Research on the Security Scenario Simulation and Evolution Path of China’s Power System Based on the SWITCH-China Model

Abstract

Accelerated climate warming has led to the frequent occurrence of extreme weather events, resulting in high-frequency, large-scale, and highly destructive power outages and electricity shortages, which serve as a wake-up call for the safe and stable operation of the power system. To predict safety risks, this study constructs a baseline scenario and five power security scenarios based on the SWITCH-China model, systematically assessing the impact of external shocks on the power system’s evolution path and carbon reduction economics. The results indicate that external shocks are the key factors influencing the power system’s installed capacity structure and generation mix. The increase in demand forces the substitution of non-fossil energy. In the demand growth scenario, by 2060, wind and solar installed capacity will be 1.034 billion kilowatts higher than in the baseline scenario. Rising fuel costs will accelerate the exit of fossil fuel units. In the fuel cost increase scenario, 765 million kilowatts of coal power were reduced cumulatively across three time points. Wind and solar outages, along with transmission failures, lead to significant local economic investments while also causing inter-provincial carbon transfer. In the wind and solar outage scenario, provinces with a high proportion of wind and solar, such as Guangdong and Guizhou, see an increase in carbon emissions of 31 million tons and 8 million tons, respectively. Conversely, provinces with a lower proportion of wind and solar, such as Inner Mongolia and Xinjiang, reduce carbon emissions by 46 million tons and 39 million tons, respectively. Energy storage development supports the expansion of non-fossil energy in the power system. The study recommends accelerating wind and solar deployment, building a storage system at the scale of hundreds of billions of kilowatt-hours, and optimizing the inter-provincial transmission network to address the dual challenges of power security and carbon neutrality.

1. Introduction

Carbon emissions from the power sector are the primary source of emissions in the energy sector, with an annual emission of approximately 4.5 billion tons, accounting for over 40% of China’s total carbon emissions [1]. Achieving low-carbon or even zero-carbon emissions in the power sector will have a significant impact on the energy sector [2]. The power sector must not only consider adjustments to installed capacity but also ensure economic feasibility. Increasing non-fossil energy is a crucial way to achieve the “3060” goal (China aims to peak its CO₂ emissions before 2030 and achieve net-zero greenhouse-gas emissions by 2060). In 2022, the share of non-fossil energy in China’s total energy consumption was 17.5%. Deepening the penetration of non-fossil energy is bound to become the focus of future energy transformation. Therefore, it is of great significance to deeply analyze the power sector’s safety transformation path under extreme security scenarios. This not only provides guidance for high-quality macroeconomic development at the national level but also holds significant theoretical and practical value for achieving carbon peak and carbon neutrality.

In October 2021, the State Council issued the Action Plan for Carbon Peak Before 2030, proposing the development of a new power system with an increasing share of renewable energy and promoting the large-scale optimal allocation of clean power resources. To achieve the goals of carbon peaking and carbon neutrality, the power system must transition from being predominantly reliant on fossil fuel generation to one primarily based on clean energy [3]. Currently, although China leads the world in installed capacity of non-fossil energy sources such as wind, solar, and nuclear power, fossil fuel generation still dominates the overall installed capacity structure [4]. This situation cannot be reversed in the short term and requires long-term efforts across various aspects of the power system, including supply, grid, storage, and demand. Measures such as demand-side modifications, improvements in end-use electrification, and the accelerated construction of transmission corridors are essential to ensuring a secure energy transition. However, during this transition, the power system inevitably faces a series of energy security challenges related to generation, grid, load, and storage. Wind and photovoltaic outputs are dictated by natural climatic conditions—wind speed and solar irradiance—whose fluctuations directly modulate the supply–demand balance in the local power grids. Any downward deviation in renewable generation or upward surge in load demand will both enlarge the supply–demand gap and reduce power system reliability. In high-renewable power systems, this gap is bridged by two principal measures-the regulation capacity of energy storage facilities, and the transmission capacity between provinces. Moreover, fuel costs remain a key determinant of long-run structure of installed capacity in the power system. Therefore, it is crucial to conduct in-depth studies on how security issues in power generation, grid infrastructure, load management, and energy storage under certain scenarios may impact the medium- and long-term stability of the power system, and to develop corresponding mitigation strategies.

Current research primarily focuses on the power system from the perspectives of generation, grid, storage, and load. From the generation side, the significant temporal and spatial fluctuations of renewable energy impact the stability and security of the power system. In particular, wind and solar power generation pose major challenges to the penetration of clean energy [5]. To address insufficient output and enhance peak regulation, strategies such as integrating energy storage on the generation side [6] or improving the flexibility of stable generation units [7] have been proposed.

Additionally, rising fossil fuel prices pose challenges to system stability [8]. The prices of coal, oil, and natural gas are highly susceptible to international political dynamics and geopolitical fluctuations [9], leading to increased costs for imported energy and significantly impacting overall system expenses. Therefore, the selection between fossil fuels and renewable energy on the generation side requires a comprehensive assessment that balances both cost considerations and system stability.

From the grid perspective, China has established extensive electricity transmission corridors between provinces [10], with high-voltage transmission primarily delivering power from inland regions to coastal cities. Due to geographical and climatic conditions, many transmission lines in China are exposed to extreme weather events. Strong winds and freezing rain can reduce transmission efficiency or even cause power outages [11]. Although power transmission increases system costs, it plays a crucial role in the efficient allocation of resources and the reduction of emissions.

From the energy storage perspective, pumped hydro storage and electrochemical storage are the two most widely applied storage technologies [12]. According to the China Energy Storage Alliance (CNESA), pumped hydro storage and electrochemical storage currently account for 37.4% and 57.6%, respectively, of China’s total installed energy-storage capacity. Numerous studies have specifically examined the economic allocation of these two storage technologies, their role in peak shaving, and their deployment at key milestones along the dual-carbon transition pathway [13,14,15]. However, there is limited research on the impact of storage development constraints on power system stability. This is particularly critical when the rapid expansion of storage requires substantial additional costs, necessitating an analysis of spatial and temporal variations across different regions of China.

From the load perspective, with rapid economic growth and increasing end-use electrification, electricity demand across society has surged, exhibiting significant spatial and temporal variations in peak and off-peak characteristics [16,17]. Factors such as daily fluctuations between morning and evening, seasonal variations throughout the year, and differences in electricity consumption among users within the same region profoundly affect load stability. Short-term abnormal load fluctuations [18] can pose serious threats to the security of the power system.

The current power system faces security risks arising from multidimensional fluctuations, yet accurately simulating extreme security scenarios and modeling the dual-scale supply–demand balance between national and provincial levels remains a significant research challenge. To address this gap, this study innovatively employs the high spatial and temporal resolution SWITCH-China model to construct a multidimensional analytical framework under carbon neutrality constraints. This framework incorporates extreme scenarios such as windless and sunless weather, sudden demand surges, and soaring fuel prices, breaking through the limitations of traditional single-risk-source analysis.

For the first time, national level capacity optimization and provincial level power flow have been incorporated into a high spatial and temporal resolution framework, enabling synchronous simulation of national-provincial dual-scale supply-demand balance. Through model-based optimization, the study derives differentiated capacity portfolios tailored to ensure energy security under each extreme scenario—such as optimal wind–solar ratios and coal phase-out pathways—as well as plans for interregional transmission expansion and provincial-level electricity cost distributions. Moreover, it reveals the interactive dynamics of interprovincial electricity support and carbon emission transfers under extreme shocks—for example, surging balancing costs in provinces with high renewable penetration and restricted export capacities in provinces with lower shares.

Based on a comprehensive system analysis, the study proposes multi-scale recommendations for power generation structure, energy storage deployment, and transmission network enhancement. These findings reveal the changing patterns of inter-provincial power mutual assistance and carbon emission transfer under the increasing normalization of climate-related risks, offering a practical scenario-based planning tool and interprovincial coordination strategies for power system design. The study thus fills a methodological gap in the quantitative assessment of energy security under extreme conditions.

2. SWITCH-China Model

2.1. Model Mechanism

SWITCH is an acronym for Solar, Wind, Hydro, Conventional Technology, and Investment. As an optimization and planning model for power systems [19,20], SWITCH-China aims to minimize the total lifecycle cost of electricity generation and transmission by coordinating the deployment and retirement strategies of generation, storage, and transmission infrastructure, both in the present and over the planning horizon. The model surpasses traditional models in both temporal (hourly resolution) and spatial (provincial geographic unit) dimensions, enabling precise simulation of the integration of high shares of variable renewable energy.

As shown in Figure 1, the model takes provincial administrative units as the basic analysis unit and constructs an optimal power system planning framework based on hourly operation data, simultaneously considering the coordinated optimization of long-term grid investment decisions and short-term operational scheduling.

During model optimization, explicit consideration was given to grid reliability, operational constraints, resource availability, and both extant and prospective climate policies and environmental regulations, with particular emphasis on the implications of carbon-peaking and carbon-neutral targets. Existing studies have utilized this model independently or in combination with other models for research. For example, some studies have improved the SWITCH-China model by incorporating cost-minimization uncertainty to conduct an in-depth analysis of China’s power sector [21]. Other studies have integrated the JIMC model to assess employment challenges faced by the coal industry during the energy transition [22]. Research has also found that declining costs of renewable energy and energy storage contribute to carbon emission reductions in the power sector [23]. Additionally, larger balancing regions can enhance the flexibility of China’s power system [24].

This study uses the SWITCH model to simulate the evolution of the power system from 2020 to 2060, setting four time scales: investment periods, months, typical days, and hours. The model includes seven five-year investment periods (2025–2030 to 2055–2060) and uses a stratified sampling method of “12 months × 2 days × 6 h,” selecting a total of 576 study hours (4 periods × 12 months × 2 days × 6 h).

Through typical hour sampling, the calculation time is reduced to 2–3 h; if the initial sampling cannot meet load demand, additional time periods are dynamically added. To represent the system’s extreme and regular operating states, we select the peak load day and median load day of each month, assigning differentiated weights: a weight of 1 for the peak day and a weight of “month days-1” for the median day. This weighting scheme plays a key role in balancing the number of days in the period, economic dispatch, and capacity assurance requirements: (1) Ensure that the total number of simulated days within each investment cycle is equal to the number of days between the start and end of that cycle; (2) Emphasize the economic efficiency of the dispatching system under “average” load conditions; (3) Guarantee that there is sufficient capacity available during peak load periods.

A load-renewable energy coupling modeling approach is used, incorporating historical meteorological data to build a spatiotemporal correlation model. For each typical day, six time slots are selected, each spaced 4 h apart (e.g., 0/4/8/12/16/20 for the median day), and the system’s operational characteristics are captured through the dynamic mapping of historical data and future investment periods.

2.2. Objective Function

This model employs a linear programming approach, aiming to minimize the total investment and operational expenditures of the power system, specifically including: (1) Capital investment in newly constructed and existing generation units; (2) Fixed operation and maintenance costs for various types of generation equipment; (3) Variable operating expenses incurred during the operation of generation units; (4) Costs associated with fuel procurement and consumption; (5) Operating costs of the transmission network and local distribution systems; (6) Capital investment and fixed operation and maintenance expenses for newly built and existing transmission lines.

SWITCH-China is optimized with the goal of cost minimization, as shown in Equation (1), which includes all model costs.

Minimize total cost:

$$\min C=C_{ca}+C_f+C_v+C_{td}$$

(1)

In the equation:

$C$ represents the total cost;

$C_{ca}$ represents the construction costs of all existing and new generation unit projects;

$C_f$ represents the fixed operation and maintenance costs for all existing generation units;

$C_v$ represents the variable costs incurred by each generation unit project, including variable operation and maintenance costs, fuel costs for generation, and fuel costs for providing spinning reserve;

$C_{td}$ represents the costs of existing and new transmission lines and distribution facilities.

The specific calculations for each cost are given by Equations (2)–(5):

$$C_{ca}={\displaystyle \sum _{T,i}{G}_{T,i}\times c_{T,i}}$$

(2)

$$C_f={\displaystyle \sum _{T,i}{G}_{p,i}\times x_{T,i}}$$

(3)

$$C_v={\displaystyle \sum _{T,i}{O}_{T,i}}\times \left(m_{T,t}+f_{T,t}+Carbo{n}_{T,i}\right)$$

(4)

$$C_{td}={\displaystyle \sum _{a,a\prime ,i}Tran{s}_{a,a\prime ,i}\times I_{a,a\prime }\times w_{a,a\prime ,i}}$$

(5)

In the equation:

$T$ represents the power generation technology; $G$ represents the power generation project; $i$ represents the time corresponding to the typical day; $t$ represents the hour point corresponding to each time point; $a$ represents the load area; $a^{\prime }$ represents another load area; $c$ represents the construction cost of the corresponding unit; $x$ represents the fixed operation and maintenance cost of the corresponding unit; $O$ represents the output power of the corresponding unit; $m$ represents the maintenance cost when the corresponding unit is in operation; $f$ represents the fuel cost when the corresponding unit is in operation; $Carbon$ represents the carbon cost when the corresponding unit is in operation; $Trans$ represents the transmission capacity between load areas $a$ and $a^{\prime }$; r represents the transmission distance between load areas $a$ and $a^{\prime }$; $w$ represents the cost of newly built transmission capacity per kilometer; $p$ represents the power generation capacity of the corresponding unit.

2.3. Constraints

The model framework includes five core constraint systems: electricity supply-demand balance constraints, system capacity margin protection constraints, operating reserve maintenance constraints, technological development goal constraints, and total carbon emission control constraints.

The electricity supply-demand balance constraint requires that the power infrastructure of each load area (covering power generation units, transmission networks, and energy storage systems) must meet electricity demand during each hourly operation. This constraint reduces the availability of the rated capacity of the grid equipment by introducing a forced outage rate, reflecting the average available generation capacity per hour during the study period. For baseload power generation units, their output levels must also be further adjusted based on the planned outage rate.

The system capacity margin constraint requires the grid to maintain an additional 15% power supply capability. Specifically, under normal operation of generators, energy storage systems, and transmission networks, each load area must have a power supply capability exceeding 15% of the actual load during each hour. When calculating reserve capacity, the available output of relevant facilities is based on the rated capacity, and the impact of forced outage rates is not considered. The SWITCH-China model uses a joint optimization algorithm to coordinate the simultaneous implementation of power balance dispatch and capacity margin allocation.

The operating reserve maintenance constraint requires that the dynamic reserve capacity maintained per hour in each balancing area must cover the sum of load forecast deviations and intermittent power source fluctuations. At least 50% of the reserve capacity must be in the form of synchronized standby spinning reserves to ensure the system has minute-level response capabilities.

The total carbon emission control constraint requires the implementation of full-cycle carbon emission cap management for the power industry, with emission targets set strictly according to national strategic goals. One of the typical control targets is to reduce carbon emissions by 80% from the 1990 baseline level by 2050, which becomes one of the core elements of the constraint system.

In addition to the carbon emission control target, the system also includes renewable energy quota systems, non-fossil energy share targets, and other objectives, all of which together form a comprehensive technological development constraint framework. This framework not only promotes the large-scale development of clean energy sources such as wind, photovoltaic, and nuclear energy, but also provides clear policy guidance for the green transformation of the power industry.

3. Scenario Setup and Input Parameters

3.1. Scenario Setup

To simulate scenarios of no wind, no sunlight, and demand growth, this study designs six scenarios: the baseline scenario (S0), no wind and no sunlight scenario (S1), demand growth scenario (S2), fuel cost increase scenario (S3), transmission line failure scenario (S4), and slow energy storage development scenario (S5).

All six scenarios are under the same carbon emission constraints, as shown in the following

Table 1. Safety Scenario Setup.

Scenario	Description
Baseline Scenario (S0)	According to the normal operation of the existing power system in each load area
No Wind and No Sunlight Scenario (S1)	Due to wind and solar weather factors, random events occur such as no sunlight for 4 continuous hours during the day or no wind for 4 continuous hours throughout the day
Demand Growth Scenario (S2)	Wind and solar output remain normal, while the peak load on the demand side increases
Fuel Cost Increase Scenario (S3)	Regional fuel costs increase due to changes in local fuel costs and the cost of purchased fuels
Transmission Line Failure Scenario (S4)	Simulate the reduction in interprovincial transmission line efficiency caused by various events
Slow Energy Storage Development Scenario (S5)	When wind and solar grow rapidly, the progress of energy storage development is slow

for scenario settings:

(1) Baseline Scenario (S0)

This scenario assumes that the existing power system operates in each load area under normal operating conditions, with all electricity generation and consumption following the current load forecast and scheduling plans, without considering any external disturbances or special events. This scenario serves as the baseline reference for the study, reflecting the normal performance of the power system under the current technology and policy framework.

(2) No Wind and No Sunlight Scenario (S1)

This scenario simulates the absence of randomness in wind and photovoltaic power generation due to weather factors, specifically scenarios where there is no solar radiation for 4 continuous hours during the day or no wind for 4 continuous hours throughout the day. The no wind or no solar radiation events are mainly set based on the statistical results in reference [25], which showed the occurrence frequency of “no wind” events lasting for 4 h and “no solar radiation” events lasting for 4 h during the day in each region of China in each season of the year. The aim of this scenario is to assess the impact of the absence of wind and photovoltaic power generation on the reliability of the power system under extreme weather conditions and evaluate the system’s ability to cope with fluctuations in natural resources.

(3) Demand Growth Scenario (S2)

In this scenario, it is assumed that wind and solar power generation remain at normal output, while the peak load on the electricity demand side increases by 20%. This scenario aims to simulate the pressure of demand growth on the power system, especially when wind and solar resources are abundant, and examine how the system responds to the electricity supply challenges caused by load growth.

(4) Fuel Cost Increase Scenario (S3)

This scenario assumes that regional fuel costs increase due to the scarcity of local fuel resources and fluctuations in the cost of purchased fuels. This scenario is used to analyze the impact of rising fuel prices on power production costs, generation structure, and electricity market prices, and assess the economic feasibility and operational stability of the power system under high fuel costs.

(5) Transmission Line Failure Scenario (S4)

This scenario simulates situations where interprovincial transmission line efficiency is reduced or failures occur due to various events, which may be caused by natural disasters, equipment aging, or human factors. The aim is to evaluate how the power system adjusts generation schedules and load distribution when transmission capacity is limited and analyze the potential risks to grid stability.

(6) Slow Energy Storage Development Scenario (S5)

This scenario assumes rapid growth of wind and photovoltaic power generation, while the construction progress of supporting energy storage facilities lags behind, resulting in energy storage capacity failing to keep up with fluctuations in renewable energy output. This scenario is used to study how to optimize power system scheduling and the stability of electricity supply in the absence of sufficient energy storage support and evaluate the importance of energy storage facilities in improving system flexibility and renewable energy utilization.

3.2. Input Parameters

3.2.1. No Wind and No Sunlight Events

This scenario primarily simulates the no-wind and no-sunlight events by setting the wind and solar output efficiencies of power generation units in each province within the input data.

The prediction for both data sets is based on the frequency of occurrence of statistical events from related research, combined with probability distributions. No-wind and no-sunlight periods occur when wind speed or horizontal radiation intensity falls below a fixed threshold, with the actual output calculated from the theoretical output.

Below is the representation of the no-wind event for wind power, and the same applies to the no-sunlight event.

The actual wind power output $P_s^{\omega }$ can be represented as:

$$P_s^{\omega }=\left\{\begin{array}{l}0,\\ P_w+\Delta {\epsilon }_w,\end{array}\right.\begin{array}{l}V\le V_{cut-in}\\ V\ge V_{cut-in}\end{array}\begin{array}{l}or\\ and\end{array}\begin{array}{l}\zeta \le P_{no-wind,r,s}\\ \zeta \ge P_{no-wind,r,s}\end{array}$$

(6)

In the equation:

$V$ represents the wind speed, and $V_{cut-in}$ represents the cut-in wind speed of the wind turbine;

$\xi \sim U(0,1)$ is a uniformly distributed random number between 0 and 1;

$P_{no-wind,r,s}$ is the probability of a continuous 4-h no-wind event occurring in a region $r$ during a specific season $s$, which can be calculated from historical statistical data:

$$P_{no-wind,r,s}={\displaystyle \frac{N_{r,s}\times 4}{T}}$$

(7)

Here, $N_{r,s}$ is the number of times a 4-h windless event occurs in region $r$ during season $s$ each year, and $T$ is the total number of hours in a quarter.

When $V\ge V_{cut-in}$ and no wind event has occurred, the wind power output still considers the impact of forecast error $\Delta {\epsilon }_w$, and its probability density function is:

$$f(\Delta {\epsilon }_w)={\displaystyle \frac{1}{\sqrt{2\pi }{\sigma }_{\omega }}}\exp (-{\displaystyle \frac{\Delta {\epsilon }_{\omega }^2}{2{\sigma }_{\omega }^2}})$$

(8)

In the equation:

${\sigma }_{\omega }$ is the standard deviation, which is proportional to the predicted output, i.e.,

$${\sigma }_{\omega }=P_{\omega }·\alpha$$

(9)

where $\alpha$ is the percentage of the standard deviation relative to the predicted output.

Wind and solar resources exhibit significant regional variations in space, and their output characteristics show instability and uncertainty over time. Therefore, when describing scenarios where wind and solar resources are reduced under extreme conditions such as “no wind, no sun,” it is necessary to precisely characterize the time and spatial scales. Based on an analysis of the generation capacity and characteristics of wind and solar power units in China, a “no-wind” event is defined when the cut-in wind speed during wind power generation is less than or equal to 3 m/s. A “no-sun” event is defined when the horizontal radiation intensity for PV modules during daytime generation is less than or equal to 30 W/m².

As shown in

Table 2. Frequency of Continuous 4-h “No Wind” Events in China.

Region	Spring	Summer	Autumn	Winter
Southwest	14	25	16	8
Northwest	36	41	49	42
South	31	38	32	31
Central China	28	33	34	29
East China	17	19	22	21
North China	15	17	25	18
Northeast	12	26	18	19

and

Table 3. Frequency of China’s 4-h “Darkness” Events (Daytime).

Region	Spring	Summer	Autumn	Winter
Southwest	0	0	0	0
Northwest	1	1	0	0
South	4	0	5	4
Central China	2	1	3	2
East China	1	2	3	1
North China	2	2	2	1
Northeast	3	2	1	0

, the primary data are sourced from the literature, providing statistical analyses of the seasonal frequency of 4-h continuous “no-wind” and daytime 4-h continuous “no-sun” events across different regions of China throughout the year.

In the SWITCH-China model, the impact of weather conditions on power generation is represented by the capacity factor of generating units, which is defined as the ratio of actual generation to theoretical generation. When a “no-wind” or “no-sun” event occurs, the capacity factor of the corresponding wind or photovoltaic unit is set to 0, indicating that the unit is operating normally but generating no electricity at that moment.

3.2.2. Electrical Demand Parameters

With the intensification of global climate change, extreme heat events have become increasingly frequent, leading to a significant rise in electricity demand. Particularly during summer, when high temperatures coincide with peak electricity consumption, short-term fluctuations in power load occur, posing unprecedented challenges to the stability and security of grid operations. For example, on 25 July 2024, at 16:38, due to prolonged high temperatures, the power load of the Guangdong grid reached a record high for the fifth time that year, peaking at 156.7 million kilowatts—an increase of 6.25% compared to the highest system load in 2023. The surge in extreme load demand necessitates not only optimizing power resource allocation on the supply side but also enhancing demand-side management to mitigate potential risks faced by the power system.

The load demand $L_t$ at time $t$ is given by the following model:

$$L_t=L_t^{base}\times (1+{\delta }_t)$$

(10)

where

$L_t^{base}$ is the baseline load curve;

${\delta }_t$ is the random fluctuation term, representing short-term stochastic variations in load, defined as:

$${\delta }_t\sim N({\mu }_t,{\sigma }_t^2)$$

(11)

For example, during high temperatures in the daytime, when demand is higher, ${\mu }_t$ = 0.02, and ${\sigma }_t=0.05$; At night, when the load is more stable, ${\mu }_t$ = 0, and ${\sigma }_t=0.02$. When a specific time point experiences an extreme load shock, the short-term random fluctuation term can be adjusted accordingly.

To further investigate the response characteristics of the power system under extreme load demand growth scenarios and the potential security challenges it may face, this study constructs a demand growth scenario model based on the existing load curve. In this scenario, it is assumed that at peak demand periods, around 16:00 each day, electricity demand increases by 10% to simulate potential demand surges under extreme high-temperature conditions. By comparing with the baseline load curve, the impact of sudden load-side demand surges on power system stability and security can be quantitatively analyzed, providing a reference for optimizing power dispatch strategies.

3.2.3. Fuel Cost Parameters

With the rapid changes in the global landscape, China, as an energy-importing country, has an external energy dependency of 9.3% for coal, 70% for oil, and 42% for natural gas. This makes domestic energy usage highly susceptible to geopolitical factors. The main reasons driving energy price increases are as follows:

Supply-demand imbalance: The global economic recovery has led to an increase in energy demand, while supply has not kept pace, especially after the COVID-19 pandemic, when many businesses resumed operations, causing a surge in energy demand; Geopolitical factors: International conflicts and political tensions, especially the conflict between Ukraine and Russia, have led many countries to ban imports of Russian oil, increasing demand for alternative oil sources and driving up prices; Exchange rate fluctuations: Energy sources like crude oil are priced in US dollars, and when local currencies depreciate against the dollar, fuel prices tend to rise; Rising production and transportation costs: Increased costs in energy extraction, production, and transportation, such as equipment maintenance, labor, and logistics, also drive up fuel prices; Insufficient investment: Insufficient investment in fossil fuel sectors can lead to supply shortages, which in turn push prices higher; Policy changes: Countries may implement stricter environmental policies to address climate change, increasing the production and usage costs of fossil fuels; Natural disasters: Natural disasters such as hurricanes and earthquakes can damage energy production facilities, leading to supply disruptions that affect fuel prices; Market impact: Fluctuations in financial markets amplify changes in fuel prices.

To account for potential fluctuations in energy prices, this scenario sets the annual coal price growth rate at 1.5% and the annual gas price growth rate at 3%, based on the coal and gas prices in the baseline scenario.

3.2.4. Line Fault Parameters

In the SWITCH-China model, the power grid transmission lines are configured for unidirectional transmission only. During each time period, the total amount of electricity transmitted must not exceed the transmission and distribution capacity. To simulate the failure of high-voltage transmission lines between provinces during natural disasters such as freezing rain, snowstorms, and strong winds, this study adjusts the line loss rates between load regions to reflect the decrease in transmission efficiency.

Under the condition of constant load, when local generation cannot meet the demand, the system may still build new transmission lines to maintain power supply even if existing transmission lines are damaged or interrupted by disasters. However, the construction cost of new transmission lines is usually higher than that of existing lines, and the transmission capacity may be limited. Therefore, within the framework of minimizing costs in the model, the system must weigh two options: first, increasing local generation capacity, i.e., investing in new units to enhance self-generation capabilities, and second, building new transmission lines to meet load demand through external power transmission.

This optimization process comprehensively considers construction costs, transmission losses, and supply-demand balance to determine the optimal power dispatch strategy under different disaster scenarios. Specific numerical settings refer to studies [26,27], with a decrease of 10–20%.

3.2.5. Constraints on Energy Storage Development

According to the “Blue Book on the Development of New Power Systems” published by the National Energy Administration, the structure of the power system has shifted from the three elements of “generation, grid, and load” to the four elements of “generation, grid, load, and storage.” Energy storage plays a crucial role in peak regulation to ensure the stability of power operation, and the development path of energy storage is also closely related to energy security issues in supply-demand balance.

According to relevant literature research [28], the installed capacity of energy storage is expected to increase from 120 million kW in 2030 to 550 million kW in 2060, with the demand for installation showing a gradual upward trend. Furthermore, the demand for new energy storage is primarily concentrated in the load centers of eastern and southern China [29].

This scenario represents the mismatch between energy storage development and the electricity system demand at different time points. The total installed capacity of energy storage in the model constraints is reduced by 20% relative to the literature research.

4. Results Analysis

4.1. Installed Capacity and Generation

This study simulates the changes in China’s power system from 2023 to 2060 under the “source-grid-storage-load” energy security issue. The most significant changes due to external shocks are observed in the power generation structure and energy mix. Figure 2 below shows the changes in installed capacity for each scenario in 2025, 2030, and 2060.

The total installed capacity in the baseline scenario increases annually, reaching 3.141 billion kW, 3.694 billion kW, and 9.593 billion kW in 2025, 2030, and 2060, respectively.

In the baseline development scenario, the proportion of wind and solar energy will increase annually, while fossil fuels dominated by coal-fired power will gradually decrease. Energy storage will play a regulatory role after the large-scale integration of renewable energy, with installed capacity increasing from 1.2% in 2025 to 10.65% in 2060.

Compared to other scenarios, the scenarios with higher total installed capacity include the no wind and no sun scenario (S1), the demand growth scenario (S2), and the fuel cost increase scenario (S3).

In the no wind and no sun scenario (S1), the output of wind and solar power decreases. However, due to certain carbon constraints and technical limitations, other units face difficulties in increasing capacity, resulting in an increase in wind and solar generation units.

In the Demand Growth Scenario (S2), due to the sudden increase in electricity demand, stable coal-fired power cannot meet the demand. Therefore, only renewable energy installations contribute to the increase in installed capacity, with wind and solar energy increasing by 0.91 billion kW, 1.16 billion kW, and 10.34 billion kW over the three investment periods.

In the Fuel Cost Increase Scenario, due to the reduced use of fossil fuel materials, there is a shift towards non-fossil fuel units. Over the three investment periods, wind and solar energy increase by 1.55 billion kW, 1.79 billion kW, and 3.07 billion kW, while the corresponding fossil units decrease by 2.5 billion kW, 2.86 billion kW, and 2.29 billion kW.

In the Transmission Line Failure Scenario, the transmission efficiency between provinces decreases. Each province, under cost optimization, chooses to build local generation units to achieve power supply and demand balance.

The generation mixes across scenarios are generally consistent. As shown in Figure 3, except for the Demand Growth Scenario (S2), which shows an increase in electricity demand, the total generation in other scenarios remains the same. The difference is that in the No Wind No Light Scenario, wind power is lower than in other scenarios, with a reduction of 479 million kW, 637 million kW, and 1.216 billion kW in the three investment periods relative to the baseline scenario.

By 2060, due to carbon constraints limiting fossil fuel units’ output, and the increased demand for energy storage with higher wind and solar power, the Safe Scenario shows a change in energy storage usage relative to the baseline scenario in 2060, with variations of 14.67%, 55.33%, 22.67%, 10.67%, and −25.33%.

4.2. Output Simulation

In different scenarios, as analyzed in Section 4.1, the installed capacity structure and generation are impacted by external conditions, leading to changes not only on an annual scale but also in terms of output power in a typical year relative to the baseline scenario.

Figure 4 shows the output situation for a typical day in 2030 under each scenario. Except for the Demand Growth Scenario, where the total demand increases at 16:00, the demand in other scenarios remains consistent. The difference lies in the combination of various energy sources at each time point.

Under a certain carbon constraint, the output of coal, gas, and nuclear power remains almost unchanged. The output of renewable energy varies, with wind power reducing significantly under the No Wind No Light scenario. Coal power output increases during peak load periods, and the demand for energy storage charging and discharging increases relative to the baseline scenario.

4.3. Inter-Provincial Transmission Simulation

Each province responds to various energy security impacts by balancing local energy output and complementing different energy types. Additionally, inter-provincial transmission also supports this balance. Figure 5 shows the local installed capacity and inter-provincial transmission capacity of each province in 2030. The size of the pie chart represents the total installed capacity of the province, with each segment representing the installed capacity of a specific energy type. The thickness of the red unidirectional lines between provinces represents the transmission capacity value.

When provinces no longer rely on power transmission, as in the transmission line failure scenario, the transmission efficiency between provinces decreases. This leads to an increase in the electricity supply cost per unit. In this case, load areas will choose to build new local plants rather than rely on transmission. As a result, provinces with a higher share of electricity imports will show an increased radius in the pie chart relative to the baseline scenario, such as Guangdong, Zhejiang, and Tianjin, while provinces exporting electricity, like Inner Mongolia, will have a reduced radius compared to the baseline scenario.

The changes in other scenarios do not significantly affect the total installed capacity. The main change is in the proportion of different types of installed capacity. For example, in the windless and sunlightless scenario, provinces with better wind and solar resources will further expand wind and solar capacity. In this scenario, Inner Mongolia’s share of wind and solar installed capacity increased by 3.25% relative to the baseline scenario.

In provinces with higher electricity demand and a smaller share of stable power sources, this scenario will suppress the development of renewable energy capacity. For example, in the Guangdong scenario S1, the share of wind and solar installed capacity decreased by 23.33% relative to the baseline scenario.

4.4. Interprovincial Carbon Emission Transfer

The emission data for the electricity sector from 2025 to 2035 is sourced from [30,31]. Data for future years is projected based on the emission reduction rate after the peak and considering the decline rate during the investment period from 2030 to 2035. By 2030, carbon emissions from the electricity sector are expected to reach approximately 5.4 billion tons of CO₂, with carbon neutrality achieved by 2060.

The comprehensive calculation is shown in Figure 6, illustrating the emission reduction path for the electricity sector. When the total carbon emissions are fixed, other scenarios relative to the baseline scenario will result in inter-provincial carbon emission transfers.

In this model study, carbon emissions primarily stem from fossil fuel-based power generation, such as coal and gas power. The demand for fossil-based electricity varies across provinces in different scenarios, leading to changes in carbon emissions driven by variations in the electricity generation mix.

Figure 7 below shows the carbon emission transfer amounts of each province relative to the baseline scenario in 2030 under different security scenarios. In the five security scenarios, fourteen provinces exhibit a positive average transfer amount. These provinces are Beijing, Fujian, Guangdong, Guangxi, Guizhou, Hebei, Jiangsu, Shaanxi, Shandong, Shanghai, Shanxi, Tianjin, Xinjiang, and Yunnan. The remaining provinces are reduction provinces. The main reasons for the positive transfer amounts in the emission-increasing provinces are twofold: one is that these provinces have a relatively high share of fossil fuels in their existing power generation mix, such as Guangdong, Shaanxi, Shanxi, and Shandong; another is that these provinces rely heavily on imported electricity, and in scenario simulations, to ensure the balance of supply and demand in the load areas, more local power generation units are needed for adjustment, as seen in Beijing, Tianjin, and Shanghai.

In the no wind and no sunlight scenario, provinces with a larger share of wind and solar installed capacity will experience higher carbon emissions, as local fossil fuel power plants will generate more electricity. For example, in the baseline scenario, the share of wind and solar capacity in provinces like Chongqing, Guangdong, Guangxi, Guizhou, and Zhejiang are 15.04%, 47.47%, 10.23%, 67.59%, and 18.42%, respectively. The higher share of renewable energy in the installed capacity structure leads to increased demand for imported electricity or local fossil-based power when the wind and solar output factors decrease. These provinces will see an increase in carbon emissions relative to the baseline scenario, amounting to 0.16 million tons, 0.31 million tons, 0.12 million tons, 0.08 million tons, and 0.07 million tons, respectively.

On the other hand, provinces with a smaller share of wind and solar capacity are less impacted by this shock. Provinces such as Anhui, Inner Mongolia, Gansu, Hubei, and Xinjiang will see reductions in emissions, with reductions of 0.58 million tons, 0.46 million tons, 0.11 million tons, 0.18 million tons, and 0.39 million tons, respectively, relative to the baseline scenario.

In the demand growth scenario, the original installed capacity structure of each province will primarily expand towards stable output units when demand exceeds supply. For example, provinces such as Anhui, Fujian, Hainan, Jiangsu, Jilin, and Ningxia will see emissions increase by 0.18 million tons, 0.08 million tons, 0.04 million tons, 0.43 million tons, 0.14 million tons, and 0.11 million tons, respectively.

The provinces with emission reductions include Chongqing, Inner Mongolia, Gansu, Henan, Liaoning, Shaanxi, and Xinjiang, with reduction amounts corresponding to 0.17 million tons, 0.32 million tons, 0.19 million tons, 0.15 million tons, 0.23 million tons, 0.15 million tons, and 0.34 million tons, respectively.

In the fuel cost increase scenario, to reduce the total electricity cost of the system while ensuring supply-demand balance, provinces with high coal and natural gas prices will reduce their consumption of fossil power, and emissions will shift to provinces with lower energy prices. For example, carbon emissions in provinces such as Inner Mongolia, Shaanxi, Xinjiang, and Yunnan will increase, with increases of 0.76 million tons, 0.44 million tons, 1.95 million tons, and 1.65 million tons, respectively, compared to the baseline scenario.

In contrast, provinces with higher energy prices or higher coal power generation costs will see a reduction in emissions. These provinces, such as Anhui, Chongqing, Henan, Hubei, Liaoning, and Hunan, will reduce emissions by 1.72 million tons, 0.64 million tons, 1.22 million tons, 0.61 million tons, 0.73 million tons, and 0.63 million tons, respectively.

In the transmission line fault scenario, the transmission efficiency between provinces decreases, and the transmission cost per unit of electricity increases. To minimize costs, the system will favor the construction of new local power plants to meet electricity demand. As a result, the power generation structure of each province will evolve compared to the baseline scenario, and emissions will further shift. This shift is particularly evident in provinces with higher external electricity imports, such as Guangdong, Hebei, Henan, Jiangsu, Shanghai, and Tianjin, where emissions will increase by 1.73 million tons, 2.51 million tons, 1.99 million tons, 1.23 million tons, 0.68 million tons, and 0.51 million tons, respectively.

In contrast, other provinces such as Anhui, Fujian, Jilin, Liaoning, Ningxia, and Shaanxi will see a reduction in emissions, with decreases of 1.03 million tons, 0.81 million tons, 0.45 million tons, 0.49 million tons, 0.56 million tons, and 0.91 million tons, respectively, compared to the baseline scenario.

In the slow development of energy storage scenario, the impact is significant on provinces with a high proportion of renewable energy electricity. The green electricity produced in these provinces cannot be fully consumed and requires energy storage planning to make good use of the resources. In this scenario, carbon emissions transfer is the least, with emissions increasing in Anhui, Fujian, Jiangsu, Ningxia, Shaanxi, and Yunnan by 0.16 million tons, 0.13 million tons, 0.14 million tons, 0.15 million tons, 0.58 million tons, and 1.45 million tons, respectively.

In contrast, emissions in other provinces decrease, which indicates that in this scenario, to meet electricity demand and reduce emissions transfer, more wind and solar power will be produced and used locally, as it is more economical than coal-fired power.

4.5. Emission Reduction Costs and Levelized Cost of Electricity

The levelized cost of electricity (LCOE) reflects the investment required by the power sector to produce one kilowatt-hour of electricity over a given investment period.

As shown in Figure 8, the change in the levelized cost of electricity (LCOE) relative to the baseline scenario is illustrated for the baseline and other scenarios. The LCOE in the baseline scenario increases steadily over the investment period, rising from 1.36 yuan in 2025 to 2.49 yuan in 2060.

In other scenarios, except for the transmission line failure scenario (S4), where costs rise, when transmission efficiency decreases, local plants will either build new units or incur higher costs to ensure power supply. In this scenario, the LCOE increases by an average of 95% compared to the baseline. In other scenarios, the changes in LCOE are not significant compared to the baseline.

In the no-wind, no-sun scenario, the costs relative to the baseline scenario will increase by more than 15% on average. In the no-wind, no-sun scenario, under carbon constraints, the output of fossil fuel units is limited, requiring additional wind, solar, and energy storage to ensure supply-demand balance during critical periods. In the demand growth scenario, the fuel cost increase scenario, and the slow energy storage development scenario, the costs gradually return to the baseline scenario in the neutral year.

In the SWITCH-China model, the carbon reduction cost refers to the additional cost required to achieve the predetermined carbon emission path. It reflects the cost incurred to meet the target in a given investment period and measures the difficulty for the power sector to achieve carbon peak and carbon neutrality goals, which is mainly determined by the emission reduction potential of fossil fuel units when meeting electricity demand.

Figure 9 shows the changes in carbon reduction costs relative to the baseline scenario for each scenario. Under the same carbon reduction target, the carbon reduction cost in the baseline scenario increases year by year, reaching 42.9 yuan/ton in 2025, 98.5 yuan/ton by 2030 (when the peak is reached), and the maximum carbon reduction cost at neutrality is 755.58 yuan/ton.

Except for the fuel cost increase scenario (S3), where the carbon reduction cost will decrease, all other scenarios will increase the system’s carbon costs. Under rising fuel costs, the demand for fossil fuels in the power system decreases, narrowing the cost gap between fossil and non-fossil electricity, and the cost required for corresponding emission reduction choices becomes smaller.

The transmission line fault scenario (S4) and the windless and sunless scenario (S1) have a significant impact on carbon reduction costs. Both scenarios require the construction of more non-fossil fuel units locally to ensure supply-demand balance, leading to a higher demand for economically efficient units.

The demand growth scenario (S2) and the slow energy storage growth scenario (S5) show some increase relative to the baseline scenario, but the impact on the existing installed capacity structure is not significant in these two scenarios.

5. Conclusions

This study uses the SWITCH-China model to simulate a baseline scenario and five power security scenarios to explore the evolutionary paths of energy sources, grids, storage, and loads under various external shocks, as well as the safety measures to ensure supply-demand balance. It demonstrates the differentiated impacts of different scenarios on power generation capacity structure, generation mix, inter-provincial carbon emissions transfer, and carbon reduction costs. The following key conclusions are drawn:

(1) External shocks have a profound impact on the power system’s capacity structure and generation mix. Demand growth and carbon total limitations force the substitution of non-fossil energy. In the demand growth scenario, the demand for wind and solar capacity in 2060 is 1.034 billion kW higher than in the baseline scenario. The increase in fuel costs accelerates the retirement of fossil fuel units. In the fuel cost increase scenario, 765 million kW of coal-fired power capacity is reduced at three time points, and the declining economics of traditional energy will impact the energy capacity structure. Transmission faults lead to extensive expansion of local units. In the case of transmission faults between provinces, when the external purchase of electricity is insufficient in the load areas, the local new units must meet the electricity demand, resulting in a significant increase in total installed capacity.

(2) Under the constraint of total carbon emissions, carbon transfer occurs between provinces due to differences in resource endowments and power structures. Carbon emissions from the power sector will peak at approximately 5.4 billion tons in 2030, and then decrease annually to zero emissions by 2060. Fluctuations in wind and solar generation will exacerbate regional carbon leakage. In the windless and cloudy scenario, provinces with a high share of wind and solar, such as Guangdong and Guizhou, will increase carbon emissions by 31 million tons and 8 million tons, respectively, while provinces with a low share of wind and solar, such as Inner Mongolia and Xinjiang, will reduce carbon emissions by 46 million tons and 39 million tons, respectively. Provinces with a large external purchase of electricity will bear the burden of both carbon transfer and economic costs. In the transmission fault scenario, regions such as Shanghai and Beijing, with higher electricity imports, will experience a shift in carbon emissions from the original purchasing provinces to local areas due to the increase in local fossil units. The rise in fuel prices will expand the carbon transfer phenomenon. In the fuel cost increase scenario, provinces with lower coal prices, such as Inner Mongolia, will see an increase of 7.6 million tons in carbon emissions, while provinces with higher coal power generation costs, such as Anhui, will reduce carbon emissions by 17.2 million tons.

(3) The impacts of safety scenarios pose a significant challenge to the economic viability of the power system. In the baseline scenario, the per-kWh cost increases annually from 1.36 RMB in 2025 to 2.49 RMB in 2060. In the transmission fault scenario, the per-kWh cost increases by an average of 95% compared to the baseline scenario, making it the scenario that has the greatest impact on the system’s economic viability. Carbon reduction costs will reach the highest value of 755.58 RMB/ton in 2060, while the fuel cost increase scenario lowers carbon reduction costs by reducing fossil energy demand. The windless and cloudy scenario and the transmission fault scenario cause the largest increases in carbon reduction costs, being 23% and 19% higher than the baseline scenario, respectively.