environmental science water research & technologyevolcke/pdf/2018 bressaniribeiro... ·...

EnvironmentalScienceWater Research & Technology

TUTORIAL REVIEW

Cite this: Environ. Sci.: Water Res.

Technol., 2018, 4, 1721

Received 20th May 2018,Accepted 30th August 2018

DOI: 10.1039/c8ew00330k

rsc.li/es-water

Trickling filters following anaerobic sewagetreatment: state of the art and perspectives

T. Bressani-Ribeiro, *ab P. G. S. Almeida,a E. I. P. Volcke b and C. A. L. Chernicharo a

High-rate anaerobic treatment of sewage using upflow anaerobic sludge blanket (UASB) reactors is a con-

solidated technology in warm climate countries. Nevertheless, since anaerobic treatment only removes or-

ganic carbon, post-treatment is required to remove nitrogen, besides residual organic carbon. Trickling fil-

ters (TFs) constitute a cost-effective post-treatment option, assuring low sludge production, low

operational costs and maintenance simplicity compared to other post-treatment technologies (e.g. acti-

vated sludge). This paper reviews the experience of the last 20 years of research, design and operation of

UASB/TF systems. Three main topics are addressed: i) the development of trickling filters for UASB reactor

effluent treatment, building on first experiences with TFs preceded by primary settlers; ii) the design criteria,

performance and empirical models for predicting the efficiency of TFs post-UASB reactors; and iii) the fu-

ture challenges associated with elimination of secondary settlers and nitrogen removal in sponge-bed

trickling filters (SBTFs).

1. Introduction

Adequate climate conditions and significant investments inresearch and development made Latin America (notably Bra-zil, Colombia and Mexico) and more recently, India becomefront-runners in using UASB reactors for sewage treatment.1

The application of such reactors for sewage treatment atlower temperatures has not yet been fully demonstrated, butlooks promising.2 Nevertheless, given that anaerobic treat-ment only removes organic carbon, post-treatment to removenitrogen and residual organic carbon is typically required tomeet effluent discharge standards. Such post-treatment facili-ties typically comprise an aerobic stage. The resulting com-bined anaerobic/aerobic systems constitute an alternative totraditional sewage treatment systems, such as activatedsludge and land-based pond systems.

The costs of a treatment plant comprising a UASB reactorfollowed by aerobic biological treatment usually allow capital

expenditure (CAPEX) savings in the range of 20–50% and op-erational expenditure (OPEX) savings above 50%, in compari-son with a conventional activated sludge plant.3,4 This is con-sidered one of the reasons for increasing sewage treatmentcoverage in Latin America.1

Post-treatment options for anaerobically treated sewageare well covered in the literature,4–9 addressing the availabletechnologies and discussing the advantages and disadvan-tages of each alternative. In summary, combined anaerobic/aerobic systems allow the achievement of the necessary effi-ciencies to comply with discharge standards in terms of car-bon and ammonium removal, mainly in developing coun-tries.1 A recent survey carried out in Brazil10 has shown thatamongst 333 investigated sewage treatment plants compris-ing UASB reactors followed by post-treatment units, tricklingfilters accounted for 25% (82 plants), as shown in Fig. 1. Interms of installed capacity, the population equivalent (PE)that can be served by the existing UASB/TF systems accountsfor approximately 3.6 million inhabitants, which represents29% of the total surveyed population.

Trickling filters (TFs) are non-submerged aerobic biofilmreactors, which were applied for sewage treatment for thefirst time in England in 1893.11 In Brazil, the first applicationwas recorded in 1910.12 A trickling filter consists basically of

Environ. Sci.: Water Res. Technol., 2018, 4, 1721–1738 | 1721This journal is © The Royal Society of Chemistry 2018

aDepartment of Sanitary and Environmental Engineering – Federal University of

Minas Gerais, Av. Antônio Carlos, 6.627 – Pampulha, Belo Horizonte – MG,

31270-901, Brazil. Tel: +55 31 3409 1946bDepartment of Green Chemistry and Technology – Ghent University, Coupure

Links 653, 9000 – Ghent, Belgium. E-mail: [email protected]

Water impact

Given that anaerobic sewage treatment only removes organic carbon, post-treatment to remove nitrogen and residual organic carbon is typically required.Trickling filters following UASB reactors have shown remarkable advantages compared to other post-treatment options. This paper gathers experience ofthe last 20 years of research, design and operation of UASB/TF systems. The design criteria, performance and main future challenges are addressed.

Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM

.

View Article OnlineView Journal | View Issue

http://crossmark.crossref.org/dialog/?doi=10.1039/c8ew00330k&domain=pdf&date_stamp=2018-10-23

http://orcid.org/0000-0002-7363-7497

http://orcid.org/0000-0002-7664-7033

http://orcid.org/0000-0001-7226-8266

http://dx.doi.org/10.1039/c8ew00330k

https://pubs.rsc.org/en/journals/journal/EW

https://pubs.rsc.org/en/journals/journal/EW?issueid=EW004011

1722 | Environ. Sci.: Water Res. Technol., 2018, 4, 1721–1738 This journal is © The Royal Society of Chemistry 2018

a tank filled with a highly-permeable material, onto whichwastewater is applied through a distribution system. Theliquid trickles downward, allowing bacterial growth (biofilm)on the surface of the packing material and air naturallymoves upward or downward.4,13 The operational simplicityand performance stability of TFs are key aspects to theirworldwide application, especially in developing countries.

Indeed, the possibility of using trickling filters post-UASBreactors has shown remarkable advantages compared to otherpost-treatment options of anaerobically treated sewage.4,14 Thetechnological choice of TFs as post-treatment units of UASBreactors ensures low sludge production and relative opera-tional and maintenance simplicity. Additionally, such a com-bined system has low operating cost, due to the reduced elec-tricity consumption and no need of chemical productdosing4 (unless P removal is a target). In this context, theUASB/TF system can play an important role in universalizingsewage treatment coverage in developing regions. Specificallyin the case of Brazil, only 42.7% of the produced sewage iscurrently treated,15 nevertheless there is a national target offull sewerage coverage and treatment until 2033.

While a comprehensive state-of-the-art review on the de-sign and operation of trickling filters preceded by primarysettlers was published by Daigger and Boltz,16 the experienceof the last 20 years of research, design and operation of trick-ling filters following UASB reactors still needs to be consoli-dated. Therefore, the structure of this review paper comprisesthree main topics: i) the development of trickling filters forUASB reactor effluent treatment, building on first experienceswith TFs preceded by primary settlers; ii) the design criteria,performance and empirical models for predicting the effi-ciency of TFs post-UASB reactors; and iii) the future chal-lenges associated with elimination of secondary settlers andnitrogen removal in sponge-bed trickling filters (SBTFs).

2. Development of trickling filtersfollowing UASB reactors for sewagetreatment

Trickling filters for sewage treatment were originally appliedfollowing primary settlers. The primary settlers were later re-

placed by UASB reactors. Over the last approximately 10years, particular attention was paid to the use of sponge-based support media. These three stages of development aredetailed below.

2.1. Trickling filters preceded by primary settlers

Trickling filters for sewage treatment were applied for thefirst time in England in 1893.11 They were initially conceivedwith rock-based support media and preceded by a primarysedimentation tank. A secondary settler following the TF wasincluded to reduce the concentration of effluent suspendedsolids and the total BOD. The resulting flowsheet (Fig. 2) istypically applied for rural and small sewage treatment plantsin cold climate regions (e.g. in the United Kingdom and Ger-many) and remains in use nowadays. Note that additionalunits for primary and secondary sludge handling are re-quired, typically for thickening, digestion and dewatering.

In this flowsheet (Fig. 2) a portion of the final effluent istypically recycled to the top of the trickling filter in order todampen BOD-loading fluctuations and consequent problemsof BOD overload and dissolved oxygen depletion.11 As BODremoval in primary settlers is typically low (<35%), effluentrecycling for feeding trickling filters in this flowsheet seeksto ensure an influent BOD concentration to the filter around100 mg L−1.3 Additionally, recirculation tends to improve sup-port media wetting efficiency.

2.2. Trickling filters following UASB reactors

Driven by the success of anaerobic sewage treatment in warmclimate regions, primary settlers were often replaced by UASBreactors, mostly in developing countries. The resulting UASB/TF flowsheet (Fig. 3) is simpler than the one with tricklingfilters preceded by primary settlers. The excess aerobic sludgefrom secondary settlers is returned to the UASB reactor3 suchthat separate units for sludge thickening and digestion arenot required. Additionally, as BOD removal in UASB reactorsis typically higher than that in primary settlers (>60%), efflu-ent recycling for feeding TFs is usually unnecessary3 settingthe admissible ranges recommended for high-rate rock-basedtrickling filter design. In the case of well-operated UASB reac-tors, even secondary settlers may turn out to be unnecessary

Fig. 1 Post-treatment technologies for UASB reactors commonly applied in Brazil.

Environmental Science: Water Research & TechnologyTutorial review

Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM

. View Article Online



(see further) depending on the operational conditions beingimposed on the trickling filters.

2.3. Trickling filter improvements through plastic- andsponge-based support media

The performance and operation of trickling filters can be im-proved with the use of plastic packing and more recently,sponges, instead of the conventional rock-based supportmedia. From the biological standpoint, the higher surfacearea for biofilm development and the ability to retain bio-mass for longer periods within the system are key points forthe performance improvements. Other relevant aspects arethe higher hydraulic capacity, better air circulation (high voidratio) and resistance to clogging.

Because the application of low hydraulic loading rates(10–20 m3 m−2 d−1) may lead to ineffective wetting of the fil-ter packing with high specific surface area, the use of plastic-

based support media (i.e. plastic rings or cross-flow) does notguarantee performance improvements, as observed in trick-ling filters preceded by primary settlers.17 Recent experimentswith random plastic rings indicated that when the organicloading is lower than 1.0 kg BOD m−3 d−1 (hydraulic loadingrates: 10–20 m3 m−2 d−1) the performance of rock- andplastic-based trickling filters for organic matter removalfollowing UASB reactors is similar.18 Whether or notrecirculation could be an operational strategy improving theperformance of plastic-bed trickling filters following UASBreactors is still a matter of further research. Interestingly, theexperiences with plastic packing reactors are seldom reportedin the literature. More frequent application of rock media asfilling material is related to the trade-offs between costs andbenefits from the perspective of more flexible legal dischargestandards for carbon and nitrogen.

A promising alternative improving the trickling filter per-formance considers the use of sponge-based support media,

Fig. 2 Sewage treatment flowsheet based on a trickling filter preceded by a primary settler. Reproduced and adapted from the work of vonSperling and Chernicharo3 with permission from IWA Publishing, Copyright 2005.

Fig. 3 Sewage treatment flowsheet based on a trickling filter following a UASB reactor. Reproduced and adapted from the work of von Sperlingand Chernicharo3 with permission from IWA Publishing, Copyright 2005.

Environmental Science: Water Research & Technology Tutorial review

Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




as discussed by Kassab.7 The use of polyurethane sponge as asupport medium in TFs following UASB reactors has begunwith the development of a system so-called ‘downflow hangingsponge (DHS)’ by the research group of Professor HidekiHarada, at the Universities of Tohoku and Nagaoka (Japan).19

A comprehensive report on the steps through the develop-ment and evolution of the DHS system can be found in thestudy of Uemura and Harada.64 From the research outcomes,a Brazilian research group of the Universidade Federal deMinas Gerais (UFMG, Brazil) has been developing self-structured sponge-based packing media since 2012.

The first study with the DHS system in Brazil wasimplemented considering the comparison with other types ofpacking materials, such as blast furnace slag, plastic rings andcorrugated plastic tubing.18 Under the same operational condi-tions, the DHS performance tended to be superior for organicmatter removal, and the possibility to operate the system with-out secondary settlers was preliminarily verified. In fact, theuse of plastic-based packing media did not contribute any per-formance improvements, considering the operational condi-tions imposed on the systems. In sequence, another study wasdeveloped in order to verify the benefits of using sponge-basedpacking media.20 The results clearly demonstrated the benefitsprovided by sponges, improving the effluent quality in termsof organic matter and also indicating a noticeable increase inammonium removal. Additionally, a molecular investigationwas implemented aiming to observe the bacterial communitycomposition by pyrosequencing.21 The results suggested thatdenitrifiers, nitrifiers, and even anammox bacteria coexistedin the reactor. Considering this finding and the observed po-tential to retain slow-growing organisms within sponge-bedtrickling filters,22,23 future improvements could also be di-rected towards total nitrogen removal (see further).

From the results obtained in Japan and in Brazil, a self-structured sponge-based packing medium was industriallyprototyped and tested on a demonstration scale under a typi-cal full-scale sewage treatment plant flow regime.24 The justi-fication adopting a self-structured design is based on the factthat intermediate underdrains are needed when randomsponge-based packing media are used, avoiding excessivecompression of the elements. Because the system wasdesigned to operate at higher organic loadings, nitrificationactivity was lower, as expected. Although nitrification de-creased at higher organic loading rates (0.45 kg BODmreactor

−3 d−1), it is important to point out that with the useof the prototyped sponge-based support media, 50% ammo-nium removal was observed, whereas it can only be achievedin rock-based trickling filters with applied volumetric loadingrates around 0.22 kg BOD mreactor

−3 d−1.13 The main steps forthe development of the sponge-based packing media(Spongepacking) at UFMG are summarized in Table 1.

Fundamental aspects related to the improvements pro-vided by sponge-based packing media. Interstitial biomass re-tention within the sponge-based media leads to a longer HRT(≈1.5–2.5 h) and sludge retention time (SRT) (>100 days), en-suring a significant improvement for the removal of organic

matter and ammonium in the post-treatment step. In addi-tion, long interstitial retention of biomass favours lysis/hy-drolysis of inactive cells, resulting in low excess of sludgeproduction (around 0.10 g SS gremoved

−1 BOD). High sludgeconcentrations, sufficient oxygen supply, adequate endoge-nous respiration rates and a high density and diversity ofmicrofauna were also relevant factors highlighted by Onoderaand colleagues,28 explaining why the sludge accumulationwas in near balance with the degradation of sludge withinthe reactor. Tandukar and colleagues29 also correlated thedegradation of sludge with heterotrophic denitrificationwithin anoxic zones of the sponges (see further). It is impor-tant to point out that a longer SRT tends to reduce the start-up period in a sponge-bed trickling filter.

In a DHS post-UASB reactor, the outer portion of thesponges tends to be predominantly aerobic.30 It is worth men-tioning that Araki et al.30 and Machdar et al.31 reported a con-centration of 1.0 mg DO L−1 even at the inner portions of thesponge (e.g. 6 mm in depth). This finding suggests an increasein aerobic niches within a sponge-bed bioreactor where slow-growers, such as nitrifiers, could also be active. Higher DO con-centration tends to increase the presence of macrofauna, whichcontributes to predation of bacteria and protozoa.28,32 However,because macrofauna overgrazing is only prominently observedon the surface of the sponge media (smaller pores tend tophysically protect biomass from snails), the presence of the re-lated organisms does not lead to significant impacts on theprocess performance.33,34 Thus, the ability of the sponge-basedmedium to protect biomass from snail overgrazing is an impor-tant aspect in maintaining the activity of microorganisms,especially the activity of slow-growers (e.g. nitrifiers).

Within the sponge pores as well as in the biofilm devel-oped on sponge fibers, a DO gradient occurs, creating anoxicniches and a suitable environment for heterotrophic denitrifi-cation. Thus, hydrolysed biomass could be utilized as an ad-ditional source of substrates in the presence of oxidizedforms of nitrogen. Because of the anoxic zones within thesponges and longer SRT, we cannot eliminate the hypothesisthat the activity of anaerobic ammonium-oxidizing bacteria(anammox bacteria) could be also a factor for nitrogen re-moval. In fact, anammox bacteria were detected in severalsamples within a sponge-bed trickling filter following a UASBreactor, as discussed by Mac Conell et al.21 However, nitrogenremoval tends to be more relevant (≈65%) within fully venti-lated systems only when effluent recycling is practiced, as ob-served by Okubo et al.35 This indicates that the presence ofreadily biodegradable organic matter and oxidized forms ofnitrogen in anoxic zones is essential for a higher heterotro-phic denitrification activity.

3. Design and performance of tricklingfilters following UASB reactors3.1. Main design criteria for TFs following UASB reactors

The main design criteria associated with trickling filters con-cern the hydraulic loading rate (HLR) and the organic


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




loading rate (OLR), primarily originating from the experienceof TFs preceded by primary settlers for sewage treatment.13,36

In this case, five categories of TFs based on the imposed HLRwere reported in the literature: low rate (1.0 to 4.0 m3 m−2

d−1), intermediate rate (3.5 to 10.0 m3 m−2 d−1), high rate(10.0 to 40.0 m3 m−2 d−1), super high rate (12.0 to 70.0 m3

m−2 d−1) and rough (45.0 to 185.0 m3 m−2 d−1).13,37,38

Design of rock-bed trickling filters following UASB reac-tors. In general, most of the full-scale units have been

designed as high-rate trickling filters, adopting organic load-ing rates between 0.5 to 1.0 kg BOD m−3 d−1. The reason foradopting higher organic loading rates is based on the lessstringent discharge standards considered in developing coun-tries to progressively accomplish the goals for sewage treat-ment expansion. Under these conditions, satisfactory removalof organic matter is expected, with very low NH4

+-N removal.The listed criteria and parameters for design rock-bed trick-ling filters as a post-treatment of UASB reactor effluents

Table 1 Development of sponge-based packing media for the improvement of trickling filters following UASB reactors: lessons learned from the UFMGexperience

Reference (focus of the research) Scale

Operational conditions(trickling filter)

Main lessons learnedHLR(m3 m−2 d−1)

OLR(kg BOD m−3 d−1)

Almeida et al.18,25 (perspective ofusing sponge-based packing media;operation without secondary settlers)

Demo-scale(V = 8.7 m3)

10–20 0.25–0.45 • From a comparison study with different mediabeing tested, the benefits of using sponges weredemonstrated in terms of organic matter removal,indicating the possibility to operate UASB/TFsystems without secondary settlers• The overall performance for the DHS system(a sponge-based medium) was above 80% for BOD,COD and TSS. However, the operational conditionsseemed to be inappropriate when ammoniumremoval above 50% was required

Almeida;26 Almeida et al.20 (effectiveinfluence of sponges improving theUASB/TF system performance;operation without secondary settlers)


10–12 0.12–0.20 • Considering the outcomes from previous studies,the influence of sponges enhancing the performanceof low-rate trickling filters following UASB reactorswas investigated• The use of sponge-based packing media wassignificantly more effective than that of plastic-basedpacking media for organic matter removal. Forammonium removal, 50% was achieved using aplastic-based medium, whereas 80–95% wasobserved with the use of sponges• The sponges notably increased the reliability ofthe system to operate without secondary settlers

Mac Conell et al.21 (microorganismsinvolved in the nitrogen cycle within asponge-bed trickling filter)


10–12 0.12–0.20 • The bacterial community composition involved inthe N-cycle within a sponge-bed trickling filtertreating UASB effluent was investigated bypyrosequencing• The results revealed that denitrifiers, nitrifiers,and anammox bacteria coexisted in the reactor,suggesting that different metabolic pathways wereinvolved in nitrogen removal within the system,including the activity of anammox bacteria

Bressani-Ribeiro;27 Bressani-Ribeiro et al.24

(performance of a UASB–SBTF systemwithout secondary settlers, consideringrealistic conditions in terms of flowregime; sponge-based packing media forindustrial prototype testing)


11.5 0.45 • The performance of the UASB/TF systemconsidering the use of sponge-based packing mediaindustrially prototyped was evaluated. A typicalfull-scale sewage treatment plant flow regime wasimposed on the system for a more realisticinvestigation. Higher organic loadings were appliedto the trickling filter in order to observe theperformance of the system under limitingoperational conditions• The possibility to operate the system withoutsecondary settlers was confirmed, even underlimiting operational conditions. Organic matterremoval was around 85–90%.As observed in previous studies, the resultsconfirmed that ammonium removal above 50%should not be expected

HLR: hydraulic loading rate; OLR: organic loading rate.


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




(Table 2) are based on the outputs of the Brazilian ResearchProgram on Basic Sanitation (PROSAB).39 The expected per-formance of rock-bed trickling filters following UASB reactorsis currently based on practical experiences, as empiricalmodels predicting organic matter and ammonium removalare not well established, as further discussed.

Dimensioning of trickling filters should take placethrough the following subsequent steps: i) determine the or-ganic loading rate (OLR) to be applied, taking into accountthe BOD and COD concentrations of the UASB reactor efflu-ent; ii) calculate the support media volume, considering arecommended OLR value; iii) select a support media height,which will be typically between 2.0 and 3.0 m; iv) calculatethe surface area of the biofilter, considering the predefinedvolume and support media height; v) verify the applied hy-draulic loading rate.

For UASB reactors treating sewage (typically diluted – BOD<400 mg L−1), the organic matter concentration in the anaer-obic effluent is typically low (<100 mg BOD L−1), implying alow OLR to the post-treatment step. As a result, the followingdesign steps of calculating the support media volume, deter-mining the height and calculating the surface area are intrin-sically correlated. After obtaining a rock-media volume for anapplied OLR, choosing heights between 2.0 and 3.0 m is suf-ficient to ensure a minimum surface area required for suit-able HLR conditions. Therefore, the HLR, and consequentlythe surface area, is usually the limiting factor for designingrock-bed trickling filters post-treating anaerobic effluents.

Design of sponge-bed trickling filters following UASB reac-tors. Important efforts have been directed towards establishingthe design criteria based on practical experiences with sponge-bed trickling filters (SBTFs) following UASB reactors.22,24,33,35,40

The downflow hanging sponge system (DHS system) is gener-ally designed for carbon and ammonium removal in a singlereaction volume, as a low-rate bioreactor. Effluent recyclinghas also been practiced in full-scale systems.35 Table 3 sum-marizes some experiences and the observed operational condi-tions and performances related to sponge-bed trickling filterstreating anaerobic effluents.

From a full-scale experience,35 the organic loading rate ap-plied to the DHS system was around 2.80 kg COD msponge

−3

d−1 (0.23 kg BOD mreactor−3 d−1), considering a theoretical hy-

draulic retention time (HRT) of 1.5 h. The sponge occupancywas around 25% of the reactor, with a useful height of 5 me-ters. The trickling filter performance for BOD and ammo-nium removal was around 90% and 80%, respectively, con-

sidering 100% effluent recirculation. In a recent lab-experience,33 a higher organic loading rate was applied to theDHS system (0.40 kg BOD mreactor

−3 d−1), with removal effi-ciencies of 85% for BOD and around 80% for ammonium,similar to those reported for the full-scale experience. Thesponge occupancy and HRT were 40% and 2 hours, respec-tively, and a useful height of 3 meters was adopted. No efflu-ent recycling was practiced during the experimental period.For similar design and operational conditions, Bressani-Ribeiro et al.24 reported removal efficiencies of 72% for BODand 44% for ammonium. The demo-scale UASB/SBTF systemwas operated under a typical flow regime, as observed in full-scale sewage treatment plants, with no effluent recycling. Theuseful reactor height was 3.5 m.

The observed differences in terms of performance canprobably be related not only to the choice of operational con-ditions, but also to the intrinsic characteristics of the supportmedia used to perform each experiment. For the design ofthe full-scale DHS system, a curtain-type packing mediumwas used, whereas for the lab- and demo-scale systems, ran-dom and vertical self-structured packing media were used, re-spectively. The pore size used for the random packing mediawas 1.6 mm, which was larger than the pore size of thepolyurethane-media previously used.33 The pore size tends toaffect the HRT and solid retention in sponge-bed trickling fil-ters. As reported by Tawfik et al.,41 an increase in HRT forsponges with 0.56 mm pores compared to those with poresranging from 0.63 to 1.92 m was observed. In addition, con-sidering the very similar operational conditions observed tolab- and demo-scale operation conditions (Table 3), deeperpenetration of oxygen (although no systematic evaluationswere performed) within the sponges and the influent flow re-gime could explain the observed differences in terms of BODand NH4

+-N removal performance.Adopting loading criteria (0.20–0.40 kg BOD mreactor

−3 d−1)to design sponge-bed trickling filters, the hydraulic retentiontime needs to be verified for the usual range. In general,most of the experiences related to sponge-bed trickling filtersfollowing UASB reactors indicate HRT ranging from 1.5 to 2.5h. Because of the uncertainties regarding operational condi-tions and the system's performance, the design of sponge-bed trickling filters following UASB reactors should be cur-rently based on practical experiences (e.g. as observed inTable 3).

Detailed trickling filter design aspects associated with ef-fluent distribution systems (rotating distributor arms),

Table 2 Criteria and parameters currently adopted for the design of high-rate rock-bed trickling filters following UASB reactors

Parameter

Operational range

Average flow Maximum daily flow Maximum hourly flow

Organic loading rate (kg BOD m−3 d−1) 0.5–1.0 0.5–1.0 0.5–1.0Hydraulic loading rate (m3 m−2 d−1) 15–30 18–22 25–30Height (m) 2.0–3.0 2.0–3.0 2.0–3.0

Note: for rock-bed trickling filters following UASB reactors the adopted sludge production rate is 0.8–1.0 kg SS kgremoved−1 BOD.


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




underdrains and natural ventilation systems fall beyond thescope of this review paper – they are described in classictextbooks.11,13

3.2. Performance of rock-, plastic- and sponge-bed tricklingfilters post-UASB reactors for carbon and ammonium removal

Experiences of trickling filters using conventional supportmedia (rock and plastic) following UASB reactors are summa-rized in Table 4, including typical OLR and HLR ranges ap-plied and the corresponding carbon and ammonium removalefficiencies. The UASB–trickling filter systems considered theoperation with secondary clarifiers.

The BOD removal efficiencies of trickling filters fed witheffluent from UASB reactors (median of 65%; Table 3) aretypically lower than those for traditional TFs following pri-mary settlers (65–80%;13). This was to be expected, sincemost of the readily biodegradable organic matter was con-sumed in the anaerobic step. Nevertheless, the overall BODremoval efficiency (UASB + trickling filter) is likely to remainsimilar to that for the traditional combination of primary set-tlers and trickling filters. From the operational results indi-cated in Table 3, it was observed that the UASB–trickling fil-ter systems produced effluents with BOD and TSSconcentrations below 40 mg BOD L−1 and 30 mg TSS L−1, re-spectively (OLR 0.45–0.90 kg BOD m−3 d−1; HLR 10–30 m3

m−2 d−1). From additional practical experiences with high-rate rock-bed trickling filters as post-treatment of UASBreactor effluents, maximum OLR and HLR values should bebetween 0.5–1.0 kg BOD m−3 d−1 and 20 and 30 m3 m−2 d−1

for BOD and SS concentrations lower than 60 mg L−1.39 Interms of ammonium removal, it can be noticed that for OLRvalues above 0.20 kg BOD m−3 d−1 (OLR ranging from 0.20–0.70 kg BOD m−3 d−1), ammonium removal around 10–42% isobserved.

Typical BOD/NH4+-N ratios fed to trickling filters following

primary settlers are approximately 6.5.38 Due to the relativelyhigher removal of readily biodegradable organic matter(especially soluble BOD fractions) in UASB reactors, theUASB effluent BOD/NH4

+-N ratio is lower, approximately 2.24

This gives more opportunities for the nitrifying bacteria,because relatively less oxygen is consumed by heterotrophs.

Denitrification could take place in addition, by implementingrecirculation of the final effluent to the inlet trickling filter(see section 4.2.1).

The experiences with sponge-bed trickling filters followingUASB reactors for sewage treatment are summarized inTable 5. Typical combined UASB/SBTF removal efficienciesare 95% for BOD, 85–90% for COD and 70–90% for TSS. Am-monium removal efficiencies above 70% are obtained for anOLR up to 2.0 kg COD msponge

−3 d−1 (around 0.76 kg CODmreactor

−3 d−1), which is higher than would be expected forrock and plastic-based TFs (ammonium removal efficiencieslower than 50% at an OLR of 0.50 kg COD m−3

reactor d−1). This

could be due to the relatively better retention of nitrifyingbiomass in sponge-bed trickling filters due to the longersludge retention time (SRT) and HRT and greater oxygenavailability.29,52

Ammonium removal efficiency increases with decreasingOLR only to a certain limit. Tawfik et al.53 observed that anOLR reduction from 2.6 to 1.6 kg COD msponge

−3 d−1 resultedin 29% increase in NH4

+-N removal efficiency. However, fur-ther OLR decrease to 1.3 kg COD msponge

−3 d−1 did not resultin better nitrification efficiency. This could be explained bysubstrate limitation (NH4

+-N and alkalinity) in the bottomcompartment of the SBTF (below 3 meters height). MacConell et al.54 observed a population reduction of ammoniaoxidizing bacteria and consequently lower nitrification ratesat the bottom compartment of a SBTF operated at OLR valuesbetween 0.45–0.55 kg COD msponge

−3 d−1.Low specific sponge volumes (msponge

3 mreactor−3) may re-

sult in lower organic matter and ammonium removal effi-ciencies considering the same applied OLR. Tawfik et al.55

determined a reduction of COD and NH4+-N removal efficien-

cies from 80 to 62% and 86 to 38%, respectively, when thesponge volume changed from 38 to 19% under the same ap-plied OLR (2.0 kg COD msponge

−3 d−1). Such an effect can beassociated with a reduction in the specific surface area andconsequently surface adsorption, which is the first step inthe sequence of organic matter degradation in sponge-bedtrickling filters.

Finally, it is worth noting that significant coliform re-moval takes place in SBTF systems (up to 4.2 log units in ademo-scale set-up, see Bressani-Ribeiro et al. study24). This is

Table 3 Summary of operational conditions and performances for BOD and ammonium removal observed for sponge-bed trickling filters followingUASB reactors (lab-, demo- and full-scale experiences)

Reference

Operational conditions SBTF performance

ScaleOLR(kg BOD m3 d−1)

HLR(m3 m−2 d−1)

HRT(h)

Spongeoccupancy (%)

Effluentrecycling

BOD removal(%)

NH4+-N

removal (%)

Okubo et al.;35 Onodera et al.40 0.23 21 1.5 25 1 : 1 90 80 Full-scaleOnodera et al.33 0.40 12 2.0 40 None 85 80 Lab-scaleBressani-Ribeiro et al.24 0.40 11.5 2.5 40 None 72 44 Demo-scale

OLR: organic loading rate; HLR: hydraulic loading rate; HRT: hydraulic retention time. Note: the reported results related to the studyof Bressani-Ribeiro et al.24 consider the operation of the UASB/SBTF system under a typical flow regime, as observed in full-scale sewagetreatment plants.


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




probably associated with the mechanism of adsorptionfollowed by predation in the SBTF systems.41

3.3. Empirical models for predicting the performance oftrickling filters following UASB reactors

BOD removal estimation. Empirical models predicting ef-fluent BOD concentrations or BOD removal efficiencies areavailable for trickling filters preceded by primary settlers.16,68

However, because UASB effluents tend to present lower read-ily biodegradable organic matter, such models should behandled with caution when dealing with UASB reactor efflu-ents.39 In any case, extrapolations outside the operationalrange should be avoided.11

The classic empirical models available for the predictionof rock-bed trickling filter performance following primary set-tlers are: i) the National Research Council (NRC) model for

BOD removal efficiency; ii) the Galler–Gotaas model for BODeffluent concentration; and iii) the Eckenfelder model forBOD effluent concentration. Detailed descriptions and evalu-ations of these and other equations for TF performance pre-diction are presented in the Manual of Practice n. 8.68 Someof the developed equations consider a certain amount of em-pirical coefficients strictly related to trickling filters precededby primary settlers, using plastic-based support media. How-ever, most of the trickling filters following UASB reactors arecurrently filled with crushed stones, and operational datafrom plastic-bed trickling filters are seldom available, as pre-viously discussed. Table 6 presents the equations typicallyused to design rock-bed trickling filters preceded by primarysettlers, as well as the parameters considered for the empiri-cal relations.

Fig. 4 presents a comparison of predicted final effluentBOD concentrations considering the usual operative range

Table 4 Main characteristics and performance of rock and plastic-bed trickling filters following UASB reactors (and followed by secondary settlers)

Scale

Operational conditions TF characteristicsUASB reactor effluent concentrations(mg L−1)

Flow (m3 d−1) OLR (kg BOD m−3 d−1) HLR (m3 m−2 d−1) Height (m) Volume (m3) Media BOD TSS NH4+-N

Pilot 0.54 0.07 3.1 2.1 0.2 Plastic — — 35Pilot 0.54 0.13 5.6 2.1 0.2 Plastic — — 30Full —c 0.09–0.22 — — — Plastic — — 12–25Demo 20.0 0.24 10.0 3.0 3.0 Plastic 105 39 37Demo 69.0 0.31 13.6 1.9 9.7 Slag 44 35 23Pilot 6.0 0.33 32.1 4.0 1.1 Rock 40 50 21Pilot 6.0 0.42 21.2 4.0 1.1 Rock 78 64 21Full 7402 0.45 8.7 2.7 2309 Rock ∼140 140 36Demo 69.1 0.50 — 1.9 11.5 Slag 88 42 —Full 155.52 0.56 13.3 2.5 3300 Rock 106 181 —Demo 69.0 0.68 13.6 1.9 9.7 Slag 96 75 30Demo 69.0 0.68 13.6 1.9 9.7 Slag 96 48 29Demo 72.0 0.89 63.7 7.0 7.9 Plastic 99 — 30Full 155.52 <1.0 16.8 2.5 3299 Rock — — —Full 14.69 — 18.5 2.5 994 Rock 114 38 —— 0.07–1.0 [0.42] 3.1–63.7 [13.6] 1.9–7.0 [2.5] — — 40–140 [96] 35–181 [49] 19–37 [30]

Scale

Trickling filter effluent concentrations(mg L−1) [removal efficiency – %] Ammonium-N removal

rate (kg N m−3 d−1) ReferenceBOD TSS NH4+-N

Pilot — — 10 [71] 0.04 Victoria42

Pilot — — 15 [50] 0.04 Victoria42

Full — — 0.8–1.6 [90] — Pearce et al.43

Demo 44 [58] 13 [67] 28 [25] 0.06 Fonseca44

Demo 23 [48] 14 [60] 19 [17] 0.03 Frade45

Pilot 18 [55] 23 [54] 17 [19] 0.03 Aisse46

Pilot 37 [53] 26 [59] 18 [14] 0.02 Aisse46

Full 30 [79] — 21 [42] 0.05 Sanepar (with Almeida26)Demo 31 [65] 19 [55] — — Pontes and Chernicharoa,47

Full 37 [65] 53 [71] — — Moraes et al.48

Demo 42 [56] 34 [55] 26 [13] 0.03 Frade45

Demo 32 [67] 22 [54] 28 [3] 0.01 Frade45

Demo 32 [68] — 21 [10] 0.03 Collivignarelli et al.49

Full 85–89b 86–89b — — Chernicharo et al.a,4,50

Full 18 [84] 11 [53] — — Lobato et al.a,51

— 18–44 [32] 48–84 [65] 11–53 [22] 53–71 [55] 1.2–28 [19] 3–90 [19] 0.01–0.06 [0.03] Observed ranges (efficiency in italic)[median]

a Flowsheet with secondary sludge return for digestion and thickening in the UASB reactor. b Overall STP removal efficiencies (PE ≈ 1 000 000inhab.), not considered for the observed ranges. c Population equivalents between 2500 and 80 000 inhabitants.


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




Table 5 Main characteristics and performance of sponge-bed trickling filters following UASB reactors (and followed by secondary settlers)

Scale

TF characteristics Operational conditions TF influent concentrations

Support media[pore (mm)]

Sponge volume(m3) [spongeoccupancy – %]

Height(m)

OLR applied

HLR(m3 m−2 d−1)

COD(mg L−1)

TSS(mg L−1)

BOD(mg L−1)

NH4+

(mg L−1)kg CODmsponge

3 d−1kg TKNm3 d−1

Pilot Cylinders [0.3–1.0] — 0.5 0.2 — — 172 — — 30Demo Spongepacking [1.0] 4.0 [40%] 3.5 2.0 0.37 11.5 218 79 123 26Full DHS-G3 [0.46] 27.7 [22.3%] 5.31 1.1 — 40.8 168 51 62 26Full DHS-G2 31.1 [24.7%] 5.31 2.84 — 21.0 177 53 56 —Pilot DHS-G3 0.102 [53%] 4.8 1.34 — — 63 33 15 6.9Pilot DHS-G6 [1.6] 0.046 [33.8%] 4.0 2.03 0.41 12.2 169 44 93 25Pilot DHS-G3 [0.63] 0.86 [53%] 4.0 0.9 0.10 — 113 33 53 27Demo Rotosponge 1.85 [49%] 4.0 0.36 — 10 160 60 60 40Demo DHS-G3/G5 [0.89] 0.93–1.62

[31–54%]— 1.4 — 14.0 147 54 — —

Pilot DHS-G1 2.4 × 10−4

[30%]2.0 1.24 — 7.0 106 — — 19

Pilot DHS-G3 [0.63] 0.024 [18%] 3.5 1.84 — 7.6 169 49 84 —Pilot DHS-G3 [0.63] 0.024 [18%] 3.5 3.2 — 10.1 — — — —Pilot DHS-G3 [0.63] 0.024 [18%] 3.5 4.8 — 15.2 — — — —Pilot DHS-G3 [0.63] 0.024 [18%] 3.5 6.2 — 30.3 — — — —Demo DHS-G3 4.3 [31%] 3.0 — — — 94 37 41 —Pilot DHS-G3 0.05 [38.2%] 3.5 2.0 0.29 14.6 226 50 — 28Pilot DHS-G3 [0.63] 0.133 [18%] 3.5 3.4 — 7.6 287 — — 21Full DHS-G2 31.1 [24.7%] 5.31 — — — 166 66 53 26Demo Rotosponge 1.90 [49%] 4.20 0.11–0.37 — 10–12 170 40 58 —Pilot DHS-G3 [0.63] 0.0516 [38%] 3.5 1.6 0.27 12.1 — — — —Pilot DHS-G3 [0.63] 0.0516 [38%] 3.5 2.6 0.36 16.3 — — — —Pilot DHS-G5 [0.63] 0.480 [55%] 4.0 2.17 0.28 11.0 227 41 136 23Pilot DHS-G3[0.63] 0.0516 [38%] 3.5 1.6 0.13 12.1 178 47 67 21Pilot DHS-G2 0.051 2.0 2.03 — 14.6 167 71 55 40Pilot DHS-G2 0.051 4.0 3.15 — 20.9 173 75 68 37Pilot DHS-G4 0.38 [39%] 4.0 2.40 0.49 18.0 — — — —Pilot DHS-G4 0.375 [39%] 4.0 2.34 0.48 18.0 195 66 78 25Pilot DHS-G2 — — — — — 161 56 51 39— — 18%–55%

[34%]2.0–5.31[3.5]

0.9–6.2[2.1]

0.1–0.49[0.29]

7–30.3[14.6]

63–227[169]

33–75[50]

15–136[56]

6.9–40[26]

ScaleTF effluent concentrations [removal efficiencies – %] Removal rate of NH4

+-NReferencesCOD (mg L−1) TSS (mg L−1) BOD (mg L−1) NH4

+ (mg L−1) g NH4+-N m−3 d−1

Pilot 46 [73%] — — 2 [93%] — Bundy et al.56

Demo 83 [62%] 30 [62%] 35 [72%] 17 [44%] 84 Bressani-Ribeiro et al.a,24

Full 40 [76%] 11 [78%] 10 [91%] 12 [54%] — Okubo et al.57

Full 37 [79%] 19 [64%] 6 [89%] — — Okubo et al.35

Pilot 25 [60%] 1 [97%] 2 [87%] 0.1 [99%] 81 Yoochatchaval et al.b,58

Pilot 48 [68%] 17 [51%] 12 [87%] 4 [84%] 252 Onodera et al.33

Pilot 36 [68%] 12 [64%] 7 [87%] 3 [89%] 180 Onodera et al.28

Demo 50 [69%] 20 [67%] 15 [75%] 8 [80%] Almeida et al.20

Demo 78 [47%] 34 [37%] — — — Tanaka et al.c,59

Pilot 14 [87%] — — 1 [95%] 203 Uemura et al.60

Pilot 50 [70%] 13 [73%] 11 [87%] 4 [83%] 115 Mahmoud et al.d,61

Pilot 74 [61%] — — ∼15 [49%] — Mahmoud et al.61

Pilot 94 [52%] — — ~20 [27%] — Mahmoud et al.61

Pilot 128 [34%] — — ~25 [13%] — Mahmoud et al.61

Demo 68 [28%] 45 [−21%] 8 [80%] — — Takahashi et al.c,62

Pilot 41 [82%] 33 [67%] — 4 [86%] 199 Tawfik et al.55

Pilot 121 [58%] — — 6 [72%] 180 Mahmoud et al.63

Full 33 [80%] 8 [88%] 6 [89%] 5 [81%] — Uemura and Haradae,64

Demo 80 [53%] 10 [75%] 20 [66%] — Chernicharo and Almeida65

Pilot — — 8 3 [88%] 196 Tawfik et al.53

Pilot 63 [88%] 21 8 [59%] 204 Tawfik et al.53

Pilot 62 [73%] 18 [56%] 17 [88%] 9 [61%] 117 Tandukar et al.66

Pilot 43 [75%] 12 [75%] 2.3 [97%] 3 [86%] 180 Tawfik et al.41



Pilot 46 [76%] 17 [74%] 8 [28%] — Tandukar et al.29


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




for trickling filters following UASB reactors. Typical values24

of effluent BODUASB concentrations, organic and hydraulicloading rates and reactor height were used to implement thegraphical representation. It was assumed that the UASB efflu-ent and biofilm formation was evenly distributed (slime dis-tribution factor – m = 0).

The empirical relation from Galler–Gotaas seems to con-siderably underpredict the final BOD effluent concentrations,whereas the NRC and Eckenfelder equations tend to overpre-dict the trickling filter performance. The low adherence ofthe model results to the expected effluent BOD concentration

(based on practical experiences) is possibly related to the factthat the models have been developed for trickling filters pre-ceded by primary settlers. However, the largest portion ofreadily biodegradable organic matter is consumed in theanaerobic step, which may reflect in the performance of trick-ling filters following UASB reactors.

Considering simple calibration for the effluent treatabilityfactor (k), the Eckenfelder formula seems to provide a goodestimation for BOD concentration in the final effluent. Whenproperly calibrated, the equation might be potentially appliedto the design of trickling filters following UASB reactors, as

Table 5 (continued)


Pilot 68 [62%] 46 [39%] 10 [83%] 15 [61%] — Machdar et al.31

— 14–128 [56] [68%] 1–46 [18] [67%] 2–21 [8] [87%] 0.1–25 [6] [73%] 81–498 [188] Observed ranges(efficiency in italic)[median]

*G1 to G6 are different configurations of the downflow hanging sponge.a Operation under a typical full-scale flow regime, without secondarysettlers. b TF directly receiving diluted sewage (BOD between 20 and 50 mg L−1). c Operation with mechanical ventilation. d Polyurethanesponge was placed in the settler compartment of the UASB reactor, improving TSS removal. e Operation with recirculation (R = 1).

Table 6 Equations typically considered to estimate rock-bed trickling filter performance following UASB reactors

Empirical model Equation Comments

NRC formula*currently used to estimatethe performance of BODremoval in TFs followingUASB reactors

E

F

1

1 0 4432. OLR

• For the model development several operational data fromrock-bed trickling filters were statistically analyzed. Thesystems operated essentially treat relatively high concentratedsewage from military bases

E = BOD removal efficiency ratio at 20 °COLR = organic loading rate (kg BOD m−3 d−1)F = recirculation factor(1 + Qrecycle/Qinfluent), where F = 1,if Qrecycle = 0

• Basically developed to estimate the BOD removal efficiency asa function of organic loading rate and effluent recycling• The formula does consider the effect of secondary clarifiers

Galler–Gotaas formulaE K v S v S

v v h r

r

r

( )

( ) ( . )

.

. . .in efl

1 19

0 78 0 67 0 250 305

• Developed from multiple regression analysis based on a largerdatabase of sewage treatment plants

Kv T

0 570 28 0 15

.. .

• The formula allows the inclusion of the geometry effect,hydraulic and organic loadings and effluent recycling

E = BOD removal efficiency ratio at 20 °C • Include intrinsic exponents not easily handled for analyticalcalibrationsv = hydraulic loading rate (m3 m−2 d−1)

vr = recycle loading rate (m3 m−2 d−1)Sin = influent BOD concentration (mg L−1)Sefl = effluent BOD concentration (mg L−1)h = filter height (m)r = radius of the filter (m)T = temperature (°C)

Eckenfelder formula*proposed formula toestimate BOD concentration S S k h

Q A

m

nefl in

0 1

1

expsup

• The semi-empirical formula considering BOD removal as afirst-order function along the trickling filter• The equation considers the effect of the wastewater treatability,also including the effect of the geometry and hydraulic andflow ratesSin = influent BOD concentration (g m−3)

Sefl = effluent BOD concentration (g m−3)A = plan-view area (ft2) • Because wastewater treatability (k) and filter medium (n) are

variables directly included, when properly calibrated, the equationmight be potentially applied to design trickling filters followingUASB reactors, even when innovative support media are considered(e.g. sponge-based media)

h = packing height (ft)Q = flow rate (gallons per min)k = factor for effluent treatabilityn = filter medium exponentm = slime distribution factor(m = 0 if evenly distributed)


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




observed in preliminary fitting plant data from a full-scaleUASB-trickling filter system (data not shown). As theEckenfelder semi-empirical equation considers an exponent re-lated to the support media used, it could be applied evenwhen innovative support media are adopted (e.g. sponge-basedmedia). Although the NRC formula may provide a reasonableapproximation for the final BOD concentrations, more effortsshould be directed towards determination of the typical rangefor filter media (n) and the effluent treatability factor (k) re-lated to trickling filters treating UASB reactor effluents.

4. Sponge-bed trickling filterspost-UASB reactors: future challenges

The future challenges regarding effluent quality improvementand further simplification of construction, operation and

maintenance of SBTFs post-UASB reactors seem to be relatedto two main topics: i) elimination of secondary settlers andii) establishing the best design and operational conditionsfor innovative nitrogen removal. These aspects are addressedbelow.

4.1. Flowsheet simplification via elimination of secondarysettlers

One of the main advantages of using sponge-based packingmedia is related to less stringent construction requirementsof the TF tank compared to rock-based systems, achieved byimplementing self-structured media. From the treatment pro-cess standpoint, the use of sponges allows the retention ofmicroorganisms for longer periods at longer hydraulic reten-tion times for the same bed volume, compared with rock orplastic-bed TFs.20 Moreover, no additional operational strate-gies (e.g. recirculation of the final effluent) are needed tomeet the discharge standards generally adopted in develop-ing countries or even additional procedures to overcome clog-ging issues or improving the wetting efficiency.

Furthermore, the increase in solid retention time (SRT)provided by the sponge media leads to high levels of endoge-nous respiration rates, contributing to a low total suspendedsolid (TSS) loading in the effluent (Table 7). In fact, TSS re-movals around 70–90% are reported from full-scale experi-ences.35 Excess sludge produced in SBTFs following UASB re-actors reported in the literature is shown in Table 7.

Low excess sludge production in SBTFs post-UASB reactorscan be observed. Additionally, even for simplified UASB/SBTFflowsheets (without secondary settlers), lower median valueshave been reported (0.28 kg TSS kgremoved

−1 COD). In fact, arelatively low excess of sludge is produced in SBTF systemscompared to the typical range of 0.25–0.88 kg TSS kgremoved

−1

Fig. 4 Graphical representation of empirical and semi-empirical rela-tions for a typical organic loading rate range applied to trickling filtersfollowing UASB reactors. The expected effluent BOD concentration (inred) is based on practical experiences.

Table 7 Excess sludge production and effluent TSS loadings in UASB/SBTF systems with and without secondary settlers

SRT (d) HRT (h)OLR (kg CODmsponge

3 d−1)Excess sludge production(kg TSS kgremoved

−1 COD)Effluent TSS loading(kg TSS mreactor

−3 d−1)Secondarysettler (Sset) Reference

— 1.2 0.8 0.16 0.09 Without Sset Bressani-Ribeiro27

— 2.0 0.82–1.33 0.06–0.20a 0.05 Almeida et al.20

— — 0.76b 0.38 — Almeida et al.18

168 2.7 1.6 0.09 0.04 Tawfik et al.41

47 2.0 2.03 0.08 0.07 With Sset Onodera et al.33

>135 3.2 0.85 0.18 0.05 Onodera et al.28

38 2.9 6.8 0.39 — Tawfik et al.c,69

64 5.8 3.6 0.26 — Tawfik et al.c,69

109 11.7 1.9 0.19 — Tawfik et al.c,69

69 — — 0.06 — Uemura and Harada64

— — 0.43 0.45 1.49 Almeida70

— — 0.24 0.26 0.61 Almeida70

90–125 2.5 2.17 0.10 0.09 Tandukar et al.66

90–100 1.3–4.0 2.03–3.15 0.27–0.4d 0.07–0.30 Tandukar et al.22

38–135 [82] 1.3–11.7 [3.1] 0.24–6.80 [2.03] 0.06–0.45 [0.19] 0.05–1.49 [0.07] Typical range: flowsheet with Sset[median]

— 2.0–2.7 [2.4] 0.82–1.60 [1.47] 0.09–0.38 [0.28] 0.04–0.50 [0.05] Typical range: flowsheet without Sset[median]

a kg VSS kgremoved−1 COD. b kg BOD msponge

−3 d−1. c DHS system treating gray water (similar to a concentrated sewage). d kg VSS kgremoved−1

BOD.


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




COD for aerobic processes (activated sludge, rotating biologi-cal contactors, submerged aerated biofilters, high-rate trick-ling filters) following UASB reactors.13,14,66,71 Moreover, thetypical yield of excess sludge from the conventionalsuspended growth process (i.e. activated sludge) ranges from0.4 to 1.7 kg TSS kgremoved

−1 COD.13

Such results could support the elimination of secondarysettlers, an important advancement towards the simplificationof UASB/SBTF systems, as schematically shown in Fig. 5. Stud-ies on the demonstration scale (equivalent population: 300–500 inhab) indicate that the conditions to design a UASB-trickling filter system without secondary settlers are compati-ble with the organic loadings required to improve nitrificationin the post-treatment step.25 In this case, the use of sponge-based support media tends to significantly improve the reli-ability of the system complying with discharge standards.26

Thus, for less stringent discharge standards and low skilledpersonnel for operation, the design of UASB/SBTF systemswithout secondary settlers is a very promising alternative.

In order to effectively implement such a proposed flow-sheet simplification, the aforementioned design criteria (sec-tion 3.1) should be taken into account. In addition, the efflu-ent TSSUASB concentration should be kept below 100 mg L−1

to avoid solid overload in the post-treatment step.20,24 Thus,for the operation of UASB-sponge-bed trickling filters withoutsecondary clarifiers the anaerobic sludge management needsto be very established in order to avoid solids washout. ForUASB/SBTF systems without secondary settlers, the typicaloverall removal efficiencies obtained are 88–97% for BOD,80–87% for COD, 78–91% for TSS, and 44–95% for NH4

+-N.20,24,65 The corresponding effluent concentrations are lessthan 90 mg L−1 for COD, less than 40 mg L−1 for BOD andTSS, and less than 20 mg L−1 for NH4

+-N. These averagevalues tend to meet the environmental discharge standardsin developing countries (e.g. in Brazil).

UASB/SBTF systems without secondary settlers could bemade even more compact by implementing the UASB reactorand the SBTF in a single treatment module. Such an optionis especially suitable either for sewage treatment in smallcommunities or as a decentralized option in densely popu-lated regions, since its area requirement is less than 0.05 m2

per inhabitant. For the management of the solid phase (i.e.anaerobic excess sludge), simplified dewatering units (i.e.drying beds) can be implemented.

4.2. Mainstream nitrogen removal considering theheterotrophic denitrification or anammox process

The effective and reliable establishment of simultaneous re-moval of residual organic carbon and nitrogen in the post-treatment step preceded by UASB reactors is currently contex-tualized as a possibility.72–74 Promising strategies have beenreported in the literature considering two different metabolicpathways for nitrogen removal: denitrification and/oranammox process. The main operational aspects of each pro-cess are summarized in Table 8.

4.2.1. N-removal via heterotrophic denitrification. Nitro-gen removal in fully ventilated sponge-bed trickling filters(SBTFs) without effluent recirculation is typically between 25and 35%.22,33 Such conditions have also been observed byAlmeida et al.20 and Bressani-Ribeiro,24 in which an increasein organic loadings led to greater N-removal. It could be asso-ciated with a substrate input (residual carbon from the UASBreactor) which may increase the activity of heterotrophs inanoxic zones of the biofilm in the presence of oxidized formsof nitrogen. When applying final effluent recirculation,N-removal could be enhanced up to values between 60 and65%,40,72 since it promotes the contact of residual carbonfrom the UASB reactor with the nitrate produced at the nitri-fying portion of the SBTF. N-removal efficiencies around 75%

Fig. 5 UASB/SBTF system without secondary settlers. Reproduced and adapted from the work of von Sperling and Chernicharo3 with permissionfrom IWA Publishing, Copyright 2005.


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




Tab

le8

Nitrogen

removalinSB

TFs:heterotrophic

den

itrific

ationan

dan

ammoxproce

ss

Referen

ceEffluen

tHRT(h)

Influen

tto

the

SBTF(m

gL−

1)

Nload

ings

(kgN

m−3

d−1)

Nremoval

(%)

Mainaspe

cts

App

lied

Rem

oved

Con

vention

alnitrogenremoval

through

(autotrop

hic)nitrification

andheterotroph

icden

itrification

Machdar

etal.31

Anaerobic(U

ASB

reactor)

2.0

COD:1

61—

—25

–31

•System

designed

forCan

dNH

4+-N

removal

post-UASB

reactor

Pilotscale

TN:5

1Onod

eraet

al.28

Anaerobic(U

ASB

reactor)

3.2

COD:1

130.23

0.07

30•DHSpo

st-UASB

reactor(w

ithou

trecirculation

).Heterotroph

icden

itrification

was

associated

with

biom

assdecay

(endog

enou

sresp

iration)

TN:3

0Pilotscale

Alm

eidaet

al.20

Anaerobic(U

ASB

reactor)

2.0

COD:2

000.44

0.11

–0.31

25–70

•System

designed

forCan

dNH

4+-N

removal

post-UASB

reactor.Higher

performan

cefor

N-rem

oval

withincreasedOLR

orefflue

nt

recirculation

TN:4

5Dem

oscale

Iked

aet

al.72

Synthetic

12.0

(4/reactor)

COD:6

00–180

01.00

0.20

–0.60

20–60

•Unus

ualconfigu

ration

forsewag

etreatm

ent,

dueto

thehighHRT

Pilotscale

TN:5

00Bun

dyet

al.56

Pre-settledsewag

e0.6–1.2

COD:1

70–220

——

40–74

•Reactor

withan

aerobicvolumefollow

edby

asu

bmergedan

oxic

compa

rtmen

tan

defflue

nt

recirculation

.Partof

thepre-settledsewag

ewas

sentto

thean

oxic

compa

rtmen

t

Pilotscale

TN:4

0–48

Onod

eraet

al.40

Anaerobic(U

ASB

reactor)

1.5

COD:1

770.42

0.33

65•Fu

ll-scale

DHSsystem

post-UASB

reactor,with

efflue

ntrecirculation

(1:1)

TN:2

6Fu

llscale

Innovativenitrogenremoval

pred

ominan

tlyba

sedon

thean

ammox

conversion

Chua

nget

al.76

Synthetic

0.7–2.0

COD:n

onap

plied

1.94

–2.98

1.84

–2.01

68–95

•Ope

ration

withan

ammox

process(w

ith

recirculation

),receivingNH

4+-N

andNO2− -N

Pilotscale

TN:4

0Sá

nch

ezGuillénet

al.73

Synthetic

1.14

–2.23

COD:n

onap

plied

2.10

–2.15

1.52

–1.60

74–78

•Ope

ration

withan

ammox

process(w

ith

recirculation

),receivingNH

4+-N

andNO2− -N

TN:1

00Pilotscale

Sánch

ezGuillénet

al.77

Synthetic

1.71

–2.96

COD:n

onap

plied

1.68

–0.95

0.77

–0.88

52–54

•Ope

ration

withpa

rtialnitritation

and

anam

mox

process(w

ithou

trecirculation

)TN:1

00–1

10Pilotscale


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




have been reported, considering the by-pass of a fraction ofpre-settled sewage (which contains readily biodegradableCOD) to anoxic compartments of SBTFs.56 The heterotrophicdenitrification was reported to be the predominant processin all of these studies. Considering the inner sponge anoxicconditions,31 as well as the typical low F/M ratio (0.032 gCOD g VSS d−1) and long sludge retention time (∼100 d), theendogenous respiration might be related to heterotrophicdenitrification within the DHS system.22,30 Thus, the use ofan additional source of organic matter by heterotrophs inanoxic zones could also contribute to the explanation how asponge-bed trickling filter produces effluents with low con-centrations of solids and even operates without secondarysettlers,28 as previously discussed by Tandukar et al.22 How-ever, because of the anoxic zones within the sponges andlonger SRT, the hypothesis that the activity of anaerobic am-monium oxidizing bacteria (anammox bacteria) was a factorfor nitrogen removal cannot be eliminated, as discussed byAlmeida et al.20

4.2.2. N-removal via the anammox process. The partialnitritation in SBTFs under controlled ventilation conditionswas proposed by Chuang et al.75 The aim was to further pro-vide a suitable environment for anammox bacteria coloniza-tion. Higher N-removal efficiencies were obtained (70–95%)and effluent recirculation (1 : 3) was applied to ensure propersupport media wetting.76 Aiming at establishing the cultiva-tion of anammox bacteria in SBTFs, Sánchez Guillén et al.73

developed a study following a methodological approach simi-lar to that used by Chuang et al.76 In this case, the perfor-mance for N-removal remained around 75–80% for tempera-tures between 20–30 °C (effluent recirculation 1 : 1). Takinginto account these promising results, a research study aimingat establishing simultaneous partial nitritation–anammoxwas developed.77 Nevertheless, the N-removal was limited to54%. In this proof of concept experiment, uncontrolled oxy-gen supply from passive aeration was probably a factor forthe lower TN removal efficiency. Therefore, it seems that inthis case ammonium oxidizers could not out-compete thenitrite oxidizers leading to a worst performance in terms ofTN removal. Regardless of such performance obtained bySánchez Guillén et al.,77 the study clearly showed the poten-tial of sponge-bed trickling filters for TN removal via theanammox process.

5. Conclusions and perspectives

The replacement of primary settlers by UASB reactors in thetechnological flowsheet of trickling filters for sewage treat-ment has brought remarkable advantages, mainly in terms ofconstruction simplification and operational requirements as-sociated with sludge handling. Following this important step,improving the performance of trickling filters by means ofimplementing sponge-based support media has shown to bean interesting strategy. In this case, a better system perfor-mance is associated with the greater biomass retention andlonger hydraulic retention time compared to those of conven-

tional rock and plastic-bed trickling filters. Additionally, self-structured support media (e.g. Spongepacking) can furthercontribute to simplification of construction, operation andmaintenance of SBTFs.

The design of sponge-bed trickling filters following UASBreactors for simultaneous removal of residual carbon andammonium should be currently based on practical experi-ences, due to the uncertainties regarding operational condi-tions and the system's performance. Therefore, adoptingloading criteria (0.20–0.40 kg BOD mreactor

−3 d−1), the hydrau-lic retention time needs to be verified for the usual range (1.5to 2.5 h). In terms of predicting effluent BOD concentrationsor BOD removal efficiencies for TFs post-UASB reactors, theavailable models still have to be properly adjusted to allow di-rect application. In this case, the Eckenfelder model could befurther improved taking into account the data from full-scaletrickling filters following UASB reactors.

Future efforts aimed at improving the UASB/SBTF technol-ogy should consider the heterotrophic denitrification as apossible strategy for N removal, considering final effluentrecirculation. On the other hand, since the sponge-based me-dia tend to increase the SRT to more than 100 days, the useof the anammox process might be a promising alternative, ifthe interaction of heterotrophic and autotrophic microorgan-isms can be managed by simply controlling the oxygen sup-ply within the SBTF.

Additionally, to the best of our knowledge, the conditionsfor UASB/SBTF operating without secondary settlers are notyet fully established, mainly because it requires rigoroussludge management control in the anaerobic reactor. Hence,low anaerobic reactor performance, typically ascribed to pooroperation and management, can jeopardize the advantages ofthe integrated UASB/SBTF system, especially considering theproposed simplified flowsheet, in which secondary settlersare not implemented.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors acknowledge the support obtained from theGhent University Special Research Fund (BOF UGent –

Funding for joint doctorate) and from the following Brazilianinstitutions: Conselho Nacional de Desenvolvimento Científicoe Tecnológico – CNPq; Fundação de Amparo à Pesquisa deMinas Gerais – FAPEMIG; Instituto Nacional de Ciência eTecnologia em Estações Sustentáveis de Tratamento de Esgoto– INCT ETEs Sustentáveis.

References

1 C. A. L. Chernicharo, J. B. Van Lier, A. Noyola and T.Bressani-Ribeiro, Anaerobic sewage treatment: state of theart, constraints and challenges, Rev. Environ. Sci. Bio/Technol., 2015, 14(4), 649–679.


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




2 E. Petropoulos, J. Dolfing, R. J. Davenport, E. J. Bowen andT. P. Curtis, Developing cold-adapted biomass for the anaer-obic treatment of domestic wastewater at low temperatures(4, 8 and 15°C) with inocula from cold environments, WaterRes., 2017, 112, 100–109.

3 M. von Sperling and C. A. L. Chernicharo, BiologicalWastewater Treatment in Warm Climate Regions, IWAPublishing, London, 2005.

4 C. A. L. Chernicharo, Post-Treatment Options for theAnaerobic Treatment of Domestic Wastewater, Rev. Environ.Sci. Bio/Technol., 2006, 5(1), 73–92.

5 E. Foresti, M. Zaiat and M. Vallero, Anaerobic processes as thecore technology for sustainable domestic wastewatertreatment: Consolidated applications, new trends, perspectives,and challenges, Rev. Environ. Sci. Bio/Technol., 2006, 5, 3–19.

6 Y. I. Chan, M. F. Chong, C. L. Law and D. G. Hassell, Areview on anaerobic–aerobic treatment of industrial andmunicipal wastewater, Chem. Eng. J., 2009, 155, 1–18.

7 G. Kassab, M. Halalsheh, A. Klapwijk, M. Fayyad and J. B.van Lier, Sequential anaerobic-aerobic treatment for domes-tic wastewater - a review, Bioresour. Technol., 2010, 101(10),3299–3310.

8 A. A. Khan, R. Z. Gaur, V. K. Tyagi, A. Khursheed, B. Lew, I.Mehrotra and A. A. Kazmi, Sustainable options of posttreatment of UASB effluent treating sewage: A review,Resour., Conserv. Recycl., 2011, 55(12), 1232–1251.

9 S. Chong, T. K. Sen, A. Kayaalp and H. M. Ang, Theperformance enhancements of upflow anaerobic sludgeblanket (UASB) reactors for domestic sludge treatment-astate-of-the-art review, Water Res., 2012, 46(11), 3434–3470.

10 C. A. L. Chernicharo, T. Bressani-Ribeiro, G. B. Garcia, A.Lermontov, C. B. Pereira, C. J. Platzer, G. R. C. Possetti,M. A. L. Leites and R. Rosseto, Panorama do tratamento deesgoto sanitário nas regiões Sul, Sudeste e Centro-Oeste doBrasil: tecnologias mais empregadas (Overview of sewagetreatment in the South, Southeast and Midwest regions ofBrazil : most used technologies), Revista DAE, 2018, vol.213(66), pp. 5–19, In Portuguese.

11 B. Rittmann and P. McCarty, Environmental Biotechnology:Principles and Applications, McGraw-hill InternationalEditions, New York, 2001.

12 E. P. Jordão and C. A. Pessoa, Tratamento de EsgotosDomésticos (Domestic sewage treatment), AssociaçãoBrasileira de Engenharia Sanitária – ABES, Rio de Janeiro,4th edn, 2005, in Portuguese.

13 Metcalf & Eddy Inc., Wastewater engineering: treatment andreuse, McGraw-Hill, New York NY, 5th edn, 2013.

14 R. F. Gonçalves, C. A. L. Chernicharo, C. O. Andrade Neto, P.Alem Sobrinho, M. T. Kato, R. H. R. Costa, M. M. Aisse andM. Zaiat, Pós-tratamento de efluentes de reatores anaeróbiospor reatores com biofilme (Post-treatment of Anaerobic efflu-ents with bioflim reactors), in Pós-tratamento de efluentes dereatores anaeróbios. FINEP/PROSAB, ed. Chernicharo, C. A.L., Rio de Janeiro, Brasil, 2001, in Portuguese.

15 SNIS - Sistema Nacional de Informações sobre Saneamento:Diagnóstico dos Serviços de Água e Esgotos – 2015 (Diagno-

sis of water and sanitation services – 2015), http://www.snis.gov.br/diagnostico-agua-e-esgotos/diagnostico-ae-2015(accessed January 2018), in Portuguese.

16 G. T. Daigger and J. P. Boltz, Trickling filter and tricklingfilter-Suspended Growth Process Design and Operation: AState-of-the-Art Review, Water Environ. Res., 2011, 83(5),388–404.

17 C. P. L. Grady, G. T. Daigger and H. D. Lim, BiologicalWastewater Treatment, 2ed, revised and expanded, MarcelDekker, New York, 1999.

18 P. G. S. Almeida, S. C. Oliveira and C. A. L. Chernicharo,Operation of trickling filters post-UASB reactors without thesecondary sedimentation stage, Eng. Sanit. Ambiental,2011, 16(2), 1–10, in Portuguese.

19 L. K. Agrawal, H. Harada and H. Okui, Treatment of rawsewage in a temperate climate using a UASB reactor and thehanging sponge cubes process, Water Sci. Technol.,1997, 36(6–7), 433–440.

20 P. G. S. Almeida, A. K. Marcus, B. E. Rittmann and C. A. L.Chernicharo, Performance of plastic- and sponge-basedtrickling filters treating effluents from an UASB reactor,Water Sci. Technol., 2013, 67(5), 1034–1042.

21 E. F. A. Mac Conell, P. G. S. Almeida, K. E. L. Martins, J. C.Araújo and C. A. L. Chernicharo, Bacterial communityinvolved in the nitrogen cycle in a down-flow sponge-basedtrickling filter treating UASB effluent, Water Sci. Technol.,2015, 72(1), 116–122.

22 M. Tandukar, I. Machdar, S. Uemura, A. Ohashi and H.Harada, Potential of a Combination of UASB and DHSReactors a Novel Sewage Treatment System for DevelopingCountries: Long-Term Evaluation, Water Sci. Technol.,2006, 53(3), 209–218.

23 K. Kubota, M. Hayashi, K. Matsunaga, A. Iguchi, A. Ohashi,Y. LI, T. Yamaguchi and H. Harada, Microbial communitycomposition of a down-flow hanging sponge (DHS) reactorcombined with an up-flow anaerobic sludge blanket (UASB)reactor for the treatment of municipal sewage, Bioresour.Technol., 2014, 151, 144–150.

24 T. Bressani-Ribeiro, E. M. F. Brandt, P. G. S. Almeida,C. A. D. Flórez and C. A. L. Chernicharo, Technologicalimprovements in compact UASB/SBTF systems fordecentralized sewage treatment in developing countries,Desalin. Water Treat., 2017, 91, 112–120.

25 P. G. S. Almeida, C. A. L. Chernicharo and C. L. Souza,Development of compact UASB-TF systems for the treatmentof domestic wastewater in small communities in Brazil,Water Sci. Technol., 2009, 59(7), 1431–1439.

26 P. G. S. Almeida, PhD Thesis, Remoção de matéria orgânica enitrogênio em filtros biológicos percoladores aplicados aopós-tratamento de efluentes de reatores UASB (Carbon andnitrogen removal in trickling filters post-UASB reactor),Universidade Federal de Minas Gerais, 2012, in Portuguese.

27 T. Bressani-Ribeiro, MSc dissertation, Sistema UASB/FBPsubmetido a hidrograma típico de vazão: avaliação do meiosuporte baseado em espuma de poliuretano e operação semdecantadores secundários (UASB/TF system submitted to a


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM


http://www.snis.gov.br/diagnostico-agua-e-esgotos/diagnostico-ae-2015

http://www.snis.gov.br/diagnostico-agua-e-esgotos/diagnostico-ae-2015



typical hydrograph flow: assessment of a sponge-based pack-ing media and operation without secondary settlers),Universidade Federal de Minas Gerais, 2015, in Portuguese.

28 T. Onodera, K. Matsunaga, K. Kubota, R. Taniguchi, H.Harada, K. Syutsubo, T. Okubo, S. Uemura, N. Araki, M.Yamada, M. Yamauchi and T. Yamaguchi, Characterizationof the retained sludge in a down-flow hanging sponge (DHS)reactor with emphasis on its low excess sludge production,Bioresour. Technol., 2013, 136, 169–175.

29 M. Tandukar, S. Uemura, A. Ohashi and H. Harada,Combining UASB and the “fourth generation” down-flowhanging sponge reactor for municipal wastewater treatment,J. Environ. Eng., 2006, 132(2), 166–172.

30 N. Araki, A. Ohashi, I. Machdar and H. Harada, Behaviour ofnitrifiers in a novel biofilm reactor employing hangingsponge-cubes as attachment site, Water Sci. Technol.,1999, 39(7), 23–31.

31 I. Machdar, H. Harada, A. Ohashi, Y. Sekiguchi, H. Okui andK. Ueki, Combination of UASB reactor and curtains typeDHS (downflow hanging sponge) reactor as a cost effectivesewage treatment system for developing countries, Water Sci.Technol., 2000, 42(3–4), 83–88.

32 T. L. G. Hendrickx, H. Temmink, H. J. H. Elissen and C. J. N.Buisman, The effect of operating conditions on aquaticworms eating waste sludge, Water Res., 2009, 43(4), 943–950.

33 T. Onodera, M. Tandukar, D. Sugiyana, S. Uemura, A.Ohashi and H. Harada, Development of a sixth-generationdown-flow hanging sponge (DHS) reactor using rigid spongemedia for post-treatment of UASB treating municipal sew-age, Bioresour. Technol., 2014, 152, 93–100.

34 T. Onodera, K. Syutsubo, W. Yoochatchaval, H. Sumino, M.Mizuochi and H. Harada, Protection of biomass from snailovergrazing in a trickling filter using sponge media as abiomass carrier: down-flow hanging sponge system, WaterSci. Technol., 2015, 71(4), 518–523.

35 T. Okubo, T. Onodera, S. Uemura, T. Yamaguchi, A. Ohashiand H. Harada, On-site evaluation of the performance of afull-scale down-flow hanging sponge reactor as a post-treatment process of an up-flow anaerobic sludge blanket re-actor for treating sewage in India, Bioresour. Technol.,2015, 194, 156–164.

36 WEF, Water Environment Federation, Operation of municipalwastewater treatment plants. Manual of Practice n° 10, v. 2,3rd edn, 1996.

37 WEF, Water Environment Federation, Design of MunicipalWastewater Treatment Plants. v.1, 3ª ed. Manual of Practicen° 8, cap 12, Alexandria, Virginia, 1992, p. 2600.

38 WEF, Water Environment Federation, Aerobic Fixed-GrowthReactors: a special publication, 2000, p. 340.

39 C. A. L. Chernicharo, Pós-tratamento de efluentes dereatores anaeróbios (Post-treatment of Anaerobic effluents).PROSAB/FINEP, 1st edn, Segrac, Rio de Janeiro, 2001, inPortuguese.

40 T. Onodera, T. Okubo, S. Uemura, T. Yamaguchi, A. Ohashiand H. Harada, Long-term performance evaluation of down-flow hanging sponge reactor regarding nitrification in a full-

scale experiment in India, Bioresour. Technol., 2016, 204,177–184.

41 A. Tawfik, F. El-Gohary, A. Ohashi and H. Harada, Theinfluence of physical–chemical and biological factors on theremoval of faecal coliform through down-flow hangingsponge (DHS) system treating UASB reactor effluent, WaterRes., 2006, 40, 1877–1883.

42 J. A. R. Victoria, PhD Thesis, Filtro biológico aeróbio-anóxicopara remoção de nitrogênio de efluentes de reatores UASB(Aerobic-anoxic biofilter for nitrogen removal from UASB reac-tor effluents), Universidade de São Paulo, 2006, in Portuguese.

43 P. Pearce and S. Jarvis, Operational experiences withstructured plastic media filters: 10 years on, Water Environ.J., 2011, 25(2), 200–207.

44 M. F. Fonseca, MSc Dissertation, Remoção de nitrogênioamoniacal em filtro biológico percolador pós-tratandoefluente de reator UASB (Ammoniun-N removal in a tricklingfilter post-treating UASB reactor effluent), Universidade Fed-eral do Rio de Janeiro, 2009, in Portuguese.

45 E. C. Frade, MSc Dissertation, Concepção e Avaliação deDesempenho de um Sistema Compacto Reator UASB/FiltroBiológico Percolador Utilizado para o Tratamento de EsgotosSanitário (Design and evaluation of a compact UASB/TFsystem for sewage treatment), Universidade Federal deMinas Gerais, 2003, in Portuguese.

46 M. M. Aisse, PhD Thesis, Tratamento de efluentes de reatoresanaeróbios (Treatment of anaerobic reactor effluents),Universidade de São Paulo, 2002, in Portuguese.

47 P. P. Pontes and C. A. L. Chernicharo, Characterization andremoval of specific organic constituents in an UASB-tricklingfilter system treating domestic wastewater, Environ. Technol.,2011, 32(3), 281–287.

48 O. J. S. Moraes, J. R. Souza, L. R. Silva, S. G. Azevedo, C. A. L.Chernicharo, L. C. S. Lobato and R. V. Silva, Long termperformance of the largest Brazilian combined anaerobic/aerobic treatment plant (PE = 1 million inhabitants), inProceedings of the X DAAL - Latin American Workshop andSymposium on Anaerobic Digestion, Ouro Preto, 2011.

49 C. Collivignarelli, G. Urbini, A. Farneti, A. Bassetti and U.Barbaresi, Anaerobic-Aerobic treatment of municipalwastewater with full-scale upflow anaerobic sludge blanketand attached biofilm reactors, Water Sci. Technol., 1990, 22(1–2), 475–482.

50 C. A. L. Chernicharo, L. C. Lobato, C. A. P. Júnior, E. V.Barbosa, J. R. Souza, L. R. Silva, M. A. M. Judice, O. J. Moraes,P. G. Almeida and T. C. F. Gonçalves, Secondary sludge returnfor thickening and digestion in UASB reactors: Case study ofONÇA STP – Brazil. Proceedings of XI Taller y SimposioLatinoamericano de Digestión Anaerobia, Cuba, 2014.

51 L. C. S. Lobato, L. C. Cardoso, M. G. Divino, H. Silva andC. A. L. Chernicharo, Influência da rotina operacional dedescarte do lodo excedente de reatores UASB na qualidadedo efluente de uma estação de tratamento de esgoto(Influence of operational routine of excess UASB sludgedischarge on effluent quality), in Proceedings of 26°Congresso Brasileiro de Engenharia Sanitária e Ambiental,


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




ABES, Porto Alegre - RS, 2011, in Portuguese.52 I. Machdar, H. Harada, A. Ohashi, Y. Sekiguchi, H. Okui and

K. Ueki, A novel and cost-effective sewage treatment systemconsisting of UASB pre-treatment and aerobic post-treatmentunits for developing countries, Water Sci. Technol.,1997, 36(12), 189–197.

53 A. Tawfik, A. Ohashi and H. Harada, Effect of spongevolume on the performance of down-flow hanging spongesystem treating UASB reactor effluent, Water Sci. Technol.,2008, 58(1), 185–194.

54 E. F. Mac Conell, P. G. S. Almeida, A. M. Zerbini, E. M. F.Brandt, J. C. Araújo and C. A. L. Chernicharo, Diversity anddynamics of ammonia-oxidizing bacterial communities in asponge-based trickling filter treating effluent from a UASBreactor, Water Sci. Technol., 2013, 68(3), 650–657.

55 A. Tawfik, F. El-Gohary, A. Ohashi and H. Harada,Optimization of the performance of an integrated anaerobic–aerobic system for domestic wastewater treatment, BioprocessBiosyst. Eng., 2010, 33, 779–785.

56 C. A. Bundy, D. Wu, M. Jong, S. T. Edwards, Z. S. Ahammadand D. W. Graham, Enhanced denitrification in DownflowHanging Sponge reactors for decentralised domesticwastewater treatment, Bioresour. Technol., 2017, 226, 1–8.

57 T. Okubo, K. Kubota, T. Yamaguchi, S. Uemura and H.Harada, Development of a new non-aeration-based sewagetreatment technology: Performance evaluation of a full-scaledown-flow hanging sponge reactor employing third-generation sponge carriers, Water Res., 2016, 102, 138–146.

58 W. Yoochatchaval, T. Onodera, H. Sumino, T. Yamaguchi,M. Mizuochi, T. Okadera and K. Syutsubo, Development of adown-flow hanging sponge reactor for the treatment of lowstrength sewage, Water Sci. Technol., 2014, 70(4), 656–663.

59 H. Tanaka, M. Takahashi, Y. Yoneyama, K. Syutsubo, K. Kato,A. Nagano, T. Yamaguchi and H. Harada, Energy saving systemwith high effluent quality for municipal sewage treatment byUASB–DHS, Water Sci. Technol., 2012, 66(6), 1186–1194.

60 S. Uemura, S. Suzuki, Y. Matuyama and H. Harada, Directtreatment of settled sewage by DHS reactors with differentsizes of sponge support media, Int. J. Environ. Res.,2012, 6(1), 25–32.

61 M. Mahmoud, A. Tawfik and F. El-Gohary, Use of down-flowhanging sponge (DHS) reactor as a promising post-treatmentsystem for municipal wastewater, Chem. Eng. J., 2011, 168,535–543.

62 M. Takahashi, A. Ohya, S. Kawakami, Y. Yoneyama, T.Onodera, K. Syutsubo, S. Yamazaki, N. Araki, A. Ohashi, H.Harada and T. Yamaguchi, Evaluation of treatmentcharacteristics and sludge properties in a UASB reactortreating municipal sewage at ambient temperature, Int. J.Environ. Res., 2011, 5(4), 821–826.

63 M. Mahmoud, A. Tawfik and F. El-Gohary, Simultaneousorganic and nutrient removal in a naturally ventilatedbiotower treating pre-settled municipal wastewater,J. Environ. Eng., 2010, 136, 301–307.

64 S. Uemura and H. Harada, Application of UASB technologyfor sewage treatment with a novel post treatment process, in

Environmental Anaerobic Technology; Applications and NewDevelopments, ed. H. H. P. Fang, Imperial College Press,Londres, 2010, p. 404.

65 C. A. L. Chernicharo and P. G. Almeida, Feasibility of UASB/trickling filter systems without final clarifiers for thetreatment of domestic wastewater in small communities inBrazil, Water Sci. Technol., 2011, 64(6), 1347–1354.

66 M. Tandukar, A. Ohashi and H. Harada, Performancecomparison of a pilot-scale UASB and DHS system and acti-vated sludge process for the treatment of municipal waste-water, Water Res., 2007, 41(12), 2697–2705.

67 M. Tandukar, S. Uemura, I. Machdar, A. Ohashi and H.Harada, A low-cost municipal sewage treatment system witha combination of UASB and the “fourth-generation”downflow hanging sponge reactors, Water Sci. Technol.,2005, 52(1–2), 323–329.

68 Water Environment Federation (WEF), Design of MunicipalWastewater Treatment Plants: WEF Manual of Practice No.8 ASCE Manuals and Reports on Engineering Practice No. 76,McGraw-Hill Professional, 5th edn, 2010, TRICKLINGFILTERS, Chapter Access Engineering.

69 A. Tawfik, R. A. Wahab, A. Al-Asmer and F. Matary, Effect ofhydraulic retention time on the performance of down-flowhanging sponge system treating grey wastewater, BioprocessBiosyst. Eng., 2011, 34, 767–776.

70 P. G. S. Almeida, MSc dissertation, Efeito de diferentes tiposde meio suporte no desempenho de filtros biológicospercoladores aplicados ao pós-tratamento de efluentes dereatores UASB, com ênfase na nitrificação (Effect of differenttypes of support media in trickling filters performanceApplied to post-treat UASB reactors effluents, focusing on ni-trification), Universidade Federal de Minas Gerais, 2007, inPortuguese.

71 R. F. Gonçalves, V. L. D. Araújo and C. A. L. Chernicharo,Association of a UASB reactor and a submerged aeratedbiofilter for domestic sewage treatment, Water Sci. Technol.,1998, 38(8–9), 189–195.

72 N. Ikeda, T. Natori, T. Okubo, A. Sugo, M. Aoki, M. Kimura,T. Yamaguchi, H. Harada, A. Ohashi and S. Uemura,Enhancement of denitrification in a DHS reactor by effluentrecirculation, Water Sci. Technol., 2013, 68(3), 591–598.

73 J. A. Sánchez -Guillén, P. R. C. Guardado, C. M. L. Vazquez,L. C. M. Cruz, D. Brdjanovic and J. B. van Lier, Anammoxcultivation in a closed sponge-bed trickling filter, Bioresour.Technol., 2015, 186, 252–260.

74 C. D. Leal, A. D. Pereira, F. T. Nunes, L. O. Ferreira, A. C.Coelho, S. K. Bicalho, E. F. A. Mac Conell, T. Bressani-Ribeiro, C. A. L. Chernicharo and J. C. Araújo, Anammox fornitrogen removal from anaerobically pre-treated municipalwastewater: Effect of COD/N ratios on process performanceand bacterial community structure, Bioresour. Technol.,2016, 211, 257–266.

75 H. P. Chuang, A. Ohashi, H. Imachi, M. Tandukar and H.Harada, Effective partial nitrification to nitrite by down-flowhanging sponge reactor under limited oxygen condition,Water Res., 2007, 41, 295–302.


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM




76 H. P. Chuang, T. Yamaguchi, H. Harada and A. Ohashi,Anoxic ammonium oxidation by application of a down-flowHanging sponge (DHS) reactor, J. Environ. Eng. Manage.,2008, 18(6), 409–417.

77 J. A. Sánchez-Guillén, L. K. M. C. B. Jayawardana, C. M. LopezVazquez, L. M. Oliveira Cruz, D. Brdjanovic and J. B. van Lier,Autotrophic nitrogen removal over nitrite in a sponge-bedtrickling filter, Bioresour. Technol., 2015, 187, 314–325.


Publ

ishe

d on

14

Sept

embe

r 20

18. D

ownl

oade

d by

642

730

on 1

0/27

/201

8 5:

56:3

3 PM



environmental science water research & technologyevolcke/pdf/2018 bressaniribeiro... ·...

Documents