Policy Analysis pubs.acs.org/est
Examining Air Pollution in China Using Production- And Consumption-Based Emissions Accounting Approaches Hong Huo,† Qiang Zhang,*,‡ Dabo Guan,‡,§ Xin Su,∥ Hongyan Zhao,‡ and Kebin He∥ †
Institute of Energy, Environment and Economy, Tsinghua University, Beijing 100084, People’s Republic of China Ministry of Education Key Laboratory for Earth System Modeling, Center for Earth System Science, Tsinghua University, Beijing 100084, People’s Republic of China § School of International Development, University of East Anglia, Norwich, NR4 7TJ, United Kingdom ∥ State Key Joint Laboratory of Environment Simulation and Pollution Control, School of Environment, Tsinghua University, Beijing 100084, People’s Republic of China ‡
S Supporting Information *
ABSTRACT: Two important reasons for China’s air pollution are the high emission factors (emission per unit of product) of pollution sources and the high emission intensity (emissions per unit of GDP) of the industrial structure. Therefore, a wide variety of policy measures, including both emission abatement technologies and economic adjustment, must be implemented. To support such measures, this study used the production- and consumption-based emissions accounting approaches to simulate the SO2, NOx, PM2.5, and VOC emissions flows among producers and consumers. This study analyzed the emissions and GDP performance of 36 production sectors. The results showed that the equipment, machinery, and devices manufacturing and construction sectors contributed more than 50% of air pollutant emissions, and most of their products were used for capital formation and export. The service sector had the lowest emission intensities, and its output was mainly consumed by households and the government. In China, the emission intensities of production activities triggered by capital formation and export were approximately twice that of the service sector triggered by final consumption expenditure. This study suggests that China should control air pollution using the following strategies: applying end-of-pipe abatement technologies and using cleaner fuels to further decrease the emission factors associated with rural cooking, electricity generation, and the transportation sector; continuing to limit highly emission-intensive but low value-added exports; developing a plan to reduce construction activities; and increasing the proportion of service GDP in the national economy.
■
INTRODUCTION
improving fuel consumption efficiencies to reduce the emissions per unit of product produced or per unit of service provided (i.e., emission factors). China has made considerable efforts in recent decades and the emission factors of the major pollution sources are decreasing rapidly. For instance, the SO2, NOx, and PM2.5 emission factors (in g/kg of coal combusted) of coal-fired power plants were decreased by 61%, 32%, and 80% from 2000 to 2012, respectively.3 Furthermore, China is trying to improve its energy mix in cities by increasing natural gas and renewable energy consumption to replace coal consumption. China’s State Council has set the BeijingTianjin-Hebei region as the most stringent air-pollution control region. The natural gas consumption plans to increase 5-fold
Recently, haze pollution in China, which is characterized by high concentrations of fine particulate matter (PM2.5), has raised substantial attention around the world. In 2013, 69 of the 74 key monitored cities in China (municipalities, major cities in the Beijing-Tianjin-Hebei region, Yangtze River Delta, the Pearl River Delta, and provincial capitals) had an annual average PM2.5 concentration over 35 μm/m3,1 the recommended interim target of the World Health Organization. The origin of PM2.5 and how to reduce pollution are questions of great concern for the scientific community, the government, and the public in China. PM2.5 can be directly emitted from fossil fuel combustion and industrial processes. It can also be formed from the reactions of gaseous pollutants, such as NOx, SO2, and VOCs, which are also released by fossil fuel combustion and/or industrial processes.2 The conventional ways to control air pollution are utilizing end-of-pipe emission control technologies and/or © 2014 American Chemical Society
Received: Revised: Accepted: Published: 14139
August 14, 2014 October 29, 2014 November 17, 2014 November 17, 2014 dx.doi.org/10.1021/es503959t | Environ. Sci. Technol. 2014, 48, 14139−14147
Environmental Science & Technology
Policy Analysis
from 12 billion m3 in 2012 to 62 billion m3 in 2017. Nevertheless, the total amount of air pollutant emissions is still increasing in China because of the rapid increases of industrial products and services, which could offset the benefit gained from the end-of-pipe emission abatement measures. Although China has ambitious plans to further decrease the emission levels from major pollutant sources, given the rapidly increasing output of industrial products and services in China, the emission control measures and strategies must be considered from multiple perspectives other than single end-of-pipe treatment approaches. China suffers from serious air pollution owing to two major issues from a production perspective: (1) the relatively high emission factors of pollution sources (e.g., power plants, vehicles, industry boilers) and (2) the high emission intensity (amount of pollutant emitted per unit of GDP) of China’s industrial structure, which is largely driven by China’s coaldominated energy-consumption pattern. China has invested 1.7 trillion Yuan (≈280 billion U.S. dollars) implementing the “Action Plan for Air Pollution Prevention and Control” to address the above two causes.4 Previous studies examined emissions from the production side, which can support policy measures to cut emissions from actual pollution sources. For example, previous research on PM in China has focused on the measurement of the chemical composition of PM and potential measures to cut emissions in Chinese cities.5−13 A detailed review of the sources and formation of air pollutants in China can be found in Chan and Yao.14 Analyzing emissions from the consumption side provides an alternative approach to understanding the fundamental causes that trigger emission discharges. Such analyses can determine the amount of the embodied emissions that are involved in intermediate production flows along production supply chains as well as the consumers of these products. This emission accounting approach provides quantitative evidence for decision makers to develop and adopt appropriate policies to reduce emissions by altering consumption practices and industrial structure. Consumption-based emission accounting was first proposed to attribute the responsibility for CO2 emissions to final consumers.15−20 Recently, research studies have been established to understand the socio-economic drivers of PM2.521,22 and mercury,23 and emissions caused by international trade in China.21,24 This paper examined SO2, NOx, PM2.5, and VOC emissions in China in 2010 from both production and consumption perspectives by simulating the emissions flows among producers and consumers and analyzing the emissions and GDP performance among different production sectors. We aim at better understanding the main causes of emissions in China from multiple perspectives and to support policy makers in addressing China’s air pollution with a wide variety of measures, ranging from emission abatement technologies to economic adjustment.
MEIC Model. We used the MEIC model (accessible at www.meicmodel.org), a published and publicly available model,25−29 developed and maintained by Tsinghua University, to generate an emission coefficient matrix. The MEIC is a unit/ technology-based, bottom-up air pollutant emission inventory for China, which provides a rich bottom-up emission data from 1990 to the present. This inventory includes SO2, NOx, CO, NMVOC, NH3, CO2, PM2.5, PM10, BC, and OC emissions from more than 700 emission sources and production categories. The MEIC is an update of the bottom-up emission inventory developed by the same group.25,26 The major updates include a unit-based power plant emission database,27 a highresolution vehicle emission inventory,28 and an explicit NMVOC emission speciation process model.29 The overall uncertainties in bottom-up emission estimates over China are 12% for SO2, 31% for NOx, 68% for NMVOC, and 107% for PM2.5.25,26 We aggregated the 700+ emission sources and production categories into 36 production sectors (see Supporting Information (SI) Table S1) based on the classification method of China’s official energy statistical yearbook. These sectors were further aggregated into 7 industry sectors (agriculture, mining, manufacturing, gas−water-electricity industry, construction, transportation, and services). Results from the MEIC model enabled production-based air pollution accounting. Environmentally Extended Input−Output Model. We used the environmentally extended input−output (IO) analysis method30,31 to analyze the emissions from the consumption perspective. This method was used to calculate the direct emissions and embodied emissions of a production sector. The mathematical structure of an IO system consists of n linear equations in n unknowns, as shown in eq 1. The equation shows that the value of total production is equal to the intermediate deliveries plus final demand for each sector.
MATERIALS AND METHODS We performed the production-based emissions analysis on the basis of a production-based, bottom-up air pollutant emission inventory of China. The Multiresolution Emission Inventory for China (MEIC) was developed in our previous work.3 We then used the environmentally extended input−output (IO) analysis method to establish the consumption-based emission accounting.
aij is a fixed relationship between a sector’s outputs and its inputs. Thus, there is an explicit definition of a linear relationship between input and output and there are no economies of scale. In matrix notation, the IO model can be written as shown in eq 3:
xi = zi1 + zi2 + ... + zij + ... + zin + yi , i = 1, 2, ..., n (1)
where n is the number of economic sectors of an economy, xi is the total output of sector i, yi is the total final demand for sector i’s product, and zij is the intermediate delivery from ith sector to the jth sector. A product of the sector i can be the feedstock of another sector j. A fundamental assumption in IO analysis is that the interindustry flows from i to j,32 represented as zij. By dividing zij by xj (the total output of the jth sector), one can obtain the ratio of input to output zij/xj, denoted as aij, which reflects the production efficiency with present technology. The so-called technical coefficient or direct requirement coefficient represents the requirement from economic sector i to produce one monetary unit of product in economic sector j. aij =
■
zij xj
(2)
x = (I − A)−1y 14140
(3)
dx.doi.org/10.1021/es503959t | Environ. Sci. Technol. 2014, 48, 14139−14147
Environmental Science & Technology
Policy Analysis
Figure 1. Emissions flows from production to consumption in China in 2010. Figure 1 starts with emissions from pollution sources (A), then separates the emissions into two segments: emissions from the consumption activities of households (B) and emissions from industrial production activities (C). Using eqs 1−5, Figure 1 illustrates the emissions flow from production industry sectors (C) to consumption industry sectors (D), then traces the consumption-based emissions (D) to five final consumption categories (E). Therefore, B and E represent emissions by consumption categories, and the shares of B and E add up to 100%.
The term (I − A)−1 accounts for the total accumulative effects, including both direct and indirect effects, on sectoral output by the changes in final demand. In other words, for every sector to deliver one unit of final demand, every sector must meet not only its own final demand but also the direct and indirect requirements for its own and the other sectors’ final demand (see further description for IO analysis in SI S1.1). The emissions of sectors are calculated by eq 4: E = Fx
consumption). We assumed that each economic sector and final demand category used sectoral imports in the same proportions (see SI section S1.2 for further description and mathematical illustration).35 Additionally, we aggregated the fixed capital investment and capital inventory changes into one “capital formation” category. Matrix F was derived from the MEIC model based on China’s energy statistics. The detailed explanation of the environmentally extended input−output analysis method is available in the SI (section S1.3). Uncertainty arises in input-output analysis in many avenues.32 The inherent weaknesses of applying input−output analysis include lack of supply side constraints, fixes prices, fixed ratios for intermediate inputs and production, etc.32 The Chinese input-output tables follow most standard formats except for a final demand column called “Others”. The error in aggregated GDP by not including “Others” is small (
