pure.southwales.ac.uk · web viewthe mppt control box was constructed to interface and drive the...
TRANSCRIPT
Date: May 27th 2016 For resubmission to Energy Technology (ente.201600191R1)
Application of maximum power point tracking to increase the power production and treatment efficiency of a continuously
operated flat-plate microbial fuel cell
Young Eun Song[a], Hitesh C. Boghani[b], Hong Suck Kim[c], Byung Goon Kim[c],
Taeho Lee[d], Byong-Hun Jeon[e], Giuliano C. Premier[b], Jung Rae Kim[a]*
[a]School of Chemical and Biomolecular Engineering, Pusan National University, Jangjeon-Dong, Geumjeong-gu, Busan, 46241, Korea
[b]Sustainable Environment Research Centre (SERC), Faculty of Computing, Engineering and Science, University of South Wales, Pontypridd, RCT, CF37 1DL, UK
[c]The MFC Research and Business Development (R&BD) Center, K-water Institute, Jeonmin-Dong, Yuseong-Gu, Daejeon, 34045, Korea
[d]Department of Civil and Environmental Engineering, Pusan National University, Busan, 46241, Korea
[e]Department of Natural Resources and Environmental Engineering, Hanyang University, Seoul, 04763, Korea
Corresponding Author:Dr. Jung Rae Kim
Tel: +82 51 510 2393Fax: +82 51 510 3943
Email address: [email protected]
1
123456789
10111213
141516171819202122232425262728293031323334353637383940
Graphical abstract
2
41
42
Highlights
1. A Maximum power point tracking (MPPT) method based on Boolean logic was developed
2. A MPPT control was applied to a continuously operating flat-plate MFC (FPM)
3. The MPPT maximized the power density and bioelectrochemical activity of FPM
4. Direct and in-situ control of microbial fuel cell system was demonstrated
3
43
44
45
46
47
48
49
Abstract
A logic-based maximum power point tracking (MPPT) and LabVIEWTM interface for
digitally controlled variable resistive load were developed and applied to a continuously
operating flat-plate microbial fuel cell (FPM). The interaction between the designed MPPT
algorithm and electrochemically active microbial performance on the electrode was
demonstrated to track the maximal performance of FPM system. MPPT could dynamically
derive the optimal performance from varied operating conditions of FPM such as organic
concentration, flow rate and sampling interval, and produce a maximum power density of
88.0 Wm-3. The results could provide essential information to build an automatic control
strategy to achieve the maximum performance from field scale microbial fuel cells for
applications to sustainable bioenergy recovery from various biomass feedstocks.
Keywords
Flat-plate MFC; Microbial fuel cells; Maximum power point tracking (MPPT); bioelectrochemical
system; process optimization
4
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
Introduction
A microbial fuel cell (MFC) is a bioreactor that can generate electricity directly using
electrochemically active microorganisms from a variety of biologically degradable organic
materials and/or wastewater [1]. Biologically catalyzed oxidation and reduction reactions
provide the potential difference to drive electron transfer from the anode to the cathode
through a (frequently resistive) external load, thus generating electrical power [2]. MFCs are
dynamic systems because the performance of the biocatalyst (i.e. bacteria) and hence the
wider state of the cell is affected significantly by the operating environment and conditions,
such as the internal and external impedances that apply to the MFC reactor, substrate
concentration, pH, and temperature. The electroactive microorganisms adapt to the reactor
state, and changes their metabolic pathways dynamically to optimize the survivability under
those conditions. On the other hand, the microbial activity also influences the cell state and
operating conditions [3].
A power density of up to 144 Wm-3 was shown to be feasible at a realistic volume of
20 L in a MFC reactor [4], whereas Yang et al. [5] reported that a miniature MFC reactor
(working volume of 30 mL) designed to minimize distance between the electrodes and reduce
power loss generated 2.08 kWm-3. Various techniques and strategies have previously been
employed in order to improve the cell voltage and power of a MFC system such as capacitor
based circuit system [6]. The electrochemically active microorganisms of the MFC interact
continuously with the electrode. Thus the power density and stability of the system can be
improved further by controlling the interaction between microbes and the electrode. At a
relatively steady state, the internal and external load resistance of the MFC, are critical
parameters for obtaining the maximum power density.
5
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
Conventionally, a fixed external load resistance has been used during enrichment for
an electrochemically active microbial community in the MFC, and also during normal
operation. On the other hand a fixed resistance may limit the power production due to
impedance mismatch without considering the dynamicity of MFC system. The maximum
performance such a system requires active control of the dynamic interactions between the
microbial metabolism and the electrode.
To maximize the MFC performance, various dynamic load shifting systems have
been developed and investigated to obtain the optimal power production using the real-time
monitoring and control [7]. Most of these systems use maximum power point tracking (MPPT)
methods that continually seek the maximum power point (MPP) of a MFC by automatically
adjusting the external resistive load using control algorithms, such as the incremental
conductance (IC) and the perturb and observe (PO) method [8]. The IC algorithm is based on
the slope of the differential of power versus potential (or current) of the power density curve,
and is intended to maintain a zero slope, thus to obtain peak power on MPP [9] (see subfigure
in Fig. 1). MPPT can detect a potential difference between the anode and cathode, and
regulate the external resistance to obtain the maximum power using a recursive algorithm. A
dynamic regulation of the load resistance by MPPT controls the MFC system to achieve a
higher performance than what is normally obtained from the fixed load resistance.
Premier et al. developed an MPPT algorithm for MFCs using a circuit board
incorporating digitally controllable resistance, which was modified from the previously
known MPPT algorithm used for solar cell [7c]. As a result, an MPPT-controlled MFC
obtained fivefold higher coulombic efficiency and power density in a two-chamber MFC,
than those obtained from a control without MPPT operation (i.e. under a fixed load
resistance). In addition, they reported that the MPPT algorithm could affect the bacterial
6
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
community of the biofilm on the electrode. Boghani et al. reported that MPPT or an
additional poised potential applied with MPPT could improve the enrichment of
electrochemically active microorganisms on the electrode [7a]. Therefore, the start-up time to
near steady state electricity generation could be reduced. A power harvesting using a
maximum power point circuit (MPPC) maintained the stable operation of a recirculating-flow
MFC for more than 18 h [10]. The energy harvested from the MPPC was 76.8 J, which was 76
times higher than an alternative charge pump power management system (PMS), while the
columbic efficiency with MPPC was 21 times higher. Alaraj et al. introduced an MFC
applicable power management system (PMS) that combines MPPT method and power
boosting/harvesting system [7d]. Park et al. also developed a PMS with a DC/DC booster
converter and hysteresis controller for parallel operation of a MFC with simultaneous energy
harvesting [11]. MPPT reduced start-up time and improved coulombic efficiency and energy
loss [12]. On the other hand, the system produced only approximately 5 Wm -3, and the effects
(or even limitations) of the MPPT on producing higher power densities are unknown.
Most previous studies for MPPT focused on the development of a circuit design and
implementation into a small scale MFC reactor. In particular, in the MFC, which use
microorganisms as the dynamic biocatalyst, the interaction between the externally regulated
resistance by the MPPT circuit, and electroactive microbial performance was not highlighted
in the previous studies. The dynamic change in power, voltage and resistance in the MFC
systems have not been investigated extensively under continuously operating systems with
MPPT. Therefore, it is difficult to extrapolate to a scaled-up continuous system, and to
employ the information in the design of a field scale application in flat plate configurations.
This paper reports the continuous operation of a flat-plate microbial fuel cell (FPM)
using an MPPT control strategy. The interaction between the designed MPPT algorithm and
7
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
electroactive microbial performance on the electrode was demonstrated to track the maximal
performance of FPM system. The effects of the organic loading rate (ORL), hydraulic
retention time (HRT) and sampling interval (SI) of FPM operation on the dynamic regulation
of power, were investigated extensively in long term operation [13]. The change in the
bioelectrochemical performance of the anode associated electroactive strains under MPPT
was studied by cyclic voltammetry.
Results and Discussion
MPPT control on different organic loading rate
The substrate concentration is an important parameter that affects the electricity
generation in MFCs. To determine the effect of MPPT operation on the dynamic change from
the influent organic concentration, the synthetic wastewater varying concentration of 0 to 20
mM acetate was supplied to the MPPT controlled FPM at 30 min HRT (5.0 mL min -1) (Fig.
2a). Each time the influent concentration was changed, the MPPT start-up resistance was
readjusted to 44.2 Ω which was selected as a conditional optimal resistance for FPM in a
separate test (Fig. 2b).
The MPPT algorithm gradually increased the load resistance with the lowest influent
acetate concentration of 0 mM while the output voltage remained negligibly low. The MPPT
algorithm recognizes the near zero voltage and power, and increases the resistance stepwise
to a maximum under the starving conditions. Similarly, when the influent concentration was
changed to 1 mM (0.13 g L-1h-1), the cell voltage increased to approximately 100 mV then
decreased slowly (Fig 2b). The power produced under MPPT operation was unsustainable
due to the low organic loading rate.
8
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
An appreciable voltage and power were produced with 2 mM of acetate (OLR of
0.25 g L-1h-1) from day 3. MPPT stabilized the external load resistance at approximately 100
Ω and the voltage at approximately 0.5 V while generating a power density of 25.3 to 33.6
Wm-3. The maximum power density achieved was 60 Wm-3 at 10 mM (OLR of 1.25 g L-1h-1)
(Fig. 2c). The load resistance was stabilized to 5 Ω during MPPT operation. A further
increase in the influent acetate concentration (20 mM) did not show a noticeable change in
the voltage, but load resistance slightly increased from 4.8 to 7.4 Ω. The power decreased
gradually probably due to saturation followed by substrate inhibition and/or pH unbalance on
the electrode surface resulted from mass transfer limitation. These results suggest that MPPT
control algorithm recognize the current voltage and power production, and continuously track
the optimal power producing condition by changing the load resistance, even when the
inhibitive and/or destructive operating condition are applied.
Acetate consumption and removal efficiency were investigated with different
influent acetate concentrations in Fig. 3. The acetate consumption increased at higher influent
concentration, but the removal efficiency decreased. The MPPT control resulted in
significantly higher acetate removal efficiency than those of the FPM with the fixed load
resistance (FLR). The results implicate that the dynamic regulation of the load by MPPT
enhances the acetate uptake by the anode microbial consortium. For example, the removal
rate of acetate and the efficiency under MPPT operation were 0.43 gL -1 and 68 %, compared
to 0.07 gL-1 and 11 % for the FLR, when provided with 0.63 gL-1h-1 OLR (influent acetate
concentration of 5 mM). The estimated CE with MPPT was higher than the FLR in the range
of acetate concentration tested (See inset figure in Fig. 3b). In continuous operation, the
electron recovery with MPPT was 33 % compared to FLR (~7 %) at 10 mM of influent
9
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
acetate concentration.
To verify the improvement of oxidation and reduction characteristics of the anode
electrode during continuous operation, CV was conducted on FPM-MPPT and FLR, while
being supplied with fixed acetate concentrations of 1 to 20 mM (OLRs of 0.13, 0.25, 0.63,
1.25, and 2.38 g L-1h-1) (see Supplementary information Fig S2a and b). The CV results
showed that a low influent acetate concentration results in low capacitance and peak current
in the system, thus weak oxidation-reduction characteristics were obtained. The
electrochemical characteristics changed more dramatically in the case of FPM-MPPT,
especially when the influent acetate concentration was increased. A higher oxidative peak
current and capacitance (area between oxidative and reductive scan) were obtained from
FPM-MPPT than from FPM-FLR. Accordingly, the peak current markedly increased with
increasing OLR.
Effect of HRT on MPPT operation
The effect of HRT was also tested with various influent flow rates using the
peristaltic pump to adjust HRT. The influent acetate concentration was selected to be 10 mM
(1.25 gL-1) based on the results from the previous OLR test with MPPT. The HRT was
decreased gradually in steps, from 1 h to 0.25 h. Each HRT was run for approximately 2 days
before the next change in HRT (Fig 4a). The MPPT control decreases the load resistance
from 110 Ω to 10 Ω during 2 days of 1 h HRT operation (Fig 4b). Consequently, the power
density increased from less than 20 Wm-3 to approximately 80 Wm-3. A further decrease in
HRT to 15 min (which was the highest flow rate to the FPM) did not improve the apparent
power density from the FPM significantly (Fig 4c). The difference in the effect of MPPT was
10
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
difficult to identify at HRTs lower than 45 min based on the apparent power density.
The electrochemical performance of FPM-MPPT and FPM-FLR induced by changes
in HRT was investigated further by CV (Fig 5). The oxidation peaks of FPM-MPPT in the
forward scan reached 0.12, 0.18, 0.30, 0.30, and 0.26 A for 60, 45, 30, 20, and 15 min HRTs,
respectively (Fig 5a) [14]. The maximum oxidation peak of FPM-MPPT was 0.3 A at 20 and
30 min HRT, which are relatively higher than those of FPM-FLR (0.1 A) at the same HRTs
(Fig 5a). The higher anode electrode capacitances were obtained between 20 and 30 min of
HRTs, while 20 min HRT (12.3 C) was slightly higher than 30 min of HRT (11.3 C). Note
that the electrochemical effect of HRT was clearer in FPM-MPPT compared to FPM-FLR in
all tested HRTs (Fig 5b). The improved electrochemical performances implicate that the mass
transport rate between bulk liquid and electrode surface is enhance at low HRT (i.e. fast flow
rate).
Overall, the MPPT operation showed an oxidation peak current and capacitance
between 2 to 3 times higher than those of the FLR operation. Both MPPT and FLR operation
show similar apparent performance at low flow rates (i.e. high HRT). On the other hand, the
difference was more distinct in the higher flow rate operation (low HRT). These results
suggest that MPPT control provides a platform to improve the anodic biofilm performance in
MFCs while the FLR does not provide conditions to reflect the dynamic characteristics of
MFCs.
The effect of MPPT on the acetate consumption at each HRT operation was
described in Fig S3. The level of acetate consumption differed greatly between FPM-MPPT
and FPM-FLR once the HRT exceeded 30 min and up to 60 min. In particular, the FPM-
11
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
MPPT consumed 5 times as much acetate as the FPM-FLR at 30 min HRT. On the other
hand, the shorter HRT in the reactor did not show appreciable differences in acetate
consumption, probably due to limitations of the residence time within the reactor. The MPPT
operation also clearly improved the bacterial substrate uptake rate and removal capacity,
especially when higher HRTs were applied.
MPPT operation with different sampling interval (SI)
Microbial fuel cells are time varying systems, in which the whole cell microbial
catalyst adapts to the operating condition, optimizing their metabolic activities to the
environment [7c]. The sampling interval (SI) defines the discrete time series at which the
MPPT control algorithm actuates the external load and adjusts the current demand from the
MFC. The SI also represents the interval between the time at which the voltage and power
from the system are sampled, and setting the new load by MPPT. The time required for the
execution of the monitoring and control algorithm are negligible (usually measured in
milliseconds) compared to the parameter of SI. When MPPT is applied to the MFC system,
therefore, SI may have a significant effect on the dynamic performance of the biocatalyst on
the electrode, and consequently the performance of the MFC. To support this hypothesis,
FPM-MPPT was tested with sampling times of 1, 3, 5 and 10 min at a acetate concentration
of 10 mM and HRT of 20 min (at 1.88 gL-1h-1) (Fig. 6).
All SIs tested tracked the optimal power generation point, as shown in Fig. 6a and b;
however, this was at different rates of convergence, resulting in dynamic trends due to the
increased delay with sampling time from the gradient driven increment or decrement in the
external load. The 1 min of SI shows rapid tracking within 20 min under MPPT operation and
produces stable power generation at 65.2 Wm-3. The longer SIs (3 to 10 min) showed a
12
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
relatively slower response; however, the final power density obtained under MPPT operation
reached a marginally higher peak power (i.e., 87.4 Wm-3 with 10 min SI) compared to 1 min
SI (Fig. 6c). The 5 min sampling time resulted in a relatively stable output voltage and power.
This result indicates that the anode bacterial community responds to the new
electrochemical environment (i.e. dynamic load control by MPPT). Therefore, different
sampling rates affects the output voltage and power density in the case of the MPPT
algorithm applied in this study. The MPPT system by their nature, will oscillate near the
convergence point (i.e. MPP). The slowing down of the tracking by changing the SI will also
slow down to achieve convergence. A step change in load will cause a slow rise/fall in the
voltage from MFC. The combined effect may be momentary oscillations with higher peak
that appears to occur in the case of 5 and 10 min of SI. If the system was allowed to stabilize
until the next change in load, the power from FPM would have settled to the levels that the
microbial community on the anode adapt to the condition.
MPPT operation with progressively decreasing organic feed concentration
From the optimized parameters of OLR, HRT and SI, it is reasonable to select
notionally near-optimal operational conditions to be 10 mM acetate, 20 min HRT (7.5 ml
min-1) and 5 min of SI under MPPT control. Fig. S4 in the Supplementary information shows
the voltage and power changes during a period when the influent concentration was
decreasing gradually from 10 mM acetate. MPPT initially traced the maximum power
generation point for 6 h and stabilized to 55 Wm-3 at 5.4 Ω, despite the acetate concentration
decreasing to half of its initial concentration. If the influent concentration was maintained at
10 mM, it is expected that the power density would have been equal to or higher than 88.0
Wm-3. A sharp decrease in voltage and power was observed at 14 h of operation by the
13
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
progressively decreasing acetate concentration. The MPPT tracked a seemingly different
peak power point from 18 hrs. This suggests that the MPPT acted to resist a power drop-off,
which resulted from the prevailing environmental conditions (i.e. the diminishing acetate
concentration in this case). A further decrease in influent acetate concentration resulted in an
increased load resistance, as might be expected because the MPPT tracked the optimal
diminishing peak power point towards the starvation condition.
Implication of MPPT control strategy for microbial fuel cell
The time variant MFC process is affected continuously by its environment (substrate
concentration, HRT, Temp and pH), as in any conventional bio-system. In addition, as the
MFCs use an anode respiring bacterial community, which are mostly attached to the
electrode, the interaction between microbe and electrode is a very important factor
determining the performance and stability of the system. The external resistance and/or
applied (or derived) potential at the electrode influences the performance significantly. MPPT
can regulate these MFC specific parameters (i.e. external load resistance and hence electrical
output sourced from the biocatalyst) through the use of a control algorithm to adjust the
external load automatically. These results successfully show that the MPPT can interact
dynamically with the operational parameters induced by the external environment, and derive
near-maximal performance from a continuously operating MFC system.
The effect of external parameters, such as OLR, HRT and SI were evaluated and
compared with respect to the performance of a FPM using MPPT and FLR. The MPPT
operation in particular achieved a higher power density and acetate removal efficiency
compared to the operation with FLR (at 100 Ω). The maximum acetate removal efficiency at
varying HRT and OLR operations was approximately 70% in performance from the
14
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
conditions of 5 mM acetate concentration and 30 min HRT under MPPT control. This is a
much higher benefit of MPPT operation because a high COD removal efficiency would have
generally required a much longer HRT with FLR. Jadhav and Ghangrekar [15], in similar flat-
plate system with a similar influent concentration on FLR, obtained a COD removal
efficiency of 70% but this was achieved using 24.55h HRT.
Another advantage of MPPT control is its ability to stabilize the output voltage
despite the fluctuations induced by the external environment, such as variation of OLR, HRT
and temperature. The stabilization of the output voltage/power will improve the quality of
electrical energy derived from the MFCs. These results also suggest that the microbial
community may adapt to the live cell-electronic circuit hybrid system, and allow stabilization
to a momentary equilibrium condition induced by the MPPT. In overall, MPPT in MFC may
act as a feedback controller and enhance biofilm formation on the electrode as it was shown
in the different external resistances [16]. Further study will be needed to validate the hypothesis
by microscopic investigation.
This strategy highlights an alternative approach from the previous optimization
typically employed in bio-processes and MFC systems. Using MPPT, the product yield
and/or treatment efficiency of the bioelectrochemical process can be regulated directly by an
electrode, and optimized automatically to approach the best achievable performance. Further
studies will be needed to develop an optimal strategy for the regulation of the anodic
community in such a time varying MFC system and an appropriate power harvesting circuit
for practical applications.
Conclusions
15
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
The performances of continuously operating flat plate microbial fuel cell under
MPPT (FPM-MPPT) and with a fixed load resistance (FPM-FLR) were compared. Different
operational parameters, such as acetate concentration, HRT, and SI were investigated for
voltage and power deployment during MPPT operation. The MPPT control strategy could
regulate the MFC’s performance and improve the power significantly compared to the FLR
operation, where the MFC was operated conventionally with a static external load. The
maximum power density obtained under MPPT operation was 88.0 Wm-3 at 5.4 Ω with 5 min
SI (10 mM acetate and 20 min HRT) in continuous operation.
Experimental Section
Flat-plate MFC configuration and startup
A flat-plate MFC (FPM) configuration consisted of two MFCs connected in parallel
that share wastewater influent through a serpentine flow passage. The anode was constructed
using carbon felt (150 mm ⅹ 200 mm NARA Cell-tech Co., Korea), and the cathode (150
mm ⅹ 200 mm) was prepared by applying four diffusion layers (air side) and a platinum
catalyst (0.5 mg Pt cm-2; liquid side) on a 20 wt. % wet-proof carbon cloth (NARA Cell-tech
Co., Korea) with 5 % Nafion solution (DE521, NARA cell-Tech Co., Korea) as a catalyst
binding agent [1a]. The continuously operating FPM reactor had a total volume of 150 mL. An
acrylic frame held an ion exchange membrane (Nafion 117, Dupont, Del. USA) between two
electrodes (anode and cathode). The assembled FPMs were inoculated with the anaerobic
secondary digester sludge (10% v/v) from the wastewater treatment plant (Suyeong WWT
Plant, Busan, Korea). The synthetic wastewater consisted of 50 mM phosphate buffer (pH
7.0) with CH3COO·Na; 3.28 gL-1, NH4Cl; 0.23 gL-1, NaCl; 0.04 gL-1, MgSO4 · H2O; 0.01 gL-
1, KCl; 0.02 gL-1, Yeast Extract; 0.02 gL-1. When the FPMs were inoculated and the operation
16
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
was started in batch mode, a 1000 Ω external load was applied. The inoculated FPMs were
placed in an incubator (SW-90S, Sangwoo, KOREA) at 30 °C. After an enrichment period of
7 days, the reactors were switched to continuous mode.
Two FPM reactors were constructed and operated; FPM with a fixed load resistance
(FLR) (FPM-FLR) and FPM with MPPT (FPM-MPPT) connected to the MPPT control, load
and interface box described below. The FPM-FLR was connected to a 100 Ω load, which was
selected as the optimal static resistance determined from a separate power curve measurement
on the FPM. The FPMs were normally operated at a HRT of 30 min (5.0 mL/min) using a
peristaltic pump (77521-50, Master Flex, USA) (See Supplementary information Fig. S1).
Design of MPPT algorithm and control box
The MPPT system consisted of a software algorithm to implement the MPPT logic
flow chart, as shown in Fig. 1, and a hardware control box to interface and implement the
algorithm (Fig. S1). The MPPT was based on a modified IC (Incremental conductance)
algorithm to control the external load shifting that is different from the conventional MPPT
algorithm, using a PC equipped with LabVIEW™ (National Instruments™, USA). The MFC
type IC algorithm can track the maximum power point by the differential of the power over
voltage value. The IC algorithm recognizes the sign of the differential of power over
potential. When the differential is positive, it indicates that the present power is located on
the lower than the MPP in the power density curve (power vs. voltage curve), thus the MPPT
control shifts to the forward by stepping up the resistance (i.e. increasing the load resistance
to increase the voltage); when the differential is negative, the algorithm decreases the
resistance for moving the power to the backward (i.e. decreasing the load resistance to reduce
17
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
the voltage) (See Fig. 1) [17]. The load was changed at the sampling interval (SI) set within the
time window of the execution of the monitoring and control software written in LabVIEW™.
The MPPT control box was constructed to interface and drive the load resistance
control circuit and data acquisition (DAQ) board. NI USB-6009 (National Instruments™,
USA) data acquisition board was employed for the monitoring and control of MPPT (See
Supplementary information Fig. S1c). The NI USB-6009 has eight analog input channels,
two analog outputs and 12 digital I/O channels. Each MPPT operating channel uses two
analog inputs, one analog output and three digital I/O channels from NI USB-6009. The load
resistance control circuit board was constructed using digitally controlled potentiometers,
Intersil® X9C102 (Future Electronics Inc., Canada) and power MOSFET IC relays
International Rectifier PVN102 (Future Electronics Inc., Canada). The X9C102
potentiometer consisted of a three-wire serial interface, 100 wiper tap points, an array
composed of 99 resistive elements and a wiper switching function driven by the serial
interface. The PVN012 consists of a single-pole, open solid-state relay.
Investigation of operating conditions
The effect on the system performance of three operational parameters were
investigated, OLR, HRT, and the SI. Before the start of the experiment, the FPMs were
operated for more than a year on synthetic wastewater, which included 40 mM acetate
supplied by a peristaltic pump (flow rate of 5.0 mL/min). The experiment that varied OLR
was performed by increasing the influent acetate concentration from 1, to 2, 5, 10, and 20
mM, which were equivalent to organic loading rates of 0.13, 0.25, 0.63, 1.25 and 2.38 g L -1h-
1, respectively. The experiment in which the HRT was varied, was carried out by gradually
changing the influent flow rate from 2.5, to 3.3, 5.0, 7.5, and 10 mL min -1, which
18
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
corresponded to HRTs of 60, 45, 30, 20, and 15 min while using 10 mM acetate. The biofilm
on the anode electrode was expected to adapt dynamically to the operational parameters
investigated, as well as to the applied external load induced by MPPT. The SI under MPPT
operation was investigated to discern any significant influence on the voltage and power
production performance of the FPM. Therefore, MPPT control was investigated at different
SI of 1, 3, 5, and 10 min.
Analyses and calculations
The voltage across the load was monitored, and taken using a computer based data
logging system (LabVIEW™, National Instruments™), as reported previously [7c]. The liquid
sample for the acetate concentration from the influent and effluent were taken regularly and
analyzed by high-performance liquid chromatography (HPLC) (1260 Infinity, Agilent
Technology, USA) equipped with a 300 mm × 7.8 mm Aminex HPX-87H column (Bio-Rad,
USA) at 65 °C using 2.5 mM H2SO4 as the mobile phase [18]. Pretreatment of the sample was
conducted by centrifuging at 10,000g for 10 min, and filtering through a nylon filter (0.22
μm, Korea biotech, Korea). Acetate removal efficiency was calculated to be RE = [(Acetate in
– Acetateout)/ Acetatein] × 100, where RE is the removal efficiency (%), Acetate in is the
influent acetate and Acetateout is the effluent acetate concentration. The coulombic efficiency
(CE, %) was calculated to be CE = Ct/Cth × 100, where Ct is the total coulombs that are
calculated by integrating the current over time, and Cth is the theoretical coulombs from the
acetate consumed. An investigation of the oxidation and reduction characteristics of a system
with its bacterial anode electrode was carried out by cyclic voltammetry (CV) using a
potentiostat (VersaSTAT3, AMETEK, USA). The scan rate was 25 mVs-1 with the scan range
from -0.8 to 0.8V vs. Ag/AgCl using Versa Studio™ software (AMETEK, USA). The peak
19
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
current (Imax) indicated maximum oxidation current, and the capacitance was obtained from
the area of the integrated CV using Versa Studio™ software.
Acknowledgments
This study was supported by the Mid-Career Researcher Program (2013069183)
through the National Research Foundation of Korea (NRF), and the MFC Research and
Business Development (R&BD) center co-funded with K-water, Hanhwa E&C and Taeyoung
E&C, and was also supported by the NERC RRfW MeteoRR project, UK (NE/L0 14106/1).
20
424
425
426
427
428
429
430
431
432
5. References
[1] a) S. Cheng, H. Liu, B. E. Logan, Electrochemistry Communications 2006, 8, 489-494; b) B. E. Logan, B. Hamelers, R. Rozendal, U. Schröder, J. Keller, S. Freguia, P. Aelterman, W. Verstraete, K. Rabaey, Environ. Sci. Technol. 2006, 40, 5181-5192; c) K. Rabaey, W. Verstraete, Trends Biotechnol. 2005, 23, 291-298.
[2] a) B. E. Logan, Nature Reviews Microbiology 2009, 7, 375-381; b) K. Rabaey, G. Lissens, W. Verstraete, Biofuels for fuel cells: renewable energy from biomass fermentation 2005, 377-399; c) S. T. Oh, J. R. Kim, G. C. Premier, T. H. Lee, C. Kim, W. T. Sloan, Biotechnol. Adv. 2010, 28, 871-881.
[3] a) H. Wang, J.-D. Park, Z. J. Ren, Environ. Sci. Technol. 2015, 49, 3267-3277; b) J. R. Kim, Y. E. Song, G. Munussami, C. Kim, B.-H. Jeon, Geosystem Eng. 2015, 1-8.
[4] A. Dekker, A. T. Heijne, M. Saakes, H. V. Hamelers, C. J. Buisman, Environ. Sci. Technol. 2009, 43, 9038-9042.
[5] F. Yang, D. Zhang, T. Shimotori, K.-C. Wang, Y. Huang, J. Power Sources 2012, 205, 86-92.
[6] a) I. Ieropoulos, C. Melhuish, J. Greenman, I. Horsfield, J. Adv. Robotic Sys. 2005, 2, 295-300; b) G. Papaharalabos, J. Greenman, A. Stinchcombe, I. Horsfield, C. Melhuish, I. Ieropoulos, J. Power Sources 2014, 272, 34-38.
[7] a) H. C. Boghani, J. R. Kim, R. M. Dinsdale, A. J. Guwy, G. C. Premier, Bioresour. Technol. 2013, 140, 277-285; b) L. Woodward, B. Tartakovsky, M. Perrier, B. Srinivasan, Biotechnol. Prog. 2009, 25, 676-682; c) G. C. Premier, J. R. Kim, I. Michie, R. M. Dinsdale, A. J. Guwy, J. Power Sources 2011, 196, 2013-2019; d) M. Alaraj, Z. J. Ren, J.-D. Park, J. Power Sources 2014, 247, 636-642.
[8] a) N. Degrenne, F. Buret, B. Allard, P. Bevilacqua, J. Power Sources 2012, 205, 188-193; b) H. C. Boghani, G. Papaharalabos, I. Michie, K. R. Fradler, R. M. Dinsdale, A. J. Guwy, I. Ieropoulos, J. Greenman, G. C. Premier, J. Power Sources 2014, 269, 363-369.
[9] D. Hohm, M. E. Ropp, Progress in photovoltaics: Research and Applications 2003, 11, 47-62.
[10] H. Wang, J.-D. Park, Z. Ren, Environ. Sci. Technol. 2012, 46, 5247-5252.
[11] J.-D. Park, Z. Ren, Energy Conversion, IEEE Transactions on 2012, 27, 715-724.
[12] D. Molognoni, S. Puig, M. D. Balaguer, A. Liberale, A. G. Capodaglio, A. Callegari, J. Colprim, J. Power Sources 2014, 269, 403-411.
[13] X. A. Walter, S. Forbes, J. Greenman, I. A. Ieropoulos, Sustain. Energy Technol. Assess. 2016, 14, 74-79.
[14] J. R. Kim, S. H. Jung, J. M. Regan, B. E. Logan, Bioresour. Technol. 2007, 98, 2568-2577.
[15] G. Jadhav, M. Ghangrekar, Bioresour. Technol. 2009, 100, 717-723.[16] L. Zhang, X. Zhu, J. Li, Q. Liao, D. Ye, J. Power Sources 2011, 196,
6029-6035.21
433
434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479
[17] A. Safari, S. Mekhilef, Ind. Elect. IEEE Trans. 2011, 58, 1154-1161.[18] S. Ashok, S. M. Raj, Y. Ko, M. Sankaranarayanan, S. Zhou, V. Kumar,
S. Park, Metabol. Eng. 2013, 15, 10-24.
22
480481482
Figure Captions
Fig 1. Maximum power point tracking (MPPT) control system flow chart. V: voltage, P:
power, R: Resistance.
Fig 2. Effect of the transition of organic loading rate with MPPT ( 30℃, pH7 ) on the
continuously operating FPM, (a) Step changes of OLR; 1mM acetate: 0.13 gL-1h-1, 2mM:
0.25 gL-1h-1, 5mM: 0.63 gL-1h-1, 10mM: 1.25 gL-1h-1, 20mM: 2.38 gL-1h-1, (b) voltage and load
change by MPPT, (c) power density under MPPT control.
Fig 3. Effect of MPPT and FLR operation on acetate removal for influent acetate
concentrations of 1, 2, 5, 10, and 20 mM, respectively (organic loading rate of 0.13, 0.25,
0.63, 1.25, and 2.38 g L-1h-1, respectively) (a) acetate consumption; (b) acetate removal
efficiency and coulombic efficiency (inset figure).
Fig 4. Effect of HRT on the power density and voltage when under MPPT control. Here, the
influent acetate concentration was 10 mM. (a) Step changes in HRT, (b) voltage and load
change by MPPT, (c) power density under MPPT control.
Fig 5. Comparison of the peak current and capacitance estimated for MPPT and FLR
operation: (a) peak current and capacitance estimated by coulomb area (b) comparison of
cyclic voltammetry of MPPT and FLR for low (20 min) and high HRT (60 min).
Fig 6. Effect of the different sampling interval (SI) on the applied external load and power
production under MPPT control (30 ℃, pH 7, 10 mM acetate, 20 min HRT: 1.88 gL-1h-1 ), (a)
voltage change according to different SI, (b) load change by MPPT, (c) power density under
MPPT control.
23
483
484485
486487488489
490491492493
494495496
497498499
500501502503
504
505
Figure 1
Fig 1. Maximum power point tracking (MPPT) control system flow chart. V: voltage, P:
power, R: Resistance.
24
506
507508
509510511512
Figure 2
Fig 2. Effect of the transition of organic loading rate with MPPT ( 30℃, pH7 ) on the
continuously operating FPM, (a) Step changes of OLR; 1mM acetate: 0.13 gL-1h-1, 2mM:
0.25 gL-1h-1, 5mM: 0.63 gL-1h-1, 10mM: 1.25 gL-1h-1, 20mM: 2.38 gL-1h-1, (b) voltage and load
change by MPPT, (c) power density under MPPT control.
25
513
514
515516517518519
Figure 3
Fig 3. Effect of MPPT and FLR operation on acetate removal for influent acetate
concentrations of 1, 2, 5, 10, and 20 mM, respectively (organic loading rate of 0.13, 0.25,
0.63, 1.25, and 2.38 gL-1h-1, respectively) (a) acetate consumption; (b) acetate removal
efficiency and coulombic efficiency (inset figure).
26
520
521
522523524525
Figure 4
Fig 4. Effect of HRT on the power density and voltage when under MPPT control. Here, the
influent acetate concentration was 10 mM. (a) Step changes in HRT, (b) voltage and load
change by MPPT, (c) power density under MPPT control.
Figure 5
27
526
527528
529530531532533
Fig 5. Comparison of the peak current and capacitance estimated for MPPT and FLR
operation: (a) peak current and capacitance estimated by coulomb area (b) comparison of
cyclic voltammetry of MPPT and FLR for low (20 min) and high HRT (60 min).
28
534
535536537
Figure 6
Fig 6. Effect of the different sampling interval (SI) on the applied external load and power
production under MPPT control (30 ℃, pH 7, 10 mM acetate, 20 min HRT: 1.88 gL-1h-1 ), (a)
voltage change according to different SI, (b) load change by MPPT, (c) power density under
MPPT control.
29
538
539540
541542543544
545