Sei sulla pagina 1di 14

e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/ecoleng

Anaerobic digesters as a pretreatment


for constructed wetlands

J.A. Alvarez , I. Ruz, M. Soto


Department of Physical Chemistry and Chemical Engineering I, Campus A Zapateira, 15008,
Faculty of Science, University of A Coruna, A Coruna, Spain

a r t i c l e i n f o a b s t r a c t

Article history: The most commonly used pretreatment technologies for constructed wetland (CW) treat-
Received 11 June 2007 ment of domestic sewage are septic tanks (ST) and Imhoff tanks (IT). These technologies
Received in revised form have frequently suffered from failures and even in normal operation they offer insufcient
25 January 2008 removal of solids. As a result, combined ST-CW or IT-CW can experience substrate clogging,
Accepted 17 February 2008 especially when high organic loads are applied. In the last 7 years, the operation of combined
systems using high-rate anaerobic digesters as a pretreatment and CW as a post-treatment
has been reported. A review of the literature indicates that CW in these combined sys-
Keywords: tems operates with a similar organic loading rate (on a chemical oxygen demand basis) but
Anaerobic digesters with a lower total suspended solid (TSS) loading rate. In these combined systems, the TSS
Constructed wetlands loading rate is 3050% less than that applied in CW combined with classical pretreatment
Municipal wastewater technologies. A low TSS loading rate could prevent substrate clogging in CW.
Clogging This work presents the results of different case studies on the treatment of municipal
wastewater with high-rate anaerobic systems. Our interest is focused on the capacity of
these systems for removing suspended solids, and therefore on their potential as an appro-
priate pretreatment to avoid clogging in constructed wetlands and to improve efciency.
Average and 95 percentile TSS concentrations of anaerobic treated wastewater were below
60 and 100 mg/l, respectively, for all congurations. Therefore, the use of high rate anaer-
obic systems as a pretreatment for constructed wetlands could delay gravel bed clogging.
Furthermore, according to the level of organic matter removal, anaerobic pretreatment pro-
vided a 3060% reduction in the required wetland area. Both treatment alternatives can
be combined to develop low-cost, robust, and long-term systems for treating municipal
wastewater.
2008 Elsevier B.V. All rights reserved.

1. Introduction structed wetlands are treatment systems with a very small


energy input, low operational cost, and low surplus sludge
Sustainability of sanitation systems should be related to low generation (Sperling, 1996; Kadlec et al., 2000; Lens et al.,
cost and low energy consumption and, in some situations, 2001; Hoffmann et al., 2002). These characteristics, together
low mechanical technology requirements. Decentralised and with low technological requirements, make them particu-
low-cost processes are considered to be a better choice for larly suitable for decentralised wastewater treatment in rural
rural areas (Lens et al., 2001). Anaerobic digesters and con- areas.


Corresponding author. Tel.: +34 981 563100x16016; fax: +34 981528050.
E-mail address: alvarezr@usc.es (J.A. Alvarez).
0925-8574/$ see front matter 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.ecoleng.2008.02.001
e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467 55

The costs of construction, installation, and operation of municipal wastewater that will later be treated in constructed
anaerobic digesters are lower than those of conventional aero- wetlands. First, a brief analysis of clogging phenomena in CW
bic units because anaerobic digesters do not require expensive is presented, and the pretreatment technologies most often
equipment for process maintenance and control. In fact, if the used in combination with CW are discussed, focusing on their
environmental conditions inside the digester are adequate, potential for reducing the quantity of suspended solids intro-
anaerobic processes are mainly self-controlled. Additionally, duced into constructed wetlands. Next, the authors review
the production of excess sludge is minimal, and energy bal- the literature on systems combining anaerobic digesters and
ances are quite favourable, even when heating is required, due CW. Finally, detailed case studies on anaerobic pretreatment
to the production of methane (Foresti, 2002). of municipal wastewater are presented, focusing on the ef-
The disadvantage of anaerobic digesters is that additional ciency of suspended solid removal and the potential of anaer-
treatment is necessary to polish and lower the pollution load. obic digesters for preventing clogging and reducing CW area.
Even in tropical regions (Sousa et al., 2001), and mainly in cool
to temperate climate regions (Alvarez et al., 2003), the efuent
of UASB (up ow anaerobic sludge blanket) systems requires 2. Substrate clogging in constructed
an efuent post-treatment to reduce organic mater, nutri- wetlands
ents, and pathogenic microorganisms. In the case of operating
temperatures below 20 C, UASB systems are good at remov- Substrate clogging encompasses several processes that lead to
ing suspended solids; however, acetic acid accumulation in a reduction of the inltration capacity of the gravel bed after
the efuent reduces the COD (chemical oxygen demand) and several years of operation. In horizontal ow (HF) wetlands,
BOD (biological oxygen demand) removal efciencies (Alvarez, apparent clogging and subsequent ponding near the inlet
2004; Alvarez et al., 2006). of the treatment cells dampen the remarkable performance
It is of great interest to combine wetland systems with of the system. This may occur after few years of operation
anaerobic digesters in order to obtain sufcient treatment ef- (Dahab and Surampalli, 2001; Caselles-Osorio et al., 2007). In
ciency. The most commonly used anaerobic technology for vertical ow (VF) wetlands, clogging of the substrate matrix
municipal wastewater treatment is the UASB (Lettinga, 2001; critically hinders the oxygen transport and therefore results
Foresti et al., 2006; Van Haandel et al., 2006). There are sev- in an extremely rapid failure of the systems ability to treat
eral studies of systems combining anaerobic pretreatment and wastewater (Langergraber et al., 2003).
constructed wetlands, which are assessed in Section 4. UASB The main parameters that inuence the substrate clogging
reactors are the referent pretreatment anaerobic technology process are the organic load and the suspended solid load.
used in these combined systems. However, other anaerobic Besides these main factors, the clogging risk is also controlled
technologies may be used as sewage pretreatment for con- by gravel size, since large gravel prevents or delays clogging
structed wetlands. The hydrolytic upow sludge bed reactor phenomena (Chazarenc and Merlin, 2005; Zhao et al., 2004).
(HUSB) is a promising alternative. Organic load is an indirect parameter leading to sludge pro-
However, constructed wetlands (CW) are land-intensive duction derived from bacterial growth. Both inuent sludge
treatment systems. The use of an appropriate anaerobic pre- and sludge generated in situ will accumulate in the gravel
treatment before constructed wetland treatment can reduce bed. Literature values for the maximal acceptable organic
the construction cost by about 3640%, due to the fact that load fall within a wide range. For example, Winter and Goetz
anaerobic treatment reduces the inuent organic matter and (2003) indicated the area of VF constructed wetlands should
therefore the area required for CW is decreased (Barros and be designed for a maximum loading rate of 20 gCOD/m2 d to
Soto, 2002). Both anaerobic and wetland treatment approaches avoid the clogging process. So, the clogging risk becomes a
are characterized by low construction and operation costs, low limitation of wetland performance.
excess sludge, and low energy demand. Therefore, both treat- On the other hand, one of the major parameters inuenc-
ment technologies are complementary and highly sustainable. ing clogging is the suspended solid load (Batchelor and Loots,
Limited organic removal efciency in anaerobic digesters 1997; Dahab and Surampalli, 2001; Winter and Goetz, 2003;
is compensated by high efciency in CW, while anaerobic Langergraber et al., 2003). Little information is available con-
digesters present minimal area requirements, generally less cerning the maximum acceptable TSS loading rates. Values
than 0.1 m2 /p.e. for UASB (Kivaisi, 2001). given are only valid for one special type of substrate and can-
Studies have shown that one of the most important oper- not be used as a general guideline. For example, Dahab and
ational handicaps of constructed wetlands is gravel bed Surampalli (2001) found clogging in a subsurface horizontal
clogging; this may occur after several years, resulting from ow constructed wetland system after 3.5 years of treating
the treatment of raw or poorly pretreated urban wastewa- wastewater with a load of 1.44 gTSS/m2 d. Winter and Goetz
ter. Suspended solids that are not removed in a pretreatment (2003) showed that in order to avoid clogging processes in
system are effectively removed by ltration and settlement a vertically constructed wetland, the average concentration
within the rst few metres beyond the inlet zone. Thus, a of TSS in the inow should not exceed 100 mg/l, while the
high level of total suspended solid (TSS) removal in anaer- suspended solid load should not exceed 5 gTSS/m2 d. These
obic pretreatment would contribute to avoiding or reducing authors thought that growth of biomass has only a minor
wetland clogging problems, reinforcing constructed wetland effect on clogging compared to the accumulation of inuent
sustainability (Vymazal, 2005; Caselles-Osorio et al., 2007). TSS.
The aim of this work is to analyse and discuss the poten- Green et al. (2006) compared two types of pretreatments:
tial of high-rate anaerobic digesters as a pretreatment for a UASB system and a primary decanter. They indicated that
56 e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467

by using UASB efuent as feeding water for a VF CW a higher ventional and natural low cost treatment approaches (Metcalf
removal rate could be achieved than by using primary decanter and Eddy, 2003). However, information about the operation
efuent, as a consequence of the relatively low TSS loading and efciency of pretreatment systems combined with CW
rate resulting from the higher removal of TSS in the UASB. is scarce. Even in many scientic reports, the TSS concentra-
These authors found that the total TSS removed in each active tion entering the CW system is not available, in contrast to
cycle (until clogging occurred) was similar for the VF CW that the frequent statement that the inuent concentration and
received either pre-settled domestic wastewater or UASB efu- loading rate of TSS are the main factors that inuence clog-
ent, while the total COD removed was about three times higher ging.
for the VF CW receiving UASB efuent. Therefore, it seems Classical sewage pretreatment technologies include a sep-
that the TSS loading rate was the most inuential parameter tic tank and Imhoff tank for small-scale installations. These
affecting the rate of bed clogging in VF CW (Green et al., 2006). systems can achieve a TSS removal of 5070%, generating pri-
Caselles-Osorio and Garcia (2007) compared the physico- mary efuent concentrations in the range of 5090 mgTSS/l
chemical pretreatment and primary settling for constructed when they are operated well (Metcalf and Eddy, 2003). Fur-
wetlands. Physico-chemical pretreatment reduced the COD to thermore, septic and Imhoff tanks stabilize the sludge by
48% and turbidity to 17% that of primary settled wastewater. anaerobic digestion, reducing the amount of sludge gener-
After 8 months of operation at similar hydraulic loading rates, ated. Another classical pretreatment alternative, which is
it was observed that the hydraulic conductivity decreased used mainly for larger installations, is the primary decanter.
by 20% in the subsurface ow (SSF) CW fed with settled Primary decanters offer similar TSS removal of 5070%, but
wastewater. The authors estimated that the physico-chemical the high amount of primary sludge produced is their largest
pretreatment extended the lifespan of the constructed wet- handicap (Metcalf and Eddy, 2003). Physico-chemical treat-
land by approximately 10 years, compared to a primary ment (coagulation and occulation followed by clarication)
decanter pretreatment. is an advanced pretreatment for domestic sewage, reach-
The effect of the inuent type (dissolved glucose or par- ing up to 90% TSS removal and 80% COD (Metcalf and Eddy,
ticulate starch) on the efciency of SSF CW was reported by 2003). However, physico-chemical pretreatment also has cer-
Caselles-Osorio and Garca (2006). The type of organic mat- tain requirements that can make this process unsuitable in
ter did not appear to inuence the COD removal efciency. the context of constructed wetlands technology; these include
However, ammonia nitrogen removal was higher in the sys- the cost of the coagulants, energy for adding and mixing coag-
tem fed with glucose than in the one fed with starch. Hydraulic ulants, and increased sludge handling (Caselles-Osorio and
conductivity was lower near the inlet of the SSF CW fed with Garcia, 2007).
glucose, despite the possible retention and accumulation of Until now, the most common wastewater pretreatments
starch particles near the inlet of the other SSF CW. The authors for CW have been the septic tank (ST) or the Imhoff tank (IT).
hypothesized the growth and development of biolm was When properly operated, ST and IT offer good pretreatment
greater in the system fed with glucose than in the system fed levels, reaching low TSS concentrations (Neralla et al., 2000;
with starch, since glucose is a readily biodegradable carbon Vymazal, 2002). However, ST and IT frequently suffer from
source. Therefore, the biolm growth could be an important failures that decreased the treatment efciency (Philippi et al.,
parameter in the evaluation of clogging phenomena, as these 1999; Mbuligwe, 2004; Caselles-Osorio et al., 2007).
authors indicated. A recent survey indicates that 86% of the constructed wet-
It is generally accepted that the application of a good land plants in operation in Spain use a septic tank or Imhoff
wastewater pretreatment is essential for sustainable, long- tank for pretreatment (Puigagut et al., 2007). This was observed
term operation of subsurface ow constructed wetlands in spite of the fact that the majority of these CW were built
(Vymazal et al., 1998; USEPA, 2000; Vymazal, 2002; Caselles- within the last 5 or 6 years. A report of recently built CW sys-
Osorio et al., 2007). On the other hand, although VF CW tems in Italy also indicated the use of Imhoff tanks (Masi et
can directly treat raw domestic wastewater (Chazarenc and al., 2006). The situation is similar in most countries where CW
Merlin, 2005), several authors also recommended wastewa- systems are being used. In the case of the Czech Republic,
ter pretreatment (Winter and Goetz, 2003; Langergraber et al., pretreatment for a small system usually consists of a septic or
2003; Green et al., 2006). settling tank, while pretreatment for larger systems usually
consists of an Imhoff tank (Vymazal, 2002). Settling tanks are
used also in Flanders (Rousseau et al., 2004a) and Denmark
3. Pretreatment alternatives for constructed (Brix and Arias, 2005).
wetlands A summary of data on wastewater pretreatment for con-
structed wetlands is presented in Table 1. The average primary
The main objective of pretreatment or primary treatment is treatment efuent concentration of SS in Czech Republic CW
the reduction of suspended solids in wastewater, although systems is 65 mg/l, while the average mass-loading rate is
additional treatment effects leading to organic content reduc- 3.6 gTSS/m2 d (Vymazal, 2002; n = 42). Data for Denmark and
tion and, in some cases, the hydrolysis and stabilization of the UK (n = 77), North America (n = 34), and Poland (n = 6),
the generated sludge are obtained. In this way, some pretreat- and the Czech Republic, indicate that the average inu-
ment technologies can reach up to 50% COD or BOD removal. ent concentration to CW after pretreatment ranges from 48
Furthermore, from a general point of view, pretreament opera- to 173 mgTSS/l and average loading rates range from 3.6
tions are considered to be a convenient means of ensuring the to 5.2 gTSS/m2 d (Vymazal, 2002). Vymazal (2005) reported
correct operation of subsequent treatment steps in both con- worldwide gures for CW, indicating an average inuent TSS
e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467 57

Table 1 Efuent concentration and efciency of TSS removal for domestic sewage pretreatment systems combined with
CW
N TSS (mg/l) TSSr (%) Reference

Septic and Imhoff tanks


Primary sedimentation (to VF CW) 1 240416 Green et al. (2006)
Septic tank (to SSF CW) 8 26114 Neralla et al. (2000)
Septic tank (to VF CW, single-house) 3 85124 Brix and Arias (2005)
Septic tank (to SSF CW) 4 90517 (261)a 35.2 Caselles-Osorio et al. (2007)
Settling pond (to FWS CW) 12 5200 (25)b Rousseau et al. (2004a)
Settling tank (to VF CW) 7 131000 (80)b Rousseau et al. (2004a)
Settling tank (to HF CW) 2 10400 (47)b Rousseau et al. (2004a)
Septic tank or Imhoff Tank (to SSF CW) 3 173 Puigagut et al. (2007)
Septic tank or Imhoff Tank (to SSF CW) 42 65 Vymazal (2002)
Imhoff Tank (to SSF CW) 1 146 73.0 Caselles-Osorio et al. (2007)
Imhoff Tank (to HF or VF CW) 3 2676 Masi et al. (2006)
Range (Average) 261000 (123)

High-rate anaerobic digesters


UASB (to VF CW) 1 124 52 Green et al. (2006)
UASB (to SSF and FWS CW) 1 59 66.5 El-Khateeb and El-Gohary (2003)
UASB (to SSF CW) 1 189 El-Hamouri et al. (2007)
UASB (to SSF CW) 1 3442 8291 Barros et al. (2006)
UASB (to SSF CW) 1 3874 (52)a 4978 (65)a Ruz et al. (2006)
Range (Average) 34189 (92) 5291 (68)

a
Range followed by the average. N is the number of studies included.
b
Range followed by 50% percentile.

concentration of 107 mg/l and an average TSS loading rate of cated that UASB removes about 73% of inuent TSS (average
5.4 g/m2 d. inuent TSS of 241 mg/l, efuent TSS of 65 mg/l, n = 127 lab
For Spanish CW-based treatment systems, TSS loading and eld applications, temperature of 21.6 C, HRT (hydraulic
rates range from 3 to 17 gTSS/m2 d (n = 6), and the aver- retention time) of 8.5 h. However, mean values for perfor-
age primary treatment efuent concentration is 173 mg/l mance of eld-only applications of UASB were lower (inuent
(n = 3) (Puigagut et al., 2007). These authors highlight the TSS of 301 mg/l, efuent TSS of 102 mg/l, n = 22, temperature
scarcity of data about TSS loading rates and inuent con- of 23.8 C, HRT of 6.9 h). This could be due to the fact that
centrations, as they surveyed a total of 39 SSF systems but UASB eld applications mainly correspond to tropical coun-
only found information on TSS for a few of these systems. tries where wastewater concentration is high. Furthermore,
Also in Spain, recent research conducted on several Catalo- higher temperatures in these countries lead to higher biogas
nian SSF CW systems (Caselles-Osorio et al., 2007) reports production that in turn increases sludge washout. However,
primary efuent from septic tanks and Imhoff tanks con- UASB offers an advanced wastewater pretreatment, which
taining 90517 mgSS-COD/l (average and standard deviation reaches about 62% COD removal and 68% BOD removal, lev-
of 238 172 mgSS-COD/l; n = 5). These SS-COD values indicate els that are maintained in eld applications. In addition, UASB
higher TSS concentrations. The authors indicated that in some systems generate very small amounts of sludge and applied
cases, septic tanks used as pretreatment systems were not HRTs are lower than those of some primary treatments such
working properly. Estimated TSS loading rates for Catalonian as septic tank or ponds.
SSF CW systems (Caselles-Osorio et al., 2007) are in the range Different congurations of anaerobic digesters had been
of 2.610 gTSS/m2 d, and are higher than the ranges indicated studied in order to treat municipal wastewater in both cold
above for other countries. and warm regions. In Section 5, some of these congurations
As indicated, high-rate anaerobic digesters have become are analysed, with special attention given to the solid removal
an alternative for sewage treatment in regions with a warm capability of anaerobic systems.
climate. As a consequence, in recent years CW systems have
been applied in some occasions as a post-treatment pro-
cess for anaerobically pretreated sewage. Section 4 deals the 4. CW post-treatment of anaerobically
operation of CW treating anaerobic efuents, while Table 1 treated sewage
summarizes available data about TSS in UASB efuents fed
to CW systems. Data from Table 1 indicate a somewhat low
Table 2 shows the main design and operating characteris-
TSS concentration in UASB efuents when compared to sep-
tics of various constructed wetlands for UASB-CW combined
tic and Imhoff tank efuents. However, it is not possible to
systems found in the literature. A dozen UASB-CW appli-
make a denitive comparison due to the scarcity of data for
cations were described, although there is only information
UASB-CW combined systems.
about inuent TSS for a few systems, as can be seen by com-
A general review of high-rate anaerobic digesters treating
paring Tables 2 and 1. In addition, the operational period
municipal wastewater (Alvarez et al., in preparation) indi-
reported in these studies is not long enough (the maximum
58
Table 2 Operation of constructed wetlands using anaerobic technology as a pretreatment
Ref.a Systemb Plant OLR (g/m2 d) (in CW sub-units system) CW system efciency (%, except FC: Log units)

TRH (d) COD TSS BOD N-NH4 TN TP COD TSS BOD N-NH4 TN TP FC

1 UASB + SSF Juncus spp. 5 1.32 0.16 82.9 65.0 70.0 70.3 89.0 3.4
2 UASB + SSF Juncus spp. 7 1.89 0.23 81.3 56.2 44.8 47.4 64.2 3.9

e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467
3 UASB + SSF Juncus spp. 10 2.64 0.32 81.6 48.0 45.1 48.0 61.1 4.5
4 UASB + SSF Juncus spp. 10 6.6 2.0 0.17 81.7 70.3 69.3 50.0 4.0
5 UASB + SSF Juncus spp. 7 9.5 2.0 0.25 76.7 66.0 44.3 47.3 3.7
6 UASB + SSF T. latifolia 5 5.513.5 1.43.3 1.74.7 0.71.8 1.33.1 0.080.19 78.0 79.7 77.8 27.3 27.3 38.2 3.3
7 UASB + SF T. latifolia 10.8 5.513.5 1.43.3 1.74.7 0.71.8 1.33.1 0.080.19 69.7 55.9 79.8 51.5 41.8 29.4 3.3
8 UASB + SSF Ph. Mauritianus 1.9 12.3 2.4 5661 2327 0.6
T. latifolia
9 UASB + SSF T. Latifolia 1.2 14.5 3.9 0.75 7580 7475 6975
Colocasia esculenta
10 UASB + VF(3x) Ph. australis 0.6 779.0 183.0 333.9 82.2 91.3 95.1
11 UASB + VF(2) + SSF Ph. australis 0.4 + 1.6 73.7 17.3 31.6 82.2 91.3 95.1
12 UASB(2) + SSF + SF Juncus spp. 5.0 7.7 1.7 5.2 1.2 2.3 0.2 7283 3252 7887 38 32 26 2.1
13 UASB + FSF + SF Juncus spp. 2.4 16.5 5.0 10.2 70.5 77.3 73
14 UASB(2) + SSF Ph. Australis
Arundo donax 0.54 130.1 64.1 74.6 21.4 21.4 3.7 7882 7980 7982 89 811 15.0 1.0
15 AT + SSF Z.b. and T.sc 1.5 4.5 1.0 71.4 86.1 95.8 90.4 93.3 100%
16 AT + SSF Z.b. and T.sc 0.75 9.0 2.0 37.5 46.1 55.2 56.6 20.9 89.7%

a
References: (1,2,3) Sousa et al. (2001), (4,5) Sousa et al. (2003), (6,7) El-Khateeb and El-Gohary (2003), (8) Kaseva (2004), (9) Mbuligwe (2004), (10,11) Green et al. (2006), (12) Barros et al. (2006), (13) Ruz
et al. (2006), (14) El-Hamouri et al. (2007), (15,16) Da Motta Marques et al., 2001.
b
System description: UASB (Upow Anaerobic Sludge Bed), SSF (Horizontal Subsurface ow constructed wetland), VF (Vertical ow constructed wetland), SF (Horizontal Surface ow constructed
wetland), and AT (Anaerobic treatment not specied). Referred units were connected in series, the number in parentheses indicates several units of the same type in the series.
c
Zizaniopsis bonariensis and Typha subalata.
e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467 59

Table 3 Comparison of the loading rate and efciency for CW treatment of efuents from UASB and from classical
pretreatment technologies
BOD5 COD TSS TP TN NH4 + -N

Worldwide experiment SSFa


Loading rate (g/m2 d) 3.9 12.0 5.4 0.39 1.76 1.06
Efciency (%) 81 71 78 32 39 34

UASB-SSF combined systemsb


Loading rate (g/m2 d) 5.5 10.8 2.9 0.49 3.01 2.02
Efciency (%) 78 73 63 54 53 53

a
Vymazal (2005), n = 66131.
b
This review: mean values obtained from data in Table 2, except for experiments 10, 11, and 14 (n = 413).

operation period was 3 years) to conclude whether anaerobic TN, and NH4 + -N) are higher for CW in UASB-SSF combined
pretreatment can prevent gravel bed clogging. Furthermore, systems, which generally also have higher nutrient removal
information about solid accumulation or hydraulic conduc- efciencies. This behaviour is in accordance with the fact that
tivity evolution in constructed wetlands combined with UASB UASB efciently removes organic mater and suspended solids,
is not included in referred bibliography. but UASB is primarily a nutrient conservative process. There-
In general, the performance of the systems is satisfactory fore, in UASB-SSF combined systems, CW will have a lower
with high removal efciencies for organic matter, suspended TSS inuent concentration but a higher nutrient concentra-
solids, nutrients and pathogens, reaching mean values (S.D.) tion. The removal of faecal coliforms has a range of 14 log
of 74 (12)% COD, 68 (17)% TSS, 83 (9)% BOD, 49 (22)% units and is clearly inuenced by the HRT applied.
TN (total nitrogen), 51 (26)% TP (total phosphorous), and 94 El-Hamouri et al. (2007) reported higher loading rates of
(13)% FC (data obtained from Table 2). These efciency val- 131 gCOD/m2 d and 64 gTSS/m2 d for a SSF CW fed with the
ues are close to those found in the literature (Vymazal, 2002; efuent from a two-step UASB system. The SSF used by
Rousseau et al., 2004a; Puigagut et al., 2007) for SSF CW treat- El-Hamouri et al. (2007) had a high depth (0.8 m) and the
ing primary settled efuents. Planted beds generally perform system reached low nutrients removal, indicating only sec-
better than unplanted ones (El-Khateeb and El-Gohary, 2003; ondary treatment objectives. Furthermore, the reported period
Sousa et al., 2003; Mbuligwe, 2004; Kaseva, 2004; El-Hamouri of operation was short (6 months) and there is no informa-
et al., 2007). Da Motta Marques et al. (2001) found that plants tion on the sustainability of this highly loaded SSF CW. Even
improve constructed wetland efciency only under high load- higher organic loading rate values were reached when UASB
ing rates. No signicant differences in efciency between efuents were treated in VF CW or in combined systems that
macrophyte species were found in UASB-CW systems treating included VF CW units (Green et al., 2006). A system includ-
domestic sewage, except in some restricted cases. ing a UASB followed by two VF CWs and one SSF CW reached
The organic load rate for horizontal ow constructed wet- a high secondary treatment efciency that had a small foot-
lands varies from 5 to 20 (mean value of 11.4) gCOD/m2 d print, equivalent to 0.9 m2 per person. An even lower footprint
and from 1.4 to 3.3 (mean value of 3.0) gTSS/m2 d, when the of 0.13 m2 per person equivalent was achieved for a scheme
study from El-Hamouri et al. (2007) is excluded. In general, that included a UASB followed by three VF CWs (Green et al.,
organic loading rates on a COD basis are similar to those 2006).
reported for SSF CW operating in several European countries
while loading rates of suspended solids are lower. As indi-
cated, Vymazal (2002) reported organic loading rates in the 5. Anaerobic congurations as CW
range of 8.612.7 gCOD/m2 d for the Czech Republic, Denmark, pretreatment: case studies
Poland, and Slovenia, and TSS loading rates in the range of
3.65.2 gTSS/m2 d for the Czech Republic, Denmark, UK, North
5.1. Anaerobic digestion processes and up ow
America, and Poland. Vymazal (2005) reported worldwide
anaerobic digesters
data, including data from Australia, Austria, Brazil, Canada,
the Czech Republic, Denmark, Germany, India, Mexico, New
The UASB reactor is the most commonly used anaerobic tech-
Zealand, Poland, Slovenia, Sweden, the USA, and the UK.
nology for domestic sewage treatment; and the hydrolytic
Table 3 compares mean worldwide values reported by
upow sludge bed (HUSB) is an option to be considered. These
Vymazal (2005) and mean values obtained from studies
digesters have similar design features, but are primarily differ-
included in Table 2 to UASB-SSF (or SF) combined systems.
entiated by their operational conditions. Both UASB and HUSB
Although the number of examples for UASB-SSF combined
can be operated as a single unit or as a combined two-step or
systems is scarce, results suggest similar organic loading rates
hybrid system (see Fig. 1).
and lower TSS loading rates for CW combined with UASB
In upow mode reactors like the HUSB and UASB, raw or
pretreatment. So, UASB reactors reduce the suspended solid
pretreated wastewater enters the bottom of the digester and
loading rate from 30 to 50% compared to classical pretreat-
goes up until it reaches the solidliquidgas (SLG) separator,
ment technologies. COD removal efciency is similar while
if it exists, and nally reaches the exit level. Sedimentation,
TSS removal efciency is lower. Nutrient loading rates (TP,
ltration, and absorption processes enable suspended solids
60 e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467

Fig. 1 Schematic representation of anaerobic systems used for laboratory, pilot, and eld scale applications (note that
Table 4 indicates which of these congurations are tested at lab, pilot, or full scale experiments). Abbreviations: UASB,
Upow Anaerobic Sludge Bed; HUSB, Hydrolytic Upow Sludge Bed; CMSS, Completely Mixed Sludge Stabilization digester;
I, Inuent; E, Efuent; G, Biogas; S, Sludge.

to be retained inside the digester, resulting in the sludge bed. of methanogenic activity in the rst step and hydrolysis in the
Because of this, suspended solids and absorbable organic mat- second step.
ter contained in wastewater have a longer solid retention time On the other hand, the anaerobic process may be stopped
(SRT) than the liquid fraction (HRT), allowing particulate mat- in the rst phase as a function of environmental and oper-
ter to be totally or partially biodegraded. In properly designed ational conditions. In this case, the one-step system will be
systems, the pass of the inuent through the sludge in up called an anaerobic hydrolytic pretreatment. The well-known
ow digesters improves contact between organic substrates UASB system is the most commonly used design for anaero-
and biomass, enhancing digester performance. bic methanogenic treatment of domestic sewage. A digester
Depending on operational conditions, the sludge held in design similar to the UASB, when used under hydrolytic (non-
the digester can reach the SLG separator and, eventually, methanogenic) conditions, is known as a HUSB reactor.
the exit level. In order to avoid the presence of great amounts The type of substrate, inuent concentration, temperature,
of suspended solids in the digester efuent, purges must be HRT, and SRT are the main operational parameters that dene
periodically practiced from a point slightly below the SLG the methanogenic or nonmethanogenic conditions. Domes-
separator or at an equivalent point. The frequency of this tic sewage is a complex substrate with only a small fraction
purge is highly variable, from once a week in the case of high of readily degradable matter in anaerobic conditions, making
load HUSB systems to a yearly purge or no purge in the case hydrolysis the limiting step of the overall process in many
of low load UASB methanogenic systems. In the case of HUSB cases. Inuent concentration and the applied HRT determine
systems, additional purges may be necessary in order to main- the maximum achievable SRT, although the actual SRT may
tain the SRT at an appropriate value, as indicated below. be reduced through a sludge purge (Alvarez et al., 2006). Lower
The anaerobic degradation process takes place in two inuent concentration and lower HRT lead to a lower SRT.
main sequential phases. Particulate organic and soluble Temperature determines the minimum required SRT for
polymers should rst be hydrolysed and subsequently acid- methanogenic conditions. Methanogenic digesters operating
ied to volatile fatty acids (known as acidogenic phase, at 1320 C need a minimum SRT of 80 and 50 d (Henze et al.,
or hydrolyticacidogenic phase). The process can continue 1995). In this way, Zeeman and Lettinga (1999) postulated that
through acetic acid generation from other volatile fatty acids a SRT higher than 75 d would be required for a UASB treating
and through methane generation from acetic acid and hydro- municipal wastewater at 15 C.
gen (known as the methanogenic phase). The overall process With dilute or very dilute sewage, the maximum achievable
for the anaerobic digestion of complex substrates may be SRT of an UASB may be equal to or below the minimum SRT
performed either in a single unit system (only one digester, required for methanogenesis. In this case, the methanogenic
single-step system) or in two separated units (two digesters processes is partial and volatile fatty acids (mainly acetic acid)
connected in series, two-step system). In two-step systems, accumulate in the efuent of the digester. In any case, the SRT
the rst step mainly deals with the substrate hydrolysis and may be reduced through a sludge purge to reduce methano-
acidication and the second step involves the acetogenic genesis and to reach predominantly hydrolyticacidogenic
and methanogenic process. However, many two-step systems conditions. In practice, hydrolytic conditions are established
respond to a partial phase separation, showing the presence by applying a low HRT and practicing an additional sludge
Table 4 Summary of the results obtained in anaerobic systems for municipal wastewater treatment
Expa Systema Volume (l) Days (samples)b T ( C) HRT (h) SRT (d) Vup (m/h) XR (gVSS/l) Efuent pH Inuent (mg/l)c Removal (%)d

COD TSS COD TSS

Single-step HUSB systems (hydrolytic pretreatment)


1 HUSBlab 2 524 (147) 20 2.24.5 1429 0.11 1015 7.38 624 (255806) 220 (110312) 3347 5568

e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467
2 HUSBpilot run 1 25500 53 (23) 1920 35 11.4 1.43 6.7 7.26 438 (291594) 294 (111605) 4659 8285
3 HUSBpilot run 2 25500 495 (250) 1320 35 22.4 1.30 11.1 7.14 282 (53565) 173 (19438) 2944 4476

Single-step UASB systems (anaerobic treatment)


4 UASBlab 2 155 (42) 20 516 33215 0.04 9.2 7.00 685 (276926) 214 (116336) 5376 6385
5 UASBpilot run 1 25500 69 (34) 1415 1011 88 0.49 11.4 7.15 282 (118451) 187 (93418) 5458 7685
6 UASBpilot run 2 25500 54 (28) 14 11 57 0.48 4.2 6.98 169 (34300) 107 (44173) 3545 5065
7 UASBpilot run 3 25500 70 (34) 2021 4.75.6 38 1.02 10.7 7.15 339 (227480) 207 (92430) 4749 8182
8 UASB (eld) 3600 325 (32) 518 17 4692 0.19 36 7.10 1354 (802700) 799 (502100) 5893 8296

UASB-CMSS systems (anaerobic treatment)


9 UASB-CMSS lab 21.6 95 (42) 20 67 3375 0.06 14.8 6.83 644 (220967) 237 (152424) 7181 7788
10 UASB-CMSS pilot 2550020000 80 (41) 1516 69 82.5 0.70 8.45.9 7.27 321 (182451) 175 (112252) 4653 6379

Two-step systems (anaerobic treatment)


11 HUSB + UASB pilot run 1 25500 + 20000 97 (42) 1421 35 + 714e 2883e 1.260.50e 11.78.3e 7.14 251 (70540) 166 (49359) 4965 8189
12 HUSB + UASB pilot run 2 25500 + 20000 60 (35) 1620 34 + 69e 2171e 1.270.52e 12.68.5e 7.21 367 (213565) 202 (99361) 5965 8689
13 UASB + UASB (eld) 3600 + 3600 252 (29) 718 24 + 24e 387 0.120.12e 4.46.5e 7.24 352 (200800) 354 (50400) 4565 7590

a
For system description, see also Fig. 1. Experiments: (1) Ligero et al. (2001a); (2) Alvarez et al. (2003); (3) Alvarez (2004); (4) Ruz et al. (1998); (5, 6, 7) Alvarez et al. (2006); (8) Barros and Soto (2002); (9)
Ruz et al. (1998); (10) Alvarez et al. (2004); (11) Alvarez et al. (2007); (12) Alvarez (2004); (13) Barros and Soto (2004).
b
Reported operation period in days, the number of samples analysed is in parentheses.
c
The average is followed by the minimum and maximum values in parentheses.
d
Removal range obtained from average removal values that corresponded to periods of different operation conditions.
e
Values corresponding to the rst and second step units (in two-step systems), respectively.

61
62 e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467

purge if necessary. In this way, a lower HRT and a lower SRT tem (Barros and Soto, 2004) consisting of two UASB, each with
differentiate HUSB from UASB systems. 3.6 m3 of active volume was used. These digesters did not have
a solidliquidgas separator.
5.2. Description of surveyed anaerobic systems Analytical methods were carried out according to Standard
methods (1995), as previously detailed (Ruz et al., 1998;
Fig. 1 shows the different anaerobic digester congurations Alvarez et al., 2006). Sampling frequency of inuent and efu-
analysed in this section, which include laboratory, pilot- ent varied from once a week for eld scale applications to four
and eld-scale applications of anaerobic digesters, single or ve times a week for pilot and lab scale digesters. The mon-
and two-step systems, and hydrolytic and methanogenic itoring period varied from 53 to 495 days depending on the
operation conditions. All applications were carried out in system considered.
Galiza, in northwest Spain. Attention has been paid to the
removal efciency and efuent concentration of suspended
5.3. Operation and efciency of anaerobic systems
solids.
The main characteristics of these systems are described
Table 4 summarises the results of the different anaerobic
below, while a detailed explanation is available in the ref-
systems studied, and includes the main design and oper-
erences indicated. All water line digesters, i.e., all digesters
ation variables such as the HRT, SRT, upow velocity, and
used except the Completely Mixed Sludge Stabilization (CMSS)
biomass concentration. The operation and efciency of these
digester, operated in an upow mode.
systems has been described in detail elsewhere (see refer-
A 2 l active volume digester was operated on a laboratory
ences in Table 4). Upow velocity of the different systems
scale UASB at a HRT of 516 h (Ruz et al., 1998). In a second
surveyed is determined by design characteristics, digester
study, the UASB reactor was operated in combination with
size, and HRT applied. Design characteristics and HRT com-
a 1.6 l active volume CMSS digester (Ruz et al., 1998). The
bined with wastewater characteristics also determined the
CMSS digester was fed with sludge drawn from UASB and an
SRT and the biomass concentration (XR ) obtained. However,
equal volume of the CMSS digester content was returned to
in some operation periods of examples 3 and 12, the SRT of
the bottom of the UASB. The CMSS digester was mechanically
the HUSB system was intentionally reduced via an additional
stirred and in a thermostat-controlled bath that was 35 C.
sludge purge. As indicated in Table 4, SRT was highly variable,
Finally, the same digester was operated as a HUSB at a HRT
while the biomass concentration was generally between 8 and
of 2.24.5 h (Ligero et al., 2001a,b). In this case, the digester
15 gVSS/l (volatile suspended solids). Lower biomass concen-
was equipped with an internal recirculation system. These
trations were registered in some cases, either when very dilute
laboratory digesters were fed with raw domestic wastewater
wastewater was treated (experiment 6) or in very low-load
collected from the main sewer of the city of A Coruna (Ruz et
digesters (experiments 8 and 13).
al., 2007).
In this paper, we carried out a comparative study of
An anaerobic pilot plant was located at the municipal
different systems focusing on the TSS removal efciency
wastewater treatment facility of Santiago de Compostela, and
and efuent quality. As indicated above, this aspect is of
it was fed with raw domestic wastewater from this city. This
great importance in preventing clogging phenomena in post-
plant had a 25.5 m3 active volume and it could treat municipal
treatment wetlands. For this purpose, original data on the
wastewater from a population of about 200300 inhabitants
inuent and efuent were used.
when operated in methanogenic conditions or about 500800
Attention is also focused on other design and operation
inhabitants when operated as a hydrolytic pretreatment. This
variables like COD removal efciency, efuent pH, biomass
pilot plant was successively operated as methanogenic UASB
activity, and surplus sludge generation. Other parameters, like
system (runs 1, 2, and 3) at a HRT from 5 to 11 h (Alvarez et
alkalinity, pathogens, fat, and oil were not measured in most
al., 2006), and as a hydrolytic HUSB reactor at a HRT of 35 h
of the research described. Pathogen removal has scarcely been
(Alvarez et al., 2003).
considered in anaerobic digesters treating municipal wastew-
In another study, this UASB was coupled with a CMSS
ater, and generally this aspect is not considered in monitoring
digester that had 20 m3 of active volume. This system is named
anaerobic digesters, although helminthic eggs were reported
as the UASB-CMSS pilot plant. In this conguration, the UASB
to be completely eliminated in UASB (Lettinga et al., 1993). In a
was operated at a HRT of 69 h and the CMSS digester at a HRT
combined UASB-CW system treating the efuent from a small
of 1627 d and 3035 C (Alvarez et al., 2004). The overall HRT
rural community, anaerobic digesters removed less than 0.5
was in the range of 10.716.1 h.
log units of faecal coliforms, while the overall system removed
A two-step pilot plant was also studied (Alvarez et al.,
about 2.0 log units.
2007), and consisted of a hydrolyticacidogenic reactor (HUSB,
25.5 m3 ) followed by a methanogenic unit (UASB, 20 m3 ). Both
digesters had a similar design, and were differentiated by their 5.3.1. Single-step HUSB systems
operating conditions. The HRT ranged from 3 to 5 h for HUSB The anaerobic hydrolysis of wastewater is a promising
and from 6 to 14 h for UASB. pretreatment with the following advantages (Wang, 1994;
A eld application of the anaerobic digester was carried out Goncalves et al., 1994; Ligero et al., 2001a,b): (a) it removes
in order to treat domestic wastewater from a small community an high percentage of SS; (b) it totally or partially stabilises
of about 30 inhabitants (Beariz, Ourense). The operation of a the sludge; and (c) it increases the biodegradability of the
single-step UASB with 3.6 m3 of active volume was checked remaining COD. The latter advantage favours the subsequent
(Barros and Soto, 2002). In a second study, a two-step sys- biological elimination of nutrients (N, P).
e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467 63

In laboratory-scale experiments with the HUSB, optimum


results were obtained at a HRT of 2.3 h (see experiment 1
in Table 4). Over 60% of inuent SS were retained in the
digester and hydrolysed. On the other hand, a pilot plant-scale
HUSB reactor treating diluted wastewater at 20 C removed
more than 82% of TSS (experiment 2). Most of the solids
removed (above 81%) were eliminated by hydrolysis. In con-
trast, at lower temperatures (1315 C), the TSS retention and
hydrolysis decreased (experiment 3). Furthermore, the HUSB
digester removed COD in an extension that varied from 30 to
60%.
The process is self-controlled in relation to operational
parameters like pH, biomass concentration, and activity. Since
acidication is a faster process than hydrolysis, the result of
hydrolytic pretreatment is the generation of VFA that reduces Fig. 2 Average efuent TSS concentration (mg/l) in the
the pH in both the sludge bed and the digester efuent. The anaerobic system studied. Bars and whiskers represent
sludge bed pH is in the range from 5.5 to 7, which is lower average values and standard deviations, respectively.
than inuent and efuent pH. The HUSB efuent contained Number of data is shown at the bottom of the columns. To
acetic acid in a range of concentrations that varied from 60 identity system, see Table 4.
to 110 mg/l, and the efuent pH was generally 0.21.0 units
lower than the inuent pH. Sludge held in HUSB reactors
showed residual methanogenic activity ranging from 0.01 to Efuent pH was generally 0.10.3 units lower than the inuent
0.02 gCH4 -COD/gVSSd, indicating partial separation of anaer- pH. Surplus organic sludge generation ranged from 0 to 30%
obic phases. A lower SRT may be reached by an additional of inuent VSS, depending mainly on organic and hydraulic
purge, which in turn enhances phase separation through loading and SRT. For example, no generation of surplus sludge
a lower biomass concentration and methanogenic activity. was found in experiments 6 and 8 (Table 4). In contrast, sur-
However, lower SRT also reduces the suspended solid hydrol- plus sludge reached 20% for inuent VSS in experiment 5 and
ysis and increases the surplus sludge generation. Previous up to 31% in experiment 7.
research (Alvarez, 2004) demonstrated that inuent wastewa- At temperatures of 20 C and particularly at tempera-
ter strength strongly inuences the overall efciency (percent tures lower than 15 C, single-step methanogenic process had
TSS, COD, and BOD removal) and acidication efciency (VFA some difculties caused by low hydrolysis rates of inuent
generation) of the HUSB reactor, while temperature only suspended solids, which accumulated in the digesters displac-
appreciably inuences the acidication efciency. The inu- ing the active methanogenic biomass (Zeeman and Lettinga,
ence of operational parameters such as HRT, SRT, and sludge 1999). Therefore, signicant amounts of volatile fatty acids
concentration on the behaviour of the HUSB system is not well remained in the digester efuent. For example, this occurred
established (Alvarez, 2004). Further research on this subject is in experiment 7, working at a low HRT of 45 h, when about
still necessary. 75 mgVFA-COD/l were registered in the treated efuent. Thus,
at a low temperature, a higher HRT must be applied in single-
5.3.2. Single-step UASB systems step UASB systems, as experiments 5, 6, and 8 described (see
The laboratory UASB system, which treats domestic waste- Table 4).
water at 20 C and HRT of 16 h, reached COD and TSS removal
efciencies of 76 and 85%, respectively. An important effect of 5.3.3. UASB-CMSS systems
HRT on the removal efciency was observed, since at the same The main aim of the CMSS digester, combined with the UASB,
temperature and 5 h of HRT, removal efciencies decreased was to enhance the biodegradation of inuent solids retained
to 53% of COD and 63% of TSS (experiment 4). In experi- in the UASB and to increase its specic methanogenic activity.
ments 5 and 6, a pilot plant UASB was operated at 1011 h The sludge drawn from the middle zone of the UASB entered
HRT and 1415 C. In experiment 5, this plant achieved TSS the upper zone of the digester and then circulated from the
and COD removals above 75 and 54%, respectively. In exper- bottom of the CMSS digester to the bottom of the UASB (Fig. 1).
iment 6, these values decreased to 58% TSS removal and The CMSS digester temperature was set at optimum values
40% COD removal. The inuent concentration explained this ranging from 30 to 35 C, while the UASB operated at ambient
behaviour, since very dilute wastewater was used in exper- temperature.
iment 6. In experiment 7, the pilot UASB reached a high The laboratory scale UASB-CMSS system (experiment 9,
level of TSS removal (8182%) but COD removal remained low Table 4) reached COD and TSS removal levels of 76% and 86%,
(4749%). In the eld application (experiment 8), the full-scale respectively, at a HRT of 6.2 h for the UASB, improving the
UASB reached a TSS removal of 8296% and a COD removal of results obtained in the single-step laboratory UASB (experi-
5893%. ment 4).
Efuent VFA (mainly acetic acid) in single-step UASB sys- The UASB-CMSS pilot plant (experiment 10, Table 4) also
tems working at low environment temperatures ranged from 0 had increased efciency compared to the single-step UASB,
to 80 mgCOD/l, while the average specic methanogenic activ- since it slightly increased the methanogenic activity of the
ity of UASB sludge ranged from 0.02 to 0.05 gCH4 -COD/gVSSd. sludge and reduced the excess sludge generation, which was
64 e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467

Fig. 3 Percentile distribution of inuent and efuent TSS for each anaerobic conguration. Legend: (1) Single-step HUSB
system, (2) Single-step UASB system, (3) UASB-CMSS system, and (4) Two-step HUSB-UASB and UASB-UASB systems. The
numbers of data included were: (1) 282, (2) 128, (3) 41, and (4) 106.

only 7% of the inuent total COD (or 11% of inuent VSS). 5.4. Efuent TSS concentration of the surveyed
The VFA concentration in UASB efuent was reduced to anaerobic systems
830 mgCOD/l. As indicated in Table 4, steady state efciency
for TSS removal was high (6379%). Furthermore, results sug- Fig. 2 shows the average TSS concentration in treated efuent
gest that the relative volume of the CMSS digester could be from each system studied. The TSS concentration of HUS-
considerably lower than the volume of the UASB, and a plug Blab efuent was the highest (87 mg/l, see Exp. 1 in Fig. 2).
up ow sludge digester could be of interest (Alvarez et al., In the HUSB pilot plant, the efuent TSS concentration was
2004). reduced to 50 and 63 mg/l depending on the operating condi-
tions (Fig. 2). These differences in TSS efuent concentration
5.3.4. Two-step anaerobic treatment systems were probably caused by the lower height of the lab scale
At temperatures below 20 C, the two-step anaerobic sys- HUSB, which reduced the distance between the top of the
tem can improve the efciency of the single digester, due to sludge bed and the efuent exit. Furthermore, the upow
the retention and hydrolysis of suspended organic matter in velocity (surface loading rate) is higher in the pilot-scale HUSB
the rst step, allowing for an increase in the methanogenic reactor (1.4 m/h compared to 0.1 m/h for the lab-scale unit)
activity of the anaerobic biomass held in the second step allowing better contact between the inuent and the sludge
digester. bed. In practice, the lab-scale digester needs efuent recircu-
The pilot-scale two-step HUSB-UASB system was operated lation in order to homogenize the sludge bed and avoid bed
at a HRT varying from 5.7 to 2.8 h for the rst step and from compaction. The pilot-scale HUSB showed a good hydraulic
13.9 to 6.5 h for the second step (experiment 11). For the overall ux distribution without the need for recirculation, as was
system, TSS and COD removals ranged from 81 to 89% and 49 outlined by experiments on hydraulic retention time distri-
to 65%, respectively. Hydrolysis of inuent VSS reached 59.7%, bution (Alvarez et al., 2003).
and surplus sludge was 22% of the incoming VSS. Although UASB systems, operating at higher HRT than HUSB sys-
COD removal efciency was inuenced by wastewater con- tems, had average efuent TSS concentrations below 50 mg/l.
centration and temperature, the efuent TSS concentration In the case of UASB-CMSS systems, values for efuent TSS
was mainly constant for inuent COD higher than 250 mg/l. were similar to those of UASB. The lowest efuent TSS con-
In the second run (experiment 12, Table 4), the efciency was centration was obtained with two-step systems, since the pilot
8689% and 5965% for TSS and COD, respectively, which was plant and eld application systems had efuent TSS concen-
slightly higher than in experiment 11 and was a consequence tration below 35 mg/l.
of a higher inuent concentration. Surplus sludge in this case Fig. 3 shows the percentile distribution of the inuent
reached 29% of the incoming VSS. The specic methanogenic and efuent TSS concentration for each anaerobic congu-
activity was 0.010.02 gCH4 -COD/gVSSd for the sludge from ration, excluding laboratory scale experiments. Efuent TSS
the rst step and 0.050.06 gCH4 -COD/gVSSd for the sludge concentration was below 100 mg/l for 95% of the data for all
from the second step. congurations. In the case of the two-step systems, this con-
The eld application of the two-step system also showed a centration was 55 mg/l for 95% of the data. Mean efuent TSS
very good efciency (experiment 13). The UASB-UASB system concentrations ranged from 35 to 63 mg/l.
operated at a HRT of 24 h for each digester and a temperature Anaerobic digesters generated pretreated efuents with a
of 718 C. The efciency of this low load system was 4565% TSS concentration that was 50% lower than that generated
and 7590% of COD and TSS, respectively (Table 4); and surplus by classical pretreatment technologies used in combination
sludge was not generated. Specic methanogenic activity was with CW, as indicated above in Table 1 (mean efuent concen-
0.01 and 0.02 gCH4 -COD/gVSSd for the sludge in the rst and tration of 123 mgTSS/l), or as reported by Vymazal (2005) for
second step UASB reactors, respectively. worldwide experiment (107 mgTSS/l). Therefore, taking into
e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467 65

account these data, all these anaerobic digester congurations eters present in Eq. (1). Finally, wetland efuent must meet
meet the general requirements for a municipal wastewater legal specications, which according to the EU is a BOD5 less
pretreatment capable of preventing clogging in a constructed than or equal to 25 mg/l.
wetland. Information on BOD5 inuent concentration and removal
The stability and reliability of these anaerobic digesters is efciency resulting from the anaerobic treatment of munic-
indicated by their behaviour when they are faced with the ipal wastewater is scarce in the literature. A general review
wide range of inuent and operational conditions that were of anaerobic digesters treating municipal wastewater (Alvarez
tested. Inuent COD varied from 34 to 2700 mg/l and inuent et al., in preparation) indicated that UASB removes about
TSS from 19 to 2100 mg/l, while the operational tempera- 67% of inuent BOD5 . Limited data for single-step UASB and
ture ranged from 5 to 21 C. Efuent quality, however, varied UASB-CMSS systems treating diluted (BOD5 about 200 mg/l)
to a lesser extent, as indicated above for efuent TSS. COD municipal wastewater at temperatures below 20 C show BOD5
removal efciency suffered from low inuent temperatures removals ranging from 50 to 70% (Alvarez et al., 2004, 2006). In
and organic loads, but values remained in the ranges indicated this case, the BOD5 entering the wetland decreases from 200
in Table 4. Pilot- and eld-scale digesters tolerated prolonged to 80 mg/l (60% reduction on average) when an UASB anaero-
periods of temperatures below 13 C. However, prolonged peri- bic pretreatment was applied. Therefore, the required wetland
ods of more than 1 month treating very dilute wastewater area will be reduced by 44%, as can be calculated using Eq. (1).
(inuent COD below 200 mg/l) clearly affected the efciency More efcient anaerobic pretreatment systems could remove
of single-step UASB, and also the stability of two-step HUSB- about 70% of BOD5 and provide a 60% reduction in wetland
UASB systems, when the biomass concentration became very area. Even if BOD5 removal decreases to 46%, as may be the
low (Alvarez, 2004). case when HUSB reactors are used as a municipal wastewater
pretreatment, the required wetland area will be 30% less.
Construction costs of CW are highly variable from place
6. Inuence of anaerobic pretreatment on to place but in many cases may be similar to those of some
constructed wetland area conventional treatment technologies or may be higher when
land costs are accounted for (Rousseau et al., 2004a; Puigagut
Anaerobic pretreatment has two important consequences for et al., 2007). The requirement of a large amount of land is one
the quality of inuent wastewater in a constructed wetland. of the limitations to widespread adoption of CW technology
The rst one is the high TSS removal and the maintenance of for wastewater treatment in both developed and developing
TSS concentration in the pretreated wastewater so that it is countries, and the need for reducing investment costs through
below 100 mg/l, as indicated above. reducing the CW area has been proposed on several occasions
A second consequence is the decrease in the inuent (Badkoubi et al., 1998; Kivaisi, 2001; Gomez Cerezo et al., 2001;
COD concentration to the wetland by an amount that varied Green et al., 2006; El-Hamouri et al., 2007). The footprint of
from 30 to 90%, depending on the type of anaerobic digester high-rate anaerobic digesters is very small, ranging from 0.005
used, wastewater characteristics, and operational conditions to 0.02 m2 /p.e. for the systems included in Table 4. Thus, anaer-
(Table 4). Horizontal ow constructed wetlands can be sized obic digesters may be combined with CW in order to reduce
in order to meet a dened supercial COD load, for exam- the overall area below 1 m2 /p.e., as previously proposed (Barros
ple 12 gCOD/m2 d (Vymazal, 2005). Therefore, generally the and Soto, 2002; Green et al., 2006). Furthermore, construction
reduction in the wetland area required when an anaerobic costs of anaerobic digesters are lower than that of CW and
pretreatment is introduced may range from 30 to 90%. operation costs are very low and are comparable to that of
However, a better method to measure constructed wet- CW (Kivaisi, 2001; Hoffmann et al., 2002). In this way, the use
lands is one that takes into consideration the BOD5 removal of high-rate anaerobic digesters as a rst treatment step may
kinetic, such as the rst order model (Rousseau et al., 2004b). be a better choice than using a high-rate vertical ow CW or a
In this case, assuming the background concentration of BOD very high load horizontal ow CW that can suffer from surface
is equal to zero, the constructed wetland area is calcu- ooding and clogging (Batchelor and Loots, 1997).
lated according to the following equation (Kadlec et al., 2000; An anaerobic system is preferable as a wetland pretreat-
Rousseau et al., 2004b): ment, compared to a primary decanter or a common septic
tank, as it reduces surplus sludge generation, it removes SS
F   BOD i 
A= kv E ln
5
(1) and BOD5 more effectively, and it offers a good way to buffer
h BOD5 e the large uctuations of municipal wastewater from a small
population.
where A is the wetland area (m2 ), F is the volumetric ow The type of anaerobic process, either a hydrolytic pre-
(m3 /d), h is the wetland depth (0.40.6 m), E is the gravel treatment or methanogenic digestion can also inuence the
bed porosity (generally, 0.3), and kv is a rst order kinetic performance and efciency of CW post-treatment as the type
constant that depends on temperature (from 0.17 to 6.11 d1 , of substrate changes. Advanced methanogenic digestion pro-
as reported by Kadlec and Knight (1996) and by Rousseau et duces an efuent that is mainly recalcitrant for anaerobic
al., 2004b). processes in CW. Therefore, post-treatment CW could be
In this way, the wetland area is proportional to the loga- designed with a lower depth in order to maximize aerobic con-
rithm of the quotient between the wetlands inuent BOD5 ditions; or VF CW may be of great interest. In the case of an
and efuent BOD5 . Anaerobic pretreatment greatly modies anaerobic hydrolytic pretreatment, as in the HUSB process,
this quotient but it does not inuence the rest of the param- most of the volatile suspended solids and readily biodegrad-
66 e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467

able matter present in raw wastewater are converted to acetic Alvarez, J.A., Ruz, I., Gomez, M., Presas, J., Soto, M., 2006. Start-up
acid. Acetic acid may be converted in both anaerobic/anoxic alternatives and performance of an UASB pilot plant treating
conditions or in aerobic conditions aiding in the nitrogen and diluted municipal wastewater at low temperature. Bioresour.
phosphorus removal process. Furthermore, biomass growth Technol. 97, 16401649.
Alvarez, J.A., Armstrong, E., Gomez, M., Soto M., 2007. Anaerobic
from acetic respiration in anaerobic conditions was lower than
treatment of low-strength municipal wastewater by a
biomass growth from complex substrates (Gujer and Zehnder, two-stage piloto plant under psychrophilic conditions. Biores.
1983). Low growth will reduce solid accumulation in CW media Technol., doi:10.1016/j.biortech.2008.01.013, in press.
and could prevent clogging phenomena. At the present time, Alvarez, J.A., Ruz, I., Soto, M. Anaerobic treatment of municipal
no research has been reported on the inuence of the type of wastewater: a review.
anaerobic pretreatment on the post-treatment CW operation; Badkoubi, A., Ganjidoust, H., Ghaderi, A., Rajabi, A., 1998.
Performance of a subsurface constructed wetland in Iran.
and there is a need for additional studies on this subject.
Water Sci. Technol. 38 (1), 345350.
Barros, P., Soto, M., 2002. Anaerobic systems for domestic
wastewater treatment in rural areas. In: Proceedings of the
7. Conclusions International Conference on Small Wastewater Technologies
and Mangement for the Mediterranean Area, Seville, Spain,
One of the most signicant handicaps of constructed wet- March 2022.
lands for urban wastewater treatment is gravel bed clogging Barros, P., Soto, M., 2004. Depuracion natural avanzada en
after a few years of operation with poor waste pretreatment nucleos rurais. In: Procceedings of the II Congreso
Internacional de Ingeniera Civil, Territorio y Medio Ambiente,
or high organic loading rates. Another disadvantage of con-
16111622, Santiago de Compostela, Spain, September 2224.
structed wetlands is that a large supercial area is required.
Barros, P., Ruz, I., Soto, M., 2006. Performance of an anaerobic
Both handicaps can be minimised with an appropriate anaer- digester-wetland system for a small community. In:
obic pretreatment. Proceedings of the 10th International Conference on Wetlands
Anaerobic plants may be operated either as hydrolytic or Systems for Water Pollution Control, Lisboa, Portugal,
methanogenic digesters. Hydrolytic digesters, at an HRT of September 2329.
35 h, remove 6585% of TSS and 3555% of COD, showing Batchelor, A., Loots, P., 1997. A critical evaluation of a pilot scale
subsurface ow wetland: 10 years after commissioning. Water
a large amount of hydrolysis and acidication of inuent
Sci. Technol. 35, 337343.
SS. Methanogenic digesters, operating at a HRT of 811 h,
Brix, H., Arias, C.A., 2005. The use of vertical ow constructed
remove 6090% of TSS and 4075% of COD. A two-step system wetlands for on-site treatment of domestic wastewater: new
(hydrolytic and methanogenic digesters in series) can remove Danish guidelines. Ecol. Eng. 25, 491500.
up to 8090% of TSS and 5065% of COD. These results corre- Caselles-Osorio, A., Garca, J., 2006. Performance of experimental
spond to applications carried out in temperate climates where horizontal subsurface ow constructed wetlands fed with
wastewater temperature ranges from 13 to 20 C, or in some dissolved or particulate organic matter. Water Res. 40,
36033611.
cases from 5 to 20 C.
Caselles-Osorio, A., Garcia, J., 2007. Effect of physico-chemical
The average and 95th percentile TSS concentrations of pretreatment on the removal efciency of horizontal
anaerobically treated wastewater were below 60 and 100 mg/l, subsurface-ow constructed wetlands. Environ. Pollut. 146,
respectively, for all congurations. Therefore, anaerobic pre- 5563.
treatment of sewage could help prevent media clogging Caselles-Osorio, A., Puigagut, J., Segu, E., Vaello, N., Granes, F.,
in constructed wetlands. Furthermore, depending on the Garcia, D., Garcia, J., 2007. Solids accumulation in six full-scale
subsurface ow constructed wetlands. Water Res. 41 (6),
amount of organic matter removed, anaerobic pretreatment
13881398.
can provide a reduction of 3060% of the wetland area.
Chazarenc, F., Merlin, G., 2005. Inuence of surface layer on
hydrology and biology of gravel bed vertical ow constructed
wetlands. Water Sci. Technol. 51 (9), 9197.
Acknowledgments Da Motta Marques, D.M.L., Leite, G.R., Giovannini, S.G.T., 2001.
Performance of two macrophyte species in experimental
This work was supported by project CTM2005-06457-C05- wetlands receiving variable loads of anaerobically treated
02/TECNO from the Ministery of Education and Science of municipal wastewater. Water Sci. Technol. 44 (1112), 311316.
Dahab, M.F., Surampalli, R.Y., 2001. Subsurface-ow constructed
Spain.
wetlands treatment in the plains: ve years of experience.
Water Sci. Technol. 44 (11), 375380.
references El-Hamouri, B., Nazih, J., Lahjouj, J., 2007. Subsurface-horizontal
ow constructed wetland for sewage treatment under
Moroccan climate conditions. Desalination 215, 153158.
El-Khateeb, M.A., El-Gohary, F.A., 2003. Combining UASB
Alvarez, J.A., Zapico, C.A., Gomez, M., Presas, J., Soto, M., 2003. technology and constructed wetland for domestic wastewater
Anaerobic hydrolysis of a municipal wastewater in a reclamation and reuse. Water Sci. Technol. 3 (4), 201208.
pilot-scale digester. Water Sci. Technol. 47 (12), 223230. Foresti, E., 2002. Anaerobic treatment of domestic sewage:
Alvarez, J.A., Armstrong, E., Gomez, M., Presas, J., Soto, M., 2004. established technologies and perspectives. Water Sci.
Performance of an UASB-Digester system treating domestic Technol. 45 (10), 181186.
wastewater. Environ. Technol. 25, 11891199. Foresti, E., Zaiat, M., Vallero, M., 2006. Anaerobic process as the
Alvarez, J.A., 2004. Tratamiento anaerobio de aguas residuales core technology for sustainable domestic wastewater
urbanas en planta piloto. PhD Thesis. Department of Physical treatment: consolidated applications, new trends,
Chemistry and Chemical Engineering I, University of A perspectives and challenges. Environ. Sci. Biotechnol. 5, 319.
Coruna, Spain.
e c o l o g i c a l e n g i n e e r i n g 3 3 ( 2 0 0 8 ) 5467 67

Goncalves, R.F., Charlier, A.C., Sammut, F., 1994. Primary Philippi, L.R., Da Costa, R.H.R., Sezerino, P.H., 1999. Domestic
fermentation of soluble and particulate organic matter for efuent treatment through integrated system of septic tank
wastewater treatment. Water Sci. Technol. 30 (6), 5362. and root zone. Water Sci. Technol. 40, 125131.
Gomez Cerezo, R., Suarez, M.L., Vidal-Abarca, M.R., 2001. The Puigagut, J., Villasenor, J., Salas, J.J., Becares, E., Garcia, J., 2007.
performance of a multi-stage system of constructed wetlands Subsurface-ow constructed wetlands in Spain for the
for urban wastewater treatment in a semiarid region of SE sanitation of small communities: a comparative study. Ecol.
Spain. Ecol. Eng. 16, 501517. Eng. 30, 312319.
Green, M., Shaul, N., Beliavski, M., Sabbah, I., Ghattas, B., Tarre, S., Rousseau, D.P.L., Vanrolleghem, P.A., Pauw, N.D., 2004a.
2006. Minimizing land requirement and evaporation in small Constructed wetlands in Flanders: a performance analysis.
wastewater treatment systems. Ecol. Eng. 26, 266271. Ecol. Eng. 23, 151163.
Gujer, W., Zehnder, A.J., 1983. Conversion processes in anaerobic Rousseau, D.P.L., Vanrolleghem, P.A., Pauw, N.D., 2004b.
digestion. Water Sci. Technol. 15, 127. Model-based design of horizontal subsurface ow constructed
Henze, H., Harremoes, P., Jansens, J., Arvin, E., 1995. Wastewater treatment wetlands: a review. Water Res. 38,
Treatment. Springer-Verlag, New York. 14841493.
Hoffmann, H., Platzer, C., Heppeler, D., Barjenbrunch, M., Ruz, I., Soto, M., Veiga, M.C., Ligero, P., Vega, A., Blazquez, R.,
Tranckner, J., Belli, P., 2002. Combination of anaerobic 1998. Performance of and biomass characterization in a UASB
treatment and nutrient removal of wastewater in Brazil. In: reactor treating domestic wastewater at ambient
Proceedings of the 3rd Water World Congress, April, temperature. Water SA 24 (3), 215221.
Melbourne, Australia, April 912. Ruz, I., Alvarez, J.A., Daz, M.A., Grana, M., Soto, M., 2006.
Kadlec, R.H., Knight, R.L., Vymazal, J., Brix, H., Cooper, P., Haberl, Municipal wastewater treatment in an anaerobic
R., 2000. Constructed Wetlands for Pollution Control: digester-constructed wetland system. In: Proceedings of 10th
Processes, Performance, Design and Operation, IWA Specialist International Conference on Wetlands Systems for Water
Group on Use of Macrophytes in Water Pollution Control, Pollution Control, Lisboa, Portugal, September 2329.
Scientic and Technical Report 8, IWA Publishing, London, Ruz, I., Blazquez, R., Soto, M., 2007. Characteristics and anaerobic
U.K. treatability of municipal and industrial estate wastewaters.
Kadlec, R.H., Knight, R.L., 1996. Treatment Wetlands. Lewis Publ. Environ. Technol. 28, 10631072.
USA, Boca Raton, FL. Sousa, J.T., Van Haandel, A.C., Guimaraes, A.A., 2001.
Kaseva, M.E., 2004. Performance of a sub-surface ow Post-treatment of anaerobic efuents in constructed wetland
constructed wetland in polish pre-treated wastewatera systems. Water Sci. Technol. 44 (4), 213219.
tropical case study. Water Res. 38, 681687. Sousa, J.T., Van Haandel, A., Lima, E.P.C., Guimaraes, A.V.A., 2003.
Kivaisi, A.K., 2001. The potential for constructed wetlands for Performance of constructed wetland systems treating
wastewater treatment and reuse in developing countries: a anaerobic efuents. Water Sci. Technol. 48 (6), 295299.
review. Ecol. Eng. 16, 545560. Sperling, von M., 1996. Comparison among the most frequently
Langergraber, G., Haberl, R., Laber, J., Pressi, A., 2003. Evaluation used systems for wastewater treatment in developing
of substrate clogging processes in vertical ow constructed countries. Water Sci. Technol. 33 (3), 5972.
wetlands. Water Sci. Technol. 48 (5), 2534. Standard Methods for the Examination of Water and Wastewater,
Lens, P., Zeeman, G., Lettinga, G., 2001. Decentralised Sanitation 1995. 19th ed. American Public Health Association/American
and Reuse: Concepts, Systems and Implementation. IWA Water Works Association/Water Environment Federation,
Publishing, London, UK. Washington, DC, USA.
Lettinga, G., de Man, A., Van der Last, A.R.M., Wiegant, W., Van USEPA, 2000. Constructed Wetland Treatment of Municipal
Knippenberg, K., Frijns, J., Van Buuren, J.C.L., 1993. Anaerobic Wastewaters. US.EPA 625/R99/010, Cincinnati, OH, USA.
treatment of domestic sewage and wastewater. Water Sci. Van Haandel, A., Kato, M.T., Cavalcanti, P.F.F., Florencio, L., 2006.
Technol. 27 (9), 6773. Anaerobic design concepts for the treatment of domestic
Lettinga, G., 2001. Digestion and degradation, air for life. Water wastewater. Environ. Sci. Biotechnol. 5, 2138.
Sci. Technol. 44 (8), 157176. Vymazal, J., Brix, H., Cooper, P.F., Green, M.B., Haberl, R., 1998.
Ligero, P., Vega, A., Soto, M., 2001a. Pre-treatment of urban Constructed Wetlands for Wastewater Treatment in Europe.
wastewaters in a hydrolytic up ow digester. Water SA 27 (3), Backhuys Publishers, Leiden, The Netherlands.
16. Vymazal, J., 2002. The use of sub-surface constructed wetlands
Ligero, P., Vega, A., Soto, M., 2001b. Inuence of HRT (hydraulic for wastewater treatment in the Czech Republic: 10 years
retention time) and SRT (solid retention time) on the experience. Ecol. Eng. 18, 633646.
hydrolytic pre-treatment of urban wastewater. Water Sci. Vymazal, J., 2005. Horizontal sub-surface ow and hybrid
Technol. 44 (4), 714. constructed wetlands systems for wastewater treatment.
Masi, F., Martinuzzi, N., Bresciani, R., Giovannelli, L., Conte, G., Ecol. Eng. 25, 478490.
2006. Tolerance to hydraulic and organic load uctuations in Wang, K., 1994. Integrated anaerobic and aerobic treatment of
constructed wetlands. In: Proceedings of 10th International sewage. Ph.D. Thesis. Department of Biotechnology,
Conference on Wetlands Systems for Water Pollution Control, University of Wageningen, The Netherlands.
Lisboa, Portugal, September 2329. Winter, K.J., Goetz, D., 2003. The impact of sewage composition
Mbuligwe, S.E., 2004. Comparative effectiveness of engineered on the soil clogging phenomena of vertical ow constructed
wetland systems in the treatment of anaerobically pre-treated wetlands. Water Sci. Technol. 48 (5), 914.
domestic wastewater. Ecol. Eng. 23, 269284. Zeeman, G., Lettinga, G., 1999. The role of anaerobic digestion of
Metcalf & Eddy. 2003. Wastewater Engineering: Treatment and domestic sewage in closing the water and nutrient cycle at
Reuse. fourth ed. McGraw-Hill, Boston, USA. community level. Water Sci. Technol. 39 (5), 187194.
Neralla, S., Weaver, R.W., Lesikar, B.J., Persyn, R.A., 2000. Zhao, Y.Q., Sun, G., Allen, S.J., 2004. Anti-sized reed bed system
Improvement of domestic wastewater quality by subsurface for animal wastewater treatment: a comparative study. Water
ow constructed wetlands. Bioresour. Technol. 75, 1925. Res. 38, 29072917.

Potrebbero piacerti anche