A review on the state-of-the-art of physical/chemical and biological technologies for biogas upgrading

1 A review on the state-of-the-art of physical/chemical and 1 biological technologies for biogas upgrading 2 3 Raúl Muñoz1,2*, Leslie Meier2, Israel Diaz1, David Jeison2 4 5 1 Department of Chemical Engineering and Environmental Technology, University of 6 Valladolid, C/Dr. Mergelina s/n, Valladolid, Spain; Phone: +34983186424, Fax: 7 34983423013. 8 2 Department of Chemical Engineering, University of La Frontera, Francisco Salazar 01145 9 Temuco, Chile; Phone: 56452325472, Fax: 56452325453. 10 *Author for correspondence: mut[email protected]. Phone: +34983186424, Fax: 11 +34983423013. 12 13 Abstract 14 The lack of tax incentives for biomethane use requires the optimization of both biogas 15 production and upgrading in order to allow the full exploitation of this renewable energy 16 source. The large number of biomethane contaminants present in biogas (CO2, H2S, H2O, 17 N2, O2, methyl siloxanes, halocarbons) has resulted in complex sequences of upgrading 18 processes based on conventional physical/chemical technologies capable of providing CH4 19 purities of 88-98 % and H2S, halocarbons and methyl siloxane removals > 99 %. 20 Unfortunately, the high consumption of energy and chemicals limits nowadays the 21 Manuscript REVISED Click here to download Manuscript: renamed_ead14.docx Click here to view linked References 2 environmental and economic sustainability of conventional biogas upgrading technologies. 22 In this context, biotechnologies can offer a low cost and environmentally friendly 23 alternative to physical/chemical biogas upgrading. Thus, biotechnologies such as H2-based 24 chemoautrophic CO2 bioconversion to CH4, microalgae-based CO2 fixation, enzymatic CO2 25 dissolution, fermentative CO2 reduction and digestion with in-situ CO2 desorption have 26 consistently shown CO2 removals of 80-100 % and CH4 purities of 88-100 %, while 27 allowing the conversion of CO2 into valuable bio-products and even a simultaneous H2S 28 removal. Likewise, H2S removals >99 % are typically reported in aerobic and anoxic 29 biotrickling filters, algal-bacterial photobioreactors and digesters under microaerophilic 30 conditions. Even, methyl siloxanes and halocarbons are potentially subject to aerobic and 31 anaerobic biodegradation. However, despite these promising results, most biotechnologies 32 still require further optimization and scale-up in order to compete with their 33 physical/chemical counterparts. This review critically presents and discusses the state of the 34 art of biogas upgrading technologies with special emphasis on biotechnologies for CO2, 35 H2S, siloxane and halocarbon removal. 36 37 Keywords: biomethane, biotechnologies, carbon dioxide removal, hydrogen sulfide 38 removal, siloxane removal, trace biogas contaminants. 39 40 1. Introduction. 41 Biogas represents a renewable energy source based on its high CH4 content. This CH4-rich 42 gas is a byproduct from the anaerobic treatment of wastewaters, the organic fraction of 43 3 municipal solid wastes (OMSW), livestock residues or organic agroindustrial wastes (Rasi, 44 2009). The composition of biogas is intrinsically determined by the carbon oxidation- 45 reduction state of the organic matter present in the waste and the type of anaerobic 46 digestion process, which in turn depend on the origin of the residue digested (Jönsson et al, 47 2003). For instance, the biogas recovered from conventional landfills is a complex mixture 48 composed of CH4 (35-65%), CO2 (15-50%), N2 (5-40%), H2O (0-5%), O2 (0-5%), H2 (0- 49 3%), CO (0-3%), H2S (0-100 ppmv), NH3 (0-5 ppmv), halogenated hydrocarbons (20-200 50 ppmv Cl-/F-), volatile organic contaminants (0-4500 mg m-3) and siloxanes (0-50 mg Si m- 51 3) (Jaffrin et al, 2003; Persson et al, 2006; Ajhar et al, 2010; Bailón and Hinge, 2012). A 52 slightly simpler biogas is typically obtained from the anaerobic degradation of sewage 53 sludge, livestock manure or agroindustrial bio-wastes: CH4 (53-70%), CO2 (30-47%), N2 54 (0-3%), H2O (5-10%), O2 (0-1%), H2S (0-10.000 ppmv), NH3 (0-100 ppmv), hydrocarbons 55 (0-200 mg m-3) and siloxanes (0-41 mg m-3) (Persson et al, 2006; Soreanu et al, 2011; 56 Bailón and Hinge, 2012). Carbon dioxide and nitrogen constitute the major contaminants of 57 biogas (N2 in the particular case of landfills), decreasing its specific calorific value and 58 therefore its Wobbe index (Ryckebosch et al, 2011). Large concentrations of O2 in the 59 biogas can entail explosion hazards, while high levels of H2S in combination with 60 condensate H2O causes corrosion in compressors, pipelines, gas storage tanks and engines. 61 Similarly, NH3 and halogenated hydrocarbons generate corrosive products during 62 combustion, which can severely damage engines and downstream pipelines (Persson et al, 63 2006; Petersson and Wellinger, 2009). Finally, methyl siloxanes combustion generates 64 silicone oxide that deposits in biogas combustion engines and valves, causing their 65 abrasion, overheating and malfunctioning (Abatzoglou and Boivin, 2009). 66 67 4 Biogas is currently used as a fuel for on-site heat, steam and electricity generation in 68 industry, as a substrate in fuel cells, as a substitute of natural gas for domestic and industrial 69 use prior injection into natural gas grids and as a vehicle fuel (Rasi, 2009; Andriani et al, 70 2014; Thrän et al, 2014). In this context, biogas production in Europe accounted for 13.4 71 million tons of oil equivalent (≈10 % increase compared to 2012), which represented 52,3 72 TWh of electricity produced and net heat sales to heating district networks of 432 megatons 73 of oil equivalent (EurObserv’ER, 2014). In addition, the actual European network of 14.000 74 anaerobic digesters is expected to increase in order to supply up to 18-20 million m3 by 75 2030 (3 % of the European gas consumption) according to the latest European Biogas 76 Association’s estimations (European Biogas Association, 2013). 77 78 The final use of biogas determines its composition and the type of upgrading process 79 required. Thus, on-site biogas use in boilers for heat generation only requires H2S removal 80 below 1000 ppmv and water removal prior to combustion (Bailón and Hinge, 2012). The 81 use of biogas in internal combustion engines for combined heat and power generation 82 (CHP) requires the removal of water, and H2S, NH3, siloxanes and halocarbons levels 83 below 200-1000 ppmv, 32-50 mg m-3, 5-28 mg m-3 and 65-100 mg m-3, respectively, 84 depending on the manufacturer. Turbines and micro-turbines for CHP generation require 85 very low contents of siloxane (0.03-0.1 ppmv) and water (pressurized dew point -6.7 ºC 86 below biogas temperature), but are able to stand high concentrations of H2S (10000-70000 87 ppmv) and halocarbon (200-1500 ppmv Cl-/F-) (Soreanu et al, 2011; Bailón and Hinge, 88 2012). However, the most stringent quality requirements are encountered in biomethane for 89 injection into natural gas grids and as a vehicle fuel, which often demands CH4 90 5 concentrations > 80- 96 %, CO2 < 2-3%, O2 < 0.2-0.5 %, H2S < 5 mg m-3, NH3 < 3-20 mg 91 m-3 and siloxanes < 5-10 mg m-3 (Table 1). 92 93 With the biogas upgrading market and technologies rapidly evolving, a more frequent 94 evaluation of the state-of-the art technologies available is necessary (Bauer et al, 2013b). In 95 this context, most physical/chemical biogas upgrading technologies are still highly energy 96 or chemical intensive, which has triggered the rapid development of biogas upgrading 97 biotechnologies based on their superior economic/environmental sustainability. This paper 98 critically reviews and discusses the state-of-the-art technologies for the removal of CO2, 99 H2S, H2O and trace biogas contaminants such as siloxanes, halocarbons, O2 and N2, with a 100 special focus on the potential and limitations of biotechnologies based on the significant 101 technological breakthroughs occurred in this field in the past 10 years. 102 103 2. Removal of Carbon dioxide. 104 CO2 removal from biogas at industrial scale is nowadays performed by physical/chemical 105 technologies based on their high degree of maturity and commercial availability, while the 106 potential of biotechnologies has been assessed only at lab or pilot scale. However, while 107 most physical/chemical units discharge the separated CO2 to the atmosphere (prior off-gas 108 post treatment to avoid the release of CH4), biotechnologies allow for the bioconversion of 109 CO2 into valuable commercial products, at significantly lower energy costs. 110 111 2.1. Physical/chemical CO2 removal technologies. 112 6 Scrubbing with water, organic solvents or chemical solutions, membrane separation, 113 pressure swing adsorption and cryogenic CO2 separation dominate the biogas upgrading 114 market nowadays. These technologies are discussed below: 115 116 2.1.1. Water Scrubbing 117 CO2 removal via scrubbing with water as selective absorbent is a classical unit operation in 118 chemical engineering based on the higher aqueous solubility of CO2 compared to that of 119 CH4 (26 times higher at 25 ºC) (Sinnott, 2005). Water scrubbing is nowadays a mature 120 technology with accounts for approximately 41 % of the global biogas upgrading market, 121 being considered the upgrading method less sensitive to biogas impurities (Thrän et al, 122 2014). The availability of a low-cost water supply of sufficient quality often determines the 123 water scrubber configuration implemented. For instance, CO2 removal from biogas 124 produced in wastewater treatment plants (WWTPs) has been performed in single-pass 125 scrubbers using pressurized treated water (6-10 bar), which after absorption is sent back to 126 the main water treatment line (Tynell et al, 2007). However, most modern units in landfills 127 or OMSW treatment facilities are constructed based on a sequential pressurized CO2 128 absorption in water (tap water quality) coupled to a two-stage stripping, which allows for 129 water regeneration (Beggel et al, 2010; Bauer et al, 2013). CO2 absorption is often carried 130 out at 6-10 bar, although pressures in the range of 10-20 bar are also used (Ryckebosch et 131 al, 2011). The first flash unit is operated at 2-4 bars, resulting in the emission of a CO2 rich 132 biogas (80-90% CO2 and 10-20 % CH4) that is returned to the absorption unit (Bauer et al, 133 2013b) (Figure 1A). Water decompression to atmospheric pressure in the second stripping 134 unit, often assisted by air injection, results in the final regeneration of the absorbent that is 135 returned to the absorption unit (Kapdi et al, 2005; Patterson et al, 2011; Ryckebosch et al, 136 7 2011). The amount of water required (m3 h-1) depends on the water pressure and 137 temperature, and can be estimated as Qbiogas /(H P), where Qbiogas (kmol h-1) represents the 138 raw molar biogas flow rate, H (M atm-1) the Henry’s Law constant and P (atm) the total 139 pressure of operation. Surprisingly, it does not depend on the pH of water or on the CO2 140 concentration in the raw biogas. Typical water flow rates of 0.1-0.2 m3water Nm-3biogas are 141 reported in single-pass scrubbers depending on the operational pressure (Persson, 2003), 142 which are comparable to the 0.18-0.23 m3water Nm-3biogas in units designed with water 143 recycling (Bauer et al, 2013b). Higher operational pressures entail lower water flow rates, 144 but higher pumping and compression costs and a reduced lifetime of the upgrading plant. 145 Despite water recycling significantly reduces water consumption, 20-200 L h-1 are 146 continuously purged to avoid the accumulation of detrimental byproducts. 147 148 Countercurrent operation is preferred regardless of the scrubbing configuration. Both 149 absorption and desorption units are typically constructed with random packings such as Pall 150 or Raschig rings to support an efficient gas-liquid mass transfer (Ryckebosch et al, 2011; 151 Bauer et al, 2013). CH4 and CO2 concentrations in the upgraded biogas are normally > 96% 152 and < 2%, respectively. CH4 losses of 1-2 % and technical plant availabilities of 95-96 % 153 are typically reported in technical literature for commercial full-scale facilities (10-10.000 154 Nm3 h-1) (Beil, 2009; Rasi, 2009; Patterson et al, 2011; Bauer et al, 2013b) (Table 2). 155 Despite manufacturers guarantee 2 % methane losses with exhaust gas recirculation, losses 156 of 8-10 % have been measured under regular operation, as a result of the non-optimized 157 operation of the flash tank (Persson, 2003). Elemental sulfur accumulation, corrosion and 158 odour nuisance also rank among the most important operational problems in water 159 8 scrubbers derived from the simultaneous absorption of H2S in water. Thus, despite this 160 technology can cope with H2S concentrations of 300-2500 ppmv (depending on the 161 manufacturer), H2S removal is highly recommended prior to water scrubbing (Persson et al, 162 2006; Thrän et al, 2014). On the other hand, microbial growth (especially when using 163 treated water in WWTPs) and foam formation in the packed bed constitute additional 164 operational problems of this technology, which result in a limited gas-liquid mass transport 165 and require the use of antifoaming agents (although their cost is marginal) (Bauer et al, 166 2013b). 167 168 Investment costs in water scrubbers linearly decrease from 5500 to 2500 € (Nm3 h-1)-1 when 169 the design treatment capacity increases from 100 to 500 Nm3 h-1, and remained relatively 170 constant at 1800-2000 € (Nm3/h)-1 for plant capacities over 1000 Nm3 h-1. On the other 171 hand, the operating costs range from 0.11-0.15 € Nm-3 (200-300 m3 h-1), which can be 172 attributed to both energy consumption (decreasing from 0.3 kWh Nm-3 at 500 Nm3 h-1 to 173 0.2 kWh Nm-3 at 2000 Nm3 h-1) and annual maintenance costs (2-3 % of the investment 174 costs), since the costs of consumables are often negligible (Urban et al, 2009; Patterson et 175 al, 2011; Bauer et al, 2013b). In this context, the major energy demanding processes are 176 gas compression (0.10-0.15 kWh Nm-3 in 6-8 bar modern facilities), water compression 177 (0.05-0.1 kWh Nm-3) and water cooling (0.01-0.05 kWh m-3). The need for an off-gas 178 treatment unit such as incinerators, activated carbon filters or biofilters to abate the H2S and 179 CH4 stripped from the desorption tank entail additional costs not considered in the above 180 discussion. 181 182 2.1.2. Organic Solvent Scrubbing 183 9 This technology, fundamentally similar to water scrubbing, uses polyethylene glycol-based 184 absorbents (commercialized under trade names such as Selexol® or Genosorb®), which 185 exhibit a higher affinity for CO2 and H2S than water. For instance, Selexol®, a mixture of 186 polyethylene glycol dimethyl ethers, has a 5 times higher affinity for CO2 than water (Tock 187 et al, 2010). These solvents allow for a decrease in both the absorbent recycling rates and 188 plant sizing, with the subsequent decrease in investment and operating costs (Petersson and 189 Wellinger, 2009; Ryckebosch et al, 2011). Unlike water scrubbing, the use of organic 190 solvents requires a gas condition step to remove water and several heating stages to 191 promote an efficient desorption of CO2 at 40 ºC (Figure 1B). Both biogas and organic 192 solvent are cooled down to 20 ºC prior absorption (Bauer et al, 2013b). The anticorrosion 193 nature of the organic solvents does not require the use of stainless steel in the scrubber. 194 Despite the advantages of this mature technology, its share in the biogas upgrading market 195 is only 6% (Thrän et al, 2014). 196 197 A biomethane with CH4 contents of 96-98.5 % can be consistently achieved in optimized 198 full scale organic solvents scrubbers with a 96-98 % technical availability (Bauer et al, 199 2013b; Thrän et al, 2014). Similarly to water scrubbing, this technology results in CH4 200 losses lower than 2 % (Persson et al, 2007). When biogas contains high concentrations of 201 H2S, solvent regeneration is conducted with steam or inert gas in order to avoid a sulfur- 202 mediated solvent deterioration (Ryckebosch et al, 2011). However, a complete H2S 203 removal using activated carbon filters is often recommended prior to organic scrubbing. 204 205 The capital costs for implementation of organic scrubbers decrease from ≈ 4500 € (Nm3 h1- 206 )-1 for 250 Nm3 h-1 plants to 2000 € (Nm3 h-1) -1 for design capacities of 1000 Nm3 h-1. 207 16 plants with capacities over 1000 Nm3 h-1. The operating costs of this technology are mainly 351 determined by membrane replacement (5-10 years lifetime), biogas compression cost (0.2- 352 0.38 kWh Nm-3) and the cost associated to biogas pre-treatment (activated carbon 353 replacement plus energy for condensation) (Benjaminsson, 2006; Beil, 2009; Bauer et al, 354 2013b). Costs in the range of 0.13-0.22 € Nm-3 are typically reported in literature (Hullu et 355 al, 2008). Membrane-based upgrading exhibits slightly higher maintenance cost (3-4 % of 356 the initial investment costs) compared to their physical chemical counterparts (2-3 %). 357 358 2.1.6. Cryogenic separation 359 The different liquefaction/solidification temperatures of the biogas components allow for a 360 selective separation of H2O, H2S, CO2 and CH4 if the temperature of biogas is stepwise 361 decreased, which even allows for the generation of a liquefied biomethane (free of O2 and 362 N2) at temperatures between -162 and -182 ºC (Bauer et al, 2013b). Cryogenic biogas 363 upgrading can be conducted at constant pressure (10 bar) using a sequential temperature 364 decrease to -25 ºC (where water, H2S, siloxanes and halogens are removed in liquid phase), 365 to -55 ºC (where most CO2 is liquefied to facilitate its withdrawal from the upgrading unit 366 and further commercialization) and finally to -85 ºC as polishing step (where the remaining 367 CO2 solidifies) (Ryckebosch et al, 2011). Process operation at high pressure avoids the 368 sudden solidification of CO2 below -78 ºC, which prevents operational problems derived 369 from clogging of pipelines and heat exchanges (Bauer et al, 2013b). The most common 370 operational procedure involves a preliminary biogas drying followed by a multistage 371 compression (with intermediate cooling) up to 80 bar (Patterson et al, 2011; Ryckebosch et 372 al, 2011). The pressurized biogas is stepwise cooled to -45 ºC and -55 ºC to promote the 373 liquefaction of most CO2, and finally expanded to 8-10 bar in a flash tank (-110 ºC) to 374 17 facilitate biomethane purification via CO2 solidification. Despite its synergies with the 375 process of biomethane liquefaction, this technology is still not reliably commercialized at 376 full scale and represents only 0.4 % of the upgrading market at a global level (Bauer et al, 377 2013; Bauer et al, 2013b; Thrän et al, 2014). 378 379 Cryogenic upgrading can provide a biomethane with a purity over 97 %, with methane 380 losses lower than 2 % (Beil, 2009; Andriani et al, 2014). The emerging nature of this 381 technology, with few operating plants in the United States, Sweden and The Netherlands, 382 does not allow yet an accurate determination of its technical availability (Petersson and 383 Wellinger, 2009; Bauer et al, 2013b). Water, H2S, siloxanes and halogens must be removed 384 prior to CO2 removal to avoid operational problems such as pipe or heat exchanger 385 clogging (Bauer et al, 2013b). On the other hand, no reliable data for investment and 386 operating costs of cryogenic upgrading plants is available, with the only estimation reported 387 by Hullu et al (2008) to 0.4 € Nm-3. There is also a large uncertainty on the estimations of 388 the energy needs for this process, with values ranging from 0.42 to 1 kWh/Nm-3 389 (Benjaminsson, 2006; Bauer et al, 2013b). 390 391 2.2 Biological CO2 removal technologies 392 CO2 mass transfer from the biogas to a microbial or enzymatic broth followed by a CO2 393 biological reduction constitutes the basis of most biotechnologies currently under research. 394 Of them, H2-assisted CO2 bioconversion, microalgae-based CO2 fixation, enzymatic CO2 395 dissolution, fermentative CO2 reduction and in-situ CO2 desorption are discussed below: 396 397 2.2.1. Chemoautotrophic biogas upgrading 398 18 The chemoautotrophic microbial conversion of CO2 to CH4 is based on the action of 399 hydrogenotrophic methanogens capable of using CO2 as their carbon source and electron 400 acceptor, and H2 as electron donor in the energy-yielding reaction described by equation 1 401 (Strevett et al, 1995): 402 403 4H2+CO2 CH4 + 2H2O (G0 = -131 KJ) (1) 404 405 The bioconversion of CO2 to CH4 using an external H2 injection has been used both in the 406 upgrading of biogas to biomethane and in the reduction of CO2 emissions from the 407 electronic industry using the on-site hydrogen produced from the electrochemical treatment 408 of its fluorhydric acid-containing wastewaters (Ju et al, 2008; Kim et al, 2013). Even 409 syngas from coal or biomass gasification processes containing CO, H2 and CO2 can be 410 upgraded to CH4 based on the ability of some methanogens to convert CO to CH4 and CO2 411 (4CO+2H2O CH4 + 3CO2). Microorganisms from the Archaeal domain such as 412 Methanobacterium sp., Methanococcus sp., Methanothermobacter sp., Methanosarcina sp., 413 Methanosaeta sp., Methanospirillum sp. and Methanoculleus sp. have been consistently 414 found in stand-alone bioreactors or anaerobic digesters upgrading CO2 to CH4 via H2 415 injection (Strevett et al, 1995; Luo et al, 2012b; Kim et al, 2013; Luo and Angelidaki, 416 2013; Wang et al, 2013). These autotrophic methanogens often exhibit an optimum pH 417 interval of 6.5-8 under both mesophilic and thermophilic conditions, and can even remove 418 part of the H2S present in the biogas by assimilation into biomass. However, while 419 thermophilic methanogens (55-88 ºC) exhibit higher bioconversion rates than their 420 mesophilic counterparts (30-40 ºC), the latter can achieve a more complete conversion of 421 19 CO2 (Strevett et al, 1995). In addition, thermophilic methanogens often present lower 422 growth yields (commonly defined as grams of biomass per mole of CH4 formed), which 423 ideally should be lower than 1 to promote the conversion of CO2 to CH4 rather than the 424 formation of biomass. In this context, chemical compounds such as cyanide or alkylhalides 425 have been shown to uncouple archaeal anabolism and catabolism, thus maximizing 426 biomethane production (Strevett et al, 1995). 427 428 Most CO2 bioconversion studies using H2 as electron donor have been carried out at lab 429 scale (0.05-100L) under mesophilic or thermophilic conditions in stirred tank, bubble 430 column, packed bed or membrane bioreactors with synthetic mixtures of CO2 and H2 431 supplied at stoichiometric ratios (1:4) (Table 3) (Kim et al, 2013). The extremely poor 432 aqueous solubility of H2 (dimensionless gas-water Henry’s law constant of 52) always 433 limited the gas-water H2 mass transfer rates and therefore the bioconversion of CO2 to CH4, 434 which is known to occur in the aqueous phase containing the methanogenic community. In 435 this regard, process operation under H2 mass transfer limitation is known to decrease the 436 efficiency of CH4 production at the expenses of an enhanced biomass formation (Strevett et 437 al, 1995). This resulted in the need to operate the process at extremely high gas residence 438 times (1-208 h) in order to achieve CH4 concentrations in the upgraded biogas over 90 %, 439 but entailed low volumetric CH4 productivities ranging from 0.65 to 5.3 L CH4/Lr d (Table 440 3). The few bioreactors reporting volumetric CH4 production capacities sufficiently high to 441 support a cost-efficient CO2 bioconversion (54-470 L CH4/Lr d) were operated during short 442 periods of time at low gas residence times (0.02-0.13 h) but yielded CH4 concentrations 443 (30-50%) not suitable for injection in natural gas grids or direct use as autogas. In this 444 context, the implementation of this bioconversion in high-mass-transfer gas phase 445 20 bioreactors such as two-phase partitioning or Taylor Flow bioreactors could support an 446 increase in the volumetric CH4 productivities of up to 1 order of magnitude, as reported 447 during the treatment of volatile organic contaminants (Kreutzer et al, 2005). 448 449 On the other hand, the studies evaluating the performance of the direct H2 injection in the 450 anaerobic digester are scarce (Luo et al, 2012b; Luo and Angelidaki, 2013). This process 451 configuration can avoid the use of an additional external bioreactor for biogas upgrading 452 (estimated to require 1/10 of the digester volume), and made the anaerobic digestion of 453 cattle manure and acidic whey more robust towards sudden increases in organic loading 454 rates, unexpectedly preventing the accumulation of Volatile Fatty Acids (VFA) likely due 455 to its associated pH increase (Luo and Angelidaki, 2013). Indeed, the addition of H2 into 456 the above described digester did not decrease the activity of the acetate kinase, a key 457 enzyme in the bioconversion of VFA to acetate, and increased the activity of the coenzyme 458 F420 (involved in hydrogenotrophic and acetoclastic methanogenesis). Likewise, the 459 injection of H2 into the digester also resulted in a significantly higher microbial activity, as 460 shown by the twice higher specific ATP content of the H2 supplemented biomass compared 461 to the mixed liquor of a similar digester deprived of H2 (Luo and Angelidaki, 2013). The 462 main limitation of this process configuration arises from the fact that anaerobic digesters 463 are not designed to maximize the gas-liquid mass transfer (excessive mixing might damage 464 the structure and functionality of anaerobic flocs), which might limit the performance of 465 this in-situ approach of CO2 bioconversion at large scale. Even small scale (0.6 L) stirred 466 tank digesters provided with fine bubble diffusers only achieved a biomethane composition 467 of 75%/6.6%/18.4% CH4/CO2/H2. In addition, the consumption of CO2 in the digester can 468 mediate inhibitory pH increases if the alkalinity of the organic fed is not properly 469 21 controlled, as reported by Luo et al (2012b) during the anaerobic digestion of cattle 470 manure. 471 472 The use of H2 to upgrade biogas entails a significant loss in energy efficiency and requires 473 the enforcement of severe safety operating procedures in anaerobic digestion plants as a 474 result of the high flammability of hydrogen. However, the use of CH4 as a fuel gas benefits 475 from both the exiting gas distribution infrastructure and well established combustion 476 technology, which represents the main reason to promote the production of CH4 over H2 477 (Wang et al, 2013). Water electrolysis from renewable energy sources (e.g. wind and solar 478 power) represents nowadays the only environmentally friendly (large-scale) method to 479 obtain H2 for bioconversion of CO2 to CH4. In this context, it must be highlighted that the 480 low density of H2 often requires high storage volumes, while the technology for H2 481 transportation and direct utilization is still under development. Therefore, H2 transformation 482 to biomethane, which can be injected into natural gas grids or employed as autogas, 483 constitutes a very attractive alternative to chemically store an energy that would be 484 otherwise lost. Finally, for chemoautotrophic biogas upgrading to be a sustainable and low 485 cost technology, H2 must be produced from water electrolysis using excess of electricity 486 (typically during the night) or as a byproduct in a nearby facility (Kim et al, 2013). 487 488 2.2.2. Photosynthetic biogas upgrading 489 Photosynthetic biogas upgrading relies on the ability of eukaryotic microalgae and 490 prokaryotic cyanobacteria (commonly referred to as microalgae) to bioconvert the CO2 491 present in the biogas into microalgae biomass using the electrons released during water 492 22 photolysis (López et al, 2013). This redox CO2 reduction process, namely oxygenic 493 photosynthesis, can be represented by the overall equation 2: 494 495 CO2 + H2O + photons + nutrients  O2 + CH1.63N0.14O0.43P0.006S0.005 + waste heat (2) 496 497 Such process requires the initial transport of the CO2 from the biogas to a microalgae- 498 containing aqueous phase. Likewise, approximately 1.8 g CO2 are required per gram of 499 microalgae produced. The low affinity for CO2 of the enzyme RubisCO in microalgae (KM 500 ≈ 1-8 mg CO2 L-1) does not entail however any technical limitation during photosynthetic 501 biogas upgrading as a result of both the relatively high levels of CO2 allowed in most 502 European biomethane legislations (3-6 %) and the presence of inorganic carbon- 503 concentrating mechanisms in most microalgae (Raven et al, 2008). Despite any microalgae 504 could eventually support photosynthetic biogas upgrading, Chlorella, Arthrospira and 505 Spirulina species have been preferentially used in the lab and pilot scale studies conducted 506 up-to-date, based on their tolerance to high CO2 and pH levels (Table 4). In this context, 507 while CO2 gas concentrations of 5 % were traditionally considered inhibitory for 508 microalgae growth, the intense research efforts conducted over the past 10 years in the field 509 of CO2-biomitigation from flue gases have resulted in the isolation of species tolerant to 510 CO2 concentrations of up to 60 % (Miyairi, 1995; Wang et al, 2008). The presence of H2S 511 in the biogas can inhibit microalgae growth, with H2S concentrations over 100 ppmv 512 exhibiting inhibitory effects on Chlorella sp. growth (Kao et al, 2012). However, the 513 synergistic occurrence of H2S oxidizing bacteria and the chemical oxidation of H2S in 514 biogas upgrading photobioreactors (operating under non-sterile conditions at high dissolved 515 oxygen concentrations) rapidly oxidizes this toxic sulfur compound into sulphate, which 516 23 eventually prevents any H2S-mediated microalgae inhibition in real applications (Bahr et al, 517 2014). On the other hand, methane does not exert any significant inhibitory effect on 518 microalgae growth in the concentration range of 20-80%, likely due to its low aqueous 519 solubility and reactivity (Kao et al, 2012). 520 521 Provided a sufficient CO2 mass transport from the biogas to the microalgal cultivation 522 broth, the rate of CO2 fixation, which itself determines the maximum biogas loading rate to 523 be applied to the upgrading unit, is governed by environmental factors such as light 524 availability, temperature, pH and dissolved O2 concentration in the cultivation medium. 525 Thus, the photosynthetic CO2 fixation rate linearly increases when increasing light intensity 526 up to a critical species-dependent saturation radiation (200-400 µE m-2 s-1), remaining 527 constant afterwards up to a critical photoinhibition value and deteriorating subsequently as 528 a result of the damage in the microalgal photosystem II at high light intensities (Tredici, 529 2009). At this point it should be highlighted that light availability does not depend 530 exclusively on the impinging light irradiation at the microalgae cultivation surface, but also 531 on the biomass density and photobioreactor configuration (Muñoz and Guieysse, 2006). 532 Most microalgae exhibit an optimum growth temperature in the range of 15 to 25ºC, 533 although some species such as Chlorella can grow optimally at 30-35ºC, which are 534 temperatures typically encountered in outdoor environments. On the other hand, while most 535 microalgae present an optimum activity at pH 7-8, process operation at pH of 9-10 (optimal 536 for cyanobacterial species such as Spirulina platensis) is desirable to maximize CO2 mass 537 transport from the biogas due to the acidic nature of this gas (Bahr et al, 2014; De Godos et 538 al, 2014). Finally, high dissolved oxygen concentrations in the cultivation broth can 539 mediate a competitive inhibition in the enzyme RubisCO (which also exhibits oxygenase 540 24 activity) and oxidative damage in the photosynthetic apparatus of microalgae due to the 541 formation of oxygen radicals. 542 543 The physical and biological mechanisms underlying CO2 removal from biogas in 544 photobioreactors are similar to those governing CO2 capture from exhaust flue gases (Yan 545 and Zheng, 2013; De Godos et al, 2014). Both processes have been implemented in open 546 and closed photobioreactors (Table 4), which are designed to maximize light distribution, 547 pH control, CO2 supply and O2 evacuation (Morweiser et al, 2010). Raceways, which 548 constitute the most common configuration of open photobioreactors, are characterized by a 549 simple construction and operation, and lower capital (2-20 € m-2) and energy requirements 550 (2-10 W m-3) than their closed counterparts (Tredici, 2009; Craggs et al, 2012). However, 551 raceways entail a poor light utilization efficiency (≈ 2 %), a high water footprint by 552 evaporation (≈ 6 L m-2 d-1) and large land requirements (López et al, 2013; De Godos et al, 553 2014). The higher photosynthetic efficiency of enclosed photobioreactors (4-6%), 554 supported by their higher illuminated surface-volume ratio and turbulence, results in 555 microalgae productivities of 0.4-1 g l-1 d-1, but at the expenses of significantly higher 556 energy consumptions (50-100 W m-3) and investment costs (500-3000 € m-2) (Acién et al, 557 2012). The number of studies evaluating the potential of microalgae-based biogas 558 upgrading in photobioreactors is scarce, most of them being conducted indoors under 559 artificial illumination and ambient temperatures (20-30 ºC) (Table 4). Bubble column and 560 horizontal tubular photobioreactors, and raceways constructed with additional biogas 561 scrubbing units rank among the preferred photobioreactor configurations evaluated. Most 562 experimental units were capable of removing CO2 with efficiencies higher than 80 %, 563 providing a biomethane with CH4 concentrations of ≈ 90% (Table 4). The gas residence 564 25 times in the absorption units ranged from 0.03-0.3 h in outdoors photobioreactors to 0.7-96 565 h in indoor set-ups, which suggests that photosynthetic activity rather than CO2 mass 566 transfer limits the biogas upgrading capacity of photobioreactors. In this context, high 567 biogas residence times in the absorption unit or a direct scrubbing in the photobioreactor 568 entails high O2 concentrations in the upgraded biomethane (5-25 %). This constitutes one of 569 the main limitations to be overcome in this novel biotechnology, due to its associated 570 explosion hazards and to the fact that most biomethane regulations require O2 levels below 571 0.5 % (Mandeno et al, 2005). In this context, the use of a 2-stage process based on biogas 572 scrubbing in an external column interconnected to the photobioreactor via a variable 573 microalgae broth recycling has been shown to support a satisfactory biogas upgrading with 574 O2 concentrations below 1 % (Bahr et al, 2014) (Figure 4). Nitrogen gas stripping from the 575 cultivation broth, which results in N2 concentration of 6-9% in the upgraded biomethane, 576 has been also identified as a technical limitation to be overcome. Thus, the removal of N2 577 from biomethane would be required in order to comply with biomethane regulations of 578 some European countries such as Sweden, Spain or Austria that require CH4 contents over 579 95 % (Persson et al, 2006; Huguen and Le Saux, 2010; Serejo et al, 2015). Finally, the CH4 580 losses derived from the mass transfer of CH4 from biogas to the recycling microalgal 581 cultivation broth and its subsequent oxidation by the methanotrophs present in this aqueous 582 medium were recently estimated to be <1% as a result of the low aqueous solubility of 583 methane (Serejo et al, 2015). 584 585 Unlike most physical/chemical CO2 absorption technologies, where CO2 is separated from 586 the biogas and discharged to the atmosphere, photosynthetic biogas upgrading allows the 587 valorization of this CO2 in the form of a valuable algal biomass. This microalgal biomass 588 32 rates (Persson et al, 2006; Abatzoglou and Boivin, 2009). Catalytic impregnation is 728 conducted by treating the carbon with a nitrogen containing reagent such as urea or 729 ammonia, while regular impregnation requires mixing of the carbon (before, during or after 730 activation) with NaHCO3, Na2CO3, NaOH, KOH, KI or KMnO4. H2S adsorption is 731 performed at high pressure (7-8 bar) and temperature (50-70ºC) with addition of air to the 732 biogas at 4-6 % in order to support the partial oxidation of H2S (equation 9) (Ryckebosch 733 et al, 2011): 734 735 2H2S + O2 2S + 2H2O (9) 736 737 Only KI or KMnO4 impregnation supports the partial oxidation of H2S in the absence of 738 O2. Carbon impregnated with these compounds is the preferred option for desulfurization 739 when biomethane is to be injected in natural gas grids or used as a vehicle fuel (Petersson 740 and Wellinger, 2009). Despite the elemental sulfur adsorbed can be desorbed at high 741 temperatures, in most cases the saturated activated carbon bed is replaced rather than 742 regenerated (Rutledge, 2005). Catalytic, impregnated and non-impregnated carbons exhibit 743 maximum adsorption capacities of 0.1, 0.15 and 0.2 g H2S g carbon-1, respectively. The 744 mechanisms underlying H2S oxidation are highly sensitive to the chemical properties of the 745 activated carbon surface, with acidic surfaces promoting H2S oxidation to SO2 and H2SO4, 746 and alkaline surfaces boosting the production of elemental sulfur (Bandosz, 2002). In 747 addition, the presence of water in the biogas severely deteriorates the performance of H2S 748 removal since this biogas component reacts with CO2, forming carbonates, and promotes 749 the formation of sulfurous acid, which can deactivate the active catalytic sites. Finally, 750 while the operating costs of activated carbon adsorption range from 0.0005 to 0.037 € m3s 751 33 (with an average impregnated activated carbon cost of ≈ 4 € kg-1), the capital cost of this 752 technology accounts for 3-120 € (m3/h)-1 (Abatzoglou and Boivin, 2009). 753 754 3.1.3 Membrane separation 755 This process is based on the selective permeability of certain membranes to H2S and the 756 corresponding retention of CH4 on the other side of the membrane. Gas-liquid membranes 757 using alkaline liquids on the other side of microporous hydrophobic membranes can 758 support H2S removal efficiencies of 98% during the desulfurization of biomethane 759 containing H2S at 2% (Ryckebosch et al, 2011). This technology is similar to that described 760 in section 2.1.5 for CO2 removal. H2S removal efficiencies of 58-94% have been recently 761 reported by Iovane et al (2014) using a Polymeric polyetheretherketone Hollow fiber 762 membrane (150 1210 mm) at biogas operating pressures of 25-41 bar. 763 764 3.1.4 H2S absorption 765 The absorption of H2S from biogas in conventional gas-liquid contactors (spray or packed 766 bed towers) can be carried out using either water or organic solvents in a process purely 767 based on physical absorption, or using aqueous chemical solutions with a conversion of 768 H2S to elemental sulfur or metal sulfides (Wellinger and Lindberg, 1999). While H2S 769 absorption in water can be implemented in both single pass and absorption-desorption 770 configurations, absorption in organic solvents such as Selexol (which entails lower liquid 771 flow rates than water scrubbing as a result of its higher affinity for H2S) requires solvent 772 regeneration based on their high cost (Ryckebosch et al, 2011). Absorption-desorption 773 configurations for H2S removal are similar to Figure 1C. Both water and organic solvent 774 34 scrubbing are suitable for the removal of low concentrations of H2S, and only competitive 775 when combined with the simultaneous removal of CO2 (Wellinger and Lindberg, 1999; 776 Kapdi et al, 2005). 777 778 The addition to the scrubbing process of chemical reagents such as NaOH, FeCl2, Fe(OH)3, 779 Fe3+/MgO, Fe3+/CuSO4 and Fe3+/EDTA can support a maximum H2S concentration 780 gradient between the biogas and the aqueous phase, thus reducing the liquid to biogas ratio 781 needed for an efficient H2S mass transfer (Abatzoglou and Boivin, 2009; Ryckebosch et al, 782 2011). The soluble salts sodium sulfide and sodium hydrogen sulfide are the end-products 783 during water scrubbing with NaOH solutions, hindering the regeneration of the NaOH 784 solution (Persson et al, 2006). However, this process is only applied for the upgrading of 785 high H2S concentrations or large biomethane flow rates based on the harsh operational 786 conditions imposed by the high concentrations of NaOH required (Petersson and Wellinger, 787 2009). In addition, the presence of CO2 in the biomethane significantly increases chemical 788 requirements. Likewise, Fe3+-based scrubbing was originally developed (and patented 789 under trademarks such as SulFerox® or LO-CAT®) for the desulfurization of sour gases 790 from oil and coal industry, and therefore only cost-effective for the upgrading of high 791 biogas flow rates with high H2S concentrations (>200 kgS d-1). This technology is highly 792 efficient, supporting final H2S biomethane concentrations of 1-10 ppmv, with an almost 793 complete regeneration of the oxidizing agent Fe3+ via aeration in a separate stage 794 (Abatzoglou and Boivin, 2009; Petersson and Wellinger, 2009). The chelated iron 795 Fe3+/EDTA (typically present at 0.2 mol L-1) is one of the most popular catalyst used for 796 H2S capture since the elemental S produced during the reduction of Fe3+ to Fe2+ according 797 to equation 10 (a first order reaction on iron and sulfur) can be easily recovered by 798 35 sedimentation prior to the regeneration of the Fe3+/EDTA solution by oxidation with air 799 according to equation 11 (Neumann and Lynn, 1984; Demmink and Beenackers, 1998): 800 801 2Fe3+ + S2-  2Fe2+ + S (10) 802 2Fe2+ + 0.5O2 + H2O  2Fe3+ + 2OH- (11) 803 804 This process can be operated at ambient pressure and temperature using gas residence times 805 (≈ 1 min) comparable to those used by their chemical adsorption counterparts (Horikawa et 806 al, 2004). Chelated iron based technologies can also remove 50-90 % of the mercaptans 807 present in the biomethane, without a significant reduction in CO2 concentration, at 808 operation costs of 0.24-0.3 € kgS-1 (Abatzoglou and Boivin, 2009). 809 810 On the other hand, the use of FeCl2 and Fe(OH)3 solutions result in the formation of the 811 insoluble salts FeS and Fe2S3 (Ryckebosch et al, 2011). Another process based on the 812 formation of intermediate insoluble metallic sulfides was originally developed by 813 Broekhuis et al (1992) for sour gas desulfurization using solutions of CuSO4 supplemented 814 with Fe3+ in a process operated at 60 ºC and gas residence times of 16-22 s. In this process, 815 H2S is transformed in a venture scrubber into CuS as described by equation 12, which is 816 further converted to elemental sulfur using Fe3+ as electron donor according to equation 13. 817 The electron donor is subsequently regenerated with air in a bubble column (equation 14): 818 819 Cu2+ + H2S +2SO42-  CuS + 2HSO4- (12) 820 CuS + 2Fe3+ Cu2+ + Fe2+ + S (13) 821 2Fe2+ + 0.5O2 + 2HSO4- 2Fe3+ + H2O + 2SO42- (14) 822 36 823 Finally, a full scale chemical scrubber using NaOH and H2O2 (as oxidizing agent) 824 supported H2S removal of 90-100 % at a plant availability of 95 % and operating cost of 825 0.03 € m-3 biogas (Miltner et al, 2012). 826 827 3.2 Biological H2S removal technologies 828 The ability of naturally occurring sulfur oxidizing bacteria (SOBs) has been used in 829 conventional biofiltration units, algal-bacterial photobioreactors and at the headspace of 830 anaerobic digesters to desulfurize biogas. 831 3.2.1 Biofiltration of H2S 832 The ability of lithoautotrophic bacteria to use H2S as electron donor and CO2 as carbon 833 source has supported the development of end-of-the pipe biotechnologies for biogas 834 upgrading (Montebello, 2013). Unfortunately, the removal of CO2 from biogas in this 835 particular technology is marginal compared to that of H2S (> 99% if properly designed) due 836 to the significantly lower H2S concentrations compared to CO2 and to the low biomass 837 yields of SOBs (YX/S ≈0.3 g VSS g S-1) (Mora et al, 2014). Oxidation of H2S using O2 as 838 the electron acceptor provides the energy required for lithotroph growth according to 839 equations 15 and 16. 840 841 H2S + 0.5O2 S +H2O (15) 842 H2S + 2O2 SO42- +2H+ (16) 843 844 37 The biological oxidation of H2S can be also carried out using NO3- (or NO2-) as electron 845 acceptors, which would avoid the contamination of biogas with O2 in the biofiltration unit, 846 via the denitrification reactions described by equations 17 and 18 (Soreanu et al, 2008): 847 848 3H2S + NO3-  3S + 0.5 N2 + 3H2O (17) 849 3H2S + 4NO3- 3SO42- + 2N2 + 6H+ (18) 850 851 Thus, low O2/S and NO3-/S ratios result in the preferential production of elemental sulfur. 852 Bacteria belonging to the genera Thiobacillus, Paracoccus, Thiomonas, Acidithiobacillus, 853 Halothiobacillus or Sulfurimonas, which are either strictly aerobes or facultative anaerobes 854 are capable of performing these H2S bioconversions. These microorganisms present 855 optimum growth temperatures in the range of 28-35 ºC. In addition, while most SOBs 856 exhibit an optimum activity at pH 6-8, extremophile species such as Acidithiobacillus 857 ferrooxidans or Acidithiobacillus thioxidans, present an optimum biocatalytic activity in the 858 low pH range (2-4) (Montebello, 2013). Strains of Acidithiobacillus thioxidans with 859 maximum sulfide oxidation rates of 21 g S g TSS-1 d-1 and tolerant to pH values as low as 860 0.2 and sulfate concentrations as high as 74 g L-1 have been reported in literature (Lee et al, 861 2006). 862 863 This end-of-the-pipe biotechnology has been mainly implemented in biotrickling filters 864 (BTF) due to their cost effectiveness, efficient gas-liquid mass transfer and easy control of 865 operational variables such as pH, temperature or nutrient supply (Estrada et al, 2012). 866 Desulfurization BTFs are packed bed columns (pall rings, HD-QPAC or polyurethane foam 867 as packing material supporting biofilm growth) operated with a recirculating aqueous phase 868 38 (at rates of 1-20 m h-1) containing the nutrients needed for SOB growth under pH controlled 869 conditions in the neutral (6-7.5) or acidic (2-3) range (Fortuny et al, 2011) (Table 5). This 870 bioreactor configuration has been successfully operated at laboratory and full scale using 871 both O2 (supplied via aeration) and NO3- as electron acceptors for the treatment of H2S 872 concentrations ranging from 500-10000 ppmv with efficiencies of 80-100 %, H2S being 873 totally depleted at concentrations below 2000 ppmv (Table 5). The high concentrations of 874 H2S present in biogas entail the operation of desulfurization BTFs at gas residence times 875 ranging from 2-16 min, which are 2 orders of magnitude larger than those typically 876 encountered in BTFs treating H2S malodorous emissions in WWTPs (Gabriel and 877 Deshusses, 2003). In this context, mass transfer limitations were recorded in desulfurization 878 BTFs operated below 120 s at a H2S concentration of 2000 ppmv (Fortuny et al, 2011). The 879 high H2S loading rate applied to these biological units, together with their satisfactory 880 desulfurization efficiency, result in ECs ranging from 40-220 gS m-3 h-1. Air is typically 881 used as O2 source based on its free availability, but results in the dilution or contamination 882 of biogas with N2 and O2 (the transfer of the latter to the liquid phase hindered by its high 883 Henry law constant). O2/H2S ratios of 2-41 have been implemented, the higher ratios 884 promoting a full oxidation of H2S to SO42- but a higher dilution of the biomethane, which 885 can limit its further applications. On the other hand, no significant differences on the 886 desulfurization performance were observed in anoxic BTFs using Ca(NO3)2, KNO3 and 887 NaNO3, although a concern exist on the potential accumulation of calcium salts (Fernández 888 et al, 2014). 889 890 H2S biofiltration exhibits a surprisingly high robustness (e.g recovery of steady state H2S 891 removal efficiencies within 4 h after a 5-days biogas supply shutdown) and lower operating 892 39 costs than physical/chemical technologies (Fortuny et al, 2011). Thus, operating costs of 893 0.013 and 0.016 € m-3 of biogas treated were estimated by Fernandez et al (2014) and 894 Tomàs et al (2009) for aerobic and anoxic biotrickling filtration, respectively, which are 895 significantly lower than the costs associated to FeCl3-mediated H2S chemical precipitation 896 or H2S chemical scrubbing (0.024 and 0.03 € m-3, respectively) (Tomàs et al, 2009; Miltner 897 et al, 2012). Packing media clogging, entailing higher operating costs derived from the 898 increase in pressure drop and the need for packing media cleaning or replacement, as a 899 result of elemental sulfur accumulation constitutes the main operational limitations of this 900 technology (Montebello et al, 2014). However, S accumulation can be minimized by either 901 the natural presence of mercaptans in biogas (as a result of the chemical reaction of 902 mercaptans with the accumulated S and the further biological oxidation of the DMDS 903 formed) or the implementation of operational strategies based on the oxygenation of the 904 packed bed in the absence of biogas supply (which has been shown to remove 80 % of the 905 accumulated S within a week) (Montebello et al, 2012; Montebello et al, 2014). 906 907 3.2.2 In-situ microaerobic H2S removal 908 Microaerobic H2S removal in the headspace of anaerobic digesters relies on the action of 909 SOBs able to grow lithoautotrophically on H2S while producing S0 under O2-limited 910 conditions according to equation 15 (Madigan et al, 2009). SOBs show diverse 911 morphological, physiological and ecological characteristics and employ primarily O2 as the 912 terminal electron acceptor, since many sulfur chemolithotrophs are aerobic (Tang et al, 913 2009). While in-situ microaerobic H2S removal has been traditionally used in anaerobic 914 digesters treating agricultural wastes based on the economic benefits of on-site biogas 915 exploitation (Schneider et al, 2002), recent research has extended its application to 916 40 anaerobic reactors treating industrial wastewaters (Rodríguez et al, 2012), WWTP sludge 917 or cow manure (Jenicek et al, 2008; Kobayashi et al, 2012). In this particular technology, 918 the headspace of anaerobic digesters acts as a H2S abatement unit where different 919 microaerophilic SOBs such as Acidithiobacillus sp., Arcobacter sp., Sulfuricuvum sp., 920 Sulfurimonas sp., Thiobacillus sp., Thiofaba sp. and Thiomonas sp. developed when a 921 limited amount of O2 is supplied (Díaz et al, 2011b; Kobayashi et al, 2012; Rodríguez et al, 922 2012). SOBs grow over the headspace walls and ceiling due to the lack of any specific 923 biomass support, thus creating superimposed laminas of S0 that act as a support material 924 (with a high specific surface area which facilitates both O2 transfer and further microbial 925 growth) (Díaz et al, 2011b; Kobayashi et al, 2012). The main advantage of in-situ H2S 926 removal is that additional end-of-pipe units for desulfurization are avoided. However, an 927 excessive S0 deposition in the digester’s headspace might impair the removal performance 928 over the time by reducing the residence time of biogas and, accordingly, the O2 transfer rate 929 to the microorganisms. This ultimately requires a periodical cleaning to maintain the H2S 930 removal efficiency. 931 932 Research studies on in-situ microaerobic H2S removal have been performed in Upflow 933 Anaerobic Sludge Blanket bioreactors, Expanded Granular Sludge Bed bioreactors and 934 fully mixed digesters under a wide range of biogas flow rates (7L d-1-250m3 h-1), H2S 935 concentrations (2500- 67000 ppmv) and operational conditions affecting O2 mass transfer 936 rate in the headspace (Table 6). The biogas residence time in the headspace was found to be 937 a key parameter determining the desulfurization efficiency. Hence, H2S removal 938 efficiencies over 97 % are typically encountered when operating at biogas residence times 939 over 5 h. Empirical observations also pointed out that higher O2 to H2S molar ratios are 940 41 required to maintain a H2S removal efficiency over 99%, when decreasing the biogas 941 residence time in the headspace. In this context, the O2 (or equivalent air) supply rate can 942 be adjusted to 0.3%-3% of the biogas production rate depending on the H2S concentration 943 and the aforementioned biogas residence time. However, a variable O2/air dosing is often 944 required in most digesters in order to minimize the residual O2 in the upgraded biogas as a 945 result of the variable biogas production rates. Hence, a residual O2 concentration of 1-1.8% 946 in the biogas can be reached by controlling the ORP in the anaerobic mixed liquor, while a 947 0.3-0.5% residual O2 concentrations were recorded when employing biogas production as 948 the control variable, despite both operational approaches supported H2S removal 949 efficiencies larger than 99% (Ramos and Fdz-Polanco, 2014). O2 can be supplied to the 950 liquid recirculation or directly to the headspace of the anaerobic digester. In this regard, 951 similar H2S removal efficiencies at equivalent O2 dosing rates were found since 952 microaerophilic SOBs seem to be favored under O2 limiting conditions (Díaz et al, 2011b; 953 Kobayashi et al, 2012; Ramos et al, 2014). In contrast, mixing conditions can be 954 manipulated to control the amount of O2 supplied and the removal of dissolved sulfide 955 (Figure 6). Thus, when anaerobic mixed liquor mixing provides a low contact between the 956 biogas and mixed liquor, i.e. by using liquid recirculation or low speed mechanical 957 agitation, H2S is removed from the biogas without altering the concentration of total 958 dissolved sulfide. On the other hand, when biogas recirculation is employed and the contact 959 between phases is larger, both H2S in the biogas and dissolved sulfide are oxidized (Díaz et 960 al, 2011b). Besides, a higher O2/H2S ratio was necessary to achieve satisfactory H2S 961 removals with biogas recirculation when compared to sludge recirculation, and the 962 concentration of more oxidized sulfur species such as S2O32- increased presumably as a 963 result of the higher O2 mass transfer rate (Díaz et al, 2011a). 964 48 6. Conclusions 1107 Physical/chemical technologies for biogas upgrading based on absorption, adsorption, 1108 chemical reaction, membrane separation or cryogenic separation are nowadays mature 1109 technologies capable of providing a biomethane suitable for injection into natural gas grids 1110 or use as autogas, with a limited room for technical and economic optimization (with the 1111 exception of membrane or cryogenic separation). However, their high energy and chemical 1112 requirements impose a severe limitation to the exploitation of the full potential of biogas as 1113 a renewable energy source. In this context, biotechnologies such as algal-bacterial 1114 photobioreactors can provide a simultaneous CO2 and H2S removal in a single process, 1115 while bioconverting CO2 into a valuable feedstock for the production of bioenergy or high 1116 added value products. The conversion of the electricity grid excess during the night into H2, 1117 and its use as electron donor in chemolitotroph-based bioreactors can bioconvert the CO2 1118 from biogas into CH4. Both technologies have been so far evaluated at lab and pilot scale, 1119 industrial scale testing and optimization being still necessary to show their full potential for 1120 biogas upgrading. Mass transfer limitations of CO2 and H2 have been identified as the main 1121 bottlenecks of algal-bacterial photobioreactor and chemolitotrophs-based bioreactors, 1122 respectively. Similarly, biotechnologies such as aerobic or anoxic biotrickling filtration 1123 and anaerobic digestion under microaerophilic conditions have been consistently shown to 1124 support H2S removal efficiencies > 99 % at significantly lower operating costs than in-situ 1125 chemical precipitation, adsorption or chemical scrubbing. These biotechnologies have 1126 undergone a rapid development over the past 20 years and are nowadays commercially 1127 available and implemented in full scale facilities. However, both biotechnologies don’t 1128 allow for a significant CO2 removal, contaminate the biomethane with O2 and N2 and still 1129 49 suffer from operational problems derive from elemental sulfur accumulation in the 1130 digester’s headspace or in the packed bed. Finally, the high catabolic potential of 1131 microorganisms allows for the biodegradation of both methyl siloxanes and halocarbons 1132 from biogas. Little research, and only restricted to lab scale feasibility tests, has been 1133 conducted in this particular field, with methyl siloxane mass transfer from the gas phase to 1134 the microorganisms being identified as the main process limitation. Based on their high 1135 biogas pollutant removal efficiencies and robustness, research on innovative biogas- 1136 microbial community mass transfer strategies and process scale-up constitute the road map 1137 to the development of cost-efficient and sustainable biotechnological process for an integral 1138 upgrading of biogas. 1139 1140 Acknowledgements 1141 The financial support of CONICYT-Chile (MEC Program Grant Nº: 80130013 and 1142 FONDECYT 1120488) is gratefully acknowledged. This work was also supported by the 1143 Regional Government of Castilla y León (Project VA024U14 and GR76) and the Spanish 1144 Ministry of Economy and Competitiveness (CTQ2012-34949 and RED NOVEDAR). 1145 1146 1147 1148 1149 50 References 1150 1151 Abatzoglou N, Boivin S (2009) A review of biogas purification processes. Biofuels, 1152 Bioprod, Bioref 3:42–71. doi:10.1002/bbb.117 1153 Accettola F, Guebitz G, Schoeftner R (2008) Siloxane removal from biogas by biofiltration: 1154 biodegradation studies. Clean Techn Environ Policy 10:211-218. doi:10.1007/s10098- 1155 007-0141-4 1156 Acién FG, Fernández JM, Magán JJ, Molina E (2012) Production cost of a real microalgae 1157 production plant and strategies to reduce it. Biotechnol Adv 30:1344–1353. 1158 doi:doi:10.1016/j.biotechadv.2012.02.005 1159 Ajhar M, Travesset M, Yüce S, Melin T (2010) Siloxane removal from landfill and digester 1160 gas - a technology overview. Bioresour Technol 101:2913-2923. 1161 doi:doi:10.1016/j.biortech.2009.12.018 1162 Ako O, Kitamura Y, Intabon K, Satake T (2008) Steady state characteristics of acclimated 1163 hydrogenotrophic methanogens on inorganic substrate in continuous chemostat 1164 reactors. Bioresour Technol 99:6305–6310. doi:10.1016/j.biortech.2007.12.016 1165 Alcántara C, García-Encina R, Muñoz R (2013) Evaluation of Mass and Energy Balances 1166 in the integrated microalgae growth-anaerobic digestion process. Chem Eng J 221:238- 1167 246. doi:10.1016/j.cej.2013.01.100 1168 51 Andriani D, Wresta A, Atmaja T, Saepudin A (2014) A Review on Optimization 1169 Production and Upgrading Biogas Through CO2 Removal Using Various Techniques. 1170 Appl Biochem Biotechnol 172:1909-1928. doi:10.1007/s12010-013-0652-x 1171 Bahr M, Díaz I, Dominguez A, González Sánchez A, Muñoz R (2014) Microalgal- 1172 biotechnology as a platform for an integral biogas upgrading and nutrient removal from 1173 anaerobic effluents. Environ Sci Technol 48:573-581. doi:10.1021/es403596m 1174 Bailón L, Hinge J (2012) Report: Biogas and bio-syngas upgrading. Danish Technological 1175 Institute. http://www.teknologisk.dk/_root/media/52679_Report- 1176 Biogas%20and%20syngas%20upgrading.pdf. Accessed 8 December 2014 1177 Bandosz TJ (2002) On the adsorption/oxidation of hydrogen sulfide on activated carbons at 1178 ambient temperatures. J Colloid Interf Sci 246:1–20. doi:10.1006/jcis.2001.7952 1179 Basu S, Khan A, Cano-Odena A, Liu C, Vankelecom I (2010) Membrane-based 1180 technologies for biogas separations. Chem Soc Rev 39:750-768. 1181 doi:10.1039/B817050A 1182 Bauer F, Hulteberg C, Persson T, Tamm D (2013b) Biogas upgrading – Review of 1183 commercial technologies. SGC Rapport 2013:270. SGC. 1184 http://vav.griffel.net/filer/C_SGC2013-270.pdf. Accessed 10 October 2014 1185 Bauer F, Persson T, Hulteberg C, Tamm D (2013) Biogas upgrading – technology 1186 overview, comparison and perspectives for the future. Biofuels, Bioprod, Bioref 7:499- 1187 511. doi:10.1002/bbb.1423 1188 52 Beggel F, Nowik IJ, Modigell M, Shalygin MG, Teplyakov VV, Zenkevitch VB (2010) A 1189 novel gas purification system for biologically produced gases. J Clean Prod 18:S43- 1190 S50. doi:10.1016/j.jclepro.2010.06.015 1191 Beil, M. Overview on biogas upgrading technologies. In: European Biomethane Fuel 1192 Conference, Goteborg, Sweden, 18 December 2009. 1193 Benjaminsson J (2006) NYA Renings - Och Uppgraderingstekniker för biogas: Rapport 1194 SGC 163. Svenskt Gastekniskt Center. 1195 http://www.sgc.se/ckfinder/userfiles/files/SGC163.pdf. Accessed 20 December 2014 1196 Berndt A (2006) Intelligent utilization of biogas - Upgrading and adding to the grid. 1197 CarboTech Engineering GmbH. http://biogas- 1198 infoboard.de/pdf/presentation_CarboTech%20Engineering%20GmbH.pdf. Accessed 1199 12 December 2014 1200 BOE (2013) Resolución de 21 de diciembre de 2012, Dirección General de Política 1201 Energética y Minas, por la que se modifica el protocolo de detalle PD-01 «Medición, 1202 Calidad y Odorización de Gas» de las normas de gestión técnica del sistema gasista. 1203 BOE N° 6 (7 January 2013). Ministerio de Industria, Energía y Turismo. 1204 Broekhuis R, Koch D, Lynn S (1992) A medium-temperature process for removal of 1205 hydrogen sulfide from sour gas streams with aqueous metal sulfate solutions. Ind Eng 1206 Chem Res 31:2635–2642. doi:10.1021/ie00012a002 1207 Bugante E, Shimomura Y, Tanaka T, Taniguchi M, Oi S (1989) Methane Production from 1208 Hydrogen and Carbon Dioxide and Monoxide in a Column Bioreactor of Thermophilic 1209 53 Methanogens by Gas Recirculation. J Ferment Bioeng 67:419-421. doi:10.1016/0922- 1210 338X(89)90148-7 1211 Burkhardt M, Busch G (2013) Methanation of hydrogen and carbon dioxide. Appl Energ 1212 111:74–79. doi:doi:10.1016/j.apenergy.2013.04.080 1213 Conde JL, Moro LE, Travieso L, Sanchez EP, Leiva A, Dupeirón R, Escobedo R (1993) 1214 Biogas purification process using intensive microalgae cultures. Biotechnol Lett 1215 15:317-320. doi:10.1007/BF00128326 1216 Converti A, Oliveira RPS, Torres BR, Lodi A, Zilli M (2009) Biogas production and 1217 valorization by means of a two-step biological process. Bioresour Technol 100:5771- 1218 5776. doi:10.1016/j.biortech.2009.05.072 1219 CO2 solutions (2014) Harnessing Nature for Efficient Carbon Capture. 1220 http://www.co2solutions.com/. Accessed 2 January 2015 1221 Craggs R, Sutherland D, Campbell H (2012) Hectare-scale demonstration of high rate algal 1222 ponds for enhanced wastewater treatment and biofuel production. J Appl Phycol 1223 24:329-337. doi:10.1007/s10811-012-9810-8 1224 De Godos I, Mendoza JL, Acién FG, Molina E, Banks J, Heaven S, Rogalla F (2014) 1225 Evaluation of carbon dioxide mass transfer in raceway reactors for microalgae culture 1226 using flue gases. Bioresour Technol 153:307–314. doi:10.1016/j.biortech.2013.11.087 1227 Deipser A, Stegmann R (1997) Biological degradation of VCCs and CFCs under simulated 1228 anaerobic landfill conditions in laboratory test digesters. Environ Sci Pollut Res Int 1229 4:209-216. 1230 54 Demmink J, Beenackers A (1998) Gas desulfurization with ferric chelates of EDTA and 1231 HEDTA: new model for the oxidative absorption of hydrogen sulfide. Ind Eng Chem 1232 Res 37:1444–1453. doi:10.1021/ie970427n 1233 Deng L, Chen H, Chen Z, Liu Y, Pu X, Song L (2009) Process of simultaneous hydrogen 1234 sulfide removal from biogas and nitrogen removal from swine wastewater. Bioresour 1235 Technol 100(23):5600–8. doi:10.1016/j.biortech.2009.06.012 1236 Díaz I, Lopes AC, Perez SI, Fdz-Polanco M (2010) Performance evaluation of oxygen, air 1237 and nitrate for the microaerobic removal of hydrogen sulphide in biogas from sludge 1238 digestion. Bioresour Technol 101:7724-7730. doi:10.1016/j.biortech.2010.04.062 1239 Díaz I, Lopes AC, Perez SI, Fdz-Polanco M (2011a) Determination of the optimal rate for 1240 the microaerobic treatment of several H2S concentrations in biogas from sludge 1241 digesters. Water Sci Technol 64:233-238. doi:10.2166/wst.2011.648 1242 Díaz I, Pérez SI, Ferrero EM, Fdz-Polanco M (2011b) Effect of oxygen dosing point and 1243 mixing on the microaerobic removal of hydrogen sulphide in sludge digesters. 1244 Bioresour Technol 102:3768-3775. doi:10.1016/j.biortech.2010.12.016 1245 Díaz I, Ramos I, Fdz-Polanco M (2015) Economic analysis of microaerobic removal of H2S 1246 from biogas in full-scale sludge digesters. Submitted for publication 1247 DMT (2014) The DMT Carborex® PWS biogas upgrading system. Dirkse-milieutechniek. 1248 http://www.dirkse-milieutechniek.com/dmt/do/webPages/200941/DMT_TS- 1249 PWS_Biogas_upgrading.html. Accessed 11 December 2014 1250 55 DMT (2014b) The DMT Carborex® MS biogas upgrading system. Dirkse-milieutechniek. 1251 http://www.dirkse- 1252 milieutechniek.com/dmt/do/webPages/202356/Biogasupgrading_small_size.html. 1253 Accessed 22 December 2014 1254 Dolejs P, Paclík L, Maca J, Pokorna D, Zabranska J, & Bartacek J (2015) Effect of S/N 1255 ratio on sulfide removal by autotrophic denitrification. Appl Microbiol 1256 Biotechnol 99(5): 2383–2392. doi:10.1007/s00253-014-6140-6 1257 Dousková I, Kastánek F, Maléterová Y, Kastánek P, Doucha J, Zachleder V (2010) 1258 Utilization of distillery stillage for energy generation and concurrent production of 1259 valuable microalgal biomass in the sequence: Biogas-cogeneration-microalgae- 1260 products. Energ Convers Manage 51:606-611. doi:10.1016/j.enconman.2009.11.008 1261 European Biogas Association (2013) Proposal for a European Biomethane Roadmap. 1262 http://european-biogas.eu/wp-content/uploads/2013/11/GGG-Biomethane-roadmap- 1263 final.pdf. Accessed 5 January 2015 1264 Eisenmann (2014) Biogas Plants. http://www.eisenmann.com/en/products-and- 1265 services/environmental-technology/biogas-plants/biogas-upgrading.html. Accessed 20 1266 December 2014 1267 Energy Transition–Creative Energy (2014) From biogas to green gas: Upgrading 1268 techniques and suppliers. 1269 http://www.rvo.nl/sites/default/files/bijlagen/From%20Biogas%20to%20Green%20Ga 1270 s%20-%20Upgrading%20techniques%20and%20suppliers.pdf. Accessed 16 December 1271 2014 1272 56 Estrada JM, Kraakman NJR, Lebrero R, Muñoz R (2012) A sensitivity analysis of process 1273 design parameters, commodity prices and robustness on the economics of odour 1274 abatement technologies. Biotechnol Adv 30:1354-1363. 1275 doi:10.1016/j.biotechadv.2012.02.010 1276 EurObserv’ER (2014) Biogas Barometer. . http://www.energies-renouvelables.org/observ- 1277 er/stat_baro/observ/baro224_Biogas_en.pdf. Accessed 20 December 2014 1278 Fernández M, Ramírez M, Gómez JM, Cantero D (2014) Biogas biodesulfurization in an 1279 anoxic biotrickling filter packed with open-pore polyurethane foam. J Hazard Mater 1280 264:529-535. doi:10.1016/j.jhazmat.2013.10.046 1281 Fortuny M, Gamisans X, Deshusses MA, Lafuente J, Casas C, Gabriel D (2011) 1282 Operational aspects of the desulfurization process of energy gases mimics in 1283 biotrickling filters. Water Res 45:5665-5674. doi:10.1016/j.watres.2011.08.029 1284 Gabriel D, Deshusses MA (2003) Retrofitting existing chemical scrubbers to biotrickling 1285 filters for H2S emission control. P Natl Acad Sci USA 100:6308-6312. 1286 doi:10.1073/pnas.0731894100 1287 Bilfinger EMS GmbH (2014) Bio-gas upgrading process. http://www.ems- 1288 clp.de/fileadmin/user_upload/pdf/BIS_EMS_Biogas_A4_EN_scrn.pdf. Accessed 16 1289 December 2014 1290 Schwelm Anlagentechnik GmbH (2014) Biogas Conditioning. http://www.schwelm- 1291 at.de/en/business-divisions/biogas/biogas-conditioning.html. Accessed 16 December 1292 2014 1293 57 Grande CA (2011) Biogas Upgrading by Pressure Swing Adsorption. Biofuel's Engineering 1294 Process Technology. doi:17476 1295 Gunnarsson I, Alvarado-Morales M, Angelidaki I (2014) Utilization of CO2 fixating 1296 bacterium Actinobacillus succinogenes 130Z for simultaneous biogas upgrading and 1297 biosuccinic acid production. Environ Sci Technol 48:12464-12468. 1298 doi:10.1021/es504000h 1299 Günther L (2007) DGE GmbH Presentation: Purification of biomethane using pressureless 1300 purification for the production of biomethane and carbon dioxide. INNOGAS. 1301 http://www.dge- 1302 wittenberg.com/english/vortraege/DGE%20Fachtagung%20WB%202006%20teil1- 1303 EN.pdf. Accessed 12 December 2014 1304 Hagmann M, Hesse E, Hentschel P, Bauer T. Purification of biogas removal of volatile 1305 silicones. In: 8th International Waste Management and Landfill Symposium, Sardinia, 1306 2001. pp 641–644 1307 Higgins V (2007) Siloxane removal process. Parker-Hannifin Corporation. US7306652 B2, 1308 11 December 2007 1309 Horikawa M, Rossi F, Gimenes M, Costa C, Silva M (2004) Chemical absorption of H2S 1310 for biogas purification. Brazilian J Chem Eng 21:415–422. doi:10.1590/S0104- 1311 66322004000300006 1312 Huguen P, Le Saux G (2010) Perspectives for a european standard on biomethane: a 1313 Biogasmax proposal. European Biogasmax project 1314 64 Morweiser M, Kruse O, Hankamer B, Posten C (2010) Developments and perspectives of 1438 photobioreactors for biofuel production. Appl Microbiol Biotechnol 87:1291-1301. 1439 doi:10.1007/s00253-010-2697-x 1440 Muñoz R, Guieysse B (2006) Algal-bacterial processes for the treatment of hazardous 1441 contaminants: A review. Water Res 40:2799-2815. doi:10.1016/j.watres.2006.06.011 1442 Neumann DW, Lynn S (1984) Oxidative adsorption of H2S and O2 by iron chelate 1443 solutions. AIChE J 30:62–69. 1444 Nordberg Å, Edström M, Uusi-Penttilä M, Rasmuson ÅC (2012) Selective desorption of 1445 carbon dioxide from sewage sludge for in-situ methane enrichment: Enrichment 1446 experiments in pilot scale. Biomass Bioenerg 37:196-204. 1447 doi:10.1016/j.biombioe.2011.12.012 1448 Patterson T, Esteves S, Dinsdale R, Guwy A (2011) An evaluation of the policy and 1449 techno-economic factors affecting the potential for biogas upgrading for transport fuel 1450 use in the UK. Energ Policy 39:1806-1816. doi:10.1016/j.enpol.2011.01.017 1451 Peillex J, Fardeau M, Boussand R, Navarro J, Belaich J (1988) Growth of Methanococcus 1452 thermolithotrophicus in batch and continuous culture on H2 and CO2: influence of 1453 agitation. Appl Microbiol Biotechnol 29:560-564. doi:10.1007/BF00260985 1454 Persson M (2003) Evaluation of upgrading techniques for biogas. Rapport SGC 142. 1455 Swedish Gas Center. 1456 https://cdm.unfccc.int/filestorage/E/6/T/E6TUR2NNQW9O83ET10CX8HTE4WXR2 1457 65 O/Evaluation%20of%20Upgrading%20Techniques%20for%20Biogas.pdf?t=YWt8bml 1458 5eTJsfDBCijpDYjFf2sE5_wGsjeuV. Accessed 5 August 2014 1459 Persson M Biogas upgrading and utilization as vehicle fuel. In: European Biogas Workshop 1460 – The Future of Biogas in Europe III. University of Southern Denmark, 2007. pp 59-64 1461 Persson M, Jönsson O, Wellinger A (2006) Biogas upgrading to vehicle fuel standards and 1462 grid injection. IEA Bioenergy. http://www.iea-biogas.net/_download/publi- 1463 task37/upgrading_report_final.pdf. Accessed 1 April 2013 1464 Persson M, Wellinger A, Rehnlund B, Rahm L (2007) Report on Technological 1465 Applicability of Existing Biogas Upgrading Processes. Biogasmax. 1466 http://www.biogasmax.eu/media/report_on_technological_2007__041639600_1025_2 1467 2052007.pdf. Accessed 20 October 2014 1468 Petersson A, Wellinger A (2009) Biogas upgrading technologies – developments and 1469 innovations. IEA Bioenergy. Task 37. http://www.iea-biogas.net/_download/publi- 1470 task37/upgrading_rz_low_final.pdf. Accessed 5 June 2014 1471 Popat S, Deshusses M (2008) Biological Removal of Siloxanes from Landfill and Digester 1472 Gases: Opportunities and Challenges. Environ Sci Technol 42:8510-8515. 1473 doi:10.1021/es801320w 1474 Puregas P (2014) Biogas upgrading. http://www.purac-puregas.com/technology/biogas- 1475 upgrading/. Accessed 15 December 2014 1476 66 Putt R, Singh M, Chinnasamy S, Das KC (2011) An efficient system for carbonation of 1477 high-rate algae pond water to enhance CO2 mass transfer. Bioresour Technol 1478 102:3240-3245. doi:10.1016/j.biortech.2010.11.029 1479 Raja R, Hemaiswarya S, Kumar N, Sridhar S, Rengasamy R (2008) A perspective on the 1480 biotechnological potential of microalgae. Crit Rev Microbiol 34:77-88. 1481 doi:10.1080/10408410802086783 1482 Ramos I, Fdz-Polanco M (2014) Microaerobic control of biogas sulphide content during 1483 sewage sludge digestion by using biogas production and hydrogen sulphide 1484 concentration. Chem Eng J 250:303-311. doi:10.1016/j.cej.2014.04.027 1485 Ramos I, Perez R, Reinoso M, Torio R, Fdz-Polanco M (2014) Microaerobic digestion of 1486 sewage sludge on an industrial-pilot scale: The efficiency of biogas desulphurisation 1487 under different configurations and the impact of O2 on the microbial communities. 1488 Bioresour Technol 164:338-346. doi:10.1016/j.biortech.2014.04.109 1489 Rasi S (2009) Biogas composition and upgrading to biomethane. University of Jyväskylä, 1490 Jyväskylä 1491 Raven JA, Cockell CS, De la Rocha CL (2008) The evolution of inorganic carbon 1492 concentrating mechanisms in photosynthesis. Phil Trans R Soc B 363:2641-2650. 1493 doi:10.1098/rstb.2008.0020 1494 Rodríguez E, Lopes A, Fdz-Polanco M, Stams AJM, García-Encina P (2012) Molecular 1495 analysis of the biomass of a fluidized bed reactor treating synthetic vinasse at 1496 67 anaerobic and micro-aerobic conditions. Appl Microbiol Biotechnol 93:2181-2191. 1497 doi:10.1007/s00253-011-3529-3 1498 Rodriguez G, Dorado AD, Fortuny M, Gabriel D, Gamisans X (2014) Biotrickling filters 1499 for biogas sweetening: Oxygen transfer improvement for a reliable operation. Process 1500 Saf Environ 92:261-268. doi:10.1016/j.psep.2013.02.002 1501 Rutledge B (2005) California Biogas Industry Assessment White Paper. WestStart- 1502 CALSTART. 1503 http://www.calstart.org/Libraries/Publications/California_Biogas_Industry_Assessmen 1504 t_White_Paper.sflb.ashx. Accessed 7 March 2014 1505 Ryckebosch E, Drouillon M, Vervaeren H (2011) Techniques for transformation of biogas 1506 to biomethane. Biomass Bioenerg 35:1633-1645. doi:10.1016/j.biombioe.2011.02.033 1507 Schmidt K, Augenstein T, Heidinger M, Ertl S, Tiehm A (2010) Aerobic biodegradation of 1508 cis-1,2-dichloroethene as sole carbon source: Stable carbon isotope fractionation and 1509 growth characteristics. Chemosphere 78:527–532. 1510 doi:10.1016/j.chemosphere.2009.11.033 1511 Schneider RL, Quicker P, Anzer T, Prechtl S, Faulstich M (2002) Grundlegende 1512 Untersuchungen zur effektiven, kostengünstigen Entfernung von Schwefelwasserstoff 1513 aus Biogas. In: Biogasanlagen Anforderungen zur Luftreinhaltung. Ausburg, 1514 Schweigkofler M, Niessner R (2001) Removal of siloxanes in biogases. J Hazard Mater 1515 B83:183–196. 1516 68 Serejo M, Posadas E, Boncz M, Blanco S, Garcia-Encina P, Muñoz R (2015) Tailoring 1517 biomass composition during the optimization of the integral upgrading of biogas in 1518 microalgal-bacterial processes. Environ Sci Technol. Submitted for publication 1519 Sinnott RK (2005) Chemical Engineering Design vol 6. 4th edn. Elsevier Butterworth- 1520 Heinemann, Oxford 1521 Soreanu G, Béland M, Falletta P, Edmonson K, Seto P (2008) Laboratory pilot scale study 1522 for H2S removal from biogas in an anoxic biotrickling filter. Water Sci Technol 1523 57:201-207. doi:10.2166/wst.2008.023 1524 Soreanu G, Béland M, Falletta P, Edmonson K, Svoboda L, Al-Jamal M, Seto P (2011) 1525 Approaches concerning siloxane removal from biogas - A review. Can Biosyst Eng 1526 53:8.1-8.18. 1527 Soreanu G, Béland M, Falletta P, Ventresca B, Seto P (2009) Evaluation of different 1528 packing media for anoxic H2S control in biogas. Environ Technol 30:1249-1259. 1529 doi:10.1080/09593330902998314 1530 Spolaore P, Joannis-Cassan C, Duran E, Isambert A (2006) Commercial applications of 1531 microalgae. J Biosci Bioeng 101:87-96. doi:10.1263/jbb.101.87 1532 Strevett KA, Vieth RF, Grasso D (1995) Chemo-autotrophic biogas purification for 1533 methane enrichment: mechanism and kinetics. Chem Eng J Bioch Eng 58:71-79. 1534 doi:10.1016/0923-0467(95)06095-2 1535 69 Tang K, Baskaran V, Nemati M (2009) Bacteria of the sulphur cycle: An overview of 1536 microbiology, biokinetics and their role in petroleum and mining industries. Biochem 1537 Eng J 44:73-94. doi:10.1016/j.bej.2008.12.011 1538 Thrän D et al. (2014) Biomethane – status and factors affecting market development and 1539 trade. IEA Task 40 and Task 37 Joint Study. 1540 http://www.bioenergytrade.org/downloads/t40-t37-biomethane-2014.pdf. Accessed 20 1541 December 2014 1542 Tock L, Gassner M, Maréchal F (2010) Thermochemical production of liquid fuels from 1543 biomass: Thermo-economic modeling, process design and process integration analysis. 1544 Biomass Bioenerg 34:1838-1854. doi:10.1016/j.biombioe.2010.07.018 1545 Tomàs M, Fortuny M, Lao C, Gabriel D, Lafuente J, Gamisans X (2009) Technical and 1546 economical study of a full-scale biotrickling filter for H2S removal from biogas. Water 1547 Pract Technol 4. doi:10.2166/wpt.2009.026 1548 Tredici MR (2009) Photobiology of microalgae mass cultures: Understanding the tools for 1549 the next green revolution. Biofuels 1:143-162. doi:10.4155/bfs.09.10 1550 Tynell Å, Börjesson G, Persson M (2007) Microbial Growth on Pall Rings: A Problem 1551 When Upgrading Biogas With the Water-Wash Absorption Technique. Appl Biochem 1552 Biotechnol 141:299 - 320. doi:10.1007/BF02729069 1553 Urban W, Girod K, Lohmann H (2009) Executive Report: The German Market for 1554 Biomethane. Deutsche Energie-Agentur GmbH (DENA), German Energy Agency. 1555 70 http://exportinitiative.dena.de/fileadmin/user_upload/Table_of_Contents_v3_Biometha 1556 n.pdf. Accessed 22 December 2014 1557 Wang B, Li Y, Wu N, Lan C (2008) CO2 bio-mitigation using microalgae. Appl Microbiol 1558 Biot 79:707-718. doi:10.1007/s00253-008-1518-y 1559 Wang W, Xie L, Luo G, Zhou Q, Angelidaki I (2013) Performance and microbial 1560 community analysis of the anaerobic reactor with coke oven gas biomethanation and in 1561 situ biogas upgrading. Bioresour Technol 146:234–239. 1562 doi:10.1016/j.biortech.2013.07.049 1563 Wellinger A, Lindberg A (1999) Biogas upgrading and utilization. IEA Bioenergy. 1564 http://www.biogasmax.eu/media/biogas_upgrading_and_utilisation__018031200_1011 1565 _24042007.pdf. Accessed 18 December 2014 1566 Xebex (2014) BGX SOLUTIONS - Biogas Plants. http://www.xebecinc.com/biogas- 1567 plants.php. Accessed 5 January 2015 1568 Yan C, Zheng Z (2013) Performance of photoperiod and light intensity on biogas upgrade 1569 and biogas effluent nutrient reduction by the microalgae Chlorella sp. Bioresour 1570 Technol 139:292–299. doi:10.1016/j.biortech.2013.04.054 1571 1572 71 Figure 1. Biogas upgrading by liquid absorption. A) Water scrubbing; B) Organic solvent scrubbing; C) Chemical scrubbing. Adapted from Bauer et al (2013b). Figure 2. Biogas upgrading by Pressure Swing Adsorption (PSA). Adapted from Bauer et al (2013b). Figure 3. Biogas upgrading by membrane separation. Different configurations of gas-gas units: I) single-pass membrane unit, II) multiple stage membrane units with internal recirculation of permeate and III) internal recirculation of retentates. Adapted from Bauer et al (2013b). Figure 4. Biogas upgrading using microalgae cultures. Adapted from Bahr et al (2014). Figure 5. CO2 removal by in-situ desorption in the anaerobic digester. Figure 6. Evolution of sulfur species under anaerobic/microaerobic conditions and the effect of mixing conditions. Figure 1. Raw biogas Upgraded biogas Air with desorbed CO2 Air Make-up water Water bleed Absorption column Flash column Desorption column Compressor Raw biogas Upgraded biogas Off-gas Stripper gas Absorption column Flash column Desorption column Compressor Condensate Cooler Heater Gas conditioning Raw biogas Upgraded biogas Absorber Stripper Cooler Heater Reboiler Hot water/oil or steam Off-gas B C A Figure 1. Biogas upgrading by liquid absorption. A) Water scrubbing; B) Organic solvent scrubbing; C) Chemical scrubbing. Adapted from Bauer et al (2013b). Figure 1 Click here to download Figure: Figure 1.doc Figure 2. Compressor Raw biogas Waste gas Vacuum pump Upgraded biogas Purge gas Adsorption Pressurization Desorption Depressurization PSA columns Figure 2. Biogas upgrading by Pressure Swing Adsorption (PSA). Adapted from Bauer et al (2013b). Figure 2 Click here to download Figure: Figure 2.doc Table 2. Table 2. Commercial upgrading technologies Technology CH4 (%) CO2 (%) H2S (%) Methane loss Costs Power consumption Examples References High pressure water scrubbing DMT Carborex®PWS P= 8-10 bar CO2 and H2S removal Solvent regeneration: Flash tank in two steps:1) 2-4 bar; 2) 1 bar. Air stripping unit and Biotrickling Filter. ˃97% ˂ 2% ˂ 2 ppmv ˂ 2% 0.105 € m-3 (250 Nm3 h-1) 0.052 € m-3 (2000 Nm3 h-1) 0.4-0.5 kWh m-3 produced gas 1) Zalaegerszeg, HU, Okoprotec (50-85 Nm3 h-1; WWTP) 2) Zwolle, NL, Nature Gas Overijssel (520 Nm3 h-1; green waste and other garbage) 3) Wijster, NL (1500 Nm3 h-1; Landfill) DMT (2014) Malmberg COMPACT® CO2 and H2S removal Capacity: 100-3000 Nm3 h-1 Methane emissions are avoided by thermal oxidation in the process air. ˃97% 1-2% ˂1% 2 ct kWh-1 (250 Nm3 h-1) 1 ct kWh-1 (2000 Nm3 h-1) 1) Stockholm Vatten, Henriksdal (1400 Nm3 h-1; WWTP) 2) Jönköping Municipality, Sweden (150 Nm3 h-1; sludge digestion) Malmberg (2014) Chemical scrubbing OASEgreen™ Process (Bilfinger EMS GmbH) Chemisorption with PuraTreat™ solvent CO2 and H2S removal Atmospheric pressure T° solvent regeneration: 106-110°C Capacity: 600- 10.000 Nm³ h-1 ˃99% ˂ 1% ˂ 4 ppmv ˂0.05% ˂ 0.01 € kWh-1 of raw biogas 1) BUP´s Verbio (2 separate plants Schwedt and Zörbig; 6000 Nm3 h-1) 2) BUP Weltec (Arneburg; 1450 Nm3 h-1) Bilfinger EMS GmbH (2014) LP Cooab-technique (Cirmac) Absorption by amines CO2 removal Atmospheric pressure Exhaust-gas treatment is not necessary 99.5% ˂0.1% 0.05 - 0.12 kWe Nm-3 raw gas Gasslosa biogas plant in Boras, Sweden Energy Transition– Creative Energy (2014) Table 2 Click here to download Table: Table 2.doc CApure™ process (Purac Puregas) Absorption by amines CO2 removal Atmospheric pressure 100 - 3000 raw biogas Nm3 h-1 99% 0.20% ˂ 0.5 ppmv ˂0.1% 0.23 - 0.26 kWh Nm-3 raw gas (with heat recovery system) Purac Puregas (2014) Organic physical scrubbing Schwelm Biogas treatment plant Capacity: 200-1600 Nm3 h-1 Absorption by polyethylene glycol. 98% ˂1% 0.21 kWh Nm-³ of raw gas Schwelm Anlagentechni k GmbH (2014) Pressure Swing adsorption Xebec PSA P= 8-11 bar 9 vessel system with a patented rotary valve Previous H2S removal Regeneration under vacuum pressure (typically 0.5 bar) Capacity: 100-10000 Nm3 h-1 Removal CO2 and water vapour 98% 1-2% 1)Scenic View Dairy, Fennville, Michigan (animal waste; 225Nm3 h-1) 2)Rumpke Landfill Cincinnati,Ohio (7000 Nm3 h-1) Xebec (2014) Membrane separation DMT Carborex® MS Previous H2S and water vapour removal P= 10 bar The off-gas contains over 99.5% CO2. Removal CO2 Gas/gas membrane 97- 99% 1-3% <0.5% 50 Nm3 h-1 (0.432 ct Nm-3); 200 Nm3 h-1 (0.211 ct Nm-3) < 0.22 kWh Nm-3 DMT (2014b) Biopower plant P = 16 bar Hollow fiber membrane Removal CO2 Gas/gas membrane 96% <1% Biopower plant in Pratteln, Switzerland (210 Nm3 h-1;high solids digestion, biowaste, yard waste) Eisenmann (2014) Table 3. Experimental studies on the chemoautotrophic CO2 conversion to CH4 Bioreactor configuration CO2:H2 (mol mol-1) Gas Residence Time (h) Maximum CH4 production CH4 (%) Reference Mesophilic sewage sludge STR digester (2 L) stirred at 200 rpm supplied with in-situ coke gas addition (92 %H2/8% CO) via bubbleless membranes 0.11-0.24 13-22 1.45 L CH4 gVS-1 d-1 0.65 L CH4 Lr-1 d-1 90-99 Wang et al (2013) Mesophilic biotrickling filter (27 L) with random packing and internal gas recycling supplied with synthetic CO2:H2 mixtures. Batchwise operation 0.25 2-10 1.17 NL CH4 Lr-1 d-1 94-98 Burkhardt and Busch (2013) Mesophilic STR (100L) stirred at 70 rpm with sparging of residual H2 and CO2 gases 0.125-0.5 (0.2)* 42-208 4.1 L CH4 Lr-1 d-1 92 Kim et al (2013) Thermophilic manure-whey STR digester (0.6 L) stirred at 150-300 rpm with in-situ H2 supply via ceramic and column diffusers. 0.25 14 0.88 L CH4 Lr-1 d-1 75 Luo and Angelidaki (2013) Thermophilic STR (0.6L) stirred 500-800 rpm with sparging of synthetic mixture of H2:CH4:CO2 (60:25:15) 0.25 1-8 5.3 L CH4 Lr-1 d-1 90-95 Luo and Angelidaki (2012a) Mesophilic STR (0.5 L) supplied with synthetic CO2:H2 mixtures 0.25 1 0.24 L CH4 gVS-1 d-1 2.4 L CH4 Lr-1 d-1 - Ako et al (2008) Mesophilic packed bed filter (7.8L) supplied with synthetic CO2:H2 mixtures 0.125-0.5 (0.2)* 3.8-6.5 1.34 L CH4 Lr-1 d-1 100 Lee et al (2012) Table 3 Click here to download Table: Table 3.doc Mesophilic Hollow Fiber biofilm membrane bioreactor (0.195 L) supplied with synthetic CO2:H2 mixtures 0.25 1.2 4.6 L CH4 Lr-1 d-1 80-90 Ju et al (2008) Thermophilic STR (2L) with sparging via membrane diffusion of synthetic biogas mixtures and H2 0.27 0.13 - 96 Strevett et al (1995) Thermophilic column packed bed reactor (0.2L) sparged with synthetic CO2:H2 mixtures 0.25 - 54 L CH4 Lr-1 d-1 - Bugante et al (1989) Thermophilic packed bed column (0.105 L) supplied downwards with a synthetic CO2:H2 mixture 0.25 0.033 105 L CH4 Lr-1 d-1 40-50 Jee et al (1988) Thermophilic STR (1.5L) stirred at 320- 1015 rpm supplied via sparging with a synthetic CO2:H2 mixture (batch and continuous) 0.25 0.012 76 L CH4 Lr-1 d-1 (continuous) 470 L CH4 Lr-1 d-1 (batch) 50% Peillex et al (1988) Thermophilic packed bed column (0.05 L) supplied downwards with a synthetic CO2:H2 mixture 0.25 0.02 144 L CH4 Lr-1 d-1 30 Jee et al (1987) *- Optimum value Table 4. Table 4. Experimental studies on biogas upgrading and CO2 removal from flue gas in microalgal photobioreactors Photobioreactor and absorption unit configuration Gas Residence Time* (h) CO2-RE (%) Microalgae productivity (g l-1 d-1) O2 (%) N2 (%) CH4 (%) Reference Indoor 180 L raceway inoculated with a microalgae consortium and interconnected to a 2.5 L bubble column (1.65 m height) via algal-broth recirculation at a liquid to biogas ratio of 1:10. Synthetic Biogas (30%/69.5%/0.5% CO2/CH4/H2S) supplied via porous diffuser. 1.4 822 0.079 1 6 88 Serejo et al (2015) Indoor 180 L raceway inoculated with Spirulina platensis and interconnected to a 0.8 L bubble column (0.6 m height) via algal-broth recirculation at a liquid to biogas ratio of 1:1. Simulated biogas (30%/69.5%/0.5% CO2/N2/H2S) supplied via porous diffuser. 0.7 865 - 0.2 - - Bahr et al (2014) Indoor 1 L column photobioreactor stirred at 100 rpm supplied with real biogas (CH4 70-72%, CO2 17-19%) and inoculated with Arthrospira platensis. 96 100 0.041 10-24 - - Converti et al (2009) Indoor 0.45 L enclosed tubular photobioreactor supplied with biogas (41%/57.5%/0.05% CO2/CH4/H2S) inoculated with Chlorella vulgaris. - 98 - 18-23 - 50-53 Mann et al (2009) Table 4 Click here to download Table: Table 4.doc Indoor 15 L algal ponds inoculated with Chlorella vulgaris using a biolift absorption unit inside the pond and supplied with real biogas (CH4 55-71%, CO2 44-48%, H2S 1 %). - 74-95 - - - 88-97 Conde et al (1993) Outdoor pilot raceway supplied with simulated biogas (40%/60% CO2/N2) using a countercurrent absorption sump (1 m deep) using a mixed microalgae population - >85 - 5.2-6 - - Mandeno et al. (2005) Indoor 0.4-6 L bubble column photobioreactor inoculated with Chlorella vulgaris supplied with real biogas (CH4-38- 80%, CO2-19-62%, H2S-0.2 %). 0.16 - 2.6-3.8 3.5 < - - Douskova et al (2010) Outdoors 50 L bubble column photobioreactor (3 m height) inoculated with a mutant Chlorella strain supplied with biogas (20%/69%/0.005% CO2/CH4/H2S) using intermittent biogas/air cycles (30 min/30 min) 0.06-0.3 74-85 0.3-0.32 - - 86-91 Kao et al (2012) Outdoor 100 m2 raceway constructed with a 0.65 m3 absorption sump (1 m deep) operated at a liquid recirculation rate of 0.22 m s-1 supplemented with flue gas (10.6 % CO2) via membrane diffuser 0.2 96 0.088 >15 - - De godos et al (2014) Outdoor 420 L raceway interconnected to a 1.4 L bubble column (3.1 m height) via water recycling from the HRAP. Abiotic experiment at pH 9-10 0.025 82-83 - - - - Putt et al (2011) Indoor 75 L open photobioreactor inoculated with Nannochloropsis gaditana and interconnected to a 0.7 L bubble column (2.2 m height) by continuous recirculation of microalgae culture at a liquid to biogas ratio of 1.8:1. Real biogas (72±2% CH4; 28±2% CO2) was supplied. 0.2 93 0.03 1.2 - - Meier et al (2015) *Gas Residence Time estimated based on the volume of the absorption unit Table 5. Table 5. Design and operation parameters of H2S biofiltration units under anoxic and aerobic conditions during biogas upgrading. Biofiltration Unit [H2S] (ppmv) Gas Residence Time (min) H2S-RE (%) Elimination Capacity (g H2S m-3 h-1) Reference Aerobic biotrickling filter (5.15 m3) packed with plastic pall rings and operated with an aeration rate of 5.6 m3 h-1 at a pH of 1.7 controlled by WWTP effluent addition 2107 151 3.8-5.9 992 5413 Rodríguez et al (2014) Aerobic unit with metal wire, plastic tubing and paper strips, inoculated with 1 L of anaerobic sludge and supplemented with real biogas and O2/H2S ratios of 2-18 2800-3700 61-100 96 40-100 Ramos et al (2013) Aerobic biotrickling filter (2 L) packed with HD-QPAC supplied with H2S/N2 synthetic mixtures simulating biogas and operated at O2/H2S ratios of 23.6 2000 3 99 55 Maestre et al (2010) Aerobic biotrickling filter (2L) packed with metallic pall rings, fed with H2S/N2/CH3SH synthetic mixtures and operated at O2/H2S ratios of 39, at a pH of 6-6.5 with air sparged at the bottom of the BTF 2000 3 99 52 Montebello et al (2012) Aerobic biotrickling filter (2 L) packed with HD-QPAC supplied with H2S/N2 synthetic mixtures simulating biogas and operated at O2/H2S ratios of 23.6 and a pH of 6-6.5 2000 2-3 98 55-82 Fortuny et al (2011) Aerobic biotrickling filter (2.4 L) packed with metallic pall rings, fed with H2S/N2 mixtures simulating biogas and operated at a pH of 2.5 and O2/H2S ratios of 8.2-41.2 2000- 10000 2.1 80-100 52-223 Montebello et al (2014) Table 5 Click here to download Table: Table 5.doc Aerobic biotrickling filter (12 m3) packed with plastic pall rings, fed with real biogas (69% CH4, 29% CO2, 1% N2) and operated at a pH of 2.7 1250-4750 1.9-9.7 99 50* Tomàs et al (2009) Anoxic biotrickling filter (2.3L) packed with polyurethane foam, fed with H2S/CH4/CO2/CH3SH synthetic mixtures and operated at a pH 7.5. NO3- was used as edonor 2000 2.7 99 59 Montebello et al (2012) Anoxic biotrickling filter (2.4L) packed with polyurethane foam, fed with real biogas (68% CH4/ 26% CO2) supplemented with H2S and operated at a pH 7.5. Ca(NO3)2, KNO3 and NaNO3 were used as eacceptor. - 2.4-3.4 99 99.8-130 Fernández et al (2014) Anoxic biotrickling filters (6.7 L) packed with polyester fibers and lava rock, supplied with synthetic biogas (65% CH4/ 35% CO2) using NO3- supplemented SBR effluent at a pH of 6.5. 500-1500 5-16 93-96 177-182 Soreanu et al (2009) *- Average elimination capacity Table 6. Table 6. Experimental studies on in-situ microaerobic H2S removal Bioreactor configuration Biogas (m3 m-3Rd-1) Biogas Residence Time in headspace (h) [H2S] (ppmv) Residual [H2S] (ppmv) H2S RE (%) O2/H2S (mol mol-1) Reactive Rate % Residual [O2] % Reference Mesophilic digester of agricultural wastes 250 m3/h 2.5 2500 < 300 > 88 1.3 -1.7 1.5 - 2 % air - Schneider et al (2002) Mesophilic digesters (2 × 1500 m3) of WWTP sludge 0.41 - 3300 30 99 3.7 5.4% air - Jenicek et al (2008) Mesophilic digester (2100 m3) of WWTP sludge 0.40 - 5600 54 99 5.5 14% air - Jenicek et al (2008) Mesophilic digester (200 L) of WWTP sludge 0.95 6.3 13000 < 50 > 98 1.1 1.4% O2 0.6 Díaz et al (2011b) Mesophilic digester (200 L) of WWTP sludge 1.07 5.3 10000 260 > 97 1 4.7% air 0.7 Diaz et al (2010) Mesophilic digester (5 m3) of WWTP sludge 1.00 9.6 2500 – 4900 < 72 > 99 0.9 - 2 0.5% (92- 98% O2) < 0.1 Ramos et al (2014) Mesophilic digester (200 L) of WWTP 0.75 8 3300 – 5000 < 10 99 1 0.3-0.5% O2 < 0.1 Ramos and Fdz-Polanco Table 6 Click here to download Table: Table 6_R.doc