Innovative conception and performance evaluation of a

Water Science & Technology Vol 54 No 2 pp 87–94 Q IWA Publishing 2006

compact on-site treatment system
V.P. Sousa and C.A.L. Chernicharo
Department of Sanitary and Environmental Engineering, Federal University of Minas Gerais, Av. Contorno,
842/701, 30110-060 - Belo Horizonte, Brazil (E-mail: calemos@desa.ufmg.br)

Abstract The purpose of this study was to develop a new configuration for a compact on-site treatment
system, which could become an attractive alternative, from technical, economic, social and environmental
viewpoints, to the technologies that are currently employed. The treatment unit consists of a cylindrical tank,
where half of the volume is used as a modified septic tank and the other half is divided between an
anaerobic hybrid reactor and a trickling filter. An intermittent feeding system was used, with minimum, mean
and maximum flowrate settings (Qmin ¼ 0.25 l.s21, Qmean ¼ 0.50 l.s21 and Qmax ¼ 1.00 l.s21), to reflect the
actual operating conditions of a compact on-site treatment system serving a typical dwelling. An average 24-
hour hydraulic detention time was used, corresponding to a flowrate of 750 l.d21. High removal efficiencies
and low concentrations of COD, BOD and TSS in the final effluent were achieved, even when the unit was
exposed to hydraulic loading peaks during feeding periods at maximum flowrate.
Keywords Anaerobic filter; anaerobic hybrid reactor; on-site treatment; septic tank; trickling filter; UASB
reactor

Introduction
In rural areas, where scattered populations and small communities prevail, it has been
common practice to employ individual sewage treatment units, usually consisting of a
septic tank and, optionally, a post-treatment and/or final disposal unit. However, despite
the high hydraulic detention times for which these systems are designed, frequent operat-
ing problems are reported, often resulting in reduced efficiency in removing organic mat-
ter (Andrade Neto, 1997).
In recent years, efforts have been made to develop new de-centralized sewage treat-
ment systems. These systems incorporate the principles of high-efficiency anaerobic reac-
tors, especially UASB and hybrid reactors (Tilche and Vieira, 1991; Bogte et al., 1993;
Elmitwalli et al., 2002, 2003). These systems differ from conventional septic tanks by
incorporating an upflow feeding structure and some solid retention device. Removal of
dissolved organic matter is much more efficient as a result of improved biomass/substrate
contact.
In addition, the association of anaerobic and aerobic reactors can also contribute to the
improvement of the final effluent quality (van Haandel and Lettinga, 1994), especially in
terms of lower organic matter concentration and higher dissolved oxygen levels. If the
aerobic post-treatment unit can maintain the same conceptual simplicity as the anaerobic
reactors, then the system can became more feasible.
Therefore, the purpose of this research was to develop and evaluate the performance
of an on-site treatment system that could be capable of ensuring effective removal of
solids and organic matter and keeping suitable levels of dissolved oxygen in the final
effluent. The system should also be an attractive alternative to the on-site technologies
currently employed in Brazil.
doi: 10.2166/wst.2006.490 87
 Methods
Experimental apparatus
Research with the compact on-site treatment system was carried out at the UFMG/CO-
PASA Experimental Campus, located next to Arrudas Sewage Treatment Plant (STP) in
Belo Horizonte City, Brazil. At Arrudas STP, raw sewage goes through a preliminary
treatment stage comprising a screening device and a sand box. Following the preliminary
treatment, a small portion of sewage was pumped to feed the experimental unit.
V.P. Sousa and C.A.L. Chernicharo

The treatment unit was made up of a cylindrical fibreglass tank, where half of the volume
was used as a modified septic tank (chamber 1) and the other half divided between an anaero-
bic hybrid reactor (chamber 2) and a trickling filter (chamber 3) (Table 1 and Figure 1). The
total volume of this unit corresponds to the volume of a septic tank, as determined according
to the Brazilian directives.
Raw sewage was fed into the system through the bottom of the chamber 1 (modified
septic tank), then flowing upwards to a lamellar sedimentation tank located in the upper
part of this chamber. Next, partly-treated sewage was directed to the bottom of chamber
2 where it was distributed and then entered in contact with the sludge bed and packing
media of the anaerobic hybrid reactor (UASB reactor þ anaerobic filter), sequentially
leaving this chamber through a lateral Y-shaped outlet positioned at the top. Within
chamber 3, the effluent flows downwards through a trickling filter where it is subjected to
final polishing. A small secondary sedimentation tank was added in the bottom of the
trickling filter.
In addition to the biological treatment unit, the experimental apparatus incorporated a
2.5 £ 2.0 m sludge drying bed.

Operating conditions
To reproduce the actual operating conditions of a compact on-site treatment system, the
experimental unit was fed with flow patterns produced by a typical dwelling. Therefore,
minimum, mean and maximum flowrates were determined, which refer to water for
interior use such as showers and toilets. The treatment unit was operated on an intermit-
tent basis, with the flowrate peaking at different times of the day.
While adjusted for an average flowrate of 750 l.d21and 24-hour hydraulic detention
time, the feed system was automated to enable pump flowrate to be varied according to
the flow pattern selected, with minimum, mean and maximum flowrate settings
(Qmin ¼ 0.25 l.s21, Qmean ¼ 0.50 l.s21 and Qmax ¼ 1.00 l.s21). Such high transient
hydraulic regime produced by a typical dwelling allowed analysis of the system’s solid
retention capacity, as it results in higher upflow velocities. To reflect the systems actually
in use in different areas of Brazil, no seed sludge was used during the start up.

Operating phases
There was a time during monitoring when raw sewage was very diluted as a result of
heavy rainfall. Because of that, monitoring was divided into three phases, allowing a bet-
ter discussion of the results: phase 1: initial monitoring period – length: 60 days; phase

Table 1 Basic details of the compact treatment unit

D (m) H (m) V total (m3) Chamber 1 Chamber 2 Chamber 3

2 3 2 3
A (m ) V (m ) A (m ) V (m ) A (m2) V (m3)

1.05 2.00 1.75 0.433 0.86 0.216 0.43 0.216 0.43

88 D ¼ diameter, H ¼ height, V ¼ volume, A ¼ surface area
 V.P. Sousa and C.A.L. Chernicharo
Figure 1 Schematics of the compact treatment unit

2: intermediate monitoring period (heavy rainfall and diluted sewage) – length: 90 days;
and phase 3: final monitoring period – length: 77 days.

Monitoring
The unit operation was monitored for the following main physical-chemical and microbio-
logical parameters: temperature, pH, DO, COD, BOD and settleable and suspended solids.
All analyses were carried out according to the procedures described in the Standard
Methods for the Examination of Water and Wastewater (AWWA/APHA/WEF, 1998).

Sampling routine. Monitoring included the collection of grab samples of raw sewage,
as well as effluents from the modified septic tank, anaerobic hybrid reactor and trickling
filter, for minimum, mean and maximum flowrates.

Results
Although the experimental unit was monitored for minimum, mean and maximum flow-
rates, only results for mean flowrate are presented and discussed in this paper. Complete
analysis of the results for all operational conditions is presented in Sousa (2004).
Figures 2– 4 show box-plot charts for results concerning COD, BOD and TSS for the
influent (Inf) and effluents from the modified septic tank (MST), anaerobic hybrid reactor
(AnHR) and trickling filter (TF).
For a more careful interpretation of these results, it should be noted that the system was
fed with different concentrations but at the same flowrate of 750 l.d21, during the three
different phases. As to phase 2, in particular, the charts show low raw sewage concen-
tration of all parameters, due to intensive rainfall and dilution of the incoming wastewater.
During phase 1, corresponding to the first 60 days of operation, removal efficiencies aver-
aged 77, 77 and 88% for COD, BOD and TSS, respectively (Figures 5–7). For phase 2,
average efficiencies were 46, 35 and 72% for COD, BOD and TSS, respectively. During
phase 3, removal efficiencies averaged 80, 81 and 86%, for COD, BOD and TSS, respect-
ively (Figures 5–7).
Upon chart analysis, large concentrations of COD, BOD and TSS for raw sewage
were observed during phase 1. Removal efficiencies were relatively high, enough to
produce a final effluent with average concentrations of 197 mgCOD.l21, 141 mgBOD.l21
and 70 mgTSS.l21. 89
V.P. Sousa and C.A.L. Chernicharo

Figure 2 Box plot of COD results for different monitoring phases. Inf: Influent, MST: Modified septic tank,
AnHR: Anaerobic hybrid reactor, TF: Trickling filter

During phase 2, as a result of the low concentrations of various raw sewage par-
ameters monitored, removal efficiencies were lower than those achieved during phase 1.
However, the final effluent from the system had average concentrations of
148 mgCOD.l21, 49 mgBOD.l21 and 57 mgTSS.l21.
In phase 3 large concentrations of COD, BOD and TSS were found for the raw sew-
90 age, similarly to what happened during phase 1. Removal efficiencies were high and
 V.P. Sousa and C.A.L. Chernicharo

Figure 3 Box plot of BOD results for different monitoring phases. Inf: Influent, MST: Modified septic tank,
AnHR: Anaerobic hybrid reactor, TF: Trickling filter

resulted in a final effluent with average concentrations of 197 mgCOD.l21,

74 mgBOD.l21 and 99 mgTSS.l21.
Since typical characteristic of raw sewage only prevail during phases 1 and 3, with
high concentrations of all parameters being observed, the discussion is thereafter focussed
on these two phases. For that, average COD, BOD and TSS concentrations and removal
efficiencies are depicted in Figures 5–7. From the figures, one can note relatively low
removal efficiencies for the modified septic tank and for the trickling filter, systematically 91
V.P. Sousa and C.A.L. Chernicharo

Figure 4 Box plot of TSS results for different monitoring phases. Inf: Influent, MST: Modified septic tank,
AnHR: Anaerobic hybrid reactor, TF: Trickling filter

below 40 and 30%, respectively. On the contrary, very high efficiencies were observed
for the anaerobic hybrid reactor, ranging from 58 to 81%.
The low efficiencies observed for the trickling filter resulted probably from its oper-
ation method, whereby it remained unfed during long periods of the day, due to the inter-
mittent feeding pattern applied to the system. This may have contributed to the biofilm
92 becoming desiccated. Another possible explanation for low TF efficiencies may be the
 V.P. Sousa and C.A.L. Chernicharo
Figure 5 Average COD concentrations and removal efficiencies

organic matter remaining from the anaerobic hybrid reactor, which is more refractory and
difficult to degrade.
The effluent maintained slight variations for the parameters analysed, irrespective of
fluctuations in the influent, thanks to the system stability. The box-plot charts show good
system stability, provided especially by the anaerobic hybrid reactor.

Sludge production
The anaerobic sludge from the unit was disposed of directly onto the drying bed on a
monthly basis. The sludge production rates were as expected for both the septic tank and
the anaerobic hybrid reactor, amounting to approximately 0.24 and 0.13 kgTSS.kgCOD21,
respectively.

Figure 6 Average BOD concentrations and removal efficiencies 93

V.P. Sousa and C.A.L. Chernicharo

Figure 7 Average TSS concentrations and removal efficiencies

Conclusions
The results obtained show that this new compact on-site treatment system configuration has
a great potential. Such system, with a volume corresponding to the volume of a septic tank,
as calculated under Brazilian directives, has produced effluent similar to that from a septic
tank þ anaerobic filter, whose volume is approximately twice as large. Organic matter
removal efficiencies were low in the modified septic tank and high in the anaerobic hybrid
reactor. The trickling filter, although showing low efficiency rates, has provided increased
system stability. The system has exhibited high removal efficiencies and low BOD, COD
and TSS concentrations in the final effluent, even when subject to hydraulic loading peaks.
Acknowledgements
The authors wish to acknowledge the support obtained from the following Brazilian insti-
tutions: Fundação Nacional de Saúde – FUNASA and Companhia de Saneamento de
Minas Gerais – COPASA.

References
Andrade Neto, C.O. (1997). Simplified systems for domestic sewage treatment: Brazilian experience. ABES,
Rio de Janeiro. (in Portuguese).
Bogte, J.J., Breure, A.M., van Andel, J.G. and Lettinga, G. (1993). Anaerobic treatment of domestic sewage
in small-scale UASB reactors. Wat. Sci. Tech., 27(9), 75 –82.
Elmitwalli, T.A., Sayed, S., Groendijk, L., Van Lier, J., Zeeman, G. and Lettinga, G. (2003). Decentralised
treatment of concentrated sewage at low temperature in two-step anaerobic system: two upflow-hybrid
septic tanks. Wat. Sci. Tech., 48(6), 219 – 226.
Elmitwalli, T.A., Sklyar, V., Zeeman, G. and Lettinga, G. (2002). Treatment of domestic sewage in a two-
step anaerobic filter/anaerobic hybrid system at low temperature. Wat. Res., 36, 2225– 2232.
Sousa, V.P. (2004). Innovative conception and performance evaluation of a compact on-site treatment system.
M.Sc. Dissertation, Faculty of Engineering, Federal University of Minas Gerais, Brazil (in Portuguese).
Standard Methods for the Examination of Water and Wastewater (1998). 20th edn, American Public Health
Association/American Water Works Association/ Water Environment Federation, Washington, DC, USA.
Tilche, A. and Vieira, S.M.M. (1991). Discussion report on reactor design of anaerobic filter and sludge bed
reactors. Wat. Sci. Tech., 24(8), 193 –206.
Van Haandel, A.C. and Lettinga, G. (1994). Anaerobic Sewage Treatment in Regions with a Hot Climate,
Wiley, Chichester, UK.
94