Bioleaching of heavy metals from sewage sludge with recirculation of the liquid phase: A mass balance model

Accepted Manuscript Bioleaching of heavy metals from sewage sludge with recirculation of the liquid phase: A mass balance model Oleksiy Marchenko, Viktor Demchenko, Galina Pshinko PII: DOI: Reference:

S1385-8947(18)30998-7 https://doi.org/10.1016/j.cej.2018.05.174 CEJ 19190

To appear in:

Chemical Engineering Journal

Please cite this article as: O. Marchenko, V. Demchenko, G. Pshinko, Bioleaching of heavy metals from sewage sludge with recirculation of the liquid phase: A mass balance model, Chemical Engineering Journal (2018), doi: https://doi.org/10.1016/j.cej.2018.05.174

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.

Bioleaching of heavy metals from sewage sludge with recirculation of the liquid phase: A mass balance model Oleksiy Marchenko∗, Viktor Demchenko, Galina Pshinko A.V. Dumansky Institute of Colloid and Water Chemistry, National Academy of Sciences of Ukraine, Acad. Vernadsky blvd. 42, Kyiv 03680, Ukraine

Abstract Bioleaching of heavy metals from sewage sludge is an effective method for decontamination of sewage sludge. The process is currently under active investigation for its improvement of sewage sludge dewatering. The bioleaching with recirculation of the liquid and solid phases of sewage sludge is of considerable interest for the development of the bioleaching method itself, and for deeper investigation of its influence on the sludge dewatering. This work presents and verifies a mass balance model of elements for bioleaching of heavy metals from sewage sludge by indigenous iron-oxidizing bacteria with recirculation of the liquid phase of the treated sludge. Good correspondence between the model predictions and the performed experiments gives ground for further development of the model to describe substances that are not conserved during the bioleaching, especially extracellular polymeric substances that impede sewage sludge dewatering. The mass balance model is presented in a form that also describes recirculation of the solid phase of sewage sludge during the bioleaching. Recirculation of the liquid and solid phases of sewage sludge during the bioleaching, together with the mass balance model, give useful tools for future ∗

Corresponding author. Email address: [email protected] (Oleksiy Marchenko)

Preprint submitted to Chemical Engineering Journal

May 30, 2018

investigation and development of dewatering of sewage sludge with the use of the bioleaching process. Keywords: Bioleaching, Heavy metals, Iron-oxidizing bacteria, Mass balance, Sewage sludge

1

1. Introduction

2

Sewage sludge is a fertilizer that is contaminated with heavy metals, and the

3

majority of metals are bound to the solid phase of the sludge [1–10]. Heavy metals

4

enter sewage treatment systems with sewage from households, industrial wastewater,

5

or urban runoff [11, 12] and bind to particulate matter present in wastewater or

6

activated sludge floc during conventional sewage treatment [2, 7]. Sewage sludge is

7

the excess activated sludge and the particulate matter, which is removed prior to

8

wastewater treatment in the activated sludge process.

9

Bioleaching of heavy metals from sewage sludge is an effective method to transfer

10

metals from the solid phase of the sludge into its liquid phase [2, 13–15]. After

11

separation of part of the liquid phase, from which metals can be recovered [13, 16, 17],

12

the treated sludge usually meets the agricultural use standards. The bioleaching by

13

iron-oxidizing bacteria is one of the types of the process. It is often performed

14

as cultivation of these bacteria in sewage sludge with addition of FeSO4 ·7H2 O and

15

H2 SO4 in flasks or continuous flow stirred tank reactors. Bacteria that are usually

16

used in this approach to the bioleaching are either cultured strains of Acidithiobacillus

17

ferrooxidans [18, 19] or iron oxidizing bacteria that are indigenous to sewage sludge

18

[20, 21]. Other main approaches to the process include the bioleaching by sulfur-

19

oxidizing bacteria with addition of elemental sulfur and the bioleaching by a combined

20

culture of iron- and sulfur-oxidizing bacteria [2, 22]. A drawback of these approaches

21

is low conversion of the added elemental sulfur and potential soil acidification from 2

22

the residual sulfur in the decontaminated sludge [2].

23

In the past several years, many of the works on bioleaching of heavy metals from

24

sewage sludge have put the emphasis on the use of the bioleaching for sewage sludge

25

dewatering [18, 19, 21, 23–33], instead of development of the method and improve-

26

ment of its efficiency. The bioleaching significantly improves dewatering properties

27

of sewage sludge. In the simplest case, the dewatering properties improve, because

28

the acidic conditions that develop during the bioleaching process change the sur-

29

face potential of the sludge floc closer to the zero value than in the untreated sludge

30

[19, 21, 28]. More specifically, the bioleaching demonstrates better results than chem-

31

ical acidification, because it significantly affects extracellular polymeric substances

32

that impede sewage sludge dewatering [25–28, 30, 32]. But there is a potentially

33

interesting development of the bioleaching method itself that has been overlooked.

34

Figure 1

35

One of the approaches to bioleaching of heavy metals from sewage sludge is

36

the bioleaching by iron-oxidizing bacteria with recycling of the solid phase [34–38].

37

The bioleaching by iron-oxidizing bacteria with recirculation of the liquid phase has

38

not been investigated. A justification for the use of this approach is the presump-

39

tion of the bioleaching mechanism that is presented in Fig. 1a. This view is based

40

on the accepted presumption that ferrous iron is oxidized by the bacteria to ferric

41

iron, which in turn forms insoluble compounds, such as jarosite, schwertmannite

42

(Fe8 O8 (OH)6 SO4 ·nH2 O), or hydroxide [2, 14, 15, 39, 40]. It is implied in Fig. 1a that

43

recirculation of the liquid phase (Fig. 1b) may improve the process performance, al-

44

though the proof of such an improvement is beyond the scope of any practical work.

45

More practically, recirculation of the liquid phase during the bioleaching allows not

46

to treat part of the liquid phase of sewage sludge and, therefore, save reagents (Fig. 3

47

1b).

48

Recirculation of the liquid phase during the bioleaching is of potential interest

49

for development of dewatering of sewage sludge. The recirculation of the liquid

50

phase (Fig. 1b) may be seen as a way to shorten hydraulic retention time of the

51

solid phase during the bioleaching. The product of the bioleaching with respect to

52

dewatering of the sludge is the acidified and flocculated solid phase of sewage sludge.

53

Removing the solid phase from the reaction volume, instead of recirculating it, may

54

be considered as an approach to shift the equilibrium of the process further toward

55

formation of the product, in accordance with Le Chatelier’s principle. It is true,

56

whether the mechanism of the influence of the bioleaching on the sludge dewatering is

57

considered to be unknown [18, 25, 28] or understood [27, 29], as it is for any chemical

58

reaction. Unfortunately, no predictions about the process kinetics can be drawn from

59

a general thermodynamic argument such as above. Additionally, the liquid phase

60

for the recirculation has to be obtained from the treated sludge by sedimentation,

61

filtration, or centrifugation, and may contain loosely bound extracellular polymeric

62

substances. Moreover, without a generally accepted mechanism of the influence of

63

the bioleaching on the sludge dewatering, it is difficult to make predictions about

64

the time-independent optimum bioleaching parameters.

65

Before a difficult task of understanding the possible influences recirculation of

66

the liquid and solid phases during the bioleaching may have on dewatering of sewage

67

sludge, a simpler task has to be solved first. An investigation into the mechanism

68

of the influence of the bioleaching on dewatering of sewage sludge should be based

69

on mass balance of substances, primarily extracellular polymeric substances. The

70

amounts of these substances change during the bioleaching [25–28, 30, 32], and it is

71

difficult to propose a mechanism of the stated influence, to formulate a mass balance

72

model based on the mechanism, and to test the model, and thus confirm or disprove 4

73

the proposed mechanism. It is simpler first to develop a mass balance model for

74

substances that are known to be conserved during the bioleaching, such as heavy

75

metals.

76

The aim of this work is to develop a mass balance model of elements for bioleach-

77

ing of heavy metals from sewage sludge by indigenous iron-oxidizing bacteria with

78

recirculation of the liquid phase of the treated sludge.

79

2. Materials and methods

80

2.1. Sewage sludge sample

81

The sample of aerobically digested secondary sewage sludge was collected from

82

“Bortnyc’ka stanciya aeracii” STP in Kyiv, Ukraine. The STP receives 0.6–0.9 mln. m3

83

of domestic sewage and pretreated industrial wastewater daily; the STP does not

84

receive urban runoff. The sludge sample was taken in PET bottles, which were pre-

85

viously soaked in distilled water. The sludge was thickened twice by decantation and

86

stored in a refrigerator close to thawing ice before the bioleaching experiments were

87

performed. This thickened sludge constituted the feed sludge for the bioleaching

88

experiments. Total solids (TS) were analyzed by the standard method [41].

89

2.2. Bioleaching experiments

90

The semi-continuous flow stirred tank reactor was made from a glass cylinder

91

with a PTFE bottom and lid and was sealed with silicone hermetic. The reactor

92

total volume was 2 L; the reaction volume – 1.8 L. The reactor lid had apertures for

93

a heater, aerator, and thermometer, and also for pH, oxidation reduction potential

94

(ORP), and dissolved oxygen (DO) electrodes. Mixing of the sludge was performed

95

by a magnetic stirrer and a PTFE sealed magnetic stir bar, heating – by an aquarium

96

heater, aeration – through a silicone tube (the aerator) with punctures; the aerator 5

97

was connected to an aquarium air pump. ORP and pH values were measured poten-

98

tiometrically, DO – polarographically.

99

The bioleaching experiments were performed in a succession as follows: (a) en-

100

richment of the indigenous iron-oxidizing bacteria in sewage sludge; (b) the semi-

101

continuous flow bioleaching with addition of 0.1% FeSO4 ·7H2 O and recirculation of

102

the liquid phase (“0.1% Fe bioleaching”); (c) the semi-continuous flow bioleaching

103

without addition of ferrous iron and with recirculation of the liquid phase (“0% Fe bi-

104

oleaching”). The enrichment was performed by addition of 2% (20 g/L) FeSO4 ·7H2 O

105

to the twice diluted feed sludge and subsequent acidification of the sludge with 1 M

106

H2 SO4 to pH 3.5. This mixture was treated in the reactor for 10 days. The transi-

107

tion from the enrichment process to 0.1% Fe bioleaching was performed in a similar

108

manner to 0.1% Fe bioleaching, but the amounts of the added FeSO4 ·7H2 O were

109

gradually reduced from 2% to 0.1%. The transition from 0.1% to 0% Fe bioleaching

110

was performed in a similar manner to 0.1% Fe bioleaching, with the exception that

111

no ferrous iron was added to the sludge. The feed sludge during 0.1% and 0% Fe

112

bioleaching processes was added as follows:

113

– twice daily 300 mL of the sludge was taken from the reactor and filtered through

114

a paper filter (12–15 µm pore diam.) to obtain 150 mL of the liquid phase of the

115

sludge (the leachate);

116

– 150 mL of the feed sludge was mixed with 150 mL of the leachate;

117

– 0.1% FeSO4 ·7H2 O or none was added on the basis of the mixture volume (300 mL);

118

– the mixture was added to the reactor;

119

– 5.2–5.4 mL of 1 M H2 SO4 was added to the sludge in the reactor (17.3–18 mM on

120

the basis of the mixture volume; 34.6–36 mM on the basis of the feed sludge volume).

121

The sludge samples for iron-oxidizing bacteria cell counts were taken, and also pH,

122

ORP and DO values were measured, before the described above sludge substitution 6

123

procedures were performed. All the described experiments were performed at 28 ◦ C).

124

The used reagents were “purissimum” grade.

125

Hydraulic retention times (HRT) for both 0.1% and 0% Fe bioleaching processes,

126

defined as the reaction volume divided by the daily inflow of the sludge, were 3 days

127

(if the recirculated leachate is included in the inflow) or 6 days (if the recirculated

128

leachate is excluded from the inflow). The bioleaching processes were considered to

129

be steady after 10 days, at which time the sludge samples were taken for elemen-

130

tal analysis. The stability of the bioleaching processes was mainly assessed by the

131

stability of pH and ORP values. HRT of the solid phase in the experimental setup

132

was set to be 3 days, and from this standpoint the steady state was obtained after

133

3×HRT, as in [34–38]. Additionally, HRT of 3 days for the solid phase is of interest

134

for dewatering of sewage sludge, as it has been shown to give good results [19, 28, 33],

135

although other works have been performed or shown good results for HRT values of

136

1 [24], 2 [23, 25], or 4 [21] days.

137

The bioleaching without addition of FeSO4 ·7H2 O was initially considered as a

138

phenomenological control. The addition of 0.1% FeSO4 ·7H2 O was used to corre-

139

spond with the works on the recycling of the solid phase [34, 38], and it is also the

140

lowest concentration of the added ferrous iron that has been reported to sustain the

141

bioleaching in a continuous process. Temperature (28 ◦ C) was set close to the values

142

in a majority of works on bioleaching of heavy metals from sewage sludge, as well as

143

the TS values.

144

2.3. Elemental analysis

145

Elemental content was analyzed in the feed sludge, the sludge from 0.1% and 0%

146

Fe bioleaching processes, and in the liquid phase of the sludge (the leachate) from

147

0.1% and 0% Fe bioleaching processes. 7

148

The leachate samples were obtained as in the bioleaching experiments, and were

149

additionally vacuum filtered through nitrocellulose membrane filters (500–700 nm

150

pore diam.). The samples of the leachate were diluted with distilled water and

151

stabilized by addition of HNO3 . The samples of the sludges were microwave digested

152

in PTFE digestion vessels with 2:1 mixture of HNO3 and HCl (oven Speedwave

153

MWS-2, vessels DAP-60K, Berghof). After the digestion, the extracts were filtered

154

and diluted with distilled water. Blank controls of the preparation of the leachate

155

samples and the digested sludges were made. All glassware and the PTFE digestion

156

vessels were prepared by soaking in diluted HNO3 , with subsequent washing and

157

soaking in distilled water. The used reagents were “purissimum” grade.

158

Elemental concentrations (Li, Be, Mn, Co, Cu, Zn, Ga, As, Rb, Sr, Cd, Ba, Tl,

159

Pb, Bi, U) in the prepared samples were measured by inductively coupled plasma

160

mass spectrometry (spectrometer Agilent-7500, Agilent Technologies).

161

The presented results include only those elements, for which the total concentra-

162

tions and the concentrations in the liquid phase were measured for both 0.1% and 0%

163

Fe bioleaching processes, and the measured values were higher than their respective

164

total uncertainties. The total uncertainties and systematic errors are calculated as

165

described in [42]. For elemental analyses these errors include 0.5% relative errors in

166

the measurements of volumes, and 5% relative errors in the measurements of elemen-

167

tal concentrations in the samples and their corresponding blank controls (calibration

168

errors).

169

2.4. Microbiological analysis

170

Plate counts of the indigenous iron-oxidizing bacteria in the feed and the treated

171

sludge were performed on silica gel plates with the ferrous sulphate medium (9K) [43].

172

The sludge samples, taken for the plate counts, were diluted with 0.01 M phosphate 8

173

buffer and shaken. Plate spreading was performed by shaking, not with a glass rod.

174

The plates were left for 11–12 days at 23–28 ◦ C, and then the grown colonies were

175

counted.

176

Silica gel was prepared from a low grade stock solution of sodium silicate by ion

177

exchange with strong cation exchange resin and weak anion exchange resin, a 3:1

178

mixture of which was regenerated with 1 M H2 SO4 , as described in [44].

179

Photographs of the indigenous iron-oxidizing bacteria were made using transmis-

180

sion electron microscopy (electron microscope JEOL JEM-1400, JEOL).

181

2.5. Mass balance model verification

182

2.5.1. Overview

183

The mass balance model (Supplement A) is verified by a comparison of the mod-

184

eled values that are based on the experimental setup with the experimentally ob-

185

served values. The comparisons of the values are analyzed by the means of statistical

186

analyses.

187

The general statement of the null hypotheses used for the statistical analyses

188

below is: the proposed mass balance model is adequate. The p-value approach is

189

used to analyze the results. “The p-value is the smallest significance level at which

190

the null hypothesis is rejected”; “the commonly used values of the significance level

191

are 0.01, 0.025, 0.05, and 0.10” [45].

192

The performed Z-tests for comparisons of modeled and experimental values are

193

performed as described in [45], and the total uncertainties of the values are used in

194

the analyses to correspond with the fact that the systematic errors are known. The

195

F -tests for the one-factor analyses of variance (ANOVA) are carried out as described

196

in [46], and include systematic errors in the calculations. The performed t-tests for

197

correlations are carried out as described in [45]. 9

198

199

2.5.2. Distribution constants and efficiencies The distribution constants KD (L/g) for each element are calculated as:

KD =

CRt,exp − CRl,exp CRl,exp

·

1 , T SR

(1)

200

where CRl,exp is the experimentally measured concentration of an element in the liquid

201

phase of the sludge in the reactor, mol/L; CRt,exp is the experimentally measured total

202

concentration of an element in sewage sludge in the reactor, mol/L; T SF is the total

203

solids content of the sludge in the reactor, g/L.

204

The efficiencies E (%) for each element are calculated as:

E= 205

206

CRl,exp · 100%, CRt,exp

(2)

where the quantities are the same as in Eq. (1). The distribution constants and efficiencies for each element observed in 0.1% and

207

0% Fe bioleaching processes are compared with the use of Z-tests.

208

2.5.3. Concentrations in the liquid phase

209

210

The predicted values of concentrations of elements in the liquid phase of the sludge in the reactor CRl,mod (mol/L) are calculated as: CRl,mod =

1 1 + KD · T SF

· CFt ,

(3)

211

where CFt is the total concentration of an element in the feed sludge, mol/L. T SF is

212

the total solids content of the feed sludge, g/L; KD is as in Eq. (1).

213

214

The calculated values of CRl,mod are compared with the experimentally obtained values CRl,exp for each element with the use of Z-tests.

10

215

216

217

2.5.4. Total concentrations The predicted values of total concentrations of elements in the sludge in the reactor CRt,mod (mol/L) are calculated as: CRt,mod = CRl,mod + (CFt − CRl,mod ) · (1 − L),

(4)

218

where CRl,mod and CFt are as in Eq. (3); L is the coefficient of recirculation of the

219

liquid phase, it is defined as part of the liquid phase of the outflow from the reactor

220

that is being recirculated, it is equal to 0.5 for the experimental setup of this work,

221

dimensionless.

222

The calculated values of CRt,mod are compared with the experimentally obtained

223

values CRt,exp for each element with the use of Z-tests.

224

2.5.5. Invariants

225

The modeled and experimental concentrations of elements have significantly dif-

226

ferent values for different elements. Therefore, it is impossible to compare directly

227

the values obtained for different elements with the aim of assessing the validity of

228

the model. To that aim it is necessary to obtain an invariant quantity that should

229

be the same for different elements from the standpoint of the model. Such invariant

230

could be used to compare the experimental and predicted concentrations of different

231

elements as one group.

232

The invariants I exp and 1/I exp are calculated for each element as:

I 233

234

235

exp

≡

CRt,exp − CRl,exp CFt − CRl,exp

,

(5)

where CFt , CRt,exp and CRl,exp are the same as in Eq. (3, 4). These invariants are compared to the model invariant I mod that is given by the experimental setup as: 11

I mod ≡ 1 − L, 236

(6)

where L is the same as in Eq. (4).

237

The comparison between I exp and I mod , and also between 1/I exp and 1/I mod is

238

made for each bioleaching process separately and for both processes as for one dataset

239

with the use of F -tests and Z-tests.

240

2.5.6. Relative concentrations

241

The relative concentrations of elements in the liquid phase of the sludge in the

242

reactor and in total RC l,mod , RC l,exp , RC t,mod and RC t,exp (all %) are calculated as: CRl,i · 100%, CFt /2

(7)

CRt,i = t · 100%, CF /2

(8)

RC l,i =

RC

t,i

243

where i denotes either modeled (mod) or experimental (exp) values; CFt is the same

244

as in Eq. (3, 4).

245

Relative concentrations are calculated with the intention to show the modeled

246

and experimental concentrations of different elements on a similar basis. To that

247

aim, total uncertainties of the concentrations in the feed sludge are not included in

248

calculation of uncertainties for Eq. (7, 8). The used basis is the total concentrations

249

of elements in the feed sludge divided by 2. This basis approximately shows the

250

total elemental concentrations in the reactor that would have been obtained during

251

bioleaching of heavy metals from the initial (not thickened) sludge without recircu-

252

lation of the liquid phase. It is presumed that the concentrations of heavy metals

253

are twice higher in the feed sludge than in the initial sludge. It is reasonable to

12

254

presume so, because it is well known that the majority of heavy metals are predom-

255

inately bound to the solid phase of sewage sludge, and would not be lost during the

256

thickening. It is obviously not so for elements that are contained in high amounts in

257

the liquid phase of sewage sludge. The invariants described previously provide much

258

more strict basis that assumes only the properties of the model, but are too abstract

259

for presentation.

260

The correlations between the modeled and experimentally obtained values of the

261

relative concentrations of elements are analyzed with the use of t-tests.

262

3. Results

263

The initial sludge had TS (g/L) value 13.25, the feed sludge – 28.73, the treated

264

sludge –18.66. Dissolved oxygen values observed during the bioleaching experiments

265

were 5.5–7.5 mgO2 /L. pH and ORP values were 3.1 and 450 mV during 0.1% Fe

266

bioleaching, during 0% Fe bioleaching the observed values were 3.9 and 410 mV

267

(Fig. SA.2). The observed ORP value during 0% Fe bioleaching was higher than was

268

expected from the extrapolation presented in [38], while the pH value was close to the

269

expected. The cell counts of indigenous iron-oxidizing bacteria in the feed sludge were

270

approx. 200 cells/mL, in the sludge in the reactor after the enrichment – 107 cells/mL,

271

in the sludge in the reactor during 0.1% Fe bioleaching – 2·106 cells/mL, during

272

0% Fe bioleaching – 105 cells/mL (Fig. SA.2, SA.3). The observed and modeled

273

elemental concentrations are shown in the form of relative concentrations in Fig. 2.

274

The correlation analyses (the t-tests) for relative concentrations for 0.1% and 0% Fe

275

bioleaching processes give the p-values 1 for all tests (Fig. SA.4).

276

Figure 2

277

The Z-tests for the distribution constants and efficiencies show that 0.1% and 13

278

0% Fe bioleaching processes had relatively similar distribution of elements between

279

the sludge phases. The tests for KD s give the average p-value 0.361, median – 0.291,

280

lower quartile – 0.032, minimum – 0.000 (Be, Cu), maximum – 0.983 (Mn), values

281

below 0.05 were obtained for Be, Cu, As, Cd, Tl, Pb. The tests for efficiencies give

282

the average p-value 0.375, median – 0.326, lower quartile –0.085, minimum – 0.000

283

(Be, Cu), maximum – 0.983 (Mn), values below 0.05 were obtained for Be, Cu, Cd,

284

Tl.

285

The Z-tests for the concentrations of elements in the liquid phase for 0.1% and

286

0% Fe bioleaching processes taken separately shows good correspondence between

287

the model and the experiment for 0.1% Fe bioleaching, and reasonable – for 0%

288

Fe bioleaching. The tests for 0.1% Fe bioleaching give the average p-value 0.485,

289

median – 0.419, lower quartile – 0.215, minimum – 0.076 (Cu), maximum – 0.981

290

(Bi), none of the elements had values below 0.05. The tests for 0% Fe bioleaching

291

give the average p-value 0.261, median – 0.084, lower quartile – 0.040, minimum –

292

0.002 (Be), maximum – 0.842 (Mn), values below 0.05 were obtained for Li, Be, Ga,

293

Rb, Tl.

294

The Z-tests for the total concentrations of elements in the sludge for 0.1% and

295

0% Fe bioleaching processes taken separately show better correspondence between

296

the predicted and the experimental concentrations than for the concentrations in the

297

liquid phase. Additionally, the test was performed for TS values. The tests for 0.1%

298

Fe bioleaching give the average p-value 0.397, median – 0.321, lower quartile – 0.114,

299

minimum – 0.000 (TS), maximum – 0.981 (Rb), values below 0.05 was obtained for

300

TS and none of the elements. The tests for 0% Fe bioleaching give the average p-value

301

0.293, median – 0.207, lower quartile – 0.052, minimum – 0.000 (TS), maximum –

302

0.984 (Mn), values below 0.05 were obtained for TS, Li, Be, Tl.

303

The F -tests for ANOVA for the invariants show that the model and the exper14

304

iments correspond well. Additionally, the invariant is calculated for the values of

305

TS. The F -test for the invariant I for 0.1% Fe bioleaching gives the p-value 0.979,

306

and for the invariant 1/I – 0.115. For 0% Fe bioleaching the F -test for the invariant

307

I gives the p-value 0.936, and for the invariant 1/I – 0.976. For 0.1% and 0% Fe

308

bioleaching processes combined as one dataset the F -test for the invariant I gives

309

the p-value 0.996, and for the invariant 1/I – 0.338.

310

The Z-tests for the invariants show that the model and the experiments corre-

311

spond well. The test for the invariant I for 0.1% Fe bioleaching gives the p-value

312

0.714, for 0% Fe bioleaching – 0.382, and for the combined dataset – 0.883. The

313

test for the invariant 1/I for 0.1% Fe bioleaching gives the p-value 0.726, for 0% Fe

314

bioleaching – 0.322, and for the combined dataset – 0.871.

315

4. Discussion

316

There are two primary benefits that arise from the use of the proposed bioleaching

317

of heavy metals with recirculation of the liquid phase. The first is the reduction of

318

the amounts of the used reagents that comes simply from the fact that part of the

319

liquid phase of the initial sewage sludge need not be treated (Fig. SA.5). The second

320

benefit is the potentially high concentrations of heavy metals in the leachate, which

321

is beneficial for its subsequent treatment. Heavy metals are predominately contained

322

in the solid phase of sewage sludge, and the sludge has high viscosity. It would be

323

highly efficient to treat as much of the solid phase as possible, but this is not possible

324

because of the high viscosity of the sludge. Specifically, it is difficult to dewater the

325

sludge, and then mix and aerate it. The proposed approach allows to treat this

326

thickened sludge in the reactor, because the sludge is diluted with the recirculated

327

liquid phase. And the main consequence of this dilution by the recirculation is the

328

high concentrations of heavy metals that are described by the model. Specifically, if 15

329

the bioleaching for some heavy metal has high efficiency, then the metal concentration

330

in the reactor would be up to its total concentration in the thickened sludge. The sum

331

total for such case is the obtained concentration of the metal that is as high, as if the

332

bioleaching process directly treated the thickened sludge. Since subsequent treatment

333

of the leachate is concentration dependent [13, 16, 17], the proposed approach to the

334

bioleaching is favorable for this necessary next step and aims to overcome one of its

335

difficulties, specifically the relatively low metal concentrations in the leachate [13].

336

The proposed mass balance model provides a robust method to calculate the con-

337

centrations of elements in sewage sludge treated by the bioleaching with recirculation

338

of the liquid phase. The model can be used as an estimate for other experimental

339

setups with recirculation of the phases, because the distribution constants for many

340

elements, but not all, are similar under different bioleaching conditions. The model

341

can also be used, after appropriate verification, for the bioleaching with recirculation

342

of the solid phase, if used in the full form given in Supplement A. The shortcoming

343

of the proposed model is the poor prediction about TS values of the sludge in the

344

reactor. Better understanding of the size fractionation of the sludge solids during

345

filtration is needed to improve the model in this respect. Another area for improve-

346

ment is to account the volumes of the solid phase in the feed sludge and in the

347

treated sludge. Much more significant improvement of the model may be obtained

348

if the dependence of the distribution constants on the bioleaching parameters was

349

investigated in more depth.

350

The bioleaching with recirculation of the liquid and solid phases [34–38] together

351

with the presented mass balance model provide exceptionally useful tools to analyze

352

and develop dewatering of sewage sludge coupled with the bioleaching. The pre-

353

sented model is general and can be extended to describe mass balance of substances.

354

In the cases of added flocculants and added acid, recirculation of the liquid phase is a 16

355

method to recycle the unreacted substance. In the case of biologically produced floc-

356

culants that are predominately contained in the solid phase of the sludge [27, 29], the

357

recirculation of the solid phase is naturally expected to produce better experimental

358

results. It is known that extracellular polymeric substances significantly influence

359

sewage sludge dewatering properties [25–28, 30, 32, 47–50]. The mass balance model

360

would provide a simple method to predict and test the amounts of flocculants and

361

sludge born extracellular polymeric substances in the treated sludge, if it was to

362

be developed further. Such development should especially account distribution of

363

extracellular polymeric substances during separation of the sludge phases. This is

364

the most significant area of future research, to which the results of this work can be

365

applied.

366

5. Conclusions

367

The work presents and verifies a mass balance model of elements for bioleaching

368

of heavy metals from sewage sludge by indigenous iron-oxidizing bacteria with re-

369

circulation of the liquid phase of the treated sludge. Good correspondence between

370

the model predictions and the performed experiments gives ground for further de-

371

velopment of the model to describe substances that are not conserved during the

372

bioleaching, especially extracellular polymeric substances that impede sewage sludge

373

dewatering. Recirculation of the liquid and solid phases of sewage sludge during

374

the bioleaching, together with the mass balance model, give useful tools for future

375

investigation and development of dewatering of sewage sludge with the use of the

376

bioleaching process.

17

377

6. Appendix A. Supplementary material

378

Supplement A provides the derivation of the mass balance model and additional

379

figures. Supplement B provides elemental concentrations data, the model calcula-

380

tions, and data on model verification.

381

7. Acknowledgements

382

ICP-MS measurements were performed in the NASU Center of Collective Us-

383

age, “Gas chromatography-mass spectrometry and inductively coupled plasma mass-

384

spectrometry”, ICWC, NASU. Electron microscopy was performed in the NASU

385

Center of Collective Usage, “Laboratory for electron microscopy”, IMV, NASU. This

386

work was supported by the NASU Department of Chemistry (0112U000040).

387

8. Conflict of interest

388

O. Marchenko and G. Pshinko are coauthors of a utility model patent UA 103623

389

U that describes bioleaching of heavy metals from sewage sludge with recirculation

390

of half of the liquid phase of the treated sludge; ICWC, NASU is the proprietary

391

rights holder for the patent.

392

References

393

[1] A. Raheem, V. Sikarwar, J. He, W. Dastyar, D. Dionysiou, W. Wang,

394

M. Zhao, Opportunities and challenges in sustainable treatment and re-

395

source reuse of sewage sludge: A review, Chem. Eng. J. 337 (2018) 616–641.

396

doi:10.1016/j.cej.2017.12.149.

18

397

[2] R. Wong, J.W.-C. Tyagi, A. Pandey (Eds.), Current Developments in Biotech-

398

nology and Bioengineering: Solid Waste Management, Elsevier, Amsterdam,

399

etc., 2017.

400

[3] B. Duan, W. Zhang, H. Zheng, C. Wu, Q. Zhang, Y. Bu, Disposal situa-

401

tion of sewage sludge from municipal wastewater treatment plants (WWTPs)

402

and assessment of the ecological risk of heavy metals for its land use in

403

Shanxi, China, Int. J. Environ. Res. Public Health 14 (7) (2017) 823.

404

doi:10.3390/ijerph14070823.

405

[4] F. Suanon, Q. Sun, X. Yang, Q. Chi, S. Mulla, D. Mama, C.-P. Yu, Assessment

406

of the occurrence, spatiotemporal variations and geoaccumulation of fifty-two

407

inorganic elements in sewage sludge: A sludge management revisit, Sci. Rep. 7

408

(2017) 5698. doi:10.1038/s41598-017-05879-9.

409

410

[5] J. Wong, R. Surampalli, T. Zhang, R. Tyagi, A. Selvam (Eds.), Sustainable Solid Waste Management, American Society of Civil Engineers, Reston, 2016.

411

[6] A. Mulchandani, P. Westerhoff, Recovery opportunities for metals and

412

energy from sewage sludges, Bioresour. Technol. 215 (2016) 215–226.

413

doi:10.1016/j.biortech.2016.03.075.

414

[7] Y. Zhou, Z. Zhang, J. Zhang, S. Xia, Understanding key constituents and feature

415

of the biopolymer in activated sludge responsible for binding heavy metals,

416

Chem. Eng. J. 304 (2016) 527–532. doi:10.1016/j.cej.2016.06.115.

417

[8] A. Marchenko, G. Pshinko, V. Demchenko, Ecological substantiation of invari-

418

ably high concentrations of heavy metals in municipal wastewater sludges, J.

419

Water Chem. Technol 37 (4) (2015) 206–210. doi:10.3103/S1063455X15040098. 19

420

[9] P. Westerhoff, S. Lee, Y. Yang, G. Gordon, K. Hristovski, R. Halden, P. Herckes,

421

Characterization, recovery opportunities, and valuation in municipal sludges

422

from U.S. wastewater treatment plants nationwide, Environ. Sci. Technol.

423

49 (16) (2015) 9479–9488. doi:10.1021/es505329q.

424

[10] Y. Qi, D. Szendrak, R. Yuen, A. Hoadley, G. Mudd, Application of sludge

425

dewatered products to soil and its effects on the leaching behaviour of heavy

426

metals, Chem. Eng. J. 166 (2011) 586–595. doi:10.1016/j.cej.2010.11.029.

427

[11] A. Hargreaves, C. Constantino, G. Dotro, E. Cartmell, P. Campo, Fate and re-

428

moval of metals in municipal wastewater treatment: a review, Environ. Technol.

429

Rev. 7 (2018) 1–18. doi:10.1080/21622515.2017.1423398.

430

[12] L. Sorme, R. Lagerkvist, Sources of heavy metals in urban wastewater in Stock-

431

holm, Sci. Total Environ 298 (2002) 131–145. doi:10.1016/S0048-9697(02)00197-

432

3.

433

[13] J. Lema, S. Suarez (Eds.), Innovative Wastewater Treatment & Resource Recov-

434

ery Technologies: Impacts on Energy, Economy and Environment, IWA Pub-

435

lishing, London, 2017.

436

[14] C. Xu, Y. Zhang, M. Zhao, H. Rong, K. Zhang, Q. Chen, Comparison of bi-

437

oleaching and electrokinetic remediation processes for removal of heavy met-

438

als from wastewater treatment sludge, Chemosphere 168 (2017) 1152–1157.

439

doi:10.1016/j.chemosphere.2016.10.086.

440

[15] F. Camargo, P. Tonello, A. dos Santos, I. Duarte, Removal of toxic metals from

441

sewage sludge through chemical, physical, and biological treatmentsa review,

442

Water Air Soil Pollut. 227 (2016) 433. doi:10.1007/s11270-016-3141-3. 20

443

[16] Z. Djedidi, M. Bouda, M. Souissi, R. Cheikh, G. Mercier, R. Tyagi, J.-F. Blais,

444

Metals removal from soil, fly ash and sewage sludge leachates by precipitation

445

and dewatering properties of the generated sludge, J. Hazard. Mater. 172 (2–3)

446

(2009) 1372–1382. doi:10.1016/j.jhazmat.2009.07.144.

447

[17] N. Meunier, P. Drogui, C. Gourvenec, G. Mercier, R. Hausler, J. Blais, Removal

448

of metals in leachate from sewage sludge using electrochemical technology, En-

449

viron. Technol. 25 (2) (2004) 235–245. doi:10.1080/09593330409355457.

450

[18] Q. Zhou, J. Gao, Y. Li, S. Zhu, L. He, W. Nie, R. Zhang, Bioleaching in batch

451

tests for improving sludge dewaterability and metal removal using Acidithiobacil-

452

lus ferrooxidans and Acidithiobacillus thiooxidans after cold acclimation, Water

453

Sci. Technol. 76 (6) (2017) 1347–1359. doi:10.2166/wst.2017.244.

454

[19] J. Gao, Y. Shen, L. Li, J. Gao, Y. Li, C. Liu, J. Chen, Enhancing dewa-

455

terability of sewage sludge by the application of tween-20 during bioleach-

456

ing: Performance evaluation and mechanistic study, Drying Technol. (2017)

457

1–10doi:10.1080/07373937.2017.1355811.

458

[20] A. Mehrotra, T. Sreekrishnan, Heavy metal bioleaching and sludge stabiliza-

459

tion in a single-stage reactor using indigenous acidophilic heterotrophs, Environ.

460

Technol.doi:10.1080/09593330.2016.1275821.

461

[21] J.-M. Fontmorin, M. Sillanpaa, Bioleaching and combined bioleaching/Fenton-

462

like processes for the treatment of urban anaerobically digested sludge: Removal

463

of heavy metals and improvement of the sludge dewaterability, Sep. Purif. Tech-

464

nol. 156 (2015) 655–664. doi:10.1016/j.seppur.2015.10.061.

465

[22] H. Li, M. Ye, L. Zheng, Y. Xu, S. Sun, Q. Du, Y. Zhong, S. Ye, D. Zhang, 21

466

Optimization of kinetics and operating parameters for the bioleaching of heavy

467

metals from sewage sludge, using co-inoculation of two Acidithiobacillus species,

468

Water Sci. Technol. (2018) wst2018167doi:10.2166/wst.2018.167.

469

[23] J. Gao, N. Ma, L. Li, S. Zhu, Y. Li, J. Chen, Y. Chen, Improvement

470

of sewage sludge dewaterability by bioleaching in a continuous plug flow

471

bioreactor: Optimization of process parameters, Water. Sci. Technol. (2018)

472

wst2018083doi:10.2166/wst.2018.083.

473

[24] G. Zheng, Z. Wang, D. Wang, L. Zhou, Enhancement of sludge dewaterabil-

474

ity by sequential inoculation of filamentous fungus Mucor circinelloides ZG-

475

3 and Acidithiobacillus ferrooxidans LX5, Chem. Eng. J. 284 (2016) 216–223.

476

doi:10.1016/j.cej.2015.08.119.

477

[25] G. Zheng, M. Huo, L. Zhou, Extracellular polymeric substances level determines

478

the sludge dewaterability in bioleaching process, J. Environ. Eng. 142 (2) (2016)

479

04015060. doi:10.1061/(ASCE)EE.1943-7870.0001008.

480

[26] J. Wong, K. Murugesan, A. Selvam, B. Ravindran, M. Kurade, S.-

481

M. Yu, Dewatering of saline sewage sludge using iron-oxidizing bacteria:

482

Effect of substrate concentration, Bioresour. Technol. 213 (2016) 31–38.

483

doi:10.1016/j.biortech.2016.03.118.

484

[27] K. Murugesan, B. Ravindran, A. Selvam, M. Kurade, S.-M. Yu, J. Wong, Fate

485

of extracellular polymeric substances of anaerobically digested sewage sludge

486

during pre-dewatering conditioning with Acidithiobacillus ferrooxidans culture,

487

Bioresour. Technol. 217 (2016) 173–178. doi:10.1016/j.biortech.2016.03.081.

488

[28] H. Liu, S. Yang, J. Shi, X. Xu, H. Liu, B. Fu, Towards understanding the 22

489

dewatering mechanism of sewage sludge improved by bioleaching processing,

490

Sep. Purif. Technol. 165 (2016) 53–59. doi:10.1016/j.seppur.2016.03.037.

491

[29] M. Kurade, K. Murugesan, A. Selvam, S.-M. Yu, J. Wong, Sludge condi-

492

tioning using biogenic flocculant produced by Acidithiobacillus ferrooxidans

493

for enhancement in dewaterability, Bioresour. Technol. 217 (2016) 179–185.

494

doi:10.1016/j.biortech.2016.02.113.

495

[30] J. Wong, J. Zhou, M. Kurade, K. Murugesan, Influence of ferrous ions on extra-

496

cellular polymeric substances content and sludge dewaterability during bioleach-

497

ing, Bioresour. Technol. 179 (2015) 78–83. doi:10.1016/j.biortech.2014.10.099.

498

[31] C. Liu, P. Zhang, C. Zeng, G. Zeng, G. Xu, Y. Huang, Feasibility of bioleaching

499

combined with Fenton oxidation to improve sewage sludge dewaterability, J.

500

Environ. Sci. 28 (2015) 37–42. doi:10.1016/j.jes.2014.05.039.

501

[32] J. Zhou, G. Zheng, X. Zhang, L. Zhou, Influences of extracellular polymeric

502

substances on the dewaterability of sewage sludge during bioleaching, PLoS

503

ONE 9 (7) (2014) e102688. doi:10.1371/journal.pone.0102688.

504

[33] Y. Song, G. Zheng, M. Huo, B. Zhao, L. Zhou, Extracellular polymeric

505

substances and bound water drastically affect bioleached sludge dewater-

506

ability at low temperatures, Environ. Technol. 35 (20) (2014) 2538–2545.

507

doi:10.1080/09593330.2014.911755.

508

[34] D. Couillard, G. Mercier, Bacterial leaching of heavy metals from sewage

509

sludge – Bioreactors comparison, Environ. Pollut. 66 (3) (1990) 237–252.

510

doi:10.1016/0269-7491(90)90004-V.

23

511

[35] D. Couillard, G. Mercier, Optimum residence time (in CSTR and airlift reactor)

512

for bacterial leaching of metals from anaerobic sewage sludge, Water Res. 25 (2)

513

(1991) 211–218. doi:10.1016/0043-1354(91)90031-K.

514

[36] D. Couillard, S. Zhu, Bacterial leaching of heavy metals from sewage sludge

515

for agricultural application, Water Air Soil Pollut. 63 (1) (1992) 67–80.

516

doi:10.1007/BF00475622.

517

[37] D. Couillard, G. Mercier, Removal of metals and fate of N and P in the bacterial

518

leaching of aerobically digested sewage sludge, Water Res. 27 (7) (1993) 1227–

519

1235. doi:10.1016/0043-1354(93)90015-A.

520

[38] D. Couillard, M. Chartier, G. Mercier, Major factors influencing bacterial leach-

521

ing of heavy metals (Cu and Zn) from anaerobic sludge, Environ. Pollut. 85 (2)

522

(1994) 175–184. doi:10.1016/0269-7491(94)90084-1.

523

[39] Y. Liao, L. Zhou, S. Bai, J. Liang, S. Wang, Occurrence of biogenic schw-

524

ertmannite in sludge bioleaching environments and its adverse effect on sol-

525

ubilization of sludge-borne metals, Appl. Geochem. 24 (9) (2009) 1739–1746.

526

doi:10.1016/j.apgeochem.2009.05.003.

527

[40] A. Marchenko, G. Pshinko, V. Demchenko, V. Goncharuk, Bioleaching of heavy

528

metals from wastewater sludge by ferrous iron oxidizing bacteria, J. Water

529

Chem. Technol. 38 (1) (2016) 51–55. doi:10.3103/S1063455X16010094.

530

531

[41] APHA, AWWA, WEF, Standard Methods for the Examination of Water and Wastewater, 20th Edition, APHA, AWWA, WEF, 1999.

532

[42] J. Taylor, An Introduction to Error Analysis. The Study of Ucertainties in

533

Physical Measurements, 2nd Edition, University Science Books, Sausalito, 1997. 24

534

535

[43] M. Dworkin (Ed.), The Prokaryotes, 3rd Edition, Vol. 5: Proteobacteria: Alpha and Beta Subclasses, Springer, Singapore, 2006.

536

[44] A. Marchenko, G. Pshinko, V. Demchenko, V. Goncharuk, Leaching heavy metal

537

from deposits of heavy metals with bacteria oxidizing elemental sulphur, J.

538

Water Chem. Technol. 37 (6) (2015) 311–316. doi:10.3103/S1063455X15060090.

539

[45] P. Mann, Introductory Statistics, 7th Edition, John Wiley & Sons, Hoboken,

540

541

542

2010. [46] S. Ross, Introductory Statistics, 3rd Edition, Academic Press, Elsevier, Amsterdam, etc., 2010.

543

[47] Y. Dai, S. Huang, J. Liang, S. Zhang, S. Sun, B. Tang, Q. Xu, Role of organic

544

compounds from different EPS fractions and their effect on sludge dewaterabil-

545

ity by combining anaerobically mesophilic digestion pre-treatment and Fentons

546

reagent/lime, Chem. Eng. J. 321 (2017) 123–138. doi:10.1016/j.cej.2017.03.107.

547

[48] J. Zhang, J. Zhang, Y. Tian, N. Li, L. Kong, L. Sun, M. Yu, W. Zuo,

548

Changes of physicochemical properties of sewage sludge during ozonation treat-

549

ment: correlation to sludge dewaterability, Chem. Eng. J. 301 (2016) 238–248.

550

doi:10.1016/j.cej.2016.04.151.

551

[49] Y. Shi, J. Yang, W. Yu, S. Zhang, S. Liang, J. Song, Q. Xu, N. Ye, S. He,

552

C. Yang, J. Hu, Synergetic conditioning of sewage sludge via Fe2+ /persulfate

553

and skeleton builder: Effect on sludge characteristics and dewaterability, Chem.

554

Eng. J. 270 (2015) 572–581. doi:10.1016/j.cej.2015.01.122.

555

[50] R. Mo, S. Huang, W. Dai, J. Liang, S. Sun, A rapid Fenton treatment

25

556

technique for sewage sludge dewatering, Chem. Eng. J. 269 (2015) 391–398.

557

doi:10.1016/j.cej.2015.02.001.

26

Figure 1: The presumed (but not tested) mechanism of bioleaching of heavy metals from sewage sludge by iron-oxidizing bacteria (a), and the setup of the bioleaching with recirculation of the liquid phase (b).

Figure 2: The experimentally observed and predicted relative concentrations of elements in total (a) and in the liquid phase (b) of sewage sludge treated in 0.1% and 0% Fe bioleaching processes.

27