Can an electronic nose assess the biomethanation process?

Gilles ADAM* and Jacques NICOLAS

Environmental Sciences and Management Department, University of Liège
Avenue de Longwy 185, 6700 Arlon, Belgium
*Gilles.Adam@doct.ulg.ac.be

E-nose Concept
Objective
Also called gas sensor array, an e-nose employs various Anaerobic digestion technology is a trendy solution for a more
complementary gas sensors, in order to improve selectivity. sustainable world. It produces biogas, a mixture of methane and carbon
Good process dioxide used in combined heat and power unit to produce renewable
energy.
Biogas
This technology is now adapted in farm plants in Belgium. Because of
Unstable process
Good
process
the diversity of substrates and sensitivity of the anaerobic digestion
S1 S2 S3 S4 S5 process to overload, most of the reactors are under-loaded to avoid
Gas sensor Gas footprint Pattern
process disturbance, resulting in lower gas yield.
Disturbed process
array recognition
Response E-nose technology offers the possibility of a new robust, simple,
sensitive and low cost tool for biogas production monitoring in order
to optimize the process and increase gas yield in small scale
Data base
agricultural plants.

Material and methods

System of anaerobic batches Measured variables
Biogas collection
Temperature regulated at 38 ± 1°C Daily analyses :
Three load strategies : Chemical analysis
1. Control : 1 g sucrose batch-1.day-1 -COV composition and
2. Slow load increase : 1 to 5 g concentration (GC-MS1)
sucrose batch-1.day-1 -[CH4] and [CO2] (IR2 sensors)
3. Punctual overloads : 1 g sucrose -[H2S], [CO] (EC3 gas sensors)
Batches batch-1.day-1 and one punctual
overload a week of 3, 9 and 12 g
sucrose batch-1.
Sample in
Experiment duration : 3 weeks
Insulated box Electronic nose analysis
6 TGS sensors (tin oxide sensors)
Twelve anaerobic batches of 1.5 L in are put in an insulated Sensors chamber regulated at 50°C
thermostatic bath. Batches were inoculated with sludge (5% TS) Dilution rate of biogas : 25x in air (10% RH at 37°C).
from the wastewater treatment station of Schifflange (Luxembourg). Biogas input of 9 minutes.

Results and discussion

3
14

Principal component analysis of the e-nose normalized data (6 variables, 83 observations) 2

Stable digestion area Overload area
9

3 1 12
10
5 14
1 11
11 109
Factor 2 (33%)

7 13
6
2 4 5
2 Overload area 95% of the variance explained 0 1 7
2 6
133 4 Area of digestion in
3 8
Two groups of data detected: -1 8
recovery process

1
-overload area
FACTOR 2 (33%)

-2 Batch 1
Batch 2
0 > 4 g sucrose.day-1 Numbers represents the days numbers after inoculating the batches
-3
-3 -2 -1 0 1 2 3
-stable digestion Factor 1 (62%)
-1 Stable digestion Overload schedule: day 4, 9 and 13( respectively 3, 9 and 12 g sucrose batch-1

< 4 g sucrose.day-1 Intermediate situation observed in the punctual

-2 Control
Slow load increase
overload data set:
Puntual overload
-Days following the overload
-3
-3 -2 -1 0 1 2 3

Recovery process detected
FACTOR 1 (62%)

Conclusions Future trends

E-nose was able to: Electronic nose demonstrated its potential in biomethanation
process monitoring. The tool was able to discriminate
1. Detect overload of the system overload situations. Nowadays, anaerobic digestion control is
2. Suggest disturbed digestion and only able using high-tech analysis, such as VFA
determination by GC-MS or HPLC. Results of these analyses
recovery time of the batches to are expensive and available one week after sampling.
return to a stable digestion E-nose allows to avoid this delay by on-line control but it
1 Gas chromatography – Mass Spectrometry (GC-MS)
needs to be tested in conditions closer to the real situation.
2 Infrared (IR)
3 Electrochemical (EC)