was submitted in fulfill-
ment of Interagency Agreement EPA
No.AD-89-F-2-A173 by the Oak Ridge
National Laboratory under the sponsor-
ship of the U.S. Environmental Protec-
tion Agency. The work described in this
report was also supported in part by the
U.S. Department of Energy.
U. S. GOVERNMENT PRINTING OFFICE;1985/559-l 11/20707
-------
-------
Michael T. Harris, TerrenceL Donaldson, RichardK. Genung, AngelL Rivera, and
Charles W. Hancher are with Oak Ridge National Laboratory, Oak Ridge, TN
37831.
Ronald F. Lewi's is the EPA Project Officer (see below).
The complete report, entitled "The Loves Creek Anaerobic, Upflow (ANFLOW)
Pilot Plant: Performance Summary," (Order No. DE 85-010 578; Cost $ 11.95,
subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Water Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati. OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/600/S2-85/040
tCU W TILLcY
RtGION V EPA
LJSPARI2N
230 S OcAS^CRN ST
CHICAGO
A 060*
------- |