Accepted Manuscript
Electrodialytic phosphorus recovery from sewage sludge ash under kinetic control Maria Villen-Guzman, Paula Guedes, Nazaré Couto, Lisbeth M. Ottosen, Alexandra B. Ribeiro, Jose M. Rodriguez-Maroto
PII: S0013-4686(18)31799-7 DOI: 10.1016/j.electacta.2018.08.032 Reference: EA 32466
To appear in: Electrochimica Acta Received Date: 1 June 2018 Revised Date: 7 August 2018 Accepted Date: 7 August 2018
Please cite this article as: M. Villen-Guzman, P. Guedes, Nazaré. Couto, L.M. Ottosen, A.B. Ribeiro, J.M. Rodriguez-Maroto, Electrodialytic phosphorus recovery from sewage sludge ash under kinetic control, Electrochimica Acta (2018), doi: 10.1016/j.electacta.2018.08.032.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
M
AN
US
CR
IP
T
AC
CE
PT
ED
Electrodialytic phosphorus recovery from sewage sludge ash under
1
kinetic control
2
Maria Villen-Guzman∗∗∗∗a, Paula Guedesb, Nazaré Coutob, Lisbeth M. Ottosenc, Alexandra
3
B. Ribeirob, Jose M. Rodriguez-Marotoa
4
a
Department of Chemical Engineering, University of Malaga, Campus de Teatinos, 5
29071-Málaga, Spain, Tel: +34 952131915, Fax: +34 952132000, e-mail: 6
b
CENSE, Faculdade de Ciências e Tecnologia, Universidade NOVA de Lisboa, Campus de 8
Caparica, 2829-516 Caparica, Portugal, Tel: +351 212948300, e-mail: [email protected] 9
c
Department of Civil Engineering, Technical University of Denmark, 2800 Lyngby, 10
Denmark, Tel: + 45 45252260, Fax: + 45 45885935, e-mail: [email protected] 11
Abstract
12
A mathematical model for simulating the electrodialytic phosphorus recovery from 13
sewage sludge ash containing heavy metal (Al, Fe, Zn, Cu, Cr, Cd and Ni) is presented. 14
The complex chemical system proposed consists of 46 species including aqueous and 15
solid species. The system setup is modelled as a four compartments: solid, liquid, anode 16
and cathode. In addition to typical phenomena; such as: electromigration of ionic, 17
simple and complex species from the liquid phase to anode and cathode through ionic 18
membranes and diffusion transport; kinetically controlled processes due to non-19
equilibrium between solid phase and bulk liquid have been incorporated. The simulation 20
results clarify the behavior of heavy metal when an electric current is applied which is 21
essential for the scaling-up of the ED technology. 22
M
AN
US
CR
IP
T
AC
CE
PT
ED
Keywords: Electrodialytic; sewage sludge ash; phosphorus recovery; model; kinetic
23
control 24
Nomenclature
25
Aij = cross-sectional area of transference for ith species between any compartment and
26
jth compartment (cm2) 27
AiS = area of transference between liquid (L) and solid (S) compartments (cm2)
28
Api = particle area (cm2)
29
c1A, c2C = concentration of H+ and OH- in anodic and cathodic compartments (mol·cm-3)
30
cij = concentration of ith species in jth volume element (mol·cm-3)
31
Di = diffusion coefficient of ith species (cm2·s-1)
32
(dci/dy) = gradient of concentration (mol·cm-4)
33
(dE/dy) = gradient of electric potential due to diffusive transport (V· cm-1) 34
F = Faraday´s constant, (96 485 C·mol-1) 35
f1λ1 = effective molar conductivity of proton through the anion-exchange membrane
36
(cm2·Ω-1·mol-1) 37
I = current intensity (A) 38
K = equilibrium constant 39
Ki = coefficient of proportionality between AiS and (MiS)2/3(cm4·mol -2/3)
40
MiS = mass of ith species in the solid compartment (S) (mol·cm-3)
41
ni = number of solid particles of ith species in the solid compartment (S)
42
Ni,j = net mass flux of ith species from any compartment into jth compartment (mol·cm
-43
2
·s-1) 44
M
AN
US
CR
IP
T
AC
CE
PT
ED
R = ideal gas constant (J·mol-1·K-1) 45
Ri = reaction rate for ith species in the jth volume element (mol·cm-3·s-1)
46
S(x) = function introduced to take into account the different movements of cations and 47 anions. 48 t = time (s) 49 T = temperature (K) 50
t1,am = effective transport number of proton through the anion-exchange membrane
51
ti,am = effective transport number of i anion through the anion-exchange membrane
52
ti,cm = effective transport number of i cation through the cation-exchange membrane
53
VA = volume of electrolyte in anode compartment (cm3)
54
VC = volume of electrolyte in cathode compartment (cm3)
55
Vj = volume of aqueous phase in the jth volume element (cm3)
56
VL= volume of liquid compartment (cm3)
57
Vpi = particle volume (cm3)
58
zi = charge number of ith species
59
η = Faradic efficiency 60
λi = molar conductivity of the ith species (cm2·Ω-1·mol-1) 61 ρi = particle density (g· cm-3) 62 ψi = particle sphericity 63 1. Introduction 64
M
AN
US
CR
IP
T
AC
CE
PT
ED
The development of new technologies for efficient phosphorus recovery from secondary 65
resources is becoming one of the most important global challenges of the 21st century 66
[1]. This is mainly caused by the decreasing quality of phosphate rock [2], the resource 67
limitations [3] and its volatile prices [4]. 68
Wastewater treatment plants (WWTPs) provide a potential opportunity for P recovery, 69
substituting an important portion of the demand for phosphate rock [5]. The traditional 70
way to valorize the nutrients from WWTPs in agriculture is the direct application of 71
stabilized and dewatered sewage sludge [6]. However, the increasing awareness about 72
contaminants, such as heavy metals, triggers the search for alternatives [7]. 73
Incineration of sewage sludge is one of the studied options since it can reduce the 74
volume and mass generated, energy can be recovered and P is retained and concentrated 75
in the ash (SSA) [8]. As heavy metals in SSA are of great concern due to their potential 76
risk for human health, the development of methods for P recovery is thus needed [9]. 77
Several authors studied the electrodialytic process (ED) for phosphorus recovery from 78
SSA, reporting promising results in the separation of P from the heavy metals [10–14]. 79
This technique, developed at the Technical University of Denmark in 1992 and patented 80
in 1995 (PK95/00209), is based on the combination of the electrokinetic movement of 81
ions in soil with the principle of electrodialysis [15]. ED has been widely studied for 82
different matrices such as treated timber waste [16,17], fly ash [18], mine tailing 83
materials[19], harbour sediments [20,21], municipal wastewater [22,23], and SSA 84
[12,24]. 85
Most published studies about the application of ED are experimentally based studies, 86
reporting results from several contaminants, matrices and operation conditions. These 87
works are essential to know the fundamentals of the technique. However, there is a clear 88
lack of theoretical results obtained by mathematical models. The theoretical research 89
M
AN
US
CR
IP
T
AC
CE
PT
ED
and the model development aims not only to allow the prediction of experimental 90
results, but also a better understanding of the processes involved to improve and 91
optimize the technique [25–28]. Also, it should be highlighted that the models are 92
essential for the implementation of the electrokinetic technology at large operating 93
scales [29–31]. 94
The present work studies the phosphorus recovery and the heavy metal removal of two 95
different sewage sludge ashes under kinetic control, induced by the combination of 96
chemical and external potential gradients. Moreover, a comparison between 97
experimental and model results is shown. 98
This research extends some studies [32,33], adding the following new features: i) the 99
incorporation of the kinetically controlled processes since the solid phase and bulk 100
liquid are not in equilibrium, ii) the development of an ED modelled system consisting 101
of four compartments applied to SSA containing metals (Al, Fe, Zn, Cu, Pb, Cr, Cd and 102
Ni) incorporating information obtained by chemical characterization, iii) the description 103
of a complex chemical system including 33 aqueous species and 13 solid species and iv) 104
the comparison between experimental and modeling results for both ashes. 105
2. Materials and methods
106
2.1. Ash characterization
107
The two SSAs samples used for this study were collected after incineration at 108
Lynettefaellesskatet, WWTP, located in Copenhagen, Denmark. One immediately after 109
the incineration process, fresh ash (SA) and the other, deposited ash, collected from the 110
disposal site (SB). The samples were stored in closed plastic containers at room 111
temperature until the experiments were carried out. 112
M
AN
US
CR
IP
T
AC
CE
PT
ED
The determination of Al, Cd, Cu, Cr, Fe, Ni, P, Pb and Zn was carried out according to 113
the DS259 method [34], using an inductive Coupled Plasma-Optical Emission 114
Spectrometer (ICP-OES), Varian 720-ES. 115
The detailed characterization of the sewage sludge ashes samples (SSAs) is given 116
elsewhere [10]. Table 1 summarizes the main physicochemical characteristics, 117
phosphorus and heavy metal content. All metal concentrations are given on a dry weight 118
basis. 119
Please, insert Table 1 120
The disposal condition has influence on the chemical parameters such as water content, 121
pH and loss on ignition. Regarding metal concentration, no important differences are 122
observed for SA and SB. 123
2.2. Electrodialytic experiments
124
The experimental setup, schematically presented in Fig. 1, consists of an ED laboratory 125
cell divided in three compartments with an internal diameter of 8 cm. All the 126
experiments were performed at constant current of 50 mA that was maintained by 127
means of a power supply unit (Hewlett-Packard E3312A). 128
The central compartment, where the SSA samples were placed, was provided with a 129
stirrer to maintain the matrix mixed. In this compartment, the SSA was suspended in the 130
initial solution, H2SO4 (0.08 M), in the ratio of 1:10 (mass:volume).
131
The central compartment, with a length of 10 cm, was separated from both electrode 132
compartments by means of commercial ion exchange membranes (anion exchange 133
membrane (AEM) AR204 SZRA B02249C and cation exchange membrane (CEM) 134
CR67 HUY N12116B). Platinum coated electrodes were used as anode and cathode. 135
The initial catholyte and anolyte solutions were 500 mL of NaNO3 (0.01 M). The pH at
M
AN
US
CR
IP
T
AC
CE
PT
ED
the electrodes compartments was kept at a constant value of 2 by the manual addition of 137
the necessary amount of 1:1 HNO3 with the aim of increasing species solubility.
138
Please, insert Figure 1. 139
The pH of the electrolytes, pH and conductivity of the central cell, voltage drop and 140
current density were measured during the experiments. The phosphorus and heavy metal 141
content from the samples obtained at the beginning and at the end of each experiment 142
were analysed. 143
At the end of the experiments, once the ashes were separated from the liquid phase, the 144
water content was measured. With the aim of metals recovery, ion exchange membranes 145
and cathode were soaked in HNO3 of different concentration (1 M and 5 M,
146
respectively) for 24 h. SSA, aqueous phases, electrolytes, membranes and cathode after 147
experiments were analysed to quantify phosphorus and heavy metal content. 148
Six ED experiments were performed, three for each kind of SSA (SA and SB), with 149
different duration (3, 7 and 14 days). 150
3. Process simulation and model description
151
The modelled system, schematically presented in Fig. 2, consists of four compartments; 152
the anode (A) and the cathode compartments (C), the solid phase in equilibrium with its 153
associated liquid phase (S) and the completely stirred bulk liquid phase (L) receiving 154
the ions from the solid compartment (S) by diffusion transport. 155
Please, insert Figure 2 156
The initial ash-sulphuric equilibrium process has been simulated using the implemented 157
model and checked by the Visual Minteq code [35]. This allows to establish the initial 158
concentration of soluble species which are ready to be removed from the liquid phase 159
and also the remaining or precipitated solid species to be decontaminated by ED, after 160
M
AN
US
CR
IP
T
AC
CE
PT
ED
their solubilization and transport. Table 2 shows the solids to be treated during the ED 161
experiments. 162
Please, insert Table 2 163
The model operation consists of two steps: i) the simulation of the kinetic processes, 164
including the transport of chemical species between (S) and (L) as well as from (L) 165
toward the electrode compartment integrating in time and also the electrochemical 166
reactions and ii) the re-establishment of the chemical equilibrium in the modelled 167
system before the next time step. The assumption of local equilibrium is widely 168
accepted and could be made due to the chemical equilibria is considered instantaneous 169
comparing with the transport phenomena [36]. Table 3 shows the chemical system 170
which is used to model the case in study. 171
Please, insert Table 3. 172
The mass conservation equation for ith species in each jth compartment, including 173
electrochemical reactions if necessary, is described by: 174 j i ij n j i,j ij j N A RV dt dc V = ∑ + =1 Eq. 1 175 i j ij n j j , i ij R V A N dt dc + ∑ = =1 Eq. 2 176
where Vj is the volume of aqueous phase in jth compartment (cm3), cij is the
177
concentration of ith species in the jth compartment (mol·cm-3), t is the time (s), Ni,j is the
178
net mass flux of ith species from other compartments into jth compartment (mol·cm-2·s -179
1
), Aij is the cross-sectional area of transference for ith species between any
180
compartment and jth compartment (cm2), and Ri is the reaction rate for ith species in the
181
jth compartment (mol·cm-3·s-1). 182
M
AN
US
CR
IP
T
AC
CE
PT
ED
3.1 Diffusive transport 183The diffusive transport of ions and non-ionic soluble compounds is the only transport 184
phenomenon considered between (S) and (L). The calculation of the flux of the ionic ith 185
species (Ni) has been based in the Nerst-Planck equation since the unidirectional flux of
186
each ion results from a combination of electrical and concentration gradients [45]: 187
( )
− − λ = D i i i 2 i i i dy dE F c z dy dc RT F z N Eq 3 188where λi, zi, F, R, T, (dci/dy) and (dE/dy) are the molar conductivity of the ith species
189
(cm2·Ω-1·mol-1), the charge number of ith species, the Faraday’s constant (96 485 190
C·mol-1), the gas constant (J·mol-1·K-1), the temperature (K) and the gradients of 191
concentration (mol·cm-4) and electric potential (V·cm-1) due to diffusive transport, 192
respectively. 193
It should be highlighted that the electric gradient is not imposed externally, but it is due 194
to the small separation of charges which result from diffusion itself. This equation, 195
applied between solid (S) and liquid compartments (L) for each of the present ions, is 196
combined to satisfy the requirement of zero current through both compartments 197 ( 0 1 =
∑
= n i i iN z ) 198The flux of the non-ionic species from the solid compartment (S) to bulk of liquid (L) is 199
calculated from the Fick law: 200 − = dy dc D N i i i Eq 4 201
where Di is the diffusion coefficient of ith species (cm2·s-1).
202
The ratio of surface to volume (AiS/Vj) used for mass transference between (S) and (L)
203
should be estimated. The total volume of liquid in the central cell is known (500 cm3) 204
M
AN
US
CR
IP
T
AC
CE
PT
ED
and the surface of transference between (S) and (L) for each ith species could be 205
approximatively related to mass as: 206
( )
(
)
23 3 2 3 2 3 1 3 2 3 1 6 6 iS i i i pi i i i pi i pi i iS K M V n V n A n A = = = ≈ρ
ψ
ρ
π
ψ
π
Eq. 5 207where ni is the number of solid particles of ith species in the solid compartment (S), Api
208
(cm2) is the area of one particle, Vpi its volume (cm3), ψi its sphericity and ρi its density
209
(g·cm-3). Although Ki is different for each ith species, on the sake of simplicity only one
210
value is considered for all the included species at the beginning of the calculations. 211
Since the liquid phase (L) is experimentally well stirred, it could be assumed to be 212
perfectly mixed. In the electrode compartments, this assumption could be also made due 213
to the recirculation of the electrolyte and the addition of acid for the pH control of the 214
system. 215
3.2 Electrochemical reactions
216
It is assumed that the only significant reaction taking place on the electrodes are the 217
reduction and oxidation of water. Some works use different theories to estimate the 218
activity coefficients [36,46]. In this case, the activity coefficients for all species are set 219
to a fixed value of 1. 220
The half-reaction at the cathode is 221
2H2O+2e− →2OH−
( )
aq +H2( )
gas E0 = −0.828V Eq. 6 222And the anode half-reaction is: 223
( )
( )
4 1.229 V 4 2 0 2O→ H+ aq +O gas + e− E = − H Eq. 7 224The electrochemical reactions (ER) are included in the mass balance equations of anode 225
and cathode compartments: 226
η
F I dt dc V dt dc V ER C C ER A A = = 1 2 Eq. 8 227M
AN
US
CR
IP
T
AC
CE
PT
ED
where VA and VC are the volumes of electrolyte in the electrode compartments
228
(≈ 500 cm3), c1A and c2C, are the concentration of H+ and OH- generated by
229
electrochemical reactions (mol·cm-3), I is the current intensity (A), F, the Faraday´s 230
constant (96 485 C·mol-1) and η the Faradic efficiency, assumed 100%. All the 231
hydroxide ions generated by electrolysis are subsequently substituted by nitrate ions due 232
to the addition of nitric acid to control pH value. 233
3.3 Electromigration transport of ions
234
Next, the electromigration (EM) of ions from the liquid compartment (L) into the 235
electrode compartments (A) and (C) is analyzed. For this, it is assumed that the cations 236
are moving out from the cell to the cathode passing through the cation exchange 237
membrane (CEM), and, simultaneously, anions are entering to the anode through the 238
anion exchange membrane (AEM). Assuming perfect mixing in the liquid phase (L) and 239
using the transport number concept, the effective transport numbers of proton, t1,am, and
240
i anion, ti,am , through the anion-exchange membrane are respectively [47,48,25,32]:
241
( )
[
]
( )
+ ∑[
( )
−( )
]
= = n i i iL i i,L A , A A , A am , c S z S c c S c f c S c f t 1 1 1 1 1 1 1 1 1 1λ
λ
λ
Eq. 9 242( )
( )
( )
[
( )
( )
]
1 1 1 1 1 1 > ∑ − + − = = i ; c S z S c c S c f c S z S c t n i i iL i i,L A , A L , i i iL i am , iλ
λ
λ
Eq. 10 243where
(
f1λ
1)
is the effective molar conductivity of proton through the anion-exchange 244membrane (cm2·Ω-1·mol-1) [49], λi is molar conductivity of ith species (ohm-1·mol
-245
1
·cm2) and zi its charge number. S(x) is a function key, S(x) = 0, if x ≤ 0, and S(x) = 1, if
246
x > 0. 247
Likewise, the transport number of i cation through the cation-exchange membrane, ti,cm,
248
is given by: 249
M
AN
US
CR
IP
T
AC
CE
PT
ED
( )
( )
[
]
( )
( )
[
]
∑ = = n i i iL i i,L L , i i iL i cm , i c S z S c c S z S c t 1λ
λ
Eq. 11 250Thus, it is possible to calculate the EM movement of each ion (Table 3) through both 251
membranes outside of the liquid compartment (L) and, therefore, the EM mass balance 252
for ith ion in VL is given by:
253
( )
(
)
F z I t t A N dt dc V i cm , i am , i iL EM i EM iL L =∑ = − Eq. 12 254 3.4. Chemical equilibria 255The concentration values obtained taking into account the transport phenomena depend 256
directly on the chemical equilibrium. It is assumed that the reversible equilibrium 257
involved is fast enough toward the products (direct) and toward the reactants (reverse). 258
These reversible equilibria play an important role on the pH values and on the 259
contaminants mobility. From the values obtained considering the transport phenomena, 260
the model recalculates new concentration values in each time step of the numerical 261
integration until chemical equilibrium is reached. The chemical system including thirty-262
three aqueous species and thirteen solid species is described in Table 3. 263
After each time step, the model verifies the presence of solid phase with the aim of 264
establishing the local concentration value for the different species present in the aqueous 265
phase. This procedure is based on an experimental fact [10]: the insoluble heavy metals 266
are present in the solid phase even after long-term electrodialytic treatments despite the 267
high percentage of ash dissolved. 268
3.5. Complete model
269
The differential equations describing the transport phenomena of ionic and non-ionic 270
chemical species together with the equilibrium equations constitute the complete model. 271
M
AN
US
CR
IP
T
AC
CE
PT
ED
The amount of phosphorus recovered and heavy metals recovered from the liquid phase 272
in the electrode compartments are given by: 273 = dt dc V dt dM iA A A Eq. 13 274 = dt dc V dt dM iC C C Eq. 14 275
The use of transport number concept [48,50] together with the zero current condition 276
and the initial neutrality of the system guarantees the electro-neutrality condition which 277
is continuously checked by the implemented model. 278
4. Results
279
In this section, the simulation results based on the modelled system described above are 280
presented and compared to the experimental results, previously published [10]. 281
Immediately before each ED experiment, the ash sample was suspended in sulphuric 282
acid with the aim of increasing species solubility. The temporal evolution of pH for the 283
liquid phase in equilibrium with solid phase (S) and for the stirred bulk liquid phase (L) 284
during the electrodialytic experiments for both SSA samples (SA and SB) is presented 285
in Fig 3. The experimental values of pH at the end of the electrodialytic experiments 286
carried out for different periods of time (3, 7 and 14 days) are also shown. The water 287
splitting at the anion exchange membrane [51] together with the transport of H+ through 288
the anion-exchange, which is not 100% ideal [49] entails a pH value decreasing during 289
ED experiments. The comparison of experimental and simulation results indicates that 290
the model predicts fairly well the pH evolution. 291
Please, insert Figure 3 292
The percentage values of elements in the different parts of the ED system (anode (A), 293
cathode (C), solid compartment (S) and the stirred bulk liquid phase (L)) against the 294
M
AN
US
CR
IP
T
AC
CE
PT
ED
treatment time obtained in the simulations for both SSA have been studied (Fig 4, 5, 6, 295
7 and 8). Phosphorus amount in the solid compartment (ash) decreased with time 296
remediation (Fig 4). P was transported mainly towards anode, with almost 40% and 297
30% collected at the cathode after 14 days. These results indicate that phosphorus could 298
have formed positively and negatively charged compounds. According to the Pourbaix 299
diagram, the predominant P-species for a pH value between 3.5 and 1.3 are H2PO4- and
300
H3PO4. It should be highlighted the inflection point found at 685 000 s and 575 000 s
301
for SA and SB, respectively, which has been related with the total dissolution of 302
hydroxyapatite. This fact is consistent with the changes observed in the pH evolution, 303
presented in Fig 3. 304
Please, insert Figure 4 305
The total dissolution of hydroxyapatite has a direct influence on the behavior of other 306
metals such as Cu, Zn and Al whose evolution is presented in Fig 5. As can be 307
observed, the lower the pH is, the higher recovery of metals in the cathode end is 308
achieved. These results are connected with the increasing species solubility with 309
decreasing pH. After 14 days, the better removal efficiency was achieved for Zn, with 310
almost 100 % being transported towards the cathode end. A good removal rate has been 311
also obtained for the Cu, with 90 % recovered in the cathode compartment while 30 % 312
and 55 % of Al was transported towards the cathode end for SA and SB samples, 313
respectively. 314
Please, insert Figure 5 315
For Cd (Fig. 6), it has been registered higher removals with approximately 70 % and 316
60 % recovery in the cathode compartment after 14 days for SA and SB, respectively. 317
Results for Ni present a similar behavior, being registered, approximately, 50 % and 318
M
AN
US
CR
IP
T
AC
CE
PT
ED
40 % of Ni in the cathode end for SA and SB, respectively. As can be observed, no 319
metal is present in the liquid phase of the central compartment. These results show that 320
the fraction of the metal solubilized is mobilized quickly towards the cathode end. 321
Please, insert Figure 6. 322
The results for Pb (Fig. 7) are directly associated with the low solubility of solid 323
species. It has been observed that after the dissolution of lead phosphate, the 324
precipitation of lead sulphate takes place. This is a direct consequence of using 325
sulphuric acid as assisting agent. Further studies using other agents, such as organic 326
acids, could be useful to improve the technique. 327
Please, insert Figure 7 328
For Cr and Fe (Fig. 8), the dissolution of solid species was no observed. So, no metals 329
were collected in the electrode compartments according to the extremely low solubility 330
of these two phosphates. 331
Please, insert Figure 8 332
The simulation and experimental results [10] for the three periods of time essayed 3, 7 333
and 14 days, are compared in Fig. 9. The x-axis represents the percentage of the metals 334
obtained experimentally in the different cell sections. Likewise, the y-axis represents the 335
results obtained with the proposed model. As can be observed, the model predicts 336
accurately the experimental results. The detected differences could be associated not 337
only with modelling assumptions but also with the known inhomogeneous distribution 338
of metals in the solid samples. 339
Please, insert Figure 9 340
5. Conclusions
M
AN
US
CR
IP
T
AC
CE
PT
ED
The implemented model for simulating the electrodialytic phosphorus recovery and 342
simultaneous heavy metal removal from sewage sludge ash reproduces satisfactorily the 343
experimental results. The experimental pH values and the evolution of metals presented 344
in the fly ash, a complex chemical system, have been fairly predicted even in the 345
presence of assisting agents. The model, based on different transport phenomena, such 346
as diffusion and electromigration, as well as electrochemical reactions, allows the 347
increasing of the overall understanding of the different processes involved. The 348
kinetically controlled processes have been demonstrated to be relevant when ED is 349
applied to SSA. The comparison of simulation and experimental results also clarifies the 350
influence of hydroxyapatite dissolution processes on the pH value and on heavy metal 351
recovery. Finally, the model application makes possible to optimize and improve some 352
key aspects of the experimental methods, such as the choice of an enhancing agent, 353
stirring rate or the current density applied. 354
Acknowledgements
355
The authors thank e.THROUGH (H2020-MSCA-RISE-2017-778045), financed by the 356
European Commission, as well as the financial support provided by CENSE-Center for 357
Environmental and Sustainability Research which is financed by national funds from 358
FCT/MEC (UID/AMB/04085/2013). Villen-Guzman also acknowledges the “Contrato 359
Puente (CI-17-010)” obtained from University of Malaga. 360
References
361
[1] P.M. Melia, A.B. Cundy, S.P. Sohi, P.S. Hooda, R. Busquets, Trends in the 362
recovery of phosphorus in bioavailable forms from wastewater, Chemosphere. 186 363
(2017) 381–395. doi:10.1016/j.chemosphere.2017.07.089. 364
M
AN
US
CR
IP
T
AC
CE
PT
ED
[2] S.S. Mar, M. Okazaki, Investigation of Cd contents in several phosphate rocks 365
used for the production of fertilizer, Microchem. J. 104 (2012) 17–21. 366
doi:10.1016/j.microc.2012.03.020. 367
[3] D. Cordell, A. Rosemarin, J.J. Schröder, A.L. Smit, Towards global phosphorus 368
security: A systems framework for phosphorus recovery and reuse options, 369
Chemosphere. 84 (2011) 747–758. doi:10.1016/j.chemosphere.2011.02.032. 370
[4] M.C. Mew, Phosphate rock costs, prices and resources interaction, Spec. Issue 371
Sustain. Phosphorus Tak. Stock Phosphorus Supply Nat. Anthropog. Pools 21st 372
Century. 542 (2016) 1008–1012. doi:10.1016/j.scitotenv.2015.08.045. 373
[5] L. Egle, H. Rechberger, J. Krampe, M. Zessner, Phosphorus recovery from 374
municipal wastewater: An integrated comparative technological, environmental 375
and economic assessment of P recovery technologies, Sci. Total Environ. 571 376
(2016) 522–542. doi:10.1016/j.scitotenv.2016.07.019. 377
[6] O.F. Schoumans, F. Bouraoui, C. Kabbe, O. Oenema, K.C. van Dijk, Phosphorus 378
management in Europe in a changing world, Ambio. 44 (2015) 180–192. 379
doi:10.1007/s13280-014-0613-9. 380
[7] A. Amann, O. Zoboli, J. Krampe, H. Rechberger, M. Zessner, L. Egle, 381
Environmental impacts of phosphorus recovery from municipal wastewater, 382
Resour. Conserv. Recycl. 130 (2018) 127–139. 383
doi:10.1016/j.resconrec.2017.11.002. 384
[8] S. Donatello, C.R. Cheeseman, Recycling and recovery routes for incinerated 385
sewage sludge ash (ISSA): A review, Waste Manag. 33 (2013) 2328–2340. 386
doi:10.1016/j.wasman.2013.05.024. 387
[9] H. Herzel, O. Krüger, L. Hermann, C. Adam, Sewage sludge ash — A promising 388
secondary phosphorus source for fertilizer production, Spec. Issue Sustain. 389
M
AN
US
CR
IP
T
AC
CE
PT
ED
Phosphorus Tak. Stock Phosphorus Supply Nat. Anthropog. Pools 21st Century. 390
542 (2016) 1136–1143. doi:10.1016/j.scitotenv.2015.08.059. 391
[10] P. Guedes, N. Couto, L.M. Ottosen, A.B. Ribeiro, Phosphorus recovery from 392
sewage sludge ash through an electrodialytic process, Waste Manag. 34 (2014) 393
886–892. doi:10.1016/j.wasman.2014.02.021. 394
[11] B. Ebbers, L.M. Ottosen, P.E. Jensen, Comparison of two different electrodialytic 395
cells for separation of phosphorus and heavy metals from sewage sludge ash, 396
Chemosphere. 125 (2015) 122–129. doi:10.1016/j.chemosphere.2014.12.013. 397
[12] L.M. Ottosen, P.E. Jensen, G.M. Kirkelund, Phosphorous recovery from sewage 398
sludge ash suspended in water in a two-compartment electrodialytic cell, Waste 399
Manag. 51 (2016) 142–148. doi:10.1016/j.wasman.2016.02.015. 400
[13] R. Parés Viader, P.E. Jensen, L.M. Ottosen, J. Ahrenfeldt, H. Hauggaard-Nielsen, 401
Sequential electrodialytic recovery of phosphorus from low-temperature 402
gasification ashes of chemically precipitated sewage sludge, Waste Manag. 60 403
(2017) 211–218. doi:10.1016/j.wasman.2016.11.030. 404
[14] P. Guedes, N. Couto, E.P. Mateus, A.B. Ribeiro, Phosphorus Recovery in Sewage 405
Sludge by Electrokinetic Based Technologies: A Multivariate and Circular 406
Economy View, Waste Biomass Valorization. (2017). doi:10.1007/s12649-017-407
9985-y. 408
[15] L.M. Ottosen, H.K. Hansen, S. Laursen, A. Villumsen, Electrodialytic remediation 409
of soil polluted with copper from wood preservation industry, Environ. Sci. 410
Technol. 31 (1997) 1711–1715. doi:10.1021/es9605883. 411
[16] A.B. Ribeiro, E.P. Mateus, L.M. Ottosen, G. Bech-Nielsen, Electrodialytic 412
removal of Cu, Cr, and As from chromated copper arsenate-treated timber waste, 413
Environ. Sci. Technol. 34 (2000) 784–788. doi:10.1021/es990442e. 414
M
AN
US
CR
IP
T
AC
CE
PT
ED
[17] J. Virkutyte, E. Velizarova, A.B. Ribeiro, M. Sillanpää, Copper and chromium 415
electrodialytic migration in CCA-treated timber waste, Water. Air. Soil Pollut. 160 416
(2005) 27–39. doi:10.1007/s11270-005-3427-3. 417
[18] W. Chen, P.E. Jensen, L.M. Ottosen, G.M. Kirkelund, Electrodialytic remediation 418
of fly ash from co-combustion of wood and straw, Electrochimica Acta. 181 419
(2015) 208–216. doi:10.1016/j.electacta.2015.04.083. 420
[19] H.K. Hansen, A.B. Ribeiro, E.P. Mateus, L.M. Ottosen, Diagnostic analysis of 421
electrodialysis in mine tailing materials, Electrochimica Acta. 52 (2007) 3406– 422
3411. doi:10.1016/j.electacta.2006.05.066. 423
[20] G.M. Nystrøm, L.M. Ottosen, A. Villumsen, Test of experimental set-ups for 424
electrodialytic removal of Cu, Zn, Pb and Cd from different contaminated harbour 425
sediments, Eng. Geol. 77 (2005) 349–357. doi:10.1016/j.enggeo.2004.07.025. 426
[21] K.B. Pedersen, P.E. Jensen, L.M. Ottosen, T. Lejon, An optimised method for 427
electrodialytic removal of heavy metals from harbour sediments, Electrochimica 428
Acta. 173 (2015) 432–439. doi:10.1016/j.electacta.2015.05.050. 429
[22] B. Ebbers, L.M. Ottosen, P.E. Jensen, Electrodialytic treatment of municipal 430
wastewater and sludge for the removal of heavy metals and recovery of 431
phosphorus, Electrochimica Acta. 181 (2015) 90–99. 432
doi:10.1016/j.electacta.2015.04.097. 433
[23] P.E. Jensen, G.M. Kirkelund, K.B. Pedersen, C. Dias-Ferreira, L.M. Ottosen, 434
Electrodialytic upgrading of three different municipal solid waste incineration 435
residue types with focus on Cr, Pb, Zn, Mn, Mo, Sb, Se, V, Cl and SO4, 436
Electrochimica Acta. 181 (2015) 167–178. doi:10.1016/j.electacta.2015.06.012. 437
[24] R. Parés Viader, P.E. Jensen, L.M. Ottosen, J. Ahrenfeldt, H. Hauggaard-Nielsen, 438
Electrodialytic extraction of phosphorus from ash of low-temperature gasification 439
M
AN
US
CR
IP
T
AC
CE
PT
ED
of sewage sludge, Electrochimica Acta. 181 (2015) 100–108. 440
doi:10.1016/j.electacta.2015.05.025. 441
[25] A.B. Ribeiro, J.M. Rodriíguez-Maroto, E.P. Mateus, H. Gomes, Removal of 442
organic contaminants from soils by an electrokinetic process: The case of atrazine. 443
Experimental and modeling, Chemosphere. 59 (2005) 1229–1239. 444
doi:10.1016/j.chemosphere.2004.11.054. 445
[26] A.T. Lima, A.B. Ribeiro, J.M. Rodríguez-Maroto, E.P. Mateus, A.M. Castro, L.M. 446
Ottosen, Experimental and modeling of the electrodialytic and dialytic treatment of 447
a fly ash containing Cd, Cu and Pb, J. Appl. Electrochem. 40 (2010) 1689–1697. 448
doi:10.1007/s10800-010-0166-y. 449
[27] A.B. Ribeiro, E.P. Mateus, J.-M. Rodríguez-Maroto, Removal of organic 450
contaminants from soils by an electrokinetic process: The case of molinate and 451
bentazone. Experimental and modeling, Sep. Purif. Technol. 79 (2011) 193–203. 452
doi:10.1016/j.seppur.2011.01.045. 453
[28] J.M. Paz-García, B. Johannesson, L.M. Ottosen, A.B. Ribeiro, J.M. Rodríguez-454
Maroto, Modeling of electrokinetic processes by finite element integration of the 455
Nernst-Planck-Poisson system of equations, Sep. Purif. Technol. 79 (2011) 183– 456
192. doi:10.1016/j.seppur.2011.02.023. 457
[29] M. Villen-Guzman, J.M. Paz-Garcia, J.M. Rodriguez-Maroto, F. Garcia-Herruzo, 458
G. Amaya-Santos, C. Gomez-Lahoz, C. Vereda-Alonso, Scaling-up the acid-459
enhanced electrokinetic remediation of a real contaminated soil, Electrochimica 460
Acta. 181 (2015) 139–145. doi:10.1016/j.electacta.2015.02.067. 461
[30] M. Villen-Guzman, C. Gomez-Lahoz, F. Garcia-Herruzo, C. Vereda-Alonso, J.M. 462
Paz-Garcia, J.M. Rodriguez-Maroto, Specific Energy Requirements in 463
Electrokinetic Remediation, (2017). doi:10.1007/s11242-017-0965-2. 464
M
AN
US
CR
IP
T
AC
CE
PT
ED
[31] R. López-Vizcaíno, C. Risco, J. Isidro, S. Rodrigo, C. Saez, P. Cañizares, V. 465
Navarro, M.A. Rodrigo, Scale-up of the electrokinetic fence technology for the 466
removal of pesticides. Part I: Some notes about the transport of inorganic species, 467
Chemosphere. 166 (2017) 540–548. doi:10.1016/j.chemosphere.2016.09.113. 468
[32] A.B. Ribeiro, J.M. Rodríguez-Maroto, E.P. Mateus, E. Velizarova, L.M. Ottosen, 469
Modeling of electrodialytic and dialytic removal of Cr, Cu and As from CCA-470
treated wood chips, Chemosphere. 66 (2007) 1716–1726. 471
doi:10.1016/j.chemosphere.2006.07.014. 472
[33] E. Velizarova, A.B. Ribeiro, E. Mateus, L.M. Ottosen, Effect of different 473
extracting solutions on the electrodialytic remediation of CCA-treated wood waste 474
Part I.: Behaviour of Cu and Cr, J. Hazard. Mater. 107 (2004) 103–113. 475
doi:10.1016/j.jhazmat.2003.09.011. 476
[34] DS259, Determination of metals in water, sludge and sediments - General 477
guidelines for determination by atomic absorption spectrophotometry in flame. 478
Dansk Standardiseringsrad (in Danish)., 2003. 479
[35] Gustafsson, J., Visual MINTEQ ver. 3.0 beta., KTH Royal Institute of 480
Technology, Dept. of Land and Water Resources Engineering, Stockholm, 481
Sweden, 2010. http://www2.lwr.kht.se/English/OurSoftware/vminteq/index.html. 482
[36] J.M. Paz-García, M. Villén-Guzmán, A. García-Rubio, S. Hall, M. Ristinmaa, C. 483
Gómez-Lahoz, A coupled reactive-transport model for electrokinetic remediation, 484
in: Electrokinet. Discip. Cont. New Strateg. Sustain. Dev., Springer International 485
Publishing, 2015: pp. 251–278. doi:10.1007/978-3-319-20179-5_13. 486
[37] J.A. Dean, N.A. Lange, Lange’s Handbook of Chemistry, McGraw-Hill, 1999. 487
https://books.google.es/books?id=56KPMQEACAAJ. 488
[38] A.E. Martell, R.M. Smith, R.J. Motekaitis, NIST Stand. Ref. Database 467. (2004). 489
M
AN
US
CR
IP
T
AC
CE
PT
ED
[39] J.D. Allison, D.S. Brown, K.L. Novo-Gradac, MINTEQA2/PRODEFA2, a 490
geochemical assessment model for environmental systems: Version 3.0 user’s 491
manual. United States Environmental Protection Agency, Office of Research and 492
Development, Washington, DC, EPA6003-91021 MTQ311. (1991). 493
[40] D.L. Parkhurst, C. Appelo, User’s guide to PHREEQC (Version 2): A computer 494
program for speciation, batch-reaction, one-dimensional transport, and inverse 495
geochemical calculations, (1999) 312. 496
[41] C. Xiao, D.J. Wesolowski, D.A. Palmer, Formation quotients of aluminum sulfate 497
complexes in NaCF3SO3 media at 10, 25, and 50 C from potentiometric titrations 498
using a mercury/mercurous sulfate electrode concentration cell, Environ. Sci. 499
Technol. 36 (2002) 166–173. 500
[42] D.K. Nordstrom, L.N. Plummer, D. Langmuir, E. Busenberg, H.M. May, B.F. 501
Jones, D.L. Parkhurst, Revised chemical equilibrium data for major water-mineral 502
reactions and their limitations, in: Chem. Model. Aqueous Syst. II DC Melchior 503
RL Bassett Eds Am. Chem. Soc. Symp. Ser. 416 31 398-413, 1990. 504
[43] W. Haynes, CRC Handbook of Chemistry and Physics, 91th Ed. (2010). 505
[44] H.M. Selim, Phosphate in Soils: Interaction with Micronutrients, Radionuclides 506
and Heavy Metals, CRC Press, 2017. 507
[45] Sherwood, T.K., Pigford, R.L., Wilke, C.L., Mass Transfer, McGraw-Hill, New 508
York, 1975. 509
[46] R. López-Vizcaíno, A. Yustres, M.J. León, C. Saez, P. Cañizares, M.A. Rodrigo, 510
V. Navarro, Multiphysics Implementation of Electrokinetic Remediation Models 511
for Natural Soils and Porewaters, Electrochimica Acta. 225 (2017) 93–104. 512
doi:10.1016/j.electacta.2016.12.102. 513
M
AN
US
CR
IP
T
AC
CE
PT
ED
[47] D.J. Wilson, J.M. Rodríguez-Maroto, C. Gómez-Lahoz, Electrokinetic 514
Remediation. I. Modeling of Simple Systems, Sep. Sci. Technol. 30 (1995) 2937– 515
2961. doi:10.1080/01496399508013121. 516
[48] D.J. Wilson, J.M. Rodriguez-Maroto, C. Gomez-Lahoz, Electrokinetic 517
remediation. II. Amphoteric metals and enhancement with a weak acid, Sep. Sci. 518
Technol. 30 (1995) 3111–3128. 519
[49] R. De Lara, J. Rodriguez-Postigo, F. Garcia-Herruzo, J. Rodriguez-Maroto, J. 520
Benavente, Decontamination of soils by membrane processes: characterization of 521
membranes under working conditions, Ind. Eng. Chem. Res. 44 (2005) 400–407. 522
[50] C. Vereda-Alonso, J.M. Rodríguez-Maroto, R.A. García-Delgado, C. Gómez-523
Lahoz, F. García-Herruzo, Two-dimensional model for soil electrokinetic 524
remediation of heavy metals: Application to a copper spiked kaolin, Chemosphere. 525
54 (2004) 895–903. doi:10.1016/j.chemosphere.2003.09.002. 526
[51] L.M. Ottosen, H.K. Hansen, C.B. Hansen, Water splitting at ion-exchange 527
membranes and potential differences in soil during electrodialytic soil remediation, 528
J. Appl. Electrochem. 30 (2000) 1199–1207. doi:10.1023/A:1026557830268. 529
M
AN
US
CR
IP
T
AC
CE
PT
ED
Reaction Log K Source
Water ionization + ↔ 13.99 [37] Acid ionization + ↔ 12.375 [38] 2 + ↔ 19.573 [38] 3 + ↔ 21.721 [38] + ↔ 1.99 [38] Complex formation + ↔ 20.01 [38] 2 + ↔ 18.98 [38] + ↔ 20.923 [38] + ↔ () 15.035 [38] + ↔ () 2.36 [38] + ↔ () 16.5 [38] + ↔ () 2.36 [38] + ↔ 22.285 [38] + ↔ () 15.475 [38] + ↔ () 2.69 [38] + ↔ () 15.69 [38] + ↔ () 2.34 [38] + 2 ↔ ( ) 3.28 [39] + 2 ↔ ( ) 3.50 [40] + 2 ↔ ( ) 5.58 [41] + 2 ↔ ( ) 5.38 [42] Precipitation-dissolution 2 + 3 ↔ () + 6 1.418 [38] + ↔ () 20.01 [43] 5 + 3 − + ↔ "( ) 44.333 [38] 3 + 2 ↔ ( ) 32.60 [39] #+ ↔ # () 2.55 [44] 3 + 2 ↔ ( ) 36.85 [39] + ↔ () 16.84 [38] 3 $%+ 2 ↔ $%( ) 36.85 [39] 3 + 2 ↔ ( ) 43.55 [38]
M
AN
US
CR
IP
T
AC
CE
PT
ED
3 + 2 ↔ ( ) 35.42 [38] + ↔ () 4.36 [42] + ↔ () 7.79 [38]M
AN
US
CR
IP
T
AC
CE
PT
ED
Parameter SA SBPhysical and chemical characteristics
pH (H2O) 12.44 ± 0.01 8.85 ± 0.03 Water content (%) 0.10 ± 0.18 16 ± 0.38 Conductivity (mS cm-1) 3.23 ± 0.51 4.81 ± 0.13 Loss on ignition (550 ºC; %) 0.15 ± 0.05 0.92 ± 0.08 Solubility in water (%) 1.8 ± 0.1 3.1 ± 0.0 Carbonate content (%) 1.57 ± 0.21 1.34 ± 0.12 Elements concentration P (g kg-1) 134 ± 1 129 ± 5 Na (g kg-1) 5.71 ± 0.03 6.73 ± 0.09 Ca (g kg-1) 163 ± 1 152 ± 2 Al (g kg-1) 22.6 ± 0.5 21.5 ± 0.7 Fe (g kg-1) 60.0 ± 1.4 62.0 ± 2.3 Zn (mg kg-1) 3335 ± 77 3157 ± 129 Cu (mg kg-1) 758 ± 5 733 ± 9 Pb (mg kg-1) 293 ± 4 297 ± 9 Cr (mg kg-1) 45.5 ± 0.4 44.9 ± 0.7 Cd (mg kg-1) 3.25 ± 0.04 3.14 ± 0.08 Ni (mg kg-1) 54.6 ± 0.6 55.7 ± 1.0
Table 1. Physico-chemical characteristics, phosphorus and heavy metal content in the
M
AN
US
CR
IP
T
AC
CE
PT
ED
SA SB Ca5(OH)(PO4)3 5.97 10-2 5.39 10-2 FePO4 10.74 10-2 11.10 10-2 AlPO4 8.37 10-2 7.96 10-2 CaSO4 8.03 10-2 9.02 10-2 Zn3(PO4)2 1.66 10-3 1.57 10-3 Cu3(PO4)2 3.77 10-4 3.65 10-4 Pb3(PO4)2 4.66 10-5 4.66 10-5 CrPO4 9.00 10-5 9.00 10-5M
AN
US
CR
IP
T
AC
CE
PT
ED
M
AN
US
CR
IP
T
AC
CE
PT
ED
M
AN
US
CR
IP
T
AC
CE
PT
ED
Fig. 3. Simulation results for the pH of the liquid phase in equilibrium with solid phase
(S) (·· ··) and for the stirred bulk liquid phase (L) (─ ·) during the electrodialytic experiments as well as the experimental results (●). a) SSA, b) SSB.
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 p H t/s·10-5
a)
b)
M
AN
US
CR
IP
T
AC
CE
PT
ED
Fig. 4. Simulation results for the evolution of P percentage in the electrodialytic cell
sections (solid (─), liquid phase (····), cathode (─ ·) and anode(─ ─)). a) SSA, b) SSB. 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 % P t/s·10-5 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 % P t/s·10-5
a)
b)
M
AN
US
CR
IP
T
AC
CE
PT
ED
0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 % C u t/s·10-5 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 % C u t/s·10-5 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 % Z n t/s·10-5 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 % Z n t/s·10-5 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 % A l t/s·10-5 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 % A l t/s·10-5a)
b)
M
AN
US
CR
IP
T
AC
CE
PT
ED
Fig. 5. Simulation results for the evolution of Cu, Zn and Al percentages in the
electrodialytic cell sections (solid (─), liquid phase (····), cathode (─ ·) and anode(─ ─)). a) SSA, b) SSB.
M
AN
US
CR
IP
T
AC
CE
PT
ED
Fig. 6. Simulation results for the evolution of Cd and Ni percentages in the
electrodialytic cell sections (solid (─), liquid phase (····), cathode (─ ·) and anode (─ ─)). a) SSA, b) SSB. 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 % C d t/s·10-5 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 % C d t/s·10-5 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 % N i t/s·10-5 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 % N i t/s·10-5
a)
b)
M
AN
US
CR
IP
T
AC
CE
PT
ED
Fig. 7. Simulation results for the evolution of Pb percentage in the electrodialytic cell
sections (solid (─), solid phase associated with lead phosphate (─ ··), liquid phase (····), cathode (─ ·) and anode (─ ─). a) SSA, b) SSB.
0 20 40 60 80 100 120 0 2 4 6 8 10 12 14 % P b t/s·10-5 0 20 40 60 80 100 120 0 2 4 6 8 10 12 14 % P b t/s·10-5
a)
b)
M
AN
US
CR
IP
T
AC
CE
PT
ED
Fig. 8. Simulation results for the evolution of Cr and Fe percentages in the
electrodialytic cell sections (solid (─), liquid phase (····), cathode (─ ·) and anode (─ ─)). a) SSA, b) SSB. 0 20 40 60 80 100 120 0 2 4 6 8 10 12 14 % C r t/s·10-5 0 20 40 60 80 100 120 0 2 4 6 8 10 12 14 % C r t/s·10-5 0 20 40 60 80 100 120 0 2 4 6 8 10 12 14 % F e t/s·10-5 0 20 40 60 80 100 120 0 2 4 6 8 10 12 14 % F e t/s·10-5
a)
b)
M
AN
US
CR
IP
T
AC
CE
PT
ED
Fig. 9. Experimental and model results for the metals analysed for different periods of
time: 3 (■), 7 (●) and 14 days (▲). a) SSA, b) SSB. 0 20 40 60 80 100 0 20 40 60 80 100 M ode l Experimental 0 20 40 60 80 100 0 20 40 60 80 100 M ode l Experimental
