Optimization of Industrial Park Integrated Energy System Considering Carbon Trading and Supply–Demand Response

Abstract

To address the challenge of the synergistic optimization of carbon reduction and economic operation in the integrated energy systems (IES) of industrial parks, this paper proposes an optimization scheduling model that incorporates carbon trading and supply–demand response (SDR) coordination mechanisms. This model is based on an IES coupling power-to-gas (P2G) and carbon capture and storage (CCS) technologies. First, the K-means clustering algorithm identifies three typical daily scenarios—transitional season, summer, and winter—from annual operation data. Then, we construct a synergistic optimization model that integrates a carbon trading mechanism, tiered carbon quota allocation, and SDR coordination. The model is solved via mixed-integer linear programming (MILP) to minimize total system operating costs. Systematic comparative analysis across six scenarios quantifies the incremental benefits: P2G–CCS coupling achieves a 15.2% cost reduction and 49.3% emission reduction during transitional seasons; supply–demand response contributes 3.5% cost and 5.6% emission reductions; technology synergies yield an additional 21.6 percentage points of emission reduction beyond individual contributions. The integrated system achieves 100% renewable energy utilization and optimizes peak-to-valley differences across electricity, heating, and cooling loads. Carbon price sensitivity analysis reveals three response stages—low sensitivity, rapid reduction, and saturation—with the saturation point at 200 CNY/t (28.6 USD/t), providing quantitative guidance for tiered carbon pricing design. This research provides theoretical support and practical guidance for achieving low-carbon economic operations in industrial parks.

1. Introduction

1.1. Research Background and Motivation

Against the backdrop of global climate change, carbon reduction has become a core issue in national energy transitions. China has proposed the strategic goals of carbon peaking and carbon neutrality; however, fossil fuels still account for a significant portion of its energy mix [1]. As concentrated areas of energy consumption and carbon emissions, industrial parks play a crucial role in achieving these dual carbon targets through low-carbon transformation [2]. Although wind and photovoltaic power installations have increased, their inherent volatility and intermittency result in the significant curtailment of renewable energy [3]. Relying solely on energy storage or fossil fuels for peak-shaving is costly and fails to meet the demands of deep decarbonization.

The synergistic application of power-to-gas (P2G) and carbon capture and storage (CCS) technologies offers a technical pathway to address these challenges [4,5]. P2G utilizes surplus renewable electricity to electrolyze water, producing hydrogen that reacts with CO₂ to synthesize methane, thereby significantly enhancing the utilization rates of renewable energy. CCS captures CO₂ emissions from fossil fuel power generation [6]. Coupling these technologies establishes an electricity–carbon–gas multi-energy cycle system, where methane produced by P2G can fuel gas turbines. At the same time, CO₂ captured by CCS serves as a carbon source for P2G, enabling the recycling of carbon resources [6]. However, capacity allocation optimization alone cannot fully unlock the potential of low-carbon technologies; market mechanism design at the system operation level is equally crucial.

Carbon trading mechanisms assign economic value to carbon emissions through market-based approaches, internalizing external environmental costs into operational expenses, and incentivizing the prioritization of clean energy within the system [7]. Traditional carbon trading studies often rely on fixed carbon prices or static quota allocations, struggling to adapt to dynamic seasonal and multi-scenario operational demands [8]. A tiered carbon pricing mechanism reflects marginal emission costs in real time through differentiated pricing, enabling more precise coordination between generation, load, and storage. This incentivizes the system to prioritize emission reduction measures during periods of high emissions [9].

Supply–demand response (SDR) effectively reduces system energy consumption and operational costs by adjusting either the supply-side output or the demand-side energy consumption patterns. The supply side achieves flexible electricity–heat output through technologies such as the Organic Rankine Cycle (ORC), while the demand side optimizes energy structures via transferable and substitutable load [10]. However, the existing research predominantly focuses on unilateral dispatch, neglecting the integration of multi-energy flow coupling characteristics with carbon cost signals. This hinders the simultaneous realization of flexibility potential and carbon emission reduction benefits [11]. The synergistic mechanism between carbon trading and supply–demand response guides dynamic responses on both sides through carbon price signals, enabling precise matching between load fluctuations and carbon emission peaks to achieve the dual optimization of environmental benefits and economic costs.

1.2. Related Work

1.2.1. Integrated Energy Systems and Low-Carbon Technologies

An integrated energy system (IES) enhances flexibility and renewable energy integration through multi-energy flow coupling, including electricity, heat, cooling, and gas [12]. Yang et al. [13] found that nearly half of IES projects were terminated due to low economic returns caused by unreasonable capacity allocation. Li et al. [14] proposed a two-stage robust optimization model to address renewable energy uncertainty, while Zhou et al. [15] developed a multi-objective optimization model for hydrogen-storage-coupled systems. Yan et al. [16] demonstrated that integrating energy storage, P2G, and electric boilers could eliminate renewable energy curtailment, albeit at high costs, revealing a trade-off between flexibility and economics.

P2G–CCS technologies exhibit inherent complementarity in carbon management. He et al. [17] integrated CCS and P2G facilities to establish a low-carbon economic dispatch model, enabling flexible system operation under high wind power penetration. Zhang et al. [18] proposed a synergistic optimization method for the planning and operation of both technologies, linking carbon reduction strategies with the utilization of renewable energy. Li et al. [19] integrated P2G and CCS into a combined heat and power (CHP) system, constructing a low-carbon economic dispatch model that demonstrated the significant role of the hybrid system in wind power integration and CO₂ reduction.

Addressing the uncertainties inherent in renewable energy output and load demand, various uncertainty modeling techniques have been developed for multi-energy system optimization. Stochastic optimization methods, such as multi-stage stochastic programming with Markov chain-based scenario propagation [20], effectively capture the temporal evolution of uncertainties but require accurate probability distributions and face computational challenges with high-dimensional problems. Distributionally robust optimization (DRO) offers an alternative by considering ambiguous probability distributions within Wasserstein distance-based ambiguity sets [21,22], achieving a balance between risk management and computational efficiency without precise distribution assumptions. Scenario-based methods employ clustering algorithms or neural network-generated scenarios [23,24] to represent diverse operational conditions, providing flexibility in uncertainty characterization while maintaining tractability. Data-driven forecasting approaches leverage probabilistic modeling and time-series prediction techniques [25,26] to reduce uncertainty impacts, though they require extensive training data and may face generalization limitations. Despite these advancements, most uncertainty modeling studies focus on operational dispatch in isolation, lacking integration with market mechanisms such as carbon trading and tiered pricing strategies.

While these studies have established solid foundations for P2G–CCS technology integration and capacity planning, most works emphasize either capacity allocation or single-technology optimization under deterministic scenarios. The multidimensional coupling between operational–level market mechanisms (e.g., dynamic carbon pricing, supply-demand response) and P2G–CCS technical characteristics across multiple timescales remains to be further explored. Specifically, existing models have not yet fully characterized how real-time carbon cost signals can simultaneously guide P2G hydrogen production, CCS capture intensity, and multi-energy flow dispatch to achieve coordinated optimization.

1.2.2. Research on Carbon Trading Mechanisms

Carbon trading mechanisms, which constrain emissions through quota allocation and market transactions, have been widely adopted in regions like the EU and China [27,28]. Huo et al. [29] developed an integrated power–heat–carbon energy hub model that incorporates seasonal carbon trading and electricity–carbon quota sharing methods, demonstrating the dual optimization potential of integrated dispatch for carbon reduction and resource allocation.

Concurrently, some scholars have explored the relationship between carbon trading policies and emission reduction performance. Lin et al. [30] found that current carbon reduction efforts rely more on government intervention than on market mechanisms, emphasizing the need to develop further and refine market mechanisms while gradually expanding the scope of carbon trading to promote better emissions reduction. Zhang et al. [31] noted that pilot carbon emissions trading significantly constrains industrial emissions, but regional variations in reduction costs necessitate optimizing resource allocation through regional coordination. Zhang et al. [32] proposed establishing a multi-tiered carbon trading market system to promote balanced development across regions with differing economic levels. Song et al. [33] and Hong et al. [34] employ a difference-in-differences model to test carbon trading through quasi-natural experiments. Results indicate that carbon trading enhances energy efficiency, while heterogeneity analysis shows that high marketization and industrial agglomeration facilitate energy efficiency gains under carbon emissions trading schemes.

Building upon these foundational studies, two methodological aspects warrant further investigation: First, most carbon trading models adopt fixed carbon prices or scenario-independent quota allocations, which may not fully capture the seasonal variations and multi-timescale operational dynamics in industrial park energy systems. Second, the quantitative linkage between tiered carbon pricing mechanisms and operational flexibility of multi-energy systems (e.g., P2G–CCS coupling, supply–demand response) remains underexplored, particularly regarding how differentiated carbon prices can dynamically guide emission reduction strategies across different load and renewable energy conditions.

1.2.3. Supply–Demand Response Research

Supply–demand response guides users to adjust electricity consumption through time-of-use pricing, economic incentives, or market signals, or by modifying the energy supply mix on the supply side, thereby achieving coordinated resource optimization across both supply and demand.

Extensive research on supply–demand response has been conducted globally. Parrish et al. [35] identified factors such as familiarity, trust, and perceived risk influencing user participation in demand response programs. Lu et al. [36] developed a real-time incentive-based demand response algorithm that combines reinforcement learning with deep neural networks to mitigate the impacts of uncertainty. Korkas et al. [37] proposed a two-level closed-loop feedback strategy integrating local and centralized control to reduce energy expenditures while ensuring user comfort.

Meanwhile, supply–demand response research has begun expanding toward multi-energy flow coupling. Wang et al. [38] proposed an IES operation method for demand response that incorporates electricity–heat synergy, demonstrating its effectiveness in reducing total costs and carbon emissions. Li et al. [39] introduced a multi-energy system that considers electric vehicles, integrating time-of-use and real-time pricing mechanisms to incentivize flexible user response. Wang et al. [40] developed a low-carbon optimization scheduling model for IES considering carbon capture and hydrogen demand, thereby enhancing multi-energy user satisfaction, IES economic efficiency, and stability while achieving carbon reduction. Wang et al. [41] proposed a multi-timescale game-based optimization scheduling model for park-level IES incorporating diverse demand response types, effectively improving economic efficiency for both operators and users.

Researchers have also explored responses on both the supply and demand sides. Tian et al. [42] proposed a day-ahead scheduling model considering demand coupling response characteristics. Leveraging the advantages of CHP, they introduced a system margin framework for coupling energy supply and demand sides, achieving coordination and optimization across both domains. Simulation analysis demonstrated significant reductions in carbon emissions and total costs for systems employing flexible scheduling on both sides.

While these studies demonstrate the effectiveness of supply–demand response in multi-energy systems, two aspects merit further development: First, most research addresses either supply-side flexibility or demand-side load management independently. The synergistic framework that simultaneously leverages supply-side electricity–heat output flexibility (e.g., ORC waste heat recovery) and demand-side dual-dimensional response (temporal shifting and inter-energy substitution) has not been fully established. Second, the integration mechanism between supply–demand response strategies and real-time carbon cost signals requires deeper investigation, particularly how tiered carbon pricing can dynamically coordinate supply–demand adjustments to achieve peak-carbon alignment and cost-emission co-optimization.

1.2.4. Research Gaps and Positioning of This Study

In summary, while the existing research has made progress in IES optimization, low-carbon technology application, and carbon trading with supply–demand response, two key methodological shortcomings remain:

(1) Unidimensional supply–demand response mechanisms: Existing supply–demand response studies predominantly focus on temporal load shifting (horizontal response) while neglecting the substitutability between different energy types enabled by multi-energy coupling (vertical response). Moreover, the synergistic coordination between supply-side flexibility and demand-side response remains unexplored. The current literature either optimizes supply-side dispatch or demand-side behavior independently, failing to establish a unified framework that captures the complementarity between flexible electricity–heat output (e.g., via ORC waste heat recovery) and multi-type load responses.

(2) Static carbon trading mechanisms with limited adaptability: Carbon trading studies predominantly employ fixed carbon prices or static quota allocations, lacking dynamic pricing mechanisms that reflect real-time marginal emission costs across different operational scenarios and timescales. This disconnect between carbon cost signals and operational flexibility prevents the simultaneous realization of economic efficiency and deep decarbonization.

To address these shortcomings, this paper proposes an operational optimization method integrating tiered carbon pricing with dual-side supply–demand responses: Based on the P2G–CCS coupled IES, tiered carbon pricing dynamically reflects marginal emission costs to guide coordinated adjustments among generation, load, and storage. It constructs a dual-dimensional response framework combining supply-side flexible electricity–heat output (ORC-based waste heat power generation) with demand-side horizontal response (temporal load shifting) and vertical response (electricity–heat substitution), achieving precise matching between load fluctuations and carbon emission peaks for the dual optimization of environmental and economic objectives.

1.3. Research Content and Key Contributions

To address the low-carbon transformation needs of industrial parks, this paper establishes an integrated energy system (IES) framework that couples P2G and CCS technologies, enabling closed-loop carbon management through hydrogen–CO₂ methanation and emission capture. A tiered carbon trading mechanism is designed to reflect marginal emission costs dynamically via differentiated pricing, guiding the system toward emission reduction during high-carbon periods. Furthermore, a supply–demand response coordination model is constructed, integrating flexible supply-side output (ORC-based waste heat recovery) with demand-side adjustments (transferable and substitutable loads), to achieve precise matching between load fluctuations and carbon emission peaks. Through the synergistic optimization of these mechanisms, the system realizes the dual objectives of environmental sustainability and economic efficiency in industrial park energy operations. Key contributions include the following:

(1) Dual-dimensional supply–demand response model: Constructs a coordinated framework integrating supply-side flexible power–heat output (ORC waste heat recovery) with demand-side horizontal response (temporal load shifting) and vertical response (electricity–heat substitution), unlocking the flexibility potential across multiple dimensions of multi-energy coupled systems.

(2) Tiered carbon pricing and response linkage mechanism: Establishes a tiered carbon pricing model based on emission brackets to reflect marginal emission costs dynamically, and proposes a linkage mechanism between carbon price signals and supply–demand response strategies. The model is solved via mixed-integer linear programming (MILP) for typical daily scenarios, establishing precise matching between load fluctuations and carbon emission peaks.

The paper’s structure is as follows: Section 2 constructs a comprehensive energy system model that couples P2G and CCS, proposes a synergistic mechanism integrating carbon trading and supply–demand response, and develops a dynamic scheduling model. Section 3 provides case descriptions and conducts a detailed analysis of the model outputs. Section 4 summarizes the paper.

2. Methodology

To achieve a low-carbon economic operation of the IES, this chapter establishes a comprehensive system model and optimization methodology framework. First, it elaborates on the IES architecture, incorporating P2G and CCS technologies. This system enables multi-energy complementarity and flexible scheduling of electricity, heat, cooling, and gas through the coordinated operation of various energy supply, conversion, and storage devices. Second, mathematical models are established for key system components, including renewable energy supply equipment (e.g., PV, wind), energy conversion equipment (e.g., waste heat recovery (WHR) unit, electric cooling, absorption refrigeration), and energy storage equipment (e.g., electrical, thermal, and cold storage). A dedicated P2G–CCS coupling mechanism model is developed to provide the theoretical foundation for system optimization.

It should be noted that the equipment capacities in this IES are predetermined based on our research team’s previous capacity planning work, which employed annual 8760 h operational data encompassing diverse scenarios, including extreme load and renewable generation conditions [43]. This ensures that the configured system can reliably accommodate full-year operational requirements. Building upon this foundation, the present study focuses on operational-level optimization that coordinates P2G–CCS coupling, tiered carbon trading mechanisms, and supply–demand response strategies to achieve low-carbon economic dispatch under the predetermined equipment configurations.

Building upon this foundation, this chapter introduces a tiered carbon trading mechanism and a supply–demand response mechanism to promote deep emissions reductions and economic operation. The tiered carbon trading mechanism employs a segmented carbon trading cost function to incentivize more proactive emissions reduction measures. Supply–demand response effectively reduces the system’s energy consumption and operational costs while enhancing flexibility by adjusting supply-side output or user energy consumption behavior. Ultimately, this chapter presents a MILP optimization model designed to minimize total operating costs. This model comprehensively considers energy procurement costs, operation and maintenance costs, carbon trading costs, supply–demand response compensation costs, and curtailment costs. It employs the K-means clustering algorithm to generate typical daily scenarios and utilizes the CPLEX solver to determine the optimal operational plans.

2.1. System Model

The IES constructed in this paper is shown in Figure 1. This system adopts an operational strategy that prioritizes ensuring electricity load coverage. It comprises photovoltaic power generation, wind power generation, P2G equipment, CCS equipment, gas turbines, gas boilers, a WHR unit, electric chillers, absorption chillers, as well as thermal storage, electrical storage, and carbon storage facilities. Through the collaborative interaction among these components, the system achieves enhanced efficiency and strengthened low-carbon performance.

P2G technology, as an innovative low-carbon energy conversion method, utilizes electrolyzers to convert electricity into hydrogen. This hydrogen can then react with CO₂ to synthesize methane, supplying natural gas to the gas system. Consequently, this technology offers high flexibility and environmental advantages. Simultaneously, P2G enhances the utilization rate of renewable energy sources, such as wind and photovoltaic power, effectively addressing the issue of energy curtailment caused by their inherent intermittency.

The CO₂ captured by the CCS unit originates from the emissions of gas turbines and gas boilers. A portion of the captured CO₂ is first transported to the P2G equipment for the reaction to produce natural gas, while the remainder is sequestered. This paper employs post-combustion capture technology, which features a simple process, low fixed investment, minimal retrofitting difficulty, and independent system flexibility.

2.2. Equipment Operation Model

The IES incorporates multiple energy supply, conversion, and storage devices. This section establishes mathematical models for these devices, providing a foundation for subsequent optimization analysis.

To balance model complexity with computational efficiency for short-term scheduling, this study simplifies the operational constraints of key equipment (CHP, gas boiler, P2G units) by temporarily omitting minimum ON/OFF time, startup/shutdown costs, and part-load efficiency curves. The simplified model retains the key input–output characteristics, operational bounds, and energy conversion efficiency of each device, which is sufficient to reflect the system’s low-carbon economic operation law and verify the proposed key contributions.

2.2.1. Energy Supply Equipment Model

Photovoltaic power generation is proportional to solar radiation intensity, with the mathematical model expressed as follows:

$$E_{\mathrm{pv},\mathrm{t}}=f{E}_{\mathrm{pv},\max }\left[{\displaystyle \frac{L_{\mathrm{P},\mathrm{t}}}{L_{\mathrm{STC}}}}\right]\left[1+\alpha \left(T_{\mathrm{p},\mathrm{t}}-T_{\mathrm{STC}}\right)\right]$$

(1)

In Equation (1), E_pv,t and E_pv,max represent the actual power and rated power of photovoltaic generation, respectively; f denotes the photovoltaic derating factor; L_P,tv and L_STC denote the actual irradiance and design irradiance, respectively, with the design irradiance set at 1 kW/m²; T_p,t and T_STC denote the actual surface temperature and design surface temperature, respectively, with the design surface temperature set at 25 °C.

The output power of a wind turbine is closely related to wind speed, with its output power equation defined as follows:

$$E_{wt,t}=\left\{\begin{array}{cc}{P}_r\left(w_t^2-w_i^2\right)/\left(w_r^2-w_i^2\right)& w_i<w<w_r\\ P_r& w_r<w<w_f\\ 0& \mathrm{Other}\end{array}\right.$$

(2)

where P_r is the rated power of the wind turbine; E_wt,t is the output power of the wind turbine at time t; w_i, w_r, and w_f are the cut-in wind speed, rated wind speed, and cut-out wind speed of the wind turbine, respectively.

Gas turbines rapidly convert the chemical energy of fuel into mechanical energy, which drives generators to produce electricity. The mathematical model is as follows:

$$E_{mt,t}={\eta }_{mt}{Q}_{mt,t}$$

(3)

In Equation (3), E_mt,t and E_mt,max represent the actual power and rated power of the gas turbine, respectively; η_mt denotes the power generation efficiency of the gas turbine; Q_mt,t indicates the gas power consumed by the gas turbine at time t.

When heat supplied by other sources in the system is insufficient, the gas boiler provides additional heat:

$$H_{b,t}=Q_{b,t}{\eta }_b$$

(4)

where H_b,t represents the heat generated by the gas boiler at time t; Q_b,t denotes the gas power consumed by the gas boiler at time t; and η_b is the efficiency of the gas boiler.

2.2.2. Energy Conversion Equipment Model

The WHR unit utilizes high-temperature exhaust gases from the gas turbine to generate heat, with the following model:

$$H_{\mathrm{re},\mathrm{t}}={\displaystyle \frac{E_{\mathrm{mt},\mathrm{t}}\left(1-{\eta }_{\mathrm{mt}}-{\eta }_{\mathrm{loss}}\right)}{{\eta }_{\mathrm{mt}}}}{\eta }_{\mathrm{hr}}$$

(5)

where H_re,t represents the heat recovered by the device at time t; η_hr denotes the thermal recovery efficiency of the device; η_loss indicates the heat loss of the device.

Absorption refrigerators utilize thermal energy to drive the refrigeration process. The relationship between refrigeration capacity and input heat is

$$C_{\mathrm{ac},t}^c=H_{\mathrm{ac},\mathrm{t}}{C}_{\mathrm{cop},\mathrm{ac}}$$

(6)

where Q_ac,t is the heat consumed by the absorption chiller at time t; $C_{ac,t}^c$ is the power of the absorption chiller at time t; C_cop,ac is the coefficient of performance (COP) of the absorption chiller.

An electric refrigeration machine consumes electrical energy for cooling, modeled as follows:

$$C_{\mathrm{ec},\mathrm{t}}^{\mathrm{c}}=E_{ec}{C}_{cop,ec}$$

(7)

where $C_{ec,t}^c$ is the cooling capacity produced by the electric refrigeration unit at time t; E_ec is the electrical power consumed by the electric refrigeration unit at time t; and C_cop,ec is the COP of the electric refrigeration unit.

2.2.3. Storage Device Model

Storage devices include batteries, thermal storage tanks, and carbon storage units. Their state of charge (SOC) changes are subject to consistent constraints, all satisfying the following conditions:

$$SO{C}_t^i=SO{C}_{t-1}^i\left(1-{\theta }^i\right)+{\eta }_{ch}^i{P}_{ch,t}^i-{\displaystyle \frac{P_{dis,t}^i}{{\eta }_{dis}^i}}$$

(8)

where ${SOC}_t^i$ represents the SOC of storage device i at time t; $P_{ch,t}^i$ and $P_{dis,t}^i$ denote the charge and discharge quantities at time t, respectively; ${\eta }_{ch}^i$ and ${\eta }_{dis}^i$ are the charge and discharge efficiencies; θ_i is the self-discharge rate of the battery.

2.2.4. P2G–CCS Coupling Mechanism

The technical process of P2G equipment is illustrated in Figure 2. Two-stage chemical reaction equations can describe the conversion of power to gas:

$$\begin{array}{l}2H_{2}O\to 2H_2+O_2\\ C{O}_2+4H_2\to C{H}_4+2H_{2}O\end{array}$$

(9)

The hydrogen produced by the electrolyzer is proportional to the input power, with the relationship expressed as follows:

$$Q_{H_2,\mathrm{t}}^{P2G}={\displaystyle \frac{{\eta }^{P2G}{E}_{e,t}^{P2G}}{H_h}}$$

(10)

In Equation (10), $E_{e,t}^{P2G}$ represents the electrical power consumed by P2G at time t; $Q_{H2,t}^{P2G}$ denotes the amount of hydrogen produced at time t; η^P2G indicates the efficiency of the electrolyzer; and H_h signifies the calorific value of hydrogen.

CO₂ and hydrogen undergo the Sabatier reaction, catalyzed by a catalyst, to produce methane and water. This reaction requires appropriate temperature and pressure conditions, with lower temperatures favoring forward reaction progression. Catalyst selection is critical for reaction efficiency and product selectivity. When synthetic CO₂ is insufficient, it can be procured from the CO₂ market. The specific mathematical model is

$$Q_{CH4,t}^M=\tau Q_{H2,t}^{P2G}$$

(11)

$$Q_{CO2,t}^{P2G}=\omega Q_{H2,t}^{P2G}$$

(12)

where $Q_{CH4,t}^M$ represents methane produced at time t; $Q_{CO2,t}^{P2G}$ denotes CO₂ consumed at time t; τ and ω are the reaction coefficients for the Sabatier reaction.

The thermal energy generated by the Sabatier reaction is also utilized to supply heating loads, as shown below:

$$H_{h,t}^M=\varphi Q_{C{H}_4,t}^M$$

(13)

where Φ represents the heat release coefficient of the Sabatier reaction.

Given the characteristics of existing technologies and the feasibility of future large-scale deployment, the carbon capture process discussed herein primarily employs post-combustion capture technology. The CO₂ captured by the CCS facility originates from the emissions of gas turbines and gas boilers. A portion of the captured CO₂ is first transported to the P2G equipment to react and produce natural gas, while the remainder is sequestered.

$$Q_{C{O}_2,t}^{CCS}=Q_{C{O}_2,t}^{P2G}+Q_{C{O}_2,t}^{CCS,f}$$

(14)

where $Q_{CO2,t}^{CCS}$ represents the amount of CO₂ captured by the system at time t; $Q_{CO2,t}^{CCS,f}$ denotes the amount of CO₂ sequestered at time t.

Power consumption during CCS capture comprises capture power consumption and fixation power consumption. Capture power consumption correlates with the amount of CO₂ captured, while fixation power consumption can be treated as a constant, set at 0.1 times the capture power consumption. The total CCS power consumption is

$$E_t^{CCS}=\left(E_{C{O}_2,t}^{CCS}+E_f^{CCS}\right)$$

(15)

$$E_{{\mathrm{CO}}_2,t}^{CCS}={\lambda }_{{\mathrm{CO}}_2}{Q}_{C{O}_2,t}^{CCS}$$

(16)

where $E_t^{CCS}$ is the total energy consumption of the capture system at time t; $E_{CO2,t}^{CCS}$ is the capture energy consumption at time t, related to the operational level of CCS, primarily including power losses from CCS absorption, decomposition, and CO₂ compression; $E_f^{CCS}$ is the fixed energy consumption of the capture system at time t, independent of CCS operational status; λ_CO2 is the energy consumption per unit of CO₂ captured by CCS equipment.

2.3. Carbon Trading Mechanism

Carbon trading is a mechanism that legally allocates carbon emission allowances and permits their trading in the market, primarily promoting carbon reduction through market-based approaches. Under the premise of controlling total emissions, the government allocates carbon emission allowances to various sources of emissions. When an emission source receives free allowances that exceed its actual emissions, it can sell the surplus allowances on the carbon trading market to generate economic benefits. Conversely, if the emissions exceed the allocated allowances, the source must purchase additional allowances on the market or face substantial penalties. The specific carbon trading process is illustrated in Figure 3.

2.3.1. Initial Carbon Emission Allowances

In China, the carbon trading market is currently undergoing comprehensive development. Initial carbon allowances are primarily allocated through three methods: free allocation, auction allocation, and a combination of free allocation and auction allocation. Currently, the power sector in China predominantly employs free allocation for initial carbon emission quotas. This chapter adopts the baseline method to determine the system’s initial carbon quotas, assuming that the initial allocation of carbon emission rights primarily involves three components: grid electricity, gas turbines, and gas boilers. The system’s initial carbon emission quotas are as follows:

$$Q_{C{O}_2,IES}^P=Q_{C{O}_2,\mathrm{buy}}^P+Q_{C{O}_2,mt}^P+Q_{C{O}_2,b}^P$$

(17)

where $Q_{CO2,IES}^P$, $Q_{CO2,buy}^P$, $Q_{CO2,mt}^P$, and $Q_{CO2,b}^P$ represent the carbon emission quotas for the system, grid power purchase, gas turbine, and gas boiler, respectively.

The carbon quota for grid-purchased electricity is calculated as follows:

$$Q_{C{O}_2,\mathrm{buy}}^P={\beta }_e{\displaystyle \sum _{t=1}^T{E}_{buy,t}}$$

(18)

where β_e represents the carbon emission quota per unit of grid electricity.

Carbon quota for gas turbines:

$$Q_{C{O}_2,mt}^P={\displaystyle \sum _{t=1}^T({\beta }_{C{H}_4,e}{E}_{mt,t}+{\beta }_{C{H}_4,h}{H}_{mt,t})}$$

(19)

where β_CH4,e represents the carbon quota per unit of electrical power for gas turbine units, and β_CH4,h represents the carbon quota per unit of thermal power for gas turbine units.

Carbon quota for gas boilers:

$$Q_{C{O}_2,b}^P={\beta }_{\mathrm{gas},h}{\displaystyle \sum _{t=1}^T{H}_{b,t}}$$

(20)

2.3.2. Actual Carbon Emission Model

Within the system, coupled CCS captures substantial CO₂, thereby reducing the system’s CO₂ emissions. Thus, the actual carbon emissions in the system are expressed as follows:

$$Q_{C{O}_2,IES}=Q_{C{O}_2,\mathrm{buy}}+Q_{C{O}_2,mt}+Q_{C{O}_2,b}-Q_{C{O}_2,CCS}$$

(21)

where Q_CO2,IES, Q_CO2,buy, Q_CO2,mt, Q_CO2,b, and Q_CO2,CCS represent the actual carbon emissions from the system, grid electricity purchases, gas turbines, and gas boilers, respectively.

Carbon emissions from grid electricity purchases:

$$Q_{C{O}_2,\mathrm{buy}}={\alpha }_e{\displaystyle \sum _{t=1}^T{E}_{buy,t}}$$

(22)

where α_e is the carbon emission factor per unit of grid electricity.

Carbon emissions from gas turbines:

$$Q_{C{O}_2,mt}={\alpha }_g{\displaystyle \sum _{t=1}^T{Q}_{mt,t}}$$

(23)

where α_g is the carbon emission factor per unit of natural gas.

Carbon emissions from gas boilers:

$$Q_{C{O}_2,b}={\alpha }_g{\displaystyle \sum _{t=1}^T{Q}_{b,t}}$$

(24)

CO₂ captured by carbon capture equipment:

$$Q_{C{O}_2,CCS}={\displaystyle \sum _{t=1}^T{C}_{C{O}_2,t}^{CCS}}$$

(25)

2.3.3. Tiered Carbon Trading Costs

To enhance carbon emission control, this chapter proposes a tiered carbon trading mechanism building upon the traditional carbon trading model. Specifically, the system divides the difference between actual carbon emissions and allocated carbon allowances into multiple intervals, applying distinct prices to calculate carbon trading costs within each interval. When the difference exceeds the length of an interval, the excess portion is traded at the carbon price of the next interval. If the difference is negative, it indicates that the system’s emissions are below its allocated quota. In this case, surplus allowances can be sold to generate revenue. This tiered pricing approach refines carbon trading costs, enabling participants to manage emissions more flexibly and efficiently based on the actual conditions. The difference between actual emissions and allocated allowances is illustrated below:

$$Q_{C{O}_2}^P=Q_{C{O}_2,IES}-Q_{C{O}_2,IES}^P$$

(26)

The calculation model for tiered carbon trading costs is as follows:

$$C_{trade}=\left\{\begin{array}{ll}-\chi (2+3\delta )L+\chi (1+3\delta )(Q_{C{O}_2}^P+2L),\hfill & Q_{C{O}_2}^P\le -2L\hfill \\ -\chi (1+\delta )L+\chi (1+2\delta )(Q_{C{O}_2}^P+L),\hfill & -2L<Q_{C{O}_2}^P\le -L\hfill \\ \chi (1+\delta )\left(Q_{C{O}_2}^P\right),\hfill & -L<Q_{C{O}_2}^P\le 0\hfill \\ \chi Q_{C{O}_2}^P,\hfill & 0<Q_{C{O}_2}^P\le L\hfill \\ \chi L+\chi (1+\theta )(Q_{C{O}_2}^P),\hfill & L<Q_{C{O}_2}^P\le 2L\hfill \\ \chi (2+\theta )L+\chi (1+2\theta )(Q_{C{O}_2}^P-2L),\hfill & 2L\le Q_{C{O}_2}^P\hfill \end{array}\right.$$

(27)

where C_trade is the carbon trading cost; χ is the benchmark price for carbon trading; θ is the growth rate of the carbon trading price; δ is the compensation coefficient; and L is the interval length of carbon emissions.

Since the aforementioned carbon trading model contains nonlinear terms, using MATLAB R2023b and YALMIP R20250626 to call the CPLEX solver requires converting the nonlinear portions into a mixed-integer linear model. First, each segment of the piecewise function is transformed into the form y = ax + b, where x represents the difference between actual carbon emissions and the carbon allowance. Next, introduce the 0–1 variables d_i to represent the activation status of each segment in the carbon trading function. Simultaneously, introduce a continuous variable z_i to represent the carbon trading volume in the i-th time period. Finally, the carbon trading function is transformed into the following Equation (28), which is equivalent to Equation (27). The purpose of introducing these two variables is to constrain the values of x and b within a specific interval.

$$\left\{\begin{array}{l}y={\displaystyle \sum _{i=1}^6(}{a}_i{z}_i+b_i{d}_i)\hfill \\ x=z_1+z_2+z_3+z_4+z_5+z_6\hfill \\ d_1+d_2+d_3+d_4+d_5+d_6=1\hfill \\ d_1{c}_1\le a_1{z}_1\le d_1{c}_2\hfill \\ d_2{c}_2\le a_2{z}_2\le d_2{c}_3\hfill \\ \begin{array}{l}{d}_3{c}_3\le a_3{z}_3\le d_3{c}_4\\ d_4{c}_4\le a_4{z}_4\le d_4{c}_5\\ d_5{c}_5\le a_5{z}_5\le d_5{c}_6\\ d_6{c}_6\le a_6{z}_6\le d_6{c}_7\end{array}\hfill \end{array}\right.$$

(28)

where c_i and c_i+1 represent the upper and lower bounds of the i-th interval.

2.4. Supply–Demand Response Mechanism

The supply–demand response schematic is illustrated in Figure 4, which is divided into two response modes: supply-side and demand-side. The supply side achieves flexible power–heat output through ORC, while the demand side includes both substitutable loads and transferable loads.

2.4.1. Supply-Side Response Model

Within the IES, gas turbines and P2G units generate substantial waste heat. While WHR systems capture part of this heat for heating or cooling, a significant portion remains unused. To address this, this chapter introduces ORC technology, which converts unused waste heat into electricity. This enhances the system flexibility and power supply capacity, enabling the production of both electricity and heat flexibly. Part of the gas turbine’s waste heat supplies loads via the WHR unit, while another portion enters the ORC for power generation. The gas turbine, WHR unit, and ORC form a CHP system, jointly enabling electricity–heat response. The model is illustrated below:

The thermal distribution between the gas turbine and P2G is defined as follows:

$$\left\{\begin{array}{l}{H}_{mt,t}=H_{mt,t}^{WHR}+H_{mt,t}^{ORC}\\ H_{h,t}^M=H_{h,t}^{WHR}+H_{h,t}^{ORC}\end{array}\right.$$

(29)

where H_mt,t, $H_{mt,t}^{WHR}$, and $H_{mt,t}^{ORC}$ represent the waste heat from the gas turbine and the heat entering the WHR unit and ORC, respectively; $H_{h,t}^M$, $H_{h,t}^{WHR}$, and $H_{h,t}^{ORC}$ represent the heat generated by the P2G unit and the heat entering the WHR unit and ORC, respectively.

The heat entering the WHR unit and ORC is, respectively,

$$\left\{\begin{array}{l}{H}_{\mathrm{in},\mathrm{t}}^{WHR}=H_{mt,t}^{WHR}+H_{h,t}^{WHR}\\ H_{\mathrm{in},\mathrm{t}}^{ORC}=H_{mt,t}^{ORC}+H_{h,t}^{ORC}\end{array}\right.$$

(30)

where $H_{in,}^{WHR}$ t and $H_{in,t}^{ORC}$ represent the heat entering the WHR and ORC, respectively.

The heat output from the WHR and the electrical output from the ORC are, respectively,

$$\left\{\begin{array}{l}{H}_{out,t}^{WHR}={\eta }_{WHB}{H}_{\mathrm{in},t}^{WHR}\\ E_{out,t}^{ORC}={\eta }_{ORC}{Q}_{\mathrm{in},t}^{ORC}\end{array}\right.$$

(31)

where η_WHR and η_ORC represent the conversion efficiencies of WHR and ORC, respectively, with values of 0.31 and 0.80; $H_{out,t}^{WHR}$ and $E_{out,t}^{ORC}$ represent the heat and electricity outputs of WHR and ORC, respectively.

The operational constraints for WHR and ORC are as follows:

$$\left\{\begin{array}{l}{H}_{min}^{WHR}\le H_{\mathrm{in},\mathrm{t}}^{WHR}\le H_{\max }^{WHR}\\ H_{min}^{ORC}\le H_{\mathrm{in},\mathrm{t}}^{ORC}\le H_{\max }^{ORC}\\ \Delta H_{\min }^{WHR}\le H_{in,t+1}^{WHR}-H_{in,t}^{WHR}\le \Delta H_{\max }^{WHR}\\ \Delta H_{\min }^{ORC}\le H_{in,t+1}^{ORC}-H_{in,t}^{ORC}\le \Delta H_{\max }^{ORC}\end{array}\right.$$

(32)

where $H_{min}^{WHR}$, $H_{max}^{WHR}$, $H_{min}^{ORC}$, and $H_{max}^{ORC}$ represent the upper and lower limits of thermal power input for WHR and ORC, respectively; Δ$H_{min}^{WHR}$, ${\Delta H}_{max}^{WHR}$, ${\Delta H}_{min}^{ORC}$, and ${\Delta H}_{max}^{ORC}$ represent the upper and lower limits of ramping power input for WHR and ORC, respectively.

The output of the CHP system, comprising the gas turbine, WHR, and ORC, is

$$\left\{\begin{array}{l}{E}_{CHP,\mathrm{t}}=E_{out,t}^{ORC}\\ H_{CHP,\mathrm{t}}=H_{out,t}^{WHR}\end{array}\right.$$

(33)

where E_CHP,t and H_CHP,t represent the electrical and thermal outputs of the CHP system, respectively.

2.4.2. Demand-Side Response Model

Demand response is categorized into horizontal demand response and vertical demand response. Horizontal demand response refers to the ability of loads within a system to shift consumption across different time periods within a scheduling cycle, specifically from peak usage periods to off-peak periods. Vertical demand response refers to the ability of different loads to substitute for one another during the same time period. Based on these horizontal and vertical response capabilities, loads are classified into three categories: fixed loads, transferable loads, and substitutable loads. This is expressed mathematically as follows:

$$P_{\mathrm{use},t}=P_{Load,t}^s+P_{Load,t}^p+P_{Load,t}^c$$

(34)

where P represents the load type; $P_{Load,t}^s$, $P_{Load,t}^p$, and $P_{Load,t}^c$ denote fixed load, transferable load, and substitutable load, respectively.

Horizontal demand response refers to the flexible adjustment of energy consumption across different time periods to achieve efficient energy utilization without significantly impacting overall system energy demand. The model is as follows:

$$\left\{\begin{array}{l}{P}_{Load,t}^{p,0}=P_{Load,t}^{\mathrm{p}}+\Delta P_{Load,t}\hfill \\ \Delta P_{Load,t}={\phi }_{p,t}^{in}{P}_{Load,t}^{in}-{\phi }_{p,t}^{out}{P}_{Load,t}^{out}\hfill \\ {\phi }_{p,t}^{in}+{\phi }_{p,t}^{out}=1\hfill \\ {\displaystyle \sum _{t=1}^T\Delta }{P}_{Load,t}=0\hfill \\ \Delta P_{Load,t}^{min}\le \Delta P_{Load,t}\le \Delta P_{Load,t}^{max}\hfill \end{array}\right.$$

(35)

where $P_{Load,t}^{p,0}$ and ΔP_Load,t denote the transferable and demand-responded quantities of load type P at time t after horizontal demand response; $P_{Load,t}^{in}$ and $P_{Load,t}^{out}$ represent the inflow and outflow power of load type P at time t; ${\phi }_{p,t}^{in}$ and ${\phi }_{p,t}^{out}$ are 0–1 variables indicating the inflow–outflow status at time t; ${\Delta P}_{Load,t}^{min}$ and ${\Delta P}_{Load,t}^{max}$ are the upper and lower limits of inflow–outflow power.

Vertical demand response refers to adjusting energy consumption strategies through technical means based on varying energy prices to select more economical supply methods while meeting end-use energy demands. Considering the characteristics of electricity, heat, and cooling, this paper focuses solely on the substitution between electric heating and electric cooling loads. When electricity prices are low, or power is abundant, electric air-source heat pumps are used for heating; when electricity prices are high, or power is scarce, absorption chillers are employed more extensively. The specific model structure is as follows:

$$\left\{\begin{array}{l}{E}_{Load,\mathrm{t}}^{c,0}=E_{Load,t}^c+\Delta E_{Load,t}^c\\ H_{Load,\mathrm{t}}^{c,0}=H_{Load,t}^c+\Delta H_{Load,t}^c\\ \Delta E_{Load,t}^c={\phi }_{P,t}^{c,in}{E}_{Load,t}^{c,in}-{\phi }_{P,t}^{c,out}(t)E_{Load,t}^{c,out}\\ \Delta H_{Load,t}^c={\phi }_{P,t}^{c,in}{H}_{Load,t}^{c,in}-{\phi }_{P,t}^{c,out}(t)H_{Load,t}^{c,out}\\ {\phi }_{p,t}^{c,in}+{\phi }_p^{c,out}=1\\ \Delta H_{Load,t}^c={\eta }_{\mathrm{eh}}\Delta E_{Load,t}^c\\ \Delta E_{Load,t}^{c,\min }\le \Delta E_{Load,t}^c\le \Delta E_{Load,t}^{c,\max }\\ \Delta H_{Load,t}^{c,\min }\le \Delta H_{Load,t}^c\le \Delta H_{Load,t}^{c,\max }\end{array}\right.$$

(36)

where $E_{Load,t}^{c,0}$ and $H_{Load,t}^{c,0}$ denote the substitute quantities of electricity and heating loads at time t; ${\Delta E}_{Load,t}^c$ and ${\Delta H}_{Load,t}^c$ represent the quantities of electricity and heating loads participating in demand response; $E_{Load,t}^{c,in}$, $E_{Load,t}^{c,out}$, $H_{Load,t}^{c,in}$, and $H_{Load,t}^{c,out}$ respectively, indicate the transfer-in and transfer-out power of electricity and heat substitute loads at time t; ${\phi }_{p,t}^{c,in}$ and ${\phi }_p^{c,out}$ are 0–1 variables indicating the transfer-in/transfer-out status of electricity and heating loads at time t; η_eh is the electricity-to-heat conversion coefficient; ${\Delta E}_{Load,t}^{c,min}$, ${\Delta E}_{Load,t}^{c,max}$, ${\Delta H}_{Load,t}^{c,min}$, and ${\Delta H}_{Load,t}^{c,max}$ are the upper and lower limits of the substitute power for electricity and heating loads.

2.5. Operational Optimization Model

2.5.1. Objective Function

The optimization objective is to minimize the total operating cost C of the IES. The total operating cost includes energy procurement cost C_buy, operation and maintenance cost C_om, carbon cost C_CO2, supply–demand response compensation cost C_SDR, and energy curtailment cost C_cur, as shown in Equation (37):

$$C=\min (C_{\mathrm{buy}}+C_{\mathrm{om}}+C_{C{O}_2}+C_{SDR}+C_{\mathrm{cur}})$$

(37)

Energy procurement costs comprise electricity and gas procurement costs, calculated as follows:

$$\left\{\begin{array}{l}{C}_{buy}=C_{buy}^{\mathrm{e}}+C_{buy}^g\\ C_{buy}^{\mathrm{e}}={\displaystyle \sum _{t=1}^T{r}_{buy,t}^e}{E}_{buy,t}\\ C_{buy}^g={\displaystyle \sum _{t=1}^T{r}_{buy,t}^{\mathrm{gas}}}{Q}_{buy,t}\end{array}\right.$$

(38)

where E_buy,t and Q_buy,t represent the power purchased for electricity and gas at time t; $r_{buy,t}^e$ and $r_{buy,t}^{gas}$ denote the unit price per power unit for electricity and gas purchased at time t.

Operational and maintenance costs refer to expenses incurred during the operation and maintenance of equipment within the system. Their purpose is to ensure regular equipment operation, extend service life, and delay equipment aging and performance degradation. The calculation method is as follows:

$$C_O={\displaystyle \sum _{\pi \in \phi }{\displaystyle \sum _{t=1}^{24}{E}_{\pi ,t}}}{O}_{\pi }$$

(39)

where E_π,t represents the output of the π unit at time t; O_π denotes the unit operating cost of the π unit.

Carbon costs encompass carbon trading costs, carbon capture costs, carbon sequestration costs, and carbon purchase costs. Carbon trading costs are shown in Equation (40). After coupling a carbon capture unit, the additional electrical energy consumption constitutes the cost of carbon capture. After supplying captured CO₂ to P2G, the remaining portion requires sequestration, incurring carbon sequestration costs. If captured CO₂ is insufficient to meet P2G demand, external CO₂ purchases are necessary, representing carbon purchase costs. Specific calculations are as follows:

$$\left\{\begin{array}{l}{C}_{C{O}_2}=C_{trade}+C_{ccs}+C_f+C_{buy}\\ C_{ccs}={\displaystyle \sum _{t=1}^T{r}_{buy,t}^e{E}_t^{CCS}}\\ C_f={\displaystyle \sum _{t=1}^T{r}_{con}{Q}_{C{O}_2,t}^{ccs,f}}\\ C_{buy}={\displaystyle \sum _{t=1}^T{r}_C{Q}_{C{O}_2,t}^{\mathrm{buy}}}\end{array}\right.$$

(40)

where C_CCS, C_f, and C_buy represent carbon capture costs, carbon sequestration costs, and carbon purchase costs, respectively.

The supply–demand response compensation cost is calculated as follows:

$$f_{SDR}={\displaystyle \sum _{t=1}^T\left(\begin{array}{l}{\lambda }_p\left(|\Delta E_{Load,t}^{p,0}|+|\Delta H_{Load,t}^{p,0}|+|\Delta C_{Load,t}^{p,0}|\right)\\ {\lambda }_c\left(|\Delta E_{Load,t}^{c,0}|+|\Delta H_{Load,t}^{c,0}|\right)\end{array}\right)}$$

(41)

where λ_p and λ_c represent the unit compensation costs for transferable load and substitutable load, respectively.

The curtailment cost refers to the cost of unused renewable electricity, as shown below.

$$C_{\mathrm{cur}}=r_{cur}\left(E_{PV,cur}+E_{WT,cur}\right)$$

(42)

where r_cur is the unit curtailment cost, set at 0.7 CNY/kWh; E_PV,cur and E_WT,cur represent curtailed wind and photovoltaic power, respectively.

2.5.2. Constraints

To ensure the stable operation of the system constructed in this chapter, the supply–demand balance constraints for electricity, heating, and cooling loads must be satisfied, preventing system instability caused by supply–demand imbalances. The specific constraints are as follows:

$$\left\{\begin{array}{l}{E}_{\mathrm{wt},t}+E_{pv,t}+E_{CHP,t}+E_{e,c,t}+E_{buy,t}=E_{\mathrm{load},t}^{DR}++E_{e,t}^{P2G}+E_t^{CCS}+E_{ec,t}+E_{e,d,t}+E_{\mathrm{fw},t}\\ H_{CHP,t}+H_{\mathrm{b},\mathrm{t}}+H_{hs,c,t}+=H_{\mathrm{load},t}^{DR}+H_{ac,t}+H_{hs,d,t}\\ C_{\mathrm{e}c,t}^{\mathrm{c}}+C_{\mathrm{ac},\mathrm{t}}^c=C_{\mathrm{load},t}^{c,DR}\end{array}\right.$$

(43)

where $E_{load,t}^{DR}$, $H_{load,t}^{DR}$, and $C_{load,t}^{c,DR}$ represent the system’s electricity, heating, and cooling demand at time t after implementing horizontal and vertical demand responses, respectively.

The SOC of storage devices must be maintained within reasonable limits to prevent excessive charging and discharging, which can damage internal materials and reduce device lifespan. Specific constraints are as follows:

$$\left\{\begin{array}{l}SO{C}_{bt,\min }\le SO{C}_{bt,t}\le SO{C}_{bt,\max }\\ H_{hs,\min }\le H_{hs,t}\le H_{hs,\max }\\ Q_{C{O}_2,\min }\le Q_{C{O}_2,t}\le Q_{C{O}_2,\max }\end{array}\right.$$

(44)

where SOC_bt,min and SOC_bt,max represent the minimum and maximum SOC for batteries, respectively; H_hs,min and H_hs,max denote the minimum and maximum thermal storage capacity for thermal storage tanks, respectively; Q_CO2,min and Q_CO2,max indicate the minimum and maximum carbon storage capacity for carbon storage devices, respectively.

Additionally, to prevent rapid power changes from affecting equipment lifespan, ramping power constraints are set for the gas turbine, P2G equipment, and storage devices. Specific constraints are as follows:

$$\left\{\begin{array}{l}{E}_{mt,\min }^{EL}\le E_{mt,t}-E_{mt,t-1}\le E_{mt,\max }^{EL}\\ E_{P2G,\min }^{EL}\le E_{P2G,t}-E_{P2G,t-1}\le E_{P2G,\max }^{EL}\\ P_{c,\min }^{EL}\le P_{c,t}-P_{c,t-1}\le P_{c,\max }^{EL}\\ P_{d,\min }^{EL}\le P_{d,t}-P_{d,t-1}\le P_{d,\max }^{EL}\end{array}\right.$$

(45)

where $E_{mt,min}^{EL}$ and $E_{mt,max}^{EL}$ represent the upper and lower limits of the gas turbine ramping power, respectively; $E_{P2G,min}^{EL}$ and $E_{P2G,max}^{EL}$ represent the upper and lower limits of the P2G equipment ramping power, respectively; $P_{c,min}^{EL}$ and $P_{c,max}^{EL}$ represent the upper and lower limits of the storage device charging ramping power, respectively; $P_{d,min}^{EL}$ and $P_{d,max}^{EL}$ represent the upper and lower limits of the storage device discharging ramping power, respectively; and where P denotes the type of storage device.

2.5.3. Generation of Typical Daily Scenarios

To reduce computational load while ensuring model representativeness, this study employs the K-means clustering algorithm to analyze 8760 h of annual load data and renewable energy output data. This generates three typical daily scenarios—transitional season, summer, and winter.

The K-means algorithm iteratively optimizes to minimize the sum of squared distances between data points and cluster centers. The algorithm steps are as follows:

(1) Randomly select K initial cluster centers.

(2) Calculate the distance from each data point to all cluster centers and assign the data point to the nearest cluster center.

(3) Recalculate the center point for each cluster.

(4) Repeat steps (2) and (3) until cluster centers remain unchanged or the maximum iteration count is reached.

Through cluster analysis, representative typical daily load profiles and renewable energy output curves can be obtained for subsequent operational optimization studies.

2.5.4. Solution Method

The operational optimization model presented in this paper is formulated as an MILP problem. The model was established using MATLAB software combined with the YALMIP toolbox, and the CPLEX solver was invoked for the solution. CPLEX is currently one of the most efficient commercial optimization solvers, capable of rapidly and accurately solving large-scale MILP problems.

The solution process is as follows: First, input system parameters and typical daily data to establish the objective function and constraints. Then, invoke the CPLEX solver to obtain the optimal operating schemes for each device and the total system operating cost. Finally, analyze the optimization results to evaluate the synergistic effects of carbon trading and supply–demand response mechanisms.

The computational performance of the proposed MILP model was evaluated on a standard computing platform (Intel Core i7, 32 GB RAM) using CPLEX 12.10. Each typical day scenario was solved to optimality within 5 min, with model scales of approximately 600 decision variables and 779 constraints. All solutions achieved relative optimality gaps below 0.01%, confirming certified optimal solutions. The consistent solve times across different seasons demonstrate computational stability regardless of load patterns. The rapid solution times validate the model’s practical applicability for day-ahead operational planning. At the same time, the MILP formulation’s linear scalability characteristics suggest extensibility to larger systems or finer temporal resolutions without prohibitive computational burden.

3. Case Studies and Discussion

3.1. Case Study and Parameters

3.1.1. Case Study

To systematically quantify the incremental benefits of each mechanism and address the reviewers’ concerns regarding baseline comparisons, six comparative scenarios are designed as shown in

Table 1. Comparative scenario configuration.

Scenario	P2G–CCS	Supply–Demand Response	Carbon Pricing
S0	No	No	Fixed
S1	Yes	No	Fixed
S2	No	Yes	Fixed
S3	Yes	Yes	Fixed
S4	No	No	Tiered
S5	Yes	Yes	Tiered

. S0 establishes the baseline without low-carbon technologies or market mechanisms. S1 and S2 isolate the contributions of P2G–CCS coupling and supply–demand response, respectively. S3 examines their synergistic integration under fixed carbon pricing. S4 isolates the effect of the tiered carbon pricing mechanism. S5 represents the complete proposed system integrating all mechanisms. All scenarios maintain identical equipment capacities, renewable energy profiles, and load characteristics to ensure comparability.

3.1.2. System Configuration and Parameters

The parameters for carbon trading and supply–demand response are presented in

Table 2. System parameters.

Category	Parameter	Value
Carbon trading benchmark price	χ (CNY/t)	200
Price growth factor	θ	0.25
Compensation factor	δ	0.20
Interval length	L (kg)	300
Gas turbine power generation benchmark	βCH4,e (tCO2/MWh)	0.3288
Gas turbine heating benchmark value	βCH4,h (tCO2/GJ)	0.0533
Grid electricity baseline value	βe (tCO2/MWh)	0.5801
Substitutable–transferable load compensation cost	λp/λc	0.2

. Gas prices, time-of-use electricity prices, and carbon emission factors are presented in

Table 3. Electricity and gas prices.

Energy Type	Time	Price
Electricity price (CNY/kWh)	8:00–11:00 4:00 P.M.–9:00 P.M.	1.1234
	11:00–16:00 9:00 P.M.–11:00 P.M.	0.7489
	11:00 P.M.–8:00 A.M.	0.3745
Natural gas purchase price (CNY/m3)	1:00 A.M.–12:00 A.M.	3.30

and

Table 4. Key technical parameters of the main equipment.

Equipment	Parameter	Value
P2G	ηP2G	0.80
	Hh(kJ/mol)	282
CCS	λCO2 (kWh/m3)	0.30
	rcon (CNY/m3)	0.10
Photovoltaic	α	−0.50
	τβ	0.90
Wind turbine	wi (m/s)	3.50
	wr (m/s)	12.00
	wf (m/s)	25.00
Gas turbine	ηmt	0.33
Gas boiler	ηb	0.90
WHR unit	ηhr	0.80
Absorption chiller	CCOP,ac	0.80
Electric refrigeration unit	CCOP,ec	4.00
Battery	θbt	0.001
	ηe,c	0.95
	ηe,d	0.95
Thermal storage tank	θhs	0.005
	ηhs,c	0.90
	ηhs,d	0.90
Carbon storage unit	θCO2	0.005
	ηCO2,c	0.90
	ηCO2,d	0.90

, respectively. All monetary values in this paper are expressed in Chinese Yuan (CNY). For international reference, the approximate exchange rate is CNY 7 ≈ USD 1.

This study examines the energy system of a high-tech park in northern China. Typical electricity, heating, and cooling load profiles derived using K-means clustering are shown in Figure 5. As the subject is a high-tech park, the electricity load curves for each typical day exhibit similar trends and values. The period between 8:00 and 18:00 represents working hours, during which electricity demand is high. Summer requires no heating and has no industrial heating demand, resulting in zero heating load during this season. The heating load is present during both transitional seasons and winter, with winter exhibiting higher demand due to lower outdoor temperatures. Concurrently, high-tech parks typically house numerous high-power devices that generate significant waste heat during operation. Consequently, cooling load demand is present across all three typical days, peaking in summer due to the high outdoor temperatures and substantial refrigeration requirements.

Wind turbine and photovoltaic power generation outputs are shown in Figure 6. Wind turbine output peaks on typical transitional season days, followed by winter days, with summer days showing the lowest output. On typical days, wind turbine output peaks in the afternoon. Photovoltaic power generation output peaks in summer, followed by transitional seasons, with winter showing the lowest output.

3.2. Results Analysis

3.2.1. Comparative System Analysis

To quantify the incremental benefits of P2G–CCS, supply–demand response, and tiered carbon pricing against suitable baseline cases, this section conducts a systematic comparison across six scenarios. The analysis focuses on transitional season results to elucidate the underlying mechanisms, with seasonal variations examined subsequently.

(1)

Techno-Economic Performance Comparison

Table 5. Performance metrics of comparative scenarios (transitional season).

Scenario	Total Cost (CNY)	Cost Reduction vs. S0	Carbon Emissions (t)	Emission Reduction vs. S0	RE Utilization (%)
S0	28,412.17	-	13.07	-	99.63
S1	24,090.17	15.2%	6.62	49.3%	100.00
S2	27,405.15	3.5%	12.34	5.6%	99.88
S3	25,948.72	8.7%	3.07	76.5%	100.00
S4	28,412.17	0%	13.07	0%	99.63
S5	25,576.80	10.0%	1.98	84.9%	100.00

presents comprehensive performance metrics for the transitional season. P2G–CCS coupling (S1) achieves the most significant contribution with 15.2% cost reduction and 49.3% emission reduction, eliminating renewable curtailment (from CNY 237.74 to zero). This stems from P2G converting surplus electricity into methane (displacing 1187 m³ daily gas procurement) while CCS transforms the system from carbon credit purchaser to seller. Supply–demand response alone (S2) contributes modestly (3.5% cost, 5.6% emission reduction), as load flexibility provides insufficient renewable absorption compared to power-to-gas conversion.

The synergy between mechanisms emerges in S3: 76.5% emission reduction exceeds the sum of S1 (49.3%) and S2 (5.6%), as supply–demand response creates flexible operational windows enabling higher P2G utilization during renewable oversupply. Critically, S4’s identical performance to S0 confirms that tiered pricing alone cannot drive improvements without enabling technologies, validating the necessity of coordinated market–technical design. The complete system S5 achieves 84.9% emission reduction, with the incremental benefit over S3 attributable to tiered pricing guiding more aggressive emission reduction during high-carbon-price periods.

(2)

Cost Structure Decomposition Analysis

Figure 7 decomposes the cost structure across scenarios, revealing the underlying economic mechanisms driving total cost variations.

Energy procurement dominates costs (84.5–87.0%), with S1 reducing it by 12.7% through fuel displacement and S2 by 5.6% via load shifting to low-price periods. Carbon trading costs transform from expenditure (CNY 504.87 in S0) to progressively increasing revenue: CNY 567.81 (S1), CNY 1424.67 (S3), and CNY 2068.35 (S5). This CNY 643.68 incremental revenue from S3 to S5 quantifies tiered pricing’s economic value—by avoiding emission levels triggering higher price tiers, the system operates CCS more aggressively, creating additional tradable allowances that transform deep decarbonization from cost burden to profit opportunity.

While P2G–CCS introduces carbon sequestration costs (CNY 243–506) and supply–demand response incurs compensation costs (CNY 588–651), procurement savings and carbon revenue offset these. Notably, S3’s procurement cost (CNY 22,578) exceeds S1’s (CNY 20,969), indicating that simultaneous supply–demand response and P2G operation create competition for renewable absorption windows, partially diminishing P2G’s fuel displacement benefits.

(3)

Seasonal Variation in Marginal Values

Table 6. Marginal benefits of individual mechanisms across seasons.

Season	Metric	P2G–CCS (S1 vs. S0)	SDR (S2 vs. S0)	Tiered Pricing (S5 vs. S3)
Transitional	Cost reduction (CNY)	4322.00 (15.2%)	1007.02 (3.5%)	−371.92 (−1.4%)
	Emission reduction (t)	6.45 (49.3%)	0.73 (5.6%)	−1.09 (−35.5%)
Summer	Cost reduction (CNY)	1294.54 (3.2%)	530.73 (1.3%)	−373.68 (−0.9%)
	Emission reduction (t)	2.04 (9.5%)	0.44 (2.0%)	−7.15 (−44.7%)
Winter	Cost reduction (CNY)	4070.54 (10.6%)	1463.91 (3.8%)	−380.05 (−1.1%)
	Emission reduction (t)	11.00 (59.6%)	1.22 (6.6%)	−2.90 (−46.3%)

quantifies the marginal values of each mechanism across three seasons, revealing significant seasonal dependencies that inform deployment prioritization strategies. P2G–CCS demonstrates the highest cost-effectiveness during transitional seasons (CNY 4322, 15.2%) and winter (CNY 4071, 10.6%), but deteriorates in summer (CNY 1295, 3.2%) due to reduced wind availability (512 kW vs. 814 kW transitional average), limiting electrolysis hours. Emission reduction peaks in winter (59.6%) due to higher baseline emissions (18.45 t) from heating loads, while summer shows the lowest effectiveness (9.5%) despite the highest baseline (21.46 t), as continuous gas turbine operation for cooling constrains CCS offset capability.

SDR’s effectiveness peaks in winter (CNY 1464, 3.8%) due to the pronounced diurnal heating variation enabling substantial load shifting (23% of 20:00–22:00 peak to midday), while summer’s flatter cooling profile limits flexibility. Comparing S5 with S3 reveals tiered pricing increases gross operational costs (CNY 372–380) through aggressive CCS deployment but generates substantially higher carbon revenue (2068 vs. 1425 CNY transitional), yielding net benefits. Summer emerges as the most challenging season for all mechanisms, suggesting that alternative strategies such as enhanced thermal storage may be needed.

3.2.2. Response Characteristics and Cost Analysis for Typical Days

(1)

Supply-Side Response Analysis

The supply-side response characteristics shown in Figure 8 demonstrate the flexible regulation of the ORC for CHP systems. During different seasons, the proportion of waste heat power generation at 10:00 significantly increases. This occurs because, although electricity demand rises at 8:00, the increase in power generation is relatively small to meet the system’s thermal requirements. After 10:00, thermal demand decreases while the amount of waste heat in the system increases. Thus, the proportion of power generation increases at this time. In summer, the heating load demand is low, requiring only sufficient heat to meet the demands of absorption chillers. Consequently, most waste heat is utilized for power generation during the summer, resulting in a significantly higher proportion of power generation compared to other seasons. In winter, the heating load demand is high, resulting in the lowest proportion of waste heat power generation during this season.

(2)

Demand-Side Response Analysis

The supply–demand response patterns for typical days are shown in Figure 9. Across the three representative days, the horizontal demand response trends for electricity load exhibit similar characteristics. Between 23:00 and 8:00, during the low-price period, the system shifts electricity consumption from higher-cost periods to this timeframe. This approach reduces peak-hour electricity usage, thereby decreasing grid power purchases and alleviating peak-period grid pressure. On the other hand, it utilizes surplus electricity from periods of low demand within the system to optimize power resource allocation. During the transitional season and winter, vertical demand response occurs between electricity and heating loads during periods of peak electricity consumption. This reduces operating costs and energy consumption while optimizing energy allocation. Due to the relatively low heating load during the transitional season, the load participating in vertical demand response is correspondingly lower.

Horizontal demand response for heating loads primarily occurs during winter, shifting a portion of thermal demand from high-demand periods to low-demand periods to optimize thermal resource allocation. Simultaneously, it substitutes for part of the peak electricity load during transitional seasons and winter, further reducing operating costs. For cooling loads, peak cooling demand is shifted to off-peak periods without compromising comfort. This reduces the use of electric chillers and absorption chillers during peak hours, thereby lowering electricity and natural gas consumption. By integrating horizontal and vertical demand response, the system achieves flexible load adjustments for electricity, heat, and cooling across different time periods. This enables optimal resource allocation and reduces overall operational costs.

(3)

Cost Analysis

Table 7. Components of total operating costs by typical day.

Cost	Unit	Transitional Season	Summer	Winter
Total Cost	CNY	25,575.27	39,643.63	34,763.56
Energy Procurement Cost	CNY	22,777.11	37,174.70	31,865.88
Carbon Trading Costs	CNY	−2071.61	−2530.72	−2328.07
Carbon Sequestration and Purchase Costs	CNY	506.96	678.53	688.12
Demand Response Compensation Cost	CNY	649.76	841.78	787.13
Energy Curtailment Costs	CNY	0	0	0
Operational Maintenance Cost	CNY	3713.05	3479.36	3750.49

shows that the total cost for a typical summer day is the highest at CNY 39,643.63, followed by the typical winter day at CNY 34,763.56. The lowest total cost is observed in the transitional season at CNY 25,575.27. The cost composition in the table indicates that energy procurement costs constitute the central portion of total expenses. Compared to the transitional season, energy procurement costs for typical summer days and typical winter days show a significant increase, resulting in higher total costs for these days. The increase in energy procurement costs stems from two factors. First, the higher wind turbine output during the transitional season reduces gas turbine output and grid electricity purchases.

Additionally, P2G output exceeds that of typical summer days, consuming part of the natural gas to meet its own demand, thereby lowering natural gas and grid electricity procurement costs. Second, lower heating and cooling load demands during the transitional season reduce energy consumption for supplying these loads. The combined effect of these factors results in significantly lower energy procurement costs for typical transitional season days compared to the other two typical days. Similarly, typical winter days exhibit higher wind turbine and P2G outputs than typical summer days, resulting in lower energy procurement costs compared to summer.

Additionally, the carbon trading cost for the summer is lower than that for the other two typical days. Although the system’s natural gas consumption increases during a typical summer day, leading to higher CO₂ emissions, the presence of CCS equipment captures the CO₂ generated within the system. This results in the system’s actual carbon emissions being lower than the allocated free carbon emission allowances, enabling the sale of surplus allowances to generate revenue. Moreover, carbon trading costs were negative across all three typical days, indicating that CCS operations effectively offset the environmental costs associated with natural gas consumption. Additionally, supply–demand response compensation costs were higher on typical summer days compared to other typical days. This stems from increased outdoor temperatures driving higher cooling demand. To optimize energy allocation and reduce operational costs, peak cooling loads were shifted, resulting in elevated compensation expenses. On other typical days, although various types of supply–demand response exist, lower cooling demand results in relatively minor volumes of cooling supply–demand response. Furthermore, electricity supply–demand response volumes vary little across different typical days. Therefore, despite the presence of heating supply–demand response, the overall supply–demand response volume remains lower than that of summer.

3.2.3. Dynamic Multi-Energy Flow Balance Analysis for Typical Days

(1)

Transitional Season Typical Day

The electricity supply–demand balance for a typical transitional season day is shown in Figure 10. From 0:00 to 8:00, the electricity demand is relatively low, and the output of wind turbines can basically meet the electricity demand. However, to ensure the stability and economy of the gas turbine, it operates at minimum load, resulting in surplus wind power generation. To reduce the curtailment rate of renewable energy electricity, the system batteries begin charging, and the P2G equipment is activated to convert excess electricity into methane. From 8:00 to 10:00, electricity demand increases. Concurrently, rising cooling loads cause electric chillers to operate at higher power, further elevating total electricity requirements. During this period, low solar irradiance results in minimal photovoltaic generation. Although CHP units increase the output, the system’s self-generated power cannot meet demand due to limitations in gas turbine ramping. Consequently, batteries discharge, and electricity is purchased from the grid. From 10:00 to 18:00, output from PV, wind turbines, and CHP units increases. Their combined power supply meets the system’s electricity needs. Afternoon PV output gradually decreases, but increased CHP output maintains a sufficient power supply. From 18:00 to 22:00, electricity demand decreases, and the power consumption of electric chillers also begins to decline. The coordinated operation of wind turbines, CHP units, and batteries can meet the system’s electricity requirements. Between 22:00 and 24:00, both electricity and cooling demands continue to decrease. The P2G equipment operates in conjunction with batteries to absorb excess renewable electricity.

The thermal and cooling energy supply–demand balance for a typical transitional season day is shown in Figure 11 and Figure 12. During this season, heating load demand is relatively low and can be met through the WHR unit from CHP units and the dynamic regulation of thermal storage tanks. Cooling demand is supplied by the coordinated operation of absorption chillers and electric chillers, with absorption chillers contributing 20% of the cooling capacity. This fully utilizes the system’s waste heat resources, enabling rational energy allocation and reducing overall system energy consumption. Through the coordinated and complementary operation of various energy equipment within the system, electrical, heating, and cooling demands are met while maximizing the use of renewable energy and minimizing dependence on the grid.

(2)

Typical Summer Day

The electricity supply–demand balance for a typical summer day is shown in Figure 13. The electricity load on a typical summer day differs little from that in the transitional season, but wind turbine output decreases due to reduced wind speeds. Therefore, between 0:00 and 8:00, although both electricity and cooling loads are relatively low, CHP units and wind turbines are unable to meet the electricity demand fully. Given that this period falls during off-peak electricity pricing, the system prioritizes purchasing electricity from the grid and storing it in batteries. Between 7:00 and 8:00, CHP output begins to increase to avoid excessive purchases of high-priced electricity in the next time slot. Between 9:00 and 18:00, both electrical and cooling loads increase, resulting in a rise in total electricity demand. Although output from PV, CHP units, and wind turbines increases, it still cannot fully meet all electricity requirements. At this point, the system employs a hybrid power supply strategy: during high-price periods, it prioritizes discharging batteries to reduce grid purchases; during flat-price periods, it prioritizes purchasing grid power while charging batteries, achieving optimized energy allocation. From 19:00 to 21:00, although total electricity demand decreases, it remains a high-price period. To minimize grid purchases, CHP units maintain high operating power while the batteries continue to discharge. From 21:00 to 24:00, total electricity demand further decreases. The CHP unit and wind turbine output can largely meet the system’s power requirements, necessitating only battery discharge and a small amount of grid electricity purchase for supplementation.

The heating and cooling energy supply–demand balance for a typical summer day is shown in Figure 14 and Figure 15. Although the heating load is zero on a typical summer day, the high ambient temperature increases the cooling demand. This drives the absorption chillers to operate and provide cooling capacity, meaning that there is still a heat demand within the system. From 0:00 to 8:00, with low cooling demand and electricity rates in the off-peak period, the required cooling is supplied by electric chillers. The waste heat recovered by the CHP units is stored in thermal storage tanks for later use. From 8:00 to 11:00, during peak electricity rates, the system activates the absorption chiller to reduce electricity load demand. The required heat is supplied by a combination of the thermal storage tank and the gas boiler. From 11:00 to 16:00, during standard electricity rates, the electric chiller increases its output while thermal demand decreases. Most of the heat is provided by the CHP unit, and the output of the gas boiler gradually decreases. From 16:00 to 21:00, as PV output gradually decreases, the system prioritizes utilizing waste heat from the CHP unit for power generation. The electric chillers meet most cooling demands during this period. From 21:00 to 24:00, despite lower electricity rates, both absorption chillers and electric chillers jointly provide cooling to maximize waste heat utilization.

(3)

Typical Winter Day

The electricity supply and demand balance for a typical winter day is illustrated in Figure 16. Compared to the two other typical days, the electricity load on winter days remains relatively stable. However, wind speeds are generally higher in winter than in summer, resulting in increased output from the wind turbines. Between 0:00 and 8:00, after meeting electricity demand, surplus output from the wind turbines and the base output of the CHP units exist. Consequently, the system activates P2G units while batteries begin charging. From 8:00 to 12:00, total electricity demand increases. Due to the ramping power limitations of the gas turbine, the CHP unit’s output rises slowly. Photovoltaic output also fails to reach its peak. Although batteries continue to discharge, they cannot fully meet demand, necessitating the purchase of power from the grid. During the 12:00–15:00 period, CHP output reaches its peak. Through the coordinated power supply from PV and wind turbines, the electricity demand is met. Surplus power is used to charge batteries, reducing external electricity purchases during high-cost periods. Between 15:00 and 22:00, the total electricity demand gradually decreases. Although PV and wind turbine output declines, the system’s self-generated electricity suffices for most periods, requiring only minor grid purchases during localized intervals. From 22:00 to 24:00, total demand further decreases. The system activates the P2G equipment, which synergizes with batteries to enhance the utilization of renewable energy.

The thermal and cooling energy supply–demand balance for a typical winter day is shown in Figure 17 and Figure 18. The heating load on a typical winter day exceeds that of the other two typical days. Thermal demand is dynamically regulated and supplied through the coordinated operation of gas boilers, CHP units, and thermal storage tanks. During the nighttime peak heating period, the CHP units operate at minimum load due to low electricity demand, resulting in reduced heat output. Thermal demand is primarily met by gas boilers, with thermal storage tanks supplementing any energy shortfall. Daytime represents the thermal off-peak period. After meeting thermal demand, excess heat is stored in the thermal storage tank. When thermal demand increases, the storage tank releases heat to reduce the fuel consumption of the gas boiler. Cooling demand is concentrated during working hours and is provided jointly by absorption chillers and electric chillers. Due to the high thermal demand on a typical winter day, less waste heat is available for utilization; therefore, electric chillers supply most of the cooling.

The three typical days demonstrate the coordinated operation of heating, cooling, and electricity, achieving 100% renewable energy utilization through cross-period synergies of P2G, energy storage, and supply–demand response—i.e., all generated renewable electricity is effectively utilized. Due to seasonal characteristics, the proportion of purchased electricity varies significantly across different typical days: 1.01% for the transitional season, 9.86% for summer, and 2.48% for winter. The proportion of purchased electricity during the typical summer day is significantly higher than during the other two typical days. This is because wind turbine output is lower in summer, while cooling demand is higher, leading to increased output from electric chillers. Consequently, the total electricity demand rises, causing the system’s self-sufficiency to fall short during peak hours, necessitating the purchase of electricity. During the summer, the system could consider implementing a dynamic pricing mechanism. Raising compensation rates during peak hours would incentivize users to engage in deep load regulation, thereby reducing the proportion of high-cost purchased electricity and lowering energy procurement costs.

3.2.4. Analysis of Carbon Price Impact on Typical Days

Figure 19 illustrates the relationship between carbon price and its impact on total operating costs, carbon trading costs, and carbon emissions across three typical days. The trends in total operating costs for these days are similar, showing an initial increase followed by a decrease as the carbon price rises. However, the carbon price at which total operating costs peak varies across days due to seasonal characteristics. Within the carbon price range of 40–240 CNY/t, peak total operating costs for the transitional season, summer, and winter occur at 80 CNY/t, 160 CNY/t, and 100 CNY/t, respectively. After these peaks, costs decline rapidly by an average of 5.9 CNY/t, 10.4 CNY/t, and 7.2 CNY/t. This provides a quantitative basis for designing tiered carbon pricing. Specifically, during the transitional season and summer, carbon trading generates profits after the inflection point, covering increased energy procurement and operational costs to reduce total operating expenses. In contrast, during a winter day, although carbon trading yields profits at the inflection point and enhances the operational flexibility of CHP units, these profits are insufficient to offset rising energy procurement and operational costs. A higher carbon price is required to achieve a reduction in total operating costs.

From the perspective of carbon trading costs, the costs for the three typical days exhibit consistent trends with rising carbon prices: at lower prices, costs increase with higher carbon prices; as prices further rise, costs decrease; and finally, when prices reach a certain threshold, the system generates profits through carbon trading. The carbon price at which profits are realized differs across typical days, determined by seasonal variations in energy supply structures. Summer and winter exhibit higher cooling and heating loads, respectively, coupled with relatively lower renewable energy generation, resulting in a greater reliance on fossil fuels compared to the transitional seasons. Consequently, these seasons require higher carbon prices to generate profits through carbon trading, thereby reducing total operating costs.

Carbon emissions exhibit three response phases as carbon trading prices increase:

• Low-Sensitivity Phase, which primarily relies on adjusting system operation strategies to reduce fossil fuel usage, yielding limited emission reductions.
• Rapid Reduction Phase: The CCS facility’s output significantly increases, capturing large volumes of CO₂ emissions, resulting in a rapid decline in carbon emissions.
• Saturation Phase: System equipment outputs stabilize, resulting in essentially unchanged carbon emissions.

These response phases exhibit seasonal characteristics: carbon prices of 80–200 CNY/t trigger rapid emission reductions during transitional seasons and winter. However, summer’s high cooling demand increases the reliance on fossil fuels, requiring higher carbon prices (100–200 CNY/t) to enter this phase. Additionally, the three typical days all enter the saturation phase at a carbon price of 200 CNY/t, beyond which further increases in carbon price show no significant additional effect on carbon reduction.

Beyond the carbon price, the interval length L and coupled parameters (compensation factor and growth factor) significantly influence system operational performance through distinct mechanisms. Figure 20 illustrates the impact of interval length L on total operating costs, carbon trading costs, and carbon emissions for the transitional season typical day. The response characteristics exhibit three distinct phases across the examined range. In the initial phase (L ≤ 400 kg), total operating costs demonstrate a gradual ascending trend while carbon trading revenue experiences a moderate decline, yet carbon emissions remain relatively stable. This behavior indicates that the moderate expansion of the interval length does not fundamentally alter the system’s operational strategy, as the carbon pricing mechanism continues to provide sufficient economic incentives for emission control. However, a critical threshold effect emerges at L = 500 kg, manifested by an abrupt escalation in total operating costs, a sharp deterioration in carbon trading revenue, and a substantial increase in carbon emissions. This discontinuity reveals that excessively large interval lengths diminish the price differentiation between adjacent intervals, thereby weakening the marginal incentive for emission reductions and prompting increased reliance on fossil fuel-based generation. The threshold at 500 kg represents a critical tipping point where the granularity of the tiered pricing mechanism becomes insufficient to maintain effective emission control. Beyond this threshold, the system enters a saturation phase where further increases in L yield negligible additional impact on operational decisions, suggesting that the carbon trading mechanism has lost its regulatory effectiveness under such parameter configurations.

Figure 21 presents the coupled influence of the compensation factor and the growth factor on system performance, revealing pronounced nonlinear interactions between these parameters. As the coupled parameters increase from [0.10, 0.15] to [0.35, 0.40], total operating costs exhibit a monotonic decline while carbon emissions demonstrate a two-stage response pattern. This bifurcated behavior originates from the dual mechanisms inherent in the tiered carbon trading structure: the compensation factor amplifies the economic penalty for exceeding emission allowances, while the growth factor intensifies the price gradient across successive intervals. Their synergistic effect manifests most prominently in the lower parameter range ([0.10, 0.15] to [0.20, 0.25]), where carbon emissions decline steeply, and carbon trading revenue increases substantially, effectively offsetting the augmented operational expenditures associated with enhanced CCS utilization. Beyond the parameter threshold of [0.20, 0.25], carbon emissions enter a plateau region, indicating that CCS equipment has approached its optimal operational intensity under prevailing energy demand constraints. Further parameter increases continue to reduce total costs through improved carbon trading performance, yet yield only marginal emission reductions. This asymmetric response characteristic underscores the existence of an optimal parameter configuration zone wherein environmental and economic objectives achieve Pareto-efficient optimization. Quantitative analysis across typical summer and winter days demonstrates consistent parametric sensitivity trends, thereby validating the robustness and generalizability of these findings across seasonal operational scenarios.

4. Conclusions

This study addresses the low-carbon economic operation requirements of IES in industrial parks. Within a framework that couples P2G and CCS technologies, it introduces a synergistic mechanism that integrates carbon trading and supply–demand response to construct a multi-timescale optimization dispatch model. The K-means clustering algorithm identified three typical daily scenarios—transitional season, summer, and winter—effectively reflecting seasonal load characteristics and patterns of renewable energy output.

Comparative scenario analysis across six progressively integrated systems quantifies the incremental benefits: (1) P2G–CCS coupling achieves 15.2% cost reduction and 49.3% emission reduction during transitional seasons, with effectiveness varying seasonally (winter 10.6%/59.6%, summer 3.2%/9.5%); (2) supply–demand response contributes 3.5% cost and 5.6% emission reductions, with winter showing the highest potential (3.8%/6.6%); (3) tiered carbon pricing requires enabling technologies, as standalone deployment (S4) shows zero improvement, but generates additional CNY 643 (USD 91.9) carbon revenue when integrated with P2G–CCS–SDR; (4) the technology synergy between P2G–CCS and supply–demand response yields 21.6 percentage points additional emission reduction beyond individual contributions; and (5) the complete integrated system achieves 84.9% emission reduction while maintaining economic viability across seasons.

Specifically, the supply–demand response mechanism reduces the peak-to-valley differences for electricity, heating, and cooling loads by 12.7%, 5.4%, and 2.1%, respectively. Renewable energy utilization rates reached 100% across all three typical days, validating the mechanism’s significant role in peak shaving, valley filling, and the integration of renewable energy. Carbon price sensitivity analysis revealed the nonlinear response characteristics between system operating costs and carbon emissions. Within the carbon price range of 40–240 CNY/t (5.7–34.3 USD/t), cost peaks occurred at 80 CNY/t (11.4 USD/t), 160 CNY/t (22.9 USD/t), and 100 CNY/t (14.3 USD/t) for typical days in the transitional season, summer, and winter, respectively. Following these peaks, the average cost reductions were 5.9 CNY/t (0.84 USD/t), 10.4 CNY/t (1.49 USD/t), and 7.2 CNY/t (1.03 USD/t), providing quantitative support for the development of seasonally differentiated, tiered carbon pricing policies. Carbon emissions exhibit three stages in response to carbon price changes: low sensitivity, rapid reduction, and saturation. The reduction saturation point occurs at 200 CNY/t (28.6 USD/t), beyond which further price increases yield negligible emission reductions, revealing the diminishing marginal effects of carbon pricing policies.

From a macroscopic perspective, this research provides actionable decision-support frameworks for multiple stakeholders in the low-carbon energy transition. For system operators, the quantified seasonal carbon price inflection points (80–160 CNY/t) enable the data-driven optimization of CHP dispatch, P2G scheduling, and CCS operation under dynamic carbon markets, while the demonstrated 100% renewable utilization validates the operational feasibility of integrating intermittent renewables with flexible conversion technologies. For industrial park managers, the dual-dimensional response framework provides practical pathways to reduce the peak electricity procurement by 12.7% and achieve 84.9% emission reductions, with carbon trading revenue (2068 CNY/day, 295.4 USD/day) offsetting CCS expenditures to transform decarbonization from a cost burden into an economic opportunity. For policymakers, the identified saturation point of 200 CNY/t (28.6 USD/t) indicates the optimal policy intervention ranges where further price escalations yield diminishing returns, while seasonal effectiveness variations (summer 3.2% vs. transitional 15.2%) highlight the necessity for differentiated quota allocation methodologies. The quantified synergy value of 21.6 percentage points between P2G–CCS and supply–demand response underscores the importance of integrated policy incentives. These findings bridge the gap between theoretical optimization and operational implementation, offering scalable solutions for industrial parks pursuing carbon neutrality under market-based frameworks.

Despite these achievements, opportunities for further investigation remain. Future research directions include the following: (1) extending the framework to multi-park collaborative optimization scenarios, where interconnected energy networks can enhance overall system efficiency through cross-park energy mutual support and carbon quota trading mechanisms; (2) incorporating stochastic or robust optimization methodologies to systematically address the uncertainties inherent in renewable energy generation and load forecasting, thereby improving dispatch reliability under high renewable penetration; and (3) developing comprehensive full-lifecycle risk assessment protocols for CCS technology, encompassing geological storage security, long-term leakage monitoring, and environmental impact evaluation. These research directions will further enhance the practical applicability and operational reliability of IES, providing more robust technical foundations for the large-scale low-carbon transformation of industrial parks.