定价：５６．００ 元
Tracing Back
the Smog:
Source Analysis and Control
Strategies for PM2.5 Pollution
in Beijing-Tianjin-Hebei
GUAN Dabo, LIU Zhu
China Environment Press
Tracing Back
the Smog:
Source Analysis and Control
Strategies for PM2.5 Pollution
in Beijing-Tianjin-Hebei
GUAN Dabo, LIU Zhu
China Environment Press · Beijing, China
Editorial Board
Editorial Board Members
Guan Dabo/ University of Leeds Liu Zhu/ Kennedy School, Harvard University
Coordinator
Liu Xiaozi
Huang Wei
Wang Qiuxia
Fang Lifeng
Lauri Myllyvirta/ Greenpeace
Photo
Chen Qinggang
Simon Lim
Liu Feiyue
Translation
Xiaozi Liu
Johannes Mauritzen
Lin Ke
Kuang Yin/ Greenpeace
Author Profiles
GUAN Dabo, PhD, is an associate professor at the University of Leeds, and a
senior researcher and director of studies in Economics of St Edmund’s College,
University of Cambridge. He has published around 50 articles in high quality
journals covering the environment, economy and geography, and been frequently
invited for comment in Nature and Science magazines. He has twice won first place
in the Leontief Prize ( 2012, 2013 ) and Top Policy Paper at Environmental Science
& Technology. He was the lead author for the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change ( IPCC ).
LIU Zhu, PhD, is a Giorgio Ruffolo fellow for the Sustainability Science Program at
Kennedy School, Harvard University. His research focuses on China’s sustainable
development and a low-carbon pathway for coping with climate change. He has
authored many academic publications in Nature Climate Change, Proceedings of the
National Academy of Sciences of the United States of America ( PNAS ) and other
journals. He was invited to write an article in Nature, titled ‘Energy policy: A lowcarbon road map for China.'
Executive Summary
II
Executive Summary
Severe air pollution and its associated health impacts
sources, little information is available on trends in
have become of major concern in China, and pollution
PM2.5 concentrations, and answers to the questions
control measures targeting heavily polluted areas
how much air pollution should be reduced and how
are top of the agenda at all levels of government.
to accomplish this reduction remain unclear. The
In September 2013, the State Council issued the
reality is that achieving the proposed PM2.5 target
Airborne Pollution Prevention and Control Action
remains a challenging task, especially considering
Plan ( 2013-17 ), pledging to improve air quality in the
the need for control measures that reflect the various
Beijing-Tianjin-Hebei area ( hereinafter referred to as
characteristics of the region. Therefore Greenpeace
the “Jingjinji” region ), the Yangtze River Delta and
has been co-operating with a team from the University
the Pearl River Delta. Soon afterwards, the Ministry of
of Leeds, UK, led by Dr. Dabo Guan, with the aim to
Environmental Protection ( MEP ) released a detailed
study PM2.5 sources and control strategies in Jingjinji,
implementation plan, aiming to reduce PM2.5 levels
since the end of 2012.
in Jingjinji by 25% and keep PM2.5 concentration in
Beijing from exceeding a level of 60 µg/m3 by 2017.
This project report is the first of its kind to
However, even if PM2.5 is reduced by 25% every five
comprehensively analyze PM2.5 sources in the Jingjinji
years, the National Air Quality Standard Level II of
region and to assess to what extend the region
3
35µg/m will not be achieved until 2030.
should do to reach the air quality targets set by the
MEP. The report aims to provide insight into PM2.5
As clear and concrete pollution reduction goals
pollution in the region, to fuel public debate, and more
at the central and local government levels are
importantly, to inform and influence decision-makers
being set, public concern for a deteriorating living
and stakeholders and provide rationale and support
environment continue to mount. Due to a lack of
for actions that reduce PM2.5 levels.
detailed analysis on PM2.5 composition and emission
Research Methodology
Specifically, this study tries to answer the following questions for the Jingjinji region:
How much are the annual PM2.5 emissions? What sources are contributing to PM2.5 in the region? Which
sectors are emission-intensive?
III
Can we become more ambitious in our aims and meet national standards in a decade instead of by 2030?
To what extend should the region be willing to make changes in order to achieve that? What key measures
are required to reduce emissions?
We have tackled these questions by first compiling a sector and fuel-specific PM2.5 emission inventory ① including
primary sources and precursor gases ( SO2, NOx, NH3 and VOCs ), then plugged this inventory into air quality
models, in order to analyze PM2.5 chemical composition, and evaluate the trends of PM2.5 concentration in different
baseline scenarios. On that basis, we have been able to suggest how much emissions to reduce so that the
national standard can be met by 2022, and recommend a list of effective control measures that are in line with subregional profiles; e.g. structure of the economy, industrial composition, energy mix and emission sources.
The detailed methodology is presented in the following flow chart:
空气质量
GAINS
Air模型
quality
model
GAINS
CMAQ
Air quality
model
PM2.5
Chemical
composition
Baseline PM2.5
concentration
2010
Emission
Inventory
Control
measure
potential
PM2.5 source
apportionment
Primary PM2.5 &
precursor gases
① Our emission inventory covers over 150 sectors and is based on data
from 2010. According to the China Energy Statistical Yearbook (2011 and
2012) and public information released by the MEP, our analysis on sector
apportionment based on the 2010 emission inventory still holds for the
situation in 2011 and 2012.
IV
Main Findings
Efforts to reduce PM2.5 level in the Jingjinji region should consider simultaneous
control of primary sources and precursor gases. Sulfate-Nitrate-Ammonium ( SNA )
aerosols that are transformed from SO2, NOx and NH3 and are the major constituents
[1]
of PM2.5, making up 50%-70% of the total mass concentration. In Beijing, primary
source emissions contribute to 40% of PM2.5 concentration, while precursors contribute
to 60% of the mass concentration. In Tianjin, the percentages are 47% vs. 53%, and
in Hebei 41% vs. 59%, respectively.
Over-reliance on coal is the most important factor of high levels of PM2.5 in Jingjinji. If
we break down the source contribution by fuel type ① , emissions from coal combustion
[2]
constitute 25% of primary PM2.5 emissions and account for 82% of SO2 and 47% of
NOx emissions in that region. In comparison, combustion of oil products accounts for
4% of primary PM2.5, 31% of NOx and 18% of VOCs emissions.
By sector ② contribution , coal-fired power generation is the largest single source
for industrial PM2.5 pollution in Jingjinji with coal-fired power stations emitting 9% of
[3]
primary PM2.5, 69% of SO2 and 47% of NOx. Industrial production of steel, cement
and brick are other major sources of PM2.5 emissions that account for 49% of primary
PM2.5, 12% of SO2 and 17% of NOx in that region.
Contribution by the transport sector to PM2.5 pollution appears more substantial in
[4]
Beijing than in Tianjin and Hebei. That sector accounts for 45% of total NOx emissions
in Beijing. This makes the transport sector the second largest single source of
industrial air pollution in the capital, following the power generation sector;
The iron and steel, cement and brick-making industries take a greater toll on ambient
[5]
air quality of Tianjin and Hebei than of Beijing. In particular, industrial processes in
Hebei province are the most important source of primary fine particles and account for
50% of direct emissions.
To meet the PM2.5 standard level of 35 µg/m3 by 2022, it is necessary to reduce PM2.5
[6]
emissions in the region by 80%, SO2 by 60%, NOx by 75%, NH3 by 85% and VOCs by
90%.
① In this research, we considered five types of fuel for
source apportionment, i.e. coal, oil, gas, biogenic
and non-combustion ( direct emissions or fugitive
emissions ).
② The research divides sector-specific sources into six
subcategories, i.e. transport, industrial processes,
energy ( coal-fired power plants ), commercial and
households, agriculture and fuel production and
others.
V
A Pathway to Blue Skies
If the Jingjinji region is to meet the National Air Quality Standard Level II ( 35 µg/m3 ), it should
focus on the following areas: 1) Limit the use of coal, especially by the utility industry, and ban
all approval of any new coal-fired power plants. In its stead the region should tap into renewable
energy sources; 2) Shutdown polluting and energy-intensive industries such as cement plants
and iron and steel mills, and replace coal-fired boilers with gas-fired ones; 3) Upgrade existing
small-scale boilers for domestic and commercial use by replacing coal-fired boilers with gas-fired
ones, increase proportion of gas consumption in the domestic sector, and ban agricultural waste
incineration; 4) Improve the quality of oil products and emission standards for vehicles.
Based on our analysis of potential control measures, we hereby put forward these suggestions for
reducing PM2.5 emissions in Beijing, Tianjin and Hebei by 2022:
Beijing
Industrial processes are a major source of primary PM2.5, and the precursor gases ( SO2, NOx and
VOCs ) are mainly emitted by the energy and transport sectors. Thus these three sectors should
be the main targets. In terms of emissions by fuel type, combustion of coal and oil products and
non-fuel emissions ( emissions during industrial processes ) are the major sources of gaseous
precursors. Emission measures should concentrate on substantial reduction of coal use,
improving quality of oil products and raising the emission standards for vehicles.
VI
Actions that should be put into place by 2022 include:
[1] Shut down coal-fired
cement plants. Install fabric filters
[4] Escalate the adoption of end-
power plants within the capital
in cement kilns and ban new
of-pipe treatments in the electric
boundary, increase percentage
cement plants;
power sector. Install flue gas
of electricity from renewable
desulfurization and de-nitration
sources in total energy mix, and
[3] Apply the National VI
equipment for all fossil fuel power
boost distributed solar and wind
Emission Standard to light-duty
plants. Use low-NOx burners and
power. Renewable energy can be
gasoline cars and heavy-duty
install fabric filters;
sourced from surrounding areas;
diesel cars, and increase the
percentage of buses and cabs
[5] Reduce VOCs emissions from
[2] Shut down all existing iron
fueled by clean energy to over
industrial processes.
and steel plants and most
40%;
Tianjin
The energy sector ( coal-fired power plants ) is the biggest source of PM2.5 emissions. In terms of fuel contribution,
emissions from coal burning and non-fuel escapement ( fugitive gas escaped from industrial processes ) play a
main role. Efforts should be targeted at reducing fuel consumption and fugitive emissions during production of
oil products.
Actions that should be put into place by 2022 include:
[1] Increase the use of wind
and oil-related products, impose
[4] Expedite the adoption of end-
power/distributed solar power
desulfurization measures, and
of-pipe treatment technologies in
and increase the use ratio,
reduce the emission of VOCs;
the electric power sector, realize
and significantly reduce the
simultaneous desulfurization
percentage of coal-fired power in
[3] Shut down the most polluting
and denitration of flue gas from
energy provision;
cement and steel plants, install
all fossil fuel power plants, use
fabric filters in existing cement
low-NOx burners, install fabric
[2] Tighten control of fugitive
kilns, and ban new cement and
filters, and shut down part of the
emissions during production of oil
steel plants;
existing coal-fired power plants.
VII
Hebei
The energy sector ( coal-fired power plants ) is the biggest PM2.5 emitters among all industrial sectors. The
industrial processes sector is the main source of primary PM2.5 aerosol, however domestic and commercial sectors
also contribute significantly. In terms of emissions by fuel combustion/use, mitigation efforts should focus on the
two main fuel types, namely coal and non-fuel ( from industrial-process sector ).
Actions that should be put into place by 2022 include:
[1] Invest in renewable power
equipment and fabric filters in
filters, and shut down part of the
generation to the maximum
iron and steel plants, and install
existing coal-fired power plants;
extent and use them to replace
fabric filters in cement kilns;
[4] Upgrade existing small-
coal-fired power plants;
[3] Expedite the adoption of end-
scale boilers for domestic and
[2] Accelerate the shutdown
of-pipe treatments in the electric
commercial use by substituting
of the most polluting iron and
power sector with the goal of
gas-fired boilers for coal-fired
steel plants, coking and cement
installing flue gas desulfurization
ones, use more gas instead of
plants and sectors plagued by
and de-nitration technology in
coal as household fuel, and ban
over-capacity problems. Also,
all fuel-fired power plants. Use
agricultural waste incineration.
install flue gas desulfurization
low-NOx burners, install fabric
A breakdown of total emissions by sub-region reveals Hebei emits the largest amount of pollutants. This is
because the region is the most dominated by heavy industries, and also the impacts of small-scale boilers for
residential and commercial purposes are greater in this region. If Hebei fails to reduce the emissions of PM2.5
and precursor gases in a timely and effective fashion, efforts to abate air pollution across the entire region will be
undermined. While Beijing, Tianjin and Hebei are each tackling air pollution on their own, they should also build
regional air pollution prevention and control mechanisms as soon as possible. Hebei province should be given
priority in terms of allocation of resources.
Ultimately, effective control of PM2.5 emissions has to integrate a full set of policy measures. These include
reducing coal use in energy provision, strict enforcement of end-of-pipe treatment technology, monitoring of major
emitting industries, upgrading small-scale boilers, and implementing regional air pollution monitoring and early
warning and emergency response systems.
Table of Contents
Background
1
Methodology
5
2.1 Emission inventories
6
2.2 Baseline scenario design
8
2.3 CMAQ air quality model
9
2.4 Choices of emission
10
reduction measures
PM2.5 source
apportionment
13
3.1 The chemical composition of PM2.5 14
in Jingjinji
3.2 Source analysis of PM2.5
PM2.5 concentrations
under business-as
-usual scenario
29
4.1 Estimation of current PM2.5
30
concentrations
4.2 Trends of PM2.5 concentrations
32
in 2010-2030
Strategies to achieve
PM2.5 standards
in Jingjinji
35
5.1 Reduction potentials of emission 36
policy measures
16
in the Jingjinji region
3.3 Source analysis of PM2.5 in Beijing 19
3.4 Source analysis of PM2.5 in Tianjin
22
3.5 Source analysis of PM2.5 in Hebei
25
5.2 A pathway to blue skies
39
Conclusions
45
Appendices
Appendix 1:
PM2.5 background information
Appendix 2:
The chemical composition of PM2.5
47
49
50
Appendix 3:
A list of sector-specific emission reduction measures
51
Bibliography
55
1
Background
2
Background
Suspended Particulate Matter ( SPM ) is a major type
developing nation, China consumes a large amount
of air pollutant. Subtypes within SPM include Total
of energy and industrial products. It is the largest
Suspended Particles ( TSP; particles with a diameter
primary energy consumer in the world. In 2008-2011,
of 100 micrometers or less ), PM10 ( particles with a
China accounted for 80% of the total global growth in
diameter of 10 micrometers or less ) and PM2.5 ( particles
the use of coal. China's output of industrial products
with a diameter of 2.5 micrometers or less ). PM10
such as cement, steel and glass also accounted for
and PM2.5 that can be easily inhaled into lungs pose
more than half of the global total. Driven by a rapid
considerable health risks.
growth rate of energy consumption and industrial
production, China has become the country with the
Because of their small diameter, toxicity, long
highest PM2.5 levels in the world ( van Donkelaar
suspension time in the air, and the long distances
et al., 2010 ), with the Jingjinji region’s level of fine
they can travel, PM2.5 can have a strong impact on
particles exceeding that of the Sahara. In the winter
human health. Relevant studies show that increases
of 2012, China experienced PM2.5 haze episodes
of PM2.5 level can significantly raise the risk of
with real-time PM2.5 levels in Beijing and some other
lung cancer and other diseases. For instance, for
regions exceeding unprecedented 1,000 µg/m3.
every 10 µg/m3 increase in PM2.5 level, lung cancer,
cardiopulmonary disease mortality rates and low birth
Shocked by the environmental and health impacts of
weight are estimated to increase by 8%, 6% and
PM2.5, the Chinese public is keen to reduce emissions.
10% respectively ( Greenpeace, 2013a ). Another
The government is also committed to resolving
Greenpeace research report ( Pan, Li and Gao, 2012 )
the issue by taking bold action. In May 2010, nine
has revealed that current PM2.5 levels causes tens
ministries including the Minister of Environmental
of thousands of early deaths in Beijing, Shanghai,
Protection jointly released the Guidance of Promoting
Guangzhou and Xi’an annually. According to the
Joint Prevention and Control of Air Pollution and
World Bank’s research, between 2003 and 2006
Improving the Regional Air Quality. In February,
air pollution-induced death and disease cost China
Minister of Environmental Protection released a
1.16%-5.0% of GDP per year. According to a recent
new ambient air quality standard ( GB3095-2012 ),
World Health Organization ( WHO ) research report,
specifying 35 µg/m3 as the limit level of PM2.5. This
atmospheric particulate pollution has become the
was the first time that the state developed a national
fourth largest cause of death in East Asia.
ambient air quality standard for PM2.5. In October
2012, the State Council promulgated the 12th Five-
As the world's most populous country and the largest
Year Plan on Air Pollution Prevention and Control
3
in Key Regions. In December, the Jingjinji region,
of pollutants from primary and secondary sources.
Yangtze River Delta and Pearl River Delta as well as
Primary PM2.5 is emitted directly from combustion and
provincial capital cities established a PM2.5 monitoring
other sources. Secondary PM2.5 is formed from the
network and have been publishing real-time air quality
emission of non-particles ( i.e. precursor gases ) -
monitoring data ever since. In June 2013, the State
such as sulfur dioxide ( SO2 ), nitrogen oxides ( NOx ),
Council announced ten measures to curb air pollution
volatile organic compounds ( VOCs ) and ammonia
and planned to invest RMB1.7 trillion in the next five
( NH3 ) - that turn into PM2.5 in the atmosphere through
years. Three months later, the State Council issued
chemical reactions or condensation ( Appendix 1 ).
the Atmospheric Pollution Prevention and Control
Previous research mainly focused on primary PM2.5
Action Plan, identifying a mix of measures and aiming
emissions without giving due attention to the analysis
to monitor PM2.5 concentrations in real time in 40
of secondary PM2.5 transformed from precursor
cities. The MEP developed a detailed implementation
gases. This report has taken a holistic approach to
plan to curb air pollution in Beijing-Tianjin-Hebei
studying primary PM2.5 and precursor gases in the
and surrounding areas, endeavoring to reduce PM2.5
most polluted Jingjinji region. We first compiled a
levels in Jingjinji by 25% and keep PM2.5 in Beijing
PM2.5 emission inventory specified by sector and by
3
below 60 µg/m by 2017.
fuel type, then applied air quality models - CMAQ and
GAINS - to simulate PM2.5 concentrations in baseline
However, whether those measures are sufficient to
scenarios. The report concludes by providing
win the battle against PM2.5 pollution is still uncertain;
emission reduction policy recommendations based on
meanwhile the public is urging the government to
the results on sector apportionment and on emission
provide a safer living environment earlier rather than
reduction potentials of each policy measure.
later. In a public opinion survey of the government’s
goal of meeting MEP level II air quality standards
This report is the first of its kind to include PM2.5
by 2030, 70% of respondents in the major cities of
source analysis, concentration simulation and
the Jingjinji region were unsatisfied, and 92% of
emission reduction policy recommendations for the
respondents believed that the standard should be met
Jingjinji region. It aims to focus public attention on
by 2020 ( Greenpeace, 2013b ). In practice, immediate
PM2.5 air pollution in the region, and more importantly,
action combating PM2.5 pollution is challenging since
to influence decision-makers and stakeholders and
few sectorial analyses of PM2.5 emissions exist.
to provide rationale and support for reducing PM2.5
levels.
PM2.5 is not a single pollutant, but rather a compound
4
5
Methodology
6
Methodology
Based on the emission inventory, we used multiple instruments including air quality models - Community Multiscale Air Quality Model (CMAQ) and Greenhouse Gas and Air Pollution Interactions and Synergies Model (GAINS)
- to estimate current PM2.5 levels and evaluate concentration changes under different scenarios. Specific research
methods are presented in the following flow chart:
空气质量
GAINS
Air模型
quality
model
GAINS
CMAQ
Air quality
model
2010
Emission
Inventory
PM2.5
Chemical
composition
Baseline PM2.5
concentration
Control
measure
potential
PM2.5 source
apportionment
Primary aerosols &
precursor gases
2.1 Emission Inventories
Although our emissions inventory is based on
A complete list of air emission sources is needed to
contributions and overall trends still hold for 2011 and
get a precise understanding of PM2.5. Factoring in the
2012. According to China Energy Statistical Yearbook
structural characteristics of emission, the research
( 2011 and 2012 ) and information released by MEP,
has developed a detailed emissions inventory that
there are some, but not significant changes in the
covers both primary and secondary sources ( sources
growth of coal and oil demand, nor has there been
of fine particles that are formed in the atmosphere
a change in the rates of industrial coal consumption
from precursor gases SO2, NOx, NH3 and VOCs ) in
and installation rates of end-of-pipe facilities for coal-
2010. The inventory is specified by sector ( a total of
fired boilers ① .
2010 national statistics, our findings on source
150 sectors ) and by fuel type.
① Statistics show that in 2011, coal consumption and oil consumption increased by 10% and 8% respectively. The percentage
of coal consumed by coal-fired power plant increased to 36% as opposed to 35% in 2010. Meanwhile, industry has
consumed a steady 22% of coal over the past two years. Installation rates of desulfurization and de-nitration facilities for
coal-fired power plants have increased by 6% and 10% in 2010-2012.
7
Total PM2.5 emissions are calculated as follows:
PM2.5
EF
Where:
are indices for sector ( j ), fuel ( k ) and pollution control technique ( m );
are the total PM2.5 emissions summed over all j, k and m;
denotes activity level of a certain emission source,
measured in ton of standard coal equivalent;
EF
is PM2.5 emission factor（ kg/kJ ）;
is the penetration rate of a control measure ( % );
is the pollutant removal efficiency of a control measure ( % ).
There are five efficiency levels per sector.
8
2.2 Baseline Scenario Design
Pollution Interactions and Synergies Model
We developed a business-as-usual ( BAU ) scenario
GAINS model is provided by the International Institute
to forecast a trend of PM2.5 concentrations in
for Applied Systems Analysis ( IIASA ) and the model
Jingjinji during 2010-2030 that assumes no further
forecasts the emissions of various pollutants in five-
interventions are taken. The result of our BAU
year intervals. Specifically, the four built-in scenarios
scenario is an average of outputs from four baseline
in the GAINS model are:
( hereinafter referred to as “GAINS model” ). The
scenarios embedded in the Greenhouse Gas and Air
[1]
[2]
UNEP ( United Nations Environment Programme ). This scenario is based on the UN
projection of China’s energy development.
IEA ( International Energy Agency ). This scenario is based on IEA projections of
China’s energy development.
[3]
ECLIPSE ( Evaluating the Climate and Air Quality Impacts of Short-lived Pollutants ).
[4]
TOP ( Tokyo Ozone project ).
Please see the IIASA website ① for detailed assumptions for each of the four scenarios.
① http://www.iiasa.ac.at/
9
2.3 CMAQ Model
used in environmental decision-making. Since June
1998, when the first version was released, the US
The PM2.5 concentrations in this report are from
simulation results using the CMAQ
①
Environmental Protection Agency ( EPA ) has been
model. CMAQ,
or Community Multi-scale Air Quality, is widely
Meteorology
model
( WRF )
Meteorology
-Chemistry
Interface
( MCIP )
Emissions
model
( SMOKE )
constantly developing and updating the CMAQ. We
used version V4.7.1 that was released in June 2010.
Boundary
condition
( BCON )
Chemical
transport model
( CCTM )
Photolysis
rates
( JPROC )
Initial
condition
( ICON )
CMAQ
Source: US EPA
http://www.epa.gov/asmdnerl/
Research/RIA/cmaq.html
① http://cmascenter.org/cmaq/
② The troposphere is the lowest portion of Earth's atmosphere and the
densest layer of the atmosphere. It contains approximately 75% of the
atmosphere's mass and nearly all of its water vapor and aerosols.
10
Many factors influence PM2.5 concentrations. The
CMAQ model sees tropospheric
①
air as a whole, and
WRF ( Weather Research and Forecasting Model )
is used to simulate meteorological fields that are
takes into account meteorological, topographic and
necessary to the operation of CMAQ. To ensure
surface conditions, as well as physical and chemical
accuracy of the boundary meteorological fields,
actions that occur during emission, transmission and
each horizontal boundary in the WRF simulation
diffusion of pollutants. The CMAQ model has five
domain has three more grids than that in the CMAQ
major modules ( see the boxes in the above diagram ).
simulation domain. On the vertical plane, there are
CCTM is the key module used to simulate process of
23 σ layers and the atmospheric pressure of the
atmospheric pollutant transport, chemical conversion
top layer is 1hPa. MODIS ( or Moderate Resolution
and aerosol settlement.
Imaging Spectroradiometer ) land use data are
amassed to calculate topographic and Earth’s surface
The simulation model is configured as follows. The
data.
vertical domain from the Earth’s surface to the top of
the troposphere is divided into 14 unequal layers ( the
atmospheric pressure of the top layer is 1hPa ).The closer
to the surface of the Earth, the more concentrated the
2.4 Choices of Emission
Reduction Measures
layers are. The model has triple-nested domains: the
To meet the National Air Quality Specification II
outer domain, with a horizontal grid spacing of 36 km,
of 35 µg/m3 by 2022, this research has developed
covers the entire East Asia region including China,
a comprehensive policy package to curb PM2.5
North Korea, South Korea and Japan; the 12 km grid-
emissions in the Jingjinji region. Compilation of the
spacing inner domain covers the relatively developed
package is based on the sensitivity test of each
eastern regions in China; the innermost domain
potential intervention policy that is embedded in the
covering Jingjinji and its surrounding areas has the
GAINS-city model ( Liu et al., 2013 ). The intervention
highest horizontal grid resolution of 4 km. The outputs
policy, specified by sector, covers measures
from the outer domains are used to provide boundary
necessary to reduce emissions of primary PM2.5 as
conditions for the inner ones by one-way nesting.
well as precursor gases that form secondary PM2.5
The outermost BCON fields are based on simulation
through a series of chemical reactions and physical
results of GEOS-Chem
②
.
processes. Broadly, the identified control policy can
be summarized as follows:
① The troposphere is the lowest portion of Earth's
atmosphere and the densest layer of the atmosphere.
It contains approximately 75% of the atmosphere's
mass and nearly all of its water vapor and aerosols.
② http://geos-chem.org/
11
• Substitute clean energy sources for coal-based
electricity: coal burning is a major source of the
• Adopt more effective end-of-pipe technology, e.g.
installation of fabric filters to collect dust;
two PM2.5 precursor gases SO2 and NOx. The use
of natural gas, wind power and PV to replace coal
would substantially reduce PM2.5 emissions;
• Improve production technology, e.g. during cement
production, replace shaft kiln production lines with
dry cement manufacturing production lines;
• Promote energy cascade use. Recycle industrial
waste energy, increase the percentage of thermal
• Limit and phase out energy-intensive industries.
plants based on combined heat and power,
Phase out and shut down pollution-intensive
and replace distributed household heating with
cement plants and iron and steel plants in the
centralized heating systems;
Jingjinji region;
• Improve the energy efficiency per unit of industrial
• Adopt stricter environmental protection standards.
production. Relevant research shows that the
For example, apply the national V and VI emission
energy efficiency of industrial production in China
standard for vehicles, and phase out all vehicles
is 30% lower than levels in developed Western
that are below the national II emission standards.
countries ( IEA, 2009 );
We have selected 79 top policy measures ( see Appendix 3 for a complete list ), based on their reduction potential
performance. The reduction potential of each policy measure is evaluated according to the following equation:
Where:
= types of pollutants, including primary PM2.5 and PM2.5 precursor gases
like SO2, NOx, NH3 and VOCs.
= total emissions of pollutant i under a business-as-usual scenario.
= emission reduction of pollutants resulting from a specific control measure.
12
13
PM2.5 Source
Apportionment
14
PM2.5 Source
Apportionment
In this section, we first show the chemical
emissions of precursor gases, i.e. SO2, NOx, NH3
composition of PM2.5 and the contributions of primary
and VOCs. CMAQ simulation results for PM2.5
and secondary sources to the total concentration
concentration reveal that in 2010 direct emissions
based on simulation results. We then present a study
of PM2.5 and precursor emissions are responsible
of primary PM2.5 emissions and PM2.5 precursor gases
for 40% and 60% of the total PM2.5 concentration in
( transformed into PM2.5 via chemical reactions ) to
Beijing, respectively. In Tianjin, the percentage is
trace the emission sources.
47% against 53%, and in Hebei, 41% against 59%.
This clearly indicates that in order to reduce PM2.5
3.1 The Chemical Composition
of PM2.5 in Jingjinji
concentrations, emission reductions of primary PM2.5
and PM2.5 precursor gases have to run in parallel.
PM2.5 comes from direct emission of fine particles,
and also aerosols transformed from the direct
Figure 3-1: Contributions to PM2.5 mass concentration by NO3- and SO42- (as percentage of concentration)
15
Differences in industrial structure and the natural
Jingjinji, making up 50%-70% of the total mass
environment can explain variations in PM2.5
concentration ( Figure A1 in Appendix 2 ). The
composition. In the following, we will focus on the
concentrations and compositions of SNA aerosols
two major PM2.5 chemical species: carbonaceous
are determined primarily by their precursor emission
aerosols ( from direct emissions and secondary
such as SO2, NOx, and NH3. A breakdown of SNA
transformation ) and inorganic aerosols ( Sulfate-
composition shows that sulfate ( SO42- ) constitutes
Nitrate-Ammonium or SNA ).
30%-33% PM2.5 mass concentration ( Figure 3-1 ),
while nitrate ( NO3- ) comprises 22%-24% and
Carbonaceous aerosol is comprised of Black
ammonium ( NH4+ ) 15%-16%.
Carbon ① ( BC ) and Organic Carbon ( OC ). Our
model indicates that BC is only responsible for 5%-
SNA exists in forms of (NH4)2SO4 or NH4NO3, but
9% of PM2.5 concentration, as shown in Figure A1 in
the former is more stable than the latter. Under most
Appendix 2. BC resulting from incomplete combustion
conditions, NH3 reacts preferentially with SO2 to form
of fossil fuels or biogenic substances is often used
ammonium sulfate ( Damberg, 2007 ) that stays in
as an indicator to measure the scale of vehicle-
the atmosphere for a long time and travels to other
induced pollution ( Huang et al. 2006 ). The mass
regions, causing regional pollution ( Zhang, 2011 ).
concentration of BC in Beijing downtown area is the
NH4NO3 is unstable and its formation is constrained
highest in the Jingjinji region, which is in part due to
strongly by temperature ( Wang et al., 2006 );
higher car ownership numbers in Beijing. With respect
e.g. the low temperature in winter is favorable for
to OC, Hebei has the highest OC mass concentration
NH4NO3 formation. If control efforts focused solely
of 20%-22% ( see Figure A1 in Appendix 2 ). OC is
on SO2 reduction, the freed NH3 in the atmosphere
both a primary and secondary pollutant compound.
would form ( if conditions allow ) two units of
Primary Organic Aerosol ( POA ) is directly emitted
NH4NO3 instead of one unit of (NH4)2SO4. In the
from fuel combustion and other sources, and
short run, it might result in a temporary rise of PM2.5
Secondary Organic Aerosol ( SOA ) is transformed
concentration ( Wang et al., 2013 ). In other words,
from semi-volatile organic compounds.
emission reduction efforts should aim at simultaneous
prevention and control of multiple pollutants.
Sulfate-Nitrate-Ammonium ( SNA ), inorganic aerosol,
is the single most important PM2.5 component in
① Black Carbon (BC) is also known as Carbon Black (CB) or Elementary
Carbon (EC).
16
3.2 Source Analysis of PM2.5 in
the Jingjinji Region
3.2.1 Contributions by Region
Our emission inventory indicates that in 2010 over 10
million tons of primary PM2.5. A breakdown of
million tons of primary PM2.5 and its precursors were
emissions shows that Beijing, Tianjin and Hebei
emitted into the atmosphere in the Jingjinji region.
emitted 130 kilotons, 140 kilotons and 1.3 million tons
By region, Hebei is the largest emitter. By sector,
respectively.
In 2010, the Jingjinji region emitted a total of 1.6
the cement, iron, steel and brick-making industries
as well as small-scale boilers for domestic and
Hebei province is not only the main emission source
commercial use are the major emitters of primary
for primary fine particles, but also their precursor
PM2.5. These sectors contribute to 80% of primary
gases. Total emissions of SO2 were about 3.5 million
PM2.5 in the region. As for precursor gases, coal-fired
tons with Beijing, Tianjin and Hebei responsible for
power plants are the major emitters, responsible for
9%, 15%, and 77% of them respectively. The total
69% and 47% of SO2 and NOx in the region. By fuel,
emissions of NOx were about 2.2 million tons with
coal burning is the major source of PM2.5, making up
Beijing, Tianjin and Hebei responsible for 12%, 13%
25% of primary PM2.5, and 82% of total emissions of
and 74%. Emissions of VOCs were about two million
SO2 and 47% of NOx.
tons in total, among which Beijing, Tianjin and Hebei
NOx
Figure 3-2: Total emissions of primary PM2.5 and precursors gases in Jingjinji, 2010
17
emitted 330 kilotons, 290 kilotons and 1.3 million
in the region, responsible for 48% of primary PM2.5
tons. Aggregate emissions of NH3 stood at one million
emissions, 12% of SO2 and 17% of NOx emissions.
tons approximately, with Beijing emitting 60 kilotons,
Tianjin 50 kilotons, and Hebei 900 kilotons.
Heating for households and businesses is the third
biggest source of pollution in Jingjinji, emitting 32%
3.2.2 Contributions by Sector
of primary PM2.5 emissions, 14% of SO2, 6% of NOx
emissions and 25% emissions of VOCs.
In terms of total emissions, coal-fired power
generation is the biggest source of pollution in
Although the transport sector emits relatively few
Jingjinji, responsible for 9% of primary PM2.5, 69% of
primary PM2.5 and specific pollutants like SO2, they
SO2 and 47% of NOx emissions.
discharge 29% of NOx and 14% of VOCs.
Industrial processes for making steel, iron and
bricks are the second biggest source of pollution
NOx
VOCs
Figure 3-3: Total emissions of primary PM2.5 and precursor gases by sector in Jingjinji, 2010
18
3.2.3 Contributions by Fuel Type
for 25% of primary PM2.5, 82% of SO2 and 47% of
NOx emissions. As for non-combustion emissions,
Four combustible fuel types have been considered
these account for 55% of primary PM2.5 in the Jingjinji
in our study: coal, oil, gas and biogenic substances.
region, 13% of SO2, 16% of NOx emissions, 98%
Emissions from non-combustion processes ( e.g.
of NH3 emissions and 53% of VOCs emissions. Oil
fugitive emissions, also particles from industrial
burning makes up 4%, 31% and 18% of primary
processes ) are classified as non-fuel or non-
PM2.5, NOx and VOCs emissions, respectively.
combustion, the fifth fuel type.
Burning biogenic materials results in 15% of primary
PM2.5 emissions and 19% of VOCs emissions.
We have shown that coal burning is the leading
source of pollution in the Jingjinji region, accounting
NOx
VOCs
Figure 3-4: Total emissions of primary PM2.5 and precursor gases by fuel type in Jingjinji, 2010
19
3.3 Source Analysis of PM2.5 in
Beijing
3.3.1 Total Emissions
Our results show that coal burning is the largest
and PM2.5 precursors, including 300 kilotons of SO2,
source of PM2.5 pollution in Beijing, followed by oil
270 kilotons of NOx, 60 kilotons of NH3 and 330
combustion from the transport sector. The sector-
kilotons of VOCs.
In 2010, Beijing emitted 130 kilotons of primary PM2.5
wide results further indicate that energy ( thermal
power generation ), transport, cement-making and
steel-making are major polluting sectors in Beijing.
NOx
VOCs
Figure 3-5: Emissions of primary PM2.5 and precursor gases by sector in Beijing, 2010
20
NOx
Figure 3-6: Contributions of the industrial processes broken down by sub-sector in Beijing, 2010
3.3.2 Emissions by Sector
Although the energy sector only accounted for 11%
of primary PM2.5, it was a key emitter of gaseous
In terms of the directly emitted PM2.5, energy sector
precursors, causing 56% of Beijing’s SO2 and 38%
( thermal power generation ) contributed the most
of NOx emissions ( Figure 3-5 ). Fuel consumption
in Beijing, followed by emissions from the transport
by transport sector is the largest source of Beijing’s
sector. The industrial processes sector ( e.g. cement-
NOx ( 45% ).
making, coking, and metal processing sectors and
smelters ) contributed 41%, making it the top source
The emissions of NH3 in Beijing are largely caused
of primary PM2.5 in Beijing. The use of small-scale
by agricultural practices ( e.g. animal husbandry)
boilers ( with capacity below 50MW ) for households
and fertilizer production. In contrast, sources of
and commercial sector shares 32% direct emissions
volatile organic compounds ( VOCs ) emissions
of fine particles.
are diversified; e.g. fugitive emissions during oil
21
NOx
VOCs
Figure 3-7: Emissions of primary PM2.5 and precursor gases by fuel type in Beijing, 2010
transportation, leaking from small coke ovens for
77% of SO2 were formed from coal burning ( mainly
domestic and commercial use, and emissions during
for coal-fired power generation, but also used in
processes.
small-scale boilers for domestic and industries ); oil
and coal burning were the key sources of NOx ( 53%
3.3.3 Emissions by Fuel Type
vs. 42% ).
Emissions from coal burning are the most important
In contrast to SO2 and NOx that are formed due to
source of PM2.5 pollution in Beijing. Of primary PM2.5,
fuel combustion, NH3 and VOCs are largely caused
non-combustion (particles formed during industrial
by fugitive emissions ( i.e. non-combustion ).
processes for making cement, glass, steel and
bricks) and coal burning played a leading role ( 47%
vs. 37% ).
22
3.4 Source Analysis of PM2.5 in
Tianjin
Tianjin had a total of 29 coal-fired power plants, which
consumed 25 million tons of coal or 52% of the total
coal usage in Tianjin.
Based on our analysis on PM2.5 chemical composition
and sector apportionment, the main source of PM2.5
3.4.1 Total Emissions
pollution in Tianjin is coal. In contrast, coal plays
a much bigger role in Tianjin than in Beijing. The
In 2010, the discharge of primary fine particles was
energy sector ( coal-fired power plants ) consumes
140 kilotons, whereas emissions by the precursor
the greatest amount of coal and is thus the major
gases was 510 kilotons for SO2, 300 kilotons for NOx,
contributor of PM2.5 pollution in Tianjin. Statistics from
50 kilotons NH3 and 290 kilotons for VOCs.
the China Electric Yearbook also reveal that in 2010,
NOx
VOCs
Figure 3-8: Emissions of primary PM2.5 and precursor gases by sector in Tianjin, 2010
23
NOx
VOCs
Figure 3-9: Contributions of the industrial processes broken down by sub-sector in Tianjin, 2010
3.4.2 Emissions by Sector
The energy sector ( coal-fired power generation ) is
In sub-sectioning the industrial processes sector of
the biggest contributor to PM2.5 pollution ( including
Tianjin, we have found that production of oil products,
primary emissions and secondary transformations ) in
cement and iron and steel are the major sources of
Tianjin. In 2010, the energy sector discharged 83% of
precursor gases.
the city's SO2 and 64% of NOx, becoming the biggest
emitter in both of the precursors, whereas in Beijing,
NOx from the transportation sector is the major source
of emissions.
24
3.4.3 Emissions by Fuel Type
Tianjin is similar to Beijing in source composition
for NH3, but is 13% higher than Beijing in the VOCs
In terms of fuel, coal is the dominant source for SO2
emitted through burning of biogenic substances ( e.g.
and NOx in Tianjin, accounting for 91% and 61% of
combustion of straws ). According to a survey ① on
the two respectively ( Figure 3-10 ).
the urbanization rate in Chinese cities, the respective
rate of Beijing is 18% higher than that of Tianjin in
2010.
NOx
VOCs
Figure 3-10: Emissions of primary PM2.5 and precursor gases by fuel type in Tianjin, 2010
① http://www.ciudsrc.com/bbs/viewthread.php?tid=8805
25
3.5 Source Analysis of PM2.5 in
Hebei
3.5.1 Total Emissions
Both chemical composition and sector contributions
(Mts) of primary fine particles and precursor gases,
suggest that the main source of PM2.5 in Hebei is coal,
which are six times that of Tianjin and seven times
a result similar to that of Tianjin. Broadly speaking,
that of Beijing. The pollutants can be ranked in order
the energy sector ( coal-fired power generation ),
of size of emissions: SO2 ( 2.7 Mts ), NOx ( 1.7 Mts ),
industrial production and residential and commercial
VOCs ( 1.4 Mts ), primary PM2.5 ( 1.3 Mts ) and NH3
sectors are the top three contributors of PM2.5.
( 1 Mts ).
In 2010, Hebei emitted a total of eight million tons
NOx
VOCs
Figure 3-11: Emissions of primary PM2.5 and precursors gases by sector in Hebei, 2010
26
3.5.2 Emissions by Sector
The contribution of industrial processes sector
( e.g. production of cement, lime, coke and bricks ) to
The energy sector ( coal-fired power generation )
Hebei’s emissions is far greater than that in Beijing
is the largest contributor to primary pollutants and
and Tianjin. Take NOx as an example, NOx emissions
precursors in Hebei, accounting for 67% and 46% of
from industrial processes are equivalent to that of the
SO2 and NOx respectively. Statistics from the China
transportation sector, accounting for 20% of the total
Electric Power Association indicates that in 2011,
NOx and higher than that of Beijing ( 9% ) and Tianjin
there were a total of 153 coal-fired power plants in
( 13% ). Moreover, industry processes is also the
Hebei, which make up 73% of the installed capacity
major source of primary PM2.5 emissions in Hebei
of thermal power generation in Beijing, Tianjin
( 50% of the total ).
and Hebei combined. Coal used in thermal power
generation is the main source of SO2 emissions.
We have found that the iron and steel sector is
Coal-based heating for households and businesses
the biggest industrial contributor to primary and
has contributed an additional 16% of SO2 emission.
secondary emissions of PM2.5 in Hebei, followed by
cement production.
NOx
VOCs
Figure 3-12: Contributions of the industrial processes sector broken down by sub-sector in Hebei, 2010
27
3.5.3 Emissions by Fuel Type
Agriculture plays a greater role in Hebei's economy
Coal is the most important fuel type in Hebei. Similar
than in Beijing or Tianjin. This is reflected in fuel
to Beijing and Tianjin, coal burning in Hebei is the
composition of VOCs emissions in Figure 3-13, which
top contributor to SO2 emissions, accounting for 81%
indicates that 24% of VOCs comes from biogenic
of the total discharge. Coal burning also contributes
sources such as agricultural residues, fugitive
to 24% of the directly emitted PM2.5 and 45% of the
methane emissions and fuel wood combustion.
precursor gas NOx.
NOx
VOCs
Figure 3-13: Emissions of primary PM2.5 and precursor gases by fuel type in Hebei, 2010
28
29
PM2.5 Concentrations under
a Business-as-usual
Scenario
30
PM2.5 Concentrations under
a Business-as-usual
Scenario
A business-as-usual scenario predicts the trend of
time-dependent. PM2.5 in this report refers to annual
fine particles emissions according to the current
average concentration. There are no official statistics
status, excluding possible further interventions. This
regarding annual average PM2.5 concentrations
chapter presents results from simulations of the
in 2010 in Jingjinji, but we managed to produce
current PM2.5 concentrations and predicts the trend
estimates based on the comparison between the
for the next 20 years.
model outputs and the actually measured values.
4.1 Estimation of Current PM2.5
Concentrations
The following is a description of methods often
used in the scientific literature to determine
PM2.5 concentrations:
The concentration of PM2.5 is highly regional and
Assume a fixed conversion rate between PM10 and PM2.5 and deduce the concentration of the
[1]
latter based on the measured concentration of the former. The MEP has applied such a method
by assuming a fixed conversion ratio of 0.65 ( see Table 4-1 ). While this method is simple, the
estimates are subject to a high degree of uncertainty ( Brauer et al., 2012 ).
From instrumental measurements. This method can accurately measure the PM2.5 concentration
[2]
in a given area, but the geographic distribution of monitoring stations is heavily biased. Only from
late 2012, key regions in China including Jingjinji began to publish real-time PM2.5 monitoring
data.
Estimates from air quality models, including the CMAQ and GAINS models used in this research.
[3]
Simulation results are sensitive to emission inventory and other factors, thus errors are expected
compared to the actual measurements.
Satellite-derived PM2.5. This method is based on satellite observation of Aerosol Optical Depth
[4]
( AOD ) to calculate ground-level concentrations of PM2.5. The AOD data provided by NASA is
only partially accessible to the public. The latest AOD data gives average PM2.5 concentrations
from 2001 to 2006 ① .
① Please see: http://www.nasa.gov/topics/earth/features/health-sapping.html
31
Table 4-1: PM2.5 concentrations in 2010 estimated from PM10 by the MEP ( µg/m3 )
PM10 concentration
Estimated PM2.5 concentration
Beijing
124
80.6
Tianjin
105
68.25
Shijiazhuang, Hebei
105
68.25
This research integrates the four methods above by verifying the simulated results from CMAQ against actual
measured values ( where possible ) and satellite-derived values. In Table 4-2, we present PM2.5 concentration
weighted by population ① and the concentration averaged over urban areas ( 20 km x 20 km ). The detailed results
for Beijing, Tianjin and Hebei are summarized in the same table.
Table 4-2: Average PM2.5 concentrations in Beijing, Tianjin and Hebei, 2010 ( µg/m3 )
PM2.5 concentration
( population-weighted )
Data sources
PM2.5 concentration
( urban average,
20 km x 20 km )
Data sources
Beijing
73.7
Brauer et al.
( 2012 )
89.2
Observed values from
Beijing Environ. Monitoring
Center & U.S. Embassy in
China
Tianjin
76.0
CMAQ
simulation
91.7
CMAQ results for
urban Tianjin
Hebei
80.1
CMAQ
simulation
112.2
CMAQ results for
urban Shijiazhuang
① The population-weighted mean concentration is acquired by multiplying the PM2.5 concentration in each grid
cell with the respective population ratio, and summing up over all grids.
32
In the table above, the average weighted by
population and urban average in Tianjin and Hebei
come from the simulation results of the CMAQ
4.2 Trends of PM2.5
concentrations in 2010-2030
modeling. Compared to the actual measurements,
With the GAINS model provided by IIASA, we
the simulation results for Beijing appear too high
have made predictions of fine particles ( PM2.5 )
( GAINS results ) or too low ( CMAQ results ).
concentration trends from 2010 to 2030 in Beijing,
In view of that, when calculating the population-
Tianjin and Hebei under the business-as-usual
weighted mean concentration for Beijing, we have
scenario.
adopted the value from Brauer et al. ( 2012 ), a
research project contributing to the Global Burden
In predicting fine particles concentrations, we need to
of Disease 2010 Project. In terms of urban average
make assumptions on future economic development,
concentration in Beijing, we have referred to the
population growth, energy demand and supply and
released measurements from the Beijing Municipal
other factors. The GAINS model has four built-
Environmental Monitoring Center and the U.S.
in baseline scenarios ( UNEP, IEA, ECLIPSE and
Embassy in Beijing.
TOP, see section 2.2 for more details ). Because the
assumptions for each of these scenarios vary, their
results on the trends of PM2.5 concentrations can
differ. Attaching equal importance to the results from
the above four baseline scenarios, we have predicted
the average trends of PM2.5 concentrations from 2010
to 2030 in Beijing, Tianjin and Hebei ( see Figure 4-1 ).
33
PM2.5
Figure 4-1: Average trends of PM2.5 concentration change during 2010-2030 in a business-as-usual scenario
Figure 4-1 shows that without further interventions,
Clearly, we cannot rely solely on the business-as-
the concentration of PM2.5 will decline gradually.
usual scenario to reach the national target of reducing
Beijing will witness a drop of 10% in PM2.5
the PM2.5 concentration level by 25% by 2017. Nor will
concentration in 2030 compared with the 2010
this non-intervention scenario take us to the National
level. The drops in Tianjin and Hebei are 13% and
II Standard, reducing fine particle concentration to
11% respectively. This decreasing trend is due to
35µg/m3 in the next two decades. Therefore, we must
the increasing role of the service industry in the
commit to doing more. The State Council has recently
economy, the growth of clean energy as spelled out
released its action plan for air pollution prevention
in the national energy planning and the expectation
and control. This marks a great leap forward in terms
that outdated production capacity will be shut down or
of escalating intervention efforts in addition to those
eliminated in the Five Year Plan.
assumed in the business-as-usual scenario.
34
35
Strategies to Achieve
PM2.5 Standards
in Jingjinji
36
Strategies to Achieve
PM2.5 Standards
in Jingjinji
The most severe air pollution in China is found in
Beijing, Tianjin and Hebei, and curbing PM2.5 levels
is high on the government agenda. According to
5.1 Reduction Potentials of
Emission Policy Measures
the Atmospheric Pollution Prevention and Control
A comprehensive policy package for emission
Action Plan issued by the State Council in September
reduction requires a careful policy evaluation on the
2013, by 2017, the PM2.5 concentration in Jingjinji
possible policy measures. Based on our analysis on
shall be reduced by 25%, and the annual average
sector apportionment and review of relevant literature
concentration of PM2.5 in Beijing shall be kept below
( e.g., Liu et al., 2013 ), we have identified 79
3
60 µg/m . In an implementation plan endorsed
measures that have high potential for reducing PM2.5.
by six ministries, a bold proposal to curb coal
These measures are distributed across domestic
consumption has been put forward for the first time.
sector, industrial combustion sector, industrial
By 2017, Beijing will reduce 13 million tons of coal
process, power plants and transport sector ( see
consumption, Tianjin 10 million tons, and Hebei 40
Appendix 3 for a complete list ). If we were to apply all
million tons.
these 79 measures in Jingjinji to achieve the 35µg/m3
target by 2022, this would mean a reduction in total
Current policies call for a 25% reduction in PM2.5
PM2.5 emissions of 50% by 2017, and furthermore, a
level, but this is still far from reaching the National
reduction of 80% of the directly emitted PM2.5, 60% of
3
II Air Quality Standards ( 35 µg/m ). A substantial
SO2, 75% of NOx, 85% of NH3 and 90% of VOCs by
gap exists between the status quo and public
2022.
expectations, and moreover, current measures have
significant limitations. With the aim of achieving
3
For Beijing, Liu et al. ( 2013 ) have conducted
the 35µg/m target in the coming decade, we have
sensitivity analysis ( in Figure 5-1 ) on the emission
proposed an alternative policy scheme in the following
reduction potentials of key measures ( either as single
sections.
measure or combination of several measures ). They
show that more active measures can reduce primary
PM2.5 emission in Beijing by over half by 2022.
37
NOx
EE stands for energy efficiency, HDDV for heavy-duty diesel
vehicles, LDDV for low-duty diesel vehicles and CFPP for coal-fired
power plants. ( Liu et al., 2013 )
Figure 5-1: Top 15 measures in Beijing by the percentage of pollutants that can be reduced
from its total emissions after applying each measure
38
Although we lack similar sensitivity analysis of policy
electricity supply from external sources. However, for
measures at the Jingjinji regional level, this Beijing-
Hebei this would mean an increase in the percentage
focused study ( Liu et al., 2013 ) is highly relevant
of renewable energy in its energy mix. Similarly,
and instrumental in reducing PM2.5 emission in
shutting down iron and steel plants would be much
Tianjin and Hebei. One justification is that Beijing
more difficult in Hebei than that in Beijing.
enjoys considerable technological advantages over
Tianjin and Hebei. If these emission measures were
We emphasize that substantial financial and policy
to be applied in Hebei and Tianjin, we could expect
support are needed to achieve the 2022 goals. This
greater reduction potentials to be achieved there
report evaluates policy measures primarily from the
as well. Nevertheless, the extent to which each
perspective of technical reduction potentials, with no
measure reduces emission will vary by city due to
consideration given to financial or economic matters.
different PM2.5 compositions. Both the design and the
Some of the proposed measures could have social
implementation of policy measures have to be tailored
implications. For instance, if the energy demands in
to local conditions. For instance, the measure of “85%
the Jingjinji region are met by shutting down power
electricity from outside” in Figure 5-1, in essence,
plants in the region and transmitting electricity from
is about the reduction of electricity generated by
elsewhere, then we are not solving the problem, but
coal-fired power plants. For Beijing, owing to its
rather shifting the pollution from one place to another.
own resource constraints, this would mean seeking
39
5.2 A Pathway to Blue Skies
If the Jingjinji region is to meet National Air Quality Specification Level II ( 35 µg/m3 ), it should focus on the
following areas:
Limit the use of coal, especially by utility industry, and halt the building of any new
[1]
coal-fired power plants. Instead, the region should replace coal-fired power generation
by renewable energy.
[2]
Shut down and rectify pollution and energy-intensive industries such as cement plants,
and iron and steel plants, and replace coal-fired boilers with gas-fired ones.
Upgrade existing small-scale boilers for domestic and commercial use by replacing
[3]
coal-fired boilers with gas-fired ones, increase the proportion of gas consumption in the
domestic sector, and ban agricultural waste incineration.
[4]
Improve the quality of oil products and emission standards for vehicles.
40
Based on our analysis of sector apportionment and evaluation of policy measures, we put forward these targeted
suggestions for Beijing, Tianjin and Hebei to achieve PM2.5 standards by 2022:
Beijing
Industrial processes are the major source of primary PM2.5, and the precursor gases ( SO2, NOx and VOCs ) are
mainly emitted by the energy and transport sectors, thus these three sectors should become the main targets.
In terms of emissions by fuel type, combustion of coal and oil products and non-fuel emissions ( emissions
during various industrial processes ) are the major sources of gaseous precursors. Emission measures should
concentrate on substantial reduction of coal use, improving quality of oil products and raising the emission
standards for vehicles.
Actions that should be put into place by 2022 include:
[1] Shut down coal-fired power plants within the
[4] Escalate the adoption of end-of-pipe solutions
capital boundary, increase the percentage of
in the electric power sector. Install flue gas
electricity from renewable sources in total energy
desulfurization and denitration equipment for all fossil
mix, and boost the development of distributed solar
fuel power plants. Use low-NOx burners and install
and wind power. Source renewable energy from the
fabric filters.
surrounding area.
[5] Reduce the emission of volatile organic
[2] Shut down all existing iron and steel plants and
most cement plants. Install fabric filters in cement
kilns and cease the building of new cement plants.
[3] Apply the National VI Emission Standard to lightduty gasoline cars and heavy-duty diesel cars, and
increase the percentage of buses and cabs fueled by
clean energy to over 40%.
compounds ( VOCs ) from industrial processes.
41
Tianjin
The energy sector ( coal-fired power plants ) is the biggest source of PM2.5 emissions in Tianjin. In terms of
fuel contribution, emission from coal burning and non-fuel escapement ( fugitive gases escaped from industrial
processes ) play the main role. Efforts should be targeting reducing fuel consumption and fugitive emissions during
the production of oil products.
Actions that should be put into place by 2022 include:
[1] Increase the use of wind power and distributed
[3] Shut down the most polluting cement and steel
solar power and their shares in the energy mix, and
plants, install fabric filters in existing cement kilns,
significantly reduce the percentage of coal-fired
and ban new cement and steel plants.
power in energy provision.
[4] Expedite the adoption of end-of-pipe technologies
[2] Tighten the control of fugitive emissions during
in the electric power sector, realize simultaneous
production of oil and oil-related products, impose
desulfurization and denitration of flue gas from all
desulfurization measures, and reduce the emission of
fossil fuel power plants, use low-NOx burners, install
VOCs.
fabric filters, and shut down part of the existing coalfired power plants.
42
Hebei
The energy sector ( coal-fired power plants ) is the biggest PM2.5 emitters among all industrial sectors in the Hebei
province. The industrial processes sector is the main source of primary PM2.5 aerosol, however domestic and
commercial sectors also contribute significantly. In terms of emissions by fuel combustion/use, mitigation efforts
should focus on the two main fuel types, namely coal and non-fuel ( from industrial processes sector ).
Actions to put in place by 2022 include:
[1] Invest heavily in renewable power generation and
the electric power sector with the goal of installing
use them to replace coal-fired power plants.
flue gas desulfurization and de-nitration technology
in all fuel-fired power plants. Use low-NOx burners,
[2] Accelerate the shutdown of the most polluting iron
install fabric filters, and shut down part of the existing
and steel plants, coking plants and cement plants,
coal-fired power plants.
and sectors plagued by over-capacity problems.
Install flue gas desulfurization equipment and fabric
[4] Upgrade existing small-scale boilers for domestic
filters in iron and steel plants, and install fabric filters
and commercial use by substituting gas-fired
in cement kilns.
boilers for coal-fired ones, use more gas instead of
coal as household fuel, and ban agricultural
[3] Expedite the adoption of end-of-pipe solutions in
waste incineration.
43
A breakdown of total emissions by sub-region reveals
own, they should also build regional mechanisms
Hebei emits the largest amount of pollutants. This
of air pollution prevention and control as soon as
is because it hosts the largest part of the heavy
possible. Hebei province should be given priority in
industries, and also because the impacts of small-
terms of allocating resources. Ultimately, effective
scale boilers for residential and commercial purposes
control of PM2.5 emissions has to integrate a full set
are greatest in this region. If Hebei fails to reduce the
of policy measures. These include reducing coal use
emissions of PM2.5 and precursor gases in a timely
in energy provision, strict enforcement of end-of-
and effective manner, the efforts to abate air pollution
pipe technology, upgrading small-scale boilers, and
of the entire region will be undermined. While Beijing,
implementing regional air pollution monitoring and
Tianjin and Hebei are tackling air pollution on their
early warning and emergency response systems.
44
45
Conclusions
46
Conclusions
Based on the sector-specific and fuel-specific
to reduce total emissions by over 80%, Beijing,
emission inventory embedded in the GAINS model,
Tianjin and Hebei have to fundamentally adjust their
this research simulates PM2.5 concentrations in
industrial structure and energy mix, substantially
Beijing, Tianjin and Hebei, using the CMAQ air quality
decrease the ratio of heavy industries, replace
model. We have also analyzed the reduction potential
coal with clean energy in the electric power sector
of various policy measures targeted at achieving
and adopt efficient and coordinated end-of-pipe
3
35µg/m PM2.5 level by 2022.
technologies.
PM2.5 emissions in Beijing, Tianjin and Hebei mainly
Considering PM2.5 precursors such as SO2 can travel
come from coal-fired power generation, industrial
far away from their sources, and moving out coal
production, combustions of small-scale boilers
power plants may deteriorate pollution elsewhere,
for commercial and domestic purposes and the
Beijing, Tianjin and Hebei should not depend on
transportation sector. In total, these sectors emit
outsourcing coal-fired power plants to other provinces
over 12 million tons of primary PM2.5 and precursor
or importing dirty electricity from these provinces.
gases such as SO2, NOx, NH3 and volatile organic
Instead, the Jingjinji region should promote the
compounds ( VOCs ) in 2010. If aggressive
development of clean and renewable energy to
measures are not adopted and reduction efforts
control and decrease the overall use of coal.
stay the same as those preceding the release of the
Action Plan by the State Council, PM2.5 in 2030 will
This report provides an important foundation for a
remain unacceptably high. In order to reduce PM2.5
science-based, systematic approach to tackling PM2.5
3
concentration to 35 µg/m in ten years, or equivalently
pollution in the Jingjinji region.
47
Appendices
48
49
Appendix 1:
PM2.5 Background Information
PM2.5 is not a single pollutant, but rather a compound of pollutants existing in various forms: filterable, condensable,
organic, inorganic, solid, and gas. PM2.5 sources can be classified into primary and secondary sources. See the
following chart:
Secondary
Precursors
SO2
NOx
NH3
VOCs
Transformation
SO42NO3NH4+
SNA
OC
OC = Organic Carbon
CB = Carbon Black
BC = Black Carbon
EC = Elementary Carbon
Source: Hu ( 2012 ) and Environment Canada ( 2001 )
PM2.5
Primary
Crust elements
Metal
BC/EC/CB
Primary and Secondary PM2.5 Sources
Primary PM2.5 refers to tiny solids or liquid droplets
nitrogen ( NOx ), ammonia ( NH3 ) and various
released either directly into the air from a variety of
hydrocarbons referred to as volatile organic compounds
sources such as cars, trucks, factories, construction
( VOCs ). These gases can result from fuel combustion
sites, agriculture, unpaved roads, stone crushing, and
in motor vehicles, at power plants, and in other
burning of wood.
industrial processes.
Secondary PM2.5 is formed in the air from the chemical
Much of the available research focuses on primary
change of gases, or indirectly formed when gases from
sources of PM2.5. Analysis of PM2.5 transformed from
burning fuels react with sunlight and water vapour.
precursors has been lacking. Our research shows that
Precursor gases involved in secondary formation
it is important to consider both of these sources.
include sulphur dioxide ( SO2 ), oxides of
50
Appendix 2:
The Chemical Composition
of PM2.5
The figure on the left shows the
percentages of Sulfate-Nitrate-Ammonium
( SNA ), Nitrate ( NH4+ ), Black Carbon
( BC ) and Organic Carbon ( OC ) in the
mass concentration of PM2.5 ( e.g. if the
figure shows the SNA in a certain area is
0.4, it means that SNA accounts for 40%
of PM2.5 mass concentration in this area ).
SNA is the dominant component in PM2.5
in the Jingjinji region, accounting for 50%70% of the total mass concentration. BC is
mainly caused by incomplete combustion
of fossil fuels or biomass, and is usually
regarded as an indicator for measuring
pollution caused by vehicles ( Huang et al.
2006 ). OC belongs both to primary and
secondary pollutants. Its sources include
primary organic aerosol ( POA ) generated
through fuel combustion and secondary
organic aerosol ( SOA ) transformed from
direct VOCs emissions.
51
Appendix 3:
A List of Sector-specific Emission
Reduction Measures
Table A1: List of sector-specific emission reduction measures
Sector
Emission reduction measures
1.Improvement of energy efficiency in the domestic sector
2.Substitution of district heating for a decentralized heat-supply
3.Phasing out of residential coal stove
Domestic
4.Growth of natural gas consumption in the domestic sector
5.Promotion of low-sulfur coal in the domestic sector
6.Installation of web scrubbers in residential coal-fired boilers
7.Installation of low-nitrogen burner ( LNB ) in residential gas-fired boilers
1.Increasing energy efficiency in the industrial-combustion sector
2.Promotion of combined heat and electricity generation
3.Phasing out of small capacity coal-fired industrial boilers
4.Substitution of gas-fired boilers for coal-fired ones
Industrial combustion
5.Promotion of low-sulfur coal in the industrial combustion sector
6.Installation of flue gas desulfurization ( FGD ) in coal-fired industrial boilers
7.Installation of FF in newly-built coal-fired industrial boilers
8.Installation of LNB in industrial boilers
9.Installation of wet scrubbers in old industrial boilers
10.Installation of fabric filter ( FF ) in existing coal-fired industrial boilers
52
Sector
Emission reduction measures
1.Increased use of electricity generated from power plants
outside the city boundary
2.Phasing out of small capacity coal power plants
3.Substitution of natural gas-fired power plants for coal-fired ones
4.Substitution of IGCC for traditional coal-fired power plants
5.Substitution of natural gas-fired heating plants for coal fired ones
6.Installation of carbon capture and storage technologies
7.Promotion of low-sulfur coal in power plants
Power plants
8.Installation of FGD in coal-fired power plants
9.Installation of FF in newly-built coal power plants
10.Installation of electrostatic precipitators ( ESP )
& FF in newly-built coal power plants
11.Installation of FF in old coal-fired power plants
12.Installation of LNB in power plants
13.Installation of selective-catalytic-reduction ( SCR )
in newly-built gas-fired power plants
14.Installation of SCR in all existing power plants
53
Sector
Emission reduction measures
1. National IV Standard for light vehicles
2. National IV Standard for heavy duty diesel vehicles ( HDDVs )
3. National V Standard for HDDVs
4. National III Standard for motorcycles
5. National II Standard for non-road vehicles ( ORVs )
6. National IV Standard for gas vehicles
7. National V Standard for light duty gasoline vehicles ( LDGVs )
8. National VI Standard for LDGVs
9. National VI Standard for HDDVs
10. National III Standard for ORVs
11. National IV Standard for ORVs
12. National V Standard for ORVs
Transportation
13. National VI Standard for ORVs
14. Scrap pre-National I gasoline vehicles
15. Scrap pre-National I diesel vehicles
16. Scrap national I gasoline vehicles
17. Scrap national I diesel vehicles
18. Scrap national II gasoline vehicles
19. Scrap national II diesel vehicles
20. Restriction on the number of private cars
21. Promotion of gas vehicles
22. Promotion of alternative energy buses
23. Promotion of alternative energy taxis
24. Promotion of alternative energy private cars
25. Installation of selective-catalytic-reduction in diesel vehicles
26. Recycling of oil vapor in gas stations
54
Sector
Emission reduction measures
1. Shutdown of cement plants with vertical kilns and non-precalciner rotary kilns
2. Shutdown of coke plant old ovens
3. Ban newly built/renovated/expanded cement plants
4. Ban newly built/renovated/expanded lime plants
5. Ban newly built/renovated/expanded brick plants
6. Ban newly built/renovated/expanded glass plants
7. Ban newly built/renovated/expanded iron and steel plants
8. Ban newly built/renovated/expanded coke plants
9. Shutdown of some existing cement plants
10. Shutdown of some existing lime plants
Industrial process
11. Shutdown of some existing tile and brick plants
12. Shutdown of some existing glass plants
13. Shutdown of some existing iron and steel plants
14. Shutdown of some existing coke plants
15. Substitution of gas-fired kilns for coal-fired ones
16. Installation of FF in cement plants
17. Installation of SNCR-DeNOx in precalciner kilns of cement plants
18. Installation of FF in lime plants
19. Installation of end-of-pipe particle control in sinter plants
20. Stricter control of fugitive emissions in sinter plants
21. Installation of end-of-pipe SO2 control in sinter plants
22. Adoption of low-VOC materials in the coating industry
Source: Liu et al. (2013)
55
Bibliography
[1] Brauer M, Amann M, Burnett R T, et al.2012. Exposure assessment for estimation of the global burden of
disease attributable to outdoor air pollution. Environ. Sci. Technol., 46(2): 652-660.
[2] Damberg R. 2007. Policies for addressing PM2.5 precursor emissions. USEPA, Office of Air Quality and
Standards. Accessed from: http://yosemite.epa.gov/oa/eab_web_docket.nsf/filings%20by%20appeal%20number/
cd5f1d01895e1b6585257719006e71bc/$file/exhibit%2027%20damberg...3.11.pdf.
[3] Environment Canada. 2001. Interim Plan 2001 on Particulate Matter and Ozone. Accessed from: http://www.
ec.gc.ca/air/default.asp?lang=En&n=0768F92F-1&offset=1&toc=show on Nov 4. 2013.
[4] Greenpeace. 2013a. The health impact from coal power plants in Beijing, Tianjin and Hebei. [report]
Greenpeace East Asia Regional Office, Beijing. http://www.greenpeace.org/china/Global/china/publications/
campaigns/climate-energy/2013/1306jingjinji-health-rpt.pdf.
[5] Greenpeace. 2013b. Survey on the public satisfaction of air quality in the Beijing, Tianjin and Hebei region.
http://www.greenpeace.org/china/zh/news/releases/climate-energy/2013/06/jingjinji-air-polling.
[6] Hu M. 2012. Source Analysis of PM2.5. http://wenku.baidu.com/view/1ad4827b8e9951e79b8927b1.html, on Nov
4, 2013.
[7] Huang X F, Yu J Z, He L Y, et al. 2006. Size distribution characteristics of elemental carbon emitted from
Chinese vehicles: results of a tunnel study and atmospheric implications. Environ. Sci. Technol., 40: 5355-5360.
[8] International Energy Agency ( IEA ). 2009. Energy technology transitions for industry: Strategies for the next
industrial revolution.
[9] Liu F, Klimont Z, Zhang Q, et al. 2013. Integrating mitigation of air pollutants and greenhouse gases in Chinese
cities: Development of GAINS-City model for Beijing. J. Clean Prod., 58: 25-33.
56
[10] Ministry of Environmental Protection, China ( MEP ). 2013. Execution program on the action plan of BeijingTianjin-Hebei and surrounding areas on air pollution prevention and control. http://www.zhb.gov.cn/gkml/hbb/
bwj/201309/W020130918412886411956.pdf.
[11] Pan X C, Li G X, Gao T. 2012. Dangerous Breathing, PM2.5: Measuring the human health and economic
impacts of on China’s largest cities. Greenpeace East Asia Regional Office, Beijing. China Environmental Science
Press.
[12] van Donkelaar A, Martin RV, Brauer M, et al. 2010. Global estimates of ambient fine particulate matter
concentrations from satellite-based aerosol optical depth: development and application. Environ. Health Persp.,
118(6): 847-855.
[13] Wang Y, Zhuang G, Zhang X, et al. 2006. The ion chemistry, seasonal cycle, and sources of PM2.5 and TSP
aerosol in Shanghai, Atmos. Environ., 40: 2935-2952.
[14] Wang Y, Zhang Q, He K, et al. 2013. Sulfate-Nitrate-Ammonium aerosols over China: response to 2000-2015
emission changes of sulfur dioxide, nitrogen oxides, and ammonia. Atmos. Chem. Phys., 13: 2635-2652
[15] Zhang Y. 2011. Chemical characteristics of secondary inorganic aerosols during a typical haze episode in
Shanghai ( in Chinese ). The Administration and Technique of Environmental Monitoring, S1.
57