Environmental Technology

Vol. 33, No. 8, April 2012, 865–872

Impact of solid retention time and nitrification capacity on the ability of activated
sludge to remove pharmaceuticals
P. Falåsa∗ , H.R. Andersenb , A. Ledinb and J. la Cour Jansena
a Waterand Environmental Engineering, Department of Chemical Engineering, Lund University, P.O. Box 124, SE-221 00 Lund,
Sweden; b Department of Environmental Engineering, Technical University of Denmark, B113, DK-2800 Kgs Lyngby, Denmark
(Received 23 December 2010; final version received 25 June 2011 )

Removal of five acidic pharmaceuticals (ibuprofen, ketoprofen, naproxen, diclofenac and clofibric acid) by activated sludge
from five municipal activated sludge treatment processes, with various sludge ages and nitrification capacities, was assessed
through batch experiments. The increase in aerobic sludge age from 1–3 to 7 d seemed to be critical for the removal of naproxen
and ketoprofen, with markedly higher rates of removal at sludge ages of 7 d or more. No removal was shown for diclofenac and
clofibric acid, whereas high rates were observed for ibuprofen in all investigated sludges. Parallel examinations of activated
sludge batches with and without allylthiourea (12 mg/L), an inhibitor of ammonia monooxygenase, showed minor to moderate
influence on the removal rates of ketoprofen and naproxen. These results suggest that the removal rates of biodegradable
pharmaceuticals in municipal activated sludge processes are strongly linked to the heterotrophic bacteria community.
Keywords: activated sludge; nitrification; solid retention time; pharmaceuticals

Introduction [9] and estrogens [10]. For pre-cultivated activated sludge

A significant amount of the pharmaceuticals used in society enriched in nitrifying bacteria, the presence of allylth-
enter the wastewater system unchanged or as metabolites. iourea (ATU), an inhibitor of ammonia monooxygenase,
The ability of wastewater treatment plants (WWTPs) to reduced the biological removal of several micropollutants
remove these substances will affect the pharmaceutical con- but it did not stop the removal completely [10–12]. These
centration in the receiving water and their adverse effects findings indicate that both nitrifying and heterotrophic
on the aquatic environment. The removal efficiency for bacteria species in activated sludge can remove micropol-
biodegradable pharmaceuticals at WWTPs seems to be lutants. Because of the pre-cultivation, little is still known
influenced by the biological treatment method [1,2], of about these bacteria groups’ individual contribution to the
which the activated sludge process is the most frequently removal of specific micropollutants in typical activated
used. sludge processes. It is therefore crucial for the optimiza-
The influence of solid retention time (SRT) on the phar- tion of pharmaceutical removal to evaluate the relative
maceutical removal in activated sludge processes is widely importance of these two bacteria groups in real activated
discussed [3–5]. For natural estrogens and some micropol- sludge with various SRTs and nitrification capacities. This
lutants, Clara et al. [6] proposed a critical SRT, above which study aims to assess the influence of SRT and the influ-
significant removal can be achieved, of 10 d at a reference ence of the nitrification capacity at different SRTs for the
temperature of 10◦ C. Owing to the lack of data for SRTs lipid regulator clofibric acid and four frequently used anti-
between 2 and 10 d, however, the critical SRT may be found inflammatory substances – ibuprofen, naproxen ketoprofen
within this range, rather than at 10 d. The temperature- and diclofenac. In order to get good coverage of SRTs and
dependent growth of Nitrosomonas europaea [7] suggests nitrification capacities, activated sludge from five Swedish
that this critical SRT range covers the SRT required for WWTPs were assessed: two with low aerobic sludge age
growth of bacteria responsible for oxidizing ammonia (1–3 d) and low nitrification capacity, and three with high
to nitrite in WWTPs with extended biological nitrogen aerobic sludge age (7–15 d) and pronounced nitrification
removal. Ammonia oxidation involves ammonia monooxy- capacity. The sludges’ ability to remove the five pharma-
genase, which has been reported to co-oxidize organic ceuticals is studied in batch reactors, with and without ATU,
substances, and pure cultures of ammonia-oxidizing bac- and the results obtained are discussed with regard to SRT,
teria have been demonstrated to oxidize a large variety of nitrifying capacity and the activity of the heterotrophic
aliphatic compounds [8], aromatic halogenated compounds bacteria, given as oxygen uptake rate.

∗ Corresponding author. Email: per.falas@chemeng.lth.se

ISSN 0959-3330 print/ISSN 1479-487X online
© 2012 Taylor & Francis
http://dx.doi.org/10.1080/09593330.2011.601764
http://www.tandfonline.com
866 P. Falås et al.

Materials and methods low-loaded activated sludge process was mainly designed
Sampling sites for BOD removal and nitrification, but it can also be
operated with partial pre-denitrification. Denitrification
Källby WWTP in Lund receives mainly domestic wastew-
occurred mainly in an MBBR process, which was sepa-
ater from 80,000 people. Incoming wastewater, with annual
rate from the activated sludge treatment. At the time of
average concentrations of approximately 180 mg/L BOD7
sampling, the wastewater temperature was 10 ◦ C, the total
(biochemical oxygen demand) and 40 mg/L TN (total nitro-
sludge age was approximately 7 d and the activated sludge
gen) was treated mechanically (screening, grit removal
basin was aerated over its full length. The HRT in the line
and sedimentation).The subsequent low-loaded activated
of sampling was 5–6 h.
sludge process was operated with enhanced biological
Öresundsverket WWTP in Helsingborg receives
phosphorous removal, pre-denitrification with side-stream
wastewater from 120,000 people and, further, a wide
hydrolysis to provide additional carbon source, and nitrifi-
range of industries. Incoming wastewater, with annual
cation in the aerated compartments. Post-precipitation was
average concentrations of approximately 180 mg/L BOD7
used as a complement to the enhanced biological phos-
and 30 mg/L TN, was treated mechanically (screening,
phorous removal. At the time of sampling, the wastewater
grit removal, and sedimentation with primary sludge
temperature was 17 ◦ C, the total sludge age was 17–18 d and
hydrolysis). The subsequent, low-loaded activated sludge
the volumetric ratio between anaerobic, anoxic and aero-
process was operated with enhanced biological phospho-
bic zones was approximately 2:3:5. The hydraulic retention
rous removal, pre-denitrification with hydrolysed primary
time (HRT) in the line of sampling was approximately 18 h.
sludge as additional carbon source, and nitrification in the
Rya WWTP in Gothenburg receives wastewater from
aerated zones. No chemicals for phosphorus removal were
630,000 people and, further, a wide range of industries.
used at the plant. At the time of sampling, the wastewa-
Incoming wastewater, with annual average concentrations
ter temperature was 10 ◦ C, the total sludge age was 25–30 d
of approximately 130 mg/L BOD7 and 30 mg/L TN, was
and the volumetric ratio between anaerobic, anoxic and aer-
treated mechanically (screening, grit removal and sedi-
obic zones was approximately 1:4:5. The HRT in the line
mentation). The subsequent high-loaded activated sludge
of sampling was approximately 14 h.
process was operated in a pre-denitrification mode with
simultaneous precipitation and addition of ethanol. The aer-
ated compartments were operated solely for BOD removal, Collection of activated sludge and initial measurements
and nitrification took place in a nitrifying trickling filter,
with recirculation to the activated sludge process. At the Activated sludge was collected at the end of the last aerated
time of sampling, the wastewater temperature was 13 ◦ C, the compartment and transported directly to the laboratory for
total sludge age was 3–4 d and the volumetric ratio between analysis of pH, NO− − +
3 -N, NO2 -N, NH4 -N, suspended solids
the anoxic and aerobic zones was approximately 1:1. The (SS), volatile suspended solids (VSS) and chemical oxygen
HRT of the activated sludge treatment was 3–4 h. demand (COD). Storage of the activated sludge at the lab-
Sjölunda WWTP in Malmö receives wastewater from oratory before the batch experiments were made at 4 ◦ C for
300,000 people and, further, a wide range of industries. less than 24 h.
Incoming wastewater, with annual average concentrations
of approximately 220 mg/L BOD7 and 40 mg/L TN, was
Batch experiments with activated sludge
treated mechanically (screening, grit removal, and sedi-
mentation with chemical precipitation). The subsequent, Activated sludge (5 L) and pharmaceuticals (100 μg/L)
high-loaded activated sludge process was operated with an were added to two completely mixed, fully aerated (O2 6–
anaerobic/anoxic zone at the inlet for denitritation of aer- 9 mg/L) parallel batch reactors, kept at 20 ± 2 ◦ C. Allylth-
obically treated reject water and an aerobic zone for BOD iourea, 12 mg/L, was added to one of the reactors. Samples,
removal at the outlet. The extended nitrogen removal at the 300 mL, for pharmaceutical analysis were withdrawn at
plant took place in a nitrifying trickling filter followed by 0.17, 1, 2, 3, 4, 6, 8, 10, 12, 15, 18 and 24 h. Samples for esti-
a moving-bed biofilm reactor (MBBR) for denitrification. mation of the nitrification (NH+ − −
4 -N, NO3 -N, NO2 -N) and
At the time of sampling, the wastewater temperature was determination of alkalinity, SS, VSS and COD were taken
19 ◦ C, the total sludge age was 2–2.5 d and the volumetric from the reactors. Oxygen and pH levels were recorded in
ratio between the anaerobic/anoxic and aerobic zones was the reactors, with electrodes connected to the meters WTW
approximately 1:3. The HRT in the line of sampling was Oxi 730 and WTW pH 320, respectively.
approximately 4 h.
Klagshamn WWTP in Malmö receives wastewater from
70,000 people. Incoming wastewater, with annual aver- Sample preparation and pharmaceutical analysis
age concentrations of 130 mg/L BOD7 and 30 mg/L TN, Samples for pharmaceutical analysis were centrifuged
was treated mechanically (screening, grit removal, and sed- and filtered through GFC-filters. Two hundred and fifty
imentation with chemical precipitation). The subsequent, millilitres of the filtered samples were acidified to
 Environmental Technology 867

pH 3 with a phosphate buffer, and an internal standard, with Dr Lange cuvettes (LCK 303 and LCK 114, respec-
mecoprop (40 μg/L), was added before extraction. The tively), whereas NO− −
2 -N, and NO3 -N were measured
SPE-cartridges, Oasis HLB 3cc, were conditioned serially either with a Technicon Autoanalyser II or Dr Lange
with 3 mL each of methanol, ethyl acetate and acidi- cuvettes (LCK 341 and LCK 339, respectively). Alka-
fied water. The extraction rates were below 2 mL/min linity was measured according to Swedish standard (ISO
and the cartridges were dried completely after extrac- 9963-1:1994) [16].
tion. Samples were eluted with 1.5 mL of ethyl acetate,
and the extracts were evaporated under a gentle stream Results and discussion
of N2 at 35 ◦ C to a volume of approximately 250 μL.
Nitrification capacity and OUR
Remaining eluates were transferred to GC vials, and
25 μL of the derivatization agent, N-(t-butyldimetylsilyl)- Activated sludge from the high-loaded processes with low
N-methyltrifluoroacetamid (MTBSTFA), was added and SRTs (2–4 d) showed insignificant nitrification capacities,
allowed to react for 60 min at 60 ◦ C. whereas sludge from the low-loaded processes with mod-
Quantification of the derivates of each pharmaceutical erate SRTs (7–30 d) showed normal to high nitrification
was made with a GC/MS, Agilent 5973N Mass Selec- capacities. Higher respiratory activities were recorded for
tive Detector. The capillary column was an Agilent HP sludges with low SRTs than for sludges with moderate SRTs
5-MS (30.0 m × 250 μm × 0.25 μm) with 1 μL injection (Table 1). Decreasing OUR at increasing SRT has also been
in splitless mode. The limit of detection for the method observed by Kristensen et al. [14].
on wastewater effluents was determined according to DIN
32645 [13] based on linearity and was found at 2.1 μg/L for Reactor conditions and performance
ibuprofen, 2.4 μg/L for naproxen, 3.0 μg/L for ketoprofen, Table 2 shows measured SS, VSS, COD, NH+ −
4 -N, and NOx -
2.7 μg/L for diclofenac and 2.0 μg/L for clofibric acid. − −
N (NO3 -N + NO2 -N) concentrations, and changes in pH
and alkalinity. For the three sludges with significant nitri-
fication capacities, the amount of NH+ 4 -N formed through
Nitrification capacity measurements
sludge hydrolysis in the presence of ATU correlates well
Nitrification capacity measurements of the sludges were with the amount of NO− x -N produced in the absence of
carried out according to Kristensen et al. [14]. A fully aer- ATU. This indicates complete inhibition of the ammonia
ated (O2 6–9 mg/L) and completely mixed reactor with oxidation when ATU is present. Furthermore, formation
400 mL of activated sludge was kept at 20 ◦ C. Nutrients of approximately 2 mg NO− x -N/(L·d) in the sludge from
and alkalinity (NaHCO3 0.672 g/L, (NH4 )2 SO4 0.236 g/L, Sjölunda WWTP and Rya WWTP suggests that ammonia-
and KH2 PO4 0.044 g/L) were added and samples were oxidizing bacteria are present even in sludge with low SRT.
taken every 15 min for 2–4 h. Samples were filtered (Munk- The SS and VSS concentrations used in the rectors corre-
tell 1002), and nitrogen fractions (NH+ −
4 -N, NO3 -N and spond to the concentrations found at the plant at the time of
−
NO2 -N) were measured. sampling.

Oxygen uptake rate (OUR) Pharmaceutical removal

Measurements of OUR were carried out according to Figure 1 shows changes in concentrations of dissolved
Hagman et al. [15]. Activated sludge (900 mL) was added pharmaceuticals in the batch reactors with sludge from
to two parallel reactors, which were mixed with magnetic Klagshamn WWTP, with and without ATU. The first
stirrers and kept at 20 ◦ C. Nutrients ((NH4 )2 SO4 0.236 g/L samples were taken after 10 min and can be expected
and KH2 PO4 0.044 g/L), alkalinity (NaHCO3 0.612 g/L) to correspond to the start concentration excluding poten-
and ATU (12 mg/L) were added. The reactors were aer- tial sorption, since the removal due to biodegradation is
ated at intervals, with 5 min of aeration followed by 5 min expected to be very limited within this short time frame.
with no aeration, and the oxygen levels were recorded with However, the observed concentration at 10 min correspond
electrodes connected to the oxygen meters, WTW Oxi 197- to the amount of pharmaceuticals added, which indicates
S and WTW Oxi 730. Acetate (80 mg/L) was added after limited removal through sorption. Sorption to activated
approximately 1 h of intermittent aeration and the total dura- sludge is a rapid process, which has been shown by Ternes
tion of the OUR recording was 4–5 h. Concentrations of SS et al. [17]. They reported that solid–water partitioning
and VSS were measured in the reactors at the end of the equilibriums were reached within 30 min for many pharma-
experiment. ceuticals, including diclofenac, ibuprofen and clofibric acid,
and <10% of these three substances sorbed to the activated
sludge at a SS concentration of 4 g/L. Based on the par-
Analysis of nitrogen fractions, COD and alkalinity titioning coefficients presented by Urase and Kikuta [18],
Nitrogen and COD fractions were measured spectrophoto- <15% sorption to activated sludge (SS 4 g/L, pH 6.7) can
metrically. Ammonium-nirogen and COD were measured be expected for naproxen, ketoprofen and diclofenac and
868 P. Falås et al.

Table 1. Recorded OURs and nitrification capacities for the sludges together with their total SRT and aerobic SRT.

Nitrification rate (max) OUR (max) OUR (endogenous) SRT (total) SRT (aerobic)
mg NO− x -N/(g VSS·h) mg O2 /(g VSS·h) mg O2 /(g VSS·h) d d

Sjölunda <0.1 29 7.1 2–3 1.5–2.5

Rya <0.1 28 6.5 3–4 1.5–2.0
Källby 4.3 11 2.7 17–18 8–9
Öresundsverket 5.4 18 5.2 25–30 12–15
Klagshamn 8.7 18 5.2 7 7

Table 2. Measured SS, VSS, COD, NH+ −

4 -N and NOx -N concentrations, pH and alkalinity. Superscript notations indicate time of
sampling: i = initial (before addition of pharmaceuticals and ATU), b = beginning of batch experiment and e = end of batch experiment.

Sjölunda Rya Källby Öresundsverket Klagshamn

No ATU With ATU No ATU With ATU No ATU With ATU No ATU With ATU No ATU With ATU

SS (g/L) 3.46 3.47 1.86 1.82 4.30 4.25 4.75 5.17 2.43 2.54
VSS (g/L) 3.05 3.02 1.36 1.34 3.16 3.14 3.47 3.69 1.93 2.01
COD (mg/L) 145i 145i 79i 79i 66i 66i 53i 53i 60i 60i
255e 172e 135e 161e 59e 78e 75e 103e 75e 101e
NH+
4 -N (mg/L) 29i 29i 10b 10b <2i <2i <2b <2b <2b <2b
37e 36e 9e 10e <2e 6e <2e 4e <2e 7e
NO−
x -N (mg/L) 0.4i 0.4i 0.1b 0.0b 2.2i 2.2i 0.6b 0.4b 8.6b 8.4b
2.8e 0.3e 2.1e 0.3e 8.6e 2.3e 6.3e 0.8e 19e 8.6e
pH 8.3b 8.3b 7.7b 7.8b 8.0b 7.9b 7.8b 8.0b 8.1b 8.1b
8.5e 8.6e 8.2e 8.4e 8.0e 8.0e 7.8e 8.2e 8.3e 8.6e
Alkalinity (mmol 8.8i 8.8i 4.1i 4.1i 6.6i 6.6i 5.1i 5.1i 4.9i 4.9i
HCO− 3 /L)
8.8e 8.8e 3.9e 4.0e 5.0e 6.5e 5.6e 6.0e 3.5e 4.9e

Figure 1. Pharmaceutical concentrations in reactors with activated sludge from Klagshamn WWTP, with and without ATU. Lines indicate
the zero and the first-order removal that gives the best fit to the removal of ibuprofen, naproxen and ketoprofen in the uninhibited sludge.

<25% for ibuprofen. According to these studies, removal in the uninhibited sludge from Klagshamn (Figure 1 and
through sorption can be expected to be negligible for all Table 3). It can be expected that zero-order kinetics gives
studied substances except ibuprofen, for which limited to good representation of the removal capacity of the sludge at
moderate sorption has been observed. high pharmaceutical concentrations, whereas the removal
No removal was observed for diclofenac and clofib- at lower concentrations is better described by first-order
ric acid in any of the sludges examined, which might be kinetics. First-order kinetics has previously been observed
explained by their chlorinated structures. Based on the R2 to give adequate representation of the removal [19]. Further-
values (Table 3), it is not clear whether the removal of more, it has been suggested that the presence of degradable
biodegradable pharmaceuticals in this concentration range substrates might suppress the removal of pharmaceuticals
is best described by zero or first-order kinetics. The R2 [18,20]. The addition of methanol (COD 150–300 mg/L)
values for the zero and first-order kinetics were gener- as the pharmaceutical solvent might therefore influence the
ally similar, and the highest discrepancy between the fit of removal regime, because of an initial surplus of organic
the zero and the first-order kinetics was seen for naproxen substrate, followed by more substrate-limited conditions.
 Environmental Technology 869

Table 3. Zero and first-order removal rate constants (K0 and K1 , respectively) obtained from batch experiments with activated sludge.
Rate constants are given as mean ±95% confidence interval with R2 values and as ∼0 when the confidence interval covers negative values.

Ibuprofen Naproxen Ketoprofen

K0 K1 K0 K1 K0 K1
μg/(g VSS·h) L/(g VSS·h) μg/(g VSS·h) L/(g VSS·h) μg/(g VSS·h) L/(g VSS·h)

Sjölunda 5.0 ± 0.5 0.09 ± 0.05 0.90 ± 0.21 0.010 ± 0.004 ∼0 ∼0

(R2 = 0.99) (R2 = 0.92) (R2 = 0.90) (R2 = 0.84)
Sjölunda (ATU) 2.4 ± 0.2 0.05 ± 0.01 0.19 ± 0.18 0.002 ± 0.002 ∼0 ∼0
(R2 = 0.99) (R2 = 0.98) (R2 = 0.39) (R2 = 0.38)
Rya 20.2 ± 5.0 0.41 ± 0.25 0.95 ± 0.31 0.012 ± 0.004 ∼0 ∼0
(R2 = 0.99) (R2 = 0.98) (R2 = 0.82) (R2 = 0.83)
Rya (ATU) 18.3 ± 3.1 0.39 ± 0.21 0.94 ± 0.31 0.012 ± 0.004 ∼0 ∼0
(R2 = 0.99) (R2 = 0.96) (R2 = 0.82) (R2 = 0.85)
Källby 9.2 ± 5.8 0.24 ± 0.09 2.30 ± 0.47 0.043 ± 0.015 0.67 ± 0.16 0.010 ± 0.003
(R2 = 0.96) (R2 = 0.99) (R2 = 0.96) (R2 = 0.93) (R2 = 0.90) (R2 = 0.89)
Källby (ATU) 9.3 ± 4.5 0.20 ± 0.10 2.32 ± 0.49 0.037 ± 0.018 0.71 ± 0.26 0.009 ± 0.004
(R2 = 0.98) (R2 = 0.99) (R2 = 0.97) (R2 = 0.91) (R2 = 0.81) (R2 = 0.79)
Öresundsverket 6.8 ± 3.4 0.19 ± 0.04 2.41 ± 0.25 0.047 ± 0.015 0.65 ± 0.18 0.007 ± 0.003
(R2 = 0.93) (R2 = 0.99) (R2 = 0.99) (R2 = 0.94) (R2 = 0.86) (R2 = 0.82)
Öresundsverket (ATU) 3.9 ± 1.4 0.10 ± 0.02 2.10 ± 0.17 0.040 ± 0.011 0.43 ± 0.14 0.005 ± 0.002
(R2 = 0.94) (R2 = 0.99) (R2 = 0.99) (R2 = 0.95) (R2 = 0.82) (R2 = 0.80)
Klagshamn 7.9 ± 2.7 0.21 ± 0.05 2.66 ± 0.36 0.040 ± 0.014 1.09 ± 0.21 0.013 ± 0.003
(R2 = 0.94) (R2 = 0.99) (R2 = 0.97) (R2 = 0.88) (R2 = 0.93) (R2 = 0.91)
Klagshamn (ATU) 8.2 ± 3.6 0.23 ± 0.05 2.52 ± 0.35 0.041 ± 0.011 1.00 ± 0.16 0.013 ± 0.002
(R2 = 0.91) (R2 = 0.99) (R2 = 0.97) (R2 = 0.92) (R2 = 0.95) (R2 = 0.96)

The observed first-order removal rates for the uninhib- minor to moderate influence on the removal rates. In con-
ited sludges were lower or similar to the rates presented by trast, Tran et al. [12] observed a 50–60% decrease in the
Joss et al. [20], Tran et al. [12], and Suarez et al. [21], but removal of ketoprofen and naproxen when ATU (10 mg/L)
were higher or similar to the rates presented by Urase and was added to a 6 d batch experiment with activated sludge
Kikuta [18], at pH 6.7. enriched in nitrifying bacteria and an initial NH+ 4 -N con-
centration of 100 mg/L. This difference may be explained
by the enrichment process or the initial addition of NH+ 4 -N
Effects of ATU inhibition, nitrification capacity and SRT as growth substrate and nitrogen source. Increased removal
The removal rates of ibuprofen, naproxen and ketoprofen of ketoprofen and naproxen was also observed by Tran et al.
were generally unaffected or moderately reduced by the [12] when the initial NH+ 4 -N concentration was increased
addition of ATU (Figure 2). Markedly impaired removal from 20 to 100 mg/L. Forrez et al. [24], on the other hand,
in the presence of ATU was only observed for ibuprofen observed no increased removal rate of ethinylestradiol in
in a nitrifying sludge and for ibuprofen and naproxen in a batch tests with an active, but ammonium-starved, nitrify-
sludge with insignificant nitrification capacity. Since ATU ing culture when a more realistic initial NH+ 4 -N dose of
addition only affected the pharmaceutical removal in one of 10.6 mg/L was added.
the two sludges with insignificant nitrification capacity, it The batch experiments indicated that inhibition of
seems more likely that ATU inhibits some enzymes of het- ammonia monooxygenase by ATU in nitrifying activated
erotrophic bacteria relevant to the pharmaceutical removal sludges from municipal full-scale installations has limited
rather than that very low abundances of ammonia-oxidizing impact on the removal rates of ketoprofen and naproxen.
bacteria can induce detectable differences in the removal. In addition, no dependency between the nitrification capac-
Allylthiourea inhibition of enzymes other than ammonia ity and the removal rates was observed for the nitrifying
monooxygenase has been observed [22], and their potential sludges (Figure 2). Consequently, the observed differences
role in ethinylestradiol removal in activated sludge has been in the removal rates of ketoprofen and naproxen between
addressed by Gaulke et al. [23]. nitrifying and non-nitrifying sludges were not primarily
The activated sludges with nitrification capacities caused by a higher nitrification capacity or a higher den-
exceeding 4 mg NO− x -N/(g VSS·h) showed considerably sity of ammonia monooxygenase, but rather to differences
higher removal of naproxen and ketoprofen than the non- in the heterotrophic bacterial community. Thus, the increase
nitrifying sludges. The increased removal of naproxen in aerobic SRT from 1–3 to 7 d seemed critical to the
and ketoprofen by the nitrifying sludges could not, how- removal of ketoprofen and naproxen (Figure 2), and fur-
ever, be attributed to the presence of available ammonia thermore this increased removal was mainly induced by
monooxygenase alone, since the addition of ATU showed the heterotrophic bacteria. This critical SRT coincided
870 P. Falås et al.

Figure 2. Relationship between the nitrification capacity and the aerobic SRT of the sludge and the first-order removal rate constants for
ibuprofen, naproxen and ketoprofen – with and without ATU. Note differences in scaling of the y-axis. T-bars indicate 95% confidence
intervals for the removal rate constants and numbers indicate sludges. (1): Rya; (2): Sjölunda; (3): Källby; (4): Öresundsverket; (5):
Klagshamn.

with a decrease in the total heterotrophic activity, given sludge age from 1–3 to 7 d is critical to the bacteria
as a decrease in both the maximum and the endogenous involved in ring cleavage of compounds structurally related
OUR. Consequently, the critical SRT for naproxen and to biphenyls. The transformation products identified for
ketoprofen was not induced by increased total biologi- naproxen and ibuprofen [25] suggests that the bacteria tar-
cal activity but by a shift in the heterotrophic bacterial gets the 2-methylpropyl end of ibuprofen and the methoxy
community. end of naproxen. The lower removal rate of naproxen com-
Based on the microbial degradation pathways and trans- pared with ibuprofen might be due to the stabilizing ether
formation products presented by Quintana et al. [25], it bond in the methoxy group.
seems likely that ketoprofen is degraded through ring For ibuprofen, no clear dependency on SRT or nitrifi-
cleavage analogous to the degradation pathway observed cation capacity was observed, and the highest and lowest
for biphenyl-related compounds. The distinct difference in removal rates were found at an SRT of approximately
ketoprofen removal between nitrifying and non-nitrifying 2 d. The removal rates for ibuprofen were, however, fairly
sludges may, consequently, indicate that the increase in consistent for all sludges with SRTs of 7 d or more.
 Environmental Technology 871

Conclusions [5] N. Kreuzinger, M. Clara, B. Strenn, and H. Kroiss, Rele-

vance of the sludge retention time (SRT) as design criteria
The removal rate of the selected pharmaceuticals depended for wastewater treatment plants for the removal of endocrine
on the intrinsic properties of the pharmaceutical and the disruptors and pharmaceuticals from wastewater, Water Sci.
bacteria composition of the sludge. Diclofenac and clofib- Technol. 50(5) (2004), pp. 149–156.
ric acid showed no removal in any of the five municipal [6] M. Clara, N. Kreuzinger, B. Strenn, O. Gans, and H.
activated sludges studied, whereas the removal rate of keto- Kroiss, The solids retention time – a suitable design param-
eter to evaluate the capacity of wastewater treatment
profen, naproxen and ibuprofen depended on the bacterial plants to remove micropollutants, Water Res. 39 (2005),
composition of the sludge. pp. 97–106.
For the nitrifying sludges, no dependency between the [7] J.H. Hunik, Engineering aspects of nitrification with
nitrification capacity and the pharmaceutical removal was immobilized cells, Ph.D. thesis, Wageningen University,
observed. 1993.
[8] M.R. Hyman, I.B. Murton, and D.J. Arp, Interaction of
The inhibition of ammonia monooxygenase by ATU had ammonia monooxygenase from Nitrosomonas europaea with
most often a limited effect on the pharmaceutical removal. alkanes, alkenes, and alkynes, Appl. Environ. Microbiol. 54
The results indicate that the removal of ibuprofen, (1988), pp. 3187–3190.
naproxen and ketoprofen in municipal activated sludge pro- [9] W.K. Keener and D.J. Arp, Transformations of aromatic
cesses is highly dependent on the heterotrophic bacteria compounds by Nitrosomonas europaea, Appl. Environ.
Microbiol. 60 (2004), pp. 1914–1920.
community and that ammonia-oxidizing bacteria may only [10] J. Shi, S. Fujisawa, S. Nakai, and M. Hosomi, Biodegra-
have a minor role. dation of natural and synthetic estrogens by nitrifying
For ketoprofen and naproxen, a correlation between the activated sludge and ammonia-oxidizing bacterium Nitro-
removal rate and the SRT was observed and the increase in somonas europaea, Water Res. 38 (2004), pp. 2323–
aerobic SRT from 1–3 to 7 d seemed critical. 2330.
[11] J.Y. Kim, K. Ryu, E.J. Kim, W.S. Choe, G.C. Cha, and I. Yoo,
Within the larger context of pharmaceutical removal at Degradation of bisphenol A and nonylphenol by nitrifying
WWTPs, these results indicate that: (1) the role of het- activated sludge, Process Biochem. 42 (2007), pp. 1470–
erotrophic bacteria are important to the optimization of the 1474.
pharmaceutical removal in activated sludge systems, (2) an [12] N.H. Tran, T. Urase, and O. Kusakabe, The characteristics
increase in aerobic SRT from 1–3 to 7 d is decisive for the of enriched nitrifier culture in the degradation of selected
pharmaceutically active compounds, J. Hazard. Mater. 171
biological removal of some pharmaceuticals, and (3) acti- (2009), pp. 1051–1057.
vated sludge treatment plants with aerobic SRTs of 1–3 d [13] DIN 32645, Chemical analysis: Decision limit, detection
might have a greater need for complementary pharmaceu- limit and determination limit; estimation in case of repeata-
tical removal than corresponding plants with SRTs of 7 d bility, terms, methods, evaluation, 1994.
or more. [14] G.H. Kristensen, P.E. Jørgensen, and M. Henze, Character-
ization of functional microorganism groups and substrate in
activated sludge and wastewater by AUR, NUR and OUR,
Acknowledgements Water Sci. Technol. 25(6) (1992), pp. 43–57.
VA SYD (Klagshamn WWTP, Sjölunda WWTP and Källby [15] M. Hagman, J.L. Nielsen, P.H. Nielsen, and J. la Cour Jansen,
WWTP), Gryaab AB (Rya WWTP) and NSVA (Öresundsver- Mixed carbon sources for nitrate reduction in activated
ket WWTP) are all acknowledged for providing sludge for the sludge-identification of bacteria and process activity studies,
experiments. This work is part of the MistraPharma project, which Water Res. 42 (2008), pp. 1539–1546.
is funded by Mistra, the Foundation for Strategic Environmental [16] ISO 9963-1:1994, Water quality – Determination of alkalin-
Research. ity – Part 1: Determination of total and composite alkalinity,
1994.
[17] T.A. Ternes, N. Herrmann, M. Bonerz, T. Knacker, H.
References Siegrist, and A. Joss, A rapid method to measure the solid–
[1] B. Kasprzyk-Hordern, R.M. Dinsdale, and A.J. Guwy, water distribution coefficient (Kd) for pharmaceuticals and
The removal of pharmaceuticals, personal care products, musk fragrances in sewage sludge, Water Res. 38 (2004),
endocrine disruptors and illicit drugs during wastewater pp. 4075–4084.
treatment and its impact on the quality of receiving waters, [18] T. Urase and T. Kikuta, Separate estimation of adsorption
Water Res. 43 (2009), pp. 363–380. and degradation of pharmaceutical substances and estro-
[2] J. Radjenović, M. Petrović, and D. Barceló, Fate and dis- gens in the activated sludge process, Water Res. 39 (2005),
tribution of pharmaceuticals in wastewater and sewage pp. 1289–1300.
sludge of the conventional activated sludge (CAS) and [19] T. Ternes and A. Joss (eds.), Human Pharmaceuticals, Hor-
advanced membrane bioreactor (MBR) treatment, Water mones and Fragrances: The Challenge of Micropollutants
Res. 43 (2009), pp. 831–842. in Urban Water Management, IWA Publishing, London,
[3] X. Zhou and J.A. Oleszkiewicz, Biodegradation of oestro- 2006.
gens in nitrifying activated sludge, Environ. Technol. 31 [20] A. Joss, S. Zabczynski, A. Göbel, B. Hoffmann, D. Löffler,
(2010), pp. 1263–1269. C.S. McArdell, T.A. Ternes, A. Thomsen, and H. Siegrist,
[4] B. Strenn, M. Clara, O. Gans, and N. Kreuzinger, Carba- Biological degradation of pharmaceuticals in municipal
mazepine, diclofenac, ibuprofen and bezafibrate – investiga- wastewater treatment: Proposing a classification scheme,
tions on the behaviour of selected pharmaceuticals during Water Res. 40 (2006), pp. 1686–1696.
wastewater treatment, Water Sci. Technol. 50(5) (2004), [21] S. Suarez, J.M. Lema, and F. Omil, Removal of
pp. 269–276. pharmaceutical and personal care products (PPCPs) under
872 P. Falås et al.

nitrifying and denitrifying conditions, Water Res. 44 (2010), [24] I. Forrez, M. Carballa, N. Boon, and W. Verstraete,
pp. 3214–3224. Biological removal of 17α-ethinylestradiol (EE2) in an
[22] N. Hamamura, R.T. Storfa, L. Semprini, and D.J Arp, aerated nitrifying fixed bed reactor during ammonium
Diversity in butane monooxygenases among butane- starvation, J. Chem. Technol. Biotechnol. 84 (2009),
grown bacteria, Appl. Environ. Microbiol. 65 (1999), pp. 119–125.
pp. 4586–4593. [25] J.B. Quintana, S. Weiss, and T. Reemtsma, Pathways and
[23] L.S. Gaulke, S.E. Strand, T.F. Kalhorn, and H.D. Stensel, metabolites of microbial degradation of selected acidic phar-
17α-ethinylestradiol transformation via abiotic nitration in maceutical and their occurrence in municipal wastewater
the presence of ammonia oxidizing bacteria, Environ. Sci. treated by a membrane bioreactor, Water Res. 39 (2005),
Technol. 42 (2008), pp. 7622–7627. pp. 2654–2664.