Water Research 36 (2002) 4079–4087

Biological nitrogen removal from municipal landfill leachate:

low-cost nitrification in biofilters and laboratory scale
in-situ denitrification
J.P.Y. Jokelaa,*, R.H. Kettunenb, K.M. Sormunena, J.A. Rintalaa
a
.
Department of Biological and Environmental Science, University of Jyvaskyl . P.O. Box 35, FIN-40351 Jyvaskyl
a, . . Finland
a,
b
Tritonet Oy, Pinninkatu 53 C, FIN-33100 Tampere, Finland
Received 2 February 2001; received in revised form 7 February 2002; accepted 15 March 2002

Abstract

The slow leaching of nitrogen from solid waste in landfills, resulting in high concentrations of ammonia in the landfill
leachate, may last for several decades. The removal of nitrogen from leachate is desirable as nitrogen can trigger
eutrophication in lakes and rivers. In the present study, a low-cost nitrification–denitrification process was developed to
reduce nitrogen load especially in leachates from small landfills. Nitrification was studied in laboratory and on-site pilot
aerobic biofilters with waste materials as filter media (crushed brick in upflow filters and bulking agent of compost in a
downflow filter) while denitrification was studied in a laboratory anoxic/anaerobic column filled with landfill waste. In
the laboratory nitrification filters, start-up of nitrification took less than 3 weeks and over 90% nitrification of leachate
(NH4-N between 60 and 170 mg N l
1, COD between 230 and 1300 mg l
1) was obtained with loading rates between 100
and 130 mgNH4-N l
1 d at 251C. In an on-site pilot study a level of nitrification of leachate (NH4-N between 160 and
270 mg N l
1, COD between 1300 and 1600 mg l
1) above 90% was achieved in a crushed brick biofilter with a loading
rate of 50 mg NH4-N l
1 d even at temperatures as low as 5–101C. Ammonium concentrations in all biofilter effluents
were usually below the detection limit. In the denitrification column, denitrification started within 2 weeks and total
oxidised nitrogen in nitrified leachate (TON between 50 and 150 mg N l
1) usually declined below the detection limit at
251C, whereas some ammonium, probably originating from the landfill waste used in the column, was detected in the
effluent. No adverse effect was observed on the methanation of waste in the denitrification column with a loading rate
of 3.8 g TON-N/t-TSwaste d. In conclusion, nitrification in a low-cost biofilter followed by denitrification in a landfill
body appears applicable for the removal of nitrogen in landfill leachate in colder climates. r 2002 Elsevier Science Ltd.
All rights reserved.

Keywords: Landfill leachate; Nitrification; Denitrification; Low temperature; Biofilter; Waste material

1. Introduction dry weight of MSW (e.g. [2]). The hydrolysis of the

In landfills, the release of soluble nitrogen from
municipal solid waste (MSW) into landfill leachate
continues over a long period compared to that of
soluble carbon compounds [1]. The main source of
nitrogen is proteins, which accounts for approx. 0.5% of
*Corresponding author. Tel.: +358-14-2602-1211; fax:
+358-14-2602-2321.
E-mail address: jari.jokela@jyu.fi (J.P.Y. Jokela).
polypeptide chain is disadvantaged in energetic terms
and this is apparently the reason for the slow kinetics of
protein hydrolysis, which in turn causes the slow release
of ammonia. Nitrogen can trigger eutrophication in
receiving watercourses and therefore its removal from
landfill leachate, e.g., by biological treatment, is desir-
able. Biological nitrogen removal is nonreversible [3]
and is carried out in two stages: aerobic nitrification
of ammonia via hydroxylamine and nitrite to nitrate,
and, subsequently, anoxic denitrification of nitrate via

0043-1354/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 3 - 1 3 5 4 ( 0 2 ) 0 0 1 2 9 - X
4080 J.P.Y. Jokela et al. / Water Research 36 (2002) 4079–4087
intermediate stages (NO3 ) NO2 ) NO ) N2O ) N2)
to nitrogen gas.
Suspended processes have been applied in full scale to
nitrification and denitrification of landfill leachate (e.g.
review by Kettunen [4]). Recently, new methods such as
suspended carrier biofilm processes, have also been
studied for leachate nitrification, even at low tempera-
tures (5–101C; e.g. [5,6]). These processes are reliable,
but they normally require major investments, whereas
there is also an evident need for low cost and low
maintenance systems. This is especially the case in low
population areas where landfills are smaller and often at
a distance from sewage systems and lack trained
personnel. Moreover, the treatment processes used
should also function in colder climates.
Normally, the conditions prevalent in a landfill body
are anaerobic/anoxic thus enabling methanogenesis and,
potentially, denitrification. Since the aerobic conditions
required for nitrification are unlikely to occur in the
landfill body, external nitrification is required. Thus, we
concluded that the removal of nitrogen by a process
consisting of denitrification in the landfill body preceded
by nitrification in a biofilter with low-cost filter media,
e.g., porous waste material, could be an attractive
option. Previously, the feasibility of compost material
as a medium for the denitrification and nitrification of
landfill leachate has been shown in laboratory-scale
columns [7]. Anaerobically, degrading MSW has also
been found to completely denitrify leachate containing
nitrate, although the leachate delayed methanogenesis
and decreased methane yield [8], apparently due to
suppression of methanogenesis by nitrate (e.g. [9–11]).
When planning the use of landfill for denitrification it
should be taken into account that denitrifiers may
outcompete methanogens [8].
The objective of this study was to develop a low-cost
method for the removal of nitrogen from landfill
leachate. For this purpose, the nitrification of landfill
leachate was studied in laboratory filters with bulking
agent of mature compost and crushed brick as the filter
material as well as in a suspended carrier biofilm process
with commercial carriers as a reference. In addition, an
on-site pilot study of nitrification using a filter of
crushed brick was carried out. Denitrification of nitrified
leachate was studied in a laboratory-scale column
containing methane-producing landfilled material. The
effect of denitrification on the methanation of landfilled
waste was also evaluated.
2. Materials and methods
2.1. Landfill leachate
Leachate was sampled for laboratory-scale experi-
ments from the main drain at Mustankorkea municipal
landfill in Jyv.askyl.a (Finland), which has been in
operation since 1963. The total volume of landfill in
1999 was approx. 5 million m3 and the annual produc-
tion of leachate was ca. 114,000 m3. In addition to
MSW, some industrial waste (e.g. fibrewaste from a
papermill, fly ash and waste grit) has been disposed off
in the landfill. Putrescible household waste has been
source sorted and composted separately since 1996. The
leachate was delivered 1–2 times per month to the
laboratory and stored at 41C. The feed for the
experiments was prepared by clarifying the leachate in
a container. Due to a deficiency of phosphorus in the
landfill leachate, 5–20 mg P l
1 (1 M H3PO4) was added
to the feed. The ammonium nitrogen (NH4-N) of the
feed was typically 60–170 mg N l
1, biochemical oxygen
demand (BOD7) was 18 mg O2 l
1 and the chemical
oxygen demand (COD) was 230–510 mg O2 l
1. Between
experimental days 135 and 156 of the runs, the leachate
contained some exceptional streams from the landfill
and, consequently, the feed had a higher COD (1300 mg
O2 l
1) and BOD7 (510 mg O2 l
1), whereas ammonium
remained between 60 and 90 mg N l
1. During the
nitrification experiment, the pH tended to rise above 8.5
resulting from loss of carbonate, which may cause loss
of ammonia by volatilisation. The feed pH was
decreased by the addition of 1 M hydrochloric acid
(HCl) to approx. 7.0–6.5 to maintain the pH of the
effluent below 8.5. Identical feed was used in all the
nitrifying laboratory reactors. During the runs, the
ammonia load was mainly controlled by adjusting the
hydraulic retention time (HRT), but it was also affected
by the changes in the feed ammonia concentration.
The leachate used in the pilot study was pumped from
the main leachate ditch at Kiertokapula municipal
landfill in Hyvink.aa. (Finland). The landfill has been in
operation since 1974 and approx. 1.3 million m3 of
MSW has been disposed off in the landfill. Also at this
site, source-segregated putrescible household waste has
been composted separately since 1996. The drainage of
the compost plant is discharged into the landfill and was
thus included in the leachate used in the experiment. The
availability of soluble phosphorus was ensured (COD:P
ratio of 100:1) by the continuous addition of 10–30 mg
phosphorus (1 M H3PO4) per litre to the leachate in a
mixing tank. The NH4-N of the leachate varied between
160 and 270 mg N l
1 (NH4-N/TKN-ratio was between
0.78 and 0.99) and the COD between 1300 and 1600 mg
O2 l
1 and the BOD7 was 5 mg O2 l
1. The pH varied
between 6.8 and 7.5.
2.2. Experimental set-up
2.2.1. Laboratory-scale experiments
Three types of reactors were used in the laboratory-
scale experiments for nitrification of landfill leachate: an
upflow filter (UF) with crushed brick as the filter
 J.P.Y. Jokela et al. / Water Research 36 (2002) 4079–4087
medium, a downflow filter (DF) with a bulking agent
(wood chips) of mature compost of source-segregated
putrescible household waste as the filter medium and a
suspended carrier biofilm process (SCBP) with a
commercial carrier. Nitrified leachate from the SCBP
was subsequently fed into a denitrification landfill
column filled with two separate horizons of old land-
filled waste of different ages. All the laboratory-scale
experiments were carried out at ambient temperature
(251C). After the laboratory experiment, an appropriate
method was selected for the nitrification pilot study,
which was carried out on a landfill site. The nitrification
filters were operated in all experiments with a loading
rate between 25 and 130 mg NH4-N g l
1 d
1, which was
considered sufficiently high for the actual cases. No
maximum loading rate of the different processes were
studied.
2.2.1.1. Nitrification. The leachate was continuously
pumped from the feed container (41C) with a peristaltic
pump (Cole Palmer, Masterflex 7553-87) into the
nitrification reactors. The UF and DF were made of
600 mm PVC pipes (total volume 9.4 l) sealed with PVC
flanges. In the UF, the leachate was fed through a
50 mm layer of grit (grain size from #5 to 15 mm) to the
bottom of the filter. The void volume of the UF was
approx. 2.4 l. In the DF filter, the leachate was
distributed to the top of the reactor through a
perforated plate. The DF medium was first washed with
tap water and sieved into particles of 10–70 mm. At the
beginning of the experiment, the nitrifying compost
column was rinsed with 8 l of tap water to reduce further
the amount of soluble material. The SCBP was made of
a 400 mm PVC pipe (total volume of reactor 7.1 l). The
bottom of the SCBP was sealed with a PVC flange and
the top covered with aluminium foil to minimise
evaporation. Forty per cent of the reactor volume was
filled with polyethylene carrier medium (Kaldnes Milj-
.
oteknologi AS, Norway, bulk density=170 kg m
3,
diameter=9.1 mm and length=7.2 mm) with an effec-
tive surface of 500 m2 m
3. PVC was used as the tubing
material in all the nitrification reactors. All three
reactors were inoculated with nitrifying activated sludge
(4 g volative suspended solids (VSS) l
1) from Viinikan-
lahti sewage treatment plant in Tampere (Finland) and
were aerated with submerged diffusers (Rena Air 200).
On account of the mixing requirements, dissolved
oxygen in the SCBP reactor was approx. 8 mg O2 l
1.
2.2.1.2. Denitrification. The denitrification column was
made of 600 mm PVC pipe (column volume 8.8 l) sealed
with PVC flanges. 220 ml of nitrified leachate from the
SCBP was fed to the top of the column with a syringe
once in a week. The total oxidised nitrogen loading rate
was calculated from the weekly added nitrified leachate.
An identical column was operated without added
4081
leachate and was used as a reference. Both columns
were filled with two layers (250 mm) of different landfill
waste. The upper layer consisted of 9-year-old landfill
waste with a maximum particle size of 50 mm. The
characteristics of the waste (753 g TS (total solids) and
345 g VS (volatile solids) in column) are reported
elsewhere [12]. The bottom layer consisted of ca. 45-
year-old MSW (657 g TS and 104 g VS in column) from
Mattilanniemi landfill in Jyv.askyl.a (Finland), which has
been closed for approx. 40 years. The 50 mm layer of grit
on the bottom of the columns prevented washout of
particles. The leachate from the denitrification column
was led from the bottom of the column through Tygon
tubing to 1 l Duran glass vessels (at 41C), which were
sealed with rubber stoppers. The gas produced was led
through Viton tubing to the gas sampling bags (Tecobag
PETP/AL/PE-12/12/75 of 10 l, Tesseraux Spezialver-
packnungen) after passing a water lock.
2.2.2. Pilot study
In the pilot study the upflow nitrification filter (pilot
UF) was constructed from a 2000 mm long HDPE pipe
(diameter 920 mm and a total volume 1.0 m3) sealed with
a HDPE flange. Crushed brick was used as the filter
medium. The medium was sieved into particles of 16–
32 mm. The influent was fed through a perforated plate
and the effluent was discharged from an overflow outlet.
The leachate was semi-continuously pumped from the
main leachate ditch with a submersible pump into a 2 m3
storage tank, from which it was led to a 100 l mixing
tank in which phosphorus was added. The leachate was
continuously pumped from the mixing tank with
Watson–Marlow (504 S/313D) peristaltic pump via a
PVC tube into the reactor. A diaphragm pump (KNF)
was used for aeration. At the start-up of the experiment,
the reactor was inoculated with nitrifying activated
sludge (2 g VSS l
1), which was supplied from Hyvink.aa.
sewage treatment plant (Finland). During the experi-
ment, dissolved oxygen in the filter (from depth of
200 mm) was between 4.2 and 9.0 mg O2 l
1. The filter
was operated for the first 5 days on batch mode to
ensure the establishment of microbial biofilm on the
filter medium.
2.3. Analysis and calculations
Ammonium nitrogen (NH4-N) and total oxidised
nitrogen (TON) were determined by using a Kjeltec
system 1002 distilling unit [13] in accordance with
Standard Methods [23]. Total Kjeldahl nitrogen (TKN)
analyses were done by following the Tecator application
procedure [13] with the same Kjeltec system after
digestion with a 2006 digestor. Nitrate and nitrite
nitrogen as well as phosphate phosphorous were
measured from filtered samples (0.2 mm filters Schleicher
.
& Schull) with Dionex DX 500 ion chromatograph.
4082 J.P.Y. Jokela et al. / Water Research 36 (2002) 4079–4087

COD (dichromate method) and BOD7 were analysed onwards. Ammonia loading reached its highest level of
according to the Finnish Standards Association [14,15]. 110–130 mg NH4-N l
1 d
1 with the shortest HRT (1.4
The volume of gas produced was measured by the days). From the beginning and during the whole run
displacement method. Gas methane concentration was effluent ammonia was below the detection limit (1 mg
analysed with a Perkin Elmer Autosystem XL gas NH4-N l
1). Nitrification efficiency ranged from 60% to
chromatograph, equipped with a flame-ionisation de- 88% until day 60 after which it was >90%. COD
tector and fitted with a PE Alumina column removal ranged typically from 26% to 62%. It was
(30 m  0.53 mm). The carbon dioxide and oxygen only with the exceptional leachate feed (between days
content was analysed with an IR-analyser (Geotechnical 135 and 156), that the nitrification efficiency dropped to
Instruments). o20% and COD removal increased to 70–75%. After
pH was measured in the laboratory with a Metrohm the replacement of the exceptional feed with normal
model 744 pH meter, whereas during the pilot experi- feed, the nitrification efficiency recovered rapidly within
ments, pH was measured with a Hanna model HI 9025 10 days and COD removal dropped. The pH in
pH meter. The dissolved oxygen was measured with a the reactor ranged from 6.5 to 7.8 throughout the
Hanna Instruments HI-9145 oxygen meter. The amount experiment, excluding the start-up period, when the pH
of immobilised biomass in the SCBP was determined by was 8.5.
mechanically scraping biomass from the surface of 12 During the pilot study, the temperature of the feed fell
carriers and determining TS and VS of the scraped gradually from 211C to 5.01C (Fig. 2) and because of the
biomass. TS and VS were analysed according to fairly short HRT in the pilot UF it was same as the on-
Standard Methods [23]. site leachate temperature. The pilot-UF was first
The nitrification efficiency was calculated by dividing operated semi-continuously with a low rate of ammonia
the effluent NO3-N by the influent NH4-N and multi- loading (26 g NH4-N m3 d
1) for 5 days, after which
plying it with 100%. continuous feeding was started with an HRT of 2.1 d
until day 21. Because ammonia was detected in the
effluent, the HRT was increased to 2.6 d on day 21, but
3. Results was adjusted back to 2.1 on day 61 owing to the
reduction of the effluent ammonia below the detection
3.1. Upflow nitrification filter with crushed brick limit. At the end of the run, the effluent ammonia was
below the detection limit and nitrification was over 90%
Nitrification of landfill leachate was studied in an UF (except day 78) with ammonia loading rates approx. 50 g
with crushed brick as the filter material in both the NH4-N m
3 d
1 even at temperatures as low as 5–101C.
laboratory (UF) and on-site pilot study (pilot UF). The removal of COD was o5% (data not shown)
First, the UF was fed once a day for 9 days with low throughout the run whereas effluent BOD7 was 3 mg
ammonia load (2.3 mg N l
1 d
1) (Fig. 1). After that, O2 l
1 as sampled on day 50. The pH in the reactor was
continuous feeding was started with an HRT of 3.8 d around 7.6 throughout the experiment. Effluent PO
3 4 -P
until day 91 and with an HRT of 1.4 d from that ranged from 0.2 to 1.6 mg P l
1 (data not shown).

1500
(mgO2 l-1)
CODCr

Feed Effluent
1000
500
0
Nitrification NH4-N&T.O.N

200 NH4-N in T.O.N. out NH4-N out

(mgN l-1)

100

0
NH4-N load &

200
(mgN l-1 d-1)

150 NH4-N-load
100 Nitrification

50
0
0 20 40 60 80 100 120 140 160 180
Time (days)
Fig. 1. Feed COD and NH4-N, effluent COD, NH4-N and TON, and NH4-N load and nitrification efficiency in a laboratory
experiment using an UF with crushed brick as the filter medium to treat municipal landfill leachate. HRT: semi-continuous feeding,
3.8 d, and 1.4 d between days 1–8, 9–91 and 92–169, respectively.
 J.P.Y. Jokela et al. / Water Research 36 (2002) 4079–4087 4083

Temperature
20

(°°C)
10 Feed
Effluent
0
400
300
Nitrogen
(mgN l-1)
200
100 NH4-N in T.O.N out TKNout NH4-N out

0
60
NH4-N load &
Nitrification
(gN m-3 d-1)

40

20 NH4-N-load
Nitrification
0
0 30 60 90
Time (days)
Fig. 2. Temperature, feed NH4-N and effluent NH4-N, NO2-N, TKN and TON and NH4-N load and nitrification efficiency during
nitrification of municipal landfill leachate in on-site pilot experiment in an UF with crushed brick used as the filter medium. HRT:
semi-continuous feeding, 2.2 d, 2.6 d, and 2.1 d between days 1–3, 4–20, 21–60, and 61–85, respectively.

1500
(mgO2 l )
-1

Feed Effluent
CODCr

1000
500
0
NH4-N&T.O.N

200 NH4-N in T.O.N. out NH4-N out

(mgN l )
-1

100

0
200
NH4-N load &
Nitrification

150 NH4-N-load
(mgN l d )
-1

100 Nitrification
-1

50
0
0 20 40 60 80 100 120 140 160 180
Time (days)

Fig. 3. Feed COD, NH4-N and effluent NH4-N, and TON and NH4-N load and nitrification efficiency during nitrification of
municipal landfill leachate in a laboratory experiment using a continuous DF with mature compost as the filter media. HRT: 9.6 d,
5.1 d and 7.3 d between days 1–58, 59–91, and 92–160, respectively.

3.2. Downflow nitrification filter with mature compost presence of denitrifiers. On day 58 HRT was shortened
to 5.1 d, but was again increased on day 91 to 7.3 d.
Nitrification of municipal landfill leachate was studied From day 70 onwards, effluent TON started to increase
in the laboratory DF with mature compost as the filter and the level of nitrification efficiency was usually above
material. The DF was first operated continuously with 90% (Fig. 3), whereas effluent NH4-N varied from o1
an HRT of 9.6 d for 58 d with an ammonia loading rate to 21 mg N l
1. This performance was achieved with
of 10 mg N l
1 d
1. During this period, both effluent ammonia loading rates of 100–125 mg NH4-N l
1 d
1.
NH4-N and TON were below the detection limit Both the feed NH4-N/TKN ratio and TON/TKN ratio
(1 mg l
1) (Fig. 3), which could indicate TON consump- of the effluent were approx. 0.9. The pH of the effluent
tion by the compost filter biomass or suggest the varied between 6.5 and 7.0 during the study. No COD
4084 J.P.Y. Jokela et al. / Water Research 36 (2002) 4079–4087

1500

(mgO2 l-1)
CODCr
Feed Effluent
1000
500
0

NH4-N&T.O.N
200 NH4-N in T.O.N. out NH4-N out

(mgN l-1)
100

0
200
NH4-N load &
Nitrification

150 NH4-N-load
(mgN l-1 d-1)

100 Nitrification
50
0
0 20 40 60 80 100 120 140 160 180
Time (days)

Fig. 4. Feed COD and NH4-N and effluent COD, NH4-N and TON and NH4-N load and nitrification efficiency during nitrification of
municipal landfill leachate in laboratory SCBP. HRT: 0.6 d, 2.6 d, 5.1 d, 2.4 d, 7.3 d and 4.6 d between days 1–13, 14–108, 109–147, 148–
169, 170–216, and 217–254, respectively.

removal occurred. On the contrary to the UF, the in the form of ammonia (data not shown) and/or
exceptional leachate feed (days 135–160) did not affect accumulated into the biomass. The pH in the reactor
the level of nitrification efficiency but COD removal rose varied from 7.5 to 8.5. On day 189, average attached
to 80%. biomass was 0.8 g VS l
1 (2.7 mg VS per carrier, approx.
300 carriers per litre) and in the suspension 0.13g VS l
1.
3.3. Nitrification in suspended carrier biofilm process Thus, the nitrification rate at the ammonia loading rates
of approx. 150 mg NH4-N l
1 d
1 corresponded to
The laboratory-scale SCBP was studied as a reference 160 mg NH4-N g VS
1 d
1 (including biomass on
for the laboratory filters. The SCBP was operated at first carriers and in suspension).
with an HRT of 9.3 d to acclimate the biomass in SCBP
to low carbon leachate. After an acclimation period of 3.4. Denitrification in landfill waste column
14 d, the HRT was decreased to between 1.4 and 3.5 d to
maintain an ammonia loading rate of between 40 and Denitrification of nitrified leachate was studied in the
73 mg NH4-N l
1 d
1 (Fig. 4). On day 170, the HRT was laboratory in a landfill waste column by feeding semi-
further decreased to 1.0 to obtain a high ammonia continuously nitrified leachate from SCBP into the
loading rate of between 87 and 160 mg NH4-N l
1 d
1. column.
The exceptional feed (days 204–225) caused inhibition of The loading rate was 0.36–0.40 mg TON l
1 d
1 until
nitrification. Also, COD removal rose to approx. 90%. day 185 which after it was increased to approx. 0.6 mg
Consequently, the HRT was increased to 1.6 d for the TON l
1 d
1. The latter value corresponds to 3.8 g TON
rest of the run and the ammonia load decreased to about t-TS
1
1
waste d . At the beginning some TON (5.5 mg N l )

1

100 mg NH4-N l
1 d
1, which after nitrification effi- was found in the effluent, while after 40 days it was no
ciency was back to >90%. longer detected (Fig. 5). The amount of ammonia in the
During the run, the effluent ammonia was mostly denitrified leachate showed a slightly increasing trend in
below the detection limit excluding the high COD the course of the run. Effluent COD was higher than
feeding period, which caused excessive growth of feed COD except in the last sample. During the run the
biomass in the reactor. During the high COD period TKN was removed by 35–71%. The NH4-N/TKN ratio
effluent ammonia was between 31 and 49 mg NH4-N l
1. of the denitrified leachate varied between 62% and 84%.
Exceptionally, some nitrite was recorded in the effluent The methane content of the gas produced and the
on day 15 (33 mg N l
1) and on day 24 (44 mg N l
1) cumulative amount of methane produced in the
probably due to incomplete interaction between nitrite denitrification column and in the control column are
producing and oxidising bacteria. Between 112 and 196 presented in Fig. 6. Both columns showed stable levels
days the nitrification efficiency was between 75% and of gas and methane production before starting the
99% and COD removal ranged from 53% to 63%. feeding of the denitrification column on day 87. During
During the experiment, in terms of nitrogen balance, the runs, the methane (50–59%) and carbon dioxide
from 10% to 30% of the nitrogen was lost by stripping (31–41%) content as well as methane production rates
 J.P.Y. Jokela et al. / Water Research 36 (2002) 4079–4087 4085

600

(mgO2 l-1)
CODCr
Feed Effluent
400
200
0
200
NH4-N&T.O.N T.O.N. in NH4-N out T.O.N out
(mgN l-1) 150

100

50

0
0 50 100 150 200
Time (days)
Fig. 5. Feed COD and TON and effluent TON and NH4-N in denitrifying waste column during denitrification of nitrified (suspended
carrier biofilm process) municipal landfill leachate in the laboratory. The column contained horizons of approx. 10- and 40-year-old
municipal landfilled waste.

Feeding period
Cumulative CH4 CH4,CO2 & O2 conc.

80
CH4 (DN) CO2 O2 CH4 (M) CO2 O2
60
40
(%)

20
0
80
production (l)

60

40
20 With nitrified feed
Without nitrified feed
0
0 50 100 150 200 250 300 350
Time (days)
Fig. 6. Methane, carbon dioxide (CO2) and oxygen concentration of biogas and cumulative methane production in denitrifying (DN)
and control (M, without leachate addition) waste column during denitrification of nitrified (SCBP) municipal landfill leachate in the
laboratory. Both columns contained horizons of approx. 10- and 40-year-old municipal landfilled waste.

(approx. 0.3 l CH4 d
1), were identical in both columns. resulted a complete denitrification with the rate equiva-
After the denitrification study was completed, biogas lent approx. to 470 and 1100 g NO3-N t-TS
1 waste d

1

composition and production were monitored for 76 days (calculated by authors). However, this resulted in a
and no difference was found between the columns. delayed onset of methanogenesis and in reduced
quantities of methane. In their study the added initial
nitrate load was about hundred times higher than in our
4. Discussion study, which probably explains the difference in the
effects on methanogenesis.
The results suggest that nitrified landfill leachate can The ammonia detected in the effluent of the deni-
be denitrified in a landfill body that contains old trification column may have originated from the landfill
methanogenic waste material. In our study, denitrifica- material used in this column. On the other hand, the
tion did not affect methanation in the denitrifying high COD/NO3-N ratio of the leachate may promote
column when compared to the reference column. In a the reduction of nitrate of leachate to ammonia [17], and
laboratory batch study by Burton and Watson-Craik as the COD/NO3-N ratio inside the column was not
[16], addition of nitrate (500–1000 mg NO3-N l
1, added measured, the possibility that nitrate removal could have
as KNO3-N) into methanogenic landfilled MSW been due to reduction to ammonia in the column cannot
4086 J.P.Y. Jokela et al. / Water Research 36 (2002) 4079–4087
be excluded. However, the few gas samples analysed
showed the presence of nitrogen (data not shown) that
supported the fact that denitrification was the major
mechanism of nitrate removal. Apparently, the waste
materials in the denitrification column contained suffi-
ciently available organic material and nutrients (espe-
cially phosphorous) for denitrification during the run.
On the other hand, during long-term operation of
denitrifying landfill body, their sufficiency should also be
considered.
Regardless of the potentially adverse effects of the
addition of nitrified effluent on the methanogenesis of
landfill, the present study suggests that the loading rate
of TON to the methane-producing part of landfill can be
at least 3.8 g N t-TS
1
waste d

1
without any adverse effect
on methanogenic activity. This means that, for example,
about 60 m3 of nitrified leachate with 100 mg TON l
1
could theoretically be recirculated daily in a 100 m2
landfill area with a depth of 20 m (density 0.8 t m
3). The
maximum denitrification capacity of the landfill material
was not studied in the present experiments, but that
would presumably be higher than the level used here as
the results of Burton and Watson-Craik [16] suggest.
However, in their study nitrate inhibited methanogenic
activity. Although the denitrification capacity of landfill
could be high, the potential of denitrifiers outcompeting
methanogens and thus suppressing the methanation
of landfill waste needs to be considered. Fang and
Zhou [10] reported that COD/NO3-N ratios lower
than 3.3 favoured denitrification and methanogens were
outcompeted in an upflow anaerobic sludge blanket
reactor treating wastewater. In the present study
the COD/TON ratio of the leachate in the feed of the
denitrification column was 1.6–3.8, whereas in the
effluent the ratios were twofold, suggesting that carbon
apparently was available inside the column for both
methanogens and denitrifiers. Denitrification could also
probably be employed during colder periods as the
temperature of the biologically active part of a landfill
body is agreeable (4–261C) even during colder periods
[18,19].
When applying denitrification in a landfill body by the
recirculation of nitrified leachate, there are a number of
technical matters that should be taken into account.
First, it should be ensured by design and selection of
leachate distribution method that the leachate is evenly
distributed to avoid destabilising the waste body by
flooding it (reviewed by Reinhart and Townsend [20]).
Secondly, the recirculation of leachate could increase its
particulate content, which might cause problems in the
form of drainage clogging. Thirdly, leachate recircula-
tion may be limited by the low hydraulic conductivity of
the landfill waste. If the hydraulic conductivity of well-
compacted waste is approx. 10
5–10
8 m s
1 (reviewed
by Saarela [21]), then only 0.1–90 m3 d
1 of leachate can
be recirculated on 100 m2 of landfill site. Fourthly, an
excessive head of leachate caused by, e.g., channel flow
of leachate or clogging of the drainage system could
result in failure of the bottom liner (reviewed by
Reinhart and Townsend [20]). Fifthly, the parts of the
landfill subject to leachate recirculation must be care-
fully selected as excessive addition of leachate by
recirculation could retard the methanation of fresh
waste [12].
The present results show that filters employing low-
cost waste materials as filter media provide for the rapid
and stable nitrification of landfill leachate. At 251C
above 90% nitrification efficiency with loadings of 100–
130 mg NH4-N l
1 d
1 were obtained for static upflow
and DFs similarly to the reference SCBP. However, it
must be noted that the maximum loading potential of
the systems is not known as that was not studied.
Previously, in an on-site biofilter study ammonia
removal from municipal landfill leachate above 90%
with an ammonia loading rate of 70 mg N l
1 d
1 was
reported using a trickling filter (commercial carrier
media) at 161C [22]. In practice, nitrification might be
more feasible in an upflow mode than downflow mode
filter due to its longer HRT, which could compensate for
the slower biological reactions at low temperatures. The
apparent porosity and subsequent large surface area of
crushed brick probably promoted the development and
presence of nitrifying biomass. In UFs crushed concrete
demolition waste—an often-available waste material—
could probably also be used as a filter medium, whereas
the applicability of compost material could be limited
owing to the possible flotation of this material in the
upflow mode. In addition, compost per se might be a
source of ammonia as shown at the start-up of the
present DF.
The feasibility of landfill leachate nitrification in low-
temperature conditions shown in the present study was
in line with the results of previous studies (e.g. [5]). In
the present study, a level of nitrification above 95% was
obtained in the pilot study UF at 51C and 101C with a
loading rate of approx. 50 mg NH4-N l
1 d
1, whereas in
a laboratory SCBP study [5] complete ammonium
removal was obtained at 51C and 101C with loading
rates of 25 and 50 mg NH4-N g l
1 d
1, respectively.
Given the similar level of loading rates applied on these
two processes, the use of waste material as biofilter
media for full-scale application might also be economic-
ally attractive.
When applying in-situ nitrogen removal, leachate
COD and ammonia concentrations and their changes
need to be taken into account, especially in landfills
containing leachate with varying characteristics and age.
This was indicated in the present study by the fact that
leachate with an exceptionally high COD value severely
depressed nitrification, probably owing the competitive
advantage of heterotrophs over nitrifiers. Onay and
Pohland [7] also concluded that the anaerobic treatment
 J.P.Y. Jokela et al. / Water Research 36 (2002) 4079–4087
of leachate prior to in-situ nitrification resulted in a
higher level of nitrification.
5. Conclusions
The present study showed that nitrogen can be removed
effectively from municipal landfill leachate by using a
nitrifying upflow biofilter with waste material as a filter
medium combined with subsequent denitrification of the
nitrified leachate in the landfill body. The loading rate of
nitrified leachate to the landfill body can be at least 3.8 g
TON t-TS
1 waste d

1
without any adverse effect on methana-
tion of waste. Crushed waste brick can be used as a carrier
material in UFs for nitrification of municipal landfill
leachate with loading rates of 100–130 and 50 mg NH4-
N l
1 d
1 at 251C and at 51C, respectively.
Acknowledgements
This study was sponsored by the Finnish Solid Waste
Association, the Tekes Technology Development Centre
(grant no 40370/99 and 40064/98 in Water Services
2001-programme) and Jyv.askyl.a Science Park. We wish
to thank the personnel at the Kiertokapula landfill for
their cooperation in the pilot study and to the staff at the
Mustankorkea and Amm. . assuo landfills for providing
the leachate and waste samples.
References
[1] Ehrig HJ. Leachate quality. In: Christensen TH, Cossu R,
Stegmann R, editors. Sanitary landfilling: Process, tech-
nology and environmental impact. London: Academic
Press, 1989. p. 213–30.
[2] Senior E, Balba MTN. Landfill biotechnology. In: Wise
DL, editor. Bioenvironmental systems, vol. II. Boca
Raton, FL: CRC Press, 1987. p. 17–65.
[3] Tiedje JM. Ecology of denitrification and dissimilatory
nitrate reduction to ammonium. In: Zehnder AJB, editor.
Biology of anaerobic microorganisms. New York: Wiley,
1988. p. 179–245.
[4] Kettunen RK. Treatment of landfill leachates by low-
temperature anaerobic, sequential anaerobic–aerobic pro-
cesses, Ph.D. thesis, Tampere University of Technology,
Finland, 1997.
[5] Hoilijoki TH, Kettunen RH, Rintala JA. Nitrification of
anaerobically pretreated municipal landfill leachate at low
temperature. Water Res 2000;34:1435–46.
[6] Welander U, Henrysson T, Welander T. Biological
nitrogen removal from municipal landfill leachate in a
pilot scale suspended carrier biofilm process. Water Res
1998;32(5):1564–70.
[7] Onay TT, Pohland FG. In situ nitrogen management
in controlled bioreactor landfills. Water Res 1998;32:
1383–92.
4087
[8] Burton SAQ, Watson-Craik IA. Ammonia production and
turnover in landfilled refuse. In: Proceedings of Sixth
International Landfill Symposium, Margherita di Pula S,
Cagliari, Sardinia 97, 13–17 October 1997, Italy, vol. 1,
1997. p. 127–36.
[9] Chen K-C, Lin Y-F. The relationship between denitrifying
bacteria and methanogenic bacteria in a mixed culture
system of acclimated sludge. Water Res 1993;27:1749–59.
[10] Fang HHP, Zhou G-M. Interactions of methanogens and
denitrifiers in degradation of phenols. J Environ Eng
1999;125:57–63.
[11] Chidthaisong A, Conrad R. Turnover of glucose and
acetate coupled to reduction of nitrate, ferric iron and
sulphate and to methanogenesis in anoxic rice field soil.
FEMS Microbiol Ecol 2000;31:73–86.
[12] Jokela JPY, Kettunen RH, Marttinen SK, Rintala JA.
Influence of waste moisture content on methane produc-
tion and leachate characteristics. In: Proceedings of
Seventh Waste Management and Landfill Symposium, S.
Margherita di Pula, Cagliari, Sardinia 99, 4–8 October
1999, Italy, vol. 1, 1999. p. 67–74.
[13] Perstorp Analytical/Tecator AB. The determination of
nitrogen according to kjeldahl using block digestion, steam
distillation. Tecator application note, 1995.
[14] SFS standard 5504. Determination of chemical oxygen
demand (CODcr) in water with the closed tube method,
oxidation with dichromate. Finnish Standard Association,
Helsinki, Finland (in Finnish), 1988.
[15] SFS standard 5508. Determination of biochemical oxygen
demand (BOD) of water, dilution method. Finnish
Standard Association, Helsinki, Finland (in Finnish),
1991.
[16] Burton SAQ, Watson-Craik IA. Nitrogen balances during
the accelerated decomposition of municipal solid waste. In:
Mata-Alvarez J, Tilche A, Cecchi F, editors. Proceedings
of the Second International Symposium on Anaerobic
Digestion of Solid Waste, Barcelona, vol. 1, Gr"afiques 92,
15–18 June, 1999. p. 212–217.
[17] Akunna JC, Bizeau C, Moletta R. Denitrification in
anaerobic digesters: possibilities and influence of waste-
water COD/NOx ratio. Environ Technol 1992;13:825–36.
[18] Airiola S, Lakso E. Formation of landfill gas in Northern
Finland landfills. Finn J Water Econ, Hydraul Agric Eng
996;5:23–5 (in Finnish).
[19] Ettala M. Short-rotation tree plantations and hydrological
aspects in landfill management. Dissertation, University of
Technology, Otaniemi, Finland, 1988.
[20] Reinhart DR, Townsend TG. Landfill bioreactor design &
operation. Boca Raton: Lewis Publishers, 1998.
[21] Saarela J. Hydraulic approximation of infiltration char-
acteristics of surface structures on closed landfills. Mono-
graphs of the Boreal Environment Research, no. 3. Finnish
Environment Institute, 139, Finland (in Finnish), 1997.
[22] Knox K. Leachate treatment with nitrification of ammo-
nia. Water Res 1985;19:895–904.
[23] American Public Health Association, American Water
Works Association & Water Environment Federation. In:
Clesceri LS, Greenberg AE, Eaton AD, editors. Standard
Methods for the Examination of Water and Wastewater,
20th edn. Washington, DC: American Public Health
Association, 1998.