Professional Documents
Culture Documents
Abstract
Predicting geotextile clogging by ne particles in suspension in water is possible by the use of an empirical model describing the
increase of head loss in the geotextile due to clogging. The present paper describes this model and shows that it is possible to t the model
parameters and nd their physical signicance by means of geotextile clogging tests. This work makes it possible to estimate a maximum
acceptable quantity of injected particles to the geotextile, before clogging. The model results are in good agreement with experimental
results obtained for different geotextiles and soils, as well as for different concentrations of ne particles in water.
r 2005 Elsevier Ltd. All rights reserved.
Keywords: Geotextile; Clogging; Model investigation; Filtration test
1. Introduction
In many ltration applications, non-woven geotextiles
are rst in contact with soft, saturated and ne soils.
Therefore, as long as this soil is not consolidated, water
easily erodes this soil and gets charged with ne particles in
suspension when passing through the geotextile.
The consequence of ne particles accumulation in the
geotextile is a progressive increase of the water head loss in
the geotextile. This increase must not reach a critical value
which will make it difcult for the water to ow through
the lter (corresponding to geotextile clogging). During the
ltration process, the surrounding soil consolidates and
reduces the water velocity and consequently the erosion.
Then clear water may reach the geotextile. The critical
water head loss must not be reached until the soil
consolidation occurs.
Predicting such a critical value of the water head loss
(corresponding to a critical quantity of retained particles in
the geotextile) should then be of great interest.
ARTICLE IN PRESS
Y.H. Faure et al. / Geotextiles and Geomembranes 24 (2006) 1120
12
Nomenclature
a, b
Co
df
k~p
p0
k~s
L
m
m1
n
N
Oo, Of
S
Tg
V
Dp
Dp0
dp
gw
mg
rf
gw VL
,
k~
(1)
(3)
ARTICLE IN PRESS
Y.H. Faure et al. / Geotextiles and Geomembranes 24 (2006) 1120
13
(a)
Dp 1 NDpp NDps ,
Obstructed pipe
(b)
(4)
Dp=Dp0 s exp
m
m1
a
for an accumulation in series;
(6)
Obstructed pipes
(c)
cake
Obstructed canals
Obstructed canals
(d)
Obstructed pipes
Fig. 1. Mechanism of particles accumulation: (a) all pipes are opened (no
one is obstructed), (b) few pipes are obstructed, (c) Most of the pipes are
obstructed and (d) Appearance of )cakes* above completely obstructed
pipes.
Flow
ki
L
Si
(a)
ki
(b)
Li
ARTICLE IN PRESS
Y.H. Faure et al. / Geotextiles and Geomembranes 24 (2006) 1120
14
3. Filtration test
The test assembly consists of the following components
(Faure et al., 1993):
1
m1
m
Fig. 3. Evolution of the pressure loss as a function of m.
3
1
0501
0501
5
6
8
p
elbow
t
PC
1 : pump
2 : suspension
3 : clear water
4 : pressure sensor
5 : geotextile
6 : piped soil in constant water level tank
7 : safety valve
8 : filtration cell
9 : stirrer
Fig. 4. Diagram of the test assembly used for ltering soils in suspension.
ARTICLE IN PRESS
Y.H. Faure et al. / Geotextiles and Geomembranes 24 (2006) 1120
(8)
2000
n1
p / V (kPa/(m/s))
1800
1600
1400
15
mg
,
rf T g
(9)
1200
1000
800
600
400
200
0
0
500
1000
1500
2000
2500
3000
3500
Cumulative mass of injected particles (g/m2)
4000
100
K35
90
C40
80
Finer by mass (%)
S40
70
S50
60
S80
50
40
30
20
10
0
0.1
10
Diameter (microns)
100
Fig. 6. Particle size range of the soil for ltration tests of soil in suspension.
1000
ARTICLE IN PRESS
Y.H. Faure et al. / Geotextiles and Geomembranes 24 (2006) 1120
16
Table 1
Characteristics of geotextiles tested
Geotextile
Structure
mg (g m2)
Tg (mm)
df (mm)
n (%)
Ooa (mm)
Ofa (mm)
CF150
CF300a
CF300b
HB300
SF300
CF600
CF800
Needle punched
Needle punched
Needle punched
Heat bonded
Needle punched
Needle punched
Needle punched
150
292
286
294
269
633
816
1.4
2.6
2.8
0.8
3.9
4.5
5.9
36
26
37
42
31
26
26
88
87.8
88.9
60.1
92.5
84.7
85
68
48
74
24
82
40
41
136
67
113
58
113
49
48
min)
40
HB300
35
Overpressure (kPa)
1 g/l
CF300a
30
25
20
15
10
CF300b
SF300
0
0
50
100
150
200
250
Time (s)
1 N exp
m
m1
N
m
m1
a
1
m
with N tanh b
.
m1
(10)
Results are presented considering dp=V so that the
results might not be affected by the velocity variations
during the test. Before considering the inuence of
the main parameters, characterizing the particles and the
geotextile, the following section proposes to analyze the
physical signicance of the model parameters.
4.1. Physical significance of model parameters
Three parameters are involved in the model: m1, a
and b. It is interesting to give a physical signicance of
each parameter, obtained from its denition (Section 2).
ARTICLE IN PRESS
Y.H. Faure et al. / Geotextiles and Geomembranes 24 (2006) 1120
17
20
soil K35
18
1 g/l
0.5 g/l
0.25 g/l
CF800
16
CF600
overpressure (kPa)
14
12
Co = 0.1g/l
10
8
6
4
2
0
0
200
400
600
800
1000
1200
1400
1600
1800
2000
time (s)
b arctanhN
1
0.9
0.8
b = 0.7
0.7
0.6
N
b = 0.5
0.5
0.4
0.3
b = 0.3
0.2
b = 0.1
0.1
0
0.1
(11)
10
100
m/m1
ARTICLE IN PRESS
Y.H. Faure et al. / Geotextiles and Geomembranes 24 (2006) 1120
18
2000
theoretical model
1800
0.1 g/l
1600
0.25 g/l
p / V (kPa/(m/s))
1400
0.50 g/l
1 g/l
1200
1000
800
600
m1
400
200
0
0
2000
4000
6000
8000
10000
12000
Fig. 10. Experimental and theoretical results for various particle concentrations with soil K35 and geotextile CF600.
p / V (kPa/(m/s))
1600
0.1 g/l
1400
0.25 g/l
1200
0.50 g/l
1 g/l
1000
800
600
m1
400
200
0
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Fig. 11. Experimental and theoretical results for various particle concentrations with soil K35 and geotextile CF800.
Table 2
Coefcients m1, b and aof Le Coqs modelsoil K35
Geotextile
Co (g/l)
m1 (g/m2)
CF600
1
0.5
0.25
0.1
3200
3200
3400
3500
7.5
6.0
5.5
5.1
0.10
0.10
0.10
0.10
CF800
1
0.5
0.25
0.1
2980
3400
2900
2900
7.2
6.7
5.5
6.2
0.10
0.10
0.10
0.10
ARTICLE IN PRESS
Y.H. Faure et al. / Geotextiles and Geomembranes 24 (2006) 1120
19
heat-bonded geotextiles than for needle-punched geotextiles. This leads to the same conclusions concerning m1
as for CF300a compared to CF600 and CF800.
However, no signicant difference can be observed
concerning other parameters a and b: the accumulation in series is very small when mom1 for both geotextiles.
4.4. Influence of soil particles granulometry on model
parameters
Table 3
Inuence of geotextile structure on model parameterssoil C40
Geotextile
Co (g/l)
m1 (g/m2)
HB300
CF300a
1
1
355
1100
15
15
105
106
2000
1800
theoretical model
P / V (kPa/(m/s))
1600
CF300a
1400
HB300
1200
1000
800
600
m1
400
m1
200
0
0
500
1000
1500
2000
2500
3000
3500
4000
Fig. 12. Experimental and theoretical results for ltration behaviour of geotextiles CF300a and HB300 with soil C40.
CF300a_0.5g/l
2000
1800
Soil S80
Soil S50
1600
p/V (kPa/m/s)
1400
1200
1000
800
600
400
m1
m1
200
0
0
500
1000
1500
2000
2500
3000
3500
Mass of injected particles (g/m2)
4000
4500
5000
Fig. 13. Experimental and theoretical results for ltration behaviour of geotextile CF300a with soils S50 and S80.
ARTICLE IN PRESS
Y.H. Faure et al. / Geotextiles and Geomembranes 24 (2006) 1120
20
Table 4
Inuence of particle size distribution on model parameters
Geotextile CF300a
Co (g/l)
m1 (g/m2)
Soil S50
Soil S80
0.5
0.5
1560
260
6.7
3.2
0.10
0.30
References
smaller than the one corresponding to the soil S50
(Table 4). If the particle size of soil increases (S50S80)
parameter b increases (0.10.3).
5. Conclusion
This study showed that it is possible to simulate the
accumulation of ne particles in a geotextile lter with the
help of an empirical model combining accumulation in
parallel and in series.
The model gives the increase in head loss as a function of
the quantity, m, of injected particles and there is a rapid
increase in head loss after a quantity, m1, has been injected
and is related to accumulation of particles in series. A
knowledge of m1 allows one to predict lter clogging
provided that one also knows all the other parameters in
the empirical model.
The geotextile clogging tests allow us to obtain these
parameters as a function of the number of dispersed
particles and soil grain size distribution and provide insight
regarding the physical signicance of all these parameters.
Amneus, J.S., 1965. A new pulp drainage model, its application and
verication. Tappi Journal 48 (11), 641647.
ASTM D-5101-01, 2003. Standard test method for measuring the soil
geotextile clogging potential by the gradient ratio. In: Annual Book of
ASTM Standards, vol. 4.13 Philadelphia, PA, USA.
Faure, Y., Gourc, J.P., Gendrin, P., 1990. Structural study of porometry
and ltration opening size of geotextiles. In: Geosynthetics, Microstructure and Performance, ASTM STP 1076, pp. 102119.
Faure, Y.H., Elamir, A., Farkouh, B., Gendrin, P., Reisinger, P., 1993.
Geotextile lter behaviour with critical ltration conditions. In:
Proceedings of the First International Conference Geo-Filters,
Karlsruhe, pp. 209116.
Faure, Y.H., Kehila, Y., Olivier, F., Paillez, S., 2000. Behaviour of nonwoven geotextiles for ltrating particles in suspension. In: Wolski, W.,
Mlynareck, J. (Eds.), Proceedings of the Geolter2000, Filters and
drainage in geotechnical engineering, pp. 5966.
Giroud, J.P., 1996. Granular lters and geotextile lters. In: Proceedings
of the Second International Conference Geo-Filters, Montreal,
pp. 565680.
Le Coq, 1996. Mise en oeuvre et modelisation de medias ltrants a` base de
bres minerales pour la ltration des huiles en avionique. The`se de
Doctorat de lInstitut national polytechnique de Grenoble, Ecole
Franc- aise de Papeterie et Industries Graphiques de Grenoble,
Novembre 1996, 280p (in French).