Economic benefits of Northeast Asia energy interconnection: A quantitative analysis based on a computable general equilibrium model

0 Introduction

Northeast Asia has maintained a massive increase in energy consumption.The total primary energy demand increased from 1.7 billion tonnes of standard coal equivalent (tce)in 2000 to 2.99 billion tce in 2017,with an average annual growth rate of 3.4% [1].Meanwhile,energy consumption of Northeast Asia has been extensively dominated by fossil fuels,resulting in severe regional environmental pollution.As climate change and other environmental considerations are becoming increasingly important,a more sustainable energy strategy is required.Northeast Asia possesses abundant renewable resources that could contribute toward its future energy needs,the resources and demand market are also mutually supportive.Russia and Mongolia are rich in clean energy resources but low in domestic demand,while China,Japan,and South Korea are short in energy resources but strong in technology and capital.Hence,regional energy cooperation could play a key role in the sustainable development of Northeast Asia.

There has been significant interest in energy cooperation of Northeast Asia [2-4].Von Hippel (2011)introduce a new concept of energy security for Northeast Asia as climate change and other international considerations have increased in importance.The new framework includes energy supply,economic,technological,environmental,socio-cultural,and military security factors [5].Von Hippel(2011)reviewed some energy infrastructure cooperation proposals in Northeast Asia,and important analysis factors which could contribute to the success or failure of infrastructure proposals [6].Although many have analyzed ultra high voltage transmission,smart grid,and renewable energy generation,studies on combining all three to form a systematic energy network were only relatively recently developed.Liu (2015)introduces a global energy interconnection (GEI)framework which mapped out the global renewable energy reserves and electricity demand,and thereby proposed the first plan for a world-wide electricity transmission network [7].Liu (2016)proposed to establish a Northeast Asia grid interconnection in an effort to facilitate optimized resource allocation under the GEI framework [8].The Global Energy Interconnection and Development Organization (GEIDCO)henceforth produced several more detailed and updated studies for a global and regional energy interconnection framework[9,10].

However,current analysis related to the energy interconnection of Northeast Asia has been strong on engineering modeling but weak on economic modeling.A few studies have used system optimization models to analyze inter-regional electricity trade [11-14].These works are important,but they only tell part of the story.The goal of energy interconnection is to promote renewable energy development that would eventually lead to better and more sustainable economic development.Understanding the potential socio-economic implications would help with better planning and execution of Northeast Asian energy cooperation.

It is possible to use existing economic modeling frameworks to analyze the benefit of energy interconnection development—albeit with appropriate modifications.To analyze the socio-economic implications,one needs a global,macroeconomic model that accounts for the interregional transmission of renewable energy sources.The most commonly used macroeconomic model for policy analysis is the computable general equilibrium (CGE)model.A large body of literature has used CGE models to analyze electricity-related issues [15-20].Jin et al.(2018)used the CGE model to study the global socio-economic implications of energy cooperation development [21].

In this study,we advance the analysis of economic benefits of energy interconnection in Northeast Asia by making progress in three key areas.First,we establish an appropriate energy-nesting structure that allows flexible substitution between electricity and fossil fuels.Second,we distinguish different types of electricity generation in our CGE model.Third,we re-estimate elasticity of substitution parameters among different types of power sources.

1 Method

1.1 Model

We start our modeling work from the energy environmental version of the standard Global Trade Analysis Project (GTAP-E)[22].The GTAP-E model is a special version of the GTAP model as it has a nesting structure that treats energy inputs as substitutes in a multilevel fuel-factor nest.In addition,GTAP-E also specifies the elasticity of substitution parameters for the constant elasticity of substitution functions that underly these nesting structures.This allows the model to substitute among different energy inputs and factors of production when their relative prices change.Our model also inherits all the elasticity of substitution parameters between fuels and factors from the GTAP-E model.We expand the GTAP-E nesting structure by adding a new three-layer nesting structure for the electricity sector,details are presented in the author’s previous research [23,24].

1.2 Database

We then incorporate the GTAP-power database into the standard GTAP-E model.The original GTAP-power has 140 industry sectors,in which the electricity sector was disaggregated into 12 different sectors [25,26].To be consistent with the energy interconnection activities mentioned afterwards and increase computation efficiency,we aggregated the original industry sectors into 21 sectors(Table 1),where energy-intense sectors and six electricity generation sectors were retained [23].

Table 1 Industry sectors in GTAP-power after aggregation

Code Industry Agri Agriculture products Coal Mining and agglomeration of coal Oil Extraction of crude petroleum and related Gas Extraction of natural gas ＆ gas distribution and related OMN Other Mining Extraction

continue

Code Industry Oil_pcts Petroleum ＆ Coke: manufacture of coke and refined petroleum products I_S Iron ＆ Steel mtp Non-Ferrous Metals CRP Manufacture of chemicals,pharmaceuticals,rubber,and plastics products nmm Manufacture of non-metallic mineral products OMF Other industry tnd Electricity transmission and distribution FFP Fossil fuel power HydroP Hydropower WindP Wind power SolarP Solar power OtherP Other power cns construction trd trade tsp Transport SER Service

1.3 Estimate of elasticity of substitution

We estimate the elasticity of substitution between fossil fuel and non-fossil power generation (σGEN)using econometric techniques.We do this only for Asia,as the current study focuses on Northeast Asia and its related neighbors in Asia.We use IEA (2018)power generation data for all Asian countries between 2000 and 2015 [27].The CES production function is specified in Eq.(1).

In (1),Elec represents the amount of total power output.Fossil fuel power (FFP)and Non-fossil fuel power (NFP)represent total fossil fuel power output and total non-fossil fuel power output,respectively.By estimating β1,β2,and β12,we could derive σGEN using Eq.(2).

The standard GTAP-E is a static model.We chose to use a dynamic model to analyze the different GEI development phases in different years.We added the Monash-style dynamisms to the GTAP-E model.The mechanisms of labor and capital markets were presented in the author’s previous research [23].The full documentation of the Monash dynamisms can be found in [28].

1.4 Base-case

We build the dynamic base-case in two phases.A dynamic CGE needs a base-case scenario.A base-case serves as the ‘business-as-usual’ scenario to which policycases are compared.A base-case should embody reasonable economic development dynamisms that not only underly the base-case,but also the policy-case.We formulate our base-case in two parts—the historical base-case and the projection base-case.The historical base-case is between 2011,the base-data year,and 2017,the year for which the latest data are available.In the historical base-case,we control GDP,population,and employment levels by endogenizing total factor productivity.We take the actual data from the leading French center for research and expertise on the world economy (CEPII).We also control private consumption and investment levels,by endogenizing the marginal propensity to consume and expected rate of return,respectively.The consumption and investment data are taken from the world development indicator (WDI)of the World Bank database [29].

The projection base-case is between 2018 and 2050.As in the historical base-case,we control GDP,employment,and population levels in the projection base-case.No actual data are available for these years,so we use projections from the literature.We use CEPII’s projection for GDP,labor,and employment,again,by endogenizing the total factor productivity.To have reasonable GEI implications,we also need reasonable energy projections.We use energy projections from the New Policy Scenario,World Energy Outlook 2018,to formulate projections for coal,oil,gas,and total electricity use,as well as projections for fossil-fuel power,hydropower,wind power,solar power,and other power.Sector production technologies are endogenized to facilitate energy sector development.

2 Energy interconnection scenarios

In the energy interconnection framework of Northeast Asia [1],only the production and trade of hydropower,wind power and solar power are considered.Production and demand of other energy commodities will be generated endogenously.

2.1 The three phases

We envision three key phases of energy interconnection development in Northeast Asia.The preparation phase is between 2018 and 2020.This phase is primarily for consensus building.The construction phase is between 2021 and 2030.Construction will occur in this period,although no additional production or trade in renewable energy will occur.Investment and financing,however,are required.The operational phase will begin in 2031.The additional renewable generation capacity and electricity network will become operational in the third phase.

2.2 Four energy interconnection activities

We identify four primary energy interconnection activities in Northeast Asia—trade,production,investment,and financing.Trade is the core of the energy interconnection activities; it refers to the additional interregional trade of energy interconnection products that are made possible by energy interconnection development.Production refers to the additional output of energy interconnection products owing to expansion of the energy networks and trade.Investment and finance refer to the capital required to realize construction and operation.A key contribution of this study is to model these activities.

We model the trade activities by processing GEIDCO results [1].Using optimization models,and by considering historical electricity generation capacity,costs,discount rates,renewable energy reserve,carbon emissions,and the balance between electricity supply and demand,GEIDCO(2019)estimated the transmission capacity and trade of GEI products in Northeast Asia [1].To distinguish the effects for different countries,we only consider country or region that has electricity trade between others and regroups this region to China,Russia,and Northeast Asia (NEAsia,including Japan,North Korea,and South Korea).Fig.1 shows the electricity trade flow in this region.Northeast Asia is projected to import a total of 220 TWh of wind power through GEI development in 2050.Of this total,120 TWh will come from Russia and 100 TWh will come from China.Using quantity data from the IEA (2018)and value data from our updated GTAP-power database,we deduced the electricity prices of electricity transactions in Northeast Asia in 2016 (Table 2).

Fig.1 Projected transmission capacity (kW)of GEI products in 2050 [1]

Table 2 Quantity and value of electricity transactions in Northeast Asia,2016

Sources: Russia China NEAsia GTAPpower Billion USD 260.8 442.5 283.3 IEA TWh 1573.8 6217.9 1643.2

Assuming the electricity transaction price remains the same as in 2016,by applying these prices,and usingour projected electricity trade values in our model,we estimate the quantity of electricity transactions under the base-case scenario in 2030.Then,with the projected trade quantities in 2030 and the estimated trade quantities in 2050,we calculate the annual percentage changes in trade of GEI products between 2031 and 2050—105% per year for imports from Russia and 115% per year from China.To accommodate such increases,we allow inter-regional electricity transmission technology to improve endogenicity in the model.

We model the production activities according to the trade activities.The large-scale increase in wind power import to the NEAsia region shall be supplied by the additional wind power capacity installed in north-east China and the far east of Russia.These additional wind power outputs will increase the exporting regions’ total wind power output accordingly.Using the same method as in trade development increases,we calculate that between 2031 and 2050,and comparing with the base-case,GEI will make China and Russia increase total wind power production by 25% and 463%,respectively.To facilitate these increases,we allow domestic production technology to improve endogenously.

We model financing and investment by processing the GEIDCO results.The GEIDCO estimates that Asia will invest a total of USD $8.5 trillion between 2018 and 2050 in GEI development [1].We assume that all of the investment and financing activities will occur in the construction phase.Thus,we divide the total investment requirements to different sub-regions and calculate the investment increase in each region.Financing shall be equal to investment.We assume investment in each region is locally financed by reducing consumption (equivalent to increasing savings).We calculated that the total investment increase,against the base-case,in China,Russia,and NEAsia are 3.38,1.59 and 11.3%,respectively.Additionally,the total reduction in consumption,against the base-case,in China,Russia and NEAsia are 1.13,0.53,and 3.89%,respectively.To allow changes in investment and consumption,we endogenize the expected rate of return to capital and marginal propensity to consume in the model.

3 Simulation results

3.1 Regional GDP,employment

Simulations show that GEI has a positive contribution to Northeast Asia.We show the cumulative real GDP percentage deviations in Fig.2.Through GEI development,the entire region would enjoy higher GDP.Northeast Asia benefits the most—its real GDP is 0.6% higher than it was in the base-case.China and Russia’s GDP are 0.05% and 0.27% higher,respectively.The primary reasons for GDP improvement are different between the electricity exporting and the electricity importing regions.Higher employment is a key contributor to higher GDP in NEAsia —the importing region for GEI products.

Fig.2 Change in GDP of China,Russia,and NEAsia.(%,cumulative deviation from the base-case)

Employment in NEAsia is 0.7% higher than it was in the base-case (Fig.3).This is higher than the GDP increase.Productivity gains in the wind power sector,however,is the main driver for the higher GDP in China and Russia—the exporting regions of GEI products (we will show this in a later subsection).Northeast Asia benefits the most within this region as it enjoys cheaper wind power imports from both China and Russia,especially when the scale of electricity trade increases after 2030.Cheaper electricity prices could reduce business costs,which allows companies to expand and hire more employees.

Fig.3 Change in employment of China,Russia,and NEAsia.(%,cumulative deviation from base-case)

3.2 Consumption,investment

Within this region,there exists a trade-off between investment and consumption (Figs.4 and 5).We assume investment equal to GEI savings and investment is financed by reducing consumption in our model.Hence,an increase in GEI development investment reduces people’s ability to consume.The trade-off is the smallest in Russia,as both the increase in investment and the decrease in consumption are relatively small.The trade-off is most obvious in NEAsia,as both the increased investment and the decreased consumption are relatively large.That said,there also exists a convergence in investment and consumption—both towards the base-case.This convergence suggests that in the long-run,both investment and consumption may approach base-case levels.In that case,the final economic demand structure will return to their base-case conditions,while total output increases.

Fig.4 Change in investment of China,Russia,and NEAsia.(%,cumulative deviation from base-case)

Fig.5 Change in consumption of China,Russia,and NEAsia.(%,cumulative deviation from base-case)

3.3 Regional output,price

3.3.1 Electricity

Total electricity output changes are much smaller (Fig.6).China’s electricity output is 0.2% lower than it was in the base-case,Russia’s is 0.05% higher,and NEAsia’s is 0.3%higher.The small decrease in China’s total electricity output is owing to the combination of two effects.In one sense,the output effect puts downward pressure on electricity use.This is because electricity output sells primarily to consumption-oriented rather than investment-oriented users.The higher consumption and lower investment in China thus have a negative impact on electricity demand.Conversely,the substitution effect has a positive impact on electricity use.Lower wind power price reduces total electricity price,which encourages users to use more electricity and less alternate energy.

Fig.6 Change in electricity power output of China,Russia,and NE Asia.(%,cumulative deviation from base-case)

Overall,the negative output effect dominates the positive substitution effect,and China’s total electricity use falls minimally.In Russia,owing to the smaller decrease in consumption (smaller negative output effect)and the larger reduction in wind power price (larger positive substitution effect),there is a smaller increase in electricity output.In NEAsia,the output effect dominates in the earlier years.As the amount of imported wind power gradually increases and becomes more influential to total electricity price,the substitution effect dominates and leads to an overall increase in total electricity use by 2050.

3.3.2 Fossil fuel

Fossil fuel power output in China,Russia and NEAsia all fall compared to the base case (Fig.7).This is as expected,because our nesting structure allows substitution away from fossil fuel power and towards cheaper,nonfossil fuel power.Fossil fuel power output in China,Russia,and NEAsia are 0.13,0.3 and 2.3% lower than the basecase,respectively,In the earlier years,the fall in fossil fuel power output is primarily because of the decrease in total power output.In the later years,especially past the 2040s,NEAsia’s FFP output decouples from its total electricity growth.The gap is obviously filled by a significant amount of wind power imported from China and Russia.

Fig.7 Change in fossil fuel power output of China,Russia,and NEAsia.(%,cumulative deviation from base-case)

The GEI development allows renewable energy to displace fossil fuel energy.Changes in fossil fuel output in China,Russia,and NEAsia are shown in Figs.8-10.Northeast Asia’s output in all three fossil fuel sources falls significantly.Wind power displaces fossil fuel in two key ways.First,wind power displaces fossil fuel power in the overall power mix.This has been shown as the decoupling of FFP from total electricity output.Second,electricity displaces fossil fuel energy.As cheaper wind power imports become available in the NEAsia market,users consume more electricity overall,and less fossil fuel energy.

Fig.8 Change in coal output of China,Russia,and NEAsia.(%,cumulative deviation from base-case)

Fig.9 Change in oil output of China,Russia,and NEAsia.(%,cumulative deviation from base-case)

Fig.10 Change in gas output of China,Russia,and NEAsia.(%,cumulative deviation from base-case)

3.3.3 Other sectors

We show output levels in non-energy sectors in Figs.11 and 12.During the construction phase,GEI infrastructure construction stimulates the construction sector growth(cns).In the GEI exporting regions,sectors,such as iron and steel (I_S),that support generation capacity expansion generally grow more.Service oriented sectors,however,contract slightly,as production factors leave to support GEI development.In NEAsia,the GEI-importing region,households are the primary electricity users,so the service sector,which primarily sells to households,expands more.

Fig.11 Changes in other sector outputs of China.(%,cumulative deviation from base-case)

Fig.12 Changes in other sector outputs.(%,cumulative deviation from base-case)

3.4 Emissions

The CO2 emissions in China,Russia,and NEAsia all decrease (Fig.13).Northeast Asia’s CO2 emissions decrease the most compare to the base-case.The CO2 emissions pattern deviation from the base-case is similar to that of the FFP output.This is because the FFP sector is a major CO2 emitter and that it has been given large,negative shocks owing to strong wind power growth.The decrease in CO2 emissions and the increase in real GDP together mean that GDP emission intensity also decreases in China,Russia,and NE Asia.

Fig.13 Change in CO2 emissions of China,Russia,and NEAsia.(%,cumulative deviation from base-case)

4 Concluding remarks

This study models the economic implications to the regions that participate in the GEI development in the Northeast Asia region.To do so,we applied the GTAP-E model,incorporated it with the GTAP-power database,established a new fuel-factor nesting structure,econometrically estimated the key CES parameters,built a base-line using data and projection from CEPII,WDI,and IEA,for the years between 2011 and 2050,and developed the GEI scenarios using GEIDCO results [1],for the years between 2018 and 2050.This is a new modeling strategy that can be used to model the economic implications of GEI development in not only the Northeast Asia region but also elsewhere in the world.

Our key findings include,first,we estimate that the constant elasticity of substitution parameters between an electricity and non-electricity bundle is 1.37 in Asia.Second,China,Russia,and NEAsia all benefit from participation in GEI development and experience higher GDP growth than their base-case levels.Northeast Asia benefits the most and in 2050 its real GDP is 0.6% higher than it was in the basecase.Third,there exists a trade-off between investment and consumption.Investment in GEI expansion increases overall investment and reduces overall consumption in the participating GEI development regions.Fourth,regions that export GEI products—wind power,in this case —benefit from expanding clean power production.Sectors that are investment-oriented,and upstream of the GEI development,such as construction and iron and steel also expand.Sectors that are consumption oriented,however,are slightly worseoff.Fifth,the GEI product importing region—NEAsia,in this case—benefit from cheaper electricity prices.Thus,primary electricity users,such as households,and their upstream sectors,such as the service sector improve.Sixth,GEI development displaces fossil fuel energy use and contributes to carbon dioxide mitigation.CO2 emissions will fall by 0.21,0.16,and 1.25% in China,Russia,and NEAsia,respectively.

Owing to the scope limitations,our study cannot fulfill every modeling task.One potential limitation is the lack of sector-level investment details,which is a database limitation.Therefore,we cannot shock investment in the power sector specifically,but must instead increase investment in the entire region.Another potential limitation is the lack of differentiation between natural resources and ordinary capital.The GEI development should turn many barren areas,such as deserts and valleys into productive natural resources.This will not create competition among capital users and will thus contain capital costs.Future studies may consider overcoming these limitations.Nevertheless,this study can be used as a guide to analyze the economic implications of GEI development in other regions elsewhere in the world.

Acknowledgements

This work was supported by the Overseas Expertise Introduction Project for Discipline Innovation (B18014);National Natural Science Foundation of China (71733002);Science and Technology Foundation of SGCC(52450018000N).

Declaration of Competing Interest

We have no conflict of interest to declare.