pure.southwales.ac.uk€¦ · web viewsources of co 2 include but are not limited to industrial...
TRANSCRIPT
Methanogenic Capacity and Robustness of Hydrogenotrophic Cultures based on Closed
Nutrient Recycling via Microbial Catabolism: Impact of Temperature and Microbial
Attachment
Savvas Savvasa*, Joanne Donnellya, Tim P. Pattersona, Zyh Siong. Chongb and Sandra. R.
Estevesa
aWales Centre of Excellence for Anaerobic Digestion, Sustainable Environment Research
Centre,University of South Wales, Pontypridd CF37 1DL, Wales UK.
bEngineering Research Centre, Faculty of Computing, Engineering and Science, University of South
Wales, Pontypridd CF37 1DL, Wales, UK
Abstract
A biological methanation system based on nutrient recycling via mixed culture microbial
catabolism was investigated at mesophilic (37o C) and thermophilic (55o C) temperatures. At
mesophilic temperatures, the formation of biofilms on two different types of material was
assessed. Results showed that with intense mixing the biofilm reactors presented
methanogenic capacities (per working volume) 50% higher than the ones operated with
suspended cultures. Gas feeding rates of 200 L/L/d were achieved at a H2/CO2 to CH4
conversion efficiency of above 90% by linking two reactors in series. Furthermore the
robustness of the cultures was assessed under a series of inhibitory conditions that simulated
possible process interferences at full scale operation. Full recovery after separate intense
oxygenation and long starvation periods was observed within 2-5 days.
Keywords: hydrogenotrophic methanogenesis; biofilm; power to gas; energy storage
* Corresponding author: Savvas Savvas ([email protected])
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
1. Introduction
The utilization of CO2 as a precursor for chemicals has recently started to gain growing
attention since it can add economic benefits to carbon sequestration (Jajesniak et al., 2014;
Styring et al., 2011). In particular, the Power to Methane (PtM) route as described in (Götz et
al., 2016) offers the additional advantage of directing a substantial percentage of renewable
electricity towards green fuel production. This presents an attractive solution to the storage of
excess renewably generated power. The methanation of CO2 can be performed thermo-
chemically via catalytic hydrogenation (Wang et al., 2011), however, through a biological
route known as hydrogenotrophic methanogenesis, high quality CH4 can also be produced at
ambient pressures and temperatures and without the need of metal catalysts (Lecker et al.,
2017). The process makes use of a distinctive microbial group that uses CO2 and H2 as their
carbon and energy source respectively. The group consists of a number of archaeal species
called hydrogenotrophic methanogens which are capable of working on their own (pure
cultures) or in conjunction with other archaeal and bacterial species (mixed cultures) (Liu and
Whitman, 2008).
Sources of CO2 include but are not limited to industrial combustion processes, distilleries,
cement production and waste water treatment. Among them, anaerobic digestion (AD) plants
are ideal candidates for the initial implementation of the PtM technology as they can produce
high quality CO2 without inhibitory for the microbes contaminants. Furthermore, due to its
composition (30-40% CO2 / 60-70% CH4) the biogas output from an anaerobic digester could
be directly upgraded to natural gas quality without the need for CO2 pre-separation (Martin et
al., 2013). Sources of H2 include among others natural gas reforming, gasification, water
electrolysis and a number of biological routes (U.S. Department of Energy, 2010). Water
electrolysis presents a number of advantages over other technologies namely, extra pure H2
streams, ease of coupling with renewable electricity streams and the added production of pure
2
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
O2 which can be used onsite (e.g. waste water treatment, oxy-combustion) (Global Carbon
Capture and Storage Institute, 2012; Patterson et al., 2017).
Significant research has been conducted over the last decade (Bassani et al., 2015; Burkhardt
et al., 2015; Guneratnam et al., 2017; Kougias et al., 2017; Luo and Angelidaki, 2012; Martin
et al., 2013; Seifert et al., 2014; Strübing et al., 2017; Yun et al., 2017) regarding
hydrogenotrophic methanogenesis and its potential as a continuous process. A number of
feasibility studies also indicate a good integration of the process within biogas plants
(Estermann et al., 2016; O’Shea et al., 2017; Patterson et al., 2017). Nevertheless,
commercialization of the process has yet to occur. This can partially be attributed to the
implication of the biological factor which adds a degree of uncertainty when it comes to long
operational periods and intermittent operation. The biochemical variables that directly affect
metabolic activity are influenced by the flowrate and composition of the gas entering the
system (Leonzio, 2016) and therefore the ability of hydrogenotrophic populations to deal
with variable feeds and inconsistent gas ratios is still disputable.
In a previous study (Savvas et al., 2017b) an ex-situ hydrogenotrophic reactor based on a
self-regenerating mixed microbiome under nutrient closed conditions was tested with results
showing that conversion efficiencies close to 100% were achievable at gas feeding rates of up
to 60 v/v/d. To the authors’ knowledge this type of microbiome has not been replicated
elsewhere apart from (Savvas et al., 2017a) where it was used to create biofilms in a plug-
flow hydrogenotrophic reactor. The present study went a step further by assessing the
robustness of such microbiome and its behaviour under a series of destabilising conditions
that can occur during operation of full scale plants. These were sudden changes in the gas
feeding rates, periods of carbon/energy starvation and oxygenation. Additionally the
evolution of the same inoculum under mesophilic (37o C) and thermophilic (55o C) conditions
was evaluated as well as its ability to form biofilms under conditions of intense agitation.
3
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
With the exception of (Savvas et al., 2017a), hydrogenotrophic biofilms have so far only been
assessed in trickle-bed type arrangements with results showing a lesser degree of gas
conversion rates to systems that depend on intense agitation (e.g. Continuously Stirred Tank
Reactors (CSTRs)) (Lecker et al., 2017). The difference is directly linked to the different gas-
liquid mass transfer rates that can be achieved by each system. Conversely, biofilms could
potentially add to the stability and robustness of the microbial catalyst as they have protective
for the microbes properties (Watnick and Kolter, 2000). The reactor type used in the present
study utilised intense gas-liquid mixing through liquid recirculation as a way to enhance gas
diffusion but also offered the possibility for the integration of biofilms thus creating a hybrid
system.
2. Materials and methods
2.1. Reactors and Inoculum
Four identical reactors were operated in parallel; one was kept at thermophilic conditions
(55±0.5oC) and three at mesophilic (37±0.5oC). No biofilm attachment media was used in one
of the mesophilic reactors (Reactor 1) or in the thermophilic reactor (Reactor 2). Biofilm
attachment media was used in the other two mesophilic reactors, Kaldnes K1 (polyethylene
wheels) in Reactor 3 and LECA (Light Expanded Clay Aggregate balls) in Reactor 4. The
two types of attachment media used in the present study had been previously assessed in
denitrification tests (Andersson et al., 2008) and were chosen based on their biofilm
formation performance among 20 different materials. The geometry, technical aspects as well
as the control and data acquisition parameters of the reactors were identical to the ones
described in (Savvas et al., 2017b).
All four reactors were filled with anaerobically digested mesophilic sewage sludge collected
from Cog Moors Wastewater Treatment Plant in Cardiff, South Wales, UK. Prior to
inoculation the sludge was filtered through a 125 μm stainless steel sieve. After inoculation
4
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
there was no addition of any solid or liquid feedstock for the whole of the operational period.
Thermophilic adaptation in Reactor 2 proceeded in one step as it had been previously found
to be advantageous to multi-step adaptation (Boušková et al., 2005).
2.2. Analytical methods
Gas composition was measured in real time with infra-red sensors (Premier Series 0-100%
Vol CO2/CH4 Voltage output 0.4-2.0V, Dynament Ltd) and by in-line hydrogen solid-state
sensors (H2Scan HY-OPTIMA 740, 0-100% Vol H2, 4-20mA output). The reliability of the
gas sensors was also periodically checked by analysis of the gas with a gas chromatograph
(Varian Inc., CP-4900) equipped with two columns, one for CO2 (Porapack Q, Varian – 10 m
x 0.15 mm) and one for CH4, H2, N2 and O2 (Molsieve 5A Plot, Varian – 10 m x 0.32 mm).
The carrier gas used was Ar. Gas flow rates were measured by custom made tip-meters and
logged in LabVIEWTM (National Intruments, UK).
Volatile Fatty Acids (VFAs) were determined according to (Cruwys et al., 2002) using a head
space autosampler gas chromatograph (Perkin Elmer, AutosystemXL) fitted with a flame
ionization detector and a Supelco Ltd. column (30 m x 0.32 mm). The carrier gas was N2. pH
was continuously measured with the use of pH electrodes HI-1001 (Hannah Instruments,
UK) connected to the reactors. Cations were determined with a Dionex ion chromatograph
equipped with an Ionpac CS12A separation column. 20 mM methansulfonic acid was used as
eluent. Microbial analysis was performed according to (Savvas et al., 2017b)
3. Results and discussion
3.1. Methanogenic activity and biofilm formation
5
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
Figure 1 displays the methanogenic activity of the four reactors recorded in real time during
180 days of operation. Online data logging started after day 10 in order for most of the
residual organics in the inoculum that could contribute to the production of CH4 to get
digested. Apart from a series of vertical drops in conversion efficiency which indicate periods
of no operation, technical issues and intentional/unintentional changes in the gas feeding
regime, it can be seen that all four reactors followed a similarly gradual increase in their
methanation capacity. As also described in (Savvas et al., 2017b) this was a surprising
outcome as it indicated a high level of nutrient recycling through microbial catabolism, since
no addition of nutrient media took place in any of the four reactors after inoculation. Specific
differences between the thermophilic and mesophilic microbiome are discussed later in the
text but the initial finding as presented by Figure 1 is of four hydrogenotrophic cultures, the
enrichment phases of which have been dictated by the inflow rate of gaseous feed.
Biofilm was formed in both the mesophilic reactors containing the two different types of
microbial attachment media (Kaldnes K1 and LECA). The high velocity of the fluid due to
recirculation (6 L/min) did not seem to have an adverse effect on biofilm formation, which
started being visible in both carrier materials after a period of two weeks. Irreversible
attachment continued for approximately 2 months after which there was a stable maturation
period, which lasted for the rest of the experiment. During maturation, biofilm thickness
seemed to remain unchanged with the naked eye which is a sign of continuous detachment
and renewal of colonisation. Actual measurement could not be performed as this would
require emptying the reactors’ contents.
The initial hypothesis was that microbial attachment would prevent the bulk majority of the
microbes passing through the pump, thus protecting them from the shear forces created by the
impeller. Two observations may be used to support this; firstly, due to the volume occupied
by the inert material the actual working volume occupied by the liquid media was lower in
6
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
the reactors with the attachment media (approx. 1L instead of 1.5L). This means that the
methanogenic capacity of the reactors containing the biofilm was approximately 50% greater
than the ones operated with suspended cultures. Also, due to attachment, lower levels of
biomass were lost through sampling. Secondly as will also be discussed later in the text the
ammonia concentration in the same reactors was always at lower levels than in the ones
containing the suspended culture Figure 4C which could be due to lower hydrolysis rates.
3.2. Oxygenation and starvation assessment
Unintentional or unavoidable exposure to oxygen can have damaging effects to strictly
anaerobic populations (Morozova and Wagner, 2007). Archaea are described as strict
anaerobes due to the common understanding that they lack two key enzymes namely
superoxide dismutase and catalase which are responsible for the neutralization of oxygen
radicals (Kato et al., 1997). Nevertheless, there is increasing evidence that oxygen toxicity
levels are not the same for all methanogenic species with some mesophilic and thermophilic
species displaying quite high tolerances (Botheju and Bakke, 2011). Also, as previously
reported (Botheju et al., 2010) mixed cultures have an advantage over pure cultures, firstly
due to the presence of facultative fermentative microbes that can scavenge dissolved oxygen
and secondly due to the shielding mechanism of biofilms where methanogens may have
protection behind layers that act as diffusion barriers.
To investigate the effect of oxygenation, the liquid media of all four reactors was sparged
with a gaseous mix of 20% CO2 / 80% air v/v for 14 hours before returning to a 78/22
H2/CO2 environment. Apart from the change in the gas feeding ratio all other conditions
remained unchanged. Figure 2A shows conversion capacity before and after oxygenation at a
constant gas throughput of 36 L/L/d.
7
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
As soon as oxygen was excluded from the gas stream there was an immediate exponential
upsurge in methane production for all three mesophilic reactors followed by a slower more
linear increase until conversion efficiencies returned to their previous values. The
thermophilic reactor however, seems to have experienced a lag phase of almost 24 hours
before responding. The reason behind this delay seems to be linked to the dissimilarity in
changes that occurred in the bacterial and methanogenic populations between the mesophilic
and the thermophilic culture as shown in Figure 2B. The percentage of increase in the
numbers of methanogenic populations two days after the end of aeration was 712% for the
mesophilic and only 36% for the thermophilic culture whereas total bacteria increased by
473% in the mesophilic culture and dropped by 55% in the thermophilic one.
Although there do not seem to be references in literature regarding differences in oxygen
sensitivity between thermophilic and mesophilic anaerobic cultures, the results here suggest
that not only the archaeal but also the bacterial populations of the thermophilic culture were
much more severely inhibited than their mesophilic counterparts. In fact, the difference in the
degree of inhibition is much higher than the one observed if we take into account that the
solubility of oxygen in water at 55oC is approximately 20% lower than at 37oC (Geng and
Duan, 2010).
The fast recovery rates of the mesophilic cultures are of significant importance since they do
not only add to the robustness of the system while in operation, but also simplify other
procedures related to the transfer and re-inoculation of reactors. In that sense pre-operational
enrichment can alternatively take place in a more controlled and supervised environment and
the culture transported to the operational site without the extra expenses that would be
required for strictly anaerobic handling.
A value of 0.016 d-1 has been reported for the death rates of mesophilic hydrogenotrophic
methanogens in mixed anaerobic cultures during starvation (Hao et al., 2012). In the case of
8
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
pure populations, experiments with Methanobacterium M-20 and Methanosarcina barkeri
strains, showed that cell survival rates where zero after just one month of starvation
(Morozova and Wagner, 2007). The difference can be explained by the fact that under gas
starvation conditions, mixed cultures can provide a more favourable environment for
hydrogenotrophic populations. This is due to a limited but constant supply of H2 and CO2
coming from the digestion of biomass, which however low in quantity, seems to be able to
support a quantifiable number of hydrogenotrophic methanogens. In pure cultures, as soon as
the external gas feed is discontinued, there is not a bacterial background in place to provide
this alternative H2/CO2 source.
Figures 2C and 2D show the CO2 to CH4 conversion efficiencies of all four reactors after two
starvation periods of 13 and 45 days, respectively. This was done in order to simulate
medium term maintenance stoppages of a week or two for renewable electricity infrastructure
as well as longer periods of over one month simulating a complete stop of wind in the
summer season. During these periods the gas supply was disconnected and the cultures were
kept under anaerobic conditions by ensuring that all entrances and exits to the system were
blocked. Mixing was also stopped and the reactors were kept at room temperature. The 13
day starvation period took place during the winter months whereas the 45 day starvation
period during the summer months.
After restart and under the same gas feeding rates of 36 L/L/d the recovery times for both
experiments were similar with an average period of less than 24 hours for all reactors to reach
80% conversion efficiency.
Hwang et al. (2010) studied the effects of prolonged starvation on mixed anaerobic
populations at mesophilic temperatures. After 4 months under starved conditions (without the
input of external substrates) there was complete recovery of the culture without any lag phase
for acid producers and a lag phase of about a month for methanogens. DGGE and qPCR
9
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
analysis also showed that although the concentrations of different strains fluctuated before,
during and after starvation, the microbial diversity remained unchanged with the same bands
appearing on the electrophoresis gel before and after starvation.
However, it must be noted that the substrate used after the starvation period was swine water
as opposed to the CO2/H2 substrate used in this experiment. This means the activity of
hydrogenotrophic methanogens was completely depended on the growth of their bacterial
precursors. In this experiment, due to the difference in the substrate used, the lag phase for
methanogenesis was of the order of hours which means that under the right conditions
selective enrichment of hydrogenotrophs is a fast process.
3.3. Sudden increase in the gas feeding rate
Towards the end of the experiment (after day 180) the conversion capacity of the
hydrogenotrophic culture of the two biofilm reactors (reactors 3 and 4) was tested by
supplying a high gas feeding rate of 200 L/L/d. The sudden change in the gas feeding
intensity was addressed by linking the two reactors in series which appeared to provide a
temporary conversion boost until the culture of the first reactor could adapt to the new
conditions. Figure 3 shows that with a change from 60.5 L/L/d to 200 L/L/d, conversion
efficiency could be kept at above 90% when reactor 4 started receiving the exhaust gases of
reactor 3. Additionally, the conversion efficiency of reactor 3 appeared to return to its initial
levels after approximately 10 hours, a fact that indicates an adaptation of the
hydrogenotrophic culture to the higher gassing rates. The instability in conversion
efficiencies after the connection of the two reactors (after hour 23) is a result of drifts away
from the 4:1 H2/CO2 ratio in the gas flowing out of reactor 3 and into reactor 4. Furthermore,
the high percentage of CH4 in this same gas mixture must have played a negative role in the
conversion capacity of reactor 4 through the displacement of H2 and CO2.
10
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
The experiment could not be continued for higher gassing rates due to the inability of the
specific liquid recirculation pumps to deal with higher gas volumes. However, it is
reasonably expected that with certain technical changes, a modular methanation system
capable of delivering constant conversion efficiencies at various gas inflow rates is realistic.
3.4. VFAs, pH and ammonium ions levels
In AD systems that rely on organic feedstocks a number of intermediates (e.g. volatile fatty
acids, alcohols etc.) play a crucial role on the formation of biogas and are part of syntrophic
relationships among numerous microbial species (Amani et al., 2010). The disruption of these
relationships (e.g. due to overloading or the presence inhibitors) can lead to the accumulation
of one or more of these intermediates which itself leads to digester inhibition and
underperformance (Li et al., 2012). Pure hydrogenotrophic cultures do not rely on the de-
polymerisation of organic substances and therefore carboxylic acids do not play a role on
methanogenesis in those systems.
In mixed enriched cultures, the presence of bacterial species has often shown the formation of
VFAs (Kougias et al., 2017; Luo and Angelidaki, 2012; Rachbauer et al., 2016). In gas fed
chemostats, due to the absence of organic feedstock these intermediates are kept at
insignificant levels and therefore are of no importance. In the present study however, the
closed nutrient conditions were of concern since the recycling of biomass had to pass through
the route of acidogenesis. Therefore, beside the hydrogenotrophic species, an adequate
number of VFAs oxidizing species should be kept in balance with their VFAs producing
counterparts so that the accumulation of these intermediates could be avoided.
Figure 4A shows that throughout the experiment such a balance was maintained for the three
reactors running at mesophilic conditions. Although the comparative performance in terms of
methane generation was similar for all four reactors, there was a significant difference in the
concentration of VFAs between the mesophilic and the thermophilic cultures.
11
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
Accumulation of VFAs in thermophilic processes has been reported in numerous studies
(Boušková et al., 2005; Rubia et al., 2002; Song et al., 2004) which usually is an indication
that the hydrolysis and acidogenesis steps progress faster than methanogenesis in
thermophilic temperatures in contrast to mesophilic ones. Of interest is the gradual build-up
of propionic acid in the thermophilic reactor with a value of 1400 mg/L towards the end of
operation. The degradation of propionic acid depends on the successful removal of its
products most notably H2 although formate and acetate may as well play a role (De Bok et al.,
2004). Consequently, in the AD system created by the recycling of cellular material, it is
logical to assume that the continuous injection of H2 must have disturbed the interspecies H2
transfer. If this is the case and why the effect was more profound in the thermophilic culture
is arguable. One possible explanation might be linked to the thermodynamics of hydrogen
consuming reactions which become less favorable at higher temperatures whereas hydrogen
formation becomes more favorable (De Bok et al., 2004; Van Lier et al., 1993).
As an important regulatory factor, pH affects metabolic activity, with different species having
different optimal ranges. Chemical pH control has also been found to help recovery from
ammonia toxicity or VFA accumulation (Chen et al., 2008). Hydrogenotrophic methanogens
prefer a neutral pH environment and therefore, buffering solutions are typically used for pH
regulation (Martin et al., 2013; Seifert et al., 2014). Since the present experiment ran for six
months without any chemical addition, pH is believed to have been controlled by the
antagonistic relation among the produced organic acids, ammonia and the carbonic acid
introduced through the dilution of CO2.
Figure 4B shows the pH trend for all four reactors for the 6 month operational period during
which, pH was controlled solely by regulating the amount of CO2 in the gaseous feeding
substrate as also explained in (Savvas et al., 2017b). Although the same substrate was used in
all reactors, it can be seen that the thermophilic culture was generally running at a higher pH
12
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
than the mesophilic ones. This could be attributed to the higher levels of NH4+ in the
thermophilic reactor throughout the experiment (Figure 4C) and the lower solubility of CO2
at 55oC. In the mesophilic reactors, the higher solubility of CO2 (in water at 40oC, 1 bar ~
0.52 L/L water) and its transition to H2CO3 proved to be more than adequate for the pH
buffering of the system.
The decline in the pH of all 4 reactors after day 150 can be possibly connected to two
parameters. Firstly, the substantial reduction in ammonia levels (Figure 4C) which until then
had helped buffer any acidic factors (e.g. VFAs/H2CO3) present in the media and secondly,
the increase in the amount of feed gas after day 150 (Figure 1) which resulted in a higher
amount of unconverted dissolved CO2.
The noteworthy difference in ammonia levels among the three mesophilic reactors may be
related to differences in the formation of biofilm as well as the different types of support
material used. Biofilm formation stopped the bulk majority of the microbes from passing
through the pump thus protecting them from the shear forces which must have enhanced
hydrolysis due to cell damage in the reactors containing the suspended culture. Since the
levels of ammonia were identical in all four reactors at start-up any reduction in those levels
was dictated by 4 factors: a) the dilution rate due to the biologically produced water, b) the
water condensation rate in each individual reactor, c) the loss through volatilization with the
produced and collected gas, d) the biomass hydrolysis rate. Hydrolysis occurring at higher
rates due to damage from shear forces might be an explanation for the higher levels of
ammonia in the suspended cultures.
Furthermore, the lower ammonia levels observed in the reactor containing the LECA balls
could be a result of adsorption. In contrast to the dense polyethylene pieces, microporous clay
aggregates have been reported to retain NH4+ ions within their structure (much like zeolites)
13
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
and have been used in the past for ammonia removal (Busenberg and Clemency, 1973; Celik
et al., 2001).
3.5. Microbial profiling of the mesophilic and thermophilic cultures
As also explained in (Savvas et al., 2017b) limited nutritional availability resulted in a self-
regenerative microbial matrix where bacterial species were responsible for the recycling of
elements through digestion of cellular matter. The methane production rates achieved towards
the end of the 6 month period (Figure 1) suggest that there was regeneration of the
methanogenic species. Figure 5A shows the gene copy numbers for total bacteria, total
methanogens and acetotrophic species at the start and towards the end of the experiment in
the mesophilic and thermophilic reactors without the attachment media (reactors 1 and 2).
Unfortunately, biofilm samples from the reactors 3 and 4 could not be obtained as this would
require emptying of their contents.
The significant observed difference in both bacterial and archaeal species between the
mesophilic and thermophilic cultures suggest a level of inhibition created by the gradual
accumulation of propionate in reactor 2 (Figure 4A). Another possible reason could be related
to the smaller diversity regarding the methanogenic populations that can thrive at
thermophilic temperatures (Jones et al., 1987; Liu and Whitman, 2008). This being the case,
although growth rates cannot be directly linked to diversity, a less diverse microbiome is
expected to exhibit less flexibility to any inhibitory conditions present in the culture; as also
observed during oxygenation (Figure 2B).
Nevertheless, a trend was found regarding the relative percentages of the acetotrophic and
hydrogenotrophic methanogens as indicated in Figure 5B. Both the mesophilic and
thermophilic reactors were initially dominated by acetotrophic microbes which seem to have
gradually given their place to hydrogenotrophic species. This was expected as the amounts of
CO2 and H2 that were continuously injected and converted were much higher than the
14
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
amounts of acetate which is suggested to have been produced solely through cell carbon.
Nevertheless, as also described in (Savvas et al., 2017a) a degree of homoacetogenesis cannot
be excluded. The notable difference in the ratio of acetotrophic to hydrogenotrophic
methanogens on day 120 between the two cultures is a result of the changes in the mesophilic
and thermophilic microbiomes during the 45 day starvation period.
3.6. Specifics of the present study
This study concentrated on investigating the biomethanation performance of four reactors
under various operational conditions. It evaluated the impact of operation at mesophilic
versus thermophilic temperatures and the growth of biofilms in packed media as opposed to
suspended cultures. The study did not evaluate the potential impact of those operational
regimes in terms of gas-liquid mass transfer rates. As indicated by a number of studies
(Ferreira et al., 2010; Ribeiro and Mewes, 2006) the complexities arising from
solid/liquid/gas interactions at different temperatures are significant and therefore no safe
decisions can be made at this point as to the differences in gas mass transfer rates among the
four different reactors. However, as biomethanation systems are rate limited by physical
parameters that influence gas transfer, bubble size and gas hold-up, it will be important for
these to be studied to enable a more complete understanding. Significant differences between
mesophilic, thermophilic, suspended or attached cultures are likely to only be felt when gas
transfer is no longer the rate limiting step. Furthermore, it must be emphasized that due to the
singular nature of the culture used in the present study which relies on nutrient recycling
through microbial catabolism, no generalizations can be made with regards to other ex-situ
hydrogenotrophic biomethanation systems.
The present study adds to the understanding of the bio-catalytic methanation system
investigated in (Savvas et al., 2017b). The findings suggest that the self-regenerating
microbial culture that was allowed to evolve is robust enough to deal with inconsistencies in
15
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
the gaseous feedstock. It was also demonstrated that the high rate mixing system through
liquid recirculation used in the present reactors allows the integration of biofilms, thus
increasing their methanation capacity. Future work should be focused on further improving
this methanation capacity by investigating ways of increasing the methanogenic/non-
methanogenic population ratio of the culture.
4. Conclusions
Four hydrogenotrophic reactors were operated in parallel. Enrichment under nutrient closed
conditions showed accumulation of VFAs in the thermophilic culture but not in the activity of
all reactors returned to its pre-starvation capacity within a time range of hours. After
oxygenation the mesophilic populations exhibited faster recovery rates than the thermophilic
one. The configurations tested allowed the formation of biofilms under intense agitation. Gas
feeding rates of 200 L/L/d were achieved at a H2/CO2 to CH4 conversion efficiency of above
90%.
E-supplementary data of this work can be found in the online version of the paper.
Acknowledgements
This research was supported by the University of South Wales, UK, through the award of a
Centenary Postgraduate Scholarship. The authors also acknowledge the European Regional
Development Funding (ERDF) support provided by the Welsh Government (WG) A4B
scheme for the Knowledge Transfer Centre for Advanced Anaerobic Processes and Biogas
Systems (Ref: HE 14 15 1009) as well as the SMART CIRCLE (Ref: 122016/COL/003).
Microbial analyses were performed as part of the ERDF and WG KESS Program (Ref.
20441). The authors would like to acknowledge Welsh Water for the provision of the initial
inoculum.
16
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
References
1 Amani, T., Nosrati, M., Mousavi, S.M., Kermanshahi, R.K., 2010. Study of syntrophic
anaerobic digestion of volatile fatty acids using enriched cultures at mesophilic
conditions. Int. J. Environ. Sci. Technol. 8, 83–96. https://doi.org/10.1007/BF03326198
2 Andersson, S., Nilsson, M., Dalhammar, G., Rajarao, G.K., 2008. Assessment of carrier
materials for biofilm formation and denitrification. Vatten 64, 201–207.
3 Bassani, I., Kougias, P.G., Treu, L., Angelidaki, I., 2015. Biogas Upgrading via
Hydrogenotrophic Methanogenesis in Two-Stage Continuous Stirred Tank Reactors at
Mesophilic and Thermophilic Conditions. Environ. Sci. Technol. 49, 12585–12593.
https://doi.org/10.1021/acs.est.5b03451
4 Botheju, D., Bakke, R., 2011. Oxygen effects in anaerobic digestion - A review. Open
Waste Manag. J. 4, 1–19.
5 Botheju, D., Lie, B., Bakke, R., 2010. Oxygen effects in anaerobic digestion - II. Model.
Identif. Control 31, 55–65. https://doi.org/10.4173/mic.2010.2.2
6 Boušková, a., Dohányos, M., Schmidt, J.E., Angelidaki, I., 2005. Strategies for changing
temperature from mesophilic to thermophilic conditions in anaerobic CSTR reactors
treating sewage sludge. Water Res. 39, 1481–1488.
https://doi.org/10.1016/j.watres.2004.12.042
7 Burkhardt, M., Koschack, T., Busch, G., 2015. Biocatalytic methanation of hydrogen and
carbon dioxide in an anaerobic three-phase system. Bioresour. Technol. 178, 330–333.
https://doi.org/10.1016/j.biortech.2014.08.023
8 Busenberg, E., Clemency, C. V, 1973. Determination of the cation exchange capacity of
clays and soil using an ammonia electrode. Clays Clay Miner. 21, 213–217.
17
9 Celik, M.S., Ozdemir, B., Turan, M., Koyuncu, I., Atesok, G., Sarikaya, H.Z., 2001.
Removal of ammonia by natural clay minerals using fixed and fluidised bed column
reactors. Water Sci. Technol. Water Supply 1, 81–88.
10 Chen, Y., Cheng, J.J., Creamer, K.S., 2008. Inhibition of anaerobic digestion process: A
review. Bioresour. Technol. 99, 4044–4064.
https://doi.org/10.1016/j.biortech.2007.01.057
11 Cruwys, J.., Dinsdale, R.., Hawkes, F.., Hawkes, D.., 2002. Development of a static
headspace gas chromatographic procedure for the routine analysis of volatile fatty acids
in wastewaters. J. Chromatogr. A 945, 195–209. https://doi.org/10.1016/S0021-
9673(01)01514-X
12 De Bok, F.A.M., Plugge, C.M., Stams, A.J.M., 2004. Interspecies electron transfer in
methanogenic propionate degrading consortia. Water Res. 38, 1368–1375.
https://doi.org/10.1016/j.watres.2003.11.028
13 Estermann, T., Newborough, M., Sterner, M., 2016. Power-to-gas systems for absorbing
excess solar power in electricity distribution networks. Int. J. Hydrogen Energy 41,
13950–13959. https://doi.org/10.1016/j.ijhydene.2016.05.278
14 Ferreira, A., Ferreira, C., Teixeira, J.A., Rocha, F., 2010. Temperature and solid
properties effects on gas–liquid mass transfer. Chem. Eng. J. 162, 743–752.
https://doi.org/10.1016/j.cej.2010.05.064
15 Geng, M., Duan, Z., 2010. Prediction of oxygen solubility in pure water and brines up to
high temperatures and pressures. Geochim. Cosmochim. Acta 74, 5631–5640.
https://doi.org/10.1016/j.gca.2010.06.034
18
16 Global Carbon Capture and Storage Institute, 2012. Oxy Combustion With CO2 Capture
[WWW Document]. CO2 Capture Technol. URL
https://www.globalccsinstitute.com/publications/co2-capture-technologies-oxy-
combustion-co2-capture
17 Götz, M., Lefebvre, J., Mörs, F., McDaniel Koch, A., Graf, F., Bajohr, S., Reimert, R.,
Kolb, T., 2016. Renewable Power-to-Gas: A technological and economic review.
Renew. Energy 85, 1371–1390. https://doi.org/10.1016/j.renene.2015.07.066
18 Guneratnam, A.J., Ahern, E., FitzGerald, J.A., Jackson, S.A., Xia, A., Dobson, A.D.W.,
Murphy, J.D., 2017. Study of the performance of a thermophilic biological methanation
system. Bioresour. Technol. 225, 308–315.
https://doi.org/10.1016/j.biortech.2016.11.066
19 Hao, X., Cai, Z., Fu, K., Zhao, D., 2012. Distinguishing activity decay and cell death
from bacterial decay for two types of methanogens. Water Res. 46, 1251–1259.
https://doi.org/10.1016/j.watres.2011.12.029
20 Hwang, K., Song, M., Kim, W., Kim, N., Hwang, S., 2010. Effects of prolonged
starvation on methanogenic population dynamics in anaerobic digestion of swine
wastewater. Bioresour. Technol. 101, S2–S6.
https://doi.org/10.1016/j.biortech.2009.03.070
21 Jajesniak, P., Ali, H.E.M.O., Wong, T.S., 2014. Carbon Dioxide Capture and Utilization
using Biological Systems: Opportunities and Challenges. J. Bioprocess. Biotech. 4.
https://doi.org/10.4172/2155-9821.1000155
22 Jones, W.J., Nagle, D.P., Whitman, W.B., 1987. Methanogens and the diversity of
archaebacteria. Microbiol. Rev. 51, 135–177
19
23 Kato, M.T., Field, J.A., Lettinga, G., 1997. Anaerobe Tolerance to Oxygen and the
Potentials of Anaerobic and Aerobic Cocultures for Wastewater Treatment. Brazilian J.
Chem. Eng. 14. https://doi.org/10.1590/S0104-66321997000400015
24 Kougias, P.G., Treu, L., Benavente, D.P., Boe, K., Campanaro, S., Angelidaki, I., 2017.
Ex-situ biogas upgrading and enhancement in different reactor systems. Bioresour.
Technol. 225, 429–437. https://doi.org/10.1016/j.biortech.2016.11.124
25 Lecker, B., Illi, L., Lemmer, A., Oechsner, H., 2017. Biological hydrogen methanation -
A review. Bioresour. Technol. https://doi.org/10.1016/j.biortech.2017.08.176
26 Leonzio, G., 2016. Process analysis of biological Sabatier reaction for bio-methane
production. Chem. Eng. J. 290, 490–498. https://doi.org/10.1016/j.cej.2016.01.068
27 Li, J., Ban, Q., Zhang, L., Jha, A.K., 2012. Syntrophic propionate degradation in
anaerobic digestion: A review. Int. J. Agric. Biol. 14, 843–850.
28 Liu, Y., Whitman, W.B., 2008. Metabolic, phylogenetic, and ecological diversity of the
methanogenic archaea. Ann. N. Y. Acad. Sci. 1125, 171–189.
https://doi.org/10.1196/annals.1419.019
29 Luo, G., Angelidaki, I., 2012. Integrated biogas upgrading and hydrogen utilization in an
anaerobic reactor containing enriched hydrogenotrophic methanogenic culture.
Biotechnol. Bioeng. 109, 2729–36. https://doi.org/10.1002/bit.24557
30 Martin, M.R., Fornero, J.J., Stark, R., Mets, L., Angenent, L.T., 2013. A single-culture
bioprocess of methanothermobacter thermautotrophicus to upgrade digester biogas by
CO2-to-CH4 conversion with H2. Archaea 2013, 11.
https://doi.org/10.1155/2013/157529
20
31 Morozova, D., Wagner, D., 2007. Stress response of methanogenic archaea from
Siberian permafrost compared with methanogens from nonpermafrost habitats. FEMS
Microbiol. Ecol. 61, 16–25. https://doi.org/10.1111/j.1574-6941.2007.00316.x
32 O’Shea, R., Wall, D.M., McDonagh, S., Murphy, J.D., 2017. The potential of power to
gas to provide green gas utilising existing CO2 sources from industries, distilleries and
wastewater treatment facilities. Renew. Energy 114, 1090–1100.
https://doi.org/10.1016/j.renene.2017.07.097
33 Patterson, T., Savvas, S., Chong, A., Law, I., Dinsdale, R., Esteves, S., 2017. Integration
of Power to Methane in a Waste Water Treatment Plant – A Feasibility Study.
Bioresour. Technol. 245, 1049–1057. https://doi.org/10.1016/j.biortech.2017.09.048
34 Rachbauer, L., Voitl, G., Bochmann, G., Fuchs, W., 2016. Biological biogas upgrading
capacity of a hydrogenotrophic community in a trickle-bed reactor. Appl. Energy 180.
https://doi.org/10.1016/j.apenergy.2016.07.109
35 Ribeiro, C.P., Mewes, D., 2006. On the effect of liquid temperature upon bubble
coalescence. Chem. Eng. Sci. 61, 5704–5716. https://doi.org/10.1016/j.ces.2006.04.043
36 Rubia, M., Perez, M., Romero, L., Sales, D., 2002. Anaerobic mesophilic and
thermophilic municipal sludge digestion. Chem, Biochem. Eng. Q. 16, 119–124.
37 Savvas, S., Donnelly, J., Patterson, T., Chong, Z.S., Esteves, S.R., 2017a. Biological
methanation of CO2 in a novel biofilm plug-flow reactor: A high rate and low parasitic
energy process. Appl. Energy 202, 238–247.
https://doi.org/10.1016/j.apenergy.2017.05.134
21
38 Savvas, S., Donnelly, J., Patterson, T., Dinsdale, R., Esteves, S.R., 2017b. Closed
nutrient recycling via microbial catabolism in an eco-engineered self regenerating mixed
anaerobic microbiome for hydrogenotrophic methanogenesis. Bioresour. Technol. 227,
93–101. https://doi.org/10.1016/j.biortech.2016.12.052
39 Seifert, A.H., Rittmann, S., Herwig, C., 2014. Analysis of process related factors to
increase volumetric productivity and quality of biomethane with Methanothermobacter
marburgensis. Appl. Energy 132, 155–162.
https://doi.org/10.1016/j.apenergy.2014.07.002
40 Song, Y.C., Kwon, S.J., Woo, J.H., 2004. Mesophilic and thermophilic temperature co-
phase anaerobic digestion compared with single-stage mesophilic- and thermophilic
digestion of sewage sludge. Water Res. 38, 1653–1662.
https://doi.org/10.1016/j.watres.2003.12.019
41 Strübing, D., Huber, B., Lebuhn, M., Drewes, J.E., Koch, K., 2017. High performance
biological methanation in a thermophilic anaerobic trickle bed reactor. Bioresour.
Technol. 0–1. https://doi.org/10.1016/j.biortech.2017.08.088
42 Styring, P., Jansen, D., de Conink, H., Reith, H., Armstrong, K., 2011. Carbon Capture
and Utilisation in the green economy: Using CO2 to manufacture fuel, chemicals and
materials. Cent. Low Carbon Futur. Report 501.
43 U.S. Department of Energy, 2010. Fuel Cell Technologies Program Record [WWW
Document]. Energy, Effic. Renew. Energy. URL
https://energy.gov/eere/fuelcells/downloads/us-department-energy-fuel-cell-
technologies-program-18th-world-hydrogen
22
44 Van Lier, J.B., Grolle, K.C.F., Frijters, C.T.M.J., Stams, A.J.M., Lettinga, G., 1993.
Effects of acetate, propionate, and butyrate on the thermophilic anaerobic degradation of
propionate by methanogenic sludge and defined cultures. Appl. Environ. Microbiol. 59,
1003–1011.
45 Wang, W., Wang, S., Ma, X., Gong, J., 2011. Recent advances in catalytic hydrogenation
of carbon dioxide. Chem. Soc. Rev. 40, 3703–3727. https://doi.org/10.1039/c1cs15008a
46 Watnick, P., Kolter, R., 2000. Biofilm, city of microbes. J. Bacteriol. 182, 2675–2679.
https://doi.org/10.1128/JB.182.10.2675-2679.2000
47 Yun, Y.M., Sung, S., Kang, S., Kim, M.S., Kim, D.H., 2017. Enrichment of
hydrogenotrophic methanogens by means of gas recycle and its application in biogas
upgrading. Energy 135, 294–302. https://doi.org/10.1016/j.energy.2017.06.133
23
Figure captions
Fig. 1 Methanation efficiency of the four reactors relative to the gaseous feed (H2/CO2) input
flowrate; (non-operational periods have been omitted).
Fig. 2 (A) Methanation efficiency before, during and after oxygenation. (B) Percentage of
recovery of the methanogenic and bacterial populations for the mesophilic and thermophilic
cultures two days after oxygenation. (C) & (D) Recovery of all four reactors as shown by
their methanation efficiency at constant gas feeding rates after 13 days and 45 days of
carbon/energy starvation respectively.
Fig. 3 Methanation efficiency of two units before and after they were linked in series at two
different gas feeding rates.
Fig. 4 (A) Concentration of acetic and propionic acids in the liquid media, (B) the pH trend
and (C) the level of ammonium ions in all four reactors during the experiment.
Fig. 5 (A), the gene copy numbers /g VS of major microbial groups and (B), the relative
quantities of acetotrophic and hydrogenotrophic methanogens in the mesophilic and
thermophilic reactors at start (day 1) and towards the end (day 171) of the experiment.
[Bact. - bacteria; T. M – total methanogens; Mst – Methanosaeta]
24
Tables and Figures
Fig. 1
Fig. 2
25
Fig. 3
26
Fig. 4
27
Fig. 5
E-supplementary data of this work can be found in the online version of the paper
Biofilm formation and the media used for microbial attachment.[(A), reactor without attachment media, the colour of the liquid media was dark black; all microbes were in suspension and continuously mixed. (B), reactor containing the attachment media; the colour of the liquid media was light brown as the majority of the microbes had been attached on the polyethylene pieces. (C), Kaldnes K1 and (D), LECA attachment media prior to use]
28