Enhancing sustainability of household water filters by mixing metallic iron with porous 
materials 
1 
2 
3 
4 
5 
6 
7 
Noubactep C.*(a,c), Caré S.(b) 
(a) Angewandte Geologie, Universität Göttingen, Goldschmidtstraße 3, D - 37077 Göttingen, Germany; 
(b) Université Paris-Est, Laboratoire Navier, Ecole des Ponts - ParisTech, LCPC, CNRS, 2 allée Kepler, 77420 
Champs sur Marne, France; 
(c) Kultur und Nachhaltige Entwicklung CDD e.V., Postfach 1502, D - 37005 Göttingen, Germany. 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
* corresponding author: e-mail: cnoubac@gwdg.de; Tel. +49 551 39 3191, Fax: +49 551 399379 
(Accepted: 08 Jun 2010) 
Abstract: This study conceptually discusses the feasibility of enhancing the sustainability of 
conventional iron/sand filter (Fe0/sand filter) for safe drinking water by partially or totally 
substituting sand (quartz) by porous materials. Relevant materials included activated carbon, 
dolomites, limestone, pumice, sandstone, and zeolites. The rational was to use the internal 
volume of porous additives as storage room for in-situ generated iron oxyhydroxides (iron 
corrosion products) and thus delay time to filter clogging. Based on previous works a filter 
with a volumetric Fe0:quartz ratio of 51:49 was used as reference system. The reference 
system is clogged upon Fe0 depletion. Results showed that totally substituting quartz by 
pumice particles having a porosity of 80 % yields to a residual porosity of 41 %. This 
encouraging result suggested that the possibility of using Fe0/MnO2/pumice systems for a 
synergic promotion of Fe0 reactivity (by MnO2) and filter permeability (by pumice) should be 
investigated in more details. 
Keywords: Drinking water, Iron/sand filter; Long-term reactivity; Pumice; Zerovalent iron. 
Capsule: Partially or totally replacing quartz by porous materials is a potential efficient tool 
to lengthen iron/sand filter service life. 
 1
1  Introduction 25 
26 
27 
28 
29 
30 
31 
32 
33 
34 
35 
36 
37 
38 
39 
40 
41 
42 
43 
44 
45 
46 
47 
48 
49 
50 
Safe drinking water is becoming worldwide an increasingly scarce resource mostly due to 
industrialization [1-3]. The situation is exacerbated in rural areas of the developing world 
where no centralized drinking water system is available. Therefore, efforts have been made to 
develop simple, efficient and affordable methods for safe drinking water production at 
household or small community level [4-9]. An ideal water treatment system for developing 
countries should remove all possible biological and chemical contaminants in a single-stage 
filtration process (Prerequisite 1). Only reverse osmosis membranes have been reported to 
achieve prerequisite 1. However, this high cost technology is not always affordable for the 
populations in need [1,10-12]. Therefore, efforts to develop simple, efficient and affordable 
filters (achieving prerequisite 1) for households and/or small communities is an area of 
ongoing active research [2,3,12-15]. 
In recent years, a great deal of work has been devoted at identifying suitable, low-cost, and 
readily available materials to be used in efficient household filters. Potential media include 
activated alumina, agricultural by-products (e.g. rice hulls), apatite, clay minerals, granular 
activated carbon (GAC), industrial by-products, iron-oxide (coated sands), manganese-oxide 
(coated sands), metallic iron (Fe0), peat and peat moss, phosphate rocks, seaweeds and their 
derivatives, wood chips, and zeolites [2,3,16,17]. Some enumerated materials are regionally 
readily available (e.g. apatite, clay minerals, zeolites) at no-expense or at a fairly low cost. 
However, only metallic iron (Fe0) is universally available. Next to its universally availability, 
the superiority of Fe0 is justified by the fact that during the dynamic process of its aqueous 
oxidative dissolution (iron corrosion), several classes of biological and chemical contaminant 
can be removed from water [14,18-23]. In particular Fe0 beds could quantitatively remove 
aqueous inorganic (e.g. MoVI) and organic (e.g. non-polar carboxylated organics) substances 
that are not readily removed by iron oxyhydroxides [24,25] Therefore, upon proper design, 
Fe0 filters necessarily achieve prerequisite 1. 
 2
Ideally, a household filter should reduce contaminant concentrations to acceptable levels 
whilst retaining adequate permeability and reactivity over extended time periods (e.g. 12 
months). Due to the volumetric expansive nature of iron corrosion [26-29], designing iron 
filters with long-term adequate permeability is a challenge for the scientific community [30]. 
In fact, very efficient Fe
51 
52 
53 
54 
55 
56 
57 
58 
59 
60 
61 
62 
63 
64 
65 
66 
67 
68 
69 
70 
71 
72 
73 
74 
75 
76 
0 filters (e.g. the 3-Kolshi filter) for safe drinking water have been 
tested in Bangladesh and Nepal for arsenic removal [31-34] but were abandoned because of 
service lives of only 6 to 24 weeks [35-37]. Hussam [37] reported that 3-Kolshi filters were 
“highly functional, but not sustainable” as the filters experienced permeability loss after 3 to 6 
months. The alternative to 3-Kolshi filters was an improved filter called SONO filters 
[13,33,37]. The heart of SONO filters is a manufactured porous composite iron matrix (CIM). 
CIM is manufactured from Fe0 and resulted filters could work for up to 11 years [37]. 
The present theoretical work is part of ongoing efforts, aiming at reviving conventional Fe0 
filters. The objective is to design efficient stand-alone Fe0 filters for households and small 
communities. A recent article [30] has shown that mixing non porous sand (quartz) with Fe0 is 
an efficient way to lengthen Fe0 filter service live while saving up to the half amount of used 
Fe0. It was shown that a filter with 51 vol-% Fe0 necessarily has a longer service live than the 
conventional iron filter (100 % Fe0), likely with the same efficiency. In the present work 
calculations will be made to access the possibility to further optimize filter efficiency by 
partly or entirely replacing quartz by porous materials. Relevant materials (Supporting 
Information) included activated carbon and natural minerals (e.g. MnO2, TiO2, zeolites) and 
rocks (e.g. dolomite, limestone, pumice, sandstone). For the sake of clarity, the Fe0/sand/H2O 
system will first be presented. 
2  The Fe0/sand/H2O system 
2.1  Descriptive aspects 
Aqueous contaminant removal in the presence of Fe0 (e.g. in Fe0/H2O systems) is an 
heterogeneous reaction ideally involving five steps: (i) contaminant mass-transfer from the 
 3
bulk solution to the Fe0 surface, (ii) contaminant adsorption on the Fe0 surface, (iii) chemical 
reaction at the Fe
77 
78 
79 
80 
81 
82 
83 
84 
85 
86 
87 
88 
89 
90 
91 
92 
93 
94 
95 
96 
97 
98 
99 
100 
101 
102 
0 surface, (iv) desorption of the reaction products from Fe0 surface, and (v) 
mass-transfer of the reaction products into the bulk solution [38,39]. The kinetics of the 
process of contaminant removal in a Fe0/H2O system are necessarily dependent on the 
availability of the Fe0 surface as well as the rate of mass-transfer of the contaminant to the 
reactive Fe0 surface (Fe0 accessibility). Accordingly, mixing an inert material (e.g. sand) and 
Fe0 is coupled with a decrease of the contaminant removal rate as the pathway to the Fe0 
surface is lengthened (Assertion 1: mixing sand and Fe0 decreases Fe0 accessibility). 
The process of aqueous iron oxidative dissolution at pH > 4.0 is characterized by the 
expansive nature of iron corrosion products (iron oxides and oxyhydroxides or simply iron 
oxyhydroxides). Depending on the nature of iron oxyhydroxide, a volumetric expansion of up 
to 6.4 has been reported [28]. In other words, an iron oxyhydroxide molecule may occupy a 
volume up to 6.4 times the volume of atomic Fe in the lattice structure. Accordingly, a pure 
iron bed (100 % Fe0) will clog sooner than a bed containing the same iron mixed with inert 
material (e.g. sand). Therefore, mixing sand and Fe0 should be regarded as a tool to sustain 
permeability of an iron bed (Assertion 2: mixing sand and Fe0 increases bed service life). 
Assertion 1 and Assertion 2 clearly show that the role of sand in an Fe0 bed is antagonistic. 
Accordingly, a well-designed Fe0/sand bed must conciliate limited Fe0 accessibility and 
extended service life. It is obvious that a critical volumetric Fe0:sand ratio exists above which 
bed clogging will occur upon Fe0 depletion. That is the Fe0 proportion for which in-situ 
generated iron oxyhydroxides will fill the inter-granular pore volume of the initial bed. The 
results of such calculations for the Fe0:quartz is recalled in the next paragraph. 
2.2  Literature review 
Iron has been mixed with inert materials (including gravel and sand) since the early stage of 
investigations regarding the applicability of iron walls for groundwater remediation [40,41]. 
However, there has been no systematic study designed to rationalize the effects of sand on the 
 4
efficiency of Fe0 beds. Most researchers employ varying Fe0:sand ratios for filtration beds in 
the laboratory and in the field on a pragmatic basis [42-48]. Fe
103 
104 
105 
106 
107 
108 
109 
110 
111 
112 
113 
114 
115 
116 
117 
118 
119 
120 
121 
122 
123 
124 
125 
126 
127 
128 
0:sand ratios are often given in 
weight percent without information on the available pore volume nor precise data on further 
operational conditions including the high of the Fe0/sand bed (Table 1). As a result, 
controversial reports for the same systems have been reported. For example, while 
investigating As removal by Fe0 in packed beds, Lien and Wilkin [46] concluded that Fe0 
alone performed better than a 50:50 sand:Fe0 mixture. In contrary, Westerhoff et al. [47] 
observed higher As removal in a Fe0:sand bed than with Fe0 alone.  
2.3  The importance of volumetric ratios 
The main concern of available data on Fe0/sand/H2O systems is their comparability. When the 
Fe0:sand ratios are given in percent it is not always specified whether it is volumetric or 
weight based. However, given the large difference between the density of sand (< 2,000 
kg/m3) and iron (7,800 kg/m3), it should always explicitly said whether given percent are 
volumetric (vol-%) or weight (wt-%). Even in some cases (e.g. ref. [42]), the Fe0:sand ratio is 
given without the high of the reactive zone nor the mass of either sand or iron.  
2.3.1  Rationale for use the volumetric ratio 
The sole calculation that could be founded to rationalize used Fe0:sand ratios is given by 
Leupin et al. [45] and mentioned by Gottinger [15,49]. The authors used 2.5 g of Fe0 which 
occupied a volume of 0.32 cm3 and lead to 4.78 g of Fe(OH)3 with a volume of 1.35 cm3 (i.e. 
4.78 cm3 as reported in the original work). Then the internal porosity of sand (40 %) is 
considered to estimate the volume of sand necessary to contain the 1.35 cm3 of iron oxides. 
Leupin et al. [45] came to the conclusion that at least 9.9 cm3 (i.e. 35 cm3) of sand should be 
used per gram of iron to avoid the clogging of more than a third of the void volume. 
The calculus of Leupin et al. [45], considered as a rough estimate by the authors, has to be 
improved because they have not taken into account the inter-granular voids in the Fe0/sand 
bed. In fact, the filter has to be considered as a granular medium with two types of porosity: 
 5
(i) the internal porosity of particle (as mentioned in ref. [45]) and (ii) the inter-granular voids 
as a function of the compactness of the granular medium. A recent result [30] has allowed 
modelling the clogging of Fe
129 
130 
131 
132 
133 
134 
135 
136 
137 
138 
139 
140 
141 
142 
143 
144 
145 
146 
147 
148 
149 
150 
151 
152 
153 
154 
0/quartz bed. Quartz is an inert, non porous material which can 
not contribute to porosity loss. This modelling gives the evolution of the inter-granular void 
as a function of the proportion of reactive Fe0 in the Fe0:quartz mixture. The results showed 
that it is possible to avoid the clogging of the filter by mixing Fe0 particles with quartz 
particles. In particular, it was shown that 51 vol-% Fe0 (25 wt-% when Fe0 is mixed with 
quartz) is the critical proportion for with system clogging (porosity loss) will occur upon Fe0 
depletion. The present paper attempts to rationalize the admixture of Fe0 particle with porous 
particles, including sandstone as used by Leupin et al. [45]. The filling of the total porosity 
(inter-particles voids and internal porosity) will be considered. The next section will discuss 
the efficiency of a Fe0/sand filter. 
3  Relevant parameters influencing the efficiency of the Fe0/sand filter 
3.1  Permeability of granular materials 
Water permeability is essential for Fe0/sand filter efficiency because it determines the rate of 
flow and, thus, the filtration ability. Many models have been proposed to relate the water 
permeability of granular medium to their microstructure characteristics [50-54]. These models 
show that permeability is controlled by the packing, shape, sorting (particle size distribution), 
and porous structure of used granular material, but it appears that porosity and tortuosity of 
granular materials are two of the primary factors that control water flow process [53-58]. In 
the most general way, permeability depends on the total porosity which is the ratio of the total 
volume of voids to the total volume of material. More importantly, however, permeability 
depends on the way in which the total porosity is distributed and thus on the effective 
porosity. The effective porosity characterizes the degree to which available pores are 
interconnected. As a rule, if all pores of a granular material are well interconnected, the total 
porosity is equal to the effective porosity. Tortuosity can be defined as the ratio of the real 
 6
155 
156 
157 
158 
159 
160 
161 
162 
163 
164 
165 
166 
167 
168 
169 
170 
171 
172 
173 
174 
175 
176 
177 
178 
179 
180 
length that the water travels inside a filter to the thickness of the filter. For a mixture of non 
porous particles, the tortuosity can be simply estimated as a function of the total porosity 
considering the solid grains as spherical inclusions in a fluid phase [59,60]. 
An idealized conventional Fe0/sand filter is made up of spherical iron and quartz particles of 
approximately equal diameter. The non porous spherical particles provide ample, unrestricted 
void spaces that are free from smaller grains and are very well linked. Consequently, a 
Fe0/sand filter initially has a high permeability which is related to the total porosity as water 
flow will occur through the inter-particular voids. In case of using porous particles with 
poorly interconnected internal voids (as for pumices, ref. [57] and ref. therein), the water 
permeability is essentially related to the effective porosity due to the inter-particular voids. In 
this case the tortuosity of the mixing only depends on the effective porosity. Indeed, part of 
the water may remain apparently stagnant in the internal pores of particles and slowly diffuses 
out of the pores. 
3.2  Efficiency of a Fe0/sand filter 
Due to the expansive nature of iron corrosion the voids are progressively filled by (i) in situ 
generated iron oxyhydroxides, (ii) immobilized contaminants and (iii) in-situ generated 
biofilms [12,23,61,62]. The ability to accurately predict the process dependent evolution of 
the permeability of a Fe0/sand filter depends on a detailed description of the processes 
yielding to porosity loss. In this study, the contribution of biofilms and contaminants to 
porosity loss is not considered. The volumetric expansive nature of iron corrosion is 
considered as the sole important path. A second simplification is necessary as iron 
oxyhydroxides are also porous in nature. It is considered that the pores of generated iron 
oxides are well interconnected. In this case, the filter permeability is solely modified by the 
process of oxide formation (pore filling). The permeability of a Fe0/sand filter typically 
decreases from the beginning of the operation to the time of complete pore filling by iron 
oxyhydroxides (porosity equal zero). 
 7
The suitability of Fe0 as reactive medium for drinking water filters relies on two essential 
characteristics: (i) the interactions of corroding iron with contaminants (adsorption, co-
precipitation/enmeshment, oxidation, reduction), and (ii) the improved size exclusion by 
virtue of the expansive nature of iron corrosion. The efficiency of an iron filter depends on 
several factors including particle size of Fe
181 
182 
183 
184 
185 
186 
187 
188 
189 
190 
191 
192 
193 
194 
195 
196 
197 
198 
199 
200 
201 
202 
203 
204 
205 
206 
0, initial contaminant concentration, and influent 
pH. These factors are determinant to the time at which the initial porosity is reduced to zero. 
The present discussion will not address how these factors affect Fe0 reactivity. The evolution 
of filter permeability due to filling of porosity by iron oxyhydroxides which is the most 
important parameter determining filter service life will be solely addressed. 
3.3  Aim of the paper 
Permeability variation in an iron filter is important for predicting filter service life. 
Understanding the dependence of filter permeability on the extent of iron consumption would 
be decisive in designing filters. An ideal iron filter is a random pack of identical spheres in a 
column. The porosity of such an ideal system has a fundamental value of 36 % assuming a 
compactness of 64 % with soft vibration [30,54-56]. The actual porosity value for a Fe0/sand 
filter will depend on the size distribution of filling particles but this porosity is a good 
approximation for such mixing. As iron consumption progresses, residual Fe0 particles 
become compacted and cemented, and the initial porosity (ideally 36 %) progressively 
decreases down to zero. At porosity zero, the filter is clogged (Figure 1). Cemented Fe0 
particles form a continuous frame which had been called “cake”. The described dynamic 
evolution of the porosity is somewhat equivalent to the formation of quartz (porosity zero) 
from sandstone (porosity 40 %) by diagenesis [63] (Figure 2). 
As described above, clogging is inherent to Fe0 filters working at near neutral pH values 
(more exactly at pH > 4.0). Therefore, lengthen filter service live initially depends on the 
ability to create additional space for in-situ generated iron oxyhydroxides while maintaining 
filter efficiency. An obvious possibility is to use porous material which internal porous 
 8
207 
208 
209 
210 
211 
212 
213 
214 
215 
216 
217 
218 
219 
220 
221 
222 
223 
224 
225 
226 
227 
228 
229 
230 
231 
structure may store in-situ generated iron iron oxyhydroxides (Fig. 1) leading to increase the 
residual effective porosity and thus the residual permeability of the mixing. 
4  Efficiency of the Fe0/sand/porous materials filter 
4.1 Background 
This work tests the hypothesis, that porous minerals and rocks have the potential to enhance 
the sustainability of metallic iron (Fe0) filter. Therefore, it is not important to work with a 
well-characterized porous material. The concept of “critical porosity” [63] will be used for the 
discussion. Critical porosity is defined as “the porosity above which the rock can exist only as 
a suspension”. The critical porosity, separates two fundamentally different domains – one of 
consolidated, frame-supported rocks, and another of fluid-supported suspensions (see ref. [63] 
for more details). Tab. 2 summarizes the critical porosity value of various rock types. The 
major feature from Tab. 2 is that the potential exists to find rocks of porosity 0 (quartz) to 80 
% (pumice). The highest value from Tab. 2 will be used to qualitatively demonstrate the 
feasibility of extending filter service live by replacing sand by porous material in a 
conventional Fe0/sand filter. Some suitable porous materials are presented in Supporting 
Information. It can be noticed that the permeability of porous materials is linked to their 
porous structure (pore connectedness). In general, due to diffusive processes in porous 
structures, slower kinetics of water flow will be observed comparative to a conventional 
Fe0/sand filter. Regardless from the pore connectedness, it is expected that material porosity 
will effectively store in-situ generated iron oxyhydroxides. 
This section presents modelling works to evaluate the feasibility of lengthening the service 
live of conventional Fe0/sand filter for safe drinking water by partially or totally substituting 
sand (quartz) by a porous material. The delay time to filter clogging is evaluated from the 
evolution of the total residual porosity. Results for pumices are presented and discussed. It 
can be noticed that the calculus can be made with other porous particles (by changing its 
 9
232 
233 
234 
235 
236 
237 
238 
239 
critical porosity); the results will be slightly different but the general conclusions will remain 
identical. 
4.2  Design and modelling 
A reactive zone (rz) of Fe0 with a thickness Hrz is introduced in the fine sand layer of a 
conventional biosand filter (Fig. 3) [7,64]. The reference reactive zone is made up of 51 vol-
% Fe0 and 49 vol-% quartz. This system was demonstrated to allow increased filter efficiency 
compared to a conventional reactive zone with 100 % Fe0 [30]. The characteristics of the 
reactive zone are given in Tab. 3. All particles are considered spherical in shape with an 
average diameter of 1.2 mm. The specific weight of Fe0 and quartz are Feρ = 7800 kg/m3 and 
ρ
240 
241 
242 
243 
244 
245 
sand = 2650 kg/m3 respectively. The biosand filter is supposed to work under anoxic 
conditions. Thus, Fe3O4 is the sole iron corrosion product with a coefficient of volumetric 
expansion  equal to Vη Fe3O4 / VFe = 2.1.  
Quartz particles are replaced by porous particles in order to increase the initial total porosity 
Φ0 of the reactive zone given by: 
 f ppppQuartz49Fe5100 .)%%( ϕ+Φ=Φ −     (1) 246 
247 
248 
249 
where the porosity Φ0 (51%FΕ−49%sand) = 1 – C. C is the compactness of the granular material (C 
= 0.64) [54-56] and “1 – C” corresponds to the inter-particular voids (porosity of the mixing 
without porous particles); 
250  ϕpp  (-) is the critical porosity of the porous particles (e.g. 0.8 for pumice, Tab. 2); 
251 
252 
253 
254 
255 
 f pp  (-) is the porous particle volume fraction determined by VVf pppp = with Vpp the 
volume of the porous particles and the V the volume of the reactive zone.  
The filling of the porosity by iron oxyhydroxides can be estimated from a simplified 
modelling (Figure 1) based on the following assumptions:  
(i) uniform corrosion: the diameter reduction of the particle is the same for all the Fe particles,  
 10
(ii) the compactness C and then the initial porosity Φ0 (51%FΕ−49%sand)  remain constant. The 
volume of the granular material is not modified by the corrosion process: no pressure induced 
by rust formation around Fe particles and no compaction of the Fe
256 
257 
258 
259 
260 
261 
262 
263 
264 
0/sand mixture during the 
corrosion process. 
(iii) iron oxyhydroxides are fluid enough to progressively fill the available pore space 
between particles (Φ0 (51%FΕ−49%sand)) and the porosity of the porous particles. The porous 
particles are enough close to the iron particles to be filled by the iron oxyhydroxides.  
Under these assumptions, the residual porosity of the mixing Fe0-quartz-porous particles is 
given by:  
 Φ
−π−η−ΦΦ
0
33
0 V
RR0341N1=
.
)(./)..(     (2) 265 
266 
267 
268 
where R0 (m) is the initial radius of the iron particle, R (m) is the residual radius of the 
consumed iron particle and η the coefficient of volumetric expansion. 
The proportion of consumed iron (% consumed Fe) is given by:  
 ).(%
R0
RR0100consumedFe 3
33 −=      (3) 269 
270 
271 
272 
273 
274 
275 
276 
277 
278 
279 
4.3  Results and discussion for Fe0/quartz/pumice mixture  
The total residual porosity of the mixture Fe0/quartz/porous particle is given in Tab. 4 and in 
Fig. 4 and 5. Calculations shown that it is possible to increase the total residual porosity by 
replacing quartz by porous pumice. For instance, totally substituting quartz by pumice 
particles having a porosity of 80 % yields to a residual porosity of 41 % when iron is depleted 
(Tab. 3, Fig. 4). For a given residual porosity (case 1, Fig. 4), more iron is consumed by 
replacing quartz by pumice. For instance for Φ/Φ0 = 60 %, the % consumed Fe is about 40 % 
for 100 % quartz (point A) and 68 % for 100 % pumice (point B). Accordingly, the long-term 
reactivity of the filter is improved. Furthermore, for a given % consumed Fe, the residual 
porosity Φ/Φ0 increases with increasing proportion of pumice. For 68 % consumed Fe0 (case 
 11
280 
281 
282 
283 
284 
285 
286 
287 
288 
289 
290 
291 
292 
293 
294 
295 
296 
297 
298 
299 
300 
301 
302 
303 
304 
305 
2), the residual porosity is about 32 % for 100 % quartz (point C) and 60 % for 100 % pumice 
(point B). The long term permeability of the filter may be improved by up to 45 %.  
As shown previously, water permeability depends on the effective porosity and not on the 
total porosity. Some water can remain more or less stagnant in the internal porous structure of 
a material exhibiting low interconnectivity of pores (or voids). Water flow in the internal 
porous structure is always slow compared to water flow in the inter-particular voids. 
Nevertheless the residual effective porosity of the mixture is increased because corrosion iron 
products are at least partly stored in the porous structure of particles and not totally in the 
inter-particular voids (see Fig. 1 and 4). Assuming that the porosity of the porous particles 
(pumice) is totally filled by iron oxyhydroxides, the residual porosity of the inter-particular 
voids (effective porosity) is 30% leading to better permeability. 
Experimental studies to validate the efficiency of porous materials to lengthen conventional 
Fe0/sand filter service life will consist to evaluate the actual permeability related to the storage 
of iron oxyhydroxides in porous particles. For instance, 3D imaging by X-ray micro 
tomography will be an efficient tool to evaluate the residual porosity of the inter-particular 
voids. 
4.4  Generalization: Fe0/quartz/porous materials 
The results discussed above for a pumice exhibiting a porosity of 80 % can be extended to 
pumices of various porosities and any other porous particles including activated carbons, 
dolomites, manganese oxides, rock salts, sandstone, and zeolites. Figure 5 depicts the general 
trend of the results are similar on the sole basis of the porosity (Supporting Information). As a 
rule, the total residual porosity of the filter increases with increasing particle porosity. For 
example a material with a grain porosity of 90 % still exhibits 44 % of the initial porosity 
upon Fe0 depletion, while a material with a grain porosity of 20 % shows a residual porosity 
of only 15 %. In practical laboratory experiments, it may be difficult to homogeneously mix 
materials of very different densities. Remember that the discussion is based on the volumetric 
 12
filling of the reactive zone by Fe0 and additives (quartz and porous materials) having the same 
size. The used mass of individual porous materials should be calculated from tabulated 
density’s values (Tab. 2). 
306 
307 
308 
309 
310 
311 
312 
313 
314 
315 
316 
317 
318 
319 
320 
321 
322 
323 
324 
325 
326 
327 
328 
329 
330 
331 
Beside the porosity, further physical and chemical properties of individual porous materials 
should be considered on a case-specific basis. For example, while rocks and activated carbons 
are inert in water, MnO1+x will be reductively dissolved in Fe0/sand/H2O systems. The 
reductive dissolution of MnO1+x by FeII (and Fe0) is necessarily coupled with a volumetric 
variation (MnO1+x is reduced to MnOOH or dissolved MnII). However, the discussion of the 
resulting volumetric variation and its impact on the filter permeability is over the scope of this 
communication. On the other hand, while using activated carbons as porous additive to 
sustain filter permeability, the pore size distribution of individual materials should be 
carefully considered. Remember that porous materials are mainly used as magazine for iron 
oxyhydroxides. Therefore, mesoporous materials are like more suitable than microporous 
materials because available pores must be accessible to iron oxyhydroxides. 
5 Concluding remarks 
The theoretical principles essential to experimentally test the use of porous materials to 
sustain Fe0/sand filters are exposed in this study. This approach was rendered possible by 
revisiting the nature of Fe0-based filters. It was recalled that a filtration system basically 
works on the size-exclusion principle [14]. Accordingly, at any time, the pore space must be 
large enough to enable the expansion/compression cycle inherent to iron corrosion. Iron 
corrosion products (iron oxyhydroxides) reduce filter porosity and thus permeability while 
increasing size-exclusion efficiency. The first task to enable long-term iron corrosion was to 
replace a part of Fe0 by an inert material (e.g. quartz) [30]. Table 1 clearly shows that the 
conventional approach of expressing the proportion of Fe0 by a weight percent is not 
consistent with the fact that pore volume availability is discussed (expansive nature of iron 
corrosion). For example, 75 wt-% Fe0 corresponds to 50.5 vol-% which is almost the 
 13
threshold value for which system clogging will occur upon Fe0 depletion (in a Fe0:quartz 
system). In other words, for systems with less than 50.5 vol-% Fe
332 
333 
334 
335 
336 
337 
338 
339 
340 
341 
342 
343 
344 
345 
346 
347 
348 
349 
350 
351 
352 
353 
354 
355 
356 
357 
0, system clogging due to 
iron corrosion alone is not likely to occur. Moreover, for much lower Fe0 contents, 
contaminant breakthrough is likely. Accordingly, some available data are to be re-evaluated. 
For example, Bi et al. [48] reported on decline in the reactivity of Fe0 for trichloroethylene 
reduction when the iron content fell below 50 wt-% (25.4 vol-%; only one halve of the 
threshold value). The discussion above has shown that this iron content is necessarily 
insufficient for quantitative contaminant removal. 
The present study positively tests the possibility to extend Fe0 reactivity by replacing quartz 
by porous materials. Substituting quartz by porous material increased the residual porosity 
from 0 to 40 % upon Fe0 depletion. It is expected, that different porous materials (minerals 
and rocks) will be tested worldwide for use in Fe0 filters. The option to synthesize porous 
materials combining permeability and reactivity sustention should be carefully checked for 
commercial Fe0 filters. However, the initial goal of this communication is to encourage 
researchers to improve Fe0/sand filter efficiency by adding readily available porous material. 
Based on the universal availability of both Fe0 and suitable porous additives, it can be 
anticipated that the Millennium Development Goal of halving “by 2015 the proportion of 
people without sustainable access to safe drinking water” can still be achieved [65].  
Supporting Information 
Selected suitable porous materials to sustain Fe0/sand filters (8 pages) 
Acknowledgments 
Dipl.-Geol. Manuela Morales Demarco (Geoscience Centre of the University of Göttingen) is 
gratefully acknowledged for the insightful discussion on the porosity of rocks and minerals. 
Sven Hellbach (student research assistant) is acknowledged for technical assistance. The 
manuscript was improved by the insightful comments of anonymous reviewers from 
Chemical Engineering Journal. 
 14
References 358 
359 
360 
361 
362 
363 
364 
365 
366 
367 
[1] M. Wegmann, B. Michen, T. Luxbacher, J. Fritsch, T. Graule, Modification of ceramic 
microfilters with colloidal zirconia to promote the adsorption of viruses from water, 
Water Res. 42 (2008), 1726–1734. 
[2] A.H. Malik, Z.M. Khan, Q. Mahmood, S. Nasreen, Z.A. Bhatti, Perspectives of low cost 
arsenic remediation of drinking water in Pakistan and other countries, J. Hazard. Mater. 
168 (2009), 1–12. 
[3] S. Wang, Y. Peng, Natural zeolites as effective adsorbents in water and wastewater 
treatment, Chem. Eng. J. 156 (2010), 11–24. 
[4] M.W. LeChevallier, K.K. Au, Water treatment and pathogen control: process efficiency in 
achieving safe drinking water. Geneva, Switzerland: WHO, 2004, http://www.cepis.ops-368 
oms.org/bvsacg/e/cd-cagua/guias/c.referencias/10.watreatpath.pdf (Access 06.06.2010). 369 
370 
371 
372 
373 
374 
375 
376 
377 
[5] M.A. Shannon, P.W. Bohn, M. Elimelech, J.G. Georgiadis, B.J. Marinas, A.M. Mayes, 
Science and technology for water purification in the coming decades Nature 452, 301–
310. 
[6] M.D. Sobsey, C.E. Stauber, L.M. Casanova, J.M. Brown, M.A. Elliott, Point of use 
household drinking water filtration: A practical, effective solution for providing 
sustained access to safe drinking water in the developing world. Environ. Sci. Technol. 
42 (2008), 4261–4267. 
[7] CAWST (2009), The Centre for Affordable Water and Sanitation Technology 
http://www.cawst.org/ (Access 07.06.2010). 378 
379 
380 
381 
382 
383 
[8] M.I. Litter, M.E. Morgada, J. Bundschuh, Possible treatments for arsenic removal in Latin 
American waters for human consumption, Environ. Pollut. 158 (2010), 1105–1118. 
[9] WHO/UNICEF 2010. Progress on sanitation and drinking-water 2010 update, Geneva, 
Switzerland: WHO/UNICEF. http://www.unwater.org/downloads/JMP_report_2010.pdf 
(Access 06.06.2010). 
 15
384 
385 
386 
387 
388 
389 
390 
391 
392 
393 
394 
395 
396 
397 
398 
399 
400 
401 
402 
403 
404 
405 
406 
407 
408 
[10] L. Leong, Removal and inactivation of viruses by treatment processes for potable water 
and wastewater – a review, Water Sci. Technol. 15 (1983), 91–114. 
[11] S. Madaeni, The application of membrane technology for water disinfection, Water Res. 
33 (1999), 301–308. 
[12] A. Zularisam, A. Ismail, R. Salim, Behaviours of natural organic matter in membrane 
filtration for surface water treatment – a review, Desalination 194 (2006), 211–231. 
[13] S. Ahamed, A.K.M. Munir, A. Hussam, Groundwater arsenic removal technologies 
based on sorbents: Field applications and sustainability, In Handbook of Water Quality 
and Water Purity, Elsevier Inc., Chapter 16 (2009), 379–417. 
[14] C. Noubactep, A. Schöner, P. Woafo, Metallic iron filters for universal access to safe 
drinking water, Clean 37 (2009), 930–937. 
[15] A.M. Gottinger, D.J. Wild, D. McMartin, B. Moldovan, D. Wang, Development of an 
iron-amended biofilter for removal of arsenic from rural Canadian prairie potable water, 
(2010) Retrieved from: http://www.mainstreamwater.com/Gottinger%20et%20al.pdf. 
(Acces 22.04.2010). 
[16] D. Mohan, C.U. Pittman Jr., Arsenic removal from water/wastewater using adsorbents – 
A critical review, J. Hazard. Mater. 142 (2007), 1–53. 
[17] V.K. Gupta, Suhas, Application of low-cost adsorbents for dye removal – A review. J. 
Environ. Manag. 90 (2009), 2313–2342. 
[18] A.D. Henderson, A.H. Demond, Long-term performance of zero-valent iron permeable 
reactive barriers: a critical review, Environ. Eng. Sci. 24 (2007), 401–423. 
[19] C. Noubactep, Processes of contaminant removal in “Fe0–H2O” systems revisited. The 
importance of co-precipitation, Open Environ. J. 1 (2007), 9–13. 
[20] C. Noubactep, A critical review on the mechanism of contaminant removal in Fe0–H2O 
systems, Environ. Technol. 29 (2008), 909–920. 
 16
409 
410 
411 
412 
413 
414 
415 
416 
417 
418 
419 
420 
421 
422 
423 
424 
425 
426 
427 
428 
429 
430 
431 
432 
[21] C. Noubactep, The suitability of metallic iron for environmental remediation, Environ. 
Progr. (2009), doi: 10.1002/ep.10406. 
[22] C. Noubactep, P. Woafo, Elemental iron (Fe0) for better drinking water in rural areas of 
developing countries, In Merkel B.J., Hasche-Berger A. (Eds.) Uranium in the 
Environment. Springer, Berlin, Heidelberg; (2008), 121–130. 
[23] V. Zolla, F.S. Freyria, R. Sethi, A. Di Molfetta, Hydrogeochemical and biological 
processes affecting the long-term performance of an iron-based permeable reactive 
barrier, J. Environ. Qual. 38 (2009), 897–908. 
[24] S. Goldberg, Influence of Soil Solution Salinity on Molybdenum Adsorption by Soils, 
Soil Sci. 174 (2009), 9–13. 
[25] Hanna, Sorption of two aromatic acids to iron oxides: Experimental study and modeling, 
J. Colloid Interf. Sci. 309 (2007), 419–428. 
[26] K. Bhargava, A.K. Ghosh, Y. Mori, S. Ramanujam, Model for cover cracking due to 
rebar corrosion in RC structures, Eng. Struct. 28 (2006), 1093–1109. 
[27] T. El. Maaddawy, K. Soudki, A model for prediction of time from corrosion initiation to 
corrosion cracking, Cem. Concr. Compos. 29 (2007), 168–175. 
[28] S. Caré, Q.T. Nguyen, V. L'Hostis, Y. Berthaud, Mechanical properties of the rust layer 
induced by impressed current method in reinforced mortar, Cement Concrete Res. 38 
(2008), 1079–1091. 
[29] S. Caré, Q.T. Nguyen, K. Beddiar, Y. Berthaud, Times to cracking in reinforced mortar 
beams subjected to accelerated corrosion tests, Mater. Struct. 43 (2010), 107–124. 
[22] C. Noubactep, S. Caré, F. Togue Kamga, A. Schöner, P. Woafo, Extending service life of 
household water filters by mixing metallic iron with sand, Clean (2010) (Submitted - 
clen.201000177). 
 17
433 
434 
435 
436 
437 
438 
439 
440 
441 
442 
443 
444 
445 
446 
447 
448 
449 
450 
451 
452 
453 
454 
455 
456 
457 
[31] A.H. Khan, S.B. Rasul, A.K.M. Munir, M. Habibuddowla, M. Alauddin, S.S. Newaz, A. 
Hussam, Appraisal of a simple arsenic removal method for groundwater of Bangladesh. 
J. Environ. Sci. Health A35 (2000), 1021–1041. 
[32] T.K.K. Ngai, S. Murcott, R.R. Shrestha, B. Dangol, M. Maharjan, Development and 
dissemination of Kanchan™ Arsenic Filter in rural Nepal, Water Sci. Technol. Water 
Supply 6 (2006), 137–146. 
[33] A. Hussam, A.K.M. Munir, A simple and effective arsenic filter based on composite iron 
matrix: Development and deployment studies for groundwater of Bangladesh, J. 
Environ. Sci. Health A 42 (2007), 1869–1878. 
[34] T.K.K. Ngai, R.R. Shrestha, B. Dangol, M. Maharjan, S.E. Murcott, Design for 
sustainable development – Household drinking water filter for arsenic and pathogen 
treatment in Nepal, J. Environ. Sci. Health A42 (2007), 1879–1888. 
[35] D. Pokhrel, B.S. Bhandari, T. Viraraghavan, Arsenic contamination of groundwater in 
the Terai region of Nepal: An overview of health concerns and treatment options, 
Environ. Int. 35 (2009), 157–161. 
[36] A. Neku, N. Tandulkar, An overview of arsenic contamination in groundwater of Nepal 
and its removal at household level, J. Phys. IV (2003), 941–944. 
[37] A. Hussam, Contending with a Development Disaster: SONO Filters Remove Arsenic 
from Well Water in Bangladesh, Innovations 4 (2009), 89–102. 
[38] H.-M. Hung, F.H. Ling, M.R. Hoffmann, Kinetics and mechanism of the enhanced 
reductive degradation of nitrobenzene by elemental iron in the presence of ultrasound, 
Environ. Sci. Technol. 34 (2000) 1758–1763. 
[39] H. Zhang, M. Jiang, Z. Wang, F. Wu, Decolorisation of CI Reactive Black 8 by zero-
valent iron powder with/without ultrasonic irradiation. Color. Technol. 123 (2007) 203–
208. 
 18
458 
459 
460 
461 
462 
463 
464 
465 
466 
467 
468 
469 
470 
471 
472 
473 
474 
475 
476 
477 
478 
479 
480 
481 
482 
[40] P. G. Tratneyk, Putting corrosion to use: Remediating contaminated groundwater with 
zero-valent metals, Chemistry & Industry 1 July (1996) 499–503. 
[41] S.F. O´Hannesin, R.W. Gillham, Long-term performance of an in situ "iron wall" for 
remediation of VOCs, Ground Water 36 (1998), 164–170. 
[42] P. Westerhoff, J. James, Nitrate removal in zero-valent iron packed columns, Wat. Res. 
37 (2003) 1818–1830. 
[43] D.I. Kaplan, T.J. Gilmore, Zero-valent iron removal rates of aqueous Cr(VI) measured 
under flow conditions, Water Air Soil pollut. 155 (2004), 21–33. 
[44] O.X. Leupin, S.J. Hug, Oxidation and removal of arsenic (III) from aerated groundwater 
by filtration through sand and zero-valent iron, Wat. Res. 39 (2005) 1729–740. 
[45] O.X. Leupin, S.J. Hug, A.B.M. Badruzzaman, Arsenic removal from Bangladesh tube 
well water with filter columns containing zerovalent iron filings and sand, Environ. Sci. 
Technol. 39 (2005) 8032–8037. 
[46] H.-L. Lien, R.T. Wilkin, High-level arsenite removal from groundwater by zero-valent 
iron. Chemosphere 59 (2005) 377–386. 
[47] P. Westerhoff, M. DeHaan, A. Martindale, M. Badruzzaman, Arsenic adsorptive media 
technology selection strategies. Water Quality Res. J. Canada 41 (2006) 171–184. 
[48] E. Bi, J.F. Devlin, B. Huang, Effects of mixing granular iron with sand on the kinetics of 
trichloroethylene reduction, Ground Water Monit. Remed. 29 (2009), 56–62. 
[49] A.M. Gottinger, Chemical-free arsenic removal from potable water with a ZVI-amended 
biofilter. Master thesis, University of Regina (Saskatchewan, Canada) (2010) 90 pp. 
[50] G.V. Chilingar, Relationship between porosity, permeability, and grain-size distribution 
of sands and sandstones. Develop. Sediment. 1 (1964), 71–75. 
[51] W. Wempe, G. Mavko, Three distinct porosity domains defined physically, 
hydraulically, electrically, and elastically, The Leading Edge 17 (2001), 198–199. 
 19
483 
484 
485 
486 
487 
488 
489 
490 
491 
492 
493 
494 
495 
496 
497 
498 
499 
500 
501 
502 
503 
504 
505 
506 
507 
[52] L.C.M. Félix, L.A.B. Muñoz, Representing a relation between porosity and permeability 
based on inductive rules, J. Petrol. Sci. Eng. 47 (2005), 23–34. 
[53] S.-H. Lee, H.Y. Jo, S.-T. Yun, Y.J. Lee, Evaluation of factors affecting performance of a 
zeolitic rock barrier to remove zinc from water, J. Hazard. Mater. 175 (2010), 224–234. 
[54] E. Zaman, P. Jalali, On hydraulic permeability of random packs of monodisperse 
spheres: Direct flow simulations versus correlations. Physica A389 (2010), 205–214. 
[55] D.J. Cumberland, R.J. Crawford, Handbook of powder technology. The packing of 
particle Vol. 6, J.C. Williams & T. Allen, Elsiever, Amsterdam (1987). 
[56] F. de Larrard, Granular structure and formulation of concretes (in French), Etudes et 
recherches des Laboratoires des ponts et chaussées 34, Laboratoire central des ponts et 
chaussées, Paris (1999), 414 pp. 
[57] P.W. Reimus, T.J. Callahan, S.D. Ware, M.J. Haga, D.A. Counce, Matrix diffusion 
coefficients in volcanic rocks at the Nevada test site: Influence of matrix porosity, 
matrix permeability, and fracture coating minerals. J. Contam. Hydrol. 93 (2007), 85–
95. 
[58] M. Nakamura, K. Otaki, S. Takeuchi, Permeability and pore-connectivity variation of 
pumices from a single pyroclastic flow eruption: Implications for partial fragmentation. 
J. Volcanol. Geoth. Res. 176 (2008), 302–314. 
[59] S. Caré, Aggregate influence on chloride ion diffusion into mortar, Cement Concrete 
Res. 33 (2003) 1021–1028.  
[60] S. Caré, E. Hervé, Application of n-phase model to the diffusion coefficient of chloride 
in mortar, Transp. Porous Media 2 (August 2004) 119–135. 
[61] M.M. Scherer, S. Richter, R.L. Valentine, P.J.J. Alvarez, Chemistry and microbiology of 
permeable reactive barriers for in situ ground water cleanup, Crit. Rev. Environ. Sci. 
Technol. 30 (2000), 363–411. 
 20
508 
509 
510 
511 
512 
513 
514 
515 
516 
[62] M. Shafahi, K. Vafai, Biofilm affected characteristics of porous structures, Int. J. Heat 
Mass Trans. 52, (2009), 574–581. 
[63] A. Nur, G. Mavko, J. Dvorkin, D. Galmudi, Critical porosity; a key to relating physical 
properties to porosity in rocks. The Leading Edge 17 (1998), 357–362. 
[64] M. Lea, Biological sand filters: low-cost bioremediation technique for production of 
clean drinking water, Curr. Prot. Microbiol. (2008) 1G.1.1-1G.1.28. 
[65] WHO, Guidelines for Drinking-water Quality, Recommendations, 3rd ed.; World Health 
Organization: Geneva, 2006; Vol. 1. 
 21
Table 1: Overview on the objectives of using iron/sand mixtures and relationship between 
percent iron mass (wt-%) and percent iron volume (vol-%). The conventional 
expression of the iron amount (wt-%) does not directly accounts for the expansive 
nature of iron corrosion. For example the often used 1:1 (50 wt-% Fe
516 
517 
518 
519 
520 
521 
522 
0) 
corresponds to 25.4 vol-% Fe0. Filter or column clogging due to iron corrosion 
alone will not occur (threshold value 51 vol-% Fe0). 
 
Iron Iron Sand Iron Scale Objective Ref. 
(wt-%) (kg) (kg) (vol-%)    
22 6100 21500 8.8 pilot study sustained permeability [31] 
20 0.60 2.40 7.8    
50 1.50 1.50 25.4 lab columns sustained reactivity [33] 
100 3.00 0.00 100.0    
25 0.02 0.06 10.2    
50 0.04 0.04 25.4    
75 0.06 0.02 50.5 lab columns Fe0 cost reduction [37] 
85 0.07 0.01 65.8    
100 0.08 0.00 100.0    
523 
524 
525 
 
 
 22
Table 2: Density and critical porosity of selected potential additives for improved reactivity 
of conventional Fe
525 
526 
527 
528 
529 
0 filters. All critical porosity’s values for rocks are from Nur et al. 
[63]. The value for activated carbon is an indicative average value from the 
literature. 
 
Material Density Average Density Critical Porosity 
 (g/cm3) (g/cm3) (%) 
Quartz 2.65 - 0 
Sandstone 2.2 - 2.8 2.50 40 
Limestone 2.3 - 2.7 2.50 40 
Dolomites 2.8 - 2.9 2.85 40 
Pumice 0.36-0.91 0.64 80 
Chalks 1.8-2.6 2.20 65 
Rock Salts 2.5 - 2.6 2.55 40 
Oceanic Basalts 2.8 - 3.0 2.90 20 
Activated carbons 0.44-2.50 1.47 55 
530 
531 
 
 23
Table 3: Composition and thickness (Hrz) of the reactive zone for 3 kg of Fe0. Fe0 and sand 
particles are 1.2 mm in diameter. The value C = 0.64 is considered for compactness. 
The residual porosity Φ/Φ
531 
532 
533 
534 
0 = 0 is obtained for 100 % consumed Fe. 
 
[Fe0]0 [Fe0]0 [sand]0 Hrz Φ/Φ0 [Fe0]∞ [Fe0]∞
(vol-%) (kg) (kg) (cm) (-) (%) (kg) 
51 3.00 0.98 1.67 0.0 0.00 0.00 
535 
536 
 
 24
Table 4: Composition of the reactive zone for 51% Fe0 (3 kg of Fe0) and 49 % of additive 
particles (quartz or porous materials). Fe
536 
537 
538 
539 
540 
0 and additives particles are 1.2 mm in 
diameter. The residual porosity Φ/Φ0 = 0 is given for 100 % consumed Fe. The 
value C = 0.64 is considered for compactness. The thickness Hrz of the reactive zone 
is 1.67 cm. The specific weight and the critical porosity of pumice are respectively 
= 640 kg/mPumiceρ 3 and ϕpumice = 0.8 (-).  541 
542  
Pumice Fe0 Quartz Pumice Φ0 Φ/Φ0
(%) (kg) (kg) (kg) (-) (-) 
0 3 0.98 0.00 0.36 0.00 
10 3 0.88 0.02 0.39 0.07 
20 3 0.78 0.05 0.41 0.12 
30 3 0.69 0.07 0.44 0.18 
40 3 0.59 0.09 0.46 0.22 
50 3 0.49 0.12 0.49 0.26 
60 3 0.39 0.14 0.51 0.30 
70 3 0.29 0.17 0.54 0.33 
80 3 0.20 0.19 0.56 0.36 
90 3 0.10 0.21 0.59 0.39 
100 3 0.00 0.24 0.61 0.41 
543 
544 
 
 25
544 
545 
Figure 1 
 
 
 546 
547 
 26
Figure 2 547 
548  
Sand  Sandstone  Quartz 
 
  
 
Figure 2: Schematic diagrams showing the evolution of the porous structure during the 
diagenesis of quartz. As diagenesis progresses. sand grains become compacted and 
cemented. The initial porosity (40 %) decreases down to zero. Modified after Nur 
et al. [63]. 
549 
550 
 
 27
Figure 3  550 
 551 
552 
553 
 
 28
Figure 4 553 
554  
0 20 40 60 80 100
0
20
40
60
80
100
C
A
B
case 2
case 1
 
3 kg Fe0
Quartz: 100 to 0 %
Pumice: 0 to 100 %
Hrz = 1.7 cm
Pumice (%)
 0
 20
 40
 60
 80
 100
Φ/
Φ 0
 / 
[%
]
consumed Fe0 / [%]
 555 
556 
557 
558 
 
 
 29
Figure 5 558 
559  
0 20 40 60 80 100
0
10
20
30
40
50
porosity (φ)
 0 %
 20 %
 40 %
 60 %
 80 %
 90 %
Φ/
Φ 0
 / 
[%
]
porous material / [%]
 560 
561 
562 
563 
 
 
 30
Figure captions 563 
564 
565 
566 
567 
568 
569 
570 
571 
572 
573 
574 
575 
576 
577 
Figure 1: Schematic diagrams showing the extend of porosity loss as influenced by the 
substitution of a part of sand by a porous material: initial stage (top) and final stage (down). 
The final stage corresponds to Fe0 depletion (100 % consumption). At Fe0 depletion, the 
residual porosity is zero for the conventional Fe0/sand filter. The effective residual porosity is 
increased for Fe0/sand/pumice filter (see text). 
Figure 2: Schematic diagrams showing the evolution of the porous structure during the 
diagenesis of quartz. As diagenesis progresses, sand grains become compacted and cemented. 
The initial porosity (40 %) decreases down to zero. Modified after ref. [63]. 
Figure 3: Schematic diagram of an iron-reactive-zone containing Biosand filter. The 
illustration highlights major principles and generic size dimensions. Modified after ref [64]. 
The thickness of a reactive layer containing 3 kg Fe0 representing 51 % (vol.) of the filling is 
1.67 cm (see Tab. 3). 
Figure 4: Evolution of the residual porosity Φ/Φ0 versus the % consumed Fe for 51% Fe0 (3 
kg of Fe0) and 49% of quartz/pumice particles. Fe0 and quartz/pumice are 1.2 mm in diameter. 
The %consumed Fe is given by ).(%
R0
RR0100consumedFe 3
33 −=  with R0 the initial radius of 
Fe
578 
579 
580 
0 and R the residual radius. The value C = 0.64 is considered for compactness. The 
thickness Hrz of the reactive zone is 1.67 cm. The specific weight and the critical porosity of 
pumice are respectively =640 kg/mPumiceρ 3 and ϕpumice =0.8 (-). 581 
582 
583 
584 
Figure 5: Evolution of the residual porosity Φ/Φ0 versus the %replaced quartz particles for 51 
% Fe0 (3 kg of Fe0) and 49 % of quartz/porous particles. Fe0 and additives are 1.2 mm in 
diameter. The residual porosity Φ/Φ0 = 0 is given for 100 % consumed Fe for various porous 
particles with porosityϕpp . The value C = 0.64 is considered for compactness. The thickness 
H
585 
586 rz of the reactive zone is 1.67 cm. 
 31