Pe o mance o pa ial ni i a ion-anammox p ocesses a
mains eam condi ions in an IFAS sys em
Alba Ped ouso, Joze T ela, Angeles Val del Rio, Anuska Mosque a-Co al and Elzbie a Plaza
Accep ed Mansuc ip
How o ci e:
Ped ouso, A., T ela, J., Val del Rio, A., Mosque a-Co al, A., & Plaza, E. (2019). Pe o mance o pa ial
ni i a ion-anammox
p ocesses a mains eam condi ions in an IFAS sys em. Jou nal O
En i onmen al Managemen , 250, 109538. doi: 10.1016/j.jen man.2019.109538
Copy igh in o ma ion:
© 2019 Else ie L d. This manusc ip e sion is made a ailable unde he CC-BY-NC-ND 4.0 license
(h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/)
1
Pe o mance o pa ial ni i a ion-anammox p ocesses a mains eam condi ions in
an IFAS sys em
Doi: h ps://doi.o g/10.1016/j.jen man.2019.109538
Alba Ped ousoa*, Joze T elab, Angeles Val del Rioa, Anuska Mosque a-Co ala and
Elzbie a Plazab
a Depa men o Chemical Enginee ing, School o Enginee ing, Uni e sidade de San iago de Compos ela,
Rúa de Lope Gómez de Ma zoa s/n, E-15782. San iago de Compos ela, Galicia, Spain. E-mail:
alba.ped[email p o ec ed], man[email p o ec ed], and anuska.mosque [email protected]
b Depa men o Sus ainable De elopmen , En i onmen al Science and Enginee ing, Royal Ins i u e o
Technology (KTH), Teknik ingen 10B, SE-10044, S ockholm, Sweden. E-mail: [email p o ec ed],
[email p o ec ed]
*Co esponding au ho : Alba Ped ouso, Email: alba.ped ous[email p o ec ed], Phone: +34 881816739
ABSTRACT:
The pa ial ni i a ion-anammox p ocesses implemen a ion in he main line o was ewa e
ea men plan s would lead hem close o he ene gy au a ky. Wi h his pu pose, an in eg a ed
ixed ilm ac i a ed sludge (IFAS) eac o was ope a ed a pilo scale. E icien ni ogen
emo al (72 ± 11 %) was achie ed o anae obically p e- ea ed municipal was ewa e a low
empe a u e (21 - 15 ºC), wi h a ni ogen emo al a e o 37 ± 3 g N/(m3·d) a 15 °C. The
ammonium oxidizing bac e ia we e mo e abundan in he ac i a ed sludge, while anammox
bac e ia we e p ima ily loca ed in bio ilm a ached on o he ca ie s su ace. Ni i e oxidizing
bac e ia (NOB) ac i i y was simila be ween bo h ac ions and i s speci ic ac i i y dec eased
mo e han ha o o he popula ions when he ope a ing empe a u e was educed. Fu he mo e,
he IFAS ope a ional s a egy (ae obic/anoxic pe iods) allowed an e icien NOB ac i i y
supp ession inside he eac o , which accoun ed only o he 10 - 20 % o he maximum
po en ial ac i i y.
Keywo ds: anammox; au o ophic ni ogen emo al; in eg a ed ixed ilm ac i a ed sludge;
mains eam; ni i a ion.
2
1. In oduc ion
The long- e m s abili y o he pa ial ni i a ion-anammox (PN/AMX) p ocesses ope a ing
a mains eam condi ions, cha ac e ized by low ammonium concen a ion (<100 mg N/L)
and low empe a u e (< 25 ºC), emains as a challenge. The main iden i ied bo lenecks
a e he achie emen o he s ingen discha ge limi s, he e ec i e supp ession o ni i e
oxidizing bac e ia (NOB) g ow h and he e ec i e anammox bac e ia e en ion inside he
eac o (Cao e al., 2017; Hoeks a e al., 2018). Di e en s a egies we e p oposed o
supp ess he NOB ac i i y (Ag awal e al., 2018; Cao e al., 2017; Ge e al., 2015). Among
hem, some au ho s sugges pe o ming he PN/AMX p ocesses in wo eac o s whe e
each p ocess is op imized sepa a ely (Dos a e al., 2015; Pé ez e al., 2015). Compa ed o
he one-s age op ion, he wo-s age PN/AMX sys em equi es mo e space o
implan a ion, a mo e complex con ol sys em, highe capi al cos s and i was epo ed o
p esumably lead o highe g eenhouse gases emissions (Lackne e al., 2014). The e o e,
he hyb id bio ilm-based echnologies and hei abili y o decouple he solid e en ion
ime (SRT) o di e en popula ions a ose as p omising app oaches o imp o e he p ocess
s abili y joining he ad an ages o bo h one-s age and wo-s age con igu a ions (Han e
al., 2016; Veuille e al., 2014).
Recen ly, hyb id sys ems like in eg a ed ixed ilm ac i a ed sludge (IFAS) eac o s
whe e sludge is a mix u e o suspended and bio ilm biomass, ha e been explo ed o
pe o m he PN/AMX p ocesses simul aneously (Veuille e al., 2014). The IFAS eac o s
ha e been ound o be mo e lexible, e icien and esis an agains bo h concen a ion
and hyd aulic shock loads han o he ac i a ed sludge p ocesses (Regmi e al., 2011). Due
o mass ans e limi a ions, anammox bac e ia p e e en ially g ow in he bio ilm while
he ammonium oxidizing bac e ia (AOB) and NOB a e mainly de eloped as suspended
biomass (Veuille e al., 2014). Thus, IFAS eac o p o ides he oppo uni y o expe ience
a mo e e sa ile ope a ion and po en ially ob ain a selec i ely NOB washou , while
anammox bac e ia emain in he sys em. Howe e , sca ce in o ma ion is a ailable
ega ding he PN/AMX p ocesses applied o he ea men o municipal was ewa e a
low empe a u es in IFAS eac o s (Table 1). Fo example, Yang e al. (2017) ope a ed a
plug- low IFAS sys em, a 24 - 26 °C, ea ing ae obically p e- ea ed was ewa e wi h a
o al ni ogen loading a e (NLR) o 125 mg N/(L·d). These au ho s achie ed an e luen
con aining o al ni ogen (TN) concen a ions lowe han 10 mg N/L (commonly
es ablished discha ge limi ) by applying a s a egy based on he esidual ammonium
3
concen a ion (3 - 5 mg N/L). Ne e heless, hese au ho s s a ed ha he long- e m p ocess
s abili y is s ill an issue equi ing he s ic con ol o he ai low a e o main ain he
mic oae obic condi ions (0.15 o 0.36 mg O2/L) (Yang e al., 2017). Fu he mo e, he use
o in e mi en ae a ion was widely used. Malo anyy e al. (2015) op imized he
in e mi en ae a ion s a egy (15 min ON/45 min OFF) ob aining a ni ogen emo al
e iciency (NRE) up o 70% o anae obically p e- ea ed was ewa e in a con inuous
IFAS sys em a 25 °C. T ojanowicz e al. (2016) un a simila sys em a 17 °C bu he
ob ained NRE was limi ed o 44%, mainly due o he ni a e p oduced by he NOB
ac i i y, p obably caused by he low ope a ional empe a u e. Lau eni e al. (2016) ea ed
low loaded was ewa e a 15 °C in a labo a o y eac o achie ing a NRE o 63 %.
Howe e , he applied mic oae obic condi ions o supp ess he NOB ac i i y limi ed he
ammonium oxida ion a e and he e o e he ni ogen emo al a e (NRR) was
app oxima ely o 23 g N/(m3·d).
Table 1. Compa ison o PN/AMX eac o s based on an IFAS sys em applied o ea municipal
was ewa e a mains eam condi ions.
Ope a ional condi ions
Sys em Pe o mance
T
(°C)
Volume
(L)
TN
(mg N/L)
sCOD/N
(g/g)
HRT
(h)
NRE
(%)
ANRE
(%)
NO3- p oduc ion
a io (%)
Re e ence
25
200
45
1.8
15
70
89
25
[1]
17
200
47
1.5
39
44
86
49
[2]
15
12
21
2.0
14
63
89
27
[3]
25
160
44
1.3
7-10
82
90
4
[4]
21 -15
200
43
2.5
19
72
81
11
This s udy
[1] Malo anyy e al. (2015); [2] T ojanowicz e al. (2016); [3] Lau eni e al. (2016); [4] Yang e al.
(2017); ANRE: ammonium ni ogen emo al e iciency; HRT: hyd aulic e en ion ime; NRE: ni ogen
emo al e iciency;.sCOD/N: soluble chemical oxygen demand o ni ogen a io; T: empe a u e; TN:
o al ni ogen.
Hence, he main objec i e o his s udy is o e alua e he pe o mance and s abili y o he
PN/AMX p ocesses aking place in a pilo -scale IFAS sys em ope a ed a mains eam
condi ions. As no el y, he eac o was ope a ed a low empe a u es (21 - 15 °C) wi h he
added challenge o ea ing anae obically p e- ea ed was ewa e wi h a soluble chemical
oxygen demand (sCOD) o ammonium ni ogen a io o 2.5 ± 0.3 g sCOD/g N.
Fu he mo e, a de ailed s udy o he mic obial popula ion seg ega ion be ween suspended
and bio ilm ac ions is p esen ed o unde s and he easons o he NOB ac i i y
limi a ion inside he eac o .
4
2. Ma e ial and Me hods
2.1. IFAS pilo plan se up
An IFAS pilo scale eac o wi h a wo king olume o 200 L was used. I was illed wi h
AnoxKaldnes K1 bio ilm ca ie s, 40% olume, (see Figu e S1 in Supplemen a y
Ma e ial) cha ac e ized by an e ec i e su ace a ea o 500 m2/m3ca ie . The suspended
sludge, lea ing he uni , se led in a sedimen a ion ank (100 L) om which he
concen a ed sludge was eci cula ed back o he eac o a a eci cula ion a io o 1 (100%
o he low a e a which was ewa e was ed o he sys em, 260 L/d). The IFAS eac o
was ope a ed as a con inuous s i ed- ank wi h in e mi en ae a ion (al e na ion be ween
ae obic and anoxic pe iods). The dissol ed oxygen (DO) concen a ion inside he eac o ,
du ing he ae a ed phases, was ixed a a se poin o 1.5 mg O2/L, al hough his se poin
was mo e di icul o main ain as he empe a u e d opped. This se poin was egula ed
by a PID (p opo ional-in eg al-de i a i e) con olle (Ce lic AB, Sweden), which
adjus ed he ai low a e. The pH was no con olled and was o 6.95 ± 0.20. The p ocess
pe o mance was moni o ed by online measu emen s o empe a u e, DO concen a ion,
pH, suspended solids (SS) concen a ion, oxida ion- educ ion po en ial (ORP),
ammonium ni ogen and ni a e ni ogen concen a ions inside he eac o . Addi ionally,
he eac o was also equipped wi h conduc i i y senso s o bo h in luen and mixed liquo .
All da a om he on-line senso s we e eco ded and collec ed in a da a acquisi ion sys em.
2.2. IFAS ope a ional condi ions
The IFAS eac o has been ope a ed o yea s be o e he beginning o he p esen esea ch
s udy. Du ing ha pe iod, i ea ed low s eng h ni ogen s eams a empe a u es o 21 -
25 °C (Malo anyy e al., 2015). The eac o was ope a ed o 100 days di ided in o h ee
di e en s ages as he empe a u e was s ep-wise dec eased om 21 ºC (S age I, days 0 -
33) o 18 °C (S age II, days 34 - 79) and inally o 15 °C in S age III (days 80 - 100). The
empe a u e was con inuously moni o ed and egula ed by bo h imme se coole and
hea e connec ed o a he mos a (Julabo AB, Sweden). Mo eo e , he in e mi en
ae a ion pa e n was changed om 15 min ON/45 min OFF pe iods o 20 min ON/40 min
OFF om day 38 onwa ds.
The IFAS eac o ope a ed a he Hamma by Sjös ads e k esea ch acili y (S ockholm,
Sweden). This acili y is loca ed di ec ly on op o he Hen iksdal WWTP and p o ided
wi h a di ec discha ge poin o municipal was ewa e , which was used as eeding. The
5
municipal was ewa e was p e- ea ed in a p ima y se le and an up low anae obic sludge
blanke (UASB) eac o (6.3 m3) ope a ed a 20 ºC o emo e he o ganic ma e
(Malo anyy e al., 2015). Then, a ac ion o he UASB e luen was il a ed h ough a
20 µm po e size il e and s o ed in an equaliza ion ank (2 m3) o educe he incoming
solid concen a ion and o mi iga e he was ewa e composi ion luc ua ions. A e wa ds,
i was con inuously ed o he IFAS uni wi h he ollowing composi ion: 38 - 48 mg
NH4+-N/L, 110 - 138 mg soluble COD/L, 235 – 283 mg CaCO3/L, 5 – 22 mg VSS/L and
a e age pH alues o 7.5 ± 0.1.
The hyd aulic e en ion ime (HRT) in he IFAS eac o was main ained cons an a 19 h
esul ing in an a e age NLR o 56 ± 3 g N/(m3·d), a ying only due o he ni ogen
concen a ion luc ua ions.
2.3. Analy ical me hods
To moni o he pilo plan pe o mance and o check and calib a e he online senso s, bo h
in luen and e luen we e pe iodically sampled and il e ed using 0.45 µm po e size
il e s. Then, he concen a ions o chemical oxygen demand (COD), o al ni ogen (TN),
ammonium, ni i e, ni a e and alkalini y we e measu ed spec opho ome ically wi h D
Lange es ki s (Hach Lange, Ge many) in a Pho olab® 6600 UV-Vis. Po assium and
chlo ine ion concen a ions we e also spec opho ome ically de e mined (wi h WTW
Spec oquan in a D Lange ION 500) o calib a e he ammonium and ni a e online
senso s, espec i ely. Concen a ions o o al suspended solids (TSS) and ola ile
suspended solids (VSS) we e de e mined acco ding o he S anda ds Me hods (Ame ican
Public Heal h Associa ion e al., 2017) in he in luen , e luen and he mixed liquo om
he bio eac o . The biomass concen a ion, TSS and VSS, co esponding o he bio ilm
was de e mined a e he mechanical de achmen o he biomass (Figu e S1 in
Supplemen a y Ma e ial) om a andomly selec ed known numbe o ca ie s.
2.4. Mic obial ac i i y ba ch es s
To ollow up he mic obial p ocess pe o mance inside he eac o , speci ic ac i i y (SA)
es s we e ca ied ou on bo h bio ilm and suspended biomass samples aken ou om he
IFAS eac o . The maximum speci ic anammox ac i i y (SAAMX), in mg N2-N/(g VSS·d),
was de e mined acco ding o Dapena-Mo a e al. (2007). A modi ica ion o he SAA
me hod was also used o de e mine he ac i i y o he e o ophic deni i ying (SAHD)
bac e ia adding as subs a e ni a e (50 mg N/L) and ace a e (100 mg COD/L). Ac i i ies
6
o ae obic he e o ophs (SAAe H), AOB (SAAOB) and NOB (SANOB) we e de e mined
acco ding o he me hod desc ibed by Su macz-Go ska e al. (1996). All hese ba ch
assays we e pe o med a he co esponding IFAS eac o ope a ional empe a u e a he
momen o he biomass collec ion. Mo eo e , SAA was also de e mined a 30 ºC,
conside ed as e e ence empe a u e o anammox bac e ia.
2.5. Iden i ica ion o mic obial popula ions
The main ac i e bac e ial popula ions p esen in he suspended sludge and bio ilm
biomass samples we e iden i ied in samples collec ed om he IFAS eac o on day 75
(S age II). The luo escence in si u hyb idiza ion (FISH) molecula echnique was applied
ollowing he p o ocol desc ibed by Amann e al. (1990). Biomass was manually de ached
om he su ace o K1 ca ie s o a oid bac e ial deac i a ion. The speci ic
oligonucleo ide p obes used (Table S1 in Supplemen a y Ma e ial) we e 5’-labelled wi h
he luo och omes FITC ( luo escei-5-isocyana e) o Cy3 (Ca bocyanine 3). De ails o
he oligonucleo ide p obes a e a ailable a p obeBase (G eu e e al., 2016). DAPI (4,6-
diamindino-2-phenylindole) was used as uni e sal dye o all DNA. Fluo escence signals
we e eco ded wi h an acquisi ion sys em (Coolsnap, Rope Scien ic Pho ome ics)
coupled o an Axioskop 2 Plus epi luo escence mic oscope (Zeiss, Ge many). The
ela i e abundances o AOB, NOB and anammox bac e ia we e es ima ed by semi-
quan i a i e coun ing o he a io be ween hei speci ic mic obial popula ions bio olume
and he o al bac e ial bio olume using he DAIME so wa e (Daims e al., 2006).
2.6. Calcula ions
S a is ical di e ences be ween he esul s ob ained in he h ee ope a ional s ages we e
es ed by one- ac o analysis o a iance (ANOVA) using he s a is ical so wa e R
e sion 3.5.2 (The R Founda ion S a is ical Compu ing). P io o ANOVA, a iance
homogenei y was con i med by Le ene’s es and no mal dis ibu ion by he Shaphi o’s
es . Then, i he ANOVA con i med he di e ence be ween mean alues, a pos hoc
analysis (Tukey’s HSD) was applied o de e mine be ween which alues he di e ence
was signi ican , conside ing a le el o signi icance o 0.05. I da a a iance homogenei y
and/o no mal dis ibu ion was no me , he non-pa ame ic K uskal-Wallis analysis was
applied and a e wa ds he Wilcoxon pos hoc one.
7
3. Resul s and discussion
3.1. Ni ogen emo al in he IFAS PN/AMX eac o a dec easing empe a u es
The ope a ional s a egy imposed o he IFAS PN/AMX pilo plan , con inuous eeding
and in e mi en ae a ion cycles, was de ined o p omo e he pe o mance o anoxic
p ocesses and exploi he lag phase o NOB bac e ia ac i i y a e oxygen s a a ion. The
NOB supp ession is c ucial du ing he ope a ion a he low empe a u es o be es ed o
21, 18 and 15 °C.
Ini ially (S age I), he daily a e age DO concen a ion was o 0.45 mg O2/L (Table 2) and
he PN/AMX p ocesses pe o mance was s able. The concen a ion o TN in he e luen
a he end o his s age was lowe han 10 mg N/L co esponding o 4.0 ± 1.5 mg NH4+-
N/L, 3.0 ± 2.2 mg NO3--N/L and less han 0.2 mg NO2--N/L (Figu e 1.A). Ammonium
ni ogen emo al e iciency (ANRE) and NRE eached alues o 80 ± 13 % and 73 ±
12%, espec i ely (Figu e 1.B and Table 2). The obse ed ni a e p oduc ion a io was
only 9 ± 5 %, which indica ed a negligible NOB ac i i y inside he eac o (Table 2).
Then, he NRE inc eased eaching a e age alues o 85 % in he las 10 days o his
pe iod.
Table 2. A e age alues o he eac o pe o mance a dec easing empe a u es.
S-I (21 ºC)
S-II (18 ºC)
S-III (15 ºC)
Phase leng h (days)
0 -33
34 - 79
80 -100
Daily DO (mg/L)
0.45 ± 0.05
0.58 ± 0.19
0.64 ± 0.07
pH
7.05 ± 0.04
6.83 ± 0.13
7.09 ± 0.23
NRR (g N/(m3·d))
40.04 ± 5.05
42.47 ± 7.80
37.35 ± 3.32
NRE (%)
72.5 ± 12.7
74.1 ± 11.4
65.8 ± 4.9
ANRE (%)
80.0 ± 13.4
84.4 ± 11.4
77.5 ± 10.5
(NO3-)p oduced/(NH4+)consumed (%)
9.4 ± 5.7
11.3 ± 9.1
12.5 ± 9.1
NO2--Ne (mg N/L)
0.10 ± 0.07
0.29 ± 0.13
0.54 ± 0.23
TNe (mg N/L)
10.9 ± 5.6
10.8 ± 5.3
15.0 ± 2.3
CODe (mg/L)
32.8 ± 0.7
44.5 ± 3.4
46.0 ± 4.9
VSS (g/L)
1.03 ± 0.12
1.05 ± 0.15
1.03 ± 0.09
ANRE: ammonium ni ogen emo al e iciency; COD: chemical oxygen demand; NRE: ni ogen emo al
e iciency; NRR: ni ogen emo al a e; TN: o al ni ogen; VSS: ola ile suspended solids. Sub-indexes
“e” and “ ” e e o concen a ion measu ed in he e luen and inside he eac o s, espec i ely.
8
On day 34 he empe a u e was dec eased o 18 °C (S age II) esul ing in an immedia e
dec ease o he NRE due o ammonium accumula ion up o 13 mg NH4+-N/L, whe eas
ni a e concen a ion emained simila o he alues a he end o S age I (< 1 mg NO3--
N/L) (Figu e 1A). The ela i ely sho oxygen supply pe iods and he low applied DO
concen a ion in hese pe iods (1.5 mg O2/L), esul ed in daily a e age DO concen a ions
o 0.49 mg O2/L. Thus, on day 39 he in e mi en ae a ion leng h was inc eased om 15
min o 20 min and he anoxic phase educed om 45 o 40 min, o diminish he AOB
ac i i y limi a ions. Then, he a e age daily DO concen a ion inc eased o 0.58 mg O2/L
(Table 2). Consequen ly, he ANRE inc eased and he NRE s abilized a a e age alues
o 74 %. Thus, he di e ences o he espec i e NRE and ANRE alues co esponding o
S age I (21 ºC) and S age II (18 ºC) we e s a is ically insigni ican acco ding o he esul s
om ANOVA (p=0.925) and K uskal-Wallis (p=0.485) analysis, espec i ely. Thus, no
e ec can be a ibu ed o he empe a u e educ ion, om 21 o 18 °C, o e he
pe o mance o he PN/AMX p ocesses.
A)
B)
Figu e 1. IFAS eac o pe o mance: A) E olu ion o he ammonium concen a ion in he in luen (●), and
e luen concen a ions o ammonium (○), ni i e (■) and ni a e (▲) in mg N/L. B) E olu ion o ammonium
ni ogen emo al e iciency (ANRE, ●) and ni ogen emo al e iciency (NRE, ○) in %.
0
5
10
15
20
25
30
35
40
45
50
010 20 30 40 50 60 70 80 90 100
Ni ogen (mg N/L)
Time (days)
S-I S-II S-III
0
10
20
30
40
50
60
70
80
90
100
010 20 30 40 50 60 70 80 90 100
Remo al e iciency (%)
Time (days)
S-I S-II S-III
15
a 21 ºC (Figu e 3.A). Thus, i can be s a ed ha ni ogen emo al in he IFAS eac o was
mos ly due o SAAMX.
The SAae HET was p esen h oughou he ope a ional pe iod a a e age alues o 100 ± 8
mg O2/(g VSS·d) and 20 ± 3 mg O2/(g VSS·d) o he suspended and bio ilm biomass,
espec i ely. Mo eo e , he SAAOB and SANOB in suspended biomass we e also highe
han in he bio ilm. In he suspended sludge, he SAAOB was highe han SANOB whe eas
he la e was highe in he bio ilm biomass (Figu e 3.B). In bo h biomass ac ions, he
SANOB was he mos a ec ed one by he ope a ional empe a u e dec ease wi h an
es ima ed descen a e o 6 mg N/(g VSS·d·ºC), which doubled ha o he SAAOB alue
(Figu e 3.B). This obse a ion con adic s he esul s om o he s udies epo ing ha
NOB supp ession is e en mo e challenging a low empe a u e (Cao e al., 2017). The
NOB a e supposed o be less empe a u e sensi i e han AOB and anammox (due o hei
lowe ac i a ion ene gy) (T ojanowicz e al., 2016). Howe e , in he p esen s udy, he
dec ease in he SANOB canno be a ibu ed only o he empe a u e descen , since he
biomass used o he assays was collec ed a di e en ope a ional days and he mic obial
popula ion dis ibu ion migh e ol e due o he di e en SRT and lowe biomass g ow h
a es due o he diminishing o empe a u e.
Conside ing he sepa a ed concen a ions o he suspended and bio ilm biomass ac ions
and he co esponding measu ed speci ic ac i i ies, he maximum achie able con e sion
capaci ies (as g N/(m3·d)) o each mic obial popula ion we e calcula ed (Figu e 4). As he
ac ion o biomass a ached o he ca ie su ace is much highe han he suspended
sludge one, he bio ilm is he main con ibu o o he TN emo al capaci y. Conside ing
he whole biomass ac ions inside he sys em, he o al AOB capaci y was 1.8 - 2.0 imes
he anammox capaci y (Figu e 4.B), which is he desi able si ua ion o each high NRE.
Addi ionally, he a io o AOB/NOB conside ing bo h biomass ac ions p og essi ely
inc eased om 0.7 in S-I (21 °C) o 1.1 in S-III indica ing he p omo ion o he AOB
while he NOB capaci y dec eased a 53% (Figu e 4). Indeed, a 15 ºC he AOB was he
popula ion wi h he highes o al capaci y wi h a e age alues o 167 g N/(m3·d) while
he NOB one was o 145 g N/(m3·d). The limi a ion o NOB ac i i y was also obse ed
du ing he eac o ope a ion since he measu ed NOB ac i i y amoun ed o 10 - 20 % o
he maximum po en ial capaci y ob ained om he ba ch es s. The much lowe
he e o ophic deni i ica ion capaci y, compa ed wi h he anammox one ( a ios lowe han
0.05 in S ages II and III, Figu e 4.B), co obo a ed ha he main ou e o ni ogen emo al
16
in he IFAS sys em was he au o ophic pa hway. Mo eo e , a much lowe con ibu ion
o he he e o ophic deni i ica ion capaci y in NRE is expec ed as he eac o is no
comple ely anoxic and COD supplied was limi ed (Table 2).
A)
B)
Figu e 4. A) Maximum con e sion capaci y, measu ed in ba ch expe imen s, o he di e en bac e ial
popula ions p esen in suspended sludge ( locs) and bio ilm biomass: anammox (AMX), ammonium
oxidizing bac e ia (AOB), ni i e oxidizing bac e ia (NOB) and he e o ophic deni i ying (HD). B)
E olu ion o he maximum con e sion capaci y a ios o he di e en bac e ial popula ions a dec easing
empe a u es 21 °C (■), 18 °C (□) and 15 °C (■).
The mic obial popula ion seg ega ion sugges ed by he di e en speci ic mic obial
ac i i ies (Figu e 3 and 4) de ec ed in each biomass ac ion was also con i med by he
mic obial cha ac e iza ion analysis o a sample collec ed in S age II (Table S2 and Figu e
S2 in Suppo ing Ma e ial). Anammox bac e ia ep esen ed app oxima ely 33% o he
ac i e bac e ial communi y in he bio ilm whe eas hey only amoun ed o 12% o he
suspended sludge ac ion. Candida us B ocadia Anammoxidans was he p edominan
anammox popula ion de ec ed ( ep esen ing 90% o he o al ac i e anammox). In he
case o he suspended sludge ac ion, AOB ela i e abundance was signi ican ly highe
(42.9 ± 6.1%) han in he bio ilm biomass (16.9 ± 4.0%). NOB we e de ec ed in bo h
bio ilm and suspended biomass. Howe e , di e ences among he dominan specie we e
obse ed. Ni ospi a was he p edominan NOB in he suspended sludge a a ela i e
0
50
100
150
200
250
S - I (21 °C) S - II (18 °C) S - III (15 °C)
Maximum capaci y (g N/(m3·d))
AMX-Bio ilm
AMX- locs
AOB-Bio ilm
AOB-Flocs
NOB-Bio ilm
NOB- locs
HD-Bio ilm
HD-Flocs
0.0
0.5
1.0
1.5
2.0
2.5
3.0
AOB/AMX AOB/NOB HD/AMX AMX/NOB
Maximum capaci y a ios
17
abundance o 11.4% whe eas in he case o he bio ilm bo h Ni ospi a and Ni obac e
spp. p esen ed simila abundances (Table S2 in Supplemen a y Ma e ial). Ni ospi a is
he mos commonly ound specie in eac o sys ems ope a ed a mains eam condi ions
as hese bac e ia p esen high a ini y o subs a es being able o su i e in ni i e limi ed
en i onmen s (Cao e al., 2017). Ni o oga a c ica was also de ec ed in bo h ac ions bu
in low pe cen ages.
3.5. O e coming AOB ac i i y limi a ion
Al hough he ob ained AOB/AMX maximum con e sion capaci y a io was almos 2
(Figu e 4.B), indica ing ha enough ni i e could be supplied, he ammonium oxida ion
was iden i ied as one o he main limi ing s eps du ing he eac o ope a ion.
AOB ac i i y in he bio ilm migh be es ic ed by DO ans e limi a ion and e en i i
was high in he locculen sludge he con ibu ion o his ac ion o he global capaci y
could be insu icien o oxidize enough ammonium o ni i e o he anammox p ocess.
Mo eo e , he leng h o he ae a ion phase o 15 o 20 min migh be also oo sho o
each his equi ed ammonium oxida ion bu inc easing he leng h o he ae a ed phase
migh also p omo e he de elopmen o NOB ac i i y, which was no desi ed. Only du ing
S age III, he obse ed accumula ion o low ni i e concen a ions (Table 2) indica es ha
AOB was no limi ing he con e sion capaci y anymo e. This ac is explained by bo h
he AOB g ow h and he NOB supp ession, meaning ha mo e ni i e is a ailable o he
anammox bac e ia.
Ano he possible ac ion would be he op imiza ion o he p e ious UASB eac o
pe o mance o dec ease he COD concen a ion o he in luen o he IFAS sys em
es ic ing he ae obic he e o ophic bac e ia compe i ion wi h AOB by oxygen.
In addi ion, in he case o PN/AMX IFAS sys ems, AOB g ow mainly as suspended
biomass. In he p esen s udy, he SAAOB in his biomass ac ion was 4 imes highe han
ha o he bio ilm (Figu e 3.B). The e o e, i suspended biomass concen a ion augmen s
he ammonium oxida ion capaci y o he sys em would inc ease oo, imp o ing he
ni ogen emo al o he sys em. Howe e , o inc ease his biomass concen a ion was
di icul , since i p esen ed poo se ling p ope ies and ended o loa in he
sedimen a ion ank p obably due o he accumula ed ni ogen gas p oduced, which was
no app op ia ely eleased.
18
3.6. Biomass seg ega ion and NOB supp ession in he PN/AMX p ocesses
To imp o e he PN/AMX p ocesses s abili y and e luen quali y o supp ess he NOB
ac i i y, wi hou a ec ing he anammox bac e ia is a equisi e. Thus, he IFAS sys ems
a ose as a p omising al e na i e o implemen ing he PN/AMX p ocesses in a one-s age
uni since ni i ying bac e ia g ow mainly in he suspended sludge while anammox
bac e ia we e ound in he bio ilm (Han e al., 2016; Veuille e al., 2014).
In he p esen s udy, he comple e supp ession o NOB ac i i y was di icul , since NOB
we e also well in eg a ed in o he bio ilm (Figu es 3, 4 and Table S2 in Suppo ing
Ma e ial). As NOB a e ae obic, hey should be in he ou e laye s o he bio ilm.
The e o e, hey migh be washed-ou by inc easing he shea s ess imposed in he sys em
o imp o e he de achmen o he loose pa s o he bio ilm su ace and adjus ing hen he
SRT o he locculen biomass. Al hough SRT was no con olled in he p esen s udy, he
NOB capaci y diminished in app oxima ely he 50% o bo h suspended sludge and
bio ilm om S age I o S age III (Figu e 4).
As an addi ional conside a ion, i was gene ally accep ed ha AOB ha e oxygen a ini ies
highe han NOB and hus PN/AMX eac o s should be ope a ed a low DO concen a ion
(Lackne e al., 2014). Howe e , a low empe a u es NOB ha e oxygen a ini ies highe
han AOB and he use o a s a egy based on jus he DO concen a ion con ol was no
sui able o NOB supp ession (Regmi e al., 2014; Val del Rio e al., 2019). Al hough
in e mi en ae a ion was widely applied o c ea e ansien anoxic condi ions, which help
he NOB ou -selec ion (Regmi e al., 2014), he induced lag phase ailed o comple ely
supp ess he NOB ac i i y a low empe a u e (Ag awal e al., 2018; Cao e al., 2017). In
he p esen s udy, he NOB ac i i y inside he eac o was limi ed o 20 % o he SANOB.
In his sense, Malo anyy e al. (2015) e alua ed he e ec o di e en a ios be ween he
leng hs o ae a ed and non-ae a ed phases o supp ess he NOB ac i i y. The bes
PN/AMX p ocesses pe o mance achie ed in hei sys em (a 25 ºC) was wi h an
in e mi en ae a ion pa e n o 15 min ON and 45 min OFF, a DO se poin o 1 mg O2/L
and a sCOD/N a io equal o 1.8 g/g, achie ing a NRR o 55 g N/(m3·d) wi h a NRE o
70 %. Howe e , he e ec i eness o he in e mi en ae a ion s a egy also depends on he
eac o ype and ope a ional condi ions. Fo example, in he p esen s udy i was
demons a ed ha a lowe empe a u es (18 and 15 °C) he ae a ion pa e n o 20 min
ON and 40 min OFF was mo e app op ia e.
19
O he s a egies o NOB supp ession ha e been p oposed like he ope a ion o he sys em
a high ee ammonium (FA) concen a ions ha lead o a good NOB supp ession since
he AOB ha e no subs a e limi a ion and NOB a e inhibi ed a lowe FA concen a ions
han AOB (Blackbu ne e al., 2007; Vadi elu e al., 2007). The use o ee ni ous acid
(FNA) o inhibi NOB as i is subs an ially mo e ha m ul o NOB han o AOB was also
p oposed (Blackbu ne e al., 2007; Vadi elu e al., 2007). Howe e , wi h he pH alue o
he eac o media ( anging om 6.64 o 7.52) nei he FA (0.1 mg NH3-N/L) no FNA
(0.02 mg HNO2-N/L) inhibi o y concen a ions we e achie ed. A e age alues we e o
0.03 ± 0.02 mg NH3-N/L and 0.10 ± 0.08 µg HNO2-N/L, espec i ely. Inhibi o y FA
concen a ions would be only possible i la ge ammonium concen a ions we e p esen
inside he eac o diminishing he NRE. To ob ain inhibi o y FNA concen a ions in a
one-s age sys em would be di icul as anammox bac e ia consume ni i e and hey a e
mo e sensi i e o his compound han NOB. This p oblem migh be a oided i he eac o
was a SBR o a plug low eac o whe e g adien concen a ions a e easily achie ed.
To main ain he s able PN/AMX p ocesses s abili y a low empe a u e is especially
di icul due o he unbalance o he AOB and anammox bac e ia ac i i ies as well as he
challenging e ec i e supp ession o NOB. Fo hese easons, in IFAS sys ems o be
compe i i e wi h he con en ional biological ni ogen emo al sys ems, he main aining
o he AOB ac i i y a long- e m ope a ion while NOB one is supp essed needs o be
u he add essed as well as he maximum NRR achie able.
In addi ion, o ope a e he PN/AMX p ocesses s able a mains eam condi ions hey need
o cope wi h he high was ewa e luc ua ions in e ms o empe a u e, low a es and
composi ion. Thus, he de ini ion o a obus and good con ol s a egy would help o ace
his challenge. Besides, his is one o he main ields o esea ch which lack o in o ma ion
as he long- e m s able PN/AMX s udies, a mains eam condi ions, we e pe o med
mos ly a labo a o y scale (Cao e al. 2017).
4. Conclusions
The easibili y o ope a ing a one-s age mains eam PN/AMX p ocesses in an IFAS
eac o con igu a ion was p o ed a pilo scale. Anae obically p e- ea ed municipal
was ewa e was ea ed a dec easing empe a u es (21 - 15 °C) achie ing an a e age NRE
o 72 ± 11%. A 15 °C, he p ocess s abili y was main ained eaching a e age NRR o 37
± 3 g N/(m3·d), compa able wi h ha obse ed in con en ional nu ien emo al sys ems
20
ope a ed a mains eam condi ions. Du ing he IFAS ope a ion NOB we e p esen bu
hei ac i i y was success ully supp essed obse ing less han he 20 % and 10 % o he
maximum po en ial NOB ac i i y inside he eac o a 21 and 15 °C, espec i ely.
Mic obial popula ion seg ega ion was obse ed being he AOB and NOB mo e abundan
in he locculen ac ion while he bio ilm was mos ly composed by he anammox
bac e ia. This ac could acili a e he selec ion o hose popula ions mainly p esen as
bio ilm and he selec i e wash ou o he locculen biomass helping o supp ess he NOB
ac i i y.
Decla a ions o in e es : none
Acknowledgmen s
This wo k was done wi hin he Pionee _STP (ID 199 (UE)/PCIN-2015-022 (AEI)/FR-
2016/0002(Fo mas)) p ojec unded by he Wa e Wo ks2014 Co unded Call (Wa e JPI/Ho izon
2020). A. Ped ouso also wan o hank he suppo by a STSM G an om COST Ac ion
Wa e _2020 (COST-STSM-ES1202-010216-076068) o he esea ch s ay a KTH in Sweden.
The au ho s om he USC belong o CRETUS S a egic Pa ne ship (ED418B 2017/075) and o
he Galician Compe i i e Resea ch G oup (GRC-ED431C 2017/29). All hese p og ams a e co-
unded by FEDER (UE) unds. The au ho s om KTH belong o VA Mäla dalen Clus e unded
by Swedish Wa e De elopmen (SVU). The au ho s would like o acknowledge he s a a
Hamma by Sjös ads e k, S ockholm (Swedish Wa e Inno a ion Cen e ) and he Swedish
En i onmen al Resea ch Ins i u e (IVL) o hei suppo .
Re e ences
Ag awal, S., Seun jens, D., Cocke , P.D., Lackne , S., Vlaeminck, S.E., 2018. Success o
mains eam pa ial ni i a ion/anammox demands in eg a ion o enginee ing, mic obiome
and modeling insigh s. Cu . Opin. Bio ech. 50, 214-221. doi:
10.1016/j.copbio.2018.01.013.
Amann, R.I., K umholz, L., S ahl, D.A., 1990. Fluo escen -oligonucleo ide p obing o
whole cells o de e mina i e, phylogene ic, and en i onmen al s udies in mic obiology.
J. Bac e iol. 172, 762-770. doi: 10.1128/jb.172.2.762-770.1990.
Ame ican Public Heal h Associa ion, Ame ican Wa e Wo ks Associa ion, Wa e
En i onmen Fede a ion, 2017. S anda d me hods o he examina ion o wa e and
was ewa e , 23 d edi ion. ed. Ame ican Public Heal h Associa ion, Washing on, DC.
21
Blackbu ne, R., Vadi elu, V.M., Yuan, Z., Kelle , J., 2007. Kine ic cha ac e isa ion o an
en iched Ni ospi a cul u e wi h compa ison o Ni obac e . Wa e Res. 41, 3033-3042.
doi: 10.1016/j.wa es.2007.01.043.
Bu on, F.L., Tchobanoglous, G., Tsuchihashi, R., S ensel, H.D., Me cal & Eddy, I.,
2014. Was ewa e enginee ing : ea men and esou ce eco e y. McG aw-Hill
Educa ion.
Cao, Y., an Loosd ech , M.C., Daigge , G.T., 2017. Mains eam pa ial ni i a ion-
anammox in municipal was ewa e ea men : s a us, bo lenecks, and u he s udies.
Appl. Mic obiol. Bio echnol. 101, 1365-1383. doi: 10.1007/s00253-016-8058-7.
Daims, H., Lücke , S., Wagne , M., 2006. Daime, a no el image analysis p og am o
mic obial ecology and bio ilm esea ch. En i on. Mic obiol. 8, 200-213. doi:
10.1111/j.1462-2920.2005.00880.x.
Dapena-Mo a, A., Fe nández, I., Campos, J.L., Mosque a-Co al, A., Méndez, R., Je en,
M.S.M., 2007. E alua ion o ac i i y and inhibi ion e ec s on Anammox p ocess by ba ch
es s based on he ni ogen gas p oduc ion. Enzyme and Mic ob. Technol. 40, 859-865.
doi: 10.1016/j.enzmic ec.2006.06.018.
Dos a, J., Vila, J., Sancho, I., Basse , N., G i oll, M., Ma a-Ál a ez, J., 2015. Two-s ep
pa ial ni i a ion/Anammox p ocess in g anula ion eac o s: S a -up ope a ion and
mic obial cha ac e iza ion. J. En i on. Manag. 164, 196-205. doi:
10.1016/j.jen man.2015.08.023.
Ge, S., Wang, S., Yang, X., Qiu, S., Li, B., Peng, Y., 2015. De ec ion o ni i ie s and
e alua ion o pa ial ni i ica ion o was ewa e ea men : A e iew. Chemosphe e 140,
85-98. doi: 10.1016/j.chemosphe e.2015.02.004.
G eu e , D., Loy, A., Ho n, M., Ra ei, T., 2016. p obeBase—an online esou ce o
RNA- a ge ed oligonucleo ide p obes and p ime s: new ea u es 2016. Nucleic Acids
Resea ch 44, D586-D589. doi: 10.1093/na /gk 1232.
Han, M., Vlaeminck, S.E., Al-Oma i, A., We , B., Bo , C., Mu hy, S., De Clippelei ,
H., 2016. Uncoupling he solids e en ion imes o locs and g anules in mains eam
deammoni ica ion: A sc een as e ec i e ou -selec ion ool o ni i e oxidizing bac e ia.
Bio esou . Technol. 221, 195-204. doi: 10.1016/j.bio ech.2016.08.115.
Hoeks a, M., Geil oe , S.P., Hend ickx, T.L.G., an E p Taalman Kip, C.S.,
Klee ebezem, R., an Loosd ech , M.C.M., 2018. Towa ds mains eam anammox:
lessons lea ned om pilo -scale esea ch a WWTP Dokha en. En i on Technol, 1-13.
doi: 10.1080/09593330.2018.1470204.
22
Lackne , S., Gilbe , E.M., Vlaeminck, S.E., Joss, A., Ho n, H., an Loosd ech , M.C.M.,
2014. Full-scale pa ial ni i a ion/anammox expe iences – An applica ion su ey. Wa e
Res. 55, 292-303. doi: 10.1016/j.wa es.2014.02.032.
Lau eni, M., Falås, P., Robin, O., Wick, A., Weissb od , D.G., Nielsen, J.L., Te nes, T.A.,
Mo gen o h, E., Joss, A., 2016. Mains eam pa ial ni i a ion and anammox: long- e m
p ocess s abili y and e luen quali y a low empe a u es. Wa e Res. 101, 628-639. doi:
10.1016/j.wa es.2016.05.005.
Lo i, T., Klee ebezem, R., Hu, Z., Ka al, B., de K euk, M.K., an E p Taalman Kip, C.,
K ui , J., Hend ickx, T.L.G., an Loosd ech , M.C.M., 2015. Pilo -scale e alua ion o
anammox-based mains eam ni ogen emo al om municipal was ewa e . En i on.
Technol. 36, 1167-1177. doi: 10.1080/09593330.2014.982722.
Malo anyy, A., T ela, J., Plaza, E., 2015. Mains eam was ewa e ea men in in eg a ed
ixed ilm ac i a ed sludge (IFAS) eac o by pa ial ni i a ion/anammox p ocess.
Bio esou . Technol. 198, 478-487. doi: 10.1016/j.bio ech.2015.08.123.
Ped ouso, A., Aia za, I., Mo ales, N., Vázquez-Padín, J.R., Rogalla, F., Campos, J.L.,
Mosque a-Co al, A., Val del Rio, A., 2018. Pilo -scale ELAN® p ocess applied o ea
p ima y se led u ban was ewa e a low empe a u e ia pa ial ni i a ion-anammox
p ocesses. Sep. Pu i . Technol. 200, 94-101. doi: 10.1016/j.seppu .2018.02.017.
Pé ez, J., Isan a, E., Ca e a, J., 2015. Would a wo-s age N- emo al be a sui able
echnology o implemen a ull scale he use o anammox o sewage ea men ? Wa e
Sci. Technol. 72, 858-864. doi: 10.2166/ws .2015.281.
Regmi, P., Mille , M.W., Holga e, B., Bunce, R., Pa k, H., Chand an, K., We , B.,
Mu hy, S., Bo , C.B., 2014. Con ol o ae a ion, ae obic SRT and COD inpu o
mains eam ni i a ion/deni i a ion. Wa e Res. 57, 162-171. doi:
10.1016/j.wa es.2014.03.035.
Regmi, P., Thomas, W., Scha an, G., Bo , C., Ru he o d, B., Wal ip, D., 2011.
Ni ogen emo al assessmen h ough ni i ica ion a es and media bio ilm accumula ion
in an IFAS p ocess demons a ion s udy. Wa e Res. 45, 6699-6708. doi:
10.1016/j.wa es.2011.10.009.
Su macz-Go ska, J., Ge naey, K., Demuynck, C., Van olleghem, P., Ve s ae e, W.,
1996. Ni i ica ion moni o ing in ac i a ed sludge by oxygen up ake a e (OUR)
measu emen s. Wa e Res. 30, 1228-1236. doi: 10.1016/0043-1354(95)00280-4.
23
T ojanowicz, K., Plaza, E., T ela, J., 2016. Pilo scale s udies on ni i a ion-anammox
p ocess o mains eam was ewa e a low empe a u e. Wa e Sci. Technol. 73, 761-768.
doi: 10.2166/ws .2015.551.
Vadi elu, V.M., Kelle , J., Yuan, Z., 2007. F ee ammonia and ee ni ous acid inhibi ion
on he anabolic and ca abolic p ocesses o Ni osomonas and Ni obac e , Wa e Sci.
Technol. 89-97.
Val del Rio, A., Campos, J.L., Da Sil a, C., Ped ouso, A., Mosque a-Co al, A., 2019.
De e mina ion o he in insic kine ic pa ame e s o ammonia-oxidizing and ni i e-
oxidizing bac e ia in g anula and locculen sludge. Sep. Pu i . Technol. 213, 571-577.
doi: 10.1016/j.seppu .2018.12.048.
Veuille , F., Lac oix, S., Bausse on, A., Gonidec, E., Ochoa, J., Ch is ensson, M.,
Lemai e, R., 2014. In eg a ed ixed- ilm ac i a ed sludge ANITA™Mox p ocess - A new
pe spec i e o ad anced ni ogen emo al. Wa e Sci. Technol. 69, 915-922. doi:
10.2166/ws .2013.786.
Yang, Y., Zhang, L., Cheng, J., Zhang, S., Li, B., Peng, Y., 2017. Achie e e icien
ni ogen emo al om eal sewage in a plug- low in eg a ed ixed- ilm ac i a ed sludge
(IFAS) eac o ia pa ial ni i a ion/anammox pa hway. Bio esou . Technol. 239, 294-
301. 10.1016/j.bio ech.2017.05.041.
