Energy–Economy–Carbon Emissions: Impacts of Energy Infrastructure Investments in Pakistan Under the China–Pakistan Economic Corridor

Abstract

Energy–economy–environment sustainability is critical in shaping energy policies, especially in developing countries facing energy shortages. Investment in energy infrastructure, such as under the China–Pakistan Economic Corridor (CPEC), provides an opportunity to explore how such investments impact economic growth, environmental quality, and energy security. This study examines the energy, economic, and environmental effects of CPEC’s energy investments in Pakistan, covering a range of power sources, including coal, hydro, solar, wind, and nuclear energy. Utilizing data from 31 CPEC energy projects and employing the GTAP-E-Power model, this research assesses these impacts through seven scenarios, comprehensively analyzing the heterogeneity of different power sources. Our findings reveal that while all types of CPEC energy infrastructure investments contribute to increasing the share of zero-emissions electricity to 49.1% and reducing CO₂ emissions by 18.61 million tons, the economic impacts vary significantly by energy source. The study suggests that it is crucial to prioritize renewable energy investments while addressing immediate power shortages to balance economic growth with environmental sustainability. Policymakers should also consider the potential inter-sectoral substitution effects when applying significant shocks to specific sectors. This analysis informs future energy investment decisions under CPEC and offers insights for other Belt and Road Initiative (BRI) countries aiming to optimize their energy strategies for sustainable development.

1. Introduction

Energy–economy–carbon emissions is a critical framework for understanding how energy infrastructure investments can influence a nation’s economic growth, environmental quality, and energy security. Energy infrastructure is not only a driver of economic development by powering industrial activities and enabling technological advancement but also a determinant of environmental outcomes due to its impact on greenhouse gas emissions and natural resource use. Developing countries like Pakistan, which face chronic energy shortages and environmental challenges, stand at the crossroads of needing to expand their energy capacity while balancing economic growth and environmental sustainability. Investments in energy infrastructure, such as those under the China–Pakistan Economic Corridor (CPEC), offer a unique opportunity to explore how different energy sources—coal, hydro, solar, wind, and nuclear—affect this delicate balance, providing valuable insights for policymakers on achieving sustainable development.

The CPEC is a strategic economic project to enhance economic connectivity between Pakistan and China. CPEC’s development has involved substantial investments over three five-year phases, focusing on energy, Gwadar port, transport infrastructure, and industrial cooperation, with a strong emphasis on energy infrastructure (as detailed in Appendix A

Table A1. CPEC projects.

		Cost (USD Million)			Projects Numbers
Project Types	Status	2016–2020	2021–2025	2026–2030	2016–2020	2021–2025	2026–2030
Energy	Completed	7899	14798	0	9	7	0
	Under Construction	0	3323	3500	0	3	1
	Under Consideration	0	0	6167	0	0	5
Transport Infrastructure	Completed	5878	692	0	5	1	0
	Under Construction	0	486	0	0	5	0
	In Pipeline	0	531	8468	0	3	5
	Long Term	0	0	3303	0	0	5
Gwadar Projects	Completed	4	602	0	1	5	0
	Under Construction	0	289	0	0	4	0
	In Pipeline	0	0	150	0	0	4
Industrial Cooperation	Under Construction	0	-	0	0	4	0
	In Pipeline	0	0	0	0	0	5

Note: While nuclear power plants are not explicitly mentioned in [63], they are included in the broader China–Pakistan collaboration under CPEC, particularly through the involvement of the China National Nuclear Corporation (CNNC). The total investment is also likely higher than the figures listed, as some projects lack specified cost estimates. Source: CPEC official website [71].

). By 2030, around USD 35.7 billion—over 50% of the total USD 56 billion CPEC investment—will be directed towards energy infrastructure, resulting in an additional installed generation capacity of 16,379.7 MW.

These significant energy infrastructure investments are crucial for Pakistan. Due to inadequate planning [1], supply shortages and distribution problems lead to costs of as much as 2% of GDP growth a year [2]. Along with the problem of power lost during transmission and distribution, Pakistan’s electricity shortfall has climbed to 6623 MW [3]. Moreover, 65.88% of Pakistan’s electricity is derived from fossil fuels, costing USD 12.33 billion annually and contributing to 28.3% of the country’s greenhouse gas emissions [4]. The government has been striving to reduce the power shortfall by attracting more investment into the power sector, and CPEC is the solution Pakistan has sought for over a decade [5]. For China, CPEC promotes technological collaboration and economic gains through construction contracts and power plant operations. Globally, CPEC is a model for how developing countries can leverage advanced technology and capital to achieve economic growth while reducing carbon emissions. Thus, analyzing how CPEC energy infrastructure investment will enhance energy security, foster economic growth, and promote environmental sustainability is crucial for providing valuable insights to policymakers.

Many studies have discussed the economic impacts of CPEC, focusing on connectivity [6,7,8], port construction [9,10], transportation infrastructure [11,12,13], and special economic zones [14]. However, the economic impact of energy infrastructure investment under CPEC remains largely uninvestigated, except for Pakistan’s vision 2025 [15] and a study by Li et al. [16]. Most scholars emphasize the impact on Pakistan’s energy development; for example, [4,5,6]. However, most related studies, except for Duan et al. [4], fail to account for heterogeneous power sources and their substitution or complementarity interactions. Additionally, the critical role of transmission lines is often neglected despite their importance for improving Pakistan’s energy system.

In contrast, the environmental impact, especially CO₂ emissions, is a key concern in the sustainability of CPEC energy investment, given the large-scale development of energy infrastructure [17] across various power sectors, from coal to nuclear energy power plants [18], which could potentially lead to significant changes in CO₂ emissions. These concerns parallel those regarding China’s investment through the BRI, which can significantly contribute to international economic activities and inevitably reshape the pattern of CO₂ emissions [19]. There is a growing consensus that hydropower cooperation [20] and nuclear cooperation [21] will bring cleaner energy to host countries (see Appendix A

Table A2. Representative studies on electric power investment.

Author(s)	Method	Sample	Main Findings
Bega and Lin (2024) [20]	In-depth literature review	China’s BRI hydropower cooperation	Benefit BRI developing countries by rendering electricity generation more efficient and less carbon-intensive
Lin et al. (2020) [21]	In-depth literature review	China’s BRI nuclear cooperation projects	Nuclear power should be a part of clean and low-carbon development solutions for fast-growing economies.
Tritto (2021) [22]	In-depth interviews	Coal plants in Indonesia under BRI	Indonesia prioritized economic growth over environmental and social sustainability
Tao et al. (2020) [17]	Collected data using a bottom-up approach and directly calculated CO2 emissions	458 power plant development projects in 15 BRI countries	Increased fossil fuel generation capacity led to higher carbon emissions
Ali et al. (2021) [23]	Direct Observation Techniques	Seven coal-fired power plants under CPEC in Pakistan	Likely to emit 14,500 metric tons of CO2 in 2020, 65% higher than the national trend
Bhandary and Gallagher (2022) [24]	In-depth interviews	Coal-fired power plant under CPEC	The fossil-fuel-intensive nature of these projects has created environmental harm
Gallagher et al. (2021) [25]	In-person interviews	Coal-fired power plants in India, Indonesia, Bangladesh, and Vietnam under BRI	Has the potential to diffuse cleaner energy technologies throughout the developing world
Lin and Bega (2021) [18]	In-depth literature review	China’s BRI coal power cooperation	Generate opportunity for the BRI participants, as coal energy can render electricity generation more efficient in these countries, mostly from the developing world, with larger and cleaner plants, providing flexibility for greater integration of renewables.
Bega and Lin (2023) [26]	SWOP-Analytic Hierarchy Process	Electric power projects under BRI	Offers cleaner and more efficient electricity generation, particularly in developing countries

Notes: 1. Some studies, for example, Munir and Khayyam (2020) [72], examined CPEC environmental impacts from a broader perspective, such as national ecological health, which are beyond the scope of our analysis. We focus specifically on carbon dioxide emissions, which aligns with most studies on BRI that address environmental effects. 2. This table highlights some representative studies on the environmental impacts of China’s BRI energy infrastructure investments, including CPEC, based on project data.

). However, significant concerns remain regarding fossil fuel power, particularly coal-fired plants. Some authors argue that these plants, especially coal-fueled, contribute substantially to environmental degradation; for example, Tritto [22] suggested that Indonesia prioritized its economic growth over environmental and social sustainability. Tao et al. [17] found that increased fossil fuel generation capacity led to higher carbon emissions, varying across different scenarios. Ali et al. [23] and Bhandary and Gallagher [24], who focused on the coal-fired power plants of CPEC, concluded that they negatively affect Pakistan’s environment. However, some studies have found different outcomes. Gallagher et al. [25] concluded that China’s BRI has the potential to become the largest means for the diffusion of cleaner energy technologies throughout the developing world, contingent on recipient countries exercising more stringent environmental regulations and changes in Chinese overseas investment policies. Findings by Lin and Bega [18] revealed that the BRI coal power projects generate opportunities for the BRI participants, as coal energy can render electricity generation more efficient in these countries, mostly developing countries, with larger and cleaner plants, providing flexibility for greater integration of renewables. Bega and Lin [26] suggested that the electric power projects under BRI offer cleaner and more efficient electricity generation, particularly in developing countries. Existing studies demonstrate that the type of energy source—whether coal or renewable—has significantly different environmental implications, with most concerns centered around coal rather than other energy sources.

Some key research gaps in this area are summarized as follows: First, most recent studies on the energy impacts of CPEC often fail to fully account for the interactions between sub-power sectors and the effects of transmission lines, which are critical for understanding the overall implications of energy infrastructure. Second, recent studies often overlook the substitution or complementarity effects between different sub-power sectors and other indirect effects, especially when conducting a partial equilibrium analysis, which limits the ability to capture indirect effects on industrial and economic development, leaving gaps in understanding the full economic impact of CPEC’s energy infrastructure development. Third, many studies on carbon emissions from energy investments under CPEC and BRI, especially coal-fired power plants, show mixed results. A new research framework is needed to explore and mitigate these inconsistencies, providing clearer implications for policymakers regarding balancing energy needs with environmental sustainability goals.

This study aims to comprehensively assess the impact of energy infrastructure investments on energy, economic, and carbon emissions. This goal is achieved by considering the heterogeneous effects and interactions between sub-electricity sectors, as well as the indirect effects of these interactions on industrial sectors and overall macroeconomic outcomes and carbon emissions, by using the GTAP-E-Power model, a global computable general equilibrium (CGE) model with detailed electricity sub-sectors developed by Peters [27,28]. This assessment will offer a comprehensive perspective for Pakistan and other developing countries, considering that CPEC is a flagship project of the BRI and is considered a model for how developing countries can achieve sustainable development through foreign investment, capital, and technology.

The major contributions of this study are as follows: First, based on project-level data, by simulating generation capacity shocks across different energy sources through seven scenarios, this study examines the heterogeneous impacts of different electricity generation sources and explores their substitution or complementarity effects. This approach provides valuable insights that address research gap 1, where prior studies often neglected to consider how various energy sources interact within the broader energy infrastructure. Second, this study investigates both the direct effects on energy production and carbon emissions, as well as the indirect effects on industrial development, economic growth, and carbon emissions. This comprehensive approach provides valuable insights into the broader implications of energy infrastructure investments, addressing research gap 2. Third, when considering the heterogeneous impacts of different electricity generation sources and the indirect economic effects, this paper addresses ongoing debates surrounding carbon emission studies. Doing so adds valuable insights to research gap 3, enhancing academic discourse and offering critical implications for policymakers involved in energy and economic planning.

The rest of the paper is structured as follows: Section 2 reviews the relevant literature, Section 3 outlines methodology and database, Section 4 presents energy projects’ data and empirical stimulation, Section 5 shows the results, Section 6 discusses findings and limitations, and Section 7 concludes the study.

2. Literature Review

2.1. Energy Infrastructure Investment and Energy-Economic Sustainable Development

Energy infrastructure is vital for developing countries as it enhances energy production and improves energy structures [29]. Investments in energy infrastructure—whether domestic or foreign, public or private—may be crucial in ensuring energy efficiency [30], supporting industrial development, and driving economic growth.

Most scholars focus on the CPEC energy infrastructure investment’s impact on Pakistan’s energy development, leaving its industrial and economic effects underdeveloped. Ahmed et al. [6] conducted a meta-analytic review and revealed strong evidence supporting the achievement of Pakistan’s energy security dream (89%), concluding that a package of energy infrastructure investment accompanies great potential in solving the ongoing energy crisis and eventually leading Pakistan towards energy security. Using a System Dynamics model, Duan et al. [4] found that CPEC projects would increase both energy and capacity growth, from 4.5% (1990–2015) to 14.6% (2016–2035) and 4.9% (1990–2015) to 12.3% (2016–2035), respectively. Ahmad et al. [5] show that many power generation projects recently commissioned under the CPEC have added sufficient installed capacity to the system, and the capacity shortfall no longer brings about power outages. Mirza et al. [31] highlighted that significant energy infrastructure investments are planned to address Pakistan’s energy crisis by utilizing existing resources. However, they focused on CPEC’s economic activities’ impact on overall energy consumption and savings potential by 2030. Except for Duan et al. [4], most studies fail to differentiate between heterogeneous power sources, resulting in limited research covering all types of energy infrastructure projects and their interactions. Additionally, few studies include the critical role of transmission lines, which are essential for the overall energy system.

When it comes to the economic effects, many studies discussed the economic impacts of CPEC, seldom focusing on the economic impacts of its energy investment. Only several articles study the economic impact of energy infrastructure investment under CPEC. Pakistan’s ‘Vision 2025’ projected a reduction in power outages due to enhancements in electricity generation capacity and investment in other production sectors of the economy. The GDP growth rate of Pakistan is targeted to increase to 7.5% by 2030, with the addition of about two million jobs in the job market [15]. In the context of BRI, as China’s international hydropower projects have been streamlined as a major focus of the BRI, Bega and Lin studied China’s BRI hydropower cooperation and found that the hydropower projects have shown many positive outcomes in promoting the energy infrastructure and local economic development in BRI countries. For energy infrastructure investments in other regions, Song et al. [32] found that energy infrastructure investment significantly promotes renewable energy generation in major developing Asian economies, and that increases in financial development and economic openness further drive renewable energy use. Zhou et al. [33], using data from 43 African countries from 2003 to 2021, applied fixed-effects models and a two-step Generalized Method of Moments (GMM) and found that China’s energy investments are closely associated with increased access to sustainable electricity in Africa, with this relationship being especially pronounced in resource-rich countries. Moreover, Schreiner and Madlener [34] conducted a macroeconomic evaluation of Germany’s planned investments in grid infrastructure through a static input-output analysis (IOA). They found that, although grid investments have a significant positive multiplier effect on total output, the impact on value-added and employment is negative, suggesting that large-scale infrastructure investments may bring mixed economic effects. Similarly, Li et al. [16] analyzed the job-creating and production-boosting effects of renewable power plant projects invested by China. However, recent studies have left the overall economic impact of various CPEC energy infrastructure investments unexplored.

2.2. Energy Infrastructure Investment and Environment Sustainable Development

Energy infrastructure is key for any modern and growing society [29]. Energy infrastructures are vulnerable to climate change, particularly in countries with poor availability [35]. Due to the lack of capital and technology, foreign energy infrastructure investment is broadly discussed in developing countries. The expansion of the BRI caused many discussions of the energy investment in the BRI in the partnering countries [16]. Guiding more countries to pursue low-carbon transition is essential to maintaining the sustainability of the BRI energy investment [18]; thus, environmental effects are hot topics in the context of BRI.

Specifically, the BRI involves large-scale infrastructure development, including energy infrastructure [17], covering a range of power sectors from coal to nuclear energy [26], and large-scale infrastructure construction. The large amount and variety of infrastructure development projects associated with the BRI means it could have a substantial influence on energy development in a large part of the world, have profound associated impacts on Earth’s environment and climate [36], and may accelerate energy consumption and deteriorate the environment of BRI countries [37]. Studies have shown that BRI has significantly contributed to international economic activities and inevitably reshaped the pattern of carbon dioxide (CO₂) emissions [19]. BRI electric power development, covering coal-fired power projects [18,22], as well as hydro, wind, and solar power projects and civil nuclear cooperation, was deemed to have a significant margin for improvements regarding carbon dioxide (CO₂) emissions reduction [17].

Consensus has been reached that renewable and nuclear energy investments will bring cleaner energy to host countries. At the same time, the biggest debates center around fossil fuel power, especially coal-fueled power, as highlighted in [17] that three-quarters of the capacity was concentrated on fossil fuels and predominantly coal power. Countries such as India, Indonesia, Bangladesh, Vietnam, and Pakistan, among the top recipients of Chinese coal financing [38], have garnered special attention. Based on fifteen in-depth interviews, Tritto [22] suggested that Indonesia prioritized its economic growth over environmental and social sustainability, highlighting the dominance of polluting subcritical plants from Chinese state-owned enterprises as a critical issue. Using a similar approach, Gallagher et al. [25] examined coal-fired power plants in India, Indonesia, Bangladesh, and Vietnam, concluding that China’s BRI has the potential to become the largest means for the diffusion of cleaner energy technologies throughout the developing world, contingent on recipient countries exercising more stringent environmental regulations and/or changes in Chinese overseas investment policies. Gu [39] studies Chinese investment in coal-fired power plants in Indonesia and suggests that coal-fired energy infrastructures lead to unequal distribution of social, environmental, and health impacts. In contrast, the findings of Lin and Bega [18] reveal that the BRI coal power projects generate opportunities for the BRI participants, as coal energy can render electricity generation more efficient in these countries, mostly from the developing world. Similarly, Li et al. [40] analyzed the pollution intensity of Chinese coal-fired power plants relative to those held by non-Chinese entities and showed that the majority of Chinese greenfield investment in coal used supercritical technologies, and this percentage is higher than that of non-Chinese coal plants.

Renewable energy infrastructure is highly interrelated to environmental sustainability and the accomplishment of sustainable development goals. For example, Song et al. [32] found that energy infrastructure investment, especially public-private partnerships (PPP), lowers fossil fuel generation, helping reduce carbon emissions and support sustainable development goals. However, as Bega and Lin [26] pointed out, the BRI renewables investments were rarely researched, while some scholars studied both coal and renewable energy cooperation. Using the Analytic Hierarchy Process (AHP), Bega and Lin [26] overviewed all types of power cooperation in coal and renewable energy projects. They found that BRI power cooperation in coal and renewable energy projects is a real chance to offer cleaner and more efficient electricity generation, especially in developing countries. Tao et al. [17] estimated the effects of CO₂ emissions from fossil fuel and renewable energy power plants in 15 representative countries across the BRI regions. They found that increased fossil fuel generation capacity led to higher carbon emissions, with variations across different scenarios, underscoring the importance and urgency of reducing fossil fuel dominance and expanding renewable energy within the BRI framework. Zhang et al. [41] implied that the reallocation of inward FDI in energy infrastructure neutralizes the negative consequences, improves energy efficiency, and consequently lowers emissions.

2.3. Methodology Literature

Various methods are employed in BRI studies, including qualitative approaches (e.g., interviews and ground analysis), quantitative methods (e.g., econometric techniques and CGE models), and mixed methods (e.g., AHP). Gu [39] was the only study using pure grounded analysis, offering a bottom-up perspective on socio-environmental conflicts surrounding coal-fired power plants in Indonesia. Thus, this study sets the foundation for future research summarizing mixed methods, econometric approaches, and CGE models.

To fully leverage qualitative and quantitative methods, Bega and Lin [26] employed the AHP, allowing for quantitative analysis of expert interviews. Some studies have used statistical descriptions to assess the evolution, challenges, and risks of energy projects [19,20,26]. Tritto [22] used a mixed-method design that prioritizes qualitative data with quantitative data, validating interview insights.

To evaluate the overall impact of BRI on energy and the environment, panel data Difference-in-Differences (DID) methods are commonly applied [19,42]. Most researchers have chosen econometric analysis for its simplicity and flexibility [43].

Each method has its attributes; however, GTAP-E-Power, an extended global CGE model, is employed to serve the purpose of this study. CGE models are economic models that produce computer-based simulations of market equilibria under different climate scenarios using a system of equations that describe the whole economy and their sectoral interactions [44]. On the global economy level, the GTAP model [45], a multiregion, multisector, computable general equilibrium model with perfect competition and constant returns to scale, and its extended models GTAP-E, GTAP-Power, and GTAP-E-Power are broadly used.

Based on the firm-level Chinese investment data and GTAP database. Burniaux and Truong [46] incorporated energy substitution structures and CO₂ emissions into the GTAP model to simulate energy-economic environmental policies, creating the GTAP-E model as a revised energy–environmental version of the GTAP model [47]. Peters [27] disaggregated the electricity sector and expanded the GTAP-Power database. Later, Peters [28] further split the electricity sector in GTAP-E into 12 sub-sectors and developed the electricity-detailed economy-wide model GTAP-E-Power, which is highly appropriate for studying the impact of climate change and energy policies [48]. The GTAP-E-Power model has been used widely for policy and energy economics assessments; for example, see [49,50,51,52,53].

3. Methodology, Database, and Procedure

This study uses the GTAP-E-Power model, as illustrated in Section 2.3. The framework of this methodology, database, and execution process is shown in Figure 1.

3.1. The GTAP-E-Power Model

The GTAP-E-Power model, as an extended GTAP model, is built on the assumptions of the standard GTAP model, where the theoretical structure is based on optimizing behavior by agents such as firms and households. Households maximize utility, firms minimize costs at zero profit conditions, and all agents are price takers [45]. Since this study focuses on power production shocks under the CPEC, the GTAP-E-Power production nest, which is most related to this study, is shown in Figure 2. Specifically, on the electricity production side, the GTAP-E-Power model subdivides the electricity sector (“ely”) into transmission, distribution, and generation, using Constant Elasticity of Substitution (CES) functions. Within the generation sub-sector, the Additive Constant Elasticity of Substitution (ACES) function is employed to minimize cost inefficiencies, representing a combination of average costs and various reliability costs associated with meeting complex demand requirements. This detailed nesting ensures that the total input equals the total output, maintaining unit consistency. The following equations are written in the standard GTAP model format.

Electricity is nested as in Equation (1) in [28]:

$$\underset{q_t^g}{\min }{U}^B={\left[\sum _{\forall t\in B}{\left(p_t^g·q_t^g\right)}^{\rho }\right]}^{{\displaystyle \frac{1}{\rho }}}$$

Subject to

$$Q^B=\sum _{\forall t\in B}{q}_t^g$$

(1)

where B is the set of baseload technologies, $q_t^g$ is the quantity of generation from technology t, $p_t^g$ is the price of generation for technology t, $Q^B$ is the total generation of baseload power, $U^B$ is the disutility of base load supply, and ρ is a parameter. The same formulation is used for peak load technologies.

Electricity production, as one of the intermediates, is nested into the production of different sectors, following the CES production nest and as outlined below:

A typical CES nested production can be written as follows in Equations (2)–(4):

$$\min C={\sum }_{i=1}^n{P}_{ir}Q{F}_{ir}$$

(2)

Subject to:

$${Q{VA}_{jr}=A\left(\sum _{i=1}^n{a}_i\left({{\lambda }_{i}QF}_{ir}^{\rho }\right)\right)}^{{\displaystyle \frac{1}{\rho }}}={A\left(\sum _{i=1}^n{a}_i\left({{\lambda }_{i}QF}_{ir}^{{\displaystyle \frac{{\sigma }_j-1}{{\sigma }_j}}}\right)\right)}^{{\displaystyle \frac{{\sigma }_j}{{\sigma }_j-1}}}$$

(3)

where C is the total cost, $P_{ir}$, $Q{F}_{ir}$, and ${QVA}_{jr}$ are, respectively, input price i, input demand i, and value-added of production sector j in region r. ${\alpha }_i$ is the so-called primal share parameters, and $\mathsf{\rho }$ is the CES exponent, ${\sigma }_j$ is the elasticity of substitution of different inputs in sector j, ${\sigma }_j=1-{\displaystyle \frac{1}{\rho }}$. A is a neutral technology shifter, ${\lambda }_i$ parameters are input-specific technology shifters (or biased technology shifters). The first-order conditions of the CES lead to the following set of equations in levels.

$${QF}_{ijr}={QVA}_{jr}{\alpha }_i{\left({\lambda }_{i}A\right)}^{{\sigma }_j-1}{\left({\displaystyle \frac{P_{ijr}}{{PV}_{jr}}}\right)}^{-{\sigma }_j}$$

(4)

where ${{\alpha }_i={a_i}^{{\sigma }_j}. PV}_{jr}$ is the weighted average price of production factors in the value-added nest. The key variable for this study is the exogenous variable A, where any shock to A will induce changes in production within the economic system. Specifically, an increase in electricity productivity, driven by expanded generation capacity, will be introduced as a shock to A and increase the output of various power subsectors and the industrial output through different CES production nest layers and a Leontief technology nest at the top [45,54], as shown in Figure 2.

On the electricity demand side, the GTAP-E-Power model segments electricity demand into firm, household, and government demand. A firm’s demand for electricity reflects general market needs and is detailed at the technological level, which is influenced not only by price and supply but also by the type of generation technology used. Household demand is influenced by income and policy, while public policy objectives and budget constraints shape government demand. Market clearing of electricity and other sectors determines the equilibrium prices and quantities in multiple markets, reaching a general equilibrium for the economy.

The GTAP-E-Power model, built upon the GTAP-E foundation, includes the treatment of detailed CO₂ emissions data (as shown in Equation (5)) and various advanced electricity generation technologies, making it highly suitable for studying the impacts of climate change and energy policies [48].

The CO₂ emissions equations are set in Equation (5):

$$C{O}_{2_{s,f,d,j,r}}={{\sigma }_{s,f,d,j,r}\times V}_{s,f,d,j,r}$$

(5)

$C{O}_{2_{s,f,d,j,r}}$ refers to ${CO}_2$ emissions released by d through industry j in region r, which combusts fossil fuel f from source s. Similarly, ${\sigma }_{s,f,d,j,r}$ refers to CO₂ emission factor, while $V_{s,f,d,j,r}$ refers to the value used of fossil fuels, respectively, where s = sources (import or domestic), f = fossil fuels (coal, oil, gas, p_c, and gassupply), j = industries (Nong et al. [55]), d = emission entity (firm, government or household) in region r. Total CO₂ emissions by source, fossil fuel, industry, region, or entity can be calculated by summing over the respective variables.

The GTAP-E-Power model offers two key advantages over other quantitative methods. First, it allows for investigating direct and indirect effects by considering all economic interactions, particularly when accounting for the input flows of different industries, which is crucial for understanding indirect economic impacts [56]. In this regard, the GTAP-E-Power model proves highly effective for analyzing the impacts of the power sector on industrial development, economic growth, and changes in CO₂ emissions throughout the process. Contrastingly, partial equilibrium models focus solely on specific sectors, and econometric analysis fails to capture industry interactions [57]. Second, the GTAP-E-Power model provides detailed insights into the electricity generation sector, covering a wide range of technologies, including several renewable energy sources. Additionally, it distinguishes between base load and peak load electricity generation and demand, reflecting the variations in electricity needs across different times of the day, seasons, and weeks [27]. Using the GTAP-E-Power model, it would be possible to analyze the effect of the change in the structure in the power sector, considering the effect of energy substitution [50].

3.2. Database

The GTAP-Power database version 11 is applied in the GTAP-E-Power model, covering 141 countries, 19 regions, and 76 sectors, with 2017 as the reference year. It includes national input-output (I-O) tables, trade, macroeconomic data, and, most importantly, detailed information on electricity generation and transmission sectors.

3.3. Assumptions and Procedure

First, this study aggregates the database into 24 sectors and 14 regions (see Appendix A,

Table A7. Aggregated sectors.

Sector Aggregation	Sectors Covered in GTAP
Agriculture	Paddy rice, wheat, cereal grains nec, vegetables, fruit, nuts, oil seeds, sugar cane, sugar beet, plant-based fibers, crops nec, bovine cattle, sheep and goats, horses, animal products nec, raw milk, wool, silkworm cocoons, forestry, fishing
Coal	Coal
Oil	Oil
Gas	Gas
Mining	Minerals nec
LightManu	Bovine meat products, meat products nec, vegetable oils and fats, dairy products, processed rice, sugar, food products nec, beverages and tobacco products, textiles, wearing apparel, leather products, wood products, paper products, publishing
P_c	Petroleum, coal products
HeavyManu	Chemical products, basic pharmaceutical products, rubber and plastic products, mineral products nec, ferrous metals, metals nec, metal products
AdvancedManu	Computer, electronic and optical products, electrical equipment, machinery and equipment nec, motor vehicles and parts, transport equipment nec, manufactures nec
TnD	Electricity: transmission and distribution
NuclearBL	Nuclear base load
CoalBL	Coal base load
GasBL	Gas base load
WindBL	Wind base load
HydroBL	Hydro base load
OilBL	Oil base load
OtherBL	Other base load
GasP	Gas peak load
HydroP	Hydro peak load
OilP	Oil peak load
SolarP	Solar peak load
Gassupply	Gas manufacture, distribution
Services	Water, construction, trade, accommodation, food and service activities, communication, financial services nec, insurance, real estate activities, business services nec, recreational and other services, public administration and defense, education, human health and social work activities, dwellings
Transport	Transport nec, water transport, air transport, warehousing and support activities

and

Table A8. Aggregated country and region.

Country/Region Aggregation	Countries/Regions Covered in GTAP
China	China
Pakistan	Pakistan
Southeast Asia	China, Hong Kong SAR, Cambodia, Indonesia, Lao People’s Democratic Republic, Malaysia, Philippines, Singapore, Thailand, Viet Nam, Rest of Southeast Asia
East Asia	Japan, Republic of Korea, Mongolia, Taiwan Province of China, Rest of East Asia, Brunei Darussalam
South Asia	Afghanistan, Bangladesh, India, Nepal, Pakistan, Sri Lanka, Rest of South Asia
Central Asia	Australia, New Zealand, Rest of Oceania
Middle East	Bahrain, Iran (Islamic Republic of), Iraq, Israel, Jordan, Kuwait, Lebanon, Oman, Palestine, Qatar, Saudi Arabia, Syrian Arab Republic, Türkiye, United Arab Emirates, Rest of Western Asia
Oceania	Canada, United States of America, Mexico, Rest of North America
North America	Argentina, Bolivia (Plurinational State of), Brazil, Chile, Colombia, Ecuador, Paraguay, Peru, Uruguay, Venezuela (Bolivarian Republic), Rest of South America, Costa Rica, Guatemala, Honduras, Nicaragua, Panama, El Salvador, Rest of Central America, Dominican Republic, Haiti, Jamaica, Puerto Rico, Trinidad and Tobago, Caribbean
Latin America	Austria, Belgium, Bulgaria, Croatia, Cyprus, Czechia, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden
EU_27	United Kingdom of Great Britain, Switzerland, Norway, Rest of EFTA, Albania, Serbia, Belarus, Russian Federation, Ukraine, Rest of Eastern Europe, Rest of Europe
Other European Countries	Kazakhstan, Kyrgyzstan, Tajikistan, Uzbekistan
Africa	Algeria, Egypt, Morocco, Tunisia, Rest of North Africa, Benin, Burkina Faso, Cameroon, Côte d’Ivoire, Ghana, Guinea, Mali, Niger, Nigeria, Senegal, Togo, Rest of Western Africa, Central African Republic, Chad, Congo, Democratic Republic of the Congo, Equatorial Guinea, Gabon, South-Central Africa, Comoros, Ethiopia, Kenya, Madagascar, Malawi, Mauritius, Mozambique, Rwanda, Sudan, United Republic of Tanzania, Uganda, Zambia, Zimbabwe, Rest of Eastern Africa, Botswana, Eswatini, Namibia, South Africa, Rest of Southern African Customs Union
Rest of the World	Countries and territories in the database not listed above

). To conduct a detailed investigation into energy cooperation, 76 industries are aggregated into 24 sectors, leaving electricity a disaggregated sector incorporating 12 sectors. The other sectors are aggregated into 12 sectors: agriculture, coal, oil, gas, p_c, gas supply, mining, light manufacturing, heavy manufacturing, advanced manufacturing, transportation, and service. Such aggregation aims to keep focus on energy and industrial change and reduce simulation load. This study aggregates 160 countries and regions into 14 regions, with Asia incorporating East Asia, South Asia, Central Asia, and Southeast Asia, to allow us to investigate CPEC’s spatial spillover effects in detail.

Second, this study conducts a comparative static simulation, as updating the I-O tables and the detailed electricity data is too complex, and a less precise update will lead to calculation errors. Since some coal, wind, and solar power projects were already completed by 2020 as early harvest or short-term projects, we set the base year as 2020. This setup allows us to focus on medium- and long-term projects rather than short-term ones and avoids disaggregating five-year plans. Thus, we updated the database from 2017 to 2020 to start the simulations. We use CEPII’s macroeconomic projections [58] and apply methods from Diffenbaugh et al. [59] and Nong et al. [55,60] to introduce shocks to key macroeconomic variables as detailed in Appendix A

Table A5. Renewable power plants under CPEC.

No	Project Name	Installed Capacity (MW)	Actual/Expected COD	Power	Technology	Estimated Cost (USD Million)
1	Quaid-e-Azam Solar Park	400	2016	Solar	PV Solar	520
2	Hydro China Dawood Wind Farm	50	2017	Wind	Wind Turbine	112.65
3	UEP Wind Farm	100	2017	Wind	Wind Turbine	250
4	Sachal Wind Farm	50	2017	Wind	Wind Turbine	134
5	Three Gorges Second and Third Wind Power Project	100	2018	Wind	Wind Turbine	150
6	Karot Hydropower Project	720	2022	Hydel	Large Hydro	1720
7	Suki Kinari Hydropower Project	870	2022	Hydel	Large Hydro	2000
8	Kohala Hydropower Project	1124	2029	Hydel	Large Hydro	2400
9	Azad Pattan Hydropower Project	700.7	2026	Hydel	Large Hydro	1600
10	Cacho Wind Power Project	50	2026	Wind	Wind Turbine	125
11	Western Energy (Pvt.) Ltd. Wind Power Project	50	2026	Wind	Wind Turbine	88.46
12	Quaid-e-Azam Solar Park	600	2024	Solar	PV Solar	781
13	Karachi 2	1145	2021	Nuclear	Hualong one unit	10,000
14	Karachi 3	1100	2022	Nuclear	Hualong one unit
15	Chashma 5	1100	by 2030	Nuclear	Hualong one unit	3500

Source: See Section 2.1.

. Figure 1 illustrates the study process.

Third, to conduct a comparative static simulation and calculate the new equilibrium, shocks will be applied to$A$in Equations (3) and (4). This study assumes the linear change in A is set as follows:

$$a_{ir}={aoall}_{ir}$$

(6)

${aoall}_{ir}$ represents output augmenting technical change in sector i of r. Specifically, to achieve the objective of this study, shocks are only applied to electricity sub-sectors comprised in CPEC: i = CoalBL, HydroBL, WindBL, SolarPL, NuclearBL, and TnD, r = Pakistan, ${aoall}_{ir}$ is shocked in different scenarios in Equation (6). Output change for each sector is derived from the whole production nest in Figure 2, and the price is determined by market clear, where electricity demand is assumed to be unshocked. CO₂ emissions are calculated from Equation (5).

Fourth, shocks are introduced step-by-step into the model to fully investigate the interactions of different power generation sources. Detailed scenario design and numerical shock calculation are explained in Section 4.

4. Investment Data and Simulation Scenarios

4.1. Data

To calculate the shocks, information for individual power plant projects under CPEC is collected from various sources (as illustrated in Appendix A

Table A3. Information sources of CPEC energy projects.

No	Sources	Note
1	CPEC energy website [73]	public
2	The Long-Term Plan for China-Pakistan Economic Corridor (2017–2030) [74]	public
3	The Overview of Pakistan’s Power Sector and Its Future Outlook [64]	public
4	The Private Participation in Infrastructure (PPI) Project Database released by the World Bank [75]	public
5	The Global Power Plant Database by the World Resources Institute [76]	public
6	The China Global Investment Tracker (CGIT)by the Boston University Global Development Policy Center (GDP Center) [77]	public

), including the official CPEC website and the Long-Term Plan for China-Pakistan Economic Corridor (2017–2030) signed in December 2017. Additional sources include the Overview of Pakistan’s Power Sector and Its Future Outlook released in September 2022, and three micro-level databases: the Private Participation in Infrastructure (PPI) Project Database released by the World Bank, the Global Power Plant Database by the World Resources Institute, and the China Global Investment Tracker (CGIT) by the American Enterprise Institute, as well as China’s Global Power (CGP) Database by the Boston University Global Development Policy Center (GDP Center). We meticulously matched and verified information from these sources, resulting in a comprehensive list of 28 projects under CPEC and three major nuclear power plants, as detailed in Appendix A,

Table A4. Coal power plants under CPEC.

No	Project Name	Installed Capacity (MW)	Actual/Expected COD	Power	Technology	Estimated Cost (USD Million)
1	Sahiwal Coal-fired Power Plant	1320	2017	Coal (Imported)	Super Critical	1912.2
2	Coal-fired Power Plant at Port Qasim Karachi	1320	2018	Coal (Imported)	Super Critical	1912.2
3	China Hub Coal Power Project	1320	2019	Coal (Imported)	Super Critical	1912.2
4	Engro Thar Coal Power Project	660	2019	Coal (Local)	Sub Critical	995.4
5	HUBCO Thar Coal Power Project	330	2022	Thar Coal	Sub Critical	497.7
6	SSRL Thar Coal Block-I 7.8 mtpa & Power Plant	1320	2023	Coal (Local)	Sub-Critical	1912.12
7	HUBCO ThalNova Thar Coal Power Project	330	2023	Thar Coal	Sub Critical	497.7
8	Coal-Fired Power Project at Gwadar	300	2023	Coal (Imported)	Super Critical	542.32
9	Thar Mine Mouth Oracle Power Plant & surface mine	1320	2026	Thar Coal	Sub Critical	1912.2

Source: See Section 2.1.

, Table A5 and

Table A6. Transmission line projects and potential electric power projects under CPEC.

No	Name	Installed Capacity (MW)	Actual/Expected COD	Estimated Cost (Million $)	Power
1	Matiari–Lahore HVDC Transmission Line	4000	2021	170	Transmission Line
Potential projects (Expected to be included in CPEC in the future)
1	Port Qasim-Faisalabad HVDC Transmission Line	4000	-	-	Transmission Line
2	Mahl HPP	640	-	-	Hydel
3	Taunsa HPP	135	-	-	Hydel
4	Toren More HPP	350	-	-	Hydel
5	Jameshill More HPP	260	-	-	Hydel
6	Phander HPP	80	-	-	Hydel

Notes: 1. Large hydro means hydropower projects whose generation is greater than 50 MW, as indicated in the China Global Investment Tracker database. 2. “-” means there is no specific information. Source: See Section 2.1.

.

By 2030, ongoing and planned projects will boost Pakistan’s generation capacity by 16,379.7 MW, with 8220 MW from coal (a cost of approximately USD 12.1 billion) and 8159.7 MW from renewables (a cost of roughly USD 23.4 billion) (Figure 3a,c), representing 50.2% and 49.8% of the total capacity increase, respectively. Figure 3b,d illustrate the province-level information. Potential future projects, including another transmission line and six hydro power plants, could also add 1465 MW to the generation capacity under CPEC (Table A6 in Appendix A). However, due to the lack of specific details, potential projects are not included in Figure 3.

4.2. Scenario and Shocks

This study sets seven scenarios with step-by-step shocks based on different power generation sources, considering both medium- and long-term perspectives, as illustrated in

Table 1. Scenario design.

Project Types	Medium-Term Projects	Medium- and Long-Term Projects
Coal	S1	S2
	• Coal power plants	• Coal power plants
REN	S3	S4
	• Hydel, solar and wind	• Hydel, solar and wind
	• Karachi 2 and 3	• Karachi 2 and 3
		• Chashma 5
Coal + REN + TnD	S5	S6
	• Coal power plants	• Coal power plants
	• Hydel, solar and wind	• Hydel, solar and wind
	• Karachi 2 and 3	• Karachi 2 and 3
	• One transmission line	• One transmission line
Coal + REN + TnD + Expected		S7
		• Coal power plant
		• Hydel, solar and wind
		• Karachi 2 and 3
		• Chashma 5
		• Expected projects
		• Two transmission lines

Notes: “M” and “L” refer to medium-term and long-term, respectively. “REN” stands for renewable energy and nuclear, “TnD” represents transmission lines, and “expected” indicates potential projects anticipated to be included in CPEC by 2030.

. Given the heterogeneity of power generation sources, from a medium-term perspective, S1 focuses on coal plants, which are of significant environmental concern, while S3 examines only renewable and nuclear power plants, and S5 evaluates the entire range of CPEC energy projects. S2, S4, and S6 take a long-term view, assessing coal, renewable, and all of the power plants, respectively. S7 includes potential projects expected to be part of CPEC by 2030.

Numerical shocks to aoall in GTAP-E-Power for each electricity sub-sector are listed in

Table 2. Scenario shocks.

	Sector	CoalBL	HydroBL	WindBL	SolarP	NuclearBL	TnD
S1	Yi,2020 (MW)	5088	10,021	1235.7	837.91	1467	27,800
	Xi,2025 (MW)	2280	-	-	-	-	-
S2	aoalli,2025	44.81	-	-	-	-	-
	Xi,2030 (MW)	3600	-	-	-	-	-
S3	aoalli,2030	70.75	-	-	-	-	-
	Xi,2025 (MW)	-	1590	0	600	2245	-
S4	aoalli,2025	-	15.87	0	71.61	153.03	-
	Xi,2030 (MW)	-	3414.7	100	600	3345	-
S5	aoalli,2030	-	34.08	8.09	71.61	228.02	-
	Xi,2025 (MW)	2280	1590	0	600	2245	4000
S6	aoalli,2025	44.81	15.87	0	71.61	153.03	14.39
	Xi,2030 (MW)	3600	3414.7	100	600	3345	4000
S7	aoalli,2030	70.75	34.08	8.09	71.61	228.02	14.39
	Xi,2030 (MW)	3600	4879.7	100	600	3345	8000
	aoalli,2030	70.75	48.69	8.09	71.61	228.02	28.78

Source: The generation capacity of CoalBL in 2020 is calculated based on Pakistan’s coal power generation capacity from the Global Power Plant Database [61], plus the increased capacity from coal projects completed by 2020 under CPEC. The generation capacity for HydroBL, WindBL, and SolarP in 2020 is sourced from the International Renewable Energy Agency (IRENA) [62], as Pakistan’s National Electric Power Regulatory Authority (NEPRA) aggregates hydro, wind, and solar as renewable energy, providing aggregated data. The generation capacity for NuclearBL and TnD in 2020 is taken from the NEPRA State of Industry Report 2021 [63].

based on project-level data shown in Appendix A, Table A4, Table A5 and Table A6, to obtain the new equilibrium status for Pakistan. The formula for calculating shocks to each electricity sub-sector generation productivity is as follows:

$${aoall}_{it}={\displaystyle \frac{X_{it}}{Y_{i,2000}}}\times 100$$

(7)

where ${aoall}_{it}$ stands for the technological augment percentage change in electricity sub-sector i in year t; $X_{it}$ represents the accumulative incremental power generation capacity of type i power generation source from 2020 to year t and $Y_{i,2000}$ represents the existing power generation capacity in 2020. i = CoalBL, HydroBL, WindBL, SolarPL, NuclearBL and TnD, t = 2025 or 2030. For example, aoall_CoalBL,2025 stands for the coal baseload sector technological augment change in 2025 induced by coal power plants’ accumulative generation capacity addition from 2021 to 2025. X_Coal,2025 means cumulative generation capacity addition by coal power plants from 2021 to 2025 (in a medium-term perspective). The growth rates in the electricity output capacity for each electricity generation sector are assumed to be fully absorbed by the output augmenting technical change variables in the model, similar to Nong et al. [55].

5. Results

5.1. Impacts on Energy Sectors

We found four major outcomes from the seven scenario results (Figure 4, Appendix A

Table A10. Change in energy output and price.

	Change in Output (%)							Value Change in Output (USD Millions)							Change in Price (%)
	S1	S2	S3	S4	S5	S6	S7	S1	S2	S3	S4	S5	S6	S7	S1	S2	S3	S4	S5	S6	S7
CoalBL	59.27	94.04	−17.82	−27.45	35.84	49.01	48.19	985	1563	−296	−456	596	815	801	−31.06	−41.70	−0.06	−0.19	−31.04	−41.72	−41.68
GasBL	−4.48	−7.32	−17.78	−27.43	−18.53	−28.99	−29.32	−188	−307	−746	−1151	−778	−1217	−1230	0.00	−0.03	−0.10	−0.20	−0.10	−0.17	−0.17
OilBL	−4.51	−7.37	−17.88	−27.59	−18.68	−29.18	−29.53	−84	−137	−332	−513	−347	−542	−549	0.01	0.01	−0.01	−0.04	0.03	0.02	0.04
NuclearBL	−4.58	−7.44	197.21	276.31	194.03	267.71	265.67	−69	−113	2992	4191	2943	4061	4030	0.07	0.07	−61.18	−70.47	−61.12	−70.41	−70.39
HydroBL	−4.65	−7.54	0.43	8.51	−0.83	5.76	21.31	−169	−275	16	310	−30	210	776	0.12	0.14	−13.51	−25.33	−13.29	−25.09	−32.43
WindBL	−4.63	−7.51	−18.15	−19.61	−19.15	−21.62	−22.14	−16	−25	−61	−66	−64	−72	−74	0.11	0.12	0.23	−7.28	0.45	−7.01	−6.87
OtherBL	−4.62	−7.51	−18.16	−27.90	−19.15	−29.70	−30.18	−12	−20	−48	−74	−51	−79	−80	0.10	0.12	0.23	0.27	0.45	0.55	0.70
OilP	1.67	2.47	4.48	7.19	8.63	12.24	15.63	56	83	150	241	289	410	523	0.01	0.01	−0.01	−0.04	0.03	0.02	0.05
SolarP	1.61	2.40	34.73	38.15	39.89	44.43	48.66	2	3	42	46	48	54	59	0.13	0.16	−41.90	−41.88	−41.74	−41.67	−41.56
TnD	1.66	2.47	5.66	8.39	9.94	13.59	17.09	108	160	366	543	644	880	1106	0.11	0.13	0.26	0.30	−12.18	−12.09	−21.86
Coal	6.29	8.65	−12.93	−19.58	−6.12	−10.97	−11.85	14	19	−28	−43	−13	−24	−26	0.73	0.97	−0.67	−1.32	0.52	0.30	0.54
Oil	−0.25	−0.34	−0.79	−1.08	−1.20	−1.59	−1.87	−4	−5	−12	−16	−18	−24	−28	−0.03	−0.07	−0.22	−0.36	−0.15	−0.27	−0.21
Gas	−0.31	−0.43	−0.95	−1.32	−1.51	−2.02	−2.42	−2	−3	−6	−9	−10	−14	−17	−0.09	−0.15	−0.37	−0.58	−0.44	−0.67	−0.72
P_c	−0.37	−0.62	−1.69	−2.52	−1.80	−2.66	−2.70	−27	−44	−121	−180	−129	−190	−193	0.01	0.00	−0.02	−0.05	0.03	0.01	0.05
Gassupply	−3.10	−4.96	−11.87	−18.11	−13.53	−20.43	−21.51	−83	−132	−316	−482	−360	−544	−573	0.19	0.24	0.50	0.63	0.91	1.14	1.42

Source: Calculated based on GTAP-E-Power model simulation results.

). First, medium- and long-term productivity shocks to electricity sub-sectors generally lead to increased output and lower prices, except for wind base load, which shows only a slight productivity increase. Second, larger shocks result in more significant output gains and price reductions, with more pronounced long-term impacts due to accumulated capacity increases. Third, in all scenarios, the increase in electricity output requires expanded transmission and distribution infrastructure, with oil and solar peak loads helping to balance demand. Fourth, step-by-step shock comparisons reveal a consistent substitution effect between different sub-types of conventional and renewable energy sources, impacting electricity and non-electricity sectors like coal, oil, and gas.

Specifically, in S1 and S2, medium- and long-term shocks increasing coal base load productivity by 44.81% and 70.75% lead to nearly 60% (or USD 985 million) and over 90% (or USD 1563 million) growth in coal electricity output, with price reductions of over 30% and 40%, respectively. Consequently, coal output rises by about 6% (or USD 14 million) and 9% (or USD 19 million), while other electricity sub-sectors, including renewables, experience declines of 4–5% and 7–8%, respectively, indicating a crowding-out effect. These findings confirm that larger shocks to one type of energy will result in more substantial output gains, price drops, and stronger substitution effects.

In S3, where medium-term productivity shocks are applied to nuclear (153.03%), solar (71.61%), and hydro (15.87%) base loads, their outputs increase by nearly 200% (or USD 2992 million), 35% (or USD 42 million), and less than 1%, respectively, while prices drop by around 60%, 40%, and 13%. Unshocked sub-sectors see 17–19% output declines, with gas baseload output value dropping the most by USD 746 million, reflecting substitution effects. Interestingly, hydro’s minimal increase is attributed to significant shifts towards nuclear and solar power. In the long term, in S4, larger productivity shocks to nuclear (228.02%), solar (71.61%), and hydro (34.08%) will boost outputs by 276% (or USD 4191 million), 38% (or USD 46 million), and 9%, with even sharper price reductions. Unshocked sub-sectors see 20–28% output declines, with coal, gas, and oil base loads decreasing by about 28%, with gas base load output dropping the most by USD 1151 million. As a result, in both S3 and S4, coal and gas supply output decreased by nearly 20%, with gas supply value dropping the most by USD 316 and USD 482 million, respectively, while oil supply saw a slight decline due to increased oil peak load.

Scenarios S5 and S6 provide full pictures by applying medium- and long-term changes in all electricity sub-sectors, respectively. Due to substitution effects, determined by the relative shock magnitudes shown in S1, S2, S3, and S4, output and price changes are moderated by the interactions among the shocked sub-sectors. As a result, the output and price changes of coal base load are smaller than those in S1 and S2, while the output and price changes of nuclear base load and hydro base load are smaller than those in S3 and S4. The output of solar peak will increase due to both the productivity shock and the need to meet peak demand. In contrast, the output of other unshocked electricity sub-sectors will significantly decrease more than when only coal or renewable shocks are introduced, with gas and oil base loads decreasing the most by approximately 20% and 30%, respectively. Consequently, the output of gas supply and oil will decrease.

If the potential projects expected to be included in CPEC are fully implemented, as captured by scenario S7, hydro baseload electricity output will increase by over 21%, and electricity transmission and distribution will rise by 17%, with prices decreasing by more than 30% and 20%, respectively. Additionally, the output of coal, nuclear, and wind baseload electricity will experience a slight decrease compared to scenario S6, indicating substitution by hydro baseload electricity. Other changes will follow the same directions as in S6, but with larger magnitudes.

Moreover, CPEC energy investment will significantly change Pakistan’s electricity output structure by driving substantial substitution effects between power sources (Figure 5a). By 2030, zero-emissions (clean) power is expected to account for 47.9% of the total power generation under S6 and around 50% under S7. Figure 5b shows more details of various power source mix changes. Nuclear power will experience the most notable increase, especially in scenarios S3 to S6, where it replaces coal, gas, and oil baseloads while slightly reducing wind, hydro, and other baseloads. In scenarios S5, S6, and S7, the shares of nuclear and coal baseloads rise, while those of gas, oil, hydro, and wind decline, leading to a cleaner electricity generation mix.

5.2. Effects on CO₂ Emissions

The environmental impact is a critical aspect of the sustainability of the BRI and CPEC energy cooperation. If shocks are applied only to coal power plants in scenarios S1 and S2, carbon emissions from coal will increase by nearly 2.5% and 3.4%, or 1.22 Mts and 1.66 Mts, respectively (Figure 6). These results align with those of Ali et al. [23], who found that net CO₂ emissions from all coal-fueled power projects under CPEC in Pakistan are likely to increase. However, using the GTAP-E-Power model with a general equilibrium analysis, our results indicate that net carbon emissions in Pakistan will decrease slightly by about 1 Mts and 1.9 Mts, respectively, primarily due to gas-to-coal substitution in electricity generation, resulting in a CO₂ reduction of 1.85 Mts and 2.96 Mts from the gas and gas supply, respectively. Notably, this conclusion may contradict the advocacy for phasing out coal plants. However, according to Pakistan’s power sector’s future outlook and CPEC energy infrastructure’s latest information on the CPEC website (see Appendix A Table A3), coal will remain an important part of energy generation, at least until 2030, within this study’s timeframe.

In scenarios S3–S7, CO₂ emissions from all fuel fossils will decrease. Long-term scenarios show greater carbon reduction than medium-term ones. CO₂ emissions from the gas supply sector will see the largest reductions, ranging from 11.73% to 21.30% or 5.59 Mts to 10.15 Mts, followed by reductions in carbon emissions from gas and coal. Specifically, in scenarios S3 and S4, the gas supply sector will have significant CO₂ emissions reductions of 11.73% and 17.89%, or 5.59 Mts and 8.52 Mts, respectively. Carbon emissions from coal will decrease by 5.73% and 8.73%, or 2.81 Mts and 4.28 Mts, respectively. In scenarios S5–S6, CO₂ emissions from the gas supply sector will decrease by 13.4% and 20.21% or 6.38 Mts and 9.63 Mts, respectively. Carbon emissions from gas will decrease by 6.44% and 9.63%, or 1.77 Mts and 2.65 Mts, respectively. Scenario S7 demonstrates the most substantial decrease in CO₂ emissions, especially CO₂ emissions from the gas supply, which will be by 21.3% or 10.15 Mts, confirming that CPEC, as a flag project of BRI, has a significant margin for improvements regarding CO₂ emissions reduction [17]. The net CO₂ in Pakistan and globally is projected to decrease by a maximum of 18.61 Mts and 16.42 Mts, respectively, in S7.

Figure 7a,b offer detailed insights into CO₂ emissions across different production sectors before the shocks and in S7, respectively. The gas base load stands out as the largest source of CO₂ emissions among the energy sectors, followed by oil peak load. The comparison of CO₂ emissions between the pre-shock period and S7 reaffirms that the primary factor driving CO₂ emissions reductions is the substitution of gas base load with alternative power sources, leading to a reduction of nearly 10 Mts of CO₂ emissions from gas base load. Direct CO₂ emissions from energy sectors, including various power generation sub-sectors, account for only a small portion of Pakistan’s total CO₂ emissions. Concerns about the environmental impact of CPEC energy infrastructure, particularly coal-fueled power plants, are often overstated, especially when compared to the two largest sources of indirect CO₂ emissions: heavy manufacturing and transportation, which are far greater than the direct emissions from electricity generation sectors.

5.3. Effects on Non-Energy Sectors

The construction and completion of CPEC energy projects will greatly impact Pakistan’s non-energy sectors by addressing electricity shortages and providing cost-competitive power, which is crucial for industrialization (Figure 8). In each scenario, energy-intensive sectors, particularly advanced and heavy manufacturing, will experience the most significant output expansion and price reductions, enhancing their competitiveness. Though with slight price increases, transport, services, and mining sectors will also benefit. The greater the electricity generation productivity shocks, the more output expansion and price decrease, indicating that improvements in Pakistan’s energy outlook will provide more opportunities for domestic and foreign manufacturers. Advanced and heavy manufacturing will achieve the highest output growth of 2.2% and 1.3%, respectively, in S7, while less energy-intensive sectors like agriculture and light manufacturing will suffer from the crowding-out effects. Overall, the impact of CPEC energy investments on industrial upgrading is positive but modest, as both supportive industrial policies and energy supply are essential for significant progress. While the improved energy supply under CPEC is vital for industrial development, more deliberate and comprehensive industrial policies will be needed moving forward.

Moreover, as shown in Figure 7, the heavy manufacturing and transport sectors, the two largest sources of CO₂ emissions, together account for over 40% of total CO₂ emissions from production. Despite the output expansion of energy-intensive sectors in S7, CO₂ emissions slightly decreased in heavy manufacturing and slightly increased in transport, resulting in a net reduction in overall CO₂ emissions. These findings indicate that CPEC can help Pakistan promote its economy while reducing the environmental burden of this development.

5.4. Effects on Macroeconomic Indicators

As a result of CPEC projects, Pakistan’s macroeconomic conditions previously constrained by electricity shortages will improve (

Table 3. Changes in Pakistan’s macroeconomic conditions.

Indicator/Scenario	S1	S2	S3	S4	S5	S6	S7
Real rental-to-capital ratio	0.23	0.32	0.70	0.98	1.22	1.59	1.97
Real rental-to-labor ratio	0.15	0.22	0.49	0.72	0.77	1.04	1.24
Change in investment	0.26	0.36	0.76	1.04	1.36	1.75	2.17
Real GDP (%)	0.16	0.25	0.59	0.88	0.92	1.26	1.52
Real income (%)	0.26	0.35	0.73	0.97	1.27	1.63	2.01
Private expenditure (%)	0.26	0.35	0.73	0.96	1.27	1.62	2.00
Government expenditure (%)	0.28	0.37	0.79	1.05	1.37	1.77	2.17
EV (million USD)	678	983	2174	3101	3625	4810	5858
GDP price index (%)	0.09	0.09	0.15	0.12	0.31	0.34	0.44
Term of trade (%)	0.13	0.16	0.32	0.39	0.59	0.72	0.90
Real import (%)	0.09	0.04	−0.25	−0.56	0.06	−0.15	0.01
Real export (%)	−0.69	−0.84	−1.66	−1.96	−2.99	−3.65	−4.51

Source: Calculated based on simulation results from the GTAP-E-Power model.

). Increased electricity generation productivity boosts capital productivity and output in the sector, raising the return on capital. In a short-run closure where capital is fixed, this higher return enhances the ratio of return to capital, driving increased investment demand and leading to GDP growth. CPEC energy investments are expected to promote the real return on capital and labor by 0.23–1.97% and 0.15–1.24%, respectively, and increase real investment by 0.26–2.17%, leading to real GDP growth of 0.16–1.52% and a rise in real income by 0.26–2.01%, along with increases in both private and government expenditures. The impact is more significant in the long term, and broader expansions across all electricity generation types lead to greater effects than improvements in just one type, such as coal or renewable energy. Under these scenarios, Pakistan’s economic welfare (EV) is projected to increase by USD 678 million to USD 5858 million.

Pakistan’s terms of trade will also improve. However, real imports are expected to rise while exports may decline due to expanding domestic consumption and production demand. Notably, the country’s imports could decrease if energy imports significantly decline, particularly when conventional electricity generation sources are significantly substituted by renewable energy in scenarios like S3, S4, and S6.

6. Discussion and Policy Implications

This study offers several important and innovative insights. First, it confirms the findings of Lin and Bega [18]. They found that coal power projects can generate opportunities by making electricity generation more efficient in developing countries constrained by power outages, like Pakistan. Coal power plants provide a viable solution to meet energy demands while minimizing costs due to the ability to substitute among different conventional energy sources. This energy shift can enhance energy security and reduce dependency on more expensive or less stable energy sources. Coal electricity can also support industrial development and drive economic growth with net carbon emissions reduction for Pakistan. This carries significant policy implications. For Pakistan, CPEC energy projects would allow Pakistan to replace gas-based power with more affordable coal-based power, reducing energy costs and enhancing the competitiveness of energy-intensive sectors such as advanced and heavy manufacturing. In the long term, this transition to coal can eventually serve as a stepping stone for Pakistan to switch to cleaner energy sources. When transitioning to cleaner energy sources, coal power should not be phased out until renewable options like nuclear, wind, hydro, and solar can fully meet power demands. Prioritizing a shift from imported coal to locally sourced Thar coal can further support energy sustainability during this period. Coal power can be an attractive option for regions with limited or expensive access to renewable energy. Coal power can help quickly scale up the energy supply and support economic growth in the short to medium term. It also has the potential to reduce net carbon emissions, challenging the common belief that CPEC or BRI energy investments harm the environment due to heavy reliance on coal. However, coal power projects should not take precedence over renewable power plants.

Second, to achieve greener energy cooperation in the future, it is essential to prioritize renewable energy investments once the immediate power shortage is addressed. A dedicated plan is needed to avoid inter-sector power source substitution effects, especially when large shocks are applied to specific sectors. According to [64], the country still faces significant challenges in meeting the projected demand of 37,129 MW by 2030. To fulfill this demand, a generation capacity of 61,112 MW must be available, which includes addressing installed capacity shortages of 6494.09 MW for solar, 3669.3 MW for wind, and 8752.3 MW for hydro, despite the capacity increases under CPEC. To meet the 2030 goal, Pakistan should strategically balance the development of solar, wind, and hydropower plants to prevent substitution effects that could undermine the overall energy mix. Solar power, in particular, is crucial as a peak load source and complements other power sources. Prioritizing solar energy development could enhance the stability and sustainability of the energy grid. Adopting a well-rounded approach to renewable energy development, which considers substitution among different subsectors, is essential for Pakistan to build a more resilient and environmentally friendly future. This approach is also applicable to the energy development of other countries.

Third, some studies have expressed concern that the engagement in BRI energy development is a debt trap due to a lack of data disclosure, a ruse towards modern neo-colonization and resource extinction in Africa [65], and posits that the BRI has been a “one-way road” [66], meaning that it benefits China as a channel for Chinese enterprises to cope with overcapacity more than other members [67]. However, this study shows that the macroeconomic conditions in Pakistan will improve rather than deteriorate. For other developing countries like Pakistan, which lack sufficient capital, utilizing foreign assistance can be essential for achieving energy and economic development goals.

Fourth, this study broadens the research on carbon intensity (as shown in Appendix A

Table A11. Change in carbon intensity and carbon elasticity.

Sectors	Change in CO2 Emissions (%)	Change in Real Output in S7 (%)	Change in Carbon Emissions Intensity (%)	Carbon Elasticity
Real GDP	−8.4411	1.5185	−9.9596	−5.5588
CoalBL	−12.7118	48.1941	−60.9059	−0.2638
GasBL	−28.9401	−29.3195	0.3794	0.9871
OilBL	−29.4392	−29.5340	0.0948	0.9968
OtherBL	−29.3044	−30.1753	0.8709	0.9711
OilP	15.7931	15.6329	0.1602	1.0102
TnD	2.3661	17.0904	−14.7243	0.1384
Coal	−20.0000	−11.8491	−8.1509	1.6879
Oil	0.0000	−1.8731	1.8731	0.0000
Gas	0.0000	−2.4172	2.4172	0.0000
P_c	−2.6934	−2.6978	0.0044	0.9984
Gassupply	−20.8578	−21.5115	0.6537	0.9696
Agriculture	−9.8431	0.0270	−9.8701	−364.5586
Mining	−8.6346	0.8297	−9.4643	−10.4069
LightManu	−9.8179	−0.4776	−9.3403	20.5568
HeavyManu	−3.3467	1.2830	−4.6297	−2.6085
AdvancedManu	−6.5570	2.1641	−8.7211	−3.0299
Transport	1.3753	0.8271	0.5482	1.6628

Notes: Carbon emissions intensity has different formulas in different study backgrounds. For example, Chen and Li [78] entail the logarithmic transformation of unitary carbon emissions of Gross Domestic Product (GDP). Zhao et al. [79] used the logarithm of the ratio of regional carbon emissions divided by real industrial added value. Xiang et al. [80] used the value of carbon emissions to calculate the production value. In this study, carbon emissions intensity for each industry in the production use is the quantity of CO₂ emissions (Mts) per unit of real industry output, while the total carbon emissions intensity for each economy is to real GDP. Thus, the change in the carbon emissions intensity for each industry is the change in CO₂ emissions deducted from the change in the real output of each industry. This study also calculated the carbon elasticity, similar to Xiang et al., as the ratio of the percentage change in CO₂ emissions to the percentage change in real output. Source: Calculated based on GTAP-E-Power model simulation results.

) and aligns with findings that, on average, China’s OFDI results in a net reduction effect on the carbon intensity of BRI countries (Wang et al. [68]). Pakistan’s carbon emissions intensity will decrease by roughly 10%, and most industry sectors will decline, with CoalBL and gas extraction experiencing the largest and the smallest decline, respectively. Carbon emissions elasticity is also calculated, showing the total elasticity is −5.56.

Fifth, while this study focuses on CPEC’s energy investment as a case study, it offers valuable insights into the broader energy–economy–environment impacts of energy investments along the BRI. Understanding the distinct impacts of different energy sources—whether coal, nuclear, solar, wind, or hydro—enables policymakers to make informed decisions that balance economic growth with environmental sustainability. The findings from this study can serve as a blueprint for other BRI countries, helping to optimize their energy strategies and contribute to a greener, more resilient global energy landscape.

7. Conclusions

Based on project-level data, this study simulated seven potential CPEC energy scenarios for Pakistan, revealing several significant outcomes. Firstly, CPEC energy investments will offer competitive electricity rates and significantly optimize Pakistan’s energy structure by replacing high-cost fuels like natural gas and oil, substantially reducing CO₂ emissions. Changes in the output of non-electricity energy sectors, such as coal, oil and gas extraction, coal and petroleum products, and gas supply, will reflect shifts in the electricity generation structure. Secondly, the construction and completion of CPEC energy projects will significantly impact Pakistan’s non-energy sectors. The improved energy outlook will create more opportunities for energy-intensive industries, with advanced manufacturing expected to see the largest gains in output and export expansion. Thirdly, due to CPEC energy projects, Pakistan’s macroeconomic conditions will improve, with real GDP, economic welfare, real income, private and government expenditures, real investment, terms of trade, and capital returns all experiencing positive growth. These positive effects persist even when considering the shocks from only coal power generation capacity. This study contributes to and enriches the literature on CPEC and energy infrastructure investment under BRI.

There are several limitations in this study. Firstly, although the scenarios and shocks were designed based on the latest project-level data, the construction and development of CPEC energy projects are subject to change, introducing potential future policy uncertainty. This study does not consider Pakistan’s Nationally Determined Contribution (NDC) [69], including a commitment to ban coal imports by 2030. Since coal for electricity generation makes up only a small portion of the total [70], and much of the coal used after 2020 has been sourced from domestic reserves like Thar coal rather than imports, this shift reduces the significance of coal imports in the national energy mix, but these effects can be investigated in the future study. Secondly, this study employs a comparative static rather than a dynamic general equilibrium analysis using the GTAP-E-Power model. The GTAP-E-Power model, a global CGE model based on each country’s input-output table and energy data, presents challenges in accurately updating dynamic data over time. Despite this, the GTAP-E-Power model provides a scientifically robust framework for investigating the impacts of various energy investments on the energy–economy–environment sustainability. While CPEC is a pivotal component of BRI and serves as a model for similar projects, future research should aim to conduct more comprehensive studies, incorporating dynamic modeling techniques to capture the evolving impacts of BRI energy investments more adequately. Expanding the scope to include a broader range of BRI projects will offer deeper insights into how different energy strategies can be optimized to achieve sustainable economic and environmental outcomes on a global scale.