Chemolithotrophic denitrification in biofilm reactors

Chemolithotrophic denitrification in biofilm reactors

Chemical Engineering Journal 280 (2015) 643–657 Contents lists available at ScienceDirect Chemical Engineering Journal journal homepage: www.elsevie...

1MB Sizes 4 Downloads 140 Views

Chemical Engineering Journal 280 (2015) 643–657

Contents lists available at ScienceDirect

Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej

Review

Chemolithotrophic denitrification in biofilm reactors Francesco Di Capua a,b,⇑, Stefano Papirio a, Piet N.L. Lens b,c, Giovanni Esposito a a

Department of Civil and Mechanical Engineering, University of Cassino and Southern Lazio, via Di Biasio 43, 03043 Cassino (FR), Italy Department of Chemistry and Bioengineering, Tampere University of Technology, P.O. Box 541, FIN-33101 Tampere, Finland c UNESCO-IHE, Institute for Water Education, PO Box 3015, 2601 DA Delft, The Netherlands b

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

 Comparative analysis of biofilm

reactors for chemolithotrophic denitrification.  Water characteristics strongly affect the bioreactor denitrification performance.  A decision tool for choosing the most appropriate bioreactor is provided.  Fluidized bed and membrane biofilm reactors have the highest denitrification rates.  Packed bed and biofilm electrode reactors are cost-effective alternatives.

a r t i c l e

i n f o

Article history: Received 16 April 2015 Received in revised form 29 May 2015 Accepted 30 May 2015 Available online 15 June 2015 Keywords: Biofilm Biofilm electrode reactor Chemolithotrophic denitrification Fluidized bed reactor Membrane biofilm reactor Packed bed reactor

a b s t r a c t Chemolithotrophic denitrification is an inexpensive and advantageous process for nitrate removal and represents a promising alternative to classical denitrification with organics. Chemolithotrophic denitrifiers are microorganisms able to reduce nitrate and nitrite using inorganic compounds as source of energy. Ferrous iron, sulfur-reduced compounds (e.g. hydrogen sulfide, elemental sulfur and thiosulfate), hydrogen gas, pyrite and arsenite have been used as inorganic electron donors resulting in diverse outcomes. In the last 40 years, a large number of engineered systems have been used to maintain chemolithotrophic denitrification and improve rate and efficiency of the process. Among them, biofilm reactors proved to be robust and high-performing technologies. Packed bed reactors are particularly suitable for the removal of low nitrate concentrations, since high retention times are required to complete denitrification. Fluidized bed and membrane biofilm reactors result in the highest denitrification rates 3 (>20 kg N-NO 3 /m d) when hydrogen gas and sulfur reduced compounds are used as electron donors. Hydrogen gas pressure and current intensity rule the performance of membrane biofilm and biofilm electrode reactors, respectively. Biofouling is the most common and detrimental issue in biofilm reactors. Bed fluidization and hydrogen supply limitation are convenient and effective solutions to mitigate biofouling. Ó 2015 Elsevier B.V. All rights reserved.

Abbreviations: 2-CP, 2-chlorophenol; AA, activated alumina; ASD, aeration denitrification system; BER, biofilm electrode reactor; COD, chemical oxygen demand; DAS, denitrification–aeration system; DBCP, dibromochloropropane; DCM, dichloromethane; DO, dissolved oxygen; EBCT, empty bed contact time; EPS, extracellular polymeric substances; FBR, fluidized bed reactor; GAC, granular activated carbon; HFMBfR, hollow-fiber-membrane biofilm reactors; HRT, hydraulic retention time; IE, ion exchange; MBR, membrane biological reactor; MBfR, membrane biofilm reactor; MF, microfiltration; ORP, oxidation–reduction potential; p-CNB, para-chloronitrobenzene; PBR, packed bed reactor; PVA, polyvinylalcohol; RO, reverse osmosis; SLAD, sulfur–limestone autotrophic denitrification; SOM, sedimentary organic matter; SPEME, solid–polymer– electrolyte membrane electrode; SRB, sulfate-reducing bacteria; TCE, trichloroethene; TON, total organic nitrogen; TSS, total suspended solids. ⇑ Corresponding author at: Department of Chemistry and Bioengineering, Tampere University of Technology, P.O. Box 541, FIN-33101 Tampere, Finland. E-mail addresses: [email protected], francesco.dicapua@tut.fi (F. Di Capua), [email protected] (S. Papirio), [email protected] (P.N.L. Lens), giovanni. [email protected] (G. Esposito). http://dx.doi.org/10.1016/j.cej.2015.05.131 1385-8947/Ó 2015 Elsevier B.V. All rights reserved.

644

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657

Contents 1. 2.

3.

4.

5.

6.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Packed bed reactor (PBR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Reactor description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1. Biofilm carrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.2. Maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Sulfur-driven denitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1. Sulfur–limestone autotrophic denitrification (SLAD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2. Sulfur-driven denitrification with bicarbonate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3. Simultaneous heterotrophic and sulfur-utilizing autotrophic denitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Hydrogenotrophic denitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.1. Carrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.2. H2 supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.3. Operating parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.4. Fe(0)-supported hydrogenotrophic denitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Arsenite oxidation coupled to nitrate reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fluidized bed reactor (FBR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. FBR vs PBR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Hydrogenotrophic denitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Sulfur-driven denitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Denitrification coupled to pyrite oxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Costs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Membrane biofilm reactor (MBfR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Potable water treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Hollow fibers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Silicone membranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4. Hydrogen gas flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5. Biofilm density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6. Membrane biofouling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7. Costs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biofilm electrode reactor (BER) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Electrode material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Current intensity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3. Cathodic surface area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4. Maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5. Costs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Guidelines for the selection of reactor technology and configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1. Packed bed reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Fluidized bed reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3. Membrane biofilm reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4. Biofilm electrode reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5. Reactor selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Funding sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction In the last 40 years the increase of nitrate (NO 3 ) concentration in groundwater has affected many areas of the world and has become an environmental problem of international interest [1– 3]. If over-discharged, NO 3 favors the eutrophication of receiving waters [4] and increases the risk of severe diseases such as methemoglobinemia [5]. Chemolithotrophic denitrification has recently gained increasing interest. Autotrophic denitrifiers are microbes capable of using inorganic reduced compounds (e.g. sulfur-reduced compounds, ferrous iron, hydrogen gas, pyrite and arsenite) as electron donors by fixing inorganic carbon [6–8]. Compared to traditional nitrate removal performed by heterotrophic denitrifiers, chemolithotrophic denitrification results in two major advantages: (1) no external organic substrate needs to be added, decreasing the operation costs [9]; (2) limited biomass production and lower biofouling [10]. Various engineered systems have been studied over the recent years to carry out chemolithotrophic denitrification. Most studies focused on lab-scale applications in continuous reactors and

644 645 645 645 645 645 646 646 646 647 647 647 647 648 648 649 649 649 649 649 649 650 650 650 651 651 651 651 652 652 652 652 653 653 653 654 654 654 654 654 654 654 654 654

different configurations and operating conditions were applied. Continuous reactors are mainly divided in two categories: attached- and suspended-growth systems. Attached-growth systems (or biofilm systems) proved to be a reliable technology with smaller footprints and lower capital and operating costs compared to suspended-growth systems [11]. Carrier materials used in these systems have an extremely high specific surface area which allows high biomass concentrations and, thus, the treatment of high nitrate loadings. Several studies report higher denitrification rates in biofilm systems such as fixed-bed or fluidized-bed reactors compared to suspended growth reactors [12–15]. A large number of studies concerning sulfur-based and hydrogenotrophic denitrification in continuous reactors exist in literature, whereas only few studies focused on ferrous iron-, pyrite- and arsenite-driven denitrification. Simultaneous autotrophic and heterotrophic denitrification was also investigated as a solution to enhance nitrate removal and limit the drawbacks of autotrophic denitrification (e.g. alkalinity consumption and high sulfate production). Reactor performance is not the unique parameter to be considered in the choice of a chemolithotrophic

645

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657 Table 1 Performances and operating conditions of PBRs maintaining sulfur-driven autotrophic denitrification. Influent type

Reactor volume [L]

Synthetic wastewater Groundwater Municipal wastewater Synthetic/real groundwater Septic tank effluent Synthetic wastewater

5

Sulfur particles

d = 2.38–19 d = 2.38–6.68 d = 0.2–0.6 cm Powder d: 2.38–4.76 mm

Carbon source

HRT/EBCT* [h]

Reference Denitrification rate 3 [kg N-NO 3 /m d]

50

20–25

2.8–18.1 20.2 8 2–3 1.5–12 3 414

0.38 0.06

16–19 28–364 15–60 30–50 as TN 60–400

NaHCO3 CaCO3 CaCO3 CaCO3 CaCO3 CaCO3 NaHCO3

20–25

100

NaHCO3 7.1–1.5

35

530–900

Organics 10.41–40.56 2.74

[32]

5–32 35

10–100 175–700

1–6 CaCO3 NaHCO3 1.5–15.4

3.3 3.84.5

[25] [30]

24 ± 1 20–25

100 138 ± 2.3

NaHCO3 1–9.5 NaHCO3 14

0.20 0.0945–0.232

[31] [104]

20–25 30 25 ± 0.2 20 22 ± 2 22 ± 2 24 ± 1

60–251 30 20–40 as N-NH4+ 30–60 250 250 30 7.6–92.4 18–102 60 600 ± 20

CaCO3 CO2 CaCO3 CaCO3 CaCO3 CaCO3 CaCO3 CaCO3 CaCO3 CaCO3 NaHCO3

0.27–0.87 0.34 0.0010.003 0.12 0.41 0.682 0.202 0.20 0.0784–0.30 0.06 0.539–3.29

[105] [94] [106] [107] [23] [24] [108] [109] [110] [111] [33]

14

NaHCO3 6–24

0.0141

[34]

75 50–75

CH3OH CaCO3

0.5 0.1

[36] [37]

20

2.8  10 1.4 7.2–7.5 1.49 7.9–8.1 37.7 9.1 (bed volume) d: 2.8–16 mm Porosity: 40% (bed porosity) Synthetic wastewater 0.250 d: 4–4.75 mm 7 ± 0.2 Porosity: 48–50% Synthetic wastewater/nitrified 3.38 d: 2.82–5.84 mm 7.5–8 leachate from a landfill Synthetic wastewater 3.2 d: 3–15 mm 6–7.5 Synthetic wastewater 1.28 d = 2.82–5.84 Porosity: 39% (bed) Groundwater 0.6837 Porosity: 43.3% (bed) Synthetic wastewater 0.25 d: 2–4 mm 7.1–7.4 Porosity: 48–50% Synthetic wastewater 3.72 d = 2.8–5.6 mm Drinking water 1.23 d = 3–4 mm Wastewater 479 7.67 Groundwater 2.7 d = 2 mm 7 Saline wastewater 9.1 d = 5.6–11.2 mm Saline wastewater 9.1 d = 5.6–11.2 mm Groundwater 0.0651 d = 2.38–4.76 mm Synthetic wastewater 1.49 d = 2.38–4.76 mm 7.8–8.5 Groundwater 0.4 Synthetic groundwater 85 d = 5–10 mm 5.23 Synthetic wastewater 7.85 d: 2.8–5.6 mm Porosity: 40% Synthetic inorganic 1 wastewater Synthetic wastewater 0.4 d: 0.5–1 mm 8 Synthetic wastewater 0.4 d: 0.5–1 mm 8–8.5 *

Feed nitrogen [mg N-NO 3 /L]

Influent T [°C] pH

15–25

30 ± 2 25 ± 2

28–30 28–30

5.47–8.86 0.5–14 10 12 14.3–30.5 3.56 2.5–10.9 1.8–30.5 24–48 2.67–26.7

3.5–8.5 11.5–31.2

[21]

0.480.77

[27] [22] [16] [101] [102]

1.42.69

[103]

0.36 0.384

Depending on the retention time used by the authors.

denitrification system. Capital, operation and management costs are important factors which determine the actual feasibility of a treatment system. Cost estimation is an indispensable tool to be used in order to compare different autotrophic denitrification systems and emphasize the cost advantages of a system over the others. This paper reviews the most used biofilm reactor configurations for maintaining chemolithotrophic denitrification. Performances, costs, potentials and drawbacks of packed bed (PBR), fluidized bed (FBR), membrane biofilm (MBfR) and biofilm electrode (BER) reactors are critically reviewed. 2. Packed bed reactor (PBR) 2.1. Reactor description In a packed bed reactor (PBR) a biofilm grows around a fixed carrier material. PBR hydrodynamics is similar to the one of a plug-flow system, in which the concentration of the substrates decreases axially from the inlet to the outlet. PBRs can be operated either in upflow or in downflow mode, according to the specific weight of the carrier used. 2.1.1. Biofilm carrier Generally, the carrier consists in a porous organic/inorganic matrix composed by particles with high specific surface area. Polymers, active carbon and natural inorganic materials such as granular sulfur or gravel are the most used carrier materials in PBR applications (Table 1). The carrier is reported to play a major

role in the performance of a PBR. Carrier porosity and particle size determine biofilm thickness and pore clogging, whereas the specific surface area of the particles determines the available surface for biomass growth [11]. High biomass concentrations in the filter media allow high treatment capacities, but result also in maintenance issues. 2.1.2. Maintenance The simple configuration of PBR results in ease of maintenance and cost-effectiveness. However, excess biofilm is responsible for clogging and channeling in the filter media, which reduce denitrification efficiency and call for periodical backwashing. Backwashing frequency also depends on chemical composition of the influent. Real groundwater is reported to foul the column at a frequency 3–6 times higher than a synthetic one [16]. N2 supersaturation of water also affects PBR operation. Nitrogen bubbles may surround the carrier granules, increasing the column pressure and causing fluctuation in the bed porosity. Moreover, nitrogen gas accumulation around the pores may also hinder nitrate mass transfer from the bulk liquid to the biofilm and cause channeling [17]. Decrease of sulfur grain size with time may further result in clogging and entrapment of nitrogen gas. Bed fluidization is indeed a solution to overcome the drawbacks related to biofouling and N2 accumulation. 2.2. Sulfur-driven denitrification Elemental sulfur, hydrogen sulfide and thiosulfate are the main electron donors used in PBR applications performing sulfur-based

646

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657

chemolithotrophic denitrification. Granular elemental sulfur, with its high specific surface area, can be used both as electron donor and biomass carrier. 2.2.1. Sulfur–limestone autotrophic denitrification (SLAD) Autotrophic denitrification in PBR packed with elemental sulfur was mainly carried out with the use of limestone in the so-called sulfur–limestone autotrophic denitrification (SLAD) process. Limestone is used to buffer the acidity produced by the process and provide an inorganic carbon source for microbial synthesis. 2.2.1.1. Influent water quality. Early studies performed SLAD in PBRs with influent nitrate concentrations up to 50 mg N-NO 3 /L [18–21]. Once tested its feasibility, SLAD was applied to the treatment of groundwater and wastewater from different sources. Kuai and Verstraete [22] achieved total organic nitrogen (TON) removal up to 86% from municipal wastewater at hydraulic retention times (HRTs) ranging from 2 to 3 h. Saline wastewater was treated in several sulfur/limestone PBRs by Gu et al. [23] and Zhao et al. [24]. 3 Denitrification rates of 0.42 and 0.68 kg N-NO 3 /m d were achieved at HRT of 14.4 and 8.8 h, respectively. The treatment of nitrate contaminated groundwater resulted in lower denitrification rates due to low feed nitrate concentration (Table 1). In lab-scale applications, the use of a synthetic influent is generally preferred to test the effect of different parameters on the process as well as the influence of inhibitors and/or toxic compounds (e.g. heavy metals) in various concentrations. 2.2.1.2. Feed nitrate and temperature. SLAD has been observed to achieve high nitrate removal efficiencies at both low and high feed nitrate. In order to keep a stable efficiency at increasing feed nitrate, increase in HRT is generally required. Zhou et al. [25] achieved nitrate removal efficiencies higher than 90% at influent nitrate concentrations ranging from 10 to 40 mg N-NO 3 /L, HRT of 3 h and temperature in the range 20–25 °C. Increase in HRT to 6 h was necessary to achieve good denitrification efficiencies when treating influent nitrate concentrations higher than 70 mg N-NO 3 /L. Also temperature plays an important role. Decrease in temperature to 5–10 °C reduced denitrification efficiency to 50% [25].

2.2.1.5. Costs. SLAD is a cost-effective process. Zhang [29] carried out a preliminary economic analysis on the upgrade of a municipal drinking water treatment system for a small community (200 people) through the addition of a SLAD unit. The economic analysis showed that the SLAD reactor system is inexpensive, resulting in a total cost ranging from 51,000 to 59,000 US$ depending on different scenarios. 2.2.2. Sulfur-driven denitrification with bicarbonate The external supplementation of limestone (CaCO3) is responsible for hardness increase in reactor effluent, which results in detrimental effects especially when high nitrate concentrations are used. Moreover, high Ca2+ concentrations may also result in Ca3(PO4)2 and CaHPO4 precipitation due to their low solubility products, which limits phosphorous bioavailability and thus bacterial growth [17]. Bicarbonate has been used as alternative to limestone as source of carbon for sulfur-driven denitrification. The influence of flow rate, sulfur particle size and influent nitrate concentration on nitrate removal has also been investigated in sulfur PBRs amended with bicarbonate [27–29]. 2.2.2.1. Nitrate loading rate. Low flow rates and influent nitrate concentrations result in high denitrification efficiencies. Denitrification efficiencies higher than 95% were achieved and maintained in three sulfur PBRs by decreasing nitrate loading rate 3 from 2.46 to 1.64 kg N-NO 3 /m d as the influent nitrate concentration was increased from 175 to 700 mg N-NO 3 /L [30]. Soares [31] achieved a complete nitrogen removal in an upflow PBR at nitrate 3 loading rates lower than 0.2 kg N-NO 3 /m d. The increase of nitrate 3 loading rate up to 0.24 kg N-NO 3 /m d caused a breakthrough in nitrate and nitrite concentrations up to 1.6 mg N-NO 3 /L and 1.4 mg N-NO 2 /L, respectively.

2.2.1.3. Sulfur grain size. Sulfur particle size affects denitrification rate. The minimum HRT required for complete nitrate removal resulted in a linear correlation with sulfur surface area [21]. Decrease in sulfur grain size permits the use of lower HRTs, increasing denitrification potential of PBRs.

2.2.2.2. Sulfur grain size. Sulfur grain size affects denitrification. Low sulfur grain sizes result in higher specific surface areas, which are beneficial for biofilm growth and enhance denitrification. Driscoll and Bisogni [21] observed that the minimum hydraulic retention time required for complete nitrate removal in a sulfur PBR decreased from 18 to 3 h when the sulfur particle size range was lowered from 19–12.7 to 2.38–6.68 mm. A similar result was obtained by Koenig and Liu [102], who achieved the highest deni3 trification rate (0.77 kg N-NO 3 /m d) using the lower sulfur particles size range (2.8–5.6 mm) among those tested. In contrast, an extremely low bed porosity may result in severe clogging and limit biofilm development. As result, the choice of sulfur grain size is of major importance to optimize the performance of a PBR.

2.2.1.4. Sulfide production. An important issue regarding the operation of a SLAD reactor is the production of sulfide. In particular, sulfide production may take place in the upper part of the sulfur/limestone bed, where anaerobic zones (no oxygen and no nitrate) may occur especially at low nitrate loading rates. In these zones, some anaerobic chemolithotrophs can grow by disproportionation of elemental sulfur to sulfate and hydrogen sulfide, whereas sulfate-reducing bacteria (SRB) can thrive using the organic carbon provided by the endogenous respiration of autotrophic bacteria as energy and carbon source [26]. Sulfide is a toxic, corrosive and malodorous compound and its emission into the environment is undesirable. The formation of sulfide in a full-scale SLAD reactor also resulted in severe clogging and head loss in the infiltration pond downstream, due to sulfide conversion to sulfur by both chemical and biological processes [27]. The use of 3 nitrate loading rates higher than 0.22 kg N-NO 3 /m d are reported to limit sulfide production. However, some nitrate breakthrough may occur at these rates [28].

2.2.3. Simultaneous heterotrophic and sulfur-utilizing autotrophic denitrification 2.2.3.1. Organic carbon. The addition of external organic carbon is reported to reduce both alkalinity consumption and sulfate production due to sulfur-driven denitrification. Methanol and sodium acetate resulted in more suitable and effective organic carbon sources than molasses and glucose, since lower concentrations were necessary to eliminate the need of external alkalinity [32]. In particular, methanol addition in concentration equivalent to 60% of its stoichiometric requirement for heterotrophic denitrification provided the alkalinity required by autotrophic denitrifiers to complete denitrification [32,33]. Wood was also used as biomass carrier and carbon source. Yamashita et al. [34] operated a PBR packed with wood to treat synthetic wastewater with a nitrate concentration of 14 mg N-NO 3 /L. Both sulfate reduction and heterotrophic denitrification took place inside the reactor and provided the electron donor (H2S) and the alkalinity needed to sulfur-driven denitrification, respectively. A maximum nitrate

647

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657 3 removal rate of 14.1 g N-NO 3 /m d and denitrification efficiency of 99% were achieved at a HRT of 24 h.

was not affected by chromate even at 20 mg Cr(VI)/L and complete Cr(VI) reduction was attained up to 10 mg Cr(VI)/L. 2.3. Hydrogenotrophic denitrification

2.2.3.2. EBCT. The effect of empty bed contact time (EBCT) on simultaneous heterotrophic and sulfur-utilizing autotrophic denitrification was investigated by Aminzadeh et al. [33] in cylindrical columns packed with sulfur beads. EBCTs from 8 to 20 h resulted in almost complete denitrification, whereas EBCT values of 2 and 4 h negatively affected denitrification efficiency. Decrease in EBCT affected both heterotrophic and autotrophic denitrification, increasing the chemical oxygen demand (COD) in the effluent and decreasing sulfate production, respectively. The contemporary increase of pH and alkalinity indicated that lower EBCTs favored heterotrophic over autotrophic denitrification. This indicates that the EBCT can control the population ratio of heterotrophic to autotrophic denitrifiers.

2.3.1. Carrier Sand, silicic gravel, glass, polyurethane and bio-ceramsite have been used as carrier materials in PBRs performing hydrogenotrophic denitrification (Table 2). Although polymeric membranes feature high specific surface areas, natural materials usually result in inexpensive solutions and optimal performances. Vasiliadou et al. [12] operated a triple stage PBR packed with silicic gravel which achieved high denitrification rates up to 3 6.2 kg N-NO at influent nitrate concentration of 3 /m d 3 2 340 mg N-NO 3 /L and hydraulic loading rates of 11.5 m /m d.

2.2.3.3. Salinity. The effect of NaCl concentration on simultaneous heterotrophic and sulfur-based autotrophic denitrification was investigated in an upflow PBR with methanol as organic carbon source. Complete nitrate removal was achieved up to 3.5% NaCl, whereas a decrease in nitrate removal efficiency to 78% and 48% was observed at 4% and 5% NaCl, respectively [33]. In a previous study, a strong inhibitory effect on methanol-driven denitrification was observed at 4% NaCl, whereas the use of acetate resulted in high nitrate removal efficiency even at 10% NaCl [35]. In another study, 3.3% NaCl did not affect the removal of 250 mg N-NO 3 /L in a SLAD column [23].

2.3.2. H2 supply In PBRs performing hydrogenotrophic denitrification hydrogen supply is carried out mostly delivering H2 in an external tank for gas adsorption or sparging H2 gas directly into the bioreactor. Direct H2 supply has been carried out with several solutions in order to improve hydrogen diffusion. Gas-permeable membranes have been widely used in PBR applications resulting in high denitrification performances [38]. Hydrogen gas can be also produced directly inside the bioreactor. Water electrolysis [39–43] and anoxic corrosion of metallic iron [44] are among the most used methods for hydrogen production in PBRs. Methanol electrolysis, producing both H2 and CO2, was coupled to autotrophic denitrification in a PBR operated by Vagheei et al. [45]. Nitrate removal effi3 ciencies above 95% and removal rates up to 0.339 kg N-NO 3 /m d were achieved at HRTs ranging from 2 to 5 h.

2.2.3.4. Chromate toxicity. Chromate toxicity on heterotrophic, sulfur-based autotrophic and mixotrophic denitrification was investigated in PBRs performing simultaneous chromate and nitrate reduction [36,37]. Autotrophic denitrification was complete and not adversely affected by chromate up to 0.5 mg Cr(VI)/L. However, mixotrophic and heterotrophic denitrification resulted in more robust processes. Mixotrophic denitrification was almost complete up to 10 mg Cr(VI)/L, whereas denitrification efficiency sharply decreased at 15 mg Cr(VI)/L. Heterotrophic denitrification

2.3.3. Operating parameters The effects of influent nitrate concentration, C/N ratio, temperature and pH on hydrogenotrophic denitrification in PBRs were evaluated by Chen et al. [46]. The gradual increase of the feed nitrate from 30 to 130 mg/L increased the denitrification rate up 3 to 0.18 kg N-NO 3 /m d, whereas hydrogenotrophic denitrification was inhibited at higher influent nitrate concentrations. C/N ratio

Table 2 Hydrogenotrophic denitrification in PBRs. Support media

Reactor volume [L]

Polypropylene carrier Polyurethane carrier Lamellar reticulated polyurethane Granulated activated carbon Aquarium rocks Iron mixed with sand Steel wool mixed with sand Pea gravel Silicic gravel Sand Hollow cylindrical media Light expanded clay aggregates Polyurethane sponge Silicic gravel Bio-ceramsite

4.2 0.27 7 0.6 0.45 0.75 0.187 4.71 4.1 2.5 0.75 2.3

Feed nitrogen [mg N-NO 3 /L]

T [°C]

HRT [h]

Carbon source

17 15–50 16–18 21–27 16.4 4.7–64.5 40 28 10–340

10 12–20

1 1.42–5.11

25–27 20

1 97.6 1–3 31.2–312 2 0.16–1.25 374 3 2–5 2 1.25 24

CO2 H2CO3 CO2 NaHCO3 HCO 3

43 27 22 100 30–130

18–23 27 ± 2 27 ± 3 18–23 23 ± 1 26 ± 1 15–30

CO2 CO2 CO2 CO2 NaHCO3 CO2 NaHCO3

H2 supply

Direct bubbling External with recirculation Water electrolysis Membrane–fed H2 Anoxic iron corrosion Anoxic iron corrosion Water electrolysis Direct bubbling Anoxic iron corrosion Porous diffusers Water electrolysis Direct bubbling Water electrolysis Direct bubbling

Denitrification rate 3 [kg N-NO 3 /m d]

Reference

0.25 0.25–0.2

0.343 1.53–6.2 0.027 0.342 0.3387 2.419 2 0.178

[112] [113] [114] [39,40] [115] [49] [50] [42] [12] [44] [38] [45] [47] [41] [46]

0.25 0.004 0.002–0.446

Table 3 PBRs performing simultaneous removal of arsenic and nitrate. Carrier

Reactor volume [L]

Feed As [mgAs(III)/L]

Feed nitrogen [mg N-NO 3 /L]

T [°C]

HRT [h]

Carbon source

As removal [%]

Reference

Sand Activated alumina (AA)

0.42 0.42

0.567 0.5

35 35

30 ± 2 30 ± 2

24 12–24

HCO 3 HCO 3

97.2 40.5–98.3

[56] [57]

[64] [65] 21.6 0.9 NaHCO3 NaHCO3 7.5 8.0 ± 0.3 21–95 16–136 2.2 2.65

22 30 ± 2

1 1.5-5.53

[66] [17] 2.5–3.4 CaCO3 CaCO3/Organics 7.2–8.4 18.2 20–700

Synthetic groundwater Domestic sewage/leachate from a sanitary landfill Synthetic groundwater Synthetic groundwater

1 2.71

10–25 20

8–19 0.15–6.12

[62] [63] 0.13 0.6–0.7 CO2 NaHCO3 4.5 0.88–6.7 6–7 25 22–25 0.72 0.8

Sand (d = 0.2–0.3 mm) Polyacrylamide-alginate copolymerspherical beads (d = 3–5 mm) Aquifer sediments Sulfur granules (d = 2–3.35 mm) Polyvinylalcohol (PVA) Sulfur-impregnated GAC Drinking water Drinking water

30 30

Carbon source

in the range 0.9–2.2 resulted in a slight influence on denitrification efficiency which was stable in the range 94–96%, whereas a 6.56% drop was observed when the C/N ratio was reduced to 0.3, which was insufficient for denitrifiers to grow. Optimal pH and temperature ranged from 7 to 8 and from 25 to 35 °C, respectively. Hydrogen-to-water flow rate ratio (Qg/Qw), depending on both hydrogen flow rate and HRT, was found to affect denitrification 3 [47]. Denitrification rates up to 2.42 kg N-NO 3 /m d and nitrate removal efficiency higher than 90% were achieved increasing Qg/Qw over 4.4. 2.3.4. Fe(0)-supported hydrogenotrophic denitrification Hydrogenotrophic denitrification assisted by anoxic Fe(0)-corrosion was first studied in flow-through columns packed with steel wool by Till et al. [48]. A low nitrate removal efficiency 3 of 61% at loading rate of 28 g N-NO 3 /m d and HRT of 2.33 d was achieved. In order to enhance nitrate removal, Westerhoff and James [49] and Biswas and Bose [50] investigated the effect of several parameters on Fe(0)-mediated denitrification in reactor columns packed with different concentrations of elemental iron and sand. 2.3.4.1. Nitrate loading rate. Increase in nitrate loading rate enhances nitrate removal rate and reduces nitrate removal efficiency. The gradual increase of nitrate loading rate from 0.04 to 3 1.55 kg N-NO 3 /m d increased nitrate removal rate from 0.03 to  3 0.45 kg N-NO3 /m d at feed nitrate and EBCT in the ranges 4.7– 64.5 mg N-NO 3 /L and 1–3 h, respectively [49]. In another study, the increase of the initial EBCT (1.3 d) enhanced nitrate removal efficiency in three columns packed with sand and different amounts of steel wool, resulting in complete nitrate removal at EBCT 13 d [50]. 2.3.4.2. pH and dissolved oxygen. pH and dissolved oxygen play an important role in nitrate removal. Low pH stimulates abiotic nitrate reduction by increasing proton availability and results in ammonia production. In a study conducted by Westerhoff and James [49], decrease in initial pH from 8.5 to 7.1 enhanced nitrate removal up to 95%. Unexpectedly, increase in dissolved oxygen (DO) up to 8 mg/L also enhanced nitrate removal, probably due to the formation of unstable surface iron precipitates (e.g. green rust, magnetite) having a role in the abiotic reduction of nitrate/nitrite to ammonia [51–55]. 2.3.4.3. Production of ammonium. Nitrate removal in Fe(0)-packed columns results in high ammonium production. In the three studies mentioned above, 50–70% of the influent nitrate was converted to ammonium. Although a high Fe(0) surface area was shown to enhance nitrate removal [49], low Fe(0) concentrations and specific surface areas as well as high HRTs are reported to limit abiotic iron corrosion and favor denitrifying activity [48,50]. In order to lower ammonium production, Sunger and Bose [44] fed nitrate post hydrogen generation in sand-packed columns, avoiding the contact with iron powder. Hydrogen was generated by the anoxic corrosion of electrolytic iron powder in a generation system where the reactor effluent was continuously recirculated. This system resulted in negligible production of ammonium and removal of nitrite and nitrate up to 95% at nitrogen loading rate of 28.9 mg N/m3 d and HRT of 15.63 d.

H2 S0

Fe2S S0

2.4. Arsenite oxidation coupled to nitrate reduction

H2 H2

Electron donor

Table 4 Chemolithotrophic denitrification in FBRs.

Feed nitrogen [mg N-NO 3 /L] Reactor volume [L] Carrier particles Influent type

Influent pH

T [°C]

HRT [h]

Reference

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657

Denitrification rate 3 [kg N-NO 3 /m d]

648

Simultaneous removal of arsenic and nitrate has also been carried out in PBR applications (Table 3). Two sand-packed PBRs were operated by Sun et al. [56] to combine denitrification with As(III) oxidation to As(V). Fe(II) was added as further electron donor and

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657

to enhance As(V) immobilization onto biogenic Fe(III) (hydr)oxides. As(III) and total arsenic removals of 99.7% and 97.2% were respectively achieved with a HRT of 1 d. In another study, total arsenic removal up to 98.3% was achieved in a column packed with activated alumina (AA) at circumneutral pH and HRT of 13 d [57]. 3. Fluidized bed reactor (FBR) A fluidized bed reactor (FBR) is an attached growth system in which the support is fluidized at high recirculation flow rates. FBRs can be operated either in up-flow or in down-flow mode, depending on the direction of the flow inside the reactor, and with several carrier materials [58]. Recently, nitrate removal in FBR has been mainly performed with classical denitrification for the treatment of low pH and heavy metal contaminated mining wastewater [59,60], whereas the use of fluidized bed reactors in chemolithotrophic denitrification applications has been exiguous (Table 4). 3.1. FBR vs PBR Unlike PBR, FBR is more similar to a continuously stirred tank reactor. Bed fluidization results in several advantages: (1) efficient contact between biomass and substrates; (2) high biomass concentrations; (3) small reactor volumes; (4) high treatment capacity; (5) short HRTs; (6) biofouling control; (7) no need of backwashing [61]. Moreover, bed fluidization enhances nitrate mass transfer through the biofilm, controlling biofilm thickness and stripping the accumulated gas [17]. The potential of recycling the produced pH-buffered water helps to maintain neutral condition in the reactor and enables the treatment of acidic/alkaline wastewaters depending on the electron donor used. 3.2. Hydrogenotrophic denitrification A cone-shaped FBR was used by Kurt et al. [62] to carry out hydrogenotrophic denitrification of drinking water. Hydrogen was delivered in an external tank in the recirculation line and sand was used as biomass carrier. Nitrate and nitrite were completely removed at HRT of 4.5 h and influent nitrate concentration of 49 mg N-NO 3 /L. The gradual decrease of HRT to 3.5 h immediately increased nitrite concentration. In more recent studies, the use of polymeric materials as biomass carriers enhanced nitrate removal. Chang et al. [63] immobilized the bacterium Alcaligenes eutrophus on a polyacrylamide and alginate copolymer achieving a maximum 3 nitrate removal rate of 0.6–0.7 kg N-NO 3 /m d at HRT of 0.88 h and influent nitrate concentration of 22 mg N-NO 3 /L. Komori and Sakakibara [64] used polyvinylalcohol (PVA) porous cubes as biomass carrier. H2 was produced by water electrolysis using a soli

649

d–polymer–electrolyte membrane electrode (SPEME). The objective was to increase production and solubility of hydrogen gas and enhance nitrate removal. High denitrification rates up to 3 2.16 kg N-NO 3 /m d were achieved at HRT of 1 h and electric currents ranging from 0.4 to 4.0 A. Increase in electric current enhanced nitrate removal rate. A double stoichiometric amount of electric current was necessary to complete denitrification. 3.3. Sulfur-driven denitrification FBR performance in maintaining sulfur-driven denitrification was observed to be higher than that achieved in PBR. Denitrification efficiencies higher than 90% were achieved in sulfur 3 FBRs at nitrate loading rates up to 2.68 kg N-NO 3 /m d, two-five times higher than those sustained by sulfur PBRs [17]. Moreover, N2O content in the produced gas was significantly lower in FBR compared to PBR. In another study, Mohammadi et al. [65] maintained nitrate removal efficiencies higher than 90% up to a nitrate 3 loading rate of 0.9 kg N-NO 3 /m d, whereas a decrease in HRT from 2.4 to 1.5 h caused a breakthrough in nitrate and nitrite concentrations. The breakage of sulfur particles due to the high fluidization rates is a serious issue in FBRs operating sulfur-driven denitrification, since it limits biofilm formation. Composite sulfur-based carriers (e.g. granular activated carbon (GAC) impregnated with elemental sulfur) may represent a solution to this problem [65]. 3.4. Denitrification coupled to pyrite oxidation Hartog and Griffioen [66] investigated the denitrification potential of pyrite-bearing anaerobic aquifer sediments amended with nitrate in FBRs. Nitrate removal was driven by sedimentary organic matter (SOM), whereas denitrification coupled to pyrite oxidation was limited by pyrite low solubility at circumneutral to slightly alkaline pH values and/or by iron oxyhydroxides precipitation on sediment surface. 3.5. Costs FBR operation results in higher operating costs compared to PBRs. Bed fluidization increases power consumption and requires additional maintenance. However, FBRs can sustain nitrate loadings five-fold higher than PBRs, resulting in lower reactor volumes and capital costs. Literature lacks a comparative economic analysis between FBRs and PBRs performing chemolithotrophic denitrification. Costs related to other parameters (e.g. carrier material, biofouling, pumping system) influence the total cost of the system and have to be taken into account. Additional research is recommended.

Fig. 1. Hollow fibers operated in (a) flow-through (b) dead end mode. According to Martin and Nerenberg [69], N2 back-diffuses in the first part of the dead-end fiber, whereas at the distal end it concentrates and diffuses back into the biofilm. No back-diffusion is observed in flow-through membranes.

650

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657

[119] [82] [120] [121] [122] [123] [124] NaHCO3/CO2 CO2 CO2

0.12–0.22 0.10–0.38 0.41 0.04–0.06 0.88–23.5 7.2 0.02 NaHCO3 NaHCO3/CO2 NaHCO3

6–10 2–9 0.63 7–18 1.5–6.7 0.06 15.4

0.071 0.12

[81] [84] [89] [86] [118] [87] 0.83–2.48 0.16–0.31

NaHCO3 CO2 CO2 HCO 3 6 7.8–14.6 4–5 194.4 24.5 4

50 50 10 30 40–50 13.4 10 0.39 atm 0.39 atm 0.17 atm 10 mL/min 1.7–2.4 atm 0.39 atm

18 mL/min 0.24–1.36 atm 0.04 atm 0.17 atm 5–10 mL/min

50–150 100 15–25 24 10–30 17–20

0.5 9–12

NaHCO3

1.1 0.02–0.19

[116] [83] [80] [85] [78] [93] [117] 0.23–0.37 0.06–0.21 0.77 0.23–0.51 NaHCO3 CO2 CO2 CO2 HCO 3 0.7 8.3 4.1–30 4.1

10–12.5 120 65–200 5–15 20–40 22 12–72 0.31–0.42 atm 20 mL/min 0.28 0.2–0.45 atm

2.2 3

0.075 0.35 3

7

8143 4200 124 1140 2800 36.4  104

H2 is an ideal electron donor for biological reduction processes in potable water, since it is clean, non-toxic, inorganic and residual-free. In recent years, MBfRs have been extensively used for the reduction of inorganic (e.g. perchlorate, chlorate, chlorite, bromate, chromate, selenate, selenite, arsenate) and organic (e.g. DCM, DBCP, TCE, p-CNB, 2-CP) contaminants from both drinking water and wastewater using nitrate, sulfate and/or oxygen as primary electron acceptors [70–77]. In order to produce potable-grade water, a post-filtration phase is often necessary to remove escaping biomass from the effluent. The use of additional membranes for post-filtration results in superior effluent water quality [78].

Perfluoropolymer coating Polyester Hydrophobic polyvinyl chloride

Hollow fiber Hollow fiber Hollow fiber Hollow fiber Tubular membrane Hollow fiber Hollow fibers

Polyethylene Polyethylene Polyvinyl chloride

4.2. Hollow fibers

6.5 1.25 0.045 1.6 0.07 14 0.92

Polyethylene/polyurethane 750 Silicon 589 Polypropylene hollow fibers potted in polysulfone fittings 3700 Polyethylene/polyurethane 750 Silicon Polyethylene/polyurethane Two hollow fiber membranes 1st: CelgardÒ X30–240 55.6 2nd: ZeeWeedÒ-1 Hollow fiber Polysulfone 1300 Multi-layer tube Dimethylsiloxane + ferro-nickel fibrous slag 163 Fiber Membrane fiber Silicone-coated, reinforced fiberglass 120 Hollow fiber Platinum cured silastic 582 Hollow fiber Polypropylene

Material Type

Hollow fiber Tube Hollow fiber Hollow fiber Tube 0.42 1.5 1.2 0.42 0.02

Reactor volume [L] Membrane

Table 5 Membrane biofilm reactors performing hydrogenotrophic denitrification.

In membrane biofilm reactors (MBfRs) gas-permeable membranes are used as both gas diffusers and biofilm carriers (Fig. 1). Pressurized gases flow through the membrane lumen and diffuse through the internal wall to a biofilm formed on the membrane external surface. Unlike membranes used in S0-based membrane bioreactors (MBRs) [67,68], no filtering is performed. Since the gas permeates the membrane in the opposite direction than water dissolved compounds, counter-gradients are established and gas use efficiency improved. The membrane specific surface area (L2 L3), defined as the ratio of membrane area to reactor volume, significantly affects reactor performances. Increase in the specific surface area of the membrane typically results in higher denitrification rates and efficiencies [69]. This emerging technology has several benefits: high gas utilization efficiency, low energy consumption and small reactor volumes [69]. 4.1. Potable water treatment

2

Surface area (cm )

H2 pressure/flow Feed nitrogen HRT [h] [mg N–NO 3 /L]

Carbon source Denitrification rate Reference 3 [kg N–NO 3 /m d]

4. Membrane biofilm reactor (MBfR)

Membranes are a key factor in determining reactor performances. Hollow fiber membranes have been widely used in MBfR performing hydrogenotrophic denitrification because of low space requirements and high performances. Outer diameter and wall thickness typically vary from 50 to 300 lm and from 5 to 60 lm, respectively. Hollow fibers can be operated in dead-end or in flow-through mode, depending on whether the distal end of the membrane is closed or open, respectively. Dead-end membrane allows 100% H2 utilization, but their use may result in the back-diffusion of nitrogen gas from external biofilm to membrane lumen (Fig. 1), which may significantly dilute hydrogen concentration and affect reactor performance. Flow-through membranes avoid N2 back-diffusion, but their use requires high consumption of gas and energy [69]. Although membrane cost is steadily decreasing, costs related to membrane replacement and maintenance and to the production and supply of hydrogen still limit the widespread of MBfRs [79,69]. Direct H2-production inside the reactor may overcome the high costs related to membrane utilization (Section 5). Hollow-fiber-membrane biofilm reactors (HFMBfRs) were operated for hydrogenotrophic denitrification with membranes of different polymeric materials (Table 5). In particular, polysulfone allows easy membrane manufacturing and its use resulted in the highest denitrification rates [80,81]. Visvanathan et al. [82] performed a simultaneous organic and nitrate removal in two double stage HFMBfRs. Two different configurations were set-up: a denitrification–aeration system (DAS) and an aeration denitrification system (ASD). Higher denitrification 3 rates and efficiencies up to 0.38 kg N-NO 3 /m d were achieved in the DAS at HRTs ranging from 2 to 3 h and influent nitrate concentration of 50 mg N-NO 3 /L. Xia et al. [124] supplied both H2 and CO2

651

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657

through two bundles of hydrophobic polyvinyl chloride hollow fibers (pore size 0.01 lm), separately. CO2 was used for pH control and as carbon source. CO2 permeation experiments were carried out in order to design number and length of membrane fibers and set the adequate flux and pressure of CO2 to keep a stable reactor pH and satisfy the carbon demand of hydrogenotrophic denitrifiers. Denitrification efficiencies higher than 95% and stable pH around 7.45 were maintained at feed nitrate 10 mg N-NO 3 /L and HRT 15.4 h with CO2 as sole carbon source.

Table 6 Main biological and electrochemical reactions in BERs [99,94]. Reaction type

e0 (V)

Cathodic

0.0 +0.401

H2 O þ 2e ! H2 þ 2OH 2H2 O ! O2 þ 4Hþ þ 4e

Anodic

+0.207 +1.23

  2NO 2 þ 5e þ 6H2 O ! N2 þ 12OH

Biological

Reaction 

O2 þ 2H2 O þ 4e ! 4OH H2 O þ 2e ! H2 þ 2OH



4.3. Silicone membranes Silicone tubes have also been broadly used to deliver H2 in MBfRs. Silicone high permeability has made these membranes a popular choice, but the outer diameter on the order of millimeters limits their specific surface area [69]. Ho et al. [83] wound a silicone membrane tube around pillars inside the reactor in order to supply a mixture of H2 and CO2 to a pure culture of A. eutrophus. Mansell and Schroeder [78] used a silicon tube to deliver hydrogen in a specific diffusion unit. Denitrification occurred by nitrate molecular diffusion through a GoreTexÒ membrane (mean pore size 0.02 lm), which separated the denitrifying culture from nitrate contaminated wastewater. Denitrification efficiencies 3 above 90% with nitrate removal rates up to 1.06 kg N-NO 3 /m d were achieved at HRT of 0.83 h, whereas shorter HRTs resulted in lower nitrate removal. Terada et al. [84] assembled a multi-layer matrix composed by a silicone membrane tube covered by ferro-nickel fibrous slag. The slag resulted effective for support, development and protection of the biofilm against shear stress. 3 Denitrification rates and efficiencies of 0.309 kg N-NO 3 /m d and 99%, respectively, were achieved when H2 pressure was increased to 50 kPa. 4.4. Hydrogen gas flow Hydrogen pressure rules denitrification efficiency. Lee and Rittmann [85] and Schnobrich et al. [86] achieved complete denitrification increasing hydrogen pressure to 0.56 atm and 1.16 atm, respectively. In a lab-scale double column packed with aquifer material and equipped with a non-porous, silicone-coated and spiral-shaped hollow fiber membrane, phosphorus addition up to 9.6 ± 1.3 mg P/L and increase in hydrogen pressure were required to complete denitrification. Nevertheless, excess in hydrogen supply may result in concomitant reduction processes.

H2, in fact, may also act as electron donor for sulfate reducing bacteria (SRB) at favorable oxidation–reduction potential (ORP) values (300 mV) [84].

4.5. Biofilm density Biofilm density is a more determinant factor than biofilm thickness in terms of reactor performance [87]. Increase in biofilm density enhances denitrifying activity, whereas biofilm thickness severely impacts nitrate mass transfer. Mixing and nitrogen gas sparging provides an adequate regime of shear forces inside the reactor and favors the increase of biofilm density over thickness.

4.6. Membrane biofouling Biofouling is the main problem affecting membrane performance. Excessive biofilm accumulation limits nitrate mass transfer and results in low removal efficiencies. On the other hand, scarce biofilm accumulation leads to low nitrate removal due to biomass limitation [69]. Effective biomass control is determinant for optimal management of MBfRs. Physical or chemical cleaning of membranes may be required to control biofouling [88], resulting in discontinuous denitrification and additional costs. Celmer et al. [89] supplied a limited flux of hydrogen (18 mL H2 per minute) in order to create starving conditions and inhibit excess production of extracellular polymeric substances (EPS). Hydrogen limitation resulted in efficient control of biomass growth and stable reactor performance. Thermophilic conditions [90] and addition of chemicals such as sodium chloride [91] and surfactants [92] were also effective in limiting membrane biofouling whilst more expensive.

Fig. 2. Configuration of a BER operated in continuous mode. Reactor components are numbered as follows: (1) influent canister; (2) influent pump; (3) DC power supply; (4) voltmeter; (5) ammeter; (6) anode; (7) cathodes; (8) effluent canister; (9) gas collector.

652

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657

0.013–0.08 0.11–0.25 NaHCO3/CO2 21 ± 2

30 25–40

25 ± 3

0.8 19.04 30–1200 6594

36 0.2 0.6 0.52

2.5 1–10 5 20 80–960 0–10 300 3–16 15 160 150

Cathode

Stainless Stainless Stainless Carbon Metal Stainless GAC Stainless steel Activated carbon fiber Palm shell GAC Stainless steel cylinder Carbon Amorphous carbon Amorphous carbon Carbon Platinum metal coated Amorphous carbon Titanium coated with platinum Carbon modified b-PbO2 Stainless steel mesh Carbon rod

Table 7 Performances and operating conditions of BERs.

5.2. Current intensity Applied current intensity results in a nearly linear correlation with nitrate removal efficiency [96] and denitrification rate [97]. Feleke et al. [96] applied electric current intensities ranging from 2 to 10 mA in three identical BERs equipped with an amorphous carbon anode and a stainless cathode on which denitrifying microorganisms were immobilized by polyurethane foam. The stoichiometric amount of electric current (5 mA) resulted in complete denitrification and highest denitrification rate 3 (0.06 kg N-NO at influent nitrate concentration of 3 /m d)  20 mg N-NO3 /L. In a similar study [98], the highest removal efficiency (98%) was achieved at current intensities between 20 and 25 mA, whereas lower current intensities resulted in a proportional decrease of nitrate removal. Increase in current intensity up to 100 mA resulted in hydrogen excess in the gas phase, methane production and inhibited denitrification.

251 251 251 42 (m2/m3) 1096 251 750 321 500

Anode Cathode Anode

Carbon electrodes are the most used because of their high conductivity, good mechanical strength and affinity to biofilm formation (Table 7). Nevertheless, no influence of electrode material on denitrification efficiency has been reported in literature so far.

Electrode surface area (cm2)

5.1. Electrode material

Electrode material

Electric current (mA)

0.205 0.205 0.205

Reactor volume [L]

In biofilm-electrode reactor (BER) electrodes are used to produce hydrogen gas by water electrolysis (Fig. 2). Once dissolved oxygen is consumed, H2 is produced on the cathode surface. Hydrogenotrophic denitrifiers use cathodic hydrogen as electron donor and form a biofilm on the cathode surface. The electrochemical oxidation of the carbonic anode generates CO2 prior of O2 and provides anoxic conditions. Moreover, the production of carbon dioxide provides a source of inorganic carbon for hydrogenotrophic bacteria and buffers the increase of alkalinity in the system. Hydrogenotrophic denitrification via bio-electrochemical H2 production was also coupled to sulfur-driven denitrification as a way to provide a buffer against the acidity produced by the process and enhance nitrite removal [94,95]. The main electrochemical and biological reactions taking place in a BER are listed in Table 6.

160

5. Biofilm electrode reactor (BER)

15 20–24 20 20 13.8–20.8 24 15 30 10–50 20 20.9–22

Feed nitrogen [mg N–NO 3 /L]

T [°C]

25 25 20–30

9 10–50 10 10–13 2–6 10 0.33 1.9–5 2.4–6 6–36 2.1–4.2

HRT [h]

CO2 CO2 CO2 NaHCO3 NaHCO3 CO2 CO2 CO2

Carbon source

0.038 0.01–0.045 0.06 0.035 0.12 0.0576 0.394 0.381

Denitrification rate 3 [kg N–NO 3 /m d]

Hydrogen supply significantly affects the process costs. Hydrogen low solubility in water may result in high reactor volumes, since long contact times are required, whereas its explosive nature is a safety issue. Measures to reduce the risk of explosions raise operating and management costs. Adham et al. [93] compared the cost of a MBfR plant with a capacity of 0.158 m3/s to the cost of existing ion exchange (IE) systems. The total cost of the plant included a hydrogen-fed MBfR, a post-aeration treatment, a multimedia filtration, construction and start-up related costs and annual operation and maintenance costs. The economic analysis resulted in a total cost of 0.29 US$/m3 that was shown to be economically competitive compared to the cost of IE systems with the same capacity (0.29–0.35 US$/m3). A comparative economic analysis on a full scale MBfR showed that the cost of membrane and electricity was the critical parameter determining the relative feasibility of conventional process over membrane based process [79]. However, the general downward trend in the market price of membranes and the steady increase in energy costs in recent years is boosting membrane competitiveness, favoring a further development of membrane-based technologies in the next years.

[125] [126] [96] [98] [127] [128] [99] [94] [129] [130] [95]

Reference

4.7. Costs

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657

653

Fig. 3. Decision tree for biofilm reactor selection.

5.3. Cathodic surface area BER operation results in lower denitrification rates compared to MBfR (Table 7) and longer HRTs are required to achieve complete nitrate removal. However, increase in cathodic surface area enhances nitrate removal and enables lower HRTs, resulting in lower reactor volumes. Wang and Qu [94] combined hydrogenand sulfur-driven denitrification in a BER with a cathodic surface area of 321 cm2 and an influent nitrate concentration of 30 mg N-NO 3 /L. Complete nitrate removal with a maximum deni3 trification rate of 0.381 kg N-NO 3 /m d was achieved at HRTs ranging from 1.9 to 5 h. A further effort to increase denitrification rate was carried out by Prosnansky et al. [99]. A multi-cathode BER was realized using an inert anode (platinum-coated titanium) and 5 porous cathodes made with GAC, resulting in an overall cathodic surface area of 750 cm2. A 0.2 lm microfiltration phase was realized after denitrification to separate excess biomass from treated 3 water. A maximum denitrification rate of 0.394 mg N-NO 3 /m d was achieved at HRT of 0.33 h and influent nitrate concentrations ranging from 15 to 40 mg N-NO 3 /L. 5.4. Maintenance The main drawback regarding BER operation is the gradual scale formation on the cathode surface, which suppresses hydrogen

production and results in a dramatic decrease in denitrification rate [100]. As for gas-permeable membranes, biofouling is a major drawback regarding BER maintenance. The increase of biofilm thickness on the cathode surface limits production and diffusion of hydrogen and results in lower reactor performances. Moreover, BER operation results in considerable loss of excess biomass in the effluent and additional treatment (e.g. microfiltration) may be required. 5.5. Costs Prosnansky et al. [99] estimated the cost of a system composed by a BER coupled to a post microfiltration phase (BER/MF). The total cost of the system included the costs of electrodes, MF modules, GAC and concrete, and the operating costs due to electric energy consumption and CO2 feeding. Increase in current density resulted in higher hydrogen production and denitrification rates and reduced the total cost of the system. Increase in current density also increased energy consumption and, thus, operating costs. A minimum cost of 0.67 US$/m3 was achieved at current density of 4.0 A/m2 and HRT of 1.4 h. This cost can be further reduced to 0.18 US$/m3, depending on electrode configuration and reactor scale. The total cost of the BER/MF system was estimated to be competitive with those of IE, reverse osmosis (RO) and conventional potable water production system.

654

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657

6. Guidelines for the selection of reactor technology and configuration In this section, potentials and flaws of the four reactor technologies reviewed in this paper are briefly summarized. A decision tool is proposed in order to help the reader to choose the most suitable reactor technology and configuration based on the specific case.

membrane biofouling without detrimental impacts on nitrate removal. The high capital and operating costs of membranes and hydrogen supply and reactor maintenance represent a considerable limit to the widespread application of MBfRs. Efforts in producing low cost hydrogen gas and membrane materials are recommended to promote full-scale MBfR operation. 6.4. Biofilm electrode reactor

6.1. Packed bed reactor PBRs are robust, easy to operate and cost-effective. Albeit natural carriers show high reliability and cost effectiveness, polymeric carriers result in a better biofilm growth. High denitrification performances are achieved when sulfur and hydrogen are used as elec3 tron donors at low nitrate loading rates (<0.4 kg N-NO 3 /m d), whereas higher loadings require a considerable increase in HRT/EBCT to complete denitrification. High concentrations of nitrate and total suspended solids (TSS) and the decrease of sulfur grain size with time may result in high biofilm thickness, N2 entrapment, clogging and channeling, which reduce denitrification performance. In SLAD process, the use of limestone is responsible for hardness increase and has to be limited. High sulfate production and existence of sulfide in the effluent may be limiting factors to SLAD application. Organic supplementation improves PBR performance, enhancing denitrification, reducing both alkalinity consumption and sulfate production and generating bacterial consortia able to tolerate high levels of contaminants. Low salinity, small sulfur grain size and temperature above 20 °C improve denitrification potential of PBRs. The use of PBRs performing chemolithotrophic denitrification is recommended for the treatment of waters containing low concentrations of nitrate and TSS (e.g. groundwater) and/or significant concentrations of toxic compounds (e.g. industrial wastewaters).

BERs are a cost-effective alternative to membranes, but less efficient. Current intensity rules denitrification resulting in a linear correlation with denitrification rate and efficiency. Low denitrifica3 tion rate (up to 0.394 kg N-NO 3 /m d), high HRT and biomass loss are the main flaws regarding BER operation. High cathodic surface areas enhance denitrification and provide a wider biofilm support, limiting the loss of excess biomass. However, detrimental effects may occur at current intensities higher than 25 mA, limiting the rate of hydrogenotrophic denitrification. BER operation results in high effluent quality if nitrate concentration in the influent is low and a solid/liquid separation treatment is provided downstream. 6.5. Reactor selection Depending on water characteristics and environmental parameters, the choice of the most appropriate bioreactor technology is of major importance for the optimization of chemolithotrophic denitrification. Fig. 3 proposes a decision tree to assist in the selection of the most appropriate treatment technology depending upon several parameters, including nitrogen loading rate, feed nitrate and pH, metal and TSS content, carbon and energy source, effluent quality, reactor cost and maintenance. Funding sources

6.2. Fluidized bed reactor FBRs result in higher denitrification performances compared to PBRs. FBRs sustain nitrate loadings up to five times higher than PBR with the same nitrate removal efficiency. A maximum denitrifica3 tion rate of 21.6 kg N-NO 3 /m d was achieved in FBRs performing 3 hydrogenotrophic denitrification, whereas 6.2 kg N-NO 3 /m d was the highest removal rate in a PBR with the same electron donor. The high resistance to inhibitors and buffering capacity due to recycling make FBRs very efficient for the treatment of acidic waters if denitrification produces alkalinity. Bed fluidization is effective in increasing nitrate mass transfer to the biofilm and reducing biofouling, but requires higher power consumption and reactor maintenance. The use of elemental sulfur as biomass carrier is not recommended in FBR applications, since high fluidization rates may cause the rapid breakage of the sulfur particles and limit biofilm formation. 6.3. Membrane biofilm reactor MBfRs show high performances and compact volumes. The use of gas-permeable membrane resulted in the fastest denitrification 3 (23.5 kg N-NO 3 /m d). Membranes allow almost complete hydrogen utilization, albeit high hydrogen pressures are often required to complete denitrification. Biofouling is the main issue affecting MBfR performance. Excess in biofilm growth and the formation of mineral and metal precipitates have a negative long-term impact on gas-permeable membranes. For this reason, the use of MBfRs is not recommended for the treatment of high-strength wastewaters and/or with high concentrations of alkaline and heavy metals. However, the use of a sustainable hydrogen flux limits

European Commission through the Erasmus Programme – grant agreement FPA no 2010-0009.

Mundus

Acknowledgments This study was carried out in the framework of the Erasmus Mundus Joint Doctorate programme ETeCoS3 (Environmental Technologies for Contaminated Solids, Soils and Sediments) funded by the Education, Audiovisual & Culture Executive Agency (EACEA) of the European Commission under the grant agreement FPA no 2010-0009. References [1] M. Soares, Biological denitrification of groundwater, Water Air Soil Pollut. 123 (2000) 183–193. [2] M. Rivett, S. Buss, P. Morgan, J. Smith, C. Bemment, Nitrate attenuation in groundwater: a review of biogeochemical controlling processes, Water Res. 42 (2008) 4215–4232. [3] European Environment Agency (EEA), Groundwater Quality and Quantity in Europe. Environmental Assessment Report No. 3, European Environment Agency, Copenhagen, 2000. [4] D. Koren, W. Gould, P. Bedard, Biological removal of ammonia and nitrate from simulated mine and mill effluents, Hydrometallurgy 56 (2000) 127–144. [5] L. Knobeloch, B. Salna, A. Hogan, J. Postle, H. Anderson, Blue babies and nitrate-contaminated well water, Environ. Health Perspect. 108 (2000) 675– 678. [6] D. Lampe, T. Zhang, Sulfur-based autotrophic denitrification for remediation of nitrate-contaminated water, in: R. Columbus (Ed.), In Situ and On-Site Bioremediation: Volume 3, Papers from the Fourth International In Situ and On-Site Bioremediation Symposium, Bettelle Press, New Orleans, 1994, pp. 423–428. [7] R. Smith, M. Ceazan, M. Brooks, Autotrophic, hydrogen-oxidizing, denitrifying bacteria in groundwater, potential agents for bioremediation of nitrate contamination, Appl. Environ. Microbiol. 60 (1994) 1949–1955.

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657 [8] K. Straub, M. Benz, B. Schink, F. Widdel, Anaerobic, nitrate-dependent microbial oxidation of ferrous iron, Appl. Environ. Microbiol. 62 (1996) 1458–1460. [9] M. Shao, T. Zhang, H. Fang, Sulfur-driven autotrophic denitrification: diversity, biochemistry, and engineering applications, Appl. Microbiol. Biotechnol. 88 (2010) 1027–1042. [10] J. Devlin, R. Eedy, B. Butler, The effects of electron donor and granular iron on nitrate transformation rates in sediments from a municipal water supply aquifer, J. Contam. Hydrol. 46 (2000) 81–97. [11] K. Karanasios, I. Vasiliadou, S. Pavlou, D. Vayenas, Hydrogenotrophic denitrification of potable water: a review, J. Hazard. Mater. 180 (2010) 1–3. [12] I. Vasiliadou, K. Karanasios, S. Pavlou, D. Vayenas, Experimental and modelling study of drinking water hydrogenotrophic denitrification in packed-bed reactors, J. Hazard. Mater. 165 (2009) 812–824. [13] B. Rezania, N. Cicek, J. Oleszkiewicz, Kinetics of hydrogen-dependent denitrification under varying pH and temperature conditions, Biotechnol. Bioeng. 92 (2005) 900–906. [14] S. Ghafari, M. Hasan, M. Aroua, Effect of carbon dioxide and bicarbonate as inorganic carbon sources on growth and adaptation of hydrogenotrophic denitrifying bacteria, J. Hazard. Mater. 162 (2009) 1507–1513. [15] S. Ghafari, M. Hasan, M. Aroua, A kinetic study of autohydrogenotrophic denitrification at the optimum pH and sodium bicarbonate dose, Bioresour. Technol. 101 (2010) 2236–2242. [16] J. Flere, T. Zhang, Nitrate removal with sulfur–limestone autotrophic denitrification processes, J. Environ. Eng. 8 (1999) 721–729. [17] H. Kim, I. Lee, J. Bae, Performance of a sulphur-utilizing fluidized bed reactor for post-denitrification, Process Biochem. 39 (2004) 1591–1597. [18] J. Kruithof, C. Van Bennekom, H. Dierx, W. Hijnen, J. Van Paassen, J. Schippers, Nitrate removal from ground water by sulphur/limestone filtration, Water Suppl. 6 (1988) 205–217. [19] B. Gayle, G. Boardman, J.B.R. Sherrard, Biological denitrification of water, J. Environ. Eng. 115 (1989) 930–943. [20] K. Hiscock, J. Lloyd, D. Lerner, Review of natural and artificial denitrification of groundwater, Water Res. 25 (1991) 1099–1111. [21] C. Driscoll, J. Bisogni, The use of sulfur and sulfide in packed bed reactors for autotrophic denitrification, J. Water Pollut. Control Fed. 50 (1978) 569–577. [22] L. Kuai, W. Verstraete, Autotrophic denitrification with elemental sulphur in small-scale wastewater treatment facilities, Environ. Technol. 20 (1999) 201– 209. [23] J. Gu, W. Qiu, A. Koenig, Y. Fan, E. Choi, Z. Yun, Removal of high NO 3 concentrations in saline water through autotrophic denitrification by the bacterium Thiobacillus denitrificans strain MP, Water Sci. Technol. 49 (2004) 105–112. [24] Z. Zhao, W. Qiu, A. Koenig, X. Fan, J. Gu, Nitrate removal from saline water using autotrophic denitrification by the bacterium Thiobacillus denitrificans MP-1, Environ. Technol. 25 (2004) 1201–1210. [25] W. Zhou, Y. Sun, B. Wu, Y. Zhang, M. Huang, T. Miyanaga, Z. Zhang, Autotrophic denitrification for nitrate and nitrite removal using sulfurlimestone, J. Environ. Sci. 23 (2011) 1761–1769. [26] T. Zhang, J. Shan, In situ septic tank effluent denitrification using a sulfur/ limestone process, Water Environ. Res. 71 (1999) 1283–1291. [27] J. van der Hoek, J. Kappelhof, W. Hijnen, Biological nitrate removal from ground water by sulphur/limestone denitrification, J. Chem. Technol. Biotechnol. 54 (1992) 197–200. [28] J. van der Hoek, W. Hijnen, C. van Bennekom, B. Mijnarends, Optimization of the sulphur–limestone filtration process for nitrate removal from groundwater, J. Water Supply Res. T. 41 (1992) 209–218. [29] T. Zhang, Development of Sulfur-limestone Autotrophic Denitrification Processes for Treatment of Nitrate-contaminated Groundwater in Small Communities: Final Report, Midwest Technology Assistance Center, Champaign, Ill, 2004. [30] J. Park, H. Shin, I. Lee, J. Bae, Denitrification of high NO 3 -N containing wastewater using elemental sulfur; nitrogen loading rate and N2O production, Environ. Technol. 23 (2002) 53–65. [31] M. Soares, Denitrification of groundwater with elemental sulfur, Water Res. 36 (2002) 1392–1395. [32] D. Lee, I. Lee, Y. Choi, J. Bae, Effects of external carbon source and empty bed contact time on simultaneous heterotrophic and sulfur-utilizing autotrophic denitrification, Process Biochem. 36 (2001) 1215–1224. [33] B. Aminzadeh, A. Torabian, A.A. Azimi, G.R. Nabi Bidhendi, N. Mehrdadi, Salt inhibition effects on simultaneous heterotrophic/autotrophic denitrification of high nitrate wastewater, Int. J. Environ. Res. Public Health 4 (2010) 255–262. [34] T. Yamashita, R. Yamamoto-Ikemoto, J. Zhu, Sulfate-reducing bacteria in a denitrification reactor packed with wood as a carbon source, Bioresour. Technol. 102 (2011) 2235–2241. [35] T. Osaka, K. Shirotani, S. Yoshie, S. Tsuneda, Effects of carbon source on denitrification efficiency and microbial community structure in a saline wastewater treatment process, Water Res. 42 (2008) 3709–3718. [36] E. Sahinkaya, A. Kilic, B. Calimlioglu, Y. Toker, Simultaneous bioreduction of nitrate and chromate using sulfur-based mixotrophic denitrification process, J. Hazard. Mater. 262 (2013) 234–239.

655

[37] E. Sahinkaya, A. Kilic, Heterotrophic and elemental-sulfur-based autotrophic denitrification processes for simultaneous nitrate and Cr(VI) reduction, Water Res. 50 (2014) 278–286. [38] C. Lu, P. Gu, P. He, G. Zhang, C. Song, Characteristics of hydrogenotrophic denitrification in a combined system of gas-permeable membrane and a biofilm reactor, J. Hazard. Mater. 168 (2009) 1581–1589. [39] S. Szekeres, I. Kiss, T. Bejerano, M. Soares, Hydrogen-dependent denitrification in a two-reactor bio-electrochemical system, Water Res. 35 (2001) 715–719. [40] S. Szekeres, I. Kiss, M. Kalman, M. Soares, Microbial population in a hydrogendependent denitrification reactor, Water Res. 36 (2002) 4088–4094. [41] K. Karanasios, M. Michailidis, I. Vasiliadou, S. Pavlou, D. Vayenas, Potable water hydrogenotrophic denitrification in packed-bed bioreactors coupled with a solar-electrolysis hydrogen production system, Desalin. Water Treat. 33 (2011) 86–96. [42] R. Smith, S. Buckwalter, D. Repert, D. Miller, Small-scale, hydrogen oxidizing denitrifying bioreactor for treatment of nitrate-contaminated drinking water, Water Res. 39 (2005) 2014–2023. [43] R. Grommen, M. Verhaege, W. Verstraete, Removal of nitrate in aquaria by means of electrochemically generated hydrogen gas as electron donor for biological denitrification, Aquacult. Eng. 34 (2006) 33–39. [44] N. Sunger, P. Bose, Autotrophic denitrification using hydrogen generated from metallic iron corrosion, Bioresour. Technol. 100 (2009) 4077–4082. [45] R. Vagheei, H. Ganjidoust, A. Azimi, B. Ayati, Nitrate removal from drinking water in a packed-bed bioreactor coupled by a methanol-based electrochemical gas generator, Environ. Prog. Sust. Energy 29 (2010) 278– 285. [46] D. Chen, K. Yang, H. Wang, B. Lv, Nitrate removal from groundwater by hydrogen-fed autotrophic denitrification in a bio-ceramsite reactor, Water Sci. Technol. 69 (2014) 2417–2422. [47] J. Lee, K. Lee, K. Park, S. Maeng, Hydrogenotrophic denitrification in a packed bed reactor: effects of hydrogen-to-water flow rate ratio, Bioresour. Technol. 101 (2010) 3940–3946. [48] B. Till, L. Weathers, P. Alvarez, Fe(0)-supported autotrophic denitrification, Environ. Sci. Technol. 32 (1998) 634–639. [49] P. Westerhoff, J. James, Nitrate removal in zero-valent iron packed columns, Water Res. 37 (2003) 1818–1830. [50] S. Biswas, P. Bose, Zero-valent iron-assisted autotrophic denitrification, J. Environ. Eng. 131 (2005) 1212–1220. [51] H. Hansen, O. Borggaard, J. Sorensen, Evaluation of free energy of formation of Fe(II)–Fe(III) hydroxide-sulphate (green rust) and its reduction of nitrate, Geochim. Cosmochim. Acta 58 (1994) 2599–2608. [52] H. Hansen, C. Koch, H. Nancke-Krogh, O. Borggaard, J. Sorensen, Abiotic nitrate reduction to ammonium: key role of green rust, Environ. Sci. Technol. 30 (1996) 2053–2056. [53] H. Hansen, C. Koch, Reduction of nitrate to ammonium by sulphate green rust: activation energy and reaction mechanism, Clay Miner. 33 (1998) 87–101. [54] H. Hansen, S. Guldberg, M. Erbs, C. Koch, Kinetics of nitrate reduction by green rusts – effects of interlayer anion and Fe(II):Fe(III) ratio, Appl. Clay Sci. 18 (2001) 81–91. [55] D. Cho, H. Song, F. Schwartz, B. Kim, B. Jeon, The role of magnetite nanoparticles in the reduction of nitrate in groundwater by zero-valent iron, Chemosphere 125 (2015) 41–49. [56] W. Sun, R. Sierra-Alvarez, L. Milner, R. Oremland, J.A. Field, Arsenite and ferrous iron oxidation linked to chemolithotrophic denitrification for the immobilization of arsenic in anoxic environments, Environ. Sci. Technol. 43 (2009) 6585–6591. [57] W. Sun, R. Sierra-Alvarez, J.A. Field, The role of denitrification on arsenite oxidation and arsenic mobility in an anoxic sediment column model with activated alumina, Biotechnol. Bioeng. 107 (2010) 786–794. [58] S. Papirio, D. Villa-Gomez, G. Esposito, F. Pirozzi, P. Lens, Acid mine drainage treatment in fluidized-bed bioreactors by sulfate-reducing bacteria: a critical review, Crit. Rev. Environ. Sci. Technol. 43 (2013) 2545–2580. [59] S. Papirio, A. Ylinen, G. Zou, M. Peltola, G. Esposito, J. Puhakka, Fluidized-bed denitrification for mine waters. Part I: Low pH and temperature operation, Biodegradation 25 (2014) 425–435. [60] G. Zou, S. Papirio, A. Ylinen, F. Di Capua, A. Lakaniemi, J. Puhakka, Fluidizedbed denitrification for mine waters. Part II: Effects of Ni and Co, Biodegradation 25 (2014) 417–423. [61] M. Green, M. Shnitzer, S. Tarre, B. Bogdan, G. Shelef, C. Sorden, Fluidized bed reactor operation for groundwater denitrification, Water Sci. Technol. 29 (1994) 509–515. [62] M. Kurt, I. Dunn, J. Bourne, Biological denitrification of drinking water using autotrophic organisms with H2 in a fluidized-bed biofilm reactor, Biotechnol. Bioeng. 29 (1987) 493–501. [63] C. Chang, S. Tseng, H. Huang, Hydrogenotrophic denitrification with immobilized Alcaligenes eutrophus for drinking water treatment, Bioresour. Technol. 69 (1999) 53–58. [64] M. Komori, Y. Sakakibara, High-rate hydrogenotrophic denitrification in a fluidized-bed biofilm reactor using solid–polymer–electrolyte membrane electrode (SPEME), Water Sci. Technol. 58 (2008) 1441–1446. [65] A. Mohammadi, H. Movahedian, M. Nikaeen, Drinking water denitrification with autotrophic denitrifying bacteria in a fluidized bed bioreactor (FBBR), Fresen. Environ. Bull. 20 (2011) 2427–2434.

656

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657

[66] N. Hartog, J. Griffioen, The response of aquifer sediments to nitrate exposure: biogeochemical controls on denitrification potential, Geophys. Res. Abstr. 5 (2003) 1. [67] K. Kimura, M. Nakamura, Y. Watanabe, Nitrate removal by a combination of elemental sulfur-based denitrification and membrane filtration, Water Res. 36 (2002) 1758–1766. [68] E. Sahinkaya, A. Yurtsever, Ö. Aktas, D. Ucar, Z. Wang, Sulfur-based autotrophic denitrification of drinking water using a membrane bioreactor, Chem. Eng. J. (2015) 180–186. [69] K. Martin, R. Nerenberg, The membrane biofilm reactor (MBfR) for water and wastewater treatment: principles, applications, and recent developments, Bioresour. Technol. 122 (2012) 83–94. [70] R. Nerenberg, B. Rittmann, Hydrogen-based, hollow-fiber membrane biofilm reactor for reduction of perchlorate and other oxidized contaminants, Water Sci Technol. (2004) 223–230. [71] J. Chung, X. Li, B. Rittmann, Bio-reduction of arsenate using a hydrogen-based membrane biofilm reactor, Chemosphere (2006) 24–34. [72] J. Chung, R. Nerenberg, B. Rittmann, Bio-reduction of soluble chromate using a hydrogen-based membrane biofilm reactor, Water Res. 40 (2006) 1634– 1642. [73] J. Chung, R. Nerenberg, B. Rittmann, Bioreduction of selenate using a hydrogen-based membrane biofilm reactor, Environ. Sci. Technol. 40 (2006) 1664–1671. [74] J. Chung, B. Rittmann, W. Wright, R. Bowman, Simultaneous bio-reduction of nitrate, perchlorate, selenate, chromate, arsenate, and dibromochloropropane using a hydrogen-based membrane biofilm reactor, Biodegradation 18 (2007) 199–209. [75] J. Chung, R. Krajmalnik-Brown, B. Rittmann, Bioreduction of trichloroethene using a hydrogen-based membrane biofilm reactor, Environ. Sci. Technol. 42 (2008) 477–483. [76] S. Xia, H. Li, Z. Zhang, Y. Zhang, X. Yang, R. Jia, K. Xie, X. Xu, Bioreduction of para-chloronitrobenzene in drinking water using a continuous stirred hydrogen-based hollow fiber membrane biofilm reactor, J. Hazard. Mater. 192 (2011) 593–598. [77] S. Xia, Z. Zhang, F. Zhong, J. Zhang, High efficiency removal of 2-chlorophenol from drinking water by a hydrogen-based polyvinyl chloride membrane biofilm reactor, J. Hazard. Mater. 186 (2011) 1367–1373. [78] B. Mansell, E. Schroeder, Hydrogenotrophic denitrification in a microporous membrane bioreactor, Water Res. 36 (2002) 4683–4690. [79] E. Casey, E. Syron, J. Shanahan, M. Semmens, Comparative Economic Analysis of Full Scale MABR Configurations, IWA Membranes 2008, IWA, Amherst, 2008. [80] S. Ergas, A. Reuss, Hydrogenotrophic denitrification of drinking water using a hollow fibre membrane bioreactor, J. Water Supply: Res. Tech. Aqua. 50 (2001) 161–171. [81] J. Shin, B. Sang, Y. Chung, Y. Choung, The removal of nitrogen using an autotrophic hybrid hollow-fiber membrane biofilm reactor, Desalination 183 (2005) 447–454. [82] C. Visvanathan, N. Hung, V. Jegatheesan, Hydrogenotrophic denitrification of synthetic aquaculture wastewater using membrane bioreactor, Process Biochem. 43 (2008) 673–682. [83] C. Ho, S. Tseng, Y. Chang, Autotrophic denitrification via a novel membraneattached biofilm reactor, Lett. Appl. Microbiol. 33 (2001) 201–205. [84] A. Terada, S. Kaku, S. Matsumoto, S. Tsuneda, Rapid autohydrogenotrophic denitrification by a membrane biofilm reactor equipped with a fibrous support around a gas-permeable membrane, Biochem. Eng. J. 31 (2006) 84– 91. [85] K. Lee, B.E. Rittmann, Applying a novel autohydrogenotrophic hollow-fiber membrane biofilm reactor for denitrification of drinking water, Water Res. 36 (2002) 2040–2052. [86] M. Schnobrich, B. Chaplin, M. Semmens, P. Novak, Stimulating hydrogenotrophic denitrification in simulated groundwater containing high dissolved oxygen and nitrate concentrations, Water Res. 41 (2007) 1869–1876. [87] D. Celmer, J. Oleszkiewicz, N. Cicek, Impact of shear force on the biofilm structure and performance of a membrane biofilm reactor for tertiary hydrogen-driven denitrification of municipal wastewater, Water Res. 42 (2008) 3057–3065. [88] P. Le-Clech, V. Chen, T. Fane, Fouling in membrane bioreactors used in wastewater treatment, J. Membr. Sci. 284 (2006) 17–53. [89] D. Celmer, J. Oleszkiewicz, N. Cicek, H. Husain, Hydrogen limitation – a method for controlling the performance of membrane biofilm reactor for autotrophic denitrification of wastewater, Water Sci. Technol. 54 (2006) 165–172. [90] B. Liao, S. Liss, A comparative study between thermophilic and mesophilic membrane aerated biofilm reactors, J. Environ. Eng. Sci. 6 (2007) 247–252. [91] L. Freitos dos Santos, A. Livingston, Membrane-attached biofilms for VOC wastewater treatment. II: Effect of biofilm thickness on performance, Biotechnol. Bioeng. 47 (1995) 90–95. [92] A. Splendiani, A. Livingston, C. Nicolella, Control of membrane-attached biofilms using surfactants, Biotechnol. Bioeng. 94 (2006) 15–23. [93] S. Adham, T. Gillogly, G. Lehman, B. Rittmann, R. Nerenberg, Membrane Biofilm Reactor Process for Nitrate and Perchlorate Removal, American Water Works Association Research Foundation, 2004. [94] H. Wang, J. Qu, Combined bioelectrochemical and sulfur autotrophic denitrification for drinking water treatment, Water Res. 37 (2003) 3767– 3775.

[95] D. Wan, H. Liu, J. Qu, P. Lei, S. Xiao, Y. Hou, Using the combined bioelectrochemical and sulfur autotrophic denitrification system for groundwater denitrification, Bioresour. Technol. 100 (2009) 142–148. [96] Z. Feleke, K. Araki, Y. Sakakibara, T. Watanabe, M. Kuroda, Selective reduction of nitrate to nitrogen gas in a biofilm-electrode reactor, Water Res. 32 (1998) 2728–2734. [97] Y. Sakakibara, M. Kuroda, Electric prompting and control of denitrification, Biotechnol. Bioeng. 42 (1993) 535–537. [98] S. Islam, M. Suidan, Electrolytic denitrification: long term performance and effect of current intensity, Water Res. 32 (1998) 528–536. [99] M. Prosnansky, Y. Sakakibara, M. Kuroda, High-rate denitrification and SS rejection by biofilm electrode reactor (BER) combined with microfiltration, Water Res. 36 (2002) 4801–4810. [100] I. Kiss, S. Szekeres, T. Bejerano, M. Soares, Hydrogen-dependent denitrification: preliminary assessment of two bio-electrochemical systems, Water Sci. Technol. 42 (2000) 373–379. [101] T. Zhang, J. Shan, In situ septic tank effluent denitrification using a sulfur– limestone process, Water Environ. Res. 71 (1999) 1283–1291. [102] A. Koenig, L. Liu, Kinetic model of autotrophic denitrification in sulfur packed-bed reactors, Water Res. 35 (2001) 1969–1978. [103] S. Oh, Y. Yoo, J. Young, I. Kim, Effect of organics on sulfur-utilizing autotrophic denitrification under mixotrophic conditions, J. Biotechnol. 92 (2001) 1–8. [104] I. Kim, S. Oh, M. Bum, J. Lee, S. Lee, Monitoring the denitrification of wastewater containing high concentrations of nitrate with methanol in a sulfur-packed reactor, Appl. Microbiol. Biotechnol. 59 (2002) 91–96. [105] A. Koenig, L. Liu, Use of limestone for pH control in autotrophic denitrification: continuous flow experiments in pilot-scale packed bed reactors, J. Biotechnol. 99 (2002) 161–171. [106] A. Bezbaruah, T. Zhang, Performance of a constructed wetland with a sulfur/ limestone denitrification section for wastewater nitrogen removal, Environ. Sci. Technol. 37 (2003) 1690–1697. [107] H. Moon, K. Ahn, S. Lee, Use of autotrophic sulfur-oxidizers to remove nitrate from bank filtrate in a permeable reactive barrier system, Environ. Pollut. 129 (2004) 499–507. [108] H. Zeng, T. Zhang, Evaluation of kinetic parameters of a sulfur–limestone autotrophic denitrification biofilm process, Water Res. 39 (2005) 4941– 4952. [109] T. Zhang, H. Zeng, Development of a response surface for prediction of nitrate removal in sulfur–limestone autotrophic denitrification fixed-bed reactors, J. Environ. Eng. 132 (2006) 1068–1072. [110] R. Sierra-Alvarez, R. Beristain-Cardoso, M. Salazar, J. Gómez, E. Razo-Flores, J.A. Field, Chemolithotrophic denitrification with elemental sulfur for groundwater treatment, Water Res. 41 (2007) 1253–1262. [111] H. Moon, Y. Shin do, K. Nam, J. Kim, A long-term performance test on an autotrophic denitrification column for application as a permeable reactive barrier, Chemosphere 73 (2008) 723–728. [112] H. Gros, G. Schnoor, P. Rutten, Biological denitrification process with hydrogen-oxidizing bacteria for drinking water treatment, Water Suppl. 6 (1998) 193–198. [113] D. Dries, J. Liessens, W. Verstraete, P. Stevens, P. de Vos, J. de Ley, Nitrate removal from drinking water by means of hydrogenotrophic denitrifiers in a polyurethane carrier reactor, Wat. Suppl. 6 (1998) 181–192. [114] J. Liessens, J. Vanbrabant, P. De Vos, K. Kersters, W. Verstraete, Mixed culture hydrogenotrophic nitrate reduction in drinking water, Microb. Ecol. 24 (1992) 271–290. [115] K. Haugen, M. Semmens, P. Novak, A novel in situ technology for the treatment of nitrate contaminated groundwater, Water Res. 36 (2002) 3497– 3506. [116] K. Lee, B. Rittmann, A novel hollow-fiber membrane biofilm reactor for autohydrogenotrophic denitrification of drinking water, Water Sci. Technol. 41 (2000) 219–226. [117] H. Mo, J. Oleszkiewicz, N. Cicek, B. Rezania, Incorporating membrane gas diffusion into a membrane bioreactor for hydrogenotrophic denitrification of groundwater, Water Sci. Technol. 51 (2005) 357–364. [118] D. Smith, T. Rector, K. Reid-Black, M. Hummerick, R. Strayer, M. Birmele, M. Roberts, J. Garland, Redox control bioreactor: a unique biological water processor, Biotechnol. Bioeng. 99 (2008) 830–845. [119] J. Shin, B. Sang, Y. Chung, Y. Choung, A novel CSTR-type of hollow fiber membrane biofilm reactor for consecutive nitrification and denitrification, Desalination 221 (2008) 526–533. [120] Y. Zhang, F. Zhong, S. Xia, X. Wang, J. Li, Autohydrogenotrophic denitrification of drinking water using a polyvinyl chloride hollow fiber membrane biofilm reactor, J. Hazard. Mater. 170 (2009) 203–209. [121] J. Hwang, N. Cicek, J. Oleszkiewicz, Inorganic precipitation during autotrophic denitrification under various operating conditions, Environ. Technol. 30 (2009) 1475–1485. [122] A. Sahu, T. Conneely, K. Nusslein, S. Ergas, Hydrogenotrophic denitrification and perchlorate reduction in ion exchange brines using membrane biofilm reactors, Biotechnol. Bioeng. 104 (2009) 483–491. [123] Y. Tang, M. Ziv-El, C. Zhou, J. Shin, C. Ahn, K. Meyer, D. Candelaria, D. Friese, R. Overstreet, R. Scott, B. Rittmann, Bioreduction of nitrate in groundwater using a pilot-scale hydrogen-based membrane biofilm reactor, Front. Environ. Sci. Eng. China 4 (2010) 280–285.

F. Di Capua et al. / Chemical Engineering Journal 280 (2015) 643–657 [124] S. Xia, C. Wang, X. Xu, Y. Tang, Z. Wang, Z. Gu, Y. Zhou, Bioreduction of nitrate in a hydrogen-based membrane biofilm reactor using CO2 for pH control and as carbon source, Chem. Eng. J. 276 (2015) 59–64. [125] Y. Sakakibara, K. Araki, T. Tanaka, T. Watanabe, M. Kuroda, Denitrification and neutralization with an electrochemical and biological reactor, Water Sci. Technol. 30 (1994) 151–155. [126] Y. Sakakibara, K. Araki, T. Watanabe, M. Kuroda, The denitrification and neutralization performance of an electrochemically activated biofilm reactor used to treat nitrate-contaminated groundwater, Water Res. 36 (1997) 61–68. [127] Y. Sakakibara, T. Nakayama, A novel multi-electrode system for electrolytic and biological water treatments: electric charge transfer and application to

657

electric charge transfer and application to denitrification, Water Res. 35 (2001) 768–778. [128] Z. Feleke, Y. Sakakibara, A bio-electrochemical reactor coupled with adsorber for the removal of nitrate and inhibitory pesticide, Water Res. 36 (2002) 3092–3102. [129] M. Zhou, W. Fu, H. Gu, L. Lei, Nitrate removal from groundwater by a novel three-dimensional electrode biofilm reactor, Electrochim. Acta 52 (2007) 6052–6059. [130] S. Ghafari, M. Hasan, M. Aroua, Nitrate remediation in a novel upflow bioelectrochemical reactor (UBER) using palm shell activated carbon as cathode material, Electrochim. Acta 54 (2009) 4164–4171.