Biological electrochemical systems (BESs) have the potential for decentralized treatment in developing countries. A 46 L, two-chamber, hydraulically partitioned microbial fuel cell (MFC) was designed to replicate low-flow scenarios leaving a composting toilet. The co-evolution of electricity and methane in this MFC was evaluated by testing two distinct waste streams: synthetic feces (Case F) and municipal primary effluent (Case W). Oxidation of organic matter was 76 ± 24% during Case F and 67 ± 21% during Case W. Methanogenesis was dominant in the anode, yielding potential power of 3.3 ± 0.64 W/m3 during Case F and 0.40 ± 0.07 W/m3 during Case W. Electrical power production was marginal, Case F = 4.7 ± 0.46 and Case W = 10.6 ± 0.39 μW/m3, although potentially useful in energy-limited areas. Complimentary batch cultivations with anode inocula yielded greater methane production in the presence of graphite. 74 ± 11% more methane was produced with graphite than suspended growth enrichments and 58 ± 10% more than enrichments with non-conductive plastic beads. The co-production of methane and electricity in an MFC may have utility in decentralized treatment. Further work is needed to optimize power from both electricity and methane.
The majority of people who lack access to sanitation systems live in developing countries (WHO/UNICEF 2015). As of 2015, only 68% of the world population used an improved sanitation facility (WHO/UNICEF 2015), and the majority of people who do practice open defecation live in rural areas. Although the most widely used systems are ventilated improved pit latrines, efforts have been made to develop incentivized sanitation systems. Examples of incentivized systems include composting latrines that produce a natural fertilizer and anaerobic digesters to capture methane gas for heating and cooking purposes (Mihelcic 2009; Surendra et al. 2014). Although anaerobic digesters and biogas toilets seem feasible in promoting sustainable means for energy recovery, high nitrogen species present in anthropogenic wastewaters have been documented to inhibit the anaerobic microbial degradation process (Fricke et al. 2007).
Bioelectrochemical systems (BESs) and, more specifically, microbial fuel cells (MFCs) have generated significant interest for energy-efficient or energy-yielding wastewater treatment approaches. MFCs decouple the electron donor and the electron acceptor, allowing for anaerobic organic degradation by microorganisms using the anode as an electron acceptor. Electrons are transported to a cathode via an electrical load, where the reduction of an electron acceptor occurs. Castro et al. (2014) previously designed and implemented a unique three-chamber MFC system that oxidized organic matter from feces in the anode chamber and utilized a biological cathode to reduce nitrate-rich effluent from a nitrification chamber fed urine. It was designed for retrofitting composting latrines. The MFC latrine incentivizes sanitation by producing compost, electricity, and treated effluent water as the three main outputs. Although the system produced all three products, electricity recovery was low.
One explanation for the reduced power production at larger scale MFC is alternative, anaerobic microbial metabolisms in the anode compartment. Some anode-respiring bacteria (ARB), like many anaerobic, chemotrophic bacteria, prefer to oxidize acetate because of its low oxidation state (−1) (Thauer et al. 1989). However, the degradation of complex organic matter requires hydrolyzation of large organic macromolecules, followed by fermentation, producing simple organic acids and hydrogen. Symbiotic relationships between ARB and fermenters have been linked to efficient conversion of organic substrates to electricity (Parameswaran et al. 2009). In contrast, methanogenesis is often cited as a barrier to full-scale implementation of MFC technologies, as methanogens and ARB are often in direct competition for end products of fermentation. Methane has been observed to be a major contributor to inefficiencies at the anode even when simple substrates, such as glucose and sucrose, are used (He et al. 2006; Jung & Regan 2011).
Methanogens have recently been documented to work synergistically with ARB. In the presence of conductive surfaces, direct interspecies electron transfer (DIET) has been reported between known anode-respiring species, Geobacter sp., and the methanogens Methanosaeta sp. and Methanosarcina sp. (Liu et al. 2012; Zhao et al. 2015). Furthermore, enhanced methane production has been documented in anaerobic digester aggregates in the presence of conductive material such as graphite (Morita et al. 2011). This suggests that the conductivity of the anode could offer a pathway to divert electrons from electricity production in favor of methanogenesis.
Although not appropriate for all MFC applications, co-evolving methane with electricity is practical for developing countries where methane could be used for heating or cooking and electricity could be stored and used for lighting. Nearly three billion people still use cooking methods that involve burning locally available biomass such as firewood, animal excreta, and kitchen waste, producing harmful indoor air pollution (Surendra et al. 2014). MFCs employed to generate both biogas and electricity at ambient temperatures could allow for on-demand electricity production and storable energy as methane.
This study evaluates the potential for methane production in MFC anode communities and the co-evolution of methane and electricity in a lab-based pilot MFC that is a one-to-one representation of the MFC paired with a composting latrine in Ghana (Castro et al. 2014). Methane production, electricity production, and treatment performance were evaluated for two different wastewater conditions: synthetic feces and municipal wastewater. Microbial enrichments from the operating pilot-scale MFC anode were incubated under ambient conditions to explore the role of conductive material on methane production. This BES delivers two additional outputs that a conventional anaerobic digester or biogas toilet cannot produce: direct electricity production and nitrogen removal.
MFC construction and startup
Batch enrichment studies
To assess enhanced methane production at room temperature in the presence of conductive surfaces, enrichment studies were performed with anode effluent from the pilot MFC. In the enrichment studies, acetate served as the carbon source in a 16 mM phosphate buffer solution. Enrichment bottles were prepared in the presence of graphite granules, non-conductive plastic beads, or with no attachment surface present (suspended). The study was conducted in 12 sealed 150 mL serum bottles, with five replicate cultivations for bottles containing graphite granules, duplicate cultivations for bottles containing plastic beads and five replicates with no additional attachment surfaces present (suspended growth). Each bottle was capped and purged with nitrogen gas prior to inoculation. The growth media consisted of a 16 mM phosphate buffer containing 1,000 mg/L of acetate and 1 mL/L of a calcium-iron and trace mineral solutions. All bottles were autoclaved and cysteine (31.5 mg/L) was added to remove any residual oxygen. All bottles were covered in foil to prevent phototrophic growth, incubated at 22 °C, and continuously shaken for 31 days.
Simultaneous electricity production and methanogenesis was evaluated using the pilot MFC reactor while treating two types of organic wastewaters: synthetic feces wastewater (Case F – 54 day operation), and municipal wastewater (Case W – 50 day operation). Case F media consisted of 8.0 g of starch, 2.50 g of casein, 4.34 g of KH2PO4, 1.09 g of Na2HPO4, 0.310 g of NH4Cl, 0.130 g of KCl, and 5.0 g of oleic acid per liter of reverse osmosis (RO) water (Du et al. 2011). The pH was 6.3 ± 0.003 and conductivity was 4,700 ± 150 μS/cm. For Case W, effluent was obtained from the primary clarifier of the Amherst WWTP and used as the influent to the anode. The pH was 7.2 ± 0.08 and conductivity was 640 ± 30 μS/cm. During both conditions, room temperature of 20–22 °C was maintained, unless otherwise noted. The cathode chamber was fed with nitrate in a 16 mM phosphate buffer with the following recipe: 0.710 g Na2HPO4, 1.50 g KH2PO4, 0.050 g MgSO4, and 0.605 g NaNO3 per liter of RO water. All anode and cathode media were purged with nitrogen gas for at least 30 minutes before being introduced into their respective compartments. Media was pumped continuously, with a hydraulic retention time (HRT) of 8 days for the anode chamber and 4 days for the cathode chamber to reflect operation when connected to a composting latrine superstructure similar to a ventilated improved pit latrine (Castro et al. 2014). Both the anode and cathode biofilms were established prior to operation under both conditions, and the MFC has been in operation for the past four years. At the completion of each wastewater scenario, visual inspection of the anode and cathode was performed by opening each of the chambers. During Case F, minimal precipitated starch at the top layer was removed when visible but the interior remained intact so as to not disturb the biofilms around the granular graphite electrodes.
Samples were collected from the inlets and outlets of the anode and cathode. All samples were filtered through 0.45 μm syringe filters prior to analysis. An ion chromatograph (850 Metrohm) was used to measure nitrite and nitrate. Chemical oxygen demand (COD) was measured according to standard methods using Hach kits (Hach Method 8000). An Agilent gas chromatograph (GC) (7890A model) was used to measure the following short-chain fatty acids (SCFAs): acetic, propionic, isobutryic, n-butyric, isovaleric, n-valeric, isocaproic, n-caproic, and heptanoic acids (standards from Matreya LLC, Pleasant Gap, PA). Liquid samples for SCFA analysis were filtered with a 0.45 μm syringe filter and acidified with 6 N of sulfuric acid for large sample quantities, or 12 N of HCl for small sample quantities, before analyzing. The GC was also used to measure methane and carbon dioxide gases using an HP-PLOT-Q column. Gas samples for methane were collected from the headspace of the anode chamber and stored in gas-tight bags before analysis. Duplicate injections were made for each sample.
Voltage production was monitored using a Keithley Model 2700 Multimeter with a 7700 Switching Module (Keithley Instruments Inc., Cleveland, OH, USA). Readings were collected every 10 minutes across the external resistance. Polarization curves for determining internal resistance were conducted by linear sweep voltammetry (LSV) using a Gamry Series G750 Potentiostat/Galvanostat/ZRA (Gamry, USA). LSV was run for three cycles at a scan rate of 1 mV/s from zero to the open circuit potential. Current was determined using Ohm's law, I = V/R, where I is the current in amps (A), V is the voltage in volts (V), and R is the resistance in ohms (Ω). Power was determined using P = I2R, where P is power in watts (W). Power densities were normalized to the anode surface area or anode liquid volume, where specified, and reported with standard errors. Reported power density from the literature was converted to the appropriate units using the ratio of reported anode surface area to anode liquid volume.
Mass balances and power production
RESULTS AND DISCUSSION
Methanogenesis within communities enriched from the anode
Increased methane production from anode microbial communities in the presence of graphite granules has implications for most MFC applications. Methane is well documented as a competing metabolism for anode-respiring communities, but the results from this study suggest that anode environments may be even more favorable to methanogenesis than conventionally thought. For optimization of energy production in MFC technologies, there are two approaches to consider. Methanogenesis could be suppressed in the anode to increase electricity production, which may be appropriate in large-scale industrial or municipal applications. Methane could also be captured as a co-evolved, value-added product, which may be useful for applications in developing areas where resources are limited and simplified reactor design and operation are required.
COD removal and nitrate reduction in the MFC
In the cathode chamber, nitrate removal was observed in all cases. Nitrate removal was greater with high organic loading rates at the anode, observing 53 ± 16% removal in Case F and only 12 ± 6.9% in Case W. The anode and cathode were hydraulically linked in this MFC configuration and effluent COD, not oxidized in the anode, was allowed to flow to the cathode, resulting in heterotrophic denitrification. This served as a COD ‘polishing’ step. For Case F, a further 9.6 ± 4.9% of anode influent COD was removed in the cathode chamber, for a total COD removal in the system of 85 ± 33%. For Case W, an additional 14 ± 6.7% was removed, for a total COD removal of 86 ± 13%. Power production was lower than expected in all cases, and minimal autotrophic denitrification occurred in the cathode chamber from current delivery to the cathode.
Electrical power production in the pilot-scale MFC
Power production from electricity was observed and sustained during both media treatment cases. Operational power outputs for Case F and Case W were 4.7 ± 0.46 and 10.6 ± 0.39 μW/m3. Polarization curves were used to determine the internal resistances and maximum power densities for both cases (42 kΩ for Case F and 214 kΩ for Case W). Case F had the greatest concentration of organic matter entering the system, at 3,800 mg COD/L, while yielding the lowest average operational power output. Low power output is frequently noted in bench-scale reactors treating complex waste streams. Starch processing waste has yielded 1.4 W/m3 (240 mW/m2) (Lu et al. 2009), 4.3 W/m3 (170 mW/m2) from swine wastewater (Min et al. 2005), and 1.74 W/m3 (26 mW/m2) from domestic wastewater (Liu et al. 2004). Recent large-scale applications of MFCs for wastewater treatment and sanitation purposes have only achieved moderate power output even when utilizing numerous stacked or sequential liter-scale MFCs (Alzate-Gaviria et al. 2016; Ge & He 2016; Ieropoulos et al. 2016). In this study, large external resistances were used to match the internal resistances of the MFC system and to maximize the voltage drop across the resistor to support an LED light and to produce methane. The large external resistances may have limited the anode's availability as an electron acceptor to ARB, which could lead to increased methanogenesis (Jung & Regan 2011).
SCFA and methane production in the MFC anode
Other microbial metabolisms in the anode were investigated. Dissolved and headspace methane gas accounted for 8.4% of electrons obtained from the oxidation of COD in Case F and 18.9% in Case W. An accumulation of SCFAs was detected in Case F and accounted for a total of 3.1% of the electrons: acetic (1.1%), propionic (0.70%), isobutyric (0.16%), n-butyric (0.24%), isovaleric (0.34%), n-valeric (0.17%), isocaproic (0.11%), n-caproic (0.11%), and heptanoic acid (0.14%). SCFAs were below the detection limit in Case W. The major SCFAs found in the effluent of the anode for case F were acetic, propionic, and n-butyric acid, suggesting that large polysaccharides derived from starch were broken down to various sugar forms such as maltose and glucose and further fermented to the simplest SCFAs. Upon inspection of the anode chamber at the end of Case F, a minor amount of starch had precipitated in the anode and contributed to COD removal. Since it was difficult to separate the small amount of precipitated starch from other volatile solids in the packed bed of graphite granules, we were unable to quantify the accumulation or its contribution to COD removal; although qualitative inspection leads us to believe is was not a significant component of COD removal.
Methane as an alternative energy source
For large-scale applications of MFC technology, methane production seems inevitable without active suppression of methanogenesis, especially when treating wastewaters with multiple complex organics. The energetic value from an MFC is typically measured by the maximum electrical energy it can produce. Although electrical energy is lowered by competition with methanogens (Torres et al. 2007), an alternate perspective includes methane as a value-added product, particularly in developing areas where biogas can be used as a cooking and heating fuel.
Enhanced methane production in the presence of graphite granules was demonstrated within mixed cultures enriched from anode communities. A pilot-scale MFC produced energetic products in the form of methane and electricity when treating complex wastewaters. The MFC was able to reliably remove greater than 70% of organics when treating two different wastewater streams: synthetic feces wastewater and municipal wastewater. Nitrate removal was also consistently observed in the cathode chamber, and was primarily due to heterotrophic denitrification. Energetic products resulting from anode respiration and methanogenesis were simultaneously observed, yielding a total energy production of 3.3 ± 0.64 W/m3 for Case F and 0.40 ± 0.07 W/m3 for Case W.
This study considers the potential for co-evolving methane and electricity in a single anode treating different strength wastewaters in a pilot MFC mirroring the operation parameters of a previous field demonstration (e.g., low and infrequent usage). For application in small rural communities, future work will evaluate the effects operational parameters have on treatment performance and energy recovery including HRT, temperature, and organic loading. Future work will also explore the major metabolic pathways for complex organic degradation to balance electrical power and methane production at the pilot scale. Simple and low-cost designs for pilot MFCs, coupled with optimization of energy recovery from all potential sources, is key to bringing this wastewater treatment technology to fruition for underserved communities.
This research was based upon work supported by the Bill & Melinda Gates Foundation (Grand Challenges Explorations Round 7) and the National Science Foundation under Grant No. S121000000211. The municipal wastewater operation was completed as part of a collaboration with Michael White of FTL Labs and Air Force SBIR Phase I Contract No. FA8501-14-P-0032.