Energy Balance Affected by ElectrolyteRecirculation and Operating Modes in Microbial

Fuel CellsKyle S. Jacobson1, Patrick T. Kelly1, Zhen He2*

ABSTRACT: Energy recovery and consumption in a microbial fuel cell

(MFC) can be significantly affected by the operating conditions. This

study investigated the effects of electrolyte recirculation and operation

mode (continuous vs sequence batch reactor) on the energy balance in a

tubular MFC. It was found that decreasing the anolyte recirculation also

decreased the energy recovery. Because of the open environment of the

cathode electrode, the catholyte recirculation consumed 10 to 50 times

more energy than the anolyte recirculation, and resulted in negative

energy balances despite the reduction of the anolyte recirculation.

Reducing the catholyte recirculation to 20% led to a positive energy

balance of 0.0288 kWh m3. The MFC operated as a sequence batchreactor generated less energy and had a lower energy balance than the

one with continuous operation. Those results encourage the further

development of MFC technology to achieve neutral or even positive

energy output. Water Environ. Res., 87, 252 (2015).

KEYWORDS: microbial fuel cells, wastewater treatment, energy,

recirculation, sequence batch reactor.

doi:10.2175/106143015X14212658613235

IntroductionMicrobial fuel cells (MFCs) are an emerging technology for

wastewater treatment with a potential to be an energy-efficient

treatment system (Logan et al., 2006). In the past decade,

intensive research has been conducted in the laboratory in

various aspects of MFCs, such as microbiology, electrode

materials, membrane materials, reactor configuration and

operation, substrates, and electrochemistry (Arends and Ver-

straete, 2012; Logan and Rabaey, 2012). Although no successful

pilot tests were reported (Logan, 2010), MFCs have been

demonstrated at a liter scale treating actual wastewater with

promising results during a long-term operation outside the

laboratory (Zhang et al., 2013b).

The primary function of MFCs designed for wastewater

treatment is recognized as contaminant removal, rather than

energy recovery (He, 2013), but energy performance is also an

important feature for establishing a successful treatment system.

Energy performance involves both energy recovery and energy

consumption. Energy recovery is represented by electricity

generation from organic contaminants through microbial

interaction with electrodes. To properly present energy recovery

and translate the research results into a language that

wastewater industry can understand, a new parameter was

proposed, called normalized energy recovery (NER). Normalized

energy recovery can be described in two units, kWh m3 basedon the volume of the treated wastewater, and kWh kg COD1

based on the amount of the removed chemical oxygen demand

(COD) (Ge et al., 2014). Energy consumption in an MFC is

mainly due to the energy requirement by the pumping system

(to feed the reactor and mixing), and aeration (if applied).

Understanding energy performance of an MFC is important to

reveal the knowledge gap for bringing MFC technology to an

energy-neutral or even positive system.

Energy performance of MFCs has been reported in a few

recent studies. For example, in a photo-bioelectrochemical

system consisting of an MFC and an algal bioreactor, the total

theoretical energy recovery from both electricity and algal

biomass was larger than the energy consumption by the

pumping system, and the produced electric energy in the MFC

was typically less than 20% of the total energy recovery (Xiao et

al., 2012). Aeration was identified as the largest contributor to

energy consumption, and produced the negative energy balance

(Zhang et al., 2013a; Zhang and He, 2012). Excluding the

aeration energy, a theoretically positive energy balance could be

achieved in an osmotic MFC (Ge, Ping, Xiao, and He, 2013).

Aeration could be eliminated by an appropriate design of the

cathode process, for example, by the use of passive air supply by

exposing the cathode electrode directly in the air. This design

has helped to achieve positive energy balances in both laboratory

and field examination of different types of MFCs (Ge, Ping, and

He, 2013; Zhang et al., 2013b).

Those previous studies indicate that MFCs have potential to

become an energy-neutral treatment process and encourage

further development of this technology. However, there is lack of

systematic investigation of energy performance of MFCs. The

authors have previously studied the effects of the MFC

dimensions and substrates on energy recovery, and found that

energy recovery did not decrease with increasing MFC size, and

domestic wastewater could be a suitable substrate for MFC

operation (Xiao et al., 2014). Herein, the authors expand the

investigation to the effects of the operating conditions in two

1 Department of Civil Engineering and Mechanics, University ofWisconsinMilwaukee, Milwaukee, Wisconsin.2 Department of Civil and Environmental Engineering, VirginiaPolytechnic Institute and State University, Blacksburg, Virginia.

* Department of Civil and Environmental Engineering, VirginiaPolytechnic Institute and State University, Blacksburg, Virginia24061; Telephone: (540) 231-1346; fax: (540) 231-7916; e-mail:zhenhe@vt.edu

252 Water Environment Research, Volume 87, Number 3

aspects: recirculation rates and operating modes (continuous vs

sequence batch reactor).

Materials and Methods

Microbial Fuel Cell Setup. The tubular MFC used in this

study was made similarly to those in the prior studies (Kelly and

He, 2014; Xiao et al., 2014), with an effective membrane length

of 50 cm and a diameter of ~5 cm, yielding an anode liquidvolume of approximately 1.1 L (Figure 1). The membrane tube

was constructed of cation exchange membrane (CEM) (Mem-

branes International, Ringwood, New Jersey). The anode

electrode was a 50-cm long carbon fiber brush, and connected

by using copper and titanium wires to the carbon cloth that

wrapped the membrane tube as the cathode electrode. Activated

Figure 1Schematic diagram and laboratory prototype of the tubular MFCs.

Figure 2Polarization results (voltage and power) of the MFC with three anolyte recirculation rates. The arrows indicate voltage (left)and power (right).

Jacobson et al.

March 2015 253

carbon powder was used as the cathode catalyst at a loading rate

of 7 mg cm2 and was coated to the carbon cloth withpolytetrafluoroethylene (PTFE) (DuPont, Wilmington, Dela-

ware). The external resistance was set at 15 X, which was closeto the internal resistance of the MFCs determined by

polarization tests.

Microbial Fuel Cell Operation. The MFC was inoculated

with a mixture of the effluent from an activated sludge tank and

the sludge from an anaerobic digester in a water resource

recovery facility (South Shore, Milwaukee, Wisconsin). The

anode was fed with a synthetic solution containing (per liter of

tap water): sodium acetate, 0.2 g; NH4Cl, 0.15 g; NaCl, 0.5 g;

MgSO4, 0.015 g; CaCl2, 0.02 g; KH2PO4, 0.53 g; K2HPO4, 1.07 g;

and trace element, 1 mL (Angenent and Sung, 2001). The anode

effluent was used as a catholyte, which rinsed the cathode

electrode. The MFC was operated under either continuous flow,

or sequence batch reactor (SBR). The continuous operation had

hydraulic retention time (HRT) of 14.1 hours. Under the SBR

operation, the anode solution was replaced by 80% with the

following procedure resulting in an HRT of 11.3 hours: the

feeding time (~6 minutes), reaction time (~635 minutes),settling time (30 minutes), and decant time (~6 minutes). Theanolyte recirculation rate was 150 mL min1, which wasdemonstrated to be effective for tubular MFCs with a similar

size (Zhang et al., 2010). The catholyte was recirculated at 50 mL

min1. The effects of the recirculation rates on energy balance

Figure 3The energy performance of the MFC affected by: (A) anolyte recirculation, and (B) catholyte recirculation. Energyconsumption consists of the energy required for anolyte recirculation and catholyte recirculation. Energy balance is the differencebetween energy recovery and the sum of the recirculation energy (consumption). Error bars were based on at least triplicatemeasurements.

Jacobson et al.

254 Water Environment Research, Volume 87, Number 3

were examined with the continuous operation. Three anolyte

recirculation rates were studied: 100% (recirculation pump was

on all the time), 33% (recirculation pump was operated with a

cycle of on for 10 minutes and off for 20 minutes), and 16%

(recirculation pump was operated with a cycle of on for 5

minutes and off for 25 minutes). The catholyte recirculation

was examined at two rates: 100 and 20% (recirculation pump

was operated with a cycle of on for 0.5 minutes and off for 2

minutes). During varying the anolyte recirculation rates, the

catholyte recirculation was maintained as 100%, and vice versa

when adjusting the catholyte recirculation.

Measurement and Analysis. The MFC voltage was recorded

every 1 minute by a digital multimeter (Keithley Instruments,

Inc., Cleveland, Ohio). The polarization curves were performed

by a potentiostat (Gamry Instruments, Warminster, Delaware) at

a scan rate of 0.1 mV s1. The power and current densities werecalculated based on the anode liquid volume. The concentra-

tions of COD were measured using a colorimeter (Hach,

Loveland, Colorado). Energy consumption was assumed to be

mainly due to the recirculation, and estimated according to a

previous study (Kim et al., 2011). The NER was calculated

according to a previous study (Xiao et al., 2014).

Results and DiscussionRecirculation of electrolyte helps to better distribute sub-

strates and improve ion transport, and thus is important to MFC

performance; on the other hand, recirculation is also a major

energy consumer in a tubular MFC system. Three anolyte

recirculation rates, 100, 33, and 16%, were applied to the MFC,

and the results confirmed the important role of recirculation in

electricity generation. According to the polarization curves

(Figure 2), the 100% anolyte recirculation resulted in the highest

maximum power density of 9.33 W m3, much higher than 6.16

W m3 (33%) and 4.17 W m3 (16%). The actual difference ofpower production among the three conditions was smaller,

because the polarization tests conducted by using a potentiostat

could inflate the power output due to a short testing time.When

the MFC was operated with external resistance for an extended

period, the MFC with 100% anolyte recirculation produced a

maximum power density of ~5.50 W m3; the actual poweroutput of the MFC with 33% anolyte recirculation was

approximately 4.94 W m3, while the one with 16% anolyterecirculation was close to the one obtained from the polarization

test.

Figure 3A shows the energy performance of the MFC affected

by the anolyte recirculation. The NER was 0.07706 0.0027 kWhm3 (100%), 0.0693 6 0.0031 kWh m3 (33%), and 0.0615 60.0017 kWh m3 (16%). The amount of energy consumption bythe anolyte recirculation was 0.0010 kWh m3 (100%), 0.0033kWh m3 (33%), and 0.0017 kWh m3 (16%). Although thecatholyte recirculation (50 mL min1) was much slower than theanolyte recirculation (150 mL min1), its energy consumptionwas 10 to 50 times higher than the anolyte recirculation, mainly

because the cathode configuration was an open environment,

and the hydraulic head loss was significantly higher than that of

the anolyte recirculation (which was a closed condition). Due to

this large energy consumption by the catholyte recirculation, the

overall energy balance of the MFC with three anolyte

recirculation rates was all negative. It seemed that higher

anolyte recirculation resulted in less negative energy balance

because of relatively higher energy production; however, the

difference of the energy balance among the three conditions was

not significant. Those results demonstrated that in the present

MFC system, the anolyte recirculation had a minor effect on

energy balance. Thus, the catholyte recirculation was examined

in the following.

Figure 4Current generation (current density) in the MFC operated continuously (Continuous) or as a sequence batch reactor(SBR).

Jacobson et al.

March 2015 255

Two catholyte recirculation rates, 100 and 20%, were tested.

The energy requirement by the 20% catholyte recirculation was

much lower at 0.0185 kWh m3 than the 100% catholyterecirculation (Figure 3B). The power output with the 20%

catholyte recirculation also became lower at ~4.0 W m3,resulting in a lower NER of 0.0573 6 0.0016 kWh m3. Theoverall energy balance with the 20% catholyte recirculation was

positive at 0.0288 kWh m3, while the 100% catholyterecirculation led to a negative balance of 0.0254 kWh m3.Clearly, the catholyte recirculation is a major energy consumer

in the present MFC system; despite the decreased energy

production at lower catholyte recirculation, the reduced demand

for energy consumption benefits the overall energy balance of

the MFC system.

In addition to electrolyte recirculation, the SBR operation of

the MFC was examined and compared with the continuous

operation. Sequence batch reactor operation is used in

wastewater treatment, but seldom adopted in MFC operation

(Peixoto et al., 2013). In this test, the anolyte recirculation rate

was 16% and the catholyte recirculation rate was 20%. Current

generation with the SBR operation was similar to that of a

typical batch operation (Figure 4). The highest current (or

power) was similar to the current (or power) with the

continuous operation, but the NER with the SBR operation

was 0.0358 6 0.0052 kWh m3, lower than 0.0489 6 0.0024kWh m3 with the continuous operation. The energy consump-tion of the electrolyte recirculation with the SBR operation was

also lower (0.0016 kWh m3 for the anolyte recirculation and0.0174 kWh m3 for the catholyte recirculation) compared withthose of the continuous operation (0.0017 kWh m3 for theanolyte recirculation and 0.0184 kWh m3 for the catholyterecirculation), likely because of no recirculation during the

periods of feeding and decanting (Figure 5). Because the

difference in the energy of the recirculation was less significant

than that of NER, the overall energy balance of the SBR

operation (0.0157 kWh m3) was still lower than that of thecontinuous operation (0.0288 kWh m3). Additionally, it wasfound that the COD removal with the continuous operation

(87.86 3.9%) was higher than that with the SBR operation (65.26 5.2%). A shorter HRT of the SBR operation (3 hours indifference) could contribute to lower COD removal efficiency.

Those results suggest that the continuous operation of an MFC

may be more advantageous than the SBR operation in terms of

energy performance.

ConclusionsThis study demonstrates that the catholyte recirculation was a

major energy consumer in the present tubular MFC system. The

open environment of the cathode electrode created a high

hydraulic head loss, resulting in significant energy requirement.

Although reducing the rate of catholyte recirculation decreased

energy recovery, the low energy consumption led to a positive

energy balance. The energy requirement by the anolyte

recirculation played a minor role in the total energy consump-

tion. The MFC could be operated as a sequence batch reactor;

however the SBR operation had lower energy recovery and

overall energy balance than the continuous operation.

AcknowledgmentThis work was financially supported by a grant from Research

Growth Initiative (RGI) at the University of Wisconsin

Milwaukee and faculty startup fund at Virginia Polytechnic

Institute and State University.

Submitted for publication March 12, 2014; accepted for

publication June 26, 2014.

ReferencesAngenent, L. T.; Sung, S. (2001) Development of Anaerobic Migrating

Blanket Reactor (AMBR), a Novel Anaerobic Treatment System.

Water Res., 35, 17391747.

Figure 5The energy performance of the MFC affected by the operating mode. Energy consumption consists of the energy required foranolyte recirculation and catholyte recirculation. Energy balance is the difference between energy recovery and the sum of therecirculation energy (consumption). Error bars were based on at least triplicate measurements.

Jacobson et al.

256 Water Environment Research, Volume 87, Number 3

Arends, J. B. A.; Verstraete, W. (2012) 100 Years of Microbial Electricity

Production: Three Concepts for the Future. Microbial Biotech., 5,

333346.

Ge, Z.; Li, J.; Xiao, L.; Tong, Y.; He, Z. (2014) Recovery of Electrical

Energy in Microbial Fuel Cells. Environ. Sci. Technol. Lett., 1, 137

141.

Ge, Z.; Ping, Q.; He, Z. (2013a) Hollow-Fiber Membrane Bioelectro-

chemical Reactor for Domestic Wastewater Treatment. J. Chem.

Technol. Biotechnol., 88, 15841590.

Ge, Z.; Ping, Q.; Xiao, L.; He, Z. (2013b) Reducing Effluent Discharge and

Recovering Bioenergy in an Osmotic Microbial Fuel Cell Treating

Domestic Wastewater. Desalination, 312, 5259.

He, Z. (2013) Microbial Fuel Cells: Now Let Us Talk About Energy.

Environ. Sci. Technol., 47, 332333.

Kelly, P. T.; He, Z. (2014) Understanding the Application Niche of

Microbial Fuel Cells in a Cheese Wastewater Treatment Process.

Bioresour. Technol., 157, 154160.

Kim, J.; Kim, K.; Ye, H.; Lee, E.; Shin, C.; McCarty, P. L.; Bae, J. (2011)

Anaerobic Fluidized Bed Membrane Bioreactor for Wastewater

Treatment. Environ. Sci. Technol., 45, 576581.

Logan, B.; Rabaey, K. (2012) Conversion of Wastes into Bioelectricity and

Chemicals Using Microbial Electrochemical Technologies. Science,

337, 686690.

Logan, B. E. (2010) Scaling Up Microbial Fuel Cells and Other

Bioelectrochemical Systems. Appl. Microbiol. Biotechnol., 85,

16651671.

Logan, B. E.; Hamelers, B.; Rozendal, R. A.; Schroder, U.; Keller, J.;

Freguia, S.; Aelterman, P.; Verstraete, W.; Rabaey, K. (2006)

Microbial Fuel Cells: Methodology and Technology. Environ. Sci.

Technol., 40, 51815192.

Peixoto, L.; Rodrigues, A. L.; Martins, G.; Nicolau, A.; Brito, A. G.; Silva,

M. M.; Parpot, P.; Nogueira, R. (2013) A Flat Microbial Fuel Cell for

Decentralized Wastewater Valorization: Process Performance and

Optimization Potential. Environ. Technol., 34, 19471956.

Xiao, L.; Ge, Z.; Kelly, P.; Zhang, F.; He, Z. (2014) Evaluation of

Normalized Energy Recovery (NER) in Microbial Fuel Cells

Affected by Reactor Dimensions and Substrates. Bioresour. Technol.,

157, 7783.

Xiao, L.; Young, E. B.; Berges, J. A.; He, Z. (2012) Integrated Photo-

Bioelectrochemical System for Contaminants Removal and Bioen-

ergy Production. Environ. Sci. Technol., 46, 1145911466.

Zhang, F.; Ge, Z.; Grimaud, J.; Hurst, J.; He, Z. (2013a) Improving

Electricity Production in Tubular Microbial Fuel Cells through

Optimizing the Anolyte Flow with Spiral Spacers. Bioresour.

Technol., 134, 251256.

Zhang, F.; Ge, Z.; Grimaud, J.; Hurst, J.; He, Z. (2013b) Long-Term

Performance of Liter-Scale Microbial Fuel Cells Installed in a

Municipal Wastewater Treatment Facility. Environ. Sci. Technol., 47,

49414948.

Zhang, F.; He, Z. (2012) Integrated Organic and Nitrogen Removal with

Electricity Generation in a Tubular Dual-Cathode Microbial Fuel

Cell. Process Biochem., 47, 21462151.

Zhang, F.; Jacobson, K. S.; Torres, P.; He, Z. (2010) Effects of Anolyte

Recirculation Rates and Catholytes on Electricity Generation in a

Liter-Scale Upflow Microbial Fuel Cell. Energy Environ. Sci., 3,

13471352.

Jacobson et al.

March 2015 257