Wat. Res. Vol. 34, No. 14, pp. 3640ą3656, 2000
7 2000 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
PII: S0043-1354(00)00114-7 0043-1354/00/$ - see front matter
www.elsevier.com/locate/watres
TREATMENT OF LANDFILL LEACHATE USING
SEQUENCING BATCH AND CONTINUOUS FLOW
UPFLOW ANAEROBIC SLUDGE BLANKET (UASB)
REACTORS
K. J. KENNEDY1*M and E. M. LENTZ2
1 2
Department of Civil Engineering, University of Ottawa, Ottawa, Canada and Kingett Mitchell
Environmental Consultants Ltd, Auckland, New Zealand
(First received 1 January 1999; accepted in revised form 1 December 1999)
Abstract�Treatment of municipal land�ll leachate was investigated/compared using sequencing batch
and continuous Żow upŻow anaerobic sludge blanket (UASB) reactors. All reactors were operated at
organic loading rates (OLRs) between 0.6ą19.7 gCOD/l d. Performance of the continuous UASB and
sequencing batch UASB reactors were very similar at low and intermediate OLRs. Continuous UASB
reactors performed more favorably at the higher OLRs than the sequencing batch UASB (AnSBR)
system. The sequencing batch UASB reactor had soluble chemical oxygen demand (COD) removal
e�ciencies ranging between 71% and 92% at hydraulic retention times (HRT) of 24, 18 and 12 h with
dilute to concentrated leachate feed. For the continuous UASB reactor, soluble COD removal e�ciency
was consistently between 77% and 91% for all HRTs and feed concentrations. At high OLRs and �ll
to react ratios of 0.5, the AnSBR system showed signs of system failure. In order to ensure successful
AnSBR treatment of municipal land�ll leachate the speci�c organic loading rate during the �ll cycle
(SOLR�ll) should not exceed 3 g COD/g VSS d. Both systems e�ectively resulted in a reduction in
1
Microtox toxicity levels following treatment. Municipal leachate treated by either continuous or
sequencing batch UASB reactor was not suitable for direct disposal into the sewerage according to
Municipal Sewer Regulations. A post treatment option such as membrane �ltration would be required.
7 2000 Elsevier Science Ltd. All rights reserved
Key words�anaerobic, sequencing batch reactor, SBR, land�ll leachate, UASB
INTRODUCTION
nation of high rate upŻow anaerobic sludge blanket
(UASB) reactors operating in sequencing batch
Leachate management and treatment has gained im-
reactor (AnSBR) mode.
portance as information and research surrounding
Sequencing batch reactors (SBRs) in general o�er
the operation of land�ll sites becomes more avail-
distinct advantages when compared to continuous
able (Barbre and Marais, 1984). At present, collec-
processes (Harty et al., 1993) including:
tion and treatment of land�ll leachates is one of the
most pressing issues surrounding the operation of
. high degree of process Żexibility in terms of cycle
land�ll sites (Lema et al., 1988). One of the avail-
time and sequence;
able options is biological leachate treatment by
. ability to incorporate aerobic, anoxic phases in a
either aerobic or anaerobic processes (Forgie, 1988).
single reactor (if desired);
However, anaerobic treatment methods are more
. sequencing batch closely resembles plug Żow op-
suitable for concentrated leachate streams, o�er
eration during the �ll cycle;
lower operating costs, the production of a useable
. near ideal quiescent settling conditions;
biogas product, and production of a pathogen free
. no separate clari�ers required;
solids residue which can be used as cover material
. elimination of short circuiting.
(Malina and Pohland, 1992). There are numerous
reports on the anaerobic treatment of leachate The greater process Żexibility of SBRs is particu-
using continuous anaerobic processes, however only larly important when considering the treatment of
a handful of reports have investigated the combi- land�ll leachates, which have a high degree of
variability in quality and quantity. Like all SBRs
the AnSBR undergoes a cycle composed of four
*Author to whom all correspondence should be addressed. distinct phases of operation: �ll, react, settle and
3640
Treatment of land�ll leachate using UASB reactors 3641
discharge (F:R:S:D). There may also be an optional treated wastewater is removed by a discharge port
idle (I) period in some systems if required. The located at the Vmin level of the reactor.
The objective of this study is to compare continu-
entire AnSBR cycle is represented in Fig. 1.
ous vs SBR anaerobic treatment of leachate and
At time to, the reactor contains a settled anaero-
provide AnSBR operational parameters and design
bic granular sludge at some minimum reactor
information for municipal land�ll leachate treat-
volume, Vmin. This is typically at the level of the
ment by evaluating the e�ect of leachate concen-
e�uent or discharge port. There is a supernatant
tration, HRT and F:R ratio.
liquid layer above the settled sludge. The beginning
Lin (1991) used a complete mixed semi-continu-
of the AnSBR cycle is marked by the �ll period.
ous anaerobic digester to investigate the ability of a
Wastewater is pumped into the reactor volume
feed and draw system to treat municipal leachate.
through a port at the bottom of the reactor to
Chemical oxygen demand (COD) removal e�cien-
some maximum volume, Vmax. Mixing of the reac-
cies of 92% were achieved with inŻuent COD con-
tor volume while �lling is accomplished by recycling
centrations of 22,750 mg COD/l and volumetric
liquid from the reactor and returning it to the reac-
organic loading rates (OLR) in the range of 1.1ą2.8
tor at the feed port. The total duration of the �ll
kg COD/m3 d. This resulted in long hydraulic
cycle is controlled by the inŻuent wastewater Żow
retention times (HRT) in the range of 8ą21 days.
rate.
Kennedy et al. (1988) compared continuous
Recycle is continued throughout the react period.
downŻow stationary �xed �lm (DSFF) and UASB
The duration of this period is determined by the
type reactors for the treatment of leachate. The
level of substrate removal desired and can be
UASB reactor achieved COD removal rates of 95%
manipulated in order to accommodate changes in
at OLRs and HRTs between 4.8ą14.5 kg COD/m3
substrate loading. The �ll to react (F:R) ratio is im-
d and 4ą1.5 days respectively. Chang (1988) also
portant in that it can be manipulated to accommo-
treated municipal leachate at a concentration of
date process changes. The settle period allows
58,400 mg COD/l using a continuous UASB type
granular anaerobic biomass to settle within the
reactor. COD removal e�ciencies were 81.7ą92.8%
reactor prior to e�uent discharge. In the case of
for OLR below 13 kg COD/m3 d but decreased to
the UASB reactor, granular sludge has especially 67.9% when the OLR was increased to 22 kg
good settling characteristics. After su�cient settling, COD/m3 d.
Fig. 1. Operational mode of the sequencing batch reactor.
3642 K. J. Kennedy and E. M. Lentz
MATERIALS AND METHODS
The UASB system is characterized by a meta-
bolically active granular sludge which has superb
Anaerobic granular sludge inocula originated from a
settling characteristics ensuring that excellent bio- fullscale UASB reactor treating chemi-thermo-mechanical
pulp (CTMP) mill e�uent. A settled volume of 1.4 l of
mass retention occurs independent of the HRT.
inocula with an associated total suspended solids (TSS)
These features make the UASB reactor ideal for
and volatile suspended solids (VSS) concentration of 50 g
SBR operation. Kennedy et al. (1991) demon-
TSS/l and 39 g VSS/l was added to each reactor for an in-
strated the applicability of the UASB reactor
itial reactor biomass concentration of 7.8 g VSS/l. The
speci�c anaerobic activity of the sludge was 0.3ą0.4 g
operated in sequencing batch mode (AnSBR) to
COD/g VSS d.
successfully treat soluble wastewaters. A soluble
Leachate was obtained from the Ottawa-Carleton Trail
wastewater treated at OLRs up to 9 g COD/l d
Road Municipal Land�ll Site, Nepean, Canada central
resulted in COD removal e�ciencies above 80%.
recirculation facility, in the spring (March) and in the fall
Low �ll to react (F:R) ratios were found to result (October). Leachate was stored frozen in plastic 200 l bar-
rels, which were thawed at room temperature and then
in COD treatment e�ciency reductions in excess
maintained at 48C prior to use. Characteristics of the un-
of 25% for the same OLR cycle. It was deter-
diluted leachate are provided in Table 1.
mined that the speci�c organic loading rate
during the �ll period (SOLR�ll) was critical and
Acclimation of reactors: initial start-up procedure
should not exceed 1 g COD/g VSS d when treating
The anaerobic consortia was acclimated over 3 months
sucrose based wastes. It was hypothesized that the
to a VFA/sucrose feed (4000 mg COD/l; Kennedy et al.,
critical SOLR�ll would rise for volatile fatty acid
1991) and leachate which was fed continuously to all reac-
(VFA) based wastewaters. The results from this tors at a 24-h HRT. At the end of the acclimation period,
biomass from all reactors was combined and 1.4 l of
study were eventually used to develop the �rst
settled granular sludge was added to each reactor to give
mathematical AnSBR model (Fernandes et al.,
7.8 g VSS/l respectively (based on total liquid volume of
1993).
reactor).
Leachate (1120ą3520 mg COD/l) from the Iowa
City municipal land�ll was treated using a UASB
Anaerobic reactors
sequencing batch reactor (358C) (Hollopeter and
Three cylindrical 15.6 cm diameter glass upŻow anaero-
Dague, 1995). The AnSBR was operated at an
bic sludge blanket reactors operating as sequencing batch
HRT of 6 h (OLRs of 1.6ą3.5 g COD/l d) with
reactors (AnSBRs 1,2,3) (Fig. 2) with total and e�ective
liquid volumes of 6.2 and 5.0 l respectively were used.
�ll, react, settle and discharge phases of 0.25,
InŻuent was introduced at the bottom of the AnSBRs
4.5, 1.0 and 0.25 h, respectively. Soluble COD
using a peristaltic pump. The level of the settled biomass
removals were consistently between 85% and
surface was just below the withdraw/e�uent port. During
90%. A decrease in the solids retention time
the withdraw phase a peristaltic pump was used to remove
(SRT) was observed which was attributed to an treated liquid from the e�uent/withdraw port at approxi-
mately the 1.5 l level of the reactor. A solenoid valve was
accumulation of inorganics within the settled
used during the withdraw phase to open the reactor to at-
sludge volume thereby forcing active biomass out
mosphere and allow e�uent to be removed without pull-
of the system. Unlike the Kennedy et al. (1991)
ing a vacuum that would a�ect biogas readings. The
study, Hollopeter and Dague's study did not
overŻow port, which was �tted with a glass manometer,
was located at a volumetric height of 5.0 l. Two recycle
have a control reactor, HRT was not varied and
ports were located equidistant between the e�uent and
�ll to react (F:R) ratios were not speci�cally
overŻow port. Recycle at a rate of 11.4 ml/min was with-
investigated
drawn from the recycle port and returned via the feed
port.
Mixed liquor from the reactor was obtained during the
Table 1. Undiluted trail road land�ll leachate
�ll and react phases. A programmable timer controlled the
solenoid valve and pumps. Biogas exited the top of the
Parameter Value, mg/lb Parameter Value, mg/lb
reactor and was measured by a Wet Flip Meter. All reac-
tors were maintained at 358C.
Conductivity 1.7ą13.5a Be < 0.001ą < 0.002
Over a period of 13 months three parameters were
pH 6.9ą9.0 Cd < 0.008ą0.012
investigated HRTs of 24, 18 and 12 h, F:R ratios of 0.5,
COD 3210ą9190 Co < 0.008ą0.011
1 and 2 and leachate concentration of 100, 66 and 33%
Al < 0.02ą0.92 Cr < 0.015ą0.092
by volume (dilution with tap water). Settle and withdraw
B 1.82ą9.63 Cu 0.008ą0.061
times were �xed at 2 h and 1 h respectively, regardless of
Ba 0.006ą0.164 Mo 0.004ą < 0.015
the conditions being investigated. For HRTs of 24, 18
Ca 2.15ą113.90 Ni 0.02ą0.27
and 12 h, the combined �ll plus react times were 21, 15
Fe 1.28ą4.90 Pb < 0.04ą < 0.06
K 161.9ą1993.8 Zn 0.035ą0.429 and 9 h respectively. Reactors were run for a minimum
Li 0.017ą0.171 V 0.31ą0.32
of 2 weeks at any particular condition and performance
Mg 181.40ą627.48 SO4 < 20ą165
was based on the average of the last 5 cycles (pseudo-
Mn 0.028ą1.541 F < 1
steadystate).
Na 672.40ą1748.46 Cl 755.91ą3035
A single laboratory scale continuously fed UASB
P < 0.5ą27.0 NO2 < 1ą5
reactor was operated as a control. UASB reactor con-
S 7.44ą36.22 Br < 5ą22.84
�guration, sludge inocula, recycle rate, start up con-
Si 3.72ą10.48 NO3 < 2ą11
ditions as well as HRTs and waste concentrations tested
Sr 0.137ą5.005 HPO4 <10
were identical to the AnSBRs. Table 2 summarizes oper-
a
ating parameters for the AnSBRs and continuous UASB
mmho/cm.
b
Unless otherwise speci�ed, units are mg/l. reactor.
Treatment of land�ll leachate using UASB reactors 3643
Fig. 2. UASB sequencing batch reactor schematic.
EXPERIMENTAL ANALYSES
man's (1979) method on a Hewlett-Packard (HP)
5840A Gas Chromatograph (GC) equipped with a
Chemical oxygen demand
packed column, Żame ionization detector, HP-7671
COD of all samples was determined using autosampler and HP-3380 integrator.
Knetchel's (1979) colorimetric method.
Biogas composition
Metallic and non-metallic elements
The CH4 and CO2 composition of the reactor
Analysis of metallic and non-metallic elements biogas, was determined using the method of Van
were performed using a Thermo-Jarrell Ash Atoms- Huyssteen (1967). A HP-5710A GC with a Poropak
can 25 Inductively Coupled Argon Plasma Atomic T porous polymer packed column and thermal con-
Emission Sequential Spectrophotometer (ICAP- ductivity detector with nitrogen carrier gas was
AES) and a Dionex DX100 Ion Chromatograph. used.
Volatile fatty acids (VFA) Toxicity
Volatile fatty acids were measured using Ack- Toxicity of inŻuent and treated e�uent was
Table 2. Summary of the operating parameters for AnSBRs and continuous UASB
AnSBR Operating scenarios: % leachate feed (by volume), ya, F:R (�ll to react ratio)
1 100% leachate feed 100% leachate feed 100% leachate feed
y=24 h y=18 h y=12 h
F:R=0b,1/2, 1/1, 2/1 F:R=0b, 1/2, 1/1, 2/1 F:R=0b, 1/2, 1/1, 2/1
2 66% leachate feed 66% leachate feed 66% leachate feed
y=24 h y=18 h y=12 h
F:R=0b, 1/2, 1/1, 2/1 F:R=0b, 1/2, 1/1, 2/1 F:R=0b, 1/2, 1/1, 2/1
3 33% leachate feed 33% leachate feed 33% leachate feed
y=24 h y=18 h y=24 h
F:R=0b, 1/2, 1/1, 2/1 F:R=0b , 1/2, 1/1, 2/1 F:R=0b, 1/2, 1/1, 2/1
a
F:R=0 refers to continuous UASB control reactor.
b
y represents HRT.
3644 K. J. Kennedy and E. M. Lentz
1
determined by the standard Microtox toxicity pro- for the continuous UASB reactor improved COD
tocol. removal performance over AnSBR performance is
evident when treating concentrated leachate (100%)
Total suspended solids and volatile suspended solids
at the shortest HRT (12 h) in SBR mode. The
Total suspended solids (TSS) and volatile sus- AnSBR treating 100% leachate at a 12 h HRT
pended solids (VSS) were determined according to showed signs of stress characterized by a marked
Standard Methods for the Examination of Water decrease in both the soluble COD removal e�ciency
and Wastewater, 17th Edition (APHA, 1989), and accumulation of VFA in the reactor (character-
Method 2540 D. All analyses were performed in istics of system failure). This is discussed later when
duplicate. considering speci�c organic loading rates (SOLRs).
AnSBR 2, treating 66% leachate feed, demonstrated
a general increase in the soluble COD removal e�-
RESULTS AND DISCUSSION
ciency with decreased OLRs. The VFA removal e�-
A general summary of the average VFA and sol- ciency was very high throughout all testing
uble COD inŻuent and e�uent concentrations for scenarios, in excess of 91%. AnSBR 3, treating
all AnSBR and continuous UASB reactor con- 33% leachate feed, irrespective of the particular
ditions tested appear in Table 3. Table 4 details the F:R ratio, had soluble COD removal e�ciencies in
relationship between soluble COD removal e�- excess of 80%. However, there did not seem to be
ciency, F:R ratio and HRT for the AnSBRs and any di�erence in removal e�ciency for similar OLR
continuous UASB reactor operated under pseudo but di�erent F:R ratios. These observations would
steadystate conditions (results of last 5 cycles). indicate that under these operating conditions the
Tables 3 and 4 indicate that, for the majority of anaerobic microbial consortia in the AnSBR was
conditions, consistently high removal e�ciencies relatively un-stressed.
were achieved for both soluble COD and VFAs. Local sewer use regulations specify that the maxi-
This is likely due to the relatively high biodegrad- mum allowable BOD5 concentration allowed for
ability of the leachate used in this study which had disposal into the regional sewerage system is
a BOD5:COD ratio of 0.86. Based on the BOD5:- 300 mg/l. For inŻuent COD concentrations in the
COD ratio of 0.86 implies those COD removal e�- range of 3210ą9190 mg/l (range in this study), the
ciencies greater than 81% were in fact degrading in corresponding BOD5 concentration based on the
excess of 95% of the biodegradable material pre- BOD5:COD ratio determined for the leachate (0.86)
sent. Table 4 and Figs 3ą5 indicate that for the ma- would be 2760ą7903 mg/l. Using performance data
jority of situations (longer HRT and lower OLR) from Table 3 and the BOD5:COD ratio, both the
VFA and COD e�uent removal e�ciencies of continuous UASB and AnSBR treating 100% lea-
AnSBRs operated at the same HRT (approx. same chate (worst-case scenario to ensure compliance)
OLR) were relatively una�ected by the F:R strategy produced a BOD5 e�uent that would exceed the
used to control the reactor. Decreasing the F:R sewer use regulations. For this particular case it
ratio from 2.0 to 0.5 while maintaining the same would seem that anaerobic treatment is best suited
HRT and similar OLR did not result in a strong as a pretreatment method.
pattern of lower removal e�ciencies when treating Individual AnSBR test scenarios evaluated the
33 and 66% leachate. Additionally, it would seem e�ect of F:R ratio while maintaining the HRT con-
that the continuous reactor behaved as well or stant (i.e. similar OLR). Soluble COD, and VFA
slightly better in terms of COD removal than concentrations over a particular reactor cycle
AnSBRs at similar loading conditions. However, formed a pro�le of the AnSBR's performance
Table 3. Overall COD and VFA inŻuent and e�uent characteristics for AnSBRs and continuous UASB reactor
Average soluble COD levels, mg/l (225 mg/l) Average VFA levelsa, mg/l (21%)
Average inŻuent Average e�uent Average inŻuent Average e�uent
concentration concentration concentration concentration
AnSBR reactors
100% leachateb 3210ą9190 650ą2100 1100ą2200 < 10ą95
66% leachate 2310ą7810 230ą1405 680ą1840 < 10ą90
33% leachate 1260ą3500 < 50ą300 430ą1650 < 10ą75
Continuous UASB
reactor
100% leachateb 4800ą9840 600ą1750 1002ą1990 < 10ą99
66% leachate 2640ą7920 200ą1288 540ą1740 < 10ą70
33% leachate 1200ą4320 < 50ą350 460ą1900 < 10ą57
a
Acetic acid, butyric acid and propionic acid.
b
Excluding results for Reactor 1 treating 100% leachate feed for a 12 h HRT (F:R ratios of 2.1, 1.0 and 0.5) which showed initial signs
of reactor failure.
Treatment of land�ll leachate using UASB reactors 3645
under speci�c HRT and F:R ratio conditions (Figs during the reaction phase, as occurred for other
6ą8). In all cases, the concentration of both VFAs operating conditions. COD and VFA levels contin-
and soluble COD within the reactor increased shar- ued to increase beyond the point when recycle was
ply during the �ll phase. The concentration of these initiated, and overall removal of VFAs and soluble
parameters increased to a maximum concentration, COD was poor. It is likely that inhibition of aceto-
which was less than the inŻuent concentration, clastic methanogens occurred resulting in the ac-
owing to dilution (varying liquid volume) and sim- cumulation of VFAs. The continuous UASB
ultaneous leachate biodegradation within the reac- reactor showed much better performance and stab-
tor. The initiation of recycle (indicated on �gures) ility indicating that the SBR mode of operation,
improved biodegradation rates as marked by a speci�cally the �ll and react phases, was inŻuencing
more rapid decrease in soluble COD and VFA con- ultimate reactor performance.
centrations, which continued throughout the react Maximum volumetric organic removal rates
phase. When operating an AnSBR, recycle should (VORRmax) and maximum speci�c organic removal
be started as early as possible within the SBR cycle. rate (SORRmax) can be determined from Figs 6ą8,
As shown in Fig. 9, the typical COD and VFA based on the react portion of the SBR cycle, which
pro�le for the AnSBR operating at a 12 h HRT represents the phase of maximum COD removal.
treating 100% leachate is erratic with elevated con- One must keep in mind that the measured
centrations of these components. The initiation of VORRmax and SORRmax are dependent not only
recycle did not lead to a decrease in COD and VFA on the concentration of the substrate in contact
Table 4. Summary of AnSBR and UASB reactor performance
HRT F:R ratio Average OLR g COD/l d Average removal e�ciency, %
VFAb COD
AnSBR reactorsa
AnSBR 1 100% leachate
24 h 0.5 3.0ą9.2 99 86
24 h 1.0 3.5ą8.0 97 81
24 h 2.0 3.2ą8.8 94 76
18 h 0.5 4.0ą11.2 91 78
18 h 1.0 4.0ą10.2 78 73
18 h 2.0 4.0ą11.9 79 71
12 h 0.5 3.0ą12.6 41 49
12 h 1.0 3.0ą18.4 43 54
12 h 2.0 3.0ą18.0 33 45
AnSBR 2 66% leachate
24 h 0.5 2.3ą6.3 97 72
24 h 1.0 2.6ą7.4 96 79
24 h 2.0 3.3ą7.6 93 85
18 h 0.5 2.8ą8.5 93 81
18 h 1.0 2.8ą9.3 92 83
18 h 2.0 2.3ą10.0 95 84
12 h 0.5 4.3ą13.5 93 78
12 h 1.0 4.4ą15.6 92 79
12 h 2.0 4.2ą15.6 91 81
AnSBR 3 33% leachate
24 h 0.5 0.6ą3.2 99 89
24 h 1.0 1.3ą3.4 92 86
24 h 2.0 1.2ą3.4 91 91
18 h 0.5 0.9ą4.6 96 88
18 h 1.0 1.5ą4.1 91 92
18 h 2.0 1.6ą4.3 94 83
12 h 0.5 2.3ą6.7 94 80
12 h 1.0 2.5ą6.7 91 83
12 h 2.0 2.5ą5.9 94 84
Continuous UASB reactora
100% leachate feed
24 h n/a 4.8ą9.8 96 78
18 h n/a 6.4ą13.1 91 81
12 h n/a 9.6ą19.7 89 77
66% leachate feed
24 h n/a 2.6ą7.9 96 84
18 h n/a 3.5ą10.6 93 79
12 h n/a 5.3ą15.8 94 89
33% leachate feed
24 h n/a 1.2ą4.3 99 91
18 h n/a 1.6ą5.8 94 89
12 h n/a 2.4ą8.6 91 84
a
Average results over study period for particular testing scenarios; n/a: not applicable.
b
Combined butyric, propionic and acetic acid.
3646 K. J. Kennedy and E. M. Lentz
Fig. 3. Soluble COD removal e�ciency vs HRT and F:R ratios: (100% leachate).
with the biomass but also a function of the SBR highest F:R ratio which was 2. Slower �ll times
operation mode. Microorganisms exposed to high resulted in less stress on the microbial consortia
OLRs during the �ll cycle (i.e. short �ll time) may with concomitant higher speci�c removal perform-
behave di�erently than microorganisms exposed to ance compared to AnSBRs receiving the same OLR
similar inŻuent wastewater concentration but oper- but operated with shorter �ll times (lower F:R
ated at longer �ll times. Table 5 summarizes ratios). As the F:R ratio was decreased the speci�c
VORRmax, SORRmax, (during react phase) and removal capacity of the sludge decreased although
maximum COD concentrations for the AnSBRs the overall COD removal performance did not drop
operated at a 24 h HRT. The highest VORRmax signi�cantly in the reactors tested. However this in-
and SORRmax in the range of 0.34ą0.45 g COD/l h formation tends to indicate that at higher volu-
and 1.04ą1.40 g COD/gVSS d respectively were metric organic loads, reactors with longer F:R
obtained when the AnSBRs were operated at the ratios would be more likely to perform better than
Fig. 4. Soluble COD removal e�ciency vs HRT and F:R ratios: (66% leachate).
Treatment of land�ll leachate using UASB reactors 3647
Fig. 5. Soluble COD removal e�ciency vs HRT and F:R ratios: (33% leachate).
reactors operated with short F:R ratios. VORRmax cycle is the same when feed and test conditions
and SORRmax were calcuated for all runs except for remain constant. All biogas pro�les bore strong
the AnSBR operated at a 12 h HRT with 100% lea- similarities in appearance in that they were charac-
chate when methanogenic inhibition was evident. terized by a slow initial production of biogas in the
Using the results from runs the maximum overall initial portion of the �ll cycle, which is attributed to
observed speci�c removal rate, was determined for the activation time required by the bacteria due to
the react phase, kreact cycle, independent of cycle the varying organic load imposed by SBR oper-
length. The kreact cycle was estimated to be 1.320.3 g ation. Biogas production increases following this ac-
COD/g VSS d. climation period and �nally, as the degradable
Results for all AnSBRs indicated the methane portion of the waste is utilized, biogas production
content of the biogas to be in the range of 82ą88% decreases. The initiation of recycle was found to
which was expected based on the high reactor pH, increase the production of biogas as was the case
approximately 8.0. For the continuous UASB reac- for the reactor soluble COD and VFA pro�les.
tor overall methane composition was above 84%. There was a marked decrease in biogas production
Again, the high methane concentration was attribu- for AnSBR1 treating 100% leachate at an HRT of
ted to the reactor pH levels, which were above 8.0. 12 h. High VFAs and low biogas production is an
Biogas production for all reactors was in the range indication of methanogenic inhibition and impend-
of 0.29ą0.34 l CH4 (STP)/g COD removed which is ing reactor failure.
in fairly good agreement with the theoretical value
InŻuence of sequencing batch mode on system per-
of 0.35 l CH4 (STP)/g COD removed. Figure 10
formance
shows 3 cycles of biogas production for a typical
AnSBR reactor scenario. The basic shape of each InŻuent and e�uent COD concentrations for
Table 5. Rate of COD removal following the end of �ll cycle: 24 h HRT
Percent F:R Maximum [COD] within Rate of COD removal following �ll Speci�c COD removal rate following �ll
leachate feed ratio reactor, mg/l cycle, g COD/l h cycle, g COD/g VSS d
100 2.0 5200 0.45 1.4
100 1.0 1750 0.14 0.43
100 0.5 2400 0.17 0.52
66 2.0 3250 0.42 1.3
66 1.0 2500 0.25 0.69
66 0.5 3300 0.35 1.04
33 2.0 2500 0.34 1.04
33 1.0 1600 0.15 0.46
33 0.5 2300 0.20 0.61
3648 K. J. Kennedy and E. M. Lentz
AnSBRs can be used to determine the overall for AnSBRs. Compared to a continuous reactor
removal e�ciencies, however they do not explain that receives inŻuent volumes equally over time, an
the operational nature of the SBR mode. AnSBR AnSBR operated at the same HRT can be exposed
organic loading rates have been based upon the to acute substrate overloading rates depending on
overall HRT for comparison with the continuous the �ll/react cycle used. The SOLRtotal and SOLR�ll
UASB. However, due to the use of varying �ll and which relate OLR to the amount of biomass in the
react phases, a further interpretive step is required reactor, also take into account that leachate feed is
Fig. 6. Typical VFA and COD concentration pro�les for AnSBR 1: 100% leachate feed, 24 h HRT (a)
F:R=2 (b) F:R=1 (c) F:R=0.5.
Treatment of land�ll leachate using UASB reactors 3649
introduced either over the whole cycle or over a operation SOLRtotal is the speci�c loading rate
portion of the cycle (i.e. �ll phase), respectively. based on the overall cycle time, the sum of the �ll,
SOLRtotal (for continuous reactors is the speci�c react, settle and discharge times. SOLRtotal is useful
loading rate based on the HRT while for AnSBR for control of continuous reactors, however for
Fig. 7. Typical VFA and COD concentration pro�les for AnSBR 2: treating 66% leachate feed, 24 h
HRT (a) F:R=2 (b) F:R=1 (c) F:R=0.5.
3650 K. J. Kennedy and E. M. Lentz
Fig. 8. Typical VFA concentration and COD concentration pro�les for AnSBR 3: treating 33% lea-
chate feed, 24 h HRT (a) F:R=2 (b) F:R=1 (c) F:R=0.5.
Treatment of land�ll leachate using UASB reactors 3651
Fig. 9. Typical VFA and COD concentration pro�les for AnSBR 1: treating 100% leachate feed, 12 h
HRT: overload case (a) F:R=2 (b) F:R=1 (c) F:R=0.5.
3652 K. J. Kennedy and E. M. Lentz
AnSBR reactors SOLR�ll has been shown to be a mined earlier and reemphasizes the importance of
critical control parameter (Kennedy et al., 1991). the SOLR�ll on the operational performance of
COD removal e�ciency based on the biodegradable the AnSBR. Kennedy et al. (1991) suggested that
fraction of the leachate) vs SOLRtotal and SOLR�ll SOLR�ll be used for design and that for sucrose
for the AnSBRs are shown in Fig. 11. There is a waste it was recommended not to exceed 1 g
decreasing trend in the COD removal e�ciency COD/g VSS d. Zeng (1995) con�rmed the �ndings
with increasing SOLR and SOLR�ll beyond 1.0 of Kennedy et al. (1991).
total
and 2.5 g COD/gVSS d, respectively. Kennedy et To ensure successful AnSBR treatment of munici-
al. (1991) and Zeng (1995) reported similar obser- pal leachate (containing a high proportion of VFA
vations to those seen in Fig. 11. For AnSBR treat- and BOD5:COD ratio in the range of 0.86) it is rec-
ment of sucrose wastewater Kennedy et al. (1991) ommended that the SOLR�ll should not exceed 2.5 g
and Zeng (1995) reported a marked decrease in COD/g VSS d. This value is in agreement with
COD removal e�ciency at SOLR�ll rates of 1.0 and speci�c removal rates values reported by Henze and
1.5 g COD/g VSS d, respectively. The literature and Harremoe� s (1982) for wastes containing short chain
work done by the authors indicates that the maxi- volatile fatty acids.
mum speci�c organic removal rate for anaerobic Changes in toxicity between inŻuent and treated
1
treatment of sucrose waste is in the range of 0.7ą leachate e�uents evaluated via Microtox (reported
1.2 g COD/g VSS d, while for VFA rich wastes as EC50 values (e�ective concentration)) are sum-
such as municipal land�ll leachate it can be in the marized in Table 6. Table 6 shows that in most situ-
range of 4ą5 g COD/g VSS d (Henze and Harre- ations, continuous UASB and AnSBRs experienced
moe� s, 1982). similar toxicity reductions (a positive increase in
Figure 12 compares all AnSBR data for EC50). For the AnSBR that began to show signs of
SORRtotal vs SOLR�ll. As SOLR�ll values exceed failure (12 h HRT), the toxicity levels in the reactor
about 2.5ą3.0 g COD/g VSS d, the slope of the e�uent actually rose indicating that more toxic
curve decreases because while the loading rate metabolic byproducts were being produced. At low
increases, the COD removal rate e�ciency can HRTs (high OLRs), the continuous UASB reactor
not be maintained. The Żattening of the curve exhibited superior toxicity removal (treatment) than
indicates that maximum removal e�ciency AnSBRs at similar loadings.
achieved by the AnSBRs decreases rapidly. Ad- InŻuent and e�uent from both AnSBR and con-
ditionally, Fig. 12 indicates that the maximum tinuous UASB reactors was analyzed for a range of
SORRtotal is in the order of 1.6ą1.8 g COD/g metallic and non-metallic substances. This was done
VSS d at SOLR�ll values of 4ą6 g COD/g VSS d. in order to ascertain the e�ectiveness of the reactors
This plot concurs with the estimate of kreact deter- in removing regulated substances which would
Fig. 10. Typical cumulative biogas production for AnSBR treating 100% leachate for 24 h HRT and
F:R ratio of 2 (OLR 4.5 g/l d): 3 cycles.
Treatment of land�ll leachate using UASB reactors 3653
e�ect disposal of treated e�uent into the municipal Concentrations of sulphates, phosphorus and zinc
sewerage system. were well below guideline concentrations in both
Selected sewer use concentration guidelines the inŻuent and treated e�uent from all reactors
(RMOC, 1994) and overall averaged inŻuent and under all conditions. Additionally, for the majority
e�uent concentration results for the AnSBR and of conditions, the AnSBRs were able to maintain
continuous UASB reactors are presented in Table 7. chloride levels in the treated e�uents below maxi-
Fig. 11. InŻuence of SOLRtotal and SOLR�ll on biodegradable COD removal e�ciency. (a)
Biodegradable fraction removal e�ciency vs SOLRtotal; (b) Biodegradable fraction removal e�ciency
vs SOLR�ll.
3654 K. J. Kennedy and E. M. Lentz
Fig. 12. SORRtotal vs SOLR�ll for AnSBRs treating land�ll leachate.
CONCLUSIONS
mum sewer use concentrations while the continuous
UASB reactor treating 100% leachate did not. The
Municipal land�ll leachate is indeed amenable to
reason for the lower chloride concentrations in the
treatment by anaerobic sequencing batch reactor
e�uent of the AnSBR compared to the continuous
technology with concomitantly consistently high
UASB was in part due to lower concentrations in
biogas production rates with high methane content.
the inŻuent to the AnSBRs. Figure 13 shows the
At low to medium OLRs, the performance of the
temporal changes in inŻuent and e�uent sulphides,
continuous UASB and AnSBR were similar. The
respectively for the AnSBR treating 100% leachate.
AnSBR had soluble COD removal e�ciencies of
In the case of sulphides, neither operating system
between 71% and 92% for three HRTs (24, 18 and
su�ciently reduced levels to within sewerage use
12 h) and three leachate feed compositions (exclud-
levels. The inability to achieve regulatory concen-
ing treatment of 100% leachate feed with a 12 h
trations would necessitate some form of leachate
HRT) at OLRs between 0.6 and 18.4 g COD/l d.
pretreatment (peat �ltration) prior to anaerobic
The continuous UASB reactor had soluble COD
digestion or of post treatment (ultra�ltration or
removal e�ciency between 77% and 91% at similar
reverse osmosis membrane treatment) before dis- HRTs, leachate concentrations and OLRs from 1.2
charge to the sewer could be considered. and 19.7 gCOD/l d.
1
Table 6. Comparison change of Microtox toxicity between inŻuent and e�uent for continuous UASB and AnSBR reactors
HRT Percentage by volume, leachate Sequencing batch reactor overall change in Continuous operation reactor increase in
(h) feed EC50 (%)a EC50 (%)b
24 100 Q67.4 59.4
66 Q58.4 56.2
33 Q20.8 n/c
18 100 Q82.3 75.1
66 Q67.3 57.7
33 Q71.6 n/c
12 100 q55.1 17.6
66 Q46.1 30.8
33 Q38.2 13.7
a
Q: increase in EC50 (reduction in toxicity); q: decrease in EC50 (increase in toxicity).
b
n/c: no detectable change in toxicity.
Treatment of land�ll leachate using UASB reactors 3655
Table 7. Overall inŻuent and e�uent characteristics for AnSBRs and continuous UASB reactor
Substance Average substance concentration over study period, mg/l
Chlorides expressed as Cl, Sulphates expressed as Phosphorus, Zinc, Sulphides expressed as
mg/l SO4, mg/l mg/l mg/l S, mg/l
Sewer guideline 1500 1500 3 2
concentrations, mg/l
AnSBRs�average values over study period (mg/l)
100% leachate:
inŻuent 1779 82 6 0.3 21
e�uent 1088 34 7 0.1 11
66% leachate:
InŻuent 1520 48 2 0.1 12
e�uent 1040 24 2 0.1 14
33% leachate:
inŻuent 851 24 3 0.05 8
e�uent 897 31 1 0.05 11
Continuous UASB reactora
100% leachate:
inŻuent 3035 165 2 0.4 35
e�uent 2352 31 2 0.1 18
66% leachate:
inŻuent 1991 71 2 0.04 12
e�uent 1125 39 1 0.09 7
33% leachate:
inŻuent 360 25 2 0.01 8
e�uent 307 40 1 0.05 6
a
Average results at all OLRs.
Continuous UASB reactors performed more of VFAs and a decrease in COD removal e�-
favorably than the AnSBR at the highest OLRs ciency.
(with corresponding low HRTs). At high OLRs The �ll:react ratio can inŻuence the removal
(12 h HRT, 100% leachate), the AnSBR system e�ciency and performance of AnSBRs. The
showed initial signs of system failure characterized SOLR�ll for AnSBR treatment of municipal lea-
by a decrease in biogas production, accumulation chate (used in this study) should not exceed 2.5 g
Fig. 13. InŻuent and e�uent sulphide concentration for AnSBR 1, treating 100% leachate (OLR 3.0ą
18.4 g COD/l d). Sulphide measured as elemental S, Municipality of Ottowa-Carleton maximum
allowable concentration for sewer discharge is 2 mg/l.
3656 K. J. Kennedy and E. M. Lentz
Conference and Exposition Anaheim, California, October
COD/g VSS d in order to ensure successful AnSBR
3ą6, Water Environment Federation, pp. 21ą31.
operation.
Henze M. and Harremoe� s P. (1982) Literature Review,
Leachate treatment in both continuous and
Anaerobic Treatment of Wastewater in Fixed Film
sequencing batch anaerobic reactors e�ectively
Reactors. IAWPR Specialised Seminar, Department of
1
reduced Microtox toxicity levels. At high OLRs Sanitary Engineering, Technical University of Denmark,
Danish IAWPR Committee, 1ą90.
(and corresponding low HRTs), the continuous sys-
Hollopeter J. A. and Dague R. R. (1995) Anaerobic
tem showed better toxicity removal.
Sequencing Batch Reactor Treatment of Land�ll Lea-
Neither operational mode was able to consist-
chate. WEF Conference 19ą95.
ently lower sulphide, chloride and BOD concen- Kennedy K. J., Hamoda M. F. and Guiot S. R. (1988)
Anaerobic treatment of leachate using �xed �lm and
trations to meet sewer use guidelines for direct
sludge bed systems. J. Water Pollut. Control Fed. 60,
wastewater disposal to the sewer. Future work
1675.
should look at peat�lters for pretreatment or ultra-
Kennedy K. J., Sanchez W. A., Hamoda M. F. and
�ltation and/or reverse osmosis as a post treatment.
Droste R. L. (1991) Performance of anaerobic sludge
blanket sequencing batch reactors. J. Wat. Pollut. Con-
trol Fed. 63, 75ą83.
REFERENCES
Knetchel R. J. (1979) A more economical method for the
determination of chemical oxygen demand. Water and
Ackman R. G. (1979) Porous polymer bead packings and
Waste Eng. 14, 25ą30.
formic acid vapor in the GLC of volatile free fatty
Lema J. M., Mendez R. and Blazquez R. (1988) Charac-
acids. J. Chromatograph Science 10, 560ą565.
teristics of land�ll leachates and alternatives for their
APHA (1989) Standard Methods for the Examination of
treatment: a review. Water Air Soil Poll. 40, 223ą250.
Water and Wastewater, 17th Edition, Method 2540D.
Lin C.-Y. (1991) Anaerobic digestion of land�ll leachate.
Barbre C. and Marais P. J. (1984) Recirculation of lea-
Water SA 17, 301ą306.
chate as a land�ll management option: bene�ts and op-
Malina J. F. and Pohland F. G. (1992) Design of Anaero-
erational problems. Q. J. Eng. Geol. London 17, 9ą29.
bic Processes for the Treatment of Industrial and Mu-
Chang J.-E. (1988) Treatment of land�ll leachate with an
nicipal Wastes. Water Quality Management Library
upŻow anaerobic reactor combining a sludge bed and a
Volume 7, Technomic Publishing Co. Inc., p. 169.
�lter. Wat. Sci. Tech. 21, 133ą143.
RMOC (Regional Municipality of Ottawa-Carleton)
Fernandes L., Kennedy K. K. and Ning Z. (1993)
Dynamic modelling of substrate degradation in sequen- (1994) Regional-Sewer Regulations. Water Protection
Division, Environmental Services Department, Regional
cing batch anaerobic reactors (SBAR). Wat. Res. 27,
Municipality of Ottawa-Carleton (RMOC), 32 pages.
1619ą1628.
Forgie D. J. (1988) Selection of the most appropriate lea- Van Huyssteen A. J. (1967) Gas chromatographic separ-
ation of anaerobic digester gases using porous polymers.
chate treatment methods�3: a decision model for the
Water Res. 1, 237ą242.
treatment train selection. Wat. Poll. Res. J. Canada 23,
341ą355. Zeng Y. (1995) Biodegradation of 2,4-dichlorophenol in
Harty D. M., Hurta G. P., Werthman P. H. and Konsella sequencing batch anaerobic reactors. In Master of
J. A. (1993) Sequencing batch reactor treatment of high Applied Science Thesis, University of Ottawa, Canada,
strength leachate: a pilot scale study. In 66th Annual Department of Civil Engineering, 122 pp.