Polycyclic Aromatic Hydrocarbons (PAHs) in the Haze from Forest Fires in Indonesia 1997
Polycyclic Aromatic Hydrocarbons (PAHs) in the Haze
from Forest Fires in Indonesia 1997
Angelika Heil, 1998
Draft , 19 June 1998
Large-scale forest fires in Indonesia 1997 have caused transboundary air pollution of a dimension never experienced before. Large amounts of particulate matter emitted by the fires accumulated in the atmosphere and were transported to neighboring countries.
Particulate pollution levels locally exceeded 3000 µg/m3 (PM10). According to the US-Environmental Agency, particulate concentrations above 150 µg/m3 (PM10) are judged as unhealthful and above 420 µg/m3 as hazardous.
Health effects do not only depend on the particulate as such, but also on the composition of toxic compounds adsorbed on their surface. In this context, carcinogenic compounds such as Polycyclic Aromatic Hydrocarbons (PAHs) which are created while combustion processes are of major concern.
Polycyclic Aromatic Hydrocarbons (PAHs): Properties and Sources
Polycyclic Aromatic Hydrocarbons (PAHs) are formed during combustion processes of organic material with insufficient oxygen supply (incomplete combustion).
PAHs comprise of more than 100 different multi-ringed compounds of which many are known to be carcinogenic.
Due to their low vapor pressure, most PAHs immediately condense and adsorb onto particulate or form very small particles themselves (average diameter 1µm).
Consequently, PAHs in the ambient air predominate in the particulate-bound phase /5/ and especially in particles smaller 10 µm diameter (PM10) /9/.
Studies performed during the last decades showed that PAHs have a significant variation in their composition for different combustion sources. The characteristic spectrum of PAH can be used as fingerprint to identify the sources /4/.
Benzo(a)pyren [BaP], a PAH known for its carcinogenic properties, occur in most emissions and is commonly used as indicator for PAHs.
In urban areas, motor vehicles are the major source of PAHs in the atmosphere, followed by residential wood/(char)-coal combustion and industrial production processes. In some areas, forest fires represent a significant source of PAHs.
Atmospheric PAHs are partly degraded by photo-oxidation. The decay process is accelerated in tropical conditions with high solar intensity and humidity /.
Background Level of PAHs and Benzo(a)pyren [BaP] in the Ambient AIr
Background levels of PAHs in the ambient air in the USA are reported to be 20 1200 ng/m3 in rural areas and 150 to 19.300 ng/m3 in urban areas /11/.
In remote areas, the background concentration of Benzo(a)pyren [BaP] (indicator parameter for PAH) in the ambient air is below 2 ng/m3 (annual average). In medium polluted areas in Germany, the annual average BaP-level is between 2 and 5 ng/m3 and in dense populated areas around 5-12 ng/m3. Locally, at locations close to traffic, coal heating or industrial areas, the daily average BaP-concentration are between 4 – 69 ng/m3 (max. 300 ng/m3) /7/
During the smog disaster in Berlin 1982, the maximal BaP-concentration was 86 ng/m3 (daily average) at a Total Particulate Level of 702 µg/m3 /12/.
The most studied representative of PAH is Benzo(a)pyren (BaP), who is known for its potential to cause cancer. Animal experiments have proven that even very low BaP concentrations increase the tumor formation rate and that there is no No-effect-Concentration. (Intratracheal application of a total dose of 350 mg/kg BaP, distributed over 18 months, induced malign lung tumors at 90 % of the treated rats /1/. The risk increases with the concentration and the duration exposed to the carcinogenic compound.
The unit-risk- value for BaP, which describes the estimates risk of cancer after a 70 year-exposition to a concentration of 1µg/m3, is 7*10-2 [1/(µg/m3)] /3/.
The effective risk is calculated out of the product of the unit-risk-value and the annual average concentration. Exposure to a BaP-concentration of 1,3 ng/m3 (annual average, exposure over 70 years) stands for an effective risk of 1:10.000.
[Statistically one person out of a population of 10.000 will get cancer due to the BaP-exposure ].
PAHs are traditionally adsorbed on very fine particulate (diameter less than 10 µm), which are not restrained in the upper respiratory tract and consequently penetrate deep into the lungs (App.1). In the lungs, the adsorbed PAH may pass into the blood stream or may react locally with the lung tissue and cause lung tumors.
A threshold of 5 ng/m3 BaP (annual average), the indicator parameter for PAH, is in discussion as an EU-wide directive. In Germany, several federal states adopted a target-BaP threshold of 1,3 ng/m3 as the annual average. This threshold is based on the recommendation of the Environmental Minister Conference (1992) to limit the total risk of cancer, which results from all air pollutants, to a level of 1:2500 /2/.
In USA, most federal state standards for BaP in the ambient air are in the 0.3-0.7 ng/m3 range as the annual average /10/.
The Occupational Safety and Health Administration (OSHA) has established a legally enforceable limit of 200 µg of all PAHs per m3 /11/.
As PAHs are predominately adsorbed on particulate, there are indirectly limited via the thresholds for particulate.
PAHs in the haze from forest fires in Indonesia 1997
Burning of low energy dense material such as wood and charcoal generally emit more particulate and hydrocarbons than burning of energy dense materials (gas, mineral oil, black/brown coal). The more incomplete combustion of wood and charcoal is favored by their higher content of water and impurities, as lignite and cellulose /10/.
The US-Environmental Agency estimates that the cancer risk from wood smoke is twelve times greater than that from equal amounts of tabacco smoke /6/.
Analysis of PAHs in the Haze from Forest Fires in Indonesia 1997
To examine the PAH-content in the haze caused by large scale forest fires in Indonesia 1997, particulate samples have been taken during intensive haze in Jambi, Sumatra in October/November 1997. The haze originated from forest and peat fires, less than 100 km far from the sampling location. For the sampling of Total Particulate Matter (TPM) or PM10 respectively a High-Volume and Low-Volume Sampler were used. The filter samples have been forwarded to Japan and Germany for PAH- analysis (HPLC+GC/MS).
Tab.1: Sampling and Analysis of Particulate Matter during the haze period in Jambi, 1997
by JICA, Japan
by JICA, Japan
by GTZ, Germany
An analysis of gaseous PAH has not taken place, which are assumed to be neglectable in relation to the adsorbed PAHs /5/.
The PAH-values of sample1 have been analyzed out of the Total Particulate Matter (TPM). Around 90% of the TPM in the forest fires haze consist of particulate smaller than 10 µm diameter (PM10). Consequently, TPM of 1707 µg/m3 corresponds to a PM10-concentration of 1536 µg/m3 (Sample 1).
Tab. 2: Concentration of adsorbed PAH in the ambient air during the haze-period in Jambi, October/November 1997 (n.a.: not analysed, b.d.: below detection limit)
in medium polluted areas*1
Summe PAH (ng/m3)
Italic : Carcinogenic or co-carcinogenic compounds
*1 Concentrations in Germany /7/, /3/
*2 Sample1: 1707 µg/m3 TPM equals 1536 µg/m3 PM10,
BaP-content in coarse particle neglectibel (assumptions),
Except Chrysene, the results correlate quite well and indicate the validity of the analysis (see also App.2).
All PAHs, except Fluoranthene, are significantly higher than the annual average PAH-level in medium polluted areas in Germany.
The PAH-indicator Benzo(a)pyren (BaP) increased with the particulate concentration. With an average value of 14 ng/m3, BaP was 5 times higher than the yearly average BaP-concentration in medium polluted areas in Germany (3 ng/m3) and more than 7 times higher than the normal background level at the sampling location which is assumed to be less than 2 ng/m3. However, compared to daily BaP-levels in polluted areas in Germany with concentrations between 4-69 ng/m3 (s.1.2), the BaPconcentration in the haze with 14 ng/m3 is relatively low.
The PAH-concentration measured in the haze refers to an exceptionally high particulate pollution level of 1300 µg/m3 PM10 (average), which is 40 times higher than the annual average particulate level in Germany.
Relating the BaP-concentration to the particulate concentration reveals that the BaP-content per mg particulate (specific BaP-content) in the haze-particulate from forest fires is 11 ng BaP per mg particulate (PM10).
The specific BaP-content of urban particulate in Germany ranges between 25 and 60 ng BaP per mg PM10. This indicates that the haze-particulate as such, compared to urban particles, contain a low amount of PAH adsorbed on their surface.
Conclusion: Health Risks due to PAHs in the Haze-Particles
Haze from forest fires does not only cause a considerable increase of particulate in the ambient air, but leads simultaneously also to an increase of the concentration of carcinogenic Polycyclic Aromatic Hydrocarbons (PAHs). The analyses of haze samples showed that the PAH-indicator Benzo(a)pyren (BaP) with 14 ng/m3 at a particulate level of 1300 µg/m3 is 5 times higher than the annual background level in medium polluted areas in Germany and 3 times the annual threshold, recommended by the EU-Directive. Compared to daily BaP-levels in polluted areas of Germany ranging between 4 and 69 ng/m3, the BaP-concentration in severe haze is relatively low.
The specific BaP-concentration of 11 ng BaP per mg particulate indicates that haze-particulate as such, compared to urban particles, contain a low amount of PAH adsorbed on their surface. Correlating the increase of BaP-concentration to the particulate concentration demonstrates that even if the particulate cause hazardous levels (421 to 600 µg/m3 PM10), the correspondent BaP-concentration is still in a range that commonly occurs in medium polluted areas in Germany.
Table 3: Short-term standards for PM10, 24 hours average (EPA) /14/
PM10-Concentration Pollution Standard Index PSI Air Quality Description Correspondent
0 50 m
£ 0.6 ng/m3
51 150 m
£ 1.7 ng/m3
151 350 m
£ 3.9 ng/m3
351 – 420 m
£ 4.6 ng/m3
421 500 m
£ 5.5 ng/m3
501 600 m
£ 6.6 ng/m3
* Assumption: BaP/PM10= 11 ng/mg
It can be concluded that compared to the health risks caused particulate matter as such, those caused PAHs seem to be of minor importance.
At haze pollution levels where the particulate concentration is categorized as hazardous and at which preventive measures are required to mitigate adverse short and long term health effects, the correspondent PAH-concentration is still in a relatively low range, compared to concentrations that occurring in Germany.
However, as particulate matter and carcinogenic compounds such as PAH represent stochastic pollutants with no “No-Effect-Level”, a risk even at very low concentrations remains.
For protection against both, the harmful effects of the particulate as such and those of adsorbed PAH, it is recommended to use face-masks and to avoid outdoor activities in times of significantly increased pollution levels.
It can be assumed that the particulate-bound PAHs-concentration decreases with the distance from the forest fires due to photochemical degradation processes.
The results are based on an analysis of 3 particulate samples from haze in Jambi 1997. For a final conclusion, further analyses are indispensable to verify the representativity of the results.
/1/ (TRGS 910-49. Benzo(a)pyren, http://www.umwelt.online.de as 28/5/98)
/2/ http://www.hamburg.de, Digitaler Umweltatlas Hamburg as 28/5/98
/3/ Landesumweltamt NRW: Luftqualitaet in Nordrhein-Westfalen, LIMES
Jahresbericht 1995, Essen 1996
/4/ Khalili, Nasrin R.; Scheff, Peter at al.: PAH Source Fingerprints for Coke Ovens,
Diesel and Gasoline Engines, Highway Tunnels and Wood Combustion Emissions,
Atmospheric Environment, Vol.29, pp.533-542, 1995
/5/ EcoChem Analytics, http://ecochem-analytics.com/Whatpa.shtml as 1/6/98
/6/ Burning Issues: Whats in Wood Smoke and other Emissions.
http://www.imaja.com/Imaja/bi/WoodSmoke.html as 1/6/98
/7/ Fresenius Institut, Stockach, Germany: Personal Information, 12/97
/8/ http://chem.leeds.ac.uk/papers/html/pahmosaic/section3_2.html as 1/6/98
/9/ Boffetta, P.: Electricity Production and Cancer Risk, Cancer Journal, Vol.6/1, 1992
/10/ Johnson, D.: Ambient Air Quality Monitoring in British Columbia, 1995,
http://www.env.gov.bc.ca/ske/skeair/pah/9503pah.html as 2/6/98
/11/ Agency for Toxic Substances and Disease Registry:Public Health Statement 1992:
PAH http://atsdr1.atsdr.cdc.gov:8080/ToxProfiles/phs9020.html as 2/6/98
/12/ TU Berlin: Skript zur LV, 1992, Berlin
/13/ McDow, S.: Sombustion Aerosol Water Content and its effect on PAH Reactivity
Athmospheric Environment Vol.29, No.7, 1995
/14/ GTZ-Integrated Forest Fire Management Project IFFM: Haze Guide, Version 2/98
as http://smd.mega.net.id/iffm/haze.html as 21/2/98
We would to thank the JICA- and EMC-Serpong-Team for their support in providing us with the analysis results and particulate samples.