See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/344433940

Development of low cost easy to clean filtration technology

Article · September 2020

CITATIONS READS

0 439

1 author:

Ababu Teklemariam Tiruneh

University of Swaziland
43 PUBLICATIONS 255 CITATIONS

SEE PROFILE

All content following this page was uploaded by Ababu Teklemariam Tiruneh on 07 October 2020.

The user has requested enhancement of the downloaded file.

 EWASH & TI Journal, 2020 Volume 4 Issue 3, Page 420-430
Environmental and Water Sciences, Public Health & Territorial Intelligence
Env.Wat. Sci. pub. H. Ter. Int. J.
ISSN Electronic Edition : 2509 - 1069
Acces on line : http://revues.imist.ma/?journal=ewash-ti/

Section : Environnent and Water Sciences

Publication type : Full paper

Development of low cost easy to clean filtration technology

Développement d'une technologie de filtration à faible coût et facile à
nettoyer
Received 16 Jul. 2020 ABABU TEKLEMARIAM TIRUNEH1 & TESFAMARIAM DEBESSAI2
Accepted 18 Aug. 2020
On line 30 Sep. 2020 (1) Department of Environmental Health Science
University of eSwatini
Swaziland.
Email : [email protected]
(2) Department of Chemistry
University of Eswatini
Kwaluseni, eSwatini, Swaziland
Email : [email protected]

KEY WORDS Abstract Filtration of water is an important unit process for treating surface
Water treatment, water sources for potable water treatment applications. The use of both slow
roughing filtration, water sand filter as well as rapid sand filter has to take into account the suspended
supply, water quality, solids load of the raw water which in tropical climate can be very high during the
turbidity, Moringa seed.
rainy seasons. An appropriate treatment alternative that can handle high
suspended solids load while being easy to clean would be desirable filtration
alternative. This research paper is a result of an experimental study of a filtration
technique that employed a two stage filtration with floating plastic media with
coarser and finer media sizes respectively. The media are packed (pressed) during
filtration and later during backwashing the packing is released to allow the media
to self-float. The treatment capability with wide range of raw water turbidities
has been tested and the system has capability to treated water to required low
water turbidities needed for subsequent disinfection. Cleaning is made easier by
the self-floating characteristics of the media. In addition, it was possible to clean
the filter using raw water rather than filtered water. High rate of filtration is
achieved through a combination of locally available coagulant of the seeds of
Moringa tree while simple flocculation using grave column flcooculator has been
employed.

MOTS CLES Résumé La filtration de l'eau est un processus unitaire important pour le
Traitement de l'eau, traitement des sources d'eau de surface pour les applications de traitement de
ébauche filtration, l'eau potable. L'utilisation à la fois d'un filtre à sable lent et d'un filtre à sable
alimentation en eau, rapide doit prendre en compte la charge de matières en suspension de l'eau
qualité de l'eau, turbidité,
brute qui, dans le climat tropical, peut être très élevée pendant les saisons des
graines de Moringa.
pluies. Une alternative de traitement appropriée qui peut gérer une charge
élevée de solides en suspension tout en étant facile à nettoyer serait une
alternative de filtration souhaitable. Ce document de recherche est le résultat
d'une étude expérimentale d'une technique de filtration qui a utilisé une filtration
en deux étapes avec des supports en plastique flottants avec des tailles de
supports plus grosses et plus fines respectivement. Les médias sont emballés
(pressés) pendant la filtration et plus tard pendant le rétrolavage, l'emballage est
libéré pour permettre aux médias de flotter automatiquement.

A.B. TIRUNEH & T. DEBESSAI 420

 EWASH & TI Journal, 2020 Volume 4 Issue 3, Page 420-430
Environmental and Water Sciences, Public Health & Territorial Intelligence
Env.Wat. Sci. pub. H. Ter. Int. J.
ISSN Electronic Edition : 2509 - 1069
Acces on line : http://revues.imist.ma/?journal=ewash-ti/

La capacité de traitement avec une large gamme de turbidités d'eau brute a été
testée et le système a la capacité de traiter l'eau aux turbidités d'eau faibles
requises pour une désinfection ultérieure. Le nettoyage est facilité par les
caractéristiques auto-flottantes du support. De plus, il était possible de nettoyer
le filtre avec de l'eau brute plutôt qu'avec de l'eau filtrée. Un taux de filtration
élevé est obtenu grâce à une combinaison de coagulant localement disponible
des graines de Moringa, tandis qu'une floculation simple utilisant un
flcooculateur à colonne grave a été utilisée.

1. Introduction side filtration or infiltration galleries, storage ponds, etc.

Surface water sources are often polluted with suspensions [7]. Intake and dynamic filters can be classified as
of silt, clay, organic matter and other pollutants. Modern roughing filters but largely act as surface filters where
water supplies are almost always accompanied by most of the deposition occurs within limited depth of the
treatment in order to render the water free from filter [3]. Roughing filters appeared in early time as early
pollutants. The treatment technologies employed as 1804 [21]. This was a river intake stone filter
nowadays are the outgrowth of over 200 years of constructed by John Gibb in Scotland [9].
engineering practice that evolved from empirical to Roughing filters since then are considered as low cost and
process based models over time [1]. In tropical climates, appropriate units that can operated without the addition
the high intensity rain produces runoff water that of chemicals and over a prolonged period of time [10],
contributes to high turbidity of river water sources. When [11].. Although primarily a solids removal technology,
such surface water sources are used as sources of water roughing filters have also been reported for removal of
supply, appropriate treatment alternatives such as slow other pollutants including organic matter and pathogenic
sand filters would require effective pretreatment to microorganisms [12]. Roughing filtration units have been
reduce the turbidity to the level required for operation of adapted for wastewater treatment such as grey water
slow sand filters. Slow sand filter work well in temperate from household with significant reduction in organic
climates because of the raw water quality containing less matter, turbidity and dissolved solids content [8].
turbidity [2] whereas in tropical climates in Africa the high Roughing filters constructed of limestone media have also
turbidity of surface water sources bedevils their been applied for the treatment of leachate waters from
operation. Absence of effective pretreatment interferes landfill site [13].
with the biological treatment of slow sand filters [3]. Roughing filters can be operated as vertical or horizontal
Many cities and towns in eSwatini depend on surface flow filters and often use coarse gravel media. However,
water sources for potable water supply. Conventional the use of other types of media such as coconut, crushed
treatment units require the usage of high dosage of plastic, charcoal, etc., has been reported [5]. Roughing
chemicals for coagulation when faced with high raw water filters rely on the large storage capacity of the coarse size
turbidities. Rapid sand filters may have to be backwashed media that minimizes the need for frequent cleaning.
frequently as a result. Slow sand filters also clog Without the use of chemicals, low rates of filtration are
frequently requiring cleaning and sand replacement at prescribed as sedimentation is the dominant mode of
uneconomic rate. Despite the abundance of surface solids removal. Several experiments revealed roughing
water, rural water supplies often avoid using these filters operate as sedimentation media [14], [15], [16],
surface water sources and opt instead for ground water [17], [18]. The media sizes for roughing filters vary
sources because of the difficulty of treating surface water between 4-20 mm sizes while the filtration rate is
with technologies that can be sustained in rural between 0.3 and 1.5 m/hr. [2]. Operating with such low
environments. The reality of life in rural areas is such the filtration rates turbidities as high as 1000 NTU can be
local situation does not support the technology, operation removed with roughing filters [6]. For a horizontal
and maintenance requirements required of conventional roughing filtration, a filtration length between 5 and 9
water treatment installations [2]. There are many meters would keep the head loss to below 30 cm [3]. The
examples of slow sand filters that were abandoned performance of roughing filters in general depends on the
overtime because of the high raw water turbidities inlet suspended solids concentration, size of the filter
experienced [4], [5]. Roughing filters provide cheaper media, length or depth of filter and rate of filtration [9].
alternative as pretreatment units to slow sand filters At high raw water turbidities, roughing filters often clog
compared to the higher tech and expensive conventional frequently. Cleaning of roughing filters is not easy.
treatment units [6]. Despite the possibility of periodic flushing through
Pretreatment technologies are available and come in the hydraulic drainage at 50-70 m/hr. horizontal flow
form of plain sedimentation tanks, roughing filters, river roughing filters eventually require the gravel being taken

A.B. TIRUNEH & T. DEBESSAI 421

 EWASH & TI Journal, 2020 Volume 4 Issue 3, Page 420-430
Environmental and Water Sciences, Public Health & Territorial Intelligence
Env.Wat. Sci. pub. H. Ter. Int. J.
ISSN Electronic Edition : 2509 - 1069
Acces on line : http://revues.imist.ma/?journal=ewash-ti/

out of the filter and washed [3]. With vertical flow removal of 30-50% when the roughing filtration alone was
roughing filters, the storage space can be partly recovered used without the coagulant. Zeng et al. [30] applied
through backwashing. However, backwashing cannot biological up flow filter for ammonia removal in which
fluidize the heavy gravel media [1]. As a result, repeated 51% removal of ammonia was obtained at a filtration rate
washing of the gravel media may not be effective over of 4 meter per hour. The turbidity removal efficiency was
time besides wasting a lot of water. In addition, the 70%.
removal efficiency is understandably less with greater rate
of filtration in addition to the high head loss. For higher 2. Material and methods
turbidity water, the vertical roughing filters need to be The objective of the research work was to test the
run at lower rates to achieve significant efficiency in effectiveness of high rate roughing filtration as a
filtration [1]. Removal efficacies of horizontal flow pretreatment alternative that makes use of floating
roughing filters decline if the suspended solids to be plastic media with minimum requirement for cleaning as
removed are largely colloidal solids unless chemical well as a locally available coagulant employed to enhance
coagulation is employed [19]. Limited run time, large solid removal through coagulation-flocculation.
space requirement, low efficiency with colloidal A floating plastic media is used that is less dense than
suspension and absence of effective cleaning mechanism water. During filtration, the media is packed with heavy
are some of the problems experienced with the operation weight and is thus prevented from floating. During
of roughing filters. cleaning the packing is lifted and the media and water are
separated allowing dirty water to be drained from the
There are a number of applications of roughing filters bottom followed by mechanical stirring and rinsing with
reported in recent times both in the areas of water small clean water or even raw water used for
treatment as well as wastewater treatment. Bakere et al., backwashing. The filter media that is employed is finer
[8] applied horizontal flow roughing filtration for the than the normally used media size for roughing filters
treatment of wastewater with resulting turbidity removal taking in to account the fact that the filter can be cleaned
of up to 90%, organic matter removal up to 70% of COD relatively easily while its performance is enhanced. The
and solids removal of up to 84% of total solids. According filtration efficiency is evaluated with respect to filtration
to Paciani [23] 85-90% of reduction of iron and media depth, media size, rate of filtration and filter run
manganese using roughing filtration operated at a rate of time. The extent of cleaning that is required is also
1.2 m/hr. with up to 95% removal when the filtration rate examined.
was decreased to 0.3 m/hr. According to Ocheing and
The pilot research setup that was used is shown in
Otieno [24], both broken brick and charcoal performed
Figure 1. It consists of a raw water tank from which water
better than gravel as a roughing filter when operated at a
is pumped up using small electric motor driven pump to
filtration rate of 0.75 m/hr. Nkwonta and Ochieng [25]
an overhead tank that is located at 4-meter height.
summarized the results of roughing filtration research in
The water from the overhead tank flows down to the
which roughing filters performed well when designed and
filters by gravity with an upstream valve along the line
operated properly. The rate of removal is higher during
used for flow control. A chemical dosing tank is located
the wet seasons than the dry season because of the easily above the overhead tank and the dosage is controlled
removal suspended solids concentration being higher in through a small tap provided along the exit pipe. In order
concentration than the dry season which tend to have to facilitate the entry of the chemical by suction through
greater proportion of smaller size colloidal solids. negative pressure, the exit pipe from the overhead tank is
According to Kusuma et al. [26], A combination of up flow positioned above the top water level in the tank so that at
roughing filter and geo textile used as ultra-filter was able the point the coagulant enters, the pressure is less than
to achieve nitrate removal of up to 72 % in the dry season atmospheric and the coagulant is readily sucked into the
and 44 % in the wet season. According to Mahvi et al. pipe (Figure 1). The system self-primes when the
[27], direct roughing filtration in which coagulant dose upstream valve is opened and the coagulated raw water
was applied to enhance solids removal yielded good readily flows from the overhead tank to the gravel column
effluent quality with turbidity less that 2 NTU with a raw flocculator.
water turbidity varying between 200 and 400 NTU.
According to Nkwonta (28], horizontal roughing filters The mixture of raw water and coagulant enters a gravel
perform better than vertical roughing filters due to pipe flocculator which is housed in a 100 mm vertical
unlimited filter length, simple layout and less susceptible Polyethylene pipe. For small water treatment plants grave
than vertical-flow filters to solid breakthroughs caused by bed flocculators provide attractive and low cost
flow rate changes in the filters. According to Hasnian and alternative to conventional mechanical flocculation
Khan [29], an up flow roughing filtration that adopted technologies [20]. The gravel media fills the flocculator
coagulation as pre-treatment prior to roughing filtration pipe to a height of 175 cm. The media sizes were arranged
from coarse to fine gravel size along the vertical flow path
achieved turbidity removal of up to 99% compared to
inside the flocculator pipe with the gravel sizes so
A.B. TIRUNEH & T. DEBESSAI 422
 EWASH & TI Journal, 2020 Volume 4 Issue 3, Page 420-430
Environmental and Water Sciences, Public Health & Territorial Intelligence
Env.Wat. Sci. pub. H. Ter. Int. J.
ISSN Electronic Edition : 2509 - 1069
Acces on line : http://revues.imist.ma/?journal=ewash-ti/

designed to achieve suitable velocity gradient required for the gauge readings for small flows. Refer to the
flocculation (see the results section below for calculation calibration section of the results for the detail.
of the velocity gradient inside the gravel column
A vertical manometer board was setup to monitor head
flocculator). The sizes of the gravel used range from 30
loss development during filtration. Sampling points were
mm at the bottom to 4 mm at the top of the floculator
provided at a height interval of 0.5 m along the filter
pipe. With a filtration rate in the gravel pipe flocculator
columns. Sampling bottles were connected through hoses
ranging between 4 and 40 m/hr, the average velocity
to the various sampling points. The sampling taps were
gradient obtained was 200 sec-1 at the top and 50 sec-1
opened slightly in drops at a minimum disturbance to the
at the bottom.
flow in the filter. The turbidity of the samples was
Two filter columns were provided in series. The first measured using a Hach model Turbidimeter. A backwash
column of 20 cm diameter has a coarse size low density line was provided to all the filter columns. In addition, an
polyethylene that has an effective size of 4.75 mm and additional filter column with Perspex glass was provided
uniformity coefficient of 1.05. The depth of the filter is to visualize the process of filtration and backwashing.
1.80 m. The second filter has a finer high density
polyethylene media of effective size 2.36 mm and
uniformity coefficient of 1.30. It was also packed to a
height of 1.80 m. Figure 2 shows the sieve analysis result
of both media used in the experiments. Both media have
good floating characteristics apart from being inert and
strong. At the top of both media a piston packing made of
perforated grid wire and a weight was used to keep the
media from floating. At the bottom of the filter columns,
gravel was provided to a depth of 0.5 m to prevent media
loss. Sampling taps were provided along the height of the
filter columns at intervals of 0.5 m to monitor head loss
and filtered water quality with respect to depth of
filtration.

Fig. 2: Grain size analysis result of the coarse and fine

polyethylene media used in the experiment performed
in this research
Source: Result of the pilot plan experiment employed in this research

Several tests were run using different raw water sources

and different rates of filtration. The coagulants used were
seeds of the Moringa Stenopetala. A jar test was run to
determine the optimum coagulant for the various raw
water sources tested. Apart from the seeds of the
Moringa tree, aluminum sulphate has also been used in
the experiment to compare the results with the Moringa
seeds. When the filters became dirty as is evidenced by
water level rise in the filters, filtration was stopped and
cleaning was done by lifting the packing from above,
Fig. 1: Water treatment setup of the pilot laboratory
floating the plastic media and displacing the dirty water
experimental apparatus used in this research consisting
and eventually using small back wash water for rinsing.
of raw water system, coagulant suction feed, gravel
The quality of the wash water was monitored during
column flocculator, filter and flow and head loss
cleaning.
measurements
Source: Pilot plant laboratory apparatus used in the research 3. Results and Discussion
An orifice flow meter was manufactured locally for flow 3.1. Calibration of orifice flow meter
measurement. The orifice has internal opening of 10 mm The orifice flow meter was constructed onsite because of
made inside a 25.4 mm GI pipe. A calibration test was run lack of availability of flow measuring device. As shown in
to determine the head discharge relationship of the Figure 3, the 1 in inch diameter GI pipe was cut into two
orifice meter. Inclined manometer was used to magnify and an orifcie plate with a circulatr openining of 10 mm

A.B. TIRUNEH & T. DEBESSAI 423

 EWASH & TI Journal, 2020 Volume 4 Issue 3, Page 420-430
Environmental and Water Sciences, Public Health & Territorial Intelligence
Env.Wat. Sci. pub. H. Ter. Int. J.
ISSN Electronic Edition : 2509 - 1069
Acces on line : http://revues.imist.ma/?journal=ewash-ti/

was welded in betweenn the two pipe sections. Once the The laminar head loss equation can be used to calculate
orifice flow meter was constructed, a claibration curve the head loss as [22]:
was prepared by running the filter sysem at different flow
rates and recording the manamomete head readings
between the upstream and downstream sections of the
orficice.

Fig. 3: Construction of orifice flow meter used in this

research
Source: Pilot plant laboratory apparatus used in this research

Figure 4 shows the calibration curve which is constructed

by regression analysis of the orifice flow head relationship
given the calibration equation of the form:

Fig. 4: Calibration curve of the orifice flow meter

Where Q is the flow in the pipe outlet initially recorded Source: Result of the pilot plant experiment performed in this research
in liters per second but converted to meters per hour in
Where h is the head loss in the gravel column flocculator,
terms of the apparent filtration rate at filter number 2
f is the friction factor,  is the shape factor (0.8), l is the
that has internal diameter of 30 cm. H is the difference
length of the flocculator column, d is the gravel size and v
between manometer readings immediately upstream and
is the face velocity in the flocculator. The friction factor f
downstream of the orifices. The regression parameters
can be calculated from [22]:
were determined by transforming the Q – H variables to
logarithm and running a linear regression analysis. The
coefficient of determination, R2, is 0.998. Additional
construction involving inclined manometer was
connected to the orifice manometer for magnifying the
Where RN is the Reynolds’s number of flow in the
flow especially at low flow rates which are difficult to
flocculator and other variables are as defined before. The
calculate with accuracy from the rather small vertical
gravel column flocculator has three different size gravel
manometer readings.
media arranged from coarse size at the top which is 20
mm gravel, middle section having gravel size of 10 mm
3.2. Velocity Gradients in the gravel column
and the bottom section with gravel size of 4 mm. Figure 5
flocuulator
shows the calculated velocity gradient for the three
The velocity gradient in gravel column flcocculator can be different flocculator sections with respect to the filtration
calculated from the equation [22]: rate in filter number 2 which is taken as the reference
filtration rate.

3.3. Solids removal modeling

Where G is the velocity gradient,  is the density, g is The removal of solids in the filters was studied with
acceleration due to gravity, h is the head loss, Q is the respect to both the initial solids load and the depth of
flow rate,  is the coefficient of dynamic viscosity of filtration. According to the analysis of the data obtained
water, V is the volume of the flocculator and α is the the solids removal is first order with respect to depth of
porosity of the gravel used in the flocculator. filtration and the rate constant varies with the initial
solids load on the filter as a second order rate reaction.

A.B. TIRUNEH & T. DEBESSAI 424

 EWASH & TI Journal, 2020 Volume 4 Issue 3, Page 420-430
Environmental and Water Sciences, Public Health & Territorial Intelligence
Env.Wat. Sci. pub. H. Ter. Int. J.
ISSN Electronic Edition : 2509 - 1069
Acces on line : http://revues.imist.ma/?journal=ewash-ti/

The equations that describe these rate variations are also removal of solids was obtained with respect to the lower
known as Fick’s law ([6] and take the form [6]: rate of filtration, namely, at a filtration rate of 1.5 m/hr.

Where C is the solids concentration at depth h and k is the

first order rate constant for removal of solids with depth.
C0 is the initial turbidity. The rate constant k varies as a
second order rate with respect to the initial turbidity
according to the equation:

The constants a and b are determined by linear

regression. Figure 6 shows a typical curve regression line
of solids removal with depth in which the rate of solids
removal varies as first order rate reaction. Figure 7 on the Fig. 6: First order solids removal model for filter #1 at
other hand is the regression model result of the variation initial kaolin clay turbidity of 350 NTU
of the rate constant with respect to the initial solids load Source: Result of the pilot plant experiment performed in this research
for the kaolinite clay suspension used as an artificial
turbidity.

Fig. 7: Second order modeling of the variation of the

rate constant with the initial turbidity
Source: Result of the pilot plant experiment performed in this research

Fig. 5: Variation of velocity gradient in the gravel column At this rate the water can be taken for disinfection
flocculator without further filtration. The use of coagulants will be
Source: Result of the pilot plant experiment performed in this research helpful if the filters have to be operated at a higher rate.
The performance of filtration is greatly enhanced by the
The removal of solids without the use of coagulants is
use of the optimum coagulant dose of 50 mg/L as is
shown in Figure 8 in which it can be seen that the rate of
shown in Figure 10. At a high rate of filtration of 4 m/hr.,
removal varies inversely with the rate of filtration.
a depth of 50-70 cm of filtration produced turbidity less
Without the use of coagulants, it is clear that the roughing
than 5 NTU that would be adequate quality for
filter will have to be run at a slower rate in order to have
subsequent disinfection. With coagulant lesser depth of
significant removal of solids. This is exemplified by the
media would be needed to achieve greater efficiency of
removal of solids at a filtration rate of 1.5 m/hr. being
filtration. In addition, the cleaning requirement would be
significantly higher than one when the filter is operated at
less difficult as shorter depth of media is used for
a higher rate of 4 m/hr. Continuation of filtration on the
filtration.
second filter produced the solids removal curves shown in
Figure 9. It is also seen once again that satisfactory

A.B. TIRUNEH & T. DEBESSAI 425

 EWASH & TI Journal, 2020 Volume 4 Issue 3, Page 420-430
Environmental and Water Sciences, Public Health & Territorial Intelligence
Env.Wat. Sci. pub. H. Ter. Int. J.
ISSN Electronic Edition : 2509 - 1069
Acces on line : http://revues.imist.ma/?journal=ewash-ti/

The effect of addition of coagulant on a lake water source solids content of the backwash water with the time of
with stable colloids is seen in Figure 11 in which filter size washing and as can be seen from the figure, within a
of 4.75 mm was used. Because of the coarser size media, washing time of 4 to 5 minutes, the media is sufficiently
the gain in performance is moderate, ranging between backwashed to be put back into operation. This result
20-40% solids removal depending on the rate of filtration suggests an important recommendation of using low
compared to only 10% removal without using a coagulant. quality water such as the raw water for backwashing
For the coarse size filter, therefore, a solids performance instead of wasting filtered water. Of course, because the
improvement of up to 30% is obtained with the media is not very clean following cleaning with raw water,
recommendation that such filters need to be operated at there has to be wastage of water for slightly longer period
lower flow rates to maximize performance. On the other at the commencement of filtration following backwashing
hand, Figure 12 shows the solids removal over filter #2 for before the filtered water is directed to subsequent
the same lake water source containing stable colloid. In treatment.
this case, a solids removal as high as 95% could be
For raw water sources such as river water that contain
obtained over the depth of filtration with filtration rate
relatively faster settling suspensions, it is possible to
being as high as 4 m/hr. compared to only about 25%
reduce the turbidity in general to below 5 NTU over a
solids removal without the use of coagulants. The
filtration depth of 1 meter and by using coarser media of
improvement in performance for the finer media is
4.75 mm effective size while the rate of filtration can be
therefore considerably higher than the coarser media. A
as high as 6-10 m/hr. The same performance can be
suitable combination would be to use the coarser media
achieved using finer media size filtration of 2.6 mm
as a sort of roughing pretreatment ahead of the finer
effective size over a relatively smaller depth of 0.5 cm
media so that the overall frequency of cleaning would be
operating at a similarly high rate of filtration. The tradeoff
reduced.
in the case of the finer media is the lesser silt storage
Pilot trial with a river water source whose turbidity has a capacity it has compared to the coarse size filter media.
faster settling characteristics because of the silt load of
Raw water sources that have very stable colloids such as
the rivers reveal that the effect of the coagulant is marked
the lake water source used in the pilot trial require a
for both the coarse and fine media filters. This is shown in
lower rate of filtration if the coarse media is to be used
Figures 13 and 14. According to the data on Figure 13, for
while the depth of filtration should be larger. The finer
the coarse media filter solids removal ranging between
media size filter is rather more effective at filtration range
60-90% is achieved for the range of filtration rates
as high as 5-10 m/ hr. with a filtration depth varying
indicated. This is compared to the 40% solids removal
between 1 and 1.5 meters.
when coagulant is not being used. The finer media
likewise achieved filtration removal of over 80% The use of coagulant to assist filtration as the
compared to only 20% removal when no coagulant is experimental data demonstrate is of great help in
used. Since large part of the turbidity is removed in the increasing the efficiency of filtration. Without the use of
first coarse media filter, the task for the second filter is coagulants, the roughing filters would have to run at
considerably reduced. Compared to the raw water source much lower filtration rate which necessitate using lager
from a lake that contains very stable colloids, the water area of media. Monitoring of flocs after the water passed
from the river source can be easily treated at a high rate through the gravel column flocculator showed that macro
of filtration. size filterable flocs were formed and the flocs have much
faster settling characteristics than the original raw water.
Cleaning of the filter media is facilitated because of the
floating characteristics of the polyethylene media used. The progression of head loss depended in general on the
The low density polyethylene, being lighter than the high filtration rate and raw water quality. The initial head loss
density polyethylene is the easier to clean. In general, the is low, between 3 and 10 cm even at a higher filtration
procedure for cleaning involved lifting the packing weight rate. The head loss is increased to between 10-50 cm of
from the top and allowing the floating media to suspend water after about 48 hours of continuous operation
at the top. It is necessary to lift the packing high up so depending upon the raw water turbidity. Generally head
that the media is completely suspended at the top. Then loss is the determining factor for the filter run time.
after the dirty water that is separated from the media and
The filtered water quality can have turbidity as low as 5
lying at the bottom is drained using the bottom valve.
NTU with a combination of the coarse and fine size media
After this, it is possible to use clean or possibly raw water
used in series over different ranges of water quality
for back washing with a period of time ranging between 1
tested. There was not much change in pH and
and 5 minutes.
conductivity as a result of the low dosage of Moringa
An experiment was conducted to test the effectiveness of coagulant used in the study. However, if the filter is
washing the media using raw water instead of filtered stopped for a longer period of time, the deposited
water. Figure 15 shows the progression of the suspended Moringa seeds start to decompose and anaerobic
A.B. TIRUNEH & T. DEBESSAI 426
 EWASH & TI Journal, 2020 Volume 4 Issue 3, Page 420-430
Environmental and Water Sciences, Public Health & Territorial Intelligence
Env.Wat. Sci. pub. H. Ter. Int. J.
ISSN Electronic Edition : 2509 - 1069
Acces on line : http://revues.imist.ma/?journal=ewash-ti/

condition arises that result in odour problem. It is,

therefore, necessary that the filters run continuously. If
there has to be stoppage of flow in the filter for longer
period of time some reason, the filters have to be drained
and possible cleaned.

Fig. 10: The effect of Moringa coagulation on the

efficiency of filtration with depth
Source: Result of the pilot plant experiment performed in this research

Fig. 8: Removal of turbidity with depth at different rates

of filtration for filter #1
Source: Result of the pilot plant experiment performed in this research

Fig. 11: The effect of coagulant on solids removal of

filter #1 with a lake water source
Source: Result of the pilot plant experiment performed in this research

Fig. 9: Removal of turbidity with depth at different rates

of filtration for filter #2
Source: Result of the pilot plant experiment performed in this research

Fig. 12: The effect of coagulant on solids removal of

filter #2 with a lake water source
Source: Result of the pilot plant experiment performed in this research

A.B. TIRUNEH & T. DEBESSAI 427

 EWASH & TI Journal, 2020 Volume 4 Issue 3, Page 420-430
Environmental and Water Sciences, Public Health & Territorial Intelligence
Env.Wat. Sci. pub. H. Ter. Int. J.
ISSN Electronic Edition : 2509 - 1069
Acces on line : http://revues.imist.ma/?journal=ewash-ti/

4. Conclusion
The experimental results demonstrate the effectiveness
of roughing filtration of different raw water sources with
the use of floating plastic media run from low to high
rates of filtration. High percentage removal of solids
ranging between 60 to 95% can be obtained when the
filtration is assisted with the use of Moringa seed as
coagulant coupled with the provision of gravel column
flocculator. The use of coagulant enables high rate of
filtration as well as high percentage removal of solids over
relatively shorter depth of the media generally varying
between 0.5 m to 1.0 m. Coarse media size filters will
have to be naturally run at lower filtration rates to
produce results that are comparable to fine media size
Fig. 13: Solids removal in roughing filter assisted with
Moringa coagulant for river water source (Filter #1) filters. However, they have greater silt storage capacity
Source: Result of the pilot plant experiment performed in this research
and as a result greater filter runs compared to the finer
size media filter. As demonstrated in the pilot
experiments, for raw water that contain stable colloidal
suspensions, the combination of coarse and fine media
arranged in series produces a filtered water quality below
turbidity of 5 NTU that would be suitable for potable
supply after disinfection. For river water sources that have
suspended solids with relatively faster settling
characteristics, both coarse and fine media filters perform
satisfactorily for solids removal at reasonable high rates of
filtration.
The removal of solids in the filter with respect to depth is
adequately modeled as a first order reaction whose rate
constant varies depending on the initial raw water
turbidity. It was demonstrated through modeling that the
variation of rate constant for solids removal with respect
Fig. 14: Solids removal in roughing filter assisted with to initial raw water turbidity can be modeled as a second
Moringa coagulant for river water source (Filter #2) order reaction. In other words, the rate constant for solids
Source: Result of the pilot plant experiment performed in this research
removal over depth varies inversely with the initial raw
water solids concentration.
The use of floating plastic media facilitates cleaning even
though the filters are run at high rates and the filter runs
as a result may be shortened. Lifting of the weight packing
and automatic floating of the media that follows allows
most of the deposited solids to separate out of the media
and be drained from the bottom. A small amount of
rinsing back wash water would be needed as a result. It
has been demonstrated through the pilot trials that it is
possible to backwash the filter with raw water in which
recovery of the solids storage capacity was obtained
within 4 to 5 minutes. As an alternative washing
mechanism, it is possible to repeatedly fill the filter and
lift the packing and drain the separated solids from the
bottom.
While the pilot experiment show that this technique of
Fig. 15: Backwash water solids content with backwashing filtration overcomes some of the problems of roughing
done using raw water filtration, it is also possible to expand the scope of
Source: Result of the pilot plant experiment performed in this research application of this filtration to irrigation and wastewater
treatment. The floating plastic media can be locally
A.B. TIRUNEH & T. DEBESSAI 428
 EWASH & TI Journal, 2020 Volume 4 Issue 3, Page 420-430
Environmental and Water Sciences, Public Health & Territorial Intelligence
Env.Wat. Sci. pub. H. Ter. Int. J.
ISSN Electronic Edition : 2509 - 1069
Acces on line : http://revues.imist.ma/?journal=ewash-ti/

obtained as part of recycling of plastic wastes with countries: A review. International Journal of Physical
reasonably low prices. Sciences, 4(9), 455-463.
[12]. Barman, R. N., Mukhopadhaya, B., Majumder, M.,
Contribution of the authors Roy, P. K., and Mazumdar, A. 2008. Estimation and
Both authors have equally contributed to the calculation of a relationship between dispersion
development of the research concept, data collection, number, Reynolds number, porosity and hydraulic
analysis and interpretation of results as well as the writing gradient in horizontal roughing filter. Journal of
of the research paper. Agricultural, Food, and Environmental Sciences.
[13]. Affam, A.C. and Adlan, M.N. 2013. Operational
References performance of vertical up flow roughing filter for
[1]. Deshpandel, V. and Salkar, V.D. 2018. Experimental pre-treatment of leachate using limestone filter
evaluation of gravel bed up-flow roughing filter for media. Journal of Urban and Environmental
pre-treatment of rural community water treatment. Engineering, 7(1), 117-125.
Civil Engineering Research Journal, 3(2), 44-52. [14]. Wegelin, M., Boller, M., Scgertenleib, R. 1987.
[2]. Wegelin, M. 1996. Surface water treatment by Particle removal by horizontal rouging filtration.
roughing filters. Swiss Federal Institute for Journal of Water Supply Research and Technology-
Environmental Science and Technology, Dubendorf, Aqua, 36, 80-90.
Switzerland. [15]. Collins, M.R., Wstersund, C.M. Cole, J.O. and
[3]. Wegelin, M., Schertenleib, R. and Boiler, M. 1991. Rocarro, J.V. 1994. Evaluation of Roughing filters
The decade of roughing filters: development of a design variables. AWWA Research Foundation and
rural water treatment process for developing American Water Works Association.
countries. Journal of Water SRT-Aqua, 40(5), 304- [16]. Pardon, M.O. 1989. Treatment of turbid surface
316. water for small community water supplies, PhD
[4]. Graham, N., and Collins N. (Eds.). 1994. Advances in Dissertation, Robens Institute, University of Surrey,
slow sand and alternative biological filtration. John Surrey, England.
Wiley & Sons Ltd Chichester, England, pp. 360-378. [17]. Del Mundo Jusi, O. 1987. Pre-treatment applications
[5]. Al-Baidhani, J.H. and Khadafy, Z.H. (2016). Treatment of horizontal flow coarse media ore filtration, M.Sc.
of water and wastewater by using roughing filter Thesis, Asian Institute of technology, Bangkok,
technology of local materials. International Journal Thailand.
of Current Engineering Technology, 6(6), 2192-2198. [18]. Boller, M. 1993. Filter mechanisms in roughing
[6]. Mukhopadhay, B., Majumder, M., Barman, R.N., filters. Journal of Water Supply research and
Kumar, P. and Mazumder, A. 2009. Verification of Technology- Aqua, 42, 174-185.
filter efficiency of horizontal roughing filter by [19]. Mahvi, A.H. Moghaddam, A., Nasseri, S. and Naddafi,
Weglin’s design criteria and Artificial Neural K. 2004. Performance of a Direct Horizontal
Network. Drinking Water Engineering and Science, 2, Roughing Filtration (DHRF) System in Treatment of
21-27. Highly Turbid Water. Iranian Journal of
[7]. Smet, J. and Visscher, J.T. 1989. Pre-treatment Environmental Health Science and Engineering,
Methods for Community Water Supply. IRC, 11(1), 1-4.
Netherlands, 1989. [20]. Kardile, J.N. 1978. Development of Simple and
[8]. Bakare, B. F., Mtsweni, S. and Rathilal, S. 2019. Pilot Economic Filtration Methods for Water Supplies.
study of a horizontal roughing filtration system Ph.D. Thesis, Nagpur University, India.
treating grey water generated from a peri-urban [21]. Baker, M.N. 1981. The quest for pure water.
community in Durban, South Africa. Journal of Water American Water Works Association.
Reuse and Desalination, 09(3), 330-337 [22]. Okun D. and Shulz, C. 1984. Surface water treatment
[9]. Nkwonta, O. 2010. A comparison of horizontal for communities in developing countries. U.S.
roughing filters and vertical roughing filters in Agency for International Development. Technical
wastewater treatment using gravel as a filter media. Report No. 89.
International Journal of the Physical Sciences, 5(8), [23]. Pacini, V. 2005. Removal of iron and manganese
1240-1247. using biological roughing up flow filtration
[10]. Onyeka I. Nkwonta, O., Olufayo, O., Ochieng, G., technology. Water Research, 399, 4463-4475.
Adeyemo, J. and Otieno, F. 2010. Turbidity removal: [24]. Ochieng G.M.and Otieno, F.A.O. 2006. Verification
Gravel and charcoal as roughing filtration media. of wegelin design criteria for horizontal flow
South African Journal of Science, 106(11/12), 1-5. roughing filters with alternative filter materials.
[11]. Nkwonta, O. and Ochieng, G. 2009. Roughing filter Watersa Vol. 32.
for water pre-treatment technology in developing

A.B. TIRUNEH & T. DEBESSAI 429

 EWASH & TI Journal, 2020 Volume 4 Issue 3, Page 420-430
Environmental and Water Sciences, Public Health & Territorial Intelligence
Env.Wat. Sci. pub. H. Ter. Int. J.
ISSN Electronic Edition : 2509 - 1069
Acces on line : http://revues.imist.ma/?journal=ewash-ti/

[25]. Nkwonta, O. and Ochieng, G. 2009. Roughing filter [28]. Nkwonta, O. 2010. A comparison of horizontal
for water pre-treatment technology in developing roughing filters and vertical roughing filters in
countries: A review. International Journal of Physical wastewater treatment using gravel as a filter media.
Sciences, 4 (9), 455-463. International Journal of the Physical Science, 5(8),
[26]. Kusuma, M.N., Oktavia, O., Fitriani, N. and Hadi, W. 1240-1247.
2016. Combination up flow roughing filter in series [29]. Hasnain, G. and Khan Z. 2014. Treatment of high
with geotextile to removal total nitrate in dry and turbidity water through plain and coagulated up-
rainy season. ARPN Journal of Engineering and flow roughing filter. Desalination and water
Applied Sciences, 11(13), 8155-8160. Treatment, 52(4-6), 1035-1043.
[27]. Mahvi, A.H., Moghaddam, M. A. , Nasseri, S. and [30]. Zeng, J., Chabi, K., Hu, Y., Zhang, S. and Yu, X. 2020.
Naddafi, K. 2004. Performance of a direct horizontal Ammonium removal of biological roughing filter for
roughing filtration (dhrf) system in treatment of rural drinking water pretreatment. Journal of Water
highly turbid water. Iranian J Env Health Sci Eng, Science and Technology: Water Supply,
1(1), 1-4. https://doi.org/10.2166/ws.2020.171.

A.B. TIRUNEH & T. DEBESSAI 430

View publication stats