Development of odorous gas model Using Municipal solid Waste Emission

Mohd. Nahar bin Othman1, Muhd Noor Muhd Yunus1 and Ku Halim Ku Hamid2.

Malaysian Nuclear Agency ,Bangi, 43000, Kajang, Malaysia

2Universiti Tecknologi MARA, Shah Alam, Malaysia

E-mail: nahar@nucIearmalaysia.gov,my

ABSTRACT: The impact of ambient odour in the vicinity of the Semenyih MSW processing plant, commonly known as RDF

plant, can be very negative to the nearby population, causing public restlessness and consequently affecting the business

operation and sustainability of the plant. The precise source of the odour, types, emission level and the meteorological conditions

are needed to predict and establish the ambient.odour level at the perimeter fence of the plant and address it with respect to the

ambient standards. To develop the odour gas model for the purpose of treatment is very compulsory because MSW odour itself

consist of many chemical components that contribute the smell. Upon modelling using an established package as well as site

measurements, the; odour level at the perimeter fence of the plant was deduced and found to be marginally high, above the normal

ambient level. Based on this issue, a study was made to model odour using Ausplume Model. This paper will address and discuss

the measurement of ambient gas odour, the dispersion modelling to establish the critical ambient emission level, as well as

experimental validation using a simulated odour. The focus will be made on exploring the use of Ausplume modelling to

develop the pattern of odour concentrations for various condition and times, as well as adapting the model for MSW odour

controls.

Abstrak

Kesan kepada udara ambien disekiran kilang MSW Semenyih, yang disebut sebagai kilang RDF menjadi sangat negatif kepada

pendudok tempatan, yang tidak menyenangkan, menyebabkan keresahan kepada persekitaran kilang.Punca kepada bau busuk,

jenis, kuantiti pelepasan dan keadaan meteorologikal perlu diramai dan diterjemahkkan kepada keadaan ambien bau sebenar pada

tahap di pagar kilang dan dipamerkan pada standard yang dibenarkan.Untuk membangunkan model gas berbau bagi tujuan

pengolahan menjadi keperluan penting disebabkan bau MSW itu sendiri mempunyai banyak komponen kimia yang memberi

sumbangan kepada bau busuk.Dengan menggunakan. model yang sudah ada, termasuk data persekitaran, tahap bau busuk di

pagar kilang sangat ketara dan didapati sangat tinggj, iebih dari tahap ambien sebenar.Berdasarkan kepada masalah ini, kajian

telah dijalankan untuk membuat model bau persekitaran menggunakan kaedah Ausplume.Kertas kerja ini akan membincangkan

teknik pengukuran bau ambien, mode! persekitaran bau untuk menentukan tahap kritikal pelepasan gas ambien,disamping ujian

validasi dan stimulasi bau.Fokus akan ditumpukan kepada penggunaan Ausplume model untuk membina paten bau mengikut

keadaan dan masa beserta model kawalan bau MSW

iiwniiifflniiiiiiiwiiiiiiiMY1204088

Keywords: Refuse Derived Fuel, odour, Emissions, ambient and Ausplume modelling

1 .INTRODU CTION

Odours can rose from several sources and most of these sources are man made. Unscientific landfill

design, increased sewage load and improper sewage treatment can produce unpleasant odour. Large

livestock operations, poultry farms, pig farms, tanneries, slaughter houses, food and meat processing

industries and bone mills are among major contributors to odour pollution. Agricultural activities like

decaying of vegetation, production and application of compost etc. also contribute towards odour

pollution. Vehicular sector also has its share in odour pollution. Rapidly growing vehicular and pungent

odours have marked effects on pedestrians a well as nearby residents.

Most commonly reported odour producing compounds are hydrogen sulphide (rotten egg odour) and

ammonia (sharp pungent odour). Carbon disulfide, mercaptans, product of decomposition of proteins

(especially of animal origin), phenols and some petroleum hydrocarbons such as dimethyl sulfide and

trimethylamine are other common odorants. Most offensive odour is created by the anaerobic decay of

wet organic matter such as flesh, manure, feed or silage. Warm temperature enhances anaerobic decay

and foul odour production. According to emission types, sources of odour can be classified as follows:

Point sources, area sources and volume sources -

i) Point sources are either a stack or a ventilation channel exhausting the odorous gases with a

known flow rate, such as a discharge stack from a slaughter house or a ventilation channel

from a bone mill.

ii) Area sources are sources where the odorous gases are emitted from a wide surface. It can be

water or solid surface, such as the water surface of a slurry storage tank, solid waste landfill,

composting or a cattle feedlot.

iii) Volume sources are sources that have a number of openings where the odorous gases escape to

the atmosphere, such as chicken and pig sheds.

MSW, if not managed properly, will be a major contributor towards odour problem. RDF plant which

is recycling solid waste had built a plant near a residential area and thus have been receiving

complaints from the residents there. The purpose of this study is to use RDF to model the odour

dispersion and try to solve the odour problem.

2. M ODELLING AND M EASUREM ENT OF ODOUR LEVEL AT RDF PLANT

In order to get more information about the surrounding area there are a few things needed to be collected

such as land use and meteorological data including wind pattern and so on,

2 .1 L a n d Use w ith in 5 -1 0 km radius

Agricultural activities are still the main activities carried out within 5 km radius of RDF plant. The land is

already gazetted for MSW land fill for MPKj in the master plan MPKj map before. Meanwhile some

new factories have already begun operating within this area. The land use within 5-10 km radius from the

RDF factory have not experienced any major changes in the north eastern part due to the extensiveness of

agriculture nursery activity and forest reserves. But the area southwest of the factory is experiencing

extensive housing and commercial development based on the planned development of the local

authorities. These areas predominantly were ex-mining land and agricultural land, in particular rubber

plantations, orchards and villages. As shown by the satellites picture in Figure 2.1

Figure 2.1: Birds eye view of Factory RDF Sdn Bhd., in red colour and Kampong Sungai Lalang area.

RDF factory, about 8 hectare, is one of the factories located in the industrial area around Sungai Lalang,

Semenyih near Kajang, District. It is near the border of Selangor and Negeri Sembilan. Near the border of

the factory there is a National Service Camp. In the north eastern terrain, there are some orang asli

villages with a small population of about 50 residents. In the south western part of the factory there is

Kampong Sungai Lalang and Kampong Pasir with quite a large number of inhabitants of about 400

residents. In Sungai Lalang industrial area only Recycle Energy Sdn. Bhd, does the processing and

recycling municipal Solid Waste and is characterized with their potential emission of odorous air

pollutants.

Meteorological data from Mac to September of the year 2007 from KLIA meteorological station getting

from Meteorological Department of Malaysia, was incorporated in this study including wind-rose

patterns, daily temperature, daily humidity and atmospheric stability classes. The wind speed, wind

direction and other meteorological parameters were used in determining the most likely direction of

dispersion of the odour from the factory as shown in Figure 2.2, The stability classes were used to

determine the fate of odour gases emitted from the stack and to accomplish the odour concentration

downwind direction.

1.

NORTH

15%

WEST

SOUTH

Figure 2.2: Stability Classes of Years 2007 (Mac-October) Hourly Data from KLIA

Meteorological Station

2.2 sampling o f odour

All involved sampling parts have to be made out o f olfactory neutral materials. Principally every sampling has to

meet the logical requirements, has to be defined, standardized, meaningful and reproducible. This is needed to make

different measurements comparable. Odorant concentrations scaled in OU/m3 are very convincing while comparing

different emissions o f different plants [5]. Because of this instead of comparing different concentrations, different

emission mass currents o f the emitted freight .are compared. The schematic diagram o f the sampling device is shown

in Figure 2.3

O d o u r s a m p l e i n t a k e

V i e w i n g W i n d o w

Figure 2.3: Odour sampling system

2.3 Data collection from RDF plan t

The data using for this modelling was getting from sampling taken around RDF factory during the survey in 2007 as

shown in Table 2.1.

T able 2.1 Data collected around RDF plant

Date Time Lom

bong

Tan

Kim

Tin

contr

ol)

Kg.

Ulu

Kaca

u

contr

ol

Kg.

Sg.Peni

ngcont

rol

Load)

nKarea

Stack

Drier

Outlet

RDF

borde

r

Kg.

Pasir

SgI.alaag

Lorry

entran

ce

inside

Lorry

entrance

outside

Sinipa

ng 3

lampu

Trafik

Remark

5.3.07 morning 184 184 184 1046 1046 515 368 368 368 MSW

2.4,07 morning 184 1046 1533 663 184 721 368 184 MSW

7.5.07 morning 184 2226 3156 184 184 1518 184 MSW

4.6.07 evening 560 3156 3156 1518 748 1046 1046 Rain

11*7.07 night 368 1046 1046 368 368 1046 368 Rain

11.9.07 down 91 515 2559 129 91 120 129 64 MSW

Results of odour test around RDF factory in OU/m J from Mach to September 2007

2 .4 D a ta A n a ly s in g

Due to the fact that neutral air and mixture are offered within two consecutive breathing periods, the panel

members has the possibility of direct comparison of gas samples. Therefore, with this data analysing

technique, the reliability of the result is increased considerably. The total measurement program is

computer controlled and run automatically with software installed in the computer. The software include

several measurement sequences, recovery breaks for panel members and also includes clean air flushing

function for dilution unit as shown in Figure 2.4,

("low i..It • finl/minjMo I E R 2 ; R 3 : R 4 R 6 R 7 1 n a | R9 R IO I M ean 1 "\4....... |' ' |' ••■■I i D ate

.....

r m34 152 4 : 152.42 152 42 152 44 152 44 152 39 1 52 43 152 49 152.49 152.47 152 44! O.D%: 0: o 12 '0 9 /0 6 00 o o i r ? :

35 151 87 151.85 151.9 151.9 151.9 151.93 151.95 151 3 5 ; 151.99 1 "51.95; ' 151 .92; o.o%: 6; u 12 ‘09/QB 00 o b i 'P38 1 51 91 151.87 151.91 151 94 151 98 152 05 152 06 152.07: 152.05 152.09^ 1 5 T 9 9 ; 0.1%: o; 0 12 '‘0 9 /0 6 00 00: W :' \37 154 29 154.34 154.4 154.45 154 46 154 44 154.44 154,43 : 154.42 154 4 154.41 0.1%: 0; ' G 12 ‘'0 9 /0 6 00 00: W \

J r f L 152 r.r; 152.84 152 G3 152.84 152 e? 1 52.7 152.7 152.7 : 152.7 152.73! 152.88 0.0%: o j Cl 12 '0 9 /0 8 00 □o: m

39 154 19 154.26 154.27; 154.25 154 24 154.2 154.18 154.2 ; 154.19 154.18; 154.22 o.o%i o; 0 12 '‘0 9 /0 8 00 00 T ' ;40 154 26 154.31 154 37- 154.38 154 35 154 33 1 5 4 33 154.34 154.32 154.28 " 1 5 * 3 9 1 o .d %; o f □ 12 '■09/06 00 oo; i T 5

........ '........... ...........lY.:. ........................................................ V.»lvi! | n . •..1 . ..<■• ......................... :..................: :::

, »:■> y i o i a l H o - i - G I . : —1. . .1 " f ~ C. . S i d i ’ l n n n o i L j . ; i . [ “ ' v i o o p ' ' ' ' r

E F R I M F R .... [ A D R ■ ' ] i I t t * | | | a . r d s S an-.o ia 1 | S arn p ft ? ' a i r t ..... .... .....................2500 : 2478; 8.1 ! o .o% ; r o % j < : 3 “ r rn ~ 2 5 3 00 0 0 2 70 00 0 9 200001250: 1 2 2 7 ; 16.3! o .i % ; £1,1 % j ; { 4 5C03D i ybui'uu 3 02 00 0 9 20 00 0

625! 609: 32 .9 : 0 .1% ; 2 .3 % ! : M . \ - h s n n . i 1290C 00 3 13 00 0 92000031 3; 301 ; EG.5 : 0 .0% ; 4 2 % | 5ZCDD 1 000C0G 3 20 00 0 9 20000156: 153 ; 131.0: o . i% ! 3.1 % | Z 5 00 00 800000 3 30 00 0 9 20000

78 1 77.7; 257.5; o . i% ; 1 S % | M s...... 2 00 00 0 6 85000 3 30 00 0 9 2000039.1 39.2 : 510.4; 0 .2% ; 2 .7 % ! : 3 2*100000 20000 3 30 00 0 9 20000

Figure 2.4: Computer program for data analysing

3. RESULTS AND DISCUSSIONS

3.1 Odour dispersion M easurement and Modelling Result

Field inspection were accomplished within the assessment area inside and around 5 km of RDF factory

for 9 months with 90 samplings measurement for odour measurement using olfactometer at 10 sampling

points. The schedule in figure 3.1, 3.2, 3.3 and 3.4 were the odour map that had been plotted using Plume

Dispersion model for the assessment area covering the months between March to November 2007. In

addition to the measurement mentioned before the border area surrounding the RDF factory, the odour is

quit disturbing. About 46.2 % of the time odour was sensed at that particular points. On the other hand,

from the Kampong Sungai Lalang area, the odour is felt intensely by the villagers during certain time and

conditions such as after rain, during early morning and at evening or night, it became worst when the

winds blows towards the village area. Many odour complaints came from village communities living near

the factory. As a matter of fact, the measurement results show the same thing, with the movement of air

together with odorous substances from the RDF factory the result confirm the unpleasant situation in the

area.

According to the Odour Dispersion Modelling at figure 3.1, 3.2, 3.3 and 3.4 the impact of odour was

observed not only in the factory area but also in the village area, where the percentage of odour presence

ranging from 6 to 40 %. The ambient odour characteristic implies that the factory should develop an

odour strategy. The odour data collected during the study, were used in Odour Ausplume Modelling.

During this process, coordinate numbers were given to each sampling points and the percentages of for

these coordinates, the contoured odour maps was plotted. The outcome of the odour maps can be seen in

figure 3.1, 3.2, 3.3 and 3.4 It was observed that, the factory odour was sensed 65 % of the time. In the

boundary of the factory area, the concentration decreased to 40 to 50 %. Considering the distance of 1.5

kilo meter away from the factory, odour concentration reduced to 20 % is good enough. The residents

who are living between 1.5 km are highly effected from odour. In order to overcome any future problem,

a setback distance, that is, a non-residential area is suggested in this study. In this case, a setback distance

with a radius of at least 1.5 km is ideal. The centre of the circle is the emission source. Radius of this

circle can be enlarge based on the capacity of the factory and this circlc can be moved towards the

opposite side of the dominant wind directions so as not to annoy the villagers.

The dispersion model of odour from RDF factory to the surrounding area can be seen from figure 3.1, 3.2,

3.3 and 3.4. Looking at the odour map, the dominant wind direction is obviously from the south east. On

the other hand, there also exists more than one parameter that affects the dispersion of odour. Wind

velocity, odour intensity, temperature and humidity are other parameters that effect odour dispersion other

than wind direction. This is the reason why the odour is not dispersing in one direction with respect to

emission source.

On the other hand, the compiled solid waste from MPK dumped around the factory is another contributor

to area source of the odour problem. The odour released from that waste was no different in

characteristics with that released from the RDF factory. Therefore, the odour map obtained is not only for

points source but also for an area source that is for the worst case scenario. As mentioned before, odour

problem around RDF factory is an every day problem because the waste is transported to the factory

daily. Because of the MPK waste disposal problem, the problem became more acute. At the end of the

sampling time most of the MPK waste had been carried out from the factory area. The odour

concentration had been reduce to 10 to 20 % . The ambient odour characteristic implied that the factory

was not a severe odour source thus, the factory administration must take into consideration the odour

abatement technology.

Figure 3.1: The odour dispersion modelling during morning hours from April to September 2007

Figure 3.2: The odour dispersion modelling during evening hour from April to September 2007

Legend,OU/nr

Figure 3.3: The odour dispersion modelling during night hours from April to September 2007

Legend,OU/nr1

Figure 3.4: The odour dispersion modelling during afternoon hours from April to September 2007

4. CONCLUSIONS AND RECONM ANDATIONS

The validation result of the Ausplume model based on the comparison with the experimental data showed

that despite of relative simplicity of adaptation procedure, the model provides reasonable result

applicable in the practise. However, in comparison with German Immission model which is considered as

a simple one, Ausplume model seems to better one. After calibrating the emission data of Ausplume

model, the expected values and the variances of the certain parameters the two odour emissions are

compatible with the empirical data.

Finally, an approach has been suggested for determining odour nuisance based on our current

understanding of atmospheric or odour dispersion modelling and taking into account of the uncertainties

inherent in this type of study. Tt is suggested that Ausplume model can be used for odour dispersion

beside air dispersion modelling.

REFERECES.

[IJAnshumali, B.C. W inkleman, and Sheth, A.C. (1997). Destruction o f Low Levels o f Volatiles Organic

Compounds in D rt Air Streams by an Electron Beam Generated Plasma, acceted for publication in the

Journal of Air & W aste M anagement Association.

[2]Chmielewski, A.G., Sun, Y,, Bulka, S., & Zimek, Z. (2007). "Review on gaseous chlorinated organic pollutants

electron beam treatment”, Radiat. Phys. Chem. 76(11-12), 1795-1801.

[3]Getoff, N. (1998). “Radiation chemistry and the environment”, IAEA-TEDOC-1023, “Radiation Technology for

Conservation of the Environment”, 121-131, Vienna, Austria.

[4]Hakoda, T., Sako, T., Shimada, A., Ishida, T., & Kojima, T. (2006). "Component analysis o f particulate products

in electron beam-irradiated xylene/air mixtures using an atmospheric pressure ionization mass

spectrometer1' Bull. Chem. Soc. Jpn. 79, 731-737.

[5]Jiang, J. (2008). Review for sampling technique, EnvironOdour A ustralia Pty LTD, Sydney, Australia.

[6]Tokunaga, O., Osada, Y. (1995). Pilot-scale test on electron beam treatment o f municipal solid waste flue gas

with spraying slaked-lime slurry. Radiat. Phys. Chem. 45 (6), 1021-1027.

[7]W iederkehr, P. (1994). Emission reduction reduction for VOC in some ORCD countries. In characterization and

control of odours and VOC in the process industries. Proceeding o f the 2nd International symposium on

characterization and control o f odours and VOC in the process industries. Louvain=la-Neuve, Belgium, 3-5

November 1993. Amsterdam: Elsevier SceinceB .V , 29.

[8]Urashima. K., and Chang J.S., (2000). Removal of Volatile Compounds from Air Streams and Industrial Flue Gas

by Non-Thermal Plasma Technology, IEEE Transactions on Dielectrics and Electrical Insulation, vol. 7, no.

5, pp. 6023-614.

[9]Zona, R., Solar, S., Getoff, N., Sehested, K., and Holcman, J. (2008). Reactivity o f OH radicals with

chlorobenzoic acids A pulse radiolysis and steady-state radiolysis study. Radiation physics and chemistry,

162-168.