1

Methane Enrichment of Biogas

Dipti Shinde, February 2011
Department of Chemical Engineering, Faculty of Engineering, Lund University


Using biogas as fuel in vehicles requires exclusion of carbon dioxide and can be accomplished through a
desorption process where air removes the carbon dioxide from the sludge, used to produce biogas. A high air
flow rate, low constant pH, high alkalinity, low pressure, high temperature, high stirrer speed and using a
packed column was expected to have an positive influence on the process and was therefore chosen to be
examined. However, experiments concerning high temperature and packed column were not performed. The
optimal value for the air flow rate resulted in 200 ml/min and the constant pH in 5.5. A higher flow rate
contributed to bubble coalescence while a lower generated an impaired desorption. The graphs for the
pressure experiments indicated that the lowest tested pressure, 0.7 atmospheres, appeared most favorable.
Nevertheless, extracting samples from the decompressed system was difficult and as a consequence,
fluctuation occurred in the result and an optimal value was not decided. Although a high stirrer speed had a
visibly desirable effect, the actual speed was not registered and optimal value remains to be determined.
Lastly, the high alkalinity was not proven to generate the positive effect expected.

INTRODUCTION
Biogas is a gaseous mixture of mainly methane and
carbon dioxide and is the results of an anaerobic
digestion of organic matter. A reduction of carbon
dioxide in the gas would enhance the efficiency of
biogas and enables the usage as fuel in vehicles
(Lindberg Rasmuson, 2006). A literature study
showed that the removal of carbon dioxide from
the sludge, producing the biogas, appeared most
beneficial. Using a desorption column where air
enters and desorbs the carbon dioxide from the
liquid enables a purified sludge to be recycled to
the digester (Lindberg Rasmuson, 2006).
Parameters
Numerous parameters were predicted to effect the
desorption of carbon dioxide and as a consequence
the aim of the study was concentrated to which
parameters would lead to an improved desorption
and what impact the design of the column has on
the process.
The air flow influences the desorption depending
on the rate. A rather low gas velocity contributes
to a homogenous flow with small bubbles while a
higher flow rate can result in bubble coalescence
(Lindberg Rasmuson, 2006). Moreover, within
the homogenous flow, the gas-liquid surface area
increases with increasing air flow rate and
consequently the amount of desorbed carbon
dioxide increases (Lei et al., 2007).
The pH in the liquid reflects the amount of carbon
dioxide available. Equations 1 to 3 illustrate the
reactions taking place in the system and are highly
influenced by the pH, stated by Lindberg
Rasmuson (2007) and Lei et al. (2007).

()

() ()

()

()

()

() ()

()

()

() ()
At lower pH values the solubility of carbon dioxide
in the liquid decreases and the reactions proceed
to the right, increasing the amount of gaseous
carbon dioxide to be carried off by the air (Richards
et al., 1994). Nevertheless, during the desorption
process the hydrogen ions, in reaction 2, are
consumed and the pH is increased (Lei et al., 2007).
The Alkalinity was expressed with equation 4
received by Tchobanoglous et al. (2003).
[

] [

] [

] [

] ()
2

Since equation 1 to 3 describes how the
consumption of one component contributes to
another one being formed, the alkalinity is not
affected by the desorption process (Summerfelt et
al., 2003). The influence of adding a strong acid is
however considered by Holmes (2002) who states
that the alkalinity will be reduced.
If the alkalinity is believed to represent the buffer
capacity of the system, the pH changes caused by
removal of carbon dioxide would not affect the
desorption when using a higher alkalinity
(Tchobanoglous et al., 2003).
The pressure, like the pH, influences the solubility
of carbon dioxide. A decompression results in a
super saturation of the liquid which as a
consequence rapidly releases the carbon dioxide to
the gas phase. However, the pressure is again
increased when the carbon dioxide is removed
from the liquid (Kierzkowska Chacuk, 2010).
The temperature is dependent on the digestion
process. Hilkiah Igoni et al. (2008) argue that an
increased temperature increases the methane
production. Moreover, Pande Fabiani (1989)
adds that the desorption rate is significantly
increased. Lastly, the time for aeration is also
expected to be reduced (Lei et al., 2007). However,
a higher temperature also results in an unstable
digestion process due to the sensitivity of the
bacteria and the mesophilic temperature range, 30
38 , is therefore commonly used (Hilkiah Igoni
et al., 2008).
The stirrer and the stone diffuser affect the mass
transfer of the carbon dioxide from the liquid to
the gas.
A stirrer creates turbulence in the liquid, increasing
the mass transfer (Kierzkowska Chacuk, 2010).
Sun et al. (2010) examined the most favorable
position of the stirrer and concluded that a stirrer
placed centrally, at the bottom of a bottle would
be most desirable. Although an increased stirrer
speed naturally would generate greater turbulence
a too high value provides an unfavorable mixing
(Kierzkowska Chacuk, 2010).
Additionally, smaller bubble diameter, as a results
of a homogenous flow, created with the help of a
stone diffuser (Lindberg Rasmuson, 2006),
increases the specific bubble surface and therefore
the mass transfer (Merkel Krauth, 1999).
A packed column can reduce the liquid back
mixing which is a major drawback of a bubble
column (Carleton et al., 1967). Moreover, reduced
bubble coalescence followed by an increased mass
transfer can be expected (Carleton et al., 1967).
Furthermore, Carleton et al. (1967) concludes that
the additional pressure drop caused by a packed
column is negligible.
Moreover, large void spaces in a structured
packing can diminish the bio fouling problems as a
result of using a packed column according to
Summerfelt et al. (2003).
EXPERIMENTAL WORK
The experiments were performed in the laboratory
of Bioprocess Control AB. The liquid containing
dissolved carbon dioxide was prepared with a soda
streamer. A 2000 ml bottle was used as a bubble
column and a pump delivered the air to the system
through a stone diffuser. The air flow rate was
measured with the help of a flow meter. A
magnetic stirrer was placed at the bottom of the
bottle. Furthermore, a probe was used to measure
pH in the liquid. A pressure sensor registered the
pressure and was regulated by a vacuum pump.
Finally, samples were extracted from the outlet
and analyzed in the gas chromatograph.
Temperature and packed column experiments
were not performed due to limitation of time and
equipment. Hence, all experiments were
conducted in ambient temperature (22 ). Each
parameter was tested one at a time.
The air flow rates used were 100 ml/min, 200
ml/min and 300 ml/min while constant pH values
of 5.5, 6.0 and 6.5 were applied. Moreover,
constant pressure values of 0.7, 0.8 and 0.9
atmosphere were examined along with a diffusion
experiment. Lastly, low and high stirrer speeds
were studied with the stone diffuser removed. All
experiments were performed in a 20 mM and an
80mM calcium carbonate buffer solution.
3

RESULTS AND DISCUSSION
Not being able to control the initial amount of
carbon dioxide in the solution resulted in
aggravated comparisons. Also, the higher alkalinity
solution generated lower maximum values of
desorbed carbon dioxide and is assumed to be a
result of a lower amount of initial carbon dioxide
added to the solution. The assumption is
confirmed when comparing the pH graphs for the
two alkalinities since the 80 mM solution starts at a
higher pH indicating less carbon dioxide.
Results from the air flow rate changes showed
that although an increase from the air flow rate
100 ml/min to 200 ml/min favored the desorption,
an increase to 300 ml/min lacked the same result,
depicted in figure 1.

Figure 1. The desorption of carbon dioxide into gas
from a 80 mM solution with different air flow rates, a
medium stirrer speed and atmospheric pressure.
Both solutions showed similar results and the air
flow rate 300 ml/min therefore seems to result in
bubble coalescence.
Results from the pH changes proved that lower pH
values generated higher maximum values and
shorter desorption times, depicted in figure 2. The
pH values were achieved by adding 3 M
hydrochloric acid, HCl, or 3 M sodium hydroxide. In
order to counteract the increase in pH during the
desorption, 0.5 M HCl was automatically added.

Figure 2. The desorption of carbon dioxide into gas
from a 20 mM solution with different pH, a medium
stirrer speed, 200 ml/min air flow rate and atmospheric
pressure.
Results from the alkalinity measurements
confirmed that the alkalinity remained unaltered
after the desorption. However, measuring the
alkalinity after adding hydrochloric acid resulted in
a reduced alkalinity.
The absence of a difference in the desorption
graphs for the two solutions indicate that the
effect of the higher alkalinity was not detectable. A
higher alkalinity solution consists of added cations
which may interact with the dissolved carbon
dioxide and can be the reason to the 80 mM
solution not resulting in an improved desorption.
Results from the pressure changes and the
diffusion are depicted in figure 3.

Figure 3. The desorption of carbon dioxide into gas
from a 20 mM solution with different pressure values
and a medium stirrer speed.
A low pressure complicated the extraction of
samples and as a result fluctuations occurred.
Nonetheless, figure 3 illustrates higher maximum
values for the lower pressure values. The diffusion
20
30
40
50
60
70
80
0 20 40 60 80
C
a
r
b
o
n

d
i
o
x
i
d
e

i
n

g
a
s

(
%
)

Time (min)
300 ml/min
200 ml/min
100 ml/min
0
20
40
60
80
100
0 50 100
C
a
r
b
o
n

d
i
o
x
i
d
e

i
n

g
a
s

(
%
)

Time (min)
pH 6.5
pH 6
pH 5.5
50
60
70
80
90
100
0 50 100
C
a
r
b
o
n

d
i
o
x
i
d
e

i
n

g
a
s

(
%
)

Time (min)
0.7 atm
0.8 atm
0.9 atm
Diffusion
4

is as desired much lower. Examining the pH graphs
however, indicates a lower amount of initial
carbon dioxide. Regardless, the marginally higher
pH is not considered to correspond to the large
difference in the desorption graphs.
Results from the stirrer speed changes showed
that higher stirrer speeds improves the desorption
in both solution although more distinctly in the
lower alkalinity solution.
CONCLUSIONS
The design of the column was considered in terms
of using a packed column instead of an empty. The
reduced liquid back mixing as well as less bubble
coalescence would improve the desorption. The
hypothesis was however not confirmed by
experiments.
The conclusions from the results for the six
parameters considered follow below:
The optimal value for the air flow rate in the
experiments was 200 ml/min since both the
higher and lower rate resulted in worsened
desorption.
The constant pH 5.5 was proven, through
experiments, to create a better desorption
than having higher values.
The alkalinity did not show any effect on the
desorption in the experiments carried out.
The graphs generated from the pressure
changes indicated that lower values are
preferable. However, an optimal value could
unfortunately not be decided since
fluctuations occurred.
The improved desorption caused by an
increased stirrer speed was apparent.
Nevertheless, an optimal value was
unattainable since no actual values were
obtained.
Finally, high temperature values are predicted
to improve the desorption. An optimal value
was not decided since the parameter was not
experimentally tested.
Future research primarily comprises testing of the
mentioned parameters in a real sludge and on a
larger scale to examine if any side reactions occur
and if the parameters generate the same results.
REFERENCES
Lindberg, Anna Rasmuson, ke C. (2006), Selective Desorption
of Carbon Dioxide From Sewage Sludge for In Situ Methane
Enrichment-Part I: Pilot-Plant Experiments. Biotechnology and
Bioengineering, Vol. 95, No. 5, pp. 794-803.
Lindberg, Anna Rasmuson, ke C. (2007), Selective Desorption
of Carbon Dioxide From Sewage Sludge for In Situ Methane
Enrichment-Part II: Modelling and Evaluation of Experiments.
Biotechnology and Bioengineering, Vol. 97, No. 5, pp. 1039-
1052.
Lei, Xiaohui Sugiura, Norio Feng, Chaunping Maekawa,
Takaaki. (2007), Pretreatment of anaerobic digestion effluent
with ammonia stripping and biogas purification. Journal of
Hazardous Materials, Vol. 145, No. 3, pp. 391-397.
Richards, Brain K. Herndon, Fredrick G. Jewell, William J.
Cummings, Robert J. White, Thomas E. (1994), In Situ
Methane Enrichment in methanogenic energy crop digesters.
Biomass and Bioenergy, Vol. 6, No. 4, pp. 275-282.
Tchobanoglous, George Burton, Franklin L. Stensel, David H.
(2003), Wastewater engineering, treatment and reuse.
McGraw-Hill, New York.
Summerfelt, Steven T. Davidson, John Waldrop, Thomas
(2003), Evaluation of full-scale carbon dioxide stripping columns
in a coldwater recirculating system. Aquacultural Engineering,
Vol. 28, No. 3-4, pp. 155-169.
Holmes-Farley, Randy (2002), Chemistry and the Aquarium:
What is Alkalinity? Advanced Aquarists Online Magazine, Vol. 1,
Issue: II.
Kierzkowska-Pawlak, Hanna Chacuk, Andrzej (2010), Carbon
Dioxide Desorption from Saturated Organic Solvents. Chemical
Engineering & Technology, Vol. 33, No 1, pp. 74-81.
Hilkiah Igoni, A. Ayotamuno, M.J. Eze, C.L. Ogaji, S.O.T.
Probert, S.D. (2008), Designs of anaerobic digesters for
producing biogas from municipal solid-waste. Applied Energy,
Vol. 85, No. 6, pp. 430-438.
Pande, D.R. Fabiani, C. (1989), Feasibility studies on the use of
a naturally occurring molecular sieve for methane enrichment
from biogas. Gas Separation & Purification, Vol. 3, No. 3, pp.
143-147.
Sun, H. Li, T. Ying, W. Fang, D. (2010), Technique on
Synthesis of Alcohol Ether Fuel in Slurry Bubble Column
Reactors (I)Experimental Study on Bubble Behavior. Energy
Sources Part A: Recovery, Utilization & Environmental Effects,
Vol.31, No. 18, pp. 1612-1626.
Merkel, Wolf Krauth, Karlheinz (1999), Mass transfer of
carbon dioxide in anaerobic reactors under dynamic substrate
loading conditions. Water Research, Vol. 33, No. 9, pp. 2011-
2020.
Carleton, A. J. Flain, R. J. Rennie, J Valentin, F.H.H. (1967),
Some properties of a packed bubble column. Chemical
Engineering Science, Vol. 22, No. 12, pp. 1839-1845.