Business as usual - Department of Industry, Innovation · Web viewThis has a dampening effect on economic growth, leading to marginally lower gross domestic product (GDP) growth. More

Key factors affecting changes in China’s demand for liquefied natural gasKey factors affecting

changes in China’s demand for liquefied natural gas 0

Key factors affecting changes in China’s demand for liquefied natural gas

Jin Liu, Xiujian Peng and Philip Adams1

April 2016

Keywords: China, LNG, gas

1 Dr Xiujian Peng and Professor Philip Adams are working at the Centre of Policy Studies

(CoPS), Victoria University, Melbourne.



For further information on this report, please contact:

Jin LiuOffice of the Chief EconomistDepartment of Industry, Innovation and ScienceGPO Box 9839Canberra ACT 2601

Further informationFor more information on other Department initiatives please see the Department’s website at www.industry.gov.au/OCE.For more information or to comment on this publication, please email [email protected].

Acknowledgements: The authors would like to acknowledge valuable comments from David Whitelaw, Kieran Bernie, Nicole Thomas, Marco Hatt, Laura Jones and Katya Golobokova in the Office of Chief Economist; the Petroleum Development and Timor Sea team in the Resources Division, Department of Industry, Innovation and Science; and Stephen Wilson, Director at Cape Otway Associates. .

The views expressed in this publication are those of the authors and do not necessarily reflect those of the Australian Government or the Department of Industry, Innovation and Science.

Commonwealth of Australia 2016

This work is copyright. Apart from use under the Copyright Act 1968, no part may be reproduced or altered by any process without prior written permission from the Australian Government. Requests and inquiries concerning reproduction and rights should be addressed to [email protected]. For more information on Office of the Chief Economist research papers, please access the Department’s website at www.industry.gov.au/OCE .

Creative Commons Licence

With the exception of the Coat of Arms, this publication is licensed under a Creative Commons Attribution 3.0 Australia Licence.

Creative Commons Attribution 3.0 Australia Licence is a standard form licence agreement that allows you to copy, distribute, transmit and adapt this publication provided that you attribute the work. A summary of the licence terms is available from http://creativecommons.org/licenses/by/3.0/au/deed.en. The full licence terms are available from http://creativecommons.org/licenses/by/3.0/au/legalcode.

The Commonwealth’s preference is that you attribute this publication (and any material sourced from it) using the following wording:

Source: Licensed from the Commonwealth of Australia under a Creative Commons Attribution 3.0 Australia Licence. The Commonwealth of Australia does not necessarily endorse the content of this publication.



http://www.industry.gov.au/OCE

mailto:[email protected]

ForewordChina faces major challenges as it seeks to rebalance its economy from investment-led growth to consumption-led growth. An implication of moving to a consumption-driven economy is a decline in the share of high energy–intensive industrial sectors and an increase in the share of low energy-intensity services. Furthermore, China’s economic rebalancing is occurring at a time when major policy initiatives relating to energy security and environmental outcomes are gaining increased attention from policy makers. One of these policies is the plan to increase the share of natural gas in China’s energy mix to obtain environmental benefits in the form of both lower atmospheric pollution and carbon dioxide (CO2) emissions.

China’s future demand for natural gas, and liquefied natural gas (LNG) in particular, is of significant interest to Australia. China is expected to be a major source of incremental global demand for LNG in the coming years, and Australia is about to become the world’s largest LNG producer.

This study aims to enhance understanding of the likely outcomes of potential policies which may affect China’s natural gas supply and demand during the next 15 years, and the likely consequences for China’s LNG demand. To capture the direct and indirect effects flowing though China’s economy in response to particular policy initiatives, the analysis uses a dynamic computable general equilibrium model of China.

The insights from the comprehensive exploratory analysis presented in this study make a valuable contribution to our understanding of the key drivers likely to affect China’s demand for LNG in the medium to long term.

Mark CullyChief EconomistDepartment of Industry, Innovation and ScienceApril 2016



Contents

Abbreviations and acronyms 5

Executive summary 6

1. Introduction 9

1.1 Study approach 10

2. The Chinese economy and energy sector 17

2.1 Rapid growth has transformed China’s economy 17

2.2 China’s energy mix is also changing 19

3. Natural gas in China 29

3.1 Consumption growing, but still low 29

3.2 Domestic production surging 29

3.3 Import reliance growing, but many options 33

4. Policy scenario one — an increase in the service sector 36

4.1 The effect of scenario one on natural gas consumption 36

4.2 The implications for energy intensity 38

5. Policy scenario two — capping coal consumption by 2020 40

5.1 The effect of scenario two on natural gas consumption 40

5.2 The implications for emission intensity 43

6. Policy scenario three — higher unconventional gas production 44

6.1 The effect of scenario three on gas supply 44

6.2 The implications for natural gas imports 45

7. A composite policy and sensitivity to oil prices 47

7.1 A composite policy and its effects 47

7.2 Crude oil prices and natural gas imports 51

8. Key findings and concluding remarks 54

Appendix A Key characteristics of CHINAGEM 57

References 73





Abbreviations and acronymsbbl barrel

bcm billion cubic meters

BTOE back of the envelope (model)

btoe billion tonnes of oil equivalent

CAGR compound annual growth rate

CES constant elasticity of substitution

CET constant elasticity of transformation

CHINAGEM a dynamic computable general equilibrium model of China

CHP combined heat and power

CO2 carbon dioxide

CoPS Centre of Policy Studies, Victoria University

FDI foreign direct investment

GW gigawatt

IEA International Energy Agency

IMF International Monetary Fund

I/O input/output

JCC Japan customs-cleared crude

kWh kilowatt hours

LNG liquefied natural gas

mmBtu million British thermal unit

mt metric tonnes

mtoe million tonnes oil equivalent

PPP purchasing power parity

RMB Chinese Yuan Renminbi

tcm trillion cubic metres

toe tonnes of oil equivalent

TWh terawatt hours

UN United Nations

WDI World Development Indicator



WGM World Gas Model (Nexant)

Executive summaryChina’s energy policy aims to reduce both energy intensity and coal dependency through greater use of cleaner fuels such as gas and renewable energy. The goal is to achieve strong economic growth with lower CO2

intensity. Achieving such an outcome ultimately involves trade-offs and challenges for policy makers.

Natural gas combustion has a lower CO2 intensity than other fossil fuels and, therefore, is likely to play an important role in an economy’s transition to lower CO2 emissions. This study explores the role of natural gas, and LNG in particular, as an energy source in China’s economy, and the effects various policies may have on its supply and use from 2015 to 2030.

China is likely to be a major source of future growth in LNG demand in the Asia–Pacific region. This will occur at a time when Australia becomes the world’s largest LNG exporter. Understanding how China’s policies may affect LNG demand is, therefore, very important to Australia. Three key initiatives that are likely to affect natural gas demand and LNG imports include policies aimed at transitioning the economy towards consumption-led growth, improving air quality and reducing greenhouse gas emissions, and increasing the production and use of unconventional gas.

This study explores key aspects of China’s future demand for energy (particularly from LNG), based on a dynamic computable general equilibrium model of China (CHINAGEM). The CHINAGEM model is used to establish a business-as-usual baseline for China’s economy (covering energy and non-energy sectors) to 2030, as well as to simulate the effects of various policy scenarios over time. The three policy scenarios considered are:

faster structural transition of its economy, through an increasing share of the service sector

stronger action to limit coal consumption, through a slowing rate of growth in coal consumption and a cap on total coal consumption

increased growth in unconventional gas production, through technological improvements and increased investment in exploration and infrastructure.

The study also explores a composite policy scenario, which integrates the three policy scenarios to help understand the potential impact of a potential combined policy approach. The composite policy scenario reflects a potential attempt by China to meet the three key objectives of economic growth, environmental improvements and energy security.

Business as usual

The baseline scenario is consistent with current trends, showing China’s economic growth rate slowing as its economy transforms and becomes less energy intensive. On the supply side, there is competition from unconventional gas production and imported pipeline gas on China’s LNG imports.



An increase in the service sector

Increasing the share of the service sector in China is found to stimulate economic growth, meaning that both industries and households can afford, and are willing to pay for, cleaner fuels such as gas. Although the service sector is less energy intensive than industry, the increased overall economic growth in this scenario results in growing total energy consumption. Higher income also supports growth in the share of gas consumption, which leads to higher gas demand, increased gas production and a faster growth in gas imports.

Capping coal consumption by 2020

Capping the growth in coal consumption leads to an increase in cost of energy to consumers. This has a dampening effect on economic growth, leading to marginally lower gross domestic product (GDP) growth.

More expensive energy also lowers total energy demand, although there is an increase in natural gas consumption of around 25 per cent by 2030 relative to baseline, attributable to switching from coal to natural gas and alternative fuels. Of particular note is the shift in gas consumption amongst sectors. The results show an increase in gas consumption in all sectors, but the relative shares of gas consumption shifts from traditional industrial sectors to emerging sectors, including the manufacture of communications equipment and the hotel industry. A further consequence of this policy is a significant reduction in both CO2 emissions and the emission intensity of the economy.

Higher unconventional gas production

The scenario modelling increased unconventional production assumes that the cost of unconventional production comes down, as a result of technological improvements and increased productivity as more unconventional gas is extracted. This reduction in costs stimulates total gas demand, and leads to a reduction in conventional gas supply, given the change in relative prices. However, total gas demand does not grow as fast as production and there is a decline in natural gas imports, reducing China’s natural gas import dependency.

A composite policy and oil price

The composite policy scenario combines the three interventions in the policy scenarios to determine the effects that might result from multiple policy interventions in China’s economy. This scenario results in a higher share of natural gas in China’s primary energy and electricity generation mix, higher LNG imports and lower carbon emissions because of a shift away from coal to gas.

Compared with the baseline scenario, the key outcomes under the composite policy scenario for the year 2030 are that:

the share of coal consumption in the primary energy mix reduces to 51 per cent (from 60 per cent), the natural gas share increases to



11 per cent (from 7 per cent) and the share of non-fossil fuels increases to 20 per cent from (15 per cent) of the primary energy mix

natural gas imports grow faster to meet the increased natural gas demand, which allows the import dependency to rise to 38 per cent (from 33 per cent)

total carbon emissions reduce to 11.2 billion tonnes (from 13.6 billion tonnes).

The sensitivity of this scenario to oil prices is also assessed, which shows that a high oil price scenario would lead to lower LNG imports and higher indigenous gas production.

It is impossible to forecast the precise policy interventions that the Chinese government will implement to balance the trade-offs between energy security, cost and emissions. However, this report finds that Chinese economic and energy policy interventions will have a large impact on China’s demand for LNG, including through changes to the structure of the economy and the energy mix.



1. IntroductionChina’s rapid economic growth has brought benefits to its and other economies. However, the activities that underpinned China’s economic development have also incurred environmental costs, such as increased emissions of air pollutants, including greenhouse gases.

Structural adjustments are transitioning China’s economy from investment-led growth to consumption-led growth. This is leading to lower headline economic growth in China, the elimination of outdated production capacity and relative declines in energy-intensive industries. Heavy industrial sectors have a lower investment share in the economy’s GDP, which implies that China is allocating capital away from these sectors (particularly low value–added activities) towards service sectors and higher value–added manufacturing. This potentially reduces the growth in CO2 emissions and the energy intensity of domestic production.

China’s economic transition, climate change policies and anti-air pollution plans have all contributed to the increased use of natural gas relative to other fossil fuels in the economy. The burning of natural gas emits less CO2 than coal and oil. In this sense, natural gas serves as a key alternative energy source for an economy seeking to achieve a balanced growth path, and implement climate change and anti-air pollution policy reform. However, significant uncertainties arise from the competition between natural gas and non-fossil fuels, and between LNG imports, pipeline imports and indigenous natural gas production. China is trying to implement an energy transition to lower- and zero-carbon energy choices, and natural gas is viewed as a viable bridge fuel to cleaner energy technologies for at least the next decade.2

Chinese policy makers face challenging trade-offs between ensuring energy security, managing economic transition costs and achieving emissions reductions. With respect to the latter, effective policy intervention across a range of sectors will be required to reverse the steep emissions growth that has been observed during recent decades. A key focus of this report is to investigate how China’s aim to increase in natural gas in its primary energy mix to achieve environmental outcomes may affect its LNG imports. This recently adopted key energy policy objective is driven by the lower relative emissions of CO2 and urban air pollutants from natural gas compared with coal, which is currently China’s main energy source. This policy objective has implications for natural gas demand in China and, hence, for imports of LNG and pipeline gas.

Several baseline trends affecting LNG demand in China are apparent. From a demand perspective, China’s economic growth rate is slowing as it goes through an economic transition — the economy is becoming less energy intensive. From the supply side, questions arise about the impact of unconventional gas and the import of pipeline gas on China’s LNG imports. The extent of China’s LNG demand is one of the key uncertainties impacting on global LNG outlooks, and is of particular interest to Australia given it aligns with the timing of our emergence as one of the world’s largest LNG exporters.

2 Wang (2015) China’s elusive shale gas boom



This study aims to improve policy makers’ and market participants’ understanding of the main forces driving changes in China’s demand for LNG. China is expected to be a major source of incremental global demand for LNG in the future. The analysis is based on a dynamic computable general equilibrium model of China (CHINAGEM), which can simulate the effects of various policy scenarios. The CHINAGEM model is well suited to analysing structural changes and sectoral shifts in the Chinese economy, and accounts for interactions between energy sectors and the rest of the economy. The model can help to identify direct and indirect effects of energy and climate change policies on energy shifts and fuel substitution.

The report includes several chapters, four of which present different policy scenarios:

Sections two and three provide the context of the energy nexus between economic growth, energy consumption and carbon emissions, and of the current state of China’s natural gas industry.

Section four examines policy scenario one, which is an increasing share of the service sector in the total economy and its effects on income, energy intensity and natural gas demand.

Section five assesses policy scenario two, which is capping coal consumption by 2020 and its effects on carbon emissions, carbon intensity and industrial gas users.

Section six analyses the third policy scenario, which is increasing unconventional gas production and its effects on China’s diversified gas supplies, including gas imports.

Section seven explores a composite policy scenario, which combines the three policy scenarios and their joint effect on aggregated gas demand and LNG imports. The effect of higher oil prices on changes in China’s LNG imports is also presented.

Section eight discusses key findings and conclusions.

1.1 Study approach

In this study, we use CHINAGEM to explore and analyse China’s energy policy issues. The Centre of Policy Studies (CoPS), an independent research unit at Victoria University, developed CHINAGEM a decade ago. Initiated and led by the Australian Government Department of Industry, Innovation and Science (the Department), CoPS and the Department have jointly developed the model further to make it more suitable for the purpose of this study. The extensions of the model allow for a more detailed examination of key factors associated with policy scenarios affecting changes in China’s demand for LNG.

CHINAGEM is a dynamic, single-country model, which deals with the long time frames associated with the issues being considered. The model is able to trace the evolution of the Chinese economy and its industries over time because it uses a dynamic approach. CHINAGEM also has a more detailed treatment of sectors than global or multicountry models. The characteristics of CHINAGEM are presented in Appendix A.



CHINAGEM is well suited to analysing structural changes and sectoral shifts in the Chinese economy, accounting for the interactions between energy sectors and the rest of the economy. The model can help to identify the direct and indirect effects of energy and climate change policies on energy shifts and fuel substitution. CHINAGEM accounts for technical changes by capturing the effects of more efficient use of, and improvements in, technologies. It also captures changes in households’ general consumption preferences for energy and other goods.

In broad terms, CHINAGEM describes the behaviour of economic agents and the linkages between sectors of the economy, as well as the linkages between China and the rest of the world. The model is based on a characterisation of the behaviour and interaction of several unique classes of economic agents. These agents include multiple producers that represent the many sectors within the Chinese economy — investors, households, exporters, importers and the government. In this sense, the model yields detailed insights into the benefits and costs of policy reform, beyond those provided by other models capable of producing macroeconomic impacts,3 such as time-series econometric models that lack sectoral detail. For example, CHINAGEM contains a climate change module for CO2 emissions accounting. Also, the model allows for substitution of different electricity-generating technologies, each of which possess different emissions intensities (see Appendix A).

Like all economic models, CHINAGEM models an imperfectly understood system that is subject to unpredictable influences. The model would benefit from having estimates generated using Chinese data, but — to the author’s knowledge — reliable and consistent estimates are not yet available. As a result, some data such as values for most substitution elasticities — including for Australia and the United States — are drawn from other sources. The modelling does not reflect or account for abrupt political changes within China, new technological developments in light and heavy manufacturing, or global events that significantly disrupt China’s economic performance.

The model allows for substitution between coal and gas, between different types of electricity generation, and between different sources of imported gas supply. However, at present, it does not allow for price-induced substitution between gas production from conventional and unconventional sources, or between biofuel (primarily wood), and fossil fuels and electricity.

Despite these limitations, the model in its current form is a powerful tool for scenario analysis. It is based on a clearly understood theoretical framework and is calibrated on the best available data. The model contains an adequate representation of broad energy-substitution possibilities, through:

equations for interfuel substitution in transport and stationary energy

equations that recognise different sources of supply of imported gas (LNG and pipeline)

separate sectors producing gas from conventional and non-conventional sources.

3 This study does not focus on the broad macroeconomic outputs of the CHINAGEM model.



The flexible closure in CHINAGEM allows insights into a wide variety of relevant real-world scenarios — for example:

a full set of energy and greenhouse gas accounts that covers each emitting agent and fuel

quantity-specific carbon taxes or prices

The model also accommodates both the economic-transition scenario and scenarios specific to the energy sector that are the focus of this study.

This study uses CHINAGEM in three modes: historical, forecast and policy. In a historical simulation, the observable variables such as GDP, production, consumption and international trade are exogenous, whereas the corresponding technical change and preference variables (such as multifactor productivity) are treated as endogenous. The model is informed of changes in the observable variables projected forward from the base year (2007) — the most recent year for which detailed input–output data are available. The model then calculates the implied changes in technology and preferences that transition the database from 2007 to 2013. To ensure relevance, the historical simulations are based on a detailed database reflecting the most recently available economic data from the Chinese government, and the most recently available data on emissions and energy usage.

Based on the model’s current database, 143 industries produce 142 commodities. Capital is industry specific. There is a single representative household and a single representative government sector. Finally, there are other countries, whose behaviour is summarised by demand curves for international exports and supply curves for international imports.

Baseline scenario

The forecast mode is used to establish a business-as-usual baseline for China’s economy (covering energy and non-energy sectors) through to 2030. The baseline is the result of a forecast simulation based on key projections (e.g. primary energy mix) from reputable organisations, such as the International Energy Agency (IEA) and the World Bank, and is consistent with the Chinese government’s strategic planning process. The baseline takes into account recent developments in electricity and gas demand, and credible assumptions for autonomous efficiency improvements in the future relating to energy and non-energy sectors.

The macroeconomic assumptions in the baseline forecast cover real GDP growth and the pattern of growth of GDP components — real consumption, real investment, government expenditure, exports and imports. Other assumptions include key sectoral growth rates for agriculture, industry and services, and the growth of the labour force. These numbers are used as inputs to CHINAGEM for the forecast simulation from 2014 to 2030 (for details of the specific forecasts used in the model, refer to Appendix A). Trends in preferences and technological change are derived from the historical simulation and serve as inputs to the forecast simulation, which is used to project the economic and energy outcomes going forward from 2013.



The baseline forecast is a business-as-usual scenario for the Chinese economy for the period 2014 to 2030. It is constructed on the assumption that there will be no changes in government policies, beyond those already announced. More specifically, the baseline is developed under the assumptions that:

1. The Chinese economy will continue to grow strongly, but following recent trends, overall growth will slowly diminish.

2. The pattern of growth will favour consumption and consumption-related industries at the expense of investment and investment-related industries.

3. Import growth will exceed export growth.

4. Growth in service sectors will exceed growth in industrial sectors.

These trends are adjusted to ensure they are consistent with the IEA’s Current Policies Scenario from its World Energy Outlook 2015, which comprises a suite of carbon, energy and industry policies (see Table 1.1). Other assumptions, such as lower growth for steel production and higher efficiencies in metallurgical coking operations, are also included in the baseline simulation.

As defined by the IEA, the Current Policies Scenario is not intended to take into account any possible, potential or even likely future policy actions from mid-2015. The scenario is designed to offer a baseline picture of how energy markets would evolve without any new policy intervention.4 Moreover, the projections of the IEA are based on extensive information on energy systems in a partial equilibrium framework. This approach does not capture the general equilibrium effects that would result from changes to other assumptions, as is the case with the CHINAGEM model.

Table 1.1: Key energy and industry assumptions/scenarios for baseline forecasts

Cross-cutting policy assumptions

Power-sector policies and measures

Industry-sector policies and measures

Implementation of measures in the 12th Five-Year Plan, including a 17 per cent cut in CO2 intensity by 2015 and a 16 per cent reduction in energy intensity by 2015 compared with 2010.Increase the share of non-fossil fuels in primary energy consumption to around 15 per cent by 2020.

Implementation of measures in the 12th Five-Year Plan.290 GW of installed hydro capacity by 2015.100 GW of installed wind capacity by 2015.35 GW of installed solar capacity by 2015.

Small plant closures and the phasing out of outdated production, including the comprehensive control of small coal-fired boilers.Mandatory adoption of coke dry-quenching and top-pressure turbines in new iron and steel plants. Support of non-blast furnace iron making.Three industries — iron smelting, steel making and steel rolling are assumed to grow at three per cent in 2015, two per cent in 2016, 1.5 per cent in 2017, one per cent in 2018 and 0.5 in 2019 and zero per cent from 2020 onwards.Due to small plant closures and the phasing out of older technologies, steel making in China will become more efficient in its use of coke. It is assumed that in every year steel producers will reduce their use of coke relative to output by five per cent.

4 IEA (2014b) World Energy Outlook 2014



The baseline forecasts are consistent with the IEA projections for primary energy consumption and electricity output by fuel type. Figure 1.1. shows that China’s coal consumption will increase from 2,144 mtoe in 2020 to 2,410 mtoe in 2030.

Figure 1.1: Forecast primary energy consumption and mix by fuel type

Primary energy consumption by fuel type Electricity output by fuel type

Coal Oil Gas Nuclear Hydro Other0

500

1,000

1,500

2,000

2,500

2020 2030

mtoe

Coal Oil Gas Nuclear Hydro Other0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

2020 2030

TWh

Notes: Based on the IEA Current Policies Scenario, which assumes no changes in policies from mid 2015

Source: IEA (2015c)

The assumptions on energy policies in the baseline scenario are consistent with the outcomes of the IEA (2015c) Current Policies Scenario:

The share of coal in primary energy mix will decline from 61.2 per cent in 2020 to 57.5 per cent in 2030.

The share of electricity output generated by coal will decline from 65 per cent in 2020 to 61.6 per cent in 2030.

The share of gas in the primary energy mix will increase from 7.2 per cent in 2020 to 8.8 per cent in 2030.

The share of electricity output generated by gas will increase from 2.1 per cent in 2014 to 4.2 per cent in 2020 and 5.8 per cent in 2030.

The share of electricity output generated by nuclear will increase from 4.2 per cent in 2020 and 5.8 per cent in 2030.

The share of non-fossil fuel in the primary energy mix will be 14.4 per cent in 2020 and 15.8 per cent in 2030.

The share of electricity output generated by non-fossil fuel will be 30.7 per cent in 2020 and 32.5 per cent in 2030.

The analysis presented in this report is structured around three new policy scenarios relative to the baseline. The aim of a policy simulation is to explore how the economy would evolve when subjected to various shocks or changes in economic policy, relative to the forecast baseline simulation.



Policy simulations play the role of attributing causation by explaining potential future events in terms of changes in naturally exogenous variables.5

Policy scenario one — increasing the share of the service sector

Policy scenario one models an increased share of the service sector in China’s GDP. This is done by accelerating economic restructuring, coupled with increases in household consumption of services, more urbanisation and preferences towards cleaner fuels such as gas. Specifically, it is assumed that the share of the service sector in GDP will be about 5 per cent higher in 2030 than in the baseline scenario (see Table 1.2).

To achieve this, the share of the service sector in the economy is treated as an exogenous variable, which frees up the proportion of income that is consumed by households. Thus, the model determines changes in consumption, which are consistent with the exogenously imposed increases in the share of services.6

Table 1.2: Change in share of sectoral contribution to GDP over time

Year Agriculture Industry Service

Baseline Policy one Baseline Policy one Baseline Policy one

2015 9.0 8.9 45.8 45.8 45.2 45.3

2020 7.7 7.2 44.6 43.7 47.7 49.2

2025 6.6 5.7 42.9 40.6 50.6 53.7

2030 5.6 4.5 40.7 36.9 53.7 58.6

Source: The World Bank (2012) and authors’ assumptions

The additional assumptions for scenario one are that:

Household preferences will shift towards services at five per cent each year. As Chinese consumers become wealthier, they will spend more money on services such as education, communication, travel and finance. Note that the shift in preferences means that, if prices and income remain at baseline values, then Chinese consumers will increase the share of services in their overall budget by five per cent a year.7

The household preference for coal will be 13 per cent lower each year than baseline (excluding indirect coal use through electricity). The rationale for this assumption is:

- As incomes rise, households will choose to consume cleaner energy — which includes an increased proportion of gas relative to coal. The

5 Dixon and Rimmer (2009) Forecasting with a CGE model: does it work?

6 By making a variable endogenous, the model is now free to determine the appropriate values

for the variable within the modelling framework.

7 This is a large increase in the service sector. It has been calibrated at a rate which would be

required to increase the service share in consumption from the current level in China to a level

consistent with the current Australian share of services in household consumption by 2030.



use of renewables could also expand, but gas is generally preferred for heating purposes.

- Rapid urbanisation means that more people will live in the cities where coal based air-pollution is a real problem, and where increasingly stringent controls are being placed on the usage of coal by households. The calibration of the shock is based on a subjective judgement that urbanisation will reduce direct household consumption of coal by around 80 per cent compared with baseline levels by 2030.8

Policy scenario two — capping coal consumption by 2020

Policy scenario two reduces the share of coal in primary energy consumption is achieved by capping coal consumption to a maximum of 4.2 billion tonnes in 2020. This results in a reduction in the coal share of the energy and electricity mix, and the CO2 intensity of GDP. To this end, it is assumed that:

The growth rate of primary coal consumption will gradually decline from 2015, and will be zero from 2021.

Electricity efficiency will be 1.5 per cent higher each year than under the baseline scenario.

Policy scenario three — higher unconventional gas production

Policy scenario three targets an increase in unconventional gas production using the IEA New Policies Scenario for China as a guide.9 This will be done by increasing overall natural gas supply by attracting both foreign direct investment (FDI) and domestic investments in gas infrastructure, and unconventional gas exploration and production.

Policy scenario four — a composite scenario

Policy scenario integrates all the assumptions in policy scenarios one, two and three. This scenario is intended to reflect the potential effects that might result from a combination of multiple policy interventions in China’s economy. This scenario is also tested for a range of different oil price scenarios, to assess how sensitive potential changes to energy and import mixes are affected by oil prices.

8 This shift is calibrated such that, initially, the household sector’s use of energy does not

change, only the mix of that energy (i.e. towards gas and away from coal). In the longer term,

the small change in the energy mix has a significant impact on the total consumption.

9 IEA (2015c) World Energy Outlook 2015



2. The Chinese economy and energy sector

2.1 Rapid growth has transformed China’s economy

The Chinese economy has been undergoing significant change since the commencement of economic reforms in 1979, growing rapidly while also undergoing structural transformation.

Structural change is underway

China’s economy can be divided into three sectors:

agriculture (or primary), which consists of farming, forestry, animal husbandry and fisheries

industry (or secondary), which comprises mining, construction and manufacturing

service (or tertiary) make up the remaining sectors, which are classed as tradable services (such as communications, business and legal practice, culture, research and education) and non-tradable services (domestic services, wholesale and retail trade).

Initially, the structural transformation of China’s economy involved the movement from an agriculture-dominated economy to an industry-dominated economy. More recently, a second phase of structural reform has seen an expansion of the service sector. This can be seen by examining changes in the shares of output and employment by sectors in the economy (Figure 2.1).

Figure 2.1: Structural transformation, output and employment shares, 1952–2014

Output share by sectors Employment share by sectors

20059

21155

22251

23346

24442

25538

26634

27729

28825

29921

31017

32112

33208

34304

35400

36495

37591

38687

39783

40878

41974

0

15

30

45

60

Primary Secondary Tertiary

Per c

ent

20059

21155

22251

23346

24442

25538

26634

27729

28825

29921

31017

32112

33208

34304

35400

36495

37591

38687

39783

40878

41974

0

15

30

45

60

75

90

Primary Secondary Tertiary

Per c

ent

Source: CEIC

During the past three decades, China’s path to industrialisation has been characterised by a lower share of the primary sector and a higher share of the industry sector in GDP and employment. From 2012, the share of the service sector in GDP grew by three per cent each year to 2014, while the share of the industry sector has been in decline. At the end of 2014, the Key factors affecting changes in China’s demand for liquefied natural gasKey factors affecting


service sector accounted for 48 per cent of GDP, with 43 per cent of GDP accounted for by the industry sector. These proportions are up by five per cent and down four per cent percentage points, respectively, relative to 2007.

The service sector generates knowledge, as well as transforming goods and services. Education, research and development, and health services are examples in the production of human capital. Services also involve the transport, distribution and sale of goods from producer to consumer. Figure 2.2 shows urban fixed capital investment doubled from 24 trillion renminbi (RMB) in 2010 to 50 trillion RMB in 2014. Additionally, the service sector attracts a greater share of FDI than the manufacturing sector. The increasing trend in the share of FDI attracted by the service sector is also apparent in Figure 2.2 (increased from 46 per cent in 2010 to 55 per cent in 2014).

In October 2015, the Fifth Plenary Session of the 18th Communist Party of China Central Committee suggested that the service sector in particular, which has huge growth potential, might be further opened to foreign investment. This would primarily impact the finance, education, culture and medical treatment sectors.



Figure 2.2: Capital investments in China, 2010–14

Urban fixed capital investment, total vs growth Foreign direct investment, by sector

2010 2011 2012 2013 20140

10

20

30

40

50

60

RMB (Trillion) Growth (per cent)

2010 2011 2012 2013 20140

20

40

60

80

100

46.1 47.6 48.252.3

55.4

46.9 44.9 43.7 38.7 33.4

Others Service Manufacturing

Per c

ent

Source: CEIC and KPMG (2015)

Income driving service sector growth

There is a relationship between growth in the service sector and income levels. Based on a sample of 21 countries, Buera and Kaboski10 found that a turning point occurs at an average per capita income of around US$7,100 purchasing power parity (PPP)11, after which the relative share of manufacturing declines and the share of the service sector continues to rise.

When GDP per capita reached US$7,554 PPP in 2008, China entered the middle-income stage of economic development. Relatively few countries that have achieved middle-income status in the past three or four decades have graduated to high-income status or achieved per capita incomes in excess of US$16,000. The International Monetary Fund (IMF) predicts that China’s GDP per capita, as measured by PPP, will be US$15,184 in 2016 (i.e. a doubling of GDP per capita over an eight-year period since 2008). By 2020, China’s GDP per capita is forecast to be US$20,494, double that of 2010. If this occurs, it will be twice the GDP per capita projected for India. From a base in 1980, when China’s GDP per capita was five per cent that of the

10 Buera and Kaboski (2008) Scale and the origins of structural change

11 The concept of PPP allows an estimation of what the exchange rate between two currencies

would have to be for the exchange to be at par with the purchasing power of the two countries'

currencies. It is a tool for converting (relative) nominal national incomes into real incomes.



United States and Japan, China’s GDP per capita is forecast to be 46 per cent of Japan’s GDP per capita and 33 per cent that of the United States by 2020 (see Figure 2.3).

Figure 2.3: China’s GDP per capita, 1980–2020

China vs India China’s GDP per capita as a percentage of Japan and the US

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

2012

2014

2016

2018

2020

0

5,000

10,000

15,000

20,000

25,000

China India

GDP

per

capit

a (P

PP)

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

2012

2014

2016

2018

2020

0

10

20

30

40

50

Japan US

Per c

ent

Source: IMF database and authors’ calculations

2.2 China’s energy mix is also changing

Over the past decade, China’s rapid economic growth has led it to be the largest energy consumer and producer in the world. China’s rapid economic development has driven ever-increasing energy use, particularly electricity generation.

The nexus between energy, growth and carbon emissions

There is a strong nexus between energy use, economic growth and CO2

emissions. Using fossil fuel–based energy facilitates economic growth, but it also generates an environmental cost. China’s energy consumption and CO2

emissions have been growing steadily since economic reforms commenced in 1979. China has now become the largest carbon emitter in the world.

Growth in electricity generation is one of the key drivers of emissions. China’s electricity use accounted for 24 per cent of global electricity generation in 2014, a three-fold increase since 2000. In comparison, China contributed to 28 per cent of the global CO2 emissions (see Figure 2.4) in 2014. The scale and age of China’s existing coal-fired power generation capacity highlights the risk of high carbon lock-in in its energy supply infrastructure. Much of China’s coal-fired power capacity has been constructed since 2000, meaning that it is technically capable of continuing to operate for decades to come,12 and must do so to yield the expected returns on investment. In addition to high CO2 emissions from power generation, large increases in the production of energy-intensive materials, such as cement and steel, have also driven China’s CO2 emissions.

12 EIA (2015a) China briefing



Figure 2.4: China’s share of global electricity and emissions, 2000–14

2000 2002 2004 2006 2008 2010 2012 2014

8.8 9.5 10.211.4

12.513.6

15.116.5 17.2

18.519.6

21.3 22.023.4 24.0

13.8 14.215.2

17.018.9

20.922.2

23.3 23.525.1 25.3

26.8 27.0 27.4 27.5

Carbon emission (per cent) Electricity generation (per cent)

Source: BP (2015) and authors’ calculations

Concerns about emissions are increasing, and China’s policy makers are facing challenging trade-offs between energy security, cost and environmental outcomes (see Figure 2.5).13 The energy policy trilemma describes the challenge for an economy to simultaneously achieve energy security, access to affordable energy service, and environmentally sensitive production and use.14

Figure 2.5: Energy policy trilemma

Source: Wensley et al. (2013)

The World Energy Council publishes an Energy Trilemma Index that ranks countries in terms of their likely ability to provide sustainable energy policies. Countries are scored based on the three dimensions of the energy trilemma:

13 Wensley et al. (2013) China’s energy demand growth and the energy policy trilemma

14 https://www.worldenergy.org/work-programme/strategic-insight/assessment-of-energy-climate-

change-policy/



Energy security: the effective management of primary energy supply from both domestic and external sources, the reliability of energy infrastructure, and the ability of participating energy firms to meet current and future demand.

Energy equity: the accessibility and affordability of energy supply across the population.

Environmental sustainability: the achievement of supply- and demand-side energy efficiencies, and the development of energy supply from renewable and other low-carbon sources.15

On the Energy Trilemma Index, China is ranked 129th out of 130 in the world for environmental impact mitigation, 79th for energy equity and 21st for energy security. Table 2.1 shows that China’s energy security is relatively strong, but environmental sustainability remains a challenge for China’s rising energy demand.

15 World Energy Council (2015) Energy Trilemma Index



Table 2.1: Energy Trilemma Index, rank for selected countries, 2015

Country Total rank Energy security

Energy equity

Environmental mitigation

Canada 7 1 2 71

United States 12 3 1 95

Germany 13 25 46 44

Australia 17 6 14 110

Japan 32 83 19 49

Brazil 37 43 78 17

China 74 21 79 129

India 107 53 104 122

Source: World Energy Council (2015)

China’s recent energy policy development focuses on reducing both energy intensity and coal dependency, and increasing the use of cleaner fuels such as gas and renewable energy.

Structural change and government policies leading a reduction in energy intensity

Energy intensity is an overall measure of how much energy is used to produce a unit of economic output (i.e. GDP). The goal for China’s sustainable economic growth is to minimise CO2 emissions and energy cost, and maximise energy security. Reducing energy intensity is a key component of that goal.

During the past five years, the trends in CO2 emissions and energy consumption have diverged. China’s energy consumption has increased while carbon intensity has decreased.16 Figure 2.6 shows the levels of economic output, energy consumption and energy intensity in China between 1979 and 2014. In 1979, China’s energy consumption was 0.4 billion tonnes of oil equivalent (btoe) but, by 1994, this had doubled. It has doubled during each decade since. From 1979 to 2014, China’s GDP grew faster than its total energy consumption. As a result, energy intensity has declined significantly. This reflects improved energy efficiencies in industrial production, and a relative decline of energy-intensive activities. Energy intensity halved from 1979 to 1993, and halved again from 1994 to 2014.

16 Carbon intensity is the amount of carbon dioxide generated (in metric tonnes) per unit of

energy consumed (in million tonnes of oil equivalent).



Figure 2.6: China’s economic output, energy consumption and energy intensity, 1979–2014

Economic output vs energy consumption Energy intensity

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

2007

2009

2011

2013

0

1

2

3

4

5

6

Total energy consumption (billion toe) GDP (2005 US$, trillion)

1979

1981 1983 1985

1987 198

9 1991 199

3 1995 1997 199

9 2001

2003

2005 2007

2009 2011 201

3

0.0

0.5

1.0

1.5

2.0

2.5

Billio

n toe

/GDP

(200

5 trill

ion U

S$)


Despite these recent improvements, China’s carbon intensity remains higher than the rest of world (see Figure 2.7).



Figure 2.7: Nexus between energy consumption and CO2 emissions in China, 1979–2014

Energy consumption vs CO2 emissions Carbon intensity

1979 1981 1983 1985 1987

1989 1991

1993

1995

1997 1999

2001

2003

2005

2007

2009 2011

2013

0.0

2.0

4.0

6.0

8.0

Energy consumption CO2 emission

1979

=1

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

2007

2009

2011

2013

2.0

2.5

3.0

3.5

4.0

China Rest of the World

CO2 (

mt)/e

nerg

y con

sump

tion (

mtoe

)

Source: BP (2015) and authors’ calculation

China’s dependence on coal is reducing, but a long way to go

China’s endowment of relatively cheap domestic coal resources makes it difficult to significantly reduce coal use for generating power.17 Electricity generation accounted for more than 50 per cent of China’s total CO2

emissions from fuel combustion, as a result of heavily reliance on coal for electricity. For a long time, coal has dominated China’s energy supply and demand, as a result of its abundance and low cost relative to other fuels. In 2014, China consumed about three btoe of coal, which comprised 66 per cent of China‘s total primary energy consumption. This was 10 percentage points higher than India, 30 percentage points higher than Australia, 38 percentage points higher than Japan and 46 percentage points higher than the United States (see Figure 2.8).

17 IEA (2012) Power Options for Low-carbon Power Generation in China: Designing an emission trading system for China’s electricity sector



Figure 2.8: Comparative total primary energy consumption and coal dependency, selected countries, 2014

Energy consumption Coal dependency

China US India Japan Germany S. Korea UK Indonesia S. Africa Australia0

500

1,000

1,500

2,000

2,500

3,000

mtoe

S. Africa China India Australia Indonesia S. Korea Japan Germany US UK0

10

20

30

40

50

60

70

80

Per c

ent


In 2014, coal accounted for 72 per cent of the electricity generation mix, although it has declined from 81 per cent in 2007. Despite this decline, the electricity output generated by coal increased from 2.7 trillion kWh in 2007 to four trillion kWh in 2014.

China now produces and consumes almost as much coal as the rest of the world combined. In 2014, China accounted for about 50 per cent of global coal consumption. Figure 2.9 shows that natural gas and nuclear consumption accounted for five per cent of the world total, respectively, in 2014. China’s oil consumption accounted for 12 per cent of the world total and renewable resources (including hydro) accounted for 44 per cent of the world total. China’s role in driving global trends is changing as it enters a much less energy–intensive phase in its development.



Figure 2.9: Share trend of China's fuel consumption in the world’s total, 1966–2014

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

2012

2014

0

10

20

30

40

50

60

Coal Oil Gas Nuclear Hydro Others*

Per c

ent

Notes: Others include solar, wind, geothermal and biofuels


The high level of domestic coal dependency in China is partly a result of the imbalance in the endowment of fossil fuels in China, with coal being more prevalent compared with other fossil fuels. Concerns for self-sufficiency and energy security, along with the low cost of coal, have led to high levels of coal usage and high rates of production from the domestic resource. Coal self-sufficiency, as measured by the ratio of domestic coal production to consumption, was 94 per cent in 2014. Although countries such as Indonesia and the United States are completely self-sufficient in coal supply and are net exporters, coal is a much smaller percentage of their total primary energy consumption because large volumes of other fossil fuel resources are available. In addition, higher gas prices in China make coal-to-gas switching far less attractive in comparison with constructing high-efficiency coal-fired plants.18

Despite the concern for self-sufficiency, coal’s share in China’s total energy consumption has fallen gradually from 87 per cent in the mid-1960s to 66 per cent in 2014, with a rapidly increasing share of natural gas and renewables in the primary energy mix (particularly since 2000). From 2000 to 2014, gas consumption increased seven-fold, as did nuclear power generation. Hydroelectric power generation increased five-fold, whereas other renewables increased 74-fold (once again from a very low base).

18 IEA (2015a) Energy and Climate Change: World Energy Outlook Special Report



Complexity in energy policy

China’s recently developed energy policies are focused on mitigating the environmental impact of air pollution and climate change. Since 2013, China has been pursuing the targets for addressing climate change set out in its 12th Five-Year Plan — implementing the action plan for controlling greenhouse gas emissions, adjusting the country’s industrial structure and increasing energy efficiency. Policies for energy efficiency include upgrading low-carbon technology, undertaking innovation and attempting to resolve overcapacity. In October 2013, the General Office of the State Council issued the Opinion on Further Strengthening Coal Mine Safety, proposing to close more than 2,000 small coal mines nationwide by the end of 2015.19 China is implementing an action plan, released by the National Energy Administration in 2015, for the clean and efficient use of coal between 2015 and 2020. The plan includes increasing coal quality and controlling residential coal use. To improve coal quality, China will invest in large-scale coal-washing capacity to ensure that 70 per cent of raw coal is washed by 2017 and more than 80 per cent by 2020, from around 40 per cent in 2015.20

In 2014, the Chinese government announced strengthened national action to address these issues. Premier Li Keqiang declared a ‘war on pollution’. President Xi Jinping called for an ‘energy revolution’, to tackle not only demand and supply bottlenecks and innovation, but also the environmental impacts of the production and consumption of energy. In the final quarter of 2014, three new national policy plans set detailed targets for 2020 on energy saving, energy strategy and climate change. A high-profile China–United States joint statement by President Xi and President Obama committed China to unilateral and bilateral action to peak CO2 emissions by 2030.

Under the 2014–20 Plan on Upgrading and Reforming Energy Saving and Emissions Reduction in Coal-fired Electricity Generation,21 the share of coal in Chinese primary energy consumption is targeted to fall below 62 per cent by 2020 — down from 66 per cent in 2013. China has taken steps to develop and construct highly efficient coal-fired power plants, and retire some of its most inefficient coal-fired power plants.22 New standards have been set for coal-power generation fleets, so that 28 per cent of coal-fired electricity generation should be combined heat and power (CHP) by 2020. Coal power plants with a capacity of more than 600 megawatts (MW) are required to achieve the efficiency target of 300g of coal equivalent per kilowatt hours (kWh) by 2020. Any new development of coal-fired power plants will no longer be approved in the major population centres of Beijing, Tianjin, the Yangtze River Delta, and Pearl River Delta regions, unless implemented with CHP. Beijing has announced that it will replace all coal-fired power with natural gas plants.

19 NDRC (2014b) 2014–20 National Plan on Climate Change

20 http://en.sxcoal.com/117736/DataShow.html

21 NDRC (2014a) Plan on Upgrading and Reforming Energy Saving and Emissions Reduction in Coal-fired Electricity Generation

22 Ibid



The Energy Development Strategic Action Plan 2014–20 reiterates the aim of the 12th Five-Year Plan to cap China’s primary energy consumption at 4.8 billion tonnes of standard coal equivalent per year by 2020.23 To achieve this, annual coal consumption will be held at 4.2 billion tonnes until 2020 (approximately 16 per cent above 2014 levels). The use of natural gas is expected to expand to about 10 per cent of primary energy consumption — in part by replacing coal in cooking and using heavier fuels for transportation. This gas objective will be supported by increased conventional and unconventional resource exploration and a target for pipeline infrastructure to total 120,000 km by 2020.

The 2014–20 National Plan on Climate Change aims for a 40–45 per cent cut from 2005 levels in CO2 emissions per unit of GDP by 2020.24 The industrial sector will play a major role in reducing emissions. Industry is expected to cut emissions by about 50 per cent per unit of GDP, and total CO2 emissions from the steel and cement sectors are expected to stablise at 2015 levels by 2020. The share of non-fossil fuels in primary energy consumption should reach 15 per cent by 2020, which will require approximately 190 terawatt hours (TWh) of renewable and nuclear power generation per year until 2020.

On 30 June 2015, China submitted its Intended Nationally Determined Contribution. China aims to employ its best efforts to achieve:

a peak in carbon emissions by 2030 at the latest

a continued reduction in carbon intensity, targeting a 60–65 per cent fall in the economy’s emission intensity by 2030 relative to 2005 levels (or conversely, increasing the amount of economic output per tonne of carbon by almost two-thirds)

an increase in non-fossil energy sources to represent at least 20 per cent of total primary energy use by 2030.

China also plans to start its national emissions trading system in 2017, which covers key industries such as iron and steel, power generation, chemicals, building materials, paper making, and non-ferrous metals. China has also committed to promote low-carbon buildings and transportation. By 2020, China’s goal is to have 50 per cent of newly built buildings be ‘green’ in cities and towns, and 30 per cent of motorised travel be on public transport in big- and medium-sized cities. It will finalise the next stage of fuel-efficiency standards for heavy-duty vehicles in 2016, and implement them in 2019.

Despite its small share of China's total energy mix, natural gas is becoming increasingly important as a result of a number of emerging policy priorities, including a greater emphasis on lowering air pollution and carbon emissions. Further exploitation of natural gas plays a substantial part in the government’s response to growing air pollution issues, and the Action Plan

23 The State Council of the People’s Republic of China (2014) The Energy Development Strategic Action Plan 2014–20

24 NDRC (2014b) 2014–20 National Plan on Climate Change



builds upon the 12th Five-Year Plan for Natural Gas Development,25 including:

Accelerate the development of natural gas and renewable energy in order to realise a clean energy supply and diversified energy mix.

Combine national natural gas pipeline networks, regional pipeline networks, LNG terminals, gas storages and other natural gas distribution projects to strengthen natural gas infrastructure construction in key regions.

Optimally allocate and use natural gas as well as develop a distributed natural gas system in accordance to the rules of the priority development of city gas, active adjustment of the industrial fuel structure, and modest development of natural gas power generation.

In China, power generation and manufacturing are the largest consumers of coal and, thus, the largest CO2 emitters. One of the key energy policy objectives in China is to reduce coal dependency in its primary energy and electricity generation mix. The Chinese government has shut down less-efficient small- and mid-sized coal plants, replacing them with large, high-efficiency units.26 In addition to making coal plants more efficient, these policies also aim to encourage a shift to alternative fuels including natural gas, which can help to reduce CO2 emissions and urban air pollution.

25 The State Council of the People’s Republic of China (2014) The Energy Development Strategic Action Plan 2014–20

26 IEA (2015a) Energy and Climate Change: World Energy Outlook Special Report



3. Natural gas in China

3.1 Consumption growing, but still low

In 2014, China was the world’s third largest gas consumer, trailing only the United States and Russia. This represents rapid growth in demand since 2000, when China ranked 21st in the world. Although the share of natural gas consumption is increasing, it remains a small share of primary energy consumption (as shown in Figure 3.1), accounting for only 5.6 per cent of China's primary energy use in 2014. This is about a quarter of the global average rate (about 24 per cent).

Figure 3.2: Trend of primary energy mix by fuel type in China, 1966–2014

Primary energy mix by fuel type: fossil fuels Primary energy mix by fuel type: non-fossil fuels

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

2012

2014

0

20

40

60

80

100

Coal Oil Gas

Per c

ent

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

2012

2014

0

2

4

6

8

10

Nuclear Hydro Others*

Per c

ent

Notes: Other includes solar, wind, geothermal and biofuels


3.2 Domestic production surging

Domestic production also surged during this time. China ranked sixth in the world in total gas production in 2014, increasing its production five-fold between 2000 and 2014. Its share of proven reserves is small, however, with less than 2 per cent of the world reserves, ranking 13th after Australia and Iraq (see Table 3.1).

China’s indigenous gas production is predominantly conventional gas, which accounted for more than 90 per cent of its total natural gas production in 2014. However, its unconventional gas production has huge potential.27 Like many other countries, such as the United States and Russia, China has rich unconventional gas resources.

27 Unconventional gas refers to gas produced from coal seams (coal seam gas or coalbed

methane), shale rocks (shale gas), and rocks with low permeability (tight gas). Once gas is

produced from these reservoirs, it has the same properties of gas produced from ‘conventional’

(i.e. sedimentary reservoirs with high porosity and permeability) sources.



Table 3.1: Global natural gas reserves, production and consumption, 2014

Proven reserves Per cent Production Per cent Consumption Per cent

Iran 18.2 United States 21.0 United States 22.4

Russian Federation 17.4 Russian Federation 16.7 Russian Federation 12.1

Qatar 13.1 Qatar 5.1 China 5.5

Turkmenistan 9.3 Iran 5.0 Iran 5.0

United States 5.2 Canada 4.7 Japan 3.3

Saudi Arabia 4.4 China 3.9 Saudi Arabia 3.2

United Arab Emirates 3.3 Norway 3.1 Canada 3.1

Venezuela 3.0 Saudi Arabia 3.1 Mexico 2.5

Nigeria 2.7 Algeria 2.4 Germany 2.1

Algeria 2.4 Indonesia 2.1 United Arab Emirates 2.0

Share of total 79.1 Share of total 67.2 Share of total 61.2

Notes: 1. The BP statistics do not fully cover proven reserves for unconventional natural gas reserves and production for some countries, including China. 2. Total proven reserves of natural gas are generally taken to be those quantities that geological and engineering information indicates, with reasonable certainty, can be recovered in the future from known reservoirs under existing economic and operating conditions. Source: BP (2015) and authors’ calculations

China’s conventional gas reserves are estimated to be 44 trillion cubic metres (tcm), which is the 13th in the world, and second largest (after Australia) in the Asia–Pacific region. These reserves are dominated by shale gas (more than 30 tcm), which accounts for almost three-quarters of the total, and coalbed methane, which accounts for 9.2 tcm in 2015 (Figure 3.2).

Figure 3.3: Unconventional gas resources in China, 2015

Unconventional gas vs selected countries China: unconventional gas by type

Canada

Australia

US

Russia

China

0 5 10 15 20 25 30 35 40 45 50

21.6

27.9

28.4

32.7

43.8

tcm

Shale gas CBM Tight gas0

5

10

15

20

25

30

35

31.6

9.2

3

tcm

Source: IEA (2015c)



Most of China’s shale gas reserves are located in Sichuan and Xinjiang (Tarim Basin) (see Map 3.1). Sichuan is densely populated and has a high level of agricultural activity with a high demand for water. It also has unstable geological conditions. Extraction and transportation costs are therefore high. Tarim Basin is located close to the Kazakhstan border and far away from the main consumption centres of natural gas in China which will require costly investment in long-distance pipelines.28

China's primary onshore natural gas–producing regions are:

Sichuan province in the southwest (Sichuan Basin)

the Xinjiang and Qinghai provinces in the northwest (Tarim, Junggar and Qaidam basins)

Shanxi province in the north (Ordos Basin).

Box 3.1: China's shale oil and gas basins

Source: Wayne (2012)

China has delved into several offshore natural gas fields located in the Bohai Basin and the Panyu complex of the Pearl River Mouth Basin (South China Sea), and is also exploring more technically challenging areas (including deep water, coalbed methane and shale gas reserves) with foreign firms.

China is currently the largest shale gas producer outside North America, but it faces significant challenges in developing its shale resources. The shale is deeper (up to 6,000 meters below ground) and tends to have more clay than United States shale. In addition, scarce water reserves in the Ordos and Tarim basins29 — where many shale gas beds lie — increase the cost of

28 Dehnavi and Yegorov (2014) Future of Shale Gas in China and its Influence on the Global Markets for Natural Gas

29 IEA (2014a) The Asian Quest for LNG in a Globalising Market



extraction in China. The sector in China is dominated by national oil companies. Although these companies are partnering with selected international firms, it is likely to take considerable time to develop effective technological solutions and commercial arrangements to produce and supply large-scale shale gas to the Chinese market.30

As Chinese firms have gained experience producing from shale, the cost of shale gas drilling has declined. By mid-2015, the cost of drilling a horizontal well in shale formations in the Sichuan Basin was between US$11.3 million and US$12.9 million per well, according to the China’s National Petroleum Corporation's Economics and Technology Research Institute.31 Sinopec, one of China's national oil firms, reports that this range was 23 per cent lower than in 2013.32 However, the cost of drilling one shale gas well in China is still more than double the drilling cost in the United States,33 which is about US$4 million per well and up to around US$7 million for more complex wells.34

There are several reasons for the higher costs of production in China, including limited economies of scale and complex geologies. For example, in southern China shale gas production is in the mountainous terrain, and water shortages pose problems in western China. Water scarcity in particular (e.g. high groundwater stress and seasonal variability) may make the cost of drilling prohibitive in the Tarim Basin in Xinjiang province, where China’s second-largest shale gas play is located.35

Declining well costs and increasing experience in developing shale gas reserves, supplemented by government investment incentives, have promoted further investment in the development of shale gas resources. In 2012, to encourage the exploration for shale gas, the Chinese government established a four-year, $1.80 per million British Thermal Units (mmBtu) subsidy programme for any Chinese firm achieving commercial production of shale gas. In mid-2015, this subsidy was extended to 2020, but at a lower rate.36 Enhanced financial incentives for investment are provided as part of the designation of shale gas as one of the nation's strategic emerging industries. However, uncertainties regarding future liberalisation of prices and third-party access, along with the absence of detailed rules to regulate shale gas activity, are among a number of factors holding back further growth.

30 Sheehan et al. (2014) The National and Regional Development of China’s Gas Market: Beyond evolutionary change?

31 Jin (2015) Opportunities and challenges of shale gas development in China: a perspective from Sinopec’s experience

32 Ibid

33 Dehnavi and Yegorov (2014) Future of shale gas in China and its influence on the global markets for natural gas


35 Wang (2015) China’s elusive shale gas boom

36 EIA (2015b) Shale gas development in China aided by government investment and decreasing well cost



Shale gas will not become a major energy source for China in the short term because of a range of technical, institutional and infrastructure constraints.37 In the long term, shale gas is expected to make a major contribution to China’s natural gas supplies, and provide the benefits of lower costs, energy security and environmental protection.

3.3 Import reliance growing, but many options

Unlike other countries in the Asia–Pacific region, such as Japan and South Korea, that are almost entirely dependent on LNG for their gas supplies, China has multiple sources of supply. It can source gas from its own indigenous resources, or import natural gas as LNG or through pipelines. Geographically, China it is well positioned to access foreign natural gas supplies both by means of marine transport from the Asia-Pacific and the Middle East region and by pipeline transport from gas-rich regions such as Central Asia and Russia.

The costs of pipelining natural gas benefit substantially from economies of scale, since large diameter pipelines carry significantly more gas than smaller diameter pipelines but at a proportionate lower cost. Pipeline costs rise linearly with distance, but, LNG — which requires liquefaction and regasification regardless of the distance travelled — has a high threshold cost, but a much lower increase in cost with distance. Thus, shorter distance transport tends to favor pipelining of natural gas, but longer distances favor LNG.

China’s strong growth in demand for natural gas has outpaced increases in domestic production, leading to greater imports sourced through both pipeline gas and LNG. In 2014, China was the world’s third largest LNG importer and the world’s sixth largest importer of pipeline gas (see Figure 3.3).

37 Liu (2014) Shale gas development and challenges in China



Figure 3.4: Top LNG and pipeline exporters and importers, 2014

LNG Pipeline gas

Qatar

Malaysia

Australia

Nigeria

Indonesia

Taiwan

India

China

South Korea

Japan

-110 -60 -10 40 90 140

Exports Imports

bcm

Russia

Norway

Canada

Netherlands

China

UK

Turkey

Italy

US

Germany

-200 -150 -100 -50 0 50 100

Exports Imports

bcm


Pipeline gas is imported from Central Asia (such as Turkmenistan, Uzbekistan and Kazakhstan) in the west, from Myanmar (from offshore fields in the Andaman Sea) in the south, and from Russia in the north and north-west. On 21 May 2014, the decade-long negotiation with Russia on gas supply reached an agreement. Gazprom will provide 38 bcm/year from eastern Siberia to China’s Bohai Bay region for 30 years, expected to start in early 2020. Six months later, a memorandum of understanding was signed between Beijing and Moscow to deliver a further 30 bcm of gas for 30 years from the western route, which is also known as the Altai pass. LNG is delivered to China’s eastern seaboard, mainly from the Middle East and Asia–Pacific regions, including Australia.

Growth in regasification capacity is an indicator of potential changes in LNG demand. China’s regasification terminals are built on three coasts near major seaports:

south coast imports go to Guangdong, Shandong, Hainan and Guangxi provinces, and Zhejiang and Shenzhen cities



east coast imports go to Shanghai, Fujian, Jiangsu and Lioning provinces, and Lianyungang city

north coast imports go to Hebei province and Tianjin city.

Although China’s south coast started its first LNG import in 2006, the regasification capacity in the area accounts for more than 42 per cent of China’s total existing regasification capacity, compared with 39 per cent in the east coast and 18 per cent in the north coast. This trend is expected to continue with the majority of new regasification capacity being constructed and planned in the south coast (see Table 3.2). This reflects that the south coast is closer to the LNG-exporting countries such as Australia. In the north, LNG imports have stronger competition from pipeline gas. Table 3.2 also shows that, by 2020, the total regasification capacity is likely to be three times the current level.

Table 3.2: LNG regasification capacity in China

Location Total capacity

Existing Construction Planned or proposed*

South coast 57.12 18.08 15.08 23.96

East coast 44.62 16.80 11.46 16.36

North coast 24.2 7.82 4.14 12.24

Total 125.94 42.70 30.68 52.56

Notes: * Planned and proposed regasification capacities that are assumed to start their

operations by 2020

Source: Nexant (2015) WGM and authors' calculations

Based on the IEA’s study in 201538, the increase in LNG imports will be led by China and the non-Organisation for Economic Co-operation and Development countries in Asia and Europe. These countries account for more than 90 per cent of incremental additions. On the supply side, new LNG supplies will come primarily from Australia and the United States, which will account for 90 per cent of additional LNG exports between 2014 and 2020.

Relatively slow growth in China’s gas production and rapidly increasing gas consumption has led to increasing gas imports during the past years. China began importing LNG in 2006. Imports of pipeline natural gas from Central Asia began in 2010, followed by imports from Myanmar in 2013. In 2014, China’s natural gas imports were more than 50 bcm. In 2007, imported LNG was around 4 bcm, and increased seven-fold from 2007 to 2014 to more than 27 bcm. Imports of pipeline gas in 2010 were less than 4 bcm, but increased eight-fold to more than 31 bcm during the period of 2010 to 2014. In 2014, pipeline gas imports accounted for 53 per cent of total imports of natural gas compared with 22 per cent in 2010. Import dependency has increased from 2 per cent in 2007 to around 30 per cent in 2014.

From 2000 to 2014, China’s gas consumption showed a seven-fold increase with a compound annual growth rate (CAGR) of more than 15 per cent

38 IEA (2015b) Medium-Term Gas Market Report 2015: Market analysis and forecasts to 2020



(Figure 3.4). During the same period, China’s gas production increased five-fold, with a CAGR of around 12 per cent.

Figure 3.5: Natural gas balance, 2000–14

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014-40

0

40

80

120

160

200

Export & Import Production Consumption

bcm

Demand > Supply

Demand < Supply




4. Policy scenario one — an increase in the service sector

Policy scenario one posits an increasing share of the service sector in China’s economy. This has immediate implications for overall economic growth, income and energy intensity.

In this scenario, economic growth is stimulated, meaning that both industry and households can afford, and are willing to pay for, cleaner fuel such as gas in industrial activities and for urban living. The modelling shows that, as total energy consumption increases, the share of gas consumption increases, which leads to higher gas demand, increased gas production and a faster growth in imports than in the baseline scenario.

4.1 The effect of scenario one on natural gas consumption

When the share of the service sector increases, the overall economy becomes richer and society shifts its focus towards consumption and away from savings. A further consequence of increased income is that households will switch from using low-grade coal for heating to gas. This gives a significant opportunity for future growth of gas demand, especially in the wealthy coastal regions.39 An increasing share of the service sector in the total economy stimulates total energy consumption and GDP growth relative to the baseline during the forecast period. Despite a reduction in energy intensity, the higher incomes seen as a result of scenario one mean that households will prefer cleaner fuels in the form of gas or non-fossil fuels for their urban living compared with the baseline scenario.

Figure 4.1 shows that China’s total energy consumption is projected to increase from 2,956 mtoe in 2015 to 4,379 mtoe in 2030. During the next 15 years, the growth in consumption of natural gas and non-fossil fuels is faster than the consumption growth for coal and oil. The compound annual growth rate (CAGR) for the forecast period of 2015–30 for both natural gas and non-fossil fuels is expected to be 5 per cent, compared with two per cent for the growth rate of coal and oil. By 2030, there is a cumulative increase in non-fossil fuel and gas consumption of nine per cent and seven per cent, respectively, relative to the baseline. This compares with cumulative declines of three per cent and four per cent for coal and oil consumption, respectively, by 2030, relative to the baseline.

39 Li (2015) Natural gas in China: a regional analysis



Figure 4.1: Primary energy consumption in policy one, 2015–30

Primary energy consumption by fuel type in policy one Cumulative changes in fuel consumptionrelative to baseline

2015 2020 2025 20300

1,500

3,000

4,500

Coal Oil Gas Non-fossil fuel

mtoe

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030-6

-4

-2

0

2

4

6

8

10

Coal Oil Gas Non-fossil fuel

Per c

ent

Source: CHINAGEM

To meet the increased demand for natural gas in this policy scenario, China’s gas imports are forecast to grow faster than natural gas production relative to the baseline. Figure 4.2 shows that natural gas consumption is projected to double in the next 15 years to 392 billion cubic metres (bcm) in 2030. To meet this increased gas demand, natural gas imports and production will double to 135 bcm and 257, respectively, by 2030. From 2015 to 2030, unconventional gas production will increase seven-fold from a low base, and its share in total gas production will rise to 60 per cent in 2030 from 12 per cent in 2015. In the policy scenario alone, the cumulative gas imports grow by 13 per cent compared with the baseline by 2030, which is much faster than the cumulative growth of 5 per cent for indigenous gas production.



Figure 4.2: Natural gas demand and supply in policy one, 2015–30

Natural gas supplies in policy one Cumulative changes from baseline

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 20300

100

200

300

400

Conventional gas Unconventional gas Pipeline gas import LNG import

bcm

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 20300

3

6

9

12

15

Gas production Gas imports

Per c

ent

Source: CHINAGEM

4.2 The implications for energy intensity

In this section, results from the CHINAGEM modelling are used to provide insights on how a larger service sector might affect changes in energy intensity and gas consumption.

In the baseline scenario, real GDP grows at around six per cent per year, and there is fairly uniform expansion in all sectors, such that the shares of each sector change little over time. In the policy simulation, exogenous changes are made to the economy to lift the share of the service sector in the GDP to be about five per cent higher in 2030 relative to the baseline scenario. In the baseline, the service sector’s share is around 54 per cent in 2030. In the policy simulation, the service sector’s share is around 59 per cent in 2030 (see Table 1.2). Comparing results for the policy simulation with those of the baseline yields projections of the effects of the increase in the share of the service sector on key variables, including energy use and energy intensity.

Increasing the share of output generated from services benefits overall employment, because the service sector is expected to generate more jobs



per unit of GDP than industry.40 This leads to rising aggregated income and increasing private consumption.

In this scenario, real household consumption grows faster than real GDP, relative to the baseline (Figure 4.3). Annual growth of real household consumption deviates by 1.6 percentage points relative to the baseline, compared with 0.6 percentage points for real GDP. Figure 4.3 also shows that the growth in investment expenditure starts to decline from 2019, despite faster growth in government demand and investment expenditure in this policy scenario. At the same time, growth in government demand continues to rise until about 2023, when it plateaus. By 2030, annual investment expenditure is 0.5 percentage points higher than for the baseline, while government demand is 1.5 percentage points higher.

Figure 4.3: Cumulative growth deviation for macroeconomic indicators in policy one from baseline, 2015–30

GDP vs household consumption Investment vs government demand

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 20300.0

0.5

1.0

1.5

2.0

Real household consumption

Real Gross national product

Perc

entag

e poin

ts

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 20300.0

0.5

1.0

1.5

2.0

Real investment expenditure Real government demands

Perc

entag

e poin

ts

Source: CHINAGEM

The service sector is less energy intensive than the secondary sector, which includes manufacturing. The secondary sector in China relies much more on coal as a source of energy than the service sector.

This policy scenario reflects a choice that the Chinese government may make in order to support sustainable economic growth and a reduction in the relative share of energy consumption. This results in lower energy intensity.

40 Rutkowski (2015) Service sector reform in China



Figure 4.4 shows that energy intensity will be reduced from 528 toe per million dollars of GDP (in 2015) to 436 toe per million dollars in 2020 and 298 toe per million dollars in 2030. China’s energy intensity reduces by three per cent in 2020 and seven per cent in 2030 relative to the baseline. For example, with a reduction of total primary energy consumption, GDP growth in this policy scenario increases by 0.6 percentage points per year in 2020 and 0.4 percentage points per year in 2030, relative to the baseline.

Figure 4.4: Changes in energy intensity and deviation from baseline, 2020 vs 2030

Changes in energy intensity: policy one vs baseline Deviation of energy intensity and consumption from baseline

2020 20300

100

200

300

400

500448

322

436

298

Baseline Policy one

toe/m

illion r

eal G

DP

Energy intensity Energy consumption-10.0

-5.0

0.0

5.0

10.0

-2.6

-1.0

-7.3

-0.7

2020 2030

Per c

ent

Notes: Energy intensity is measured by primary energy consumption in toe per million dollars of GDP in 2005 prices

Source: CHINAGEM



5. Policy scenario two — capping coal consumption by 2020

The second policy scenario involves slowing the growth rate in coal consumption, with a cap on total coal consumption in place from 2020. This reflects a potential policy goal to further reduce CO2 emissions through aggressive reductions in coal consumption. The growth rate of primary coal consumption gradually decreases from 2015 and is zero after 2020. Total coal consumption peaks at 4.2 billion tonnes in 2020. This peak is assumed to be achieved by a demand-side policy intervention using the government’s regulatory powers.

Setting 2020 as the year for China’s coal consumption to peak is one of the policies designed to reduce CO2 emissions and air pollution. Although a national cap on coal consumption builds on other efforts to protect the environment, it also conserves resources and provides a basis for future growth in clean energy industries. In the absence of other policies, peaking coal consumption by 2020 results in a reduction in both total energy consumption and GDP growth relative to the baseline.

This policy has a number of implications for economic growth and energy consumption in China. Firstly, the modelling shows that GDP growth is marginally lower than in the baseline scenario, as this scenario lacks any countervailing drivers of economic growth, such as a shift to a larger service sector. Secondly, total energy demand is also lower than the baseline, although there is growth in natural gas consumption of around 25 per cent by 2030 relative to baseline as a result of switching from coal to natural gas and alternative fuels.

Of particular note is the shift in gas consumption among sectors. The results show an increase in consumption overall in all sectors, but the relative share of gas consumption increases in emerging sectors, such as the manufacture of communications equipment. A further consequence of this policy is a significant reduction in CO2 emissions and in the emission intensity of the economy.

5.1 The effect of scenario two on natural gas consumption

In this policy scenario, the growth rate of primary coal consumption gradually decreases from 2015 and is zero after 2020. Total coal consumption peaks by 2020, which is assumed to be achieved by a demand-side policy intervention using the government’s regulatory powers.

Figure 5.1 shows that coal consumption in this policy scenario reduces by 128 mtoe in 2020 and by 611 mtoe in 2030, relative to the baseline. As a result, projected coal consumption is 23 per cent lower than in the baseline scenario in 2030, whereas projected gas consumption is 25 per cent higher. The simulation results also show that GDP growth is marginally lower relative to the baseline in 2030.

This study does not explicitly examine how introducing a tax on coal and energy pricing will affect the coal share in the primary energy mix and CO2



emissions. The revenues from a coal tax could partly be applied towards clean energy development, with a higher potential for fueling economic growth.41 China is set to introduce an emissions trading scheme in 2017 covering the power sector and heavy industry. This scheme will help to curb the appetite for coal, and slow the rate of growth in China’s CO2 emissions before they peak around 2030.42 Nevertheless, taxes and price will influence investment decisions on coal and other alternative fuels.

Figure 5.1: Coal and gas consumption in policy two, 2015–30

Changes in coal and gas consumption in policy two Cumulative deviation of coal and gas consumption from baseline

2020 2025 2030-700

-600

-500

-400

-300

-200

-100

0

100

200

Coal Gas

mto

e

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030-30

-20

-10

0

10

20

30

Coal Gas

Per c

ent

Source: CHINAGEM

Achieving a national coal cap would depend predominantly on the industrial sector. Key coal-consuming industries include power, iron and steel, cement, and building. In the industrial sector, natural gas typically competes against coal, oil products and electricity. Natural gas is also used by industry for non-energy purposes, mainly as a feedstock for the manufacture of fertilisers and petrochemical products.

The top 10 traditional gas users in China are:

extraction of petroleum and natural gas

41 Green and Stern (2015) China’s ‘new normal’: structural change, better growth, and peak emissions




mining of ferrous metal ores

coking

manufacture of basic chemical raw materials

manufacture of fertilisers

manufacture of synthetic materials

manufacture of chemical fibres

rolling of steel



electricity generation by gas

gas supply (to residential and transport sectors).

In policy scenario two, the gas used in non-energy sectors as an intermediate input increases for non-traditional industrial gas users (relative to the baseline), and replaces petroleum products. This is coupled with a reduction in the use of gas in some energy-intensive sectors, such as steel and coking production. Replacing petroleum products with gas products reflects that, in the transport sector, use of natural gas vehicles can significantly improve air quality, because natural gas vehicles have lower NOx and SOx emissions. Using natural gas for transport could reduce petrol and diesel consumption, which is a key driver of China's oil products demand.

Zero growth in primary coal consumption after 2020 has a major impact on natural gas consumption for industrial users. Figure 5.2 shows the 10 industries with the largest growth in natural gas consumption in 2030, relative to the baseline. Most are non-traditional or emerging industrial gas sectors, such as hotels, restaurants and computers, and manufacturing that is less energy intensive. Figure 5.2 also shows the cumulative change in gas consumption for the 10 largest traditional industrial gas users. Gas consumption is nine per cent higher than the baseline, but the share of these users’ gas consumption is 11 per cent lower by 2030. The top 10 traditional industrial gas users account for 71 per cent of total industrial gas consumption by 2030 under this policy scenario, which represents a decline from 79 per cent in the baseline scenario.

Figure 5.2: Cumulative deviation from baseline of primary gas used by industries in policy two

Deviation for 10 emerging industrial gas users, 2030 Deviation trend for the top 10 traditional industrial gas users, 2015–30

ElecCommsEqp

Hotels

FishProc

VegetOils

Ships

ElctronParts

Computers

Restaurant

GasSupply

ElecGas

0 50 100

Per cent

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030-15

-10

-5

0

5

10

Gas consumption Share of gas consumption

Per c

ent

Notes: ElecGas refers the sector of electricity generated by gas. Restaurant is the sector of catering services. Computers refers

the sector of Manufacture of Computers. ElctronParts refers the sector of Manufacture of Electronic Component. Ships refers

the sector of Manufacture of Boats and Ships and Floating Devices. VegetOils refers the sector of Refining of Vegetable Oil.

FishProc refers the sector of Processing of Aquatic Product. ElecCommsEqp refers the sector of Manufacture of

Communication Equipment.

Source: CHINAGEM



5.2 The implications for emission intensity

A policy of peaking coal consumption in 2020 will reduce total carbon emissions from coal, as the reduction in emissions from coal is greater than the increase in emissions from gas. In this policy scenario, by 2030:

total carbon emissions decline to 11 billion tonnes from the baseline 13.6 billion tonnes

- the projected decline of carbon emissions from coal consumption is more than two-fold, but the carbon emissions from natural gas doubles relative to the baseline

total emissions intensity declines by 61 per cent (from 2007) compared with a deduction of 57 per cent in the baseline.

Figure 5.3 shows that, by 2030, coal-based carbon emissions in this scenario are 5.6 billion tonnes lower than the baseline, and natural gas-based carbon emissions are 0.7 billion tonnes higher than the baseline. Gas consumption is higher — not only in power generation, but also in industrial and residential uses. The small increase in CO2 emissions from petrol refineries reflects a high level of substitution between coal, and gas and renewables, and an upsurge in the small amount of oil used for primary energy consumption. As a result, emissions intensity is 17 per cent lower and carbon intensity (defined as the ratio of carbon emissions to total energy consumption) is eight per cent lower than the baseline by 2030.



Figure 5.3: Carbon emissions and carbon intensity in policy two, 2015–30

Changes in carbon emissions, selected years Cumulative deviation of emission intensity and carbon intensity from baseline

2015 2020 2030-8

-6

-4

-2

0

2

4

-2.3-3.1

-5.6

0.3 0.4 0.7

2.0 2.2 2.3

Coal Gas Petrol refining

Billio

n to

nes

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030-18

-16

-14

-12

-10

-8

-6

-4

-2

0

2

Emission intensity Carbon intensity

Per c

ent

Source: CHINAGEM



6. Policy scenario three — higher unconventional gas production

The third policy scenario involves an increase in unconventional gas production. The increased unconventional gas production reflects technological improvements and increases in investment for exploration and infrastructure. These factors reduce the average cost of gas production and increase productivity as more unconventional gas is extracted.

This will have implications for both the level of total gas supply and the quantity of natural gas imports into China. Given the current state of technology in the production of unconventional gas in China, there is significant scope for productivity gains over time. Any productivity improvement lowers the unit cost of production and increases the competitiveness of unconventional gas in the national gas market. As the competitiveness of unconventional gas improves, so does its share in the national market.

One result of accelerating unconventional gas production is a reduction in supply from conventional gas sources relative to the baseline. Nevertheless, total indigenous gas production is higher in this scenario, as gas demand is stimulated by the reduction in gas production costs. However, demand does not grow as fast as production and, hence, there will be a decline in natural gas imports. This means that, in this scenario, the import dependency of natural gas will decline relative to the baseline.

6.1 The effect of scenario three on natural gas supply

The increase in indigenous unconventional gas production leads to technological improvements and applications, and increases the scale of economies involved in drilling and infrastructure. The increase in production comes partly at the expense of reduced production from conventional sources. This change in the relative quantity produced is in response to a change in the relative price of output. The change in relative price lowers the price of unconventional gas relative to conventional gas through productivity improvements, which reduces the average cost of production of unconventional shale oil gas.

Figure 6.1 shows how unconventional gas production rises and conventional gas production declines post-2020. The production of unconventional gas overtakes that of conventional gas in 2030, accounting for 55 per cent of total gas production, compared with the current share of 10 per cent. This is 15 percentage points higher than the baseline share by 2030. Figure 5.2 also shows that conventional gas production grows in the short-run to 2020, but declines relative to baseline in the long-run to 2030. In all scenarios, unconventional gas production grows faster than conventional gas production.



Figure 6.6: Changes in indigenous gas production: conventional vs unconventional, 2015–30

Gas production in policy three Gas production growth: policy three vs baseline

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 20300

50

100

150

200

250

300

Conventional gas Unconventional gas

bcm

2015-2020 2021-2030-5

0

5

10

15

20

25

3.8

0.4

4.4

-1.3

15.6

12.4

19.0

15.8

Conventional gas (baseline) Conventional gas (policy three) Unconventional gas (baseline)

Unconventional gas (policy three)

Per c

ent

Source: CHINAGEM

6.2 The implications for natural gas imports

The modelling results show that a policy to accelerate production of unconventional gas would lead to greater overall gas production and greater gas demand. However, the rate of growth of gas demand will trail the growth in production. This means that gas imports will grow marginally slower than in the baseline scenario.

Both indigenous gas production and gas imports will grow steadily in the next 15 years, but import growth is lower. Figure 6.2 shows that total gas production in scenario three will reach 286 bcm in 2030 compared with 246 bcm in the baseline, and that total gas imports will decline to 105 bcm from 119 bcm. The gas imports under this scenario are 12 per cent lower and gas production is 16 per cent higher than in the baseline scenario by 2030.



Figure 6.7: Changes from baseline in gas imports and production in policy three

Gas imports and production in policy three vs baseline in 2030 Cumulative deviation in policy three from baseline, 2015–30

Baseline Policy three0

100

200

300

400

Production Imports

bcm

2015201620172018201920202021202220232024202520262027202820292030

-15 -10 -5 0 5 10 15 20

Imports Production

Per cent

Source: CHINAGEM

Import dependency is the percentage of gas imports to total gas consumption, and it is relevant to the policy goal of energy security. China’s dependency on natural gas will continue to increase as it shifts away from coal as an energy source. This drives the increase in total gas imports and import dependency exhibited in the first two policy scenarios. However, in the unconventional gas policy scenario, increasing unconventional gas production has a negative effect on the total gas imports and import dependency.

Although this policy scenario implies lower conventional gas production relative to the baseline, total indigenous gas production rises as unconventional gas production grows at a faster rate post-2020. The declining rate of growth in gas consumption results in a decline in the growth rate of total gas imports.

Figure 6.3 compares the impact of each of the three policy scenarios on gas import dependency between 2020 and 2030. The baseline import dependency is 32.2 per cent in 2020 and 32.6 per cent in 2030. The import dependency in the unconventional gas policy (scenario three) decreases from 30 per cent in 2020 to 27 per cent in 2030. This results in import dependency declining by two percentage points in 2020 and six percentage



points in 2030, respectively, relative to the baseline. The import dependency in the coal capping policy (scenario two) and the increase in the service sector policy (scenario one) are four percentage points and two percentage points higher than in the baseline, respectively, by 2030. Therefore, compared with the other two policies, increasing unconventional gas production strongly reduces gas import dependency. This policy results in a 10 percentage point reduction in import dependency compared with a policy of capping coal (37 per cent) and seven percentage points lower than a policy to increase the service sector (34 per cent) by 2030.

Figure 6.8: Changes in import dependency of natural gas in various policy scenarios, 2020 vs 2030

Import dependency in various policy scenarios Changes of import dependency relative to baseline

Policy one Policy two Policy three0.0

10.0

20.0

30.0

40.0

33.0 33.6

30.3

34.336.5

26.9

2020 2030

Per c

ent

Policy one Policy two Policy three-6.0

-4.0

-2.0

0.0

2.0

4.0

6.0

0.71.4

-1.9

1.8

3.9

-5.72020 2030

Per c

ent

Source: CHINAGEM

7. A composite policy and sensitivity to oil prices

In this section, the three policy scenarios are analysed jointly to determine the overall impact of multiple policy interventions in China’s economy. In addition, this analysis is supplemented by an examination of the impact of a range of future oil price scenarios on the level of gas imports and on the incentives for greater indigenous gas production in China.



Relative to the three individual policy scenarios, the composite policy scenario models an attempt to balance economic growth, environmental benefits and energy security. Modelling of the composite policy results in a higher share of natural gas in the primary energy and electricity generation mixes, higher LNG imports, improved air quality from the lower coal consumption, and lower carbon emissions from the shift away from coal to gas.

An analysis of oil prices, in conjunction with the composite policy, shows that higher oil prices linked to LNG prices will lead to lower LNG imports, but will provide higher returns to capital investment and incentivise indigenous gas production.

7.1 A composite policy and its effects

China is likely to reform a variety of policies during the forecast period, with impacts on economic growth, and coal and gas production and consumption. Therefore, not only do we need to gain insight on the key factors associated with each policy in isolation, but it is also important to consider the interaction of the three policies in combination.

China’s natural gas imports are expected to expand to meet the continually increasing gas demand from the power, industrial, residential and transport sectors. Unconventional gas production increases, but the volume of production over time is uncertain. Despite the effects of oil price volatility, gas supply availability and capital intensive infrastructure, it is pricing reform, and government policy and funding to promote natural gas over other fuels that are the key factors affecting the speed at which the switch to gas and LNG occurs.

In this section, the discussion focuses on the main comparative effects from the composite policy relative to each of the three distinct economic and energy policies presented previously.

Scenario one — an increase in the service sector

Compared with scenario one, the composite policy results in higher economic growth, higher income and lower total primary energy consumption. China has less of a need to produce fossil fuels from resources in the economy and energy intensity is lower. Higher gas demand is met by both higher indigenous gas production and gas imports.

Scenario two — capping coal consumption by 2020

Compared with the scenario two, the composite policy shows continued reduction in carbon emissions combined with higher economic growth and higher total energy consumption. Emissions intensity is lower and gas imports are higher, but there is only a small reduction in indigenous gas production.

Scenario three — higher unconventional gas production

Compared with the scenario three, the composite policy results in lower total energy consumption and higher gas demand. The accelerated



increase in unconventional gas production crowds out small production of conventional gas. There is less total indigenous gas production and more gas imports, leading to higher gas import dependency (12 percentage points higher).

The impact of all four policy scenarios on total natural gas consumption is shown in Figure 7.1. It is clear that the composite policy leads to significantly higher gas consumption (29 per cent) than the baseline scenario. This is only marginally greater than the results for scenario two alone (coal consumption capped), but significantly higher than the results for the other two scenarios.

Figure 7.9: Comparative total gas consumption in various policy scenarios in 2030

Composite policy Policy two Policy one Policy three Baseline0

100

200

300

400

500 472456

392 391365

bcm

Source: CHINAGEM

Figure 7.2 shows the shares of fuel consumption in the primary energy mix and in the electricity generation mix over time under the composite scenario. From 2015 to 2030, the share of coal consumption in the energy mix reduces from 65 per cent in 2015 to 51 per cent in 2030. This decline is balanced by an increase in the gas share from 6 per cent in 2015 to 11 per cent in 2030, and an even greater increase in non-fossil fuels from 11 per cent in 2015 to 20 per cent in 2030.

In the electricity generation mix, gas remains a minor input, growing to only seven per cent by 2030. The main growth area is in non-fossil fuels, which increases to 44 per cent of the electricity generation mix in 2030, up by 17 percentage points in 2015. Coal consumption for electricity generation declines by 20 percentage points to 50 per cent in 2030.



Figure 7.2: Gas share in the primary energy mix and electricity generation mix in the composite policy, 2015–30

Primary energy mix Electricity generation mix

Coal Oil Gas Non-fossil fuel0

10

20

30

40

50

60

70

2015 2020 2030

Per c

ent

Coal Oil Gas Non-fossil fuel0

10

20

30

40

50

60

70

2015 2020 2030

Per c

ent

Source: CHINAGEM

Figure 7.3 shows the effects of the composite policy on LNG and pipeline imports of gas, relative to the baseline. The composite policy stimulates higher demand for natural gas, and a higher share of natural gas in the primary energy and electricity generation mix, resulting in higher imports.

In the baseline scenario, an increase in LNG imports is limited because of the lower total gas demand relative to the new policy scenarios. In the composite policy in 2030, LNG imports are 85 bcm and pipeline gas imports are 95 bcm, which are more than 50 per cent higher than the respective LNG and pipeline gas import in the baseline scenario. Dependency on imported natural gas is 38 per cent, higher than in other policy scenarios by 2030 (see Figure 6.3).

CHINAGEM has not been designed to predict the sources of China’s LNG imports during the forecast period. However, in 2014, the main sources for China’s total LNG imports (27.1 bcm) were Qatar (34 per cent), Australia (19 per cent), Malaysia (15 per cent) and Indonesia (13 per cent). It is expected that the share of LNG imports from Australia will increase during the next five years, mainly at the expense of Qatari imports.

Under the composite policy, the share of coal in the energy mix declines, whereas the share of gas and non-fossil fuels increases. This has implications for carbon emissions and greenhouse gas policy. Natural gas



combustion is relatively cleaner and emits less CO2 than other fossil fuels, which makes it favourable in terms of reducing greenhouse gas emissions.

Figure 7.3: Changes in natural gas imports in composite policy vs baseline, 2010–30

LNG imports Pipeline gas imports

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

0

20

40

60

80

100

56

85

Baseline Composite policy

LNG

(bcm)

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

0

20

40

60

80

100

63

95


Pipe

line

gas (

bcm)

Source: CHINAGEM

Figure 7.4 shows that, in the composite policy, total CO2 emissions grow very slowly from 2020 to 2030, with 0.4 per cent CAGR, compared with around two per cent in the baseline scenario during the same period. As a result, by 2030, total CO2 emissions under the composite policy are 11.1 billion tonnes, which is more than 2.5 billion tonnes below the emissions in the baseline (13.6 billion tonnes). With an annual growth rate of only 0.3 per cent in 2030, total carbon emissions are very close to peaking by 2030, in line with commitments from the Chinese government.

The simulated results also show that, among the total CO2 emissions, coal emissions reduce to 8.1 billion tonnes in 2030 in the composite policy from 10.7 billion tonnes in the baseline, a reduction of 25 per cent. In the opposite direction, gas CO2 emissions increase to 0.8 billion tonnes in the composite policy from 0.6 billion tonnes in the baseline, an increase of 29 per cent in 2030.



Figure 7.4: Changes in carbon emissions and intensity: composite policy vs baseline, 2015–30

Increasing trend of carbon emissions Carbon emissions, by fuel type, 2030

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 20308

9

10

11

12

13

1413.6

11.1


Giga

tonne

s

Coal Petrol refinery Gas0

2

4

6

8

10

1210.7

0.6

8.1

0.8


Giga

tonne

s

Source: CHINAGEM

7.2 Crude oil prices and natural gas imports

LNG will play a key role in China’s energy mix, as it helps to maintain a diversified fuel supply mix and feed the gas-hungry coastal regions, which remain distant from pipeline or most domestic gas supplies.43 An upwards oil price shock would affect gas imports through the oil price indexing of gas contracts for LNG and pipeline gas. This section examines the effect on China’s LNG imports from a rise in crude oil prices.

China’s LNG is sourced mainly under long-term contracts, with the price linked to the price of oil; spot trading has also increased over time. The LNG contracts are typically linked to the price of Japanese customs-cleared crude oil (JCC), which reflects the average price of crude oil imported into Japan and, which in turn, is highly correlated with the price of Brent crude oil (see Figure 7.5). The contract price is also determined by an S-curve44, including the coefficient (the pricing slope) that determines the sensitivity of LNG prices to changes in the oil price benchmark, factoring in the conversion between US$/barrel (bbl) and US$/mmBtu.

43 Corbeau et al. (2014) The Asian Quest for LNG in a Globalising Market

44 Typically S-curve pricing formula is expressed as P = A*JCC+B. S-curves are intended to

reduce price risks by mitigating the impact of either rapidly rising or falling oil prices. The sellers

require a price floor to protect their liquefaction projects from oil price collapse. As a trade-off,

buyers want upside protection. Floor and ceiling prices can be set to offset such risks.



Figure 7.5: Brent oil price versus Japanese customs-cleared crude oil, 2005 (quarter 1)to 2015 (quarter 3)

Q1 200

5

Q3 200

5

Q1 200

6

Q3 200

6

Q1 200

7

Q3 200

7

Q1 200

8

Q3 200

8

Q1 200

9

Q3 200

9

Q1 201

0

Q3 201

0

Q1 201

1

Q3 201

1

Q1 201

2

Q3 201

2

Q1 201

3

Q3 201

3

Q1 201

4

Q3 201

4

Q1 201

5

Q3 201

5

40

60

80

100

120

140

Brent JCC

US$

/bbl

Source: Nexant (2015) WGM

The LNG pricing slope for imports into China reportedly ranges between 13 and 14 per cent. The LNG contract price also includes a constant, largely reflecting transportation cost, of around US$1/mmBtu for China’s imported LNG. Pipeline gas imports into China are determined by a Brent oil–linked price with a lower slope of around 9 per cent and a higher constant of US$2.5/mmBtu, reflecting a higher transportation tariff component.45 This implies that LNG imports are more sensitive to oil prices than pipeline imports. As a result, an increase in the oil price will raise the contracted prices of LNG and pipeline for China, but to different extents.

CHINAGEM has been used to analyse the impact of a range of oil price scenarios on natural gas supply. To simulate the effects of the oil price on gas imports (LNG and pipeline), oil prices are modelled in US$10 increments from US$80/bbl to US$120/bbl by 2030, in comparison to the baseline of US$55 — equivalent to the average price in 2015. The analysis assumes that all of China’s current and future LNG imports are based on oil-linked contracts. Although this may be a simplification of actual contract terms, the general results are still valid. Table 7.1 shows the indicative changes in China’s import prices for LNG and pipeline gas for each oil price scenario.

45 The price difference between pipeline gas and LNG imports reflects the higher tariff involved in

transporting pipeline gas from the border regions (e.g. north-western China) to the major gas

consuming cities in the north-eastern regions. It also reflects the lower regasification costs for

LNG delivered at China's LNG terminals (which are mainly located relatively proximate to the

major gas-consuming regions (e.g. Guangdong, Pearl Delta and Yangtze Delta).



Table 7.1: Scenarios of oil price in the world market: indexed LNG and pipeline gas prices

Brent oil (US$/bbl) LNG (US$/mmBtu) Pipeline gas (US$/mmBtu)

55 8.69 7.42

80 12.09 10.79

90 13.45 12.13

100 14.81 13.48

110 16.17 14.83

120 17.52 16.18

Notes: US$55/bbl was the average Brent oil price in 2015Source: Nexant (2015) WGM and authors’ calculations

A low oil price (e.g. US$55/bbl) extending to 2030 reflects an assumption that there would be continued increasing supply from new discoveries of oil. In addition, increases in the productivity of oil production with advanced technologies that offset the high cost of greenfields oil production could also lead to lower oil prices. On other hand, growth in demand for oil is declining because of slowing economic growth, changes in economic structure and fuel substitutions (e.g. the effects of climate change policies).

Higher imported gas prices linked to higher oil prices will be less favourable to gas imports, but will provide incentives to invest in indigenous gas production. In general, low oil prices will have the opposite effect on investment and production. This means increased indigenous gas production is offset by declining imports when there is an upwards trend in the oil price path, from US$55/bbl to US$80/bbl, to US$100/bbl and to US$120/bbl. For example, Figure 7.6 shows that with the higher oil price path from US$55/bbl to US$80/bbl, gas production increases from 292 bcm to 307 bcm (by five per cent) and gas imports decline from 180 bcm to 165 bcm (by eight per cent) in 2030.



Figure 7.6: Natural gas supplies at various oil prices under composite policy scenario, 2030

Gas production and imports Deviation of gas production and imports

US$55 US$80 US$100 US$1200

100

200

300

400

500


bcm

US$80 US$100 US$120-30

-20

-10

0

10

20


Per c

ent

Notes: The baseline is the composite policy with an oil price of US$55/bbl

Source: CHINAGEM

China’s imported LNG is expected to be 85 bcm per year by 2030 compared with 95 bcm for imported pipeline gas. Figure 7.7 indicates that a higher oil price under the composite policy scenario leads to significant falls in both LNG and pipeline imports. Higher LNG prices resulting from higher oil prices reduce gas imports, but provide an incentive to invest in indigenous gas production.



Figure 7.7: Natural gas imports at various oil prices under composite policy scenario, 2030

Gas imports at various prices Deviation of gas imports at higher oil prices

US$55 US$80 US$100 US$1200

20

40

60

80

100 95

87

77

69

8578

7063

Pipeline gas LNG

bcm

US$80 US$100 US$120-30

-25

-20

-15

-10

-5

0

-8.8

-19.3

-27.2

-7.7

-17.9

-25.7

Pipeline gas LNG

Per c

ent

Notes: The baseline is the composite policy with an oil price of US$55/bbl

Source: CHINAGEM



8. Key findings and concluding remarks This report provides insights into effects of a range of policy scenarios on dynamics of China’s demand for LNG to year 2030. China’s economic and energy policies are key factors in its LNG demand. The composite scenario provides an assessment of a potential future policy framework which attempts to address the needs for structural transformations to support economic growth, the need to reduce CO2 emissions through reducing coal consumption, and the need for energy security through encouraging the development of China’s significant onshore unconventional gas resources.

Compared with the baseline scenario, the key outcomes under the composite policy scenario for the year 2030 are that:

The share of coal consumption in the primary energy mix reduces to 51 per cent (from 60 per cent), the natural gas share increases to 11 per cent (from 7 per cent) and the share of non-fossil fuels increases to 20 per cent from (15 per cent) of the primary energy mix.

Natural gas imports grow faster to meet the increased natural gas demand, which allows the import dependency to rise to 38 per cent (from 33 per cent).

Total carbon emissions reduce to 11.2 billion tonnes (from 13.6 billion tonnes).

This study explores key aspects of China’s future demand for energy — particularly from LNG — under three distinct policy scenarios:

scenario one (increasing share of service sector in the economy)

scenario two (capping coal consumption by 2020

scenario three (accelerating unconventional gas production).

The analysis is based on CHINAGEM to simulate the effects of various policy scenarios over time. The forecast mode is used to establish a business-as-usual baseline for China’s economy (covering energy and non-energy sectors) through to 2030. The modelling does not reflect or account for abrupt political changes within China, new technological developments in light and heavy manufacturing, or global events that significantly disrupt China’s economic performance.

A final policy simulation (i.e. composite policy scenario) integrates the three separate policy scenarios. Compared with the three individual scenarios, the composite policy is an attempt to balance between the objectives of strong economic growth, lower carbon emissions, environmental benefits and energy security. The composite policy scenario leads to:

a higher share of natural gas in China’s primary energy and electricity generation mix

higher LNG imports

improved air quality because of the lower coal consumption

lower carbon emissions as a result of the shift from coal to gas.



These beneficial outcomes are matched with an increased trend in GDP growth relative to the baseline scenario.

Compared with scenario one alone, the composite policy predicts higher economic growth, higher income and lower total primary energy consumption. This means there is a reduced need to produce fossil fuels from resources in the economy, and a lower energy intensity. There is higher gas demand, which is supplied by higher indigenous gas production and gas imports.

Compared with scenario two alone, the composite policy shows continued reduction in carbon emissions combined with higher economic growth and higher total energy consumption. There are lower emissions intensity, higher gas imports and a small reduction in indigenous gas production.

The composite policy results in lower total energy consumption and higher gas demand compared with scenario three alone. The accelerated increase in unconventional gas production crowds out a small amount of production of conventional gas. The import dependency of natural gas declines relative to the baseline.

Table 8.1 summarises the impact on energy demand of each of the four policy scenarios compared with baseline in 2030.

Table 8.1: Indicators of changes in energy demand at various policy scenarios in 2030

Demand Baseline Scenario 1 Scenario 2 Scenario 3 Composite scenario

Primary energy mix (per cent)

Coal 60.3 58.9 52.7 60.3 51.1

Gas 7.5 8.1 10.5 7.9 10.7

Non-fossil fuels 14.6 16.0 16.9 14.3 19.7

Electricity generation mix (per cent)

Coal 63.9 63.7 54.4 64.8 49.7

Gas 5.0 4.8 6.7 4.7 6.7

Non-fossil fuels 31.1 31.4 38.9 30.5 43.5

Carbon emissions (billion tonnes)

Coal 10.7 10.5 8.0 10.7 8.1

Gas 0.6 0.6 0.7 0.7 0.8

Petroleum refining 2.3 2.3 2.3 2.3 2.3

Gas demand

Gas consumption (bcm) 365 392 456 391 472

LNG imports (bcm) 56 64 79 50 85

Import dependency (per cent) 33 34 37 27 38

Source: CHINAGEM and authors’ calculations



The composite policy scenario is sensitive to changes in oil price. Given the link with LNG prices, higher oil prices lead to lower LNG imports, but provide incentives for capital investment in indigenous gas production.

China’s economic and energy policies aim to deliver three key objectives; economic growth, environmental improvements and energy security. Although we cannot predict the precise policies that China will use in response to these three objectives, this report has considered the outcomes of a potential policy approach which attempts to achieve a balance between all three. This composite policy scenario results in a reduction in coal dependency for China’s total energy consumption, and a reduction in CO2

emissions. The scenario results in increasing share of natural gas in China’s primary energy consumption during the next 15 years, and a large increase in the volumes of LNG imports.



Appendix A

Key characteristics of CHINAGEM

A.1 Key elements

The dynamic CGE approach emphasises the economics of change, in contrast with a time series econometric approach, which focuses on historical evidence to predict the future. In a country such as China, credible time series data for anything beyond the broad macro-aggregates are generally not available. However, even if they were available, it is difficult to see how econometrics could deal with some of the huge structural changes affecting the Chinese economy now and in the future.

In recursive-dynamic mode, CHINAGEM produces sequences of annual solutions connected by dynamic relationships such as physical capital accumulation, net foreign liability accumulation and lagged labour market adjustment. Policy analysis with CHINAGEM conducted in a dynamic setting compares two alternative sequences of solutions, one generated without the policy change and the other with the policy change in place.

The model includes a number of satellite modules providing more detail on the model’s government finance accounts, household income accounts, population and demography, and energy and greenhouse gas emissions. Each of the satellite modules is linked into other parts of the model, so that projections from the model core can feed through into relevant parts of a module and changes in a module can feed back into the model core.

There are two broad classes of equation in CHINAGEM. The first comprises accounting relationships, which ensure that receipts and expenditures of every agent in the economy balance. The second class comprises behavioural equations, based on microeconomic theory, that specify the behaviour of profit- or utility-maximising agents.

In CHINAGEM, the production technology of each industry is modelled using constant elasticity of substitution (CES) and Leontief production functions. The CES functions allow for substitution in production between domestic and imported varieties of each input and between primary factors (i.e. labour, capital, and land for sectors such as agriculture). Composite intermediate inputs are combined with composite factor inputs in fixed proportions. Several different commodity outputs may be produced by a single industry. There is a constant elasticity of transformation (CET) between the outputs of each industry. Production functions are subject to constant returns to scale.

Each industry comprises many identical firms that take prices of inputs and outputs as given. That is, they operate in perfectly competitive markets. Subject to their production functions, as described above, firms choose their mix of inputs and outputs to minimise costs and maximise revenues. The level of output of each industry is determined such that total costs equal total revenues: firms earn zero pure profits. Household demand is modelled by the linear expenditure system.



Supply–demand balance is linked to the past through the accumulation of assets from past investments. Capital stocks in this model can either grow at a constant or flexible rate. The amount of investment determines the rate of capital growth. Investment, on the other hand, is determined by its rate of return. Capital is expected to grow annually and the rate of growth is equal to net investment, defined as the rate of investment at the beginning of the year less the rate of depreciation for that year.

Real wages can also adjust in response to changes in employment levels. The adjustment mechanism is such that if the end of the year employment level is above the baseline trend level, real wages adjust in inverse proportion. Given that the rate of employment falls when wages increase and rises when the real wage decreases, the adjustment mechanism ensures that the level of employment gradually returns to baseline trend. The adjustment parameter is chosen so that the effects of a temporary shock to labour demand neutralise over a time scale of about 10 years.

A.2 Behavioural parameters

Behavioural equations incorporate a range of coefficients that must be estimated, the most important of which are the substitution elasticities. The key elasticities are as follows:

Elasticity of substitution between domestic and imported goods. CHINAGEM adopts the assumption that imports are imperfect substitutes for domestic supplies. For each domestic demand agent (industries, capital creators, households and government), imports of a commodity were specified as CES substitutes for the corresponding domestic commodity. Normally, these parameters are empirically assessed by time series and econometric analysis. However, because of data limitations in the field, the elasticities in this paper have been given generally accepted values taken from a range of studies.

CES between factor inputs. The composite input of primary factors is a CES combination of land, capital and composite labour. The value of elasticity substitution between factor inputs is 0.5.

CET for industries' output mix. CHINAGEM allows for each industry to produce more than one commodity. The value of the transformation elasticity is 0.5.

CET for the commodities used between export and local market. CHINAGEM allows for the differentiation of domestically produced goods that are destined for export, and those that are intended for domestic use. Conversion of an undifferentiated commodity into goods for both destinations is governed by a CET frontier.

Export elasticities for individual commodities. Export demand elasticities give the percentage change in world demand for Chinese exports with respect to a one per cent increase in world price of the Chinese product. The value of the elasticities for China’s main exported products is –4.

Household expenditure elasticities. An expenditure elasticity is required for each commodity consumed by the representative household. Values for these elasticities in the CHINAGEM model have been drawn from



empirical research. The value of the expenditure elasticity for each commodity in CHINAGEM model varies from 0 to 1.517.

A.3 The model modifications

Our starting point is the previous version of CHINAGEM, with a database reflecting the year 2007. The previous version lacks the capacity to model energy issues in detail. Primary energy is supplied by only two industries: the coal industry, and a crude oil and gas industry. There are two secondary energy industries, which produce refined oil products and electricity.

We extended this version of CHINAGEM from the original 135 industries to 143 industries. The 143 industries, together with households, comprise the set of fuel users (144 in total, counting households as one fuel user). Key modifications of the model structure include:

splitting the crude oil and gas industry into the crude oil industry and the gas industry

disaggregating domestic gas production into conventional and unconventional gas production

separating two sources of imported natural gas — pipeline gas and LNG

disaggregating electricity generation across six unique fuel technologies — coal, gas, oil, nuclear, hydro and other renewables — and introducing inter-fuel substitution in electricity generation

defining cost-responsive changes in the relative supplies of conventional and unconventional gas

defining price-responsive substitution possibilities in demand for the two alternative sources of imported gas supply

introducing a climate change module in the model (energy accounting and CO2 emissions accounting).

Disaggregation of oil and gas

The initial database recognised crude oil and gas production as a single industry producing a single product. We separate this industry into three parts: crude oil, conventional gas and unconventional gas. We separate the single commodity into crude oil and gas. The crude oil industry produces only crude oil and is the only domestic industry that does so. The two gas industries each produce a single product (gas) and are the only domestic industries that do so. Oil and gas are also imported. Total supply of each commodity equals their domestic production plus imports.

For splitting sales of the crude oil and gas composite, we make the simple but intuitive assumption that those sales of the composite product to the petroleum refinery sector is of crude oil only, while all other sales are of gas only. The same rule is applied to imported as well as domestically produced oil and gas. This approach is based on the assumption that ‘oil’ consists solely of crude oil, which cannot be used without refining.



For splitting production costs of the crude oil and gas sector into three (crude oil, conventional gas and unconventional gas), we assumed identical technologies. This means that cost shares are the same across each new industry. Clearly this is not the case, but there are currently insufficient data to allow a more accurate disaggregation.

Cost-responsive changes in relative supplies of conventional and unconventional gas

By introducing a two-industry structure for gas supply, we provide the model with the means for cost-responsive changes in relative supplies. This is illustrated in Figure A1.

Figure A1: Cost-responsive changes in relative supplies of conventional and unconventional gas

Initially, the intersection of market supply and demand determines an overall quantity of gas supplied (Q) to match demand at the market price (P). At this price, the amount of unconventional gas supplied is Quncon and the amount of conventional gas supplied is Qconv. Now, suppose that there is a cost-reducing change in technology for producing unconventional gas. As shown in the diagram, this shifts the supply schedule for unconventional gas to the right. With the supply schedule for conventional gas unchanged, there will be a shift out in market supply. With the market demand schedule unchanged, this leads to a lower market price and increased overall supply. The share of unconventional gas in total supply rises and the share of conventional gas falls.



Two sources of supply of gas imports

China’s demand for gas is met from domestic production and imports. As explained above, domestic gas is supplied from conventional and unconventional producers. There are two primary sources of import supply: pipeline gas and shipments of LNG.

To model two sources of gas supply, we require data on expenditure by source of supply for each gas user recognised in the model. Currently, these data for individual users are not available. Accordingly, we use national shares to allocate purchases of imported gas of each user to pipeline and LNG sources.

Demand for gas from alternative imported sources

To model the alternative sources of imported gas, we assume that pipeline gas is an imperfect substitute for LNG. Thus, the landed cost, insurance and freight (CIF) price of pipeline gas can differ from the landed CIF price of LNG, and that any change in the relative price will lead to a change in ratio of use. This is illustrated in Figure A2, which shows the input structure for gas for a typical gas user in the model. At the top level, overall demand for gas is met from a combination of domestically produced gas and imported gas. The aggregator function has a CES form. As explained above, domestic gas can be sourced from conventional and unconventional producers.

The idea that domestically produced gas is an imperfect substitute for imported gas is part of the existing model structure. To this we add the new specification that allows for imported gas to be sourced from LNG and pipeline gas. Again, the aggregator has a CES form.

Figure A2: Demand for gas from alternative imported sources

Electricity disaggregated by generation type and supply

The previous CHINAGEM model recognised one electricity sector that generates electricity, and provides services associated with transmission, distribution and retailing. Intermediate inputs to electricity, including fuels, are



combined in fixed proportions. Accordingly, there is no possibility of inter-fuel substitution in electricity generation.

We correct this by introducing inter-fuel substitution in electricity generation using the ‘technology bundle’ approach. In the revised model, we split the composite electricity sector into generation and supply. Electricity-generating industries are distinguished based on the type of fuel used. The end-use supplier (electricity supply) purchases generation and provides electricity to electricity users. In purchasing electricity, it can substitute between the different generation technologies in response to changes in generation costs. Such substitution is price induced, with the elasticity of substitution between the technologies typically set at five.

The model distinguishes six types of electricity generation: coal, oil, gas, nuclear, hydro, and other renewables. It treats each type of electricity generation as one industry with a unique output, such that electricity produced by different fuels may, and indeed are likely, to have different prices in different scenarios. All electricity generation industries sell only to the electricity supply industry. The electricity supply industry sources from these electricity generation industries according to a CES substitution. In configuration, we set the value of this substitution variable to be five (if the value of the substitution variable is zero, then effectively setting the production structure would be Leontief. This reflects the fact that the dispatching order in China’s electricity market reacts more to administrative orders than price signals).

Energy accounting

Four industries produce three types of primary fossil fuels. The coal and mining products industry produces coal, the crude oil industry produces crude oil, and the conventional gas and unconventional gas industries produce gas. Table A1 displays the primary fossil fuels use — including the fuels used in the conversion activities, such as the use of coal in electricity generation and the use of crude oil in petroleum refineries in the base year 2007. Initially, we allocate total primary energy by fuel type (in the base year 2007) to different fuel users (144 fuels users in the model including 143 industries and one household) according to the share of the basic value used by each fuel user, respectively. Then total primary energy by fuel type used by each fuel user is updated based on the growth of real output of each fuel user (for the 143 industries) and the growth of household real consumption (for household).

Table A1: Total primary fossil fuel consumption, 2007

Fuel type Consumption (mtoe)

Coal 1,396

Crude oil 369Gas 65Total 1,830



Source: China Statistics Press (2010) and authors’ calculations

Seven industries produce six types of final energy:

coal and mining products (produces coal)

conventional gas (produces gas)

unconventional gas (produces gas)

petroleum refinery (produces petroleum refinery products)

coking (produces coke)

electricity supply (supplies electricity)

gas supply (supplies gas).

The final energy consumption includes the use of primary (e.g. coal) and secondary fuels (e.g. automotive petrol), but does not include the energy used in conversion (e.g. the electricity generation, refinery production). The final energy use in 2007 is displayed in Table A2. Like primary fossil fuels, the initial quantities (in year 2007) of final energy consumed by fuel users are allocated according to the share of the basic value used by each fuel user, respectively. Then, total final energy use by each fuel user is updated based on the growth of real output of each fuel user (for the 143 industries) and the growth of household real consumption (for household). However there is one exception. The industries of coking, electricity supply and gas supply use coal to produce final energy outputs and, therefore, these industries are not counted as the final fuel users of coal. Thus, the total quantity of coal as final fuel is allocated among 141 (144–3) fuel users.

Table A2: Total final energy consumption, 2007

Category Consumption (mtoe)

Coal 316Gas 54Petroleum refining 336Coke 139Electricity supply 820Gas supply 5.6Total 1,670

Source: China Statistics Press (2010) and authors’ calculations

Carbon dioxide emissions accounting

CHINAGEM accounts for all CO2 emissions from fossil fuel combustion. The initial levels of emissions are linked to three types of CO2-emitting fossil fuels — coal as primary energy, gas as final energy and petroleum refinery products as final energy (Table A3). Coal as primary energy is distributed among all fuel users except the coking industry, which does not use coal



combustion in its production process. Non-coal final energy sources are distributed among all fuel users according to their value shares. Hence, by dividing the total emissions of each CO2-emitting fuel by the corresponding total quantities of CO2-emitting fuels used by all the fuel users, we obtain the overall emission intensity factor for each CO2-emitting fuel (a three by one matrix). Each of these factors is then multiplied by the quantities of CO 2-emitting fuels used by every fuel user, which gives the approximated levels of CO2 emissions from each CO2-emitting fuels by every fuel user.

Table A3: Total emissions from fossil fuel combustion, 2007

Category Emissions (mt)

Coal 5,330

Gas 115

Petroleum refining 1,347

Total 6,792

Source: Authors’ calculations based on the World Bank (2015) WDI database

A.4 The model scenarios and simulations

CHINAGEM can be used for four modes of analysis: historical, decomposition, forecast and policy.

1. Historical simulations are used to analyse past trends in production technologies, preferences and other naturally exogenous, but unobservable, variables.

2. Decomposition simulations explain historical episodes and place policy effects in historical context.

3. Forecast simulations provide baselines using extrapolated trends from historical simulations together with specialist forecasts.

4. Policy simulations are used to analyse the economic impact of policies as deviations from a baseline economy scenario.46

This study uses CHINAGEM in three modes: historical, forecast and policy. Each mode is referred to as a simulation, with each simulation run using a distinct closure. This is because the number of variables in CHINAGEM is larger than the number of equations. When the equation system possesses a set of unknown variables that are equal in number to the number of equations in the model, CHINAGEM can be used to solve for changes in these endogenous variables. Variables that are not endogenous are called exogenous variables, the values of which must be specified by the user. The classification of endogenous/exogenous variables — which are to be solved by CHINAGEM — is flexible and is what we refer to as the closure. One variable can be endogenous in one simulation and exogenous in another.

46 Dixon et al. (2013) The MONASH style of computable general equilibrium modeling: a framework for practical policy analysis



Historical simulation from 2007 to 2013

In a historical simulation, the observable variables such as GDP, production, consumption and international trade are set as exogenous, whereas the corresponding technical change and preference variables (such as multi-factor productivity) are endogenous. We inform the model of changes in the observable variables from the base year, for which detailed input–output (I/O) data are available (i.e. 2007). The model then calculates implied changes in technology and preferences that transitions the database from 2007 to 2013, which is the most recent year that the statistics for most of observable variables are available.

The CHINAGEM model is a very large model with hundreds of equations and thousands of variables. The detail is necessary for answering practical policy questions. However, the equations listed in Box A1 provide a good understanding of the fundamentals of the model. Together they form the Back-of-the-Envelope (BOTE) model 47, which is very useful for explaining simulation results. In the following paragraphs, we use the BOTE model presented in Box A1 to explain the methodology of the historical simulation conducted for this project.

The modelling starts with a database that describes the structure of the Chinese economy in 2007. As discussed, the 2007 China I/O table is the main source of input–output data for the CHINAGEM model. The I/O table shows the flow of goods and services from one industry to all other industries (described in rows) and the inputs required by an industry to produce its outputs (described in columns) in detail. The major components and their common notations in an I/O table include:

Intermediate transaction (xij) describes the intermediate deliveries of product from one sector to all sectors (including itself). For example, x ij

represents the delivery of product x from sector i to sector j.

Final demand (yi) shows the sales of product to end users. Sales can be made to households, the government and foreigners (via exports), and is also considered in terms of an increase in inventories.

Value added (wi) represents the other inputs to production, including employees’ compensation and tax.

The premise of the historical simulation is to describe how the Chinese economy evolved from 2007 to 2013. We provide the model with the observed growth in the following macroeconomic variables for 2007–13:

The components of GDP from the expenditure side: consumption (C), investment (I), government expenditure (G), exports (X) and imports (M), and GDP price index (Pg).

The three macro sectors: agriculture, industry and service.

Population (POP) and employment (L).

47 Dixon et al. (1982) ORANI: a multisector model of the Australian economy and Dixon and

Rimmer (2002) Dynamic general equilibrium modelling for forecasting and policy: a practical guide and documentation of MONASH



The model has dynamic equations that link the economy from one year to the next (see Box A1). When we provide the model with the observed growth of C, I, G, E and M from equation (1), the model will calculate the implied rate of GDP growth (Y).

Box A2: The BOTE model

The two most important relationships in the CHINAGEM model are the GDP identity and the aggregate production function:

Y=C=I+G+X−M (1)

Y= 1A∗F (K , L) (2)

where Y is GDP;C is consumption;I is investment;G is government expenditure;X is exports;M is imports;K is aggregate capital stock;L is aggregate employment; and decreases in A allow for technological progress.Equilibrium in the capital market requires the real cost of capital to be equal to the marginal physical product of capital. Hence:

QPg

= 1A

∗F k (K /L) (3)

whereQ is the rental per unit of capital;

Pg is the price of a unit of GDP; and

Fk is the partial derivative of F with respect to K. We write F kas a function of K/L under the

assumption that F is homogenous of degree one. Labour-market equilibrium requires:

WPg

= 1A

∗FL(K /L) (4)

whereW is the wage rate; and

FL is the partial derivative of F with respect to L.

The final equation in our BOTE model explains capital in the current plus one year as the sum of net capital in the current year plus investment. Hence:

K1=K+ I−D (5)

where

K and K 1are the capital stock in the current and following year respectively; and

I and D are investment and depreciation in the current year.

Source: Mai, Dixon & Rimmer (2010)

Given the growth in investment (I), the growth in aggregate capital stock (K) is then determined by equation (5), which models the accumulation of physical capital. The capital stock in the next year equals the capital stock in the current year, plus investment in the current year minus depreciation.



After we inform the model of growth of Y, L and I (and thus K), changes in technology (A) will be solved by the production function — see equation (2).

Since growth in the GDP price index (Pg) is also tied down, equation (3) will solve for the capital rental (Q), and equation (4) will solve for the wage level (W).

At the industry level, we provide the model with the historical growth paths for output (Yi) of the three macro sectors. Consequently, the industry versions of the aggregate production function (2) and factor market equilibrium conditions (2) and (3) can jointly solve for industry specific capital stock (Ki), rental rates (Qi) and technology (Ai).

For this project, the main data source used for the growth rates of real GDP and its expenditure-side components is the World Bank World Development Indicators (WDIs). Table A4 shows the annual growth rates of real GDP, consumption, investment, government expenditure, exports and imports from 2008 to 2013 from the WDIs.

Before proceeding, we check whether the real GDP growth and the growth rates of its components satisfy the GDP identity (equation (1) in Box A1, where Y, C, I, G, X, and M are levels of GDP, consumption, investment, government expenditure, exports and imports, respectively. This is achieved by multiplying the levels of C, I, G, X and M in the CHINAGEM database with the respective WDI growth rates.

From this process, we determined that the observed growth rates of the real GDP components implied a growth rate for real GDP that was a little different from that forecast by the WDIs. This was because the initial CHINAGEM levels data, to which the individual growth rates were applied, were a little different from the initial levels assumed for the WDI forecasts. Accordingly, the growth rates for individual components were scaled proportionally to account for this inconsistency. Note, however, the trend behaviour apparent in the WDI data is preserved. For example, consumption grew at a slower rate than real GDP, while investment grew faster than real GDP in 2008; this pattern of behaviour is preserved in our re-scaling. Additionally, although trade grew faster than real GDP across most years in the sample, we note that, in 2008, trade growth lagged real GDP growth. As before, these relative growth rates are preserved.

Table A4 shows calculated growth rates of GDP components. They are used as inputs into CHINAGEM for the historical simulation from 2008 to 2013. From Table A4, China displayed a high rate of GDP growth in excess of 9 per cent per year from 2008 to 2011. From 2012, the rate of growth declined to less than 8 per cent. We also note that real investment (or gross fixed capital formation) grew faster than real private and government consumption (except 2011), indicating a high savings rate in China. From 2011, China’s import growth exceeded export growth, driving down China’s current account surplus.

As discussed, we provide the model with the growth rates of three macro sectors — agriculture, industry and service. As before, the historical data for these three sectors are from the WDIs. To ensure data consistency, we use the same method to account for data inconsistencies as that applied to Key factors affecting changes in China’s demand for liquefied natural gasKey factors affecting


balance the growth rates in the components of GDP. Table A5 summarises the scaled annual growth rates of three macro sectors from 2008 to 2013 used in the model.

Table A6 shows the observed growth of employment (L) and the broader population, as well as the change in the GDP deflator and the import price index (CIF). These data are sourced from the China Statistical Yearbook and the WDIs.

Table A4: Historical simulation: scaled growth rates of GDP components, 2008–13

GDP component (per cent) 2008 2009 2010 2011 2012 2013

Real consumption 8.08 10.27 6.58 10.75 7.70 7.19

Real investment 9.46 24.77 9.30 8.91 8.42 8.96

Government expenditure 8.43 8.94 8.78 9.24 7.94 7.80

Export 9.28 –11.13 22.32 10.08 6.45 8.28

Import 5.11 5.28 16.43 11.69 7.50 10.28

Real GDP 9.63 9.21 10.45 9.30 7.65 7.67

Source: WDIs and authors’ calculations

Table A5: Historical simulation: scaled growth rates of value added by industry group, 2008–13

Industry group (per cent) 2008 2009 2010 2011 2012 2013

Agriculture 5.43 4.22 4.28 4.23 3.47 4.02

Industry 10.51 9.64 9.76 9.38 7.69 8.30

Service 9.98 10.02 12.27 10.25 8.40 7.88

Real GDP 9.63 9.21 10.45 9.30 7.65 7.67

Source: WDIs and authors’ calculations

Table A6: Historical simulation: GDP deflator, and growth rates of employment and population, 2008–13

Year (per cent) Employment Population GDP deflator Import price index

2008 0.32 0.51 8.10 15.80

2009 0.35 0.49 1.08 –12.50

2010 0.37 0.48 4.61 13.00

2011 0.41 0.48 7.35 4.00

2012 0.37 0.50 3.97 4.00

2013 0.36 0.49 2.00 1.00

Source: Growths of employment and population are from China Statistics Press (2014). GDP deflator is from WDI online. Price

of agricultural products is from China Statistics Press (2014)

Forecast simulation from 2014 to 2030

Preference and technical change trends derived from the historical simulations serve as inputs to our forecast simulation, which are used to



project forward the solution in 2013 (the final year for which data are available). Given historical trends, and assuming these trends continue to prevail in future years, the forecast simulation yields the likely path for the Chinese economy.

In this sense, the forecast mode is used to produce a baseline for the future evolution of the Chinese economy. The classification of endogenous/exogenous variables — which are solved by CHINAGEM — is flexible and is referred as the closure.48 In the historical closure, the set of exogenous variables is determined by variables for which actual historical data are available (e.g. investment growth, private consumption growth). In the forecast closure, the set of exogenous variables include exports by commodity (for which reliable forecasts may be available), and demographic variables (for which forecasts are provided by official organizations), and technology, preference and trade variables that can be given shocks informed by trends from the historical simulation.49

As discussed in Section 1, the pattern of China’s economic growth is changing; any such changes should therefore be reflected in the forecast simulation for CHINAGEM. In the context of these targets, President Xi reiterated the official targets of doubling the GDP and income per capita by 2020 (from 2010 levels) when he first took office at the 1st plenum session of the 18th Communist Party Central Committee in November 2012. This implies average annual GDP and income growth of 6.6–6.7 per cent from 2016 to 2020.

Based on China’s changing growth pattern, for the 2014–30 forecast period, real GDP is assumed to settle down to a slower longer-term trend of around seven per cent from 2015 to 2016, and 6.5 per cent from 2017 to 2020, six per cent from 2021 to 2025 and 5.5 per cent from 2026 to 2030 (see Table A7). The pattern of growth of GDP components is slightly different to the historical pattern: growth of consumption is higher than that of investment, whereas growth of imports exceeds growth of exports. Both imports and exports are assumed to continually grow faster than real GDP, as has been the case historically. Table A7 also shows the calibrated growth rates of GDP components. These numbers are used as inputs to CHINAGEM under the forecast closure, yielding the assumed growth of real GDP for the forecast simulation from 2014 to 2030.

48 CHINAGEM has many more variables than equations. The number of variables that can be

determined is equal to the number of equations. Model-determined variables are called

endogenous variables. The remaining variables have values determined by the user. These

variables are called exogenous variables. The user has considerable freedom in choosing the

variables that are to be endogenous and that are to be exogenous. This flexibility of choice

(also known as closure) is the key attribute that gives CHINAGEM the capacity to run the type

of simulations required for this study.

49 Dixon et al. (2013) The MONASH style of computable general equilibrium modeling: a framework for practical policy analysis



Table A7: Forecast simulation: growth rates of real GDP and GDP components, 2014–30

GDP component Average annual growth rate (per cent)

2014 2015–16 2017–20 2021–25 2026–30

Real consumption 8.84 8.37 7.79 7.22 6.64

Real investment 7.10 6.72 6.25 5.79 5.33

Government expenditure 7.70 7.29 6.79 6.28 5.79

Export 8.17 7.73 7.20 6.67 6.14

Import 10.15 9.60 8.94 8.28 7.62

Real GDP 7.40 7.00 6.50 6.00 5.50

Source: Authors’ calculations

Table A8 displays the growth rates of value added by industry groups during the 2014–30 forecast period. The growth of agriculture (as a whole) is assumed to follow its historical trend, such that growth is slower than that of the industry and service groups. As assumed, the service group grows faster than the industry group beyond 2015.

Table A8: Forecast simulation: growth rates of value added by industry group, 2014–30

Industry group Average annual growth rate (per cent)

2014 2015–16 2017–20 2021–25 2026–30

Agriculture 3.87 3.88 3.73 3.57 3.40

Industry 8.00 7.61 7.31 7.00 6.68

Service 7.59 8.01 7.70 7.38 7.04

Real GDP 7.40 7.00 6.50 6.00 5.50

Source: Authors’ calculations

Based on information from the United Nations’ (UN) medium variant population projection, Table A9 shows the growth of population for the forecast period. For employment, as assumed, the labour force participation rate and unemployment rate from 2015 to 2030 will be the same as in 2014, so we can take the growth of employment to be the same as that of the working age population, as derived from the UN population projection.

The overall population growth rate will, however, trend lower than the historically observed growth rate (Table A9), which is driven primarily by a continuing trend of declining fertility rates. In addition, the size of the population aged 65 and over is forecast to increase dramatically, from 116 million in 2010 to an estimated 250 million in 2030 (UN 2012). When combined with reduced rates of fertility, this points to a strong decline in the proportion of the Chinese population that are of working age (15 to 65 years) during the forecast period.

In this context, continued expansion of the capital stock (as a result of strong growth in investment) and ongoing improvements in primary factor augmenting productivity will be the key drivers of China’s economic growth during the forecast period.Key factors affecting changes in China’s demand for liquefied natural gasKey factors affecting


Table A9: Forecast simulation: growth rates of employment and population, and GDP deflator and import price index, 2014–30

Component Average annual growth rate (per cent)

2014 2015–19 2020–24 2025–30

Employment 0.35 –0.21 –0.04 –0.29

Population 0.42 0.44 0.22 0.06

GDP deflator 2.00 2.00 2.00 2.00

Import price index 1.00 1.00 1.00 1.00

Source: Employment and population data are from the UN (2012), GDP deflator data are from the World Bank (2015) WDI

online. The change of GDP deflator and import price index are authors’ assumptions

Policy simulation from 2014 to 2030

The policy simulation is conducted with a so-called policy closure, in which naturally exogenous variables are exogenous and naturally endogenous variables are endogenous. Naturally endogenous means that variables are typically thought of as model-determined in a CGE framework. These include large macro aggregates such as real GDP, national employment, national consumption and the trade balance; and micro variables such as industry output and price. Naturally exogenous means that other variables in the model are not determined by an equation or are controlled by policy makers. Good examples are rates of technological progress, the positions of world demand curves and tax rates.

In the policy simulations, nearly all of the exogenous variables adopt the values in the forecast simulation. The only exceptions are the policy variables. For example, if we are interested in the effects of a reduction in coal consumption, the relevant coal consumption variable is shocked with a value that differs from its baseline forecast path. The effects of coal consumption reduction on macro variables, CO2 emissions and other endogenous variables are calculated by comparing their paths in the policy simulation with their paths in the baseline (forecast) simulation.

Additional assumptions

In the CHINAGEM model, China can influence the world price of any product for which it has significant market power. This applies to most markets to which it exports, and in some markets from which it imports. Good examples of the latter are the markets for coal and iron ore (see Table A10). For these commodities, the model assumes that China faces market pressures such that any increase in China demand leads to an increase in world price and vice versa.

China imports natural gas as LNG delivered by ship and directly via pipelines from Central Asian countries and elsewhere. In the modelling, the CIF foreign-currency prices of LNG and pipeline gas are, for the most part, exogenous. This reflects an absence of empirical data for the elasticity of Chinese demand for imported gas. Ideally, given the importance of China in the world gas market, there should be an endogenous positive relationship, with a sustained increase in demand



leading to an increase in price (and vice versa). Note that in the modelling of some of the scenarios reported, the price of imported gas does change. For example, in the modelling of a higher world oil price, the foreign-currency price of gas rises in line with the world price of oil (see the described details in Section 7.2). In other simulations, the share of imported gas in total gas usage in China is targeted through model-determined changes in the price of imported gas.

Table A10: China’s import shares in global markets: iron ore, coal, oil and gas, 2007–14

Year (per cent) Iron ore Crude oil Coal Gas

2007 45.8 6.9 4.7 0.4

2008 49.1 7.6 3.7 0.5

2009 67.0 9.0 11.8 0.8

2010 59.2 10.4 14.6 1.5

2011 61.6 11.1 16.4 2.8

2012 63.7 11.8 19.7 3.9

2013 65.3 12.6 21.0 4.8

2014 67.7 13.4 18.3 5.3

Source: UNCTAD (2014) for iron ore, IEA (2015d) for crude oil, coal and gas



ReferencesBP (2015) Statistical review of world energy 2015, BP

Buera J & Kaboski J (2008) Scale and the origins of structural change, Federal Reserve Bank of Chicago Working Paper WP2008-06, Chicago, Federal Reserve Bank

China Statistics Press (2010) China statistical yearbook, http://www.stats.gov.cn/tjsj/ndsj/2010/indexeh.htm

China Statistics Press (2014) China statistical yearbook, http://www.stats.gov.cn/tjsj/ndsj/2010/indexeh.htm

Corbeau A, Braaksma A, Hussin F, Yagoto Y & Yamamoto T (2014) The Asian quest for LNG in a globalising market, IEA

Dehnavi J & Yegorov Y (2014) Future of shale gas in China and its influence on the global markets for natural gas, ResearchGate, https://www.researchgate.net/publication/272350555_Future_of_Shale_Gas_in_China_and_its_Influence_on_the_Global_Markets_for_Natural_Gas

Dixon PB, Parmenter BR, Sutton J & Vincent DP (1982) ORANI: a multisector model of the Australian economy, North-Holland, Amsterdam

Dixon PB & Rimmer M T (2002) Dynamic general equilibrium modelling for forecasting and policy: a practical guide and documentation of MONASH, North-Holland Publishing Company, Amsterdam

Dixon PB & Rimmer MT (2009) Forecasting with a CGE model: does it work? General Paper No. G-197, Centre of Policy Studies, Monash University, Melbourne

Dixon PB, Koopman RB & Rimmer MT (2013) The MONASH style of computable general equilibrium modeling: a framework for practical policy analysis, in Dixon PB & DW Jorgenson (eds.), Handbook of computable general equilibrium modeling, Volume 1A, North-Holland, Oxford, UK, pp 23–104

Green F & Stern N (2015) China’s ‘new normal’: structural change, better growth, and peak emissions, Policy Brief, Centre for Climate Change Economies and Policy, and Grantham Research Institute on Climate Change and the Environment

International Energy Agency (IEA) (2012) Power options for low-carbon power generation in China, designing an emission trading system for China’s electricity sector, IEA, Paris

International Energy Agency (IEA) (2014a) The Asian quest for LNG in a globalizing market, IEA, Paris

International Energy Agency (IEA) (2014b) World Energy Outlook 2014, IEA, Paris



http://www.stats.gov.cn/tjsj/ndsj/2010/indexeh.htm

http://www.stats.gov.cn/tjsj/ndsj/2010/indexeh.htm

International Energy Agency (IEA) (2015a) Energy and climate change: world energy outlook special report, IEA, Paris

International Energy Agency (IEA) (2015b) Gas medium-term market report 2015, market analysis and forecasts to 2020, IEA, Paris

International Energy Agency (IEA) (2015c) World Energy Outlook 2015, IEA, Paris

International Energy Agency (IEA) (2015d) IEA energy database, IEA, Paris

Jin Z (2015) Opportunities and challenges of shale gas development in China: a perspective from Sinopec’s experience, presentation at the 3rd IEA Unconventional Gas Forum, Chengdu, China, 10 April 2015

KPMG (2015) China Outlook 2015

Li X (2015) Natural gas in China: a regional analysis, The Oxford Institute for Energy Studies, UK

Liu X (2014) Shale gas development and challenges in China, Oxford Energy Forum, a quarterly journal for debating energy issues and policies, The Oxford Institute for Energy Studies, UK

Mai Y, Dixon PB & Rimmer MT (2010) CHINAGEM: A Monash-Styled Dynamic CGE Model of China, Working paper G-201, Centre of Policy Studies, Victoria University, Melbourne, Australia

National Development and Reform Commission (NDRC) (2014a) 2014–20 plan on upgrading and reforming energy saving and emissions reduction in coal-fired electricity generation, NDRC, Beijing, www.sdpc.gov.cn/gzdt/201409/W020140919603712991447.doc

National Development and Reform Commission (NDRC) (2014b) 2014–20 national plan on climate change, NDRC, Beijing, http://www.china.org.cn/china/2014-09/19/content_33560895.htm

Nexant (2015) World Gas Model, Nexant

Rutkowski R (2015) Service sector reform in China, Policy Brief, Peterson Institute for International Economies, Washington DC

Sheehan P, English A, Sun F & Cheng E (2014) The national and regional development of China’s gas market: beyond evolutionary change? report to Australia China Natural Gas Technology Partnership Fund, Centre for Strategic Economic Studies, Victoria University, Melbourne

The State Council of the People’s Republic of China (2014) The energy development strategic action plan 2014–20, State Council of the People’s Republic of China, Beijing, www.gov.cn/zhengce/content/2014- 11/19/content_9222.htm

The World Bank (2012) China 2030, World Bank, Washington DC

United Nations (UN) (2012) United Nations’ world population prospects: the 2012 revision



http://www.gov.cn/zhengce/content/2014-%2011/19/content_9222.htm

http://www.gov.cn/zhengce/content/2014-%2011/19/content_9222.htm

http://www.china.org.cn/china/2014-09/19/content_33560895.htm

http://www.sdpc.gov.cn/gzdt/201409/W020140919603712991447.doc

United Nations Conference on Trade and Development (UNCTAD) (2014) World Trade

US Energy Information Administration (EIA) (2015a) China briefing, EIA, Washington DC

US Energy Information Administration (EIA) (2015b) Shale gas development in China aided by government investment and decreasing well cost, EIA, Washington DC

Wang Z (2015) China’s elusive shale gas boom, Paulson Papers on Energy and Environment, The Paulson Institute, Chicago, http://www.paulsoninstitute.org/think-tank/paulson-papers-energy-environment/2015/03/12/chinas-elusive-shale-gas-boom/

Wayne MA (2012) China's shale-gas boom slow to start, Wall Street Journal, 3 December 2012, viewed 17 March 2016, http://online.wsj.com/news/articles/SB10001424127887323401904578156710038647662

Wensley S, Wilson S & Kuang J (2013) China’s energy demand growth and the energy policy trilemma, China Update 2014

World Energy Council (2015) 2015 Energy Trilemma Index: benchmarking the sustainability of national energy systems, WEC, London, UK



http://online.wsj.com/news/articles/SB10001424127887323401904578156710038647662

http://online.wsj.com/news/articles/SB10001424127887323401904578156710038647662

http://www.paulsoninstitute.org/think-tank/paulson-papers-energy-environment/2015/03/12/chinas-elusive-shale-gas-boom/

http://www.paulsoninstitute.org/think-tank/paulson-papers-energy-environment/2015/03/12/chinas-elusive-shale-gas-boom/

Documents

Business as usual - Department of Industry, Innovation · Web viewThis has a dampening effect on economic growth, leading to marginally lower gross domestic product (GDP) growth. More