full papers proceeding The 9th International Conference on Environmental Engineering, Science and Management_Final

- 186 -

particles to obtain the resolution of 4 cm-1 between the Infrared (IR) range of 600 to 4000 cm-1. The collected
spectrum was compared to the reference spectra of OMNIC polymer database provided by Thermo Fisher.
Particles with higher than 70% similarity index were accepted as MPs and below 70% were assumed as non-
plastic [8].

RESULTS AND DISCUSSION
Abundance and distribution of microplastics
MPs were found in all stations of the study area with the mean concentration of 2.33  105 particles/km2 (2.33
particles/m3) which is lower than the concentration in Japan Sea (1.7106 particles/km2) but much higher than the
concentration in East China Sea (0.167 ± 0.138 particles/m3) [9][10]. Globally accepted standard procedures are
not yet set for sampling and identifying of MPs so the inconsistent methodologies are the main challenge for
comparing with other findings from different regions of the globe. Moreover, there is no universal unit for
presenting MPs concentration which is expressed as particles in area or volume of water. However, the
comparison provides the qualitative information on the severe threat of MPs to the organisms in the study area.
The standardization of methodology as well as harmonization of unit are required for comparing and reporting of
MPs concentration globally.

All collected MPs were classified according to the size ranges, shapes and colors. MPs concentration of 3.24  105
particles/km2 was observed from ST1 which located 0.6 km from the nearest land. The concentration of ST2 (1.7
km) and the last, ST3, (4.0 km) were 2.14  105 particles/km2, and 1.62  105 particles/km2 respectively. The
results showed that the highest concentration point was the nearest to the land and the lowest one was the farthest
from the land. According to the results, the abundance of MPs decreased with the distance far from the land and it
indicated that collected MPs were derived from land-based sources.

Figure 4 shows the collected MPs sorted into four different size ranges (335 to 515 microns, 516 to 990 microns,
991 to 2100 microns and 2101 to 5000 microns) to know more clear distribution form of MPs in the study area.
The number of collected smallest range (335 to 515 microns) MPs in ST1 was 172  103 particles, the second
range was 127  103 particles, the third range was 20  103 particles, and the last and largest range was 4  103
particles respectively. For ST2, the smallest range was 155  103 particles, the second range was 33  103
particles, the third range was 20  103 particles, and the last was 6  103 particles. For the last station (ST3), the
smallest range was 149  103 particles, the second range was 10  103 particles, the third range was 3  103
particles, and the last range was not found in this station. According to the results, the abundance of larger size
ranges significantly decreased by the distance far from land but the smallest size range remained nearly the same
quantity.

Figure 4 Abundance of microplastics in each station of the study area
9th International Conference on Environmental Engineering, Science and Management

The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 187 -

The results indicated that the average abundance size was the smallest range with 68% of the total collected MPs.
The second, third and last size ranges accounted for 24%, 6%, and 1% respectively. The concentration of MPs
with the smallest size range was significantly higher than the other sizes. The smallest size range of MPs are
similar size with zooplankton and the high threat to the other aquatic organisms because of high probability of
mistaken ingestion as their food [11]. The result indicated that the possibility of ingestion increased to the aquatic
organisms in the study area that may not only stress to their organ systems and affect to their bioavailability but
also threaten to the connected food chain.

Physical identification of microplastics
The results from physical identification shown that the shapes of MPs in ST1 were 215  103 fragments, 53  103
pellets, 38  103 films, and 18  103 fibers. For ST2, the shapes were 125  103 fragments, 59  103 pellets, 31 
103 films and no fibers were found. The shapes in the last station (ST3) were 109  103 fragments, 26  103
pellets, 14  103 films, and 13  103 fibers. The shape of MPs represents their origin (primary or secondary MPs).
Fragments and films were the product of photo-chemical degradation and mechanically break down of larger
pieces of plastic waste so they were secondary microplastics. Pellets were likely to be considered as primary
microplastics which mostly used as the industry feedstock for plastic material productions. Fibers were considered
as the secondary microplastics from the wastewater of washing machine and degradation of the fishing net.
According to the results, the predominant shape in the study area was the fragment with 64% of total collected
MPs which follow by pellet (20%), film (12%), and fiber (4%) respectively (Figure 5). This result indicated that
combination of fragment, film and fiber accounted 80% of total collected MPs were secondary MPs derived from
the fragmentation of larger plastics pieces and the rest 20% were primary plastic probably generated from the
industrial sector.

4%

12%

20% Fragment Pellet

64%

Fragment Pellet Film Fiber

Film Fiber
Figure 5 Shapes of microplastics found in the study area and proportions of the shapes

Physical identification showed that the number of MPs in colors followed as a decreasing order: white >
transparent > blue > red > brown > black. Among these, white and transparent were dominant colors with 84% of
total collected MPs. 10% of total MPs were blue and the rest were 2% of each black, red and brown colors
respectively (Table 2). White and transparent MPs were mostly in film and fragment shape and these colors were
widely used in the packaging industry that indicated such MPs were derived from the fragmentation of plastic
waste from the land as the secondary MPs. White and blue were dominant in the form of fiber, respectively which
could be derived from the degradation of fishing lines and nets. Predominant of the blue color fibers in the gut of
fish in Goiana Estuary, Brazil has been reported [12]. The rest colored MPs were also very harmful to the aquatic
organisms for the reason that the color of MPs attract the predators to increase misidentification as their food [13].

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 188 -

Table 2 Proportion of the observed colors of microplastics from the study area

Color Total percent of
each color (%)
White
Transparent 56
28
Blue 10
Red 2
Brown 2
Black 2
Total 100

Chemical identification of microplastics
The results from FTIR analysis showed that the proportions of chemical composition in ST1 were 68% of
polypropylene (PP), 17% of low-density polyethylene (LDPE), and 15% of polyethylene (PE) respectively.
For ST2, the proportions were 55% of PP, 15% of LDPE, 29% of PE, and 1% of polystyrene (PS)
respectively. The proportions of last station (ST3) were 59% of PP, 27% of PE, 7% of PS, and 7% of nylon
respectively. Among these compositions, the highest abundance of average proportion in the study area was
PP with 62% of total collected MPs which follow by 22% of PE, 12% of LDPE, 2% of PS and 2% of nylon
respectively (Figure 6). Polypropylene (PP) was significantly higher compared to other types in the study
area. It is most widely produced plastic in the world. Moreover, it is used in a wide variety of applications,
especially in the packaging and labelling industries [14]. The high proportion of PE and LDPE was also not
very surprising due to their wide application in daily life and industrial sectors. The low specific density and
high buoyant properties of PS allows to float and widespread distribution in the aquatic environment and it is
widely use in the food packaging container and protective material for packaging. Nylon is commonly used
in the textile and fishing net. The results indicated that MPs collected from the study area were mostly
derived from the land base mismanaged plastic waste.

Figure 6 Proportion of chemical composition of collected microplastics from study area

CONCLUSIONS
MPs were found in all sample collection stations of the Chao Phraya River Estuary with the mean
concentration of 2.3  105 particles/km2. The abundance of larger size ranges of MPs significantly decreased
with distance far from land but the smallest size range remained nearly same quantity. The dominant shapes
of MPs were film and fragment with white and transparent colors indicated that these were derived from the
fragmentation of mismanaged plastic waste from the land as secondary MPs. Polypropylene (PP) was the

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 189 -

most abundance in the study area and other abundance plastics (PE, LDPE, PS, and nylon) are mostly
derived from the degradation of packaging, labelling, daily use materials and industrial application as
secondary MPs. Chao Phraya River is the main river system of Thailand and flows to the Gulf of Thailand so
its estuary is a large habitat for many aquatic organisms. Therefore, the government and environmental
organizations need to enforce to reduce the plastics usage and to improve the solid waste management to
prevent the plastic debris entering into the estuary.

ACKNOWLEDGEMENT
The author would like to express sincere thanks to the Head of Regional Harbor Office (Samut Prakan
Branch), officers and staff for supporting the required documents and the marine vessel. This study was
supported by “Capacity Building for Initiative for Myanmar” jointly sponsored by Mahidol University
and Norwegian Government. This research was partially supported by On-site Laboratory Initiative of
Graduate School of Global Environmental Studies, Kyoto University.

REFERENCES
[1] Moore, C. J. (2008). Synthetic polymers in the marine environment: A rapidly increasing, long-term

threat. Environmental Research, 108(2), 131-139. doi:https://doi.org/10.1016/j.envres.2008.07.025.
[2] Hammer, J., Kraak, M., & Parsons, J. (2012). Plastics in the Marine Environment: The Dark Side of a

Modern Gift. Reviews of environmental contamination and toxicology, 220, 1-44. doi:10.1007/978-1-
4614-3414-6-1.
[3] Sadri, S. S., & Thompson, R. C. (2014). On the quantity and composition of floating plastic debris
entering and leaving the Tamar Estuary, Southwest England. Marine Pollution Bulletin, 81(1), 55-60.
doi:https://doi.org/10.1016/j.marpolbul.2014.02.020.
[4] Wang, Y., Zou, X., Peng, C., Qiao, S., Wang, T., Yu, W., . . . Kornkanitnan, N. (2020). Occurrence
and distribution of microplastics in surface sediments from the Gulf of Thailand. Marine Pollution
Bulletin, 152, 110916. doi:https://doi.org/10.1016/j.marpolbul.2020.110916.
[5] Thushari, G. G. N., Senevirathna, J. D. M., Yakupitiyage, A., & Chavanich, S. (2017). Effects of
microplastics on sessile invertebrates in the eastern coast of Thailand: An approach to coastal zone
conservation. Marine Pollution Bulletin, 124(1), 349-355. doi:https://doi.org/10.1016/j.marpolbul.
2017.06.010.
[6] McLaren, R., Kanjanapa, K., Navasumrit, P., Gooneratne, R., & Ruchirawat, M. (2004). Cadmium in
the Water and Sediments of the Chao Phraya River and Associated Waterways, Bangkok, Thailand.
Water Air and Soil Pollution, 154, 385-398. doi:10.1023/B: WATE. 0000022990. 80129.
[7] Port Authority of Thailand (P.A.T.) (1993). Map of The Chao Phraya River from Pom Phrachul to
Memorial Bridge, surveyed by the Marine Survey Division, Marine Department, P.A.T. Bangkok.
[8] Frias, J. P. G. L., Gago, J., Otero, V., & Sobral, P. (2016). Microplastics in coastal sediments from
Southern Portuguese shelf waters. Marine Environmental Research, 114, 24-30.
doi:https://doi.org/10.1016/j.marenvres.2015.12.006.
[9] Isobe, A., Uchida, K., Tokai, T., & Iwasaki, S. (2015). East Asian seas: A hot spot of pelagic
microplastics. Marine Pollution Bulletin, 101. doi:10.1016/j.marpolbul.2015.10.042
[10] Zhao, S., Zhu, L., Wang, T., & Li, D. (2014). Suspended microplastics in the surface water of the
Yangtze Estuary System, China: First observations on occurrence, distribution. Marine Pollution
Bulletin, 86(1), 562-568. doi:https://doi.org/10.1016/j.marpolbul.2014.06.032.
[11] Auta, H., Emenike, C., & Fauziah, S. H. (2017). Distribution and importance of microplastics in the
marine environment: A review of the sources, fate, effects, and potential solutions. Environment
international, 102. doi:10.1016/j.envint.2017.02.013.
[12] Possatto, F. E., Barletta, M., Costa, M. F., Ivar do Sul, J. A., & Dantas, D. V. (2011). Plastic debris
ingestion by marine catfish: An unexpected fisheries impact. Marine Pollution Bulletin, 62(5), 1098-
1102. doi:https://doi.org/10.1016/j.marpolbul.2011.01.036.
[13] Aliabad, M. K., Nassiri, M., & Kor, K. (2019). Microplastics in the surface seawaters of Chabahar
Bay, Gulf of Oman (Makran Coasts). Marine Pollution Bulletin, 143, 125-133.
doi:https://doi.org/10.1016/j.marpolbul.2019.04.037.
[14] Heinrich Böll Foundation. (2019). Plastics Atlas 2019 (First). Berlin.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 190- I 034

Electricity generation using wind stack together with carbon dioxide capture

Rawat Yimvilai1, Suwanna Kitpati Boontanon2 and Narin Boontanon3

1Graduate student, Faculty of Graduate Studies, Mahidol University; 2Associate Professor,
Department of Civil and Environmental Engineering, Faculty of Engineering, Mahidol University;

3Lecturer, Faculty of Environment and Resource Studies, Mahidol University,
Nakhon Phathom 73170, Thailand

*Phone : 024415000 ext. 2211; E-mail : [email protected]

ABSTRACT
Carbon dioxide emission is increasing from the expansion of the economy and human demand. Huge amount
of CO2 emission from combustion using fossil fuels for energy production. After burning, exhaust gas will
be released into the stack and floating up to the atmosphere. Therefore, using air Plumes from the

smokestacks may benefit because hot air has unique properties such as high temperature, low density, and
high velocity. Those properties are changed to buoyant force for electric generation coupling with CO2
capture. The result showed that the shape of the bubble collector as half-spherical cylinder shape was the
most appropriate for electricity generation due to less resistance during movement in the water. The motion

design was a vertical ellipse which provides the highest efficiency because it can maintain the mechanical
energy (buoyancy) for a long time. The average electricity generation is 254 watts using an air volume of 3.5
cubic meters per minute from 500Watt low RPM generator (BPE-PMG500). Furthermore, during electricity

generation, CO2 gas in the air was react with calcium hydroxide solution and generated calcium carbonate
which easily to remove from the system as solid phase. Carbon dioxide capture efficiency is highest after 10
minutes of the reaction which is about 65 mgCO2 per hour under test conditions. This research is a new
innovative idea for electricity generation from waste wind stack together with CO2 reduction. These idea is
an eco-friendly and will be a part of the global warming problem reduction.

Keywords : Buoyant Force, electricity generation, carbon dioxide capture, waste to worth, wind stack

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 191-

INTRODUCTION
Global warming is currently worst and tend to more serious due to economic growth and increasing of
human demand. Industrial sector has emitted the largest greenhouse gases (GHGs), especially carbon dioxide
(CO2). In 2017, approximately 42% of the global GHGs emission is CO2 emission from fossil fuels
combustion. [1] Burning fossil fuels in boiler and gasifier result to steam and carbon dioxide which were
released through the stack into the atmosphere. However, the wind stack, hot air, has unique properties such
as high temperature, low density and high velocity. Thus, it might have the potential to use as the source of
energy as the concept of “waste to worth”. Furthermore, CO2 in wind stack might be reduced during its
utilization as well. This research aims to invent a prototype of electricity generation using wind stack
coupling with chemical carbon dioxide capture for waste utilization and minimization.
Research objectives design and build a prototype machine electricity from wind stack together with capture
carbon dioxide by calcium hydroxide solution is new alternative energy from the concept waste to worth
together with reducing environmental impact

METHODOLOGY
This research is an invention. The operating location is the Faculty of Environment and Resource Studies,
Mahidol University. Study method to divide into 4 steps. The electricity generation and CO2 trapping
prototype were invented following study:

Step 1: The selection of three different shapes of air caption as a cylindrical cross-section design the shape of
the air bubble blade to receive 3 shapes of air bubbles used in the test 1.1) Cylindrical, cross-section 1.2)
Horizontal cylinder and 1.3) Half spherical cylinder. The results were obtained from the time of upward
movement in the water under same distance, material, weight and air volume. The shortest time will be
chosen as selected shape and use for next step.

1.1) Cylindrical, cross-section 1.2) Horizontal cylinder 1.3) Half spherical cylinder

Figure 1 The shape design and when there is air entering the air bubble blade .

Step 2: The design of circular motion for electric generation using air bubble trapping. The device is
allowing the air trapping in the appropriated shape from step 1 and make the buoyancy force driving the
chain which connecting to wheel and electric generator. The number of air bubble trapping blade were
investigated by feeding 3.5 cubic meters per minute. of air volume into the 200-liters plastic tank. Then the
revolution rotation per min of the blades was observed.
Step 3: The electricity generation testing using 600 watt low RPM generator (BPE-PMG500). Electricity
voltage and current were recorded every 10 minutes for an hour by using. Digital multimeter for the
measurement of electric current (A) voltage (V) produced and calculated electricity power using the formula

P = V*I (1)

where P = The electric power (Watts), W
V = Electric potential difference (Volt), V
I = Electric current (Amp), A

9thInternational Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May27-29, 2020

- 192-

Step 4 : The CO2 caption performance was estimated using previous studied and CO2 content in the wind
stack. The estimation will be focus on CO2 reduction under the chemical reaction of CO2 in the wind stack
and calcium hydroxide solution under experiment conditions. The reaction is shown in following equation :

Ca(OH)2 + CO2  CaCO3 + H2O (2)

This step was performed by simulating of the electricity generation parameters (such as flowrate) but
downscale (from 200 to 45 liters tank) the size to save chemicals and more easier to handle.

That system was using a 45-liter tank and standard CO2 gas concentration 10,000 ppm with a circular
flowrate of 1,000 l/min. Calcium hydroxide solution 5,000 ppm was used as a trapping solution in the tank.
The sample was collecting 10 minutes interval to measurements pH, turbidity and suspended solids (SS) by
weighing, drying after filtering the water. Those results will be used to estimate the CO2 caption efficiency.

RESULTS AND DISCUSSIONS
Step 1: Air bubble blade design Is constructed from the same material, weight but making different shapes
The results of the experiment are as in table 1

Table 1 The results of the experiment with the speed of the air bubble blade

The Blade shape Time of up to the surface water, (seconds) average
1.1) Half-spherical cylinder (seconds)
Sample 1 Sample 2 Sample 3
0.71
0.73 0.68 0.72

1.2) Cross-section cylindrical

0.92 0.88 0.91 0.90

1.3) Horizontal cylinder

1.18 0.98 1.10 1.07

Step 1 results follows 1.1) Half-spherical cylinder average speed 0.71 seconds 1.2) Cross-section cylindrical
average speed 0.90 seconds 1.3) Horizontal cylinder average speed 1.07 seconds respective.

Time (Seconds)

1.1) Half-spherical 1.2) Cross-section 1.3) Horizontal

cylinder cylindrical cylinder

Figure 2 Buoyancy speed of the air bubble blade.

Step 1: According to the results, the half spherical cylinder shape was found as the fastest upwelling speed as
approximately 0.71 seconds (Fig.1). Therefore, this shape was selected for the further step. The time
differentiation of each shape is depending on the resistance drag force [2] at the front which slow down the
buoyancy motion.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 193-

Step 2: According to the results, the motion design was a vertical ellipse which provides the highest efficiency
because it can maintain the mechanical energy (buoyancy) for a long time along with the vertical distance. The
vertical ellipse shape air trapping consists of a half-spherical cylinder air trapping blade made by aluminum with 9
cm. high, 6.5 cm. diameter and 72 blades due to the limitation of water tank capacity (200-liters). The design was
shown in figure 3.

Electric power

Generator Generator

Vertical-ellipse shape Buoyancy force

Air pass
liquid
Reaction
capture a
CO2

Figure 3 Experimental model and process electricity generator using wind stack together with carbon
dioxide capture.

The principle of electricity generation in this research is turning the buoyancy force to kinetic energy using
the properties of a low density of hot air in the stack. Those buoyancy force is an upward force exerted by a
fluid that opposes the weight of immersed object [5]. In this experiment we use an exhaust gas to stimulated
the air bubble at the bottom of the tank. Those air bubble was collected by an air blade to generated the
buoyancy force. Those obtain energy was transfer to the rotating belt connect to the generator. According to
the exhaust gas are containing with high concentration of CO2. During air bubble was generated in the
system, those portion of CO2 gas could react with any chemical in the solution. I such conditions, we use a
calcium hydroxide (Ca(OH)2) solution to react with CO2 and produced a particle of calcium carbonate
(CaCO3) as shown in equation 2.
The vertical-ellipse shape system was used because it can be keeping mechanical energy from buoyancy forces of
the air bubbles for a long time. As it corresponds to the buoyancy of the air bubble, which floats up to the surface
water in a straight line. We will notice that vertical ellipse can hold force for a long time affect performance power
generation as shown in figure. 4 and similar to the research Hossein Samadi-Boroujeni al [5].

(1)Vertical-ellipse shape Air babble F4
= F1 + F2 + F3 + F4 +….Fx F3
F2
F1

Figure 4 Compare energy storage between Vertical-ellipse shape and circle shape.

9thInternational Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May27-29, 2020

- 194-

Step 3: The efficiency of electric generation using buoyancy force of wind stack was The average electricity 254
watts at an air volume of 3.5 cubic meters per minute. If calculated in hours was about 1.2 watts/cubic meter per
hours and the average revolution is 304 rpm as shown in figure 5.

700

Power generation Watt, RPM 600 306 315 302 311 301 304
500

400
RPM

300 Watt

254 261 250 258 250 252

200

100

Time (min)

0
10 20 30 40 50 60

Figure 5 The electric power generation and revolution of the system.

Under the experiment design, the electric power obtained from the system is just about 43% of the generator
performance. It is due to the revolution of the system which could not reach to 600 rpm under testing
condition. There is a limitation on the height of experimental model tanks or the limitation of air flow that

affect to the buoyancy force. However, increasing of the height or increase air bubble blade frequency may
resulting of higher power obtained from that generator. The extension size of the air bubble blade also helps

to increase the power to produce electricity or reducing rotation speed to get the full efficiency of the

generator.

Step 4: Trap performance test Carbon dioxide

Line Output Air tank 45 L CO2 10,000 ppm

Valve Feed Ca(OH)2
Valve Water sample

Line input Samping 5 minute of 1 hour

• pH pen Mater

• Turbidity Mater

Air Pump • Suspended solids
1,000 L/Min Calcium hydroxide solution (45 L)

Ratio Ca(OH)2: CO2
5,000 ppm : 10,000 ppm

Figure 6 The diagram of experiment for carbon dioxide caption.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 195-

Figure 7 The experiment system for carbon dioxide caption performance.

Step 4: According from the results for 10 minutes intervial, From the test found that every 10 minutes, the
suspended solids values were 65, 44, 41, 40, 38 and 38 mg as CO2/L, respectively (figure.8), turbidity 58, 49,
48, 46, 46 and 46 NTU, respectively (Fig. 9)

It was found that the precipitation was highest after 10 minutes reaction as weight of 65 mg CO2 and
turbidity 58 NTU. Then decreased to 50 minutes at 38 mg CO2 and turbidity at 46 NTU. The decreasing may
be due to the agitation of the impeller, resulting in the dissolution of some previous leaf trapping or stirring
may affect sediment measurement and trap efficiency of 8%. The reason is due to some air that stay in the
blade, not exposed to chemicals therefore resulting in decreased efficiency. And the air has a very fast flow
rate which can affect the capture, in accordance with the research study. The electricity decreased because of
collecting water samples causing the water level to drop affect the spin cycle in electricity production from
reduce buoyancy force.

CO2 Suspended Solids mg/L 80 44 41 40 38 38
65

60

40

20

0
10 20 30 40 50 60 Time (min)

Figure 8 Calcium carbonate produced as suspended solids.

9thInternational Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May27-29, 2020

- 196-

Turbidity : NTU 80

60 58 49 48 46 46 46
40

20

0

10 20 30 40 50 60 Time (min)

Figure 9 Turbidity of the system during CO2 caption.

* NTU (Nephelometric Turbidity Units)

7 6.9

6.8 6.7 6.6 6.6 6.6

6.6 6.5

pH 6.4

6.2 Time (min)
10 20 30 40 50 60

Figure 10 pH value from trapping CO2.

CONCLUSION

This research proposes an alternative electricity generation using an exhaust gas in wind stack. which are
considered sources of air pollution. Electricity was generated from the 72 half-round air bubble blades that
are arranged in a vertical ellipse, forcing 600 revolutions per minute to drive the generator. In our experiment
model, it can generate about 254 watts using an air volume of 3.5 cubic meters per minute. Simultaneously,

CO2 gas in the air will be eliminate up to 65 mgCO2 in per hour using calcium hydroxide solution during

electric generation.This prototype experiment demonstrated has an ability and possibility of turning waste
into value in terms of energy conservation along with waste reduction (CO2). We believe that our ideas will

be a part in creating a better environment in the future.

ACKNOWLEDGEMENT
This research is supported by National Research and Innovation Information System (NRIS) number

1141821, Graduate Studies of Mahidol University Alumni Association.

REFERENCES
[1] International Energy Agency (IEA). CO2 emissions from fuel combustion highlights (2017 edition),
France: International Energy Agency, 2017.
[2] Pao-ChiChen al. Optimum conditions for the capture of carbon dioxide with a bubble-column scrubber.
International Journal of Greenhouse Gas Control, 2015.
[3] Tuangporn Paenpoom. Chapter 4 Suspensions (Suspensions: SS). Water sample test guide Laboratory of
Regional Environment Office 6 (Nonthaburi Province), 2014.
[4] Kai Cui1 al. Aerodynamic Performance Comparison of Head Shapes for High-Speed Train at 500KPH.
Advances in Civil, Environmental, and Materials Research (ACEM’ 12). Seoul Korea. August 26-30, 2012.
[5] Hossein Samadi-Boroujeni al. Application of buoyancy-power generator for compressed air energy
storage using a fluid–air displacement system. Journal of Energy Storage. Volume 26 December. 100926,
2019.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 197 - I 035

Daytime and Nighttime Variation of Chemical Characteristics of
PM1.0 and Gaseous Pollutants in the Suburbs of Taiwan in Autumn

Sasipim Somphan1* Khajornsak Sopajaree2 and Ying I. Tsai3

1*Graduate student, Graduate Master’s Degree Program in Environmental Engineering,
Faculty of Engineering, Chiang Mai University, Chiang Mai 50000, Thailand;

2Associate Professor, Environmental Engineering, Faculty of Engineering, Chiang Mai University,
Chiang Mai 50000, Thailand

3Professor, Department of Environmental Engineering and Science,
Chia Nan University of Pharmacy and Science, Taiwan

*Phone : 0848107762 E-mail : [email protected]

ABSTRACT
The characteristics of particulate matter and gaseous pollutants differ depending on the human activities and
natural emissions that occurred during that time. In this study, chemcomb speciation sampling cartridges
used to have collected PM1.0 and gaseous pollutants from suburbs of Taiwan during the autumn 2018. The
chemical characteristics of PM1.0 including organic matter (carboxylates, saccharides, anhydrosugars and
water-soluble ions) and gaseous pollutants were investigated. HCl, HNO3, SO2 and gaseous oxalic acid were
higher in the daytime than in the nighttime due to photochemical reaction and more activities in daytime.
Lactate was the most abundant carboxylate species and higher in the daytime than nighttime with an average
concentration of 0.769±0.355 µg/m3. Similarity, SO42- was the most dominant ionic species and higher in the
daytime than the nighttime with an average concentration of 4.75 ± 1.32 µg/m3 due to industrial activities
and photochemical reaction. Levoglucosan was higher in nighttime than in the day time with an average
concentration of 0.049±0.030 µg/m3 indicating that biomass burning activities especially more frequently
existed at night.

Keywords: Gaseous pollutants; PM1 aerosol; Chemical characteristics; biomass burning

INTRODUCTION
Particulate matter (PM) is a common indicator for air pollution. PM more effects on people than any other
pollutants. The major components of PM are sulfate, nitrates, ammonia, sodium chloride, black carbon,
mineral dust and water. Natural and human activities are the major sources of PM such as forest fires,
incomplete combustion of biomass, fossil fuels, industrial process and agriculture (Huang B, et al. 2013; Tsai
YI, et al. 2013). In addition, PM especially fine particles such as PM2.5 and PM1.0 have attracted more and
more attention due to its important roles in air pollution, health effects and global climate change. (Zhou X,
et al. 2016). Furthermore, the effects of PM, gaseous pollutants also such as CO, SO2, NO2, and O3.
Moreover, effects of PM tend to increase human health especially respiratory diseases and leading to
hematological problems and cancer (Li Q, et al. 2017). According to previous studied, Srimuruganandam
and Nagendra studied about PM emissions from the heterogeneous traffic near an urban roadway in India
and showed that the PM1.0 concentration higher at 8 a.m.-11 a.m. and 5 p.m.-9 p.m. due to traffic hours
(Srimuruganandam B and Shiva Nagendra SM, et al. 2010). The same as, B.C. Beh studied about
distribution of PM in Penang Island, Malaysia and showed that all PM’s distribution (PM10, PM2.5 and PM1.0)
were the highest at daytime during 11 a.m. to 1 p.m. due to lunchtime and more transportation on the road
(Beh BC, et al. 2013) Moreover, Buczynska founded that daily temperature, wind speed and relative
humidity are associated with the concentrations of PM2.5 and PM1.0, which the lowest temperature, the lowest
wind speed and the highest relative humidity are elevated PM concentration. (Buczynska AJ, et al. 2014).
Zhou studied characteristics of PM1.0 over Shanghai during 2015 and 2016 and founded that the PM1.0
concentrations was higher during early morning and near noon. The concentrations of PM1.0 are increase
with low relative humidity, low temperature and low wind speed. SO2, NO2, and NO were high correlations
with PM1.0 (Zhou G, et al. 2018). Dhananjay studied water soluble ions in PM2.5 and PM1.0 in Durg, India and
founded that SO42- was the highest concentration in PM1.0 followed by NO3-, Cl- and NH4+ due to motor
vehicles, commercial activities and services in Durg city (Dhananjay K, et al. 2011). The aim of this study is
to characterize the chemical properties of PM1.0 during daytime and nighttime in suburban including ions,

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 198 -

organic acids, anhydrosugars, saccharides and gaseous pollutants. To determine the relationships between
chemical species and gaseous pollutants on particulate matters during daytime and nighttime in suburban.

METHODOLOGY
In this study, chemical species characteristics of PM1.0 and gaseous pollutants were investigated. The
relationships between chemical species of PM and gaseous pollutants during daytime and nighttime in the
autumn (September 2018 - October 2018) were also statistically explored. All samples were collected in the
suburban of Tainan City, southern Taiwan, during the daytime (8 a.m. to 5 p.m.) and nighttime. (5 p.m. to 8
a.m.). The samples of PM1.0 and gaseous pollutant were collected by Chemcomb 3500 Speciation Collection
Cartridge. The samples were analyzed by ion chromatography.

RESULTS AND DISCUSSIONS
In the Figure 1, the concentrations of PM1.0 were divided into two phases, namely the gas phase and the particle
phase. In the gas phase accounted for 64.3% of total PM1.0 and 35.7% for the particulate phase. However, particle
phase include carboxylates (3.8% of total particle phase), water soluble ions (31.6% of total particle phase) and
anhydrosugars. (0.3% of total particle phase)

Particle phase
Gas phase

Figure 1 Percentage of the particulate phase and gas phase

The average concentrations of gaseous pollutants are shown in Table 1. NH3 was the most abundant gaseous
pollutants with an average concentration of 7.96 ± 2.62 µg/m3 followed by HCl and SO2 with average
concentrations of 6.72 ± 5.30 µg/m3 and 5.05 ± 6.56 µg/m3, respectively. However, the average concentration of
gaseous pollutants was different during daytime and nighttime. HCl and SO2 were higher in the daytime with
average concentrations of 8.70 ± 6.72 µg/m3 and 6.87 ± 9.03 µg/m3, respectively, due to combustion in industries
and transportation in the daytime. Compared to the daytime HNO2 of 0.288 ± 0.260 µg/m3, HNO2 were higher in
the nighttime with an average concentration of 0.879 ± 0.339 µg/m3. Meanwhile, the daytime HNO3 of 3.86 ±
0.971 µg/m3 was higher than the nighttime with an average concentration of 1.89 ± 0.656 µg/m3 due to the
photochemical reaction that HNO2 transformed to HNO3 in the daytime, the detail was shown in Figure 2. On the
other hand, compared to the daytime NH3 of 7.12 ± 2.35 µg/m3, the nighttime NH3 were higher with an average
concentration of 8.79 ± 2.70 µg/m3 (Fig. 2). PM1.0 NH4+ was 2.14 ± 0.639 µg/m3 in the daytime and 1.80 ± 0.747
µg/m3 in the nighttime (Fig. 3), indicating that NH3 was converted to PM1.0 ammonium by photochemical
reaction during the daytime. Oxalate was abundant dicarboxylate species in PM1.0 and a little higher in the
nighttime with an average concentration 0.229 ± 0.078 µg/m3, compared to 0.218 ± 0.059 µg/m3 in the daytime
(Fig. 4). The average concentration of gaseous oxalic acid during the day is 0.420 ±0.159 µg/m3, while the
average concentration of gaseous oxalic acid at night is 0.255 ± 0.101 µg/m3 (Fig. 2). The gaseous oxalic acid in
the daytime was always higher than that in the nighttime, apparently indicating there is more oxalic acid produced
during the day by biomass burning and motor vehicles.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 199 -

Table 1 Average concentration of gaseous pollutants

Gaseous Pollutants Average concentration (µg/m3)

NH3 7.96 ± 2.62
HCl 6.72 ± 5.30

HNO2 0.584 ± 0.422

HNO3 2.88 ± 1.29

SO2 3.70 ± 1.60
0.337 ± 0.155
Oxalic acid 0.404 ± 0.180
PO43-

n=24

Figure 2 Average concentrations of gaseous pollutants in PM1.0
in daytime and nighttime

Figure 3 Average concentrations of water soluble ions in PM1.0
in daytime and nighttime

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 200 -

Figure 4 Average concentrations of carboxylates in PM1.0
in daytime and nighttime

Carboxylates bounded on particulate matter. In case of carboxylates on PM1.0, lactate was the most abundant with
an average concentration of 0.620 ± 0.322 µg/m3 followed by acetate and oxalate with average concentration of
0.238 ± 0.085 µg/m3 and 0.224 ± 0.068 µg/m3, respectively (Table 2). Whereas, the less amount species are

Maleate and phthalate so on. In average daytime and nighttime the results showed that most species in the

daytime are larger than in the nighttime especially lactate and acetate with average concentration 0.769 ± 0.355
µg/m3 and 0.283 ± 0.080 µg/m3 (Fig. 4), respectively. Carboxylates sources emitted directly into atmosphere via

human activities or biogenic activities and as secondary by products of photochemical oxidation of organic
precursors (Thepnuan D, et al. 2019).

Table 2 Average concentration of carboxylates

Carboxylates PM1.0
Average concentration (µg/m3)
Lactate
Acetate 0.620±0.322
Formate 0.238±0.085
Pyluvate 0.065±0.031
Succinate 0.057±0.021
Malate 0.065±0.050
Malonate 0.034±0.021
Maleate 0.043±0.015
Fumarate 0.010±0.004
Oxalate 0.037±0.024
Phthalate 0.224±0.068
0.019±0.012

n=24

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 201 -

The chemical analysis water soluble ions, the water soluble ion that are defined as anion and cation. SO42- was the
most dominant species and the highest species in PM1.0 with an average concentration of 4.55 ± 1.47 µg/m3 and
accounted for 34.2% of total particle phase followed by NO3- (17.1% of total particle phase) and NH4+ (14.8% of
total particle phase). In contrast, F- and NO2- were less amounts species (Table 3). SO42- in PM1.0 that higher in the
daytime than in the nighttime with an average concentration of 4.75 ± 1.32 µg/m3 due to photochemical and
combustion such as engine exhaust and industrial activities in the daytime (Fig. 3). Furthermore, non-sea salt
sulfate (nss-sulfate) accounted for 95.6% of total SO42- and higher in the daytime with an average concentration of
4.52±1.32 µg/m3 due to industrial process and human activities such as transportation and residential heating
(Fig.5). Similarly, Cheng studied about ion characteristic of PM1.0 in Wuda and Tian Hong at Wuhan, China
and reported that SO42- , NO3-, and NH4+ were the three major ion species in both site during haze day and
normal day according to SO42-, NO3-,and NH4+ as secondary pollution, which results from the transformation
of their precursors SO2, NO2, and NH3, respectively (Cheng H, et al. 2014).

Table 3 Average concentration of water soluble ions

Water soluble PM1.0
ions Average concentration

F- (µg/m3)
Cl- Anions
NO2-
NO3- 0.057±0.021
SO42- 0.92±0.35
PO43-
0.114±0.056
Na+
NH4+ 2.28±0.822

K+ 4.55±1.47
Mg2+
Ca2+ 0.119±0.080
n=24 Cations

0.803±0245
1.96±0.700

0.270±0.154
0.125±0.082
0.564±0.270

Figure 5 Average concentrations of sea salt sulfate and non-sea salt sulfate
in PM1.0 in daytime and nighttime

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 202 -

For the result of anhydrosugar founded that levoglucosan has large amount concentration on PM1.0 with an
average concentration of 0.037 ± 0.025 µg/m3 followed by glucose and trehalose with average concentration of
0.015 ± 0.007 µg/m3 and 0.014 ± 0.013 µg/m3, respectively. Moreover, mannosan and mannose have less amount
species in anhydrosugars (Table 4). In addition, the PM1.0 levoglucosan concentrations were 0.024 ± 0.11 µg/m3
and 0.049 ± 0.030 µg/m3 at day and night (Fig. 6), respectively, indicating that biomass burning contributed to the
atmosphere, especially at night with higher biomass burning. Additionally, Křůmal K studied about organic

compound in PM1.0 in Czech and showed that levoglucosan was the most dominant saccharide species with
an average concentration of 423 ng/m3 followed by mannosan (69.4 ng/m3) and galactosan (20.1 ng/m3)

indicates that more biomass burning such as wood especially softwood that used for residential heating
around the sampling site (Křůmal K, et al. 2017).

Table 4 Average concentration of anhydrosugars

Anhydrosugars PM1.0
Average concentration (µg/m3)
Xylitol
Levoglucosan 0.012±0.009 (n=11)
0.037±0.025 (n=24)
Arabitol 0.013±0.002 (n=10)
Sorbitol 0.004±0.004 (n=24)
Mannosan 0.0021±0.0009 (n=7)
Trehalose 0.014±0.013 (n=24)
Mannitol 0.007±0.004 (n=10)
Mannose 0.0025±0.0009 (n=10)
Glucose 0.015±0.007 (n=24)

Figure 6 Average concentrations of anhydrosugars in PM1.0
in daytime and nighttime

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 203 -

CONCLUSION
This study outlines the chemical characterization and gaseous pollutants with PM1.0 that are generated
during daytime and nighttime. HCl, HNO3, SO2 and gaseous oxalic acid were higher in daytime than in the
night time due to photochemical reaction and more activities in daytime. Whereas, NH3 and HNO2 were
higher in nighttime than in the day time due to photochemical reaction that HNO2 transformed to HNO3 in
the day time and NH3 converted to NH4 during daytime. Among the organic matters, Lactate of carboxylates,
SO42- of ions and levoglucosan of anhydrosugars were the highest concentrations on bound PM1.0. Lactate
and SO42- were higher during daytime than in the night time due to photochemical reaction, burning biomass
and industrial activities. However, levoglucosan was higher in nighttime than in the day time due to more
biomass burning in nighttime than in daytime.

REFERENCE
[1] Huang B, Liu M, Ren Z, Bi X, Zhang G, Sheng G, et al. Chemical composition, diurnal variation and

sources of PM2.5 at two industrial sites of South China. Atmospheric Pollution Research.
2013;4(3):298-305.
[2] Tsai YI, Sopajaree K, Chotruksa A, Wu H-C, Kuo S-C. Source indicators of biomass burning
associated with inorganic salts and carboxylates in dry season ambient aerosol in Chiang Mai Basin,
Thailand. Atmospheric Environment. 2013;78:93-104.
[3] Zhou X, Cao Z, Ma Y, Wang L, Wu R, Wang W. Concentrations, correlations and chemical species of
PM2.5/PM10 based on published data in China: Potential implications for the revised particulate
standard. Chemosphere. 2016;144:518-26.
[4] Li Q, Wang Y, Luo C, Li J, Zhang G. Characteristics and potential sources of polychlorinated
biphenyl pollution in a suburban area of Guangzhou, southern China. Atmospheric Environment.
2017;156:70-6.
[5] Srimuruganandam B, Shiva Nagendra SM. Analysis and interpretation of particulate matter – PM10,
PM2.5 and PM1 emissions from the heterogeneous traffic near an urban roadway. Atmospheric
Pollution Research. 2010;1(3):184-94.
[6] Beh BC, Tan F, Tan CH, Syahreza S, Mat Jafri MZ, Lim HS. PM10, PM2.5 and PM1 distribution in
Penang Island, Malaysia. 2013. p. 146-50.
[7] Buczynska AJ, Krata A, Van Grieken R, Brown A, Polezer G, De Wael K, et al. Composition of
PM2.5 and PM1 on high and low pollution event days and its relation to indoor air quality in a home
for the elderly. Sci Total Environ. 2014;490:134-43.
[8] Zhou G, Xu J, Gao W, Gu Y, Mao Z, Cui L. Characteristics of PM 1 over Shanghai, relationships with
precursors and meteorological variables and impacts on visibility. Atmospheric Environment.
2018;184:224-32.
[9] Dhananjay K. Deshmukh1, Manas K. Deb1, Ying I. Tsai, Stelyus L. Mkoma. Water Soluble Ions in
PM2.5 and PM1 Aerosols in Durg City, Chhattisgarh, India. Aerosol and Air Quality Research,
2011;11: 696–708.
[10] Thepnuan D, Chantara S, Lee CT, Lin NH, Tsai YI. Molecular markers for biomass burning
associated with the characterization of PM2.5 and component sources during dry season haze episodes
in Upper South East Asia. Sci Total Environ. 2019;658:708-22
[11] Cheng H, Gong W, Wang Z, Zhang F, Wang X, Lv X, et al. Ionic composition of submicron particles
(PM1.0) during the long-lasting haze period in January 2013 in Wuhan, central China. Journal of
Environmental Sciences. 2014;26(4):810-7.
[12] Křůmal K, Mikuška P, Večeřa Z. Characterization of organic compounds in winter PM1 aerosols in a
small industrial town. Atmospheric Pollution Research. 2017;8(5):930-9.

Electronic Database
[1] Beh, B. C., Tan, F., Tan, C. H., Syahreza, S., Mat Jafri, M. Z., & Lim, H. S. 2013. PM10, PM2.5 and

PM1 distribution in Penang Island, Malaysia. From www.aip.scitation.org/doi/abs/10.1063/1.4803585,
29 march 2020.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 204 - I 036

A Study on Influence of Meteorological Factors on PM2.5 over
Bangkok, Thailand: Case study of Bang Na Station

Angkhana Ketjalan1* Usa Humphries2 and Warawut Suadee3

1*Graduate Student of Environmental Technology, The Joint Graduate School of Energy and Environment,
King Mongkut’s University of Technology Thonburi, Bangkok.

2Associate Professor Dr., Department of Mathematics, Faculty of Science, King Mongkut’s University of
Technology Thonburi, Bangkok.

3Associate Professor Dr., Faculty of Public Health, Thammasat University, Pathum Thani.
*Phone : 090-980-0908, E-mail : [email protected].

ABSTRACT
PM2.5 has 5 major components such as organic matter (OM), elemental carbon (EC), crustal material,
ammonium sulfate((NH4)2SO4) and ammonium nitrate (NH3NO3). PM2.5 can be primary pollutant (releasing
from source) and secondary pollutants (resulting from precursor gases which are SO2, NOx, VOCs, and NH3)
[1]. Particulate matters in the air can affect to human health particularly in height concentration. PM10 or
particulate matter with diameter 10 micron or less than can pass into and stay in lung but PM2.5 can pass
through the lung barrier and go inside blood system. Long time exposure relates to cardiovascular
development, respiratory diseases and lung cancer [2]. WHO air quality guideline defines that annual mean
should not above 10 µg/m3 and 24 hours mean should not above 25 µg/m3. However, PM2.5 concentration
does not vary on only emission from source but also vary on meteorological parameters such as temperature,
wind speed, wind direction, relative humidity and planetary boundary layer height [3]. A study on influence
of meteorological factors on PM2.5 over Bangkok, Thailand: case study of Bang Na station aims to
investigate relationship of meteorological parameters and PM2.5. If its concentration does not increase from
source (including primary and secondary source), the meteorological parameters that is the major factor in
increasing concentration. The result of this study can help in providing or considering the mitigation or
policy in air pollution management in the future. The PM2.5 concentration possibly has influence from
meteorological more than precursor gas and the PM2.5 concentration has strong positive relation with relative
humidity in the most date and strong negative relative with wind speed and temperature respectively.
Moreover, PM2.5 also has medium relation with planetary boundary layer. However, in November 2016,
there are negative relation between SO2 and PM2.5 that should be investigated more.

Keywords: NO2; SO2; Planetary Boundary Layer Height; Correlation Coefficient

INTRODUCTION

PM2.5 is a harmful atmospheric composition if it is in height concentration that affect to environment and
health. it consists of 5 major components such as organic matter (OM), elemental carbon (EC), crustal
material, ammonium sulfate((NH4)2SO4) and ammonium nitrate (NH3NO3). PM2.5 can be primary pollutant
(releasing from source) and secondary pollutants (resulting from precursor gases which are SO2, NOx, VOCs,
and NH3) [1]. The primary sources of PM2.5 in general are incomplete combustion, automobile emission, dust
and cooking and secondary emission sources is chemical reaction in the atmosphere. Particulate matters in
the air can affect to human health particularly in height concentration. PM10 or particulate matter with
diameter 10 micron or less than can pass into and stay in lung but PM2.5 can pass through the lung barrier and
go inside blood system. Long time exposure relates to cardiovascular development, respiratory diseases and
lung cancer [2]. WHO air quality guideline defines that annual mean should not above 10 µg/m3 and 24 hours
mean should not (over) than 25 µg/m3. However, PM2.5 concentration does not vary on only emission from
source but also vary on meteorological parameters such as temperature, wind speed, wind direction, relative
humidity and planetary boundary layer height [3].

According to Air Quality Index (AQI) of Pollution Control Department(PCD), Ministry of Natural Resource
and Environment of Thailand, PM2.5 is a crisis issue of Bangkok nowadays because its concentration when
comparing to AQI is in range between 100-200 (orange color means that it starts to affect to popular health)

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 205 -
or 51-90 µg/m3 of 24 hours concentration. And in some day, the 24 hours of PM2.5 concentration is above 91
µg/m3 or more than 200 of AQI that surely impact to human health in the society [4].
A study on influence of meteorological factors on PM2.5 over Bangkok, Thailand: case study of Bang Na
station aims to investigate relationship of meteorological parameters and PM2.5. If its concentration does not
increase from source (including primary and secondary source), the meteorological parameters that is the
major factor in increasing concentration. The result of this study can help in providing or considering the
mitigation or policy in air pollution management in the future.
METHODOLOGY
This study uses statistical method by using correction coefficient to analysis relationship in pollutant (PM2.5
and Precursor gases (SO2 and NO2)) and relationship of PM2.5 and meteorological parameters at Bang Na air
pollution monitoring station of PCD.
The period in study was addressed in January November and December 2016 to 2018 due to peak of 24
hours average concentration as show in figure 1.1 that shows 12 months of 2016, 2017 and 2018 and
limitation of ambient hourly PM2.5 concentration.

Figure1 24-hour average of PM2.5 concentration in 2016, 2017 and 2018 (12 months)

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 206 -

RESULTS AND DISCUSSIONS
From correlation coefficient analysis to finding relationship between hourly PM2.5 concentration and
precursor gases (NO2 and SO2) and meteorological parameters (wind speed, wind direction, temperature,
relative humidity and planetary boundary layer height) in the day that have AQI more than 100 or
concentration more than 50 g/m3. The correlation coefficient value was calculated as table1but they are not
significant when using regression method(R2).

1) Relationship between NO2, SO2 and PM2.5
From correlation coefficient analysis of NO2 and PM2.5 was shown in table1. The result show correlation
coefficient (r) between 0.101 to 0.867. Due to NO2 is the precursor gas that can convert to PM2.5 (NO2 is a
substrate in formation of NH4NO3 that is a major component of PM2.5[1]) that correlation should be strong negative
because when the PM2.5 increase, the NO2 should decrease to produce the PM2.5. Therefore, the positive
correlation means that increasing of PM2.5 does not come from NO2 formation.

From SO2 and PM2.5 relationship analysis, the r value is between -0.540 to 0.832. The negative correlations were
found on 4,5,6 and 21 November 2016 but there is strong negative relation on 04 November 2016 (r = -0.540).
The negative correlation of SO2 and PM2.5 on 6 November 2016 (the other days have positive correlation) means
that PM2.5 was synthesized from SO2 (SO2 in can convert to be (NH4)2SO4 that is a major composition of PM2.5 [1])
because negative correlation means that a PM2.5 increase when the SO2 decrease. And the other days show positive
relation that mean that the PM2.5 come from the other factors.

Table 1. correlation coefficient value between hourly PM2.5 concentration and precursor gases (NO2 and SO2)

Year Month Date NO2 SO2
2016 November 04 0.838 0.827
0.523
2017 December 05 0.822 -0.540
2018 0.487
January 06 0.223 0.000
December 0.719
January 21 0.277 0.532
December 0.353
07 0.123 0.621
0.071
08 0.529 0.537
0.612
09 0.833 0.651
0.598
10 0.270 0.695
0.476
11 0.487 0.579
0.6772
12 0.474 0.2228
0.7042
13 0.759 0.7203
0.5200
14 0.211 0.7638
0.1070
19 0.582 0.5398
0.1010
20 0.801 0.8072
0.8670
21 0.573 0.6998
0.8411
25 0.861 0.5745
0.7846
26 0.704 0.6178
0.6737
15 0.4402 0.8121
0.6726
25 0.4976 0.2484
0.2288
29 0.2728

30 0.7019

31 0.4639

11 0.0344

22 0.1930

23 0.2716

26 -0.1822

16 0.3942

17 0.6016

19 0.4912

21 0.4863

22 0.0427

20 0.7142

21 0.8320

22 -0.0727

25 -0.2869

26 0.7954

27 0.5414

28 0.4577

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 207 -
a) b)

c) d)

Figure2 a) to d) show PM2.5 with NO2 and SO2 concentration trend on 4,5,6 and 21 November 2016 respectively
Relationship between meteorological factor and PM2.5
1)Wind Speed and PM2.5 relationship
The correlation between wind speed and PM2.5 was found between -0.777 to 0.406 and shows negative value 30
days from 38 selection days that have value between -0.777 to -0.003 as show in table 2. Moreover, the strong
negative correlation values were found on 15 November and 12,13 December in 2016, 29,31 January and 11
December in 2017. The relationship means that PM2.5 increase in low level of wind speed.
a) c)

b) d)

Figure3 a) and c) PM2.5concentration and wind speed trend on 4 and 5 November 2016 respectively. b) and
d) NO2 Concentration on 4 and 5 November 2016 respectively
2)Wind direction
The value of correlation coefficients (r) are between -0.543 to 0.690 and show the strong positive correlation on
23 December 2017 (r = 0.69) that mean that the PM2.5 concentration mostly influence from northwest wind in this
day. From figure below, the trend of wind direction is similar to NO2 more than PM2.5.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 208 -

c)
a)

b) d)

Figure4 a) and c) PM2.5concentration and wind direction trend on 4 and 5 November 2016 respectively. b)
and d) NO2 Concentration on 4 and 5 November 2016 respectively

3) Temperature and PM2.5 relationship
The correlation coefficient of temperature and PM2.5 are between -0.883 to 0.612. The r values have negative 33
days from 38 selection days. There is strong negative correlation between -0.883 to -0.697 on 5 November, 12,13
December 2016 and 29,31 January 2017, 11,22 December 2017, 17,19 January 2018 and 18,20 December 2018.
The strong negative correlation due to decreasing the transforming in nitrate and volatile organic component in
particle phase to gas phrase and in the winter, there are enormous of nitrate [5]. The figure shown the trend
between PM2.5 and temperature compare with NO2 (histogram) of the day that have.

a)

b)

Figure5 a) PM2.5concentration and wind direction trend on 5 November
2016 and b) NO2 Concentration on 5 November 2016.
4) Relative Humidity (R.H)
The R.H. positvely relate to PM2.5 in the most of selection date (34days from 38days). However, the
r values are between -0.335 to 0.878. In addition, there are 13 days of strong positive relation (the
value of r = 0.724 to 0.878) on 5,6 November and 12,13,19 December in 2016, 29,31 January 2017,
18 January and 20,26,27 December in 2018. The strong positive relation mean that R.H. affect to
water content of nitrate particulate matter[5]. The figures below show the trend of PM2.5, R.H and
NO2 that are similar.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 209 -
a) c)

b) d)

Figure3 a) and c) PM2.5concentration and relative humidity trend on 4 and 5 November 2016 respectively.
b) and d) NO2 concentration on 4 and 5 November 2016 respectively

5) Planetary Boundary Layer Height (PBLH) and PM2.5 relationship
The correlation values(r) are between -0.514 to 0.867 and show the medium negative relation on 22
December 2018. The different dates that consist of weak negative, strong and strong positive
ar)elation was show in figure below. Actually, the PM2.5 usually increase when the PBLH decrease
but the most case in this study shoe the positive correlation that mean PM2.5 influence from the
other factor except the day that have negative correlation.

c) e)
b)

d) f)

R = -0.347 R= -0.514 R=0.867

Figure3 a), c) and e) PM2.5concentration and PBLH trend on 4 November 2016, 22 and 27 December 2018
respectively. b), d) and f) NO2 concentration on 4 November 2016, 22 and 27 December 2018 respectively

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 210 -

Table 2 Correlation coefficient value between meteorological parameter and PM2.5

Year Month Date Wind speed Wind direction Temperature R.H PBLH
0.5464 -0.3471
04 -0.4964 0.5719 -0.1579 0.8380 0.4865
0.7325 0.3834
November 05 -0.7770 -0.5336 -0.8350 -0.2153 0.3805
06 -0.4939 0.2047 -0.6460 -0.0401 -0.2028
0.3978 -0.1316
21 0.1443 0.3395 0.1122 0.5810 -0.0635
0.2401 -0.3087
07 0.4065 0.1766 0.0619 0.2782 0.3584
0.8175 0.3241
08 -0.0026 0.0806 -0.1360 0.7451 0.6557
0.1558 0.4005
09 -0.6426 -0.1340 -0.4876 0.7525 0.3695
0.5630 -0.0676
10 -0.3618 -0.1727 -0.2136 0.4777 0.3675
0.6666 0.3218
2016 11 -0.3910 -0.1532 -0.2628 0.6148 0.1449
0.5055 0.6356
12 -0.7625 -0.3062 -0.8063 0.3127 0.5337
0.8382 0.7051
December 13 -0.6936 -0.4575 -0.7332 0.0265 -0.1340
0.8000 0.5833
14 0.2504 -0.0954 0.6125 0.7242 0.4191
0.6741 0.7774
19 -0.4873 0.3537 -0.6973 0.6046 0.4148
0.1912 0.7438
20 -0.2909 0.1476 -0.4061 0.2812 -0.3348
0.7899 0.4720
21 -0.2663 0.2547 -0.3829 0.8047 0.7531
0.4877 -0.1098
25 -0.5706 0.4120 -0.6297 0.5763 0.4108
0.7921 0.0146
26 -0.2786 0.3256 -0.4632 0.4080 0.7887
-0.1120 -0.5135
15 -0.3127 0.2712 -0.5096 0.6081 -0.3904
0.8784 0.2662
25 0.0114 0.2536 -0.5462 0.7297 0.8668
-0.3347 -0.0641
January 29 -0.7678 -0.4263 -0.8241

30 -0.0176 0.3155 -0.0185

2017 31 -0.7047 0.0598 -0.8834

11 -0.7658 -0.0703 -0.7772

December 22 -0.3009 0.3552 -0.7761
23 0.0261 0.6900 -0.4960

26 0.1013 0.1906 -0.1020

16 -0.2719 -0.0695 -0.0926

17 -0.6064 0.3577 -0.8173

January 19 -0.4015 0.1082 -0.7292

21 -0.3542 0.5997 -0.4830

22 -0.5683 0.1725 -0.6359

2018 20 -0.6574 0.2570 -0.8090
21 -0.6604 0.1428 -0.3373

22 -0.0241 0.1669 0.1684

December 25 -0.4912 -0.2036 -0.4631

26 -0.4152 0.2739 -0.6672

27 -0.3964 -0.0959 -0.6888

28 0.3023 0.2396 0.1456

CONCLUSION
The PM2.5 concentration possibly has influence from meteorological more than precursor gas and the PM2.5
concentration has strong positive relation with relative humidity in the most date and strong negative relative
with wind speed and temperature respectively. Moreover, PM2.5 also has medium relation with planetary
boundary layer. However, in November 2016, there are negative relation between SO2 and PM2.5 that should
be investigated more.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 211 -

ACKNOWLEDGEMENT
This study was success because of encouragement in scholarship from the Joint Graduate School of Energy
and Environment (JGSEE), monitoring data of air pollutants from Pollution Control Department (PCD),
WRF-Chem model from University Corporation for Atmospheric Research (UCAR), Meteorological dataset
from National Center for Atmospheric Research (NCAR). I also acknowledge Associate Professor Dr. Usa
Humphries, my advisor, for advising, providing education resource and supporting opportunity in
participating advantage workshop including activity that relate to my work including to this conference.
Moreover, I have to give the special thank Assoc. Prof. Dr. Warawut Suedee from Thammasat University in
helping me in knowledge in air pollutants and planetary boundary layer height.
REFERENCE
[1] United States Environmental Protection Agency. 2019. Fine Particulate Matter (PM2.5) Precursor

Demonstration Guidance. Research Park., New York.
[2] World Health Organization. Ambient (outdoor) air pollution. April, 2020, 01, from World Health

Organization: Ambient (outdoor) air pollution.
Web site: https://www.who.int/news-room/fact-sheets/detail/ambient-(outdoor)-air-quality-and-health.
[3] Fahimeh, H. and Azadeh, H. Influence of meteorological parameters on air pollution in Isfahan. At 3rd
International Conference on Biology, Environment and Chemistry. IPCBEE. 46(2012):7-12.
[4] Pollution Control Department. Thailand’s air quality information. March, 2020, 28, from Pollution
Control Department: air quality and noise managements bureau
Web site: http://air4thai.pcd.go.th/webV2/aqi_info.php.
[5] Liang, P. Jianming, X. Xuexi, T. Xiaoqing, M. Wei, G. and Lyu, C. Long-term measurements of
plantary boundary layer height and interactions with PM2.5 in Shanghai, China. Atmospheric Pollution
Research. 10(2019): 982-996.
[6] Danial, J. and Darrell, A. Effect of climate change on air quality. Atmospheric Environment.
43(2009): 51-63.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 212 -

I 037

Determination of Emission Factors and Chemical Properties of
Particulate Matter from Biomass Burning

Siripitch Songsompun1* Khajornsak Sopajaree2 and Ying I. Tsai3

1* Graduate student ; 2Associate Professor, Department of Environmental Engineering,

Faculty of Engineering, Chiang Mai University, Chiang Mai 50000, Thailand
3Professor, Department of Environmental Engineering and Science,

Chia Nan University of Pharmacy and Science, Taiwan

*Phone : 085-7236245, E-mail : [email protected]

ABSTRACT

One of the serious environmental problems is air pollution that the main source of this problem mostly

comes from biomass burning. Consequently, this study concentrated on collected and investigated the

particulate matter especially PM2.5 that released from the different types of biomass species that were three

types of road tree leaves and their pruning waste (Madagascar almond, Camphor and Horsetail Tree) and two

type of joss paper (burning at home and temple) were selected to burn in an open combustion chamber.

The XC-5000 Automated Isokinetic Sampling Console was used to collect the particulate matters from

biomass burning smoke. The chemical compositions, water soluble ions, carboxylic acids and total

saccharides were analyzed by ion chromatograph (IC). Results showed that the highest EF of PM2.5 was

obtained from Horsetail tree (6.51 ± 5.41 g/kg), while the lowest EF was obtained from joss paper that uses

in the temple (1.72 ± 0.23 g/kg). Moisture contents of all biomass samples were found insignificant

difference. They were in the range of 2.26% to 2.63%. CO2 was the highest gases pollutant followed by CO.

Except for Horsetail tree with organic carbon (OC) as the main carbon content in PM2.5, the other 4 types of

biomass burning was found elemental carbon (EC) as the main carbon content in PM2.5. For water-soluble

ions, the major forms of water-soluble ions are chloride (Cl-), potassium (K+), sulfate -) and sodium

(Na+). Levoglucosan was the dominant species of total saccharides in PM2.5. The highest EF of Levoglucosan

is from Madagascar almond burning (265.53 ± 0.12 mg/kg) and the lowest EF is from Camphor burning

(29.34 ± 0.03 mg/kg). In addition, the result on carboxylate expressed that Lactic acid was the dominant

species that bound on PM2.5. Based on this study, knowing the emission factor and the chemical components

can further be identified the source of particulate matters in the ambient air.

Keywords：PM2.5; emission factor; biomass burning; chemical composition

INTRODUCTION

Air pollution is one of the most serious problems in the world. It can emit from source such as transportation,
industrial process, biomass burning emission from agricultural and residential house, photochemical oxidant
etc. Many researchs pay attention to air pollution release from biomass burning that is the major source of air
pollution especially in Asia. Biomass burning is the combustion of organic matter. Burning can be from
natural or manmade fires including the burning of crop stubble, forest residues and vegetation burnt for land
clearing, burning of joss paper and burning of biomass for fuel. Biomass burning is a major source of many
air-borne particles and trace gases that influence for environment.

The important air pollutant is particulate matter (PM). PM can cause of air quality and environmentally
destructive considering its role in smog formation, decreases visibility, the cause damage to ecosystems
(Watson,2002), direct and indirect climate forcing. Furthermore, air pollution can cause long-term and short-
term health effects. It's found that the elderly and young children are more affected by air pollution. It is
potential impact on human health since it can penetrate deep into respiratory system after inhalation
(Oberdörster and Utell,2002;Knibbs et al.,2001) by enter and deposit in the lung that may infiltrate into

- 213 -

blood system. PM can cause cardiovascular health problem, increase the risk of myocardial in fraction
(Koton et al.,2013; Madrigano et al., 2013) and premature mortality.

Particulate matters consist of heterogeneous compounds in terms of chemical composition, solid or liquid
state and size (characterized in particular by the diameter of the PM). PM can divided in to 3 size that are
particle which are 2.5 to 10 µm in diameter are called coarse particles (PM10), particle less than 2.5 µm in
diameter are called fine particle (PM2.5) and include ultrafine particles of less than 0.1 µm (PM0.1). The
particle size is a very important characteristic of particulate matter to find source and determines the effect of
them. In addition, PM are different in size, they are different in chemical composition depending on origin.
Their chemical composition likes salts, organic carbon compounds, trace elements and black carbon. The
typical chemical composition of biomass combustion are CO, CH4, VOCs, water soluble ion, heavy metals,
PAHs, etc (Radzi Bin Abas et al., 2004).

Since air pollution from biomass burning is become the most important problem nowadays. Many
researchers have been studied about air pollution that emitted from Agricultural waste but there are not
much the researches that pay attention to the air pollution which come from the burning of the dry leaves and
branch that fall from the roadside tree. Consequently, this study concentrated on collected and investigated
the particulate matter especially PM2.5 that released from the different types of biomass species that were
three types of road tree leaves and their pruning waste (Madagascar almond, Camphor and Horsetail Tree)
and two type of joss paper (burning at home and temple) were selected to burning in an open combustion
chamber. In addition, the XC-5000 Automated Isokinetic Sampling Console will be used to collect the
particulate matters from biomass burning smoke. The chemical compositions, water soluble ions, carboxylic
acids and total saccharides, will be analyzed by ion chromatograph (IC).

METHODOLOGY

1. Biomass burning experiment

For the experiment, There are 5 types of material that were chosen to study in this research. There are 3 types
of road tree leave and their pruning waste that are Madagascar almond, camphor and horsetail tree. There are
the abundant plant in Tainan city. Moreover, there are 2 types of joss paper that used for burning at home or
commercial and burning in the temple. The samples were burning in an open combustion chamber (Figure
1.) that had been controlled on the combustion conditions. The burning process was taken around 3-20
minutes and weight of biomass samples were about 1-4 kg. depending on biomass types. Which descripted
as the following :

Table 1 Weight, combustion time and number of biomass sample burning.

Madagascar almond Weight (kg) Time (min) Number
Camphor 1.1 - 3.1 6 - 11 3
Horsetail Tree 2.8 - 3.4 16 - 20 2
Joss paper- Home 1.4 - 2.2 3-7 3
Joss paper- Temple 3.3 - 4.0 20 2
3.3 - 4.0 20 2

- 214 -

Figure 1 An open combustion chamber

2. PM2.5 collection and gas measurement

The collection of smoke for analysis of particulate matter and gas from biomass burning is shown in Figure
2. The analysis of total particulate matter and PM2.5 were based on the method of the US Environmental
Protection Agency's Method 201A. The equipment used for sampling this sample was the XC-5000
Automated Isokinetic Sampling Console from APEX Instruments, USA. The total particulate matter (TPM)
were included the filterable PM2.5 (FPM) and the condensable PM2.5 (CPM) that were collected by 47 mm
quartz filter paper.

The exhaust gases from combustion of biomass burning were collected and monitored using continuous
analyzers brand HORIBA model PG-250A to measure SO2, NOx, CO2 and CO. THC was measured by THC
/ SRI / 8610C analyzer

FPM SOx/NOx/O2/
CO/CO2
analyzer

THC
analyzer

投料口 Isokinetic
進氣控制 Sampler

CPM

Figure 2 Diagram of PM2.5 and gas sampling process

- 215 -

3 Chemical analysis of PM2.5 sample

3.1 Analysis of total carbon (TC), elemental carbon (EC) and organic carbon (OC)

The carbon content were analyzed by elemental analyzer refer Environmental Protection Agency Standard

Method M403.01B Elemental Analysis Method, Brand / Model: ELEMENTAR / Vario EL cube. The

experiment were divided into two parts. The first part of sample that direct analysis of carbon content, the
measured results represent the total carbon. The other part of sample were analyzed after incubated 350 oC,

30 minutes, the measured results represent the elemental carbon.

3.2 Analysis of water soluble ion

The sampling filters were extracted to find anion (F-, Cl-, -, - , - and -) and cation (Na+, ,
K+, Ca+ and Mg+) by ion Chromatograph. Anion was analyzed by an ion chromatography system (IC) model

Dionex ICS-5000 DC. For cation, an ion chromatography system model ICS-1100 was used.

3.3 Analysis of total saccharide.

There are 13 types of sugar that were analyzed in this experiment. They can divide into three major
functional groups that were anhydrosugar (Levoglucosan, Mannosan and Galactosan), sugar (Trehalose,
Mannose, Glucose and Galactose) and sugar alcohol (myo-Inositol, Erythrital, Xylitol, Arabitol, Mannital
and Sorbitol). The total saccharide were analyzed by an ion chromatography system (IC) model Dionex ICS-
5000 DC.

3.4 Analysis of carboxylic acid

There are 15 species of carboxylic acid that used to identify in this experiment. Fifteen carboxylic acid
functional group including Lactate, Acetate, Formate, Pyrurate, Glyoxylate, Glutarate, Succinate, Malate,
Malonate, Tartate, Maleate, Fumarate, Oxalate, Phthalate and Citrate were identified and quantified by an
ion chromatography system (IC) model Dionex ICS-5000 DC.

RESULT AND DISCUSSION

1. Moisture content and emission factor of pollutant gases

Moisture content of all types of biomass samples are insignificant difference (as show in Table 2). It ranges
from 2.18% to 2.77%. Horsetail tree samples contain higher moisture content (2.61 ± 0.16%) than other
sample types. On the other hand, Madagascar almond contain lower moisture content (2.26 ± 0.07%) than
other samples types.

(Permadi et al., 2012) studied on biomass combustion in Asia, the result showed that large amounts of CO
and CO2 are emitted during biomass combustion, both of which account for more than 90% of the total
emissions. In this study, the emission factors of gaseous pollutants of five types of biomass sample show in
Table 2. CO2 is the highest gases pollutant following by CO. The emission factor of CO2 from Camphor is
higher than the other biomass sample types (1,393.27 ± 174.62 g/kg).

Table 2 % Moisture content and emission factor of gas pollution from different type of biomass
burning samples (g/kg).

Madagascar Camphor (n=2) Horsetail tree Joss paper – Joss paper –
almond (n=3) (n=3) Home (n=2) Temple (n=2)
2.50 ± 0.11%
Moisture 2.26 ± 0.07% 1393.27 ± 174.62 2.61 ± 0.16% 2.49 ± 0.35% 2.47 ± 0.29%
Content
CO2 762.56 ± 333.94 37.13 ± 1.86 789.98 ± 649.33 764.08 ± 235.13 988.05 ± 207.18
CO 31.23 ± 14.90 0.78 ± 0.10 15.90 ± 16.48 11.52 ± 0.74 16.57 ± 0.58
2.44 ± 0.34 0.30 ± 0.28 0.49 ± 0.02 0.80 ± 0.31
SO2 0.49 ± 0.15 4.75 ± 0.95 0.75 ± 0.76 0.85 ± 0.02 0.99 ± 0.15
2.08 ± 0.87
NOx as NO2 3.37 ± 2.02 2.40 ± 1.83 0.85 ± 0.60
TNMHC as 3.43 ± 1.38
CH4

- 216 -

2. Particle mass emission factor

Experimental discussion the mass emission factor of total particulate matter (TPM), filterable PM2.5 (FPM)
and condensable PM2.5 (CPM) are show in table 3. For three of side-road trees, the highest emission factor of
PM2.5 is from Horsetail tree (6.51 ± 1.35 g/kg) and the lowest is from Camphor (2.58 ± 1.66 g/kg). As for
joss paper, there are no significant difference are found in emission factor.

Table 3 Emission factor of PM2.5 in different type of biomass burning samples (g/kg).

Madagascar almond (n=3) FPM Emission factor (g/kg) TPM
Camphor (n=2) 2.64 ± 1.35 CPM 3.53 ± 1.23
Horsetail Tree (n=3) 2.52 ± 1.67 2.58 ± 1.66
Joss paper- Home (n=2) 4.58 ± 3.90 0.89 ± 0.70 6.51 ± 5.41
Joss paper- Temple (n=2) 1.55 ± 0.39 0.06 ± 0.01 1.79 ± 0.65
1.41 ± 0.16 1.93 ± 1.68 1.72 ± 0.23
0.23 ± 0.25
0.31 ± 0.39

3. Total carbon

The result of total carbon (TC) emission factor for five types of biomass burning is show in Table 4. Except
for Horsetail tree burning emitted organic carbon (OC) as the main carbon content that bound on PM2.5, the
other 4 types of biomass burning are found elemental carbon (EC) as the main carbon content that bound on
PM2.5.

Table 4 Emission factor of TC, EC and OC in different types of biomass burning samples (mg/kg).

Emission factor (mg/kg)

TC EC OC

Madagascar almond (n=3) 1537.57 ± 230.91 944.13 ± 282.69 593.43 ± 53.13
Camphor (n=2)
Horsetail Tree (n=3) 1382.65 ± 916.68 794.85 ± 423.14 587.80 ± 493.54
Joss paper- Home (n=2)
Joss paper- Temple (n=2) 1984.07 ± 1324.43 509.15 ± 275.48 1474.92 ± 1331.28

881.67 ± 198.47 651.68 ± 115.98 230.00 ± 82.49

1230.85 ± 127.77 804.27 ± 147.29 426.58 ± 19.52

4. Emission factor of water soluble ion

The results show that major forms of water-soluble ions are chloride (Cl-), potassium (K+), sulfate ( -) and
sodium (Na+) emitted from biomass burning (as show in Table 5), which is consistent with (Chantara, S et

al.,2019). But, it is different in water soluble ion in PM2.5 in ambient air. While in ambient air is the

dominant species of cation and - is the dominant species of anion (Cheng, Y.-H et al., 2010). K+ and Cl-

are the dominant species in three of road-side trees samples, while Na+ and - are the dominant species in
joss paper samples. For Horse tail tree, Na+ and Cl- are the dominant species, which can be explain that

Horsetail tree is planted near to the sea shore, so it will have influence of NaCl from the sea.

- 217 -

Table 5 Emission factor of water soluble ion from different types of biomass burning samples (mg/kg).

Madagascar Camphor (n=2) Horsetail tree Joss paper – Joss paper –
almond (n=3) (n=3) Home (n=2) Temple (n=2)

Cation 6.55 ± 2.24 2.40 ± 1.07 267.27 ± 215.84 54.76 ± 23.93 38.84 ± 2.16
Na+ 17.32 ± 0.33 11.76 ± 10.73
11.66 ± 2.49
39.28 ± 25.46 5.18 ± 0.68 63.33 ± 55.25 0.22 ± 0.16 6.88 ± 0.17
1.91 ± 0.48 0.15 ± 0.12
K+ 73.07 ± 33.75 214.51 ± 97.87 63.95 ± 53.04 1.21 ± 0.07
Mg2+ 0.61 ± 0.36 0.17 ± 0.02 0.76 ± 0.34 NDa
63.82 ± 27.51 NDa
Ca2+ 2.35 ± 1.31 1.49 ± 0.92 5.95 ± 4.74 37.73 ± 13.49
0.06 ± 0.03
Anion 3.07 ± 1.27 0.11 ± 0.11
107.10 ± 20.44 5.06 ± 4.79
F- NDa NDa NDa 2.30 ± 0.28 50.46 ± 57.74
0.67 ± 0.54
Cl- 164.73 ± 65.06 171.08 ± 85.67 714.63 ± 451.23

0.17 ± 0.09 0.10 ± 0.08 0.17 ± 0.17

9.66 ± 8.86 8.10 ± 1.78 5.27 ± 1.41

43.86 ± 11.73 25.51 ± 33.31 45.69 ± 4.43

0.77 ± 0.20 0.57 ± 0.04 0.85 ± 0.16

a. ND denotes not detected or lower than background level.

5. Emission factors of Total saccharide.

Levoglucosan is the dominant species of total saccharides that bound on PM2.5. The highest EF of
Levoglucosan is from Madagascar almond burning (265.53 ± 125.25 mg/kg) and the lowest EF is from
Camphor burning (12.27 ± 4.83 mg/kg). The ratio between levoglucosan and mannosan from the three types
of side-road trees show that there is the highest in Horsetail tree and the lowest in Camphor, indicate the
higher cellulose in hardwood. However, there is a difference in the ratio in joss paper, indicate that the
sources of materials of joss paper are more diverse.

Table 6 Emission factor of levoglucosan and mannosan (mg/kg) and ratio of levoglucosan and
mannosan

Madagascar almond (n=3) Levoglucosan Mannosan Levoglucosan/mannosan
Camphor (n=2) 265.53 ± 125.25 12.04 ± 6.32 22.06
Horsetail Tree (n=3) 0.85 ± 0.66 14.46
Joss paper- Home (n=2) 12.27 ± 4.83 4.72 ± 4.42 35.54
Joss paper- Temple (n=2) 167.64 ± 173.61 1.10 ± 0.40 45.94
50.31 ± 18.59 2.96 ± 0.92 17.95
53.20 ± 50.52

6. Emission factor of Carboxylate

Figure 3 show the emission factor of total carboxylate from different type of biomass samples.
Madagascar almond emit the highest of total carboxylate and the lowest of total carboxylate is from Joss
paper-temple. Total carboxylate mostly found in the three types of side-road tree samples more than the joss
paper sample. Moreover, Lactate is the dominate species of carboxylate that bound on PM2.5.

- 218 -

Total Carboxylate

Emission factor (mg/kg) 30.00
25.00
20.00
15.00
10.00

5.00
0.00

Madagascar Camphor Horsetail Joss paper- Joss paper-
almond Tree Home Temple

Figure 3 Emission factor of total carboxylate in different types of biomass burning sample

CONCLUSION

In this study, biomass burning sample were burned in an open combustion chamber. PM2.5 and gases emitted
from biomass burning samples were collected in order to estimate the emission factors (EF) of various
chemical properties. The highest EF of PM2.5 is from Horsetail tree (6.51 ± 1.35 g/kg) while the lowest EF is
obtained from joss paper that uses in the temple (1.72 ± 0.23 g/kg). For water soluble ion, K+ and Cl- are the
dominant species in side-road tree waste samples, while Na+ and ( -) are the dominant species in joss
paper samples. Levoglucosan is the dominant species of total saccharide, so levoglucosan has been
recommended as a tracer for biomass combustion. Moreover, lactate is the dominant species of total
carboxylate. The total carboxylate mostly emitted from the three of side-road tree, indicate that carboxylate
is include in organism.

REFERENCE

[1] Chantara, S., Thepnuan, D., Wiriya, W., Prawan, S., & Tsai, Y. I. 2019. Emissions of pollutant
gases, fine particulate matters and their significant tracers from biomass burning in an open-system
combustion chamber. Chemosphere, 224: 407-416.

[2] Cheng, Y.-H., & Li, Y.-S. 2010. Influences of Traffic Emissions and Meteorological Conditions on
Ambient PM10 and PM2.5 Levels at a Highway Toll Station. Aerosol and Air Quality Research,
10(5): 456-462.

[3] Knibbs, L. D., Cole-Hunter, T., & Morawska, L. 2011. A review of commuter exposure to ultrafine
particles and its health effects. Atmospheric Environment, 45(16): 2611-2622.

[4] Koton, S., Molshatzki, N., Yuval, Myers, V., Broday, D. M., Drory, Y., . . . Gerber, Y. 2013.
Cumulative exposure to particulate matter air pollution and long-term post-myocardial infarction
outcomes. Preventive Medicine, 57(4): 339-344.

[5] Madrigano, J., Kloog, I., Goldberg, R., Coull, B. A., Mittleman, M. A., & Schwartz, J. 2013. Long-
term exposure to PM2.5 and incidence of acute myocardial infarction. Environmental Health
Perspectives, 121(2): 192-196.

[6] Oberdörster, G., & Utell, M. J. 2002. Ultrafine particles in the urban air: To the respiratory tract -
Ang beyond? Environmental Health Perspectives, 110(8): A440-A441.

[7] Permadi, D. A., & Kim Oanh, N. T. 2013. Assessment of biomass open burning emissions in
Indonesia and potential climate forcing impact. Atmospheric Environment, 78: 250-258.

[8] Radzi Bin Abas, M., Rahman, N. A., Omar, N. Y. M. J., Maah, M. J., Abu Samah, A., Oros, D. R.,.
Simoneit, B. R. T. 2004. Organic composition of aerosol particulate matter during a haze episode in
Kuala Lumpur, Malaysia. Atmospheric Environment, 38(25): 4223-4241.

[9] Watson, J. G. 2002. Visibility: Science and Regulation. Journal of the Air & Waste Management
Association, 52(6): 628-713.

- 219 - I 038

Effect of Salinity on Chlorella vulgaris for Increasing Lipid Content

Vicheka Keo1 and Thaniya Kaosol2*

1Graduate student, Department of Civil Engineering, Faculty of Engineering,
Prince of Songkla University, Songkhla 90110, Thailand.

2*Assoc.Prof., Department of Civil Engineering, Faculty of Engineering,
Prince of Songkla University, Songkhla 90110, Thailand.

*Contact No: 6674-287-136 Fax: 6674-287012 e-mail: [email protected]

ABSTRACT

Microalgae growth with effluent from the frozen seafood wastewater treatment plant can provide some
benefits such as produce high biomass and lipid for biodiesel production. Chlorella vulgaris is considered as
the best feedstock of energy to produce biodiesel. The objective of this study is to increase lipid content
under salinity stress. There are three reactors including R1, R2 and R3 for cultivating microalgae. The R1
used the effluent of seafood processing wastewater treatment plant without adding the NaCl. The R2 and R3
were added with the NaCl for increasing the salinity. The various NaCl concentrations of R1, R2 and R3
were 1.34 g/L (0.023 M), 2.9 g/L (0.050 M) and 4.4 g/L (0.075 M), respectively. In this study, Chlorella
vulgaris growth in R1 and R2 was reached the maximum of DCW (Dry Cell Weight) about 1.02 g/L and
1.16 g/L on Day-3, respectively. While the R3 was reached the maximum of DCW about 1.47 g/L on Day-4.
Furthermore, the total lipid content of Chlorella vulgaris was increased in different concentration of salinity.
The total lipid content in R2 was lower than in R3, of which R2 and R3 contained 1.84% and 3.09% of lipid
content, respectively. However, both reactors were lower than the lipid content of R1 which was 4.60% of
lipid content. It could be concluded that the lipid content in Chlorella vulgaris strain was enhanced slightly
between various concentrations of salinity. Therefore, the effluent from frozen seafood factory was suitable
for growth Chlorella vulgaris without adding NaCl. The salinity content in the effluent from frozen seafood

factory was enough for microalgae growth and the nutrient contained in the effluent was also removed by
microalgae cultivation.

Keyword: Microalgae growth; Chlorella vulgaris; Lipid; Salinity

INTRODUCTION

The high requirement of energy in the world is the crucial crisis faced nowadays, and the fossil fuel is
gradually decreased. The biomass is the sustainable energy source which can be utilized and substituted the
carbon from fossil fuel source to produce the production of carbon-base. The production of carbon-base
includes chemicals, raw materials, and liquid fuels [1]. The consumption of fossil fuel causes many problems

such as energy requirement, economic issue, rising of fuel utilization and fuel price and the releasing of
pollution gases [2]. Moreover, fossil fuel burning induce to rise the greenhouse gas releasing and the
environmental issue to the earth [3]. Therefore, the effect of fossil fuel burning has been seriously concerned
nowadays [4]. Thus, the bioenergy has been interested for energy production. Microalgae are considered as
the best feedstock for converting to biodiesel production [5]. There are many advantages of microalgae

bioenergy which produce high lipid content, high growth rate, the precious of chemical production, great

ability to absorb carbon dioxide and being able to combine with wastewater treatment for energy production
[6]. Moreover, the microalgae cultivation does not need large land to grow compared to vegetable
cultivation, which is also the renewable source for biodiesel production but requires large area [7]. The
microalgae can produce lipid content and biomass productivity which are the most important parameters.

They require several limited conditions to increase lipid accumulation such as light, temperature, pH and
nutrient [8]. Some researchers reported the lipid increment of microalgae cultivating in different conditions,
such as the stress condition induced high lipid content in microalgae cell [2, 9]. Salinity is the one essential
stress component for microalgae. It leads to change the metabolic in nutrient absorption, increase toxic ions,
create osmose stress, and make oxidative stress [10]. The salinity stress induces to increase lipid productivity
is suggested by previous researchers [11, 12]. The seafood processing wastewater contains high organic
content which comes from blood, fish heads, intestine, and meat residuals [13]. In addition, microalgae are

known that can use organic and inorganic nutrient in wastewater and produce biomass for biofuel

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 220 -

production. Meanwhile, microalgae have ability to uptake nutrient which are phosphorus and nitrogen
contained in wastewater. Furthermore, there are few researches that have evaluated the potential of effluent
reuse [14].

In this study, the effluent of frozen seafood processing and microalgae Chlorella vulgaris strain are utilized
for cultivation due to high organic contained in effluent. Thus, the objective was to study the effect of
salinity stress on Chlorella vulgaris for lipid content increment.

METHODOLOGY

1. Microalgae strain and pre-cultivation

The Chlorella vulgaris strain was obtained from the Coastal Aquaculture Research Institute in Songkhla
province, Thailand. The Chlorella vulgaris strain was cultivated in Urea fertilizer which contain Urea (0.2
g/L), Diammonium phosphate (0.003 g/L), CaO (0.2 g/L) and Glutamic mother liquid (0.8 mL/L). The
sample was incubated in 5 L of bottle at the ambient temperature with light intensity at 6,000 Lux.

2. Microalgae growth determination
Cell density was determined by measuring optical density (OD) using spectrophotometer at 680 nm. The

microalgae were harvested by centrifugation at 5,000 rpm for 10 min. After that, the microalgae are rinsed
with distilled water two times to take medium off. Then the purified microalgae were dried in the oven at

105ºC until obtaining constant weight and cooled in the desiccator. The linear relationship between OD680
and dry cell weight (DCW, g dry weight/L) is shown as equation (1).

( / )= . R2 = 0.9996 (1)

Where: DCW1 and DCW2 are dry cell weight (g/L) at time t1 and t2, respectively. The biomass productivity
(Pbiomass, g/L/d) is determined by equation (2).

= - ()
-

The growth rate per day (µ, d-1) is determined by equation (3):

µ= ( - ) ()
-

3. Experimental design
The effluent of frozen seafood processing factory was used for microalgae cultivation. Furthermore, the

batch experiment was used in the photobioreactor (length 60 cm, diameter 20 cm) and filled with 10 L of
effluent into each reactor. There were three reactors such as R1, R2 and R3 for cultivating microalgae. The

R1 was used the effluent of seafood processing wastewater treatment plant without adding NaCl. The R2 and
R3 were added NaCl for increasing salinity. The various NaCl concentrations of R1, R2 and R3 were 1.34
g/L (0.023 M), 2.9 g/L (0.050 M) and 4.4 g/L (0.075 M), respectively. The microalgae were cultivated until

reach the stationary phase. The photobioreactor is used for microalgae cultivation as shown in Figure 1.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 221 -
Diameter 20 cm

Sample volume 10L Pipe

Length 50 cm

Air plate Air pump
diffuser

Figure 1 Photobioreactor Used for Microalgae Cultivation

4. Lipid extraction
The lipid of microalgae was extracted from microalgae following the Folch method [15]. The biomass dry

weight was homogenized with Chloroform: Methanol (2:1 by volume) in the fume hood. The mixture was
put in ultrasonic cleaner for 30 min. The solvent mixture was separated into two phases by centrifugation at

5,000 rpm for 15 min. The upper phase contained all the non-lipid compounds and these compounds were
evaporated in the fume hood. The lower phase contained the solid which was contained lipid and taken it for
repeat again and again until obtaining the clear water. The lipid content was calculated by equation (4).

(%) = () % ()
()

RESULTS AND DISCUSSIONS

1. Effluent properties before cultivation

In this study, the properties of frozen seafood effluent are shown as nutrient for microalgae growth such as
TKN, TP, and COD were measured. Moreover, the turbidity, salinity, TDS and pH were also measured
before growing microalgae (Table 1). These parameters were analyzed to recognize the ability of Chlorella

vulgaris growth of using nutrient in effluent and to compare between the initial and final concentration of
effluent after microalgae cultivation.

Table 1 Initial concentration of effluent from frozen seafood factory

TKN TP COD Turbidity Salinity TDS pH
(mg/L) (mg/L) (mg/L) (NTU) (ppt) (mg/L) 7.7
152.3 8.4 1.6
8.1 128 1,061

2. Monitoring parameters
Salinity and pH were observed for the changing of their quantity during cultivation time. In this study, the quantity

of salinity in R1 and R2 were decreased steadily until Day-3 and increased on Day-4. However, The R3 was
decreased until Day-2 and increased from Day-3 to Day-4. The quantities of salinity of all reactors were decreased

again at last day of cultivation (Figure 2A). The decreasing of salinity indicated that microalgae consumed it for
their growth. Salinity was the major parameter that provided the development of plant surviving and made the

blocking on metabolic activity of photosynthesis [16]. The previous study reported that the salinity could be
increased because of the evaporation [17]. Even the salinity could block the photosynthesis but Chlorella vulgaris
was able to grow normally due to it was cultivated in the suitable concentration of salinity.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 222 -

Moreover, pH was also changed as its natural in the cultivation reactor. In this study, pH in all reactors were
increased from the first day until Day-3. For three days, pH of R1, R2, and R3 was increased from 7.95 to 8.74,

8.11 to 8.89 and 8.15 to 9.07, respectively. The previous study reported that pH increased gradually when
microalgae synthesized and consumed CO2 at daytime [18]. Additionally, the other research described the cell of
microalgae excreted hydroxyl ion to the cultivation medium during it used HCO3 as CO2. The reducing of HCO3
led to increase the pH [19]. However, it was demonstrated that the respiration process at night time induces pH
decreasing [18].

5 R1 R2 R3 9.5 R1 R2 R3

4
9

Salinity (ppt)
pH
3
8.5

2

8
1

0 7.5 1234 5
012345 0 Cultivation time (day)
Cultivation time (day)

Figure 2 Results of Salinity (A) and pH (B)

3. Microalgae growth

3.1 Dry cell weight of Chlorella vulgaris strain
The Chlorella vulgaris strain was cultivated for 5 days in this study. The dry cell weight (DCW) was a
parameter which was used for estimating microalgae growth. The microalgae growth in R1 and R2 were
reached the maximum DCW about 1.02 and 1.16 g/L on Day-3, respectively. While the microalgae growth in
R3 was reached the maximum DCW about 1.47 g/L on Day-4 (figure 3). The other research reported that
Chlorella vulgaris growth was decreased in various high concentrations of salinity which the salinity
concentrations were 0, 0.26, 0.51 and 0.77 M [20]. However, this study observed that Chlorella vulgaris could
increase their growth in increasing of salinity at 0.05 and 0.075 M. The previous study demonstrated that
microalgae were able adapt in the rank of salinity at 0.05, 0.15 and 0.20 M [21]. According to the previous
study, the rank of salinity in this study was grown in the suitable condition of salinity.

Dry Cell Weight (g/L) 1.6 R1 R2 R3
1.2
0.8 1234 5
0.4 Cultivation time (day)
0.0

0

Figure 3 Chlorella Vulgaris Growth as Dry Cell Weight

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 223 -

3.2 Biomass productivity of Chlorella vulgaris strain
Moreover, the highest biomass productivity of Chlorella vulgaris for R1 and R2 was reached on Day-2,
whereas R3 was reached on Day-3. In addition, the highest biomass productivity of R1, R2 and R3 were
0.395, 0.424 and 0.322 g/L/d, respectively (figure 4).

Biomass Productivity 0.5 R1 R2 R3
(g/L/d) 0.4 3
0.3 12
0.2 Cultivation time (day)
0.1
0.0

0

Figure 4 Chlorella Vulgaris Growth as Biomass Productivity

3.3 Specific growth rate of Chlorella vulgaris
Furthermore, the specific growth rate was determined by calculating with dry cell weight curve. The
maximum growth rate was observed on Day-2 for R1, R2 and R3 which was 0.525, 0.595 and 0.405 d-1,
respectively. These microalgae growth rates were decreased gradually until last day of cultivation (figure 5).

0.6 R1 R2 R3
0.5
Specific growth rate (µ/d) 0.4
0.3
0.2 12 3
0.1 Cultivation time (day)
0.0

0

Figure 5 Chlorella Vulgaris Growth as Specific Growth Rate

4. Effect of salinity on lipid content
Total lipid content of Chlorella vulgaris was increased in different concentration of salinity. In this study, the
total lipid content in R2 was lower than R3, of which the R2 and R3 contained 1.84% and 3.09% of lipid
content, respectively. However, the lipid content of both reactors were lower than the lipid content of R1
which was 4.60% of lipid content (figure 6). Similarly, the previous study showed that the lipid content was
increased depending on NaCl concentration, but it was lower than effluent without adding NaCl [22]. In
addition, the other studies reported that total lipid content was increased slightly in high concentration of

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 224 -

salinity. However, the previous study showed that the increasing salinity concentration was able to
accumulate lipid content for Chorella sp. which was increased from 1.92 ± 0.012% to 3.49 ± 0.016% [23].

Lipid content (%) 5

4 R3

3

2

1

0
R1 R2
Reactor

Figure 6 Effect of Salinity on Lipid Content of Chlorella Vulgaris Growth as Specific Growth Rate

5. The efficiency of nutrient uptake from effluent
Not all of kind and source of wastewater are suitable for microalgae to growth. The different microalgae can
adapt with different type of wastewater; it is also depended on the characteristic of wastewater conditions.

Moreover, Chlorophyte microalgae can utilize the various kind of wastewater and have the capacity to
remove nutrient waste from wastewater as well [3]. In this study, Chlorella vulgaris was a species of
Chlorophyte which was used for growing in the frozen seafood effluent. As mentioned above, the microalgae
used nutrient from frozen seafood effluent as their food such as TKN, TP, and COD. These parameters were
analyzed after cultivating to know how many percent that microalgae could uptake. The final concentration

of each reactor are shown in Table 2, presenting that the microalgae could uptake TKN, TP, and COD in R1
were 51.7%, 24.5% and 25.0%, respectively. While the efficiency of uptake TKN, TP, and COD in R2 were
61.5%, 26.0% and 43.8%, respectively. Whereas the efficiency of uptake TKN, TP, and COD in R3 were
48.9%, 23.6% and 31.8%. The efficiency of nutrient uptake in each reactor are shown in Table 3. The

efficiency of nutrient uptake was lower than the previous research which TKN and TP could be removed in
the rank from 72% to 85% and 57% to 77%, respectively [24]. The previous research reported that total
nitrogen (TN) and total phosphorus (TP) could be removed from 11.9% to 74.3% and 22.5% to 94.8%,
respectively [25]. Thus, it could be assumed that Chlorella vulgaris had highly efficient to uptake the
nutrient from the frozen seafood effluent. The phytoremediation is the common wastewater treatment by
microalgae for pollutants removal such as organic and inorganic [26].

Table 2 Final concentration of effluent after growing microalgae

Reactor TKN TP COD Turbidity Salinity TDS pH
(mg/L) (mg/L) (mg/L) (NTU) (ppt) (mg/L)
1 8.97
2 73.6 6.1 96.0 3.7 1.2 970 8.46
3 58.7 6.0 72.0 3.4 2.8 8.94
78.1 6.2 87.2 4.2 4.4 2,079

3,803

Table 3 Efficiency of nutrient removal

Reactor TKN Efficiency of nutrient removal (%) COD
51.7 25.0
1 TP
2 61.5 24.5 43.8
3 26.0
48.9 23.6 31.8

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 225 -

CONCLUSION
The effect of salinity concentration on lipid content of Chlorella vulgaris was observed that the salinity
concentration of 0.023, 0.05 and 0.075 M induced to increase lipid content about 4.60%, 1.84% and 3.09%,
respectively. The effluent from frozen seafood factory contained 0.023 M of salinity, the growth of Chlorella
vulgaris with this effluent provided 4.60% of lipid content increasing. Then, the effluent from frozen seafood
factory was suitable for microalgae growth. It could be concluded that the salinity increment led to enhance
lipid content in Chlorella vulgaris strain. But it was slightly enhanced among various concentrations of
salinity.

ACKNOWLEGEMENTS

The authors would like to thank the Royal Scholarship under Her Royal Highness Princess Maha Chakri
Sirindhorn Education Project to Kingdom of Cambodia, Faculty of Engineering (PSU) and PSU graduate
school for providing financial support.

REFERENCES
[1] Veronesi, D., Idà, A., D’Imporzano, G. and Adani, F. 2015. Microalgae cultivation: nutrient recovery

from digestate for producing algae biomass. Chemical Engineering Transactions, 43: 1201-1206.
[2] Chi, N.T.L., Duc, P.A., Mathimani, T. and Pugazhendhi, A. 2019. Evaluating the potential of green

alga Chlorella sp. for high biomass and lipid production in biodiesel viewpoint. Biocatalysis and
Agricultural Biotechnology, 17: 184-188.
[3] Chui, S.Y., Kao, C.Y., Chen, T.Y., Chang, Y.B., Kuo, C.M. and Lin, C.S. 2015. Cultivation of
microalgal Chlorella for biomass and lipid production using wastewater as nutrient resource.
Bioresource Technology, 184: 179-189.
[4] Gonçalves, A.L., Pire, J.C. and Simões, M. 2013.Lipid production of Chlorella vulgaris and
Pseudokirchneriella subcapitata. International Journal of Energy and Environmental Engineering,
4(1): 1-6.
[5] Danquah, M.K., Gladman, B., Moheimani, N., and Forde, G.M. 2009. Microalgal growth
characteristics and subsequent influence on dewatering efficiency. Chemical Engineering Journal, 151:
73-78.
[6] Wu, Y.H., Hu, H.Y., Yu, Y., Zhang, T.Y., Zhu, S.F., Zhuang, L.L., Zhang, X., and Lu, Y. 2014.
Microalgal species for sustainable biomass/lipid production using wastewater as resource: A review.
Renewable and Sustainable Energy Reviews, 33, 675-688.
[7] Danquah, M.K., Gladman, B., Moheimani, N. and Forde, G.M. 2009. Microalgae growth
characteristics and subsequent influent on dewatering efficiency. Chemical Engineering Journal, 151:
73-78.
[8] Shin, Y.S., Choi, H.I., Choi, J.W., Lee, J.S., Sung, Y.J. and Sim, S.J. 2018. Multilateral approach on
enhancing economic viability of lipid production from microalgae: A review. Bioresource
Technology, 258: 335-344.
[9] Yoo, C., Jun, S.Y., Lee, J.Y., Ahn, C.Y., and Oh, H.M. 2009. Selection of microalgae for lipid
production under high levels carbon dioxide. Bioresource Technology, 101: S71–S74.
[10] Kebeish, R., El-Ayouty, Y. and Hussein, A. 2014. Effect of salinity on biochemical traits and
photosynthesis-related gene transcription in Chlorella vulgaris. Egyptian Journal of Botany, 54(2):
281- 294.
[11] Ben Moussa-Dahmen, I., Chtourou, H., Rezgui, F., Sayadi, S. and Dhouib, A. 2016. Salinity stress
increases lipid, secondary metabolites and enzyme activity inAmphora subtropicaandDunaliellasp. for
biodiesel production. Bioresource Technology, 218: 816-825.
[12] Mohan, S.V. and Devi, M.P. 2014. Salinity stress induced lipid synthesis to harness biodiesel during
dual mode cultivation of mixotrophic microalgae. Bioresource Technology, 165: 288-294.
[13] Jayashree, C., Tamilarasan, K., Rajkumar, M., Arulazhagan, P., Yogalakshmi, K.N., Srikanth, M. and
Banu, J.R. 2016. Treatment of seafood processing wastewater using upflow microbial fuel cell for
power generation and identification of bacterial community in anodic biofilm. Journal of
Environmental Management, 180: 351-358.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 226 -

[14] Guimarães, J.T., Souza, A.L.M., Brígida, A.I.S., Furtado, A.A.L., Chicrala, P.C.M.S., Santos, V.R.V.,
Rosiana, R.A., Luiz, D.B. and Mesquita, E.F.M. 2018. Quantification and characterization of effluents
from the seafood processing industry aiming at water reuse: A pilot study. Journal of Water Process
Engineering, 26: 138-145.

[15] Folch, J., Lees, M., and Sloane Stanley, G.H. 1957. A simple method for the isolation and purification
of total lipides from animal tissues. The Journal of Biological Chemistry. 226(1): 497–509.

[16] Adenan, N. S., Yusoff, F. M. and Shariff, M. 2013. Effect of Salinity and Temperature on the Growth
of Diatoms and Green algae. Journal of Fisheries and Aquatic Science 8 (2): 397- 404.

[17] Ahmad, A., Buang, A. and Bhat, A.H. 2016. Renewable and sustainable bioenergy production from
microalgal co-cultivation with palm oil mill effluent (POME): A review. Renewable and Sustainable
Energy Reviews, 65: 214-234.

[18] Sajjadi, B., Chen, W.Y., Raman, A.A.A. and Ibrahim, S. 2018. Microalgae lipid and biomass for
biofuel production: A comprehensive review on lipid enhancement strategies and their effects on fatty
acid composition. Renewable and Sustainable Energy Reviews, 97: 200-232.

[19] Mandalam, R. K. and Palsso, B. 1995. Chlorella vulgaris (Chlorellaceae) does not screte autoinhibitors
at high cell densities. American Journal of Botany, 82(8): 955-96.

[20] Church, J., Hwang, J.H., Kim, K.T., McLean, R., Oh, Y.K., Nam, B., Joo, J.C. and Lee, W.H. 2017.
Effect of salt type and concentration on the growth and lipid content of Chlorella vulgaris in synthetic
saline wastewater for biofuel production. Bioresource Technology, 243: 147-153.

[21] Kebeish, R., EI-Aouty, Y. and Hussein, A. 2014. Effect of salinity on biochemical traits and
photosynthesis-related gene transcription in Chlorella vugaris. Egyptian Journal of Botany, 54(2):
281-294.

[22] Yun, C.J., Hwang, K.O., Han, S.S. and Ri, H.G. 2019. The effect of salinity stress on the biofuel
production potential of freshwater microalgae Chlorella vulgaris YH703. Biomass and Bioenergy,
127: 1-7.

[23] Bajwa, K. and Bishnoi, N.R. 2015. Osmotic stress induced by salinity for lipid overproduction in
batch culture of Chlorella pyrenoidosa and effect on others physiological as well as physicochemical
attributes. Journal of Algae Biomass Utilization, 6(4): 26- 34.

[24] Sreesai, S. and Pakpain, P. 2007. Nutrient Recycling by Chlorella vulgaris from Septage Effluent of
the Bangkok City, Thailand. Science Asia, 33: 293-299.

[25] Khoo, C.G., Woo, M.H., Yury, N., Lam, M.K. and Lee, K.T. 2017. Dual role of Chlorella vulgaris
wastewater treatment for biodiesel production: growth optimization and nutrient removal study.
Journal of the Japan Institute of Energy, 96(8): 290-299.

[26] Cuellar-Bermudez, S.P., Aleman-Nava, G.S., Chandra, R., Garcia-Perez, J.S., Contreras-Angulo, J.R.,
Markou, G., Muylaert, K., Rittamann, B.E. and Parra-Saldivar, R. 2017. Nutrients utilization and
contaminants removal, A review of two approaches of algae and cyanobacterial in wastewater. Algal
Research, 24: 438-449.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 227 - I 040

Optimization of Washing Conditions and Adsorption Process for
Petroleum Hydrocarbon Removal from Drill Cuttings Byproduct

Theary Peng Orng1,3 Thaksina Poyai1 Nattawin Chawaloesphonsiya1 Saret Bun3 and
Pisut Painmanakul1,2,4*

1Department of Environmental Engineering, Faculty of Engineering,
Chulalongkorn University, Bangkok 10330, Thailand

2Center of Excellence on Hazardous Substance Management (HSM), Bangkok 10330, Thailand
3Water and Environmental Engineering, Faculty of Hydrology and Water Resources Engineering,

Institute of Technology of Cambodia, Phnom Penh 12156, Cambodia
4Research Unit on Technology for Oil Spill and Contamination Management,

Chulalongkorn University, Bangkok 10330, Thailand
*Phone : 02-218-6671, Fax : 02-218-6666, E-mail : [email protected]

ABSTRACT
Drill cuttings contaminated with total petroleum hydrocarbon (TPH) are generated from oil and gas
exploration and production. The treatment of drill cuttings through washing process has been applied due to
its high efficiency and less energy consumption. However, this process generates washing solution
containing TPH as a petroleum waste, which requires further management. Therefore, the purpose of this
study is to optimize the drill cuttings washing process using ethyl lactate (EL) as a green washing agent.
Afterwards, the spent washing solution was purified through an adsorption process using two adsorbents:
coal-based and coconut shell-based granular activated carbon (GAC). The result showed that liquid-to-solid
(L/S) ratio was the most influential factor affecting the removal of TPH from drill cuttings by EL. The higher
volume of EL used, the higher TPH extraction capacity. The mixing speed followed by washing time and
rinse-to-solid (R/S) ratio also significantly affected the TPH removal efficiency, whereas the rinsing time
was statistically insignificant. The optimum washing conditions were L/S ratio of 10 mL/g, washing time of
20 min, mixing speed of 100 rpm, R/S ratio of 10 mL/g, and rinsing time of 1 min, from which the TPH
removal of 87.6% was achieved. For the adsorption experiments, coal-based GAC performed better in
adsorbing TPH from the spent washing solution compared to coconut shell-based GAC. Thus, the overall
results suggest that EL was a promising agent for removing TPH from drill cuttings, and coal-based GAC
could be a potential adsorbent for spent EL purification and recovery.

Keywords: drill cuttings; ethyl lactate; washing process; total petroleum hydrocarbon; activated carbon

INTRODUCTION
In oil and gas exploration and production, drilling is a key process conducted by the aid of drilling mud,
which is a synthetic oil containing several additive chemicals for controlling and stabilizing the borehole.
During drilling operation, a drill bit is used to grind the rocks while drilling mud is pumped down the well
through the hollow drill string and returned carrying the drill cuttings (i.e., crushed rock pieces). Once drill
cuttings are taken to the surface, the drilling mud is then separated through the solid-control equipment for
reusing in drilling process and drill cuttings become wastes. Drill cuttings contaminated with drilling mud
and petroleum hydrocarbon are recognized as hazardous wastes. Generally, a single drilling well produces
more than a thousand tons of drill cuttings [1,2]. Therefore, the proper management and disposal of drill
cuttings are necessarily required. Currently, washing process has been widely applied for remediating
hydrocarbon-contaminated soil due to its simple and rapid operation, less power consumption, and high
removal efficiency [3]. Washing process can be an in-situ or ex-situ remediation process in which an
extracting agent is used to remove the contaminants from soils. Green solvent, ethyl lactate (EL) has drawn
much attention in hydrocarbon-polluted soil washing process due to its high efficiency, biodegradability, and
low toxic properties [4]. Hence, EL is expected to exhibit high removal efficiency for hydrocarbon liberation
from drill cuttings.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 228 -

Despite cleaned drill cuttings as a product, the process remains washing waste (or washing solution) holding
high concentration of solvent and hydrocarbon pollutants. Thus, it is essential to recover and reuse the
washing solution in order to minimize the chemical waste volume as well as the overall operational cost.
Various treatment techniques have been studied on hydrocarbon removal from non-aqueous solutions such as
membrane filtration [5], distillation [6], and solvent extraction process [7]. Nevertheless, several problems
are commonly found from these methods such as high treatment cost, high energy consumption, and
complex operation [8]. Besides these technologies, an adsorption process could be another technique to
remove oil from the spent washing solution [9-11]. Moreover, the process exhibits its easy operation, low
maintenance cost, and small footprint, compared to other treatment processes [12]. Therefore, this work aims
to optimize the washing conditions for total petroleum hydrocarbons (TPH) removal from drill cuttings using
EL as an extracting agent. Subsequently, the washing solution is purified through an adsorption process in
order to determine the optimal adsorbent type for solvent recovery.

METHODOLOGY
Drill Cuttings
Drill cuttings samples were collected from offshore petroleum exploration and production. The samples were
prepared by air-dried for two days at room temperature before screening by a standard sieve #7 (equivalent
to 2.8 mm) to remove large grains as it is hard to react with other constituents. Physicochemical properties of
drill cuttings including heavy metals, chloride, pH, electrical conductivity (EC), cation exchange capacity
(CEC), TPH concentration, bulk density, moisture content, organic matter, and particle size distribution,
were analyzed.

Design of Experiment and Data Analysis
Response Surface Methodology with Central Composite Design (RSM-CCD) was used to design the
experimental conditions and investigate the effects of the washing parameters on the treatment performance.
There are several influenced factors on the TPH removal from drill cuttings washing process including
liquid-to-solid ratio (L/S), washing time, mixing speed, rinse-to-solid ratio (R/S), and rinsing time. The low,
middle and high levels of each factor were coded as -1, 0 and +1, respectively as shown in Table 1.
According to five factors with three levels each, the 32 experimental runs consisting of 16 cube points, 10
axial points, and 6 replicated at the center points, were provided for performing the experiments. The TPH
removal percentage calculated by Equation (1) was a response variable to be optimized.

TPH removal (%) = Co - Ce  100 (1)
Co

Where Co is the initial TPH concentration (mg/kg) and Ce is the residual TPH concentration (mg/kg).

Table 1 Experimental design for washing process

Factors Unit Factor levels +1
-1 0 10.0
L/S ratio mL/g 3.0 6.5 20.0
Washing time min 1.0 10.5 150
Mixing speed rpm 50 100 10.0
R/S ratio mL/g 1.0 5.5 10.0
Rinsing time min 1.0 5.5

Analysis of variance (ANOVA) was also employed to validate the predicted model and evaluate the

statistical effect of each factor. The experimental results were analyzed using Minitab software and fitted

into the empirical second-order polynomial equation as given in Equation (2) in order to optimize the drill
cuttings washing conditions [13]. The model quality is estimated by the correlation coefficient (R2) and the

result analysis is carried out using F-test and p-value with 95% confidence level of statistical analysis.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 229 -

55 45
   y = β0 + βixi + βiixi2 + βijxix j
(2)

i=1 i=1 i=1 j=i+1

Where y is the predicted response by the model, xi and xj are the independent factors, βo is a constant, βi is
linear coefficients, βii is second-order coefficients, and βij is interaction coefficients.

Washing Procedures
Two grams of drill cuttings were added with a pre-determined EL volume, and the washing process was
conducted under room temperature following the conditions obtained from RSM-CCD using a shaker. The
washing solution was drained out after settling down for a certain time. Then, drill cuttings were rinsed with
tap water by shaking at the previous speed of EL washing. Once rinsing water was removed, cleaned drill
cuttings were dried in the oven at 60oC for 6 hours and analyzed for the residual TPH concentration through
solvent extraction and detection by gas chromatography equipped with flame ionization detector (GC-FID,
Agilent 6890N, USA), following the U.S. EPA 8015 method [14].

Adsorbent Materials

To remove TPH from the washing solution, two types of commercial granular activated carbon (GAC), i.e.,

coal-based (0.60-2.36 mm) and coconut shell-based (1.18-2.36 mm) were employed. The GAC samples were
washed four times with deionized water and then dried at 60oC overnight prior to use.

Adsorption Experiment
In this part, the synthetic solutions representing of the washing solution generated from drill cuttings
washing process were prepared by dissolving diesel oil in EL at the calculated amount in order to get three
solution concentrations. The adsorption experiment was conducted in a 150 mL glass vial screw-top holding
50 mL of washing solution and 0.2 g of GAC. The bottles were therefore sealed and agitated in a water bath
shaker under a speed of 150 rpm at 25oC for 24 hours to reach the equilibrium. After adsorption, the
solutions were filtrated and analyzed for the residual TPH concentration using UV-Vis spectrophotometer at
260 nm wavelength.

RESULTS AND DISCUSSIONS

Drill Cuttings Characterization

The characteristics of offshore drill cuttings are displayed in Table 2. The drill cuttings sample contained an

acceptable level of heavy metals except for the arsenic concentration, which was found beyond the limit of
residential and agricultural soil quality (3.9 mg/kg) based on the standard of Thailand’s Pollution Control

Department (PCD). Iron (1.73%) and calcium (0.06%) were two dominant elements found in drill cuttings similar
to general soils. In addition, the CEC of 7.29 cmol/kg was measured and fallen in a range of 6 – 12 cmol/kg,

indicating that the drill cuttings were low reactive in nature [15]. The TPH concentration of approximately

236,000 mg/kg was detected and regarded as the initial TPH concentration used in this study. The particle size

distribution analysis showed that the drill cuttings particles mostly existed as silt, sand, and clay, respectively.
Moreover, a specific surface area up to 389.8 m2/kg was measured from the sample.

Table 2 Physicochemical properties of drill cuttings

Parameters Value Parameters Value
Fe (mg/kg) 17 330 271
Ca (mg/kg) Chloride (mg/L) 8.0
Hg (mg/kg) 596 pH 0.106
Pb (mg/kg) 0.521 EC (mS/cm) 7.29
As (mg/kg) 30.870 CEC (cmol/kg) 236,000
25.550 TPH concentration (mg/kg) 1.24
Cd (mg/kg) Bulk density (g/cm3) 5.0
Cr (mg/kg) 0.183 Moisture content (%) 15.31
Mg (mg/kg) 25.430 Organic matter (%) 12.88
Mn (mg/kg) 4 685 Sand (2000 – 50 µm) (%) 83.45
Ni (mg/kg) 1 254 Silt (50 – 2 µm) (%) 3.67
Zn (mg/kg) 23.430 Clay (< 2 µm) (%)
60.490

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 230 -

Optimization of Washing Conditions
Based on the experimental results, the optimum washing condition was found at L/S of 10 mL/g, washing time of

20 min, mixing speed of 100 rpm, R/S ratio of 10 mL/g, and rinsing time of 1 min, in which the maximum TPH

removal efficiency of 87.6% was achieved. The validity of the model is evaluated by ANOVA and the second-

order polynomial equation of the TPH removal efficiency as a function of L/S ratio (A), washing time (B), mixing

speed (C), R/S ratio (D), and rinsing time (E) are given in Table 3 and Equation (3), respectively. According to the
F-test with 95% confidence level, the ―p-value‖ less than 0.05 designates the parameter statistically significant. As

shown in Table 3, the F-value of 12.91 and the p-value well below 0.05 indicate that the model prediction was
statistically significant. Moreover, the linear terms (A, B, C, D), square term (C2), and 2-way interaction terms

(AB, AD, BC, DE) had significant impact on the washing performance since the p-values were less than 0.05. It

can be observed that L/S ratio had the highest influence on the TPH removal efficiency followed by mixing speed,

washing time, and R/S ratio, respectively. The large p-value of Lack-of-Fit (0.177) also indicates that the
experimental error was not statistically significant. Additionally, the fitness of the model R2 = 0.9591 was

obtained.

%TPH removal = 32.78 + 5.14A  0.426B + 0.565C + 0.65D  1.14E  0.228A2 + 0.0398B2 (3)
 0.003738C2 + 0.0459D2 + 0.1313E2  0.0727AB + 0.00828AC  0.1093AD

+ 0.0676AE + 0.00406BC + 0.0002BD  0.0271BE + 0.00562CD + 0.00157CE

 0.1199DE

Table 3 ANOVA result for the TPH removal percentage

Source DF Adj SS Adj MS F-value p-value

Model 20 1870.06 93.503 12.91 0.000 Significant
5 1098.20 219.640 30.32 0.000
Linear 1 889.72 889.717 122.82 0.000
1 62.533 8.63 0.013
A: L/S ratio 1 62.53 100.017 13.81 0.003
1 100.02 45.379 6.26 0.029
B: Washing time 1 45.38 0.08 0.787
5 0.555 10.39 0.001
C: Mixing speed 1 0.55 75.303 2.66 0.131
1 375.51 19.235 4.39 0.060
D: R/S ratio 1 19.24 31.788 29.67 0.000
1 31.79 214.936 0.29 0.599
E: Rinsing time 1 214.94 2.125 2.40 0.149
10 17.401 5.46 0.005
Square 1 2.12 39.534 12.89 0.004
1 17.40 93.364 4.64 0.054
A2 1 395.34 33.611 6.55 0.027
B2 1 93.36 47.438 2.50 0.142
C2 1 33.61 18.126 8.21 0.015
D2 1 47.44 59.483 0.00 0.992
E2 1 18.13 0.001 2.96 0.114
1 59.48 21.414 3.53 0.087
2-Way Interaction 1 0.00 25.578 0.28 0.610
1 21.41 1.995 13.02 0.004
AB 11 25.58 94.333
6 2.00 7.244 2.41
AC 5 94.33 9.868
31 79.69 4.096
AD 59.21
20.48
AE 1949.74

BC

BD

BE

CD

CE

DE

Error

Lack-of-Fit 0.177 Not significant

Pure Error

Total

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 231 -

The predicted responses (%TPH removal) given by Equation (3) were plotted against the experimental values as
displayed in Figure 1 to evaluate the precision and competency of the model. The high correlation R2 = 0.9574
indicated a great goodness-of-fit of the model to the experimental data. Additionally, the main effects plot of each
factor and the surface plot of interactive effects were illustrated in Figures 2 and 3, respectively. The results
exhibited that L/S ratio was the main influential factor on the TPH removal efficiency as confirmed by the least p-
value shown in Table 3. The increase of L/S ratio from 3 to 10 had a positive effect on the washing performance,
indicating that the amount of TPH removed was directly influenced by the volume of EL added. This can be
explained that the higher L/S ratio, the greater capacity of EL to mobilize the TPH molecules existing in drill
cuttings.

Figure 1 Experimental values against predicted values of the TPH removal efficiency

Despite the high impact of L/S ratio, the effects of mixing speed on TPH removal could not be neglected. An
effective washing performance could be reached at a velocity that the surface area of drill cuttings was completely
contacted with EL. It can be observed in Figure 2 that the TPH removal efficiency has a tendency to increase with
increasing mixing speed until it raised to 100 rpm. However, when the speed continued to increase more, the TPH
removal was dropped significantly. This result indicates that a mixing speed of 100 rpm was adequate to obtain
the maximum TPH extraction potential. For the washing time, it corresponds to the degree of mass transfer rate

between TPH and EL. The proceMss caoiunldEbfefeenchtanscPedlobyt pfroolronTgPinHg wRasehmingotivmaelto%20 min.

Fitted Means

L/S ratio Washing time (min) Mixing speed (rpm) R/S ratio Rinsing time (min)
85
Mean of TPH Removal %
80

75

70

3.0 6.5 10.0 1 10 20 50 100 150 1 5 10 1 5 10

Factor level of each investigated factor

Figure 2 Main effects plot of each factor

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 232 -

Another finding to point out is the effect of R/S ratio, rinsing drill cuttings with water is required in order to
remove the remaining TPH and EL solution adhered to drill cuttings. Since EL contains both polar and nonpolar
properties, it is completely miscible in hydrocarbons and water [16]. Therefore, after water rinsing, the cleaned
drill cuttings would become more purified. Based on the main effects plot, the TPH removal was improved as
increasing R/S ratio from 1 to 10 mL/g. Hence, the higher R/S ratio resulted in a larger volume of water used for
liberating the residual solution from drill cuttings. Nevertheless, the rinsing time was found statistically
insignificant as the p-value was higher than 0.05 (Table 3). A prolonged rinsing time normally exhibits better
exposure between drill cuttings and washing solution. However, in this experimental result, an extended rinsing
time from 1 to 10 min did not noticeably improve the TPH removal. This is probably due to the emulsified
mixture that formed from the co-exist of TPH, EL, and water. The longer rinsing time might produce more
emulsions which were difficult to remove from drill cuttings, and thus resulting in less TPH removal efficiency.

Figure 3 Surface plots of (a) L/S ratio and washing time, (b) L/S ratio and mixing speed, (c) washing
time and mixing speed, and (d) R/S ratio and rinsing time against %TPH removal

Characterization of Adsorbents

The chemical element composition of coal-based GAC and coconut shell-based GAC were analyzed through

energy dispersive x-ray spectrometer (EDS) as shown in Table 4. Both coal-based and coconut shell-based GAC

contained carbon and oxygen as the main compositions. According to the BET analysis, the specific surface area
of 850 m2/g and 1000-1130 m2/g were measured from the coconut shell-based and coal-based GAC, respectively.

Table 4 EDS analysis of coal-based and coconut shell-based GAC

GAC Element analysis (wt%)

Coal C O Ca K Other
Coconut shell 84.61 1.77
90.13 12.19 1.43 -
-
6.47 0.82 2.58

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 233 -

Adsorption of TPH from Washing Solution
Considering on TPH adsorption performance, Figure 4 clearly demonstrates that the TPH removal
efficiencies of coal-based and coconut shell-based GAC decreased from 33.3% to 13.7%, and from 20.2% to
6.1%, respectively, as increasing TPH concentration from 1600 mg/L to 6300 mg/L. Moreover, the amount
of TPH adsorbed with coal-based GAC was higher than that of coconut shell-based for all concentrations. As
presented in the adsorbent characteristics, the coal-based GAC holding larger specific surface area which
could enhance the interaction between TPH molecules and adsorbent. Therefore, at the TPH concentration of
1600 mg/L, the coal-based GAC could remove TPH from the washing solution about 33.3%, whereas only
20.2% of TPH was removed by coconut shell-based GAC. The visual appearance of the washing solution
before and after adsorption experiment by coconut shell-based and coal-based GAC is also presented in
Figure 5. The washing solution changed from yellow to clear solution since the TPH in solution had been
removed by GAC. Additionally, the solution adsorbed by coal-based GAC was much clearer, indicating the
higher TPH removal than that of coconut shell-based GAC.

Figure 4 TPH removal efficiency at the different concentrations for both GAC

Figure 5 Appearance of washing solution before and after adsorption with GAC
CONCLUSION
The treatment of drill cuttings using EL washing process and subsequent adsorption of TPH from washing
solution was investigated. The result indicated that drill cuttings washing performance was largely
influenced by L/S ratio followed by mixing speed, washing time, and R/S ratio. The maximum TPH removal
efficiency of 87.6% was achieved at the optimum conditions: L/S ratio of 10 mL/g, washing time of 20 min,
mixing speed of 100 rpm, R/S ratio of 10 mL/g, and rinsing time of 1 min. Moreover, the coal-based GAC
exhibited better performance for removing TPH from the spent washing solution compared to coconut shell-
based GAC for all solution concentrations. It was due to the coal-based GAC containing higher specific
surface area than that of coconut shell-based GAC. Therefore, coal-based GAC is recommended as an

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 234 -

alternative material to recovery the spent EL generated from drill cuttings washing to reduce the chemical
waste volume as well as the treatment cost.

ACKNOWLEDGEMENT
This work was supported by the Department of Environmental Engineering, Faculty of Engineering,
Chulalongkorn University for the research grant and laboratory facilities. Additionally, the authors are
grateful to the Research Unit on Technology for Oil Spill and Contamination Management, Chulalongkorn
University and AUN/SEED-Net of JICA for the research fund.

REFERENCES
[1] Huang, X., Jiang, G. and Deng, Z. 2015. Oil Extraction from Oil-Contaminated Drill Cuttings Using a

Recyclable Single-Phase O/W Microemulsion. Tenside Surfactants Detergents. 52(6): 454-463.
[2] Huang, Z., Xu, Z., Quan, Y., Jia, H., Li, J., Li, Q., Chen, Z. and Pu, K. 2018, July. A review of

treatment methods for oil-based drill cuttings. In IOP Conference Series: Earth and Environmental
Science. 170 (2): 22-74. IOP Publishing.
[3] Befkadu, A.A. and Quanyuan, C.H.E.N. 2018. Surfactant-enhanced soil washing for removal of
petroleum hydrocarbons from contaminated soils: a review. Pedosphere. 28(3): 383-410.
[4] Yap, C.L., Gan, S. and Ng, H.K. 2012. Evaluation of solubility of polycyclic aromatic hydrocarbons in
ethyl lactate/water versus ethanol/water mixtures for contaminated soil remediation applications.
Journal of Environmental Sciences. 24(6): 1064-1075.
[5] Firman, L. R., Ochoa, N. A., Marchese, J. and Pagliero, C. L. 2013. Deacidification and solvent
recovery of soybean oil by nanofiltration membranes. Journal of membrane science. 431: 187-196.
[6] Rundquist, E. M., Pink, C. J. and Livingston, A. G. 2012. Organic solvent nanofiltration: a potential
alternative to distillation for solvent recovery from crystallization mother liquors. Green Chemistry.
14(8): 2197-2205.
[7] Berk, Z. 2013. Chapter 11—Extraction. Food process engineering and technology, 287-309.
[8] Liu, J., Chen, J., Jiang, L. and Yin, X. 2014. Adsorption of mixed polycyclic aromatic hydrocarbons in
surfactant solutions by activated carbon. Journal of Industrial and Engineering Chemistry. 20(2): 616-
623.
[9] Ahn, C. K., Kim, Y. M., Woo, S. H. and Park, J. M. 2007. Selective adsorption of phenanthrene
dissolved in surfactant solution using activated carbon. Chemosphere. 69(11): 1681-1688.
[10] Dowaidar, A. M., El‐Shahawi, M. S. and Ashour, I. 2007. Adsorption of Polycyclic Aromatic
Hydrocarbons onto Activated Carbon from Non‐Aqueous Media: 1. The Influence of the Organic
Solvent Polarity. Separation Science and Technology. 42(16): 3609-3622.
[11] Liu, J., Chen, J., Jiang, L. and Yin, X. 2014. Adsorption of mixed polycyclic aromatic hydrocarbons in
surfactant solutions by activated carbon. Journal of Industrial and Engineering Chemistry. 20(2): 616-
623.
[12] Rosas, J.M., Santos, A. and Romero, A. 2013. Soil-washing effluent treatment by selective adsorption
of toxic organic contaminants on activated carbon. Water, Air, and Soil Pollution. 224(5):1506.
[13] Montgomery, D. C. 2017. Design and analysis of experiments. John wiley & sons.
[14] U.S. EPA 2010. Guidance for Assessing Petroleum Hydrocarbons in Soil. Report No. DERR-00-DI-
033.
[15] Babajide, A. A., Adebowale, O., Fadairo, A. S. A., Ako, C. T. and Ifechukwu, M. 2016. Effects of
Temperature and Pressure on Shale Cuttings Dispersion in Water Base Mud WBM Using NACL,
CACL2, KCL Salts as Primary Inhibiting Agents and Polymer XCD Xanthan Gum as secondary
Inhibiting Agent. SPE-184310-MS, 1-15.
[16] Kua, Y. L., Gan, S., Morris, A. and Ng, H. K. 2016. Ethyl lactate as a potential green solvent to extract
hydrophilic (polar) and lipophilic (non-polar) phytonutrients simultaneously from fruit and vegetable
by-products. Sustainable Chemistry and Pharmacy. 4: 21-31.

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020

- 235 - I 041

Potential of Pollutant Transport from Petroleum Platform Area in the
Gulf of Thailand for Monitoring Plan

Viranya Kittivarakul1 and Pichet Chaiwiwatworakul2*

1Master student ; 2*Assistant Professor, Department of Environmental Engineering, Faculty of Engineering,
King Mongkut's University of Technology Thonburi, Bangkok 10140, Thailand.
*Phone : +66-(02)4709163, Fax :- , E-mail : [email protected]

ABSTRACT
The Gulf of Thailand is the major natural gas production area in Thailand. As common as other offshore oil
field around the world, petroleum activities can reintroduce a soluble pollutant back to the aquatic
environment. As a result, the soluble pollutant contaminated may transport from the petroleum source due to
wave and stream which are variable by seasons and can be contaminating the benthic fish and enter to human
via the food chain which may affect to the human health. The computer model was used to simulate the
transport of pollutant due to the complexation of the sea current behaviors. The water level and wind data
from year 2009 to 2015 were selected for model calibration and validation and Chezy coefficient was an
adjustment parameter. Then, the sediment transports were simulated using averaged annual hydrodynamic
condition. The consideration was done by separating the condition into four periods according to water
current behaviors. Mean absolute error (MAE), coefficient of determination (R^2), and Nash-Sutcliffe model
efficiency coefficient (NSE) were calculated to estimate the model accuracy by comparing the simulated
water level with the measured water level at Laem Ngop station, Mae Klong station, and Sichon station. The
soluble pollutant concentration was assumed to be equal to 0.1 mg/l based on mercury, which was the most
concerned heavy metal type. The result showed that the farthest soluble pollutant distant was 52.5 km to the
northwest direction during the second-inter monsoon. The plume concentration near the source area region
was lower than 0.000025 mg/l and reduced to around 0.00001 mg/l - 0.000015 mg/l. After that, the plume
concentration dropped down to 0.00001 and continued to reduce until the concentration was undetectable.
Moreover, the soluble pollutant transport was affected by the seasonal currents.

Keywords : Pollutant transport simulation; Soluble pollutant transport

INTRODUCTION
The Gulf of Thailand has many valuable natural resources, such as fishery resource and petroleum. Since
1968, petroleum sources were surveyed, and petroleum platforms were installed to produced petroleum. In
present, there are 10 petroleum sources in Gulf of Thailand [1]. As common as other offshore oil field in the
world, the petroleum collection process may cause the contamination of soluble pollutant in the aquatic
environment. The most common pollutants were cadmium (Cd), chromium (Cr), copper (Cu), mercury (Hg),
manganese (Mn), nickel (Ni), lead (Pb), and zinc (Zn) [2]. The soluble pollutant might be transport due to
the wave and stream that are variable by seasons [3] and contaminated benthic fish and enter to human via
the food chain which may affect to the human health [4],[5]. The Gulf of Thailand is the shallow gulf which
the average depth is around 58 m. and the maximum depth is 85 m which is at middle of lower part of the
gulf [6]. Moreover, the main sediment type in the Gulf of Thailand was sand which the silt type mostly
covers the center area of the gulf. In addition, the major constituent of the sediment particle was quartz , and
the highest amount of surface bottom sediment was about 5 m [7]. According to the studies of hydrodynamic
condition in the Gulf of Thailand, hydrodynamic conditions were divided in to 4 conditions which were
varied under northeast monsoon, first-inter monsoon, southwest monsoon, and second-inter monsoon. Under
the northeast monsoon, the mean sea currents entered from the South China Sea and exited through the east
of the lower part of the gulf while the mean sea currents flowed to the opposite direction under the southwest
monsoon. Under the first, and second inter-monsoon, the mean sea currents entered to the gulf from the
mouth of the gulf. However, the sea water flow exited through different direction which through the South
China Sea under the first inter-monsoon, and through the west of the lower part of the gulf under the second
inter-monsoon as shown in figure 1 [8], [9], [10], [11], [12].

9th International Conference on Environmental Engineering, Science and Management
The Heritage Chiang Rai, Thailand, May 27-29, 2020