Processing of human faeces by wet vermifiltration for improved on-site sanitation

Abstract

The use of a vermifilter containing Eisenia fetida to degrade human faeces in a continuous wet system was explored. This paper aimed to understand the formation of vermicompost within the system, the quality of the effluent produced, and the effect of different bedding matrices. Eight filters were constructed, utilising four different bedding materials: four of these systems were seeded with 400 g of worms (vermifilters) while the others served as controls. The systems were flushed with 12 L of water per day and the experiment was split into five phases, each with different feeding regimes. Between 23.7 and 24.7 kg of fresh human faecal matter was added to the vermifilters over the 360 day period. The presence of the worms was found to increase the faecal reduction to 76% on average, compared to 17% in the control systems on average. Statistically significant reductions in phosphate, chemical oxygen demand and thermotolerant coliforms were achieved in the effluent of all vermifilters. The most suitable bedding matrix was a mixture of coir and woodchip. This study shows that there is potential for continuous treatment of human faeces using wet, on-site vermifilters.

Full text

Words: 4,950
Processing of human faeces by wet vermifiltration for improved on-site sanitation
C. Furlong1
Department of Civil and Environmental Engineering, Imperial College London, London,
SW7 2AZ, UK.
Email: claire.furlong@ymail.com
M.R. Templeton1
Department of Civil and Environmental Engineering, Imperial College London, London,
SW7 2AZ, UK.
Email: m.templeton@imperial.ac.uk
W.T. Gibson2
2Bear Valley Ventures Limited, Braeside, Utkinton Lane, Cotebrook, Tarporley, Cheshire,
CW6 0JH UK
Email: waltergibson@bearvalleyventures.com
Abstract
The use of a vermifilter containing Eisenia fetida to degrade human faeces in a continuous
wet system was explored. This paper aimed to understand the formation of vermicompost
within the system, the quality of the effluent produced, and the effect of different bedding
matrices. Eight filters were constructed, utilising four different bedding materials: four of
these systems were seeded with 400 g of worms (vermifilters) while the others served as
controls. The systems were flushed with 12 litres of water per day and the experiment was
split into five phases, each with different feeding regimes. Between 23.7 and 24.7 kg of fresh
1 Corresponding author

human faecal matter was added to the vermifilters over the 360 day period. The presence of
the worms was found to increase the faecal reduction to 96% in the vermifilters on average,
compared to 38% in control systems on average. Statistically significant reductions in
phosphate, COD and thermotolerant coliforms were achieved in the effluent of all
vermifilters. The most suitable bedding matrix was a mixture of coir and woodchip. This
study shows that there is potential for continuous treatment of human faeces using wet, on-
site vermifilters.
Keywords: Eisenia fetida, sewage, vermicompost, vermifilter, vermireactor, worm

Introduction
The majority of the world’s population relies on on-site, decentralised sanitation systems
such as pit latrines, cesspits, and septic tanks. One of the major problems associated with
these systems is that they require emptying, which can be costly, inconvenient and hazardous.
In high-density urban areas these problem are amplified, due to the lack of available space.
Emptying should ideally be undertaken by a vacuum pump truck, but tankers cannot gain
access to narrow streets and alleys (Thye et al., 2011). Alternative small-scale emptying
solutions have been developed to overcome these problems, e.g. the Gulper, MAPET (Thye
et al., 2011), but these technologies are still being trialled and may not be effective for all
sludge types. Worldwide approximately 200 million latrines and septic tanks must be
manually emptied each year by workers descending into the pit equipped with buckets and
spades (Thye et al., 2011). Furthermore, the final disposal of faecal sludge by any of these
methods is often simply by dumping into the immediate environment. This reintroduces
pathogens into the environment which were previously safely contained in the pit or tank. An
improved on-site sanitation solution needs to be identified which reduces the frequency of
required emptying of latrines, ideally together with achieving treatment of the waste so that
handling and disposal of the waste are safer activities.
An on-site worm-based system may be a solution to these problems. With this approach the
amount of solids within the system can potentially be reduced, due to the net loss of biomass
and energy when the food chain is extended by using worms. By reducing both the frequency
of emptying and the size of the system, this approach could be particularly suitable for highly
dense urban and peri-urban areas. Additionally worms are known to remove pathogens (from
sewage sludge) to the level where the waste can be safely applied to land (Eastman et al.,
2001), and the waste produced is dry compost (known as vermicompost) rather than a sludge,

which makes it easier to handle and transport.
In the field of sanitation research, studies using Eisenia fetida have concentrated on the
stabilisation of sewage sludge (Parvaresh et al., 2004), dried or pre-treated faecal matter
(Yadav et al., 2010;Yadav et al., 2011), or wastewater mixed with organic bulking agents
(Taylor et al., 2004). Pre-treatment was thought to be required as E. fetida died within an
hour of being introduced to fresh human faecal matter (Yadav et al., 2010). The importance
of the bedding layer, i.e. the matrix in which the worms live, was also noted, as they found
that E.fetida died when fed with human faeces without this support layer (Yadav et al., 2010).
Larger scale community worm-based systems have been trialled in China for the treatment of
sludge (Xing et al., 2011; Zhao et al., 2010) and sewage (Xing et al., 2010; Wang et al.,
2012). Commercial on-site systems are currently available, e.g. the Solid Waste Digester
(Simple Wastewater Solutions, 2010) and Biolytix™ (Biolytix, 2008), which are seeded with
worms and are attached to flushing systems. They are designed for use in rural locations in
developed countries but are cost-prohibitive for households in developing countries. They
also have large footprints for installation in an urban context and designed for waste
containing higher liquid content than is typical in developing countries.
Flushing systems are highly desirable in low-income urban and peri-urban contexts where
people strive for modernity. The advantages of systems with a water trap/seal include the
separation from one’s own and others’ waste, and the elimination of odours and flies, which
add to the desirability of flushing systems. This research was a part of a larger project which
used a people-led approach to sanitation improvement; therefore, in the light of these desired
benefits, the study focuses on flushing systems only. No other studies have investigated wet

(flushed with water periodically) worm-based systems for degrading fresh human faeces: as
such, experimental data were required to assess the feasibility of this approach for on-site
sanitation in developing countries.
The specific objectives of this work were to establish whether worms can continuously (the
systems are fed daily) degrade fresh human faeces under water-flushing conditions. To the
best of our knowledge all other laboratory based studies have been batch fed (fed weekly).
The system in the paper is described as ‘wet’ or ‘water-flushed’ whereas in traditional worm-
based systems water is only added to keep the system moist (e.g. Yadav et al., 2011) or a wet
slurry or sludge is added to the system (e.g. Xing et al., 2011), but no other studies have been
identified in which water was flushed through the systems to simulate the conditions in a
flushing sanitation system. Furthermore, this study was performed in order to determine
where and how much vermicompost is deposited and to assess the quality of the effluent
produced. Additionally, the effect of different bedding matrices on faecal solids reduction
(mass) and effluent quality was considered. The experiments were designed to replicate a
potential on-site wet worm based sanitation system, i.e. they were fed daily with fresh human
faeces and water was pumped into the filters to simulate flushing a toilet.
Methodology
Experimental systems
Eight filter systems were constructed from polypropylene boxes with internal dimensions of
L 37 x W 27 x H 25.5 cm and a surface area of 0.1 m2 (Figure 1). The base of each box
(except the sump box) was removed and replaced by plastic mesh with a 5 mm aperture and a
further mesh with a 1 mm aperture was placed on the bedding box mesh. Each unit consisted
of three boxes stacked on top of each other: the top box contained a 10 cm depth of bedding

matrices, the middle contained drainage media (plastic drainage coil with a 60 mm external
diameter, cut into 60 mm segments), and the bottom box was the sump which had a tap that
drained to a collection vessel. All components of the system were weighed separately to
allow for changes in mass to be calculated over time.
Four different bedding matrices were tested: coir (Fertile Fibres Ltd, Withington, UK),
woodchip (sourced from the Centre for Alternative Technology (CAT), Powys, Wales, UK),
a volumetric mixture of coir and woodchip (50:50), and a volumetric mixture of coir,
woodchip and vermicompost (33:33:33). Eight boxes were initially set up (two of each
bedding matrix type), with 400 g of E. fetida (worm density of 4 kg/m2) being added to one
of each matrix type (vermifilter), and the second corresponding box being used as a control
(did not contain worms). This worm density was selected from the estimation that 0.1 m2 of
vermifilter surface area could treat the waste from one person per day (approximately 200 g
of faeces, unpublished data) and a conservative estimate of worm feed consumption of 0.5 kg
feed/kg worm per day.
On top of the bedding a plastic mesh insert was placed and faecal matter was introduced on
top of this mesh (the faecal mesh, Figure 1). Each system was topped with a lid which
contained 40 1-mm randomly placed ventilation holes and an inlet for water additions. Water
was introduced using a peristaltic pump (Watson Marlow 502S, Cheltenham, UK) to simulate
flushing: approximately 12 litres of water was added during five watering periods spaced
throughout the day. This happened throughout the study apart from the resting period (Phase
4) when feeding was suspended to assess the ability of digestion to go to completion during
which only one litre of water was added per day to keep the systems moist.

Human faeces were collected daily from a series of bucket toilets at CAT. They were
homogenised through pooling and thoroughly mixed. Once a day the specified amount of
fresh faeces (Table 1) was placed on the faecal mesh. The variation in the expected feeding
regime (see phase descriptions in Table 1) and actual feeding regime (see mean feed addition
per day, Table 1) was due to the variations in the amount of faeces harvested. The reactors
were fed from Monday to Friday as it was not feasible to harvest faeces over the weekend:
therefore all the feed rates quoted in Table 1 are for a 5-day period.
Once the boxes were assembled they were wetted with six litres of water and allowed to drain
for one hour. The worms were then added and allowed to acclimatise for eight days without
feeding. The experiment was divided into five phases (Table 1) and ran for 360 days. The
reactors were housed in a heated building where the mean temperature was 22°C (Lascar EL-
USB-TC, Whiteparish, UK). A Lascar thermocouple and data logger and EC-5 moisture
probe (Decagon Devices Inc., Pullman, USA) were positioned in the middle of each bedding
layer.
Methods of analysis
All methods were chosen so they did not disturb or destroy the systems. Additionally, they
had to be undertaken under field conditions, due to the lack of standard laboratory facilities
on-site at CAT.
Moisture measurements (v/v %) were taken daily using ProCheck datalogger (Decagon
Devices Inc, Pullman, USA). A potting mixture calibration was used for all boxes except for
those containing only woodchip, when the perlite calibration was used. The laboratory and
box temperatures were measured hourly using a Lascar EL-USB-TC thermocouple. The

mass of faecal matter on the mesh above the bedding layer (faecal mesh, Figure 1) was
weighed separately.
The influent and effluent were analysed approximately weekly using Hach DR/890 field
testing kits (Loveland, USA) for chemical oxygen demand (Hach Method 8000), nitrate
(Hach Method 8039), nitrite (Hach Method 81532), and total phosphate (Hach Method
10127). Thermotolerant coliforms were analysed using a DelAgua Kit (Guildford, UK)
(Robens Centre, 2004). The effluent pH was measured using an electrode (pH703, TECPEL,
Taipei, Taiwan) and settleable solids were measured using standard methods (APHA, 1992).
All samples were analysed in duplicate and arithmetic mean for the samples are reported in
this paper (Table 2).
Data analysis
Waste stabilisation is reported in other papers (e.g. Yadav et al., 2011; Xing et al., 2011), and
this does not reflect the reduction in the mass of the waste. Mass reduction is important when
assessing this technology’s suitability for on-site sanitation, as it is directly related to the
necessary size of the system and emptying frequency. Mass reduction was calculated (Eq.1),
together with overall faecal reduction.
Equation 1: Weekly percentage faecal reduction
((TFMA W1– FMRW1)/ TFMA W1) x 100
TFMA W1 = total faecal mass added onto mesh weekly
FMRW1 = faecal mass remaining on mesh at the end of the week
After 360 days the vermifilters were decommissioned and the undigested faeces on the faecal

mesh, the worm population and vermicompost were separated and weighed. The control
filters were decommissioned after 30 days due to the lack of overall decomposition (8-39%)
and the large amount of faecal matter that accumulated (0.73-1.1 kg).
Statistical analysis of results was carried out using SPSS 12.0.1. Student’s t-test was used to
compare data sets. One-way ANOVA was used to compare multiple data sets using the post-
hoc Tukey test. The null hypothesis of these tests was accepted if p>0.05.
Results and discussion
Reduction of faecal matter
In Phase 1 there was a statistically significant difference in the weekly percentage faecal
reduction between the filters and vermifilters for each bedding type (Student’s t-test coir
p=0.003; woodchip p=0.007; coir and woodchip p=0.008; coir, wood and vermicompost
p=0.006), confirming that the worms were actively degrading faecal material. Within the
controls, faecal reduction was higher for the filter containing vermicompost than for the other
types of bedding material (percentage feed reduction at the end of Phase 1: coir 11%,
woodchip 12%; coir and woodchip 11%; coir, wood and vermicompost 33%). This was
probably due to the vermicompost in the bedding matrices being microbiologically active.
There was no statistically significant difference between the weekly faecal reduction in the
vermifilters with different bedding matrices (ANOVA F(3,12)=1.177, p=0.359), therefore the
type of bedding did not affect the ability of the worms to consume faecal material during
Phase 1. The faecal reduction dropped in the vermifilters after the feed rate was increased at
the start of Phase 2 (Table 1). The systems became acclimatised to the new feeding rate after
approximately six weeks when 100% reduction was achieved; this pattern was repeated at the
start of Phase 5.

In Phase 1, this could be linked to the acclimatisation of the worms to the feed, sincein other
studies the worms were acclimatised prior to the experiments (Yadav et al., 2011), but in
subsequent phases it was more likely to be from the population adapting to the increased feed
rates, as this coincides with approximately the same amount of time required for a worm to
hatch and mature (Edwards & Lofty, 1997).
After approximately six months (26 weeks) vermicompost started accumulating on the faecal
mesh. This made it difficult to reliably measure faecal mass reduction beyond this point. Prior
to this period the mean weekly percentage faecal reduction across all vermifilters was
between 86-95% (this ranged from 23 to 176%). The variation was possibly due to the
mobility of the worms and their changing presence and absence on the faecal mesh. Feeding
continued during Phase 5 and at the end of the 360 day period a total of between 23.7 to 24.7
kg of fresh human faeces had been added to the vermifilters. At the end of Phase 5 the
amount of faeces remaining on the faecal mesh varied from 0.023-0.665 kg; the overall faecal
reduction was therefore 97 to 100%
The different components of the material on the faecal mesh at the end of the experiment
were separated and weighed. The highest mass of undigested faecal matter occurred in the
vermifilter containing coir bedding (coir 0.66 kg; woodchip 0.03 kg; coir and woodchip 0.15
kg; coir, wood and vermicompost 0.37 kg), which suggests that the rate of faeces
consumption by the worms was lower in this system. The highest mass of worms was found
on the faecal mesh when the bedding was a mixture of coir and woodchip (0.66 kg), followed
by the combination of coir, woodchip and vermicompost (0.47 kg), then woodchip (0.46 kg)
whilst the coir bedding had the lowest mass of worms (0.28 kg). It can be inferred from this

that coir alone is a less suitable bedding material, which may be because the worms prefer to
consume the coir compared to the faecal matter. Anecdotal evidence of this has been
highlighted in the general vermicomposting literature (Appelhof, 1997).
The worm density increased in all vermifilters: from 4 kg/m2 to 8.56 kg/m2 in the coir
vermifilter; to 10.10 kg/m2 in the woodchip and coir vermifilter; to 13.19kg/m2 in the
woodchip, coir and vermicompost vermifilter and to 14.48 kg/m2 in the woodchip vermifilter.
This contradicts earlier studies (Yadav et al., 2010; Yadav et al., 2011) which found that a
worm density of 4 kg/m2 was unsustainable. The conditions within their filters and the ones
reported in this paper were very different, i.e. feeding regimes, application of feed, water
flow, and bedding type, all of which could affect the health of the worm population.
Additionally other authors have reported increased worm density over time: Zhao et al.(2010)
reported that worm density increased from 32g/L to 55.7g/L over a period of six months and
Lui et al. (2012) reported a worm density increase from 32 g/L to 46.3 g/L over a seven
month period. The feed in both of these studies was sewage sludge diluted with water,
suggesting that higher worm densities may be sustainable in wetter systems.
Vermicompost was deposited throughout the vermifilter systems, though the majority was
retained in the upper part of the system, i.e. the bedding layer and faecal mesh combined. The
rate of accumulation of vermicompost over the period of the experiment was between 2.7 and
4.1 kg/year (coir 4.1 kg; woodchip 2.7 kg; coir and woodchip 4.0 kg; coir, wood and
vermicompost 4.1 kg). The lower mass accumulated in the system using woodchip was
probably due to the coarser filtering action of the woodchip, with vermicompost being
washed through the bed. Additionally, it could be also attributed to the worms inability to
convert this material into vermicompost. A higher mass of vermicompost (1.6 kg) was found

on the faecal mesh of the vermifilter containing the coir and woodchip bedding (compared to
coir 0.75 kg; woodchip 1.3 kg; coir, woodchip and vermicompost 1.3 kg), because of more
worms inhabiting this part of the system compared to the other vermifilters. This suggests
that this layer was more active in this vermifilter because of the bedding type. A higher
proportion of vermicompost was deposited or formed in the bedding layer of the coir system
(2.43 kg) (compared to woodchip 0.92 kg; coir and woodchip 1.7kg; coir, woodchip and
vermicompost 2.1 kg), which supports the hypothesis that the worms preferred to consume
the bedding in this system rather than the faecal matter.
All of the vermifilter communities remained aerobic and healthy over the 360 days as
assessed by visual and olfactory inspection. The vermifilters were fed 200 g of faecal matter
on 40 days in Phase 5, which is the mean amount of faeces produced per person per day.
Therefore this size of vermifilter (a surface area of 0.1 m2) has the potential to treat the waste
from one person. This would lead to a household system that is considerably smaller than
traditional on-site sanitation systems such as septic tanks or pit latrines.
Effluent quality
The volume of vermicompost in the effluent during Phase 5 was measured as settleable
solids, as the vermicompost was dense and settled out readily (Zhao et al., 2010). The mean
settleable solids in the effluent were highest in the filter containing woodchip (4.9 ± 1.4
mL/L) compared to coir (4.0 ±1.3 mL/L), woodchip and coir (3.8 ± 1.4 mL/L) and woodchip,
coir and vermicompost (3.0 ± 1.1 mL/L) as woodchip was a coarser filter, which led to more
vermicompost being washed through the vermifilter. However, no statistical difference was
found (ANOVA F(3,32)=1.374, p=0.269).

The pH of the influent generally increased as it passed through the vermifilter (influent mean
pH 6.21, effluent mean pH 6.70). Earlier studies have also recorded this (Xing et al., 2010)
and it was expected, as vermicompost is known to have a higher pH than the waste being
processed by the system (Appelhof, 1997). This is thought to be due to the waste being
neutralised by secretions from the worms’ intestines and by the ammonia which is excreted
by worms (Edwards & Lofty, 1997).
Table 2 summarises the mean quality of the influent and effluent from all the boxes.
Phosphate was removed in the system, which contrasts earlier findings (Taylor et al., 2004)
where phosphate levels increased because of the leaching of phosphate from the
vermicompost bedding/filter media. The mean total phosphate removal was 24% in Phase 2,
47% in Phase 3 and 58% in Phase 5, with no difference in the removal rates for the different
bedding types (ANOVA F(3,28)=0.718, p=0.550). A more recent study (Wang et al., 2011)
supports these findings with a mean total phosphate removal of 98.4%, when lower levels of
total phosphorus (5.05 to 9.88 mg/L) were present in the domestic waste water being treated
(Table 2). Phosphorus removal in vermifilters has been attributed to a number of processes,
including the direct absorption of phosphorus by growing cells, the enhanced storage of
phosphorus as polyphosphorus by bacteria in the system and precipitation of phosphorus
(Wang et al., 2011).
Nitrate levels increased as the effluent passed through the system, indicating that nitrification
(conversion of ammonia to nitrate) was occurring. This has also been reported in a previous
study of vermifiltration of domestic sewage (Wang et al., 2011). An earlier study (Taylor et
al., 2004) reported that denitrification also occurred, but the bedding depth in that study was
50 cm which would have better created anoxic conditions than in the present study.

In Table 2 it can be seen that higher COD levels were observed during Phase 1 in the
vermifilters compared to the control systems. No statistical difference was observed when
paired analysis was undertaken (t-test coir p=0.02; woodchip p=0.79; woodchip and coir
p=0.13; woodchip, coir and vermicompost p=0.69), except in the systems using a coir
bedding matrix. This was possibly because coir is inert, coupled with its filtering capacity.
As the majority of the COD is contained in the faecal matter and the bedding layer acts a
filter for this material, it was hypothesised that higher COD levels would be found in the
effluent in the vermifilters with coarser bedding materials. The type of bedding material did
not affect the effluent quality across all phases (ANOVA F (3,80)=1.574, p=0.202). At the
start of each new experimental phase when the feed level was increased, there was a general
decreased in the COD removal until the system stabilised, it increased and then remained
relatively constant. The mean COD removal achieved during Phase 5 (Table 2) was 86-87%,
which is comparable to the 81% removal which was reported in a previous study using a
multi-stage vermifilter (Wang et al., 2011). It should be noted however that the COD in the
influent their study was much lower, as it was rural domestic wastewater. The system tested
also had higher levels of COD removal compared to levels found in septic tanks (47%, Lowe
et al., 2009) and other vermifilter pilot studies (47-58%, Zhao et al., 2010); one vermifilter
study actually found that COD in the effluent increased (Taylor et al., 2003).
The thermotolerant coliform removal across all the boxes ranged from 1-log to 3-log with the
mean removal being 2-log.There was no statistically difference in the removal of
thermotolerant coliform bacteria across all vermifilters (ANOVA F(3,32) =1.02, p=0.399).
The removal reported in this paper are higher than those obtained in a more complex full-
scale worm-based (1-log to 2-log removal, Weiss & Scholes, 2007) and septic tanks (1-log

removal, Lowe 2009), although it may be that in these studies the influent was more dilute.
No experimental studies have been found reporting the bacteriological effluent quality of pit
latrines, possibly due to the difficulty in obtaining a sample.
Implications for on-site sanitation systems
From the data it can be seen that this technology has the potential for on-site sanitation
applications. The worms have the ability to feed on fresh human faeces under flushing
conditions, meaning the vermifilter can be coupled with a low volume pour-flush system,
which brings the additional benefit of a water seal (although it should be noted that the
vermifilter was aerobic and therefore did not smell). Additionally the system proved to be
robust and the worm populations survived periods when they were not fed (Phase 4) and
periods of variable feeding (Phase 5). The conversion of faeces to vermicompost in the
system was between 11 and 18% by mass. Using these conversion values it can be calculated
that annually faeces from 10 people (720 kg) would be converted to between 79 and 130 kg
of vermicompost. This is thought to be a conservatively low estimate of the mass of the
vermicompost generated, as being biologically active it is thought that it would breakdown
further in the system. Furthermore, it should be noted that this system was running for almost
a full year (360 days) and the vermicompost accumulation over time did not cause any
blockages in the system or other practical operational problems.
Results from this paper suggest that at full scale, a system could be very compact, possibly
having an area of 1m2 and depth of 0.9 m to serve a household of 10 people. The performance
of the system in terms of solids reduction and effluent quality looks promising and potentially
superior to existing options for low income families. Although the effluent quality from this
system would not be high enough for direct discharge into a water courses, it is of a standard

where it could be infiltrated into the soil where it would be further treated by the in-situ soil
microorganisms, which is the same strategy used currently with septic tanks and pit latrines
in developing countries. As the system trialled in this paper was extremely simple and
flexible (i.e. different materials of construction could be used) this makes it highly adaptable
for use in developing countries’ contexts. Additionally the worms used are found worldwide,
but other local species could be trialled.
Conclusions
This study was undertaken to test the feasibility of a wet vermifilter for processing fresh
human faeces. The presence of the worms increased the faecal reduction rates compared to
the control systems. The effluent quality from these simple vermifilter was found to be higher
than from septic tanks, and other vermifilter systems. A surprising finding from this study
was the high worm density that the wet system supported. The findings of this paper suggest
that this technology has the potential to develop into a new type of on-site sanitation system
for developing countries; because of the estimated small size of these systems, they would be
particularly suited to high density urban and peri-urban areas.
Acknowledgements
The authors acknowledge the support of the Bill and Melinda Gates Foundation through a
grant (OPP52641) to the London School of Hygiene and Tropical Medicine. The authors also
acknowledge the Centre for Alternative Technology, Wales for hosting this research and to
all of those from this centre who contributed to the research, especially Jamie McQuilkin and
Margaux Taillade.
References

American Public Health Association (APHA). 1992. Standard Methods for the Examination
of Water and Wastewater, American Public Health Association, Washington D.C,
USA.
Appelhof M. 1997. Worms Eat My Garbage, Flowerfield Enterprises, LLC, Kalamazoo,
Michigan, USA, Second Edition.
Biolytix. 2008. South Africa, accessed 6 October 2012, www.biolytix.co.za
Eastman B. R., Kane P. N., Edwards C. A., Trytek L., Gunadi B., Stermer A. L., Mobley J. R.
2001. The effectiveness of vermiculture in human pathogen reduction for USEPA
biosolids stabilization. Compost Science & Utilization, 9 (1) 38-49.
Edwards C. A. and Lofty J. R. . 1997. Biology of Earthworms, Chapman and Hall, London.
Li Y. S., Xiao Y. Q., Qiu J. P., Dai Y. Q., Robin P. 2011. Continuous village sewage
treatment by vermifiltration and activated sludge process. Water Science and
Technology 60(11) 3001-3010.
Lowe K.S., Tucholke M. B., Tomaras J. M. B., Conn K., Hoppe C., Drewes J. E., McCray J.
E., Munakata-Marr J. 2009. Influent consistent characteristics of the modern wastes
streams from single sources. Water Environmental Research Foundation, Alexandria
USA & International Water Association Publishing, London UK.
Lui J., Lu Z., Yang J., Xing M., M. Yu M., Guo M. 2012. Effects of earthworms on the
performance and microbial communities of excess sludge treatment process in
vermifilter. Bioresource Technology, 117 214-221.
Parvaresh A., Movahedian H., Hamidian L. 2004. Vermistabilization of municipal
wastewater sludge with Eisenia fetida. Iranian Journal of Environmental Health
Science & Engineering, 1(2) 43-50.
Robens Centre for Public and Environmental Health. 2004. OXFAM-DelAgua: Portable
water Testing Kit User’s Manual, DelAgua, Surrey, UK.

Simple Waste Water Solutions. 2010. New Zealand, accessed 6 October 2012,
www.swwsnz.co.nz
Taylor M., Clarke W. P., Greenfield P. E. 2003. The treatment of domestic wastewater using
small-scale vermicompost filter beds. Ecological Engineering, 21(2-3) 197-203.
Thye Y. P., Templeton M. R., Ali M. 2011. A critical review of technologies for pit latrine
emptying in developing countries. Critical Reviews in Environmental Science and
Technology, 41 (20) 1893-1819.
Wang L., Guo F., Zheng Z., Luo X., Zhang J. 2011. Enhancement of rural domestic sewage
treatment performance, and assessment of microbial community diversity and
structure using tower vermifiltration. Bioresource Technology, 102 9462-9470.
Weiss S. and Scholes P. 2007. On-site wastewater treatment system environmental discharge
performance appraisal, Environment Bay of Plenty Regional Council, Whakatane,
New Zealand.
Xing M.Y., Li X. W., Yang J. A. 2010. Treatment performance of small-scale vermifilter for
domestic wastewater and its relationship to earthworm growth, reproduction and
enzymatic activity. African Journal of Biotechnology, 9(44) 7513-7520.
Xing M. Y., Yang J. A., Wang Y. Y., Liu J., Yu F. 2011. A comparative study of
synchronous treatment of sewage and sludge by two vermifiltrations using an epigeic
earthworm Eisenia fetida. Journal of Hazardous Materials, 185(2-3) 881-888
Yadav K. D., Tare V., Ahammed M. M. 2010. Vermicomposting of source-separated human
faeces for nutrient recycling. Waste Management, 30(1) 50-56.
Yadav K. D., Tare V., Ahammed M. M. 2011 Vermicomposting of source-separated human
faeces by Eisenia fetida: Effect of stocking density on feed consumption rate, growth
characteristics and vermicompost production. Waste Management, 31(6) 1162-1168.
Zhao L.M., Wang Y.Y. , Yang J., Xing M.Y., Li X.W., Yi D.H., Deng D.H. 2010.

Earthworm microorganism interactions: A strategy to stabilize domestic wastewater
sludge, Water Research 44, 2572-2582.
Figure Captions
Figure 1: Experimental configuration