Two-phase anaerobic digestion for production of hydrogen–methane mixtures

Download Two-phase anaerobic digestion for production of hydrogen–methane mixtures

Post on 21-Jun-2016




0 download

Embed Size (px)


<ul><li><p>iget</p><p>Ch</p><p>Ea</p><p>gy,</p><p>No</p><p>ber13</p><p>Abstract</p><p>1. Introduction converted to H2 fuel (Das and Veziroglu, 2001; Van Ginkel</p><p>butyric and lactic, as well as alcohols and ketones, are alsoformed during the breakdown of the organic substrates bythe highly diverse populations generically known as acido-gens. However, in a well operating process, these productsare mostly converted to acetic acid and H2, which, in turn,are then converted to methane gas. The key to this process</p><p>* Corresponding author. Present address: Hawaii Natural EnergyInstitute, University of Hawaii, 1680 EastWest Road POST 109,Honolulu HI 96822, United States. Tel.: +1 808 956 7337; fax: +1 808956 2336.</p><p>E-mail address: (M. Cooney).</p><p>Bioresource Technology 98 (2Biological hydrogen fuel production is a challengingarea of biotechnology. Photobiological processes, by whichmicroalgae or photosynthetic bacteria produce H2 fromsunlight and either water or organic substrates, respec-tively, have been studied for several decades, but arelimited by many practical and fundamental limitations(Benemann, 1997; Levin et al., 2004; Nath and Das,2004). Dark bacterial hydrogen fermentations or organicsubstrates also have limitations, principally the relativelylow yields of hydrogen obtained until now, with typicallyonly 1020%, and at most 30%, of the substrate energy</p><p>and Logan, 2005). This compares to 8090% yieldsobtained in commercial ethanol or methane fermentations(Claassen et al., 1999). Higher yields may be achievable, inprinciple, through metabolic engineering (Hallenback andBenemann, 2002; Keasling et al., 1998) but, thus far, nomajor improvements in yields have been reported.</p><p>Methane fermentations, also called anaerobic digestion,involve consortia of two major types of bacteria: the so-called acidogenic bacteria that break down the substratesinto mainly H2, acetic acid and CO2, and the methanogenicbacteria, that convert acetic acid, H2 and CO2 to methanegas. A variety of higher organic acids, such as propionic,An anaerobic digestion process to produce hydrogen and methane in two sequential stages was investigated, using two bioreactors of2 and 15 L working volume, respectively. This relative volume ratio (and shorter retention time in the second, CH4-producing reactor)was selected, in part, to test the assumption that separation of phase can enhance metabolism in the second methane producing reactor.The reactor system was seeded with conventional anaerobic digester sludge, fed with a glucoseyeast extract peptone medium and oper-ated under conditions of relatively low mixing, to simulate full scale operation. A total of nine steady states were investigated, spanning arange of feed concentrations, dilution rates, feed carbon to nitrogen ratios and degree of integration of the two stages. The performanceof this two-stage process and potential practical applications for the production of clean-burning hydrogenmethane mixtures arediscussed. 2006 Elsevier Ltd. All rights reserved.</p><p>Keywords: Biological hydrogen; Methane; Anaerobic digestion; Renewable energy; Acetogen; MethanogenTwo-phase anaerobic dof hydrogenm</p><p>Michael Cooney a,b,*, Nathan Maynard a,a Hawaii Natural Energy Institute, University of Hawaii, 1680</p><p>b Center for Space Research, Massachusetts Institute of Technoloc Benemann Associates, 3434 Tice Creek Dr.</p><p>Received in revised form 11 SeptemAvailable online0960-8524/$ - see front matter 2006 Elsevier Ltd. All rights reserved.doi:10.1016/j.biortech.2006.09.054estion for productionhane mixtures</p><p>ristopher Cannizzaro b, John Benemann c</p><p>stWest Road POST 109, Honolulu HI 96822, United States</p><p>77 Massachusetts Avenue, Cambridge MA 02139, United States</p><p>.1, Walnut Creek, CA 94595, United States</p><p>2006; accepted 16 September 2006December 2006</p><p>007) 26412651</p></li><li><p>Tecis that the H2 produced by the acetogenic bacteria isremoved to a very low partial pressure, typically in thenanomolar range, by the methanogenic bacteria, whichallows the otherwise thermodynamically unfavorablemetabolism of the higher organic acids and alcohols to ace-tic acid and H2 (Belaich et al., 1990). The result of thiscomensal relationship is an overall high yield conversionof fermentable substrates to methane fuel with at mosttrace amounts of H2 present in the gas phase (Ferris,1993). Indeed, even a small amount of H2 (&gt;0.1%) in thegas phase indicates a malfunctioning process, due to over-load, toxicity, or other factors unbalancing the comensalrelationships, generally followed by a cessation of CH4-production. However, the two reactions, formation oforganic acids and H2 and methane-production, can, at leastpartially be separated into separate bioreactors in series, inwhich the rst, smaller, reactor produces organic acids, H2and CO2, while the second, much larger, reactor producesCH4 and CO2. Such two-phase anaerobic digestions wereproposed as a way to optimize for the growth of each typeof bacteria in the separate reactors, specically by growingthe acetogenic bacteria at a lower pH (e.g., 56) and shorthydraulic residence time (typically 12 days) in the rststage, while the slower growing methanogenic bacteriastage, requiring a more neutral pH, were preferentially cul-tured in the second stage with a much longer hydraulic res-idence time (typically 1020 days), (Blonskaja et al., 2003;Demirel and Yenigun, 2002; Pohland and Ghosh, 1971).However, as noted above, anaerobic digestion involves acomensal interaction of the two general types of bacterialpopulations, in which the methanogens feed on and e-ciently remove, the waste products (H2 and acetic acid)of the acidogenic bacteria. Thus separating these two basicprocesses will not generally signicantly accelerate orincrease overall methane-production, although it can beof some advantage in making the process more resistantto shock loading.</p><p>In earlier work on two-phase anaerobic digestion, thegas, H2 and CO2, produced in the rst stage was trans-ferred to the second stage to be converted to CH4, and,indeed, few measurements on gas production from the rstphase are reported in the literature. However, due to theburgeoning interest in H2 fuels and fuel cells, there hasbeen a great deal of research in recent years on dark H2 fer-mentations (Das and Veziroglu, 2001; Hallenback andBenemann, 2002; Levin et al., 2004) which, essentially, cor-respond to the rst phase of such a two-phase anaerobicdigestion process. However, as stated above, in all suchstudies, the overall H2 yields are low, only 1020% of thesubstrate energy being converted to H2 fuel with theremainder converted to organic acids, and other products.A number of authors have proposed converting these by-products to H2 fuel using photosynthetic bacteria, whichcan exhibit high yields, though at very low solar conversioneciencies (Miyake, 1998), which makes such an approach</p><p>2642 M. Cooney et al. / Bioresourceimpractical. Producing methane gas in a second reactor hasbeen proposed as another option (Benemann, 1998), andsome work on such dual H2 and CH4-production hasappeared recently (Benemann et al., 2004; Kramer andBagley, 2004). However, if H2-production is the objective,it is more direct and plausible to produce methane gas bynormal anaerobic digestion and then convert this fuel toH2 through a conventional reformer process.</p><p>A potential near-term practical application of two-phaseanaerobic digestion is the production of H2CH4 methanemixtures. H2CH4 mixtures, in the range of 1030% H2and 9070% CH4, on a volumetric basis, are known toburn with much lower NOx emission in internal combus-tion engines, and this allows the use of such fuels in regionswhere NOx emissions are strictly regulated (Bauer and For-est, 2001; Collier et al., 1996). Here we address the co-pro-duction of H2 and CH4 mixtures in a two phase anaerobicdigestion process, using a simulated high carbohydratewastewater with a mixed bacterial population obtainedfrom a conventional anaerobic digester under operatingconditions designed to simulate a scaled-up process.</p><p>2. Methods</p><p>2.1. Two-stage anaerobic bioreactor system</p><p>The two-stage anaerobic bioreactor system (Fig. 1)consisted of a 4.7 L polycarbonate jar with cover(DS5300-9609, Nalgene, Rochester, NY) for the rst(hydrogen-production,) reactor and a 18.8 L polycar-bonate jar and cover (DS5300-9212, Nalgene) for the sec-ond (methane-production) reactor. The workingvolumes were 2.0 and 15.0 L, respectively. All connectionswere made with Teon tubing (890 FEP, Nalgene) andstainless steel or nylon compression ttings (SwagelokCo., Solon, Ohio). The lids for the reactors were sealedby compression against an oring using two steel platesplaced above and below the reactor and bolted in placealong four corners. The assembled reactors were indepen-dently tested for gas leakage by introducing nitrogen gasvia a gas sparging port and with an exhaust tube placedat the bottom of a graduated cylinder lled with 45.0 cmof water (corresponding to a pressure head of about0.05 bar or 0.67 psi). After the nitrogen ow was stoppedthe gas level in the submersed exhaust tube remained con-stant for at least 1 h, indicating no signicant gas leakage.</p><p>The reactors were mixed by placing each reactor ona magnetic stirrer (PC-310, Corning Inc., Corning, NY)and stirring with 1.5 in. magnetic stir bars. When thehydrogen and methane-production reactors were operatedindependently (non-integrated operation), the euentfrom each reactor was peristaltically pumped to separatewaste bottles. When operated as a two-phase process(integrated operation), the euent from the rst reactorwas pumped into the second reactor. In both cases, peri-staltic pumping of reactor media was through a tube thatwas placed below the surface of the reactor liquid, to avoid</p><p>hnology 98 (2007) 26412651the exchange of gases between the two reactors, foam frac-tionation and related artifacts. The liquid height and thus</p></li><li><p>e frto esc</p><p>ed t</p><p>TecN2 </p><p>Cryostat </p><p>Feed Tank </p><p>Base </p><p>Acid</p><p>The relativdelivered reactor is d</p><p>Fig. 1. Two-stage anaerobic reactor system design. The rst stage is referrCH4-production reactor.</p><p>M. Cooney et al. / Bioresourcereactor volume in both reactors was thus controlled by aconductivity sensor which activated a peristaltic euentpump when the liquid reached the tip the of the sensor.In all experiments, in the rst (hydrogen-production) reac-tors the conductivity sensor was set to a height that main-tained a working volume of 2.0 L while in the second(methane-production) reactors the height was set to aworking volume of 15.0 L.</p><p>The feed medium was held in a polyethylene tank with acapacity of 114.0 L, maintained at 5 C with an internalaluminum coil heat exchanger (EX11, Aquatic Eco-systemsInc., Apopka, FL), connected to an external, setpoint con-trolled cryostat (Ultratemp, 2000, Julabo Labortecknic,Germany) and mixed with a magnetic stir bar. Mixing inboth reactors was deliberately low to better reect condi-tions of industrial-scale systems, where high mixing ratesnormally used in experimental bioreactors are not applica-ble. The feed tubings were replaced daily or every 2 days, asneeded. Nitrogen gas was bubbled continuously throughthe medium to help maintain an anaerobic environment.The feed media was delivered to the reactors by peristalticpumping through Teon tubing (101U/R, WatsonMar-low Ltd., Cornwall, England). The pumps were calibratedand their feed rates periodically veried gravimetrically.</p><p>Temperature and pH in both reactors were controlledwith a MicroDCU unit (B. Braun Biotech Inc., Allentown,PA) and continuously logged through a RS232 serial inter-face connected to an external Dell 4100 computer (Dell,Texas USA). Temperature was measured in situ using PT-L </p><p>T</p><p>pH</p><p>T </p><p> pH </p><p>P </p><p>P L</p><p>action of feed media or effluentthe second CH4-production</p><p>ribed in Table 2. </p><p>H2 Rxr </p><p>CH4 Rxr</p><p>o as the H2-production reactor while the second stage is referred to as the</p><p>hnology 98 (2007) 26412651 2643100 probes and the pH with gel-based probes (Mettler-Toledo, Greifensee, Switzerland) and maintained at35 C 0.1 C in both reactors using 200 W and 450 Wcartridge heaters (McMaster-Carr, Los Angeles, CA), inthe rst and second reactors, respectively. pH was con-trolled at 5.5 in the rst reactor and at 7.0 in the secondreactor through automated addition of sodium hydroxide.The concentration of base was 2.0 M, but was increased to4.0 and 6.0 M during the experiments using the highfeed concentrations. The total base consumed was calcu-lated from the recorded on-time of the base pump andthe base consumption rate was calculated by leastsquares regression of the data with respect to time over a60 min s interval.</p><p>Data was calculated in 1 or 2 min intervals and loggedusing Lab View. For each steady state, rates (gas evolution,base addition) were averaged over a 24 h period that dem-onstrated consistent data and did not exhibit mechanical orother interruptions (foaming, clogged lines, lack of mixing,or power outages). Steady state was assumed to have beenreached after at least two full residence times (i.e. reactorsvolumes) had passed and the gas evolution rate (estimatedover a 2024 h period) had steadied to the same value over3 days.</p><p>2.2. Media, inoculum, and start-up</p><p>The standard media used for the batch and continuousexperiments consisted of (per liter): 10.0 g glucose, 1.5 g</p></li><li><p>TecKH2PO4, 1.67 Na2HPO4, 0.5 NH4Cl, 0.18 MgCl 6H2O,2.0 g yeast extract, 2.0 g peptone, 0.02 ml antifoam A(Sigma). In some experiments this media was modied byincreasing the feed glucose concentration or by doublingthe concentration of peptone and yeast extract per liter(termed 2N as opposed to 1N for the standard medium)as described. The carbon to nitrogen ratio calculated foreach media formulation was calculated by summing thegrams of carbon and nitrogen (per liter) provided by theglucose, ammonium chloride, peptone, and yeast extract.The contribution of nitrogen and carbon (per liter) frompeptone (15.4% N, 31.5% C) and yeast extract (18% Nand 32% C) was estimated by summing the contributionsof carbon and nitrogen from the components (i.e., aminoacids, casein, etc.) in each as per the compositional analysisprovided by the manufacturers. The C:N mol mol1 ratioof the standard medium was 7.59.</p><p>Anaerobic digester sludge was collected from the HawaiiKai Waste Water Treatment Plant (Hawaii Kai, HI) andstored at 4 C until 5.0 ml of the well-mixed sludge wasused to innoculate both reactors. No pre-treatment (e.g.,heating to help select for spore-forming anaerobes) wasapplied, to better reect larger-scale processes where self-selecting bacterial populations would likely dominate aftera relatively short initial period. To establish the methano-genic culture, the second reactor was initially operatedindependently of the rst, by directly feeding it standardmedia at increasingly greater ow rates. Once a methanevolume percentage of 50% was achieved, euent fromthe rst hydrogen-production reactor was used to slowlyreplace the second reactor media feed, eventually achievinga fully integrated operation of a two-phase anaerobic...</p></li></ul>


View more >