-------
action to protect the health and safety of U.S. citizens in the area. Requests for assistance in dealing
with the overall situation were also received by the U.S. Government (USG) from national
governments in the region and international organizations.
A USG response coordinated by the U.S. Department'of State, involved cooperative efforts
among the U.S. Environmental Protection Agency (U.S. EPA), the Centers for Disease Control
(CDC), the Department of Defense (DOD), the U.S. Forest Service (U.S. FS), the National Oceanic
and Atmospheric Administration (NOAA), and other agencies. It was agreed that the USG response
would consist of essentially two parts: (1) a fire-fighting effort involving DOD and the U.S. FS, and
(2) an air monitoring-health study effort involving the U.S. EPA and the CDC. As part of the second
effort, an air monitoring sampling plan was developed and air quality measurements were made by
the U.S. EPA in Malaysia and Indonesia to obtain information about the size distribution and
composition of particles in the haze for possible use in health assessments. The U.S. EPA
participation was funded by the U.S.-Asia Environmental Partnership (AEP), a U.S. AID sponsored
program. Unfortunately, it was not possible for the CDC to collect respiratory health data as part
of a parallel health measurement field study. The present report (1) focuses mainly on the results
of the U.S. EPA air monitoring effort; (2) discusses potential health implications of findings from
that effort and certain other pertinent air quality information obtained from other sources; and (3)
makes recommendations for future actions.
U.S. EPA AIR MONITORING CAMPAIGN
GOALS OF THE STUDY
The main emphasis in the U.S. EPA monitoring effort was placed on measuring PM10 and
PM25 concentrations because of their evident elevated levels in the biomass burning haze relative
to allowable U.S. levels (as indicated in Figure 1). PM10 particles having aerodynamic diameters less
than 10 micrometers (also known as thoracic particles) can be inhaled past the nose and throat into
lower respiratory tract areas, including the lungs. PM10 particles consist of two main groups:
(1) PM2.5 particles having aerodynamic diameters less than 2.5 micrometers, and (2) PM,0_2 5 particles
having aerodynamic diameters between 2.5 and 10 micrometers. PM25 particles are often referred
to as "fine" particles, whereas PM,0.25 particles are often referred to as "coarse" particles. PM25
-------
particles include substances derived from high temperature processes (e.g., combustion, smelting,
etc.,.). They can reach lower regions of the lung, and are of much concern with regard to a variety
of potential adverse health outcomes. Coarse (PM10.2.s) particles include substances such as
suspended crustal material, plant and insect debris, mold spores, etc., some of which may also be of
health concern (e.g. they may exacerbate asthma symptoms). The efficiency of penetration of
particles of different sizes deep into the lungs depends on the health of the individual and the
ventilation rate. Increasing exercise (activity) levels, for example, result in increased ventilation
rates and enhanced delivery of inhaled particles to lower respiratory tract areas.
In addition to determining PM10 and PM2.5 mass concentrations, efforts were also focused on
evaluation of the composition of the ambient particles sampled. Thus, airborne concentrations of
trace elements (e.g., potassium, lead, etc.) and concentrations of total organic and elemental ("black"
or "soot") carbon in the particles were measured. Images of particles on selected samples were also
obtained by scanning electron microscopy. The concentrations of PAHs (polycyclic aromatic
hydrocarbons) in the gas phase and as part of particulate matter were also measured. All data
reported here were based on 24-hr sampling periods. See Table 1, for more information on the
sampling and analysis methods used, all of which have been used extensively by the U.S. EPA
(Pinto et al., 1998; Winberry et al., 1990; U.S. Environmental Protection Agency, 1998) and are
considered to be EPA recommended methods.
SAMPLING LOCATIONS
The locations of the U.S. EPA monitoring sites in Indonesia and Malaysia are shown in
Figure 2, which also shows haze and wind conditions during the U.S. EPA sampling campaign. The
types of sampling sites chosen in both countries were similar (i.e., measurements were made in both
countries at urban sites and at more rural background sites). The "background" measurements were
meant to capture the composition of the particles produced by the biomass burning, while the urban
measurements were meant to capture the composition of the particles produced by local sources in
addition to those produced by biomass burning. The two sets of monitoring sites were deployed
along the prevailing wind direction in both countries (i.e., along the Kelang Valley in Malaysia, and
south to north in Sumatra from Inderalaya to Palembang).
-------
TABLE 1. Sampling and Analysis Methods used in the November, 1997 U.S. EPA
Air-Monitoring Effort in SE Asia (at Malaysia and Indonesia sites).
Sample Collected' Method
PM2.5i PM,0 Modified virtual impactor containing 47mm Teflon filters
Organic carbon, elemental carbon Collected on 47mm Quartz filers mounted in same virtual
impactor
PAHs Collected on 47mm Quartz filer for aerosol phase and
polyurethane foam trap for gas phase
Trace Elements (Na-Pb) Collected on 47mm Teflon filters
Sample Analysis Method
PM25 PM10 Gravimetric analysis of deposit on Teflon filters
Organic carbon, elemental carbon Thermo-optical analysis of deposit on quartz filters
PAHs2 Gas chromatography using FID/MS detection for analysis
of extracts from quartz filters and polyurethane foam traps.
Trace Elements (Na-Pb)3 X-ray fluorescence analysis of particles collected on Teflon
filters
'All samples collected for 24-hr periods
2A11 analyses performed by Midwest Research Institute
3A11 analyses performed by ManTech Environmental Technology, Inc.
MALAYSIA RESULTS
Measurements of airborne particle concentrations were made at existing Malaysian
Department of Environment (MDOE) monitoring sites in Petaling Jaya and Shah Alam, in the
vicinity of Kuala Lumpur, as shown in Figure 3. The U.S. Army, Center for Health Promotion and
Preventative Medicine (CHPPM) also provided equipment to monitor concentrations of fine particles
(PM25) at Shah Alam and Petaling Jaya. The 24-hr PM,0 concentrations measured during
November 2 to November 11, 1997 are shown in Figure 4. It rained on several days during the
sampling period, resulting, in part, in the day to day variability seen in the PM10 levels shown in
Figure 4. Winds were mainly from the northeast during this period, except for a brief time early on
when they had originated in the southwest and may have brought in contributions of the biomass
burning particles from Sumatra. Therefore, the results presented here may be most appropriately
viewed as representing contributions from local sources to the measured ambient particles as much
as being due to smoke from the Indonesian biomass fires. The measurements made on Sumatra,
discussed below, are more representative of the composition of the biomass burning emissions.
-------
100E
105E
110E
115E
»- Low Level Wind
(/^ Hot Spots
O Moderate to thick
smoke/haze
O Slight to moderate
smoke/haze
^ Monitoring Stations
Figure 2. Wind and haze conditions on November 5,1997 (9:30 a.m.) during the U.S. EPA
sampling campaign, as indicated on Internet sites reporting on status of biomass
fire "hot spots" and haze conditions, based on information provided by various
government authorities in the region. The U.S. EPA monitoring locations near
Kuala Lumpur and on Sumatra are indicated by *.
Another goal of the U.S. EPA effort was to compare the EPA results for PM10 concentrations
to those obtained at MDOE monitoring stations operated by staff of Alam Sekitar Malaysia (ASMA)
working under contract to the MDOE. The daily 24-hr PM,0 concentrations obtained at Petaling
Jaya with U.S. EPA equipment are shown in Figure 4, along with results from collocated MDOE
monitoring equipment. The average PM10 levels recorded by the MDOE monitoring instrument
(82.4 ug/m3) were about 11 ug/m3 (15%) higher than those obtained using U.S. EPA equipment
(71.1 ug/m3). This difference is probably not significant, although a longer record is needed to draw
more definitive conclusions. The corresponding PM2.5 concentrations recorded by the U.S. Army
and U.S. EPA at Petaling Jaya are shown in Figure 5. The average PM25 concentrations measured
-------
Figure 3. U.S. EPA sampling locations in Kuala Lumpur, Malaysia (KL) and vicinity
(Sampling locations indicated by *).
using U.S. EPA (59.2 ug/m3) and U.S. Army (59.7 ug/m3) equipment were in very close agreement,
although there were differences of several per cent between the two methods on a few days. The
PM2.5 size fraction accounted for approximately 80% of the PMIO mass measured at Petaling Jaya.
During the November 2 to 11 period of EPA and U.S. Army monitoring, the PM10 values measured
by MDOE at the Shah Alam site averaged about 30% lower than those measured by MDOE at the
Petaling Jaya site, and PM25 concentrations at Shah Alam were determined by U.S. Army
monitoring to be about 20% lower than those measured at Petaling Jaya.
Information about the overall composition of the PM2 5 particles is summarized in pie-chart
form in Figure 6. As can be seen from Figure 6, most of the aerosol mass was in the form of
organic material, with smaller amounts of sulfate, elemental or "soot" carbon, crustal elements such
as aluminum (Al), silicon (Si), calcium (Ca), titanium (Ti), and iron (Fe), and other trace metals.
-------
c
o
*-<
2
*-
0)
o
c
o
O
160
150
140
120-
100-
80-
60-
40-
20
X \
\
* EPA
--*- ASMA
11/3
PM10 NAAQS
11/5
11/7
Date
11/9
11/11
Figure 4. PM10 concentrations in micrograms per cubic meter, (ug/m3), obtained using
U.S. EPA and ASMA equipment at Petaling Jaya, in KL vicinity.
in
CM'
Q_
160
140-
c
.0 120
"ro
100-
(D
o
3 80
65
60
40-
20
11/3
EPA
O- Army
PM2 5 NAAQS
11/5
11/7
Date
11/9
11/11
Figure 5. PM2 5 concentrations in micrograms per cubic meter, (ug/m3), obtained using
U.S. EPA and U.S. Army equipment at Petaling Jaya, in KL vicinity.
-------
There are many possible sources (e.g. motor vehicles, vegetation burning, plant and animal debris,
pollen, fungal spores, organic compounds which condensed onto existing particles) for the
carbonaceous constituents that were sampled. The value shown for organic carbon also reflects a
rough estimate of the amounts of organic compounds containing hydrogen, nitrogen and oxygen.
Selected filter samples were analyzed by scanning electron microscopy to obtain information about
the nature of the carbonaceous particles collected on them. Most of the larger particles were mold
spores, with a few particles present that are typical of diesel exhaust. The smaller particles were
probably generated by combustion by motor vehicles, power plants, and perhaps by vegetation
burning. These types of particles are to be expected, given the proximity of the monitoring site to
vegetation and to a nearby road. Sulfate in aerosol samples collected in the United States is typically
associated with the oxidation of sulfur dioxide (SO2) emitted by power plants and to a lesser extent
by motor vehicles. Similar sources may also have contributed to the sulfate seen in these samples,
as shown in Figure 6. Wind blown dust suspended from roads, construction sites, and natural
surfaces probably represents the major source of the crustal elements shown in Figure 6. The heavy
metals shown in Figure 6 may originate from a variety of industrial processes such as incineration,
manufacturing, smelting, etc.
The results of more detailed chemical analyses of the composition of the organic and the
inorganic fraction of the fine (PM2 5) and coarse (PM10.2.s) particles are summarized in Table 2, along
with comparison of the fine particle composition obtained at Petaling Jaya in the Kuala Lumpur area
with that obtained in selected U.S. cities (Los Angeles, CA, Philadelphia, PA, and Roanoke, VA).
These comparisons are shown to help place the air pollution values obtained in Malaysia in
perspective in relation to those encountered in U.S. cities. Los Angeles and Philadelphia were
chosen because they are both large urban areas, and Roanoke because it is impacted by wood smoke.
The mean concentration of lead (Pb) in Petaling Jaya was 39 ng/m3, compared to the U.S. National
Ambient Air Quality Standard for lead of 1500 ng/m3 (1.5 ug/m3 90-day ave.). The concentrations
of other heavy metals such as nickel (Ni), copper (Cu), and zinc (Zn) were all substantially lower,
and cobalt (Co) and cadmium (Cd) were not detected. As can be seen from the tables, concentrations
of the trace metals are well within the range of values routinely observed in U.S. cities, except that
the abundance of potassium (K) was higher in Petaling Jaya than in either Los Angeles or
Philadelphia. High levels of potassium are usually associated with wood burning, as seen in
-------
Organic Carbon
42%
Water
25%
Heavy Metals
0.1%
Crustal Elements
3.1%
Ammonium Sulfate
16%
Elemental "Black" Carbon
4.3%
Figure 6. Gross composition of PM2.5 particles sampled in Petaling Jaya, in KL vicinity.
Roanoke during the winter (Stevens et al., 1993). However, it is difficult to say how much of the
fine particle mass in the Kuala Lumpur area was due to local vegetation burning or transport from
Indonesia biomass fire areas without knowing background levels of potassium produced by other
sources. The composition of the coarse particles, i.e., with aerodynamic diameters between 2.5 and
10 micrometers (PM^.s), was dominated by soil particles and some biological material, such as
mold spores.
Concentrations of polycyclic aromatic hydrocarbons (PAHs) measured at Petaling Jaya and
similar data from cities in the Czech Republic are shown in Table 3. PAHs are produced by the
incomplete combustion of biomass, motor vehicle fuels (e.g., diesel fuels are rich sources of PAHs),
and other fossil fuels such as coal and oil. They may be present in either the gas phase or attached
to particles, depending mainly on temperature. However, some PAHs (mainly napthalene to
anthracene) partially evaporate from particle deposits during sampling, so it is not possible to say
10
-------
TABLE 2. Mean Aerosol Composition Measured by U.S. EPA at Petaling Jaya in the Kuala
Lumpur Area (November 1997) and in Selected U.S. Cities.
Species
Particulate Matter (^g/mj)
No. of Samples
Total Mass
Organic Carbon
Elemental Carbon
Metal Oxides
Sulfate
Trace Elements (ng/m3)
Al
Si
S
Cl
K
Ca
Ti
V
Cr
Mn
Fe
Ni
Cu
Zn
As
Se
Br
Pb
Petaling Jaya
(1997)
Fine
(PM2,)
9
62.1
26.1
1.9
10.0
10.0
bd2
160
2400
70
280
98
27
9.3
0.2
4.5
120
2.2
9.3
34.3
2.3
0.7
9.8
39
Coarse
(PM10.2.5)
9
11.9
no
no
1.0
1.0
630
1270
235
83
160
580
55
1.5
2.4
3.9
310
1.0
10.2
13
0.5
bd2
1.0
19
Los Angeles, CA
(1987)
Fine
(PM2.5)
11
41
8.3
2.4
11.8
35
52
2830
93
41
22
5
6
22
16
99
5
63
90
22
13
13
38
Philadelphia, PA
(1994)
Fine
(PM2.5)
21
32
4.5
0.8
13.8
114
165
3300
26
60
58
<42
<13
bd
3
127
7
7
41
bd2
<2
9
19
Roanoke, VA
(1988)
Fine
(PM2.5)
20
7.3
1.5
4.9
18
77
1180
53
177
47
bd2
1.7
1
12
114
bd2
7
83
2
2
5
27
'no = not obtained
2bd = beneath detection limit
11
-------
accurately what fraction of these species was present initially in the atmosphere in either the particle
or gas phase.
As for placing the observed results for PAH compounds in perspective, Menichini (1992)
reviewed airborne levels of PAHs measured since the mid-1970's in approximately 60 urban areas
world-wide. According to Menichini, one of the more notable PAHs of potential health concern, i.e.,
the strongly carcinogenic compound benzo[a]pyrene (BaP), typically has been measured in the range
of 1-20 ng/m3 in Europe and around 1 ng/m3 in the United States since the mid-1970's. The
concentrations of other individual PAH compounds were generally found in the range of: 1-50 ng/m3
in Europe; 0.1-1 ng/m3 in North and South America and Australia; 1-10 ng/m3 in Japan; and 10-100
ng/m3 in the Indian city of Calcutta and a highly polluted urban area of New Zealand. Measurements
of PAHs in Western Europe and America were typically higher from the 1950's to the early 1970's,
before substantial reductions in emissions from widespread coal burning were accomplished.
European BaP concentrations, for example, ranged from ~ 5 to 100 ng/m3 in winter (some up to 600
ng/m3) and a few ng/m3 in summer (some up to 30 ng/m3), whereas concentrations at least an order
of magnitude lower were found in the United States. Earlier (pre-mid 70's) concentrations of other
individual PAHs ranged from a few to about 50 ng/m3 in European towns and 1-10 ng/m3 in U.S.
and Australian towns. Different analytical techniques used in the above-noted studies make precise
quantitative comparisons across different time periods and geographic areas difficult to obtain.
However, a rough comparison of the values reported here indicates that the concentrations of most
all of the PAHs measured in the KL area (including BaP values) fall well within the lower ranges
typically reported for most urban areas sampled world-wide since the mid-1970's.
Results allowing for more confident direct comparisons to those obtained here for Petaling
Jaya were derived in Los Angeles (as reported by Grosjean, 1983) and from analogous air
monitoring efforts carried out by the U.S. EPA and the Czech Ministries of the Environment and
Hygiene in two Czech cities, Teplice and Prachatice, in the early 1990's (as reported by Pinto et al.,
1998). Concentrations of the PAHs anthracene, fluoranthene, and pyrene in Los Angeles air were
distinctly lower than those measured in Petaling Jaya. It is not clear if values for other PAHs
(signified by " < " for Petaling Jaya) were also lower in Los Angeles although the concentrations
for several of the PAHs were probably much lower than the detection limit values (signified by" <")
in Petaling Jaya. Concentrations of various PAHs found in the two Czech
12
-------
TABLE 3. Mean Polycyclic Aromatic Hydrocarbon (PAH) Concentrations (ng/m3) Measured
in November 1997 by U.S. EPA at Petaling Jaya in the Kuala Lumpur Area,
Compared to Range of Seasonal Means Observed in Los Angeles1 and Selected
Czech Cities2.
Compound
Petaling Jaya
napthalene
acenapthylene
acenapthene
fluorene
phenanthrene
anthracene
fluoranthene
pyrene
benz[a]anthracene
chrysene
benzo[6]fluoranthene
benzo[£]fluoranthene
benzo[e]pyrene
benzo[a]pyrene
indeno[7, 2, 3-o/|pyrene
dibenzo[a, /j]anthracene
benzo[g, h, /Jpery lene
coronene
perylene
Total PAHs
<4.9
6.0
3.9
12.5
50.9
12.1
10.9
11.9
<4.1
<4.1
<6.5
<4.0
<4.0
<3.8
<3.6
<4.4
<4.1
<4.7
<4.0
135.03
Los
(Summer)
0.0
0.8
1.5
0.2
0.6
0.4
0.2
0.1
0.2
0.3
1.4
5.73
Angeles
(Winter)
0.8
1.0
1.7
0.6
1.2
1.2
0.4
0.6
0.6
4.5
4.7
17.33
Czech Cities
(Summer 1993) (Winters 1993
&94)
10.7-11.7 30.8-61.1
4.2-5.6 75.3-80.1
3.4-4.0 18.3-38.6
2.5-2.6 15.3-37.9
0.4-0.5 5.8-11.6
nr nr - 6.6
nr nr - 6.7
0.1-0.3 3.6-7.2
0.1-0.5 3.4-8.0
0.4 - 0.5 4.5 - 7.7
nr nr- 1.4
0.3 - 0.4 2.4 - 5.4
'
24 - 273 84 -2783
"Values reported by Grosjean (1983) for Los Angeles.
2Range of values reported by Pinto et al. (1998) for Prachatice and Teplice, Czech Republic during non-heating
(summer 1993) and heating (winters 1993 & 94) seasons, reflecting the impact of combustion of lignite (brown coal)
for heating.
'Based only on values shown above. True value may be higher because of non analyzable compounds and because of
the evaporation of some of the lighter PAHs (shown in the upper portion of the table) during sampling.
nr = data not reported; = not measured in Czech cities
13
-------
cities during the summer months of 1993 tended to be somewhat lower than those reported here for
Petaling Jaya). However, the concentrations for many of the PAHs were distinctly higher in the
Czech cities during the winters of 1993 and 1994, because of higher PAH emissions from the
combustion of lignite (peat-like brown coal) used primarily in small hand fired furnaces for heating.
INDONESIA RESULTS
Kalimantan and Sumatra were the areas most severely affected by the biomass fires of 1997.
The U.S. EPA conducted monitoring for particulate matter (PM) at two sites on Sumatra from
November 4 to 8, 1997. The first site was located in Palembang, a heavily industrialized city; and
the second was located on the campus of Sriwijaya University, in Inderalaya, about 30 km southwest
of Palembang. Both sites were relatively close to "hot spots", where some biomass burning was still
continuing in southern Sumatra during the EPA sampling period. Sampling was conducted by staff
from the U.S. EPA, the Indonesian Ministry of the Environment (Bapedal) and the faculty of
Sriwijaya University, using U.S. EPA equipment. Both PM10 and PM25 concentrations were
measured in Palembang, but only PM2 5 particle concentrations were measured on the campus of
Sriwijaya University. The composition of the organic and inorganic chemical components of the
particles was evaluated for samples collected at both sites.
The mean daily (24-hr) PMi0 concentration at Palembang was 402 ug/m3. The mean daily
fine particle (PM25) concentration was 341 ug/m3 at Palembang and 264 ug/m3 at Sriwijaya
University. Because of equipment problems, the PM2 5 mass concentration in Palembang had to be
estimated on the basis of the concentrations of the measured components: carbon, sulfate, and trace
elements. This procedure is expected to underestimate the true level by no more than 10%. The
concentrations of PM10 and PM2 5 exceeded the 24-hr U.S. National Ambient Air Quality Standards
(NAAQS) for both PM10 (150 ug/m3) and PM25 (65 ug/m3) by large margins on all five days. The
PM10 average corresponded to a U.S. EPA PSI value of about 300 for PM10 levels, with values for
several individual days reaching higher levels categorized as "Hazardous". Approximately 85% of
the mass of the particles was concentrated in the PM2 5 (fine size) fraction at Palembang. Since the
Indonesian government does not routinely monitor airborne particulate matter levels at the Sumatra
14
-------
sites sampled by U.S. EPA, no intercomparisons with their equipment or evaluation of their
techniques could be performed.
Information about the overall composition of the PM2.5 particles collected in Palembang is
summarized in pie-chart form in Figure 7. As shown by Figure 7, most of the particle mass was in
the form of organic matter, with smaller amounts being in the form of: sulfate, elemental or "soot"
carbon; crustal elements such as aluminum (Al), silicon (Si), calcium (Ca), titanium (Ti), and iron
(Fe); and other trace metals. Images obtained by scanning electron microscopy indicate that the
particles were composed mainly of hygroscopic carbon compounds. The value shown for organic
carbon also reflects a rough estimate of the amounts of organic compounds containing hydrogen,
nitrogen and oxygen. Small amounts of the organic compounds could have also been produced by
motor vehicle emissions and the condensation of organic vapors. Mold spores and plant and animal
debris were also present. Sulfate in aerosol samples collected in the United States is typically
associated with the oxidation of sulfur dioxide (SO2) emitted by power plants and to a lesser extent
by motor vehicles. Similar sources may also have contributed to the sulfate seen in Figure 7. Wind
blown dust suspended from roads, construction sites, and natural surfaces probably represents the
major source of the crustal elements shown in Figure 7. Heavy metals shown in Figure 7 originate
from a variety of industrial processes such as incineration, manufacturing, smelting, etc.
The results of more detailed chemical analyses of the composition of the organic and the
inorganic fraction of the PM25 particles are summarized in Table 4 for Palembang and Table 5 for
Sriwijaya University. The values shown in Tables 4 and 5 may be compared to the values for
representative U.S. cities shown earlier in Table 2. As can be seen, the values for most trace metals
are similar, but concentrations of chlorine (Cl) and potassium (K) are much higher for the Indonesian
sites. Potassium (K) is frequently used as a tracer for biomass burning, because of its high
abundance in wood. Ratios of K to total mass ranged from 0.5 to 1.0% in the PM2.5 samples
collected. These values are characteristic of wood burning emissions. Most of the mass of the
emissions from biomass burning is typically in the form of organic compounds, and these were
elevated in both the Palembang and Sriwijaya sites. Thus, the organic matter and the overall PM2 5
particle composition at both Indonesian sites appear to be dominated by biomass burning emissions.
The composition of the coarse particles at Palembang reflects mainly soil, perhaps suspended by
motor vehicle traffic, with additions of biomass burning products.
15
-------
Organic Carbon
83%
Heavy Metals
.05%
Elemental "Black" Carbon
1.6%
Crustal Elements
4.4%
Ammonium Sulfate
13%
Figure 7. Gross composition of PM2.S particles sampled in Palembang, Indonesia. Sum of
percentages given exceeds 100% because of analytical uncertainties.
PAH levels measured on Sumatra (c.f. Table 6) were generally higher than those measured
in the Kuala Lumpur area and in the Czech cities shown in Table 3. Some of the PAHs were found
associated with particles in the Palembang samples but not in the samples collected at Sriwijaya
University. It is not clear why this occurred. Perhaps the samples collected at Palembang were more
strongly influenced by local diesel emissions or by other incomplete combustion sources, which
produce PAHs that adhere to particles. In any event, the PAHs at both sites are indicative of the
incomplete combustion of biomass, motor vehicle fuels and other types of fossil fuels.
POTENTIAL HEALTH IMPLICATIONS
MALAYSIA
The potential health implications of the air monitoring data discussed above are difficult to
assess because of the very limited data set. Based on data collected by MDOE, the short-term
excursions of ambient PM10 concentrations to levels notably above the U.S. EPA and Malaysia daily
PM10 standards nearly every day during the last two weeks of September, 1997 must certainly be
16
-------
TABLE 4. Mean Aerosol Composition at Palembang, Indonesia, Based on U.S. EPA
Measurements (November 4-8,1997).
Species
Particulate Matter fag/m3)
No. of Samples
Total Mass
Organic Carbon
Elemental Carbon
Metal Oxides
Sulfate
Trace Elements (ng/m3)
Al
Si
S
Cl
K
Ca
Ti
V
Cr
Mn
Fe
Ni
Cu
Zn
As
Se
Br
Pb
Fine
(PM2.5)
5
341
282
5.4
15
44
bd2
200
11000
4500
1400
79
11
bd2
bd2
1.7
83
3.8
2.1
13
1.2
3.9
95
64
Coarse
(PM1M.5)
5
61
no1
no1
4.4
1300
3700
1100
1200
420
1400
100
3.0
1.3
17
1000
0.4
2.1
20
0.5
<0.1
11
15
"no = not obtained
2bd = beneath detection limit
17
-------
TABLE 5. Mean Aerosol Composition at Sriwijaya University, Indonesia, Based on U.S. EPA
Measurements (November 4-8,1997).
Species
Particulate Matter (jj-glm*)
No. of Samples
Total Mass
Organic Carbon
Elemental Carbon
Metal Oxides
Sulfate
Trace Elements (ng/m3)
Al
Si
S
Cl
K
Ca
Ti
V
Cr
Mn
Fe
Ni
Cu
Zn
As
Se
Br
Pb
Fine
(PM2.5)
5
264
200
3.2
13
29
bd1
115
6900
4600
1500
47
6.5
bd1
bd1
bd1
71
O.I
3.9
6.4
1.3
1.4
72
7.7
bd = beneath detection limit
18
-------
TABLE 6. Mean Polycyclic Aromatic Hydrocarbon (PAH) Concentrations (ng/m3) measured
By U.S. EPA in Sumatra (November 4-8,1997).
Compound
napthalene
acenapthylene
acenapthene
fluorene
phenanthrene
anthracene
fluoranthene
pyrene
benzo[o]anthracene
chrysene
benzo[6]fluoranthene
benzo[£]fluoranthene
benzo[e]pyrene
benzo[a]pyrene
indeno[7, 2, 3-cd]pyrene
dibenzo[a,/j]anthracene
benzo[g, h, /]perylene
coronehe
perylene
Total PAHs
Palembang
12.1
11.8
<2.8
47.8
187.8
32.8
38.5
42.8
8.7'
11.7'
10.4'
<4.0
6.3'
7.1'
4.8'
<4.4
5.8'
<4.7
<4.0
318.02
Sriwijaya University
<4.9
<4.0
<2.8
26.9
108.6
17.1
21.0
18.1
<4.1
<4.9
<6.5
<4.0
<4.0
<4.0
<3.6
<4.4
<4.1
<4.7
<4.0
222.02
'Present mainly in aerosol phase
2Based only on values shown above. True value may be higher because of non analyzable compounds, and because of
the evaporation of some of the lighter PAHs (shown in the upper portion of the table) during sampling.
viewed as having posed some increased acute health risks for the general population in the Kuala
Lumpur area. This is especially likely, given indications from the U.S. EPA air monitoring efforts
that a substantial proportion (-80%) of the PM10 mass in the Kuala Lumpur area appeared to be
small sized, fine (PM2.5) particles. Available epidemiologic studies, reviewed in the recent U.S. EPA
air quality criteria document for particulate matter (U.S. EPA, 1996) and by the World Health
Organization (WHO, 1998), indicate that short-term (24-hr) exposures to ambient particles measured
19
-------
as PM10 or PM25 are associated with some increased risk of mortality and morbidity (measured as
increased respiratory symptoms, hospital admissions, etc.,.), especially among the elderly(s65 yrs
old) and persons with preexisting cardiorespiratory disease conditions. Some increased risk for
exacerbation of asthma symptoms in asthmatic individuals or, possibly, worsening of acute
respiratory disease impacts (e.g. in the case of acute respiratory infections or pneumonia), could also
occur based on available PM epidemiologic studies.
As for the health implications of specific PM constituents (e.g. elemental or organic carbon,
crustal materials or heavy metals, sulfate, etc.) or PAHs, the EPA monitoring results for KL were
not indicative of much, if any, increase in acute exposure health risks. This outcome is not
necessarily very meaningful, however, given that the EPA monitoring data were obtained during a
distinctly lower air pollution period after the peak period of haze from the biomass fires had passed.
Also, increased risks of possible cancer and non-cancer health risks associated with the specific
compounds measured are typically of most concern in relation to prolonged chronic exposures to
ambient air concentrations encountered by the general public (versus usually much higher acute
exposure levels often experienced in occupational settings). Such chronic environmental exposure
health risks are generally assessed and quantified based on the assumptions of daily (24-hr)
exposures over an entire lifetime (70 yrs average) for susceptible individualsa scenario clearly not
met by the relatively brief increased exposures in Malaysia to biomass fire emissions components
in Fall 1997. Repeated, more prolonged exposures to biomass fire smoke constituents every few
years, however, might result in cumulative doses projected to be associated with increased health
risks. Much more extensive data and assessment efforts would be necessary to attempt any more
specific estimation of potential health impacts of increased air pollution concentrations in Malaysia
due to haze from biomass fires.
The risks of any adverse effects due to exposure to biomass burning haze components are
likely to be substantially less for U.S. Embassy personnel and their dependents in the Kuala Lumpur
area than for the general local population, to the extent that individuals falling in special risk
categories of most concern (i.e., elderly individuals over 65 yrs. old, individuals with preexisting
chronic respiratory diseases or cardiac problems, etc.) were screened out by medical examinations
prior to KL posting. Compliance with recommendations to reduce activity levels and/or remain
indoors during high API days among U.S. Embassy staff and dependents would also tend to reduce
20
-------
health risks for U.S. personnel compared to the general population. On the other hand any
U.S. personnel with undiagnosed asthma or those progressing toward more severe asthma from mild
forms readily amenable to effective medication control, or persons with acute respiratory infections
(e.g. pneumonia), could be placed at increased risk for rapid onset of worsening of respiratory
systems and lung function declines due to short-term acute exposure to the haze produced by the
biomass fires.
INDONESIA
The situation in Indonesian areas impacted by haze from the biomass fires, especially those
relatively close to "hot spots" where concentrations of PM and certain other air pollutants are likely
at their highest, almost certainly posed substantially greater risk for the local general population.
Taking the EPA Indonesia monitoring results as an example, daily PM,0 levels in Palembang
approached or exceeded levels that would be deemed to be "Hazardous" in terms of the U.S. EPA
Pollutant Standard Index (PSI) or analogous Malaysian API values shown in Figure 1. Furthermore,
given that the EPA monitoring effort occurred after the extent and intensity of the fires had already
been substantially reduced from earlier peak levels, the local population in Palembang and its
vicinity were likely exposed on a daily basis to even higher PM levels during the preceding months
of September and October.
Of particular concern is the very high percentages (i.e., 85%) of the PM)0 mass attributable
to PM2 5, the "fine particle" size fraction thought to be most clearly implicated in increasing risks of
mortality and morbidity due to exposures to PM of ambient origin, as evaluated by both U.S. EPA
(1996) and WHO (1998). Assuming that the average PM25 mass concentration of 264 jig/m3
detected by EPA monitoring at Sriwijaya University largely reflects the impact of smoke from
nearby biomass burning and that similar biomass smoke input levels contributed to the average
341 ng/m3 PM2.5 mass found by EPA monitoring in Palembang, it can be estimated that an average
increment of about 250 ng/m3 PM2 5 from the biomass burning haze was added to the daily average
of about 75 ng/m3 of PM25 generated from other local sources in Palembang. That increment of
haze-related particle exposure in excess of background particle levels due to local sources can be
projected (based, for example, on quantitative risk estimates published by WHO, 1998) to have
likely contributed to detectable increases in respiratory symptoms and hospital admissions for
21
-------
respiratory problems among the local general population and, even possibly, some mortality among
the elderly and those with preexisting chronic lung diseases or cardiac conditions. Again, much
more extensive information and assessment efforts would be needed to attempt even a rough
estimate of potential increases in mortality or morbidity among members of the general population
in Palembang or other Indonesian areas impacted by the haze from the biomass burning. The public
health impacts could be fairly substantial given the PM concentrations involved and the size of the
likely affected Indonesian populations in Sumatra, Kalimantan, etc. Even larger, very substantial
public health impacts would be proj ected if Jakarta and other parts of densely populated Java (largely
much less impacted by the 1997 biomass burning haze) were to experience any prolonged periods
of haze from biomass burning.
KEY FINDINGS AND RECOMMENDATIONS
MALAYSIA
Key Findings
1. Based on MDOE air monitoring results during the last half of September, 1997, daily PM,0
concentrations in Kuala Lumpur (KL) dramatically increased to levels well in excess of U.S. and
Malaysian PM10 air standards, on some days approaching levels judged to be "Hazardous" in
terms of U.S. PSI or Malaysian API values, largely as the result of haze transported from
Indonesian biomass fires.
2. U.S. EPA air monitoring results, obtained during the first two weeks of November, 1997 (after
the peak period of biomass fire haze over Malaysia had passed), indicated good agreement with
MDOE results (within ca. 15%) obtained from collocated PM10 monitoring at Petaling Jaya (a KL
suburb). Approximately 80% of the PM10 mass at Petaling Jaya was attributable to PM2 5
concentrations monitored there by U.S. EPA.
3. The PM10 levels measured by MDOE during late September, probably represent short-term
incremental PM contributions from the biomass fires of, on average, about 250/ug/m3 above
background PM levels from local KL sources. Based on current knowledge of the sizes of
particles produced by biomass burning and the results in Sumatra, most of PMIO was probably
made up of PM2.5 during the haze episodes.
22
-------
4. Recent U.S. EPA and WHO evaluations of available PM epidemiology studies indicate that such
incremental acute exposures to daily (24-hr) PM concentrations in the above range are likely
associated with increased risk of adverse health effects (e.g. increased respiratory symptoms and
hospital admissions of the elderly, those with preexisting chronic cardiorespiratory disease, and
asthmatic persons) among the general local population.
5. Probably lower adverse health effects risks were posed, overall, for U.S. Embassy staff and
dependents in KL, to the extent that susceptible individuals (65 + yrs. old, these with preexisting
cardiac or chronic lung diseases, etc.) were screened out from overseas posting to KL.
6. The concentrations of specific chemical components monitored in the KL area (e.g. trace metals,
PAH compounds, etc.,) by U.S. EPA were not found to be particularly remarkable, individually,
in being likely to pose much, if any, health threat, unless repeated prolonged exposures over
several weeks or months every few years occurred. Such repeated exposures could result in
cumulative doses associated with increased risk of adverse cancer or non-cancer health effects.
However, the concentrations reported here were observed after peak exposures to haze
components had passed.
Recommendations
1. Retrospective epidemiologic analyses relating KL health statistics (e.g. hospital admissions,
mortality rates, etc.) to PMIO levels measured by MDOE during the August-November, 1997
period appear warranted to evaluate, more quantitatively, the potential health impacts of the
biomass burning haze on the general population.
2. Ambient PM2 5 concentrations should also be routinely measured at MDOE monitoring sites in
KL and if possible, elsewhere in Malaysia. Consideration should also be given to air quality
monitoring at the U.S. Embassy, especially during haze events, provided that appropriately
trained personnel are available to conduct the monitoring and the analyses can be performed
locally.
3. The siting of existing MDOE air monitoring equipment should be reviewed to ensure consistency
with U.S. EPA guidelines to minimize potential artifacts in sample collection/measurement.
23
-------
INDONESIA
Key Findings
1. The lack of any well-established, routinely operating air monitoring network in Indonesia
precludes having data available by which to attempt analyses analogous to the one presented here
based on MDOE PM data for KL during August-November, 1997.
2. Nevertheless, even the brief period of U.S. EPA monitoring in early November at two sites near
biomass fire hot spots on Sumatra indicates that ambient levels of both PMIO and PM2 5 markedly
exceeded U.S. PM Standards and approached or exceeded 24-hr PM10 levels designated as
"Hazardous" in terms of U.S. PSI values. Most of the particles were in the PM25 fraction, and
the specific composition of the particles and presence of particular PAH compounds are
characteristic of wood smoke.
3. The daily (24-hr) PM10 and PM2.5 concentrations measured by U.S. EPA at the Sumatra sites
(Palembang, and Sriwijaya University) probably posed increased public health risks for the local
general population. The U.S. EPA monitoring was conducted after reductions in the number of
hot spots seen in satellite images had occurred. Thus, exposures of the general population to even
higher PM levels for prolonged periods of time (weeks, months) likely occurred, pointing toward
even more substantial public health impacts in Indonesia being associated with exposures to haze
from the biomass burning.
Recommendations
1. Strong measures should be taken to control the setting and spread of the biomass fires in
Indonesia and elsewhere in the S.E. Asia region. Also, public education programs should be
conducted to better inform the populations in the region about the detrimental health and
environmental impacts of uncontrolled biomass burning.
2. Air quality monitoring should be conducted in Indonesia on a regular basis in major urban areas
and in areas likely to be impacted by biomass burning emissions. Monitoring the so-called
criteria pollutants (PM10, PM2 5, O3, CO, SO2, NOX, and Pb) should be given high priority, with
measurements of PM10 and PM2 5 given highest priority.
24
-------
3. An effective air quality index system analogous to the U.S. EPA PSI system should be
implemented to better inform Indonesian civil authorities and citizens about unhealthy air quality
conditions and to assist in taking appropriate actions to avert or lessen public health impacts.
4. Possibilities for conducting future health effects measurement studies to evaluate health impacts
of recurring biomass fires should be explored.
Acknowledgement:
EPA would like to acknowledge the AEP technical representatives in Kuala Lumpur and Jakarta,
the U.S. Embassies in Malaysia and Indonesia, and the Malaysian Department of the Environment
(MDOE) and the Indonesian Ministry of Environment (Bapedal) for their support and assistance in
this monitoring effort.
REFERENCES
Chia, L. S. and Wong, M. L. (1977) Atmospheric Pollution and its control in Singapore. Proc.
Symp. on our Environment, Singapore, pp. 100-123
Grosjean, D. (1983) Polycyclic aromatic hydrocarbons in Los Angeles from samples collected on
teflon, glass and quartz filters. Atmospheric Environment, 17: 2565-2573.
Menichini, E. (1992) Urban air pollution by polycyclic aromatic hydrocarbons: levels and sources
of variability. Sci. Total Environ. 116: 109-135.
Nichol, J. (1997) Bioclimatic impacts of the 1994 smoke haze event in southeast Asia. Atmos.
Environ. 31: 1209-1219.
Pinto, J. P.; Stevens, R. K.; Willis, R. D.; Kellogg, R.; Mamane, Y.; Novak, J.; Santroch, J.;
Bene§, I.; Lenidek, J.; Bures, V. (1998) Czech air quality monitoring and receptor modeling study.
Environ. Sci. Technol. 32: 843-854.
Samah, A. A and Tick, L. J. (1992) Investigation of transport of air pollution in the Klang Valley,
Kuala Lumpur, Malaysia. J. Fiz. Mai. 13(18):! 8-28.
Stevens, R. K.; Hoffman, A. J.; Baugh, J. D.; Lewis, C. W.; Zweidinger, R. B.; Cupitt, L. T.;
Kellogg, R. B.; Simonson, J. H. (1993) A comparison of air quality measurements in Roanoke, VA,
and other integrated air cancer project monitoring locations. In: Measurement of toxic and related
air pollutants: proceedings of the 1993 U.S. EPA/A&WMA international symposium; May; Durham,
25
-------
NC. Pittsburgh, PA: Air & Waste Management Association; pp. 185-189. (A&WMA specialty
conference VIP-34).
U.S. Environmental Protection Agency. (1998) Compendium of Methods for the Determination of
Inorganic Compounds in Ambient Air; in review.
U.S. Environmental Protection Agency. (1996) Air quality criteria for particulate matter. Research
Triangle Park, NC: National Center for Environmental Assessment-RTP Office; report nos.
EPA/600/P-95/001aF-cF. 3v. Available from: NTIS, Springfield, VA; PB96-168224.
Winberry, W. T.; Murphy, N. T.; and R. M. Riggan (1990) Methods for Determination of Toxic
Organic Compounds in Air: EPA Methods, Noyes Data Corporation, Park Ridge, NJ.
World Health Organization. (1998) Air quality guidelines for Europe. 2nd ed. Copenhagen,
Denmark: Regional Office for Europe; in press.
26
-------