See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/345652195

Main Sources of Air Pollution in DKI Jakarta

Technical Report · September 2020

CITATIONS READS

0 518

3 authors, including:

Puji Lestari
Bandung Institute of Technology
63 PUBLICATIONS   1,477 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Air Quality Modeling Surrounding Iron and Steel Industry in Cilegon, Banten Province View project

SPARTAN Project View project

All content following this page was uploaded by Vivian C. Pun on 10 November 2020.

The user has requested enhancement of the downloaded file.

Main Sources
of Air Pollution
in DKI Jakarta
Technical Brief
 Main Sources of Air Pollution
in DKI Jakarta
Technical Brief

Overview

Air pollution caused 4.9 million premature deaths globally in 2017,

with ambient fine particulate matter (PM2.5) pollution accounting for 3
millions of those deaths.1 PM2.5 formation varies considerably across
References
different emission sources, geographical region and seasons, making
it a challenging pollutant to characterize and regulate for adequate 1 IHME. GBD Compare. https://
public health protection. vizhub.healthdata.org/gbd-compare
/#. Published 2019. Accessed
January 8, 2019.
An effective air quality management starts with knowing the leading
sources of pollution in the city. Information on what share of air 2 Vital Strategies. Accelerating City
Progress on Clean Air - Innovation
pollution problem is attributable to different sources (e.g., traffic) and Action Guide. New York, NY;
allows effective clean air actions to be prioritized.2 Generally, this 2020. https://www.vitalstrategies.org
involves identifying leading sources of air pollution using two comple- /wp-content/uploads/Accelerating
CityProgressCleanAir.pdf.
mentary approaches: 1) the source-based (or bottom-up) approach
that uses emission inventory of sources and meteorological data to 3 The United States Department of
State. AirNow. https://airnow.gov/
simulate source contributions to ambient pollution concentration; and index.cfm?action=airnow.global_
2) the receptor-based (or top-down) approach that collects ambient summary. Accessed January 8, 2019.
air filter samples at air quality monitoring sites (i.e., receptor), analyzes
4 Kementerian Lingkungan Hidup
the chemical composition in the samples, and matches the chemical Indonesia. Status Lingkungan
profiles with those of emissions from different fuel types.2 This Hidup Indonesia 2010.; 2012.
doi:10.1146/annurev.energy.28
approach is complementary to and can be used to identify potential .050302.105509
discrepancies with findings from the source-based approach.
5 Lestari P. Inventarisasi Emisi PM2.5,
CO, NOx, SO2, BC, Dan GHG Di
Despite the fact the annual average PM2.5 concentrations in Jakarta are Jakarta.; 2019.
routinely four to five times the World Health Organization (WHO)
6 Breathe Easy J. Factsheet 3:
health-based Air Quality Guidelines (10 µg/m3 for annual PM2.5 average)3, Emission Inventory. http://www.
reliable and recent information on leading sources of air pollution is urbanemissions.info/wp-content/
uploads/docs/2017-01-Jakarta
scarce. Over the past decades, three ad-hoc air pollution emission
-Facsheet3-Emissions-Inventory.
inventories for Jakarta were compiled4–6; only two included an pdf. Published 2016.
inventory for PM2.5.5,6 Similarly, results of the only receptor-based
7 Santoso M, Lestiani DD, Markwitz
source apportionment study for Jakarta, published in 2008, are quite A. Characterization of airborne
dated now, and likely biased, as the study estimated the contributing particulate matter collected at
PM2.5 sources near a roadside monitor in the city.7 To address this Jakarta roadside of an arterial road.
J Radioanal Nucl Chem. 2013;
major evidence gap, Vital Strategies worked with the Bandung 297(2): 165-169. doi:10.1007/
Institute of Technology (Institut Teknologi Bandung or ITB) to expand s10967-012-2350-5
an ongoing receptor-based source apportionment study in progress
to better assess the contributing sources of ambient PM2.5 level in two
seasons at three urban locations around Jakarta.

2
 Main Sources of Air Pollution in DKI Jakarta

Approach

In order to understand the PM2.5 chemical content (composition) and

subsequently its leading sources across Jakarta, ITB collected 24-hour
filter samples of PM2.5 at three urban background sampling sites that
were interspersed in different districts of Jakarta: Gelora Bung Karno
(GBK), Kebon Jeruk and Lubang Buaya (Figure 1). These sites were
selected based on land use, topography, weather (e.g. wind pattern)
and other considerations to capture potential variation in air pollution
sources, and were co-located with government-owned air quality
monitoring stations at the provincial and national level (Figure 1).
Monitoring at two of the three sites (i.e., Kebun Jeruk and Lubang
Buaya sites) were an expansion of activities set up by ITB previously
under the Toyota Clean Air Program (TCAP) project.

Figure 1
West residential site (1 km away from intercity
road to Banten, co-located with provincial
environmental agency monitor – DKI-5)

Wet: 12 samples; Dry: 10 samples

Central mixed site (located in an open field

besides a stadium and next to major road, and
co-located with national Ministry of Environ-
ment and Forestry monitor)

Wet: 32* samples; Dry: 9 samples

* Samples collected with financial support
from TCAP.

East mixed site (located in a park next to small

road, co-located with provincial environmen-
tal agency monitor – DKI-4)

Wet: 9 samples; Dry: 16† samples

† Additional samples were collected due to
sample contamination

There were two sampling periods: one wet season (October 2018 to
March 2019) and one dry season (July to September 2019). Sampling
was generally conducted on a daily basis for on average 10 days, on
both weekdays and weekends in order to capture all emission source
patterns throughout the week (e.g., less traffic on the weekend), with
the exception of rainy days when no sampling was conducted.

On each sampling day, two Air Mini Volume samplers equipped with
a 47mm Teflon filter and a 47 mm Quartz filter, respectively, were
used simultaneously to collect PM2.5 samples. Meteorological param-
eters such as temperature, humidity, wind speed and wind direction
were also measured simultaneously using anemometer, barometer,
compass and sling psychrometer, respectively. Sample collection at
the three sampling sites was staggered due to limited equipment
availability.

3 Technical Brief
 Main Sources of Air Pollution in DKI Jakarta

Chemical species characterization

All PM2.5 samples collected were stored in the refrigerator before
chemical composition analysis. The mass concentration of PM2.5 was
measured by the reference gravimetric method, that is, by weighing
filters before and after sampling.

To enable leading sources of pollution to be identified, filter samples

were sent to local and international laboratories, where specific
chemical components were detected. Specifically, black carbon
(BC), 11 metal elements, and five ions were speciated from Teflon
filters; whereas organic carbon (OC) and elemental carbon (EC) were
speciated from Quartz filters. The 11 metal elements identified were
sodium (Na), aluminum (Al), silicon (Si), sulfur (S), chlorine (Cl), potas-
sium (K), calcium (Ca), titanium (Ti), manganese (Mn), iron (Fe), nickel
(Ni), copper (Cu), zinc (Zn) and lead (Pb). Five ions identified were
sodium ion (Na+), chloride ion (Cl–), nitrate (NO3–), sulfate (SO42–), and
ammonium ion (NH4+). Five ions speciated were sodium ion (Na+),
chloride ion (Cl–), nitrate (NO3–), sulfate (SO42–), and ammonium ion
(NH4+). Different instruments were used to determine the concentra-
tions of PM2.5 chemical composition (Table 1); these results will be
reported elsewhere.

Table 1. Sample Analytic Instruments

Parameter Instrument
PM2.5 Gravimetric
BC EEL Smoke Stain Reflectometer Model 43D
Metal element XRF
Ions Ion chromatography
EC, OC Thermal-optical transmittance

Source apportionment analysis References

A statistical method that takes information on key chemical compo-
8. U.S. EPA (Environmental Protection Agency).
nents and source profiles was then used to scientifically estimate the
Chemical Mass Balance User Manual.; 2004.
relative contribution of leading emission sources to ambient PM2.5 https://www3.epa.gov/ttn/scram/models/
mass. In this study, the U.S. Environmental Protection Agency’s Chemical receptor/EPA-CMB82Manual.pdf.

Mass Balance (CMB), version 8.2 (https://www3.epa.gov/scram001/

receptor_cmb.htm), was applied. At least 11 known source profiles
(unique chemical signature) were used as input data, along with
measurements of PM2.5 chemical composition. Details on CMB
modeling have been described previously.8 All filter measurements
from the same sampling site were combined and analyzed in one
CMB model, resulting in three models developed for the three sites.

4 Technical Brief
 Main Sources of Air Pollution in DKI Jakarta

Six to nine factors were evaluated using goodness-of-fit parameters

(i.e., chi-square value and reconstructed mass) to assess how well the
model fits the input composition data, and optimum number of
sources, which can be estimated with scientific certainty, was
subsequently determined.

Results
Findings will inform policymakers on the leading sources of air
pollution in the city, and also identify potential consistencies and
discrepancies with findings from earlier emission inventory results.

Air Pollution Levels in Jakarta

The average daily PM2.5 concentrations over the entire sampling
period range from 39 µg/m3 in central Jakarta (GBK) to 56 µg/m3 in
west of Jakarta (Lubang Buaya); all of which exceeded the WHO’s Air
Quality Guideline of 25 µg/m3 for 24-hour PM2.5 average, though the
Indonesian National Ambient Air Quality Standard based on
Government Regulation Number 41. year 1999 is 65 µg/m3 (24-hour
average). In general, daily PM2.5 levels were higher in the dry season
than the wet season, but the variability in pollution levels across the
city was greater in the wet season than the dry season (Figure 2),
possibly due to more influence of transported PM2.5 during the dry
season. The prevailing wind for the wet season came from West and
Southwest of Jakarta, whereas that for the dry season was from East
and Northeast of Jakarta.

Figure 2. Average daily PM2.5 concentrations in Jakarta

80
70 Indonesia PM2.5 24-hr standard
Concentration (ug/m3)

60 63
58 58
50
49
40
30 WHO PM2.5 24-hr guideline
20 31
21
10
0
Wet Season Dry Season

GBK KJ LB

5 Technical Brief
 Main Sources of Air Pollution in DKI Jakarta

Of the PM2.5 chemical components speciated, OC generally

accounted for nearly 40% of PM2.5 mass across the city, except for
central Jakarta where OC only took up 25% (Table 2). EC was another
major component of PM2.5 (7% to 13%). In central Jakarta (GBK),
assorted ions, comprising of Cl–, NO3–, SO42– and others, took up the
largest percentage (30%) of PM2.5 mass, and assorted metals, com-
prising of S, Pb, Zn and other, also accounted for 21% of PM2.5.

Table 2 Average daily PM2.5 mass concentration and its chemical

composition by sampling sites.

Kebon Lubang
GBK
Jeruk Buaya
PM2.5 mass 39 µg/m3 41 µg/m3 56 µg/m3
OC 25% 38% 37%
EC 7% 13% 8%
BC* 9%* 9%* 8%*
Ions 30% 17% 11%
Cl –
7% 2% 1%
NO 3
–
7% 5% 4%
SO42– 7% 4% 3%
Others^ 10% 6% 3%
Metals 21% 18% 15% * Since BC and EC were used interchangeably,
BC was not counted towards total PM2.5 mass
S 9% 6% 6% in the presence of EC.
Pb 2% 4% 1%
^ Other ions include Na+ and NH4+.
Zn 2% 2% 1%
# Other metals include Al, As, Br, Ca, Cl, Co,
Others# 8% 6% 7% Cu, Fe, Hg, K, Mg, Mn, Na, Ni, P, Si, Ti.

The sum of PM2.5 chemical composition concentration was about 68%

to 91% of the measured concentration of PM2.5 mass, depending on
the sampling sites and seasons (data not shown). In general, the higher
the percentage, the better the source apportionment results at
depicting the actual emission sources.

6 Technical Brief
 Main Sources of Air Pollution in DKI Jakarta

Main Sources of Air Pollution in Jakarta

Overall, six to seven factors were identified from each sampling site,
with known source profiles obtained from literature or references
(Table 3). The leading primary sources of ambient PM2.5 pollution in
Jakarta were vehicle exhaust (32%–57%), non-vehicular emissions
(17%–46% including those with natural origins such as sea salt), and
secondary formation of particles (1%–16%; secondary aerosols)
(Figure 3).

Table 3 PM2.5 chemical “trace” components associated with each

emission source.

Source Trace components

Vehicle exhaust
Gasoline vehicles OC, Zn, Ni
Diesel vehicles OC, EC, Zn
Paved road dust Fe i Secondary inorganic aerosols (e.g.,
ammonium nitrate or sulfate) are
Construction activities S, Al, Si, EC
formed when precursor gaseous
Soil Al, Si, Ti pollutants (e.g., sulfur oxides and
nitrogen oxides) undergo chemical
Opening burning OC
reactions within the atmosphere. On
Secondary aerosol i
NO3–, SO42– the other hand, primary pollutants
Coal combustion Al, Ca, Mn, Cu occur as a result of direct emissions
from an air pollution source.
Sea salt ii
Na, Cl, Na+
ii Sea salts, a natural marine emission,
Diesel generator Fe, Ni, Cu
are formed due to wind action at the
Residential LPG EC, NO3– ocean surface.

Figure 3. Source contribution (in %) to ambient PM2.5

concentrations in Jakarta

100% 2%
10%
22% 17%
26% 1%
31%
80% 7%
6%
57%
16% 43%
11%
60%
38% 42%

40% 41%
32%

46%
42%
20% 34% 34%
25%
17%
0%
GBK KJ LB GBK KJ LB

Wet Season Dry Season

Unidentified Secondary aerosols
Vehicle exhaust Non-vehicular primary sources

7 Technical Brief
 Main Sources of Air Pollution in DKI Jakarta

Figures 4 and 5 show the percentage contribution of vehicular and

non-vehicular sources in Jakarta. For vehicle exhaust, gasoline
vehicles accounted for more than half of the vehicular emissions
across the city with the exception of Lubang Buaya site during the dry
season (Figure 4). However, it is unclear how much of these came from
on-road vehicles and how much from off-road emissions (e.g., con-
struction equipment). For non-vehicular primary sources, coal
combustion, (non-combustion) activities and open burning were the
leading anthropogenic pollution sources, whereas resuspended soil
and sea salt were the major natural sources identified.

Figure 4. Vehicular emission contribution (%) to Figure 5. Non-vehicular primary source contribution
ambient PM2.5 concentrations in Jakarta (%) to ambient PM2.5 concentrations in Jakarta

Vehicular Emission Non-vehicular primary sources

60% 50%

50%
40%
28% 19%
40%
30% 22%
18% 14%
30% 19% 7% 13% 25% 1% 22% 9%
20%
20% 9%
11% 10%
26% 24% 28% 10%
22% 24% 10% 18%
10% 18% 1% 13% 11% 12%
6% 9%
0% 1%
0%
GBK KJ LB GBK KJ LB GBK KJ LB GBK KJ LB
Wet Season Dry Season Wet Season Dry Season

Gasoline vehicle Diesel vehicle Paved road dust Construction Soil

Open burning Sea salt Coal combustion

The contribution of leading sources to ambient PM2.5 concentration

varied across seasons and by location, which may be explained by
variation in local activities, or regional sources of pollution, depending
on weather conditions (e.g., upwind emission from adjacent cities).

During the wet season, exhaust from gasoline and diesel vehicles was
a leading source of air pollution, contributing to 32% to 41% of ambient
particulate level across the city. In central Jakarta (GBK), secondary iii A sensitivity analysis using Positive Matrix
Factorization (PMF) model on the 32 samples
aerosols, sea salt and paved road dust accounted for 11%, 10% and 6%,
collected at Kebon Jeruk site during the wet
respectively. In west (Kebon Jeruk) and east Jakarta (Lubang Buaya), season (under the TCAP project) found that
open burning of biomass or other fuels was another major source of contributions from vehicle exhaust, sea salt,
open burning, construction activities, and
ambient PM2.5 levels (11%), following by secondary aerosols and sea salt.
secondary aerosol were generally in agreement
Construction activities also contributed 13% of PM2.5 level in Kebon with those from the CMB model. In addition, PMF
Jerukiii; whereas coal combustion (14%) was observed to be the second also identified coal combustion mixed with
aged sea salt as anofther contributing source.
leading source in Lubang Buaya after vehicle exhaust. Secondary
aerosols accounted for 6–16% of ambient PM2.5 level, and source of
22% to 31% of PM2.5 mass remained unidentifiable in the wet season.

8 Technical Brief
 Main Sources of Air Pollution in DKI Jakarta

Unlike the wet season, pollution sources in the dry season were fairly
uniform across the city, with vehicle exhaust, sea salt, resuspended
soil being the leading sources. Contribution from exhaust from gasoline
and diesel vehicles across the city was higher in the dry season
(42%–57%) than in the wet season (Figure 4). Sea salt was another
major source of ambient PM2.5 level, contributing 19% to 22%, fol-
lowed by resuspended soil particles (10%–18%). In west Jakarta
(Kebon Jeruk), paved road dust (9%) was also found to contribute to
PM2.5 level. Open burning (9%) remains a major source of polluted air
in east Jakarta (Lubang Buaya). Secondary aerosols accounted for
1–7% of ambient PM2.5 level. Overall, there was <1% to 17% of PM2.5
mass with unidentified origins in the dry season.

Discussion

This is the first study to examine contributing sources to ambient

PM2.5 level in multiple locations with different land use mix across
Jakarta. We found vehicle exhaust, coal combustion, open burning,
construction, road dust, resuspended soil particles and sea salt to be
significant sources of ambient particulate pollution in Jakarta.

This study cannot pinpoint the exact place of emission of these

pollution sources, nor can it determine the relative contributions
from local and regional sources. Knowledge on the emission activities
and locations of point- and area-sources around the sampling sites
and neighborhood area (e.g., Greater Jakarta) may shed light on the
potential emission sources. For instance, open burning identified
during wet season at Kebon Jeruk site may be indicative of biomass
burning in a large area of paddy fields to the west of the sampling site.
Secondary aerosols (e.g., ammonium sulfate and ammonium nitrate)
represent mainly long-range, transboundary contribution to ambient
PM2.5 pollution, and can be controlled by addressing regional emis-
sions of their precursor gaseous pollutants (i.e., sulfur oxides and
nitrogen oxides). Similarly, the identification of coal combustion and
secondary aerosols may be indicative of emission from coal-fired
power plant (CFPP), given that 80% of coal consumption in Indonesia
is for electricity generation, and coal-burning is the main emission
source of sulfur oxides.9 Although there is no CFPP located within
Jakarta’s city boundary, there are clusters of CFPP units east (~20km) References
and west (~55km) of Jakarta. Given that Indonesia plans to construct
9. Ministry of Energy and Mineral Resource. Coal
more CFPPs within 100km of Jakarta, their contribution to ambient Consumption for Power Plants Continues to
PM2.5 pollution in Jakarta will increase over time. Another possible Increase.https://www.liputan6.com/bisnis/
origin of coal combustion may be textile factories that may consume read/3867935/konsumsi-batu-bara-un-
tuk-pembangkit-listrik-terus-meningkat.
coal, though the amount of coal consumed remains unknown.

9 Technical Brief
 Main Sources of Air Pollution in DKI Jakarta

This receptor-based source apportionment study suggests that there

are many leading sources of ambient PM2.5 pollution in and around
Jakarta, and that comprehensive air pollution control measures are
needed to tackle these leading sources in order to improve air quality
in the city. Results from this study should be interpreted with caution
given the small sample size, and limited data on scarcity in known
source profiles for CMB analysis. Future research might include
source-based approach that encompasses emission inventory and
modeling analysis for the Greater Jakarta Area, as well as recep-
tor-based approach with more specific marker types (e.g. molecular
markers that could more effectively identify sources like coal) and
the use of larger datasets.

Acknowledgements

This study was made possible with financial support from Bloomberg
Philanthropies and ClimateWorks Foundation to supplement an
ongoing project of Bandung Institute of Technology funded by the
Toyota Clean Air Project (TCAP).

For further details, including policy recommendations, visit

www.vitalstrategies.org/source-apportionment-report.

10 Technical Brief

View publication stats