-
FACULTY OF BIOSCIENCE ENGINEERING
CENTRE FOR ENVIRONMENTAL SCIENCE AND TECHNOLOGY
Academic year 2015-2016
CHARACTERIZATION OF INDUSTRIAL VOLATILE ORGANIC
COMPOUNDS EMISSION IN RWANDA AND BIOFILTRATION
OF ACETONE, DIMETHYL SULFIDE AND HEXANE
Juvenal MUKURARINDA
Promoter: Prof. dr. ir. Herman VAN LANGENHOVE
Tutors: Dr. ir. Christophe WALGRAEVE
Ir. Joren BRUNEEL
Master’s dissertation submitted in partial fulfillment of the
requirements for the degree of
Master of Science in ENVIRONMENTAL SANITATION
-
i
COPYRIGHT
The author and promoter give permission to use this thesis for
consultation and to copy parts of it
for personal use. Every other use is subject to the laws of
copyright; more specifically the source
must be extensively specified when using results from this
dissertation.
Gent, June 2016.
Juvenal MUKURARINDA (author)
Ir. Joren BRUNEEL (tutor)
Dr. ir. Christophe WALGRAEVE (tutor)
Prof. dr. ir. Herman VAN LANGENHOVE (promoter)
-
ii
ACKNOWLEDGEMENTS
First of all, I want to say thank you to God Almighty for his
faithfulness, mercy, provision,
protection and support during my entire study period.
I would never have been able to finish my thesis without
guidance from my tutors and professor.
My deepest gratitude goes to my promoter Prof. dr. ir. Herman
Van LANGENHOVE for allowing
me to do research at the EnVOC lab. I am also particularly
grateful to him for his scholastic
guidance, innovative suggestions, and supervision throughout the
period of research work.
I gratefully thank my tutors, Dr. ir. Christophe WALGRAEVE and
Ir. Joren BRUNEEL for their helpful
attitude, constant encouragement, providing information
constructive comments and great
endurance throughout the research and manuscript writing. They
consistently allowed this paper
to be my own work, but steered me in the right direction
whenever they thought I needed it.
I also gratefully acknowledge the valuable comments and
suggestions from
Prof. dr. ir. Kristof Demeestere during the EnVOC presentation
seminars.
I wish to extend my gratitude to all members of the EnVOC family
especially Lore and Patrick for
their kind assistance during the research time.
My sincere gratitude and cordial respect to Prof. dr.ir. Peter
Goethals to have allowed me to join
the challenging but wonderful program (Master of Science in
Environmental Sanitation). My
sincere thanks to the coordinators of the program: Sylvie,
Veerle for their kind cooperation,
valuable advice and continuous encouragement during the entire
study period.
I would like to like to thank the Flemish Interuniversity
Council i.e. Vlaams Interuniversitaire Raad
(VLIR-OUS) for offering me a scholarship to pursue higher
education at Ghent University, Belgium
as well as for their blessed aim of transferring knowledge
towards developing countries such as
Rwanda.
Finally, I must express my very profound gratitude to my
parents, sisters and brother for providing
me with unfailing support and continuous encouragement
throughout my years of study.
-
iii
ABSTRACT
Rwanda’s economic transformation is based on the service
delivery, mining sector and industrial
activities. Technologies to handle the emissions in industries
from production processes especially
VOC are yet to be established. In addition to that, no studies
have been conducted before to
check the status of emissions made in different local
manufacturing industries in Rwanda. VOC
are organic compounds which have adverse effects on human health
as well as the
environment when exposed to high concentrations for long
time.
This study was divided into two parts: In the first part,
samples by means of Tenax TA sorbent
tubes were collected indoor and outdoor in three different
industries, Sulfo Rwanda industry,
AMEKI color and Inyange industries. Samples were analyzed by
TD-GC-MS and a total new data
of 45 VOCs concentrations levels were monitored to both indoor
and outdoor environment of
the three local manufacturing industries. In Sulfo, soap
production unit, the TVOCs indoor and
outdoor were 3.38 103 and 3.51 103 μg.m-3 respectively. Still at
Sulfo, cosmetic production unit,
the TVOCs was 0.13 103 μg.m-3 for indoor and 0.06 103 μg.m-3 for
outdoor. In AMEKI color, the
indoor and outdoor was 39.2 103 and 0.02 103 μg.m-3
respectively. At Inyange, the TVOCs
encountered were 0.45 103 μg.m-3 indoor and 0.02 103 μg.m-3
outdoor. The second part of the
study investigated the performance of a biofilter contaminated
by three compounds with
different physical chemical properties, acetone, DMS and hexane.
By means of SIFT-MS, VOC
concentrations were measured at different position along the BF
and perform dynamic
partitioning coefficient of BF packing materials. The
performance assessment of the biofilter was
done by comparing inlet concentrations (IL), elimination
capacity (EC) and removal efficiency
(RE). The maximum RE of the mixed target total VOC was 74 % to
IL of 22.4 ± 4.80 mg C.m-3.min-1
and EC of 16.6 ± 4.07 mg C.m-3.min-1 at an EBRT of 57 s. The
highest maximum RE for individual
contaminants was 99.9 % for acetone at an IL of 20.9
mg.m-3.min-1 and 20.9 mg.m-3.min-1 EC. The
maximum RE of DMS was 74 % at IL of 34.9 mg.m-3.min-1 and 25.72
mg.m-3.min-1 EC. The maximum
RE for hexane was 47 % at IL of 67.33 mg.m-3.min-1 and 31.68
mg.m-3.min-1 EC.
Based on this performance, biofiltration can be seen as an
urgent technology for the treatment
of the target VOC in manufacturing and production industries
where technology for VOC
treatment is yet to be implemented.
Keywords: VOC, thermal desorption, gas chromatography, mass
spectroscopy, biofiltration,
selected ion flow tubes mass spectroscopy (SIFT-MS), Ion
Chromatography.
-
iv
TABLE OF CONTENTS
COPYRIGHT
...................................................................................................................................................
i
ACKNOWLEDGEMENTS................................................................................................................................
ii
ABSTRACT
....................................................................................................................................................
iii
GENERAL INTRODUCTION
...........................................................................................................................
1
PROBLEM STATEMENT
...................................................................................................................................
1
CHAPTER 1 LITERATURE REVIEW
...................................................................................................................
2
1.1 Volatile organic compounds
.....................................................................................................................
2
1.2 Sources of Volatile organic compounds
................................................................................................
2
1.2.1 Natural sources
.......................................................................................................................................
3
1.2.2 Anthropogenic sources
........................................................................................................................
3
1.3 Hazards of VOCs
............................................................................................................................................
4
1.3.1 Human health effect
............................................................................................................................
4
1.3.2 Tropospheric photochemical ozone formation
............................................................................
5
1.3.3 Stratospheric ozone depletion
...........................................................................................................
6
1.3.4 Global Greenhouse effect
..................................................................................................................
6
1.4 Air pollution control technologies for VOCs
...........................................................................................
7
1.4.1 Non-biological technology for VOCs
...............................................................................................
9
1.4.2 Biological treatment of VOCS
............................................................................................................
9
1.4.2.1 Biotrickling filters
............................................................................................................................
10
1.4.2.2 Bioscrubber
....................................................................................................................................
11
1.4.2.3 Biofilter
.............................................................................................................................................
12
1.4.3 The biodegradation of hydrophobic compounds
.....................................................................
15
1.4.4 Parameters used to check the performance of the biological
systems .............................. 16
1.5 Scope and objectives
................................................................................................................................
16
CHAPTER 2 MATERIALS AND METHODS
....................................................................................................
18
PART I: ANALYSIS OF INDUSTRIAL VOC CONCENTRATIONS IN RWANDA
............................................ 18
2.1 Tenax TA tubes
.............................................................................................................................................
18
2.2 Conditioning
.................................................................................................................................................
18
2.2.1 Loading with internal standards
......................................................................................................
19
2.2.1.1 Chemicals
......................................................................................................................................
19
2.2.1.2 Preparation of the closed two-phase system
......................................................................
19
2.2.1.3 Calculation of the headspace concentration
....................................................................
20
-
v
2.2.1.4 Loading
...........................................................................................................................................
20
2.2.2 Pump Calibration
.................................................................................................................................
20
2.3 Sampling Campaigns
.................................................................................................................................
20
2.3.1 Description of the sampling
locations............................................................................................
22
2.3.1.1 Sulfo Rwanda Industries
..............................................................................................................
22
2.3.1.2 AMEKI Color
...................................................................................................................................
23
2.3.1.3 Inyange Industries
........................................................................................................................
23
2.4 Analysis of Tenax TA sampling tubes
......................................................................................................
24
2.4.1 TD-GC-MS
...............................................................................................................................................
24
2.4.2 Calibration of the TD-GC-MS
............................................................................................................
25
2.5 Quantification
..............................................................................................................................................
26
2.5.1 Calculation of the analyte concentration
...................................................................................
26
PART II: ABATEMENT TECHNOLOGY
.........................................................................................................
28
2.6 BIOFILTRATION
...............................................................................................................................................
28
2.6.1 Physical chemical properties of the representative VOC
compounds ............................... 28
2.6.2 Biofiltration process
..............................................................................................................................
28
2.6.2.1 Biofiltration design
........................................................................................................................
28
2.6.2.2 Biofiltration setup
..........................................................................................................................
28
2.6.3 Characterization of the packing materials
..................................................................................
30
2.6.3.1 Bulk Density
....................................................................................................................................
31
2.6.3.2 Moisture content
..........................................................................................................................
31
2.6.3.3 Water holding capacity
.............................................................................................................
31
2.6.3.4 Porosity
............................................................................................................................................
31
2.6.4 Environmental conditions of the filter bed
....................................................................................
34
2.6.4.1 Temperature
..................................................................................................................................
34
2.6.4.2 pH
.....................................................................................................................................................
34
2.6.4.3 Nutrients
..........................................................................................................................................
34
2.6.4.3 Pressure drop
.................................................................................................................................
34
2.6.5. Analytical instrumentation
...............................................................................................................
35
2.6.5.1 Analysis with SIFT-MS
....................................................................................................................
35
2.6.5.2 Analysis with Ion chromatography
..........................................................................................
35
-
vi
CHAPTER 3 RESULTS AND DISCUSSION
.....................................................................................................
36
PART I: INDUSTRIAL VOC ANALYSIS IN RWANDA
....................................................................................
36
3.1 Results
.............................................................................................................................................................
36
3.2 Discussion
.......................................................................................................................................................
41
3.2.1 General discussion
...............................................................................................................................
41
3.2.1 Indoor to outdoor concentrations of the sampling sites
.......................................................... 44
Part II: BIOFILTRATION OF VOC
................................................................................................................
46
3.3 Results and discussion
................................................................................................................................
46
3.3.1 Partition coefficient of the pollutants to the packing
materials ............................................. 46
3.3.2 Biological oxidation of pollutants.
...................................................................................................
48
3.3.3 Bioreactor bed
.....................................................................................................................................
49
3.3.4 The Carbon dioxide (CO2) and Elimination capacity (EC)
...................................................... 51
3.3.5 The effect of pH on the removal of target VOC
.........................................................................
52
3.3.6 Sulfate measurement
.........................................................................................................................
53
3.3.7 Effect of Silicon on the removal of hexane
..................................................................................
54
3.3.8 Inhibitory effect for hexane degradation
.....................................................................................
54
CHAPTER 4 CONCLUSION AND RECOMMENDATION
.............................................................................
56
4.1 CONCLUSION
...............................................................................................................................................
56
4.2
RECOMMENDATION....................................................................................................................................
57
REFERENCES
................................................................................................................................................
58
APPENDIX I
.................................................................................................................................................
69
APPENDIX II
................................................................................................................................................
72
A. Breakthrough curves for dry silicon foam
...............................................................................................
72
B. Breakthrough curves for dry wood chips
................................................................................................
72
C. Breakthrough curves for dry compost
....................................................................................................
73
D. Breakthrough curve of compost at normal and dry condition.
...................................................... 73
-
vii
LIST OF FIGURES
Figure 1: Industrial sector VOC emissions in EU-27
.......................................................................................
4
Figure 2: A tree diagram of the VOC emissions abatement
technology (Khan & Ghoshal, 2000). . 8
Figure 3: Application limit of flow rate vs VOC concentrations
of different air pollution
technologies control.
.....................................................................................................................
8
Figure 4 : Biotrickling filter setup .
..................................................................................................................
11
Figure 5: Bioscrubber setup.
.........................................................................................................................
12
Figure 6: A typical setup of biofilter.
............................................................................................................
12
Figure 7: Conditioning oven
.........................................................................................................................
19
Figure 8: Tol-d8 structure.
................................................................................................................................
19
Figure 9 : Sampling locations on the map of Kigali
city...........................................................................
21
Figure 10: Soap production unit. Figure 11: Cosmetics production
unit. .......................... 22
Figure 12: Paint production.
..........................................................................................................................
23
Figure 13: Juice and milk production unit.
.................................................................................................
24
Figure 14 : The TD-GC-MS with (1) the TD (2) the transfer line
from GC to MS, (3) GC and (4) MS. 24
Figure 15: Schematic diagram of the biofiltration setup. it.
....................................................................
29
Figure 16: Actual setup of the biofilter
........................................................................................................
30
Figure 17: Packing materials used in biofiltration process.
......................................................................
30
Figure 18: The peak injection experiment of acetone, DMS and
hexane. .......................................... 32
Figure 19: The peak injection experiment using methane gas.
.............................................................
32
Figure 20: Breakthrough curve of the pollutant to the packing
material and blank. ....................... 33
Figure 21: Pressure drop in the filter bed.
....................................................................................................
34
Figure 22: The total indoor and outdoor concentrations of four
sampling sites. ................................ 41
Figure 23: The indoor total VOC concentrations of chemical
groups in four sampling sites. .......... 42
Figure 24: The outdoor total VOC concentrations of chemical
groups in four sampling sites. ....... 42
Figure 25: Indoor target groups’ abundances in four sampled
sites. ................................................... 43
Figure 26: Outdoor target groups’ abundance in four sampled
sites. ................................................. 43
Figure 27: Partitioning coefficient of acetone, DMS and Hexane
......................................................... 47
Figure 28: The normalized start up concentrations of acetone,
DMS and hexane at EBRT of 57 s.49
Figure 29: The Total inlet concentrations ( ) and total removal
efficiency ( ) of the three
pollutants at EBRT of 57 s.
............................................................................................................
50
Figure 30: EC in function of IL of acetone, DMS and hexane at an
EBRT of 57 s. ............................... 51
Figure 31: Produced CO2 in function of total EC at an EBRT of 57
s. ..................................................... 52
Figure 32: RE in function of pH of acetone, DMS and hexane at an
EBRT of 57 s. ............................. 53
Figure 34: The EC in function of IL for hexane in mixture and
hexane only at EBRT of 57 s. .............. 55
-
viii
LIST OF TABLES
Table 1: Related health effects to exposure of high VOC
concentrations. .......................................... 4
Table 2: Classification of vapor phase biotechnology systems.
............................................................ 10
Table 3: Performance parameters used in biological treatment
systems. .......................................... 16
Table 4: The overview information of the VOC sampling campaign
collected at three local
industries in Rwanda.
.......................................................................................................................
22
Table 5: Physical chemical properties of acetone, dimethyl
sulfide and hexane. ............................ 28
Table 6: Calculated physical chemical properties of the packing
materials. .................................... 32
Table 7: The precursor and products ion used to measure
concentrations in SIFT-MS. .................... 35
Table 8: Indoor VOC concentrations (μg.m-3) measured at four
sampling sites, Kigali, Rwanda. . 37
Table 9: Outdoor VOC concentrations (μg.m-3) measured at four
sampling sites, Kigali, Rwanda.
....................................................................................................................................................................
39
Table 10: Indoor to Outdoor ratio concentrations of the four
sampling sites. .................................... 44
Table 11: Calculated partitioning coefficients of the dry
packing material ....................................... 47
Table 12: Performance parameters of the biofilter.
................................................................................
49
-
ix
LIST OF ABBREVIATION
AMEKI
Atelier des Meubles de Kigali
BF
Biofilter
Cin Inlet Concentration
Cout
Outlet concentration
CTS Closed two phase system
DMS
Dimethyl Sulfide
EBRT Empty Bed Residence Time
EC Elimination Capacity
EPA Environment Program Agency
GC Gas Chromatography
GWP Global Warming Potential
I/O Inlet to Outlet ratio
IL Inlet Load
IS Internal Standards
MINICOFIN
Ministry of Finance and Economic Planning, Rwanda
NIST National Institute of Standard and Technology
NOx Nitrogen Oxides
Q Flow
RF Response Factor
RSRF Relative Sample Response Factor
SIFT-MS Selected Ion Flow Tubes Mass Spectroscopy
TD Thermal Desorption
TVOCs Total Volatile Organic Compounds
V Volume
VOC Volatile Organic Compounds
-
1
GENERAL INTRODUCTION
PROBLEM STATEMENT
Air is an important free available commodity which defines life
on earth but due to mostly
human activities the quality of air is changing and this reflect
negative effects to human health
as well on environment.
The atmospheric emissions trends in developing countries are
increasing mainly because of their
rapid economic transformation especially in urban places. In the
last decade, emissions in
developed countries are reported to have decreased but some are
still in higher concentrations
than the air quality standards for the protection of human
health (Guerreiro et al., 2014). The
World Health Organizations (WHO) report that about seven million
death globally attributed by
both indoor and outdoor air quality (WHO, 2014).
Atmospheric emissions are composed of (in) organic compounds and
particulate matter.
Volatile organic compounds (VOCs) are part of organic compounds.
They are harmful
pollutants with the ability to form the undesired photochemical
tropospheric ozone smog and
potentially carcinogenic and mutagenic (Mohammed et al., 2013).
Also VOCS participate in
destruction of stratospheric ozone which protects us from UV
radiation (Mohammed et al., 2013).
Rwanda is an African developing country striving to transform
its economy on average to 11.5 %
of Gross Domestic Products (GDP) growth by 2018 (MINECOFIN
2013). To attain that goal,
industries are increasing day to day in the country but know-how
of handling emissions from
industrial activities is still lacking and there are no
available air pollution control technologies. To
the best of our knowledge, so far in Rwanda no studies have been
conducted to check the air
quality status in local manufacturing industries.
Therefore, to start bridging the gap, it is fortunate to
characterize VOCs emitted from local
manufacturing industries in Rwanda and evaluate the performance
of the cost effective
abatement technology which can be used to handle emission
emitted during production
processes.
-
2
CHAPTER 1 LITERATURE REVIEW
1.1 Volatile organic compounds
Air is an essential component of life on earth. Therefore, air
pollution is seen as a serious threat to
human being and the environment. Volatile organic compounds also
commonly shorten as
VOCs, are organic compounds usually distinguished based on two
groups definition; effect
definition and definition based on physical chemical properties
(Demeestere et al., 2007).
Firstly, effect definition, US EPA define VOC as any compound
containing at least one atom of
carbon, excluding carbon monoxide, carbon dioxide, carbonic
acid, metallic carbides or
carbonates, and ammonium carbonates which participate in
atmospheric photochemical
reactions (EPA, 2016). Secondly, based on physical and chemical
properties, Solvent Emission
Directive (SED) defines VOC as any organic compound having at 20
°C a vapor pressure of
0.01 kPa (Directive 1999/13/EC).
Methane is often viewed separately due to non-absolute
reactivity in the troposphere and
different concentrations range observed in different part of the
atmosphere
(Demeestere et al. 2007). Therefore, it is imperative to take
care of VOCs due to damage they
cause to both human health and as well as environment. They
contribute to major
environmental problems to mention, global warming, stratospheric
ozone depletion and
photochemical smog (Do et al., 2015). In presence of light, VOC
react with nitrogen oxides to
form tropospheric ozone which in high concentrations cause human
health problem
(Do et al., 2015). Many other VOCs like styrene and benzene are
said to be responsible for
numerous adverse health effects, mainly respiratory, heart
disorders and carcinogenic
(Stoji et al., 2015).
1.2 Sources of Volatile organic compounds
Sources of VOCs are divided into anthropogenic and natural
sources. Anthropogenic sources
are the man-made VOCs while natural are emitted naturally mostly
from vegetation. Often a
term biogenic is used to describe natural emissions of
non-methane hydrocarbons (Evuti, 2013).
Emissions distribution depends on the industrial activities,
climate and vegetation and varies
region to region (Evuti, 2013).
Globally, the biogenic VOCs, 1150 106 ton.yr-1 (Goldstein &
Galbally, 2007), emissions in remote
areas are almost 10 times higher than the anthropogenic VOCs,
142.106 ton.yr-1 (Müller, 1992),
per carbon per year in the forms of VOCs. The inverse happen in
urban places where
anthropogenic surpasses biogenic emissions concentrations
(Burrows et al., 2007).
-
3
1.2.1 Natural sources
VOCs are naturally emitted from vegetation (Guenther et al.,
2006). They account isoprenoids
(terpenes and monoterpenes) as well as alkanes, alkenes,
carbonyls, alcohols, esters, ethers and
acids (Guenther et al., 2006) .
The concentrations of the emitted compounds reveal isoprenoids
to be most prominent
compounds followed by alcohol and carbonyl compounds
(Kesselmeier and Staudt, 1999).
The oxidation of the biogenic VOCs produce products with low
volatility which participate in the
formation of Secondary Organic Aerosols (SOAs) (Kavouras et al.,
1998; O’Dowd et al., 2002;
Kanakidou et al., 2005; Jimenez et al., 2009).
SOAs have an important impact on air quality and climate (Fiore
et al., 2012; Scott et al., 2014).
To climate, SOA absorb and scatter solar radiation and they
indirectly affect the cloud
condensation (Gouw, 2009).
Methane is not accounted in the oxidation process although it is
produced naturally from
wetlands, rice field, livestock, landfills, biomass burning,
forests, termites and oceans; it’s total
emissions is in between 145 to 260 106 ton.yr-1 (EPA, 2016).
1.2.2 Anthropogenic sources
Human made emissions encounter indoor and outdoor environment
and they vary from various
sources (Bari et al., 2015). Indoor VOC concentrations are
generally found in higher levels than
the ambient outdoor levels (Fellin et al., 1994; Spengler, 1995;
Zhu et al., 2005; Heroux et al., 2008;
Stocco et al., 2008 ). Indoor VOCs are most emitted from
building materials (e.g, floor and wall
coverings, carpet, insulation, paint), combustion processes
(e.g, smoking, cooking, home
heating), consumer products (e.g, cleaners, solvents, air
fresheners, and mothballs), attached
garages, dry-cleaned clothing, municipal tap water, or personal
care products
(Wallace et al., 1987; Batterman et al., 2007; Stocco et al.,
2008; wheeler et al.,2013;
Ye et al., 2014).
According to European Environment Agency, the main sectors
involved in high VOC emissions
for the EU-27 are solvent and product use (41 %), the road and
no road transportation
(18%), and commercial, institutional and household associated
emissions (14 %)
(European environment Agency, 2010). Still in the EU-27 at the
industry level, the most occurred
VOC sources are in; (1) energy (41 %), (2) chemical industry (22
%) and (3) coating and surface
treatment activities (18 %) (European Pollutant Release and
Transfer Register, 2016) (Figure 1).
Ambient outdoor sources combine natural (e.g, vegetation and
fires) and anthropogenic
sources (e.g, evaporation processes associated with industry and
transportation, or paints and
solvents use) (Watson et al.,2001; Liu et al., 2008). In urban
atmosphere, motor vehicles exhaust
-
4
and evaporative emissions are reported to have higher VOCs
emissions concentrations than
other sources (Cetin et al.,2003; Lin et al., 2004).
Figure 1: Industrial sector VOC emissions in EU-27 (Adapted from
European Pollutant
Release and Transfer Register, 2016).
1.3 Hazards of VOCs
1.3.1 Human health effect
The exposure to higher permissible limit of VOCs concentrations
lead to acute or chronic health
effects (e.g, exposure to ceiling concentration for an 8 hours
shift than 500 ppm of toluene
causes headache and dizziness (Jiang et al., 2005) (Table 1).
But biochemical pathways and
physiological functions of most VOCs to human health are still
uncertain (Rudnicka et al., 2014).
Table 1: Related health effects to exposure of high VOC
concentrations.
Chemical compound Health effect
Benzene Carcinogenic
Ethers Producing peroxides, affecting the reproductive
system
Xylene Eye and respiratory tract irritation, narcotic
effect,
nervous system depression and death
Chloroform Affect central nervous system causing depression,
dizziness, liver and kidney damages, skin infection
Acetone and Acetaldehyde Respiratory and eye irritation
Phenol Offensive odor and toxicity
Epoxides Toxic, carcinogenic, explosive
N-containing compounds
(Amines)
Bad Odor, carcinogenic (affecting urinary bladder)
Source: Viswanathan et al (2007).
41%
18%
8% 1%
22%
6% 1% 3%
Energy Sector
Coating & Surface treatment
activitiesProduction and processing of metals
Mineral Industry
Chemical Industry
Paper and wood production
Waste and Wastewater
management
-
5
1.3.2 Tropospheric photochemical ozone formation
Troposphere is the region of the Earth’s atmosphere where people
reside and in which most
chemical compounds are emitted as a result of human activities
(Atkinson 2000). Nitrogen
oxides (NOx= NO + NO2), VOCs and sulfur compounds lead the
chemical and physical
transformation which results in the formation of tropospheric
ozone globally (Logan, 1994), acid
deposition (Schwartz, 1989). The production of ozone in
troposphere relies on the photolysis of
NO2 (Equation 1) and the subsequent association of the
photoproducts O(3P) with O2 via
(Equation 2) through the molecular reaction with the third body
(M being used to present any
third body co reactants, i.e N2) (Monks et al. 2015).
The mechanism reaction of tropospheric ozone formation is a
complex branched chain reaction
between the VOCs and NOx in the presence of light (Evuti 2013).
Equation 3 to 9 depicts the
mechanism reaction of tropospheric reaction. Ozone is first used
as source of hydroxyl radicals
(OH) (Monks et al., 2015) through
O3 + ℎ𝑣 → O2 + O(1D) (Eq.3)
O(1D) + H2O → 2OH∙ (Eq.4)
Where O(1D) is the electronic excited state atomic oxygen formed
through photolysis at
wavelengths
-
6
The tropospheric O3 produced from hydrocarbon reactions as well
as other sources is reportedly
to harm plants by reducing their growth due to limitation of
carbon dioxide in stomata of
vegetation (Felzer et al., 2007).
1.3.3 Stratospheric ozone depletion
Stratospheric ozone layer is known to protect lower part of the
earth’s atmosphere from high
frequency Ultraviolet (UV) light (Albritton, 1998). The ozone
layer is reduced as a result of
imbalance between the formation and loss of ozone, where
destruction is higher than
production (T et al., 2011).
Chlorine and Bromine released from man-made compounds such as
chlorofluorocarbons
(CFCs) (example of CFC is dichlorodifluorocarbon [CCl2F2]) prone
highly to the destruction of
stratospheric ozone (Angell et al., 2005). CFCs have long life
time in the atmosphere (10 to 120
years) (Angell et al., 2005). As a matter of fact, CFCs are
transported to the stratosphere where
they are eventually broken down by UV rays forming free chlorine
(Equation 10) which reduce
ozone to oxygen molecule (equation 11 to 13).
CCl2F2 + ℎ𝑣 → Cl∙ + CF2Cl
. (Eq. 10)
Cl. + O3 → ClO. + O2 (Eq. 11)
ClO. + O → Cl. + O2 (Eq. 12)
O3 + 0 → 2O2 (Eq. 13)
1.3.4 Global Greenhouse effect
Earth has the capacity to balance the absorption and emission of
solar radiations (Evuti, 2013). It
absorbs the energy in the form of ultraviolet, visible light and
infrared and emits the infrared to
outer space(Mohammed et al., 2012). Any process which interfere
with this balance result in the
phenomenon of global warming also termed as climate change or
greenhouse effect
(AEA group, 2007).
The Infrared (IR) absorption of atmospheric trace gases, water
vapor and carbon dioxide
(Derwent, 1995; Mohammed et al., 2012) disturbs the radiative
balance. Therefore, earth’s
surface and the atmosphere react to the disturbance by warming
to restore the radiative
balance. This process is termed as radiative forcing and the
warming is the greenhouse effect.
Halogenated compounds are also claimed to be powerful greenhouse
gases and deplete
stratospheric ozone, they are ozone depleting substances (ODSs)
especially compounds which
have chlorine and bromine attached on, hence, causing global
warming (Myhre et al. 2013)
-
7
The effect of the compounds to cause global warming compared to
carbon dioxide is
expressed in term of Global Warming Potentials (GWPs)(Evuti,
2013) (Table 2). The GWP is defined
as a ratio of the radiative forcing from a given mass emission
of the trace gas compared to that
from the same mass emission of carbon dioxide, integrated over a
given time horizon
(Mohammed et al., 2012).
Table 2: Global warming potential (GWP) of some VOCs in a
100-year time horizon.
Source: AEA group (2007).
1.4 Air pollution control technologies for VOCs
Many technologies have been introduced for VOC emission control.
The available techniques
are basically classified into two different categories: (i)
process and equipment modification and
(ii) add on control technique (Khan & Ghoshal, 2000) (Figure
2). In the first category, control of
VOC emissions are made by modifying the process equipment, raw
material, and / or change
the process (Khan & Ghoshal, 2000).
On the other hand the latter category, require an additional
control method to regulate the
VOC emissions. It has two subgroups dubbed destruction and
recovery of VOCs
(Khan & Ghoshal, 2000; Delhoménie & Heitz, 2005).
VOC GWP VOC GWP
Carbon dioxide 1 Dimethylether 1
Bromomethane 5 propylene 4.9
Propane 6.3 ethylene 6.8
Butane 7 1,1- Difluoroethane 122
Ethane 8.4 Difluoromethane 670
Dichloromethane 10 1,1,1,3,3,-Pentafluorobutane 782
Chloromethane 16 1,1,1,3,3-Pentafluoropropane 1020
Dichlorotrifluoroethane 76 1,1,1,2-Tetrafluoethane 1410
Dichloropentafluoropropane 120
1,1,1,2,3,4,4,5,5,5-Decafluoropentane 1610
1,1,1-Trichloroethane 144 1,1,1,2,3,3,3 Heptafluoropropane
3140
Dichlorotetrafluoroethane 599 Pentanfluoromethane 3450
Dichlorodifluoroethane 713 1,1,1-Trifluoroethane 4400
Chlorodifluoromethane 1780 1,1,1,3,3,3- Hexafluoropropane
9500
Chlorodifluoroethane 2270 Trifluoromethane 14310
-
8
*RFR: Reverse Flow Reactor
Figure 2: A tree diagram of the VOC emissions abatement
technology (Khan & Ghoshal, 2000).
The adaptation or choice of technology lies on the operating
conditions (flow rate,
temperature, humidity and VOC concentrations) and the pollutants
physico-chemical
characteristics (solubility, vapor pressure, biodegradability
level and inflammability)
(Crocker & Schnelle, 1998). An illustration is given in
Figure 3, of the application limit for flow rate
in function of VOC concentrations of destruction and recovery
technologies.
Figure 3: Application limit of flow rate vs VOC concentrations
of different air pollution
technologies control (Delhoménie and Heitz 2005).
VOC removal technique
Process and equipment
modification
Condensation Oxidation
Add on control techniques
Destruction Recovery
Absorption Adsorption Biofiltration Membrane
separation
Thermal
oxidatio
n
RFR*
8
Catalytic
oxidation
Activated carbon based adsorption Zeolite based adsorption
-
9
Biofiltration is the only biological VOC treatment technology
found in recovery technologies. The
remaining ones found in recovery and all destruction control
techniques are non-biological
treatment technologies.
1.4.1 Non-biological technology for VOCs
As indicated in Figure 3, the non- biological technologies are
physical chemical technologies
generally applied to reduce off-gas with high VOC emission
concentrations. The lower VOC
concentrations in the flue gases the higher energy input will be
required to get rid of the VOC
especially in the oxidation processes (incinerations)(Khan and
Ghoshal, 2000). In terms of cost,
physical chemical treatment technologies for VOCs emission
involve higher cost than biological
treatments (Font and Artola 2011).
1.4.2 Biological treatment of VOCS
The biological treatment technology (biotechnology) for VOCs
emissions was introduced first by
the European countries (Germany followed by The Netherlands), in
1960 (Leson & Winer, 1991;
Cloirec et al, 2005). The fundamental purpose of that biological
treatment technology was to
handle odor and VOC emission at the industrial scale
(Álvarez-hornos et al., 2011).
In 2003, the European IPPC reported that vapor-phase
biotechnologies, including biofilters,
biotrickling filters and bioscrubbers, have proven to be more
environmental friendly and chosen
as best available technologies for the reduction of the VOC
emissions in chemical sector
(European commission, 2003).
Hence, biological gas treatment (biotechnologies) are seen as
potential alternative to the
conventional physico-chemical processes for removal of VOCs with
high flow rate emission
streams with relative low VOC concentrations; conditions
observed more particular in painting,
coating and printing processes (Álvarez-hornos et al.,
2011).
Biotechnologies or biological VOC treatment technologies rely on
the capacity of
microorganisms of using their metabolisms to translate the
organic pollutants to less harmful
compounds. Since the pollutants are in gas phase, they have to
be transferred in aqueous
phase to be ready and used by microorganism (Álvarez-hornos et
al. 2004).
The overall degradation process of the biofiltration is
presented in Equation 14
( Álvarez-hornos et al., 2011; Font & Artola, 2011).
Organic pollutant + O2 CO2 + H2O + heat + biomass + other
byproduct (Eq. 14)
microbes
-
10
The main types of biological treatment of VOC emissions include
biofilters, biotrickling filters and
bioscrubbers (Delhoménie and Heitz 2005). The basic idea for the
removal of VOCs mechanism
for these three biological technologies is similar but there
notable differences with regards to the
aqueous phase and microorganism growth (Álvarez-hornos et al.,
2011) (table 2).
Table 2: Classification of vapor phase biotechnology
systems.
Biotechnology system Microorganism growth Aqueous phase
Biofiltration Attached growth Stationary
Biotrickling filter Attached growth Flowing
Bioscrubber Suspended growth Flowing
Source: Álvarez-hornos et al (2011).
1.4.2.1 Biotrickling filters
In biotrickling filters, biodegrdation happen when the gas is
first transferred to the biofilm which
grow to the packing materials. The packing materials are made
from chemical inert materials
such as plastic rings (Waweru et al,. 2000), resins, ceramics,
celite, polyurethane foam
(Yamashima and Kitagawa, 1998), and no nutrients are available
in such materials for
microorganism to grow. Nutrients are continuously supplied from
the top to bottom in
countercurrent with the flue gas, the leachate is collected at
the bottom and recycled back up
(Berenjian et al., 2012)( Figure 4). This feeding process
facilitates control of biological operating
parameters like nutrients and pH (Muñoz et al., 2015). Soluble
VOC are reported to be highly
removed by biotrickling flters (Berenjian et al., 2012).
The major bottleneck of this system is the clogging of excess
biomass in the filter bed and
research has developed three major solutions, mechanical,
chemical and biological
(Delhoménie and Heitz 2005). Mechanical by bed stirring (Wübker
et al., 1997;
Laurenzis et al., 1998) or bed backwashing with water which
allows drainage of the excess
accumulated biomass (Smith et al., 1996).
Chemical treatment to breakdown the chemical bindings between
biomass and bed particle
by using disinfecting reagents (Diks et al.,1994; Schönduve et
al., 1996; Cox and Deshusses, 1999;
Armon et al., 2000; Chen and Stewart, 2000). Biological use
biomass predators such as protozoa
(Cox and Deshusses, 1997). Amongst all these solutions
mechanical treatment using water for
backwashing is claimed to be most efficient and at least
friendly to the ecosystem
(Cai et al., 2004).
-
11
Figure 4 : Biotrickling filter setup (Delhoménie and Heitz,
2005).
1.4.2.2 Bioscrubber
The bioscrubber contains two reactors, the absorption tower and
bioreactor. In the absorption
tower, the gas is absorbed or diffused into aqueous solution via
the countercurrent gas-liquid
flow through the inert packing materials. Packing material
within the absorption tower provides
a better surface transfer between VOC and aqueous phase (Kellner
and Flauger, 1998)
(Figure 5). The washed off or clean gas flow to the top and the
contaminated liquid is pumped
in the bioreactor (Berenjian, Chan, and Malmiri 2012). The
bioreactor is inoculated with
degrading constrains in aqueous phase and contains nutrients
essential for their growth and
maintenance
(Delhoménie and Heitz 2005). The major limitation of
bioscrubbing system is that they are applied
to only soluble contaminants with low Henry’s constant (
-
12
Figure 5: Bioscrubber setup (Delhoménie and Heitz, 2005).
1.4.2.3 Biofilter
This is the most basic biological treatment process that uses
organic packing materials in which
culture of microorganisms are developed to degrade pollutants
into less harmful compounds.
The contaminated air pass through a biofilter packed with
organic carrier materials where
biofilm are fixed (Figure 6). Before the inlet gas stream enters
the filter bed, it is pre-humidified to
avoid clogging in filter bed (Waweru et al., 2000).
Biodegradation happen when the pollutant is
first transferred from gas to liquid phase. In the liquid phase,
the pollutant is either absorbed in
water or adsorbed on the packing material. The unavailable
pollutants for biofilm diffuse through
the filter bed.
Figure 6: A typical setup of biofilter (Delhoménie and Heitz
2005).
Aqueous solution Clean air
Bioreactor
Waste solutions containing pollutants
Polluted air
Activated sludge, suspended in
nutrient solution
Absorption
column
Treated air
Nutrient solution
Occasional irrigation
Waste solutions possible recycling Polluted air
Bed packed
with organic
materials
-
13
The successfulness of microorganisms to degrade pollutants
depends on a good follow up of
physical, chemical and biological parameters of the biological
system:
(I) Filter bed, is an important part of the biological treatment
process because they support the
growth of microorganism communities responsible for pollutants
degradation and increase the
contact between the gas and biofilm (Iranpour et al., 2005;
Kennes et al., 2009).
A good packing material should have a high specific area
favorable for microbial activity, good
water retention to avoid dehydration, high porosity to provide a
homogeneous gas distribution
entirely into bed, availability of intrinsic for nutrients and
diverse microflora
(Delhoménie & Heitz, 2005; Berenjian et al., 2012).
The most used organic packing media are compost, peat, soil, and
at smaller scale woodchips
and bark(Easter et al., 2005; Delhoménie & Heitz, 2005;
Gabriel et al., 2007). Studies made on
woodchips or bark found that these packing materials are less
satisfactory as compared to peat
and compost because of their low pH buffering capacity, low
specific area and nutrient
availability (Smet et al., 1996a; Smet et al., 1999; Hong &
Park, 2004).
(II) Moisture content is a crucial parameter for effective
filter bed as microorganisms need water
to carry their metabolic activity (Shareefdeen et al., 2005).
Less bed moisture content lead to
dehydration and gas channeling which affect particularly the
microflora
(Delhoménie and Heitz 2005). On the contrary too much water in
the filter bed cause flooding
which leads to compaction and anaerobic conditions (Delhoménie
and Heitz 2005). The
moisture content of the overall carrier materials must have a
value between 40 and 60 (w/w)
(Ottengraf 1986; Waweru et al., 2000).
(III) Temperature, microbial activity also depends on the
biofilter operating temperature. The
microbial growth in biological systems works at a temperature
between 10 and 40oC
(Cloirec et al., 2005). Most of microorganisms grow in the
biofilter are mesophilic
(Kennes & Thalasso, 1998) at a temperature ranging between
20 and 40 oC. And this
temperature ranges define the optimum temperature in biofilters
(Delhoménie and Heitz 2005).
(IV) pH, to support the microbial growth a pH range from 5 to 9
is normally used and the stability
of this parameter in the biofilter increase the microbial
activity (Cloirec et al. 2005). The optimum
pH is around neutrality, pH≈7 (Delhoménie and Heitz 2005).
Compounds containing heteroatoms
(sulfur, chlorine and nitrogen) are oxidized to acid by-products
which in turn lower the pH of the
biofilter (Devinny and Hodge, 1995; Christen et al., 2002). The
effect of pH on biofiltration
efficiency depends on types of microorganisms (Clark et al.,
2004). Fungi has the ability to grow
-
14
at both neutral as well as acidic medium conditions and they are
metabolically active at pH
approximately between 2 and 7 (Delhoménie and Heitz 2005). On
the other hand, bacteria are
very sensitive to pH, they are less tolerant to pH below 7
(Kumar et al., 2011). Two methods used
by authors to maintain pH to neutrality are either to irrigate
the biofilter by nutrients solution
which have buffer capacity or insert the buffer materials in
biofilters
(Delhoménie and Heitz 2005). Nevertheless, the ideal pH of the
biofilter medium depends on the
pollutant being treated and the characteristics of the microbial
ecosystem (Kumar et al., 2011) .
(V)Nutrient requirement, aerobic microorganisms’ performance in
biofilter depends on the
availability of nutrients. The most elements needed for the
growth of the biomass are nitrogen,
phosphorous, potassium, sulfur and trace elements in additional
to oxygen and carbon
(Álvarez-hornos et al., 2011). For the long term performance of
the biofilter an additional of
nutrients is required (Yang et al., 2002). Due to the importance
of nitrogen towards biomass
growth, an additional of nitrogen to the biofitler media is
reported to enhance the performance
of the biofilter (Morales et al., 1998).
(VI) Bed porosity, this is an essential parameter which
maintains even air flow rate and decrease
the pressure drop across biofilter (Álvarez-hornos et al.,
2011). The filter bed which used only
compost as packing material report to be 44.4 % for dry compost
and 39.6 % for wet compost
(Douglas & Devinny, 1997). To increase the porosity and
decrease degree of compaction in the
bed filter, a mixture of packing materials are used (Bohn,
1992).
(VII) Inlet pollutant concentration, obviously biofilter perform
best for treating pollutant which are
in concentration less than 1000 ppm. High inlet VOC
concentration in the biofilter lead to
inhibition of microbial activity (Álvarez-hornos et al., 2011).
Also, high inlet concentrations lead to
insufficient oxygen availability in biofilter (Ottengraf, 1987).
Studies have found that 30 ppm of
toluene had 99% removal efficiency but when doubled its
concentration, the removal get down
to 82% removal efficiency (Álvarez-hornos et al., 2011).
(VIII) Microorganisms and acclimation time, the natural organic
packing material used in bed
media parent microorganism in biofiltration. Microorganism such
bacteria and fungi are used for
the degradation of VOCs (Kumar et al., 2011). The degradation of
the pollutant depend on the
nature of the filtering materials and the biodegradability level
of VOC to be treated
(Kumar et al., 2011). A single type of microorganism is enough
to degrade certain pollutants and
for certain group of pollutants or even a culture of
microorganism is used (Nanda et al., 2012).
-
15
Compost has been reported to use bacteria belonging to a group
of Proteobacteria,
Actinobacteria, Bacteroidetes and Firmicutes (Chung, 2007).
An acclimation, time required obtaining stable high removal
efficiency over a long time, for
microorganism to handle new substrate environment may take 10
days to 10 weeks
(Ralebitso et al., 2012). Introduction of inoculum to the bed
media can shorten the lag phase
(Álvarez-hornos et al., 2011). A typical biofilter usually
contains 106-1010 cfu of bacteria and
actinomycetes per gram of bed and fungi in the range of 103-106
cfu per gram of bed
(Ottengraf, 1987). Degrading species in a biofilter are normally
between 1 and 15 % of the total
microbial population (Pedersen et al., 1997; Delhomenie et al.,
2001).
(IX) Empty bed residence time (EBRT), both air flow rate and
EBRT are important parameters with
reasonable impact on the biodegradation performance of the
biofilter (Elmrini et al., 2004).
Increasing EBRT will produce high removal efficiency. EBRT can
be relied on to increase the
biofiltration performance and should be greater the time needed
for diffusion processes for low
operating flow rate (Álvarez-hornos et al., 2011).
1.4.3 The biodegradation of hydrophobic compounds
Hydrophobic compounds have high Henry’s constant as compared to
hydrophilic compounds.
That said, they are less soluble in water than hydrophilic
compounds a factor which make them
often hard to reach the biofilm layer in biofilter thus
providing low removal efficiency.
The composition of the filter materials is a critical parameter
for effective biofilter toward the
removal efficiency of the hydrophobic and less soluble
compounds. Studies conducted on
biofiltration of hydrophobic compounds proposed that improved
adsorbing materials such as
granular activated carbon (GAC) may have characteristics that
may promote higher
elimination capacity particularly for compounds with low
solubility that emitted in variable loads
(Tonekaboni, 1998). Also, a way suggested by researchers to
reduce solubility and transport of
hydrophobic compounds into filter bed was the use of Fungi
(Woertz et al., 2001;
García-Peña et al., 2001). García-Peña et al (2001) described
elimination capacity for toluene
up to six times higher than usually reported for bacteria using
Paecilomyces variotii. Again for
hexane, which is around 100 times less soluble than toluene, EC
between 100 and 150g.m-3.h-1
were obtained by Aspergillus níger (Spigno et al., 2003), while
only between 10 and 60g.m-3.h-1
have been reported with bacterial consortia (Budwill and
Coleman, 1999; Paca et al., 2001;
Kibazohi et al., 2004).
-
16
1.4.4 Parameters used to check the performance of the biological
systems
The most common parameters used to check and compare the
performance of biological
systems are summarized below (table 3)
Table 3: Performance parameters used in biological treatment
systems.
Parameter Formula [unit] Description
EBRT EBRT =
V
Q[𝑠]
EBRT is the time taken by a gas in the biofilter.
Where V= Volume of the reactor (m3)and Q =
the flow of the gas (m3.h-1)
Inlet Load (IL) IL =
Q
Vx Cin[g. m
−3. h−1]
This is the amount of the pollutant introduced in
biofilter per unit volume per time. Where Cin is
concentration of pollutants in the inlet gas
stream (g.m-3)
Elimination
capacity
(EC)
EC
=Q
V (Cin − Cout)[g. m
−3 . h−1]
This is the amount of the pollutant removed per
volume of a filter bed per unit time
Removal
efficiency
(RE)
RE =(Cin−cout)
Cin x 100 [ %] This is the amount of the pollutant removed
in
fraction converted in percentage.
Source: Waweru et al (2000).
1.5 Scope and objectives
Rwanda is a landlocked country whose economy has shown to be
increasing since the tragedy
of the 1994 Tutsi’s Genocide (MINECOFIN 2013). Despite the
tragedy of Tutsi’s Genocide,
population density (people per km2) is increasing year to year
(449 in 2010 and 460 in 2015)
(World Bank, 2016). The basic country’s economic transformation
is helped by the industrial,
service delivery and mining sector (Rwanda national institute of
statistics, 2011). Industrial
activities and traffic are believed to be the main contributors
of high atmospheric emissions in
the country especially in the capital city, Kigali.
Prior to industrial emissions, there are no available abatement
technologies for the already
implemented industries. In addition to that, no studies have
been conducted before to check
the status of emissions made in different local manufacturing
industries. This is a common
problem shared by almost all African countries where data on the
concern of air quality status
are hardly or not even found (Do et al. 2013).
-
17
To start bridging the gaps, a VOC study was conducted to make a
new qualitative and
quantitative data in three different local industries namely
Sulfo Rwanda industries producing
soap and cosmetic, Atelier Des Meubles de Kigali (AMEKI) making
paints and Inyange industries
producing milk and juices by means of active sampling using
Tenax TA tubes and TD-GC-MS
analysis.
This study is divided into two main objectives:
1. To characterize the VOC emitted in three local manufacturing
industries, Sulfo Rwanda
Industries, AMEKI color and Inyange industries. Specific
objectives on this first part are:
To characterize VOCs emitted in three industries
To identify and comparing the most occurring compounds from
Indoor to outdoor in all
industries.
2. To evaluate biofiltration for the cost effective treatment of
the waste gases containing
important pollutants, focus given to acetone, dimethyl sulfide
and hexane.
Specific objectives on are:
To compare removal efficiencies of the three compounds (acetone,
dimethyl sulfide and
hexane) in a biofilter packed with compost, silicon foam and
wood chips.
To check the effect of using adsorbing materials, silicon foam,
for the removal of
hydrophobic organic, hexane and assess inhibitory effects.
To check the partition coefficient of target VOC to the packing
materials.
-
18
CHAPTER 2 MATERIALS AND METHODS
This chapter is split into two parts; Part (I) is the analysis
of the industrial VOC concentrations
sampled in Rwanda. Samples were taken at three different local
manufacturing industries by
means of active sampling using Tenax TA sorbent tubes. After
sampling, they were transported
to the environment organic chemistry and technology lab for
analysis. Part (II) is the Technology
based part, where biofiltration a cost effective abatement
technology was used to evaluate the
removal efficiency of VOC where focus was given to acetone,
dimethyl sulfide (DMS) and
hexane as representative VOC.
PART I: ANALYSIS OF INDUSTRIAL VOC CONCENTRATIONS IN RWANDA
2.1 Tenax TA tubes
Tenax TA tubes are tubes filled with sorbent resin (2, 6
diphenylene oxide) to capture VOCs and
semi-VOCs. They have standard dimensions of 1/4 inch (6.4 mm
outer diameter x 5 mm internal
diameter), the length of 3.5 inch( 89 mm) and are filled with
200 mg of Tenax TA) (Anon, 2011).
Tenax TA tube can be heated up to 350 °c, has low affinity for
water and has a specific surface
area of 35 m2.g-1 and average pore size of 200 nm based on
Scientific Instrument Services
(SIS, 2016). The tubes are closed with 1/4 inch brass closure
caps (Anon, 2011). The cap has a
white Teflon ferrule (Alltech SF-400T) that creates a better
airtight seal. Each tube has an external
groove which indicates the sampling side (Do et al, 2009). The
tube can be used more than 100
and after that period the resin should be replaced out of
precaution (SIS, 2016).
2.2 Conditioning
Prior to sampling, the Tenax TA tubes should be conditioned to
make sure that no residual
components remain on sorbents. Stainless steels, Tenax TA tubes,
were put in an oven at 300 oc
for an hour under the flow of 10-50 ml.min-1 helium, to remove
all residual components. During
heating oxygen should be avoided to enter since it is
detrimental to the adsorbent resin. The
Tenax TA tubes in the oven are positioned with the sampling side
mounted out. The oven
(Carlo Erba Instruments, MFC 500) is capable of heating nine
tubes all at once.
-
19
Figure 7: Conditioning oven
2.2.1 Loading with internal standards
2.2.1.1 Chemicals
Deuterated Toluene (Tol-d8; 99.5%; Acros organics, Geel,
Belgium) was used as an internal
standard (IS) (Figure 8).
Figure 8: Tol-d8 structure.
The solvent used for IS is methanol (LC-MS grade, 99.5%,
Biosolve, Valkenswaard, Netherlands).
Tenax TA sorbet tubes have low affinity for methanol that’s why
methanol was used as a solvent.
The stock solution of 223.7 µg.ml-1 was prepared by putting 24
µL of Tol-d8 in 100 mL of methanol
then kept in total darkness at -18 oC.
2.2.1.2 Preparation of the closed two-phase system
To prepare a gaseous Tol-d8, 20 μL of stock solution was added
to 20 ml of deionized water
present in 119.8 mL of a glass bottle. Then, the bottle
containing a mixture of stock solution and
deionized water was gas tightly sealed with a mininert valve
(Alltech, Lokeren, Belgium) and
incubated in a thermostatic water bath at 25 ± 0.2 oC for 12
hours to assure the equilibrium
between gas and the liquid phase is reached.
D
CD3
D
D
D
D
-
20
2.2.1.3 Calculation of the headspace concentration
A given gas and water volumes at a specific temperature with
known total mass and Henry’s
law constant of Tol-d8 (Dewulf et al., 1996), the headspace
concentration of IS can be
calculated from the mass balance equilibrium. First, the total
mass of Tol-d8 (m total) added to
the CTS was equal to 4474 ng. Based on the mass balance and
Henry constant of Tol-d 8 at
25 oc (H= 0.183), we can derive,
Equation 15 can be rewritten into Equation. 17
4474 ng =Cair
0.183∗ Vwater + Cair ∗ Vair
(Eq. 17)
Then with Vwater= 20 mL; Vair =99.8 mL, the concentration of
Tol-d8 in the headspace can be
calculated as 21.4 (ng.mL-1). This means that 0.5 mL air in the
CTS contains 10.7 ng of Tol-d8.
2.2.1.4 Loading
In CTS, 0.5 mL of headspace was taken by 0.5 ml gastight
pressure lock VICI precision analytical
syringe (Series A, Alltech). The desired volume was loaded onto
the sorbent tubes through an
injection system flushed with helium (He) (flow rate of 100
mL.min-1). And finally, the He stream
was held on for 3 min before the tubes were sealed with ¼ inch
brass long term storage
endcaps, equipped with ¼ inch one-piece PTFE ferrules.
2.2.2 Pump Calibration
A Gil Air sampling pump was calibrated before use by Gilibrator
to make sure the targeted flow
rate is at least repeatedly obtained, and it was regulated on an
average flow rate of 100 ± 0.5 %
mL.min-1 (n = 8).
2.3 Sampling Campaigns
The sampling campaign of VOCs in Rwanda was held at three
different local industries namely
Sulfo Rwanda industries, AMEKI color and Inyange industries
producing cosmetics, paints and
beverages respectively (Figure 9).
mtotal = mair + mwater=cair × Vair+cwater × Vwater ( Eq. 15)
H =Cair
CWater=0.183 mol.L−1 mol.L−1⁄ ( Eq. 16)
-
21
The sampling campaign was performed by means of active sampling
using Tenax TA sorbent
tubes (n = 1) at indoor and outdoor of the three local
manufacturing industries on 16th July and
30th July 2015 (Table 4). Two samples (one for three minutes and
another for 30 minutes) were
sampled indoor and outdoor at each sampling site and two blanks
were among sampling Tenax
tubes which remained closed entirely the whole campaign. The
three minutes samples are the
one which were analyzed as they were found to be loaded with
enough VOC concentrations
for measurements.
(1) Sulfo soap production unit (2) Sulfo cosmetics production
unit, (3) AMEKI color for paints (4) Inyange for
juice and milk processing (beverages)
Figure 9 : Sampling locations on the map of Kigali city.
2
1
3
4
1
2
3
4
40 km
-
22
Table 4: The overview information of the VOC sampling campaign
collected at three local
industries in Rwanda.
Indoor Outdoor Date Time Sample
size
Sampling times
1. Sulfo Rwanda Industries
16-07-2015
11:00-13:50
8 I. Soap production 30 30
3 3
II. Cosmetics 30 30
3 3
2. AMEKI color 30 30 16-07-2015 14:30-16:00 4
3 3
3. Inyange Industries 30 30 30-07-2015 16:00-17:45 4
3 3
2.3.1 Description of the sampling locations
2.3.1.1 Sulfo Rwanda Industries
Sulfo Rwanda Industries is a local manufacturing industry
producing drinking water, hard and soft
soap, and cosmetics (body lotion, glycerin and toilet soap). The
industry has four production
units. Samples were taken in late morning between 11:00 and
13:50 on 16th July, 2015 at two
manufacturing units, hard soap production and cosmetics
unit.
The two production units are built in the middle of a busy place
downtown in the capital city; on
the street of soap production unit there is a lot of traffic,
big public parking lot at the back and
like in 300 m there is also a big city market hall and other
many retailing business around it. The
cosmetic production unit is close to a national museum and
prison (Figure 10) and (Figure 11).
Figure 10: Soap production unit. Figure 11: Cosmetics production
unit.
-
23
2.3.1.2 AMEKI Color
AMEKI color is a paint production company located at the
industrial park; it neighbors different
industries and is no far away from a polyclinic and petro
station. Samples were taken in
afternoon between 14:30 and 16:00 on 16th July, 2015.
AMEKI color is the leading paint local manufacturing industry in
Rwanda. Paints produced are;
(1) latex matt, based on styrene acrylic emulsion, suitable for
ceilings and walls internally or
externally whether new or previously painted and (2) Silk Vinyl
emulsion, water-based emulsion,
noted to be environmental friendly (Figure 12).
The company is mostly manual based paints production and
permanent employees are in
direct contact with the raw products for paint making. Diesel is
entirely used for cleaning all used
materials in the process.
Figure 12: Paint production.
2.3.1.3 Inyange Industries
The Inyange industry is located outside the capital, there is no
high traffic as compared to the
city center and it is built in the lowland close to marchland.
There is little habitation across the
industry. Inyange is the first ranked industry for the
production of beverages in the country. Their
daily production is drinking water, juices, milk packaging and
milk processing. All production
units are combined in the same site. After the production and
packaging, caustic soda is
passed in the tanks for cleaning purposes. Samples were
collected in the afternoon from 16:00 to
17:45 on 30th July, 2015 in the milk and juice production units
(Figure 13).
-
24
Figure 13: Juice and milk production unit.
2.4 Analysis of Tenax TA sampling tubes
2.4.1 TD-GC-MS
Tenax TA tubes used in three sampling campaign location in
Rwanda were transported to
laboratory and analyzed by TD-GC-MS in a method described by Do
et al (2009) (Figure 14).
Figure 14 : The TD-GC-MS with (1) the TD (2) the transfer line
from GC to MS,
(3) GC and (4) MS.
The desorption of analytes pre-concentrated on the Tenax TA
sorbent tubes was performed by
a unity 2 thermal Desorption system (Markes, Llantrisant, UK) at
260 oC with 20 mL.min-1 helium
flow for 10 minutes. Each Tenax TA tube was put in the TD system
equipped with two special
diffusion caps.
Desorption process was first pre-purged at 34 oC for two minutes
to make sure that water vapor is
eliminated inside Tenax TA tubes and then a temperature of 260
oC for 10 min was set to tube
desorption. Next, analytes were refocused on a microtrap 100 %
Tenax TA (noC-TNXTA)
(Markes, Llantrisant, UK) cooled at -10 oC. The Tenax TA tubes
were heated up sharply from
-10 oC to 280 oC within three minutes.
2
4
3
1
-
25
Analytes were carried by a helium flow and injected onto a 30 m
factor four VF-1 ms low bleed
bounded phase capillary GC column (Varian, Sint-Katelijne-Waver,
Belgium;
100 % polydimethylsiloxane, internal diameter 0.25 mm, film
thickness 1µm), after splitting helium
flow at 5 ml.min-1. The column head pressure was set at 50 kPa,
resulting into a flow of
1.0 mL.min-1 (at 130 oC) through the GC column.
The GC (Focus GC, Thermo Scientific, Italy) oven temperature was
initially set at 35 oC for
10 minutes. Next, the temperature in the GC was increased
gradually up to 240 oC into four
stages (1) from 35 to 60 oC (2 oC.min-1), (2) from 60 to 170 oC
(8 oC.min-1), (3) from 170 to 240 oC
(15 oC. min-1) and (4) (240 oC) was held for 10 minutes before
cooling down to 35 oC. Even
though GC was cooled down, the transfer line from GC to MS was
kept at 240 oC.
Mass from m/z 29 to 300 were recorded in full scan mode (200 ms
per scan) on a DSQ II.
Quadrupole MS (Thermo Scientific, Austin, TX, USA), hyphenated
to the GC, and operating at an
electron impact energy of 70 eV. Chromatograms and mass spectra
were processed using
Xcalibur software (Thermo Finnigan, version 2.2)
Compound identifications were predicted based on (i) their
fragmentation patterns and by
comparison of their mass spectra with the US National Institute
of Science and Technology (NIST,
Gaithersburg, MD, USA) 2.0 database [NIST/US Environmental
Protection Agency (EPA)/US
National Institute of Health (NIH) Mass Spectra Library], and
(ii) comparison of retention time with
the standards.
2.4.2 Calibration of the TD-GC-MS
To have the correct concentration values, calibration was
performed. A total set of 78 LC-MS
grade standard VOCs were used for the calibration.
These VOCs were purchased at the Acros Organics (Geel, Belgium)
and/or at Sigma-Aldrich
(Bornen, Belgium) and all had a purity of at least 99.8 %.
Methanol (LC-MC grade, 99.95 %,
Biosolve, Valkenswaard, Netherlands) served as a solvent for all
standard compounds.
The 78 standards compounds were prepared and divided into three
stock solutions (A, B and C)
(in methanol) together with a known mass of Tol-d8. A known
volume 1µL of the stock solution
was loaded on two Tenax TA sampling tubes corresponding with a
loaded mass between 31.3
and 81.2 ng. For each calibration of the TD-GC-MS, there were
six calibration files corresponding
with 2x3 sampling tubes. The six tubes were then analyzed in the
TD-GC-MS in full scan mode.
The quantification, data were processed by an extracted ion
chromatogram.
-
26
The sample response factor (SRFi) in the chromatography is
defined as the signal output per unit
mass of the substance injected. Therefore, SRFi can be
calculated based on the Equation 18.
Where, Ai is the peak area and mi the mass (ng) of the substance
i on the sorbent tube. Based on
the concept of SRF, the RSRF (Relative Sample Response Factor)
is defined as the ratio of sample
response factor of the analyte (SRFa) and Tol-d8 (SRFst).
It is worth noting that RSRF is dimensionless. During the
calibration of the TD-GC-MS, both
analyses and Tol-d8 were loaded from the liquid phase.
Demeestere et al. (2008) have found RSRF L,L (both loaded from
liquid phase) and RSRF G,G (both
loaded from air phase) are the same.
2.5 Quantification
2.5.1 Calculation of the analyte concentration
Since we know (i) the mass of the IS (mst = 10.7 ng as
calculated in section 2.2.1.3 from CTS (ii) the
peak areas of the analyte and the IS and (iii) RSRFL,L from the
calibration of TD-GC-MS we can
depict the mass of the target compound (ma) on the sorbent in
active sampling.
The Volume of the sampled air can be obtained from Eq.23.
SRFi =Aimi
(Eq.18)
RSRF =SRFaSRFst
(Eq.19)
RSRFL,L =SRFaSRFst
(Eq.20)
RSRFL,L ≈ RSRFG,G =SRFaSRFst
(Eq.21)
ma =mast × Aa
Ast × SRFL,L
(Eq.22)
-
27
And Qsample is the flow rate of the sampling pump and tsample is
the sampling residence time. And
the concentration can be calculated as:
All components’ peaks were determined from respective extracted
ion chromatographs and
the RSRFL, L was determined from the calibration of the
TD-GC-MS. A blank correction was also
performed during the concentration calculations.
V = QSample × tsample (Eq.23)
Ca =maV
(Eq.24)
-
28
PART II: ABATEMENT TECHNOLOGY
2.6 BIOFILTRATION
Given state of the art cost effective abatement technology,
biofiltration was used to evaluate
the removal efficiency of VOCs where focus was given to three
compounds, acetone, dimethyl
sulfide and hexane.
2.6.1 Physical chemical properties of the representative VOC
compounds
The most important physical chemical properties describing
acetone, dimethyl sulfide and
hexane are summarized in Table 5.
Table 5: Physical chemical properties of acetone, dimethyl
sulfide and hexane.
Parameters Acetone DMS Hexane
Functional group Ketone Sulfur compound Alkane
Molecular weight (g.mol-1) 58.08 62.13 84.17
Odor threshold(ppmv)2 42 0.003 1.5
Boiling point (0C)2 46.5 29.5 68.5
Vapor pressure at 25 0C (mmHg)3 231 502 153
Solubility in H2O at 25 oC( g.L-1)2 94 45 0.016
Henry’s law constant (-) (Cg/CL)1 0.012 0.048 44
1Calculated using the solubility and vapor pressure
2 (Nagata, 2003); 3 (SciFinder, 2016); (Daubert, 1989)
2.6.2 Biofiltration process
2.6.2.1 Biofiltration design
The biofiltrer was built using six identical cylindrical modules
of Plexiglas and the total height of
the setup was 1.2 m with internal diameter of 10 cm. The packing
materials in the biofilter
occupied only 1 m height and the remained 0.2 m at the bottom
was occupied by glassbits
purposed to homogenize gas flow streams before enter the filter
bed.
2.6.2.2 Biofiltration setup
The biofiltration setup used in the experiment was divided into
three major parts: (A) generation
of the flow air controlled by mass flow controllers (B) the
filter bed equally packed by compost,
woodchips and silicon foam and (C) analysis of VOC
concentrations and CO2 produced by
-
29
SIFT-MS (Syft technology, the Voice 200®, Christchurch, New
Zealand) and Vaisala CARBOCAP®
hand held carbon dioxide analyzer (GM70 model, vaisala, Finland)
respectively (Figure 15).
Figure 15: Schematic diagram of the biofiltration setup. (1)
pressure regulator (PR), (2) mass flow
controllers (MFC1; Q1 = 5 L.mim-1, MFC2 ; Q2 = 2.5 L.min-1,
MCF3; Q5 = 0.2 L.min-1 and
Q3= Q1+Q2) (3) acetone bottle, (4) DMS bottle, (5) hexane
bottle, (6) pressure control valve, (7)
humidifier column, (8) filter bed, (L1) leachate collection
port, (P1) inlet port, (P2), (P3), (P4) and
(P5) are intermediate ports, (P6) outlet port, (9) flow
monitoring valve, (10) rotameter, (11) clean
gas exit.
The aim of the setup was to measure the overall performance of
the filter bed polluted with
acetone, DMS and hexane. The polluted air flow streams were
generated by passing streams of
air into capillaries attached to liquid bottles filled with
acetone, DMS and Hexane
(Acros Organics) controlled by mass flow controllers (Brooks
Instruments, Mass flow controllers®,
Hartfield, USA). The capillary used for acetone and hexane was 5
cm long and 1/8 inch
diameter, and DMS was 10 cm long and 1/8 inch diameter. They
were connected to ¼ inch
white Teflon PFTE tubing which connected the streams in the
entire system. The main
Q4
Q3
Q5
Q3
Q1
Q2
2
11
9
1
1
2
2
6
3 4 5
7
L1
1
P1
10
P6
P2
P3
P4
P5
MFC 1
SIFT-MS
CO2
8
A B
b
C
MFC 2
MFC 3
-
30
contaminated stream flow (Q3) was pre-humidified in humidifier
filled with water before getting
in the filter bed packed equally in volume by compost, woodchips
and silicon foam.
The stream (Q3) flows upward in the filter bed which had six
measuring ports (i.e. inlet, outlet and
four intermediate measuring ports). The leachate collection port
was at the bottom of the filter
bed. Valve on measuring port was manually switched to start a
small flow (Q4) that was
analyzed from the big stream (Q3) in the BF. This stream flow
(Q4) controlled by valves and
measured with rotameter was diluted with a stream flow (Q5) of
nitrogen before reaching
measuring instruments to avoid high concentration in the
SIFT-MS. Figure 16 represent the actual
experimental set up for the study
Figure 16: Actual setup of the biofilter (1) PR, (2) polluted
bottles, (3) humidifier column,
(4) biofilter, (5) SIFT-MS.
2.6.3 Characterization of the packing materials
The packing materials used for the experiments were compost
(port grand, Belgium), woodchips
and silicon foam (sponge cord®, Netherlands). They were mixed
and put in a filter bed at equal
volume fractions (1/3 v/v) (Figure 17). The physical chemical
properties conducted for the
packing materials were density, moisture content, water holding
capacity and porosity.
Figure 17: Packing materials used in biofiltration process. (1)
Compost, (2) woodchips and
(3) Silicon foam.
5
2 3
4
1
1
2
3
-
31
2.6.3.1 Bulk Density
The apparent density of the three packing materials was measured
by weighting packing
materials at ambient conditions into a known volume dimension.
The Equation (25) was used to
calculate the bulk density, where mP and VB are weight of the
packing material and volume of
used column respectively.
Bulk density =mPVB
(Eq.25)
2.6.3.2 Moisture content
The moisture content of the packing materials was calculated
based on Equation (26) with mP
and mDP the mass of the packing material at ambient conditions
and the mass of the dry
packing material after 72 h at 358 K in oven.
Moisture content =mP − mDP
mpx100 (Eq.26)
2.6.3.3 Water holding capacity
The water holding capacity of the packing materials was
calculated by applying Equation (27)
with mw and mWP the amount water poured on the dried packing
material and the mass of the
packing material 15 min after pouring water.
2.6.3.4 Porosity
The porosity of the mixed packing materials (1/1/1 volume ratio)
was calculated based on an
online method using SIFT-MS. It was calculated as the ratio of
net residence time (NRT) over
empty bed residence time (EBRT) (Equation 28). Compounds with
high Henry’s law constants
behave as inert compound into biofilter and no degradation and
absorption might happen to
such compounds (Volckaert, 2014). Dynamic experiment to
calculate the porosity in a first
attempt was done by injecting 10 μL of hexane liquid into the
biofilter. No peaks were found at
the outlet port for acetone, DMS and hexane. This can be cause
by the interaction of the
packing materials with the contaminants (Figure 18). The second
experiment, a 15 ml of
methane (± 35000 ppm) gas was injected and the peaks were
recorded by the SIFT-MS to both
Water holding capacity =mWP −